diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,100856 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 14402,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 6.943480072212192e-05,
+      "grad_norm": 25.743069053425643,
+      "learning_rate": 2.3094688221709007e-08,
+      "loss": 3.1826,
+      "step": 1
+    },
+    {
+      "epoch": 0.00013886960144424384,
+      "grad_norm": 24.166175221666883,
+      "learning_rate": 4.6189376443418014e-08,
+      "loss": 3.1349,
+      "step": 2
+    },
+    {
+      "epoch": 0.00020830440216636578,
+      "grad_norm": 22.440605356482912,
+      "learning_rate": 6.928406466512703e-08,
+      "loss": 2.5422,
+      "step": 3
+    },
+    {
+      "epoch": 0.0002777392028884877,
+      "grad_norm": 30.98926697054666,
+      "learning_rate": 9.237875288683603e-08,
+      "loss": 3.2239,
+      "step": 4
+    },
+    {
+      "epoch": 0.00034717400361060965,
+      "grad_norm": 29.170489475620762,
+      "learning_rate": 1.1547344110854505e-07,
+      "loss": 3.3828,
+      "step": 5
+    },
+    {
+      "epoch": 0.00041660880433273156,
+      "grad_norm": 36.73531608789288,
+      "learning_rate": 1.3856812933025406e-07,
+      "loss": 3.4907,
+      "step": 6
+    },
+    {
+      "epoch": 0.0004860436050548535,
+      "grad_norm": 32.23771588791372,
+      "learning_rate": 1.6166281755196307e-07,
+      "loss": 3.6359,
+      "step": 7
+    },
+    {
+      "epoch": 0.0005554784057769754,
+      "grad_norm": 32.510747196956984,
+      "learning_rate": 1.8475750577367206e-07,
+      "loss": 3.5005,
+      "step": 8
+    },
+    {
+      "epoch": 0.0006249132064990974,
+      "grad_norm": 34.70408112594647,
+      "learning_rate": 2.0785219399538107e-07,
+      "loss": 3.8527,
+      "step": 9
+    },
+    {
+      "epoch": 0.0006943480072212193,
+      "grad_norm": 28.577339992701063,
+      "learning_rate": 2.309468822170901e-07,
+      "loss": 3.6636,
+      "step": 10
+    },
+    {
+      "epoch": 0.0007637828079433412,
+      "grad_norm": 26.801568183338322,
+      "learning_rate": 2.540415704387991e-07,
+      "loss": 2.9264,
+      "step": 11
+    },
+    {
+      "epoch": 0.0008332176086654631,
+      "grad_norm": 26.62357260357466,
+      "learning_rate": 2.771362586605081e-07,
+      "loss": 3.1676,
+      "step": 12
+    },
+    {
+      "epoch": 0.000902652409387585,
+      "grad_norm": 32.93000361263025,
+      "learning_rate": 3.0023094688221713e-07,
+      "loss": 3.3359,
+      "step": 13
+    },
+    {
+      "epoch": 0.000972087210109707,
+      "grad_norm": 35.826058883271884,
+      "learning_rate": 3.2332563510392614e-07,
+      "loss": 3.7214,
+      "step": 14
+    },
+    {
+      "epoch": 0.001041522010831829,
+      "grad_norm": 23.840167416609845,
+      "learning_rate": 3.464203233256351e-07,
+      "loss": 2.682,
+      "step": 15
+    },
+    {
+      "epoch": 0.0011109568115539507,
+      "grad_norm": 29.234444647064542,
+      "learning_rate": 3.695150115473441e-07,
+      "loss": 3.3909,
+      "step": 16
+    },
+    {
+      "epoch": 0.0011803916122760728,
+      "grad_norm": 28.281741162159857,
+      "learning_rate": 3.926096997690532e-07,
+      "loss": 3.0161,
+      "step": 17
+    },
+    {
+      "epoch": 0.0012498264129981948,
+      "grad_norm": 25.458988024367404,
+      "learning_rate": 4.1570438799076213e-07,
+      "loss": 2.5692,
+      "step": 18
+    },
+    {
+      "epoch": 0.0013192612137203166,
+      "grad_norm": 27.49871744760481,
+      "learning_rate": 4.3879907621247114e-07,
+      "loss": 3.1641,
+      "step": 19
+    },
+    {
+      "epoch": 0.0013886960144424386,
+      "grad_norm": 26.594539253723877,
+      "learning_rate": 4.618937644341802e-07,
+      "loss": 2.8514,
+      "step": 20
+    },
+    {
+      "epoch": 0.0014581308151645604,
+      "grad_norm": 24.857691441667104,
+      "learning_rate": 4.849884526558892e-07,
+      "loss": 2.754,
+      "step": 21
+    },
+    {
+      "epoch": 0.0015275656158866824,
+      "grad_norm": 37.539728438369586,
+      "learning_rate": 5.080831408775982e-07,
+      "loss": 4.1185,
+      "step": 22
+    },
+    {
+      "epoch": 0.0015970004166088044,
+      "grad_norm": 28.29778430699994,
+      "learning_rate": 5.311778290993072e-07,
+      "loss": 3.48,
+      "step": 23
+    },
+    {
+      "epoch": 0.0016664352173309262,
+      "grad_norm": 24.33092393333535,
+      "learning_rate": 5.542725173210162e-07,
+      "loss": 2.6638,
+      "step": 24
+    },
+    {
+      "epoch": 0.0017358700180530482,
+      "grad_norm": 25.256558995989188,
+      "learning_rate": 5.773672055427252e-07,
+      "loss": 2.582,
+      "step": 25
+    },
+    {
+      "epoch": 0.00180530481877517,
+      "grad_norm": 25.760249406840487,
+      "learning_rate": 6.004618937644343e-07,
+      "loss": 2.0152,
+      "step": 26
+    },
+    {
+      "epoch": 0.001874739619497292,
+      "grad_norm": 23.489047916347662,
+      "learning_rate": 6.235565819861433e-07,
+      "loss": 2.1558,
+      "step": 27
+    },
+    {
+      "epoch": 0.001944174420219414,
+      "grad_norm": 31.235978398880896,
+      "learning_rate": 6.466512702078523e-07,
+      "loss": 3.5227,
+      "step": 28
+    },
+    {
+      "epoch": 0.002013609220941536,
+      "grad_norm": 23.888543064125773,
+      "learning_rate": 6.697459584295613e-07,
+      "loss": 2.648,
+      "step": 29
+    },
+    {
+      "epoch": 0.002083044021663658,
+      "grad_norm": 36.33417629402573,
+      "learning_rate": 6.928406466512702e-07,
+      "loss": 4.6123,
+      "step": 30
+    },
+    {
+      "epoch": 0.00215247882238578,
+      "grad_norm": 28.010977611629272,
+      "learning_rate": 7.159353348729793e-07,
+      "loss": 2.9833,
+      "step": 31
+    },
+    {
+      "epoch": 0.0022219136231079015,
+      "grad_norm": 24.297214528180493,
+      "learning_rate": 7.390300230946882e-07,
+      "loss": 2.6014,
+      "step": 32
+    },
+    {
+      "epoch": 0.0022913484238300235,
+      "grad_norm": 25.52625918335633,
+      "learning_rate": 7.621247113163972e-07,
+      "loss": 2.4657,
+      "step": 33
+    },
+    {
+      "epoch": 0.0023607832245521455,
+      "grad_norm": 26.47100750061038,
+      "learning_rate": 7.852193995381063e-07,
+      "loss": 2.7159,
+      "step": 34
+    },
+    {
+      "epoch": 0.0024302180252742675,
+      "grad_norm": 29.18671888965165,
+      "learning_rate": 8.083140877598153e-07,
+      "loss": 3.403,
+      "step": 35
+    },
+    {
+      "epoch": 0.0024996528259963896,
+      "grad_norm": 23.36608661775619,
+      "learning_rate": 8.314087759815243e-07,
+      "loss": 2.6611,
+      "step": 36
+    },
+    {
+      "epoch": 0.002569087626718511,
+      "grad_norm": 25.488314032085164,
+      "learning_rate": 8.545034642032334e-07,
+      "loss": 2.7589,
+      "step": 37
+    },
+    {
+      "epoch": 0.002638522427440633,
+      "grad_norm": 22.56654378362558,
+      "learning_rate": 8.775981524249423e-07,
+      "loss": 3.0017,
+      "step": 38
+    },
+    {
+      "epoch": 0.002707957228162755,
+      "grad_norm": 27.823323684714733,
+      "learning_rate": 9.006928406466514e-07,
+      "loss": 3.1453,
+      "step": 39
+    },
+    {
+      "epoch": 0.002777392028884877,
+      "grad_norm": 22.146653263912004,
+      "learning_rate": 9.237875288683604e-07,
+      "loss": 2.6011,
+      "step": 40
+    },
+    {
+      "epoch": 0.002846826829606999,
+      "grad_norm": 24.048448823267766,
+      "learning_rate": 9.468822170900693e-07,
+      "loss": 2.637,
+      "step": 41
+    },
+    {
+      "epoch": 0.002916261630329121,
+      "grad_norm": 24.07555341942036,
+      "learning_rate": 9.699769053117784e-07,
+      "loss": 3.051,
+      "step": 42
+    },
+    {
+      "epoch": 0.002985696431051243,
+      "grad_norm": 19.982728293116626,
+      "learning_rate": 9.930715935334874e-07,
+      "loss": 2.0515,
+      "step": 43
+    },
+    {
+      "epoch": 0.003055131231773365,
+      "grad_norm": 22.50681926212181,
+      "learning_rate": 1.0161662817551965e-06,
+      "loss": 2.7645,
+      "step": 44
+    },
+    {
+      "epoch": 0.003124566032495487,
+      "grad_norm": 17.11870138045247,
+      "learning_rate": 1.0392609699769055e-06,
+      "loss": 1.7083,
+      "step": 45
+    },
+    {
+      "epoch": 0.003194000833217609,
+      "grad_norm": 24.713318785274932,
+      "learning_rate": 1.0623556581986145e-06,
+      "loss": 2.381,
+      "step": 46
+    },
+    {
+      "epoch": 0.0032634356339397304,
+      "grad_norm": 17.501124546350336,
+      "learning_rate": 1.0854503464203233e-06,
+      "loss": 1.9371,
+      "step": 47
+    },
+    {
+      "epoch": 0.0033328704346618524,
+      "grad_norm": 16.206932101084885,
+      "learning_rate": 1.1085450346420325e-06,
+      "loss": 1.9531,
+      "step": 48
+    },
+    {
+      "epoch": 0.0034023052353839745,
+      "grad_norm": 14.423982498032824,
+      "learning_rate": 1.1316397228637415e-06,
+      "loss": 1.3662,
+      "step": 49
+    },
+    {
+      "epoch": 0.0034717400361060965,
+      "grad_norm": 15.341809929206708,
+      "learning_rate": 1.1547344110854503e-06,
+      "loss": 1.4146,
+      "step": 50
+    },
+    {
+      "epoch": 0.0035411748368282185,
+      "grad_norm": 19.41155990733793,
+      "learning_rate": 1.1778290993071595e-06,
+      "loss": 1.4623,
+      "step": 51
+    },
+    {
+      "epoch": 0.00361060963755034,
+      "grad_norm": 16.373297361743013,
+      "learning_rate": 1.2009237875288685e-06,
+      "loss": 1.186,
+      "step": 52
+    },
+    {
+      "epoch": 0.003680044438272462,
+      "grad_norm": 16.275983136764896,
+      "learning_rate": 1.2240184757505773e-06,
+      "loss": 1.5246,
+      "step": 53
+    },
+    {
+      "epoch": 0.003749479238994584,
+      "grad_norm": 20.327434114179265,
+      "learning_rate": 1.2471131639722866e-06,
+      "loss": 2.194,
+      "step": 54
+    },
+    {
+      "epoch": 0.003818914039716706,
+      "grad_norm": 17.578348797589847,
+      "learning_rate": 1.2702078521939956e-06,
+      "loss": 1.8619,
+      "step": 55
+    },
+    {
+      "epoch": 0.003888348840438828,
+      "grad_norm": 9.54970331285728,
+      "learning_rate": 1.2933025404157046e-06,
+      "loss": 0.9079,
+      "step": 56
+    },
+    {
+      "epoch": 0.00395778364116095,
+      "grad_norm": 12.279321817700035,
+      "learning_rate": 1.3163972286374136e-06,
+      "loss": 0.8782,
+      "step": 57
+    },
+    {
+      "epoch": 0.004027218441883072,
+      "grad_norm": 12.137562385951293,
+      "learning_rate": 1.3394919168591226e-06,
+      "loss": 1.5448,
+      "step": 58
+    },
+    {
+      "epoch": 0.004096653242605193,
+      "grad_norm": 12.568245754461351,
+      "learning_rate": 1.3625866050808314e-06,
+      "loss": 1.9133,
+      "step": 59
+    },
+    {
+      "epoch": 0.004166088043327316,
+      "grad_norm": 12.048296413369126,
+      "learning_rate": 1.3856812933025404e-06,
+      "loss": 1.5706,
+      "step": 60
+    },
+    {
+      "epoch": 0.004235522844049437,
+      "grad_norm": 12.750101727107443,
+      "learning_rate": 1.4087759815242496e-06,
+      "loss": 1.3996,
+      "step": 61
+    },
+    {
+      "epoch": 0.00430495764477156,
+      "grad_norm": 8.718744059951307,
+      "learning_rate": 1.4318706697459586e-06,
+      "loss": 1.0854,
+      "step": 62
+    },
+    {
+      "epoch": 0.004374392445493681,
+      "grad_norm": 11.089197498297837,
+      "learning_rate": 1.4549653579676676e-06,
+      "loss": 1.3503,
+      "step": 63
+    },
+    {
+      "epoch": 0.004443827246215803,
+      "grad_norm": 10.842418149670134,
+      "learning_rate": 1.4780600461893764e-06,
+      "loss": 1.1687,
+      "step": 64
+    },
+    {
+      "epoch": 0.004513262046937925,
+      "grad_norm": 8.753486708154485,
+      "learning_rate": 1.5011547344110855e-06,
+      "loss": 1.1534,
+      "step": 65
+    },
+    {
+      "epoch": 0.004582696847660047,
+      "grad_norm": 9.47527019885148,
+      "learning_rate": 1.5242494226327945e-06,
+      "loss": 0.8791,
+      "step": 66
+    },
+    {
+      "epoch": 0.0046521316483821695,
+      "grad_norm": 8.67619044254941,
+      "learning_rate": 1.5473441108545037e-06,
+      "loss": 1.1002,
+      "step": 67
+    },
+    {
+      "epoch": 0.004721566449104291,
+      "grad_norm": 8.999343453644995,
+      "learning_rate": 1.5704387990762127e-06,
+      "loss": 1.0985,
+      "step": 68
+    },
+    {
+      "epoch": 0.004791001249826413,
+      "grad_norm": 10.493490047629653,
+      "learning_rate": 1.5935334872979217e-06,
+      "loss": 1.1387,
+      "step": 69
+    },
+    {
+      "epoch": 0.004860436050548535,
+      "grad_norm": 10.356976905673072,
+      "learning_rate": 1.6166281755196305e-06,
+      "loss": 1.2244,
+      "step": 70
+    },
+    {
+      "epoch": 0.004929870851270657,
+      "grad_norm": 9.000559087830059,
+      "learning_rate": 1.6397228637413395e-06,
+      "loss": 1.5649,
+      "step": 71
+    },
+    {
+      "epoch": 0.004999305651992779,
+      "grad_norm": 7.338640699431165,
+      "learning_rate": 1.6628175519630485e-06,
+      "loss": 0.6841,
+      "step": 72
+    },
+    {
+      "epoch": 0.005068740452714901,
+      "grad_norm": 8.905079207582279,
+      "learning_rate": 1.6859122401847578e-06,
+      "loss": 0.9226,
+      "step": 73
+    },
+    {
+      "epoch": 0.005138175253437022,
+      "grad_norm": 11.250404321011482,
+      "learning_rate": 1.7090069284064668e-06,
+      "loss": 0.8143,
+      "step": 74
+    },
+    {
+      "epoch": 0.005207610054159145,
+      "grad_norm": 10.356677078236354,
+      "learning_rate": 1.7321016166281756e-06,
+      "loss": 1.0377,
+      "step": 75
+    },
+    {
+      "epoch": 0.005277044854881266,
+      "grad_norm": 10.138981484769202,
+      "learning_rate": 1.7551963048498846e-06,
+      "loss": 0.9871,
+      "step": 76
+    },
+    {
+      "epoch": 0.005346479655603389,
+      "grad_norm": 9.680459763079574,
+      "learning_rate": 1.7782909930715936e-06,
+      "loss": 1.0687,
+      "step": 77
+    },
+    {
+      "epoch": 0.00541591445632551,
+      "grad_norm": 9.502236507628096,
+      "learning_rate": 1.8013856812933028e-06,
+      "loss": 0.576,
+      "step": 78
+    },
+    {
+      "epoch": 0.005485349257047632,
+      "grad_norm": 8.543038322969451,
+      "learning_rate": 1.8244803695150118e-06,
+      "loss": 0.9264,
+      "step": 79
+    },
+    {
+      "epoch": 0.005554784057769754,
+      "grad_norm": 13.08679735538333,
+      "learning_rate": 1.8475750577367208e-06,
+      "loss": 0.9738,
+      "step": 80
+    },
+    {
+      "epoch": 0.005624218858491876,
+      "grad_norm": 6.243650078421165,
+      "learning_rate": 1.8706697459584296e-06,
+      "loss": 0.4676,
+      "step": 81
+    },
+    {
+      "epoch": 0.005693653659213998,
+      "grad_norm": 7.518004315662655,
+      "learning_rate": 1.8937644341801386e-06,
+      "loss": 0.5602,
+      "step": 82
+    },
+    {
+      "epoch": 0.00576308845993612,
+      "grad_norm": 10.026994351790133,
+      "learning_rate": 1.9168591224018476e-06,
+      "loss": 0.953,
+      "step": 83
+    },
+    {
+      "epoch": 0.005832523260658242,
+      "grad_norm": 7.471787475858601,
+      "learning_rate": 1.939953810623557e-06,
+      "loss": 0.6896,
+      "step": 84
+    },
+    {
+      "epoch": 0.005901958061380364,
+      "grad_norm": 8.112109679718,
+      "learning_rate": 1.9630484988452657e-06,
+      "loss": 1.2848,
+      "step": 85
+    },
+    {
+      "epoch": 0.005971392862102486,
+      "grad_norm": 8.901612937354118,
+      "learning_rate": 1.986143187066975e-06,
+      "loss": 0.8456,
+      "step": 86
+    },
+    {
+      "epoch": 0.006040827662824608,
+      "grad_norm": 6.632958150924269,
+      "learning_rate": 2.0092378752886837e-06,
+      "loss": 0.4002,
+      "step": 87
+    },
+    {
+      "epoch": 0.00611026246354673,
+      "grad_norm": 9.513902698937503,
+      "learning_rate": 2.032332563510393e-06,
+      "loss": 0.953,
+      "step": 88
+    },
+    {
+      "epoch": 0.006179697264268851,
+      "grad_norm": 8.977874587093943,
+      "learning_rate": 2.0554272517321017e-06,
+      "loss": 0.8093,
+      "step": 89
+    },
+    {
+      "epoch": 0.006249132064990974,
+      "grad_norm": 7.877280918683887,
+      "learning_rate": 2.078521939953811e-06,
+      "loss": 0.9446,
+      "step": 90
+    },
+    {
+      "epoch": 0.006318566865713095,
+      "grad_norm": 8.682888539294106,
+      "learning_rate": 2.1016166281755197e-06,
+      "loss": 0.825,
+      "step": 91
+    },
+    {
+      "epoch": 0.006388001666435218,
+      "grad_norm": 10.503719122240408,
+      "learning_rate": 2.124711316397229e-06,
+      "loss": 0.8066,
+      "step": 92
+    },
+    {
+      "epoch": 0.006457436467157339,
+      "grad_norm": 7.049284652644519,
+      "learning_rate": 2.1478060046189377e-06,
+      "loss": 0.7004,
+      "step": 93
+    },
+    {
+      "epoch": 0.006526871267879461,
+      "grad_norm": 8.004087065974892,
+      "learning_rate": 2.1709006928406465e-06,
+      "loss": 0.7037,
+      "step": 94
+    },
+    {
+      "epoch": 0.006596306068601583,
+      "grad_norm": 9.634970345519243,
+      "learning_rate": 2.1939953810623558e-06,
+      "loss": 0.8039,
+      "step": 95
+    },
+    {
+      "epoch": 0.006665740869323705,
+      "grad_norm": 7.392604521492039,
+      "learning_rate": 2.217090069284065e-06,
+      "loss": 0.6265,
+      "step": 96
+    },
+    {
+      "epoch": 0.006735175670045827,
+      "grad_norm": 8.341040957707106,
+      "learning_rate": 2.2401847575057738e-06,
+      "loss": 0.9059,
+      "step": 97
+    },
+    {
+      "epoch": 0.006804610470767949,
+      "grad_norm": 8.51503350737925,
+      "learning_rate": 2.263279445727483e-06,
+      "loss": 0.6902,
+      "step": 98
+    },
+    {
+      "epoch": 0.0068740452714900705,
+      "grad_norm": 8.552520045378948,
+      "learning_rate": 2.286374133949192e-06,
+      "loss": 0.9884,
+      "step": 99
+    },
+    {
+      "epoch": 0.006943480072212193,
+      "grad_norm": 6.329427847739734,
+      "learning_rate": 2.3094688221709006e-06,
+      "loss": 0.8089,
+      "step": 100
+    },
+    {
+      "epoch": 0.0070129148729343145,
+      "grad_norm": 8.998937337841024,
+      "learning_rate": 2.33256351039261e-06,
+      "loss": 0.7129,
+      "step": 101
+    },
+    {
+      "epoch": 0.007082349673656437,
+      "grad_norm": 7.786626046748243,
+      "learning_rate": 2.355658198614319e-06,
+      "loss": 0.7022,
+      "step": 102
+    },
+    {
+      "epoch": 0.007151784474378559,
+      "grad_norm": 6.622691089398595,
+      "learning_rate": 2.378752886836028e-06,
+      "loss": 0.5738,
+      "step": 103
+    },
+    {
+      "epoch": 0.00722121927510068,
+      "grad_norm": 8.000336760770878,
+      "learning_rate": 2.401847575057737e-06,
+      "loss": 0.8833,
+      "step": 104
+    },
+    {
+      "epoch": 0.007290654075822803,
+      "grad_norm": 8.085605236794848,
+      "learning_rate": 2.424942263279446e-06,
+      "loss": 0.6113,
+      "step": 105
+    },
+    {
+      "epoch": 0.007360088876544924,
+      "grad_norm": 7.052447042160323,
+      "learning_rate": 2.4480369515011547e-06,
+      "loss": 0.5477,
+      "step": 106
+    },
+    {
+      "epoch": 0.007429523677267047,
+      "grad_norm": 7.347218311552021,
+      "learning_rate": 2.471131639722864e-06,
+      "loss": 0.7281,
+      "step": 107
+    },
+    {
+      "epoch": 0.007498958477989168,
+      "grad_norm": 8.165754998974737,
+      "learning_rate": 2.494226327944573e-06,
+      "loss": 0.76,
+      "step": 108
+    },
+    {
+      "epoch": 0.00756839327871129,
+      "grad_norm": 9.20995143835554,
+      "learning_rate": 2.517321016166282e-06,
+      "loss": 0.8136,
+      "step": 109
+    },
+    {
+      "epoch": 0.007637828079433412,
+      "grad_norm": 8.04080671703488,
+      "learning_rate": 2.540415704387991e-06,
+      "loss": 0.5286,
+      "step": 110
+    },
+    {
+      "epoch": 0.007707262880155534,
+      "grad_norm": 7.9324158285274375,
+      "learning_rate": 2.5635103926097e-06,
+      "loss": 0.521,
+      "step": 111
+    },
+    {
+      "epoch": 0.007776697680877656,
+      "grad_norm": 9.96336583788214,
+      "learning_rate": 2.586605080831409e-06,
+      "loss": 1.0405,
+      "step": 112
+    },
+    {
+      "epoch": 0.007846132481599778,
+      "grad_norm": 6.353347273278931,
+      "learning_rate": 2.609699769053118e-06,
+      "loss": 0.445,
+      "step": 113
+    },
+    {
+      "epoch": 0.0079155672823219,
+      "grad_norm": 8.588038325108936,
+      "learning_rate": 2.632794457274827e-06,
+      "loss": 0.9429,
+      "step": 114
+    },
+    {
+      "epoch": 0.007985002083044021,
+      "grad_norm": 8.058255637313328,
+      "learning_rate": 2.6558891454965355e-06,
+      "loss": 0.5134,
+      "step": 115
+    },
+    {
+      "epoch": 0.008054436883766143,
+      "grad_norm": 7.483094127899339,
+      "learning_rate": 2.678983833718245e-06,
+      "loss": 0.5747,
+      "step": 116
+    },
+    {
+      "epoch": 0.008123871684488266,
+      "grad_norm": 9.349145562118684,
+      "learning_rate": 2.7020785219399544e-06,
+      "loss": 0.9157,
+      "step": 117
+    },
+    {
+      "epoch": 0.008193306485210387,
+      "grad_norm": 8.163396168943548,
+      "learning_rate": 2.725173210161663e-06,
+      "loss": 0.6872,
+      "step": 118
+    },
+    {
+      "epoch": 0.008262741285932509,
+      "grad_norm": 8.23598924390474,
+      "learning_rate": 2.748267898383372e-06,
+      "loss": 0.8795,
+      "step": 119
+    },
+    {
+      "epoch": 0.008332176086654632,
+      "grad_norm": 7.404893421679968,
+      "learning_rate": 2.771362586605081e-06,
+      "loss": 0.9829,
+      "step": 120
+    },
+    {
+      "epoch": 0.008401610887376754,
+      "grad_norm": 6.7105027768007925,
+      "learning_rate": 2.79445727482679e-06,
+      "loss": 0.5319,
+      "step": 121
+    },
+    {
+      "epoch": 0.008471045688098875,
+      "grad_norm": 8.298025773152233,
+      "learning_rate": 2.8175519630484993e-06,
+      "loss": 0.7877,
+      "step": 122
+    },
+    {
+      "epoch": 0.008540480488820997,
+      "grad_norm": 6.80474408690692,
+      "learning_rate": 2.840646651270208e-06,
+      "loss": 0.837,
+      "step": 123
+    },
+    {
+      "epoch": 0.00860991528954312,
+      "grad_norm": 7.907476042663422,
+      "learning_rate": 2.8637413394919173e-06,
+      "loss": 0.7648,
+      "step": 124
+    },
+    {
+      "epoch": 0.00867935009026524,
+      "grad_norm": 8.430192083729516,
+      "learning_rate": 2.886836027713626e-06,
+      "loss": 0.5931,
+      "step": 125
+    },
+    {
+      "epoch": 0.008748784890987363,
+      "grad_norm": 6.334908481261137,
+      "learning_rate": 2.9099307159353353e-06,
+      "loss": 0.423,
+      "step": 126
+    },
+    {
+      "epoch": 0.008818219691709485,
+      "grad_norm": 6.664750534802436,
+      "learning_rate": 2.9330254041570437e-06,
+      "loss": 0.7361,
+      "step": 127
+    },
+    {
+      "epoch": 0.008887654492431606,
+      "grad_norm": 6.316517517639449,
+      "learning_rate": 2.956120092378753e-06,
+      "loss": 0.6497,
+      "step": 128
+    },
+    {
+      "epoch": 0.008957089293153728,
+      "grad_norm": 9.000839744781612,
+      "learning_rate": 2.979214780600462e-06,
+      "loss": 0.6092,
+      "step": 129
+    },
+    {
+      "epoch": 0.00902652409387585,
+      "grad_norm": 9.454946006631703,
+      "learning_rate": 3.002309468822171e-06,
+      "loss": 0.9259,
+      "step": 130
+    },
+    {
+      "epoch": 0.009095958894597973,
+      "grad_norm": 6.471047428954251,
+      "learning_rate": 3.02540415704388e-06,
+      "loss": 0.3845,
+      "step": 131
+    },
+    {
+      "epoch": 0.009165393695320094,
+      "grad_norm": 9.146262598600908,
+      "learning_rate": 3.048498845265589e-06,
+      "loss": 0.9956,
+      "step": 132
+    },
+    {
+      "epoch": 0.009234828496042216,
+      "grad_norm": 4.973471332963633,
+      "learning_rate": 3.071593533487298e-06,
+      "loss": 0.3136,
+      "step": 133
+    },
+    {
+      "epoch": 0.009304263296764339,
+      "grad_norm": 5.037309599011863,
+      "learning_rate": 3.0946882217090074e-06,
+      "loss": 0.2538,
+      "step": 134
+    },
+    {
+      "epoch": 0.00937369809748646,
+      "grad_norm": 8.846441039525033,
+      "learning_rate": 3.117782909930716e-06,
+      "loss": 0.8734,
+      "step": 135
+    },
+    {
+      "epoch": 0.009443132898208582,
+      "grad_norm": 8.741796894358815,
+      "learning_rate": 3.1408775981524254e-06,
+      "loss": 0.7593,
+      "step": 136
+    },
+    {
+      "epoch": 0.009512567698930705,
+      "grad_norm": 8.388844859539322,
+      "learning_rate": 3.163972286374134e-06,
+      "loss": 0.5611,
+      "step": 137
+    },
+    {
+      "epoch": 0.009582002499652825,
+      "grad_norm": 8.440278770747923,
+      "learning_rate": 3.1870669745958434e-06,
+      "loss": 0.5961,
+      "step": 138
+    },
+    {
+      "epoch": 0.009651437300374948,
+      "grad_norm": 9.155870099159435,
+      "learning_rate": 3.2101616628175526e-06,
+      "loss": 0.6409,
+      "step": 139
+    },
+    {
+      "epoch": 0.00972087210109707,
+      "grad_norm": 7.165109260560067,
+      "learning_rate": 3.233256351039261e-06,
+      "loss": 0.5397,
+      "step": 140
+    },
+    {
+      "epoch": 0.009790306901819193,
+      "grad_norm": 7.666755660133118,
+      "learning_rate": 3.2563510392609702e-06,
+      "loss": 0.6154,
+      "step": 141
+    },
+    {
+      "epoch": 0.009859741702541313,
+      "grad_norm": 9.113626284488772,
+      "learning_rate": 3.279445727482679e-06,
+      "loss": 0.9999,
+      "step": 142
+    },
+    {
+      "epoch": 0.009929176503263436,
+      "grad_norm": 7.561342490361833,
+      "learning_rate": 3.3025404157043883e-06,
+      "loss": 0.5588,
+      "step": 143
+    },
+    {
+      "epoch": 0.009998611303985558,
+      "grad_norm": 7.466591873255865,
+      "learning_rate": 3.325635103926097e-06,
+      "loss": 0.5744,
+      "step": 144
+    },
+    {
+      "epoch": 0.010068046104707679,
+      "grad_norm": 11.61710081432426,
+      "learning_rate": 3.3487297921478063e-06,
+      "loss": 0.7473,
+      "step": 145
+    },
+    {
+      "epoch": 0.010137480905429801,
+      "grad_norm": 7.8725926074307,
+      "learning_rate": 3.3718244803695155e-06,
+      "loss": 0.5119,
+      "step": 146
+    },
+    {
+      "epoch": 0.010206915706151924,
+      "grad_norm": 7.972460125637176,
+      "learning_rate": 3.3949191685912243e-06,
+      "loss": 0.4512,
+      "step": 147
+    },
+    {
+      "epoch": 0.010276350506874045,
+      "grad_norm": 8.724357984745915,
+      "learning_rate": 3.4180138568129335e-06,
+      "loss": 0.6453,
+      "step": 148
+    },
+    {
+      "epoch": 0.010345785307596167,
+      "grad_norm": 7.961249184519657,
+      "learning_rate": 3.441108545034642e-06,
+      "loss": 0.5433,
+      "step": 149
+    },
+    {
+      "epoch": 0.01041522010831829,
+      "grad_norm": 6.215403593926059,
+      "learning_rate": 3.464203233256351e-06,
+      "loss": 0.7723,
+      "step": 150
+    },
+    {
+      "epoch": 0.010484654909040412,
+      "grad_norm": 8.758460490427854,
+      "learning_rate": 3.4872979214780608e-06,
+      "loss": 0.9119,
+      "step": 151
+    },
+    {
+      "epoch": 0.010554089709762533,
+      "grad_norm": 10.176733999827276,
+      "learning_rate": 3.510392609699769e-06,
+      "loss": 0.8499,
+      "step": 152
+    },
+    {
+      "epoch": 0.010623524510484655,
+      "grad_norm": 6.876447113139471,
+      "learning_rate": 3.5334872979214784e-06,
+      "loss": 0.4014,
+      "step": 153
+    },
+    {
+      "epoch": 0.010692959311206778,
+      "grad_norm": 7.746090493668413,
+      "learning_rate": 3.556581986143187e-06,
+      "loss": 0.7376,
+      "step": 154
+    },
+    {
+      "epoch": 0.010762394111928898,
+      "grad_norm": 9.048510962138742,
+      "learning_rate": 3.5796766743648964e-06,
+      "loss": 0.6867,
+      "step": 155
+    },
+    {
+      "epoch": 0.01083182891265102,
+      "grad_norm": 7.736614251182222,
+      "learning_rate": 3.6027713625866056e-06,
+      "loss": 0.6344,
+      "step": 156
+    },
+    {
+      "epoch": 0.010901263713373143,
+      "grad_norm": 7.090764783531742,
+      "learning_rate": 3.6258660508083144e-06,
+      "loss": 0.7005,
+      "step": 157
+    },
+    {
+      "epoch": 0.010970698514095264,
+      "grad_norm": 7.706643936289208,
+      "learning_rate": 3.6489607390300236e-06,
+      "loss": 0.4559,
+      "step": 158
+    },
+    {
+      "epoch": 0.011040133314817386,
+      "grad_norm": 7.776544216058004,
+      "learning_rate": 3.6720554272517324e-06,
+      "loss": 0.5341,
+      "step": 159
+    },
+    {
+      "epoch": 0.011109568115539509,
+      "grad_norm": 8.019923216313877,
+      "learning_rate": 3.6951501154734416e-06,
+      "loss": 0.4778,
+      "step": 160
+    },
+    {
+      "epoch": 0.011179002916261631,
+      "grad_norm": 9.161321307245682,
+      "learning_rate": 3.71824480369515e-06,
+      "loss": 1.0299,
+      "step": 161
+    },
+    {
+      "epoch": 0.011248437716983752,
+      "grad_norm": 8.433578559862811,
+      "learning_rate": 3.7413394919168592e-06,
+      "loss": 0.6039,
+      "step": 162
+    },
+    {
+      "epoch": 0.011317872517705874,
+      "grad_norm": 7.695087311710237,
+      "learning_rate": 3.7644341801385685e-06,
+      "loss": 0.7305,
+      "step": 163
+    },
+    {
+      "epoch": 0.011387307318427997,
+      "grad_norm": 7.655838073611998,
+      "learning_rate": 3.7875288683602773e-06,
+      "loss": 0.6541,
+      "step": 164
+    },
+    {
+      "epoch": 0.011456742119150118,
+      "grad_norm": 10.467888289091613,
+      "learning_rate": 3.8106235565819865e-06,
+      "loss": 0.5579,
+      "step": 165
+    },
+    {
+      "epoch": 0.01152617691987224,
+      "grad_norm": 8.032530258362842,
+      "learning_rate": 3.833718244803695e-06,
+      "loss": 0.5161,
+      "step": 166
+    },
+    {
+      "epoch": 0.011595611720594362,
+      "grad_norm": 6.628730962263039,
+      "learning_rate": 3.8568129330254045e-06,
+      "loss": 0.5773,
+      "step": 167
+    },
+    {
+      "epoch": 0.011665046521316483,
+      "grad_norm": 8.296067825933568,
+      "learning_rate": 3.879907621247114e-06,
+      "loss": 0.6489,
+      "step": 168
+    },
+    {
+      "epoch": 0.011734481322038606,
+      "grad_norm": 7.901972751907289,
+      "learning_rate": 3.903002309468822e-06,
+      "loss": 0.597,
+      "step": 169
+    },
+    {
+      "epoch": 0.011803916122760728,
+      "grad_norm": 6.6234571772396595,
+      "learning_rate": 3.926096997690531e-06,
+      "loss": 0.4176,
+      "step": 170
+    },
+    {
+      "epoch": 0.011873350923482849,
+      "grad_norm": 4.951919107747259,
+      "learning_rate": 3.9491916859122405e-06,
+      "loss": 0.4214,
+      "step": 171
+    },
+    {
+      "epoch": 0.011942785724204971,
+      "grad_norm": 8.268020533640861,
+      "learning_rate": 3.97228637413395e-06,
+      "loss": 0.6889,
+      "step": 172
+    },
+    {
+      "epoch": 0.012012220524927094,
+      "grad_norm": 9.395095407345556,
+      "learning_rate": 3.995381062355658e-06,
+      "loss": 0.6425,
+      "step": 173
+    },
+    {
+      "epoch": 0.012081655325649216,
+      "grad_norm": 7.691579921237829,
+      "learning_rate": 4.018475750577367e-06,
+      "loss": 0.6374,
+      "step": 174
+    },
+    {
+      "epoch": 0.012151090126371337,
+      "grad_norm": 6.493290820973727,
+      "learning_rate": 4.041570438799077e-06,
+      "loss": 0.5608,
+      "step": 175
+    },
+    {
+      "epoch": 0.01222052492709346,
+      "grad_norm": 7.197257103887168,
+      "learning_rate": 4.064665127020786e-06,
+      "loss": 0.8264,
+      "step": 176
+    },
+    {
+      "epoch": 0.012289959727815582,
+      "grad_norm": 10.38255815605028,
+      "learning_rate": 4.087759815242495e-06,
+      "loss": 0.9234,
+      "step": 177
+    },
+    {
+      "epoch": 0.012359394528537702,
+      "grad_norm": 7.244668465776146,
+      "learning_rate": 4.110854503464203e-06,
+      "loss": 0.851,
+      "step": 178
+    },
+    {
+      "epoch": 0.012428829329259825,
+      "grad_norm": 9.357321140568963,
+      "learning_rate": 4.133949191685913e-06,
+      "loss": 0.6943,
+      "step": 179
+    },
+    {
+      "epoch": 0.012498264129981947,
+      "grad_norm": 7.631312363698161,
+      "learning_rate": 4.157043879907622e-06,
+      "loss": 0.7102,
+      "step": 180
+    },
+    {
+      "epoch": 0.012567698930704068,
+      "grad_norm": 7.586090088158953,
+      "learning_rate": 4.18013856812933e-06,
+      "loss": 0.7218,
+      "step": 181
+    },
+    {
+      "epoch": 0.01263713373142619,
+      "grad_norm": 7.414171322152922,
+      "learning_rate": 4.2032332563510394e-06,
+      "loss": 0.6047,
+      "step": 182
+    },
+    {
+      "epoch": 0.012706568532148313,
+      "grad_norm": 7.896374712841234,
+      "learning_rate": 4.226327944572749e-06,
+      "loss": 0.7097,
+      "step": 183
+    },
+    {
+      "epoch": 0.012776003332870435,
+      "grad_norm": 9.595437263954201,
+      "learning_rate": 4.249422632794458e-06,
+      "loss": 0.6796,
+      "step": 184
+    },
+    {
+      "epoch": 0.012845438133592556,
+      "grad_norm": 8.399978659382272,
+      "learning_rate": 4.272517321016167e-06,
+      "loss": 0.6807,
+      "step": 185
+    },
+    {
+      "epoch": 0.012914872934314679,
+      "grad_norm": 7.078793603147643,
+      "learning_rate": 4.2956120092378755e-06,
+      "loss": 0.4348,
+      "step": 186
+    },
+    {
+      "epoch": 0.012984307735036801,
+      "grad_norm": 9.295339273821106,
+      "learning_rate": 4.318706697459585e-06,
+      "loss": 0.7936,
+      "step": 187
+    },
+    {
+      "epoch": 0.013053742535758922,
+      "grad_norm": 8.695839431216239,
+      "learning_rate": 4.341801385681293e-06,
+      "loss": 0.9782,
+      "step": 188
+    },
+    {
+      "epoch": 0.013123177336481044,
+      "grad_norm": 6.366062648654598,
+      "learning_rate": 4.364896073903003e-06,
+      "loss": 0.559,
+      "step": 189
+    },
+    {
+      "epoch": 0.013192612137203167,
+      "grad_norm": 7.521188061274255,
+      "learning_rate": 4.3879907621247115e-06,
+      "loss": 0.4666,
+      "step": 190
+    },
+    {
+      "epoch": 0.013262046937925287,
+      "grad_norm": 7.0531940973021445,
+      "learning_rate": 4.411085450346421e-06,
+      "loss": 0.6315,
+      "step": 191
+    },
+    {
+      "epoch": 0.01333148173864741,
+      "grad_norm": 9.519300206969021,
+      "learning_rate": 4.43418013856813e-06,
+      "loss": 1.2702,
+      "step": 192
+    },
+    {
+      "epoch": 0.013400916539369532,
+      "grad_norm": 7.036508094027822,
+      "learning_rate": 4.457274826789838e-06,
+      "loss": 0.647,
+      "step": 193
+    },
+    {
+      "epoch": 0.013470351340091655,
+      "grad_norm": 7.526637565291745,
+      "learning_rate": 4.4803695150115476e-06,
+      "loss": 0.5896,
+      "step": 194
+    },
+    {
+      "epoch": 0.013539786140813775,
+      "grad_norm": 8.06785613897329,
+      "learning_rate": 4.503464203233257e-06,
+      "loss": 0.672,
+      "step": 195
+    },
+    {
+      "epoch": 0.013609220941535898,
+      "grad_norm": 8.16332388619368,
+      "learning_rate": 4.526558891454966e-06,
+      "loss": 0.4016,
+      "step": 196
+    },
+    {
+      "epoch": 0.01367865574225802,
+      "grad_norm": 7.5665928541979905,
+      "learning_rate": 4.549653579676675e-06,
+      "loss": 0.6695,
+      "step": 197
+    },
+    {
+      "epoch": 0.013748090542980141,
+      "grad_norm": 7.885979985250841,
+      "learning_rate": 4.572748267898384e-06,
+      "loss": 0.6387,
+      "step": 198
+    },
+    {
+      "epoch": 0.013817525343702263,
+      "grad_norm": 7.996267726925191,
+      "learning_rate": 4.595842956120093e-06,
+      "loss": 0.6419,
+      "step": 199
+    },
+    {
+      "epoch": 0.013886960144424386,
+      "grad_norm": 7.886509937063648,
+      "learning_rate": 4.618937644341801e-06,
+      "loss": 0.7893,
+      "step": 200
+    },
+    {
+      "epoch": 0.013956394945146507,
+      "grad_norm": 10.137581562013832,
+      "learning_rate": 4.6420323325635104e-06,
+      "loss": 0.8724,
+      "step": 201
+    },
+    {
+      "epoch": 0.014025829745868629,
+      "grad_norm": 8.866437627812644,
+      "learning_rate": 4.66512702078522e-06,
+      "loss": 0.9136,
+      "step": 202
+    },
+    {
+      "epoch": 0.014095264546590752,
+      "grad_norm": 7.240950884521287,
+      "learning_rate": 4.688221709006929e-06,
+      "loss": 0.5637,
+      "step": 203
+    },
+    {
+      "epoch": 0.014164699347312874,
+      "grad_norm": 6.705868275639191,
+      "learning_rate": 4.711316397228638e-06,
+      "loss": 0.3435,
+      "step": 204
+    },
+    {
+      "epoch": 0.014234134148034995,
+      "grad_norm": 7.7813978578362,
+      "learning_rate": 4.7344110854503465e-06,
+      "loss": 0.87,
+      "step": 205
+    },
+    {
+      "epoch": 0.014303568948757117,
+      "grad_norm": 6.34798697499396,
+      "learning_rate": 4.757505773672056e-06,
+      "loss": 0.5861,
+      "step": 206
+    },
+    {
+      "epoch": 0.01437300374947924,
+      "grad_norm": 5.762108611077589,
+      "learning_rate": 4.780600461893765e-06,
+      "loss": 0.3837,
+      "step": 207
+    },
+    {
+      "epoch": 0.01444243855020136,
+      "grad_norm": 6.776809589591002,
+      "learning_rate": 4.803695150115474e-06,
+      "loss": 0.5251,
+      "step": 208
+    },
+    {
+      "epoch": 0.014511873350923483,
+      "grad_norm": 7.252967084559288,
+      "learning_rate": 4.826789838337183e-06,
+      "loss": 0.4171,
+      "step": 209
+    },
+    {
+      "epoch": 0.014581308151645605,
+      "grad_norm": 6.267014708319897,
+      "learning_rate": 4.849884526558892e-06,
+      "loss": 0.4742,
+      "step": 210
+    },
+    {
+      "epoch": 0.014650742952367726,
+      "grad_norm": 7.906051619813322,
+      "learning_rate": 4.872979214780601e-06,
+      "loss": 0.7827,
+      "step": 211
+    },
+    {
+      "epoch": 0.014720177753089848,
+      "grad_norm": 8.613978998958547,
+      "learning_rate": 4.896073903002309e-06,
+      "loss": 0.6629,
+      "step": 212
+    },
+    {
+      "epoch": 0.01478961255381197,
+      "grad_norm": 6.203115949991411,
+      "learning_rate": 4.9191685912240186e-06,
+      "loss": 0.4997,
+      "step": 213
+    },
+    {
+      "epoch": 0.014859047354534093,
+      "grad_norm": 5.755868901799211,
+      "learning_rate": 4.942263279445728e-06,
+      "loss": 0.3926,
+      "step": 214
+    },
+    {
+      "epoch": 0.014928482155256214,
+      "grad_norm": 7.203269526065333,
+      "learning_rate": 4.965357967667437e-06,
+      "loss": 0.5858,
+      "step": 215
+    },
+    {
+      "epoch": 0.014997916955978336,
+      "grad_norm": 8.07108253770573,
+      "learning_rate": 4.988452655889146e-06,
+      "loss": 0.7231,
+      "step": 216
+    },
+    {
+      "epoch": 0.015067351756700459,
+      "grad_norm": 6.9279162382934665,
+      "learning_rate": 5.0115473441108554e-06,
+      "loss": 0.6657,
+      "step": 217
+    },
+    {
+      "epoch": 0.01513678655742258,
+      "grad_norm": 7.73020488394158,
+      "learning_rate": 5.034642032332564e-06,
+      "loss": 0.8388,
+      "step": 218
+    },
+    {
+      "epoch": 0.015206221358144702,
+      "grad_norm": 7.902878886808829,
+      "learning_rate": 5.057736720554273e-06,
+      "loss": 0.4083,
+      "step": 219
+    },
+    {
+      "epoch": 0.015275656158866825,
+      "grad_norm": 10.711205506478693,
+      "learning_rate": 5.080831408775982e-06,
+      "loss": 0.4782,
+      "step": 220
+    },
+    {
+      "epoch": 0.015345090959588945,
+      "grad_norm": 6.880081162040072,
+      "learning_rate": 5.1039260969976915e-06,
+      "loss": 0.2855,
+      "step": 221
+    },
+    {
+      "epoch": 0.015414525760311068,
+      "grad_norm": 6.380662974163436,
+      "learning_rate": 5.1270207852194e-06,
+      "loss": 0.4787,
+      "step": 222
+    },
+    {
+      "epoch": 0.01548396056103319,
+      "grad_norm": 7.439845565309826,
+      "learning_rate": 5.150115473441108e-06,
+      "loss": 0.5416,
+      "step": 223
+    },
+    {
+      "epoch": 0.015553395361755313,
+      "grad_norm": 8.091993541925458,
+      "learning_rate": 5.173210161662818e-06,
+      "loss": 0.3417,
+      "step": 224
+    },
+    {
+      "epoch": 0.015622830162477433,
+      "grad_norm": 7.49630391835408,
+      "learning_rate": 5.196304849884527e-06,
+      "loss": 0.7003,
+      "step": 225
+    },
+    {
+      "epoch": 0.015692264963199556,
+      "grad_norm": 7.939654978211845,
+      "learning_rate": 5.219399538106236e-06,
+      "loss": 0.7269,
+      "step": 226
+    },
+    {
+      "epoch": 0.015761699763921676,
+      "grad_norm": 7.345411689485614,
+      "learning_rate": 5.242494226327945e-06,
+      "loss": 0.5719,
+      "step": 227
+    },
+    {
+      "epoch": 0.0158311345646438,
+      "grad_norm": 10.482862789614648,
+      "learning_rate": 5.265588914549654e-06,
+      "loss": 0.9187,
+      "step": 228
+    },
+    {
+      "epoch": 0.01590056936536592,
+      "grad_norm": 8.171438197822155,
+      "learning_rate": 5.288683602771363e-06,
+      "loss": 0.4687,
+      "step": 229
+    },
+    {
+      "epoch": 0.015970004166088042,
+      "grad_norm": 7.223450030240171,
+      "learning_rate": 5.311778290993071e-06,
+      "loss": 0.6092,
+      "step": 230
+    },
+    {
+      "epoch": 0.016039438966810166,
+      "grad_norm": 7.068081165092103,
+      "learning_rate": 5.334872979214781e-06,
+      "loss": 0.5466,
+      "step": 231
+    },
+    {
+      "epoch": 0.016108873767532287,
+      "grad_norm": 5.874106446122856,
+      "learning_rate": 5.35796766743649e-06,
+      "loss": 0.4628,
+      "step": 232
+    },
+    {
+      "epoch": 0.016178308568254408,
+      "grad_norm": 7.108867549566205,
+      "learning_rate": 5.381062355658199e-06,
+      "loss": 0.3716,
+      "step": 233
+    },
+    {
+      "epoch": 0.016247743368976532,
+      "grad_norm": 5.632178509844155,
+      "learning_rate": 5.404157043879909e-06,
+      "loss": 0.5437,
+      "step": 234
+    },
+    {
+      "epoch": 0.016317178169698653,
+      "grad_norm": 7.789452575297797,
+      "learning_rate": 5.427251732101617e-06,
+      "loss": 0.9047,
+      "step": 235
+    },
+    {
+      "epoch": 0.016386612970420773,
+      "grad_norm": 6.313918989909171,
+      "learning_rate": 5.450346420323326e-06,
+      "loss": 0.8541,
+      "step": 236
+    },
+    {
+      "epoch": 0.016456047771142897,
+      "grad_norm": 8.792205404881122,
+      "learning_rate": 5.473441108545036e-06,
+      "loss": 0.5568,
+      "step": 237
+    },
+    {
+      "epoch": 0.016525482571865018,
+      "grad_norm": 7.093517449702753,
+      "learning_rate": 5.496535796766744e-06,
+      "loss": 0.2775,
+      "step": 238
+    },
+    {
+      "epoch": 0.016594917372587142,
+      "grad_norm": 7.030211419043433,
+      "learning_rate": 5.519630484988453e-06,
+      "loss": 0.3933,
+      "step": 239
+    },
+    {
+      "epoch": 0.016664352173309263,
+      "grad_norm": 7.109727530662788,
+      "learning_rate": 5.542725173210162e-06,
+      "loss": 0.6273,
+      "step": 240
+    },
+    {
+      "epoch": 0.016733786974031384,
+      "grad_norm": 9.235528657829454,
+      "learning_rate": 5.565819861431872e-06,
+      "loss": 0.9034,
+      "step": 241
+    },
+    {
+      "epoch": 0.016803221774753508,
+      "grad_norm": 5.60568755918627,
+      "learning_rate": 5.58891454965358e-06,
+      "loss": 0.2712,
+      "step": 242
+    },
+    {
+      "epoch": 0.01687265657547563,
+      "grad_norm": 7.894937100207131,
+      "learning_rate": 5.6120092378752884e-06,
+      "loss": 0.4055,
+      "step": 243
+    },
+    {
+      "epoch": 0.01694209137619775,
+      "grad_norm": 7.242054568582779,
+      "learning_rate": 5.6351039260969985e-06,
+      "loss": 0.673,
+      "step": 244
+    },
+    {
+      "epoch": 0.017011526176919874,
+      "grad_norm": 7.741044483359136,
+      "learning_rate": 5.658198614318708e-06,
+      "loss": 0.5858,
+      "step": 245
+    },
+    {
+      "epoch": 0.017080960977641994,
+      "grad_norm": 6.561504313361452,
+      "learning_rate": 5.681293302540416e-06,
+      "loss": 0.5438,
+      "step": 246
+    },
+    {
+      "epoch": 0.017150395778364115,
+      "grad_norm": 7.82850665938877,
+      "learning_rate": 5.7043879907621245e-06,
+      "loss": 0.6689,
+      "step": 247
+    },
+    {
+      "epoch": 0.01721983057908624,
+      "grad_norm": 6.818172850670444,
+      "learning_rate": 5.7274826789838346e-06,
+      "loss": 0.9547,
+      "step": 248
+    },
+    {
+      "epoch": 0.01728926537980836,
+      "grad_norm": 6.798584154668491,
+      "learning_rate": 5.750577367205543e-06,
+      "loss": 0.4316,
+      "step": 249
+    },
+    {
+      "epoch": 0.01735870018053048,
+      "grad_norm": 8.379369709812018,
+      "learning_rate": 5.773672055427252e-06,
+      "loss": 0.8569,
+      "step": 250
+    },
+    {
+      "epoch": 0.017428134981252605,
+      "grad_norm": 5.911403153707157,
+      "learning_rate": 5.796766743648961e-06,
+      "loss": 0.2661,
+      "step": 251
+    },
+    {
+      "epoch": 0.017497569781974726,
+      "grad_norm": 7.432621330431931,
+      "learning_rate": 5.819861431870671e-06,
+      "loss": 0.5803,
+      "step": 252
+    },
+    {
+      "epoch": 0.017567004582696846,
+      "grad_norm": 6.981417976237189,
+      "learning_rate": 5.842956120092379e-06,
+      "loss": 0.6352,
+      "step": 253
+    },
+    {
+      "epoch": 0.01763643938341897,
+      "grad_norm": 7.268127620571004,
+      "learning_rate": 5.866050808314087e-06,
+      "loss": 0.5386,
+      "step": 254
+    },
+    {
+      "epoch": 0.01770587418414109,
+      "grad_norm": 6.779749442363237,
+      "learning_rate": 5.889145496535797e-06,
+      "loss": 0.419,
+      "step": 255
+    },
+    {
+      "epoch": 0.017775308984863212,
+      "grad_norm": 6.048719139098827,
+      "learning_rate": 5.912240184757506e-06,
+      "loss": 0.398,
+      "step": 256
+    },
+    {
+      "epoch": 0.017844743785585336,
+      "grad_norm": 8.916395897895692,
+      "learning_rate": 5.935334872979215e-06,
+      "loss": 0.733,
+      "step": 257
+    },
+    {
+      "epoch": 0.017914178586307457,
+      "grad_norm": 5.763008183819964,
+      "learning_rate": 5.958429561200924e-06,
+      "loss": 0.5008,
+      "step": 258
+    },
+    {
+      "epoch": 0.017983613387029578,
+      "grad_norm": 5.506019659210942,
+      "learning_rate": 5.9815242494226335e-06,
+      "loss": 0.3564,
+      "step": 259
+    },
+    {
+      "epoch": 0.0180530481877517,
+      "grad_norm": 6.2034451104650365,
+      "learning_rate": 6.004618937644342e-06,
+      "loss": 0.4833,
+      "step": 260
+    },
+    {
+      "epoch": 0.018122482988473822,
+      "grad_norm": 7.430924159177619,
+      "learning_rate": 6.027713625866052e-06,
+      "loss": 0.432,
+      "step": 261
+    },
+    {
+      "epoch": 0.018191917789195947,
+      "grad_norm": 6.850038257928346,
+      "learning_rate": 6.05080831408776e-06,
+      "loss": 0.494,
+      "step": 262
+    },
+    {
+      "epoch": 0.018261352589918067,
+      "grad_norm": 6.721300418423033,
+      "learning_rate": 6.0739030023094695e-06,
+      "loss": 0.6773,
+      "step": 263
+    },
+    {
+      "epoch": 0.018330787390640188,
+      "grad_norm": 6.841477097403693,
+      "learning_rate": 6.096997690531178e-06,
+      "loss": 0.4801,
+      "step": 264
+    },
+    {
+      "epoch": 0.018400222191362312,
+      "grad_norm": 8.708883828325524,
+      "learning_rate": 6.120092378752888e-06,
+      "loss": 0.7887,
+      "step": 265
+    },
+    {
+      "epoch": 0.018469656992084433,
+      "grad_norm": 6.575860342200785,
+      "learning_rate": 6.143187066974596e-06,
+      "loss": 0.478,
+      "step": 266
+    },
+    {
+      "epoch": 0.018539091792806554,
+      "grad_norm": 7.144096209557728,
+      "learning_rate": 6.166281755196305e-06,
+      "loss": 0.6448,
+      "step": 267
+    },
+    {
+      "epoch": 0.018608526593528678,
+      "grad_norm": 8.178977160679182,
+      "learning_rate": 6.189376443418015e-06,
+      "loss": 0.7615,
+      "step": 268
+    },
+    {
+      "epoch": 0.0186779613942508,
+      "grad_norm": 8.376766153462068,
+      "learning_rate": 6.212471131639723e-06,
+      "loss": 0.7756,
+      "step": 269
+    },
+    {
+      "epoch": 0.01874739619497292,
+      "grad_norm": 8.048363295250732,
+      "learning_rate": 6.235565819861432e-06,
+      "loss": 0.45,
+      "step": 270
+    },
+    {
+      "epoch": 0.018816830995695043,
+      "grad_norm": 7.907170737872879,
+      "learning_rate": 6.258660508083141e-06,
+      "loss": 0.8838,
+      "step": 271
+    },
+    {
+      "epoch": 0.018886265796417164,
+      "grad_norm": 5.689841682331542,
+      "learning_rate": 6.281755196304851e-06,
+      "loss": 0.3793,
+      "step": 272
+    },
+    {
+      "epoch": 0.018955700597139285,
+      "grad_norm": 7.235933537441556,
+      "learning_rate": 6.304849884526559e-06,
+      "loss": 0.5064,
+      "step": 273
+    },
+    {
+      "epoch": 0.01902513539786141,
+      "grad_norm": 8.573491732383479,
+      "learning_rate": 6.327944572748268e-06,
+      "loss": 0.7878,
+      "step": 274
+    },
+    {
+      "epoch": 0.01909457019858353,
+      "grad_norm": 7.384608978851236,
+      "learning_rate": 6.351039260969978e-06,
+      "loss": 0.6261,
+      "step": 275
+    },
+    {
+      "epoch": 0.01916400499930565,
+      "grad_norm": 6.8300163065760735,
+      "learning_rate": 6.374133949191687e-06,
+      "loss": 0.5075,
+      "step": 276
+    },
+    {
+      "epoch": 0.019233439800027775,
+      "grad_norm": 7.20469012733151,
+      "learning_rate": 6.397228637413395e-06,
+      "loss": 0.6793,
+      "step": 277
+    },
+    {
+      "epoch": 0.019302874600749895,
+      "grad_norm": 7.66091465195689,
+      "learning_rate": 6.420323325635105e-06,
+      "loss": 0.6588,
+      "step": 278
+    },
+    {
+      "epoch": 0.019372309401472016,
+      "grad_norm": 6.1393678300018335,
+      "learning_rate": 6.443418013856814e-06,
+      "loss": 0.3072,
+      "step": 279
+    },
+    {
+      "epoch": 0.01944174420219414,
+      "grad_norm": 6.955873617409415,
+      "learning_rate": 6.466512702078522e-06,
+      "loss": 0.7568,
+      "step": 280
+    },
+    {
+      "epoch": 0.01951117900291626,
+      "grad_norm": 7.3958652182056674,
+      "learning_rate": 6.489607390300231e-06,
+      "loss": 0.6083,
+      "step": 281
+    },
+    {
+      "epoch": 0.019580613803638385,
+      "grad_norm": 7.248430019318075,
+      "learning_rate": 6.5127020785219405e-06,
+      "loss": 0.6296,
+      "step": 282
+    },
+    {
+      "epoch": 0.019650048604360506,
+      "grad_norm": 7.261412925078724,
+      "learning_rate": 6.53579676674365e-06,
+      "loss": 0.6231,
+      "step": 283
+    },
+    {
+      "epoch": 0.019719483405082627,
+      "grad_norm": 7.115747329622963,
+      "learning_rate": 6.558891454965358e-06,
+      "loss": 0.7525,
+      "step": 284
+    },
+    {
+      "epoch": 0.01978891820580475,
+      "grad_norm": 4.884646393329919,
+      "learning_rate": 6.581986143187068e-06,
+      "loss": 0.4707,
+      "step": 285
+    },
+    {
+      "epoch": 0.01985835300652687,
+      "grad_norm": 7.031418572019309,
+      "learning_rate": 6.6050808314087765e-06,
+      "loss": 0.6885,
+      "step": 286
+    },
+    {
+      "epoch": 0.019927787807248992,
+      "grad_norm": 8.727511101239376,
+      "learning_rate": 6.628175519630486e-06,
+      "loss": 0.8772,
+      "step": 287
+    },
+    {
+      "epoch": 0.019997222607971116,
+      "grad_norm": 6.825765458286112,
+      "learning_rate": 6.651270207852194e-06,
+      "loss": 0.6025,
+      "step": 288
+    },
+    {
+      "epoch": 0.020066657408693237,
+      "grad_norm": 7.803096499060419,
+      "learning_rate": 6.674364896073904e-06,
+      "loss": 0.587,
+      "step": 289
+    },
+    {
+      "epoch": 0.020136092209415358,
+      "grad_norm": 8.051493066179713,
+      "learning_rate": 6.6974595842956126e-06,
+      "loss": 0.7166,
+      "step": 290
+    },
+    {
+      "epoch": 0.020205527010137482,
+      "grad_norm": 7.528003888149653,
+      "learning_rate": 6.720554272517321e-06,
+      "loss": 0.7072,
+      "step": 291
+    },
+    {
+      "epoch": 0.020274961810859603,
+      "grad_norm": 5.872670048523211,
+      "learning_rate": 6.743648960739031e-06,
+      "loss": 0.6164,
+      "step": 292
+    },
+    {
+      "epoch": 0.020344396611581723,
+      "grad_norm": 7.247530452394163,
+      "learning_rate": 6.766743648960739e-06,
+      "loss": 0.6691,
+      "step": 293
+    },
+    {
+      "epoch": 0.020413831412303848,
+      "grad_norm": 7.148143712801787,
+      "learning_rate": 6.789838337182449e-06,
+      "loss": 0.6739,
+      "step": 294
+    },
+    {
+      "epoch": 0.02048326621302597,
+      "grad_norm": 7.654533786734221,
+      "learning_rate": 6.812933025404158e-06,
+      "loss": 0.7014,
+      "step": 295
+    },
+    {
+      "epoch": 0.02055270101374809,
+      "grad_norm": 6.616408505516904,
+      "learning_rate": 6.836027713625867e-06,
+      "loss": 0.5719,
+      "step": 296
+    },
+    {
+      "epoch": 0.020622135814470213,
+      "grad_norm": 7.042103483214846,
+      "learning_rate": 6.859122401847575e-06,
+      "loss": 0.7897,
+      "step": 297
+    },
+    {
+      "epoch": 0.020691570615192334,
+      "grad_norm": 8.395484946348608,
+      "learning_rate": 6.882217090069284e-06,
+      "loss": 0.6979,
+      "step": 298
+    },
+    {
+      "epoch": 0.020761005415914455,
+      "grad_norm": 6.4000256002916,
+      "learning_rate": 6.905311778290994e-06,
+      "loss": 0.7118,
+      "step": 299
+    },
+    {
+      "epoch": 0.02083044021663658,
+      "grad_norm": 6.564081158509931,
+      "learning_rate": 6.928406466512702e-06,
+      "loss": 0.4898,
+      "step": 300
+    },
+    {
+      "epoch": 0.0208998750173587,
+      "grad_norm": 7.506242231861854,
+      "learning_rate": 6.9515011547344115e-06,
+      "loss": 0.7916,
+      "step": 301
+    },
+    {
+      "epoch": 0.020969309818080824,
+      "grad_norm": 6.565232756783366,
+      "learning_rate": 6.9745958429561215e-06,
+      "loss": 0.5997,
+      "step": 302
+    },
+    {
+      "epoch": 0.021038744618802944,
+      "grad_norm": 5.535471339222495,
+      "learning_rate": 6.99769053117783e-06,
+      "loss": 0.5694,
+      "step": 303
+    },
+    {
+      "epoch": 0.021108179419525065,
+      "grad_norm": 6.245243607682892,
+      "learning_rate": 7.020785219399538e-06,
+      "loss": 0.3781,
+      "step": 304
+    },
+    {
+      "epoch": 0.02117761422024719,
+      "grad_norm": 7.5399123022411585,
+      "learning_rate": 7.0438799076212475e-06,
+      "loss": 0.7124,
+      "step": 305
+    },
+    {
+      "epoch": 0.02124704902096931,
+      "grad_norm": 7.235601607230619,
+      "learning_rate": 7.066974595842957e-06,
+      "loss": 0.6458,
+      "step": 306
+    },
+    {
+      "epoch": 0.02131648382169143,
+      "grad_norm": 6.05366644187799,
+      "learning_rate": 7.090069284064666e-06,
+      "loss": 0.293,
+      "step": 307
+    },
+    {
+      "epoch": 0.021385918622413555,
+      "grad_norm": 6.8668146431598345,
+      "learning_rate": 7.113163972286374e-06,
+      "loss": 0.7551,
+      "step": 308
+    },
+    {
+      "epoch": 0.021455353423135676,
+      "grad_norm": 6.736934020837996,
+      "learning_rate": 7.136258660508084e-06,
+      "loss": 0.6981,
+      "step": 309
+    },
+    {
+      "epoch": 0.021524788223857796,
+      "grad_norm": 5.249739320162398,
+      "learning_rate": 7.159353348729793e-06,
+      "loss": 0.3149,
+      "step": 310
+    },
+    {
+      "epoch": 0.02159422302457992,
+      "grad_norm": 6.8273055750049645,
+      "learning_rate": 7.182448036951501e-06,
+      "loss": 0.2502,
+      "step": 311
+    },
+    {
+      "epoch": 0.02166365782530204,
+      "grad_norm": 8.320322755579785,
+      "learning_rate": 7.205542725173211e-06,
+      "loss": 0.8918,
+      "step": 312
+    },
+    {
+      "epoch": 0.021733092626024162,
+      "grad_norm": 8.442613851593565,
+      "learning_rate": 7.22863741339492e-06,
+      "loss": 0.7903,
+      "step": 313
+    },
+    {
+      "epoch": 0.021802527426746286,
+      "grad_norm": 8.45217220190808,
+      "learning_rate": 7.251732101616629e-06,
+      "loss": 0.8665,
+      "step": 314
+    },
+    {
+      "epoch": 0.021871962227468407,
+      "grad_norm": 7.5277451108055935,
+      "learning_rate": 7.274826789838337e-06,
+      "loss": 0.6209,
+      "step": 315
+    },
+    {
+      "epoch": 0.021941397028190528,
+      "grad_norm": 8.248503159595833,
+      "learning_rate": 7.297921478060047e-06,
+      "loss": 0.9061,
+      "step": 316
+    },
+    {
+      "epoch": 0.022010831828912652,
+      "grad_norm": 7.2776793516288105,
+      "learning_rate": 7.321016166281756e-06,
+      "loss": 0.6603,
+      "step": 317
+    },
+    {
+      "epoch": 0.022080266629634773,
+      "grad_norm": 5.965637859359356,
+      "learning_rate": 7.344110854503465e-06,
+      "loss": 0.4422,
+      "step": 318
+    },
+    {
+      "epoch": 0.022149701430356893,
+      "grad_norm": 7.5312840452806284,
+      "learning_rate": 7.367205542725174e-06,
+      "loss": 0.6134,
+      "step": 319
+    },
+    {
+      "epoch": 0.022219136231079017,
+      "grad_norm": 5.186297170080935,
+      "learning_rate": 7.390300230946883e-06,
+      "loss": 0.3595,
+      "step": 320
+    },
+    {
+      "epoch": 0.022288571031801138,
+      "grad_norm": 5.828307732663609,
+      "learning_rate": 7.413394919168592e-06,
+      "loss": 0.8684,
+      "step": 321
+    },
+    {
+      "epoch": 0.022358005832523262,
+      "grad_norm": 7.883420149475018,
+      "learning_rate": 7.4364896073903e-06,
+      "loss": 0.7388,
+      "step": 322
+    },
+    {
+      "epoch": 0.022427440633245383,
+      "grad_norm": 5.958168620052153,
+      "learning_rate": 7.45958429561201e-06,
+      "loss": 0.6228,
+      "step": 323
+    },
+    {
+      "epoch": 0.022496875433967504,
+      "grad_norm": 7.586493157542109,
+      "learning_rate": 7.4826789838337185e-06,
+      "loss": 0.6236,
+      "step": 324
+    },
+    {
+      "epoch": 0.022566310234689628,
+      "grad_norm": 5.070739679339793,
+      "learning_rate": 7.505773672055428e-06,
+      "loss": 0.2893,
+      "step": 325
+    },
+    {
+      "epoch": 0.02263574503541175,
+      "grad_norm": 7.297666983970221,
+      "learning_rate": 7.528868360277137e-06,
+      "loss": 0.6513,
+      "step": 326
+    },
+    {
+      "epoch": 0.02270517983613387,
+      "grad_norm": 6.458388113960575,
+      "learning_rate": 7.551963048498846e-06,
+      "loss": 0.5596,
+      "step": 327
+    },
+    {
+      "epoch": 0.022774614636855994,
+      "grad_norm": 5.436275519046475,
+      "learning_rate": 7.5750577367205545e-06,
+      "loss": 0.3816,
+      "step": 328
+    },
+    {
+      "epoch": 0.022844049437578114,
+      "grad_norm": 7.137994819236926,
+      "learning_rate": 7.598152424942264e-06,
+      "loss": 0.7518,
+      "step": 329
+    },
+    {
+      "epoch": 0.022913484238300235,
+      "grad_norm": 7.055767976438735,
+      "learning_rate": 7.621247113163973e-06,
+      "loss": 0.5626,
+      "step": 330
+    },
+    {
+      "epoch": 0.02298291903902236,
+      "grad_norm": 6.997292636102885,
+      "learning_rate": 7.644341801385682e-06,
+      "loss": 0.824,
+      "step": 331
+    },
+    {
+      "epoch": 0.02305235383974448,
+      "grad_norm": 6.212786801754549,
+      "learning_rate": 7.66743648960739e-06,
+      "loss": 0.8321,
+      "step": 332
+    },
+    {
+      "epoch": 0.0231217886404666,
+      "grad_norm": 6.9599198959217174,
+      "learning_rate": 7.6905311778291e-06,
+      "loss": 0.7084,
+      "step": 333
+    },
+    {
+      "epoch": 0.023191223441188725,
+      "grad_norm": 7.06848038393813,
+      "learning_rate": 7.713625866050809e-06,
+      "loss": 0.53,
+      "step": 334
+    },
+    {
+      "epoch": 0.023260658241910846,
+      "grad_norm": 6.266825349856001,
+      "learning_rate": 7.736720554272517e-06,
+      "loss": 0.5645,
+      "step": 335
+    },
+    {
+      "epoch": 0.023330093042632966,
+      "grad_norm": 6.155799681888197,
+      "learning_rate": 7.759815242494227e-06,
+      "loss": 0.3237,
+      "step": 336
+    },
+    {
+      "epoch": 0.02339952784335509,
+      "grad_norm": 5.917148975534441,
+      "learning_rate": 7.782909930715936e-06,
+      "loss": 0.6142,
+      "step": 337
+    },
+    {
+      "epoch": 0.02346896264407721,
+      "grad_norm": 6.109046201222556,
+      "learning_rate": 7.806004618937644e-06,
+      "loss": 0.5259,
+      "step": 338
+    },
+    {
+      "epoch": 0.023538397444799332,
+      "grad_norm": 5.117697228312204,
+      "learning_rate": 7.829099307159354e-06,
+      "loss": 0.287,
+      "step": 339
+    },
+    {
+      "epoch": 0.023607832245521456,
+      "grad_norm": 6.01514587593891,
+      "learning_rate": 7.852193995381063e-06,
+      "loss": 0.5645,
+      "step": 340
+    },
+    {
+      "epoch": 0.023677267046243577,
+      "grad_norm": 6.802472620112372,
+      "learning_rate": 7.875288683602773e-06,
+      "loss": 0.498,
+      "step": 341
+    },
+    {
+      "epoch": 0.023746701846965697,
+      "grad_norm": 8.104846417351672,
+      "learning_rate": 7.898383371824481e-06,
+      "loss": 0.806,
+      "step": 342
+    },
+    {
+      "epoch": 0.02381613664768782,
+      "grad_norm": 8.340285985024336,
+      "learning_rate": 7.921478060046191e-06,
+      "loss": 0.7905,
+      "step": 343
+    },
+    {
+      "epoch": 0.023885571448409942,
+      "grad_norm": 6.537761804549327,
+      "learning_rate": 7.9445727482679e-06,
+      "loss": 0.6177,
+      "step": 344
+    },
+    {
+      "epoch": 0.023955006249132067,
+      "grad_norm": 6.988147662879393,
+      "learning_rate": 7.967667436489608e-06,
+      "loss": 0.5674,
+      "step": 345
+    },
+    {
+      "epoch": 0.024024441049854187,
+      "grad_norm": 5.990117475928348,
+      "learning_rate": 7.990762124711316e-06,
+      "loss": 0.6938,
+      "step": 346
+    },
+    {
+      "epoch": 0.024093875850576308,
+      "grad_norm": 5.538492973429358,
+      "learning_rate": 8.013856812933026e-06,
+      "loss": 0.5507,
+      "step": 347
+    },
+    {
+      "epoch": 0.024163310651298432,
+      "grad_norm": 6.067197910485304,
+      "learning_rate": 8.036951501154735e-06,
+      "loss": 0.6371,
+      "step": 348
+    },
+    {
+      "epoch": 0.024232745452020553,
+      "grad_norm": 8.345060036141263,
+      "learning_rate": 8.060046189376443e-06,
+      "loss": 0.8673,
+      "step": 349
+    },
+    {
+      "epoch": 0.024302180252742674,
+      "grad_norm": 7.0232190023213965,
+      "learning_rate": 8.083140877598153e-06,
+      "loss": 0.6343,
+      "step": 350
+    },
+    {
+      "epoch": 0.024371615053464798,
+      "grad_norm": 7.299012686964988,
+      "learning_rate": 8.106235565819862e-06,
+      "loss": 0.4694,
+      "step": 351
+    },
+    {
+      "epoch": 0.02444104985418692,
+      "grad_norm": 5.699938535693354,
+      "learning_rate": 8.129330254041572e-06,
+      "loss": 0.4333,
+      "step": 352
+    },
+    {
+      "epoch": 0.02451048465490904,
+      "grad_norm": 6.64751936730123,
+      "learning_rate": 8.15242494226328e-06,
+      "loss": 0.6341,
+      "step": 353
+    },
+    {
+      "epoch": 0.024579919455631163,
+      "grad_norm": 7.358249173848904,
+      "learning_rate": 8.17551963048499e-06,
+      "loss": 0.7334,
+      "step": 354
+    },
+    {
+      "epoch": 0.024649354256353284,
+      "grad_norm": 6.947107744962895,
+      "learning_rate": 8.198614318706698e-06,
+      "loss": 0.4811,
+      "step": 355
+    },
+    {
+      "epoch": 0.024718789057075405,
+      "grad_norm": 7.9905042925361105,
+      "learning_rate": 8.221709006928407e-06,
+      "loss": 0.808,
+      "step": 356
+    },
+    {
+      "epoch": 0.02478822385779753,
+      "grad_norm": 7.074925682535921,
+      "learning_rate": 8.244803695150117e-06,
+      "loss": 0.6418,
+      "step": 357
+    },
+    {
+      "epoch": 0.02485765865851965,
+      "grad_norm": 6.493440715600506,
+      "learning_rate": 8.267898383371825e-06,
+      "loss": 0.6527,
+      "step": 358
+    },
+    {
+      "epoch": 0.02492709345924177,
+      "grad_norm": 6.252875541049399,
+      "learning_rate": 8.290993071593534e-06,
+      "loss": 0.5673,
+      "step": 359
+    },
+    {
+      "epoch": 0.024996528259963895,
+      "grad_norm": 6.680630293609687,
+      "learning_rate": 8.314087759815244e-06,
+      "loss": 0.859,
+      "step": 360
+    },
+    {
+      "epoch": 0.025065963060686015,
+      "grad_norm": 5.061991943083161,
+      "learning_rate": 8.337182448036952e-06,
+      "loss": 0.4473,
+      "step": 361
+    },
+    {
+      "epoch": 0.025135397861408136,
+      "grad_norm": 6.7000922277432124,
+      "learning_rate": 8.36027713625866e-06,
+      "loss": 0.8466,
+      "step": 362
+    },
+    {
+      "epoch": 0.02520483266213026,
+      "grad_norm": 7.104739112728029,
+      "learning_rate": 8.38337182448037e-06,
+      "loss": 0.5612,
+      "step": 363
+    },
+    {
+      "epoch": 0.02527426746285238,
+      "grad_norm": 5.660476424452743,
+      "learning_rate": 8.406466512702079e-06,
+      "loss": 0.5105,
+      "step": 364
+    },
+    {
+      "epoch": 0.025343702263574505,
+      "grad_norm": 6.204625626329607,
+      "learning_rate": 8.429561200923789e-06,
+      "loss": 0.7786,
+      "step": 365
+    },
+    {
+      "epoch": 0.025413137064296626,
+      "grad_norm": 7.189824870870765,
+      "learning_rate": 8.452655889145497e-06,
+      "loss": 0.7959,
+      "step": 366
+    },
+    {
+      "epoch": 0.025482571865018747,
+      "grad_norm": 5.4005025338163835,
+      "learning_rate": 8.475750577367207e-06,
+      "loss": 0.5015,
+      "step": 367
+    },
+    {
+      "epoch": 0.02555200666574087,
+      "grad_norm": 6.179153950657576,
+      "learning_rate": 8.498845265588916e-06,
+      "loss": 0.8045,
+      "step": 368
+    },
+    {
+      "epoch": 0.02562144146646299,
+      "grad_norm": 6.380841571553142,
+      "learning_rate": 8.521939953810624e-06,
+      "loss": 0.614,
+      "step": 369
+    },
+    {
+      "epoch": 0.025690876267185112,
+      "grad_norm": 5.98211190707568,
+      "learning_rate": 8.545034642032334e-06,
+      "loss": 0.7181,
+      "step": 370
+    },
+    {
+      "epoch": 0.025760311067907236,
+      "grad_norm": 6.716203480188986,
+      "learning_rate": 8.568129330254043e-06,
+      "loss": 0.9596,
+      "step": 371
+    },
+    {
+      "epoch": 0.025829745868629357,
+      "grad_norm": 4.688742121775829,
+      "learning_rate": 8.591224018475751e-06,
+      "loss": 0.3099,
+      "step": 372
+    },
+    {
+      "epoch": 0.025899180669351478,
+      "grad_norm": 6.838852405968659,
+      "learning_rate": 8.61431870669746e-06,
+      "loss": 0.7038,
+      "step": 373
+    },
+    {
+      "epoch": 0.025968615470073602,
+      "grad_norm": 7.882614321083732,
+      "learning_rate": 8.63741339491917e-06,
+      "loss": 0.7109,
+      "step": 374
+    },
+    {
+      "epoch": 0.026038050270795723,
+      "grad_norm": 5.892903512026724,
+      "learning_rate": 8.660508083140878e-06,
+      "loss": 0.659,
+      "step": 375
+    },
+    {
+      "epoch": 0.026107485071517843,
+      "grad_norm": 6.44911963805991,
+      "learning_rate": 8.683602771362586e-06,
+      "loss": 0.8151,
+      "step": 376
+    },
+    {
+      "epoch": 0.026176919872239968,
+      "grad_norm": 4.527131866714667,
+      "learning_rate": 8.706697459584296e-06,
+      "loss": 0.7085,
+      "step": 377
+    },
+    {
+      "epoch": 0.02624635467296209,
+      "grad_norm": 6.9135618427276615,
+      "learning_rate": 8.729792147806006e-06,
+      "loss": 0.2928,
+      "step": 378
+    },
+    {
+      "epoch": 0.02631578947368421,
+      "grad_norm": 6.760408838578817,
+      "learning_rate": 8.752886836027715e-06,
+      "loss": 0.7458,
+      "step": 379
+    },
+    {
+      "epoch": 0.026385224274406333,
+      "grad_norm": 9.24616524555643,
+      "learning_rate": 8.775981524249423e-06,
+      "loss": 0.7244,
+      "step": 380
+    },
+    {
+      "epoch": 0.026454659075128454,
+      "grad_norm": 7.711115269193827,
+      "learning_rate": 8.799076212471133e-06,
+      "loss": 0.3853,
+      "step": 381
+    },
+    {
+      "epoch": 0.026524093875850575,
+      "grad_norm": 7.176188599274549,
+      "learning_rate": 8.822170900692842e-06,
+      "loss": 0.8815,
+      "step": 382
+    },
+    {
+      "epoch": 0.0265935286765727,
+      "grad_norm": 6.619997606508476,
+      "learning_rate": 8.84526558891455e-06,
+      "loss": 0.4085,
+      "step": 383
+    },
+    {
+      "epoch": 0.02666296347729482,
+      "grad_norm": 4.8904274112495045,
+      "learning_rate": 8.86836027713626e-06,
+      "loss": 0.4159,
+      "step": 384
+    },
+    {
+      "epoch": 0.026732398278016944,
+      "grad_norm": 5.867374157660708,
+      "learning_rate": 8.891454965357968e-06,
+      "loss": 0.788,
+      "step": 385
+    },
+    {
+      "epoch": 0.026801833078739064,
+      "grad_norm": 5.958051232927474,
+      "learning_rate": 8.914549653579677e-06,
+      "loss": 0.428,
+      "step": 386
+    },
+    {
+      "epoch": 0.026871267879461185,
+      "grad_norm": 7.0537509318282785,
+      "learning_rate": 8.937644341801387e-06,
+      "loss": 0.7907,
+      "step": 387
+    },
+    {
+      "epoch": 0.02694070268018331,
+      "grad_norm": 6.6452728123702745,
+      "learning_rate": 8.960739030023095e-06,
+      "loss": 0.6033,
+      "step": 388
+    },
+    {
+      "epoch": 0.02701013748090543,
+      "grad_norm": 7.503647623096358,
+      "learning_rate": 8.983833718244804e-06,
+      "loss": 0.7108,
+      "step": 389
+    },
+    {
+      "epoch": 0.02707957228162755,
+      "grad_norm": 4.752411000913283,
+      "learning_rate": 9.006928406466514e-06,
+      "loss": 0.3046,
+      "step": 390
+    },
+    {
+      "epoch": 0.027149007082349675,
+      "grad_norm": 5.876081179515414,
+      "learning_rate": 9.030023094688222e-06,
+      "loss": 0.6925,
+      "step": 391
+    },
+    {
+      "epoch": 0.027218441883071796,
+      "grad_norm": 6.819150079586152,
+      "learning_rate": 9.053117782909932e-06,
+      "loss": 0.8163,
+      "step": 392
+    },
+    {
+      "epoch": 0.027287876683793916,
+      "grad_norm": 6.362870296106029,
+      "learning_rate": 9.07621247113164e-06,
+      "loss": 0.4485,
+      "step": 393
+    },
+    {
+      "epoch": 0.02735731148451604,
+      "grad_norm": 6.918324256538554,
+      "learning_rate": 9.09930715935335e-06,
+      "loss": 0.9359,
+      "step": 394
+    },
+    {
+      "epoch": 0.02742674628523816,
+      "grad_norm": 6.246425940430203,
+      "learning_rate": 9.122401847575059e-06,
+      "loss": 0.3473,
+      "step": 395
+    },
+    {
+      "epoch": 0.027496181085960282,
+      "grad_norm": 6.6533307700734365,
+      "learning_rate": 9.145496535796767e-06,
+      "loss": 0.9051,
+      "step": 396
+    },
+    {
+      "epoch": 0.027565615886682406,
+      "grad_norm": 4.522431898254659,
+      "learning_rate": 9.168591224018476e-06,
+      "loss": 0.4896,
+      "step": 397
+    },
+    {
+      "epoch": 0.027635050687404527,
+      "grad_norm": 5.62359784174198,
+      "learning_rate": 9.191685912240186e-06,
+      "loss": 0.6406,
+      "step": 398
+    },
+    {
+      "epoch": 0.027704485488126648,
+      "grad_norm": 6.554786252346115,
+      "learning_rate": 9.214780600461894e-06,
+      "loss": 0.7186,
+      "step": 399
+    },
+    {
+      "epoch": 0.027773920288848772,
+      "grad_norm": 6.470728191349876,
+      "learning_rate": 9.237875288683602e-06,
+      "loss": 0.5706,
+      "step": 400
+    },
+    {
+      "epoch": 0.027843355089570893,
+      "grad_norm": 5.696581976585349,
+      "learning_rate": 9.260969976905312e-06,
+      "loss": 0.6381,
+      "step": 401
+    },
+    {
+      "epoch": 0.027912789890293013,
+      "grad_norm": 6.827037938775935,
+      "learning_rate": 9.284064665127021e-06,
+      "loss": 0.7833,
+      "step": 402
+    },
+    {
+      "epoch": 0.027982224691015137,
+      "grad_norm": 6.560358522649497,
+      "learning_rate": 9.307159353348731e-06,
+      "loss": 0.6933,
+      "step": 403
+    },
+    {
+      "epoch": 0.028051659491737258,
+      "grad_norm": 6.647170924819252,
+      "learning_rate": 9.33025404157044e-06,
+      "loss": 0.4499,
+      "step": 404
+    },
+    {
+      "epoch": 0.028121094292459382,
+      "grad_norm": 5.130789644873473,
+      "learning_rate": 9.35334872979215e-06,
+      "loss": 0.4588,
+      "step": 405
+    },
+    {
+      "epoch": 0.028190529093181503,
+      "grad_norm": 4.860021733661967,
+      "learning_rate": 9.376443418013858e-06,
+      "loss": 0.2815,
+      "step": 406
+    },
+    {
+      "epoch": 0.028259963893903624,
+      "grad_norm": 4.90820253480832,
+      "learning_rate": 9.399538106235566e-06,
+      "loss": 0.4098,
+      "step": 407
+    },
+    {
+      "epoch": 0.028329398694625748,
+      "grad_norm": 4.941040654728571,
+      "learning_rate": 9.422632794457276e-06,
+      "loss": 0.4002,
+      "step": 408
+    },
+    {
+      "epoch": 0.02839883349534787,
+      "grad_norm": 7.675709003678412,
+      "learning_rate": 9.445727482678985e-06,
+      "loss": 0.9723,
+      "step": 409
+    },
+    {
+      "epoch": 0.02846826829606999,
+      "grad_norm": 8.592025328623098,
+      "learning_rate": 9.468822170900693e-06,
+      "loss": 0.2338,
+      "step": 410
+    },
+    {
+      "epoch": 0.028537703096792114,
+      "grad_norm": 5.993286898711553,
+      "learning_rate": 9.491916859122403e-06,
+      "loss": 0.5384,
+      "step": 411
+    },
+    {
+      "epoch": 0.028607137897514234,
+      "grad_norm": 5.8652312391691055,
+      "learning_rate": 9.515011547344111e-06,
+      "loss": 0.4496,
+      "step": 412
+    },
+    {
+      "epoch": 0.028676572698236355,
+      "grad_norm": 4.901801928340526,
+      "learning_rate": 9.53810623556582e-06,
+      "loss": 0.4712,
+      "step": 413
+    },
+    {
+      "epoch": 0.02874600749895848,
+      "grad_norm": 7.614399384579382,
+      "learning_rate": 9.56120092378753e-06,
+      "loss": 0.6751,
+      "step": 414
+    },
+    {
+      "epoch": 0.0288154422996806,
+      "grad_norm": 4.412664038311866,
+      "learning_rate": 9.584295612009238e-06,
+      "loss": 0.3466,
+      "step": 415
+    },
+    {
+      "epoch": 0.02888487710040272,
+      "grad_norm": 7.238038321499026,
+      "learning_rate": 9.607390300230948e-06,
+      "loss": 0.9787,
+      "step": 416
+    },
+    {
+      "epoch": 0.028954311901124845,
+      "grad_norm": 6.94095866789776,
+      "learning_rate": 9.630484988452657e-06,
+      "loss": 0.9159,
+      "step": 417
+    },
+    {
+      "epoch": 0.029023746701846966,
+      "grad_norm": 5.653221161555569,
+      "learning_rate": 9.653579676674367e-06,
+      "loss": 0.4889,
+      "step": 418
+    },
+    {
+      "epoch": 0.029093181502569086,
+      "grad_norm": 5.047780685821234,
+      "learning_rate": 9.676674364896075e-06,
+      "loss": 0.4796,
+      "step": 419
+    },
+    {
+      "epoch": 0.02916261630329121,
+      "grad_norm": 6.069627117715251,
+      "learning_rate": 9.699769053117783e-06,
+      "loss": 0.4571,
+      "step": 420
+    },
+    {
+      "epoch": 0.02923205110401333,
+      "grad_norm": 6.496502890553569,
+      "learning_rate": 9.722863741339492e-06,
+      "loss": 0.8334,
+      "step": 421
+    },
+    {
+      "epoch": 0.029301485904735452,
+      "grad_norm": 5.687072942120927,
+      "learning_rate": 9.745958429561202e-06,
+      "loss": 0.6352,
+      "step": 422
+    },
+    {
+      "epoch": 0.029370920705457576,
+      "grad_norm": 6.279007178298347,
+      "learning_rate": 9.76905311778291e-06,
+      "loss": 0.7161,
+      "step": 423
+    },
+    {
+      "epoch": 0.029440355506179697,
+      "grad_norm": 4.882604390084902,
+      "learning_rate": 9.792147806004619e-06,
+      "loss": 0.5925,
+      "step": 424
+    },
+    {
+      "epoch": 0.029509790306901817,
+      "grad_norm": 5.593126759659433,
+      "learning_rate": 9.815242494226329e-06,
+      "loss": 0.5873,
+      "step": 425
+    },
+    {
+      "epoch": 0.02957922510762394,
+      "grad_norm": 6.904887221887285,
+      "learning_rate": 9.838337182448037e-06,
+      "loss": 0.5615,
+      "step": 426
+    },
+    {
+      "epoch": 0.029648659908346062,
+      "grad_norm": 7.180240302055389,
+      "learning_rate": 9.861431870669747e-06,
+      "loss": 0.5184,
+      "step": 427
+    },
+    {
+      "epoch": 0.029718094709068187,
+      "grad_norm": 7.498208464056241,
+      "learning_rate": 9.884526558891456e-06,
+      "loss": 0.7807,
+      "step": 428
+    },
+    {
+      "epoch": 0.029787529509790307,
+      "grad_norm": 4.391965369061992,
+      "learning_rate": 9.907621247113166e-06,
+      "loss": 0.4396,
+      "step": 429
+    },
+    {
+      "epoch": 0.029856964310512428,
+      "grad_norm": 7.582962848348381,
+      "learning_rate": 9.930715935334874e-06,
+      "loss": 0.7123,
+      "step": 430
+    },
+    {
+      "epoch": 0.029926399111234552,
+      "grad_norm": 5.231346626173306,
+      "learning_rate": 9.953810623556582e-06,
+      "loss": 0.4976,
+      "step": 431
+    },
+    {
+      "epoch": 0.029995833911956673,
+      "grad_norm": 5.0576953003156255,
+      "learning_rate": 9.976905311778292e-06,
+      "loss": 0.5557,
+      "step": 432
+    },
+    {
+      "epoch": 0.030065268712678794,
+      "grad_norm": 6.9069570660612145,
+      "learning_rate": 1e-05,
+      "loss": 0.7385,
+      "step": 433
+    },
+    {
+      "epoch": 0.030134703513400918,
+      "grad_norm": 7.778669181247511,
+      "learning_rate": 9.99999987355283e-06,
+      "loss": 0.6871,
+      "step": 434
+    },
+    {
+      "epoch": 0.03020413831412304,
+      "grad_norm": 5.357549502532245,
+      "learning_rate": 9.999999494211324e-06,
+      "loss": 0.8859,
+      "step": 435
+    },
+    {
+      "epoch": 0.03027357311484516,
+      "grad_norm": 6.284819787134612,
+      "learning_rate": 9.9999988619755e-06,
+      "loss": 0.5724,
+      "step": 436
+    },
+    {
+      "epoch": 0.030343007915567283,
+      "grad_norm": 6.831598996626386,
+      "learning_rate": 9.999997976845392e-06,
+      "loss": 0.6657,
+      "step": 437
+    },
+    {
+      "epoch": 0.030412442716289404,
+      "grad_norm": 6.27780986608783,
+      "learning_rate": 9.999996838821044e-06,
+      "loss": 0.7065,
+      "step": 438
+    },
+    {
+      "epoch": 0.030481877517011525,
+      "grad_norm": 5.045048202924045,
+      "learning_rate": 9.999995447902514e-06,
+      "loss": 0.5953,
+      "step": 439
+    },
+    {
+      "epoch": 0.03055131231773365,
+      "grad_norm": 5.842654195692098,
+      "learning_rate": 9.999993804089873e-06,
+      "loss": 0.6818,
+      "step": 440
+    },
+    {
+      "epoch": 0.03062074711845577,
+      "grad_norm": 6.460290689885131,
+      "learning_rate": 9.999991907383202e-06,
+      "loss": 0.6786,
+      "step": 441
+    },
+    {
+      "epoch": 0.03069018191917789,
+      "grad_norm": 2.861495331541182,
+      "learning_rate": 9.999989757782599e-06,
+      "loss": 0.1267,
+      "step": 442
+    },
+    {
+      "epoch": 0.030759616719900015,
+      "grad_norm": 5.029345903596466,
+      "learning_rate": 9.999987355288172e-06,
+      "loss": 0.4737,
+      "step": 443
+    },
+    {
+      "epoch": 0.030829051520622135,
+      "grad_norm": 6.080178030290366,
+      "learning_rate": 9.999984699900042e-06,
+      "loss": 0.5977,
+      "step": 444
+    },
+    {
+      "epoch": 0.030898486321344256,
+      "grad_norm": 6.755300927136362,
+      "learning_rate": 9.999981791618343e-06,
+      "loss": 0.8255,
+      "step": 445
+    },
+    {
+      "epoch": 0.03096792112206638,
+      "grad_norm": 6.768465028511879,
+      "learning_rate": 9.999978630443223e-06,
+      "loss": 0.5011,
+      "step": 446
+    },
+    {
+      "epoch": 0.0310373559227885,
+      "grad_norm": 6.5119485992982185,
+      "learning_rate": 9.999975216374843e-06,
+      "loss": 0.5937,
+      "step": 447
+    },
+    {
+      "epoch": 0.031106790723510625,
+      "grad_norm": 3.6145781303228954,
+      "learning_rate": 9.999971549413374e-06,
+      "loss": 0.2656,
+      "step": 448
+    },
+    {
+      "epoch": 0.031176225524232746,
+      "grad_norm": 5.395036276138725,
+      "learning_rate": 9.999967629559002e-06,
+      "loss": 0.6524,
+      "step": 449
+    },
+    {
+      "epoch": 0.031245660324954867,
+      "grad_norm": 6.3267192533831915,
+      "learning_rate": 9.999963456811926e-06,
+      "loss": 0.6883,
+      "step": 450
+    },
+    {
+      "epoch": 0.03131509512567699,
+      "grad_norm": 5.323953199320294,
+      "learning_rate": 9.999959031172353e-06,
+      "loss": 0.6204,
+      "step": 451
+    },
+    {
+      "epoch": 0.03138452992639911,
+      "grad_norm": 6.599704105261902,
+      "learning_rate": 9.999954352640512e-06,
+      "loss": 0.7781,
+      "step": 452
+    },
+    {
+      "epoch": 0.03145396472712123,
+      "grad_norm": 4.986954276383524,
+      "learning_rate": 9.999949421216638e-06,
+      "loss": 0.4345,
+      "step": 453
+    },
+    {
+      "epoch": 0.03152339952784335,
+      "grad_norm": 5.247116121837747,
+      "learning_rate": 9.999944236900981e-06,
+      "loss": 0.61,
+      "step": 454
+    },
+    {
+      "epoch": 0.031592834328565474,
+      "grad_norm": 6.211583434456717,
+      "learning_rate": 9.9999387996938e-06,
+      "loss": 0.6007,
+      "step": 455
+    },
+    {
+      "epoch": 0.0316622691292876,
+      "grad_norm": 5.668713988535656,
+      "learning_rate": 9.999933109595375e-06,
+      "loss": 0.4547,
+      "step": 456
+    },
+    {
+      "epoch": 0.03173170393000972,
+      "grad_norm": 6.079668473959285,
+      "learning_rate": 9.999927166605988e-06,
+      "loss": 0.6103,
+      "step": 457
+    },
+    {
+      "epoch": 0.03180113873073184,
+      "grad_norm": 6.064849792893844,
+      "learning_rate": 9.999920970725945e-06,
+      "loss": 0.7693,
+      "step": 458
+    },
+    {
+      "epoch": 0.03187057353145396,
+      "grad_norm": 5.738259981227205,
+      "learning_rate": 9.999914521955557e-06,
+      "loss": 0.5279,
+      "step": 459
+    },
+    {
+      "epoch": 0.031940008332176084,
+      "grad_norm": 4.5610230068039925,
+      "learning_rate": 9.99990782029515e-06,
+      "loss": 0.4056,
+      "step": 460
+    },
+    {
+      "epoch": 0.03200944313289821,
+      "grad_norm": 6.936309822383197,
+      "learning_rate": 9.999900865745061e-06,
+      "loss": 0.818,
+      "step": 461
+    },
+    {
+      "epoch": 0.03207887793362033,
+      "grad_norm": 6.286528207571545,
+      "learning_rate": 9.999893658305646e-06,
+      "loss": 0.6776,
+      "step": 462
+    },
+    {
+      "epoch": 0.03214831273434245,
+      "grad_norm": 6.176488771304935,
+      "learning_rate": 9.999886197977267e-06,
+      "loss": 0.5587,
+      "step": 463
+    },
+    {
+      "epoch": 0.032217747535064574,
+      "grad_norm": 5.627431279771351,
+      "learning_rate": 9.999878484760304e-06,
+      "loss": 0.5378,
+      "step": 464
+    },
+    {
+      "epoch": 0.032287182335786695,
+      "grad_norm": 5.519314095876356,
+      "learning_rate": 9.999870518655142e-06,
+      "loss": 0.3596,
+      "step": 465
+    },
+    {
+      "epoch": 0.032356617136508815,
+      "grad_norm": 5.402724973265845,
+      "learning_rate": 9.999862299662188e-06,
+      "loss": 0.3665,
+      "step": 466
+    },
+    {
+      "epoch": 0.03242605193723094,
+      "grad_norm": 6.3812711140790075,
+      "learning_rate": 9.999853827781857e-06,
+      "loss": 0.8771,
+      "step": 467
+    },
+    {
+      "epoch": 0.032495486737953064,
+      "grad_norm": 5.9338189617621415,
+      "learning_rate": 9.999845103014577e-06,
+      "loss": 0.6081,
+      "step": 468
+    },
+    {
+      "epoch": 0.032564921538675184,
+      "grad_norm": 5.883826060939156,
+      "learning_rate": 9.999836125360789e-06,
+      "loss": 0.6843,
+      "step": 469
+    },
+    {
+      "epoch": 0.032634356339397305,
+      "grad_norm": 5.411641552716915,
+      "learning_rate": 9.999826894820947e-06,
+      "loss": 0.4745,
+      "step": 470
+    },
+    {
+      "epoch": 0.032703791140119426,
+      "grad_norm": 6.692910460347541,
+      "learning_rate": 9.99981741139552e-06,
+      "loss": 0.9,
+      "step": 471
+    },
+    {
+      "epoch": 0.03277322594084155,
+      "grad_norm": 5.961849396614783,
+      "learning_rate": 9.999807675084984e-06,
+      "loss": 0.5387,
+      "step": 472
+    },
+    {
+      "epoch": 0.032842660741563674,
+      "grad_norm": 6.689824686070395,
+      "learning_rate": 9.999797685889836e-06,
+      "loss": 0.6502,
+      "step": 473
+    },
+    {
+      "epoch": 0.032912095542285795,
+      "grad_norm": 6.058811668661387,
+      "learning_rate": 9.999787443810576e-06,
+      "loss": 0.6155,
+      "step": 474
+    },
+    {
+      "epoch": 0.032981530343007916,
+      "grad_norm": 5.9874867549465405,
+      "learning_rate": 9.999776948847727e-06,
+      "loss": 0.6768,
+      "step": 475
+    },
+    {
+      "epoch": 0.033050965143730036,
+      "grad_norm": 5.651512684905336,
+      "learning_rate": 9.999766201001817e-06,
+      "loss": 0.4318,
+      "step": 476
+    },
+    {
+      "epoch": 0.03312039994445216,
+      "grad_norm": 5.028687361869108,
+      "learning_rate": 9.999755200273389e-06,
+      "loss": 0.5417,
+      "step": 477
+    },
+    {
+      "epoch": 0.033189834745174285,
+      "grad_norm": 5.126187229503819,
+      "learning_rate": 9.999743946663002e-06,
+      "loss": 0.3153,
+      "step": 478
+    },
+    {
+      "epoch": 0.033259269545896405,
+      "grad_norm": 5.257781775734921,
+      "learning_rate": 9.999732440171223e-06,
+      "loss": 0.5462,
+      "step": 479
+    },
+    {
+      "epoch": 0.033328704346618526,
+      "grad_norm": 5.255353219812936,
+      "learning_rate": 9.999720680798635e-06,
+      "loss": 0.5243,
+      "step": 480
+    },
+    {
+      "epoch": 0.03339813914734065,
+      "grad_norm": 5.9253259357897985,
+      "learning_rate": 9.999708668545831e-06,
+      "loss": 0.5819,
+      "step": 481
+    },
+    {
+      "epoch": 0.03346757394806277,
+      "grad_norm": 6.184543080330575,
+      "learning_rate": 9.999696403413423e-06,
+      "loss": 0.6708,
+      "step": 482
+    },
+    {
+      "epoch": 0.03353700874878489,
+      "grad_norm": 6.024552485805366,
+      "learning_rate": 9.999683885402026e-06,
+      "loss": 0.5302,
+      "step": 483
+    },
+    {
+      "epoch": 0.033606443549507016,
+      "grad_norm": 7.399753261509762,
+      "learning_rate": 9.999671114512277e-06,
+      "loss": 0.4729,
+      "step": 484
+    },
+    {
+      "epoch": 0.03367587835022914,
+      "grad_norm": 5.55269525780665,
+      "learning_rate": 9.999658090744818e-06,
+      "loss": 0.4701,
+      "step": 485
+    },
+    {
+      "epoch": 0.03374531315095126,
+      "grad_norm": 5.199497229607478,
+      "learning_rate": 9.999644814100313e-06,
+      "loss": 0.3629,
+      "step": 486
+    },
+    {
+      "epoch": 0.03381474795167338,
+      "grad_norm": 6.080533607320299,
+      "learning_rate": 9.999631284579433e-06,
+      "loss": 0.7568,
+      "step": 487
+    },
+    {
+      "epoch": 0.0338841827523955,
+      "grad_norm": 6.018959158711198,
+      "learning_rate": 9.999617502182854e-06,
+      "loss": 0.5558,
+      "step": 488
+    },
+    {
+      "epoch": 0.03395361755311762,
+      "grad_norm": 5.674459181219298,
+      "learning_rate": 9.999603466911284e-06,
+      "loss": 0.4269,
+      "step": 489
+    },
+    {
+      "epoch": 0.03402305235383975,
+      "grad_norm": 6.274947389477818,
+      "learning_rate": 9.999589178765426e-06,
+      "loss": 0.5257,
+      "step": 490
+    },
+    {
+      "epoch": 0.03409248715456187,
+      "grad_norm": 4.986999759560102,
+      "learning_rate": 9.999574637746006e-06,
+      "loss": 0.3431,
+      "step": 491
+    },
+    {
+      "epoch": 0.03416192195528399,
+      "grad_norm": 5.965796935867004,
+      "learning_rate": 9.999559843853756e-06,
+      "loss": 0.8307,
+      "step": 492
+    },
+    {
+      "epoch": 0.03423135675600611,
+      "grad_norm": 6.066166717319229,
+      "learning_rate": 9.999544797089428e-06,
+      "loss": 0.4083,
+      "step": 493
+    },
+    {
+      "epoch": 0.03430079155672823,
+      "grad_norm": 6.165297196657152,
+      "learning_rate": 9.999529497453782e-06,
+      "loss": 0.741,
+      "step": 494
+    },
+    {
+      "epoch": 0.03437022635745035,
+      "grad_norm": 7.944318113381601,
+      "learning_rate": 9.999513944947591e-06,
+      "loss": 0.8734,
+      "step": 495
+    },
+    {
+      "epoch": 0.03443966115817248,
+      "grad_norm": 7.618278080480178,
+      "learning_rate": 9.999498139571642e-06,
+      "loss": 0.3951,
+      "step": 496
+    },
+    {
+      "epoch": 0.0345090959588946,
+      "grad_norm": 5.624323598064029,
+      "learning_rate": 9.999482081326734e-06,
+      "loss": 0.4671,
+      "step": 497
+    },
+    {
+      "epoch": 0.03457853075961672,
+      "grad_norm": 6.163382158823257,
+      "learning_rate": 9.99946577021368e-06,
+      "loss": 0.7706,
+      "step": 498
+    },
+    {
+      "epoch": 0.03464796556033884,
+      "grad_norm": 5.860245938439233,
+      "learning_rate": 9.999449206233305e-06,
+      "loss": 0.4915,
+      "step": 499
+    },
+    {
+      "epoch": 0.03471740036106096,
+      "grad_norm": 6.531519831041931,
+      "learning_rate": 9.999432389386445e-06,
+      "loss": 0.6789,
+      "step": 500
+    },
+    {
+      "epoch": 0.03478683516178309,
+      "grad_norm": 4.352611016790576,
+      "learning_rate": 9.999415319673953e-06,
+      "loss": 0.5366,
+      "step": 501
+    },
+    {
+      "epoch": 0.03485626996250521,
+      "grad_norm": 6.771348467977927,
+      "learning_rate": 9.999397997096692e-06,
+      "loss": 0.8456,
+      "step": 502
+    },
+    {
+      "epoch": 0.03492570476322733,
+      "grad_norm": 4.854548394309112,
+      "learning_rate": 9.999380421655535e-06,
+      "loss": 0.336,
+      "step": 503
+    },
+    {
+      "epoch": 0.03499513956394945,
+      "grad_norm": 5.748395146004432,
+      "learning_rate": 9.999362593351376e-06,
+      "loss": 0.6206,
+      "step": 504
+    },
+    {
+      "epoch": 0.03506457436467157,
+      "grad_norm": 7.4122792634318895,
+      "learning_rate": 9.999344512185114e-06,
+      "loss": 0.7026,
+      "step": 505
+    },
+    {
+      "epoch": 0.03513400916539369,
+      "grad_norm": 5.46496670740742,
+      "learning_rate": 9.999326178157662e-06,
+      "loss": 0.6925,
+      "step": 506
+    },
+    {
+      "epoch": 0.03520344396611582,
+      "grad_norm": 6.658261234910614,
+      "learning_rate": 9.999307591269951e-06,
+      "loss": 0.5679,
+      "step": 507
+    },
+    {
+      "epoch": 0.03527287876683794,
+      "grad_norm": 5.145984865551088,
+      "learning_rate": 9.999288751522918e-06,
+      "loss": 0.4936,
+      "step": 508
+    },
+    {
+      "epoch": 0.03534231356756006,
+      "grad_norm": 6.230757230212693,
+      "learning_rate": 9.999269658917517e-06,
+      "loss": 0.8232,
+      "step": 509
+    },
+    {
+      "epoch": 0.03541174836828218,
+      "grad_norm": 4.704307621881029,
+      "learning_rate": 9.999250313454714e-06,
+      "loss": 0.33,
+      "step": 510
+    },
+    {
+      "epoch": 0.0354811831690043,
+      "grad_norm": 6.333297779913553,
+      "learning_rate": 9.999230715135488e-06,
+      "loss": 0.7925,
+      "step": 511
+    },
+    {
+      "epoch": 0.035550617969726424,
+      "grad_norm": 4.931566885802384,
+      "learning_rate": 9.99921086396083e-06,
+      "loss": 0.5662,
+      "step": 512
+    },
+    {
+      "epoch": 0.03562005277044855,
+      "grad_norm": 6.6737933855309075,
+      "learning_rate": 9.999190759931741e-06,
+      "loss": 0.7374,
+      "step": 513
+    },
+    {
+      "epoch": 0.03568948757117067,
+      "grad_norm": 5.588899464065289,
+      "learning_rate": 9.99917040304924e-06,
+      "loss": 0.6212,
+      "step": 514
+    },
+    {
+      "epoch": 0.03575892237189279,
+      "grad_norm": 6.2235116182363175,
+      "learning_rate": 9.99914979331436e-06,
+      "loss": 0.5285,
+      "step": 515
+    },
+    {
+      "epoch": 0.035828357172614914,
+      "grad_norm": 6.292876061387508,
+      "learning_rate": 9.999128930728138e-06,
+      "loss": 0.5399,
+      "step": 516
+    },
+    {
+      "epoch": 0.035897791973337034,
+      "grad_norm": 5.228767276828957,
+      "learning_rate": 9.999107815291632e-06,
+      "loss": 0.4125,
+      "step": 517
+    },
+    {
+      "epoch": 0.035967226774059155,
+      "grad_norm": 5.313882172315606,
+      "learning_rate": 9.99908644700591e-06,
+      "loss": 0.5639,
+      "step": 518
+    },
+    {
+      "epoch": 0.03603666157478128,
+      "grad_norm": 5.772639537012846,
+      "learning_rate": 9.999064825872052e-06,
+      "loss": 0.6747,
+      "step": 519
+    },
+    {
+      "epoch": 0.0361060963755034,
+      "grad_norm": 6.22948282300013,
+      "learning_rate": 9.999042951891152e-06,
+      "loss": 0.5887,
+      "step": 520
+    },
+    {
+      "epoch": 0.036175531176225524,
+      "grad_norm": 6.013811875298543,
+      "learning_rate": 9.999020825064318e-06,
+      "loss": 0.6529,
+      "step": 521
+    },
+    {
+      "epoch": 0.036244965976947645,
+      "grad_norm": 5.525633323920022,
+      "learning_rate": 9.998998445392665e-06,
+      "loss": 0.6372,
+      "step": 522
+    },
+    {
+      "epoch": 0.036314400777669766,
+      "grad_norm": 5.230982960847445,
+      "learning_rate": 9.998975812877328e-06,
+      "loss": 0.5117,
+      "step": 523
+    },
+    {
+      "epoch": 0.03638383557839189,
+      "grad_norm": 5.419072106730317,
+      "learning_rate": 9.998952927519452e-06,
+      "loss": 0.5176,
+      "step": 524
+    },
+    {
+      "epoch": 0.036453270379114014,
+      "grad_norm": 6.494550656664596,
+      "learning_rate": 9.998929789320192e-06,
+      "loss": 0.7149,
+      "step": 525
+    },
+    {
+      "epoch": 0.036522705179836135,
+      "grad_norm": 6.302959404371616,
+      "learning_rate": 9.998906398280723e-06,
+      "loss": 0.5371,
+      "step": 526
+    },
+    {
+      "epoch": 0.036592139980558255,
+      "grad_norm": 5.494105037980103,
+      "learning_rate": 9.998882754402221e-06,
+      "loss": 0.3644,
+      "step": 527
+    },
+    {
+      "epoch": 0.036661574781280376,
+      "grad_norm": 6.778447536483514,
+      "learning_rate": 9.99885885768589e-06,
+      "loss": 0.779,
+      "step": 528
+    },
+    {
+      "epoch": 0.0367310095820025,
+      "grad_norm": 3.822710549909421,
+      "learning_rate": 9.998834708132932e-06,
+      "loss": 0.2716,
+      "step": 529
+    },
+    {
+      "epoch": 0.036800444382724624,
+      "grad_norm": 5.709501580025093,
+      "learning_rate": 9.998810305744571e-06,
+      "loss": 0.4167,
+      "step": 530
+    },
+    {
+      "epoch": 0.036869879183446745,
+      "grad_norm": 5.083773330678954,
+      "learning_rate": 9.998785650522041e-06,
+      "loss": 0.5782,
+      "step": 531
+    },
+    {
+      "epoch": 0.036939313984168866,
+      "grad_norm": 5.079229343723481,
+      "learning_rate": 9.998760742466591e-06,
+      "loss": 0.2723,
+      "step": 532
+    },
+    {
+      "epoch": 0.03700874878489099,
+      "grad_norm": 7.414963449830528,
+      "learning_rate": 9.998735581579478e-06,
+      "loss": 0.8686,
+      "step": 533
+    },
+    {
+      "epoch": 0.03707818358561311,
+      "grad_norm": 5.017802446829348,
+      "learning_rate": 9.998710167861976e-06,
+      "loss": 0.4596,
+      "step": 534
+    },
+    {
+      "epoch": 0.03714761838633523,
+      "grad_norm": 5.1978148470449685,
+      "learning_rate": 9.99868450131537e-06,
+      "loss": 0.6486,
+      "step": 535
+    },
+    {
+      "epoch": 0.037217053187057356,
+      "grad_norm": 6.365903071705479,
+      "learning_rate": 9.998658581940959e-06,
+      "loss": 0.6638,
+      "step": 536
+    },
+    {
+      "epoch": 0.037286487987779476,
+      "grad_norm": 6.304925677062259,
+      "learning_rate": 9.998632409740052e-06,
+      "loss": 0.6632,
+      "step": 537
+    },
+    {
+      "epoch": 0.0373559227885016,
+      "grad_norm": 4.71526936573796,
+      "learning_rate": 9.998605984713975e-06,
+      "loss": 0.5156,
+      "step": 538
+    },
+    {
+      "epoch": 0.03742535758922372,
+      "grad_norm": 5.306462620702109,
+      "learning_rate": 9.998579306864062e-06,
+      "loss": 0.4895,
+      "step": 539
+    },
+    {
+      "epoch": 0.03749479238994584,
+      "grad_norm": 6.640155197930903,
+      "learning_rate": 9.998552376191665e-06,
+      "loss": 0.76,
+      "step": 540
+    },
+    {
+      "epoch": 0.037564227190667966,
+      "grad_norm": 6.407428092082328,
+      "learning_rate": 9.998525192698148e-06,
+      "loss": 0.6922,
+      "step": 541
+    },
+    {
+      "epoch": 0.03763366199139009,
+      "grad_norm": 6.319311262754528,
+      "learning_rate": 9.99849775638488e-06,
+      "loss": 0.718,
+      "step": 542
+    },
+    {
+      "epoch": 0.03770309679211221,
+      "grad_norm": 5.646072684540412,
+      "learning_rate": 9.998470067253251e-06,
+      "loss": 0.4708,
+      "step": 543
+    },
+    {
+      "epoch": 0.03777253159283433,
+      "grad_norm": 6.386419015831803,
+      "learning_rate": 9.998442125304664e-06,
+      "loss": 0.6598,
+      "step": 544
+    },
+    {
+      "epoch": 0.03784196639355645,
+      "grad_norm": 4.795924051829016,
+      "learning_rate": 9.99841393054053e-06,
+      "loss": 0.4261,
+      "step": 545
+    },
+    {
+      "epoch": 0.03791140119427857,
+      "grad_norm": 6.026562026792236,
+      "learning_rate": 9.998385482962274e-06,
+      "loss": 0.8194,
+      "step": 546
+    },
+    {
+      "epoch": 0.0379808359950007,
+      "grad_norm": 4.335606456591261,
+      "learning_rate": 9.998356782571336e-06,
+      "loss": 0.3986,
+      "step": 547
+    },
+    {
+      "epoch": 0.03805027079572282,
+      "grad_norm": 5.1976334497067995,
+      "learning_rate": 9.99832782936917e-06,
+      "loss": 0.5406,
+      "step": 548
+    },
+    {
+      "epoch": 0.03811970559644494,
+      "grad_norm": 6.08563476776115,
+      "learning_rate": 9.998298623357238e-06,
+      "loss": 0.7269,
+      "step": 549
+    },
+    {
+      "epoch": 0.03818914039716706,
+      "grad_norm": 5.65762912255376,
+      "learning_rate": 9.998269164537018e-06,
+      "loss": 0.3486,
+      "step": 550
+    },
+    {
+      "epoch": 0.03825857519788918,
+      "grad_norm": 6.651051283287734,
+      "learning_rate": 9.998239452909998e-06,
+      "loss": 0.6524,
+      "step": 551
+    },
+    {
+      "epoch": 0.0383280099986113,
+      "grad_norm": 6.3433191654376815,
+      "learning_rate": 9.998209488477683e-06,
+      "loss": 0.9663,
+      "step": 552
+    },
+    {
+      "epoch": 0.03839744479933343,
+      "grad_norm": 4.658874637603492,
+      "learning_rate": 9.998179271241587e-06,
+      "loss": 0.4584,
+      "step": 553
+    },
+    {
+      "epoch": 0.03846687960005555,
+      "grad_norm": 5.003904426291276,
+      "learning_rate": 9.998148801203242e-06,
+      "loss": 0.564,
+      "step": 554
+    },
+    {
+      "epoch": 0.03853631440077767,
+      "grad_norm": 5.813184614200642,
+      "learning_rate": 9.998118078364186e-06,
+      "loss": 0.5276,
+      "step": 555
+    },
+    {
+      "epoch": 0.03860574920149979,
+      "grad_norm": 6.217352664495251,
+      "learning_rate": 9.99808710272597e-06,
+      "loss": 0.7469,
+      "step": 556
+    },
+    {
+      "epoch": 0.03867518400222191,
+      "grad_norm": 6.362511327994057,
+      "learning_rate": 9.998055874290166e-06,
+      "loss": 0.693,
+      "step": 557
+    },
+    {
+      "epoch": 0.03874461880294403,
+      "grad_norm": 5.697222739415906,
+      "learning_rate": 9.998024393058353e-06,
+      "loss": 0.4795,
+      "step": 558
+    },
+    {
+      "epoch": 0.03881405360366616,
+      "grad_norm": 5.584251313085637,
+      "learning_rate": 9.99799265903212e-06,
+      "loss": 0.6306,
+      "step": 559
+    },
+    {
+      "epoch": 0.03888348840438828,
+      "grad_norm": 5.907968337914951,
+      "learning_rate": 9.997960672213076e-06,
+      "loss": 0.5679,
+      "step": 560
+    },
+    {
+      "epoch": 0.0389529232051104,
+      "grad_norm": 5.654380218521138,
+      "learning_rate": 9.997928432602834e-06,
+      "loss": 0.6453,
+      "step": 561
+    },
+    {
+      "epoch": 0.03902235800583252,
+      "grad_norm": 5.579514915789915,
+      "learning_rate": 9.99789594020303e-06,
+      "loss": 0.58,
+      "step": 562
+    },
+    {
+      "epoch": 0.03909179280655464,
+      "grad_norm": 3.797666493065679,
+      "learning_rate": 9.997863195015303e-06,
+      "loss": 0.4575,
+      "step": 563
+    },
+    {
+      "epoch": 0.03916122760727677,
+      "grad_norm": 5.883390941564609,
+      "learning_rate": 9.997830197041312e-06,
+      "loss": 0.3515,
+      "step": 564
+    },
+    {
+      "epoch": 0.03923066240799889,
+      "grad_norm": 5.94720735536341,
+      "learning_rate": 9.997796946282725e-06,
+      "loss": 0.7877,
+      "step": 565
+    },
+    {
+      "epoch": 0.03930009720872101,
+      "grad_norm": 7.15133606832761,
+      "learning_rate": 9.997763442741225e-06,
+      "loss": 0.9724,
+      "step": 566
+    },
+    {
+      "epoch": 0.03936953200944313,
+      "grad_norm": 5.575033621973364,
+      "learning_rate": 9.997729686418502e-06,
+      "loss": 0.441,
+      "step": 567
+    },
+    {
+      "epoch": 0.03943896681016525,
+      "grad_norm": 4.833553273250969,
+      "learning_rate": 9.99769567731627e-06,
+      "loss": 0.5724,
+      "step": 568
+    },
+    {
+      "epoch": 0.039508401610887374,
+      "grad_norm": 6.013917168462498,
+      "learning_rate": 9.997661415436244e-06,
+      "loss": 0.7292,
+      "step": 569
+    },
+    {
+      "epoch": 0.0395778364116095,
+      "grad_norm": 4.3930702335539955,
+      "learning_rate": 9.99762690078016e-06,
+      "loss": 0.3436,
+      "step": 570
+    },
+    {
+      "epoch": 0.03964727121233162,
+      "grad_norm": 5.7270457492432145,
+      "learning_rate": 9.997592133349761e-06,
+      "loss": 0.5271,
+      "step": 571
+    },
+    {
+      "epoch": 0.03971670601305374,
+      "grad_norm": 5.407928229591425,
+      "learning_rate": 9.997557113146808e-06,
+      "loss": 0.6306,
+      "step": 572
+    },
+    {
+      "epoch": 0.039786140813775864,
+      "grad_norm": 7.0475262448294025,
+      "learning_rate": 9.997521840173073e-06,
+      "loss": 0.6983,
+      "step": 573
+    },
+    {
+      "epoch": 0.039855575614497984,
+      "grad_norm": 5.745509503266648,
+      "learning_rate": 9.997486314430336e-06,
+      "loss": 0.525,
+      "step": 574
+    },
+    {
+      "epoch": 0.039925010415220105,
+      "grad_norm": 7.301695667997415,
+      "learning_rate": 9.997450535920397e-06,
+      "loss": 0.7057,
+      "step": 575
+    },
+    {
+      "epoch": 0.03999444521594223,
+      "grad_norm": 6.551702374771802,
+      "learning_rate": 9.997414504645066e-06,
+      "loss": 1.1228,
+      "step": 576
+    },
+    {
+      "epoch": 0.040063880016664354,
+      "grad_norm": 5.596269245405725,
+      "learning_rate": 9.997378220606162e-06,
+      "loss": 0.6745,
+      "step": 577
+    },
+    {
+      "epoch": 0.040133314817386474,
+      "grad_norm": 3.63566918066838,
+      "learning_rate": 9.997341683805525e-06,
+      "loss": 0.3247,
+      "step": 578
+    },
+    {
+      "epoch": 0.040202749618108595,
+      "grad_norm": 6.543727455537456,
+      "learning_rate": 9.997304894245e-06,
+      "loss": 0.6967,
+      "step": 579
+    },
+    {
+      "epoch": 0.040272184418830716,
+      "grad_norm": 4.388741000208411,
+      "learning_rate": 9.997267851926446e-06,
+      "loss": 0.3417,
+      "step": 580
+    },
+    {
+      "epoch": 0.040341619219552836,
+      "grad_norm": 7.020275206785026,
+      "learning_rate": 9.997230556851742e-06,
+      "loss": 0.8831,
+      "step": 581
+    },
+    {
+      "epoch": 0.040411054020274964,
+      "grad_norm": 6.189141402572488,
+      "learning_rate": 9.997193009022768e-06,
+      "loss": 0.7277,
+      "step": 582
+    },
+    {
+      "epoch": 0.040480488820997085,
+      "grad_norm": 5.595771903656124,
+      "learning_rate": 9.997155208441429e-06,
+      "loss": 0.3981,
+      "step": 583
+    },
+    {
+      "epoch": 0.040549923621719206,
+      "grad_norm": 5.840701643978763,
+      "learning_rate": 9.997117155109633e-06,
+      "loss": 0.6628,
+      "step": 584
+    },
+    {
+      "epoch": 0.040619358422441326,
+      "grad_norm": 6.484513350752704,
+      "learning_rate": 9.997078849029307e-06,
+      "loss": 0.6173,
+      "step": 585
+    },
+    {
+      "epoch": 0.04068879322316345,
+      "grad_norm": 4.832760592513056,
+      "learning_rate": 9.997040290202385e-06,
+      "loss": 0.4516,
+      "step": 586
+    },
+    {
+      "epoch": 0.040758228023885575,
+      "grad_norm": 4.768738192266674,
+      "learning_rate": 9.99700147863082e-06,
+      "loss": 0.4889,
+      "step": 587
+    },
+    {
+      "epoch": 0.040827662824607695,
+      "grad_norm": 6.458325425739323,
+      "learning_rate": 9.996962414316578e-06,
+      "loss": 0.522,
+      "step": 588
+    },
+    {
+      "epoch": 0.040897097625329816,
+      "grad_norm": 5.057316576087953,
+      "learning_rate": 9.99692309726163e-06,
+      "loss": 0.8534,
+      "step": 589
+    },
+    {
+      "epoch": 0.04096653242605194,
+      "grad_norm": 4.470762887557778,
+      "learning_rate": 9.996883527467965e-06,
+      "loss": 0.4179,
+      "step": 590
+    },
+    {
+      "epoch": 0.04103596722677406,
+      "grad_norm": 6.765726662496632,
+      "learning_rate": 9.996843704937587e-06,
+      "loss": 0.5624,
+      "step": 591
+    },
+    {
+      "epoch": 0.04110540202749618,
+      "grad_norm": 4.990741793173546,
+      "learning_rate": 9.996803629672506e-06,
+      "loss": 0.5746,
+      "step": 592
+    },
+    {
+      "epoch": 0.041174836828218306,
+      "grad_norm": 5.331974763473855,
+      "learning_rate": 9.996763301674754e-06,
+      "loss": 0.7352,
+      "step": 593
+    },
+    {
+      "epoch": 0.04124427162894043,
+      "grad_norm": 4.8496698570403005,
+      "learning_rate": 9.996722720946368e-06,
+      "loss": 0.492,
+      "step": 594
+    },
+    {
+      "epoch": 0.04131370642966255,
+      "grad_norm": 5.3242084775693845,
+      "learning_rate": 9.9966818874894e-06,
+      "loss": 0.2868,
+      "step": 595
+    },
+    {
+      "epoch": 0.04138314123038467,
+      "grad_norm": 4.398737550239831,
+      "learning_rate": 9.996640801305917e-06,
+      "loss": 0.5425,
+      "step": 596
+    },
+    {
+      "epoch": 0.04145257603110679,
+      "grad_norm": 4.2244766870285355,
+      "learning_rate": 9.996599462397997e-06,
+      "loss": 0.2951,
+      "step": 597
+    },
+    {
+      "epoch": 0.04152201083182891,
+      "grad_norm": 4.3066984828806465,
+      "learning_rate": 9.99655787076773e-06,
+      "loss": 0.4191,
+      "step": 598
+    },
+    {
+      "epoch": 0.04159144563255104,
+      "grad_norm": 5.688637374956635,
+      "learning_rate": 9.99651602641722e-06,
+      "loss": 0.6695,
+      "step": 599
+    },
+    {
+      "epoch": 0.04166088043327316,
+      "grad_norm": 4.844937401261487,
+      "learning_rate": 9.996473929348583e-06,
+      "loss": 0.6308,
+      "step": 600
+    },
+    {
+      "epoch": 0.04173031523399528,
+      "grad_norm": 4.938870732358953,
+      "learning_rate": 9.996431579563947e-06,
+      "loss": 0.8417,
+      "step": 601
+    },
+    {
+      "epoch": 0.0417997500347174,
+      "grad_norm": 4.0279850075318056,
+      "learning_rate": 9.99638897706546e-06,
+      "loss": 0.2207,
+      "step": 602
+    },
+    {
+      "epoch": 0.04186918483543952,
+      "grad_norm": 6.600400974556559,
+      "learning_rate": 9.996346121855266e-06,
+      "loss": 0.7621,
+      "step": 603
+    },
+    {
+      "epoch": 0.04193861963616165,
+      "grad_norm": 6.486423792075853,
+      "learning_rate": 9.996303013935545e-06,
+      "loss": 0.2614,
+      "step": 604
+    },
+    {
+      "epoch": 0.04200805443688377,
+      "grad_norm": 4.910369192845338,
+      "learning_rate": 9.996259653308466e-06,
+      "loss": 0.5273,
+      "step": 605
+    },
+    {
+      "epoch": 0.04207748923760589,
+      "grad_norm": 5.247117900741713,
+      "learning_rate": 9.996216039976231e-06,
+      "loss": 0.6113,
+      "step": 606
+    },
+    {
+      "epoch": 0.04214692403832801,
+      "grad_norm": 5.242787371137216,
+      "learning_rate": 9.99617217394104e-06,
+      "loss": 0.7545,
+      "step": 607
+    },
+    {
+      "epoch": 0.04221635883905013,
+      "grad_norm": 7.1495405431187855,
+      "learning_rate": 9.996128055205117e-06,
+      "loss": 0.8877,
+      "step": 608
+    },
+    {
+      "epoch": 0.04228579363977225,
+      "grad_norm": 5.234100840499188,
+      "learning_rate": 9.996083683770688e-06,
+      "loss": 0.41,
+      "step": 609
+    },
+    {
+      "epoch": 0.04235522844049438,
+      "grad_norm": 5.792141595305357,
+      "learning_rate": 9.996039059640002e-06,
+      "loss": 0.6725,
+      "step": 610
+    },
+    {
+      "epoch": 0.0424246632412165,
+      "grad_norm": 5.836904810788175,
+      "learning_rate": 9.99599418281531e-06,
+      "loss": 0.8989,
+      "step": 611
+    },
+    {
+      "epoch": 0.04249409804193862,
+      "grad_norm": 5.84631547095041,
+      "learning_rate": 9.995949053298889e-06,
+      "loss": 0.8644,
+      "step": 612
+    },
+    {
+      "epoch": 0.04256353284266074,
+      "grad_norm": 5.025589687994242,
+      "learning_rate": 9.995903671093017e-06,
+      "loss": 0.3413,
+      "step": 613
+    },
+    {
+      "epoch": 0.04263296764338286,
+      "grad_norm": 5.786687074635843,
+      "learning_rate": 9.99585803619999e-06,
+      "loss": 0.7646,
+      "step": 614
+    },
+    {
+      "epoch": 0.04270240244410498,
+      "grad_norm": 6.002052351059022,
+      "learning_rate": 9.995812148622117e-06,
+      "loss": 0.732,
+      "step": 615
+    },
+    {
+      "epoch": 0.04277183724482711,
+      "grad_norm": 5.278909763533911,
+      "learning_rate": 9.99576600836172e-06,
+      "loss": 0.4222,
+      "step": 616
+    },
+    {
+      "epoch": 0.04284127204554923,
+      "grad_norm": 6.083448501144449,
+      "learning_rate": 9.995719615421129e-06,
+      "loss": 0.5847,
+      "step": 617
+    },
+    {
+      "epoch": 0.04291070684627135,
+      "grad_norm": 5.1123005644540145,
+      "learning_rate": 9.995672969802694e-06,
+      "loss": 0.6799,
+      "step": 618
+    },
+    {
+      "epoch": 0.04298014164699347,
+      "grad_norm": 7.301556957688154,
+      "learning_rate": 9.995626071508774e-06,
+      "loss": 0.8487,
+      "step": 619
+    },
+    {
+      "epoch": 0.04304957644771559,
+      "grad_norm": 6.706804976629173,
+      "learning_rate": 9.99557892054174e-06,
+      "loss": 0.9211,
+      "step": 620
+    },
+    {
+      "epoch": 0.043119011248437714,
+      "grad_norm": 5.616566740842066,
+      "learning_rate": 9.995531516903976e-06,
+      "loss": 0.5703,
+      "step": 621
+    },
+    {
+      "epoch": 0.04318844604915984,
+      "grad_norm": 5.373063444175822,
+      "learning_rate": 9.995483860597881e-06,
+      "loss": 0.8442,
+      "step": 622
+    },
+    {
+      "epoch": 0.04325788084988196,
+      "grad_norm": 4.895218416623463,
+      "learning_rate": 9.995435951625866e-06,
+      "loss": 0.2844,
+      "step": 623
+    },
+    {
+      "epoch": 0.04332731565060408,
+      "grad_norm": 6.1419272721031675,
+      "learning_rate": 9.995387789990352e-06,
+      "loss": 0.4993,
+      "step": 624
+    },
+    {
+      "epoch": 0.0433967504513262,
+      "grad_norm": 4.435039491462992,
+      "learning_rate": 9.995339375693778e-06,
+      "loss": 0.4735,
+      "step": 625
+    },
+    {
+      "epoch": 0.043466185252048324,
+      "grad_norm": 5.436198258551347,
+      "learning_rate": 9.99529070873859e-06,
+      "loss": 0.6172,
+      "step": 626
+    },
+    {
+      "epoch": 0.04353562005277045,
+      "grad_norm": 5.275137828235285,
+      "learning_rate": 9.99524178912725e-06,
+      "loss": 0.304,
+      "step": 627
+    },
+    {
+      "epoch": 0.04360505485349257,
+      "grad_norm": 5.757922178376902,
+      "learning_rate": 9.995192616862232e-06,
+      "loss": 0.5592,
+      "step": 628
+    },
+    {
+      "epoch": 0.04367448965421469,
+      "grad_norm": 5.250581508439782,
+      "learning_rate": 9.995143191946026e-06,
+      "loss": 0.5363,
+      "step": 629
+    },
+    {
+      "epoch": 0.043743924454936814,
+      "grad_norm": 6.3283396001910885,
+      "learning_rate": 9.99509351438113e-06,
+      "loss": 0.6936,
+      "step": 630
+    },
+    {
+      "epoch": 0.043813359255658935,
+      "grad_norm": 5.086353651351679,
+      "learning_rate": 9.995043584170054e-06,
+      "loss": 0.2576,
+      "step": 631
+    },
+    {
+      "epoch": 0.043882794056381055,
+      "grad_norm": 6.0505639592194775,
+      "learning_rate": 9.994993401315328e-06,
+      "loss": 0.6608,
+      "step": 632
+    },
+    {
+      "epoch": 0.04395222885710318,
+      "grad_norm": 4.266165030814394,
+      "learning_rate": 9.994942965819487e-06,
+      "loss": 0.3557,
+      "step": 633
+    },
+    {
+      "epoch": 0.044021663657825304,
+      "grad_norm": 6.289600481450322,
+      "learning_rate": 9.994892277685082e-06,
+      "loss": 0.5873,
+      "step": 634
+    },
+    {
+      "epoch": 0.044091098458547424,
+      "grad_norm": 6.085378916307771,
+      "learning_rate": 9.994841336914678e-06,
+      "loss": 0.7853,
+      "step": 635
+    },
+    {
+      "epoch": 0.044160533259269545,
+      "grad_norm": 5.5657995175392365,
+      "learning_rate": 9.994790143510853e-06,
+      "loss": 0.5294,
+      "step": 636
+    },
+    {
+      "epoch": 0.044229968059991666,
+      "grad_norm": 5.848892549029386,
+      "learning_rate": 9.994738697476194e-06,
+      "loss": 0.6143,
+      "step": 637
+    },
+    {
+      "epoch": 0.04429940286071379,
+      "grad_norm": 6.647537434047508,
+      "learning_rate": 9.994686998813302e-06,
+      "loss": 0.7232,
+      "step": 638
+    },
+    {
+      "epoch": 0.044368837661435914,
+      "grad_norm": 5.117303661423518,
+      "learning_rate": 9.994635047524794e-06,
+      "loss": 0.5613,
+      "step": 639
+    },
+    {
+      "epoch": 0.044438272462158035,
+      "grad_norm": 5.082880709591779,
+      "learning_rate": 9.994582843613299e-06,
+      "loss": 0.5221,
+      "step": 640
+    },
+    {
+      "epoch": 0.044507707262880156,
+      "grad_norm": 5.752944526950684,
+      "learning_rate": 9.994530387081454e-06,
+      "loss": 0.6203,
+      "step": 641
+    },
+    {
+      "epoch": 0.044577142063602276,
+      "grad_norm": 5.43113467137701,
+      "learning_rate": 9.994477677931912e-06,
+      "loss": 0.5625,
+      "step": 642
+    },
+    {
+      "epoch": 0.0446465768643244,
+      "grad_norm": 5.710043828784028,
+      "learning_rate": 9.994424716167344e-06,
+      "loss": 0.5739,
+      "step": 643
+    },
+    {
+      "epoch": 0.044716011665046525,
+      "grad_norm": 4.461680661965069,
+      "learning_rate": 9.994371501790423e-06,
+      "loss": 0.4762,
+      "step": 644
+    },
+    {
+      "epoch": 0.044785446465768645,
+      "grad_norm": 6.681123482451172,
+      "learning_rate": 9.994318034803845e-06,
+      "loss": 0.9257,
+      "step": 645
+    },
+    {
+      "epoch": 0.044854881266490766,
+      "grad_norm": 3.8356650513087773,
+      "learning_rate": 9.994264315210311e-06,
+      "loss": 0.3248,
+      "step": 646
+    },
+    {
+      "epoch": 0.04492431606721289,
+      "grad_norm": 4.905152621803148,
+      "learning_rate": 9.99421034301254e-06,
+      "loss": 0.4264,
+      "step": 647
+    },
+    {
+      "epoch": 0.04499375086793501,
+      "grad_norm": 4.706602343878299,
+      "learning_rate": 9.994156118213261e-06,
+      "loss": 0.4325,
+      "step": 648
+    },
+    {
+      "epoch": 0.04506318566865713,
+      "grad_norm": 4.8466194806818335,
+      "learning_rate": 9.994101640815216e-06,
+      "loss": 0.7114,
+      "step": 649
+    },
+    {
+      "epoch": 0.045132620469379256,
+      "grad_norm": 4.839562791247465,
+      "learning_rate": 9.994046910821163e-06,
+      "loss": 0.4455,
+      "step": 650
+    },
+    {
+      "epoch": 0.04520205527010138,
+      "grad_norm": 5.3547027734374035,
+      "learning_rate": 9.993991928233866e-06,
+      "loss": 0.5754,
+      "step": 651
+    },
+    {
+      "epoch": 0.0452714900708235,
+      "grad_norm": 5.930482608884689,
+      "learning_rate": 9.99393669305611e-06,
+      "loss": 0.6949,
+      "step": 652
+    },
+    {
+      "epoch": 0.04534092487154562,
+      "grad_norm": 5.946020152314283,
+      "learning_rate": 9.993881205290685e-06,
+      "loss": 0.9579,
+      "step": 653
+    },
+    {
+      "epoch": 0.04541035967226774,
+      "grad_norm": 4.130156922877328,
+      "learning_rate": 9.993825464940402e-06,
+      "loss": 0.425,
+      "step": 654
+    },
+    {
+      "epoch": 0.04547979447298986,
+      "grad_norm": 5.762412804957007,
+      "learning_rate": 9.993769472008077e-06,
+      "loss": 0.6353,
+      "step": 655
+    },
+    {
+      "epoch": 0.04554922927371199,
+      "grad_norm": 5.055458561677381,
+      "learning_rate": 9.993713226496543e-06,
+      "loss": 0.4816,
+      "step": 656
+    },
+    {
+      "epoch": 0.04561866407443411,
+      "grad_norm": 5.523202897008252,
+      "learning_rate": 9.993656728408645e-06,
+      "loss": 0.8008,
+      "step": 657
+    },
+    {
+      "epoch": 0.04568809887515623,
+      "grad_norm": 5.99476635847649,
+      "learning_rate": 9.993599977747239e-06,
+      "loss": 0.6256,
+      "step": 658
+    },
+    {
+      "epoch": 0.04575753367587835,
+      "grad_norm": 5.921879574170726,
+      "learning_rate": 9.993542974515197e-06,
+      "loss": 0.7431,
+      "step": 659
+    },
+    {
+      "epoch": 0.04582696847660047,
+      "grad_norm": 6.626727056619924,
+      "learning_rate": 9.993485718715405e-06,
+      "loss": 0.5809,
+      "step": 660
+    },
+    {
+      "epoch": 0.04589640327732259,
+      "grad_norm": 4.925078096480798,
+      "learning_rate": 9.993428210350752e-06,
+      "loss": 0.6023,
+      "step": 661
+    },
+    {
+      "epoch": 0.04596583807804472,
+      "grad_norm": 6.49043590894669,
+      "learning_rate": 9.993370449424153e-06,
+      "loss": 0.7246,
+      "step": 662
+    },
+    {
+      "epoch": 0.04603527287876684,
+      "grad_norm": 5.571292684133112,
+      "learning_rate": 9.993312435938528e-06,
+      "loss": 0.7258,
+      "step": 663
+    },
+    {
+      "epoch": 0.04610470767948896,
+      "grad_norm": 5.895604140670428,
+      "learning_rate": 9.993254169896807e-06,
+      "loss": 0.7167,
+      "step": 664
+    },
+    {
+      "epoch": 0.04617414248021108,
+      "grad_norm": 4.26586213385751,
+      "learning_rate": 9.993195651301942e-06,
+      "loss": 0.3716,
+      "step": 665
+    },
+    {
+      "epoch": 0.0462435772809332,
+      "grad_norm": 4.53897670668393,
+      "learning_rate": 9.99313688015689e-06,
+      "loss": 0.4756,
+      "step": 666
+    },
+    {
+      "epoch": 0.04631301208165533,
+      "grad_norm": 3.294704919455413,
+      "learning_rate": 9.993077856464627e-06,
+      "loss": 0.2139,
+      "step": 667
+    },
+    {
+      "epoch": 0.04638244688237745,
+      "grad_norm": 5.6353877185186425,
+      "learning_rate": 9.993018580228135e-06,
+      "loss": 0.6015,
+      "step": 668
+    },
+    {
+      "epoch": 0.04645188168309957,
+      "grad_norm": 5.134723049069843,
+      "learning_rate": 9.992959051450413e-06,
+      "loss": 0.7059,
+      "step": 669
+    },
+    {
+      "epoch": 0.04652131648382169,
+      "grad_norm": 4.430640977255875,
+      "learning_rate": 9.992899270134473e-06,
+      "loss": 0.347,
+      "step": 670
+    },
+    {
+      "epoch": 0.04659075128454381,
+      "grad_norm": 5.709162874616909,
+      "learning_rate": 9.992839236283336e-06,
+      "loss": 0.7561,
+      "step": 671
+    },
+    {
+      "epoch": 0.04666018608526593,
+      "grad_norm": 4.589483738764916,
+      "learning_rate": 9.992778949900042e-06,
+      "loss": 0.3547,
+      "step": 672
+    },
+    {
+      "epoch": 0.04672962088598806,
+      "grad_norm": 6.6286852586880505,
+      "learning_rate": 9.992718410987636e-06,
+      "loss": 0.7243,
+      "step": 673
+    },
+    {
+      "epoch": 0.04679905568671018,
+      "grad_norm": 5.472732735300226,
+      "learning_rate": 9.992657619549184e-06,
+      "loss": 0.6439,
+      "step": 674
+    },
+    {
+      "epoch": 0.0468684904874323,
+      "grad_norm": 5.186320391604662,
+      "learning_rate": 9.99259657558776e-06,
+      "loss": 0.6176,
+      "step": 675
+    },
+    {
+      "epoch": 0.04693792528815442,
+      "grad_norm": 6.525902983235542,
+      "learning_rate": 9.99253527910645e-06,
+      "loss": 0.5332,
+      "step": 676
+    },
+    {
+      "epoch": 0.04700736008887654,
+      "grad_norm": 5.230737744483306,
+      "learning_rate": 9.992473730108354e-06,
+      "loss": 0.489,
+      "step": 677
+    },
+    {
+      "epoch": 0.047076794889598664,
+      "grad_norm": 4.095183710870003,
+      "learning_rate": 9.992411928596588e-06,
+      "loss": 0.3977,
+      "step": 678
+    },
+    {
+      "epoch": 0.04714622969032079,
+      "grad_norm": 6.148857555475747,
+      "learning_rate": 9.992349874574274e-06,
+      "loss": 0.5967,
+      "step": 679
+    },
+    {
+      "epoch": 0.04721566449104291,
+      "grad_norm": 6.046034821291769,
+      "learning_rate": 9.992287568044554e-06,
+      "loss": 0.9083,
+      "step": 680
+    },
+    {
+      "epoch": 0.04728509929176503,
+      "grad_norm": 5.501001092681591,
+      "learning_rate": 9.992225009010576e-06,
+      "loss": 0.6978,
+      "step": 681
+    },
+    {
+      "epoch": 0.047354534092487154,
+      "grad_norm": 4.773098237371324,
+      "learning_rate": 9.99216219747551e-06,
+      "loss": 0.4296,
+      "step": 682
+    },
+    {
+      "epoch": 0.047423968893209274,
+      "grad_norm": 4.7886233652203245,
+      "learning_rate": 9.992099133442525e-06,
+      "loss": 0.3198,
+      "step": 683
+    },
+    {
+      "epoch": 0.047493403693931395,
+      "grad_norm": 3.7095350557337192,
+      "learning_rate": 9.992035816914816e-06,
+      "loss": 0.3297,
+      "step": 684
+    },
+    {
+      "epoch": 0.04756283849465352,
+      "grad_norm": 5.671251028019554,
+      "learning_rate": 9.991972247895585e-06,
+      "loss": 0.5346,
+      "step": 685
+    },
+    {
+      "epoch": 0.04763227329537564,
+      "grad_norm": 4.691379188955125,
+      "learning_rate": 9.991908426388046e-06,
+      "loss": 0.7333,
+      "step": 686
+    },
+    {
+      "epoch": 0.047701708096097764,
+      "grad_norm": 4.650332561327247,
+      "learning_rate": 9.991844352395429e-06,
+      "loss": 0.4452,
+      "step": 687
+    },
+    {
+      "epoch": 0.047771142896819885,
+      "grad_norm": 5.699856464232138,
+      "learning_rate": 9.991780025920972e-06,
+      "loss": 0.5472,
+      "step": 688
+    },
+    {
+      "epoch": 0.047840577697542006,
+      "grad_norm": 6.2112708294475665,
+      "learning_rate": 9.991715446967928e-06,
+      "loss": 0.6031,
+      "step": 689
+    },
+    {
+      "epoch": 0.04791001249826413,
+      "grad_norm": 4.909061303795624,
+      "learning_rate": 9.991650615539567e-06,
+      "loss": 0.425,
+      "step": 690
+    },
+    {
+      "epoch": 0.047979447298986254,
+      "grad_norm": 6.21423437032649,
+      "learning_rate": 9.991585531639167e-06,
+      "loss": 0.7406,
+      "step": 691
+    },
+    {
+      "epoch": 0.048048882099708375,
+      "grad_norm": 4.382090372038084,
+      "learning_rate": 9.99152019527002e-06,
+      "loss": 0.3912,
+      "step": 692
+    },
+    {
+      "epoch": 0.048118316900430495,
+      "grad_norm": 5.6078304679304365,
+      "learning_rate": 9.991454606435427e-06,
+      "loss": 0.5902,
+      "step": 693
+    },
+    {
+      "epoch": 0.048187751701152616,
+      "grad_norm": 5.853339406196495,
+      "learning_rate": 9.99138876513871e-06,
+      "loss": 0.5891,
+      "step": 694
+    },
+    {
+      "epoch": 0.04825718650187474,
+      "grad_norm": 5.8455438851216135,
+      "learning_rate": 9.991322671383197e-06,
+      "loss": 0.6937,
+      "step": 695
+    },
+    {
+      "epoch": 0.048326621302596864,
+      "grad_norm": 6.649374833206646,
+      "learning_rate": 9.991256325172232e-06,
+      "loss": 0.8635,
+      "step": 696
+    },
+    {
+      "epoch": 0.048396056103318985,
+      "grad_norm": 6.012925621088737,
+      "learning_rate": 9.99118972650917e-06,
+      "loss": 1.0924,
+      "step": 697
+    },
+    {
+      "epoch": 0.048465490904041106,
+      "grad_norm": 4.977710728227318,
+      "learning_rate": 9.991122875397377e-06,
+      "loss": 0.3265,
+      "step": 698
+    },
+    {
+      "epoch": 0.04853492570476323,
+      "grad_norm": 7.086722062528404,
+      "learning_rate": 9.991055771840239e-06,
+      "loss": 0.6656,
+      "step": 699
+    },
+    {
+      "epoch": 0.04860436050548535,
+      "grad_norm": 6.182449126065321,
+      "learning_rate": 9.990988415841147e-06,
+      "loss": 0.7988,
+      "step": 700
+    },
+    {
+      "epoch": 0.04867379530620747,
+      "grad_norm": 4.305440514230685,
+      "learning_rate": 9.99092080740351e-06,
+      "loss": 0.3211,
+      "step": 701
+    },
+    {
+      "epoch": 0.048743230106929596,
+      "grad_norm": 6.50311447978165,
+      "learning_rate": 9.990852946530746e-06,
+      "loss": 0.789,
+      "step": 702
+    },
+    {
+      "epoch": 0.048812664907651716,
+      "grad_norm": 4.73488664658267,
+      "learning_rate": 9.990784833226285e-06,
+      "loss": 0.6239,
+      "step": 703
+    },
+    {
+      "epoch": 0.04888209970837384,
+      "grad_norm": 5.548278431037929,
+      "learning_rate": 9.990716467493577e-06,
+      "loss": 0.6327,
+      "step": 704
+    },
+    {
+      "epoch": 0.04895153450909596,
+      "grad_norm": 5.744145323311782,
+      "learning_rate": 9.990647849336076e-06,
+      "loss": 0.7761,
+      "step": 705
+    },
+    {
+      "epoch": 0.04902096930981808,
+      "grad_norm": 4.502047106110494,
+      "learning_rate": 9.990578978757257e-06,
+      "loss": 0.5615,
+      "step": 706
+    },
+    {
+      "epoch": 0.049090404110540206,
+      "grad_norm": 4.924566567066488,
+      "learning_rate": 9.990509855760598e-06,
+      "loss": 0.4088,
+      "step": 707
+    },
+    {
+      "epoch": 0.04915983891126233,
+      "grad_norm": 4.124872574793904,
+      "learning_rate": 9.990440480349599e-06,
+      "loss": 0.3375,
+      "step": 708
+    },
+    {
+      "epoch": 0.04922927371198445,
+      "grad_norm": 6.252079440192097,
+      "learning_rate": 9.990370852527767e-06,
+      "loss": 0.6336,
+      "step": 709
+    },
+    {
+      "epoch": 0.04929870851270657,
+      "grad_norm": 5.117180459969774,
+      "learning_rate": 9.990300972298624e-06,
+      "loss": 0.5806,
+      "step": 710
+    },
+    {
+      "epoch": 0.04936814331342869,
+      "grad_norm": 5.049624789535316,
+      "learning_rate": 9.990230839665704e-06,
+      "loss": 0.541,
+      "step": 711
+    },
+    {
+      "epoch": 0.04943757811415081,
+      "grad_norm": 5.229084678687425,
+      "learning_rate": 9.990160454632554e-06,
+      "loss": 0.5403,
+      "step": 712
+    },
+    {
+      "epoch": 0.04950701291487294,
+      "grad_norm": 5.490394347351558,
+      "learning_rate": 9.990089817202737e-06,
+      "loss": 0.5434,
+      "step": 713
+    },
+    {
+      "epoch": 0.04957644771559506,
+      "grad_norm": 4.8100647031152315,
+      "learning_rate": 9.990018927379823e-06,
+      "loss": 0.4469,
+      "step": 714
+    },
+    {
+      "epoch": 0.04964588251631718,
+      "grad_norm": 5.473783736933605,
+      "learning_rate": 9.9899477851674e-06,
+      "loss": 0.618,
+      "step": 715
+    },
+    {
+      "epoch": 0.0497153173170393,
+      "grad_norm": 5.249865227981726,
+      "learning_rate": 9.989876390569062e-06,
+      "loss": 0.5261,
+      "step": 716
+    },
+    {
+      "epoch": 0.04978475211776142,
+      "grad_norm": 4.708877748911251,
+      "learning_rate": 9.989804743588423e-06,
+      "loss": 0.624,
+      "step": 717
+    },
+    {
+      "epoch": 0.04985418691848354,
+      "grad_norm": 6.115422646427735,
+      "learning_rate": 9.989732844229107e-06,
+      "loss": 0.7497,
+      "step": 718
+    },
+    {
+      "epoch": 0.04992362171920567,
+      "grad_norm": 4.260508105711181,
+      "learning_rate": 9.989660692494751e-06,
+      "loss": 0.4474,
+      "step": 719
+    },
+    {
+      "epoch": 0.04999305651992779,
+      "grad_norm": 7.51725860076484,
+      "learning_rate": 9.989588288389003e-06,
+      "loss": 0.6335,
+      "step": 720
+    },
+    {
+      "epoch": 0.05006249132064991,
+      "grad_norm": 5.783938157840532,
+      "learning_rate": 9.989515631915524e-06,
+      "loss": 0.7792,
+      "step": 721
+    },
+    {
+      "epoch": 0.05013192612137203,
+      "grad_norm": 4.364848401501063,
+      "learning_rate": 9.989442723077991e-06,
+      "loss": 0.5082,
+      "step": 722
+    },
+    {
+      "epoch": 0.05020136092209415,
+      "grad_norm": 3.9802923239712875,
+      "learning_rate": 9.989369561880091e-06,
+      "loss": 0.2611,
+      "step": 723
+    },
+    {
+      "epoch": 0.05027079572281627,
+      "grad_norm": 5.755484380666307,
+      "learning_rate": 9.989296148325525e-06,
+      "loss": 0.5454,
+      "step": 724
+    },
+    {
+      "epoch": 0.0503402305235384,
+      "grad_norm": 3.5439999881225317,
+      "learning_rate": 9.989222482418006e-06,
+      "loss": 0.3573,
+      "step": 725
+    },
+    {
+      "epoch": 0.05040966532426052,
+      "grad_norm": 6.618609773149128,
+      "learning_rate": 9.989148564161258e-06,
+      "loss": 0.911,
+      "step": 726
+    },
+    {
+      "epoch": 0.05047910012498264,
+      "grad_norm": 5.069257498600404,
+      "learning_rate": 9.989074393559022e-06,
+      "loss": 0.613,
+      "step": 727
+    },
+    {
+      "epoch": 0.05054853492570476,
+      "grad_norm": 5.864054610037569,
+      "learning_rate": 9.98899997061505e-06,
+      "loss": 0.8184,
+      "step": 728
+    },
+    {
+      "epoch": 0.05061796972642688,
+      "grad_norm": 4.517963581897608,
+      "learning_rate": 9.988925295333103e-06,
+      "loss": 0.5548,
+      "step": 729
+    },
+    {
+      "epoch": 0.05068740452714901,
+      "grad_norm": 4.618753160369594,
+      "learning_rate": 9.98885036771696e-06,
+      "loss": 0.5531,
+      "step": 730
+    },
+    {
+      "epoch": 0.05075683932787113,
+      "grad_norm": 4.376843038128889,
+      "learning_rate": 9.988775187770413e-06,
+      "loss": 0.4608,
+      "step": 731
+    },
+    {
+      "epoch": 0.05082627412859325,
+      "grad_norm": 4.866481954747598,
+      "learning_rate": 9.988699755497258e-06,
+      "loss": 0.3514,
+      "step": 732
+    },
+    {
+      "epoch": 0.05089570892931537,
+      "grad_norm": 4.1040280996416865,
+      "learning_rate": 9.988624070901318e-06,
+      "loss": 0.6799,
+      "step": 733
+    },
+    {
+      "epoch": 0.05096514373003749,
+      "grad_norm": 4.682159071004771,
+      "learning_rate": 9.988548133986415e-06,
+      "loss": 0.5379,
+      "step": 734
+    },
+    {
+      "epoch": 0.051034578530759614,
+      "grad_norm": 4.966606443743591,
+      "learning_rate": 9.988471944756395e-06,
+      "loss": 0.4636,
+      "step": 735
+    },
+    {
+      "epoch": 0.05110401333148174,
+      "grad_norm": 4.711458416668277,
+      "learning_rate": 9.988395503215107e-06,
+      "loss": 0.3936,
+      "step": 736
+    },
+    {
+      "epoch": 0.05117344813220386,
+      "grad_norm": 4.709532737592508,
+      "learning_rate": 9.98831880936642e-06,
+      "loss": 0.4481,
+      "step": 737
+    },
+    {
+      "epoch": 0.05124288293292598,
+      "grad_norm": 5.057946309627939,
+      "learning_rate": 9.988241863214212e-06,
+      "loss": 0.4989,
+      "step": 738
+    },
+    {
+      "epoch": 0.051312317733648104,
+      "grad_norm": 6.029719457481402,
+      "learning_rate": 9.988164664762374e-06,
+      "loss": 0.5903,
+      "step": 739
+    },
+    {
+      "epoch": 0.051381752534370224,
+      "grad_norm": 4.348655708656128,
+      "learning_rate": 9.988087214014815e-06,
+      "loss": 0.5438,
+      "step": 740
+    },
+    {
+      "epoch": 0.051451187335092345,
+      "grad_norm": 6.263846152222153,
+      "learning_rate": 9.988009510975446e-06,
+      "loss": 0.608,
+      "step": 741
+    },
+    {
+      "epoch": 0.05152062213581447,
+      "grad_norm": 5.341494767010231,
+      "learning_rate": 9.9879315556482e-06,
+      "loss": 0.6357,
+      "step": 742
+    },
+    {
+      "epoch": 0.051590056936536594,
+      "grad_norm": 4.882202875476889,
+      "learning_rate": 9.987853348037022e-06,
+      "loss": 0.3735,
+      "step": 743
+    },
+    {
+      "epoch": 0.051659491737258714,
+      "grad_norm": 4.492912656255205,
+      "learning_rate": 9.987774888145867e-06,
+      "loss": 0.4875,
+      "step": 744
+    },
+    {
+      "epoch": 0.051728926537980835,
+      "grad_norm": 3.773069207931385,
+      "learning_rate": 9.987696175978698e-06,
+      "loss": 0.3732,
+      "step": 745
+    },
+    {
+      "epoch": 0.051798361338702956,
+      "grad_norm": 4.712267786494523,
+      "learning_rate": 9.987617211539505e-06,
+      "loss": 0.4418,
+      "step": 746
+    },
+    {
+      "epoch": 0.051867796139425076,
+      "grad_norm": 5.502889018503653,
+      "learning_rate": 9.987537994832275e-06,
+      "loss": 0.5774,
+      "step": 747
+    },
+    {
+      "epoch": 0.051937230940147204,
+      "grad_norm": 4.20299459003689,
+      "learning_rate": 9.987458525861016e-06,
+      "loss": 0.4716,
+      "step": 748
+    },
+    {
+      "epoch": 0.052006665740869325,
+      "grad_norm": 6.591786713702156,
+      "learning_rate": 9.98737880462975e-06,
+      "loss": 0.6916,
+      "step": 749
+    },
+    {
+      "epoch": 0.052076100541591445,
+      "grad_norm": 5.425698032187315,
+      "learning_rate": 9.987298831142507e-06,
+      "loss": 0.6178,
+      "step": 750
+    },
+    {
+      "epoch": 0.052145535342313566,
+      "grad_norm": 5.878407309301575,
+      "learning_rate": 9.987218605403332e-06,
+      "loss": 0.5453,
+      "step": 751
+    },
+    {
+      "epoch": 0.05221497014303569,
+      "grad_norm": 5.08698484079951,
+      "learning_rate": 9.987138127416285e-06,
+      "loss": 0.5782,
+      "step": 752
+    },
+    {
+      "epoch": 0.052284404943757815,
+      "grad_norm": 5.879980805190245,
+      "learning_rate": 9.987057397185433e-06,
+      "loss": 0.9409,
+      "step": 753
+    },
+    {
+      "epoch": 0.052353839744479935,
+      "grad_norm": 7.0619768626850155,
+      "learning_rate": 9.986976414714863e-06,
+      "loss": 0.6808,
+      "step": 754
+    },
+    {
+      "epoch": 0.052423274545202056,
+      "grad_norm": 4.113634903601791,
+      "learning_rate": 9.98689518000867e-06,
+      "loss": 0.5029,
+      "step": 755
+    },
+    {
+      "epoch": 0.05249270934592418,
+      "grad_norm": 6.365846519761652,
+      "learning_rate": 9.986813693070958e-06,
+      "loss": 0.8543,
+      "step": 756
+    },
+    {
+      "epoch": 0.0525621441466463,
+      "grad_norm": 5.337016275372264,
+      "learning_rate": 9.986731953905854e-06,
+      "loss": 0.998,
+      "step": 757
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 4.482510077691802,
+      "learning_rate": 9.986649962517491e-06,
+      "loss": 0.4713,
+      "step": 758
+    },
+    {
+      "epoch": 0.052701013748090546,
+      "grad_norm": 5.99016367767901,
+      "learning_rate": 9.986567718910017e-06,
+      "loss": 0.8063,
+      "step": 759
+    },
+    {
+      "epoch": 0.052770448548812667,
+      "grad_norm": 4.790256336151655,
+      "learning_rate": 9.98648522308759e-06,
+      "loss": 0.4555,
+      "step": 760
+    },
+    {
+      "epoch": 0.05283988334953479,
+      "grad_norm": 4.081396326845866,
+      "learning_rate": 9.986402475054382e-06,
+      "loss": 0.3434,
+      "step": 761
+    },
+    {
+      "epoch": 0.05290931815025691,
+      "grad_norm": 4.731669740723197,
+      "learning_rate": 9.98631947481458e-06,
+      "loss": 0.7651,
+      "step": 762
+    },
+    {
+      "epoch": 0.05297875295097903,
+      "grad_norm": 4.94347253864801,
+      "learning_rate": 9.986236222372379e-06,
+      "loss": 0.4651,
+      "step": 763
+    },
+    {
+      "epoch": 0.05304818775170115,
+      "grad_norm": 5.960552828652827,
+      "learning_rate": 9.986152717731994e-06,
+      "loss": 0.7553,
+      "step": 764
+    },
+    {
+      "epoch": 0.05311762255242328,
+      "grad_norm": 4.693546729412604,
+      "learning_rate": 9.986068960897648e-06,
+      "loss": 0.5428,
+      "step": 765
+    },
+    {
+      "epoch": 0.0531870573531454,
+      "grad_norm": 4.82511924849523,
+      "learning_rate": 9.985984951873575e-06,
+      "loss": 0.4141,
+      "step": 766
+    },
+    {
+      "epoch": 0.05325649215386752,
+      "grad_norm": 4.8448179849610105,
+      "learning_rate": 9.985900690664025e-06,
+      "loss": 0.5061,
+      "step": 767
+    },
+    {
+      "epoch": 0.05332592695458964,
+      "grad_norm": 4.430891864955989,
+      "learning_rate": 9.98581617727326e-06,
+      "loss": 0.346,
+      "step": 768
+    },
+    {
+      "epoch": 0.05339536175531176,
+      "grad_norm": 5.11006363764043,
+      "learning_rate": 9.985731411705554e-06,
+      "loss": 0.7828,
+      "step": 769
+    },
+    {
+      "epoch": 0.05346479655603389,
+      "grad_norm": 4.718222616321539,
+      "learning_rate": 9.985646393965196e-06,
+      "loss": 0.4868,
+      "step": 770
+    },
+    {
+      "epoch": 0.05353423135675601,
+      "grad_norm": 5.691807304395084,
+      "learning_rate": 9.985561124056484e-06,
+      "loss": 0.7973,
+      "step": 771
+    },
+    {
+      "epoch": 0.05360366615747813,
+      "grad_norm": 4.392238674725026,
+      "learning_rate": 9.985475601983733e-06,
+      "loss": 0.3112,
+      "step": 772
+    },
+    {
+      "epoch": 0.05367310095820025,
+      "grad_norm": 5.799717920285053,
+      "learning_rate": 9.985389827751266e-06,
+      "loss": 0.6339,
+      "step": 773
+    },
+    {
+      "epoch": 0.05374253575892237,
+      "grad_norm": 6.3274326905982035,
+      "learning_rate": 9.985303801363423e-06,
+      "loss": 0.8679,
+      "step": 774
+    },
+    {
+      "epoch": 0.05381197055964449,
+      "grad_norm": 4.955774779022077,
+      "learning_rate": 9.985217522824558e-06,
+      "loss": 0.5535,
+      "step": 775
+    },
+    {
+      "epoch": 0.05388140536036662,
+      "grad_norm": 5.371640328381147,
+      "learning_rate": 9.98513099213903e-06,
+      "loss": 0.511,
+      "step": 776
+    },
+    {
+      "epoch": 0.05395084016108874,
+      "grad_norm": 4.93937416310012,
+      "learning_rate": 9.985044209311217e-06,
+      "loss": 0.4582,
+      "step": 777
+    },
+    {
+      "epoch": 0.05402027496181086,
+      "grad_norm": 5.832216246395709,
+      "learning_rate": 9.984957174345509e-06,
+      "loss": 0.7087,
+      "step": 778
+    },
+    {
+      "epoch": 0.05408970976253298,
+      "grad_norm": 4.859658162849972,
+      "learning_rate": 9.984869887246307e-06,
+      "loss": 0.5068,
+      "step": 779
+    },
+    {
+      "epoch": 0.0541591445632551,
+      "grad_norm": 5.465886543799718,
+      "learning_rate": 9.984782348018027e-06,
+      "loss": 0.6644,
+      "step": 780
+    },
+    {
+      "epoch": 0.05422857936397722,
+      "grad_norm": 4.912474934027746,
+      "learning_rate": 9.984694556665098e-06,
+      "loss": 0.5801,
+      "step": 781
+    },
+    {
+      "epoch": 0.05429801416469935,
+      "grad_norm": 4.8307962171364585,
+      "learning_rate": 9.984606513191959e-06,
+      "loss": 0.4253,
+      "step": 782
+    },
+    {
+      "epoch": 0.05436744896542147,
+      "grad_norm": 3.78195117183482,
+      "learning_rate": 9.984518217603062e-06,
+      "loss": 0.4946,
+      "step": 783
+    },
+    {
+      "epoch": 0.05443688376614359,
+      "grad_norm": 6.1860148087415485,
+      "learning_rate": 9.984429669902874e-06,
+      "loss": 0.753,
+      "step": 784
+    },
+    {
+      "epoch": 0.05450631856686571,
+      "grad_norm": 5.452248410668079,
+      "learning_rate": 9.984340870095874e-06,
+      "loss": 0.8678,
+      "step": 785
+    },
+    {
+      "epoch": 0.05457575336758783,
+      "grad_norm": 5.377308015234931,
+      "learning_rate": 9.984251818186552e-06,
+      "loss": 0.6031,
+      "step": 786
+    },
+    {
+      "epoch": 0.054645188168309954,
+      "grad_norm": 5.170839150857725,
+      "learning_rate": 9.984162514179414e-06,
+      "loss": 0.4633,
+      "step": 787
+    },
+    {
+      "epoch": 0.05471462296903208,
+      "grad_norm": 5.2700664222346285,
+      "learning_rate": 9.984072958078977e-06,
+      "loss": 0.6024,
+      "step": 788
+    },
+    {
+      "epoch": 0.0547840577697542,
+      "grad_norm": 5.277357561535729,
+      "learning_rate": 9.983983149889768e-06,
+      "loss": 0.516,
+      "step": 789
+    },
+    {
+      "epoch": 0.05485349257047632,
+      "grad_norm": 4.943264751845346,
+      "learning_rate": 9.983893089616332e-06,
+      "loss": 0.3824,
+      "step": 790
+    },
+    {
+      "epoch": 0.05492292737119844,
+      "grad_norm": 4.460423425633663,
+      "learning_rate": 9.983802777263223e-06,
+      "loss": 0.4777,
+      "step": 791
+    },
+    {
+      "epoch": 0.054992362171920564,
+      "grad_norm": 5.333019121676857,
+      "learning_rate": 9.983712212835009e-06,
+      "loss": 0.6833,
+      "step": 792
+    },
+    {
+      "epoch": 0.05506179697264269,
+      "grad_norm": 4.715970705130396,
+      "learning_rate": 9.98362139633627e-06,
+      "loss": 0.4223,
+      "step": 793
+    },
+    {
+      "epoch": 0.05513123177336481,
+      "grad_norm": 6.100854587077938,
+      "learning_rate": 9.9835303277716e-06,
+      "loss": 0.8033,
+      "step": 794
+    },
+    {
+      "epoch": 0.05520066657408693,
+      "grad_norm": 5.214988274486931,
+      "learning_rate": 9.983439007145606e-06,
+      "loss": 0.512,
+      "step": 795
+    },
+    {
+      "epoch": 0.055270101374809054,
+      "grad_norm": 5.454061849419084,
+      "learning_rate": 9.983347434462906e-06,
+      "loss": 0.8152,
+      "step": 796
+    },
+    {
+      "epoch": 0.055339536175531175,
+      "grad_norm": 4.889922018041808,
+      "learning_rate": 9.98325560972813e-06,
+      "loss": 0.5321,
+      "step": 797
+    },
+    {
+      "epoch": 0.055408970976253295,
+      "grad_norm": 5.953590913340075,
+      "learning_rate": 9.983163532945927e-06,
+      "loss": 0.4376,
+      "step": 798
+    },
+    {
+      "epoch": 0.05547840577697542,
+      "grad_norm": 5.366688377270275,
+      "learning_rate": 9.98307120412095e-06,
+      "loss": 0.7865,
+      "step": 799
+    },
+    {
+      "epoch": 0.055547840577697544,
+      "grad_norm": 4.87553821918744,
+      "learning_rate": 9.98297862325787e-06,
+      "loss": 0.6314,
+      "step": 800
+    },
+    {
+      "epoch": 0.055617275378419664,
+      "grad_norm": 4.9894506888616315,
+      "learning_rate": 9.982885790361372e-06,
+      "loss": 0.6921,
+      "step": 801
+    },
+    {
+      "epoch": 0.055686710179141785,
+      "grad_norm": 4.753296030077526,
+      "learning_rate": 9.982792705436147e-06,
+      "loss": 0.4623,
+      "step": 802
+    },
+    {
+      "epoch": 0.055756144979863906,
+      "grad_norm": 4.68018376892769,
+      "learning_rate": 9.982699368486907e-06,
+      "loss": 0.313,
+      "step": 803
+    },
+    {
+      "epoch": 0.05582557978058603,
+      "grad_norm": 4.894882821522259,
+      "learning_rate": 9.982605779518369e-06,
+      "loss": 0.3951,
+      "step": 804
+    },
+    {
+      "epoch": 0.055895014581308154,
+      "grad_norm": 7.351167666092212,
+      "learning_rate": 9.98251193853527e-06,
+      "loss": 0.6325,
+      "step": 805
+    },
+    {
+      "epoch": 0.055964449382030275,
+      "grad_norm": 3.850195782085043,
+      "learning_rate": 9.982417845542357e-06,
+      "loss": 0.4287,
+      "step": 806
+    },
+    {
+      "epoch": 0.056033884182752396,
+      "grad_norm": 4.818134158620349,
+      "learning_rate": 9.982323500544385e-06,
+      "loss": 0.4672,
+      "step": 807
+    },
+    {
+      "epoch": 0.056103318983474516,
+      "grad_norm": 6.533269444355503,
+      "learning_rate": 9.982228903546129e-06,
+      "loss": 0.9613,
+      "step": 808
+    },
+    {
+      "epoch": 0.05617275378419664,
+      "grad_norm": 5.0910109719802135,
+      "learning_rate": 9.982134054552373e-06,
+      "loss": 0.3717,
+      "step": 809
+    },
+    {
+      "epoch": 0.056242188584918765,
+      "grad_norm": 3.6367939709812354,
+      "learning_rate": 9.982038953567914e-06,
+      "loss": 0.4374,
+      "step": 810
+    },
+    {
+      "epoch": 0.056311623385640885,
+      "grad_norm": 4.19897794140048,
+      "learning_rate": 9.981943600597562e-06,
+      "loss": 0.4336,
+      "step": 811
+    },
+    {
+      "epoch": 0.056381058186363006,
+      "grad_norm": 5.771272428000438,
+      "learning_rate": 9.981847995646142e-06,
+      "loss": 0.6326,
+      "step": 812
+    },
+    {
+      "epoch": 0.05645049298708513,
+      "grad_norm": 5.559806523706431,
+      "learning_rate": 9.981752138718484e-06,
+      "loss": 0.616,
+      "step": 813
+    },
+    {
+      "epoch": 0.05651992778780725,
+      "grad_norm": 4.6925742039467675,
+      "learning_rate": 9.981656029819444e-06,
+      "loss": 0.5655,
+      "step": 814
+    },
+    {
+      "epoch": 0.05658936258852937,
+      "grad_norm": 5.543281182192348,
+      "learning_rate": 9.981559668953876e-06,
+      "loss": 0.8608,
+      "step": 815
+    },
+    {
+      "epoch": 0.056658797389251496,
+      "grad_norm": 5.447741188044963,
+      "learning_rate": 9.98146305612666e-06,
+      "loss": 0.524,
+      "step": 816
+    },
+    {
+      "epoch": 0.05672823218997362,
+      "grad_norm": 4.758266373977329,
+      "learning_rate": 9.981366191342677e-06,
+      "loss": 0.5684,
+      "step": 817
+    },
+    {
+      "epoch": 0.05679766699069574,
+      "grad_norm": 4.362557076905672,
+      "learning_rate": 9.98126907460683e-06,
+      "loss": 0.479,
+      "step": 818
+    },
+    {
+      "epoch": 0.05686710179141786,
+      "grad_norm": 4.414720533516325,
+      "learning_rate": 9.981171705924029e-06,
+      "loss": 0.312,
+      "step": 819
+    },
+    {
+      "epoch": 0.05693653659213998,
+      "grad_norm": 4.599478915544888,
+      "learning_rate": 9.9810740852992e-06,
+      "loss": 0.6729,
+      "step": 820
+    },
+    {
+      "epoch": 0.0570059713928621,
+      "grad_norm": 4.51681507491556,
+      "learning_rate": 9.98097621273728e-06,
+      "loss": 0.5332,
+      "step": 821
+    },
+    {
+      "epoch": 0.05707540619358423,
+      "grad_norm": 4.458634891794412,
+      "learning_rate": 9.98087808824322e-06,
+      "loss": 0.3974,
+      "step": 822
+    },
+    {
+      "epoch": 0.05714484099430635,
+      "grad_norm": 5.679804026500894,
+      "learning_rate": 9.980779711821983e-06,
+      "loss": 0.5288,
+      "step": 823
+    },
+    {
+      "epoch": 0.05721427579502847,
+      "grad_norm": 4.5083719310795685,
+      "learning_rate": 9.980681083478543e-06,
+      "loss": 0.4033,
+      "step": 824
+    },
+    {
+      "epoch": 0.05728371059575059,
+      "grad_norm": 6.986410721750218,
+      "learning_rate": 9.980582203217893e-06,
+      "loss": 0.7277,
+      "step": 825
+    },
+    {
+      "epoch": 0.05735314539647271,
+      "grad_norm": 4.915844057565651,
+      "learning_rate": 9.980483071045028e-06,
+      "loss": 0.7306,
+      "step": 826
+    },
+    {
+      "epoch": 0.05742258019719483,
+      "grad_norm": 6.065689651606107,
+      "learning_rate": 9.980383686964966e-06,
+      "loss": 0.8851,
+      "step": 827
+    },
+    {
+      "epoch": 0.05749201499791696,
+      "grad_norm": 5.903824276691955,
+      "learning_rate": 9.980284050982731e-06,
+      "loss": 0.7023,
+      "step": 828
+    },
+    {
+      "epoch": 0.05756144979863908,
+      "grad_norm": 5.782728474541114,
+      "learning_rate": 9.980184163103368e-06,
+      "loss": 0.636,
+      "step": 829
+    },
+    {
+      "epoch": 0.0576308845993612,
+      "grad_norm": 5.023634802144964,
+      "learning_rate": 9.980084023331921e-06,
+      "loss": 0.5108,
+      "step": 830
+    },
+    {
+      "epoch": 0.05770031940008332,
+      "grad_norm": 4.44219042772744,
+      "learning_rate": 9.979983631673463e-06,
+      "loss": 0.3715,
+      "step": 831
+    },
+    {
+      "epoch": 0.05776975420080544,
+      "grad_norm": 4.087772742765401,
+      "learning_rate": 9.979882988133068e-06,
+      "loss": 0.2826,
+      "step": 832
+    },
+    {
+      "epoch": 0.05783918900152757,
+      "grad_norm": 5.251898639356191,
+      "learning_rate": 9.979782092715825e-06,
+      "loss": 0.4231,
+      "step": 833
+    },
+    {
+      "epoch": 0.05790862380224969,
+      "grad_norm": 4.9695246622775,
+      "learning_rate": 9.97968094542684e-06,
+      "loss": 0.5981,
+      "step": 834
+    },
+    {
+      "epoch": 0.05797805860297181,
+      "grad_norm": 5.834848117543546,
+      "learning_rate": 9.979579546271225e-06,
+      "loss": 0.7907,
+      "step": 835
+    },
+    {
+      "epoch": 0.05804749340369393,
+      "grad_norm": 5.454160042170218,
+      "learning_rate": 9.979477895254113e-06,
+      "loss": 0.821,
+      "step": 836
+    },
+    {
+      "epoch": 0.05811692820441605,
+      "grad_norm": 4.46656251099197,
+      "learning_rate": 9.979375992380645e-06,
+      "loss": 0.3572,
+      "step": 837
+    },
+    {
+      "epoch": 0.05818636300513817,
+      "grad_norm": 4.993120846382734,
+      "learning_rate": 9.979273837655973e-06,
+      "loss": 0.6124,
+      "step": 838
+    },
+    {
+      "epoch": 0.0582557978058603,
+      "grad_norm": 3.911172763154037,
+      "learning_rate": 9.979171431085264e-06,
+      "loss": 0.2543,
+      "step": 839
+    },
+    {
+      "epoch": 0.05832523260658242,
+      "grad_norm": 5.432654817736759,
+      "learning_rate": 9.979068772673698e-06,
+      "loss": 0.5199,
+      "step": 840
+    },
+    {
+      "epoch": 0.05839466740730454,
+      "grad_norm": 4.9650495752362565,
+      "learning_rate": 9.978965862426471e-06,
+      "loss": 0.5773,
+      "step": 841
+    },
+    {
+      "epoch": 0.05846410220802666,
+      "grad_norm": 3.7169989933920498,
+      "learning_rate": 9.978862700348782e-06,
+      "loss": 0.2395,
+      "step": 842
+    },
+    {
+      "epoch": 0.05853353700874878,
+      "grad_norm": 6.561747574338523,
+      "learning_rate": 9.978759286445851e-06,
+      "loss": 0.7589,
+      "step": 843
+    },
+    {
+      "epoch": 0.058602971809470904,
+      "grad_norm": 5.024869506221608,
+      "learning_rate": 9.97865562072291e-06,
+      "loss": 0.2439,
+      "step": 844
+    },
+    {
+      "epoch": 0.05867240661019303,
+      "grad_norm": 5.204125246126191,
+      "learning_rate": 9.978551703185202e-06,
+      "loss": 0.762,
+      "step": 845
+    },
+    {
+      "epoch": 0.05874184141091515,
+      "grad_norm": 4.768023281407063,
+      "learning_rate": 9.978447533837981e-06,
+      "loss": 0.4073,
+      "step": 846
+    },
+    {
+      "epoch": 0.05881127621163727,
+      "grad_norm": 5.311523694767885,
+      "learning_rate": 9.978343112686519e-06,
+      "loss": 0.4754,
+      "step": 847
+    },
+    {
+      "epoch": 0.058880711012359394,
+      "grad_norm": 3.6270137334563155,
+      "learning_rate": 9.978238439736093e-06,
+      "loss": 0.3285,
+      "step": 848
+    },
+    {
+      "epoch": 0.058950145813081514,
+      "grad_norm": 5.811385562564773,
+      "learning_rate": 9.978133514992004e-06,
+      "loss": 0.9314,
+      "step": 849
+    },
+    {
+      "epoch": 0.059019580613803635,
+      "grad_norm": 5.220765594471936,
+      "learning_rate": 9.978028338459553e-06,
+      "loss": 0.5203,
+      "step": 850
+    },
+    {
+      "epoch": 0.05908901541452576,
+      "grad_norm": 4.284268000302469,
+      "learning_rate": 9.977922910144061e-06,
+      "loss": 0.3385,
+      "step": 851
+    },
+    {
+      "epoch": 0.05915845021524788,
+      "grad_norm": 3.778962136553783,
+      "learning_rate": 9.977817230050861e-06,
+      "loss": 0.2803,
+      "step": 852
+    },
+    {
+      "epoch": 0.059227885015970004,
+      "grad_norm": 5.990180129947594,
+      "learning_rate": 9.977711298185299e-06,
+      "loss": 0.7038,
+      "step": 853
+    },
+    {
+      "epoch": 0.059297319816692125,
+      "grad_norm": 5.2643320168657315,
+      "learning_rate": 9.977605114552734e-06,
+      "loss": 0.4497,
+      "step": 854
+    },
+    {
+      "epoch": 0.059366754617414245,
+      "grad_norm": 5.566776662591551,
+      "learning_rate": 9.977498679158532e-06,
+      "loss": 0.4686,
+      "step": 855
+    },
+    {
+      "epoch": 0.05943618941813637,
+      "grad_norm": 5.081580337441843,
+      "learning_rate": 9.977391992008082e-06,
+      "loss": 0.5855,
+      "step": 856
+    },
+    {
+      "epoch": 0.059505624218858494,
+      "grad_norm": 6.290713035797393,
+      "learning_rate": 9.977285053106773e-06,
+      "loss": 0.9367,
+      "step": 857
+    },
+    {
+      "epoch": 0.059575059019580615,
+      "grad_norm": 4.852832831147473,
+      "learning_rate": 9.977177862460021e-06,
+      "loss": 0.5509,
+      "step": 858
+    },
+    {
+      "epoch": 0.059644493820302735,
+      "grad_norm": 5.8750319175016745,
+      "learning_rate": 9.977070420073245e-06,
+      "loss": 0.7002,
+      "step": 859
+    },
+    {
+      "epoch": 0.059713928621024856,
+      "grad_norm": 5.366944502014877,
+      "learning_rate": 9.976962725951878e-06,
+      "loss": 0.683,
+      "step": 860
+    },
+    {
+      "epoch": 0.05978336342174698,
+      "grad_norm": 6.393163947069777,
+      "learning_rate": 9.97685478010137e-06,
+      "loss": 0.912,
+      "step": 861
+    },
+    {
+      "epoch": 0.059852798222469104,
+      "grad_norm": 4.401148216867979,
+      "learning_rate": 9.976746582527177e-06,
+      "loss": 0.5825,
+      "step": 862
+    },
+    {
+      "epoch": 0.059922233023191225,
+      "grad_norm": 4.654659423810781,
+      "learning_rate": 9.976638133234773e-06,
+      "loss": 0.6291,
+      "step": 863
+    },
+    {
+      "epoch": 0.059991667823913346,
+      "grad_norm": 5.470211469742042,
+      "learning_rate": 9.976529432229645e-06,
+      "loss": 0.8138,
+      "step": 864
+    },
+    {
+      "epoch": 0.060061102624635467,
+      "grad_norm": 4.825491035778757,
+      "learning_rate": 9.97642047951729e-06,
+      "loss": 0.3348,
+      "step": 865
+    },
+    {
+      "epoch": 0.06013053742535759,
+      "grad_norm": 3.6568859412625394,
+      "learning_rate": 9.97631127510322e-06,
+      "loss": 0.4965,
+      "step": 866
+    },
+    {
+      "epoch": 0.06019997222607971,
+      "grad_norm": 5.407199196874604,
+      "learning_rate": 9.976201818992952e-06,
+      "loss": 0.7079,
+      "step": 867
+    },
+    {
+      "epoch": 0.060269407026801836,
+      "grad_norm": 4.101236946933585,
+      "learning_rate": 9.97609211119203e-06,
+      "loss": 0.4677,
+      "step": 868
+    },
+    {
+      "epoch": 0.060338841827523956,
+      "grad_norm": 4.238608913628256,
+      "learning_rate": 9.975982151705998e-06,
+      "loss": 0.4338,
+      "step": 869
+    },
+    {
+      "epoch": 0.06040827662824608,
+      "grad_norm": 3.586802823860022,
+      "learning_rate": 9.975871940540422e-06,
+      "loss": 0.3797,
+      "step": 870
+    },
+    {
+      "epoch": 0.0604777114289682,
+      "grad_norm": 5.015545499131453,
+      "learning_rate": 9.975761477700873e-06,
+      "loss": 0.7338,
+      "step": 871
+    },
+    {
+      "epoch": 0.06054714622969032,
+      "grad_norm": 4.348134558601847,
+      "learning_rate": 9.975650763192938e-06,
+      "loss": 0.4368,
+      "step": 872
+    },
+    {
+      "epoch": 0.060616581030412446,
+      "grad_norm": 5.17596403133597,
+      "learning_rate": 9.975539797022218e-06,
+      "loss": 0.4639,
+      "step": 873
+    },
+    {
+      "epoch": 0.06068601583113457,
+      "grad_norm": 6.9953767658247665,
+      "learning_rate": 9.975428579194327e-06,
+      "loss": 0.7597,
+      "step": 874
+    },
+    {
+      "epoch": 0.06075545063185669,
+      "grad_norm": 5.340506592070401,
+      "learning_rate": 9.975317109714886e-06,
+      "loss": 0.6475,
+      "step": 875
+    },
+    {
+      "epoch": 0.06082488543257881,
+      "grad_norm": 5.29047765463008,
+      "learning_rate": 9.975205388589537e-06,
+      "loss": 0.5791,
+      "step": 876
+    },
+    {
+      "epoch": 0.06089432023330093,
+      "grad_norm": 5.6974687377834625,
+      "learning_rate": 9.975093415823928e-06,
+      "loss": 0.6237,
+      "step": 877
+    },
+    {
+      "epoch": 0.06096375503402305,
+      "grad_norm": 5.548071497529065,
+      "learning_rate": 9.974981191423725e-06,
+      "loss": 0.468,
+      "step": 878
+    },
+    {
+      "epoch": 0.06103318983474518,
+      "grad_norm": 3.106728793995161,
+      "learning_rate": 9.974868715394604e-06,
+      "loss": 0.2516,
+      "step": 879
+    },
+    {
+      "epoch": 0.0611026246354673,
+      "grad_norm": 4.0955982347025905,
+      "learning_rate": 9.97475598774225e-06,
+      "loss": 0.5117,
+      "step": 880
+    },
+    {
+      "epoch": 0.06117205943618942,
+      "grad_norm": 5.513738555852737,
+      "learning_rate": 9.97464300847237e-06,
+      "loss": 0.7659,
+      "step": 881
+    },
+    {
+      "epoch": 0.06124149423691154,
+      "grad_norm": 4.742117963553443,
+      "learning_rate": 9.974529777590674e-06,
+      "loss": 0.5422,
+      "step": 882
+    },
+    {
+      "epoch": 0.06131092903763366,
+      "grad_norm": 3.159027201489317,
+      "learning_rate": 9.974416295102892e-06,
+      "loss": 0.1298,
+      "step": 883
+    },
+    {
+      "epoch": 0.06138036383835578,
+      "grad_norm": 4.869362082844647,
+      "learning_rate": 9.974302561014762e-06,
+      "loss": 0.6021,
+      "step": 884
+    },
+    {
+      "epoch": 0.06144979863907791,
+      "grad_norm": 3.059106279202786,
+      "learning_rate": 9.974188575332036e-06,
+      "loss": 0.2314,
+      "step": 885
+    },
+    {
+      "epoch": 0.06151923343980003,
+      "grad_norm": 4.763636407962317,
+      "learning_rate": 9.974074338060482e-06,
+      "loss": 0.5019,
+      "step": 886
+    },
+    {
+      "epoch": 0.06158866824052215,
+      "grad_norm": 4.987420676154112,
+      "learning_rate": 9.973959849205876e-06,
+      "loss": 0.448,
+      "step": 887
+    },
+    {
+      "epoch": 0.06165810304124427,
+      "grad_norm": 5.604948309749669,
+      "learning_rate": 9.973845108774012e-06,
+      "loss": 0.2817,
+      "step": 888
+    },
+    {
+      "epoch": 0.06172753784196639,
+      "grad_norm": 4.761072252438883,
+      "learning_rate": 9.973730116770688e-06,
+      "loss": 0.4862,
+      "step": 889
+    },
+    {
+      "epoch": 0.06179697264268851,
+      "grad_norm": 5.300757562179495,
+      "learning_rate": 9.973614873201722e-06,
+      "loss": 0.5307,
+      "step": 890
+    },
+    {
+      "epoch": 0.06186640744341064,
+      "grad_norm": 5.713485720989508,
+      "learning_rate": 9.973499378072947e-06,
+      "loss": 0.5329,
+      "step": 891
+    },
+    {
+      "epoch": 0.06193584224413276,
+      "grad_norm": 6.090220816575116,
+      "learning_rate": 9.973383631390199e-06,
+      "loss": 0.7065,
+      "step": 892
+    },
+    {
+      "epoch": 0.06200527704485488,
+      "grad_norm": 2.959788149743685,
+      "learning_rate": 9.973267633159335e-06,
+      "loss": 0.3194,
+      "step": 893
+    },
+    {
+      "epoch": 0.062074711845577,
+      "grad_norm": 5.114553588918628,
+      "learning_rate": 9.973151383386222e-06,
+      "loss": 0.6698,
+      "step": 894
+    },
+    {
+      "epoch": 0.06214414664629912,
+      "grad_norm": 4.152304745216256,
+      "learning_rate": 9.97303488207674e-06,
+      "loss": 0.4091,
+      "step": 895
+    },
+    {
+      "epoch": 0.06221358144702125,
+      "grad_norm": 3.9643779803590977,
+      "learning_rate": 9.972918129236779e-06,
+      "loss": 0.3985,
+      "step": 896
+    },
+    {
+      "epoch": 0.06228301624774337,
+      "grad_norm": 4.546765992530377,
+      "learning_rate": 9.972801124872248e-06,
+      "loss": 0.5113,
+      "step": 897
+    },
+    {
+      "epoch": 0.06235245104846549,
+      "grad_norm": 4.2010299821223125,
+      "learning_rate": 9.972683868989063e-06,
+      "loss": 0.322,
+      "step": 898
+    },
+    {
+      "epoch": 0.06242188584918761,
+      "grad_norm": 4.9671207282442555,
+      "learning_rate": 9.972566361593155e-06,
+      "loss": 0.4954,
+      "step": 899
+    },
+    {
+      "epoch": 0.06249132064990973,
+      "grad_norm": 5.127850447152676,
+      "learning_rate": 9.972448602690467e-06,
+      "loss": 0.5614,
+      "step": 900
+    },
+    {
+      "epoch": 0.06256075545063186,
+      "grad_norm": 5.924564244264348,
+      "learning_rate": 9.972330592286956e-06,
+      "loss": 0.8143,
+      "step": 901
+    },
+    {
+      "epoch": 0.06263019025135398,
+      "grad_norm": 4.5368959262939015,
+      "learning_rate": 9.97221233038859e-06,
+      "loss": 0.4361,
+      "step": 902
+    },
+    {
+      "epoch": 0.0626996250520761,
+      "grad_norm": 4.916949533125268,
+      "learning_rate": 9.97209381700135e-06,
+      "loss": 0.5876,
+      "step": 903
+    },
+    {
+      "epoch": 0.06276905985279822,
+      "grad_norm": 5.1309711609392865,
+      "learning_rate": 9.971975052131231e-06,
+      "loss": 0.7273,
+      "step": 904
+    },
+    {
+      "epoch": 0.06283849465352034,
+      "grad_norm": 5.080705813535925,
+      "learning_rate": 9.97185603578424e-06,
+      "loss": 0.6639,
+      "step": 905
+    },
+    {
+      "epoch": 0.06290792945424246,
+      "grad_norm": 3.9934500093197767,
+      "learning_rate": 9.971736767966397e-06,
+      "loss": 0.4592,
+      "step": 906
+    },
+    {
+      "epoch": 0.06297736425496459,
+      "grad_norm": 6.43875890376898,
+      "learning_rate": 9.971617248683736e-06,
+      "loss": 0.6645,
+      "step": 907
+    },
+    {
+      "epoch": 0.0630467990556867,
+      "grad_norm": 4.067354097287484,
+      "learning_rate": 9.971497477942297e-06,
+      "loss": 0.5227,
+      "step": 908
+    },
+    {
+      "epoch": 0.06311623385640883,
+      "grad_norm": 4.861073689208835,
+      "learning_rate": 9.971377455748144e-06,
+      "loss": 0.4918,
+      "step": 909
+    },
+    {
+      "epoch": 0.06318566865713095,
+      "grad_norm": 5.665398584991399,
+      "learning_rate": 9.971257182107343e-06,
+      "loss": 0.5597,
+      "step": 910
+    },
+    {
+      "epoch": 0.06325510345785308,
+      "grad_norm": 5.2107106156001235,
+      "learning_rate": 9.97113665702598e-06,
+      "loss": 0.7108,
+      "step": 911
+    },
+    {
+      "epoch": 0.0633245382585752,
+      "grad_norm": 5.74240152950369,
+      "learning_rate": 9.97101588051015e-06,
+      "loss": 0.7618,
+      "step": 912
+    },
+    {
+      "epoch": 0.06339397305929732,
+      "grad_norm": 4.99449985710756,
+      "learning_rate": 9.970894852565963e-06,
+      "loss": 0.6625,
+      "step": 913
+    },
+    {
+      "epoch": 0.06346340786001944,
+      "grad_norm": 4.8980195397179855,
+      "learning_rate": 9.970773573199537e-06,
+      "loss": 0.7337,
+      "step": 914
+    },
+    {
+      "epoch": 0.06353284266074156,
+      "grad_norm": 4.134093868592578,
+      "learning_rate": 9.97065204241701e-06,
+      "loss": 0.5792,
+      "step": 915
+    },
+    {
+      "epoch": 0.06360227746146369,
+      "grad_norm": 5.694789794956021,
+      "learning_rate": 9.970530260224526e-06,
+      "loss": 0.7233,
+      "step": 916
+    },
+    {
+      "epoch": 0.0636717122621858,
+      "grad_norm": 5.498143501458203,
+      "learning_rate": 9.970408226628248e-06,
+      "loss": 0.5447,
+      "step": 917
+    },
+    {
+      "epoch": 0.06374114706290793,
+      "grad_norm": 5.410190167568352,
+      "learning_rate": 9.970285941634346e-06,
+      "loss": 0.7237,
+      "step": 918
+    },
+    {
+      "epoch": 0.06381058186363005,
+      "grad_norm": 5.539925312538419,
+      "learning_rate": 9.970163405249004e-06,
+      "loss": 0.6118,
+      "step": 919
+    },
+    {
+      "epoch": 0.06388001666435217,
+      "grad_norm": 5.876427794128868,
+      "learning_rate": 9.97004061747842e-06,
+      "loss": 0.5688,
+      "step": 920
+    },
+    {
+      "epoch": 0.06394945146507429,
+      "grad_norm": 4.2317853639985366,
+      "learning_rate": 9.969917578328808e-06,
+      "loss": 0.3539,
+      "step": 921
+    },
+    {
+      "epoch": 0.06401888626579642,
+      "grad_norm": 4.5892420963011515,
+      "learning_rate": 9.969794287806389e-06,
+      "loss": 0.5841,
+      "step": 922
+    },
+    {
+      "epoch": 0.06408832106651854,
+      "grad_norm": 3.3146340769070424,
+      "learning_rate": 9.969670745917396e-06,
+      "loss": 0.2819,
+      "step": 923
+    },
+    {
+      "epoch": 0.06415775586724067,
+      "grad_norm": 5.081862364209531,
+      "learning_rate": 9.969546952668084e-06,
+      "loss": 0.6818,
+      "step": 924
+    },
+    {
+      "epoch": 0.06422719066796279,
+      "grad_norm": 4.631114001452725,
+      "learning_rate": 9.969422908064707e-06,
+      "loss": 0.523,
+      "step": 925
+    },
+    {
+      "epoch": 0.0642966254686849,
+      "grad_norm": 4.901189625878395,
+      "learning_rate": 9.969298612113544e-06,
+      "loss": 0.6004,
+      "step": 926
+    },
+    {
+      "epoch": 0.06436606026940703,
+      "grad_norm": 5.299401241730944,
+      "learning_rate": 9.96917406482088e-06,
+      "loss": 0.7762,
+      "step": 927
+    },
+    {
+      "epoch": 0.06443549507012915,
+      "grad_norm": 5.542163696763225,
+      "learning_rate": 9.969049266193015e-06,
+      "loss": 0.4865,
+      "step": 928
+    },
+    {
+      "epoch": 0.06450492987085127,
+      "grad_norm": 4.465404114208305,
+      "learning_rate": 9.96892421623626e-06,
+      "loss": 0.4133,
+      "step": 929
+    },
+    {
+      "epoch": 0.06457436467157339,
+      "grad_norm": 6.35869173710654,
+      "learning_rate": 9.968798914956943e-06,
+      "loss": 0.892,
+      "step": 930
+    },
+    {
+      "epoch": 0.06464379947229551,
+      "grad_norm": 6.9370680148147335,
+      "learning_rate": 9.968673362361399e-06,
+      "loss": 0.6944,
+      "step": 931
+    },
+    {
+      "epoch": 0.06471323427301763,
+      "grad_norm": 5.661517076298483,
+      "learning_rate": 9.968547558455978e-06,
+      "loss": 0.5798,
+      "step": 932
+    },
+    {
+      "epoch": 0.06478266907373975,
+      "grad_norm": 4.9951964780944005,
+      "learning_rate": 9.968421503247043e-06,
+      "loss": 0.7221,
+      "step": 933
+    },
+    {
+      "epoch": 0.06485210387446189,
+      "grad_norm": 4.0016157505836665,
+      "learning_rate": 9.968295196740972e-06,
+      "loss": 0.3895,
+      "step": 934
+    },
+    {
+      "epoch": 0.064921538675184,
+      "grad_norm": 9.559840478550852,
+      "learning_rate": 9.968168638944152e-06,
+      "loss": 0.2989,
+      "step": 935
+    },
+    {
+      "epoch": 0.06499097347590613,
+      "grad_norm": 4.935931500016278,
+      "learning_rate": 9.968041829862982e-06,
+      "loss": 0.5947,
+      "step": 936
+    },
+    {
+      "epoch": 0.06506040827662825,
+      "grad_norm": 5.587065543745426,
+      "learning_rate": 9.96791476950388e-06,
+      "loss": 0.5525,
+      "step": 937
+    },
+    {
+      "epoch": 0.06512984307735037,
+      "grad_norm": 3.8792570713408026,
+      "learning_rate": 9.967787457873271e-06,
+      "loss": 0.4604,
+      "step": 938
+    },
+    {
+      "epoch": 0.06519927787807249,
+      "grad_norm": 5.445166715053491,
+      "learning_rate": 9.967659894977593e-06,
+      "loss": 0.6125,
+      "step": 939
+    },
+    {
+      "epoch": 0.06526871267879461,
+      "grad_norm": 5.6259443769761805,
+      "learning_rate": 9.967532080823296e-06,
+      "loss": 0.3986,
+      "step": 940
+    },
+    {
+      "epoch": 0.06533814747951673,
+      "grad_norm": 5.069460638270786,
+      "learning_rate": 9.967404015416852e-06,
+      "loss": 0.4489,
+      "step": 941
+    },
+    {
+      "epoch": 0.06540758228023885,
+      "grad_norm": 3.889412231608481,
+      "learning_rate": 9.967275698764731e-06,
+      "loss": 0.2919,
+      "step": 942
+    },
+    {
+      "epoch": 0.06547701708096097,
+      "grad_norm": 6.258305167921691,
+      "learning_rate": 9.967147130873428e-06,
+      "loss": 0.679,
+      "step": 943
+    },
+    {
+      "epoch": 0.0655464518816831,
+      "grad_norm": 4.935509271523038,
+      "learning_rate": 9.967018311749442e-06,
+      "loss": 0.3051,
+      "step": 944
+    },
+    {
+      "epoch": 0.06561588668240523,
+      "grad_norm": 4.599972173731018,
+      "learning_rate": 9.966889241399292e-06,
+      "loss": 0.5479,
+      "step": 945
+    },
+    {
+      "epoch": 0.06568532148312735,
+      "grad_norm": 5.257587420403792,
+      "learning_rate": 9.966759919829504e-06,
+      "loss": 0.4262,
+      "step": 946
+    },
+    {
+      "epoch": 0.06575475628384947,
+      "grad_norm": 3.300766154660848,
+      "learning_rate": 9.966630347046618e-06,
+      "loss": 0.2644,
+      "step": 947
+    },
+    {
+      "epoch": 0.06582419108457159,
+      "grad_norm": 4.686274300288148,
+      "learning_rate": 9.96650052305719e-06,
+      "loss": 0.448,
+      "step": 948
+    },
+    {
+      "epoch": 0.06589362588529371,
+      "grad_norm": 5.923122079322992,
+      "learning_rate": 9.966370447867786e-06,
+      "loss": 0.9193,
+      "step": 949
+    },
+    {
+      "epoch": 0.06596306068601583,
+      "grad_norm": 4.929882151993082,
+      "learning_rate": 9.966240121484986e-06,
+      "loss": 0.3997,
+      "step": 950
+    },
+    {
+      "epoch": 0.06603249548673795,
+      "grad_norm": 5.271498456006508,
+      "learning_rate": 9.966109543915379e-06,
+      "loss": 0.4331,
+      "step": 951
+    },
+    {
+      "epoch": 0.06610193028746007,
+      "grad_norm": 4.840715873537683,
+      "learning_rate": 9.96597871516557e-06,
+      "loss": 0.5001,
+      "step": 952
+    },
+    {
+      "epoch": 0.0661713650881822,
+      "grad_norm": 5.10183667515691,
+      "learning_rate": 9.965847635242178e-06,
+      "loss": 0.507,
+      "step": 953
+    },
+    {
+      "epoch": 0.06624079988890431,
+      "grad_norm": 5.414356583087017,
+      "learning_rate": 9.96571630415183e-06,
+      "loss": 0.5994,
+      "step": 954
+    },
+    {
+      "epoch": 0.06631023468962643,
+      "grad_norm": 4.8547297969582015,
+      "learning_rate": 9.965584721901174e-06,
+      "loss": 0.4566,
+      "step": 955
+    },
+    {
+      "epoch": 0.06637966949034857,
+      "grad_norm": 5.432584003569413,
+      "learning_rate": 9.965452888496858e-06,
+      "loss": 0.6977,
+      "step": 956
+    },
+    {
+      "epoch": 0.06644910429107069,
+      "grad_norm": 5.769937502432863,
+      "learning_rate": 9.965320803945556e-06,
+      "loss": 0.6047,
+      "step": 957
+    },
+    {
+      "epoch": 0.06651853909179281,
+      "grad_norm": 3.5877495567883657,
+      "learning_rate": 9.965188468253946e-06,
+      "loss": 0.403,
+      "step": 958
+    },
+    {
+      "epoch": 0.06658797389251493,
+      "grad_norm": 4.3328059547176725,
+      "learning_rate": 9.965055881428722e-06,
+      "loss": 0.475,
+      "step": 959
+    },
+    {
+      "epoch": 0.06665740869323705,
+      "grad_norm": 5.070560454634842,
+      "learning_rate": 9.964923043476588e-06,
+      "loss": 0.9678,
+      "step": 960
+    },
+    {
+      "epoch": 0.06672684349395917,
+      "grad_norm": 4.199185495674789,
+      "learning_rate": 9.964789954404266e-06,
+      "loss": 0.4211,
+      "step": 961
+    },
+    {
+      "epoch": 0.0667962782946813,
+      "grad_norm": 5.000263599913347,
+      "learning_rate": 9.964656614218487e-06,
+      "loss": 0.6436,
+      "step": 962
+    },
+    {
+      "epoch": 0.06686571309540341,
+      "grad_norm": 5.248422134006128,
+      "learning_rate": 9.964523022925994e-06,
+      "loss": 0.3279,
+      "step": 963
+    },
+    {
+      "epoch": 0.06693514789612554,
+      "grad_norm": 4.63320215509228,
+      "learning_rate": 9.964389180533542e-06,
+      "loss": 0.559,
+      "step": 964
+    },
+    {
+      "epoch": 0.06700458269684766,
+      "grad_norm": 4.808317605138757,
+      "learning_rate": 9.964255087047905e-06,
+      "loss": 0.3749,
+      "step": 965
+    },
+    {
+      "epoch": 0.06707401749756978,
+      "grad_norm": 4.748395187083906,
+      "learning_rate": 9.964120742475862e-06,
+      "loss": 0.4182,
+      "step": 966
+    },
+    {
+      "epoch": 0.0671434522982919,
+      "grad_norm": 4.618444970402005,
+      "learning_rate": 9.963986146824212e-06,
+      "loss": 0.5677,
+      "step": 967
+    },
+    {
+      "epoch": 0.06721288709901403,
+      "grad_norm": 4.3391769180757915,
+      "learning_rate": 9.963851300099755e-06,
+      "loss": 0.4017,
+      "step": 968
+    },
+    {
+      "epoch": 0.06728232189973615,
+      "grad_norm": 4.5085605044239125,
+      "learning_rate": 9.96371620230932e-06,
+      "loss": 0.4572,
+      "step": 969
+    },
+    {
+      "epoch": 0.06735175670045827,
+      "grad_norm": 4.498879617709213,
+      "learning_rate": 9.963580853459736e-06,
+      "loss": 0.403,
+      "step": 970
+    },
+    {
+      "epoch": 0.0674211915011804,
+      "grad_norm": 6.085207824059735,
+      "learning_rate": 9.963445253557849e-06,
+      "loss": 0.3908,
+      "step": 971
+    },
+    {
+      "epoch": 0.06749062630190251,
+      "grad_norm": 6.119907990219186,
+      "learning_rate": 9.963309402610516e-06,
+      "loss": 0.9212,
+      "step": 972
+    },
+    {
+      "epoch": 0.06756006110262464,
+      "grad_norm": 5.649927276214212,
+      "learning_rate": 9.963173300624611e-06,
+      "loss": 0.6599,
+      "step": 973
+    },
+    {
+      "epoch": 0.06762949590334676,
+      "grad_norm": 5.90755139735476,
+      "learning_rate": 9.963036947607015e-06,
+      "loss": 0.9232,
+      "step": 974
+    },
+    {
+      "epoch": 0.06769893070406888,
+      "grad_norm": 4.7525889773397205,
+      "learning_rate": 9.962900343564628e-06,
+      "loss": 0.5801,
+      "step": 975
+    },
+    {
+      "epoch": 0.067768365504791,
+      "grad_norm": 4.805003349380725,
+      "learning_rate": 9.962763488504358e-06,
+      "loss": 0.5882,
+      "step": 976
+    },
+    {
+      "epoch": 0.06783780030551312,
+      "grad_norm": 4.258515303354054,
+      "learning_rate": 9.962626382433126e-06,
+      "loss": 0.4745,
+      "step": 977
+    },
+    {
+      "epoch": 0.06790723510623524,
+      "grad_norm": 5.428223264656458,
+      "learning_rate": 9.962489025357866e-06,
+      "loss": 0.5739,
+      "step": 978
+    },
+    {
+      "epoch": 0.06797666990695737,
+      "grad_norm": 4.483352979682293,
+      "learning_rate": 9.962351417285527e-06,
+      "loss": 0.5748,
+      "step": 979
+    },
+    {
+      "epoch": 0.0680461047076795,
+      "grad_norm": 5.206350310861984,
+      "learning_rate": 9.962213558223069e-06,
+      "loss": 0.6655,
+      "step": 980
+    },
+    {
+      "epoch": 0.06811553950840162,
+      "grad_norm": 4.289404062911407,
+      "learning_rate": 9.962075448177464e-06,
+      "loss": 0.4287,
+      "step": 981
+    },
+    {
+      "epoch": 0.06818497430912374,
+      "grad_norm": 4.524158306888315,
+      "learning_rate": 9.961937087155697e-06,
+      "loss": 0.3906,
+      "step": 982
+    },
+    {
+      "epoch": 0.06825440910984586,
+      "grad_norm": 4.946946002356679,
+      "learning_rate": 9.96179847516477e-06,
+      "loss": 0.7494,
+      "step": 983
+    },
+    {
+      "epoch": 0.06832384391056798,
+      "grad_norm": 4.4642887682414045,
+      "learning_rate": 9.961659612211687e-06,
+      "loss": 0.3617,
+      "step": 984
+    },
+    {
+      "epoch": 0.0683932787112901,
+      "grad_norm": 5.500827820213797,
+      "learning_rate": 9.961520498303478e-06,
+      "loss": 0.8504,
+      "step": 985
+    },
+    {
+      "epoch": 0.06846271351201222,
+      "grad_norm": 5.136642431754584,
+      "learning_rate": 9.961381133447175e-06,
+      "loss": 0.5318,
+      "step": 986
+    },
+    {
+      "epoch": 0.06853214831273434,
+      "grad_norm": 5.52539288210066,
+      "learning_rate": 9.96124151764983e-06,
+      "loss": 0.6852,
+      "step": 987
+    },
+    {
+      "epoch": 0.06860158311345646,
+      "grad_norm": 4.428777704571156,
+      "learning_rate": 9.9611016509185e-06,
+      "loss": 0.3122,
+      "step": 988
+    },
+    {
+      "epoch": 0.06867101791417858,
+      "grad_norm": 5.2708835637065246,
+      "learning_rate": 9.960961533260266e-06,
+      "loss": 0.6886,
+      "step": 989
+    },
+    {
+      "epoch": 0.0687404527149007,
+      "grad_norm": 3.9931575066170146,
+      "learning_rate": 9.96082116468221e-06,
+      "loss": 0.4194,
+      "step": 990
+    },
+    {
+      "epoch": 0.06880988751562284,
+      "grad_norm": 6.832481155647511,
+      "learning_rate": 9.960680545191433e-06,
+      "loss": 0.9276,
+      "step": 991
+    },
+    {
+      "epoch": 0.06887932231634496,
+      "grad_norm": 5.7151866707559895,
+      "learning_rate": 9.960539674795047e-06,
+      "loss": 0.5466,
+      "step": 992
+    },
+    {
+      "epoch": 0.06894875711706708,
+      "grad_norm": 5.2636203243830995,
+      "learning_rate": 9.960398553500178e-06,
+      "loss": 0.6021,
+      "step": 993
+    },
+    {
+      "epoch": 0.0690181919177892,
+      "grad_norm": 3.9032444533172472,
+      "learning_rate": 9.960257181313963e-06,
+      "loss": 0.4737,
+      "step": 994
+    },
+    {
+      "epoch": 0.06908762671851132,
+      "grad_norm": 4.64332598624562,
+      "learning_rate": 9.960115558243553e-06,
+      "loss": 0.4973,
+      "step": 995
+    },
+    {
+      "epoch": 0.06915706151923344,
+      "grad_norm": 4.938738787736154,
+      "learning_rate": 9.95997368429611e-06,
+      "loss": 0.6196,
+      "step": 996
+    },
+    {
+      "epoch": 0.06922649631995556,
+      "grad_norm": 3.041256838796851,
+      "learning_rate": 9.959831559478811e-06,
+      "loss": 0.122,
+      "step": 997
+    },
+    {
+      "epoch": 0.06929593112067768,
+      "grad_norm": 4.056889867993404,
+      "learning_rate": 9.959689183798843e-06,
+      "loss": 0.405,
+      "step": 998
+    },
+    {
+      "epoch": 0.0693653659213998,
+      "grad_norm": 5.030799742114715,
+      "learning_rate": 9.95954655726341e-06,
+      "loss": 0.4943,
+      "step": 999
+    },
+    {
+      "epoch": 0.06943480072212192,
+      "grad_norm": 5.411608639754731,
+      "learning_rate": 9.959403679879724e-06,
+      "loss": 0.6233,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06950423552284404,
+      "grad_norm": 3.6928322172334913,
+      "learning_rate": 9.959260551655012e-06,
+      "loss": 0.328,
+      "step": 1001
+    },
+    {
+      "epoch": 0.06957367032356618,
+      "grad_norm": 4.59437761930444,
+      "learning_rate": 9.959117172596513e-06,
+      "loss": 0.3028,
+      "step": 1002
+    },
+    {
+      "epoch": 0.0696431051242883,
+      "grad_norm": 5.292544934425304,
+      "learning_rate": 9.958973542711479e-06,
+      "loss": 0.6095,
+      "step": 1003
+    },
+    {
+      "epoch": 0.06971253992501042,
+      "grad_norm": 4.085850586704651,
+      "learning_rate": 9.958829662007175e-06,
+      "loss": 0.512,
+      "step": 1004
+    },
+    {
+      "epoch": 0.06978197472573254,
+      "grad_norm": 5.21134476478826,
+      "learning_rate": 9.958685530490877e-06,
+      "loss": 0.4792,
+      "step": 1005
+    },
+    {
+      "epoch": 0.06985140952645466,
+      "grad_norm": 4.760194137837527,
+      "learning_rate": 9.958541148169876e-06,
+      "loss": 0.4877,
+      "step": 1006
+    },
+    {
+      "epoch": 0.06992084432717678,
+      "grad_norm": 5.168300952803078,
+      "learning_rate": 9.958396515051475e-06,
+      "loss": 0.5909,
+      "step": 1007
+    },
+    {
+      "epoch": 0.0699902791278989,
+      "grad_norm": 4.984849652097988,
+      "learning_rate": 9.95825163114299e-06,
+      "loss": 0.4613,
+      "step": 1008
+    },
+    {
+      "epoch": 0.07005971392862102,
+      "grad_norm": 4.76983048599734,
+      "learning_rate": 9.958106496451747e-06,
+      "loss": 0.663,
+      "step": 1009
+    },
+    {
+      "epoch": 0.07012914872934314,
+      "grad_norm": 5.510563683412331,
+      "learning_rate": 9.957961110985087e-06,
+      "loss": 0.7078,
+      "step": 1010
+    },
+    {
+      "epoch": 0.07019858353006526,
+      "grad_norm": 5.450244464940457,
+      "learning_rate": 9.957815474750367e-06,
+      "loss": 0.5241,
+      "step": 1011
+    },
+    {
+      "epoch": 0.07026801833078739,
+      "grad_norm": 7.293180876619849,
+      "learning_rate": 9.957669587754949e-06,
+      "loss": 1.0173,
+      "step": 1012
+    },
+    {
+      "epoch": 0.0703374531315095,
+      "grad_norm": 4.776561652084263,
+      "learning_rate": 9.957523450006213e-06,
+      "loss": 0.7421,
+      "step": 1013
+    },
+    {
+      "epoch": 0.07040688793223164,
+      "grad_norm": 4.386340269433218,
+      "learning_rate": 9.95737706151155e-06,
+      "loss": 0.4798,
+      "step": 1014
+    },
+    {
+      "epoch": 0.07047632273295376,
+      "grad_norm": 5.420171081285744,
+      "learning_rate": 9.957230422278366e-06,
+      "loss": 0.5972,
+      "step": 1015
+    },
+    {
+      "epoch": 0.07054575753367588,
+      "grad_norm": 4.709384652308194,
+      "learning_rate": 9.957083532314078e-06,
+      "loss": 0.4396,
+      "step": 1016
+    },
+    {
+      "epoch": 0.070615192334398,
+      "grad_norm": 4.594456997365588,
+      "learning_rate": 9.956936391626113e-06,
+      "loss": 0.4249,
+      "step": 1017
+    },
+    {
+      "epoch": 0.07068462713512012,
+      "grad_norm": 5.417561171225137,
+      "learning_rate": 9.956789000221913e-06,
+      "loss": 0.5429,
+      "step": 1018
+    },
+    {
+      "epoch": 0.07075406193584224,
+      "grad_norm": 5.458146186881246,
+      "learning_rate": 9.956641358108936e-06,
+      "loss": 0.6841,
+      "step": 1019
+    },
+    {
+      "epoch": 0.07082349673656436,
+      "grad_norm": 6.2015737535658095,
+      "learning_rate": 9.956493465294647e-06,
+      "loss": 0.5538,
+      "step": 1020
+    },
+    {
+      "epoch": 0.07089293153728649,
+      "grad_norm": 4.2977119537080215,
+      "learning_rate": 9.956345321786529e-06,
+      "loss": 0.5605,
+      "step": 1021
+    },
+    {
+      "epoch": 0.0709623663380086,
+      "grad_norm": 4.307382266167953,
+      "learning_rate": 9.956196927592071e-06,
+      "loss": 0.3595,
+      "step": 1022
+    },
+    {
+      "epoch": 0.07103180113873073,
+      "grad_norm": 5.911114503193186,
+      "learning_rate": 9.956048282718782e-06,
+      "loss": 0.7524,
+      "step": 1023
+    },
+    {
+      "epoch": 0.07110123593945285,
+      "grad_norm": 4.29391786236575,
+      "learning_rate": 9.955899387174179e-06,
+      "loss": 0.4595,
+      "step": 1024
+    },
+    {
+      "epoch": 0.07117067074017498,
+      "grad_norm": 4.7298040688777485,
+      "learning_rate": 9.955750240965792e-06,
+      "loss": 0.4059,
+      "step": 1025
+    },
+    {
+      "epoch": 0.0712401055408971,
+      "grad_norm": 6.669446193839085,
+      "learning_rate": 9.955600844101166e-06,
+      "loss": 0.9742,
+      "step": 1026
+    },
+    {
+      "epoch": 0.07130954034161922,
+      "grad_norm": 3.9502386263885723,
+      "learning_rate": 9.955451196587857e-06,
+      "loss": 0.3498,
+      "step": 1027
+    },
+    {
+      "epoch": 0.07137897514234134,
+      "grad_norm": 4.074939068376195,
+      "learning_rate": 9.955301298433435e-06,
+      "loss": 0.4876,
+      "step": 1028
+    },
+    {
+      "epoch": 0.07144840994306347,
+      "grad_norm": 5.170854111381556,
+      "learning_rate": 9.955151149645479e-06,
+      "loss": 0.5523,
+      "step": 1029
+    },
+    {
+      "epoch": 0.07151784474378559,
+      "grad_norm": 4.187257809490731,
+      "learning_rate": 9.955000750231586e-06,
+      "loss": 0.589,
+      "step": 1030
+    },
+    {
+      "epoch": 0.0715872795445077,
+      "grad_norm": 5.16127155546752,
+      "learning_rate": 9.954850100199361e-06,
+      "loss": 0.5855,
+      "step": 1031
+    },
+    {
+      "epoch": 0.07165671434522983,
+      "grad_norm": 4.964547894200507,
+      "learning_rate": 9.954699199556426e-06,
+      "loss": 0.6431,
+      "step": 1032
+    },
+    {
+      "epoch": 0.07172614914595195,
+      "grad_norm": 4.2512883460282245,
+      "learning_rate": 9.954548048310411e-06,
+      "loss": 0.6329,
+      "step": 1033
+    },
+    {
+      "epoch": 0.07179558394667407,
+      "grad_norm": 3.8370726730536355,
+      "learning_rate": 9.954396646468962e-06,
+      "loss": 0.3079,
+      "step": 1034
+    },
+    {
+      "epoch": 0.07186501874739619,
+      "grad_norm": 4.975439229505246,
+      "learning_rate": 9.95424499403974e-06,
+      "loss": 0.5805,
+      "step": 1035
+    },
+    {
+      "epoch": 0.07193445354811831,
+      "grad_norm": 3.961435572419675,
+      "learning_rate": 9.954093091030408e-06,
+      "loss": 0.3708,
+      "step": 1036
+    },
+    {
+      "epoch": 0.07200388834884044,
+      "grad_norm": 5.925678724743737,
+      "learning_rate": 9.953940937448657e-06,
+      "loss": 0.6303,
+      "step": 1037
+    },
+    {
+      "epoch": 0.07207332314956257,
+      "grad_norm": 4.686833690892003,
+      "learning_rate": 9.953788533302176e-06,
+      "loss": 0.703,
+      "step": 1038
+    },
+    {
+      "epoch": 0.07214275795028469,
+      "grad_norm": 7.3699263513723885,
+      "learning_rate": 9.953635878598678e-06,
+      "loss": 0.717,
+      "step": 1039
+    },
+    {
+      "epoch": 0.0722121927510068,
+      "grad_norm": 4.452765672993177,
+      "learning_rate": 9.953482973345883e-06,
+      "loss": 0.5205,
+      "step": 1040
+    },
+    {
+      "epoch": 0.07228162755172893,
+      "grad_norm": 4.960614394841153,
+      "learning_rate": 9.953329817551524e-06,
+      "loss": 0.4376,
+      "step": 1041
+    },
+    {
+      "epoch": 0.07235106235245105,
+      "grad_norm": 4.589168201737519,
+      "learning_rate": 9.953176411223348e-06,
+      "loss": 0.5071,
+      "step": 1042
+    },
+    {
+      "epoch": 0.07242049715317317,
+      "grad_norm": 4.555086111883653,
+      "learning_rate": 9.953022754369115e-06,
+      "loss": 0.614,
+      "step": 1043
+    },
+    {
+      "epoch": 0.07248993195389529,
+      "grad_norm": 3.154823390915928,
+      "learning_rate": 9.952868846996594e-06,
+      "loss": 0.356,
+      "step": 1044
+    },
+    {
+      "epoch": 0.07255936675461741,
+      "grad_norm": 4.718248397747654,
+      "learning_rate": 9.952714689113572e-06,
+      "loss": 0.6445,
+      "step": 1045
+    },
+    {
+      "epoch": 0.07262880155533953,
+      "grad_norm": 4.581650676814528,
+      "learning_rate": 9.952560280727846e-06,
+      "loss": 0.5306,
+      "step": 1046
+    },
+    {
+      "epoch": 0.07269823635606165,
+      "grad_norm": 3.6644459171202963,
+      "learning_rate": 9.952405621847226e-06,
+      "loss": 0.2656,
+      "step": 1047
+    },
+    {
+      "epoch": 0.07276767115678379,
+      "grad_norm": 4.046147335314258,
+      "learning_rate": 9.95225071247953e-06,
+      "loss": 0.5103,
+      "step": 1048
+    },
+    {
+      "epoch": 0.07283710595750591,
+      "grad_norm": 5.355061234498256,
+      "learning_rate": 9.9520955526326e-06,
+      "loss": 0.7253,
+      "step": 1049
+    },
+    {
+      "epoch": 0.07290654075822803,
+      "grad_norm": 4.333199765343125,
+      "learning_rate": 9.951940142314281e-06,
+      "loss": 0.4539,
+      "step": 1050
+    },
+    {
+      "epoch": 0.07297597555895015,
+      "grad_norm": 4.070974835280359,
+      "learning_rate": 9.951784481532431e-06,
+      "loss": 0.2147,
+      "step": 1051
+    },
+    {
+      "epoch": 0.07304541035967227,
+      "grad_norm": 4.756207890337684,
+      "learning_rate": 9.951628570294927e-06,
+      "loss": 0.513,
+      "step": 1052
+    },
+    {
+      "epoch": 0.07311484516039439,
+      "grad_norm": 5.073521508665398,
+      "learning_rate": 9.951472408609652e-06,
+      "loss": 0.6551,
+      "step": 1053
+    },
+    {
+      "epoch": 0.07318427996111651,
+      "grad_norm": 4.719688942934508,
+      "learning_rate": 9.951315996484504e-06,
+      "loss": 0.5327,
+      "step": 1054
+    },
+    {
+      "epoch": 0.07325371476183863,
+      "grad_norm": 5.243207290798345,
+      "learning_rate": 9.9511593339274e-06,
+      "loss": 0.6048,
+      "step": 1055
+    },
+    {
+      "epoch": 0.07332314956256075,
+      "grad_norm": 3.944595594133608,
+      "learning_rate": 9.951002420946256e-06,
+      "loss": 0.3478,
+      "step": 1056
+    },
+    {
+      "epoch": 0.07339258436328287,
+      "grad_norm": 4.9035345496008125,
+      "learning_rate": 9.950845257549014e-06,
+      "loss": 0.5161,
+      "step": 1057
+    },
+    {
+      "epoch": 0.073462019164005,
+      "grad_norm": 6.162208538319855,
+      "learning_rate": 9.950687843743619e-06,
+      "loss": 0.4798,
+      "step": 1058
+    },
+    {
+      "epoch": 0.07353145396472713,
+      "grad_norm": 5.163469182764663,
+      "learning_rate": 9.950530179538036e-06,
+      "loss": 0.4314,
+      "step": 1059
+    },
+    {
+      "epoch": 0.07360088876544925,
+      "grad_norm": 4.908100864939862,
+      "learning_rate": 9.95037226494024e-06,
+      "loss": 0.4189,
+      "step": 1060
+    },
+    {
+      "epoch": 0.07367032356617137,
+      "grad_norm": 6.257497350136744,
+      "learning_rate": 9.950214099958212e-06,
+      "loss": 0.7988,
+      "step": 1061
+    },
+    {
+      "epoch": 0.07373975836689349,
+      "grad_norm": 4.509034251468079,
+      "learning_rate": 9.950055684599961e-06,
+      "loss": 0.4832,
+      "step": 1062
+    },
+    {
+      "epoch": 0.07380919316761561,
+      "grad_norm": 5.589379940824678,
+      "learning_rate": 9.949897018873493e-06,
+      "loss": 0.5365,
+      "step": 1063
+    },
+    {
+      "epoch": 0.07387862796833773,
+      "grad_norm": 5.1175159247520785,
+      "learning_rate": 9.949738102786836e-06,
+      "loss": 0.6402,
+      "step": 1064
+    },
+    {
+      "epoch": 0.07394806276905985,
+      "grad_norm": 5.40432800157665,
+      "learning_rate": 9.949578936348025e-06,
+      "loss": 0.7799,
+      "step": 1065
+    },
+    {
+      "epoch": 0.07401749756978197,
+      "grad_norm": 4.683268087825423,
+      "learning_rate": 9.949419519565113e-06,
+      "loss": 0.5868,
+      "step": 1066
+    },
+    {
+      "epoch": 0.0740869323705041,
+      "grad_norm": 4.636203716811295,
+      "learning_rate": 9.949259852446163e-06,
+      "loss": 0.4692,
+      "step": 1067
+    },
+    {
+      "epoch": 0.07415636717122621,
+      "grad_norm": 5.595424409010696,
+      "learning_rate": 9.949099934999252e-06,
+      "loss": 0.7494,
+      "step": 1068
+    },
+    {
+      "epoch": 0.07422580197194834,
+      "grad_norm": 4.2697931512413865,
+      "learning_rate": 9.948939767232463e-06,
+      "loss": 0.5135,
+      "step": 1069
+    },
+    {
+      "epoch": 0.07429523677267046,
+      "grad_norm": 5.485031794407256,
+      "learning_rate": 9.948779349153902e-06,
+      "loss": 0.5376,
+      "step": 1070
+    },
+    {
+      "epoch": 0.07436467157339259,
+      "grad_norm": 4.133166228380315,
+      "learning_rate": 9.948618680771682e-06,
+      "loss": 0.5123,
+      "step": 1071
+    },
+    {
+      "epoch": 0.07443410637411471,
+      "grad_norm": 5.639850381119093,
+      "learning_rate": 9.948457762093929e-06,
+      "loss": 0.7541,
+      "step": 1072
+    },
+    {
+      "epoch": 0.07450354117483683,
+      "grad_norm": 4.5853703821480725,
+      "learning_rate": 9.948296593128779e-06,
+      "loss": 0.6902,
+      "step": 1073
+    },
+    {
+      "epoch": 0.07457297597555895,
+      "grad_norm": 5.284811095457621,
+      "learning_rate": 9.94813517388439e-06,
+      "loss": 0.6339,
+      "step": 1074
+    },
+    {
+      "epoch": 0.07464241077628107,
+      "grad_norm": 6.459569378219492,
+      "learning_rate": 9.947973504368923e-06,
+      "loss": 0.7999,
+      "step": 1075
+    },
+    {
+      "epoch": 0.0747118455770032,
+      "grad_norm": 4.195725082843013,
+      "learning_rate": 9.947811584590553e-06,
+      "loss": 0.3348,
+      "step": 1076
+    },
+    {
+      "epoch": 0.07478128037772531,
+      "grad_norm": 5.218944847957534,
+      "learning_rate": 9.947649414557472e-06,
+      "loss": 0.7348,
+      "step": 1077
+    },
+    {
+      "epoch": 0.07485071517844744,
+      "grad_norm": 5.523518245898167,
+      "learning_rate": 9.947486994277884e-06,
+      "loss": 0.9064,
+      "step": 1078
+    },
+    {
+      "epoch": 0.07492014997916956,
+      "grad_norm": 4.306449068834788,
+      "learning_rate": 9.947324323760002e-06,
+      "loss": 0.3249,
+      "step": 1079
+    },
+    {
+      "epoch": 0.07498958477989168,
+      "grad_norm": 7.200146486890911,
+      "learning_rate": 9.947161403012053e-06,
+      "loss": 0.6001,
+      "step": 1080
+    },
+    {
+      "epoch": 0.0750590195806138,
+      "grad_norm": 4.376901264600628,
+      "learning_rate": 9.94699823204228e-06,
+      "loss": 0.6751,
+      "step": 1081
+    },
+    {
+      "epoch": 0.07512845438133593,
+      "grad_norm": 5.522702194373978,
+      "learning_rate": 9.946834810858932e-06,
+      "loss": 0.8449,
+      "step": 1082
+    },
+    {
+      "epoch": 0.07519788918205805,
+      "grad_norm": 5.180661175604792,
+      "learning_rate": 9.946671139470278e-06,
+      "loss": 0.4985,
+      "step": 1083
+    },
+    {
+      "epoch": 0.07526732398278017,
+      "grad_norm": 5.190318398189906,
+      "learning_rate": 9.946507217884597e-06,
+      "loss": 0.6379,
+      "step": 1084
+    },
+    {
+      "epoch": 0.0753367587835023,
+      "grad_norm": 3.7854202097064378,
+      "learning_rate": 9.946343046110176e-06,
+      "loss": 0.4363,
+      "step": 1085
+    },
+    {
+      "epoch": 0.07540619358422442,
+      "grad_norm": 5.111445069774006,
+      "learning_rate": 9.946178624155322e-06,
+      "loss": 0.6724,
+      "step": 1086
+    },
+    {
+      "epoch": 0.07547562838494654,
+      "grad_norm": 3.3378423785766316,
+      "learning_rate": 9.946013952028351e-06,
+      "loss": 0.2684,
+      "step": 1087
+    },
+    {
+      "epoch": 0.07554506318566866,
+      "grad_norm": 3.755569890286403,
+      "learning_rate": 9.945849029737588e-06,
+      "loss": 0.3951,
+      "step": 1088
+    },
+    {
+      "epoch": 0.07561449798639078,
+      "grad_norm": 4.570432242900106,
+      "learning_rate": 9.94568385729138e-06,
+      "loss": 0.4258,
+      "step": 1089
+    },
+    {
+      "epoch": 0.0756839327871129,
+      "grad_norm": 4.479712289815535,
+      "learning_rate": 9.94551843469808e-06,
+      "loss": 0.5424,
+      "step": 1090
+    },
+    {
+      "epoch": 0.07575336758783502,
+      "grad_norm": 5.011187815002273,
+      "learning_rate": 9.945352761966052e-06,
+      "loss": 0.6591,
+      "step": 1091
+    },
+    {
+      "epoch": 0.07582280238855714,
+      "grad_norm": 3.569937393157766,
+      "learning_rate": 9.945186839103678e-06,
+      "loss": 0.2467,
+      "step": 1092
+    },
+    {
+      "epoch": 0.07589223718927926,
+      "grad_norm": 4.958672460824238,
+      "learning_rate": 9.94502066611935e-06,
+      "loss": 0.3021,
+      "step": 1093
+    },
+    {
+      "epoch": 0.0759616719900014,
+      "grad_norm": 4.634282130003978,
+      "learning_rate": 9.94485424302147e-06,
+      "loss": 0.5594,
+      "step": 1094
+    },
+    {
+      "epoch": 0.07603110679072352,
+      "grad_norm": 2.6541089030073906,
+      "learning_rate": 9.94468756981846e-06,
+      "loss": 0.1268,
+      "step": 1095
+    },
+    {
+      "epoch": 0.07610054159144564,
+      "grad_norm": 4.818636885587487,
+      "learning_rate": 9.944520646518747e-06,
+      "loss": 0.5112,
+      "step": 1096
+    },
+    {
+      "epoch": 0.07616997639216776,
+      "grad_norm": 5.535042864904982,
+      "learning_rate": 9.944353473130775e-06,
+      "loss": 0.6944,
+      "step": 1097
+    },
+    {
+      "epoch": 0.07623941119288988,
+      "grad_norm": 4.683414080245664,
+      "learning_rate": 9.944186049663e-06,
+      "loss": 0.4795,
+      "step": 1098
+    },
+    {
+      "epoch": 0.076308845993612,
+      "grad_norm": 5.978393538903464,
+      "learning_rate": 9.94401837612389e-06,
+      "loss": 0.9775,
+      "step": 1099
+    },
+    {
+      "epoch": 0.07637828079433412,
+      "grad_norm": 3.59374866009973,
+      "learning_rate": 9.943850452521924e-06,
+      "loss": 0.3919,
+      "step": 1100
+    },
+    {
+      "epoch": 0.07644771559505624,
+      "grad_norm": 4.147217453959496,
+      "learning_rate": 9.943682278865596e-06,
+      "loss": 0.4297,
+      "step": 1101
+    },
+    {
+      "epoch": 0.07651715039577836,
+      "grad_norm": 6.053280441092893,
+      "learning_rate": 9.943513855163413e-06,
+      "loss": 0.734,
+      "step": 1102
+    },
+    {
+      "epoch": 0.07658658519650048,
+      "grad_norm": 4.760608243644429,
+      "learning_rate": 9.943345181423893e-06,
+      "loss": 0.5406,
+      "step": 1103
+    },
+    {
+      "epoch": 0.0766560199972226,
+      "grad_norm": 4.4103870343030405,
+      "learning_rate": 9.943176257655567e-06,
+      "loss": 0.4588,
+      "step": 1104
+    },
+    {
+      "epoch": 0.07672545479794474,
+      "grad_norm": 3.9809512648571355,
+      "learning_rate": 9.94300708386698e-06,
+      "loss": 0.3698,
+      "step": 1105
+    },
+    {
+      "epoch": 0.07679488959866686,
+      "grad_norm": 4.961850212307463,
+      "learning_rate": 9.94283766006669e-06,
+      "loss": 0.4751,
+      "step": 1106
+    },
+    {
+      "epoch": 0.07686432439938898,
+      "grad_norm": 5.0068795355998414,
+      "learning_rate": 9.942667986263261e-06,
+      "loss": 0.3645,
+      "step": 1107
+    },
+    {
+      "epoch": 0.0769337592001111,
+      "grad_norm": 4.303559660098102,
+      "learning_rate": 9.94249806246528e-06,
+      "loss": 0.4047,
+      "step": 1108
+    },
+    {
+      "epoch": 0.07700319400083322,
+      "grad_norm": 6.892718668179105,
+      "learning_rate": 9.942327888681339e-06,
+      "loss": 0.8255,
+      "step": 1109
+    },
+    {
+      "epoch": 0.07707262880155534,
+      "grad_norm": 4.148080933294095,
+      "learning_rate": 9.942157464920047e-06,
+      "loss": 0.4765,
+      "step": 1110
+    },
+    {
+      "epoch": 0.07714206360227746,
+      "grad_norm": 7.266802126398682,
+      "learning_rate": 9.941986791190022e-06,
+      "loss": 0.6487,
+      "step": 1111
+    },
+    {
+      "epoch": 0.07721149840299958,
+      "grad_norm": 5.4413906685860205,
+      "learning_rate": 9.941815867499897e-06,
+      "loss": 0.5206,
+      "step": 1112
+    },
+    {
+      "epoch": 0.0772809332037217,
+      "grad_norm": 5.564253666958973,
+      "learning_rate": 9.94164469385832e-06,
+      "loss": 1.0296,
+      "step": 1113
+    },
+    {
+      "epoch": 0.07735036800444382,
+      "grad_norm": 4.911221303387646,
+      "learning_rate": 9.941473270273943e-06,
+      "loss": 0.6206,
+      "step": 1114
+    },
+    {
+      "epoch": 0.07741980280516594,
+      "grad_norm": 3.530029882723889,
+      "learning_rate": 9.941301596755442e-06,
+      "loss": 0.2404,
+      "step": 1115
+    },
+    {
+      "epoch": 0.07748923760588806,
+      "grad_norm": 4.582876379929587,
+      "learning_rate": 9.941129673311497e-06,
+      "loss": 0.4813,
+      "step": 1116
+    },
+    {
+      "epoch": 0.0775586724066102,
+      "grad_norm": 4.692535545190693,
+      "learning_rate": 9.940957499950803e-06,
+      "loss": 0.7174,
+      "step": 1117
+    },
+    {
+      "epoch": 0.07762810720733232,
+      "grad_norm": 5.2011626352892515,
+      "learning_rate": 9.940785076682071e-06,
+      "loss": 0.4203,
+      "step": 1118
+    },
+    {
+      "epoch": 0.07769754200805444,
+      "grad_norm": 4.954208635513522,
+      "learning_rate": 9.940612403514021e-06,
+      "loss": 0.8683,
+      "step": 1119
+    },
+    {
+      "epoch": 0.07776697680877656,
+      "grad_norm": 4.16477203321742,
+      "learning_rate": 9.940439480455386e-06,
+      "loss": 0.4857,
+      "step": 1120
+    },
+    {
+      "epoch": 0.07783641160949868,
+      "grad_norm": 3.840970906816745,
+      "learning_rate": 9.940266307514912e-06,
+      "loss": 0.5077,
+      "step": 1121
+    },
+    {
+      "epoch": 0.0779058464102208,
+      "grad_norm": 4.133212698035837,
+      "learning_rate": 9.94009288470136e-06,
+      "loss": 0.4221,
+      "step": 1122
+    },
+    {
+      "epoch": 0.07797528121094292,
+      "grad_norm": 5.725692012663836,
+      "learning_rate": 9.939919212023499e-06,
+      "loss": 0.8626,
+      "step": 1123
+    },
+    {
+      "epoch": 0.07804471601166504,
+      "grad_norm": 4.820829613993883,
+      "learning_rate": 9.939745289490115e-06,
+      "loss": 0.6214,
+      "step": 1124
+    },
+    {
+      "epoch": 0.07811415081238716,
+      "grad_norm": 4.128199766906215,
+      "learning_rate": 9.939571117110002e-06,
+      "loss": 0.3708,
+      "step": 1125
+    },
+    {
+      "epoch": 0.07818358561310929,
+      "grad_norm": 4.221331125263477,
+      "learning_rate": 9.939396694891973e-06,
+      "loss": 0.5674,
+      "step": 1126
+    },
+    {
+      "epoch": 0.0782530204138314,
+      "grad_norm": 4.875702544809185,
+      "learning_rate": 9.939222022844848e-06,
+      "loss": 0.5837,
+      "step": 1127
+    },
+    {
+      "epoch": 0.07832245521455354,
+      "grad_norm": 4.298183264117277,
+      "learning_rate": 9.939047100977462e-06,
+      "loss": 0.4794,
+      "step": 1128
+    },
+    {
+      "epoch": 0.07839189001527566,
+      "grad_norm": 5.807473231071572,
+      "learning_rate": 9.938871929298662e-06,
+      "loss": 0.8932,
+      "step": 1129
+    },
+    {
+      "epoch": 0.07846132481599778,
+      "grad_norm": 4.073882014124974,
+      "learning_rate": 9.938696507817308e-06,
+      "loss": 0.2716,
+      "step": 1130
+    },
+    {
+      "epoch": 0.0785307596167199,
+      "grad_norm": 5.124283166435162,
+      "learning_rate": 9.938520836542276e-06,
+      "loss": 0.5822,
+      "step": 1131
+    },
+    {
+      "epoch": 0.07860019441744202,
+      "grad_norm": 4.47390840162405,
+      "learning_rate": 9.938344915482447e-06,
+      "loss": 0.4653,
+      "step": 1132
+    },
+    {
+      "epoch": 0.07866962921816414,
+      "grad_norm": 4.0013654048100795,
+      "learning_rate": 9.93816874464672e-06,
+      "loss": 0.5078,
+      "step": 1133
+    },
+    {
+      "epoch": 0.07873906401888627,
+      "grad_norm": 9.37207396481838,
+      "learning_rate": 9.937992324044006e-06,
+      "loss": 0.5636,
+      "step": 1134
+    },
+    {
+      "epoch": 0.07880849881960839,
+      "grad_norm": 4.91779721369415,
+      "learning_rate": 9.937815653683228e-06,
+      "loss": 0.4476,
+      "step": 1135
+    },
+    {
+      "epoch": 0.0788779336203305,
+      "grad_norm": 4.73370765935559,
+      "learning_rate": 9.937638733573321e-06,
+      "loss": 0.4221,
+      "step": 1136
+    },
+    {
+      "epoch": 0.07894736842105263,
+      "grad_norm": 5.259205142365437,
+      "learning_rate": 9.937461563723235e-06,
+      "loss": 0.4458,
+      "step": 1137
+    },
+    {
+      "epoch": 0.07901680322177475,
+      "grad_norm": 5.020464343693775,
+      "learning_rate": 9.93728414414193e-06,
+      "loss": 0.6043,
+      "step": 1138
+    },
+    {
+      "epoch": 0.07908623802249687,
+      "grad_norm": 3.643359009351639,
+      "learning_rate": 9.937106474838381e-06,
+      "loss": 0.3207,
+      "step": 1139
+    },
+    {
+      "epoch": 0.079155672823219,
+      "grad_norm": 5.091720282289669,
+      "learning_rate": 9.936928555821571e-06,
+      "loss": 0.8982,
+      "step": 1140
+    },
+    {
+      "epoch": 0.07922510762394112,
+      "grad_norm": 6.745025353426768,
+      "learning_rate": 9.936750387100502e-06,
+      "loss": 0.6157,
+      "step": 1141
+    },
+    {
+      "epoch": 0.07929454242466324,
+      "grad_norm": 5.502258543117327,
+      "learning_rate": 9.936571968684185e-06,
+      "loss": 0.6599,
+      "step": 1142
+    },
+    {
+      "epoch": 0.07936397722538537,
+      "grad_norm": 3.696056151196825,
+      "learning_rate": 9.936393300581643e-06,
+      "loss": 0.467,
+      "step": 1143
+    },
+    {
+      "epoch": 0.07943341202610749,
+      "grad_norm": 4.523466954561496,
+      "learning_rate": 9.936214382801917e-06,
+      "loss": 0.3725,
+      "step": 1144
+    },
+    {
+      "epoch": 0.0795028468268296,
+      "grad_norm": 6.360211010176474,
+      "learning_rate": 9.93603521535405e-06,
+      "loss": 0.8839,
+      "step": 1145
+    },
+    {
+      "epoch": 0.07957228162755173,
+      "grad_norm": 4.115518547206297,
+      "learning_rate": 9.935855798247107e-06,
+      "loss": 0.6318,
+      "step": 1146
+    },
+    {
+      "epoch": 0.07964171642827385,
+      "grad_norm": 5.309615110085079,
+      "learning_rate": 9.935676131490166e-06,
+      "loss": 0.7017,
+      "step": 1147
+    },
+    {
+      "epoch": 0.07971115122899597,
+      "grad_norm": 5.060261671222787,
+      "learning_rate": 9.935496215092308e-06,
+      "loss": 0.6353,
+      "step": 1148
+    },
+    {
+      "epoch": 0.07978058602971809,
+      "grad_norm": 5.54925450882981,
+      "learning_rate": 9.935316049062639e-06,
+      "loss": 0.6805,
+      "step": 1149
+    },
+    {
+      "epoch": 0.07985002083044021,
+      "grad_norm": 3.1926776436690436,
+      "learning_rate": 9.935135633410267e-06,
+      "loss": 0.288,
+      "step": 1150
+    },
+    {
+      "epoch": 0.07991945563116234,
+      "grad_norm": 4.360632216589036,
+      "learning_rate": 9.934954968144322e-06,
+      "loss": 0.6064,
+      "step": 1151
+    },
+    {
+      "epoch": 0.07998889043188447,
+      "grad_norm": 4.716966721412772,
+      "learning_rate": 9.934774053273936e-06,
+      "loss": 0.1244,
+      "step": 1152
+    },
+    {
+      "epoch": 0.08005832523260659,
+      "grad_norm": 4.342857660779656,
+      "learning_rate": 9.934592888808263e-06,
+      "loss": 0.7548,
+      "step": 1153
+    },
+    {
+      "epoch": 0.08012776003332871,
+      "grad_norm": 4.507632700787169,
+      "learning_rate": 9.934411474756467e-06,
+      "loss": 0.7116,
+      "step": 1154
+    },
+    {
+      "epoch": 0.08019719483405083,
+      "grad_norm": 4.346453634404164,
+      "learning_rate": 9.93422981112772e-06,
+      "loss": 0.456,
+      "step": 1155
+    },
+    {
+      "epoch": 0.08026662963477295,
+      "grad_norm": 4.49186851032777,
+      "learning_rate": 9.934047897931213e-06,
+      "loss": 0.5934,
+      "step": 1156
+    },
+    {
+      "epoch": 0.08033606443549507,
+      "grad_norm": 5.759450283585912,
+      "learning_rate": 9.933865735176149e-06,
+      "loss": 0.7882,
+      "step": 1157
+    },
+    {
+      "epoch": 0.08040549923621719,
+      "grad_norm": 2.8159045510153993,
+      "learning_rate": 9.933683322871737e-06,
+      "loss": 0.2394,
+      "step": 1158
+    },
+    {
+      "epoch": 0.08047493403693931,
+      "grad_norm": 4.852447298704516,
+      "learning_rate": 9.933500661027206e-06,
+      "loss": 0.5212,
+      "step": 1159
+    },
+    {
+      "epoch": 0.08054436883766143,
+      "grad_norm": 4.696042661976838,
+      "learning_rate": 9.933317749651794e-06,
+      "loss": 0.5441,
+      "step": 1160
+    },
+    {
+      "epoch": 0.08061380363838355,
+      "grad_norm": 5.711331618036688,
+      "learning_rate": 9.933134588754754e-06,
+      "loss": 0.6229,
+      "step": 1161
+    },
+    {
+      "epoch": 0.08068323843910567,
+      "grad_norm": 4.654595481876398,
+      "learning_rate": 9.932951178345347e-06,
+      "loss": 0.5526,
+      "step": 1162
+    },
+    {
+      "epoch": 0.08075267323982781,
+      "grad_norm": 4.226448844502341,
+      "learning_rate": 9.932767518432852e-06,
+      "loss": 0.3289,
+      "step": 1163
+    },
+    {
+      "epoch": 0.08082210804054993,
+      "grad_norm": 3.9605988151296385,
+      "learning_rate": 9.932583609026558e-06,
+      "loss": 0.3378,
+      "step": 1164
+    },
+    {
+      "epoch": 0.08089154284127205,
+      "grad_norm": 4.402353041515778,
+      "learning_rate": 9.932399450135765e-06,
+      "loss": 0.4658,
+      "step": 1165
+    },
+    {
+      "epoch": 0.08096097764199417,
+      "grad_norm": 5.8049205495542875,
+      "learning_rate": 9.932215041769792e-06,
+      "loss": 0.696,
+      "step": 1166
+    },
+    {
+      "epoch": 0.08103041244271629,
+      "grad_norm": 5.431009987685382,
+      "learning_rate": 9.932030383937963e-06,
+      "loss": 0.7269,
+      "step": 1167
+    },
+    {
+      "epoch": 0.08109984724343841,
+      "grad_norm": 4.635621321944196,
+      "learning_rate": 9.931845476649616e-06,
+      "loss": 0.3618,
+      "step": 1168
+    },
+    {
+      "epoch": 0.08116928204416053,
+      "grad_norm": 3.8109762867769903,
+      "learning_rate": 9.931660319914108e-06,
+      "loss": 0.3818,
+      "step": 1169
+    },
+    {
+      "epoch": 0.08123871684488265,
+      "grad_norm": 4.435645286892863,
+      "learning_rate": 9.931474913740799e-06,
+      "loss": 0.3451,
+      "step": 1170
+    },
+    {
+      "epoch": 0.08130815164560477,
+      "grad_norm": 3.7349159513651795,
+      "learning_rate": 9.93128925813907e-06,
+      "loss": 0.2724,
+      "step": 1171
+    },
+    {
+      "epoch": 0.0813775864463269,
+      "grad_norm": 4.017916522379333,
+      "learning_rate": 9.93110335311831e-06,
+      "loss": 0.2894,
+      "step": 1172
+    },
+    {
+      "epoch": 0.08144702124704901,
+      "grad_norm": 4.853429578781107,
+      "learning_rate": 9.930917198687924e-06,
+      "loss": 0.4605,
+      "step": 1173
+    },
+    {
+      "epoch": 0.08151645604777115,
+      "grad_norm": 5.338552505513365,
+      "learning_rate": 9.930730794857325e-06,
+      "loss": 0.9494,
+      "step": 1174
+    },
+    {
+      "epoch": 0.08158589084849327,
+      "grad_norm": 5.15412834138613,
+      "learning_rate": 9.930544141635943e-06,
+      "loss": 0.491,
+      "step": 1175
+    },
+    {
+      "epoch": 0.08165532564921539,
+      "grad_norm": 3.7818909823758045,
+      "learning_rate": 9.930357239033214e-06,
+      "loss": 0.4604,
+      "step": 1176
+    },
+    {
+      "epoch": 0.08172476044993751,
+      "grad_norm": 4.758105948067564,
+      "learning_rate": 9.9301700870586e-06,
+      "loss": 0.6408,
+      "step": 1177
+    },
+    {
+      "epoch": 0.08179419525065963,
+      "grad_norm": 3.8783143730941156,
+      "learning_rate": 9.929982685721557e-06,
+      "loss": 0.2288,
+      "step": 1178
+    },
+    {
+      "epoch": 0.08186363005138175,
+      "grad_norm": 5.901015286597707,
+      "learning_rate": 9.929795035031572e-06,
+      "loss": 0.7598,
+      "step": 1179
+    },
+    {
+      "epoch": 0.08193306485210387,
+      "grad_norm": 5.00760882401945,
+      "learning_rate": 9.929607134998132e-06,
+      "loss": 0.5429,
+      "step": 1180
+    },
+    {
+      "epoch": 0.082002499652826,
+      "grad_norm": 4.79910973031596,
+      "learning_rate": 9.929418985630739e-06,
+      "loss": 0.4473,
+      "step": 1181
+    },
+    {
+      "epoch": 0.08207193445354811,
+      "grad_norm": 4.055070365184233,
+      "learning_rate": 9.929230586938914e-06,
+      "loss": 0.3835,
+      "step": 1182
+    },
+    {
+      "epoch": 0.08214136925427024,
+      "grad_norm": 5.303129382891074,
+      "learning_rate": 9.929041938932183e-06,
+      "loss": 0.7743,
+      "step": 1183
+    },
+    {
+      "epoch": 0.08221080405499236,
+      "grad_norm": 4.151930400806376,
+      "learning_rate": 9.928853041620088e-06,
+      "loss": 0.3956,
+      "step": 1184
+    },
+    {
+      "epoch": 0.08228023885571449,
+      "grad_norm": 5.0833415244397955,
+      "learning_rate": 9.928663895012186e-06,
+      "loss": 0.4021,
+      "step": 1185
+    },
+    {
+      "epoch": 0.08234967365643661,
+      "grad_norm": 4.533648861832096,
+      "learning_rate": 9.928474499118037e-06,
+      "loss": 0.4614,
+      "step": 1186
+    },
+    {
+      "epoch": 0.08241910845715873,
+      "grad_norm": 4.863303884673229,
+      "learning_rate": 9.928284853947227e-06,
+      "loss": 0.7539,
+      "step": 1187
+    },
+    {
+      "epoch": 0.08248854325788085,
+      "grad_norm": 4.235326812692163,
+      "learning_rate": 9.928094959509347e-06,
+      "loss": 0.5374,
+      "step": 1188
+    },
+    {
+      "epoch": 0.08255797805860297,
+      "grad_norm": 5.28669080191623,
+      "learning_rate": 9.927904815813997e-06,
+      "loss": 0.6986,
+      "step": 1189
+    },
+    {
+      "epoch": 0.0826274128593251,
+      "grad_norm": 4.561071694807699,
+      "learning_rate": 9.927714422870802e-06,
+      "loss": 0.4995,
+      "step": 1190
+    },
+    {
+      "epoch": 0.08269684766004722,
+      "grad_norm": 4.234473595771569,
+      "learning_rate": 9.927523780689385e-06,
+      "loss": 0.3524,
+      "step": 1191
+    },
+    {
+      "epoch": 0.08276628246076934,
+      "grad_norm": 4.3233534021613265,
+      "learning_rate": 9.927332889279391e-06,
+      "loss": 0.6914,
+      "step": 1192
+    },
+    {
+      "epoch": 0.08283571726149146,
+      "grad_norm": 5.27325821916644,
+      "learning_rate": 9.927141748650478e-06,
+      "loss": 0.463,
+      "step": 1193
+    },
+    {
+      "epoch": 0.08290515206221358,
+      "grad_norm": 5.806077967321281,
+      "learning_rate": 9.926950358812308e-06,
+      "loss": 0.7533,
+      "step": 1194
+    },
+    {
+      "epoch": 0.0829745868629357,
+      "grad_norm": 5.159543358673594,
+      "learning_rate": 9.926758719774563e-06,
+      "loss": 0.7184,
+      "step": 1195
+    },
+    {
+      "epoch": 0.08304402166365782,
+      "grad_norm": 4.639061796688853,
+      "learning_rate": 9.92656683154694e-06,
+      "loss": 0.7285,
+      "step": 1196
+    },
+    {
+      "epoch": 0.08311345646437995,
+      "grad_norm": 4.32519579020212,
+      "learning_rate": 9.926374694139139e-06,
+      "loss": 0.4814,
+      "step": 1197
+    },
+    {
+      "epoch": 0.08318289126510207,
+      "grad_norm": 10.13789421276774,
+      "learning_rate": 9.926182307560881e-06,
+      "loss": 0.6947,
+      "step": 1198
+    },
+    {
+      "epoch": 0.0832523260658242,
+      "grad_norm": 5.295061374579348,
+      "learning_rate": 9.925989671821898e-06,
+      "loss": 0.6968,
+      "step": 1199
+    },
+    {
+      "epoch": 0.08332176086654632,
+      "grad_norm": 3.862797799062758,
+      "learning_rate": 9.92579678693193e-06,
+      "loss": 0.4694,
+      "step": 1200
+    },
+    {
+      "epoch": 0.08339119566726844,
+      "grad_norm": 4.7552534411439,
+      "learning_rate": 9.925603652900734e-06,
+      "loss": 0.3565,
+      "step": 1201
+    },
+    {
+      "epoch": 0.08346063046799056,
+      "grad_norm": 4.323811257444635,
+      "learning_rate": 9.92541026973808e-06,
+      "loss": 0.5622,
+      "step": 1202
+    },
+    {
+      "epoch": 0.08353006526871268,
+      "grad_norm": 6.211969349580078,
+      "learning_rate": 9.925216637453747e-06,
+      "loss": 0.888,
+      "step": 1203
+    },
+    {
+      "epoch": 0.0835995000694348,
+      "grad_norm": 6.270521654628784,
+      "learning_rate": 9.925022756057532e-06,
+      "loss": 0.7173,
+      "step": 1204
+    },
+    {
+      "epoch": 0.08366893487015692,
+      "grad_norm": 6.2410427118124,
+      "learning_rate": 9.924828625559238e-06,
+      "loss": 0.6177,
+      "step": 1205
+    },
+    {
+      "epoch": 0.08373836967087904,
+      "grad_norm": 5.295255252914425,
+      "learning_rate": 9.924634245968682e-06,
+      "loss": 0.563,
+      "step": 1206
+    },
+    {
+      "epoch": 0.08380780447160116,
+      "grad_norm": 5.005831086986683,
+      "learning_rate": 9.924439617295702e-06,
+      "loss": 0.4043,
+      "step": 1207
+    },
+    {
+      "epoch": 0.0838772392723233,
+      "grad_norm": 4.531180198877733,
+      "learning_rate": 9.924244739550137e-06,
+      "loss": 0.7454,
+      "step": 1208
+    },
+    {
+      "epoch": 0.08394667407304542,
+      "grad_norm": 5.572767558579299,
+      "learning_rate": 9.924049612741846e-06,
+      "loss": 0.5127,
+      "step": 1209
+    },
+    {
+      "epoch": 0.08401610887376754,
+      "grad_norm": 4.966498900014144,
+      "learning_rate": 9.923854236880698e-06,
+      "loss": 0.665,
+      "step": 1210
+    },
+    {
+      "epoch": 0.08408554367448966,
+      "grad_norm": 4.80091328092973,
+      "learning_rate": 9.923658611976573e-06,
+      "loss": 0.286,
+      "step": 1211
+    },
+    {
+      "epoch": 0.08415497847521178,
+      "grad_norm": 4.860568416232989,
+      "learning_rate": 9.923462738039368e-06,
+      "loss": 0.735,
+      "step": 1212
+    },
+    {
+      "epoch": 0.0842244132759339,
+      "grad_norm": 4.324215783686278,
+      "learning_rate": 9.923266615078988e-06,
+      "loss": 0.3958,
+      "step": 1213
+    },
+    {
+      "epoch": 0.08429384807665602,
+      "grad_norm": 4.912697561852342,
+      "learning_rate": 9.923070243105355e-06,
+      "loss": 0.4456,
+      "step": 1214
+    },
+    {
+      "epoch": 0.08436328287737814,
+      "grad_norm": 4.757053456316674,
+      "learning_rate": 9.922873622128399e-06,
+      "loss": 0.6899,
+      "step": 1215
+    },
+    {
+      "epoch": 0.08443271767810026,
+      "grad_norm": 5.518080019572764,
+      "learning_rate": 9.922676752158064e-06,
+      "loss": 0.7354,
+      "step": 1216
+    },
+    {
+      "epoch": 0.08450215247882238,
+      "grad_norm": 4.254881855936205,
+      "learning_rate": 9.922479633204312e-06,
+      "loss": 0.5505,
+      "step": 1217
+    },
+    {
+      "epoch": 0.0845715872795445,
+      "grad_norm": 4.579494108975009,
+      "learning_rate": 9.92228226527711e-06,
+      "loss": 0.4096,
+      "step": 1218
+    },
+    {
+      "epoch": 0.08464102208026662,
+      "grad_norm": 4.142602362561205,
+      "learning_rate": 9.92208464838644e-06,
+      "loss": 0.4461,
+      "step": 1219
+    },
+    {
+      "epoch": 0.08471045688098876,
+      "grad_norm": 4.15269701171399,
+      "learning_rate": 9.921886782542299e-06,
+      "loss": 0.4628,
+      "step": 1220
+    },
+    {
+      "epoch": 0.08477989168171088,
+      "grad_norm": 5.198638042777074,
+      "learning_rate": 9.921688667754694e-06,
+      "loss": 0.5251,
+      "step": 1221
+    },
+    {
+      "epoch": 0.084849326482433,
+      "grad_norm": 4.690806376901492,
+      "learning_rate": 9.921490304033646e-06,
+      "loss": 0.6629,
+      "step": 1222
+    },
+    {
+      "epoch": 0.08491876128315512,
+      "grad_norm": 4.604497296462114,
+      "learning_rate": 9.921291691389186e-06,
+      "loss": 0.4228,
+      "step": 1223
+    },
+    {
+      "epoch": 0.08498819608387724,
+      "grad_norm": 3.564469973159241,
+      "learning_rate": 9.92109282983136e-06,
+      "loss": 0.3682,
+      "step": 1224
+    },
+    {
+      "epoch": 0.08505763088459936,
+      "grad_norm": 4.320918600116759,
+      "learning_rate": 9.92089371937023e-06,
+      "loss": 0.3814,
+      "step": 1225
+    },
+    {
+      "epoch": 0.08512706568532148,
+      "grad_norm": 5.6883056591861045,
+      "learning_rate": 9.920694360015864e-06,
+      "loss": 0.5792,
+      "step": 1226
+    },
+    {
+      "epoch": 0.0851965004860436,
+      "grad_norm": 5.157495893261966,
+      "learning_rate": 9.920494751778344e-06,
+      "loss": 0.5797,
+      "step": 1227
+    },
+    {
+      "epoch": 0.08526593528676572,
+      "grad_norm": 3.9554364424016972,
+      "learning_rate": 9.920294894667767e-06,
+      "loss": 0.5692,
+      "step": 1228
+    },
+    {
+      "epoch": 0.08533537008748784,
+      "grad_norm": 4.85975284832723,
+      "learning_rate": 9.920094788694243e-06,
+      "loss": 0.6154,
+      "step": 1229
+    },
+    {
+      "epoch": 0.08540480488820996,
+      "grad_norm": 11.290487963439533,
+      "learning_rate": 9.919894433867892e-06,
+      "loss": 0.3245,
+      "step": 1230
+    },
+    {
+      "epoch": 0.0854742396889321,
+      "grad_norm": 5.025840201185536,
+      "learning_rate": 9.919693830198847e-06,
+      "loss": 0.4431,
+      "step": 1231
+    },
+    {
+      "epoch": 0.08554367448965422,
+      "grad_norm": 4.649628926064188,
+      "learning_rate": 9.919492977697256e-06,
+      "loss": 0.4839,
+      "step": 1232
+    },
+    {
+      "epoch": 0.08561310929037634,
+      "grad_norm": 3.909887107218836,
+      "learning_rate": 9.919291876373276e-06,
+      "loss": 0.3956,
+      "step": 1233
+    },
+    {
+      "epoch": 0.08568254409109846,
+      "grad_norm": 5.579771580702045,
+      "learning_rate": 9.91909052623708e-06,
+      "loss": 0.8172,
+      "step": 1234
+    },
+    {
+      "epoch": 0.08575197889182058,
+      "grad_norm": 4.25543454312069,
+      "learning_rate": 9.918888927298851e-06,
+      "loss": 0.5459,
+      "step": 1235
+    },
+    {
+      "epoch": 0.0858214136925427,
+      "grad_norm": 3.578453570913025,
+      "learning_rate": 9.918687079568788e-06,
+      "loss": 0.4142,
+      "step": 1236
+    },
+    {
+      "epoch": 0.08589084849326482,
+      "grad_norm": 5.0251449494192375,
+      "learning_rate": 9.918484983057095e-06,
+      "loss": 0.6628,
+      "step": 1237
+    },
+    {
+      "epoch": 0.08596028329398694,
+      "grad_norm": 4.000402175452813,
+      "learning_rate": 9.918282637774e-06,
+      "loss": 0.3298,
+      "step": 1238
+    },
+    {
+      "epoch": 0.08602971809470907,
+      "grad_norm": 5.565048302466547,
+      "learning_rate": 9.918080043729734e-06,
+      "loss": 0.5422,
+      "step": 1239
+    },
+    {
+      "epoch": 0.08609915289543119,
+      "grad_norm": 2.866303794001422,
+      "learning_rate": 9.917877200934544e-06,
+      "loss": 0.1285,
+      "step": 1240
+    },
+    {
+      "epoch": 0.0861685876961533,
+      "grad_norm": 3.2177502069660804,
+      "learning_rate": 9.917674109398692e-06,
+      "loss": 0.3212,
+      "step": 1241
+    },
+    {
+      "epoch": 0.08623802249687543,
+      "grad_norm": 4.29930280407711,
+      "learning_rate": 9.917470769132446e-06,
+      "loss": 0.396,
+      "step": 1242
+    },
+    {
+      "epoch": 0.08630745729759756,
+      "grad_norm": 3.834965187966803,
+      "learning_rate": 9.917267180146094e-06,
+      "loss": 0.3565,
+      "step": 1243
+    },
+    {
+      "epoch": 0.08637689209831968,
+      "grad_norm": 4.444698731519701,
+      "learning_rate": 9.917063342449931e-06,
+      "loss": 0.3961,
+      "step": 1244
+    },
+    {
+      "epoch": 0.0864463268990418,
+      "grad_norm": 5.439025184387969,
+      "learning_rate": 9.91685925605427e-06,
+      "loss": 0.7406,
+      "step": 1245
+    },
+    {
+      "epoch": 0.08651576169976392,
+      "grad_norm": 3.989220358625172,
+      "learning_rate": 9.91665492096943e-06,
+      "loss": 0.4845,
+      "step": 1246
+    },
+    {
+      "epoch": 0.08658519650048604,
+      "grad_norm": 3.943214901159205,
+      "learning_rate": 9.91645033720575e-06,
+      "loss": 0.4204,
+      "step": 1247
+    },
+    {
+      "epoch": 0.08665463130120817,
+      "grad_norm": 4.669685128962917,
+      "learning_rate": 9.916245504773572e-06,
+      "loss": 0.5363,
+      "step": 1248
+    },
+    {
+      "epoch": 0.08672406610193029,
+      "grad_norm": 3.7840896969144455,
+      "learning_rate": 9.916040423683262e-06,
+      "loss": 0.3577,
+      "step": 1249
+    },
+    {
+      "epoch": 0.0867935009026524,
+      "grad_norm": 4.9497348902264235,
+      "learning_rate": 9.91583509394519e-06,
+      "loss": 0.5895,
+      "step": 1250
+    },
+    {
+      "epoch": 0.08686293570337453,
+      "grad_norm": 4.606669931296728,
+      "learning_rate": 9.91562951556974e-06,
+      "loss": 0.3642,
+      "step": 1251
+    },
+    {
+      "epoch": 0.08693237050409665,
+      "grad_norm": 4.669192355729503,
+      "learning_rate": 9.915423688567314e-06,
+      "loss": 0.6443,
+      "step": 1252
+    },
+    {
+      "epoch": 0.08700180530481877,
+      "grad_norm": 5.237051267799775,
+      "learning_rate": 9.915217612948317e-06,
+      "loss": 0.4671,
+      "step": 1253
+    },
+    {
+      "epoch": 0.0870712401055409,
+      "grad_norm": 4.155042572238659,
+      "learning_rate": 9.915011288723178e-06,
+      "loss": 0.39,
+      "step": 1254
+    },
+    {
+      "epoch": 0.08714067490626302,
+      "grad_norm": 6.347686450090276,
+      "learning_rate": 9.914804715902328e-06,
+      "loss": 0.5465,
+      "step": 1255
+    },
+    {
+      "epoch": 0.08721010970698514,
+      "grad_norm": 4.959711362763308,
+      "learning_rate": 9.914597894496218e-06,
+      "loss": 0.8028,
+      "step": 1256
+    },
+    {
+      "epoch": 0.08727954450770727,
+      "grad_norm": 4.018123637413376,
+      "learning_rate": 9.914390824515307e-06,
+      "loss": 0.3598,
+      "step": 1257
+    },
+    {
+      "epoch": 0.08734897930842939,
+      "grad_norm": 4.2754949293590965,
+      "learning_rate": 9.91418350597007e-06,
+      "loss": 0.527,
+      "step": 1258
+    },
+    {
+      "epoch": 0.08741841410915151,
+      "grad_norm": 3.653540912737491,
+      "learning_rate": 9.91397593887099e-06,
+      "loss": 0.3612,
+      "step": 1259
+    },
+    {
+      "epoch": 0.08748784890987363,
+      "grad_norm": 5.655544159304154,
+      "learning_rate": 9.91376812322857e-06,
+      "loss": 0.6405,
+      "step": 1260
+    },
+    {
+      "epoch": 0.08755728371059575,
+      "grad_norm": 4.323575821022961,
+      "learning_rate": 9.913560059053318e-06,
+      "loss": 0.5584,
+      "step": 1261
+    },
+    {
+      "epoch": 0.08762671851131787,
+      "grad_norm": 4.641284609826128,
+      "learning_rate": 9.913351746355758e-06,
+      "loss": 0.4738,
+      "step": 1262
+    },
+    {
+      "epoch": 0.08769615331203999,
+      "grad_norm": 5.671778146857839,
+      "learning_rate": 9.913143185146426e-06,
+      "loss": 0.7733,
+      "step": 1263
+    },
+    {
+      "epoch": 0.08776558811276211,
+      "grad_norm": 4.17883334991228,
+      "learning_rate": 9.91293437543587e-06,
+      "loss": 0.5649,
+      "step": 1264
+    },
+    {
+      "epoch": 0.08783502291348423,
+      "grad_norm": 4.772101358865596,
+      "learning_rate": 9.912725317234655e-06,
+      "loss": 0.6759,
+      "step": 1265
+    },
+    {
+      "epoch": 0.08790445771420637,
+      "grad_norm": 3.574186501565113,
+      "learning_rate": 9.912516010553352e-06,
+      "loss": 0.4596,
+      "step": 1266
+    },
+    {
+      "epoch": 0.08797389251492849,
+      "grad_norm": 3.0729789014084137,
+      "learning_rate": 9.912306455402548e-06,
+      "loss": 0.1449,
+      "step": 1267
+    },
+    {
+      "epoch": 0.08804332731565061,
+      "grad_norm": 4.594586914437889,
+      "learning_rate": 9.912096651792842e-06,
+      "loss": 0.4047,
+      "step": 1268
+    },
+    {
+      "epoch": 0.08811276211637273,
+      "grad_norm": 4.148315010228122,
+      "learning_rate": 9.911886599734845e-06,
+      "loss": 0.5129,
+      "step": 1269
+    },
+    {
+      "epoch": 0.08818219691709485,
+      "grad_norm": 4.118703464794777,
+      "learning_rate": 9.911676299239184e-06,
+      "loss": 0.326,
+      "step": 1270
+    },
+    {
+      "epoch": 0.08825163171781697,
+      "grad_norm": 4.939462868697017,
+      "learning_rate": 9.911465750316492e-06,
+      "loss": 0.6523,
+      "step": 1271
+    },
+    {
+      "epoch": 0.08832106651853909,
+      "grad_norm": 4.493972667108219,
+      "learning_rate": 9.91125495297742e-06,
+      "loss": 0.3893,
+      "step": 1272
+    },
+    {
+      "epoch": 0.08839050131926121,
+      "grad_norm": 5.53983544924788,
+      "learning_rate": 9.91104390723263e-06,
+      "loss": 0.6655,
+      "step": 1273
+    },
+    {
+      "epoch": 0.08845993611998333,
+      "grad_norm": 5.434130743391432,
+      "learning_rate": 9.910832613092797e-06,
+      "loss": 0.6191,
+      "step": 1274
+    },
+    {
+      "epoch": 0.08852937092070545,
+      "grad_norm": 4.343286838992886,
+      "learning_rate": 9.910621070568607e-06,
+      "loss": 0.337,
+      "step": 1275
+    },
+    {
+      "epoch": 0.08859880572142757,
+      "grad_norm": 4.020449921602951,
+      "learning_rate": 9.91040927967076e-06,
+      "loss": 0.4356,
+      "step": 1276
+    },
+    {
+      "epoch": 0.08866824052214971,
+      "grad_norm": 4.336008748857587,
+      "learning_rate": 9.910197240409969e-06,
+      "loss": 0.5097,
+      "step": 1277
+    },
+    {
+      "epoch": 0.08873767532287183,
+      "grad_norm": 5.388843466595745,
+      "learning_rate": 9.909984952796959e-06,
+      "loss": 0.8222,
+      "step": 1278
+    },
+    {
+      "epoch": 0.08880711012359395,
+      "grad_norm": 4.074947137892525,
+      "learning_rate": 9.909772416842463e-06,
+      "loss": 0.4551,
+      "step": 1279
+    },
+    {
+      "epoch": 0.08887654492431607,
+      "grad_norm": 4.69379941576267,
+      "learning_rate": 9.909559632557236e-06,
+      "loss": 0.3801,
+      "step": 1280
+    },
+    {
+      "epoch": 0.08894597972503819,
+      "grad_norm": 4.753870459916141,
+      "learning_rate": 9.909346599952037e-06,
+      "loss": 0.7037,
+      "step": 1281
+    },
+    {
+      "epoch": 0.08901541452576031,
+      "grad_norm": 4.650182234514726,
+      "learning_rate": 9.909133319037643e-06,
+      "loss": 0.5504,
+      "step": 1282
+    },
+    {
+      "epoch": 0.08908484932648243,
+      "grad_norm": 4.184575958201373,
+      "learning_rate": 9.90891978982484e-06,
+      "loss": 0.5578,
+      "step": 1283
+    },
+    {
+      "epoch": 0.08915428412720455,
+      "grad_norm": 4.795798844662197,
+      "learning_rate": 9.908706012324431e-06,
+      "loss": 0.7351,
+      "step": 1284
+    },
+    {
+      "epoch": 0.08922371892792667,
+      "grad_norm": 4.285945627437185,
+      "learning_rate": 9.908491986547224e-06,
+      "loss": 0.5288,
+      "step": 1285
+    },
+    {
+      "epoch": 0.0892931537286488,
+      "grad_norm": 5.304631654318846,
+      "learning_rate": 9.908277712504049e-06,
+      "loss": 0.6029,
+      "step": 1286
+    },
+    {
+      "epoch": 0.08936258852937091,
+      "grad_norm": 3.6115709814775716,
+      "learning_rate": 9.908063190205739e-06,
+      "loss": 0.4159,
+      "step": 1287
+    },
+    {
+      "epoch": 0.08943202333009305,
+      "grad_norm": 6.516016620520026,
+      "learning_rate": 9.907848419663148e-06,
+      "loss": 0.5863,
+      "step": 1288
+    },
+    {
+      "epoch": 0.08950145813081517,
+      "grad_norm": 4.480860582873414,
+      "learning_rate": 9.907633400887138e-06,
+      "loss": 0.5386,
+      "step": 1289
+    },
+    {
+      "epoch": 0.08957089293153729,
+      "grad_norm": 4.394487387306154,
+      "learning_rate": 9.907418133888582e-06,
+      "loss": 0.5631,
+      "step": 1290
+    },
+    {
+      "epoch": 0.08964032773225941,
+      "grad_norm": 5.105504320990119,
+      "learning_rate": 9.907202618678371e-06,
+      "loss": 0.5023,
+      "step": 1291
+    },
+    {
+      "epoch": 0.08970976253298153,
+      "grad_norm": 4.9202791224563205,
+      "learning_rate": 9.906986855267405e-06,
+      "loss": 0.7332,
+      "step": 1292
+    },
+    {
+      "epoch": 0.08977919733370365,
+      "grad_norm": 4.536169532580502,
+      "learning_rate": 9.906770843666595e-06,
+      "loss": 0.4669,
+      "step": 1293
+    },
+    {
+      "epoch": 0.08984863213442577,
+      "grad_norm": 3.5264698946332196,
+      "learning_rate": 9.90655458388687e-06,
+      "loss": 0.4169,
+      "step": 1294
+    },
+    {
+      "epoch": 0.0899180669351479,
+      "grad_norm": 5.059279789467382,
+      "learning_rate": 9.906338075939166e-06,
+      "loss": 0.5981,
+      "step": 1295
+    },
+    {
+      "epoch": 0.08998750173587002,
+      "grad_norm": 4.496010782071867,
+      "learning_rate": 9.906121319834432e-06,
+      "loss": 0.6819,
+      "step": 1296
+    },
+    {
+      "epoch": 0.09005693653659214,
+      "grad_norm": 4.399011002827967,
+      "learning_rate": 9.905904315583636e-06,
+      "loss": 0.6472,
+      "step": 1297
+    },
+    {
+      "epoch": 0.09012637133731426,
+      "grad_norm": 4.527553330493005,
+      "learning_rate": 9.90568706319775e-06,
+      "loss": 0.4522,
+      "step": 1298
+    },
+    {
+      "epoch": 0.09019580613803638,
+      "grad_norm": 4.879407890269408,
+      "learning_rate": 9.905469562687763e-06,
+      "loss": 0.7609,
+      "step": 1299
+    },
+    {
+      "epoch": 0.09026524093875851,
+      "grad_norm": 3.866646781536308,
+      "learning_rate": 9.905251814064677e-06,
+      "loss": 0.372,
+      "step": 1300
+    },
+    {
+      "epoch": 0.09033467573948063,
+      "grad_norm": 3.8932277886797735,
+      "learning_rate": 9.905033817339504e-06,
+      "loss": 0.5858,
+      "step": 1301
+    },
+    {
+      "epoch": 0.09040411054020275,
+      "grad_norm": 5.20472257612294,
+      "learning_rate": 9.904815572523272e-06,
+      "loss": 0.3693,
+      "step": 1302
+    },
+    {
+      "epoch": 0.09047354534092487,
+      "grad_norm": 3.6715186269153683,
+      "learning_rate": 9.904597079627018e-06,
+      "loss": 0.5368,
+      "step": 1303
+    },
+    {
+      "epoch": 0.090542980141647,
+      "grad_norm": 3.115208846097869,
+      "learning_rate": 9.904378338661795e-06,
+      "loss": 0.1586,
+      "step": 1304
+    },
+    {
+      "epoch": 0.09061241494236912,
+      "grad_norm": 4.457582419352622,
+      "learning_rate": 9.904159349638665e-06,
+      "loss": 0.4126,
+      "step": 1305
+    },
+    {
+      "epoch": 0.09068184974309124,
+      "grad_norm": 4.636940100048691,
+      "learning_rate": 9.903940112568701e-06,
+      "loss": 0.7778,
+      "step": 1306
+    },
+    {
+      "epoch": 0.09075128454381336,
+      "grad_norm": 5.386520201347107,
+      "learning_rate": 9.903720627463e-06,
+      "loss": 0.8934,
+      "step": 1307
+    },
+    {
+      "epoch": 0.09082071934453548,
+      "grad_norm": 4.660555687696774,
+      "learning_rate": 9.903500894332655e-06,
+      "loss": 0.5957,
+      "step": 1308
+    },
+    {
+      "epoch": 0.0908901541452576,
+      "grad_norm": 4.751308921371486,
+      "learning_rate": 9.903280913188786e-06,
+      "loss": 0.479,
+      "step": 1309
+    },
+    {
+      "epoch": 0.09095958894597972,
+      "grad_norm": 5.343201001230028,
+      "learning_rate": 9.903060684042516e-06,
+      "loss": 0.8311,
+      "step": 1310
+    },
+    {
+      "epoch": 0.09102902374670185,
+      "grad_norm": 5.4340435221876024,
+      "learning_rate": 9.902840206904984e-06,
+      "loss": 0.8404,
+      "step": 1311
+    },
+    {
+      "epoch": 0.09109845854742397,
+      "grad_norm": 3.4760932067568318,
+      "learning_rate": 9.902619481787344e-06,
+      "loss": 0.2382,
+      "step": 1312
+    },
+    {
+      "epoch": 0.0911678933481461,
+      "grad_norm": 4.283967096608612,
+      "learning_rate": 9.902398508700757e-06,
+      "loss": 0.4666,
+      "step": 1313
+    },
+    {
+      "epoch": 0.09123732814886822,
+      "grad_norm": 4.4100162091719675,
+      "learning_rate": 9.9021772876564e-06,
+      "loss": 0.6359,
+      "step": 1314
+    },
+    {
+      "epoch": 0.09130676294959034,
+      "grad_norm": 4.335130693770736,
+      "learning_rate": 9.901955818665464e-06,
+      "loss": 0.6145,
+      "step": 1315
+    },
+    {
+      "epoch": 0.09137619775031246,
+      "grad_norm": 4.093762037805684,
+      "learning_rate": 9.901734101739152e-06,
+      "loss": 0.519,
+      "step": 1316
+    },
+    {
+      "epoch": 0.09144563255103458,
+      "grad_norm": 4.725196910642091,
+      "learning_rate": 9.901512136888672e-06,
+      "loss": 0.5798,
+      "step": 1317
+    },
+    {
+      "epoch": 0.0915150673517567,
+      "grad_norm": 4.5440338595334975,
+      "learning_rate": 9.901289924125257e-06,
+      "loss": 0.5259,
+      "step": 1318
+    },
+    {
+      "epoch": 0.09158450215247882,
+      "grad_norm": 4.658909332328787,
+      "learning_rate": 9.901067463460142e-06,
+      "loss": 0.5461,
+      "step": 1319
+    },
+    {
+      "epoch": 0.09165393695320094,
+      "grad_norm": 4.158419941967739,
+      "learning_rate": 9.900844754904581e-06,
+      "loss": 0.4688,
+      "step": 1320
+    },
+    {
+      "epoch": 0.09172337175392306,
+      "grad_norm": 4.88658882490353,
+      "learning_rate": 9.90062179846984e-06,
+      "loss": 0.7088,
+      "step": 1321
+    },
+    {
+      "epoch": 0.09179280655464518,
+      "grad_norm": 2.9143450946453284,
+      "learning_rate": 9.90039859416719e-06,
+      "loss": 0.2995,
+      "step": 1322
+    },
+    {
+      "epoch": 0.09186224135536732,
+      "grad_norm": 3.943203609196513,
+      "learning_rate": 9.900175142007929e-06,
+      "loss": 0.4954,
+      "step": 1323
+    },
+    {
+      "epoch": 0.09193167615608944,
+      "grad_norm": 4.47572314076152,
+      "learning_rate": 9.89995144200335e-06,
+      "loss": 0.7639,
+      "step": 1324
+    },
+    {
+      "epoch": 0.09200111095681156,
+      "grad_norm": 4.969523424797767,
+      "learning_rate": 9.899727494164774e-06,
+      "loss": 0.7059,
+      "step": 1325
+    },
+    {
+      "epoch": 0.09207054575753368,
+      "grad_norm": 4.120087734712923,
+      "learning_rate": 9.899503298503525e-06,
+      "loss": 0.3763,
+      "step": 1326
+    },
+    {
+      "epoch": 0.0921399805582558,
+      "grad_norm": 5.067988309702797,
+      "learning_rate": 9.899278855030944e-06,
+      "loss": 0.585,
+      "step": 1327
+    },
+    {
+      "epoch": 0.09220941535897792,
+      "grad_norm": 4.039966234840161,
+      "learning_rate": 9.899054163758383e-06,
+      "loss": 0.4453,
+      "step": 1328
+    },
+    {
+      "epoch": 0.09227885015970004,
+      "grad_norm": 4.624625701293212,
+      "learning_rate": 9.898829224697206e-06,
+      "loss": 0.5645,
+      "step": 1329
+    },
+    {
+      "epoch": 0.09234828496042216,
+      "grad_norm": 5.177213800487478,
+      "learning_rate": 9.89860403785879e-06,
+      "loss": 0.605,
+      "step": 1330
+    },
+    {
+      "epoch": 0.09241771976114428,
+      "grad_norm": 4.658634603397049,
+      "learning_rate": 9.898378603254524e-06,
+      "loss": 0.6117,
+      "step": 1331
+    },
+    {
+      "epoch": 0.0924871545618664,
+      "grad_norm": 4.666195866979545,
+      "learning_rate": 9.898152920895811e-06,
+      "loss": 0.412,
+      "step": 1332
+    },
+    {
+      "epoch": 0.09255658936258852,
+      "grad_norm": 4.376930942206212,
+      "learning_rate": 9.897926990794066e-06,
+      "loss": 0.4629,
+      "step": 1333
+    },
+    {
+      "epoch": 0.09262602416331066,
+      "grad_norm": 4.855977029944248,
+      "learning_rate": 9.897700812960717e-06,
+      "loss": 0.5679,
+      "step": 1334
+    },
+    {
+      "epoch": 0.09269545896403278,
+      "grad_norm": 6.169285730188878,
+      "learning_rate": 9.897474387407202e-06,
+      "loss": 0.7218,
+      "step": 1335
+    },
+    {
+      "epoch": 0.0927648937647549,
+      "grad_norm": 5.178209339476342,
+      "learning_rate": 9.897247714144975e-06,
+      "loss": 0.5485,
+      "step": 1336
+    },
+    {
+      "epoch": 0.09283432856547702,
+      "grad_norm": 4.985863476100043,
+      "learning_rate": 9.897020793185501e-06,
+      "loss": 0.5302,
+      "step": 1337
+    },
+    {
+      "epoch": 0.09290376336619914,
+      "grad_norm": 3.7767209090906046,
+      "learning_rate": 9.896793624540256e-06,
+      "loss": 0.2786,
+      "step": 1338
+    },
+    {
+      "epoch": 0.09297319816692126,
+      "grad_norm": 5.220556557392699,
+      "learning_rate": 9.896566208220729e-06,
+      "loss": 0.5117,
+      "step": 1339
+    },
+    {
+      "epoch": 0.09304263296764338,
+      "grad_norm": 3.741201155047686,
+      "learning_rate": 9.896338544238426e-06,
+      "loss": 0.3969,
+      "step": 1340
+    },
+    {
+      "epoch": 0.0931120677683655,
+      "grad_norm": 4.447324193413887,
+      "learning_rate": 9.896110632604858e-06,
+      "loss": 0.3914,
+      "step": 1341
+    },
+    {
+      "epoch": 0.09318150256908762,
+      "grad_norm": 5.00842964667195,
+      "learning_rate": 9.895882473331557e-06,
+      "loss": 0.7516,
+      "step": 1342
+    },
+    {
+      "epoch": 0.09325093736980974,
+      "grad_norm": 4.3887508796747765,
+      "learning_rate": 9.89565406643006e-06,
+      "loss": 0.5764,
+      "step": 1343
+    },
+    {
+      "epoch": 0.09332037217053187,
+      "grad_norm": 4.471200980128816,
+      "learning_rate": 9.89542541191192e-06,
+      "loss": 0.5304,
+      "step": 1344
+    },
+    {
+      "epoch": 0.09338980697125399,
+      "grad_norm": 4.497355139791726,
+      "learning_rate": 9.8951965097887e-06,
+      "loss": 0.687,
+      "step": 1345
+    },
+    {
+      "epoch": 0.09345924177197612,
+      "grad_norm": 4.617979234372066,
+      "learning_rate": 9.894967360071982e-06,
+      "loss": 0.481,
+      "step": 1346
+    },
+    {
+      "epoch": 0.09352867657269824,
+      "grad_norm": 5.369970066031287,
+      "learning_rate": 9.894737962773354e-06,
+      "loss": 0.6619,
+      "step": 1347
+    },
+    {
+      "epoch": 0.09359811137342036,
+      "grad_norm": 4.226961802603493,
+      "learning_rate": 9.894508317904418e-06,
+      "loss": 0.5354,
+      "step": 1348
+    },
+    {
+      "epoch": 0.09366754617414248,
+      "grad_norm": 4.946606876667253,
+      "learning_rate": 9.89427842547679e-06,
+      "loss": 0.7071,
+      "step": 1349
+    },
+    {
+      "epoch": 0.0937369809748646,
+      "grad_norm": 5.5561916504720275,
+      "learning_rate": 9.894048285502098e-06,
+      "loss": 0.6732,
+      "step": 1350
+    },
+    {
+      "epoch": 0.09380641577558672,
+      "grad_norm": 6.913186391522856,
+      "learning_rate": 9.89381789799198e-06,
+      "loss": 0.7755,
+      "step": 1351
+    },
+    {
+      "epoch": 0.09387585057630884,
+      "grad_norm": 3.7737207055580773,
+      "learning_rate": 9.893587262958093e-06,
+      "loss": 0.379,
+      "step": 1352
+    },
+    {
+      "epoch": 0.09394528537703097,
+      "grad_norm": 4.624893208379848,
+      "learning_rate": 9.893356380412097e-06,
+      "loss": 0.5548,
+      "step": 1353
+    },
+    {
+      "epoch": 0.09401472017775309,
+      "grad_norm": 5.159897332703197,
+      "learning_rate": 9.893125250365674e-06,
+      "loss": 0.6081,
+      "step": 1354
+    },
+    {
+      "epoch": 0.0940841549784752,
+      "grad_norm": 5.173391268751992,
+      "learning_rate": 9.892893872830514e-06,
+      "loss": 0.5876,
+      "step": 1355
+    },
+    {
+      "epoch": 0.09415358977919733,
+      "grad_norm": 3.649056165755258,
+      "learning_rate": 9.892662247818317e-06,
+      "loss": 0.4209,
+      "step": 1356
+    },
+    {
+      "epoch": 0.09422302457991946,
+      "grad_norm": 4.9779515279684645,
+      "learning_rate": 9.8924303753408e-06,
+      "loss": 0.8236,
+      "step": 1357
+    },
+    {
+      "epoch": 0.09429245938064158,
+      "grad_norm": 5.157859649265562,
+      "learning_rate": 9.892198255409691e-06,
+      "loss": 0.4799,
+      "step": 1358
+    },
+    {
+      "epoch": 0.0943618941813637,
+      "grad_norm": 3.8127157576257775,
+      "learning_rate": 9.89196588803673e-06,
+      "loss": 0.421,
+      "step": 1359
+    },
+    {
+      "epoch": 0.09443132898208582,
+      "grad_norm": 3.2938190476061187,
+      "learning_rate": 9.89173327323367e-06,
+      "loss": 0.4079,
+      "step": 1360
+    },
+    {
+      "epoch": 0.09450076378280794,
+      "grad_norm": 4.099457305916928,
+      "learning_rate": 9.891500411012275e-06,
+      "loss": 0.293,
+      "step": 1361
+    },
+    {
+      "epoch": 0.09457019858353007,
+      "grad_norm": 4.723280762128105,
+      "learning_rate": 9.891267301384327e-06,
+      "loss": 0.692,
+      "step": 1362
+    },
+    {
+      "epoch": 0.09463963338425219,
+      "grad_norm": 4.431120674952316,
+      "learning_rate": 9.891033944361613e-06,
+      "loss": 0.4724,
+      "step": 1363
+    },
+    {
+      "epoch": 0.09470906818497431,
+      "grad_norm": 5.292143597371503,
+      "learning_rate": 9.890800339955935e-06,
+      "loss": 0.7446,
+      "step": 1364
+    },
+    {
+      "epoch": 0.09477850298569643,
+      "grad_norm": 4.322797194365144,
+      "learning_rate": 9.890566488179111e-06,
+      "loss": 0.2078,
+      "step": 1365
+    },
+    {
+      "epoch": 0.09484793778641855,
+      "grad_norm": 3.675068748583708,
+      "learning_rate": 9.89033238904297e-06,
+      "loss": 0.4216,
+      "step": 1366
+    },
+    {
+      "epoch": 0.09491737258714067,
+      "grad_norm": 4.107590015210328,
+      "learning_rate": 9.890098042559348e-06,
+      "loss": 0.4276,
+      "step": 1367
+    },
+    {
+      "epoch": 0.09498680738786279,
+      "grad_norm": 4.635964410226979,
+      "learning_rate": 9.889863448740102e-06,
+      "loss": 0.475,
+      "step": 1368
+    },
+    {
+      "epoch": 0.09505624218858492,
+      "grad_norm": 3.7456902203348283,
+      "learning_rate": 9.889628607597095e-06,
+      "loss": 0.3446,
+      "step": 1369
+    },
+    {
+      "epoch": 0.09512567698930705,
+      "grad_norm": 4.552400080916885,
+      "learning_rate": 9.889393519142207e-06,
+      "loss": 0.6254,
+      "step": 1370
+    },
+    {
+      "epoch": 0.09519511179002917,
+      "grad_norm": 4.038329545370405,
+      "learning_rate": 9.889158183387326e-06,
+      "loss": 0.5735,
+      "step": 1371
+    },
+    {
+      "epoch": 0.09526454659075129,
+      "grad_norm": 5.414999401779139,
+      "learning_rate": 9.888922600344359e-06,
+      "loss": 0.6119,
+      "step": 1372
+    },
+    {
+      "epoch": 0.09533398139147341,
+      "grad_norm": 4.7755519402171585,
+      "learning_rate": 9.888686770025217e-06,
+      "loss": 0.4937,
+      "step": 1373
+    },
+    {
+      "epoch": 0.09540341619219553,
+      "grad_norm": 4.206937521522721,
+      "learning_rate": 9.888450692441832e-06,
+      "loss": 0.3964,
+      "step": 1374
+    },
+    {
+      "epoch": 0.09547285099291765,
+      "grad_norm": 4.089789260062656,
+      "learning_rate": 9.88821436760614e-06,
+      "loss": 0.6167,
+      "step": 1375
+    },
+    {
+      "epoch": 0.09554228579363977,
+      "grad_norm": 4.620450270970714,
+      "learning_rate": 9.8879777955301e-06,
+      "loss": 0.5489,
+      "step": 1376
+    },
+    {
+      "epoch": 0.09561172059436189,
+      "grad_norm": 5.325258261901468,
+      "learning_rate": 9.88774097622567e-06,
+      "loss": 0.5987,
+      "step": 1377
+    },
+    {
+      "epoch": 0.09568115539508401,
+      "grad_norm": 3.705544240782535,
+      "learning_rate": 9.887503909704835e-06,
+      "loss": 0.3993,
+      "step": 1378
+    },
+    {
+      "epoch": 0.09575059019580613,
+      "grad_norm": 4.283317732522032,
+      "learning_rate": 9.887266595979583e-06,
+      "loss": 0.5186,
+      "step": 1379
+    },
+    {
+      "epoch": 0.09582002499652827,
+      "grad_norm": 5.645905596427402,
+      "learning_rate": 9.887029035061915e-06,
+      "loss": 0.8449,
+      "step": 1380
+    },
+    {
+      "epoch": 0.09588945979725039,
+      "grad_norm": 5.10412900196531,
+      "learning_rate": 9.88679122696385e-06,
+      "loss": 0.7481,
+      "step": 1381
+    },
+    {
+      "epoch": 0.09595889459797251,
+      "grad_norm": 3.948881952467173,
+      "learning_rate": 9.886553171697415e-06,
+      "loss": 0.4464,
+      "step": 1382
+    },
+    {
+      "epoch": 0.09602832939869463,
+      "grad_norm": 3.9647976545727763,
+      "learning_rate": 9.886314869274649e-06,
+      "loss": 0.3703,
+      "step": 1383
+    },
+    {
+      "epoch": 0.09609776419941675,
+      "grad_norm": 4.904702878550367,
+      "learning_rate": 9.886076319707607e-06,
+      "loss": 0.5627,
+      "step": 1384
+    },
+    {
+      "epoch": 0.09616719900013887,
+      "grad_norm": 6.458986495164821,
+      "learning_rate": 9.885837523008353e-06,
+      "loss": 0.7612,
+      "step": 1385
+    },
+    {
+      "epoch": 0.09623663380086099,
+      "grad_norm": 4.737654809639494,
+      "learning_rate": 9.885598479188966e-06,
+      "loss": 0.4233,
+      "step": 1386
+    },
+    {
+      "epoch": 0.09630606860158311,
+      "grad_norm": 4.793801421196134,
+      "learning_rate": 9.885359188261536e-06,
+      "loss": 0.523,
+      "step": 1387
+    },
+    {
+      "epoch": 0.09637550340230523,
+      "grad_norm": 4.773298687718501,
+      "learning_rate": 9.885119650238167e-06,
+      "loss": 0.4104,
+      "step": 1388
+    },
+    {
+      "epoch": 0.09644493820302735,
+      "grad_norm": 4.770502384270093,
+      "learning_rate": 9.884879865130973e-06,
+      "loss": 0.6155,
+      "step": 1389
+    },
+    {
+      "epoch": 0.09651437300374947,
+      "grad_norm": 4.365994878344466,
+      "learning_rate": 9.884639832952085e-06,
+      "loss": 0.5617,
+      "step": 1390
+    },
+    {
+      "epoch": 0.09658380780447161,
+      "grad_norm": 5.233463610741223,
+      "learning_rate": 9.884399553713639e-06,
+      "loss": 0.6585,
+      "step": 1391
+    },
+    {
+      "epoch": 0.09665324260519373,
+      "grad_norm": 4.2868594230134285,
+      "learning_rate": 9.884159027427793e-06,
+      "loss": 0.5836,
+      "step": 1392
+    },
+    {
+      "epoch": 0.09672267740591585,
+      "grad_norm": 5.06131148639139,
+      "learning_rate": 9.883918254106709e-06,
+      "loss": 0.5357,
+      "step": 1393
+    },
+    {
+      "epoch": 0.09679211220663797,
+      "grad_norm": 4.168288755435476,
+      "learning_rate": 9.883677233762567e-06,
+      "loss": 0.4407,
+      "step": 1394
+    },
+    {
+      "epoch": 0.09686154700736009,
+      "grad_norm": 4.661055665697704,
+      "learning_rate": 9.883435966407555e-06,
+      "loss": 0.557,
+      "step": 1395
+    },
+    {
+      "epoch": 0.09693098180808221,
+      "grad_norm": 5.102820698152733,
+      "learning_rate": 9.88319445205388e-06,
+      "loss": 0.7647,
+      "step": 1396
+    },
+    {
+      "epoch": 0.09700041660880433,
+      "grad_norm": 4.311361601834448,
+      "learning_rate": 9.882952690713754e-06,
+      "loss": 0.646,
+      "step": 1397
+    },
+    {
+      "epoch": 0.09706985140952645,
+      "grad_norm": 5.199056766179924,
+      "learning_rate": 9.882710682399407e-06,
+      "loss": 0.3538,
+      "step": 1398
+    },
+    {
+      "epoch": 0.09713928621024857,
+      "grad_norm": 5.567304990121936,
+      "learning_rate": 9.882468427123079e-06,
+      "loss": 0.673,
+      "step": 1399
+    },
+    {
+      "epoch": 0.0972087210109707,
+      "grad_norm": 4.5797556508897665,
+      "learning_rate": 9.882225924897022e-06,
+      "loss": 0.5623,
+      "step": 1400
+    },
+    {
+      "epoch": 0.09727815581169282,
+      "grad_norm": 4.653949428324637,
+      "learning_rate": 9.881983175733503e-06,
+      "loss": 0.4494,
+      "step": 1401
+    },
+    {
+      "epoch": 0.09734759061241494,
+      "grad_norm": 5.009193093287199,
+      "learning_rate": 9.881740179644799e-06,
+      "loss": 0.5073,
+      "step": 1402
+    },
+    {
+      "epoch": 0.09741702541313707,
+      "grad_norm": 5.101340255142069,
+      "learning_rate": 9.8814969366432e-06,
+      "loss": 0.6643,
+      "step": 1403
+    },
+    {
+      "epoch": 0.09748646021385919,
+      "grad_norm": 5.522477413238814,
+      "learning_rate": 9.88125344674101e-06,
+      "loss": 0.7866,
+      "step": 1404
+    },
+    {
+      "epoch": 0.09755589501458131,
+      "grad_norm": 6.1313161603201785,
+      "learning_rate": 9.881009709950547e-06,
+      "loss": 0.9623,
+      "step": 1405
+    },
+    {
+      "epoch": 0.09762532981530343,
+      "grad_norm": 5.378298622525633,
+      "learning_rate": 9.880765726284133e-06,
+      "loss": 0.661,
+      "step": 1406
+    },
+    {
+      "epoch": 0.09769476461602555,
+      "grad_norm": 4.430002476911952,
+      "learning_rate": 9.880521495754112e-06,
+      "loss": 0.4735,
+      "step": 1407
+    },
+    {
+      "epoch": 0.09776419941674767,
+      "grad_norm": 4.837863791741688,
+      "learning_rate": 9.880277018372837e-06,
+      "loss": 0.4609,
+      "step": 1408
+    },
+    {
+      "epoch": 0.0978336342174698,
+      "grad_norm": 5.2930098943589945,
+      "learning_rate": 9.880032294152673e-06,
+      "loss": 0.6534,
+      "step": 1409
+    },
+    {
+      "epoch": 0.09790306901819192,
+      "grad_norm": 3.9025667504026536,
+      "learning_rate": 9.879787323105996e-06,
+      "loss": 0.3865,
+      "step": 1410
+    },
+    {
+      "epoch": 0.09797250381891404,
+      "grad_norm": 4.257170339438434,
+      "learning_rate": 9.8795421052452e-06,
+      "loss": 0.3296,
+      "step": 1411
+    },
+    {
+      "epoch": 0.09804193861963616,
+      "grad_norm": 5.111117920768885,
+      "learning_rate": 9.879296640582686e-06,
+      "loss": 0.6103,
+      "step": 1412
+    },
+    {
+      "epoch": 0.09811137342035828,
+      "grad_norm": 5.301777133455815,
+      "learning_rate": 9.879050929130867e-06,
+      "loss": 0.6157,
+      "step": 1413
+    },
+    {
+      "epoch": 0.09818080822108041,
+      "grad_norm": 4.960311362655914,
+      "learning_rate": 9.878804970902175e-06,
+      "loss": 0.5044,
+      "step": 1414
+    },
+    {
+      "epoch": 0.09825024302180253,
+      "grad_norm": 4.569294337625017,
+      "learning_rate": 9.878558765909048e-06,
+      "loss": 0.3892,
+      "step": 1415
+    },
+    {
+      "epoch": 0.09831967782252465,
+      "grad_norm": 5.7908651558011135,
+      "learning_rate": 9.878312314163938e-06,
+      "loss": 0.7408,
+      "step": 1416
+    },
+    {
+      "epoch": 0.09838911262324677,
+      "grad_norm": 4.66316644897711,
+      "learning_rate": 9.878065615679313e-06,
+      "loss": 0.669,
+      "step": 1417
+    },
+    {
+      "epoch": 0.0984585474239689,
+      "grad_norm": 4.575384290322658,
+      "learning_rate": 9.877818670467647e-06,
+      "loss": 0.4646,
+      "step": 1418
+    },
+    {
+      "epoch": 0.09852798222469102,
+      "grad_norm": 3.795580118089873,
+      "learning_rate": 9.877571478541434e-06,
+      "loss": 0.4155,
+      "step": 1419
+    },
+    {
+      "epoch": 0.09859741702541314,
+      "grad_norm": 4.828266354344776,
+      "learning_rate": 9.877324039913173e-06,
+      "loss": 0.616,
+      "step": 1420
+    },
+    {
+      "epoch": 0.09866685182613526,
+      "grad_norm": 5.198681724913463,
+      "learning_rate": 9.877076354595385e-06,
+      "loss": 0.6337,
+      "step": 1421
+    },
+    {
+      "epoch": 0.09873628662685738,
+      "grad_norm": 3.349218705869498,
+      "learning_rate": 9.87682842260059e-06,
+      "loss": 0.3862,
+      "step": 1422
+    },
+    {
+      "epoch": 0.0988057214275795,
+      "grad_norm": 5.360275399041558,
+      "learning_rate": 9.876580243941333e-06,
+      "loss": 0.7028,
+      "step": 1423
+    },
+    {
+      "epoch": 0.09887515622830162,
+      "grad_norm": 3.295308978859412,
+      "learning_rate": 9.876331818630168e-06,
+      "loss": 0.5054,
+      "step": 1424
+    },
+    {
+      "epoch": 0.09894459102902374,
+      "grad_norm": 5.354463514674654,
+      "learning_rate": 9.876083146679654e-06,
+      "loss": 0.8394,
+      "step": 1425
+    },
+    {
+      "epoch": 0.09901402582974587,
+      "grad_norm": 3.9354718132436743,
+      "learning_rate": 9.875834228102374e-06,
+      "loss": 0.6806,
+      "step": 1426
+    },
+    {
+      "epoch": 0.099083460630468,
+      "grad_norm": 2.9775044809519655,
+      "learning_rate": 9.875585062910916e-06,
+      "loss": 0.3313,
+      "step": 1427
+    },
+    {
+      "epoch": 0.09915289543119012,
+      "grad_norm": 5.914369367608152,
+      "learning_rate": 9.875335651117882e-06,
+      "loss": 0.4899,
+      "step": 1428
+    },
+    {
+      "epoch": 0.09922233023191224,
+      "grad_norm": 4.651429212607517,
+      "learning_rate": 9.875085992735886e-06,
+      "loss": 0.689,
+      "step": 1429
+    },
+    {
+      "epoch": 0.09929176503263436,
+      "grad_norm": 4.071094062518788,
+      "learning_rate": 9.87483608777756e-06,
+      "loss": 0.2934,
+      "step": 1430
+    },
+    {
+      "epoch": 0.09936119983335648,
+      "grad_norm": 4.072330746610035,
+      "learning_rate": 9.87458593625554e-06,
+      "loss": 0.581,
+      "step": 1431
+    },
+    {
+      "epoch": 0.0994306346340786,
+      "grad_norm": 4.434073271049279,
+      "learning_rate": 9.874335538182478e-06,
+      "loss": 0.5349,
+      "step": 1432
+    },
+    {
+      "epoch": 0.09950006943480072,
+      "grad_norm": 4.837339591680378,
+      "learning_rate": 9.87408489357104e-06,
+      "loss": 0.5878,
+      "step": 1433
+    },
+    {
+      "epoch": 0.09956950423552284,
+      "grad_norm": 5.0955652705818695,
+      "learning_rate": 9.873834002433907e-06,
+      "loss": 0.5706,
+      "step": 1434
+    },
+    {
+      "epoch": 0.09963893903624496,
+      "grad_norm": 4.472690709419307,
+      "learning_rate": 9.873582864783762e-06,
+      "loss": 0.6272,
+      "step": 1435
+    },
+    {
+      "epoch": 0.09970837383696708,
+      "grad_norm": 5.469142792153725,
+      "learning_rate": 9.873331480633312e-06,
+      "loss": 0.7154,
+      "step": 1436
+    },
+    {
+      "epoch": 0.09977780863768922,
+      "grad_norm": 4.6354564244692416,
+      "learning_rate": 9.87307984999527e-06,
+      "loss": 0.4549,
+      "step": 1437
+    },
+    {
+      "epoch": 0.09984724343841134,
+      "grad_norm": 4.074484653508504,
+      "learning_rate": 9.872827972882362e-06,
+      "loss": 0.6131,
+      "step": 1438
+    },
+    {
+      "epoch": 0.09991667823913346,
+      "grad_norm": 3.465181877612099,
+      "learning_rate": 9.872575849307332e-06,
+      "loss": 0.3152,
+      "step": 1439
+    },
+    {
+      "epoch": 0.09998611303985558,
+      "grad_norm": 5.1034914967571,
+      "learning_rate": 9.872323479282926e-06,
+      "loss": 0.6223,
+      "step": 1440
+    },
+    {
+      "epoch": 0.1000555478405777,
+      "grad_norm": 3.6859642808034816,
+      "learning_rate": 9.872070862821914e-06,
+      "loss": 0.248,
+      "step": 1441
+    },
+    {
+      "epoch": 0.10012498264129982,
+      "grad_norm": 4.213779463270209,
+      "learning_rate": 9.87181799993707e-06,
+      "loss": 0.5969,
+      "step": 1442
+    },
+    {
+      "epoch": 0.10019441744202194,
+      "grad_norm": 4.839763005401552,
+      "learning_rate": 9.871564890641184e-06,
+      "loss": 0.4299,
+      "step": 1443
+    },
+    {
+      "epoch": 0.10026385224274406,
+      "grad_norm": 6.4194272334589275,
+      "learning_rate": 9.87131153494706e-06,
+      "loss": 1.0267,
+      "step": 1444
+    },
+    {
+      "epoch": 0.10033328704346618,
+      "grad_norm": 5.405520777809143,
+      "learning_rate": 9.871057932867509e-06,
+      "loss": 0.657,
+      "step": 1445
+    },
+    {
+      "epoch": 0.1004027218441883,
+      "grad_norm": 3.510282552520804,
+      "learning_rate": 9.87080408441536e-06,
+      "loss": 0.4098,
+      "step": 1446
+    },
+    {
+      "epoch": 0.10047215664491042,
+      "grad_norm": 5.018641436959001,
+      "learning_rate": 9.870549989603453e-06,
+      "loss": 0.7437,
+      "step": 1447
+    },
+    {
+      "epoch": 0.10054159144563254,
+      "grad_norm": 4.175765925711543,
+      "learning_rate": 9.870295648444639e-06,
+      "loss": 0.4091,
+      "step": 1448
+    },
+    {
+      "epoch": 0.10061102624635468,
+      "grad_norm": 4.507699369394679,
+      "learning_rate": 9.870041060951782e-06,
+      "loss": 0.3783,
+      "step": 1449
+    },
+    {
+      "epoch": 0.1006804610470768,
+      "grad_norm": 4.645013840426378,
+      "learning_rate": 9.869786227137758e-06,
+      "loss": 0.6014,
+      "step": 1450
+    },
+    {
+      "epoch": 0.10074989584779892,
+      "grad_norm": 5.24303950064517,
+      "learning_rate": 9.869531147015459e-06,
+      "loss": 0.6051,
+      "step": 1451
+    },
+    {
+      "epoch": 0.10081933064852104,
+      "grad_norm": 5.433163687740451,
+      "learning_rate": 9.869275820597783e-06,
+      "loss": 0.7789,
+      "step": 1452
+    },
+    {
+      "epoch": 0.10088876544924316,
+      "grad_norm": 4.843308164396283,
+      "learning_rate": 9.869020247897648e-06,
+      "loss": 0.6026,
+      "step": 1453
+    },
+    {
+      "epoch": 0.10095820024996528,
+      "grad_norm": 4.576815581219095,
+      "learning_rate": 9.868764428927977e-06,
+      "loss": 0.4794,
+      "step": 1454
+    },
+    {
+      "epoch": 0.1010276350506874,
+      "grad_norm": 3.6229477668361936,
+      "learning_rate": 9.86850836370171e-06,
+      "loss": 0.2725,
+      "step": 1455
+    },
+    {
+      "epoch": 0.10109706985140952,
+      "grad_norm": 4.058156320114755,
+      "learning_rate": 9.8682520522318e-06,
+      "loss": 0.2962,
+      "step": 1456
+    },
+    {
+      "epoch": 0.10116650465213164,
+      "grad_norm": 3.805415086298999,
+      "learning_rate": 9.86799549453121e-06,
+      "loss": 0.4468,
+      "step": 1457
+    },
+    {
+      "epoch": 0.10123593945285377,
+      "grad_norm": 4.682132470477474,
+      "learning_rate": 9.867738690612916e-06,
+      "loss": 0.4124,
+      "step": 1458
+    },
+    {
+      "epoch": 0.10130537425357589,
+      "grad_norm": 4.417322092609537,
+      "learning_rate": 9.867481640489907e-06,
+      "loss": 0.7896,
+      "step": 1459
+    },
+    {
+      "epoch": 0.10137480905429802,
+      "grad_norm": 5.370521293032936,
+      "learning_rate": 9.867224344175185e-06,
+      "loss": 0.8068,
+      "step": 1460
+    },
+    {
+      "epoch": 0.10144424385502014,
+      "grad_norm": 3.9229433283414443,
+      "learning_rate": 9.866966801681764e-06,
+      "loss": 0.2458,
+      "step": 1461
+    },
+    {
+      "epoch": 0.10151367865574226,
+      "grad_norm": 4.9491396772186365,
+      "learning_rate": 9.866709013022669e-06,
+      "loss": 0.6703,
+      "step": 1462
+    },
+    {
+      "epoch": 0.10158311345646438,
+      "grad_norm": 5.513867226692305,
+      "learning_rate": 9.866450978210938e-06,
+      "loss": 0.6185,
+      "step": 1463
+    },
+    {
+      "epoch": 0.1016525482571865,
+      "grad_norm": 6.46850910579035,
+      "learning_rate": 9.866192697259625e-06,
+      "loss": 0.61,
+      "step": 1464
+    },
+    {
+      "epoch": 0.10172198305790862,
+      "grad_norm": 3.7946975651332737,
+      "learning_rate": 9.86593417018179e-06,
+      "loss": 0.5239,
+      "step": 1465
+    },
+    {
+      "epoch": 0.10179141785863074,
+      "grad_norm": 4.216608452908571,
+      "learning_rate": 9.865675396990513e-06,
+      "loss": 0.49,
+      "step": 1466
+    },
+    {
+      "epoch": 0.10186085265935287,
+      "grad_norm": 4.225718064339089,
+      "learning_rate": 9.86541637769888e-06,
+      "loss": 0.3515,
+      "step": 1467
+    },
+    {
+      "epoch": 0.10193028746007499,
+      "grad_norm": 5.022422290702395,
+      "learning_rate": 9.865157112319992e-06,
+      "loss": 0.7022,
+      "step": 1468
+    },
+    {
+      "epoch": 0.10199972226079711,
+      "grad_norm": 4.4170400213203385,
+      "learning_rate": 9.864897600866961e-06,
+      "loss": 0.5751,
+      "step": 1469
+    },
+    {
+      "epoch": 0.10206915706151923,
+      "grad_norm": 4.274323488898478,
+      "learning_rate": 9.864637843352916e-06,
+      "loss": 0.4817,
+      "step": 1470
+    },
+    {
+      "epoch": 0.10213859186224135,
+      "grad_norm": 4.408698994386973,
+      "learning_rate": 9.864377839790992e-06,
+      "loss": 0.594,
+      "step": 1471
+    },
+    {
+      "epoch": 0.10220802666296348,
+      "grad_norm": 4.244252469750399,
+      "learning_rate": 9.864117590194342e-06,
+      "loss": 0.4003,
+      "step": 1472
+    },
+    {
+      "epoch": 0.1022774614636856,
+      "grad_norm": 6.39113269552439,
+      "learning_rate": 9.863857094576129e-06,
+      "loss": 0.8399,
+      "step": 1473
+    },
+    {
+      "epoch": 0.10234689626440772,
+      "grad_norm": 3.9532686514251383,
+      "learning_rate": 9.863596352949528e-06,
+      "loss": 0.3364,
+      "step": 1474
+    },
+    {
+      "epoch": 0.10241633106512985,
+      "grad_norm": 5.12506003554793,
+      "learning_rate": 9.863335365327726e-06,
+      "loss": 0.6353,
+      "step": 1475
+    },
+    {
+      "epoch": 0.10248576586585197,
+      "grad_norm": 5.4996458875676515,
+      "learning_rate": 9.863074131723926e-06,
+      "loss": 0.6631,
+      "step": 1476
+    },
+    {
+      "epoch": 0.10255520066657409,
+      "grad_norm": 4.204707386388905,
+      "learning_rate": 9.86281265215134e-06,
+      "loss": 0.5784,
+      "step": 1477
+    },
+    {
+      "epoch": 0.10262463546729621,
+      "grad_norm": 3.1744307981162536,
+      "learning_rate": 9.86255092662319e-06,
+      "loss": 0.3978,
+      "step": 1478
+    },
+    {
+      "epoch": 0.10269407026801833,
+      "grad_norm": 5.584314613196682,
+      "learning_rate": 9.862288955152716e-06,
+      "loss": 0.5764,
+      "step": 1479
+    },
+    {
+      "epoch": 0.10276350506874045,
+      "grad_norm": 3.7756388145140103,
+      "learning_rate": 9.862026737753172e-06,
+      "loss": 0.4831,
+      "step": 1480
+    },
+    {
+      "epoch": 0.10283293986946257,
+      "grad_norm": 3.9668559875166354,
+      "learning_rate": 9.861764274437815e-06,
+      "loss": 0.2832,
+      "step": 1481
+    },
+    {
+      "epoch": 0.10290237467018469,
+      "grad_norm": 5.383119243692636,
+      "learning_rate": 9.861501565219924e-06,
+      "loss": 0.8845,
+      "step": 1482
+    },
+    {
+      "epoch": 0.10297180947090682,
+      "grad_norm": 4.715036109096074,
+      "learning_rate": 9.861238610112784e-06,
+      "loss": 0.5594,
+      "step": 1483
+    },
+    {
+      "epoch": 0.10304124427162895,
+      "grad_norm": 3.6892227513070686,
+      "learning_rate": 9.860975409129696e-06,
+      "loss": 0.384,
+      "step": 1484
+    },
+    {
+      "epoch": 0.10311067907235107,
+      "grad_norm": 4.335314531370543,
+      "learning_rate": 9.860711962283972e-06,
+      "loss": 0.6645,
+      "step": 1485
+    },
+    {
+      "epoch": 0.10318011387307319,
+      "grad_norm": 4.291383789553024,
+      "learning_rate": 9.860448269588938e-06,
+      "loss": 0.483,
+      "step": 1486
+    },
+    {
+      "epoch": 0.10324954867379531,
+      "grad_norm": 5.582990820965034,
+      "learning_rate": 9.86018433105793e-06,
+      "loss": 0.6673,
+      "step": 1487
+    },
+    {
+      "epoch": 0.10331898347451743,
+      "grad_norm": 4.981697082701538,
+      "learning_rate": 9.859920146704299e-06,
+      "loss": 0.6948,
+      "step": 1488
+    },
+    {
+      "epoch": 0.10338841827523955,
+      "grad_norm": 3.987023159443007,
+      "learning_rate": 9.859655716541405e-06,
+      "loss": 0.3902,
+      "step": 1489
+    },
+    {
+      "epoch": 0.10345785307596167,
+      "grad_norm": 5.194073800640568,
+      "learning_rate": 9.859391040582625e-06,
+      "loss": 0.8335,
+      "step": 1490
+    },
+    {
+      "epoch": 0.10352728787668379,
+      "grad_norm": 3.872956527428385,
+      "learning_rate": 9.859126118841345e-06,
+      "loss": 0.2913,
+      "step": 1491
+    },
+    {
+      "epoch": 0.10359672267740591,
+      "grad_norm": 2.404685406215228,
+      "learning_rate": 9.858860951330964e-06,
+      "loss": 0.1651,
+      "step": 1492
+    },
+    {
+      "epoch": 0.10366615747812803,
+      "grad_norm": 4.387267208777114,
+      "learning_rate": 9.858595538064895e-06,
+      "loss": 0.656,
+      "step": 1493
+    },
+    {
+      "epoch": 0.10373559227885015,
+      "grad_norm": 3.972165891023337,
+      "learning_rate": 9.85832987905656e-06,
+      "loss": 0.5746,
+      "step": 1494
+    },
+    {
+      "epoch": 0.10380502707957229,
+      "grad_norm": 4.893645957387192,
+      "learning_rate": 9.858063974319399e-06,
+      "loss": 0.5839,
+      "step": 1495
+    },
+    {
+      "epoch": 0.10387446188029441,
+      "grad_norm": 4.703251458920722,
+      "learning_rate": 9.857797823866858e-06,
+      "loss": 0.3784,
+      "step": 1496
+    },
+    {
+      "epoch": 0.10394389668101653,
+      "grad_norm": 4.921201400380296,
+      "learning_rate": 9.857531427712402e-06,
+      "loss": 0.5057,
+      "step": 1497
+    },
+    {
+      "epoch": 0.10401333148173865,
+      "grad_norm": 4.907933484566761,
+      "learning_rate": 9.8572647858695e-06,
+      "loss": 0.6931,
+      "step": 1498
+    },
+    {
+      "epoch": 0.10408276628246077,
+      "grad_norm": 4.575797357221193,
+      "learning_rate": 9.856997898351644e-06,
+      "loss": 0.5452,
+      "step": 1499
+    },
+    {
+      "epoch": 0.10415220108318289,
+      "grad_norm": 4.890072875984711,
+      "learning_rate": 9.85673076517233e-06,
+      "loss": 0.4678,
+      "step": 1500
+    },
+    {
+      "epoch": 0.10422163588390501,
+      "grad_norm": 5.854166444131935,
+      "learning_rate": 9.856463386345068e-06,
+      "loss": 0.4742,
+      "step": 1501
+    },
+    {
+      "epoch": 0.10429107068462713,
+      "grad_norm": 4.66304438296173,
+      "learning_rate": 9.856195761883385e-06,
+      "loss": 0.4911,
+      "step": 1502
+    },
+    {
+      "epoch": 0.10436050548534925,
+      "grad_norm": 3.8920834612027053,
+      "learning_rate": 9.855927891800814e-06,
+      "loss": 0.4804,
+      "step": 1503
+    },
+    {
+      "epoch": 0.10442994028607137,
+      "grad_norm": 4.111242119672071,
+      "learning_rate": 9.855659776110908e-06,
+      "loss": 0.5079,
+      "step": 1504
+    },
+    {
+      "epoch": 0.1044993750867935,
+      "grad_norm": 4.374730703835033,
+      "learning_rate": 9.855391414827223e-06,
+      "loss": 0.3222,
+      "step": 1505
+    },
+    {
+      "epoch": 0.10456880988751563,
+      "grad_norm": 5.764078149491738,
+      "learning_rate": 9.855122807963334e-06,
+      "loss": 0.837,
+      "step": 1506
+    },
+    {
+      "epoch": 0.10463824468823775,
+      "grad_norm": 4.5850401913252945,
+      "learning_rate": 9.854853955532829e-06,
+      "loss": 0.5047,
+      "step": 1507
+    },
+    {
+      "epoch": 0.10470767948895987,
+      "grad_norm": 4.498032863862511,
+      "learning_rate": 9.854584857549303e-06,
+      "loss": 0.5032,
+      "step": 1508
+    },
+    {
+      "epoch": 0.10477711428968199,
+      "grad_norm": 4.202398976515563,
+      "learning_rate": 9.85431551402637e-06,
+      "loss": 0.4368,
+      "step": 1509
+    },
+    {
+      "epoch": 0.10484654909040411,
+      "grad_norm": 4.952828072459177,
+      "learning_rate": 9.854045924977648e-06,
+      "loss": 0.63,
+      "step": 1510
+    },
+    {
+      "epoch": 0.10491598389112623,
+      "grad_norm": 4.425120166048361,
+      "learning_rate": 9.85377609041678e-06,
+      "loss": 0.6482,
+      "step": 1511
+    },
+    {
+      "epoch": 0.10498541869184835,
+      "grad_norm": 4.529078257756628,
+      "learning_rate": 9.853506010357408e-06,
+      "loss": 0.6276,
+      "step": 1512
+    },
+    {
+      "epoch": 0.10505485349257047,
+      "grad_norm": 4.637433011654157,
+      "learning_rate": 9.853235684813193e-06,
+      "loss": 0.6854,
+      "step": 1513
+    },
+    {
+      "epoch": 0.1051242882932926,
+      "grad_norm": 4.661170672388385,
+      "learning_rate": 9.85296511379781e-06,
+      "loss": 0.6506,
+      "step": 1514
+    },
+    {
+      "epoch": 0.10519372309401472,
+      "grad_norm": 6.071298315668853,
+      "learning_rate": 9.852694297324943e-06,
+      "loss": 0.9429,
+      "step": 1515
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 3.94590979840419,
+      "learning_rate": 9.85242323540829e-06,
+      "loss": 0.4408,
+      "step": 1516
+    },
+    {
+      "epoch": 0.10533259269545897,
+      "grad_norm": 4.332637500242891,
+      "learning_rate": 9.85215192806156e-06,
+      "loss": 0.296,
+      "step": 1517
+    },
+    {
+      "epoch": 0.10540202749618109,
+      "grad_norm": 4.480137032902055,
+      "learning_rate": 9.851880375298476e-06,
+      "loss": 0.4728,
+      "step": 1518
+    },
+    {
+      "epoch": 0.10547146229690321,
+      "grad_norm": 4.731982617974487,
+      "learning_rate": 9.851608577132772e-06,
+      "loss": 0.5655,
+      "step": 1519
+    },
+    {
+      "epoch": 0.10554089709762533,
+      "grad_norm": 6.520421620226633,
+      "learning_rate": 9.851336533578197e-06,
+      "loss": 0.582,
+      "step": 1520
+    },
+    {
+      "epoch": 0.10561033189834745,
+      "grad_norm": 5.8580211585506525,
+      "learning_rate": 9.851064244648512e-06,
+      "loss": 0.6654,
+      "step": 1521
+    },
+    {
+      "epoch": 0.10567976669906957,
+      "grad_norm": 4.118247584013102,
+      "learning_rate": 9.850791710357484e-06,
+      "loss": 0.4173,
+      "step": 1522
+    },
+    {
+      "epoch": 0.1057492014997917,
+      "grad_norm": 3.846588091553713,
+      "learning_rate": 9.8505189307189e-06,
+      "loss": 0.4547,
+      "step": 1523
+    },
+    {
+      "epoch": 0.10581863630051382,
+      "grad_norm": 3.780827342457809,
+      "learning_rate": 9.85024590574656e-06,
+      "loss": 0.3835,
+      "step": 1524
+    },
+    {
+      "epoch": 0.10588807110123594,
+      "grad_norm": 4.142471460006247,
+      "learning_rate": 9.849972635454269e-06,
+      "loss": 0.3718,
+      "step": 1525
+    },
+    {
+      "epoch": 0.10595750590195806,
+      "grad_norm": 5.30455870557488,
+      "learning_rate": 9.849699119855849e-06,
+      "loss": 0.644,
+      "step": 1526
+    },
+    {
+      "epoch": 0.10602694070268018,
+      "grad_norm": 5.125261613370212,
+      "learning_rate": 9.849425358965137e-06,
+      "loss": 0.5728,
+      "step": 1527
+    },
+    {
+      "epoch": 0.1060963755034023,
+      "grad_norm": 4.036633810686275,
+      "learning_rate": 9.849151352795978e-06,
+      "loss": 0.521,
+      "step": 1528
+    },
+    {
+      "epoch": 0.10616581030412443,
+      "grad_norm": 4.023083043024948,
+      "learning_rate": 9.848877101362229e-06,
+      "loss": 0.4578,
+      "step": 1529
+    },
+    {
+      "epoch": 0.10623524510484655,
+      "grad_norm": 6.097939891037617,
+      "learning_rate": 9.848602604677764e-06,
+      "loss": 0.683,
+      "step": 1530
+    },
+    {
+      "epoch": 0.10630467990556867,
+      "grad_norm": 5.428831392012675,
+      "learning_rate": 9.848327862756466e-06,
+      "loss": 0.5905,
+      "step": 1531
+    },
+    {
+      "epoch": 0.1063741147062908,
+      "grad_norm": 4.038482225978325,
+      "learning_rate": 9.848052875612232e-06,
+      "loss": 0.3259,
+      "step": 1532
+    },
+    {
+      "epoch": 0.10644354950701292,
+      "grad_norm": 4.774959460651911,
+      "learning_rate": 9.84777764325897e-06,
+      "loss": 0.5392,
+      "step": 1533
+    },
+    {
+      "epoch": 0.10651298430773504,
+      "grad_norm": 4.649973700708671,
+      "learning_rate": 9.847502165710596e-06,
+      "loss": 0.5627,
+      "step": 1534
+    },
+    {
+      "epoch": 0.10658241910845716,
+      "grad_norm": 5.712074584432798,
+      "learning_rate": 9.84722644298105e-06,
+      "loss": 0.4408,
+      "step": 1535
+    },
+    {
+      "epoch": 0.10665185390917928,
+      "grad_norm": 6.35996853697112,
+      "learning_rate": 9.846950475084279e-06,
+      "loss": 0.9047,
+      "step": 1536
+    },
+    {
+      "epoch": 0.1067212887099014,
+      "grad_norm": 5.33043706443409,
+      "learning_rate": 9.846674262034234e-06,
+      "loss": 0.66,
+      "step": 1537
+    },
+    {
+      "epoch": 0.10679072351062352,
+      "grad_norm": 4.937533137775621,
+      "learning_rate": 9.84639780384489e-06,
+      "loss": 0.4078,
+      "step": 1538
+    },
+    {
+      "epoch": 0.10686015831134564,
+      "grad_norm": 5.910042035005478,
+      "learning_rate": 9.846121100530229e-06,
+      "loss": 0.5399,
+      "step": 1539
+    },
+    {
+      "epoch": 0.10692959311206778,
+      "grad_norm": 4.818395582616907,
+      "learning_rate": 9.845844152104248e-06,
+      "loss": 0.5133,
+      "step": 1540
+    },
+    {
+      "epoch": 0.1069990279127899,
+      "grad_norm": 5.285148066265314,
+      "learning_rate": 9.84556695858095e-06,
+      "loss": 0.8738,
+      "step": 1541
+    },
+    {
+      "epoch": 0.10706846271351202,
+      "grad_norm": 5.475571325031809,
+      "learning_rate": 9.845289519974362e-06,
+      "loss": 0.5095,
+      "step": 1542
+    },
+    {
+      "epoch": 0.10713789751423414,
+      "grad_norm": 6.184994607743289,
+      "learning_rate": 9.84501183629851e-06,
+      "loss": 0.4073,
+      "step": 1543
+    },
+    {
+      "epoch": 0.10720733231495626,
+      "grad_norm": 3.8931914758363138,
+      "learning_rate": 9.844733907567444e-06,
+      "loss": 0.3666,
+      "step": 1544
+    },
+    {
+      "epoch": 0.10727676711567838,
+      "grad_norm": 4.297116611492556,
+      "learning_rate": 9.844455733795217e-06,
+      "loss": 0.4009,
+      "step": 1545
+    },
+    {
+      "epoch": 0.1073462019164005,
+      "grad_norm": 4.192064205735997,
+      "learning_rate": 9.844177314995902e-06,
+      "loss": 0.402,
+      "step": 1546
+    },
+    {
+      "epoch": 0.10741563671712262,
+      "grad_norm": 3.7897208059550316,
+      "learning_rate": 9.84389865118358e-06,
+      "loss": 0.2806,
+      "step": 1547
+    },
+    {
+      "epoch": 0.10748507151784474,
+      "grad_norm": 4.3203932120933315,
+      "learning_rate": 9.843619742372345e-06,
+      "loss": 0.5323,
+      "step": 1548
+    },
+    {
+      "epoch": 0.10755450631856686,
+      "grad_norm": 3.9939091673572746,
+      "learning_rate": 9.843340588576306e-06,
+      "loss": 0.5407,
+      "step": 1549
+    },
+    {
+      "epoch": 0.10762394111928898,
+      "grad_norm": 5.150457779299713,
+      "learning_rate": 9.84306118980958e-06,
+      "loss": 0.7399,
+      "step": 1550
+    },
+    {
+      "epoch": 0.1076933759200111,
+      "grad_norm": 4.911652375244797,
+      "learning_rate": 9.842781546086298e-06,
+      "loss": 0.6613,
+      "step": 1551
+    },
+    {
+      "epoch": 0.10776281072073324,
+      "grad_norm": 4.881668184883155,
+      "learning_rate": 9.842501657420606e-06,
+      "loss": 0.7577,
+      "step": 1552
+    },
+    {
+      "epoch": 0.10783224552145536,
+      "grad_norm": 5.202801671014035,
+      "learning_rate": 9.84222152382666e-06,
+      "loss": 0.5879,
+      "step": 1553
+    },
+    {
+      "epoch": 0.10790168032217748,
+      "grad_norm": 2.8899270835761808,
+      "learning_rate": 9.841941145318628e-06,
+      "loss": 0.2977,
+      "step": 1554
+    },
+    {
+      "epoch": 0.1079711151228996,
+      "grad_norm": 4.569706152644165,
+      "learning_rate": 9.841660521910692e-06,
+      "loss": 0.4627,
+      "step": 1555
+    },
+    {
+      "epoch": 0.10804054992362172,
+      "grad_norm": 4.687682745828268,
+      "learning_rate": 9.841379653617048e-06,
+      "loss": 0.4952,
+      "step": 1556
+    },
+    {
+      "epoch": 0.10810998472434384,
+      "grad_norm": 4.252852647572385,
+      "learning_rate": 9.841098540451897e-06,
+      "loss": 0.4663,
+      "step": 1557
+    },
+    {
+      "epoch": 0.10817941952506596,
+      "grad_norm": 4.719937487685151,
+      "learning_rate": 9.840817182429459e-06,
+      "loss": 0.6428,
+      "step": 1558
+    },
+    {
+      "epoch": 0.10824885432578808,
+      "grad_norm": 4.459412855425336,
+      "learning_rate": 9.840535579563969e-06,
+      "loss": 0.5252,
+      "step": 1559
+    },
+    {
+      "epoch": 0.1083182891265102,
+      "grad_norm": 5.054094951706585,
+      "learning_rate": 9.840253731869664e-06,
+      "loss": 0.597,
+      "step": 1560
+    },
+    {
+      "epoch": 0.10838772392723232,
+      "grad_norm": 3.1400965121412687,
+      "learning_rate": 9.839971639360804e-06,
+      "loss": 0.2184,
+      "step": 1561
+    },
+    {
+      "epoch": 0.10845715872795444,
+      "grad_norm": 3.969700058327271,
+      "learning_rate": 9.839689302051654e-06,
+      "loss": 0.4614,
+      "step": 1562
+    },
+    {
+      "epoch": 0.10852659352867658,
+      "grad_norm": 4.0392657255352855,
+      "learning_rate": 9.839406719956496e-06,
+      "loss": 0.3555,
+      "step": 1563
+    },
+    {
+      "epoch": 0.1085960283293987,
+      "grad_norm": 3.495674736955305,
+      "learning_rate": 9.839123893089622e-06,
+      "loss": 0.2343,
+      "step": 1564
+    },
+    {
+      "epoch": 0.10866546313012082,
+      "grad_norm": 4.667452749454958,
+      "learning_rate": 9.838840821465338e-06,
+      "loss": 0.643,
+      "step": 1565
+    },
+    {
+      "epoch": 0.10873489793084294,
+      "grad_norm": 4.490900119758754,
+      "learning_rate": 9.83855750509796e-06,
+      "loss": 0.5119,
+      "step": 1566
+    },
+    {
+      "epoch": 0.10880433273156506,
+      "grad_norm": 3.314578752852845,
+      "learning_rate": 9.83827394400182e-06,
+      "loss": 0.3871,
+      "step": 1567
+    },
+    {
+      "epoch": 0.10887376753228718,
+      "grad_norm": 4.26856821223684,
+      "learning_rate": 9.837990138191258e-06,
+      "loss": 0.4262,
+      "step": 1568
+    },
+    {
+      "epoch": 0.1089432023330093,
+      "grad_norm": 3.9969153602034533,
+      "learning_rate": 9.837706087680628e-06,
+      "loss": 0.4782,
+      "step": 1569
+    },
+    {
+      "epoch": 0.10901263713373142,
+      "grad_norm": 3.6387732467998637,
+      "learning_rate": 9.8374217924843e-06,
+      "loss": 0.3585,
+      "step": 1570
+    },
+    {
+      "epoch": 0.10908207193445354,
+      "grad_norm": 3.358258342037373,
+      "learning_rate": 9.83713725261665e-06,
+      "loss": 0.2208,
+      "step": 1571
+    },
+    {
+      "epoch": 0.10915150673517567,
+      "grad_norm": 3.8320195121228413,
+      "learning_rate": 9.836852468092075e-06,
+      "loss": 0.4853,
+      "step": 1572
+    },
+    {
+      "epoch": 0.10922094153589779,
+      "grad_norm": 4.593973320215978,
+      "learning_rate": 9.836567438924972e-06,
+      "loss": 0.4529,
+      "step": 1573
+    },
+    {
+      "epoch": 0.10929037633661991,
+      "grad_norm": 3.217217836563474,
+      "learning_rate": 9.836282165129763e-06,
+      "loss": 0.377,
+      "step": 1574
+    },
+    {
+      "epoch": 0.10935981113734204,
+      "grad_norm": 4.5976554124817905,
+      "learning_rate": 9.835996646720873e-06,
+      "loss": 0.5341,
+      "step": 1575
+    },
+    {
+      "epoch": 0.10942924593806416,
+      "grad_norm": 3.7926232650431015,
+      "learning_rate": 9.835710883712746e-06,
+      "loss": 0.4095,
+      "step": 1576
+    },
+    {
+      "epoch": 0.10949868073878628,
+      "grad_norm": 4.174914908042183,
+      "learning_rate": 9.835424876119833e-06,
+      "loss": 0.4987,
+      "step": 1577
+    },
+    {
+      "epoch": 0.1095681155395084,
+      "grad_norm": 2.8208369748292643,
+      "learning_rate": 9.835138623956603e-06,
+      "loss": 0.3247,
+      "step": 1578
+    },
+    {
+      "epoch": 0.10963755034023052,
+      "grad_norm": 3.5268928991424877,
+      "learning_rate": 9.834852127237532e-06,
+      "loss": 0.4655,
+      "step": 1579
+    },
+    {
+      "epoch": 0.10970698514095265,
+      "grad_norm": 5.3490427375905885,
+      "learning_rate": 9.83456538597711e-06,
+      "loss": 0.8395,
+      "step": 1580
+    },
+    {
+      "epoch": 0.10977641994167477,
+      "grad_norm": 4.528018993315913,
+      "learning_rate": 9.834278400189845e-06,
+      "loss": 0.5525,
+      "step": 1581
+    },
+    {
+      "epoch": 0.10984585474239689,
+      "grad_norm": 4.932849394101731,
+      "learning_rate": 9.833991169890247e-06,
+      "loss": 0.5518,
+      "step": 1582
+    },
+    {
+      "epoch": 0.10991528954311901,
+      "grad_norm": 4.714738264369017,
+      "learning_rate": 9.833703695092845e-06,
+      "loss": 0.5196,
+      "step": 1583
+    },
+    {
+      "epoch": 0.10998472434384113,
+      "grad_norm": 4.408660557303999,
+      "learning_rate": 9.83341597581218e-06,
+      "loss": 0.4204,
+      "step": 1584
+    },
+    {
+      "epoch": 0.11005415914456325,
+      "grad_norm": 3.743320813706743,
+      "learning_rate": 9.833128012062805e-06,
+      "loss": 0.5559,
+      "step": 1585
+    },
+    {
+      "epoch": 0.11012359394528538,
+      "grad_norm": 3.5811830990667435,
+      "learning_rate": 9.832839803859285e-06,
+      "loss": 0.4234,
+      "step": 1586
+    },
+    {
+      "epoch": 0.1101930287460075,
+      "grad_norm": 5.104278667578373,
+      "learning_rate": 9.832551351216195e-06,
+      "loss": 0.6434,
+      "step": 1587
+    },
+    {
+      "epoch": 0.11026246354672962,
+      "grad_norm": 6.72400788502477,
+      "learning_rate": 9.832262654148126e-06,
+      "loss": 0.8587,
+      "step": 1588
+    },
+    {
+      "epoch": 0.11033189834745175,
+      "grad_norm": 4.161496348110367,
+      "learning_rate": 9.83197371266968e-06,
+      "loss": 0.5506,
+      "step": 1589
+    },
+    {
+      "epoch": 0.11040133314817387,
+      "grad_norm": 3.776560693539345,
+      "learning_rate": 9.831684526795471e-06,
+      "loss": 0.4692,
+      "step": 1590
+    },
+    {
+      "epoch": 0.11047076794889599,
+      "grad_norm": 4.498195818207967,
+      "learning_rate": 9.831395096540128e-06,
+      "loss": 0.5355,
+      "step": 1591
+    },
+    {
+      "epoch": 0.11054020274961811,
+      "grad_norm": 3.7454207291080124,
+      "learning_rate": 9.831105421918287e-06,
+      "loss": 0.4557,
+      "step": 1592
+    },
+    {
+      "epoch": 0.11060963755034023,
+      "grad_norm": 4.390260306066933,
+      "learning_rate": 9.8308155029446e-06,
+      "loss": 0.4584,
+      "step": 1593
+    },
+    {
+      "epoch": 0.11067907235106235,
+      "grad_norm": 3.0579293731284625,
+      "learning_rate": 9.830525339633732e-06,
+      "loss": 0.2558,
+      "step": 1594
+    },
+    {
+      "epoch": 0.11074850715178447,
+      "grad_norm": 4.870705655984856,
+      "learning_rate": 9.830234932000358e-06,
+      "loss": 0.6957,
+      "step": 1595
+    },
+    {
+      "epoch": 0.11081794195250659,
+      "grad_norm": 3.823578666763392,
+      "learning_rate": 9.829944280059166e-06,
+      "loss": 0.5417,
+      "step": 1596
+    },
+    {
+      "epoch": 0.11088737675322871,
+      "grad_norm": 3.966919167976425,
+      "learning_rate": 9.82965338382486e-06,
+      "loss": 0.3847,
+      "step": 1597
+    },
+    {
+      "epoch": 0.11095681155395085,
+      "grad_norm": 4.842282801747006,
+      "learning_rate": 9.829362243312151e-06,
+      "loss": 0.4381,
+      "step": 1598
+    },
+    {
+      "epoch": 0.11102624635467297,
+      "grad_norm": 3.58162065625632,
+      "learning_rate": 9.829070858535761e-06,
+      "loss": 0.3113,
+      "step": 1599
+    },
+    {
+      "epoch": 0.11109568115539509,
+      "grad_norm": 4.492226767601615,
+      "learning_rate": 9.828779229510434e-06,
+      "loss": 0.5637,
+      "step": 1600
+    },
+    {
+      "epoch": 0.11116511595611721,
+      "grad_norm": 4.387054802836475,
+      "learning_rate": 9.82848735625092e-06,
+      "loss": 0.4992,
+      "step": 1601
+    },
+    {
+      "epoch": 0.11123455075683933,
+      "grad_norm": 5.118782618547968,
+      "learning_rate": 9.828195238771976e-06,
+      "loss": 0.5997,
+      "step": 1602
+    },
+    {
+      "epoch": 0.11130398555756145,
+      "grad_norm": 5.251109347835807,
+      "learning_rate": 9.82790287708838e-06,
+      "loss": 0.683,
+      "step": 1603
+    },
+    {
+      "epoch": 0.11137342035828357,
+      "grad_norm": 3.5240909002017995,
+      "learning_rate": 9.827610271214921e-06,
+      "loss": 0.351,
+      "step": 1604
+    },
+    {
+      "epoch": 0.11144285515900569,
+      "grad_norm": 4.875392596363631,
+      "learning_rate": 9.827317421166397e-06,
+      "loss": 0.502,
+      "step": 1605
+    },
+    {
+      "epoch": 0.11151228995972781,
+      "grad_norm": 5.8870866083624795,
+      "learning_rate": 9.827024326957621e-06,
+      "loss": 0.5,
+      "step": 1606
+    },
+    {
+      "epoch": 0.11158172476044993,
+      "grad_norm": 4.090982102166375,
+      "learning_rate": 9.826730988603416e-06,
+      "loss": 0.5652,
+      "step": 1607
+    },
+    {
+      "epoch": 0.11165115956117205,
+      "grad_norm": 5.348751817782694,
+      "learning_rate": 9.82643740611862e-06,
+      "loss": 0.9469,
+      "step": 1608
+    },
+    {
+      "epoch": 0.11172059436189419,
+      "grad_norm": 3.050607589608507,
+      "learning_rate": 9.82614357951808e-06,
+      "loss": 0.3071,
+      "step": 1609
+    },
+    {
+      "epoch": 0.11179002916261631,
+      "grad_norm": 5.4928244049893555,
+      "learning_rate": 9.825849508816661e-06,
+      "loss": 0.7792,
+      "step": 1610
+    },
+    {
+      "epoch": 0.11185946396333843,
+      "grad_norm": 4.209022350640694,
+      "learning_rate": 9.825555194029232e-06,
+      "loss": 0.5517,
+      "step": 1611
+    },
+    {
+      "epoch": 0.11192889876406055,
+      "grad_norm": 4.092041768635105,
+      "learning_rate": 9.825260635170684e-06,
+      "loss": 0.5325,
+      "step": 1612
+    },
+    {
+      "epoch": 0.11199833356478267,
+      "grad_norm": 4.9619086965080434,
+      "learning_rate": 9.824965832255913e-06,
+      "loss": 0.743,
+      "step": 1613
+    },
+    {
+      "epoch": 0.11206776836550479,
+      "grad_norm": 4.421598831619101,
+      "learning_rate": 9.82467078529983e-06,
+      "loss": 0.3902,
+      "step": 1614
+    },
+    {
+      "epoch": 0.11213720316622691,
+      "grad_norm": 3.9773891500369074,
+      "learning_rate": 9.824375494317358e-06,
+      "loss": 0.4186,
+      "step": 1615
+    },
+    {
+      "epoch": 0.11220663796694903,
+      "grad_norm": 3.723363296604242,
+      "learning_rate": 9.824079959323431e-06,
+      "loss": 0.3506,
+      "step": 1616
+    },
+    {
+      "epoch": 0.11227607276767115,
+      "grad_norm": 4.030108433675491,
+      "learning_rate": 9.823784180333e-06,
+      "loss": 0.4958,
+      "step": 1617
+    },
+    {
+      "epoch": 0.11234550756839327,
+      "grad_norm": 4.9857469711080515,
+      "learning_rate": 9.823488157361023e-06,
+      "loss": 0.5275,
+      "step": 1618
+    },
+    {
+      "epoch": 0.1124149423691154,
+      "grad_norm": 5.174116321636986,
+      "learning_rate": 9.823191890422474e-06,
+      "loss": 0.6698,
+      "step": 1619
+    },
+    {
+      "epoch": 0.11248437716983753,
+      "grad_norm": 3.2327252267581024,
+      "learning_rate": 9.822895379532336e-06,
+      "loss": 0.3652,
+      "step": 1620
+    },
+    {
+      "epoch": 0.11255381197055965,
+      "grad_norm": 3.132738327208875,
+      "learning_rate": 9.822598624705609e-06,
+      "loss": 0.2778,
+      "step": 1621
+    },
+    {
+      "epoch": 0.11262324677128177,
+      "grad_norm": 3.5805252444323457,
+      "learning_rate": 9.822301625957299e-06,
+      "loss": 0.2494,
+      "step": 1622
+    },
+    {
+      "epoch": 0.11269268157200389,
+      "grad_norm": 2.768350140469748,
+      "learning_rate": 9.82200438330243e-06,
+      "loss": 0.2753,
+      "step": 1623
+    },
+    {
+      "epoch": 0.11276211637272601,
+      "grad_norm": 3.9852206986210277,
+      "learning_rate": 9.821706896756035e-06,
+      "loss": 0.4955,
+      "step": 1624
+    },
+    {
+      "epoch": 0.11283155117344813,
+      "grad_norm": 5.609812176569715,
+      "learning_rate": 9.821409166333163e-06,
+      "loss": 0.6614,
+      "step": 1625
+    },
+    {
+      "epoch": 0.11290098597417025,
+      "grad_norm": 4.3938758193827265,
+      "learning_rate": 9.82111119204887e-06,
+      "loss": 0.4627,
+      "step": 1626
+    },
+    {
+      "epoch": 0.11297042077489237,
+      "grad_norm": 5.5237164783105195,
+      "learning_rate": 9.820812973918229e-06,
+      "loss": 0.6492,
+      "step": 1627
+    },
+    {
+      "epoch": 0.1130398555756145,
+      "grad_norm": 4.5435492063896215,
+      "learning_rate": 9.820514511956322e-06,
+      "loss": 0.5198,
+      "step": 1628
+    },
+    {
+      "epoch": 0.11310929037633662,
+      "grad_norm": 4.994011109844973,
+      "learning_rate": 9.820215806178247e-06,
+      "loss": 0.6478,
+      "step": 1629
+    },
+    {
+      "epoch": 0.11317872517705874,
+      "grad_norm": 2.462600576237636,
+      "learning_rate": 9.81991685659911e-06,
+      "loss": 0.3581,
+      "step": 1630
+    },
+    {
+      "epoch": 0.11324815997778086,
+      "grad_norm": 4.173065566970126,
+      "learning_rate": 9.819617663234033e-06,
+      "loss": 0.5546,
+      "step": 1631
+    },
+    {
+      "epoch": 0.11331759477850299,
+      "grad_norm": 4.713183084165724,
+      "learning_rate": 9.819318226098147e-06,
+      "loss": 0.755,
+      "step": 1632
+    },
+    {
+      "epoch": 0.11338702957922511,
+      "grad_norm": 4.484860173800317,
+      "learning_rate": 9.8190185452066e-06,
+      "loss": 0.5616,
+      "step": 1633
+    },
+    {
+      "epoch": 0.11345646437994723,
+      "grad_norm": 5.0019413446242815,
+      "learning_rate": 9.818718620574547e-06,
+      "loss": 0.4998,
+      "step": 1634
+    },
+    {
+      "epoch": 0.11352589918066935,
+      "grad_norm": 4.243186151910063,
+      "learning_rate": 9.81841845221716e-06,
+      "loss": 0.4372,
+      "step": 1635
+    },
+    {
+      "epoch": 0.11359533398139147,
+      "grad_norm": 5.811550641860544,
+      "learning_rate": 9.81811804014962e-06,
+      "loss": 0.799,
+      "step": 1636
+    },
+    {
+      "epoch": 0.1136647687821136,
+      "grad_norm": 4.018389265290418,
+      "learning_rate": 9.817817384387123e-06,
+      "loss": 0.5037,
+      "step": 1637
+    },
+    {
+      "epoch": 0.11373420358283572,
+      "grad_norm": 4.737743263934409,
+      "learning_rate": 9.817516484944871e-06,
+      "loss": 0.7064,
+      "step": 1638
+    },
+    {
+      "epoch": 0.11380363838355784,
+      "grad_norm": 4.318047725515127,
+      "learning_rate": 9.817215341838089e-06,
+      "loss": 0.5675,
+      "step": 1639
+    },
+    {
+      "epoch": 0.11387307318427996,
+      "grad_norm": 4.18738630195801,
+      "learning_rate": 9.816913955082006e-06,
+      "loss": 0.6433,
+      "step": 1640
+    },
+    {
+      "epoch": 0.11394250798500208,
+      "grad_norm": 4.359658266159575,
+      "learning_rate": 9.816612324691865e-06,
+      "loss": 0.4907,
+      "step": 1641
+    },
+    {
+      "epoch": 0.1140119427857242,
+      "grad_norm": 2.930211588564268,
+      "learning_rate": 9.816310450682923e-06,
+      "loss": 0.3517,
+      "step": 1642
+    },
+    {
+      "epoch": 0.11408137758644633,
+      "grad_norm": 4.153235784652996,
+      "learning_rate": 9.816008333070449e-06,
+      "loss": 0.5227,
+      "step": 1643
+    },
+    {
+      "epoch": 0.11415081238716845,
+      "grad_norm": 5.435588917114712,
+      "learning_rate": 9.815705971869722e-06,
+      "loss": 0.6447,
+      "step": 1644
+    },
+    {
+      "epoch": 0.11422024718789058,
+      "grad_norm": 5.193230487598492,
+      "learning_rate": 9.815403367096038e-06,
+      "loss": 0.7624,
+      "step": 1645
+    },
+    {
+      "epoch": 0.1142896819886127,
+      "grad_norm": 3.510205620724451,
+      "learning_rate": 9.8151005187647e-06,
+      "loss": 0.4098,
+      "step": 1646
+    },
+    {
+      "epoch": 0.11435911678933482,
+      "grad_norm": 5.170449628229908,
+      "learning_rate": 9.814797426891026e-06,
+      "loss": 0.558,
+      "step": 1647
+    },
+    {
+      "epoch": 0.11442855159005694,
+      "grad_norm": 4.271438949125593,
+      "learning_rate": 9.814494091490347e-06,
+      "loss": 0.4687,
+      "step": 1648
+    },
+    {
+      "epoch": 0.11449798639077906,
+      "grad_norm": 4.5198704923220285,
+      "learning_rate": 9.814190512578003e-06,
+      "loss": 0.5943,
+      "step": 1649
+    },
+    {
+      "epoch": 0.11456742119150118,
+      "grad_norm": 4.780131795922946,
+      "learning_rate": 9.813886690169353e-06,
+      "loss": 0.6866,
+      "step": 1650
+    },
+    {
+      "epoch": 0.1146368559922233,
+      "grad_norm": 4.082262998535368,
+      "learning_rate": 9.81358262427976e-06,
+      "loss": 0.5692,
+      "step": 1651
+    },
+    {
+      "epoch": 0.11470629079294542,
+      "grad_norm": 4.184962174339737,
+      "learning_rate": 9.813278314924606e-06,
+      "loss": 0.4397,
+      "step": 1652
+    },
+    {
+      "epoch": 0.11477572559366754,
+      "grad_norm": 3.863851143741417,
+      "learning_rate": 9.812973762119282e-06,
+      "loss": 0.2746,
+      "step": 1653
+    },
+    {
+      "epoch": 0.11484516039438966,
+      "grad_norm": 4.636606359121546,
+      "learning_rate": 9.812668965879189e-06,
+      "loss": 0.5567,
+      "step": 1654
+    },
+    {
+      "epoch": 0.1149145951951118,
+      "grad_norm": 3.9603695271694153,
+      "learning_rate": 9.812363926219747e-06,
+      "loss": 0.5271,
+      "step": 1655
+    },
+    {
+      "epoch": 0.11498402999583392,
+      "grad_norm": 4.384958577879751,
+      "learning_rate": 9.812058643156383e-06,
+      "loss": 0.5356,
+      "step": 1656
+    },
+    {
+      "epoch": 0.11505346479655604,
+      "grad_norm": 4.941713028545739,
+      "learning_rate": 9.811753116704538e-06,
+      "loss": 0.677,
+      "step": 1657
+    },
+    {
+      "epoch": 0.11512289959727816,
+      "grad_norm": 3.7051703814617416,
+      "learning_rate": 9.811447346879666e-06,
+      "loss": 0.3111,
+      "step": 1658
+    },
+    {
+      "epoch": 0.11519233439800028,
+      "grad_norm": 4.507083872550089,
+      "learning_rate": 9.811141333697229e-06,
+      "loss": 0.3968,
+      "step": 1659
+    },
+    {
+      "epoch": 0.1152617691987224,
+      "grad_norm": 3.8764165037350873,
+      "learning_rate": 9.81083507717271e-06,
+      "loss": 0.5018,
+      "step": 1660
+    },
+    {
+      "epoch": 0.11533120399944452,
+      "grad_norm": 4.690159655133787,
+      "learning_rate": 9.810528577321596e-06,
+      "loss": 0.5007,
+      "step": 1661
+    },
+    {
+      "epoch": 0.11540063880016664,
+      "grad_norm": 4.75494499096628,
+      "learning_rate": 9.81022183415939e-06,
+      "loss": 0.4398,
+      "step": 1662
+    },
+    {
+      "epoch": 0.11547007360088876,
+      "grad_norm": 4.152409365538443,
+      "learning_rate": 9.809914847701608e-06,
+      "loss": 0.6402,
+      "step": 1663
+    },
+    {
+      "epoch": 0.11553950840161088,
+      "grad_norm": 4.836386581933431,
+      "learning_rate": 9.809607617963774e-06,
+      "loss": 0.6558,
+      "step": 1664
+    },
+    {
+      "epoch": 0.115608943202333,
+      "grad_norm": 4.879353539847682,
+      "learning_rate": 9.809300144961428e-06,
+      "loss": 0.5634,
+      "step": 1665
+    },
+    {
+      "epoch": 0.11567837800305514,
+      "grad_norm": 3.8245068287320474,
+      "learning_rate": 9.808992428710128e-06,
+      "loss": 0.3772,
+      "step": 1666
+    },
+    {
+      "epoch": 0.11574781280377726,
+      "grad_norm": 4.933973318157725,
+      "learning_rate": 9.808684469225427e-06,
+      "loss": 0.5334,
+      "step": 1667
+    },
+    {
+      "epoch": 0.11581724760449938,
+      "grad_norm": 3.582216719386233,
+      "learning_rate": 9.80837626652291e-06,
+      "loss": 0.4699,
+      "step": 1668
+    },
+    {
+      "epoch": 0.1158866824052215,
+      "grad_norm": 5.042437378079707,
+      "learning_rate": 9.808067820618162e-06,
+      "loss": 0.8377,
+      "step": 1669
+    },
+    {
+      "epoch": 0.11595611720594362,
+      "grad_norm": 3.096714235307455,
+      "learning_rate": 9.807759131526784e-06,
+      "loss": 0.3174,
+      "step": 1670
+    },
+    {
+      "epoch": 0.11602555200666574,
+      "grad_norm": 4.083449398549343,
+      "learning_rate": 9.807450199264388e-06,
+      "loss": 0.4748,
+      "step": 1671
+    },
+    {
+      "epoch": 0.11609498680738786,
+      "grad_norm": 5.066613552249841,
+      "learning_rate": 9.807141023846602e-06,
+      "loss": 0.5623,
+      "step": 1672
+    },
+    {
+      "epoch": 0.11616442160810998,
+      "grad_norm": 2.829723536089631,
+      "learning_rate": 9.806831605289064e-06,
+      "loss": 0.298,
+      "step": 1673
+    },
+    {
+      "epoch": 0.1162338564088321,
+      "grad_norm": 3.7123490835591877,
+      "learning_rate": 9.806521943607421e-06,
+      "loss": 0.4025,
+      "step": 1674
+    },
+    {
+      "epoch": 0.11630329120955422,
+      "grad_norm": 5.29058880790818,
+      "learning_rate": 9.806212038817337e-06,
+      "loss": 0.9478,
+      "step": 1675
+    },
+    {
+      "epoch": 0.11637272601027634,
+      "grad_norm": 4.362487781678488,
+      "learning_rate": 9.805901890934489e-06,
+      "loss": 0.4006,
+      "step": 1676
+    },
+    {
+      "epoch": 0.11644216081099847,
+      "grad_norm": 4.251015453973613,
+      "learning_rate": 9.80559149997456e-06,
+      "loss": 0.5686,
+      "step": 1677
+    },
+    {
+      "epoch": 0.1165115956117206,
+      "grad_norm": 4.249796639268927,
+      "learning_rate": 9.80528086595325e-06,
+      "loss": 0.6019,
+      "step": 1678
+    },
+    {
+      "epoch": 0.11658103041244272,
+      "grad_norm": 5.358643557038087,
+      "learning_rate": 9.804969988886272e-06,
+      "loss": 0.5723,
+      "step": 1679
+    },
+    {
+      "epoch": 0.11665046521316484,
+      "grad_norm": 3.3736961957766143,
+      "learning_rate": 9.804658868789349e-06,
+      "loss": 0.3636,
+      "step": 1680
+    },
+    {
+      "epoch": 0.11671990001388696,
+      "grad_norm": 4.01816723815723,
+      "learning_rate": 9.80434750567822e-06,
+      "loss": 0.4961,
+      "step": 1681
+    },
+    {
+      "epoch": 0.11678933481460908,
+      "grad_norm": 4.302250390946548,
+      "learning_rate": 9.804035899568626e-06,
+      "loss": 0.4491,
+      "step": 1682
+    },
+    {
+      "epoch": 0.1168587696153312,
+      "grad_norm": 4.781545779769371,
+      "learning_rate": 9.803724050476335e-06,
+      "loss": 0.8137,
+      "step": 1683
+    },
+    {
+      "epoch": 0.11692820441605332,
+      "grad_norm": 4.094319295725907,
+      "learning_rate": 9.803411958417118e-06,
+      "loss": 0.5435,
+      "step": 1684
+    },
+    {
+      "epoch": 0.11699763921677545,
+      "grad_norm": 4.313402821208586,
+      "learning_rate": 9.803099623406757e-06,
+      "loss": 0.5524,
+      "step": 1685
+    },
+    {
+      "epoch": 0.11706707401749757,
+      "grad_norm": 5.282461204485641,
+      "learning_rate": 9.802787045461056e-06,
+      "loss": 0.645,
+      "step": 1686
+    },
+    {
+      "epoch": 0.11713650881821969,
+      "grad_norm": 3.672488360157587,
+      "learning_rate": 9.802474224595818e-06,
+      "loss": 0.3966,
+      "step": 1687
+    },
+    {
+      "epoch": 0.11720594361894181,
+      "grad_norm": 5.0527815413364845,
+      "learning_rate": 9.802161160826868e-06,
+      "loss": 0.8493,
+      "step": 1688
+    },
+    {
+      "epoch": 0.11727537841966394,
+      "grad_norm": 4.0520212681991365,
+      "learning_rate": 9.801847854170042e-06,
+      "loss": 0.3701,
+      "step": 1689
+    },
+    {
+      "epoch": 0.11734481322038606,
+      "grad_norm": 4.116258360688445,
+      "learning_rate": 9.801534304641184e-06,
+      "loss": 0.5303,
+      "step": 1690
+    },
+    {
+      "epoch": 0.11741424802110818,
+      "grad_norm": 3.4463356608170654,
+      "learning_rate": 9.801220512256155e-06,
+      "loss": 0.3537,
+      "step": 1691
+    },
+    {
+      "epoch": 0.1174836828218303,
+      "grad_norm": 4.944817681385152,
+      "learning_rate": 9.800906477030825e-06,
+      "loss": 0.6223,
+      "step": 1692
+    },
+    {
+      "epoch": 0.11755311762255242,
+      "grad_norm": 5.04652776842049,
+      "learning_rate": 9.800592198981077e-06,
+      "loss": 0.2809,
+      "step": 1693
+    },
+    {
+      "epoch": 0.11762255242327455,
+      "grad_norm": 3.208818525830077,
+      "learning_rate": 9.80027767812281e-06,
+      "loss": 0.4328,
+      "step": 1694
+    },
+    {
+      "epoch": 0.11769198722399667,
+      "grad_norm": 5.067224331419792,
+      "learning_rate": 9.799962914471927e-06,
+      "loss": 0.8128,
+      "step": 1695
+    },
+    {
+      "epoch": 0.11776142202471879,
+      "grad_norm": 3.3737912896805367,
+      "learning_rate": 9.799647908044355e-06,
+      "loss": 0.3542,
+      "step": 1696
+    },
+    {
+      "epoch": 0.11783085682544091,
+      "grad_norm": 4.270624352759668,
+      "learning_rate": 9.79933265885602e-06,
+      "loss": 0.4514,
+      "step": 1697
+    },
+    {
+      "epoch": 0.11790029162616303,
+      "grad_norm": 5.09791387047133,
+      "learning_rate": 9.79901716692287e-06,
+      "loss": 0.7495,
+      "step": 1698
+    },
+    {
+      "epoch": 0.11796972642688515,
+      "grad_norm": 4.066840800561533,
+      "learning_rate": 9.798701432260865e-06,
+      "loss": 0.3086,
+      "step": 1699
+    },
+    {
+      "epoch": 0.11803916122760727,
+      "grad_norm": 3.8076512506229694,
+      "learning_rate": 9.798385454885968e-06,
+      "loss": 0.468,
+      "step": 1700
+    },
+    {
+      "epoch": 0.1181085960283294,
+      "grad_norm": 3.7979216614517215,
+      "learning_rate": 9.798069234814167e-06,
+      "loss": 0.3929,
+      "step": 1701
+    },
+    {
+      "epoch": 0.11817803082905153,
+      "grad_norm": 2.574675289203647,
+      "learning_rate": 9.797752772061453e-06,
+      "loss": 0.2818,
+      "step": 1702
+    },
+    {
+      "epoch": 0.11824746562977365,
+      "grad_norm": 3.7641269228487366,
+      "learning_rate": 9.797436066643831e-06,
+      "loss": 0.379,
+      "step": 1703
+    },
+    {
+      "epoch": 0.11831690043049577,
+      "grad_norm": 5.46268523545497,
+      "learning_rate": 9.797119118577323e-06,
+      "loss": 0.8578,
+      "step": 1704
+    },
+    {
+      "epoch": 0.11838633523121789,
+      "grad_norm": 4.641459046859287,
+      "learning_rate": 9.79680192787796e-06,
+      "loss": 0.7158,
+      "step": 1705
+    },
+    {
+      "epoch": 0.11845577003194001,
+      "grad_norm": 4.289730359366163,
+      "learning_rate": 9.79648449456178e-06,
+      "loss": 0.6539,
+      "step": 1706
+    },
+    {
+      "epoch": 0.11852520483266213,
+      "grad_norm": 5.343407314120756,
+      "learning_rate": 9.796166818644843e-06,
+      "loss": 0.7804,
+      "step": 1707
+    },
+    {
+      "epoch": 0.11859463963338425,
+      "grad_norm": 4.778188601029679,
+      "learning_rate": 9.795848900143217e-06,
+      "loss": 0.4941,
+      "step": 1708
+    },
+    {
+      "epoch": 0.11866407443410637,
+      "grad_norm": 4.340651735285887,
+      "learning_rate": 9.795530739072979e-06,
+      "loss": 0.5169,
+      "step": 1709
+    },
+    {
+      "epoch": 0.11873350923482849,
+      "grad_norm": 4.799767648773151,
+      "learning_rate": 9.795212335450224e-06,
+      "loss": 0.598,
+      "step": 1710
+    },
+    {
+      "epoch": 0.11880294403555061,
+      "grad_norm": 4.447824518809235,
+      "learning_rate": 9.794893689291054e-06,
+      "loss": 0.4449,
+      "step": 1711
+    },
+    {
+      "epoch": 0.11887237883627275,
+      "grad_norm": 4.846488817884635,
+      "learning_rate": 9.794574800611588e-06,
+      "loss": 0.4941,
+      "step": 1712
+    },
+    {
+      "epoch": 0.11894181363699487,
+      "grad_norm": 3.5224235747276644,
+      "learning_rate": 9.794255669427952e-06,
+      "loss": 0.3266,
+      "step": 1713
+    },
+    {
+      "epoch": 0.11901124843771699,
+      "grad_norm": 4.827867599440925,
+      "learning_rate": 9.793936295756292e-06,
+      "loss": 0.5808,
+      "step": 1714
+    },
+    {
+      "epoch": 0.11908068323843911,
+      "grad_norm": 5.040162625644233,
+      "learning_rate": 9.793616679612756e-06,
+      "loss": 0.4275,
+      "step": 1715
+    },
+    {
+      "epoch": 0.11915011803916123,
+      "grad_norm": 4.876503092833679,
+      "learning_rate": 9.793296821013515e-06,
+      "loss": 0.8744,
+      "step": 1716
+    },
+    {
+      "epoch": 0.11921955283988335,
+      "grad_norm": 5.165187735288729,
+      "learning_rate": 9.792976719974744e-06,
+      "loss": 0.7894,
+      "step": 1717
+    },
+    {
+      "epoch": 0.11928898764060547,
+      "grad_norm": 3.5660370719822523,
+      "learning_rate": 9.792656376512635e-06,
+      "loss": 0.4372,
+      "step": 1718
+    },
+    {
+      "epoch": 0.11935842244132759,
+      "grad_norm": 4.716438597867215,
+      "learning_rate": 9.792335790643387e-06,
+      "loss": 0.4876,
+      "step": 1719
+    },
+    {
+      "epoch": 0.11942785724204971,
+      "grad_norm": 3.6117060114192534,
+      "learning_rate": 9.792014962383221e-06,
+      "loss": 0.4416,
+      "step": 1720
+    },
+    {
+      "epoch": 0.11949729204277183,
+      "grad_norm": 4.308363500310813,
+      "learning_rate": 9.79169389174836e-06,
+      "loss": 0.5443,
+      "step": 1721
+    },
+    {
+      "epoch": 0.11956672684349395,
+      "grad_norm": 4.461284097438357,
+      "learning_rate": 9.791372578755044e-06,
+      "loss": 0.3977,
+      "step": 1722
+    },
+    {
+      "epoch": 0.11963616164421609,
+      "grad_norm": 3.8188951864622127,
+      "learning_rate": 9.791051023419522e-06,
+      "loss": 0.4713,
+      "step": 1723
+    },
+    {
+      "epoch": 0.11970559644493821,
+      "grad_norm": 4.870742333437552,
+      "learning_rate": 9.790729225758064e-06,
+      "loss": 0.535,
+      "step": 1724
+    },
+    {
+      "epoch": 0.11977503124566033,
+      "grad_norm": 3.6518919420761073,
+      "learning_rate": 9.79040718578694e-06,
+      "loss": 0.2369,
+      "step": 1725
+    },
+    {
+      "epoch": 0.11984446604638245,
+      "grad_norm": 4.302248853074158,
+      "learning_rate": 9.790084903522444e-06,
+      "loss": 0.4053,
+      "step": 1726
+    },
+    {
+      "epoch": 0.11991390084710457,
+      "grad_norm": 3.6731511625912674,
+      "learning_rate": 9.789762378980874e-06,
+      "loss": 0.4124,
+      "step": 1727
+    },
+    {
+      "epoch": 0.11998333564782669,
+      "grad_norm": 4.702931511904636,
+      "learning_rate": 9.789439612178541e-06,
+      "loss": 0.5971,
+      "step": 1728
+    },
+    {
+      "epoch": 0.12005277044854881,
+      "grad_norm": 4.457762639450602,
+      "learning_rate": 9.789116603131773e-06,
+      "loss": 0.3287,
+      "step": 1729
+    },
+    {
+      "epoch": 0.12012220524927093,
+      "grad_norm": 3.875600980576659,
+      "learning_rate": 9.788793351856906e-06,
+      "loss": 0.3351,
+      "step": 1730
+    },
+    {
+      "epoch": 0.12019164004999305,
+      "grad_norm": 5.06322354064501,
+      "learning_rate": 9.788469858370292e-06,
+      "loss": 0.5239,
+      "step": 1731
+    },
+    {
+      "epoch": 0.12026107485071517,
+      "grad_norm": 5.732136164714776,
+      "learning_rate": 9.788146122688289e-06,
+      "loss": 0.5588,
+      "step": 1732
+    },
+    {
+      "epoch": 0.1203305096514373,
+      "grad_norm": 3.7812660863601515,
+      "learning_rate": 9.787822144827272e-06,
+      "loss": 0.4961,
+      "step": 1733
+    },
+    {
+      "epoch": 0.12039994445215942,
+      "grad_norm": 4.509362506992422,
+      "learning_rate": 9.78749792480363e-06,
+      "loss": 0.5193,
+      "step": 1734
+    },
+    {
+      "epoch": 0.12046937925288155,
+      "grad_norm": 4.7182803821630275,
+      "learning_rate": 9.78717346263376e-06,
+      "loss": 0.5364,
+      "step": 1735
+    },
+    {
+      "epoch": 0.12053881405360367,
+      "grad_norm": 5.197037297552331,
+      "learning_rate": 9.786848758334075e-06,
+      "loss": 0.8427,
+      "step": 1736
+    },
+    {
+      "epoch": 0.12060824885432579,
+      "grad_norm": 5.3935377930770265,
+      "learning_rate": 9.786523811920995e-06,
+      "loss": 0.6223,
+      "step": 1737
+    },
+    {
+      "epoch": 0.12067768365504791,
+      "grad_norm": 4.019467386284341,
+      "learning_rate": 9.786198623410958e-06,
+      "loss": 0.6432,
+      "step": 1738
+    },
+    {
+      "epoch": 0.12074711845577003,
+      "grad_norm": 3.8245309085058263,
+      "learning_rate": 9.785873192820411e-06,
+      "loss": 0.3583,
+      "step": 1739
+    },
+    {
+      "epoch": 0.12081655325649215,
+      "grad_norm": 4.697094246983288,
+      "learning_rate": 9.785547520165813e-06,
+      "loss": 0.5494,
+      "step": 1740
+    },
+    {
+      "epoch": 0.12088598805721427,
+      "grad_norm": 5.71910329169793,
+      "learning_rate": 9.785221605463636e-06,
+      "loss": 0.714,
+      "step": 1741
+    },
+    {
+      "epoch": 0.1209554228579364,
+      "grad_norm": 4.623625250642741,
+      "learning_rate": 9.784895448730366e-06,
+      "loss": 0.6718,
+      "step": 1742
+    },
+    {
+      "epoch": 0.12102485765865852,
+      "grad_norm": 3.6374407594337455,
+      "learning_rate": 9.784569049982497e-06,
+      "loss": 0.3915,
+      "step": 1743
+    },
+    {
+      "epoch": 0.12109429245938064,
+      "grad_norm": 5.221373600544769,
+      "learning_rate": 9.78424240923654e-06,
+      "loss": 0.5445,
+      "step": 1744
+    },
+    {
+      "epoch": 0.12116372726010276,
+      "grad_norm": 5.024479846322383,
+      "learning_rate": 9.783915526509016e-06,
+      "loss": 0.5985,
+      "step": 1745
+    },
+    {
+      "epoch": 0.12123316206082489,
+      "grad_norm": 4.64495320212532,
+      "learning_rate": 9.78358840181646e-06,
+      "loss": 0.4452,
+      "step": 1746
+    },
+    {
+      "epoch": 0.12130259686154701,
+      "grad_norm": 5.8191971805827185,
+      "learning_rate": 9.783261035175413e-06,
+      "loss": 0.685,
+      "step": 1747
+    },
+    {
+      "epoch": 0.12137203166226913,
+      "grad_norm": 2.8720120588204,
+      "learning_rate": 9.782933426602436e-06,
+      "loss": 0.2025,
+      "step": 1748
+    },
+    {
+      "epoch": 0.12144146646299125,
+      "grad_norm": 3.265447671056633,
+      "learning_rate": 9.7826055761141e-06,
+      "loss": 0.3473,
+      "step": 1749
+    },
+    {
+      "epoch": 0.12151090126371338,
+      "grad_norm": 5.093259054752176,
+      "learning_rate": 9.782277483726984e-06,
+      "loss": 0.8673,
+      "step": 1750
+    },
+    {
+      "epoch": 0.1215803360644355,
+      "grad_norm": 5.420568448158613,
+      "learning_rate": 9.781949149457686e-06,
+      "loss": 0.6675,
+      "step": 1751
+    },
+    {
+      "epoch": 0.12164977086515762,
+      "grad_norm": 3.3841345040308566,
+      "learning_rate": 9.78162057332281e-06,
+      "loss": 0.4042,
+      "step": 1752
+    },
+    {
+      "epoch": 0.12171920566587974,
+      "grad_norm": 3.0575130032376565,
+      "learning_rate": 9.781291755338975e-06,
+      "loss": 0.2801,
+      "step": 1753
+    },
+    {
+      "epoch": 0.12178864046660186,
+      "grad_norm": 3.8175123389163264,
+      "learning_rate": 9.780962695522815e-06,
+      "loss": 0.3632,
+      "step": 1754
+    },
+    {
+      "epoch": 0.12185807526732398,
+      "grad_norm": 5.414279726866016,
+      "learning_rate": 9.780633393890972e-06,
+      "loss": 0.6495,
+      "step": 1755
+    },
+    {
+      "epoch": 0.1219275100680461,
+      "grad_norm": 4.355123148881332,
+      "learning_rate": 9.780303850460102e-06,
+      "loss": 0.447,
+      "step": 1756
+    },
+    {
+      "epoch": 0.12199694486876822,
+      "grad_norm": 4.569654746824713,
+      "learning_rate": 9.779974065246872e-06,
+      "loss": 0.3025,
+      "step": 1757
+    },
+    {
+      "epoch": 0.12206637966949035,
+      "grad_norm": 4.862548918327962,
+      "learning_rate": 9.779644038267962e-06,
+      "loss": 0.6075,
+      "step": 1758
+    },
+    {
+      "epoch": 0.12213581447021248,
+      "grad_norm": 2.546465191252702,
+      "learning_rate": 9.779313769540066e-06,
+      "loss": 0.3151,
+      "step": 1759
+    },
+    {
+      "epoch": 0.1222052492709346,
+      "grad_norm": 3.9465978019141494,
+      "learning_rate": 9.778983259079886e-06,
+      "loss": 0.461,
+      "step": 1760
+    },
+    {
+      "epoch": 0.12227468407165672,
+      "grad_norm": 4.489098934043291,
+      "learning_rate": 9.778652506904142e-06,
+      "loss": 0.8292,
+      "step": 1761
+    },
+    {
+      "epoch": 0.12234411887237884,
+      "grad_norm": 4.0408485298165875,
+      "learning_rate": 9.778321513029564e-06,
+      "loss": 0.4274,
+      "step": 1762
+    },
+    {
+      "epoch": 0.12241355367310096,
+      "grad_norm": 4.273000573877079,
+      "learning_rate": 9.777990277472887e-06,
+      "loss": 0.3045,
+      "step": 1763
+    },
+    {
+      "epoch": 0.12248298847382308,
+      "grad_norm": 2.751467069584683,
+      "learning_rate": 9.777658800250868e-06,
+      "loss": 0.3264,
+      "step": 1764
+    },
+    {
+      "epoch": 0.1225524232745452,
+      "grad_norm": 4.155573166386027,
+      "learning_rate": 9.777327081380277e-06,
+      "loss": 0.4696,
+      "step": 1765
+    },
+    {
+      "epoch": 0.12262185807526732,
+      "grad_norm": 4.263280283126045,
+      "learning_rate": 9.776995120877886e-06,
+      "loss": 0.6178,
+      "step": 1766
+    },
+    {
+      "epoch": 0.12269129287598944,
+      "grad_norm": 4.067418782200545,
+      "learning_rate": 9.776662918760487e-06,
+      "loss": 0.4424,
+      "step": 1767
+    },
+    {
+      "epoch": 0.12276072767671156,
+      "grad_norm": 5.116189955115654,
+      "learning_rate": 9.776330475044883e-06,
+      "loss": 0.8439,
+      "step": 1768
+    },
+    {
+      "epoch": 0.1228301624774337,
+      "grad_norm": 4.352649030954681,
+      "learning_rate": 9.775997789747888e-06,
+      "loss": 0.6252,
+      "step": 1769
+    },
+    {
+      "epoch": 0.12289959727815582,
+      "grad_norm": 4.826425145081024,
+      "learning_rate": 9.775664862886331e-06,
+      "loss": 0.6233,
+      "step": 1770
+    },
+    {
+      "epoch": 0.12296903207887794,
+      "grad_norm": 4.184276031581763,
+      "learning_rate": 9.775331694477047e-06,
+      "loss": 0.566,
+      "step": 1771
+    },
+    {
+      "epoch": 0.12303846687960006,
+      "grad_norm": 4.0849420870839905,
+      "learning_rate": 9.774998284536892e-06,
+      "loss": 0.4232,
+      "step": 1772
+    },
+    {
+      "epoch": 0.12310790168032218,
+      "grad_norm": 3.7174346496372945,
+      "learning_rate": 9.774664633082724e-06,
+      "loss": 0.4009,
+      "step": 1773
+    },
+    {
+      "epoch": 0.1231773364810443,
+      "grad_norm": 4.186400117553004,
+      "learning_rate": 9.774330740131424e-06,
+      "loss": 0.4533,
+      "step": 1774
+    },
+    {
+      "epoch": 0.12324677128176642,
+      "grad_norm": 4.057531497930357,
+      "learning_rate": 9.773996605699876e-06,
+      "loss": 0.4657,
+      "step": 1775
+    },
+    {
+      "epoch": 0.12331620608248854,
+      "grad_norm": 4.868401522510038,
+      "learning_rate": 9.773662229804984e-06,
+      "loss": 0.5264,
+      "step": 1776
+    },
+    {
+      "epoch": 0.12338564088321066,
+      "grad_norm": 4.606773982600506,
+      "learning_rate": 9.773327612463656e-06,
+      "loss": 0.5314,
+      "step": 1777
+    },
+    {
+      "epoch": 0.12345507568393278,
+      "grad_norm": 3.7509095111252355,
+      "learning_rate": 9.772992753692821e-06,
+      "loss": 0.3741,
+      "step": 1778
+    },
+    {
+      "epoch": 0.1235245104846549,
+      "grad_norm": 4.4145984250899675,
+      "learning_rate": 9.772657653509412e-06,
+      "loss": 0.7466,
+      "step": 1779
+    },
+    {
+      "epoch": 0.12359394528537702,
+      "grad_norm": 4.283551128463392,
+      "learning_rate": 9.77232231193038e-06,
+      "loss": 0.4786,
+      "step": 1780
+    },
+    {
+      "epoch": 0.12366338008609916,
+      "grad_norm": 4.5007352726891225,
+      "learning_rate": 9.771986728972684e-06,
+      "loss": 0.3799,
+      "step": 1781
+    },
+    {
+      "epoch": 0.12373281488682128,
+      "grad_norm": 5.258005770112338,
+      "learning_rate": 9.771650904653302e-06,
+      "loss": 0.7578,
+      "step": 1782
+    },
+    {
+      "epoch": 0.1238022496875434,
+      "grad_norm": 3.1455804420815454,
+      "learning_rate": 9.771314838989215e-06,
+      "loss": 0.3213,
+      "step": 1783
+    },
+    {
+      "epoch": 0.12387168448826552,
+      "grad_norm": 4.204647422245746,
+      "learning_rate": 9.770978531997423e-06,
+      "loss": 0.5212,
+      "step": 1784
+    },
+    {
+      "epoch": 0.12394111928898764,
+      "grad_norm": 5.062125799064982,
+      "learning_rate": 9.770641983694935e-06,
+      "loss": 0.8633,
+      "step": 1785
+    },
+    {
+      "epoch": 0.12401055408970976,
+      "grad_norm": 5.350570800154277,
+      "learning_rate": 9.770305194098774e-06,
+      "loss": 0.5724,
+      "step": 1786
+    },
+    {
+      "epoch": 0.12407998889043188,
+      "grad_norm": 5.038263677308377,
+      "learning_rate": 9.769968163225975e-06,
+      "loss": 0.7449,
+      "step": 1787
+    },
+    {
+      "epoch": 0.124149423691154,
+      "grad_norm": 3.522527657097216,
+      "learning_rate": 9.769630891093583e-06,
+      "loss": 0.4395,
+      "step": 1788
+    },
+    {
+      "epoch": 0.12421885849187612,
+      "grad_norm": 4.836971375159003,
+      "learning_rate": 9.769293377718658e-06,
+      "loss": 0.8029,
+      "step": 1789
+    },
+    {
+      "epoch": 0.12428829329259825,
+      "grad_norm": 2.6164357002141467,
+      "learning_rate": 9.768955623118273e-06,
+      "loss": 0.3406,
+      "step": 1790
+    },
+    {
+      "epoch": 0.12435772809332037,
+      "grad_norm": 4.253442700254873,
+      "learning_rate": 9.768617627309505e-06,
+      "loss": 0.672,
+      "step": 1791
+    },
+    {
+      "epoch": 0.1244271628940425,
+      "grad_norm": 3.9975407517883763,
+      "learning_rate": 9.768279390309457e-06,
+      "loss": 0.4902,
+      "step": 1792
+    },
+    {
+      "epoch": 0.12449659769476462,
+      "grad_norm": 3.7615750603518534,
+      "learning_rate": 9.76794091213523e-06,
+      "loss": 0.4263,
+      "step": 1793
+    },
+    {
+      "epoch": 0.12456603249548674,
+      "grad_norm": 3.4481392328629727,
+      "learning_rate": 9.767602192803948e-06,
+      "loss": 0.3783,
+      "step": 1794
+    },
+    {
+      "epoch": 0.12463546729620886,
+      "grad_norm": 3.6165197489946315,
+      "learning_rate": 9.767263232332744e-06,
+      "loss": 0.3354,
+      "step": 1795
+    },
+    {
+      "epoch": 0.12470490209693098,
+      "grad_norm": 5.258884734657244,
+      "learning_rate": 9.766924030738758e-06,
+      "loss": 0.7447,
+      "step": 1796
+    },
+    {
+      "epoch": 0.1247743368976531,
+      "grad_norm": 7.004388794378471,
+      "learning_rate": 9.76658458803915e-06,
+      "loss": 0.6935,
+      "step": 1797
+    },
+    {
+      "epoch": 0.12484377169837522,
+      "grad_norm": 4.324621150544358,
+      "learning_rate": 9.766244904251087e-06,
+      "loss": 0.5644,
+      "step": 1798
+    },
+    {
+      "epoch": 0.12491320649909735,
+      "grad_norm": 4.43271847031387,
+      "learning_rate": 9.765904979391749e-06,
+      "loss": 0.4704,
+      "step": 1799
+    },
+    {
+      "epoch": 0.12498264129981947,
+      "grad_norm": 4.556371616407555,
+      "learning_rate": 9.765564813478332e-06,
+      "loss": 0.4905,
+      "step": 1800
+    },
+    {
+      "epoch": 0.1250520761005416,
+      "grad_norm": 3.172510936296308,
+      "learning_rate": 9.765224406528037e-06,
+      "loss": 0.2841,
+      "step": 1801
+    },
+    {
+      "epoch": 0.12512151090126372,
+      "grad_norm": 4.892158968318668,
+      "learning_rate": 9.764883758558086e-06,
+      "loss": 0.4777,
+      "step": 1802
+    },
+    {
+      "epoch": 0.12519094570198583,
+      "grad_norm": 3.5949272436689097,
+      "learning_rate": 9.764542869585705e-06,
+      "loss": 0.3131,
+      "step": 1803
+    },
+    {
+      "epoch": 0.12526038050270796,
+      "grad_norm": 4.962337124079528,
+      "learning_rate": 9.764201739628136e-06,
+      "loss": 0.5364,
+      "step": 1804
+    },
+    {
+      "epoch": 0.12532981530343007,
+      "grad_norm": 5.73503737180503,
+      "learning_rate": 9.763860368702638e-06,
+      "loss": 0.6831,
+      "step": 1805
+    },
+    {
+      "epoch": 0.1253992501041522,
+      "grad_norm": 2.978771160888224,
+      "learning_rate": 9.76351875682647e-06,
+      "loss": 0.3032,
+      "step": 1806
+    },
+    {
+      "epoch": 0.1254686849048743,
+      "grad_norm": 3.9247400718152354,
+      "learning_rate": 9.763176904016914e-06,
+      "loss": 0.5178,
+      "step": 1807
+    },
+    {
+      "epoch": 0.12553811970559645,
+      "grad_norm": 4.341900589246379,
+      "learning_rate": 9.762834810291259e-06,
+      "loss": 0.3188,
+      "step": 1808
+    },
+    {
+      "epoch": 0.12560755450631858,
+      "grad_norm": 5.784727291764583,
+      "learning_rate": 9.762492475666811e-06,
+      "loss": 0.846,
+      "step": 1809
+    },
+    {
+      "epoch": 0.1256769893070407,
+      "grad_norm": 3.310205368605784,
+      "learning_rate": 9.762149900160881e-06,
+      "loss": 0.3303,
+      "step": 1810
+    },
+    {
+      "epoch": 0.12574642410776282,
+      "grad_norm": 4.342776512321875,
+      "learning_rate": 9.761807083790799e-06,
+      "loss": 0.4553,
+      "step": 1811
+    },
+    {
+      "epoch": 0.12581585890848493,
+      "grad_norm": 5.163083634892525,
+      "learning_rate": 9.761464026573903e-06,
+      "loss": 0.742,
+      "step": 1812
+    },
+    {
+      "epoch": 0.12588529370920706,
+      "grad_norm": 3.7469870277442197,
+      "learning_rate": 9.761120728527544e-06,
+      "loss": 0.4785,
+      "step": 1813
+    },
+    {
+      "epoch": 0.12595472850992917,
+      "grad_norm": 5.094743515886799,
+      "learning_rate": 9.760777189669087e-06,
+      "loss": 0.6243,
+      "step": 1814
+    },
+    {
+      "epoch": 0.1260241633106513,
+      "grad_norm": 5.3348630084182895,
+      "learning_rate": 9.760433410015906e-06,
+      "loss": 0.757,
+      "step": 1815
+    },
+    {
+      "epoch": 0.1260935981113734,
+      "grad_norm": 4.218047348561074,
+      "learning_rate": 9.76008938958539e-06,
+      "loss": 0.4869,
+      "step": 1816
+    },
+    {
+      "epoch": 0.12616303291209555,
+      "grad_norm": 3.434367644257062,
+      "learning_rate": 9.75974512839494e-06,
+      "loss": 0.3012,
+      "step": 1817
+    },
+    {
+      "epoch": 0.12623246771281765,
+      "grad_norm": 4.290369411886893,
+      "learning_rate": 9.759400626461966e-06,
+      "loss": 0.5693,
+      "step": 1818
+    },
+    {
+      "epoch": 0.1263019025135398,
+      "grad_norm": 4.177294307507477,
+      "learning_rate": 9.759055883803896e-06,
+      "loss": 0.5852,
+      "step": 1819
+    },
+    {
+      "epoch": 0.1263713373142619,
+      "grad_norm": 4.12931338919358,
+      "learning_rate": 9.758710900438162e-06,
+      "loss": 0.4204,
+      "step": 1820
+    },
+    {
+      "epoch": 0.12644077211498403,
+      "grad_norm": 5.25275519473393,
+      "learning_rate": 9.758365676382218e-06,
+      "loss": 0.5688,
+      "step": 1821
+    },
+    {
+      "epoch": 0.12651020691570616,
+      "grad_norm": 4.761388335309499,
+      "learning_rate": 9.75802021165352e-06,
+      "loss": 0.6799,
+      "step": 1822
+    },
+    {
+      "epoch": 0.12657964171642827,
+      "grad_norm": 4.01693446713266,
+      "learning_rate": 9.757674506269545e-06,
+      "loss": 0.3772,
+      "step": 1823
+    },
+    {
+      "epoch": 0.1266490765171504,
+      "grad_norm": 4.754351366456632,
+      "learning_rate": 9.757328560247777e-06,
+      "loss": 0.5844,
+      "step": 1824
+    },
+    {
+      "epoch": 0.1267185113178725,
+      "grad_norm": 4.2931784997254505,
+      "learning_rate": 9.756982373605715e-06,
+      "loss": 0.6067,
+      "step": 1825
+    },
+    {
+      "epoch": 0.12678794611859465,
+      "grad_norm": 3.762348714479913,
+      "learning_rate": 9.756635946360867e-06,
+      "loss": 0.5811,
+      "step": 1826
+    },
+    {
+      "epoch": 0.12685738091931675,
+      "grad_norm": 4.360583412209665,
+      "learning_rate": 9.756289278530754e-06,
+      "loss": 0.5136,
+      "step": 1827
+    },
+    {
+      "epoch": 0.1269268157200389,
+      "grad_norm": 5.868709425117602,
+      "learning_rate": 9.75594237013291e-06,
+      "loss": 0.7776,
+      "step": 1828
+    },
+    {
+      "epoch": 0.126996250520761,
+      "grad_norm": 3.936409560835182,
+      "learning_rate": 9.755595221184887e-06,
+      "loss": 0.4506,
+      "step": 1829
+    },
+    {
+      "epoch": 0.12706568532148313,
+      "grad_norm": 4.707009052812682,
+      "learning_rate": 9.755247831704236e-06,
+      "loss": 0.7754,
+      "step": 1830
+    },
+    {
+      "epoch": 0.12713512012220524,
+      "grad_norm": 3.8013013531831272,
+      "learning_rate": 9.754900201708532e-06,
+      "loss": 0.3204,
+      "step": 1831
+    },
+    {
+      "epoch": 0.12720455492292737,
+      "grad_norm": 3.617958678969524,
+      "learning_rate": 9.754552331215354e-06,
+      "loss": 0.4046,
+      "step": 1832
+    },
+    {
+      "epoch": 0.1272739897236495,
+      "grad_norm": 4.630773013586941,
+      "learning_rate": 9.754204220242303e-06,
+      "loss": 0.5782,
+      "step": 1833
+    },
+    {
+      "epoch": 0.1273434245243716,
+      "grad_norm": 4.569606148302303,
+      "learning_rate": 9.753855868806978e-06,
+      "loss": 0.8256,
+      "step": 1834
+    },
+    {
+      "epoch": 0.12741285932509375,
+      "grad_norm": 3.843743315787575,
+      "learning_rate": 9.753507276927005e-06,
+      "loss": 0.3633,
+      "step": 1835
+    },
+    {
+      "epoch": 0.12748229412581585,
+      "grad_norm": 4.787536272790188,
+      "learning_rate": 9.753158444620013e-06,
+      "loss": 0.6483,
+      "step": 1836
+    },
+    {
+      "epoch": 0.127551728926538,
+      "grad_norm": 4.7951874554992076,
+      "learning_rate": 9.752809371903645e-06,
+      "loss": 0.4415,
+      "step": 1837
+    },
+    {
+      "epoch": 0.1276211637272601,
+      "grad_norm": 4.61157766036463,
+      "learning_rate": 9.752460058795558e-06,
+      "loss": 0.5129,
+      "step": 1838
+    },
+    {
+      "epoch": 0.12769059852798223,
+      "grad_norm": 3.5377934082312588,
+      "learning_rate": 9.752110505313419e-06,
+      "loss": 0.3025,
+      "step": 1839
+    },
+    {
+      "epoch": 0.12776003332870434,
+      "grad_norm": 5.0961048792962265,
+      "learning_rate": 9.751760711474907e-06,
+      "loss": 0.6219,
+      "step": 1840
+    },
+    {
+      "epoch": 0.12782946812942647,
+      "grad_norm": 3.330631129108507,
+      "learning_rate": 9.751410677297714e-06,
+      "loss": 0.2671,
+      "step": 1841
+    },
+    {
+      "epoch": 0.12789890293014858,
+      "grad_norm": 5.846752266653663,
+      "learning_rate": 9.751060402799548e-06,
+      "loss": 0.5587,
+      "step": 1842
+    },
+    {
+      "epoch": 0.1279683377308707,
+      "grad_norm": 5.190908305769495,
+      "learning_rate": 9.750709887998123e-06,
+      "loss": 0.4654,
+      "step": 1843
+    },
+    {
+      "epoch": 0.12803777253159285,
+      "grad_norm": 4.15082659834225,
+      "learning_rate": 9.750359132911165e-06,
+      "loss": 0.49,
+      "step": 1844
+    },
+    {
+      "epoch": 0.12810720733231495,
+      "grad_norm": 5.725666465139423,
+      "learning_rate": 9.75000813755642e-06,
+      "loss": 0.4922,
+      "step": 1845
+    },
+    {
+      "epoch": 0.1281766421330371,
+      "grad_norm": 5.884857658741604,
+      "learning_rate": 9.749656901951637e-06,
+      "loss": 0.6666,
+      "step": 1846
+    },
+    {
+      "epoch": 0.1282460769337592,
+      "grad_norm": 5.1766292626022885,
+      "learning_rate": 9.749305426114583e-06,
+      "loss": 0.4867,
+      "step": 1847
+    },
+    {
+      "epoch": 0.12831551173448133,
+      "grad_norm": 4.466225087796118,
+      "learning_rate": 9.748953710063033e-06,
+      "loss": 0.5987,
+      "step": 1848
+    },
+    {
+      "epoch": 0.12838494653520344,
+      "grad_norm": 4.115171277959829,
+      "learning_rate": 9.748601753814782e-06,
+      "loss": 0.4159,
+      "step": 1849
+    },
+    {
+      "epoch": 0.12845438133592557,
+      "grad_norm": 5.458812641659435,
+      "learning_rate": 9.748249557387624e-06,
+      "loss": 0.687,
+      "step": 1850
+    },
+    {
+      "epoch": 0.12852381613664768,
+      "grad_norm": 6.452974341830799,
+      "learning_rate": 9.747897120799379e-06,
+      "loss": 0.5378,
+      "step": 1851
+    },
+    {
+      "epoch": 0.1285932509373698,
+      "grad_norm": 2.873414426968416,
+      "learning_rate": 9.74754444406787e-06,
+      "loss": 0.2825,
+      "step": 1852
+    },
+    {
+      "epoch": 0.12866268573809192,
+      "grad_norm": 6.296518029614031,
+      "learning_rate": 9.747191527210932e-06,
+      "loss": 1.0101,
+      "step": 1853
+    },
+    {
+      "epoch": 0.12873212053881405,
+      "grad_norm": 4.89988189496897,
+      "learning_rate": 9.746838370246421e-06,
+      "loss": 0.6278,
+      "step": 1854
+    },
+    {
+      "epoch": 0.1288015553395362,
+      "grad_norm": 4.199057286242706,
+      "learning_rate": 9.746484973192196e-06,
+      "loss": 0.5515,
+      "step": 1855
+    },
+    {
+      "epoch": 0.1288709901402583,
+      "grad_norm": 5.0511284510622865,
+      "learning_rate": 9.746131336066134e-06,
+      "loss": 0.5065,
+      "step": 1856
+    },
+    {
+      "epoch": 0.12894042494098043,
+      "grad_norm": 5.2781893494311625,
+      "learning_rate": 9.745777458886118e-06,
+      "loss": 0.5325,
+      "step": 1857
+    },
+    {
+      "epoch": 0.12900985974170254,
+      "grad_norm": 3.4295011204261514,
+      "learning_rate": 9.745423341670048e-06,
+      "loss": 0.4434,
+      "step": 1858
+    },
+    {
+      "epoch": 0.12907929454242467,
+      "grad_norm": 3.048474282969756,
+      "learning_rate": 9.745068984435835e-06,
+      "loss": 0.2637,
+      "step": 1859
+    },
+    {
+      "epoch": 0.12914872934314678,
+      "grad_norm": 2.6657393938115,
+      "learning_rate": 9.744714387201403e-06,
+      "loss": 0.2568,
+      "step": 1860
+    },
+    {
+      "epoch": 0.1292181641438689,
+      "grad_norm": 3.8898655774511472,
+      "learning_rate": 9.744359549984687e-06,
+      "loss": 0.4741,
+      "step": 1861
+    },
+    {
+      "epoch": 0.12928759894459102,
+      "grad_norm": 3.9785119962155897,
+      "learning_rate": 9.744004472803631e-06,
+      "loss": 0.5367,
+      "step": 1862
+    },
+    {
+      "epoch": 0.12935703374531315,
+      "grad_norm": 4.634057145852632,
+      "learning_rate": 9.7436491556762e-06,
+      "loss": 0.6075,
+      "step": 1863
+    },
+    {
+      "epoch": 0.12942646854603526,
+      "grad_norm": 3.437199570037133,
+      "learning_rate": 9.74329359862036e-06,
+      "loss": 0.3462,
+      "step": 1864
+    },
+    {
+      "epoch": 0.1294959033467574,
+      "grad_norm": 4.0292865726279095,
+      "learning_rate": 9.7429378016541e-06,
+      "loss": 0.5191,
+      "step": 1865
+    },
+    {
+      "epoch": 0.1295653381474795,
+      "grad_norm": 5.342536280165616,
+      "learning_rate": 9.742581764795411e-06,
+      "loss": 0.5465,
+      "step": 1866
+    },
+    {
+      "epoch": 0.12963477294820164,
+      "grad_norm": 4.56825366376222,
+      "learning_rate": 9.742225488062302e-06,
+      "loss": 0.6381,
+      "step": 1867
+    },
+    {
+      "epoch": 0.12970420774892377,
+      "grad_norm": 3.8985818298212855,
+      "learning_rate": 9.741868971472797e-06,
+      "loss": 0.4097,
+      "step": 1868
+    },
+    {
+      "epoch": 0.12977364254964588,
+      "grad_norm": 4.464710102008608,
+      "learning_rate": 9.741512215044922e-06,
+      "loss": 0.5609,
+      "step": 1869
+    },
+    {
+      "epoch": 0.129843077350368,
+      "grad_norm": 4.350304441014219,
+      "learning_rate": 9.741155218796728e-06,
+      "loss": 0.5476,
+      "step": 1870
+    },
+    {
+      "epoch": 0.12991251215109012,
+      "grad_norm": 3.596483734332975,
+      "learning_rate": 9.740797982746264e-06,
+      "loss": 0.3873,
+      "step": 1871
+    },
+    {
+      "epoch": 0.12998194695181225,
+      "grad_norm": 3.7896218631946983,
+      "learning_rate": 9.740440506911606e-06,
+      "loss": 0.373,
+      "step": 1872
+    },
+    {
+      "epoch": 0.13005138175253436,
+      "grad_norm": 3.3792507105550134,
+      "learning_rate": 9.74008279131083e-06,
+      "loss": 0.2868,
+      "step": 1873
+    },
+    {
+      "epoch": 0.1301208165532565,
+      "grad_norm": 3.9974235776743607,
+      "learning_rate": 9.73972483596203e-06,
+      "loss": 0.5502,
+      "step": 1874
+    },
+    {
+      "epoch": 0.1301902513539786,
+      "grad_norm": 3.0903653975396135,
+      "learning_rate": 9.73936664088331e-06,
+      "loss": 0.1898,
+      "step": 1875
+    },
+    {
+      "epoch": 0.13025968615470074,
+      "grad_norm": 3.5899348209446544,
+      "learning_rate": 9.73900820609279e-06,
+      "loss": 0.3208,
+      "step": 1876
+    },
+    {
+      "epoch": 0.13032912095542284,
+      "grad_norm": 4.788605177701336,
+      "learning_rate": 9.738649531608597e-06,
+      "loss": 0.7412,
+      "step": 1877
+    },
+    {
+      "epoch": 0.13039855575614498,
+      "grad_norm": 4.741830076454195,
+      "learning_rate": 9.738290617448874e-06,
+      "loss": 0.4594,
+      "step": 1878
+    },
+    {
+      "epoch": 0.13046799055686711,
+      "grad_norm": 5.789563584742074,
+      "learning_rate": 9.737931463631771e-06,
+      "loss": 0.6426,
+      "step": 1879
+    },
+    {
+      "epoch": 0.13053742535758922,
+      "grad_norm": 8.694320870163759,
+      "learning_rate": 9.737572070175458e-06,
+      "loss": 0.3904,
+      "step": 1880
+    },
+    {
+      "epoch": 0.13060686015831136,
+      "grad_norm": 3.6509275315791023,
+      "learning_rate": 9.73721243709811e-06,
+      "loss": 0.3424,
+      "step": 1881
+    },
+    {
+      "epoch": 0.13067629495903346,
+      "grad_norm": 4.495602856398995,
+      "learning_rate": 9.736852564417916e-06,
+      "loss": 0.6616,
+      "step": 1882
+    },
+    {
+      "epoch": 0.1307457297597556,
+      "grad_norm": 5.707796752110541,
+      "learning_rate": 9.736492452153081e-06,
+      "loss": 0.784,
+      "step": 1883
+    },
+    {
+      "epoch": 0.1308151645604777,
+      "grad_norm": 5.191613763869277,
+      "learning_rate": 9.736132100321817e-06,
+      "loss": 0.7692,
+      "step": 1884
+    },
+    {
+      "epoch": 0.13088459936119984,
+      "grad_norm": 3.112531467470029,
+      "learning_rate": 9.735771508942352e-06,
+      "loss": 0.2483,
+      "step": 1885
+    },
+    {
+      "epoch": 0.13095403416192195,
+      "grad_norm": 4.02630979495226,
+      "learning_rate": 9.73541067803292e-06,
+      "loss": 0.5709,
+      "step": 1886
+    },
+    {
+      "epoch": 0.13102346896264408,
+      "grad_norm": 4.101849453728972,
+      "learning_rate": 9.735049607611776e-06,
+      "loss": 0.3765,
+      "step": 1887
+    },
+    {
+      "epoch": 0.1310929037633662,
+      "grad_norm": 4.921690609383153,
+      "learning_rate": 9.73468829769718e-06,
+      "loss": 0.6169,
+      "step": 1888
+    },
+    {
+      "epoch": 0.13116233856408832,
+      "grad_norm": 4.87811784102188,
+      "learning_rate": 9.73432674830741e-06,
+      "loss": 0.7865,
+      "step": 1889
+    },
+    {
+      "epoch": 0.13123177336481046,
+      "grad_norm": 3.9286496306094993,
+      "learning_rate": 9.733964959460749e-06,
+      "loss": 0.5547,
+      "step": 1890
+    },
+    {
+      "epoch": 0.13130120816553256,
+      "grad_norm": 5.0123117686077645,
+      "learning_rate": 9.733602931175496e-06,
+      "loss": 0.7967,
+      "step": 1891
+    },
+    {
+      "epoch": 0.1313706429662547,
+      "grad_norm": 4.534716974227372,
+      "learning_rate": 9.733240663469965e-06,
+      "loss": 0.5185,
+      "step": 1892
+    },
+    {
+      "epoch": 0.1314400777669768,
+      "grad_norm": 4.380307948647215,
+      "learning_rate": 9.732878156362478e-06,
+      "loss": 0.6317,
+      "step": 1893
+    },
+    {
+      "epoch": 0.13150951256769894,
+      "grad_norm": 4.567734919423023,
+      "learning_rate": 9.732515409871366e-06,
+      "loss": 0.7239,
+      "step": 1894
+    },
+    {
+      "epoch": 0.13157894736842105,
+      "grad_norm": 3.7090731747763743,
+      "learning_rate": 9.732152424014981e-06,
+      "loss": 0.3383,
+      "step": 1895
+    },
+    {
+      "epoch": 0.13164838216914318,
+      "grad_norm": 5.134006435308687,
+      "learning_rate": 9.731789198811683e-06,
+      "loss": 0.6073,
+      "step": 1896
+    },
+    {
+      "epoch": 0.1317178169698653,
+      "grad_norm": 5.397993275656045,
+      "learning_rate": 9.73142573427984e-06,
+      "loss": 0.955,
+      "step": 1897
+    },
+    {
+      "epoch": 0.13178725177058742,
+      "grad_norm": 3.724827370248617,
+      "learning_rate": 9.73106203043784e-06,
+      "loss": 0.3382,
+      "step": 1898
+    },
+    {
+      "epoch": 0.13185668657130953,
+      "grad_norm": 4.364961604860191,
+      "learning_rate": 9.730698087304073e-06,
+      "loss": 0.8464,
+      "step": 1899
+    },
+    {
+      "epoch": 0.13192612137203166,
+      "grad_norm": 4.146868527795593,
+      "learning_rate": 9.730333904896952e-06,
+      "loss": 0.4927,
+      "step": 1900
+    },
+    {
+      "epoch": 0.1319955561727538,
+      "grad_norm": 3.391759332252334,
+      "learning_rate": 9.729969483234894e-06,
+      "loss": 0.2921,
+      "step": 1901
+    },
+    {
+      "epoch": 0.1320649909734759,
+      "grad_norm": 5.060121522366206,
+      "learning_rate": 9.72960482233633e-06,
+      "loss": 0.6825,
+      "step": 1902
+    },
+    {
+      "epoch": 0.13213442577419804,
+      "grad_norm": 4.4889983555212005,
+      "learning_rate": 9.729239922219708e-06,
+      "loss": 0.522,
+      "step": 1903
+    },
+    {
+      "epoch": 0.13220386057492015,
+      "grad_norm": 3.7224531263254885,
+      "learning_rate": 9.728874782903481e-06,
+      "loss": 0.5061,
+      "step": 1904
+    },
+    {
+      "epoch": 0.13227329537564228,
+      "grad_norm": 4.988375686835955,
+      "learning_rate": 9.728509404406121e-06,
+      "loss": 0.9253,
+      "step": 1905
+    },
+    {
+      "epoch": 0.1323427301763644,
+      "grad_norm": 4.068864570832338,
+      "learning_rate": 9.728143786746104e-06,
+      "loss": 0.5936,
+      "step": 1906
+    },
+    {
+      "epoch": 0.13241216497708652,
+      "grad_norm": 4.320068356195618,
+      "learning_rate": 9.727777929941924e-06,
+      "loss": 0.5795,
+      "step": 1907
+    },
+    {
+      "epoch": 0.13248159977780863,
+      "grad_norm": 4.221897186521596,
+      "learning_rate": 9.727411834012087e-06,
+      "loss": 0.4937,
+      "step": 1908
+    },
+    {
+      "epoch": 0.13255103457853076,
+      "grad_norm": 5.686093395727419,
+      "learning_rate": 9.72704549897511e-06,
+      "loss": 0.6904,
+      "step": 1909
+    },
+    {
+      "epoch": 0.13262046937925287,
+      "grad_norm": 5.013842655693153,
+      "learning_rate": 9.72667892484952e-06,
+      "loss": 0.5298,
+      "step": 1910
+    },
+    {
+      "epoch": 0.132689904179975,
+      "grad_norm": 5.3998698912562535,
+      "learning_rate": 9.726312111653857e-06,
+      "loss": 0.837,
+      "step": 1911
+    },
+    {
+      "epoch": 0.13275933898069714,
+      "grad_norm": 4.56287932313851,
+      "learning_rate": 9.725945059406677e-06,
+      "loss": 0.5392,
+      "step": 1912
+    },
+    {
+      "epoch": 0.13282877378141925,
+      "grad_norm": 3.9830599159927362,
+      "learning_rate": 9.725577768126544e-06,
+      "loss": 0.3005,
+      "step": 1913
+    },
+    {
+      "epoch": 0.13289820858214138,
+      "grad_norm": 4.828242305496214,
+      "learning_rate": 9.725210237832034e-06,
+      "loss": 0.6916,
+      "step": 1914
+    },
+    {
+      "epoch": 0.1329676433828635,
+      "grad_norm": 4.324068302429372,
+      "learning_rate": 9.724842468541737e-06,
+      "loss": 0.6229,
+      "step": 1915
+    },
+    {
+      "epoch": 0.13303707818358562,
+      "grad_norm": 4.202831357224146,
+      "learning_rate": 9.724474460274255e-06,
+      "loss": 0.4579,
+      "step": 1916
+    },
+    {
+      "epoch": 0.13310651298430773,
+      "grad_norm": 4.306499401557309,
+      "learning_rate": 9.7241062130482e-06,
+      "loss": 0.5066,
+      "step": 1917
+    },
+    {
+      "epoch": 0.13317594778502986,
+      "grad_norm": 38.021407239355845,
+      "learning_rate": 9.7237377268822e-06,
+      "loss": 0.3939,
+      "step": 1918
+    },
+    {
+      "epoch": 0.13324538258575197,
+      "grad_norm": 4.526755364466834,
+      "learning_rate": 9.723369001794891e-06,
+      "loss": 0.4766,
+      "step": 1919
+    },
+    {
+      "epoch": 0.1333148173864741,
+      "grad_norm": 2.6962370702294547,
+      "learning_rate": 9.723000037804922e-06,
+      "loss": 0.276,
+      "step": 1920
+    },
+    {
+      "epoch": 0.1333842521871962,
+      "grad_norm": 4.040355251708409,
+      "learning_rate": 9.722630834930955e-06,
+      "loss": 0.3421,
+      "step": 1921
+    },
+    {
+      "epoch": 0.13345368698791835,
+      "grad_norm": 4.900738668115288,
+      "learning_rate": 9.722261393191664e-06,
+      "loss": 0.8399,
+      "step": 1922
+    },
+    {
+      "epoch": 0.13352312178864045,
+      "grad_norm": 5.011107757373327,
+      "learning_rate": 9.721891712605737e-06,
+      "loss": 0.6696,
+      "step": 1923
+    },
+    {
+      "epoch": 0.1335925565893626,
+      "grad_norm": 4.401838554184152,
+      "learning_rate": 9.721521793191868e-06,
+      "loss": 0.4223,
+      "step": 1924
+    },
+    {
+      "epoch": 0.13366199139008472,
+      "grad_norm": 4.249245121301882,
+      "learning_rate": 9.721151634968773e-06,
+      "loss": 0.5972,
+      "step": 1925
+    },
+    {
+      "epoch": 0.13373142619080683,
+      "grad_norm": 4.82963563937356,
+      "learning_rate": 9.720781237955168e-06,
+      "loss": 0.8872,
+      "step": 1926
+    },
+    {
+      "epoch": 0.13380086099152896,
+      "grad_norm": 4.406076926274472,
+      "learning_rate": 9.72041060216979e-06,
+      "loss": 0.5614,
+      "step": 1927
+    },
+    {
+      "epoch": 0.13387029579225107,
+      "grad_norm": 3.7875716576048513,
+      "learning_rate": 9.720039727631387e-06,
+      "loss": 0.4767,
+      "step": 1928
+    },
+    {
+      "epoch": 0.1339397305929732,
+      "grad_norm": 4.927149061673529,
+      "learning_rate": 9.719668614358713e-06,
+      "loss": 0.7639,
+      "step": 1929
+    },
+    {
+      "epoch": 0.1340091653936953,
+      "grad_norm": 4.192310511719908,
+      "learning_rate": 9.719297262370545e-06,
+      "loss": 0.4583,
+      "step": 1930
+    },
+    {
+      "epoch": 0.13407860019441745,
+      "grad_norm": 3.3052351400874542,
+      "learning_rate": 9.718925671685658e-06,
+      "loss": 0.2644,
+      "step": 1931
+    },
+    {
+      "epoch": 0.13414803499513955,
+      "grad_norm": 5.242148073229841,
+      "learning_rate": 9.718553842322851e-06,
+      "loss": 0.5569,
+      "step": 1932
+    },
+    {
+      "epoch": 0.1342174697958617,
+      "grad_norm": 4.770046058495332,
+      "learning_rate": 9.718181774300931e-06,
+      "loss": 0.581,
+      "step": 1933
+    },
+    {
+      "epoch": 0.1342869045965838,
+      "grad_norm": 3.887634159569658,
+      "learning_rate": 9.717809467638716e-06,
+      "loss": 0.5577,
+      "step": 1934
+    },
+    {
+      "epoch": 0.13435633939730593,
+      "grad_norm": 3.7053507283037117,
+      "learning_rate": 9.717436922355035e-06,
+      "loss": 0.4089,
+      "step": 1935
+    },
+    {
+      "epoch": 0.13442577419802806,
+      "grad_norm": 3.9579773677299155,
+      "learning_rate": 9.717064138468735e-06,
+      "loss": 0.4999,
+      "step": 1936
+    },
+    {
+      "epoch": 0.13449520899875017,
+      "grad_norm": 3.9169762105431714,
+      "learning_rate": 9.716691115998667e-06,
+      "loss": 0.5036,
+      "step": 1937
+    },
+    {
+      "epoch": 0.1345646437994723,
+      "grad_norm": 3.8036819067690493,
+      "learning_rate": 9.7163178549637e-06,
+      "loss": 0.593,
+      "step": 1938
+    },
+    {
+      "epoch": 0.1346340786001944,
+      "grad_norm": 4.388330531145816,
+      "learning_rate": 9.715944355382714e-06,
+      "loss": 0.5321,
+      "step": 1939
+    },
+    {
+      "epoch": 0.13470351340091655,
+      "grad_norm": 2.699600558969193,
+      "learning_rate": 9.715570617274596e-06,
+      "loss": 0.2605,
+      "step": 1940
+    },
+    {
+      "epoch": 0.13477294820163865,
+      "grad_norm": 4.983772464641617,
+      "learning_rate": 9.715196640658256e-06,
+      "loss": 0.7274,
+      "step": 1941
+    },
+    {
+      "epoch": 0.1348423830023608,
+      "grad_norm": 4.712923103359737,
+      "learning_rate": 9.714822425552604e-06,
+      "loss": 0.6865,
+      "step": 1942
+    },
+    {
+      "epoch": 0.1349118178030829,
+      "grad_norm": 4.527389000413363,
+      "learning_rate": 9.714447971976568e-06,
+      "loss": 0.6267,
+      "step": 1943
+    },
+    {
+      "epoch": 0.13498125260380503,
+      "grad_norm": 3.3735229378086427,
+      "learning_rate": 9.714073279949089e-06,
+      "loss": 0.3693,
+      "step": 1944
+    },
+    {
+      "epoch": 0.13505068740452714,
+      "grad_norm": 4.547484488255602,
+      "learning_rate": 9.713698349489117e-06,
+      "loss": 0.2491,
+      "step": 1945
+    },
+    {
+      "epoch": 0.13512012220524927,
+      "grad_norm": 4.689896667038806,
+      "learning_rate": 9.713323180615617e-06,
+      "loss": 0.6146,
+      "step": 1946
+    },
+    {
+      "epoch": 0.1351895570059714,
+      "grad_norm": 3.801806427659175,
+      "learning_rate": 9.712947773347564e-06,
+      "loss": 0.3315,
+      "step": 1947
+    },
+    {
+      "epoch": 0.1352589918066935,
+      "grad_norm": 3.442317975792722,
+      "learning_rate": 9.712572127703945e-06,
+      "loss": 0.3588,
+      "step": 1948
+    },
+    {
+      "epoch": 0.13532842660741565,
+      "grad_norm": 4.232493005297604,
+      "learning_rate": 9.712196243703762e-06,
+      "loss": 0.6084,
+      "step": 1949
+    },
+    {
+      "epoch": 0.13539786140813775,
+      "grad_norm": 4.422228233628285,
+      "learning_rate": 9.711820121366025e-06,
+      "loss": 0.4592,
+      "step": 1950
+    },
+    {
+      "epoch": 0.1354672962088599,
+      "grad_norm": 4.100366819007629,
+      "learning_rate": 9.711443760709757e-06,
+      "loss": 0.4624,
+      "step": 1951
+    },
+    {
+      "epoch": 0.135536731009582,
+      "grad_norm": 3.854244024748049,
+      "learning_rate": 9.711067161753995e-06,
+      "loss": 0.4878,
+      "step": 1952
+    },
+    {
+      "epoch": 0.13560616581030413,
+      "grad_norm": 3.7245586980659318,
+      "learning_rate": 9.710690324517788e-06,
+      "loss": 0.4548,
+      "step": 1953
+    },
+    {
+      "epoch": 0.13567560061102624,
+      "grad_norm": 4.781606643938577,
+      "learning_rate": 9.710313249020193e-06,
+      "loss": 0.5182,
+      "step": 1954
+    },
+    {
+      "epoch": 0.13574503541174837,
+      "grad_norm": 4.755514193717609,
+      "learning_rate": 9.709935935280286e-06,
+      "loss": 0.6849,
+      "step": 1955
+    },
+    {
+      "epoch": 0.13581447021247048,
+      "grad_norm": 4.548151837119778,
+      "learning_rate": 9.709558383317148e-06,
+      "loss": 0.7407,
+      "step": 1956
+    },
+    {
+      "epoch": 0.1358839050131926,
+      "grad_norm": 3.7754027807517723,
+      "learning_rate": 9.709180593149877e-06,
+      "loss": 0.3542,
+      "step": 1957
+    },
+    {
+      "epoch": 0.13595333981391475,
+      "grad_norm": 10.771841496731694,
+      "learning_rate": 9.70880256479758e-06,
+      "loss": 0.5798,
+      "step": 1958
+    },
+    {
+      "epoch": 0.13602277461463685,
+      "grad_norm": 4.277629931100356,
+      "learning_rate": 9.708424298279377e-06,
+      "loss": 0.4424,
+      "step": 1959
+    },
+    {
+      "epoch": 0.136092209415359,
+      "grad_norm": 3.1059775129904885,
+      "learning_rate": 9.708045793614402e-06,
+      "loss": 0.3527,
+      "step": 1960
+    },
+    {
+      "epoch": 0.1361616442160811,
+      "grad_norm": 3.959252139166287,
+      "learning_rate": 9.707667050821796e-06,
+      "loss": 0.4787,
+      "step": 1961
+    },
+    {
+      "epoch": 0.13623107901680323,
+      "grad_norm": 4.627029172037978,
+      "learning_rate": 9.70728806992072e-06,
+      "loss": 0.6027,
+      "step": 1962
+    },
+    {
+      "epoch": 0.13630051381752534,
+      "grad_norm": 6.462599389150344,
+      "learning_rate": 9.70690885093034e-06,
+      "loss": 0.8626,
+      "step": 1963
+    },
+    {
+      "epoch": 0.13636994861824747,
+      "grad_norm": 5.128555094824931,
+      "learning_rate": 9.706529393869835e-06,
+      "loss": 0.8223,
+      "step": 1964
+    },
+    {
+      "epoch": 0.13643938341896958,
+      "grad_norm": 4.522585983654513,
+      "learning_rate": 9.706149698758401e-06,
+      "loss": 0.6136,
+      "step": 1965
+    },
+    {
+      "epoch": 0.1365088182196917,
+      "grad_norm": 3.582141328723532,
+      "learning_rate": 9.705769765615239e-06,
+      "loss": 0.3099,
+      "step": 1966
+    },
+    {
+      "epoch": 0.13657825302041382,
+      "grad_norm": 4.216729129685716,
+      "learning_rate": 9.705389594459566e-06,
+      "loss": 0.4292,
+      "step": 1967
+    },
+    {
+      "epoch": 0.13664768782113595,
+      "grad_norm": 4.2008671713960055,
+      "learning_rate": 9.705009185310615e-06,
+      "loss": 0.4613,
+      "step": 1968
+    },
+    {
+      "epoch": 0.13671712262185806,
+      "grad_norm": 4.3301690302084195,
+      "learning_rate": 9.704628538187621e-06,
+      "loss": 0.6703,
+      "step": 1969
+    },
+    {
+      "epoch": 0.1367865574225802,
+      "grad_norm": 4.090942632933998,
+      "learning_rate": 9.704247653109843e-06,
+      "loss": 0.421,
+      "step": 1970
+    },
+    {
+      "epoch": 0.13685599222330233,
+      "grad_norm": 3.715539963891412,
+      "learning_rate": 9.703866530096538e-06,
+      "loss": 0.3299,
+      "step": 1971
+    },
+    {
+      "epoch": 0.13692542702402444,
+      "grad_norm": 4.646310454673358,
+      "learning_rate": 9.703485169166988e-06,
+      "loss": 0.6391,
+      "step": 1972
+    },
+    {
+      "epoch": 0.13699486182474657,
+      "grad_norm": 12.191677580913243,
+      "learning_rate": 9.70310357034048e-06,
+      "loss": 0.4313,
+      "step": 1973
+    },
+    {
+      "epoch": 0.13706429662546868,
+      "grad_norm": 4.414715314025958,
+      "learning_rate": 9.702721733636315e-06,
+      "loss": 0.7022,
+      "step": 1974
+    },
+    {
+      "epoch": 0.1371337314261908,
+      "grad_norm": 4.552738634733077,
+      "learning_rate": 9.702339659073807e-06,
+      "loss": 0.8359,
+      "step": 1975
+    },
+    {
+      "epoch": 0.13720316622691292,
+      "grad_norm": 6.309611887311083,
+      "learning_rate": 9.70195734667228e-06,
+      "loss": 0.7154,
+      "step": 1976
+    },
+    {
+      "epoch": 0.13727260102763506,
+      "grad_norm": 5.880815314177445,
+      "learning_rate": 9.701574796451071e-06,
+      "loss": 0.598,
+      "step": 1977
+    },
+    {
+      "epoch": 0.13734203582835716,
+      "grad_norm": 3.780988900300168,
+      "learning_rate": 9.701192008429529e-06,
+      "loss": 0.4191,
+      "step": 1978
+    },
+    {
+      "epoch": 0.1374114706290793,
+      "grad_norm": 3.9939162496094687,
+      "learning_rate": 9.700808982627015e-06,
+      "loss": 0.5581,
+      "step": 1979
+    },
+    {
+      "epoch": 0.1374809054298014,
+      "grad_norm": 3.905331085573648,
+      "learning_rate": 9.700425719062902e-06,
+      "loss": 0.4617,
+      "step": 1980
+    },
+    {
+      "epoch": 0.13755034023052354,
+      "grad_norm": 4.13987907064628,
+      "learning_rate": 9.700042217756575e-06,
+      "loss": 0.4853,
+      "step": 1981
+    },
+    {
+      "epoch": 0.13761977503124567,
+      "grad_norm": 4.359510947687961,
+      "learning_rate": 9.699658478727431e-06,
+      "loss": 0.5342,
+      "step": 1982
+    },
+    {
+      "epoch": 0.13768920983196778,
+      "grad_norm": 4.740046279301326,
+      "learning_rate": 9.69927450199488e-06,
+      "loss": 0.4983,
+      "step": 1983
+    },
+    {
+      "epoch": 0.13775864463268991,
+      "grad_norm": 4.811088123538485,
+      "learning_rate": 9.69889028757834e-06,
+      "loss": 0.5284,
+      "step": 1984
+    },
+    {
+      "epoch": 0.13782807943341202,
+      "grad_norm": 4.431822772844443,
+      "learning_rate": 9.698505835497249e-06,
+      "loss": 0.608,
+      "step": 1985
+    },
+    {
+      "epoch": 0.13789751423413416,
+      "grad_norm": 4.542012058952836,
+      "learning_rate": 9.698121145771048e-06,
+      "loss": 0.5729,
+      "step": 1986
+    },
+    {
+      "epoch": 0.13796694903485626,
+      "grad_norm": 4.201730581065927,
+      "learning_rate": 9.697736218419196e-06,
+      "loss": 0.5455,
+      "step": 1987
+    },
+    {
+      "epoch": 0.1380363838355784,
+      "grad_norm": 4.452249724962726,
+      "learning_rate": 9.697351053461163e-06,
+      "loss": 0.5571,
+      "step": 1988
+    },
+    {
+      "epoch": 0.1381058186363005,
+      "grad_norm": 3.910507386314688,
+      "learning_rate": 9.696965650916427e-06,
+      "loss": 0.3761,
+      "step": 1989
+    },
+    {
+      "epoch": 0.13817525343702264,
+      "grad_norm": 5.152119323324873,
+      "learning_rate": 9.696580010804486e-06,
+      "loss": 0.6357,
+      "step": 1990
+    },
+    {
+      "epoch": 0.13824468823774475,
+      "grad_norm": 4.530585079973892,
+      "learning_rate": 9.69619413314484e-06,
+      "loss": 0.5589,
+      "step": 1991
+    },
+    {
+      "epoch": 0.13831412303846688,
+      "grad_norm": 4.95239284919159,
+      "learning_rate": 9.695808017957012e-06,
+      "loss": 0.5127,
+      "step": 1992
+    },
+    {
+      "epoch": 0.13838355783918901,
+      "grad_norm": 5.747623942083142,
+      "learning_rate": 9.695421665260525e-06,
+      "loss": 0.7764,
+      "step": 1993
+    },
+    {
+      "epoch": 0.13845299263991112,
+      "grad_norm": 4.825390983709328,
+      "learning_rate": 9.695035075074924e-06,
+      "loss": 0.6451,
+      "step": 1994
+    },
+    {
+      "epoch": 0.13852242744063326,
+      "grad_norm": 4.187502716995179,
+      "learning_rate": 9.694648247419762e-06,
+      "loss": 0.381,
+      "step": 1995
+    },
+    {
+      "epoch": 0.13859186224135536,
+      "grad_norm": 4.222532622492172,
+      "learning_rate": 9.694261182314605e-06,
+      "loss": 0.6246,
+      "step": 1996
+    },
+    {
+      "epoch": 0.1386612970420775,
+      "grad_norm": 3.8711348324921375,
+      "learning_rate": 9.69387387977903e-06,
+      "loss": 0.455,
+      "step": 1997
+    },
+    {
+      "epoch": 0.1387307318427996,
+      "grad_norm": 5.504042372192557,
+      "learning_rate": 9.693486339832623e-06,
+      "loss": 0.6149,
+      "step": 1998
+    },
+    {
+      "epoch": 0.13880016664352174,
+      "grad_norm": 4.073917769821312,
+      "learning_rate": 9.693098562494988e-06,
+      "loss": 0.3421,
+      "step": 1999
+    },
+    {
+      "epoch": 0.13886960144424385,
+      "grad_norm": 3.9402395095896137,
+      "learning_rate": 9.69271054778574e-06,
+      "loss": 0.6305,
+      "step": 2000
+    },
+    {
+      "epoch": 0.13893903624496598,
+      "grad_norm": 4.6535730345142055,
+      "learning_rate": 9.692322295724503e-06,
+      "loss": 0.4873,
+      "step": 2001
+    },
+    {
+      "epoch": 0.1390084710456881,
+      "grad_norm": 4.155673564835397,
+      "learning_rate": 9.691933806330913e-06,
+      "loss": 0.6034,
+      "step": 2002
+    },
+    {
+      "epoch": 0.13907790584641022,
+      "grad_norm": 3.4917887981696367,
+      "learning_rate": 9.69154507962462e-06,
+      "loss": 0.3722,
+      "step": 2003
+    },
+    {
+      "epoch": 0.13914734064713236,
+      "grad_norm": 3.1936535776092247,
+      "learning_rate": 9.691156115625287e-06,
+      "loss": 0.4081,
+      "step": 2004
+    },
+    {
+      "epoch": 0.13921677544785446,
+      "grad_norm": 5.370418533866856,
+      "learning_rate": 9.690766914352586e-06,
+      "loss": 0.625,
+      "step": 2005
+    },
+    {
+      "epoch": 0.1392862102485766,
+      "grad_norm": 3.72225268254193,
+      "learning_rate": 9.690377475826201e-06,
+      "loss": 0.3724,
+      "step": 2006
+    },
+    {
+      "epoch": 0.1393556450492987,
+      "grad_norm": 4.275777785811119,
+      "learning_rate": 9.68998780006583e-06,
+      "loss": 0.4789,
+      "step": 2007
+    },
+    {
+      "epoch": 0.13942507985002084,
+      "grad_norm": 5.11604047161193,
+      "learning_rate": 9.689597887091186e-06,
+      "loss": 0.3912,
+      "step": 2008
+    },
+    {
+      "epoch": 0.13949451465074295,
+      "grad_norm": 3.493829860584996,
+      "learning_rate": 9.689207736921986e-06,
+      "loss": 0.4383,
+      "step": 2009
+    },
+    {
+      "epoch": 0.13956394945146508,
+      "grad_norm": 3.814691586632843,
+      "learning_rate": 9.688817349577966e-06,
+      "loss": 0.5006,
+      "step": 2010
+    },
+    {
+      "epoch": 0.1396333842521872,
+      "grad_norm": 5.296556849163979,
+      "learning_rate": 9.688426725078868e-06,
+      "loss": 0.956,
+      "step": 2011
+    },
+    {
+      "epoch": 0.13970281905290932,
+      "grad_norm": 3.9901503441249515,
+      "learning_rate": 9.688035863444453e-06,
+      "loss": 0.6011,
+      "step": 2012
+    },
+    {
+      "epoch": 0.13977225385363143,
+      "grad_norm": 3.865976012220175,
+      "learning_rate": 9.687644764694488e-06,
+      "loss": 0.4383,
+      "step": 2013
+    },
+    {
+      "epoch": 0.13984168865435356,
+      "grad_norm": 4.998681857518684,
+      "learning_rate": 9.687253428848757e-06,
+      "loss": 0.4956,
+      "step": 2014
+    },
+    {
+      "epoch": 0.1399111234550757,
+      "grad_norm": 3.0371246949613733,
+      "learning_rate": 9.68686185592705e-06,
+      "loss": 0.2869,
+      "step": 2015
+    },
+    {
+      "epoch": 0.1399805582557978,
+      "grad_norm": 4.71332296829214,
+      "learning_rate": 9.686470045949175e-06,
+      "loss": 0.4438,
+      "step": 2016
+    },
+    {
+      "epoch": 0.14004999305651994,
+      "grad_norm": 4.703425836493707,
+      "learning_rate": 9.686077998934948e-06,
+      "loss": 0.5444,
+      "step": 2017
+    },
+    {
+      "epoch": 0.14011942785724205,
+      "grad_norm": 4.785108338916558,
+      "learning_rate": 9.685685714904198e-06,
+      "loss": 0.5929,
+      "step": 2018
+    },
+    {
+      "epoch": 0.14018886265796418,
+      "grad_norm": 4.844972255457499,
+      "learning_rate": 9.685293193876766e-06,
+      "loss": 0.6697,
+      "step": 2019
+    },
+    {
+      "epoch": 0.1402582974586863,
+      "grad_norm": 4.3003844144559045,
+      "learning_rate": 9.684900435872507e-06,
+      "loss": 0.5971,
+      "step": 2020
+    },
+    {
+      "epoch": 0.14032773225940842,
+      "grad_norm": 3.5184866007677904,
+      "learning_rate": 9.684507440911286e-06,
+      "loss": 0.4103,
+      "step": 2021
+    },
+    {
+      "epoch": 0.14039716706013053,
+      "grad_norm": 4.439364471675243,
+      "learning_rate": 9.684114209012979e-06,
+      "loss": 0.4947,
+      "step": 2022
+    },
+    {
+      "epoch": 0.14046660186085266,
+      "grad_norm": 5.981733828806437,
+      "learning_rate": 9.683720740197474e-06,
+      "loss": 0.6097,
+      "step": 2023
+    },
+    {
+      "epoch": 0.14053603666157477,
+      "grad_norm": 4.799351121715541,
+      "learning_rate": 9.683327034484674e-06,
+      "loss": 0.4424,
+      "step": 2024
+    },
+    {
+      "epoch": 0.1406054714622969,
+      "grad_norm": 4.494609203450447,
+      "learning_rate": 9.682933091894495e-06,
+      "loss": 0.5255,
+      "step": 2025
+    },
+    {
+      "epoch": 0.140674906263019,
+      "grad_norm": 3.162269364031781,
+      "learning_rate": 9.682538912446857e-06,
+      "loss": 0.1532,
+      "step": 2026
+    },
+    {
+      "epoch": 0.14074434106374115,
+      "grad_norm": 3.3210971369734326,
+      "learning_rate": 9.6821444961617e-06,
+      "loss": 0.4505,
+      "step": 2027
+    },
+    {
+      "epoch": 0.14081377586446328,
+      "grad_norm": 3.690271639091779,
+      "learning_rate": 9.681749843058973e-06,
+      "loss": 0.3989,
+      "step": 2028
+    },
+    {
+      "epoch": 0.1408832106651854,
+      "grad_norm": 4.545355226660664,
+      "learning_rate": 9.681354953158636e-06,
+      "loss": 0.6355,
+      "step": 2029
+    },
+    {
+      "epoch": 0.14095264546590752,
+      "grad_norm": 4.253528297667697,
+      "learning_rate": 9.680959826480664e-06,
+      "loss": 0.7317,
+      "step": 2030
+    },
+    {
+      "epoch": 0.14102208026662963,
+      "grad_norm": 4.041190115096981,
+      "learning_rate": 9.680564463045041e-06,
+      "loss": 0.2772,
+      "step": 2031
+    },
+    {
+      "epoch": 0.14109151506735176,
+      "grad_norm": 3.6680877541888206,
+      "learning_rate": 9.680168862871762e-06,
+      "loss": 0.3295,
+      "step": 2032
+    },
+    {
+      "epoch": 0.14116094986807387,
+      "grad_norm": 4.086416824923184,
+      "learning_rate": 9.67977302598084e-06,
+      "loss": 0.6584,
+      "step": 2033
+    },
+    {
+      "epoch": 0.141230384668796,
+      "grad_norm": 4.183636055332061,
+      "learning_rate": 9.679376952392293e-06,
+      "loss": 0.4658,
+      "step": 2034
+    },
+    {
+      "epoch": 0.1412998194695181,
+      "grad_norm": 4.68064726128267,
+      "learning_rate": 9.678980642126155e-06,
+      "loss": 0.4912,
+      "step": 2035
+    },
+    {
+      "epoch": 0.14136925427024025,
+      "grad_norm": 4.362028975453703,
+      "learning_rate": 9.678584095202468e-06,
+      "loss": 0.4873,
+      "step": 2036
+    },
+    {
+      "epoch": 0.14143868907096235,
+      "grad_norm": 3.3070382344840104,
+      "learning_rate": 9.678187311641295e-06,
+      "loss": 0.2924,
+      "step": 2037
+    },
+    {
+      "epoch": 0.1415081238716845,
+      "grad_norm": 3.5990291695230945,
+      "learning_rate": 9.6777902914627e-06,
+      "loss": 0.5233,
+      "step": 2038
+    },
+    {
+      "epoch": 0.14157755867240662,
+      "grad_norm": 3.300326116228172,
+      "learning_rate": 9.677393034686766e-06,
+      "loss": 0.3191,
+      "step": 2039
+    },
+    {
+      "epoch": 0.14164699347312873,
+      "grad_norm": 7.058213958288835,
+      "learning_rate": 9.676995541333585e-06,
+      "loss": 0.7709,
+      "step": 2040
+    },
+    {
+      "epoch": 0.14171642827385086,
+      "grad_norm": 3.7259821793919197,
+      "learning_rate": 9.676597811423264e-06,
+      "loss": 0.4154,
+      "step": 2041
+    },
+    {
+      "epoch": 0.14178586307457297,
+      "grad_norm": 4.556570792476836,
+      "learning_rate": 9.676199844975914e-06,
+      "loss": 0.482,
+      "step": 2042
+    },
+    {
+      "epoch": 0.1418552978752951,
+      "grad_norm": 2.9546063447011495,
+      "learning_rate": 9.675801642011669e-06,
+      "loss": 0.2433,
+      "step": 2043
+    },
+    {
+      "epoch": 0.1419247326760172,
+      "grad_norm": 4.142795416249976,
+      "learning_rate": 9.675403202550667e-06,
+      "loss": 0.384,
+      "step": 2044
+    },
+    {
+      "epoch": 0.14199416747673935,
+      "grad_norm": 4.235361392547599,
+      "learning_rate": 9.675004526613062e-06,
+      "loss": 0.5672,
+      "step": 2045
+    },
+    {
+      "epoch": 0.14206360227746145,
+      "grad_norm": 4.262726126556873,
+      "learning_rate": 9.674605614219019e-06,
+      "loss": 0.6718,
+      "step": 2046
+    },
+    {
+      "epoch": 0.1421330370781836,
+      "grad_norm": 7.903242075261548,
+      "learning_rate": 9.674206465388713e-06,
+      "loss": 0.9077,
+      "step": 2047
+    },
+    {
+      "epoch": 0.1422024718789057,
+      "grad_norm": 4.2459558452820785,
+      "learning_rate": 9.673807080142334e-06,
+      "loss": 0.4331,
+      "step": 2048
+    },
+    {
+      "epoch": 0.14227190667962783,
+      "grad_norm": 4.239551182077637,
+      "learning_rate": 9.67340745850008e-06,
+      "loss": 0.579,
+      "step": 2049
+    },
+    {
+      "epoch": 0.14234134148034996,
+      "grad_norm": 8.038026349729106,
+      "learning_rate": 9.673007600482168e-06,
+      "loss": 0.5906,
+      "step": 2050
+    },
+    {
+      "epoch": 0.14241077628107207,
+      "grad_norm": 4.002059659042826,
+      "learning_rate": 9.672607506108817e-06,
+      "loss": 0.5827,
+      "step": 2051
+    },
+    {
+      "epoch": 0.1424802110817942,
+      "grad_norm": 3.6532412751703083,
+      "learning_rate": 9.672207175400265e-06,
+      "loss": 0.3607,
+      "step": 2052
+    },
+    {
+      "epoch": 0.1425496458825163,
+      "grad_norm": 5.39727124970411,
+      "learning_rate": 9.671806608376763e-06,
+      "loss": 0.8619,
+      "step": 2053
+    },
+    {
+      "epoch": 0.14261908068323845,
+      "grad_norm": 3.7304817664753545,
+      "learning_rate": 9.671405805058569e-06,
+      "loss": 0.4787,
+      "step": 2054
+    },
+    {
+      "epoch": 0.14268851548396055,
+      "grad_norm": 5.571774798894162,
+      "learning_rate": 9.671004765465955e-06,
+      "loss": 0.518,
+      "step": 2055
+    },
+    {
+      "epoch": 0.1427579502846827,
+      "grad_norm": 5.013494184550903,
+      "learning_rate": 9.670603489619206e-06,
+      "loss": 0.6222,
+      "step": 2056
+    },
+    {
+      "epoch": 0.1428273850854048,
+      "grad_norm": 3.7019157226706985,
+      "learning_rate": 9.670201977538616e-06,
+      "loss": 0.4918,
+      "step": 2057
+    },
+    {
+      "epoch": 0.14289681988612693,
+      "grad_norm": 4.621294263187852,
+      "learning_rate": 9.669800229244497e-06,
+      "loss": 0.6814,
+      "step": 2058
+    },
+    {
+      "epoch": 0.14296625468684904,
+      "grad_norm": 3.2022282426342716,
+      "learning_rate": 9.669398244757164e-06,
+      "loss": 0.3462,
+      "step": 2059
+    },
+    {
+      "epoch": 0.14303568948757117,
+      "grad_norm": 4.504873090753799,
+      "learning_rate": 9.668996024096955e-06,
+      "loss": 0.4152,
+      "step": 2060
+    },
+    {
+      "epoch": 0.1431051242882933,
+      "grad_norm": 4.714806660737154,
+      "learning_rate": 9.668593567284208e-06,
+      "loss": 0.804,
+      "step": 2061
+    },
+    {
+      "epoch": 0.1431745590890154,
+      "grad_norm": 3.9891302249264045,
+      "learning_rate": 9.668190874339281e-06,
+      "loss": 0.4326,
+      "step": 2062
+    },
+    {
+      "epoch": 0.14324399388973755,
+      "grad_norm": 3.683347926804802,
+      "learning_rate": 9.667787945282544e-06,
+      "loss": 0.4933,
+      "step": 2063
+    },
+    {
+      "epoch": 0.14331342869045965,
+      "grad_norm": 4.1253775336132374,
+      "learning_rate": 9.667384780134373e-06,
+      "loss": 0.4991,
+      "step": 2064
+    },
+    {
+      "epoch": 0.1433828634911818,
+      "grad_norm": 4.45048989616051,
+      "learning_rate": 9.666981378915161e-06,
+      "loss": 0.5493,
+      "step": 2065
+    },
+    {
+      "epoch": 0.1434522982919039,
+      "grad_norm": 3.0058904010812255,
+      "learning_rate": 9.666577741645313e-06,
+      "loss": 0.3457,
+      "step": 2066
+    },
+    {
+      "epoch": 0.14352173309262603,
+      "grad_norm": 4.49495986414482,
+      "learning_rate": 9.666173868345243e-06,
+      "loss": 0.6551,
+      "step": 2067
+    },
+    {
+      "epoch": 0.14359116789334814,
+      "grad_norm": 3.7112993964502934,
+      "learning_rate": 9.665769759035379e-06,
+      "loss": 0.2404,
+      "step": 2068
+    },
+    {
+      "epoch": 0.14366060269407027,
+      "grad_norm": 6.237392857371296,
+      "learning_rate": 9.66536541373616e-06,
+      "loss": 0.6418,
+      "step": 2069
+    },
+    {
+      "epoch": 0.14373003749479238,
+      "grad_norm": 4.61966544142573,
+      "learning_rate": 9.664960832468038e-06,
+      "loss": 0.5599,
+      "step": 2070
+    },
+    {
+      "epoch": 0.1437994722955145,
+      "grad_norm": 3.9083094106465857,
+      "learning_rate": 9.664556015251476e-06,
+      "loss": 0.2786,
+      "step": 2071
+    },
+    {
+      "epoch": 0.14386890709623662,
+      "grad_norm": 5.103914771144059,
+      "learning_rate": 9.664150962106948e-06,
+      "loss": 0.3883,
+      "step": 2072
+    },
+    {
+      "epoch": 0.14393834189695875,
+      "grad_norm": 4.899108425182691,
+      "learning_rate": 9.663745673054943e-06,
+      "loss": 0.672,
+      "step": 2073
+    },
+    {
+      "epoch": 0.1440077766976809,
+      "grad_norm": 4.55170172296448,
+      "learning_rate": 9.663340148115961e-06,
+      "loss": 0.6498,
+      "step": 2074
+    },
+    {
+      "epoch": 0.144077211498403,
+      "grad_norm": 3.8242001460057105,
+      "learning_rate": 9.66293438731051e-06,
+      "loss": 0.2487,
+      "step": 2075
+    },
+    {
+      "epoch": 0.14414664629912513,
+      "grad_norm": 3.9355486909118977,
+      "learning_rate": 9.662528390659111e-06,
+      "loss": 0.415,
+      "step": 2076
+    },
+    {
+      "epoch": 0.14421608109984724,
+      "grad_norm": 4.624334921565409,
+      "learning_rate": 9.662122158182306e-06,
+      "loss": 0.5791,
+      "step": 2077
+    },
+    {
+      "epoch": 0.14428551590056937,
+      "grad_norm": 4.222531085801669,
+      "learning_rate": 9.661715689900636e-06,
+      "loss": 0.5788,
+      "step": 2078
+    },
+    {
+      "epoch": 0.14435495070129148,
+      "grad_norm": 4.9143621463126035,
+      "learning_rate": 9.661308985834661e-06,
+      "loss": 0.7304,
+      "step": 2079
+    },
+    {
+      "epoch": 0.1444243855020136,
+      "grad_norm": 5.022547185570577,
+      "learning_rate": 9.660902046004954e-06,
+      "loss": 0.5861,
+      "step": 2080
+    },
+    {
+      "epoch": 0.14449382030273572,
+      "grad_norm": 4.659990300546865,
+      "learning_rate": 9.660494870432094e-06,
+      "loss": 0.6541,
+      "step": 2081
+    },
+    {
+      "epoch": 0.14456325510345786,
+      "grad_norm": 4.30139623958207,
+      "learning_rate": 9.660087459136678e-06,
+      "loss": 0.4429,
+      "step": 2082
+    },
+    {
+      "epoch": 0.14463268990417996,
+      "grad_norm": 4.016953235363166,
+      "learning_rate": 9.65967981213931e-06,
+      "loss": 0.4234,
+      "step": 2083
+    },
+    {
+      "epoch": 0.1447021247049021,
+      "grad_norm": 3.3875678485971736,
+      "learning_rate": 9.659271929460612e-06,
+      "loss": 0.4571,
+      "step": 2084
+    },
+    {
+      "epoch": 0.14477155950562423,
+      "grad_norm": 3.0449316946739238,
+      "learning_rate": 9.658863811121212e-06,
+      "loss": 0.2103,
+      "step": 2085
+    },
+    {
+      "epoch": 0.14484099430634634,
+      "grad_norm": 3.8184834853338927,
+      "learning_rate": 9.658455457141751e-06,
+      "loss": 0.3131,
+      "step": 2086
+    },
+    {
+      "epoch": 0.14491042910706847,
+      "grad_norm": 4.873143529200353,
+      "learning_rate": 9.658046867542885e-06,
+      "loss": 0.5708,
+      "step": 2087
+    },
+    {
+      "epoch": 0.14497986390779058,
+      "grad_norm": 3.558694303086474,
+      "learning_rate": 9.65763804234528e-06,
+      "loss": 0.303,
+      "step": 2088
+    },
+    {
+      "epoch": 0.14504929870851271,
+      "grad_norm": 6.1269836643224105,
+      "learning_rate": 9.657228981569612e-06,
+      "loss": 0.8868,
+      "step": 2089
+    },
+    {
+      "epoch": 0.14511873350923482,
+      "grad_norm": 4.571678745494512,
+      "learning_rate": 9.656819685236574e-06,
+      "loss": 0.5841,
+      "step": 2090
+    },
+    {
+      "epoch": 0.14518816830995696,
+      "grad_norm": 4.1197659416405905,
+      "learning_rate": 9.656410153366864e-06,
+      "loss": 0.4262,
+      "step": 2091
+    },
+    {
+      "epoch": 0.14525760311067906,
+      "grad_norm": 3.928091041737535,
+      "learning_rate": 9.656000385981198e-06,
+      "loss": 0.335,
+      "step": 2092
+    },
+    {
+      "epoch": 0.1453270379114012,
+      "grad_norm": 5.2877342602906285,
+      "learning_rate": 9.655590383100302e-06,
+      "loss": 0.7814,
+      "step": 2093
+    },
+    {
+      "epoch": 0.1453964727121233,
+      "grad_norm": 3.1919939465047484,
+      "learning_rate": 9.655180144744912e-06,
+      "loss": 0.2093,
+      "step": 2094
+    },
+    {
+      "epoch": 0.14546590751284544,
+      "grad_norm": 4.982244355002265,
+      "learning_rate": 9.654769670935778e-06,
+      "loss": 0.7627,
+      "step": 2095
+    },
+    {
+      "epoch": 0.14553534231356757,
+      "grad_norm": 4.441459683627491,
+      "learning_rate": 9.654358961693661e-06,
+      "loss": 0.547,
+      "step": 2096
+    },
+    {
+      "epoch": 0.14560477711428968,
+      "grad_norm": 4.507316410103269,
+      "learning_rate": 9.653948017039335e-06,
+      "loss": 0.5081,
+      "step": 2097
+    },
+    {
+      "epoch": 0.14567421191501181,
+      "grad_norm": 4.559209379079217,
+      "learning_rate": 9.653536836993584e-06,
+      "loss": 0.4905,
+      "step": 2098
+    },
+    {
+      "epoch": 0.14574364671573392,
+      "grad_norm": 3.319787945490247,
+      "learning_rate": 9.653125421577204e-06,
+      "loss": 0.4427,
+      "step": 2099
+    },
+    {
+      "epoch": 0.14581308151645606,
+      "grad_norm": 3.640337076457778,
+      "learning_rate": 9.65271377081101e-06,
+      "loss": 0.2775,
+      "step": 2100
+    },
+    {
+      "epoch": 0.14588251631717816,
+      "grad_norm": 4.125116930941902,
+      "learning_rate": 9.652301884715815e-06,
+      "loss": 0.4386,
+      "step": 2101
+    },
+    {
+      "epoch": 0.1459519511179003,
+      "grad_norm": 4.771023132367938,
+      "learning_rate": 9.651889763312456e-06,
+      "loss": 0.5129,
+      "step": 2102
+    },
+    {
+      "epoch": 0.1460213859186224,
+      "grad_norm": 2.963634209392736,
+      "learning_rate": 9.651477406621776e-06,
+      "loss": 0.4573,
+      "step": 2103
+    },
+    {
+      "epoch": 0.14609082071934454,
+      "grad_norm": 4.768417154334674,
+      "learning_rate": 9.651064814664633e-06,
+      "loss": 0.7566,
+      "step": 2104
+    },
+    {
+      "epoch": 0.14616025552006665,
+      "grad_norm": 2.9122861978174406,
+      "learning_rate": 9.650651987461896e-06,
+      "loss": 0.1471,
+      "step": 2105
+    },
+    {
+      "epoch": 0.14622969032078878,
+      "grad_norm": 5.859515755877611,
+      "learning_rate": 9.650238925034441e-06,
+      "loss": 0.5179,
+      "step": 2106
+    },
+    {
+      "epoch": 0.14629912512151091,
+      "grad_norm": 3.964267190676561,
+      "learning_rate": 9.649825627403166e-06,
+      "loss": 0.4821,
+      "step": 2107
+    },
+    {
+      "epoch": 0.14636855992223302,
+      "grad_norm": 4.54913469478474,
+      "learning_rate": 9.649412094588971e-06,
+      "loss": 0.4727,
+      "step": 2108
+    },
+    {
+      "epoch": 0.14643799472295516,
+      "grad_norm": 4.860896334877277,
+      "learning_rate": 9.648998326612774e-06,
+      "loss": 0.6699,
+      "step": 2109
+    },
+    {
+      "epoch": 0.14650742952367726,
+      "grad_norm": 3.5387954041003713,
+      "learning_rate": 9.648584323495502e-06,
+      "loss": 0.3415,
+      "step": 2110
+    },
+    {
+      "epoch": 0.1465768643243994,
+      "grad_norm": 2.9956984849515123,
+      "learning_rate": 9.648170085258094e-06,
+      "loss": 0.2599,
+      "step": 2111
+    },
+    {
+      "epoch": 0.1466462991251215,
+      "grad_norm": 5.281627270863168,
+      "learning_rate": 9.647755611921505e-06,
+      "loss": 0.6811,
+      "step": 2112
+    },
+    {
+      "epoch": 0.14671573392584364,
+      "grad_norm": 3.313124716405168,
+      "learning_rate": 9.647340903506697e-06,
+      "loss": 0.4289,
+      "step": 2113
+    },
+    {
+      "epoch": 0.14678516872656575,
+      "grad_norm": 5.621777066423199,
+      "learning_rate": 9.646925960034643e-06,
+      "loss": 0.7604,
+      "step": 2114
+    },
+    {
+      "epoch": 0.14685460352728788,
+      "grad_norm": 2.7928116904997076,
+      "learning_rate": 9.646510781526332e-06,
+      "loss": 0.2051,
+      "step": 2115
+    },
+    {
+      "epoch": 0.14692403832801,
+      "grad_norm": 3.8886256796163754,
+      "learning_rate": 9.646095368002765e-06,
+      "loss": 0.45,
+      "step": 2116
+    },
+    {
+      "epoch": 0.14699347312873212,
+      "grad_norm": 2.5670306658311794,
+      "learning_rate": 9.64567971948495e-06,
+      "loss": 0.2052,
+      "step": 2117
+    },
+    {
+      "epoch": 0.14706290792945426,
+      "grad_norm": 7.5435510105966905,
+      "learning_rate": 9.645263835993914e-06,
+      "loss": 0.3888,
+      "step": 2118
+    },
+    {
+      "epoch": 0.14713234273017636,
+      "grad_norm": 4.542963662982822,
+      "learning_rate": 9.644847717550686e-06,
+      "loss": 0.7395,
+      "step": 2119
+    },
+    {
+      "epoch": 0.1472017775308985,
+      "grad_norm": 3.4930400126079584,
+      "learning_rate": 9.64443136417632e-06,
+      "loss": 0.4938,
+      "step": 2120
+    },
+    {
+      "epoch": 0.1472712123316206,
+      "grad_norm": 3.6627284340003055,
+      "learning_rate": 9.644014775891871e-06,
+      "loss": 0.3185,
+      "step": 2121
+    },
+    {
+      "epoch": 0.14734064713234274,
+      "grad_norm": 4.796525961331801,
+      "learning_rate": 9.643597952718409e-06,
+      "loss": 0.8325,
+      "step": 2122
+    },
+    {
+      "epoch": 0.14741008193306485,
+      "grad_norm": 5.5445109667379775,
+      "learning_rate": 9.643180894677015e-06,
+      "loss": 0.7546,
+      "step": 2123
+    },
+    {
+      "epoch": 0.14747951673378698,
+      "grad_norm": 2.7253182669581175,
+      "learning_rate": 9.642763601788788e-06,
+      "loss": 0.1883,
+      "step": 2124
+    },
+    {
+      "epoch": 0.1475489515345091,
+      "grad_norm": 5.122835293498146,
+      "learning_rate": 9.64234607407483e-06,
+      "loss": 0.7419,
+      "step": 2125
+    },
+    {
+      "epoch": 0.14761838633523122,
+      "grad_norm": 4.271000985188102,
+      "learning_rate": 9.641928311556263e-06,
+      "loss": 0.6409,
+      "step": 2126
+    },
+    {
+      "epoch": 0.14768782113595333,
+      "grad_norm": 4.915586052957864,
+      "learning_rate": 9.641510314254214e-06,
+      "loss": 0.7303,
+      "step": 2127
+    },
+    {
+      "epoch": 0.14775725593667546,
+      "grad_norm": 5.174385425179848,
+      "learning_rate": 9.641092082189824e-06,
+      "loss": 0.5006,
+      "step": 2128
+    },
+    {
+      "epoch": 0.14782669073739757,
+      "grad_norm": 2.409503078200908,
+      "learning_rate": 9.640673615384248e-06,
+      "loss": 0.1535,
+      "step": 2129
+    },
+    {
+      "epoch": 0.1478961255381197,
+      "grad_norm": 4.201531251957234,
+      "learning_rate": 9.640254913858653e-06,
+      "loss": 0.5363,
+      "step": 2130
+    },
+    {
+      "epoch": 0.14796556033884184,
+      "grad_norm": 4.754898519537631,
+      "learning_rate": 9.639835977634216e-06,
+      "loss": 0.6777,
+      "step": 2131
+    },
+    {
+      "epoch": 0.14803499513956395,
+      "grad_norm": 4.458480151582614,
+      "learning_rate": 9.639416806732124e-06,
+      "loss": 0.6779,
+      "step": 2132
+    },
+    {
+      "epoch": 0.14810442994028608,
+      "grad_norm": 3.495242742808231,
+      "learning_rate": 9.638997401173581e-06,
+      "loss": 0.3626,
+      "step": 2133
+    },
+    {
+      "epoch": 0.1481738647410082,
+      "grad_norm": 4.003974587317677,
+      "learning_rate": 9.638577760979799e-06,
+      "loss": 0.3901,
+      "step": 2134
+    },
+    {
+      "epoch": 0.14824329954173032,
+      "grad_norm": 4.009722704641441,
+      "learning_rate": 9.638157886172001e-06,
+      "loss": 0.6799,
+      "step": 2135
+    },
+    {
+      "epoch": 0.14831273434245243,
+      "grad_norm": 4.8117589946830694,
+      "learning_rate": 9.637737776771428e-06,
+      "loss": 0.8886,
+      "step": 2136
+    },
+    {
+      "epoch": 0.14838216914317456,
+      "grad_norm": 3.8628286078524794,
+      "learning_rate": 9.637317432799324e-06,
+      "loss": 0.627,
+      "step": 2137
+    },
+    {
+      "epoch": 0.14845160394389667,
+      "grad_norm": 4.806849215836901,
+      "learning_rate": 9.636896854276954e-06,
+      "loss": 0.7726,
+      "step": 2138
+    },
+    {
+      "epoch": 0.1485210387446188,
+      "grad_norm": 4.35950051490445,
+      "learning_rate": 9.636476041225586e-06,
+      "loss": 0.6371,
+      "step": 2139
+    },
+    {
+      "epoch": 0.1485904735453409,
+      "grad_norm": 3.2712297665561283,
+      "learning_rate": 9.636054993666507e-06,
+      "loss": 0.441,
+      "step": 2140
+    },
+    {
+      "epoch": 0.14865990834606305,
+      "grad_norm": 4.355060290067269,
+      "learning_rate": 9.635633711621014e-06,
+      "loss": 0.653,
+      "step": 2141
+    },
+    {
+      "epoch": 0.14872934314678518,
+      "grad_norm": 3.8338356810863337,
+      "learning_rate": 9.63521219511041e-06,
+      "loss": 0.4056,
+      "step": 2142
+    },
+    {
+      "epoch": 0.1487987779475073,
+      "grad_norm": 3.671166130351197,
+      "learning_rate": 9.634790444156021e-06,
+      "loss": 0.5412,
+      "step": 2143
+    },
+    {
+      "epoch": 0.14886821274822942,
+      "grad_norm": 5.13140177444575,
+      "learning_rate": 9.634368458779175e-06,
+      "loss": 0.6732,
+      "step": 2144
+    },
+    {
+      "epoch": 0.14893764754895153,
+      "grad_norm": 3.140632701933939,
+      "learning_rate": 9.633946239001217e-06,
+      "loss": 0.1796,
+      "step": 2145
+    },
+    {
+      "epoch": 0.14900708234967366,
+      "grad_norm": 4.080931149673151,
+      "learning_rate": 9.633523784843502e-06,
+      "loss": 0.4658,
+      "step": 2146
+    },
+    {
+      "epoch": 0.14907651715039577,
+      "grad_norm": 4.0423618324455095,
+      "learning_rate": 9.633101096327396e-06,
+      "loss": 0.6895,
+      "step": 2147
+    },
+    {
+      "epoch": 0.1491459519511179,
+      "grad_norm": 4.199203475655609,
+      "learning_rate": 9.63267817347428e-06,
+      "loss": 0.5258,
+      "step": 2148
+    },
+    {
+      "epoch": 0.14921538675184,
+      "grad_norm": 4.135762111929506,
+      "learning_rate": 9.632255016305545e-06,
+      "loss": 0.5535,
+      "step": 2149
+    },
+    {
+      "epoch": 0.14928482155256215,
+      "grad_norm": 3.55350284521541,
+      "learning_rate": 9.631831624842591e-06,
+      "loss": 0.3804,
+      "step": 2150
+    },
+    {
+      "epoch": 0.14935425635328425,
+      "grad_norm": 3.560772374066174,
+      "learning_rate": 9.631407999106837e-06,
+      "loss": 0.4919,
+      "step": 2151
+    },
+    {
+      "epoch": 0.1494236911540064,
+      "grad_norm": 5.428231971785874,
+      "learning_rate": 9.630984139119706e-06,
+      "loss": 0.7099,
+      "step": 2152
+    },
+    {
+      "epoch": 0.14949312595472852,
+      "grad_norm": 4.283387069547894,
+      "learning_rate": 9.630560044902637e-06,
+      "loss": 0.5947,
+      "step": 2153
+    },
+    {
+      "epoch": 0.14956256075545063,
+      "grad_norm": 3.7978832595563463,
+      "learning_rate": 9.630135716477083e-06,
+      "loss": 0.328,
+      "step": 2154
+    },
+    {
+      "epoch": 0.14963199555617276,
+      "grad_norm": 3.8698691820464473,
+      "learning_rate": 9.629711153864501e-06,
+      "loss": 0.5876,
+      "step": 2155
+    },
+    {
+      "epoch": 0.14970143035689487,
+      "grad_norm": 3.7171291787596044,
+      "learning_rate": 9.62928635708637e-06,
+      "loss": 0.3968,
+      "step": 2156
+    },
+    {
+      "epoch": 0.149770865157617,
+      "grad_norm": 4.010113480677708,
+      "learning_rate": 9.628861326164172e-06,
+      "loss": 0.4367,
+      "step": 2157
+    },
+    {
+      "epoch": 0.1498402999583391,
+      "grad_norm": 5.170529051538263,
+      "learning_rate": 9.628436061119407e-06,
+      "loss": 0.7093,
+      "step": 2158
+    },
+    {
+      "epoch": 0.14990973475906125,
+      "grad_norm": 5.42154246122125,
+      "learning_rate": 9.628010561973583e-06,
+      "loss": 0.8771,
+      "step": 2159
+    },
+    {
+      "epoch": 0.14997916955978335,
+      "grad_norm": 4.560966698215717,
+      "learning_rate": 9.62758482874822e-06,
+      "loss": 0.5701,
+      "step": 2160
+    },
+    {
+      "epoch": 0.1500486043605055,
+      "grad_norm": 4.1083092882864545,
+      "learning_rate": 9.627158861464856e-06,
+      "loss": 0.4977,
+      "step": 2161
+    },
+    {
+      "epoch": 0.1501180391612276,
+      "grad_norm": 4.162130983711981,
+      "learning_rate": 9.62673266014503e-06,
+      "loss": 0.4592,
+      "step": 2162
+    },
+    {
+      "epoch": 0.15018747396194973,
+      "grad_norm": 4.150643483025613,
+      "learning_rate": 9.626306224810304e-06,
+      "loss": 0.5144,
+      "step": 2163
+    },
+    {
+      "epoch": 0.15025690876267186,
+      "grad_norm": 3.52445630944535,
+      "learning_rate": 9.625879555482243e-06,
+      "loss": 0.2949,
+      "step": 2164
+    },
+    {
+      "epoch": 0.15032634356339397,
+      "grad_norm": 4.34367618299048,
+      "learning_rate": 9.625452652182428e-06,
+      "loss": 0.5263,
+      "step": 2165
+    },
+    {
+      "epoch": 0.1503957783641161,
+      "grad_norm": 5.7225432767828535,
+      "learning_rate": 9.625025514932454e-06,
+      "loss": 0.8177,
+      "step": 2166
+    },
+    {
+      "epoch": 0.1504652131648382,
+      "grad_norm": 4.411352366839603,
+      "learning_rate": 9.624598143753921e-06,
+      "loss": 0.6157,
+      "step": 2167
+    },
+    {
+      "epoch": 0.15053464796556035,
+      "grad_norm": 4.672624854610096,
+      "learning_rate": 9.624170538668448e-06,
+      "loss": 0.6514,
+      "step": 2168
+    },
+    {
+      "epoch": 0.15060408276628245,
+      "grad_norm": 5.338776565245194,
+      "learning_rate": 9.62374269969766e-06,
+      "loss": 0.5284,
+      "step": 2169
+    },
+    {
+      "epoch": 0.1506735175670046,
+      "grad_norm": 4.144303880694495,
+      "learning_rate": 9.623314626863199e-06,
+      "loss": 0.669,
+      "step": 2170
+    },
+    {
+      "epoch": 0.1507429523677267,
+      "grad_norm": 6.997013068353923,
+      "learning_rate": 9.622886320186717e-06,
+      "loss": 0.7822,
+      "step": 2171
+    },
+    {
+      "epoch": 0.15081238716844883,
+      "grad_norm": 3.4879408606678557,
+      "learning_rate": 9.622457779689876e-06,
+      "loss": 0.3417,
+      "step": 2172
+    },
+    {
+      "epoch": 0.15088182196917094,
+      "grad_norm": 6.638254983399789,
+      "learning_rate": 9.622029005394351e-06,
+      "loss": 0.6465,
+      "step": 2173
+    },
+    {
+      "epoch": 0.15095125676989307,
+      "grad_norm": 4.599317854997834,
+      "learning_rate": 9.621599997321829e-06,
+      "loss": 0.6965,
+      "step": 2174
+    },
+    {
+      "epoch": 0.15102069157061518,
+      "grad_norm": 4.042166253060516,
+      "learning_rate": 9.621170755494007e-06,
+      "loss": 0.53,
+      "step": 2175
+    },
+    {
+      "epoch": 0.1510901263713373,
+      "grad_norm": 4.5173353425869776,
+      "learning_rate": 9.6207412799326e-06,
+      "loss": 0.6743,
+      "step": 2176
+    },
+    {
+      "epoch": 0.15115956117205945,
+      "grad_norm": 4.099354829361525,
+      "learning_rate": 9.620311570659326e-06,
+      "loss": 0.4366,
+      "step": 2177
+    },
+    {
+      "epoch": 0.15122899597278155,
+      "grad_norm": 3.7265663868190066,
+      "learning_rate": 9.619881627695921e-06,
+      "loss": 0.3429,
+      "step": 2178
+    },
+    {
+      "epoch": 0.1512984307735037,
+      "grad_norm": 3.5766231935228876,
+      "learning_rate": 9.619451451064132e-06,
+      "loss": 0.4632,
+      "step": 2179
+    },
+    {
+      "epoch": 0.1513678655742258,
+      "grad_norm": 4.191523906066392,
+      "learning_rate": 9.619021040785717e-06,
+      "loss": 0.2871,
+      "step": 2180
+    },
+    {
+      "epoch": 0.15143730037494793,
+      "grad_norm": 5.094071493747692,
+      "learning_rate": 9.618590396882443e-06,
+      "loss": 0.7728,
+      "step": 2181
+    },
+    {
+      "epoch": 0.15150673517567004,
+      "grad_norm": 4.585236149998841,
+      "learning_rate": 9.618159519376094e-06,
+      "loss": 0.6826,
+      "step": 2182
+    },
+    {
+      "epoch": 0.15157616997639217,
+      "grad_norm": 3.561424487335961,
+      "learning_rate": 9.61772840828846e-06,
+      "loss": 0.3847,
+      "step": 2183
+    },
+    {
+      "epoch": 0.15164560477711428,
+      "grad_norm": 4.479997799178344,
+      "learning_rate": 9.617297063641352e-06,
+      "loss": 0.554,
+      "step": 2184
+    },
+    {
+      "epoch": 0.1517150395778364,
+      "grad_norm": 4.894157220215697,
+      "learning_rate": 9.61686548545658e-06,
+      "loss": 0.6575,
+      "step": 2185
+    },
+    {
+      "epoch": 0.15178447437855852,
+      "grad_norm": 3.1848964285422348,
+      "learning_rate": 9.616433673755977e-06,
+      "loss": 0.4898,
+      "step": 2186
+    },
+    {
+      "epoch": 0.15185390917928066,
+      "grad_norm": 4.301072185107848,
+      "learning_rate": 9.616001628561383e-06,
+      "loss": 0.5825,
+      "step": 2187
+    },
+    {
+      "epoch": 0.1519233439800028,
+      "grad_norm": 4.001166271475615,
+      "learning_rate": 9.61556934989465e-06,
+      "loss": 0.6727,
+      "step": 2188
+    },
+    {
+      "epoch": 0.1519927787807249,
+      "grad_norm": 4.029510516138517,
+      "learning_rate": 9.615136837777642e-06,
+      "loss": 0.5502,
+      "step": 2189
+    },
+    {
+      "epoch": 0.15206221358144703,
+      "grad_norm": 4.045130807084963,
+      "learning_rate": 9.614704092232236e-06,
+      "loss": 0.4201,
+      "step": 2190
+    },
+    {
+      "epoch": 0.15213164838216914,
+      "grad_norm": 4.799132886434701,
+      "learning_rate": 9.614271113280317e-06,
+      "loss": 0.6717,
+      "step": 2191
+    },
+    {
+      "epoch": 0.15220108318289127,
+      "grad_norm": 4.417803179709248,
+      "learning_rate": 9.61383790094379e-06,
+      "loss": 0.5045,
+      "step": 2192
+    },
+    {
+      "epoch": 0.15227051798361338,
+      "grad_norm": 3.0958310419921995,
+      "learning_rate": 9.613404455244559e-06,
+      "loss": 0.2164,
+      "step": 2193
+    },
+    {
+      "epoch": 0.15233995278433551,
+      "grad_norm": 4.987880669311369,
+      "learning_rate": 9.612970776204553e-06,
+      "loss": 0.7546,
+      "step": 2194
+    },
+    {
+      "epoch": 0.15240938758505762,
+      "grad_norm": 3.1490928292903564,
+      "learning_rate": 9.612536863845704e-06,
+      "loss": 0.3057,
+      "step": 2195
+    },
+    {
+      "epoch": 0.15247882238577976,
+      "grad_norm": 4.859826456677941,
+      "learning_rate": 9.612102718189962e-06,
+      "loss": 0.6992,
+      "step": 2196
+    },
+    {
+      "epoch": 0.15254825718650186,
+      "grad_norm": 3.759370550433794,
+      "learning_rate": 9.611668339259281e-06,
+      "loss": 0.4494,
+      "step": 2197
+    },
+    {
+      "epoch": 0.152617691987224,
+      "grad_norm": 4.696407139683973,
+      "learning_rate": 9.611233727075635e-06,
+      "loss": 0.59,
+      "step": 2198
+    },
+    {
+      "epoch": 0.15268712678794613,
+      "grad_norm": 4.251295120929709,
+      "learning_rate": 9.610798881661005e-06,
+      "loss": 0.554,
+      "step": 2199
+    },
+    {
+      "epoch": 0.15275656158866824,
+      "grad_norm": 5.010389993922083,
+      "learning_rate": 9.610363803037384e-06,
+      "loss": 0.2998,
+      "step": 2200
+    },
+    {
+      "epoch": 0.15282599638939037,
+      "grad_norm": 4.599848443036542,
+      "learning_rate": 9.60992849122678e-06,
+      "loss": 0.446,
+      "step": 2201
+    },
+    {
+      "epoch": 0.15289543119011248,
+      "grad_norm": 4.504389324676448,
+      "learning_rate": 9.60949294625121e-06,
+      "loss": 0.4754,
+      "step": 2202
+    },
+    {
+      "epoch": 0.15296486599083461,
+      "grad_norm": 3.983256582971974,
+      "learning_rate": 9.609057168132701e-06,
+      "loss": 0.5499,
+      "step": 2203
+    },
+    {
+      "epoch": 0.15303430079155672,
+      "grad_norm": 5.459218118067474,
+      "learning_rate": 9.608621156893298e-06,
+      "loss": 0.6484,
+      "step": 2204
+    },
+    {
+      "epoch": 0.15310373559227886,
+      "grad_norm": 4.489706173966664,
+      "learning_rate": 9.60818491255505e-06,
+      "loss": 0.5891,
+      "step": 2205
+    },
+    {
+      "epoch": 0.15317317039300096,
+      "grad_norm": 4.859178105499087,
+      "learning_rate": 9.607748435140025e-06,
+      "loss": 0.6211,
+      "step": 2206
+    },
+    {
+      "epoch": 0.1532426051937231,
+      "grad_norm": 5.534765668202554,
+      "learning_rate": 9.607311724670297e-06,
+      "loss": 0.701,
+      "step": 2207
+    },
+    {
+      "epoch": 0.1533120399944452,
+      "grad_norm": 4.723858289525773,
+      "learning_rate": 9.606874781167957e-06,
+      "loss": 0.5543,
+      "step": 2208
+    },
+    {
+      "epoch": 0.15338147479516734,
+      "grad_norm": 4.39705125249002,
+      "learning_rate": 9.606437604655103e-06,
+      "loss": 0.5024,
+      "step": 2209
+    },
+    {
+      "epoch": 0.15345090959588947,
+      "grad_norm": 4.221118384057801,
+      "learning_rate": 9.606000195153846e-06,
+      "loss": 0.5979,
+      "step": 2210
+    },
+    {
+      "epoch": 0.15352034439661158,
+      "grad_norm": 4.093665115879449,
+      "learning_rate": 9.605562552686312e-06,
+      "loss": 0.4684,
+      "step": 2211
+    },
+    {
+      "epoch": 0.15358977919733371,
+      "grad_norm": 5.669613904245084,
+      "learning_rate": 9.605124677274637e-06,
+      "loss": 0.6409,
+      "step": 2212
+    },
+    {
+      "epoch": 0.15365921399805582,
+      "grad_norm": 3.7993841474952244,
+      "learning_rate": 9.604686568940966e-06,
+      "loss": 0.3656,
+      "step": 2213
+    },
+    {
+      "epoch": 0.15372864879877796,
+      "grad_norm": 3.2532120597641616,
+      "learning_rate": 9.604248227707457e-06,
+      "loss": 0.2739,
+      "step": 2214
+    },
+    {
+      "epoch": 0.15379808359950006,
+      "grad_norm": 4.865153529138012,
+      "learning_rate": 9.603809653596286e-06,
+      "loss": 0.7091,
+      "step": 2215
+    },
+    {
+      "epoch": 0.1538675184002222,
+      "grad_norm": 4.698515233693294,
+      "learning_rate": 9.60337084662963e-06,
+      "loss": 0.7606,
+      "step": 2216
+    },
+    {
+      "epoch": 0.1539369532009443,
+      "grad_norm": 4.328576777087992,
+      "learning_rate": 9.602931806829688e-06,
+      "loss": 0.585,
+      "step": 2217
+    },
+    {
+      "epoch": 0.15400638800166644,
+      "grad_norm": 4.772455909519958,
+      "learning_rate": 9.602492534218662e-06,
+      "loss": 0.6278,
+      "step": 2218
+    },
+    {
+      "epoch": 0.15407582280238855,
+      "grad_norm": 3.090391041170871,
+      "learning_rate": 9.602053028818772e-06,
+      "loss": 0.2159,
+      "step": 2219
+    },
+    {
+      "epoch": 0.15414525760311068,
+      "grad_norm": 3.8649101984275194,
+      "learning_rate": 9.601613290652246e-06,
+      "loss": 0.5631,
+      "step": 2220
+    },
+    {
+      "epoch": 0.15421469240383281,
+      "grad_norm": 8.536454886869532,
+      "learning_rate": 9.601173319741327e-06,
+      "loss": 0.6514,
+      "step": 2221
+    },
+    {
+      "epoch": 0.15428412720455492,
+      "grad_norm": 4.58777928140159,
+      "learning_rate": 9.60073311610827e-06,
+      "loss": 0.6224,
+      "step": 2222
+    },
+    {
+      "epoch": 0.15435356200527706,
+      "grad_norm": 5.352664238632613,
+      "learning_rate": 9.600292679775338e-06,
+      "loss": 0.7644,
+      "step": 2223
+    },
+    {
+      "epoch": 0.15442299680599916,
+      "grad_norm": 4.90504377939357,
+      "learning_rate": 9.599852010764808e-06,
+      "loss": 0.6574,
+      "step": 2224
+    },
+    {
+      "epoch": 0.1544924316067213,
+      "grad_norm": 4.18308356178959,
+      "learning_rate": 9.599411109098967e-06,
+      "loss": 0.4716,
+      "step": 2225
+    },
+    {
+      "epoch": 0.1545618664074434,
+      "grad_norm": 4.6087451639829045,
+      "learning_rate": 9.598969974800116e-06,
+      "loss": 0.5011,
+      "step": 2226
+    },
+    {
+      "epoch": 0.15463130120816554,
+      "grad_norm": 4.833187987561192,
+      "learning_rate": 9.598528607890569e-06,
+      "loss": 0.749,
+      "step": 2227
+    },
+    {
+      "epoch": 0.15470073600888765,
+      "grad_norm": 3.6491229759270984,
+      "learning_rate": 9.59808700839265e-06,
+      "loss": 0.4544,
+      "step": 2228
+    },
+    {
+      "epoch": 0.15477017080960978,
+      "grad_norm": 4.974073027682453,
+      "learning_rate": 9.597645176328691e-06,
+      "loss": 0.5836,
+      "step": 2229
+    },
+    {
+      "epoch": 0.1548396056103319,
+      "grad_norm": 2.9360282418807655,
+      "learning_rate": 9.597203111721042e-06,
+      "loss": 0.213,
+      "step": 2230
+    },
+    {
+      "epoch": 0.15490904041105402,
+      "grad_norm": 5.232706077129518,
+      "learning_rate": 9.596760814592061e-06,
+      "loss": 0.7803,
+      "step": 2231
+    },
+    {
+      "epoch": 0.15497847521177613,
+      "grad_norm": 4.371667142036753,
+      "learning_rate": 9.59631828496412e-06,
+      "loss": 0.6815,
+      "step": 2232
+    },
+    {
+      "epoch": 0.15504791001249826,
+      "grad_norm": 4.854097833053744,
+      "learning_rate": 9.595875522859604e-06,
+      "loss": 0.7341,
+      "step": 2233
+    },
+    {
+      "epoch": 0.1551173448132204,
+      "grad_norm": 5.009857229455537,
+      "learning_rate": 9.5954325283009e-06,
+      "loss": 0.456,
+      "step": 2234
+    },
+    {
+      "epoch": 0.1551867796139425,
+      "grad_norm": 4.086937530145662,
+      "learning_rate": 9.594989301310423e-06,
+      "loss": 0.4557,
+      "step": 2235
+    },
+    {
+      "epoch": 0.15525621441466464,
+      "grad_norm": 4.628705263050036,
+      "learning_rate": 9.594545841910586e-06,
+      "loss": 0.6023,
+      "step": 2236
+    },
+    {
+      "epoch": 0.15532564921538675,
+      "grad_norm": 4.3001149926052875,
+      "learning_rate": 9.594102150123816e-06,
+      "loss": 0.5026,
+      "step": 2237
+    },
+    {
+      "epoch": 0.15539508401610888,
+      "grad_norm": 4.127495344339525,
+      "learning_rate": 9.593658225972562e-06,
+      "loss": 0.599,
+      "step": 2238
+    },
+    {
+      "epoch": 0.155464518816831,
+      "grad_norm": 5.060900644131906,
+      "learning_rate": 9.593214069479272e-06,
+      "loss": 0.5588,
+      "step": 2239
+    },
+    {
+      "epoch": 0.15553395361755312,
+      "grad_norm": 5.4176957997946715,
+      "learning_rate": 9.592769680666409e-06,
+      "loss": 0.6048,
+      "step": 2240
+    },
+    {
+      "epoch": 0.15560338841827523,
+      "grad_norm": 4.793367140954558,
+      "learning_rate": 9.592325059556455e-06,
+      "loss": 0.623,
+      "step": 2241
+    },
+    {
+      "epoch": 0.15567282321899736,
+      "grad_norm": 4.3226670598415184,
+      "learning_rate": 9.591880206171896e-06,
+      "loss": 0.657,
+      "step": 2242
+    },
+    {
+      "epoch": 0.15574225801971947,
+      "grad_norm": 3.261919850240303,
+      "learning_rate": 9.591435120535231e-06,
+      "loss": 0.1683,
+      "step": 2243
+    },
+    {
+      "epoch": 0.1558116928204416,
+      "grad_norm": 2.3377488856361692,
+      "learning_rate": 9.590989802668975e-06,
+      "loss": 0.2288,
+      "step": 2244
+    },
+    {
+      "epoch": 0.15588112762116374,
+      "grad_norm": 4.095308226757974,
+      "learning_rate": 9.59054425259565e-06,
+      "loss": 0.4485,
+      "step": 2245
+    },
+    {
+      "epoch": 0.15595056242188585,
+      "grad_norm": 2.840755720801834,
+      "learning_rate": 9.59009847033779e-06,
+      "loss": 0.2508,
+      "step": 2246
+    },
+    {
+      "epoch": 0.15601999722260798,
+      "grad_norm": 5.410580650052158,
+      "learning_rate": 9.589652455917943e-06,
+      "loss": 0.5681,
+      "step": 2247
+    },
+    {
+      "epoch": 0.1560894320233301,
+      "grad_norm": 3.7555928328296013,
+      "learning_rate": 9.58920620935867e-06,
+      "loss": 0.5221,
+      "step": 2248
+    },
+    {
+      "epoch": 0.15615886682405222,
+      "grad_norm": 4.371211165196956,
+      "learning_rate": 9.588759730682538e-06,
+      "loss": 0.695,
+      "step": 2249
+    },
+    {
+      "epoch": 0.15622830162477433,
+      "grad_norm": 3.6418905183885792,
+      "learning_rate": 9.588313019912132e-06,
+      "loss": 0.5032,
+      "step": 2250
+    },
+    {
+      "epoch": 0.15629773642549646,
+      "grad_norm": 4.747795675500118,
+      "learning_rate": 9.587866077070046e-06,
+      "loss": 0.7966,
+      "step": 2251
+    },
+    {
+      "epoch": 0.15636717122621857,
+      "grad_norm": 5.318563700813509,
+      "learning_rate": 9.587418902178885e-06,
+      "loss": 0.4884,
+      "step": 2252
+    },
+    {
+      "epoch": 0.1564366060269407,
+      "grad_norm": 4.468146392052785,
+      "learning_rate": 9.586971495261268e-06,
+      "loss": 0.6217,
+      "step": 2253
+    },
+    {
+      "epoch": 0.1565060408276628,
+      "grad_norm": 4.499484728799713,
+      "learning_rate": 9.586523856339822e-06,
+      "loss": 0.6122,
+      "step": 2254
+    },
+    {
+      "epoch": 0.15657547562838495,
+      "grad_norm": 4.502491887358075,
+      "learning_rate": 9.58607598543719e-06,
+      "loss": 0.5106,
+      "step": 2255
+    },
+    {
+      "epoch": 0.15664491042910708,
+      "grad_norm": 4.498915940861701,
+      "learning_rate": 9.585627882576023e-06,
+      "loss": 0.553,
+      "step": 2256
+    },
+    {
+      "epoch": 0.1567143452298292,
+      "grad_norm": 3.350750672659533,
+      "learning_rate": 9.585179547778987e-06,
+      "loss": 0.2702,
+      "step": 2257
+    },
+    {
+      "epoch": 0.15678378003055132,
+      "grad_norm": 3.931956938849642,
+      "learning_rate": 9.584730981068759e-06,
+      "loss": 0.4332,
+      "step": 2258
+    },
+    {
+      "epoch": 0.15685321483127343,
+      "grad_norm": 2.8802322444599913,
+      "learning_rate": 9.584282182468025e-06,
+      "loss": 0.1681,
+      "step": 2259
+    },
+    {
+      "epoch": 0.15692264963199556,
+      "grad_norm": 4.728179444858324,
+      "learning_rate": 9.583833151999488e-06,
+      "loss": 0.5674,
+      "step": 2260
+    },
+    {
+      "epoch": 0.15699208443271767,
+      "grad_norm": 4.505436592965721,
+      "learning_rate": 9.583383889685854e-06,
+      "loss": 0.5343,
+      "step": 2261
+    },
+    {
+      "epoch": 0.1570615192334398,
+      "grad_norm": 4.50883175774027,
+      "learning_rate": 9.58293439554985e-06,
+      "loss": 0.5165,
+      "step": 2262
+    },
+    {
+      "epoch": 0.1571309540341619,
+      "grad_norm": 3.347169107573412,
+      "learning_rate": 9.582484669614212e-06,
+      "loss": 0.4138,
+      "step": 2263
+    },
+    {
+      "epoch": 0.15720038883488405,
+      "grad_norm": 14.352319034529115,
+      "learning_rate": 9.582034711901684e-06,
+      "loss": 0.5213,
+      "step": 2264
+    },
+    {
+      "epoch": 0.15726982363560615,
+      "grad_norm": 4.016815832051416,
+      "learning_rate": 9.581584522435025e-06,
+      "loss": 0.3601,
+      "step": 2265
+    },
+    {
+      "epoch": 0.1573392584363283,
+      "grad_norm": 4.760764139599011,
+      "learning_rate": 9.581134101237004e-06,
+      "loss": 0.7454,
+      "step": 2266
+    },
+    {
+      "epoch": 0.15740869323705042,
+      "grad_norm": 4.032330445073442,
+      "learning_rate": 9.580683448330405e-06,
+      "loss": 0.3894,
+      "step": 2267
+    },
+    {
+      "epoch": 0.15747812803777253,
+      "grad_norm": 5.274412689107398,
+      "learning_rate": 9.58023256373802e-06,
+      "loss": 0.6202,
+      "step": 2268
+    },
+    {
+      "epoch": 0.15754756283849466,
+      "grad_norm": 4.023481654255643,
+      "learning_rate": 9.579781447482657e-06,
+      "loss": 0.4831,
+      "step": 2269
+    },
+    {
+      "epoch": 0.15761699763921677,
+      "grad_norm": 4.11254151064439,
+      "learning_rate": 9.57933009958713e-06,
+      "loss": 0.5913,
+      "step": 2270
+    },
+    {
+      "epoch": 0.1576864324399389,
+      "grad_norm": 4.192353973138778,
+      "learning_rate": 9.578878520074268e-06,
+      "loss": 0.5148,
+      "step": 2271
+    },
+    {
+      "epoch": 0.157755867240661,
+      "grad_norm": 4.462575874594395,
+      "learning_rate": 9.57842670896691e-06,
+      "loss": 0.587,
+      "step": 2272
+    },
+    {
+      "epoch": 0.15782530204138315,
+      "grad_norm": 4.357870644536954,
+      "learning_rate": 9.577974666287913e-06,
+      "loss": 0.5564,
+      "step": 2273
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 4.8292456288199945,
+      "learning_rate": 9.577522392060134e-06,
+      "loss": 0.5367,
+      "step": 2274
+    },
+    {
+      "epoch": 0.1579641716428274,
+      "grad_norm": 4.401807915568853,
+      "learning_rate": 9.577069886306455e-06,
+      "loss": 0.6625,
+      "step": 2275
+    },
+    {
+      "epoch": 0.1580336064435495,
+      "grad_norm": 2.9934394086346443,
+      "learning_rate": 9.57661714904976e-06,
+      "loss": 0.3232,
+      "step": 2276
+    },
+    {
+      "epoch": 0.15810304124427163,
+      "grad_norm": 4.82092681484257,
+      "learning_rate": 9.57616418031295e-06,
+      "loss": 0.7151,
+      "step": 2277
+    },
+    {
+      "epoch": 0.15817247604499374,
+      "grad_norm": 7.375034875308307,
+      "learning_rate": 9.575710980118931e-06,
+      "loss": 0.5792,
+      "step": 2278
+    },
+    {
+      "epoch": 0.15824191084571587,
+      "grad_norm": 5.080238516115881,
+      "learning_rate": 9.57525754849063e-06,
+      "loss": 0.7074,
+      "step": 2279
+    },
+    {
+      "epoch": 0.158311345646438,
+      "grad_norm": 4.848069078435341,
+      "learning_rate": 9.574803885450979e-06,
+      "loss": 0.7688,
+      "step": 2280
+    },
+    {
+      "epoch": 0.1583807804471601,
+      "grad_norm": 4.3755574281481495,
+      "learning_rate": 9.574349991022924e-06,
+      "loss": 0.3341,
+      "step": 2281
+    },
+    {
+      "epoch": 0.15845021524788225,
+      "grad_norm": 2.9668206570989697,
+      "learning_rate": 9.573895865229424e-06,
+      "loss": 0.3524,
+      "step": 2282
+    },
+    {
+      "epoch": 0.15851965004860435,
+      "grad_norm": 3.5247035725124625,
+      "learning_rate": 9.573441508093445e-06,
+      "loss": 0.5685,
+      "step": 2283
+    },
+    {
+      "epoch": 0.1585890848493265,
+      "grad_norm": 3.604211417509853,
+      "learning_rate": 9.572986919637972e-06,
+      "loss": 0.5509,
+      "step": 2284
+    },
+    {
+      "epoch": 0.1586585196500486,
+      "grad_norm": 3.4357426443151473,
+      "learning_rate": 9.572532099885992e-06,
+      "loss": 0.2446,
+      "step": 2285
+    },
+    {
+      "epoch": 0.15872795445077073,
+      "grad_norm": 4.267981730472809,
+      "learning_rate": 9.572077048860516e-06,
+      "loss": 0.5516,
+      "step": 2286
+    },
+    {
+      "epoch": 0.15879738925149284,
+      "grad_norm": 4.705510938624109,
+      "learning_rate": 9.571621766584554e-06,
+      "loss": 0.5019,
+      "step": 2287
+    },
+    {
+      "epoch": 0.15886682405221497,
+      "grad_norm": 4.077655349108888,
+      "learning_rate": 9.571166253081137e-06,
+      "loss": 0.4756,
+      "step": 2288
+    },
+    {
+      "epoch": 0.15893625885293708,
+      "grad_norm": 4.622230707108691,
+      "learning_rate": 9.570710508373304e-06,
+      "loss": 0.6148,
+      "step": 2289
+    },
+    {
+      "epoch": 0.1590056936536592,
+      "grad_norm": 3.1458093793763147,
+      "learning_rate": 9.570254532484106e-06,
+      "loss": 0.2311,
+      "step": 2290
+    },
+    {
+      "epoch": 0.15907512845438135,
+      "grad_norm": 3.7905227092981617,
+      "learning_rate": 9.569798325436603e-06,
+      "loss": 0.5566,
+      "step": 2291
+    },
+    {
+      "epoch": 0.15914456325510346,
+      "grad_norm": 4.410040109309413,
+      "learning_rate": 9.569341887253875e-06,
+      "loss": 0.5267,
+      "step": 2292
+    },
+    {
+      "epoch": 0.1592139980558256,
+      "grad_norm": 4.052569758457331,
+      "learning_rate": 9.568885217959002e-06,
+      "loss": 0.3068,
+      "step": 2293
+    },
+    {
+      "epoch": 0.1592834328565477,
+      "grad_norm": 4.754984411285499,
+      "learning_rate": 9.568428317575086e-06,
+      "loss": 0.548,
+      "step": 2294
+    },
+    {
+      "epoch": 0.15935286765726983,
+      "grad_norm": 4.581153026015169,
+      "learning_rate": 9.567971186125236e-06,
+      "loss": 0.7509,
+      "step": 2295
+    },
+    {
+      "epoch": 0.15942230245799194,
+      "grad_norm": 4.615198944833523,
+      "learning_rate": 9.567513823632573e-06,
+      "loss": 0.2786,
+      "step": 2296
+    },
+    {
+      "epoch": 0.15949173725871407,
+      "grad_norm": 4.113398900341135,
+      "learning_rate": 9.567056230120227e-06,
+      "loss": 0.465,
+      "step": 2297
+    },
+    {
+      "epoch": 0.15956117205943618,
+      "grad_norm": 3.6507246939887468,
+      "learning_rate": 9.566598405611344e-06,
+      "loss": 0.5308,
+      "step": 2298
+    },
+    {
+      "epoch": 0.15963060686015831,
+      "grad_norm": 4.113783316427625,
+      "learning_rate": 9.566140350129084e-06,
+      "loss": 0.4454,
+      "step": 2299
+    },
+    {
+      "epoch": 0.15970004166088042,
+      "grad_norm": 4.174566137807122,
+      "learning_rate": 9.565682063696611e-06,
+      "loss": 0.4666,
+      "step": 2300
+    },
+    {
+      "epoch": 0.15976947646160256,
+      "grad_norm": 3.289926752244745,
+      "learning_rate": 9.565223546337106e-06,
+      "loss": 0.32,
+      "step": 2301
+    },
+    {
+      "epoch": 0.1598389112623247,
+      "grad_norm": 3.481615028270102,
+      "learning_rate": 9.56476479807376e-06,
+      "loss": 0.463,
+      "step": 2302
+    },
+    {
+      "epoch": 0.1599083460630468,
+      "grad_norm": 3.2731810528337966,
+      "learning_rate": 9.564305818929776e-06,
+      "loss": 0.3808,
+      "step": 2303
+    },
+    {
+      "epoch": 0.15997778086376893,
+      "grad_norm": 3.617734992367984,
+      "learning_rate": 9.563846608928366e-06,
+      "loss": 0.3764,
+      "step": 2304
+    },
+    {
+      "epoch": 0.16004721566449104,
+      "grad_norm": 3.9754828886928397,
+      "learning_rate": 9.563387168092762e-06,
+      "loss": 0.298,
+      "step": 2305
+    },
+    {
+      "epoch": 0.16011665046521317,
+      "grad_norm": 4.30938466323922,
+      "learning_rate": 9.562927496446197e-06,
+      "loss": 0.6277,
+      "step": 2306
+    },
+    {
+      "epoch": 0.16018608526593528,
+      "grad_norm": 3.544408251240936,
+      "learning_rate": 9.562467594011924e-06,
+      "loss": 0.3815,
+      "step": 2307
+    },
+    {
+      "epoch": 0.16025552006665741,
+      "grad_norm": 6.458064427857517,
+      "learning_rate": 9.562007460813201e-06,
+      "loss": 0.8546,
+      "step": 2308
+    },
+    {
+      "epoch": 0.16032495486737952,
+      "grad_norm": 4.2848439665283165,
+      "learning_rate": 9.561547096873305e-06,
+      "loss": 0.5042,
+      "step": 2309
+    },
+    {
+      "epoch": 0.16039438966810166,
+      "grad_norm": 4.672207480846155,
+      "learning_rate": 9.561086502215517e-06,
+      "loss": 0.6461,
+      "step": 2310
+    },
+    {
+      "epoch": 0.16046382446882376,
+      "grad_norm": 3.849411890887942,
+      "learning_rate": 9.560625676863135e-06,
+      "loss": 0.4879,
+      "step": 2311
+    },
+    {
+      "epoch": 0.1605332592695459,
+      "grad_norm": 4.409426410111942,
+      "learning_rate": 9.560164620839468e-06,
+      "loss": 0.49,
+      "step": 2312
+    },
+    {
+      "epoch": 0.16060269407026803,
+      "grad_norm": 4.271218248722517,
+      "learning_rate": 9.559703334167834e-06,
+      "loss": 0.2877,
+      "step": 2313
+    },
+    {
+      "epoch": 0.16067212887099014,
+      "grad_norm": 5.275863020656294,
+      "learning_rate": 9.559241816871565e-06,
+      "loss": 0.6568,
+      "step": 2314
+    },
+    {
+      "epoch": 0.16074156367171227,
+      "grad_norm": 4.718131238979712,
+      "learning_rate": 9.558780068974006e-06,
+      "loss": 0.7289,
+      "step": 2315
+    },
+    {
+      "epoch": 0.16081099847243438,
+      "grad_norm": 4.208251730114253,
+      "learning_rate": 9.558318090498507e-06,
+      "loss": 0.4908,
+      "step": 2316
+    },
+    {
+      "epoch": 0.16088043327315651,
+      "grad_norm": 3.5384967955225304,
+      "learning_rate": 9.557855881468437e-06,
+      "loss": 0.4776,
+      "step": 2317
+    },
+    {
+      "epoch": 0.16094986807387862,
+      "grad_norm": 3.38593928192316,
+      "learning_rate": 9.557393441907175e-06,
+      "loss": 0.4885,
+      "step": 2318
+    },
+    {
+      "epoch": 0.16101930287460076,
+      "grad_norm": 4.467286604842763,
+      "learning_rate": 9.55693077183811e-06,
+      "loss": 0.6762,
+      "step": 2319
+    },
+    {
+      "epoch": 0.16108873767532286,
+      "grad_norm": 3.923019104578957,
+      "learning_rate": 9.556467871284643e-06,
+      "loss": 0.2991,
+      "step": 2320
+    },
+    {
+      "epoch": 0.161158172476045,
+      "grad_norm": 3.8786260218939463,
+      "learning_rate": 9.556004740270188e-06,
+      "loss": 0.3058,
+      "step": 2321
+    },
+    {
+      "epoch": 0.1612276072767671,
+      "grad_norm": 3.4827282875526056,
+      "learning_rate": 9.555541378818168e-06,
+      "loss": 0.2017,
+      "step": 2322
+    },
+    {
+      "epoch": 0.16129704207748924,
+      "grad_norm": 3.893391384235062,
+      "learning_rate": 9.55507778695202e-06,
+      "loss": 0.4303,
+      "step": 2323
+    },
+    {
+      "epoch": 0.16136647687821135,
+      "grad_norm": 5.119219559535707,
+      "learning_rate": 9.554613964695189e-06,
+      "loss": 0.6286,
+      "step": 2324
+    },
+    {
+      "epoch": 0.16143591167893348,
+      "grad_norm": 4.024923499510395,
+      "learning_rate": 9.55414991207114e-06,
+      "loss": 0.7174,
+      "step": 2325
+    },
+    {
+      "epoch": 0.16150534647965561,
+      "grad_norm": 5.2836129756977295,
+      "learning_rate": 9.553685629103344e-06,
+      "loss": 0.4814,
+      "step": 2326
+    },
+    {
+      "epoch": 0.16157478128037772,
+      "grad_norm": 3.2169923076170575,
+      "learning_rate": 9.553221115815278e-06,
+      "loss": 0.2713,
+      "step": 2327
+    },
+    {
+      "epoch": 0.16164421608109986,
+      "grad_norm": 2.655304326401075,
+      "learning_rate": 9.552756372230442e-06,
+      "loss": 0.2459,
+      "step": 2328
+    },
+    {
+      "epoch": 0.16171365088182196,
+      "grad_norm": 4.579757588695799,
+      "learning_rate": 9.552291398372342e-06,
+      "loss": 0.3437,
+      "step": 2329
+    },
+    {
+      "epoch": 0.1617830856825441,
+      "grad_norm": 4.225758348469034,
+      "learning_rate": 9.551826194264493e-06,
+      "loss": 0.4447,
+      "step": 2330
+    },
+    {
+      "epoch": 0.1618525204832662,
+      "grad_norm": 3.5043011752638007,
+      "learning_rate": 9.551360759930424e-06,
+      "loss": 0.3426,
+      "step": 2331
+    },
+    {
+      "epoch": 0.16192195528398834,
+      "grad_norm": 2.9642269184920425,
+      "learning_rate": 9.550895095393679e-06,
+      "loss": 0.2659,
+      "step": 2332
+    },
+    {
+      "epoch": 0.16199139008471045,
+      "grad_norm": 4.60677120752096,
+      "learning_rate": 9.550429200677811e-06,
+      "loss": 0.6668,
+      "step": 2333
+    },
+    {
+      "epoch": 0.16206082488543258,
+      "grad_norm": 3.425591902262055,
+      "learning_rate": 9.549963075806383e-06,
+      "loss": 0.3882,
+      "step": 2334
+    },
+    {
+      "epoch": 0.1621302596861547,
+      "grad_norm": 4.1161195158159245,
+      "learning_rate": 9.549496720802971e-06,
+      "loss": 0.5028,
+      "step": 2335
+    },
+    {
+      "epoch": 0.16219969448687682,
+      "grad_norm": 3.5550127838083743,
+      "learning_rate": 9.549030135691161e-06,
+      "loss": 0.4077,
+      "step": 2336
+    },
+    {
+      "epoch": 0.16226912928759896,
+      "grad_norm": 4.53908049728783,
+      "learning_rate": 9.548563320494557e-06,
+      "loss": 0.6878,
+      "step": 2337
+    },
+    {
+      "epoch": 0.16233856408832106,
+      "grad_norm": 3.709617337055653,
+      "learning_rate": 9.548096275236766e-06,
+      "loss": 0.6337,
+      "step": 2338
+    },
+    {
+      "epoch": 0.1624079988890432,
+      "grad_norm": 4.3936284301072135,
+      "learning_rate": 9.547628999941413e-06,
+      "loss": 0.5906,
+      "step": 2339
+    },
+    {
+      "epoch": 0.1624774336897653,
+      "grad_norm": 4.046642486700777,
+      "learning_rate": 9.54716149463213e-06,
+      "loss": 0.618,
+      "step": 2340
+    },
+    {
+      "epoch": 0.16254686849048744,
+      "grad_norm": 3.4111402312648753,
+      "learning_rate": 9.546693759332564e-06,
+      "loss": 0.3449,
+      "step": 2341
+    },
+    {
+      "epoch": 0.16261630329120955,
+      "grad_norm": 2.786881927263152,
+      "learning_rate": 9.546225794066373e-06,
+      "loss": 0.2162,
+      "step": 2342
+    },
+    {
+      "epoch": 0.16268573809193168,
+      "grad_norm": 3.730035142246005,
+      "learning_rate": 9.545757598857227e-06,
+      "loss": 0.4596,
+      "step": 2343
+    },
+    {
+      "epoch": 0.1627551728926538,
+      "grad_norm": 3.098781151664003,
+      "learning_rate": 9.545289173728805e-06,
+      "loss": 0.472,
+      "step": 2344
+    },
+    {
+      "epoch": 0.16282460769337592,
+      "grad_norm": 3.9739797971503217,
+      "learning_rate": 9.544820518704798e-06,
+      "loss": 0.5474,
+      "step": 2345
+    },
+    {
+      "epoch": 0.16289404249409803,
+      "grad_norm": 4.808088174047386,
+      "learning_rate": 9.544351633808914e-06,
+      "loss": 0.283,
+      "step": 2346
+    },
+    {
+      "epoch": 0.16296347729482016,
+      "grad_norm": 4.018741365451562,
+      "learning_rate": 9.543882519064866e-06,
+      "loss": 0.401,
+      "step": 2347
+    },
+    {
+      "epoch": 0.1630329120955423,
+      "grad_norm": 4.679918028713415,
+      "learning_rate": 9.543413174496381e-06,
+      "loss": 0.6496,
+      "step": 2348
+    },
+    {
+      "epoch": 0.1631023468962644,
+      "grad_norm": 4.798289975448442,
+      "learning_rate": 9.542943600127201e-06,
+      "loss": 0.6941,
+      "step": 2349
+    },
+    {
+      "epoch": 0.16317178169698654,
+      "grad_norm": 4.462042661379931,
+      "learning_rate": 9.542473795981074e-06,
+      "loss": 0.6275,
+      "step": 2350
+    },
+    {
+      "epoch": 0.16324121649770865,
+      "grad_norm": 3.4013343660971582,
+      "learning_rate": 9.54200376208176e-06,
+      "loss": 0.4502,
+      "step": 2351
+    },
+    {
+      "epoch": 0.16331065129843078,
+      "grad_norm": 4.275023703120487,
+      "learning_rate": 9.541533498453037e-06,
+      "loss": 0.5999,
+      "step": 2352
+    },
+    {
+      "epoch": 0.1633800860991529,
+      "grad_norm": 4.728477156008202,
+      "learning_rate": 9.54106300511869e-06,
+      "loss": 0.8082,
+      "step": 2353
+    },
+    {
+      "epoch": 0.16344952089987502,
+      "grad_norm": 3.1344819277455946,
+      "learning_rate": 9.540592282102512e-06,
+      "loss": 0.2498,
+      "step": 2354
+    },
+    {
+      "epoch": 0.16351895570059713,
+      "grad_norm": 4.3756885617936625,
+      "learning_rate": 9.540121329428316e-06,
+      "loss": 0.6281,
+      "step": 2355
+    },
+    {
+      "epoch": 0.16358839050131926,
+      "grad_norm": 2.864864701435672,
+      "learning_rate": 9.53965014711992e-06,
+      "loss": 0.2832,
+      "step": 2356
+    },
+    {
+      "epoch": 0.16365782530204137,
+      "grad_norm": 3.5960456133555043,
+      "learning_rate": 9.539178735201157e-06,
+      "loss": 0.4716,
+      "step": 2357
+    },
+    {
+      "epoch": 0.1637272601027635,
+      "grad_norm": 4.0279930454859745,
+      "learning_rate": 9.53870709369587e-06,
+      "loss": 0.5604,
+      "step": 2358
+    },
+    {
+      "epoch": 0.16379669490348564,
+      "grad_norm": 3.6287455510808275,
+      "learning_rate": 9.538235222627914e-06,
+      "loss": 0.4891,
+      "step": 2359
+    },
+    {
+      "epoch": 0.16386612970420775,
+      "grad_norm": 3.950405636673243,
+      "learning_rate": 9.537763122021158e-06,
+      "loss": 0.4139,
+      "step": 2360
+    },
+    {
+      "epoch": 0.16393556450492988,
+      "grad_norm": 3.9847880187786355,
+      "learning_rate": 9.537290791899475e-06,
+      "loss": 0.3308,
+      "step": 2361
+    },
+    {
+      "epoch": 0.164004999305652,
+      "grad_norm": 3.512379434738461,
+      "learning_rate": 9.53681823228676e-06,
+      "loss": 0.4483,
+      "step": 2362
+    },
+    {
+      "epoch": 0.16407443410637412,
+      "grad_norm": 2.8985876093539957,
+      "learning_rate": 9.536345443206912e-06,
+      "loss": 0.2381,
+      "step": 2363
+    },
+    {
+      "epoch": 0.16414386890709623,
+      "grad_norm": 2.8708160775213196,
+      "learning_rate": 9.535872424683842e-06,
+      "loss": 0.2271,
+      "step": 2364
+    },
+    {
+      "epoch": 0.16421330370781836,
+      "grad_norm": 8.298354364388473,
+      "learning_rate": 9.53539917674148e-06,
+      "loss": 0.4437,
+      "step": 2365
+    },
+    {
+      "epoch": 0.16428273850854047,
+      "grad_norm": 4.536971426062279,
+      "learning_rate": 9.53492569940376e-06,
+      "loss": 0.5058,
+      "step": 2366
+    },
+    {
+      "epoch": 0.1643521733092626,
+      "grad_norm": 4.99397177022692,
+      "learning_rate": 9.53445199269463e-06,
+      "loss": 0.5501,
+      "step": 2367
+    },
+    {
+      "epoch": 0.1644216081099847,
+      "grad_norm": 4.706911104442606,
+      "learning_rate": 9.53397805663805e-06,
+      "loss": 0.4867,
+      "step": 2368
+    },
+    {
+      "epoch": 0.16449104291070685,
+      "grad_norm": 5.052019938687275,
+      "learning_rate": 9.533503891257989e-06,
+      "loss": 0.8182,
+      "step": 2369
+    },
+    {
+      "epoch": 0.16456047771142898,
+      "grad_norm": 4.918998269727205,
+      "learning_rate": 9.53302949657843e-06,
+      "loss": 0.4797,
+      "step": 2370
+    },
+    {
+      "epoch": 0.1646299125121511,
+      "grad_norm": 4.000032134556176,
+      "learning_rate": 9.53255487262337e-06,
+      "loss": 0.3573,
+      "step": 2371
+    },
+    {
+      "epoch": 0.16469934731287322,
+      "grad_norm": 3.2125303798885234,
+      "learning_rate": 9.532080019416814e-06,
+      "loss": 0.2463,
+      "step": 2372
+    },
+    {
+      "epoch": 0.16476878211359533,
+      "grad_norm": 4.740107079059526,
+      "learning_rate": 9.531604936982777e-06,
+      "loss": 0.6076,
+      "step": 2373
+    },
+    {
+      "epoch": 0.16483821691431746,
+      "grad_norm": 4.7335859724997045,
+      "learning_rate": 9.531129625345292e-06,
+      "loss": 0.6571,
+      "step": 2374
+    },
+    {
+      "epoch": 0.16490765171503957,
+      "grad_norm": 3.7862317288224836,
+      "learning_rate": 9.530654084528397e-06,
+      "loss": 0.6106,
+      "step": 2375
+    },
+    {
+      "epoch": 0.1649770865157617,
+      "grad_norm": 5.038323788202095,
+      "learning_rate": 9.530178314556146e-06,
+      "loss": 0.696,
+      "step": 2376
+    },
+    {
+      "epoch": 0.1650465213164838,
+      "grad_norm": 4.4971577390890545,
+      "learning_rate": 9.529702315452603e-06,
+      "loss": 0.6284,
+      "step": 2377
+    },
+    {
+      "epoch": 0.16511595611720595,
+      "grad_norm": 3.421754441211268,
+      "learning_rate": 9.52922608724184e-06,
+      "loss": 0.3289,
+      "step": 2378
+    },
+    {
+      "epoch": 0.16518539091792805,
+      "grad_norm": 4.311718973436578,
+      "learning_rate": 9.528749629947947e-06,
+      "loss": 0.7679,
+      "step": 2379
+    },
+    {
+      "epoch": 0.1652548257186502,
+      "grad_norm": 4.398435652130076,
+      "learning_rate": 9.528272943595023e-06,
+      "loss": 0.6875,
+      "step": 2380
+    },
+    {
+      "epoch": 0.1653242605193723,
+      "grad_norm": 4.0594377040309135,
+      "learning_rate": 9.52779602820718e-06,
+      "loss": 0.4043,
+      "step": 2381
+    },
+    {
+      "epoch": 0.16539369532009443,
+      "grad_norm": 3.9290419341809315,
+      "learning_rate": 9.527318883808535e-06,
+      "loss": 0.6643,
+      "step": 2382
+    },
+    {
+      "epoch": 0.16546313012081657,
+      "grad_norm": 4.76096238194875,
+      "learning_rate": 9.526841510423223e-06,
+      "loss": 0.7886,
+      "step": 2383
+    },
+    {
+      "epoch": 0.16553256492153867,
+      "grad_norm": 3.1263155832019445,
+      "learning_rate": 9.526363908075391e-06,
+      "loss": 0.3825,
+      "step": 2384
+    },
+    {
+      "epoch": 0.1656019997222608,
+      "grad_norm": 4.468461768655371,
+      "learning_rate": 9.525886076789195e-06,
+      "loss": 0.593,
+      "step": 2385
+    },
+    {
+      "epoch": 0.1656714345229829,
+      "grad_norm": 5.311994081990736,
+      "learning_rate": 9.525408016588802e-06,
+      "loss": 0.6052,
+      "step": 2386
+    },
+    {
+      "epoch": 0.16574086932370505,
+      "grad_norm": 3.8546551951799155,
+      "learning_rate": 9.524929727498393e-06,
+      "loss": 0.3102,
+      "step": 2387
+    },
+    {
+      "epoch": 0.16581030412442715,
+      "grad_norm": 4.552532562299914,
+      "learning_rate": 9.524451209542157e-06,
+      "loss": 0.8007,
+      "step": 2388
+    },
+    {
+      "epoch": 0.1658797389251493,
+      "grad_norm": 3.838363034448072,
+      "learning_rate": 9.5239724627443e-06,
+      "loss": 0.4641,
+      "step": 2389
+    },
+    {
+      "epoch": 0.1659491737258714,
+      "grad_norm": 3.618034399924716,
+      "learning_rate": 9.523493487129035e-06,
+      "loss": 0.5616,
+      "step": 2390
+    },
+    {
+      "epoch": 0.16601860852659353,
+      "grad_norm": 3.3708671723703425,
+      "learning_rate": 9.523014282720587e-06,
+      "loss": 0.2932,
+      "step": 2391
+    },
+    {
+      "epoch": 0.16608804332731564,
+      "grad_norm": 3.7174677994420113,
+      "learning_rate": 9.522534849543198e-06,
+      "loss": 0.492,
+      "step": 2392
+    },
+    {
+      "epoch": 0.16615747812803777,
+      "grad_norm": 4.176829289780711,
+      "learning_rate": 9.52205518762111e-06,
+      "loss": 0.5082,
+      "step": 2393
+    },
+    {
+      "epoch": 0.1662269129287599,
+      "grad_norm": 4.615197202309693,
+      "learning_rate": 9.52157529697859e-06,
+      "loss": 0.6968,
+      "step": 2394
+    },
+    {
+      "epoch": 0.166296347729482,
+      "grad_norm": 3.37278870722557,
+      "learning_rate": 9.521095177639905e-06,
+      "loss": 0.4376,
+      "step": 2395
+    },
+    {
+      "epoch": 0.16636578253020415,
+      "grad_norm": 2.8535398286940605,
+      "learning_rate": 9.520614829629344e-06,
+      "loss": 0.3118,
+      "step": 2396
+    },
+    {
+      "epoch": 0.16643521733092626,
+      "grad_norm": 4.043381898561645,
+      "learning_rate": 9.520134252971201e-06,
+      "loss": 0.4571,
+      "step": 2397
+    },
+    {
+      "epoch": 0.1665046521316484,
+      "grad_norm": 3.694026701209633,
+      "learning_rate": 9.519653447689782e-06,
+      "loss": 0.4708,
+      "step": 2398
+    },
+    {
+      "epoch": 0.1665740869323705,
+      "grad_norm": 4.10414508332498,
+      "learning_rate": 9.519172413809405e-06,
+      "loss": 0.6155,
+      "step": 2399
+    },
+    {
+      "epoch": 0.16664352173309263,
+      "grad_norm": 4.552246253317947,
+      "learning_rate": 9.518691151354403e-06,
+      "loss": 0.4564,
+      "step": 2400
+    },
+    {
+      "epoch": 0.16671295653381474,
+      "grad_norm": 4.528597680302501,
+      "learning_rate": 9.518209660349114e-06,
+      "loss": 0.4594,
+      "step": 2401
+    },
+    {
+      "epoch": 0.16678239133453687,
+      "grad_norm": 3.9139058579759443,
+      "learning_rate": 9.517727940817893e-06,
+      "loss": 0.6658,
+      "step": 2402
+    },
+    {
+      "epoch": 0.16685182613525898,
+      "grad_norm": 4.468573084623158,
+      "learning_rate": 9.517245992785107e-06,
+      "loss": 0.5455,
+      "step": 2403
+    },
+    {
+      "epoch": 0.16692126093598111,
+      "grad_norm": 3.7708704107736986,
+      "learning_rate": 9.516763816275128e-06,
+      "loss": 0.5429,
+      "step": 2404
+    },
+    {
+      "epoch": 0.16699069573670325,
+      "grad_norm": 4.131363708080557,
+      "learning_rate": 9.516281411312347e-06,
+      "loss": 0.5911,
+      "step": 2405
+    },
+    {
+      "epoch": 0.16706013053742536,
+      "grad_norm": 3.4773982007532065,
+      "learning_rate": 9.515798777921163e-06,
+      "loss": 0.3994,
+      "step": 2406
+    },
+    {
+      "epoch": 0.1671295653381475,
+      "grad_norm": 4.3617090557855,
+      "learning_rate": 9.515315916125987e-06,
+      "loss": 0.516,
+      "step": 2407
+    },
+    {
+      "epoch": 0.1671990001388696,
+      "grad_norm": 3.747226463171343,
+      "learning_rate": 9.51483282595124e-06,
+      "loss": 0.497,
+      "step": 2408
+    },
+    {
+      "epoch": 0.16726843493959173,
+      "grad_norm": 3.7462229647053786,
+      "learning_rate": 9.514349507421359e-06,
+      "loss": 0.4571,
+      "step": 2409
+    },
+    {
+      "epoch": 0.16733786974031384,
+      "grad_norm": 3.6885696168511917,
+      "learning_rate": 9.513865960560788e-06,
+      "loss": 0.5938,
+      "step": 2410
+    },
+    {
+      "epoch": 0.16740730454103597,
+      "grad_norm": 4.6336556338151045,
+      "learning_rate": 9.513382185393984e-06,
+      "loss": 0.7066,
+      "step": 2411
+    },
+    {
+      "epoch": 0.16747673934175808,
+      "grad_norm": 3.5338751812132054,
+      "learning_rate": 9.512898181945418e-06,
+      "loss": 0.3413,
+      "step": 2412
+    },
+    {
+      "epoch": 0.16754617414248021,
+      "grad_norm": 4.066274858770111,
+      "learning_rate": 9.512413950239566e-06,
+      "loss": 0.3728,
+      "step": 2413
+    },
+    {
+      "epoch": 0.16761560894320232,
+      "grad_norm": 4.5401311505861655,
+      "learning_rate": 9.511929490300924e-06,
+      "loss": 0.5686,
+      "step": 2414
+    },
+    {
+      "epoch": 0.16768504374392446,
+      "grad_norm": 4.494541541384937,
+      "learning_rate": 9.511444802153995e-06,
+      "loss": 0.8168,
+      "step": 2415
+    },
+    {
+      "epoch": 0.1677544785446466,
+      "grad_norm": 3.9445172571290996,
+      "learning_rate": 9.510959885823289e-06,
+      "loss": 0.5884,
+      "step": 2416
+    },
+    {
+      "epoch": 0.1678239133453687,
+      "grad_norm": 3.8153756099880005,
+      "learning_rate": 9.510474741333341e-06,
+      "loss": 0.3322,
+      "step": 2417
+    },
+    {
+      "epoch": 0.16789334814609083,
+      "grad_norm": 4.269161473790331,
+      "learning_rate": 9.509989368708684e-06,
+      "loss": 0.7182,
+      "step": 2418
+    },
+    {
+      "epoch": 0.16796278294681294,
+      "grad_norm": 3.3826141088984127,
+      "learning_rate": 9.509503767973866e-06,
+      "loss": 0.3696,
+      "step": 2419
+    },
+    {
+      "epoch": 0.16803221774753507,
+      "grad_norm": 4.15557468886651,
+      "learning_rate": 9.509017939153452e-06,
+      "loss": 0.3244,
+      "step": 2420
+    },
+    {
+      "epoch": 0.16810165254825718,
+      "grad_norm": 3.6718491976743928,
+      "learning_rate": 9.50853188227201e-06,
+      "loss": 0.5639,
+      "step": 2421
+    },
+    {
+      "epoch": 0.16817108734897931,
+      "grad_norm": 4.6583416673393065,
+      "learning_rate": 9.50804559735413e-06,
+      "loss": 0.6644,
+      "step": 2422
+    },
+    {
+      "epoch": 0.16824052214970142,
+      "grad_norm": 3.79596881146758,
+      "learning_rate": 9.507559084424404e-06,
+      "loss": 0.5245,
+      "step": 2423
+    },
+    {
+      "epoch": 0.16830995695042356,
+      "grad_norm": 4.156108840786895,
+      "learning_rate": 9.50707234350744e-06,
+      "loss": 0.3648,
+      "step": 2424
+    },
+    {
+      "epoch": 0.16837939175114566,
+      "grad_norm": 3.689455522010027,
+      "learning_rate": 9.506585374627857e-06,
+      "loss": 0.3116,
+      "step": 2425
+    },
+    {
+      "epoch": 0.1684488265518678,
+      "grad_norm": 3.755881942223915,
+      "learning_rate": 9.506098177810286e-06,
+      "loss": 0.3292,
+      "step": 2426
+    },
+    {
+      "epoch": 0.1685182613525899,
+      "grad_norm": 2.8024359395239467,
+      "learning_rate": 9.505610753079367e-06,
+      "loss": 0.2049,
+      "step": 2427
+    },
+    {
+      "epoch": 0.16858769615331204,
+      "grad_norm": 4.49714945130422,
+      "learning_rate": 9.505123100459755e-06,
+      "loss": 0.7371,
+      "step": 2428
+    },
+    {
+      "epoch": 0.16865713095403417,
+      "grad_norm": 5.312037461605566,
+      "learning_rate": 9.504635219976114e-06,
+      "loss": 0.7069,
+      "step": 2429
+    },
+    {
+      "epoch": 0.16872656575475628,
+      "grad_norm": 4.205291103104701,
+      "learning_rate": 9.504147111653123e-06,
+      "loss": 0.6004,
+      "step": 2430
+    },
+    {
+      "epoch": 0.16879600055547841,
+      "grad_norm": 5.245704109499341,
+      "learning_rate": 9.503658775515466e-06,
+      "loss": 0.6133,
+      "step": 2431
+    },
+    {
+      "epoch": 0.16886543535620052,
+      "grad_norm": 4.978854456845057,
+      "learning_rate": 9.503170211587845e-06,
+      "loss": 0.5544,
+      "step": 2432
+    },
+    {
+      "epoch": 0.16893487015692266,
+      "grad_norm": 3.6776073538204703,
+      "learning_rate": 9.50268141989497e-06,
+      "loss": 0.3496,
+      "step": 2433
+    },
+    {
+      "epoch": 0.16900430495764476,
+      "grad_norm": 3.0046077463745084,
+      "learning_rate": 9.502192400461565e-06,
+      "loss": 0.2767,
+      "step": 2434
+    },
+    {
+      "epoch": 0.1690737397583669,
+      "grad_norm": 4.282518730750572,
+      "learning_rate": 9.501703153312361e-06,
+      "loss": 0.5636,
+      "step": 2435
+    },
+    {
+      "epoch": 0.169143174559089,
+      "grad_norm": 4.661711579009234,
+      "learning_rate": 9.501213678472107e-06,
+      "loss": 0.5752,
+      "step": 2436
+    },
+    {
+      "epoch": 0.16921260935981114,
+      "grad_norm": 4.7422146279079795,
+      "learning_rate": 9.500723975965558e-06,
+      "loss": 0.6866,
+      "step": 2437
+    },
+    {
+      "epoch": 0.16928204416053325,
+      "grad_norm": 4.3380421612000974,
+      "learning_rate": 9.500234045817485e-06,
+      "loss": 0.5878,
+      "step": 2438
+    },
+    {
+      "epoch": 0.16935147896125538,
+      "grad_norm": 4.60943920474025,
+      "learning_rate": 9.499743888052663e-06,
+      "loss": 0.6599,
+      "step": 2439
+    },
+    {
+      "epoch": 0.16942091376197752,
+      "grad_norm": 3.867257208741439,
+      "learning_rate": 9.49925350269589e-06,
+      "loss": 0.6893,
+      "step": 2440
+    },
+    {
+      "epoch": 0.16949034856269962,
+      "grad_norm": 2.4610368979572685,
+      "learning_rate": 9.498762889771964e-06,
+      "loss": 0.1735,
+      "step": 2441
+    },
+    {
+      "epoch": 0.16955978336342176,
+      "grad_norm": 4.408139180335056,
+      "learning_rate": 9.498272049305703e-06,
+      "loss": 0.3538,
+      "step": 2442
+    },
+    {
+      "epoch": 0.16962921816414386,
+      "grad_norm": 3.799621706654066,
+      "learning_rate": 9.497780981321931e-06,
+      "loss": 0.4171,
+      "step": 2443
+    },
+    {
+      "epoch": 0.169698652964866,
+      "grad_norm": 4.464408553967309,
+      "learning_rate": 9.497289685845487e-06,
+      "loss": 0.5159,
+      "step": 2444
+    },
+    {
+      "epoch": 0.1697680877655881,
+      "grad_norm": 4.876331346663356,
+      "learning_rate": 9.49679816290122e-06,
+      "loss": 0.4691,
+      "step": 2445
+    },
+    {
+      "epoch": 0.16983752256631024,
+      "grad_norm": 3.189575232540383,
+      "learning_rate": 9.496306412513989e-06,
+      "loss": 0.3184,
+      "step": 2446
+    },
+    {
+      "epoch": 0.16990695736703235,
+      "grad_norm": 4.890500059741749,
+      "learning_rate": 9.49581443470867e-06,
+      "loss": 0.4038,
+      "step": 2447
+    },
+    {
+      "epoch": 0.16997639216775448,
+      "grad_norm": 3.7978752822500517,
+      "learning_rate": 9.495322229510142e-06,
+      "loss": 0.4084,
+      "step": 2448
+    },
+    {
+      "epoch": 0.1700458269684766,
+      "grad_norm": 4.3587212771452375,
+      "learning_rate": 9.494829796943304e-06,
+      "loss": 0.6418,
+      "step": 2449
+    },
+    {
+      "epoch": 0.17011526176919872,
+      "grad_norm": 5.007592297039119,
+      "learning_rate": 9.49433713703306e-06,
+      "loss": 0.7855,
+      "step": 2450
+    },
+    {
+      "epoch": 0.17018469656992086,
+      "grad_norm": 4.04901020610556,
+      "learning_rate": 9.493844249804328e-06,
+      "loss": 0.5663,
+      "step": 2451
+    },
+    {
+      "epoch": 0.17025413137064296,
+      "grad_norm": 4.310272379992436,
+      "learning_rate": 9.493351135282043e-06,
+      "loss": 0.5723,
+      "step": 2452
+    },
+    {
+      "epoch": 0.1703235661713651,
+      "grad_norm": 3.8364400491606734,
+      "learning_rate": 9.492857793491138e-06,
+      "loss": 0.2925,
+      "step": 2453
+    },
+    {
+      "epoch": 0.1703930009720872,
+      "grad_norm": 3.63145789351474,
+      "learning_rate": 9.492364224456574e-06,
+      "loss": 0.3312,
+      "step": 2454
+    },
+    {
+      "epoch": 0.17046243577280934,
+      "grad_norm": 4.77388187844432,
+      "learning_rate": 9.491870428203306e-06,
+      "loss": 0.7356,
+      "step": 2455
+    },
+    {
+      "epoch": 0.17053187057353145,
+      "grad_norm": 3.277398461369125,
+      "learning_rate": 9.491376404756318e-06,
+      "loss": 0.2545,
+      "step": 2456
+    },
+    {
+      "epoch": 0.17060130537425358,
+      "grad_norm": 3.2382704299524687,
+      "learning_rate": 9.490882154140594e-06,
+      "loss": 0.4156,
+      "step": 2457
+    },
+    {
+      "epoch": 0.1706707401749757,
+      "grad_norm": 3.9234597465667274,
+      "learning_rate": 9.490387676381132e-06,
+      "loss": 0.4654,
+      "step": 2458
+    },
+    {
+      "epoch": 0.17074017497569782,
+      "grad_norm": 3.8773634209709313,
+      "learning_rate": 9.489892971502941e-06,
+      "loss": 0.3416,
+      "step": 2459
+    },
+    {
+      "epoch": 0.17080960977641993,
+      "grad_norm": 4.126082767822525,
+      "learning_rate": 9.489398039531045e-06,
+      "loss": 0.6697,
+      "step": 2460
+    },
+    {
+      "epoch": 0.17087904457714206,
+      "grad_norm": 4.509680058161002,
+      "learning_rate": 9.488902880490476e-06,
+      "loss": 0.3857,
+      "step": 2461
+    },
+    {
+      "epoch": 0.1709484793778642,
+      "grad_norm": 3.2936954499385185,
+      "learning_rate": 9.48840749440628e-06,
+      "loss": 0.4885,
+      "step": 2462
+    },
+    {
+      "epoch": 0.1710179141785863,
+      "grad_norm": 4.446826991135186,
+      "learning_rate": 9.487911881303512e-06,
+      "loss": 0.4549,
+      "step": 2463
+    },
+    {
+      "epoch": 0.17108734897930844,
+      "grad_norm": 5.616094797782135,
+      "learning_rate": 9.487416041207239e-06,
+      "loss": 0.679,
+      "step": 2464
+    },
+    {
+      "epoch": 0.17115678378003055,
+      "grad_norm": 5.334464652359958,
+      "learning_rate": 9.486919974142539e-06,
+      "loss": 0.8263,
+      "step": 2465
+    },
+    {
+      "epoch": 0.17122621858075268,
+      "grad_norm": 4.012462243737707,
+      "learning_rate": 9.486423680134506e-06,
+      "loss": 0.5833,
+      "step": 2466
+    },
+    {
+      "epoch": 0.1712956533814748,
+      "grad_norm": 4.2629563988376376,
+      "learning_rate": 9.485927159208239e-06,
+      "loss": 0.5212,
+      "step": 2467
+    },
+    {
+      "epoch": 0.17136508818219692,
+      "grad_norm": 4.307555934858232,
+      "learning_rate": 9.485430411388852e-06,
+      "loss": 0.712,
+      "step": 2468
+    },
+    {
+      "epoch": 0.17143452298291903,
+      "grad_norm": 4.572750886075486,
+      "learning_rate": 9.484933436701471e-06,
+      "loss": 0.7104,
+      "step": 2469
+    },
+    {
+      "epoch": 0.17150395778364116,
+      "grad_norm": 4.613429790628599,
+      "learning_rate": 9.484436235171232e-06,
+      "loss": 0.5473,
+      "step": 2470
+    },
+    {
+      "epoch": 0.17157339258436327,
+      "grad_norm": 4.332762239260598,
+      "learning_rate": 9.483938806823283e-06,
+      "loss": 0.3874,
+      "step": 2471
+    },
+    {
+      "epoch": 0.1716428273850854,
+      "grad_norm": 4.965580954508247,
+      "learning_rate": 9.483441151682781e-06,
+      "loss": 0.6581,
+      "step": 2472
+    },
+    {
+      "epoch": 0.17171226218580754,
+      "grad_norm": 4.37400242255445,
+      "learning_rate": 9.482943269774903e-06,
+      "loss": 0.5639,
+      "step": 2473
+    },
+    {
+      "epoch": 0.17178169698652965,
+      "grad_norm": 4.329458222806838,
+      "learning_rate": 9.482445161124824e-06,
+      "loss": 0.774,
+      "step": 2474
+    },
+    {
+      "epoch": 0.17185113178725178,
+      "grad_norm": 4.182391102070757,
+      "learning_rate": 9.481946825757742e-06,
+      "loss": 0.4937,
+      "step": 2475
+    },
+    {
+      "epoch": 0.1719205665879739,
+      "grad_norm": 4.558755052925949,
+      "learning_rate": 9.48144826369886e-06,
+      "loss": 0.6855,
+      "step": 2476
+    },
+    {
+      "epoch": 0.17199000138869602,
+      "grad_norm": 4.116640408597874,
+      "learning_rate": 9.480949474973398e-06,
+      "loss": 0.5586,
+      "step": 2477
+    },
+    {
+      "epoch": 0.17205943618941813,
+      "grad_norm": 6.134955614054857,
+      "learning_rate": 9.48045045960658e-06,
+      "loss": 0.6961,
+      "step": 2478
+    },
+    {
+      "epoch": 0.17212887099014026,
+      "grad_norm": 5.5211758815305485,
+      "learning_rate": 9.479951217623649e-06,
+      "loss": 0.9109,
+      "step": 2479
+    },
+    {
+      "epoch": 0.17219830579086237,
+      "grad_norm": 4.403960580482242,
+      "learning_rate": 9.479451749049855e-06,
+      "loss": 0.626,
+      "step": 2480
+    },
+    {
+      "epoch": 0.1722677405915845,
+      "grad_norm": 4.3508048321288495,
+      "learning_rate": 9.47895205391046e-06,
+      "loss": 0.5596,
+      "step": 2481
+    },
+    {
+      "epoch": 0.1723371753923066,
+      "grad_norm": 4.237715170632942,
+      "learning_rate": 9.478452132230738e-06,
+      "loss": 0.3406,
+      "step": 2482
+    },
+    {
+      "epoch": 0.17240661019302875,
+      "grad_norm": 3.7245643613229786,
+      "learning_rate": 9.477951984035975e-06,
+      "loss": 0.4676,
+      "step": 2483
+    },
+    {
+      "epoch": 0.17247604499375085,
+      "grad_norm": 4.791425834786883,
+      "learning_rate": 9.477451609351467e-06,
+      "loss": 0.8774,
+      "step": 2484
+    },
+    {
+      "epoch": 0.172545479794473,
+      "grad_norm": 3.7346370031303193,
+      "learning_rate": 9.476951008202523e-06,
+      "loss": 0.4457,
+      "step": 2485
+    },
+    {
+      "epoch": 0.17261491459519512,
+      "grad_norm": 4.216735665638584,
+      "learning_rate": 9.476450180614465e-06,
+      "loss": 0.5258,
+      "step": 2486
+    },
+    {
+      "epoch": 0.17268434939591723,
+      "grad_norm": 4.705078524019482,
+      "learning_rate": 9.475949126612622e-06,
+      "loss": 0.6157,
+      "step": 2487
+    },
+    {
+      "epoch": 0.17275378419663937,
+      "grad_norm": 3.126093149768665,
+      "learning_rate": 9.475447846222336e-06,
+      "loss": 0.3631,
+      "step": 2488
+    },
+    {
+      "epoch": 0.17282321899736147,
+      "grad_norm": 4.768709168842983,
+      "learning_rate": 9.474946339468963e-06,
+      "loss": 0.2688,
+      "step": 2489
+    },
+    {
+      "epoch": 0.1728926537980836,
+      "grad_norm": 4.248480545995738,
+      "learning_rate": 9.474444606377869e-06,
+      "loss": 0.4894,
+      "step": 2490
+    },
+    {
+      "epoch": 0.1729620885988057,
+      "grad_norm": 3.7852532303774353,
+      "learning_rate": 9.473942646974428e-06,
+      "loss": 0.5209,
+      "step": 2491
+    },
+    {
+      "epoch": 0.17303152339952785,
+      "grad_norm": 3.274787675820715,
+      "learning_rate": 9.473440461284032e-06,
+      "loss": 0.302,
+      "step": 2492
+    },
+    {
+      "epoch": 0.17310095820024995,
+      "grad_norm": 4.10101652755142,
+      "learning_rate": 9.472938049332077e-06,
+      "loss": 0.5499,
+      "step": 2493
+    },
+    {
+      "epoch": 0.1731703930009721,
+      "grad_norm": 5.698311623012093,
+      "learning_rate": 9.472435411143979e-06,
+      "loss": 0.7219,
+      "step": 2494
+    },
+    {
+      "epoch": 0.1732398278016942,
+      "grad_norm": 3.022957291163817,
+      "learning_rate": 9.471932546745157e-06,
+      "loss": 0.3453,
+      "step": 2495
+    },
+    {
+      "epoch": 0.17330926260241633,
+      "grad_norm": 5.570720864179199,
+      "learning_rate": 9.47142945616105e-06,
+      "loss": 0.5649,
+      "step": 2496
+    },
+    {
+      "epoch": 0.17337869740313847,
+      "grad_norm": 3.690666704461808,
+      "learning_rate": 9.470926139417099e-06,
+      "loss": 0.5439,
+      "step": 2497
+    },
+    {
+      "epoch": 0.17344813220386057,
+      "grad_norm": 4.8021056071034165,
+      "learning_rate": 9.470422596538762e-06,
+      "loss": 0.5627,
+      "step": 2498
+    },
+    {
+      "epoch": 0.1735175670045827,
+      "grad_norm": 3.799113231612468,
+      "learning_rate": 9.469918827551511e-06,
+      "loss": 0.4677,
+      "step": 2499
+    },
+    {
+      "epoch": 0.1735870018053048,
+      "grad_norm": 4.167959430426168,
+      "learning_rate": 9.469414832480824e-06,
+      "loss": 0.7912,
+      "step": 2500
+    },
+    {
+      "epoch": 0.17365643660602695,
+      "grad_norm": 3.539133572913732,
+      "learning_rate": 9.468910611352192e-06,
+      "loss": 0.4062,
+      "step": 2501
+    },
+    {
+      "epoch": 0.17372587140674906,
+      "grad_norm": 3.621334290097632,
+      "learning_rate": 9.468406164191115e-06,
+      "loss": 0.5838,
+      "step": 2502
+    },
+    {
+      "epoch": 0.1737953062074712,
+      "grad_norm": 4.572593454779693,
+      "learning_rate": 9.467901491023113e-06,
+      "loss": 0.7584,
+      "step": 2503
+    },
+    {
+      "epoch": 0.1738647410081933,
+      "grad_norm": 5.2394872572667275,
+      "learning_rate": 9.46739659187371e-06,
+      "loss": 0.7588,
+      "step": 2504
+    },
+    {
+      "epoch": 0.17393417580891543,
+      "grad_norm": 3.9627653974160157,
+      "learning_rate": 9.466891466768442e-06,
+      "loss": 0.5346,
+      "step": 2505
+    },
+    {
+      "epoch": 0.17400361060963754,
+      "grad_norm": 4.825983525269473,
+      "learning_rate": 9.466386115732857e-06,
+      "loss": 0.7055,
+      "step": 2506
+    },
+    {
+      "epoch": 0.17407304541035967,
+      "grad_norm": 3.7413348095963697,
+      "learning_rate": 9.465880538792519e-06,
+      "loss": 0.3747,
+      "step": 2507
+    },
+    {
+      "epoch": 0.1741424802110818,
+      "grad_norm": 3.3402827620686835,
+      "learning_rate": 9.465374735972994e-06,
+      "loss": 0.5207,
+      "step": 2508
+    },
+    {
+      "epoch": 0.17421191501180391,
+      "grad_norm": 3.8828334658948593,
+      "learning_rate": 9.46486870729987e-06,
+      "loss": 0.4463,
+      "step": 2509
+    },
+    {
+      "epoch": 0.17428134981252605,
+      "grad_norm": 4.1078424063160375,
+      "learning_rate": 9.464362452798738e-06,
+      "loss": 0.5495,
+      "step": 2510
+    },
+    {
+      "epoch": 0.17435078461324816,
+      "grad_norm": 4.4492975938808375,
+      "learning_rate": 9.463855972495204e-06,
+      "loss": 0.5307,
+      "step": 2511
+    },
+    {
+      "epoch": 0.1744202194139703,
+      "grad_norm": 3.317158215582718,
+      "learning_rate": 9.463349266414888e-06,
+      "loss": 0.2606,
+      "step": 2512
+    },
+    {
+      "epoch": 0.1744896542146924,
+      "grad_norm": 4.140430321705507,
+      "learning_rate": 9.462842334583415e-06,
+      "loss": 0.2702,
+      "step": 2513
+    },
+    {
+      "epoch": 0.17455908901541453,
+      "grad_norm": 3.553630993493244,
+      "learning_rate": 9.462335177026428e-06,
+      "loss": 0.4736,
+      "step": 2514
+    },
+    {
+      "epoch": 0.17462852381613664,
+      "grad_norm": 4.754250410093507,
+      "learning_rate": 9.461827793769576e-06,
+      "loss": 0.4211,
+      "step": 2515
+    },
+    {
+      "epoch": 0.17469795861685877,
+      "grad_norm": 3.5810850361491946,
+      "learning_rate": 9.461320184838525e-06,
+      "loss": 0.3673,
+      "step": 2516
+    },
+    {
+      "epoch": 0.17476739341758088,
+      "grad_norm": 3.1074727212077518,
+      "learning_rate": 9.460812350258946e-06,
+      "loss": 0.2814,
+      "step": 2517
+    },
+    {
+      "epoch": 0.17483682821830301,
+      "grad_norm": 5.625353929190155,
+      "learning_rate": 9.460304290056527e-06,
+      "loss": 0.5769,
+      "step": 2518
+    },
+    {
+      "epoch": 0.17490626301902515,
+      "grad_norm": 4.39464902168474,
+      "learning_rate": 9.459796004256963e-06,
+      "loss": 0.6023,
+      "step": 2519
+    },
+    {
+      "epoch": 0.17497569781974726,
+      "grad_norm": 3.8670630639759933,
+      "learning_rate": 9.459287492885965e-06,
+      "loss": 0.395,
+      "step": 2520
+    },
+    {
+      "epoch": 0.1750451326204694,
+      "grad_norm": 4.975056511938133,
+      "learning_rate": 9.458778755969251e-06,
+      "loss": 0.5925,
+      "step": 2521
+    },
+    {
+      "epoch": 0.1751145674211915,
+      "grad_norm": 3.6703256448193344,
+      "learning_rate": 9.458269793532553e-06,
+      "loss": 0.3675,
+      "step": 2522
+    },
+    {
+      "epoch": 0.17518400222191363,
+      "grad_norm": 4.046397273775209,
+      "learning_rate": 9.457760605601613e-06,
+      "loss": 0.583,
+      "step": 2523
+    },
+    {
+      "epoch": 0.17525343702263574,
+      "grad_norm": 4.414716799176455,
+      "learning_rate": 9.457251192202185e-06,
+      "loss": 0.5963,
+      "step": 2524
+    },
+    {
+      "epoch": 0.17532287182335787,
+      "grad_norm": 4.914761629735875,
+      "learning_rate": 9.456741553360039e-06,
+      "loss": 0.6278,
+      "step": 2525
+    },
+    {
+      "epoch": 0.17539230662407998,
+      "grad_norm": 3.9052324029947716,
+      "learning_rate": 9.456231689100945e-06,
+      "loss": 0.5533,
+      "step": 2526
+    },
+    {
+      "epoch": 0.17546174142480211,
+      "grad_norm": 5.117364200082593,
+      "learning_rate": 9.455721599450696e-06,
+      "loss": 0.7875,
+      "step": 2527
+    },
+    {
+      "epoch": 0.17553117622552422,
+      "grad_norm": 5.289025429004056,
+      "learning_rate": 9.45521128443509e-06,
+      "loss": 0.6239,
+      "step": 2528
+    },
+    {
+      "epoch": 0.17560061102624636,
+      "grad_norm": 5.204444854119443,
+      "learning_rate": 9.454700744079941e-06,
+      "loss": 0.5305,
+      "step": 2529
+    },
+    {
+      "epoch": 0.17567004582696846,
+      "grad_norm": 2.6537194441694814,
+      "learning_rate": 9.454189978411068e-06,
+      "loss": 0.2434,
+      "step": 2530
+    },
+    {
+      "epoch": 0.1757394806276906,
+      "grad_norm": 4.3928848938700975,
+      "learning_rate": 9.453678987454305e-06,
+      "loss": 0.7166,
+      "step": 2531
+    },
+    {
+      "epoch": 0.17580891542841273,
+      "grad_norm": 3.1494994019101097,
+      "learning_rate": 9.453167771235499e-06,
+      "loss": 0.3704,
+      "step": 2532
+    },
+    {
+      "epoch": 0.17587835022913484,
+      "grad_norm": 4.165953584897143,
+      "learning_rate": 9.452656329780506e-06,
+      "loss": 0.3486,
+      "step": 2533
+    },
+    {
+      "epoch": 0.17594778502985697,
+      "grad_norm": 3.0588125029061946,
+      "learning_rate": 9.452144663115195e-06,
+      "loss": 0.3192,
+      "step": 2534
+    },
+    {
+      "epoch": 0.17601721983057908,
+      "grad_norm": 4.829807215733037,
+      "learning_rate": 9.451632771265444e-06,
+      "loss": 0.6823,
+      "step": 2535
+    },
+    {
+      "epoch": 0.17608665463130121,
+      "grad_norm": 5.503397799908635,
+      "learning_rate": 9.451120654257144e-06,
+      "loss": 0.8862,
+      "step": 2536
+    },
+    {
+      "epoch": 0.17615608943202332,
+      "grad_norm": 4.419571922372808,
+      "learning_rate": 9.4506083121162e-06,
+      "loss": 0.3456,
+      "step": 2537
+    },
+    {
+      "epoch": 0.17622552423274546,
+      "grad_norm": 3.435600822609814,
+      "learning_rate": 9.450095744868523e-06,
+      "loss": 0.342,
+      "step": 2538
+    },
+    {
+      "epoch": 0.17629495903346756,
+      "grad_norm": 3.9251691797279262,
+      "learning_rate": 9.44958295254004e-06,
+      "loss": 0.4531,
+      "step": 2539
+    },
+    {
+      "epoch": 0.1763643938341897,
+      "grad_norm": 3.5249949602663575,
+      "learning_rate": 9.449069935156684e-06,
+      "loss": 0.2855,
+      "step": 2540
+    },
+    {
+      "epoch": 0.1764338286349118,
+      "grad_norm": 3.8553430441706396,
+      "learning_rate": 9.448556692744406e-06,
+      "loss": 0.472,
+      "step": 2541
+    },
+    {
+      "epoch": 0.17650326343563394,
+      "grad_norm": 4.695961105175266,
+      "learning_rate": 9.448043225329166e-06,
+      "loss": 0.7451,
+      "step": 2542
+    },
+    {
+      "epoch": 0.17657269823635607,
+      "grad_norm": 3.1447620209924754,
+      "learning_rate": 9.447529532936932e-06,
+      "loss": 0.2998,
+      "step": 2543
+    },
+    {
+      "epoch": 0.17664213303707818,
+      "grad_norm": 5.034252418815933,
+      "learning_rate": 9.447015615593687e-06,
+      "loss": 0.5559,
+      "step": 2544
+    },
+    {
+      "epoch": 0.17671156783780032,
+      "grad_norm": 4.652850602310597,
+      "learning_rate": 9.446501473325422e-06,
+      "loss": 0.6781,
+      "step": 2545
+    },
+    {
+      "epoch": 0.17678100263852242,
+      "grad_norm": 3.791145253430967,
+      "learning_rate": 9.445987106158148e-06,
+      "loss": 0.532,
+      "step": 2546
+    },
+    {
+      "epoch": 0.17685043743924456,
+      "grad_norm": 3.645616754843282,
+      "learning_rate": 9.445472514117874e-06,
+      "loss": 0.5799,
+      "step": 2547
+    },
+    {
+      "epoch": 0.17691987223996666,
+      "grad_norm": 4.469863345440935,
+      "learning_rate": 9.444957697230633e-06,
+      "loss": 0.4935,
+      "step": 2548
+    },
+    {
+      "epoch": 0.1769893070406888,
+      "grad_norm": 3.6638120238355993,
+      "learning_rate": 9.44444265552246e-06,
+      "loss": 0.3045,
+      "step": 2549
+    },
+    {
+      "epoch": 0.1770587418414109,
+      "grad_norm": 6.4355196428719506,
+      "learning_rate": 9.443927389019407e-06,
+      "loss": 0.5495,
+      "step": 2550
+    },
+    {
+      "epoch": 0.17712817664213304,
+      "grad_norm": 4.828843155329521,
+      "learning_rate": 9.443411897747534e-06,
+      "loss": 0.7995,
+      "step": 2551
+    },
+    {
+      "epoch": 0.17719761144285515,
+      "grad_norm": 3.3413988283878817,
+      "learning_rate": 9.442896181732918e-06,
+      "loss": 0.2981,
+      "step": 2552
+    },
+    {
+      "epoch": 0.17726704624357728,
+      "grad_norm": 4.828586393264978,
+      "learning_rate": 9.44238024100164e-06,
+      "loss": 0.5184,
+      "step": 2553
+    },
+    {
+      "epoch": 0.17733648104429942,
+      "grad_norm": 4.374276013777648,
+      "learning_rate": 9.441864075579794e-06,
+      "loss": 0.6367,
+      "step": 2554
+    },
+    {
+      "epoch": 0.17740591584502152,
+      "grad_norm": 4.602794787741462,
+      "learning_rate": 9.44134768549349e-06,
+      "loss": 0.6283,
+      "step": 2555
+    },
+    {
+      "epoch": 0.17747535064574366,
+      "grad_norm": 3.825073669974407,
+      "learning_rate": 9.440831070768848e-06,
+      "loss": 0.518,
+      "step": 2556
+    },
+    {
+      "epoch": 0.17754478544646576,
+      "grad_norm": 3.512590050106786,
+      "learning_rate": 9.440314231431993e-06,
+      "loss": 0.3142,
+      "step": 2557
+    },
+    {
+      "epoch": 0.1776142202471879,
+      "grad_norm": 4.9653758107614525,
+      "learning_rate": 9.43979716750907e-06,
+      "loss": 0.4623,
+      "step": 2558
+    },
+    {
+      "epoch": 0.17768365504791,
+      "grad_norm": 4.221497483096389,
+      "learning_rate": 9.439279879026231e-06,
+      "loss": 0.5783,
+      "step": 2559
+    },
+    {
+      "epoch": 0.17775308984863214,
+      "grad_norm": 3.580331688334701,
+      "learning_rate": 9.438762366009638e-06,
+      "loss": 0.2322,
+      "step": 2560
+    },
+    {
+      "epoch": 0.17782252464935425,
+      "grad_norm": 6.075770423491884,
+      "learning_rate": 9.438244628485466e-06,
+      "loss": 0.7625,
+      "step": 2561
+    },
+    {
+      "epoch": 0.17789195945007638,
+      "grad_norm": 5.8891010305906715,
+      "learning_rate": 9.437726666479906e-06,
+      "loss": 0.8774,
+      "step": 2562
+    },
+    {
+      "epoch": 0.1779613942507985,
+      "grad_norm": 4.437590072464308,
+      "learning_rate": 9.437208480019152e-06,
+      "loss": 0.3963,
+      "step": 2563
+    },
+    {
+      "epoch": 0.17803082905152062,
+      "grad_norm": 4.195769191252464,
+      "learning_rate": 9.436690069129413e-06,
+      "loss": 0.3887,
+      "step": 2564
+    },
+    {
+      "epoch": 0.17810026385224276,
+      "grad_norm": 4.623620089034488,
+      "learning_rate": 9.436171433836911e-06,
+      "loss": 0.5898,
+      "step": 2565
+    },
+    {
+      "epoch": 0.17816969865296486,
+      "grad_norm": 3.1361681145364133,
+      "learning_rate": 9.435652574167877e-06,
+      "loss": 0.2388,
+      "step": 2566
+    },
+    {
+      "epoch": 0.178239133453687,
+      "grad_norm": 3.945517694619383,
+      "learning_rate": 9.435133490148557e-06,
+      "loss": 0.4957,
+      "step": 2567
+    },
+    {
+      "epoch": 0.1783085682544091,
+      "grad_norm": 4.789306683601022,
+      "learning_rate": 9.434614181805203e-06,
+      "loss": 0.5541,
+      "step": 2568
+    },
+    {
+      "epoch": 0.17837800305513124,
+      "grad_norm": 4.211993493003448,
+      "learning_rate": 9.434094649164081e-06,
+      "loss": 0.5206,
+      "step": 2569
+    },
+    {
+      "epoch": 0.17844743785585335,
+      "grad_norm": 3.895569156452278,
+      "learning_rate": 9.43357489225147e-06,
+      "loss": 0.5212,
+      "step": 2570
+    },
+    {
+      "epoch": 0.17851687265657548,
+      "grad_norm": 3.8098220015554523,
+      "learning_rate": 9.433054911093658e-06,
+      "loss": 0.28,
+      "step": 2571
+    },
+    {
+      "epoch": 0.1785863074572976,
+      "grad_norm": 3.094355393548992,
+      "learning_rate": 9.432534705716945e-06,
+      "loss": 0.2913,
+      "step": 2572
+    },
+    {
+      "epoch": 0.17865574225801972,
+      "grad_norm": 4.075841825753508,
+      "learning_rate": 9.43201427614764e-06,
+      "loss": 0.4941,
+      "step": 2573
+    },
+    {
+      "epoch": 0.17872517705874183,
+      "grad_norm": 4.06290239638234,
+      "learning_rate": 9.431493622412072e-06,
+      "loss": 0.4982,
+      "step": 2574
+    },
+    {
+      "epoch": 0.17879461185946396,
+      "grad_norm": 3.1330324957417583,
+      "learning_rate": 9.430972744536568e-06,
+      "loss": 0.4332,
+      "step": 2575
+    },
+    {
+      "epoch": 0.1788640466601861,
+      "grad_norm": 4.604136494096769,
+      "learning_rate": 9.430451642547478e-06,
+      "loss": 0.5489,
+      "step": 2576
+    },
+    {
+      "epoch": 0.1789334814609082,
+      "grad_norm": 4.149784820732293,
+      "learning_rate": 9.429930316471158e-06,
+      "loss": 0.6032,
+      "step": 2577
+    },
+    {
+      "epoch": 0.17900291626163034,
+      "grad_norm": 3.383524892839391,
+      "learning_rate": 9.429408766333974e-06,
+      "loss": 0.5634,
+      "step": 2578
+    },
+    {
+      "epoch": 0.17907235106235245,
+      "grad_norm": 4.9137357438599025,
+      "learning_rate": 9.428886992162307e-06,
+      "loss": 0.7109,
+      "step": 2579
+    },
+    {
+      "epoch": 0.17914178586307458,
+      "grad_norm": 5.114699088246984,
+      "learning_rate": 9.428364993982548e-06,
+      "loss": 1.0056,
+      "step": 2580
+    },
+    {
+      "epoch": 0.1792112206637967,
+      "grad_norm": 3.678788204054091,
+      "learning_rate": 9.427842771821099e-06,
+      "loss": 0.3424,
+      "step": 2581
+    },
+    {
+      "epoch": 0.17928065546451882,
+      "grad_norm": 4.201820132548689,
+      "learning_rate": 9.427320325704372e-06,
+      "loss": 0.8085,
+      "step": 2582
+    },
+    {
+      "epoch": 0.17935009026524093,
+      "grad_norm": 2.4798176732912096,
+      "learning_rate": 9.426797655658793e-06,
+      "loss": 0.1866,
+      "step": 2583
+    },
+    {
+      "epoch": 0.17941952506596306,
+      "grad_norm": 4.660639319594791,
+      "learning_rate": 9.426274761710799e-06,
+      "loss": 0.8595,
+      "step": 2584
+    },
+    {
+      "epoch": 0.17948895986668517,
+      "grad_norm": 4.765098857506229,
+      "learning_rate": 9.425751643886835e-06,
+      "loss": 0.4,
+      "step": 2585
+    },
+    {
+      "epoch": 0.1795583946674073,
+      "grad_norm": 3.771592890418262,
+      "learning_rate": 9.42522830221336e-06,
+      "loss": 0.2954,
+      "step": 2586
+    },
+    {
+      "epoch": 0.1796278294681294,
+      "grad_norm": 3.8275180226428698,
+      "learning_rate": 9.424704736716847e-06,
+      "loss": 0.6602,
+      "step": 2587
+    },
+    {
+      "epoch": 0.17969726426885155,
+      "grad_norm": 4.30850383427892,
+      "learning_rate": 9.424180947423774e-06,
+      "loss": 0.6607,
+      "step": 2588
+    },
+    {
+      "epoch": 0.17976669906957368,
+      "grad_norm": 4.79549595289997,
+      "learning_rate": 9.423656934360637e-06,
+      "loss": 0.672,
+      "step": 2589
+    },
+    {
+      "epoch": 0.1798361338702958,
+      "grad_norm": 3.3059563850323648,
+      "learning_rate": 9.423132697553937e-06,
+      "loss": 0.3413,
+      "step": 2590
+    },
+    {
+      "epoch": 0.17990556867101792,
+      "grad_norm": 4.016048624923509,
+      "learning_rate": 9.422608237030189e-06,
+      "loss": 0.4841,
+      "step": 2591
+    },
+    {
+      "epoch": 0.17997500347174003,
+      "grad_norm": 3.5180965141476728,
+      "learning_rate": 9.422083552815922e-06,
+      "loss": 0.3386,
+      "step": 2592
+    },
+    {
+      "epoch": 0.18004443827246217,
+      "grad_norm": 4.25964721047945,
+      "learning_rate": 9.421558644937672e-06,
+      "loss": 0.3766,
+      "step": 2593
+    },
+    {
+      "epoch": 0.18011387307318427,
+      "grad_norm": 3.785716005792397,
+      "learning_rate": 9.42103351342199e-06,
+      "loss": 0.4496,
+      "step": 2594
+    },
+    {
+      "epoch": 0.1801833078739064,
+      "grad_norm": 4.182574179469882,
+      "learning_rate": 9.420508158295434e-06,
+      "loss": 0.5564,
+      "step": 2595
+    },
+    {
+      "epoch": 0.1802527426746285,
+      "grad_norm": 2.996819485798898,
+      "learning_rate": 9.419982579584578e-06,
+      "loss": 0.2496,
+      "step": 2596
+    },
+    {
+      "epoch": 0.18032217747535065,
+      "grad_norm": 3.5301387664346193,
+      "learning_rate": 9.419456777316006e-06,
+      "loss": 0.3734,
+      "step": 2597
+    },
+    {
+      "epoch": 0.18039161227607275,
+      "grad_norm": 3.821625108864334,
+      "learning_rate": 9.41893075151631e-06,
+      "loss": 0.4334,
+      "step": 2598
+    },
+    {
+      "epoch": 0.1804610470767949,
+      "grad_norm": 2.8667496492111826,
+      "learning_rate": 9.418404502212098e-06,
+      "loss": 0.2425,
+      "step": 2599
+    },
+    {
+      "epoch": 0.18053048187751702,
+      "grad_norm": 4.644936523177233,
+      "learning_rate": 9.417878029429985e-06,
+      "loss": 0.4482,
+      "step": 2600
+    },
+    {
+      "epoch": 0.18059991667823913,
+      "grad_norm": 4.578899850787322,
+      "learning_rate": 9.417351333196602e-06,
+      "loss": 0.4448,
+      "step": 2601
+    },
+    {
+      "epoch": 0.18066935147896127,
+      "grad_norm": 3.9513336860389603,
+      "learning_rate": 9.416824413538585e-06,
+      "loss": 0.4418,
+      "step": 2602
+    },
+    {
+      "epoch": 0.18073878627968337,
+      "grad_norm": 3.9088965785236907,
+      "learning_rate": 9.416297270482589e-06,
+      "loss": 0.2561,
+      "step": 2603
+    },
+    {
+      "epoch": 0.1808082210804055,
+      "grad_norm": 4.97371150714962,
+      "learning_rate": 9.415769904055274e-06,
+      "loss": 0.7128,
+      "step": 2604
+    },
+    {
+      "epoch": 0.1808776558811276,
+      "grad_norm": 4.7474894520208215,
+      "learning_rate": 9.415242314283312e-06,
+      "loss": 0.6806,
+      "step": 2605
+    },
+    {
+      "epoch": 0.18094709068184975,
+      "grad_norm": 4.621623965915265,
+      "learning_rate": 9.414714501193393e-06,
+      "loss": 0.7374,
+      "step": 2606
+    },
+    {
+      "epoch": 0.18101652548257186,
+      "grad_norm": 3.4151302464723146,
+      "learning_rate": 9.414186464812208e-06,
+      "loss": 0.5039,
+      "step": 2607
+    },
+    {
+      "epoch": 0.181085960283294,
+      "grad_norm": 3.867066680550426,
+      "learning_rate": 9.413658205166467e-06,
+      "loss": 0.4512,
+      "step": 2608
+    },
+    {
+      "epoch": 0.1811553950840161,
+      "grad_norm": 3.7529630769956643,
+      "learning_rate": 9.41312972228289e-06,
+      "loss": 0.482,
+      "step": 2609
+    },
+    {
+      "epoch": 0.18122482988473823,
+      "grad_norm": 3.752713128661412,
+      "learning_rate": 9.412601016188204e-06,
+      "loss": 0.525,
+      "step": 2610
+    },
+    {
+      "epoch": 0.18129426468546037,
+      "grad_norm": 4.270473284220056,
+      "learning_rate": 9.412072086909152e-06,
+      "loss": 0.3234,
+      "step": 2611
+    },
+    {
+      "epoch": 0.18136369948618247,
+      "grad_norm": 4.835264370030915,
+      "learning_rate": 9.411542934472487e-06,
+      "loss": 0.7415,
+      "step": 2612
+    },
+    {
+      "epoch": 0.1814331342869046,
+      "grad_norm": 3.986420311435972,
+      "learning_rate": 9.41101355890497e-06,
+      "loss": 0.3977,
+      "step": 2613
+    },
+    {
+      "epoch": 0.18150256908762671,
+      "grad_norm": 6.500336539729755,
+      "learning_rate": 9.41048396023338e-06,
+      "loss": 0.5897,
+      "step": 2614
+    },
+    {
+      "epoch": 0.18157200388834885,
+      "grad_norm": 4.478610637460873,
+      "learning_rate": 9.409954138484502e-06,
+      "loss": 0.59,
+      "step": 2615
+    },
+    {
+      "epoch": 0.18164143868907096,
+      "grad_norm": 4.059114296021768,
+      "learning_rate": 9.409424093685136e-06,
+      "loss": 0.6371,
+      "step": 2616
+    },
+    {
+      "epoch": 0.1817108734897931,
+      "grad_norm": 4.712199768856567,
+      "learning_rate": 9.408893825862087e-06,
+      "loss": 0.5524,
+      "step": 2617
+    },
+    {
+      "epoch": 0.1817803082905152,
+      "grad_norm": 4.829087755183378,
+      "learning_rate": 9.408363335042176e-06,
+      "loss": 0.5665,
+      "step": 2618
+    },
+    {
+      "epoch": 0.18184974309123733,
+      "grad_norm": 4.23453680908939,
+      "learning_rate": 9.407832621252238e-06,
+      "loss": 0.6399,
+      "step": 2619
+    },
+    {
+      "epoch": 0.18191917789195944,
+      "grad_norm": 4.948459135831155,
+      "learning_rate": 9.407301684519114e-06,
+      "loss": 0.4331,
+      "step": 2620
+    },
+    {
+      "epoch": 0.18198861269268157,
+      "grad_norm": 3.4723506064251817,
+      "learning_rate": 9.406770524869657e-06,
+      "loss": 0.4795,
+      "step": 2621
+    },
+    {
+      "epoch": 0.1820580474934037,
+      "grad_norm": 5.035648561846344,
+      "learning_rate": 9.406239142330734e-06,
+      "loss": 0.4367,
+      "step": 2622
+    },
+    {
+      "epoch": 0.18212748229412581,
+      "grad_norm": 3.863234836385909,
+      "learning_rate": 9.405707536929221e-06,
+      "loss": 0.4996,
+      "step": 2623
+    },
+    {
+      "epoch": 0.18219691709484795,
+      "grad_norm": 4.065753722644651,
+      "learning_rate": 9.405175708692005e-06,
+      "loss": 0.493,
+      "step": 2624
+    },
+    {
+      "epoch": 0.18226635189557006,
+      "grad_norm": 3.8633272158548846,
+      "learning_rate": 9.404643657645988e-06,
+      "loss": 0.4596,
+      "step": 2625
+    },
+    {
+      "epoch": 0.1823357866962922,
+      "grad_norm": 4.2708432403884675,
+      "learning_rate": 9.40411138381808e-06,
+      "loss": 0.3629,
+      "step": 2626
+    },
+    {
+      "epoch": 0.1824052214970143,
+      "grad_norm": 4.455077326166294,
+      "learning_rate": 9.403578887235198e-06,
+      "loss": 0.6898,
+      "step": 2627
+    },
+    {
+      "epoch": 0.18247465629773643,
+      "grad_norm": 2.871854933908505,
+      "learning_rate": 9.403046167924282e-06,
+      "loss": 0.4401,
+      "step": 2628
+    },
+    {
+      "epoch": 0.18254409109845854,
+      "grad_norm": 3.6874690350425587,
+      "learning_rate": 9.402513225912273e-06,
+      "loss": 0.5697,
+      "step": 2629
+    },
+    {
+      "epoch": 0.18261352589918067,
+      "grad_norm": 4.189768546522508,
+      "learning_rate": 9.401980061226126e-06,
+      "loss": 0.4704,
+      "step": 2630
+    },
+    {
+      "epoch": 0.18268296069990278,
+      "grad_norm": 4.209606084958585,
+      "learning_rate": 9.40144667389281e-06,
+      "loss": 0.4992,
+      "step": 2631
+    },
+    {
+      "epoch": 0.18275239550062491,
+      "grad_norm": 4.368302939734846,
+      "learning_rate": 9.400913063939299e-06,
+      "loss": 0.6804,
+      "step": 2632
+    },
+    {
+      "epoch": 0.18282183030134702,
+      "grad_norm": 3.755068095445998,
+      "learning_rate": 9.400379231392587e-06,
+      "loss": 0.4957,
+      "step": 2633
+    },
+    {
+      "epoch": 0.18289126510206916,
+      "grad_norm": 3.6279847827056257,
+      "learning_rate": 9.399845176279673e-06,
+      "loss": 0.5565,
+      "step": 2634
+    },
+    {
+      "epoch": 0.1829606999027913,
+      "grad_norm": 4.830164038300566,
+      "learning_rate": 9.39931089862757e-06,
+      "loss": 0.6452,
+      "step": 2635
+    },
+    {
+      "epoch": 0.1830301347035134,
+      "grad_norm": 3.842295137568818,
+      "learning_rate": 9.398776398463296e-06,
+      "loss": 0.3495,
+      "step": 2636
+    },
+    {
+      "epoch": 0.18309956950423553,
+      "grad_norm": 3.8495804633275235,
+      "learning_rate": 9.398241675813893e-06,
+      "loss": 0.6371,
+      "step": 2637
+    },
+    {
+      "epoch": 0.18316900430495764,
+      "grad_norm": 4.103967734243168,
+      "learning_rate": 9.397706730706402e-06,
+      "loss": 0.5425,
+      "step": 2638
+    },
+    {
+      "epoch": 0.18323843910567977,
+      "grad_norm": 4.6841532757175415,
+      "learning_rate": 9.39717156316788e-06,
+      "loss": 0.5694,
+      "step": 2639
+    },
+    {
+      "epoch": 0.18330787390640188,
+      "grad_norm": 5.649685095767387,
+      "learning_rate": 9.396636173225398e-06,
+      "loss": 0.8654,
+      "step": 2640
+    },
+    {
+      "epoch": 0.18337730870712401,
+      "grad_norm": 4.061522963793848,
+      "learning_rate": 9.39610056090603e-06,
+      "loss": 0.5021,
+      "step": 2641
+    },
+    {
+      "epoch": 0.18344674350784612,
+      "grad_norm": 3.464055033847907,
+      "learning_rate": 9.395564726236872e-06,
+      "loss": 0.3587,
+      "step": 2642
+    },
+    {
+      "epoch": 0.18351617830856826,
+      "grad_norm": 5.045865020505513,
+      "learning_rate": 9.395028669245025e-06,
+      "loss": 0.9348,
+      "step": 2643
+    },
+    {
+      "epoch": 0.18358561310929036,
+      "grad_norm": 4.455677300812232,
+      "learning_rate": 9.394492389957602e-06,
+      "loss": 0.7207,
+      "step": 2644
+    },
+    {
+      "epoch": 0.1836550479100125,
+      "grad_norm": 3.4874376683372055,
+      "learning_rate": 9.393955888401725e-06,
+      "loss": 0.3084,
+      "step": 2645
+    },
+    {
+      "epoch": 0.18372448271073463,
+      "grad_norm": 3.5953519400125398,
+      "learning_rate": 9.39341916460453e-06,
+      "loss": 0.4434,
+      "step": 2646
+    },
+    {
+      "epoch": 0.18379391751145674,
+      "grad_norm": 4.362552652175736,
+      "learning_rate": 9.392882218593166e-06,
+      "loss": 0.7288,
+      "step": 2647
+    },
+    {
+      "epoch": 0.18386335231217887,
+      "grad_norm": 4.2848285050376695,
+      "learning_rate": 9.392345050394792e-06,
+      "loss": 0.5396,
+      "step": 2648
+    },
+    {
+      "epoch": 0.18393278711290098,
+      "grad_norm": 3.746814563243955,
+      "learning_rate": 9.391807660036575e-06,
+      "loss": 0.3151,
+      "step": 2649
+    },
+    {
+      "epoch": 0.18400222191362312,
+      "grad_norm": 3.4428445199161866,
+      "learning_rate": 9.391270047545695e-06,
+      "loss": 0.3143,
+      "step": 2650
+    },
+    {
+      "epoch": 0.18407165671434522,
+      "grad_norm": 3.615423917648399,
+      "learning_rate": 9.390732212949345e-06,
+      "loss": 0.3328,
+      "step": 2651
+    },
+    {
+      "epoch": 0.18414109151506736,
+      "grad_norm": 3.8539542676126404,
+      "learning_rate": 9.39019415627473e-06,
+      "loss": 0.4763,
+      "step": 2652
+    },
+    {
+      "epoch": 0.18421052631578946,
+      "grad_norm": 3.589699469516653,
+      "learning_rate": 9.389655877549062e-06,
+      "loss": 0.5355,
+      "step": 2653
+    },
+    {
+      "epoch": 0.1842799611165116,
+      "grad_norm": 4.7557986634837395,
+      "learning_rate": 9.389117376799565e-06,
+      "loss": 0.6126,
+      "step": 2654
+    },
+    {
+      "epoch": 0.1843493959172337,
+      "grad_norm": 4.252885564188266,
+      "learning_rate": 9.38857865405348e-06,
+      "loss": 0.6426,
+      "step": 2655
+    },
+    {
+      "epoch": 0.18441883071795584,
+      "grad_norm": 4.850771133809857,
+      "learning_rate": 9.38803970933805e-06,
+      "loss": 0.6627,
+      "step": 2656
+    },
+    {
+      "epoch": 0.18448826551867797,
+      "grad_norm": 4.745332883283432,
+      "learning_rate": 9.38750054268054e-06,
+      "loss": 0.7279,
+      "step": 2657
+    },
+    {
+      "epoch": 0.18455770031940008,
+      "grad_norm": 4.998169672481638,
+      "learning_rate": 9.386961154108216e-06,
+      "loss": 0.8132,
+      "step": 2658
+    },
+    {
+      "epoch": 0.18462713512012222,
+      "grad_norm": 4.216403963800552,
+      "learning_rate": 9.386421543648361e-06,
+      "loss": 0.429,
+      "step": 2659
+    },
+    {
+      "epoch": 0.18469656992084432,
+      "grad_norm": 3.285685174883587,
+      "learning_rate": 9.385881711328268e-06,
+      "loss": 0.2602,
+      "step": 2660
+    },
+    {
+      "epoch": 0.18476600472156646,
+      "grad_norm": 4.2732567440425875,
+      "learning_rate": 9.38534165717524e-06,
+      "loss": 0.4432,
+      "step": 2661
+    },
+    {
+      "epoch": 0.18483543952228856,
+      "grad_norm": 4.515064034558809,
+      "learning_rate": 9.384801381216595e-06,
+      "loss": 0.3799,
+      "step": 2662
+    },
+    {
+      "epoch": 0.1849048743230107,
+      "grad_norm": 3.999345716429541,
+      "learning_rate": 9.384260883479657e-06,
+      "loss": 0.5192,
+      "step": 2663
+    },
+    {
+      "epoch": 0.1849743091237328,
+      "grad_norm": 3.093273078533135,
+      "learning_rate": 9.383720163991763e-06,
+      "loss": 0.4324,
+      "step": 2664
+    },
+    {
+      "epoch": 0.18504374392445494,
+      "grad_norm": 3.253106532777847,
+      "learning_rate": 9.383179222780266e-06,
+      "loss": 0.3638,
+      "step": 2665
+    },
+    {
+      "epoch": 0.18511317872517705,
+      "grad_norm": 4.4860566801124175,
+      "learning_rate": 9.382638059872523e-06,
+      "loss": 0.3877,
+      "step": 2666
+    },
+    {
+      "epoch": 0.18518261352589918,
+      "grad_norm": 4.272186455965108,
+      "learning_rate": 9.382096675295906e-06,
+      "loss": 0.5467,
+      "step": 2667
+    },
+    {
+      "epoch": 0.18525204832662132,
+      "grad_norm": 4.63151004344994,
+      "learning_rate": 9.381555069077799e-06,
+      "loss": 0.5893,
+      "step": 2668
+    },
+    {
+      "epoch": 0.18532148312734342,
+      "grad_norm": 3.971234968677336,
+      "learning_rate": 9.381013241245593e-06,
+      "loss": 0.5765,
+      "step": 2669
+    },
+    {
+      "epoch": 0.18539091792806556,
+      "grad_norm": 4.62659117515569,
+      "learning_rate": 9.380471191826696e-06,
+      "loss": 0.8164,
+      "step": 2670
+    },
+    {
+      "epoch": 0.18546035272878766,
+      "grad_norm": 4.661653313611526,
+      "learning_rate": 9.379928920848523e-06,
+      "loss": 0.452,
+      "step": 2671
+    },
+    {
+      "epoch": 0.1855297875295098,
+      "grad_norm": 3.9862730516995115,
+      "learning_rate": 9.379386428338499e-06,
+      "loss": 0.3762,
+      "step": 2672
+    },
+    {
+      "epoch": 0.1855992223302319,
+      "grad_norm": 3.3959861824044153,
+      "learning_rate": 9.378843714324069e-06,
+      "loss": 0.3968,
+      "step": 2673
+    },
+    {
+      "epoch": 0.18566865713095404,
+      "grad_norm": 4.130213865502688,
+      "learning_rate": 9.378300778832676e-06,
+      "loss": 0.5228,
+      "step": 2674
+    },
+    {
+      "epoch": 0.18573809193167615,
+      "grad_norm": 4.000332228479085,
+      "learning_rate": 9.377757621891786e-06,
+      "loss": 0.6585,
+      "step": 2675
+    },
+    {
+      "epoch": 0.18580752673239828,
+      "grad_norm": 4.360601849195899,
+      "learning_rate": 9.377214243528868e-06,
+      "loss": 0.7967,
+      "step": 2676
+    },
+    {
+      "epoch": 0.1858769615331204,
+      "grad_norm": 3.7506733547067315,
+      "learning_rate": 9.376670643771405e-06,
+      "loss": 0.5865,
+      "step": 2677
+    },
+    {
+      "epoch": 0.18594639633384252,
+      "grad_norm": 4.293988253689358,
+      "learning_rate": 9.376126822646895e-06,
+      "loss": 0.6054,
+      "step": 2678
+    },
+    {
+      "epoch": 0.18601583113456466,
+      "grad_norm": 4.565326671518356,
+      "learning_rate": 9.375582780182843e-06,
+      "loss": 0.581,
+      "step": 2679
+    },
+    {
+      "epoch": 0.18608526593528676,
+      "grad_norm": 3.667405717599369,
+      "learning_rate": 9.375038516406765e-06,
+      "loss": 0.4897,
+      "step": 2680
+    },
+    {
+      "epoch": 0.1861547007360089,
+      "grad_norm": 4.674126851451416,
+      "learning_rate": 9.374494031346189e-06,
+      "loss": 0.5175,
+      "step": 2681
+    },
+    {
+      "epoch": 0.186224135536731,
+      "grad_norm": 4.693795148754813,
+      "learning_rate": 9.373949325028655e-06,
+      "loss": 0.6797,
+      "step": 2682
+    },
+    {
+      "epoch": 0.18629357033745314,
+      "grad_norm": 4.517987558032381,
+      "learning_rate": 9.373404397481715e-06,
+      "loss": 0.6698,
+      "step": 2683
+    },
+    {
+      "epoch": 0.18636300513817525,
+      "grad_norm": 3.306546482658048,
+      "learning_rate": 9.372859248732928e-06,
+      "loss": 0.3668,
+      "step": 2684
+    },
+    {
+      "epoch": 0.18643243993889738,
+      "grad_norm": 3.4324960784679415,
+      "learning_rate": 9.372313878809868e-06,
+      "loss": 0.4192,
+      "step": 2685
+    },
+    {
+      "epoch": 0.1865018747396195,
+      "grad_norm": 3.893627845899475,
+      "learning_rate": 9.371768287740121e-06,
+      "loss": 0.4881,
+      "step": 2686
+    },
+    {
+      "epoch": 0.18657130954034162,
+      "grad_norm": 4.637353267076004,
+      "learning_rate": 9.371222475551281e-06,
+      "loss": 0.6366,
+      "step": 2687
+    },
+    {
+      "epoch": 0.18664074434106373,
+      "grad_norm": 5.496973660559162,
+      "learning_rate": 9.370676442270953e-06,
+      "loss": 0.4352,
+      "step": 2688
+    },
+    {
+      "epoch": 0.18671017914178586,
+      "grad_norm": 4.704562350998911,
+      "learning_rate": 9.370130187926759e-06,
+      "loss": 0.5153,
+      "step": 2689
+    },
+    {
+      "epoch": 0.18677961394250797,
+      "grad_norm": 4.099705667440132,
+      "learning_rate": 9.369583712546322e-06,
+      "loss": 0.5575,
+      "step": 2690
+    },
+    {
+      "epoch": 0.1868490487432301,
+      "grad_norm": 3.9142226362209414,
+      "learning_rate": 9.369037016157286e-06,
+      "loss": 0.6525,
+      "step": 2691
+    },
+    {
+      "epoch": 0.18691848354395224,
+      "grad_norm": 5.092744404946171,
+      "learning_rate": 9.368490098787302e-06,
+      "loss": 0.7738,
+      "step": 2692
+    },
+    {
+      "epoch": 0.18698791834467435,
+      "grad_norm": 4.366350397356477,
+      "learning_rate": 9.367942960464034e-06,
+      "loss": 0.6119,
+      "step": 2693
+    },
+    {
+      "epoch": 0.18705735314539648,
+      "grad_norm": 3.7796989173002964,
+      "learning_rate": 9.36739560121515e-06,
+      "loss": 0.3767,
+      "step": 2694
+    },
+    {
+      "epoch": 0.1871267879461186,
+      "grad_norm": 4.465157904727939,
+      "learning_rate": 9.366848021068341e-06,
+      "loss": 0.4584,
+      "step": 2695
+    },
+    {
+      "epoch": 0.18719622274684072,
+      "grad_norm": 3.8234237737463452,
+      "learning_rate": 9.366300220051301e-06,
+      "loss": 0.5005,
+      "step": 2696
+    },
+    {
+      "epoch": 0.18726565754756283,
+      "grad_norm": 3.750841668592917,
+      "learning_rate": 9.365752198191735e-06,
+      "loss": 0.4426,
+      "step": 2697
+    },
+    {
+      "epoch": 0.18733509234828497,
+      "grad_norm": 3.553119612657192,
+      "learning_rate": 9.365203955517366e-06,
+      "loss": 0.4838,
+      "step": 2698
+    },
+    {
+      "epoch": 0.18740452714900707,
+      "grad_norm": 4.569877706251379,
+      "learning_rate": 9.364655492055917e-06,
+      "loss": 0.5901,
+      "step": 2699
+    },
+    {
+      "epoch": 0.1874739619497292,
+      "grad_norm": 3.264565579017233,
+      "learning_rate": 9.364106807835134e-06,
+      "loss": 0.3152,
+      "step": 2700
+    },
+    {
+      "epoch": 0.1875433967504513,
+      "grad_norm": 3.911946645297428,
+      "learning_rate": 9.363557902882768e-06,
+      "loss": 0.4852,
+      "step": 2701
+    },
+    {
+      "epoch": 0.18761283155117345,
+      "grad_norm": 4.057238251950517,
+      "learning_rate": 9.363008777226579e-06,
+      "loss": 0.598,
+      "step": 2702
+    },
+    {
+      "epoch": 0.18768226635189558,
+      "grad_norm": 4.942989454162959,
+      "learning_rate": 9.362459430894342e-06,
+      "loss": 0.9231,
+      "step": 2703
+    },
+    {
+      "epoch": 0.1877517011526177,
+      "grad_norm": 3.4189552785359854,
+      "learning_rate": 9.361909863913845e-06,
+      "loss": 0.5374,
+      "step": 2704
+    },
+    {
+      "epoch": 0.18782113595333982,
+      "grad_norm": 3.300158387504638,
+      "learning_rate": 9.361360076312885e-06,
+      "loss": 0.379,
+      "step": 2705
+    },
+    {
+      "epoch": 0.18789057075406193,
+      "grad_norm": 4.269859189069414,
+      "learning_rate": 9.360810068119263e-06,
+      "loss": 0.6118,
+      "step": 2706
+    },
+    {
+      "epoch": 0.18796000555478407,
+      "grad_norm": 4.414861276635057,
+      "learning_rate": 9.360259839360806e-06,
+      "loss": 0.5451,
+      "step": 2707
+    },
+    {
+      "epoch": 0.18802944035550617,
+      "grad_norm": 4.971972510749617,
+      "learning_rate": 9.35970939006534e-06,
+      "loss": 0.5853,
+      "step": 2708
+    },
+    {
+      "epoch": 0.1880988751562283,
+      "grad_norm": 3.2139045384928284,
+      "learning_rate": 9.359158720260704e-06,
+      "loss": 0.2629,
+      "step": 2709
+    },
+    {
+      "epoch": 0.1881683099569504,
+      "grad_norm": 4.2835668957995585,
+      "learning_rate": 9.358607829974755e-06,
+      "loss": 0.598,
+      "step": 2710
+    },
+    {
+      "epoch": 0.18823774475767255,
+      "grad_norm": 4.576594729965632,
+      "learning_rate": 9.358056719235353e-06,
+      "loss": 0.6368,
+      "step": 2711
+    },
+    {
+      "epoch": 0.18830717955839466,
+      "grad_norm": 3.649299382433351,
+      "learning_rate": 9.357505388070374e-06,
+      "loss": 0.6356,
+      "step": 2712
+    },
+    {
+      "epoch": 0.1883766143591168,
+      "grad_norm": 4.190225111288776,
+      "learning_rate": 9.356953836507702e-06,
+      "loss": 0.5943,
+      "step": 2713
+    },
+    {
+      "epoch": 0.18844604915983892,
+      "grad_norm": 3.799433608007023,
+      "learning_rate": 9.356402064575237e-06,
+      "loss": 0.4415,
+      "step": 2714
+    },
+    {
+      "epoch": 0.18851548396056103,
+      "grad_norm": 3.2704213302503367,
+      "learning_rate": 9.355850072300885e-06,
+      "loss": 0.4086,
+      "step": 2715
+    },
+    {
+      "epoch": 0.18858491876128317,
+      "grad_norm": 4.809849727129228,
+      "learning_rate": 9.355297859712565e-06,
+      "loss": 0.517,
+      "step": 2716
+    },
+    {
+      "epoch": 0.18865435356200527,
+      "grad_norm": 3.51281557191898,
+      "learning_rate": 9.354745426838208e-06,
+      "loss": 0.3684,
+      "step": 2717
+    },
+    {
+      "epoch": 0.1887237883627274,
+      "grad_norm": 3.9953390577791876,
+      "learning_rate": 9.354192773705755e-06,
+      "loss": 0.4533,
+      "step": 2718
+    },
+    {
+      "epoch": 0.18879322316344951,
+      "grad_norm": 3.4078418732593767,
+      "learning_rate": 9.353639900343156e-06,
+      "loss": 0.2911,
+      "step": 2719
+    },
+    {
+      "epoch": 0.18886265796417165,
+      "grad_norm": 3.6734345955257854,
+      "learning_rate": 9.35308680677838e-06,
+      "loss": 0.4506,
+      "step": 2720
+    },
+    {
+      "epoch": 0.18893209276489376,
+      "grad_norm": 3.4191989686856177,
+      "learning_rate": 9.352533493039399e-06,
+      "loss": 0.3259,
+      "step": 2721
+    },
+    {
+      "epoch": 0.1890015275656159,
+      "grad_norm": 3.6567529470743425,
+      "learning_rate": 9.351979959154198e-06,
+      "loss": 0.491,
+      "step": 2722
+    },
+    {
+      "epoch": 0.189070962366338,
+      "grad_norm": 4.624552391559659,
+      "learning_rate": 9.351426205150778e-06,
+      "loss": 0.7231,
+      "step": 2723
+    },
+    {
+      "epoch": 0.18914039716706013,
+      "grad_norm": 5.913573668756369,
+      "learning_rate": 9.350872231057143e-06,
+      "loss": 0.6042,
+      "step": 2724
+    },
+    {
+      "epoch": 0.18920983196778227,
+      "grad_norm": 2.6099992463084014,
+      "learning_rate": 9.35031803690131e-06,
+      "loss": 0.3139,
+      "step": 2725
+    },
+    {
+      "epoch": 0.18927926676850437,
+      "grad_norm": 4.50995402186975,
+      "learning_rate": 9.349763622711318e-06,
+      "loss": 0.6777,
+      "step": 2726
+    },
+    {
+      "epoch": 0.1893487015692265,
+      "grad_norm": 3.2216086594793483,
+      "learning_rate": 9.349208988515204e-06,
+      "loss": 0.3374,
+      "step": 2727
+    },
+    {
+      "epoch": 0.18941813636994861,
+      "grad_norm": 4.107081382539988,
+      "learning_rate": 9.34865413434102e-06,
+      "loss": 0.6015,
+      "step": 2728
+    },
+    {
+      "epoch": 0.18948757117067075,
+      "grad_norm": 3.61101862039446,
+      "learning_rate": 9.34809906021683e-06,
+      "loss": 0.337,
+      "step": 2729
+    },
+    {
+      "epoch": 0.18955700597139286,
+      "grad_norm": 4.354232240522092,
+      "learning_rate": 9.34754376617071e-06,
+      "loss": 0.3765,
+      "step": 2730
+    },
+    {
+      "epoch": 0.189626440772115,
+      "grad_norm": 4.351544470282709,
+      "learning_rate": 9.346988252230746e-06,
+      "loss": 0.4047,
+      "step": 2731
+    },
+    {
+      "epoch": 0.1896958755728371,
+      "grad_norm": 3.7045585897261866,
+      "learning_rate": 9.346432518425035e-06,
+      "loss": 0.4443,
+      "step": 2732
+    },
+    {
+      "epoch": 0.18976531037355923,
+      "grad_norm": 3.5778556173733738,
+      "learning_rate": 9.345876564781687e-06,
+      "loss": 0.4219,
+      "step": 2733
+    },
+    {
+      "epoch": 0.18983474517428134,
+      "grad_norm": 3.6630542223987717,
+      "learning_rate": 9.345320391328818e-06,
+      "loss": 0.3879,
+      "step": 2734
+    },
+    {
+      "epoch": 0.18990417997500347,
+      "grad_norm": 4.429881979430322,
+      "learning_rate": 9.344763998094561e-06,
+      "loss": 0.4893,
+      "step": 2735
+    },
+    {
+      "epoch": 0.18997361477572558,
+      "grad_norm": 4.055937117026432,
+      "learning_rate": 9.344207385107058e-06,
+      "loss": 0.5636,
+      "step": 2736
+    },
+    {
+      "epoch": 0.19004304957644771,
+      "grad_norm": 3.7728912695566645,
+      "learning_rate": 9.343650552394461e-06,
+      "loss": 0.3829,
+      "step": 2737
+    },
+    {
+      "epoch": 0.19011248437716985,
+      "grad_norm": 3.253544783639077,
+      "learning_rate": 9.343093499984934e-06,
+      "loss": 0.3914,
+      "step": 2738
+    },
+    {
+      "epoch": 0.19018191917789196,
+      "grad_norm": 2.821709914668401,
+      "learning_rate": 9.342536227906653e-06,
+      "loss": 0.1828,
+      "step": 2739
+    },
+    {
+      "epoch": 0.1902513539786141,
+      "grad_norm": 3.812753702751062,
+      "learning_rate": 9.341978736187805e-06,
+      "loss": 0.4588,
+      "step": 2740
+    },
+    {
+      "epoch": 0.1903207887793362,
+      "grad_norm": 3.9297013449581657,
+      "learning_rate": 9.341421024856582e-06,
+      "loss": 0.5823,
+      "step": 2741
+    },
+    {
+      "epoch": 0.19039022358005833,
+      "grad_norm": 3.4883306366601654,
+      "learning_rate": 9.3408630939412e-06,
+      "loss": 0.5591,
+      "step": 2742
+    },
+    {
+      "epoch": 0.19045965838078044,
+      "grad_norm": 3.511011886453169,
+      "learning_rate": 9.340304943469871e-06,
+      "loss": 0.3234,
+      "step": 2743
+    },
+    {
+      "epoch": 0.19052909318150257,
+      "grad_norm": 4.479663853988132,
+      "learning_rate": 9.339746573470832e-06,
+      "loss": 0.7253,
+      "step": 2744
+    },
+    {
+      "epoch": 0.19059852798222468,
+      "grad_norm": 4.487626675708971,
+      "learning_rate": 9.339187983972321e-06,
+      "loss": 0.5299,
+      "step": 2745
+    },
+    {
+      "epoch": 0.19066796278294681,
+      "grad_norm": 3.7987376491743134,
+      "learning_rate": 9.338629175002592e-06,
+      "loss": 0.3924,
+      "step": 2746
+    },
+    {
+      "epoch": 0.19073739758366892,
+      "grad_norm": 5.6539131576751345,
+      "learning_rate": 9.33807014658991e-06,
+      "loss": 0.7562,
+      "step": 2747
+    },
+    {
+      "epoch": 0.19080683238439106,
+      "grad_norm": 3.290502924696365,
+      "learning_rate": 9.337510898762548e-06,
+      "loss": 0.2409,
+      "step": 2748
+    },
+    {
+      "epoch": 0.1908762671851132,
+      "grad_norm": 3.562466193535862,
+      "learning_rate": 9.336951431548794e-06,
+      "loss": 0.3545,
+      "step": 2749
+    },
+    {
+      "epoch": 0.1909457019858353,
+      "grad_norm": 3.460301785349842,
+      "learning_rate": 9.336391744976942e-06,
+      "loss": 0.3328,
+      "step": 2750
+    },
+    {
+      "epoch": 0.19101513678655743,
+      "grad_norm": 5.162659018358059,
+      "learning_rate": 9.335831839075303e-06,
+      "loss": 0.7052,
+      "step": 2751
+    },
+    {
+      "epoch": 0.19108457158727954,
+      "grad_norm": 4.929602906230562,
+      "learning_rate": 9.335271713872198e-06,
+      "loss": 0.669,
+      "step": 2752
+    },
+    {
+      "epoch": 0.19115400638800167,
+      "grad_norm": 4.430727543953501,
+      "learning_rate": 9.334711369395953e-06,
+      "loss": 0.6739,
+      "step": 2753
+    },
+    {
+      "epoch": 0.19122344118872378,
+      "grad_norm": 3.061095759664949,
+      "learning_rate": 9.334150805674913e-06,
+      "loss": 0.3962,
+      "step": 2754
+    },
+    {
+      "epoch": 0.19129287598944592,
+      "grad_norm": 5.1623950974380675,
+      "learning_rate": 9.333590022737428e-06,
+      "loss": 0.5439,
+      "step": 2755
+    },
+    {
+      "epoch": 0.19136231079016802,
+      "grad_norm": 4.463649858992726,
+      "learning_rate": 9.333029020611865e-06,
+      "loss": 0.4312,
+      "step": 2756
+    },
+    {
+      "epoch": 0.19143174559089016,
+      "grad_norm": 4.8136711203666875,
+      "learning_rate": 9.332467799326599e-06,
+      "loss": 0.8508,
+      "step": 2757
+    },
+    {
+      "epoch": 0.19150118039161226,
+      "grad_norm": 4.544231933795228,
+      "learning_rate": 9.331906358910012e-06,
+      "loss": 0.7869,
+      "step": 2758
+    },
+    {
+      "epoch": 0.1915706151923344,
+      "grad_norm": 4.003804009963702,
+      "learning_rate": 9.331344699390504e-06,
+      "loss": 0.3863,
+      "step": 2759
+    },
+    {
+      "epoch": 0.19164004999305653,
+      "grad_norm": 3.6890801937053013,
+      "learning_rate": 9.330782820796482e-06,
+      "loss": 0.4428,
+      "step": 2760
+    },
+    {
+      "epoch": 0.19170948479377864,
+      "grad_norm": 3.791804860372565,
+      "learning_rate": 9.330220723156365e-06,
+      "loss": 0.3654,
+      "step": 2761
+    },
+    {
+      "epoch": 0.19177891959450077,
+      "grad_norm": 4.019905593335045,
+      "learning_rate": 9.329658406498586e-06,
+      "loss": 0.3392,
+      "step": 2762
+    },
+    {
+      "epoch": 0.19184835439522288,
+      "grad_norm": 3.4847390614949942,
+      "learning_rate": 9.329095870851581e-06,
+      "loss": 0.454,
+      "step": 2763
+    },
+    {
+      "epoch": 0.19191778919594502,
+      "grad_norm": 3.8882326450944285,
+      "learning_rate": 9.328533116243807e-06,
+      "loss": 0.5431,
+      "step": 2764
+    },
+    {
+      "epoch": 0.19198722399666712,
+      "grad_norm": 3.578616256686992,
+      "learning_rate": 9.327970142703726e-06,
+      "loss": 0.3556,
+      "step": 2765
+    },
+    {
+      "epoch": 0.19205665879738926,
+      "grad_norm": 4.471841768198546,
+      "learning_rate": 9.327406950259814e-06,
+      "loss": 0.5637,
+      "step": 2766
+    },
+    {
+      "epoch": 0.19212609359811136,
+      "grad_norm": 4.314400921717331,
+      "learning_rate": 9.326843538940553e-06,
+      "loss": 0.5535,
+      "step": 2767
+    },
+    {
+      "epoch": 0.1921955283988335,
+      "grad_norm": 4.252126876824009,
+      "learning_rate": 9.326279908774444e-06,
+      "loss": 0.564,
+      "step": 2768
+    },
+    {
+      "epoch": 0.1922649631995556,
+      "grad_norm": 3.9393460822026127,
+      "learning_rate": 9.32571605978999e-06,
+      "loss": 0.7253,
+      "step": 2769
+    },
+    {
+      "epoch": 0.19233439800027774,
+      "grad_norm": 3.4707706948114545,
+      "learning_rate": 9.325151992015715e-06,
+      "loss": 0.2873,
+      "step": 2770
+    },
+    {
+      "epoch": 0.19240383280099987,
+      "grad_norm": 4.311575867253222,
+      "learning_rate": 9.324587705480145e-06,
+      "loss": 0.5869,
+      "step": 2771
+    },
+    {
+      "epoch": 0.19247326760172198,
+      "grad_norm": 4.989532099298601,
+      "learning_rate": 9.324023200211825e-06,
+      "loss": 0.4649,
+      "step": 2772
+    },
+    {
+      "epoch": 0.19254270240244412,
+      "grad_norm": 3.6178106807347805,
+      "learning_rate": 9.323458476239302e-06,
+      "loss": 0.3059,
+      "step": 2773
+    },
+    {
+      "epoch": 0.19261213720316622,
+      "grad_norm": 4.909673010947121,
+      "learning_rate": 9.322893533591144e-06,
+      "loss": 0.5213,
+      "step": 2774
+    },
+    {
+      "epoch": 0.19268157200388836,
+      "grad_norm": 3.4305574362653704,
+      "learning_rate": 9.32232837229592e-06,
+      "loss": 0.3366,
+      "step": 2775
+    },
+    {
+      "epoch": 0.19275100680461046,
+      "grad_norm": 4.390128546793546,
+      "learning_rate": 9.32176299238222e-06,
+      "loss": 0.6442,
+      "step": 2776
+    },
+    {
+      "epoch": 0.1928204416053326,
+      "grad_norm": 3.0614079651599946,
+      "learning_rate": 9.321197393878637e-06,
+      "loss": 0.1788,
+      "step": 2777
+    },
+    {
+      "epoch": 0.1928898764060547,
+      "grad_norm": 4.285743652857991,
+      "learning_rate": 9.320631576813779e-06,
+      "loss": 0.485,
+      "step": 2778
+    },
+    {
+      "epoch": 0.19295931120677684,
+      "grad_norm": 3.8545005090056277,
+      "learning_rate": 9.320065541216265e-06,
+      "loss": 0.3717,
+      "step": 2779
+    },
+    {
+      "epoch": 0.19302874600749895,
+      "grad_norm": 5.4509651800760635,
+      "learning_rate": 9.319499287114726e-06,
+      "loss": 0.5684,
+      "step": 2780
+    },
+    {
+      "epoch": 0.19309818080822108,
+      "grad_norm": 3.2767933437028454,
+      "learning_rate": 9.318932814537801e-06,
+      "loss": 0.4264,
+      "step": 2781
+    },
+    {
+      "epoch": 0.19316761560894322,
+      "grad_norm": 3.173529415205206,
+      "learning_rate": 9.31836612351414e-06,
+      "loss": 0.2306,
+      "step": 2782
+    },
+    {
+      "epoch": 0.19323705040966532,
+      "grad_norm": 4.374773340902683,
+      "learning_rate": 9.317799214072408e-06,
+      "loss": 0.4269,
+      "step": 2783
+    },
+    {
+      "epoch": 0.19330648521038746,
+      "grad_norm": 3.3408269185084256,
+      "learning_rate": 9.317232086241277e-06,
+      "loss": 0.2131,
+      "step": 2784
+    },
+    {
+      "epoch": 0.19337592001110956,
+      "grad_norm": 5.302543156896799,
+      "learning_rate": 9.316664740049433e-06,
+      "loss": 0.5825,
+      "step": 2785
+    },
+    {
+      "epoch": 0.1934453548118317,
+      "grad_norm": 3.168899584533589,
+      "learning_rate": 9.316097175525571e-06,
+      "loss": 0.4335,
+      "step": 2786
+    },
+    {
+      "epoch": 0.1935147896125538,
+      "grad_norm": 3.9844736151978393,
+      "learning_rate": 9.315529392698396e-06,
+      "loss": 0.6811,
+      "step": 2787
+    },
+    {
+      "epoch": 0.19358422441327594,
+      "grad_norm": 4.663782641145318,
+      "learning_rate": 9.314961391596631e-06,
+      "loss": 0.5962,
+      "step": 2788
+    },
+    {
+      "epoch": 0.19365365921399805,
+      "grad_norm": 5.314800989511133,
+      "learning_rate": 9.314393172249e-06,
+      "loss": 0.7669,
+      "step": 2789
+    },
+    {
+      "epoch": 0.19372309401472018,
+      "grad_norm": 5.251449302011684,
+      "learning_rate": 9.313824734684243e-06,
+      "loss": 0.7401,
+      "step": 2790
+    },
+    {
+      "epoch": 0.1937925288154423,
+      "grad_norm": 3.515634392366791,
+      "learning_rate": 9.313256078931115e-06,
+      "loss": 0.3952,
+      "step": 2791
+    },
+    {
+      "epoch": 0.19386196361616442,
+      "grad_norm": 4.161358828944821,
+      "learning_rate": 9.312687205018373e-06,
+      "loss": 0.4989,
+      "step": 2792
+    },
+    {
+      "epoch": 0.19393139841688653,
+      "grad_norm": 4.544955284737086,
+      "learning_rate": 9.312118112974792e-06,
+      "loss": 0.4924,
+      "step": 2793
+    },
+    {
+      "epoch": 0.19400083321760866,
+      "grad_norm": 4.4341974414983,
+      "learning_rate": 9.31154880282916e-06,
+      "loss": 0.6021,
+      "step": 2794
+    },
+    {
+      "epoch": 0.1940702680183308,
+      "grad_norm": 3.671305050461036,
+      "learning_rate": 9.310979274610266e-06,
+      "loss": 0.3005,
+      "step": 2795
+    },
+    {
+      "epoch": 0.1941397028190529,
+      "grad_norm": 4.483897711088626,
+      "learning_rate": 9.310409528346917e-06,
+      "loss": 0.6386,
+      "step": 2796
+    },
+    {
+      "epoch": 0.19420913761977504,
+      "grad_norm": 4.731574327185386,
+      "learning_rate": 9.309839564067932e-06,
+      "loss": 0.6518,
+      "step": 2797
+    },
+    {
+      "epoch": 0.19427857242049715,
+      "grad_norm": 4.300848137354891,
+      "learning_rate": 9.309269381802142e-06,
+      "loss": 0.5616,
+      "step": 2798
+    },
+    {
+      "epoch": 0.19434800722121928,
+      "grad_norm": 2.9043070145178507,
+      "learning_rate": 9.30869898157838e-06,
+      "loss": 0.2264,
+      "step": 2799
+    },
+    {
+      "epoch": 0.1944174420219414,
+      "grad_norm": 4.5630022657246485,
+      "learning_rate": 9.3081283634255e-06,
+      "loss": 0.5008,
+      "step": 2800
+    },
+    {
+      "epoch": 0.19448687682266352,
+      "grad_norm": 3.8082555706484094,
+      "learning_rate": 9.307557527372362e-06,
+      "loss": 0.4062,
+      "step": 2801
+    },
+    {
+      "epoch": 0.19455631162338563,
+      "grad_norm": 4.221238460729788,
+      "learning_rate": 9.306986473447839e-06,
+      "loss": 0.3888,
+      "step": 2802
+    },
+    {
+      "epoch": 0.19462574642410777,
+      "grad_norm": 2.3523711258331708,
+      "learning_rate": 9.306415201680814e-06,
+      "loss": 0.2021,
+      "step": 2803
+    },
+    {
+      "epoch": 0.19469518122482987,
+      "grad_norm": 3.6494852125817734,
+      "learning_rate": 9.30584371210018e-06,
+      "loss": 0.5035,
+      "step": 2804
+    },
+    {
+      "epoch": 0.194764616025552,
+      "grad_norm": 5.314266633581966,
+      "learning_rate": 9.305272004734844e-06,
+      "loss": 0.6419,
+      "step": 2805
+    },
+    {
+      "epoch": 0.19483405082627414,
+      "grad_norm": 2.90490224754564,
+      "learning_rate": 9.304700079613724e-06,
+      "loss": 0.317,
+      "step": 2806
+    },
+    {
+      "epoch": 0.19490348562699625,
+      "grad_norm": 3.129894216355529,
+      "learning_rate": 9.304127936765742e-06,
+      "loss": 0.2155,
+      "step": 2807
+    },
+    {
+      "epoch": 0.19497292042771838,
+      "grad_norm": 4.419536740165044,
+      "learning_rate": 9.30355557621984e-06,
+      "loss": 0.3615,
+      "step": 2808
+    },
+    {
+      "epoch": 0.1950423552284405,
+      "grad_norm": 5.153098610474333,
+      "learning_rate": 9.302982998004969e-06,
+      "loss": 0.7054,
+      "step": 2809
+    },
+    {
+      "epoch": 0.19511179002916262,
+      "grad_norm": 3.6023131061547438,
+      "learning_rate": 9.302410202150085e-06,
+      "loss": 0.2973,
+      "step": 2810
+    },
+    {
+      "epoch": 0.19518122482988473,
+      "grad_norm": 4.036006808990755,
+      "learning_rate": 9.30183718868416e-06,
+      "loss": 0.4837,
+      "step": 2811
+    },
+    {
+      "epoch": 0.19525065963060687,
+      "grad_norm": 4.018437137749635,
+      "learning_rate": 9.30126395763618e-06,
+      "loss": 0.4471,
+      "step": 2812
+    },
+    {
+      "epoch": 0.19532009443132897,
+      "grad_norm": 3.773208313710731,
+      "learning_rate": 9.300690509035137e-06,
+      "loss": 0.4447,
+      "step": 2813
+    },
+    {
+      "epoch": 0.1953895292320511,
+      "grad_norm": 4.380705912628938,
+      "learning_rate": 9.300116842910033e-06,
+      "loss": 0.4603,
+      "step": 2814
+    },
+    {
+      "epoch": 0.1954589640327732,
+      "grad_norm": 3.7126717256336,
+      "learning_rate": 9.299542959289886e-06,
+      "loss": 0.5517,
+      "step": 2815
+    },
+    {
+      "epoch": 0.19552839883349535,
+      "grad_norm": 4.478601610383789,
+      "learning_rate": 9.29896885820372e-06,
+      "loss": 0.4721,
+      "step": 2816
+    },
+    {
+      "epoch": 0.19559783363421748,
+      "grad_norm": 4.070336578517675,
+      "learning_rate": 9.298394539680575e-06,
+      "loss": 0.4748,
+      "step": 2817
+    },
+    {
+      "epoch": 0.1956672684349396,
+      "grad_norm": 3.8628663005563113,
+      "learning_rate": 9.297820003749499e-06,
+      "loss": 0.4625,
+      "step": 2818
+    },
+    {
+      "epoch": 0.19573670323566172,
+      "grad_norm": 5.244829616893765,
+      "learning_rate": 9.297245250439548e-06,
+      "loss": 0.8626,
+      "step": 2819
+    },
+    {
+      "epoch": 0.19580613803638383,
+      "grad_norm": 3.323327387207132,
+      "learning_rate": 9.296670279779797e-06,
+      "loss": 0.5417,
+      "step": 2820
+    },
+    {
+      "epoch": 0.19587557283710597,
+      "grad_norm": 3.8631300637906065,
+      "learning_rate": 9.296095091799324e-06,
+      "loss": 0.5871,
+      "step": 2821
+    },
+    {
+      "epoch": 0.19594500763782807,
+      "grad_norm": 3.4227619997658096,
+      "learning_rate": 9.295519686527224e-06,
+      "loss": 0.4033,
+      "step": 2822
+    },
+    {
+      "epoch": 0.1960144424385502,
+      "grad_norm": 4.651250344923784,
+      "learning_rate": 9.294944063992597e-06,
+      "loss": 0.6142,
+      "step": 2823
+    },
+    {
+      "epoch": 0.19608387723927231,
+      "grad_norm": 2.965613086222905,
+      "learning_rate": 9.294368224224562e-06,
+      "loss": 0.389,
+      "step": 2824
+    },
+    {
+      "epoch": 0.19615331203999445,
+      "grad_norm": 4.593639847556587,
+      "learning_rate": 9.29379216725224e-06,
+      "loss": 0.5812,
+      "step": 2825
+    },
+    {
+      "epoch": 0.19622274684071656,
+      "grad_norm": 5.101824592425328,
+      "learning_rate": 9.293215893104767e-06,
+      "loss": 0.7988,
+      "step": 2826
+    },
+    {
+      "epoch": 0.1962921816414387,
+      "grad_norm": 3.9791459098321065,
+      "learning_rate": 9.292639401811295e-06,
+      "loss": 0.457,
+      "step": 2827
+    },
+    {
+      "epoch": 0.19636161644216082,
+      "grad_norm": 3.894962360423816,
+      "learning_rate": 9.292062693400979e-06,
+      "loss": 0.2939,
+      "step": 2828
+    },
+    {
+      "epoch": 0.19643105124288293,
+      "grad_norm": 3.8883014281747132,
+      "learning_rate": 9.291485767902988e-06,
+      "loss": 0.592,
+      "step": 2829
+    },
+    {
+      "epoch": 0.19650048604360507,
+      "grad_norm": 4.358102151108144,
+      "learning_rate": 9.290908625346505e-06,
+      "loss": 0.638,
+      "step": 2830
+    },
+    {
+      "epoch": 0.19656992084432717,
+      "grad_norm": 3.263443770950024,
+      "learning_rate": 9.290331265760717e-06,
+      "loss": 0.3062,
+      "step": 2831
+    },
+    {
+      "epoch": 0.1966393556450493,
+      "grad_norm": 3.722123562750182,
+      "learning_rate": 9.289753689174829e-06,
+      "loss": 0.3017,
+      "step": 2832
+    },
+    {
+      "epoch": 0.19670879044577141,
+      "grad_norm": 4.170869769936464,
+      "learning_rate": 9.289175895618053e-06,
+      "loss": 0.5381,
+      "step": 2833
+    },
+    {
+      "epoch": 0.19677822524649355,
+      "grad_norm": 4.056698633598476,
+      "learning_rate": 9.288597885119614e-06,
+      "loss": 0.4596,
+      "step": 2834
+    },
+    {
+      "epoch": 0.19684766004721566,
+      "grad_norm": 3.543849811122119,
+      "learning_rate": 9.288019657708748e-06,
+      "loss": 0.5461,
+      "step": 2835
+    },
+    {
+      "epoch": 0.1969170948479378,
+      "grad_norm": 2.6729923932583275,
+      "learning_rate": 9.287441213414699e-06,
+      "loss": 0.1371,
+      "step": 2836
+    },
+    {
+      "epoch": 0.1969865296486599,
+      "grad_norm": 2.998740615867098,
+      "learning_rate": 9.286862552266724e-06,
+      "loss": 0.1333,
+      "step": 2837
+    },
+    {
+      "epoch": 0.19705596444938203,
+      "grad_norm": 3.931708523755941,
+      "learning_rate": 9.286283674294094e-06,
+      "loss": 0.5478,
+      "step": 2838
+    },
+    {
+      "epoch": 0.19712539925010414,
+      "grad_norm": 3.5670240798862833,
+      "learning_rate": 9.285704579526086e-06,
+      "loss": 0.2976,
+      "step": 2839
+    },
+    {
+      "epoch": 0.19719483405082627,
+      "grad_norm": 2.5788393953324453,
+      "learning_rate": 9.285125267991989e-06,
+      "loss": 0.2778,
+      "step": 2840
+    },
+    {
+      "epoch": 0.1972642688515484,
+      "grad_norm": 3.186000849080737,
+      "learning_rate": 9.284545739721105e-06,
+      "loss": 0.3043,
+      "step": 2841
+    },
+    {
+      "epoch": 0.19733370365227051,
+      "grad_norm": 4.570872250529393,
+      "learning_rate": 9.283965994742746e-06,
+      "loss": 0.5433,
+      "step": 2842
+    },
+    {
+      "epoch": 0.19740313845299265,
+      "grad_norm": 4.073703574615141,
+      "learning_rate": 9.283386033086235e-06,
+      "loss": 0.4412,
+      "step": 2843
+    },
+    {
+      "epoch": 0.19747257325371476,
+      "grad_norm": 4.3701470762613095,
+      "learning_rate": 9.282805854780906e-06,
+      "loss": 0.3853,
+      "step": 2844
+    },
+    {
+      "epoch": 0.1975420080544369,
+      "grad_norm": 4.822016343162825,
+      "learning_rate": 9.2822254598561e-06,
+      "loss": 0.4714,
+      "step": 2845
+    },
+    {
+      "epoch": 0.197611442855159,
+      "grad_norm": 4.480888160800615,
+      "learning_rate": 9.281644848341178e-06,
+      "loss": 0.4886,
+      "step": 2846
+    },
+    {
+      "epoch": 0.19768087765588113,
+      "grad_norm": 4.967095534526796,
+      "learning_rate": 9.281064020265504e-06,
+      "loss": 0.6235,
+      "step": 2847
+    },
+    {
+      "epoch": 0.19775031245660324,
+      "grad_norm": 3.5552872352014426,
+      "learning_rate": 9.280482975658457e-06,
+      "loss": 0.5288,
+      "step": 2848
+    },
+    {
+      "epoch": 0.19781974725732537,
+      "grad_norm": 3.071121353381513,
+      "learning_rate": 9.279901714549427e-06,
+      "loss": 0.3055,
+      "step": 2849
+    },
+    {
+      "epoch": 0.19788918205804748,
+      "grad_norm": 5.192346239082116,
+      "learning_rate": 9.279320236967808e-06,
+      "loss": 0.8085,
+      "step": 2850
+    },
+    {
+      "epoch": 0.19795861685876961,
+      "grad_norm": 4.338361991320224,
+      "learning_rate": 9.278738542943014e-06,
+      "loss": 0.7138,
+      "step": 2851
+    },
+    {
+      "epoch": 0.19802805165949175,
+      "grad_norm": 5.426002678737996,
+      "learning_rate": 9.278156632504467e-06,
+      "loss": 0.6292,
+      "step": 2852
+    },
+    {
+      "epoch": 0.19809748646021386,
+      "grad_norm": 4.900933285310241,
+      "learning_rate": 9.2775745056816e-06,
+      "loss": 0.5777,
+      "step": 2853
+    },
+    {
+      "epoch": 0.198166921260936,
+      "grad_norm": 4.341100405697076,
+      "learning_rate": 9.276992162503851e-06,
+      "loss": 0.7086,
+      "step": 2854
+    },
+    {
+      "epoch": 0.1982363560616581,
+      "grad_norm": 4.712246236287635,
+      "learning_rate": 9.276409603000683e-06,
+      "loss": 0.7005,
+      "step": 2855
+    },
+    {
+      "epoch": 0.19830579086238023,
+      "grad_norm": 4.55468193033358,
+      "learning_rate": 9.275826827201553e-06,
+      "loss": 0.6057,
+      "step": 2856
+    },
+    {
+      "epoch": 0.19837522566310234,
+      "grad_norm": 3.868273286292925,
+      "learning_rate": 9.27524383513594e-06,
+      "loss": 0.4427,
+      "step": 2857
+    },
+    {
+      "epoch": 0.19844466046382447,
+      "grad_norm": 4.427906210640304,
+      "learning_rate": 9.274660626833334e-06,
+      "loss": 0.6805,
+      "step": 2858
+    },
+    {
+      "epoch": 0.19851409526454658,
+      "grad_norm": 4.440817530845498,
+      "learning_rate": 9.27407720232323e-06,
+      "loss": 0.5951,
+      "step": 2859
+    },
+    {
+      "epoch": 0.19858353006526872,
+      "grad_norm": 4.1146360643677165,
+      "learning_rate": 9.273493561635136e-06,
+      "loss": 0.4785,
+      "step": 2860
+    },
+    {
+      "epoch": 0.19865296486599082,
+      "grad_norm": 3.2975626630161465,
+      "learning_rate": 9.272909704798575e-06,
+      "loss": 0.3489,
+      "step": 2861
+    },
+    {
+      "epoch": 0.19872239966671296,
+      "grad_norm": 2.966735840236814,
+      "learning_rate": 9.272325631843076e-06,
+      "loss": 0.3953,
+      "step": 2862
+    },
+    {
+      "epoch": 0.1987918344674351,
+      "grad_norm": 5.006951664533371,
+      "learning_rate": 9.271741342798181e-06,
+      "loss": 0.6096,
+      "step": 2863
+    },
+    {
+      "epoch": 0.1988612692681572,
+      "grad_norm": 4.093955024075473,
+      "learning_rate": 9.271156837693442e-06,
+      "loss": 0.541,
+      "step": 2864
+    },
+    {
+      "epoch": 0.19893070406887933,
+      "grad_norm": 3.8209976522536717,
+      "learning_rate": 9.270572116558426e-06,
+      "loss": 0.4163,
+      "step": 2865
+    },
+    {
+      "epoch": 0.19900013886960144,
+      "grad_norm": 3.7051121713552573,
+      "learning_rate": 9.269987179422702e-06,
+      "loss": 0.5604,
+      "step": 2866
+    },
+    {
+      "epoch": 0.19906957367032357,
+      "grad_norm": 3.867872448330714,
+      "learning_rate": 9.269402026315859e-06,
+      "loss": 0.3634,
+      "step": 2867
+    },
+    {
+      "epoch": 0.19913900847104568,
+      "grad_norm": 4.560730970668807,
+      "learning_rate": 9.268816657267493e-06,
+      "loss": 0.6268,
+      "step": 2868
+    },
+    {
+      "epoch": 0.19920844327176782,
+      "grad_norm": 4.252405331903572,
+      "learning_rate": 9.26823107230721e-06,
+      "loss": 0.5711,
+      "step": 2869
+    },
+    {
+      "epoch": 0.19927787807248992,
+      "grad_norm": 4.10706863231667,
+      "learning_rate": 9.26764527146463e-06,
+      "loss": 0.3839,
+      "step": 2870
+    },
+    {
+      "epoch": 0.19934731287321206,
+      "grad_norm": 16.72678210183898,
+      "learning_rate": 9.26705925476938e-06,
+      "loss": 0.6435,
+      "step": 2871
+    },
+    {
+      "epoch": 0.19941674767393416,
+      "grad_norm": 3.2039398430779125,
+      "learning_rate": 9.266473022251103e-06,
+      "loss": 0.2513,
+      "step": 2872
+    },
+    {
+      "epoch": 0.1994861824746563,
+      "grad_norm": 4.798987822597973,
+      "learning_rate": 9.265886573939448e-06,
+      "loss": 0.6732,
+      "step": 2873
+    },
+    {
+      "epoch": 0.19955561727537843,
+      "grad_norm": 3.9599936457517213,
+      "learning_rate": 9.265299909864076e-06,
+      "loss": 0.5171,
+      "step": 2874
+    },
+    {
+      "epoch": 0.19962505207610054,
+      "grad_norm": 4.647332580827222,
+      "learning_rate": 9.264713030054664e-06,
+      "loss": 0.5541,
+      "step": 2875
+    },
+    {
+      "epoch": 0.19969448687682267,
+      "grad_norm": 4.043905424604964,
+      "learning_rate": 9.26412593454089e-06,
+      "loss": 0.7019,
+      "step": 2876
+    },
+    {
+      "epoch": 0.19976392167754478,
+      "grad_norm": 2.7922678307577247,
+      "learning_rate": 9.263538623352453e-06,
+      "loss": 0.3012,
+      "step": 2877
+    },
+    {
+      "epoch": 0.19983335647826692,
+      "grad_norm": 3.3536319431468264,
+      "learning_rate": 9.262951096519056e-06,
+      "loss": 0.2987,
+      "step": 2878
+    },
+    {
+      "epoch": 0.19990279127898902,
+      "grad_norm": 2.8726638702647964,
+      "learning_rate": 9.262363354070416e-06,
+      "loss": 0.3041,
+      "step": 2879
+    },
+    {
+      "epoch": 0.19997222607971116,
+      "grad_norm": 5.233352087729568,
+      "learning_rate": 9.261775396036262e-06,
+      "loss": 0.6618,
+      "step": 2880
+    },
+    {
+      "epoch": 0.20004166088043326,
+      "grad_norm": 4.6701881707817305,
+      "learning_rate": 9.26118722244633e-06,
+      "loss": 0.558,
+      "step": 2881
+    },
+    {
+      "epoch": 0.2001110956811554,
+      "grad_norm": 3.015451627342665,
+      "learning_rate": 9.260598833330372e-06,
+      "loss": 0.2588,
+      "step": 2882
+    },
+    {
+      "epoch": 0.2001805304818775,
+      "grad_norm": 3.995105038264002,
+      "learning_rate": 9.260010228718144e-06,
+      "loss": 0.5419,
+      "step": 2883
+    },
+    {
+      "epoch": 0.20024996528259964,
+      "grad_norm": 4.266776646372009,
+      "learning_rate": 9.259421408639421e-06,
+      "loss": 0.7082,
+      "step": 2884
+    },
+    {
+      "epoch": 0.20031940008332177,
+      "grad_norm": 5.044144064454226,
+      "learning_rate": 9.25883237312398e-06,
+      "loss": 0.5746,
+      "step": 2885
+    },
+    {
+      "epoch": 0.20038883488404388,
+      "grad_norm": 4.2600481731678865,
+      "learning_rate": 9.25824312220162e-06,
+      "loss": 0.7338,
+      "step": 2886
+    },
+    {
+      "epoch": 0.20045826968476602,
+      "grad_norm": 3.679001086251324,
+      "learning_rate": 9.25765365590214e-06,
+      "loss": 0.4332,
+      "step": 2887
+    },
+    {
+      "epoch": 0.20052770448548812,
+      "grad_norm": 4.523112344384331,
+      "learning_rate": 9.257063974255357e-06,
+      "loss": 0.4191,
+      "step": 2888
+    },
+    {
+      "epoch": 0.20059713928621026,
+      "grad_norm": 4.589641026024377,
+      "learning_rate": 9.256474077291095e-06,
+      "loss": 0.4619,
+      "step": 2889
+    },
+    {
+      "epoch": 0.20066657408693236,
+      "grad_norm": 4.753899821995365,
+      "learning_rate": 9.25588396503919e-06,
+      "loss": 0.6826,
+      "step": 2890
+    },
+    {
+      "epoch": 0.2007360088876545,
+      "grad_norm": 4.3996051461422105,
+      "learning_rate": 9.25529363752949e-06,
+      "loss": 0.6415,
+      "step": 2891
+    },
+    {
+      "epoch": 0.2008054436883766,
+      "grad_norm": 3.171732823770511,
+      "learning_rate": 9.254703094791853e-06,
+      "loss": 0.326,
+      "step": 2892
+    },
+    {
+      "epoch": 0.20087487848909874,
+      "grad_norm": 3.5869900559025067,
+      "learning_rate": 9.254112336856148e-06,
+      "loss": 0.4881,
+      "step": 2893
+    },
+    {
+      "epoch": 0.20094431328982085,
+      "grad_norm": 3.889098136905742,
+      "learning_rate": 9.253521363752254e-06,
+      "loss": 0.483,
+      "step": 2894
+    },
+    {
+      "epoch": 0.20101374809054298,
+      "grad_norm": 4.2068532276195185,
+      "learning_rate": 9.252930175510064e-06,
+      "loss": 0.6299,
+      "step": 2895
+    },
+    {
+      "epoch": 0.2010831828912651,
+      "grad_norm": 3.3606570952293953,
+      "learning_rate": 9.252338772159478e-06,
+      "loss": 0.3605,
+      "step": 2896
+    },
+    {
+      "epoch": 0.20115261769198722,
+      "grad_norm": 3.06567646335898,
+      "learning_rate": 9.251747153730409e-06,
+      "loss": 0.2652,
+      "step": 2897
+    },
+    {
+      "epoch": 0.20122205249270936,
+      "grad_norm": 4.680752764109054,
+      "learning_rate": 9.251155320252779e-06,
+      "loss": 0.6425,
+      "step": 2898
+    },
+    {
+      "epoch": 0.20129148729343146,
+      "grad_norm": 3.6014270678149205,
+      "learning_rate": 9.250563271756526e-06,
+      "loss": 0.4729,
+      "step": 2899
+    },
+    {
+      "epoch": 0.2013609220941536,
+      "grad_norm": 4.313883512704269,
+      "learning_rate": 9.24997100827159e-06,
+      "loss": 0.6282,
+      "step": 2900
+    },
+    {
+      "epoch": 0.2014303568948757,
+      "grad_norm": 3.6981417396891216,
+      "learning_rate": 9.249378529827931e-06,
+      "loss": 0.5673,
+      "step": 2901
+    },
+    {
+      "epoch": 0.20149979169559784,
+      "grad_norm": 3.5908118208917,
+      "learning_rate": 9.248785836455512e-06,
+      "loss": 0.3073,
+      "step": 2902
+    },
+    {
+      "epoch": 0.20156922649631995,
+      "grad_norm": 4.159934861635394,
+      "learning_rate": 9.248192928184315e-06,
+      "loss": 0.5134,
+      "step": 2903
+    },
+    {
+      "epoch": 0.20163866129704208,
+      "grad_norm": 4.209681562138876,
+      "learning_rate": 9.247599805044328e-06,
+      "loss": 0.4735,
+      "step": 2904
+    },
+    {
+      "epoch": 0.2017080960977642,
+      "grad_norm": 4.1803551890617605,
+      "learning_rate": 9.247006467065547e-06,
+      "loss": 0.6779,
+      "step": 2905
+    },
+    {
+      "epoch": 0.20177753089848632,
+      "grad_norm": 4.783665529861281,
+      "learning_rate": 9.246412914277986e-06,
+      "loss": 0.7115,
+      "step": 2906
+    },
+    {
+      "epoch": 0.20184696569920843,
+      "grad_norm": 4.223882427887777,
+      "learning_rate": 9.245819146711665e-06,
+      "loss": 0.5924,
+      "step": 2907
+    },
+    {
+      "epoch": 0.20191640049993057,
+      "grad_norm": 4.009003404468766,
+      "learning_rate": 9.245225164396616e-06,
+      "loss": 0.4349,
+      "step": 2908
+    },
+    {
+      "epoch": 0.2019858353006527,
+      "grad_norm": 7.910085825559944,
+      "learning_rate": 9.244630967362881e-06,
+      "loss": 0.9379,
+      "step": 2909
+    },
+    {
+      "epoch": 0.2020552701013748,
+      "grad_norm": 3.2015646197293606,
+      "learning_rate": 9.244036555640517e-06,
+      "loss": 0.4615,
+      "step": 2910
+    },
+    {
+      "epoch": 0.20212470490209694,
+      "grad_norm": 3.918230465847634,
+      "learning_rate": 9.243441929259585e-06,
+      "loss": 0.6131,
+      "step": 2911
+    },
+    {
+      "epoch": 0.20219413970281905,
+      "grad_norm": 4.217653758541043,
+      "learning_rate": 9.242847088250161e-06,
+      "loss": 0.5875,
+      "step": 2912
+    },
+    {
+      "epoch": 0.20226357450354118,
+      "grad_norm": 5.31502105223746,
+      "learning_rate": 9.242252032642334e-06,
+      "loss": 0.6769,
+      "step": 2913
+    },
+    {
+      "epoch": 0.2023330093042633,
+      "grad_norm": 4.741691150932073,
+      "learning_rate": 9.241656762466197e-06,
+      "loss": 0.5952,
+      "step": 2914
+    },
+    {
+      "epoch": 0.20240244410498542,
+      "grad_norm": 3.963954480353428,
+      "learning_rate": 9.241061277751863e-06,
+      "loss": 0.5107,
+      "step": 2915
+    },
+    {
+      "epoch": 0.20247187890570753,
+      "grad_norm": 3.944650357171821,
+      "learning_rate": 9.240465578529449e-06,
+      "loss": 0.5586,
+      "step": 2916
+    },
+    {
+      "epoch": 0.20254131370642967,
+      "grad_norm": 4.278504472374262,
+      "learning_rate": 9.239869664829084e-06,
+      "loss": 0.7028,
+      "step": 2917
+    },
+    {
+      "epoch": 0.20261074850715177,
+      "grad_norm": 4.947280904543761,
+      "learning_rate": 9.23927353668091e-06,
+      "loss": 0.8043,
+      "step": 2918
+    },
+    {
+      "epoch": 0.2026801833078739,
+      "grad_norm": 4.269798383634381,
+      "learning_rate": 9.238677194115075e-06,
+      "loss": 0.516,
+      "step": 2919
+    },
+    {
+      "epoch": 0.20274961810859604,
+      "grad_norm": 3.7163544619123883,
+      "learning_rate": 9.238080637161747e-06,
+      "loss": 0.4065,
+      "step": 2920
+    },
+    {
+      "epoch": 0.20281905290931815,
+      "grad_norm": 3.297777424087066,
+      "learning_rate": 9.237483865851093e-06,
+      "loss": 0.4312,
+      "step": 2921
+    },
+    {
+      "epoch": 0.20288848771004028,
+      "grad_norm": 3.725025719387034,
+      "learning_rate": 9.236886880213303e-06,
+      "loss": 0.4778,
+      "step": 2922
+    },
+    {
+      "epoch": 0.2029579225107624,
+      "grad_norm": 3.7811309441109793,
+      "learning_rate": 9.236289680278567e-06,
+      "loss": 0.6261,
+      "step": 2923
+    },
+    {
+      "epoch": 0.20302735731148452,
+      "grad_norm": 3.7628040791675716,
+      "learning_rate": 9.235692266077094e-06,
+      "loss": 0.3686,
+      "step": 2924
+    },
+    {
+      "epoch": 0.20309679211220663,
+      "grad_norm": 2.7595855424126223,
+      "learning_rate": 9.235094637639099e-06,
+      "loss": 0.2416,
+      "step": 2925
+    },
+    {
+      "epoch": 0.20316622691292877,
+      "grad_norm": 3.581946861493472,
+      "learning_rate": 9.234496794994809e-06,
+      "loss": 0.4834,
+      "step": 2926
+    },
+    {
+      "epoch": 0.20323566171365087,
+      "grad_norm": 3.5601245772492685,
+      "learning_rate": 9.233898738174461e-06,
+      "loss": 0.4425,
+      "step": 2927
+    },
+    {
+      "epoch": 0.203305096514373,
+      "grad_norm": 3.6285639308142783,
+      "learning_rate": 9.233300467208309e-06,
+      "loss": 0.3101,
+      "step": 2928
+    },
+    {
+      "epoch": 0.20337453131509511,
+      "grad_norm": 4.1234962126204495,
+      "learning_rate": 9.232701982126608e-06,
+      "loss": 0.5094,
+      "step": 2929
+    },
+    {
+      "epoch": 0.20344396611581725,
+      "grad_norm": 4.325504527190692,
+      "learning_rate": 9.23210328295963e-06,
+      "loss": 0.5721,
+      "step": 2930
+    },
+    {
+      "epoch": 0.20351340091653938,
+      "grad_norm": 5.669834090129811,
+      "learning_rate": 9.231504369737658e-06,
+      "loss": 0.563,
+      "step": 2931
+    },
+    {
+      "epoch": 0.2035828357172615,
+      "grad_norm": 4.328193778458343,
+      "learning_rate": 9.230905242490981e-06,
+      "loss": 0.7028,
+      "step": 2932
+    },
+    {
+      "epoch": 0.20365227051798362,
+      "grad_norm": 5.317017198970767,
+      "learning_rate": 9.230305901249907e-06,
+      "loss": 0.5494,
+      "step": 2933
+    },
+    {
+      "epoch": 0.20372170531870573,
+      "grad_norm": 4.432966571197587,
+      "learning_rate": 9.229706346044749e-06,
+      "loss": 0.6649,
+      "step": 2934
+    },
+    {
+      "epoch": 0.20379114011942787,
+      "grad_norm": 4.728486635701924,
+      "learning_rate": 9.229106576905827e-06,
+      "loss": 0.5592,
+      "step": 2935
+    },
+    {
+      "epoch": 0.20386057492014997,
+      "grad_norm": 4.555859436240858,
+      "learning_rate": 9.22850659386348e-06,
+      "loss": 0.5771,
+      "step": 2936
+    },
+    {
+      "epoch": 0.2039300097208721,
+      "grad_norm": 3.775546039879905,
+      "learning_rate": 9.227906396948055e-06,
+      "loss": 0.4177,
+      "step": 2937
+    },
+    {
+      "epoch": 0.20399944452159421,
+      "grad_norm": 5.554831271820678,
+      "learning_rate": 9.227305986189909e-06,
+      "loss": 0.6755,
+      "step": 2938
+    },
+    {
+      "epoch": 0.20406887932231635,
+      "grad_norm": 4.106718235307273,
+      "learning_rate": 9.22670536161941e-06,
+      "loss": 0.4106,
+      "step": 2939
+    },
+    {
+      "epoch": 0.20413831412303846,
+      "grad_norm": 4.431333948988815,
+      "learning_rate": 9.226104523266935e-06,
+      "loss": 0.7343,
+      "step": 2940
+    },
+    {
+      "epoch": 0.2042077489237606,
+      "grad_norm": 3.5587412526022506,
+      "learning_rate": 9.225503471162878e-06,
+      "loss": 0.435,
+      "step": 2941
+    },
+    {
+      "epoch": 0.2042771837244827,
+      "grad_norm": 3.1837939612170736,
+      "learning_rate": 9.224902205337635e-06,
+      "loss": 0.3048,
+      "step": 2942
+    },
+    {
+      "epoch": 0.20434661852520483,
+      "grad_norm": 3.292866601586631,
+      "learning_rate": 9.22430072582162e-06,
+      "loss": 0.339,
+      "step": 2943
+    },
+    {
+      "epoch": 0.20441605332592697,
+      "grad_norm": 2.679715320239534,
+      "learning_rate": 9.223699032645253e-06,
+      "loss": 0.2536,
+      "step": 2944
+    },
+    {
+      "epoch": 0.20448548812664907,
+      "grad_norm": 3.958460500456027,
+      "learning_rate": 9.223097125838969e-06,
+      "loss": 0.2439,
+      "step": 2945
+    },
+    {
+      "epoch": 0.2045549229273712,
+      "grad_norm": 3.399172119980921,
+      "learning_rate": 9.222495005433211e-06,
+      "loss": 0.3214,
+      "step": 2946
+    },
+    {
+      "epoch": 0.20462435772809331,
+      "grad_norm": 5.016057539963686,
+      "learning_rate": 9.221892671458437e-06,
+      "loss": 0.7045,
+      "step": 2947
+    },
+    {
+      "epoch": 0.20469379252881545,
+      "grad_norm": 4.262393861374148,
+      "learning_rate": 9.221290123945107e-06,
+      "loss": 0.4399,
+      "step": 2948
+    },
+    {
+      "epoch": 0.20476322732953756,
+      "grad_norm": 4.422303838711451,
+      "learning_rate": 9.220687362923696e-06,
+      "loss": 0.598,
+      "step": 2949
+    },
+    {
+      "epoch": 0.2048326621302597,
+      "grad_norm": 3.975302639975611,
+      "learning_rate": 9.220084388424698e-06,
+      "loss": 0.4963,
+      "step": 2950
+    },
+    {
+      "epoch": 0.2049020969309818,
+      "grad_norm": 4.372213926050992,
+      "learning_rate": 9.219481200478606e-06,
+      "loss": 0.4883,
+      "step": 2951
+    },
+    {
+      "epoch": 0.20497153173170393,
+      "grad_norm": 4.816523240614509,
+      "learning_rate": 9.218877799115929e-06,
+      "loss": 0.5668,
+      "step": 2952
+    },
+    {
+      "epoch": 0.20504096653242604,
+      "grad_norm": 4.5302116964549395,
+      "learning_rate": 9.218274184367187e-06,
+      "loss": 0.5612,
+      "step": 2953
+    },
+    {
+      "epoch": 0.20511040133314817,
+      "grad_norm": 3.026730972312457,
+      "learning_rate": 9.217670356262911e-06,
+      "loss": 0.4173,
+      "step": 2954
+    },
+    {
+      "epoch": 0.2051798361338703,
+      "grad_norm": 2.703287260040338,
+      "learning_rate": 9.217066314833641e-06,
+      "loss": 0.2322,
+      "step": 2955
+    },
+    {
+      "epoch": 0.20524927093459241,
+      "grad_norm": 4.015242042178952,
+      "learning_rate": 9.216462060109927e-06,
+      "loss": 0.5551,
+      "step": 2956
+    },
+    {
+      "epoch": 0.20531870573531455,
+      "grad_norm": 5.154877499724889,
+      "learning_rate": 9.215857592122334e-06,
+      "loss": 0.414,
+      "step": 2957
+    },
+    {
+      "epoch": 0.20538814053603666,
+      "grad_norm": 3.998525625006187,
+      "learning_rate": 9.215252910901436e-06,
+      "loss": 0.3237,
+      "step": 2958
+    },
+    {
+      "epoch": 0.2054575753367588,
+      "grad_norm": 4.098215810781089,
+      "learning_rate": 9.214648016477814e-06,
+      "loss": 0.4172,
+      "step": 2959
+    },
+    {
+      "epoch": 0.2055270101374809,
+      "grad_norm": 2.698607671722973,
+      "learning_rate": 9.214042908882065e-06,
+      "loss": 0.131,
+      "step": 2960
+    },
+    {
+      "epoch": 0.20559644493820303,
+      "grad_norm": 3.641773156612639,
+      "learning_rate": 9.213437588144793e-06,
+      "loss": 0.44,
+      "step": 2961
+    },
+    {
+      "epoch": 0.20566587973892514,
+      "grad_norm": 4.620562233300561,
+      "learning_rate": 9.212832054296618e-06,
+      "loss": 0.4445,
+      "step": 2962
+    },
+    {
+      "epoch": 0.20573531453964727,
+      "grad_norm": 3.7285650618439576,
+      "learning_rate": 9.212226307368164e-06,
+      "loss": 0.2866,
+      "step": 2963
+    },
+    {
+      "epoch": 0.20580474934036938,
+      "grad_norm": 1.5409635407663693,
+      "learning_rate": 9.211620347390068e-06,
+      "loss": 0.1435,
+      "step": 2964
+    },
+    {
+      "epoch": 0.20587418414109152,
+      "grad_norm": 4.431454260697495,
+      "learning_rate": 9.211014174392983e-06,
+      "loss": 0.5389,
+      "step": 2965
+    },
+    {
+      "epoch": 0.20594361894181365,
+      "grad_norm": 4.088549974359027,
+      "learning_rate": 9.210407788407564e-06,
+      "loss": 0.5235,
+      "step": 2966
+    },
+    {
+      "epoch": 0.20601305374253576,
+      "grad_norm": 3.6081366578461123,
+      "learning_rate": 9.209801189464486e-06,
+      "loss": 0.3829,
+      "step": 2967
+    },
+    {
+      "epoch": 0.2060824885432579,
+      "grad_norm": 5.040971164870507,
+      "learning_rate": 9.209194377594427e-06,
+      "loss": 0.5671,
+      "step": 2968
+    },
+    {
+      "epoch": 0.20615192334398,
+      "grad_norm": 5.876744985494463,
+      "learning_rate": 9.208587352828077e-06,
+      "loss": 0.3555,
+      "step": 2969
+    },
+    {
+      "epoch": 0.20622135814470213,
+      "grad_norm": 2.9949444063169133,
+      "learning_rate": 9.207980115196145e-06,
+      "loss": 0.2446,
+      "step": 2970
+    },
+    {
+      "epoch": 0.20629079294542424,
+      "grad_norm": 3.3200413999387512,
+      "learning_rate": 9.207372664729337e-06,
+      "loss": 0.3733,
+      "step": 2971
+    },
+    {
+      "epoch": 0.20636022774614637,
+      "grad_norm": 2.1710359423043917,
+      "learning_rate": 9.206765001458381e-06,
+      "loss": 0.2332,
+      "step": 2972
+    },
+    {
+      "epoch": 0.20642966254686848,
+      "grad_norm": 3.6244245685484553,
+      "learning_rate": 9.206157125414013e-06,
+      "loss": 0.6052,
+      "step": 2973
+    },
+    {
+      "epoch": 0.20649909734759062,
+      "grad_norm": 3.5936789816901316,
+      "learning_rate": 9.205549036626976e-06,
+      "loss": 0.4288,
+      "step": 2974
+    },
+    {
+      "epoch": 0.20656853214831272,
+      "grad_norm": 3.34349110424724,
+      "learning_rate": 9.204940735128028e-06,
+      "loss": 0.375,
+      "step": 2975
+    },
+    {
+      "epoch": 0.20663796694903486,
+      "grad_norm": 4.007977871492753,
+      "learning_rate": 9.204332220947936e-06,
+      "loss": 0.4116,
+      "step": 2976
+    },
+    {
+      "epoch": 0.206707401749757,
+      "grad_norm": 5.015516193720728,
+      "learning_rate": 9.203723494117479e-06,
+      "loss": 0.8886,
+      "step": 2977
+    },
+    {
+      "epoch": 0.2067768365504791,
+      "grad_norm": 3.992597594019516,
+      "learning_rate": 9.203114554667442e-06,
+      "loss": 0.3354,
+      "step": 2978
+    },
+    {
+      "epoch": 0.20684627135120123,
+      "grad_norm": 3.3863435882181934,
+      "learning_rate": 9.202505402628628e-06,
+      "loss": 0.4754,
+      "step": 2979
+    },
+    {
+      "epoch": 0.20691570615192334,
+      "grad_norm": 4.859684467840073,
+      "learning_rate": 9.201896038031847e-06,
+      "loss": 0.6911,
+      "step": 2980
+    },
+    {
+      "epoch": 0.20698514095264547,
+      "grad_norm": 3.8974634220770175,
+      "learning_rate": 9.201286460907918e-06,
+      "loss": 0.5883,
+      "step": 2981
+    },
+    {
+      "epoch": 0.20705457575336758,
+      "grad_norm": 4.202218565871006,
+      "learning_rate": 9.200676671287674e-06,
+      "loss": 0.5944,
+      "step": 2982
+    },
+    {
+      "epoch": 0.20712401055408972,
+      "grad_norm": 4.404358746071678,
+      "learning_rate": 9.200066669201958e-06,
+      "loss": 0.6628,
+      "step": 2983
+    },
+    {
+      "epoch": 0.20719344535481182,
+      "grad_norm": 4.257972988606241,
+      "learning_rate": 9.199456454681623e-06,
+      "loss": 0.5393,
+      "step": 2984
+    },
+    {
+      "epoch": 0.20726288015553396,
+      "grad_norm": 3.845464549637663,
+      "learning_rate": 9.19884602775753e-06,
+      "loss": 0.4122,
+      "step": 2985
+    },
+    {
+      "epoch": 0.20733231495625606,
+      "grad_norm": 4.379398247964277,
+      "learning_rate": 9.198235388460558e-06,
+      "loss": 0.6378,
+      "step": 2986
+    },
+    {
+      "epoch": 0.2074017497569782,
+      "grad_norm": 3.235536582148886,
+      "learning_rate": 9.197624536821591e-06,
+      "loss": 0.2484,
+      "step": 2987
+    },
+    {
+      "epoch": 0.2074711845577003,
+      "grad_norm": 4.0465118409422205,
+      "learning_rate": 9.197013472871526e-06,
+      "loss": 0.3766,
+      "step": 2988
+    },
+    {
+      "epoch": 0.20754061935842244,
+      "grad_norm": 4.190572616191252,
+      "learning_rate": 9.196402196641268e-06,
+      "loss": 0.3946,
+      "step": 2989
+    },
+    {
+      "epoch": 0.20761005415914457,
+      "grad_norm": 2.4794056426187585,
+      "learning_rate": 9.195790708161733e-06,
+      "loss": 0.2265,
+      "step": 2990
+    },
+    {
+      "epoch": 0.20767948895986668,
+      "grad_norm": 4.432359829642248,
+      "learning_rate": 9.195179007463855e-06,
+      "loss": 0.69,
+      "step": 2991
+    },
+    {
+      "epoch": 0.20774892376058882,
+      "grad_norm": 3.8711637191924084,
+      "learning_rate": 9.194567094578569e-06,
+      "loss": 0.3126,
+      "step": 2992
+    },
+    {
+      "epoch": 0.20781835856131092,
+      "grad_norm": 4.510346441785111,
+      "learning_rate": 9.193954969536825e-06,
+      "loss": 0.6388,
+      "step": 2993
+    },
+    {
+      "epoch": 0.20788779336203306,
+      "grad_norm": 4.696288262956791,
+      "learning_rate": 9.193342632369584e-06,
+      "loss": 0.6456,
+      "step": 2994
+    },
+    {
+      "epoch": 0.20795722816275516,
+      "grad_norm": 3.9114869935332957,
+      "learning_rate": 9.19273008310782e-06,
+      "loss": 0.3599,
+      "step": 2995
+    },
+    {
+      "epoch": 0.2080266629634773,
+      "grad_norm": 3.991378834077295,
+      "learning_rate": 9.19211732178251e-06,
+      "loss": 0.4249,
+      "step": 2996
+    },
+    {
+      "epoch": 0.2080960977641994,
+      "grad_norm": 3.935206347909007,
+      "learning_rate": 9.19150434842465e-06,
+      "loss": 0.4334,
+      "step": 2997
+    },
+    {
+      "epoch": 0.20816553256492154,
+      "grad_norm": 4.179782509145216,
+      "learning_rate": 9.190891163065246e-06,
+      "loss": 0.5651,
+      "step": 2998
+    },
+    {
+      "epoch": 0.20823496736564365,
+      "grad_norm": 4.461541853580165,
+      "learning_rate": 9.190277765735307e-06,
+      "loss": 0.6242,
+      "step": 2999
+    },
+    {
+      "epoch": 0.20830440216636578,
+      "grad_norm": 3.787464471347851,
+      "learning_rate": 9.18966415646586e-06,
+      "loss": 0.4826,
+      "step": 3000
+    },
+    {
+      "epoch": 0.20837383696708792,
+      "grad_norm": 4.996412982875818,
+      "learning_rate": 9.189050335287941e-06,
+      "loss": 0.7873,
+      "step": 3001
+    },
+    {
+      "epoch": 0.20844327176781002,
+      "grad_norm": 3.810718241960916,
+      "learning_rate": 9.188436302232598e-06,
+      "loss": 0.4396,
+      "step": 3002
+    },
+    {
+      "epoch": 0.20851270656853216,
+      "grad_norm": 2.3939946214001777,
+      "learning_rate": 9.187822057330883e-06,
+      "loss": 0.0925,
+      "step": 3003
+    },
+    {
+      "epoch": 0.20858214136925426,
+      "grad_norm": 3.49864888648758,
+      "learning_rate": 9.18720760061387e-06,
+      "loss": 0.3737,
+      "step": 3004
+    },
+    {
+      "epoch": 0.2086515761699764,
+      "grad_norm": 3.376209598562402,
+      "learning_rate": 9.186592932112634e-06,
+      "loss": 0.422,
+      "step": 3005
+    },
+    {
+      "epoch": 0.2087210109706985,
+      "grad_norm": 3.718660065251022,
+      "learning_rate": 9.185978051858265e-06,
+      "loss": 0.4118,
+      "step": 3006
+    },
+    {
+      "epoch": 0.20879044577142064,
+      "grad_norm": 3.945956979543894,
+      "learning_rate": 9.185362959881862e-06,
+      "loss": 0.5775,
+      "step": 3007
+    },
+    {
+      "epoch": 0.20885988057214275,
+      "grad_norm": 4.461495896998751,
+      "learning_rate": 9.18474765621454e-06,
+      "loss": 0.4672,
+      "step": 3008
+    },
+    {
+      "epoch": 0.20892931537286488,
+      "grad_norm": 3.595031843733969,
+      "learning_rate": 9.184132140887414e-06,
+      "loss": 0.404,
+      "step": 3009
+    },
+    {
+      "epoch": 0.208998750173587,
+      "grad_norm": 3.8326058447841085,
+      "learning_rate": 9.183516413931618e-06,
+      "loss": 0.4135,
+      "step": 3010
+    },
+    {
+      "epoch": 0.20906818497430912,
+      "grad_norm": 7.490574766918556,
+      "learning_rate": 9.182900475378297e-06,
+      "loss": 0.4537,
+      "step": 3011
+    },
+    {
+      "epoch": 0.20913761977503126,
+      "grad_norm": 4.996906096220394,
+      "learning_rate": 9.182284325258603e-06,
+      "loss": 0.5429,
+      "step": 3012
+    },
+    {
+      "epoch": 0.20920705457575337,
+      "grad_norm": 4.987454899082449,
+      "learning_rate": 9.1816679636037e-06,
+      "loss": 0.8081,
+      "step": 3013
+    },
+    {
+      "epoch": 0.2092764893764755,
+      "grad_norm": 3.253770813941013,
+      "learning_rate": 9.181051390444765e-06,
+      "loss": 0.2422,
+      "step": 3014
+    },
+    {
+      "epoch": 0.2093459241771976,
+      "grad_norm": 5.243902509584505,
+      "learning_rate": 9.18043460581298e-06,
+      "loss": 0.6212,
+      "step": 3015
+    },
+    {
+      "epoch": 0.20941535897791974,
+      "grad_norm": 3.4425282557283547,
+      "learning_rate": 9.179817609739545e-06,
+      "loss": 0.431,
+      "step": 3016
+    },
+    {
+      "epoch": 0.20948479377864185,
+      "grad_norm": 4.4298823483233924,
+      "learning_rate": 9.179200402255664e-06,
+      "loss": 0.5552,
+      "step": 3017
+    },
+    {
+      "epoch": 0.20955422857936398,
+      "grad_norm": 4.905425205483724,
+      "learning_rate": 9.178582983392556e-06,
+      "loss": 0.7432,
+      "step": 3018
+    },
+    {
+      "epoch": 0.2096236633800861,
+      "grad_norm": 4.692332724087887,
+      "learning_rate": 9.177965353181448e-06,
+      "loss": 0.6348,
+      "step": 3019
+    },
+    {
+      "epoch": 0.20969309818080822,
+      "grad_norm": 3.8306559690106794,
+      "learning_rate": 9.17734751165358e-06,
+      "loss": 0.3818,
+      "step": 3020
+    },
+    {
+      "epoch": 0.20976253298153033,
+      "grad_norm": 4.629743094497733,
+      "learning_rate": 9.176729458840203e-06,
+      "loss": 0.4196,
+      "step": 3021
+    },
+    {
+      "epoch": 0.20983196778225247,
+      "grad_norm": 4.200183941718307,
+      "learning_rate": 9.176111194772577e-06,
+      "loss": 0.7091,
+      "step": 3022
+    },
+    {
+      "epoch": 0.2099014025829746,
+      "grad_norm": 3.629942279609985,
+      "learning_rate": 9.175492719481971e-06,
+      "loss": 0.5527,
+      "step": 3023
+    },
+    {
+      "epoch": 0.2099708373836967,
+      "grad_norm": 4.2507857358365975,
+      "learning_rate": 9.174874032999668e-06,
+      "loss": 0.4814,
+      "step": 3024
+    },
+    {
+      "epoch": 0.21004027218441884,
+      "grad_norm": 5.183896950097042,
+      "learning_rate": 9.174255135356963e-06,
+      "loss": 0.7816,
+      "step": 3025
+    },
+    {
+      "epoch": 0.21010970698514095,
+      "grad_norm": 4.695165086044882,
+      "learning_rate": 9.173636026585154e-06,
+      "loss": 0.8463,
+      "step": 3026
+    },
+    {
+      "epoch": 0.21017914178586308,
+      "grad_norm": 4.080712897436878,
+      "learning_rate": 9.173016706715559e-06,
+      "loss": 0.4484,
+      "step": 3027
+    },
+    {
+      "epoch": 0.2102485765865852,
+      "grad_norm": 3.8730389445143376,
+      "learning_rate": 9.1723971757795e-06,
+      "loss": 0.4386,
+      "step": 3028
+    },
+    {
+      "epoch": 0.21031801138730732,
+      "grad_norm": 5.03105743399599,
+      "learning_rate": 9.171777433808313e-06,
+      "loss": 0.8397,
+      "step": 3029
+    },
+    {
+      "epoch": 0.21038744618802943,
+      "grad_norm": 5.116355636263747,
+      "learning_rate": 9.171157480833345e-06,
+      "loss": 0.6534,
+      "step": 3030
+    },
+    {
+      "epoch": 0.21045688098875157,
+      "grad_norm": 2.6440019186833057,
+      "learning_rate": 9.170537316885953e-06,
+      "loss": 0.4631,
+      "step": 3031
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 4.02988983940807,
+      "learning_rate": 9.1699169419975e-06,
+      "loss": 0.4753,
+      "step": 3032
+    },
+    {
+      "epoch": 0.2105957505901958,
+      "grad_norm": 3.8867657627818892,
+      "learning_rate": 9.169296356199368e-06,
+      "loss": 0.6021,
+      "step": 3033
+    },
+    {
+      "epoch": 0.21066518539091794,
+      "grad_norm": 4.0898968108518,
+      "learning_rate": 9.168675559522944e-06,
+      "loss": 0.4906,
+      "step": 3034
+    },
+    {
+      "epoch": 0.21073462019164005,
+      "grad_norm": 3.7192952105480126,
+      "learning_rate": 9.168054551999627e-06,
+      "loss": 0.5023,
+      "step": 3035
+    },
+    {
+      "epoch": 0.21080405499236218,
+      "grad_norm": 4.965968860732031,
+      "learning_rate": 9.167433333660828e-06,
+      "loss": 0.9098,
+      "step": 3036
+    },
+    {
+      "epoch": 0.2108734897930843,
+      "grad_norm": 3.881113663567508,
+      "learning_rate": 9.166811904537967e-06,
+      "loss": 0.5022,
+      "step": 3037
+    },
+    {
+      "epoch": 0.21094292459380642,
+      "grad_norm": 4.284229542078125,
+      "learning_rate": 9.166190264662473e-06,
+      "loss": 0.4457,
+      "step": 3038
+    },
+    {
+      "epoch": 0.21101235939452853,
+      "grad_norm": 3.9295783925446393,
+      "learning_rate": 9.165568414065792e-06,
+      "loss": 0.4882,
+      "step": 3039
+    },
+    {
+      "epoch": 0.21108179419525067,
+      "grad_norm": 4.216564450240247,
+      "learning_rate": 9.164946352779373e-06,
+      "loss": 0.6646,
+      "step": 3040
+    },
+    {
+      "epoch": 0.21115122899597277,
+      "grad_norm": 4.130812573234921,
+      "learning_rate": 9.16432408083468e-06,
+      "loss": 0.4314,
+      "step": 3041
+    },
+    {
+      "epoch": 0.2112206637966949,
+      "grad_norm": 3.8034464243990493,
+      "learning_rate": 9.163701598263188e-06,
+      "loss": 0.3907,
+      "step": 3042
+    },
+    {
+      "epoch": 0.21129009859741701,
+      "grad_norm": 4.298404027631847,
+      "learning_rate": 9.16307890509638e-06,
+      "loss": 0.5337,
+      "step": 3043
+    },
+    {
+      "epoch": 0.21135953339813915,
+      "grad_norm": 3.7488555212177417,
+      "learning_rate": 9.162456001365753e-06,
+      "loss": 0.5129,
+      "step": 3044
+    },
+    {
+      "epoch": 0.21142896819886126,
+      "grad_norm": 3.5679628953252567,
+      "learning_rate": 9.161832887102813e-06,
+      "loss": 0.7084,
+      "step": 3045
+    },
+    {
+      "epoch": 0.2114984029995834,
+      "grad_norm": 3.591279135741178,
+      "learning_rate": 9.161209562339073e-06,
+      "loss": 0.3695,
+      "step": 3046
+    },
+    {
+      "epoch": 0.21156783780030552,
+      "grad_norm": 4.072850313228122,
+      "learning_rate": 9.160586027106062e-06,
+      "loss": 0.5386,
+      "step": 3047
+    },
+    {
+      "epoch": 0.21163727260102763,
+      "grad_norm": 3.42282009376979,
+      "learning_rate": 9.159962281435318e-06,
+      "loss": 0.3228,
+      "step": 3048
+    },
+    {
+      "epoch": 0.21170670740174977,
+      "grad_norm": 4.462195846448209,
+      "learning_rate": 9.159338325358389e-06,
+      "loss": 0.5386,
+      "step": 3049
+    },
+    {
+      "epoch": 0.21177614220247187,
+      "grad_norm": 4.440935303686969,
+      "learning_rate": 9.158714158906834e-06,
+      "loss": 0.6398,
+      "step": 3050
+    },
+    {
+      "epoch": 0.211845577003194,
+      "grad_norm": 3.676460675767309,
+      "learning_rate": 9.158089782112223e-06,
+      "loss": 0.4207,
+      "step": 3051
+    },
+    {
+      "epoch": 0.21191501180391611,
+      "grad_norm": 4.759515778061281,
+      "learning_rate": 9.157465195006137e-06,
+      "loss": 0.6801,
+      "step": 3052
+    },
+    {
+      "epoch": 0.21198444660463825,
+      "grad_norm": 2.9339254048991856,
+      "learning_rate": 9.156840397620164e-06,
+      "loss": 0.1921,
+      "step": 3053
+    },
+    {
+      "epoch": 0.21205388140536036,
+      "grad_norm": 4.113335174822009,
+      "learning_rate": 9.15621538998591e-06,
+      "loss": 0.5049,
+      "step": 3054
+    },
+    {
+      "epoch": 0.2121233162060825,
+      "grad_norm": 3.421880643667943,
+      "learning_rate": 9.155590172134983e-06,
+      "loss": 0.3204,
+      "step": 3055
+    },
+    {
+      "epoch": 0.2121927510068046,
+      "grad_norm": 4.970850030457273,
+      "learning_rate": 9.154964744099006e-06,
+      "loss": 0.6536,
+      "step": 3056
+    },
+    {
+      "epoch": 0.21226218580752673,
+      "grad_norm": 4.107373472600804,
+      "learning_rate": 9.154339105909615e-06,
+      "loss": 0.6691,
+      "step": 3057
+    },
+    {
+      "epoch": 0.21233162060824887,
+      "grad_norm": 3.1683352666123916,
+      "learning_rate": 9.153713257598453e-06,
+      "loss": 0.4187,
+      "step": 3058
+    },
+    {
+      "epoch": 0.21240105540897097,
+      "grad_norm": 3.2769558879244687,
+      "learning_rate": 9.153087199197175e-06,
+      "loss": 0.5633,
+      "step": 3059
+    },
+    {
+      "epoch": 0.2124704902096931,
+      "grad_norm": 3.9345365387921034,
+      "learning_rate": 9.152460930737448e-06,
+      "loss": 0.2962,
+      "step": 3060
+    },
+    {
+      "epoch": 0.21253992501041521,
+      "grad_norm": 3.703587921502508,
+      "learning_rate": 9.151834452250942e-06,
+      "loss": 0.3698,
+      "step": 3061
+    },
+    {
+      "epoch": 0.21260935981113735,
+      "grad_norm": 3.732968760068661,
+      "learning_rate": 9.151207763769352e-06,
+      "loss": 0.3339,
+      "step": 3062
+    },
+    {
+      "epoch": 0.21267879461185946,
+      "grad_norm": 4.56938222565007,
+      "learning_rate": 9.150580865324368e-06,
+      "loss": 0.4469,
+      "step": 3063
+    },
+    {
+      "epoch": 0.2127482294125816,
+      "grad_norm": 5.158801847966025,
+      "learning_rate": 9.1499537569477e-06,
+      "loss": 0.8767,
+      "step": 3064
+    },
+    {
+      "epoch": 0.2128176642133037,
+      "grad_norm": 4.1474335913631855,
+      "learning_rate": 9.149326438671067e-06,
+      "loss": 0.4557,
+      "step": 3065
+    },
+    {
+      "epoch": 0.21288709901402583,
+      "grad_norm": 4.340886894224082,
+      "learning_rate": 9.148698910526199e-06,
+      "loss": 0.8676,
+      "step": 3066
+    },
+    {
+      "epoch": 0.21295653381474794,
+      "grad_norm": 3.68725084153512,
+      "learning_rate": 9.148071172544835e-06,
+      "loss": 0.5116,
+      "step": 3067
+    },
+    {
+      "epoch": 0.21302596861547007,
+      "grad_norm": 3.8678573821943125,
+      "learning_rate": 9.147443224758725e-06,
+      "loss": 0.311,
+      "step": 3068
+    },
+    {
+      "epoch": 0.2130954034161922,
+      "grad_norm": 2.927458804768907,
+      "learning_rate": 9.146815067199628e-06,
+      "loss": 0.4113,
+      "step": 3069
+    },
+    {
+      "epoch": 0.21316483821691432,
+      "grad_norm": 3.3600972357085057,
+      "learning_rate": 9.14618669989932e-06,
+      "loss": 0.3246,
+      "step": 3070
+    },
+    {
+      "epoch": 0.21323427301763645,
+      "grad_norm": 5.328649073457586,
+      "learning_rate": 9.145558122889579e-06,
+      "loss": 0.4158,
+      "step": 3071
+    },
+    {
+      "epoch": 0.21330370781835856,
+      "grad_norm": 3.54143183878089,
+      "learning_rate": 9.144929336202199e-06,
+      "loss": 0.4415,
+      "step": 3072
+    },
+    {
+      "epoch": 0.2133731426190807,
+      "grad_norm": 3.839025124560028,
+      "learning_rate": 9.144300339868983e-06,
+      "loss": 0.5091,
+      "step": 3073
+    },
+    {
+      "epoch": 0.2134425774198028,
+      "grad_norm": 3.3880694412616466,
+      "learning_rate": 9.143671133921746e-06,
+      "loss": 0.3608,
+      "step": 3074
+    },
+    {
+      "epoch": 0.21351201222052493,
+      "grad_norm": 4.5241255698422425,
+      "learning_rate": 9.143041718392313e-06,
+      "loss": 0.5161,
+      "step": 3075
+    },
+    {
+      "epoch": 0.21358144702124704,
+      "grad_norm": 5.692720021469452,
+      "learning_rate": 9.142412093312516e-06,
+      "loss": 0.658,
+      "step": 3076
+    },
+    {
+      "epoch": 0.21365088182196917,
+      "grad_norm": 5.132239433808138,
+      "learning_rate": 9.141782258714205e-06,
+      "loss": 0.5712,
+      "step": 3077
+    },
+    {
+      "epoch": 0.21372031662269128,
+      "grad_norm": 4.335707092185283,
+      "learning_rate": 9.141152214629231e-06,
+      "loss": 0.6008,
+      "step": 3078
+    },
+    {
+      "epoch": 0.21378975142341342,
+      "grad_norm": 3.1903193374390875,
+      "learning_rate": 9.140521961089468e-06,
+      "loss": 0.3771,
+      "step": 3079
+    },
+    {
+      "epoch": 0.21385918622413555,
+      "grad_norm": 4.3278411634058145,
+      "learning_rate": 9.139891498126787e-06,
+      "loss": 0.5265,
+      "step": 3080
+    },
+    {
+      "epoch": 0.21392862102485766,
+      "grad_norm": 4.087571064325811,
+      "learning_rate": 9.139260825773079e-06,
+      "loss": 0.4189,
+      "step": 3081
+    },
+    {
+      "epoch": 0.2139980558255798,
+      "grad_norm": 4.4525788973645035,
+      "learning_rate": 9.138629944060243e-06,
+      "loss": 0.5436,
+      "step": 3082
+    },
+    {
+      "epoch": 0.2140674906263019,
+      "grad_norm": 3.447727968723044,
+      "learning_rate": 9.137998853020187e-06,
+      "loss": 0.3647,
+      "step": 3083
+    },
+    {
+      "epoch": 0.21413692542702403,
+      "grad_norm": 5.936355902588573,
+      "learning_rate": 9.13736755268483e-06,
+      "loss": 0.6887,
+      "step": 3084
+    },
+    {
+      "epoch": 0.21420636022774614,
+      "grad_norm": 3.8405917051082423,
+      "learning_rate": 9.136736043086107e-06,
+      "loss": 0.3461,
+      "step": 3085
+    },
+    {
+      "epoch": 0.21427579502846827,
+      "grad_norm": 3.319732761861475,
+      "learning_rate": 9.136104324255955e-06,
+      "loss": 0.4738,
+      "step": 3086
+    },
+    {
+      "epoch": 0.21434522982919038,
+      "grad_norm": 4.85617079122543,
+      "learning_rate": 9.135472396226325e-06,
+      "loss": 0.6516,
+      "step": 3087
+    },
+    {
+      "epoch": 0.21441466462991252,
+      "grad_norm": 4.511568375550234,
+      "learning_rate": 9.134840259029183e-06,
+      "loss": 0.426,
+      "step": 3088
+    },
+    {
+      "epoch": 0.21448409943063462,
+      "grad_norm": 3.18494414858035,
+      "learning_rate": 9.134207912696497e-06,
+      "loss": 0.4717,
+      "step": 3089
+    },
+    {
+      "epoch": 0.21455353423135676,
+      "grad_norm": 4.008745512177627,
+      "learning_rate": 9.133575357260254e-06,
+      "loss": 0.4918,
+      "step": 3090
+    },
+    {
+      "epoch": 0.21462296903207886,
+      "grad_norm": 5.293775798695443,
+      "learning_rate": 9.132942592752447e-06,
+      "loss": 0.6368,
+      "step": 3091
+    },
+    {
+      "epoch": 0.214692403832801,
+      "grad_norm": 4.942373543541338,
+      "learning_rate": 9.132309619205081e-06,
+      "loss": 0.5674,
+      "step": 3092
+    },
+    {
+      "epoch": 0.21476183863352313,
+      "grad_norm": 5.116525949864272,
+      "learning_rate": 9.13167643665017e-06,
+      "loss": 0.79,
+      "step": 3093
+    },
+    {
+      "epoch": 0.21483127343424524,
+      "grad_norm": 3.03424408274422,
+      "learning_rate": 9.131043045119741e-06,
+      "loss": 0.4294,
+      "step": 3094
+    },
+    {
+      "epoch": 0.21490070823496737,
+      "grad_norm": 3.7602860195400267,
+      "learning_rate": 9.130409444645829e-06,
+      "loss": 0.4811,
+      "step": 3095
+    },
+    {
+      "epoch": 0.21497014303568948,
+      "grad_norm": 3.2834263800866617,
+      "learning_rate": 9.12977563526048e-06,
+      "loss": 0.3168,
+      "step": 3096
+    },
+    {
+      "epoch": 0.21503957783641162,
+      "grad_norm": 4.3501943240532155,
+      "learning_rate": 9.129141616995754e-06,
+      "loss": 0.5386,
+      "step": 3097
+    },
+    {
+      "epoch": 0.21510901263713372,
+      "grad_norm": 4.596567151953131,
+      "learning_rate": 9.128507389883716e-06,
+      "loss": 0.6332,
+      "step": 3098
+    },
+    {
+      "epoch": 0.21517844743785586,
+      "grad_norm": 5.802297573056446,
+      "learning_rate": 9.127872953956447e-06,
+      "loss": 0.9454,
+      "step": 3099
+    },
+    {
+      "epoch": 0.21524788223857796,
+      "grad_norm": 4.453147883072212,
+      "learning_rate": 9.127238309246036e-06,
+      "loss": 0.5962,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2153173170393001,
+      "grad_norm": 2.921091746462143,
+      "learning_rate": 9.12660345578458e-06,
+      "loss": 0.4112,
+      "step": 3101
+    },
+    {
+      "epoch": 0.2153867518400222,
+      "grad_norm": 3.4713130535078305,
+      "learning_rate": 9.12596839360419e-06,
+      "loss": 0.3862,
+      "step": 3102
+    },
+    {
+      "epoch": 0.21545618664074434,
+      "grad_norm": 3.73070738233506,
+      "learning_rate": 9.125333122736987e-06,
+      "loss": 0.4481,
+      "step": 3103
+    },
+    {
+      "epoch": 0.21552562144146648,
+      "grad_norm": 4.043128920618464,
+      "learning_rate": 9.124697643215105e-06,
+      "loss": 0.6725,
+      "step": 3104
+    },
+    {
+      "epoch": 0.21559505624218858,
+      "grad_norm": 3.937878555945242,
+      "learning_rate": 9.124061955070685e-06,
+      "loss": 0.5313,
+      "step": 3105
+    },
+    {
+      "epoch": 0.21566449104291072,
+      "grad_norm": 4.535071609588837,
+      "learning_rate": 9.123426058335877e-06,
+      "loss": 0.8052,
+      "step": 3106
+    },
+    {
+      "epoch": 0.21573392584363282,
+      "grad_norm": 4.535369551654566,
+      "learning_rate": 9.122789953042843e-06,
+      "loss": 0.613,
+      "step": 3107
+    },
+    {
+      "epoch": 0.21580336064435496,
+      "grad_norm": 5.094929868553323,
+      "learning_rate": 9.122153639223758e-06,
+      "loss": 0.651,
+      "step": 3108
+    },
+    {
+      "epoch": 0.21587279544507706,
+      "grad_norm": 2.8565849666850784,
+      "learning_rate": 9.121517116910808e-06,
+      "loss": 0.2055,
+      "step": 3109
+    },
+    {
+      "epoch": 0.2159422302457992,
+      "grad_norm": 4.212783117165023,
+      "learning_rate": 9.120880386136186e-06,
+      "loss": 0.7382,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2160116650465213,
+      "grad_norm": 4.077992027163121,
+      "learning_rate": 9.120243446932099e-06,
+      "loss": 0.3683,
+      "step": 3111
+    },
+    {
+      "epoch": 0.21608109984724344,
+      "grad_norm": 3.6622676370870297,
+      "learning_rate": 9.119606299330757e-06,
+      "loss": 0.3335,
+      "step": 3112
+    },
+    {
+      "epoch": 0.21615053464796555,
+      "grad_norm": 4.509152574725696,
+      "learning_rate": 9.118968943364393e-06,
+      "loss": 0.4701,
+      "step": 3113
+    },
+    {
+      "epoch": 0.21621996944868768,
+      "grad_norm": 3.822939585896521,
+      "learning_rate": 9.11833137906524e-06,
+      "loss": 0.4954,
+      "step": 3114
+    },
+    {
+      "epoch": 0.21628940424940982,
+      "grad_norm": 4.630548226738602,
+      "learning_rate": 9.117693606465544e-06,
+      "loss": 0.6423,
+      "step": 3115
+    },
+    {
+      "epoch": 0.21635883905013192,
+      "grad_norm": 3.615145934213717,
+      "learning_rate": 9.117055625597567e-06,
+      "loss": 0.3944,
+      "step": 3116
+    },
+    {
+      "epoch": 0.21642827385085406,
+      "grad_norm": 4.231738578002882,
+      "learning_rate": 9.116417436493574e-06,
+      "loss": 0.4536,
+      "step": 3117
+    },
+    {
+      "epoch": 0.21649770865157617,
+      "grad_norm": 4.484160279718635,
+      "learning_rate": 9.115779039185847e-06,
+      "loss": 0.5683,
+      "step": 3118
+    },
+    {
+      "epoch": 0.2165671434522983,
+      "grad_norm": 3.3916200399309084,
+      "learning_rate": 9.115140433706673e-06,
+      "loss": 0.4321,
+      "step": 3119
+    },
+    {
+      "epoch": 0.2166365782530204,
+      "grad_norm": 5.573636579364532,
+      "learning_rate": 9.114501620088352e-06,
+      "loss": 0.9509,
+      "step": 3120
+    },
+    {
+      "epoch": 0.21670601305374254,
+      "grad_norm": 4.056580604098502,
+      "learning_rate": 9.113862598363195e-06,
+      "loss": 0.4642,
+      "step": 3121
+    },
+    {
+      "epoch": 0.21677544785446465,
+      "grad_norm": 4.07908743687115,
+      "learning_rate": 9.113223368563521e-06,
+      "loss": 0.6822,
+      "step": 3122
+    },
+    {
+      "epoch": 0.21684488265518678,
+      "grad_norm": 2.770760683959282,
+      "learning_rate": 9.112583930721666e-06,
+      "loss": 0.3059,
+      "step": 3123
+    },
+    {
+      "epoch": 0.2169143174559089,
+      "grad_norm": 3.3734619221312214,
+      "learning_rate": 9.111944284869969e-06,
+      "loss": 0.2395,
+      "step": 3124
+    },
+    {
+      "epoch": 0.21698375225663102,
+      "grad_norm": 4.299023993253073,
+      "learning_rate": 9.111304431040782e-06,
+      "loss": 0.5411,
+      "step": 3125
+    },
+    {
+      "epoch": 0.21705318705735316,
+      "grad_norm": 4.213144221304634,
+      "learning_rate": 9.11066436926647e-06,
+      "loss": 0.7102,
+      "step": 3126
+    },
+    {
+      "epoch": 0.21712262185807527,
+      "grad_norm": 3.725235422971617,
+      "learning_rate": 9.110024099579407e-06,
+      "loss": 0.3095,
+      "step": 3127
+    },
+    {
+      "epoch": 0.2171920566587974,
+      "grad_norm": 3.0490560910198083,
+      "learning_rate": 9.109383622011973e-06,
+      "loss": 0.4276,
+      "step": 3128
+    },
+    {
+      "epoch": 0.2172614914595195,
+      "grad_norm": 4.351720996469051,
+      "learning_rate": 9.108742936596567e-06,
+      "loss": 0.7454,
+      "step": 3129
+    },
+    {
+      "epoch": 0.21733092626024164,
+      "grad_norm": 3.693640954962,
+      "learning_rate": 9.108102043365594e-06,
+      "loss": 0.5545,
+      "step": 3130
+    },
+    {
+      "epoch": 0.21740036106096375,
+      "grad_norm": 3.932869010382426,
+      "learning_rate": 9.107460942351467e-06,
+      "loss": 0.4209,
+      "step": 3131
+    },
+    {
+      "epoch": 0.21746979586168588,
+      "grad_norm": 4.529151509314926,
+      "learning_rate": 9.106819633586612e-06,
+      "loss": 0.5477,
+      "step": 3132
+    },
+    {
+      "epoch": 0.217539230662408,
+      "grad_norm": 7.065562075850721,
+      "learning_rate": 9.106178117103468e-06,
+      "loss": 0.5704,
+      "step": 3133
+    },
+    {
+      "epoch": 0.21760866546313012,
+      "grad_norm": 3.445804052243519,
+      "learning_rate": 9.105536392934483e-06,
+      "loss": 0.5002,
+      "step": 3134
+    },
+    {
+      "epoch": 0.21767810026385223,
+      "grad_norm": 4.298564030661732,
+      "learning_rate": 9.10489446111211e-06,
+      "loss": 0.4868,
+      "step": 3135
+    },
+    {
+      "epoch": 0.21774753506457437,
+      "grad_norm": 4.187801444503477,
+      "learning_rate": 9.104252321668823e-06,
+      "loss": 0.5251,
+      "step": 3136
+    },
+    {
+      "epoch": 0.2178169698652965,
+      "grad_norm": 3.6736762743810614,
+      "learning_rate": 9.103609974637097e-06,
+      "loss": 0.4618,
+      "step": 3137
+    },
+    {
+      "epoch": 0.2178864046660186,
+      "grad_norm": 4.253603379700095,
+      "learning_rate": 9.102967420049421e-06,
+      "loss": 0.5018,
+      "step": 3138
+    },
+    {
+      "epoch": 0.21795583946674074,
+      "grad_norm": 3.9327163022409075,
+      "learning_rate": 9.102324657938297e-06,
+      "loss": 0.3432,
+      "step": 3139
+    },
+    {
+      "epoch": 0.21802527426746285,
+      "grad_norm": 3.597650195560748,
+      "learning_rate": 9.101681688336233e-06,
+      "loss": 0.4101,
+      "step": 3140
+    },
+    {
+      "epoch": 0.21809470906818498,
+      "grad_norm": 4.571305699698438,
+      "learning_rate": 9.10103851127575e-06,
+      "loss": 0.6379,
+      "step": 3141
+    },
+    {
+      "epoch": 0.2181641438689071,
+      "grad_norm": 5.020259576371573,
+      "learning_rate": 9.10039512678938e-06,
+      "loss": 0.7497,
+      "step": 3142
+    },
+    {
+      "epoch": 0.21823357866962922,
+      "grad_norm": 4.440256458876575,
+      "learning_rate": 9.099751534909666e-06,
+      "loss": 0.4613,
+      "step": 3143
+    },
+    {
+      "epoch": 0.21830301347035133,
+      "grad_norm": 3.3436968701459557,
+      "learning_rate": 9.099107735669157e-06,
+      "loss": 0.371,
+      "step": 3144
+    },
+    {
+      "epoch": 0.21837244827107347,
+      "grad_norm": 4.269541620196962,
+      "learning_rate": 9.098463729100418e-06,
+      "loss": 0.6622,
+      "step": 3145
+    },
+    {
+      "epoch": 0.21844188307179557,
+      "grad_norm": 3.8250525133325137,
+      "learning_rate": 9.097819515236022e-06,
+      "loss": 0.5138,
+      "step": 3146
+    },
+    {
+      "epoch": 0.2185113178725177,
+      "grad_norm": 4.198308017536499,
+      "learning_rate": 9.09717509410855e-06,
+      "loss": 0.4571,
+      "step": 3147
+    },
+    {
+      "epoch": 0.21858075267323981,
+      "grad_norm": 4.306533146874031,
+      "learning_rate": 9.096530465750599e-06,
+      "loss": 0.7002,
+      "step": 3148
+    },
+    {
+      "epoch": 0.21865018747396195,
+      "grad_norm": 4.10067975244862,
+      "learning_rate": 9.095885630194774e-06,
+      "loss": 0.3935,
+      "step": 3149
+    },
+    {
+      "epoch": 0.21871962227468408,
+      "grad_norm": 4.29636125111369,
+      "learning_rate": 9.095240587473687e-06,
+      "loss": 0.6706,
+      "step": 3150
+    },
+    {
+      "epoch": 0.2187890570754062,
+      "grad_norm": 3.8332562629814184,
+      "learning_rate": 9.094595337619964e-06,
+      "loss": 0.5684,
+      "step": 3151
+    },
+    {
+      "epoch": 0.21885849187612832,
+      "grad_norm": 3.8241695664611233,
+      "learning_rate": 9.093949880666244e-06,
+      "loss": 0.418,
+      "step": 3152
+    },
+    {
+      "epoch": 0.21892792667685043,
+      "grad_norm": 4.31222951366557,
+      "learning_rate": 9.09330421664517e-06,
+      "loss": 0.601,
+      "step": 3153
+    },
+    {
+      "epoch": 0.21899736147757257,
+      "grad_norm": 4.041826293521384,
+      "learning_rate": 9.092658345589404e-06,
+      "loss": 0.4314,
+      "step": 3154
+    },
+    {
+      "epoch": 0.21906679627829467,
+      "grad_norm": 2.6831636813476956,
+      "learning_rate": 9.092012267531607e-06,
+      "loss": 0.1996,
+      "step": 3155
+    },
+    {
+      "epoch": 0.2191362310790168,
+      "grad_norm": 4.690498104972807,
+      "learning_rate": 9.091365982504461e-06,
+      "loss": 0.8055,
+      "step": 3156
+    },
+    {
+      "epoch": 0.21920566587973891,
+      "grad_norm": 3.6573197529870876,
+      "learning_rate": 9.090719490540654e-06,
+      "loss": 0.4624,
+      "step": 3157
+    },
+    {
+      "epoch": 0.21927510068046105,
+      "grad_norm": 4.648329256333156,
+      "learning_rate": 9.090072791672881e-06,
+      "loss": 0.4473,
+      "step": 3158
+    },
+    {
+      "epoch": 0.21934453548118316,
+      "grad_norm": 4.431520922017104,
+      "learning_rate": 9.089425885933858e-06,
+      "loss": 0.8378,
+      "step": 3159
+    },
+    {
+      "epoch": 0.2194139702819053,
+      "grad_norm": 3.642096154107174,
+      "learning_rate": 9.088778773356299e-06,
+      "loss": 0.5909,
+      "step": 3160
+    },
+    {
+      "epoch": 0.21948340508262743,
+      "grad_norm": 3.7035281519428684,
+      "learning_rate": 9.088131453972938e-06,
+      "loss": 0.3521,
+      "step": 3161
+    },
+    {
+      "epoch": 0.21955283988334953,
+      "grad_norm": 3.958560969195321,
+      "learning_rate": 9.087483927816513e-06,
+      "loss": 0.4117,
+      "step": 3162
+    },
+    {
+      "epoch": 0.21962227468407167,
+      "grad_norm": 3.4362757123754744,
+      "learning_rate": 9.086836194919777e-06,
+      "loss": 0.5871,
+      "step": 3163
+    },
+    {
+      "epoch": 0.21969170948479377,
+      "grad_norm": 5.062514970008441,
+      "learning_rate": 9.08618825531549e-06,
+      "loss": 0.5547,
+      "step": 3164
+    },
+    {
+      "epoch": 0.2197611442855159,
+      "grad_norm": 3.7799760996796348,
+      "learning_rate": 9.085540109036426e-06,
+      "loss": 0.5471,
+      "step": 3165
+    },
+    {
+      "epoch": 0.21983057908623801,
+      "grad_norm": 3.4072184801236856,
+      "learning_rate": 9.084891756115365e-06,
+      "loss": 0.3145,
+      "step": 3166
+    },
+    {
+      "epoch": 0.21990001388696015,
+      "grad_norm": 4.313631465829302,
+      "learning_rate": 9.084243196585101e-06,
+      "loss": 0.4822,
+      "step": 3167
+    },
+    {
+      "epoch": 0.21996944868768226,
+      "grad_norm": 3.3087178110709976,
+      "learning_rate": 9.08359443047844e-06,
+      "loss": 0.2751,
+      "step": 3168
+    },
+    {
+      "epoch": 0.2200388834884044,
+      "grad_norm": 4.733265203565386,
+      "learning_rate": 9.082945457828191e-06,
+      "loss": 0.7757,
+      "step": 3169
+    },
+    {
+      "epoch": 0.2201083182891265,
+      "grad_norm": 3.576480283422527,
+      "learning_rate": 9.082296278667184e-06,
+      "loss": 0.4656,
+      "step": 3170
+    },
+    {
+      "epoch": 0.22017775308984863,
+      "grad_norm": 3.1839757828048962,
+      "learning_rate": 9.081646893028248e-06,
+      "loss": 0.2634,
+      "step": 3171
+    },
+    {
+      "epoch": 0.22024718789057077,
+      "grad_norm": 5.321503059839413,
+      "learning_rate": 9.080997300944232e-06,
+      "loss": 0.8959,
+      "step": 3172
+    },
+    {
+      "epoch": 0.22031662269129287,
+      "grad_norm": 3.3603177372770303,
+      "learning_rate": 9.08034750244799e-06,
+      "loss": 0.3754,
+      "step": 3173
+    },
+    {
+      "epoch": 0.220386057492015,
+      "grad_norm": 4.16192163000631,
+      "learning_rate": 9.079697497572389e-06,
+      "loss": 0.467,
+      "step": 3174
+    },
+    {
+      "epoch": 0.22045549229273712,
+      "grad_norm": 4.213405818720391,
+      "learning_rate": 9.079047286350304e-06,
+      "loss": 0.6342,
+      "step": 3175
+    },
+    {
+      "epoch": 0.22052492709345925,
+      "grad_norm": 3.9366326505046887,
+      "learning_rate": 9.078396868814625e-06,
+      "loss": 0.4861,
+      "step": 3176
+    },
+    {
+      "epoch": 0.22059436189418136,
+      "grad_norm": 3.5387391629559555,
+      "learning_rate": 9.077746244998246e-06,
+      "loss": 0.3605,
+      "step": 3177
+    },
+    {
+      "epoch": 0.2206637966949035,
+      "grad_norm": 4.952069669021886,
+      "learning_rate": 9.077095414934076e-06,
+      "loss": 0.7291,
+      "step": 3178
+    },
+    {
+      "epoch": 0.2207332314956256,
+      "grad_norm": 4.989549346045639,
+      "learning_rate": 9.076444378655034e-06,
+      "loss": 0.9141,
+      "step": 3179
+    },
+    {
+      "epoch": 0.22080266629634773,
+      "grad_norm": 5.1225030074973015,
+      "learning_rate": 9.07579313619405e-06,
+      "loss": 0.7268,
+      "step": 3180
+    },
+    {
+      "epoch": 0.22087210109706984,
+      "grad_norm": 2.7799762186087116,
+      "learning_rate": 9.075141687584056e-06,
+      "loss": 0.2322,
+      "step": 3181
+    },
+    {
+      "epoch": 0.22094153589779197,
+      "grad_norm": 3.296339146768401,
+      "learning_rate": 9.074490032858013e-06,
+      "loss": 0.3359,
+      "step": 3182
+    },
+    {
+      "epoch": 0.2210109706985141,
+      "grad_norm": 2.410597587949289,
+      "learning_rate": 9.07383817204887e-06,
+      "loss": 0.3118,
+      "step": 3183
+    },
+    {
+      "epoch": 0.22108040549923622,
+      "grad_norm": 4.430216342465204,
+      "learning_rate": 9.073186105189605e-06,
+      "loss": 0.3594,
+      "step": 3184
+    },
+    {
+      "epoch": 0.22114984029995835,
+      "grad_norm": 3.810111461541002,
+      "learning_rate": 9.072533832313194e-06,
+      "loss": 0.351,
+      "step": 3185
+    },
+    {
+      "epoch": 0.22121927510068046,
+      "grad_norm": 5.231565937218624,
+      "learning_rate": 9.071881353452632e-06,
+      "loss": 0.8923,
+      "step": 3186
+    },
+    {
+      "epoch": 0.2212887099014026,
+      "grad_norm": 4.3664163065531465,
+      "learning_rate": 9.071228668640918e-06,
+      "loss": 0.5511,
+      "step": 3187
+    },
+    {
+      "epoch": 0.2213581447021247,
+      "grad_norm": 4.395651104645602,
+      "learning_rate": 9.070575777911063e-06,
+      "loss": 0.6657,
+      "step": 3188
+    },
+    {
+      "epoch": 0.22142757950284683,
+      "grad_norm": 3.86060901837731,
+      "learning_rate": 9.069922681296094e-06,
+      "loss": 0.2892,
+      "step": 3189
+    },
+    {
+      "epoch": 0.22149701430356894,
+      "grad_norm": 4.846235252109851,
+      "learning_rate": 9.06926937882904e-06,
+      "loss": 0.5847,
+      "step": 3190
+    },
+    {
+      "epoch": 0.22156644910429107,
+      "grad_norm": 5.186671314569035,
+      "learning_rate": 9.068615870542946e-06,
+      "loss": 0.5685,
+      "step": 3191
+    },
+    {
+      "epoch": 0.22163588390501318,
+      "grad_norm": 4.398646902131686,
+      "learning_rate": 9.067962156470865e-06,
+      "loss": 0.633,
+      "step": 3192
+    },
+    {
+      "epoch": 0.22170531870573532,
+      "grad_norm": 4.210034692604905,
+      "learning_rate": 9.06730823664586e-06,
+      "loss": 0.2974,
+      "step": 3193
+    },
+    {
+      "epoch": 0.22177475350645742,
+      "grad_norm": 4.185427526192808,
+      "learning_rate": 9.066654111101009e-06,
+      "loss": 0.5827,
+      "step": 3194
+    },
+    {
+      "epoch": 0.22184418830717956,
+      "grad_norm": 4.920770646954655,
+      "learning_rate": 9.065999779869394e-06,
+      "loss": 0.6647,
+      "step": 3195
+    },
+    {
+      "epoch": 0.2219136231079017,
+      "grad_norm": 5.472157451721229,
+      "learning_rate": 9.06534524298411e-06,
+      "loss": 0.4983,
+      "step": 3196
+    },
+    {
+      "epoch": 0.2219830579086238,
+      "grad_norm": 4.207769175036573,
+      "learning_rate": 9.064690500478266e-06,
+      "loss": 0.3872,
+      "step": 3197
+    },
+    {
+      "epoch": 0.22205249270934593,
+      "grad_norm": 4.248045562387065,
+      "learning_rate": 9.064035552384975e-06,
+      "loss": 0.6654,
+      "step": 3198
+    },
+    {
+      "epoch": 0.22212192751006804,
+      "grad_norm": 3.0786019399998996,
+      "learning_rate": 9.063380398737364e-06,
+      "loss": 0.3686,
+      "step": 3199
+    },
+    {
+      "epoch": 0.22219136231079017,
+      "grad_norm": 3.4721083765118705,
+      "learning_rate": 9.06272503956857e-06,
+      "loss": 0.3282,
+      "step": 3200
+    },
+    {
+      "epoch": 0.22226079711151228,
+      "grad_norm": 3.0602153438617465,
+      "learning_rate": 9.062069474911742e-06,
+      "loss": 0.3121,
+      "step": 3201
+    },
+    {
+      "epoch": 0.22233023191223442,
+      "grad_norm": 3.418533436434518,
+      "learning_rate": 9.061413704800037e-06,
+      "loss": 0.484,
+      "step": 3202
+    },
+    {
+      "epoch": 0.22239966671295652,
+      "grad_norm": 3.7860165484407484,
+      "learning_rate": 9.06075772926662e-06,
+      "loss": 0.53,
+      "step": 3203
+    },
+    {
+      "epoch": 0.22246910151367866,
+      "grad_norm": 4.405256710961151,
+      "learning_rate": 9.060101548344675e-06,
+      "loss": 0.3916,
+      "step": 3204
+    },
+    {
+      "epoch": 0.22253853631440076,
+      "grad_norm": 4.031819357590662,
+      "learning_rate": 9.059445162067386e-06,
+      "loss": 0.447,
+      "step": 3205
+    },
+    {
+      "epoch": 0.2226079711151229,
+      "grad_norm": 5.452168709658814,
+      "learning_rate": 9.058788570467956e-06,
+      "loss": 0.7057,
+      "step": 3206
+    },
+    {
+      "epoch": 0.22267740591584503,
+      "grad_norm": 3.190825358038926,
+      "learning_rate": 9.058131773579592e-06,
+      "loss": 0.3113,
+      "step": 3207
+    },
+    {
+      "epoch": 0.22274684071656714,
+      "grad_norm": 6.781812916756434,
+      "learning_rate": 9.057474771435516e-06,
+      "loss": 0.6399,
+      "step": 3208
+    },
+    {
+      "epoch": 0.22281627551728928,
+      "grad_norm": 4.026958928228671,
+      "learning_rate": 9.056817564068956e-06,
+      "loss": 0.5334,
+      "step": 3209
+    },
+    {
+      "epoch": 0.22288571031801138,
+      "grad_norm": 3.755392518967289,
+      "learning_rate": 9.056160151513154e-06,
+      "loss": 0.4576,
+      "step": 3210
+    },
+    {
+      "epoch": 0.22295514511873352,
+      "grad_norm": 4.258764120726826,
+      "learning_rate": 9.055502533801361e-06,
+      "loss": 0.6035,
+      "step": 3211
+    },
+    {
+      "epoch": 0.22302457991945562,
+      "grad_norm": 4.790069014847633,
+      "learning_rate": 9.05484471096684e-06,
+      "loss": 0.8171,
+      "step": 3212
+    },
+    {
+      "epoch": 0.22309401472017776,
+      "grad_norm": 3.4455477515667283,
+      "learning_rate": 9.054186683042863e-06,
+      "loss": 0.3367,
+      "step": 3213
+    },
+    {
+      "epoch": 0.22316344952089986,
+      "grad_norm": 5.080014953759017,
+      "learning_rate": 9.05352845006271e-06,
+      "loss": 0.813,
+      "step": 3214
+    },
+    {
+      "epoch": 0.223232884321622,
+      "grad_norm": 3.125153708248819,
+      "learning_rate": 9.052870012059676e-06,
+      "loss": 0.3587,
+      "step": 3215
+    },
+    {
+      "epoch": 0.2233023191223441,
+      "grad_norm": 4.201543242961882,
+      "learning_rate": 9.05221136906706e-06,
+      "loss": 0.4502,
+      "step": 3216
+    },
+    {
+      "epoch": 0.22337175392306624,
+      "grad_norm": 4.223503254330185,
+      "learning_rate": 9.051552521118181e-06,
+      "loss": 0.4295,
+      "step": 3217
+    },
+    {
+      "epoch": 0.22344118872378838,
+      "grad_norm": 3.911988669682742,
+      "learning_rate": 9.05089346824636e-06,
+      "loss": 0.4821,
+      "step": 3218
+    },
+    {
+      "epoch": 0.22351062352451048,
+      "grad_norm": 4.11918215030611,
+      "learning_rate": 9.05023421048493e-06,
+      "loss": 0.514,
+      "step": 3219
+    },
+    {
+      "epoch": 0.22358005832523262,
+      "grad_norm": 3.6457240653186638,
+      "learning_rate": 9.049574747867239e-06,
+      "loss": 0.3099,
+      "step": 3220
+    },
+    {
+      "epoch": 0.22364949312595472,
+      "grad_norm": 4.362450941916721,
+      "learning_rate": 9.048915080426637e-06,
+      "loss": 0.4088,
+      "step": 3221
+    },
+    {
+      "epoch": 0.22371892792667686,
+      "grad_norm": 3.934809977691779,
+      "learning_rate": 9.048255208196492e-06,
+      "loss": 0.4095,
+      "step": 3222
+    },
+    {
+      "epoch": 0.22378836272739897,
+      "grad_norm": 4.851740634140416,
+      "learning_rate": 9.04759513121018e-06,
+      "loss": 0.6844,
+      "step": 3223
+    },
+    {
+      "epoch": 0.2238577975281211,
+      "grad_norm": 3.558937352019959,
+      "learning_rate": 9.046934849501087e-06,
+      "loss": 0.4246,
+      "step": 3224
+    },
+    {
+      "epoch": 0.2239272323288432,
+      "grad_norm": 3.666594750735752,
+      "learning_rate": 9.04627436310261e-06,
+      "loss": 0.6376,
+      "step": 3225
+    },
+    {
+      "epoch": 0.22399666712956534,
+      "grad_norm": 4.244613573097513,
+      "learning_rate": 9.045613672048152e-06,
+      "loss": 0.4875,
+      "step": 3226
+    },
+    {
+      "epoch": 0.22406610193028745,
+      "grad_norm": 4.0717711785397706,
+      "learning_rate": 9.044952776371134e-06,
+      "loss": 0.4137,
+      "step": 3227
+    },
+    {
+      "epoch": 0.22413553673100958,
+      "grad_norm": 4.080885430378061,
+      "learning_rate": 9.044291676104981e-06,
+      "loss": 0.4361,
+      "step": 3228
+    },
+    {
+      "epoch": 0.22420497153173172,
+      "grad_norm": 3.7138611550110108,
+      "learning_rate": 9.043630371283131e-06,
+      "loss": 0.4302,
+      "step": 3229
+    },
+    {
+      "epoch": 0.22427440633245382,
+      "grad_norm": 4.146470582755441,
+      "learning_rate": 9.042968861939034e-06,
+      "loss": 0.536,
+      "step": 3230
+    },
+    {
+      "epoch": 0.22434384113317596,
+      "grad_norm": 3.4519970328055987,
+      "learning_rate": 9.042307148106146e-06,
+      "loss": 0.3918,
+      "step": 3231
+    },
+    {
+      "epoch": 0.22441327593389807,
+      "grad_norm": 4.480878934569296,
+      "learning_rate": 9.041645229817936e-06,
+      "loss": 0.6384,
+      "step": 3232
+    },
+    {
+      "epoch": 0.2244827107346202,
+      "grad_norm": 3.7571363646842455,
+      "learning_rate": 9.040983107107884e-06,
+      "loss": 0.5155,
+      "step": 3233
+    },
+    {
+      "epoch": 0.2245521455353423,
+      "grad_norm": 4.437106872771973,
+      "learning_rate": 9.04032078000948e-06,
+      "loss": 0.5441,
+      "step": 3234
+    },
+    {
+      "epoch": 0.22462158033606444,
+      "grad_norm": 3.4468186919304937,
+      "learning_rate": 9.039658248556222e-06,
+      "loss": 0.2585,
+      "step": 3235
+    },
+    {
+      "epoch": 0.22469101513678655,
+      "grad_norm": 3.086416253982117,
+      "learning_rate": 9.03899551278162e-06,
+      "loss": 0.2822,
+      "step": 3236
+    },
+    {
+      "epoch": 0.22476044993750868,
+      "grad_norm": 5.550854900193479,
+      "learning_rate": 9.038332572719197e-06,
+      "loss": 0.8222,
+      "step": 3237
+    },
+    {
+      "epoch": 0.2248298847382308,
+      "grad_norm": 4.370674228309681,
+      "learning_rate": 9.037669428402482e-06,
+      "loss": 0.5625,
+      "step": 3238
+    },
+    {
+      "epoch": 0.22489931953895292,
+      "grad_norm": 4.248561478430586,
+      "learning_rate": 9.037006079865017e-06,
+      "loss": 0.487,
+      "step": 3239
+    },
+    {
+      "epoch": 0.22496875433967506,
+      "grad_norm": 4.1471973778933,
+      "learning_rate": 9.036342527140352e-06,
+      "loss": 0.556,
+      "step": 3240
+    },
+    {
+      "epoch": 0.22503818914039717,
+      "grad_norm": 5.027461615377546,
+      "learning_rate": 9.035678770262049e-06,
+      "loss": 0.6528,
+      "step": 3241
+    },
+    {
+      "epoch": 0.2251076239411193,
+      "grad_norm": 4.55207968453122,
+      "learning_rate": 9.035014809263681e-06,
+      "loss": 0.5913,
+      "step": 3242
+    },
+    {
+      "epoch": 0.2251770587418414,
+      "grad_norm": 4.170993253745282,
+      "learning_rate": 9.034350644178829e-06,
+      "loss": 0.4206,
+      "step": 3243
+    },
+    {
+      "epoch": 0.22524649354256354,
+      "grad_norm": 5.388173689516993,
+      "learning_rate": 9.033686275041088e-06,
+      "loss": 0.4943,
+      "step": 3244
+    },
+    {
+      "epoch": 0.22531592834328565,
+      "grad_norm": 3.5239911456343282,
+      "learning_rate": 9.033021701884057e-06,
+      "loss": 0.5444,
+      "step": 3245
+    },
+    {
+      "epoch": 0.22538536314400778,
+      "grad_norm": 4.871107000153552,
+      "learning_rate": 9.032356924741354e-06,
+      "loss": 0.6076,
+      "step": 3246
+    },
+    {
+      "epoch": 0.2254547979447299,
+      "grad_norm": 5.69230000963202,
+      "learning_rate": 9.031691943646602e-06,
+      "loss": 0.8392,
+      "step": 3247
+    },
+    {
+      "epoch": 0.22552423274545202,
+      "grad_norm": 5.381434365494375,
+      "learning_rate": 9.03102675863343e-06,
+      "loss": 0.642,
+      "step": 3248
+    },
+    {
+      "epoch": 0.22559366754617413,
+      "grad_norm": 4.120124775108753,
+      "learning_rate": 9.030361369735488e-06,
+      "loss": 0.5776,
+      "step": 3249
+    },
+    {
+      "epoch": 0.22566310234689627,
+      "grad_norm": 3.8530163948326495,
+      "learning_rate": 9.029695776986428e-06,
+      "loss": 0.4595,
+      "step": 3250
+    },
+    {
+      "epoch": 0.22573253714761837,
+      "grad_norm": 3.4757791229564754,
+      "learning_rate": 9.029029980419917e-06,
+      "loss": 0.4112,
+      "step": 3251
+    },
+    {
+      "epoch": 0.2258019719483405,
+      "grad_norm": 4.12374353009636,
+      "learning_rate": 9.028363980069628e-06,
+      "loss": 0.4492,
+      "step": 3252
+    },
+    {
+      "epoch": 0.22587140674906264,
+      "grad_norm": 3.721077081011987,
+      "learning_rate": 9.027697775969246e-06,
+      "loss": 0.5169,
+      "step": 3253
+    },
+    {
+      "epoch": 0.22594084154978475,
+      "grad_norm": 3.6983501167657207,
+      "learning_rate": 9.02703136815247e-06,
+      "loss": 0.5111,
+      "step": 3254
+    },
+    {
+      "epoch": 0.22601027635050688,
+      "grad_norm": 5.066074521070494,
+      "learning_rate": 9.026364756653004e-06,
+      "loss": 0.763,
+      "step": 3255
+    },
+    {
+      "epoch": 0.226079711151229,
+      "grad_norm": 3.1084805793498775,
+      "learning_rate": 9.025697941504564e-06,
+      "loss": 0.4527,
+      "step": 3256
+    },
+    {
+      "epoch": 0.22614914595195112,
+      "grad_norm": 4.020763489521019,
+      "learning_rate": 9.025030922740879e-06,
+      "loss": 0.4474,
+      "step": 3257
+    },
+    {
+      "epoch": 0.22621858075267323,
+      "grad_norm": 3.1682521011323046,
+      "learning_rate": 9.024363700395683e-06,
+      "loss": 0.3735,
+      "step": 3258
+    },
+    {
+      "epoch": 0.22628801555339537,
+      "grad_norm": 3.532028945786179,
+      "learning_rate": 9.023696274502725e-06,
+      "loss": 0.4093,
+      "step": 3259
+    },
+    {
+      "epoch": 0.22635745035411747,
+      "grad_norm": 3.6075348518352155,
+      "learning_rate": 9.02302864509576e-06,
+      "loss": 0.4937,
+      "step": 3260
+    },
+    {
+      "epoch": 0.2264268851548396,
+      "grad_norm": 2.9671341943831693,
+      "learning_rate": 9.022360812208564e-06,
+      "loss": 0.2126,
+      "step": 3261
+    },
+    {
+      "epoch": 0.22649631995556171,
+      "grad_norm": 4.166635242307017,
+      "learning_rate": 9.021692775874906e-06,
+      "loss": 0.6409,
+      "step": 3262
+    },
+    {
+      "epoch": 0.22656575475628385,
+      "grad_norm": 4.194861020198873,
+      "learning_rate": 9.02102453612858e-06,
+      "loss": 0.516,
+      "step": 3263
+    },
+    {
+      "epoch": 0.22663518955700598,
+      "grad_norm": 4.169700348540964,
+      "learning_rate": 9.020356093003381e-06,
+      "loss": 0.4806,
+      "step": 3264
+    },
+    {
+      "epoch": 0.2267046243577281,
+      "grad_norm": 3.5001225912535925,
+      "learning_rate": 9.019687446533122e-06,
+      "loss": 0.3067,
+      "step": 3265
+    },
+    {
+      "epoch": 0.22677405915845023,
+      "grad_norm": 4.0189903287019355,
+      "learning_rate": 9.01901859675162e-06,
+      "loss": 0.4979,
+      "step": 3266
+    },
+    {
+      "epoch": 0.22684349395917233,
+      "grad_norm": 3.4185651570264945,
+      "learning_rate": 9.018349543692706e-06,
+      "loss": 0.457,
+      "step": 3267
+    },
+    {
+      "epoch": 0.22691292875989447,
+      "grad_norm": 3.2986901878932637,
+      "learning_rate": 9.017680287390217e-06,
+      "loss": 0.4494,
+      "step": 3268
+    },
+    {
+      "epoch": 0.22698236356061657,
+      "grad_norm": 3.0910952370393887,
+      "learning_rate": 9.017010827878008e-06,
+      "loss": 0.3756,
+      "step": 3269
+    },
+    {
+      "epoch": 0.2270517983613387,
+      "grad_norm": 4.087155541552879,
+      "learning_rate": 9.016341165189935e-06,
+      "loss": 0.6461,
+      "step": 3270
+    },
+    {
+      "epoch": 0.22712123316206081,
+      "grad_norm": 3.473777532901928,
+      "learning_rate": 9.015671299359871e-06,
+      "loss": 0.3388,
+      "step": 3271
+    },
+    {
+      "epoch": 0.22719066796278295,
+      "grad_norm": 3.43156124132588,
+      "learning_rate": 9.015001230421698e-06,
+      "loss": 0.4307,
+      "step": 3272
+    },
+    {
+      "epoch": 0.22726010276350506,
+      "grad_norm": 3.845123770791104,
+      "learning_rate": 9.014330958409306e-06,
+      "loss": 0.5157,
+      "step": 3273
+    },
+    {
+      "epoch": 0.2273295375642272,
+      "grad_norm": 3.8474545722598683,
+      "learning_rate": 9.013660483356594e-06,
+      "loss": 0.4224,
+      "step": 3274
+    },
+    {
+      "epoch": 0.22739897236494933,
+      "grad_norm": 4.209494611295047,
+      "learning_rate": 9.012989805297479e-06,
+      "loss": 0.4056,
+      "step": 3275
+    },
+    {
+      "epoch": 0.22746840716567143,
+      "grad_norm": 4.41777787432151,
+      "learning_rate": 9.01231892426588e-06,
+      "loss": 0.3616,
+      "step": 3276
+    },
+    {
+      "epoch": 0.22753784196639357,
+      "grad_norm": 3.7129817690982936,
+      "learning_rate": 9.011647840295729e-06,
+      "loss": 0.5377,
+      "step": 3277
+    },
+    {
+      "epoch": 0.22760727676711567,
+      "grad_norm": 3.3732750482280642,
+      "learning_rate": 9.010976553420972e-06,
+      "loss": 0.3097,
+      "step": 3278
+    },
+    {
+      "epoch": 0.2276767115678378,
+      "grad_norm": 3.3035935992133534,
+      "learning_rate": 9.010305063675557e-06,
+      "loss": 0.3631,
+      "step": 3279
+    },
+    {
+      "epoch": 0.22774614636855992,
+      "grad_norm": 3.5745572365909277,
+      "learning_rate": 9.009633371093452e-06,
+      "loss": 0.328,
+      "step": 3280
+    },
+    {
+      "epoch": 0.22781558116928205,
+      "grad_norm": 2.959695718181537,
+      "learning_rate": 9.008961475708628e-06,
+      "loss": 0.4074,
+      "step": 3281
+    },
+    {
+      "epoch": 0.22788501597000416,
+      "grad_norm": 4.657067745681218,
+      "learning_rate": 9.008289377555068e-06,
+      "loss": 0.6203,
+      "step": 3282
+    },
+    {
+      "epoch": 0.2279544507707263,
+      "grad_norm": 4.206290552074754,
+      "learning_rate": 9.007617076666768e-06,
+      "loss": 0.6147,
+      "step": 3283
+    },
+    {
+      "epoch": 0.2280238855714484,
+      "grad_norm": 4.150764225909826,
+      "learning_rate": 9.006944573077729e-06,
+      "loss": 0.5385,
+      "step": 3284
+    },
+    {
+      "epoch": 0.22809332037217053,
+      "grad_norm": 3.5266526916199794,
+      "learning_rate": 9.006271866821969e-06,
+      "loss": 0.2318,
+      "step": 3285
+    },
+    {
+      "epoch": 0.22816275517289267,
+      "grad_norm": 4.338667428327251,
+      "learning_rate": 9.005598957933513e-06,
+      "loss": 0.5261,
+      "step": 3286
+    },
+    {
+      "epoch": 0.22823218997361477,
+      "grad_norm": 4.523497214010638,
+      "learning_rate": 9.004925846446393e-06,
+      "loss": 0.4,
+      "step": 3287
+    },
+    {
+      "epoch": 0.2283016247743369,
+      "grad_norm": 3.4181914217814127,
+      "learning_rate": 9.004252532394654e-06,
+      "loss": 0.3159,
+      "step": 3288
+    },
+    {
+      "epoch": 0.22837105957505902,
+      "grad_norm": 3.3918555790964926,
+      "learning_rate": 9.003579015812356e-06,
+      "loss": 0.338,
+      "step": 3289
+    },
+    {
+      "epoch": 0.22844049437578115,
+      "grad_norm": 4.093306528153707,
+      "learning_rate": 9.002905296733561e-06,
+      "loss": 0.3947,
+      "step": 3290
+    },
+    {
+      "epoch": 0.22850992917650326,
+      "grad_norm": 4.879188339408868,
+      "learning_rate": 9.002231375192346e-06,
+      "loss": 0.5572,
+      "step": 3291
+    },
+    {
+      "epoch": 0.2285793639772254,
+      "grad_norm": 4.647888178501052,
+      "learning_rate": 9.001557251222794e-06,
+      "loss": 0.6671,
+      "step": 3292
+    },
+    {
+      "epoch": 0.2286487987779475,
+      "grad_norm": 4.3152061567601105,
+      "learning_rate": 9.000882924859007e-06,
+      "loss": 0.3607,
+      "step": 3293
+    },
+    {
+      "epoch": 0.22871823357866963,
+      "grad_norm": 5.509443840881015,
+      "learning_rate": 9.000208396135088e-06,
+      "loss": 0.532,
+      "step": 3294
+    },
+    {
+      "epoch": 0.22878766837939174,
+      "grad_norm": 3.030300980135691,
+      "learning_rate": 8.999533665085155e-06,
+      "loss": 0.3756,
+      "step": 3295
+    },
+    {
+      "epoch": 0.22885710318011387,
+      "grad_norm": 3.8930106321551032,
+      "learning_rate": 8.998858731743335e-06,
+      "loss": 0.3733,
+      "step": 3296
+    },
+    {
+      "epoch": 0.22892653798083598,
+      "grad_norm": 4.533993772398375,
+      "learning_rate": 8.998183596143765e-06,
+      "loss": 0.7729,
+      "step": 3297
+    },
+    {
+      "epoch": 0.22899597278155812,
+      "grad_norm": 3.140762447363481,
+      "learning_rate": 8.997508258320593e-06,
+      "loss": 0.3236,
+      "step": 3298
+    },
+    {
+      "epoch": 0.22906540758228025,
+      "grad_norm": 5.233020907170975,
+      "learning_rate": 8.996832718307979e-06,
+      "loss": 0.3306,
+      "step": 3299
+    },
+    {
+      "epoch": 0.22913484238300236,
+      "grad_norm": 4.264765678673607,
+      "learning_rate": 8.996156976140088e-06,
+      "loss": 0.4723,
+      "step": 3300
+    },
+    {
+      "epoch": 0.2292042771837245,
+      "grad_norm": 3.2513966117048323,
+      "learning_rate": 8.995481031851097e-06,
+      "loss": 0.33,
+      "step": 3301
+    },
+    {
+      "epoch": 0.2292737119844466,
+      "grad_norm": 2.7407812579739916,
+      "learning_rate": 8.994804885475197e-06,
+      "loss": 0.2695,
+      "step": 3302
+    },
+    {
+      "epoch": 0.22934314678516873,
+      "grad_norm": 3.788670678873121,
+      "learning_rate": 8.994128537046587e-06,
+      "loss": 0.4258,
+      "step": 3303
+    },
+    {
+      "epoch": 0.22941258158589084,
+      "grad_norm": 4.835697299970811,
+      "learning_rate": 8.993451986599474e-06,
+      "loss": 0.5633,
+      "step": 3304
+    },
+    {
+      "epoch": 0.22948201638661297,
+      "grad_norm": 4.317329183551526,
+      "learning_rate": 8.99277523416808e-06,
+      "loss": 0.4643,
+      "step": 3305
+    },
+    {
+      "epoch": 0.22955145118733508,
+      "grad_norm": 3.8886125378452236,
+      "learning_rate": 8.992098279786634e-06,
+      "loss": 0.4422,
+      "step": 3306
+    },
+    {
+      "epoch": 0.22962088598805722,
+      "grad_norm": 4.398256003664236,
+      "learning_rate": 8.991421123489373e-06,
+      "loss": 0.5,
+      "step": 3307
+    },
+    {
+      "epoch": 0.22969032078877932,
+      "grad_norm": 4.9720455301084,
+      "learning_rate": 8.990743765310547e-06,
+      "loss": 0.5397,
+      "step": 3308
+    },
+    {
+      "epoch": 0.22975975558950146,
+      "grad_norm": 4.433124513989263,
+      "learning_rate": 8.990066205284417e-06,
+      "loss": 0.6294,
+      "step": 3309
+    },
+    {
+      "epoch": 0.2298291903902236,
+      "grad_norm": 4.664645622465553,
+      "learning_rate": 8.989388443445256e-06,
+      "loss": 0.6,
+      "step": 3310
+    },
+    {
+      "epoch": 0.2298986251909457,
+      "grad_norm": 3.8341873829765674,
+      "learning_rate": 8.98871047982734e-06,
+      "loss": 0.5526,
+      "step": 3311
+    },
+    {
+      "epoch": 0.22996805999166783,
+      "grad_norm": 4.255273053935012,
+      "learning_rate": 8.988032314464963e-06,
+      "loss": 0.6691,
+      "step": 3312
+    },
+    {
+      "epoch": 0.23003749479238994,
+      "grad_norm": 4.1662027455166255,
+      "learning_rate": 8.987353947392422e-06,
+      "loss": 0.5802,
+      "step": 3313
+    },
+    {
+      "epoch": 0.23010692959311208,
+      "grad_norm": 5.45198074243529,
+      "learning_rate": 8.986675378644032e-06,
+      "loss": 0.4117,
+      "step": 3314
+    },
+    {
+      "epoch": 0.23017636439383418,
+      "grad_norm": 2.2447038933412173,
+      "learning_rate": 8.985996608254113e-06,
+      "loss": 0.1665,
+      "step": 3315
+    },
+    {
+      "epoch": 0.23024579919455632,
+      "grad_norm": 4.192959016608387,
+      "learning_rate": 8.985317636256994e-06,
+      "loss": 0.4812,
+      "step": 3316
+    },
+    {
+      "epoch": 0.23031523399527842,
+      "grad_norm": 4.3753098058211775,
+      "learning_rate": 8.98463846268702e-06,
+      "loss": 0.6928,
+      "step": 3317
+    },
+    {
+      "epoch": 0.23038466879600056,
+      "grad_norm": 3.5556490756304346,
+      "learning_rate": 8.983959087578542e-06,
+      "loss": 0.4773,
+      "step": 3318
+    },
+    {
+      "epoch": 0.23045410359672266,
+      "grad_norm": 3.418455644042121,
+      "learning_rate": 8.98327951096592e-06,
+      "loss": 0.4675,
+      "step": 3319
+    },
+    {
+      "epoch": 0.2305235383974448,
+      "grad_norm": 5.472887332941176,
+      "learning_rate": 8.982599732883528e-06,
+      "loss": 0.8142,
+      "step": 3320
+    },
+    {
+      "epoch": 0.23059297319816693,
+      "grad_norm": 2.8833440781651283,
+      "learning_rate": 8.981919753365748e-06,
+      "loss": 0.2592,
+      "step": 3321
+    },
+    {
+      "epoch": 0.23066240799888904,
+      "grad_norm": 3.3418810146106117,
+      "learning_rate": 8.981239572446975e-06,
+      "loss": 0.308,
+      "step": 3322
+    },
+    {
+      "epoch": 0.23073184279961118,
+      "grad_norm": 3.0048785547972487,
+      "learning_rate": 8.980559190161607e-06,
+      "loss": 0.2899,
+      "step": 3323
+    },
+    {
+      "epoch": 0.23080127760033328,
+      "grad_norm": 5.295175041185349,
+      "learning_rate": 8.979878606544061e-06,
+      "loss": 0.5676,
+      "step": 3324
+    },
+    {
+      "epoch": 0.23087071240105542,
+      "grad_norm": 2.3756071796125333,
+      "learning_rate": 8.979197821628758e-06,
+      "loss": 0.1594,
+      "step": 3325
+    },
+    {
+      "epoch": 0.23094014720177752,
+      "grad_norm": 4.123512813892296,
+      "learning_rate": 8.97851683545013e-06,
+      "loss": 0.5433,
+      "step": 3326
+    },
+    {
+      "epoch": 0.23100958200249966,
+      "grad_norm": 2.315339577945573,
+      "learning_rate": 8.977835648042626e-06,
+      "loss": 0.2494,
+      "step": 3327
+    },
+    {
+      "epoch": 0.23107901680322177,
+      "grad_norm": 4.115183519525345,
+      "learning_rate": 8.977154259440695e-06,
+      "loss": 0.479,
+      "step": 3328
+    },
+    {
+      "epoch": 0.2311484516039439,
+      "grad_norm": 3.7364876316647617,
+      "learning_rate": 8.976472669678801e-06,
+      "loss": 0.2769,
+      "step": 3329
+    },
+    {
+      "epoch": 0.231217886404666,
+      "grad_norm": 3.650959086572034,
+      "learning_rate": 8.97579087879142e-06,
+      "loss": 0.5645,
+      "step": 3330
+    },
+    {
+      "epoch": 0.23128732120538814,
+      "grad_norm": 4.4933665091326915,
+      "learning_rate": 8.975108886813035e-06,
+      "loss": 0.9186,
+      "step": 3331
+    },
+    {
+      "epoch": 0.23135675600611028,
+      "grad_norm": 3.74900299916282,
+      "learning_rate": 8.97442669377814e-06,
+      "loss": 0.653,
+      "step": 3332
+    },
+    {
+      "epoch": 0.23142619080683238,
+      "grad_norm": 3.718195321957163,
+      "learning_rate": 8.97374429972124e-06,
+      "loss": 0.3182,
+      "step": 3333
+    },
+    {
+      "epoch": 0.23149562560755452,
+      "grad_norm": 3.798842222962264,
+      "learning_rate": 8.97306170467685e-06,
+      "loss": 0.4301,
+      "step": 3334
+    },
+    {
+      "epoch": 0.23156506040827662,
+      "grad_norm": 3.2507870132983534,
+      "learning_rate": 8.972378908679496e-06,
+      "loss": 0.268,
+      "step": 3335
+    },
+    {
+      "epoch": 0.23163449520899876,
+      "grad_norm": 3.5743744707341465,
+      "learning_rate": 8.971695911763712e-06,
+      "loss": 0.3383,
+      "step": 3336
+    },
+    {
+      "epoch": 0.23170393000972087,
+      "grad_norm": 3.98713149391989,
+      "learning_rate": 8.971012713964043e-06,
+      "loss": 0.4376,
+      "step": 3337
+    },
+    {
+      "epoch": 0.231773364810443,
+      "grad_norm": 4.803120078072777,
+      "learning_rate": 8.970329315315042e-06,
+      "loss": 0.6666,
+      "step": 3338
+    },
+    {
+      "epoch": 0.2318427996111651,
+      "grad_norm": 3.408056810018496,
+      "learning_rate": 8.969645715851279e-06,
+      "loss": 0.5395,
+      "step": 3339
+    },
+    {
+      "epoch": 0.23191223441188724,
+      "grad_norm": 4.152958202627536,
+      "learning_rate": 8.968961915607327e-06,
+      "loss": 0.3961,
+      "step": 3340
+    },
+    {
+      "epoch": 0.23198166921260935,
+      "grad_norm": 2.749034230274466,
+      "learning_rate": 8.968277914617773e-06,
+      "loss": 0.3492,
+      "step": 3341
+    },
+    {
+      "epoch": 0.23205110401333148,
+      "grad_norm": 4.200202739207985,
+      "learning_rate": 8.967593712917213e-06,
+      "loss": 0.4991,
+      "step": 3342
+    },
+    {
+      "epoch": 0.23212053881405362,
+      "grad_norm": 4.467870838964765,
+      "learning_rate": 8.966909310540251e-06,
+      "loss": 0.7526,
+      "step": 3343
+    },
+    {
+      "epoch": 0.23218997361477572,
+      "grad_norm": 2.8735861593028713,
+      "learning_rate": 8.966224707521508e-06,
+      "loss": 0.4459,
+      "step": 3344
+    },
+    {
+      "epoch": 0.23225940841549786,
+      "grad_norm": 8.423817411811424,
+      "learning_rate": 8.965539903895603e-06,
+      "loss": 0.5026,
+      "step": 3345
+    },
+    {
+      "epoch": 0.23232884321621997,
+      "grad_norm": 4.716740028027578,
+      "learning_rate": 8.96485489969718e-06,
+      "loss": 0.6625,
+      "step": 3346
+    },
+    {
+      "epoch": 0.2323982780169421,
+      "grad_norm": 4.9332017759714795,
+      "learning_rate": 8.964169694960882e-06,
+      "loss": 0.6211,
+      "step": 3347
+    },
+    {
+      "epoch": 0.2324677128176642,
+      "grad_norm": 2.6922439074272724,
+      "learning_rate": 8.963484289721366e-06,
+      "loss": 0.4367,
+      "step": 3348
+    },
+    {
+      "epoch": 0.23253714761838634,
+      "grad_norm": 3.0381392436537102,
+      "learning_rate": 8.9627986840133e-06,
+      "loss": 0.3502,
+      "step": 3349
+    },
+    {
+      "epoch": 0.23260658241910845,
+      "grad_norm": 5.328153369560276,
+      "learning_rate": 8.96211287787136e-06,
+      "loss": 0.736,
+      "step": 3350
+    },
+    {
+      "epoch": 0.23267601721983058,
+      "grad_norm": 4.235209959488185,
+      "learning_rate": 8.961426871330236e-06,
+      "loss": 0.5585,
+      "step": 3351
+    },
+    {
+      "epoch": 0.2327454520205527,
+      "grad_norm": 4.197855303305832,
+      "learning_rate": 8.960740664424622e-06,
+      "loss": 0.6997,
+      "step": 3352
+    },
+    {
+      "epoch": 0.23281488682127482,
+      "grad_norm": 4.291113952435272,
+      "learning_rate": 8.960054257189228e-06,
+      "loss": 0.7383,
+      "step": 3353
+    },
+    {
+      "epoch": 0.23288432162199693,
+      "grad_norm": 3.8358199007491205,
+      "learning_rate": 8.95936764965877e-06,
+      "loss": 0.2604,
+      "step": 3354
+    },
+    {
+      "epoch": 0.23295375642271907,
+      "grad_norm": 3.7635641552551484,
+      "learning_rate": 8.958680841867977e-06,
+      "loss": 0.304,
+      "step": 3355
+    },
+    {
+      "epoch": 0.2330231912234412,
+      "grad_norm": 4.27048773599421,
+      "learning_rate": 8.957993833851587e-06,
+      "loss": 0.542,
+      "step": 3356
+    },
+    {
+      "epoch": 0.2330926260241633,
+      "grad_norm": 3.9157452887181488,
+      "learning_rate": 8.957306625644347e-06,
+      "loss": 0.4503,
+      "step": 3357
+    },
+    {
+      "epoch": 0.23316206082488544,
+      "grad_norm": 3.4182495579614045,
+      "learning_rate": 8.956619217281016e-06,
+      "loss": 0.2536,
+      "step": 3358
+    },
+    {
+      "epoch": 0.23323149562560755,
+      "grad_norm": 4.185413196427375,
+      "learning_rate": 8.955931608796362e-06,
+      "loss": 0.6125,
+      "step": 3359
+    },
+    {
+      "epoch": 0.23330093042632968,
+      "grad_norm": 3.896797424380583,
+      "learning_rate": 8.955243800225164e-06,
+      "loss": 0.293,
+      "step": 3360
+    },
+    {
+      "epoch": 0.2333703652270518,
+      "grad_norm": 3.798300902174269,
+      "learning_rate": 8.95455579160221e-06,
+      "loss": 0.6208,
+      "step": 3361
+    },
+    {
+      "epoch": 0.23343980002777392,
+      "grad_norm": 3.4009955175680773,
+      "learning_rate": 8.9538675829623e-06,
+      "loss": 0.3601,
+      "step": 3362
+    },
+    {
+      "epoch": 0.23350923482849603,
+      "grad_norm": 4.543279021141069,
+      "learning_rate": 8.953179174340241e-06,
+      "loss": 0.7166,
+      "step": 3363
+    },
+    {
+      "epoch": 0.23357866962921817,
+      "grad_norm": 3.839788504616365,
+      "learning_rate": 8.952490565770853e-06,
+      "loss": 0.4993,
+      "step": 3364
+    },
+    {
+      "epoch": 0.23364810442994027,
+      "grad_norm": 5.475177320637296,
+      "learning_rate": 8.951801757288962e-06,
+      "loss": 0.8026,
+      "step": 3365
+    },
+    {
+      "epoch": 0.2337175392306624,
+      "grad_norm": 4.835838292295004,
+      "learning_rate": 8.951112748929414e-06,
+      "loss": 0.4774,
+      "step": 3366
+    },
+    {
+      "epoch": 0.23378697403138454,
+      "grad_norm": 4.378367624530631,
+      "learning_rate": 8.950423540727051e-06,
+      "loss": 0.5953,
+      "step": 3367
+    },
+    {
+      "epoch": 0.23385640883210665,
+      "grad_norm": 4.858037252958956,
+      "learning_rate": 8.949734132716737e-06,
+      "loss": 0.4573,
+      "step": 3368
+    },
+    {
+      "epoch": 0.23392584363282878,
+      "grad_norm": 4.137376308407242,
+      "learning_rate": 8.949044524933342e-06,
+      "loss": 0.6468,
+      "step": 3369
+    },
+    {
+      "epoch": 0.2339952784335509,
+      "grad_norm": 3.936538430847476,
+      "learning_rate": 8.94835471741174e-06,
+      "loss": 0.5321,
+      "step": 3370
+    },
+    {
+      "epoch": 0.23406471323427303,
+      "grad_norm": 4.094439928988578,
+      "learning_rate": 8.947664710186825e-06,
+      "loss": 0.609,
+      "step": 3371
+    },
+    {
+      "epoch": 0.23413414803499513,
+      "grad_norm": 3.310639094320513,
+      "learning_rate": 8.946974503293497e-06,
+      "loss": 0.37,
+      "step": 3372
+    },
+    {
+      "epoch": 0.23420358283571727,
+      "grad_norm": 4.932059504385731,
+      "learning_rate": 8.946284096766664e-06,
+      "loss": 0.9065,
+      "step": 3373
+    },
+    {
+      "epoch": 0.23427301763643937,
+      "grad_norm": 4.670274968641153,
+      "learning_rate": 8.945593490641248e-06,
+      "loss": 0.4096,
+      "step": 3374
+    },
+    {
+      "epoch": 0.2343424524371615,
+      "grad_norm": 4.3246680480367,
+      "learning_rate": 8.944902684952178e-06,
+      "loss": 0.6245,
+      "step": 3375
+    },
+    {
+      "epoch": 0.23441188723788361,
+      "grad_norm": 4.647584938884255,
+      "learning_rate": 8.944211679734394e-06,
+      "loss": 0.5449,
+      "step": 3376
+    },
+    {
+      "epoch": 0.23448132203860575,
+      "grad_norm": 3.6210826047017073,
+      "learning_rate": 8.943520475022846e-06,
+      "loss": 0.3787,
+      "step": 3377
+    },
+    {
+      "epoch": 0.23455075683932788,
+      "grad_norm": 3.8531015668948907,
+      "learning_rate": 8.942829070852495e-06,
+      "loss": 0.4568,
+      "step": 3378
+    },
+    {
+      "epoch": 0.23462019164005,
+      "grad_norm": 5.032536681824429,
+      "learning_rate": 8.942137467258312e-06,
+      "loss": 0.8581,
+      "step": 3379
+    },
+    {
+      "epoch": 0.23468962644077213,
+      "grad_norm": 4.877122835383702,
+      "learning_rate": 8.941445664275274e-06,
+      "loss": 0.8188,
+      "step": 3380
+    },
+    {
+      "epoch": 0.23475906124149423,
+      "grad_norm": 4.355878772229833,
+      "learning_rate": 8.940753661938377e-06,
+      "loss": 0.4724,
+      "step": 3381
+    },
+    {
+      "epoch": 0.23482849604221637,
+      "grad_norm": 3.476655711401579,
+      "learning_rate": 8.940061460282619e-06,
+      "loss": 0.4733,
+      "step": 3382
+    },
+    {
+      "epoch": 0.23489793084293847,
+      "grad_norm": 4.220378142969899,
+      "learning_rate": 8.939369059343008e-06,
+      "loss": 0.4638,
+      "step": 3383
+    },
+    {
+      "epoch": 0.2349673656436606,
+      "grad_norm": 4.086210202368822,
+      "learning_rate": 8.93867645915457e-06,
+      "loss": 0.5657,
+      "step": 3384
+    },
+    {
+      "epoch": 0.23503680044438272,
+      "grad_norm": 4.625643542404187,
+      "learning_rate": 8.937983659752335e-06,
+      "loss": 0.7502,
+      "step": 3385
+    },
+    {
+      "epoch": 0.23510623524510485,
+      "grad_norm": 3.6129740585457695,
+      "learning_rate": 8.937290661171341e-06,
+      "loss": 0.4391,
+      "step": 3386
+    },
+    {
+      "epoch": 0.23517567004582696,
+      "grad_norm": 4.954692914858558,
+      "learning_rate": 8.936597463446639e-06,
+      "loss": 0.842,
+      "step": 3387
+    },
+    {
+      "epoch": 0.2352451048465491,
+      "grad_norm": 3.231737859588972,
+      "learning_rate": 8.935904066613295e-06,
+      "loss": 0.3204,
+      "step": 3388
+    },
+    {
+      "epoch": 0.23531453964727123,
+      "grad_norm": 3.229392515818551,
+      "learning_rate": 8.935210470706374e-06,
+      "loss": 0.2561,
+      "step": 3389
+    },
+    {
+      "epoch": 0.23538397444799333,
+      "grad_norm": 3.756300216832138,
+      "learning_rate": 8.934516675760962e-06,
+      "loss": 0.6203,
+      "step": 3390
+    },
+    {
+      "epoch": 0.23545340924871547,
+      "grad_norm": 3.953787631441059,
+      "learning_rate": 8.93382268181215e-06,
+      "loss": 0.6087,
+      "step": 3391
+    },
+    {
+      "epoch": 0.23552284404943757,
+      "grad_norm": 3.448353761025936,
+      "learning_rate": 8.933128488895036e-06,
+      "loss": 0.3747,
+      "step": 3392
+    },
+    {
+      "epoch": 0.2355922788501597,
+      "grad_norm": 3.2720514376138494,
+      "learning_rate": 8.932434097044735e-06,
+      "loss": 0.369,
+      "step": 3393
+    },
+    {
+      "epoch": 0.23566171365088182,
+      "grad_norm": 3.649427377454302,
+      "learning_rate": 8.931739506296366e-06,
+      "loss": 0.4324,
+      "step": 3394
+    },
+    {
+      "epoch": 0.23573114845160395,
+      "grad_norm": 3.9630549818490164,
+      "learning_rate": 8.931044716685063e-06,
+      "loss": 0.4788,
+      "step": 3395
+    },
+    {
+      "epoch": 0.23580058325232606,
+      "grad_norm": 3.937715497607778,
+      "learning_rate": 8.930349728245965e-06,
+      "loss": 0.429,
+      "step": 3396
+    },
+    {
+      "epoch": 0.2358700180530482,
+      "grad_norm": 4.31892630707009,
+      "learning_rate": 8.929654541014227e-06,
+      "loss": 0.4829,
+      "step": 3397
+    },
+    {
+      "epoch": 0.2359394528537703,
+      "grad_norm": 3.708245271854308,
+      "learning_rate": 8.928959155025008e-06,
+      "loss": 0.1864,
+      "step": 3398
+    },
+    {
+      "epoch": 0.23600888765449243,
+      "grad_norm": 4.174177054413822,
+      "learning_rate": 8.928263570313483e-06,
+      "loss": 0.6108,
+      "step": 3399
+    },
+    {
+      "epoch": 0.23607832245521454,
+      "grad_norm": 4.85156222682328,
+      "learning_rate": 8.92756778691483e-06,
+      "loss": 0.6724,
+      "step": 3400
+    },
+    {
+      "epoch": 0.23614775725593667,
+      "grad_norm": 2.9856238664889143,
+      "learning_rate": 8.926871804864242e-06,
+      "loss": 0.3353,
+      "step": 3401
+    },
+    {
+      "epoch": 0.2362171920566588,
+      "grad_norm": 3.7501304671640083,
+      "learning_rate": 8.926175624196921e-06,
+      "loss": 0.3548,
+      "step": 3402
+    },
+    {
+      "epoch": 0.23628662685738092,
+      "grad_norm": 3.1277728620823306,
+      "learning_rate": 8.925479244948081e-06,
+      "loss": 0.3819,
+      "step": 3403
+    },
+    {
+      "epoch": 0.23635606165810305,
+      "grad_norm": 3.738017512608187,
+      "learning_rate": 8.924782667152943e-06,
+      "loss": 0.4004,
+      "step": 3404
+    },
+    {
+      "epoch": 0.23642549645882516,
+      "grad_norm": 3.770749830226631,
+      "learning_rate": 8.92408589084674e-06,
+      "loss": 0.4789,
+      "step": 3405
+    },
+    {
+      "epoch": 0.2364949312595473,
+      "grad_norm": 3.9300022084069752,
+      "learning_rate": 8.92338891606471e-06,
+      "loss": 0.4008,
+      "step": 3406
+    },
+    {
+      "epoch": 0.2365643660602694,
+      "grad_norm": 3.1195924790060263,
+      "learning_rate": 8.922691742842112e-06,
+      "loss": 0.2789,
+      "step": 3407
+    },
+    {
+      "epoch": 0.23663380086099153,
+      "grad_norm": 4.93983735552413,
+      "learning_rate": 8.921994371214202e-06,
+      "loss": 0.6781,
+      "step": 3408
+    },
+    {
+      "epoch": 0.23670323566171364,
+      "grad_norm": 3.621970808135141,
+      "learning_rate": 8.921296801216255e-06,
+      "loss": 0.4139,
+      "step": 3409
+    },
+    {
+      "epoch": 0.23677267046243577,
+      "grad_norm": 3.8903205996193853,
+      "learning_rate": 8.920599032883553e-06,
+      "loss": 0.5832,
+      "step": 3410
+    },
+    {
+      "epoch": 0.23684210526315788,
+      "grad_norm": 3.252887920455228,
+      "learning_rate": 8.919901066251389e-06,
+      "loss": 0.4231,
+      "step": 3411
+    },
+    {
+      "epoch": 0.23691154006388002,
+      "grad_norm": 2.96533854518334,
+      "learning_rate": 8.919202901355064e-06,
+      "loss": 0.3032,
+      "step": 3412
+    },
+    {
+      "epoch": 0.23698097486460215,
+      "grad_norm": 4.348096951179054,
+      "learning_rate": 8.91850453822989e-06,
+      "loss": 0.5009,
+      "step": 3413
+    },
+    {
+      "epoch": 0.23705040966532426,
+      "grad_norm": 3.0525439725841887,
+      "learning_rate": 8.917805976911194e-06,
+      "loss": 0.2848,
+      "step": 3414
+    },
+    {
+      "epoch": 0.2371198444660464,
+      "grad_norm": 4.687705473686893,
+      "learning_rate": 8.917107217434302e-06,
+      "loss": 0.7724,
+      "step": 3415
+    },
+    {
+      "epoch": 0.2371892792667685,
+      "grad_norm": 3.871926767307289,
+      "learning_rate": 8.916408259834562e-06,
+      "loss": 0.5542,
+      "step": 3416
+    },
+    {
+      "epoch": 0.23725871406749063,
+      "grad_norm": 3.7909469279179953,
+      "learning_rate": 8.91570910414732e-06,
+      "loss": 0.5268,
+      "step": 3417
+    },
+    {
+      "epoch": 0.23732814886821274,
+      "grad_norm": 3.8419992792217426,
+      "learning_rate": 8.915009750407947e-06,
+      "loss": 0.4926,
+      "step": 3418
+    },
+    {
+      "epoch": 0.23739758366893488,
+      "grad_norm": 4.135368143245829,
+      "learning_rate": 8.914310198651807e-06,
+      "loss": 0.477,
+      "step": 3419
+    },
+    {
+      "epoch": 0.23746701846965698,
+      "grad_norm": 4.452521067169484,
+      "learning_rate": 8.91361044891429e-06,
+      "loss": 0.5635,
+      "step": 3420
+    },
+    {
+      "epoch": 0.23753645327037912,
+      "grad_norm": 3.7541742409623815,
+      "learning_rate": 8.912910501230782e-06,
+      "loss": 0.4751,
+      "step": 3421
+    },
+    {
+      "epoch": 0.23760588807110122,
+      "grad_norm": 3.8371193885839014,
+      "learning_rate": 8.91221035563669e-06,
+      "loss": 0.5039,
+      "step": 3422
+    },
+    {
+      "epoch": 0.23767532287182336,
+      "grad_norm": 5.077340597136206,
+      "learning_rate": 8.911510012167425e-06,
+      "loss": 0.6132,
+      "step": 3423
+    },
+    {
+      "epoch": 0.2377447576725455,
+      "grad_norm": 3.141818035172513,
+      "learning_rate": 8.91080947085841e-06,
+      "loss": 0.3086,
+      "step": 3424
+    },
+    {
+      "epoch": 0.2378141924732676,
+      "grad_norm": 5.615593639178157,
+      "learning_rate": 8.910108731745078e-06,
+      "loss": 0.7889,
+      "step": 3425
+    },
+    {
+      "epoch": 0.23788362727398973,
+      "grad_norm": 4.70487180473826,
+      "learning_rate": 8.90940779486287e-06,
+      "loss": 0.6064,
+      "step": 3426
+    },
+    {
+      "epoch": 0.23795306207471184,
+      "grad_norm": 3.602691751691279,
+      "learning_rate": 8.908706660247239e-06,
+      "loss": 0.4251,
+      "step": 3427
+    },
+    {
+      "epoch": 0.23802249687543398,
+      "grad_norm": 4.228402901905052,
+      "learning_rate": 8.90800532793365e-06,
+      "loss": 0.5611,
+      "step": 3428
+    },
+    {
+      "epoch": 0.23809193167615608,
+      "grad_norm": 4.405210442623311,
+      "learning_rate": 8.907303797957571e-06,
+      "loss": 0.5285,
+      "step": 3429
+    },
+    {
+      "epoch": 0.23816136647687822,
+      "grad_norm": 3.505146206411121,
+      "learning_rate": 8.906602070354488e-06,
+      "loss": 0.3053,
+      "step": 3430
+    },
+    {
+      "epoch": 0.23823080127760032,
+      "grad_norm": 4.601475378960544,
+      "learning_rate": 8.905900145159894e-06,
+      "loss": 0.6146,
+      "step": 3431
+    },
+    {
+      "epoch": 0.23830023607832246,
+      "grad_norm": 4.418610404683968,
+      "learning_rate": 8.905198022409292e-06,
+      "loss": 0.5927,
+      "step": 3432
+    },
+    {
+      "epoch": 0.23836967087904457,
+      "grad_norm": 3.9570299858963445,
+      "learning_rate": 8.904495702138189e-06,
+      "loss": 0.4316,
+      "step": 3433
+    },
+    {
+      "epoch": 0.2384391056797667,
+      "grad_norm": 4.3308963416843245,
+      "learning_rate": 8.903793184382115e-06,
+      "loss": 0.623,
+      "step": 3434
+    },
+    {
+      "epoch": 0.23850854048048883,
+      "grad_norm": 4.8135802585804095,
+      "learning_rate": 8.903090469176598e-06,
+      "loss": 0.6551,
+      "step": 3435
+    },
+    {
+      "epoch": 0.23857797528121094,
+      "grad_norm": 4.253986366506154,
+      "learning_rate": 8.902387556557182e-06,
+      "loss": 0.535,
+      "step": 3436
+    },
+    {
+      "epoch": 0.23864741008193308,
+      "grad_norm": 4.492071024323041,
+      "learning_rate": 8.90168444655942e-06,
+      "loss": 0.4903,
+      "step": 3437
+    },
+    {
+      "epoch": 0.23871684488265518,
+      "grad_norm": 4.059494433365717,
+      "learning_rate": 8.900981139218872e-06,
+      "loss": 0.689,
+      "step": 3438
+    },
+    {
+      "epoch": 0.23878627968337732,
+      "grad_norm": 3.417187640754658,
+      "learning_rate": 8.900277634571113e-06,
+      "loss": 0.4187,
+      "step": 3439
+    },
+    {
+      "epoch": 0.23885571448409942,
+      "grad_norm": 4.655441852707797,
+      "learning_rate": 8.899573932651723e-06,
+      "loss": 0.4867,
+      "step": 3440
+    },
+    {
+      "epoch": 0.23892514928482156,
+      "grad_norm": 4.566931576308445,
+      "learning_rate": 8.898870033496299e-06,
+      "loss": 0.7239,
+      "step": 3441
+    },
+    {
+      "epoch": 0.23899458408554367,
+      "grad_norm": 4.561754223053981,
+      "learning_rate": 8.898165937140439e-06,
+      "loss": 0.7939,
+      "step": 3442
+    },
+    {
+      "epoch": 0.2390640188862658,
+      "grad_norm": 3.7492515932581174,
+      "learning_rate": 8.897461643619758e-06,
+      "loss": 0.3455,
+      "step": 3443
+    },
+    {
+      "epoch": 0.2391334536869879,
+      "grad_norm": 3.9796283953703826,
+      "learning_rate": 8.896757152969878e-06,
+      "loss": 0.5,
+      "step": 3444
+    },
+    {
+      "epoch": 0.23920288848771004,
+      "grad_norm": 4.256454843044218,
+      "learning_rate": 8.896052465226427e-06,
+      "loss": 0.5779,
+      "step": 3445
+    },
+    {
+      "epoch": 0.23927232328843218,
+      "grad_norm": 3.5169316922912373,
+      "learning_rate": 8.895347580425056e-06,
+      "loss": 0.5087,
+      "step": 3446
+    },
+    {
+      "epoch": 0.23934175808915428,
+      "grad_norm": 4.651881746849845,
+      "learning_rate": 8.894642498601409e-06,
+      "loss": 0.5431,
+      "step": 3447
+    },
+    {
+      "epoch": 0.23941119288987642,
+      "grad_norm": 4.529313098759388,
+      "learning_rate": 8.893937219791151e-06,
+      "loss": 0.482,
+      "step": 3448
+    },
+    {
+      "epoch": 0.23948062769059852,
+      "grad_norm": 5.094807145100306,
+      "learning_rate": 8.893231744029956e-06,
+      "loss": 0.5853,
+      "step": 3449
+    },
+    {
+      "epoch": 0.23955006249132066,
+      "grad_norm": 3.8647826896018502,
+      "learning_rate": 8.892526071353504e-06,
+      "loss": 0.6818,
+      "step": 3450
+    },
+    {
+      "epoch": 0.23961949729204277,
+      "grad_norm": 4.028957408905017,
+      "learning_rate": 8.89182020179749e-06,
+      "loss": 0.499,
+      "step": 3451
+    },
+    {
+      "epoch": 0.2396889320927649,
+      "grad_norm": 4.135513025509712,
+      "learning_rate": 8.891114135397614e-06,
+      "loss": 0.4164,
+      "step": 3452
+    },
+    {
+      "epoch": 0.239758366893487,
+      "grad_norm": 3.831839688098623,
+      "learning_rate": 8.890407872189587e-06,
+      "loss": 0.3281,
+      "step": 3453
+    },
+    {
+      "epoch": 0.23982780169420914,
+      "grad_norm": 5.540201262194585,
+      "learning_rate": 8.889701412209131e-06,
+      "loss": 0.7008,
+      "step": 3454
+    },
+    {
+      "epoch": 0.23989723649493125,
+      "grad_norm": 4.466899840394317,
+      "learning_rate": 8.88899475549198e-06,
+      "loss": 0.5492,
+      "step": 3455
+    },
+    {
+      "epoch": 0.23996667129565338,
+      "grad_norm": 3.8026019918869927,
+      "learning_rate": 8.888287902073875e-06,
+      "loss": 0.5855,
+      "step": 3456
+    },
+    {
+      "epoch": 0.2400361060963755,
+      "grad_norm": 4.562941758582137,
+      "learning_rate": 8.887580851990568e-06,
+      "loss": 0.7271,
+      "step": 3457
+    },
+    {
+      "epoch": 0.24010554089709762,
+      "grad_norm": 3.74035408592314,
+      "learning_rate": 8.886873605277822e-06,
+      "loss": 0.4254,
+      "step": 3458
+    },
+    {
+      "epoch": 0.24017497569781976,
+      "grad_norm": 4.428535303961577,
+      "learning_rate": 8.886166161971404e-06,
+      "loss": 0.6783,
+      "step": 3459
+    },
+    {
+      "epoch": 0.24024441049854187,
+      "grad_norm": 4.413689275496993,
+      "learning_rate": 8.885458522107102e-06,
+      "loss": 0.6705,
+      "step": 3460
+    },
+    {
+      "epoch": 0.240313845299264,
+      "grad_norm": 3.0964466053211765,
+      "learning_rate": 8.884750685720703e-06,
+      "loss": 0.3354,
+      "step": 3461
+    },
+    {
+      "epoch": 0.2403832800999861,
+      "grad_norm": 4.743670843049077,
+      "learning_rate": 8.884042652848011e-06,
+      "loss": 0.6877,
+      "step": 3462
+    },
+    {
+      "epoch": 0.24045271490070824,
+      "grad_norm": 3.5843044436040934,
+      "learning_rate": 8.883334423524837e-06,
+      "loss": 0.3225,
+      "step": 3463
+    },
+    {
+      "epoch": 0.24052214970143035,
+      "grad_norm": 3.992543955167913,
+      "learning_rate": 8.882625997787e-06,
+      "loss": 0.4635,
+      "step": 3464
+    },
+    {
+      "epoch": 0.24059158450215248,
+      "grad_norm": 4.101986971823068,
+      "learning_rate": 8.881917375670336e-06,
+      "loss": 0.4948,
+      "step": 3465
+    },
+    {
+      "epoch": 0.2406610193028746,
+      "grad_norm": 4.0560980395492425,
+      "learning_rate": 8.881208557210683e-06,
+      "loss": 0.339,
+      "step": 3466
+    },
+    {
+      "epoch": 0.24073045410359672,
+      "grad_norm": 4.285602605052266,
+      "learning_rate": 8.880499542443892e-06,
+      "loss": 0.3424,
+      "step": 3467
+    },
+    {
+      "epoch": 0.24079988890431883,
+      "grad_norm": 4.03833138974363,
+      "learning_rate": 8.879790331405827e-06,
+      "loss": 0.6373,
+      "step": 3468
+    },
+    {
+      "epoch": 0.24086932370504097,
+      "grad_norm": 4.343961404758944,
+      "learning_rate": 8.879080924132357e-06,
+      "loss": 0.6724,
+      "step": 3469
+    },
+    {
+      "epoch": 0.2409387585057631,
+      "grad_norm": 2.76921893582247,
+      "learning_rate": 8.878371320659362e-06,
+      "loss": 0.2177,
+      "step": 3470
+    },
+    {
+      "epoch": 0.2410081933064852,
+      "grad_norm": 3.5264093015772207,
+      "learning_rate": 8.877661521022735e-06,
+      "loss": 0.3222,
+      "step": 3471
+    },
+    {
+      "epoch": 0.24107762810720734,
+      "grad_norm": 3.231164985060828,
+      "learning_rate": 8.876951525258377e-06,
+      "loss": 0.4275,
+      "step": 3472
+    },
+    {
+      "epoch": 0.24114706290792945,
+      "grad_norm": 3.629759068444725,
+      "learning_rate": 8.876241333402197e-06,
+      "loss": 0.3833,
+      "step": 3473
+    },
+    {
+      "epoch": 0.24121649770865158,
+      "grad_norm": 3.5329451781239816,
+      "learning_rate": 8.875530945490118e-06,
+      "loss": 0.4025,
+      "step": 3474
+    },
+    {
+      "epoch": 0.2412859325093737,
+      "grad_norm": 3.8651259222070564,
+      "learning_rate": 8.874820361558068e-06,
+      "loss": 0.4571,
+      "step": 3475
+    },
+    {
+      "epoch": 0.24135536731009583,
+      "grad_norm": 4.584271283635883,
+      "learning_rate": 8.874109581641987e-06,
+      "loss": 0.6936,
+      "step": 3476
+    },
+    {
+      "epoch": 0.24142480211081793,
+      "grad_norm": 4.75277884644892,
+      "learning_rate": 8.87339860577783e-06,
+      "loss": 0.6095,
+      "step": 3477
+    },
+    {
+      "epoch": 0.24149423691154007,
+      "grad_norm": 3.2282265627761513,
+      "learning_rate": 8.872687434001554e-06,
+      "loss": 0.3934,
+      "step": 3478
+    },
+    {
+      "epoch": 0.24156367171226217,
+      "grad_norm": 3.9686150297828546,
+      "learning_rate": 8.871976066349131e-06,
+      "loss": 0.6361,
+      "step": 3479
+    },
+    {
+      "epoch": 0.2416331065129843,
+      "grad_norm": 3.6858671517398953,
+      "learning_rate": 8.871264502856538e-06,
+      "loss": 0.4687,
+      "step": 3480
+    },
+    {
+      "epoch": 0.24170254131370644,
+      "grad_norm": 3.1962732865078567,
+      "learning_rate": 8.870552743559766e-06,
+      "loss": 0.3877,
+      "step": 3481
+    },
+    {
+      "epoch": 0.24177197611442855,
+      "grad_norm": 5.115025850842917,
+      "learning_rate": 8.86984078849482e-06,
+      "loss": 0.5889,
+      "step": 3482
+    },
+    {
+      "epoch": 0.24184141091515068,
+      "grad_norm": 2.7936324764430127,
+      "learning_rate": 8.869128637697702e-06,
+      "loss": 0.3243,
+      "step": 3483
+    },
+    {
+      "epoch": 0.2419108457158728,
+      "grad_norm": 4.525926732330099,
+      "learning_rate": 8.868416291204437e-06,
+      "loss": 0.5303,
+      "step": 3484
+    },
+    {
+      "epoch": 0.24198028051659493,
+      "grad_norm": 5.115483855169774,
+      "learning_rate": 8.867703749051054e-06,
+      "loss": 0.3235,
+      "step": 3485
+    },
+    {
+      "epoch": 0.24204971531731703,
+      "grad_norm": 3.749797035021384,
+      "learning_rate": 8.86699101127359e-06,
+      "loss": 0.6413,
+      "step": 3486
+    },
+    {
+      "epoch": 0.24211915011803917,
+      "grad_norm": 2.3784005825318384,
+      "learning_rate": 8.8662780779081e-06,
+      "loss": 0.1788,
+      "step": 3487
+    },
+    {
+      "epoch": 0.24218858491876127,
+      "grad_norm": 4.448423098383032,
+      "learning_rate": 8.865564948990637e-06,
+      "loss": 0.4768,
+      "step": 3488
+    },
+    {
+      "epoch": 0.2422580197194834,
+      "grad_norm": 4.008890554114843,
+      "learning_rate": 8.864851624557275e-06,
+      "loss": 0.6424,
+      "step": 3489
+    },
+    {
+      "epoch": 0.24232745452020552,
+      "grad_norm": 3.717288120750269,
+      "learning_rate": 8.86413810464409e-06,
+      "loss": 0.5609,
+      "step": 3490
+    },
+    {
+      "epoch": 0.24239688932092765,
+      "grad_norm": 5.368235057613119,
+      "learning_rate": 8.863424389287172e-06,
+      "loss": 0.6841,
+      "step": 3491
+    },
+    {
+      "epoch": 0.24246632412164978,
+      "grad_norm": 4.556455129869119,
+      "learning_rate": 8.86271047852262e-06,
+      "loss": 0.828,
+      "step": 3492
+    },
+    {
+      "epoch": 0.2425357589223719,
+      "grad_norm": 4.923971417499381,
+      "learning_rate": 8.861996372386544e-06,
+      "loss": 0.507,
+      "step": 3493
+    },
+    {
+      "epoch": 0.24260519372309403,
+      "grad_norm": 5.507383727474723,
+      "learning_rate": 8.861282070915062e-06,
+      "loss": 0.9531,
+      "step": 3494
+    },
+    {
+      "epoch": 0.24267462852381613,
+      "grad_norm": 4.001556448279561,
+      "learning_rate": 8.860567574144301e-06,
+      "loss": 0.62,
+      "step": 3495
+    },
+    {
+      "epoch": 0.24274406332453827,
+      "grad_norm": 5.171409257461122,
+      "learning_rate": 8.8598528821104e-06,
+      "loss": 0.6338,
+      "step": 3496
+    },
+    {
+      "epoch": 0.24281349812526037,
+      "grad_norm": 3.7008047484375424,
+      "learning_rate": 8.85913799484951e-06,
+      "loss": 0.5443,
+      "step": 3497
+    },
+    {
+      "epoch": 0.2428829329259825,
+      "grad_norm": 3.561741785124683,
+      "learning_rate": 8.858422912397785e-06,
+      "loss": 0.4276,
+      "step": 3498
+    },
+    {
+      "epoch": 0.24295236772670462,
+      "grad_norm": 4.36708606561379,
+      "learning_rate": 8.857707634791398e-06,
+      "loss": 0.6536,
+      "step": 3499
+    },
+    {
+      "epoch": 0.24302180252742675,
+      "grad_norm": 3.4563224610533343,
+      "learning_rate": 8.856992162066522e-06,
+      "loss": 0.4809,
+      "step": 3500
+    },
+    {
+      "epoch": 0.24309123732814886,
+      "grad_norm": 4.422443847250424,
+      "learning_rate": 8.856276494259345e-06,
+      "loss": 0.6138,
+      "step": 3501
+    },
+    {
+      "epoch": 0.243160672128871,
+      "grad_norm": 5.3523206109867765,
+      "learning_rate": 8.85556063140607e-06,
+      "loss": 0.6802,
+      "step": 3502
+    },
+    {
+      "epoch": 0.2432301069295931,
+      "grad_norm": 4.042651987928296,
+      "learning_rate": 8.8548445735429e-06,
+      "loss": 0.6427,
+      "step": 3503
+    },
+    {
+      "epoch": 0.24329954173031523,
+      "grad_norm": 3.472452935412435,
+      "learning_rate": 8.854128320706054e-06,
+      "loss": 0.3488,
+      "step": 3504
+    },
+    {
+      "epoch": 0.24336897653103737,
+      "grad_norm": 4.257518770969773,
+      "learning_rate": 8.853411872931758e-06,
+      "loss": 0.4204,
+      "step": 3505
+    },
+    {
+      "epoch": 0.24343841133175947,
+      "grad_norm": 4.236966141025333,
+      "learning_rate": 8.85269523025625e-06,
+      "loss": 0.6052,
+      "step": 3506
+    },
+    {
+      "epoch": 0.2435078461324816,
+      "grad_norm": 3.6469696271218295,
+      "learning_rate": 8.851978392715776e-06,
+      "loss": 0.4649,
+      "step": 3507
+    },
+    {
+      "epoch": 0.24357728093320372,
+      "grad_norm": 4.946733980228326,
+      "learning_rate": 8.851261360346596e-06,
+      "loss": 0.4172,
+      "step": 3508
+    },
+    {
+      "epoch": 0.24364671573392585,
+      "grad_norm": 11.474036104532104,
+      "learning_rate": 8.850544133184973e-06,
+      "loss": 0.5279,
+      "step": 3509
+    },
+    {
+      "epoch": 0.24371615053464796,
+      "grad_norm": 4.828537435557371,
+      "learning_rate": 8.849826711267185e-06,
+      "loss": 0.4246,
+      "step": 3510
+    },
+    {
+      "epoch": 0.2437855853353701,
+      "grad_norm": 4.196021493476455,
+      "learning_rate": 8.849109094629517e-06,
+      "loss": 0.4524,
+      "step": 3511
+    },
+    {
+      "epoch": 0.2438550201360922,
+      "grad_norm": 4.702989804911505,
+      "learning_rate": 8.848391283308269e-06,
+      "loss": 0.4107,
+      "step": 3512
+    },
+    {
+      "epoch": 0.24392445493681433,
+      "grad_norm": 5.169319101376446,
+      "learning_rate": 8.847673277339744e-06,
+      "loss": 0.3251,
+      "step": 3513
+    },
+    {
+      "epoch": 0.24399388973753644,
+      "grad_norm": 4.5793182066594005,
+      "learning_rate": 8.846955076760257e-06,
+      "loss": 0.5392,
+      "step": 3514
+    },
+    {
+      "epoch": 0.24406332453825857,
+      "grad_norm": 4.57674884002615,
+      "learning_rate": 8.846236681606137e-06,
+      "loss": 0.5417,
+      "step": 3515
+    },
+    {
+      "epoch": 0.2441327593389807,
+      "grad_norm": 4.017780847209307,
+      "learning_rate": 8.845518091913718e-06,
+      "loss": 0.5056,
+      "step": 3516
+    },
+    {
+      "epoch": 0.24420219413970282,
+      "grad_norm": 3.100508784587696,
+      "learning_rate": 8.844799307719346e-06,
+      "loss": 0.2766,
+      "step": 3517
+    },
+    {
+      "epoch": 0.24427162894042495,
+      "grad_norm": 3.100303861920758,
+      "learning_rate": 8.844080329059373e-06,
+      "loss": 0.3772,
+      "step": 3518
+    },
+    {
+      "epoch": 0.24434106374114706,
+      "grad_norm": 4.14780103820431,
+      "learning_rate": 8.84336115597017e-06,
+      "loss": 0.5875,
+      "step": 3519
+    },
+    {
+      "epoch": 0.2444104985418692,
+      "grad_norm": 5.198249679127238,
+      "learning_rate": 8.842641788488106e-06,
+      "loss": 0.544,
+      "step": 3520
+    },
+    {
+      "epoch": 0.2444799333425913,
+      "grad_norm": 4.1218422602524765,
+      "learning_rate": 8.84192222664957e-06,
+      "loss": 0.4322,
+      "step": 3521
+    },
+    {
+      "epoch": 0.24454936814331343,
+      "grad_norm": 4.611596621501159,
+      "learning_rate": 8.841202470490956e-06,
+      "loss": 0.3419,
+      "step": 3522
+    },
+    {
+      "epoch": 0.24461880294403554,
+      "grad_norm": 4.294724805593306,
+      "learning_rate": 8.840482520048665e-06,
+      "loss": 0.6214,
+      "step": 3523
+    },
+    {
+      "epoch": 0.24468823774475768,
+      "grad_norm": 4.397296886629712,
+      "learning_rate": 8.839762375359116e-06,
+      "loss": 0.5415,
+      "step": 3524
+    },
+    {
+      "epoch": 0.24475767254547978,
+      "grad_norm": 4.119329509408553,
+      "learning_rate": 8.83904203645873e-06,
+      "loss": 0.6755,
+      "step": 3525
+    },
+    {
+      "epoch": 0.24482710734620192,
+      "grad_norm": 2.690446440938124,
+      "learning_rate": 8.838321503383944e-06,
+      "loss": 0.1978,
+      "step": 3526
+    },
+    {
+      "epoch": 0.24489654214692405,
+      "grad_norm": 4.663776912336376,
+      "learning_rate": 8.837600776171197e-06,
+      "loss": 0.6693,
+      "step": 3527
+    },
+    {
+      "epoch": 0.24496597694764616,
+      "grad_norm": 3.603460056303718,
+      "learning_rate": 8.836879854856947e-06,
+      "loss": 0.4874,
+      "step": 3528
+    },
+    {
+      "epoch": 0.2450354117483683,
+      "grad_norm": 3.9241289718369714,
+      "learning_rate": 8.836158739477656e-06,
+      "loss": 0.3979,
+      "step": 3529
+    },
+    {
+      "epoch": 0.2451048465490904,
+      "grad_norm": 3.86377067525096,
+      "learning_rate": 8.835437430069796e-06,
+      "loss": 0.4273,
+      "step": 3530
+    },
+    {
+      "epoch": 0.24517428134981253,
+      "grad_norm": 3.69691209136962,
+      "learning_rate": 8.83471592666985e-06,
+      "loss": 0.4975,
+      "step": 3531
+    },
+    {
+      "epoch": 0.24524371615053464,
+      "grad_norm": 4.470364631154342,
+      "learning_rate": 8.833994229314313e-06,
+      "loss": 0.5501,
+      "step": 3532
+    },
+    {
+      "epoch": 0.24531315095125678,
+      "grad_norm": 4.146740815152786,
+      "learning_rate": 8.833272338039686e-06,
+      "loss": 0.4969,
+      "step": 3533
+    },
+    {
+      "epoch": 0.24538258575197888,
+      "grad_norm": 3.5208109814199,
+      "learning_rate": 8.832550252882482e-06,
+      "loss": 0.3429,
+      "step": 3534
+    },
+    {
+      "epoch": 0.24545202055270102,
+      "grad_norm": 3.7044745887001223,
+      "learning_rate": 8.831827973879224e-06,
+      "loss": 0.4798,
+      "step": 3535
+    },
+    {
+      "epoch": 0.24552145535342312,
+      "grad_norm": 4.324288618207141,
+      "learning_rate": 8.83110550106644e-06,
+      "loss": 0.3887,
+      "step": 3536
+    },
+    {
+      "epoch": 0.24559089015414526,
+      "grad_norm": 5.115329305867889,
+      "learning_rate": 8.830382834480678e-06,
+      "loss": 0.7651,
+      "step": 3537
+    },
+    {
+      "epoch": 0.2456603249548674,
+      "grad_norm": 4.4921222253130475,
+      "learning_rate": 8.829659974158483e-06,
+      "loss": 0.4086,
+      "step": 3538
+    },
+    {
+      "epoch": 0.2457297597555895,
+      "grad_norm": 3.568785533279242,
+      "learning_rate": 8.828936920136424e-06,
+      "loss": 0.5179,
+      "step": 3539
+    },
+    {
+      "epoch": 0.24579919455631163,
+      "grad_norm": 4.524499382115759,
+      "learning_rate": 8.828213672451066e-06,
+      "loss": 0.5334,
+      "step": 3540
+    },
+    {
+      "epoch": 0.24586862935703374,
+      "grad_norm": 3.5895766528792445,
+      "learning_rate": 8.827490231138993e-06,
+      "loss": 0.4367,
+      "step": 3541
+    },
+    {
+      "epoch": 0.24593806415775588,
+      "grad_norm": 3.366393593268394,
+      "learning_rate": 8.826766596236797e-06,
+      "loss": 0.3875,
+      "step": 3542
+    },
+    {
+      "epoch": 0.24600749895847798,
+      "grad_norm": 3.21275650851757,
+      "learning_rate": 8.826042767781075e-06,
+      "loss": 0.2262,
+      "step": 3543
+    },
+    {
+      "epoch": 0.24607693375920012,
+      "grad_norm": 4.645932879839337,
+      "learning_rate": 8.82531874580844e-06,
+      "loss": 0.4215,
+      "step": 3544
+    },
+    {
+      "epoch": 0.24614636855992222,
+      "grad_norm": 3.210595562013852,
+      "learning_rate": 8.82459453035551e-06,
+      "loss": 0.3727,
+      "step": 3545
+    },
+    {
+      "epoch": 0.24621580336064436,
+      "grad_norm": 4.294531871288308,
+      "learning_rate": 8.823870121458919e-06,
+      "loss": 0.4362,
+      "step": 3546
+    },
+    {
+      "epoch": 0.24628523816136647,
+      "grad_norm": 5.672633157989422,
+      "learning_rate": 8.823145519155303e-06,
+      "loss": 0.9266,
+      "step": 3547
+    },
+    {
+      "epoch": 0.2463546729620886,
+      "grad_norm": 4.065258773350984,
+      "learning_rate": 8.822420723481312e-06,
+      "loss": 0.4598,
+      "step": 3548
+    },
+    {
+      "epoch": 0.24642410776281073,
+      "grad_norm": 3.3069900874278564,
+      "learning_rate": 8.821695734473607e-06,
+      "loss": 0.4232,
+      "step": 3549
+    },
+    {
+      "epoch": 0.24649354256353284,
+      "grad_norm": 4.030398967772835,
+      "learning_rate": 8.820970552168854e-06,
+      "loss": 0.435,
+      "step": 3550
+    },
+    {
+      "epoch": 0.24656297736425498,
+      "grad_norm": 3.3095629617353906,
+      "learning_rate": 8.820245176603737e-06,
+      "loss": 0.3608,
+      "step": 3551
+    },
+    {
+      "epoch": 0.24663241216497708,
+      "grad_norm": 5.006761832837181,
+      "learning_rate": 8.819519607814942e-06,
+      "loss": 0.4751,
+      "step": 3552
+    },
+    {
+      "epoch": 0.24670184696569922,
+      "grad_norm": 3.918058284618927,
+      "learning_rate": 8.818793845839166e-06,
+      "loss": 0.4008,
+      "step": 3553
+    },
+    {
+      "epoch": 0.24677128176642132,
+      "grad_norm": 4.884306130574844,
+      "learning_rate": 8.81806789071312e-06,
+      "loss": 0.3733,
+      "step": 3554
+    },
+    {
+      "epoch": 0.24684071656714346,
+      "grad_norm": 3.579634229650744,
+      "learning_rate": 8.817341742473518e-06,
+      "loss": 0.4559,
+      "step": 3555
+    },
+    {
+      "epoch": 0.24691015136786557,
+      "grad_norm": 4.7129044348561315,
+      "learning_rate": 8.816615401157091e-06,
+      "loss": 0.8162,
+      "step": 3556
+    },
+    {
+      "epoch": 0.2469795861685877,
+      "grad_norm": 4.6627084496188935,
+      "learning_rate": 8.815888866800577e-06,
+      "loss": 0.3568,
+      "step": 3557
+    },
+    {
+      "epoch": 0.2470490209693098,
+      "grad_norm": 3.5072675528545196,
+      "learning_rate": 8.815162139440723e-06,
+      "loss": 0.4519,
+      "step": 3558
+    },
+    {
+      "epoch": 0.24711845577003194,
+      "grad_norm": 4.301887818243652,
+      "learning_rate": 8.814435219114283e-06,
+      "loss": 0.4241,
+      "step": 3559
+    },
+    {
+      "epoch": 0.24718789057075405,
+      "grad_norm": 4.132318148586803,
+      "learning_rate": 8.813708105858028e-06,
+      "loss": 0.4976,
+      "step": 3560
+    },
+    {
+      "epoch": 0.24725732537147618,
+      "grad_norm": 3.962017022498468,
+      "learning_rate": 8.81298079970873e-06,
+      "loss": 0.4165,
+      "step": 3561
+    },
+    {
+      "epoch": 0.24732676017219832,
+      "grad_norm": 3.5190888435760024,
+      "learning_rate": 8.81225330070318e-06,
+      "loss": 0.3577,
+      "step": 3562
+    },
+    {
+      "epoch": 0.24739619497292042,
+      "grad_norm": 3.2543413803605206,
+      "learning_rate": 8.81152560887817e-06,
+      "loss": 0.2487,
+      "step": 3563
+    },
+    {
+      "epoch": 0.24746562977364256,
+      "grad_norm": 3.230279552489996,
+      "learning_rate": 8.810797724270508e-06,
+      "loss": 0.3464,
+      "step": 3564
+    },
+    {
+      "epoch": 0.24753506457436467,
+      "grad_norm": 3.9236631580115295,
+      "learning_rate": 8.81006964691701e-06,
+      "loss": 0.504,
+      "step": 3565
+    },
+    {
+      "epoch": 0.2476044993750868,
+      "grad_norm": 3.023801008147578,
+      "learning_rate": 8.809341376854502e-06,
+      "loss": 0.3255,
+      "step": 3566
+    },
+    {
+      "epoch": 0.2476739341758089,
+      "grad_norm": 3.9637064407598244,
+      "learning_rate": 8.808612914119817e-06,
+      "loss": 0.4151,
+      "step": 3567
+    },
+    {
+      "epoch": 0.24774336897653104,
+      "grad_norm": 2.906557934543004,
+      "learning_rate": 8.807884258749799e-06,
+      "loss": 0.3187,
+      "step": 3568
+    },
+    {
+      "epoch": 0.24781280377725315,
+      "grad_norm": 4.099038071958367,
+      "learning_rate": 8.807155410781306e-06,
+      "loss": 0.5943,
+      "step": 3569
+    },
+    {
+      "epoch": 0.24788223857797528,
+      "grad_norm": 3.39973451392086,
+      "learning_rate": 8.806426370251198e-06,
+      "loss": 0.4191,
+      "step": 3570
+    },
+    {
+      "epoch": 0.2479516733786974,
+      "grad_norm": 4.495282181749335,
+      "learning_rate": 8.805697137196354e-06,
+      "loss": 0.6804,
+      "step": 3571
+    },
+    {
+      "epoch": 0.24802110817941952,
+      "grad_norm": 5.832171965523165,
+      "learning_rate": 8.804967711653654e-06,
+      "loss": 0.6672,
+      "step": 3572
+    },
+    {
+      "epoch": 0.24809054298014166,
+      "grad_norm": 4.71914727258289,
+      "learning_rate": 8.80423809365999e-06,
+      "loss": 0.4282,
+      "step": 3573
+    },
+    {
+      "epoch": 0.24815997778086377,
+      "grad_norm": 2.6071709743470697,
+      "learning_rate": 8.803508283252273e-06,
+      "loss": 0.234,
+      "step": 3574
+    },
+    {
+      "epoch": 0.2482294125815859,
+      "grad_norm": 3.8902674295237625,
+      "learning_rate": 8.802778280467408e-06,
+      "loss": 0.552,
+      "step": 3575
+    },
+    {
+      "epoch": 0.248298847382308,
+      "grad_norm": 4.594274108845213,
+      "learning_rate": 8.802048085342323e-06,
+      "loss": 0.499,
+      "step": 3576
+    },
+    {
+      "epoch": 0.24836828218303014,
+      "grad_norm": 4.561424017198423,
+      "learning_rate": 8.801317697913945e-06,
+      "loss": 0.5721,
+      "step": 3577
+    },
+    {
+      "epoch": 0.24843771698375225,
+      "grad_norm": 4.684597155688811,
+      "learning_rate": 8.80058711821922e-06,
+      "loss": 0.4497,
+      "step": 3578
+    },
+    {
+      "epoch": 0.24850715178447438,
+      "grad_norm": 4.287908890262907,
+      "learning_rate": 8.799856346295102e-06,
+      "loss": 0.3877,
+      "step": 3579
+    },
+    {
+      "epoch": 0.2485765865851965,
+      "grad_norm": 5.083252578486347,
+      "learning_rate": 8.799125382178549e-06,
+      "loss": 0.5935,
+      "step": 3580
+    },
+    {
+      "epoch": 0.24864602138591863,
+      "grad_norm": 4.73830918312085,
+      "learning_rate": 8.798394225906532e-06,
+      "loss": 0.3947,
+      "step": 3581
+    },
+    {
+      "epoch": 0.24871545618664073,
+      "grad_norm": 5.015842549006712,
+      "learning_rate": 8.797662877516033e-06,
+      "loss": 0.8407,
+      "step": 3582
+    },
+    {
+      "epoch": 0.24878489098736287,
+      "grad_norm": 3.7213335249810404,
+      "learning_rate": 8.796931337044044e-06,
+      "loss": 0.4483,
+      "step": 3583
+    },
+    {
+      "epoch": 0.248854325788085,
+      "grad_norm": 3.151143230861929,
+      "learning_rate": 8.796199604527565e-06,
+      "loss": 0.3466,
+      "step": 3584
+    },
+    {
+      "epoch": 0.2489237605888071,
+      "grad_norm": 4.06314198418967,
+      "learning_rate": 8.795467680003604e-06,
+      "loss": 0.4826,
+      "step": 3585
+    },
+    {
+      "epoch": 0.24899319538952924,
+      "grad_norm": 3.6776099803374422,
+      "learning_rate": 8.794735563509183e-06,
+      "loss": 0.4429,
+      "step": 3586
+    },
+    {
+      "epoch": 0.24906263019025135,
+      "grad_norm": 4.917035515170583,
+      "learning_rate": 8.794003255081332e-06,
+      "loss": 0.7712,
+      "step": 3587
+    },
+    {
+      "epoch": 0.24913206499097348,
+      "grad_norm": 2.4468320380180373,
+      "learning_rate": 8.793270754757089e-06,
+      "loss": 0.0983,
+      "step": 3588
+    },
+    {
+      "epoch": 0.2492014997916956,
+      "grad_norm": 4.398217835711658,
+      "learning_rate": 8.792538062573503e-06,
+      "loss": 0.4024,
+      "step": 3589
+    },
+    {
+      "epoch": 0.24927093459241773,
+      "grad_norm": 4.822895929890573,
+      "learning_rate": 8.791805178567633e-06,
+      "loss": 0.7639,
+      "step": 3590
+    },
+    {
+      "epoch": 0.24934036939313983,
+      "grad_norm": 4.625758633902606,
+      "learning_rate": 8.79107210277655e-06,
+      "loss": 0.7146,
+      "step": 3591
+    },
+    {
+      "epoch": 0.24940980419386197,
+      "grad_norm": 3.2865196102545786,
+      "learning_rate": 8.79033883523733e-06,
+      "loss": 0.3829,
+      "step": 3592
+    },
+    {
+      "epoch": 0.24947923899458407,
+      "grad_norm": 3.3042104829087036,
+      "learning_rate": 8.789605375987058e-06,
+      "loss": 0.2158,
+      "step": 3593
+    },
+    {
+      "epoch": 0.2495486737953062,
+      "grad_norm": 3.6189334451907618,
+      "learning_rate": 8.788871725062836e-06,
+      "loss": 0.3905,
+      "step": 3594
+    },
+    {
+      "epoch": 0.24961810859602834,
+      "grad_norm": 5.909954686312657,
+      "learning_rate": 8.78813788250177e-06,
+      "loss": 0.8676,
+      "step": 3595
+    },
+    {
+      "epoch": 0.24968754339675045,
+      "grad_norm": 3.7474712826596486,
+      "learning_rate": 8.787403848340976e-06,
+      "loss": 0.5168,
+      "step": 3596
+    },
+    {
+      "epoch": 0.24975697819747258,
+      "grad_norm": 2.892563194478331,
+      "learning_rate": 8.78666962261758e-06,
+      "loss": 0.3759,
+      "step": 3597
+    },
+    {
+      "epoch": 0.2498264129981947,
+      "grad_norm": 3.540121142107857,
+      "learning_rate": 8.78593520536872e-06,
+      "loss": 0.6119,
+      "step": 3598
+    },
+    {
+      "epoch": 0.24989584779891683,
+      "grad_norm": 4.144610718526077,
+      "learning_rate": 8.785200596631543e-06,
+      "loss": 0.4343,
+      "step": 3599
+    },
+    {
+      "epoch": 0.24996528259963893,
+      "grad_norm": 4.000802200521305,
+      "learning_rate": 8.784465796443201e-06,
+      "loss": 0.5409,
+      "step": 3600
+    },
+    {
+      "epoch": 0.25003471740036104,
+      "grad_norm": 3.5132985526432554,
+      "learning_rate": 8.783730804840861e-06,
+      "loss": 0.3216,
+      "step": 3601
+    },
+    {
+      "epoch": 0.2501041522010832,
+      "grad_norm": 4.502903541809305,
+      "learning_rate": 8.782995621861702e-06,
+      "loss": 0.664,
+      "step": 3602
+    },
+    {
+      "epoch": 0.2501735870018053,
+      "grad_norm": 3.102606855236283,
+      "learning_rate": 8.782260247542902e-06,
+      "loss": 0.3258,
+      "step": 3603
+    },
+    {
+      "epoch": 0.25024302180252744,
+      "grad_norm": 4.597347976442723,
+      "learning_rate": 8.78152468192166e-06,
+      "loss": 0.5729,
+      "step": 3604
+    },
+    {
+      "epoch": 0.2503124566032495,
+      "grad_norm": 5.805953369370509,
+      "learning_rate": 8.780788925035178e-06,
+      "loss": 0.6389,
+      "step": 3605
+    },
+    {
+      "epoch": 0.25038189140397166,
+      "grad_norm": 2.92902719939337,
+      "learning_rate": 8.78005297692067e-06,
+      "loss": 0.2813,
+      "step": 3606
+    },
+    {
+      "epoch": 0.2504513262046938,
+      "grad_norm": 3.937389409005245,
+      "learning_rate": 8.77931683761536e-06,
+      "loss": 0.6427,
+      "step": 3607
+    },
+    {
+      "epoch": 0.2505207610054159,
+      "grad_norm": 4.649954172796734,
+      "learning_rate": 8.778580507156482e-06,
+      "loss": 0.4159,
+      "step": 3608
+    },
+    {
+      "epoch": 0.25059019580613806,
+      "grad_norm": 3.838746173915422,
+      "learning_rate": 8.777843985581278e-06,
+      "loss": 0.3741,
+      "step": 3609
+    },
+    {
+      "epoch": 0.25065963060686014,
+      "grad_norm": 3.549208890333452,
+      "learning_rate": 8.777107272926999e-06,
+      "loss": 0.4535,
+      "step": 3610
+    },
+    {
+      "epoch": 0.2507290654075823,
+      "grad_norm": 3.859925793358068,
+      "learning_rate": 8.776370369230909e-06,
+      "loss": 0.4632,
+      "step": 3611
+    },
+    {
+      "epoch": 0.2507985002083044,
+      "grad_norm": 4.610844773680847,
+      "learning_rate": 8.775633274530278e-06,
+      "loss": 0.622,
+      "step": 3612
+    },
+    {
+      "epoch": 0.25086793500902654,
+      "grad_norm": 4.139742428804724,
+      "learning_rate": 8.774895988862388e-06,
+      "loss": 0.4407,
+      "step": 3613
+    },
+    {
+      "epoch": 0.2509373698097486,
+      "grad_norm": 3.943339084228374,
+      "learning_rate": 8.774158512264532e-06,
+      "loss": 0.5416,
+      "step": 3614
+    },
+    {
+      "epoch": 0.25100680461047076,
+      "grad_norm": 3.230383698727962,
+      "learning_rate": 8.773420844774008e-06,
+      "loss": 0.3282,
+      "step": 3615
+    },
+    {
+      "epoch": 0.2510762394111929,
+      "grad_norm": 4.323247617295156,
+      "learning_rate": 8.772682986428127e-06,
+      "loss": 0.6489,
+      "step": 3616
+    },
+    {
+      "epoch": 0.251145674211915,
+      "grad_norm": 2.932844242013293,
+      "learning_rate": 8.771944937264211e-06,
+      "loss": 0.2246,
+      "step": 3617
+    },
+    {
+      "epoch": 0.25121510901263716,
+      "grad_norm": 3.36799511813649,
+      "learning_rate": 8.771206697319589e-06,
+      "loss": 0.525,
+      "step": 3618
+    },
+    {
+      "epoch": 0.25128454381335924,
+      "grad_norm": 4.053378883153455,
+      "learning_rate": 8.770468266631597e-06,
+      "loss": 0.4624,
+      "step": 3619
+    },
+    {
+      "epoch": 0.2513539786140814,
+      "grad_norm": 3.7672796058655966,
+      "learning_rate": 8.76972964523759e-06,
+      "loss": 0.5482,
+      "step": 3620
+    },
+    {
+      "epoch": 0.2514234134148035,
+      "grad_norm": 4.479030793614153,
+      "learning_rate": 8.768990833174922e-06,
+      "loss": 0.8307,
+      "step": 3621
+    },
+    {
+      "epoch": 0.25149284821552564,
+      "grad_norm": 4.474930832897893,
+      "learning_rate": 8.768251830480961e-06,
+      "loss": 0.5424,
+      "step": 3622
+    },
+    {
+      "epoch": 0.2515622830162477,
+      "grad_norm": 4.492412094694208,
+      "learning_rate": 8.767512637193087e-06,
+      "loss": 0.4893,
+      "step": 3623
+    },
+    {
+      "epoch": 0.25163171781696986,
+      "grad_norm": 3.184243977779598,
+      "learning_rate": 8.766773253348688e-06,
+      "loss": 0.3811,
+      "step": 3624
+    },
+    {
+      "epoch": 0.251701152617692,
+      "grad_norm": 4.637354344562509,
+      "learning_rate": 8.76603367898516e-06,
+      "loss": 0.5443,
+      "step": 3625
+    },
+    {
+      "epoch": 0.2517705874184141,
+      "grad_norm": 4.079590444048545,
+      "learning_rate": 8.76529391413991e-06,
+      "loss": 0.5648,
+      "step": 3626
+    },
+    {
+      "epoch": 0.2518400222191362,
+      "grad_norm": 3.0484044303861886,
+      "learning_rate": 8.764553958850355e-06,
+      "loss": 0.3502,
+      "step": 3627
+    },
+    {
+      "epoch": 0.25190945701985834,
+      "grad_norm": 3.0354564220285494,
+      "learning_rate": 8.76381381315392e-06,
+      "loss": 0.4384,
+      "step": 3628
+    },
+    {
+      "epoch": 0.2519788918205805,
+      "grad_norm": 4.366812456871676,
+      "learning_rate": 8.763073477088042e-06,
+      "loss": 0.5984,
+      "step": 3629
+    },
+    {
+      "epoch": 0.2520483266213026,
+      "grad_norm": 4.226622389450282,
+      "learning_rate": 8.762332950690164e-06,
+      "loss": 0.4704,
+      "step": 3630
+    },
+    {
+      "epoch": 0.25211776142202474,
+      "grad_norm": 4.040621303424189,
+      "learning_rate": 8.761592233997745e-06,
+      "loss": 0.4489,
+      "step": 3631
+    },
+    {
+      "epoch": 0.2521871962227468,
+      "grad_norm": 4.203514471193115,
+      "learning_rate": 8.760851327048246e-06,
+      "loss": 0.5196,
+      "step": 3632
+    },
+    {
+      "epoch": 0.25225663102346896,
+      "grad_norm": 4.2215117460974,
+      "learning_rate": 8.760110229879141e-06,
+      "loss": 0.8041,
+      "step": 3633
+    },
+    {
+      "epoch": 0.2523260658241911,
+      "grad_norm": 3.946850381200541,
+      "learning_rate": 8.75936894252792e-06,
+      "loss": 0.4112,
+      "step": 3634
+    },
+    {
+      "epoch": 0.2523955006249132,
+      "grad_norm": 3.7897335125099327,
+      "learning_rate": 8.758627465032067e-06,
+      "loss": 0.4262,
+      "step": 3635
+    },
+    {
+      "epoch": 0.2524649354256353,
+      "grad_norm": 2.8736133562246757,
+      "learning_rate": 8.757885797429094e-06,
+      "loss": 0.2942,
+      "step": 3636
+    },
+    {
+      "epoch": 0.25253437022635744,
+      "grad_norm": 3.785259565516119,
+      "learning_rate": 8.757143939756508e-06,
+      "loss": 0.4697,
+      "step": 3637
+    },
+    {
+      "epoch": 0.2526038050270796,
+      "grad_norm": 4.08405023953646,
+      "learning_rate": 8.756401892051833e-06,
+      "loss": 0.5857,
+      "step": 3638
+    },
+    {
+      "epoch": 0.2526732398278017,
+      "grad_norm": 4.0307755030621735,
+      "learning_rate": 8.755659654352599e-06,
+      "loss": 0.4941,
+      "step": 3639
+    },
+    {
+      "epoch": 0.2527426746285238,
+      "grad_norm": 3.9516246259651386,
+      "learning_rate": 8.754917226696354e-06,
+      "loss": 0.4915,
+      "step": 3640
+    },
+    {
+      "epoch": 0.2528121094292459,
+      "grad_norm": 4.884640595768288,
+      "learning_rate": 8.754174609120641e-06,
+      "loss": 0.4803,
+      "step": 3641
+    },
+    {
+      "epoch": 0.25288154422996806,
+      "grad_norm": 4.212373541346412,
+      "learning_rate": 8.753431801663026e-06,
+      "loss": 0.571,
+      "step": 3642
+    },
+    {
+      "epoch": 0.2529509790306902,
+      "grad_norm": 4.202355914645096,
+      "learning_rate": 8.752688804361076e-06,
+      "loss": 0.5444,
+      "step": 3643
+    },
+    {
+      "epoch": 0.2530204138314123,
+      "grad_norm": 3.0879241793157015,
+      "learning_rate": 8.751945617252375e-06,
+      "loss": 0.428,
+      "step": 3644
+    },
+    {
+      "epoch": 0.2530898486321344,
+      "grad_norm": 4.728177083807009,
+      "learning_rate": 8.751202240374509e-06,
+      "loss": 0.6275,
+      "step": 3645
+    },
+    {
+      "epoch": 0.25315928343285654,
+      "grad_norm": 4.118359739641388,
+      "learning_rate": 8.750458673765079e-06,
+      "loss": 0.2562,
+      "step": 3646
+    },
+    {
+      "epoch": 0.2532287182335787,
+      "grad_norm": 3.295194792987422,
+      "learning_rate": 8.74971491746169e-06,
+      "loss": 0.418,
+      "step": 3647
+    },
+    {
+      "epoch": 0.2532981530343008,
+      "grad_norm": 3.6907412271185565,
+      "learning_rate": 8.748970971501969e-06,
+      "loss": 0.2962,
+      "step": 3648
+    },
+    {
+      "epoch": 0.2533675878350229,
+      "grad_norm": 4.381140820125764,
+      "learning_rate": 8.748226835923536e-06,
+      "loss": 0.4977,
+      "step": 3649
+    },
+    {
+      "epoch": 0.253437022635745,
+      "grad_norm": 3.9172406466866456,
+      "learning_rate": 8.747482510764031e-06,
+      "loss": 0.5541,
+      "step": 3650
+    },
+    {
+      "epoch": 0.25350645743646716,
+      "grad_norm": 3.405115914085888,
+      "learning_rate": 8.746737996061101e-06,
+      "loss": 0.4783,
+      "step": 3651
+    },
+    {
+      "epoch": 0.2535758922371893,
+      "grad_norm": 3.7074751347279156,
+      "learning_rate": 8.745993291852403e-06,
+      "loss": 0.4945,
+      "step": 3652
+    },
+    {
+      "epoch": 0.25364532703791143,
+      "grad_norm": 4.250690387344984,
+      "learning_rate": 8.745248398175603e-06,
+      "loss": 0.4796,
+      "step": 3653
+    },
+    {
+      "epoch": 0.2537147618386335,
+      "grad_norm": 3.7414967435958997,
+      "learning_rate": 8.744503315068378e-06,
+      "loss": 0.368,
+      "step": 3654
+    },
+    {
+      "epoch": 0.25378419663935564,
+      "grad_norm": 4.075277359632381,
+      "learning_rate": 8.743758042568411e-06,
+      "loss": 0.3835,
+      "step": 3655
+    },
+    {
+      "epoch": 0.2538536314400778,
+      "grad_norm": 3.8935116917625407,
+      "learning_rate": 8.743012580713399e-06,
+      "loss": 0.4072,
+      "step": 3656
+    },
+    {
+      "epoch": 0.2539230662407999,
+      "grad_norm": 5.225153267260068,
+      "learning_rate": 8.742266929541047e-06,
+      "loss": 0.4731,
+      "step": 3657
+    },
+    {
+      "epoch": 0.253992501041522,
+      "grad_norm": 4.466770558968311,
+      "learning_rate": 8.741521089089067e-06,
+      "loss": 0.6497,
+      "step": 3658
+    },
+    {
+      "epoch": 0.2540619358422441,
+      "grad_norm": 3.7053352967803677,
+      "learning_rate": 8.740775059395186e-06,
+      "loss": 0.5237,
+      "step": 3659
+    },
+    {
+      "epoch": 0.25413137064296626,
+      "grad_norm": 3.57390461731123,
+      "learning_rate": 8.740028840497135e-06,
+      "loss": 0.4618,
+      "step": 3660
+    },
+    {
+      "epoch": 0.2542008054436884,
+      "grad_norm": 3.724980180152874,
+      "learning_rate": 8.739282432432656e-06,
+      "loss": 0.5086,
+      "step": 3661
+    },
+    {
+      "epoch": 0.2542702402444105,
+      "grad_norm": 4.675003657936353,
+      "learning_rate": 8.738535835239503e-06,
+      "loss": 0.7822,
+      "step": 3662
+    },
+    {
+      "epoch": 0.2543396750451326,
+      "grad_norm": 4.16525664946798,
+      "learning_rate": 8.737789048955441e-06,
+      "loss": 0.5575,
+      "step": 3663
+    },
+    {
+      "epoch": 0.25440910984585474,
+      "grad_norm": 3.482066131114158,
+      "learning_rate": 8.737042073618235e-06,
+      "loss": 0.3992,
+      "step": 3664
+    },
+    {
+      "epoch": 0.2544785446465769,
+      "grad_norm": 3.3182608724070213,
+      "learning_rate": 8.73629490926567e-06,
+      "loss": 0.4392,
+      "step": 3665
+    },
+    {
+      "epoch": 0.254547979447299,
+      "grad_norm": 4.1851825580433895,
+      "learning_rate": 8.735547555935538e-06,
+      "loss": 0.4706,
+      "step": 3666
+    },
+    {
+      "epoch": 0.2546174142480211,
+      "grad_norm": 3.5759274157686423,
+      "learning_rate": 8.734800013665636e-06,
+      "loss": 0.3779,
+      "step": 3667
+    },
+    {
+      "epoch": 0.2546868490487432,
+      "grad_norm": 4.036421183320925,
+      "learning_rate": 8.734052282493777e-06,
+      "loss": 0.4628,
+      "step": 3668
+    },
+    {
+      "epoch": 0.25475628384946536,
+      "grad_norm": 3.057088943827132,
+      "learning_rate": 8.73330436245778e-06,
+      "loss": 0.3298,
+      "step": 3669
+    },
+    {
+      "epoch": 0.2548257186501875,
+      "grad_norm": 3.9774513532894225,
+      "learning_rate": 8.73255625359547e-06,
+      "loss": 0.4872,
+      "step": 3670
+    },
+    {
+      "epoch": 0.2548951534509096,
+      "grad_norm": 3.9196377120040458,
+      "learning_rate": 8.731807955944688e-06,
+      "loss": 0.4498,
+      "step": 3671
+    },
+    {
+      "epoch": 0.2549645882516317,
+      "grad_norm": 2.9761825268456295,
+      "learning_rate": 8.731059469543285e-06,
+      "loss": 0.1846,
+      "step": 3672
+    },
+    {
+      "epoch": 0.25503402305235384,
+      "grad_norm": 3.692458981632782,
+      "learning_rate": 8.730310794429115e-06,
+      "loss": 0.556,
+      "step": 3673
+    },
+    {
+      "epoch": 0.255103457853076,
+      "grad_norm": 24.86622286813129,
+      "learning_rate": 8.729561930640048e-06,
+      "loss": 0.2963,
+      "step": 3674
+    },
+    {
+      "epoch": 0.2551728926537981,
+      "grad_norm": 3.851109028921879,
+      "learning_rate": 8.728812878213954e-06,
+      "loss": 0.3692,
+      "step": 3675
+    },
+    {
+      "epoch": 0.2552423274545202,
+      "grad_norm": 4.162831173573109,
+      "learning_rate": 8.728063637188728e-06,
+      "loss": 0.6184,
+      "step": 3676
+    },
+    {
+      "epoch": 0.2553117622552423,
+      "grad_norm": 3.6093935617377224,
+      "learning_rate": 8.727314207602261e-06,
+      "loss": 0.3999,
+      "step": 3677
+    },
+    {
+      "epoch": 0.25538119705596446,
+      "grad_norm": 4.285372398943112,
+      "learning_rate": 8.72656458949246e-06,
+      "loss": 0.4057,
+      "step": 3678
+    },
+    {
+      "epoch": 0.2554506318566866,
+      "grad_norm": 3.3743181014105144,
+      "learning_rate": 8.725814782897237e-06,
+      "loss": 0.4756,
+      "step": 3679
+    },
+    {
+      "epoch": 0.2555200666574087,
+      "grad_norm": 4.471286202051784,
+      "learning_rate": 8.72506478785452e-06,
+      "loss": 0.5942,
+      "step": 3680
+    },
+    {
+      "epoch": 0.2555895014581308,
+      "grad_norm": 5.726229220612755,
+      "learning_rate": 8.72431460440224e-06,
+      "loss": 0.7285,
+      "step": 3681
+    },
+    {
+      "epoch": 0.25565893625885294,
+      "grad_norm": 3.8798350765523564,
+      "learning_rate": 8.723564232578341e-06,
+      "loss": 0.6202,
+      "step": 3682
+    },
+    {
+      "epoch": 0.2557283710595751,
+      "grad_norm": 6.404690843120875,
+      "learning_rate": 8.722813672420777e-06,
+      "loss": 0.4851,
+      "step": 3683
+    },
+    {
+      "epoch": 0.25579780586029716,
+      "grad_norm": 4.297676062369375,
+      "learning_rate": 8.722062923967512e-06,
+      "loss": 0.6379,
+      "step": 3684
+    },
+    {
+      "epoch": 0.2558672406610193,
+      "grad_norm": 3.8458737779939236,
+      "learning_rate": 8.721311987256515e-06,
+      "loss": 0.52,
+      "step": 3685
+    },
+    {
+      "epoch": 0.2559366754617414,
+      "grad_norm": 4.669469687906569,
+      "learning_rate": 8.72056086232577e-06,
+      "loss": 0.7007,
+      "step": 3686
+    },
+    {
+      "epoch": 0.25600611026246356,
+      "grad_norm": 3.6919988310888714,
+      "learning_rate": 8.719809549213263e-06,
+      "loss": 0.4704,
+      "step": 3687
+    },
+    {
+      "epoch": 0.2560755450631857,
+      "grad_norm": 3.3944859547390185,
+      "learning_rate": 8.719058047957e-06,
+      "loss": 0.5059,
+      "step": 3688
+    },
+    {
+      "epoch": 0.2561449798639078,
+      "grad_norm": 4.471839233972908,
+      "learning_rate": 8.718306358594992e-06,
+      "loss": 0.7689,
+      "step": 3689
+    },
+    {
+      "epoch": 0.2562144146646299,
+      "grad_norm": 4.0410926864541254,
+      "learning_rate": 8.717554481165253e-06,
+      "loss": 0.1858,
+      "step": 3690
+    },
+    {
+      "epoch": 0.25628384946535204,
+      "grad_norm": 3.9511930998372895,
+      "learning_rate": 8.716802415705817e-06,
+      "loss": 0.5426,
+      "step": 3691
+    },
+    {
+      "epoch": 0.2563532842660742,
+      "grad_norm": 4.538200525076151,
+      "learning_rate": 8.716050162254719e-06,
+      "loss": 0.8154,
+      "step": 3692
+    },
+    {
+      "epoch": 0.25642271906679626,
+      "grad_norm": 4.022554026893457,
+      "learning_rate": 8.715297720850008e-06,
+      "loss": 0.602,
+      "step": 3693
+    },
+    {
+      "epoch": 0.2564921538675184,
+      "grad_norm": 3.4645986290554394,
+      "learning_rate": 8.714545091529744e-06,
+      "loss": 0.5585,
+      "step": 3694
+    },
+    {
+      "epoch": 0.2565615886682405,
+      "grad_norm": 4.195554527251548,
+      "learning_rate": 8.713792274331992e-06,
+      "loss": 0.4479,
+      "step": 3695
+    },
+    {
+      "epoch": 0.25663102346896266,
+      "grad_norm": 4.464151820857455,
+      "learning_rate": 8.71303926929483e-06,
+      "loss": 0.7188,
+      "step": 3696
+    },
+    {
+      "epoch": 0.25670045826968474,
+      "grad_norm": 3.3854129015094743,
+      "learning_rate": 8.712286076456343e-06,
+      "loss": 0.5404,
+      "step": 3697
+    },
+    {
+      "epoch": 0.2567698930704069,
+      "grad_norm": 3.698147523172282,
+      "learning_rate": 8.711532695854626e-06,
+      "loss": 0.4118,
+      "step": 3698
+    },
+    {
+      "epoch": 0.256839327871129,
+      "grad_norm": 6.089872241714095,
+      "learning_rate": 8.710779127527787e-06,
+      "loss": 0.761,
+      "step": 3699
+    },
+    {
+      "epoch": 0.25690876267185114,
+      "grad_norm": 3.0248302688147124,
+      "learning_rate": 8.710025371513938e-06,
+      "loss": 0.2765,
+      "step": 3700
+    },
+    {
+      "epoch": 0.2569781974725733,
+      "grad_norm": 4.7582300074726325,
+      "learning_rate": 8.709271427851202e-06,
+      "loss": 0.8491,
+      "step": 3701
+    },
+    {
+      "epoch": 0.25704763227329536,
+      "grad_norm": 3.927282465707881,
+      "learning_rate": 8.708517296577717e-06,
+      "loss": 0.6237,
+      "step": 3702
+    },
+    {
+      "epoch": 0.2571170670740175,
+      "grad_norm": 3.9276460844862453,
+      "learning_rate": 8.707762977731623e-06,
+      "loss": 0.5451,
+      "step": 3703
+    },
+    {
+      "epoch": 0.2571865018747396,
+      "grad_norm": 3.7129210863975564,
+      "learning_rate": 8.707008471351072e-06,
+      "loss": 0.4854,
+      "step": 3704
+    },
+    {
+      "epoch": 0.25725593667546176,
+      "grad_norm": 4.378542857487529,
+      "learning_rate": 8.706253777474229e-06,
+      "loss": 0.5442,
+      "step": 3705
+    },
+    {
+      "epoch": 0.25732537147618384,
+      "grad_norm": 2.7478404921210866,
+      "learning_rate": 8.705498896139262e-06,
+      "loss": 0.3281,
+      "step": 3706
+    },
+    {
+      "epoch": 0.257394806276906,
+      "grad_norm": 3.764600016738008,
+      "learning_rate": 8.704743827384354e-06,
+      "loss": 0.4428,
+      "step": 3707
+    },
+    {
+      "epoch": 0.2574642410776281,
+      "grad_norm": 3.5789090920691433,
+      "learning_rate": 8.703988571247697e-06,
+      "loss": 0.5114,
+      "step": 3708
+    },
+    {
+      "epoch": 0.25753367587835024,
+      "grad_norm": 3.4778798247844627,
+      "learning_rate": 8.70323312776749e-06,
+      "loss": 0.5107,
+      "step": 3709
+    },
+    {
+      "epoch": 0.2576031106790724,
+      "grad_norm": 4.705778284769779,
+      "learning_rate": 8.70247749698194e-06,
+      "loss": 0.8659,
+      "step": 3710
+    },
+    {
+      "epoch": 0.25767254547979446,
+      "grad_norm": 2.9687088745549928,
+      "learning_rate": 8.70172167892927e-06,
+      "loss": 0.3563,
+      "step": 3711
+    },
+    {
+      "epoch": 0.2577419802805166,
+      "grad_norm": 4.061336723632215,
+      "learning_rate": 8.700965673647703e-06,
+      "loss": 0.3727,
+      "step": 3712
+    },
+    {
+      "epoch": 0.2578114150812387,
+      "grad_norm": 5.665995477290148,
+      "learning_rate": 8.700209481175482e-06,
+      "loss": 0.7981,
+      "step": 3713
+    },
+    {
+      "epoch": 0.25788084988196086,
+      "grad_norm": 4.2363723457374585,
+      "learning_rate": 8.699453101550851e-06,
+      "loss": 0.5755,
+      "step": 3714
+    },
+    {
+      "epoch": 0.25795028468268294,
+      "grad_norm": 4.403971077616883,
+      "learning_rate": 8.69869653481207e-06,
+      "loss": 0.7409,
+      "step": 3715
+    },
+    {
+      "epoch": 0.2580197194834051,
+      "grad_norm": 3.5425542790285633,
+      "learning_rate": 8.697939780997403e-06,
+      "loss": 0.3264,
+      "step": 3716
+    },
+    {
+      "epoch": 0.2580891542841272,
+      "grad_norm": 3.5325835683909896,
+      "learning_rate": 8.697182840145129e-06,
+      "loss": 0.5285,
+      "step": 3717
+    },
+    {
+      "epoch": 0.25815858908484934,
+      "grad_norm": 3.669265364198408,
+      "learning_rate": 8.696425712293528e-06,
+      "loss": 0.5511,
+      "step": 3718
+    },
+    {
+      "epoch": 0.2582280238855714,
+      "grad_norm": 3.6563355371365134,
+      "learning_rate": 8.695668397480896e-06,
+      "loss": 0.4521,
+      "step": 3719
+    },
+    {
+      "epoch": 0.25829745868629356,
+      "grad_norm": 3.2591664195204846,
+      "learning_rate": 8.69491089574554e-06,
+      "loss": 0.4233,
+      "step": 3720
+    },
+    {
+      "epoch": 0.2583668934870157,
+      "grad_norm": 3.2601829432722345,
+      "learning_rate": 8.694153207125773e-06,
+      "loss": 0.3865,
+      "step": 3721
+    },
+    {
+      "epoch": 0.2584363282877378,
+      "grad_norm": 4.1589999012102945,
+      "learning_rate": 8.693395331659915e-06,
+      "loss": 0.6507,
+      "step": 3722
+    },
+    {
+      "epoch": 0.25850576308845996,
+      "grad_norm": 3.8521434799643637,
+      "learning_rate": 8.692637269386301e-06,
+      "loss": 0.4159,
+      "step": 3723
+    },
+    {
+      "epoch": 0.25857519788918204,
+      "grad_norm": 4.223491911413969,
+      "learning_rate": 8.691879020343275e-06,
+      "loss": 0.4368,
+      "step": 3724
+    },
+    {
+      "epoch": 0.2586446326899042,
+      "grad_norm": 3.8970764252375183,
+      "learning_rate": 8.691120584569182e-06,
+      "loss": 0.3934,
+      "step": 3725
+    },
+    {
+      "epoch": 0.2587140674906263,
+      "grad_norm": 4.544287573826822,
+      "learning_rate": 8.690361962102389e-06,
+      "loss": 0.5748,
+      "step": 3726
+    },
+    {
+      "epoch": 0.25878350229134844,
+      "grad_norm": 3.4477499550117856,
+      "learning_rate": 8.689603152981262e-06,
+      "loss": 0.5012,
+      "step": 3727
+    },
+    {
+      "epoch": 0.2588529370920705,
+      "grad_norm": 4.679099911045072,
+      "learning_rate": 8.688844157244185e-06,
+      "loss": 0.6416,
+      "step": 3728
+    },
+    {
+      "epoch": 0.25892237189279266,
+      "grad_norm": 2.8864586149355462,
+      "learning_rate": 8.688084974929542e-06,
+      "loss": 0.3298,
+      "step": 3729
+    },
+    {
+      "epoch": 0.2589918066935148,
+      "grad_norm": 3.823790535025746,
+      "learning_rate": 8.687325606075735e-06,
+      "loss": 0.4523,
+      "step": 3730
+    },
+    {
+      "epoch": 0.2590612414942369,
+      "grad_norm": 4.238353655221532,
+      "learning_rate": 8.686566050721174e-06,
+      "loss": 0.619,
+      "step": 3731
+    },
+    {
+      "epoch": 0.259130676294959,
+      "grad_norm": 4.356311785181821,
+      "learning_rate": 8.685806308904272e-06,
+      "loss": 0.6209,
+      "step": 3732
+    },
+    {
+      "epoch": 0.25920011109568114,
+      "grad_norm": 4.558367248298183,
+      "learning_rate": 8.685046380663457e-06,
+      "loss": 0.6248,
+      "step": 3733
+    },
+    {
+      "epoch": 0.2592695458964033,
+      "grad_norm": 3.7073873196327547,
+      "learning_rate": 8.684286266037168e-06,
+      "loss": 0.3233,
+      "step": 3734
+    },
+    {
+      "epoch": 0.2593389806971254,
+      "grad_norm": 3.3628312008263204,
+      "learning_rate": 8.683525965063845e-06,
+      "loss": 0.3987,
+      "step": 3735
+    },
+    {
+      "epoch": 0.25940841549784754,
+      "grad_norm": 3.679609783276532,
+      "learning_rate": 8.68276547778195e-06,
+      "loss": 0.298,
+      "step": 3736
+    },
+    {
+      "epoch": 0.2594778502985696,
+      "grad_norm": 3.9455659140466603,
+      "learning_rate": 8.682004804229945e-06,
+      "loss": 0.303,
+      "step": 3737
+    },
+    {
+      "epoch": 0.25954728509929176,
+      "grad_norm": 3.103492199022302,
+      "learning_rate": 8.6812439444463e-06,
+      "loss": 0.1908,
+      "step": 3738
+    },
+    {
+      "epoch": 0.2596167199000139,
+      "grad_norm": 3.341647308622433,
+      "learning_rate": 8.680482898469504e-06,
+      "loss": 0.2076,
+      "step": 3739
+    },
+    {
+      "epoch": 0.259686154700736,
+      "grad_norm": 3.3935584487925445,
+      "learning_rate": 8.679721666338048e-06,
+      "loss": 0.4279,
+      "step": 3740
+    },
+    {
+      "epoch": 0.2597555895014581,
+      "grad_norm": 4.834601659346408,
+      "learning_rate": 8.678960248090433e-06,
+      "loss": 0.6755,
+      "step": 3741
+    },
+    {
+      "epoch": 0.25982502430218024,
+      "grad_norm": 3.672041143407877,
+      "learning_rate": 8.678198643765174e-06,
+      "loss": 0.4076,
+      "step": 3742
+    },
+    {
+      "epoch": 0.2598944591029024,
+      "grad_norm": 4.688737112950332,
+      "learning_rate": 8.677436853400786e-06,
+      "loss": 0.5583,
+      "step": 3743
+    },
+    {
+      "epoch": 0.2599638939036245,
+      "grad_norm": 3.1196529498052517,
+      "learning_rate": 8.676674877035806e-06,
+      "loss": 0.3651,
+      "step": 3744
+    },
+    {
+      "epoch": 0.26003332870434664,
+      "grad_norm": 3.535350411870906,
+      "learning_rate": 8.675912714708771e-06,
+      "loss": 0.4869,
+      "step": 3745
+    },
+    {
+      "epoch": 0.2601027635050687,
+      "grad_norm": 4.41781643681792,
+      "learning_rate": 8.67515036645823e-06,
+      "loss": 0.6239,
+      "step": 3746
+    },
+    {
+      "epoch": 0.26017219830579086,
+      "grad_norm": 3.9610319128571203,
+      "learning_rate": 8.674387832322744e-06,
+      "loss": 0.5729,
+      "step": 3747
+    },
+    {
+      "epoch": 0.260241633106513,
+      "grad_norm": 3.820885237594903,
+      "learning_rate": 8.673625112340876e-06,
+      "loss": 0.4452,
+      "step": 3748
+    },
+    {
+      "epoch": 0.2603110679072351,
+      "grad_norm": 4.001776286416796,
+      "learning_rate": 8.672862206551209e-06,
+      "loss": 0.5465,
+      "step": 3749
+    },
+    {
+      "epoch": 0.2603805027079572,
+      "grad_norm": 3.352673362378982,
+      "learning_rate": 8.672099114992327e-06,
+      "loss": 0.5004,
+      "step": 3750
+    },
+    {
+      "epoch": 0.26044993750867934,
+      "grad_norm": 4.064581245372834,
+      "learning_rate": 8.671335837702827e-06,
+      "loss": 0.7259,
+      "step": 3751
+    },
+    {
+      "epoch": 0.2605193723094015,
+      "grad_norm": 4.37582457625417,
+      "learning_rate": 8.670572374721316e-06,
+      "loss": 0.6717,
+      "step": 3752
+    },
+    {
+      "epoch": 0.2605888071101236,
+      "grad_norm": 3.2761967136565553,
+      "learning_rate": 8.669808726086407e-06,
+      "loss": 0.2703,
+      "step": 3753
+    },
+    {
+      "epoch": 0.2606582419108457,
+      "grad_norm": 3.4886537153901784,
+      "learning_rate": 8.669044891836723e-06,
+      "loss": 0.2042,
+      "step": 3754
+    },
+    {
+      "epoch": 0.2607276767115678,
+      "grad_norm": 4.673566482489093,
+      "learning_rate": 8.668280872010902e-06,
+      "loss": 0.6267,
+      "step": 3755
+    },
+    {
+      "epoch": 0.26079711151228996,
+      "grad_norm": 3.6465630390547363,
+      "learning_rate": 8.667516666647586e-06,
+      "loss": 0.3612,
+      "step": 3756
+    },
+    {
+      "epoch": 0.2608665463130121,
+      "grad_norm": 3.2626920658503207,
+      "learning_rate": 8.666752275785427e-06,
+      "loss": 0.4037,
+      "step": 3757
+    },
+    {
+      "epoch": 0.26093598111373423,
+      "grad_norm": 4.606720147918789,
+      "learning_rate": 8.665987699463087e-06,
+      "loss": 0.4988,
+      "step": 3758
+    },
+    {
+      "epoch": 0.2610054159144563,
+      "grad_norm": 3.25884793127794,
+      "learning_rate": 8.665222937719237e-06,
+      "loss": 0.37,
+      "step": 3759
+    },
+    {
+      "epoch": 0.26107485071517844,
+      "grad_norm": 5.113012219061018,
+      "learning_rate": 8.664457990592559e-06,
+      "loss": 0.547,
+      "step": 3760
+    },
+    {
+      "epoch": 0.2611442855159006,
+      "grad_norm": 3.1486922647506543,
+      "learning_rate": 8.663692858121741e-06,
+      "loss": 0.298,
+      "step": 3761
+    },
+    {
+      "epoch": 0.2612137203166227,
+      "grad_norm": 3.836045705085945,
+      "learning_rate": 8.662927540345485e-06,
+      "loss": 0.3853,
+      "step": 3762
+    },
+    {
+      "epoch": 0.2612831551173448,
+      "grad_norm": 3.4533157945695168,
+      "learning_rate": 8.662162037302498e-06,
+      "loss": 0.2393,
+      "step": 3763
+    },
+    {
+      "epoch": 0.2613525899180669,
+      "grad_norm": 4.700753035554289,
+      "learning_rate": 8.661396349031502e-06,
+      "loss": 0.582,
+      "step": 3764
+    },
+    {
+      "epoch": 0.26142202471878906,
+      "grad_norm": 4.927499997392843,
+      "learning_rate": 8.660630475571217e-06,
+      "loss": 0.4884,
+      "step": 3765
+    },
+    {
+      "epoch": 0.2614914595195112,
+      "grad_norm": 2.6440646874593394,
+      "learning_rate": 8.65986441696039e-06,
+      "loss": 0.2278,
+      "step": 3766
+    },
+    {
+      "epoch": 0.26156089432023333,
+      "grad_norm": 2.927078536762051,
+      "learning_rate": 8.65909817323776e-06,
+      "loss": 0.324,
+      "step": 3767
+    },
+    {
+      "epoch": 0.2616303291209554,
+      "grad_norm": 3.402025023092969,
+      "learning_rate": 8.658331744442083e-06,
+      "loss": 0.422,
+      "step": 3768
+    },
+    {
+      "epoch": 0.26169976392167754,
+      "grad_norm": 3.9085888250836014,
+      "learning_rate": 8.657565130612129e-06,
+      "loss": 0.4297,
+      "step": 3769
+    },
+    {
+      "epoch": 0.2617691987223997,
+      "grad_norm": 4.704031019647921,
+      "learning_rate": 8.656798331786667e-06,
+      "loss": 0.7359,
+      "step": 3770
+    },
+    {
+      "epoch": 0.2618386335231218,
+      "grad_norm": 3.1802009589598046,
+      "learning_rate": 8.656031348004485e-06,
+      "loss": 0.2815,
+      "step": 3771
+    },
+    {
+      "epoch": 0.2619080683238439,
+      "grad_norm": 2.998361920106762,
+      "learning_rate": 8.655264179304374e-06,
+      "loss": 0.4179,
+      "step": 3772
+    },
+    {
+      "epoch": 0.261977503124566,
+      "grad_norm": 4.423547920011755,
+      "learning_rate": 8.654496825725138e-06,
+      "loss": 0.5624,
+      "step": 3773
+    },
+    {
+      "epoch": 0.26204693792528816,
+      "grad_norm": 3.57355921739013,
+      "learning_rate": 8.653729287305587e-06,
+      "loss": 0.3846,
+      "step": 3774
+    },
+    {
+      "epoch": 0.2621163727260103,
+      "grad_norm": 4.350798646433192,
+      "learning_rate": 8.652961564084544e-06,
+      "loss": 0.5279,
+      "step": 3775
+    },
+    {
+      "epoch": 0.2621858075267324,
+      "grad_norm": 3.9354246477590733,
+      "learning_rate": 8.652193656100837e-06,
+      "loss": 0.6322,
+      "step": 3776
+    },
+    {
+      "epoch": 0.2622552423274545,
+      "grad_norm": 4.3015505002426755,
+      "learning_rate": 8.651425563393307e-06,
+      "loss": 0.7676,
+      "step": 3777
+    },
+    {
+      "epoch": 0.26232467712817664,
+      "grad_norm": 3.947754858390564,
+      "learning_rate": 8.650657286000808e-06,
+      "loss": 0.6426,
+      "step": 3778
+    },
+    {
+      "epoch": 0.2623941119288988,
+      "grad_norm": 3.514071081103875,
+      "learning_rate": 8.649888823962191e-06,
+      "loss": 0.4179,
+      "step": 3779
+    },
+    {
+      "epoch": 0.2624635467296209,
+      "grad_norm": 3.801685335216067,
+      "learning_rate": 8.649120177316328e-06,
+      "loss": 0.3758,
+      "step": 3780
+    },
+    {
+      "epoch": 0.262532981530343,
+      "grad_norm": 6.882026619986459,
+      "learning_rate": 8.648351346102095e-06,
+      "loss": 0.7294,
+      "step": 3781
+    },
+    {
+      "epoch": 0.2626024163310651,
+      "grad_norm": 3.1114883394645783,
+      "learning_rate": 8.647582330358381e-06,
+      "loss": 0.2768,
+      "step": 3782
+    },
+    {
+      "epoch": 0.26267185113178726,
+      "grad_norm": 3.5751433738350284,
+      "learning_rate": 8.64681313012408e-06,
+      "loss": 0.3243,
+      "step": 3783
+    },
+    {
+      "epoch": 0.2627412859325094,
+      "grad_norm": 6.539711888238014,
+      "learning_rate": 8.646043745438097e-06,
+      "loss": 0.9965,
+      "step": 3784
+    },
+    {
+      "epoch": 0.2628107207332315,
+      "grad_norm": 2.8124836461727023,
+      "learning_rate": 8.645274176339346e-06,
+      "loss": 0.5004,
+      "step": 3785
+    },
+    {
+      "epoch": 0.2628801555339536,
+      "grad_norm": 2.788695676727389,
+      "learning_rate": 8.644504422866755e-06,
+      "loss": 0.311,
+      "step": 3786
+    },
+    {
+      "epoch": 0.26294959033467574,
+      "grad_norm": 4.296597862178043,
+      "learning_rate": 8.643734485059251e-06,
+      "loss": 0.767,
+      "step": 3787
+    },
+    {
+      "epoch": 0.2630190251353979,
+      "grad_norm": 4.746730540125942,
+      "learning_rate": 8.642964362955781e-06,
+      "loss": 0.5542,
+      "step": 3788
+    },
+    {
+      "epoch": 0.26308845993611996,
+      "grad_norm": 2.914023897910806,
+      "learning_rate": 8.642194056595296e-06,
+      "loss": 0.317,
+      "step": 3789
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 4.54338720894241,
+      "learning_rate": 8.64142356601676e-06,
+      "loss": 0.762,
+      "step": 3790
+    },
+    {
+      "epoch": 0.2632273295375642,
+      "grad_norm": 3.584984396457684,
+      "learning_rate": 8.640652891259137e-06,
+      "loss": 0.5233,
+      "step": 3791
+    },
+    {
+      "epoch": 0.26329676433828636,
+      "grad_norm": 3.408992675298201,
+      "learning_rate": 8.639882032361411e-06,
+      "loss": 0.4627,
+      "step": 3792
+    },
+    {
+      "epoch": 0.2633661991390085,
+      "grad_norm": 3.8739063791524706,
+      "learning_rate": 8.63911098936257e-06,
+      "loss": 0.4694,
+      "step": 3793
+    },
+    {
+      "epoch": 0.2634356339397306,
+      "grad_norm": 3.5742839483160447,
+      "learning_rate": 8.638339762301616e-06,
+      "loss": 0.3901,
+      "step": 3794
+    },
+    {
+      "epoch": 0.2635050687404527,
+      "grad_norm": 3.3205130963177987,
+      "learning_rate": 8.63756835121755e-06,
+      "loss": 0.4456,
+      "step": 3795
+    },
+    {
+      "epoch": 0.26357450354117484,
+      "grad_norm": 3.6653432946672933,
+      "learning_rate": 8.636796756149396e-06,
+      "loss": 0.4762,
+      "step": 3796
+    },
+    {
+      "epoch": 0.263643938341897,
+      "grad_norm": 3.32571864693562,
+      "learning_rate": 8.636024977136177e-06,
+      "loss": 0.3565,
+      "step": 3797
+    },
+    {
+      "epoch": 0.26371337314261906,
+      "grad_norm": 4.290371587467403,
+      "learning_rate": 8.635253014216928e-06,
+      "loss": 0.5047,
+      "step": 3798
+    },
+    {
+      "epoch": 0.2637828079433412,
+      "grad_norm": 3.374953103251966,
+      "learning_rate": 8.634480867430695e-06,
+      "loss": 0.4623,
+      "step": 3799
+    },
+    {
+      "epoch": 0.2638522427440633,
+      "grad_norm": 3.9390408204088514,
+      "learning_rate": 8.633708536816534e-06,
+      "loss": 0.5433,
+      "step": 3800
+    },
+    {
+      "epoch": 0.26392167754478546,
+      "grad_norm": 4.712009496590449,
+      "learning_rate": 8.632936022413505e-06,
+      "loss": 0.4005,
+      "step": 3801
+    },
+    {
+      "epoch": 0.2639911123455076,
+      "grad_norm": 3.688528639183929,
+      "learning_rate": 8.632163324260684e-06,
+      "loss": 0.6237,
+      "step": 3802
+    },
+    {
+      "epoch": 0.2640605471462297,
+      "grad_norm": 4.170705360892748,
+      "learning_rate": 8.631390442397152e-06,
+      "loss": 0.6521,
+      "step": 3803
+    },
+    {
+      "epoch": 0.2641299819469518,
+      "grad_norm": 4.08723631012054,
+      "learning_rate": 8.630617376862e-06,
+      "loss": 0.5216,
+      "step": 3804
+    },
+    {
+      "epoch": 0.26419941674767394,
+      "grad_norm": 3.4493718510062004,
+      "learning_rate": 8.629844127694332e-06,
+      "loss": 0.2866,
+      "step": 3805
+    },
+    {
+      "epoch": 0.2642688515483961,
+      "grad_norm": 3.302268233080713,
+      "learning_rate": 8.629070694933253e-06,
+      "loss": 0.4769,
+      "step": 3806
+    },
+    {
+      "epoch": 0.26433828634911816,
+      "grad_norm": 3.776633995435359,
+      "learning_rate": 8.628297078617885e-06,
+      "loss": 0.5345,
+      "step": 3807
+    },
+    {
+      "epoch": 0.2644077211498403,
+      "grad_norm": 3.8120901718808957,
+      "learning_rate": 8.627523278787357e-06,
+      "loss": 0.4971,
+      "step": 3808
+    },
+    {
+      "epoch": 0.2644771559505624,
+      "grad_norm": 3.4268250674306353,
+      "learning_rate": 8.626749295480805e-06,
+      "loss": 0.4652,
+      "step": 3809
+    },
+    {
+      "epoch": 0.26454659075128456,
+      "grad_norm": 3.970691406011485,
+      "learning_rate": 8.62597512873738e-06,
+      "loss": 0.475,
+      "step": 3810
+    },
+    {
+      "epoch": 0.26461602555200664,
+      "grad_norm": 4.718484252031654,
+      "learning_rate": 8.625200778596234e-06,
+      "loss": 0.6127,
+      "step": 3811
+    },
+    {
+      "epoch": 0.2646854603527288,
+      "grad_norm": 3.9584262285431713,
+      "learning_rate": 8.624426245096535e-06,
+      "loss": 0.5658,
+      "step": 3812
+    },
+    {
+      "epoch": 0.2647548951534509,
+      "grad_norm": 3.2644966313755113,
+      "learning_rate": 8.623651528277459e-06,
+      "loss": 0.4205,
+      "step": 3813
+    },
+    {
+      "epoch": 0.26482432995417304,
+      "grad_norm": 4.556259436948974,
+      "learning_rate": 8.62287662817819e-06,
+      "loss": 0.6465,
+      "step": 3814
+    },
+    {
+      "epoch": 0.2648937647548952,
+      "grad_norm": 2.8780325686080412,
+      "learning_rate": 8.622101544837917e-06,
+      "loss": 0.335,
+      "step": 3815
+    },
+    {
+      "epoch": 0.26496319955561726,
+      "grad_norm": 4.100471378746066,
+      "learning_rate": 8.621326278295849e-06,
+      "loss": 0.5263,
+      "step": 3816
+    },
+    {
+      "epoch": 0.2650326343563394,
+      "grad_norm": 2.224855121211458,
+      "learning_rate": 8.620550828591197e-06,
+      "loss": 0.2391,
+      "step": 3817
+    },
+    {
+      "epoch": 0.2651020691570615,
+      "grad_norm": 3.6960904514928727,
+      "learning_rate": 8.61977519576318e-06,
+      "loss": 0.3606,
+      "step": 3818
+    },
+    {
+      "epoch": 0.26517150395778366,
+      "grad_norm": 3.9057121774521533,
+      "learning_rate": 8.618999379851028e-06,
+      "loss": 0.575,
+      "step": 3819
+    },
+    {
+      "epoch": 0.26524093875850574,
+      "grad_norm": 3.8153213463117206,
+      "learning_rate": 8.618223380893984e-06,
+      "loss": 0.415,
+      "step": 3820
+    },
+    {
+      "epoch": 0.2653103735592279,
+      "grad_norm": 5.335332175421417,
+      "learning_rate": 8.617447198931295e-06,
+      "loss": 0.6197,
+      "step": 3821
+    },
+    {
+      "epoch": 0.26537980835995,
+      "grad_norm": 4.481521131599379,
+      "learning_rate": 8.61667083400222e-06,
+      "loss": 0.6084,
+      "step": 3822
+    },
+    {
+      "epoch": 0.26544924316067214,
+      "grad_norm": 5.224329864438417,
+      "learning_rate": 8.615894286146027e-06,
+      "loss": 0.6111,
+      "step": 3823
+    },
+    {
+      "epoch": 0.2655186779613943,
+      "grad_norm": 3.800069033108678,
+      "learning_rate": 8.615117555401994e-06,
+      "loss": 0.4478,
+      "step": 3824
+    },
+    {
+      "epoch": 0.26558811276211636,
+      "grad_norm": 5.239495041191937,
+      "learning_rate": 8.614340641809405e-06,
+      "loss": 0.6503,
+      "step": 3825
+    },
+    {
+      "epoch": 0.2656575475628385,
+      "grad_norm": 4.039016755031907,
+      "learning_rate": 8.613563545407554e-06,
+      "loss": 0.6781,
+      "step": 3826
+    },
+    {
+      "epoch": 0.2657269823635606,
+      "grad_norm": 4.231362451096744,
+      "learning_rate": 8.612786266235752e-06,
+      "loss": 0.4616,
+      "step": 3827
+    },
+    {
+      "epoch": 0.26579641716428276,
+      "grad_norm": 4.371438656833525,
+      "learning_rate": 8.612008804333307e-06,
+      "loss": 0.4893,
+      "step": 3828
+    },
+    {
+      "epoch": 0.26586585196500484,
+      "grad_norm": 3.213769006806774,
+      "learning_rate": 8.611231159739544e-06,
+      "loss": 0.2833,
+      "step": 3829
+    },
+    {
+      "epoch": 0.265935286765727,
+      "grad_norm": 4.112808008409419,
+      "learning_rate": 8.610453332493793e-06,
+      "loss": 0.5225,
+      "step": 3830
+    },
+    {
+      "epoch": 0.2660047215664491,
+      "grad_norm": 4.308154137605331,
+      "learning_rate": 8.6096753226354e-06,
+      "loss": 0.837,
+      "step": 3831
+    },
+    {
+      "epoch": 0.26607415636717124,
+      "grad_norm": 2.4229927303720684,
+      "learning_rate": 8.608897130203714e-06,
+      "loss": 0.2418,
+      "step": 3832
+    },
+    {
+      "epoch": 0.2661435911678933,
+      "grad_norm": 3.6925050085731472,
+      "learning_rate": 8.608118755238095e-06,
+      "loss": 0.5549,
+      "step": 3833
+    },
+    {
+      "epoch": 0.26621302596861546,
+      "grad_norm": 4.193244552810381,
+      "learning_rate": 8.607340197777911e-06,
+      "loss": 0.5605,
+      "step": 3834
+    },
+    {
+      "epoch": 0.2662824607693376,
+      "grad_norm": 2.994637132759814,
+      "learning_rate": 8.606561457862543e-06,
+      "loss": 0.1686,
+      "step": 3835
+    },
+    {
+      "epoch": 0.2663518955700597,
+      "grad_norm": 4.25942134858396,
+      "learning_rate": 8.605782535531377e-06,
+      "loss": 0.5894,
+      "step": 3836
+    },
+    {
+      "epoch": 0.26642133037078186,
+      "grad_norm": 3.292129482183229,
+      "learning_rate": 8.605003430823811e-06,
+      "loss": 0.3642,
+      "step": 3837
+    },
+    {
+      "epoch": 0.26649076517150394,
+      "grad_norm": 5.713508352119714,
+      "learning_rate": 8.60422414377925e-06,
+      "loss": 0.7843,
+      "step": 3838
+    },
+    {
+      "epoch": 0.2665601999722261,
+      "grad_norm": 3.861083019718728,
+      "learning_rate": 8.603444674437111e-06,
+      "loss": 0.7057,
+      "step": 3839
+    },
+    {
+      "epoch": 0.2666296347729482,
+      "grad_norm": 4.103552462258168,
+      "learning_rate": 8.602665022836819e-06,
+      "loss": 0.315,
+      "step": 3840
+    },
+    {
+      "epoch": 0.26669906957367034,
+      "grad_norm": 5.847778639386878,
+      "learning_rate": 8.601885189017805e-06,
+      "loss": 0.8071,
+      "step": 3841
+    },
+    {
+      "epoch": 0.2667685043743924,
+      "grad_norm": 3.577010243957329,
+      "learning_rate": 8.601105173019515e-06,
+      "loss": 0.3759,
+      "step": 3842
+    },
+    {
+      "epoch": 0.26683793917511456,
+      "grad_norm": 2.63615955770399,
+      "learning_rate": 8.6003249748814e-06,
+      "loss": 0.1789,
+      "step": 3843
+    },
+    {
+      "epoch": 0.2669073739758367,
+      "grad_norm": 3.9319838727918874,
+      "learning_rate": 8.599544594642921e-06,
+      "loss": 0.6455,
+      "step": 3844
+    },
+    {
+      "epoch": 0.2669768087765588,
+      "grad_norm": 4.437684699802833,
+      "learning_rate": 8.598764032343551e-06,
+      "loss": 0.6937,
+      "step": 3845
+    },
+    {
+      "epoch": 0.2670462435772809,
+      "grad_norm": 3.421097589180136,
+      "learning_rate": 8.597983288022766e-06,
+      "loss": 0.4942,
+      "step": 3846
+    },
+    {
+      "epoch": 0.26711567837800304,
+      "grad_norm": 3.046556726355791,
+      "learning_rate": 8.59720236172006e-06,
+      "loss": 0.3484,
+      "step": 3847
+    },
+    {
+      "epoch": 0.2671851131787252,
+      "grad_norm": 3.550665537742162,
+      "learning_rate": 8.59642125347493e-06,
+      "loss": 0.6028,
+      "step": 3848
+    },
+    {
+      "epoch": 0.2672545479794473,
+      "grad_norm": 4.26848441595344,
+      "learning_rate": 8.59563996332688e-06,
+      "loss": 0.6712,
+      "step": 3849
+    },
+    {
+      "epoch": 0.26732398278016944,
+      "grad_norm": 3.518828961044441,
+      "learning_rate": 8.59485849131543e-06,
+      "loss": 0.4858,
+      "step": 3850
+    },
+    {
+      "epoch": 0.2673934175808915,
+      "grad_norm": 5.818887364146886,
+      "learning_rate": 8.594076837480105e-06,
+      "loss": 0.3209,
+      "step": 3851
+    },
+    {
+      "epoch": 0.26746285238161366,
+      "grad_norm": 2.8048280036065534,
+      "learning_rate": 8.593295001860443e-06,
+      "loss": 0.4392,
+      "step": 3852
+    },
+    {
+      "epoch": 0.2675322871823358,
+      "grad_norm": 3.089201713948864,
+      "learning_rate": 8.592512984495984e-06,
+      "loss": 0.4257,
+      "step": 3853
+    },
+    {
+      "epoch": 0.2676017219830579,
+      "grad_norm": 3.663746925968469,
+      "learning_rate": 8.591730785426282e-06,
+      "loss": 0.462,
+      "step": 3854
+    },
+    {
+      "epoch": 0.26767115678378,
+      "grad_norm": 3.569923215691886,
+      "learning_rate": 8.590948404690904e-06,
+      "loss": 0.3531,
+      "step": 3855
+    },
+    {
+      "epoch": 0.26774059158450214,
+      "grad_norm": 3.9023213865239192,
+      "learning_rate": 8.59016584232942e-06,
+      "loss": 0.5818,
+      "step": 3856
+    },
+    {
+      "epoch": 0.2678100263852243,
+      "grad_norm": 3.2719269519988448,
+      "learning_rate": 8.589383098381408e-06,
+      "loss": 0.2685,
+      "step": 3857
+    },
+    {
+      "epoch": 0.2678794611859464,
+      "grad_norm": 4.158223664845444,
+      "learning_rate": 8.58860017288646e-06,
+      "loss": 0.7703,
+      "step": 3858
+    },
+    {
+      "epoch": 0.26794889598666854,
+      "grad_norm": 2.829621025541304,
+      "learning_rate": 8.587817065884177e-06,
+      "loss": 0.32,
+      "step": 3859
+    },
+    {
+      "epoch": 0.2680183307873906,
+      "grad_norm": 4.087408740164263,
+      "learning_rate": 8.587033777414167e-06,
+      "loss": 0.3155,
+      "step": 3860
+    },
+    {
+      "epoch": 0.26808776558811276,
+      "grad_norm": 5.557694376384905,
+      "learning_rate": 8.586250307516046e-06,
+      "loss": 0.9102,
+      "step": 3861
+    },
+    {
+      "epoch": 0.2681572003888349,
+      "grad_norm": 4.0097911468739165,
+      "learning_rate": 8.585466656229445e-06,
+      "loss": 0.576,
+      "step": 3862
+    },
+    {
+      "epoch": 0.26822663518955703,
+      "grad_norm": 3.8949080584365805,
+      "learning_rate": 8.584682823593996e-06,
+      "loss": 0.5071,
+      "step": 3863
+    },
+    {
+      "epoch": 0.2682960699902791,
+      "grad_norm": 3.81584766261813,
+      "learning_rate": 8.583898809649345e-06,
+      "loss": 0.471,
+      "step": 3864
+    },
+    {
+      "epoch": 0.26836550479100124,
+      "grad_norm": 4.392704051084909,
+      "learning_rate": 8.58311461443515e-06,
+      "loss": 0.5164,
+      "step": 3865
+    },
+    {
+      "epoch": 0.2684349395917234,
+      "grad_norm": 3.0818068184538605,
+      "learning_rate": 8.582330237991072e-06,
+      "loss": 0.3706,
+      "step": 3866
+    },
+    {
+      "epoch": 0.2685043743924455,
+      "grad_norm": 3.1203525685382076,
+      "learning_rate": 8.581545680356783e-06,
+      "loss": 0.215,
+      "step": 3867
+    },
+    {
+      "epoch": 0.2685738091931676,
+      "grad_norm": 5.03731914711839,
+      "learning_rate": 8.580760941571968e-06,
+      "loss": 0.786,
+      "step": 3868
+    },
+    {
+      "epoch": 0.2686432439938897,
+      "grad_norm": 3.6466514811477366,
+      "learning_rate": 8.579976021676315e-06,
+      "loss": 0.3877,
+      "step": 3869
+    },
+    {
+      "epoch": 0.26871267879461186,
+      "grad_norm": 3.8792744694470773,
+      "learning_rate": 8.579190920709527e-06,
+      "loss": 0.473,
+      "step": 3870
+    },
+    {
+      "epoch": 0.268782113595334,
+      "grad_norm": 3.509184200615901,
+      "learning_rate": 8.57840563871131e-06,
+      "loss": 0.3894,
+      "step": 3871
+    },
+    {
+      "epoch": 0.26885154839605613,
+      "grad_norm": 3.2675731402737838,
+      "learning_rate": 8.577620175721388e-06,
+      "loss": 0.4483,
+      "step": 3872
+    },
+    {
+      "epoch": 0.2689209831967782,
+      "grad_norm": 6.009858957663147,
+      "learning_rate": 8.576834531779484e-06,
+      "loss": 0.6879,
+      "step": 3873
+    },
+    {
+      "epoch": 0.26899041799750034,
+      "grad_norm": 3.582344873853896,
+      "learning_rate": 8.576048706925338e-06,
+      "loss": 0.2515,
+      "step": 3874
+    },
+    {
+      "epoch": 0.2690598527982225,
+      "grad_norm": 3.4651956502576295,
+      "learning_rate": 8.575262701198694e-06,
+      "loss": 0.3092,
+      "step": 3875
+    },
+    {
+      "epoch": 0.2691292875989446,
+      "grad_norm": 5.552752785621959,
+      "learning_rate": 8.574476514639309e-06,
+      "loss": 0.9008,
+      "step": 3876
+    },
+    {
+      "epoch": 0.2691987223996667,
+      "grad_norm": 3.796075710995265,
+      "learning_rate": 8.573690147286947e-06,
+      "loss": 0.4454,
+      "step": 3877
+    },
+    {
+      "epoch": 0.2692681572003888,
+      "grad_norm": 4.513370285562939,
+      "learning_rate": 8.57290359918138e-06,
+      "loss": 0.5987,
+      "step": 3878
+    },
+    {
+      "epoch": 0.26933759200111096,
+      "grad_norm": 2.825597766741668,
+      "learning_rate": 8.572116870362392e-06,
+      "loss": 0.2787,
+      "step": 3879
+    },
+    {
+      "epoch": 0.2694070268018331,
+      "grad_norm": 5.247271576577546,
+      "learning_rate": 8.571329960869776e-06,
+      "loss": 0.4598,
+      "step": 3880
+    },
+    {
+      "epoch": 0.26947646160255523,
+      "grad_norm": 4.003190775525243,
+      "learning_rate": 8.570542870743333e-06,
+      "loss": 0.3947,
+      "step": 3881
+    },
+    {
+      "epoch": 0.2695458964032773,
+      "grad_norm": 4.009250111141734,
+      "learning_rate": 8.569755600022869e-06,
+      "loss": 0.3646,
+      "step": 3882
+    },
+    {
+      "epoch": 0.26961533120399944,
+      "grad_norm": 3.2573358114813353,
+      "learning_rate": 8.568968148748208e-06,
+      "loss": 0.3189,
+      "step": 3883
+    },
+    {
+      "epoch": 0.2696847660047216,
+      "grad_norm": 3.6530009168342485,
+      "learning_rate": 8.568180516959175e-06,
+      "loss": 0.5447,
+      "step": 3884
+    },
+    {
+      "epoch": 0.2697542008054437,
+      "grad_norm": 3.0762048773545834,
+      "learning_rate": 8.567392704695612e-06,
+      "loss": 0.245,
+      "step": 3885
+    },
+    {
+      "epoch": 0.2698236356061658,
+      "grad_norm": 3.504238387281911,
+      "learning_rate": 8.56660471199736e-06,
+      "loss": 0.4517,
+      "step": 3886
+    },
+    {
+      "epoch": 0.2698930704068879,
+      "grad_norm": 4.826330345087875,
+      "learning_rate": 8.56581653890428e-06,
+      "loss": 0.6409,
+      "step": 3887
+    },
+    {
+      "epoch": 0.26996250520761006,
+      "grad_norm": 4.098379604927787,
+      "learning_rate": 8.56502818545623e-06,
+      "loss": 0.5758,
+      "step": 3888
+    },
+    {
+      "epoch": 0.2700319400083322,
+      "grad_norm": 4.813088830645753,
+      "learning_rate": 8.564239651693092e-06,
+      "loss": 0.3833,
+      "step": 3889
+    },
+    {
+      "epoch": 0.2701013748090543,
+      "grad_norm": 3.952636111842192,
+      "learning_rate": 8.563450937654746e-06,
+      "loss": 0.3664,
+      "step": 3890
+    },
+    {
+      "epoch": 0.2701708096097764,
+      "grad_norm": 4.279041753253571,
+      "learning_rate": 8.562662043381083e-06,
+      "loss": 0.4426,
+      "step": 3891
+    },
+    {
+      "epoch": 0.27024024441049854,
+      "grad_norm": 4.359928729625919,
+      "learning_rate": 8.561872968912005e-06,
+      "loss": 0.5343,
+      "step": 3892
+    },
+    {
+      "epoch": 0.2703096792112207,
+      "grad_norm": 4.2040795690099335,
+      "learning_rate": 8.561083714287423e-06,
+      "loss": 0.5281,
+      "step": 3893
+    },
+    {
+      "epoch": 0.2703791140119428,
+      "grad_norm": 3.340529601914484,
+      "learning_rate": 8.560294279547256e-06,
+      "loss": 0.2504,
+      "step": 3894
+    },
+    {
+      "epoch": 0.2704485488126649,
+      "grad_norm": 4.169057448270038,
+      "learning_rate": 8.559504664731432e-06,
+      "loss": 0.4156,
+      "step": 3895
+    },
+    {
+      "epoch": 0.270517983613387,
+      "grad_norm": 3.6196791851967562,
+      "learning_rate": 8.558714869879892e-06,
+      "loss": 0.3248,
+      "step": 3896
+    },
+    {
+      "epoch": 0.27058741841410916,
+      "grad_norm": 2.8569923021341577,
+      "learning_rate": 8.55792489503258e-06,
+      "loss": 0.3679,
+      "step": 3897
+    },
+    {
+      "epoch": 0.2706568532148313,
+      "grad_norm": 3.1882548681120553,
+      "learning_rate": 8.557134740229453e-06,
+      "loss": 0.3203,
+      "step": 3898
+    },
+    {
+      "epoch": 0.2707262880155534,
+      "grad_norm": 4.082105513197329,
+      "learning_rate": 8.556344405510478e-06,
+      "loss": 0.5906,
+      "step": 3899
+    },
+    {
+      "epoch": 0.2707957228162755,
+      "grad_norm": 3.5623634899337175,
+      "learning_rate": 8.555553890915625e-06,
+      "loss": 0.516,
+      "step": 3900
+    },
+    {
+      "epoch": 0.27086515761699764,
+      "grad_norm": 4.02893261001425,
+      "learning_rate": 8.55476319648488e-06,
+      "loss": 0.5333,
+      "step": 3901
+    },
+    {
+      "epoch": 0.2709345924177198,
+      "grad_norm": 5.080311954976574,
+      "learning_rate": 8.553972322258235e-06,
+      "loss": 0.6897,
+      "step": 3902
+    },
+    {
+      "epoch": 0.27100402721844186,
+      "grad_norm": 4.51289904146512,
+      "learning_rate": 8.553181268275691e-06,
+      "loss": 0.7505,
+      "step": 3903
+    },
+    {
+      "epoch": 0.271073462019164,
+      "grad_norm": 3.566562049272631,
+      "learning_rate": 8.55239003457726e-06,
+      "loss": 0.4698,
+      "step": 3904
+    },
+    {
+      "epoch": 0.2711428968198861,
+      "grad_norm": 2.4311598191921244,
+      "learning_rate": 8.55159862120296e-06,
+      "loss": 0.2783,
+      "step": 3905
+    },
+    {
+      "epoch": 0.27121233162060826,
+      "grad_norm": 4.35402246560189,
+      "learning_rate": 8.55080702819282e-06,
+      "loss": 0.5469,
+      "step": 3906
+    },
+    {
+      "epoch": 0.2712817664213304,
+      "grad_norm": 2.131067693260364,
+      "learning_rate": 8.550015255586881e-06,
+      "loss": 0.2056,
+      "step": 3907
+    },
+    {
+      "epoch": 0.2713512012220525,
+      "grad_norm": 3.3862729869829495,
+      "learning_rate": 8.549223303425187e-06,
+      "loss": 0.4187,
+      "step": 3908
+    },
+    {
+      "epoch": 0.2714206360227746,
+      "grad_norm": 4.079833994777261,
+      "learning_rate": 8.548431171747793e-06,
+      "loss": 0.455,
+      "step": 3909
+    },
+    {
+      "epoch": 0.27149007082349674,
+      "grad_norm": 3.5826218475095297,
+      "learning_rate": 8.547638860594765e-06,
+      "loss": 0.3455,
+      "step": 3910
+    },
+    {
+      "epoch": 0.2715595056242189,
+      "grad_norm": 3.0111521465501183,
+      "learning_rate": 8.54684637000618e-06,
+      "loss": 0.2479,
+      "step": 3911
+    },
+    {
+      "epoch": 0.27162894042494096,
+      "grad_norm": 3.007210227602982,
+      "learning_rate": 8.546053700022118e-06,
+      "loss": 0.457,
+      "step": 3912
+    },
+    {
+      "epoch": 0.2716983752256631,
+      "grad_norm": 3.1536558948414273,
+      "learning_rate": 8.545260850682673e-06,
+      "loss": 0.3014,
+      "step": 3913
+    },
+    {
+      "epoch": 0.2717678100263852,
+      "grad_norm": 3.9974074625601457,
+      "learning_rate": 8.544467822027946e-06,
+      "loss": 0.5332,
+      "step": 3914
+    },
+    {
+      "epoch": 0.27183724482710736,
+      "grad_norm": 4.546530981497959,
+      "learning_rate": 8.543674614098046e-06,
+      "loss": 0.6505,
+      "step": 3915
+    },
+    {
+      "epoch": 0.2719066796278295,
+      "grad_norm": 3.7122579297131852,
+      "learning_rate": 8.542881226933095e-06,
+      "loss": 0.4536,
+      "step": 3916
+    },
+    {
+      "epoch": 0.2719761144285516,
+      "grad_norm": 4.141981177387173,
+      "learning_rate": 8.54208766057322e-06,
+      "loss": 0.5269,
+      "step": 3917
+    },
+    {
+      "epoch": 0.2720455492292737,
+      "grad_norm": 3.5738918931496393,
+      "learning_rate": 8.541293915058561e-06,
+      "loss": 0.5078,
+      "step": 3918
+    },
+    {
+      "epoch": 0.27211498402999584,
+      "grad_norm": 3.873632814578405,
+      "learning_rate": 8.540499990429262e-06,
+      "loss": 0.3717,
+      "step": 3919
+    },
+    {
+      "epoch": 0.272184418830718,
+      "grad_norm": 4.4500103806219835,
+      "learning_rate": 8.53970588672548e-06,
+      "loss": 0.7431,
+      "step": 3920
+    },
+    {
+      "epoch": 0.27225385363144006,
+      "grad_norm": 3.2402467822452876,
+      "learning_rate": 8.538911603987379e-06,
+      "loss": 0.1798,
+      "step": 3921
+    },
+    {
+      "epoch": 0.2723232884321622,
+      "grad_norm": 4.4644854094805675,
+      "learning_rate": 8.538117142255133e-06,
+      "loss": 0.5751,
+      "step": 3922
+    },
+    {
+      "epoch": 0.2723927232328843,
+      "grad_norm": 3.8541168607766143,
+      "learning_rate": 8.537322501568927e-06,
+      "loss": 0.5901,
+      "step": 3923
+    },
+    {
+      "epoch": 0.27246215803360646,
+      "grad_norm": 3.2966949927377187,
+      "learning_rate": 8.536527681968951e-06,
+      "loss": 0.3975,
+      "step": 3924
+    },
+    {
+      "epoch": 0.27253159283432854,
+      "grad_norm": 4.459296143148606,
+      "learning_rate": 8.535732683495408e-06,
+      "loss": 0.6631,
+      "step": 3925
+    },
+    {
+      "epoch": 0.2726010276350507,
+      "grad_norm": 2.7370043017821066,
+      "learning_rate": 8.534937506188505e-06,
+      "loss": 0.2215,
+      "step": 3926
+    },
+    {
+      "epoch": 0.2726704624357728,
+      "grad_norm": 3.6387638061871885,
+      "learning_rate": 8.534142150088465e-06,
+      "loss": 0.4871,
+      "step": 3927
+    },
+    {
+      "epoch": 0.27273989723649494,
+      "grad_norm": 3.3401665506071017,
+      "learning_rate": 8.533346615235511e-06,
+      "loss": 0.3502,
+      "step": 3928
+    },
+    {
+      "epoch": 0.2728093320372171,
+      "grad_norm": 3.902736803443321,
+      "learning_rate": 8.532550901669888e-06,
+      "loss": 0.4743,
+      "step": 3929
+    },
+    {
+      "epoch": 0.27287876683793916,
+      "grad_norm": 5.121392973387632,
+      "learning_rate": 8.531755009431834e-06,
+      "loss": 0.6266,
+      "step": 3930
+    },
+    {
+      "epoch": 0.2729482016386613,
+      "grad_norm": 4.6875362862297765,
+      "learning_rate": 8.530958938561608e-06,
+      "loss": 0.602,
+      "step": 3931
+    },
+    {
+      "epoch": 0.2730176364393834,
+      "grad_norm": 3.5698927408132506,
+      "learning_rate": 8.530162689099478e-06,
+      "loss": 0.3078,
+      "step": 3932
+    },
+    {
+      "epoch": 0.27308707124010556,
+      "grad_norm": 3.842575041333963,
+      "learning_rate": 8.529366261085709e-06,
+      "loss": 0.3684,
+      "step": 3933
+    },
+    {
+      "epoch": 0.27315650604082764,
+      "grad_norm": 3.3087525843160663,
+      "learning_rate": 8.528569654560592e-06,
+      "loss": 0.4151,
+      "step": 3934
+    },
+    {
+      "epoch": 0.2732259408415498,
+      "grad_norm": 3.492061981393543,
+      "learning_rate": 8.527772869564412e-06,
+      "loss": 0.4676,
+      "step": 3935
+    },
+    {
+      "epoch": 0.2732953756422719,
+      "grad_norm": 3.7733239444621827,
+      "learning_rate": 8.526975906137476e-06,
+      "loss": 0.4144,
+      "step": 3936
+    },
+    {
+      "epoch": 0.27336481044299404,
+      "grad_norm": 4.19321264976016,
+      "learning_rate": 8.526178764320086e-06,
+      "loss": 0.6234,
+      "step": 3937
+    },
+    {
+      "epoch": 0.2734342452437161,
+      "grad_norm": 2.610959778680088,
+      "learning_rate": 8.525381444152566e-06,
+      "loss": 0.2507,
+      "step": 3938
+    },
+    {
+      "epoch": 0.27350368004443826,
+      "grad_norm": 4.553971601585641,
+      "learning_rate": 8.524583945675243e-06,
+      "loss": 0.2621,
+      "step": 3939
+    },
+    {
+      "epoch": 0.2735731148451604,
+      "grad_norm": 4.1710120958577255,
+      "learning_rate": 8.523786268928449e-06,
+      "loss": 0.6324,
+      "step": 3940
+    },
+    {
+      "epoch": 0.2736425496458825,
+      "grad_norm": 4.37450159491833,
+      "learning_rate": 8.522988413952535e-06,
+      "loss": 0.3864,
+      "step": 3941
+    },
+    {
+      "epoch": 0.27371198444660466,
+      "grad_norm": 3.8745029075193185,
+      "learning_rate": 8.522190380787853e-06,
+      "loss": 0.5334,
+      "step": 3942
+    },
+    {
+      "epoch": 0.27378141924732674,
+      "grad_norm": 4.221566270834122,
+      "learning_rate": 8.521392169474766e-06,
+      "loss": 0.4507,
+      "step": 3943
+    },
+    {
+      "epoch": 0.2738508540480489,
+      "grad_norm": 4.096753753463659,
+      "learning_rate": 8.520593780053653e-06,
+      "loss": 0.595,
+      "step": 3944
+    },
+    {
+      "epoch": 0.273920288848771,
+      "grad_norm": 3.8841435059680607,
+      "learning_rate": 8.519795212564886e-06,
+      "loss": 0.5323,
+      "step": 3945
+    },
+    {
+      "epoch": 0.27398972364949314,
+      "grad_norm": 4.497939338855549,
+      "learning_rate": 8.51899646704886e-06,
+      "loss": 0.6577,
+      "step": 3946
+    },
+    {
+      "epoch": 0.2740591584502152,
+      "grad_norm": 3.2301426627009513,
+      "learning_rate": 8.518197543545975e-06,
+      "loss": 0.2191,
+      "step": 3947
+    },
+    {
+      "epoch": 0.27412859325093736,
+      "grad_norm": 3.43826617069453,
+      "learning_rate": 8.51739844209664e-06,
+      "loss": 0.3505,
+      "step": 3948
+    },
+    {
+      "epoch": 0.2741980280516595,
+      "grad_norm": 2.8496398595137156,
+      "learning_rate": 8.516599162741272e-06,
+      "loss": 0.3319,
+      "step": 3949
+    },
+    {
+      "epoch": 0.2742674628523816,
+      "grad_norm": 3.9102946648353747,
+      "learning_rate": 8.515799705520298e-06,
+      "loss": 0.3348,
+      "step": 3950
+    },
+    {
+      "epoch": 0.27433689765310376,
+      "grad_norm": 4.780337176054412,
+      "learning_rate": 8.515000070474152e-06,
+      "loss": 0.7771,
+      "step": 3951
+    },
+    {
+      "epoch": 0.27440633245382584,
+      "grad_norm": 4.0990320522954455,
+      "learning_rate": 8.514200257643281e-06,
+      "loss": 0.5314,
+      "step": 3952
+    },
+    {
+      "epoch": 0.274475767254548,
+      "grad_norm": 4.193847029044622,
+      "learning_rate": 8.513400267068137e-06,
+      "loss": 0.3235,
+      "step": 3953
+    },
+    {
+      "epoch": 0.2745452020552701,
+      "grad_norm": 2.9117500427935106,
+      "learning_rate": 8.512600098789184e-06,
+      "loss": 0.3009,
+      "step": 3954
+    },
+    {
+      "epoch": 0.27461463685599224,
+      "grad_norm": 4.360164559496565,
+      "learning_rate": 8.511799752846893e-06,
+      "loss": 0.5964,
+      "step": 3955
+    },
+    {
+      "epoch": 0.2746840716567143,
+      "grad_norm": 4.941288939914137,
+      "learning_rate": 8.510999229281742e-06,
+      "loss": 0.7531,
+      "step": 3956
+    },
+    {
+      "epoch": 0.27475350645743646,
+      "grad_norm": 4.5160464079108635,
+      "learning_rate": 8.510198528134224e-06,
+      "loss": 0.5213,
+      "step": 3957
+    },
+    {
+      "epoch": 0.2748229412581586,
+      "grad_norm": 4.866185133989373,
+      "learning_rate": 8.509397649444837e-06,
+      "loss": 0.5629,
+      "step": 3958
+    },
+    {
+      "epoch": 0.2748923760588807,
+      "grad_norm": 3.659014231939476,
+      "learning_rate": 8.508596593254088e-06,
+      "loss": 0.3265,
+      "step": 3959
+    },
+    {
+      "epoch": 0.2749618108596028,
+      "grad_norm": 5.648875352908416,
+      "learning_rate": 8.507795359602493e-06,
+      "loss": 0.5491,
+      "step": 3960
+    },
+    {
+      "epoch": 0.27503124566032494,
+      "grad_norm": 4.328783987942329,
+      "learning_rate": 8.50699394853058e-06,
+      "loss": 0.4401,
+      "step": 3961
+    },
+    {
+      "epoch": 0.2751006804610471,
+      "grad_norm": 5.173724432264053,
+      "learning_rate": 8.50619236007888e-06,
+      "loss": 0.6956,
+      "step": 3962
+    },
+    {
+      "epoch": 0.2751701152617692,
+      "grad_norm": 5.5432746400195585,
+      "learning_rate": 8.505390594287939e-06,
+      "loss": 0.4899,
+      "step": 3963
+    },
+    {
+      "epoch": 0.27523955006249134,
+      "grad_norm": 36.555204148244876,
+      "learning_rate": 8.504588651198307e-06,
+      "loss": 0.6905,
+      "step": 3964
+    },
+    {
+      "epoch": 0.2753089848632134,
+      "grad_norm": 4.510116141813476,
+      "learning_rate": 8.503786530850545e-06,
+      "loss": 0.3931,
+      "step": 3965
+    },
+    {
+      "epoch": 0.27537841966393556,
+      "grad_norm": 5.804165046003108,
+      "learning_rate": 8.50298423328523e-06,
+      "loss": 0.6384,
+      "step": 3966
+    },
+    {
+      "epoch": 0.2754478544646577,
+      "grad_norm": 3.5509848381216984,
+      "learning_rate": 8.502181758542934e-06,
+      "loss": 0.5456,
+      "step": 3967
+    },
+    {
+      "epoch": 0.27551728926537983,
+      "grad_norm": 3.2958677391363818,
+      "learning_rate": 8.501379106664247e-06,
+      "loss": 0.2616,
+      "step": 3968
+    },
+    {
+      "epoch": 0.2755867240661019,
+      "grad_norm": 4.333922625407603,
+      "learning_rate": 8.500576277689768e-06,
+      "loss": 0.5866,
+      "step": 3969
+    },
+    {
+      "epoch": 0.27565615886682404,
+      "grad_norm": 3.4166556645241015,
+      "learning_rate": 8.4997732716601e-06,
+      "loss": 0.1816,
+      "step": 3970
+    },
+    {
+      "epoch": 0.2757255936675462,
+      "grad_norm": 3.1016646177239564,
+      "learning_rate": 8.498970088615861e-06,
+      "loss": 0.2755,
+      "step": 3971
+    },
+    {
+      "epoch": 0.2757950284682683,
+      "grad_norm": 3.128761465521594,
+      "learning_rate": 8.498166728597675e-06,
+      "loss": 0.2171,
+      "step": 3972
+    },
+    {
+      "epoch": 0.27586446326899045,
+      "grad_norm": 3.4957978795719913,
+      "learning_rate": 8.497363191646174e-06,
+      "loss": 0.4962,
+      "step": 3973
+    },
+    {
+      "epoch": 0.2759338980697125,
+      "grad_norm": 3.7163187490375336,
+      "learning_rate": 8.496559477801999e-06,
+      "loss": 0.3683,
+      "step": 3974
+    },
+    {
+      "epoch": 0.27600333287043466,
+      "grad_norm": 4.556605352713522,
+      "learning_rate": 8.495755587105805e-06,
+      "loss": 0.7999,
+      "step": 3975
+    },
+    {
+      "epoch": 0.2760727676711568,
+      "grad_norm": 3.5344759807280424,
+      "learning_rate": 8.494951519598247e-06,
+      "loss": 0.4172,
+      "step": 3976
+    },
+    {
+      "epoch": 0.27614220247187893,
+      "grad_norm": 5.475560162721737,
+      "learning_rate": 8.494147275319997e-06,
+      "loss": 0.6752,
+      "step": 3977
+    },
+    {
+      "epoch": 0.276211637272601,
+      "grad_norm": 4.064245414278551,
+      "learning_rate": 8.493342854311732e-06,
+      "loss": 0.5148,
+      "step": 3978
+    },
+    {
+      "epoch": 0.27628107207332314,
+      "grad_norm": 2.8269240789283794,
+      "learning_rate": 8.492538256614138e-06,
+      "loss": 0.2589,
+      "step": 3979
+    },
+    {
+      "epoch": 0.2763505068740453,
+      "grad_norm": 4.478320409797524,
+      "learning_rate": 8.49173348226791e-06,
+      "loss": 0.5475,
+      "step": 3980
+    },
+    {
+      "epoch": 0.2764199416747674,
+      "grad_norm": 4.383672258718609,
+      "learning_rate": 8.490928531313754e-06,
+      "loss": 0.5026,
+      "step": 3981
+    },
+    {
+      "epoch": 0.2764893764754895,
+      "grad_norm": 3.1383803432342288,
+      "learning_rate": 8.490123403792385e-06,
+      "loss": 0.3206,
+      "step": 3982
+    },
+    {
+      "epoch": 0.2765588112762116,
+      "grad_norm": 2.686453896282508,
+      "learning_rate": 8.48931809974452e-06,
+      "loss": 0.211,
+      "step": 3983
+    },
+    {
+      "epoch": 0.27662824607693376,
+      "grad_norm": 4.1028639941544505,
+      "learning_rate": 8.488512619210898e-06,
+      "loss": 0.5724,
+      "step": 3984
+    },
+    {
+      "epoch": 0.2766976808776559,
+      "grad_norm": 3.5597639153926828,
+      "learning_rate": 8.487706962232252e-06,
+      "loss": 0.409,
+      "step": 3985
+    },
+    {
+      "epoch": 0.27676711567837803,
+      "grad_norm": 3.8829886712348727,
+      "learning_rate": 8.486901128849337e-06,
+      "loss": 0.627,
+      "step": 3986
+    },
+    {
+      "epoch": 0.2768365504791001,
+      "grad_norm": 3.638811596020324,
+      "learning_rate": 8.486095119102907e-06,
+      "loss": 0.3147,
+      "step": 3987
+    },
+    {
+      "epoch": 0.27690598527982224,
+      "grad_norm": 4.594077141814216,
+      "learning_rate": 8.48528893303373e-06,
+      "loss": 0.4465,
+      "step": 3988
+    },
+    {
+      "epoch": 0.2769754200805444,
+      "grad_norm": 4.525517072607048,
+      "learning_rate": 8.484482570682584e-06,
+      "loss": 0.7978,
+      "step": 3989
+    },
+    {
+      "epoch": 0.2770448548812665,
+      "grad_norm": 4.158662107570065,
+      "learning_rate": 8.483676032090252e-06,
+      "loss": 0.5915,
+      "step": 3990
+    },
+    {
+      "epoch": 0.2771142896819886,
+      "grad_norm": 4.117731456261733,
+      "learning_rate": 8.482869317297529e-06,
+      "loss": 0.5039,
+      "step": 3991
+    },
+    {
+      "epoch": 0.2771837244827107,
+      "grad_norm": 3.518711943853984,
+      "learning_rate": 8.482062426345217e-06,
+      "loss": 0.2799,
+      "step": 3992
+    },
+    {
+      "epoch": 0.27725315928343286,
+      "grad_norm": 3.475791772361931,
+      "learning_rate": 8.481255359274127e-06,
+      "loss": 0.3491,
+      "step": 3993
+    },
+    {
+      "epoch": 0.277322594084155,
+      "grad_norm": 4.2651457753923205,
+      "learning_rate": 8.48044811612508e-06,
+      "loss": 0.4764,
+      "step": 3994
+    },
+    {
+      "epoch": 0.2773920288848771,
+      "grad_norm": 3.8506198038834603,
+      "learning_rate": 8.479640696938906e-06,
+      "loss": 0.3734,
+      "step": 3995
+    },
+    {
+      "epoch": 0.2774614636855992,
+      "grad_norm": 5.688336871992002,
+      "learning_rate": 8.478833101756444e-06,
+      "loss": 0.5026,
+      "step": 3996
+    },
+    {
+      "epoch": 0.27753089848632134,
+      "grad_norm": 3.8932229558853,
+      "learning_rate": 8.478025330618541e-06,
+      "loss": 0.5945,
+      "step": 3997
+    },
+    {
+      "epoch": 0.2776003332870435,
+      "grad_norm": 3.1092978864306047,
+      "learning_rate": 8.47721738356605e-06,
+      "loss": 0.3724,
+      "step": 3998
+    },
+    {
+      "epoch": 0.2776697680877656,
+      "grad_norm": 3.17830692841414,
+      "learning_rate": 8.476409260639838e-06,
+      "loss": 0.4245,
+      "step": 3999
+    },
+    {
+      "epoch": 0.2777392028884877,
+      "grad_norm": 3.689409196680556,
+      "learning_rate": 8.475600961880781e-06,
+      "loss": 0.5328,
+      "step": 4000
+    },
+    {
+      "epoch": 0.2778086376892098,
+      "grad_norm": 3.217836477647881,
+      "learning_rate": 8.474792487329761e-06,
+      "loss": 0.3163,
+      "step": 4001
+    },
+    {
+      "epoch": 0.27787807248993196,
+      "grad_norm": 3.4252580676287905,
+      "learning_rate": 8.473983837027668e-06,
+      "loss": 0.3379,
+      "step": 4002
+    },
+    {
+      "epoch": 0.2779475072906541,
+      "grad_norm": 4.147392996962823,
+      "learning_rate": 8.473175011015405e-06,
+      "loss": 0.7043,
+      "step": 4003
+    },
+    {
+      "epoch": 0.2780169420913762,
+      "grad_norm": 5.565553678850914,
+      "learning_rate": 8.472366009333877e-06,
+      "loss": 0.8151,
+      "step": 4004
+    },
+    {
+      "epoch": 0.2780863768920983,
+      "grad_norm": 3.2255866601641023,
+      "learning_rate": 8.471556832024008e-06,
+      "loss": 0.3363,
+      "step": 4005
+    },
+    {
+      "epoch": 0.27815581169282044,
+      "grad_norm": 3.929571808171114,
+      "learning_rate": 8.470747479126721e-06,
+      "loss": 0.6045,
+      "step": 4006
+    },
+    {
+      "epoch": 0.2782252464935426,
+      "grad_norm": 4.042282621854157,
+      "learning_rate": 8.469937950682956e-06,
+      "loss": 0.5686,
+      "step": 4007
+    },
+    {
+      "epoch": 0.2782946812942647,
+      "grad_norm": 4.016366702539411,
+      "learning_rate": 8.469128246733655e-06,
+      "loss": 0.238,
+      "step": 4008
+    },
+    {
+      "epoch": 0.2783641160949868,
+      "grad_norm": 4.417358218370244,
+      "learning_rate": 8.468318367319772e-06,
+      "loss": 0.4056,
+      "step": 4009
+    },
+    {
+      "epoch": 0.2784335508957089,
+      "grad_norm": 3.9453179882589793,
+      "learning_rate": 8.46750831248227e-06,
+      "loss": 0.583,
+      "step": 4010
+    },
+    {
+      "epoch": 0.27850298569643106,
+      "grad_norm": 5.159194775574404,
+      "learning_rate": 8.466698082262125e-06,
+      "loss": 0.7985,
+      "step": 4011
+    },
+    {
+      "epoch": 0.2785724204971532,
+      "grad_norm": 3.724096746460141,
+      "learning_rate": 8.465887676700311e-06,
+      "loss": 0.4227,
+      "step": 4012
+    },
+    {
+      "epoch": 0.2786418552978753,
+      "grad_norm": 2.5674478051109877,
+      "learning_rate": 8.46507709583782e-06,
+      "loss": 0.1777,
+      "step": 4013
+    },
+    {
+      "epoch": 0.2787112900985974,
+      "grad_norm": 4.350340933384753,
+      "learning_rate": 8.464266339715652e-06,
+      "loss": 0.4746,
+      "step": 4014
+    },
+    {
+      "epoch": 0.27878072489931954,
+      "grad_norm": 2.4926137147585155,
+      "learning_rate": 8.463455408374812e-06,
+      "loss": 0.2077,
+      "step": 4015
+    },
+    {
+      "epoch": 0.2788501597000417,
+      "grad_norm": 4.068468642851547,
+      "learning_rate": 8.462644301856318e-06,
+      "loss": 0.4392,
+      "step": 4016
+    },
+    {
+      "epoch": 0.27891959450076376,
+      "grad_norm": 3.786098716381982,
+      "learning_rate": 8.461833020201191e-06,
+      "loss": 0.2454,
+      "step": 4017
+    },
+    {
+      "epoch": 0.2789890293014859,
+      "grad_norm": 3.584446980298048,
+      "learning_rate": 8.461021563450469e-06,
+      "loss": 0.3159,
+      "step": 4018
+    },
+    {
+      "epoch": 0.279058464102208,
+      "grad_norm": 3.4579509412790883,
+      "learning_rate": 8.460209931645191e-06,
+      "loss": 0.3715,
+      "step": 4019
+    },
+    {
+      "epoch": 0.27912789890293016,
+      "grad_norm": 4.6930096329954205,
+      "learning_rate": 8.45939812482641e-06,
+      "loss": 0.5753,
+      "step": 4020
+    },
+    {
+      "epoch": 0.2791973337036523,
+      "grad_norm": 3.794703968397586,
+      "learning_rate": 8.458586143035188e-06,
+      "loss": 0.3902,
+      "step": 4021
+    },
+    {
+      "epoch": 0.2792667685043744,
+      "grad_norm": 3.7278509551810712,
+      "learning_rate": 8.457773986312592e-06,
+      "loss": 0.3939,
+      "step": 4022
+    },
+    {
+      "epoch": 0.2793362033050965,
+      "grad_norm": 3.7231569350685474,
+      "learning_rate": 8.4569616546997e-06,
+      "loss": 0.4265,
+      "step": 4023
+    },
+    {
+      "epoch": 0.27940563810581864,
+      "grad_norm": 3.800563153929813,
+      "learning_rate": 8.4561491482376e-06,
+      "loss": 0.3028,
+      "step": 4024
+    },
+    {
+      "epoch": 0.2794750729065408,
+      "grad_norm": 7.606980929230774,
+      "learning_rate": 8.455336466967387e-06,
+      "loss": 0.4427,
+      "step": 4025
+    },
+    {
+      "epoch": 0.27954450770726286,
+      "grad_norm": 3.656863151236229,
+      "learning_rate": 8.454523610930165e-06,
+      "loss": 0.4883,
+      "step": 4026
+    },
+    {
+      "epoch": 0.279613942507985,
+      "grad_norm": 3.8517080664818066,
+      "learning_rate": 8.45371058016705e-06,
+      "loss": 0.399,
+      "step": 4027
+    },
+    {
+      "epoch": 0.2796833773087071,
+      "grad_norm": 4.129533158578845,
+      "learning_rate": 8.45289737471916e-06,
+      "loss": 0.4375,
+      "step": 4028
+    },
+    {
+      "epoch": 0.27975281210942926,
+      "grad_norm": 3.5278347102911156,
+      "learning_rate": 8.452083994627629e-06,
+      "loss": 0.316,
+      "step": 4029
+    },
+    {
+      "epoch": 0.2798222469101514,
+      "grad_norm": 3.5101761845345,
+      "learning_rate": 8.451270439933595e-06,
+      "loss": 0.4084,
+      "step": 4030
+    },
+    {
+      "epoch": 0.2798916817108735,
+      "grad_norm": 3.8681780688047662,
+      "learning_rate": 8.450456710678205e-06,
+      "loss": 0.4593,
+      "step": 4031
+    },
+    {
+      "epoch": 0.2799611165115956,
+      "grad_norm": 5.160114530704963,
+      "learning_rate": 8.449642806902623e-06,
+      "loss": 0.4941,
+      "step": 4032
+    },
+    {
+      "epoch": 0.28003055131231774,
+      "grad_norm": 2.939221119622317,
+      "learning_rate": 8.44882872864801e-06,
+      "loss": 0.376,
+      "step": 4033
+    },
+    {
+      "epoch": 0.2800999861130399,
+      "grad_norm": 4.070923858188423,
+      "learning_rate": 8.44801447595554e-06,
+      "loss": 0.3259,
+      "step": 4034
+    },
+    {
+      "epoch": 0.28016942091376196,
+      "grad_norm": 4.677821186775436,
+      "learning_rate": 8.447200048866404e-06,
+      "loss": 0.5048,
+      "step": 4035
+    },
+    {
+      "epoch": 0.2802388557144841,
+      "grad_norm": 4.006686975978456,
+      "learning_rate": 8.446385447421788e-06,
+      "loss": 0.4987,
+      "step": 4036
+    },
+    {
+      "epoch": 0.2803082905152062,
+      "grad_norm": 3.348060900891605,
+      "learning_rate": 8.445570671662897e-06,
+      "loss": 0.4799,
+      "step": 4037
+    },
+    {
+      "epoch": 0.28037772531592836,
+      "grad_norm": 4.516830064846909,
+      "learning_rate": 8.44475572163094e-06,
+      "loss": 0.7503,
+      "step": 4038
+    },
+    {
+      "epoch": 0.28044716011665044,
+      "grad_norm": 5.02513680369139,
+      "learning_rate": 8.443940597367135e-06,
+      "loss": 0.7388,
+      "step": 4039
+    },
+    {
+      "epoch": 0.2805165949173726,
+      "grad_norm": 3.4592416976367395,
+      "learning_rate": 8.443125298912713e-06,
+      "loss": 0.3575,
+      "step": 4040
+    },
+    {
+      "epoch": 0.2805860297180947,
+      "grad_norm": 3.4929084233964676,
+      "learning_rate": 8.442309826308909e-06,
+      "loss": 0.4165,
+      "step": 4041
+    },
+    {
+      "epoch": 0.28065546451881684,
+      "grad_norm": 3.301489716982089,
+      "learning_rate": 8.441494179596969e-06,
+      "loss": 0.401,
+      "step": 4042
+    },
+    {
+      "epoch": 0.280724899319539,
+      "grad_norm": 3.9330085755571256,
+      "learning_rate": 8.440678358818149e-06,
+      "loss": 0.4365,
+      "step": 4043
+    },
+    {
+      "epoch": 0.28079433412026106,
+      "grad_norm": 4.107595698802399,
+      "learning_rate": 8.43986236401371e-06,
+      "loss": 0.4836,
+      "step": 4044
+    },
+    {
+      "epoch": 0.2808637689209832,
+      "grad_norm": 4.022721885102071,
+      "learning_rate": 8.439046195224925e-06,
+      "loss": 0.4467,
+      "step": 4045
+    },
+    {
+      "epoch": 0.2809332037217053,
+      "grad_norm": 4.01047608546661,
+      "learning_rate": 8.438229852493077e-06,
+      "loss": 0.514,
+      "step": 4046
+    },
+    {
+      "epoch": 0.28100263852242746,
+      "grad_norm": 3.3408439635940472,
+      "learning_rate": 8.43741333585945e-06,
+      "loss": 0.3194,
+      "step": 4047
+    },
+    {
+      "epoch": 0.28107207332314954,
+      "grad_norm": 4.533620842178191,
+      "learning_rate": 8.436596645365348e-06,
+      "loss": 0.4275,
+      "step": 4048
+    },
+    {
+      "epoch": 0.2811415081238717,
+      "grad_norm": 3.6033551587674872,
+      "learning_rate": 8.435779781052076e-06,
+      "loss": 0.543,
+      "step": 4049
+    },
+    {
+      "epoch": 0.2812109429245938,
+      "grad_norm": 4.482075488511583,
+      "learning_rate": 8.434962742960952e-06,
+      "loss": 0.7,
+      "step": 4050
+    },
+    {
+      "epoch": 0.28128037772531594,
+      "grad_norm": 3.8171253109635694,
+      "learning_rate": 8.434145531133298e-06,
+      "loss": 0.4778,
+      "step": 4051
+    },
+    {
+      "epoch": 0.281349812526038,
+      "grad_norm": 4.695338771151494,
+      "learning_rate": 8.433328145610452e-06,
+      "loss": 0.6964,
+      "step": 4052
+    },
+    {
+      "epoch": 0.28141924732676016,
+      "grad_norm": 2.7371675297922575,
+      "learning_rate": 8.43251058643375e-06,
+      "loss": 0.2559,
+      "step": 4053
+    },
+    {
+      "epoch": 0.2814886821274823,
+      "grad_norm": 4.527182311600126,
+      "learning_rate": 8.431692853644547e-06,
+      "loss": 0.6303,
+      "step": 4054
+    },
+    {
+      "epoch": 0.2815581169282044,
+      "grad_norm": 4.098375824420949,
+      "learning_rate": 8.430874947284204e-06,
+      "loss": 0.4423,
+      "step": 4055
+    },
+    {
+      "epoch": 0.28162755172892656,
+      "grad_norm": 3.4640817191640783,
+      "learning_rate": 8.430056867394087e-06,
+      "loss": 0.3723,
+      "step": 4056
+    },
+    {
+      "epoch": 0.28169698652964864,
+      "grad_norm": 2.719024683888056,
+      "learning_rate": 8.429238614015575e-06,
+      "loss": 0.2484,
+      "step": 4057
+    },
+    {
+      "epoch": 0.2817664213303708,
+      "grad_norm": 4.010671736206263,
+      "learning_rate": 8.428420187190057e-06,
+      "loss": 0.5737,
+      "step": 4058
+    },
+    {
+      "epoch": 0.2818358561310929,
+      "grad_norm": 4.307792403034726,
+      "learning_rate": 8.427601586958923e-06,
+      "loss": 0.7896,
+      "step": 4059
+    },
+    {
+      "epoch": 0.28190529093181504,
+      "grad_norm": 6.6604089718524,
+      "learning_rate": 8.42678281336358e-06,
+      "loss": 0.4925,
+      "step": 4060
+    },
+    {
+      "epoch": 0.2819747257325371,
+      "grad_norm": 4.95119217689486,
+      "learning_rate": 8.42596386644544e-06,
+      "loss": 0.4747,
+      "step": 4061
+    },
+    {
+      "epoch": 0.28204416053325926,
+      "grad_norm": 4.386565032445933,
+      "learning_rate": 8.425144746245925e-06,
+      "loss": 0.7824,
+      "step": 4062
+    },
+    {
+      "epoch": 0.2821135953339814,
+      "grad_norm": 4.022928500527897,
+      "learning_rate": 8.424325452806463e-06,
+      "loss": 0.4656,
+      "step": 4063
+    },
+    {
+      "epoch": 0.2821830301347035,
+      "grad_norm": 3.730351474349969,
+      "learning_rate": 8.423505986168497e-06,
+      "loss": 0.5111,
+      "step": 4064
+    },
+    {
+      "epoch": 0.28225246493542566,
+      "grad_norm": 3.855507767179331,
+      "learning_rate": 8.42268634637347e-06,
+      "loss": 0.3539,
+      "step": 4065
+    },
+    {
+      "epoch": 0.28232189973614774,
+      "grad_norm": 3.236919774150023,
+      "learning_rate": 8.421866533462841e-06,
+      "loss": 0.2889,
+      "step": 4066
+    },
+    {
+      "epoch": 0.2823913345368699,
+      "grad_norm": 3.413153721224843,
+      "learning_rate": 8.421046547478074e-06,
+      "loss": 0.465,
+      "step": 4067
+    },
+    {
+      "epoch": 0.282460769337592,
+      "grad_norm": 4.702650215184505,
+      "learning_rate": 8.420226388460647e-06,
+      "loss": 0.6201,
+      "step": 4068
+    },
+    {
+      "epoch": 0.28253020413831414,
+      "grad_norm": 4.891329942037103,
+      "learning_rate": 8.419406056452035e-06,
+      "loss": 0.5424,
+      "step": 4069
+    },
+    {
+      "epoch": 0.2825996389390362,
+      "grad_norm": 3.760425391817513,
+      "learning_rate": 8.418585551493736e-06,
+      "loss": 0.5935,
+      "step": 4070
+    },
+    {
+      "epoch": 0.28266907373975836,
+      "grad_norm": 4.930744264288468,
+      "learning_rate": 8.41776487362725e-06,
+      "loss": 1.0282,
+      "step": 4071
+    },
+    {
+      "epoch": 0.2827385085404805,
+      "grad_norm": 2.9914054410193662,
+      "learning_rate": 8.416944022894082e-06,
+      "loss": 0.3457,
+      "step": 4072
+    },
+    {
+      "epoch": 0.28280794334120263,
+      "grad_norm": 3.226931845016531,
+      "learning_rate": 8.416122999335754e-06,
+      "loss": 0.4147,
+      "step": 4073
+    },
+    {
+      "epoch": 0.2828773781419247,
+      "grad_norm": 3.7500736773139987,
+      "learning_rate": 8.415301802993787e-06,
+      "loss": 0.5911,
+      "step": 4074
+    },
+    {
+      "epoch": 0.28294681294264684,
+      "grad_norm": 3.4392896032730147,
+      "learning_rate": 8.414480433909721e-06,
+      "loss": 0.4706,
+      "step": 4075
+    },
+    {
+      "epoch": 0.283016247743369,
+      "grad_norm": 3.9289255040722453,
+      "learning_rate": 8.413658892125099e-06,
+      "loss": 0.51,
+      "step": 4076
+    },
+    {
+      "epoch": 0.2830856825440911,
+      "grad_norm": 3.9453723048772043,
+      "learning_rate": 8.412837177681474e-06,
+      "loss": 0.5765,
+      "step": 4077
+    },
+    {
+      "epoch": 0.28315511734481325,
+      "grad_norm": 3.8218365106942445,
+      "learning_rate": 8.412015290620405e-06,
+      "loss": 0.3521,
+      "step": 4078
+    },
+    {
+      "epoch": 0.2832245521455353,
+      "grad_norm": 4.773772181346369,
+      "learning_rate": 8.411193230983463e-06,
+      "loss": 0.5365,
+      "step": 4079
+    },
+    {
+      "epoch": 0.28329398694625746,
+      "grad_norm": 3.623292174646872,
+      "learning_rate": 8.41037099881223e-06,
+      "loss": 0.3438,
+      "step": 4080
+    },
+    {
+      "epoch": 0.2833634217469796,
+      "grad_norm": 3.575940874712963,
+      "learning_rate": 8.409548594148289e-06,
+      "loss": 0.3277,
+      "step": 4081
+    },
+    {
+      "epoch": 0.28343285654770173,
+      "grad_norm": 3.4379510268769806,
+      "learning_rate": 8.408726017033237e-06,
+      "loss": 0.5344,
+      "step": 4082
+    },
+    {
+      "epoch": 0.2835022913484238,
+      "grad_norm": 3.6130402586921346,
+      "learning_rate": 8.407903267508681e-06,
+      "loss": 0.2961,
+      "step": 4083
+    },
+    {
+      "epoch": 0.28357172614914594,
+      "grad_norm": 4.297132562603742,
+      "learning_rate": 8.407080345616236e-06,
+      "loss": 0.5173,
+      "step": 4084
+    },
+    {
+      "epoch": 0.2836411609498681,
+      "grad_norm": 3.4587881960059406,
+      "learning_rate": 8.406257251397522e-06,
+      "loss": 0.4273,
+      "step": 4085
+    },
+    {
+      "epoch": 0.2837105957505902,
+      "grad_norm": 4.004248957218643,
+      "learning_rate": 8.405433984894169e-06,
+      "loss": 0.5476,
+      "step": 4086
+    },
+    {
+      "epoch": 0.2837800305513123,
+      "grad_norm": 3.2863216151252006,
+      "learning_rate": 8.40461054614782e-06,
+      "loss": 0.3802,
+      "step": 4087
+    },
+    {
+      "epoch": 0.2838494653520344,
+      "grad_norm": 3.0987867162244966,
+      "learning_rate": 8.403786935200121e-06,
+      "loss": 0.4116,
+      "step": 4088
+    },
+    {
+      "epoch": 0.28391890015275656,
+      "grad_norm": 4.509268572891679,
+      "learning_rate": 8.40296315209273e-06,
+      "loss": 0.6288,
+      "step": 4089
+    },
+    {
+      "epoch": 0.2839883349534787,
+      "grad_norm": 4.253878629277526,
+      "learning_rate": 8.402139196867317e-06,
+      "loss": 0.5307,
+      "step": 4090
+    },
+    {
+      "epoch": 0.28405776975420083,
+      "grad_norm": 3.8227184164159977,
+      "learning_rate": 8.40131506956555e-06,
+      "loss": 0.4914,
+      "step": 4091
+    },
+    {
+      "epoch": 0.2841272045549229,
+      "grad_norm": 4.528993929907842,
+      "learning_rate": 8.400490770229115e-06,
+      "loss": 0.5349,
+      "step": 4092
+    },
+    {
+      "epoch": 0.28419663935564504,
+      "grad_norm": 4.99315911486536,
+      "learning_rate": 8.399666298899706e-06,
+      "loss": 0.6564,
+      "step": 4093
+    },
+    {
+      "epoch": 0.2842660741563672,
+      "grad_norm": 4.7847903010010855,
+      "learning_rate": 8.398841655619024e-06,
+      "loss": 0.8837,
+      "step": 4094
+    },
+    {
+      "epoch": 0.2843355089570893,
+      "grad_norm": 4.487162351864438,
+      "learning_rate": 8.398016840428776e-06,
+      "loss": 0.5113,
+      "step": 4095
+    },
+    {
+      "epoch": 0.2844049437578114,
+      "grad_norm": 3.58871113688031,
+      "learning_rate": 8.39719185337068e-06,
+      "loss": 0.5797,
+      "step": 4096
+    },
+    {
+      "epoch": 0.2844743785585335,
+      "grad_norm": 3.9166396000144355,
+      "learning_rate": 8.396366694486466e-06,
+      "loss": 0.4529,
+      "step": 4097
+    },
+    {
+      "epoch": 0.28454381335925566,
+      "grad_norm": 3.42310090575979,
+      "learning_rate": 8.395541363817868e-06,
+      "loss": 0.4789,
+      "step": 4098
+    },
+    {
+      "epoch": 0.2846132481599778,
+      "grad_norm": 4.081190130535716,
+      "learning_rate": 8.394715861406628e-06,
+      "loss": 0.624,
+      "step": 4099
+    },
+    {
+      "epoch": 0.28468268296069993,
+      "grad_norm": 3.951274934045624,
+      "learning_rate": 8.393890187294504e-06,
+      "loss": 0.4355,
+      "step": 4100
+    },
+    {
+      "epoch": 0.284752117761422,
+      "grad_norm": 3.998624590371367,
+      "learning_rate": 8.393064341523253e-06,
+      "loss": 0.6357,
+      "step": 4101
+    },
+    {
+      "epoch": 0.28482155256214414,
+      "grad_norm": 2.9383962896418834,
+      "learning_rate": 8.392238324134646e-06,
+      "loss": 0.3114,
+      "step": 4102
+    },
+    {
+      "epoch": 0.2848909873628663,
+      "grad_norm": 3.6078407916489503,
+      "learning_rate": 8.391412135170467e-06,
+      "loss": 0.5298,
+      "step": 4103
+    },
+    {
+      "epoch": 0.2849604221635884,
+      "grad_norm": 3.4590017908215427,
+      "learning_rate": 8.390585774672494e-06,
+      "loss": 0.5292,
+      "step": 4104
+    },
+    {
+      "epoch": 0.2850298569643105,
+      "grad_norm": 3.3611177969892645,
+      "learning_rate": 8.389759242682534e-06,
+      "loss": 0.3649,
+      "step": 4105
+    },
+    {
+      "epoch": 0.2850992917650326,
+      "grad_norm": 5.057511271296704,
+      "learning_rate": 8.388932539242387e-06,
+      "loss": 0.5748,
+      "step": 4106
+    },
+    {
+      "epoch": 0.28516872656575476,
+      "grad_norm": 3.410356217128276,
+      "learning_rate": 8.388105664393866e-06,
+      "loss": 0.5103,
+      "step": 4107
+    },
+    {
+      "epoch": 0.2852381613664769,
+      "grad_norm": 1.63942889339211,
+      "learning_rate": 8.387278618178794e-06,
+      "loss": 0.1453,
+      "step": 4108
+    },
+    {
+      "epoch": 0.285307596167199,
+      "grad_norm": 3.5302353437661846,
+      "learning_rate": 8.386451400639002e-06,
+      "loss": 0.5076,
+      "step": 4109
+    },
+    {
+      "epoch": 0.2853770309679211,
+      "grad_norm": 3.6999585792976784,
+      "learning_rate": 8.385624011816332e-06,
+      "loss": 0.5428,
+      "step": 4110
+    },
+    {
+      "epoch": 0.28544646576864324,
+      "grad_norm": 3.3792429300962734,
+      "learning_rate": 8.384796451752627e-06,
+      "loss": 0.278,
+      "step": 4111
+    },
+    {
+      "epoch": 0.2855159005693654,
+      "grad_norm": 3.8187010986034737,
+      "learning_rate": 8.383968720489752e-06,
+      "loss": 0.4468,
+      "step": 4112
+    },
+    {
+      "epoch": 0.2855853353700875,
+      "grad_norm": 4.9680420258857545,
+      "learning_rate": 8.383140818069566e-06,
+      "loss": 0.2834,
+      "step": 4113
+    },
+    {
+      "epoch": 0.2856547701708096,
+      "grad_norm": 3.8259416083641393,
+      "learning_rate": 8.382312744533947e-06,
+      "loss": 0.5547,
+      "step": 4114
+    },
+    {
+      "epoch": 0.2857242049715317,
+      "grad_norm": 4.6858429745538945,
+      "learning_rate": 8.381484499924775e-06,
+      "loss": 0.5408,
+      "step": 4115
+    },
+    {
+      "epoch": 0.28579363977225386,
+      "grad_norm": 2.685338585231715,
+      "learning_rate": 8.380656084283945e-06,
+      "loss": 0.3507,
+      "step": 4116
+    },
+    {
+      "epoch": 0.285863074572976,
+      "grad_norm": 3.2446294873849277,
+      "learning_rate": 8.379827497653354e-06,
+      "loss": 0.4532,
+      "step": 4117
+    },
+    {
+      "epoch": 0.2859325093736981,
+      "grad_norm": 3.4901822940792955,
+      "learning_rate": 8.378998740074916e-06,
+      "loss": 0.385,
+      "step": 4118
+    },
+    {
+      "epoch": 0.2860019441744202,
+      "grad_norm": 2.8884029656115198,
+      "learning_rate": 8.378169811590542e-06,
+      "loss": 0.2963,
+      "step": 4119
+    },
+    {
+      "epoch": 0.28607137897514234,
+      "grad_norm": 3.4122874840941764,
+      "learning_rate": 8.377340712242161e-06,
+      "loss": 0.3827,
+      "step": 4120
+    },
+    {
+      "epoch": 0.2861408137758645,
+      "grad_norm": 3.509702570015625,
+      "learning_rate": 8.376511442071711e-06,
+      "loss": 0.3106,
+      "step": 4121
+    },
+    {
+      "epoch": 0.2862102485765866,
+      "grad_norm": 3.8734074011445365,
+      "learning_rate": 8.375682001121132e-06,
+      "loss": 0.5135,
+      "step": 4122
+    },
+    {
+      "epoch": 0.2862796833773087,
+      "grad_norm": 3.033185463260702,
+      "learning_rate": 8.374852389432378e-06,
+      "loss": 0.3503,
+      "step": 4123
+    },
+    {
+      "epoch": 0.2863491181780308,
+      "grad_norm": 3.7384426901948555,
+      "learning_rate": 8.374022607047409e-06,
+      "loss": 0.4274,
+      "step": 4124
+    },
+    {
+      "epoch": 0.28641855297875296,
+      "grad_norm": 2.8769429499840355,
+      "learning_rate": 8.373192654008194e-06,
+      "loss": 0.1879,
+      "step": 4125
+    },
+    {
+      "epoch": 0.2864879877794751,
+      "grad_norm": 4.174490762887198,
+      "learning_rate": 8.372362530356711e-06,
+      "loss": 0.4198,
+      "step": 4126
+    },
+    {
+      "epoch": 0.2865574225801972,
+      "grad_norm": 3.2563087393739174,
+      "learning_rate": 8.371532236134947e-06,
+      "loss": 0.5054,
+      "step": 4127
+    },
+    {
+      "epoch": 0.2866268573809193,
+      "grad_norm": 3.7420856353625744,
+      "learning_rate": 8.3707017713849e-06,
+      "loss": 0.243,
+      "step": 4128
+    },
+    {
+      "epoch": 0.28669629218164144,
+      "grad_norm": 2.9865918320588025,
+      "learning_rate": 8.36987113614857e-06,
+      "loss": 0.2106,
+      "step": 4129
+    },
+    {
+      "epoch": 0.2867657269823636,
+      "grad_norm": 4.617791887214233,
+      "learning_rate": 8.36904033046797e-06,
+      "loss": 0.5665,
+      "step": 4130
+    },
+    {
+      "epoch": 0.28683516178308566,
+      "grad_norm": 3.5611140232804797,
+      "learning_rate": 8.368209354385124e-06,
+      "loss": 0.3485,
+      "step": 4131
+    },
+    {
+      "epoch": 0.2869045965838078,
+      "grad_norm": 4.28164494757258,
+      "learning_rate": 8.36737820794206e-06,
+      "loss": 0.4814,
+      "step": 4132
+    },
+    {
+      "epoch": 0.2869740313845299,
+      "grad_norm": 3.640662571704023,
+      "learning_rate": 8.366546891180817e-06,
+      "loss": 0.5143,
+      "step": 4133
+    },
+    {
+      "epoch": 0.28704346618525206,
+      "grad_norm": 3.7269083350101777,
+      "learning_rate": 8.365715404143442e-06,
+      "loss": 0.4231,
+      "step": 4134
+    },
+    {
+      "epoch": 0.2871129009859742,
+      "grad_norm": 4.859879121718179,
+      "learning_rate": 8.36488374687199e-06,
+      "loss": 0.8108,
+      "step": 4135
+    },
+    {
+      "epoch": 0.2871823357866963,
+      "grad_norm": 3.854094664920411,
+      "learning_rate": 8.364051919408524e-06,
+      "loss": 0.3596,
+      "step": 4136
+    },
+    {
+      "epoch": 0.2872517705874184,
+      "grad_norm": 4.492058715186599,
+      "learning_rate": 8.363219921795121e-06,
+      "loss": 0.5469,
+      "step": 4137
+    },
+    {
+      "epoch": 0.28732120538814054,
+      "grad_norm": 4.0033171953003945,
+      "learning_rate": 8.36238775407386e-06,
+      "loss": 0.6207,
+      "step": 4138
+    },
+    {
+      "epoch": 0.2873906401888627,
+      "grad_norm": 3.2335879049724516,
+      "learning_rate": 8.361555416286829e-06,
+      "loss": 0.4234,
+      "step": 4139
+    },
+    {
+      "epoch": 0.28746007498958476,
+      "grad_norm": 3.8102254306126495,
+      "learning_rate": 8.36072290847613e-06,
+      "loss": 0.6541,
+      "step": 4140
+    },
+    {
+      "epoch": 0.2875295097903069,
+      "grad_norm": 3.715870180617638,
+      "learning_rate": 8.35989023068387e-06,
+      "loss": 0.2695,
+      "step": 4141
+    },
+    {
+      "epoch": 0.287598944591029,
+      "grad_norm": 4.643962878105529,
+      "learning_rate": 8.359057382952162e-06,
+      "loss": 0.6681,
+      "step": 4142
+    },
+    {
+      "epoch": 0.28766837939175116,
+      "grad_norm": 3.863724060574862,
+      "learning_rate": 8.358224365323135e-06,
+      "loss": 0.5237,
+      "step": 4143
+    },
+    {
+      "epoch": 0.28773781419247324,
+      "grad_norm": 2.321551725792302,
+      "learning_rate": 8.35739117783892e-06,
+      "loss": 0.1565,
+      "step": 4144
+    },
+    {
+      "epoch": 0.2878072489931954,
+      "grad_norm": 4.356858765544578,
+      "learning_rate": 8.356557820541658e-06,
+      "loss": 0.6937,
+      "step": 4145
+    },
+    {
+      "epoch": 0.2878766837939175,
+      "grad_norm": 4.281247251295548,
+      "learning_rate": 8.355724293473498e-06,
+      "loss": 0.6678,
+      "step": 4146
+    },
+    {
+      "epoch": 0.28794611859463964,
+      "grad_norm": 3.753505011902372,
+      "learning_rate": 8.354890596676601e-06,
+      "loss": 0.5815,
+      "step": 4147
+    },
+    {
+      "epoch": 0.2880155533953618,
+      "grad_norm": 5.487036051490371,
+      "learning_rate": 8.354056730193136e-06,
+      "loss": 0.5379,
+      "step": 4148
+    },
+    {
+      "epoch": 0.28808498819608386,
+      "grad_norm": 4.166525356439661,
+      "learning_rate": 8.353222694065273e-06,
+      "loss": 0.6189,
+      "step": 4149
+    },
+    {
+      "epoch": 0.288154422996806,
+      "grad_norm": 3.691179444867459,
+      "learning_rate": 8.352388488335203e-06,
+      "loss": 0.2824,
+      "step": 4150
+    },
+    {
+      "epoch": 0.2882238577975281,
+      "grad_norm": 3.8384098543521077,
+      "learning_rate": 8.351554113045118e-06,
+      "loss": 0.4719,
+      "step": 4151
+    },
+    {
+      "epoch": 0.28829329259825026,
+      "grad_norm": 3.380193810299827,
+      "learning_rate": 8.350719568237216e-06,
+      "loss": 0.491,
+      "step": 4152
+    },
+    {
+      "epoch": 0.28836272739897234,
+      "grad_norm": 2.07452743196578,
+      "learning_rate": 8.349884853953712e-06,
+      "loss": 0.1928,
+      "step": 4153
+    },
+    {
+      "epoch": 0.2884321621996945,
+      "grad_norm": 4.000157877126909,
+      "learning_rate": 8.349049970236822e-06,
+      "loss": 0.5222,
+      "step": 4154
+    },
+    {
+      "epoch": 0.2885015970004166,
+      "grad_norm": 4.797546814876716,
+      "learning_rate": 8.348214917128773e-06,
+      "loss": 0.5449,
+      "step": 4155
+    },
+    {
+      "epoch": 0.28857103180113874,
+      "grad_norm": 4.106034294061708,
+      "learning_rate": 8.347379694671803e-06,
+      "loss": 0.396,
+      "step": 4156
+    },
+    {
+      "epoch": 0.2886404666018609,
+      "grad_norm": 4.209357182159019,
+      "learning_rate": 8.346544302908158e-06,
+      "loss": 0.5141,
+      "step": 4157
+    },
+    {
+      "epoch": 0.28870990140258296,
+      "grad_norm": 3.414874014017322,
+      "learning_rate": 8.345708741880086e-06,
+      "loss": 0.3919,
+      "step": 4158
+    },
+    {
+      "epoch": 0.2887793362033051,
+      "grad_norm": 4.338356040043959,
+      "learning_rate": 8.344873011629852e-06,
+      "loss": 0.5006,
+      "step": 4159
+    },
+    {
+      "epoch": 0.2888487710040272,
+      "grad_norm": 4.588598281808183,
+      "learning_rate": 8.344037112199727e-06,
+      "loss": 0.706,
+      "step": 4160
+    },
+    {
+      "epoch": 0.28891820580474936,
+      "grad_norm": 4.541332212418766,
+      "learning_rate": 8.34320104363199e-06,
+      "loss": 0.5058,
+      "step": 4161
+    },
+    {
+      "epoch": 0.28898764060547144,
+      "grad_norm": 4.90621851037414,
+      "learning_rate": 8.342364805968925e-06,
+      "loss": 0.6953,
+      "step": 4162
+    },
+    {
+      "epoch": 0.2890570754061936,
+      "grad_norm": 4.0128998344196285,
+      "learning_rate": 8.341528399252832e-06,
+      "loss": 0.54,
+      "step": 4163
+    },
+    {
+      "epoch": 0.2891265102069157,
+      "grad_norm": 4.290941781276502,
+      "learning_rate": 8.340691823526013e-06,
+      "loss": 0.7322,
+      "step": 4164
+    },
+    {
+      "epoch": 0.28919594500763784,
+      "grad_norm": 3.216600272784175,
+      "learning_rate": 8.339855078830781e-06,
+      "loss": 0.4447,
+      "step": 4165
+    },
+    {
+      "epoch": 0.2892653798083599,
+      "grad_norm": 3.829790720469176,
+      "learning_rate": 8.33901816520946e-06,
+      "loss": 0.4243,
+      "step": 4166
+    },
+    {
+      "epoch": 0.28933481460908206,
+      "grad_norm": 3.213559204009377,
+      "learning_rate": 8.338181082704377e-06,
+      "loss": 0.3345,
+      "step": 4167
+    },
+    {
+      "epoch": 0.2894042494098042,
+      "grad_norm": 3.521831945790899,
+      "learning_rate": 8.337343831357874e-06,
+      "loss": 0.4695,
+      "step": 4168
+    },
+    {
+      "epoch": 0.2894736842105263,
+      "grad_norm": 4.575263814331052,
+      "learning_rate": 8.336506411212295e-06,
+      "loss": 0.4924,
+      "step": 4169
+    },
+    {
+      "epoch": 0.28954311901124846,
+      "grad_norm": 3.7462252735678603,
+      "learning_rate": 8.335668822309997e-06,
+      "loss": 0.481,
+      "step": 4170
+    },
+    {
+      "epoch": 0.28961255381197054,
+      "grad_norm": 3.2283012195956124,
+      "learning_rate": 8.334831064693348e-06,
+      "loss": 0.5578,
+      "step": 4171
+    },
+    {
+      "epoch": 0.2896819886126927,
+      "grad_norm": 2.959677690055825,
+      "learning_rate": 8.333993138404712e-06,
+      "loss": 0.3459,
+      "step": 4172
+    },
+    {
+      "epoch": 0.2897514234134148,
+      "grad_norm": 3.88611831804658,
+      "learning_rate": 8.333155043486479e-06,
+      "loss": 0.3662,
+      "step": 4173
+    },
+    {
+      "epoch": 0.28982085821413694,
+      "grad_norm": 4.063428224272474,
+      "learning_rate": 8.332316779981035e-06,
+      "loss": 0.7501,
+      "step": 4174
+    },
+    {
+      "epoch": 0.289890293014859,
+      "grad_norm": 3.866401215203011,
+      "learning_rate": 8.33147834793078e-06,
+      "loss": 0.447,
+      "step": 4175
+    },
+    {
+      "epoch": 0.28995972781558116,
+      "grad_norm": 3.5902627745509177,
+      "learning_rate": 8.33063974737812e-06,
+      "loss": 0.4886,
+      "step": 4176
+    },
+    {
+      "epoch": 0.2900291626163033,
+      "grad_norm": 3.129036605595824,
+      "learning_rate": 8.329800978365468e-06,
+      "loss": 0.5396,
+      "step": 4177
+    },
+    {
+      "epoch": 0.29009859741702543,
+      "grad_norm": 3.550193816314146,
+      "learning_rate": 8.328962040935253e-06,
+      "loss": 0.3312,
+      "step": 4178
+    },
+    {
+      "epoch": 0.29016803221774756,
+      "grad_norm": 3.283254816387682,
+      "learning_rate": 8.328122935129902e-06,
+      "loss": 0.3887,
+      "step": 4179
+    },
+    {
+      "epoch": 0.29023746701846964,
+      "grad_norm": 3.5394317275438953,
+      "learning_rate": 8.327283660991861e-06,
+      "loss": 0.4123,
+      "step": 4180
+    },
+    {
+      "epoch": 0.2903069018191918,
+      "grad_norm": 3.0358270804421403,
+      "learning_rate": 8.326444218563578e-06,
+      "loss": 0.2669,
+      "step": 4181
+    },
+    {
+      "epoch": 0.2903763366199139,
+      "grad_norm": 3.6605537521434233,
+      "learning_rate": 8.325604607887509e-06,
+      "loss": 0.3005,
+      "step": 4182
+    },
+    {
+      "epoch": 0.29044577142063605,
+      "grad_norm": 4.828173302742655,
+      "learning_rate": 8.324764829006125e-06,
+      "loss": 0.848,
+      "step": 4183
+    },
+    {
+      "epoch": 0.2905152062213581,
+      "grad_norm": 4.931233995571509,
+      "learning_rate": 8.323924881961895e-06,
+      "loss": 0.7007,
+      "step": 4184
+    },
+    {
+      "epoch": 0.29058464102208026,
+      "grad_norm": 3.8660106780032115,
+      "learning_rate": 8.323084766797307e-06,
+      "loss": 0.5526,
+      "step": 4185
+    },
+    {
+      "epoch": 0.2906540758228024,
+      "grad_norm": 4.193697788739468,
+      "learning_rate": 8.322244483554852e-06,
+      "loss": 0.845,
+      "step": 4186
+    },
+    {
+      "epoch": 0.29072351062352453,
+      "grad_norm": 3.5803052432590787,
+      "learning_rate": 8.321404032277031e-06,
+      "loss": 0.3493,
+      "step": 4187
+    },
+    {
+      "epoch": 0.2907929454242466,
+      "grad_norm": 3.858407318636596,
+      "learning_rate": 8.32056341300635e-06,
+      "loss": 0.5402,
+      "step": 4188
+    },
+    {
+      "epoch": 0.29086238022496874,
+      "grad_norm": 4.383166192621695,
+      "learning_rate": 8.319722625785332e-06,
+      "loss": 0.6952,
+      "step": 4189
+    },
+    {
+      "epoch": 0.2909318150256909,
+      "grad_norm": 3.453880230865078,
+      "learning_rate": 8.318881670656499e-06,
+      "loss": 0.3548,
+      "step": 4190
+    },
+    {
+      "epoch": 0.291001249826413,
+      "grad_norm": 5.265988078628725,
+      "learning_rate": 8.318040547662386e-06,
+      "loss": 0.6534,
+      "step": 4191
+    },
+    {
+      "epoch": 0.29107068462713515,
+      "grad_norm": 4.6395359570415025,
+      "learning_rate": 8.317199256845536e-06,
+      "loss": 0.5682,
+      "step": 4192
+    },
+    {
+      "epoch": 0.2911401194278572,
+      "grad_norm": 4.39344953712397,
+      "learning_rate": 8.316357798248504e-06,
+      "loss": 0.2198,
+      "step": 4193
+    },
+    {
+      "epoch": 0.29120955422857936,
+      "grad_norm": 4.656650297559044,
+      "learning_rate": 8.315516171913845e-06,
+      "loss": 0.6717,
+      "step": 4194
+    },
+    {
+      "epoch": 0.2912789890293015,
+      "grad_norm": 4.548866008408077,
+      "learning_rate": 8.314674377884131e-06,
+      "loss": 0.6236,
+      "step": 4195
+    },
+    {
+      "epoch": 0.29134842383002363,
+      "grad_norm": 4.396075149918093,
+      "learning_rate": 8.313832416201937e-06,
+      "loss": 0.5899,
+      "step": 4196
+    },
+    {
+      "epoch": 0.2914178586307457,
+      "grad_norm": 4.086374431699724,
+      "learning_rate": 8.31299028690985e-06,
+      "loss": 0.4253,
+      "step": 4197
+    },
+    {
+      "epoch": 0.29148729343146784,
+      "grad_norm": 3.2217776417070767,
+      "learning_rate": 8.312147990050464e-06,
+      "loss": 0.2348,
+      "step": 4198
+    },
+    {
+      "epoch": 0.29155672823219,
+      "grad_norm": 5.312734753181491,
+      "learning_rate": 8.311305525666377e-06,
+      "loss": 1.0386,
+      "step": 4199
+    },
+    {
+      "epoch": 0.2916261630329121,
+      "grad_norm": 3.28999700181382,
+      "learning_rate": 8.310462893800207e-06,
+      "loss": 0.4045,
+      "step": 4200
+    },
+    {
+      "epoch": 0.2916955978336342,
+      "grad_norm": 2.940133710946343,
+      "learning_rate": 8.309620094494568e-06,
+      "loss": 0.2961,
+      "step": 4201
+    },
+    {
+      "epoch": 0.2917650326343563,
+      "grad_norm": 4.200256456724963,
+      "learning_rate": 8.30877712779209e-06,
+      "loss": 0.62,
+      "step": 4202
+    },
+    {
+      "epoch": 0.29183446743507846,
+      "grad_norm": 3.996763767766416,
+      "learning_rate": 8.30793399373541e-06,
+      "loss": 0.3832,
+      "step": 4203
+    },
+    {
+      "epoch": 0.2919039022358006,
+      "grad_norm": 2.696016977270572,
+      "learning_rate": 8.307090692367172e-06,
+      "loss": 0.3309,
+      "step": 4204
+    },
+    {
+      "epoch": 0.29197333703652273,
+      "grad_norm": 2.8540688523409674,
+      "learning_rate": 8.306247223730028e-06,
+      "loss": 0.3036,
+      "step": 4205
+    },
+    {
+      "epoch": 0.2920427718372448,
+      "grad_norm": 4.293694836120316,
+      "learning_rate": 8.30540358786664e-06,
+      "loss": 0.3466,
+      "step": 4206
+    },
+    {
+      "epoch": 0.29211220663796694,
+      "grad_norm": 4.22118610771197,
+      "learning_rate": 8.30455978481968e-06,
+      "loss": 0.4281,
+      "step": 4207
+    },
+    {
+      "epoch": 0.2921816414386891,
+      "grad_norm": 3.818633440563336,
+      "learning_rate": 8.303715814631825e-06,
+      "loss": 0.3839,
+      "step": 4208
+    },
+    {
+      "epoch": 0.2922510762394112,
+      "grad_norm": 3.9241789444038084,
+      "learning_rate": 8.302871677345763e-06,
+      "loss": 0.4339,
+      "step": 4209
+    },
+    {
+      "epoch": 0.2923205110401333,
+      "grad_norm": 4.1984709809832665,
+      "learning_rate": 8.302027373004189e-06,
+      "loss": 0.4579,
+      "step": 4210
+    },
+    {
+      "epoch": 0.2923899458408554,
+      "grad_norm": 5.0609571834216664,
+      "learning_rate": 8.301182901649806e-06,
+      "loss": 0.9839,
+      "step": 4211
+    },
+    {
+      "epoch": 0.29245938064157756,
+      "grad_norm": 3.326461378639018,
+      "learning_rate": 8.30033826332533e-06,
+      "loss": 0.2321,
+      "step": 4212
+    },
+    {
+      "epoch": 0.2925288154422997,
+      "grad_norm": 3.751149001931698,
+      "learning_rate": 8.299493458073475e-06,
+      "loss": 0.4685,
+      "step": 4213
+    },
+    {
+      "epoch": 0.29259825024302183,
+      "grad_norm": 4.277311702710431,
+      "learning_rate": 8.298648485936978e-06,
+      "loss": 0.6342,
+      "step": 4214
+    },
+    {
+      "epoch": 0.2926676850437439,
+      "grad_norm": 5.030596122069059,
+      "learning_rate": 8.29780334695857e-06,
+      "loss": 0.6863,
+      "step": 4215
+    },
+    {
+      "epoch": 0.29273711984446604,
+      "grad_norm": 5.348911723394327,
+      "learning_rate": 8.296958041181004e-06,
+      "loss": 0.4601,
+      "step": 4216
+    },
+    {
+      "epoch": 0.2928065546451882,
+      "grad_norm": 4.314339788885729,
+      "learning_rate": 8.29611256864703e-06,
+      "loss": 0.4571,
+      "step": 4217
+    },
+    {
+      "epoch": 0.2928759894459103,
+      "grad_norm": 3.4090964308541576,
+      "learning_rate": 8.29526692939941e-06,
+      "loss": 0.2755,
+      "step": 4218
+    },
+    {
+      "epoch": 0.2929454242466324,
+      "grad_norm": 4.276628955857692,
+      "learning_rate": 8.294421123480921e-06,
+      "loss": 0.6232,
+      "step": 4219
+    },
+    {
+      "epoch": 0.2930148590473545,
+      "grad_norm": 3.6513031781668888,
+      "learning_rate": 8.293575150934337e-06,
+      "loss": 0.4808,
+      "step": 4220
+    },
+    {
+      "epoch": 0.29308429384807666,
+      "grad_norm": 3.550774567699921,
+      "learning_rate": 8.292729011802449e-06,
+      "loss": 0.2929,
+      "step": 4221
+    },
+    {
+      "epoch": 0.2931537286487988,
+      "grad_norm": 2.8749819415608338,
+      "learning_rate": 8.291882706128054e-06,
+      "loss": 0.183,
+      "step": 4222
+    },
+    {
+      "epoch": 0.2932231634495209,
+      "grad_norm": 4.175315650986765,
+      "learning_rate": 8.291036233953957e-06,
+      "loss": 0.698,
+      "step": 4223
+    },
+    {
+      "epoch": 0.293292598250243,
+      "grad_norm": 4.845579174633432,
+      "learning_rate": 8.29018959532297e-06,
+      "loss": 0.4577,
+      "step": 4224
+    },
+    {
+      "epoch": 0.29336203305096514,
+      "grad_norm": 4.204962210770397,
+      "learning_rate": 8.289342790277918e-06,
+      "loss": 0.595,
+      "step": 4225
+    },
+    {
+      "epoch": 0.2934314678516873,
+      "grad_norm": 4.7096260595511845,
+      "learning_rate": 8.288495818861631e-06,
+      "loss": 0.609,
+      "step": 4226
+    },
+    {
+      "epoch": 0.2935009026524094,
+      "grad_norm": 3.615911736413674,
+      "learning_rate": 8.287648681116945e-06,
+      "loss": 0.5239,
+      "step": 4227
+    },
+    {
+      "epoch": 0.2935703374531315,
+      "grad_norm": 3.94130606686846,
+      "learning_rate": 8.286801377086708e-06,
+      "loss": 0.5995,
+      "step": 4228
+    },
+    {
+      "epoch": 0.2936397722538536,
+      "grad_norm": 3.555702495229184,
+      "learning_rate": 8.285953906813778e-06,
+      "loss": 0.4366,
+      "step": 4229
+    },
+    {
+      "epoch": 0.29370920705457576,
+      "grad_norm": 4.204786601128588,
+      "learning_rate": 8.285106270341017e-06,
+      "loss": 0.441,
+      "step": 4230
+    },
+    {
+      "epoch": 0.2937786418552979,
+      "grad_norm": 3.8905135890219213,
+      "learning_rate": 8.284258467711297e-06,
+      "loss": 0.4538,
+      "step": 4231
+    },
+    {
+      "epoch": 0.29384807665602,
+      "grad_norm": 3.6774063273554147,
+      "learning_rate": 8.283410498967502e-06,
+      "loss": 0.3105,
+      "step": 4232
+    },
+    {
+      "epoch": 0.2939175114567421,
+      "grad_norm": 5.488546702178218,
+      "learning_rate": 8.282562364152519e-06,
+      "loss": 0.9099,
+      "step": 4233
+    },
+    {
+      "epoch": 0.29398694625746424,
+      "grad_norm": 2.871320157189596,
+      "learning_rate": 8.281714063309246e-06,
+      "loss": 0.3757,
+      "step": 4234
+    },
+    {
+      "epoch": 0.2940563810581864,
+      "grad_norm": 3.3211984939217523,
+      "learning_rate": 8.28086559648059e-06,
+      "loss": 0.2074,
+      "step": 4235
+    },
+    {
+      "epoch": 0.2941258158589085,
+      "grad_norm": 3.603076061294814,
+      "learning_rate": 8.280016963709463e-06,
+      "loss": 0.4063,
+      "step": 4236
+    },
+    {
+      "epoch": 0.2941952506596306,
+      "grad_norm": 3.3592424320612837,
+      "learning_rate": 8.27916816503879e-06,
+      "loss": 0.386,
+      "step": 4237
+    },
+    {
+      "epoch": 0.2942646854603527,
+      "grad_norm": 5.16203435031679,
+      "learning_rate": 8.278319200511501e-06,
+      "loss": 0.4563,
+      "step": 4238
+    },
+    {
+      "epoch": 0.29433412026107486,
+      "grad_norm": 4.038750884560467,
+      "learning_rate": 8.277470070170538e-06,
+      "loss": 0.481,
+      "step": 4239
+    },
+    {
+      "epoch": 0.294403555061797,
+      "grad_norm": 3.3213110674815947,
+      "learning_rate": 8.276620774058847e-06,
+      "loss": 0.3534,
+      "step": 4240
+    },
+    {
+      "epoch": 0.2944729898625191,
+      "grad_norm": 4.172828393178461,
+      "learning_rate": 8.275771312219387e-06,
+      "loss": 0.6278,
+      "step": 4241
+    },
+    {
+      "epoch": 0.2945424246632412,
+      "grad_norm": 3.606952062268344,
+      "learning_rate": 8.274921684695119e-06,
+      "loss": 0.2889,
+      "step": 4242
+    },
+    {
+      "epoch": 0.29461185946396334,
+      "grad_norm": 4.538497812827897,
+      "learning_rate": 8.27407189152902e-06,
+      "loss": 0.6587,
+      "step": 4243
+    },
+    {
+      "epoch": 0.2946812942646855,
+      "grad_norm": 3.610867670562362,
+      "learning_rate": 8.273221932764065e-06,
+      "loss": 0.4255,
+      "step": 4244
+    },
+    {
+      "epoch": 0.29475072906540756,
+      "grad_norm": 3.7750422561544634,
+      "learning_rate": 8.27237180844325e-06,
+      "loss": 0.3736,
+      "step": 4245
+    },
+    {
+      "epoch": 0.2948201638661297,
+      "grad_norm": 3.9269836477596214,
+      "learning_rate": 8.271521518609576e-06,
+      "loss": 0.5608,
+      "step": 4246
+    },
+    {
+      "epoch": 0.2948895986668518,
+      "grad_norm": 4.025751950641143,
+      "learning_rate": 8.270671063306042e-06,
+      "loss": 0.3022,
+      "step": 4247
+    },
+    {
+      "epoch": 0.29495903346757396,
+      "grad_norm": 3.796040347308218,
+      "learning_rate": 8.269820442575667e-06,
+      "loss": 0.4126,
+      "step": 4248
+    },
+    {
+      "epoch": 0.2950284682682961,
+      "grad_norm": 4.81411674855092,
+      "learning_rate": 8.268969656461475e-06,
+      "loss": 0.6856,
+      "step": 4249
+    },
+    {
+      "epoch": 0.2950979030690182,
+      "grad_norm": 3.268076174439191,
+      "learning_rate": 8.268118705006495e-06,
+      "loss": 0.3138,
+      "step": 4250
+    },
+    {
+      "epoch": 0.2951673378697403,
+      "grad_norm": 3.201372526533332,
+      "learning_rate": 8.267267588253771e-06,
+      "loss": 0.2161,
+      "step": 4251
+    },
+    {
+      "epoch": 0.29523677267046244,
+      "grad_norm": 3.86100659409227,
+      "learning_rate": 8.26641630624635e-06,
+      "loss": 0.3946,
+      "step": 4252
+    },
+    {
+      "epoch": 0.2953062074711846,
+      "grad_norm": 4.009248968458765,
+      "learning_rate": 8.265564859027287e-06,
+      "loss": 0.3914,
+      "step": 4253
+    },
+    {
+      "epoch": 0.29537564227190666,
+      "grad_norm": 4.818749270345499,
+      "learning_rate": 8.264713246639648e-06,
+      "loss": 0.5471,
+      "step": 4254
+    },
+    {
+      "epoch": 0.2954450770726288,
+      "grad_norm": 3.500759855445182,
+      "learning_rate": 8.263861469126508e-06,
+      "loss": 0.322,
+      "step": 4255
+    },
+    {
+      "epoch": 0.2955145118733509,
+      "grad_norm": 3.71540834259118,
+      "learning_rate": 8.263009526530949e-06,
+      "loss": 0.4293,
+      "step": 4256
+    },
+    {
+      "epoch": 0.29558394667407306,
+      "grad_norm": 4.496607737120699,
+      "learning_rate": 8.262157418896059e-06,
+      "loss": 0.5934,
+      "step": 4257
+    },
+    {
+      "epoch": 0.29565338147479514,
+      "grad_norm": 4.899336327526058,
+      "learning_rate": 8.26130514626494e-06,
+      "loss": 0.7202,
+      "step": 4258
+    },
+    {
+      "epoch": 0.2957228162755173,
+      "grad_norm": 3.291350312788581,
+      "learning_rate": 8.260452708680695e-06,
+      "loss": 0.458,
+      "step": 4259
+    },
+    {
+      "epoch": 0.2957922510762394,
+      "grad_norm": 4.099156702875364,
+      "learning_rate": 8.259600106186444e-06,
+      "loss": 0.5491,
+      "step": 4260
+    },
+    {
+      "epoch": 0.29586168587696154,
+      "grad_norm": 4.462024193148102,
+      "learning_rate": 8.258747338825306e-06,
+      "loss": 0.7147,
+      "step": 4261
+    },
+    {
+      "epoch": 0.2959311206776837,
+      "grad_norm": 3.3575359608426254,
+      "learning_rate": 8.257894406640416e-06,
+      "loss": 0.412,
+      "step": 4262
+    },
+    {
+      "epoch": 0.29600055547840576,
+      "grad_norm": 3.630188523235875,
+      "learning_rate": 8.257041309674913e-06,
+      "loss": 0.6407,
+      "step": 4263
+    },
+    {
+      "epoch": 0.2960699902791279,
+      "grad_norm": 3.790671663036885,
+      "learning_rate": 8.256188047971948e-06,
+      "loss": 0.5687,
+      "step": 4264
+    },
+    {
+      "epoch": 0.29613942507985,
+      "grad_norm": 4.133396061038882,
+      "learning_rate": 8.255334621574673e-06,
+      "loss": 0.6824,
+      "step": 4265
+    },
+    {
+      "epoch": 0.29620885988057216,
+      "grad_norm": 3.8646728399417576,
+      "learning_rate": 8.254481030526258e-06,
+      "loss": 0.4167,
+      "step": 4266
+    },
+    {
+      "epoch": 0.29627829468129424,
+      "grad_norm": 5.225141120871583,
+      "learning_rate": 8.253627274869875e-06,
+      "loss": 0.5329,
+      "step": 4267
+    },
+    {
+      "epoch": 0.2963477294820164,
+      "grad_norm": 2.7938501843439356,
+      "learning_rate": 8.252773354648707e-06,
+      "loss": 0.2769,
+      "step": 4268
+    },
+    {
+      "epoch": 0.2964171642827385,
+      "grad_norm": 3.6474052755829103,
+      "learning_rate": 8.251919269905944e-06,
+      "loss": 0.4024,
+      "step": 4269
+    },
+    {
+      "epoch": 0.29648659908346064,
+      "grad_norm": 3.1463226898160834,
+      "learning_rate": 8.25106502068478e-06,
+      "loss": 0.3972,
+      "step": 4270
+    },
+    {
+      "epoch": 0.2965560338841828,
+      "grad_norm": 3.4947158724567124,
+      "learning_rate": 8.25021060702843e-06,
+      "loss": 0.6515,
+      "step": 4271
+    },
+    {
+      "epoch": 0.29662546868490486,
+      "grad_norm": 3.2985506696609166,
+      "learning_rate": 8.249356028980105e-06,
+      "loss": 0.3299,
+      "step": 4272
+    },
+    {
+      "epoch": 0.296694903485627,
+      "grad_norm": 3.4846761551504004,
+      "learning_rate": 8.248501286583029e-06,
+      "loss": 0.3759,
+      "step": 4273
+    },
+    {
+      "epoch": 0.2967643382863491,
+      "grad_norm": 3.540839269402786,
+      "learning_rate": 8.247646379880433e-06,
+      "loss": 0.4681,
+      "step": 4274
+    },
+    {
+      "epoch": 0.29683377308707126,
+      "grad_norm": 3.1937217035626473,
+      "learning_rate": 8.246791308915558e-06,
+      "loss": 0.246,
+      "step": 4275
+    },
+    {
+      "epoch": 0.29690320788779334,
+      "grad_norm": 4.303737318051593,
+      "learning_rate": 8.245936073731654e-06,
+      "loss": 0.7185,
+      "step": 4276
+    },
+    {
+      "epoch": 0.2969726426885155,
+      "grad_norm": 2.8997555707519087,
+      "learning_rate": 8.245080674371975e-06,
+      "loss": 0.2243,
+      "step": 4277
+    },
+    {
+      "epoch": 0.2970420774892376,
+      "grad_norm": 3.4736187592007743,
+      "learning_rate": 8.244225110879787e-06,
+      "loss": 0.4692,
+      "step": 4278
+    },
+    {
+      "epoch": 0.29711151228995974,
+      "grad_norm": 2.7724146308399726,
+      "learning_rate": 8.243369383298365e-06,
+      "loss": 0.3082,
+      "step": 4279
+    },
+    {
+      "epoch": 0.2971809470906818,
+      "grad_norm": 3.7652902144440166,
+      "learning_rate": 8.24251349167099e-06,
+      "loss": 0.3723,
+      "step": 4280
+    },
+    {
+      "epoch": 0.29725038189140396,
+      "grad_norm": 3.222825020753176,
+      "learning_rate": 8.24165743604095e-06,
+      "loss": 0.3276,
+      "step": 4281
+    },
+    {
+      "epoch": 0.2973198166921261,
+      "grad_norm": 4.759606560201149,
+      "learning_rate": 8.240801216451548e-06,
+      "loss": 0.6407,
+      "step": 4282
+    },
+    {
+      "epoch": 0.29738925149284823,
+      "grad_norm": 4.654161386667701,
+      "learning_rate": 8.239944832946085e-06,
+      "loss": 0.7273,
+      "step": 4283
+    },
+    {
+      "epoch": 0.29745868629357036,
+      "grad_norm": 3.2456572284308627,
+      "learning_rate": 8.23908828556788e-06,
+      "loss": 0.3964,
+      "step": 4284
+    },
+    {
+      "epoch": 0.29752812109429244,
+      "grad_norm": 4.475221443050549,
+      "learning_rate": 8.238231574360256e-06,
+      "loss": 0.4329,
+      "step": 4285
+    },
+    {
+      "epoch": 0.2975975558950146,
+      "grad_norm": 3.6377426839872715,
+      "learning_rate": 8.23737469936654e-06,
+      "loss": 0.4988,
+      "step": 4286
+    },
+    {
+      "epoch": 0.2976669906957367,
+      "grad_norm": 4.138112859934414,
+      "learning_rate": 8.236517660630077e-06,
+      "loss": 0.5797,
+      "step": 4287
+    },
+    {
+      "epoch": 0.29773642549645885,
+      "grad_norm": 3.804048089992206,
+      "learning_rate": 8.235660458194213e-06,
+      "loss": 0.5082,
+      "step": 4288
+    },
+    {
+      "epoch": 0.2978058602971809,
+      "grad_norm": 3.667172197218969,
+      "learning_rate": 8.234803092102305e-06,
+      "loss": 0.5346,
+      "step": 4289
+    },
+    {
+      "epoch": 0.29787529509790306,
+      "grad_norm": 3.749333566257004,
+      "learning_rate": 8.233945562397716e-06,
+      "loss": 0.3514,
+      "step": 4290
+    },
+    {
+      "epoch": 0.2979447298986252,
+      "grad_norm": 4.46242032534881,
+      "learning_rate": 8.23308786912382e-06,
+      "loss": 0.6833,
+      "step": 4291
+    },
+    {
+      "epoch": 0.29801416469934733,
+      "grad_norm": 4.368434279930634,
+      "learning_rate": 8.232230012323998e-06,
+      "loss": 0.5334,
+      "step": 4292
+    },
+    {
+      "epoch": 0.2980835995000694,
+      "grad_norm": 4.0289765023981925,
+      "learning_rate": 8.23137199204164e-06,
+      "loss": 0.5327,
+      "step": 4293
+    },
+    {
+      "epoch": 0.29815303430079154,
+      "grad_norm": 6.840398878298236,
+      "learning_rate": 8.230513808320144e-06,
+      "loss": 0.8362,
+      "step": 4294
+    },
+    {
+      "epoch": 0.2982224691015137,
+      "grad_norm": 3.051279309231908,
+      "learning_rate": 8.229655461202913e-06,
+      "loss": 0.4083,
+      "step": 4295
+    },
+    {
+      "epoch": 0.2982919039022358,
+      "grad_norm": 4.559531897290236,
+      "learning_rate": 8.228796950733364e-06,
+      "loss": 0.6167,
+      "step": 4296
+    },
+    {
+      "epoch": 0.29836133870295795,
+      "grad_norm": 3.018156734934253,
+      "learning_rate": 8.22793827695492e-06,
+      "loss": 0.2414,
+      "step": 4297
+    },
+    {
+      "epoch": 0.29843077350368,
+      "grad_norm": 2.991476571051541,
+      "learning_rate": 8.22707943991101e-06,
+      "loss": 0.2886,
+      "step": 4298
+    },
+    {
+      "epoch": 0.29850020830440216,
+      "grad_norm": 4.276136082720696,
+      "learning_rate": 8.226220439645074e-06,
+      "loss": 0.5661,
+      "step": 4299
+    },
+    {
+      "epoch": 0.2985696431051243,
+      "grad_norm": 4.1525673361261415,
+      "learning_rate": 8.225361276200559e-06,
+      "loss": 0.5241,
+      "step": 4300
+    },
+    {
+      "epoch": 0.29863907790584643,
+      "grad_norm": 4.226823036820217,
+      "learning_rate": 8.22450194962092e-06,
+      "loss": 0.7547,
+      "step": 4301
+    },
+    {
+      "epoch": 0.2987085127065685,
+      "grad_norm": 4.325073579184161,
+      "learning_rate": 8.22364245994962e-06,
+      "loss": 0.6188,
+      "step": 4302
+    },
+    {
+      "epoch": 0.29877794750729064,
+      "grad_norm": 2.8589516055818227,
+      "learning_rate": 8.222782807230135e-06,
+      "loss": 0.3185,
+      "step": 4303
+    },
+    {
+      "epoch": 0.2988473823080128,
+      "grad_norm": 4.27184594695777,
+      "learning_rate": 8.22192299150594e-06,
+      "loss": 0.5827,
+      "step": 4304
+    },
+    {
+      "epoch": 0.2989168171087349,
+      "grad_norm": 3.955433825000662,
+      "learning_rate": 8.221063012820527e-06,
+      "loss": 0.4045,
+      "step": 4305
+    },
+    {
+      "epoch": 0.29898625190945705,
+      "grad_norm": 3.298826754643124,
+      "learning_rate": 8.22020287121739e-06,
+      "loss": 0.3766,
+      "step": 4306
+    },
+    {
+      "epoch": 0.2990556867101791,
+      "grad_norm": 3.3878754313463446,
+      "learning_rate": 8.219342566740036e-06,
+      "loss": 0.3628,
+      "step": 4307
+    },
+    {
+      "epoch": 0.29912512151090126,
+      "grad_norm": 2.740792065932492,
+      "learning_rate": 8.21848209943198e-06,
+      "loss": 0.3518,
+      "step": 4308
+    },
+    {
+      "epoch": 0.2991945563116234,
+      "grad_norm": 4.578593118588602,
+      "learning_rate": 8.217621469336741e-06,
+      "loss": 0.7125,
+      "step": 4309
+    },
+    {
+      "epoch": 0.29926399111234553,
+      "grad_norm": 3.9051337858976773,
+      "learning_rate": 8.216760676497849e-06,
+      "loss": 0.3083,
+      "step": 4310
+    },
+    {
+      "epoch": 0.2993334259130676,
+      "grad_norm": 2.535543703457087,
+      "learning_rate": 8.21589972095884e-06,
+      "loss": 0.2933,
+      "step": 4311
+    },
+    {
+      "epoch": 0.29940286071378974,
+      "grad_norm": 4.2050038967646515,
+      "learning_rate": 8.215038602763264e-06,
+      "loss": 0.7416,
+      "step": 4312
+    },
+    {
+      "epoch": 0.2994722955145119,
+      "grad_norm": 4.5563772719799776,
+      "learning_rate": 8.214177321954672e-06,
+      "loss": 0.7102,
+      "step": 4313
+    },
+    {
+      "epoch": 0.299541730315234,
+      "grad_norm": 4.057365451691864,
+      "learning_rate": 8.21331587857663e-06,
+      "loss": 0.5877,
+      "step": 4314
+    },
+    {
+      "epoch": 0.2996111651159561,
+      "grad_norm": 4.434078074434157,
+      "learning_rate": 8.212454272672705e-06,
+      "loss": 0.4333,
+      "step": 4315
+    },
+    {
+      "epoch": 0.2996805999166782,
+      "grad_norm": 4.384880336635134,
+      "learning_rate": 8.211592504286479e-06,
+      "loss": 0.6168,
+      "step": 4316
+    },
+    {
+      "epoch": 0.29975003471740036,
+      "grad_norm": 4.150694608944207,
+      "learning_rate": 8.210730573461539e-06,
+      "loss": 0.5901,
+      "step": 4317
+    },
+    {
+      "epoch": 0.2998194695181225,
+      "grad_norm": 3.10361656018691,
+      "learning_rate": 8.209868480241479e-06,
+      "loss": 0.2815,
+      "step": 4318
+    },
+    {
+      "epoch": 0.29988890431884463,
+      "grad_norm": 3.7841265385971696,
+      "learning_rate": 8.209006224669904e-06,
+      "loss": 0.473,
+      "step": 4319
+    },
+    {
+      "epoch": 0.2999583391195667,
+      "grad_norm": 3.171945248596576,
+      "learning_rate": 8.208143806790425e-06,
+      "loss": 0.4406,
+      "step": 4320
+    },
+    {
+      "epoch": 0.30002777392028884,
+      "grad_norm": 3.6923611028444654,
+      "learning_rate": 8.207281226646662e-06,
+      "loss": 0.423,
+      "step": 4321
+    },
+    {
+      "epoch": 0.300097208721011,
+      "grad_norm": 3.184410139087182,
+      "learning_rate": 8.206418484282245e-06,
+      "loss": 0.4724,
+      "step": 4322
+    },
+    {
+      "epoch": 0.3001666435217331,
+      "grad_norm": 4.316900703355333,
+      "learning_rate": 8.20555557974081e-06,
+      "loss": 0.4758,
+      "step": 4323
+    },
+    {
+      "epoch": 0.3002360783224552,
+      "grad_norm": 5.695773996229779,
+      "learning_rate": 8.204692513065999e-06,
+      "loss": 0.6536,
+      "step": 4324
+    },
+    {
+      "epoch": 0.3003055131231773,
+      "grad_norm": 3.5423146186576067,
+      "learning_rate": 8.203829284301468e-06,
+      "loss": 0.4574,
+      "step": 4325
+    },
+    {
+      "epoch": 0.30037494792389946,
+      "grad_norm": 4.455268562250085,
+      "learning_rate": 8.202965893490877e-06,
+      "loss": 0.6349,
+      "step": 4326
+    },
+    {
+      "epoch": 0.3004443827246216,
+      "grad_norm": 3.9058474034541066,
+      "learning_rate": 8.202102340677896e-06,
+      "loss": 0.6202,
+      "step": 4327
+    },
+    {
+      "epoch": 0.30051381752534373,
+      "grad_norm": 5.310037267389661,
+      "learning_rate": 8.2012386259062e-06,
+      "loss": 0.6712,
+      "step": 4328
+    },
+    {
+      "epoch": 0.3005832523260658,
+      "grad_norm": 3.5520677768569446,
+      "learning_rate": 8.200374749219479e-06,
+      "loss": 0.402,
+      "step": 4329
+    },
+    {
+      "epoch": 0.30065268712678794,
+      "grad_norm": 3.8463060959578526,
+      "learning_rate": 8.199510710661424e-06,
+      "loss": 0.5665,
+      "step": 4330
+    },
+    {
+      "epoch": 0.3007221219275101,
+      "grad_norm": 3.3525048423676402,
+      "learning_rate": 8.198646510275738e-06,
+      "loss": 0.325,
+      "step": 4331
+    },
+    {
+      "epoch": 0.3007915567282322,
+      "grad_norm": 5.075693294061668,
+      "learning_rate": 8.19778214810613e-06,
+      "loss": 0.7931,
+      "step": 4332
+    },
+    {
+      "epoch": 0.3008609915289543,
+      "grad_norm": 3.240218155371128,
+      "learning_rate": 8.196917624196322e-06,
+      "loss": 0.2891,
+      "step": 4333
+    },
+    {
+      "epoch": 0.3009304263296764,
+      "grad_norm": 4.2114369093971,
+      "learning_rate": 8.196052938590036e-06,
+      "loss": 0.6269,
+      "step": 4334
+    },
+    {
+      "epoch": 0.30099986113039856,
+      "grad_norm": 3.5805273775393993,
+      "learning_rate": 8.19518809133101e-06,
+      "loss": 0.4408,
+      "step": 4335
+    },
+    {
+      "epoch": 0.3010692959311207,
+      "grad_norm": 2.928402808157151,
+      "learning_rate": 8.194323082462984e-06,
+      "loss": 0.5091,
+      "step": 4336
+    },
+    {
+      "epoch": 0.3011387307318428,
+      "grad_norm": 4.177516793762414,
+      "learning_rate": 8.193457912029713e-06,
+      "loss": 0.5763,
+      "step": 4337
+    },
+    {
+      "epoch": 0.3012081655325649,
+      "grad_norm": 4.0466022234030685,
+      "learning_rate": 8.192592580074954e-06,
+      "loss": 0.5488,
+      "step": 4338
+    },
+    {
+      "epoch": 0.30127760033328704,
+      "grad_norm": 3.2846480886267555,
+      "learning_rate": 8.191727086642475e-06,
+      "loss": 0.4344,
+      "step": 4339
+    },
+    {
+      "epoch": 0.3013470351340092,
+      "grad_norm": 3.9710449387973212,
+      "learning_rate": 8.190861431776051e-06,
+      "loss": 0.4172,
+      "step": 4340
+    },
+    {
+      "epoch": 0.3014164699347313,
+      "grad_norm": 4.970611595831778,
+      "learning_rate": 8.189995615519467e-06,
+      "loss": 0.7167,
+      "step": 4341
+    },
+    {
+      "epoch": 0.3014859047354534,
+      "grad_norm": 3.9192354518779866,
+      "learning_rate": 8.189129637916515e-06,
+      "loss": 0.4873,
+      "step": 4342
+    },
+    {
+      "epoch": 0.3015553395361755,
+      "grad_norm": 4.241166958266135,
+      "learning_rate": 8.188263499010993e-06,
+      "loss": 0.4216,
+      "step": 4343
+    },
+    {
+      "epoch": 0.30162477433689766,
+      "grad_norm": 4.231452814182144,
+      "learning_rate": 8.187397198846714e-06,
+      "loss": 0.5637,
+      "step": 4344
+    },
+    {
+      "epoch": 0.3016942091376198,
+      "grad_norm": 3.42862198225882,
+      "learning_rate": 8.186530737467488e-06,
+      "loss": 0.3464,
+      "step": 4345
+    },
+    {
+      "epoch": 0.3017636439383419,
+      "grad_norm": 3.3195594374595845,
+      "learning_rate": 8.185664114917146e-06,
+      "loss": 0.4057,
+      "step": 4346
+    },
+    {
+      "epoch": 0.301833078739064,
+      "grad_norm": 5.168701984619272,
+      "learning_rate": 8.184797331239515e-06,
+      "loss": 0.5304,
+      "step": 4347
+    },
+    {
+      "epoch": 0.30190251353978614,
+      "grad_norm": 3.5979356319784097,
+      "learning_rate": 8.183930386478442e-06,
+      "loss": 0.3188,
+      "step": 4348
+    },
+    {
+      "epoch": 0.3019719483405083,
+      "grad_norm": 4.535023099458533,
+      "learning_rate": 8.18306328067777e-06,
+      "loss": 0.7025,
+      "step": 4349
+    },
+    {
+      "epoch": 0.30204138314123036,
+      "grad_norm": 4.083349874596811,
+      "learning_rate": 8.18219601388136e-06,
+      "loss": 0.4164,
+      "step": 4350
+    },
+    {
+      "epoch": 0.3021108179419525,
+      "grad_norm": 3.3088679511209684,
+      "learning_rate": 8.181328586133078e-06,
+      "loss": 0.1994,
+      "step": 4351
+    },
+    {
+      "epoch": 0.3021802527426746,
+      "grad_norm": 3.45560539032056,
+      "learning_rate": 8.180460997476796e-06,
+      "loss": 0.346,
+      "step": 4352
+    },
+    {
+      "epoch": 0.30224968754339676,
+      "grad_norm": 3.8546845112624477,
+      "learning_rate": 8.179593247956394e-06,
+      "loss": 0.4427,
+      "step": 4353
+    },
+    {
+      "epoch": 0.3023191223441189,
+      "grad_norm": 2.5079896357619123,
+      "learning_rate": 8.178725337615764e-06,
+      "loss": 0.178,
+      "step": 4354
+    },
+    {
+      "epoch": 0.302388557144841,
+      "grad_norm": 2.6722945743559783,
+      "learning_rate": 8.177857266498804e-06,
+      "loss": 0.222,
+      "step": 4355
+    },
+    {
+      "epoch": 0.3024579919455631,
+      "grad_norm": 4.088999251563087,
+      "learning_rate": 8.176989034649419e-06,
+      "loss": 0.4666,
+      "step": 4356
+    },
+    {
+      "epoch": 0.30252742674628524,
+      "grad_norm": 4.582462299929816,
+      "learning_rate": 8.176120642111523e-06,
+      "loss": 0.5749,
+      "step": 4357
+    },
+    {
+      "epoch": 0.3025968615470074,
+      "grad_norm": 4.506313844374007,
+      "learning_rate": 8.17525208892904e-06,
+      "loss": 0.6591,
+      "step": 4358
+    },
+    {
+      "epoch": 0.30266629634772946,
+      "grad_norm": 4.668941509840434,
+      "learning_rate": 8.1743833751459e-06,
+      "loss": 0.8776,
+      "step": 4359
+    },
+    {
+      "epoch": 0.3027357311484516,
+      "grad_norm": 2.853682001796087,
+      "learning_rate": 8.173514500806039e-06,
+      "loss": 0.3936,
+      "step": 4360
+    },
+    {
+      "epoch": 0.3028051659491737,
+      "grad_norm": 3.2192634537315388,
+      "learning_rate": 8.172645465953405e-06,
+      "loss": 0.3946,
+      "step": 4361
+    },
+    {
+      "epoch": 0.30287460074989586,
+      "grad_norm": 2.8210479254113032,
+      "learning_rate": 8.171776270631958e-06,
+      "loss": 0.4015,
+      "step": 4362
+    },
+    {
+      "epoch": 0.302944035550618,
+      "grad_norm": 3.8432029232622846,
+      "learning_rate": 8.170906914885651e-06,
+      "loss": 0.4156,
+      "step": 4363
+    },
+    {
+      "epoch": 0.3030134703513401,
+      "grad_norm": 3.272333744479109,
+      "learning_rate": 8.170037398758464e-06,
+      "loss": 0.4669,
+      "step": 4364
+    },
+    {
+      "epoch": 0.3030829051520622,
+      "grad_norm": 3.055422708425608,
+      "learning_rate": 8.16916772229437e-06,
+      "loss": 0.3,
+      "step": 4365
+    },
+    {
+      "epoch": 0.30315233995278434,
+      "grad_norm": 4.141494613464518,
+      "learning_rate": 8.16829788553736e-06,
+      "loss": 0.5087,
+      "step": 4366
+    },
+    {
+      "epoch": 0.3032217747535065,
+      "grad_norm": 4.179485824864375,
+      "learning_rate": 8.167427888531429e-06,
+      "loss": 0.5049,
+      "step": 4367
+    },
+    {
+      "epoch": 0.30329120955422856,
+      "grad_norm": 3.544204503499928,
+      "learning_rate": 8.166557731320577e-06,
+      "loss": 0.5535,
+      "step": 4368
+    },
+    {
+      "epoch": 0.3033606443549507,
+      "grad_norm": 5.7066224761601125,
+      "learning_rate": 8.165687413948821e-06,
+      "loss": 0.6478,
+      "step": 4369
+    },
+    {
+      "epoch": 0.3034300791556728,
+      "grad_norm": 3.7911779992224637,
+      "learning_rate": 8.164816936460176e-06,
+      "loss": 0.5755,
+      "step": 4370
+    },
+    {
+      "epoch": 0.30349951395639496,
+      "grad_norm": 3.5305719064361503,
+      "learning_rate": 8.163946298898673e-06,
+      "loss": 0.4962,
+      "step": 4371
+    },
+    {
+      "epoch": 0.30356894875711704,
+      "grad_norm": 2.6845894481267187,
+      "learning_rate": 8.163075501308345e-06,
+      "loss": 0.4123,
+      "step": 4372
+    },
+    {
+      "epoch": 0.3036383835578392,
+      "grad_norm": 3.76521304095693,
+      "learning_rate": 8.162204543733238e-06,
+      "loss": 0.4075,
+      "step": 4373
+    },
+    {
+      "epoch": 0.3037078183585613,
+      "grad_norm": 4.606683924172447,
+      "learning_rate": 8.161333426217404e-06,
+      "loss": 0.7127,
+      "step": 4374
+    },
+    {
+      "epoch": 0.30377725315928344,
+      "grad_norm": 3.3956887826231203,
+      "learning_rate": 8.160462148804902e-06,
+      "loss": 0.4297,
+      "step": 4375
+    },
+    {
+      "epoch": 0.3038466879600056,
+      "grad_norm": 3.0227703429094697,
+      "learning_rate": 8.159590711539802e-06,
+      "loss": 0.3099,
+      "step": 4376
+    },
+    {
+      "epoch": 0.30391612276072766,
+      "grad_norm": 3.6803607657322917,
+      "learning_rate": 8.158719114466176e-06,
+      "loss": 0.3429,
+      "step": 4377
+    },
+    {
+      "epoch": 0.3039855575614498,
+      "grad_norm": 3.8038208520649572,
+      "learning_rate": 8.157847357628115e-06,
+      "loss": 0.4823,
+      "step": 4378
+    },
+    {
+      "epoch": 0.3040549923621719,
+      "grad_norm": 3.235762413282526,
+      "learning_rate": 8.156975441069705e-06,
+      "loss": 0.3557,
+      "step": 4379
+    },
+    {
+      "epoch": 0.30412442716289406,
+      "grad_norm": 3.7474106346540084,
+      "learning_rate": 8.156103364835052e-06,
+      "loss": 0.337,
+      "step": 4380
+    },
+    {
+      "epoch": 0.30419386196361614,
+      "grad_norm": 3.70037265898967,
+      "learning_rate": 8.155231128968262e-06,
+      "loss": 0.4728,
+      "step": 4381
+    },
+    {
+      "epoch": 0.3042632967643383,
+      "grad_norm": 3.614290543234798,
+      "learning_rate": 8.15435873351345e-06,
+      "loss": 0.4484,
+      "step": 4382
+    },
+    {
+      "epoch": 0.3043327315650604,
+      "grad_norm": 4.517794589641213,
+      "learning_rate": 8.153486178514744e-06,
+      "loss": 0.5234,
+      "step": 4383
+    },
+    {
+      "epoch": 0.30440216636578254,
+      "grad_norm": 3.6952257771516246,
+      "learning_rate": 8.152613464016278e-06,
+      "loss": 0.5045,
+      "step": 4384
+    },
+    {
+      "epoch": 0.3044716011665047,
+      "grad_norm": 3.382801037918305,
+      "learning_rate": 8.151740590062187e-06,
+      "loss": 0.3345,
+      "step": 4385
+    },
+    {
+      "epoch": 0.30454103596722676,
+      "grad_norm": 4.118293481703411,
+      "learning_rate": 8.150867556696623e-06,
+      "loss": 0.3631,
+      "step": 4386
+    },
+    {
+      "epoch": 0.3046104707679489,
+      "grad_norm": 5.414131044206265,
+      "learning_rate": 8.149994363963745e-06,
+      "loss": 0.486,
+      "step": 4387
+    },
+    {
+      "epoch": 0.30467990556867103,
+      "grad_norm": 4.4491637788055085,
+      "learning_rate": 8.149121011907715e-06,
+      "loss": 0.5188,
+      "step": 4388
+    },
+    {
+      "epoch": 0.30474934036939316,
+      "grad_norm": 2.8998483139193536,
+      "learning_rate": 8.14824750057271e-06,
+      "loss": 0.3033,
+      "step": 4389
+    },
+    {
+      "epoch": 0.30481877517011524,
+      "grad_norm": 4.5882046790101585,
+      "learning_rate": 8.147373830002906e-06,
+      "loss": 0.475,
+      "step": 4390
+    },
+    {
+      "epoch": 0.3048882099708374,
+      "grad_norm": 3.837718364234041,
+      "learning_rate": 8.1465000002425e-06,
+      "loss": 0.5431,
+      "step": 4391
+    },
+    {
+      "epoch": 0.3049576447715595,
+      "grad_norm": 4.3215187227720735,
+      "learning_rate": 8.14562601133568e-06,
+      "loss": 0.4061,
+      "step": 4392
+    },
+    {
+      "epoch": 0.30502707957228165,
+      "grad_norm": 4.082962065203325,
+      "learning_rate": 8.144751863326656e-06,
+      "loss": 0.375,
+      "step": 4393
+    },
+    {
+      "epoch": 0.3050965143730037,
+      "grad_norm": 3.9799476281434676,
+      "learning_rate": 8.143877556259644e-06,
+      "loss": 0.3558,
+      "step": 4394
+    },
+    {
+      "epoch": 0.30516594917372586,
+      "grad_norm": 4.333491636663673,
+      "learning_rate": 8.143003090178861e-06,
+      "loss": 0.6461,
+      "step": 4395
+    },
+    {
+      "epoch": 0.305235383974448,
+      "grad_norm": 4.31820612823554,
+      "learning_rate": 8.142128465128539e-06,
+      "loss": 0.5984,
+      "step": 4396
+    },
+    {
+      "epoch": 0.30530481877517013,
+      "grad_norm": 3.9168145078000594,
+      "learning_rate": 8.141253681152913e-06,
+      "loss": 0.3759,
+      "step": 4397
+    },
+    {
+      "epoch": 0.30537425357589226,
+      "grad_norm": 3.470542531161544,
+      "learning_rate": 8.140378738296233e-06,
+      "loss": 0.4397,
+      "step": 4398
+    },
+    {
+      "epoch": 0.30544368837661434,
+      "grad_norm": 5.082617596426652,
+      "learning_rate": 8.139503636602748e-06,
+      "loss": 0.4643,
+      "step": 4399
+    },
+    {
+      "epoch": 0.3055131231773365,
+      "grad_norm": 2.75953376479929,
+      "learning_rate": 8.138628376116722e-06,
+      "loss": 0.2343,
+      "step": 4400
+    },
+    {
+      "epoch": 0.3055825579780586,
+      "grad_norm": 4.262094642901102,
+      "learning_rate": 8.137752956882425e-06,
+      "loss": 0.6372,
+      "step": 4401
+    },
+    {
+      "epoch": 0.30565199277878075,
+      "grad_norm": 3.854959305590892,
+      "learning_rate": 8.136877378944133e-06,
+      "loss": 0.534,
+      "step": 4402
+    },
+    {
+      "epoch": 0.3057214275795028,
+      "grad_norm": 3.8173774445742996,
+      "learning_rate": 8.136001642346134e-06,
+      "loss": 0.4997,
+      "step": 4403
+    },
+    {
+      "epoch": 0.30579086238022496,
+      "grad_norm": 4.256830494731312,
+      "learning_rate": 8.13512574713272e-06,
+      "loss": 0.4724,
+      "step": 4404
+    },
+    {
+      "epoch": 0.3058602971809471,
+      "grad_norm": 4.219029971034703,
+      "learning_rate": 8.134249693348194e-06,
+      "loss": 0.4348,
+      "step": 4405
+    },
+    {
+      "epoch": 0.30592973198166923,
+      "grad_norm": 3.03421053916804,
+      "learning_rate": 8.133373481036863e-06,
+      "loss": 0.3317,
+      "step": 4406
+    },
+    {
+      "epoch": 0.3059991667823913,
+      "grad_norm": 2.4838249242367634,
+      "learning_rate": 8.13249711024305e-06,
+      "loss": 0.2369,
+      "step": 4407
+    },
+    {
+      "epoch": 0.30606860158311344,
+      "grad_norm": 4.476396723391165,
+      "learning_rate": 8.131620581011078e-06,
+      "loss": 0.7447,
+      "step": 4408
+    },
+    {
+      "epoch": 0.3061380363838356,
+      "grad_norm": 4.272066139148233,
+      "learning_rate": 8.130743893385278e-06,
+      "loss": 0.5377,
+      "step": 4409
+    },
+    {
+      "epoch": 0.3062074711845577,
+      "grad_norm": 3.582838350403838,
+      "learning_rate": 8.129867047409996e-06,
+      "loss": 0.4204,
+      "step": 4410
+    },
+    {
+      "epoch": 0.30627690598527985,
+      "grad_norm": 4.530648710603671,
+      "learning_rate": 8.128990043129578e-06,
+      "loss": 0.3536,
+      "step": 4411
+    },
+    {
+      "epoch": 0.3063463407860019,
+      "grad_norm": 4.3601479853972105,
+      "learning_rate": 8.128112880588388e-06,
+      "loss": 0.6034,
+      "step": 4412
+    },
+    {
+      "epoch": 0.30641577558672406,
+      "grad_norm": 4.241356024877396,
+      "learning_rate": 8.127235559830787e-06,
+      "loss": 0.5378,
+      "step": 4413
+    },
+    {
+      "epoch": 0.3064852103874462,
+      "grad_norm": 4.41454313466964,
+      "learning_rate": 8.12635808090115e-06,
+      "loss": 0.5095,
+      "step": 4414
+    },
+    {
+      "epoch": 0.30655464518816833,
+      "grad_norm": 3.1461283435761143,
+      "learning_rate": 8.125480443843856e-06,
+      "loss": 0.2991,
+      "step": 4415
+    },
+    {
+      "epoch": 0.3066240799888904,
+      "grad_norm": 3.619036081279081,
+      "learning_rate": 8.124602648703302e-06,
+      "loss": 0.4409,
+      "step": 4416
+    },
+    {
+      "epoch": 0.30669351478961254,
+      "grad_norm": 3.8710041152815564,
+      "learning_rate": 8.12372469552388e-06,
+      "loss": 0.3634,
+      "step": 4417
+    },
+    {
+      "epoch": 0.3067629495903347,
+      "grad_norm": 10.470330163352108,
+      "learning_rate": 8.122846584349997e-06,
+      "loss": 0.4854,
+      "step": 4418
+    },
+    {
+      "epoch": 0.3068323843910568,
+      "grad_norm": 5.409820845049516,
+      "learning_rate": 8.12196831522607e-06,
+      "loss": 0.566,
+      "step": 4419
+    },
+    {
+      "epoch": 0.30690181919177895,
+      "grad_norm": 3.7227677108899293,
+      "learning_rate": 8.121089888196517e-06,
+      "loss": 0.5143,
+      "step": 4420
+    },
+    {
+      "epoch": 0.306971253992501,
+      "grad_norm": 2.316100758041508,
+      "learning_rate": 8.120211303305767e-06,
+      "loss": 0.2651,
+      "step": 4421
+    },
+    {
+      "epoch": 0.30704068879322316,
+      "grad_norm": 3.0144149025319704,
+      "learning_rate": 8.119332560598263e-06,
+      "loss": 0.3067,
+      "step": 4422
+    },
+    {
+      "epoch": 0.3071101235939453,
+      "grad_norm": 4.561779940617743,
+      "learning_rate": 8.118453660118446e-06,
+      "loss": 0.5702,
+      "step": 4423
+    },
+    {
+      "epoch": 0.30717955839466743,
+      "grad_norm": 3.4037221623850034,
+      "learning_rate": 8.117574601910773e-06,
+      "loss": 0.3958,
+      "step": 4424
+    },
+    {
+      "epoch": 0.3072489931953895,
+      "grad_norm": 5.497384671760635,
+      "learning_rate": 8.116695386019703e-06,
+      "loss": 0.4654,
+      "step": 4425
+    },
+    {
+      "epoch": 0.30731842799611164,
+      "grad_norm": 2.5014560796022423,
+      "learning_rate": 8.115816012489709e-06,
+      "loss": 0.1476,
+      "step": 4426
+    },
+    {
+      "epoch": 0.3073878627968338,
+      "grad_norm": 3.628303282130237,
+      "learning_rate": 8.114936481365266e-06,
+      "loss": 0.3464,
+      "step": 4427
+    },
+    {
+      "epoch": 0.3074572975975559,
+      "grad_norm": 4.20300144097651,
+      "learning_rate": 8.114056792690861e-06,
+      "loss": 0.5402,
+      "step": 4428
+    },
+    {
+      "epoch": 0.307526732398278,
+      "grad_norm": 4.154491168365554,
+      "learning_rate": 8.113176946510985e-06,
+      "loss": 0.6373,
+      "step": 4429
+    },
+    {
+      "epoch": 0.3075961671990001,
+      "grad_norm": 4.590013238096061,
+      "learning_rate": 8.112296942870145e-06,
+      "loss": 0.4922,
+      "step": 4430
+    },
+    {
+      "epoch": 0.30766560199972226,
+      "grad_norm": 2.839191476875827,
+      "learning_rate": 8.111416781812844e-06,
+      "loss": 0.2131,
+      "step": 4431
+    },
+    {
+      "epoch": 0.3077350368004444,
+      "grad_norm": 3.916844429240888,
+      "learning_rate": 8.110536463383607e-06,
+      "loss": 0.4405,
+      "step": 4432
+    },
+    {
+      "epoch": 0.30780447160116653,
+      "grad_norm": 3.0358787420557767,
+      "learning_rate": 8.109655987626952e-06,
+      "loss": 0.213,
+      "step": 4433
+    },
+    {
+      "epoch": 0.3078739064018886,
+      "grad_norm": 3.2277461176625004,
+      "learning_rate": 8.108775354587419e-06,
+      "loss": 0.2478,
+      "step": 4434
+    },
+    {
+      "epoch": 0.30794334120261074,
+      "grad_norm": 3.8289680055955713,
+      "learning_rate": 8.107894564309544e-06,
+      "loss": 0.8028,
+      "step": 4435
+    },
+    {
+      "epoch": 0.3080127760033329,
+      "grad_norm": 3.1632155161137456,
+      "learning_rate": 8.107013616837879e-06,
+      "loss": 0.4484,
+      "step": 4436
+    },
+    {
+      "epoch": 0.308082210804055,
+      "grad_norm": 5.217563547397045,
+      "learning_rate": 8.106132512216982e-06,
+      "loss": 0.6494,
+      "step": 4437
+    },
+    {
+      "epoch": 0.3081516456047771,
+      "grad_norm": 3.618649134408479,
+      "learning_rate": 8.105251250491416e-06,
+      "loss": 0.4117,
+      "step": 4438
+    },
+    {
+      "epoch": 0.3082210804054992,
+      "grad_norm": 4.378405825404889,
+      "learning_rate": 8.104369831705757e-06,
+      "loss": 0.649,
+      "step": 4439
+    },
+    {
+      "epoch": 0.30829051520622136,
+      "grad_norm": 2.656974482680879,
+      "learning_rate": 8.103488255904584e-06,
+      "loss": 0.1773,
+      "step": 4440
+    },
+    {
+      "epoch": 0.3083599500069435,
+      "grad_norm": 3.2092800781291064,
+      "learning_rate": 8.102606523132487e-06,
+      "loss": 0.3493,
+      "step": 4441
+    },
+    {
+      "epoch": 0.30842938480766563,
+      "grad_norm": 3.4566222925875874,
+      "learning_rate": 8.101724633434065e-06,
+      "loss": 0.6609,
+      "step": 4442
+    },
+    {
+      "epoch": 0.3084988196083877,
+      "grad_norm": 2.7678484656225226,
+      "learning_rate": 8.100842586853918e-06,
+      "loss": 0.3951,
+      "step": 4443
+    },
+    {
+      "epoch": 0.30856825440910984,
+      "grad_norm": 3.017962517665393,
+      "learning_rate": 8.099960383436663e-06,
+      "loss": 0.4165,
+      "step": 4444
+    },
+    {
+      "epoch": 0.308637689209832,
+      "grad_norm": 3.7899308794899644,
+      "learning_rate": 8.099078023226919e-06,
+      "loss": 0.416,
+      "step": 4445
+    },
+    {
+      "epoch": 0.3087071240105541,
+      "grad_norm": 4.036789949575463,
+      "learning_rate": 8.098195506269315e-06,
+      "loss": 0.6015,
+      "step": 4446
+    },
+    {
+      "epoch": 0.3087765588112762,
+      "grad_norm": 3.813199671432458,
+      "learning_rate": 8.097312832608488e-06,
+      "loss": 0.3948,
+      "step": 4447
+    },
+    {
+      "epoch": 0.3088459936119983,
+      "grad_norm": 2.7095700296841194,
+      "learning_rate": 8.096430002289086e-06,
+      "loss": 0.1121,
+      "step": 4448
+    },
+    {
+      "epoch": 0.30891542841272046,
+      "grad_norm": 4.094781410560263,
+      "learning_rate": 8.095547015355756e-06,
+      "loss": 0.5487,
+      "step": 4449
+    },
+    {
+      "epoch": 0.3089848632134426,
+      "grad_norm": 3.3463116924288405,
+      "learning_rate": 8.094663871853162e-06,
+      "loss": 0.5388,
+      "step": 4450
+    },
+    {
+      "epoch": 0.3090542980141647,
+      "grad_norm": 4.412867885211758,
+      "learning_rate": 8.093780571825968e-06,
+      "loss": 0.675,
+      "step": 4451
+    },
+    {
+      "epoch": 0.3091237328148868,
+      "grad_norm": 3.52994549400418,
+      "learning_rate": 8.092897115318857e-06,
+      "loss": 0.4467,
+      "step": 4452
+    },
+    {
+      "epoch": 0.30919316761560894,
+      "grad_norm": 4.995930469254473,
+      "learning_rate": 8.092013502376509e-06,
+      "loss": 0.7392,
+      "step": 4453
+    },
+    {
+      "epoch": 0.3092626024163311,
+      "grad_norm": 5.2865955633498345,
+      "learning_rate": 8.091129733043614e-06,
+      "loss": 0.7246,
+      "step": 4454
+    },
+    {
+      "epoch": 0.3093320372170532,
+      "grad_norm": 4.551841109687148,
+      "learning_rate": 8.090245807364879e-06,
+      "loss": 0.4455,
+      "step": 4455
+    },
+    {
+      "epoch": 0.3094014720177753,
+      "grad_norm": 4.032816901609972,
+      "learning_rate": 8.089361725385005e-06,
+      "loss": 0.649,
+      "step": 4456
+    },
+    {
+      "epoch": 0.3094709068184974,
+      "grad_norm": 3.6388332366288583,
+      "learning_rate": 8.088477487148712e-06,
+      "loss": 0.3065,
+      "step": 4457
+    },
+    {
+      "epoch": 0.30954034161921956,
+      "grad_norm": 4.44412661526409,
+      "learning_rate": 8.087593092700723e-06,
+      "loss": 0.6061,
+      "step": 4458
+    },
+    {
+      "epoch": 0.3096097764199417,
+      "grad_norm": 2.271353875797235,
+      "learning_rate": 8.086708542085769e-06,
+      "loss": 0.1833,
+      "step": 4459
+    },
+    {
+      "epoch": 0.3096792112206638,
+      "grad_norm": 5.52872821102468,
+      "learning_rate": 8.085823835348588e-06,
+      "loss": 0.6503,
+      "step": 4460
+    },
+    {
+      "epoch": 0.3097486460213859,
+      "grad_norm": 4.005711313915286,
+      "learning_rate": 8.084938972533931e-06,
+      "loss": 0.5572,
+      "step": 4461
+    },
+    {
+      "epoch": 0.30981808082210804,
+      "grad_norm": 5.070902280390715,
+      "learning_rate": 8.084053953686551e-06,
+      "loss": 0.5378,
+      "step": 4462
+    },
+    {
+      "epoch": 0.3098875156228302,
+      "grad_norm": 3.8672260068133943,
+      "learning_rate": 8.083168778851213e-06,
+      "loss": 0.5637,
+      "step": 4463
+    },
+    {
+      "epoch": 0.30995695042355226,
+      "grad_norm": 5.090046857239955,
+      "learning_rate": 8.082283448072685e-06,
+      "loss": 0.5059,
+      "step": 4464
+    },
+    {
+      "epoch": 0.3100263852242744,
+      "grad_norm": 2.24476687530852,
+      "learning_rate": 8.08139796139575e-06,
+      "loss": 0.2829,
+      "step": 4465
+    },
+    {
+      "epoch": 0.3100958200249965,
+      "grad_norm": 3.288728483761353,
+      "learning_rate": 8.080512318865192e-06,
+      "loss": 0.6081,
+      "step": 4466
+    },
+    {
+      "epoch": 0.31016525482571866,
+      "grad_norm": 3.651776211041886,
+      "learning_rate": 8.079626520525808e-06,
+      "loss": 0.4627,
+      "step": 4467
+    },
+    {
+      "epoch": 0.3102346896264408,
+      "grad_norm": 3.444922896922857,
+      "learning_rate": 8.078740566422398e-06,
+      "loss": 0.5138,
+      "step": 4468
+    },
+    {
+      "epoch": 0.3103041244271629,
+      "grad_norm": 3.9692262858277667,
+      "learning_rate": 8.077854456599775e-06,
+      "loss": 0.3492,
+      "step": 4469
+    },
+    {
+      "epoch": 0.310373559227885,
+      "grad_norm": 3.4782587058965344,
+      "learning_rate": 8.076968191102754e-06,
+      "loss": 0.4013,
+      "step": 4470
+    },
+    {
+      "epoch": 0.31044299402860714,
+      "grad_norm": 3.854553662645897,
+      "learning_rate": 8.076081769976166e-06,
+      "loss": 0.4763,
+      "step": 4471
+    },
+    {
+      "epoch": 0.3105124288293293,
+      "grad_norm": 4.063362219286672,
+      "learning_rate": 8.075195193264843e-06,
+      "loss": 0.4785,
+      "step": 4472
+    },
+    {
+      "epoch": 0.31058186363005136,
+      "grad_norm": 4.5851089788534605,
+      "learning_rate": 8.074308461013626e-06,
+      "loss": 0.4994,
+      "step": 4473
+    },
+    {
+      "epoch": 0.3106512984307735,
+      "grad_norm": 4.376715667161838,
+      "learning_rate": 8.073421573267366e-06,
+      "loss": 0.57,
+      "step": 4474
+    },
+    {
+      "epoch": 0.3107207332314956,
+      "grad_norm": 3.4490719851174343,
+      "learning_rate": 8.072534530070922e-06,
+      "loss": 0.3916,
+      "step": 4475
+    },
+    {
+      "epoch": 0.31079016803221776,
+      "grad_norm": 4.509297679238457,
+      "learning_rate": 8.071647331469157e-06,
+      "loss": 0.3915,
+      "step": 4476
+    },
+    {
+      "epoch": 0.3108596028329399,
+      "grad_norm": 2.8812626526753298,
+      "learning_rate": 8.070759977506945e-06,
+      "loss": 0.2958,
+      "step": 4477
+    },
+    {
+      "epoch": 0.310929037633662,
+      "grad_norm": 4.039628430094393,
+      "learning_rate": 8.06987246822917e-06,
+      "loss": 0.3646,
+      "step": 4478
+    },
+    {
+      "epoch": 0.3109984724343841,
+      "grad_norm": 2.977710598727337,
+      "learning_rate": 8.068984803680717e-06,
+      "loss": 0.193,
+      "step": 4479
+    },
+    {
+      "epoch": 0.31106790723510624,
+      "grad_norm": 3.672612287269218,
+      "learning_rate": 8.068096983906488e-06,
+      "loss": 0.5416,
+      "step": 4480
+    },
+    {
+      "epoch": 0.3111373420358284,
+      "grad_norm": 3.432180510890564,
+      "learning_rate": 8.067209008951382e-06,
+      "loss": 0.5045,
+      "step": 4481
+    },
+    {
+      "epoch": 0.31120677683655046,
+      "grad_norm": 3.146804182232353,
+      "learning_rate": 8.06632087886032e-06,
+      "loss": 0.2903,
+      "step": 4482
+    },
+    {
+      "epoch": 0.3112762116372726,
+      "grad_norm": 3.20689131971911,
+      "learning_rate": 8.065432593678213e-06,
+      "loss": 0.4862,
+      "step": 4483
+    },
+    {
+      "epoch": 0.3113456464379947,
+      "grad_norm": 4.2783633822559235,
+      "learning_rate": 8.064544153449996e-06,
+      "loss": 0.5001,
+      "step": 4484
+    },
+    {
+      "epoch": 0.31141508123871686,
+      "grad_norm": 3.6604235538562166,
+      "learning_rate": 8.063655558220605e-06,
+      "loss": 0.4118,
+      "step": 4485
+    },
+    {
+      "epoch": 0.31148451603943894,
+      "grad_norm": 3.797398685293877,
+      "learning_rate": 8.06276680803498e-06,
+      "loss": 0.4201,
+      "step": 4486
+    },
+    {
+      "epoch": 0.3115539508401611,
+      "grad_norm": 4.250788222197798,
+      "learning_rate": 8.061877902938077e-06,
+      "loss": 0.5273,
+      "step": 4487
+    },
+    {
+      "epoch": 0.3116233856408832,
+      "grad_norm": 3.3677202341977908,
+      "learning_rate": 8.060988842974853e-06,
+      "loss": 0.4621,
+      "step": 4488
+    },
+    {
+      "epoch": 0.31169282044160534,
+      "grad_norm": 3.6594708925218193,
+      "learning_rate": 8.06009962819028e-06,
+      "loss": 0.4634,
+      "step": 4489
+    },
+    {
+      "epoch": 0.3117622552423275,
+      "grad_norm": 3.384579206101019,
+      "learning_rate": 8.059210258629328e-06,
+      "loss": 0.5538,
+      "step": 4490
+    },
+    {
+      "epoch": 0.31183169004304956,
+      "grad_norm": 4.3945471726361,
+      "learning_rate": 8.058320734336984e-06,
+      "loss": 0.7264,
+      "step": 4491
+    },
+    {
+      "epoch": 0.3119011248437717,
+      "grad_norm": 4.594607014140406,
+      "learning_rate": 8.057431055358238e-06,
+      "loss": 0.6429,
+      "step": 4492
+    },
+    {
+      "epoch": 0.31197055964449383,
+      "grad_norm": 3.4497983979041433,
+      "learning_rate": 8.056541221738088e-06,
+      "loss": 0.4277,
+      "step": 4493
+    },
+    {
+      "epoch": 0.31203999444521596,
+      "grad_norm": 4.362289135572665,
+      "learning_rate": 8.055651233521543e-06,
+      "loss": 0.5336,
+      "step": 4494
+    },
+    {
+      "epoch": 0.31210942924593804,
+      "grad_norm": 4.509164446041097,
+      "learning_rate": 8.054761090753615e-06,
+      "loss": 0.6592,
+      "step": 4495
+    },
+    {
+      "epoch": 0.3121788640466602,
+      "grad_norm": 3.1237688586523187,
+      "learning_rate": 8.053870793479329e-06,
+      "loss": 0.3181,
+      "step": 4496
+    },
+    {
+      "epoch": 0.3122482988473823,
+      "grad_norm": 3.6116413598140897,
+      "learning_rate": 8.052980341743712e-06,
+      "loss": 0.482,
+      "step": 4497
+    },
+    {
+      "epoch": 0.31231773364810445,
+      "grad_norm": 4.313299276718832,
+      "learning_rate": 8.052089735591806e-06,
+      "loss": 0.4442,
+      "step": 4498
+    },
+    {
+      "epoch": 0.3123871684488265,
+      "grad_norm": 10.795401691382107,
+      "learning_rate": 8.051198975068655e-06,
+      "loss": 0.5234,
+      "step": 4499
+    },
+    {
+      "epoch": 0.31245660324954866,
+      "grad_norm": 3.8421364182217945,
+      "learning_rate": 8.050308060219311e-06,
+      "loss": 0.7851,
+      "step": 4500
+    },
+    {
+      "epoch": 0.3125260380502708,
+      "grad_norm": 4.173483350252587,
+      "learning_rate": 8.049416991088838e-06,
+      "loss": 0.5298,
+      "step": 4501
+    },
+    {
+      "epoch": 0.31259547285099293,
+      "grad_norm": 3.9334632226309623,
+      "learning_rate": 8.048525767722304e-06,
+      "loss": 0.659,
+      "step": 4502
+    },
+    {
+      "epoch": 0.31266490765171506,
+      "grad_norm": 3.3854651646308405,
+      "learning_rate": 8.047634390164787e-06,
+      "loss": 0.4844,
+      "step": 4503
+    },
+    {
+      "epoch": 0.31273434245243714,
+      "grad_norm": 3.168982435519991,
+      "learning_rate": 8.04674285846137e-06,
+      "loss": 0.358,
+      "step": 4504
+    },
+    {
+      "epoch": 0.3128037772531593,
+      "grad_norm": 4.045872058279486,
+      "learning_rate": 8.045851172657146e-06,
+      "loss": 0.7503,
+      "step": 4505
+    },
+    {
+      "epoch": 0.3128732120538814,
+      "grad_norm": 4.824018499992043,
+      "learning_rate": 8.044959332797217e-06,
+      "loss": 0.5927,
+      "step": 4506
+    },
+    {
+      "epoch": 0.31294264685460355,
+      "grad_norm": 3.4837328009379145,
+      "learning_rate": 8.044067338926693e-06,
+      "loss": 0.5333,
+      "step": 4507
+    },
+    {
+      "epoch": 0.3130120816553256,
+      "grad_norm": 4.12526070940582,
+      "learning_rate": 8.043175191090687e-06,
+      "loss": 0.5585,
+      "step": 4508
+    },
+    {
+      "epoch": 0.31308151645604776,
+      "grad_norm": 3.922028171082318,
+      "learning_rate": 8.042282889334322e-06,
+      "loss": 0.6927,
+      "step": 4509
+    },
+    {
+      "epoch": 0.3131509512567699,
+      "grad_norm": 3.228177792195968,
+      "learning_rate": 8.041390433702729e-06,
+      "loss": 0.3174,
+      "step": 4510
+    },
+    {
+      "epoch": 0.31322038605749203,
+      "grad_norm": 3.421018627605359,
+      "learning_rate": 8.040497824241051e-06,
+      "loss": 0.3538,
+      "step": 4511
+    },
+    {
+      "epoch": 0.31328982085821416,
+      "grad_norm": 3.697949005925923,
+      "learning_rate": 8.039605060994435e-06,
+      "loss": 0.4627,
+      "step": 4512
+    },
+    {
+      "epoch": 0.31335925565893624,
+      "grad_norm": 2.7125115776848157,
+      "learning_rate": 8.038712144008035e-06,
+      "loss": 0.2444,
+      "step": 4513
+    },
+    {
+      "epoch": 0.3134286904596584,
+      "grad_norm": 5.499710724462778,
+      "learning_rate": 8.037819073327012e-06,
+      "loss": 0.4499,
+      "step": 4514
+    },
+    {
+      "epoch": 0.3134981252603805,
+      "grad_norm": 4.5252127951966825,
+      "learning_rate": 8.036925848996537e-06,
+      "loss": 0.463,
+      "step": 4515
+    },
+    {
+      "epoch": 0.31356756006110265,
+      "grad_norm": 3.6768390267608715,
+      "learning_rate": 8.03603247106179e-06,
+      "loss": 0.437,
+      "step": 4516
+    },
+    {
+      "epoch": 0.3136369948618247,
+      "grad_norm": 3.734811904427,
+      "learning_rate": 8.035138939567955e-06,
+      "loss": 0.5122,
+      "step": 4517
+    },
+    {
+      "epoch": 0.31370642966254686,
+      "grad_norm": 3.010988053258721,
+      "learning_rate": 8.03424525456023e-06,
+      "loss": 0.3149,
+      "step": 4518
+    },
+    {
+      "epoch": 0.313775864463269,
+      "grad_norm": 3.9902219394831295,
+      "learning_rate": 8.03335141608381e-06,
+      "loss": 0.5176,
+      "step": 4519
+    },
+    {
+      "epoch": 0.31384529926399113,
+      "grad_norm": 3.501876809299136,
+      "learning_rate": 8.032457424183909e-06,
+      "loss": 0.3673,
+      "step": 4520
+    },
+    {
+      "epoch": 0.3139147340647132,
+      "grad_norm": 3.034950270406639,
+      "learning_rate": 8.031563278905744e-06,
+      "loss": 0.3731,
+      "step": 4521
+    },
+    {
+      "epoch": 0.31398416886543534,
+      "grad_norm": 3.785104588218375,
+      "learning_rate": 8.030668980294539e-06,
+      "loss": 0.3874,
+      "step": 4522
+    },
+    {
+      "epoch": 0.3140536036661575,
+      "grad_norm": 4.376463199172921,
+      "learning_rate": 8.029774528395523e-06,
+      "loss": 0.7204,
+      "step": 4523
+    },
+    {
+      "epoch": 0.3141230384668796,
+      "grad_norm": 5.318149157963461,
+      "learning_rate": 8.028879923253943e-06,
+      "loss": 0.6724,
+      "step": 4524
+    },
+    {
+      "epoch": 0.31419247326760175,
+      "grad_norm": 2.7053782739354886,
+      "learning_rate": 8.027985164915044e-06,
+      "loss": 0.2198,
+      "step": 4525
+    },
+    {
+      "epoch": 0.3142619080683238,
+      "grad_norm": 3.889788717869216,
+      "learning_rate": 8.02709025342408e-06,
+      "loss": 0.4828,
+      "step": 4526
+    },
+    {
+      "epoch": 0.31433134286904596,
+      "grad_norm": 3.603170873120601,
+      "learning_rate": 8.026195188826318e-06,
+      "loss": 0.4517,
+      "step": 4527
+    },
+    {
+      "epoch": 0.3144007776697681,
+      "grad_norm": 4.352725565702803,
+      "learning_rate": 8.025299971167026e-06,
+      "loss": 0.5557,
+      "step": 4528
+    },
+    {
+      "epoch": 0.31447021247049023,
+      "grad_norm": 6.346075373124257,
+      "learning_rate": 8.024404600491485e-06,
+      "loss": 1.0009,
+      "step": 4529
+    },
+    {
+      "epoch": 0.3145396472712123,
+      "grad_norm": 2.7681282881066087,
+      "learning_rate": 8.023509076844984e-06,
+      "loss": 0.2242,
+      "step": 4530
+    },
+    {
+      "epoch": 0.31460908207193444,
+      "grad_norm": 4.342935971872531,
+      "learning_rate": 8.022613400272813e-06,
+      "loss": 0.5686,
+      "step": 4531
+    },
+    {
+      "epoch": 0.3146785168726566,
+      "grad_norm": 3.7165907157696405,
+      "learning_rate": 8.021717570820278e-06,
+      "loss": 0.5998,
+      "step": 4532
+    },
+    {
+      "epoch": 0.3147479516733787,
+      "grad_norm": 3.5314554016819524,
+      "learning_rate": 8.020821588532685e-06,
+      "loss": 0.3413,
+      "step": 4533
+    },
+    {
+      "epoch": 0.31481738647410085,
+      "grad_norm": 4.270056155307777,
+      "learning_rate": 8.019925453455357e-06,
+      "loss": 0.4416,
+      "step": 4534
+    },
+    {
+      "epoch": 0.3148868212748229,
+      "grad_norm": 3.282529859459503,
+      "learning_rate": 8.019029165633617e-06,
+      "loss": 0.3327,
+      "step": 4535
+    },
+    {
+      "epoch": 0.31495625607554506,
+      "grad_norm": 4.855975144877896,
+      "learning_rate": 8.018132725112796e-06,
+      "loss": 0.5824,
+      "step": 4536
+    },
+    {
+      "epoch": 0.3150256908762672,
+      "grad_norm": 3.218366639815275,
+      "learning_rate": 8.017236131938238e-06,
+      "loss": 0.3953,
+      "step": 4537
+    },
+    {
+      "epoch": 0.31509512567698933,
+      "grad_norm": 4.418789432625289,
+      "learning_rate": 8.01633938615529e-06,
+      "loss": 0.7477,
+      "step": 4538
+    },
+    {
+      "epoch": 0.3151645604777114,
+      "grad_norm": 3.558490107985622,
+      "learning_rate": 8.015442487809311e-06,
+      "loss": 0.3256,
+      "step": 4539
+    },
+    {
+      "epoch": 0.31523399527843354,
+      "grad_norm": 3.3864253570776777,
+      "learning_rate": 8.014545436945662e-06,
+      "loss": 0.4903,
+      "step": 4540
+    },
+    {
+      "epoch": 0.3153034300791557,
+      "grad_norm": 4.368700704511398,
+      "learning_rate": 8.013648233609715e-06,
+      "loss": 0.6201,
+      "step": 4541
+    },
+    {
+      "epoch": 0.3153728648798778,
+      "grad_norm": 3.711882282295784,
+      "learning_rate": 8.012750877846853e-06,
+      "loss": 0.5054,
+      "step": 4542
+    },
+    {
+      "epoch": 0.3154422996805999,
+      "grad_norm": 3.6020104268932513,
+      "learning_rate": 8.011853369702458e-06,
+      "loss": 0.4528,
+      "step": 4543
+    },
+    {
+      "epoch": 0.315511734481322,
+      "grad_norm": 3.5006019312694976,
+      "learning_rate": 8.010955709221929e-06,
+      "loss": 0.4258,
+      "step": 4544
+    },
+    {
+      "epoch": 0.31558116928204416,
+      "grad_norm": 4.429251798225652,
+      "learning_rate": 8.010057896450668e-06,
+      "loss": 0.6561,
+      "step": 4545
+    },
+    {
+      "epoch": 0.3156506040827663,
+      "grad_norm": 4.20194708811198,
+      "learning_rate": 8.009159931434082e-06,
+      "loss": 0.7396,
+      "step": 4546
+    },
+    {
+      "epoch": 0.31572003888348843,
+      "grad_norm": 4.685831937197169,
+      "learning_rate": 8.008261814217596e-06,
+      "loss": 0.6368,
+      "step": 4547
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 3.738448314997094,
+      "learning_rate": 8.007363544846628e-06,
+      "loss": 0.3896,
+      "step": 4548
+    },
+    {
+      "epoch": 0.31585890848493264,
+      "grad_norm": 4.906875865520298,
+      "learning_rate": 8.006465123366616e-06,
+      "loss": 0.6939,
+      "step": 4549
+    },
+    {
+      "epoch": 0.3159283432856548,
+      "grad_norm": 4.105792076079457,
+      "learning_rate": 8.005566549823e-06,
+      "loss": 0.4661,
+      "step": 4550
+    },
+    {
+      "epoch": 0.3159977780863769,
+      "grad_norm": 3.73387453417385,
+      "learning_rate": 8.00466782426123e-06,
+      "loss": 0.5312,
+      "step": 4551
+    },
+    {
+      "epoch": 0.316067212887099,
+      "grad_norm": 3.8873780417617807,
+      "learning_rate": 8.00376894672676e-06,
+      "loss": 0.5587,
+      "step": 4552
+    },
+    {
+      "epoch": 0.3161366476878211,
+      "grad_norm": 2.1738424812859884,
+      "learning_rate": 8.002869917265056e-06,
+      "loss": 0.2238,
+      "step": 4553
+    },
+    {
+      "epoch": 0.31620608248854326,
+      "grad_norm": 3.279184022679103,
+      "learning_rate": 8.001970735921591e-06,
+      "loss": 0.4829,
+      "step": 4554
+    },
+    {
+      "epoch": 0.3162755172892654,
+      "grad_norm": 4.519534670663577,
+      "learning_rate": 8.001071402741843e-06,
+      "loss": 0.4506,
+      "step": 4555
+    },
+    {
+      "epoch": 0.3163449520899875,
+      "grad_norm": 5.004565282506696,
+      "learning_rate": 8.000171917771297e-06,
+      "loss": 0.544,
+      "step": 4556
+    },
+    {
+      "epoch": 0.3164143868907096,
+      "grad_norm": 4.211814064690637,
+      "learning_rate": 7.999272281055452e-06,
+      "loss": 0.5274,
+      "step": 4557
+    },
+    {
+      "epoch": 0.31648382169143174,
+      "grad_norm": 3.85219753702993,
+      "learning_rate": 7.99837249263981e-06,
+      "loss": 0.4346,
+      "step": 4558
+    },
+    {
+      "epoch": 0.3165532564921539,
+      "grad_norm": 3.7765355623234136,
+      "learning_rate": 7.997472552569877e-06,
+      "loss": 0.4536,
+      "step": 4559
+    },
+    {
+      "epoch": 0.316622691292876,
+      "grad_norm": 2.97148807572818,
+      "learning_rate": 7.996572460891176e-06,
+      "loss": 0.3746,
+      "step": 4560
+    },
+    {
+      "epoch": 0.3166921260935981,
+      "grad_norm": 3.696867565621778,
+      "learning_rate": 7.995672217649231e-06,
+      "loss": 0.6967,
+      "step": 4561
+    },
+    {
+      "epoch": 0.3167615608943202,
+      "grad_norm": 4.64768828842926,
+      "learning_rate": 7.994771822889575e-06,
+      "loss": 0.4421,
+      "step": 4562
+    },
+    {
+      "epoch": 0.31683099569504236,
+      "grad_norm": 3.646561863846733,
+      "learning_rate": 7.99387127665775e-06,
+      "loss": 0.6111,
+      "step": 4563
+    },
+    {
+      "epoch": 0.3169004304957645,
+      "grad_norm": 4.225993571261041,
+      "learning_rate": 7.992970578999303e-06,
+      "loss": 0.5535,
+      "step": 4564
+    },
+    {
+      "epoch": 0.3169698652964866,
+      "grad_norm": 4.045295345049359,
+      "learning_rate": 7.99206972995979e-06,
+      "loss": 0.5822,
+      "step": 4565
+    },
+    {
+      "epoch": 0.3170393000972087,
+      "grad_norm": 3.9962793551379185,
+      "learning_rate": 7.991168729584775e-06,
+      "loss": 0.4423,
+      "step": 4566
+    },
+    {
+      "epoch": 0.31710873489793084,
+      "grad_norm": 4.475517415464021,
+      "learning_rate": 7.990267577919833e-06,
+      "loss": 0.4236,
+      "step": 4567
+    },
+    {
+      "epoch": 0.317178169698653,
+      "grad_norm": 3.67563569380176,
+      "learning_rate": 7.98936627501054e-06,
+      "loss": 0.3612,
+      "step": 4568
+    },
+    {
+      "epoch": 0.3172476044993751,
+      "grad_norm": 3.592914863446774,
+      "learning_rate": 7.988464820902482e-06,
+      "loss": 0.4735,
+      "step": 4569
+    },
+    {
+      "epoch": 0.3173170393000972,
+      "grad_norm": 2.59805260898133,
+      "learning_rate": 7.987563215641256e-06,
+      "loss": 0.204,
+      "step": 4570
+    },
+    {
+      "epoch": 0.3173864741008193,
+      "grad_norm": 4.154543370907658,
+      "learning_rate": 7.986661459272463e-06,
+      "loss": 0.5733,
+      "step": 4571
+    },
+    {
+      "epoch": 0.31745590890154146,
+      "grad_norm": 3.5861481067565735,
+      "learning_rate": 7.985759551841711e-06,
+      "loss": 0.4414,
+      "step": 4572
+    },
+    {
+      "epoch": 0.3175253437022636,
+      "grad_norm": 3.3827084904939437,
+      "learning_rate": 7.984857493394623e-06,
+      "loss": 0.4369,
+      "step": 4573
+    },
+    {
+      "epoch": 0.3175947785029857,
+      "grad_norm": 3.2560966029796687,
+      "learning_rate": 7.98395528397682e-06,
+      "loss": 0.4787,
+      "step": 4574
+    },
+    {
+      "epoch": 0.3176642133037078,
+      "grad_norm": 3.991344325912637,
+      "learning_rate": 7.983052923633933e-06,
+      "loss": 0.5226,
+      "step": 4575
+    },
+    {
+      "epoch": 0.31773364810442994,
+      "grad_norm": 4.704753002267821,
+      "learning_rate": 7.982150412411607e-06,
+      "loss": 0.7063,
+      "step": 4576
+    },
+    {
+      "epoch": 0.3178030829051521,
+      "grad_norm": 4.499960285252185,
+      "learning_rate": 7.981247750355486e-06,
+      "loss": 0.5527,
+      "step": 4577
+    },
+    {
+      "epoch": 0.31787251770587416,
+      "grad_norm": 2.684181899305464,
+      "learning_rate": 7.980344937511229e-06,
+      "loss": 0.3847,
+      "step": 4578
+    },
+    {
+      "epoch": 0.3179419525065963,
+      "grad_norm": 2.0736181021625506,
+      "learning_rate": 7.979441973924496e-06,
+      "loss": 0.2376,
+      "step": 4579
+    },
+    {
+      "epoch": 0.3180113873073184,
+      "grad_norm": 3.545523347003672,
+      "learning_rate": 7.97853885964096e-06,
+      "loss": 0.4363,
+      "step": 4580
+    },
+    {
+      "epoch": 0.31808082210804056,
+      "grad_norm": 3.395274430805249,
+      "learning_rate": 7.977635594706298e-06,
+      "loss": 0.466,
+      "step": 4581
+    },
+    {
+      "epoch": 0.3181502569087627,
+      "grad_norm": 4.3213454506717195,
+      "learning_rate": 7.9767321791662e-06,
+      "loss": 0.5375,
+      "step": 4582
+    },
+    {
+      "epoch": 0.3182196917094848,
+      "grad_norm": 3.8104639067360657,
+      "learning_rate": 7.975828613066354e-06,
+      "loss": 0.4245,
+      "step": 4583
+    },
+    {
+      "epoch": 0.3182891265102069,
+      "grad_norm": 3.9666997296154505,
+      "learning_rate": 7.974924896452466e-06,
+      "loss": 0.4862,
+      "step": 4584
+    },
+    {
+      "epoch": 0.31835856131092904,
+      "grad_norm": 5.364981480732536,
+      "learning_rate": 7.974021029370242e-06,
+      "loss": 0.7239,
+      "step": 4585
+    },
+    {
+      "epoch": 0.3184279961116512,
+      "grad_norm": 3.08931924921442,
+      "learning_rate": 7.973117011865398e-06,
+      "loss": 0.3552,
+      "step": 4586
+    },
+    {
+      "epoch": 0.31849743091237326,
+      "grad_norm": 3.10929712553979,
+      "learning_rate": 7.972212843983662e-06,
+      "loss": 0.3238,
+      "step": 4587
+    },
+    {
+      "epoch": 0.3185668657130954,
+      "grad_norm": 2.50848245218919,
+      "learning_rate": 7.971308525770763e-06,
+      "loss": 0.1952,
+      "step": 4588
+    },
+    {
+      "epoch": 0.3186363005138175,
+      "grad_norm": 4.6272823142786885,
+      "learning_rate": 7.970404057272441e-06,
+      "loss": 0.5662,
+      "step": 4589
+    },
+    {
+      "epoch": 0.31870573531453966,
+      "grad_norm": 4.317798080943889,
+      "learning_rate": 7.969499438534445e-06,
+      "loss": 0.5852,
+      "step": 4590
+    },
+    {
+      "epoch": 0.3187751701152618,
+      "grad_norm": 3.758440421974862,
+      "learning_rate": 7.968594669602525e-06,
+      "loss": 0.5574,
+      "step": 4591
+    },
+    {
+      "epoch": 0.3188446049159839,
+      "grad_norm": 5.321842531780306,
+      "learning_rate": 7.967689750522447e-06,
+      "loss": 0.6716,
+      "step": 4592
+    },
+    {
+      "epoch": 0.318914039716706,
+      "grad_norm": 4.271673454246642,
+      "learning_rate": 7.966784681339978e-06,
+      "loss": 0.5639,
+      "step": 4593
+    },
+    {
+      "epoch": 0.31898347451742814,
+      "grad_norm": 4.166481008501559,
+      "learning_rate": 7.965879462100898e-06,
+      "loss": 0.6935,
+      "step": 4594
+    },
+    {
+      "epoch": 0.3190529093181503,
+      "grad_norm": 3.6344461696172634,
+      "learning_rate": 7.964974092850992e-06,
+      "loss": 0.3892,
+      "step": 4595
+    },
+    {
+      "epoch": 0.31912234411887236,
+      "grad_norm": 3.6726102318285117,
+      "learning_rate": 7.964068573636051e-06,
+      "loss": 0.4744,
+      "step": 4596
+    },
+    {
+      "epoch": 0.3191917789195945,
+      "grad_norm": 4.173658715763125,
+      "learning_rate": 7.963162904501874e-06,
+      "loss": 0.5806,
+      "step": 4597
+    },
+    {
+      "epoch": 0.31926121372031663,
+      "grad_norm": 3.1698397010838977,
+      "learning_rate": 7.962257085494273e-06,
+      "loss": 0.3242,
+      "step": 4598
+    },
+    {
+      "epoch": 0.31933064852103876,
+      "grad_norm": 3.1721050466740297,
+      "learning_rate": 7.96135111665906e-06,
+      "loss": 0.2268,
+      "step": 4599
+    },
+    {
+      "epoch": 0.31940008332176084,
+      "grad_norm": 3.6555288223855484,
+      "learning_rate": 7.960444998042057e-06,
+      "loss": 0.1804,
+      "step": 4600
+    },
+    {
+      "epoch": 0.319469518122483,
+      "grad_norm": 5.5226825700954025,
+      "learning_rate": 7.959538729689095e-06,
+      "loss": 0.8646,
+      "step": 4601
+    },
+    {
+      "epoch": 0.3195389529232051,
+      "grad_norm": 2.9690630029429204,
+      "learning_rate": 7.958632311646015e-06,
+      "loss": 0.28,
+      "step": 4602
+    },
+    {
+      "epoch": 0.31960838772392725,
+      "grad_norm": 3.665297740589691,
+      "learning_rate": 7.957725743958662e-06,
+      "loss": 0.4565,
+      "step": 4603
+    },
+    {
+      "epoch": 0.3196778225246494,
+      "grad_norm": 3.806323776510999,
+      "learning_rate": 7.956819026672885e-06,
+      "loss": 0.5064,
+      "step": 4604
+    },
+    {
+      "epoch": 0.31974725732537146,
+      "grad_norm": 3.321615353592721,
+      "learning_rate": 7.955912159834549e-06,
+      "loss": 0.4375,
+      "step": 4605
+    },
+    {
+      "epoch": 0.3198166921260936,
+      "grad_norm": 3.1440187834910986,
+      "learning_rate": 7.95500514348952e-06,
+      "loss": 0.3389,
+      "step": 4606
+    },
+    {
+      "epoch": 0.31988612692681573,
+      "grad_norm": 3.550764641168834,
+      "learning_rate": 7.954097977683675e-06,
+      "loss": 0.3334,
+      "step": 4607
+    },
+    {
+      "epoch": 0.31995556172753786,
+      "grad_norm": 5.012512197261403,
+      "learning_rate": 7.953190662462897e-06,
+      "loss": 0.5701,
+      "step": 4608
+    },
+    {
+      "epoch": 0.32002499652825994,
+      "grad_norm": 2.7699909374771035,
+      "learning_rate": 7.952283197873077e-06,
+      "loss": 0.2799,
+      "step": 4609
+    },
+    {
+      "epoch": 0.3200944313289821,
+      "grad_norm": 4.770496178497335,
+      "learning_rate": 7.951375583960114e-06,
+      "loss": 0.6681,
+      "step": 4610
+    },
+    {
+      "epoch": 0.3201638661297042,
+      "grad_norm": 4.361290994687142,
+      "learning_rate": 7.950467820769914e-06,
+      "loss": 0.5187,
+      "step": 4611
+    },
+    {
+      "epoch": 0.32023330093042635,
+      "grad_norm": 4.435622470136309,
+      "learning_rate": 7.949559908348389e-06,
+      "loss": 0.6855,
+      "step": 4612
+    },
+    {
+      "epoch": 0.3203027357311484,
+      "grad_norm": 4.5653730498547755,
+      "learning_rate": 7.948651846741461e-06,
+      "loss": 0.5833,
+      "step": 4613
+    },
+    {
+      "epoch": 0.32037217053187056,
+      "grad_norm": 4.503956814882699,
+      "learning_rate": 7.947743635995062e-06,
+      "loss": 0.8082,
+      "step": 4614
+    },
+    {
+      "epoch": 0.3204416053325927,
+      "grad_norm": 3.1523464676738393,
+      "learning_rate": 7.946835276155123e-06,
+      "loss": 0.4796,
+      "step": 4615
+    },
+    {
+      "epoch": 0.32051104013331483,
+      "grad_norm": 3.9845614660329645,
+      "learning_rate": 7.945926767267591e-06,
+      "loss": 0.4445,
+      "step": 4616
+    },
+    {
+      "epoch": 0.32058047493403696,
+      "grad_norm": 4.213881643869688,
+      "learning_rate": 7.945018109378417e-06,
+      "loss": 0.7245,
+      "step": 4617
+    },
+    {
+      "epoch": 0.32064990973475904,
+      "grad_norm": 3.2213069592082957,
+      "learning_rate": 7.944109302533559e-06,
+      "loss": 0.2825,
+      "step": 4618
+    },
+    {
+      "epoch": 0.3207193445354812,
+      "grad_norm": 6.864401109934016,
+      "learning_rate": 7.943200346778986e-06,
+      "loss": 0.6383,
+      "step": 4619
+    },
+    {
+      "epoch": 0.3207887793362033,
+      "grad_norm": 3.805601480082595,
+      "learning_rate": 7.942291242160667e-06,
+      "loss": 0.434,
+      "step": 4620
+    },
+    {
+      "epoch": 0.32085821413692545,
+      "grad_norm": 4.174043821867604,
+      "learning_rate": 7.941381988724588e-06,
+      "loss": 0.5853,
+      "step": 4621
+    },
+    {
+      "epoch": 0.3209276489376475,
+      "grad_norm": 3.913797952737136,
+      "learning_rate": 7.940472586516736e-06,
+      "loss": 0.5452,
+      "step": 4622
+    },
+    {
+      "epoch": 0.32099708373836966,
+      "grad_norm": 4.592740016667488,
+      "learning_rate": 7.939563035583109e-06,
+      "loss": 0.7351,
+      "step": 4623
+    },
+    {
+      "epoch": 0.3210665185390918,
+      "grad_norm": 3.854015029196456,
+      "learning_rate": 7.938653335969708e-06,
+      "loss": 0.6549,
+      "step": 4624
+    },
+    {
+      "epoch": 0.32113595333981393,
+      "grad_norm": 4.83387939563592,
+      "learning_rate": 7.937743487722546e-06,
+      "loss": 0.4381,
+      "step": 4625
+    },
+    {
+      "epoch": 0.32120538814053606,
+      "grad_norm": 3.7113110721342855,
+      "learning_rate": 7.936833490887645e-06,
+      "loss": 0.507,
+      "step": 4626
+    },
+    {
+      "epoch": 0.32127482294125814,
+      "grad_norm": 3.864979484085414,
+      "learning_rate": 7.93592334551103e-06,
+      "loss": 0.4661,
+      "step": 4627
+    },
+    {
+      "epoch": 0.3213442577419803,
+      "grad_norm": 4.7423085173980875,
+      "learning_rate": 7.935013051638732e-06,
+      "loss": 0.5067,
+      "step": 4628
+    },
+    {
+      "epoch": 0.3214136925427024,
+      "grad_norm": 4.047918617692219,
+      "learning_rate": 7.934102609316796e-06,
+      "loss": 0.6548,
+      "step": 4629
+    },
+    {
+      "epoch": 0.32148312734342455,
+      "grad_norm": 4.9278533171062495,
+      "learning_rate": 7.93319201859127e-06,
+      "loss": 0.3736,
+      "step": 4630
+    },
+    {
+      "epoch": 0.3215525621441466,
+      "grad_norm": 4.657116767590137,
+      "learning_rate": 7.932281279508211e-06,
+      "loss": 0.5713,
+      "step": 4631
+    },
+    {
+      "epoch": 0.32162199694486876,
+      "grad_norm": 4.508238235123358,
+      "learning_rate": 7.931370392113684e-06,
+      "loss": 0.6774,
+      "step": 4632
+    },
+    {
+      "epoch": 0.3216914317455909,
+      "grad_norm": 3.792245474840228,
+      "learning_rate": 7.930459356453757e-06,
+      "loss": 0.5425,
+      "step": 4633
+    },
+    {
+      "epoch": 0.32176086654631303,
+      "grad_norm": 4.444086239333384,
+      "learning_rate": 7.929548172574515e-06,
+      "loss": 0.5196,
+      "step": 4634
+    },
+    {
+      "epoch": 0.3218303013470351,
+      "grad_norm": 2.8759854741317192,
+      "learning_rate": 7.92863684052204e-06,
+      "loss": 0.2465,
+      "step": 4635
+    },
+    {
+      "epoch": 0.32189973614775724,
+      "grad_norm": 3.777057895864684,
+      "learning_rate": 7.927725360342426e-06,
+      "loss": 0.4267,
+      "step": 4636
+    },
+    {
+      "epoch": 0.3219691709484794,
+      "grad_norm": 3.596245613784351,
+      "learning_rate": 7.926813732081778e-06,
+      "loss": 0.5298,
+      "step": 4637
+    },
+    {
+      "epoch": 0.3220386057492015,
+      "grad_norm": 2.5888536577246923,
+      "learning_rate": 7.925901955786203e-06,
+      "loss": 0.2207,
+      "step": 4638
+    },
+    {
+      "epoch": 0.32210804054992365,
+      "grad_norm": 3.430925414938377,
+      "learning_rate": 7.924990031501818e-06,
+      "loss": 0.4014,
+      "step": 4639
+    },
+    {
+      "epoch": 0.3221774753506457,
+      "grad_norm": 3.6567542735206833,
+      "learning_rate": 7.924077959274748e-06,
+      "loss": 0.3783,
+      "step": 4640
+    },
+    {
+      "epoch": 0.32224691015136786,
+      "grad_norm": 4.284605663095535,
+      "learning_rate": 7.923165739151123e-06,
+      "loss": 0.5944,
+      "step": 4641
+    },
+    {
+      "epoch": 0.32231634495209,
+      "grad_norm": 4.158889384928871,
+      "learning_rate": 7.922253371177081e-06,
+      "loss": 0.3697,
+      "step": 4642
+    },
+    {
+      "epoch": 0.32238577975281213,
+      "grad_norm": 6.163553069002673,
+      "learning_rate": 7.92134085539877e-06,
+      "loss": 0.8389,
+      "step": 4643
+    },
+    {
+      "epoch": 0.3224552145535342,
+      "grad_norm": 4.416300654922108,
+      "learning_rate": 7.920428191862349e-06,
+      "loss": 0.5493,
+      "step": 4644
+    },
+    {
+      "epoch": 0.32252464935425634,
+      "grad_norm": 4.884184625388962,
+      "learning_rate": 7.91951538061397e-06,
+      "loss": 0.7031,
+      "step": 4645
+    },
+    {
+      "epoch": 0.3225940841549785,
+      "grad_norm": 4.898345432005847,
+      "learning_rate": 7.918602421699808e-06,
+      "loss": 0.6119,
+      "step": 4646
+    },
+    {
+      "epoch": 0.3226635189557006,
+      "grad_norm": 4.394891147265864,
+      "learning_rate": 7.917689315166039e-06,
+      "loss": 0.6013,
+      "step": 4647
+    },
+    {
+      "epoch": 0.3227329537564227,
+      "grad_norm": 4.692922459589077,
+      "learning_rate": 7.916776061058846e-06,
+      "loss": 0.7299,
+      "step": 4648
+    },
+    {
+      "epoch": 0.3228023885571448,
+      "grad_norm": 3.8570564366702866,
+      "learning_rate": 7.915862659424419e-06,
+      "loss": 0.4548,
+      "step": 4649
+    },
+    {
+      "epoch": 0.32287182335786696,
+      "grad_norm": 6.502251882988549,
+      "learning_rate": 7.914949110308959e-06,
+      "loss": 0.7557,
+      "step": 4650
+    },
+    {
+      "epoch": 0.3229412581585891,
+      "grad_norm": 3.8412091538633804,
+      "learning_rate": 7.914035413758671e-06,
+      "loss": 0.3041,
+      "step": 4651
+    },
+    {
+      "epoch": 0.32301069295931123,
+      "grad_norm": 5.0526817022030555,
+      "learning_rate": 7.91312156981977e-06,
+      "loss": 0.664,
+      "step": 4652
+    },
+    {
+      "epoch": 0.3230801277600333,
+      "grad_norm": 4.2386209293253865,
+      "learning_rate": 7.912207578538478e-06,
+      "loss": 0.5079,
+      "step": 4653
+    },
+    {
+      "epoch": 0.32314956256075544,
+      "grad_norm": 4.059204685867363,
+      "learning_rate": 7.91129343996102e-06,
+      "loss": 0.4375,
+      "step": 4654
+    },
+    {
+      "epoch": 0.3232189973614776,
+      "grad_norm": 4.165574699128413,
+      "learning_rate": 7.910379154133634e-06,
+      "loss": 0.6977,
+      "step": 4655
+    },
+    {
+      "epoch": 0.3232884321621997,
+      "grad_norm": 3.743241076099596,
+      "learning_rate": 7.909464721102564e-06,
+      "loss": 0.5233,
+      "step": 4656
+    },
+    {
+      "epoch": 0.3233578669629218,
+      "grad_norm": 4.3367259118941535,
+      "learning_rate": 7.90855014091406e-06,
+      "loss": 0.5999,
+      "step": 4657
+    },
+    {
+      "epoch": 0.3234273017636439,
+      "grad_norm": 3.6506698632173835,
+      "learning_rate": 7.907635413614383e-06,
+      "loss": 0.4829,
+      "step": 4658
+    },
+    {
+      "epoch": 0.32349673656436606,
+      "grad_norm": 2.508147130562036,
+      "learning_rate": 7.906720539249795e-06,
+      "loss": 0.2944,
+      "step": 4659
+    },
+    {
+      "epoch": 0.3235661713650882,
+      "grad_norm": 3.8889760772615203,
+      "learning_rate": 7.905805517866572e-06,
+      "loss": 0.6976,
+      "step": 4660
+    },
+    {
+      "epoch": 0.32363560616581033,
+      "grad_norm": 4.054815641364154,
+      "learning_rate": 7.904890349510994e-06,
+      "loss": 0.63,
+      "step": 4661
+    },
+    {
+      "epoch": 0.3237050409665324,
+      "grad_norm": 3.8996557833561702,
+      "learning_rate": 7.90397503422935e-06,
+      "loss": 0.5693,
+      "step": 4662
+    },
+    {
+      "epoch": 0.32377447576725454,
+      "grad_norm": 3.8678112981275903,
+      "learning_rate": 7.903059572067933e-06,
+      "loss": 0.5863,
+      "step": 4663
+    },
+    {
+      "epoch": 0.3238439105679767,
+      "grad_norm": 4.09382195206162,
+      "learning_rate": 7.90214396307305e-06,
+      "loss": 0.4455,
+      "step": 4664
+    },
+    {
+      "epoch": 0.3239133453686988,
+      "grad_norm": 3.190047969818176,
+      "learning_rate": 7.90122820729101e-06,
+      "loss": 0.365,
+      "step": 4665
+    },
+    {
+      "epoch": 0.3239827801694209,
+      "grad_norm": 3.914207039534262,
+      "learning_rate": 7.900312304768127e-06,
+      "loss": 0.4955,
+      "step": 4666
+    },
+    {
+      "epoch": 0.324052214970143,
+      "grad_norm": 4.258849357837336,
+      "learning_rate": 7.899396255550733e-06,
+      "loss": 0.5264,
+      "step": 4667
+    },
+    {
+      "epoch": 0.32412164977086516,
+      "grad_norm": 3.734791926017694,
+      "learning_rate": 7.898480059685154e-06,
+      "loss": 0.4164,
+      "step": 4668
+    },
+    {
+      "epoch": 0.3241910845715873,
+      "grad_norm": 3.938144601132886,
+      "learning_rate": 7.897563717217736e-06,
+      "loss": 0.5074,
+      "step": 4669
+    },
+    {
+      "epoch": 0.3242605193723094,
+      "grad_norm": 4.458406724746656,
+      "learning_rate": 7.896647228194823e-06,
+      "loss": 0.7585,
+      "step": 4670
+    },
+    {
+      "epoch": 0.3243299541730315,
+      "grad_norm": 3.672883063241422,
+      "learning_rate": 7.895730592662772e-06,
+      "loss": 0.3493,
+      "step": 4671
+    },
+    {
+      "epoch": 0.32439938897375364,
+      "grad_norm": 4.024615462551951,
+      "learning_rate": 7.894813810667942e-06,
+      "loss": 0.4803,
+      "step": 4672
+    },
+    {
+      "epoch": 0.3244688237744758,
+      "grad_norm": 4.5435348849871495,
+      "learning_rate": 7.893896882256707e-06,
+      "loss": 0.775,
+      "step": 4673
+    },
+    {
+      "epoch": 0.3245382585751979,
+      "grad_norm": 3.321601545819084,
+      "learning_rate": 7.892979807475443e-06,
+      "loss": 0.3376,
+      "step": 4674
+    },
+    {
+      "epoch": 0.32460769337592,
+      "grad_norm": 3.218001652286279,
+      "learning_rate": 7.892062586370533e-06,
+      "loss": 0.3474,
+      "step": 4675
+    },
+    {
+      "epoch": 0.3246771281766421,
+      "grad_norm": 3.505112846495203,
+      "learning_rate": 7.891145218988369e-06,
+      "loss": 0.445,
+      "step": 4676
+    },
+    {
+      "epoch": 0.32474656297736426,
+      "grad_norm": 3.449205586774331,
+      "learning_rate": 7.890227705375352e-06,
+      "loss": 0.5055,
+      "step": 4677
+    },
+    {
+      "epoch": 0.3248159977780864,
+      "grad_norm": 3.394558783722113,
+      "learning_rate": 7.889310045577889e-06,
+      "loss": 0.2543,
+      "step": 4678
+    },
+    {
+      "epoch": 0.3248854325788085,
+      "grad_norm": 4.194647851846029,
+      "learning_rate": 7.888392239642392e-06,
+      "loss": 0.5433,
+      "step": 4679
+    },
+    {
+      "epoch": 0.3249548673795306,
+      "grad_norm": 3.190438123651363,
+      "learning_rate": 7.887474287615285e-06,
+      "loss": 0.3345,
+      "step": 4680
+    },
+    {
+      "epoch": 0.32502430218025274,
+      "grad_norm": 4.548012915209548,
+      "learning_rate": 7.886556189542995e-06,
+      "loss": 0.4888,
+      "step": 4681
+    },
+    {
+      "epoch": 0.3250937369809749,
+      "grad_norm": 4.340228071042583,
+      "learning_rate": 7.88563794547196e-06,
+      "loss": 0.3026,
+      "step": 4682
+    },
+    {
+      "epoch": 0.325163171781697,
+      "grad_norm": 4.777459926891092,
+      "learning_rate": 7.884719555448623e-06,
+      "loss": 0.5877,
+      "step": 4683
+    },
+    {
+      "epoch": 0.3252326065824191,
+      "grad_norm": 4.557339607602754,
+      "learning_rate": 7.883801019519436e-06,
+      "loss": 0.5472,
+      "step": 4684
+    },
+    {
+      "epoch": 0.3253020413831412,
+      "grad_norm": 4.030100785730653,
+      "learning_rate": 7.882882337730855e-06,
+      "loss": 0.4973,
+      "step": 4685
+    },
+    {
+      "epoch": 0.32537147618386336,
+      "grad_norm": 5.210024784273937,
+      "learning_rate": 7.881963510129349e-06,
+      "loss": 0.6391,
+      "step": 4686
+    },
+    {
+      "epoch": 0.3254409109845855,
+      "grad_norm": 3.7215092746754004,
+      "learning_rate": 7.881044536761389e-06,
+      "loss": 0.5831,
+      "step": 4687
+    },
+    {
+      "epoch": 0.3255103457853076,
+      "grad_norm": 4.517778416764318,
+      "learning_rate": 7.880125417673457e-06,
+      "loss": 0.6917,
+      "step": 4688
+    },
+    {
+      "epoch": 0.3255797805860297,
+      "grad_norm": 3.8345256405396246,
+      "learning_rate": 7.87920615291204e-06,
+      "loss": 0.3102,
+      "step": 4689
+    },
+    {
+      "epoch": 0.32564921538675184,
+      "grad_norm": 3.6555974450580218,
+      "learning_rate": 7.878286742523634e-06,
+      "loss": 0.5172,
+      "step": 4690
+    },
+    {
+      "epoch": 0.325718650187474,
+      "grad_norm": 4.658409032547017,
+      "learning_rate": 7.877367186554741e-06,
+      "loss": 0.674,
+      "step": 4691
+    },
+    {
+      "epoch": 0.32578808498819606,
+      "grad_norm": 1.9841531113872088,
+      "learning_rate": 7.876447485051872e-06,
+      "loss": 0.183,
+      "step": 4692
+    },
+    {
+      "epoch": 0.3258575197889182,
+      "grad_norm": 3.8532623541618416,
+      "learning_rate": 7.875527638061544e-06,
+      "loss": 0.5343,
+      "step": 4693
+    },
+    {
+      "epoch": 0.3259269545896403,
+      "grad_norm": 3.2099282592893004,
+      "learning_rate": 7.874607645630283e-06,
+      "loss": 0.296,
+      "step": 4694
+    },
+    {
+      "epoch": 0.32599638939036246,
+      "grad_norm": 4.038551422799916,
+      "learning_rate": 7.873687507804619e-06,
+      "loss": 0.6481,
+      "step": 4695
+    },
+    {
+      "epoch": 0.3260658241910846,
+      "grad_norm": 4.193105456308426,
+      "learning_rate": 7.872767224631093e-06,
+      "loss": 0.4533,
+      "step": 4696
+    },
+    {
+      "epoch": 0.3261352589918067,
+      "grad_norm": 5.690814272373825,
+      "learning_rate": 7.871846796156251e-06,
+      "loss": 0.7202,
+      "step": 4697
+    },
+    {
+      "epoch": 0.3262046937925288,
+      "grad_norm": 3.2513754020271644,
+      "learning_rate": 7.87092622242665e-06,
+      "loss": 0.186,
+      "step": 4698
+    },
+    {
+      "epoch": 0.32627412859325094,
+      "grad_norm": 2.3770210229974658,
+      "learning_rate": 7.870005503488847e-06,
+      "loss": 0.1447,
+      "step": 4699
+    },
+    {
+      "epoch": 0.3263435633939731,
+      "grad_norm": 4.136432066972335,
+      "learning_rate": 7.869084639389412e-06,
+      "loss": 0.5305,
+      "step": 4700
+    },
+    {
+      "epoch": 0.32641299819469516,
+      "grad_norm": 3.610216002697001,
+      "learning_rate": 7.868163630174926e-06,
+      "loss": 0.6002,
+      "step": 4701
+    },
+    {
+      "epoch": 0.3264824329954173,
+      "grad_norm": 3.3223454580462897,
+      "learning_rate": 7.867242475891968e-06,
+      "loss": 0.5736,
+      "step": 4702
+    },
+    {
+      "epoch": 0.32655186779613943,
+      "grad_norm": 3.7511849357144924,
+      "learning_rate": 7.866321176587129e-06,
+      "loss": 0.424,
+      "step": 4703
+    },
+    {
+      "epoch": 0.32662130259686156,
+      "grad_norm": 5.260149073115975,
+      "learning_rate": 7.865399732307008e-06,
+      "loss": 0.7874,
+      "step": 4704
+    },
+    {
+      "epoch": 0.32669073739758364,
+      "grad_norm": 3.3025411040532116,
+      "learning_rate": 7.864478143098212e-06,
+      "loss": 0.3391,
+      "step": 4705
+    },
+    {
+      "epoch": 0.3267601721983058,
+      "grad_norm": 4.20691907899176,
+      "learning_rate": 7.863556409007355e-06,
+      "loss": 0.7692,
+      "step": 4706
+    },
+    {
+      "epoch": 0.3268296069990279,
+      "grad_norm": 4.748310725321278,
+      "learning_rate": 7.862634530081053e-06,
+      "loss": 0.8968,
+      "step": 4707
+    },
+    {
+      "epoch": 0.32689904179975005,
+      "grad_norm": 3.588014528880619,
+      "learning_rate": 7.861712506365934e-06,
+      "loss": 0.4014,
+      "step": 4708
+    },
+    {
+      "epoch": 0.3269684766004722,
+      "grad_norm": 3.70231044481188,
+      "learning_rate": 7.860790337908638e-06,
+      "loss": 0.4725,
+      "step": 4709
+    },
+    {
+      "epoch": 0.32703791140119426,
+      "grad_norm": 4.813706646372495,
+      "learning_rate": 7.859868024755802e-06,
+      "loss": 0.605,
+      "step": 4710
+    },
+    {
+      "epoch": 0.3271073462019164,
+      "grad_norm": 3.59113578779413,
+      "learning_rate": 7.858945566954076e-06,
+      "loss": 0.4283,
+      "step": 4711
+    },
+    {
+      "epoch": 0.32717678100263853,
+      "grad_norm": 4.1891251646625145,
+      "learning_rate": 7.858022964550122e-06,
+      "loss": 0.7248,
+      "step": 4712
+    },
+    {
+      "epoch": 0.32724621580336066,
+      "grad_norm": 3.466639358787095,
+      "learning_rate": 7.857100217590599e-06,
+      "loss": 0.3072,
+      "step": 4713
+    },
+    {
+      "epoch": 0.32731565060408274,
+      "grad_norm": 3.635187172134428,
+      "learning_rate": 7.856177326122179e-06,
+      "loss": 0.4059,
+      "step": 4714
+    },
+    {
+      "epoch": 0.3273850854048049,
+      "grad_norm": 4.835500747168967,
+      "learning_rate": 7.855254290191543e-06,
+      "loss": 0.7096,
+      "step": 4715
+    },
+    {
+      "epoch": 0.327454520205527,
+      "grad_norm": 3.172193259516325,
+      "learning_rate": 7.854331109845376e-06,
+      "loss": 0.3931,
+      "step": 4716
+    },
+    {
+      "epoch": 0.32752395500624915,
+      "grad_norm": 3.7820031310325692,
+      "learning_rate": 7.853407785130371e-06,
+      "loss": 0.5338,
+      "step": 4717
+    },
+    {
+      "epoch": 0.3275933898069713,
+      "grad_norm": 3.343050140866542,
+      "learning_rate": 7.852484316093227e-06,
+      "loss": 0.4621,
+      "step": 4718
+    },
+    {
+      "epoch": 0.32766282460769336,
+      "grad_norm": 3.9915269148585972,
+      "learning_rate": 7.851560702780658e-06,
+      "loss": 0.4657,
+      "step": 4719
+    },
+    {
+      "epoch": 0.3277322594084155,
+      "grad_norm": 4.2046483794872005,
+      "learning_rate": 7.850636945239373e-06,
+      "loss": 0.7303,
+      "step": 4720
+    },
+    {
+      "epoch": 0.32780169420913763,
+      "grad_norm": 4.004710359783434,
+      "learning_rate": 7.849713043516097e-06,
+      "loss": 0.6645,
+      "step": 4721
+    },
+    {
+      "epoch": 0.32787112900985976,
+      "grad_norm": 4.337612345542736,
+      "learning_rate": 7.848788997657561e-06,
+      "loss": 0.5615,
+      "step": 4722
+    },
+    {
+      "epoch": 0.32794056381058184,
+      "grad_norm": 3.9725592507520737,
+      "learning_rate": 7.8478648077105e-06,
+      "loss": 0.4277,
+      "step": 4723
+    },
+    {
+      "epoch": 0.328009998611304,
+      "grad_norm": 3.8417761165988633,
+      "learning_rate": 7.84694047372166e-06,
+      "loss": 0.4871,
+      "step": 4724
+    },
+    {
+      "epoch": 0.3280794334120261,
+      "grad_norm": 5.023924331115942,
+      "learning_rate": 7.846015995737792e-06,
+      "loss": 0.7762,
+      "step": 4725
+    },
+    {
+      "epoch": 0.32814886821274825,
+      "grad_norm": 3.8734636495564474,
+      "learning_rate": 7.845091373805656e-06,
+      "loss": 0.3322,
+      "step": 4726
+    },
+    {
+      "epoch": 0.3282183030134703,
+      "grad_norm": 3.9789398545757715,
+      "learning_rate": 7.844166607972018e-06,
+      "loss": 0.5292,
+      "step": 4727
+    },
+    {
+      "epoch": 0.32828773781419246,
+      "grad_norm": 4.649774119117001,
+      "learning_rate": 7.843241698283652e-06,
+      "loss": 0.6158,
+      "step": 4728
+    },
+    {
+      "epoch": 0.3283571726149146,
+      "grad_norm": 4.185950464704938,
+      "learning_rate": 7.842316644787336e-06,
+      "loss": 0.4567,
+      "step": 4729
+    },
+    {
+      "epoch": 0.32842660741563673,
+      "grad_norm": 3.6639796818769503,
+      "learning_rate": 7.841391447529862e-06,
+      "loss": 0.5612,
+      "step": 4730
+    },
+    {
+      "epoch": 0.32849604221635886,
+      "grad_norm": 3.2403976095474807,
+      "learning_rate": 7.840466106558025e-06,
+      "loss": 0.3463,
+      "step": 4731
+    },
+    {
+      "epoch": 0.32856547701708094,
+      "grad_norm": 4.194396170098813,
+      "learning_rate": 7.839540621918625e-06,
+      "loss": 0.506,
+      "step": 4732
+    },
+    {
+      "epoch": 0.3286349118178031,
+      "grad_norm": 3.824496801777473,
+      "learning_rate": 7.838614993658474e-06,
+      "loss": 0.4243,
+      "step": 4733
+    },
+    {
+      "epoch": 0.3287043466185252,
+      "grad_norm": 4.014966399424347,
+      "learning_rate": 7.837689221824388e-06,
+      "loss": 0.6806,
+      "step": 4734
+    },
+    {
+      "epoch": 0.32877378141924735,
+      "grad_norm": 4.2485326973969615,
+      "learning_rate": 7.836763306463191e-06,
+      "loss": 0.5998,
+      "step": 4735
+    },
+    {
+      "epoch": 0.3288432162199694,
+      "grad_norm": 4.00131314388345,
+      "learning_rate": 7.835837247621718e-06,
+      "loss": 0.395,
+      "step": 4736
+    },
+    {
+      "epoch": 0.32891265102069156,
+      "grad_norm": 4.039448615559206,
+      "learning_rate": 7.834911045346805e-06,
+      "loss": 0.2687,
+      "step": 4737
+    },
+    {
+      "epoch": 0.3289820858214137,
+      "grad_norm": 3.3593159033491093,
+      "learning_rate": 7.8339846996853e-06,
+      "loss": 0.4399,
+      "step": 4738
+    },
+    {
+      "epoch": 0.32905152062213583,
+      "grad_norm": 3.742343244644939,
+      "learning_rate": 7.833058210684055e-06,
+      "loss": 0.2418,
+      "step": 4739
+    },
+    {
+      "epoch": 0.32912095542285796,
+      "grad_norm": 3.4910570011367748,
+      "learning_rate": 7.832131578389932e-06,
+      "loss": 0.4679,
+      "step": 4740
+    },
+    {
+      "epoch": 0.32919039022358004,
+      "grad_norm": 3.7221313590050853,
+      "learning_rate": 7.8312048028498e-06,
+      "loss": 0.4242,
+      "step": 4741
+    },
+    {
+      "epoch": 0.3292598250243022,
+      "grad_norm": 3.437734492948652,
+      "learning_rate": 7.830277884110531e-06,
+      "loss": 0.3457,
+      "step": 4742
+    },
+    {
+      "epoch": 0.3293292598250243,
+      "grad_norm": 3.187803973487701,
+      "learning_rate": 7.829350822219008e-06,
+      "loss": 0.5265,
+      "step": 4743
+    },
+    {
+      "epoch": 0.32939869462574645,
+      "grad_norm": 4.5191883912061765,
+      "learning_rate": 7.828423617222125e-06,
+      "loss": 0.4803,
+      "step": 4744
+    },
+    {
+      "epoch": 0.3294681294264685,
+      "grad_norm": 3.0449015249764235,
+      "learning_rate": 7.827496269166774e-06,
+      "loss": 0.2957,
+      "step": 4745
+    },
+    {
+      "epoch": 0.32953756422719066,
+      "grad_norm": 4.796420550732095,
+      "learning_rate": 7.826568778099862e-06,
+      "loss": 0.4296,
+      "step": 4746
+    },
+    {
+      "epoch": 0.3296069990279128,
+      "grad_norm": 3.906738951938081,
+      "learning_rate": 7.825641144068299e-06,
+      "loss": 0.4259,
+      "step": 4747
+    },
+    {
+      "epoch": 0.32967643382863493,
+      "grad_norm": 3.8564593328589925,
+      "learning_rate": 7.824713367119006e-06,
+      "loss": 0.3805,
+      "step": 4748
+    },
+    {
+      "epoch": 0.329745868629357,
+      "grad_norm": 2.8244329478739427,
+      "learning_rate": 7.823785447298906e-06,
+      "loss": 0.2499,
+      "step": 4749
+    },
+    {
+      "epoch": 0.32981530343007914,
+      "grad_norm": 2.44415300355179,
+      "learning_rate": 7.822857384654934e-06,
+      "loss": 0.197,
+      "step": 4750
+    },
+    {
+      "epoch": 0.3298847382308013,
+      "grad_norm": 4.004220519866786,
+      "learning_rate": 7.82192917923403e-06,
+      "loss": 0.5909,
+      "step": 4751
+    },
+    {
+      "epoch": 0.3299541730315234,
+      "grad_norm": 4.008199758870167,
+      "learning_rate": 7.821000831083139e-06,
+      "loss": 0.5831,
+      "step": 4752
+    },
+    {
+      "epoch": 0.33002360783224555,
+      "grad_norm": 2.6884398092751107,
+      "learning_rate": 7.82007234024922e-06,
+      "loss": 0.3029,
+      "step": 4753
+    },
+    {
+      "epoch": 0.3300930426329676,
+      "grad_norm": 2.9098692925158036,
+      "learning_rate": 7.819143706779233e-06,
+      "loss": 0.2855,
+      "step": 4754
+    },
+    {
+      "epoch": 0.33016247743368976,
+      "grad_norm": 3.649353618292213,
+      "learning_rate": 7.818214930720149e-06,
+      "loss": 0.3424,
+      "step": 4755
+    },
+    {
+      "epoch": 0.3302319122344119,
+      "grad_norm": 4.2660052893192795,
+      "learning_rate": 7.817286012118941e-06,
+      "loss": 0.4426,
+      "step": 4756
+    },
+    {
+      "epoch": 0.33030134703513403,
+      "grad_norm": 4.269290924481466,
+      "learning_rate": 7.816356951022594e-06,
+      "loss": 0.55,
+      "step": 4757
+    },
+    {
+      "epoch": 0.3303707818358561,
+      "grad_norm": 3.588995613667248,
+      "learning_rate": 7.8154277474781e-06,
+      "loss": 0.3982,
+      "step": 4758
+    },
+    {
+      "epoch": 0.33044021663657824,
+      "grad_norm": 3.8191343968506777,
+      "learning_rate": 7.814498401532456e-06,
+      "loss": 0.5013,
+      "step": 4759
+    },
+    {
+      "epoch": 0.3305096514373004,
+      "grad_norm": 3.7814610227534966,
+      "learning_rate": 7.81356891323267e-06,
+      "loss": 0.4199,
+      "step": 4760
+    },
+    {
+      "epoch": 0.3305790862380225,
+      "grad_norm": 2.542225823851452,
+      "learning_rate": 7.812639282625748e-06,
+      "loss": 0.2038,
+      "step": 4761
+    },
+    {
+      "epoch": 0.3306485210387446,
+      "grad_norm": 3.791506059479617,
+      "learning_rate": 7.811709509758719e-06,
+      "loss": 0.4281,
+      "step": 4762
+    },
+    {
+      "epoch": 0.3307179558394667,
+      "grad_norm": 3.4062761318242503,
+      "learning_rate": 7.810779594678599e-06,
+      "loss": 0.3439,
+      "step": 4763
+    },
+    {
+      "epoch": 0.33078739064018886,
+      "grad_norm": 3.5833761857715327,
+      "learning_rate": 7.809849537432432e-06,
+      "loss": 0.2861,
+      "step": 4764
+    },
+    {
+      "epoch": 0.330856825440911,
+      "grad_norm": 3.6414608191704176,
+      "learning_rate": 7.808919338067254e-06,
+      "loss": 0.4716,
+      "step": 4765
+    },
+    {
+      "epoch": 0.33092626024163313,
+      "grad_norm": 2.397444582857908,
+      "learning_rate": 7.807988996630112e-06,
+      "loss": 0.1842,
+      "step": 4766
+    },
+    {
+      "epoch": 0.3309956950423552,
+      "grad_norm": 3.8983879458180914,
+      "learning_rate": 7.807058513168066e-06,
+      "loss": 0.2066,
+      "step": 4767
+    },
+    {
+      "epoch": 0.33106512984307734,
+      "grad_norm": 4.768107241605324,
+      "learning_rate": 7.806127887728176e-06,
+      "loss": 0.407,
+      "step": 4768
+    },
+    {
+      "epoch": 0.3311345646437995,
+      "grad_norm": 3.9123896238914835,
+      "learning_rate": 7.805197120357515e-06,
+      "loss": 0.567,
+      "step": 4769
+    },
+    {
+      "epoch": 0.3312039994445216,
+      "grad_norm": 4.553800062624672,
+      "learning_rate": 7.804266211103155e-06,
+      "loss": 0.5677,
+      "step": 4770
+    },
+    {
+      "epoch": 0.3312734342452437,
+      "grad_norm": 4.992154002882232,
+      "learning_rate": 7.803335160012187e-06,
+      "loss": 0.7013,
+      "step": 4771
+    },
+    {
+      "epoch": 0.3313428690459658,
+      "grad_norm": 3.0707185097279734,
+      "learning_rate": 7.802403967131696e-06,
+      "loss": 0.3535,
+      "step": 4772
+    },
+    {
+      "epoch": 0.33141230384668796,
+      "grad_norm": 3.8771915613240666,
+      "learning_rate": 7.801472632508783e-06,
+      "loss": 0.5341,
+      "step": 4773
+    },
+    {
+      "epoch": 0.3314817386474101,
+      "grad_norm": 3.6241602583643573,
+      "learning_rate": 7.800541156190556e-06,
+      "loss": 0.4209,
+      "step": 4774
+    },
+    {
+      "epoch": 0.33155117344813223,
+      "grad_norm": 3.325742162634742,
+      "learning_rate": 7.799609538224124e-06,
+      "loss": 0.2486,
+      "step": 4775
+    },
+    {
+      "epoch": 0.3316206082488543,
+      "grad_norm": 3.5883195483550456,
+      "learning_rate": 7.798677778656613e-06,
+      "loss": 0.4583,
+      "step": 4776
+    },
+    {
+      "epoch": 0.33169004304957644,
+      "grad_norm": 3.620266839793677,
+      "learning_rate": 7.797745877535144e-06,
+      "loss": 0.4232,
+      "step": 4777
+    },
+    {
+      "epoch": 0.3317594778502986,
+      "grad_norm": 5.59170449305756,
+      "learning_rate": 7.796813834906856e-06,
+      "loss": 0.5315,
+      "step": 4778
+    },
+    {
+      "epoch": 0.3318289126510207,
+      "grad_norm": 2.922557851954977,
+      "learning_rate": 7.795881650818889e-06,
+      "loss": 0.2239,
+      "step": 4779
+    },
+    {
+      "epoch": 0.3318983474517428,
+      "grad_norm": 4.9684873010448785,
+      "learning_rate": 7.79494932531839e-06,
+      "loss": 0.3773,
+      "step": 4780
+    },
+    {
+      "epoch": 0.3319677822524649,
+      "grad_norm": 3.2627132718022143,
+      "learning_rate": 7.794016858452519e-06,
+      "loss": 0.3741,
+      "step": 4781
+    },
+    {
+      "epoch": 0.33203721705318706,
+      "grad_norm": 2.707711150782449,
+      "learning_rate": 7.793084250268435e-06,
+      "loss": 0.2963,
+      "step": 4782
+    },
+    {
+      "epoch": 0.3321066518539092,
+      "grad_norm": 3.4103403754598727,
+      "learning_rate": 7.792151500813311e-06,
+      "loss": 0.4594,
+      "step": 4783
+    },
+    {
+      "epoch": 0.3321760866546313,
+      "grad_norm": 4.4959981043701704,
+      "learning_rate": 7.791218610134324e-06,
+      "loss": 0.3072,
+      "step": 4784
+    },
+    {
+      "epoch": 0.3322455214553534,
+      "grad_norm": 3.078185755865892,
+      "learning_rate": 7.79028557827866e-06,
+      "loss": 0.3279,
+      "step": 4785
+    },
+    {
+      "epoch": 0.33231495625607554,
+      "grad_norm": 4.011355266928748,
+      "learning_rate": 7.789352405293506e-06,
+      "loss": 0.4368,
+      "step": 4786
+    },
+    {
+      "epoch": 0.3323843910567977,
+      "grad_norm": 4.371563747043994,
+      "learning_rate": 7.788419091226065e-06,
+      "loss": 0.5169,
+      "step": 4787
+    },
+    {
+      "epoch": 0.3324538258575198,
+      "grad_norm": 3.139082235794632,
+      "learning_rate": 7.78748563612354e-06,
+      "loss": 0.3709,
+      "step": 4788
+    },
+    {
+      "epoch": 0.3325232606582419,
+      "grad_norm": 3.335298063370998,
+      "learning_rate": 7.786552040033147e-06,
+      "loss": 0.4315,
+      "step": 4789
+    },
+    {
+      "epoch": 0.332592695458964,
+      "grad_norm": 3.8218843478523303,
+      "learning_rate": 7.785618303002106e-06,
+      "loss": 0.4739,
+      "step": 4790
+    },
+    {
+      "epoch": 0.33266213025968616,
+      "grad_norm": 3.1891259571647748,
+      "learning_rate": 7.784684425077641e-06,
+      "loss": 0.3058,
+      "step": 4791
+    },
+    {
+      "epoch": 0.3327315650604083,
+      "grad_norm": 3.7348887067457808,
+      "learning_rate": 7.783750406306992e-06,
+      "loss": 0.4707,
+      "step": 4792
+    },
+    {
+      "epoch": 0.3328009998611304,
+      "grad_norm": 4.192637608330812,
+      "learning_rate": 7.782816246737394e-06,
+      "loss": 0.5349,
+      "step": 4793
+    },
+    {
+      "epoch": 0.3328704346618525,
+      "grad_norm": 5.0784938113673785,
+      "learning_rate": 7.781881946416103e-06,
+      "loss": 0.6123,
+      "step": 4794
+    },
+    {
+      "epoch": 0.33293986946257464,
+      "grad_norm": 3.2915472326496795,
+      "learning_rate": 7.780947505390369e-06,
+      "loss": 0.4767,
+      "step": 4795
+    },
+    {
+      "epoch": 0.3330093042632968,
+      "grad_norm": 5.1741757841899085,
+      "learning_rate": 7.780012923707457e-06,
+      "loss": 0.6191,
+      "step": 4796
+    },
+    {
+      "epoch": 0.3330787390640189,
+      "grad_norm": 3.5247167295447985,
+      "learning_rate": 7.779078201414638e-06,
+      "loss": 0.6964,
+      "step": 4797
+    },
+    {
+      "epoch": 0.333148173864741,
+      "grad_norm": 2.914965141077988,
+      "learning_rate": 7.778143338559186e-06,
+      "loss": 0.2555,
+      "step": 4798
+    },
+    {
+      "epoch": 0.3332176086654631,
+      "grad_norm": 2.4532326864591054,
+      "learning_rate": 7.77720833518839e-06,
+      "loss": 0.1711,
+      "step": 4799
+    },
+    {
+      "epoch": 0.33328704346618526,
+      "grad_norm": 4.270107820155478,
+      "learning_rate": 7.776273191349537e-06,
+      "loss": 0.4523,
+      "step": 4800
+    },
+    {
+      "epoch": 0.3333564782669074,
+      "grad_norm": 3.8697806017873653,
+      "learning_rate": 7.77533790708993e-06,
+      "loss": 0.6405,
+      "step": 4801
+    },
+    {
+      "epoch": 0.3334259130676295,
+      "grad_norm": 4.150335586804698,
+      "learning_rate": 7.774402482456872e-06,
+      "loss": 0.5049,
+      "step": 4802
+    },
+    {
+      "epoch": 0.3334953478683516,
+      "grad_norm": 3.8577780744431234,
+      "learning_rate": 7.773466917497674e-06,
+      "loss": 0.4201,
+      "step": 4803
+    },
+    {
+      "epoch": 0.33356478266907374,
+      "grad_norm": 3.824038058402649,
+      "learning_rate": 7.772531212259657e-06,
+      "loss": 0.3057,
+      "step": 4804
+    },
+    {
+      "epoch": 0.3336342174697959,
+      "grad_norm": 4.295519952819993,
+      "learning_rate": 7.77159536679015e-06,
+      "loss": 0.7575,
+      "step": 4805
+    },
+    {
+      "epoch": 0.33370365227051796,
+      "grad_norm": 4.015597405122909,
+      "learning_rate": 7.770659381136487e-06,
+      "loss": 0.3624,
+      "step": 4806
+    },
+    {
+      "epoch": 0.3337730870712401,
+      "grad_norm": 4.987754175448214,
+      "learning_rate": 7.769723255346006e-06,
+      "loss": 0.712,
+      "step": 4807
+    },
+    {
+      "epoch": 0.33384252187196223,
+      "grad_norm": 3.212356078967072,
+      "learning_rate": 7.768786989466058e-06,
+      "loss": 0.2764,
+      "step": 4808
+    },
+    {
+      "epoch": 0.33391195667268436,
+      "grad_norm": 3.858440909073672,
+      "learning_rate": 7.767850583543997e-06,
+      "loss": 0.4117,
+      "step": 4809
+    },
+    {
+      "epoch": 0.3339813914734065,
+      "grad_norm": 4.076307036761403,
+      "learning_rate": 7.766914037627184e-06,
+      "loss": 0.4454,
+      "step": 4810
+    },
+    {
+      "epoch": 0.3340508262741286,
+      "grad_norm": 3.6527313254892415,
+      "learning_rate": 7.76597735176299e-06,
+      "loss": 0.3244,
+      "step": 4811
+    },
+    {
+      "epoch": 0.3341202610748507,
+      "grad_norm": 3.8134843308834148,
+      "learning_rate": 7.765040525998794e-06,
+      "loss": 0.4097,
+      "step": 4812
+    },
+    {
+      "epoch": 0.33418969587557285,
+      "grad_norm": 3.0115712133756074,
+      "learning_rate": 7.764103560381976e-06,
+      "loss": 0.3068,
+      "step": 4813
+    },
+    {
+      "epoch": 0.334259130676295,
+      "grad_norm": 2.7229885429355005,
+      "learning_rate": 7.763166454959926e-06,
+      "loss": 0.2451,
+      "step": 4814
+    },
+    {
+      "epoch": 0.33432856547701706,
+      "grad_norm": 3.8717244842302927,
+      "learning_rate": 7.762229209780045e-06,
+      "loss": 0.2722,
+      "step": 4815
+    },
+    {
+      "epoch": 0.3343980002777392,
+      "grad_norm": 4.080329307689948,
+      "learning_rate": 7.761291824889736e-06,
+      "loss": 0.5056,
+      "step": 4816
+    },
+    {
+      "epoch": 0.33446743507846133,
+      "grad_norm": 3.511412978592732,
+      "learning_rate": 7.760354300336411e-06,
+      "loss": 0.2584,
+      "step": 4817
+    },
+    {
+      "epoch": 0.33453686987918346,
+      "grad_norm": 3.945974759117404,
+      "learning_rate": 7.759416636167488e-06,
+      "loss": 0.5245,
+      "step": 4818
+    },
+    {
+      "epoch": 0.33460630467990554,
+      "grad_norm": 3.4154578367973722,
+      "learning_rate": 7.758478832430396e-06,
+      "loss": 0.202,
+      "step": 4819
+    },
+    {
+      "epoch": 0.3346757394806277,
+      "grad_norm": 3.7451621818757252,
+      "learning_rate": 7.757540889172565e-06,
+      "loss": 0.395,
+      "step": 4820
+    },
+    {
+      "epoch": 0.3347451742813498,
+      "grad_norm": 3.919663514284129,
+      "learning_rate": 7.756602806441436e-06,
+      "loss": 0.5694,
+      "step": 4821
+    },
+    {
+      "epoch": 0.33481460908207195,
+      "grad_norm": 3.3997221928671637,
+      "learning_rate": 7.755664584284456e-06,
+      "loss": 0.5812,
+      "step": 4822
+    },
+    {
+      "epoch": 0.3348840438827941,
+      "grad_norm": 4.7651089585699875,
+      "learning_rate": 7.75472622274908e-06,
+      "loss": 0.6252,
+      "step": 4823
+    },
+    {
+      "epoch": 0.33495347868351616,
+      "grad_norm": 3.671551309405416,
+      "learning_rate": 7.75378772188277e-06,
+      "loss": 0.4007,
+      "step": 4824
+    },
+    {
+      "epoch": 0.3350229134842383,
+      "grad_norm": 3.780443719607119,
+      "learning_rate": 7.752849081732993e-06,
+      "loss": 0.495,
+      "step": 4825
+    },
+    {
+      "epoch": 0.33509234828496043,
+      "grad_norm": 4.809225419934876,
+      "learning_rate": 7.751910302347223e-06,
+      "loss": 0.5839,
+      "step": 4826
+    },
+    {
+      "epoch": 0.33516178308568256,
+      "grad_norm": 3.3365582124055027,
+      "learning_rate": 7.750971383772946e-06,
+      "loss": 0.4372,
+      "step": 4827
+    },
+    {
+      "epoch": 0.33523121788640464,
+      "grad_norm": 4.333372309869496,
+      "learning_rate": 7.750032326057647e-06,
+      "loss": 0.4893,
+      "step": 4828
+    },
+    {
+      "epoch": 0.3353006526871268,
+      "grad_norm": 2.926101144649098,
+      "learning_rate": 7.749093129248826e-06,
+      "loss": 0.4039,
+      "step": 4829
+    },
+    {
+      "epoch": 0.3353700874878489,
+      "grad_norm": 3.298689571862438,
+      "learning_rate": 7.748153793393986e-06,
+      "loss": 0.3748,
+      "step": 4830
+    },
+    {
+      "epoch": 0.33543952228857105,
+      "grad_norm": 3.3854606003846386,
+      "learning_rate": 7.747214318540635e-06,
+      "loss": 0.378,
+      "step": 4831
+    },
+    {
+      "epoch": 0.3355089570892932,
+      "grad_norm": 3.244868505274605,
+      "learning_rate": 7.746274704736294e-06,
+      "loss": 0.4066,
+      "step": 4832
+    },
+    {
+      "epoch": 0.33557839189001526,
+      "grad_norm": 3.9848895295749402,
+      "learning_rate": 7.745334952028486e-06,
+      "loss": 0.3815,
+      "step": 4833
+    },
+    {
+      "epoch": 0.3356478266907374,
+      "grad_norm": 3.2972008927057708,
+      "learning_rate": 7.744395060464742e-06,
+      "loss": 0.3845,
+      "step": 4834
+    },
+    {
+      "epoch": 0.33571726149145953,
+      "grad_norm": 4.46255539624817,
+      "learning_rate": 7.743455030092602e-06,
+      "loss": 0.8389,
+      "step": 4835
+    },
+    {
+      "epoch": 0.33578669629218166,
+      "grad_norm": 4.231880794881921,
+      "learning_rate": 7.74251486095961e-06,
+      "loss": 0.6917,
+      "step": 4836
+    },
+    {
+      "epoch": 0.33585613109290374,
+      "grad_norm": 4.304026469753395,
+      "learning_rate": 7.74157455311332e-06,
+      "loss": 0.6669,
+      "step": 4837
+    },
+    {
+      "epoch": 0.3359255658936259,
+      "grad_norm": 4.387056685524457,
+      "learning_rate": 7.740634106601292e-06,
+      "loss": 0.6693,
+      "step": 4838
+    },
+    {
+      "epoch": 0.335995000694348,
+      "grad_norm": 3.964334597191936,
+      "learning_rate": 7.739693521471093e-06,
+      "loss": 0.7126,
+      "step": 4839
+    },
+    {
+      "epoch": 0.33606443549507015,
+      "grad_norm": 3.4139242945942487,
+      "learning_rate": 7.738752797770294e-06,
+      "loss": 0.3177,
+      "step": 4840
+    },
+    {
+      "epoch": 0.3361338702957922,
+      "grad_norm": 4.1495741046687655,
+      "learning_rate": 7.737811935546478e-06,
+      "loss": 0.4242,
+      "step": 4841
+    },
+    {
+      "epoch": 0.33620330509651436,
+      "grad_norm": 3.027435921769314,
+      "learning_rate": 7.736870934847233e-06,
+      "loss": 0.2829,
+      "step": 4842
+    },
+    {
+      "epoch": 0.3362727398972365,
+      "grad_norm": 4.339870815562016,
+      "learning_rate": 7.735929795720151e-06,
+      "loss": 0.547,
+      "step": 4843
+    },
+    {
+      "epoch": 0.33634217469795863,
+      "grad_norm": 3.4442966852245105,
+      "learning_rate": 7.734988518212838e-06,
+      "loss": 0.3925,
+      "step": 4844
+    },
+    {
+      "epoch": 0.33641160949868076,
+      "grad_norm": 3.564799783814901,
+      "learning_rate": 7.7340471023729e-06,
+      "loss": 0.4119,
+      "step": 4845
+    },
+    {
+      "epoch": 0.33648104429940284,
+      "grad_norm": 2.6823870065591033,
+      "learning_rate": 7.733105548247954e-06,
+      "loss": 0.2629,
+      "step": 4846
+    },
+    {
+      "epoch": 0.336550479100125,
+      "grad_norm": 2.8283188997926776,
+      "learning_rate": 7.732163855885621e-06,
+      "loss": 0.3227,
+      "step": 4847
+    },
+    {
+      "epoch": 0.3366199139008471,
+      "grad_norm": 6.76110201181669,
+      "learning_rate": 7.731222025333534e-06,
+      "loss": 0.362,
+      "step": 4848
+    },
+    {
+      "epoch": 0.33668934870156925,
+      "grad_norm": 4.464878096866039,
+      "learning_rate": 7.730280056639326e-06,
+      "loss": 0.6647,
+      "step": 4849
+    },
+    {
+      "epoch": 0.3367587835022913,
+      "grad_norm": 4.255593172137241,
+      "learning_rate": 7.729337949850641e-06,
+      "loss": 0.3596,
+      "step": 4850
+    },
+    {
+      "epoch": 0.33682821830301346,
+      "grad_norm": 4.405526751743151,
+      "learning_rate": 7.728395705015133e-06,
+      "loss": 0.7481,
+      "step": 4851
+    },
+    {
+      "epoch": 0.3368976531037356,
+      "grad_norm": 4.3179207544808955,
+      "learning_rate": 7.727453322180455e-06,
+      "loss": 0.4988,
+      "step": 4852
+    },
+    {
+      "epoch": 0.33696708790445773,
+      "grad_norm": 2.4158164033993463,
+      "learning_rate": 7.726510801394279e-06,
+      "loss": 0.1601,
+      "step": 4853
+    },
+    {
+      "epoch": 0.3370365227051798,
+      "grad_norm": 2.8664915975517427,
+      "learning_rate": 7.725568142704267e-06,
+      "loss": 0.3752,
+      "step": 4854
+    },
+    {
+      "epoch": 0.33710595750590194,
+      "grad_norm": 4.171164121618114,
+      "learning_rate": 7.724625346158105e-06,
+      "loss": 0.4222,
+      "step": 4855
+    },
+    {
+      "epoch": 0.3371753923066241,
+      "grad_norm": 4.223995004071627,
+      "learning_rate": 7.723682411803476e-06,
+      "loss": 0.6088,
+      "step": 4856
+    },
+    {
+      "epoch": 0.3372448271073462,
+      "grad_norm": 3.205020896793061,
+      "learning_rate": 7.722739339688072e-06,
+      "loss": 0.323,
+      "step": 4857
+    },
+    {
+      "epoch": 0.33731426190806835,
+      "grad_norm": 5.280527854329665,
+      "learning_rate": 7.721796129859594e-06,
+      "loss": 0.9124,
+      "step": 4858
+    },
+    {
+      "epoch": 0.3373836967087904,
+      "grad_norm": 3.694301264452014,
+      "learning_rate": 7.720852782365747e-06,
+      "loss": 0.3079,
+      "step": 4859
+    },
+    {
+      "epoch": 0.33745313150951256,
+      "grad_norm": 5.397706369740218,
+      "learning_rate": 7.719909297254244e-06,
+      "loss": 0.5943,
+      "step": 4860
+    },
+    {
+      "epoch": 0.3375225663102347,
+      "grad_norm": 3.546554630183329,
+      "learning_rate": 7.718965674572808e-06,
+      "loss": 0.2817,
+      "step": 4861
+    },
+    {
+      "epoch": 0.33759200111095683,
+      "grad_norm": 3.9138020027590827,
+      "learning_rate": 7.718021914369167e-06,
+      "loss": 0.578,
+      "step": 4862
+    },
+    {
+      "epoch": 0.3376614359116789,
+      "grad_norm": 3.6946986227486502,
+      "learning_rate": 7.71707801669105e-06,
+      "loss": 0.4071,
+      "step": 4863
+    },
+    {
+      "epoch": 0.33773087071240104,
+      "grad_norm": 4.188903698589989,
+      "learning_rate": 7.716133981586202e-06,
+      "loss": 0.6166,
+      "step": 4864
+    },
+    {
+      "epoch": 0.3378003055131232,
+      "grad_norm": 3.056632682117543,
+      "learning_rate": 7.715189809102372e-06,
+      "loss": 0.4631,
+      "step": 4865
+    },
+    {
+      "epoch": 0.3378697403138453,
+      "grad_norm": 4.77857611494246,
+      "learning_rate": 7.714245499287312e-06,
+      "loss": 0.583,
+      "step": 4866
+    },
+    {
+      "epoch": 0.33793917511456745,
+      "grad_norm": 4.429316357593643,
+      "learning_rate": 7.713301052188787e-06,
+      "loss": 0.4545,
+      "step": 4867
+    },
+    {
+      "epoch": 0.3380086099152895,
+      "grad_norm": 3.1279844753637747,
+      "learning_rate": 7.712356467854565e-06,
+      "loss": 0.3131,
+      "step": 4868
+    },
+    {
+      "epoch": 0.33807804471601166,
+      "grad_norm": 4.283684134010977,
+      "learning_rate": 7.711411746332423e-06,
+      "loss": 0.4626,
+      "step": 4869
+    },
+    {
+      "epoch": 0.3381474795167338,
+      "grad_norm": 4.719123702265057,
+      "learning_rate": 7.710466887670141e-06,
+      "loss": 0.7056,
+      "step": 4870
+    },
+    {
+      "epoch": 0.33821691431745593,
+      "grad_norm": 4.072975065691144,
+      "learning_rate": 7.709521891915513e-06,
+      "loss": 0.3944,
+      "step": 4871
+    },
+    {
+      "epoch": 0.338286349118178,
+      "grad_norm": 3.6388609438877166,
+      "learning_rate": 7.708576759116333e-06,
+      "loss": 0.4477,
+      "step": 4872
+    },
+    {
+      "epoch": 0.33835578391890014,
+      "grad_norm": 4.384558956802952,
+      "learning_rate": 7.707631489320407e-06,
+      "loss": 0.6101,
+      "step": 4873
+    },
+    {
+      "epoch": 0.3384252187196223,
+      "grad_norm": 3.2073475885225586,
+      "learning_rate": 7.706686082575541e-06,
+      "loss": 0.3848,
+      "step": 4874
+    },
+    {
+      "epoch": 0.3384946535203444,
+      "grad_norm": 4.954920602307715,
+      "learning_rate": 7.705740538929559e-06,
+      "loss": 0.6687,
+      "step": 4875
+    },
+    {
+      "epoch": 0.3385640883210665,
+      "grad_norm": 26.01129372291356,
+      "learning_rate": 7.70479485843028e-06,
+      "loss": 0.4851,
+      "step": 4876
+    },
+    {
+      "epoch": 0.3386335231217886,
+      "grad_norm": 3.6228150558132475,
+      "learning_rate": 7.703849041125538e-06,
+      "loss": 0.4666,
+      "step": 4877
+    },
+    {
+      "epoch": 0.33870295792251076,
+      "grad_norm": 4.280092026607003,
+      "learning_rate": 7.702903087063171e-06,
+      "loss": 0.7155,
+      "step": 4878
+    },
+    {
+      "epoch": 0.3387723927232329,
+      "grad_norm": 4.969199188899165,
+      "learning_rate": 7.701956996291025e-06,
+      "loss": 0.7843,
+      "step": 4879
+    },
+    {
+      "epoch": 0.33884182752395503,
+      "grad_norm": 3.9546061084672504,
+      "learning_rate": 7.701010768856951e-06,
+      "loss": 0.3964,
+      "step": 4880
+    },
+    {
+      "epoch": 0.3389112623246771,
+      "grad_norm": 3.0491401168525485,
+      "learning_rate": 7.700064404808812e-06,
+      "loss": 0.3486,
+      "step": 4881
+    },
+    {
+      "epoch": 0.33898069712539924,
+      "grad_norm": 3.2153289763417376,
+      "learning_rate": 7.699117904194466e-06,
+      "loss": 0.2309,
+      "step": 4882
+    },
+    {
+      "epoch": 0.3390501319261214,
+      "grad_norm": 3.9754318015786776,
+      "learning_rate": 7.698171267061793e-06,
+      "loss": 0.5292,
+      "step": 4883
+    },
+    {
+      "epoch": 0.3391195667268435,
+      "grad_norm": 4.320458148405599,
+      "learning_rate": 7.69722449345867e-06,
+      "loss": 0.5635,
+      "step": 4884
+    },
+    {
+      "epoch": 0.3391890015275656,
+      "grad_norm": 4.1237640758285234,
+      "learning_rate": 7.696277583432985e-06,
+      "loss": 0.7333,
+      "step": 4885
+    },
+    {
+      "epoch": 0.3392584363282877,
+      "grad_norm": 4.557901532739388,
+      "learning_rate": 7.695330537032629e-06,
+      "loss": 0.6619,
+      "step": 4886
+    },
+    {
+      "epoch": 0.33932787112900986,
+      "grad_norm": 3.0344752806553577,
+      "learning_rate": 7.694383354305505e-06,
+      "loss": 0.2528,
+      "step": 4887
+    },
+    {
+      "epoch": 0.339397305929732,
+      "grad_norm": 3.364559343118975,
+      "learning_rate": 7.693436035299522e-06,
+      "loss": 0.2991,
+      "step": 4888
+    },
+    {
+      "epoch": 0.33946674073045413,
+      "grad_norm": 2.778081617168606,
+      "learning_rate": 7.692488580062589e-06,
+      "loss": 0.2362,
+      "step": 4889
+    },
+    {
+      "epoch": 0.3395361755311762,
+      "grad_norm": 4.1799237534769995,
+      "learning_rate": 7.691540988642635e-06,
+      "loss": 0.4254,
+      "step": 4890
+    },
+    {
+      "epoch": 0.33960561033189834,
+      "grad_norm": 3.5868929205412288,
+      "learning_rate": 7.690593261087579e-06,
+      "loss": 0.3424,
+      "step": 4891
+    },
+    {
+      "epoch": 0.3396750451326205,
+      "grad_norm": 4.652041887723486,
+      "learning_rate": 7.689645397445362e-06,
+      "loss": 0.6067,
+      "step": 4892
+    },
+    {
+      "epoch": 0.3397444799333426,
+      "grad_norm": 4.096387104685421,
+      "learning_rate": 7.688697397763924e-06,
+      "loss": 0.5055,
+      "step": 4893
+    },
+    {
+      "epoch": 0.3398139147340647,
+      "grad_norm": 2.8088127192920456,
+      "learning_rate": 7.687749262091215e-06,
+      "loss": 0.2793,
+      "step": 4894
+    },
+    {
+      "epoch": 0.3398833495347868,
+      "grad_norm": 4.007960699932023,
+      "learning_rate": 7.686800990475189e-06,
+      "loss": 0.6868,
+      "step": 4895
+    },
+    {
+      "epoch": 0.33995278433550896,
+      "grad_norm": 3.373245232530958,
+      "learning_rate": 7.685852582963808e-06,
+      "loss": 0.255,
+      "step": 4896
+    },
+    {
+      "epoch": 0.3400222191362311,
+      "grad_norm": 4.2758864730429345,
+      "learning_rate": 7.684904039605046e-06,
+      "loss": 0.4856,
+      "step": 4897
+    },
+    {
+      "epoch": 0.3400916539369532,
+      "grad_norm": 3.686511076928021,
+      "learning_rate": 7.683955360446873e-06,
+      "loss": 0.4835,
+      "step": 4898
+    },
+    {
+      "epoch": 0.3401610887376753,
+      "grad_norm": 5.084192417096578,
+      "learning_rate": 7.683006545537274e-06,
+      "loss": 0.6442,
+      "step": 4899
+    },
+    {
+      "epoch": 0.34023052353839744,
+      "grad_norm": 4.834501091308603,
+      "learning_rate": 7.682057594924244e-06,
+      "loss": 0.7456,
+      "step": 4900
+    },
+    {
+      "epoch": 0.3402999583391196,
+      "grad_norm": 3.2085964702561247,
+      "learning_rate": 7.681108508655772e-06,
+      "loss": 0.2815,
+      "step": 4901
+    },
+    {
+      "epoch": 0.3403693931398417,
+      "grad_norm": 5.140670939533525,
+      "learning_rate": 7.680159286779866e-06,
+      "loss": 0.9264,
+      "step": 4902
+    },
+    {
+      "epoch": 0.3404388279405638,
+      "grad_norm": 3.8884503924099363,
+      "learning_rate": 7.67920992934454e-06,
+      "loss": 0.5157,
+      "step": 4903
+    },
+    {
+      "epoch": 0.3405082627412859,
+      "grad_norm": 3.666289792784984,
+      "learning_rate": 7.678260436397805e-06,
+      "loss": 0.4411,
+      "step": 4904
+    },
+    {
+      "epoch": 0.34057769754200806,
+      "grad_norm": 3.6330060998524103,
+      "learning_rate": 7.677310807987685e-06,
+      "loss": 0.4306,
+      "step": 4905
+    },
+    {
+      "epoch": 0.3406471323427302,
+      "grad_norm": 2.1483564135560993,
+      "learning_rate": 7.676361044162217e-06,
+      "loss": 0.141,
+      "step": 4906
+    },
+    {
+      "epoch": 0.3407165671434523,
+      "grad_norm": 4.252595432998503,
+      "learning_rate": 7.675411144969436e-06,
+      "loss": 0.591,
+      "step": 4907
+    },
+    {
+      "epoch": 0.3407860019441744,
+      "grad_norm": 4.761579563032916,
+      "learning_rate": 7.674461110457385e-06,
+      "loss": 0.6819,
+      "step": 4908
+    },
+    {
+      "epoch": 0.34085543674489654,
+      "grad_norm": 3.286738879183506,
+      "learning_rate": 7.673510940674118e-06,
+      "loss": 0.3453,
+      "step": 4909
+    },
+    {
+      "epoch": 0.3409248715456187,
+      "grad_norm": 4.109525049249137,
+      "learning_rate": 7.672560635667694e-06,
+      "loss": 0.4241,
+      "step": 4910
+    },
+    {
+      "epoch": 0.34099430634634076,
+      "grad_norm": 3.7447246825401264,
+      "learning_rate": 7.671610195486177e-06,
+      "loss": 0.3232,
+      "step": 4911
+    },
+    {
+      "epoch": 0.3410637411470629,
+      "grad_norm": 4.146407333780971,
+      "learning_rate": 7.670659620177638e-06,
+      "loss": 0.5391,
+      "step": 4912
+    },
+    {
+      "epoch": 0.34113317594778503,
+      "grad_norm": 3.648471926251393,
+      "learning_rate": 7.66970890979016e-06,
+      "loss": 0.2495,
+      "step": 4913
+    },
+    {
+      "epoch": 0.34120261074850716,
+      "grad_norm": 4.485083472890363,
+      "learning_rate": 7.668758064371824e-06,
+      "loss": 0.4666,
+      "step": 4914
+    },
+    {
+      "epoch": 0.3412720455492293,
+      "grad_norm": 4.238565405560386,
+      "learning_rate": 7.667807083970727e-06,
+      "loss": 0.4392,
+      "step": 4915
+    },
+    {
+      "epoch": 0.3413414803499514,
+      "grad_norm": 4.671663469781545,
+      "learning_rate": 7.666855968634964e-06,
+      "loss": 0.5653,
+      "step": 4916
+    },
+    {
+      "epoch": 0.3414109151506735,
+      "grad_norm": 4.11868155642245,
+      "learning_rate": 7.665904718412646e-06,
+      "loss": 0.579,
+      "step": 4917
+    },
+    {
+      "epoch": 0.34148034995139565,
+      "grad_norm": 4.192327929046221,
+      "learning_rate": 7.664953333351882e-06,
+      "loss": 0.2979,
+      "step": 4918
+    },
+    {
+      "epoch": 0.3415497847521178,
+      "grad_norm": 3.5876036510924116,
+      "learning_rate": 7.664001813500794e-06,
+      "loss": 0.5838,
+      "step": 4919
+    },
+    {
+      "epoch": 0.34161921955283986,
+      "grad_norm": 3.9145969710148454,
+      "learning_rate": 7.66305015890751e-06,
+      "loss": 0.613,
+      "step": 4920
+    },
+    {
+      "epoch": 0.341688654353562,
+      "grad_norm": 4.9276723599203756,
+      "learning_rate": 7.662098369620164e-06,
+      "loss": 0.5713,
+      "step": 4921
+    },
+    {
+      "epoch": 0.34175808915428413,
+      "grad_norm": 3.968461559555649,
+      "learning_rate": 7.661146445686893e-06,
+      "loss": 0.6693,
+      "step": 4922
+    },
+    {
+      "epoch": 0.34182752395500626,
+      "grad_norm": 3.9380937338729782,
+      "learning_rate": 7.660194387155846e-06,
+      "loss": 0.4176,
+      "step": 4923
+    },
+    {
+      "epoch": 0.3418969587557284,
+      "grad_norm": 2.9929718911102867,
+      "learning_rate": 7.659242194075178e-06,
+      "loss": 0.2875,
+      "step": 4924
+    },
+    {
+      "epoch": 0.3419663935564505,
+      "grad_norm": 6.360276240506255,
+      "learning_rate": 7.658289866493047e-06,
+      "loss": 0.604,
+      "step": 4925
+    },
+    {
+      "epoch": 0.3420358283571726,
+      "grad_norm": 3.9855764884951603,
+      "learning_rate": 7.657337404457626e-06,
+      "loss": 0.5393,
+      "step": 4926
+    },
+    {
+      "epoch": 0.34210526315789475,
+      "grad_norm": 4.775029293113738,
+      "learning_rate": 7.656384808017083e-06,
+      "loss": 0.7523,
+      "step": 4927
+    },
+    {
+      "epoch": 0.3421746979586169,
+      "grad_norm": 3.2276019636249687,
+      "learning_rate": 7.655432077219602e-06,
+      "loss": 0.438,
+      "step": 4928
+    },
+    {
+      "epoch": 0.34224413275933896,
+      "grad_norm": 2.929945487154513,
+      "learning_rate": 7.654479212113373e-06,
+      "loss": 0.2248,
+      "step": 4929
+    },
+    {
+      "epoch": 0.3423135675600611,
+      "grad_norm": 3.0992219229547033,
+      "learning_rate": 7.653526212746589e-06,
+      "loss": 0.227,
+      "step": 4930
+    },
+    {
+      "epoch": 0.34238300236078323,
+      "grad_norm": 3.380113063638498,
+      "learning_rate": 7.652573079167453e-06,
+      "loss": 0.466,
+      "step": 4931
+    },
+    {
+      "epoch": 0.34245243716150536,
+      "grad_norm": 3.4433183463704102,
+      "learning_rate": 7.65161981142417e-06,
+      "loss": 0.4475,
+      "step": 4932
+    },
+    {
+      "epoch": 0.34252187196222744,
+      "grad_norm": 3.166585173445886,
+      "learning_rate": 7.650666409564958e-06,
+      "loss": 0.2795,
+      "step": 4933
+    },
+    {
+      "epoch": 0.3425913067629496,
+      "grad_norm": 3.271855559538518,
+      "learning_rate": 7.649712873638039e-06,
+      "loss": 0.3896,
+      "step": 4934
+    },
+    {
+      "epoch": 0.3426607415636717,
+      "grad_norm": 4.860060135648331,
+      "learning_rate": 7.64875920369164e-06,
+      "loss": 0.6546,
+      "step": 4935
+    },
+    {
+      "epoch": 0.34273017636439385,
+      "grad_norm": 3.0421181789493708,
+      "learning_rate": 7.647805399774e-06,
+      "loss": 0.3974,
+      "step": 4936
+    },
+    {
+      "epoch": 0.342799611165116,
+      "grad_norm": 4.166106379730964,
+      "learning_rate": 7.646851461933357e-06,
+      "loss": 0.6803,
+      "step": 4937
+    },
+    {
+      "epoch": 0.34286904596583806,
+      "grad_norm": 4.351358842219458,
+      "learning_rate": 7.645897390217964e-06,
+      "loss": 0.3591,
+      "step": 4938
+    },
+    {
+      "epoch": 0.3429384807665602,
+      "grad_norm": 3.7871030000973946,
+      "learning_rate": 7.644943184676074e-06,
+      "loss": 0.4997,
+      "step": 4939
+    },
+    {
+      "epoch": 0.34300791556728233,
+      "grad_norm": 4.31251538164162,
+      "learning_rate": 7.643988845355951e-06,
+      "loss": 0.6772,
+      "step": 4940
+    },
+    {
+      "epoch": 0.34307735036800446,
+      "grad_norm": 2.657027498019386,
+      "learning_rate": 7.643034372305864e-06,
+      "loss": 0.266,
+      "step": 4941
+    },
+    {
+      "epoch": 0.34314678516872654,
+      "grad_norm": 3.3323060956473114,
+      "learning_rate": 7.642079765574089e-06,
+      "loss": 0.3742,
+      "step": 4942
+    },
+    {
+      "epoch": 0.3432162199694487,
+      "grad_norm": 3.377416305664132,
+      "learning_rate": 7.641125025208909e-06,
+      "loss": 0.4202,
+      "step": 4943
+    },
+    {
+      "epoch": 0.3432856547701708,
+      "grad_norm": 2.924600904250069,
+      "learning_rate": 7.640170151258614e-06,
+      "loss": 0.2127,
+      "step": 4944
+    },
+    {
+      "epoch": 0.34335508957089295,
+      "grad_norm": 4.4982830758182155,
+      "learning_rate": 7.6392151437715e-06,
+      "loss": 0.8252,
+      "step": 4945
+    },
+    {
+      "epoch": 0.3434245243716151,
+      "grad_norm": 3.5556512517355694,
+      "learning_rate": 7.638260002795871e-06,
+      "loss": 0.383,
+      "step": 4946
+    },
+    {
+      "epoch": 0.34349395917233716,
+      "grad_norm": 4.193044201234624,
+      "learning_rate": 7.637304728380036e-06,
+      "loss": 0.5593,
+      "step": 4947
+    },
+    {
+      "epoch": 0.3435633939730593,
+      "grad_norm": 4.326857867834154,
+      "learning_rate": 7.636349320572313e-06,
+      "loss": 0.7883,
+      "step": 4948
+    },
+    {
+      "epoch": 0.34363282877378143,
+      "grad_norm": 4.667291806371995,
+      "learning_rate": 7.635393779421025e-06,
+      "loss": 0.7982,
+      "step": 4949
+    },
+    {
+      "epoch": 0.34370226357450356,
+      "grad_norm": 3.645378722455744,
+      "learning_rate": 7.634438104974499e-06,
+      "loss": 0.4164,
+      "step": 4950
+    },
+    {
+      "epoch": 0.34377169837522564,
+      "grad_norm": 4.403122593060269,
+      "learning_rate": 7.633482297281078e-06,
+      "loss": 0.48,
+      "step": 4951
+    },
+    {
+      "epoch": 0.3438411331759478,
+      "grad_norm": 3.4079372096549374,
+      "learning_rate": 7.6325263563891e-06,
+      "loss": 0.2931,
+      "step": 4952
+    },
+    {
+      "epoch": 0.3439105679766699,
+      "grad_norm": 4.853677766783322,
+      "learning_rate": 7.631570282346918e-06,
+      "loss": 0.7206,
+      "step": 4953
+    },
+    {
+      "epoch": 0.34398000277739205,
+      "grad_norm": 3.4232188042669653,
+      "learning_rate": 7.63061407520289e-06,
+      "loss": 0.454,
+      "step": 4954
+    },
+    {
+      "epoch": 0.3440494375781141,
+      "grad_norm": 3.2896641669605104,
+      "learning_rate": 7.629657735005381e-06,
+      "loss": 0.2678,
+      "step": 4955
+    },
+    {
+      "epoch": 0.34411887237883626,
+      "grad_norm": 4.090866257333853,
+      "learning_rate": 7.628701261802756e-06,
+      "loss": 0.4528,
+      "step": 4956
+    },
+    {
+      "epoch": 0.3441883071795584,
+      "grad_norm": 3.860838640162453,
+      "learning_rate": 7.627744655643396e-06,
+      "loss": 0.527,
+      "step": 4957
+    },
+    {
+      "epoch": 0.34425774198028053,
+      "grad_norm": 4.2395130551207565,
+      "learning_rate": 7.626787916575686e-06,
+      "loss": 0.5153,
+      "step": 4958
+    },
+    {
+      "epoch": 0.34432717678100266,
+      "grad_norm": 2.832207217493252,
+      "learning_rate": 7.625831044648014e-06,
+      "loss": 0.3489,
+      "step": 4959
+    },
+    {
+      "epoch": 0.34439661158172474,
+      "grad_norm": 5.1999940724480735,
+      "learning_rate": 7.624874039908781e-06,
+      "loss": 0.578,
+      "step": 4960
+    },
+    {
+      "epoch": 0.3444660463824469,
+      "grad_norm": 4.435956837479547,
+      "learning_rate": 7.623916902406388e-06,
+      "loss": 0.7175,
+      "step": 4961
+    },
+    {
+      "epoch": 0.344535481183169,
+      "grad_norm": 4.744054261416868,
+      "learning_rate": 7.622959632189248e-06,
+      "loss": 0.7576,
+      "step": 4962
+    },
+    {
+      "epoch": 0.34460491598389115,
+      "grad_norm": 3.9423970598658853,
+      "learning_rate": 7.6220022293057784e-06,
+      "loss": 0.4735,
+      "step": 4963
+    },
+    {
+      "epoch": 0.3446743507846132,
+      "grad_norm": 4.504747855627425,
+      "learning_rate": 7.6210446938044016e-06,
+      "loss": 0.7768,
+      "step": 4964
+    },
+    {
+      "epoch": 0.34474378558533536,
+      "grad_norm": 4.186346682172334,
+      "learning_rate": 7.620087025733552e-06,
+      "loss": 0.5777,
+      "step": 4965
+    },
+    {
+      "epoch": 0.3448132203860575,
+      "grad_norm": 3.3698277111604957,
+      "learning_rate": 7.619129225141664e-06,
+      "loss": 0.4934,
+      "step": 4966
+    },
+    {
+      "epoch": 0.34488265518677963,
+      "grad_norm": 3.4966782911616705,
+      "learning_rate": 7.618171292077183e-06,
+      "loss": 0.3665,
+      "step": 4967
+    },
+    {
+      "epoch": 0.3449520899875017,
+      "grad_norm": 4.033482080154214,
+      "learning_rate": 7.617213226588561e-06,
+      "loss": 0.7563,
+      "step": 4968
+    },
+    {
+      "epoch": 0.34502152478822384,
+      "grad_norm": 3.298472226325857,
+      "learning_rate": 7.6162550287242565e-06,
+      "loss": 0.4609,
+      "step": 4969
+    },
+    {
+      "epoch": 0.345090959588946,
+      "grad_norm": 2.6373381605452764,
+      "learning_rate": 7.615296698532732e-06,
+      "loss": 0.298,
+      "step": 4970
+    },
+    {
+      "epoch": 0.3451603943896681,
+      "grad_norm": 3.358468531528095,
+      "learning_rate": 7.614338236062462e-06,
+      "loss": 0.4216,
+      "step": 4971
+    },
+    {
+      "epoch": 0.34522982919039025,
+      "grad_norm": 3.5707197631501284,
+      "learning_rate": 7.613379641361921e-06,
+      "loss": 0.3622,
+      "step": 4972
+    },
+    {
+      "epoch": 0.3452992639911123,
+      "grad_norm": 4.567646691117784,
+      "learning_rate": 7.612420914479595e-06,
+      "loss": 0.5992,
+      "step": 4973
+    },
+    {
+      "epoch": 0.34536869879183446,
+      "grad_norm": 4.270655759432465,
+      "learning_rate": 7.611462055463974e-06,
+      "loss": 0.5079,
+      "step": 4974
+    },
+    {
+      "epoch": 0.3454381335925566,
+      "grad_norm": 3.284226924917533,
+      "learning_rate": 7.610503064363559e-06,
+      "loss": 0.3587,
+      "step": 4975
+    },
+    {
+      "epoch": 0.34550756839327873,
+      "grad_norm": 4.000699191645003,
+      "learning_rate": 7.609543941226852e-06,
+      "loss": 0.529,
+      "step": 4976
+    },
+    {
+      "epoch": 0.3455770031940008,
+      "grad_norm": 2.919561897917775,
+      "learning_rate": 7.608584686102367e-06,
+      "loss": 0.1985,
+      "step": 4977
+    },
+    {
+      "epoch": 0.34564643799472294,
+      "grad_norm": 3.376862122032198,
+      "learning_rate": 7.607625299038619e-06,
+      "loss": 0.462,
+      "step": 4978
+    },
+    {
+      "epoch": 0.3457158727954451,
+      "grad_norm": 4.0853830925223145,
+      "learning_rate": 7.606665780084134e-06,
+      "loss": 0.3814,
+      "step": 4979
+    },
+    {
+      "epoch": 0.3457853075961672,
+      "grad_norm": 3.2076017186018198,
+      "learning_rate": 7.6057061292874436e-06,
+      "loss": 0.3978,
+      "step": 4980
+    },
+    {
+      "epoch": 0.34585474239688935,
+      "grad_norm": 2.702879077998113,
+      "learning_rate": 7.604746346697087e-06,
+      "loss": 0.3163,
+      "step": 4981
+    },
+    {
+      "epoch": 0.3459241771976114,
+      "grad_norm": 3.486118826670658,
+      "learning_rate": 7.603786432361606e-06,
+      "loss": 0.5266,
+      "step": 4982
+    },
+    {
+      "epoch": 0.34599361199833356,
+      "grad_norm": 3.155462987189284,
+      "learning_rate": 7.6028263863295535e-06,
+      "loss": 0.2953,
+      "step": 4983
+    },
+    {
+      "epoch": 0.3460630467990557,
+      "grad_norm": 5.789916715940687,
+      "learning_rate": 7.601866208649489e-06,
+      "loss": 1.0424,
+      "step": 4984
+    },
+    {
+      "epoch": 0.34613248159977783,
+      "grad_norm": 2.326342116146704,
+      "learning_rate": 7.600905899369976e-06,
+      "loss": 0.1081,
+      "step": 4985
+    },
+    {
+      "epoch": 0.3462019164004999,
+      "grad_norm": 3.308308103644391,
+      "learning_rate": 7.599945458539584e-06,
+      "loss": 0.4041,
+      "step": 4986
+    },
+    {
+      "epoch": 0.34627135120122204,
+      "grad_norm": 9.524943640380164,
+      "learning_rate": 7.598984886206895e-06,
+      "loss": 0.4442,
+      "step": 4987
+    },
+    {
+      "epoch": 0.3463407860019442,
+      "grad_norm": 3.2353082029263613,
+      "learning_rate": 7.598024182420491e-06,
+      "loss": 0.454,
+      "step": 4988
+    },
+    {
+      "epoch": 0.3464102208026663,
+      "grad_norm": 4.20705455551874,
+      "learning_rate": 7.597063347228961e-06,
+      "loss": 0.7184,
+      "step": 4989
+    },
+    {
+      "epoch": 0.3464796556033884,
+      "grad_norm": 4.266489095068036,
+      "learning_rate": 7.5961023806809075e-06,
+      "loss": 0.7042,
+      "step": 4990
+    },
+    {
+      "epoch": 0.3465490904041105,
+      "grad_norm": 3.9716891467774142,
+      "learning_rate": 7.595141282824934e-06,
+      "loss": 0.3931,
+      "step": 4991
+    },
+    {
+      "epoch": 0.34661852520483266,
+      "grad_norm": 3.8347544817365398,
+      "learning_rate": 7.59418005370965e-06,
+      "loss": 0.3965,
+      "step": 4992
+    },
+    {
+      "epoch": 0.3466879600055548,
+      "grad_norm": 3.516202644540482,
+      "learning_rate": 7.593218693383675e-06,
+      "loss": 0.378,
+      "step": 4993
+    },
+    {
+      "epoch": 0.34675739480627693,
+      "grad_norm": 4.101569105223676,
+      "learning_rate": 7.592257201895633e-06,
+      "loss": 0.4947,
+      "step": 4994
+    },
+    {
+      "epoch": 0.346826829606999,
+      "grad_norm": 3.203913055585752,
+      "learning_rate": 7.591295579294153e-06,
+      "loss": 0.2467,
+      "step": 4995
+    },
+    {
+      "epoch": 0.34689626440772114,
+      "grad_norm": 3.185677118317719,
+      "learning_rate": 7.590333825627878e-06,
+      "loss": 0.4108,
+      "step": 4996
+    },
+    {
+      "epoch": 0.3469656992084433,
+      "grad_norm": 3.1309434348068663,
+      "learning_rate": 7.589371940945449e-06,
+      "loss": 0.2656,
+      "step": 4997
+    },
+    {
+      "epoch": 0.3470351340091654,
+      "grad_norm": 3.9095524796559755,
+      "learning_rate": 7.588409925295515e-06,
+      "loss": 0.488,
+      "step": 4998
+    },
+    {
+      "epoch": 0.3471045688098875,
+      "grad_norm": 3.6282229472152787,
+      "learning_rate": 7.587447778726737e-06,
+      "loss": 0.3676,
+      "step": 4999
+    },
+    {
+      "epoch": 0.3471740036106096,
+      "grad_norm": 3.087622169003717,
+      "learning_rate": 7.586485501287779e-06,
+      "loss": 0.2439,
+      "step": 5000
+    },
+    {
+      "epoch": 0.34724343841133176,
+      "grad_norm": 4.809241947674745,
+      "learning_rate": 7.58552309302731e-06,
+      "loss": 0.4697,
+      "step": 5001
+    },
+    {
+      "epoch": 0.3473128732120539,
+      "grad_norm": 3.6147559490717263,
+      "learning_rate": 7.584560553994009e-06,
+      "loss": 0.6059,
+      "step": 5002
+    },
+    {
+      "epoch": 0.34738230801277603,
+      "grad_norm": 3.3760245492471146,
+      "learning_rate": 7.583597884236561e-06,
+      "loss": 0.3014,
+      "step": 5003
+    },
+    {
+      "epoch": 0.3474517428134981,
+      "grad_norm": 3.679397960901014,
+      "learning_rate": 7.582635083803656e-06,
+      "loss": 0.4794,
+      "step": 5004
+    },
+    {
+      "epoch": 0.34752117761422024,
+      "grad_norm": 4.145391730854627,
+      "learning_rate": 7.5816721527439875e-06,
+      "loss": 0.5993,
+      "step": 5005
+    },
+    {
+      "epoch": 0.3475906124149424,
+      "grad_norm": 6.34470076593736,
+      "learning_rate": 7.5807090911062655e-06,
+      "loss": 0.9443,
+      "step": 5006
+    },
+    {
+      "epoch": 0.3476600472156645,
+      "grad_norm": 3.2650439849410966,
+      "learning_rate": 7.579745898939197e-06,
+      "loss": 0.3352,
+      "step": 5007
+    },
+    {
+      "epoch": 0.3477294820163866,
+      "grad_norm": 3.1197404095547934,
+      "learning_rate": 7.578782576291501e-06,
+      "loss": 0.3294,
+      "step": 5008
+    },
+    {
+      "epoch": 0.3477989168171087,
+      "grad_norm": 4.5173950421168385,
+      "learning_rate": 7.577819123211898e-06,
+      "loss": 0.5264,
+      "step": 5009
+    },
+    {
+      "epoch": 0.34786835161783086,
+      "grad_norm": 3.9040792945982354,
+      "learning_rate": 7.5768555397491236e-06,
+      "loss": 0.4207,
+      "step": 5010
+    },
+    {
+      "epoch": 0.347937786418553,
+      "grad_norm": 3.470578091050898,
+      "learning_rate": 7.575891825951911e-06,
+      "loss": 0.4245,
+      "step": 5011
+    },
+    {
+      "epoch": 0.3480072212192751,
+      "grad_norm": 2.7550899871253955,
+      "learning_rate": 7.574927981869005e-06,
+      "loss": 0.2478,
+      "step": 5012
+    },
+    {
+      "epoch": 0.3480766560199972,
+      "grad_norm": 3.8536817438992768,
+      "learning_rate": 7.5739640075491546e-06,
+      "loss": 0.6882,
+      "step": 5013
+    },
+    {
+      "epoch": 0.34814609082071934,
+      "grad_norm": 3.7317183606845536,
+      "learning_rate": 7.572999903041117e-06,
+      "loss": 0.4475,
+      "step": 5014
+    },
+    {
+      "epoch": 0.3482155256214415,
+      "grad_norm": 3.3051209417986795,
+      "learning_rate": 7.572035668393657e-06,
+      "loss": 0.3982,
+      "step": 5015
+    },
+    {
+      "epoch": 0.3482849604221636,
+      "grad_norm": 3.3081172837977055,
+      "learning_rate": 7.5710713036555425e-06,
+      "loss": 0.5134,
+      "step": 5016
+    },
+    {
+      "epoch": 0.3483543952228857,
+      "grad_norm": 3.9990490961232554,
+      "learning_rate": 7.570106808875552e-06,
+      "loss": 0.5308,
+      "step": 5017
+    },
+    {
+      "epoch": 0.34842383002360783,
+      "grad_norm": 3.7941985634494055,
+      "learning_rate": 7.5691421841024645e-06,
+      "loss": 0.5223,
+      "step": 5018
+    },
+    {
+      "epoch": 0.34849326482432996,
+      "grad_norm": 3.4464714701542642,
+      "learning_rate": 7.5681774293850756e-06,
+      "loss": 0.471,
+      "step": 5019
+    },
+    {
+      "epoch": 0.3485626996250521,
+      "grad_norm": 3.6754481072251006,
+      "learning_rate": 7.567212544772178e-06,
+      "loss": 0.3892,
+      "step": 5020
+    },
+    {
+      "epoch": 0.3486321344257742,
+      "grad_norm": 3.9245735878475654,
+      "learning_rate": 7.566247530312574e-06,
+      "loss": 0.4625,
+      "step": 5021
+    },
+    {
+      "epoch": 0.3487015692264963,
+      "grad_norm": 3.691761014546558,
+      "learning_rate": 7.565282386055075e-06,
+      "loss": 0.5374,
+      "step": 5022
+    },
+    {
+      "epoch": 0.34877100402721845,
+      "grad_norm": 3.844108036938108,
+      "learning_rate": 7.5643171120484935e-06,
+      "loss": 0.5432,
+      "step": 5023
+    },
+    {
+      "epoch": 0.3488404388279406,
+      "grad_norm": 3.5590915042686344,
+      "learning_rate": 7.563351708341657e-06,
+      "loss": 0.6736,
+      "step": 5024
+    },
+    {
+      "epoch": 0.34890987362866266,
+      "grad_norm": 3.6418640736255856,
+      "learning_rate": 7.562386174983389e-06,
+      "loss": 0.4483,
+      "step": 5025
+    },
+    {
+      "epoch": 0.3489793084293848,
+      "grad_norm": 3.810707503487819,
+      "learning_rate": 7.56142051202253e-06,
+      "loss": 0.6407,
+      "step": 5026
+    },
+    {
+      "epoch": 0.34904874323010693,
+      "grad_norm": 5.125669805091938,
+      "learning_rate": 7.560454719507918e-06,
+      "loss": 0.6929,
+      "step": 5027
+    },
+    {
+      "epoch": 0.34911817803082906,
+      "grad_norm": 3.028233717407423,
+      "learning_rate": 7.559488797488404e-06,
+      "loss": 0.2888,
+      "step": 5028
+    },
+    {
+      "epoch": 0.3491876128315512,
+      "grad_norm": 3.728650296358148,
+      "learning_rate": 7.558522746012845e-06,
+      "loss": 0.4351,
+      "step": 5029
+    },
+    {
+      "epoch": 0.3492570476322733,
+      "grad_norm": 3.227338174062697,
+      "learning_rate": 7.5575565651300985e-06,
+      "loss": 0.5075,
+      "step": 5030
+    },
+    {
+      "epoch": 0.3493264824329954,
+      "grad_norm": 3.657684218138805,
+      "learning_rate": 7.556590254889036e-06,
+      "loss": 0.4699,
+      "step": 5031
+    },
+    {
+      "epoch": 0.34939591723371755,
+      "grad_norm": 3.678752571902783,
+      "learning_rate": 7.555623815338531e-06,
+      "loss": 0.449,
+      "step": 5032
+    },
+    {
+      "epoch": 0.3494653520344397,
+      "grad_norm": 3.9535238145747975,
+      "learning_rate": 7.554657246527466e-06,
+      "loss": 0.3857,
+      "step": 5033
+    },
+    {
+      "epoch": 0.34953478683516176,
+      "grad_norm": 2.433561894281368,
+      "learning_rate": 7.553690548504726e-06,
+      "loss": 0.242,
+      "step": 5034
+    },
+    {
+      "epoch": 0.3496042216358839,
+      "grad_norm": 3.589812847724724,
+      "learning_rate": 7.552723721319211e-06,
+      "loss": 0.3105,
+      "step": 5035
+    },
+    {
+      "epoch": 0.34967365643660603,
+      "grad_norm": 3.7792969099608404,
+      "learning_rate": 7.5517567650198155e-06,
+      "loss": 0.3774,
+      "step": 5036
+    },
+    {
+      "epoch": 0.34974309123732816,
+      "grad_norm": 3.4980878278701226,
+      "learning_rate": 7.550789679655453e-06,
+      "loss": 0.379,
+      "step": 5037
+    },
+    {
+      "epoch": 0.3498125260380503,
+      "grad_norm": 3.5502703084386806,
+      "learning_rate": 7.549822465275034e-06,
+      "loss": 0.3263,
+      "step": 5038
+    },
+    {
+      "epoch": 0.3498819608387724,
+      "grad_norm": 4.333475887675362,
+      "learning_rate": 7.54885512192748e-06,
+      "loss": 0.5408,
+      "step": 5039
+    },
+    {
+      "epoch": 0.3499513956394945,
+      "grad_norm": 3.442654049145105,
+      "learning_rate": 7.547887649661718e-06,
+      "loss": 0.1839,
+      "step": 5040
+    },
+    {
+      "epoch": 0.35002083044021665,
+      "grad_norm": 3.1848949796924795,
+      "learning_rate": 7.546920048526681e-06,
+      "loss": 0.3377,
+      "step": 5041
+    },
+    {
+      "epoch": 0.3500902652409388,
+      "grad_norm": 6.289694110036782,
+      "learning_rate": 7.545952318571312e-06,
+      "loss": 0.6819,
+      "step": 5042
+    },
+    {
+      "epoch": 0.35015970004166086,
+      "grad_norm": 3.0939061190239823,
+      "learning_rate": 7.544984459844553e-06,
+      "loss": 0.3019,
+      "step": 5043
+    },
+    {
+      "epoch": 0.350229134842383,
+      "grad_norm": 3.7466443840417316,
+      "learning_rate": 7.5440164723953625e-06,
+      "loss": 0.4437,
+      "step": 5044
+    },
+    {
+      "epoch": 0.35029856964310513,
+      "grad_norm": 5.217260454900235,
+      "learning_rate": 7.5430483562726964e-06,
+      "loss": 0.8,
+      "step": 5045
+    },
+    {
+      "epoch": 0.35036800444382726,
+      "grad_norm": 4.098974076584689,
+      "learning_rate": 7.542080111525523e-06,
+      "loss": 0.4995,
+      "step": 5046
+    },
+    {
+      "epoch": 0.35043743924454934,
+      "grad_norm": 4.532379127037582,
+      "learning_rate": 7.541111738202813e-06,
+      "loss": 0.3306,
+      "step": 5047
+    },
+    {
+      "epoch": 0.3505068740452715,
+      "grad_norm": 3.1755955248933696,
+      "learning_rate": 7.540143236353548e-06,
+      "loss": 0.2972,
+      "step": 5048
+    },
+    {
+      "epoch": 0.3505763088459936,
+      "grad_norm": 3.5783267316309493,
+      "learning_rate": 7.539174606026711e-06,
+      "loss": 0.4031,
+      "step": 5049
+    },
+    {
+      "epoch": 0.35064574364671575,
+      "grad_norm": 2.679469539207239,
+      "learning_rate": 7.538205847271295e-06,
+      "loss": 0.1788,
+      "step": 5050
+    },
+    {
+      "epoch": 0.3507151784474379,
+      "grad_norm": 5.226601990932275,
+      "learning_rate": 7.5372369601363e-06,
+      "loss": 0.8277,
+      "step": 5051
+    },
+    {
+      "epoch": 0.35078461324815996,
+      "grad_norm": 3.3555124295917755,
+      "learning_rate": 7.536267944670732e-06,
+      "loss": 0.4122,
+      "step": 5052
+    },
+    {
+      "epoch": 0.3508540480488821,
+      "grad_norm": 3.96976715454651,
+      "learning_rate": 7.535298800923601e-06,
+      "loss": 0.3753,
+      "step": 5053
+    },
+    {
+      "epoch": 0.35092348284960423,
+      "grad_norm": 3.558249892869726,
+      "learning_rate": 7.534329528943925e-06,
+      "loss": 0.3054,
+      "step": 5054
+    },
+    {
+      "epoch": 0.35099291765032636,
+      "grad_norm": 2.9080808699755023,
+      "learning_rate": 7.533360128780729e-06,
+      "loss": 0.3005,
+      "step": 5055
+    },
+    {
+      "epoch": 0.35106235245104844,
+      "grad_norm": 2.6411945128578,
+      "learning_rate": 7.532390600483046e-06,
+      "loss": 0.1862,
+      "step": 5056
+    },
+    {
+      "epoch": 0.3511317872517706,
+      "grad_norm": 3.528977537067989,
+      "learning_rate": 7.531420944099909e-06,
+      "loss": 0.4028,
+      "step": 5057
+    },
+    {
+      "epoch": 0.3512012220524927,
+      "grad_norm": 6.73408172114143,
+      "learning_rate": 7.530451159680369e-06,
+      "loss": 0.5786,
+      "step": 5058
+    },
+    {
+      "epoch": 0.35127065685321485,
+      "grad_norm": 3.3743823596867846,
+      "learning_rate": 7.529481247273469e-06,
+      "loss": 0.3801,
+      "step": 5059
+    },
+    {
+      "epoch": 0.3513400916539369,
+      "grad_norm": 3.363252840984427,
+      "learning_rate": 7.528511206928272e-06,
+      "loss": 0.3785,
+      "step": 5060
+    },
+    {
+      "epoch": 0.35140952645465906,
+      "grad_norm": 5.028617899263473,
+      "learning_rate": 7.527541038693839e-06,
+      "loss": 0.7898,
+      "step": 5061
+    },
+    {
+      "epoch": 0.3514789612553812,
+      "grad_norm": 3.8429369821287334,
+      "learning_rate": 7.52657074261924e-06,
+      "loss": 0.3131,
+      "step": 5062
+    },
+    {
+      "epoch": 0.35154839605610333,
+      "grad_norm": 4.255881097598719,
+      "learning_rate": 7.525600318753552e-06,
+      "loss": 0.3449,
+      "step": 5063
+    },
+    {
+      "epoch": 0.35161783085682546,
+      "grad_norm": 5.109844173204817,
+      "learning_rate": 7.524629767145858e-06,
+      "loss": 0.6087,
+      "step": 5064
+    },
+    {
+      "epoch": 0.35168726565754754,
+      "grad_norm": 3.332935811236292,
+      "learning_rate": 7.523659087845247e-06,
+      "loss": 0.3957,
+      "step": 5065
+    },
+    {
+      "epoch": 0.3517567004582697,
+      "grad_norm": 3.849874430905618,
+      "learning_rate": 7.522688280900814e-06,
+      "loss": 0.3064,
+      "step": 5066
+    },
+    {
+      "epoch": 0.3518261352589918,
+      "grad_norm": 3.9627379927505966,
+      "learning_rate": 7.521717346361663e-06,
+      "loss": 0.5379,
+      "step": 5067
+    },
+    {
+      "epoch": 0.35189557005971395,
+      "grad_norm": 4.115828743175035,
+      "learning_rate": 7.520746284276903e-06,
+      "loss": 0.6845,
+      "step": 5068
+    },
+    {
+      "epoch": 0.351965004860436,
+      "grad_norm": 3.4220087497853826,
+      "learning_rate": 7.51977509469565e-06,
+      "loss": 0.4625,
+      "step": 5069
+    },
+    {
+      "epoch": 0.35203443966115816,
+      "grad_norm": 4.2939019877036655,
+      "learning_rate": 7.518803777667019e-06,
+      "loss": 0.3672,
+      "step": 5070
+    },
+    {
+      "epoch": 0.3521038744618803,
+      "grad_norm": 3.707428777977297,
+      "learning_rate": 7.517832333240148e-06,
+      "loss": 0.4247,
+      "step": 5071
+    },
+    {
+      "epoch": 0.35217330926260243,
+      "grad_norm": 4.580632139232148,
+      "learning_rate": 7.516860761464164e-06,
+      "loss": 0.5729,
+      "step": 5072
+    },
+    {
+      "epoch": 0.35224274406332456,
+      "grad_norm": 4.026010416083216,
+      "learning_rate": 7.515889062388212e-06,
+      "loss": 0.3215,
+      "step": 5073
+    },
+    {
+      "epoch": 0.35231217886404664,
+      "grad_norm": 3.9329882551314803,
+      "learning_rate": 7.514917236061439e-06,
+      "loss": 0.4484,
+      "step": 5074
+    },
+    {
+      "epoch": 0.3523816136647688,
+      "grad_norm": 3.9763464010319183,
+      "learning_rate": 7.513945282532995e-06,
+      "loss": 0.5524,
+      "step": 5075
+    },
+    {
+      "epoch": 0.3524510484654909,
+      "grad_norm": 4.432501531128317,
+      "learning_rate": 7.512973201852045e-06,
+      "loss": 0.2751,
+      "step": 5076
+    },
+    {
+      "epoch": 0.35252048326621305,
+      "grad_norm": 2.858450525586634,
+      "learning_rate": 7.512000994067754e-06,
+      "loss": 0.4788,
+      "step": 5077
+    },
+    {
+      "epoch": 0.3525899180669351,
+      "grad_norm": 3.5100143463642364,
+      "learning_rate": 7.511028659229296e-06,
+      "loss": 0.6,
+      "step": 5078
+    },
+    {
+      "epoch": 0.35265935286765726,
+      "grad_norm": 3.5667520858997,
+      "learning_rate": 7.510056197385849e-06,
+      "loss": 0.452,
+      "step": 5079
+    },
+    {
+      "epoch": 0.3527287876683794,
+      "grad_norm": 4.20181025896451,
+      "learning_rate": 7.5090836085866e-06,
+      "loss": 0.5358,
+      "step": 5080
+    },
+    {
+      "epoch": 0.35279822246910153,
+      "grad_norm": 3.670930227059418,
+      "learning_rate": 7.508110892880743e-06,
+      "loss": 0.4909,
+      "step": 5081
+    },
+    {
+      "epoch": 0.3528676572698236,
+      "grad_norm": 3.7585188611870564,
+      "learning_rate": 7.507138050317471e-06,
+      "loss": 0.5334,
+      "step": 5082
+    },
+    {
+      "epoch": 0.35293709207054574,
+      "grad_norm": 3.6898727863433916,
+      "learning_rate": 7.506165080945997e-06,
+      "loss": 0.4755,
+      "step": 5083
+    },
+    {
+      "epoch": 0.3530065268712679,
+      "grad_norm": 4.801593865012166,
+      "learning_rate": 7.505191984815528e-06,
+      "loss": 0.5129,
+      "step": 5084
+    },
+    {
+      "epoch": 0.35307596167199,
+      "grad_norm": 4.345113521359982,
+      "learning_rate": 7.504218761975284e-06,
+      "loss": 0.6738,
+      "step": 5085
+    },
+    {
+      "epoch": 0.35314539647271215,
+      "grad_norm": 3.05013337366833,
+      "learning_rate": 7.503245412474487e-06,
+      "loss": 0.4401,
+      "step": 5086
+    },
+    {
+      "epoch": 0.3532148312734342,
+      "grad_norm": 3.1992875001686207,
+      "learning_rate": 7.502271936362371e-06,
+      "loss": 0.4152,
+      "step": 5087
+    },
+    {
+      "epoch": 0.35328426607415636,
+      "grad_norm": 3.6865908335844577,
+      "learning_rate": 7.501298333688171e-06,
+      "loss": 0.4075,
+      "step": 5088
+    },
+    {
+      "epoch": 0.3533537008748785,
+      "grad_norm": 3.030499961151148,
+      "learning_rate": 7.5003246045011325e-06,
+      "loss": 0.2025,
+      "step": 5089
+    },
+    {
+      "epoch": 0.35342313567560063,
+      "grad_norm": 2.4092893967900553,
+      "learning_rate": 7.499350748850504e-06,
+      "loss": 0.1898,
+      "step": 5090
+    },
+    {
+      "epoch": 0.3534925704763227,
+      "grad_norm": 4.050137950372166,
+      "learning_rate": 7.498376766785544e-06,
+      "loss": 0.31,
+      "step": 5091
+    },
+    {
+      "epoch": 0.35356200527704484,
+      "grad_norm": 3.7560851289282993,
+      "learning_rate": 7.497402658355514e-06,
+      "loss": 0.4685,
+      "step": 5092
+    },
+    {
+      "epoch": 0.353631440077767,
+      "grad_norm": 4.5650354346721285,
+      "learning_rate": 7.496428423609683e-06,
+      "loss": 0.7466,
+      "step": 5093
+    },
+    {
+      "epoch": 0.3537008748784891,
+      "grad_norm": 4.708974663696075,
+      "learning_rate": 7.495454062597327e-06,
+      "loss": 0.6004,
+      "step": 5094
+    },
+    {
+      "epoch": 0.35377030967921125,
+      "grad_norm": 2.740239467897368,
+      "learning_rate": 7.49447957536773e-06,
+      "loss": 0.267,
+      "step": 5095
+    },
+    {
+      "epoch": 0.3538397444799333,
+      "grad_norm": 3.8750891205585645,
+      "learning_rate": 7.493504961970176e-06,
+      "loss": 0.4118,
+      "step": 5096
+    },
+    {
+      "epoch": 0.35390917928065546,
+      "grad_norm": 3.220071268920196,
+      "learning_rate": 7.492530222453965e-06,
+      "loss": 0.4078,
+      "step": 5097
+    },
+    {
+      "epoch": 0.3539786140813776,
+      "grad_norm": 3.797214305271728,
+      "learning_rate": 7.491555356868394e-06,
+      "loss": 0.477,
+      "step": 5098
+    },
+    {
+      "epoch": 0.35404804888209973,
+      "grad_norm": 3.526919580558146,
+      "learning_rate": 7.490580365262774e-06,
+      "loss": 0.3859,
+      "step": 5099
+    },
+    {
+      "epoch": 0.3541174836828218,
+      "grad_norm": 4.36592115231672,
+      "learning_rate": 7.489605247686418e-06,
+      "loss": 0.6633,
+      "step": 5100
+    },
+    {
+      "epoch": 0.35418691848354394,
+      "grad_norm": 4.306410318277185,
+      "learning_rate": 7.488630004188644e-06,
+      "loss": 0.5159,
+      "step": 5101
+    },
+    {
+      "epoch": 0.3542563532842661,
+      "grad_norm": 4.2094183032909775,
+      "learning_rate": 7.487654634818779e-06,
+      "loss": 0.6105,
+      "step": 5102
+    },
+    {
+      "epoch": 0.3543257880849882,
+      "grad_norm": 3.152924946354207,
+      "learning_rate": 7.486679139626162e-06,
+      "loss": 0.4262,
+      "step": 5103
+    },
+    {
+      "epoch": 0.3543952228857103,
+      "grad_norm": 4.873150838666274,
+      "learning_rate": 7.485703518660125e-06,
+      "loss": 0.7435,
+      "step": 5104
+    },
+    {
+      "epoch": 0.3544646576864324,
+      "grad_norm": 3.4442813140711044,
+      "learning_rate": 7.484727771970018e-06,
+      "loss": 0.3982,
+      "step": 5105
+    },
+    {
+      "epoch": 0.35453409248715456,
+      "grad_norm": 3.6019492370140065,
+      "learning_rate": 7.48375189960519e-06,
+      "loss": 0.3958,
+      "step": 5106
+    },
+    {
+      "epoch": 0.3546035272878767,
+      "grad_norm": 3.82644381910013,
+      "learning_rate": 7.482775901615004e-06,
+      "loss": 0.4856,
+      "step": 5107
+    },
+    {
+      "epoch": 0.35467296208859883,
+      "grad_norm": 3.477387646264915,
+      "learning_rate": 7.48179977804882e-06,
+      "loss": 0.4396,
+      "step": 5108
+    },
+    {
+      "epoch": 0.3547423968893209,
+      "grad_norm": 3.51182238002862,
+      "learning_rate": 7.480823528956014e-06,
+      "loss": 0.414,
+      "step": 5109
+    },
+    {
+      "epoch": 0.35481183169004304,
+      "grad_norm": 4.452272184424566,
+      "learning_rate": 7.4798471543859596e-06,
+      "loss": 0.681,
+      "step": 5110
+    },
+    {
+      "epoch": 0.3548812664907652,
+      "grad_norm": 3.946796502272044,
+      "learning_rate": 7.478870654388042e-06,
+      "loss": 0.5638,
+      "step": 5111
+    },
+    {
+      "epoch": 0.3549507012914873,
+      "grad_norm": 3.05091184749042,
+      "learning_rate": 7.477894029011652e-06,
+      "loss": 0.3458,
+      "step": 5112
+    },
+    {
+      "epoch": 0.3550201360922094,
+      "grad_norm": 3.830585148424721,
+      "learning_rate": 7.476917278306187e-06,
+      "loss": 0.3243,
+      "step": 5113
+    },
+    {
+      "epoch": 0.3550895708929315,
+      "grad_norm": 2.44485822516337,
+      "learning_rate": 7.475940402321047e-06,
+      "loss": 0.2004,
+      "step": 5114
+    },
+    {
+      "epoch": 0.35515900569365366,
+      "grad_norm": 3.3391141212712507,
+      "learning_rate": 7.474963401105645e-06,
+      "loss": 0.3286,
+      "step": 5115
+    },
+    {
+      "epoch": 0.3552284404943758,
+      "grad_norm": 3.442540217833721,
+      "learning_rate": 7.473986274709394e-06,
+      "loss": 0.4583,
+      "step": 5116
+    },
+    {
+      "epoch": 0.3552978752950979,
+      "grad_norm": 4.63407256960876,
+      "learning_rate": 7.473009023181716e-06,
+      "loss": 0.816,
+      "step": 5117
+    },
+    {
+      "epoch": 0.35536731009582,
+      "grad_norm": 3.822461600354311,
+      "learning_rate": 7.472031646572041e-06,
+      "loss": 0.4705,
+      "step": 5118
+    },
+    {
+      "epoch": 0.35543674489654214,
+      "grad_norm": 6.352055086655258,
+      "learning_rate": 7.471054144929803e-06,
+      "loss": 0.7226,
+      "step": 5119
+    },
+    {
+      "epoch": 0.3555061796972643,
+      "grad_norm": 3.6755774766232947,
+      "learning_rate": 7.470076518304442e-06,
+      "loss": 0.4753,
+      "step": 5120
+    },
+    {
+      "epoch": 0.3555756144979864,
+      "grad_norm": 3.425734697930106,
+      "learning_rate": 7.469098766745405e-06,
+      "loss": 0.336,
+      "step": 5121
+    },
+    {
+      "epoch": 0.3556450492987085,
+      "grad_norm": 4.181907890672625,
+      "learning_rate": 7.4681208903021475e-06,
+      "loss": 0.5783,
+      "step": 5122
+    },
+    {
+      "epoch": 0.35571448409943063,
+      "grad_norm": 3.6531353706307828,
+      "learning_rate": 7.467142889024128e-06,
+      "loss": 0.4971,
+      "step": 5123
+    },
+    {
+      "epoch": 0.35578391890015276,
+      "grad_norm": 4.7086077798523585,
+      "learning_rate": 7.466164762960813e-06,
+      "loss": 0.7279,
+      "step": 5124
+    },
+    {
+      "epoch": 0.3558533537008749,
+      "grad_norm": 3.575866276327494,
+      "learning_rate": 7.465186512161673e-06,
+      "loss": 0.3507,
+      "step": 5125
+    },
+    {
+      "epoch": 0.355922788501597,
+      "grad_norm": 3.8811157899343454,
+      "learning_rate": 7.464208136676192e-06,
+      "loss": 0.3347,
+      "step": 5126
+    },
+    {
+      "epoch": 0.3559922233023191,
+      "grad_norm": 3.373350801542921,
+      "learning_rate": 7.463229636553849e-06,
+      "loss": 0.3407,
+      "step": 5127
+    },
+    {
+      "epoch": 0.35606165810304125,
+      "grad_norm": 4.580205091734243,
+      "learning_rate": 7.46225101184414e-06,
+      "loss": 0.5791,
+      "step": 5128
+    },
+    {
+      "epoch": 0.3561310929037634,
+      "grad_norm": 3.679391786235105,
+      "learning_rate": 7.4612722625965615e-06,
+      "loss": 0.4474,
+      "step": 5129
+    },
+    {
+      "epoch": 0.3562005277044855,
+      "grad_norm": 2.927897980835335,
+      "learning_rate": 7.460293388860616e-06,
+      "loss": 0.3082,
+      "step": 5130
+    },
+    {
+      "epoch": 0.3562699625052076,
+      "grad_norm": 3.3387537979858486,
+      "learning_rate": 7.459314390685815e-06,
+      "loss": 0.4871,
+      "step": 5131
+    },
+    {
+      "epoch": 0.35633939730592973,
+      "grad_norm": 3.4735151813449425,
+      "learning_rate": 7.458335268121676e-06,
+      "loss": 0.3503,
+      "step": 5132
+    },
+    {
+      "epoch": 0.35640883210665186,
+      "grad_norm": 3.2819009685628955,
+      "learning_rate": 7.45735602121772e-06,
+      "loss": 0.3309,
+      "step": 5133
+    },
+    {
+      "epoch": 0.356478266907374,
+      "grad_norm": 3.6017808184563407,
+      "learning_rate": 7.456376650023476e-06,
+      "loss": 0.4008,
+      "step": 5134
+    },
+    {
+      "epoch": 0.3565477017080961,
+      "grad_norm": 3.583515084334034,
+      "learning_rate": 7.4553971545884815e-06,
+      "loss": 0.4022,
+      "step": 5135
+    },
+    {
+      "epoch": 0.3566171365088182,
+      "grad_norm": 3.817768644533773,
+      "learning_rate": 7.454417534962279e-06,
+      "loss": 0.4465,
+      "step": 5136
+    },
+    {
+      "epoch": 0.35668657130954035,
+      "grad_norm": 3.4986597516410245,
+      "learning_rate": 7.453437791194414e-06,
+      "loss": 0.2979,
+      "step": 5137
+    },
+    {
+      "epoch": 0.3567560061102625,
+      "grad_norm": 3.076940286097457,
+      "learning_rate": 7.452457923334441e-06,
+      "loss": 0.2533,
+      "step": 5138
+    },
+    {
+      "epoch": 0.35682544091098456,
+      "grad_norm": 5.651024902883033,
+      "learning_rate": 7.451477931431922e-06,
+      "loss": 0.6201,
+      "step": 5139
+    },
+    {
+      "epoch": 0.3568948757117067,
+      "grad_norm": 3.0586800679149917,
+      "learning_rate": 7.450497815536424e-06,
+      "loss": 0.425,
+      "step": 5140
+    },
+    {
+      "epoch": 0.35696431051242883,
+      "grad_norm": 3.752184734875135,
+      "learning_rate": 7.449517575697517e-06,
+      "loss": 0.4455,
+      "step": 5141
+    },
+    {
+      "epoch": 0.35703374531315096,
+      "grad_norm": 3.837818998558299,
+      "learning_rate": 7.448537211964786e-06,
+      "loss": 0.4091,
+      "step": 5142
+    },
+    {
+      "epoch": 0.3571031801138731,
+      "grad_norm": 3.8101585730952463,
+      "learning_rate": 7.44755672438781e-06,
+      "loss": 0.4156,
+      "step": 5143
+    },
+    {
+      "epoch": 0.3571726149145952,
+      "grad_norm": 3.0879600080824745,
+      "learning_rate": 7.446576113016188e-06,
+      "loss": 0.235,
+      "step": 5144
+    },
+    {
+      "epoch": 0.3572420497153173,
+      "grad_norm": 4.241898492276191,
+      "learning_rate": 7.4455953778995125e-06,
+      "loss": 0.3403,
+      "step": 5145
+    },
+    {
+      "epoch": 0.35731148451603945,
+      "grad_norm": 4.927193058056566,
+      "learning_rate": 7.444614519087391e-06,
+      "loss": 0.2474,
+      "step": 5146
+    },
+    {
+      "epoch": 0.3573809193167616,
+      "grad_norm": 3.2232857521381137,
+      "learning_rate": 7.4436335366294334e-06,
+      "loss": 0.3312,
+      "step": 5147
+    },
+    {
+      "epoch": 0.35745035411748366,
+      "grad_norm": 3.3998370588161744,
+      "learning_rate": 7.442652430575257e-06,
+      "loss": 0.3784,
+      "step": 5148
+    },
+    {
+      "epoch": 0.3575197889182058,
+      "grad_norm": 3.92296589317403,
+      "learning_rate": 7.441671200974483e-06,
+      "loss": 0.3995,
+      "step": 5149
+    },
+    {
+      "epoch": 0.35758922371892793,
+      "grad_norm": 3.7896101974754783,
+      "learning_rate": 7.440689847876743e-06,
+      "loss": 0.5331,
+      "step": 5150
+    },
+    {
+      "epoch": 0.35765865851965006,
+      "grad_norm": 2.894088286185405,
+      "learning_rate": 7.439708371331673e-06,
+      "loss": 0.4172,
+      "step": 5151
+    },
+    {
+      "epoch": 0.3577280933203722,
+      "grad_norm": 3.7119832001087274,
+      "learning_rate": 7.438726771388915e-06,
+      "loss": 0.3261,
+      "step": 5152
+    },
+    {
+      "epoch": 0.3577975281210943,
+      "grad_norm": 3.7623974683311845,
+      "learning_rate": 7.437745048098116e-06,
+      "loss": 0.3755,
+      "step": 5153
+    },
+    {
+      "epoch": 0.3578669629218164,
+      "grad_norm": 4.261158375084097,
+      "learning_rate": 7.436763201508931e-06,
+      "loss": 0.5336,
+      "step": 5154
+    },
+    {
+      "epoch": 0.35793639772253855,
+      "grad_norm": 2.354268980583923,
+      "learning_rate": 7.43578123167102e-06,
+      "loss": 0.1809,
+      "step": 5155
+    },
+    {
+      "epoch": 0.3580058325232607,
+      "grad_norm": 2.6639528134118375,
+      "learning_rate": 7.434799138634051e-06,
+      "loss": 0.2497,
+      "step": 5156
+    },
+    {
+      "epoch": 0.35807526732398276,
+      "grad_norm": 2.9806344497623813,
+      "learning_rate": 7.433816922447697e-06,
+      "loss": 0.188,
+      "step": 5157
+    },
+    {
+      "epoch": 0.3581447021247049,
+      "grad_norm": 4.251981101823567,
+      "learning_rate": 7.432834583161637e-06,
+      "loss": 0.5382,
+      "step": 5158
+    },
+    {
+      "epoch": 0.35821413692542703,
+      "grad_norm": 3.8351636666552893,
+      "learning_rate": 7.4318521208255555e-06,
+      "loss": 0.5478,
+      "step": 5159
+    },
+    {
+      "epoch": 0.35828357172614916,
+      "grad_norm": 4.359071933954251,
+      "learning_rate": 7.430869535489147e-06,
+      "loss": 0.6244,
+      "step": 5160
+    },
+    {
+      "epoch": 0.35835300652687124,
+      "grad_norm": 4.231283947676127,
+      "learning_rate": 7.429886827202108e-06,
+      "loss": 0.6036,
+      "step": 5161
+    },
+    {
+      "epoch": 0.3584224413275934,
+      "grad_norm": 3.691277181649569,
+      "learning_rate": 7.428903996014143e-06,
+      "loss": 0.3578,
+      "step": 5162
+    },
+    {
+      "epoch": 0.3584918761283155,
+      "grad_norm": 4.061539534427449,
+      "learning_rate": 7.427921041974961e-06,
+      "loss": 0.5851,
+      "step": 5163
+    },
+    {
+      "epoch": 0.35856131092903765,
+      "grad_norm": 4.283109853697811,
+      "learning_rate": 7.426937965134282e-06,
+      "loss": 0.4067,
+      "step": 5164
+    },
+    {
+      "epoch": 0.3586307457297598,
+      "grad_norm": 4.03655546508861,
+      "learning_rate": 7.4259547655418265e-06,
+      "loss": 0.3808,
+      "step": 5165
+    },
+    {
+      "epoch": 0.35870018053048186,
+      "grad_norm": 4.346470175112034,
+      "learning_rate": 7.424971443247323e-06,
+      "loss": 0.4962,
+      "step": 5166
+    },
+    {
+      "epoch": 0.358769615331204,
+      "grad_norm": 3.431830100018635,
+      "learning_rate": 7.423987998300508e-06,
+      "loss": 0.3864,
+      "step": 5167
+    },
+    {
+      "epoch": 0.35883905013192613,
+      "grad_norm": 3.6227313882576118,
+      "learning_rate": 7.423004430751124e-06,
+      "loss": 0.4748,
+      "step": 5168
+    },
+    {
+      "epoch": 0.35890848493264826,
+      "grad_norm": 3.6309923782791307,
+      "learning_rate": 7.4220207406489176e-06,
+      "loss": 0.421,
+      "step": 5169
+    },
+    {
+      "epoch": 0.35897791973337034,
+      "grad_norm": 4.557882501971525,
+      "learning_rate": 7.421036928043643e-06,
+      "loss": 0.5369,
+      "step": 5170
+    },
+    {
+      "epoch": 0.3590473545340925,
+      "grad_norm": 4.170451569158213,
+      "learning_rate": 7.4200529929850596e-06,
+      "loss": 0.4801,
+      "step": 5171
+    },
+    {
+      "epoch": 0.3591167893348146,
+      "grad_norm": 4.8303186161802145,
+      "learning_rate": 7.419068935522935e-06,
+      "loss": 0.5724,
+      "step": 5172
+    },
+    {
+      "epoch": 0.35918622413553675,
+      "grad_norm": 3.7128687380257452,
+      "learning_rate": 7.418084755707039e-06,
+      "loss": 0.3838,
+      "step": 5173
+    },
+    {
+      "epoch": 0.3592556589362588,
+      "grad_norm": 3.764641466905656,
+      "learning_rate": 7.417100453587155e-06,
+      "loss": 0.3514,
+      "step": 5174
+    },
+    {
+      "epoch": 0.35932509373698096,
+      "grad_norm": 2.6466446309278227,
+      "learning_rate": 7.416116029213063e-06,
+      "loss": 0.2088,
+      "step": 5175
+    },
+    {
+      "epoch": 0.3593945285377031,
+      "grad_norm": 3.74943102285954,
+      "learning_rate": 7.415131482634558e-06,
+      "loss": 0.4627,
+      "step": 5176
+    },
+    {
+      "epoch": 0.35946396333842523,
+      "grad_norm": 3.971065791394491,
+      "learning_rate": 7.414146813901434e-06,
+      "loss": 0.404,
+      "step": 5177
+    },
+    {
+      "epoch": 0.35953339813914736,
+      "grad_norm": 4.7189982526175225,
+      "learning_rate": 7.413162023063497e-06,
+      "loss": 0.585,
+      "step": 5178
+    },
+    {
+      "epoch": 0.35960283293986944,
+      "grad_norm": 4.464242563749744,
+      "learning_rate": 7.412177110170555e-06,
+      "loss": 0.5609,
+      "step": 5179
+    },
+    {
+      "epoch": 0.3596722677405916,
+      "grad_norm": 4.351045089001082,
+      "learning_rate": 7.411192075272424e-06,
+      "loss": 0.5843,
+      "step": 5180
+    },
+    {
+      "epoch": 0.3597417025413137,
+      "grad_norm": 3.7251552536797847,
+      "learning_rate": 7.410206918418927e-06,
+      "loss": 0.4888,
+      "step": 5181
+    },
+    {
+      "epoch": 0.35981113734203585,
+      "grad_norm": 3.013973264601099,
+      "learning_rate": 7.409221639659892e-06,
+      "loss": 0.4285,
+      "step": 5182
+    },
+    {
+      "epoch": 0.3598805721427579,
+      "grad_norm": 4.599783170174827,
+      "learning_rate": 7.4082362390451514e-06,
+      "loss": 0.7125,
+      "step": 5183
+    },
+    {
+      "epoch": 0.35995000694348006,
+      "grad_norm": 5.373362477637537,
+      "learning_rate": 7.407250716624548e-06,
+      "loss": 0.6862,
+      "step": 5184
+    },
+    {
+      "epoch": 0.3600194417442022,
+      "grad_norm": 4.1047433762547225,
+      "learning_rate": 7.406265072447927e-06,
+      "loss": 0.5084,
+      "step": 5185
+    },
+    {
+      "epoch": 0.36008887654492433,
+      "grad_norm": 3.9460208601691815,
+      "learning_rate": 7.4052793065651406e-06,
+      "loss": 0.6067,
+      "step": 5186
+    },
+    {
+      "epoch": 0.36015831134564646,
+      "grad_norm": 3.811237232718791,
+      "learning_rate": 7.404293419026049e-06,
+      "loss": 0.4171,
+      "step": 5187
+    },
+    {
+      "epoch": 0.36022774614636854,
+      "grad_norm": 2.906746769387508,
+      "learning_rate": 7.403307409880518e-06,
+      "loss": 0.4269,
+      "step": 5188
+    },
+    {
+      "epoch": 0.3602971809470907,
+      "grad_norm": 4.9345261219228,
+      "learning_rate": 7.402321279178416e-06,
+      "loss": 0.5333,
+      "step": 5189
+    },
+    {
+      "epoch": 0.3603666157478128,
+      "grad_norm": 4.195846774321302,
+      "learning_rate": 7.401335026969624e-06,
+      "loss": 0.5604,
+      "step": 5190
+    },
+    {
+      "epoch": 0.36043605054853495,
+      "grad_norm": 3.0865426179268205,
+      "learning_rate": 7.400348653304022e-06,
+      "loss": 0.4003,
+      "step": 5191
+    },
+    {
+      "epoch": 0.360505485349257,
+      "grad_norm": 3.9512045337731916,
+      "learning_rate": 7.3993621582315026e-06,
+      "loss": 0.5755,
+      "step": 5192
+    },
+    {
+      "epoch": 0.36057492014997916,
+      "grad_norm": 4.95496577552601,
+      "learning_rate": 7.398375541801958e-06,
+      "loss": 0.6059,
+      "step": 5193
+    },
+    {
+      "epoch": 0.3606443549507013,
+      "grad_norm": 4.49146802581522,
+      "learning_rate": 7.397388804065295e-06,
+      "loss": 0.4936,
+      "step": 5194
+    },
+    {
+      "epoch": 0.36071378975142343,
+      "grad_norm": 4.262702809089655,
+      "learning_rate": 7.396401945071418e-06,
+      "loss": 0.4605,
+      "step": 5195
+    },
+    {
+      "epoch": 0.3607832245521455,
+      "grad_norm": 2.4272195273363124,
+      "learning_rate": 7.3954149648702425e-06,
+      "loss": 0.1331,
+      "step": 5196
+    },
+    {
+      "epoch": 0.36085265935286764,
+      "grad_norm": 3.875595990027703,
+      "learning_rate": 7.39442786351169e-06,
+      "loss": 0.3984,
+      "step": 5197
+    },
+    {
+      "epoch": 0.3609220941535898,
+      "grad_norm": 3.551575066844826,
+      "learning_rate": 7.393440641045684e-06,
+      "loss": 0.5079,
+      "step": 5198
+    },
+    {
+      "epoch": 0.3609915289543119,
+      "grad_norm": 3.866348054455549,
+      "learning_rate": 7.392453297522159e-06,
+      "loss": 0.4711,
+      "step": 5199
+    },
+    {
+      "epoch": 0.36106096375503405,
+      "grad_norm": 3.3568438154277995,
+      "learning_rate": 7.391465832991054e-06,
+      "loss": 0.34,
+      "step": 5200
+    },
+    {
+      "epoch": 0.3611303985557561,
+      "grad_norm": 3.671695421773663,
+      "learning_rate": 7.390478247502313e-06,
+      "loss": 0.3849,
+      "step": 5201
+    },
+    {
+      "epoch": 0.36119983335647826,
+      "grad_norm": 3.8577267975490765,
+      "learning_rate": 7.389490541105886e-06,
+      "loss": 0.5757,
+      "step": 5202
+    },
+    {
+      "epoch": 0.3612692681572004,
+      "grad_norm": 3.8988334185466287,
+      "learning_rate": 7.388502713851733e-06,
+      "loss": 0.4949,
+      "step": 5203
+    },
+    {
+      "epoch": 0.36133870295792253,
+      "grad_norm": 2.636997471803004,
+      "learning_rate": 7.387514765789815e-06,
+      "loss": 0.2375,
+      "step": 5204
+    },
+    {
+      "epoch": 0.3614081377586446,
+      "grad_norm": 3.869454872048885,
+      "learning_rate": 7.386526696970101e-06,
+      "loss": 0.4365,
+      "step": 5205
+    },
+    {
+      "epoch": 0.36147757255936674,
+      "grad_norm": 4.00359029345579,
+      "learning_rate": 7.385538507442568e-06,
+      "loss": 0.4625,
+      "step": 5206
+    },
+    {
+      "epoch": 0.3615470073600889,
+      "grad_norm": 4.804503710614609,
+      "learning_rate": 7.3845501972571965e-06,
+      "loss": 0.8323,
+      "step": 5207
+    },
+    {
+      "epoch": 0.361616442160811,
+      "grad_norm": 4.409764481915239,
+      "learning_rate": 7.3835617664639745e-06,
+      "loss": 0.3822,
+      "step": 5208
+    },
+    {
+      "epoch": 0.36168587696153315,
+      "grad_norm": 3.09672297085788,
+      "learning_rate": 7.382573215112896e-06,
+      "loss": 0.2411,
+      "step": 5209
+    },
+    {
+      "epoch": 0.3617553117622552,
+      "grad_norm": 3.934619100792288,
+      "learning_rate": 7.38158454325396e-06,
+      "loss": 0.5692,
+      "step": 5210
+    },
+    {
+      "epoch": 0.36182474656297736,
+      "grad_norm": 4.4710745413826825,
+      "learning_rate": 7.380595750937173e-06,
+      "loss": 0.7858,
+      "step": 5211
+    },
+    {
+      "epoch": 0.3618941813636995,
+      "grad_norm": 4.044929621972786,
+      "learning_rate": 7.379606838212545e-06,
+      "loss": 0.3569,
+      "step": 5212
+    },
+    {
+      "epoch": 0.36196361616442163,
+      "grad_norm": 4.043225176635758,
+      "learning_rate": 7.378617805130099e-06,
+      "loss": 0.5504,
+      "step": 5213
+    },
+    {
+      "epoch": 0.3620330509651437,
+      "grad_norm": 3.884783666682311,
+      "learning_rate": 7.377628651739853e-06,
+      "loss": 0.2867,
+      "step": 5214
+    },
+    {
+      "epoch": 0.36210248576586584,
+      "grad_norm": 4.181176039200803,
+      "learning_rate": 7.376639378091841e-06,
+      "loss": 0.4164,
+      "step": 5215
+    },
+    {
+      "epoch": 0.362171920566588,
+      "grad_norm": 3.641751353475488,
+      "learning_rate": 7.3756499842361e-06,
+      "loss": 0.5453,
+      "step": 5216
+    },
+    {
+      "epoch": 0.3622413553673101,
+      "grad_norm": 4.0412756899372875,
+      "learning_rate": 7.3746604702226696e-06,
+      "loss": 0.4528,
+      "step": 5217
+    },
+    {
+      "epoch": 0.3623107901680322,
+      "grad_norm": 2.758086316360966,
+      "learning_rate": 7.3736708361016e-06,
+      "loss": 0.2696,
+      "step": 5218
+    },
+    {
+      "epoch": 0.3623802249687543,
+      "grad_norm": 3.7266699836166013,
+      "learning_rate": 7.372681081922947e-06,
+      "loss": 0.5316,
+      "step": 5219
+    },
+    {
+      "epoch": 0.36244965976947646,
+      "grad_norm": 4.357410868657281,
+      "learning_rate": 7.371691207736769e-06,
+      "loss": 0.476,
+      "step": 5220
+    },
+    {
+      "epoch": 0.3625190945701986,
+      "grad_norm": 3.48191688827045,
+      "learning_rate": 7.370701213593131e-06,
+      "loss": 0.3644,
+      "step": 5221
+    },
+    {
+      "epoch": 0.36258852937092073,
+      "grad_norm": 4.3173703480943955,
+      "learning_rate": 7.369711099542111e-06,
+      "loss": 0.5097,
+      "step": 5222
+    },
+    {
+      "epoch": 0.3626579641716428,
+      "grad_norm": 4.529308761032398,
+      "learning_rate": 7.368720865633784e-06,
+      "loss": 0.7696,
+      "step": 5223
+    },
+    {
+      "epoch": 0.36272739897236494,
+      "grad_norm": 4.746701413286964,
+      "learning_rate": 7.367730511918236e-06,
+      "loss": 0.67,
+      "step": 5224
+    },
+    {
+      "epoch": 0.3627968337730871,
+      "grad_norm": 3.364463320412421,
+      "learning_rate": 7.366740038445558e-06,
+      "loss": 0.2047,
+      "step": 5225
+    },
+    {
+      "epoch": 0.3628662685738092,
+      "grad_norm": 3.446472059365696,
+      "learning_rate": 7.365749445265848e-06,
+      "loss": 0.4153,
+      "step": 5226
+    },
+    {
+      "epoch": 0.3629357033745313,
+      "grad_norm": 3.670606176534461,
+      "learning_rate": 7.364758732429207e-06,
+      "loss": 0.4569,
+      "step": 5227
+    },
+    {
+      "epoch": 0.36300513817525343,
+      "grad_norm": 3.106071035396148,
+      "learning_rate": 7.363767899985745e-06,
+      "loss": 0.363,
+      "step": 5228
+    },
+    {
+      "epoch": 0.36307457297597556,
+      "grad_norm": 3.0759031907010272,
+      "learning_rate": 7.362776947985579e-06,
+      "loss": 0.2045,
+      "step": 5229
+    },
+    {
+      "epoch": 0.3631440077766977,
+      "grad_norm": 2.7909201927851286,
+      "learning_rate": 7.361785876478827e-06,
+      "loss": 0.3142,
+      "step": 5230
+    },
+    {
+      "epoch": 0.3632134425774198,
+      "grad_norm": 3.935041565733543,
+      "learning_rate": 7.360794685515619e-06,
+      "loss": 0.5807,
+      "step": 5231
+    },
+    {
+      "epoch": 0.3632828773781419,
+      "grad_norm": 4.476043050234251,
+      "learning_rate": 7.359803375146086e-06,
+      "loss": 0.5882,
+      "step": 5232
+    },
+    {
+      "epoch": 0.36335231217886405,
+      "grad_norm": 4.3698002466110815,
+      "learning_rate": 7.35881194542037e-06,
+      "loss": 0.5222,
+      "step": 5233
+    },
+    {
+      "epoch": 0.3634217469795862,
+      "grad_norm": 4.014297716606658,
+      "learning_rate": 7.357820396388613e-06,
+      "loss": 0.5179,
+      "step": 5234
+    },
+    {
+      "epoch": 0.3634911817803083,
+      "grad_norm": 4.902661988077576,
+      "learning_rate": 7.356828728100971e-06,
+      "loss": 0.5505,
+      "step": 5235
+    },
+    {
+      "epoch": 0.3635606165810304,
+      "grad_norm": 4.732035972737129,
+      "learning_rate": 7.355836940607598e-06,
+      "loss": 0.7167,
+      "step": 5236
+    },
+    {
+      "epoch": 0.36363005138175253,
+      "grad_norm": 3.8664416142644855,
+      "learning_rate": 7.354845033958657e-06,
+      "loss": 0.5711,
+      "step": 5237
+    },
+    {
+      "epoch": 0.36369948618247466,
+      "grad_norm": 3.1375484152522084,
+      "learning_rate": 7.35385300820432e-06,
+      "loss": 0.3431,
+      "step": 5238
+    },
+    {
+      "epoch": 0.3637689209831968,
+      "grad_norm": 3.870866541431729,
+      "learning_rate": 7.352860863394762e-06,
+      "loss": 0.627,
+      "step": 5239
+    },
+    {
+      "epoch": 0.3638383557839189,
+      "grad_norm": 3.8516168247849816,
+      "learning_rate": 7.351868599580163e-06,
+      "loss": 0.3579,
+      "step": 5240
+    },
+    {
+      "epoch": 0.363907790584641,
+      "grad_norm": 3.607571547776457,
+      "learning_rate": 7.350876216810711e-06,
+      "loss": 0.451,
+      "step": 5241
+    },
+    {
+      "epoch": 0.36397722538536315,
+      "grad_norm": 4.559476636915713,
+      "learning_rate": 7.349883715136601e-06,
+      "loss": 0.5555,
+      "step": 5242
+    },
+    {
+      "epoch": 0.3640466601860853,
+      "grad_norm": 4.121016620171231,
+      "learning_rate": 7.348891094608031e-06,
+      "loss": 0.6644,
+      "step": 5243
+    },
+    {
+      "epoch": 0.3641160949868074,
+      "grad_norm": 3.957438445150098,
+      "learning_rate": 7.347898355275208e-06,
+      "loss": 0.4973,
+      "step": 5244
+    },
+    {
+      "epoch": 0.3641855297875295,
+      "grad_norm": 4.144260264535891,
+      "learning_rate": 7.346905497188344e-06,
+      "loss": 0.4929,
+      "step": 5245
+    },
+    {
+      "epoch": 0.36425496458825163,
+      "grad_norm": 2.842187093455992,
+      "learning_rate": 7.345912520397655e-06,
+      "loss": 0.2226,
+      "step": 5246
+    },
+    {
+      "epoch": 0.36432439938897376,
+      "grad_norm": 3.758283283670636,
+      "learning_rate": 7.344919424953365e-06,
+      "loss": 0.4126,
+      "step": 5247
+    },
+    {
+      "epoch": 0.3643938341896959,
+      "grad_norm": 3.947415827646632,
+      "learning_rate": 7.343926210905703e-06,
+      "loss": 0.6244,
+      "step": 5248
+    },
+    {
+      "epoch": 0.364463268990418,
+      "grad_norm": 4.62861053883523,
+      "learning_rate": 7.342932878304907e-06,
+      "loss": 0.3936,
+      "step": 5249
+    },
+    {
+      "epoch": 0.3645327037911401,
+      "grad_norm": 3.8111531731136408,
+      "learning_rate": 7.341939427201216e-06,
+      "loss": 0.5124,
+      "step": 5250
+    },
+    {
+      "epoch": 0.36460213859186225,
+      "grad_norm": 3.0192700686366267,
+      "learning_rate": 7.34094585764488e-06,
+      "loss": 0.4593,
+      "step": 5251
+    },
+    {
+      "epoch": 0.3646715733925844,
+      "grad_norm": 3.5492621108719176,
+      "learning_rate": 7.3399521696861505e-06,
+      "loss": 0.3993,
+      "step": 5252
+    },
+    {
+      "epoch": 0.36474100819330646,
+      "grad_norm": 3.5282137789283623,
+      "learning_rate": 7.338958363375289e-06,
+      "loss": 0.3456,
+      "step": 5253
+    },
+    {
+      "epoch": 0.3648104429940286,
+      "grad_norm": 4.795810296475828,
+      "learning_rate": 7.337964438762561e-06,
+      "loss": 0.6379,
+      "step": 5254
+    },
+    {
+      "epoch": 0.36487987779475073,
+      "grad_norm": 3.895704498200813,
+      "learning_rate": 7.336970395898235e-06,
+      "loss": 0.5505,
+      "step": 5255
+    },
+    {
+      "epoch": 0.36494931259547286,
+      "grad_norm": 3.7751176813765284,
+      "learning_rate": 7.3359762348325924e-06,
+      "loss": 0.4449,
+      "step": 5256
+    },
+    {
+      "epoch": 0.365018747396195,
+      "grad_norm": 3.0246909841024086,
+      "learning_rate": 7.334981955615914e-06,
+      "loss": 0.2935,
+      "step": 5257
+    },
+    {
+      "epoch": 0.3650881821969171,
+      "grad_norm": 4.258508689344232,
+      "learning_rate": 7.333987558298492e-06,
+      "loss": 0.536,
+      "step": 5258
+    },
+    {
+      "epoch": 0.3651576169976392,
+      "grad_norm": 3.4538149863428282,
+      "learning_rate": 7.33299304293062e-06,
+      "loss": 0.4596,
+      "step": 5259
+    },
+    {
+      "epoch": 0.36522705179836135,
+      "grad_norm": 3.8057677283709945,
+      "learning_rate": 7.331998409562598e-06,
+      "loss": 0.5072,
+      "step": 5260
+    },
+    {
+      "epoch": 0.3652964865990835,
+      "grad_norm": 3.391090788268474,
+      "learning_rate": 7.331003658244738e-06,
+      "loss": 0.3337,
+      "step": 5261
+    },
+    {
+      "epoch": 0.36536592139980556,
+      "grad_norm": 3.7674882006690207,
+      "learning_rate": 7.3300087890273496e-06,
+      "loss": 0.5325,
+      "step": 5262
+    },
+    {
+      "epoch": 0.3654353562005277,
+      "grad_norm": 4.780023535577388,
+      "learning_rate": 7.329013801960752e-06,
+      "loss": 0.7462,
+      "step": 5263
+    },
+    {
+      "epoch": 0.36550479100124983,
+      "grad_norm": 3.2825605342517714,
+      "learning_rate": 7.328018697095274e-06,
+      "loss": 0.3117,
+      "step": 5264
+    },
+    {
+      "epoch": 0.36557422580197196,
+      "grad_norm": 4.230946027704936,
+      "learning_rate": 7.327023474481243e-06,
+      "loss": 0.4621,
+      "step": 5265
+    },
+    {
+      "epoch": 0.36564366060269404,
+      "grad_norm": 2.9427390772138278,
+      "learning_rate": 7.326028134168998e-06,
+      "loss": 0.2859,
+      "step": 5266
+    },
+    {
+      "epoch": 0.3657130954034162,
+      "grad_norm": 3.6615293487189127,
+      "learning_rate": 7.325032676208882e-06,
+      "loss": 0.7188,
+      "step": 5267
+    },
+    {
+      "epoch": 0.3657825302041383,
+      "grad_norm": 4.537087291950671,
+      "learning_rate": 7.324037100651245e-06,
+      "loss": 0.6963,
+      "step": 5268
+    },
+    {
+      "epoch": 0.36585196500486045,
+      "grad_norm": 4.155695048588974,
+      "learning_rate": 7.32304140754644e-06,
+      "loss": 0.6195,
+      "step": 5269
+    },
+    {
+      "epoch": 0.3659213998055826,
+      "grad_norm": 4.023666585952381,
+      "learning_rate": 7.3220455969448315e-06,
+      "loss": 0.2538,
+      "step": 5270
+    },
+    {
+      "epoch": 0.36599083460630466,
+      "grad_norm": 3.8526617838127613,
+      "learning_rate": 7.321049668896783e-06,
+      "loss": 0.5782,
+      "step": 5271
+    },
+    {
+      "epoch": 0.3660602694070268,
+      "grad_norm": 2.9281759798026505,
+      "learning_rate": 7.320053623452668e-06,
+      "loss": 0.361,
+      "step": 5272
+    },
+    {
+      "epoch": 0.36612970420774893,
+      "grad_norm": 3.5749795127803807,
+      "learning_rate": 7.319057460662866e-06,
+      "loss": 0.2888,
+      "step": 5273
+    },
+    {
+      "epoch": 0.36619913900847106,
+      "grad_norm": 5.041645832718997,
+      "learning_rate": 7.318061180577763e-06,
+      "loss": 0.5884,
+      "step": 5274
+    },
+    {
+      "epoch": 0.36626857380919314,
+      "grad_norm": 3.408888256001804,
+      "learning_rate": 7.317064783247748e-06,
+      "loss": 0.2853,
+      "step": 5275
+    },
+    {
+      "epoch": 0.3663380086099153,
+      "grad_norm": 4.4212115541538815,
+      "learning_rate": 7.3160682687232186e-06,
+      "loss": 0.4673,
+      "step": 5276
+    },
+    {
+      "epoch": 0.3664074434106374,
+      "grad_norm": 4.353473922522457,
+      "learning_rate": 7.315071637054576e-06,
+      "loss": 0.5888,
+      "step": 5277
+    },
+    {
+      "epoch": 0.36647687821135955,
+      "grad_norm": 5.260715173062866,
+      "learning_rate": 7.314074888292231e-06,
+      "loss": 0.8462,
+      "step": 5278
+    },
+    {
+      "epoch": 0.3665463130120817,
+      "grad_norm": 2.8620908715304836,
+      "learning_rate": 7.313078022486596e-06,
+      "loss": 0.2493,
+      "step": 5279
+    },
+    {
+      "epoch": 0.36661574781280376,
+      "grad_norm": 3.759821781758709,
+      "learning_rate": 7.3120810396880925e-06,
+      "loss": 0.5773,
+      "step": 5280
+    },
+    {
+      "epoch": 0.3666851826135259,
+      "grad_norm": 3.573085953884359,
+      "learning_rate": 7.311083939947146e-06,
+      "loss": 0.4299,
+      "step": 5281
+    },
+    {
+      "epoch": 0.36675461741424803,
+      "grad_norm": 3.0366561594339925,
+      "learning_rate": 7.310086723314187e-06,
+      "loss": 0.3328,
+      "step": 5282
+    },
+    {
+      "epoch": 0.36682405221497016,
+      "grad_norm": 4.588819409680726,
+      "learning_rate": 7.309089389839658e-06,
+      "loss": 0.6397,
+      "step": 5283
+    },
+    {
+      "epoch": 0.36689348701569224,
+      "grad_norm": 3.316497831756882,
+      "learning_rate": 7.308091939574e-06,
+      "loss": 0.3728,
+      "step": 5284
+    },
+    {
+      "epoch": 0.3669629218164144,
+      "grad_norm": 4.253174149251256,
+      "learning_rate": 7.307094372567665e-06,
+      "loss": 0.5961,
+      "step": 5285
+    },
+    {
+      "epoch": 0.3670323566171365,
+      "grad_norm": 5.918741004012838,
+      "learning_rate": 7.306096688871105e-06,
+      "loss": 0.4941,
+      "step": 5286
+    },
+    {
+      "epoch": 0.36710179141785865,
+      "grad_norm": 2.6927485888703426,
+      "learning_rate": 7.305098888534785e-06,
+      "loss": 0.2656,
+      "step": 5287
+    },
+    {
+      "epoch": 0.3671712262185807,
+      "grad_norm": 3.4814665984798094,
+      "learning_rate": 7.304100971609171e-06,
+      "loss": 0.3898,
+      "step": 5288
+    },
+    {
+      "epoch": 0.36724066101930286,
+      "grad_norm": 6.339991452732976,
+      "learning_rate": 7.3031029381447375e-06,
+      "loss": 0.4089,
+      "step": 5289
+    },
+    {
+      "epoch": 0.367310095820025,
+      "grad_norm": 3.9694920096492936,
+      "learning_rate": 7.302104788191964e-06,
+      "loss": 0.6052,
+      "step": 5290
+    },
+    {
+      "epoch": 0.36737953062074713,
+      "grad_norm": 4.767434801543367,
+      "learning_rate": 7.3011065218013335e-06,
+      "loss": 0.7248,
+      "step": 5291
+    },
+    {
+      "epoch": 0.36744896542146926,
+      "grad_norm": 3.148374933859908,
+      "learning_rate": 7.300108139023341e-06,
+      "loss": 0.1739,
+      "step": 5292
+    },
+    {
+      "epoch": 0.36751840022219134,
+      "grad_norm": 3.7982139615036394,
+      "learning_rate": 7.299109639908482e-06,
+      "loss": 0.4263,
+      "step": 5293
+    },
+    {
+      "epoch": 0.3675878350229135,
+      "grad_norm": 4.081060341811651,
+      "learning_rate": 7.298111024507257e-06,
+      "loss": 0.6496,
+      "step": 5294
+    },
+    {
+      "epoch": 0.3676572698236356,
+      "grad_norm": 4.000938839097152,
+      "learning_rate": 7.297112292870177e-06,
+      "loss": 0.4879,
+      "step": 5295
+    },
+    {
+      "epoch": 0.36772670462435775,
+      "grad_norm": 2.483007804669054,
+      "learning_rate": 7.296113445047758e-06,
+      "loss": 0.2168,
+      "step": 5296
+    },
+    {
+      "epoch": 0.3677961394250798,
+      "grad_norm": 4.982133218500219,
+      "learning_rate": 7.295114481090519e-06,
+      "loss": 0.8941,
+      "step": 5297
+    },
+    {
+      "epoch": 0.36786557422580196,
+      "grad_norm": 3.505457970585262,
+      "learning_rate": 7.294115401048985e-06,
+      "loss": 0.5145,
+      "step": 5298
+    },
+    {
+      "epoch": 0.3679350090265241,
+      "grad_norm": 4.122571104457174,
+      "learning_rate": 7.293116204973691e-06,
+      "loss": 0.4536,
+      "step": 5299
+    },
+    {
+      "epoch": 0.36800444382724623,
+      "grad_norm": 6.8168397526491535,
+      "learning_rate": 7.292116892915175e-06,
+      "loss": 0.442,
+      "step": 5300
+    },
+    {
+      "epoch": 0.36807387862796836,
+      "grad_norm": 4.371468614081541,
+      "learning_rate": 7.29111746492398e-06,
+      "loss": 0.7694,
+      "step": 5301
+    },
+    {
+      "epoch": 0.36814331342869044,
+      "grad_norm": 3.5083510698018165,
+      "learning_rate": 7.290117921050657e-06,
+      "loss": 0.3829,
+      "step": 5302
+    },
+    {
+      "epoch": 0.3682127482294126,
+      "grad_norm": 3.2086290955614385,
+      "learning_rate": 7.289118261345759e-06,
+      "loss": 0.4295,
+      "step": 5303
+    },
+    {
+      "epoch": 0.3682821830301347,
+      "grad_norm": 4.1133573463170405,
+      "learning_rate": 7.288118485859851e-06,
+      "loss": 0.2946,
+      "step": 5304
+    },
+    {
+      "epoch": 0.36835161783085685,
+      "grad_norm": 4.831428665242624,
+      "learning_rate": 7.287118594643497e-06,
+      "loss": 0.4511,
+      "step": 5305
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 3.9774320312836013,
+      "learning_rate": 7.286118587747275e-06,
+      "loss": 0.3544,
+      "step": 5306
+    },
+    {
+      "epoch": 0.36849048743230106,
+      "grad_norm": 3.949426377053619,
+      "learning_rate": 7.285118465221761e-06,
+      "loss": 0.4384,
+      "step": 5307
+    },
+    {
+      "epoch": 0.3685599222330232,
+      "grad_norm": 2.5959500956444965,
+      "learning_rate": 7.284118227117541e-06,
+      "loss": 0.2793,
+      "step": 5308
+    },
+    {
+      "epoch": 0.36862935703374533,
+      "grad_norm": 3.5261209281636376,
+      "learning_rate": 7.283117873485206e-06,
+      "loss": 0.3596,
+      "step": 5309
+    },
+    {
+      "epoch": 0.3686987918344674,
+      "grad_norm": 4.22786861052139,
+      "learning_rate": 7.282117404375352e-06,
+      "loss": 0.4395,
+      "step": 5310
+    },
+    {
+      "epoch": 0.36876822663518954,
+      "grad_norm": 4.4298154480332546,
+      "learning_rate": 7.281116819838582e-06,
+      "loss": 0.5589,
+      "step": 5311
+    },
+    {
+      "epoch": 0.3688376614359117,
+      "grad_norm": 3.3456104838391303,
+      "learning_rate": 7.280116119925504e-06,
+      "loss": 0.3398,
+      "step": 5312
+    },
+    {
+      "epoch": 0.3689070962366338,
+      "grad_norm": 2.6710461073306893,
+      "learning_rate": 7.2791153046867344e-06,
+      "loss": 0.14,
+      "step": 5313
+    },
+    {
+      "epoch": 0.36897653103735595,
+      "grad_norm": 4.170125909785202,
+      "learning_rate": 7.27811437417289e-06,
+      "loss": 0.4526,
+      "step": 5314
+    },
+    {
+      "epoch": 0.369045965838078,
+      "grad_norm": 4.355060077280559,
+      "learning_rate": 7.2771133284345984e-06,
+      "loss": 0.4518,
+      "step": 5315
+    },
+    {
+      "epoch": 0.36911540063880016,
+      "grad_norm": 3.504673707567617,
+      "learning_rate": 7.2761121675224926e-06,
+      "loss": 0.4897,
+      "step": 5316
+    },
+    {
+      "epoch": 0.3691848354395223,
+      "grad_norm": 4.272493027386857,
+      "learning_rate": 7.275110891487208e-06,
+      "loss": 0.3938,
+      "step": 5317
+    },
+    {
+      "epoch": 0.36925427024024443,
+      "grad_norm": 3.885785869257253,
+      "learning_rate": 7.274109500379389e-06,
+      "loss": 0.2657,
+      "step": 5318
+    },
+    {
+      "epoch": 0.3693237050409665,
+      "grad_norm": 3.818801198535173,
+      "learning_rate": 7.2731079942496865e-06,
+      "loss": 0.2898,
+      "step": 5319
+    },
+    {
+      "epoch": 0.36939313984168864,
+      "grad_norm": 4.51990971306465,
+      "learning_rate": 7.272106373148753e-06,
+      "loss": 0.609,
+      "step": 5320
+    },
+    {
+      "epoch": 0.3694625746424108,
+      "grad_norm": 3.212455743026587,
+      "learning_rate": 7.271104637127248e-06,
+      "loss": 0.2053,
+      "step": 5321
+    },
+    {
+      "epoch": 0.3695320094431329,
+      "grad_norm": 3.6747098250465444,
+      "learning_rate": 7.270102786235842e-06,
+      "loss": 0.5695,
+      "step": 5322
+    },
+    {
+      "epoch": 0.369601444243855,
+      "grad_norm": 4.8643152895152095,
+      "learning_rate": 7.269100820525207e-06,
+      "loss": 0.6991,
+      "step": 5323
+    },
+    {
+      "epoch": 0.3696708790445771,
+      "grad_norm": 3.3344190435829195,
+      "learning_rate": 7.268098740046019e-06,
+      "loss": 0.2979,
+      "step": 5324
+    },
+    {
+      "epoch": 0.36974031384529926,
+      "grad_norm": 3.9345070348116917,
+      "learning_rate": 7.267096544848964e-06,
+      "loss": 0.4614,
+      "step": 5325
+    },
+    {
+      "epoch": 0.3698097486460214,
+      "grad_norm": 4.397042030236148,
+      "learning_rate": 7.26609423498473e-06,
+      "loss": 0.2951,
+      "step": 5326
+    },
+    {
+      "epoch": 0.36987918344674353,
+      "grad_norm": 3.619514368007099,
+      "learning_rate": 7.265091810504015e-06,
+      "loss": 0.4345,
+      "step": 5327
+    },
+    {
+      "epoch": 0.3699486182474656,
+      "grad_norm": 3.8687171577500457,
+      "learning_rate": 7.264089271457517e-06,
+      "loss": 0.4041,
+      "step": 5328
+    },
+    {
+      "epoch": 0.37001805304818775,
+      "grad_norm": 3.5983570895417527,
+      "learning_rate": 7.263086617895949e-06,
+      "loss": 0.2985,
+      "step": 5329
+    },
+    {
+      "epoch": 0.3700874878489099,
+      "grad_norm": 3.5343622140015527,
+      "learning_rate": 7.2620838498700185e-06,
+      "loss": 0.3636,
+      "step": 5330
+    },
+    {
+      "epoch": 0.370156922649632,
+      "grad_norm": 3.0571042132074706,
+      "learning_rate": 7.261080967430447e-06,
+      "loss": 0.4412,
+      "step": 5331
+    },
+    {
+      "epoch": 0.3702263574503541,
+      "grad_norm": 4.100132079238135,
+      "learning_rate": 7.2600779706279585e-06,
+      "loss": 0.7203,
+      "step": 5332
+    },
+    {
+      "epoch": 0.37029579225107623,
+      "grad_norm": 3.2362760911462245,
+      "learning_rate": 7.259074859513284e-06,
+      "loss": 0.401,
+      "step": 5333
+    },
+    {
+      "epoch": 0.37036522705179836,
+      "grad_norm": 3.421911420071634,
+      "learning_rate": 7.2580716341371594e-06,
+      "loss": 0.4264,
+      "step": 5334
+    },
+    {
+      "epoch": 0.3704346618525205,
+      "grad_norm": 5.187443005402242,
+      "learning_rate": 7.257068294550328e-06,
+      "loss": 0.5395,
+      "step": 5335
+    },
+    {
+      "epoch": 0.37050409665324263,
+      "grad_norm": 5.114298487349007,
+      "learning_rate": 7.256064840803535e-06,
+      "loss": 0.4778,
+      "step": 5336
+    },
+    {
+      "epoch": 0.3705735314539647,
+      "grad_norm": 4.08902035128331,
+      "learning_rate": 7.255061272947535e-06,
+      "loss": 0.4919,
+      "step": 5337
+    },
+    {
+      "epoch": 0.37064296625468685,
+      "grad_norm": 4.266926616486269,
+      "learning_rate": 7.254057591033088e-06,
+      "loss": 0.4322,
+      "step": 5338
+    },
+    {
+      "epoch": 0.370712401055409,
+      "grad_norm": 3.5956776393265417,
+      "learning_rate": 7.253053795110959e-06,
+      "loss": 0.3342,
+      "step": 5339
+    },
+    {
+      "epoch": 0.3707818358561311,
+      "grad_norm": 4.127598444649088,
+      "learning_rate": 7.252049885231919e-06,
+      "loss": 0.5547,
+      "step": 5340
+    },
+    {
+      "epoch": 0.3708512706568532,
+      "grad_norm": 4.0647867038519925,
+      "learning_rate": 7.251045861446741e-06,
+      "loss": 0.5805,
+      "step": 5341
+    },
+    {
+      "epoch": 0.37092070545757533,
+      "grad_norm": 3.2870579504611563,
+      "learning_rate": 7.250041723806215e-06,
+      "loss": 0.3859,
+      "step": 5342
+    },
+    {
+      "epoch": 0.37099014025829746,
+      "grad_norm": 4.002343069734852,
+      "learning_rate": 7.249037472361122e-06,
+      "loss": 0.5498,
+      "step": 5343
+    },
+    {
+      "epoch": 0.3710595750590196,
+      "grad_norm": 5.989080017540409,
+      "learning_rate": 7.24803310716226e-06,
+      "loss": 0.4263,
+      "step": 5344
+    },
+    {
+      "epoch": 0.3711290098597417,
+      "grad_norm": 3.336577746460094,
+      "learning_rate": 7.247028628260427e-06,
+      "loss": 0.3076,
+      "step": 5345
+    },
+    {
+      "epoch": 0.3711984446604638,
+      "grad_norm": 3.782180950500412,
+      "learning_rate": 7.246024035706427e-06,
+      "loss": 0.6037,
+      "step": 5346
+    },
+    {
+      "epoch": 0.37126787946118595,
+      "grad_norm": 3.523336924712128,
+      "learning_rate": 7.245019329551075e-06,
+      "loss": 0.5285,
+      "step": 5347
+    },
+    {
+      "epoch": 0.3713373142619081,
+      "grad_norm": 2.909187582974306,
+      "learning_rate": 7.244014509845185e-06,
+      "loss": 0.3129,
+      "step": 5348
+    },
+    {
+      "epoch": 0.3714067490626302,
+      "grad_norm": 3.9257869886414127,
+      "learning_rate": 7.24300957663958e-06,
+      "loss": 0.4827,
+      "step": 5349
+    },
+    {
+      "epoch": 0.3714761838633523,
+      "grad_norm": 4.759376921279855,
+      "learning_rate": 7.242004529985088e-06,
+      "loss": 0.7958,
+      "step": 5350
+    },
+    {
+      "epoch": 0.37154561866407443,
+      "grad_norm": 3.0704874727802753,
+      "learning_rate": 7.2409993699325455e-06,
+      "loss": 0.3401,
+      "step": 5351
+    },
+    {
+      "epoch": 0.37161505346479656,
+      "grad_norm": 3.4863473330233274,
+      "learning_rate": 7.239994096532791e-06,
+      "loss": 0.3516,
+      "step": 5352
+    },
+    {
+      "epoch": 0.3716844882655187,
+      "grad_norm": 3.859641586497657,
+      "learning_rate": 7.238988709836667e-06,
+      "loss": 0.5821,
+      "step": 5353
+    },
+    {
+      "epoch": 0.3717539230662408,
+      "grad_norm": 3.750376212473087,
+      "learning_rate": 7.23798320989503e-06,
+      "loss": 0.5067,
+      "step": 5354
+    },
+    {
+      "epoch": 0.3718233578669629,
+      "grad_norm": 2.6504672904390043,
+      "learning_rate": 7.236977596758734e-06,
+      "loss": 0.2996,
+      "step": 5355
+    },
+    {
+      "epoch": 0.37189279266768505,
+      "grad_norm": 4.798207729253736,
+      "learning_rate": 7.235971870478643e-06,
+      "loss": 0.5289,
+      "step": 5356
+    },
+    {
+      "epoch": 0.3719622274684072,
+      "grad_norm": 3.3210465521389483,
+      "learning_rate": 7.234966031105624e-06,
+      "loss": 0.3565,
+      "step": 5357
+    },
+    {
+      "epoch": 0.3720316622691293,
+      "grad_norm": 5.1063542531716575,
+      "learning_rate": 7.233960078690554e-06,
+      "loss": 0.6655,
+      "step": 5358
+    },
+    {
+      "epoch": 0.3721010970698514,
+      "grad_norm": 3.8372303135424457,
+      "learning_rate": 7.232954013284309e-06,
+      "loss": 0.5025,
+      "step": 5359
+    },
+    {
+      "epoch": 0.37217053187057353,
+      "grad_norm": 3.87013071111723,
+      "learning_rate": 7.231947834937778e-06,
+      "loss": 0.5926,
+      "step": 5360
+    },
+    {
+      "epoch": 0.37223996667129566,
+      "grad_norm": 4.338086203255544,
+      "learning_rate": 7.230941543701852e-06,
+      "loss": 0.2548,
+      "step": 5361
+    },
+    {
+      "epoch": 0.3723094014720178,
+      "grad_norm": 4.280938990520116,
+      "learning_rate": 7.229935139627425e-06,
+      "loss": 0.3738,
+      "step": 5362
+    },
+    {
+      "epoch": 0.3723788362727399,
+      "grad_norm": 3.779119701968508,
+      "learning_rate": 7.228928622765403e-06,
+      "loss": 0.5819,
+      "step": 5363
+    },
+    {
+      "epoch": 0.372448271073462,
+      "grad_norm": 3.6944573076621623,
+      "learning_rate": 7.227921993166693e-06,
+      "loss": 0.4794,
+      "step": 5364
+    },
+    {
+      "epoch": 0.37251770587418415,
+      "grad_norm": 3.5566956861035712,
+      "learning_rate": 7.2269152508822105e-06,
+      "loss": 0.3298,
+      "step": 5365
+    },
+    {
+      "epoch": 0.3725871406749063,
+      "grad_norm": 4.043473612760919,
+      "learning_rate": 7.225908395962874e-06,
+      "loss": 0.636,
+      "step": 5366
+    },
+    {
+      "epoch": 0.37265657547562836,
+      "grad_norm": 3.005836441072363,
+      "learning_rate": 7.22490142845961e-06,
+      "loss": 0.3953,
+      "step": 5367
+    },
+    {
+      "epoch": 0.3727260102763505,
+      "grad_norm": 3.815252297419189,
+      "learning_rate": 7.223894348423348e-06,
+      "loss": 0.3678,
+      "step": 5368
+    },
+    {
+      "epoch": 0.37279544507707263,
+      "grad_norm": 9.738091458011734,
+      "learning_rate": 7.2228871559050285e-06,
+      "loss": 0.5951,
+      "step": 5369
+    },
+    {
+      "epoch": 0.37286487987779476,
+      "grad_norm": 3.375737299053214,
+      "learning_rate": 7.221879850955591e-06,
+      "loss": 0.4138,
+      "step": 5370
+    },
+    {
+      "epoch": 0.3729343146785169,
+      "grad_norm": 4.526199130311125,
+      "learning_rate": 7.220872433625985e-06,
+      "loss": 0.5608,
+      "step": 5371
+    },
+    {
+      "epoch": 0.373003749479239,
+      "grad_norm": 3.4127951358586457,
+      "learning_rate": 7.219864903967165e-06,
+      "loss": 0.3672,
+      "step": 5372
+    },
+    {
+      "epoch": 0.3730731842799611,
+      "grad_norm": 3.9487718542003947,
+      "learning_rate": 7.218857262030088e-06,
+      "loss": 0.5505,
+      "step": 5373
+    },
+    {
+      "epoch": 0.37314261908068325,
+      "grad_norm": 4.7702697981777025,
+      "learning_rate": 7.217849507865724e-06,
+      "loss": 0.8822,
+      "step": 5374
+    },
+    {
+      "epoch": 0.3732120538814054,
+      "grad_norm": 4.007945215007396,
+      "learning_rate": 7.21684164152504e-06,
+      "loss": 0.6258,
+      "step": 5375
+    },
+    {
+      "epoch": 0.37328148868212746,
+      "grad_norm": 2.2098324262295783,
+      "learning_rate": 7.215833663059014e-06,
+      "loss": 0.1477,
+      "step": 5376
+    },
+    {
+      "epoch": 0.3733509234828496,
+      "grad_norm": 3.786374816914992,
+      "learning_rate": 7.21482557251863e-06,
+      "loss": 0.6749,
+      "step": 5377
+    },
+    {
+      "epoch": 0.37342035828357173,
+      "grad_norm": 3.0939635771475724,
+      "learning_rate": 7.213817369954874e-06,
+      "loss": 0.3183,
+      "step": 5378
+    },
+    {
+      "epoch": 0.37348979308429386,
+      "grad_norm": 3.344641634130665,
+      "learning_rate": 7.212809055418741e-06,
+      "loss": 0.5366,
+      "step": 5379
+    },
+    {
+      "epoch": 0.37355922788501594,
+      "grad_norm": 4.892370232721007,
+      "learning_rate": 7.211800628961229e-06,
+      "loss": 0.6668,
+      "step": 5380
+    },
+    {
+      "epoch": 0.3736286626857381,
+      "grad_norm": 3.0016041194240533,
+      "learning_rate": 7.210792090633346e-06,
+      "loss": 0.4618,
+      "step": 5381
+    },
+    {
+      "epoch": 0.3736980974864602,
+      "grad_norm": 2.750773129489912,
+      "learning_rate": 7.209783440486099e-06,
+      "loss": 0.2897,
+      "step": 5382
+    },
+    {
+      "epoch": 0.37376753228718235,
+      "grad_norm": 3.6787301172973454,
+      "learning_rate": 7.208774678570507e-06,
+      "loss": 0.4531,
+      "step": 5383
+    },
+    {
+      "epoch": 0.3738369670879045,
+      "grad_norm": 36.762015953860036,
+      "learning_rate": 7.20776580493759e-06,
+      "loss": 0.3385,
+      "step": 5384
+    },
+    {
+      "epoch": 0.37390640188862656,
+      "grad_norm": 4.241586539949382,
+      "learning_rate": 7.206756819638379e-06,
+      "loss": 0.5358,
+      "step": 5385
+    },
+    {
+      "epoch": 0.3739758366893487,
+      "grad_norm": 3.8746411551401603,
+      "learning_rate": 7.205747722723903e-06,
+      "loss": 0.6325,
+      "step": 5386
+    },
+    {
+      "epoch": 0.37404527149007083,
+      "grad_norm": 3.577317496582448,
+      "learning_rate": 7.204738514245205e-06,
+      "loss": 0.3492,
+      "step": 5387
+    },
+    {
+      "epoch": 0.37411470629079296,
+      "grad_norm": 4.033481043176407,
+      "learning_rate": 7.203729194253326e-06,
+      "loss": 0.8111,
+      "step": 5388
+    },
+    {
+      "epoch": 0.37418414109151504,
+      "grad_norm": 2.464725638679876,
+      "learning_rate": 7.202719762799317e-06,
+      "loss": 0.2851,
+      "step": 5389
+    },
+    {
+      "epoch": 0.3742535758922372,
+      "grad_norm": 3.8253699518459614,
+      "learning_rate": 7.201710219934238e-06,
+      "loss": 0.5091,
+      "step": 5390
+    },
+    {
+      "epoch": 0.3743230106929593,
+      "grad_norm": 4.028008669758356,
+      "learning_rate": 7.200700565709146e-06,
+      "loss": 0.5012,
+      "step": 5391
+    },
+    {
+      "epoch": 0.37439244549368145,
+      "grad_norm": 3.5943057061521633,
+      "learning_rate": 7.19969080017511e-06,
+      "loss": 0.4352,
+      "step": 5392
+    },
+    {
+      "epoch": 0.3744618802944036,
+      "grad_norm": 4.7692909311279195,
+      "learning_rate": 7.1986809233832025e-06,
+      "loss": 0.7374,
+      "step": 5393
+    },
+    {
+      "epoch": 0.37453131509512566,
+      "grad_norm": 4.040817038949215,
+      "learning_rate": 7.197670935384501e-06,
+      "loss": 0.5026,
+      "step": 5394
+    },
+    {
+      "epoch": 0.3746007498958478,
+      "grad_norm": 4.471130086693506,
+      "learning_rate": 7.196660836230091e-06,
+      "loss": 0.3923,
+      "step": 5395
+    },
+    {
+      "epoch": 0.37467018469656993,
+      "grad_norm": 4.589928742971696,
+      "learning_rate": 7.195650625971061e-06,
+      "loss": 0.5248,
+      "step": 5396
+    },
+    {
+      "epoch": 0.37473961949729206,
+      "grad_norm": 3.773999316956789,
+      "learning_rate": 7.194640304658509e-06,
+      "loss": 0.3507,
+      "step": 5397
+    },
+    {
+      "epoch": 0.37480905429801414,
+      "grad_norm": 4.510569538960003,
+      "learning_rate": 7.193629872343532e-06,
+      "loss": 0.4856,
+      "step": 5398
+    },
+    {
+      "epoch": 0.3748784890987363,
+      "grad_norm": 4.080387751371503,
+      "learning_rate": 7.192619329077238e-06,
+      "loss": 0.5011,
+      "step": 5399
+    },
+    {
+      "epoch": 0.3749479238994584,
+      "grad_norm": 3.6624930053487477,
+      "learning_rate": 7.191608674910741e-06,
+      "loss": 0.6582,
+      "step": 5400
+    },
+    {
+      "epoch": 0.37501735870018055,
+      "grad_norm": 48.04198102537829,
+      "learning_rate": 7.1905979098951564e-06,
+      "loss": 0.4876,
+      "step": 5401
+    },
+    {
+      "epoch": 0.3750867935009026,
+      "grad_norm": 4.6597997271489335,
+      "learning_rate": 7.189587034081607e-06,
+      "loss": 0.2864,
+      "step": 5402
+    },
+    {
+      "epoch": 0.37515622830162476,
+      "grad_norm": 4.146369748773861,
+      "learning_rate": 7.188576047521225e-06,
+      "loss": 0.4768,
+      "step": 5403
+    },
+    {
+      "epoch": 0.3752256631023469,
+      "grad_norm": 5.027219434348952,
+      "learning_rate": 7.187564950265142e-06,
+      "loss": 0.4588,
+      "step": 5404
+    },
+    {
+      "epoch": 0.37529509790306903,
+      "grad_norm": 2.904670121686568,
+      "learning_rate": 7.1865537423644995e-06,
+      "loss": 0.3735,
+      "step": 5405
+    },
+    {
+      "epoch": 0.37536453270379116,
+      "grad_norm": 3.3301729604548296,
+      "learning_rate": 7.185542423870444e-06,
+      "loss": 0.2556,
+      "step": 5406
+    },
+    {
+      "epoch": 0.37543396750451324,
+      "grad_norm": 3.14772773265129,
+      "learning_rate": 7.184530994834125e-06,
+      "loss": 0.3296,
+      "step": 5407
+    },
+    {
+      "epoch": 0.3755034023052354,
+      "grad_norm": 3.3737242326943875,
+      "learning_rate": 7.183519455306699e-06,
+      "loss": 0.3524,
+      "step": 5408
+    },
+    {
+      "epoch": 0.3755728371059575,
+      "grad_norm": 3.1365187846686298,
+      "learning_rate": 7.182507805339331e-06,
+      "loss": 0.2187,
+      "step": 5409
+    },
+    {
+      "epoch": 0.37564227190667965,
+      "grad_norm": 4.533940165369903,
+      "learning_rate": 7.181496044983189e-06,
+      "loss": 0.697,
+      "step": 5410
+    },
+    {
+      "epoch": 0.3757117067074017,
+      "grad_norm": 4.985615856094159,
+      "learning_rate": 7.1804841742894435e-06,
+      "loss": 0.2645,
+      "step": 5411
+    },
+    {
+      "epoch": 0.37578114150812386,
+      "grad_norm": 2.856238086371595,
+      "learning_rate": 7.179472193309276e-06,
+      "loss": 0.2854,
+      "step": 5412
+    },
+    {
+      "epoch": 0.375850576308846,
+      "grad_norm": 3.9332383109636284,
+      "learning_rate": 7.178460102093873e-06,
+      "loss": 0.5405,
+      "step": 5413
+    },
+    {
+      "epoch": 0.37592001110956813,
+      "grad_norm": 3.3940188691844666,
+      "learning_rate": 7.177447900694421e-06,
+      "loss": 0.326,
+      "step": 5414
+    },
+    {
+      "epoch": 0.3759894459102902,
+      "grad_norm": 4.015251333125671,
+      "learning_rate": 7.176435589162119e-06,
+      "loss": 0.4991,
+      "step": 5415
+    },
+    {
+      "epoch": 0.37605888071101234,
+      "grad_norm": 3.134419853083636,
+      "learning_rate": 7.175423167548168e-06,
+      "loss": 0.3621,
+      "step": 5416
+    },
+    {
+      "epoch": 0.3761283155117345,
+      "grad_norm": 2.997903168259594,
+      "learning_rate": 7.174410635903776e-06,
+      "loss": 0.3495,
+      "step": 5417
+    },
+    {
+      "epoch": 0.3761977503124566,
+      "grad_norm": 4.222126231618553,
+      "learning_rate": 7.173397994280153e-06,
+      "loss": 0.6791,
+      "step": 5418
+    },
+    {
+      "epoch": 0.37626718511317875,
+      "grad_norm": 4.953936237039759,
+      "learning_rate": 7.172385242728519e-06,
+      "loss": 0.7217,
+      "step": 5419
+    },
+    {
+      "epoch": 0.3763366199139008,
+      "grad_norm": 4.25229243122479,
+      "learning_rate": 7.171372381300098e-06,
+      "loss": 0.5345,
+      "step": 5420
+    },
+    {
+      "epoch": 0.37640605471462296,
+      "grad_norm": 3.6334790829745294,
+      "learning_rate": 7.1703594100461175e-06,
+      "loss": 0.3446,
+      "step": 5421
+    },
+    {
+      "epoch": 0.3764754895153451,
+      "grad_norm": 4.820496968330673,
+      "learning_rate": 7.169346329017816e-06,
+      "loss": 0.6859,
+      "step": 5422
+    },
+    {
+      "epoch": 0.37654492431606723,
+      "grad_norm": 3.2716435536380457,
+      "learning_rate": 7.168333138266431e-06,
+      "loss": 0.2991,
+      "step": 5423
+    },
+    {
+      "epoch": 0.3766143591167893,
+      "grad_norm": 5.7735154924825425,
+      "learning_rate": 7.167319837843209e-06,
+      "loss": 0.3938,
+      "step": 5424
+    },
+    {
+      "epoch": 0.37668379391751144,
+      "grad_norm": 4.281871000185905,
+      "learning_rate": 7.1663064277994035e-06,
+      "loss": 0.3823,
+      "step": 5425
+    },
+    {
+      "epoch": 0.3767532287182336,
+      "grad_norm": 3.9901090798417784,
+      "learning_rate": 7.16529290818627e-06,
+      "loss": 0.48,
+      "step": 5426
+    },
+    {
+      "epoch": 0.3768226635189557,
+      "grad_norm": 2.8722216748194054,
+      "learning_rate": 7.164279279055071e-06,
+      "loss": 0.292,
+      "step": 5427
+    },
+    {
+      "epoch": 0.37689209831967785,
+      "grad_norm": 4.298053844479467,
+      "learning_rate": 7.163265540457074e-06,
+      "loss": 0.6027,
+      "step": 5428
+    },
+    {
+      "epoch": 0.3769615331203999,
+      "grad_norm": 3.6531964034251043,
+      "learning_rate": 7.162251692443556e-06,
+      "loss": 0.3515,
+      "step": 5429
+    },
+    {
+      "epoch": 0.37703096792112206,
+      "grad_norm": 3.601747961125408,
+      "learning_rate": 7.161237735065791e-06,
+      "loss": 0.3833,
+      "step": 5430
+    },
+    {
+      "epoch": 0.3771004027218442,
+      "grad_norm": 3.620878583847293,
+      "learning_rate": 7.16022366837507e-06,
+      "loss": 0.2218,
+      "step": 5431
+    },
+    {
+      "epoch": 0.37716983752256633,
+      "grad_norm": 3.7121192672800842,
+      "learning_rate": 7.15920949242268e-06,
+      "loss": 0.3214,
+      "step": 5432
+    },
+    {
+      "epoch": 0.3772392723232884,
+      "grad_norm": 3.3064207946155495,
+      "learning_rate": 7.158195207259916e-06,
+      "loss": 0.4449,
+      "step": 5433
+    },
+    {
+      "epoch": 0.37730870712401055,
+      "grad_norm": 3.6858261543834967,
+      "learning_rate": 7.157180812938081e-06,
+      "loss": 0.6709,
+      "step": 5434
+    },
+    {
+      "epoch": 0.3773781419247327,
+      "grad_norm": 5.675707380817739,
+      "learning_rate": 7.156166309508482e-06,
+      "loss": 1.1867,
+      "step": 5435
+    },
+    {
+      "epoch": 0.3774475767254548,
+      "grad_norm": 3.89372978194419,
+      "learning_rate": 7.155151697022431e-06,
+      "loss": 0.3408,
+      "step": 5436
+    },
+    {
+      "epoch": 0.3775170115261769,
+      "grad_norm": 3.1947594731323297,
+      "learning_rate": 7.154136975531244e-06,
+      "loss": 0.2899,
+      "step": 5437
+    },
+    {
+      "epoch": 0.37758644632689903,
+      "grad_norm": 3.2902161588855328,
+      "learning_rate": 7.153122145086248e-06,
+      "loss": 0.3016,
+      "step": 5438
+    },
+    {
+      "epoch": 0.37765588112762116,
+      "grad_norm": 4.811639373918064,
+      "learning_rate": 7.15210720573877e-06,
+      "loss": 0.5154,
+      "step": 5439
+    },
+    {
+      "epoch": 0.3777253159283433,
+      "grad_norm": 3.8598214055407736,
+      "learning_rate": 7.151092157540145e-06,
+      "loss": 0.6604,
+      "step": 5440
+    },
+    {
+      "epoch": 0.37779475072906543,
+      "grad_norm": 3.123640282521276,
+      "learning_rate": 7.150077000541713e-06,
+      "loss": 0.4385,
+      "step": 5441
+    },
+    {
+      "epoch": 0.3778641855297875,
+      "grad_norm": 3.0631929095056063,
+      "learning_rate": 7.149061734794819e-06,
+      "loss": 0.323,
+      "step": 5442
+    },
+    {
+      "epoch": 0.37793362033050965,
+      "grad_norm": 3.3844866492038723,
+      "learning_rate": 7.148046360350814e-06,
+      "loss": 0.4488,
+      "step": 5443
+    },
+    {
+      "epoch": 0.3780030551312318,
+      "grad_norm": 5.284883612692944,
+      "learning_rate": 7.147030877261054e-06,
+      "loss": 0.5551,
+      "step": 5444
+    },
+    {
+      "epoch": 0.3780724899319539,
+      "grad_norm": 2.8763180907193617,
+      "learning_rate": 7.146015285576904e-06,
+      "loss": 0.202,
+      "step": 5445
+    },
+    {
+      "epoch": 0.378141924732676,
+      "grad_norm": 4.854673765515048,
+      "learning_rate": 7.144999585349728e-06,
+      "loss": 0.6433,
+      "step": 5446
+    },
+    {
+      "epoch": 0.37821135953339813,
+      "grad_norm": 2.6366804395372103,
+      "learning_rate": 7.143983776630899e-06,
+      "loss": 0.2114,
+      "step": 5447
+    },
+    {
+      "epoch": 0.37828079433412026,
+      "grad_norm": 4.2391565104810125,
+      "learning_rate": 7.142967859471799e-06,
+      "loss": 0.573,
+      "step": 5448
+    },
+    {
+      "epoch": 0.3783502291348424,
+      "grad_norm": 3.5483377237472067,
+      "learning_rate": 7.1419518339238085e-06,
+      "loss": 0.3062,
+      "step": 5449
+    },
+    {
+      "epoch": 0.37841966393556453,
+      "grad_norm": 3.5619749492052675,
+      "learning_rate": 7.140935700038316e-06,
+      "loss": 0.4807,
+      "step": 5450
+    },
+    {
+      "epoch": 0.3784890987362866,
+      "grad_norm": 5.327809889321563,
+      "learning_rate": 7.139919457866722e-06,
+      "loss": 0.6001,
+      "step": 5451
+    },
+    {
+      "epoch": 0.37855853353700875,
+      "grad_norm": 5.0798829698977075,
+      "learning_rate": 7.1389031074604224e-06,
+      "loss": 0.5704,
+      "step": 5452
+    },
+    {
+      "epoch": 0.3786279683377309,
+      "grad_norm": 3.0281918073975733,
+      "learning_rate": 7.1378866488708224e-06,
+      "loss": 0.2863,
+      "step": 5453
+    },
+    {
+      "epoch": 0.378697403138453,
+      "grad_norm": 4.0301619454856334,
+      "learning_rate": 7.136870082149335e-06,
+      "loss": 0.6036,
+      "step": 5454
+    },
+    {
+      "epoch": 0.3787668379391751,
+      "grad_norm": 3.8643286923765743,
+      "learning_rate": 7.135853407347378e-06,
+      "loss": 0.6244,
+      "step": 5455
+    },
+    {
+      "epoch": 0.37883627273989723,
+      "grad_norm": 5.1446638069854655,
+      "learning_rate": 7.134836624516372e-06,
+      "loss": 0.6473,
+      "step": 5456
+    },
+    {
+      "epoch": 0.37890570754061936,
+      "grad_norm": 3.1444419308767335,
+      "learning_rate": 7.133819733707746e-06,
+      "loss": 0.3929,
+      "step": 5457
+    },
+    {
+      "epoch": 0.3789751423413415,
+      "grad_norm": 4.856174748892138,
+      "learning_rate": 7.132802734972932e-06,
+      "loss": 0.5114,
+      "step": 5458
+    },
+    {
+      "epoch": 0.3790445771420636,
+      "grad_norm": 3.93955621427381,
+      "learning_rate": 7.131785628363368e-06,
+      "loss": 0.4291,
+      "step": 5459
+    },
+    {
+      "epoch": 0.3791140119427857,
+      "grad_norm": 3.9711469027849926,
+      "learning_rate": 7.1307684139305e-06,
+      "loss": 0.4916,
+      "step": 5460
+    },
+    {
+      "epoch": 0.37918344674350785,
+      "grad_norm": 4.051329616446848,
+      "learning_rate": 7.129751091725778e-06,
+      "loss": 0.4507,
+      "step": 5461
+    },
+    {
+      "epoch": 0.37925288154423,
+      "grad_norm": 3.6734105585592363,
+      "learning_rate": 7.1287336618006546e-06,
+      "loss": 0.3443,
+      "step": 5462
+    },
+    {
+      "epoch": 0.3793223163449521,
+      "grad_norm": 3.7985443047913594,
+      "learning_rate": 7.127716124206591e-06,
+      "loss": 0.5239,
+      "step": 5463
+    },
+    {
+      "epoch": 0.3793917511456742,
+      "grad_norm": 5.660547669819418,
+      "learning_rate": 7.126698478995054e-06,
+      "loss": 0.5432,
+      "step": 5464
+    },
+    {
+      "epoch": 0.37946118594639633,
+      "grad_norm": 3.282673283643919,
+      "learning_rate": 7.125680726217515e-06,
+      "loss": 0.3606,
+      "step": 5465
+    },
+    {
+      "epoch": 0.37953062074711846,
+      "grad_norm": 4.718901484530516,
+      "learning_rate": 7.124662865925449e-06,
+      "loss": 0.7748,
+      "step": 5466
+    },
+    {
+      "epoch": 0.3796000555478406,
+      "grad_norm": 3.371075923614832,
+      "learning_rate": 7.12364489817034e-06,
+      "loss": 0.4036,
+      "step": 5467
+    },
+    {
+      "epoch": 0.3796694903485627,
+      "grad_norm": 3.5911201724504793,
+      "learning_rate": 7.122626823003677e-06,
+      "loss": 0.5538,
+      "step": 5468
+    },
+    {
+      "epoch": 0.3797389251492848,
+      "grad_norm": 3.427194827738846,
+      "learning_rate": 7.121608640476948e-06,
+      "loss": 0.4817,
+      "step": 5469
+    },
+    {
+      "epoch": 0.37980835995000695,
+      "grad_norm": 3.291335486451126,
+      "learning_rate": 7.120590350641656e-06,
+      "loss": 0.3871,
+      "step": 5470
+    },
+    {
+      "epoch": 0.3798777947507291,
+      "grad_norm": 4.329793331690985,
+      "learning_rate": 7.119571953549305e-06,
+      "loss": 0.3769,
+      "step": 5471
+    },
+    {
+      "epoch": 0.37994722955145116,
+      "grad_norm": 5.010495033096167,
+      "learning_rate": 7.118553449251402e-06,
+      "loss": 0.4888,
+      "step": 5472
+    },
+    {
+      "epoch": 0.3800166643521733,
+      "grad_norm": 3.7227629233246335,
+      "learning_rate": 7.117534837799461e-06,
+      "loss": 0.2788,
+      "step": 5473
+    },
+    {
+      "epoch": 0.38008609915289543,
+      "grad_norm": 2.62153569879019,
+      "learning_rate": 7.116516119245007e-06,
+      "loss": 0.2344,
+      "step": 5474
+    },
+    {
+      "epoch": 0.38015553395361756,
+      "grad_norm": 4.28439689435706,
+      "learning_rate": 7.115497293639561e-06,
+      "loss": 0.7034,
+      "step": 5475
+    },
+    {
+      "epoch": 0.3802249687543397,
+      "grad_norm": 4.000524668621709,
+      "learning_rate": 7.1144783610346555e-06,
+      "loss": 0.6261,
+      "step": 5476
+    },
+    {
+      "epoch": 0.3802944035550618,
+      "grad_norm": 3.7112984820800374,
+      "learning_rate": 7.113459321481828e-06,
+      "loss": 0.4558,
+      "step": 5477
+    },
+    {
+      "epoch": 0.3803638383557839,
+      "grad_norm": 3.222052983332126,
+      "learning_rate": 7.11244017503262e-06,
+      "loss": 0.5072,
+      "step": 5478
+    },
+    {
+      "epoch": 0.38043327315650605,
+      "grad_norm": 4.168095180536021,
+      "learning_rate": 7.111420921738579e-06,
+      "loss": 0.4549,
+      "step": 5479
+    },
+    {
+      "epoch": 0.3805027079572282,
+      "grad_norm": 5.098181002899938,
+      "learning_rate": 7.1104015616512555e-06,
+      "loss": 0.9791,
+      "step": 5480
+    },
+    {
+      "epoch": 0.38057214275795026,
+      "grad_norm": 4.786685032851522,
+      "learning_rate": 7.1093820948222095e-06,
+      "loss": 0.8421,
+      "step": 5481
+    },
+    {
+      "epoch": 0.3806415775586724,
+      "grad_norm": 3.493777368207477,
+      "learning_rate": 7.108362521303003e-06,
+      "loss": 0.1085,
+      "step": 5482
+    },
+    {
+      "epoch": 0.38071101235939453,
+      "grad_norm": 4.243236676366345,
+      "learning_rate": 7.107342841145207e-06,
+      "loss": 0.4051,
+      "step": 5483
+    },
+    {
+      "epoch": 0.38078044716011666,
+      "grad_norm": 2.8855855180416308,
+      "learning_rate": 7.106323054400395e-06,
+      "loss": 0.3682,
+      "step": 5484
+    },
+    {
+      "epoch": 0.3808498819608388,
+      "grad_norm": 4.209320509335512,
+      "learning_rate": 7.105303161120146e-06,
+      "loss": 0.4553,
+      "step": 5485
+    },
+    {
+      "epoch": 0.3809193167615609,
+      "grad_norm": 3.842062861389581,
+      "learning_rate": 7.104283161356046e-06,
+      "loss": 0.5366,
+      "step": 5486
+    },
+    {
+      "epoch": 0.380988751562283,
+      "grad_norm": 3.8359673042540803,
+      "learning_rate": 7.103263055159684e-06,
+      "loss": 0.3823,
+      "step": 5487
+    },
+    {
+      "epoch": 0.38105818636300515,
+      "grad_norm": 4.4044078559972615,
+      "learning_rate": 7.102242842582658e-06,
+      "loss": 0.5356,
+      "step": 5488
+    },
+    {
+      "epoch": 0.3811276211637273,
+      "grad_norm": 3.268262353247625,
+      "learning_rate": 7.101222523676565e-06,
+      "loss": 0.2907,
+      "step": 5489
+    },
+    {
+      "epoch": 0.38119705596444936,
+      "grad_norm": 4.055831248213965,
+      "learning_rate": 7.100202098493017e-06,
+      "loss": 0.5756,
+      "step": 5490
+    },
+    {
+      "epoch": 0.3812664907651715,
+      "grad_norm": 2.825264798556411,
+      "learning_rate": 7.099181567083622e-06,
+      "loss": 0.3259,
+      "step": 5491
+    },
+    {
+      "epoch": 0.38133592556589363,
+      "grad_norm": 3.6947763603750765,
+      "learning_rate": 7.0981609294999994e-06,
+      "loss": 0.4522,
+      "step": 5492
+    },
+    {
+      "epoch": 0.38140536036661576,
+      "grad_norm": 2.8581579799599797,
+      "learning_rate": 7.097140185793771e-06,
+      "loss": 0.2625,
+      "step": 5493
+    },
+    {
+      "epoch": 0.38147479516733784,
+      "grad_norm": 3.982201048535751,
+      "learning_rate": 7.096119336016567e-06,
+      "loss": 0.4934,
+      "step": 5494
+    },
+    {
+      "epoch": 0.38154422996806,
+      "grad_norm": 4.771532021786225,
+      "learning_rate": 7.0950983802200165e-06,
+      "loss": 0.4685,
+      "step": 5495
+    },
+    {
+      "epoch": 0.3816136647687821,
+      "grad_norm": 3.632382522157554,
+      "learning_rate": 7.094077318455762e-06,
+      "loss": 0.4472,
+      "step": 5496
+    },
+    {
+      "epoch": 0.38168309956950425,
+      "grad_norm": 3.749120895626868,
+      "learning_rate": 7.093056150775446e-06,
+      "loss": 0.5335,
+      "step": 5497
+    },
+    {
+      "epoch": 0.3817525343702264,
+      "grad_norm": 3.7952118509261643,
+      "learning_rate": 7.092034877230717e-06,
+      "loss": 0.2069,
+      "step": 5498
+    },
+    {
+      "epoch": 0.38182196917094846,
+      "grad_norm": 4.5287487518908645,
+      "learning_rate": 7.091013497873232e-06,
+      "loss": 0.4937,
+      "step": 5499
+    },
+    {
+      "epoch": 0.3818914039716706,
+      "grad_norm": 5.142826672831739,
+      "learning_rate": 7.0899920127546514e-06,
+      "loss": 0.3951,
+      "step": 5500
+    },
+    {
+      "epoch": 0.38196083877239273,
+      "grad_norm": 3.433837542335874,
+      "learning_rate": 7.088970421926638e-06,
+      "loss": 0.3134,
+      "step": 5501
+    },
+    {
+      "epoch": 0.38203027357311486,
+      "grad_norm": 3.535113941660185,
+      "learning_rate": 7.087948725440867e-06,
+      "loss": 0.3935,
+      "step": 5502
+    },
+    {
+      "epoch": 0.38209970837383694,
+      "grad_norm": 3.8778537412139378,
+      "learning_rate": 7.086926923349011e-06,
+      "loss": 0.5613,
+      "step": 5503
+    },
+    {
+      "epoch": 0.3821691431745591,
+      "grad_norm": 3.100672672292877,
+      "learning_rate": 7.085905015702752e-06,
+      "loss": 0.3661,
+      "step": 5504
+    },
+    {
+      "epoch": 0.3822385779752812,
+      "grad_norm": 3.4649361142514876,
+      "learning_rate": 7.084883002553777e-06,
+      "loss": 0.5329,
+      "step": 5505
+    },
+    {
+      "epoch": 0.38230801277600335,
+      "grad_norm": 3.9367063808967306,
+      "learning_rate": 7.083860883953782e-06,
+      "loss": 0.3842,
+      "step": 5506
+    },
+    {
+      "epoch": 0.3823774475767255,
+      "grad_norm": 4.248300396352316,
+      "learning_rate": 7.082838659954459e-06,
+      "loss": 0.5383,
+      "step": 5507
+    },
+    {
+      "epoch": 0.38244688237744756,
+      "grad_norm": 4.619421953652103,
+      "learning_rate": 7.0818163306075135e-06,
+      "loss": 0.8405,
+      "step": 5508
+    },
+    {
+      "epoch": 0.3825163171781697,
+      "grad_norm": 3.7866627141924836,
+      "learning_rate": 7.080793895964655e-06,
+      "loss": 0.3803,
+      "step": 5509
+    },
+    {
+      "epoch": 0.38258575197889183,
+      "grad_norm": 3.5134780945763544,
+      "learning_rate": 7.079771356077595e-06,
+      "loss": 0.5887,
+      "step": 5510
+    },
+    {
+      "epoch": 0.38265518677961396,
+      "grad_norm": 3.6462447497931056,
+      "learning_rate": 7.078748710998053e-06,
+      "loss": 0.3331,
+      "step": 5511
+    },
+    {
+      "epoch": 0.38272462158033604,
+      "grad_norm": 3.5840849815821696,
+      "learning_rate": 7.0777259607777534e-06,
+      "loss": 0.3431,
+      "step": 5512
+    },
+    {
+      "epoch": 0.3827940563810582,
+      "grad_norm": 3.1535235820731193,
+      "learning_rate": 7.076703105468428e-06,
+      "loss": 0.3658,
+      "step": 5513
+    },
+    {
+      "epoch": 0.3828634911817803,
+      "grad_norm": 3.3595035133197473,
+      "learning_rate": 7.075680145121806e-06,
+      "loss": 0.3113,
+      "step": 5514
+    },
+    {
+      "epoch": 0.38293292598250245,
+      "grad_norm": 3.7859810344569085,
+      "learning_rate": 7.074657079789633e-06,
+      "loss": 0.4559,
+      "step": 5515
+    },
+    {
+      "epoch": 0.3830023607832245,
+      "grad_norm": 5.642204639597304,
+      "learning_rate": 7.073633909523653e-06,
+      "loss": 0.5079,
+      "step": 5516
+    },
+    {
+      "epoch": 0.38307179558394666,
+      "grad_norm": 4.35707192959465,
+      "learning_rate": 7.072610634375615e-06,
+      "loss": 0.6242,
+      "step": 5517
+    },
+    {
+      "epoch": 0.3831412303846688,
+      "grad_norm": 4.5397980050325,
+      "learning_rate": 7.071587254397278e-06,
+      "loss": 0.6895,
+      "step": 5518
+    },
+    {
+      "epoch": 0.38321066518539093,
+      "grad_norm": 3.5104830199609816,
+      "learning_rate": 7.070563769640401e-06,
+      "loss": 0.4902,
+      "step": 5519
+    },
+    {
+      "epoch": 0.38328009998611307,
+      "grad_norm": 4.792848552316435,
+      "learning_rate": 7.0695401801567505e-06,
+      "loss": 0.4915,
+      "step": 5520
+    },
+    {
+      "epoch": 0.38334953478683514,
+      "grad_norm": 4.953515852254025,
+      "learning_rate": 7.0685164859981e-06,
+      "loss": 0.6449,
+      "step": 5521
+    },
+    {
+      "epoch": 0.3834189695875573,
+      "grad_norm": 5.08388751143556,
+      "learning_rate": 7.067492687216226e-06,
+      "loss": 0.664,
+      "step": 5522
+    },
+    {
+      "epoch": 0.3834884043882794,
+      "grad_norm": 3.4777921883020797,
+      "learning_rate": 7.066468783862913e-06,
+      "loss": 0.2175,
+      "step": 5523
+    },
+    {
+      "epoch": 0.38355783918900155,
+      "grad_norm": 3.746619376685117,
+      "learning_rate": 7.065444775989947e-06,
+      "loss": 0.5288,
+      "step": 5524
+    },
+    {
+      "epoch": 0.3836272739897236,
+      "grad_norm": 4.590216361801467,
+      "learning_rate": 7.064420663649121e-06,
+      "loss": 0.4268,
+      "step": 5525
+    },
+    {
+      "epoch": 0.38369670879044576,
+      "grad_norm": 4.499353396360326,
+      "learning_rate": 7.063396446892235e-06,
+      "loss": 0.5877,
+      "step": 5526
+    },
+    {
+      "epoch": 0.3837661435911679,
+      "grad_norm": 4.775062380016619,
+      "learning_rate": 7.06237212577109e-06,
+      "loss": 0.6094,
+      "step": 5527
+    },
+    {
+      "epoch": 0.38383557839189003,
+      "grad_norm": 4.560545407395623,
+      "learning_rate": 7.0613477003374955e-06,
+      "loss": 0.5627,
+      "step": 5528
+    },
+    {
+      "epoch": 0.3839050131926121,
+      "grad_norm": 4.692329623314918,
+      "learning_rate": 7.06032317064327e-06,
+      "loss": 0.7517,
+      "step": 5529
+    },
+    {
+      "epoch": 0.38397444799333424,
+      "grad_norm": 5.156727926138188,
+      "learning_rate": 7.0592985367402264e-06,
+      "loss": 0.8359,
+      "step": 5530
+    },
+    {
+      "epoch": 0.3840438827940564,
+      "grad_norm": 3.5543404613190934,
+      "learning_rate": 7.058273798680195e-06,
+      "loss": 0.544,
+      "step": 5531
+    },
+    {
+      "epoch": 0.3841133175947785,
+      "grad_norm": 4.497729168733534,
+      "learning_rate": 7.057248956515005e-06,
+      "loss": 0.6553,
+      "step": 5532
+    },
+    {
+      "epoch": 0.38418275239550065,
+      "grad_norm": 4.585107535495594,
+      "learning_rate": 7.0562240102964894e-06,
+      "loss": 0.4542,
+      "step": 5533
+    },
+    {
+      "epoch": 0.3842521871962227,
+      "grad_norm": 4.276323784272024,
+      "learning_rate": 7.05519896007649e-06,
+      "loss": 0.5539,
+      "step": 5534
+    },
+    {
+      "epoch": 0.38432162199694486,
+      "grad_norm": 3.729650412082024,
+      "learning_rate": 7.0541738059068545e-06,
+      "loss": 0.4812,
+      "step": 5535
+    },
+    {
+      "epoch": 0.384391056797667,
+      "grad_norm": 3.413983618224626,
+      "learning_rate": 7.053148547839431e-06,
+      "loss": 0.5296,
+      "step": 5536
+    },
+    {
+      "epoch": 0.38446049159838913,
+      "grad_norm": 3.517262575650505,
+      "learning_rate": 7.052123185926077e-06,
+      "loss": 0.5011,
+      "step": 5537
+    },
+    {
+      "epoch": 0.3845299263991112,
+      "grad_norm": 6.451468643657486,
+      "learning_rate": 7.051097720218653e-06,
+      "loss": 1.024,
+      "step": 5538
+    },
+    {
+      "epoch": 0.38459936119983335,
+      "grad_norm": 4.692449896647971,
+      "learning_rate": 7.050072150769031e-06,
+      "loss": 0.6824,
+      "step": 5539
+    },
+    {
+      "epoch": 0.3846687960005555,
+      "grad_norm": 3.248528350180921,
+      "learning_rate": 7.049046477629077e-06,
+      "loss": 0.3442,
+      "step": 5540
+    },
+    {
+      "epoch": 0.3847382308012776,
+      "grad_norm": 5.052686408667282,
+      "learning_rate": 7.04802070085067e-06,
+      "loss": 0.75,
+      "step": 5541
+    },
+    {
+      "epoch": 0.38480766560199975,
+      "grad_norm": 3.719787411831542,
+      "learning_rate": 7.046994820485696e-06,
+      "loss": 0.4893,
+      "step": 5542
+    },
+    {
+      "epoch": 0.38487710040272183,
+      "grad_norm": 4.245469196283293,
+      "learning_rate": 7.045968836586039e-06,
+      "loss": 0.5536,
+      "step": 5543
+    },
+    {
+      "epoch": 0.38494653520344396,
+      "grad_norm": 2.5570080486129316,
+      "learning_rate": 7.044942749203593e-06,
+      "loss": 0.2408,
+      "step": 5544
+    },
+    {
+      "epoch": 0.3850159700041661,
+      "grad_norm": 3.2690842814637073,
+      "learning_rate": 7.043916558390258e-06,
+      "loss": 0.4979,
+      "step": 5545
+    },
+    {
+      "epoch": 0.38508540480488823,
+      "grad_norm": 3.453828813289731,
+      "learning_rate": 7.042890264197935e-06,
+      "loss": 0.441,
+      "step": 5546
+    },
+    {
+      "epoch": 0.3851548396056103,
+      "grad_norm": 4.055282436906723,
+      "learning_rate": 7.041863866678535e-06,
+      "loss": 0.5536,
+      "step": 5547
+    },
+    {
+      "epoch": 0.38522427440633245,
+      "grad_norm": 2.7667159071586314,
+      "learning_rate": 7.04083736588397e-06,
+      "loss": 0.2381,
+      "step": 5548
+    },
+    {
+      "epoch": 0.3852937092070546,
+      "grad_norm": 3.2417702348858817,
+      "learning_rate": 7.039810761866162e-06,
+      "loss": 0.3266,
+      "step": 5549
+    },
+    {
+      "epoch": 0.3853631440077767,
+      "grad_norm": 3.714801108540632,
+      "learning_rate": 7.038784054677033e-06,
+      "loss": 0.6171,
+      "step": 5550
+    },
+    {
+      "epoch": 0.3854325788084988,
+      "grad_norm": 2.4978252546068855,
+      "learning_rate": 7.037757244368514e-06,
+      "loss": 0.1695,
+      "step": 5551
+    },
+    {
+      "epoch": 0.38550201360922093,
+      "grad_norm": 3.7866144629789855,
+      "learning_rate": 7.036730330992538e-06,
+      "loss": 0.4029,
+      "step": 5552
+    },
+    {
+      "epoch": 0.38557144840994306,
+      "grad_norm": 4.732409396782647,
+      "learning_rate": 7.035703314601048e-06,
+      "loss": 0.5831,
+      "step": 5553
+    },
+    {
+      "epoch": 0.3856408832106652,
+      "grad_norm": 3.6181020923228013,
+      "learning_rate": 7.034676195245988e-06,
+      "loss": 0.4055,
+      "step": 5554
+    },
+    {
+      "epoch": 0.38571031801138733,
+      "grad_norm": 4.055514673968562,
+      "learning_rate": 7.033648972979308e-06,
+      "loss": 0.6137,
+      "step": 5555
+    },
+    {
+      "epoch": 0.3857797528121094,
+      "grad_norm": 4.5921542454519315,
+      "learning_rate": 7.032621647852964e-06,
+      "loss": 0.4806,
+      "step": 5556
+    },
+    {
+      "epoch": 0.38584918761283155,
+      "grad_norm": 4.198207802605822,
+      "learning_rate": 7.031594219918916e-06,
+      "loss": 0.3423,
+      "step": 5557
+    },
+    {
+      "epoch": 0.3859186224135537,
+      "grad_norm": 4.624994704116025,
+      "learning_rate": 7.030566689229133e-06,
+      "loss": 0.4937,
+      "step": 5558
+    },
+    {
+      "epoch": 0.3859880572142758,
+      "grad_norm": 2.9322745459068833,
+      "learning_rate": 7.0295390558355846e-06,
+      "loss": 0.3031,
+      "step": 5559
+    },
+    {
+      "epoch": 0.3860574920149979,
+      "grad_norm": 3.8228415611583246,
+      "learning_rate": 7.028511319790247e-06,
+      "loss": 0.5731,
+      "step": 5560
+    },
+    {
+      "epoch": 0.38612692681572003,
+      "grad_norm": 3.190057175191812,
+      "learning_rate": 7.027483481145102e-06,
+      "loss": 0.3455,
+      "step": 5561
+    },
+    {
+      "epoch": 0.38619636161644216,
+      "grad_norm": 4.1901651611560355,
+      "learning_rate": 7.026455539952136e-06,
+      "loss": 0.5482,
+      "step": 5562
+    },
+    {
+      "epoch": 0.3862657964171643,
+      "grad_norm": 5.919078359061228,
+      "learning_rate": 7.025427496263344e-06,
+      "loss": 0.6511,
+      "step": 5563
+    },
+    {
+      "epoch": 0.38633523121788643,
+      "grad_norm": 3.7727643451859714,
+      "learning_rate": 7.02439935013072e-06,
+      "loss": 0.4331,
+      "step": 5564
+    },
+    {
+      "epoch": 0.3864046660186085,
+      "grad_norm": 4.679877728127092,
+      "learning_rate": 7.023371101606267e-06,
+      "loss": 0.6575,
+      "step": 5565
+    },
+    {
+      "epoch": 0.38647410081933065,
+      "grad_norm": 3.3039380722154883,
+      "learning_rate": 7.022342750741993e-06,
+      "loss": 0.4654,
+      "step": 5566
+    },
+    {
+      "epoch": 0.3865435356200528,
+      "grad_norm": 6.645272198616555,
+      "learning_rate": 7.021314297589913e-06,
+      "loss": 0.4788,
+      "step": 5567
+    },
+    {
+      "epoch": 0.3866129704207749,
+      "grad_norm": 4.822798338025373,
+      "learning_rate": 7.0202857422020424e-06,
+      "loss": 0.5537,
+      "step": 5568
+    },
+    {
+      "epoch": 0.386682405221497,
+      "grad_norm": 3.2169703775892917,
+      "learning_rate": 7.019257084630404e-06,
+      "loss": 0.3326,
+      "step": 5569
+    },
+    {
+      "epoch": 0.38675184002221913,
+      "grad_norm": 3.1645760359552937,
+      "learning_rate": 7.018228324927027e-06,
+      "loss": 0.3592,
+      "step": 5570
+    },
+    {
+      "epoch": 0.38682127482294126,
+      "grad_norm": 4.018737878555018,
+      "learning_rate": 7.017199463143946e-06,
+      "loss": 0.4987,
+      "step": 5571
+    },
+    {
+      "epoch": 0.3868907096236634,
+      "grad_norm": 2.8093752159617997,
+      "learning_rate": 7.0161704993332e-06,
+      "loss": 0.4107,
+      "step": 5572
+    },
+    {
+      "epoch": 0.3869601444243855,
+      "grad_norm": 4.174055932966737,
+      "learning_rate": 7.015141433546829e-06,
+      "loss": 0.471,
+      "step": 5573
+    },
+    {
+      "epoch": 0.3870295792251076,
+      "grad_norm": 3.8015600366779454,
+      "learning_rate": 7.014112265836887e-06,
+      "loss": 0.4382,
+      "step": 5574
+    },
+    {
+      "epoch": 0.38709901402582975,
+      "grad_norm": 3.8721827049204744,
+      "learning_rate": 7.013082996255425e-06,
+      "loss": 0.3839,
+      "step": 5575
+    },
+    {
+      "epoch": 0.3871684488265519,
+      "grad_norm": 7.348275241277975,
+      "learning_rate": 7.012053624854503e-06,
+      "loss": 0.4103,
+      "step": 5576
+    },
+    {
+      "epoch": 0.387237883627274,
+      "grad_norm": 3.71490420962562,
+      "learning_rate": 7.011024151686185e-06,
+      "loss": 0.4849,
+      "step": 5577
+    },
+    {
+      "epoch": 0.3873073184279961,
+      "grad_norm": 3.717905137044848,
+      "learning_rate": 7.009994576802541e-06,
+      "loss": 0.3523,
+      "step": 5578
+    },
+    {
+      "epoch": 0.38737675322871823,
+      "grad_norm": 3.507506038575984,
+      "learning_rate": 7.008964900255646e-06,
+      "loss": 0.3487,
+      "step": 5579
+    },
+    {
+      "epoch": 0.38744618802944036,
+      "grad_norm": 3.340294639954744,
+      "learning_rate": 7.00793512209758e-06,
+      "loss": 0.2975,
+      "step": 5580
+    },
+    {
+      "epoch": 0.3875156228301625,
+      "grad_norm": 3.9263552243048125,
+      "learning_rate": 7.006905242380426e-06,
+      "loss": 0.5627,
+      "step": 5581
+    },
+    {
+      "epoch": 0.3875850576308846,
+      "grad_norm": 3.439278262923715,
+      "learning_rate": 7.005875261156276e-06,
+      "loss": 0.3546,
+      "step": 5582
+    },
+    {
+      "epoch": 0.3876544924316067,
+      "grad_norm": 4.274673593606483,
+      "learning_rate": 7.004845178477227e-06,
+      "loss": 0.6202,
+      "step": 5583
+    },
+    {
+      "epoch": 0.38772392723232885,
+      "grad_norm": 3.489918092315722,
+      "learning_rate": 7.003814994395378e-06,
+      "loss": 0.3566,
+      "step": 5584
+    },
+    {
+      "epoch": 0.387793362033051,
+      "grad_norm": 2.708956815756573,
+      "learning_rate": 7.002784708962832e-06,
+      "loss": 0.2368,
+      "step": 5585
+    },
+    {
+      "epoch": 0.38786279683377306,
+      "grad_norm": 4.315090205916436,
+      "learning_rate": 7.001754322231702e-06,
+      "loss": 0.358,
+      "step": 5586
+    },
+    {
+      "epoch": 0.3879322316344952,
+      "grad_norm": 5.724749626709329,
+      "learning_rate": 7.0007238342541035e-06,
+      "loss": 0.9594,
+      "step": 5587
+    },
+    {
+      "epoch": 0.38800166643521733,
+      "grad_norm": 7.165319967867814,
+      "learning_rate": 6.99969324508216e-06,
+      "loss": 0.6098,
+      "step": 5588
+    },
+    {
+      "epoch": 0.38807110123593946,
+      "grad_norm": 4.6523906368739825,
+      "learning_rate": 6.998662554767991e-06,
+      "loss": 0.764,
+      "step": 5589
+    },
+    {
+      "epoch": 0.3881405360366616,
+      "grad_norm": 3.4303971874574115,
+      "learning_rate": 6.9976317633637344e-06,
+      "loss": 0.3919,
+      "step": 5590
+    },
+    {
+      "epoch": 0.3882099708373837,
+      "grad_norm": 3.6018037527804823,
+      "learning_rate": 6.996600870921522e-06,
+      "loss": 0.3691,
+      "step": 5591
+    },
+    {
+      "epoch": 0.3882794056381058,
+      "grad_norm": 19.431526586127553,
+      "learning_rate": 6.995569877493498e-06,
+      "loss": 0.3072,
+      "step": 5592
+    },
+    {
+      "epoch": 0.38834884043882795,
+      "grad_norm": 3.86914378857059,
+      "learning_rate": 6.994538783131808e-06,
+      "loss": 0.5025,
+      "step": 5593
+    },
+    {
+      "epoch": 0.3884182752395501,
+      "grad_norm": 3.5008499951980943,
+      "learning_rate": 6.993507587888602e-06,
+      "loss": 0.324,
+      "step": 5594
+    },
+    {
+      "epoch": 0.38848771004027216,
+      "grad_norm": 3.68821762968492,
+      "learning_rate": 6.992476291816039e-06,
+      "loss": 0.5203,
+      "step": 5595
+    },
+    {
+      "epoch": 0.3885571448409943,
+      "grad_norm": 4.420635092630538,
+      "learning_rate": 6.991444894966279e-06,
+      "loss": 0.5721,
+      "step": 5596
+    },
+    {
+      "epoch": 0.38862657964171643,
+      "grad_norm": 4.009226130677225,
+      "learning_rate": 6.990413397391491e-06,
+      "loss": 0.5236,
+      "step": 5597
+    },
+    {
+      "epoch": 0.38869601444243856,
+      "grad_norm": 3.7039231455641466,
+      "learning_rate": 6.989381799143844e-06,
+      "loss": 0.4255,
+      "step": 5598
+    },
+    {
+      "epoch": 0.3887654492431607,
+      "grad_norm": 2.554638186231122,
+      "learning_rate": 6.9883501002755175e-06,
+      "loss": 0.1971,
+      "step": 5599
+    },
+    {
+      "epoch": 0.3888348840438828,
+      "grad_norm": 3.8927945369170023,
+      "learning_rate": 6.987318300838695e-06,
+      "loss": 0.6669,
+      "step": 5600
+    },
+    {
+      "epoch": 0.3889043188446049,
+      "grad_norm": 3.1765062018794157,
+      "learning_rate": 6.986286400885558e-06,
+      "loss": 0.4318,
+      "step": 5601
+    },
+    {
+      "epoch": 0.38897375364532705,
+      "grad_norm": 4.712948994945242,
+      "learning_rate": 6.985254400468305e-06,
+      "loss": 0.6575,
+      "step": 5602
+    },
+    {
+      "epoch": 0.3890431884460492,
+      "grad_norm": 4.662775590173519,
+      "learning_rate": 6.984222299639129e-06,
+      "loss": 0.6473,
+      "step": 5603
+    },
+    {
+      "epoch": 0.38911262324677126,
+      "grad_norm": 4.340494894506122,
+      "learning_rate": 6.983190098450235e-06,
+      "loss": 0.4358,
+      "step": 5604
+    },
+    {
+      "epoch": 0.3891820580474934,
+      "grad_norm": 2.5312331926975187,
+      "learning_rate": 6.98215779695383e-06,
+      "loss": 0.2963,
+      "step": 5605
+    },
+    {
+      "epoch": 0.38925149284821553,
+      "grad_norm": 4.308918574518346,
+      "learning_rate": 6.981125395202128e-06,
+      "loss": 0.6722,
+      "step": 5606
+    },
+    {
+      "epoch": 0.38932092764893766,
+      "grad_norm": 4.860550664419271,
+      "learning_rate": 6.980092893247344e-06,
+      "loss": 0.823,
+      "step": 5607
+    },
+    {
+      "epoch": 0.38939036244965974,
+      "grad_norm": 4.371834572942315,
+      "learning_rate": 6.979060291141702e-06,
+      "loss": 0.6467,
+      "step": 5608
+    },
+    {
+      "epoch": 0.3894597972503819,
+      "grad_norm": 3.5341300384562904,
+      "learning_rate": 6.9780275889374305e-06,
+      "loss": 0.5019,
+      "step": 5609
+    },
+    {
+      "epoch": 0.389529232051104,
+      "grad_norm": 4.056427588271549,
+      "learning_rate": 6.976994786686761e-06,
+      "loss": 0.4976,
+      "step": 5610
+    },
+    {
+      "epoch": 0.38959866685182615,
+      "grad_norm": 3.776327125606446,
+      "learning_rate": 6.975961884441933e-06,
+      "loss": 0.4237,
+      "step": 5611
+    },
+    {
+      "epoch": 0.3896681016525483,
+      "grad_norm": 2.8983650089052917,
+      "learning_rate": 6.974928882255189e-06,
+      "loss": 0.4039,
+      "step": 5612
+    },
+    {
+      "epoch": 0.38973753645327036,
+      "grad_norm": 4.204035839604379,
+      "learning_rate": 6.973895780178776e-06,
+      "loss": 0.4017,
+      "step": 5613
+    },
+    {
+      "epoch": 0.3898069712539925,
+      "grad_norm": 2.880651194525737,
+      "learning_rate": 6.9728625782649495e-06,
+      "loss": 0.2168,
+      "step": 5614
+    },
+    {
+      "epoch": 0.38987640605471463,
+      "grad_norm": 4.44286454813566,
+      "learning_rate": 6.971829276565964e-06,
+      "loss": 0.5611,
+      "step": 5615
+    },
+    {
+      "epoch": 0.38994584085543676,
+      "grad_norm": 4.001242223190559,
+      "learning_rate": 6.970795875134088e-06,
+      "loss": 0.5831,
+      "step": 5616
+    },
+    {
+      "epoch": 0.39001527565615884,
+      "grad_norm": 2.382641154340544,
+      "learning_rate": 6.969762374021585e-06,
+      "loss": 0.1705,
+      "step": 5617
+    },
+    {
+      "epoch": 0.390084710456881,
+      "grad_norm": 4.263278390666253,
+      "learning_rate": 6.96872877328073e-06,
+      "loss": 0.4376,
+      "step": 5618
+    },
+    {
+      "epoch": 0.3901541452576031,
+      "grad_norm": 3.289177508651344,
+      "learning_rate": 6.967695072963802e-06,
+      "loss": 0.3664,
+      "step": 5619
+    },
+    {
+      "epoch": 0.39022358005832525,
+      "grad_norm": 3.4325002230462744,
+      "learning_rate": 6.966661273123083e-06,
+      "loss": 0.6771,
+      "step": 5620
+    },
+    {
+      "epoch": 0.3902930148590473,
+      "grad_norm": 3.8667911670375363,
+      "learning_rate": 6.965627373810863e-06,
+      "loss": 0.483,
+      "step": 5621
+    },
+    {
+      "epoch": 0.39036244965976946,
+      "grad_norm": 3.718252455372809,
+      "learning_rate": 6.964593375079435e-06,
+      "loss": 0.2595,
+      "step": 5622
+    },
+    {
+      "epoch": 0.3904318844604916,
+      "grad_norm": 3.929784413371683,
+      "learning_rate": 6.963559276981095e-06,
+      "loss": 0.5785,
+      "step": 5623
+    },
+    {
+      "epoch": 0.39050131926121373,
+      "grad_norm": 3.5372707404233057,
+      "learning_rate": 6.962525079568151e-06,
+      "loss": 0.459,
+      "step": 5624
+    },
+    {
+      "epoch": 0.39057075406193587,
+      "grad_norm": 5.114464347641802,
+      "learning_rate": 6.961490782892908e-06,
+      "loss": 0.686,
+      "step": 5625
+    },
+    {
+      "epoch": 0.39064018886265794,
+      "grad_norm": 3.92918219508859,
+      "learning_rate": 6.960456387007682e-06,
+      "loss": 0.4886,
+      "step": 5626
+    },
+    {
+      "epoch": 0.3907096236633801,
+      "grad_norm": 4.884722372586597,
+      "learning_rate": 6.95942189196479e-06,
+      "loss": 0.5602,
+      "step": 5627
+    },
+    {
+      "epoch": 0.3907790584641022,
+      "grad_norm": 4.150849891568653,
+      "learning_rate": 6.9583872978165534e-06,
+      "loss": 0.7654,
+      "step": 5628
+    },
+    {
+      "epoch": 0.39084849326482435,
+      "grad_norm": 4.8439816219451535,
+      "learning_rate": 6.957352604615307e-06,
+      "loss": 0.7586,
+      "step": 5629
+    },
+    {
+      "epoch": 0.3909179280655464,
+      "grad_norm": 4.348139707625982,
+      "learning_rate": 6.956317812413378e-06,
+      "loss": 0.3261,
+      "step": 5630
+    },
+    {
+      "epoch": 0.39098736286626856,
+      "grad_norm": 4.469305718657237,
+      "learning_rate": 6.955282921263109e-06,
+      "loss": 0.5415,
+      "step": 5631
+    },
+    {
+      "epoch": 0.3910567976669907,
+      "grad_norm": 3.849947284197216,
+      "learning_rate": 6.9542479312168436e-06,
+      "loss": 0.4641,
+      "step": 5632
+    },
+    {
+      "epoch": 0.39112623246771283,
+      "grad_norm": 3.4145016591109085,
+      "learning_rate": 6.953212842326927e-06,
+      "loss": 0.3429,
+      "step": 5633
+    },
+    {
+      "epoch": 0.39119566726843497,
+      "grad_norm": 3.2896478984667,
+      "learning_rate": 6.952177654645717e-06,
+      "loss": 0.3205,
+      "step": 5634
+    },
+    {
+      "epoch": 0.39126510206915704,
+      "grad_norm": 3.974683380064343,
+      "learning_rate": 6.95114236822557e-06,
+      "loss": 0.4984,
+      "step": 5635
+    },
+    {
+      "epoch": 0.3913345368698792,
+      "grad_norm": 5.161437814693137,
+      "learning_rate": 6.950106983118848e-06,
+      "loss": 0.5833,
+      "step": 5636
+    },
+    {
+      "epoch": 0.3914039716706013,
+      "grad_norm": 3.073619093326718,
+      "learning_rate": 6.949071499377922e-06,
+      "loss": 0.2787,
+      "step": 5637
+    },
+    {
+      "epoch": 0.39147340647132345,
+      "grad_norm": 3.0741303349864437,
+      "learning_rate": 6.9480359170551655e-06,
+      "loss": 0.3635,
+      "step": 5638
+    },
+    {
+      "epoch": 0.3915428412720455,
+      "grad_norm": 5.1197127438759535,
+      "learning_rate": 6.947000236202958e-06,
+      "loss": 0.417,
+      "step": 5639
+    },
+    {
+      "epoch": 0.39161227607276766,
+      "grad_norm": 4.851343354653931,
+      "learning_rate": 6.9459644568736805e-06,
+      "loss": 0.4673,
+      "step": 5640
+    },
+    {
+      "epoch": 0.3916817108734898,
+      "grad_norm": 4.34458764056399,
+      "learning_rate": 6.944928579119723e-06,
+      "loss": 0.6188,
+      "step": 5641
+    },
+    {
+      "epoch": 0.39175114567421193,
+      "grad_norm": 1.892028433648046,
+      "learning_rate": 6.943892602993478e-06,
+      "loss": 0.0983,
+      "step": 5642
+    },
+    {
+      "epoch": 0.391820580474934,
+      "grad_norm": 4.16143930583073,
+      "learning_rate": 6.942856528547346e-06,
+      "loss": 0.3165,
+      "step": 5643
+    },
+    {
+      "epoch": 0.39189001527565615,
+      "grad_norm": 3.368090511867243,
+      "learning_rate": 6.941820355833728e-06,
+      "loss": 0.3535,
+      "step": 5644
+    },
+    {
+      "epoch": 0.3919594500763783,
+      "grad_norm": 3.2248980383005175,
+      "learning_rate": 6.940784084905037e-06,
+      "loss": 0.4341,
+      "step": 5645
+    },
+    {
+      "epoch": 0.3920288848771004,
+      "grad_norm": 4.304220058262542,
+      "learning_rate": 6.939747715813679e-06,
+      "loss": 0.6458,
+      "step": 5646
+    },
+    {
+      "epoch": 0.39209831967782255,
+      "grad_norm": 3.6405002356229637,
+      "learning_rate": 6.938711248612078e-06,
+      "loss": 0.2951,
+      "step": 5647
+    },
+    {
+      "epoch": 0.39216775447854463,
+      "grad_norm": 3.1226667823843623,
+      "learning_rate": 6.9376746833526565e-06,
+      "loss": 0.1876,
+      "step": 5648
+    },
+    {
+      "epoch": 0.39223718927926676,
+      "grad_norm": 3.7314782416779204,
+      "learning_rate": 6.936638020087842e-06,
+      "loss": 0.5664,
+      "step": 5649
+    },
+    {
+      "epoch": 0.3923066240799889,
+      "grad_norm": 3.748428245801171,
+      "learning_rate": 6.935601258870067e-06,
+      "loss": 0.3878,
+      "step": 5650
+    },
+    {
+      "epoch": 0.39237605888071103,
+      "grad_norm": 3.651862971980098,
+      "learning_rate": 6.934564399751772e-06,
+      "loss": 0.57,
+      "step": 5651
+    },
+    {
+      "epoch": 0.3924454936814331,
+      "grad_norm": 4.286191481096184,
+      "learning_rate": 6.933527442785397e-06,
+      "loss": 0.2867,
+      "step": 5652
+    },
+    {
+      "epoch": 0.39251492848215525,
+      "grad_norm": 4.31259829474449,
+      "learning_rate": 6.932490388023393e-06,
+      "loss": 0.5817,
+      "step": 5653
+    },
+    {
+      "epoch": 0.3925843632828774,
+      "grad_norm": 4.059726610183235,
+      "learning_rate": 6.931453235518212e-06,
+      "loss": 0.4275,
+      "step": 5654
+    },
+    {
+      "epoch": 0.3926537980835995,
+      "grad_norm": 4.147220750858426,
+      "learning_rate": 6.930415985322312e-06,
+      "loss": 0.8832,
+      "step": 5655
+    },
+    {
+      "epoch": 0.39272323288432165,
+      "grad_norm": 3.6406309757150543,
+      "learning_rate": 6.929378637488156e-06,
+      "loss": 0.466,
+      "step": 5656
+    },
+    {
+      "epoch": 0.39279266768504373,
+      "grad_norm": 2.528006067096268,
+      "learning_rate": 6.928341192068211e-06,
+      "loss": 0.2144,
+      "step": 5657
+    },
+    {
+      "epoch": 0.39286210248576586,
+      "grad_norm": 3.5934751826163924,
+      "learning_rate": 6.9273036491149506e-06,
+      "loss": 0.56,
+      "step": 5658
+    },
+    {
+      "epoch": 0.392931537286488,
+      "grad_norm": 4.103672026055777,
+      "learning_rate": 6.926266008680853e-06,
+      "loss": 0.4996,
+      "step": 5659
+    },
+    {
+      "epoch": 0.39300097208721013,
+      "grad_norm": 3.7559624270861898,
+      "learning_rate": 6.9252282708183995e-06,
+      "loss": 0.4962,
+      "step": 5660
+    },
+    {
+      "epoch": 0.3930704068879322,
+      "grad_norm": 3.941428510568149,
+      "learning_rate": 6.92419043558008e-06,
+      "loss": 0.5785,
+      "step": 5661
+    },
+    {
+      "epoch": 0.39313984168865435,
+      "grad_norm": 4.499825759436411,
+      "learning_rate": 6.923152503018384e-06,
+      "loss": 0.7415,
+      "step": 5662
+    },
+    {
+      "epoch": 0.3932092764893765,
+      "grad_norm": 3.122930677618771,
+      "learning_rate": 6.922114473185812e-06,
+      "loss": 0.4815,
+      "step": 5663
+    },
+    {
+      "epoch": 0.3932787112900986,
+      "grad_norm": 3.2400275808225887,
+      "learning_rate": 6.9210763461348655e-06,
+      "loss": 0.4084,
+      "step": 5664
+    },
+    {
+      "epoch": 0.3933481460908207,
+      "grad_norm": 3.5740210898751714,
+      "learning_rate": 6.920038121918049e-06,
+      "loss": 0.3103,
+      "step": 5665
+    },
+    {
+      "epoch": 0.39341758089154283,
+      "grad_norm": 3.242407917280528,
+      "learning_rate": 6.9189998005878776e-06,
+      "loss": 0.2125,
+      "step": 5666
+    },
+    {
+      "epoch": 0.39348701569226496,
+      "grad_norm": 3.3168840406506113,
+      "learning_rate": 6.9179613821968695e-06,
+      "loss": 0.3734,
+      "step": 5667
+    },
+    {
+      "epoch": 0.3935564504929871,
+      "grad_norm": 3.822325628234687,
+      "learning_rate": 6.916922866797544e-06,
+      "loss": 0.4396,
+      "step": 5668
+    },
+    {
+      "epoch": 0.39362588529370923,
+      "grad_norm": 2.979766358154854,
+      "learning_rate": 6.915884254442429e-06,
+      "loss": 0.2746,
+      "step": 5669
+    },
+    {
+      "epoch": 0.3936953200944313,
+      "grad_norm": 3.7979847729435434,
+      "learning_rate": 6.914845545184056e-06,
+      "loss": 0.3301,
+      "step": 5670
+    },
+    {
+      "epoch": 0.39376475489515345,
+      "grad_norm": 4.384429971176278,
+      "learning_rate": 6.913806739074962e-06,
+      "loss": 0.4338,
+      "step": 5671
+    },
+    {
+      "epoch": 0.3938341896958756,
+      "grad_norm": 5.4920167477982265,
+      "learning_rate": 6.9127678361676896e-06,
+      "loss": 0.5428,
+      "step": 5672
+    },
+    {
+      "epoch": 0.3939036244965977,
+      "grad_norm": 4.32585380348086,
+      "learning_rate": 6.911728836514784e-06,
+      "loss": 0.665,
+      "step": 5673
+    },
+    {
+      "epoch": 0.3939730592973198,
+      "grad_norm": 3.8688090807236257,
+      "learning_rate": 6.910689740168798e-06,
+      "loss": 0.3709,
+      "step": 5674
+    },
+    {
+      "epoch": 0.39404249409804193,
+      "grad_norm": 3.916852042272188,
+      "learning_rate": 6.909650547182286e-06,
+      "loss": 0.3611,
+      "step": 5675
+    },
+    {
+      "epoch": 0.39411192889876406,
+      "grad_norm": 3.250147561970421,
+      "learning_rate": 6.90861125760781e-06,
+      "loss": 0.2622,
+      "step": 5676
+    },
+    {
+      "epoch": 0.3941813636994862,
+      "grad_norm": 3.0542572483972177,
+      "learning_rate": 6.9075718714979375e-06,
+      "loss": 0.3407,
+      "step": 5677
+    },
+    {
+      "epoch": 0.3942507985002083,
+      "grad_norm": 2.744717701354929,
+      "learning_rate": 6.906532388905237e-06,
+      "loss": 0.1949,
+      "step": 5678
+    },
+    {
+      "epoch": 0.3943202333009304,
+      "grad_norm": 3.7925690102687164,
+      "learning_rate": 6.905492809882286e-06,
+      "loss": 0.5389,
+      "step": 5679
+    },
+    {
+      "epoch": 0.39438966810165255,
+      "grad_norm": 3.4525734113339515,
+      "learning_rate": 6.9044531344816655e-06,
+      "loss": 0.4237,
+      "step": 5680
+    },
+    {
+      "epoch": 0.3944591029023747,
+      "grad_norm": 3.073162612849862,
+      "learning_rate": 6.903413362755961e-06,
+      "loss": 0.257,
+      "step": 5681
+    },
+    {
+      "epoch": 0.3945285377030968,
+      "grad_norm": 4.2202133215026985,
+      "learning_rate": 6.902373494757761e-06,
+      "loss": 0.4427,
+      "step": 5682
+    },
+    {
+      "epoch": 0.3945979725038189,
+      "grad_norm": 4.706320370791534,
+      "learning_rate": 6.901333530539662e-06,
+      "loss": 0.7231,
+      "step": 5683
+    },
+    {
+      "epoch": 0.39466740730454103,
+      "grad_norm": 3.7565743947962935,
+      "learning_rate": 6.900293470154267e-06,
+      "loss": 0.494,
+      "step": 5684
+    },
+    {
+      "epoch": 0.39473684210526316,
+      "grad_norm": 3.7580221314488385,
+      "learning_rate": 6.899253313654176e-06,
+      "loss": 0.4698,
+      "step": 5685
+    },
+    {
+      "epoch": 0.3948062769059853,
+      "grad_norm": 4.283663052222066,
+      "learning_rate": 6.898213061092003e-06,
+      "loss": 0.4602,
+      "step": 5686
+    },
+    {
+      "epoch": 0.3948757117067074,
+      "grad_norm": 3.128887990541601,
+      "learning_rate": 6.897172712520361e-06,
+      "loss": 0.3616,
+      "step": 5687
+    },
+    {
+      "epoch": 0.3949451465074295,
+      "grad_norm": 2.3506992899499646,
+      "learning_rate": 6.896132267991871e-06,
+      "loss": 0.2614,
+      "step": 5688
+    },
+    {
+      "epoch": 0.39501458130815165,
+      "grad_norm": 3.2703354348261233,
+      "learning_rate": 6.895091727559153e-06,
+      "loss": 0.4132,
+      "step": 5689
+    },
+    {
+      "epoch": 0.3950840161088738,
+      "grad_norm": 3.0196151914221456,
+      "learning_rate": 6.894051091274845e-06,
+      "loss": 0.4032,
+      "step": 5690
+    },
+    {
+      "epoch": 0.3951534509095959,
+      "grad_norm": 3.699039380100635,
+      "learning_rate": 6.893010359191572e-06,
+      "loss": 0.5582,
+      "step": 5691
+    },
+    {
+      "epoch": 0.395222885710318,
+      "grad_norm": 4.283148583781468,
+      "learning_rate": 6.891969531361979e-06,
+      "loss": 0.5212,
+      "step": 5692
+    },
+    {
+      "epoch": 0.39529232051104013,
+      "grad_norm": 3.403303670303957,
+      "learning_rate": 6.890928607838708e-06,
+      "loss": 0.3134,
+      "step": 5693
+    },
+    {
+      "epoch": 0.39536175531176226,
+      "grad_norm": 3.0504556519539663,
+      "learning_rate": 6.889887588674407e-06,
+      "loss": 0.2064,
+      "step": 5694
+    },
+    {
+      "epoch": 0.3954311901124844,
+      "grad_norm": 3.5711469171689862,
+      "learning_rate": 6.888846473921731e-06,
+      "loss": 0.5361,
+      "step": 5695
+    },
+    {
+      "epoch": 0.3955006249132065,
+      "grad_norm": 2.7723326193925906,
+      "learning_rate": 6.887805263633337e-06,
+      "loss": 0.2763,
+      "step": 5696
+    },
+    {
+      "epoch": 0.3955700597139286,
+      "grad_norm": 3.4297016123590174,
+      "learning_rate": 6.886763957861889e-06,
+      "loss": 0.4801,
+      "step": 5697
+    },
+    {
+      "epoch": 0.39563949451465075,
+      "grad_norm": 3.7451022007737786,
+      "learning_rate": 6.885722556660055e-06,
+      "loss": 0.4784,
+      "step": 5698
+    },
+    {
+      "epoch": 0.3957089293153729,
+      "grad_norm": 5.7728960260090405,
+      "learning_rate": 6.884681060080508e-06,
+      "loss": 0.5532,
+      "step": 5699
+    },
+    {
+      "epoch": 0.39577836411609496,
+      "grad_norm": 3.6806774963214837,
+      "learning_rate": 6.883639468175926e-06,
+      "loss": 0.2997,
+      "step": 5700
+    },
+    {
+      "epoch": 0.3958477989168171,
+      "grad_norm": 4.2502706128469505,
+      "learning_rate": 6.882597780998991e-06,
+      "loss": 0.5236,
+      "step": 5701
+    },
+    {
+      "epoch": 0.39591723371753923,
+      "grad_norm": 4.807145252930331,
+      "learning_rate": 6.8815559986023905e-06,
+      "loss": 0.6043,
+      "step": 5702
+    },
+    {
+      "epoch": 0.39598666851826136,
+      "grad_norm": 2.2817857365450567,
+      "learning_rate": 6.880514121038817e-06,
+      "loss": 0.2168,
+      "step": 5703
+    },
+    {
+      "epoch": 0.3960561033189835,
+      "grad_norm": 2.627355321915552,
+      "learning_rate": 6.8794721483609675e-06,
+      "loss": 0.1961,
+      "step": 5704
+    },
+    {
+      "epoch": 0.3961255381197056,
+      "grad_norm": 3.481442034571039,
+      "learning_rate": 6.878430080621542e-06,
+      "loss": 0.3637,
+      "step": 5705
+    },
+    {
+      "epoch": 0.3961949729204277,
+      "grad_norm": 4.109572521635142,
+      "learning_rate": 6.877387917873251e-06,
+      "loss": 0.545,
+      "step": 5706
+    },
+    {
+      "epoch": 0.39626440772114985,
+      "grad_norm": 4.529743450045566,
+      "learning_rate": 6.8763456601688016e-06,
+      "loss": 0.7499,
+      "step": 5707
+    },
+    {
+      "epoch": 0.396333842521872,
+      "grad_norm": 3.7313108160221815,
+      "learning_rate": 6.875303307560912e-06,
+      "loss": 0.4675,
+      "step": 5708
+    },
+    {
+      "epoch": 0.39640327732259406,
+      "grad_norm": 3.380360971431975,
+      "learning_rate": 6.874260860102304e-06,
+      "loss": 0.4114,
+      "step": 5709
+    },
+    {
+      "epoch": 0.3964727121233162,
+      "grad_norm": 4.414549822904587,
+      "learning_rate": 6.873218317845702e-06,
+      "loss": 0.5324,
+      "step": 5710
+    },
+    {
+      "epoch": 0.39654214692403833,
+      "grad_norm": 3.8894110901931835,
+      "learning_rate": 6.872175680843838e-06,
+      "loss": 0.7147,
+      "step": 5711
+    },
+    {
+      "epoch": 0.39661158172476046,
+      "grad_norm": 2.909511194718064,
+      "learning_rate": 6.871132949149446e-06,
+      "loss": 0.3169,
+      "step": 5712
+    },
+    {
+      "epoch": 0.3966810165254826,
+      "grad_norm": 3.804742124671051,
+      "learning_rate": 6.8700901228152676e-06,
+      "loss": 0.451,
+      "step": 5713
+    },
+    {
+      "epoch": 0.3967504513262047,
+      "grad_norm": 2.962890617307222,
+      "learning_rate": 6.869047201894044e-06,
+      "loss": 0.299,
+      "step": 5714
+    },
+    {
+      "epoch": 0.3968198861269268,
+      "grad_norm": 2.9025558713792647,
+      "learning_rate": 6.868004186438531e-06,
+      "loss": 0.3069,
+      "step": 5715
+    },
+    {
+      "epoch": 0.39688932092764895,
+      "grad_norm": 3.9509765110919703,
+      "learning_rate": 6.866961076501479e-06,
+      "loss": 0.4061,
+      "step": 5716
+    },
+    {
+      "epoch": 0.3969587557283711,
+      "grad_norm": 3.249056966087549,
+      "learning_rate": 6.8659178721356475e-06,
+      "loss": 0.5183,
+      "step": 5717
+    },
+    {
+      "epoch": 0.39702819052909316,
+      "grad_norm": 4.238189500010239,
+      "learning_rate": 6.864874573393803e-06,
+      "loss": 0.5372,
+      "step": 5718
+    },
+    {
+      "epoch": 0.3970976253298153,
+      "grad_norm": 3.5986005768163736,
+      "learning_rate": 6.863831180328713e-06,
+      "loss": 0.5113,
+      "step": 5719
+    },
+    {
+      "epoch": 0.39716706013053743,
+      "grad_norm": 4.118884875098035,
+      "learning_rate": 6.862787692993149e-06,
+      "loss": 0.6818,
+      "step": 5720
+    },
+    {
+      "epoch": 0.39723649493125956,
+      "grad_norm": 3.7038356343748293,
+      "learning_rate": 6.861744111439892e-06,
+      "loss": 0.4052,
+      "step": 5721
+    },
+    {
+      "epoch": 0.39730592973198164,
+      "grad_norm": 3.0483853210755716,
+      "learning_rate": 6.8607004357217265e-06,
+      "loss": 0.477,
+      "step": 5722
+    },
+    {
+      "epoch": 0.3973753645327038,
+      "grad_norm": 3.346963108688846,
+      "learning_rate": 6.8596566658914365e-06,
+      "loss": 0.3977,
+      "step": 5723
+    },
+    {
+      "epoch": 0.3974447993334259,
+      "grad_norm": 4.5038583792909055,
+      "learning_rate": 6.858612802001817e-06,
+      "loss": 0.6457,
+      "step": 5724
+    },
+    {
+      "epoch": 0.39751423413414805,
+      "grad_norm": 3.5704439308258538,
+      "learning_rate": 6.857568844105665e-06,
+      "loss": 0.3142,
+      "step": 5725
+    },
+    {
+      "epoch": 0.3975836689348702,
+      "grad_norm": 3.352613496352378,
+      "learning_rate": 6.856524792255783e-06,
+      "loss": 0.2664,
+      "step": 5726
+    },
+    {
+      "epoch": 0.39765310373559226,
+      "grad_norm": 4.935145846798327,
+      "learning_rate": 6.855480646504978e-06,
+      "loss": 0.8577,
+      "step": 5727
+    },
+    {
+      "epoch": 0.3977225385363144,
+      "grad_norm": 4.235744910050579,
+      "learning_rate": 6.854436406906062e-06,
+      "loss": 0.7081,
+      "step": 5728
+    },
+    {
+      "epoch": 0.39779197333703653,
+      "grad_norm": 3.088193450647446,
+      "learning_rate": 6.853392073511851e-06,
+      "loss": 0.2973,
+      "step": 5729
+    },
+    {
+      "epoch": 0.39786140813775867,
+      "grad_norm": 4.656690591011675,
+      "learning_rate": 6.852347646375165e-06,
+      "loss": 0.8398,
+      "step": 5730
+    },
+    {
+      "epoch": 0.39793084293848074,
+      "grad_norm": 4.014157080101815,
+      "learning_rate": 6.851303125548832e-06,
+      "loss": 0.3277,
+      "step": 5731
+    },
+    {
+      "epoch": 0.3980002777392029,
+      "grad_norm": 3.563924655667442,
+      "learning_rate": 6.850258511085681e-06,
+      "loss": 0.5612,
+      "step": 5732
+    },
+    {
+      "epoch": 0.398069712539925,
+      "grad_norm": 3.8358159251406607,
+      "learning_rate": 6.849213803038549e-06,
+      "loss": 0.5555,
+      "step": 5733
+    },
+    {
+      "epoch": 0.39813914734064715,
+      "grad_norm": 3.664411173842289,
+      "learning_rate": 6.848169001460276e-06,
+      "loss": 0.4186,
+      "step": 5734
+    },
+    {
+      "epoch": 0.3982085821413692,
+      "grad_norm": 3.3591401129858527,
+      "learning_rate": 6.847124106403705e-06,
+      "loss": 0.4952,
+      "step": 5735
+    },
+    {
+      "epoch": 0.39827801694209136,
+      "grad_norm": 4.627110099346733,
+      "learning_rate": 6.846079117921688e-06,
+      "loss": 0.5048,
+      "step": 5736
+    },
+    {
+      "epoch": 0.3983474517428135,
+      "grad_norm": 2.5849897856317483,
+      "learning_rate": 6.8450340360670764e-06,
+      "loss": 0.2822,
+      "step": 5737
+    },
+    {
+      "epoch": 0.39841688654353563,
+      "grad_norm": 2.4415686443865092,
+      "learning_rate": 6.843988860892733e-06,
+      "loss": 0.2027,
+      "step": 5738
+    },
+    {
+      "epoch": 0.39848632134425777,
+      "grad_norm": 4.293609472217636,
+      "learning_rate": 6.8429435924515184e-06,
+      "loss": 0.7884,
+      "step": 5739
+    },
+    {
+      "epoch": 0.39855575614497984,
+      "grad_norm": 3.840276490537884,
+      "learning_rate": 6.841898230796302e-06,
+      "loss": 0.3357,
+      "step": 5740
+    },
+    {
+      "epoch": 0.398625190945702,
+      "grad_norm": 6.584551300194193,
+      "learning_rate": 6.840852775979958e-06,
+      "loss": 0.5519,
+      "step": 5741
+    },
+    {
+      "epoch": 0.3986946257464241,
+      "grad_norm": 4.155648557139466,
+      "learning_rate": 6.839807228055364e-06,
+      "loss": 0.6205,
+      "step": 5742
+    },
+    {
+      "epoch": 0.39876406054714625,
+      "grad_norm": 3.9743670204833865,
+      "learning_rate": 6.838761587075402e-06,
+      "loss": 0.5917,
+      "step": 5743
+    },
+    {
+      "epoch": 0.3988334953478683,
+      "grad_norm": 3.2416449999328663,
+      "learning_rate": 6.837715853092959e-06,
+      "loss": 0.519,
+      "step": 5744
+    },
+    {
+      "epoch": 0.39890293014859046,
+      "grad_norm": 3.556474667587179,
+      "learning_rate": 6.83667002616093e-06,
+      "loss": 0.4837,
+      "step": 5745
+    },
+    {
+      "epoch": 0.3989723649493126,
+      "grad_norm": 4.604360051831115,
+      "learning_rate": 6.835624106332206e-06,
+      "loss": 0.5177,
+      "step": 5746
+    },
+    {
+      "epoch": 0.39904179975003473,
+      "grad_norm": 2.7360673463044645,
+      "learning_rate": 6.834578093659694e-06,
+      "loss": 0.334,
+      "step": 5747
+    },
+    {
+      "epoch": 0.39911123455075687,
+      "grad_norm": 4.14502600402918,
+      "learning_rate": 6.833531988196298e-06,
+      "loss": 0.634,
+      "step": 5748
+    },
+    {
+      "epoch": 0.39918066935147895,
+      "grad_norm": 3.5076968645883717,
+      "learning_rate": 6.8324857899949285e-06,
+      "loss": 0.4232,
+      "step": 5749
+    },
+    {
+      "epoch": 0.3992501041522011,
+      "grad_norm": 4.138182094059944,
+      "learning_rate": 6.831439499108501e-06,
+      "loss": 0.3635,
+      "step": 5750
+    },
+    {
+      "epoch": 0.3993195389529232,
+      "grad_norm": 3.012203454688716,
+      "learning_rate": 6.830393115589936e-06,
+      "loss": 0.3278,
+      "step": 5751
+    },
+    {
+      "epoch": 0.39938897375364535,
+      "grad_norm": 3.756862358094155,
+      "learning_rate": 6.8293466394921595e-06,
+      "loss": 0.3345,
+      "step": 5752
+    },
+    {
+      "epoch": 0.39945840855436743,
+      "grad_norm": 3.407170541963846,
+      "learning_rate": 6.828300070868098e-06,
+      "loss": 0.407,
+      "step": 5753
+    },
+    {
+      "epoch": 0.39952784335508956,
+      "grad_norm": 11.645297484858022,
+      "learning_rate": 6.827253409770689e-06,
+      "loss": 0.2956,
+      "step": 5754
+    },
+    {
+      "epoch": 0.3995972781558117,
+      "grad_norm": 3.4056288681042672,
+      "learning_rate": 6.82620665625287e-06,
+      "loss": 0.3145,
+      "step": 5755
+    },
+    {
+      "epoch": 0.39966671295653383,
+      "grad_norm": 3.5116101156797828,
+      "learning_rate": 6.825159810367585e-06,
+      "loss": 0.5053,
+      "step": 5756
+    },
+    {
+      "epoch": 0.3997361477572559,
+      "grad_norm": 4.5644216140444565,
+      "learning_rate": 6.8241128721677825e-06,
+      "loss": 0.4335,
+      "step": 5757
+    },
+    {
+      "epoch": 0.39980558255797805,
+      "grad_norm": 4.104930749672699,
+      "learning_rate": 6.823065841706415e-06,
+      "loss": 0.6777,
+      "step": 5758
+    },
+    {
+      "epoch": 0.3998750173587002,
+      "grad_norm": 2.7627837793095114,
+      "learning_rate": 6.822018719036439e-06,
+      "loss": 0.2426,
+      "step": 5759
+    },
+    {
+      "epoch": 0.3999444521594223,
+      "grad_norm": 3.805381598568102,
+      "learning_rate": 6.820971504210818e-06,
+      "loss": 0.4598,
+      "step": 5760
+    },
+    {
+      "epoch": 0.40001388696014445,
+      "grad_norm": 4.175901637400691,
+      "learning_rate": 6.819924197282521e-06,
+      "loss": 0.4677,
+      "step": 5761
+    },
+    {
+      "epoch": 0.40008332176086653,
+      "grad_norm": 4.489454257235447,
+      "learning_rate": 6.818876798304516e-06,
+      "loss": 0.5574,
+      "step": 5762
+    },
+    {
+      "epoch": 0.40015275656158866,
+      "grad_norm": 3.858224049626737,
+      "learning_rate": 6.817829307329781e-06,
+      "loss": 0.482,
+      "step": 5763
+    },
+    {
+      "epoch": 0.4002221913623108,
+      "grad_norm": 4.165831120904523,
+      "learning_rate": 6.816781724411296e-06,
+      "loss": 0.5596,
+      "step": 5764
+    },
+    {
+      "epoch": 0.40029162616303293,
+      "grad_norm": 4.035193447051069,
+      "learning_rate": 6.815734049602048e-06,
+      "loss": 0.6447,
+      "step": 5765
+    },
+    {
+      "epoch": 0.400361060963755,
+      "grad_norm": 3.2983159184469812,
+      "learning_rate": 6.814686282955026e-06,
+      "loss": 0.3218,
+      "step": 5766
+    },
+    {
+      "epoch": 0.40043049576447715,
+      "grad_norm": 2.23306171781236,
+      "learning_rate": 6.813638424523226e-06,
+      "loss": 0.1493,
+      "step": 5767
+    },
+    {
+      "epoch": 0.4004999305651993,
+      "grad_norm": 3.65837406842894,
+      "learning_rate": 6.812590474359646e-06,
+      "loss": 0.3744,
+      "step": 5768
+    },
+    {
+      "epoch": 0.4005693653659214,
+      "grad_norm": 2.6245185124045602,
+      "learning_rate": 6.81154243251729e-06,
+      "loss": 0.2114,
+      "step": 5769
+    },
+    {
+      "epoch": 0.40063880016664355,
+      "grad_norm": 3.177351778945432,
+      "learning_rate": 6.8104942990491694e-06,
+      "loss": 0.4593,
+      "step": 5770
+    },
+    {
+      "epoch": 0.40070823496736563,
+      "grad_norm": 4.050022592112915,
+      "learning_rate": 6.809446074008295e-06,
+      "loss": 0.4178,
+      "step": 5771
+    },
+    {
+      "epoch": 0.40077766976808776,
+      "grad_norm": 2.617423169864227,
+      "learning_rate": 6.808397757447687e-06,
+      "loss": 0.3036,
+      "step": 5772
+    },
+    {
+      "epoch": 0.4008471045688099,
+      "grad_norm": 3.095521385865647,
+      "learning_rate": 6.807349349420363e-06,
+      "loss": 0.3924,
+      "step": 5773
+    },
+    {
+      "epoch": 0.40091653936953203,
+      "grad_norm": 3.7010854328070404,
+      "learning_rate": 6.806300849979359e-06,
+      "loss": 0.2906,
+      "step": 5774
+    },
+    {
+      "epoch": 0.4009859741702541,
+      "grad_norm": 4.00599377103177,
+      "learning_rate": 6.8052522591776995e-06,
+      "loss": 0.4504,
+      "step": 5775
+    },
+    {
+      "epoch": 0.40105540897097625,
+      "grad_norm": 4.104953870184914,
+      "learning_rate": 6.804203577068422e-06,
+      "loss": 0.5452,
+      "step": 5776
+    },
+    {
+      "epoch": 0.4011248437716984,
+      "grad_norm": 3.7746641240075545,
+      "learning_rate": 6.803154803704572e-06,
+      "loss": 0.4431,
+      "step": 5777
+    },
+    {
+      "epoch": 0.4011942785724205,
+      "grad_norm": 4.219222140235012,
+      "learning_rate": 6.802105939139192e-06,
+      "loss": 0.4746,
+      "step": 5778
+    },
+    {
+      "epoch": 0.4012637133731426,
+      "grad_norm": 4.294727074809847,
+      "learning_rate": 6.801056983425331e-06,
+      "loss": 0.4823,
+      "step": 5779
+    },
+    {
+      "epoch": 0.40133314817386473,
+      "grad_norm": 3.416377662392865,
+      "learning_rate": 6.800007936616048e-06,
+      "loss": 0.3535,
+      "step": 5780
+    },
+    {
+      "epoch": 0.40140258297458686,
+      "grad_norm": 3.1025720405610833,
+      "learning_rate": 6.798958798764399e-06,
+      "loss": 0.3409,
+      "step": 5781
+    },
+    {
+      "epoch": 0.401472017775309,
+      "grad_norm": 3.386928657199894,
+      "learning_rate": 6.797909569923449e-06,
+      "loss": 0.4774,
+      "step": 5782
+    },
+    {
+      "epoch": 0.40154145257603113,
+      "grad_norm": 3.7299661140802507,
+      "learning_rate": 6.7968602501462696e-06,
+      "loss": 0.2652,
+      "step": 5783
+    },
+    {
+      "epoch": 0.4016108873767532,
+      "grad_norm": 3.43756468272369,
+      "learning_rate": 6.795810839485931e-06,
+      "loss": 0.4033,
+      "step": 5784
+    },
+    {
+      "epoch": 0.40168032217747535,
+      "grad_norm": 4.07862473644909,
+      "learning_rate": 6.7947613379955114e-06,
+      "loss": 0.4492,
+      "step": 5785
+    },
+    {
+      "epoch": 0.4017497569781975,
+      "grad_norm": 3.0305183260104065,
+      "learning_rate": 6.7937117457280955e-06,
+      "loss": 0.3218,
+      "step": 5786
+    },
+    {
+      "epoch": 0.4018191917789196,
+      "grad_norm": 3.7100359060060866,
+      "learning_rate": 6.792662062736768e-06,
+      "loss": 0.4045,
+      "step": 5787
+    },
+    {
+      "epoch": 0.4018886265796417,
+      "grad_norm": 3.089562047783084,
+      "learning_rate": 6.791612289074623e-06,
+      "loss": 0.2837,
+      "step": 5788
+    },
+    {
+      "epoch": 0.40195806138036383,
+      "grad_norm": 2.5363350142854926,
+      "learning_rate": 6.790562424794754e-06,
+      "loss": 0.1855,
+      "step": 5789
+    },
+    {
+      "epoch": 0.40202749618108596,
+      "grad_norm": 4.793365727999766,
+      "learning_rate": 6.789512469950265e-06,
+      "loss": 0.5399,
+      "step": 5790
+    },
+    {
+      "epoch": 0.4020969309818081,
+      "grad_norm": 3.205774524442511,
+      "learning_rate": 6.788462424594262e-06,
+      "loss": 0.3727,
+      "step": 5791
+    },
+    {
+      "epoch": 0.4021663657825302,
+      "grad_norm": 4.38948906733427,
+      "learning_rate": 6.78741228877985e-06,
+      "loss": 0.545,
+      "step": 5792
+    },
+    {
+      "epoch": 0.4022358005832523,
+      "grad_norm": 3.4987646306342093,
+      "learning_rate": 6.78636206256015e-06,
+      "loss": 0.3759,
+      "step": 5793
+    },
+    {
+      "epoch": 0.40230523538397445,
+      "grad_norm": 4.258628120400825,
+      "learning_rate": 6.785311745988276e-06,
+      "loss": 0.5523,
+      "step": 5794
+    },
+    {
+      "epoch": 0.4023746701846966,
+      "grad_norm": 5.674581198108077,
+      "learning_rate": 6.784261339117355e-06,
+      "loss": 0.5127,
+      "step": 5795
+    },
+    {
+      "epoch": 0.4024441049854187,
+      "grad_norm": 4.434488510400074,
+      "learning_rate": 6.783210842000515e-06,
+      "loss": 0.6336,
+      "step": 5796
+    },
+    {
+      "epoch": 0.4025135397861408,
+      "grad_norm": 5.663311344543477,
+      "learning_rate": 6.7821602546908884e-06,
+      "loss": 0.7357,
+      "step": 5797
+    },
+    {
+      "epoch": 0.40258297458686293,
+      "grad_norm": 3.287741894596281,
+      "learning_rate": 6.7811095772416116e-06,
+      "loss": 0.3827,
+      "step": 5798
+    },
+    {
+      "epoch": 0.40265240938758506,
+      "grad_norm": 4.051059167590617,
+      "learning_rate": 6.780058809705827e-06,
+      "loss": 0.5029,
+      "step": 5799
+    },
+    {
+      "epoch": 0.4027218441883072,
+      "grad_norm": 4.724989369449416,
+      "learning_rate": 6.7790079521366856e-06,
+      "loss": 0.5271,
+      "step": 5800
+    },
+    {
+      "epoch": 0.4027912789890293,
+      "grad_norm": 3.4425312288034995,
+      "learning_rate": 6.777957004587332e-06,
+      "loss": 0.4991,
+      "step": 5801
+    },
+    {
+      "epoch": 0.4028607137897514,
+      "grad_norm": 4.834440885483448,
+      "learning_rate": 6.776905967110927e-06,
+      "loss": 0.4887,
+      "step": 5802
+    },
+    {
+      "epoch": 0.40293014859047355,
+      "grad_norm": 3.7880383589498443,
+      "learning_rate": 6.775854839760627e-06,
+      "loss": 0.4994,
+      "step": 5803
+    },
+    {
+      "epoch": 0.4029995833911957,
+      "grad_norm": 3.961305044181077,
+      "learning_rate": 6.774803622589601e-06,
+      "loss": 0.489,
+      "step": 5804
+    },
+    {
+      "epoch": 0.4030690181919178,
+      "grad_norm": 4.279406877610342,
+      "learning_rate": 6.773752315651014e-06,
+      "loss": 0.6022,
+      "step": 5805
+    },
+    {
+      "epoch": 0.4031384529926399,
+      "grad_norm": 4.321755319561774,
+      "learning_rate": 6.772700918998043e-06,
+      "loss": 0.6169,
+      "step": 5806
+    },
+    {
+      "epoch": 0.40320788779336203,
+      "grad_norm": 3.286839295697232,
+      "learning_rate": 6.771649432683866e-06,
+      "loss": 0.3646,
+      "step": 5807
+    },
+    {
+      "epoch": 0.40327732259408416,
+      "grad_norm": 4.034684149937206,
+      "learning_rate": 6.770597856761664e-06,
+      "loss": 0.347,
+      "step": 5808
+    },
+    {
+      "epoch": 0.4033467573948063,
+      "grad_norm": 3.169383136423241,
+      "learning_rate": 6.7695461912846265e-06,
+      "loss": 0.2739,
+      "step": 5809
+    },
+    {
+      "epoch": 0.4034161921955284,
+      "grad_norm": 4.05296309115515,
+      "learning_rate": 6.768494436305946e-06,
+      "loss": 0.4014,
+      "step": 5810
+    },
+    {
+      "epoch": 0.4034856269962505,
+      "grad_norm": 4.280778602000879,
+      "learning_rate": 6.767442591878817e-06,
+      "loss": 0.4736,
+      "step": 5811
+    },
+    {
+      "epoch": 0.40355506179697265,
+      "grad_norm": 2.8342990434506956,
+      "learning_rate": 6.766390658056442e-06,
+      "loss": 0.2555,
+      "step": 5812
+    },
+    {
+      "epoch": 0.4036244965976948,
+      "grad_norm": 3.655331189146411,
+      "learning_rate": 6.765338634892027e-06,
+      "loss": 0.3159,
+      "step": 5813
+    },
+    {
+      "epoch": 0.40369393139841686,
+      "grad_norm": 2.689161924230765,
+      "learning_rate": 6.764286522438779e-06,
+      "loss": 0.3332,
+      "step": 5814
+    },
+    {
+      "epoch": 0.403763366199139,
+      "grad_norm": 3.8326487770047533,
+      "learning_rate": 6.763234320749917e-06,
+      "loss": 0.3385,
+      "step": 5815
+    },
+    {
+      "epoch": 0.40383280099986113,
+      "grad_norm": 3.0041302486605157,
+      "learning_rate": 6.76218202987866e-06,
+      "loss": 0.3453,
+      "step": 5816
+    },
+    {
+      "epoch": 0.40390223580058326,
+      "grad_norm": 3.309314467306892,
+      "learning_rate": 6.7611296498782264e-06,
+      "loss": 0.3419,
+      "step": 5817
+    },
+    {
+      "epoch": 0.4039716706013054,
+      "grad_norm": 3.6594349458454065,
+      "learning_rate": 6.76007718080185e-06,
+      "loss": 0.3718,
+      "step": 5818
+    },
+    {
+      "epoch": 0.4040411054020275,
+      "grad_norm": 4.845944680218625,
+      "learning_rate": 6.759024622702761e-06,
+      "loss": 0.5115,
+      "step": 5819
+    },
+    {
+      "epoch": 0.4041105402027496,
+      "grad_norm": 4.622515238858883,
+      "learning_rate": 6.757971975634199e-06,
+      "loss": 0.6267,
+      "step": 5820
+    },
+    {
+      "epoch": 0.40417997500347175,
+      "grad_norm": 5.081849847481002,
+      "learning_rate": 6.7569192396493995e-06,
+      "loss": 0.5335,
+      "step": 5821
+    },
+    {
+      "epoch": 0.4042494098041939,
+      "grad_norm": 3.385219280763738,
+      "learning_rate": 6.755866414801617e-06,
+      "loss": 0.3789,
+      "step": 5822
+    },
+    {
+      "epoch": 0.40431884460491596,
+      "grad_norm": 4.214311919192354,
+      "learning_rate": 6.754813501144096e-06,
+      "loss": 0.6231,
+      "step": 5823
+    },
+    {
+      "epoch": 0.4043882794056381,
+      "grad_norm": 3.484737866613268,
+      "learning_rate": 6.753760498730095e-06,
+      "loss": 0.2889,
+      "step": 5824
+    },
+    {
+      "epoch": 0.40445771420636023,
+      "grad_norm": 4.11068589343891,
+      "learning_rate": 6.7527074076128716e-06,
+      "loss": 0.6504,
+      "step": 5825
+    },
+    {
+      "epoch": 0.40452714900708237,
+      "grad_norm": 6.988661685370977,
+      "learning_rate": 6.751654227845692e-06,
+      "loss": 0.9208,
+      "step": 5826
+    },
+    {
+      "epoch": 0.40459658380780444,
+      "grad_norm": 3.671144525136897,
+      "learning_rate": 6.750600959481822e-06,
+      "loss": 0.4628,
+      "step": 5827
+    },
+    {
+      "epoch": 0.4046660186085266,
+      "grad_norm": 3.259757705149907,
+      "learning_rate": 6.749547602574537e-06,
+      "loss": 0.339,
+      "step": 5828
+    },
+    {
+      "epoch": 0.4047354534092487,
+      "grad_norm": 4.357809595916123,
+      "learning_rate": 6.748494157177114e-06,
+      "loss": 0.584,
+      "step": 5829
+    },
+    {
+      "epoch": 0.40480488820997085,
+      "grad_norm": 4.380261056317368,
+      "learning_rate": 6.747440623342835e-06,
+      "loss": 0.5702,
+      "step": 5830
+    },
+    {
+      "epoch": 0.404874323010693,
+      "grad_norm": 4.215734879340497,
+      "learning_rate": 6.746387001124987e-06,
+      "loss": 0.6637,
+      "step": 5831
+    },
+    {
+      "epoch": 0.40494375781141506,
+      "grad_norm": 4.567374733643246,
+      "learning_rate": 6.745333290576863e-06,
+      "loss": 0.4943,
+      "step": 5832
+    },
+    {
+      "epoch": 0.4050131926121372,
+      "grad_norm": 2.897764971602156,
+      "learning_rate": 6.744279491751753e-06,
+      "loss": 0.3262,
+      "step": 5833
+    },
+    {
+      "epoch": 0.40508262741285933,
+      "grad_norm": 3.411687608336122,
+      "learning_rate": 6.743225604702961e-06,
+      "loss": 0.4164,
+      "step": 5834
+    },
+    {
+      "epoch": 0.40515206221358147,
+      "grad_norm": 4.958622337848191,
+      "learning_rate": 6.74217162948379e-06,
+      "loss": 0.6992,
+      "step": 5835
+    },
+    {
+      "epoch": 0.40522149701430354,
+      "grad_norm": 4.084297381718783,
+      "learning_rate": 6.741117566147551e-06,
+      "loss": 0.6105,
+      "step": 5836
+    },
+    {
+      "epoch": 0.4052909318150257,
+      "grad_norm": 4.605534157165234,
+      "learning_rate": 6.740063414747553e-06,
+      "loss": 0.5499,
+      "step": 5837
+    },
+    {
+      "epoch": 0.4053603666157478,
+      "grad_norm": 3.2796469338575327,
+      "learning_rate": 6.739009175337119e-06,
+      "loss": 0.2975,
+      "step": 5838
+    },
+    {
+      "epoch": 0.40542980141646995,
+      "grad_norm": 4.318903704469736,
+      "learning_rate": 6.7379548479695655e-06,
+      "loss": 0.5882,
+      "step": 5839
+    },
+    {
+      "epoch": 0.4054992362171921,
+      "grad_norm": 2.6683786096547664,
+      "learning_rate": 6.736900432698226e-06,
+      "loss": 0.244,
+      "step": 5840
+    },
+    {
+      "epoch": 0.40556867101791416,
+      "grad_norm": 2.8937504455784286,
+      "learning_rate": 6.735845929576425e-06,
+      "loss": 0.2444,
+      "step": 5841
+    },
+    {
+      "epoch": 0.4056381058186363,
+      "grad_norm": 4.6382827542910725,
+      "learning_rate": 6.734791338657505e-06,
+      "loss": 0.3108,
+      "step": 5842
+    },
+    {
+      "epoch": 0.40570754061935843,
+      "grad_norm": 3.43594571415395,
+      "learning_rate": 6.733736659994799e-06,
+      "loss": 0.2983,
+      "step": 5843
+    },
+    {
+      "epoch": 0.40577697542008057,
+      "grad_norm": 3.05742724271604,
+      "learning_rate": 6.732681893641655e-06,
+      "loss": 0.2937,
+      "step": 5844
+    },
+    {
+      "epoch": 0.40584641022080264,
+      "grad_norm": 2.7629751818574095,
+      "learning_rate": 6.731627039651423e-06,
+      "loss": 0.2259,
+      "step": 5845
+    },
+    {
+      "epoch": 0.4059158450215248,
+      "grad_norm": 5.038036689219401,
+      "learning_rate": 6.7305720980774525e-06,
+      "loss": 0.2739,
+      "step": 5846
+    },
+    {
+      "epoch": 0.4059852798222469,
+      "grad_norm": 3.788951390224794,
+      "learning_rate": 6.729517068973106e-06,
+      "loss": 0.4617,
+      "step": 5847
+    },
+    {
+      "epoch": 0.40605471462296905,
+      "grad_norm": 3.525201091172801,
+      "learning_rate": 6.728461952391742e-06,
+      "loss": 0.394,
+      "step": 5848
+    },
+    {
+      "epoch": 0.4061241494236911,
+      "grad_norm": 3.8455628464808966,
+      "learning_rate": 6.727406748386728e-06,
+      "loss": 0.5284,
+      "step": 5849
+    },
+    {
+      "epoch": 0.40619358422441326,
+      "grad_norm": 4.295510608276293,
+      "learning_rate": 6.726351457011437e-06,
+      "loss": 0.6841,
+      "step": 5850
+    },
+    {
+      "epoch": 0.4062630190251354,
+      "grad_norm": 4.647480687292294,
+      "learning_rate": 6.725296078319241e-06,
+      "loss": 0.5685,
+      "step": 5851
+    },
+    {
+      "epoch": 0.40633245382585753,
+      "grad_norm": 3.9092716055630294,
+      "learning_rate": 6.724240612363521e-06,
+      "loss": 0.4853,
+      "step": 5852
+    },
+    {
+      "epoch": 0.40640188862657967,
+      "grad_norm": 2.997315850629966,
+      "learning_rate": 6.723185059197663e-06,
+      "loss": 0.3067,
+      "step": 5853
+    },
+    {
+      "epoch": 0.40647132342730175,
+      "grad_norm": 3.0311459090407276,
+      "learning_rate": 6.7221294188750555e-06,
+      "loss": 0.32,
+      "step": 5854
+    },
+    {
+      "epoch": 0.4065407582280239,
+      "grad_norm": 4.281366380606873,
+      "learning_rate": 6.72107369144909e-06,
+      "loss": 0.4652,
+      "step": 5855
+    },
+    {
+      "epoch": 0.406610193028746,
+      "grad_norm": 5.490409653600115,
+      "learning_rate": 6.720017876973165e-06,
+      "loss": 0.2866,
+      "step": 5856
+    },
+    {
+      "epoch": 0.40667962782946815,
+      "grad_norm": 3.3538252204615118,
+      "learning_rate": 6.718961975500682e-06,
+      "loss": 0.2365,
+      "step": 5857
+    },
+    {
+      "epoch": 0.40674906263019023,
+      "grad_norm": 4.367921246647332,
+      "learning_rate": 6.7179059870850475e-06,
+      "loss": 0.6269,
+      "step": 5858
+    },
+    {
+      "epoch": 0.40681849743091236,
+      "grad_norm": 4.617943948107425,
+      "learning_rate": 6.716849911779672e-06,
+      "loss": 0.7698,
+      "step": 5859
+    },
+    {
+      "epoch": 0.4068879322316345,
+      "grad_norm": 4.250952030354651,
+      "learning_rate": 6.71579374963797e-06,
+      "loss": 0.5047,
+      "step": 5860
+    },
+    {
+      "epoch": 0.40695736703235663,
+      "grad_norm": 4.180005023947055,
+      "learning_rate": 6.714737500713363e-06,
+      "loss": 0.6449,
+      "step": 5861
+    },
+    {
+      "epoch": 0.40702680183307877,
+      "grad_norm": 2.078575843162596,
+      "learning_rate": 6.713681165059271e-06,
+      "loss": 0.1761,
+      "step": 5862
+    },
+    {
+      "epoch": 0.40709623663380085,
+      "grad_norm": 4.0723398700108095,
+      "learning_rate": 6.712624742729128e-06,
+      "loss": 0.5921,
+      "step": 5863
+    },
+    {
+      "epoch": 0.407165671434523,
+      "grad_norm": 2.9012724208563734,
+      "learning_rate": 6.711568233776362e-06,
+      "loss": 0.3326,
+      "step": 5864
+    },
+    {
+      "epoch": 0.4072351062352451,
+      "grad_norm": 5.925823425340588,
+      "learning_rate": 6.710511638254412e-06,
+      "loss": 0.7255,
+      "step": 5865
+    },
+    {
+      "epoch": 0.40730454103596725,
+      "grad_norm": 3.9631755882419233,
+      "learning_rate": 6.709454956216719e-06,
+      "loss": 0.5727,
+      "step": 5866
+    },
+    {
+      "epoch": 0.40737397583668933,
+      "grad_norm": 3.900004258123668,
+      "learning_rate": 6.708398187716728e-06,
+      "loss": 0.4965,
+      "step": 5867
+    },
+    {
+      "epoch": 0.40744341063741146,
+      "grad_norm": 4.313163358290437,
+      "learning_rate": 6.7073413328078904e-06,
+      "loss": 0.4028,
+      "step": 5868
+    },
+    {
+      "epoch": 0.4075128454381336,
+      "grad_norm": 3.1658095815284657,
+      "learning_rate": 6.706284391543659e-06,
+      "loss": 0.2757,
+      "step": 5869
+    },
+    {
+      "epoch": 0.40758228023885573,
+      "grad_norm": 4.438429410335899,
+      "learning_rate": 6.705227363977495e-06,
+      "loss": 0.54,
+      "step": 5870
+    },
+    {
+      "epoch": 0.4076517150395778,
+      "grad_norm": 4.573335138664659,
+      "learning_rate": 6.70417025016286e-06,
+      "loss": 0.5805,
+      "step": 5871
+    },
+    {
+      "epoch": 0.40772114984029995,
+      "grad_norm": 6.075891757690346,
+      "learning_rate": 6.703113050153222e-06,
+      "loss": 0.4729,
+      "step": 5872
+    },
+    {
+      "epoch": 0.4077905846410221,
+      "grad_norm": 7.154696697144242,
+      "learning_rate": 6.702055764002055e-06,
+      "loss": 0.5836,
+      "step": 5873
+    },
+    {
+      "epoch": 0.4078600194417442,
+      "grad_norm": 3.62723589599387,
+      "learning_rate": 6.700998391762832e-06,
+      "loss": 0.3056,
+      "step": 5874
+    },
+    {
+      "epoch": 0.40792945424246635,
+      "grad_norm": 4.995113186134501,
+      "learning_rate": 6.6999409334890355e-06,
+      "loss": 0.7423,
+      "step": 5875
+    },
+    {
+      "epoch": 0.40799888904318843,
+      "grad_norm": 3.192289291959845,
+      "learning_rate": 6.698883389234149e-06,
+      "loss": 0.3186,
+      "step": 5876
+    },
+    {
+      "epoch": 0.40806832384391056,
+      "grad_norm": 4.027991287662262,
+      "learning_rate": 6.697825759051665e-06,
+      "loss": 0.5152,
+      "step": 5877
+    },
+    {
+      "epoch": 0.4081377586446327,
+      "grad_norm": 3.8166397144513824,
+      "learning_rate": 6.696768042995073e-06,
+      "loss": 0.435,
+      "step": 5878
+    },
+    {
+      "epoch": 0.40820719344535483,
+      "grad_norm": 3.8838489575156774,
+      "learning_rate": 6.695710241117874e-06,
+      "loss": 0.4021,
+      "step": 5879
+    },
+    {
+      "epoch": 0.4082766282460769,
+      "grad_norm": 3.921780602818124,
+      "learning_rate": 6.694652353473572e-06,
+      "loss": 0.4075,
+      "step": 5880
+    },
+    {
+      "epoch": 0.40834606304679905,
+      "grad_norm": 5.618549817364158,
+      "learning_rate": 6.69359438011567e-06,
+      "loss": 0.6882,
+      "step": 5881
+    },
+    {
+      "epoch": 0.4084154978475212,
+      "grad_norm": 4.279394435667104,
+      "learning_rate": 6.6925363210976805e-06,
+      "loss": 0.715,
+      "step": 5882
+    },
+    {
+      "epoch": 0.4084849326482433,
+      "grad_norm": 3.352895743852748,
+      "learning_rate": 6.691478176473119e-06,
+      "loss": 0.2688,
+      "step": 5883
+    },
+    {
+      "epoch": 0.4085543674489654,
+      "grad_norm": 4.787252541492366,
+      "learning_rate": 6.6904199462955066e-06,
+      "loss": 0.4539,
+      "step": 5884
+    },
+    {
+      "epoch": 0.40862380224968753,
+      "grad_norm": 4.2170197641690415,
+      "learning_rate": 6.689361630618365e-06,
+      "loss": 0.5739,
+      "step": 5885
+    },
+    {
+      "epoch": 0.40869323705040966,
+      "grad_norm": 3.0514195585222166,
+      "learning_rate": 6.6883032294952235e-06,
+      "loss": 0.3535,
+      "step": 5886
+    },
+    {
+      "epoch": 0.4087626718511318,
+      "grad_norm": 3.174092047219007,
+      "learning_rate": 6.6872447429796165e-06,
+      "loss": 0.3313,
+      "step": 5887
+    },
+    {
+      "epoch": 0.40883210665185393,
+      "grad_norm": 3.9586476320541824,
+      "learning_rate": 6.686186171125078e-06,
+      "loss": 0.2986,
+      "step": 5888
+    },
+    {
+      "epoch": 0.408901541452576,
+      "grad_norm": 4.707054949895881,
+      "learning_rate": 6.685127513985153e-06,
+      "loss": 0.8038,
+      "step": 5889
+    },
+    {
+      "epoch": 0.40897097625329815,
+      "grad_norm": 5.222779371017647,
+      "learning_rate": 6.6840687716133836e-06,
+      "loss": 0.4819,
+      "step": 5890
+    },
+    {
+      "epoch": 0.4090404110540203,
+      "grad_norm": 5.914431568774466,
+      "learning_rate": 6.683009944063322e-06,
+      "loss": 0.975,
+      "step": 5891
+    },
+    {
+      "epoch": 0.4091098458547424,
+      "grad_norm": 3.6476355790144117,
+      "learning_rate": 6.681951031388522e-06,
+      "loss": 0.4605,
+      "step": 5892
+    },
+    {
+      "epoch": 0.4091792806554645,
+      "grad_norm": 4.1983743182352296,
+      "learning_rate": 6.680892033642542e-06,
+      "loss": 0.4921,
+      "step": 5893
+    },
+    {
+      "epoch": 0.40924871545618663,
+      "grad_norm": 3.763099160781569,
+      "learning_rate": 6.6798329508789435e-06,
+      "loss": 0.5579,
+      "step": 5894
+    },
+    {
+      "epoch": 0.40931815025690876,
+      "grad_norm": 3.8992747205643394,
+      "learning_rate": 6.678773783151298e-06,
+      "loss": 0.266,
+      "step": 5895
+    },
+    {
+      "epoch": 0.4093875850576309,
+      "grad_norm": 4.806555818325423,
+      "learning_rate": 6.677714530513172e-06,
+      "loss": 0.739,
+      "step": 5896
+    },
+    {
+      "epoch": 0.40945701985835303,
+      "grad_norm": 4.052946994876943,
+      "learning_rate": 6.676655193018145e-06,
+      "loss": 0.5116,
+      "step": 5897
+    },
+    {
+      "epoch": 0.4095264546590751,
+      "grad_norm": 4.395198902434428,
+      "learning_rate": 6.675595770719794e-06,
+      "loss": 0.5286,
+      "step": 5898
+    },
+    {
+      "epoch": 0.40959588945979725,
+      "grad_norm": 4.162047009414529,
+      "learning_rate": 6.674536263671705e-06,
+      "loss": 0.7285,
+      "step": 5899
+    },
+    {
+      "epoch": 0.4096653242605194,
+      "grad_norm": 3.237450897307577,
+      "learning_rate": 6.673476671927469e-06,
+      "loss": 0.3734,
+      "step": 5900
+    },
+    {
+      "epoch": 0.4097347590612415,
+      "grad_norm": 3.155129636765505,
+      "learning_rate": 6.672416995540674e-06,
+      "loss": 0.3517,
+      "step": 5901
+    },
+    {
+      "epoch": 0.4098041938619636,
+      "grad_norm": 4.660231469046866,
+      "learning_rate": 6.671357234564922e-06,
+      "loss": 0.6399,
+      "step": 5902
+    },
+    {
+      "epoch": 0.40987362866268573,
+      "grad_norm": 3.076170155508563,
+      "learning_rate": 6.67029738905381e-06,
+      "loss": 0.4378,
+      "step": 5903
+    },
+    {
+      "epoch": 0.40994306346340786,
+      "grad_norm": 4.208727675373513,
+      "learning_rate": 6.669237459060949e-06,
+      "loss": 0.5492,
+      "step": 5904
+    },
+    {
+      "epoch": 0.41001249826413,
+      "grad_norm": 3.803122969220418,
+      "learning_rate": 6.668177444639943e-06,
+      "loss": 0.5074,
+      "step": 5905
+    },
+    {
+      "epoch": 0.4100819330648521,
+      "grad_norm": 4.055629583275958,
+      "learning_rate": 6.667117345844412e-06,
+      "loss": 0.5432,
+      "step": 5906
+    },
+    {
+      "epoch": 0.4101513678655742,
+      "grad_norm": 2.9202481356540666,
+      "learning_rate": 6.666057162727971e-06,
+      "loss": 0.3016,
+      "step": 5907
+    },
+    {
+      "epoch": 0.41022080266629635,
+      "grad_norm": 3.759433899415458,
+      "learning_rate": 6.664996895344244e-06,
+      "loss": 0.4002,
+      "step": 5908
+    },
+    {
+      "epoch": 0.4102902374670185,
+      "grad_norm": 4.450768686666573,
+      "learning_rate": 6.663936543746859e-06,
+      "loss": 0.4668,
+      "step": 5909
+    },
+    {
+      "epoch": 0.4103596722677406,
+      "grad_norm": 3.231435287481911,
+      "learning_rate": 6.6628761079894455e-06,
+      "loss": 0.4188,
+      "step": 5910
+    },
+    {
+      "epoch": 0.4104291070684627,
+      "grad_norm": 3.679610150834794,
+      "learning_rate": 6.66181558812564e-06,
+      "loss": 0.4736,
+      "step": 5911
+    },
+    {
+      "epoch": 0.41049854186918483,
+      "grad_norm": 1.9954281910558032,
+      "learning_rate": 6.660754984209083e-06,
+      "loss": 0.1997,
+      "step": 5912
+    },
+    {
+      "epoch": 0.41056797666990696,
+      "grad_norm": 5.289002001172379,
+      "learning_rate": 6.659694296293419e-06,
+      "loss": 0.7583,
+      "step": 5913
+    },
+    {
+      "epoch": 0.4106374114706291,
+      "grad_norm": 4.273182986741465,
+      "learning_rate": 6.658633524432293e-06,
+      "loss": 0.5964,
+      "step": 5914
+    },
+    {
+      "epoch": 0.4107068462713512,
+      "grad_norm": 3.8681286637149257,
+      "learning_rate": 6.657572668679362e-06,
+      "loss": 0.6502,
+      "step": 5915
+    },
+    {
+      "epoch": 0.4107762810720733,
+      "grad_norm": 3.4056447014614992,
+      "learning_rate": 6.656511729088283e-06,
+      "loss": 0.3841,
+      "step": 5916
+    },
+    {
+      "epoch": 0.41084571587279545,
+      "grad_norm": 3.2910703824932046,
+      "learning_rate": 6.655450705712712e-06,
+      "loss": 0.2944,
+      "step": 5917
+    },
+    {
+      "epoch": 0.4109151506735176,
+      "grad_norm": 5.194844303241585,
+      "learning_rate": 6.654389598606319e-06,
+      "loss": 0.5354,
+      "step": 5918
+    },
+    {
+      "epoch": 0.4109845854742397,
+      "grad_norm": 4.982133942194317,
+      "learning_rate": 6.653328407822772e-06,
+      "loss": 0.5852,
+      "step": 5919
+    },
+    {
+      "epoch": 0.4110540202749618,
+      "grad_norm": 4.331500222866751,
+      "learning_rate": 6.652267133415745e-06,
+      "loss": 0.4463,
+      "step": 5920
+    },
+    {
+      "epoch": 0.41112345507568393,
+      "grad_norm": 3.5967563353571497,
+      "learning_rate": 6.651205775438915e-06,
+      "loss": 0.5262,
+      "step": 5921
+    },
+    {
+      "epoch": 0.41119288987640606,
+      "grad_norm": 3.518169563269043,
+      "learning_rate": 6.650144333945967e-06,
+      "loss": 0.4601,
+      "step": 5922
+    },
+    {
+      "epoch": 0.4112623246771282,
+      "grad_norm": 4.184806884149068,
+      "learning_rate": 6.6490828089905854e-06,
+      "loss": 0.3972,
+      "step": 5923
+    },
+    {
+      "epoch": 0.4113317594778503,
+      "grad_norm": 4.829607267750261,
+      "learning_rate": 6.648021200626461e-06,
+      "loss": 0.5794,
+      "step": 5924
+    },
+    {
+      "epoch": 0.4114011942785724,
+      "grad_norm": 3.8021323096553172,
+      "learning_rate": 6.646959508907289e-06,
+      "loss": 0.381,
+      "step": 5925
+    },
+    {
+      "epoch": 0.41147062907929455,
+      "grad_norm": 3.2063745931897927,
+      "learning_rate": 6.645897733886769e-06,
+      "loss": 0.3945,
+      "step": 5926
+    },
+    {
+      "epoch": 0.4115400638800167,
+      "grad_norm": 4.092857132261174,
+      "learning_rate": 6.644835875618605e-06,
+      "loss": 0.3987,
+      "step": 5927
+    },
+    {
+      "epoch": 0.41160949868073876,
+      "grad_norm": 3.7730263416638468,
+      "learning_rate": 6.643773934156502e-06,
+      "loss": 0.3898,
+      "step": 5928
+    },
+    {
+      "epoch": 0.4116789334814609,
+      "grad_norm": 3.9907095817209117,
+      "learning_rate": 6.6427119095541745e-06,
+      "loss": 0.5007,
+      "step": 5929
+    },
+    {
+      "epoch": 0.41174836828218303,
+      "grad_norm": 4.058541369300089,
+      "learning_rate": 6.641649801865336e-06,
+      "loss": 0.5124,
+      "step": 5930
+    },
+    {
+      "epoch": 0.41181780308290517,
+      "grad_norm": 3.6012011650320703,
+      "learning_rate": 6.64058761114371e-06,
+      "loss": 0.5843,
+      "step": 5931
+    },
+    {
+      "epoch": 0.4118872378836273,
+      "grad_norm": 3.5267397701417655,
+      "learning_rate": 6.63952533744302e-06,
+      "loss": 0.3361,
+      "step": 5932
+    },
+    {
+      "epoch": 0.4119566726843494,
+      "grad_norm": 3.2973009106072024,
+      "learning_rate": 6.638462980816991e-06,
+      "loss": 0.451,
+      "step": 5933
+    },
+    {
+      "epoch": 0.4120261074850715,
+      "grad_norm": 3.6014745606413405,
+      "learning_rate": 6.63740054131936e-06,
+      "loss": 0.4682,
+      "step": 5934
+    },
+    {
+      "epoch": 0.41209554228579365,
+      "grad_norm": 5.1985078424696365,
+      "learning_rate": 6.636338019003862e-06,
+      "loss": 0.6571,
+      "step": 5935
+    },
+    {
+      "epoch": 0.4121649770865158,
+      "grad_norm": 3.5711542056633547,
+      "learning_rate": 6.635275413924238e-06,
+      "loss": 0.2541,
+      "step": 5936
+    },
+    {
+      "epoch": 0.41223441188723786,
+      "grad_norm": 3.7969106283580007,
+      "learning_rate": 6.634212726134235e-06,
+      "loss": 0.5522,
+      "step": 5937
+    },
+    {
+      "epoch": 0.41230384668796,
+      "grad_norm": 3.7764966457648703,
+      "learning_rate": 6.633149955687602e-06,
+      "loss": 0.3602,
+      "step": 5938
+    },
+    {
+      "epoch": 0.41237328148868213,
+      "grad_norm": 3.5961277066134283,
+      "learning_rate": 6.632087102638091e-06,
+      "loss": 0.4246,
+      "step": 5939
+    },
+    {
+      "epoch": 0.41244271628940427,
+      "grad_norm": 4.283934639118943,
+      "learning_rate": 6.631024167039461e-06,
+      "loss": 0.4702,
+      "step": 5940
+    },
+    {
+      "epoch": 0.41251215109012634,
+      "grad_norm": 4.507665720805658,
+      "learning_rate": 6.629961148945475e-06,
+      "loss": 0.6833,
+      "step": 5941
+    },
+    {
+      "epoch": 0.4125815858908485,
+      "grad_norm": 3.336149228634475,
+      "learning_rate": 6.6288980484098994e-06,
+      "loss": 0.2978,
+      "step": 5942
+    },
+    {
+      "epoch": 0.4126510206915706,
+      "grad_norm": 3.6698902634904136,
+      "learning_rate": 6.627834865486502e-06,
+      "loss": 0.4876,
+      "step": 5943
+    },
+    {
+      "epoch": 0.41272045549229275,
+      "grad_norm": 4.214610520727473,
+      "learning_rate": 6.626771600229059e-06,
+      "loss": 0.7225,
+      "step": 5944
+    },
+    {
+      "epoch": 0.4127898902930149,
+      "grad_norm": 5.163373777729271,
+      "learning_rate": 6.6257082526913515e-06,
+      "loss": 0.6017,
+      "step": 5945
+    },
+    {
+      "epoch": 0.41285932509373696,
+      "grad_norm": 2.590636025921402,
+      "learning_rate": 6.624644822927159e-06,
+      "loss": 0.1924,
+      "step": 5946
+    },
+    {
+      "epoch": 0.4129287598944591,
+      "grad_norm": 4.752575031449287,
+      "learning_rate": 6.623581310990269e-06,
+      "loss": 0.7692,
+      "step": 5947
+    },
+    {
+      "epoch": 0.41299819469518123,
+      "grad_norm": 4.148498231048011,
+      "learning_rate": 6.6225177169344756e-06,
+      "loss": 0.4368,
+      "step": 5948
+    },
+    {
+      "epoch": 0.41306762949590337,
+      "grad_norm": 3.3461980276016012,
+      "learning_rate": 6.621454040813569e-06,
+      "loss": 0.319,
+      "step": 5949
+    },
+    {
+      "epoch": 0.41313706429662544,
+      "grad_norm": 3.4121787640450085,
+      "learning_rate": 6.620390282681355e-06,
+      "loss": 0.2289,
+      "step": 5950
+    },
+    {
+      "epoch": 0.4132064990973476,
+      "grad_norm": 3.076598224519381,
+      "learning_rate": 6.6193264425916325e-06,
+      "loss": 0.4939,
+      "step": 5951
+    },
+    {
+      "epoch": 0.4132759338980697,
+      "grad_norm": 3.24952883063537,
+      "learning_rate": 6.618262520598211e-06,
+      "loss": 0.4662,
+      "step": 5952
+    },
+    {
+      "epoch": 0.41334536869879185,
+      "grad_norm": 3.273221906035571,
+      "learning_rate": 6.617198516754901e-06,
+      "loss": 0.1984,
+      "step": 5953
+    },
+    {
+      "epoch": 0.413414803499514,
+      "grad_norm": 3.9943603386298974,
+      "learning_rate": 6.616134431115523e-06,
+      "loss": 0.4798,
+      "step": 5954
+    },
+    {
+      "epoch": 0.41348423830023606,
+      "grad_norm": 3.432414428143647,
+      "learning_rate": 6.615070263733891e-06,
+      "loss": 0.2774,
+      "step": 5955
+    },
+    {
+      "epoch": 0.4135536731009582,
+      "grad_norm": 3.8023889458774183,
+      "learning_rate": 6.6140060146638355e-06,
+      "loss": 0.4927,
+      "step": 5956
+    },
+    {
+      "epoch": 0.41362310790168033,
+      "grad_norm": 4.107753646599888,
+      "learning_rate": 6.612941683959181e-06,
+      "loss": 0.2787,
+      "step": 5957
+    },
+    {
+      "epoch": 0.41369254270240247,
+      "grad_norm": 2.862977909829752,
+      "learning_rate": 6.611877271673761e-06,
+      "loss": 0.3095,
+      "step": 5958
+    },
+    {
+      "epoch": 0.41376197750312455,
+      "grad_norm": 4.08107664800648,
+      "learning_rate": 6.6108127778614126e-06,
+      "loss": 0.3139,
+      "step": 5959
+    },
+    {
+      "epoch": 0.4138314123038467,
+      "grad_norm": 3.283768566514904,
+      "learning_rate": 6.609748202575976e-06,
+      "loss": 0.2975,
+      "step": 5960
+    },
+    {
+      "epoch": 0.4139008471045688,
+      "grad_norm": 4.548084471201058,
+      "learning_rate": 6.6086835458712994e-06,
+      "loss": 0.5423,
+      "step": 5961
+    },
+    {
+      "epoch": 0.41397028190529095,
+      "grad_norm": 4.756708378913505,
+      "learning_rate": 6.607618807801227e-06,
+      "loss": 0.483,
+      "step": 5962
+    },
+    {
+      "epoch": 0.41403971670601303,
+      "grad_norm": 4.271132643086925,
+      "learning_rate": 6.606553988419615e-06,
+      "loss": 0.6305,
+      "step": 5963
+    },
+    {
+      "epoch": 0.41410915150673516,
+      "grad_norm": 3.7117500649357322,
+      "learning_rate": 6.60548908778032e-06,
+      "loss": 0.3805,
+      "step": 5964
+    },
+    {
+      "epoch": 0.4141785863074573,
+      "grad_norm": 5.839294345960325,
+      "learning_rate": 6.604424105937205e-06,
+      "loss": 0.647,
+      "step": 5965
+    },
+    {
+      "epoch": 0.41424802110817943,
+      "grad_norm": 5.026592636058434,
+      "learning_rate": 6.603359042944133e-06,
+      "loss": 0.7131,
+      "step": 5966
+    },
+    {
+      "epoch": 0.41431745590890157,
+      "grad_norm": 4.099199914534691,
+      "learning_rate": 6.6022938988549755e-06,
+      "loss": 0.5236,
+      "step": 5967
+    },
+    {
+      "epoch": 0.41438689070962365,
+      "grad_norm": 3.2668865112559056,
+      "learning_rate": 6.6012286737236056e-06,
+      "loss": 0.2629,
+      "step": 5968
+    },
+    {
+      "epoch": 0.4144563255103458,
+      "grad_norm": 5.500072938008738,
+      "learning_rate": 6.6001633676039e-06,
+      "loss": 0.724,
+      "step": 5969
+    },
+    {
+      "epoch": 0.4145257603110679,
+      "grad_norm": 3.564160387367135,
+      "learning_rate": 6.599097980549744e-06,
+      "loss": 0.4112,
+      "step": 5970
+    },
+    {
+      "epoch": 0.41459519511179005,
+      "grad_norm": 5.935853775523354,
+      "learning_rate": 6.59803251261502e-06,
+      "loss": 0.4225,
+      "step": 5971
+    },
+    {
+      "epoch": 0.41466462991251213,
+      "grad_norm": 3.4353260115452873,
+      "learning_rate": 6.596966963853621e-06,
+      "loss": 0.3019,
+      "step": 5972
+    },
+    {
+      "epoch": 0.41473406471323426,
+      "grad_norm": 4.752866054882168,
+      "learning_rate": 6.59590133431944e-06,
+      "loss": 0.7035,
+      "step": 5973
+    },
+    {
+      "epoch": 0.4148034995139564,
+      "grad_norm": 3.8998604835877315,
+      "learning_rate": 6.594835624066375e-06,
+      "loss": 0.6103,
+      "step": 5974
+    },
+    {
+      "epoch": 0.41487293431467853,
+      "grad_norm": 4.039922923711053,
+      "learning_rate": 6.5937698331483295e-06,
+      "loss": 0.4017,
+      "step": 5975
+    },
+    {
+      "epoch": 0.4149423691154006,
+      "grad_norm": 3.791477448342948,
+      "learning_rate": 6.592703961619207e-06,
+      "loss": 0.3924,
+      "step": 5976
+    },
+    {
+      "epoch": 0.41501180391612275,
+      "grad_norm": 5.979169761290956,
+      "learning_rate": 6.591638009532923e-06,
+      "loss": 0.5755,
+      "step": 5977
+    },
+    {
+      "epoch": 0.4150812387168449,
+      "grad_norm": 3.999373012322334,
+      "learning_rate": 6.590571976943387e-06,
+      "loss": 0.5232,
+      "step": 5978
+    },
+    {
+      "epoch": 0.415150673517567,
+      "grad_norm": 5.013172341464592,
+      "learning_rate": 6.589505863904523e-06,
+      "loss": 0.7576,
+      "step": 5979
+    },
+    {
+      "epoch": 0.41522010831828915,
+      "grad_norm": 2.707524023017276,
+      "learning_rate": 6.58843967047025e-06,
+      "loss": 0.2317,
+      "step": 5980
+    },
+    {
+      "epoch": 0.41528954311901123,
+      "grad_norm": 4.43508441736149,
+      "learning_rate": 6.587373396694496e-06,
+      "loss": 0.6286,
+      "step": 5981
+    },
+    {
+      "epoch": 0.41535897791973336,
+      "grad_norm": 4.7917072832940715,
+      "learning_rate": 6.586307042631192e-06,
+      "loss": 0.4474,
+      "step": 5982
+    },
+    {
+      "epoch": 0.4154284127204555,
+      "grad_norm": 7.629401313170102,
+      "learning_rate": 6.585240608334272e-06,
+      "loss": 0.32,
+      "step": 5983
+    },
+    {
+      "epoch": 0.41549784752117763,
+      "grad_norm": 4.045749950822491,
+      "learning_rate": 6.584174093857676e-06,
+      "loss": 0.6604,
+      "step": 5984
+    },
+    {
+      "epoch": 0.4155672823218997,
+      "grad_norm": 4.100293467743907,
+      "learning_rate": 6.5831074992553465e-06,
+      "loss": 0.4401,
+      "step": 5985
+    },
+    {
+      "epoch": 0.41563671712262185,
+      "grad_norm": 3.5283461686205664,
+      "learning_rate": 6.582040824581233e-06,
+      "loss": 0.4108,
+      "step": 5986
+    },
+    {
+      "epoch": 0.415706151923344,
+      "grad_norm": 4.061756782528265,
+      "learning_rate": 6.580974069889283e-06,
+      "loss": 0.3636,
+      "step": 5987
+    },
+    {
+      "epoch": 0.4157755867240661,
+      "grad_norm": 3.667780322548429,
+      "learning_rate": 6.579907235233454e-06,
+      "loss": 0.615,
+      "step": 5988
+    },
+    {
+      "epoch": 0.41584502152478825,
+      "grad_norm": 4.50694731920286,
+      "learning_rate": 6.578840320667705e-06,
+      "loss": 0.8167,
+      "step": 5989
+    },
+    {
+      "epoch": 0.41591445632551033,
+      "grad_norm": 4.018970295523801,
+      "learning_rate": 6.5777733262459995e-06,
+      "loss": 0.4205,
+      "step": 5990
+    },
+    {
+      "epoch": 0.41598389112623246,
+      "grad_norm": 4.256155107757088,
+      "learning_rate": 6.576706252022304e-06,
+      "loss": 0.4638,
+      "step": 5991
+    },
+    {
+      "epoch": 0.4160533259269546,
+      "grad_norm": 3.9799483208314923,
+      "learning_rate": 6.57563909805059e-06,
+      "loss": 0.3579,
+      "step": 5992
+    },
+    {
+      "epoch": 0.41612276072767673,
+      "grad_norm": 5.250308741263798,
+      "learning_rate": 6.574571864384835e-06,
+      "loss": 0.4732,
+      "step": 5993
+    },
+    {
+      "epoch": 0.4161921955283988,
+      "grad_norm": 3.5431734289452934,
+      "learning_rate": 6.573504551079015e-06,
+      "loss": 0.4609,
+      "step": 5994
+    },
+    {
+      "epoch": 0.41626163032912095,
+      "grad_norm": 4.086440270062286,
+      "learning_rate": 6.5724371581871175e-06,
+      "loss": 0.6302,
+      "step": 5995
+    },
+    {
+      "epoch": 0.4163310651298431,
+      "grad_norm": 4.630688252009449,
+      "learning_rate": 6.571369685763126e-06,
+      "loss": 0.6693,
+      "step": 5996
+    },
+    {
+      "epoch": 0.4164004999305652,
+      "grad_norm": 4.355906978789595,
+      "learning_rate": 6.570302133861034e-06,
+      "loss": 0.5666,
+      "step": 5997
+    },
+    {
+      "epoch": 0.4164699347312873,
+      "grad_norm": 4.106175428632236,
+      "learning_rate": 6.569234502534838e-06,
+      "loss": 0.5607,
+      "step": 5998
+    },
+    {
+      "epoch": 0.41653936953200943,
+      "grad_norm": 4.417857746977174,
+      "learning_rate": 6.568166791838536e-06,
+      "loss": 0.7039,
+      "step": 5999
+    },
+    {
+      "epoch": 0.41660880433273156,
+      "grad_norm": 4.135282633525598,
+      "learning_rate": 6.567099001826133e-06,
+      "loss": 0.2914,
+      "step": 6000
+    },
+    {
+      "epoch": 0.4166782391334537,
+      "grad_norm": 3.678670965085961,
+      "learning_rate": 6.566031132551634e-06,
+      "loss": 0.3352,
+      "step": 6001
+    },
+    {
+      "epoch": 0.41674767393417583,
+      "grad_norm": 4.819477573572015,
+      "learning_rate": 6.564963184069053e-06,
+      "loss": 0.6571,
+      "step": 6002
+    },
+    {
+      "epoch": 0.4168171087348979,
+      "grad_norm": 4.308681583605821,
+      "learning_rate": 6.563895156432405e-06,
+      "loss": 0.5728,
+      "step": 6003
+    },
+    {
+      "epoch": 0.41688654353562005,
+      "grad_norm": 2.8802495366572396,
+      "learning_rate": 6.56282704969571e-06,
+      "loss": 0.2767,
+      "step": 6004
+    },
+    {
+      "epoch": 0.4169559783363422,
+      "grad_norm": 4.055709462249376,
+      "learning_rate": 6.561758863912991e-06,
+      "loss": 0.4474,
+      "step": 6005
+    },
+    {
+      "epoch": 0.4170254131370643,
+      "grad_norm": 4.740713992840043,
+      "learning_rate": 6.560690599138277e-06,
+      "loss": 0.9646,
+      "step": 6006
+    },
+    {
+      "epoch": 0.4170948479377864,
+      "grad_norm": 3.869120748204385,
+      "learning_rate": 6.559622255425598e-06,
+      "loss": 0.43,
+      "step": 6007
+    },
+    {
+      "epoch": 0.41716428273850853,
+      "grad_norm": 4.631126825076075,
+      "learning_rate": 6.55855383282899e-06,
+      "loss": 0.7757,
+      "step": 6008
+    },
+    {
+      "epoch": 0.41723371753923066,
+      "grad_norm": 3.5143602200022994,
+      "learning_rate": 6.557485331402494e-06,
+      "loss": 0.4444,
+      "step": 6009
+    },
+    {
+      "epoch": 0.4173031523399528,
+      "grad_norm": 3.2979481071024743,
+      "learning_rate": 6.5564167512001496e-06,
+      "loss": 0.2892,
+      "step": 6010
+    },
+    {
+      "epoch": 0.41737258714067493,
+      "grad_norm": 7.261492177131983,
+      "learning_rate": 6.555348092276008e-06,
+      "loss": 0.5031,
+      "step": 6011
+    },
+    {
+      "epoch": 0.417442021941397,
+      "grad_norm": 3.974404138460476,
+      "learning_rate": 6.554279354684121e-06,
+      "loss": 0.4757,
+      "step": 6012
+    },
+    {
+      "epoch": 0.41751145674211915,
+      "grad_norm": 3.983571118866898,
+      "learning_rate": 6.553210538478542e-06,
+      "loss": 0.4088,
+      "step": 6013
+    },
+    {
+      "epoch": 0.4175808915428413,
+      "grad_norm": 3.5801342877607665,
+      "learning_rate": 6.552141643713334e-06,
+      "loss": 0.3163,
+      "step": 6014
+    },
+    {
+      "epoch": 0.4176503263435634,
+      "grad_norm": 3.4938435226373725,
+      "learning_rate": 6.551072670442554e-06,
+      "loss": 0.5077,
+      "step": 6015
+    },
+    {
+      "epoch": 0.4177197611442855,
+      "grad_norm": 3.479033959102143,
+      "learning_rate": 6.550003618720277e-06,
+      "loss": 0.4043,
+      "step": 6016
+    },
+    {
+      "epoch": 0.41778919594500763,
+      "grad_norm": 4.204367460846488,
+      "learning_rate": 6.548934488600569e-06,
+      "loss": 0.5408,
+      "step": 6017
+    },
+    {
+      "epoch": 0.41785863074572976,
+      "grad_norm": 4.360862455777884,
+      "learning_rate": 6.547865280137507e-06,
+      "loss": 0.5516,
+      "step": 6018
+    },
+    {
+      "epoch": 0.4179280655464519,
+      "grad_norm": 2.8065093647682264,
+      "learning_rate": 6.546795993385172e-06,
+      "loss": 0.1874,
+      "step": 6019
+    },
+    {
+      "epoch": 0.417997500347174,
+      "grad_norm": 3.699325362917268,
+      "learning_rate": 6.545726628397646e-06,
+      "loss": 0.531,
+      "step": 6020
+    },
+    {
+      "epoch": 0.4180669351478961,
+      "grad_norm": 3.9946305771491617,
+      "learning_rate": 6.544657185229014e-06,
+      "loss": 0.6079,
+      "step": 6021
+    },
+    {
+      "epoch": 0.41813636994861825,
+      "grad_norm": 3.9575612081875935,
+      "learning_rate": 6.543587663933373e-06,
+      "loss": 0.5549,
+      "step": 6022
+    },
+    {
+      "epoch": 0.4182058047493404,
+      "grad_norm": 2.7961794737288055,
+      "learning_rate": 6.5425180645648125e-06,
+      "loss": 0.2726,
+      "step": 6023
+    },
+    {
+      "epoch": 0.4182752395500625,
+      "grad_norm": 4.570486094680746,
+      "learning_rate": 6.541448387177434e-06,
+      "loss": 0.678,
+      "step": 6024
+    },
+    {
+      "epoch": 0.4183446743507846,
+      "grad_norm": 3.899465307463808,
+      "learning_rate": 6.540378631825339e-06,
+      "loss": 0.513,
+      "step": 6025
+    },
+    {
+      "epoch": 0.41841410915150673,
+      "grad_norm": 2.9027411671370875,
+      "learning_rate": 6.539308798562638e-06,
+      "loss": 0.2125,
+      "step": 6026
+    },
+    {
+      "epoch": 0.41848354395222886,
+      "grad_norm": 3.618225188840961,
+      "learning_rate": 6.538238887443439e-06,
+      "loss": 0.3345,
+      "step": 6027
+    },
+    {
+      "epoch": 0.418552978752951,
+      "grad_norm": 3.600633353446839,
+      "learning_rate": 6.537168898521857e-06,
+      "loss": 0.3684,
+      "step": 6028
+    },
+    {
+      "epoch": 0.4186224135536731,
+      "grad_norm": 3.7134990189042694,
+      "learning_rate": 6.536098831852013e-06,
+      "loss": 0.5795,
+      "step": 6029
+    },
+    {
+      "epoch": 0.4186918483543952,
+      "grad_norm": 4.089622555002528,
+      "learning_rate": 6.535028687488028e-06,
+      "loss": 0.4647,
+      "step": 6030
+    },
+    {
+      "epoch": 0.41876128315511735,
+      "grad_norm": 3.5980417513099034,
+      "learning_rate": 6.5339584654840285e-06,
+      "loss": 0.422,
+      "step": 6031
+    },
+    {
+      "epoch": 0.4188307179558395,
+      "grad_norm": 4.5279715614629,
+      "learning_rate": 6.5328881658941465e-06,
+      "loss": 0.5197,
+      "step": 6032
+    },
+    {
+      "epoch": 0.41890015275656156,
+      "grad_norm": 3.679356878002001,
+      "learning_rate": 6.531817788772514e-06,
+      "loss": 0.342,
+      "step": 6033
+    },
+    {
+      "epoch": 0.4189695875572837,
+      "grad_norm": 4.512231380692237,
+      "learning_rate": 6.5307473341732716e-06,
+      "loss": 0.5692,
+      "step": 6034
+    },
+    {
+      "epoch": 0.41903902235800583,
+      "grad_norm": 4.115720243011601,
+      "learning_rate": 6.529676802150562e-06,
+      "loss": 0.6132,
+      "step": 6035
+    },
+    {
+      "epoch": 0.41910845715872797,
+      "grad_norm": 3.2013356523126517,
+      "learning_rate": 6.52860619275853e-06,
+      "loss": 0.3609,
+      "step": 6036
+    },
+    {
+      "epoch": 0.4191778919594501,
+      "grad_norm": 5.542840427215459,
+      "learning_rate": 6.5275355060513256e-06,
+      "loss": 0.3734,
+      "step": 6037
+    },
+    {
+      "epoch": 0.4192473267601722,
+      "grad_norm": 4.268606719949882,
+      "learning_rate": 6.526464742083106e-06,
+      "loss": 0.7196,
+      "step": 6038
+    },
+    {
+      "epoch": 0.4193167615608943,
+      "grad_norm": 3.201347655982047,
+      "learning_rate": 6.525393900908027e-06,
+      "loss": 0.3387,
+      "step": 6039
+    },
+    {
+      "epoch": 0.41938619636161645,
+      "grad_norm": 2.5681047723346166,
+      "learning_rate": 6.524322982580249e-06,
+      "loss": 0.3211,
+      "step": 6040
+    },
+    {
+      "epoch": 0.4194556311623386,
+      "grad_norm": 5.038397113646505,
+      "learning_rate": 6.52325198715394e-06,
+      "loss": 0.4734,
+      "step": 6041
+    },
+    {
+      "epoch": 0.41952506596306066,
+      "grad_norm": 3.8208049712749186,
+      "learning_rate": 6.522180914683269e-06,
+      "loss": 0.2741,
+      "step": 6042
+    },
+    {
+      "epoch": 0.4195945007637828,
+      "grad_norm": 2.9651733411660572,
+      "learning_rate": 6.5211097652224095e-06,
+      "loss": 0.1394,
+      "step": 6043
+    },
+    {
+      "epoch": 0.41966393556450493,
+      "grad_norm": 2.7721057672388594,
+      "learning_rate": 6.52003853882554e-06,
+      "loss": 0.376,
+      "step": 6044
+    },
+    {
+      "epoch": 0.41973337036522707,
+      "grad_norm": 3.3019578493470996,
+      "learning_rate": 6.5189672355468415e-06,
+      "loss": 0.4997,
+      "step": 6045
+    },
+    {
+      "epoch": 0.4198028051659492,
+      "grad_norm": 4.714418992647794,
+      "learning_rate": 6.517895855440498e-06,
+      "loss": 0.5589,
+      "step": 6046
+    },
+    {
+      "epoch": 0.4198722399666713,
+      "grad_norm": 4.078158973383872,
+      "learning_rate": 6.516824398560701e-06,
+      "loss": 0.3377,
+      "step": 6047
+    },
+    {
+      "epoch": 0.4199416747673934,
+      "grad_norm": 6.592595413578484,
+      "learning_rate": 6.515752864961642e-06,
+      "loss": 0.8498,
+      "step": 6048
+    },
+    {
+      "epoch": 0.42001110956811555,
+      "grad_norm": 4.520407060254008,
+      "learning_rate": 6.514681254697517e-06,
+      "loss": 0.2933,
+      "step": 6049
+    },
+    {
+      "epoch": 0.4200805443688377,
+      "grad_norm": 4.5882533122976685,
+      "learning_rate": 6.513609567822529e-06,
+      "loss": 0.6601,
+      "step": 6050
+    },
+    {
+      "epoch": 0.42014997916955976,
+      "grad_norm": 1.8845564112630744,
+      "learning_rate": 6.512537804390883e-06,
+      "loss": 0.0911,
+      "step": 6051
+    },
+    {
+      "epoch": 0.4202194139702819,
+      "grad_norm": 4.111633155370928,
+      "learning_rate": 6.511465964456785e-06,
+      "loss": 0.6183,
+      "step": 6052
+    },
+    {
+      "epoch": 0.42028884877100403,
+      "grad_norm": 4.327947265337648,
+      "learning_rate": 6.510394048074449e-06,
+      "loss": 0.4192,
+      "step": 6053
+    },
+    {
+      "epoch": 0.42035828357172617,
+      "grad_norm": 3.3528841845329613,
+      "learning_rate": 6.509322055298092e-06,
+      "loss": 0.3139,
+      "step": 6054
+    },
+    {
+      "epoch": 0.42042771837244824,
+      "grad_norm": 3.202619124518965,
+      "learning_rate": 6.508249986181931e-06,
+      "loss": 0.3789,
+      "step": 6055
+    },
+    {
+      "epoch": 0.4204971531731704,
+      "grad_norm": 3.8857586249414524,
+      "learning_rate": 6.5071778407801955e-06,
+      "loss": 0.4773,
+      "step": 6056
+    },
+    {
+      "epoch": 0.4205665879738925,
+      "grad_norm": 4.261108104256129,
+      "learning_rate": 6.506105619147108e-06,
+      "loss": 0.5262,
+      "step": 6057
+    },
+    {
+      "epoch": 0.42063602277461465,
+      "grad_norm": 5.682191542890119,
+      "learning_rate": 6.5050333213369044e-06,
+      "loss": 0.6677,
+      "step": 6058
+    },
+    {
+      "epoch": 0.4207054575753368,
+      "grad_norm": 3.777291835713108,
+      "learning_rate": 6.503960947403818e-06,
+      "loss": 0.4188,
+      "step": 6059
+    },
+    {
+      "epoch": 0.42077489237605886,
+      "grad_norm": 3.8156496824070323,
+      "learning_rate": 6.502888497402087e-06,
+      "loss": 0.6032,
+      "step": 6060
+    },
+    {
+      "epoch": 0.420844327176781,
+      "grad_norm": 3.1060753127707854,
+      "learning_rate": 6.501815971385959e-06,
+      "loss": 0.2598,
+      "step": 6061
+    },
+    {
+      "epoch": 0.42091376197750313,
+      "grad_norm": 4.019809038108703,
+      "learning_rate": 6.500743369409676e-06,
+      "loss": 0.6035,
+      "step": 6062
+    },
+    {
+      "epoch": 0.42098319677822527,
+      "grad_norm": 4.271632667044726,
+      "learning_rate": 6.4996706915274935e-06,
+      "loss": 0.5402,
+      "step": 6063
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 4.2791206431556175,
+      "learning_rate": 6.498597937793665e-06,
+      "loss": 0.4962,
+      "step": 6064
+    },
+    {
+      "epoch": 0.4211220663796695,
+      "grad_norm": 3.863856827027266,
+      "learning_rate": 6.497525108262446e-06,
+      "loss": 0.426,
+      "step": 6065
+    },
+    {
+      "epoch": 0.4211915011803916,
+      "grad_norm": 3.3488976404895543,
+      "learning_rate": 6.496452202988104e-06,
+      "loss": 0.3959,
+      "step": 6066
+    },
+    {
+      "epoch": 0.42126093598111375,
+      "grad_norm": 3.501628951042371,
+      "learning_rate": 6.495379222024901e-06,
+      "loss": 0.5029,
+      "step": 6067
+    },
+    {
+      "epoch": 0.4213303707818359,
+      "grad_norm": 3.609672961257711,
+      "learning_rate": 6.494306165427111e-06,
+      "loss": 0.3534,
+      "step": 6068
+    },
+    {
+      "epoch": 0.42139980558255796,
+      "grad_norm": 4.653207558954737,
+      "learning_rate": 6.493233033249004e-06,
+      "loss": 0.3919,
+      "step": 6069
+    },
+    {
+      "epoch": 0.4214692403832801,
+      "grad_norm": 3.8126400644483818,
+      "learning_rate": 6.492159825544861e-06,
+      "loss": 0.4411,
+      "step": 6070
+    },
+    {
+      "epoch": 0.42153867518400223,
+      "grad_norm": 3.427054617712613,
+      "learning_rate": 6.4910865423689626e-06,
+      "loss": 0.5088,
+      "step": 6071
+    },
+    {
+      "epoch": 0.42160810998472437,
+      "grad_norm": 5.068917800583965,
+      "learning_rate": 6.490013183775594e-06,
+      "loss": 0.6713,
+      "step": 6072
+    },
+    {
+      "epoch": 0.42167754478544645,
+      "grad_norm": 2.866985555881974,
+      "learning_rate": 6.488939749819045e-06,
+      "loss": 0.3681,
+      "step": 6073
+    },
+    {
+      "epoch": 0.4217469795861686,
+      "grad_norm": 4.196937653836931,
+      "learning_rate": 6.487866240553608e-06,
+      "loss": 0.6518,
+      "step": 6074
+    },
+    {
+      "epoch": 0.4218164143868907,
+      "grad_norm": 3.9649150266569557,
+      "learning_rate": 6.4867926560335805e-06,
+      "loss": 0.6451,
+      "step": 6075
+    },
+    {
+      "epoch": 0.42188584918761285,
+      "grad_norm": 3.7453206291006342,
+      "learning_rate": 6.485718996313262e-06,
+      "loss": 0.5116,
+      "step": 6076
+    },
+    {
+      "epoch": 0.42195528398833493,
+      "grad_norm": 3.823194618011116,
+      "learning_rate": 6.484645261446959e-06,
+      "loss": 0.5023,
+      "step": 6077
+    },
+    {
+      "epoch": 0.42202471878905706,
+      "grad_norm": 3.009890919251052,
+      "learning_rate": 6.483571451488977e-06,
+      "loss": 0.2257,
+      "step": 6078
+    },
+    {
+      "epoch": 0.4220941535897792,
+      "grad_norm": 3.2477430597453893,
+      "learning_rate": 6.4824975664936305e-06,
+      "loss": 0.3676,
+      "step": 6079
+    },
+    {
+      "epoch": 0.42216358839050133,
+      "grad_norm": 4.068978198132351,
+      "learning_rate": 6.4814236065152355e-06,
+      "loss": 0.4056,
+      "step": 6080
+    },
+    {
+      "epoch": 0.42223302319122347,
+      "grad_norm": 3.7073421207993302,
+      "learning_rate": 6.480349571608111e-06,
+      "loss": 0.3759,
+      "step": 6081
+    },
+    {
+      "epoch": 0.42230245799194555,
+      "grad_norm": 3.441880660842226,
+      "learning_rate": 6.479275461826579e-06,
+      "loss": 0.4573,
+      "step": 6082
+    },
+    {
+      "epoch": 0.4223718927926677,
+      "grad_norm": 2.769291940772123,
+      "learning_rate": 6.478201277224969e-06,
+      "loss": 0.367,
+      "step": 6083
+    },
+    {
+      "epoch": 0.4224413275933898,
+      "grad_norm": 2.6160906696091972,
+      "learning_rate": 6.477127017857611e-06,
+      "loss": 0.1689,
+      "step": 6084
+    },
+    {
+      "epoch": 0.42251076239411195,
+      "grad_norm": 3.895257025321492,
+      "learning_rate": 6.4760526837788384e-06,
+      "loss": 0.3702,
+      "step": 6085
+    },
+    {
+      "epoch": 0.42258019719483403,
+      "grad_norm": 4.112193591869085,
+      "learning_rate": 6.474978275042994e-06,
+      "loss": 0.4545,
+      "step": 6086
+    },
+    {
+      "epoch": 0.42264963199555616,
+      "grad_norm": 2.1516462197723025,
+      "learning_rate": 6.4739037917044145e-06,
+      "loss": 0.2185,
+      "step": 6087
+    },
+    {
+      "epoch": 0.4227190667962783,
+      "grad_norm": 2.939015898205002,
+      "learning_rate": 6.472829233817452e-06,
+      "loss": 0.1933,
+      "step": 6088
+    },
+    {
+      "epoch": 0.42278850159700043,
+      "grad_norm": 3.624853366272666,
+      "learning_rate": 6.471754601436452e-06,
+      "loss": 0.4424,
+      "step": 6089
+    },
+    {
+      "epoch": 0.4228579363977225,
+      "grad_norm": 5.082966205152242,
+      "learning_rate": 6.47067989461577e-06,
+      "loss": 0.5578,
+      "step": 6090
+    },
+    {
+      "epoch": 0.42292737119844465,
+      "grad_norm": 4.224631542291856,
+      "learning_rate": 6.4696051134097624e-06,
+      "loss": 0.7495,
+      "step": 6091
+    },
+    {
+      "epoch": 0.4229968059991668,
+      "grad_norm": 3.43386118208664,
+      "learning_rate": 6.46853025787279e-06,
+      "loss": 0.3722,
+      "step": 6092
+    },
+    {
+      "epoch": 0.4230662407998889,
+      "grad_norm": 1.9495776387695833,
+      "learning_rate": 6.467455328059222e-06,
+      "loss": 0.1817,
+      "step": 6093
+    },
+    {
+      "epoch": 0.42313567560061105,
+      "grad_norm": 4.181882839237819,
+      "learning_rate": 6.466380324023423e-06,
+      "loss": 0.396,
+      "step": 6094
+    },
+    {
+      "epoch": 0.42320511040133313,
+      "grad_norm": 3.1779226403844603,
+      "learning_rate": 6.465305245819765e-06,
+      "loss": 0.6808,
+      "step": 6095
+    },
+    {
+      "epoch": 0.42327454520205526,
+      "grad_norm": 4.439537409913754,
+      "learning_rate": 6.464230093502627e-06,
+      "loss": 0.5226,
+      "step": 6096
+    },
+    {
+      "epoch": 0.4233439800027774,
+      "grad_norm": 3.4228970498265068,
+      "learning_rate": 6.4631548671263876e-06,
+      "loss": 0.5751,
+      "step": 6097
+    },
+    {
+      "epoch": 0.42341341480349953,
+      "grad_norm": 3.804722097950861,
+      "learning_rate": 6.4620795667454315e-06,
+      "loss": 0.615,
+      "step": 6098
+    },
+    {
+      "epoch": 0.4234828496042216,
+      "grad_norm": 3.053084757824642,
+      "learning_rate": 6.4610041924141445e-06,
+      "loss": 0.3096,
+      "step": 6099
+    },
+    {
+      "epoch": 0.42355228440494375,
+      "grad_norm": 3.2253592180049804,
+      "learning_rate": 6.459928744186919e-06,
+      "loss": 0.441,
+      "step": 6100
+    },
+    {
+      "epoch": 0.4236217192056659,
+      "grad_norm": 3.2217589147523253,
+      "learning_rate": 6.4588532221181484e-06,
+      "loss": 0.3218,
+      "step": 6101
+    },
+    {
+      "epoch": 0.423691154006388,
+      "grad_norm": 4.160929195466393,
+      "learning_rate": 6.457777626262234e-06,
+      "loss": 0.4574,
+      "step": 6102
+    },
+    {
+      "epoch": 0.42376058880711015,
+      "grad_norm": 3.363528054234572,
+      "learning_rate": 6.456701956673577e-06,
+      "loss": 0.3042,
+      "step": 6103
+    },
+    {
+      "epoch": 0.42383002360783223,
+      "grad_norm": 3.4362607872843194,
+      "learning_rate": 6.455626213406583e-06,
+      "loss": 0.2377,
+      "step": 6104
+    },
+    {
+      "epoch": 0.42389945840855436,
+      "grad_norm": 4.450401785729174,
+      "learning_rate": 6.454550396515661e-06,
+      "loss": 0.5959,
+      "step": 6105
+    },
+    {
+      "epoch": 0.4239688932092765,
+      "grad_norm": 4.395041436749567,
+      "learning_rate": 6.453474506055228e-06,
+      "loss": 0.4338,
+      "step": 6106
+    },
+    {
+      "epoch": 0.42403832800999863,
+      "grad_norm": 3.4028518171125057,
+      "learning_rate": 6.4523985420796984e-06,
+      "loss": 0.3306,
+      "step": 6107
+    },
+    {
+      "epoch": 0.4241077628107207,
+      "grad_norm": 5.809384522899484,
+      "learning_rate": 6.451322504643493e-06,
+      "loss": 0.5098,
+      "step": 6108
+    },
+    {
+      "epoch": 0.42417719761144285,
+      "grad_norm": 2.8713062600082533,
+      "learning_rate": 6.45024639380104e-06,
+      "loss": 0.3193,
+      "step": 6109
+    },
+    {
+      "epoch": 0.424246632412165,
+      "grad_norm": 2.7163570111851105,
+      "learning_rate": 6.4491702096067634e-06,
+      "loss": 0.2602,
+      "step": 6110
+    },
+    {
+      "epoch": 0.4243160672128871,
+      "grad_norm": 4.115612960663711,
+      "learning_rate": 6.448093952115098e-06,
+      "loss": 0.3214,
+      "step": 6111
+    },
+    {
+      "epoch": 0.4243855020136092,
+      "grad_norm": 3.865730155726266,
+      "learning_rate": 6.447017621380479e-06,
+      "loss": 0.3075,
+      "step": 6112
+    },
+    {
+      "epoch": 0.42445493681433133,
+      "grad_norm": 3.7544772635090333,
+      "learning_rate": 6.4459412174573464e-06,
+      "loss": 0.4696,
+      "step": 6113
+    },
+    {
+      "epoch": 0.42452437161505346,
+      "grad_norm": 3.7601954440388607,
+      "learning_rate": 6.444864740400143e-06,
+      "loss": 0.4165,
+      "step": 6114
+    },
+    {
+      "epoch": 0.4245938064157756,
+      "grad_norm": 3.511231598413274,
+      "learning_rate": 6.443788190263316e-06,
+      "loss": 0.498,
+      "step": 6115
+    },
+    {
+      "epoch": 0.42466324121649773,
+      "grad_norm": 3.9591369495328705,
+      "learning_rate": 6.4427115671013165e-06,
+      "loss": 0.6344,
+      "step": 6116
+    },
+    {
+      "epoch": 0.4247326760172198,
+      "grad_norm": 3.8396890711328138,
+      "learning_rate": 6.441634870968597e-06,
+      "loss": 0.4813,
+      "step": 6117
+    },
+    {
+      "epoch": 0.42480211081794195,
+      "grad_norm": 4.641531558147842,
+      "learning_rate": 6.440558101919619e-06,
+      "loss": 0.5964,
+      "step": 6118
+    },
+    {
+      "epoch": 0.4248715456186641,
+      "grad_norm": 3.562442047518287,
+      "learning_rate": 6.4394812600088415e-06,
+      "loss": 0.3657,
+      "step": 6119
+    },
+    {
+      "epoch": 0.4249409804193862,
+      "grad_norm": 3.540489536958872,
+      "learning_rate": 6.4384043452907305e-06,
+      "loss": 0.5034,
+      "step": 6120
+    },
+    {
+      "epoch": 0.4250104152201083,
+      "grad_norm": 3.7734022411787715,
+      "learning_rate": 6.4373273578197545e-06,
+      "loss": 0.6397,
+      "step": 6121
+    },
+    {
+      "epoch": 0.42507985002083043,
+      "grad_norm": 3.2766099514332834,
+      "learning_rate": 6.43625029765039e-06,
+      "loss": 0.368,
+      "step": 6122
+    },
+    {
+      "epoch": 0.42514928482155256,
+      "grad_norm": 3.203916049624322,
+      "learning_rate": 6.435173164837109e-06,
+      "loss": 0.5466,
+      "step": 6123
+    },
+    {
+      "epoch": 0.4252187196222747,
+      "grad_norm": 3.3270231272103334,
+      "learning_rate": 6.434095959434392e-06,
+      "loss": 0.3635,
+      "step": 6124
+    },
+    {
+      "epoch": 0.42528815442299683,
+      "grad_norm": 4.024554458056902,
+      "learning_rate": 6.433018681496728e-06,
+      "loss": 0.3972,
+      "step": 6125
+    },
+    {
+      "epoch": 0.4253575892237189,
+      "grad_norm": 4.436967559525287,
+      "learning_rate": 6.4319413310785965e-06,
+      "loss": 0.5474,
+      "step": 6126
+    },
+    {
+      "epoch": 0.42542702402444105,
+      "grad_norm": 3.441141697285122,
+      "learning_rate": 6.430863908234495e-06,
+      "loss": 0.4243,
+      "step": 6127
+    },
+    {
+      "epoch": 0.4254964588251632,
+      "grad_norm": 2.9312991765415015,
+      "learning_rate": 6.429786413018915e-06,
+      "loss": 0.2505,
+      "step": 6128
+    },
+    {
+      "epoch": 0.4255658936258853,
+      "grad_norm": 3.39208262346195,
+      "learning_rate": 6.428708845486358e-06,
+      "loss": 0.4172,
+      "step": 6129
+    },
+    {
+      "epoch": 0.4256353284266074,
+      "grad_norm": 3.6616072857145037,
+      "learning_rate": 6.4276312056913226e-06,
+      "loss": 0.4142,
+      "step": 6130
+    },
+    {
+      "epoch": 0.42570476322732953,
+      "grad_norm": 4.053630312940221,
+      "learning_rate": 6.426553493688315e-06,
+      "loss": 0.3978,
+      "step": 6131
+    },
+    {
+      "epoch": 0.42577419802805166,
+      "grad_norm": 4.025600586076838,
+      "learning_rate": 6.42547570953185e-06,
+      "loss": 0.4863,
+      "step": 6132
+    },
+    {
+      "epoch": 0.4258436328287738,
+      "grad_norm": 4.595075601590586,
+      "learning_rate": 6.424397853276433e-06,
+      "loss": 0.4768,
+      "step": 6133
+    },
+    {
+      "epoch": 0.4259130676294959,
+      "grad_norm": 3.647108490783121,
+      "learning_rate": 6.423319924976586e-06,
+      "loss": 0.489,
+      "step": 6134
+    },
+    {
+      "epoch": 0.425982502430218,
+      "grad_norm": 4.5010524737115185,
+      "learning_rate": 6.422241924686827e-06,
+      "loss": 0.604,
+      "step": 6135
+    },
+    {
+      "epoch": 0.42605193723094015,
+      "grad_norm": 3.9702804083668957,
+      "learning_rate": 6.421163852461682e-06,
+      "loss": 0.6158,
+      "step": 6136
+    },
+    {
+      "epoch": 0.4261213720316623,
+      "grad_norm": 3.738192682727602,
+      "learning_rate": 6.420085708355676e-06,
+      "loss": 0.4821,
+      "step": 6137
+    },
+    {
+      "epoch": 0.4261908068323844,
+      "grad_norm": 2.7251659153171075,
+      "learning_rate": 6.419007492423344e-06,
+      "loss": 0.2899,
+      "step": 6138
+    },
+    {
+      "epoch": 0.4262602416331065,
+      "grad_norm": 3.942330946782234,
+      "learning_rate": 6.417929204719218e-06,
+      "loss": 0.5805,
+      "step": 6139
+    },
+    {
+      "epoch": 0.42632967643382863,
+      "grad_norm": 3.5241678269643626,
+      "learning_rate": 6.416850845297835e-06,
+      "loss": 0.4847,
+      "step": 6140
+    },
+    {
+      "epoch": 0.42639911123455077,
+      "grad_norm": 3.872661473144205,
+      "learning_rate": 6.415772414213741e-06,
+      "loss": 0.3761,
+      "step": 6141
+    },
+    {
+      "epoch": 0.4264685460352729,
+      "grad_norm": 4.423697086361109,
+      "learning_rate": 6.414693911521481e-06,
+      "loss": 0.5974,
+      "step": 6142
+    },
+    {
+      "epoch": 0.426537980835995,
+      "grad_norm": 4.151335729475707,
+      "learning_rate": 6.413615337275604e-06,
+      "loss": 0.4959,
+      "step": 6143
+    },
+    {
+      "epoch": 0.4266074156367171,
+      "grad_norm": 3.471652507019552,
+      "learning_rate": 6.412536691530662e-06,
+      "loss": 0.383,
+      "step": 6144
+    },
+    {
+      "epoch": 0.42667685043743925,
+      "grad_norm": 3.374826871939644,
+      "learning_rate": 6.411457974341213e-06,
+      "loss": 0.1825,
+      "step": 6145
+    },
+    {
+      "epoch": 0.4267462852381614,
+      "grad_norm": 5.995279899736149,
+      "learning_rate": 6.410379185761816e-06,
+      "loss": 0.4838,
+      "step": 6146
+    },
+    {
+      "epoch": 0.42681572003888346,
+      "grad_norm": 3.34778237916351,
+      "learning_rate": 6.409300325847036e-06,
+      "loss": 0.4385,
+      "step": 6147
+    },
+    {
+      "epoch": 0.4268851548396056,
+      "grad_norm": 3.7724298815392494,
+      "learning_rate": 6.408221394651441e-06,
+      "loss": 0.4342,
+      "step": 6148
+    },
+    {
+      "epoch": 0.42695458964032773,
+      "grad_norm": 4.057265289047125,
+      "learning_rate": 6.407142392229601e-06,
+      "loss": 0.4915,
+      "step": 6149
+    },
+    {
+      "epoch": 0.42702402444104987,
+      "grad_norm": 3.4926721763659225,
+      "learning_rate": 6.406063318636091e-06,
+      "loss": 0.2714,
+      "step": 6150
+    },
+    {
+      "epoch": 0.427093459241772,
+      "grad_norm": 3.1131552205320245,
+      "learning_rate": 6.40498417392549e-06,
+      "loss": 0.3838,
+      "step": 6151
+    },
+    {
+      "epoch": 0.4271628940424941,
+      "grad_norm": 3.361258014981579,
+      "learning_rate": 6.403904958152378e-06,
+      "loss": 0.1894,
+      "step": 6152
+    },
+    {
+      "epoch": 0.4272323288432162,
+      "grad_norm": 3.218553734409216,
+      "learning_rate": 6.402825671371342e-06,
+      "loss": 0.4672,
+      "step": 6153
+    },
+    {
+      "epoch": 0.42730176364393835,
+      "grad_norm": 4.273292563547726,
+      "learning_rate": 6.401746313636973e-06,
+      "loss": 0.331,
+      "step": 6154
+    },
+    {
+      "epoch": 0.4273711984446605,
+      "grad_norm": 4.316483317142835,
+      "learning_rate": 6.400666885003861e-06,
+      "loss": 0.6188,
+      "step": 6155
+    },
+    {
+      "epoch": 0.42744063324538256,
+      "grad_norm": 3.314495097353974,
+      "learning_rate": 6.399587385526601e-06,
+      "loss": 0.4089,
+      "step": 6156
+    },
+    {
+      "epoch": 0.4275100680461047,
+      "grad_norm": 3.044933522252987,
+      "learning_rate": 6.398507815259796e-06,
+      "loss": 0.2628,
+      "step": 6157
+    },
+    {
+      "epoch": 0.42757950284682683,
+      "grad_norm": 3.628670345301784,
+      "learning_rate": 6.397428174258048e-06,
+      "loss": 0.4727,
+      "step": 6158
+    },
+    {
+      "epoch": 0.42764893764754897,
+      "grad_norm": 4.498950609684083,
+      "learning_rate": 6.3963484625759645e-06,
+      "loss": 0.4911,
+      "step": 6159
+    },
+    {
+      "epoch": 0.4277183724482711,
+      "grad_norm": 8.158931495231998,
+      "learning_rate": 6.3952686802681565e-06,
+      "loss": 0.4679,
+      "step": 6160
+    },
+    {
+      "epoch": 0.4277878072489932,
+      "grad_norm": 3.521487697685304,
+      "learning_rate": 6.394188827389236e-06,
+      "loss": 0.6341,
+      "step": 6161
+    },
+    {
+      "epoch": 0.4278572420497153,
+      "grad_norm": 4.238024758148727,
+      "learning_rate": 6.3931089039938225e-06,
+      "loss": 0.8481,
+      "step": 6162
+    },
+    {
+      "epoch": 0.42792667685043745,
+      "grad_norm": 2.9021670290139414,
+      "learning_rate": 6.3920289101365375e-06,
+      "loss": 0.3003,
+      "step": 6163
+    },
+    {
+      "epoch": 0.4279961116511596,
+      "grad_norm": 2.0578058299711364,
+      "learning_rate": 6.390948845872007e-06,
+      "loss": 0.2322,
+      "step": 6164
+    },
+    {
+      "epoch": 0.42806554645188166,
+      "grad_norm": 3.4358771304708915,
+      "learning_rate": 6.389868711254855e-06,
+      "loss": 0.523,
+      "step": 6165
+    },
+    {
+      "epoch": 0.4281349812526038,
+      "grad_norm": 3.802293289421713,
+      "learning_rate": 6.388788506339718e-06,
+      "loss": 0.58,
+      "step": 6166
+    },
+    {
+      "epoch": 0.42820441605332593,
+      "grad_norm": 3.9697112986486953,
+      "learning_rate": 6.387708231181229e-06,
+      "loss": 0.5044,
+      "step": 6167
+    },
+    {
+      "epoch": 0.42827385085404807,
+      "grad_norm": 3.8060208513613256,
+      "learning_rate": 6.3866278858340295e-06,
+      "loss": 0.5174,
+      "step": 6168
+    },
+    {
+      "epoch": 0.42834328565477015,
+      "grad_norm": 4.411844931449198,
+      "learning_rate": 6.385547470352758e-06,
+      "loss": 0.5321,
+      "step": 6169
+    },
+    {
+      "epoch": 0.4284127204554923,
+      "grad_norm": 4.516817406719567,
+      "learning_rate": 6.384466984792066e-06,
+      "loss": 0.6663,
+      "step": 6170
+    },
+    {
+      "epoch": 0.4284821552562144,
+      "grad_norm": 4.313660558214254,
+      "learning_rate": 6.3833864292065985e-06,
+      "loss": 0.5489,
+      "step": 6171
+    },
+    {
+      "epoch": 0.42855159005693655,
+      "grad_norm": 4.453285933251541,
+      "learning_rate": 6.3823058036510124e-06,
+      "loss": 0.6331,
+      "step": 6172
+    },
+    {
+      "epoch": 0.4286210248576587,
+      "grad_norm": 2.6972715093604607,
+      "learning_rate": 6.3812251081799615e-06,
+      "loss": 0.4027,
+      "step": 6173
+    },
+    {
+      "epoch": 0.42869045965838076,
+      "grad_norm": 2.6017593600327458,
+      "learning_rate": 6.380144342848109e-06,
+      "loss": 0.1743,
+      "step": 6174
+    },
+    {
+      "epoch": 0.4287598944591029,
+      "grad_norm": 4.3261312890547785,
+      "learning_rate": 6.379063507710117e-06,
+      "loss": 0.5511,
+      "step": 6175
+    },
+    {
+      "epoch": 0.42882932925982503,
+      "grad_norm": 4.05977718061093,
+      "learning_rate": 6.377982602820651e-06,
+      "loss": 0.5268,
+      "step": 6176
+    },
+    {
+      "epoch": 0.42889876406054717,
+      "grad_norm": 3.206955776170202,
+      "learning_rate": 6.376901628234389e-06,
+      "loss": 0.432,
+      "step": 6177
+    },
+    {
+      "epoch": 0.42896819886126925,
+      "grad_norm": 2.4015639265780817,
+      "learning_rate": 6.375820584005996e-06,
+      "loss": 0.168,
+      "step": 6178
+    },
+    {
+      "epoch": 0.4290376336619914,
+      "grad_norm": 3.6136295905364353,
+      "learning_rate": 6.374739470190157e-06,
+      "loss": 0.3063,
+      "step": 6179
+    },
+    {
+      "epoch": 0.4291070684627135,
+      "grad_norm": 4.166383954490124,
+      "learning_rate": 6.373658286841552e-06,
+      "loss": 0.5429,
+      "step": 6180
+    },
+    {
+      "epoch": 0.42917650326343565,
+      "grad_norm": 3.2563161863746743,
+      "learning_rate": 6.372577034014863e-06,
+      "loss": 0.4107,
+      "step": 6181
+    },
+    {
+      "epoch": 0.42924593806415773,
+      "grad_norm": 3.6736664605662046,
+      "learning_rate": 6.371495711764783e-06,
+      "loss": 0.5004,
+      "step": 6182
+    },
+    {
+      "epoch": 0.42931537286487986,
+      "grad_norm": 3.2907558128364474,
+      "learning_rate": 6.370414320146e-06,
+      "loss": 0.3884,
+      "step": 6183
+    },
+    {
+      "epoch": 0.429384807665602,
+      "grad_norm": 5.364322915852689,
+      "learning_rate": 6.3693328592132135e-06,
+      "loss": 0.6337,
+      "step": 6184
+    },
+    {
+      "epoch": 0.42945424246632413,
+      "grad_norm": 4.457239379309641,
+      "learning_rate": 6.368251329021118e-06,
+      "loss": 0.6233,
+      "step": 6185
+    },
+    {
+      "epoch": 0.42952367726704627,
+      "grad_norm": 3.7179634433302873,
+      "learning_rate": 6.3671697296244195e-06,
+      "loss": 0.6133,
+      "step": 6186
+    },
+    {
+      "epoch": 0.42959311206776835,
+      "grad_norm": 3.2921191653155533,
+      "learning_rate": 6.366088061077823e-06,
+      "loss": 0.4374,
+      "step": 6187
+    },
+    {
+      "epoch": 0.4296625468684905,
+      "grad_norm": 3.2560945071811114,
+      "learning_rate": 6.365006323436039e-06,
+      "loss": 0.4529,
+      "step": 6188
+    },
+    {
+      "epoch": 0.4297319816692126,
+      "grad_norm": 2.7333213474512426,
+      "learning_rate": 6.363924516753779e-06,
+      "loss": 0.3384,
+      "step": 6189
+    },
+    {
+      "epoch": 0.42980141646993475,
+      "grad_norm": 3.598629968181966,
+      "learning_rate": 6.3628426410857615e-06,
+      "loss": 0.5157,
+      "step": 6190
+    },
+    {
+      "epoch": 0.42987085127065683,
+      "grad_norm": 3.426960860786126,
+      "learning_rate": 6.361760696486704e-06,
+      "loss": 0.3785,
+      "step": 6191
+    },
+    {
+      "epoch": 0.42994028607137896,
+      "grad_norm": 2.7514027894688455,
+      "learning_rate": 6.360678683011332e-06,
+      "loss": 0.2937,
+      "step": 6192
+    },
+    {
+      "epoch": 0.4300097208721011,
+      "grad_norm": 3.3941879927613594,
+      "learning_rate": 6.359596600714373e-06,
+      "loss": 0.323,
+      "step": 6193
+    },
+    {
+      "epoch": 0.43007915567282323,
+      "grad_norm": 3.2927727007323058,
+      "learning_rate": 6.358514449650554e-06,
+      "loss": 0.339,
+      "step": 6194
+    },
+    {
+      "epoch": 0.43014859047354537,
+      "grad_norm": 4.527619586004049,
+      "learning_rate": 6.3574322298746126e-06,
+      "loss": 0.6325,
+      "step": 6195
+    },
+    {
+      "epoch": 0.43021802527426745,
+      "grad_norm": 3.3550187720318934,
+      "learning_rate": 6.356349941441286e-06,
+      "loss": 0.3778,
+      "step": 6196
+    },
+    {
+      "epoch": 0.4302874600749896,
+      "grad_norm": 3.657734296225048,
+      "learning_rate": 6.355267584405314e-06,
+      "loss": 0.4068,
+      "step": 6197
+    },
+    {
+      "epoch": 0.4303568948757117,
+      "grad_norm": 4.569850246635278,
+      "learning_rate": 6.35418515882144e-06,
+      "loss": 0.5493,
+      "step": 6198
+    },
+    {
+      "epoch": 0.43042632967643385,
+      "grad_norm": 4.198647490754046,
+      "learning_rate": 6.353102664744413e-06,
+      "loss": 0.4748,
+      "step": 6199
+    },
+    {
+      "epoch": 0.43049576447715593,
+      "grad_norm": 3.158135947762238,
+      "learning_rate": 6.352020102228985e-06,
+      "loss": 0.4843,
+      "step": 6200
+    },
+    {
+      "epoch": 0.43056519927787806,
+      "grad_norm": 4.387471013540285,
+      "learning_rate": 6.35093747132991e-06,
+      "loss": 0.7032,
+      "step": 6201
+    },
+    {
+      "epoch": 0.4306346340786002,
+      "grad_norm": 3.2150422721582994,
+      "learning_rate": 6.349854772101946e-06,
+      "loss": 0.251,
+      "step": 6202
+    },
+    {
+      "epoch": 0.43070406887932233,
+      "grad_norm": 4.196692955772125,
+      "learning_rate": 6.348772004599856e-06,
+      "loss": 0.6914,
+      "step": 6203
+    },
+    {
+      "epoch": 0.4307735036800444,
+      "grad_norm": 3.820219439777216,
+      "learning_rate": 6.3476891688784036e-06,
+      "loss": 0.5693,
+      "step": 6204
+    },
+    {
+      "epoch": 0.43084293848076655,
+      "grad_norm": 5.856131972127276,
+      "learning_rate": 6.346606264992359e-06,
+      "loss": 0.6002,
+      "step": 6205
+    },
+    {
+      "epoch": 0.4309123732814887,
+      "grad_norm": 3.8306785474543825,
+      "learning_rate": 6.345523292996492e-06,
+      "loss": 0.3749,
+      "step": 6206
+    },
+    {
+      "epoch": 0.4309818080822108,
+      "grad_norm": 4.531593760807623,
+      "learning_rate": 6.344440252945581e-06,
+      "loss": 0.6542,
+      "step": 6207
+    },
+    {
+      "epoch": 0.43105124288293295,
+      "grad_norm": 3.0504734588578475,
+      "learning_rate": 6.3433571448943995e-06,
+      "loss": 0.3293,
+      "step": 6208
+    },
+    {
+      "epoch": 0.43112067768365503,
+      "grad_norm": 3.5630384644858895,
+      "learning_rate": 6.342273968897739e-06,
+      "loss": 0.2986,
+      "step": 6209
+    },
+    {
+      "epoch": 0.43119011248437716,
+      "grad_norm": 4.162995106672686,
+      "learning_rate": 6.341190725010376e-06,
+      "loss": 0.4781,
+      "step": 6210
+    },
+    {
+      "epoch": 0.4312595472850993,
+      "grad_norm": 3.528974969779685,
+      "learning_rate": 6.340107413287105e-06,
+      "loss": 0.4632,
+      "step": 6211
+    },
+    {
+      "epoch": 0.43132898208582143,
+      "grad_norm": 3.830091631203563,
+      "learning_rate": 6.3390240337827195e-06,
+      "loss": 0.508,
+      "step": 6212
+    },
+    {
+      "epoch": 0.4313984168865435,
+      "grad_norm": 2.965138919729062,
+      "learning_rate": 6.337940586552012e-06,
+      "loss": 0.4046,
+      "step": 6213
+    },
+    {
+      "epoch": 0.43146785168726565,
+      "grad_norm": 3.6198752805128795,
+      "learning_rate": 6.336857071649783e-06,
+      "loss": 0.4246,
+      "step": 6214
+    },
+    {
+      "epoch": 0.4315372864879878,
+      "grad_norm": 3.276994350501958,
+      "learning_rate": 6.335773489130836e-06,
+      "loss": 0.4339,
+      "step": 6215
+    },
+    {
+      "epoch": 0.4316067212887099,
+      "grad_norm": 3.8300853748447836,
+      "learning_rate": 6.334689839049979e-06,
+      "loss": 0.3931,
+      "step": 6216
+    },
+    {
+      "epoch": 0.43167615608943205,
+      "grad_norm": 3.126244155762975,
+      "learning_rate": 6.333606121462018e-06,
+      "loss": 0.3783,
+      "step": 6217
+    },
+    {
+      "epoch": 0.43174559089015413,
+      "grad_norm": 3.4044486176369273,
+      "learning_rate": 6.3325223364217695e-06,
+      "loss": 0.3964,
+      "step": 6218
+    },
+    {
+      "epoch": 0.43181502569087626,
+      "grad_norm": 3.842516779526409,
+      "learning_rate": 6.331438483984049e-06,
+      "loss": 0.4021,
+      "step": 6219
+    },
+    {
+      "epoch": 0.4318844604915984,
+      "grad_norm": 3.857044452304051,
+      "learning_rate": 6.330354564203678e-06,
+      "loss": 0.6023,
+      "step": 6220
+    },
+    {
+      "epoch": 0.43195389529232053,
+      "grad_norm": 3.203766751879517,
+      "learning_rate": 6.329270577135477e-06,
+      "loss": 0.4275,
+      "step": 6221
+    },
+    {
+      "epoch": 0.4320233300930426,
+      "grad_norm": 3.854780059053013,
+      "learning_rate": 6.328186522834276e-06,
+      "loss": 0.3416,
+      "step": 6222
+    },
+    {
+      "epoch": 0.43209276489376475,
+      "grad_norm": 4.31001493319814,
+      "learning_rate": 6.327102401354902e-06,
+      "loss": 0.3666,
+      "step": 6223
+    },
+    {
+      "epoch": 0.4321621996944869,
+      "grad_norm": 3.993518923811869,
+      "learning_rate": 6.3260182127521895e-06,
+      "loss": 0.4039,
+      "step": 6224
+    },
+    {
+      "epoch": 0.432231634495209,
+      "grad_norm": 5.170902919387802,
+      "learning_rate": 6.3249339570809785e-06,
+      "loss": 0.8354,
+      "step": 6225
+    },
+    {
+      "epoch": 0.4323010692959311,
+      "grad_norm": 3.7512461289643557,
+      "learning_rate": 6.323849634396105e-06,
+      "loss": 0.4974,
+      "step": 6226
+    },
+    {
+      "epoch": 0.43237050409665323,
+      "grad_norm": 5.063321421563801,
+      "learning_rate": 6.322765244752417e-06,
+      "loss": 0.5905,
+      "step": 6227
+    },
+    {
+      "epoch": 0.43243993889737536,
+      "grad_norm": 2.9651251060680037,
+      "learning_rate": 6.3216807882047585e-06,
+      "loss": 0.4106,
+      "step": 6228
+    },
+    {
+      "epoch": 0.4325093736980975,
+      "grad_norm": 3.864325585647238,
+      "learning_rate": 6.320596264807982e-06,
+      "loss": 0.4079,
+      "step": 6229
+    },
+    {
+      "epoch": 0.43257880849881963,
+      "grad_norm": 4.174936246643424,
+      "learning_rate": 6.319511674616941e-06,
+      "loss": 0.6877,
+      "step": 6230
+    },
+    {
+      "epoch": 0.4326482432995417,
+      "grad_norm": 4.871606079294851,
+      "learning_rate": 6.318427017686492e-06,
+      "loss": 0.529,
+      "step": 6231
+    },
+    {
+      "epoch": 0.43271767810026385,
+      "grad_norm": 4.022767134671024,
+      "learning_rate": 6.317342294071497e-06,
+      "loss": 0.6402,
+      "step": 6232
+    },
+    {
+      "epoch": 0.432787112900986,
+      "grad_norm": 4.015514653397711,
+      "learning_rate": 6.316257503826818e-06,
+      "loss": 0.538,
+      "step": 6233
+    },
+    {
+      "epoch": 0.4328565477017081,
+      "grad_norm": 4.211121252807417,
+      "learning_rate": 6.315172647007325e-06,
+      "loss": 0.5353,
+      "step": 6234
+    },
+    {
+      "epoch": 0.4329259825024302,
+      "grad_norm": 4.155505020982948,
+      "learning_rate": 6.314087723667888e-06,
+      "loss": 0.4662,
+      "step": 6235
+    },
+    {
+      "epoch": 0.43299541730315233,
+      "grad_norm": 3.309727161091552,
+      "learning_rate": 6.313002733863381e-06,
+      "loss": 0.3797,
+      "step": 6236
+    },
+    {
+      "epoch": 0.43306485210387446,
+      "grad_norm": 3.273735534154873,
+      "learning_rate": 6.311917677648679e-06,
+      "loss": 0.3887,
+      "step": 6237
+    },
+    {
+      "epoch": 0.4331342869045966,
+      "grad_norm": 4.952631455792074,
+      "learning_rate": 6.3108325550786684e-06,
+      "loss": 0.5046,
+      "step": 6238
+    },
+    {
+      "epoch": 0.4332037217053187,
+      "grad_norm": 5.177352625785571,
+      "learning_rate": 6.3097473662082295e-06,
+      "loss": 0.6613,
+      "step": 6239
+    },
+    {
+      "epoch": 0.4332731565060408,
+      "grad_norm": 2.8794287861606738,
+      "learning_rate": 6.308662111092251e-06,
+      "loss": 0.212,
+      "step": 6240
+    },
+    {
+      "epoch": 0.43334259130676295,
+      "grad_norm": 3.149121066036248,
+      "learning_rate": 6.307576789785625e-06,
+      "loss": 0.5001,
+      "step": 6241
+    },
+    {
+      "epoch": 0.4334120261074851,
+      "grad_norm": 3.9306306089303598,
+      "learning_rate": 6.306491402343242e-06,
+      "loss": 0.4868,
+      "step": 6242
+    },
+    {
+      "epoch": 0.4334814609082072,
+      "grad_norm": 4.537332670125194,
+      "learning_rate": 6.305405948820004e-06,
+      "loss": 0.5738,
+      "step": 6243
+    },
+    {
+      "epoch": 0.4335508957089293,
+      "grad_norm": 3.5635309940963964,
+      "learning_rate": 6.304320429270809e-06,
+      "loss": 0.4863,
+      "step": 6244
+    },
+    {
+      "epoch": 0.43362033050965143,
+      "grad_norm": 5.213153857910934,
+      "learning_rate": 6.303234843750564e-06,
+      "loss": 0.679,
+      "step": 6245
+    },
+    {
+      "epoch": 0.43368976531037357,
+      "grad_norm": 3.595521705175229,
+      "learning_rate": 6.3021491923141754e-06,
+      "loss": 0.5134,
+      "step": 6246
+    },
+    {
+      "epoch": 0.4337592001110957,
+      "grad_norm": 3.656660661179552,
+      "learning_rate": 6.301063475016551e-06,
+      "loss": 0.278,
+      "step": 6247
+    },
+    {
+      "epoch": 0.4338286349118178,
+      "grad_norm": 4.630509349384864,
+      "learning_rate": 6.299977691912613e-06,
+      "loss": 0.6721,
+      "step": 6248
+    },
+    {
+      "epoch": 0.4338980697125399,
+      "grad_norm": 3.682352105783668,
+      "learning_rate": 6.2988918430572706e-06,
+      "loss": 0.4432,
+      "step": 6249
+    },
+    {
+      "epoch": 0.43396750451326205,
+      "grad_norm": 3.666991064415426,
+      "learning_rate": 6.29780592850545e-06,
+      "loss": 0.1883,
+      "step": 6250
+    },
+    {
+      "epoch": 0.4340369393139842,
+      "grad_norm": 4.613861818537799,
+      "learning_rate": 6.296719948312074e-06,
+      "loss": 0.4703,
+      "step": 6251
+    },
+    {
+      "epoch": 0.4341063741147063,
+      "grad_norm": 4.381498376832106,
+      "learning_rate": 6.295633902532071e-06,
+      "loss": 0.609,
+      "step": 6252
+    },
+    {
+      "epoch": 0.4341758089154284,
+      "grad_norm": 3.8712790668841013,
+      "learning_rate": 6.29454779122037e-06,
+      "loss": 0.5603,
+      "step": 6253
+    },
+    {
+      "epoch": 0.43424524371615053,
+      "grad_norm": 3.7259353022541917,
+      "learning_rate": 6.2934616144319096e-06,
+      "loss": 0.4209,
+      "step": 6254
+    },
+    {
+      "epoch": 0.43431467851687267,
+      "grad_norm": 3.7251965620997183,
+      "learning_rate": 6.292375372221622e-06,
+      "loss": 0.3717,
+      "step": 6255
+    },
+    {
+      "epoch": 0.4343841133175948,
+      "grad_norm": 3.540529929298755,
+      "learning_rate": 6.29128906464445e-06,
+      "loss": 0.48,
+      "step": 6256
+    },
+    {
+      "epoch": 0.4344535481183169,
+      "grad_norm": 3.3442460579804005,
+      "learning_rate": 6.290202691755338e-06,
+      "loss": 0.4438,
+      "step": 6257
+    },
+    {
+      "epoch": 0.434522982919039,
+      "grad_norm": 5.512456991686067,
+      "learning_rate": 6.289116253609235e-06,
+      "loss": 0.6124,
+      "step": 6258
+    },
+    {
+      "epoch": 0.43459241771976115,
+      "grad_norm": 3.9172157549382742,
+      "learning_rate": 6.288029750261091e-06,
+      "loss": 0.3657,
+      "step": 6259
+    },
+    {
+      "epoch": 0.4346618525204833,
+      "grad_norm": 4.407281520310914,
+      "learning_rate": 6.28694318176586e-06,
+      "loss": 0.5637,
+      "step": 6260
+    },
+    {
+      "epoch": 0.43473128732120536,
+      "grad_norm": 5.406711611027732,
+      "learning_rate": 6.285856548178498e-06,
+      "loss": 0.6946,
+      "step": 6261
+    },
+    {
+      "epoch": 0.4348007221219275,
+      "grad_norm": 4.374091927254284,
+      "learning_rate": 6.284769849553967e-06,
+      "loss": 0.5585,
+      "step": 6262
+    },
+    {
+      "epoch": 0.43487015692264963,
+      "grad_norm": 4.018837631950021,
+      "learning_rate": 6.283683085947231e-06,
+      "loss": 0.5537,
+      "step": 6263
+    },
+    {
+      "epoch": 0.43493959172337177,
+      "grad_norm": 3.091788051910182,
+      "learning_rate": 6.282596257413258e-06,
+      "loss": 0.4993,
+      "step": 6264
+    },
+    {
+      "epoch": 0.4350090265240939,
+      "grad_norm": 4.355272553388547,
+      "learning_rate": 6.2815093640070156e-06,
+      "loss": 0.4774,
+      "step": 6265
+    },
+    {
+      "epoch": 0.435078461324816,
+      "grad_norm": 3.1155100456252254,
+      "learning_rate": 6.280422405783482e-06,
+      "loss": 0.2782,
+      "step": 6266
+    },
+    {
+      "epoch": 0.4351478961255381,
+      "grad_norm": 5.624575123183313,
+      "learning_rate": 6.2793353827976314e-06,
+      "loss": 0.5197,
+      "step": 6267
+    },
+    {
+      "epoch": 0.43521733092626025,
+      "grad_norm": 3.1778482473440715,
+      "learning_rate": 6.278248295104443e-06,
+      "loss": 0.4633,
+      "step": 6268
+    },
+    {
+      "epoch": 0.4352867657269824,
+      "grad_norm": 4.762522964150873,
+      "learning_rate": 6.277161142758903e-06,
+      "loss": 0.7185,
+      "step": 6269
+    },
+    {
+      "epoch": 0.43535620052770446,
+      "grad_norm": 3.193375504660091,
+      "learning_rate": 6.2760739258159995e-06,
+      "loss": 0.2972,
+      "step": 6270
+    },
+    {
+      "epoch": 0.4354256353284266,
+      "grad_norm": 3.908466918004387,
+      "learning_rate": 6.274986644330719e-06,
+      "loss": 0.514,
+      "step": 6271
+    },
+    {
+      "epoch": 0.43549507012914873,
+      "grad_norm": 3.7058670048466946,
+      "learning_rate": 6.273899298358057e-06,
+      "loss": 0.4089,
+      "step": 6272
+    },
+    {
+      "epoch": 0.43556450492987087,
+      "grad_norm": 4.342551125251282,
+      "learning_rate": 6.272811887953009e-06,
+      "loss": 0.5115,
+      "step": 6273
+    },
+    {
+      "epoch": 0.435633939730593,
+      "grad_norm": 3.5946400576630113,
+      "learning_rate": 6.271724413170579e-06,
+      "loss": 0.4494,
+      "step": 6274
+    },
+    {
+      "epoch": 0.4357033745313151,
+      "grad_norm": 3.6199582975699824,
+      "learning_rate": 6.270636874065766e-06,
+      "loss": 0.5463,
+      "step": 6275
+    },
+    {
+      "epoch": 0.4357728093320372,
+      "grad_norm": 4.911996207750501,
+      "learning_rate": 6.269549270693576e-06,
+      "loss": 0.5419,
+      "step": 6276
+    },
+    {
+      "epoch": 0.43584224413275935,
+      "grad_norm": 2.768496308600077,
+      "learning_rate": 6.268461603109023e-06,
+      "loss": 0.342,
+      "step": 6277
+    },
+    {
+      "epoch": 0.4359116789334815,
+      "grad_norm": 3.5732622311759186,
+      "learning_rate": 6.267373871367115e-06,
+      "loss": 0.4356,
+      "step": 6278
+    },
+    {
+      "epoch": 0.43598111373420356,
+      "grad_norm": 4.430813893048166,
+      "learning_rate": 6.266286075522872e-06,
+      "loss": 0.4786,
+      "step": 6279
+    },
+    {
+      "epoch": 0.4360505485349257,
+      "grad_norm": 3.9345620310893143,
+      "learning_rate": 6.265198215631313e-06,
+      "loss": 0.5277,
+      "step": 6280
+    },
+    {
+      "epoch": 0.43611998333564783,
+      "grad_norm": 3.521946262248316,
+      "learning_rate": 6.264110291747458e-06,
+      "loss": 0.4195,
+      "step": 6281
+    },
+    {
+      "epoch": 0.43618941813636997,
+      "grad_norm": 3.1177796546296315,
+      "learning_rate": 6.263022303926336e-06,
+      "loss": 0.3275,
+      "step": 6282
+    },
+    {
+      "epoch": 0.43625885293709205,
+      "grad_norm": 3.0610177893834667,
+      "learning_rate": 6.261934252222975e-06,
+      "loss": 0.2638,
+      "step": 6283
+    },
+    {
+      "epoch": 0.4363282877378142,
+      "grad_norm": 4.368808448241052,
+      "learning_rate": 6.260846136692408e-06,
+      "loss": 0.4377,
+      "step": 6284
+    },
+    {
+      "epoch": 0.4363977225385363,
+      "grad_norm": 3.685414709462015,
+      "learning_rate": 6.259757957389669e-06,
+      "loss": 0.4031,
+      "step": 6285
+    },
+    {
+      "epoch": 0.43646715733925845,
+      "grad_norm": 5.406926886525559,
+      "learning_rate": 6.258669714369799e-06,
+      "loss": 0.6408,
+      "step": 6286
+    },
+    {
+      "epoch": 0.4365365921399806,
+      "grad_norm": 3.43114825479396,
+      "learning_rate": 6.257581407687838e-06,
+      "loss": 0.4048,
+      "step": 6287
+    },
+    {
+      "epoch": 0.43660602694070266,
+      "grad_norm": 3.3213039720473776,
+      "learning_rate": 6.256493037398834e-06,
+      "loss": 0.3204,
+      "step": 6288
+    },
+    {
+      "epoch": 0.4366754617414248,
+      "grad_norm": 4.0424244421176025,
+      "learning_rate": 6.255404603557833e-06,
+      "loss": 0.5476,
+      "step": 6289
+    },
+    {
+      "epoch": 0.43674489654214693,
+      "grad_norm": 9.350532773026421,
+      "learning_rate": 6.254316106219887e-06,
+      "loss": 0.5336,
+      "step": 6290
+    },
+    {
+      "epoch": 0.43681433134286907,
+      "grad_norm": 3.500586088293062,
+      "learning_rate": 6.253227545440054e-06,
+      "loss": 0.4152,
+      "step": 6291
+    },
+    {
+      "epoch": 0.43688376614359115,
+      "grad_norm": 3.4747199313887505,
+      "learning_rate": 6.252138921273387e-06,
+      "loss": 0.3851,
+      "step": 6292
+    },
+    {
+      "epoch": 0.4369532009443133,
+      "grad_norm": 3.9514436963721535,
+      "learning_rate": 6.251050233774953e-06,
+      "loss": 0.7342,
+      "step": 6293
+    },
+    {
+      "epoch": 0.4370226357450354,
+      "grad_norm": 3.7221958352180886,
+      "learning_rate": 6.249961482999812e-06,
+      "loss": 0.4237,
+      "step": 6294
+    },
+    {
+      "epoch": 0.43709207054575755,
+      "grad_norm": 3.1242007263475964,
+      "learning_rate": 6.248872669003034e-06,
+      "loss": 0.3739,
+      "step": 6295
+    },
+    {
+      "epoch": 0.43716150534647963,
+      "grad_norm": 4.213094697758579,
+      "learning_rate": 6.247783791839691e-06,
+      "loss": 0.6812,
+      "step": 6296
+    },
+    {
+      "epoch": 0.43723094014720176,
+      "grad_norm": 3.666798381574115,
+      "learning_rate": 6.246694851564854e-06,
+      "loss": 0.354,
+      "step": 6297
+    },
+    {
+      "epoch": 0.4373003749479239,
+      "grad_norm": 3.6823700115746933,
+      "learning_rate": 6.245605848233602e-06,
+      "loss": 0.3702,
+      "step": 6298
+    },
+    {
+      "epoch": 0.43736980974864603,
+      "grad_norm": 3.0808060315751757,
+      "learning_rate": 6.244516781901018e-06,
+      "loss": 0.3097,
+      "step": 6299
+    },
+    {
+      "epoch": 0.43743924454936817,
+      "grad_norm": 3.2591826788743683,
+      "learning_rate": 6.2434276526221814e-06,
+      "loss": 0.3494,
+      "step": 6300
+    },
+    {
+      "epoch": 0.43750867935009025,
+      "grad_norm": 4.251471875757505,
+      "learning_rate": 6.242338460452181e-06,
+      "loss": 0.7552,
+      "step": 6301
+    },
+    {
+      "epoch": 0.4375781141508124,
+      "grad_norm": 3.6652863451656787,
+      "learning_rate": 6.241249205446107e-06,
+      "loss": 0.4499,
+      "step": 6302
+    },
+    {
+      "epoch": 0.4376475489515345,
+      "grad_norm": 4.3385412988438725,
+      "learning_rate": 6.240159887659054e-06,
+      "loss": 0.6957,
+      "step": 6303
+    },
+    {
+      "epoch": 0.43771698375225665,
+      "grad_norm": 3.6317915155946343,
+      "learning_rate": 6.2390705071461165e-06,
+      "loss": 0.3951,
+      "step": 6304
+    },
+    {
+      "epoch": 0.43778641855297873,
+      "grad_norm": 3.86560993896574,
+      "learning_rate": 6.237981063962395e-06,
+      "loss": 0.4704,
+      "step": 6305
+    },
+    {
+      "epoch": 0.43785585335370086,
+      "grad_norm": 3.023131027607578,
+      "learning_rate": 6.236891558162993e-06,
+      "loss": 0.3241,
+      "step": 6306
+    },
+    {
+      "epoch": 0.437925288154423,
+      "grad_norm": 3.6746180644952826,
+      "learning_rate": 6.235801989803015e-06,
+      "loss": 0.4107,
+      "step": 6307
+    },
+    {
+      "epoch": 0.43799472295514513,
+      "grad_norm": 4.491708145064552,
+      "learning_rate": 6.234712358937569e-06,
+      "loss": 0.6872,
+      "step": 6308
+    },
+    {
+      "epoch": 0.43806415775586727,
+      "grad_norm": 3.857875349762248,
+      "learning_rate": 6.2336226656217724e-06,
+      "loss": 0.5015,
+      "step": 6309
+    },
+    {
+      "epoch": 0.43813359255658935,
+      "grad_norm": 3.2917308362868787,
+      "learning_rate": 6.232532909910735e-06,
+      "loss": 0.4069,
+      "step": 6310
+    },
+    {
+      "epoch": 0.4382030273573115,
+      "grad_norm": 2.31715582346272,
+      "learning_rate": 6.231443091859578e-06,
+      "loss": 0.1419,
+      "step": 6311
+    },
+    {
+      "epoch": 0.4382724621580336,
+      "grad_norm": 4.364501155687156,
+      "learning_rate": 6.2303532115234235e-06,
+      "loss": 0.671,
+      "step": 6312
+    },
+    {
+      "epoch": 0.43834189695875575,
+      "grad_norm": 3.2326852329743465,
+      "learning_rate": 6.229263268957395e-06,
+      "loss": 0.4033,
+      "step": 6313
+    },
+    {
+      "epoch": 0.43841133175947783,
+      "grad_norm": 3.15497422172834,
+      "learning_rate": 6.228173264216622e-06,
+      "loss": 0.3396,
+      "step": 6314
+    },
+    {
+      "epoch": 0.43848076656019996,
+      "grad_norm": 4.325934785550777,
+      "learning_rate": 6.227083197356235e-06,
+      "loss": 0.6066,
+      "step": 6315
+    },
+    {
+      "epoch": 0.4385502013609221,
+      "grad_norm": 3.227545766795584,
+      "learning_rate": 6.2259930684313685e-06,
+      "loss": 0.2638,
+      "step": 6316
+    },
+    {
+      "epoch": 0.43861963616164423,
+      "grad_norm": 2.7997864323913526,
+      "learning_rate": 6.224902877497159e-06,
+      "loss": 0.3507,
+      "step": 6317
+    },
+    {
+      "epoch": 0.4386890709623663,
+      "grad_norm": 4.036725519723322,
+      "learning_rate": 6.2238126246087495e-06,
+      "loss": 0.6026,
+      "step": 6318
+    },
+    {
+      "epoch": 0.43875850576308845,
+      "grad_norm": 3.1885803647718665,
+      "learning_rate": 6.222722309821281e-06,
+      "loss": 0.265,
+      "step": 6319
+    },
+    {
+      "epoch": 0.4388279405638106,
+      "grad_norm": 3.4711771196752,
+      "learning_rate": 6.221631933189903e-06,
+      "loss": 0.3755,
+      "step": 6320
+    },
+    {
+      "epoch": 0.4388973753645327,
+      "grad_norm": 3.4957236006543977,
+      "learning_rate": 6.220541494769763e-06,
+      "loss": 0.4657,
+      "step": 6321
+    },
+    {
+      "epoch": 0.43896681016525485,
+      "grad_norm": 4.027063362632856,
+      "learning_rate": 6.219450994616015e-06,
+      "loss": 0.4228,
+      "step": 6322
+    },
+    {
+      "epoch": 0.43903624496597693,
+      "grad_norm": 3.762994585218503,
+      "learning_rate": 6.218360432783816e-06,
+      "loss": 0.5389,
+      "step": 6323
+    },
+    {
+      "epoch": 0.43910567976669906,
+      "grad_norm": 5.590656972490438,
+      "learning_rate": 6.217269809328326e-06,
+      "loss": 0.87,
+      "step": 6324
+    },
+    {
+      "epoch": 0.4391751145674212,
+      "grad_norm": 2.648865681721893,
+      "learning_rate": 6.2161791243047065e-06,
+      "loss": 0.2843,
+      "step": 6325
+    },
+    {
+      "epoch": 0.43924454936814333,
+      "grad_norm": 4.014279071194499,
+      "learning_rate": 6.215088377768121e-06,
+      "loss": 0.6164,
+      "step": 6326
+    },
+    {
+      "epoch": 0.4393139841688654,
+      "grad_norm": 3.9078202668686175,
+      "learning_rate": 6.213997569773742e-06,
+      "loss": 0.6364,
+      "step": 6327
+    },
+    {
+      "epoch": 0.43938341896958755,
+      "grad_norm": 3.698323126595387,
+      "learning_rate": 6.2129067003767395e-06,
+      "loss": 0.655,
+      "step": 6328
+    },
+    {
+      "epoch": 0.4394528537703097,
+      "grad_norm": 4.439157217765335,
+      "learning_rate": 6.2118157696322875e-06,
+      "loss": 0.5008,
+      "step": 6329
+    },
+    {
+      "epoch": 0.4395222885710318,
+      "grad_norm": 3.984877774849983,
+      "learning_rate": 6.210724777595567e-06,
+      "loss": 0.4757,
+      "step": 6330
+    },
+    {
+      "epoch": 0.43959172337175395,
+      "grad_norm": 3.1734151789047895,
+      "learning_rate": 6.209633724321756e-06,
+      "loss": 0.3155,
+      "step": 6331
+    },
+    {
+      "epoch": 0.43966115817247603,
+      "grad_norm": 3.1802912935557552,
+      "learning_rate": 6.208542609866041e-06,
+      "loss": 0.4019,
+      "step": 6332
+    },
+    {
+      "epoch": 0.43973059297319816,
+      "grad_norm": 3.8702403510380625,
+      "learning_rate": 6.207451434283607e-06,
+      "loss": 0.4299,
+      "step": 6333
+    },
+    {
+      "epoch": 0.4398000277739203,
+      "grad_norm": 4.900151106451885,
+      "learning_rate": 6.206360197629646e-06,
+      "loss": 0.6643,
+      "step": 6334
+    },
+    {
+      "epoch": 0.43986946257464243,
+      "grad_norm": 3.7047540569029005,
+      "learning_rate": 6.2052688999593525e-06,
+      "loss": 0.5053,
+      "step": 6335
+    },
+    {
+      "epoch": 0.4399388973753645,
+      "grad_norm": 3.9766997133529656,
+      "learning_rate": 6.204177541327922e-06,
+      "loss": 0.5562,
+      "step": 6336
+    },
+    {
+      "epoch": 0.44000833217608665,
+      "grad_norm": 5.037806722750332,
+      "learning_rate": 6.203086121790553e-06,
+      "loss": 0.6293,
+      "step": 6337
+    },
+    {
+      "epoch": 0.4400777669768088,
+      "grad_norm": 3.8981757847394674,
+      "learning_rate": 6.201994641402449e-06,
+      "loss": 0.4743,
+      "step": 6338
+    },
+    {
+      "epoch": 0.4401472017775309,
+      "grad_norm": 2.9708228590882637,
+      "learning_rate": 6.200903100218817e-06,
+      "loss": 0.2783,
+      "step": 6339
+    },
+    {
+      "epoch": 0.440216636578253,
+      "grad_norm": 3.3625020209834364,
+      "learning_rate": 6.199811498294864e-06,
+      "loss": 0.4026,
+      "step": 6340
+    },
+    {
+      "epoch": 0.44028607137897513,
+      "grad_norm": 3.2066176367403627,
+      "learning_rate": 6.198719835685805e-06,
+      "loss": 0.2333,
+      "step": 6341
+    },
+    {
+      "epoch": 0.44035550617969726,
+      "grad_norm": 4.689883205711582,
+      "learning_rate": 6.1976281124468515e-06,
+      "loss": 0.5064,
+      "step": 6342
+    },
+    {
+      "epoch": 0.4404249409804194,
+      "grad_norm": 4.086958159371678,
+      "learning_rate": 6.1965363286332236e-06,
+      "loss": 0.6102,
+      "step": 6343
+    },
+    {
+      "epoch": 0.44049437578114153,
+      "grad_norm": 3.5330187007705955,
+      "learning_rate": 6.195444484300143e-06,
+      "loss": 0.3546,
+      "step": 6344
+    },
+    {
+      "epoch": 0.4405638105818636,
+      "grad_norm": 3.363498297895666,
+      "learning_rate": 6.194352579502832e-06,
+      "loss": 0.5109,
+      "step": 6345
+    },
+    {
+      "epoch": 0.44063324538258575,
+      "grad_norm": 3.3883380947062847,
+      "learning_rate": 6.19326061429652e-06,
+      "loss": 0.3446,
+      "step": 6346
+    },
+    {
+      "epoch": 0.4407026801833079,
+      "grad_norm": 4.076007675652173,
+      "learning_rate": 6.192168588736436e-06,
+      "loss": 0.5026,
+      "step": 6347
+    },
+    {
+      "epoch": 0.44077211498403,
+      "grad_norm": 3.2341208733171998,
+      "learning_rate": 6.1910765028778145e-06,
+      "loss": 0.3634,
+      "step": 6348
+    },
+    {
+      "epoch": 0.4408415497847521,
+      "grad_norm": 2.382854252484818,
+      "learning_rate": 6.18998435677589e-06,
+      "loss": 0.1706,
+      "step": 6349
+    },
+    {
+      "epoch": 0.44091098458547423,
+      "grad_norm": 5.975773241650764,
+      "learning_rate": 6.188892150485904e-06,
+      "loss": 0.4608,
+      "step": 6350
+    },
+    {
+      "epoch": 0.44098041938619637,
+      "grad_norm": 3.6089222822727667,
+      "learning_rate": 6.187799884063098e-06,
+      "loss": 0.4476,
+      "step": 6351
+    },
+    {
+      "epoch": 0.4410498541869185,
+      "grad_norm": 3.432907937474817,
+      "learning_rate": 6.186707557562718e-06,
+      "loss": 0.4611,
+      "step": 6352
+    },
+    {
+      "epoch": 0.4411192889876406,
+      "grad_norm": 3.003570337475251,
+      "learning_rate": 6.185615171040013e-06,
+      "loss": 0.3365,
+      "step": 6353
+    },
+    {
+      "epoch": 0.4411887237883627,
+      "grad_norm": 3.9255578306433025,
+      "learning_rate": 6.184522724550235e-06,
+      "loss": 0.4916,
+      "step": 6354
+    },
+    {
+      "epoch": 0.44125815858908485,
+      "grad_norm": 3.54023018165851,
+      "learning_rate": 6.183430218148636e-06,
+      "loss": 0.3805,
+      "step": 6355
+    },
+    {
+      "epoch": 0.441327593389807,
+      "grad_norm": 3.075331779595415,
+      "learning_rate": 6.182337651890477e-06,
+      "loss": 0.4181,
+      "step": 6356
+    },
+    {
+      "epoch": 0.4413970281905291,
+      "grad_norm": 3.410112647264346,
+      "learning_rate": 6.181245025831017e-06,
+      "loss": 0.517,
+      "step": 6357
+    },
+    {
+      "epoch": 0.4414664629912512,
+      "grad_norm": 3.0456582535383947,
+      "learning_rate": 6.18015234002552e-06,
+      "loss": 0.3997,
+      "step": 6358
+    },
+    {
+      "epoch": 0.44153589779197333,
+      "grad_norm": 3.0990238339398872,
+      "learning_rate": 6.179059594529255e-06,
+      "loss": 0.3283,
+      "step": 6359
+    },
+    {
+      "epoch": 0.44160533259269547,
+      "grad_norm": 3.997714601804692,
+      "learning_rate": 6.177966789397487e-06,
+      "loss": 0.3722,
+      "step": 6360
+    },
+    {
+      "epoch": 0.4416747673934176,
+      "grad_norm": 3.9771153178117773,
+      "learning_rate": 6.1768739246854936e-06,
+      "loss": 0.583,
+      "step": 6361
+    },
+    {
+      "epoch": 0.4417442021941397,
+      "grad_norm": 3.829096924210249,
+      "learning_rate": 6.175781000448549e-06,
+      "loss": 0.4084,
+      "step": 6362
+    },
+    {
+      "epoch": 0.4418136369948618,
+      "grad_norm": 3.88167808298752,
+      "learning_rate": 6.17468801674193e-06,
+      "loss": 0.4129,
+      "step": 6363
+    },
+    {
+      "epoch": 0.44188307179558395,
+      "grad_norm": 4.051326177626964,
+      "learning_rate": 6.173594973620923e-06,
+      "loss": 0.4728,
+      "step": 6364
+    },
+    {
+      "epoch": 0.4419525065963061,
+      "grad_norm": 4.687033519131718,
+      "learning_rate": 6.172501871140808e-06,
+      "loss": 0.6342,
+      "step": 6365
+    },
+    {
+      "epoch": 0.4420219413970282,
+      "grad_norm": 4.518954459965344,
+      "learning_rate": 6.171408709356876e-06,
+      "loss": 0.6057,
+      "step": 6366
+    },
+    {
+      "epoch": 0.4420913761977503,
+      "grad_norm": 3.7419127766649143,
+      "learning_rate": 6.170315488324417e-06,
+      "loss": 0.3315,
+      "step": 6367
+    },
+    {
+      "epoch": 0.44216081099847243,
+      "grad_norm": 3.3656677522079748,
+      "learning_rate": 6.169222208098725e-06,
+      "loss": 0.4091,
+      "step": 6368
+    },
+    {
+      "epoch": 0.44223024579919457,
+      "grad_norm": 3.7711204979932065,
+      "learning_rate": 6.168128868735096e-06,
+      "loss": 0.3937,
+      "step": 6369
+    },
+    {
+      "epoch": 0.4422996805999167,
+      "grad_norm": 3.5046429865618083,
+      "learning_rate": 6.167035470288832e-06,
+      "loss": 0.4633,
+      "step": 6370
+    },
+    {
+      "epoch": 0.4423691154006388,
+      "grad_norm": 4.002858070146452,
+      "learning_rate": 6.1659420128152345e-06,
+      "loss": 0.3966,
+      "step": 6371
+    },
+    {
+      "epoch": 0.4424385502013609,
+      "grad_norm": 2.673015930655621,
+      "learning_rate": 6.1648484963696075e-06,
+      "loss": 0.3356,
+      "step": 6372
+    },
+    {
+      "epoch": 0.44250798500208305,
+      "grad_norm": 4.350660933679766,
+      "learning_rate": 6.163754921007264e-06,
+      "loss": 0.5224,
+      "step": 6373
+    },
+    {
+      "epoch": 0.4425774198028052,
+      "grad_norm": 2.4947880868381023,
+      "learning_rate": 6.162661286783512e-06,
+      "loss": 0.1623,
+      "step": 6374
+    },
+    {
+      "epoch": 0.44264685460352726,
+      "grad_norm": 4.6702841150731444,
+      "learning_rate": 6.161567593753668e-06,
+      "loss": 0.4899,
+      "step": 6375
+    },
+    {
+      "epoch": 0.4427162894042494,
+      "grad_norm": 4.350387679809751,
+      "learning_rate": 6.16047384197305e-06,
+      "loss": 0.5456,
+      "step": 6376
+    },
+    {
+      "epoch": 0.44278572420497153,
+      "grad_norm": 3.629506047659448,
+      "learning_rate": 6.159380031496978e-06,
+      "loss": 0.5402,
+      "step": 6377
+    },
+    {
+      "epoch": 0.44285515900569367,
+      "grad_norm": 5.27740101340111,
+      "learning_rate": 6.158286162380776e-06,
+      "loss": 0.5268,
+      "step": 6378
+    },
+    {
+      "epoch": 0.4429245938064158,
+      "grad_norm": 3.7441763676539765,
+      "learning_rate": 6.15719223467977e-06,
+      "loss": 0.5073,
+      "step": 6379
+    },
+    {
+      "epoch": 0.4429940286071379,
+      "grad_norm": 4.610113321999732,
+      "learning_rate": 6.1560982484492915e-06,
+      "loss": 0.3578,
+      "step": 6380
+    },
+    {
+      "epoch": 0.44306346340786,
+      "grad_norm": 4.000932635852753,
+      "learning_rate": 6.15500420374467e-06,
+      "loss": 0.4669,
+      "step": 6381
+    },
+    {
+      "epoch": 0.44313289820858215,
+      "grad_norm": 4.3067229951273855,
+      "learning_rate": 6.153910100621243e-06,
+      "loss": 0.6057,
+      "step": 6382
+    },
+    {
+      "epoch": 0.4432023330093043,
+      "grad_norm": 2.87851573004484,
+      "learning_rate": 6.15281593913435e-06,
+      "loss": 0.2247,
+      "step": 6383
+    },
+    {
+      "epoch": 0.44327176781002636,
+      "grad_norm": 3.087343219556703,
+      "learning_rate": 6.151721719339332e-06,
+      "loss": 0.2943,
+      "step": 6384
+    },
+    {
+      "epoch": 0.4433412026107485,
+      "grad_norm": 3.8609677856199043,
+      "learning_rate": 6.15062744129153e-06,
+      "loss": 0.4063,
+      "step": 6385
+    },
+    {
+      "epoch": 0.44341063741147063,
+      "grad_norm": 4.7439968169337225,
+      "learning_rate": 6.1495331050462966e-06,
+      "loss": 0.6875,
+      "step": 6386
+    },
+    {
+      "epoch": 0.44348007221219277,
+      "grad_norm": 3.1960805493483213,
+      "learning_rate": 6.148438710658979e-06,
+      "loss": 0.2667,
+      "step": 6387
+    },
+    {
+      "epoch": 0.44354950701291485,
+      "grad_norm": 3.550867256927177,
+      "learning_rate": 6.14734425818493e-06,
+      "loss": 0.2779,
+      "step": 6388
+    },
+    {
+      "epoch": 0.443618941813637,
+      "grad_norm": 3.716546086324014,
+      "learning_rate": 6.146249747679507e-06,
+      "loss": 0.7447,
+      "step": 6389
+    },
+    {
+      "epoch": 0.4436883766143591,
+      "grad_norm": 4.456961776346747,
+      "learning_rate": 6.145155179198069e-06,
+      "loss": 0.5507,
+      "step": 6390
+    },
+    {
+      "epoch": 0.44375781141508125,
+      "grad_norm": 4.229950912810634,
+      "learning_rate": 6.144060552795978e-06,
+      "loss": 0.7541,
+      "step": 6391
+    },
+    {
+      "epoch": 0.4438272462158034,
+      "grad_norm": 3.0815503832105957,
+      "learning_rate": 6.142965868528598e-06,
+      "loss": 0.1598,
+      "step": 6392
+    },
+    {
+      "epoch": 0.44389668101652546,
+      "grad_norm": 3.22997200764313,
+      "learning_rate": 6.1418711264513e-06,
+      "loss": 0.4398,
+      "step": 6393
+    },
+    {
+      "epoch": 0.4439661158172476,
+      "grad_norm": 3.589137294057625,
+      "learning_rate": 6.140776326619451e-06,
+      "loss": 0.5427,
+      "step": 6394
+    },
+    {
+      "epoch": 0.44403555061796973,
+      "grad_norm": 3.5423850584550665,
+      "learning_rate": 6.1396814690884254e-06,
+      "loss": 0.3325,
+      "step": 6395
+    },
+    {
+      "epoch": 0.44410498541869187,
+      "grad_norm": 3.8327896687178287,
+      "learning_rate": 6.138586553913604e-06,
+      "loss": 0.3546,
+      "step": 6396
+    },
+    {
+      "epoch": 0.44417442021941395,
+      "grad_norm": 3.62283166119116,
+      "learning_rate": 6.1374915811503585e-06,
+      "loss": 0.3159,
+      "step": 6397
+    },
+    {
+      "epoch": 0.4442438550201361,
+      "grad_norm": 5.1428058433119945,
+      "learning_rate": 6.136396550854079e-06,
+      "loss": 0.8069,
+      "step": 6398
+    },
+    {
+      "epoch": 0.4443132898208582,
+      "grad_norm": 4.1747025944983,
+      "learning_rate": 6.135301463080147e-06,
+      "loss": 0.455,
+      "step": 6399
+    },
+    {
+      "epoch": 0.44438272462158035,
+      "grad_norm": 4.245236770873,
+      "learning_rate": 6.134206317883953e-06,
+      "loss": 0.6182,
+      "step": 6400
+    },
+    {
+      "epoch": 0.4444521594223025,
+      "grad_norm": 3.5473758719600834,
+      "learning_rate": 6.133111115320884e-06,
+      "loss": 0.3919,
+      "step": 6401
+    },
+    {
+      "epoch": 0.44452159422302456,
+      "grad_norm": 3.94802805498725,
+      "learning_rate": 6.132015855446339e-06,
+      "loss": 0.4976,
+      "step": 6402
+    },
+    {
+      "epoch": 0.4445910290237467,
+      "grad_norm": 3.9741885838752844,
+      "learning_rate": 6.130920538315712e-06,
+      "loss": 0.5679,
+      "step": 6403
+    },
+    {
+      "epoch": 0.44466046382446883,
+      "grad_norm": 3.008557601102426,
+      "learning_rate": 6.129825163984406e-06,
+      "loss": 0.4498,
+      "step": 6404
+    },
+    {
+      "epoch": 0.44472989862519097,
+      "grad_norm": 4.710685062964972,
+      "learning_rate": 6.12872973250782e-06,
+      "loss": 0.5374,
+      "step": 6405
+    },
+    {
+      "epoch": 0.44479933342591305,
+      "grad_norm": 3.337555939249492,
+      "learning_rate": 6.127634243941361e-06,
+      "loss": 0.4082,
+      "step": 6406
+    },
+    {
+      "epoch": 0.4448687682266352,
+      "grad_norm": 4.878930678027899,
+      "learning_rate": 6.12653869834044e-06,
+      "loss": 0.692,
+      "step": 6407
+    },
+    {
+      "epoch": 0.4449382030273573,
+      "grad_norm": 3.847734244466304,
+      "learning_rate": 6.125443095760464e-06,
+      "loss": 0.4965,
+      "step": 6408
+    },
+    {
+      "epoch": 0.44500763782807945,
+      "grad_norm": 3.0639586629229654,
+      "learning_rate": 6.124347436256851e-06,
+      "loss": 0.3744,
+      "step": 6409
+    },
+    {
+      "epoch": 0.44507707262880153,
+      "grad_norm": 2.99496622389321,
+      "learning_rate": 6.123251719885016e-06,
+      "loss": 0.3575,
+      "step": 6410
+    },
+    {
+      "epoch": 0.44514650742952366,
+      "grad_norm": 2.5733126593609366,
+      "learning_rate": 6.122155946700381e-06,
+      "loss": 0.2548,
+      "step": 6411
+    },
+    {
+      "epoch": 0.4452159422302458,
+      "grad_norm": 3.2347342789073847,
+      "learning_rate": 6.121060116758369e-06,
+      "loss": 0.4612,
+      "step": 6412
+    },
+    {
+      "epoch": 0.44528537703096793,
+      "grad_norm": 2.49083332535861,
+      "learning_rate": 6.119964230114403e-06,
+      "loss": 0.3196,
+      "step": 6413
+    },
+    {
+      "epoch": 0.44535481183169007,
+      "grad_norm": 3.515091511047168,
+      "learning_rate": 6.1188682868239135e-06,
+      "loss": 0.6336,
+      "step": 6414
+    },
+    {
+      "epoch": 0.44542424663241215,
+      "grad_norm": 3.2685478058427773,
+      "learning_rate": 6.117772286942334e-06,
+      "loss": 0.3536,
+      "step": 6415
+    },
+    {
+      "epoch": 0.4454936814331343,
+      "grad_norm": 4.767681147435549,
+      "learning_rate": 6.116676230525096e-06,
+      "loss": 0.5015,
+      "step": 6416
+    },
+    {
+      "epoch": 0.4455631162338564,
+      "grad_norm": 3.200942202515901,
+      "learning_rate": 6.115580117627636e-06,
+      "loss": 0.436,
+      "step": 6417
+    },
+    {
+      "epoch": 0.44563255103457855,
+      "grad_norm": 3.420496222217815,
+      "learning_rate": 6.114483948305399e-06,
+      "loss": 0.5002,
+      "step": 6418
+    },
+    {
+      "epoch": 0.44570198583530063,
+      "grad_norm": 3.119375724754164,
+      "learning_rate": 6.1133877226138235e-06,
+      "loss": 0.452,
+      "step": 6419
+    },
+    {
+      "epoch": 0.44577142063602276,
+      "grad_norm": 2.6876541812729395,
+      "learning_rate": 6.112291440608357e-06,
+      "loss": 0.2981,
+      "step": 6420
+    },
+    {
+      "epoch": 0.4458408554367449,
+      "grad_norm": 3.915636930526553,
+      "learning_rate": 6.111195102344448e-06,
+      "loss": 0.4211,
+      "step": 6421
+    },
+    {
+      "epoch": 0.44591029023746703,
+      "grad_norm": 2.749146971685769,
+      "learning_rate": 6.110098707877549e-06,
+      "loss": 0.2795,
+      "step": 6422
+    },
+    {
+      "epoch": 0.44597972503818917,
+      "grad_norm": 4.0801914307425005,
+      "learning_rate": 6.1090022572631125e-06,
+      "loss": 0.2346,
+      "step": 6423
+    },
+    {
+      "epoch": 0.44604915983891125,
+      "grad_norm": 3.868554075643408,
+      "learning_rate": 6.107905750556597e-06,
+      "loss": 0.5105,
+      "step": 6424
+    },
+    {
+      "epoch": 0.4461185946396334,
+      "grad_norm": 3.9398585562685526,
+      "learning_rate": 6.106809187813464e-06,
+      "loss": 0.369,
+      "step": 6425
+    },
+    {
+      "epoch": 0.4461880294403555,
+      "grad_norm": 2.933579552805726,
+      "learning_rate": 6.105712569089171e-06,
+      "loss": 0.2349,
+      "step": 6426
+    },
+    {
+      "epoch": 0.44625746424107765,
+      "grad_norm": 4.449118146734304,
+      "learning_rate": 6.104615894439191e-06,
+      "loss": 0.5248,
+      "step": 6427
+    },
+    {
+      "epoch": 0.44632689904179973,
+      "grad_norm": 2.7846045603929617,
+      "learning_rate": 6.103519163918987e-06,
+      "loss": 0.2892,
+      "step": 6428
+    },
+    {
+      "epoch": 0.44639633384252186,
+      "grad_norm": 4.263506635889358,
+      "learning_rate": 6.102422377584033e-06,
+      "loss": 0.5561,
+      "step": 6429
+    },
+    {
+      "epoch": 0.446465768643244,
+      "grad_norm": 3.9825871560879884,
+      "learning_rate": 6.101325535489804e-06,
+      "loss": 0.5442,
+      "step": 6430
+    },
+    {
+      "epoch": 0.44653520344396613,
+      "grad_norm": 4.072769761096678,
+      "learning_rate": 6.100228637691774e-06,
+      "loss": 0.3631,
+      "step": 6431
+    },
+    {
+      "epoch": 0.4466046382446882,
+      "grad_norm": 3.810580430723449,
+      "learning_rate": 6.099131684245425e-06,
+      "loss": 0.5508,
+      "step": 6432
+    },
+    {
+      "epoch": 0.44667407304541035,
+      "grad_norm": 1.7417939742780761,
+      "learning_rate": 6.098034675206238e-06,
+      "loss": 0.1271,
+      "step": 6433
+    },
+    {
+      "epoch": 0.4467435078461325,
+      "grad_norm": 2.27168586854458,
+      "learning_rate": 6.096937610629701e-06,
+      "loss": 0.1419,
+      "step": 6434
+    },
+    {
+      "epoch": 0.4468129426468546,
+      "grad_norm": 4.316612777870374,
+      "learning_rate": 6.0958404905712994e-06,
+      "loss": 0.608,
+      "step": 6435
+    },
+    {
+      "epoch": 0.44688237744757675,
+      "grad_norm": 4.5478349034822845,
+      "learning_rate": 6.094743315086528e-06,
+      "loss": 0.4858,
+      "step": 6436
+    },
+    {
+      "epoch": 0.44695181224829883,
+      "grad_norm": 4.06929033117874,
+      "learning_rate": 6.093646084230878e-06,
+      "loss": 0.6528,
+      "step": 6437
+    },
+    {
+      "epoch": 0.44702124704902096,
+      "grad_norm": 3.248152342239782,
+      "learning_rate": 6.092548798059845e-06,
+      "loss": 0.3749,
+      "step": 6438
+    },
+    {
+      "epoch": 0.4470906818497431,
+      "grad_norm": 3.6207689253100357,
+      "learning_rate": 6.091451456628931e-06,
+      "loss": 0.4234,
+      "step": 6439
+    },
+    {
+      "epoch": 0.44716011665046523,
+      "grad_norm": 3.855759779381679,
+      "learning_rate": 6.090354059993637e-06,
+      "loss": 0.4233,
+      "step": 6440
+    },
+    {
+      "epoch": 0.4472295514511873,
+      "grad_norm": 4.024103861891615,
+      "learning_rate": 6.089256608209471e-06,
+      "loss": 0.5426,
+      "step": 6441
+    },
+    {
+      "epoch": 0.44729898625190945,
+      "grad_norm": 4.092167397978343,
+      "learning_rate": 6.0881591013319355e-06,
+      "loss": 0.4418,
+      "step": 6442
+    },
+    {
+      "epoch": 0.4473684210526316,
+      "grad_norm": 2.9814363639974117,
+      "learning_rate": 6.087061539416545e-06,
+      "loss": 0.3078,
+      "step": 6443
+    },
+    {
+      "epoch": 0.4474378558533537,
+      "grad_norm": 3.273519534715452,
+      "learning_rate": 6.085963922518812e-06,
+      "loss": 0.3414,
+      "step": 6444
+    },
+    {
+      "epoch": 0.4475072906540758,
+      "grad_norm": 3.837182545658379,
+      "learning_rate": 6.084866250694252e-06,
+      "loss": 0.4782,
+      "step": 6445
+    },
+    {
+      "epoch": 0.44757672545479793,
+      "grad_norm": 3.079212197627249,
+      "learning_rate": 6.083768523998385e-06,
+      "loss": 0.3565,
+      "step": 6446
+    },
+    {
+      "epoch": 0.44764616025552006,
+      "grad_norm": 4.4872532259516,
+      "learning_rate": 6.0826707424867316e-06,
+      "loss": 0.5113,
+      "step": 6447
+    },
+    {
+      "epoch": 0.4477155950562422,
+      "grad_norm": 4.563572380269752,
+      "learning_rate": 6.0815729062148195e-06,
+      "loss": 0.5979,
+      "step": 6448
+    },
+    {
+      "epoch": 0.44778502985696433,
+      "grad_norm": 3.274635695211956,
+      "learning_rate": 6.080475015238172e-06,
+      "loss": 0.5138,
+      "step": 6449
+    },
+    {
+      "epoch": 0.4478544646576864,
+      "grad_norm": 6.66466284073973,
+      "learning_rate": 6.079377069612321e-06,
+      "loss": 0.627,
+      "step": 6450
+    },
+    {
+      "epoch": 0.44792389945840855,
+      "grad_norm": 3.9203767325609675,
+      "learning_rate": 6.0782790693928e-06,
+      "loss": 0.6612,
+      "step": 6451
+    },
+    {
+      "epoch": 0.4479933342591307,
+      "grad_norm": 4.0129877039263535,
+      "learning_rate": 6.0771810146351425e-06,
+      "loss": 0.3677,
+      "step": 6452
+    },
+    {
+      "epoch": 0.4480627690598528,
+      "grad_norm": 4.066409085454307,
+      "learning_rate": 6.076082905394889e-06,
+      "loss": 0.5723,
+      "step": 6453
+    },
+    {
+      "epoch": 0.4481322038605749,
+      "grad_norm": 4.8410763683185,
+      "learning_rate": 6.07498474172758e-06,
+      "loss": 0.4041,
+      "step": 6454
+    },
+    {
+      "epoch": 0.44820163866129703,
+      "grad_norm": 3.761606450929358,
+      "learning_rate": 6.073886523688759e-06,
+      "loss": 0.3268,
+      "step": 6455
+    },
+    {
+      "epoch": 0.44827107346201917,
+      "grad_norm": 3.8963442184339656,
+      "learning_rate": 6.072788251333972e-06,
+      "loss": 0.4744,
+      "step": 6456
+    },
+    {
+      "epoch": 0.4483405082627413,
+      "grad_norm": 2.3374548369309753,
+      "learning_rate": 6.071689924718771e-06,
+      "loss": 0.2232,
+      "step": 6457
+    },
+    {
+      "epoch": 0.44840994306346343,
+      "grad_norm": 2.697756012898419,
+      "learning_rate": 6.070591543898704e-06,
+      "loss": 0.2619,
+      "step": 6458
+    },
+    {
+      "epoch": 0.4484793778641855,
+      "grad_norm": 3.7144267313974746,
+      "learning_rate": 6.06949310892933e-06,
+      "loss": 0.3687,
+      "step": 6459
+    },
+    {
+      "epoch": 0.44854881266490765,
+      "grad_norm": 4.884429454142245,
+      "learning_rate": 6.068394619866203e-06,
+      "loss": 0.492,
+      "step": 6460
+    },
+    {
+      "epoch": 0.4486182474656298,
+      "grad_norm": 4.355845168427054,
+      "learning_rate": 6.067296076764887e-06,
+      "loss": 0.569,
+      "step": 6461
+    },
+    {
+      "epoch": 0.4486876822663519,
+      "grad_norm": 3.569563726022667,
+      "learning_rate": 6.066197479680942e-06,
+      "loss": 0.4869,
+      "step": 6462
+    },
+    {
+      "epoch": 0.448757117067074,
+      "grad_norm": 3.644916133422157,
+      "learning_rate": 6.0650988286699355e-06,
+      "loss": 0.4771,
+      "step": 6463
+    },
+    {
+      "epoch": 0.44882655186779613,
+      "grad_norm": 5.587910492731703,
+      "learning_rate": 6.064000123787436e-06,
+      "loss": 0.6544,
+      "step": 6464
+    },
+    {
+      "epoch": 0.44889598666851827,
+      "grad_norm": 3.8562091816907516,
+      "learning_rate": 6.062901365089012e-06,
+      "loss": 0.3854,
+      "step": 6465
+    },
+    {
+      "epoch": 0.4489654214692404,
+      "grad_norm": 3.696616018881992,
+      "learning_rate": 6.061802552630242e-06,
+      "loss": 0.3737,
+      "step": 6466
+    },
+    {
+      "epoch": 0.4490348562699625,
+      "grad_norm": 3.7386340102730693,
+      "learning_rate": 6.0607036864667e-06,
+      "loss": 0.4022,
+      "step": 6467
+    },
+    {
+      "epoch": 0.4491042910706846,
+      "grad_norm": 3.4263693065715293,
+      "learning_rate": 6.059604766653966e-06,
+      "loss": 0.3583,
+      "step": 6468
+    },
+    {
+      "epoch": 0.44917372587140675,
+      "grad_norm": 3.3489562660500956,
+      "learning_rate": 6.05850579324762e-06,
+      "loss": 0.5038,
+      "step": 6469
+    },
+    {
+      "epoch": 0.4492431606721289,
+      "grad_norm": 3.5459014461282408,
+      "learning_rate": 6.057406766303251e-06,
+      "loss": 0.5291,
+      "step": 6470
+    },
+    {
+      "epoch": 0.449312595472851,
+      "grad_norm": 2.410279609674972,
+      "learning_rate": 6.056307685876443e-06,
+      "loss": 0.2024,
+      "step": 6471
+    },
+    {
+      "epoch": 0.4493820302735731,
+      "grad_norm": 4.366032966181556,
+      "learning_rate": 6.0552085520227875e-06,
+      "loss": 0.5726,
+      "step": 6472
+    },
+    {
+      "epoch": 0.44945146507429523,
+      "grad_norm": 4.947445276836761,
+      "learning_rate": 6.054109364797879e-06,
+      "loss": 0.7236,
+      "step": 6473
+    },
+    {
+      "epoch": 0.44952089987501737,
+      "grad_norm": 3.4253802678475003,
+      "learning_rate": 6.05301012425731e-06,
+      "loss": 0.4136,
+      "step": 6474
+    },
+    {
+      "epoch": 0.4495903346757395,
+      "grad_norm": 4.091537607190693,
+      "learning_rate": 6.051910830456682e-06,
+      "loss": 0.539,
+      "step": 6475
+    },
+    {
+      "epoch": 0.4496597694764616,
+      "grad_norm": 5.540231509868604,
+      "learning_rate": 6.050811483451593e-06,
+      "loss": 0.7037,
+      "step": 6476
+    },
+    {
+      "epoch": 0.4497292042771837,
+      "grad_norm": 3.4410372298804117,
+      "learning_rate": 6.04971208329765e-06,
+      "loss": 0.449,
+      "step": 6477
+    },
+    {
+      "epoch": 0.44979863907790585,
+      "grad_norm": 3.296300143342138,
+      "learning_rate": 6.0486126300504555e-06,
+      "loss": 0.512,
+      "step": 6478
+    },
+    {
+      "epoch": 0.449868073878628,
+      "grad_norm": 4.098748643240273,
+      "learning_rate": 6.047513123765622e-06,
+      "loss": 0.6582,
+      "step": 6479
+    },
+    {
+      "epoch": 0.4499375086793501,
+      "grad_norm": 5.096787399807512,
+      "learning_rate": 6.04641356449876e-06,
+      "loss": 0.4871,
+      "step": 6480
+    },
+    {
+      "epoch": 0.4500069434800722,
+      "grad_norm": 4.330007712410489,
+      "learning_rate": 6.045313952305483e-06,
+      "loss": 0.3909,
+      "step": 6481
+    },
+    {
+      "epoch": 0.45007637828079433,
+      "grad_norm": 2.9042193196347115,
+      "learning_rate": 6.04421428724141e-06,
+      "loss": 0.4092,
+      "step": 6482
+    },
+    {
+      "epoch": 0.45014581308151647,
+      "grad_norm": 4.132846516220287,
+      "learning_rate": 6.04311456936216e-06,
+      "loss": 0.5942,
+      "step": 6483
+    },
+    {
+      "epoch": 0.4502152478822386,
+      "grad_norm": 4.462373672900015,
+      "learning_rate": 6.0420147987233556e-06,
+      "loss": 0.5726,
+      "step": 6484
+    },
+    {
+      "epoch": 0.4502846826829607,
+      "grad_norm": 2.8726902520630895,
+      "learning_rate": 6.040914975380621e-06,
+      "loss": 0.3279,
+      "step": 6485
+    },
+    {
+      "epoch": 0.4503541174836828,
+      "grad_norm": 4.56609613262533,
+      "learning_rate": 6.039815099389587e-06,
+      "loss": 0.5645,
+      "step": 6486
+    },
+    {
+      "epoch": 0.45042355228440495,
+      "grad_norm": 3.518723474580144,
+      "learning_rate": 6.03871517080588e-06,
+      "loss": 0.3518,
+      "step": 6487
+    },
+    {
+      "epoch": 0.4504929870851271,
+      "grad_norm": 3.9695216053766282,
+      "learning_rate": 6.037615189685134e-06,
+      "loss": 0.3725,
+      "step": 6488
+    },
+    {
+      "epoch": 0.45056242188584916,
+      "grad_norm": 3.317152102645948,
+      "learning_rate": 6.036515156082987e-06,
+      "loss": 0.3224,
+      "step": 6489
+    },
+    {
+      "epoch": 0.4506318566865713,
+      "grad_norm": 3.4119277128209355,
+      "learning_rate": 6.035415070055077e-06,
+      "loss": 0.3289,
+      "step": 6490
+    },
+    {
+      "epoch": 0.45070129148729343,
+      "grad_norm": 4.382366282335321,
+      "learning_rate": 6.0343149316570435e-06,
+      "loss": 0.5411,
+      "step": 6491
+    },
+    {
+      "epoch": 0.45077072628801557,
+      "grad_norm": 3.438983864807977,
+      "learning_rate": 6.033214740944532e-06,
+      "loss": 0.4609,
+      "step": 6492
+    },
+    {
+      "epoch": 0.4508401610887377,
+      "grad_norm": 3.921548772247009,
+      "learning_rate": 6.032114497973188e-06,
+      "loss": 0.5714,
+      "step": 6493
+    },
+    {
+      "epoch": 0.4509095958894598,
+      "grad_norm": 3.799233028095503,
+      "learning_rate": 6.03101420279866e-06,
+      "loss": 0.6858,
+      "step": 6494
+    },
+    {
+      "epoch": 0.4509790306901819,
+      "grad_norm": 4.723661022129299,
+      "learning_rate": 6.0299138554766005e-06,
+      "loss": 0.7396,
+      "step": 6495
+    },
+    {
+      "epoch": 0.45104846549090405,
+      "grad_norm": 3.546524755539137,
+      "learning_rate": 6.028813456062665e-06,
+      "loss": 0.5713,
+      "step": 6496
+    },
+    {
+      "epoch": 0.4511179002916262,
+      "grad_norm": 3.943418532059535,
+      "learning_rate": 6.027713004612506e-06,
+      "loss": 0.5071,
+      "step": 6497
+    },
+    {
+      "epoch": 0.45118733509234826,
+      "grad_norm": 4.2650291057796945,
+      "learning_rate": 6.026612501181788e-06,
+      "loss": 0.6585,
+      "step": 6498
+    },
+    {
+      "epoch": 0.4512567698930704,
+      "grad_norm": 2.7630657741099425,
+      "learning_rate": 6.025511945826171e-06,
+      "loss": 0.2768,
+      "step": 6499
+    },
+    {
+      "epoch": 0.45132620469379253,
+      "grad_norm": 2.6287336200063374,
+      "learning_rate": 6.024411338601321e-06,
+      "loss": 0.2516,
+      "step": 6500
+    },
+    {
+      "epoch": 0.45139563949451467,
+      "grad_norm": 3.449307383703831,
+      "learning_rate": 6.023310679562904e-06,
+      "loss": 0.3791,
+      "step": 6501
+    },
+    {
+      "epoch": 0.45146507429523675,
+      "grad_norm": 3.781952534402416,
+      "learning_rate": 6.022209968766591e-06,
+      "loss": 0.3955,
+      "step": 6502
+    },
+    {
+      "epoch": 0.4515345090959589,
+      "grad_norm": 3.4150039885564665,
+      "learning_rate": 6.021109206268054e-06,
+      "loss": 0.5224,
+      "step": 6503
+    },
+    {
+      "epoch": 0.451603943896681,
+      "grad_norm": 2.5463190703288245,
+      "learning_rate": 6.020008392122969e-06,
+      "loss": 0.1787,
+      "step": 6504
+    },
+    {
+      "epoch": 0.45167337869740315,
+      "grad_norm": 3.821476738345288,
+      "learning_rate": 6.018907526387014e-06,
+      "loss": 0.3652,
+      "step": 6505
+    },
+    {
+      "epoch": 0.4517428134981253,
+      "grad_norm": 3.4833280093390493,
+      "learning_rate": 6.017806609115869e-06,
+      "loss": 0.3346,
+      "step": 6506
+    },
+    {
+      "epoch": 0.45181224829884736,
+      "grad_norm": 3.1699036982286204,
+      "learning_rate": 6.016705640365219e-06,
+      "loss": 0.3922,
+      "step": 6507
+    },
+    {
+      "epoch": 0.4518816830995695,
+      "grad_norm": 4.1190980363933605,
+      "learning_rate": 6.015604620190746e-06,
+      "loss": 0.6074,
+      "step": 6508
+    },
+    {
+      "epoch": 0.45195111790029163,
+      "grad_norm": 3.169103594163511,
+      "learning_rate": 6.0145035486481415e-06,
+      "loss": 0.3396,
+      "step": 6509
+    },
+    {
+      "epoch": 0.45202055270101377,
+      "grad_norm": 4.036900529603892,
+      "learning_rate": 6.013402425793095e-06,
+      "loss": 0.4181,
+      "step": 6510
+    },
+    {
+      "epoch": 0.45208998750173585,
+      "grad_norm": 3.68484214207789,
+      "learning_rate": 6.012301251681302e-06,
+      "loss": 0.5354,
+      "step": 6511
+    },
+    {
+      "epoch": 0.452159422302458,
+      "grad_norm": 3.511488133133981,
+      "learning_rate": 6.011200026368457e-06,
+      "loss": 0.3993,
+      "step": 6512
+    },
+    {
+      "epoch": 0.4522288571031801,
+      "grad_norm": 4.155115814296944,
+      "learning_rate": 6.010098749910257e-06,
+      "loss": 0.5291,
+      "step": 6513
+    },
+    {
+      "epoch": 0.45229829190390225,
+      "grad_norm": 3.2489209591183132,
+      "learning_rate": 6.008997422362406e-06,
+      "loss": 0.4399,
+      "step": 6514
+    },
+    {
+      "epoch": 0.4523677267046244,
+      "grad_norm": 3.408908623445261,
+      "learning_rate": 6.007896043780608e-06,
+      "loss": 0.3544,
+      "step": 6515
+    },
+    {
+      "epoch": 0.45243716150534646,
+      "grad_norm": 3.2671560823917365,
+      "learning_rate": 6.006794614220568e-06,
+      "loss": 0.3208,
+      "step": 6516
+    },
+    {
+      "epoch": 0.4525065963060686,
+      "grad_norm": 4.477703750694288,
+      "learning_rate": 6.005693133737996e-06,
+      "loss": 0.3418,
+      "step": 6517
+    },
+    {
+      "epoch": 0.45257603110679073,
+      "grad_norm": 2.9776327947096437,
+      "learning_rate": 6.004591602388602e-06,
+      "loss": 0.3322,
+      "step": 6518
+    },
+    {
+      "epoch": 0.45264546590751287,
+      "grad_norm": 3.085286799322782,
+      "learning_rate": 6.0034900202281034e-06,
+      "loss": 0.3832,
+      "step": 6519
+    },
+    {
+      "epoch": 0.45271490070823495,
+      "grad_norm": 3.5592489441454225,
+      "learning_rate": 6.0023883873122145e-06,
+      "loss": 0.2628,
+      "step": 6520
+    },
+    {
+      "epoch": 0.4527843355089571,
+      "grad_norm": 4.593385658713493,
+      "learning_rate": 6.001286703696655e-06,
+      "loss": 0.6076,
+      "step": 6521
+    },
+    {
+      "epoch": 0.4528537703096792,
+      "grad_norm": 3.956838336465241,
+      "learning_rate": 6.0001849694371475e-06,
+      "loss": 0.5129,
+      "step": 6522
+    },
+    {
+      "epoch": 0.45292320511040135,
+      "grad_norm": 4.738356196009445,
+      "learning_rate": 5.9990831845894145e-06,
+      "loss": 0.4617,
+      "step": 6523
+    },
+    {
+      "epoch": 0.45299263991112343,
+      "grad_norm": 3.34740383763524,
+      "learning_rate": 5.997981349209186e-06,
+      "loss": 0.3613,
+      "step": 6524
+    },
+    {
+      "epoch": 0.45306207471184556,
+      "grad_norm": 3.0664907865648865,
+      "learning_rate": 5.996879463352191e-06,
+      "loss": 0.2668,
+      "step": 6525
+    },
+    {
+      "epoch": 0.4531315095125677,
+      "grad_norm": 4.0369270552444405,
+      "learning_rate": 5.995777527074158e-06,
+      "loss": 0.5807,
+      "step": 6526
+    },
+    {
+      "epoch": 0.45320094431328983,
+      "grad_norm": 4.5037001228108045,
+      "learning_rate": 5.9946755404308265e-06,
+      "loss": 0.6769,
+      "step": 6527
+    },
+    {
+      "epoch": 0.45327037911401197,
+      "grad_norm": 4.175441549008766,
+      "learning_rate": 5.993573503477932e-06,
+      "loss": 0.496,
+      "step": 6528
+    },
+    {
+      "epoch": 0.45333981391473405,
+      "grad_norm": 3.7016753651379415,
+      "learning_rate": 5.992471416271213e-06,
+      "loss": 0.6437,
+      "step": 6529
+    },
+    {
+      "epoch": 0.4534092487154562,
+      "grad_norm": 4.732485345547643,
+      "learning_rate": 5.9913692788664136e-06,
+      "loss": 0.668,
+      "step": 6530
+    },
+    {
+      "epoch": 0.4534786835161783,
+      "grad_norm": 4.19583377204447,
+      "learning_rate": 5.9902670913192774e-06,
+      "loss": 0.4971,
+      "step": 6531
+    },
+    {
+      "epoch": 0.45354811831690045,
+      "grad_norm": 4.805159208359288,
+      "learning_rate": 5.989164853685552e-06,
+      "loss": 0.7558,
+      "step": 6532
+    },
+    {
+      "epoch": 0.45361755311762253,
+      "grad_norm": 3.985902015990341,
+      "learning_rate": 5.988062566020987e-06,
+      "loss": 0.7363,
+      "step": 6533
+    },
+    {
+      "epoch": 0.45368698791834466,
+      "grad_norm": 3.809640709870053,
+      "learning_rate": 5.9869602283813365e-06,
+      "loss": 0.4559,
+      "step": 6534
+    },
+    {
+      "epoch": 0.4537564227190668,
+      "grad_norm": 4.527125327110629,
+      "learning_rate": 5.985857840822355e-06,
+      "loss": 0.6012,
+      "step": 6535
+    },
+    {
+      "epoch": 0.45382585751978893,
+      "grad_norm": 3.9570781769808736,
+      "learning_rate": 5.984755403399799e-06,
+      "loss": 0.5308,
+      "step": 6536
+    },
+    {
+      "epoch": 0.45389529232051107,
+      "grad_norm": 4.306501037389282,
+      "learning_rate": 5.983652916169429e-06,
+      "loss": 0.4533,
+      "step": 6537
+    },
+    {
+      "epoch": 0.45396472712123315,
+      "grad_norm": 3.83909339676864,
+      "learning_rate": 5.982550379187008e-06,
+      "loss": 0.5921,
+      "step": 6538
+    },
+    {
+      "epoch": 0.4540341619219553,
+      "grad_norm": 4.157353401461698,
+      "learning_rate": 5.9814477925083e-06,
+      "loss": 0.3536,
+      "step": 6539
+    },
+    {
+      "epoch": 0.4541035967226774,
+      "grad_norm": 4.4099439795831685,
+      "learning_rate": 5.980345156189072e-06,
+      "loss": 0.4417,
+      "step": 6540
+    },
+    {
+      "epoch": 0.45417303152339955,
+      "grad_norm": 3.247225112953667,
+      "learning_rate": 5.979242470285098e-06,
+      "loss": 0.3139,
+      "step": 6541
+    },
+    {
+      "epoch": 0.45424246632412163,
+      "grad_norm": 3.7456758799817043,
+      "learning_rate": 5.978139734852146e-06,
+      "loss": 0.5087,
+      "step": 6542
+    },
+    {
+      "epoch": 0.45431190112484376,
+      "grad_norm": 3.3363107932578955,
+      "learning_rate": 5.977036949945993e-06,
+      "loss": 0.3107,
+      "step": 6543
+    },
+    {
+      "epoch": 0.4543813359255659,
+      "grad_norm": 4.03767880613357,
+      "learning_rate": 5.9759341156224185e-06,
+      "loss": 0.2792,
+      "step": 6544
+    },
+    {
+      "epoch": 0.45445077072628803,
+      "grad_norm": 4.033706516517885,
+      "learning_rate": 5.974831231937199e-06,
+      "loss": 0.5703,
+      "step": 6545
+    },
+    {
+      "epoch": 0.4545202055270101,
+      "grad_norm": 4.333455404575776,
+      "learning_rate": 5.97372829894612e-06,
+      "loss": 0.5762,
+      "step": 6546
+    },
+    {
+      "epoch": 0.45458964032773225,
+      "grad_norm": 4.30149083955346,
+      "learning_rate": 5.972625316704965e-06,
+      "loss": 0.5668,
+      "step": 6547
+    },
+    {
+      "epoch": 0.4546590751284544,
+      "grad_norm": 2.446215639209645,
+      "learning_rate": 5.971522285269522e-06,
+      "loss": 0.2339,
+      "step": 6548
+    },
+    {
+      "epoch": 0.4547285099291765,
+      "grad_norm": 3.013153979220055,
+      "learning_rate": 5.970419204695581e-06,
+      "loss": 0.2908,
+      "step": 6549
+    },
+    {
+      "epoch": 0.45479794472989865,
+      "grad_norm": 3.7870980705655897,
+      "learning_rate": 5.9693160750389344e-06,
+      "loss": 0.632,
+      "step": 6550
+    },
+    {
+      "epoch": 0.45486737953062073,
+      "grad_norm": 2.848945364913972,
+      "learning_rate": 5.968212896355379e-06,
+      "loss": 0.3146,
+      "step": 6551
+    },
+    {
+      "epoch": 0.45493681433134286,
+      "grad_norm": 6.099280860262838,
+      "learning_rate": 5.9671096687007105e-06,
+      "loss": 0.5378,
+      "step": 6552
+    },
+    {
+      "epoch": 0.455006249132065,
+      "grad_norm": 4.45201272408518,
+      "learning_rate": 5.966006392130729e-06,
+      "loss": 0.376,
+      "step": 6553
+    },
+    {
+      "epoch": 0.45507568393278713,
+      "grad_norm": 4.14188576035099,
+      "learning_rate": 5.9649030667012376e-06,
+      "loss": 0.5746,
+      "step": 6554
+    },
+    {
+      "epoch": 0.4551451187335092,
+      "grad_norm": 3.9794506830324523,
+      "learning_rate": 5.96379969246804e-06,
+      "loss": 0.3662,
+      "step": 6555
+    },
+    {
+      "epoch": 0.45521455353423135,
+      "grad_norm": 4.409586180191458,
+      "learning_rate": 5.962696269486944e-06,
+      "loss": 0.5659,
+      "step": 6556
+    },
+    {
+      "epoch": 0.4552839883349535,
+      "grad_norm": 3.8602612967368306,
+      "learning_rate": 5.961592797813762e-06,
+      "loss": 0.4451,
+      "step": 6557
+    },
+    {
+      "epoch": 0.4553534231356756,
+      "grad_norm": 3.1593327881630273,
+      "learning_rate": 5.960489277504303e-06,
+      "loss": 0.3196,
+      "step": 6558
+    },
+    {
+      "epoch": 0.4554228579363977,
+      "grad_norm": 4.346761784878238,
+      "learning_rate": 5.959385708614384e-06,
+      "loss": 0.5645,
+      "step": 6559
+    },
+    {
+      "epoch": 0.45549229273711983,
+      "grad_norm": 3.727137767881669,
+      "learning_rate": 5.9582820911998216e-06,
+      "loss": 0.4348,
+      "step": 6560
+    },
+    {
+      "epoch": 0.45556172753784197,
+      "grad_norm": 3.489938655162405,
+      "learning_rate": 5.957178425316434e-06,
+      "loss": 0.3333,
+      "step": 6561
+    },
+    {
+      "epoch": 0.4556311623385641,
+      "grad_norm": 3.3614363516798464,
+      "learning_rate": 5.956074711020047e-06,
+      "loss": 0.3874,
+      "step": 6562
+    },
+    {
+      "epoch": 0.45570059713928623,
+      "grad_norm": 3.5719187450217653,
+      "learning_rate": 5.9549709483664805e-06,
+      "loss": 0.3903,
+      "step": 6563
+    },
+    {
+      "epoch": 0.4557700319400083,
+      "grad_norm": 4.8100724622581525,
+      "learning_rate": 5.953867137411566e-06,
+      "loss": 0.4796,
+      "step": 6564
+    },
+    {
+      "epoch": 0.45583946674073045,
+      "grad_norm": 3.91707932192862,
+      "learning_rate": 5.95276327821113e-06,
+      "loss": 0.3591,
+      "step": 6565
+    },
+    {
+      "epoch": 0.4559089015414526,
+      "grad_norm": 4.609547147627957,
+      "learning_rate": 5.951659370821007e-06,
+      "loss": 0.7442,
+      "step": 6566
+    },
+    {
+      "epoch": 0.4559783363421747,
+      "grad_norm": 4.727461112063819,
+      "learning_rate": 5.950555415297028e-06,
+      "loss": 0.6209,
+      "step": 6567
+    },
+    {
+      "epoch": 0.4560477711428968,
+      "grad_norm": 3.5596629380724654,
+      "learning_rate": 5.949451411695033e-06,
+      "loss": 0.3818,
+      "step": 6568
+    },
+    {
+      "epoch": 0.45611720594361893,
+      "grad_norm": 3.7511576048959334,
+      "learning_rate": 5.9483473600708595e-06,
+      "loss": 0.4017,
+      "step": 6569
+    },
+    {
+      "epoch": 0.45618664074434107,
+      "grad_norm": 3.555544630357605,
+      "learning_rate": 5.94724326048035e-06,
+      "loss": 0.3939,
+      "step": 6570
+    },
+    {
+      "epoch": 0.4562560755450632,
+      "grad_norm": 4.569068145967377,
+      "learning_rate": 5.946139112979348e-06,
+      "loss": 0.6096,
+      "step": 6571
+    },
+    {
+      "epoch": 0.45632551034578533,
+      "grad_norm": 4.6059989735404665,
+      "learning_rate": 5.9450349176237e-06,
+      "loss": 0.5843,
+      "step": 6572
+    },
+    {
+      "epoch": 0.4563949451465074,
+      "grad_norm": 4.383031629464643,
+      "learning_rate": 5.943930674469255e-06,
+      "loss": 0.7014,
+      "step": 6573
+    },
+    {
+      "epoch": 0.45646437994722955,
+      "grad_norm": 3.8334233196955627,
+      "learning_rate": 5.942826383571865e-06,
+      "loss": 0.359,
+      "step": 6574
+    },
+    {
+      "epoch": 0.4565338147479517,
+      "grad_norm": 4.466379465098667,
+      "learning_rate": 5.941722044987384e-06,
+      "loss": 0.6505,
+      "step": 6575
+    },
+    {
+      "epoch": 0.4566032495486738,
+      "grad_norm": 3.7776982502947374,
+      "learning_rate": 5.940617658771667e-06,
+      "loss": 0.5938,
+      "step": 6576
+    },
+    {
+      "epoch": 0.4566726843493959,
+      "grad_norm": 3.1530693263154332,
+      "learning_rate": 5.939513224980573e-06,
+      "loss": 0.2954,
+      "step": 6577
+    },
+    {
+      "epoch": 0.45674211915011803,
+      "grad_norm": 3.922452185488341,
+      "learning_rate": 5.938408743669965e-06,
+      "loss": 0.6157,
+      "step": 6578
+    },
+    {
+      "epoch": 0.45681155395084017,
+      "grad_norm": 3.365699573063296,
+      "learning_rate": 5.937304214895702e-06,
+      "loss": 0.492,
+      "step": 6579
+    },
+    {
+      "epoch": 0.4568809887515623,
+      "grad_norm": 4.9646814560035954,
+      "learning_rate": 5.936199638713655e-06,
+      "loss": 0.5575,
+      "step": 6580
+    },
+    {
+      "epoch": 0.4569504235522844,
+      "grad_norm": 3.7922676630089254,
+      "learning_rate": 5.935095015179686e-06,
+      "loss": 0.5089,
+      "step": 6581
+    },
+    {
+      "epoch": 0.4570198583530065,
+      "grad_norm": 3.540303227263665,
+      "learning_rate": 5.933990344349671e-06,
+      "loss": 0.6771,
+      "step": 6582
+    },
+    {
+      "epoch": 0.45708929315372865,
+      "grad_norm": 3.8816923951348197,
+      "learning_rate": 5.932885626279483e-06,
+      "loss": 0.4535,
+      "step": 6583
+    },
+    {
+      "epoch": 0.4571587279544508,
+      "grad_norm": 2.663319807428733,
+      "learning_rate": 5.931780861024995e-06,
+      "loss": 0.3249,
+      "step": 6584
+    },
+    {
+      "epoch": 0.4572281627551729,
+      "grad_norm": 4.029833467360081,
+      "learning_rate": 5.930676048642083e-06,
+      "loss": 0.5332,
+      "step": 6585
+    },
+    {
+      "epoch": 0.457297597555895,
+      "grad_norm": 3.217052442293713,
+      "learning_rate": 5.929571189186634e-06,
+      "loss": 0.5392,
+      "step": 6586
+    },
+    {
+      "epoch": 0.45736703235661713,
+      "grad_norm": 3.542865908716853,
+      "learning_rate": 5.928466282714523e-06,
+      "loss": 0.3384,
+      "step": 6587
+    },
+    {
+      "epoch": 0.45743646715733927,
+      "grad_norm": 4.324082415072232,
+      "learning_rate": 5.927361329281638e-06,
+      "loss": 0.4848,
+      "step": 6588
+    },
+    {
+      "epoch": 0.4575059019580614,
+      "grad_norm": 3.989650807065453,
+      "learning_rate": 5.926256328943867e-06,
+      "loss": 0.3371,
+      "step": 6589
+    },
+    {
+      "epoch": 0.4575753367587835,
+      "grad_norm": 4.686273767322977,
+      "learning_rate": 5.925151281757099e-06,
+      "loss": 0.6548,
+      "step": 6590
+    },
+    {
+      "epoch": 0.4576447715595056,
+      "grad_norm": 7.052148073713641,
+      "learning_rate": 5.924046187777226e-06,
+      "loss": 0.4253,
+      "step": 6591
+    },
+    {
+      "epoch": 0.45771420636022775,
+      "grad_norm": 4.923526941618985,
+      "learning_rate": 5.922941047060142e-06,
+      "loss": 0.6459,
+      "step": 6592
+    },
+    {
+      "epoch": 0.4577836411609499,
+      "grad_norm": 4.407521206810167,
+      "learning_rate": 5.921835859661746e-06,
+      "loss": 0.6013,
+      "step": 6593
+    },
+    {
+      "epoch": 0.45785307596167196,
+      "grad_norm": 4.113004088596926,
+      "learning_rate": 5.920730625637934e-06,
+      "loss": 0.5125,
+      "step": 6594
+    },
+    {
+      "epoch": 0.4579225107623941,
+      "grad_norm": 2.7146217597963176,
+      "learning_rate": 5.919625345044607e-06,
+      "loss": 0.2922,
+      "step": 6595
+    },
+    {
+      "epoch": 0.45799194556311623,
+      "grad_norm": 3.1315540708269416,
+      "learning_rate": 5.918520017937674e-06,
+      "loss": 0.2947,
+      "step": 6596
+    },
+    {
+      "epoch": 0.45806138036383837,
+      "grad_norm": 4.573486605305715,
+      "learning_rate": 5.9174146443730355e-06,
+      "loss": 0.5861,
+      "step": 6597
+    },
+    {
+      "epoch": 0.4581308151645605,
+      "grad_norm": 3.1044043197639914,
+      "learning_rate": 5.916309224406604e-06,
+      "loss": 0.4049,
+      "step": 6598
+    },
+    {
+      "epoch": 0.4582002499652826,
+      "grad_norm": 3.0094861406810622,
+      "learning_rate": 5.9152037580942875e-06,
+      "loss": 0.2895,
+      "step": 6599
+    },
+    {
+      "epoch": 0.4582696847660047,
+      "grad_norm": 4.028224506489351,
+      "learning_rate": 5.914098245492002e-06,
+      "loss": 0.4288,
+      "step": 6600
+    },
+    {
+      "epoch": 0.45833911956672685,
+      "grad_norm": 5.119873227376004,
+      "learning_rate": 5.912992686655659e-06,
+      "loss": 0.7436,
+      "step": 6601
+    },
+    {
+      "epoch": 0.458408554367449,
+      "grad_norm": 4.066444758992238,
+      "learning_rate": 5.911887081641182e-06,
+      "loss": 0.5358,
+      "step": 6602
+    },
+    {
+      "epoch": 0.45847798916817106,
+      "grad_norm": 3.208447167405556,
+      "learning_rate": 5.910781430504488e-06,
+      "loss": 0.2359,
+      "step": 6603
+    },
+    {
+      "epoch": 0.4585474239688932,
+      "grad_norm": 4.28542725376236,
+      "learning_rate": 5.9096757333014975e-06,
+      "loss": 0.4515,
+      "step": 6604
+    },
+    {
+      "epoch": 0.45861685876961533,
+      "grad_norm": 4.655296430066386,
+      "learning_rate": 5.90856999008814e-06,
+      "loss": 0.578,
+      "step": 6605
+    },
+    {
+      "epoch": 0.45868629357033747,
+      "grad_norm": 4.127636141596106,
+      "learning_rate": 5.90746420092034e-06,
+      "loss": 0.4842,
+      "step": 6606
+    },
+    {
+      "epoch": 0.4587557283710596,
+      "grad_norm": 2.7212380661110913,
+      "learning_rate": 5.906358365854027e-06,
+      "loss": 0.2188,
+      "step": 6607
+    },
+    {
+      "epoch": 0.4588251631717817,
+      "grad_norm": 4.225683300705746,
+      "learning_rate": 5.905252484945133e-06,
+      "loss": 0.4258,
+      "step": 6608
+    },
+    {
+      "epoch": 0.4588945979725038,
+      "grad_norm": 4.179077956454522,
+      "learning_rate": 5.904146558249595e-06,
+      "loss": 0.4166,
+      "step": 6609
+    },
+    {
+      "epoch": 0.45896403277322595,
+      "grad_norm": 4.205205141165408,
+      "learning_rate": 5.903040585823344e-06,
+      "loss": 0.3628,
+      "step": 6610
+    },
+    {
+      "epoch": 0.4590334675739481,
+      "grad_norm": 3.8088258038483924,
+      "learning_rate": 5.901934567722324e-06,
+      "loss": 0.5675,
+      "step": 6611
+    },
+    {
+      "epoch": 0.45910290237467016,
+      "grad_norm": 3.2565375886806764,
+      "learning_rate": 5.9008285040024746e-06,
+      "loss": 0.3089,
+      "step": 6612
+    },
+    {
+      "epoch": 0.4591723371753923,
+      "grad_norm": 3.995322723745124,
+      "learning_rate": 5.899722394719738e-06,
+      "loss": 0.4555,
+      "step": 6613
+    },
+    {
+      "epoch": 0.45924177197611443,
+      "grad_norm": 2.794780689661263,
+      "learning_rate": 5.898616239930061e-06,
+      "loss": 0.2991,
+      "step": 6614
+    },
+    {
+      "epoch": 0.45931120677683657,
+      "grad_norm": 3.475949973943329,
+      "learning_rate": 5.897510039689391e-06,
+      "loss": 0.5611,
+      "step": 6615
+    },
+    {
+      "epoch": 0.45938064157755865,
+      "grad_norm": 5.972298618868445,
+      "learning_rate": 5.896403794053679e-06,
+      "loss": 0.8622,
+      "step": 6616
+    },
+    {
+      "epoch": 0.4594500763782808,
+      "grad_norm": 3.0693407187702637,
+      "learning_rate": 5.8952975030788775e-06,
+      "loss": 0.4123,
+      "step": 6617
+    },
+    {
+      "epoch": 0.4595195111790029,
+      "grad_norm": 4.31122841974351,
+      "learning_rate": 5.894191166820941e-06,
+      "loss": 0.5879,
+      "step": 6618
+    },
+    {
+      "epoch": 0.45958894597972505,
+      "grad_norm": 4.738422659036301,
+      "learning_rate": 5.893084785335829e-06,
+      "loss": 0.3926,
+      "step": 6619
+    },
+    {
+      "epoch": 0.4596583807804472,
+      "grad_norm": 4.1612585082901346,
+      "learning_rate": 5.891978358679497e-06,
+      "loss": 0.5911,
+      "step": 6620
+    },
+    {
+      "epoch": 0.45972781558116926,
+      "grad_norm": 3.7838889007136944,
+      "learning_rate": 5.89087188690791e-06,
+      "loss": 0.5357,
+      "step": 6621
+    },
+    {
+      "epoch": 0.4597972503818914,
+      "grad_norm": 3.5449771796564318,
+      "learning_rate": 5.889765370077031e-06,
+      "loss": 0.3802,
+      "step": 6622
+    },
+    {
+      "epoch": 0.45986668518261353,
+      "grad_norm": 4.432373441933868,
+      "learning_rate": 5.888658808242825e-06,
+      "loss": 0.6374,
+      "step": 6623
+    },
+    {
+      "epoch": 0.45993611998333567,
+      "grad_norm": 4.409254883126733,
+      "learning_rate": 5.887552201461263e-06,
+      "loss": 0.5281,
+      "step": 6624
+    },
+    {
+      "epoch": 0.46000555478405775,
+      "grad_norm": 2.826211397557647,
+      "learning_rate": 5.886445549788317e-06,
+      "loss": 0.3517,
+      "step": 6625
+    },
+    {
+      "epoch": 0.4600749895847799,
+      "grad_norm": 3.729620339797882,
+      "learning_rate": 5.885338853279955e-06,
+      "loss": 0.5066,
+      "step": 6626
+    },
+    {
+      "epoch": 0.460144424385502,
+      "grad_norm": 3.047776291328894,
+      "learning_rate": 5.884232111992156e-06,
+      "loss": 0.3146,
+      "step": 6627
+    },
+    {
+      "epoch": 0.46021385918622415,
+      "grad_norm": 5.75168144838073,
+      "learning_rate": 5.8831253259809005e-06,
+      "loss": 0.8869,
+      "step": 6628
+    },
+    {
+      "epoch": 0.4602832939869463,
+      "grad_norm": 3.561639654290226,
+      "learning_rate": 5.882018495302162e-06,
+      "loss": 0.5262,
+      "step": 6629
+    },
+    {
+      "epoch": 0.46035272878766836,
+      "grad_norm": 2.1712020449143203,
+      "learning_rate": 5.880911620011928e-06,
+      "loss": 0.1913,
+      "step": 6630
+    },
+    {
+      "epoch": 0.4604221635883905,
+      "grad_norm": 2.736258385069976,
+      "learning_rate": 5.879804700166181e-06,
+      "loss": 0.329,
+      "step": 6631
+    },
+    {
+      "epoch": 0.46049159838911263,
+      "grad_norm": 2.8498743726928732,
+      "learning_rate": 5.878697735820906e-06,
+      "loss": 0.245,
+      "step": 6632
+    },
+    {
+      "epoch": 0.46056103318983477,
+      "grad_norm": 3.4948686960130906,
+      "learning_rate": 5.877590727032094e-06,
+      "loss": 0.4492,
+      "step": 6633
+    },
+    {
+      "epoch": 0.46063046799055685,
+      "grad_norm": 4.090872153049156,
+      "learning_rate": 5.876483673855737e-06,
+      "loss": 0.6287,
+      "step": 6634
+    },
+    {
+      "epoch": 0.460699902791279,
+      "grad_norm": 3.4405019259144027,
+      "learning_rate": 5.875376576347828e-06,
+      "loss": 0.5276,
+      "step": 6635
+    },
+    {
+      "epoch": 0.4607693375920011,
+      "grad_norm": 3.5094811197045437,
+      "learning_rate": 5.8742694345643625e-06,
+      "loss": 0.631,
+      "step": 6636
+    },
+    {
+      "epoch": 0.46083877239272325,
+      "grad_norm": 3.766766950755339,
+      "learning_rate": 5.873162248561338e-06,
+      "loss": 0.4954,
+      "step": 6637
+    },
+    {
+      "epoch": 0.46090820719344533,
+      "grad_norm": 3.3403051802824923,
+      "learning_rate": 5.872055018394756e-06,
+      "loss": 0.4203,
+      "step": 6638
+    },
+    {
+      "epoch": 0.46097764199416746,
+      "grad_norm": 3.6911561933127945,
+      "learning_rate": 5.870947744120616e-06,
+      "loss": 0.3949,
+      "step": 6639
+    },
+    {
+      "epoch": 0.4610470767948896,
+      "grad_norm": 3.985577438347011,
+      "learning_rate": 5.869840425794925e-06,
+      "loss": 0.3917,
+      "step": 6640
+    },
+    {
+      "epoch": 0.46111651159561173,
+      "grad_norm": 3.860375539750852,
+      "learning_rate": 5.868733063473691e-06,
+      "loss": 0.4095,
+      "step": 6641
+    },
+    {
+      "epoch": 0.46118594639633387,
+      "grad_norm": 3.6974773248108974,
+      "learning_rate": 5.86762565721292e-06,
+      "loss": 0.4307,
+      "step": 6642
+    },
+    {
+      "epoch": 0.46125538119705595,
+      "grad_norm": 3.2431074952069388,
+      "learning_rate": 5.866518207068626e-06,
+      "loss": 0.3737,
+      "step": 6643
+    },
+    {
+      "epoch": 0.4613248159977781,
+      "grad_norm": 2.7567039492981036,
+      "learning_rate": 5.865410713096821e-06,
+      "loss": 0.2014,
+      "step": 6644
+    },
+    {
+      "epoch": 0.4613942507985002,
+      "grad_norm": 3.8317909858615806,
+      "learning_rate": 5.864303175353522e-06,
+      "loss": 0.5063,
+      "step": 6645
+    },
+    {
+      "epoch": 0.46146368559922235,
+      "grad_norm": 3.5735269688620876,
+      "learning_rate": 5.863195593894746e-06,
+      "loss": 0.4775,
+      "step": 6646
+    },
+    {
+      "epoch": 0.46153312039994443,
+      "grad_norm": 3.8841543513824406,
+      "learning_rate": 5.862087968776514e-06,
+      "loss": 0.3998,
+      "step": 6647
+    },
+    {
+      "epoch": 0.46160255520066656,
+      "grad_norm": 2.7758799757518022,
+      "learning_rate": 5.860980300054848e-06,
+      "loss": 0.2009,
+      "step": 6648
+    },
+    {
+      "epoch": 0.4616719900013887,
+      "grad_norm": 2.832289046012341,
+      "learning_rate": 5.859872587785772e-06,
+      "loss": 0.2976,
+      "step": 6649
+    },
+    {
+      "epoch": 0.46174142480211083,
+      "grad_norm": 3.1672537257080458,
+      "learning_rate": 5.8587648320253145e-06,
+      "loss": 0.4843,
+      "step": 6650
+    },
+    {
+      "epoch": 0.4618108596028329,
+      "grad_norm": 3.4557574481962745,
+      "learning_rate": 5.857657032829503e-06,
+      "loss": 0.3487,
+      "step": 6651
+    },
+    {
+      "epoch": 0.46188029440355505,
+      "grad_norm": 3.4546219135484737,
+      "learning_rate": 5.85654919025437e-06,
+      "loss": 0.3744,
+      "step": 6652
+    },
+    {
+      "epoch": 0.4619497292042772,
+      "grad_norm": 4.428602052283512,
+      "learning_rate": 5.855441304355947e-06,
+      "loss": 0.5319,
+      "step": 6653
+    },
+    {
+      "epoch": 0.4620191640049993,
+      "grad_norm": 4.015501160374193,
+      "learning_rate": 5.854333375190272e-06,
+      "loss": 0.5409,
+      "step": 6654
+    },
+    {
+      "epoch": 0.46208859880572145,
+      "grad_norm": 2.297262350223988,
+      "learning_rate": 5.853225402813381e-06,
+      "loss": 0.1324,
+      "step": 6655
+    },
+    {
+      "epoch": 0.46215803360644353,
+      "grad_norm": 4.041103395628692,
+      "learning_rate": 5.852117387281313e-06,
+      "loss": 0.4155,
+      "step": 6656
+    },
+    {
+      "epoch": 0.46222746840716566,
+      "grad_norm": 4.726834220060106,
+      "learning_rate": 5.851009328650114e-06,
+      "loss": 0.7386,
+      "step": 6657
+    },
+    {
+      "epoch": 0.4622969032078878,
+      "grad_norm": 3.6000841593577366,
+      "learning_rate": 5.849901226975826e-06,
+      "loss": 0.3482,
+      "step": 6658
+    },
+    {
+      "epoch": 0.46236633800860993,
+      "grad_norm": 4.207718458711362,
+      "learning_rate": 5.848793082314496e-06,
+      "loss": 0.4331,
+      "step": 6659
+    },
+    {
+      "epoch": 0.462435772809332,
+      "grad_norm": 4.143866872790647,
+      "learning_rate": 5.8476848947221705e-06,
+      "loss": 0.4656,
+      "step": 6660
+    },
+    {
+      "epoch": 0.46250520761005415,
+      "grad_norm": 3.7324131604448945,
+      "learning_rate": 5.846576664254904e-06,
+      "loss": 0.4365,
+      "step": 6661
+    },
+    {
+      "epoch": 0.4625746424107763,
+      "grad_norm": 3.9620947477267396,
+      "learning_rate": 5.8454683909687474e-06,
+      "loss": 0.4761,
+      "step": 6662
+    },
+    {
+      "epoch": 0.4626440772114984,
+      "grad_norm": 3.477942988100187,
+      "learning_rate": 5.844360074919756e-06,
+      "loss": 0.3271,
+      "step": 6663
+    },
+    {
+      "epoch": 0.46271351201222055,
+      "grad_norm": 3.9479927716406507,
+      "learning_rate": 5.8432517161639875e-06,
+      "loss": 0.4262,
+      "step": 6664
+    },
+    {
+      "epoch": 0.46278294681294263,
+      "grad_norm": 4.062232175232798,
+      "learning_rate": 5.842143314757501e-06,
+      "loss": 0.4303,
+      "step": 6665
+    },
+    {
+      "epoch": 0.46285238161366477,
+      "grad_norm": 3.4832001985872107,
+      "learning_rate": 5.841034870756359e-06,
+      "loss": 0.3871,
+      "step": 6666
+    },
+    {
+      "epoch": 0.4629218164143869,
+      "grad_norm": 3.3878096761522176,
+      "learning_rate": 5.839926384216625e-06,
+      "loss": 0.4959,
+      "step": 6667
+    },
+    {
+      "epoch": 0.46299125121510903,
+      "grad_norm": 3.492659507729984,
+      "learning_rate": 5.838817855194365e-06,
+      "loss": 0.4153,
+      "step": 6668
+    },
+    {
+      "epoch": 0.4630606860158311,
+      "grad_norm": 3.9940103368909803,
+      "learning_rate": 5.837709283745647e-06,
+      "loss": 0.6313,
+      "step": 6669
+    },
+    {
+      "epoch": 0.46313012081655325,
+      "grad_norm": 2.6808185278652563,
+      "learning_rate": 5.836600669926542e-06,
+      "loss": 0.2365,
+      "step": 6670
+    },
+    {
+      "epoch": 0.4631995556172754,
+      "grad_norm": 2.7508931180559597,
+      "learning_rate": 5.835492013793122e-06,
+      "loss": 0.2514,
+      "step": 6671
+    },
+    {
+      "epoch": 0.4632689904179975,
+      "grad_norm": 3.685683231221985,
+      "learning_rate": 5.834383315401458e-06,
+      "loss": 0.4754,
+      "step": 6672
+    },
+    {
+      "epoch": 0.4633384252187196,
+      "grad_norm": 3.417746281882585,
+      "learning_rate": 5.833274574807635e-06,
+      "loss": 0.4691,
+      "step": 6673
+    },
+    {
+      "epoch": 0.46340786001944173,
+      "grad_norm": 4.33086735426182,
+      "learning_rate": 5.832165792067723e-06,
+      "loss": 0.6191,
+      "step": 6674
+    },
+    {
+      "epoch": 0.46347729482016387,
+      "grad_norm": 3.2378244575517536,
+      "learning_rate": 5.831056967237808e-06,
+      "loss": 0.4363,
+      "step": 6675
+    },
+    {
+      "epoch": 0.463546729620886,
+      "grad_norm": 3.695386462880349,
+      "learning_rate": 5.829948100373973e-06,
+      "loss": 0.3119,
+      "step": 6676
+    },
+    {
+      "epoch": 0.46361616442160813,
+      "grad_norm": 3.4320593305452602,
+      "learning_rate": 5.828839191532301e-06,
+      "loss": 0.422,
+      "step": 6677
+    },
+    {
+      "epoch": 0.4636855992223302,
+      "grad_norm": 3.3667652941611483,
+      "learning_rate": 5.827730240768882e-06,
+      "loss": 0.4156,
+      "step": 6678
+    },
+    {
+      "epoch": 0.46375503402305235,
+      "grad_norm": 3.4692293085957506,
+      "learning_rate": 5.826621248139802e-06,
+      "loss": 0.321,
+      "step": 6679
+    },
+    {
+      "epoch": 0.4638244688237745,
+      "grad_norm": 3.3936836447616847,
+      "learning_rate": 5.8255122137011575e-06,
+      "loss": 0.352,
+      "step": 6680
+    },
+    {
+      "epoch": 0.4638939036244966,
+      "grad_norm": 4.925315055984609,
+      "learning_rate": 5.824403137509037e-06,
+      "loss": 0.5505,
+      "step": 6681
+    },
+    {
+      "epoch": 0.4639633384252187,
+      "grad_norm": 4.537521705974809,
+      "learning_rate": 5.82329401961954e-06,
+      "loss": 0.565,
+      "step": 6682
+    },
+    {
+      "epoch": 0.46403277322594083,
+      "grad_norm": 4.119392642260565,
+      "learning_rate": 5.822184860088762e-06,
+      "loss": 0.6114,
+      "step": 6683
+    },
+    {
+      "epoch": 0.46410220802666297,
+      "grad_norm": 3.9129910628138074,
+      "learning_rate": 5.821075658972806e-06,
+      "loss": 0.5601,
+      "step": 6684
+    },
+    {
+      "epoch": 0.4641716428273851,
+      "grad_norm": 4.607533160977391,
+      "learning_rate": 5.81996641632777e-06,
+      "loss": 0.612,
+      "step": 6685
+    },
+    {
+      "epoch": 0.46424107762810723,
+      "grad_norm": 3.749344479185912,
+      "learning_rate": 5.818857132209762e-06,
+      "loss": 0.3412,
+      "step": 6686
+    },
+    {
+      "epoch": 0.4643105124288293,
+      "grad_norm": 3.570205156664813,
+      "learning_rate": 5.817747806674886e-06,
+      "loss": 0.2557,
+      "step": 6687
+    },
+    {
+      "epoch": 0.46437994722955145,
+      "grad_norm": 4.72649116886234,
+      "learning_rate": 5.8166384397792505e-06,
+      "loss": 0.9041,
+      "step": 6688
+    },
+    {
+      "epoch": 0.4644493820302736,
+      "grad_norm": 4.679178087118482,
+      "learning_rate": 5.815529031578968e-06,
+      "loss": 0.5121,
+      "step": 6689
+    },
+    {
+      "epoch": 0.4645188168309957,
+      "grad_norm": 4.70697121397895,
+      "learning_rate": 5.814419582130151e-06,
+      "loss": 0.6076,
+      "step": 6690
+    },
+    {
+      "epoch": 0.4645882516317178,
+      "grad_norm": 3.6296495806519657,
+      "learning_rate": 5.813310091488912e-06,
+      "loss": 0.5333,
+      "step": 6691
+    },
+    {
+      "epoch": 0.46465768643243993,
+      "grad_norm": 4.4130051528915475,
+      "learning_rate": 5.812200559711369e-06,
+      "loss": 0.5203,
+      "step": 6692
+    },
+    {
+      "epoch": 0.46472712123316207,
+      "grad_norm": 3.7941963268806247,
+      "learning_rate": 5.81109098685364e-06,
+      "loss": 0.4889,
+      "step": 6693
+    },
+    {
+      "epoch": 0.4647965560338842,
+      "grad_norm": 3.6223899791016367,
+      "learning_rate": 5.8099813729718475e-06,
+      "loss": 0.5496,
+      "step": 6694
+    },
+    {
+      "epoch": 0.4648659908346063,
+      "grad_norm": 3.1419047673968565,
+      "learning_rate": 5.808871718122113e-06,
+      "loss": 0.3057,
+      "step": 6695
+    },
+    {
+      "epoch": 0.4649354256353284,
+      "grad_norm": 3.8295425330406547,
+      "learning_rate": 5.807762022360564e-06,
+      "loss": 0.3902,
+      "step": 6696
+    },
+    {
+      "epoch": 0.46500486043605055,
+      "grad_norm": 4.182559159749013,
+      "learning_rate": 5.806652285743324e-06,
+      "loss": 0.4894,
+      "step": 6697
+    },
+    {
+      "epoch": 0.4650742952367727,
+      "grad_norm": 4.076081553611543,
+      "learning_rate": 5.805542508326525e-06,
+      "loss": 0.6503,
+      "step": 6698
+    },
+    {
+      "epoch": 0.4651437300374948,
+      "grad_norm": 2.9983328607869764,
+      "learning_rate": 5.804432690166298e-06,
+      "loss": 0.3046,
+      "step": 6699
+    },
+    {
+      "epoch": 0.4652131648382169,
+      "grad_norm": 5.488522566022272,
+      "learning_rate": 5.803322831318775e-06,
+      "loss": 0.7442,
+      "step": 6700
+    },
+    {
+      "epoch": 0.46528259963893903,
+      "grad_norm": 5.179497085190627,
+      "learning_rate": 5.802212931840092e-06,
+      "loss": 0.7672,
+      "step": 6701
+    },
+    {
+      "epoch": 0.46535203443966117,
+      "grad_norm": 2.7521223726909962,
+      "learning_rate": 5.8011029917863885e-06,
+      "loss": 0.2702,
+      "step": 6702
+    },
+    {
+      "epoch": 0.4654214692403833,
+      "grad_norm": 4.165229888218057,
+      "learning_rate": 5.7999930112138e-06,
+      "loss": 0.692,
+      "step": 6703
+    },
+    {
+      "epoch": 0.4654909040411054,
+      "grad_norm": 3.8098471065517643,
+      "learning_rate": 5.798882990178471e-06,
+      "loss": 0.3947,
+      "step": 6704
+    },
+    {
+      "epoch": 0.4655603388418275,
+      "grad_norm": 3.64607339157889,
+      "learning_rate": 5.797772928736543e-06,
+      "loss": 0.4162,
+      "step": 6705
+    },
+    {
+      "epoch": 0.46562977364254965,
+      "grad_norm": 4.178100418160067,
+      "learning_rate": 5.796662826944166e-06,
+      "loss": 0.7043,
+      "step": 6706
+    },
+    {
+      "epoch": 0.4656992084432718,
+      "grad_norm": 3.6683335120860754,
+      "learning_rate": 5.795552684857483e-06,
+      "loss": 0.4398,
+      "step": 6707
+    },
+    {
+      "epoch": 0.46576864324399386,
+      "grad_norm": 3.78707414379052,
+      "learning_rate": 5.794442502532646e-06,
+      "loss": 0.3695,
+      "step": 6708
+    },
+    {
+      "epoch": 0.465838078044716,
+      "grad_norm": 2.788736482323854,
+      "learning_rate": 5.793332280025805e-06,
+      "loss": 0.3224,
+      "step": 6709
+    },
+    {
+      "epoch": 0.46590751284543813,
+      "grad_norm": 3.8239043489708493,
+      "learning_rate": 5.792222017393116e-06,
+      "loss": 0.5014,
+      "step": 6710
+    },
+    {
+      "epoch": 0.46597694764616027,
+      "grad_norm": 3.811727094077433,
+      "learning_rate": 5.791111714690733e-06,
+      "loss": 0.5067,
+      "step": 6711
+    },
+    {
+      "epoch": 0.4660463824468824,
+      "grad_norm": 3.69668910315829,
+      "learning_rate": 5.790001371974816e-06,
+      "loss": 0.5259,
+      "step": 6712
+    },
+    {
+      "epoch": 0.4661158172476045,
+      "grad_norm": 3.069166811558484,
+      "learning_rate": 5.7888909893015225e-06,
+      "loss": 0.3392,
+      "step": 6713
+    },
+    {
+      "epoch": 0.4661852520483266,
+      "grad_norm": 2.5221099442599595,
+      "learning_rate": 5.787780566727016e-06,
+      "loss": 0.208,
+      "step": 6714
+    },
+    {
+      "epoch": 0.46625468684904875,
+      "grad_norm": 3.4365418730793795,
+      "learning_rate": 5.786670104307459e-06,
+      "loss": 0.2855,
+      "step": 6715
+    },
+    {
+      "epoch": 0.4663241216497709,
+      "grad_norm": 3.3902813035115065,
+      "learning_rate": 5.785559602099019e-06,
+      "loss": 0.3721,
+      "step": 6716
+    },
+    {
+      "epoch": 0.46639355645049296,
+      "grad_norm": 3.0301535449471046,
+      "learning_rate": 5.784449060157863e-06,
+      "loss": 0.2714,
+      "step": 6717
+    },
+    {
+      "epoch": 0.4664629912512151,
+      "grad_norm": 4.9984355552320485,
+      "learning_rate": 5.7833384785401624e-06,
+      "loss": 0.5896,
+      "step": 6718
+    },
+    {
+      "epoch": 0.46653242605193723,
+      "grad_norm": 4.142349935239867,
+      "learning_rate": 5.782227857302086e-06,
+      "loss": 0.4898,
+      "step": 6719
+    },
+    {
+      "epoch": 0.46660186085265937,
+      "grad_norm": 4.047557623426405,
+      "learning_rate": 5.78111719649981e-06,
+      "loss": 0.5157,
+      "step": 6720
+    },
+    {
+      "epoch": 0.4666712956533815,
+      "grad_norm": 3.616305102156895,
+      "learning_rate": 5.7800064961895105e-06,
+      "loss": 0.5131,
+      "step": 6721
+    },
+    {
+      "epoch": 0.4667407304541036,
+      "grad_norm": 3.752536341097849,
+      "learning_rate": 5.778895756427366e-06,
+      "loss": 0.4905,
+      "step": 6722
+    },
+    {
+      "epoch": 0.4668101652548257,
+      "grad_norm": 3.8501381635620637,
+      "learning_rate": 5.777784977269555e-06,
+      "loss": 0.517,
+      "step": 6723
+    },
+    {
+      "epoch": 0.46687960005554785,
+      "grad_norm": 4.882629815374773,
+      "learning_rate": 5.776674158772258e-06,
+      "loss": 0.5395,
+      "step": 6724
+    },
+    {
+      "epoch": 0.46694903485627,
+      "grad_norm": 3.4776902371729954,
+      "learning_rate": 5.775563300991664e-06,
+      "loss": 0.3463,
+      "step": 6725
+    },
+    {
+      "epoch": 0.46701846965699206,
+      "grad_norm": 3.202491584594151,
+      "learning_rate": 5.774452403983955e-06,
+      "loss": 0.4431,
+      "step": 6726
+    },
+    {
+      "epoch": 0.4670879044577142,
+      "grad_norm": 4.622260807498098,
+      "learning_rate": 5.773341467805319e-06,
+      "loss": 0.7356,
+      "step": 6727
+    },
+    {
+      "epoch": 0.46715733925843633,
+      "grad_norm": 3.9173617656489603,
+      "learning_rate": 5.772230492511948e-06,
+      "loss": 0.6255,
+      "step": 6728
+    },
+    {
+      "epoch": 0.46722677405915847,
+      "grad_norm": 4.476440996838808,
+      "learning_rate": 5.771119478160031e-06,
+      "loss": 0.6822,
+      "step": 6729
+    },
+    {
+      "epoch": 0.46729620885988055,
+      "grad_norm": 3.7597108114218156,
+      "learning_rate": 5.7700084248057645e-06,
+      "loss": 0.4639,
+      "step": 6730
+    },
+    {
+      "epoch": 0.4673656436606027,
+      "grad_norm": 4.106728518521276,
+      "learning_rate": 5.768897332505342e-06,
+      "loss": 0.4522,
+      "step": 6731
+    },
+    {
+      "epoch": 0.4674350784613248,
+      "grad_norm": 4.394082044207434,
+      "learning_rate": 5.767786201314964e-06,
+      "loss": 0.4963,
+      "step": 6732
+    },
+    {
+      "epoch": 0.46750451326204695,
+      "grad_norm": 3.4248351268552844,
+      "learning_rate": 5.766675031290826e-06,
+      "loss": 0.3746,
+      "step": 6733
+    },
+    {
+      "epoch": 0.4675739480627691,
+      "grad_norm": 3.6680265088772726,
+      "learning_rate": 5.765563822489134e-06,
+      "loss": 0.4904,
+      "step": 6734
+    },
+    {
+      "epoch": 0.46764338286349116,
+      "grad_norm": 3.824219600192621,
+      "learning_rate": 5.764452574966092e-06,
+      "loss": 0.6171,
+      "step": 6735
+    },
+    {
+      "epoch": 0.4677128176642133,
+      "grad_norm": 3.6950687864063645,
+      "learning_rate": 5.763341288777902e-06,
+      "loss": 0.3641,
+      "step": 6736
+    },
+    {
+      "epoch": 0.46778225246493543,
+      "grad_norm": 5.007832218064642,
+      "learning_rate": 5.7622299639807735e-06,
+      "loss": 0.2299,
+      "step": 6737
+    },
+    {
+      "epoch": 0.46785168726565757,
+      "grad_norm": 4.248901369538338,
+      "learning_rate": 5.761118600630915e-06,
+      "loss": 0.4186,
+      "step": 6738
+    },
+    {
+      "epoch": 0.46792112206637965,
+      "grad_norm": 3.6113643843421,
+      "learning_rate": 5.7600071987845405e-06,
+      "loss": 0.5514,
+      "step": 6739
+    },
+    {
+      "epoch": 0.4679905568671018,
+      "grad_norm": 4.005009135098748,
+      "learning_rate": 5.7588957584978615e-06,
+      "loss": 0.3797,
+      "step": 6740
+    },
+    {
+      "epoch": 0.4680599916678239,
+      "grad_norm": 3.3860626758666412,
+      "learning_rate": 5.757784279827094e-06,
+      "loss": 0.4078,
+      "step": 6741
+    },
+    {
+      "epoch": 0.46812942646854605,
+      "grad_norm": 2.8505917984201203,
+      "learning_rate": 5.756672762828454e-06,
+      "loss": 0.3937,
+      "step": 6742
+    },
+    {
+      "epoch": 0.46819886126926813,
+      "grad_norm": 3.4658380243430456,
+      "learning_rate": 5.7555612075581625e-06,
+      "loss": 0.4083,
+      "step": 6743
+    },
+    {
+      "epoch": 0.46826829606999026,
+      "grad_norm": 3.424347254084609,
+      "learning_rate": 5.75444961407244e-06,
+      "loss": 0.4948,
+      "step": 6744
+    },
+    {
+      "epoch": 0.4683377308707124,
+      "grad_norm": 3.2542726910236643,
+      "learning_rate": 5.753337982427511e-06,
+      "loss": 0.318,
+      "step": 6745
+    },
+    {
+      "epoch": 0.46840716567143453,
+      "grad_norm": 3.5009869627425534,
+      "learning_rate": 5.7522263126795986e-06,
+      "loss": 0.4804,
+      "step": 6746
+    },
+    {
+      "epoch": 0.46847660047215667,
+      "grad_norm": 5.005366740932165,
+      "learning_rate": 5.75111460488493e-06,
+      "loss": 0.5303,
+      "step": 6747
+    },
+    {
+      "epoch": 0.46854603527287875,
+      "grad_norm": 3.4413564729407455,
+      "learning_rate": 5.750002859099735e-06,
+      "loss": 0.3209,
+      "step": 6748
+    },
+    {
+      "epoch": 0.4686154700736009,
+      "grad_norm": 3.769728505265845,
+      "learning_rate": 5.748891075380243e-06,
+      "loss": 0.4922,
+      "step": 6749
+    },
+    {
+      "epoch": 0.468684904874323,
+      "grad_norm": 4.897230734198281,
+      "learning_rate": 5.747779253782688e-06,
+      "loss": 0.4281,
+      "step": 6750
+    },
+    {
+      "epoch": 0.46875433967504515,
+      "grad_norm": 4.0411639440870015,
+      "learning_rate": 5.746667394363308e-06,
+      "loss": 0.5228,
+      "step": 6751
+    },
+    {
+      "epoch": 0.46882377447576723,
+      "grad_norm": 4.407968313966592,
+      "learning_rate": 5.745555497178332e-06,
+      "loss": 0.8793,
+      "step": 6752
+    },
+    {
+      "epoch": 0.46889320927648936,
+      "grad_norm": 4.110558483720758,
+      "learning_rate": 5.744443562284003e-06,
+      "loss": 0.5292,
+      "step": 6753
+    },
+    {
+      "epoch": 0.4689626440772115,
+      "grad_norm": 4.751684836318537,
+      "learning_rate": 5.743331589736563e-06,
+      "loss": 0.7217,
+      "step": 6754
+    },
+    {
+      "epoch": 0.46903207887793363,
+      "grad_norm": 2.948645157842378,
+      "learning_rate": 5.742219579592252e-06,
+      "loss": 0.4631,
+      "step": 6755
+    },
+    {
+      "epoch": 0.46910151367865577,
+      "grad_norm": 4.1488207129295125,
+      "learning_rate": 5.741107531907313e-06,
+      "loss": 0.545,
+      "step": 6756
+    },
+    {
+      "epoch": 0.46917094847937785,
+      "grad_norm": 2.254468199220277,
+      "learning_rate": 5.739995446737995e-06,
+      "loss": 0.1934,
+      "step": 6757
+    },
+    {
+      "epoch": 0.4692403832801,
+      "grad_norm": 3.0475527326654333,
+      "learning_rate": 5.7388833241405425e-06,
+      "loss": 0.3873,
+      "step": 6758
+    },
+    {
+      "epoch": 0.4693098180808221,
+      "grad_norm": 2.948671689881155,
+      "learning_rate": 5.737771164171211e-06,
+      "loss": 0.2264,
+      "step": 6759
+    },
+    {
+      "epoch": 0.46937925288154425,
+      "grad_norm": 3.507086333045732,
+      "learning_rate": 5.736658966886246e-06,
+      "loss": 0.3489,
+      "step": 6760
+    },
+    {
+      "epoch": 0.46944868768226633,
+      "grad_norm": 4.008319428035398,
+      "learning_rate": 5.735546732341906e-06,
+      "loss": 0.6018,
+      "step": 6761
+    },
+    {
+      "epoch": 0.46951812248298846,
+      "grad_norm": 4.046157183997936,
+      "learning_rate": 5.734434460594443e-06,
+      "loss": 0.4384,
+      "step": 6762
+    },
+    {
+      "epoch": 0.4695875572837106,
+      "grad_norm": 5.363941253718464,
+      "learning_rate": 5.733322151700118e-06,
+      "loss": 0.4157,
+      "step": 6763
+    },
+    {
+      "epoch": 0.46965699208443273,
+      "grad_norm": 2.636461927496305,
+      "learning_rate": 5.732209805715186e-06,
+      "loss": 0.2345,
+      "step": 6764
+    },
+    {
+      "epoch": 0.4697264268851548,
+      "grad_norm": 4.17068679258831,
+      "learning_rate": 5.7310974226959115e-06,
+      "loss": 0.5451,
+      "step": 6765
+    },
+    {
+      "epoch": 0.46979586168587695,
+      "grad_norm": 3.6115089350808605,
+      "learning_rate": 5.729985002698557e-06,
+      "loss": 0.41,
+      "step": 6766
+    },
+    {
+      "epoch": 0.4698652964865991,
+      "grad_norm": 3.6762860746970683,
+      "learning_rate": 5.728872545779387e-06,
+      "loss": 0.4221,
+      "step": 6767
+    },
+    {
+      "epoch": 0.4699347312873212,
+      "grad_norm": 2.9521455638614755,
+      "learning_rate": 5.727760051994669e-06,
+      "loss": 0.3671,
+      "step": 6768
+    },
+    {
+      "epoch": 0.47000416608804335,
+      "grad_norm": 3.6669555104162908,
+      "learning_rate": 5.726647521400671e-06,
+      "loss": 0.3742,
+      "step": 6769
+    },
+    {
+      "epoch": 0.47007360088876543,
+      "grad_norm": 3.678585168560078,
+      "learning_rate": 5.725534954053663e-06,
+      "loss": 0.5591,
+      "step": 6770
+    },
+    {
+      "epoch": 0.47014303568948757,
+      "grad_norm": 3.696206814664136,
+      "learning_rate": 5.7244223500099185e-06,
+      "loss": 0.4748,
+      "step": 6771
+    },
+    {
+      "epoch": 0.4702124704902097,
+      "grad_norm": 3.287582383772018,
+      "learning_rate": 5.72330970932571e-06,
+      "loss": 0.406,
+      "step": 6772
+    },
+    {
+      "epoch": 0.47028190529093183,
+      "grad_norm": 5.912194450755426,
+      "learning_rate": 5.722197032057317e-06,
+      "loss": 0.6034,
+      "step": 6773
+    },
+    {
+      "epoch": 0.4703513400916539,
+      "grad_norm": 3.9379744916679433,
+      "learning_rate": 5.7210843182610134e-06,
+      "loss": 0.5249,
+      "step": 6774
+    },
+    {
+      "epoch": 0.47042077489237605,
+      "grad_norm": 3.236000134104256,
+      "learning_rate": 5.719971567993081e-06,
+      "loss": 0.2431,
+      "step": 6775
+    },
+    {
+      "epoch": 0.4704902096930982,
+      "grad_norm": 4.312539027789056,
+      "learning_rate": 5.718858781309803e-06,
+      "loss": 0.4646,
+      "step": 6776
+    },
+    {
+      "epoch": 0.4705596444938203,
+      "grad_norm": 3.2141708309090546,
+      "learning_rate": 5.7177459582674595e-06,
+      "loss": 0.3563,
+      "step": 6777
+    },
+    {
+      "epoch": 0.47062907929454245,
+      "grad_norm": 4.147081450269217,
+      "learning_rate": 5.716633098922339e-06,
+      "loss": 0.5529,
+      "step": 6778
+    },
+    {
+      "epoch": 0.47069851409526453,
+      "grad_norm": 1.9551920820933877,
+      "learning_rate": 5.715520203330727e-06,
+      "loss": 0.1531,
+      "step": 6779
+    },
+    {
+      "epoch": 0.47076794889598667,
+      "grad_norm": 4.009777669391415,
+      "learning_rate": 5.714407271548913e-06,
+      "loss": 0.5255,
+      "step": 6780
+    },
+    {
+      "epoch": 0.4708373836967088,
+      "grad_norm": 3.035458823122768,
+      "learning_rate": 5.713294303633185e-06,
+      "loss": 0.2727,
+      "step": 6781
+    },
+    {
+      "epoch": 0.47090681849743093,
+      "grad_norm": 4.3646039928532305,
+      "learning_rate": 5.712181299639842e-06,
+      "loss": 0.6682,
+      "step": 6782
+    },
+    {
+      "epoch": 0.470976253298153,
+      "grad_norm": 3.8693201715519807,
+      "learning_rate": 5.711068259625173e-06,
+      "loss": 0.5833,
+      "step": 6783
+    },
+    {
+      "epoch": 0.47104568809887515,
+      "grad_norm": 3.3023294017727522,
+      "learning_rate": 5.709955183645477e-06,
+      "loss": 0.3254,
+      "step": 6784
+    },
+    {
+      "epoch": 0.4711151228995973,
+      "grad_norm": 4.282266506793972,
+      "learning_rate": 5.70884207175705e-06,
+      "loss": 0.4258,
+      "step": 6785
+    },
+    {
+      "epoch": 0.4711845577003194,
+      "grad_norm": 3.6253810184418565,
+      "learning_rate": 5.7077289240161935e-06,
+      "loss": 0.3366,
+      "step": 6786
+    },
+    {
+      "epoch": 0.4712539925010415,
+      "grad_norm": 3.635647098225617,
+      "learning_rate": 5.7066157404792085e-06,
+      "loss": 0.3975,
+      "step": 6787
+    },
+    {
+      "epoch": 0.47132342730176363,
+      "grad_norm": 4.3401213376896814,
+      "learning_rate": 5.705502521202399e-06,
+      "loss": 0.6986,
+      "step": 6788
+    },
+    {
+      "epoch": 0.47139286210248577,
+      "grad_norm": 3.757513211767109,
+      "learning_rate": 5.704389266242072e-06,
+      "loss": 0.4602,
+      "step": 6789
+    },
+    {
+      "epoch": 0.4714622969032079,
+      "grad_norm": 3.8871209182798205,
+      "learning_rate": 5.703275975654531e-06,
+      "loss": 0.583,
+      "step": 6790
+    },
+    {
+      "epoch": 0.47153173170393003,
+      "grad_norm": 4.058279494308542,
+      "learning_rate": 5.702162649496088e-06,
+      "loss": 0.6407,
+      "step": 6791
+    },
+    {
+      "epoch": 0.4716011665046521,
+      "grad_norm": 3.8852291641208136,
+      "learning_rate": 5.701049287823052e-06,
+      "loss": 0.2628,
+      "step": 6792
+    },
+    {
+      "epoch": 0.47167060130537425,
+      "grad_norm": 4.190713855430629,
+      "learning_rate": 5.699935890691737e-06,
+      "loss": 0.6037,
+      "step": 6793
+    },
+    {
+      "epoch": 0.4717400361060964,
+      "grad_norm": 5.154799364062615,
+      "learning_rate": 5.6988224581584565e-06,
+      "loss": 0.631,
+      "step": 6794
+    },
+    {
+      "epoch": 0.4718094709068185,
+      "grad_norm": 4.245105837555174,
+      "learning_rate": 5.697708990279525e-06,
+      "loss": 0.6525,
+      "step": 6795
+    },
+    {
+      "epoch": 0.4718789057075406,
+      "grad_norm": 3.7951904360769957,
+      "learning_rate": 5.696595487111266e-06,
+      "loss": 0.6016,
+      "step": 6796
+    },
+    {
+      "epoch": 0.47194834050826273,
+      "grad_norm": 3.6192700235332254,
+      "learning_rate": 5.6954819487099924e-06,
+      "loss": 0.5459,
+      "step": 6797
+    },
+    {
+      "epoch": 0.47201777530898487,
+      "grad_norm": 3.392157843751449,
+      "learning_rate": 5.69436837513203e-06,
+      "loss": 0.3827,
+      "step": 6798
+    },
+    {
+      "epoch": 0.472087210109707,
+      "grad_norm": 3.044128786177679,
+      "learning_rate": 5.693254766433701e-06,
+      "loss": 0.3309,
+      "step": 6799
+    },
+    {
+      "epoch": 0.4721566449104291,
+      "grad_norm": 4.477221368726688,
+      "learning_rate": 5.692141122671331e-06,
+      "loss": 0.4217,
+      "step": 6800
+    },
+    {
+      "epoch": 0.4722260797111512,
+      "grad_norm": 3.4798803931166598,
+      "learning_rate": 5.691027443901245e-06,
+      "loss": 0.3819,
+      "step": 6801
+    },
+    {
+      "epoch": 0.47229551451187335,
+      "grad_norm": 4.5430920123767,
+      "learning_rate": 5.689913730179773e-06,
+      "loss": 0.7103,
+      "step": 6802
+    },
+    {
+      "epoch": 0.4723649493125955,
+      "grad_norm": 4.530463191157134,
+      "learning_rate": 5.688799981563246e-06,
+      "loss": 0.3467,
+      "step": 6803
+    },
+    {
+      "epoch": 0.4724343841133176,
+      "grad_norm": 4.089373455616298,
+      "learning_rate": 5.687686198107993e-06,
+      "loss": 0.5019,
+      "step": 6804
+    },
+    {
+      "epoch": 0.4725038189140397,
+      "grad_norm": 5.819093724568025,
+      "learning_rate": 5.686572379870353e-06,
+      "loss": 0.5005,
+      "step": 6805
+    },
+    {
+      "epoch": 0.47257325371476183,
+      "grad_norm": 3.6884016709420235,
+      "learning_rate": 5.685458526906659e-06,
+      "loss": 0.5948,
+      "step": 6806
+    },
+    {
+      "epoch": 0.47264268851548397,
+      "grad_norm": 4.31751103056666,
+      "learning_rate": 5.6843446392732475e-06,
+      "loss": 0.5589,
+      "step": 6807
+    },
+    {
+      "epoch": 0.4727121233162061,
+      "grad_norm": 2.2621620334272534,
+      "learning_rate": 5.6832307170264575e-06,
+      "loss": 0.2189,
+      "step": 6808
+    },
+    {
+      "epoch": 0.4727815581169282,
+      "grad_norm": 3.673482004790837,
+      "learning_rate": 5.6821167602226325e-06,
+      "loss": 0.3734,
+      "step": 6809
+    },
+    {
+      "epoch": 0.4728509929176503,
+      "grad_norm": 2.9728019538039954,
+      "learning_rate": 5.681002768918114e-06,
+      "loss": 0.2858,
+      "step": 6810
+    },
+    {
+      "epoch": 0.47292042771837245,
+      "grad_norm": 3.2473891024229116,
+      "learning_rate": 5.679888743169243e-06,
+      "loss": 0.4056,
+      "step": 6811
+    },
+    {
+      "epoch": 0.4729898625190946,
+      "grad_norm": 4.000645328095766,
+      "learning_rate": 5.678774683032372e-06,
+      "loss": 0.7438,
+      "step": 6812
+    },
+    {
+      "epoch": 0.4730592973198167,
+      "grad_norm": 4.006431534255663,
+      "learning_rate": 5.677660588563843e-06,
+      "loss": 0.5656,
+      "step": 6813
+    },
+    {
+      "epoch": 0.4731287321205388,
+      "grad_norm": 5.849799306472783,
+      "learning_rate": 5.676546459820011e-06,
+      "loss": 0.5468,
+      "step": 6814
+    },
+    {
+      "epoch": 0.47319816692126093,
+      "grad_norm": 4.527294094890077,
+      "learning_rate": 5.675432296857223e-06,
+      "loss": 0.4095,
+      "step": 6815
+    },
+    {
+      "epoch": 0.47326760172198307,
+      "grad_norm": 3.7941751354602267,
+      "learning_rate": 5.6743180997318345e-06,
+      "loss": 0.5521,
+      "step": 6816
+    },
+    {
+      "epoch": 0.4733370365227052,
+      "grad_norm": 3.982629729252017,
+      "learning_rate": 5.673203868500198e-06,
+      "loss": 0.6308,
+      "step": 6817
+    },
+    {
+      "epoch": 0.4734064713234273,
+      "grad_norm": 3.9194971217227548,
+      "learning_rate": 5.672089603218673e-06,
+      "loss": 0.4781,
+      "step": 6818
+    },
+    {
+      "epoch": 0.4734759061241494,
+      "grad_norm": 2.8979209591414845,
+      "learning_rate": 5.670975303943617e-06,
+      "loss": 0.4298,
+      "step": 6819
+    },
+    {
+      "epoch": 0.47354534092487155,
+      "grad_norm": 3.6549389968316777,
+      "learning_rate": 5.669860970731388e-06,
+      "loss": 0.4107,
+      "step": 6820
+    },
+    {
+      "epoch": 0.4736147757255937,
+      "grad_norm": 4.495714563876899,
+      "learning_rate": 5.668746603638349e-06,
+      "loss": 0.5784,
+      "step": 6821
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 3.9551237790749463,
+      "learning_rate": 5.667632202720864e-06,
+      "loss": 0.4847,
+      "step": 6822
+    },
+    {
+      "epoch": 0.4737536453270379,
+      "grad_norm": 4.229460570977114,
+      "learning_rate": 5.666517768035299e-06,
+      "loss": 0.4761,
+      "step": 6823
+    },
+    {
+      "epoch": 0.47382308012776003,
+      "grad_norm": 4.086441366644287,
+      "learning_rate": 5.665403299638017e-06,
+      "loss": 0.45,
+      "step": 6824
+    },
+    {
+      "epoch": 0.47389251492848217,
+      "grad_norm": 3.230781295872515,
+      "learning_rate": 5.664288797585391e-06,
+      "loss": 0.3526,
+      "step": 6825
+    },
+    {
+      "epoch": 0.4739619497292043,
+      "grad_norm": 3.413271420975861,
+      "learning_rate": 5.663174261933789e-06,
+      "loss": 0.449,
+      "step": 6826
+    },
+    {
+      "epoch": 0.4740313845299264,
+      "grad_norm": 3.2315376008947574,
+      "learning_rate": 5.662059692739582e-06,
+      "loss": 0.4001,
+      "step": 6827
+    },
+    {
+      "epoch": 0.4741008193306485,
+      "grad_norm": 3.4688587117706517,
+      "learning_rate": 5.660945090059147e-06,
+      "loss": 0.3994,
+      "step": 6828
+    },
+    {
+      "epoch": 0.47417025413137065,
+      "grad_norm": 3.9483486991960137,
+      "learning_rate": 5.6598304539488555e-06,
+      "loss": 0.5032,
+      "step": 6829
+    },
+    {
+      "epoch": 0.4742396889320928,
+      "grad_norm": 4.879392566075788,
+      "learning_rate": 5.6587157844650865e-06,
+      "loss": 0.6989,
+      "step": 6830
+    },
+    {
+      "epoch": 0.47430912373281486,
+      "grad_norm": 4.231585181179906,
+      "learning_rate": 5.657601081664219e-06,
+      "loss": 0.545,
+      "step": 6831
+    },
+    {
+      "epoch": 0.474378558533537,
+      "grad_norm": 4.044458555523051,
+      "learning_rate": 5.656486345602633e-06,
+      "loss": 0.6012,
+      "step": 6832
+    },
+    {
+      "epoch": 0.47444799333425913,
+      "grad_norm": 3.5959176189119795,
+      "learning_rate": 5.65537157633671e-06,
+      "loss": 0.5022,
+      "step": 6833
+    },
+    {
+      "epoch": 0.47451742813498127,
+      "grad_norm": 2.687586781158489,
+      "learning_rate": 5.654256773922835e-06,
+      "loss": 0.2664,
+      "step": 6834
+    },
+    {
+      "epoch": 0.4745868629357034,
+      "grad_norm": 3.63289143660525,
+      "learning_rate": 5.653141938417393e-06,
+      "loss": 0.4087,
+      "step": 6835
+    },
+    {
+      "epoch": 0.4746562977364255,
+      "grad_norm": 2.8785834644162605,
+      "learning_rate": 5.652027069876769e-06,
+      "loss": 0.2605,
+      "step": 6836
+    },
+    {
+      "epoch": 0.4747257325371476,
+      "grad_norm": 2.941691162408179,
+      "learning_rate": 5.650912168357355e-06,
+      "loss": 0.2483,
+      "step": 6837
+    },
+    {
+      "epoch": 0.47479516733786975,
+      "grad_norm": 3.648821669470954,
+      "learning_rate": 5.649797233915539e-06,
+      "loss": 0.5812,
+      "step": 6838
+    },
+    {
+      "epoch": 0.4748646021385919,
+      "grad_norm": 4.041615463098413,
+      "learning_rate": 5.648682266607715e-06,
+      "loss": 0.5791,
+      "step": 6839
+    },
+    {
+      "epoch": 0.47493403693931396,
+      "grad_norm": 3.9356304671476674,
+      "learning_rate": 5.647567266490276e-06,
+      "loss": 0.5223,
+      "step": 6840
+    },
+    {
+      "epoch": 0.4750034717400361,
+      "grad_norm": 3.6974502860589493,
+      "learning_rate": 5.646452233619619e-06,
+      "loss": 0.3477,
+      "step": 6841
+    },
+    {
+      "epoch": 0.47507290654075823,
+      "grad_norm": 4.16395822953417,
+      "learning_rate": 5.6453371680521375e-06,
+      "loss": 0.5913,
+      "step": 6842
+    },
+    {
+      "epoch": 0.47514234134148037,
+      "grad_norm": 3.7122191065307835,
+      "learning_rate": 5.644222069844232e-06,
+      "loss": 0.31,
+      "step": 6843
+    },
+    {
+      "epoch": 0.47521177614220245,
+      "grad_norm": 3.807580927204713,
+      "learning_rate": 5.643106939052305e-06,
+      "loss": 0.4784,
+      "step": 6844
+    },
+    {
+      "epoch": 0.4752812109429246,
+      "grad_norm": 2.511359419128291,
+      "learning_rate": 5.641991775732756e-06,
+      "loss": 0.217,
+      "step": 6845
+    },
+    {
+      "epoch": 0.4753506457436467,
+      "grad_norm": 4.493740716968294,
+      "learning_rate": 5.640876579941988e-06,
+      "loss": 0.578,
+      "step": 6846
+    },
+    {
+      "epoch": 0.47542008054436885,
+      "grad_norm": 3.450773154357236,
+      "learning_rate": 5.63976135173641e-06,
+      "loss": 0.3787,
+      "step": 6847
+    },
+    {
+      "epoch": 0.475489515345091,
+      "grad_norm": 2.063527958346874,
+      "learning_rate": 5.638646091172426e-06,
+      "loss": 0.165,
+      "step": 6848
+    },
+    {
+      "epoch": 0.47555895014581306,
+      "grad_norm": 3.4050633134737542,
+      "learning_rate": 5.637530798306445e-06,
+      "loss": 0.3151,
+      "step": 6849
+    },
+    {
+      "epoch": 0.4756283849465352,
+      "grad_norm": 3.025376782644657,
+      "learning_rate": 5.6364154731948765e-06,
+      "loss": 0.2757,
+      "step": 6850
+    },
+    {
+      "epoch": 0.47569781974725733,
+      "grad_norm": 3.607501158083564,
+      "learning_rate": 5.635300115894135e-06,
+      "loss": 0.613,
+      "step": 6851
+    },
+    {
+      "epoch": 0.47576725454797947,
+      "grad_norm": 4.109430756620664,
+      "learning_rate": 5.634184726460632e-06,
+      "loss": 0.6593,
+      "step": 6852
+    },
+    {
+      "epoch": 0.47583668934870155,
+      "grad_norm": 4.186070690159678,
+      "learning_rate": 5.633069304950784e-06,
+      "loss": 0.6286,
+      "step": 6853
+    },
+    {
+      "epoch": 0.4759061241494237,
+      "grad_norm": 3.6984068720825234,
+      "learning_rate": 5.631953851421007e-06,
+      "loss": 0.2947,
+      "step": 6854
+    },
+    {
+      "epoch": 0.4759755589501458,
+      "grad_norm": 4.498567239354485,
+      "learning_rate": 5.6308383659277184e-06,
+      "loss": 0.7054,
+      "step": 6855
+    },
+    {
+      "epoch": 0.47604499375086795,
+      "grad_norm": 4.037436376071036,
+      "learning_rate": 5.629722848527338e-06,
+      "loss": 0.4119,
+      "step": 6856
+    },
+    {
+      "epoch": 0.47611442855159003,
+      "grad_norm": 3.0672514631053494,
+      "learning_rate": 5.628607299276292e-06,
+      "loss": 0.1917,
+      "step": 6857
+    },
+    {
+      "epoch": 0.47618386335231216,
+      "grad_norm": 3.76752336053928,
+      "learning_rate": 5.627491718230997e-06,
+      "loss": 0.5562,
+      "step": 6858
+    },
+    {
+      "epoch": 0.4762532981530343,
+      "grad_norm": 4.140813254773085,
+      "learning_rate": 5.6263761054478814e-06,
+      "loss": 0.6648,
+      "step": 6859
+    },
+    {
+      "epoch": 0.47632273295375643,
+      "grad_norm": 3.4528271235043815,
+      "learning_rate": 5.6252604609833725e-06,
+      "loss": 0.5038,
+      "step": 6860
+    },
+    {
+      "epoch": 0.47639216775447857,
+      "grad_norm": 5.6185496371337145,
+      "learning_rate": 5.624144784893896e-06,
+      "loss": 0.6925,
+      "step": 6861
+    },
+    {
+      "epoch": 0.47646160255520065,
+      "grad_norm": 5.18851194358585,
+      "learning_rate": 5.623029077235883e-06,
+      "loss": 0.6733,
+      "step": 6862
+    },
+    {
+      "epoch": 0.4765310373559228,
+      "grad_norm": 3.6075112143165744,
+      "learning_rate": 5.621913338065764e-06,
+      "loss": 0.4313,
+      "step": 6863
+    },
+    {
+      "epoch": 0.4766004721566449,
+      "grad_norm": 3.638654786300206,
+      "learning_rate": 5.620797567439973e-06,
+      "loss": 0.5565,
+      "step": 6864
+    },
+    {
+      "epoch": 0.47666990695736705,
+      "grad_norm": 3.8049830169194077,
+      "learning_rate": 5.6196817654149415e-06,
+      "loss": 0.3336,
+      "step": 6865
+    },
+    {
+      "epoch": 0.47673934175808913,
+      "grad_norm": 4.11430271998524,
+      "learning_rate": 5.618565932047109e-06,
+      "loss": 0.5104,
+      "step": 6866
+    },
+    {
+      "epoch": 0.47680877655881126,
+      "grad_norm": 3.145703709202706,
+      "learning_rate": 5.6174500673929126e-06,
+      "loss": 0.2909,
+      "step": 6867
+    },
+    {
+      "epoch": 0.4768782113595334,
+      "grad_norm": 3.9987729475673666,
+      "learning_rate": 5.6163341715087884e-06,
+      "loss": 0.6205,
+      "step": 6868
+    },
+    {
+      "epoch": 0.47694764616025553,
+      "grad_norm": 4.392068622771916,
+      "learning_rate": 5.6152182444511795e-06,
+      "loss": 0.6173,
+      "step": 6869
+    },
+    {
+      "epoch": 0.47701708096097767,
+      "grad_norm": 3.0889245040899054,
+      "learning_rate": 5.6141022862765285e-06,
+      "loss": 0.204,
+      "step": 6870
+    },
+    {
+      "epoch": 0.47708651576169975,
+      "grad_norm": 2.878083209137268,
+      "learning_rate": 5.6129862970412785e-06,
+      "loss": 0.4188,
+      "step": 6871
+    },
+    {
+      "epoch": 0.4771559505624219,
+      "grad_norm": 2.7472741013184274,
+      "learning_rate": 5.611870276801875e-06,
+      "loss": 0.3325,
+      "step": 6872
+    },
+    {
+      "epoch": 0.477225385363144,
+      "grad_norm": 3.653764927358615,
+      "learning_rate": 5.610754225614767e-06,
+      "loss": 0.48,
+      "step": 6873
+    },
+    {
+      "epoch": 0.47729482016386615,
+      "grad_norm": 3.916384246052207,
+      "learning_rate": 5.609638143536399e-06,
+      "loss": 0.502,
+      "step": 6874
+    },
+    {
+      "epoch": 0.47736425496458823,
+      "grad_norm": 3.54344228155308,
+      "learning_rate": 5.608522030623224e-06,
+      "loss": 0.3945,
+      "step": 6875
+    },
+    {
+      "epoch": 0.47743368976531037,
+      "grad_norm": 4.152623121685118,
+      "learning_rate": 5.607405886931695e-06,
+      "loss": 0.422,
+      "step": 6876
+    },
+    {
+      "epoch": 0.4775031245660325,
+      "grad_norm": 3.1658735291811633,
+      "learning_rate": 5.606289712518262e-06,
+      "loss": 0.3394,
+      "step": 6877
+    },
+    {
+      "epoch": 0.47757255936675463,
+      "grad_norm": 3.627879497229283,
+      "learning_rate": 5.605173507439383e-06,
+      "loss": 0.5644,
+      "step": 6878
+    },
+    {
+      "epoch": 0.4776419941674767,
+      "grad_norm": 3.430171247443991,
+      "learning_rate": 5.604057271751512e-06,
+      "loss": 0.4958,
+      "step": 6879
+    },
+    {
+      "epoch": 0.47771142896819885,
+      "grad_norm": 3.6680342882311185,
+      "learning_rate": 5.602941005511108e-06,
+      "loss": 0.376,
+      "step": 6880
+    },
+    {
+      "epoch": 0.477780863768921,
+      "grad_norm": 3.4917121726224667,
+      "learning_rate": 5.601824708774628e-06,
+      "loss": 0.3612,
+      "step": 6881
+    },
+    {
+      "epoch": 0.4778502985696431,
+      "grad_norm": 2.878605142881803,
+      "learning_rate": 5.6007083815985374e-06,
+      "loss": 0.2539,
+      "step": 6882
+    },
+    {
+      "epoch": 0.47791973337036525,
+      "grad_norm": 4.05497237255503,
+      "learning_rate": 5.599592024039297e-06,
+      "loss": 0.6409,
+      "step": 6883
+    },
+    {
+      "epoch": 0.47798916817108733,
+      "grad_norm": 3.9288315105684544,
+      "learning_rate": 5.598475636153369e-06,
+      "loss": 0.5477,
+      "step": 6884
+    },
+    {
+      "epoch": 0.47805860297180947,
+      "grad_norm": 3.595512501740939,
+      "learning_rate": 5.597359217997222e-06,
+      "loss": 0.4696,
+      "step": 6885
+    },
+    {
+      "epoch": 0.4781280377725316,
+      "grad_norm": 4.35193461591855,
+      "learning_rate": 5.5962427696273206e-06,
+      "loss": 0.5946,
+      "step": 6886
+    },
+    {
+      "epoch": 0.47819747257325373,
+      "grad_norm": 4.466316643509731,
+      "learning_rate": 5.595126291100134e-06,
+      "loss": 0.5,
+      "step": 6887
+    },
+    {
+      "epoch": 0.4782669073739758,
+      "grad_norm": 3.2935441115985187,
+      "learning_rate": 5.594009782472134e-06,
+      "loss": 0.3136,
+      "step": 6888
+    },
+    {
+      "epoch": 0.47833634217469795,
+      "grad_norm": 3.442847498412359,
+      "learning_rate": 5.592893243799793e-06,
+      "loss": 0.2528,
+      "step": 6889
+    },
+    {
+      "epoch": 0.4784057769754201,
+      "grad_norm": 4.074573609257723,
+      "learning_rate": 5.59177667513958e-06,
+      "loss": 0.4578,
+      "step": 6890
+    },
+    {
+      "epoch": 0.4784752117761422,
+      "grad_norm": 3.3762103261527923,
+      "learning_rate": 5.590660076547974e-06,
+      "loss": 0.3324,
+      "step": 6891
+    },
+    {
+      "epoch": 0.47854464657686435,
+      "grad_norm": 2.445749368959596,
+      "learning_rate": 5.589543448081448e-06,
+      "loss": 0.1821,
+      "step": 6892
+    },
+    {
+      "epoch": 0.47861408137758643,
+      "grad_norm": 3.748354979387605,
+      "learning_rate": 5.588426789796483e-06,
+      "loss": 0.4633,
+      "step": 6893
+    },
+    {
+      "epoch": 0.47868351617830857,
+      "grad_norm": 4.7330246567560055,
+      "learning_rate": 5.587310101749557e-06,
+      "loss": 0.357,
+      "step": 6894
+    },
+    {
+      "epoch": 0.4787529509790307,
+      "grad_norm": 3.8172098401676298,
+      "learning_rate": 5.586193383997149e-06,
+      "loss": 0.5969,
+      "step": 6895
+    },
+    {
+      "epoch": 0.47882238577975283,
+      "grad_norm": 3.1169255852256432,
+      "learning_rate": 5.585076636595744e-06,
+      "loss": 0.3476,
+      "step": 6896
+    },
+    {
+      "epoch": 0.4788918205804749,
+      "grad_norm": 2.9050568090816697,
+      "learning_rate": 5.583959859601824e-06,
+      "loss": 0.3122,
+      "step": 6897
+    },
+    {
+      "epoch": 0.47896125538119705,
+      "grad_norm": 4.254835872587314,
+      "learning_rate": 5.582843053071877e-06,
+      "loss": 0.5182,
+      "step": 6898
+    },
+    {
+      "epoch": 0.4790306901819192,
+      "grad_norm": 5.191033855448743,
+      "learning_rate": 5.5817262170623865e-06,
+      "loss": 0.6578,
+      "step": 6899
+    },
+    {
+      "epoch": 0.4791001249826413,
+      "grad_norm": 4.192838803153526,
+      "learning_rate": 5.580609351629842e-06,
+      "loss": 0.5865,
+      "step": 6900
+    },
+    {
+      "epoch": 0.4791695597833634,
+      "grad_norm": 3.2599993604117774,
+      "learning_rate": 5.5794924568307355e-06,
+      "loss": 0.31,
+      "step": 6901
+    },
+    {
+      "epoch": 0.47923899458408553,
+      "grad_norm": 3.8075429259870304,
+      "learning_rate": 5.578375532721554e-06,
+      "loss": 0.4033,
+      "step": 6902
+    },
+    {
+      "epoch": 0.47930842938480767,
+      "grad_norm": 4.398786312918532,
+      "learning_rate": 5.577258579358794e-06,
+      "loss": 0.5978,
+      "step": 6903
+    },
+    {
+      "epoch": 0.4793778641855298,
+      "grad_norm": 3.496795971192161,
+      "learning_rate": 5.5761415967989464e-06,
+      "loss": 0.4616,
+      "step": 6904
+    },
+    {
+      "epoch": 0.47944729898625194,
+      "grad_norm": 3.9163787725176835,
+      "learning_rate": 5.57502458509851e-06,
+      "loss": 0.512,
+      "step": 6905
+    },
+    {
+      "epoch": 0.479516733786974,
+      "grad_norm": 5.538483485289079,
+      "learning_rate": 5.573907544313981e-06,
+      "loss": 0.7099,
+      "step": 6906
+    },
+    {
+      "epoch": 0.47958616858769615,
+      "grad_norm": 3.3088933948981696,
+      "learning_rate": 5.572790474501857e-06,
+      "loss": 0.3226,
+      "step": 6907
+    },
+    {
+      "epoch": 0.4796556033884183,
+      "grad_norm": 3.554231417179007,
+      "learning_rate": 5.5716733757186406e-06,
+      "loss": 0.5112,
+      "step": 6908
+    },
+    {
+      "epoch": 0.4797250381891404,
+      "grad_norm": 3.7105077137238043,
+      "learning_rate": 5.57055624802083e-06,
+      "loss": 0.3253,
+      "step": 6909
+    },
+    {
+      "epoch": 0.4797944729898625,
+      "grad_norm": 3.5395388443250146,
+      "learning_rate": 5.569439091464931e-06,
+      "loss": 0.3743,
+      "step": 6910
+    },
+    {
+      "epoch": 0.47986390779058463,
+      "grad_norm": 3.0983258613078632,
+      "learning_rate": 5.5683219061074455e-06,
+      "loss": 0.3293,
+      "step": 6911
+    },
+    {
+      "epoch": 0.47993334259130677,
+      "grad_norm": 3.7405967197390018,
+      "learning_rate": 5.5672046920048835e-06,
+      "loss": 0.499,
+      "step": 6912
+    },
+    {
+      "epoch": 0.4800027773920289,
+      "grad_norm": 3.876461888308058,
+      "learning_rate": 5.566087449213748e-06,
+      "loss": 0.4959,
+      "step": 6913
+    },
+    {
+      "epoch": 0.480072212192751,
+      "grad_norm": 2.640293489427926,
+      "learning_rate": 5.564970177790552e-06,
+      "loss": 0.1982,
+      "step": 6914
+    },
+    {
+      "epoch": 0.4801416469934731,
+      "grad_norm": 3.822609780965002,
+      "learning_rate": 5.5638528777918024e-06,
+      "loss": 0.423,
+      "step": 6915
+    },
+    {
+      "epoch": 0.48021108179419525,
+      "grad_norm": 3.0085274032615064,
+      "learning_rate": 5.562735549274012e-06,
+      "loss": 0.3137,
+      "step": 6916
+    },
+    {
+      "epoch": 0.4802805165949174,
+      "grad_norm": 4.830303009526612,
+      "learning_rate": 5.561618192293695e-06,
+      "loss": 0.8478,
+      "step": 6917
+    },
+    {
+      "epoch": 0.4803499513956395,
+      "grad_norm": 4.388400865237029,
+      "learning_rate": 5.5605008069073655e-06,
+      "loss": 0.6717,
+      "step": 6918
+    },
+    {
+      "epoch": 0.4804193861963616,
+      "grad_norm": 3.541469917585858,
+      "learning_rate": 5.559383393171538e-06,
+      "loss": 0.463,
+      "step": 6919
+    },
+    {
+      "epoch": 0.48048882099708373,
+      "grad_norm": 3.4132899966746404,
+      "learning_rate": 5.5582659511427315e-06,
+      "loss": 0.4368,
+      "step": 6920
+    },
+    {
+      "epoch": 0.48055825579780587,
+      "grad_norm": 2.9292842685575953,
+      "learning_rate": 5.557148480877467e-06,
+      "loss": 0.3308,
+      "step": 6921
+    },
+    {
+      "epoch": 0.480627690598528,
+      "grad_norm": 2.819719721077523,
+      "learning_rate": 5.556030982432261e-06,
+      "loss": 0.3713,
+      "step": 6922
+    },
+    {
+      "epoch": 0.4806971253992501,
+      "grad_norm": 3.092139446288678,
+      "learning_rate": 5.5549134558636375e-06,
+      "loss": 0.4073,
+      "step": 6923
+    },
+    {
+      "epoch": 0.4807665601999722,
+      "grad_norm": 3.276199820665668,
+      "learning_rate": 5.553795901228121e-06,
+      "loss": 0.3676,
+      "step": 6924
+    },
+    {
+      "epoch": 0.48083599500069435,
+      "grad_norm": 3.2317018216467934,
+      "learning_rate": 5.552678318582232e-06,
+      "loss": 0.4514,
+      "step": 6925
+    },
+    {
+      "epoch": 0.4809054298014165,
+      "grad_norm": 3.210481088970257,
+      "learning_rate": 5.551560707982501e-06,
+      "loss": 0.3899,
+      "step": 6926
+    },
+    {
+      "epoch": 0.4809748646021386,
+      "grad_norm": 3.5502074859850983,
+      "learning_rate": 5.550443069485452e-06,
+      "loss": 0.529,
+      "step": 6927
+    },
+    {
+      "epoch": 0.4810442994028607,
+      "grad_norm": 3.5994570695687513,
+      "learning_rate": 5.549325403147616e-06,
+      "loss": 0.43,
+      "step": 6928
+    },
+    {
+      "epoch": 0.48111373420358283,
+      "grad_norm": 3.6930306118730245,
+      "learning_rate": 5.5482077090255215e-06,
+      "loss": 0.4677,
+      "step": 6929
+    },
+    {
+      "epoch": 0.48118316900430497,
+      "grad_norm": 3.1986492975802547,
+      "learning_rate": 5.5470899871757034e-06,
+      "loss": 0.3978,
+      "step": 6930
+    },
+    {
+      "epoch": 0.4812526038050271,
+      "grad_norm": 3.6768283482726956,
+      "learning_rate": 5.545972237654692e-06,
+      "loss": 0.3953,
+      "step": 6931
+    },
+    {
+      "epoch": 0.4813220386057492,
+      "grad_norm": 3.0719332372228862,
+      "learning_rate": 5.5448544605190224e-06,
+      "loss": 0.4218,
+      "step": 6932
+    },
+    {
+      "epoch": 0.4813914734064713,
+      "grad_norm": 4.882881498690462,
+      "learning_rate": 5.54373665582523e-06,
+      "loss": 0.7839,
+      "step": 6933
+    },
+    {
+      "epoch": 0.48146090820719345,
+      "grad_norm": 3.639410359653315,
+      "learning_rate": 5.5426188236298565e-06,
+      "loss": 0.3294,
+      "step": 6934
+    },
+    {
+      "epoch": 0.4815303430079156,
+      "grad_norm": 3.626441459686348,
+      "learning_rate": 5.541500963989433e-06,
+      "loss": 0.3336,
+      "step": 6935
+    },
+    {
+      "epoch": 0.48159977780863766,
+      "grad_norm": 3.953041244999979,
+      "learning_rate": 5.540383076960505e-06,
+      "loss": 0.4123,
+      "step": 6936
+    },
+    {
+      "epoch": 0.4816692126093598,
+      "grad_norm": 4.526534768297793,
+      "learning_rate": 5.539265162599612e-06,
+      "loss": 0.5172,
+      "step": 6937
+    },
+    {
+      "epoch": 0.48173864741008193,
+      "grad_norm": 4.899549171352108,
+      "learning_rate": 5.538147220963297e-06,
+      "loss": 0.6023,
+      "step": 6938
+    },
+    {
+      "epoch": 0.48180808221080407,
+      "grad_norm": 3.766134267984721,
+      "learning_rate": 5.537029252108106e-06,
+      "loss": 0.359,
+      "step": 6939
+    },
+    {
+      "epoch": 0.4818775170115262,
+      "grad_norm": 2.7574986754881907,
+      "learning_rate": 5.53591125609058e-06,
+      "loss": 0.2612,
+      "step": 6940
+    },
+    {
+      "epoch": 0.4819469518122483,
+      "grad_norm": 3.968142391790459,
+      "learning_rate": 5.534793232967273e-06,
+      "loss": 0.5711,
+      "step": 6941
+    },
+    {
+      "epoch": 0.4820163866129704,
+      "grad_norm": 3.8524787646436485,
+      "learning_rate": 5.533675182794728e-06,
+      "loss": 0.6372,
+      "step": 6942
+    },
+    {
+      "epoch": 0.48208582141369255,
+      "grad_norm": 8.165968067738195,
+      "learning_rate": 5.5325571056294945e-06,
+      "loss": 0.5467,
+      "step": 6943
+    },
+    {
+      "epoch": 0.4821552562144147,
+      "grad_norm": 4.701594522723309,
+      "learning_rate": 5.531439001528128e-06,
+      "loss": 0.3691,
+      "step": 6944
+    },
+    {
+      "epoch": 0.48222469101513676,
+      "grad_norm": 3.6774611846374565,
+      "learning_rate": 5.5303208705471756e-06,
+      "loss": 0.5579,
+      "step": 6945
+    },
+    {
+      "epoch": 0.4822941258158589,
+      "grad_norm": 4.402197250710555,
+      "learning_rate": 5.529202712743194e-06,
+      "loss": 0.5021,
+      "step": 6946
+    },
+    {
+      "epoch": 0.48236356061658103,
+      "grad_norm": 3.2362805677283575,
+      "learning_rate": 5.52808452817274e-06,
+      "loss": 0.3996,
+      "step": 6947
+    },
+    {
+      "epoch": 0.48243299541730317,
+      "grad_norm": 4.881821744103387,
+      "learning_rate": 5.526966316892367e-06,
+      "loss": 0.5232,
+      "step": 6948
+    },
+    {
+      "epoch": 0.48250243021802525,
+      "grad_norm": 4.898920162475988,
+      "learning_rate": 5.525848078958634e-06,
+      "loss": 0.3947,
+      "step": 6949
+    },
+    {
+      "epoch": 0.4825718650187474,
+      "grad_norm": 2.557136591505448,
+      "learning_rate": 5.5247298144280995e-06,
+      "loss": 0.261,
+      "step": 6950
+    },
+    {
+      "epoch": 0.4826412998194695,
+      "grad_norm": 3.456637004090075,
+      "learning_rate": 5.523611523357326e-06,
+      "loss": 0.4461,
+      "step": 6951
+    },
+    {
+      "epoch": 0.48271073462019165,
+      "grad_norm": 5.136043732008318,
+      "learning_rate": 5.522493205802873e-06,
+      "loss": 0.8836,
+      "step": 6952
+    },
+    {
+      "epoch": 0.4827801694209138,
+      "grad_norm": 4.149510195418904,
+      "learning_rate": 5.5213748618213055e-06,
+      "loss": 0.5754,
+      "step": 6953
+    },
+    {
+      "epoch": 0.48284960422163586,
+      "grad_norm": 5.0614104046568515,
+      "learning_rate": 5.520256491469187e-06,
+      "loss": 0.5501,
+      "step": 6954
+    },
+    {
+      "epoch": 0.482919039022358,
+      "grad_norm": 2.6423079631453086,
+      "learning_rate": 5.519138094803085e-06,
+      "loss": 0.2684,
+      "step": 6955
+    },
+    {
+      "epoch": 0.48298847382308013,
+      "grad_norm": 3.0468470414439945,
+      "learning_rate": 5.518019671879564e-06,
+      "loss": 0.471,
+      "step": 6956
+    },
+    {
+      "epoch": 0.48305790862380227,
+      "grad_norm": 5.515769984123761,
+      "learning_rate": 5.516901222755195e-06,
+      "loss": 0.5988,
+      "step": 6957
+    },
+    {
+      "epoch": 0.48312734342452435,
+      "grad_norm": 3.2011807681663793,
+      "learning_rate": 5.515782747486549e-06,
+      "loss": 0.2984,
+      "step": 6958
+    },
+    {
+      "epoch": 0.4831967782252465,
+      "grad_norm": 2.2121177030803167,
+      "learning_rate": 5.5146642461301915e-06,
+      "loss": 0.2455,
+      "step": 6959
+    },
+    {
+      "epoch": 0.4832662130259686,
+      "grad_norm": 4.208613049488054,
+      "learning_rate": 5.513545718742702e-06,
+      "loss": 0.5201,
+      "step": 6960
+    },
+    {
+      "epoch": 0.48333564782669075,
+      "grad_norm": 3.055382881476904,
+      "learning_rate": 5.512427165380649e-06,
+      "loss": 0.2731,
+      "step": 6961
+    },
+    {
+      "epoch": 0.4834050826274129,
+      "grad_norm": 4.185969415502752,
+      "learning_rate": 5.511308586100611e-06,
+      "loss": 0.5186,
+      "step": 6962
+    },
+    {
+      "epoch": 0.48347451742813496,
+      "grad_norm": 3.27396345127389,
+      "learning_rate": 5.510189980959163e-06,
+      "loss": 0.4021,
+      "step": 6963
+    },
+    {
+      "epoch": 0.4835439522288571,
+      "grad_norm": 3.1102008975673914,
+      "learning_rate": 5.509071350012883e-06,
+      "loss": 0.3534,
+      "step": 6964
+    },
+    {
+      "epoch": 0.48361338702957923,
+      "grad_norm": 4.328759916473042,
+      "learning_rate": 5.5079526933183496e-06,
+      "loss": 0.465,
+      "step": 6965
+    },
+    {
+      "epoch": 0.48368282183030137,
+      "grad_norm": 3.2923007372167263,
+      "learning_rate": 5.506834010932145e-06,
+      "loss": 0.3065,
+      "step": 6966
+    },
+    {
+      "epoch": 0.48375225663102345,
+      "grad_norm": 5.536271458988812,
+      "learning_rate": 5.50571530291085e-06,
+      "loss": 0.731,
+      "step": 6967
+    },
+    {
+      "epoch": 0.4838216914317456,
+      "grad_norm": 3.5332932561127572,
+      "learning_rate": 5.504596569311044e-06,
+      "loss": 0.2962,
+      "step": 6968
+    },
+    {
+      "epoch": 0.4838911262324677,
+      "grad_norm": 4.360482451786159,
+      "learning_rate": 5.503477810189318e-06,
+      "loss": 0.5836,
+      "step": 6969
+    },
+    {
+      "epoch": 0.48396056103318985,
+      "grad_norm": 4.375020346718677,
+      "learning_rate": 5.502359025602253e-06,
+      "loss": 0.578,
+      "step": 6970
+    },
+    {
+      "epoch": 0.48402999583391193,
+      "grad_norm": 4.625206674837379,
+      "learning_rate": 5.501240215606437e-06,
+      "loss": 0.6806,
+      "step": 6971
+    },
+    {
+      "epoch": 0.48409943063463406,
+      "grad_norm": 2.6111694213058114,
+      "learning_rate": 5.500121380258456e-06,
+      "loss": 0.2446,
+      "step": 6972
+    },
+    {
+      "epoch": 0.4841688654353562,
+      "grad_norm": 4.931910887739966,
+      "learning_rate": 5.499002519614906e-06,
+      "loss": 0.6414,
+      "step": 6973
+    },
+    {
+      "epoch": 0.48423830023607833,
+      "grad_norm": 4.453377758831862,
+      "learning_rate": 5.497883633732369e-06,
+      "loss": 0.4173,
+      "step": 6974
+    },
+    {
+      "epoch": 0.48430773503680047,
+      "grad_norm": 3.4753371181469683,
+      "learning_rate": 5.4967647226674435e-06,
+      "loss": 0.4151,
+      "step": 6975
+    },
+    {
+      "epoch": 0.48437716983752255,
+      "grad_norm": 4.983755723896898,
+      "learning_rate": 5.49564578647672e-06,
+      "loss": 0.5652,
+      "step": 6976
+    },
+    {
+      "epoch": 0.4844466046382447,
+      "grad_norm": 3.2378244701895116,
+      "learning_rate": 5.494526825216793e-06,
+      "loss": 0.4206,
+      "step": 6977
+    },
+    {
+      "epoch": 0.4845160394389668,
+      "grad_norm": 2.254154446435162,
+      "learning_rate": 5.493407838944259e-06,
+      "loss": 0.1622,
+      "step": 6978
+    },
+    {
+      "epoch": 0.48458547423968895,
+      "grad_norm": 3.2083847517799122,
+      "learning_rate": 5.492288827715715e-06,
+      "loss": 0.2658,
+      "step": 6979
+    },
+    {
+      "epoch": 0.48465490904041103,
+      "grad_norm": 3.198020158176356,
+      "learning_rate": 5.491169791587759e-06,
+      "loss": 0.498,
+      "step": 6980
+    },
+    {
+      "epoch": 0.48472434384113317,
+      "grad_norm": 3.92769984715268,
+      "learning_rate": 5.490050730616991e-06,
+      "loss": 0.5434,
+      "step": 6981
+    },
+    {
+      "epoch": 0.4847937786418553,
+      "grad_norm": 3.6286286735741005,
+      "learning_rate": 5.4889316448600104e-06,
+      "loss": 0.3701,
+      "step": 6982
+    },
+    {
+      "epoch": 0.48486321344257743,
+      "grad_norm": 5.12574482288133,
+      "learning_rate": 5.487812534373423e-06,
+      "loss": 0.5249,
+      "step": 6983
+    },
+    {
+      "epoch": 0.48493264824329957,
+      "grad_norm": 4.939031066254939,
+      "learning_rate": 5.486693399213827e-06,
+      "loss": 0.5723,
+      "step": 6984
+    },
+    {
+      "epoch": 0.48500208304402165,
+      "grad_norm": 3.508050947216148,
+      "learning_rate": 5.485574239437831e-06,
+      "loss": 0.356,
+      "step": 6985
+    },
+    {
+      "epoch": 0.4850715178447438,
+      "grad_norm": 3.2832477202887205,
+      "learning_rate": 5.484455055102039e-06,
+      "loss": 0.3478,
+      "step": 6986
+    },
+    {
+      "epoch": 0.4851409526454659,
+      "grad_norm": 4.419111943834761,
+      "learning_rate": 5.483335846263059e-06,
+      "loss": 0.5915,
+      "step": 6987
+    },
+    {
+      "epoch": 0.48521038744618805,
+      "grad_norm": 3.022562895674062,
+      "learning_rate": 5.482216612977497e-06,
+      "loss": 0.4529,
+      "step": 6988
+    },
+    {
+      "epoch": 0.48527982224691013,
+      "grad_norm": 4.44636186440409,
+      "learning_rate": 5.481097355301968e-06,
+      "loss": 0.3865,
+      "step": 6989
+    },
+    {
+      "epoch": 0.48534925704763227,
+      "grad_norm": 3.2376941666003205,
+      "learning_rate": 5.4799780732930755e-06,
+      "loss": 0.3118,
+      "step": 6990
+    },
+    {
+      "epoch": 0.4854186918483544,
+      "grad_norm": 15.034008485773793,
+      "learning_rate": 5.478858767007436e-06,
+      "loss": 0.5178,
+      "step": 6991
+    },
+    {
+      "epoch": 0.48548812664907653,
+      "grad_norm": 3.2225523327747587,
+      "learning_rate": 5.477739436501663e-06,
+      "loss": 0.4585,
+      "step": 6992
+    },
+    {
+      "epoch": 0.4855575614497986,
+      "grad_norm": 4.02782040763264,
+      "learning_rate": 5.47662008183237e-06,
+      "loss": 0.5577,
+      "step": 6993
+    },
+    {
+      "epoch": 0.48562699625052075,
+      "grad_norm": 3.52370522902014,
+      "learning_rate": 5.475500703056173e-06,
+      "loss": 0.4555,
+      "step": 6994
+    },
+    {
+      "epoch": 0.4856964310512429,
+      "grad_norm": 3.3619368402007113,
+      "learning_rate": 5.474381300229686e-06,
+      "loss": 0.4282,
+      "step": 6995
+    },
+    {
+      "epoch": 0.485765865851965,
+      "grad_norm": 3.7271603759356418,
+      "learning_rate": 5.473261873409532e-06,
+      "loss": 0.5351,
+      "step": 6996
+    },
+    {
+      "epoch": 0.48583530065268715,
+      "grad_norm": 4.789424603339742,
+      "learning_rate": 5.472142422652327e-06,
+      "loss": 0.6059,
+      "step": 6997
+    },
+    {
+      "epoch": 0.48590473545340923,
+      "grad_norm": 2.6327874954895716,
+      "learning_rate": 5.471022948014692e-06,
+      "loss": 0.2564,
+      "step": 6998
+    },
+    {
+      "epoch": 0.48597417025413137,
+      "grad_norm": 3.8982884075469695,
+      "learning_rate": 5.46990344955325e-06,
+      "loss": 0.4103,
+      "step": 6999
+    },
+    {
+      "epoch": 0.4860436050548535,
+      "grad_norm": 3.530450155813866,
+      "learning_rate": 5.4687839273246236e-06,
+      "loss": 0.5684,
+      "step": 7000
+    },
+    {
+      "epoch": 0.48611303985557563,
+      "grad_norm": 4.467241391084448,
+      "learning_rate": 5.467664381385437e-06,
+      "loss": 0.5692,
+      "step": 7001
+    },
+    {
+      "epoch": 0.4861824746562977,
+      "grad_norm": 3.6942966954454532,
+      "learning_rate": 5.466544811792314e-06,
+      "loss": 0.4233,
+      "step": 7002
+    },
+    {
+      "epoch": 0.48625190945701985,
+      "grad_norm": 3.957394547192311,
+      "learning_rate": 5.465425218601883e-06,
+      "loss": 0.1946,
+      "step": 7003
+    },
+    {
+      "epoch": 0.486321344257742,
+      "grad_norm": 3.441764463297761,
+      "learning_rate": 5.46430560187077e-06,
+      "loss": 0.5798,
+      "step": 7004
+    },
+    {
+      "epoch": 0.4863907790584641,
+      "grad_norm": 3.2716238644687867,
+      "learning_rate": 5.463185961655607e-06,
+      "loss": 0.36,
+      "step": 7005
+    },
+    {
+      "epoch": 0.4864602138591862,
+      "grad_norm": 4.549098908901638,
+      "learning_rate": 5.46206629801302e-06,
+      "loss": 0.7096,
+      "step": 7006
+    },
+    {
+      "epoch": 0.48652964865990833,
+      "grad_norm": 3.9242384803116432,
+      "learning_rate": 5.4609466109996445e-06,
+      "loss": 0.4479,
+      "step": 7007
+    },
+    {
+      "epoch": 0.48659908346063047,
+      "grad_norm": 3.482857185747351,
+      "learning_rate": 5.459826900672109e-06,
+      "loss": 0.1778,
+      "step": 7008
+    },
+    {
+      "epoch": 0.4866685182613526,
+      "grad_norm": 3.267806434218714,
+      "learning_rate": 5.4587071670870515e-06,
+      "loss": 0.3487,
+      "step": 7009
+    },
+    {
+      "epoch": 0.48673795306207474,
+      "grad_norm": 4.008501552147815,
+      "learning_rate": 5.457587410301104e-06,
+      "loss": 0.5819,
+      "step": 7010
+    },
+    {
+      "epoch": 0.4868073878627968,
+      "grad_norm": 3.721885054933105,
+      "learning_rate": 5.4564676303709e-06,
+      "loss": 0.3322,
+      "step": 7011
+    },
+    {
+      "epoch": 0.48687682266351895,
+      "grad_norm": 4.037027769775607,
+      "learning_rate": 5.455347827353085e-06,
+      "loss": 0.5758,
+      "step": 7012
+    },
+    {
+      "epoch": 0.4869462574642411,
+      "grad_norm": 11.805811688208767,
+      "learning_rate": 5.454228001304288e-06,
+      "loss": 0.413,
+      "step": 7013
+    },
+    {
+      "epoch": 0.4870156922649632,
+      "grad_norm": 4.482429578626988,
+      "learning_rate": 5.453108152281154e-06,
+      "loss": 0.6947,
+      "step": 7014
+    },
+    {
+      "epoch": 0.4870851270656853,
+      "grad_norm": 2.997358411330258,
+      "learning_rate": 5.451988280340322e-06,
+      "loss": 0.4322,
+      "step": 7015
+    },
+    {
+      "epoch": 0.48715456186640743,
+      "grad_norm": 3.810653214437037,
+      "learning_rate": 5.450868385538436e-06,
+      "loss": 0.3427,
+      "step": 7016
+    },
+    {
+      "epoch": 0.48722399666712957,
+      "grad_norm": 4.146119243655524,
+      "learning_rate": 5.4497484679321355e-06,
+      "loss": 0.5509,
+      "step": 7017
+    },
+    {
+      "epoch": 0.4872934314678517,
+      "grad_norm": 3.3805430736576456,
+      "learning_rate": 5.4486285275780674e-06,
+      "loss": 0.3999,
+      "step": 7018
+    },
+    {
+      "epoch": 0.48736286626857384,
+      "grad_norm": 3.1038843143646258,
+      "learning_rate": 5.447508564532875e-06,
+      "loss": 0.4051,
+      "step": 7019
+    },
+    {
+      "epoch": 0.4874323010692959,
+      "grad_norm": 3.544914678857529,
+      "learning_rate": 5.446388578853205e-06,
+      "loss": 0.4009,
+      "step": 7020
+    },
+    {
+      "epoch": 0.48750173587001805,
+      "grad_norm": 2.9970421466413257,
+      "learning_rate": 5.4452685705957084e-06,
+      "loss": 0.1605,
+      "step": 7021
+    },
+    {
+      "epoch": 0.4875711706707402,
+      "grad_norm": 3.7623921304847836,
+      "learning_rate": 5.44414853981703e-06,
+      "loss": 0.4198,
+      "step": 7022
+    },
+    {
+      "epoch": 0.4876406054714623,
+      "grad_norm": 3.6739995613831193,
+      "learning_rate": 5.443028486573821e-06,
+      "loss": 0.3413,
+      "step": 7023
+    },
+    {
+      "epoch": 0.4877100402721844,
+      "grad_norm": 3.042382801032805,
+      "learning_rate": 5.441908410922734e-06,
+      "loss": 0.2823,
+      "step": 7024
+    },
+    {
+      "epoch": 0.48777947507290653,
+      "grad_norm": 3.689622993727875,
+      "learning_rate": 5.440788312920418e-06,
+      "loss": 0.6014,
+      "step": 7025
+    },
+    {
+      "epoch": 0.48784890987362867,
+      "grad_norm": 4.233184047969592,
+      "learning_rate": 5.4396681926235285e-06,
+      "loss": 0.6397,
+      "step": 7026
+    },
+    {
+      "epoch": 0.4879183446743508,
+      "grad_norm": 3.0752474160202716,
+      "learning_rate": 5.4385480500887175e-06,
+      "loss": 0.4699,
+      "step": 7027
+    },
+    {
+      "epoch": 0.4879877794750729,
+      "grad_norm": 3.5066890091344343,
+      "learning_rate": 5.437427885372645e-06,
+      "loss": 0.5761,
+      "step": 7028
+    },
+    {
+      "epoch": 0.488057214275795,
+      "grad_norm": 3.772400078707538,
+      "learning_rate": 5.436307698531963e-06,
+      "loss": 0.4267,
+      "step": 7029
+    },
+    {
+      "epoch": 0.48812664907651715,
+      "grad_norm": 4.399794687980241,
+      "learning_rate": 5.435187489623332e-06,
+      "loss": 0.532,
+      "step": 7030
+    },
+    {
+      "epoch": 0.4881960838772393,
+      "grad_norm": 4.479420280086334,
+      "learning_rate": 5.434067258703411e-06,
+      "loss": 0.606,
+      "step": 7031
+    },
+    {
+      "epoch": 0.4882655186779614,
+      "grad_norm": 3.025849777195786,
+      "learning_rate": 5.432947005828859e-06,
+      "loss": 0.3548,
+      "step": 7032
+    },
+    {
+      "epoch": 0.4883349534786835,
+      "grad_norm": 4.850528342891305,
+      "learning_rate": 5.431826731056337e-06,
+      "loss": 0.641,
+      "step": 7033
+    },
+    {
+      "epoch": 0.48840438827940563,
+      "grad_norm": 3.2368770609065187,
+      "learning_rate": 5.430706434442508e-06,
+      "loss": 0.271,
+      "step": 7034
+    },
+    {
+      "epoch": 0.48847382308012777,
+      "grad_norm": 3.86410741783333,
+      "learning_rate": 5.429586116044034e-06,
+      "loss": 0.3402,
+      "step": 7035
+    },
+    {
+      "epoch": 0.4885432578808499,
+      "grad_norm": 4.846426014110778,
+      "learning_rate": 5.4284657759175805e-06,
+      "loss": 0.8363,
+      "step": 7036
+    },
+    {
+      "epoch": 0.488612692681572,
+      "grad_norm": 4.2225166411818,
+      "learning_rate": 5.427345414119813e-06,
+      "loss": 0.6361,
+      "step": 7037
+    },
+    {
+      "epoch": 0.4886821274822941,
+      "grad_norm": 3.1896809850729566,
+      "learning_rate": 5.4262250307074e-06,
+      "loss": 0.357,
+      "step": 7038
+    },
+    {
+      "epoch": 0.48875156228301625,
+      "grad_norm": 2.828416229092935,
+      "learning_rate": 5.425104625737004e-06,
+      "loss": 0.2406,
+      "step": 7039
+    },
+    {
+      "epoch": 0.4888209970837384,
+      "grad_norm": 4.359043070900398,
+      "learning_rate": 5.4239841992653e-06,
+      "loss": 0.5679,
+      "step": 7040
+    },
+    {
+      "epoch": 0.4888904318844605,
+      "grad_norm": 3.829085878821352,
+      "learning_rate": 5.4228637513489535e-06,
+      "loss": 0.4839,
+      "step": 7041
+    },
+    {
+      "epoch": 0.4889598666851826,
+      "grad_norm": 3.8646972208938988,
+      "learning_rate": 5.421743282044637e-06,
+      "loss": 0.4884,
+      "step": 7042
+    },
+    {
+      "epoch": 0.48902930148590473,
+      "grad_norm": 4.510938864107892,
+      "learning_rate": 5.420622791409022e-06,
+      "loss": 0.4178,
+      "step": 7043
+    },
+    {
+      "epoch": 0.48909873628662687,
+      "grad_norm": 3.827014090785199,
+      "learning_rate": 5.419502279498784e-06,
+      "loss": 0.397,
+      "step": 7044
+    },
+    {
+      "epoch": 0.489168171087349,
+      "grad_norm": 3.1468740688734402,
+      "learning_rate": 5.418381746370593e-06,
+      "loss": 0.4166,
+      "step": 7045
+    },
+    {
+      "epoch": 0.4892376058880711,
+      "grad_norm": 3.355730729530613,
+      "learning_rate": 5.417261192081128e-06,
+      "loss": 0.5955,
+      "step": 7046
+    },
+    {
+      "epoch": 0.4893070406887932,
+      "grad_norm": 4.179213487328587,
+      "learning_rate": 5.4161406166870645e-06,
+      "loss": 0.587,
+      "step": 7047
+    },
+    {
+      "epoch": 0.48937647548951535,
+      "grad_norm": 5.122705864014512,
+      "learning_rate": 5.415020020245079e-06,
+      "loss": 0.3448,
+      "step": 7048
+    },
+    {
+      "epoch": 0.4894459102902375,
+      "grad_norm": 4.599408780890138,
+      "learning_rate": 5.41389940281185e-06,
+      "loss": 0.5739,
+      "step": 7049
+    },
+    {
+      "epoch": 0.48951534509095956,
+      "grad_norm": 3.823790023557841,
+      "learning_rate": 5.412778764444058e-06,
+      "loss": 0.4686,
+      "step": 7050
+    },
+    {
+      "epoch": 0.4895847798916817,
+      "grad_norm": 4.289588729872633,
+      "learning_rate": 5.411658105198384e-06,
+      "loss": 0.5063,
+      "step": 7051
+    },
+    {
+      "epoch": 0.48965421469240383,
+      "grad_norm": 4.817521982435509,
+      "learning_rate": 5.410537425131507e-06,
+      "loss": 0.668,
+      "step": 7052
+    },
+    {
+      "epoch": 0.48972364949312597,
+      "grad_norm": 4.310958725884966,
+      "learning_rate": 5.409416724300113e-06,
+      "loss": 0.6766,
+      "step": 7053
+    },
+    {
+      "epoch": 0.4897930842938481,
+      "grad_norm": 4.1585112975107466,
+      "learning_rate": 5.408296002760885e-06,
+      "loss": 0.3002,
+      "step": 7054
+    },
+    {
+      "epoch": 0.4898625190945702,
+      "grad_norm": 4.4878297730743295,
+      "learning_rate": 5.407175260570506e-06,
+      "loss": 0.549,
+      "step": 7055
+    },
+    {
+      "epoch": 0.4899319538952923,
+      "grad_norm": 3.714329462055869,
+      "learning_rate": 5.406054497785662e-06,
+      "loss": 0.525,
+      "step": 7056
+    },
+    {
+      "epoch": 0.49000138869601445,
+      "grad_norm": 3.667374424833374,
+      "learning_rate": 5.404933714463043e-06,
+      "loss": 0.3344,
+      "step": 7057
+    },
+    {
+      "epoch": 0.4900708234967366,
+      "grad_norm": 3.936808443607699,
+      "learning_rate": 5.403812910659335e-06,
+      "loss": 0.4177,
+      "step": 7058
+    },
+    {
+      "epoch": 0.49014025829745866,
+      "grad_norm": 4.622057090956318,
+      "learning_rate": 5.402692086431225e-06,
+      "loss": 0.777,
+      "step": 7059
+    },
+    {
+      "epoch": 0.4902096930981808,
+      "grad_norm": 3.7568078116979398,
+      "learning_rate": 5.401571241835407e-06,
+      "loss": 0.4949,
+      "step": 7060
+    },
+    {
+      "epoch": 0.49027912789890293,
+      "grad_norm": 3.503833511660692,
+      "learning_rate": 5.400450376928569e-06,
+      "loss": 0.404,
+      "step": 7061
+    },
+    {
+      "epoch": 0.49034856269962507,
+      "grad_norm": 3.8503778219293006,
+      "learning_rate": 5.399329491767405e-06,
+      "loss": 0.5039,
+      "step": 7062
+    },
+    {
+      "epoch": 0.49041799750034715,
+      "grad_norm": 3.889654729978222,
+      "learning_rate": 5.398208586408605e-06,
+      "loss": 0.2657,
+      "step": 7063
+    },
+    {
+      "epoch": 0.4904874323010693,
+      "grad_norm": 4.136377780270438,
+      "learning_rate": 5.397087660908867e-06,
+      "loss": 0.2129,
+      "step": 7064
+    },
+    {
+      "epoch": 0.4905568671017914,
+      "grad_norm": 3.196631418474297,
+      "learning_rate": 5.395966715324883e-06,
+      "loss": 0.2706,
+      "step": 7065
+    },
+    {
+      "epoch": 0.49062630190251355,
+      "grad_norm": 3.843932926563781,
+      "learning_rate": 5.394845749713352e-06,
+      "loss": 0.6307,
+      "step": 7066
+    },
+    {
+      "epoch": 0.4906957367032357,
+      "grad_norm": 3.0776744817711017,
+      "learning_rate": 5.393724764130971e-06,
+      "loss": 0.5121,
+      "step": 7067
+    },
+    {
+      "epoch": 0.49076517150395776,
+      "grad_norm": 2.7512322894592893,
+      "learning_rate": 5.392603758634435e-06,
+      "loss": 0.1882,
+      "step": 7068
+    },
+    {
+      "epoch": 0.4908346063046799,
+      "grad_norm": 4.403634363135146,
+      "learning_rate": 5.391482733280446e-06,
+      "loss": 0.3319,
+      "step": 7069
+    },
+    {
+      "epoch": 0.49090404110540203,
+      "grad_norm": 3.0689038484707956,
+      "learning_rate": 5.390361688125702e-06,
+      "loss": 0.3281,
+      "step": 7070
+    },
+    {
+      "epoch": 0.49097347590612417,
+      "grad_norm": 3.626154411725304,
+      "learning_rate": 5.389240623226907e-06,
+      "loss": 0.3713,
+      "step": 7071
+    },
+    {
+      "epoch": 0.49104291070684625,
+      "grad_norm": 3.377471853637553,
+      "learning_rate": 5.388119538640761e-06,
+      "loss": 0.4627,
+      "step": 7072
+    },
+    {
+      "epoch": 0.4911123455075684,
+      "grad_norm": 2.0119859117946515,
+      "learning_rate": 5.386998434423969e-06,
+      "loss": 0.1082,
+      "step": 7073
+    },
+    {
+      "epoch": 0.4911817803082905,
+      "grad_norm": 3.4463036968993155,
+      "learning_rate": 5.385877310633233e-06,
+      "loss": 0.2909,
+      "step": 7074
+    },
+    {
+      "epoch": 0.49125121510901265,
+      "grad_norm": 4.524175940638843,
+      "learning_rate": 5.384756167325259e-06,
+      "loss": 0.7222,
+      "step": 7075
+    },
+    {
+      "epoch": 0.4913206499097348,
+      "grad_norm": 5.814402119204693,
+      "learning_rate": 5.383635004556754e-06,
+      "loss": 0.8433,
+      "step": 7076
+    },
+    {
+      "epoch": 0.49139008471045686,
+      "grad_norm": 2.670521714896633,
+      "learning_rate": 5.382513822384424e-06,
+      "loss": 0.2156,
+      "step": 7077
+    },
+    {
+      "epoch": 0.491459519511179,
+      "grad_norm": 3.6336452436456557,
+      "learning_rate": 5.381392620864979e-06,
+      "loss": 0.4604,
+      "step": 7078
+    },
+    {
+      "epoch": 0.49152895431190113,
+      "grad_norm": 3.3957624242968216,
+      "learning_rate": 5.380271400055126e-06,
+      "loss": 0.5248,
+      "step": 7079
+    },
+    {
+      "epoch": 0.49159838911262327,
+      "grad_norm": 4.21256576472506,
+      "learning_rate": 5.379150160011576e-06,
+      "loss": 0.5441,
+      "step": 7080
+    },
+    {
+      "epoch": 0.49166782391334535,
+      "grad_norm": 4.19586458485088,
+      "learning_rate": 5.378028900791039e-06,
+      "loss": 0.4626,
+      "step": 7081
+    },
+    {
+      "epoch": 0.4917372587140675,
+      "grad_norm": 4.064079229713481,
+      "learning_rate": 5.376907622450229e-06,
+      "loss": 0.6011,
+      "step": 7082
+    },
+    {
+      "epoch": 0.4918066935147896,
+      "grad_norm": 4.355155217395831,
+      "learning_rate": 5.37578632504586e-06,
+      "loss": 0.5049,
+      "step": 7083
+    },
+    {
+      "epoch": 0.49187612831551175,
+      "grad_norm": 3.170112863918857,
+      "learning_rate": 5.37466500863464e-06,
+      "loss": 0.2759,
+      "step": 7084
+    },
+    {
+      "epoch": 0.49194556311623383,
+      "grad_norm": 3.6284582460962804,
+      "learning_rate": 5.373543673273289e-06,
+      "loss": 0.5731,
+      "step": 7085
+    },
+    {
+      "epoch": 0.49201499791695597,
+      "grad_norm": 3.360994326288128,
+      "learning_rate": 5.3724223190185235e-06,
+      "loss": 0.3512,
+      "step": 7086
+    },
+    {
+      "epoch": 0.4920844327176781,
+      "grad_norm": 3.1923036025832463,
+      "learning_rate": 5.371300945927057e-06,
+      "loss": 0.4018,
+      "step": 7087
+    },
+    {
+      "epoch": 0.49215386751840023,
+      "grad_norm": 4.543802119796461,
+      "learning_rate": 5.370179554055608e-06,
+      "loss": 0.3276,
+      "step": 7088
+    },
+    {
+      "epoch": 0.49222330231912237,
+      "grad_norm": 3.5772570153764645,
+      "learning_rate": 5.369058143460899e-06,
+      "loss": 0.4548,
+      "step": 7089
+    },
+    {
+      "epoch": 0.49229273711984445,
+      "grad_norm": 3.836663798380964,
+      "learning_rate": 5.367936714199643e-06,
+      "loss": 0.5101,
+      "step": 7090
+    },
+    {
+      "epoch": 0.4923621719205666,
+      "grad_norm": 3.4147183342396357,
+      "learning_rate": 5.366815266328567e-06,
+      "loss": 0.359,
+      "step": 7091
+    },
+    {
+      "epoch": 0.4924316067212887,
+      "grad_norm": 3.718788778802708,
+      "learning_rate": 5.365693799904389e-06,
+      "loss": 0.5331,
+      "step": 7092
+    },
+    {
+      "epoch": 0.49250104152201085,
+      "grad_norm": 3.82234695435734,
+      "learning_rate": 5.364572314983833e-06,
+      "loss": 0.2334,
+      "step": 7093
+    },
+    {
+      "epoch": 0.49257047632273293,
+      "grad_norm": 3.6650577667817386,
+      "learning_rate": 5.363450811623622e-06,
+      "loss": 0.5805,
+      "step": 7094
+    },
+    {
+      "epoch": 0.49263991112345507,
+      "grad_norm": 3.2618651506345344,
+      "learning_rate": 5.362329289880479e-06,
+      "loss": 0.4341,
+      "step": 7095
+    },
+    {
+      "epoch": 0.4927093459241772,
+      "grad_norm": 3.7045229958678574,
+      "learning_rate": 5.361207749811131e-06,
+      "loss": 0.4515,
+      "step": 7096
+    },
+    {
+      "epoch": 0.49277878072489933,
+      "grad_norm": 3.793494209936031,
+      "learning_rate": 5.360086191472303e-06,
+      "loss": 0.3455,
+      "step": 7097
+    },
+    {
+      "epoch": 0.49284821552562147,
+      "grad_norm": 4.543797999111898,
+      "learning_rate": 5.358964614920724e-06,
+      "loss": 0.5888,
+      "step": 7098
+    },
+    {
+      "epoch": 0.49291765032634355,
+      "grad_norm": 3.526633109201806,
+      "learning_rate": 5.3578430202131215e-06,
+      "loss": 0.4215,
+      "step": 7099
+    },
+    {
+      "epoch": 0.4929870851270657,
+      "grad_norm": 3.4893156897797626,
+      "learning_rate": 5.356721407406223e-06,
+      "loss": 0.4875,
+      "step": 7100
+    },
+    {
+      "epoch": 0.4930565199277878,
+      "grad_norm": 3.5379761901320665,
+      "learning_rate": 5.355599776556759e-06,
+      "loss": 0.4249,
+      "step": 7101
+    },
+    {
+      "epoch": 0.49312595472850995,
+      "grad_norm": 4.2541936107763725,
+      "learning_rate": 5.354478127721462e-06,
+      "loss": 0.3551,
+      "step": 7102
+    },
+    {
+      "epoch": 0.49319538952923203,
+      "grad_norm": 3.8310424033390107,
+      "learning_rate": 5.353356460957062e-06,
+      "loss": 0.3181,
+      "step": 7103
+    },
+    {
+      "epoch": 0.49326482432995417,
+      "grad_norm": 3.7087785114890433,
+      "learning_rate": 5.352234776320291e-06,
+      "loss": 0.5541,
+      "step": 7104
+    },
+    {
+      "epoch": 0.4933342591306763,
+      "grad_norm": 4.4491572652855975,
+      "learning_rate": 5.351113073867885e-06,
+      "loss": 0.4496,
+      "step": 7105
+    },
+    {
+      "epoch": 0.49340369393139843,
+      "grad_norm": 3.084191534893803,
+      "learning_rate": 5.349991353656575e-06,
+      "loss": 0.2512,
+      "step": 7106
+    },
+    {
+      "epoch": 0.4934731287321205,
+      "grad_norm": 3.35212575027703,
+      "learning_rate": 5.3488696157431005e-06,
+      "loss": 0.4875,
+      "step": 7107
+    },
+    {
+      "epoch": 0.49354256353284265,
+      "grad_norm": 4.689501310777951,
+      "learning_rate": 5.347747860184195e-06,
+      "loss": 0.5149,
+      "step": 7108
+    },
+    {
+      "epoch": 0.4936119983335648,
+      "grad_norm": 2.603225165651721,
+      "learning_rate": 5.346626087036596e-06,
+      "loss": 0.272,
+      "step": 7109
+    },
+    {
+      "epoch": 0.4936814331342869,
+      "grad_norm": 3.101240869300462,
+      "learning_rate": 5.345504296357042e-06,
+      "loss": 0.3345,
+      "step": 7110
+    },
+    {
+      "epoch": 0.49375086793500905,
+      "grad_norm": 3.1128146752482295,
+      "learning_rate": 5.344382488202271e-06,
+      "loss": 0.4328,
+      "step": 7111
+    },
+    {
+      "epoch": 0.49382030273573113,
+      "grad_norm": 3.3555256220325975,
+      "learning_rate": 5.343260662629025e-06,
+      "loss": 0.3468,
+      "step": 7112
+    },
+    {
+      "epoch": 0.49388973753645327,
+      "grad_norm": 3.8189806418256564,
+      "learning_rate": 5.3421388196940405e-06,
+      "loss": 0.534,
+      "step": 7113
+    },
+    {
+      "epoch": 0.4939591723371754,
+      "grad_norm": 3.120467300028105,
+      "learning_rate": 5.341016959454064e-06,
+      "loss": 0.422,
+      "step": 7114
+    },
+    {
+      "epoch": 0.49402860713789754,
+      "grad_norm": 6.044010655463055,
+      "learning_rate": 5.339895081965835e-06,
+      "loss": 0.7997,
+      "step": 7115
+    },
+    {
+      "epoch": 0.4940980419386196,
+      "grad_norm": 4.056277184607958,
+      "learning_rate": 5.338773187286097e-06,
+      "loss": 0.4908,
+      "step": 7116
+    },
+    {
+      "epoch": 0.49416747673934175,
+      "grad_norm": 3.5275031696452324,
+      "learning_rate": 5.337651275471595e-06,
+      "loss": 0.26,
+      "step": 7117
+    },
+    {
+      "epoch": 0.4942369115400639,
+      "grad_norm": 2.2069894139170403,
+      "learning_rate": 5.336529346579073e-06,
+      "loss": 0.1885,
+      "step": 7118
+    },
+    {
+      "epoch": 0.494306346340786,
+      "grad_norm": 4.580578305307517,
+      "learning_rate": 5.335407400665278e-06,
+      "loss": 0.5559,
+      "step": 7119
+    },
+    {
+      "epoch": 0.4943757811415081,
+      "grad_norm": 4.291592562011333,
+      "learning_rate": 5.334285437786955e-06,
+      "loss": 0.5441,
+      "step": 7120
+    },
+    {
+      "epoch": 0.49444521594223023,
+      "grad_norm": 3.396321083622744,
+      "learning_rate": 5.333163458000854e-06,
+      "loss": 0.4067,
+      "step": 7121
+    },
+    {
+      "epoch": 0.49451465074295237,
+      "grad_norm": 3.610393560243008,
+      "learning_rate": 5.332041461363721e-06,
+      "loss": 0.5126,
+      "step": 7122
+    },
+    {
+      "epoch": 0.4945840855436745,
+      "grad_norm": 4.240141452919585,
+      "learning_rate": 5.330919447932309e-06,
+      "loss": 0.544,
+      "step": 7123
+    },
+    {
+      "epoch": 0.49465352034439664,
+      "grad_norm": 2.9656862054456883,
+      "learning_rate": 5.329797417763365e-06,
+      "loss": 0.3652,
+      "step": 7124
+    },
+    {
+      "epoch": 0.4947229551451187,
+      "grad_norm": 2.9987462203281803,
+      "learning_rate": 5.32867537091364e-06,
+      "loss": 0.2055,
+      "step": 7125
+    },
+    {
+      "epoch": 0.49479238994584085,
+      "grad_norm": 3.255309700091123,
+      "learning_rate": 5.327553307439887e-06,
+      "loss": 0.3632,
+      "step": 7126
+    },
+    {
+      "epoch": 0.494861824746563,
+      "grad_norm": 3.8199439769958374,
+      "learning_rate": 5.326431227398857e-06,
+      "loss": 0.5408,
+      "step": 7127
+    },
+    {
+      "epoch": 0.4949312595472851,
+      "grad_norm": 5.576465449820448,
+      "learning_rate": 5.325309130847307e-06,
+      "loss": 0.3894,
+      "step": 7128
+    },
+    {
+      "epoch": 0.4950006943480072,
+      "grad_norm": 3.799649889176957,
+      "learning_rate": 5.3241870178419885e-06,
+      "loss": 0.4203,
+      "step": 7129
+    },
+    {
+      "epoch": 0.49507012914872933,
+      "grad_norm": 4.3278162787920085,
+      "learning_rate": 5.323064888439657e-06,
+      "loss": 0.6124,
+      "step": 7130
+    },
+    {
+      "epoch": 0.49513956394945147,
+      "grad_norm": 4.242917064519879,
+      "learning_rate": 5.321942742697069e-06,
+      "loss": 0.5528,
+      "step": 7131
+    },
+    {
+      "epoch": 0.4952089987501736,
+      "grad_norm": 4.507205001983563,
+      "learning_rate": 5.320820580670982e-06,
+      "loss": 0.7479,
+      "step": 7132
+    },
+    {
+      "epoch": 0.49527843355089574,
+      "grad_norm": 3.8092447295342855,
+      "learning_rate": 5.319698402418154e-06,
+      "loss": 0.4543,
+      "step": 7133
+    },
+    {
+      "epoch": 0.4953478683516178,
+      "grad_norm": 3.7839617013593854,
+      "learning_rate": 5.318576207995342e-06,
+      "loss": 0.5772,
+      "step": 7134
+    },
+    {
+      "epoch": 0.49541730315233995,
+      "grad_norm": 4.512963681277308,
+      "learning_rate": 5.3174539974593055e-06,
+      "loss": 0.6533,
+      "step": 7135
+    },
+    {
+      "epoch": 0.4954867379530621,
+      "grad_norm": 3.75318013876013,
+      "learning_rate": 5.316331770866803e-06,
+      "loss": 0.3972,
+      "step": 7136
+    },
+    {
+      "epoch": 0.4955561727537842,
+      "grad_norm": 3.1119616258028784,
+      "learning_rate": 5.3152095282746e-06,
+      "loss": 0.2738,
+      "step": 7137
+    },
+    {
+      "epoch": 0.4956256075545063,
+      "grad_norm": 3.460554031361484,
+      "learning_rate": 5.3140872697394555e-06,
+      "loss": 0.4003,
+      "step": 7138
+    },
+    {
+      "epoch": 0.49569504235522843,
+      "grad_norm": 4.974105861641772,
+      "learning_rate": 5.312964995318132e-06,
+      "loss": 0.7448,
+      "step": 7139
+    },
+    {
+      "epoch": 0.49576447715595057,
+      "grad_norm": 5.708659071362712,
+      "learning_rate": 5.311842705067392e-06,
+      "loss": 0.2668,
+      "step": 7140
+    },
+    {
+      "epoch": 0.4958339119566727,
+      "grad_norm": 3.8930872092586766,
+      "learning_rate": 5.310720399044002e-06,
+      "loss": 0.6422,
+      "step": 7141
+    },
+    {
+      "epoch": 0.4959033467573948,
+      "grad_norm": 4.108552338881843,
+      "learning_rate": 5.3095980773047265e-06,
+      "loss": 0.4942,
+      "step": 7142
+    },
+    {
+      "epoch": 0.4959727815581169,
+      "grad_norm": 3.763301652610291,
+      "learning_rate": 5.308475739906329e-06,
+      "loss": 0.4892,
+      "step": 7143
+    },
+    {
+      "epoch": 0.49604221635883905,
+      "grad_norm": 3.4021805320908554,
+      "learning_rate": 5.307353386905579e-06,
+      "loss": 0.455,
+      "step": 7144
+    },
+    {
+      "epoch": 0.4961116511595612,
+      "grad_norm": 2.520922171228968,
+      "learning_rate": 5.306231018359242e-06,
+      "loss": 0.2056,
+      "step": 7145
+    },
+    {
+      "epoch": 0.4961810859602833,
+      "grad_norm": 3.8460211549509062,
+      "learning_rate": 5.305108634324087e-06,
+      "loss": 0.4752,
+      "step": 7146
+    },
+    {
+      "epoch": 0.4962505207610054,
+      "grad_norm": 2.8812511183880996,
+      "learning_rate": 5.303986234856882e-06,
+      "loss": 0.262,
+      "step": 7147
+    },
+    {
+      "epoch": 0.49631995556172753,
+      "grad_norm": 3.190455172760793,
+      "learning_rate": 5.302863820014396e-06,
+      "loss": 0.3286,
+      "step": 7148
+    },
+    {
+      "epoch": 0.49638939036244967,
+      "grad_norm": 4.792910080450556,
+      "learning_rate": 5.301741389853403e-06,
+      "loss": 0.6328,
+      "step": 7149
+    },
+    {
+      "epoch": 0.4964588251631718,
+      "grad_norm": 4.752682503127999,
+      "learning_rate": 5.300618944430671e-06,
+      "loss": 0.86,
+      "step": 7150
+    },
+    {
+      "epoch": 0.4965282599638939,
+      "grad_norm": 7.409280634904407,
+      "learning_rate": 5.2994964838029736e-06,
+      "loss": 0.3251,
+      "step": 7151
+    },
+    {
+      "epoch": 0.496597694764616,
+      "grad_norm": 3.8859552868736054,
+      "learning_rate": 5.298374008027082e-06,
+      "loss": 0.4129,
+      "step": 7152
+    },
+    {
+      "epoch": 0.49666712956533815,
+      "grad_norm": 3.321691512534654,
+      "learning_rate": 5.29725151715977e-06,
+      "loss": 0.1707,
+      "step": 7153
+    },
+    {
+      "epoch": 0.4967365643660603,
+      "grad_norm": 4.804087252609198,
+      "learning_rate": 5.296129011257814e-06,
+      "loss": 0.6742,
+      "step": 7154
+    },
+    {
+      "epoch": 0.49680599916678236,
+      "grad_norm": 4.369310117087779,
+      "learning_rate": 5.295006490377987e-06,
+      "loss": 0.6106,
+      "step": 7155
+    },
+    {
+      "epoch": 0.4968754339675045,
+      "grad_norm": 4.554791532773279,
+      "learning_rate": 5.293883954577066e-06,
+      "loss": 0.5304,
+      "step": 7156
+    },
+    {
+      "epoch": 0.49694486876822663,
+      "grad_norm": 4.020452555715454,
+      "learning_rate": 5.292761403911828e-06,
+      "loss": 0.4763,
+      "step": 7157
+    },
+    {
+      "epoch": 0.49701430356894877,
+      "grad_norm": 4.341033573823346,
+      "learning_rate": 5.2916388384390485e-06,
+      "loss": 0.4591,
+      "step": 7158
+    },
+    {
+      "epoch": 0.4970837383696709,
+      "grad_norm": 3.55171904675619,
+      "learning_rate": 5.290516258215505e-06,
+      "loss": 0.4488,
+      "step": 7159
+    },
+    {
+      "epoch": 0.497153173170393,
+      "grad_norm": 5.6200575089314055,
+      "learning_rate": 5.289393663297981e-06,
+      "loss": 0.6136,
+      "step": 7160
+    },
+    {
+      "epoch": 0.4972226079711151,
+      "grad_norm": 3.4192579088732624,
+      "learning_rate": 5.28827105374325e-06,
+      "loss": 0.3172,
+      "step": 7161
+    },
+    {
+      "epoch": 0.49729204277183725,
+      "grad_norm": 3.4421862406591974,
+      "learning_rate": 5.287148429608098e-06,
+      "loss": 0.4058,
+      "step": 7162
+    },
+    {
+      "epoch": 0.4973614775725594,
+      "grad_norm": 3.494628996611738,
+      "learning_rate": 5.286025790949302e-06,
+      "loss": 0.3734,
+      "step": 7163
+    },
+    {
+      "epoch": 0.49743091237328146,
+      "grad_norm": 3.8188799078379594,
+      "learning_rate": 5.2849031378236445e-06,
+      "loss": 0.5037,
+      "step": 7164
+    },
+    {
+      "epoch": 0.4975003471740036,
+      "grad_norm": 3.5239823969836026,
+      "learning_rate": 5.2837804702879105e-06,
+      "loss": 0.3853,
+      "step": 7165
+    },
+    {
+      "epoch": 0.49756978197472573,
+      "grad_norm": 4.128526274174642,
+      "learning_rate": 5.2826577883988805e-06,
+      "loss": 0.5364,
+      "step": 7166
+    },
+    {
+      "epoch": 0.49763921677544787,
+      "grad_norm": 4.125547750480884,
+      "learning_rate": 5.281535092213339e-06,
+      "loss": 0.3785,
+      "step": 7167
+    },
+    {
+      "epoch": 0.49770865157617,
+      "grad_norm": 3.924061951826711,
+      "learning_rate": 5.28041238178807e-06,
+      "loss": 0.5469,
+      "step": 7168
+    },
+    {
+      "epoch": 0.4977780863768921,
+      "grad_norm": 4.769729360190772,
+      "learning_rate": 5.279289657179862e-06,
+      "loss": 0.7011,
+      "step": 7169
+    },
+    {
+      "epoch": 0.4978475211776142,
+      "grad_norm": 4.728584667652929,
+      "learning_rate": 5.278166918445499e-06,
+      "loss": 0.5105,
+      "step": 7170
+    },
+    {
+      "epoch": 0.49791695597833635,
+      "grad_norm": 3.3970712513989882,
+      "learning_rate": 5.277044165641767e-06,
+      "loss": 0.3371,
+      "step": 7171
+    },
+    {
+      "epoch": 0.4979863907790585,
+      "grad_norm": 4.008725301555456,
+      "learning_rate": 5.2759213988254544e-06,
+      "loss": 0.5655,
+      "step": 7172
+    },
+    {
+      "epoch": 0.49805582557978056,
+      "grad_norm": 6.875714919516217,
+      "learning_rate": 5.274798618053352e-06,
+      "loss": 0.9303,
+      "step": 7173
+    },
+    {
+      "epoch": 0.4981252603805027,
+      "grad_norm": 3.8747047293998595,
+      "learning_rate": 5.273675823382245e-06,
+      "loss": 0.4319,
+      "step": 7174
+    },
+    {
+      "epoch": 0.49819469518122483,
+      "grad_norm": 2.9699327859739992,
+      "learning_rate": 5.272553014868924e-06,
+      "loss": 0.3657,
+      "step": 7175
+    },
+    {
+      "epoch": 0.49826412998194697,
+      "grad_norm": 2.371481669411635,
+      "learning_rate": 5.271430192570181e-06,
+      "loss": 0.246,
+      "step": 7176
+    },
+    {
+      "epoch": 0.49833356478266905,
+      "grad_norm": 4.078410693745689,
+      "learning_rate": 5.270307356542805e-06,
+      "loss": 0.4913,
+      "step": 7177
+    },
+    {
+      "epoch": 0.4984029995833912,
+      "grad_norm": 4.309350950486881,
+      "learning_rate": 5.269184506843589e-06,
+      "loss": 0.6337,
+      "step": 7178
+    },
+    {
+      "epoch": 0.4984724343841133,
+      "grad_norm": 4.594534756780209,
+      "learning_rate": 5.268061643529326e-06,
+      "loss": 0.5938,
+      "step": 7179
+    },
+    {
+      "epoch": 0.49854186918483545,
+      "grad_norm": 3.998845901976261,
+      "learning_rate": 5.266938766656807e-06,
+      "loss": 0.4451,
+      "step": 7180
+    },
+    {
+      "epoch": 0.4986113039855576,
+      "grad_norm": 3.2176590880594107,
+      "learning_rate": 5.265815876282829e-06,
+      "loss": 0.3404,
+      "step": 7181
+    },
+    {
+      "epoch": 0.49868073878627966,
+      "grad_norm": 3.6939987142670985,
+      "learning_rate": 5.264692972464183e-06,
+      "loss": 0.5014,
+      "step": 7182
+    },
+    {
+      "epoch": 0.4987501735870018,
+      "grad_norm": 4.129135520229409,
+      "learning_rate": 5.263570055257668e-06,
+      "loss": 0.6594,
+      "step": 7183
+    },
+    {
+      "epoch": 0.49881960838772393,
+      "grad_norm": 2.9283959666625576,
+      "learning_rate": 5.262447124720076e-06,
+      "loss": 0.2899,
+      "step": 7184
+    },
+    {
+      "epoch": 0.49888904318844607,
+      "grad_norm": 4.875773599050903,
+      "learning_rate": 5.2613241809082065e-06,
+      "loss": 0.5409,
+      "step": 7185
+    },
+    {
+      "epoch": 0.49895847798916815,
+      "grad_norm": 3.793237469910889,
+      "learning_rate": 5.260201223878855e-06,
+      "loss": 0.5597,
+      "step": 7186
+    },
+    {
+      "epoch": 0.4990279127898903,
+      "grad_norm": 2.80112940842597,
+      "learning_rate": 5.25907825368882e-06,
+      "loss": 0.1872,
+      "step": 7187
+    },
+    {
+      "epoch": 0.4990973475906124,
+      "grad_norm": 3.5622148434070793,
+      "learning_rate": 5.2579552703949e-06,
+      "loss": 0.3903,
+      "step": 7188
+    },
+    {
+      "epoch": 0.49916678239133455,
+      "grad_norm": 3.6302453123686633,
+      "learning_rate": 5.256832274053896e-06,
+      "loss": 0.5845,
+      "step": 7189
+    },
+    {
+      "epoch": 0.4992362171920567,
+      "grad_norm": 3.8159433823095594,
+      "learning_rate": 5.255709264722605e-06,
+      "loss": 0.4908,
+      "step": 7190
+    },
+    {
+      "epoch": 0.49930565199277877,
+      "grad_norm": 3.935442865805352,
+      "learning_rate": 5.2545862424578276e-06,
+      "loss": 0.6225,
+      "step": 7191
+    },
+    {
+      "epoch": 0.4993750867935009,
+      "grad_norm": 3.251944791967001,
+      "learning_rate": 5.2534632073163675e-06,
+      "loss": 0.3343,
+      "step": 7192
+    },
+    {
+      "epoch": 0.49944452159422303,
+      "grad_norm": 4.172297095128883,
+      "learning_rate": 5.252340159355024e-06,
+      "loss": 0.5531,
+      "step": 7193
+    },
+    {
+      "epoch": 0.49951395639494517,
+      "grad_norm": 4.600623832685623,
+      "learning_rate": 5.251217098630602e-06,
+      "loss": 0.6608,
+      "step": 7194
+    },
+    {
+      "epoch": 0.49958339119566725,
+      "grad_norm": 3.2144871832961814,
+      "learning_rate": 5.250094025199903e-06,
+      "loss": 0.3431,
+      "step": 7195
+    },
+    {
+      "epoch": 0.4996528259963894,
+      "grad_norm": 3.654094390946151,
+      "learning_rate": 5.248970939119732e-06,
+      "loss": 0.5365,
+      "step": 7196
+    },
+    {
+      "epoch": 0.4997222607971115,
+      "grad_norm": 2.3868768317250066,
+      "learning_rate": 5.24784784044689e-06,
+      "loss": 0.2038,
+      "step": 7197
+    },
+    {
+      "epoch": 0.49979169559783365,
+      "grad_norm": 2.9928772026534056,
+      "learning_rate": 5.246724729238188e-06,
+      "loss": 0.1364,
+      "step": 7198
+    },
+    {
+      "epoch": 0.49986113039855573,
+      "grad_norm": 1.9501663736045818,
+      "learning_rate": 5.2456016055504274e-06,
+      "loss": 0.2002,
+      "step": 7199
+    },
+    {
+      "epoch": 0.49993056519927787,
+      "grad_norm": 4.660104696791479,
+      "learning_rate": 5.244478469440414e-06,
+      "loss": 0.7036,
+      "step": 7200
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.746836592493186,
+      "learning_rate": 5.243355320964958e-06,
+      "loss": 0.4439,
+      "step": 7201
+    },
+    {
+      "epoch": 0.5000694348007221,
+      "grad_norm": 3.6720621338020467,
+      "learning_rate": 5.2422321601808646e-06,
+      "loss": 0.519,
+      "step": 7202
+    },
+    {
+      "epoch": 0.5001388696014443,
+      "grad_norm": 3.619176276475976,
+      "learning_rate": 5.241108987144943e-06,
+      "loss": 0.4642,
+      "step": 7203
+    },
+    {
+      "epoch": 0.5002083044021663,
+      "grad_norm": 3.477907487767762,
+      "learning_rate": 5.2399858019140005e-06,
+      "loss": 0.4744,
+      "step": 7204
+    },
+    {
+      "epoch": 0.5002777392028885,
+      "grad_norm": 3.4226791179143947,
+      "learning_rate": 5.238862604544849e-06,
+      "loss": 0.4601,
+      "step": 7205
+    },
+    {
+      "epoch": 0.5003471740036106,
+      "grad_norm": 4.212163114215492,
+      "learning_rate": 5.237739395094296e-06,
+      "loss": 0.5548,
+      "step": 7206
+    },
+    {
+      "epoch": 0.5004166088043327,
+      "grad_norm": 3.5845944353631563,
+      "learning_rate": 5.236616173619155e-06,
+      "loss": 0.2511,
+      "step": 7207
+    },
+    {
+      "epoch": 0.5004860436050549,
+      "grad_norm": 4.317463510068775,
+      "learning_rate": 5.235492940176235e-06,
+      "loss": 0.702,
+      "step": 7208
+    },
+    {
+      "epoch": 0.500555478405777,
+      "grad_norm": 3.5267036628826256,
+      "learning_rate": 5.234369694822348e-06,
+      "loss": 0.5953,
+      "step": 7209
+    },
+    {
+      "epoch": 0.500624913206499,
+      "grad_norm": 4.603833257260333,
+      "learning_rate": 5.233246437614308e-06,
+      "loss": 0.6984,
+      "step": 7210
+    },
+    {
+      "epoch": 0.5006943480072212,
+      "grad_norm": 3.119370751023977,
+      "learning_rate": 5.232123168608926e-06,
+      "loss": 0.2596,
+      "step": 7211
+    },
+    {
+      "epoch": 0.5007637828079433,
+      "grad_norm": 3.594355511152304,
+      "learning_rate": 5.230999887863018e-06,
+      "loss": 0.3693,
+      "step": 7212
+    },
+    {
+      "epoch": 0.5008332176086655,
+      "grad_norm": 4.278470400684505,
+      "learning_rate": 5.229876595433395e-06,
+      "loss": 0.5303,
+      "step": 7213
+    },
+    {
+      "epoch": 0.5009026524093876,
+      "grad_norm": 2.9327968403421445,
+      "learning_rate": 5.228753291376875e-06,
+      "loss": 0.3906,
+      "step": 7214
+    },
+    {
+      "epoch": 0.5009720872101097,
+      "grad_norm": 3.0158133264807714,
+      "learning_rate": 5.227629975750274e-06,
+      "loss": 0.3153,
+      "step": 7215
+    },
+    {
+      "epoch": 0.5010415220108319,
+      "grad_norm": 3.940323533922246,
+      "learning_rate": 5.226506648610404e-06,
+      "loss": 0.2882,
+      "step": 7216
+    },
+    {
+      "epoch": 0.5011109568115539,
+      "grad_norm": 3.527280735717722,
+      "learning_rate": 5.225383310014086e-06,
+      "loss": 0.4358,
+      "step": 7217
+    },
+    {
+      "epoch": 0.5011803916122761,
+      "grad_norm": 3.543320125475908,
+      "learning_rate": 5.224259960018135e-06,
+      "loss": 0.6069,
+      "step": 7218
+    },
+    {
+      "epoch": 0.5012498264129982,
+      "grad_norm": 3.725611661775222,
+      "learning_rate": 5.223136598679368e-06,
+      "loss": 0.4378,
+      "step": 7219
+    },
+    {
+      "epoch": 0.5013192612137203,
+      "grad_norm": 3.804259297246827,
+      "learning_rate": 5.2220132260546045e-06,
+      "loss": 0.5409,
+      "step": 7220
+    },
+    {
+      "epoch": 0.5013886960144425,
+      "grad_norm": 4.327521733094509,
+      "learning_rate": 5.2208898422006634e-06,
+      "loss": 0.4174,
+      "step": 7221
+    },
+    {
+      "epoch": 0.5014581308151645,
+      "grad_norm": 3.6486906051113133,
+      "learning_rate": 5.219766447174364e-06,
+      "loss": 0.3962,
+      "step": 7222
+    },
+    {
+      "epoch": 0.5015275656158866,
+      "grad_norm": 2.0114211985789847,
+      "learning_rate": 5.218643041032526e-06,
+      "loss": 0.1319,
+      "step": 7223
+    },
+    {
+      "epoch": 0.5015970004166088,
+      "grad_norm": 3.083387601193856,
+      "learning_rate": 5.217519623831971e-06,
+      "loss": 0.2501,
+      "step": 7224
+    },
+    {
+      "epoch": 0.5016664352173309,
+      "grad_norm": 2.98307443287388,
+      "learning_rate": 5.2163961956295196e-06,
+      "loss": 0.3029,
+      "step": 7225
+    },
+    {
+      "epoch": 0.5017358700180531,
+      "grad_norm": 3.767359754295796,
+      "learning_rate": 5.215272756481994e-06,
+      "loss": 0.397,
+      "step": 7226
+    },
+    {
+      "epoch": 0.5018053048187752,
+      "grad_norm": 3.7242593811855165,
+      "learning_rate": 5.214149306446215e-06,
+      "loss": 0.4136,
+      "step": 7227
+    },
+    {
+      "epoch": 0.5018747396194972,
+      "grad_norm": 3.5941509184210534,
+      "learning_rate": 5.2130258455790075e-06,
+      "loss": 0.5528,
+      "step": 7228
+    },
+    {
+      "epoch": 0.5019441744202194,
+      "grad_norm": 4.336523150469935,
+      "learning_rate": 5.211902373937192e-06,
+      "loss": 0.6431,
+      "step": 7229
+    },
+    {
+      "epoch": 0.5020136092209415,
+      "grad_norm": 4.142841145177535,
+      "learning_rate": 5.210778891577595e-06,
+      "loss": 0.5995,
+      "step": 7230
+    },
+    {
+      "epoch": 0.5020830440216637,
+      "grad_norm": 4.208289591673076,
+      "learning_rate": 5.209655398557041e-06,
+      "loss": 0.3495,
+      "step": 7231
+    },
+    {
+      "epoch": 0.5021524788223858,
+      "grad_norm": 4.3535542251548796,
+      "learning_rate": 5.208531894932354e-06,
+      "loss": 0.5908,
+      "step": 7232
+    },
+    {
+      "epoch": 0.5022219136231079,
+      "grad_norm": 3.3485461579942655,
+      "learning_rate": 5.207408380760358e-06,
+      "loss": 0.3773,
+      "step": 7233
+    },
+    {
+      "epoch": 0.50229134842383,
+      "grad_norm": 4.239984321277322,
+      "learning_rate": 5.206284856097883e-06,
+      "loss": 0.533,
+      "step": 7234
+    },
+    {
+      "epoch": 0.5023607832245521,
+      "grad_norm": 3.4166096039427867,
+      "learning_rate": 5.205161321001753e-06,
+      "loss": 0.5408,
+      "step": 7235
+    },
+    {
+      "epoch": 0.5024302180252743,
+      "grad_norm": 3.954747818849586,
+      "learning_rate": 5.204037775528794e-06,
+      "loss": 0.5229,
+      "step": 7236
+    },
+    {
+      "epoch": 0.5024996528259964,
+      "grad_norm": 3.9410199089888955,
+      "learning_rate": 5.202914219735836e-06,
+      "loss": 0.5692,
+      "step": 7237
+    },
+    {
+      "epoch": 0.5025690876267185,
+      "grad_norm": 3.972022698930772,
+      "learning_rate": 5.201790653679707e-06,
+      "loss": 0.4713,
+      "step": 7238
+    },
+    {
+      "epoch": 0.5026385224274407,
+      "grad_norm": 3.967305803979107,
+      "learning_rate": 5.200667077417234e-06,
+      "loss": 0.6258,
+      "step": 7239
+    },
+    {
+      "epoch": 0.5027079572281627,
+      "grad_norm": 5.519282879703435,
+      "learning_rate": 5.199543491005249e-06,
+      "loss": 0.7682,
+      "step": 7240
+    },
+    {
+      "epoch": 0.5027773920288848,
+      "grad_norm": 3.2535404542959037,
+      "learning_rate": 5.198419894500579e-06,
+      "loss": 0.3684,
+      "step": 7241
+    },
+    {
+      "epoch": 0.502846826829607,
+      "grad_norm": 3.3846768529165505,
+      "learning_rate": 5.197296287960055e-06,
+      "loss": 0.3128,
+      "step": 7242
+    },
+    {
+      "epoch": 0.5029162616303291,
+      "grad_norm": 4.429698524310986,
+      "learning_rate": 5.196172671440508e-06,
+      "loss": 0.502,
+      "step": 7243
+    },
+    {
+      "epoch": 0.5029856964310513,
+      "grad_norm": 2.7409830824998163,
+      "learning_rate": 5.19504904499877e-06,
+      "loss": 0.3191,
+      "step": 7244
+    },
+    {
+      "epoch": 0.5030551312317734,
+      "grad_norm": 4.208283805060563,
+      "learning_rate": 5.193925408691671e-06,
+      "loss": 0.7864,
+      "step": 7245
+    },
+    {
+      "epoch": 0.5031245660324954,
+      "grad_norm": 4.575571167096378,
+      "learning_rate": 5.192801762576045e-06,
+      "loss": 0.4382,
+      "step": 7246
+    },
+    {
+      "epoch": 0.5031940008332176,
+      "grad_norm": 4.7332623190695475,
+      "learning_rate": 5.191678106708724e-06,
+      "loss": 0.6272,
+      "step": 7247
+    },
+    {
+      "epoch": 0.5032634356339397,
+      "grad_norm": 4.878381439411061,
+      "learning_rate": 5.190554441146541e-06,
+      "loss": 0.5952,
+      "step": 7248
+    },
+    {
+      "epoch": 0.5033328704346619,
+      "grad_norm": 3.5523760281111154,
+      "learning_rate": 5.189430765946331e-06,
+      "loss": 0.5221,
+      "step": 7249
+    },
+    {
+      "epoch": 0.503402305235384,
+      "grad_norm": 2.836863366208607,
+      "learning_rate": 5.188307081164926e-06,
+      "loss": 0.3598,
+      "step": 7250
+    },
+    {
+      "epoch": 0.5034717400361061,
+      "grad_norm": 3.7142630293813452,
+      "learning_rate": 5.187183386859162e-06,
+      "loss": 0.329,
+      "step": 7251
+    },
+    {
+      "epoch": 0.5035411748368283,
+      "grad_norm": 2.6228456310122588,
+      "learning_rate": 5.186059683085874e-06,
+      "loss": 0.1799,
+      "step": 7252
+    },
+    {
+      "epoch": 0.5036106096375503,
+      "grad_norm": 3.7124971982354755,
+      "learning_rate": 5.1849359699018974e-06,
+      "loss": 0.3587,
+      "step": 7253
+    },
+    {
+      "epoch": 0.5036800444382724,
+      "grad_norm": 3.7997365969125294,
+      "learning_rate": 5.18381224736407e-06,
+      "loss": 0.598,
+      "step": 7254
+    },
+    {
+      "epoch": 0.5037494792389946,
+      "grad_norm": 3.2179499113229424,
+      "learning_rate": 5.182688515529227e-06,
+      "loss": 0.156,
+      "step": 7255
+    },
+    {
+      "epoch": 0.5038189140397167,
+      "grad_norm": 4.255157356847298,
+      "learning_rate": 5.181564774454205e-06,
+      "loss": 0.4395,
+      "step": 7256
+    },
+    {
+      "epoch": 0.5038883488404389,
+      "grad_norm": 2.768835391935664,
+      "learning_rate": 5.180441024195843e-06,
+      "loss": 0.1541,
+      "step": 7257
+    },
+    {
+      "epoch": 0.503957783641161,
+      "grad_norm": 4.988180931525735,
+      "learning_rate": 5.179317264810977e-06,
+      "loss": 0.7126,
+      "step": 7258
+    },
+    {
+      "epoch": 0.504027218441883,
+      "grad_norm": 3.634251859391279,
+      "learning_rate": 5.178193496356446e-06,
+      "loss": 0.4069,
+      "step": 7259
+    },
+    {
+      "epoch": 0.5040966532426052,
+      "grad_norm": 3.4521715364683736,
+      "learning_rate": 5.177069718889093e-06,
+      "loss": 0.5582,
+      "step": 7260
+    },
+    {
+      "epoch": 0.5041660880433273,
+      "grad_norm": 3.6005713469618175,
+      "learning_rate": 5.17594593246575e-06,
+      "loss": 0.4963,
+      "step": 7261
+    },
+    {
+      "epoch": 0.5042355228440495,
+      "grad_norm": 4.2469378512974005,
+      "learning_rate": 5.174822137143263e-06,
+      "loss": 0.4855,
+      "step": 7262
+    },
+    {
+      "epoch": 0.5043049576447716,
+      "grad_norm": 3.8001143595842377,
+      "learning_rate": 5.173698332978469e-06,
+      "loss": 0.3286,
+      "step": 7263
+    },
+    {
+      "epoch": 0.5043743924454936,
+      "grad_norm": 4.392261018707386,
+      "learning_rate": 5.1725745200282114e-06,
+      "loss": 0.564,
+      "step": 7264
+    },
+    {
+      "epoch": 0.5044438272462158,
+      "grad_norm": 3.7688785137422776,
+      "learning_rate": 5.171450698349329e-06,
+      "loss": 0.4837,
+      "step": 7265
+    },
+    {
+      "epoch": 0.5045132620469379,
+      "grad_norm": 4.346456270743364,
+      "learning_rate": 5.170326867998665e-06,
+      "loss": 0.5479,
+      "step": 7266
+    },
+    {
+      "epoch": 0.50458269684766,
+      "grad_norm": 2.2108039164863627,
+      "learning_rate": 5.16920302903306e-06,
+      "loss": 0.1792,
+      "step": 7267
+    },
+    {
+      "epoch": 0.5046521316483822,
+      "grad_norm": 3.140518888563056,
+      "learning_rate": 5.168079181509357e-06,
+      "loss": 0.3722,
+      "step": 7268
+    },
+    {
+      "epoch": 0.5047215664491043,
+      "grad_norm": 3.938158284869595,
+      "learning_rate": 5.166955325484398e-06,
+      "loss": 0.292,
+      "step": 7269
+    },
+    {
+      "epoch": 0.5047910012498265,
+      "grad_norm": 4.015870184255071,
+      "learning_rate": 5.165831461015031e-06,
+      "loss": 0.5239,
+      "step": 7270
+    },
+    {
+      "epoch": 0.5048604360505485,
+      "grad_norm": 2.780512823474123,
+      "learning_rate": 5.1647075881580935e-06,
+      "loss": 0.2994,
+      "step": 7271
+    },
+    {
+      "epoch": 0.5049298708512706,
+      "grad_norm": 2.833534239456496,
+      "learning_rate": 5.163583706970433e-06,
+      "loss": 0.2391,
+      "step": 7272
+    },
+    {
+      "epoch": 0.5049993056519928,
+      "grad_norm": 4.251828879688886,
+      "learning_rate": 5.162459817508896e-06,
+      "loss": 0.5862,
+      "step": 7273
+    },
+    {
+      "epoch": 0.5050687404527149,
+      "grad_norm": 3.613754087450469,
+      "learning_rate": 5.161335919830322e-06,
+      "loss": 0.4699,
+      "step": 7274
+    },
+    {
+      "epoch": 0.5051381752534371,
+      "grad_norm": 3.090862377531071,
+      "learning_rate": 5.160212013991561e-06,
+      "loss": 0.1662,
+      "step": 7275
+    },
+    {
+      "epoch": 0.5052076100541592,
+      "grad_norm": 3.139315243955183,
+      "learning_rate": 5.15908810004946e-06,
+      "loss": 0.3352,
+      "step": 7276
+    },
+    {
+      "epoch": 0.5052770448548812,
+      "grad_norm": 3.6782138058658727,
+      "learning_rate": 5.157964178060859e-06,
+      "loss": 0.215,
+      "step": 7277
+    },
+    {
+      "epoch": 0.5053464796556034,
+      "grad_norm": 4.07282405471761,
+      "learning_rate": 5.156840248082611e-06,
+      "loss": 0.7034,
+      "step": 7278
+    },
+    {
+      "epoch": 0.5054159144563255,
+      "grad_norm": 3.504236977706975,
+      "learning_rate": 5.15571631017156e-06,
+      "loss": 0.37,
+      "step": 7279
+    },
+    {
+      "epoch": 0.5054853492570476,
+      "grad_norm": 3.586813778622438,
+      "learning_rate": 5.154592364384554e-06,
+      "loss": 0.3322,
+      "step": 7280
+    },
+    {
+      "epoch": 0.5055547840577698,
+      "grad_norm": 2.3498077722842603,
+      "learning_rate": 5.153468410778441e-06,
+      "loss": 0.1749,
+      "step": 7281
+    },
+    {
+      "epoch": 0.5056242188584918,
+      "grad_norm": 5.903836002185436,
+      "learning_rate": 5.152344449410069e-06,
+      "loss": 0.765,
+      "step": 7282
+    },
+    {
+      "epoch": 0.505693653659214,
+      "grad_norm": 3.7105459873810984,
+      "learning_rate": 5.1512204803362895e-06,
+      "loss": 0.3497,
+      "step": 7283
+    },
+    {
+      "epoch": 0.5057630884599361,
+      "grad_norm": 4.523459835859084,
+      "learning_rate": 5.150096503613947e-06,
+      "loss": 0.7992,
+      "step": 7284
+    },
+    {
+      "epoch": 0.5058325232606582,
+      "grad_norm": 5.172702330291772,
+      "learning_rate": 5.1489725192998955e-06,
+      "loss": 0.6305,
+      "step": 7285
+    },
+    {
+      "epoch": 0.5059019580613804,
+      "grad_norm": 4.0215923493998496,
+      "learning_rate": 5.14784852745098e-06,
+      "loss": 0.3395,
+      "step": 7286
+    },
+    {
+      "epoch": 0.5059713928621025,
+      "grad_norm": 4.326317382938955,
+      "learning_rate": 5.146724528124055e-06,
+      "loss": 0.4767,
+      "step": 7287
+    },
+    {
+      "epoch": 0.5060408276628247,
+      "grad_norm": 3.785356305000666,
+      "learning_rate": 5.1456005213759674e-06,
+      "loss": 0.4784,
+      "step": 7288
+    },
+    {
+      "epoch": 0.5061102624635467,
+      "grad_norm": 3.213885380427425,
+      "learning_rate": 5.144476507263573e-06,
+      "loss": 0.4045,
+      "step": 7289
+    },
+    {
+      "epoch": 0.5061796972642688,
+      "grad_norm": 4.232182679619171,
+      "learning_rate": 5.143352485843719e-06,
+      "loss": 0.2863,
+      "step": 7290
+    },
+    {
+      "epoch": 0.506249132064991,
+      "grad_norm": 3.4622279350739253,
+      "learning_rate": 5.142228457173257e-06,
+      "loss": 0.5058,
+      "step": 7291
+    },
+    {
+      "epoch": 0.5063185668657131,
+      "grad_norm": 4.010560856621945,
+      "learning_rate": 5.141104421309043e-06,
+      "loss": 0.6036,
+      "step": 7292
+    },
+    {
+      "epoch": 0.5063880016664353,
+      "grad_norm": 4.454136884777085,
+      "learning_rate": 5.139980378307926e-06,
+      "loss": 0.5393,
+      "step": 7293
+    },
+    {
+      "epoch": 0.5064574364671574,
+      "grad_norm": 4.022361922257905,
+      "learning_rate": 5.138856328226759e-06,
+      "loss": 0.5726,
+      "step": 7294
+    },
+    {
+      "epoch": 0.5065268712678794,
+      "grad_norm": 6.854155959548227,
+      "learning_rate": 5.137732271122398e-06,
+      "loss": 0.4179,
+      "step": 7295
+    },
+    {
+      "epoch": 0.5065963060686016,
+      "grad_norm": 4.435266811393554,
+      "learning_rate": 5.1366082070516935e-06,
+      "loss": 0.4173,
+      "step": 7296
+    },
+    {
+      "epoch": 0.5066657408693237,
+      "grad_norm": 3.188885397761379,
+      "learning_rate": 5.135484136071499e-06,
+      "loss": 0.2022,
+      "step": 7297
+    },
+    {
+      "epoch": 0.5067351756700458,
+      "grad_norm": 3.9331959380564467,
+      "learning_rate": 5.134360058238672e-06,
+      "loss": 0.4341,
+      "step": 7298
+    },
+    {
+      "epoch": 0.506804610470768,
+      "grad_norm": 4.152511559922815,
+      "learning_rate": 5.1332359736100655e-06,
+      "loss": 0.4742,
+      "step": 7299
+    },
+    {
+      "epoch": 0.50687404527149,
+      "grad_norm": 4.397799624821824,
+      "learning_rate": 5.132111882242533e-06,
+      "loss": 0.6547,
+      "step": 7300
+    },
+    {
+      "epoch": 0.5069434800722122,
+      "grad_norm": 9.070957585781645,
+      "learning_rate": 5.130987784192932e-06,
+      "loss": 0.3703,
+      "step": 7301
+    },
+    {
+      "epoch": 0.5070129148729343,
+      "grad_norm": 4.2665091829176065,
+      "learning_rate": 5.129863679518116e-06,
+      "loss": 0.3536,
+      "step": 7302
+    },
+    {
+      "epoch": 0.5070823496736564,
+      "grad_norm": 4.678782200414299,
+      "learning_rate": 5.1287395682749444e-06,
+      "loss": 0.6875,
+      "step": 7303
+    },
+    {
+      "epoch": 0.5071517844743786,
+      "grad_norm": 3.6715605865168413,
+      "learning_rate": 5.127615450520268e-06,
+      "loss": 0.4513,
+      "step": 7304
+    },
+    {
+      "epoch": 0.5072212192751007,
+      "grad_norm": 3.7407373989866155,
+      "learning_rate": 5.126491326310949e-06,
+      "loss": 0.4072,
+      "step": 7305
+    },
+    {
+      "epoch": 0.5072906540758229,
+      "grad_norm": 2.8431098758028597,
+      "learning_rate": 5.125367195703841e-06,
+      "loss": 0.3618,
+      "step": 7306
+    },
+    {
+      "epoch": 0.5073600888765449,
+      "grad_norm": 5.255642422136762,
+      "learning_rate": 5.124243058755801e-06,
+      "loss": 0.5029,
+      "step": 7307
+    },
+    {
+      "epoch": 0.507429523677267,
+      "grad_norm": 5.0063359438651185,
+      "learning_rate": 5.123118915523689e-06,
+      "loss": 0.5127,
+      "step": 7308
+    },
+    {
+      "epoch": 0.5074989584779892,
+      "grad_norm": 3.7628705433561844,
+      "learning_rate": 5.121994766064361e-06,
+      "loss": 0.5409,
+      "step": 7309
+    },
+    {
+      "epoch": 0.5075683932787113,
+      "grad_norm": 4.046950010828195,
+      "learning_rate": 5.120870610434676e-06,
+      "loss": 0.6884,
+      "step": 7310
+    },
+    {
+      "epoch": 0.5076378280794334,
+      "grad_norm": 3.2299242185746277,
+      "learning_rate": 5.1197464486914925e-06,
+      "loss": 0.4736,
+      "step": 7311
+    },
+    {
+      "epoch": 0.5077072628801556,
+      "grad_norm": 3.3990792276475648,
+      "learning_rate": 5.11862228089167e-06,
+      "loss": 0.442,
+      "step": 7312
+    },
+    {
+      "epoch": 0.5077766976808776,
+      "grad_norm": 4.03911631318669,
+      "learning_rate": 5.117498107092064e-06,
+      "loss": 0.5153,
+      "step": 7313
+    },
+    {
+      "epoch": 0.5078461324815998,
+      "grad_norm": 4.619454303163214,
+      "learning_rate": 5.116373927349538e-06,
+      "loss": 0.5535,
+      "step": 7314
+    },
+    {
+      "epoch": 0.5079155672823219,
+      "grad_norm": 4.215895479172811,
+      "learning_rate": 5.115249741720953e-06,
+      "loss": 0.4981,
+      "step": 7315
+    },
+    {
+      "epoch": 0.507985002083044,
+      "grad_norm": 5.1518174910545245,
+      "learning_rate": 5.1141255502631626e-06,
+      "loss": 0.5276,
+      "step": 7316
+    },
+    {
+      "epoch": 0.5080544368837662,
+      "grad_norm": 3.639036208755218,
+      "learning_rate": 5.113001353033034e-06,
+      "loss": 0.4965,
+      "step": 7317
+    },
+    {
+      "epoch": 0.5081238716844882,
+      "grad_norm": 4.997077914934165,
+      "learning_rate": 5.111877150087423e-06,
+      "loss": 0.5622,
+      "step": 7318
+    },
+    {
+      "epoch": 0.5081933064852104,
+      "grad_norm": 3.3227340752957364,
+      "learning_rate": 5.110752941483192e-06,
+      "loss": 0.4569,
+      "step": 7319
+    },
+    {
+      "epoch": 0.5082627412859325,
+      "grad_norm": 3.333572433367519,
+      "learning_rate": 5.109628727277204e-06,
+      "loss": 0.3045,
+      "step": 7320
+    },
+    {
+      "epoch": 0.5083321760866546,
+      "grad_norm": 3.869795359563918,
+      "learning_rate": 5.10850450752632e-06,
+      "loss": 0.3898,
+      "step": 7321
+    },
+    {
+      "epoch": 0.5084016108873768,
+      "grad_norm": 3.8734084260986616,
+      "learning_rate": 5.107380282287397e-06,
+      "loss": 0.5122,
+      "step": 7322
+    },
+    {
+      "epoch": 0.5084710456880989,
+      "grad_norm": 2.6485488193135343,
+      "learning_rate": 5.106256051617303e-06,
+      "loss": 0.3834,
+      "step": 7323
+    },
+    {
+      "epoch": 0.508540480488821,
+      "grad_norm": 4.331602587561121,
+      "learning_rate": 5.105131815572899e-06,
+      "loss": 0.6433,
+      "step": 7324
+    },
+    {
+      "epoch": 0.5086099152895431,
+      "grad_norm": 3.7808476035374707,
+      "learning_rate": 5.104007574211046e-06,
+      "loss": 0.2841,
+      "step": 7325
+    },
+    {
+      "epoch": 0.5086793500902652,
+      "grad_norm": 4.351890799148692,
+      "learning_rate": 5.102883327588608e-06,
+      "loss": 0.4539,
+      "step": 7326
+    },
+    {
+      "epoch": 0.5087487848909874,
+      "grad_norm": 3.7487331575200065,
+      "learning_rate": 5.101759075762446e-06,
+      "loss": 0.237,
+      "step": 7327
+    },
+    {
+      "epoch": 0.5088182196917095,
+      "grad_norm": 3.239051486277959,
+      "learning_rate": 5.100634818789427e-06,
+      "loss": 0.2157,
+      "step": 7328
+    },
+    {
+      "epoch": 0.5088876544924316,
+      "grad_norm": 3.9147030633068347,
+      "learning_rate": 5.099510556726411e-06,
+      "loss": 0.7202,
+      "step": 7329
+    },
+    {
+      "epoch": 0.5089570892931538,
+      "grad_norm": 3.515192339882417,
+      "learning_rate": 5.098386289630266e-06,
+      "loss": 0.4358,
+      "step": 7330
+    },
+    {
+      "epoch": 0.5090265240938758,
+      "grad_norm": 3.1993609765652256,
+      "learning_rate": 5.097262017557853e-06,
+      "loss": 0.2289,
+      "step": 7331
+    },
+    {
+      "epoch": 0.509095958894598,
+      "grad_norm": 3.3784722802459024,
+      "learning_rate": 5.096137740566036e-06,
+      "loss": 0.2266,
+      "step": 7332
+    },
+    {
+      "epoch": 0.5091653936953201,
+      "grad_norm": 3.6923315489913824,
+      "learning_rate": 5.095013458711682e-06,
+      "loss": 0.4823,
+      "step": 7333
+    },
+    {
+      "epoch": 0.5092348284960422,
+      "grad_norm": 4.338774770405565,
+      "learning_rate": 5.093889172051655e-06,
+      "loss": 0.3231,
+      "step": 7334
+    },
+    {
+      "epoch": 0.5093042632967644,
+      "grad_norm": 3.4343717968562437,
+      "learning_rate": 5.092764880642819e-06,
+      "loss": 0.4444,
+      "step": 7335
+    },
+    {
+      "epoch": 0.5093736980974864,
+      "grad_norm": 2.209798312569309,
+      "learning_rate": 5.091640584542039e-06,
+      "loss": 0.2428,
+      "step": 7336
+    },
+    {
+      "epoch": 0.5094431328982085,
+      "grad_norm": 3.583212744776147,
+      "learning_rate": 5.090516283806185e-06,
+      "loss": 0.4566,
+      "step": 7337
+    },
+    {
+      "epoch": 0.5095125676989307,
+      "grad_norm": 3.648913615352976,
+      "learning_rate": 5.089391978492118e-06,
+      "loss": 0.3551,
+      "step": 7338
+    },
+    {
+      "epoch": 0.5095820024996528,
+      "grad_norm": 4.457411859698555,
+      "learning_rate": 5.0882676686567056e-06,
+      "loss": 0.5513,
+      "step": 7339
+    },
+    {
+      "epoch": 0.509651437300375,
+      "grad_norm": 4.651116125006129,
+      "learning_rate": 5.087143354356816e-06,
+      "loss": 0.5883,
+      "step": 7340
+    },
+    {
+      "epoch": 0.5097208721010971,
+      "grad_norm": 3.9702899954139377,
+      "learning_rate": 5.086019035649313e-06,
+      "loss": 0.571,
+      "step": 7341
+    },
+    {
+      "epoch": 0.5097903069018191,
+      "grad_norm": 4.366411952162467,
+      "learning_rate": 5.084894712591065e-06,
+      "loss": 0.6354,
+      "step": 7342
+    },
+    {
+      "epoch": 0.5098597417025413,
+      "grad_norm": 4.039745825946049,
+      "learning_rate": 5.083770385238938e-06,
+      "loss": 0.6537,
+      "step": 7343
+    },
+    {
+      "epoch": 0.5099291765032634,
+      "grad_norm": 2.7223353738746323,
+      "learning_rate": 5.0826460536498e-06,
+      "loss": 0.235,
+      "step": 7344
+    },
+    {
+      "epoch": 0.5099986113039856,
+      "grad_norm": 3.892021564779901,
+      "learning_rate": 5.081521717880517e-06,
+      "loss": 0.3636,
+      "step": 7345
+    },
+    {
+      "epoch": 0.5100680461047077,
+      "grad_norm": 3.3003064540332017,
+      "learning_rate": 5.080397377987959e-06,
+      "loss": 0.509,
+      "step": 7346
+    },
+    {
+      "epoch": 0.5101374809054298,
+      "grad_norm": 4.099239393762142,
+      "learning_rate": 5.079273034028993e-06,
+      "loss": 0.3808,
+      "step": 7347
+    },
+    {
+      "epoch": 0.510206915706152,
+      "grad_norm": 5.045927676357197,
+      "learning_rate": 5.078148686060485e-06,
+      "loss": 0.6752,
+      "step": 7348
+    },
+    {
+      "epoch": 0.510276350506874,
+      "grad_norm": 3.9211210312711975,
+      "learning_rate": 5.077024334139307e-06,
+      "loss": 0.5902,
+      "step": 7349
+    },
+    {
+      "epoch": 0.5103457853075962,
+      "grad_norm": 2.5422465456688115,
+      "learning_rate": 5.075899978322324e-06,
+      "loss": 0.2332,
+      "step": 7350
+    },
+    {
+      "epoch": 0.5104152201083183,
+      "grad_norm": 3.7084603078155274,
+      "learning_rate": 5.074775618666407e-06,
+      "loss": 0.3367,
+      "step": 7351
+    },
+    {
+      "epoch": 0.5104846549090404,
+      "grad_norm": 3.1723032394773703,
+      "learning_rate": 5.073651255228422e-06,
+      "loss": 0.1995,
+      "step": 7352
+    },
+    {
+      "epoch": 0.5105540897097626,
+      "grad_norm": 4.87057584060196,
+      "learning_rate": 5.072526888065241e-06,
+      "loss": 0.3142,
+      "step": 7353
+    },
+    {
+      "epoch": 0.5106235245104846,
+      "grad_norm": 5.192692942379798,
+      "learning_rate": 5.071402517233732e-06,
+      "loss": 0.6636,
+      "step": 7354
+    },
+    {
+      "epoch": 0.5106929593112067,
+      "grad_norm": 4.121587746243678,
+      "learning_rate": 5.070278142790765e-06,
+      "loss": 0.6396,
+      "step": 7355
+    },
+    {
+      "epoch": 0.5107623941119289,
+      "grad_norm": 4.72840724797986,
+      "learning_rate": 5.069153764793208e-06,
+      "loss": 0.5234,
+      "step": 7356
+    },
+    {
+      "epoch": 0.510831828912651,
+      "grad_norm": 4.898512609125218,
+      "learning_rate": 5.068029383297934e-06,
+      "loss": 0.5852,
+      "step": 7357
+    },
+    {
+      "epoch": 0.5109012637133732,
+      "grad_norm": 4.567178052927413,
+      "learning_rate": 5.066904998361808e-06,
+      "loss": 0.3001,
+      "step": 7358
+    },
+    {
+      "epoch": 0.5109706985140953,
+      "grad_norm": 3.367337365241316,
+      "learning_rate": 5.065780610041703e-06,
+      "loss": 0.2904,
+      "step": 7359
+    },
+    {
+      "epoch": 0.5110401333148173,
+      "grad_norm": 3.912805513220785,
+      "learning_rate": 5.064656218394491e-06,
+      "loss": 0.5044,
+      "step": 7360
+    },
+    {
+      "epoch": 0.5111095681155395,
+      "grad_norm": 3.650919678174253,
+      "learning_rate": 5.0635318234770394e-06,
+      "loss": 0.637,
+      "step": 7361
+    },
+    {
+      "epoch": 0.5111790029162616,
+      "grad_norm": 2.868649650562126,
+      "learning_rate": 5.062407425346221e-06,
+      "loss": 0.2507,
+      "step": 7362
+    },
+    {
+      "epoch": 0.5112484377169838,
+      "grad_norm": 4.067788908101779,
+      "learning_rate": 5.061283024058905e-06,
+      "loss": 0.4835,
+      "step": 7363
+    },
+    {
+      "epoch": 0.5113178725177059,
+      "grad_norm": 4.138576359209665,
+      "learning_rate": 5.060158619671963e-06,
+      "loss": 0.3445,
+      "step": 7364
+    },
+    {
+      "epoch": 0.511387307318428,
+      "grad_norm": 4.09788121012581,
+      "learning_rate": 5.059034212242266e-06,
+      "loss": 0.4249,
+      "step": 7365
+    },
+    {
+      "epoch": 0.5114567421191502,
+      "grad_norm": 3.6792889648849867,
+      "learning_rate": 5.057909801826686e-06,
+      "loss": 0.4177,
+      "step": 7366
+    },
+    {
+      "epoch": 0.5115261769198722,
+      "grad_norm": 3.9235997644277627,
+      "learning_rate": 5.056785388482094e-06,
+      "loss": 0.4392,
+      "step": 7367
+    },
+    {
+      "epoch": 0.5115956117205943,
+      "grad_norm": 3.6891302097659935,
+      "learning_rate": 5.055660972265359e-06,
+      "loss": 0.5416,
+      "step": 7368
+    },
+    {
+      "epoch": 0.5116650465213165,
+      "grad_norm": 3.7346039332093857,
+      "learning_rate": 5.054536553233357e-06,
+      "loss": 0.5202,
+      "step": 7369
+    },
+    {
+      "epoch": 0.5117344813220386,
+      "grad_norm": 4.190990611199103,
+      "learning_rate": 5.053412131442957e-06,
+      "loss": 0.6197,
+      "step": 7370
+    },
+    {
+      "epoch": 0.5118039161227608,
+      "grad_norm": 4.6552316316834546,
+      "learning_rate": 5.052287706951033e-06,
+      "loss": 0.6024,
+      "step": 7371
+    },
+    {
+      "epoch": 0.5118733509234829,
+      "grad_norm": 2.668848292089595,
+      "learning_rate": 5.051163279814456e-06,
+      "loss": 0.2663,
+      "step": 7372
+    },
+    {
+      "epoch": 0.5119427857242049,
+      "grad_norm": 4.849814525151152,
+      "learning_rate": 5.050038850090097e-06,
+      "loss": 0.6434,
+      "step": 7373
+    },
+    {
+      "epoch": 0.5120122205249271,
+      "grad_norm": 5.532234751676624,
+      "learning_rate": 5.048914417834831e-06,
+      "loss": 0.5771,
+      "step": 7374
+    },
+    {
+      "epoch": 0.5120816553256492,
+      "grad_norm": 5.894770271013601,
+      "learning_rate": 5.047789983105527e-06,
+      "loss": 0.5901,
+      "step": 7375
+    },
+    {
+      "epoch": 0.5121510901263714,
+      "grad_norm": 3.8916712185232156,
+      "learning_rate": 5.046665545959062e-06,
+      "loss": 0.5251,
+      "step": 7376
+    },
+    {
+      "epoch": 0.5122205249270935,
+      "grad_norm": 3.643642701607656,
+      "learning_rate": 5.045541106452304e-06,
+      "loss": 0.4344,
+      "step": 7377
+    },
+    {
+      "epoch": 0.5122899597278155,
+      "grad_norm": 3.927717597354102,
+      "learning_rate": 5.0444166646421296e-06,
+      "loss": 0.415,
+      "step": 7378
+    },
+    {
+      "epoch": 0.5123593945285377,
+      "grad_norm": 4.5091329258044786,
+      "learning_rate": 5.0432922205854105e-06,
+      "loss": 0.6988,
+      "step": 7379
+    },
+    {
+      "epoch": 0.5124288293292598,
+      "grad_norm": 3.8226146647026296,
+      "learning_rate": 5.042167774339021e-06,
+      "loss": 0.3855,
+      "step": 7380
+    },
+    {
+      "epoch": 0.5124982641299819,
+      "grad_norm": 4.27580246810149,
+      "learning_rate": 5.041043325959832e-06,
+      "loss": 0.6088,
+      "step": 7381
+    },
+    {
+      "epoch": 0.5125676989307041,
+      "grad_norm": 5.026632666519692,
+      "learning_rate": 5.039918875504719e-06,
+      "loss": 0.7263,
+      "step": 7382
+    },
+    {
+      "epoch": 0.5126371337314262,
+      "grad_norm": 3.929315988947903,
+      "learning_rate": 5.038794423030554e-06,
+      "loss": 0.4225,
+      "step": 7383
+    },
+    {
+      "epoch": 0.5127065685321484,
+      "grad_norm": 3.3843586499910154,
+      "learning_rate": 5.03766996859421e-06,
+      "loss": 0.2893,
+      "step": 7384
+    },
+    {
+      "epoch": 0.5127760033328704,
+      "grad_norm": 4.21622162747944,
+      "learning_rate": 5.036545512252562e-06,
+      "loss": 0.4854,
+      "step": 7385
+    },
+    {
+      "epoch": 0.5128454381335925,
+      "grad_norm": 3.1419041804649,
+      "learning_rate": 5.0354210540624835e-06,
+      "loss": 0.3465,
+      "step": 7386
+    },
+    {
+      "epoch": 0.5129148729343147,
+      "grad_norm": 3.726252964432927,
+      "learning_rate": 5.034296594080849e-06,
+      "loss": 0.5058,
+      "step": 7387
+    },
+    {
+      "epoch": 0.5129843077350368,
+      "grad_norm": 4.551023190621925,
+      "learning_rate": 5.03317213236453e-06,
+      "loss": 0.6602,
+      "step": 7388
+    },
+    {
+      "epoch": 0.513053742535759,
+      "grad_norm": 3.9417216256808816,
+      "learning_rate": 5.032047668970403e-06,
+      "loss": 0.3407,
+      "step": 7389
+    },
+    {
+      "epoch": 0.513123177336481,
+      "grad_norm": 3.614828451274167,
+      "learning_rate": 5.030923203955342e-06,
+      "loss": 0.4358,
+      "step": 7390
+    },
+    {
+      "epoch": 0.5131926121372031,
+      "grad_norm": 3.439788279758886,
+      "learning_rate": 5.0297987373762195e-06,
+      "loss": 0.2892,
+      "step": 7391
+    },
+    {
+      "epoch": 0.5132620469379253,
+      "grad_norm": 3.410438631041953,
+      "learning_rate": 5.02867426928991e-06,
+      "loss": 0.2713,
+      "step": 7392
+    },
+    {
+      "epoch": 0.5133314817386474,
+      "grad_norm": 3.1960815787517345,
+      "learning_rate": 5.027549799753289e-06,
+      "loss": 0.3394,
+      "step": 7393
+    },
+    {
+      "epoch": 0.5134009165393695,
+      "grad_norm": 4.049103593529985,
+      "learning_rate": 5.02642532882323e-06,
+      "loss": 0.5214,
+      "step": 7394
+    },
+    {
+      "epoch": 0.5134703513400917,
+      "grad_norm": 3.503746421951426,
+      "learning_rate": 5.025300856556609e-06,
+      "loss": 0.3548,
+      "step": 7395
+    },
+    {
+      "epoch": 0.5135397861408137,
+      "grad_norm": 3.5221906436998074,
+      "learning_rate": 5.024176383010299e-06,
+      "loss": 0.3894,
+      "step": 7396
+    },
+    {
+      "epoch": 0.5136092209415359,
+      "grad_norm": 4.367991582801302,
+      "learning_rate": 5.023051908241173e-06,
+      "loss": 0.4638,
+      "step": 7397
+    },
+    {
+      "epoch": 0.513678655742258,
+      "grad_norm": 3.663850545934597,
+      "learning_rate": 5.021927432306108e-06,
+      "loss": 0.3295,
+      "step": 7398
+    },
+    {
+      "epoch": 0.5137480905429801,
+      "grad_norm": 3.936735311414971,
+      "learning_rate": 5.020802955261981e-06,
+      "loss": 0.424,
+      "step": 7399
+    },
+    {
+      "epoch": 0.5138175253437023,
+      "grad_norm": 4.893341683518537,
+      "learning_rate": 5.019678477165661e-06,
+      "loss": 0.6211,
+      "step": 7400
+    },
+    {
+      "epoch": 0.5138869601444244,
+      "grad_norm": 3.8650943334381473,
+      "learning_rate": 5.018553998074027e-06,
+      "loss": 0.6089,
+      "step": 7401
+    },
+    {
+      "epoch": 0.5139563949451466,
+      "grad_norm": 5.110811447140414,
+      "learning_rate": 5.0174295180439535e-06,
+      "loss": 0.6568,
+      "step": 7402
+    },
+    {
+      "epoch": 0.5140258297458686,
+      "grad_norm": 3.4259390688196474,
+      "learning_rate": 5.016305037132314e-06,
+      "loss": 0.2707,
+      "step": 7403
+    },
+    {
+      "epoch": 0.5140952645465907,
+      "grad_norm": 3.6219259196108067,
+      "learning_rate": 5.015180555395983e-06,
+      "loss": 0.5879,
+      "step": 7404
+    },
+    {
+      "epoch": 0.5141646993473129,
+      "grad_norm": 4.4679404760022265,
+      "learning_rate": 5.01405607289184e-06,
+      "loss": 0.4615,
+      "step": 7405
+    },
+    {
+      "epoch": 0.514234134148035,
+      "grad_norm": 4.862082444895862,
+      "learning_rate": 5.012931589676754e-06,
+      "loss": 0.2659,
+      "step": 7406
+    },
+    {
+      "epoch": 0.5143035689487571,
+      "grad_norm": 4.829242437920496,
+      "learning_rate": 5.0118071058076026e-06,
+      "loss": 0.5855,
+      "step": 7407
+    },
+    {
+      "epoch": 0.5143730037494793,
+      "grad_norm": 3.9486886250318687,
+      "learning_rate": 5.010682621341262e-06,
+      "loss": 0.3947,
+      "step": 7408
+    },
+    {
+      "epoch": 0.5144424385502013,
+      "grad_norm": 3.103357133665367,
+      "learning_rate": 5.009558136334607e-06,
+      "loss": 0.3305,
+      "step": 7409
+    },
+    {
+      "epoch": 0.5145118733509235,
+      "grad_norm": 3.650317343219308,
+      "learning_rate": 5.008433650844512e-06,
+      "loss": 0.4124,
+      "step": 7410
+    },
+    {
+      "epoch": 0.5145813081516456,
+      "grad_norm": 3.8131190801382377,
+      "learning_rate": 5.007309164927853e-06,
+      "loss": 0.5105,
+      "step": 7411
+    },
+    {
+      "epoch": 0.5146507429523677,
+      "grad_norm": 3.532756246143908,
+      "learning_rate": 5.006184678641504e-06,
+      "loss": 0.4849,
+      "step": 7412
+    },
+    {
+      "epoch": 0.5147201777530899,
+      "grad_norm": 4.222547184845305,
+      "learning_rate": 5.005060192042339e-06,
+      "loss": 0.4929,
+      "step": 7413
+    },
+    {
+      "epoch": 0.514789612553812,
+      "grad_norm": 3.919410398359665,
+      "learning_rate": 5.003935705187237e-06,
+      "loss": 0.5032,
+      "step": 7414
+    },
+    {
+      "epoch": 0.5148590473545341,
+      "grad_norm": 3.1342244339931993,
+      "learning_rate": 5.0028112181330736e-06,
+      "loss": 0.347,
+      "step": 7415
+    },
+    {
+      "epoch": 0.5149284821552562,
+      "grad_norm": 3.5028107039265497,
+      "learning_rate": 5.001686730936719e-06,
+      "loss": 0.4159,
+      "step": 7416
+    },
+    {
+      "epoch": 0.5149979169559783,
+      "grad_norm": 3.1337309621149907,
+      "learning_rate": 5.0005622436550525e-06,
+      "loss": 0.4034,
+      "step": 7417
+    },
+    {
+      "epoch": 0.5150673517567005,
+      "grad_norm": 3.9667189110597776,
+      "learning_rate": 4.999437756344948e-06,
+      "loss": 0.4024,
+      "step": 7418
+    },
+    {
+      "epoch": 0.5151367865574226,
+      "grad_norm": 3.396356699479342,
+      "learning_rate": 4.998313269063282e-06,
+      "loss": 0.5023,
+      "step": 7419
+    },
+    {
+      "epoch": 0.5152062213581448,
+      "grad_norm": 2.0701880851423495,
+      "learning_rate": 4.997188781866929e-06,
+      "loss": 0.1875,
+      "step": 7420
+    },
+    {
+      "epoch": 0.5152756561588668,
+      "grad_norm": 2.9196388401222144,
+      "learning_rate": 4.9960642948127635e-06,
+      "loss": 0.2814,
+      "step": 7421
+    },
+    {
+      "epoch": 0.5153450909595889,
+      "grad_norm": 4.0452288604675894,
+      "learning_rate": 4.994939807957661e-06,
+      "loss": 0.5527,
+      "step": 7422
+    },
+    {
+      "epoch": 0.5154145257603111,
+      "grad_norm": 3.813480093911458,
+      "learning_rate": 4.993815321358499e-06,
+      "loss": 0.4133,
+      "step": 7423
+    },
+    {
+      "epoch": 0.5154839605610332,
+      "grad_norm": 4.631406447952932,
+      "learning_rate": 4.992690835072149e-06,
+      "loss": 0.4669,
+      "step": 7424
+    },
+    {
+      "epoch": 0.5155533953617553,
+      "grad_norm": 3.601168122937942,
+      "learning_rate": 4.991566349155489e-06,
+      "loss": 0.4162,
+      "step": 7425
+    },
+    {
+      "epoch": 0.5156228301624775,
+      "grad_norm": 4.761889546969619,
+      "learning_rate": 4.990441863665393e-06,
+      "loss": 0.4695,
+      "step": 7426
+    },
+    {
+      "epoch": 0.5156922649631995,
+      "grad_norm": 3.6994665527818253,
+      "learning_rate": 4.989317378658739e-06,
+      "loss": 0.4137,
+      "step": 7427
+    },
+    {
+      "epoch": 0.5157616997639217,
+      "grad_norm": 4.779711746647431,
+      "learning_rate": 4.988192894192398e-06,
+      "loss": 0.322,
+      "step": 7428
+    },
+    {
+      "epoch": 0.5158311345646438,
+      "grad_norm": 3.739015475062009,
+      "learning_rate": 4.987068410323248e-06,
+      "loss": 0.4807,
+      "step": 7429
+    },
+    {
+      "epoch": 0.5159005693653659,
+      "grad_norm": 3.3389499584328077,
+      "learning_rate": 4.985943927108163e-06,
+      "loss": 0.4506,
+      "step": 7430
+    },
+    {
+      "epoch": 0.5159700041660881,
+      "grad_norm": 4.732029473768147,
+      "learning_rate": 4.984819444604018e-06,
+      "loss": 0.6887,
+      "step": 7431
+    },
+    {
+      "epoch": 0.5160394389668101,
+      "grad_norm": 3.9530305200250453,
+      "learning_rate": 4.983694962867687e-06,
+      "loss": 0.4963,
+      "step": 7432
+    },
+    {
+      "epoch": 0.5161088737675323,
+      "grad_norm": 4.139718479421074,
+      "learning_rate": 4.982570481956047e-06,
+      "loss": 0.6287,
+      "step": 7433
+    },
+    {
+      "epoch": 0.5161783085682544,
+      "grad_norm": 3.526725979520689,
+      "learning_rate": 4.9814460019259745e-06,
+      "loss": 0.411,
+      "step": 7434
+    },
+    {
+      "epoch": 0.5162477433689765,
+      "grad_norm": 3.133034095546358,
+      "learning_rate": 4.9803215228343405e-06,
+      "loss": 0.433,
+      "step": 7435
+    },
+    {
+      "epoch": 0.5163171781696987,
+      "grad_norm": 3.6174425989802415,
+      "learning_rate": 4.97919704473802e-06,
+      "loss": 0.4467,
+      "step": 7436
+    },
+    {
+      "epoch": 0.5163866129704208,
+      "grad_norm": 4.031198835084751,
+      "learning_rate": 4.978072567693892e-06,
+      "loss": 0.3667,
+      "step": 7437
+    },
+    {
+      "epoch": 0.5164560477711428,
+      "grad_norm": 3.3930918882064818,
+      "learning_rate": 4.9769480917588295e-06,
+      "loss": 0.3938,
+      "step": 7438
+    },
+    {
+      "epoch": 0.516525482571865,
+      "grad_norm": 4.076396016414378,
+      "learning_rate": 4.975823616989704e-06,
+      "loss": 0.4294,
+      "step": 7439
+    },
+    {
+      "epoch": 0.5165949173725871,
+      "grad_norm": 4.615457457825244,
+      "learning_rate": 4.974699143443393e-06,
+      "loss": 0.6725,
+      "step": 7440
+    },
+    {
+      "epoch": 0.5166643521733093,
+      "grad_norm": 3.882061475766921,
+      "learning_rate": 4.973574671176772e-06,
+      "loss": 0.4237,
+      "step": 7441
+    },
+    {
+      "epoch": 0.5167337869740314,
+      "grad_norm": 3.1425675779098405,
+      "learning_rate": 4.972450200246712e-06,
+      "loss": 0.3549,
+      "step": 7442
+    },
+    {
+      "epoch": 0.5168032217747535,
+      "grad_norm": 4.475548458954799,
+      "learning_rate": 4.971325730710091e-06,
+      "loss": 0.3639,
+      "step": 7443
+    },
+    {
+      "epoch": 0.5168726565754757,
+      "grad_norm": 4.987274369695312,
+      "learning_rate": 4.970201262623781e-06,
+      "loss": 0.5691,
+      "step": 7444
+    },
+    {
+      "epoch": 0.5169420913761977,
+      "grad_norm": 3.7843322770991743,
+      "learning_rate": 4.96907679604466e-06,
+      "loss": 0.4522,
+      "step": 7445
+    },
+    {
+      "epoch": 0.5170115261769199,
+      "grad_norm": 5.852420954155457,
+      "learning_rate": 4.9679523310295976e-06,
+      "loss": 0.52,
+      "step": 7446
+    },
+    {
+      "epoch": 0.517080960977642,
+      "grad_norm": 3.795972864003024,
+      "learning_rate": 4.966827867635471e-06,
+      "loss": 0.4762,
+      "step": 7447
+    },
+    {
+      "epoch": 0.5171503957783641,
+      "grad_norm": 3.599465245664327,
+      "learning_rate": 4.965703405919154e-06,
+      "loss": 0.3878,
+      "step": 7448
+    },
+    {
+      "epoch": 0.5172198305790863,
+      "grad_norm": 4.076955225766076,
+      "learning_rate": 4.964578945937517e-06,
+      "loss": 0.3919,
+      "step": 7449
+    },
+    {
+      "epoch": 0.5172892653798083,
+      "grad_norm": 4.967015797190261,
+      "learning_rate": 4.963454487747439e-06,
+      "loss": 0.5377,
+      "step": 7450
+    },
+    {
+      "epoch": 0.5173587001805304,
+      "grad_norm": 4.488148372757971,
+      "learning_rate": 4.962330031405791e-06,
+      "loss": 0.5394,
+      "step": 7451
+    },
+    {
+      "epoch": 0.5174281349812526,
+      "grad_norm": 3.7525627444133955,
+      "learning_rate": 4.961205576969449e-06,
+      "loss": 0.4171,
+      "step": 7452
+    },
+    {
+      "epoch": 0.5174975697819747,
+      "grad_norm": 3.137059576876212,
+      "learning_rate": 4.960081124495283e-06,
+      "loss": 0.3239,
+      "step": 7453
+    },
+    {
+      "epoch": 0.5175670045826969,
+      "grad_norm": 3.508016311545466,
+      "learning_rate": 4.958956674040169e-06,
+      "loss": 0.5235,
+      "step": 7454
+    },
+    {
+      "epoch": 0.517636439383419,
+      "grad_norm": 3.8043459093943346,
+      "learning_rate": 4.957832225660982e-06,
+      "loss": 0.3742,
+      "step": 7455
+    },
+    {
+      "epoch": 0.517705874184141,
+      "grad_norm": 3.6020784514952737,
+      "learning_rate": 4.95670777941459e-06,
+      "loss": 0.4566,
+      "step": 7456
+    },
+    {
+      "epoch": 0.5177753089848632,
+      "grad_norm": 4.047470226598575,
+      "learning_rate": 4.955583335357871e-06,
+      "loss": 0.5954,
+      "step": 7457
+    },
+    {
+      "epoch": 0.5178447437855853,
+      "grad_norm": 3.803362490498877,
+      "learning_rate": 4.954458893547696e-06,
+      "loss": 0.4045,
+      "step": 7458
+    },
+    {
+      "epoch": 0.5179141785863075,
+      "grad_norm": 4.258490219661581,
+      "learning_rate": 4.953334454040941e-06,
+      "loss": 0.497,
+      "step": 7459
+    },
+    {
+      "epoch": 0.5179836133870296,
+      "grad_norm": 3.1665407656401183,
+      "learning_rate": 4.952210016894475e-06,
+      "loss": 0.351,
+      "step": 7460
+    },
+    {
+      "epoch": 0.5180530481877517,
+      "grad_norm": 7.124607720989545,
+      "learning_rate": 4.951085582165171e-06,
+      "loss": 0.504,
+      "step": 7461
+    },
+    {
+      "epoch": 0.5181224829884739,
+      "grad_norm": 3.7920177226079876,
+      "learning_rate": 4.9499611499099044e-06,
+      "loss": 0.4604,
+      "step": 7462
+    },
+    {
+      "epoch": 0.5181919177891959,
+      "grad_norm": 4.3638013575300665,
+      "learning_rate": 4.948836720185547e-06,
+      "loss": 0.5517,
+      "step": 7463
+    },
+    {
+      "epoch": 0.518261352589918,
+      "grad_norm": 3.8830018687896573,
+      "learning_rate": 4.947712293048968e-06,
+      "loss": 0.4623,
+      "step": 7464
+    },
+    {
+      "epoch": 0.5183307873906402,
+      "grad_norm": 4.021336543863978,
+      "learning_rate": 4.9465878685570426e-06,
+      "loss": 0.5641,
+      "step": 7465
+    },
+    {
+      "epoch": 0.5184002221913623,
+      "grad_norm": 3.6199174824948535,
+      "learning_rate": 4.9454634467666445e-06,
+      "loss": 0.4274,
+      "step": 7466
+    },
+    {
+      "epoch": 0.5184696569920845,
+      "grad_norm": 3.043527862789783,
+      "learning_rate": 4.944339027734642e-06,
+      "loss": 0.2518,
+      "step": 7467
+    },
+    {
+      "epoch": 0.5185390917928066,
+      "grad_norm": 4.590053773232761,
+      "learning_rate": 4.943214611517907e-06,
+      "loss": 0.6955,
+      "step": 7468
+    },
+    {
+      "epoch": 0.5186085265935286,
+      "grad_norm": 2.9521522873524786,
+      "learning_rate": 4.9420901981733155e-06,
+      "loss": 0.1785,
+      "step": 7469
+    },
+    {
+      "epoch": 0.5186779613942508,
+      "grad_norm": 3.690117446195559,
+      "learning_rate": 4.940965787757736e-06,
+      "loss": 0.3715,
+      "step": 7470
+    },
+    {
+      "epoch": 0.5187473961949729,
+      "grad_norm": 3.67230083518739,
+      "learning_rate": 4.939841380328038e-06,
+      "loss": 0.4515,
+      "step": 7471
+    },
+    {
+      "epoch": 0.5188168309956951,
+      "grad_norm": 4.435123467769752,
+      "learning_rate": 4.938716975941096e-06,
+      "loss": 0.7382,
+      "step": 7472
+    },
+    {
+      "epoch": 0.5188862657964172,
+      "grad_norm": 5.936918097764168,
+      "learning_rate": 4.937592574653782e-06,
+      "loss": 0.5296,
+      "step": 7473
+    },
+    {
+      "epoch": 0.5189557005971392,
+      "grad_norm": 2.619471109127567,
+      "learning_rate": 4.936468176522962e-06,
+      "loss": 0.2328,
+      "step": 7474
+    },
+    {
+      "epoch": 0.5190251353978614,
+      "grad_norm": 3.5656751582580815,
+      "learning_rate": 4.93534378160551e-06,
+      "loss": 0.6072,
+      "step": 7475
+    },
+    {
+      "epoch": 0.5190945701985835,
+      "grad_norm": 3.848373382293671,
+      "learning_rate": 4.9342193899582975e-06,
+      "loss": 0.4654,
+      "step": 7476
+    },
+    {
+      "epoch": 0.5191640049993057,
+      "grad_norm": 3.129756283013653,
+      "learning_rate": 4.933095001638195e-06,
+      "loss": 0.3167,
+      "step": 7477
+    },
+    {
+      "epoch": 0.5192334398000278,
+      "grad_norm": 4.480671072898298,
+      "learning_rate": 4.931970616702069e-06,
+      "loss": 0.505,
+      "step": 7478
+    },
+    {
+      "epoch": 0.5193028746007499,
+      "grad_norm": 5.173018447424363,
+      "learning_rate": 4.930846235206793e-06,
+      "loss": 0.7989,
+      "step": 7479
+    },
+    {
+      "epoch": 0.519372309401472,
+      "grad_norm": 3.570656650148614,
+      "learning_rate": 4.929721857209238e-06,
+      "loss": 0.4606,
+      "step": 7480
+    },
+    {
+      "epoch": 0.5194417442021941,
+      "grad_norm": 4.372306302253888,
+      "learning_rate": 4.928597482766269e-06,
+      "loss": 0.3936,
+      "step": 7481
+    },
+    {
+      "epoch": 0.5195111790029162,
+      "grad_norm": 3.6346792488532222,
+      "learning_rate": 4.92747311193476e-06,
+      "loss": 0.3943,
+      "step": 7482
+    },
+    {
+      "epoch": 0.5195806138036384,
+      "grad_norm": 3.8986357997135075,
+      "learning_rate": 4.926348744771579e-06,
+      "loss": 0.3916,
+      "step": 7483
+    },
+    {
+      "epoch": 0.5196500486043605,
+      "grad_norm": 4.321729184597143,
+      "learning_rate": 4.9252243813335956e-06,
+      "loss": 0.3067,
+      "step": 7484
+    },
+    {
+      "epoch": 0.5197194834050827,
+      "grad_norm": 3.309782373956979,
+      "learning_rate": 4.924100021677677e-06,
+      "loss": 0.2252,
+      "step": 7485
+    },
+    {
+      "epoch": 0.5197889182058048,
+      "grad_norm": 4.7120736422940315,
+      "learning_rate": 4.922975665860694e-06,
+      "loss": 0.5368,
+      "step": 7486
+    },
+    {
+      "epoch": 0.5198583530065268,
+      "grad_norm": 3.629825826108819,
+      "learning_rate": 4.921851313939515e-06,
+      "loss": 0.335,
+      "step": 7487
+    },
+    {
+      "epoch": 0.519927787807249,
+      "grad_norm": 2.697189788751329,
+      "learning_rate": 4.9207269659710085e-06,
+      "loss": 0.3812,
+      "step": 7488
+    },
+    {
+      "epoch": 0.5199972226079711,
+      "grad_norm": 4.065926088847972,
+      "learning_rate": 4.919602622012042e-06,
+      "loss": 0.3214,
+      "step": 7489
+    },
+    {
+      "epoch": 0.5200666574086933,
+      "grad_norm": 2.8749544035743337,
+      "learning_rate": 4.9184782821194835e-06,
+      "loss": 0.304,
+      "step": 7490
+    },
+    {
+      "epoch": 0.5201360922094154,
+      "grad_norm": 4.989590159116698,
+      "learning_rate": 4.9173539463502025e-06,
+      "loss": 0.8215,
+      "step": 7491
+    },
+    {
+      "epoch": 0.5202055270101374,
+      "grad_norm": 2.9707239020785052,
+      "learning_rate": 4.916229614761065e-06,
+      "loss": 0.2269,
+      "step": 7492
+    },
+    {
+      "epoch": 0.5202749618108596,
+      "grad_norm": 5.967053870860305,
+      "learning_rate": 4.9151052874089365e-06,
+      "loss": 0.5471,
+      "step": 7493
+    },
+    {
+      "epoch": 0.5203443966115817,
+      "grad_norm": 4.346946189581933,
+      "learning_rate": 4.913980964350688e-06,
+      "loss": 0.4469,
+      "step": 7494
+    },
+    {
+      "epoch": 0.5204138314123038,
+      "grad_norm": 5.292851305489597,
+      "learning_rate": 4.912856645643186e-06,
+      "loss": 0.7722,
+      "step": 7495
+    },
+    {
+      "epoch": 0.520483266213026,
+      "grad_norm": 3.499532189367827,
+      "learning_rate": 4.911732331343295e-06,
+      "loss": 0.3472,
+      "step": 7496
+    },
+    {
+      "epoch": 0.5205527010137481,
+      "grad_norm": 4.1228641227892435,
+      "learning_rate": 4.910608021507883e-06,
+      "loss": 0.5266,
+      "step": 7497
+    },
+    {
+      "epoch": 0.5206221358144703,
+      "grad_norm": 3.294174014222333,
+      "learning_rate": 4.909483716193817e-06,
+      "loss": 0.3738,
+      "step": 7498
+    },
+    {
+      "epoch": 0.5206915706151923,
+      "grad_norm": 3.425465900396751,
+      "learning_rate": 4.908359415457962e-06,
+      "loss": 0.3754,
+      "step": 7499
+    },
+    {
+      "epoch": 0.5207610054159144,
+      "grad_norm": 3.852591340405903,
+      "learning_rate": 4.907235119357183e-06,
+      "loss": 0.4138,
+      "step": 7500
+    },
+    {
+      "epoch": 0.5208304402166366,
+      "grad_norm": 4.701215699862018,
+      "learning_rate": 4.906110827948347e-06,
+      "loss": 0.7044,
+      "step": 7501
+    },
+    {
+      "epoch": 0.5208998750173587,
+      "grad_norm": 3.046167050709071,
+      "learning_rate": 4.9049865412883205e-06,
+      "loss": 0.2782,
+      "step": 7502
+    },
+    {
+      "epoch": 0.5209693098180809,
+      "grad_norm": 3.700247199629093,
+      "learning_rate": 4.903862259433965e-06,
+      "loss": 0.4541,
+      "step": 7503
+    },
+    {
+      "epoch": 0.521038744618803,
+      "grad_norm": 3.488366701342752,
+      "learning_rate": 4.902737982442148e-06,
+      "loss": 0.6177,
+      "step": 7504
+    },
+    {
+      "epoch": 0.521108179419525,
+      "grad_norm": 4.21476650065884,
+      "learning_rate": 4.9016137103697356e-06,
+      "loss": 0.8293,
+      "step": 7505
+    },
+    {
+      "epoch": 0.5211776142202472,
+      "grad_norm": 3.57792653700439,
+      "learning_rate": 4.90048944327359e-06,
+      "loss": 0.3288,
+      "step": 7506
+    },
+    {
+      "epoch": 0.5212470490209693,
+      "grad_norm": 3.973096155203936,
+      "learning_rate": 4.899365181210574e-06,
+      "loss": 0.4767,
+      "step": 7507
+    },
+    {
+      "epoch": 0.5213164838216914,
+      "grad_norm": 4.678196918418609,
+      "learning_rate": 4.898240924237554e-06,
+      "loss": 0.5879,
+      "step": 7508
+    },
+    {
+      "epoch": 0.5213859186224136,
+      "grad_norm": 3.637778059472762,
+      "learning_rate": 4.897116672411395e-06,
+      "loss": 0.637,
+      "step": 7509
+    },
+    {
+      "epoch": 0.5214553534231356,
+      "grad_norm": 3.3309840268946096,
+      "learning_rate": 4.8959924257889554e-06,
+      "loss": 0.211,
+      "step": 7510
+    },
+    {
+      "epoch": 0.5215247882238578,
+      "grad_norm": 3.71746568758711,
+      "learning_rate": 4.894868184427102e-06,
+      "loss": 0.3039,
+      "step": 7511
+    },
+    {
+      "epoch": 0.5215942230245799,
+      "grad_norm": 3.5051974282577003,
+      "learning_rate": 4.893743948382696e-06,
+      "loss": 0.5485,
+      "step": 7512
+    },
+    {
+      "epoch": 0.521663657825302,
+      "grad_norm": 3.8216843200357498,
+      "learning_rate": 4.8926197177126035e-06,
+      "loss": 0.3697,
+      "step": 7513
+    },
+    {
+      "epoch": 0.5217330926260242,
+      "grad_norm": 3.4712902380385993,
+      "learning_rate": 4.891495492473683e-06,
+      "loss": 0.2865,
+      "step": 7514
+    },
+    {
+      "epoch": 0.5218025274267463,
+      "grad_norm": 4.220174423767041,
+      "learning_rate": 4.890371272722797e-06,
+      "loss": 0.5742,
+      "step": 7515
+    },
+    {
+      "epoch": 0.5218719622274685,
+      "grad_norm": 3.179004269920638,
+      "learning_rate": 4.8892470585168095e-06,
+      "loss": 0.4484,
+      "step": 7516
+    },
+    {
+      "epoch": 0.5219413970281905,
+      "grad_norm": 3.5736651577557743,
+      "learning_rate": 4.888122849912579e-06,
+      "loss": 0.3727,
+      "step": 7517
+    },
+    {
+      "epoch": 0.5220108318289126,
+      "grad_norm": 2.9745945323294856,
+      "learning_rate": 4.886998646966968e-06,
+      "loss": 0.2614,
+      "step": 7518
+    },
+    {
+      "epoch": 0.5220802666296348,
+      "grad_norm": 3.4109334530506255,
+      "learning_rate": 4.885874449736837e-06,
+      "loss": 0.3683,
+      "step": 7519
+    },
+    {
+      "epoch": 0.5221497014303569,
+      "grad_norm": 3.8946956171958402,
+      "learning_rate": 4.88475025827905e-06,
+      "loss": 0.4094,
+      "step": 7520
+    },
+    {
+      "epoch": 0.522219136231079,
+      "grad_norm": 3.963858319031122,
+      "learning_rate": 4.883626072650463e-06,
+      "loss": 0.4972,
+      "step": 7521
+    },
+    {
+      "epoch": 0.5222885710318012,
+      "grad_norm": 5.976793283812102,
+      "learning_rate": 4.882501892907936e-06,
+      "loss": 0.6885,
+      "step": 7522
+    },
+    {
+      "epoch": 0.5223580058325232,
+      "grad_norm": 3.7249661146290753,
+      "learning_rate": 4.8813777191083335e-06,
+      "loss": 0.5074,
+      "step": 7523
+    },
+    {
+      "epoch": 0.5224274406332454,
+      "grad_norm": 4.1624662574957325,
+      "learning_rate": 4.880253551308508e-06,
+      "loss": 0.3992,
+      "step": 7524
+    },
+    {
+      "epoch": 0.5224968754339675,
+      "grad_norm": 3.1828458714283197,
+      "learning_rate": 4.879129389565325e-06,
+      "loss": 0.1703,
+      "step": 7525
+    },
+    {
+      "epoch": 0.5225663102346896,
+      "grad_norm": 4.1379185093262745,
+      "learning_rate": 4.878005233935639e-06,
+      "loss": 0.3613,
+      "step": 7526
+    },
+    {
+      "epoch": 0.5226357450354118,
+      "grad_norm": 2.654738806001754,
+      "learning_rate": 4.876881084476312e-06,
+      "loss": 0.276,
+      "step": 7527
+    },
+    {
+      "epoch": 0.5227051798361338,
+      "grad_norm": 3.759988743770085,
+      "learning_rate": 4.8757569412442e-06,
+      "loss": 0.4422,
+      "step": 7528
+    },
+    {
+      "epoch": 0.522774614636856,
+      "grad_norm": 4.68881995870071,
+      "learning_rate": 4.874632804296161e-06,
+      "loss": 0.5652,
+      "step": 7529
+    },
+    {
+      "epoch": 0.5228440494375781,
+      "grad_norm": 3.656255301243486,
+      "learning_rate": 4.873508673689053e-06,
+      "loss": 0.4475,
+      "step": 7530
+    },
+    {
+      "epoch": 0.5229134842383002,
+      "grad_norm": 3.0593216618017802,
+      "learning_rate": 4.872384549479733e-06,
+      "loss": 0.2277,
+      "step": 7531
+    },
+    {
+      "epoch": 0.5229829190390224,
+      "grad_norm": 2.915545148413807,
+      "learning_rate": 4.871260431725058e-06,
+      "loss": 0.3215,
+      "step": 7532
+    },
+    {
+      "epoch": 0.5230523538397445,
+      "grad_norm": 3.7709763705217973,
+      "learning_rate": 4.870136320481884e-06,
+      "loss": 0.5412,
+      "step": 7533
+    },
+    {
+      "epoch": 0.5231217886404667,
+      "grad_norm": 3.5937118194443407,
+      "learning_rate": 4.869012215807069e-06,
+      "loss": 0.3742,
+      "step": 7534
+    },
+    {
+      "epoch": 0.5231912234411887,
+      "grad_norm": 4.355929676260524,
+      "learning_rate": 4.867888117757468e-06,
+      "loss": 0.5215,
+      "step": 7535
+    },
+    {
+      "epoch": 0.5232606582419108,
+      "grad_norm": 4.128021594091155,
+      "learning_rate": 4.866764026389935e-06,
+      "loss": 0.6572,
+      "step": 7536
+    },
+    {
+      "epoch": 0.523330093042633,
+      "grad_norm": 3.3808461552816205,
+      "learning_rate": 4.865639941761329e-06,
+      "loss": 0.4563,
+      "step": 7537
+    },
+    {
+      "epoch": 0.5233995278433551,
+      "grad_norm": 4.173449901292679,
+      "learning_rate": 4.8645158639285025e-06,
+      "loss": 0.4999,
+      "step": 7538
+    },
+    {
+      "epoch": 0.5234689626440772,
+      "grad_norm": 4.599080494001665,
+      "learning_rate": 4.863391792948308e-06,
+      "loss": 0.6843,
+      "step": 7539
+    },
+    {
+      "epoch": 0.5235383974447994,
+      "grad_norm": 3.9274742877234643,
+      "learning_rate": 4.862267728877604e-06,
+      "loss": 0.5224,
+      "step": 7540
+    },
+    {
+      "epoch": 0.5236078322455214,
+      "grad_norm": 5.140539743114913,
+      "learning_rate": 4.861143671773243e-06,
+      "loss": 0.6705,
+      "step": 7541
+    },
+    {
+      "epoch": 0.5236772670462436,
+      "grad_norm": 4.033299633213148,
+      "learning_rate": 4.860019621692076e-06,
+      "loss": 0.5198,
+      "step": 7542
+    },
+    {
+      "epoch": 0.5237467018469657,
+      "grad_norm": 3.3158099968009216,
+      "learning_rate": 4.858895578690958e-06,
+      "loss": 0.3143,
+      "step": 7543
+    },
+    {
+      "epoch": 0.5238161366476878,
+      "grad_norm": 3.3562223728878218,
+      "learning_rate": 4.857771542826743e-06,
+      "loss": 0.3163,
+      "step": 7544
+    },
+    {
+      "epoch": 0.52388557144841,
+      "grad_norm": 3.6998718390310827,
+      "learning_rate": 4.856647514156284e-06,
+      "loss": 0.3125,
+      "step": 7545
+    },
+    {
+      "epoch": 0.523955006249132,
+      "grad_norm": 3.4450661664028988,
+      "learning_rate": 4.855523492736429e-06,
+      "loss": 0.5172,
+      "step": 7546
+    },
+    {
+      "epoch": 0.5240244410498542,
+      "grad_norm": 3.4132250383845832,
+      "learning_rate": 4.8543994786240325e-06,
+      "loss": 0.4554,
+      "step": 7547
+    },
+    {
+      "epoch": 0.5240938758505763,
+      "grad_norm": 4.657955207740734,
+      "learning_rate": 4.8532754718759475e-06,
+      "loss": 0.6324,
+      "step": 7548
+    },
+    {
+      "epoch": 0.5241633106512984,
+      "grad_norm": 4.274213761246854,
+      "learning_rate": 4.852151472549021e-06,
+      "loss": 0.5593,
+      "step": 7549
+    },
+    {
+      "epoch": 0.5242327454520206,
+      "grad_norm": 3.8313621344046433,
+      "learning_rate": 4.851027480700107e-06,
+      "loss": 0.4469,
+      "step": 7550
+    },
+    {
+      "epoch": 0.5243021802527427,
+      "grad_norm": 3.4638806955300536,
+      "learning_rate": 4.849903496386052e-06,
+      "loss": 0.4327,
+      "step": 7551
+    },
+    {
+      "epoch": 0.5243716150534647,
+      "grad_norm": 3.634738002569708,
+      "learning_rate": 4.848779519663712e-06,
+      "loss": 0.4888,
+      "step": 7552
+    },
+    {
+      "epoch": 0.5244410498541869,
+      "grad_norm": 3.777626123264553,
+      "learning_rate": 4.8476555505899314e-06,
+      "loss": 0.2734,
+      "step": 7553
+    },
+    {
+      "epoch": 0.524510484654909,
+      "grad_norm": 2.909240926714825,
+      "learning_rate": 4.8465315892215595e-06,
+      "loss": 0.3226,
+      "step": 7554
+    },
+    {
+      "epoch": 0.5245799194556312,
+      "grad_norm": 4.260706526750719,
+      "learning_rate": 4.845407635615448e-06,
+      "loss": 0.4559,
+      "step": 7555
+    },
+    {
+      "epoch": 0.5246493542563533,
+      "grad_norm": 3.526066805594896,
+      "learning_rate": 4.844283689828441e-06,
+      "loss": 0.2152,
+      "step": 7556
+    },
+    {
+      "epoch": 0.5247187890570754,
+      "grad_norm": 3.281158577202449,
+      "learning_rate": 4.843159751917391e-06,
+      "loss": 0.3405,
+      "step": 7557
+    },
+    {
+      "epoch": 0.5247882238577976,
+      "grad_norm": 3.703555121517168,
+      "learning_rate": 4.842035821939141e-06,
+      "loss": 0.5067,
+      "step": 7558
+    },
+    {
+      "epoch": 0.5248576586585196,
+      "grad_norm": 2.2924320746164124,
+      "learning_rate": 4.8409118999505435e-06,
+      "loss": 0.2614,
+      "step": 7559
+    },
+    {
+      "epoch": 0.5249270934592418,
+      "grad_norm": 3.277476625894575,
+      "learning_rate": 4.83978798600844e-06,
+      "loss": 0.3365,
+      "step": 7560
+    },
+    {
+      "epoch": 0.5249965282599639,
+      "grad_norm": 4.826216468081945,
+      "learning_rate": 4.8386640801696786e-06,
+      "loss": 0.458,
+      "step": 7561
+    },
+    {
+      "epoch": 0.525065963060686,
+      "grad_norm": 3.3247688162565274,
+      "learning_rate": 4.8375401824911064e-06,
+      "loss": 0.245,
+      "step": 7562
+    },
+    {
+      "epoch": 0.5251353978614082,
+      "grad_norm": 4.1948600140664585,
+      "learning_rate": 4.836416293029569e-06,
+      "loss": 0.4996,
+      "step": 7563
+    },
+    {
+      "epoch": 0.5252048326621302,
+      "grad_norm": 4.355933448161882,
+      "learning_rate": 4.835292411841907e-06,
+      "loss": 0.5548,
+      "step": 7564
+    },
+    {
+      "epoch": 0.5252742674628523,
+      "grad_norm": 4.6402188376246905,
+      "learning_rate": 4.83416853898497e-06,
+      "loss": 0.7553,
+      "step": 7565
+    },
+    {
+      "epoch": 0.5253437022635745,
+      "grad_norm": 2.9360575364270263,
+      "learning_rate": 4.833044674515603e-06,
+      "loss": 0.3392,
+      "step": 7566
+    },
+    {
+      "epoch": 0.5254131370642966,
+      "grad_norm": 2.948554253163831,
+      "learning_rate": 4.831920818490645e-06,
+      "loss": 0.3724,
+      "step": 7567
+    },
+    {
+      "epoch": 0.5254825718650188,
+      "grad_norm": 4.466094060872265,
+      "learning_rate": 4.830796970966942e-06,
+      "loss": 0.3714,
+      "step": 7568
+    },
+    {
+      "epoch": 0.5255520066657409,
+      "grad_norm": 4.7105727525794014,
+      "learning_rate": 4.829673132001337e-06,
+      "loss": 0.4727,
+      "step": 7569
+    },
+    {
+      "epoch": 0.525621441466463,
+      "grad_norm": 4.557010346927838,
+      "learning_rate": 4.828549301650673e-06,
+      "loss": 0.6636,
+      "step": 7570
+    },
+    {
+      "epoch": 0.5256908762671851,
+      "grad_norm": 4.24206942199176,
+      "learning_rate": 4.82742547997179e-06,
+      "loss": 0.4691,
+      "step": 7571
+    },
+    {
+      "epoch": 0.5257603110679072,
+      "grad_norm": 3.560349152013464,
+      "learning_rate": 4.826301667021532e-06,
+      "loss": 0.4934,
+      "step": 7572
+    },
+    {
+      "epoch": 0.5258297458686294,
+      "grad_norm": 2.848386125812199,
+      "learning_rate": 4.8251778628567395e-06,
+      "loss": 0.2237,
+      "step": 7573
+    },
+    {
+      "epoch": 0.5258991806693515,
+      "grad_norm": 3.041182188505059,
+      "learning_rate": 4.824054067534251e-06,
+      "loss": 0.2207,
+      "step": 7574
+    },
+    {
+      "epoch": 0.5259686154700736,
+      "grad_norm": 4.026251639351946,
+      "learning_rate": 4.822930281110909e-06,
+      "loss": 0.391,
+      "step": 7575
+    },
+    {
+      "epoch": 0.5260380502707958,
+      "grad_norm": 4.02501985722207,
+      "learning_rate": 4.821806503643554e-06,
+      "loss": 0.503,
+      "step": 7576
+    },
+    {
+      "epoch": 0.5261074850715178,
+      "grad_norm": 4.426648874782034,
+      "learning_rate": 4.820682735189025e-06,
+      "loss": 0.5981,
+      "step": 7577
+    },
+    {
+      "epoch": 0.5261769198722399,
+      "grad_norm": 3.7160441872205165,
+      "learning_rate": 4.819558975804159e-06,
+      "loss": 0.3262,
+      "step": 7578
+    },
+    {
+      "epoch": 0.5262463546729621,
+      "grad_norm": 2.970748354883217,
+      "learning_rate": 4.818435225545796e-06,
+      "loss": 0.3478,
+      "step": 7579
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 3.9234364557102133,
+      "learning_rate": 4.817311484470776e-06,
+      "loss": 0.3742,
+      "step": 7580
+    },
+    {
+      "epoch": 0.5263852242744064,
+      "grad_norm": 3.209848973909349,
+      "learning_rate": 4.816187752635931e-06,
+      "loss": 0.32,
+      "step": 7581
+    },
+    {
+      "epoch": 0.5264546590751285,
+      "grad_norm": 15.826325501719342,
+      "learning_rate": 4.815064030098103e-06,
+      "loss": 0.2911,
+      "step": 7582
+    },
+    {
+      "epoch": 0.5265240938758505,
+      "grad_norm": 2.92216938282492,
+      "learning_rate": 4.813940316914127e-06,
+      "loss": 0.316,
+      "step": 7583
+    },
+    {
+      "epoch": 0.5265935286765727,
+      "grad_norm": 3.3176866862206076,
+      "learning_rate": 4.81281661314084e-06,
+      "loss": 0.4231,
+      "step": 7584
+    },
+    {
+      "epoch": 0.5266629634772948,
+      "grad_norm": 2.8538681242744928,
+      "learning_rate": 4.811692918835075e-06,
+      "loss": 0.2186,
+      "step": 7585
+    },
+    {
+      "epoch": 0.526732398278017,
+      "grad_norm": 2.5661296163764065,
+      "learning_rate": 4.810569234053671e-06,
+      "loss": 0.2495,
+      "step": 7586
+    },
+    {
+      "epoch": 0.5268018330787391,
+      "grad_norm": 4.428315134144549,
+      "learning_rate": 4.809445558853459e-06,
+      "loss": 0.4703,
+      "step": 7587
+    },
+    {
+      "epoch": 0.5268712678794611,
+      "grad_norm": 2.953286961936975,
+      "learning_rate": 4.8083218932912776e-06,
+      "loss": 0.3682,
+      "step": 7588
+    },
+    {
+      "epoch": 0.5269407026801833,
+      "grad_norm": 3.8433508467282036,
+      "learning_rate": 4.807198237423957e-06,
+      "loss": 0.5439,
+      "step": 7589
+    },
+    {
+      "epoch": 0.5270101374809054,
+      "grad_norm": 4.538520863946223,
+      "learning_rate": 4.80607459130833e-06,
+      "loss": 0.4865,
+      "step": 7590
+    },
+    {
+      "epoch": 0.5270795722816276,
+      "grad_norm": 3.178012338165663,
+      "learning_rate": 4.804950955001232e-06,
+      "loss": 0.2528,
+      "step": 7591
+    },
+    {
+      "epoch": 0.5271490070823497,
+      "grad_norm": 4.176085519729535,
+      "learning_rate": 4.803827328559494e-06,
+      "loss": 0.3465,
+      "step": 7592
+    },
+    {
+      "epoch": 0.5272184418830718,
+      "grad_norm": 3.376475931668156,
+      "learning_rate": 4.802703712039946e-06,
+      "loss": 0.5196,
+      "step": 7593
+    },
+    {
+      "epoch": 0.527287876683794,
+      "grad_norm": 3.8417306555503066,
+      "learning_rate": 4.801580105499422e-06,
+      "loss": 0.3458,
+      "step": 7594
+    },
+    {
+      "epoch": 0.527357311484516,
+      "grad_norm": 3.8340198618244594,
+      "learning_rate": 4.800456508994752e-06,
+      "loss": 0.4164,
+      "step": 7595
+    },
+    {
+      "epoch": 0.5274267462852381,
+      "grad_norm": 4.601995105973886,
+      "learning_rate": 4.799332922582767e-06,
+      "loss": 0.6385,
+      "step": 7596
+    },
+    {
+      "epoch": 0.5274961810859603,
+      "grad_norm": 4.629812581629832,
+      "learning_rate": 4.7982093463202935e-06,
+      "loss": 0.4247,
+      "step": 7597
+    },
+    {
+      "epoch": 0.5275656158866824,
+      "grad_norm": 4.193230211695929,
+      "learning_rate": 4.797085780264166e-06,
+      "loss": 0.3899,
+      "step": 7598
+    },
+    {
+      "epoch": 0.5276350506874046,
+      "grad_norm": 3.6981027121400376,
+      "learning_rate": 4.7959622244712075e-06,
+      "loss": 0.4659,
+      "step": 7599
+    },
+    {
+      "epoch": 0.5277044854881267,
+      "grad_norm": 5.287201246304581,
+      "learning_rate": 4.794838678998249e-06,
+      "loss": 0.9563,
+      "step": 7600
+    },
+    {
+      "epoch": 0.5277739202888487,
+      "grad_norm": 4.908797808238697,
+      "learning_rate": 4.793715143902119e-06,
+      "loss": 0.6463,
+      "step": 7601
+    },
+    {
+      "epoch": 0.5278433550895709,
+      "grad_norm": 3.0148535977925417,
+      "learning_rate": 4.792591619239644e-06,
+      "loss": 0.3651,
+      "step": 7602
+    },
+    {
+      "epoch": 0.527912789890293,
+      "grad_norm": 3.4765397181752475,
+      "learning_rate": 4.791468105067648e-06,
+      "loss": 0.5398,
+      "step": 7603
+    },
+    {
+      "epoch": 0.5279822246910152,
+      "grad_norm": 3.6370504942297273,
+      "learning_rate": 4.79034460144296e-06,
+      "loss": 0.4001,
+      "step": 7604
+    },
+    {
+      "epoch": 0.5280516594917373,
+      "grad_norm": 3.587560445602243,
+      "learning_rate": 4.789221108422407e-06,
+      "loss": 0.537,
+      "step": 7605
+    },
+    {
+      "epoch": 0.5281210942924593,
+      "grad_norm": 2.5616184767807617,
+      "learning_rate": 4.78809762606281e-06,
+      "loss": 0.2123,
+      "step": 7606
+    },
+    {
+      "epoch": 0.5281905290931815,
+      "grad_norm": 4.546211942773034,
+      "learning_rate": 4.786974154420994e-06,
+      "loss": 0.5091,
+      "step": 7607
+    },
+    {
+      "epoch": 0.5282599638939036,
+      "grad_norm": 3.4442057259691006,
+      "learning_rate": 4.785850693553786e-06,
+      "loss": 0.4746,
+      "step": 7608
+    },
+    {
+      "epoch": 0.5283293986946257,
+      "grad_norm": 4.490318720086395,
+      "learning_rate": 4.784727243518009e-06,
+      "loss": 0.4585,
+      "step": 7609
+    },
+    {
+      "epoch": 0.5283988334953479,
+      "grad_norm": 4.633471245654575,
+      "learning_rate": 4.783603804370481e-06,
+      "loss": 0.6022,
+      "step": 7610
+    },
+    {
+      "epoch": 0.52846826829607,
+      "grad_norm": 4.316620574197479,
+      "learning_rate": 4.78248037616803e-06,
+      "loss": 0.6819,
+      "step": 7611
+    },
+    {
+      "epoch": 0.5285377030967922,
+      "grad_norm": 4.17672287241586,
+      "learning_rate": 4.781356958967474e-06,
+      "loss": 0.7453,
+      "step": 7612
+    },
+    {
+      "epoch": 0.5286071378975142,
+      "grad_norm": 4.364221692216325,
+      "learning_rate": 4.780233552825638e-06,
+      "loss": 0.6576,
+      "step": 7613
+    },
+    {
+      "epoch": 0.5286765726982363,
+      "grad_norm": 3.184957002896794,
+      "learning_rate": 4.779110157799337e-06,
+      "loss": 0.4321,
+      "step": 7614
+    },
+    {
+      "epoch": 0.5287460074989585,
+      "grad_norm": 5.264073496133862,
+      "learning_rate": 4.777986773945396e-06,
+      "loss": 0.7561,
+      "step": 7615
+    },
+    {
+      "epoch": 0.5288154422996806,
+      "grad_norm": 6.599535217025938,
+      "learning_rate": 4.776863401320635e-06,
+      "loss": 0.8908,
+      "step": 7616
+    },
+    {
+      "epoch": 0.5288848771004028,
+      "grad_norm": 3.5496142758851956,
+      "learning_rate": 4.775740039981867e-06,
+      "loss": 0.4606,
+      "step": 7617
+    },
+    {
+      "epoch": 0.5289543119011249,
+      "grad_norm": 3.5300314840032723,
+      "learning_rate": 4.774616689985916e-06,
+      "loss": 0.382,
+      "step": 7618
+    },
+    {
+      "epoch": 0.5290237467018469,
+      "grad_norm": 3.6451272995713273,
+      "learning_rate": 4.773493351389596e-06,
+      "loss": 0.5546,
+      "step": 7619
+    },
+    {
+      "epoch": 0.5290931815025691,
+      "grad_norm": 4.482994235308365,
+      "learning_rate": 4.772370024249728e-06,
+      "loss": 0.7301,
+      "step": 7620
+    },
+    {
+      "epoch": 0.5291626163032912,
+      "grad_norm": 3.3104289437322407,
+      "learning_rate": 4.7712467086231255e-06,
+      "loss": 0.4218,
+      "step": 7621
+    },
+    {
+      "epoch": 0.5292320511040133,
+      "grad_norm": 4.323829170807633,
+      "learning_rate": 4.770123404566605e-06,
+      "loss": 0.6127,
+      "step": 7622
+    },
+    {
+      "epoch": 0.5293014859047355,
+      "grad_norm": 2.729079520417261,
+      "learning_rate": 4.769000112136985e-06,
+      "loss": 0.147,
+      "step": 7623
+    },
+    {
+      "epoch": 0.5293709207054575,
+      "grad_norm": 3.6893090363571086,
+      "learning_rate": 4.767876831391075e-06,
+      "loss": 0.3154,
+      "step": 7624
+    },
+    {
+      "epoch": 0.5294403555061797,
+      "grad_norm": 3.971447487848771,
+      "learning_rate": 4.7667535623856945e-06,
+      "loss": 0.3691,
+      "step": 7625
+    },
+    {
+      "epoch": 0.5295097903069018,
+      "grad_norm": 4.281544982294868,
+      "learning_rate": 4.765630305177652e-06,
+      "loss": 0.6053,
+      "step": 7626
+    },
+    {
+      "epoch": 0.5295792251076239,
+      "grad_norm": 3.81483841632714,
+      "learning_rate": 4.764507059823767e-06,
+      "loss": 0.4556,
+      "step": 7627
+    },
+    {
+      "epoch": 0.5296486599083461,
+      "grad_norm": 3.0557442397083245,
+      "learning_rate": 4.763383826380848e-06,
+      "loss": 0.3772,
+      "step": 7628
+    },
+    {
+      "epoch": 0.5297180947090682,
+      "grad_norm": 3.8671314498525673,
+      "learning_rate": 4.762260604905704e-06,
+      "loss": 0.5598,
+      "step": 7629
+    },
+    {
+      "epoch": 0.5297875295097904,
+      "grad_norm": 2.3940355617309,
+      "learning_rate": 4.761137395455153e-06,
+      "loss": 0.1093,
+      "step": 7630
+    },
+    {
+      "epoch": 0.5298569643105124,
+      "grad_norm": 4.148900277553757,
+      "learning_rate": 4.760014198086001e-06,
+      "loss": 0.4845,
+      "step": 7631
+    },
+    {
+      "epoch": 0.5299263991112345,
+      "grad_norm": 4.487238401018542,
+      "learning_rate": 4.758891012855059e-06,
+      "loss": 0.5781,
+      "step": 7632
+    },
+    {
+      "epoch": 0.5299958339119567,
+      "grad_norm": 3.6223210433942787,
+      "learning_rate": 4.757767839819137e-06,
+      "loss": 0.2489,
+      "step": 7633
+    },
+    {
+      "epoch": 0.5300652687126788,
+      "grad_norm": 3.279842795484512,
+      "learning_rate": 4.756644679035045e-06,
+      "loss": 0.4866,
+      "step": 7634
+    },
+    {
+      "epoch": 0.5301347035134009,
+      "grad_norm": 3.482369081542808,
+      "learning_rate": 4.755521530559587e-06,
+      "loss": 0.4453,
+      "step": 7635
+    },
+    {
+      "epoch": 0.530204138314123,
+      "grad_norm": 3.7935855730939383,
+      "learning_rate": 4.754398394449573e-06,
+      "loss": 0.6607,
+      "step": 7636
+    },
+    {
+      "epoch": 0.5302735731148451,
+      "grad_norm": 2.7895429220366323,
+      "learning_rate": 4.753275270761814e-06,
+      "loss": 0.2803,
+      "step": 7637
+    },
+    {
+      "epoch": 0.5303430079155673,
+      "grad_norm": 4.264072129495534,
+      "learning_rate": 4.752152159553111e-06,
+      "loss": 0.3608,
+      "step": 7638
+    },
+    {
+      "epoch": 0.5304124427162894,
+      "grad_norm": 3.4616402154084134,
+      "learning_rate": 4.7510290608802695e-06,
+      "loss": 0.4841,
+      "step": 7639
+    },
+    {
+      "epoch": 0.5304818775170115,
+      "grad_norm": 3.3924460325273893,
+      "learning_rate": 4.749905974800098e-06,
+      "loss": 0.3477,
+      "step": 7640
+    },
+    {
+      "epoch": 0.5305513123177337,
+      "grad_norm": 3.8219978207962946,
+      "learning_rate": 4.7487829013694e-06,
+      "loss": 0.4219,
+      "step": 7641
+    },
+    {
+      "epoch": 0.5306207471184557,
+      "grad_norm": 3.4408200776477345,
+      "learning_rate": 4.747659840644977e-06,
+      "loss": 0.2711,
+      "step": 7642
+    },
+    {
+      "epoch": 0.5306901819191779,
+      "grad_norm": 4.012072261374291,
+      "learning_rate": 4.746536792683635e-06,
+      "loss": 0.6345,
+      "step": 7643
+    },
+    {
+      "epoch": 0.5307596167199,
+      "grad_norm": 4.129265164399618,
+      "learning_rate": 4.745413757542173e-06,
+      "loss": 0.223,
+      "step": 7644
+    },
+    {
+      "epoch": 0.5308290515206221,
+      "grad_norm": 3.7250093689686774,
+      "learning_rate": 4.744290735277398e-06,
+      "loss": 0.1672,
+      "step": 7645
+    },
+    {
+      "epoch": 0.5308984863213443,
+      "grad_norm": 3.4691721261954522,
+      "learning_rate": 4.7431677259461065e-06,
+      "loss": 0.3186,
+      "step": 7646
+    },
+    {
+      "epoch": 0.5309679211220664,
+      "grad_norm": 2.9930076097453617,
+      "learning_rate": 4.7420447296051005e-06,
+      "loss": 0.3577,
+      "step": 7647
+    },
+    {
+      "epoch": 0.5310373559227886,
+      "grad_norm": 3.0937243809183848,
+      "learning_rate": 4.740921746311183e-06,
+      "loss": 0.2283,
+      "step": 7648
+    },
+    {
+      "epoch": 0.5311067907235106,
+      "grad_norm": 2.6757122504765234,
+      "learning_rate": 4.739798776121147e-06,
+      "loss": 0.3811,
+      "step": 7649
+    },
+    {
+      "epoch": 0.5311762255242327,
+      "grad_norm": 4.6182376935736995,
+      "learning_rate": 4.738675819091795e-06,
+      "loss": 0.3601,
+      "step": 7650
+    },
+    {
+      "epoch": 0.5312456603249549,
+      "grad_norm": 3.8754991444887446,
+      "learning_rate": 4.7375528752799245e-06,
+      "loss": 0.4221,
+      "step": 7651
+    },
+    {
+      "epoch": 0.531315095125677,
+      "grad_norm": 3.482607727905127,
+      "learning_rate": 4.736429944742335e-06,
+      "loss": 0.3851,
+      "step": 7652
+    },
+    {
+      "epoch": 0.5313845299263991,
+      "grad_norm": 3.5451539100047142,
+      "learning_rate": 4.735307027535819e-06,
+      "loss": 0.4364,
+      "step": 7653
+    },
+    {
+      "epoch": 0.5314539647271213,
+      "grad_norm": 3.0033916656801196,
+      "learning_rate": 4.734184123717172e-06,
+      "loss": 0.2232,
+      "step": 7654
+    },
+    {
+      "epoch": 0.5315233995278433,
+      "grad_norm": 3.8129059864485484,
+      "learning_rate": 4.733061233343195e-06,
+      "loss": 0.4561,
+      "step": 7655
+    },
+    {
+      "epoch": 0.5315928343285655,
+      "grad_norm": 4.463133125395323,
+      "learning_rate": 4.731938356470675e-06,
+      "loss": 0.5236,
+      "step": 7656
+    },
+    {
+      "epoch": 0.5316622691292876,
+      "grad_norm": 3.5203712411289576,
+      "learning_rate": 4.730815493156412e-06,
+      "loss": 0.3622,
+      "step": 7657
+    },
+    {
+      "epoch": 0.5317317039300097,
+      "grad_norm": 3.809319442148734,
+      "learning_rate": 4.729692643457196e-06,
+      "loss": 0.3419,
+      "step": 7658
+    },
+    {
+      "epoch": 0.5318011387307319,
+      "grad_norm": 3.1573776806496032,
+      "learning_rate": 4.728569807429821e-06,
+      "loss": 0.3349,
+      "step": 7659
+    },
+    {
+      "epoch": 0.531870573531454,
+      "grad_norm": 2.6179847221489996,
+      "learning_rate": 4.727446985131078e-06,
+      "loss": 0.1373,
+      "step": 7660
+    },
+    {
+      "epoch": 0.5319400083321761,
+      "grad_norm": 3.1789125892452685,
+      "learning_rate": 4.726324176617757e-06,
+      "loss": 0.3723,
+      "step": 7661
+    },
+    {
+      "epoch": 0.5320094431328982,
+      "grad_norm": 5.235128873493515,
+      "learning_rate": 4.72520138194665e-06,
+      "loss": 0.5147,
+      "step": 7662
+    },
+    {
+      "epoch": 0.5320788779336203,
+      "grad_norm": 4.3924821990763485,
+      "learning_rate": 4.724078601174546e-06,
+      "loss": 0.5038,
+      "step": 7663
+    },
+    {
+      "epoch": 0.5321483127343425,
+      "grad_norm": 3.3363166469608854,
+      "learning_rate": 4.722955834358234e-06,
+      "loss": 0.327,
+      "step": 7664
+    },
+    {
+      "epoch": 0.5322177475350646,
+      "grad_norm": 3.4566621635125263,
+      "learning_rate": 4.721833081554502e-06,
+      "loss": 0.5327,
+      "step": 7665
+    },
+    {
+      "epoch": 0.5322871823357866,
+      "grad_norm": 3.4473483043462965,
+      "learning_rate": 4.72071034282014e-06,
+      "loss": 0.2266,
+      "step": 7666
+    },
+    {
+      "epoch": 0.5323566171365088,
+      "grad_norm": 3.781971859630506,
+      "learning_rate": 4.719587618211931e-06,
+      "loss": 0.6296,
+      "step": 7667
+    },
+    {
+      "epoch": 0.5324260519372309,
+      "grad_norm": 3.6206056143832575,
+      "learning_rate": 4.718464907786662e-06,
+      "loss": 0.2911,
+      "step": 7668
+    },
+    {
+      "epoch": 0.5324954867379531,
+      "grad_norm": 3.7255293777568648,
+      "learning_rate": 4.717342211601121e-06,
+      "loss": 0.3603,
+      "step": 7669
+    },
+    {
+      "epoch": 0.5325649215386752,
+      "grad_norm": 4.319241156800124,
+      "learning_rate": 4.716219529712092e-06,
+      "loss": 0.4791,
+      "step": 7670
+    },
+    {
+      "epoch": 0.5326343563393973,
+      "grad_norm": 3.866847180484654,
+      "learning_rate": 4.715096862176356e-06,
+      "loss": 0.3733,
+      "step": 7671
+    },
+    {
+      "epoch": 0.5327037911401195,
+      "grad_norm": 4.7365965469779585,
+      "learning_rate": 4.713974209050699e-06,
+      "loss": 0.4856,
+      "step": 7672
+    },
+    {
+      "epoch": 0.5327732259408415,
+      "grad_norm": 2.869318549935514,
+      "learning_rate": 4.712851570391905e-06,
+      "loss": 0.1557,
+      "step": 7673
+    },
+    {
+      "epoch": 0.5328426607415637,
+      "grad_norm": 3.3944422745507588,
+      "learning_rate": 4.711728946256751e-06,
+      "loss": 0.3371,
+      "step": 7674
+    },
+    {
+      "epoch": 0.5329120955422858,
+      "grad_norm": 3.2953735733597553,
+      "learning_rate": 4.71060633670202e-06,
+      "loss": 0.2186,
+      "step": 7675
+    },
+    {
+      "epoch": 0.5329815303430079,
+      "grad_norm": 4.044717213004769,
+      "learning_rate": 4.7094837417844946e-06,
+      "loss": 0.6493,
+      "step": 7676
+    },
+    {
+      "epoch": 0.5330509651437301,
+      "grad_norm": 4.385837450981106,
+      "learning_rate": 4.708361161560954e-06,
+      "loss": 0.609,
+      "step": 7677
+    },
+    {
+      "epoch": 0.5331203999444522,
+      "grad_norm": 3.381586300161956,
+      "learning_rate": 4.707238596088173e-06,
+      "loss": 0.3276,
+      "step": 7678
+    },
+    {
+      "epoch": 0.5331898347451742,
+      "grad_norm": 3.327144351858488,
+      "learning_rate": 4.706116045422935e-06,
+      "loss": 0.4117,
+      "step": 7679
+    },
+    {
+      "epoch": 0.5332592695458964,
+      "grad_norm": 4.5291930560703735,
+      "learning_rate": 4.704993509622014e-06,
+      "loss": 0.3867,
+      "step": 7680
+    },
+    {
+      "epoch": 0.5333287043466185,
+      "grad_norm": 4.488588201939917,
+      "learning_rate": 4.703870988742187e-06,
+      "loss": 0.4146,
+      "step": 7681
+    },
+    {
+      "epoch": 0.5333981391473407,
+      "grad_norm": 3.849611191232792,
+      "learning_rate": 4.702748482840231e-06,
+      "loss": 0.3757,
+      "step": 7682
+    },
+    {
+      "epoch": 0.5334675739480628,
+      "grad_norm": 2.903008397490526,
+      "learning_rate": 4.701625991972919e-06,
+      "loss": 0.2998,
+      "step": 7683
+    },
+    {
+      "epoch": 0.5335370087487848,
+      "grad_norm": 2.5369271531814204,
+      "learning_rate": 4.700503516197029e-06,
+      "loss": 0.1545,
+      "step": 7684
+    },
+    {
+      "epoch": 0.533606443549507,
+      "grad_norm": 3.4001771146595092,
+      "learning_rate": 4.69938105556933e-06,
+      "loss": 0.212,
+      "step": 7685
+    },
+    {
+      "epoch": 0.5336758783502291,
+      "grad_norm": 3.30503283862454,
+      "learning_rate": 4.6982586101465985e-06,
+      "loss": 0.3887,
+      "step": 7686
+    },
+    {
+      "epoch": 0.5337453131509513,
+      "grad_norm": 2.230267232281318,
+      "learning_rate": 4.697136179985605e-06,
+      "loss": 0.2099,
+      "step": 7687
+    },
+    {
+      "epoch": 0.5338147479516734,
+      "grad_norm": 4.1876891419692575,
+      "learning_rate": 4.69601376514312e-06,
+      "loss": 0.5612,
+      "step": 7688
+    },
+    {
+      "epoch": 0.5338841827523955,
+      "grad_norm": 5.073850885955729,
+      "learning_rate": 4.694891365675915e-06,
+      "loss": 0.5868,
+      "step": 7689
+    },
+    {
+      "epoch": 0.5339536175531177,
+      "grad_norm": 4.44110986471076,
+      "learning_rate": 4.693768981640758e-06,
+      "loss": 0.5934,
+      "step": 7690
+    },
+    {
+      "epoch": 0.5340230523538397,
+      "grad_norm": 3.250650351202611,
+      "learning_rate": 4.692646613094422e-06,
+      "loss": 0.3002,
+      "step": 7691
+    },
+    {
+      "epoch": 0.5340924871545618,
+      "grad_norm": 3.9513613706719224,
+      "learning_rate": 4.691524260093672e-06,
+      "loss": 0.5048,
+      "step": 7692
+    },
+    {
+      "epoch": 0.534161921955284,
+      "grad_norm": 4.705778722531619,
+      "learning_rate": 4.690401922695274e-06,
+      "loss": 0.3941,
+      "step": 7693
+    },
+    {
+      "epoch": 0.5342313567560061,
+      "grad_norm": 3.8176042294210735,
+      "learning_rate": 4.689279600955998e-06,
+      "loss": 0.393,
+      "step": 7694
+    },
+    {
+      "epoch": 0.5343007915567283,
+      "grad_norm": 4.641355428099003,
+      "learning_rate": 4.688157294932608e-06,
+      "loss": 0.5573,
+      "step": 7695
+    },
+    {
+      "epoch": 0.5343702263574504,
+      "grad_norm": 3.4015256019912607,
+      "learning_rate": 4.6870350046818705e-06,
+      "loss": 0.4906,
+      "step": 7696
+    },
+    {
+      "epoch": 0.5344396611581724,
+      "grad_norm": 3.7177063890413615,
+      "learning_rate": 4.685912730260545e-06,
+      "loss": 0.6134,
+      "step": 7697
+    },
+    {
+      "epoch": 0.5345090959588946,
+      "grad_norm": 3.5773231876750087,
+      "learning_rate": 4.684790471725402e-06,
+      "loss": 0.346,
+      "step": 7698
+    },
+    {
+      "epoch": 0.5345785307596167,
+      "grad_norm": 3.0379603806792947,
+      "learning_rate": 4.683668229133198e-06,
+      "loss": 0.4674,
+      "step": 7699
+    },
+    {
+      "epoch": 0.5346479655603389,
+      "grad_norm": 4.032505367124036,
+      "learning_rate": 4.682546002540697e-06,
+      "loss": 0.5441,
+      "step": 7700
+    },
+    {
+      "epoch": 0.534717400361061,
+      "grad_norm": 3.527588768043789,
+      "learning_rate": 4.68142379200466e-06,
+      "loss": 0.5461,
+      "step": 7701
+    },
+    {
+      "epoch": 0.534786835161783,
+      "grad_norm": 3.5795773442064904,
+      "learning_rate": 4.680301597581849e-06,
+      "loss": 0.365,
+      "step": 7702
+    },
+    {
+      "epoch": 0.5348562699625052,
+      "grad_norm": 3.4524466502656987,
+      "learning_rate": 4.679179419329019e-06,
+      "loss": 0.4785,
+      "step": 7703
+    },
+    {
+      "epoch": 0.5349257047632273,
+      "grad_norm": 3.5518359435491127,
+      "learning_rate": 4.678057257302931e-06,
+      "loss": 0.4363,
+      "step": 7704
+    },
+    {
+      "epoch": 0.5349951395639495,
+      "grad_norm": 5.136777772980805,
+      "learning_rate": 4.6769351115603445e-06,
+      "loss": 0.4172,
+      "step": 7705
+    },
+    {
+      "epoch": 0.5350645743646716,
+      "grad_norm": 3.9765180566040943,
+      "learning_rate": 4.675812982158013e-06,
+      "loss": 0.5559,
+      "step": 7706
+    },
+    {
+      "epoch": 0.5351340091653937,
+      "grad_norm": 26.38968430614968,
+      "learning_rate": 4.674690869152694e-06,
+      "loss": 0.3006,
+      "step": 7707
+    },
+    {
+      "epoch": 0.5352034439661159,
+      "grad_norm": 2.838714000868335,
+      "learning_rate": 4.673568772601143e-06,
+      "loss": 0.2575,
+      "step": 7708
+    },
+    {
+      "epoch": 0.5352728787668379,
+      "grad_norm": 3.4687534111998404,
+      "learning_rate": 4.672446692560116e-06,
+      "loss": 0.4209,
+      "step": 7709
+    },
+    {
+      "epoch": 0.53534231356756,
+      "grad_norm": 3.7265551307033853,
+      "learning_rate": 4.671324629086362e-06,
+      "loss": 0.2799,
+      "step": 7710
+    },
+    {
+      "epoch": 0.5354117483682822,
+      "grad_norm": 3.305952283692755,
+      "learning_rate": 4.670202582236637e-06,
+      "loss": 0.3792,
+      "step": 7711
+    },
+    {
+      "epoch": 0.5354811831690043,
+      "grad_norm": 4.291653013827585,
+      "learning_rate": 4.669080552067694e-06,
+      "loss": 0.5356,
+      "step": 7712
+    },
+    {
+      "epoch": 0.5355506179697265,
+      "grad_norm": 3.9962102549505794,
+      "learning_rate": 4.6679585386362795e-06,
+      "loss": 0.5538,
+      "step": 7713
+    },
+    {
+      "epoch": 0.5356200527704486,
+      "grad_norm": 3.017980999955834,
+      "learning_rate": 4.666836541999148e-06,
+      "loss": 0.3825,
+      "step": 7714
+    },
+    {
+      "epoch": 0.5356894875711706,
+      "grad_norm": 5.094235537153492,
+      "learning_rate": 4.665714562213045e-06,
+      "loss": 0.7998,
+      "step": 7715
+    },
+    {
+      "epoch": 0.5357589223718928,
+      "grad_norm": 3.5668450671897385,
+      "learning_rate": 4.664592599334725e-06,
+      "loss": 0.4263,
+      "step": 7716
+    },
+    {
+      "epoch": 0.5358283571726149,
+      "grad_norm": 3.3347559110882745,
+      "learning_rate": 4.6634706534209284e-06,
+      "loss": 0.4169,
+      "step": 7717
+    },
+    {
+      "epoch": 0.5358977919733371,
+      "grad_norm": 3.5257045728119025,
+      "learning_rate": 4.662348724528407e-06,
+      "loss": 0.4106,
+      "step": 7718
+    },
+    {
+      "epoch": 0.5359672267740592,
+      "grad_norm": 4.4474454759232405,
+      "learning_rate": 4.661226812713903e-06,
+      "loss": 0.6376,
+      "step": 7719
+    },
+    {
+      "epoch": 0.5360366615747812,
+      "grad_norm": 3.2134895603726723,
+      "learning_rate": 4.660104918034167e-06,
+      "loss": 0.2785,
+      "step": 7720
+    },
+    {
+      "epoch": 0.5361060963755034,
+      "grad_norm": 3.523472452190626,
+      "learning_rate": 4.658983040545938e-06,
+      "loss": 0.2902,
+      "step": 7721
+    },
+    {
+      "epoch": 0.5361755311762255,
+      "grad_norm": 3.410977745571911,
+      "learning_rate": 4.6578611803059594e-06,
+      "loss": 0.3642,
+      "step": 7722
+    },
+    {
+      "epoch": 0.5362449659769476,
+      "grad_norm": 2.5124714232768786,
+      "learning_rate": 4.656739337370978e-06,
+      "loss": 0.2007,
+      "step": 7723
+    },
+    {
+      "epoch": 0.5363144007776698,
+      "grad_norm": 3.745830488325483,
+      "learning_rate": 4.655617511797731e-06,
+      "loss": 0.4836,
+      "step": 7724
+    },
+    {
+      "epoch": 0.5363838355783919,
+      "grad_norm": 4.295268882219355,
+      "learning_rate": 4.654495703642959e-06,
+      "loss": 0.6661,
+      "step": 7725
+    },
+    {
+      "epoch": 0.5364532703791141,
+      "grad_norm": 4.076473026278984,
+      "learning_rate": 4.653373912963404e-06,
+      "loss": 0.4766,
+      "step": 7726
+    },
+    {
+      "epoch": 0.5365227051798361,
+      "grad_norm": 4.340279129528889,
+      "learning_rate": 4.652252139815807e-06,
+      "loss": 0.4945,
+      "step": 7727
+    },
+    {
+      "epoch": 0.5365921399805582,
+      "grad_norm": 3.5886032831450483,
+      "learning_rate": 4.651130384256901e-06,
+      "loss": 0.3673,
+      "step": 7728
+    },
+    {
+      "epoch": 0.5366615747812804,
+      "grad_norm": 3.453286487693722,
+      "learning_rate": 4.650008646343425e-06,
+      "loss": 0.3705,
+      "step": 7729
+    },
+    {
+      "epoch": 0.5367310095820025,
+      "grad_norm": 3.7908337695937786,
+      "learning_rate": 4.6488869261321175e-06,
+      "loss": 0.4063,
+      "step": 7730
+    },
+    {
+      "epoch": 0.5368004443827247,
+      "grad_norm": 2.955141196344212,
+      "learning_rate": 4.647765223679711e-06,
+      "loss": 0.3931,
+      "step": 7731
+    },
+    {
+      "epoch": 0.5368698791834468,
+      "grad_norm": 3.258880556476825,
+      "learning_rate": 4.64664353904294e-06,
+      "loss": 0.3063,
+      "step": 7732
+    },
+    {
+      "epoch": 0.5369393139841688,
+      "grad_norm": 3.6619120641441336,
+      "learning_rate": 4.6455218722785395e-06,
+      "loss": 0.4918,
+      "step": 7733
+    },
+    {
+      "epoch": 0.537008748784891,
+      "grad_norm": 5.004413325584721,
+      "learning_rate": 4.644400223443243e-06,
+      "loss": 0.6756,
+      "step": 7734
+    },
+    {
+      "epoch": 0.5370781835856131,
+      "grad_norm": 3.5466729757842073,
+      "learning_rate": 4.643278592593778e-06,
+      "loss": 0.4134,
+      "step": 7735
+    },
+    {
+      "epoch": 0.5371476183863352,
+      "grad_norm": 4.159772931134172,
+      "learning_rate": 4.642156979786879e-06,
+      "loss": 0.505,
+      "step": 7736
+    },
+    {
+      "epoch": 0.5372170531870574,
+      "grad_norm": 3.0506058871701685,
+      "learning_rate": 4.641035385079277e-06,
+      "loss": 0.3657,
+      "step": 7737
+    },
+    {
+      "epoch": 0.5372864879877794,
+      "grad_norm": 4.004025663918991,
+      "learning_rate": 4.6399138085276985e-06,
+      "loss": 0.3494,
+      "step": 7738
+    },
+    {
+      "epoch": 0.5373559227885016,
+      "grad_norm": 4.100908851495084,
+      "learning_rate": 4.6387922501888706e-06,
+      "loss": 0.6365,
+      "step": 7739
+    },
+    {
+      "epoch": 0.5374253575892237,
+      "grad_norm": 4.2170936361681495,
+      "learning_rate": 4.637670710119523e-06,
+      "loss": 0.6643,
+      "step": 7740
+    },
+    {
+      "epoch": 0.5374947923899458,
+      "grad_norm": 3.028322461942456,
+      "learning_rate": 4.636549188376381e-06,
+      "loss": 0.416,
+      "step": 7741
+    },
+    {
+      "epoch": 0.537564227190668,
+      "grad_norm": 4.591543547514798,
+      "learning_rate": 4.635427685016169e-06,
+      "loss": 0.5728,
+      "step": 7742
+    },
+    {
+      "epoch": 0.5376336619913901,
+      "grad_norm": 3.696756224258373,
+      "learning_rate": 4.6343062000956125e-06,
+      "loss": 0.4103,
+      "step": 7743
+    },
+    {
+      "epoch": 0.5377030967921123,
+      "grad_norm": 2.980852100919446,
+      "learning_rate": 4.633184733671433e-06,
+      "loss": 0.2286,
+      "step": 7744
+    },
+    {
+      "epoch": 0.5377725315928343,
+      "grad_norm": 3.0010691153029545,
+      "learning_rate": 4.632063285800358e-06,
+      "loss": 0.2447,
+      "step": 7745
+    },
+    {
+      "epoch": 0.5378419663935564,
+      "grad_norm": 4.291870237707775,
+      "learning_rate": 4.630941856539103e-06,
+      "loss": 0.4616,
+      "step": 7746
+    },
+    {
+      "epoch": 0.5379114011942786,
+      "grad_norm": 3.1002540774051073,
+      "learning_rate": 4.629820445944392e-06,
+      "loss": 0.2843,
+      "step": 7747
+    },
+    {
+      "epoch": 0.5379808359950007,
+      "grad_norm": 3.9640776032418827,
+      "learning_rate": 4.628699054072945e-06,
+      "loss": 0.5418,
+      "step": 7748
+    },
+    {
+      "epoch": 0.5380502707957228,
+      "grad_norm": 4.119176862089531,
+      "learning_rate": 4.627577680981478e-06,
+      "loss": 0.6691,
+      "step": 7749
+    },
+    {
+      "epoch": 0.538119705596445,
+      "grad_norm": 3.4080735134052036,
+      "learning_rate": 4.626456326726712e-06,
+      "loss": 0.2981,
+      "step": 7750
+    },
+    {
+      "epoch": 0.538189140397167,
+      "grad_norm": 2.3168713393825735,
+      "learning_rate": 4.625334991365361e-06,
+      "loss": 0.1928,
+      "step": 7751
+    },
+    {
+      "epoch": 0.5382585751978892,
+      "grad_norm": 3.056506382606844,
+      "learning_rate": 4.624213674954144e-06,
+      "loss": 0.2702,
+      "step": 7752
+    },
+    {
+      "epoch": 0.5383280099986113,
+      "grad_norm": 2.8041572825177146,
+      "learning_rate": 4.623092377549772e-06,
+      "loss": 0.2992,
+      "step": 7753
+    },
+    {
+      "epoch": 0.5383974447993334,
+      "grad_norm": 4.057377334564871,
+      "learning_rate": 4.621971099208961e-06,
+      "loss": 0.6518,
+      "step": 7754
+    },
+    {
+      "epoch": 0.5384668796000556,
+      "grad_norm": 3.4762119089042693,
+      "learning_rate": 4.620849839988426e-06,
+      "loss": 0.3528,
+      "step": 7755
+    },
+    {
+      "epoch": 0.5385363144007776,
+      "grad_norm": 4.943733974828772,
+      "learning_rate": 4.6197285999448755e-06,
+      "loss": 0.6247,
+      "step": 7756
+    },
+    {
+      "epoch": 0.5386057492014998,
+      "grad_norm": 4.372872721685097,
+      "learning_rate": 4.618607379135023e-06,
+      "loss": 0.4667,
+      "step": 7757
+    },
+    {
+      "epoch": 0.5386751840022219,
+      "grad_norm": 4.231452392392652,
+      "learning_rate": 4.617486177615576e-06,
+      "loss": 0.5437,
+      "step": 7758
+    },
+    {
+      "epoch": 0.538744618802944,
+      "grad_norm": 5.030879270436095,
+      "learning_rate": 4.616364995443248e-06,
+      "loss": 0.327,
+      "step": 7759
+    },
+    {
+      "epoch": 0.5388140536036662,
+      "grad_norm": 4.232128076290483,
+      "learning_rate": 4.615243832674743e-06,
+      "loss": 0.4787,
+      "step": 7760
+    },
+    {
+      "epoch": 0.5388834884043883,
+      "grad_norm": 3.8312330615921155,
+      "learning_rate": 4.614122689366769e-06,
+      "loss": 0.5087,
+      "step": 7761
+    },
+    {
+      "epoch": 0.5389529232051105,
+      "grad_norm": 4.372523599526809,
+      "learning_rate": 4.613001565576034e-06,
+      "loss": 0.7023,
+      "step": 7762
+    },
+    {
+      "epoch": 0.5390223580058325,
+      "grad_norm": 2.4090663749630292,
+      "learning_rate": 4.611880461359241e-06,
+      "loss": 0.2228,
+      "step": 7763
+    },
+    {
+      "epoch": 0.5390917928065546,
+      "grad_norm": 3.1187785104429095,
+      "learning_rate": 4.610759376773095e-06,
+      "loss": 0.3343,
+      "step": 7764
+    },
+    {
+      "epoch": 0.5391612276072768,
+      "grad_norm": 3.9413439002826975,
+      "learning_rate": 4.6096383118742986e-06,
+      "loss": 0.5463,
+      "step": 7765
+    },
+    {
+      "epoch": 0.5392306624079989,
+      "grad_norm": 3.348581764306762,
+      "learning_rate": 4.608517266719556e-06,
+      "loss": 0.3457,
+      "step": 7766
+    },
+    {
+      "epoch": 0.539300097208721,
+      "grad_norm": 5.236299105298392,
+      "learning_rate": 4.607396241365567e-06,
+      "loss": 0.6456,
+      "step": 7767
+    },
+    {
+      "epoch": 0.5393695320094432,
+      "grad_norm": 4.852551301895995,
+      "learning_rate": 4.606275235869031e-06,
+      "loss": 0.5471,
+      "step": 7768
+    },
+    {
+      "epoch": 0.5394389668101652,
+      "grad_norm": 3.63275820650788,
+      "learning_rate": 4.605154250286648e-06,
+      "loss": 0.4642,
+      "step": 7769
+    },
+    {
+      "epoch": 0.5395084016108874,
+      "grad_norm": 4.466145826048087,
+      "learning_rate": 4.604033284675118e-06,
+      "loss": 0.5292,
+      "step": 7770
+    },
+    {
+      "epoch": 0.5395778364116095,
+      "grad_norm": 3.8139071239048175,
+      "learning_rate": 4.602912339091134e-06,
+      "loss": 0.4908,
+      "step": 7771
+    },
+    {
+      "epoch": 0.5396472712123316,
+      "grad_norm": 4.4039152789258855,
+      "learning_rate": 4.601791413591396e-06,
+      "loss": 0.3488,
+      "step": 7772
+    },
+    {
+      "epoch": 0.5397167060130538,
+      "grad_norm": 3.0673191210392856,
+      "learning_rate": 4.600670508232599e-06,
+      "loss": 0.3073,
+      "step": 7773
+    },
+    {
+      "epoch": 0.5397861408137758,
+      "grad_norm": 3.359753617284215,
+      "learning_rate": 4.599549623071433e-06,
+      "loss": 0.31,
+      "step": 7774
+    },
+    {
+      "epoch": 0.539855575614498,
+      "grad_norm": 4.390795619418485,
+      "learning_rate": 4.598428758164594e-06,
+      "loss": 0.605,
+      "step": 7775
+    },
+    {
+      "epoch": 0.5399250104152201,
+      "grad_norm": 3.511963318432062,
+      "learning_rate": 4.597307913568775e-06,
+      "loss": 0.3442,
+      "step": 7776
+    },
+    {
+      "epoch": 0.5399944452159422,
+      "grad_norm": 4.0537972734714876,
+      "learning_rate": 4.5961870893406675e-06,
+      "loss": 0.5945,
+      "step": 7777
+    },
+    {
+      "epoch": 0.5400638800166644,
+      "grad_norm": 3.5042964140299055,
+      "learning_rate": 4.595066285536958e-06,
+      "loss": 0.4823,
+      "step": 7778
+    },
+    {
+      "epoch": 0.5401333148173865,
+      "grad_norm": 3.804070367279913,
+      "learning_rate": 4.593945502214339e-06,
+      "loss": 0.4052,
+      "step": 7779
+    },
+    {
+      "epoch": 0.5402027496181085,
+      "grad_norm": 2.877467191129682,
+      "learning_rate": 4.592824739429497e-06,
+      "loss": 0.2685,
+      "step": 7780
+    },
+    {
+      "epoch": 0.5402721844188307,
+      "grad_norm": 4.161668333031738,
+      "learning_rate": 4.591703997239118e-06,
+      "loss": 0.4799,
+      "step": 7781
+    },
+    {
+      "epoch": 0.5403416192195528,
+      "grad_norm": 3.993144969802978,
+      "learning_rate": 4.590583275699888e-06,
+      "loss": 0.5943,
+      "step": 7782
+    },
+    {
+      "epoch": 0.540411054020275,
+      "grad_norm": 3.3958560165694087,
+      "learning_rate": 4.589462574868493e-06,
+      "loss": 0.4757,
+      "step": 7783
+    },
+    {
+      "epoch": 0.5404804888209971,
+      "grad_norm": 2.999212949651576,
+      "learning_rate": 4.588341894801619e-06,
+      "loss": 0.2654,
+      "step": 7784
+    },
+    {
+      "epoch": 0.5405499236217192,
+      "grad_norm": 3.8120488966796438,
+      "learning_rate": 4.587221235555942e-06,
+      "loss": 0.6412,
+      "step": 7785
+    },
+    {
+      "epoch": 0.5406193584224414,
+      "grad_norm": 3.992937347172324,
+      "learning_rate": 4.586100597188151e-06,
+      "loss": 0.4797,
+      "step": 7786
+    },
+    {
+      "epoch": 0.5406887932231634,
+      "grad_norm": 4.580935596231863,
+      "learning_rate": 4.584979979754924e-06,
+      "loss": 0.5467,
+      "step": 7787
+    },
+    {
+      "epoch": 0.5407582280238856,
+      "grad_norm": 3.4949970486162525,
+      "learning_rate": 4.583859383312937e-06,
+      "loss": 0.4276,
+      "step": 7788
+    },
+    {
+      "epoch": 0.5408276628246077,
+      "grad_norm": 3.5389538263131177,
+      "learning_rate": 4.582738807918873e-06,
+      "loss": 0.3806,
+      "step": 7789
+    },
+    {
+      "epoch": 0.5408970976253298,
+      "grad_norm": 4.785179912872509,
+      "learning_rate": 4.581618253629408e-06,
+      "loss": 0.5611,
+      "step": 7790
+    },
+    {
+      "epoch": 0.540966532426052,
+      "grad_norm": 2.647674476657264,
+      "learning_rate": 4.580497720501219e-06,
+      "loss": 0.2475,
+      "step": 7791
+    },
+    {
+      "epoch": 0.541035967226774,
+      "grad_norm": 4.569117088219584,
+      "learning_rate": 4.57937720859098e-06,
+      "loss": 0.5439,
+      "step": 7792
+    },
+    {
+      "epoch": 0.5411054020274961,
+      "grad_norm": 3.95336936303858,
+      "learning_rate": 4.5782567179553645e-06,
+      "loss": 0.4036,
+      "step": 7793
+    },
+    {
+      "epoch": 0.5411748368282183,
+      "grad_norm": 4.391317913258363,
+      "learning_rate": 4.5771362486510464e-06,
+      "loss": 0.5631,
+      "step": 7794
+    },
+    {
+      "epoch": 0.5412442716289404,
+      "grad_norm": 4.082782999119234,
+      "learning_rate": 4.5760158007347024e-06,
+      "loss": 0.435,
+      "step": 7795
+    },
+    {
+      "epoch": 0.5413137064296626,
+      "grad_norm": 3.4416321255622924,
+      "learning_rate": 4.5748953742629965e-06,
+      "loss": 0.3792,
+      "step": 7796
+    },
+    {
+      "epoch": 0.5413831412303847,
+      "grad_norm": 3.9509235767153603,
+      "learning_rate": 4.573774969292601e-06,
+      "loss": 0.4394,
+      "step": 7797
+    },
+    {
+      "epoch": 0.5414525760311067,
+      "grad_norm": 4.059817996063551,
+      "learning_rate": 4.572654585880188e-06,
+      "loss": 0.5338,
+      "step": 7798
+    },
+    {
+      "epoch": 0.5415220108318289,
+      "grad_norm": 2.851290477009252,
+      "learning_rate": 4.571534224082421e-06,
+      "loss": 0.207,
+      "step": 7799
+    },
+    {
+      "epoch": 0.541591445632551,
+      "grad_norm": 3.9023556018040484,
+      "learning_rate": 4.570413883955967e-06,
+      "loss": 0.4765,
+      "step": 7800
+    },
+    {
+      "epoch": 0.5416608804332732,
+      "grad_norm": 2.877499804618252,
+      "learning_rate": 4.569293565557494e-06,
+      "loss": 0.3124,
+      "step": 7801
+    },
+    {
+      "epoch": 0.5417303152339953,
+      "grad_norm": 2.6779613446884825,
+      "learning_rate": 4.568173268943666e-06,
+      "loss": 0.2407,
+      "step": 7802
+    },
+    {
+      "epoch": 0.5417997500347174,
+      "grad_norm": 3.944465727418278,
+      "learning_rate": 4.567052994171142e-06,
+      "loss": 0.4367,
+      "step": 7803
+    },
+    {
+      "epoch": 0.5418691848354396,
+      "grad_norm": 3.8489436943239403,
+      "learning_rate": 4.56593274129659e-06,
+      "loss": 0.3924,
+      "step": 7804
+    },
+    {
+      "epoch": 0.5419386196361616,
+      "grad_norm": 5.346934484377518,
+      "learning_rate": 4.56481251037667e-06,
+      "loss": 0.7692,
+      "step": 7805
+    },
+    {
+      "epoch": 0.5420080544368837,
+      "grad_norm": 3.741945537500703,
+      "learning_rate": 4.563692301468038e-06,
+      "loss": 0.4884,
+      "step": 7806
+    },
+    {
+      "epoch": 0.5420774892376059,
+      "grad_norm": 3.343131613175298,
+      "learning_rate": 4.562572114627356e-06,
+      "loss": 0.2434,
+      "step": 7807
+    },
+    {
+      "epoch": 0.542146924038328,
+      "grad_norm": 4.323468480977452,
+      "learning_rate": 4.561451949911283e-06,
+      "loss": 0.614,
+      "step": 7808
+    },
+    {
+      "epoch": 0.5422163588390502,
+      "grad_norm": 4.175738385090236,
+      "learning_rate": 4.560331807376475e-06,
+      "loss": 0.31,
+      "step": 7809
+    },
+    {
+      "epoch": 0.5422857936397723,
+      "grad_norm": 4.091246703256656,
+      "learning_rate": 4.559211687079584e-06,
+      "loss": 0.4352,
+      "step": 7810
+    },
+    {
+      "epoch": 0.5423552284404943,
+      "grad_norm": 4.560568703125746,
+      "learning_rate": 4.558091589077268e-06,
+      "loss": 0.3646,
+      "step": 7811
+    },
+    {
+      "epoch": 0.5424246632412165,
+      "grad_norm": 3.578650024481626,
+      "learning_rate": 4.556971513426181e-06,
+      "loss": 0.298,
+      "step": 7812
+    },
+    {
+      "epoch": 0.5424940980419386,
+      "grad_norm": 4.787485870006128,
+      "learning_rate": 4.555851460182971e-06,
+      "loss": 0.6568,
+      "step": 7813
+    },
+    {
+      "epoch": 0.5425635328426608,
+      "grad_norm": 3.309158398136236,
+      "learning_rate": 4.554731429404293e-06,
+      "loss": 0.4776,
+      "step": 7814
+    },
+    {
+      "epoch": 0.5426329676433829,
+      "grad_norm": 4.460722245154039,
+      "learning_rate": 4.553611421146794e-06,
+      "loss": 0.5775,
+      "step": 7815
+    },
+    {
+      "epoch": 0.542702402444105,
+      "grad_norm": 3.9798967845419244,
+      "learning_rate": 4.552491435467126e-06,
+      "loss": 0.4332,
+      "step": 7816
+    },
+    {
+      "epoch": 0.5427718372448271,
+      "grad_norm": 6.24690165336592,
+      "learning_rate": 4.551371472421933e-06,
+      "loss": 0.5835,
+      "step": 7817
+    },
+    {
+      "epoch": 0.5428412720455492,
+      "grad_norm": 3.821024194378232,
+      "learning_rate": 4.550251532067865e-06,
+      "loss": 0.3577,
+      "step": 7818
+    },
+    {
+      "epoch": 0.5429107068462713,
+      "grad_norm": 3.2908906248024246,
+      "learning_rate": 4.549131614461564e-06,
+      "loss": 0.3094,
+      "step": 7819
+    },
+    {
+      "epoch": 0.5429801416469935,
+      "grad_norm": 4.509235617357524,
+      "learning_rate": 4.548011719659679e-06,
+      "loss": 0.3551,
+      "step": 7820
+    },
+    {
+      "epoch": 0.5430495764477156,
+      "grad_norm": 3.8538095685754064,
+      "learning_rate": 4.5468918477188474e-06,
+      "loss": 0.347,
+      "step": 7821
+    },
+    {
+      "epoch": 0.5431190112484378,
+      "grad_norm": 3.5227639195684954,
+      "learning_rate": 4.545771998695712e-06,
+      "loss": 0.4202,
+      "step": 7822
+    },
+    {
+      "epoch": 0.5431884460491598,
+      "grad_norm": 3.6155845215261824,
+      "learning_rate": 4.5446521726469185e-06,
+      "loss": 0.4865,
+      "step": 7823
+    },
+    {
+      "epoch": 0.5432578808498819,
+      "grad_norm": 3.6685302920354173,
+      "learning_rate": 4.5435323696291005e-06,
+      "loss": 0.4639,
+      "step": 7824
+    },
+    {
+      "epoch": 0.5433273156506041,
+      "grad_norm": 3.658261813317143,
+      "learning_rate": 4.542412589698898e-06,
+      "loss": 0.5128,
+      "step": 7825
+    },
+    {
+      "epoch": 0.5433967504513262,
+      "grad_norm": 4.296781719352591,
+      "learning_rate": 4.541292832912949e-06,
+      "loss": 0.4389,
+      "step": 7826
+    },
+    {
+      "epoch": 0.5434661852520484,
+      "grad_norm": 4.627301915242828,
+      "learning_rate": 4.540173099327892e-06,
+      "loss": 0.4105,
+      "step": 7827
+    },
+    {
+      "epoch": 0.5435356200527705,
+      "grad_norm": 4.397616599142978,
+      "learning_rate": 4.539053389000357e-06,
+      "loss": 0.8035,
+      "step": 7828
+    },
+    {
+      "epoch": 0.5436050548534925,
+      "grad_norm": 3.7429011610219924,
+      "learning_rate": 4.53793370198698e-06,
+      "loss": 0.4082,
+      "step": 7829
+    },
+    {
+      "epoch": 0.5436744896542147,
+      "grad_norm": 3.7622088692591458,
+      "learning_rate": 4.536814038344396e-06,
+      "loss": 0.5232,
+      "step": 7830
+    },
+    {
+      "epoch": 0.5437439244549368,
+      "grad_norm": 2.565176565544614,
+      "learning_rate": 4.535694398129232e-06,
+      "loss": 0.1621,
+      "step": 7831
+    },
+    {
+      "epoch": 0.543813359255659,
+      "grad_norm": 4.699676542639202,
+      "learning_rate": 4.534574781398119e-06,
+      "loss": 0.7175,
+      "step": 7832
+    },
+    {
+      "epoch": 0.5438827940563811,
+      "grad_norm": 3.857790470826273,
+      "learning_rate": 4.533455188207688e-06,
+      "loss": 0.4953,
+      "step": 7833
+    },
+    {
+      "epoch": 0.5439522288571031,
+      "grad_norm": 4.108780476502911,
+      "learning_rate": 4.5323356186145665e-06,
+      "loss": 0.4502,
+      "step": 7834
+    },
+    {
+      "epoch": 0.5440216636578253,
+      "grad_norm": 3.902921243273005,
+      "learning_rate": 4.531216072675378e-06,
+      "loss": 0.5658,
+      "step": 7835
+    },
+    {
+      "epoch": 0.5440910984585474,
+      "grad_norm": 4.072623988526568,
+      "learning_rate": 4.53009655044675e-06,
+      "loss": 0.4544,
+      "step": 7836
+    },
+    {
+      "epoch": 0.5441605332592695,
+      "grad_norm": 4.973514606390276,
+      "learning_rate": 4.5289770519853094e-06,
+      "loss": 0.628,
+      "step": 7837
+    },
+    {
+      "epoch": 0.5442299680599917,
+      "grad_norm": 3.2988403859413453,
+      "learning_rate": 4.527857577347675e-06,
+      "loss": 0.4276,
+      "step": 7838
+    },
+    {
+      "epoch": 0.5442994028607138,
+      "grad_norm": 3.531401392772517,
+      "learning_rate": 4.526738126590469e-06,
+      "loss": 0.4412,
+      "step": 7839
+    },
+    {
+      "epoch": 0.544368837661436,
+      "grad_norm": 4.188836507060861,
+      "learning_rate": 4.525618699770314e-06,
+      "loss": 0.5406,
+      "step": 7840
+    },
+    {
+      "epoch": 0.544438272462158,
+      "grad_norm": 2.9878048815041467,
+      "learning_rate": 4.524499296943831e-06,
+      "loss": 0.3095,
+      "step": 7841
+    },
+    {
+      "epoch": 0.5445077072628801,
+      "grad_norm": 3.1216665902706184,
+      "learning_rate": 4.523379918167632e-06,
+      "loss": 0.2296,
+      "step": 7842
+    },
+    {
+      "epoch": 0.5445771420636023,
+      "grad_norm": 3.823838070691442,
+      "learning_rate": 4.522260563498338e-06,
+      "loss": 0.4619,
+      "step": 7843
+    },
+    {
+      "epoch": 0.5446465768643244,
+      "grad_norm": 3.6633734361941426,
+      "learning_rate": 4.521141232992566e-06,
+      "loss": 0.4079,
+      "step": 7844
+    },
+    {
+      "epoch": 0.5447160116650466,
+      "grad_norm": 3.237974871365177,
+      "learning_rate": 4.520021926706926e-06,
+      "loss": 0.3856,
+      "step": 7845
+    },
+    {
+      "epoch": 0.5447854464657687,
+      "grad_norm": 4.531569033342239,
+      "learning_rate": 4.518902644698035e-06,
+      "loss": 0.7278,
+      "step": 7846
+    },
+    {
+      "epoch": 0.5448548812664907,
+      "grad_norm": 3.3474516602436326,
+      "learning_rate": 4.517783387022503e-06,
+      "loss": 0.2808,
+      "step": 7847
+    },
+    {
+      "epoch": 0.5449243160672129,
+      "grad_norm": 3.841947251692044,
+      "learning_rate": 4.516664153736943e-06,
+      "loss": 0.4473,
+      "step": 7848
+    },
+    {
+      "epoch": 0.544993750867935,
+      "grad_norm": 5.024935846692401,
+      "learning_rate": 4.515544944897962e-06,
+      "loss": 0.7742,
+      "step": 7849
+    },
+    {
+      "epoch": 0.5450631856686571,
+      "grad_norm": 2.451779319537501,
+      "learning_rate": 4.5144257605621705e-06,
+      "loss": 0.169,
+      "step": 7850
+    },
+    {
+      "epoch": 0.5451326204693793,
+      "grad_norm": 3.2673103920403133,
+      "learning_rate": 4.5133066007861735e-06,
+      "loss": 0.3717,
+      "step": 7851
+    },
+    {
+      "epoch": 0.5452020552701013,
+      "grad_norm": 2.865751525649564,
+      "learning_rate": 4.512187465626579e-06,
+      "loss": 0.2934,
+      "step": 7852
+    },
+    {
+      "epoch": 0.5452714900708235,
+      "grad_norm": 2.436368662225953,
+      "learning_rate": 4.511068355139991e-06,
+      "loss": 0.1529,
+      "step": 7853
+    },
+    {
+      "epoch": 0.5453409248715456,
+      "grad_norm": 4.112247132082978,
+      "learning_rate": 4.50994926938301e-06,
+      "loss": 0.4986,
+      "step": 7854
+    },
+    {
+      "epoch": 0.5454103596722677,
+      "grad_norm": 2.93869574583043,
+      "learning_rate": 4.5088302084122436e-06,
+      "loss": 0.3858,
+      "step": 7855
+    },
+    {
+      "epoch": 0.5454797944729899,
+      "grad_norm": 3.3623534927603385,
+      "learning_rate": 4.507711172284286e-06,
+      "loss": 0.4445,
+      "step": 7856
+    },
+    {
+      "epoch": 0.545549229273712,
+      "grad_norm": 4.335016490223838,
+      "learning_rate": 4.506592161055742e-06,
+      "loss": 0.4385,
+      "step": 7857
+    },
+    {
+      "epoch": 0.5456186640744342,
+      "grad_norm": 3.496126976370751,
+      "learning_rate": 4.505473174783207e-06,
+      "loss": 0.5206,
+      "step": 7858
+    },
+    {
+      "epoch": 0.5456880988751562,
+      "grad_norm": 3.993937390005536,
+      "learning_rate": 4.504354213523282e-06,
+      "loss": 0.2683,
+      "step": 7859
+    },
+    {
+      "epoch": 0.5457575336758783,
+      "grad_norm": 3.9507671069411776,
+      "learning_rate": 4.503235277332558e-06,
+      "loss": 0.3484,
+      "step": 7860
+    },
+    {
+      "epoch": 0.5458269684766005,
+      "grad_norm": 3.2029148736822406,
+      "learning_rate": 4.502116366267631e-06,
+      "loss": 0.3091,
+      "step": 7861
+    },
+    {
+      "epoch": 0.5458964032773226,
+      "grad_norm": 4.583294644459198,
+      "learning_rate": 4.500997480385097e-06,
+      "loss": 0.6528,
+      "step": 7862
+    },
+    {
+      "epoch": 0.5459658380780447,
+      "grad_norm": 3.1356845502109882,
+      "learning_rate": 4.499878619741545e-06,
+      "loss": 0.237,
+      "step": 7863
+    },
+    {
+      "epoch": 0.5460352728787669,
+      "grad_norm": 3.5063987998937542,
+      "learning_rate": 4.498759784393564e-06,
+      "loss": 0.241,
+      "step": 7864
+    },
+    {
+      "epoch": 0.5461047076794889,
+      "grad_norm": 3.4683401464472445,
+      "learning_rate": 4.497640974397747e-06,
+      "loss": 0.3361,
+      "step": 7865
+    },
+    {
+      "epoch": 0.5461741424802111,
+      "grad_norm": 3.557318186266412,
+      "learning_rate": 4.496522189810683e-06,
+      "loss": 0.454,
+      "step": 7866
+    },
+    {
+      "epoch": 0.5462435772809332,
+      "grad_norm": 3.8530556362666597,
+      "learning_rate": 4.495403430688956e-06,
+      "loss": 0.5746,
+      "step": 7867
+    },
+    {
+      "epoch": 0.5463130120816553,
+      "grad_norm": 5.0421371621566955,
+      "learning_rate": 4.494284697089152e-06,
+      "loss": 0.9102,
+      "step": 7868
+    },
+    {
+      "epoch": 0.5463824468823775,
+      "grad_norm": 4.0578908552130795,
+      "learning_rate": 4.493165989067857e-06,
+      "loss": 0.4863,
+      "step": 7869
+    },
+    {
+      "epoch": 0.5464518816830995,
+      "grad_norm": 3.4959425216219624,
+      "learning_rate": 4.492047306681651e-06,
+      "loss": 0.4289,
+      "step": 7870
+    },
+    {
+      "epoch": 0.5465213164838217,
+      "grad_norm": 3.7829928586282735,
+      "learning_rate": 4.490928649987118e-06,
+      "loss": 0.4026,
+      "step": 7871
+    },
+    {
+      "epoch": 0.5465907512845438,
+      "grad_norm": 4.177019234742986,
+      "learning_rate": 4.489810019040839e-06,
+      "loss": 0.6519,
+      "step": 7872
+    },
+    {
+      "epoch": 0.5466601860852659,
+      "grad_norm": 5.38908905030776,
+      "learning_rate": 4.4886914138993914e-06,
+      "loss": 0.818,
+      "step": 7873
+    },
+    {
+      "epoch": 0.5467296208859881,
+      "grad_norm": 2.7556952788455993,
+      "learning_rate": 4.4875728346193526e-06,
+      "loss": 0.4404,
+      "step": 7874
+    },
+    {
+      "epoch": 0.5467990556867102,
+      "grad_norm": 4.9663409906876,
+      "learning_rate": 4.4864542812573e-06,
+      "loss": 0.5904,
+      "step": 7875
+    },
+    {
+      "epoch": 0.5468684904874322,
+      "grad_norm": 4.14203694200325,
+      "learning_rate": 4.4853357538698085e-06,
+      "loss": 0.4948,
+      "step": 7876
+    },
+    {
+      "epoch": 0.5469379252881544,
+      "grad_norm": 5.130982168646041,
+      "learning_rate": 4.484217252513454e-06,
+      "loss": 0.9528,
+      "step": 7877
+    },
+    {
+      "epoch": 0.5470073600888765,
+      "grad_norm": 5.258988293292949,
+      "learning_rate": 4.483098777244806e-06,
+      "loss": 0.6053,
+      "step": 7878
+    },
+    {
+      "epoch": 0.5470767948895987,
+      "grad_norm": 4.486548574093626,
+      "learning_rate": 4.4819803281204365e-06,
+      "loss": 0.5033,
+      "step": 7879
+    },
+    {
+      "epoch": 0.5471462296903208,
+      "grad_norm": 4.433264916177619,
+      "learning_rate": 4.480861905196917e-06,
+      "loss": 0.5708,
+      "step": 7880
+    },
+    {
+      "epoch": 0.5472156644910429,
+      "grad_norm": 6.16111650136586,
+      "learning_rate": 4.479743508530814e-06,
+      "loss": 0.58,
+      "step": 7881
+    },
+    {
+      "epoch": 0.547285099291765,
+      "grad_norm": 3.4379582286320725,
+      "learning_rate": 4.478625138178696e-06,
+      "loss": 0.4042,
+      "step": 7882
+    },
+    {
+      "epoch": 0.5473545340924871,
+      "grad_norm": 4.298553254120449,
+      "learning_rate": 4.477506794197127e-06,
+      "loss": 0.4323,
+      "step": 7883
+    },
+    {
+      "epoch": 0.5474239688932093,
+      "grad_norm": 3.4021884041417505,
+      "learning_rate": 4.476388476642676e-06,
+      "loss": 0.3333,
+      "step": 7884
+    },
+    {
+      "epoch": 0.5474934036939314,
+      "grad_norm": 3.615758129567429,
+      "learning_rate": 4.475270185571902e-06,
+      "loss": 0.4725,
+      "step": 7885
+    },
+    {
+      "epoch": 0.5475628384946535,
+      "grad_norm": 3.706068996605706,
+      "learning_rate": 4.474151921041367e-06,
+      "loss": 0.3636,
+      "step": 7886
+    },
+    {
+      "epoch": 0.5476322732953757,
+      "grad_norm": 3.594665121381344,
+      "learning_rate": 4.4730336831076355e-06,
+      "loss": 0.4392,
+      "step": 7887
+    },
+    {
+      "epoch": 0.5477017080960978,
+      "grad_norm": 4.651491386235978,
+      "learning_rate": 4.471915471827262e-06,
+      "loss": 0.5858,
+      "step": 7888
+    },
+    {
+      "epoch": 0.5477711428968199,
+      "grad_norm": 3.869906140473688,
+      "learning_rate": 4.4707972872568065e-06,
+      "loss": 0.439,
+      "step": 7889
+    },
+    {
+      "epoch": 0.547840577697542,
+      "grad_norm": 4.14370410379573,
+      "learning_rate": 4.469679129452824e-06,
+      "loss": 0.3423,
+      "step": 7890
+    },
+    {
+      "epoch": 0.5479100124982641,
+      "grad_norm": 3.593538761836902,
+      "learning_rate": 4.4685609984718745e-06,
+      "loss": 0.285,
+      "step": 7891
+    },
+    {
+      "epoch": 0.5479794472989863,
+      "grad_norm": 3.8856291822531186,
+      "learning_rate": 4.467442894370506e-06,
+      "loss": 0.2935,
+      "step": 7892
+    },
+    {
+      "epoch": 0.5480488820997084,
+      "grad_norm": 4.006517617151857,
+      "learning_rate": 4.466324817205274e-06,
+      "loss": 0.5322,
+      "step": 7893
+    },
+    {
+      "epoch": 0.5481183169004304,
+      "grad_norm": 3.806870167661226,
+      "learning_rate": 4.465206767032729e-06,
+      "loss": 0.4075,
+      "step": 7894
+    },
+    {
+      "epoch": 0.5481877517011526,
+      "grad_norm": 4.125140631790992,
+      "learning_rate": 4.464088743909421e-06,
+      "loss": 0.4935,
+      "step": 7895
+    },
+    {
+      "epoch": 0.5482571865018747,
+      "grad_norm": 4.290221451239905,
+      "learning_rate": 4.462970747891896e-06,
+      "loss": 0.6986,
+      "step": 7896
+    },
+    {
+      "epoch": 0.5483266213025969,
+      "grad_norm": 4.3913599570855375,
+      "learning_rate": 4.461852779036703e-06,
+      "loss": 0.5693,
+      "step": 7897
+    },
+    {
+      "epoch": 0.548396056103319,
+      "grad_norm": 3.7434284310815085,
+      "learning_rate": 4.46073483740039e-06,
+      "loss": 0.5228,
+      "step": 7898
+    },
+    {
+      "epoch": 0.5484654909040411,
+      "grad_norm": 3.2362813759181686,
+      "learning_rate": 4.459616923039497e-06,
+      "loss": 0.2786,
+      "step": 7899
+    },
+    {
+      "epoch": 0.5485349257047633,
+      "grad_norm": 4.258901384452667,
+      "learning_rate": 4.458499036010568e-06,
+      "loss": 0.4919,
+      "step": 7900
+    },
+    {
+      "epoch": 0.5486043605054853,
+      "grad_norm": 3.968102879658112,
+      "learning_rate": 4.457381176370146e-06,
+      "loss": 0.3607,
+      "step": 7901
+    },
+    {
+      "epoch": 0.5486737953062075,
+      "grad_norm": 4.148065141826787,
+      "learning_rate": 4.456263344174771e-06,
+      "loss": 0.4299,
+      "step": 7902
+    },
+    {
+      "epoch": 0.5487432301069296,
+      "grad_norm": 3.4507670866569047,
+      "learning_rate": 4.455145539480978e-06,
+      "loss": 0.2755,
+      "step": 7903
+    },
+    {
+      "epoch": 0.5488126649076517,
+      "grad_norm": 3.1032518910106757,
+      "learning_rate": 4.45402776234531e-06,
+      "loss": 0.2928,
+      "step": 7904
+    },
+    {
+      "epoch": 0.5488820997083739,
+      "grad_norm": 4.1950812626960925,
+      "learning_rate": 4.452910012824299e-06,
+      "loss": 0.5158,
+      "step": 7905
+    },
+    {
+      "epoch": 0.548951534509096,
+      "grad_norm": 3.9567106677171178,
+      "learning_rate": 4.451792290974479e-06,
+      "loss": 0.5323,
+      "step": 7906
+    },
+    {
+      "epoch": 0.549020969309818,
+      "grad_norm": 4.305575843093653,
+      "learning_rate": 4.4506745968523855e-06,
+      "loss": 0.367,
+      "step": 7907
+    },
+    {
+      "epoch": 0.5490904041105402,
+      "grad_norm": 5.131052091892561,
+      "learning_rate": 4.449556930514549e-06,
+      "loss": 0.3009,
+      "step": 7908
+    },
+    {
+      "epoch": 0.5491598389112623,
+      "grad_norm": 4.096069285898798,
+      "learning_rate": 4.448439292017502e-06,
+      "loss": 0.5735,
+      "step": 7909
+    },
+    {
+      "epoch": 0.5492292737119845,
+      "grad_norm": 4.430343650529863,
+      "learning_rate": 4.4473216814177685e-06,
+      "loss": 0.4035,
+      "step": 7910
+    },
+    {
+      "epoch": 0.5492987085127066,
+      "grad_norm": 3.0240317406961656,
+      "learning_rate": 4.446204098771881e-06,
+      "loss": 0.2514,
+      "step": 7911
+    },
+    {
+      "epoch": 0.5493681433134286,
+      "grad_norm": 3.8256352346244302,
+      "learning_rate": 4.445086544136364e-06,
+      "loss": 0.5403,
+      "step": 7912
+    },
+    {
+      "epoch": 0.5494375781141508,
+      "grad_norm": 5.794244850334644,
+      "learning_rate": 4.443969017567741e-06,
+      "loss": 0.7466,
+      "step": 7913
+    },
+    {
+      "epoch": 0.5495070129148729,
+      "grad_norm": 5.474179751459487,
+      "learning_rate": 4.4428515191225354e-06,
+      "loss": 0.7446,
+      "step": 7914
+    },
+    {
+      "epoch": 0.5495764477155951,
+      "grad_norm": 3.9035724592946592,
+      "learning_rate": 4.4417340488572685e-06,
+      "loss": 0.4756,
+      "step": 7915
+    },
+    {
+      "epoch": 0.5496458825163172,
+      "grad_norm": 2.415844427581755,
+      "learning_rate": 4.440616606828464e-06,
+      "loss": 0.2236,
+      "step": 7916
+    },
+    {
+      "epoch": 0.5497153173170393,
+      "grad_norm": 3.5292265895432378,
+      "learning_rate": 4.439499193092637e-06,
+      "loss": 0.4957,
+      "step": 7917
+    },
+    {
+      "epoch": 0.5497847521177615,
+      "grad_norm": 2.855162428387412,
+      "learning_rate": 4.438381807706306e-06,
+      "loss": 0.3292,
+      "step": 7918
+    },
+    {
+      "epoch": 0.5498541869184835,
+      "grad_norm": 3.7998163312577633,
+      "learning_rate": 4.4372644507259905e-06,
+      "loss": 0.3757,
+      "step": 7919
+    },
+    {
+      "epoch": 0.5499236217192056,
+      "grad_norm": 3.482455272048503,
+      "learning_rate": 4.4361471222082e-06,
+      "loss": 0.5242,
+      "step": 7920
+    },
+    {
+      "epoch": 0.5499930565199278,
+      "grad_norm": 4.41984722393816,
+      "learning_rate": 4.43502982220945e-06,
+      "loss": 0.4838,
+      "step": 7921
+    },
+    {
+      "epoch": 0.5500624913206499,
+      "grad_norm": 5.535387869736434,
+      "learning_rate": 4.4339125507862516e-06,
+      "loss": 0.5463,
+      "step": 7922
+    },
+    {
+      "epoch": 0.5501319261213721,
+      "grad_norm": 3.817464795803458,
+      "learning_rate": 4.432795307995118e-06,
+      "loss": 0.368,
+      "step": 7923
+    },
+    {
+      "epoch": 0.5502013609220942,
+      "grad_norm": 3.676068914334841,
+      "learning_rate": 4.431678093892555e-06,
+      "loss": 0.3689,
+      "step": 7924
+    },
+    {
+      "epoch": 0.5502707957228162,
+      "grad_norm": 3.019774095957323,
+      "learning_rate": 4.43056090853507e-06,
+      "loss": 0.3959,
+      "step": 7925
+    },
+    {
+      "epoch": 0.5503402305235384,
+      "grad_norm": 3.181964786539587,
+      "learning_rate": 4.42944375197917e-06,
+      "loss": 0.3957,
+      "step": 7926
+    },
+    {
+      "epoch": 0.5504096653242605,
+      "grad_norm": 3.6707778627188437,
+      "learning_rate": 4.428326624281362e-06,
+      "loss": 0.3619,
+      "step": 7927
+    },
+    {
+      "epoch": 0.5504791001249827,
+      "grad_norm": 3.5896921107083184,
+      "learning_rate": 4.427209525498144e-06,
+      "loss": 0.3856,
+      "step": 7928
+    },
+    {
+      "epoch": 0.5505485349257048,
+      "grad_norm": 4.395115483168356,
+      "learning_rate": 4.42609245568602e-06,
+      "loss": 0.6444,
+      "step": 7929
+    },
+    {
+      "epoch": 0.5506179697264268,
+      "grad_norm": 5.509488388574817,
+      "learning_rate": 4.424975414901492e-06,
+      "loss": 0.8689,
+      "step": 7930
+    },
+    {
+      "epoch": 0.550687404527149,
+      "grad_norm": 3.149073901880616,
+      "learning_rate": 4.423858403201055e-06,
+      "loss": 0.3253,
+      "step": 7931
+    },
+    {
+      "epoch": 0.5507568393278711,
+      "grad_norm": 3.6854112896197857,
+      "learning_rate": 4.422741420641208e-06,
+      "loss": 0.3073,
+      "step": 7932
+    },
+    {
+      "epoch": 0.5508262741285932,
+      "grad_norm": 4.840887949330318,
+      "learning_rate": 4.421624467278447e-06,
+      "loss": 0.5498,
+      "step": 7933
+    },
+    {
+      "epoch": 0.5508957089293154,
+      "grad_norm": 2.5902444660353283,
+      "learning_rate": 4.420507543169268e-06,
+      "loss": 0.2945,
+      "step": 7934
+    },
+    {
+      "epoch": 0.5509651437300375,
+      "grad_norm": 3.927002084019557,
+      "learning_rate": 4.419390648370159e-06,
+      "loss": 0.6209,
+      "step": 7935
+    },
+    {
+      "epoch": 0.5510345785307597,
+      "grad_norm": 3.7969213503396917,
+      "learning_rate": 4.4182737829376135e-06,
+      "loss": 0.3833,
+      "step": 7936
+    },
+    {
+      "epoch": 0.5511040133314817,
+      "grad_norm": 5.087144198340094,
+      "learning_rate": 4.417156946928125e-06,
+      "loss": 0.6636,
+      "step": 7937
+    },
+    {
+      "epoch": 0.5511734481322038,
+      "grad_norm": 4.1873270410429555,
+      "learning_rate": 4.416040140398177e-06,
+      "loss": 0.416,
+      "step": 7938
+    },
+    {
+      "epoch": 0.551242882932926,
+      "grad_norm": 4.023183231306474,
+      "learning_rate": 4.414923363404257e-06,
+      "loss": 0.5194,
+      "step": 7939
+    },
+    {
+      "epoch": 0.5513123177336481,
+      "grad_norm": 2.7471062583481105,
+      "learning_rate": 4.413806616002852e-06,
+      "loss": 0.3136,
+      "step": 7940
+    },
+    {
+      "epoch": 0.5513817525343703,
+      "grad_norm": 4.369241212270322,
+      "learning_rate": 4.412689898250446e-06,
+      "loss": 0.3908,
+      "step": 7941
+    },
+    {
+      "epoch": 0.5514511873350924,
+      "grad_norm": 3.700229070054268,
+      "learning_rate": 4.4115732102035186e-06,
+      "loss": 0.415,
+      "step": 7942
+    },
+    {
+      "epoch": 0.5515206221358144,
+      "grad_norm": 4.458270099068796,
+      "learning_rate": 4.410456551918553e-06,
+      "loss": 0.5552,
+      "step": 7943
+    },
+    {
+      "epoch": 0.5515900569365366,
+      "grad_norm": 2.4179399159846326,
+      "learning_rate": 4.40933992345203e-06,
+      "loss": 0.2123,
+      "step": 7944
+    },
+    {
+      "epoch": 0.5516594917372587,
+      "grad_norm": 4.634787773946215,
+      "learning_rate": 4.408223324860422e-06,
+      "loss": 0.7405,
+      "step": 7945
+    },
+    {
+      "epoch": 0.5517289265379809,
+      "grad_norm": 3.185264285833028,
+      "learning_rate": 4.407106756200209e-06,
+      "loss": 0.3842,
+      "step": 7946
+    },
+    {
+      "epoch": 0.551798361338703,
+      "grad_norm": 2.4773409308235594,
+      "learning_rate": 4.405990217527866e-06,
+      "loss": 0.2298,
+      "step": 7947
+    },
+    {
+      "epoch": 0.551867796139425,
+      "grad_norm": 4.418729088339476,
+      "learning_rate": 4.404873708899867e-06,
+      "loss": 0.5424,
+      "step": 7948
+    },
+    {
+      "epoch": 0.5519372309401472,
+      "grad_norm": 3.7655194930842253,
+      "learning_rate": 4.403757230372681e-06,
+      "loss": 0.4721,
+      "step": 7949
+    },
+    {
+      "epoch": 0.5520066657408693,
+      "grad_norm": 3.5031912701625787,
+      "learning_rate": 4.40264078200278e-06,
+      "loss": 0.3636,
+      "step": 7950
+    },
+    {
+      "epoch": 0.5520761005415914,
+      "grad_norm": 3.3385631527255244,
+      "learning_rate": 4.401524363846631e-06,
+      "loss": 0.3839,
+      "step": 7951
+    },
+    {
+      "epoch": 0.5521455353423136,
+      "grad_norm": 4.13453418697,
+      "learning_rate": 4.400407975960705e-06,
+      "loss": 0.4541,
+      "step": 7952
+    },
+    {
+      "epoch": 0.5522149701430357,
+      "grad_norm": 4.875365416167943,
+      "learning_rate": 4.399291618401464e-06,
+      "loss": 0.7122,
+      "step": 7953
+    },
+    {
+      "epoch": 0.5522844049437579,
+      "grad_norm": 3.716012812244266,
+      "learning_rate": 4.3981752912253725e-06,
+      "loss": 0.3943,
+      "step": 7954
+    },
+    {
+      "epoch": 0.5523538397444799,
+      "grad_norm": 3.5602949624444227,
+      "learning_rate": 4.397058994488895e-06,
+      "loss": 0.3792,
+      "step": 7955
+    },
+    {
+      "epoch": 0.552423274545202,
+      "grad_norm": 3.9412152639985045,
+      "learning_rate": 4.3959427282484895e-06,
+      "loss": 0.5076,
+      "step": 7956
+    },
+    {
+      "epoch": 0.5524927093459242,
+      "grad_norm": 3.1688328648719755,
+      "learning_rate": 4.394826492560618e-06,
+      "loss": 0.3416,
+      "step": 7957
+    },
+    {
+      "epoch": 0.5525621441466463,
+      "grad_norm": 4.013513182224877,
+      "learning_rate": 4.393710287481737e-06,
+      "loss": 0.3626,
+      "step": 7958
+    },
+    {
+      "epoch": 0.5526315789473685,
+      "grad_norm": 3.8322144817976946,
+      "learning_rate": 4.392594113068306e-06,
+      "loss": 0.5341,
+      "step": 7959
+    },
+    {
+      "epoch": 0.5527010137480906,
+      "grad_norm": 4.080909115889743,
+      "learning_rate": 4.391477969376777e-06,
+      "loss": 0.4921,
+      "step": 7960
+    },
+    {
+      "epoch": 0.5527704485488126,
+      "grad_norm": 4.154841665087655,
+      "learning_rate": 4.390361856463601e-06,
+      "loss": 0.347,
+      "step": 7961
+    },
+    {
+      "epoch": 0.5528398833495348,
+      "grad_norm": 3.7389184847166272,
+      "learning_rate": 4.389245774385235e-06,
+      "loss": 0.3459,
+      "step": 7962
+    },
+    {
+      "epoch": 0.5529093181502569,
+      "grad_norm": 4.498990352242344,
+      "learning_rate": 4.388129723198126e-06,
+      "loss": 0.578,
+      "step": 7963
+    },
+    {
+      "epoch": 0.552978752950979,
+      "grad_norm": 5.6533854135584365,
+      "learning_rate": 4.387013702958722e-06,
+      "loss": 0.3347,
+      "step": 7964
+    },
+    {
+      "epoch": 0.5530481877517012,
+      "grad_norm": 4.461771946469103,
+      "learning_rate": 4.385897713723472e-06,
+      "loss": 0.4243,
+      "step": 7965
+    },
+    {
+      "epoch": 0.5531176225524232,
+      "grad_norm": 3.8490168128715125,
+      "learning_rate": 4.384781755548823e-06,
+      "loss": 0.2819,
+      "step": 7966
+    },
+    {
+      "epoch": 0.5531870573531454,
+      "grad_norm": 4.128367880781366,
+      "learning_rate": 4.383665828491214e-06,
+      "loss": 0.3998,
+      "step": 7967
+    },
+    {
+      "epoch": 0.5532564921538675,
+      "grad_norm": 2.525011343418954,
+      "learning_rate": 4.382549932607089e-06,
+      "loss": 0.1469,
+      "step": 7968
+    },
+    {
+      "epoch": 0.5533259269545896,
+      "grad_norm": 2.8130810609876242,
+      "learning_rate": 4.381434067952893e-06,
+      "loss": 0.2523,
+      "step": 7969
+    },
+    {
+      "epoch": 0.5533953617553118,
+      "grad_norm": 4.468741439854196,
+      "learning_rate": 4.38031823458506e-06,
+      "loss": 0.7451,
+      "step": 7970
+    },
+    {
+      "epoch": 0.5534647965560339,
+      "grad_norm": 2.87534946199836,
+      "learning_rate": 4.379202432560029e-06,
+      "loss": 0.2741,
+      "step": 7971
+    },
+    {
+      "epoch": 0.5535342313567561,
+      "grad_norm": 3.2750310603315587,
+      "learning_rate": 4.378086661934237e-06,
+      "loss": 0.3077,
+      "step": 7972
+    },
+    {
+      "epoch": 0.5536036661574781,
+      "grad_norm": 4.270109746229505,
+      "learning_rate": 4.376970922764119e-06,
+      "loss": 0.4675,
+      "step": 7973
+    },
+    {
+      "epoch": 0.5536731009582002,
+      "grad_norm": 3.4226562554158124,
+      "learning_rate": 4.375855215106105e-06,
+      "loss": 0.3705,
+      "step": 7974
+    },
+    {
+      "epoch": 0.5537425357589224,
+      "grad_norm": 4.00250215696577,
+      "learning_rate": 4.374739539016629e-06,
+      "loss": 0.502,
+      "step": 7975
+    },
+    {
+      "epoch": 0.5538119705596445,
+      "grad_norm": 4.991099343483806,
+      "learning_rate": 4.3736238945521185e-06,
+      "loss": 0.4291,
+      "step": 7976
+    },
+    {
+      "epoch": 0.5538814053603666,
+      "grad_norm": 2.8092616961154255,
+      "learning_rate": 4.372508281769004e-06,
+      "loss": 0.2455,
+      "step": 7977
+    },
+    {
+      "epoch": 0.5539508401610888,
+      "grad_norm": 3.1087895334313167,
+      "learning_rate": 4.3713927007237104e-06,
+      "loss": 0.1976,
+      "step": 7978
+    },
+    {
+      "epoch": 0.5540202749618108,
+      "grad_norm": 4.780987592212305,
+      "learning_rate": 4.3702771514726615e-06,
+      "loss": 0.6574,
+      "step": 7979
+    },
+    {
+      "epoch": 0.554089709762533,
+      "grad_norm": 4.441327826193259,
+      "learning_rate": 4.369161634072284e-06,
+      "loss": 0.757,
+      "step": 7980
+    },
+    {
+      "epoch": 0.5541591445632551,
+      "grad_norm": 3.3393032626235737,
+      "learning_rate": 4.368046148578995e-06,
+      "loss": 0.364,
+      "step": 7981
+    },
+    {
+      "epoch": 0.5542285793639772,
+      "grad_norm": 3.546433096390069,
+      "learning_rate": 4.366930695049217e-06,
+      "loss": 0.3617,
+      "step": 7982
+    },
+    {
+      "epoch": 0.5542980141646994,
+      "grad_norm": 3.836359145991592,
+      "learning_rate": 4.365815273539368e-06,
+      "loss": 0.511,
+      "step": 7983
+    },
+    {
+      "epoch": 0.5543674489654214,
+      "grad_norm": 3.3392862422919194,
+      "learning_rate": 4.364699884105866e-06,
+      "loss": 0.3546,
+      "step": 7984
+    },
+    {
+      "epoch": 0.5544368837661436,
+      "grad_norm": 2.745442412705354,
+      "learning_rate": 4.363584526805124e-06,
+      "loss": 0.3373,
+      "step": 7985
+    },
+    {
+      "epoch": 0.5545063185668657,
+      "grad_norm": 3.325892497304828,
+      "learning_rate": 4.362469201693556e-06,
+      "loss": 0.2191,
+      "step": 7986
+    },
+    {
+      "epoch": 0.5545757533675878,
+      "grad_norm": 2.6735411206784683,
+      "learning_rate": 4.361353908827577e-06,
+      "loss": 0.2911,
+      "step": 7987
+    },
+    {
+      "epoch": 0.55464518816831,
+      "grad_norm": 3.977619051697231,
+      "learning_rate": 4.360238648263591e-06,
+      "loss": 0.4674,
+      "step": 7988
+    },
+    {
+      "epoch": 0.5547146229690321,
+      "grad_norm": 3.7267064974992037,
+      "learning_rate": 4.359123420058013e-06,
+      "loss": 0.4593,
+      "step": 7989
+    },
+    {
+      "epoch": 0.5547840577697541,
+      "grad_norm": 4.999904390531836,
+      "learning_rate": 4.358008224267245e-06,
+      "loss": 0.5662,
+      "step": 7990
+    },
+    {
+      "epoch": 0.5548534925704763,
+      "grad_norm": 4.479396806272263,
+      "learning_rate": 4.3568930609476964e-06,
+      "loss": 0.5029,
+      "step": 7991
+    },
+    {
+      "epoch": 0.5549229273711984,
+      "grad_norm": 2.6431742753672096,
+      "learning_rate": 4.355777930155769e-06,
+      "loss": 0.1754,
+      "step": 7992
+    },
+    {
+      "epoch": 0.5549923621719206,
+      "grad_norm": 3.6006094853421353,
+      "learning_rate": 4.354662831947863e-06,
+      "loss": 0.3914,
+      "step": 7993
+    },
+    {
+      "epoch": 0.5550617969726427,
+      "grad_norm": 4.026048817642806,
+      "learning_rate": 4.353547766380384e-06,
+      "loss": 0.3545,
+      "step": 7994
+    },
+    {
+      "epoch": 0.5551312317733648,
+      "grad_norm": 3.9100487941225617,
+      "learning_rate": 4.352432733509725e-06,
+      "loss": 0.5687,
+      "step": 7995
+    },
+    {
+      "epoch": 0.555200666574087,
+      "grad_norm": 3.872365307883858,
+      "learning_rate": 4.3513177333922855e-06,
+      "loss": 0.4288,
+      "step": 7996
+    },
+    {
+      "epoch": 0.555270101374809,
+      "grad_norm": 4.934567948767992,
+      "learning_rate": 4.3502027660844606e-06,
+      "loss": 0.6546,
+      "step": 7997
+    },
+    {
+      "epoch": 0.5553395361755312,
+      "grad_norm": 4.308362687182634,
+      "learning_rate": 4.3490878316426464e-06,
+      "loss": 0.5472,
+      "step": 7998
+    },
+    {
+      "epoch": 0.5554089709762533,
+      "grad_norm": 3.933495045189264,
+      "learning_rate": 4.347972930123232e-06,
+      "loss": 0.3982,
+      "step": 7999
+    },
+    {
+      "epoch": 0.5554784057769754,
+      "grad_norm": 4.539997064667931,
+      "learning_rate": 4.346858061582608e-06,
+      "loss": 0.4613,
+      "step": 8000
+    },
+    {
+      "epoch": 0.5555478405776976,
+      "grad_norm": 3.159671633508985,
+      "learning_rate": 4.345743226077166e-06,
+      "loss": 0.355,
+      "step": 8001
+    },
+    {
+      "epoch": 0.5556172753784197,
+      "grad_norm": 3.9236063272338515,
+      "learning_rate": 4.344628423663292e-06,
+      "loss": 0.4882,
+      "step": 8002
+    },
+    {
+      "epoch": 0.5556867101791418,
+      "grad_norm": 3.9446194270922357,
+      "learning_rate": 4.343513654397368e-06,
+      "loss": 0.2749,
+      "step": 8003
+    },
+    {
+      "epoch": 0.5557561449798639,
+      "grad_norm": 3.63565619146985,
+      "learning_rate": 4.3423989183357815e-06,
+      "loss": 0.3144,
+      "step": 8004
+    },
+    {
+      "epoch": 0.555825579780586,
+      "grad_norm": 3.7745700007046206,
+      "learning_rate": 4.341284215534915e-06,
+      "loss": 0.5633,
+      "step": 8005
+    },
+    {
+      "epoch": 0.5558950145813082,
+      "grad_norm": 3.8122967868522304,
+      "learning_rate": 4.340169546051146e-06,
+      "loss": 0.3741,
+      "step": 8006
+    },
+    {
+      "epoch": 0.5559644493820303,
+      "grad_norm": 3.6017088553048207,
+      "learning_rate": 4.339054909940854e-06,
+      "loss": 0.4157,
+      "step": 8007
+    },
+    {
+      "epoch": 0.5560338841827523,
+      "grad_norm": 4.171330297258857,
+      "learning_rate": 4.337940307260418e-06,
+      "loss": 0.6017,
+      "step": 8008
+    },
+    {
+      "epoch": 0.5561033189834745,
+      "grad_norm": 2.546143847923642,
+      "learning_rate": 4.336825738066213e-06,
+      "loss": 0.2734,
+      "step": 8009
+    },
+    {
+      "epoch": 0.5561727537841966,
+      "grad_norm": 5.051997531763219,
+      "learning_rate": 4.3357112024146095e-06,
+      "loss": 0.7547,
+      "step": 8010
+    },
+    {
+      "epoch": 0.5562421885849188,
+      "grad_norm": 3.1660691836084416,
+      "learning_rate": 4.334596700361984e-06,
+      "loss": 0.231,
+      "step": 8011
+    },
+    {
+      "epoch": 0.5563116233856409,
+      "grad_norm": 3.570132483906301,
+      "learning_rate": 4.333482231964704e-06,
+      "loss": 0.4546,
+      "step": 8012
+    },
+    {
+      "epoch": 0.556381058186363,
+      "grad_norm": 4.673398747368056,
+      "learning_rate": 4.3323677972791365e-06,
+      "loss": 0.5107,
+      "step": 8013
+    },
+    {
+      "epoch": 0.5564504929870852,
+      "grad_norm": 3.209563382113947,
+      "learning_rate": 4.331253396361652e-06,
+      "loss": 0.257,
+      "step": 8014
+    },
+    {
+      "epoch": 0.5565199277878072,
+      "grad_norm": 2.551323817402932,
+      "learning_rate": 4.330139029268613e-06,
+      "loss": 0.2315,
+      "step": 8015
+    },
+    {
+      "epoch": 0.5565893625885294,
+      "grad_norm": 3.9244732258550434,
+      "learning_rate": 4.3290246960563855e-06,
+      "loss": 0.359,
+      "step": 8016
+    },
+    {
+      "epoch": 0.5566587973892515,
+      "grad_norm": 3.8565560404401933,
+      "learning_rate": 4.3279103967813275e-06,
+      "loss": 0.3186,
+      "step": 8017
+    },
+    {
+      "epoch": 0.5567282321899736,
+      "grad_norm": 3.8660123835589433,
+      "learning_rate": 4.326796131499802e-06,
+      "loss": 0.5323,
+      "step": 8018
+    },
+    {
+      "epoch": 0.5567976669906958,
+      "grad_norm": 3.821889060541428,
+      "learning_rate": 4.325681900268168e-06,
+      "loss": 0.4549,
+      "step": 8019
+    },
+    {
+      "epoch": 0.5568671017914179,
+      "grad_norm": 2.651366966879796,
+      "learning_rate": 4.324567703142779e-06,
+      "loss": 0.1881,
+      "step": 8020
+    },
+    {
+      "epoch": 0.5569365365921399,
+      "grad_norm": 4.25735707991398,
+      "learning_rate": 4.323453540179991e-06,
+      "loss": 0.4554,
+      "step": 8021
+    },
+    {
+      "epoch": 0.5570059713928621,
+      "grad_norm": 4.6344108468425755,
+      "learning_rate": 4.322339411436157e-06,
+      "loss": 0.5277,
+      "step": 8022
+    },
+    {
+      "epoch": 0.5570754061935842,
+      "grad_norm": 3.338843726031172,
+      "learning_rate": 4.32122531696763e-06,
+      "loss": 0.459,
+      "step": 8023
+    },
+    {
+      "epoch": 0.5571448409943064,
+      "grad_norm": 4.0649598247326475,
+      "learning_rate": 4.320111256830758e-06,
+      "loss": 0.337,
+      "step": 8024
+    },
+    {
+      "epoch": 0.5572142757950285,
+      "grad_norm": 4.8411121173564435,
+      "learning_rate": 4.318997231081888e-06,
+      "loss": 0.5425,
+      "step": 8025
+    },
+    {
+      "epoch": 0.5572837105957505,
+      "grad_norm": 3.7479473571589232,
+      "learning_rate": 4.3178832397773675e-06,
+      "loss": 0.409,
+      "step": 8026
+    },
+    {
+      "epoch": 0.5573531453964727,
+      "grad_norm": 3.5850288105319428,
+      "learning_rate": 4.316769282973543e-06,
+      "loss": 0.419,
+      "step": 8027
+    },
+    {
+      "epoch": 0.5574225801971948,
+      "grad_norm": 3.6079457080658313,
+      "learning_rate": 4.315655360726755e-06,
+      "loss": 0.4074,
+      "step": 8028
+    },
+    {
+      "epoch": 0.557492014997917,
+      "grad_norm": 3.6309558812798164,
+      "learning_rate": 4.314541473093342e-06,
+      "loss": 0.3185,
+      "step": 8029
+    },
+    {
+      "epoch": 0.5575614497986391,
+      "grad_norm": 3.8587085769601885,
+      "learning_rate": 4.3134276201296484e-06,
+      "loss": 0.6381,
+      "step": 8030
+    },
+    {
+      "epoch": 0.5576308845993612,
+      "grad_norm": 3.6111334703464117,
+      "learning_rate": 4.312313801892008e-06,
+      "loss": 0.4805,
+      "step": 8031
+    },
+    {
+      "epoch": 0.5577003194000834,
+      "grad_norm": 3.791047079299569,
+      "learning_rate": 4.311200018436755e-06,
+      "loss": 0.3241,
+      "step": 8032
+    },
+    {
+      "epoch": 0.5577697542008054,
+      "grad_norm": 3.4350727058912027,
+      "learning_rate": 4.310086269820228e-06,
+      "loss": 0.4461,
+      "step": 8033
+    },
+    {
+      "epoch": 0.5578391890015275,
+      "grad_norm": 4.000253154079633,
+      "learning_rate": 4.308972556098758e-06,
+      "loss": 0.4685,
+      "step": 8034
+    },
+    {
+      "epoch": 0.5579086238022497,
+      "grad_norm": 3.033555607443416,
+      "learning_rate": 4.307858877328671e-06,
+      "loss": 0.4714,
+      "step": 8035
+    },
+    {
+      "epoch": 0.5579780586029718,
+      "grad_norm": 2.8589462096367972,
+      "learning_rate": 4.3067452335663e-06,
+      "loss": 0.2226,
+      "step": 8036
+    },
+    {
+      "epoch": 0.558047493403694,
+      "grad_norm": 3.785152531452116,
+      "learning_rate": 4.305631624867971e-06,
+      "loss": 0.3644,
+      "step": 8037
+    },
+    {
+      "epoch": 0.558116928204416,
+      "grad_norm": 4.041312525851003,
+      "learning_rate": 4.304518051290009e-06,
+      "loss": 0.579,
+      "step": 8038
+    },
+    {
+      "epoch": 0.5581863630051381,
+      "grad_norm": 3.456910964301835,
+      "learning_rate": 4.303404512888736e-06,
+      "loss": 0.406,
+      "step": 8039
+    },
+    {
+      "epoch": 0.5582557978058603,
+      "grad_norm": 4.003596298550473,
+      "learning_rate": 4.3022910097204746e-06,
+      "loss": 0.5526,
+      "step": 8040
+    },
+    {
+      "epoch": 0.5583252326065824,
+      "grad_norm": 3.440118120043311,
+      "learning_rate": 4.301177541841546e-06,
+      "loss": 0.483,
+      "step": 8041
+    },
+    {
+      "epoch": 0.5583946674073046,
+      "grad_norm": 2.6851504514986217,
+      "learning_rate": 4.300064109308265e-06,
+      "loss": 0.3009,
+      "step": 8042
+    },
+    {
+      "epoch": 0.5584641022080267,
+      "grad_norm": 4.312399885079339,
+      "learning_rate": 4.298950712176949e-06,
+      "loss": 0.4678,
+      "step": 8043
+    },
+    {
+      "epoch": 0.5585335370087487,
+      "grad_norm": 4.759977007532914,
+      "learning_rate": 4.297837350503915e-06,
+      "loss": 0.4394,
+      "step": 8044
+    },
+    {
+      "epoch": 0.5586029718094709,
+      "grad_norm": 4.037399607168306,
+      "learning_rate": 4.296724024345471e-06,
+      "loss": 0.5861,
+      "step": 8045
+    },
+    {
+      "epoch": 0.558672406610193,
+      "grad_norm": 4.221280452644258,
+      "learning_rate": 4.295610733757929e-06,
+      "loss": 0.684,
+      "step": 8046
+    },
+    {
+      "epoch": 0.5587418414109151,
+      "grad_norm": 4.75674460599765,
+      "learning_rate": 4.2944974787976015e-06,
+      "loss": 0.4667,
+      "step": 8047
+    },
+    {
+      "epoch": 0.5588112762116373,
+      "grad_norm": 3.8070836015104073,
+      "learning_rate": 4.293384259520793e-06,
+      "loss": 0.4039,
+      "step": 8048
+    },
+    {
+      "epoch": 0.5588807110123594,
+      "grad_norm": 4.655021697529806,
+      "learning_rate": 4.292271075983807e-06,
+      "loss": 0.5336,
+      "step": 8049
+    },
+    {
+      "epoch": 0.5589501458130816,
+      "grad_norm": 3.822841050726024,
+      "learning_rate": 4.2911579282429514e-06,
+      "loss": 0.4689,
+      "step": 8050
+    },
+    {
+      "epoch": 0.5590195806138036,
+      "grad_norm": 5.005227463254897,
+      "learning_rate": 4.290044816354526e-06,
+      "loss": 0.7394,
+      "step": 8051
+    },
+    {
+      "epoch": 0.5590890154145257,
+      "grad_norm": 3.5699912229478286,
+      "learning_rate": 4.288931740374828e-06,
+      "loss": 0.2747,
+      "step": 8052
+    },
+    {
+      "epoch": 0.5591584502152479,
+      "grad_norm": 4.759385720995197,
+      "learning_rate": 4.28781870036016e-06,
+      "loss": 0.4766,
+      "step": 8053
+    },
+    {
+      "epoch": 0.55922788501597,
+      "grad_norm": 3.651996214781099,
+      "learning_rate": 4.286705696366814e-06,
+      "loss": 0.3897,
+      "step": 8054
+    },
+    {
+      "epoch": 0.5592973198166922,
+      "grad_norm": 4.0622158308199126,
+      "learning_rate": 4.28559272845109e-06,
+      "loss": 0.6018,
+      "step": 8055
+    },
+    {
+      "epoch": 0.5593667546174143,
+      "grad_norm": 5.5779982527649725,
+      "learning_rate": 4.284479796669276e-06,
+      "loss": 0.6662,
+      "step": 8056
+    },
+    {
+      "epoch": 0.5594361894181363,
+      "grad_norm": 2.9891069608549317,
+      "learning_rate": 4.283366901077662e-06,
+      "loss": 0.1936,
+      "step": 8057
+    },
+    {
+      "epoch": 0.5595056242188585,
+      "grad_norm": 3.467276993891775,
+      "learning_rate": 4.28225404173254e-06,
+      "loss": 0.3669,
+      "step": 8058
+    },
+    {
+      "epoch": 0.5595750590195806,
+      "grad_norm": 2.9775066106144528,
+      "learning_rate": 4.281141218690199e-06,
+      "loss": 0.3,
+      "step": 8059
+    },
+    {
+      "epoch": 0.5596444938203028,
+      "grad_norm": 4.222788936038443,
+      "learning_rate": 4.28002843200692e-06,
+      "loss": 0.4675,
+      "step": 8060
+    },
+    {
+      "epoch": 0.5597139286210249,
+      "grad_norm": 2.747053951464099,
+      "learning_rate": 4.278915681738987e-06,
+      "loss": 0.2577,
+      "step": 8061
+    },
+    {
+      "epoch": 0.559783363421747,
+      "grad_norm": 3.7261632998639644,
+      "learning_rate": 4.277802967942686e-06,
+      "loss": 0.2379,
+      "step": 8062
+    },
+    {
+      "epoch": 0.5598527982224691,
+      "grad_norm": 3.731873232279424,
+      "learning_rate": 4.276690290674292e-06,
+      "loss": 0.386,
+      "step": 8063
+    },
+    {
+      "epoch": 0.5599222330231912,
+      "grad_norm": 4.5001940344366,
+      "learning_rate": 4.275577649990083e-06,
+      "loss": 0.6856,
+      "step": 8064
+    },
+    {
+      "epoch": 0.5599916678239133,
+      "grad_norm": 3.0482818554032476,
+      "learning_rate": 4.274465045946338e-06,
+      "loss": 0.2661,
+      "step": 8065
+    },
+    {
+      "epoch": 0.5600611026246355,
+      "grad_norm": 3.784758115346529,
+      "learning_rate": 4.273352478599332e-06,
+      "loss": 0.3525,
+      "step": 8066
+    },
+    {
+      "epoch": 0.5601305374253576,
+      "grad_norm": 4.759387961089601,
+      "learning_rate": 4.272239948005332e-06,
+      "loss": 0.3067,
+      "step": 8067
+    },
+    {
+      "epoch": 0.5601999722260798,
+      "grad_norm": 4.621699454443586,
+      "learning_rate": 4.271127454220613e-06,
+      "loss": 0.5957,
+      "step": 8068
+    },
+    {
+      "epoch": 0.5602694070268018,
+      "grad_norm": 3.9103628684669545,
+      "learning_rate": 4.2700149973014445e-06,
+      "loss": 0.4948,
+      "step": 8069
+    },
+    {
+      "epoch": 0.5603388418275239,
+      "grad_norm": 5.138704526475872,
+      "learning_rate": 4.26890257730409e-06,
+      "loss": 0.6486,
+      "step": 8070
+    },
+    {
+      "epoch": 0.5604082766282461,
+      "grad_norm": 2.957036252937091,
+      "learning_rate": 4.267790194284815e-06,
+      "loss": 0.3464,
+      "step": 8071
+    },
+    {
+      "epoch": 0.5604777114289682,
+      "grad_norm": 5.764796852455868,
+      "learning_rate": 4.266677848299884e-06,
+      "loss": 0.7154,
+      "step": 8072
+    },
+    {
+      "epoch": 0.5605471462296904,
+      "grad_norm": 4.573464369438991,
+      "learning_rate": 4.265565539405559e-06,
+      "loss": 0.4377,
+      "step": 8073
+    },
+    {
+      "epoch": 0.5606165810304125,
+      "grad_norm": 4.342873209204397,
+      "learning_rate": 4.264453267658096e-06,
+      "loss": 0.8664,
+      "step": 8074
+    },
+    {
+      "epoch": 0.5606860158311345,
+      "grad_norm": 3.689379978303678,
+      "learning_rate": 4.263341033113754e-06,
+      "loss": 0.5167,
+      "step": 8075
+    },
+    {
+      "epoch": 0.5607554506318567,
+      "grad_norm": 3.728726126436983,
+      "learning_rate": 4.262228835828792e-06,
+      "loss": 0.4355,
+      "step": 8076
+    },
+    {
+      "epoch": 0.5608248854325788,
+      "grad_norm": 3.5475220770009157,
+      "learning_rate": 4.261116675859458e-06,
+      "loss": 0.3629,
+      "step": 8077
+    },
+    {
+      "epoch": 0.5608943202333009,
+      "grad_norm": 3.8060252779515347,
+      "learning_rate": 4.260004553262006e-06,
+      "loss": 0.5349,
+      "step": 8078
+    },
+    {
+      "epoch": 0.5609637550340231,
+      "grad_norm": 2.539140125577981,
+      "learning_rate": 4.258892468092688e-06,
+      "loss": 0.2316,
+      "step": 8079
+    },
+    {
+      "epoch": 0.5610331898347451,
+      "grad_norm": 6.392793929203389,
+      "learning_rate": 4.257780420407751e-06,
+      "loss": 1.0245,
+      "step": 8080
+    },
+    {
+      "epoch": 0.5611026246354673,
+      "grad_norm": 4.560841650768598,
+      "learning_rate": 4.256668410263438e-06,
+      "loss": 0.5007,
+      "step": 8081
+    },
+    {
+      "epoch": 0.5611720594361894,
+      "grad_norm": 4.846623830495395,
+      "learning_rate": 4.255556437715998e-06,
+      "loss": 0.4971,
+      "step": 8082
+    },
+    {
+      "epoch": 0.5612414942369115,
+      "grad_norm": 9.84865015745314,
+      "learning_rate": 4.254444502821669e-06,
+      "loss": 0.3536,
+      "step": 8083
+    },
+    {
+      "epoch": 0.5613109290376337,
+      "grad_norm": 4.034626626972785,
+      "learning_rate": 4.253332605636695e-06,
+      "loss": 0.5029,
+      "step": 8084
+    },
+    {
+      "epoch": 0.5613803638383558,
+      "grad_norm": 8.492252176999823,
+      "learning_rate": 4.2522207462173124e-06,
+      "loss": 0.5979,
+      "step": 8085
+    },
+    {
+      "epoch": 0.561449798639078,
+      "grad_norm": 4.39176848080515,
+      "learning_rate": 4.2511089246197575e-06,
+      "loss": 0.6506,
+      "step": 8086
+    },
+    {
+      "epoch": 0.5615192334398,
+      "grad_norm": 4.582675213777285,
+      "learning_rate": 4.2499971409002675e-06,
+      "loss": 0.629,
+      "step": 8087
+    },
+    {
+      "epoch": 0.5615886682405221,
+      "grad_norm": 3.734472841424299,
+      "learning_rate": 4.248885395115072e-06,
+      "loss": 0.3284,
+      "step": 8088
+    },
+    {
+      "epoch": 0.5616581030412443,
+      "grad_norm": 3.6625870446563185,
+      "learning_rate": 4.247773687320403e-06,
+      "loss": 0.344,
+      "step": 8089
+    },
+    {
+      "epoch": 0.5617275378419664,
+      "grad_norm": 3.2084158119287145,
+      "learning_rate": 4.246662017572489e-06,
+      "loss": 0.2704,
+      "step": 8090
+    },
+    {
+      "epoch": 0.5617969726426885,
+      "grad_norm": 3.455062142003421,
+      "learning_rate": 4.245550385927561e-06,
+      "loss": 0.3266,
+      "step": 8091
+    },
+    {
+      "epoch": 0.5618664074434107,
+      "grad_norm": 4.928357066696252,
+      "learning_rate": 4.244438792441838e-06,
+      "loss": 0.5406,
+      "step": 8092
+    },
+    {
+      "epoch": 0.5619358422441327,
+      "grad_norm": 3.5877828963060976,
+      "learning_rate": 4.243327237171546e-06,
+      "loss": 0.3744,
+      "step": 8093
+    },
+    {
+      "epoch": 0.5620052770448549,
+      "grad_norm": 2.5498584729273333,
+      "learning_rate": 4.242215720172908e-06,
+      "loss": 0.2633,
+      "step": 8094
+    },
+    {
+      "epoch": 0.562074711845577,
+      "grad_norm": 4.4099002854716165,
+      "learning_rate": 4.241104241502141e-06,
+      "loss": 0.4558,
+      "step": 8095
+    },
+    {
+      "epoch": 0.5621441466462991,
+      "grad_norm": 3.656778387369396,
+      "learning_rate": 4.23999280121546e-06,
+      "loss": 0.3621,
+      "step": 8096
+    },
+    {
+      "epoch": 0.5622135814470213,
+      "grad_norm": 4.191310171800451,
+      "learning_rate": 4.238881399369085e-06,
+      "loss": 0.2192,
+      "step": 8097
+    },
+    {
+      "epoch": 0.5622830162477434,
+      "grad_norm": 2.9876215132200725,
+      "learning_rate": 4.237770036019229e-06,
+      "loss": 0.3389,
+      "step": 8098
+    },
+    {
+      "epoch": 0.5623524510484655,
+      "grad_norm": 4.058141304487534,
+      "learning_rate": 4.2366587112221e-06,
+      "loss": 0.5927,
+      "step": 8099
+    },
+    {
+      "epoch": 0.5624218858491876,
+      "grad_norm": 3.151403944126138,
+      "learning_rate": 4.235547425033909e-06,
+      "loss": 0.4508,
+      "step": 8100
+    },
+    {
+      "epoch": 0.5624913206499097,
+      "grad_norm": 3.3437092798908945,
+      "learning_rate": 4.2344361775108675e-06,
+      "loss": 0.3172,
+      "step": 8101
+    },
+    {
+      "epoch": 0.5625607554506319,
+      "grad_norm": 3.701951222376618,
+      "learning_rate": 4.2333249687091755e-06,
+      "loss": 0.4957,
+      "step": 8102
+    },
+    {
+      "epoch": 0.562630190251354,
+      "grad_norm": 3.231759933245245,
+      "learning_rate": 4.232213798685039e-06,
+      "loss": 0.3692,
+      "step": 8103
+    },
+    {
+      "epoch": 0.562699625052076,
+      "grad_norm": 3.3931137239978866,
+      "learning_rate": 4.23110266749466e-06,
+      "loss": 0.3564,
+      "step": 8104
+    },
+    {
+      "epoch": 0.5627690598527982,
+      "grad_norm": 3.2259419099428177,
+      "learning_rate": 4.229991575194239e-06,
+      "loss": 0.2396,
+      "step": 8105
+    },
+    {
+      "epoch": 0.5628384946535203,
+      "grad_norm": 2.647385183411293,
+      "learning_rate": 4.228880521839971e-06,
+      "loss": 0.4323,
+      "step": 8106
+    },
+    {
+      "epoch": 0.5629079294542425,
+      "grad_norm": 4.167743942983958,
+      "learning_rate": 4.227769507488053e-06,
+      "loss": 0.5169,
+      "step": 8107
+    },
+    {
+      "epoch": 0.5629773642549646,
+      "grad_norm": 4.180563368683919,
+      "learning_rate": 4.226658532194682e-06,
+      "loss": 0.4443,
+      "step": 8108
+    },
+    {
+      "epoch": 0.5630467990556867,
+      "grad_norm": 3.690142800221043,
+      "learning_rate": 4.225547596016047e-06,
+      "loss": 0.5928,
+      "step": 8109
+    },
+    {
+      "epoch": 0.5631162338564089,
+      "grad_norm": 4.903539476942145,
+      "learning_rate": 4.2244366990083365e-06,
+      "loss": 0.4863,
+      "step": 8110
+    },
+    {
+      "epoch": 0.5631856686571309,
+      "grad_norm": 4.76017789991848,
+      "learning_rate": 4.223325841227741e-06,
+      "loss": 0.5019,
+      "step": 8111
+    },
+    {
+      "epoch": 0.5632551034578531,
+      "grad_norm": 3.0784452277795746,
+      "learning_rate": 4.222215022730447e-06,
+      "loss": 0.2069,
+      "step": 8112
+    },
+    {
+      "epoch": 0.5633245382585752,
+      "grad_norm": 4.429703459975234,
+      "learning_rate": 4.221104243572635e-06,
+      "loss": 0.4352,
+      "step": 8113
+    },
+    {
+      "epoch": 0.5633939730592973,
+      "grad_norm": 3.7123668615264998,
+      "learning_rate": 4.21999350381049e-06,
+      "loss": 0.5006,
+      "step": 8114
+    },
+    {
+      "epoch": 0.5634634078600195,
+      "grad_norm": 3.5757844972252277,
+      "learning_rate": 4.21888280350019e-06,
+      "loss": 0.4769,
+      "step": 8115
+    },
+    {
+      "epoch": 0.5635328426607416,
+      "grad_norm": 2.6086103075963103,
+      "learning_rate": 4.217772142697916e-06,
+      "loss": 0.1998,
+      "step": 8116
+    },
+    {
+      "epoch": 0.5636022774614637,
+      "grad_norm": 3.9128340499299585,
+      "learning_rate": 4.216661521459839e-06,
+      "loss": 0.3663,
+      "step": 8117
+    },
+    {
+      "epoch": 0.5636717122621858,
+      "grad_norm": 3.3748034214434766,
+      "learning_rate": 4.215550939842138e-06,
+      "loss": 0.3027,
+      "step": 8118
+    },
+    {
+      "epoch": 0.5637411470629079,
+      "grad_norm": 3.025397680232139,
+      "learning_rate": 4.214440397900983e-06,
+      "loss": 0.318,
+      "step": 8119
+    },
+    {
+      "epoch": 0.5638105818636301,
+      "grad_norm": 3.4050101831654214,
+      "learning_rate": 4.213329895692542e-06,
+      "loss": 0.3515,
+      "step": 8120
+    },
+    {
+      "epoch": 0.5638800166643522,
+      "grad_norm": 4.36811476668919,
+      "learning_rate": 4.212219433272986e-06,
+      "loss": 0.5281,
+      "step": 8121
+    },
+    {
+      "epoch": 0.5639494514650742,
+      "grad_norm": 2.7638261172320795,
+      "learning_rate": 4.211109010698478e-06,
+      "loss": 0.2208,
+      "step": 8122
+    },
+    {
+      "epoch": 0.5640188862657964,
+      "grad_norm": 3.769782909893306,
+      "learning_rate": 4.209998628025186e-06,
+      "loss": 0.3619,
+      "step": 8123
+    },
+    {
+      "epoch": 0.5640883210665185,
+      "grad_norm": 4.071857298714483,
+      "learning_rate": 4.208888285309268e-06,
+      "loss": 0.5466,
+      "step": 8124
+    },
+    {
+      "epoch": 0.5641577558672407,
+      "grad_norm": 3.182954396456208,
+      "learning_rate": 4.207777982606885e-06,
+      "loss": 0.295,
+      "step": 8125
+    },
+    {
+      "epoch": 0.5642271906679628,
+      "grad_norm": 4.602885597909247,
+      "learning_rate": 4.2066677199741965e-06,
+      "loss": 0.577,
+      "step": 8126
+    },
+    {
+      "epoch": 0.5642966254686849,
+      "grad_norm": 4.068994059212368,
+      "learning_rate": 4.205557497467356e-06,
+      "loss": 0.4953,
+      "step": 8127
+    },
+    {
+      "epoch": 0.564366060269407,
+      "grad_norm": 5.018801198612864,
+      "learning_rate": 4.204447315142519e-06,
+      "loss": 0.7809,
+      "step": 8128
+    },
+    {
+      "epoch": 0.5644354950701291,
+      "grad_norm": 4.362779075125485,
+      "learning_rate": 4.203337173055835e-06,
+      "loss": 0.4866,
+      "step": 8129
+    },
+    {
+      "epoch": 0.5645049298708513,
+      "grad_norm": 3.6279325163334257,
+      "learning_rate": 4.2022270712634576e-06,
+      "loss": 0.4996,
+      "step": 8130
+    },
+    {
+      "epoch": 0.5645743646715734,
+      "grad_norm": 3.460649103174236,
+      "learning_rate": 4.201117009821531e-06,
+      "loss": 0.4026,
+      "step": 8131
+    },
+    {
+      "epoch": 0.5646437994722955,
+      "grad_norm": 3.6763728338597885,
+      "learning_rate": 4.200006988786201e-06,
+      "loss": 0.5306,
+      "step": 8132
+    },
+    {
+      "epoch": 0.5647132342730177,
+      "grad_norm": 2.315921030171904,
+      "learning_rate": 4.198897008213614e-06,
+      "loss": 0.189,
+      "step": 8133
+    },
+    {
+      "epoch": 0.5647826690737398,
+      "grad_norm": 4.327994743204927,
+      "learning_rate": 4.19778706815991e-06,
+      "loss": 0.7816,
+      "step": 8134
+    },
+    {
+      "epoch": 0.5648521038744618,
+      "grad_norm": 2.526784547402994,
+      "learning_rate": 4.196677168681226e-06,
+      "loss": 0.1978,
+      "step": 8135
+    },
+    {
+      "epoch": 0.564921538675184,
+      "grad_norm": 2.610381848846158,
+      "learning_rate": 4.195567309833703e-06,
+      "loss": 0.3588,
+      "step": 8136
+    },
+    {
+      "epoch": 0.5649909734759061,
+      "grad_norm": 3.7727991895008213,
+      "learning_rate": 4.194457491673477e-06,
+      "loss": 0.4962,
+      "step": 8137
+    },
+    {
+      "epoch": 0.5650604082766283,
+      "grad_norm": 4.059890289448165,
+      "learning_rate": 4.1933477142566775e-06,
+      "loss": 0.5952,
+      "step": 8138
+    },
+    {
+      "epoch": 0.5651298430773504,
+      "grad_norm": 4.232470839620361,
+      "learning_rate": 4.192237977639437e-06,
+      "loss": 0.3745,
+      "step": 8139
+    },
+    {
+      "epoch": 0.5651992778780724,
+      "grad_norm": 3.1302027887411916,
+      "learning_rate": 4.191128281877887e-06,
+      "loss": 0.453,
+      "step": 8140
+    },
+    {
+      "epoch": 0.5652687126787946,
+      "grad_norm": 3.518813226779574,
+      "learning_rate": 4.190018627028154e-06,
+      "loss": 0.3857,
+      "step": 8141
+    },
+    {
+      "epoch": 0.5653381474795167,
+      "grad_norm": 3.58982894310111,
+      "learning_rate": 4.188909013146361e-06,
+      "loss": 0.4728,
+      "step": 8142
+    },
+    {
+      "epoch": 0.5654075822802389,
+      "grad_norm": 4.510067031770324,
+      "learning_rate": 4.187799440288632e-06,
+      "loss": 0.7264,
+      "step": 8143
+    },
+    {
+      "epoch": 0.565477017080961,
+      "grad_norm": 4.110441271929254,
+      "learning_rate": 4.186689908511091e-06,
+      "loss": 0.412,
+      "step": 8144
+    },
+    {
+      "epoch": 0.5655464518816831,
+      "grad_norm": 3.638411301016271,
+      "learning_rate": 4.185580417869851e-06,
+      "loss": 0.3549,
+      "step": 8145
+    },
+    {
+      "epoch": 0.5656158866824053,
+      "grad_norm": 5.9781689104685425,
+      "learning_rate": 4.184470968421033e-06,
+      "loss": 0.2968,
+      "step": 8146
+    },
+    {
+      "epoch": 0.5656853214831273,
+      "grad_norm": 3.477545016170227,
+      "learning_rate": 4.183361560220749e-06,
+      "loss": 0.3887,
+      "step": 8147
+    },
+    {
+      "epoch": 0.5657547562838494,
+      "grad_norm": 3.675754948258682,
+      "learning_rate": 4.182252193325116e-06,
+      "loss": 0.3711,
+      "step": 8148
+    },
+    {
+      "epoch": 0.5658241910845716,
+      "grad_norm": 3.7872778159432876,
+      "learning_rate": 4.18114286779024e-06,
+      "loss": 0.3829,
+      "step": 8149
+    },
+    {
+      "epoch": 0.5658936258852937,
+      "grad_norm": 3.349732192396922,
+      "learning_rate": 4.180033583672231e-06,
+      "loss": 0.3943,
+      "step": 8150
+    },
+    {
+      "epoch": 0.5659630606860159,
+      "grad_norm": 3.9521781762702917,
+      "learning_rate": 4.178924341027197e-06,
+      "loss": 0.4907,
+      "step": 8151
+    },
+    {
+      "epoch": 0.566032495486738,
+      "grad_norm": 5.12389887956287,
+      "learning_rate": 4.177815139911239e-06,
+      "loss": 0.682,
+      "step": 8152
+    },
+    {
+      "epoch": 0.56610193028746,
+      "grad_norm": 3.792130725067246,
+      "learning_rate": 4.176705980380462e-06,
+      "loss": 0.3857,
+      "step": 8153
+    },
+    {
+      "epoch": 0.5661713650881822,
+      "grad_norm": 4.0475833522663,
+      "learning_rate": 4.175596862490963e-06,
+      "loss": 0.5314,
+      "step": 8154
+    },
+    {
+      "epoch": 0.5662407998889043,
+      "grad_norm": 4.077816106461868,
+      "learning_rate": 4.174487786298845e-06,
+      "loss": 0.3961,
+      "step": 8155
+    },
+    {
+      "epoch": 0.5663102346896265,
+      "grad_norm": 2.799849899290536,
+      "learning_rate": 4.173378751860199e-06,
+      "loss": 0.4125,
+      "step": 8156
+    },
+    {
+      "epoch": 0.5663796694903486,
+      "grad_norm": 2.4496052269283792,
+      "learning_rate": 4.172269759231119e-06,
+      "loss": 0.2269,
+      "step": 8157
+    },
+    {
+      "epoch": 0.5664491042910706,
+      "grad_norm": 4.368187198904308,
+      "learning_rate": 4.171160808467699e-06,
+      "loss": 0.3796,
+      "step": 8158
+    },
+    {
+      "epoch": 0.5665185390917928,
+      "grad_norm": 3.659180489482833,
+      "learning_rate": 4.170051899626029e-06,
+      "loss": 0.293,
+      "step": 8159
+    },
+    {
+      "epoch": 0.5665879738925149,
+      "grad_norm": 3.493171828669165,
+      "learning_rate": 4.168943032762193e-06,
+      "loss": 0.3841,
+      "step": 8160
+    },
+    {
+      "epoch": 0.566657408693237,
+      "grad_norm": 4.391921531075759,
+      "learning_rate": 4.167834207932278e-06,
+      "loss": 0.6614,
+      "step": 8161
+    },
+    {
+      "epoch": 0.5667268434939592,
+      "grad_norm": 4.374730041533376,
+      "learning_rate": 4.1667254251923685e-06,
+      "loss": 0.5358,
+      "step": 8162
+    },
+    {
+      "epoch": 0.5667962782946813,
+      "grad_norm": 3.965302754317366,
+      "learning_rate": 4.165616684598543e-06,
+      "loss": 0.4544,
+      "step": 8163
+    },
+    {
+      "epoch": 0.5668657130954035,
+      "grad_norm": 4.1689496812477564,
+      "learning_rate": 4.164507986206881e-06,
+      "loss": 0.4955,
+      "step": 8164
+    },
+    {
+      "epoch": 0.5669351478961255,
+      "grad_norm": 4.04818930650821,
+      "learning_rate": 4.1633993300734595e-06,
+      "loss": 0.6027,
+      "step": 8165
+    },
+    {
+      "epoch": 0.5670045826968476,
+      "grad_norm": 4.426784094949415,
+      "learning_rate": 4.162290716254355e-06,
+      "loss": 0.5542,
+      "step": 8166
+    },
+    {
+      "epoch": 0.5670740174975698,
+      "grad_norm": 4.467901552743312,
+      "learning_rate": 4.1611821448056364e-06,
+      "loss": 0.5287,
+      "step": 8167
+    },
+    {
+      "epoch": 0.5671434522982919,
+      "grad_norm": 3.5207290458031424,
+      "learning_rate": 4.160073615783375e-06,
+      "loss": 0.399,
+      "step": 8168
+    },
+    {
+      "epoch": 0.5672128870990141,
+      "grad_norm": 6.4365259725437625,
+      "learning_rate": 4.158965129243643e-06,
+      "loss": 0.3712,
+      "step": 8169
+    },
+    {
+      "epoch": 0.5672823218997362,
+      "grad_norm": 2.9158785655865356,
+      "learning_rate": 4.1578566852425e-06,
+      "loss": 0.3681,
+      "step": 8170
+    },
+    {
+      "epoch": 0.5673517567004582,
+      "grad_norm": 4.835608873758036,
+      "learning_rate": 4.156748283836013e-06,
+      "loss": 0.6262,
+      "step": 8171
+    },
+    {
+      "epoch": 0.5674211915011804,
+      "grad_norm": 3.98491407674,
+      "learning_rate": 4.155639925080245e-06,
+      "loss": 0.3814,
+      "step": 8172
+    },
+    {
+      "epoch": 0.5674906263019025,
+      "grad_norm": 4.147869131977386,
+      "learning_rate": 4.154531609031255e-06,
+      "loss": 0.2885,
+      "step": 8173
+    },
+    {
+      "epoch": 0.5675600611026246,
+      "grad_norm": 3.8741148789186943,
+      "learning_rate": 4.153423335745097e-06,
+      "loss": 0.3922,
+      "step": 8174
+    },
+    {
+      "epoch": 0.5676294959033468,
+      "grad_norm": 4.183791892786589,
+      "learning_rate": 4.15231510527783e-06,
+      "loss": 0.5087,
+      "step": 8175
+    },
+    {
+      "epoch": 0.5676989307040688,
+      "grad_norm": 4.202114691470736,
+      "learning_rate": 4.151206917685508e-06,
+      "loss": 0.6302,
+      "step": 8176
+    },
+    {
+      "epoch": 0.567768365504791,
+      "grad_norm": 3.0739438071394924,
+      "learning_rate": 4.150098773024175e-06,
+      "loss": 0.2406,
+      "step": 8177
+    },
+    {
+      "epoch": 0.5678378003055131,
+      "grad_norm": 5.166071564980779,
+      "learning_rate": 4.148990671349887e-06,
+      "loss": 0.5116,
+      "step": 8178
+    },
+    {
+      "epoch": 0.5679072351062352,
+      "grad_norm": 3.0959499547210756,
+      "learning_rate": 4.147882612718687e-06,
+      "loss": 0.4007,
+      "step": 8179
+    },
+    {
+      "epoch": 0.5679766699069574,
+      "grad_norm": 3.2367790976698916,
+      "learning_rate": 4.146774597186622e-06,
+      "loss": 0.4305,
+      "step": 8180
+    },
+    {
+      "epoch": 0.5680461047076795,
+      "grad_norm": 3.886395889701452,
+      "learning_rate": 4.14566662480973e-06,
+      "loss": 0.5084,
+      "step": 8181
+    },
+    {
+      "epoch": 0.5681155395084017,
+      "grad_norm": 3.868854214754719,
+      "learning_rate": 4.144558695644054e-06,
+      "loss": 0.5114,
+      "step": 8182
+    },
+    {
+      "epoch": 0.5681849743091237,
+      "grad_norm": 3.5779261248020635,
+      "learning_rate": 4.143450809745631e-06,
+      "loss": 0.3348,
+      "step": 8183
+    },
+    {
+      "epoch": 0.5682544091098458,
+      "grad_norm": 4.118830080027121,
+      "learning_rate": 4.142342967170498e-06,
+      "loss": 0.5417,
+      "step": 8184
+    },
+    {
+      "epoch": 0.568323843910568,
+      "grad_norm": 4.436253411587737,
+      "learning_rate": 4.141235167974687e-06,
+      "loss": 0.5689,
+      "step": 8185
+    },
+    {
+      "epoch": 0.5683932787112901,
+      "grad_norm": 3.329951400517597,
+      "learning_rate": 4.1401274122142285e-06,
+      "loss": 0.2797,
+      "step": 8186
+    },
+    {
+      "epoch": 0.5684627135120123,
+      "grad_norm": 4.648025047439982,
+      "learning_rate": 4.139019699945154e-06,
+      "loss": 0.5371,
+      "step": 8187
+    },
+    {
+      "epoch": 0.5685321483127344,
+      "grad_norm": 3.6962183467864995,
+      "learning_rate": 4.137912031223487e-06,
+      "loss": 0.5129,
+      "step": 8188
+    },
+    {
+      "epoch": 0.5686015831134564,
+      "grad_norm": 3.6309138034480872,
+      "learning_rate": 4.136804406105256e-06,
+      "loss": 0.5365,
+      "step": 8189
+    },
+    {
+      "epoch": 0.5686710179141786,
+      "grad_norm": 3.710523533558838,
+      "learning_rate": 4.135696824646479e-06,
+      "loss": 0.4167,
+      "step": 8190
+    },
+    {
+      "epoch": 0.5687404527149007,
+      "grad_norm": 4.871071732358399,
+      "learning_rate": 4.134589286903181e-06,
+      "loss": 0.6011,
+      "step": 8191
+    },
+    {
+      "epoch": 0.5688098875156228,
+      "grad_norm": 4.371993090295723,
+      "learning_rate": 4.133481792931376e-06,
+      "loss": 0.7042,
+      "step": 8192
+    },
+    {
+      "epoch": 0.568879322316345,
+      "grad_norm": 4.752809887192429,
+      "learning_rate": 4.132374342787081e-06,
+      "loss": 0.6542,
+      "step": 8193
+    },
+    {
+      "epoch": 0.568948757117067,
+      "grad_norm": 4.10511814127952,
+      "learning_rate": 4.131266936526312e-06,
+      "loss": 0.2812,
+      "step": 8194
+    },
+    {
+      "epoch": 0.5690181919177892,
+      "grad_norm": 4.256562734406948,
+      "learning_rate": 4.130159574205077e-06,
+      "loss": 0.4694,
+      "step": 8195
+    },
+    {
+      "epoch": 0.5690876267185113,
+      "grad_norm": 2.757829124983594,
+      "learning_rate": 4.1290522558793855e-06,
+      "loss": 0.2625,
+      "step": 8196
+    },
+    {
+      "epoch": 0.5691570615192334,
+      "grad_norm": 4.976259576743357,
+      "learning_rate": 4.127944981605245e-06,
+      "loss": 0.5731,
+      "step": 8197
+    },
+    {
+      "epoch": 0.5692264963199556,
+      "grad_norm": 3.067717985681503,
+      "learning_rate": 4.126837751438663e-06,
+      "loss": 0.1835,
+      "step": 8198
+    },
+    {
+      "epoch": 0.5692959311206777,
+      "grad_norm": 4.603090658263815,
+      "learning_rate": 4.125730565435639e-06,
+      "loss": 0.61,
+      "step": 8199
+    },
+    {
+      "epoch": 0.5693653659213999,
+      "grad_norm": 5.141820677497581,
+      "learning_rate": 4.124623423652172e-06,
+      "loss": 0.6183,
+      "step": 8200
+    },
+    {
+      "epoch": 0.5694348007221219,
+      "grad_norm": 5.334562339426687,
+      "learning_rate": 4.123516326144264e-06,
+      "loss": 0.527,
+      "step": 8201
+    },
+    {
+      "epoch": 0.569504235522844,
+      "grad_norm": 4.938105013990115,
+      "learning_rate": 4.1224092729679065e-06,
+      "loss": 0.3932,
+      "step": 8202
+    },
+    {
+      "epoch": 0.5695736703235662,
+      "grad_norm": 4.886546812371977,
+      "learning_rate": 4.121302264179095e-06,
+      "loss": 0.5181,
+      "step": 8203
+    },
+    {
+      "epoch": 0.5696431051242883,
+      "grad_norm": 3.8297936736361935,
+      "learning_rate": 4.120195299833821e-06,
+      "loss": 0.5723,
+      "step": 8204
+    },
+    {
+      "epoch": 0.5697125399250104,
+      "grad_norm": 4.881437063272225,
+      "learning_rate": 4.1190883799880745e-06,
+      "loss": 0.6236,
+      "step": 8205
+    },
+    {
+      "epoch": 0.5697819747257326,
+      "grad_norm": 6.653667897876564,
+      "learning_rate": 4.117981504697839e-06,
+      "loss": 0.8079,
+      "step": 8206
+    },
+    {
+      "epoch": 0.5698514095264546,
+      "grad_norm": 3.86727325025805,
+      "learning_rate": 4.116874674019101e-06,
+      "loss": 0.3782,
+      "step": 8207
+    },
+    {
+      "epoch": 0.5699208443271768,
+      "grad_norm": 3.1006918668145316,
+      "learning_rate": 4.1157678880078436e-06,
+      "loss": 0.4621,
+      "step": 8208
+    },
+    {
+      "epoch": 0.5699902791278989,
+      "grad_norm": 3.8350282099357487,
+      "learning_rate": 4.114661146720047e-06,
+      "loss": 0.5566,
+      "step": 8209
+    },
+    {
+      "epoch": 0.570059713928621,
+      "grad_norm": 3.7037370882502816,
+      "learning_rate": 4.113554450211685e-06,
+      "loss": 0.383,
+      "step": 8210
+    },
+    {
+      "epoch": 0.5701291487293432,
+      "grad_norm": 4.151300314907731,
+      "learning_rate": 4.112447798538737e-06,
+      "loss": 0.2447,
+      "step": 8211
+    },
+    {
+      "epoch": 0.5701985835300653,
+      "grad_norm": 3.510790263352245,
+      "learning_rate": 4.111341191757176e-06,
+      "loss": 0.33,
+      "step": 8212
+    },
+    {
+      "epoch": 0.5702680183307874,
+      "grad_norm": 3.8732529086503704,
+      "learning_rate": 4.110234629922971e-06,
+      "loss": 0.2227,
+      "step": 8213
+    },
+    {
+      "epoch": 0.5703374531315095,
+      "grad_norm": 4.174888923701188,
+      "learning_rate": 4.109128113092092e-06,
+      "loss": 0.7096,
+      "step": 8214
+    },
+    {
+      "epoch": 0.5704068879322316,
+      "grad_norm": 3.6537448416740186,
+      "learning_rate": 4.108021641320504e-06,
+      "loss": 0.3924,
+      "step": 8215
+    },
+    {
+      "epoch": 0.5704763227329538,
+      "grad_norm": 3.731109541635737,
+      "learning_rate": 4.106915214664174e-06,
+      "loss": 0.4461,
+      "step": 8216
+    },
+    {
+      "epoch": 0.5705457575336759,
+      "grad_norm": 3.374042079141516,
+      "learning_rate": 4.10580883317906e-06,
+      "loss": 0.3449,
+      "step": 8217
+    },
+    {
+      "epoch": 0.570615192334398,
+      "grad_norm": 4.190076660298445,
+      "learning_rate": 4.104702496921123e-06,
+      "loss": 0.3763,
+      "step": 8218
+    },
+    {
+      "epoch": 0.5706846271351201,
+      "grad_norm": 3.576769621090016,
+      "learning_rate": 4.103596205946323e-06,
+      "loss": 0.4509,
+      "step": 8219
+    },
+    {
+      "epoch": 0.5707540619358422,
+      "grad_norm": 3.012672379462364,
+      "learning_rate": 4.1024899603106105e-06,
+      "loss": 0.3729,
+      "step": 8220
+    },
+    {
+      "epoch": 0.5708234967365644,
+      "grad_norm": 4.032490793076996,
+      "learning_rate": 4.101383760069941e-06,
+      "loss": 0.5286,
+      "step": 8221
+    },
+    {
+      "epoch": 0.5708929315372865,
+      "grad_norm": 4.497212335074443,
+      "learning_rate": 4.100277605280263e-06,
+      "loss": 0.5735,
+      "step": 8222
+    },
+    {
+      "epoch": 0.5709623663380086,
+      "grad_norm": 4.339599970119813,
+      "learning_rate": 4.099171495997528e-06,
+      "loss": 0.3435,
+      "step": 8223
+    },
+    {
+      "epoch": 0.5710318011387308,
+      "grad_norm": 4.227598600975215,
+      "learning_rate": 4.0980654322776775e-06,
+      "loss": 0.4741,
+      "step": 8224
+    },
+    {
+      "epoch": 0.5711012359394528,
+      "grad_norm": 4.323423024859913,
+      "learning_rate": 4.096959414176656e-06,
+      "loss": 0.7926,
+      "step": 8225
+    },
+    {
+      "epoch": 0.571170670740175,
+      "grad_norm": 3.5309036089927983,
+      "learning_rate": 4.0958534417504085e-06,
+      "loss": 0.3851,
+      "step": 8226
+    },
+    {
+      "epoch": 0.5712401055408971,
+      "grad_norm": 4.016654047361848,
+      "learning_rate": 4.094747515054869e-06,
+      "loss": 0.671,
+      "step": 8227
+    },
+    {
+      "epoch": 0.5713095403416192,
+      "grad_norm": 5.486159824766313,
+      "learning_rate": 4.093641634145974e-06,
+      "loss": 0.5476,
+      "step": 8228
+    },
+    {
+      "epoch": 0.5713789751423414,
+      "grad_norm": 3.107279755666281,
+      "learning_rate": 4.092535799079661e-06,
+      "loss": 0.4037,
+      "step": 8229
+    },
+    {
+      "epoch": 0.5714484099430635,
+      "grad_norm": 4.0484560224919655,
+      "learning_rate": 4.091430009911862e-06,
+      "loss": 0.3758,
+      "step": 8230
+    },
+    {
+      "epoch": 0.5715178447437855,
+      "grad_norm": 5.049242392125939,
+      "learning_rate": 4.090324266698504e-06,
+      "loss": 0.4354,
+      "step": 8231
+    },
+    {
+      "epoch": 0.5715872795445077,
+      "grad_norm": 4.945383010019044,
+      "learning_rate": 4.089218569495514e-06,
+      "loss": 0.4905,
+      "step": 8232
+    },
+    {
+      "epoch": 0.5716567143452298,
+      "grad_norm": 3.90100571268338,
+      "learning_rate": 4.088112918358819e-06,
+      "loss": 0.2238,
+      "step": 8233
+    },
+    {
+      "epoch": 0.571726149145952,
+      "grad_norm": 4.149724421606885,
+      "learning_rate": 4.087007313344342e-06,
+      "loss": 0.3803,
+      "step": 8234
+    },
+    {
+      "epoch": 0.5717955839466741,
+      "grad_norm": 3.676290486778727,
+      "learning_rate": 4.085901754508e-06,
+      "loss": 0.3579,
+      "step": 8235
+    },
+    {
+      "epoch": 0.5718650187473961,
+      "grad_norm": 3.6656625537944003,
+      "learning_rate": 4.084796241905713e-06,
+      "loss": 0.5401,
+      "step": 8236
+    },
+    {
+      "epoch": 0.5719344535481183,
+      "grad_norm": 3.2580458218522432,
+      "learning_rate": 4.083690775593399e-06,
+      "loss": 0.2925,
+      "step": 8237
+    },
+    {
+      "epoch": 0.5720038883488404,
+      "grad_norm": 3.9363137629387666,
+      "learning_rate": 4.082585355626965e-06,
+      "loss": 0.5388,
+      "step": 8238
+    },
+    {
+      "epoch": 0.5720733231495626,
+      "grad_norm": 4.000743846331227,
+      "learning_rate": 4.0814799820623275e-06,
+      "loss": 0.7363,
+      "step": 8239
+    },
+    {
+      "epoch": 0.5721427579502847,
+      "grad_norm": 3.8465567973638835,
+      "learning_rate": 4.080374654955393e-06,
+      "loss": 0.3689,
+      "step": 8240
+    },
+    {
+      "epoch": 0.5722121927510068,
+      "grad_norm": 4.64989596743859,
+      "learning_rate": 4.0792693743620695e-06,
+      "loss": 0.6305,
+      "step": 8241
+    },
+    {
+      "epoch": 0.572281627551729,
+      "grad_norm": 4.194235239910773,
+      "learning_rate": 4.078164140338256e-06,
+      "loss": 0.7083,
+      "step": 8242
+    },
+    {
+      "epoch": 0.572351062352451,
+      "grad_norm": 3.3921206302533835,
+      "learning_rate": 4.077058952939859e-06,
+      "loss": 0.3234,
+      "step": 8243
+    },
+    {
+      "epoch": 0.5724204971531732,
+      "grad_norm": 3.3887926702927524,
+      "learning_rate": 4.075953812222777e-06,
+      "loss": 0.2621,
+      "step": 8244
+    },
+    {
+      "epoch": 0.5724899319538953,
+      "grad_norm": 3.596906679910927,
+      "learning_rate": 4.074848718242903e-06,
+      "loss": 0.4348,
+      "step": 8245
+    },
+    {
+      "epoch": 0.5725593667546174,
+      "grad_norm": 4.848947362706893,
+      "learning_rate": 4.073743671056135e-06,
+      "loss": 0.5158,
+      "step": 8246
+    },
+    {
+      "epoch": 0.5726288015553396,
+      "grad_norm": 2.829421388768847,
+      "learning_rate": 4.072638670718363e-06,
+      "loss": 0.317,
+      "step": 8247
+    },
+    {
+      "epoch": 0.5726982363560617,
+      "grad_norm": 3.758014021769186,
+      "learning_rate": 4.07153371728548e-06,
+      "loss": 0.3952,
+      "step": 8248
+    },
+    {
+      "epoch": 0.5727676711567837,
+      "grad_norm": 4.1082237048614285,
+      "learning_rate": 4.070428810813369e-06,
+      "loss": 0.6192,
+      "step": 8249
+    },
+    {
+      "epoch": 0.5728371059575059,
+      "grad_norm": 2.948796228396795,
+      "learning_rate": 4.069323951357917e-06,
+      "loss": 0.1813,
+      "step": 8250
+    },
+    {
+      "epoch": 0.572906540758228,
+      "grad_norm": 2.9276160374647873,
+      "learning_rate": 4.068219138975008e-06,
+      "loss": 0.3392,
+      "step": 8251
+    },
+    {
+      "epoch": 0.5729759755589502,
+      "grad_norm": 4.598374429548369,
+      "learning_rate": 4.067114373720518e-06,
+      "loss": 0.5748,
+      "step": 8252
+    },
+    {
+      "epoch": 0.5730454103596723,
+      "grad_norm": 3.207341570442264,
+      "learning_rate": 4.0660096556503295e-06,
+      "loss": 0.3085,
+      "step": 8253
+    },
+    {
+      "epoch": 0.5731148451603943,
+      "grad_norm": 3.4050989042323794,
+      "learning_rate": 4.064904984820313e-06,
+      "loss": 0.3733,
+      "step": 8254
+    },
+    {
+      "epoch": 0.5731842799611165,
+      "grad_norm": 3.359431054972266,
+      "learning_rate": 4.063800361286349e-06,
+      "loss": 0.3887,
+      "step": 8255
+    },
+    {
+      "epoch": 0.5732537147618386,
+      "grad_norm": 4.063161873144555,
+      "learning_rate": 4.062695785104301e-06,
+      "loss": 0.504,
+      "step": 8256
+    },
+    {
+      "epoch": 0.5733231495625608,
+      "grad_norm": 4.1074015338232375,
+      "learning_rate": 4.061591256330038e-06,
+      "loss": 0.6453,
+      "step": 8257
+    },
+    {
+      "epoch": 0.5733925843632829,
+      "grad_norm": 3.5124201203747085,
+      "learning_rate": 4.060486775019429e-06,
+      "loss": 0.41,
+      "step": 8258
+    },
+    {
+      "epoch": 0.573462019164005,
+      "grad_norm": 4.2381340516835,
+      "learning_rate": 4.059382341228334e-06,
+      "loss": 0.4797,
+      "step": 8259
+    },
+    {
+      "epoch": 0.5735314539647272,
+      "grad_norm": 3.037417725016238,
+      "learning_rate": 4.058277955012617e-06,
+      "loss": 0.385,
+      "step": 8260
+    },
+    {
+      "epoch": 0.5736008887654492,
+      "grad_norm": 3.9205697561469934,
+      "learning_rate": 4.057173616428135e-06,
+      "loss": 0.5236,
+      "step": 8261
+    },
+    {
+      "epoch": 0.5736703235661713,
+      "grad_norm": 2.872065804430074,
+      "learning_rate": 4.056069325530746e-06,
+      "loss": 0.3457,
+      "step": 8262
+    },
+    {
+      "epoch": 0.5737397583668935,
+      "grad_norm": 3.4163167632435223,
+      "learning_rate": 4.054965082376302e-06,
+      "loss": 0.425,
+      "step": 8263
+    },
+    {
+      "epoch": 0.5738091931676156,
+      "grad_norm": 10.761068023792847,
+      "learning_rate": 4.053860887020653e-06,
+      "loss": 0.5428,
+      "step": 8264
+    },
+    {
+      "epoch": 0.5738786279683378,
+      "grad_norm": 5.886325752572919,
+      "learning_rate": 4.052756739519652e-06,
+      "loss": 0.4421,
+      "step": 8265
+    },
+    {
+      "epoch": 0.5739480627690599,
+      "grad_norm": 4.085505623340654,
+      "learning_rate": 4.051652639929143e-06,
+      "loss": 0.5554,
+      "step": 8266
+    },
+    {
+      "epoch": 0.5740174975697819,
+      "grad_norm": 4.136658766300606,
+      "learning_rate": 4.050548588304968e-06,
+      "loss": 0.3957,
+      "step": 8267
+    },
+    {
+      "epoch": 0.5740869323705041,
+      "grad_norm": 3.906387490028484,
+      "learning_rate": 4.049444584702972e-06,
+      "loss": 0.3925,
+      "step": 8268
+    },
+    {
+      "epoch": 0.5741563671712262,
+      "grad_norm": 4.196516660486522,
+      "learning_rate": 4.048340629178995e-06,
+      "loss": 0.52,
+      "step": 8269
+    },
+    {
+      "epoch": 0.5742258019719484,
+      "grad_norm": 2.8985905758339476,
+      "learning_rate": 4.047236721788871e-06,
+      "loss": 0.3754,
+      "step": 8270
+    },
+    {
+      "epoch": 0.5742952367726705,
+      "grad_norm": 6.115894965211508,
+      "learning_rate": 4.046132862588435e-06,
+      "loss": 0.4109,
+      "step": 8271
+    },
+    {
+      "epoch": 0.5743646715733925,
+      "grad_norm": 3.607390801562125,
+      "learning_rate": 4.0450290516335195e-06,
+      "loss": 0.4645,
+      "step": 8272
+    },
+    {
+      "epoch": 0.5744341063741147,
+      "grad_norm": 4.569251003926501,
+      "learning_rate": 4.043925288979956e-06,
+      "loss": 0.6653,
+      "step": 8273
+    },
+    {
+      "epoch": 0.5745035411748368,
+      "grad_norm": 3.487761169646035,
+      "learning_rate": 4.042821574683567e-06,
+      "loss": 0.4078,
+      "step": 8274
+    },
+    {
+      "epoch": 0.5745729759755589,
+      "grad_norm": 4.280770182506097,
+      "learning_rate": 4.04171790880018e-06,
+      "loss": 0.3805,
+      "step": 8275
+    },
+    {
+      "epoch": 0.5746424107762811,
+      "grad_norm": 3.883991240322071,
+      "learning_rate": 4.040614291385619e-06,
+      "loss": 0.3446,
+      "step": 8276
+    },
+    {
+      "epoch": 0.5747118455770032,
+      "grad_norm": 2.9467998866383995,
+      "learning_rate": 4.039510722495698e-06,
+      "loss": 0.3023,
+      "step": 8277
+    },
+    {
+      "epoch": 0.5747812803777254,
+      "grad_norm": 3.236039428818039,
+      "learning_rate": 4.038407202186239e-06,
+      "loss": 0.3014,
+      "step": 8278
+    },
+    {
+      "epoch": 0.5748507151784474,
+      "grad_norm": 3.14964037069236,
+      "learning_rate": 4.037303730513056e-06,
+      "loss": 0.281,
+      "step": 8279
+    },
+    {
+      "epoch": 0.5749201499791695,
+      "grad_norm": 3.8918177628176442,
+      "learning_rate": 4.036200307531962e-06,
+      "loss": 0.5459,
+      "step": 8280
+    },
+    {
+      "epoch": 0.5749895847798917,
+      "grad_norm": 2.9748550421965843,
+      "learning_rate": 4.035096933298764e-06,
+      "loss": 0.2267,
+      "step": 8281
+    },
+    {
+      "epoch": 0.5750590195806138,
+      "grad_norm": 4.169971190126039,
+      "learning_rate": 4.033993607869273e-06,
+      "loss": 0.3896,
+      "step": 8282
+    },
+    {
+      "epoch": 0.575128454381336,
+      "grad_norm": 2.344897522663576,
+      "learning_rate": 4.032890331299292e-06,
+      "loss": 0.2403,
+      "step": 8283
+    },
+    {
+      "epoch": 0.575197889182058,
+      "grad_norm": 3.4844068283426006,
+      "learning_rate": 4.031787103644623e-06,
+      "loss": 0.3134,
+      "step": 8284
+    },
+    {
+      "epoch": 0.5752673239827801,
+      "grad_norm": 3.7517548270607093,
+      "learning_rate": 4.030683924961066e-06,
+      "loss": 0.4406,
+      "step": 8285
+    },
+    {
+      "epoch": 0.5753367587835023,
+      "grad_norm": 3.2762846799811007,
+      "learning_rate": 4.02958079530442e-06,
+      "loss": 0.3804,
+      "step": 8286
+    },
+    {
+      "epoch": 0.5754061935842244,
+      "grad_norm": 3.3957480830337214,
+      "learning_rate": 4.02847771473048e-06,
+      "loss": 0.444,
+      "step": 8287
+    },
+    {
+      "epoch": 0.5754756283849465,
+      "grad_norm": 3.437316311920712,
+      "learning_rate": 4.027374683295036e-06,
+      "loss": 0.4216,
+      "step": 8288
+    },
+    {
+      "epoch": 0.5755450631856687,
+      "grad_norm": 6.355746239357887,
+      "learning_rate": 4.026271701053881e-06,
+      "loss": 0.7195,
+      "step": 8289
+    },
+    {
+      "epoch": 0.5756144979863907,
+      "grad_norm": 5.734150169101989,
+      "learning_rate": 4.025168768062801e-06,
+      "loss": 0.8573,
+      "step": 8290
+    },
+    {
+      "epoch": 0.5756839327871129,
+      "grad_norm": 3.7545094896807294,
+      "learning_rate": 4.024065884377583e-06,
+      "loss": 0.541,
+      "step": 8291
+    },
+    {
+      "epoch": 0.575753367587835,
+      "grad_norm": 5.287401368554198,
+      "learning_rate": 4.0229630500540075e-06,
+      "loss": 0.4149,
+      "step": 8292
+    },
+    {
+      "epoch": 0.5758228023885571,
+      "grad_norm": 2.623034989930876,
+      "learning_rate": 4.021860265147854e-06,
+      "loss": 0.3252,
+      "step": 8293
+    },
+    {
+      "epoch": 0.5758922371892793,
+      "grad_norm": 3.303578251172555,
+      "learning_rate": 4.020757529714904e-06,
+      "loss": 0.2271,
+      "step": 8294
+    },
+    {
+      "epoch": 0.5759616719900014,
+      "grad_norm": 4.450790331947789,
+      "learning_rate": 4.019654843810929e-06,
+      "loss": 0.5132,
+      "step": 8295
+    },
+    {
+      "epoch": 0.5760311067907236,
+      "grad_norm": 5.229785866138677,
+      "learning_rate": 4.018552207491702e-06,
+      "loss": 0.7396,
+      "step": 8296
+    },
+    {
+      "epoch": 0.5761005415914456,
+      "grad_norm": 4.4838989114748165,
+      "learning_rate": 4.017449620812992e-06,
+      "loss": 0.4912,
+      "step": 8297
+    },
+    {
+      "epoch": 0.5761699763921677,
+      "grad_norm": 3.8068326485199884,
+      "learning_rate": 4.016347083830572e-06,
+      "loss": 0.5299,
+      "step": 8298
+    },
+    {
+      "epoch": 0.5762394111928899,
+      "grad_norm": 3.1571259414289674,
+      "learning_rate": 4.015244596600202e-06,
+      "loss": 0.3301,
+      "step": 8299
+    },
+    {
+      "epoch": 0.576308845993612,
+      "grad_norm": 5.927453176704675,
+      "learning_rate": 4.014142159177645e-06,
+      "loss": 0.6132,
+      "step": 8300
+    },
+    {
+      "epoch": 0.5763782807943342,
+      "grad_norm": 3.993876816356581,
+      "learning_rate": 4.013039771618664e-06,
+      "loss": 0.5652,
+      "step": 8301
+    },
+    {
+      "epoch": 0.5764477155950563,
+      "grad_norm": 4.538385168103553,
+      "learning_rate": 4.011937433979014e-06,
+      "loss": 0.5237,
+      "step": 8302
+    },
+    {
+      "epoch": 0.5765171503957783,
+      "grad_norm": 4.813324538972775,
+      "learning_rate": 4.010835146314449e-06,
+      "loss": 0.4832,
+      "step": 8303
+    },
+    {
+      "epoch": 0.5765865851965005,
+      "grad_norm": 3.97776369235501,
+      "learning_rate": 4.009732908680724e-06,
+      "loss": 0.557,
+      "step": 8304
+    },
+    {
+      "epoch": 0.5766560199972226,
+      "grad_norm": 3.6619011202153247,
+      "learning_rate": 4.00863072113359e-06,
+      "loss": 0.4126,
+      "step": 8305
+    },
+    {
+      "epoch": 0.5767254547979447,
+      "grad_norm": 4.47160206177307,
+      "learning_rate": 4.007528583728788e-06,
+      "loss": 0.4051,
+      "step": 8306
+    },
+    {
+      "epoch": 0.5767948895986669,
+      "grad_norm": 4.8179334562298735,
+      "learning_rate": 4.00642649652207e-06,
+      "loss": 0.4593,
+      "step": 8307
+    },
+    {
+      "epoch": 0.576864324399389,
+      "grad_norm": 3.1833440261712727,
+      "learning_rate": 4.005324459569176e-06,
+      "loss": 0.2362,
+      "step": 8308
+    },
+    {
+      "epoch": 0.5769337592001111,
+      "grad_norm": 4.141055657701225,
+      "learning_rate": 4.004222472925843e-06,
+      "loss": 0.2366,
+      "step": 8309
+    },
+    {
+      "epoch": 0.5770031940008332,
+      "grad_norm": 3.0514158939574108,
+      "learning_rate": 4.003120536647811e-06,
+      "loss": 0.4718,
+      "step": 8310
+    },
+    {
+      "epoch": 0.5770726288015553,
+      "grad_norm": 3.4230497066950556,
+      "learning_rate": 4.002018650790815e-06,
+      "loss": 0.2625,
+      "step": 8311
+    },
+    {
+      "epoch": 0.5771420636022775,
+      "grad_norm": 4.803066722102279,
+      "learning_rate": 4.000916815410587e-06,
+      "loss": 0.7524,
+      "step": 8312
+    },
+    {
+      "epoch": 0.5772114984029996,
+      "grad_norm": 3.4898718891381,
+      "learning_rate": 3.999815030562855e-06,
+      "loss": 0.5221,
+      "step": 8313
+    },
+    {
+      "epoch": 0.5772809332037218,
+      "grad_norm": 3.6619907669021785,
+      "learning_rate": 3.998713296303347e-06,
+      "loss": 0.4321,
+      "step": 8314
+    },
+    {
+      "epoch": 0.5773503680044438,
+      "grad_norm": 3.729229086518605,
+      "learning_rate": 3.997611612687786e-06,
+      "loss": 0.4449,
+      "step": 8315
+    },
+    {
+      "epoch": 0.5774198028051659,
+      "grad_norm": 2.8968451595450944,
+      "learning_rate": 3.996509979771899e-06,
+      "loss": 0.2336,
+      "step": 8316
+    },
+    {
+      "epoch": 0.5774892376058881,
+      "grad_norm": 3.542188430787019,
+      "learning_rate": 3.995408397611399e-06,
+      "loss": 0.4427,
+      "step": 8317
+    },
+    {
+      "epoch": 0.5775586724066102,
+      "grad_norm": 10.303692356523682,
+      "learning_rate": 3.994306866262005e-06,
+      "loss": 0.6655,
+      "step": 8318
+    },
+    {
+      "epoch": 0.5776281072073323,
+      "grad_norm": 2.678691479838158,
+      "learning_rate": 3.993205385779434e-06,
+      "loss": 0.2019,
+      "step": 8319
+    },
+    {
+      "epoch": 0.5776975420080545,
+      "grad_norm": 3.0774692979372746,
+      "learning_rate": 3.992103956219393e-06,
+      "loss": 0.3375,
+      "step": 8320
+    },
+    {
+      "epoch": 0.5777669768087765,
+      "grad_norm": 4.43823016801703,
+      "learning_rate": 3.991002577637595e-06,
+      "loss": 0.6069,
+      "step": 8321
+    },
+    {
+      "epoch": 0.5778364116094987,
+      "grad_norm": 4.0164346791106125,
+      "learning_rate": 3.989901250089743e-06,
+      "loss": 0.5368,
+      "step": 8322
+    },
+    {
+      "epoch": 0.5779058464102208,
+      "grad_norm": 3.925388077439334,
+      "learning_rate": 3.988799973631546e-06,
+      "loss": 0.4042,
+      "step": 8323
+    },
+    {
+      "epoch": 0.5779752812109429,
+      "grad_norm": 2.2756170192592777,
+      "learning_rate": 3.9876987483187005e-06,
+      "loss": 0.1972,
+      "step": 8324
+    },
+    {
+      "epoch": 0.5780447160116651,
+      "grad_norm": 3.8199855051037748,
+      "learning_rate": 3.9865975742069055e-06,
+      "loss": 0.3483,
+      "step": 8325
+    },
+    {
+      "epoch": 0.5781141508123872,
+      "grad_norm": 3.367689396144764,
+      "learning_rate": 3.98549645135186e-06,
+      "loss": 0.3895,
+      "step": 8326
+    },
+    {
+      "epoch": 0.5781835856131093,
+      "grad_norm": 3.6054070777508493,
+      "learning_rate": 3.984395379809256e-06,
+      "loss": 0.4893,
+      "step": 8327
+    },
+    {
+      "epoch": 0.5782530204138314,
+      "grad_norm": 5.23973456659734,
+      "learning_rate": 3.983294359634783e-06,
+      "loss": 0.6332,
+      "step": 8328
+    },
+    {
+      "epoch": 0.5783224552145535,
+      "grad_norm": 3.564335850316491,
+      "learning_rate": 3.98219339088413e-06,
+      "loss": 0.3913,
+      "step": 8329
+    },
+    {
+      "epoch": 0.5783918900152757,
+      "grad_norm": 3.2602457904661364,
+      "learning_rate": 3.981092473612987e-06,
+      "loss": 0.3611,
+      "step": 8330
+    },
+    {
+      "epoch": 0.5784613248159978,
+      "grad_norm": 6.553837489491782,
+      "learning_rate": 3.979991607877032e-06,
+      "loss": 0.9449,
+      "step": 8331
+    },
+    {
+      "epoch": 0.5785307596167198,
+      "grad_norm": 4.596923151282438,
+      "learning_rate": 3.978890793731947e-06,
+      "loss": 0.4192,
+      "step": 8332
+    },
+    {
+      "epoch": 0.578600194417442,
+      "grad_norm": 3.7962532720516124,
+      "learning_rate": 3.977790031233411e-06,
+      "loss": 0.5468,
+      "step": 8333
+    },
+    {
+      "epoch": 0.5786696292181641,
+      "grad_norm": 5.051442999233338,
+      "learning_rate": 3.976689320437098e-06,
+      "loss": 0.5829,
+      "step": 8334
+    },
+    {
+      "epoch": 0.5787390640188863,
+      "grad_norm": 3.125875125123957,
+      "learning_rate": 3.9755886613986806e-06,
+      "loss": 0.3667,
+      "step": 8335
+    },
+    {
+      "epoch": 0.5788084988196084,
+      "grad_norm": 4.08034835110671,
+      "learning_rate": 3.97448805417383e-06,
+      "loss": 0.5218,
+      "step": 8336
+    },
+    {
+      "epoch": 0.5788779336203305,
+      "grad_norm": 5.200081542687821,
+      "learning_rate": 3.9733874988182145e-06,
+      "loss": 0.7339,
+      "step": 8337
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 4.351381081037609,
+      "learning_rate": 3.972286995387495e-06,
+      "loss": 0.6607,
+      "step": 8338
+    },
+    {
+      "epoch": 0.5790168032217747,
+      "grad_norm": 2.748537906854643,
+      "learning_rate": 3.971186543937337e-06,
+      "loss": 0.1615,
+      "step": 8339
+    },
+    {
+      "epoch": 0.5790862380224969,
+      "grad_norm": 3.2936182164727166,
+      "learning_rate": 3.9700861445234e-06,
+      "loss": 0.3194,
+      "step": 8340
+    },
+    {
+      "epoch": 0.579155672823219,
+      "grad_norm": 4.5301561613284855,
+      "learning_rate": 3.968985797201343e-06,
+      "loss": 0.5029,
+      "step": 8341
+    },
+    {
+      "epoch": 0.5792251076239411,
+      "grad_norm": 4.713040705120686,
+      "learning_rate": 3.9678855020268135e-06,
+      "loss": 0.2712,
+      "step": 8342
+    },
+    {
+      "epoch": 0.5792945424246633,
+      "grad_norm": 4.391819122280112,
+      "learning_rate": 3.966785259055469e-06,
+      "loss": 0.5121,
+      "step": 8343
+    },
+    {
+      "epoch": 0.5793639772253854,
+      "grad_norm": 4.020215422494957,
+      "learning_rate": 3.965685068342958e-06,
+      "loss": 0.4846,
+      "step": 8344
+    },
+    {
+      "epoch": 0.5794334120261074,
+      "grad_norm": 3.8145602613541683,
+      "learning_rate": 3.9645849299449245e-06,
+      "loss": 0.4079,
+      "step": 8345
+    },
+    {
+      "epoch": 0.5795028468268296,
+      "grad_norm": 5.172869668220643,
+      "learning_rate": 3.963484843917014e-06,
+      "loss": 0.7602,
+      "step": 8346
+    },
+    {
+      "epoch": 0.5795722816275517,
+      "grad_norm": 4.389851392994651,
+      "learning_rate": 3.962384810314867e-06,
+      "loss": 0.6465,
+      "step": 8347
+    },
+    {
+      "epoch": 0.5796417164282739,
+      "grad_norm": 4.421636816440425,
+      "learning_rate": 3.9612848291941234e-06,
+      "loss": 0.4587,
+      "step": 8348
+    },
+    {
+      "epoch": 0.579711151228996,
+      "grad_norm": 3.7393576479878536,
+      "learning_rate": 3.960184900610416e-06,
+      "loss": 0.5233,
+      "step": 8349
+    },
+    {
+      "epoch": 0.579780586029718,
+      "grad_norm": 4.315731618211668,
+      "learning_rate": 3.95908502461938e-06,
+      "loss": 0.5783,
+      "step": 8350
+    },
+    {
+      "epoch": 0.5798500208304402,
+      "grad_norm": 4.116040934602692,
+      "learning_rate": 3.957985201276647e-06,
+      "loss": 0.4575,
+      "step": 8351
+    },
+    {
+      "epoch": 0.5799194556311623,
+      "grad_norm": 4.828259289999425,
+      "learning_rate": 3.956885430637841e-06,
+      "loss": 0.4806,
+      "step": 8352
+    },
+    {
+      "epoch": 0.5799888904318845,
+      "grad_norm": 3.9071177219333877,
+      "learning_rate": 3.955785712758591e-06,
+      "loss": 0.5295,
+      "step": 8353
+    },
+    {
+      "epoch": 0.5800583252326066,
+      "grad_norm": 3.6242505495354087,
+      "learning_rate": 3.954686047694517e-06,
+      "loss": 0.4358,
+      "step": 8354
+    },
+    {
+      "epoch": 0.5801277600333287,
+      "grad_norm": 3.8431766203035234,
+      "learning_rate": 3.9535864355012425e-06,
+      "loss": 0.5996,
+      "step": 8355
+    },
+    {
+      "epoch": 0.5801971948340509,
+      "grad_norm": 5.256780190710877,
+      "learning_rate": 3.95248687623438e-06,
+      "loss": 0.663,
+      "step": 8356
+    },
+    {
+      "epoch": 0.5802666296347729,
+      "grad_norm": 4.930484442314402,
+      "learning_rate": 3.951387369949545e-06,
+      "loss": 0.5178,
+      "step": 8357
+    },
+    {
+      "epoch": 0.5803360644354951,
+      "grad_norm": 2.889109550262033,
+      "learning_rate": 3.950287916702354e-06,
+      "loss": 0.2985,
+      "step": 8358
+    },
+    {
+      "epoch": 0.5804054992362172,
+      "grad_norm": 4.232154749773528,
+      "learning_rate": 3.949188516548409e-06,
+      "loss": 0.3386,
+      "step": 8359
+    },
+    {
+      "epoch": 0.5804749340369393,
+      "grad_norm": 3.2420368922988847,
+      "learning_rate": 3.9480891695433195e-06,
+      "loss": 0.317,
+      "step": 8360
+    },
+    {
+      "epoch": 0.5805443688376615,
+      "grad_norm": 4.291089469216785,
+      "learning_rate": 3.94698987574269e-06,
+      "loss": 0.3614,
+      "step": 8361
+    },
+    {
+      "epoch": 0.5806138036383836,
+      "grad_norm": 3.520656868587251,
+      "learning_rate": 3.945890635202123e-06,
+      "loss": 0.3863,
+      "step": 8362
+    },
+    {
+      "epoch": 0.5806832384391056,
+      "grad_norm": 3.3824923807433773,
+      "learning_rate": 3.944791447977213e-06,
+      "loss": 0.2196,
+      "step": 8363
+    },
+    {
+      "epoch": 0.5807526732398278,
+      "grad_norm": 4.406458554674027,
+      "learning_rate": 3.9436923141235574e-06,
+      "loss": 0.5993,
+      "step": 8364
+    },
+    {
+      "epoch": 0.5808221080405499,
+      "grad_norm": 2.8893878532168,
+      "learning_rate": 3.94259323369675e-06,
+      "loss": 0.3231,
+      "step": 8365
+    },
+    {
+      "epoch": 0.5808915428412721,
+      "grad_norm": 3.401012218895298,
+      "learning_rate": 3.941494206752381e-06,
+      "loss": 0.3107,
+      "step": 8366
+    },
+    {
+      "epoch": 0.5809609776419942,
+      "grad_norm": 3.6532482587450565,
+      "learning_rate": 3.940395233346036e-06,
+      "loss": 0.397,
+      "step": 8367
+    },
+    {
+      "epoch": 0.5810304124427162,
+      "grad_norm": 4.010710991936554,
+      "learning_rate": 3.939296313533301e-06,
+      "loss": 0.5426,
+      "step": 8368
+    },
+    {
+      "epoch": 0.5810998472434384,
+      "grad_norm": 2.777008502644784,
+      "learning_rate": 3.938197447369759e-06,
+      "loss": 0.2758,
+      "step": 8369
+    },
+    {
+      "epoch": 0.5811692820441605,
+      "grad_norm": 4.163724887190737,
+      "learning_rate": 3.937098634910989e-06,
+      "loss": 0.5743,
+      "step": 8370
+    },
+    {
+      "epoch": 0.5812387168448827,
+      "grad_norm": 3.3126008796877575,
+      "learning_rate": 3.9359998762125655e-06,
+      "loss": 0.3639,
+      "step": 8371
+    },
+    {
+      "epoch": 0.5813081516456048,
+      "grad_norm": 3.3345488838460353,
+      "learning_rate": 3.934901171330065e-06,
+      "loss": 0.4139,
+      "step": 8372
+    },
+    {
+      "epoch": 0.5813775864463269,
+      "grad_norm": 4.64065549538711,
+      "learning_rate": 3.93380252031906e-06,
+      "loss": 0.4962,
+      "step": 8373
+    },
+    {
+      "epoch": 0.5814470212470491,
+      "grad_norm": 4.018089025717462,
+      "learning_rate": 3.932703923235115e-06,
+      "loss": 0.5307,
+      "step": 8374
+    },
+    {
+      "epoch": 0.5815164560477711,
+      "grad_norm": 4.165649964919355,
+      "learning_rate": 3.9316053801337975e-06,
+      "loss": 0.4472,
+      "step": 8375
+    },
+    {
+      "epoch": 0.5815858908484932,
+      "grad_norm": 3.341719043285829,
+      "learning_rate": 3.930506891070673e-06,
+      "loss": 0.3559,
+      "step": 8376
+    },
+    {
+      "epoch": 0.5816553256492154,
+      "grad_norm": 3.935040521240501,
+      "learning_rate": 3.929408456101298e-06,
+      "loss": 0.3765,
+      "step": 8377
+    },
+    {
+      "epoch": 0.5817247604499375,
+      "grad_norm": 5.183353394951048,
+      "learning_rate": 3.9283100752812306e-06,
+      "loss": 0.7425,
+      "step": 8378
+    },
+    {
+      "epoch": 0.5817941952506597,
+      "grad_norm": 3.8219290875100342,
+      "learning_rate": 3.927211748666029e-06,
+      "loss": 0.5006,
+      "step": 8379
+    },
+    {
+      "epoch": 0.5818636300513818,
+      "grad_norm": 4.245770193129816,
+      "learning_rate": 3.926113476311244e-06,
+      "loss": 0.6895,
+      "step": 8380
+    },
+    {
+      "epoch": 0.5819330648521038,
+      "grad_norm": 3.8657273928962828,
+      "learning_rate": 3.925015258272422e-06,
+      "loss": 0.568,
+      "step": 8381
+    },
+    {
+      "epoch": 0.582002499652826,
+      "grad_norm": 2.510081813948898,
+      "learning_rate": 3.923917094605113e-06,
+      "loss": 0.2302,
+      "step": 8382
+    },
+    {
+      "epoch": 0.5820719344535481,
+      "grad_norm": 3.9545930185537284,
+      "learning_rate": 3.92281898536486e-06,
+      "loss": 0.5249,
+      "step": 8383
+    },
+    {
+      "epoch": 0.5821413692542703,
+      "grad_norm": 4.138970728580292,
+      "learning_rate": 3.921720930607203e-06,
+      "loss": 0.5818,
+      "step": 8384
+    },
+    {
+      "epoch": 0.5822108040549924,
+      "grad_norm": 2.657277571267311,
+      "learning_rate": 3.920622930387681e-06,
+      "loss": 0.1763,
+      "step": 8385
+    },
+    {
+      "epoch": 0.5822802388557144,
+      "grad_norm": 3.5027391466759363,
+      "learning_rate": 3.9195249847618285e-06,
+      "loss": 0.3472,
+      "step": 8386
+    },
+    {
+      "epoch": 0.5823496736564366,
+      "grad_norm": 4.819812076049637,
+      "learning_rate": 3.918427093785183e-06,
+      "loss": 0.4851,
+      "step": 8387
+    },
+    {
+      "epoch": 0.5824191084571587,
+      "grad_norm": 3.4196915628035707,
+      "learning_rate": 3.917329257513269e-06,
+      "loss": 0.2543,
+      "step": 8388
+    },
+    {
+      "epoch": 0.5824885432578808,
+      "grad_norm": 5.104625973904496,
+      "learning_rate": 3.916231476001616e-06,
+      "loss": 0.8003,
+      "step": 8389
+    },
+    {
+      "epoch": 0.582557978058603,
+      "grad_norm": 2.715210096401347,
+      "learning_rate": 3.915133749305748e-06,
+      "loss": 0.3185,
+      "step": 8390
+    },
+    {
+      "epoch": 0.5826274128593251,
+      "grad_norm": 3.275802814641016,
+      "learning_rate": 3.91403607748119e-06,
+      "loss": 0.2854,
+      "step": 8391
+    },
+    {
+      "epoch": 0.5826968476600473,
+      "grad_norm": 4.9067210119764395,
+      "learning_rate": 3.912938460583456e-06,
+      "loss": 0.3857,
+      "step": 8392
+    },
+    {
+      "epoch": 0.5827662824607693,
+      "grad_norm": 4.004462018097176,
+      "learning_rate": 3.911840898668065e-06,
+      "loss": 0.4379,
+      "step": 8393
+    },
+    {
+      "epoch": 0.5828357172614914,
+      "grad_norm": 5.375110249648641,
+      "learning_rate": 3.910743391790532e-06,
+      "loss": 0.7087,
+      "step": 8394
+    },
+    {
+      "epoch": 0.5829051520622136,
+      "grad_norm": 3.3378333116121435,
+      "learning_rate": 3.909645940006364e-06,
+      "loss": 0.404,
+      "step": 8395
+    },
+    {
+      "epoch": 0.5829745868629357,
+      "grad_norm": 4.329949587818784,
+      "learning_rate": 3.908548543371069e-06,
+      "loss": 0.5316,
+      "step": 8396
+    },
+    {
+      "epoch": 0.5830440216636579,
+      "grad_norm": 3.0871269555376424,
+      "learning_rate": 3.9074512019401555e-06,
+      "loss": 0.2353,
+      "step": 8397
+    },
+    {
+      "epoch": 0.58311345646438,
+      "grad_norm": 4.657783021310161,
+      "learning_rate": 3.906353915769125e-06,
+      "loss": 0.6441,
+      "step": 8398
+    },
+    {
+      "epoch": 0.583182891265102,
+      "grad_norm": 3.1496711144035108,
+      "learning_rate": 3.905256684913473e-06,
+      "loss": 0.3655,
+      "step": 8399
+    },
+    {
+      "epoch": 0.5832523260658242,
+      "grad_norm": 2.620243311916356,
+      "learning_rate": 3.9041595094287e-06,
+      "loss": 0.2785,
+      "step": 8400
+    },
+    {
+      "epoch": 0.5833217608665463,
+      "grad_norm": 3.6432526829370118,
+      "learning_rate": 3.903062389370301e-06,
+      "loss": 0.3487,
+      "step": 8401
+    },
+    {
+      "epoch": 0.5833911956672684,
+      "grad_norm": 4.440160697921049,
+      "learning_rate": 3.901965324793764e-06,
+      "loss": 0.3752,
+      "step": 8402
+    },
+    {
+      "epoch": 0.5834606304679906,
+      "grad_norm": 5.492622623658774,
+      "learning_rate": 3.9008683157545765e-06,
+      "loss": 0.5273,
+      "step": 8403
+    },
+    {
+      "epoch": 0.5835300652687126,
+      "grad_norm": 4.439879393608571,
+      "learning_rate": 3.899771362308228e-06,
+      "loss": 0.6175,
+      "step": 8404
+    },
+    {
+      "epoch": 0.5835995000694348,
+      "grad_norm": 6.350907617978958,
+      "learning_rate": 3.8986744645101994e-06,
+      "loss": 0.6469,
+      "step": 8405
+    },
+    {
+      "epoch": 0.5836689348701569,
+      "grad_norm": 3.3756604517968545,
+      "learning_rate": 3.897577622415968e-06,
+      "loss": 0.3683,
+      "step": 8406
+    },
+    {
+      "epoch": 0.583738369670879,
+      "grad_norm": 3.2427457915381033,
+      "learning_rate": 3.896480836081014e-06,
+      "loss": 0.3068,
+      "step": 8407
+    },
+    {
+      "epoch": 0.5838078044716012,
+      "grad_norm": 3.96406869259864,
+      "learning_rate": 3.895384105560812e-06,
+      "loss": 0.4709,
+      "step": 8408
+    },
+    {
+      "epoch": 0.5838772392723233,
+      "grad_norm": 3.650824203760492,
+      "learning_rate": 3.894287430910829e-06,
+      "loss": 0.2619,
+      "step": 8409
+    },
+    {
+      "epoch": 0.5839466740730455,
+      "grad_norm": 3.7244571852663095,
+      "learning_rate": 3.893190812186538e-06,
+      "loss": 0.5548,
+      "step": 8410
+    },
+    {
+      "epoch": 0.5840161088737675,
+      "grad_norm": 4.751765833161116,
+      "learning_rate": 3.892094249443403e-06,
+      "loss": 0.4457,
+      "step": 8411
+    },
+    {
+      "epoch": 0.5840855436744896,
+      "grad_norm": 4.12967003817428,
+      "learning_rate": 3.890997742736889e-06,
+      "loss": 0.6108,
+      "step": 8412
+    },
+    {
+      "epoch": 0.5841549784752118,
+      "grad_norm": 3.6533504670571486,
+      "learning_rate": 3.889901292122452e-06,
+      "loss": 0.4312,
+      "step": 8413
+    },
+    {
+      "epoch": 0.5842244132759339,
+      "grad_norm": 3.3699815215688744,
+      "learning_rate": 3.888804897655553e-06,
+      "loss": 0.3667,
+      "step": 8414
+    },
+    {
+      "epoch": 0.5842938480766561,
+      "grad_norm": 3.696356017414049,
+      "learning_rate": 3.887708559391643e-06,
+      "loss": 0.4391,
+      "step": 8415
+    },
+    {
+      "epoch": 0.5843632828773782,
+      "grad_norm": 3.7280304797598873,
+      "learning_rate": 3.886612277386178e-06,
+      "loss": 0.4402,
+      "step": 8416
+    },
+    {
+      "epoch": 0.5844327176781002,
+      "grad_norm": 4.858284882106675,
+      "learning_rate": 3.885516051694604e-06,
+      "loss": 0.7441,
+      "step": 8417
+    },
+    {
+      "epoch": 0.5845021524788224,
+      "grad_norm": 4.323521178526095,
+      "learning_rate": 3.884419882372364e-06,
+      "loss": 0.5589,
+      "step": 8418
+    },
+    {
+      "epoch": 0.5845715872795445,
+      "grad_norm": 4.402313337475904,
+      "learning_rate": 3.883323769474907e-06,
+      "loss": 0.6068,
+      "step": 8419
+    },
+    {
+      "epoch": 0.5846410220802666,
+      "grad_norm": 3.645314357194898,
+      "learning_rate": 3.882227713057668e-06,
+      "loss": 0.2983,
+      "step": 8420
+    },
+    {
+      "epoch": 0.5847104568809888,
+      "grad_norm": 17.097228147996045,
+      "learning_rate": 3.881131713176087e-06,
+      "loss": 0.3851,
+      "step": 8421
+    },
+    {
+      "epoch": 0.5847798916817109,
+      "grad_norm": 3.790211358874174,
+      "learning_rate": 3.880035769885597e-06,
+      "loss": 0.3715,
+      "step": 8422
+    },
+    {
+      "epoch": 0.584849326482433,
+      "grad_norm": 4.874024090775505,
+      "learning_rate": 3.878939883241633e-06,
+      "loss": 0.6526,
+      "step": 8423
+    },
+    {
+      "epoch": 0.5849187612831551,
+      "grad_norm": 3.8993941694969596,
+      "learning_rate": 3.87784405329962e-06,
+      "loss": 0.3321,
+      "step": 8424
+    },
+    {
+      "epoch": 0.5849881960838772,
+      "grad_norm": 3.1606081153851804,
+      "learning_rate": 3.876748280114984e-06,
+      "loss": 0.4448,
+      "step": 8425
+    },
+    {
+      "epoch": 0.5850576308845994,
+      "grad_norm": 3.3479964256189514,
+      "learning_rate": 3.875652563743151e-06,
+      "loss": 0.3126,
+      "step": 8426
+    },
+    {
+      "epoch": 0.5851270656853215,
+      "grad_norm": 3.2693408452751327,
+      "learning_rate": 3.874556904239538e-06,
+      "loss": 0.2758,
+      "step": 8427
+    },
+    {
+      "epoch": 0.5851965004860437,
+      "grad_norm": 4.942275814622401,
+      "learning_rate": 3.873461301659562e-06,
+      "loss": 0.4931,
+      "step": 8428
+    },
+    {
+      "epoch": 0.5852659352867657,
+      "grad_norm": 3.7595652371294213,
+      "learning_rate": 3.872365756058639e-06,
+      "loss": 0.4454,
+      "step": 8429
+    },
+    {
+      "epoch": 0.5853353700874878,
+      "grad_norm": 4.166533526352181,
+      "learning_rate": 3.8712702674921815e-06,
+      "loss": 0.6004,
+      "step": 8430
+    },
+    {
+      "epoch": 0.58540480488821,
+      "grad_norm": 4.791549355396392,
+      "learning_rate": 3.870174836015596e-06,
+      "loss": 0.5471,
+      "step": 8431
+    },
+    {
+      "epoch": 0.5854742396889321,
+      "grad_norm": 3.252195116406542,
+      "learning_rate": 3.869079461684287e-06,
+      "loss": 0.498,
+      "step": 8432
+    },
+    {
+      "epoch": 0.5855436744896542,
+      "grad_norm": 3.5098078464936293,
+      "learning_rate": 3.867984144553662e-06,
+      "loss": 0.3072,
+      "step": 8433
+    },
+    {
+      "epoch": 0.5856131092903764,
+      "grad_norm": 3.8295261714452975,
+      "learning_rate": 3.866888884679117e-06,
+      "loss": 0.3309,
+      "step": 8434
+    },
+    {
+      "epoch": 0.5856825440910984,
+      "grad_norm": 3.9296071218312236,
+      "learning_rate": 3.8657936821160495e-06,
+      "loss": 0.5009,
+      "step": 8435
+    },
+    {
+      "epoch": 0.5857519788918206,
+      "grad_norm": 3.5504965955494425,
+      "learning_rate": 3.864698536919854e-06,
+      "loss": 0.3203,
+      "step": 8436
+    },
+    {
+      "epoch": 0.5858214136925427,
+      "grad_norm": 2.5504847698761615,
+      "learning_rate": 3.863603449145923e-06,
+      "loss": 0.2454,
+      "step": 8437
+    },
+    {
+      "epoch": 0.5858908484932648,
+      "grad_norm": 4.111766292954278,
+      "learning_rate": 3.862508418849642e-06,
+      "loss": 0.4455,
+      "step": 8438
+    },
+    {
+      "epoch": 0.585960283293987,
+      "grad_norm": 3.6683038046789465,
+      "learning_rate": 3.861413446086398e-06,
+      "loss": 0.3525,
+      "step": 8439
+    },
+    {
+      "epoch": 0.586029718094709,
+      "grad_norm": 3.608241214120669,
+      "learning_rate": 3.8603185309115745e-06,
+      "loss": 0.4096,
+      "step": 8440
+    },
+    {
+      "epoch": 0.5860991528954312,
+      "grad_norm": 2.252712947091438,
+      "learning_rate": 3.859223673380551e-06,
+      "loss": 0.2396,
+      "step": 8441
+    },
+    {
+      "epoch": 0.5861685876961533,
+      "grad_norm": 3.005488730931568,
+      "learning_rate": 3.858128873548701e-06,
+      "loss": 0.2475,
+      "step": 8442
+    },
+    {
+      "epoch": 0.5862380224968754,
+      "grad_norm": 3.8407954942218097,
+      "learning_rate": 3.8570341314714025e-06,
+      "loss": 0.4409,
+      "step": 8443
+    },
+    {
+      "epoch": 0.5863074572975976,
+      "grad_norm": 2.495524239940476,
+      "learning_rate": 3.855939447204024e-06,
+      "loss": 0.1602,
+      "step": 8444
+    },
+    {
+      "epoch": 0.5863768920983197,
+      "grad_norm": 2.7912954588430563,
+      "learning_rate": 3.854844820801932e-06,
+      "loss": 0.3244,
+      "step": 8445
+    },
+    {
+      "epoch": 0.5864463268990417,
+      "grad_norm": 4.950443307519895,
+      "learning_rate": 3.853750252320494e-06,
+      "loss": 0.6244,
+      "step": 8446
+    },
+    {
+      "epoch": 0.5865157616997639,
+      "grad_norm": 4.754334454270179,
+      "learning_rate": 3.852655741815071e-06,
+      "loss": 0.5453,
+      "step": 8447
+    },
+    {
+      "epoch": 0.586585196500486,
+      "grad_norm": 4.2891910884307265,
+      "learning_rate": 3.851561289341023e-06,
+      "loss": 0.4327,
+      "step": 8448
+    },
+    {
+      "epoch": 0.5866546313012082,
+      "grad_norm": 2.4214264081507855,
+      "learning_rate": 3.850466894953704e-06,
+      "loss": 0.2041,
+      "step": 8449
+    },
+    {
+      "epoch": 0.5867240661019303,
+      "grad_norm": 3.0842758703989257,
+      "learning_rate": 3.84937255870847e-06,
+      "loss": 0.2756,
+      "step": 8450
+    },
+    {
+      "epoch": 0.5867935009026524,
+      "grad_norm": 2.002186032555324,
+      "learning_rate": 3.848278280660671e-06,
+      "loss": 0.1039,
+      "step": 8451
+    },
+    {
+      "epoch": 0.5868629357033746,
+      "grad_norm": 4.1827634175033195,
+      "learning_rate": 3.847184060865651e-06,
+      "loss": 0.4986,
+      "step": 8452
+    },
+    {
+      "epoch": 0.5869323705040966,
+      "grad_norm": 4.051641478625084,
+      "learning_rate": 3.846089899378758e-06,
+      "loss": 0.3005,
+      "step": 8453
+    },
+    {
+      "epoch": 0.5870018053048188,
+      "grad_norm": 4.689006398047894,
+      "learning_rate": 3.8449957962553305e-06,
+      "loss": 0.6752,
+      "step": 8454
+    },
+    {
+      "epoch": 0.5870712401055409,
+      "grad_norm": 3.95345387433119,
+      "learning_rate": 3.843901751550712e-06,
+      "loss": 0.5682,
+      "step": 8455
+    },
+    {
+      "epoch": 0.587140674906263,
+      "grad_norm": 4.860830166962689,
+      "learning_rate": 3.842807765320233e-06,
+      "loss": 0.5492,
+      "step": 8456
+    },
+    {
+      "epoch": 0.5872101097069852,
+      "grad_norm": 2.8187573834130037,
+      "learning_rate": 3.841713837619225e-06,
+      "loss": 0.4363,
+      "step": 8457
+    },
+    {
+      "epoch": 0.5872795445077073,
+      "grad_norm": 4.741577810534145,
+      "learning_rate": 3.840619968503025e-06,
+      "loss": 0.4092,
+      "step": 8458
+    },
+    {
+      "epoch": 0.5873489793084293,
+      "grad_norm": 3.873465295308872,
+      "learning_rate": 3.839526158026951e-06,
+      "loss": 0.2452,
+      "step": 8459
+    },
+    {
+      "epoch": 0.5874184141091515,
+      "grad_norm": 3.1930337554913097,
+      "learning_rate": 3.838432406246333e-06,
+      "loss": 0.3823,
+      "step": 8460
+    },
+    {
+      "epoch": 0.5874878489098736,
+      "grad_norm": 4.108479806624415,
+      "learning_rate": 3.837338713216489e-06,
+      "loss": 0.4078,
+      "step": 8461
+    },
+    {
+      "epoch": 0.5875572837105958,
+      "grad_norm": 3.813778921813502,
+      "learning_rate": 3.836245078992738e-06,
+      "loss": 0.5565,
+      "step": 8462
+    },
+    {
+      "epoch": 0.5876267185113179,
+      "grad_norm": 4.279888073831197,
+      "learning_rate": 3.835151503630393e-06,
+      "loss": 0.6473,
+      "step": 8463
+    },
+    {
+      "epoch": 0.58769615331204,
+      "grad_norm": 3.7070688370117675,
+      "learning_rate": 3.834057987184766e-06,
+      "loss": 0.3695,
+      "step": 8464
+    },
+    {
+      "epoch": 0.5877655881127621,
+      "grad_norm": 3.9476571650736454,
+      "learning_rate": 3.83296452971117e-06,
+      "loss": 0.5541,
+      "step": 8465
+    },
+    {
+      "epoch": 0.5878350229134842,
+      "grad_norm": 5.015061827729214,
+      "learning_rate": 3.831871131264905e-06,
+      "loss": 0.6399,
+      "step": 8466
+    },
+    {
+      "epoch": 0.5879044577142064,
+      "grad_norm": 4.876442935476198,
+      "learning_rate": 3.830777791901276e-06,
+      "loss": 0.3994,
+      "step": 8467
+    },
+    {
+      "epoch": 0.5879738925149285,
+      "grad_norm": 3.748252251791791,
+      "learning_rate": 3.8296845116755835e-06,
+      "loss": 0.506,
+      "step": 8468
+    },
+    {
+      "epoch": 0.5880433273156506,
+      "grad_norm": 4.075130910895806,
+      "learning_rate": 3.8285912906431255e-06,
+      "loss": 0.3991,
+      "step": 8469
+    },
+    {
+      "epoch": 0.5881127621163728,
+      "grad_norm": 3.7209239906883975,
+      "learning_rate": 3.827498128859193e-06,
+      "loss": 0.3816,
+      "step": 8470
+    },
+    {
+      "epoch": 0.5881821969170948,
+      "grad_norm": 4.733018665176041,
+      "learning_rate": 3.826405026379078e-06,
+      "loss": 0.4629,
+      "step": 8471
+    },
+    {
+      "epoch": 0.588251631717817,
+      "grad_norm": 4.195344191506253,
+      "learning_rate": 3.82531198325807e-06,
+      "loss": 0.6005,
+      "step": 8472
+    },
+    {
+      "epoch": 0.5883210665185391,
+      "grad_norm": 3.977346420240632,
+      "learning_rate": 3.824218999551454e-06,
+      "loss": 0.5097,
+      "step": 8473
+    },
+    {
+      "epoch": 0.5883905013192612,
+      "grad_norm": 3.8309195728138135,
+      "learning_rate": 3.823126075314508e-06,
+      "loss": 0.4314,
+      "step": 8474
+    },
+    {
+      "epoch": 0.5884599361199834,
+      "grad_norm": 3.9785861384914085,
+      "learning_rate": 3.822033210602514e-06,
+      "loss": 0.4829,
+      "step": 8475
+    },
+    {
+      "epoch": 0.5885293709207055,
+      "grad_norm": 3.696269388990146,
+      "learning_rate": 3.820940405470749e-06,
+      "loss": 0.3352,
+      "step": 8476
+    },
+    {
+      "epoch": 0.5885988057214275,
+      "grad_norm": 2.9841407377205145,
+      "learning_rate": 3.819847659974481e-06,
+      "loss": 0.3525,
+      "step": 8477
+    },
+    {
+      "epoch": 0.5886682405221497,
+      "grad_norm": 4.15582494760309,
+      "learning_rate": 3.818754974168985e-06,
+      "loss": 0.5331,
+      "step": 8478
+    },
+    {
+      "epoch": 0.5887376753228718,
+      "grad_norm": 3.0474825899216573,
+      "learning_rate": 3.817662348109524e-06,
+      "loss": 0.3317,
+      "step": 8479
+    },
+    {
+      "epoch": 0.588807110123594,
+      "grad_norm": 4.288021191793315,
+      "learning_rate": 3.816569781851366e-06,
+      "loss": 0.5627,
+      "step": 8480
+    },
+    {
+      "epoch": 0.5888765449243161,
+      "grad_norm": 4.980832700153119,
+      "learning_rate": 3.815477275449767e-06,
+      "loss": 0.7722,
+      "step": 8481
+    },
+    {
+      "epoch": 0.5889459797250381,
+      "grad_norm": 3.859889456936487,
+      "learning_rate": 3.8143848289599885e-06,
+      "loss": 0.446,
+      "step": 8482
+    },
+    {
+      "epoch": 0.5890154145257603,
+      "grad_norm": 3.6758535871594593,
+      "learning_rate": 3.813292442437284e-06,
+      "loss": 0.3974,
+      "step": 8483
+    },
+    {
+      "epoch": 0.5890848493264824,
+      "grad_norm": 2.975107543897671,
+      "learning_rate": 3.8122001159369038e-06,
+      "loss": 0.3408,
+      "step": 8484
+    },
+    {
+      "epoch": 0.5891542841272046,
+      "grad_norm": 4.080592256913607,
+      "learning_rate": 3.811107849514098e-06,
+      "loss": 0.6018,
+      "step": 8485
+    },
+    {
+      "epoch": 0.5892237189279267,
+      "grad_norm": 4.191330199340877,
+      "learning_rate": 3.810015643224111e-06,
+      "loss": 0.4093,
+      "step": 8486
+    },
+    {
+      "epoch": 0.5892931537286488,
+      "grad_norm": 3.9594657119985364,
+      "learning_rate": 3.8089234971221884e-06,
+      "loss": 0.3774,
+      "step": 8487
+    },
+    {
+      "epoch": 0.589362588529371,
+      "grad_norm": 4.767078790143186,
+      "learning_rate": 3.807831411263566e-06,
+      "loss": 0.7029,
+      "step": 8488
+    },
+    {
+      "epoch": 0.589432023330093,
+      "grad_norm": 3.327620621157235,
+      "learning_rate": 3.806739385703481e-06,
+      "loss": 0.4763,
+      "step": 8489
+    },
+    {
+      "epoch": 0.5895014581308151,
+      "grad_norm": 5.143335390960936,
+      "learning_rate": 3.80564742049717e-06,
+      "loss": 0.7385,
+      "step": 8490
+    },
+    {
+      "epoch": 0.5895708929315373,
+      "grad_norm": 4.276557970754809,
+      "learning_rate": 3.804555515699859e-06,
+      "loss": 0.7023,
+      "step": 8491
+    },
+    {
+      "epoch": 0.5896403277322594,
+      "grad_norm": 6.804498934939847,
+      "learning_rate": 3.8034636713667777e-06,
+      "loss": 0.5053,
+      "step": 8492
+    },
+    {
+      "epoch": 0.5897097625329816,
+      "grad_norm": 4.7116579864499215,
+      "learning_rate": 3.802371887553149e-06,
+      "loss": 0.5227,
+      "step": 8493
+    },
+    {
+      "epoch": 0.5897791973337037,
+      "grad_norm": 3.368677250784847,
+      "learning_rate": 3.801280164314197e-06,
+      "loss": 0.4151,
+      "step": 8494
+    },
+    {
+      "epoch": 0.5898486321344257,
+      "grad_norm": 2.2769359342258837,
+      "learning_rate": 3.8001885017051377e-06,
+      "loss": 0.1154,
+      "step": 8495
+    },
+    {
+      "epoch": 0.5899180669351479,
+      "grad_norm": 3.5989338637236257,
+      "learning_rate": 3.7990968997811847e-06,
+      "loss": 0.3009,
+      "step": 8496
+    },
+    {
+      "epoch": 0.58998750173587,
+      "grad_norm": 4.838549550301688,
+      "learning_rate": 3.7980053585975526e-06,
+      "loss": 0.4369,
+      "step": 8497
+    },
+    {
+      "epoch": 0.5900569365365922,
+      "grad_norm": 3.0852712330663654,
+      "learning_rate": 3.79691387820945e-06,
+      "loss": 0.3329,
+      "step": 8498
+    },
+    {
+      "epoch": 0.5901263713373143,
+      "grad_norm": 3.740370702252939,
+      "learning_rate": 3.7958224586720805e-06,
+      "loss": 0.4645,
+      "step": 8499
+    },
+    {
+      "epoch": 0.5901958061380363,
+      "grad_norm": 3.5197247511630367,
+      "learning_rate": 3.7947311000406483e-06,
+      "loss": 0.3749,
+      "step": 8500
+    },
+    {
+      "epoch": 0.5902652409387585,
+      "grad_norm": 3.441555325258468,
+      "learning_rate": 3.793639802370355e-06,
+      "loss": 0.4317,
+      "step": 8501
+    },
+    {
+      "epoch": 0.5903346757394806,
+      "grad_norm": 3.435896582663432,
+      "learning_rate": 3.7925485657163946e-06,
+      "loss": 0.2947,
+      "step": 8502
+    },
+    {
+      "epoch": 0.5904041105402027,
+      "grad_norm": 3.8504637330436027,
+      "learning_rate": 3.7914573901339602e-06,
+      "loss": 0.5698,
+      "step": 8503
+    },
+    {
+      "epoch": 0.5904735453409249,
+      "grad_norm": 4.944167743574787,
+      "learning_rate": 3.790366275678245e-06,
+      "loss": 0.632,
+      "step": 8504
+    },
+    {
+      "epoch": 0.590542980141647,
+      "grad_norm": 3.126626931646617,
+      "learning_rate": 3.7892752224044353e-06,
+      "loss": 0.287,
+      "step": 8505
+    },
+    {
+      "epoch": 0.5906124149423692,
+      "grad_norm": 3.1152108949424897,
+      "learning_rate": 3.788184230367713e-06,
+      "loss": 0.3999,
+      "step": 8506
+    },
+    {
+      "epoch": 0.5906818497430912,
+      "grad_norm": 3.534621815375558,
+      "learning_rate": 3.787093299623262e-06,
+      "loss": 0.3545,
+      "step": 8507
+    },
+    {
+      "epoch": 0.5907512845438133,
+      "grad_norm": 3.81013552083832,
+      "learning_rate": 3.7860024302262603e-06,
+      "loss": 0.3806,
+      "step": 8508
+    },
+    {
+      "epoch": 0.5908207193445355,
+      "grad_norm": 4.033634866429512,
+      "learning_rate": 3.7849116222318803e-06,
+      "loss": 0.4994,
+      "step": 8509
+    },
+    {
+      "epoch": 0.5908901541452576,
+      "grad_norm": 5.071912557673434,
+      "learning_rate": 3.7838208756952956e-06,
+      "loss": 0.517,
+      "step": 8510
+    },
+    {
+      "epoch": 0.5909595889459798,
+      "grad_norm": 3.6541344130061715,
+      "learning_rate": 3.782730190671675e-06,
+      "loss": 0.2126,
+      "step": 8511
+    },
+    {
+      "epoch": 0.5910290237467019,
+      "grad_norm": 4.402263908175208,
+      "learning_rate": 3.7816395672161852e-06,
+      "loss": 0.5274,
+      "step": 8512
+    },
+    {
+      "epoch": 0.5910984585474239,
+      "grad_norm": 4.79185426522873,
+      "learning_rate": 3.780549005383986e-06,
+      "loss": 0.6416,
+      "step": 8513
+    },
+    {
+      "epoch": 0.5911678933481461,
+      "grad_norm": 3.7500921710261137,
+      "learning_rate": 3.779458505230239e-06,
+      "loss": 0.5415,
+      "step": 8514
+    },
+    {
+      "epoch": 0.5912373281488682,
+      "grad_norm": 3.2405816830650713,
+      "learning_rate": 3.7783680668101003e-06,
+      "loss": 0.3399,
+      "step": 8515
+    },
+    {
+      "epoch": 0.5913067629495903,
+      "grad_norm": 4.04246279272713,
+      "learning_rate": 3.7772776901787206e-06,
+      "loss": 0.689,
+      "step": 8516
+    },
+    {
+      "epoch": 0.5913761977503125,
+      "grad_norm": 3.520377505867746,
+      "learning_rate": 3.7761873753912526e-06,
+      "loss": 0.4181,
+      "step": 8517
+    },
+    {
+      "epoch": 0.5914456325510346,
+      "grad_norm": 6.778310251607479,
+      "learning_rate": 3.7750971225028415e-06,
+      "loss": 0.411,
+      "step": 8518
+    },
+    {
+      "epoch": 0.5915150673517567,
+      "grad_norm": 4.107538884274487,
+      "learning_rate": 3.774006931568634e-06,
+      "loss": 0.4638,
+      "step": 8519
+    },
+    {
+      "epoch": 0.5915845021524788,
+      "grad_norm": 4.386435497473944,
+      "learning_rate": 3.7729168026437663e-06,
+      "loss": 0.4391,
+      "step": 8520
+    },
+    {
+      "epoch": 0.5916539369532009,
+      "grad_norm": 3.218257816738442,
+      "learning_rate": 3.7718267357833794e-06,
+      "loss": 0.334,
+      "step": 8521
+    },
+    {
+      "epoch": 0.5917233717539231,
+      "grad_norm": 4.273498719851935,
+      "learning_rate": 3.770736731042605e-06,
+      "loss": 0.5073,
+      "step": 8522
+    },
+    {
+      "epoch": 0.5917928065546452,
+      "grad_norm": 3.2483792076016558,
+      "learning_rate": 3.769646788476578e-06,
+      "loss": 0.245,
+      "step": 8523
+    },
+    {
+      "epoch": 0.5918622413553674,
+      "grad_norm": 2.726052675213289,
+      "learning_rate": 3.768556908140424e-06,
+      "loss": 0.2128,
+      "step": 8524
+    },
+    {
+      "epoch": 0.5919316761560894,
+      "grad_norm": 4.011885320113991,
+      "learning_rate": 3.7674670900892656e-06,
+      "loss": 0.4947,
+      "step": 8525
+    },
+    {
+      "epoch": 0.5920011109568115,
+      "grad_norm": 3.79163239049805,
+      "learning_rate": 3.76637733437823e-06,
+      "loss": 0.434,
+      "step": 8526
+    },
+    {
+      "epoch": 0.5920705457575337,
+      "grad_norm": 4.100593410796163,
+      "learning_rate": 3.7652876410624316e-06,
+      "loss": 0.6677,
+      "step": 8527
+    },
+    {
+      "epoch": 0.5921399805582558,
+      "grad_norm": 3.894478103531174,
+      "learning_rate": 3.7641980101969865e-06,
+      "loss": 0.3234,
+      "step": 8528
+    },
+    {
+      "epoch": 0.5922094153589779,
+      "grad_norm": 4.279489061700669,
+      "learning_rate": 3.7631084418370075e-06,
+      "loss": 0.515,
+      "step": 8529
+    },
+    {
+      "epoch": 0.5922788501597,
+      "grad_norm": 2.749537735965704,
+      "learning_rate": 3.762018936037606e-06,
+      "loss": 0.2081,
+      "step": 8530
+    },
+    {
+      "epoch": 0.5923482849604221,
+      "grad_norm": 4.787935202757789,
+      "learning_rate": 3.7609294928538843e-06,
+      "loss": 0.6375,
+      "step": 8531
+    },
+    {
+      "epoch": 0.5924177197611443,
+      "grad_norm": 4.510729870770608,
+      "learning_rate": 3.7598401123409462e-06,
+      "loss": 0.5762,
+      "step": 8532
+    },
+    {
+      "epoch": 0.5924871545618664,
+      "grad_norm": 4.107892817168008,
+      "learning_rate": 3.758750794553894e-06,
+      "loss": 0.6634,
+      "step": 8533
+    },
+    {
+      "epoch": 0.5925565893625885,
+      "grad_norm": 4.098512660753035,
+      "learning_rate": 3.7576615395478206e-06,
+      "loss": 0.4491,
+      "step": 8534
+    },
+    {
+      "epoch": 0.5926260241633107,
+      "grad_norm": 4.49218611839039,
+      "learning_rate": 3.75657234737782e-06,
+      "loss": 0.4436,
+      "step": 8535
+    },
+    {
+      "epoch": 0.5926954589640328,
+      "grad_norm": 3.5015570819070736,
+      "learning_rate": 3.755483218098984e-06,
+      "loss": 0.3831,
+      "step": 8536
+    },
+    {
+      "epoch": 0.5927648937647549,
+      "grad_norm": 3.0685222467576274,
+      "learning_rate": 3.7543941517663997e-06,
+      "loss": 0.2457,
+      "step": 8537
+    },
+    {
+      "epoch": 0.592834328565477,
+      "grad_norm": 3.03573312759437,
+      "learning_rate": 3.7533051484351473e-06,
+      "loss": 0.4012,
+      "step": 8538
+    },
+    {
+      "epoch": 0.5929037633661991,
+      "grad_norm": 4.2552610664733885,
+      "learning_rate": 3.7522162081603096e-06,
+      "loss": 0.6107,
+      "step": 8539
+    },
+    {
+      "epoch": 0.5929731981669213,
+      "grad_norm": 4.033852957496691,
+      "learning_rate": 3.751127330996967e-06,
+      "loss": 0.5426,
+      "step": 8540
+    },
+    {
+      "epoch": 0.5930426329676434,
+      "grad_norm": 3.962614328504431,
+      "learning_rate": 3.7500385170001897e-06,
+      "loss": 0.4882,
+      "step": 8541
+    },
+    {
+      "epoch": 0.5931120677683656,
+      "grad_norm": 13.860643685321643,
+      "learning_rate": 3.7489497662250483e-06,
+      "loss": 0.4452,
+      "step": 8542
+    },
+    {
+      "epoch": 0.5931815025690876,
+      "grad_norm": 4.14015735607899,
+      "learning_rate": 3.7478610787266134e-06,
+      "loss": 0.4337,
+      "step": 8543
+    },
+    {
+      "epoch": 0.5932509373698097,
+      "grad_norm": 3.5139651110165206,
+      "learning_rate": 3.7467724545599486e-06,
+      "loss": 0.4353,
+      "step": 8544
+    },
+    {
+      "epoch": 0.5933203721705319,
+      "grad_norm": 3.587939671686392,
+      "learning_rate": 3.745683893780113e-06,
+      "loss": 0.2525,
+      "step": 8545
+    },
+    {
+      "epoch": 0.593389806971254,
+      "grad_norm": 4.5587159138361315,
+      "learning_rate": 3.744595396442169e-06,
+      "loss": 0.4313,
+      "step": 8546
+    },
+    {
+      "epoch": 0.5934592417719761,
+      "grad_norm": 4.723224048193733,
+      "learning_rate": 3.7435069626011667e-06,
+      "loss": 0.5722,
+      "step": 8547
+    },
+    {
+      "epoch": 0.5935286765726983,
+      "grad_norm": 4.433577581060515,
+      "learning_rate": 3.7424185923121635e-06,
+      "loss": 0.4098,
+      "step": 8548
+    },
+    {
+      "epoch": 0.5935981113734203,
+      "grad_norm": 4.083520708320628,
+      "learning_rate": 3.741330285630202e-06,
+      "loss": 0.5824,
+      "step": 8549
+    },
+    {
+      "epoch": 0.5936675461741425,
+      "grad_norm": 3.636996752293765,
+      "learning_rate": 3.740242042610332e-06,
+      "loss": 0.4044,
+      "step": 8550
+    },
+    {
+      "epoch": 0.5937369809748646,
+      "grad_norm": 4.958901777944066,
+      "learning_rate": 3.7391538633075947e-06,
+      "loss": 0.4734,
+      "step": 8551
+    },
+    {
+      "epoch": 0.5938064157755867,
+      "grad_norm": 3.997282738181894,
+      "learning_rate": 3.7380657477770267e-06,
+      "loss": 0.3476,
+      "step": 8552
+    },
+    {
+      "epoch": 0.5938758505763089,
+      "grad_norm": 3.300794959287424,
+      "learning_rate": 3.736977696073665e-06,
+      "loss": 0.4325,
+      "step": 8553
+    },
+    {
+      "epoch": 0.593945285377031,
+      "grad_norm": 3.165498133305783,
+      "learning_rate": 3.7358897082525425e-06,
+      "loss": 0.2107,
+      "step": 8554
+    },
+    {
+      "epoch": 0.5940147201777531,
+      "grad_norm": 3.6111078339872043,
+      "learning_rate": 3.7348017843686896e-06,
+      "loss": 0.3428,
+      "step": 8555
+    },
+    {
+      "epoch": 0.5940841549784752,
+      "grad_norm": 4.250980776360398,
+      "learning_rate": 3.73371392447713e-06,
+      "loss": 0.4566,
+      "step": 8556
+    },
+    {
+      "epoch": 0.5941535897791973,
+      "grad_norm": 6.508266462831398,
+      "learning_rate": 3.7326261286328856e-06,
+      "loss": 0.5029,
+      "step": 8557
+    },
+    {
+      "epoch": 0.5942230245799195,
+      "grad_norm": 4.712878495269399,
+      "learning_rate": 3.73153839689098e-06,
+      "loss": 0.7578,
+      "step": 8558
+    },
+    {
+      "epoch": 0.5942924593806416,
+      "grad_norm": 3.6910113551378876,
+      "learning_rate": 3.7304507293064265e-06,
+      "loss": 0.3324,
+      "step": 8559
+    },
+    {
+      "epoch": 0.5943618941813636,
+      "grad_norm": 2.5879518247745446,
+      "learning_rate": 3.729363125934236e-06,
+      "loss": 0.2129,
+      "step": 8560
+    },
+    {
+      "epoch": 0.5944313289820858,
+      "grad_norm": 4.48830414233212,
+      "learning_rate": 3.728275586829422e-06,
+      "loss": 0.57,
+      "step": 8561
+    },
+    {
+      "epoch": 0.5945007637828079,
+      "grad_norm": 3.9150188324557775,
+      "learning_rate": 3.727188112046991e-06,
+      "loss": 0.4332,
+      "step": 8562
+    },
+    {
+      "epoch": 0.5945701985835301,
+      "grad_norm": 3.5938955720013115,
+      "learning_rate": 3.7261007016419448e-06,
+      "loss": 0.3939,
+      "step": 8563
+    },
+    {
+      "epoch": 0.5946396333842522,
+      "grad_norm": 5.252194547524712,
+      "learning_rate": 3.725013355669282e-06,
+      "loss": 0.4298,
+      "step": 8564
+    },
+    {
+      "epoch": 0.5947090681849743,
+      "grad_norm": 3.7305220747712746,
+      "learning_rate": 3.7239260741840035e-06,
+      "loss": 0.4066,
+      "step": 8565
+    },
+    {
+      "epoch": 0.5947785029856965,
+      "grad_norm": 4.869072208500784,
+      "learning_rate": 3.7228388572410986e-06,
+      "loss": 0.6933,
+      "step": 8566
+    },
+    {
+      "epoch": 0.5948479377864185,
+      "grad_norm": 3.0820242862668015,
+      "learning_rate": 3.7217517048955576e-06,
+      "loss": 0.2294,
+      "step": 8567
+    },
+    {
+      "epoch": 0.5949173725871407,
+      "grad_norm": 3.3258705131241473,
+      "learning_rate": 3.7206646172023707e-06,
+      "loss": 0.3889,
+      "step": 8568
+    },
+    {
+      "epoch": 0.5949868073878628,
+      "grad_norm": 3.10738452228029,
+      "learning_rate": 3.7195775942165213e-06,
+      "loss": 0.2476,
+      "step": 8569
+    },
+    {
+      "epoch": 0.5950562421885849,
+      "grad_norm": 4.5507049766775065,
+      "learning_rate": 3.7184906359929853e-06,
+      "loss": 0.5887,
+      "step": 8570
+    },
+    {
+      "epoch": 0.5951256769893071,
+      "grad_norm": 4.562776831312148,
+      "learning_rate": 3.717403742586743e-06,
+      "loss": 0.5362,
+      "step": 8571
+    },
+    {
+      "epoch": 0.5951951117900292,
+      "grad_norm": 3.5971891373729483,
+      "learning_rate": 3.7163169140527694e-06,
+      "loss": 0.3734,
+      "step": 8572
+    },
+    {
+      "epoch": 0.5952645465907512,
+      "grad_norm": 3.618300196006156,
+      "learning_rate": 3.7152301504460348e-06,
+      "loss": 0.496,
+      "step": 8573
+    },
+    {
+      "epoch": 0.5953339813914734,
+      "grad_norm": 3.623724703184307,
+      "learning_rate": 3.714143451821503e-06,
+      "loss": 0.2985,
+      "step": 8574
+    },
+    {
+      "epoch": 0.5954034161921955,
+      "grad_norm": 3.4334741352813283,
+      "learning_rate": 3.7130568182341414e-06,
+      "loss": 0.3509,
+      "step": 8575
+    },
+    {
+      "epoch": 0.5954728509929177,
+      "grad_norm": 4.283910729230446,
+      "learning_rate": 3.711970249738911e-06,
+      "loss": 0.6051,
+      "step": 8576
+    },
+    {
+      "epoch": 0.5955422857936398,
+      "grad_norm": 3.9642111495467622,
+      "learning_rate": 3.7108837463907656e-06,
+      "loss": 0.5265,
+      "step": 8577
+    },
+    {
+      "epoch": 0.5956117205943618,
+      "grad_norm": 3.452648313979229,
+      "learning_rate": 3.7097973082446625e-06,
+      "loss": 0.3549,
+      "step": 8578
+    },
+    {
+      "epoch": 0.595681155395084,
+      "grad_norm": 2.7891606040482464,
+      "learning_rate": 3.7087109353555507e-06,
+      "loss": 0.2149,
+      "step": 8579
+    },
+    {
+      "epoch": 0.5957505901958061,
+      "grad_norm": 4.167676805754961,
+      "learning_rate": 3.7076246277783805e-06,
+      "loss": 0.4808,
+      "step": 8580
+    },
+    {
+      "epoch": 0.5958200249965283,
+      "grad_norm": 1.9841120185185133,
+      "learning_rate": 3.7065383855680925e-06,
+      "loss": 0.1229,
+      "step": 8581
+    },
+    {
+      "epoch": 0.5958894597972504,
+      "grad_norm": 4.108624975799672,
+      "learning_rate": 3.7054522087796297e-06,
+      "loss": 0.3976,
+      "step": 8582
+    },
+    {
+      "epoch": 0.5959588945979725,
+      "grad_norm": 2.9633584808048736,
+      "learning_rate": 3.704366097467931e-06,
+      "loss": 0.3312,
+      "step": 8583
+    },
+    {
+      "epoch": 0.5960283293986947,
+      "grad_norm": 2.74523931436068,
+      "learning_rate": 3.703280051687926e-06,
+      "loss": 0.3181,
+      "step": 8584
+    },
+    {
+      "epoch": 0.5960977641994167,
+      "grad_norm": 3.925660047105778,
+      "learning_rate": 3.702194071494551e-06,
+      "loss": 0.4,
+      "step": 8585
+    },
+    {
+      "epoch": 0.5961671990001388,
+      "grad_norm": 3.1899026899110092,
+      "learning_rate": 3.7011081569427294e-06,
+      "loss": 0.3294,
+      "step": 8586
+    },
+    {
+      "epoch": 0.596236633800861,
+      "grad_norm": 3.526590953350226,
+      "learning_rate": 3.70002230808739e-06,
+      "loss": 0.4775,
+      "step": 8587
+    },
+    {
+      "epoch": 0.5963060686015831,
+      "grad_norm": 3.1109440167875366,
+      "learning_rate": 3.6989365249834495e-06,
+      "loss": 0.3854,
+      "step": 8588
+    },
+    {
+      "epoch": 0.5963755034023053,
+      "grad_norm": 3.8144028837813106,
+      "learning_rate": 3.6978508076858266e-06,
+      "loss": 0.4626,
+      "step": 8589
+    },
+    {
+      "epoch": 0.5964449382030274,
+      "grad_norm": 3.738120378184053,
+      "learning_rate": 3.696765156249439e-06,
+      "loss": 0.3759,
+      "step": 8590
+    },
+    {
+      "epoch": 0.5965143730037494,
+      "grad_norm": 6.135987536518666,
+      "learning_rate": 3.695679570729192e-06,
+      "loss": 0.5693,
+      "step": 8591
+    },
+    {
+      "epoch": 0.5965838078044716,
+      "grad_norm": 3.768880593779302,
+      "learning_rate": 3.694594051179998e-06,
+      "loss": 0.3858,
+      "step": 8592
+    },
+    {
+      "epoch": 0.5966532426051937,
+      "grad_norm": 3.1042479796598417,
+      "learning_rate": 3.6935085976567585e-06,
+      "loss": 0.3175,
+      "step": 8593
+    },
+    {
+      "epoch": 0.5967226774059159,
+      "grad_norm": 3.0095981842626243,
+      "learning_rate": 3.6924232102143782e-06,
+      "loss": 0.2138,
+      "step": 8594
+    },
+    {
+      "epoch": 0.596792112206638,
+      "grad_norm": 3.6272129571237284,
+      "learning_rate": 3.691337888907751e-06,
+      "loss": 0.4748,
+      "step": 8595
+    },
+    {
+      "epoch": 0.59686154700736,
+      "grad_norm": 4.631128190662742,
+      "learning_rate": 3.690252633791771e-06,
+      "loss": 0.648,
+      "step": 8596
+    },
+    {
+      "epoch": 0.5969309818080822,
+      "grad_norm": 2.165215058226834,
+      "learning_rate": 3.689167444921332e-06,
+      "loss": 0.135,
+      "step": 8597
+    },
+    {
+      "epoch": 0.5970004166088043,
+      "grad_norm": 3.243582544786251,
+      "learning_rate": 3.6880823223513217e-06,
+      "loss": 0.383,
+      "step": 8598
+    },
+    {
+      "epoch": 0.5970698514095265,
+      "grad_norm": 2.9742764972041877,
+      "learning_rate": 3.6869972661366214e-06,
+      "loss": 0.377,
+      "step": 8599
+    },
+    {
+      "epoch": 0.5971392862102486,
+      "grad_norm": 4.191603759978091,
+      "learning_rate": 3.685912276332113e-06,
+      "loss": 0.6315,
+      "step": 8600
+    },
+    {
+      "epoch": 0.5972087210109707,
+      "grad_norm": 3.6064965100774433,
+      "learning_rate": 3.6848273529926764e-06,
+      "loss": 0.4051,
+      "step": 8601
+    },
+    {
+      "epoch": 0.5972781558116929,
+      "grad_norm": 5.332571548741159,
+      "learning_rate": 3.683742496173184e-06,
+      "loss": 0.3432,
+      "step": 8602
+    },
+    {
+      "epoch": 0.5973475906124149,
+      "grad_norm": 4.259959098088565,
+      "learning_rate": 3.6826577059285045e-06,
+      "loss": 0.3901,
+      "step": 8603
+    },
+    {
+      "epoch": 0.597417025413137,
+      "grad_norm": 5.193812928929078,
+      "learning_rate": 3.6815729823135094e-06,
+      "loss": 0.6605,
+      "step": 8604
+    },
+    {
+      "epoch": 0.5974864602138592,
+      "grad_norm": 3.977917847461278,
+      "learning_rate": 3.6804883253830614e-06,
+      "loss": 0.4712,
+      "step": 8605
+    },
+    {
+      "epoch": 0.5975558950145813,
+      "grad_norm": 4.036382772770804,
+      "learning_rate": 3.6794037351920194e-06,
+      "loss": 0.388,
+      "step": 8606
+    },
+    {
+      "epoch": 0.5976253298153035,
+      "grad_norm": 4.085163502269295,
+      "learning_rate": 3.6783192117952427e-06,
+      "loss": 0.4378,
+      "step": 8607
+    },
+    {
+      "epoch": 0.5976947646160256,
+      "grad_norm": 4.487342755027042,
+      "learning_rate": 3.6772347552475857e-06,
+      "loss": 0.664,
+      "step": 8608
+    },
+    {
+      "epoch": 0.5977641994167476,
+      "grad_norm": 4.042795182086229,
+      "learning_rate": 3.676150365603896e-06,
+      "loss": 0.5911,
+      "step": 8609
+    },
+    {
+      "epoch": 0.5978336342174698,
+      "grad_norm": 3.8372546831076444,
+      "learning_rate": 3.6750660429190227e-06,
+      "loss": 0.3467,
+      "step": 8610
+    },
+    {
+      "epoch": 0.5979030690181919,
+      "grad_norm": 3.991659105143138,
+      "learning_rate": 3.6739817872478113e-06,
+      "loss": 0.366,
+      "step": 8611
+    },
+    {
+      "epoch": 0.5979725038189141,
+      "grad_norm": 3.270534536783417,
+      "learning_rate": 3.672897598645101e-06,
+      "loss": 0.3272,
+      "step": 8612
+    },
+    {
+      "epoch": 0.5980419386196362,
+      "grad_norm": 3.206931713408161,
+      "learning_rate": 3.6718134771657265e-06,
+      "loss": 0.2669,
+      "step": 8613
+    },
+    {
+      "epoch": 0.5981113734203582,
+      "grad_norm": 4.957500645650893,
+      "learning_rate": 3.6707294228645242e-06,
+      "loss": 0.5807,
+      "step": 8614
+    },
+    {
+      "epoch": 0.5981808082210804,
+      "grad_norm": 3.2041602940830503,
+      "learning_rate": 3.669645435796325e-06,
+      "loss": 0.3509,
+      "step": 8615
+    },
+    {
+      "epoch": 0.5982502430218025,
+      "grad_norm": 3.995778250124758,
+      "learning_rate": 3.6685615160159514e-06,
+      "loss": 0.4554,
+      "step": 8616
+    },
+    {
+      "epoch": 0.5983196778225246,
+      "grad_norm": 2.4261517232059244,
+      "learning_rate": 3.6674776635782317e-06,
+      "loss": 0.2522,
+      "step": 8617
+    },
+    {
+      "epoch": 0.5983891126232468,
+      "grad_norm": 3.7473313929810304,
+      "learning_rate": 3.6663938785379826e-06,
+      "loss": 0.246,
+      "step": 8618
+    },
+    {
+      "epoch": 0.5984585474239689,
+      "grad_norm": 3.4107790283875157,
+      "learning_rate": 3.665310160950024e-06,
+      "loss": 0.445,
+      "step": 8619
+    },
+    {
+      "epoch": 0.5985279822246911,
+      "grad_norm": 3.848285208237695,
+      "learning_rate": 3.664226510869165e-06,
+      "loss": 0.2575,
+      "step": 8620
+    },
+    {
+      "epoch": 0.5985974170254131,
+      "grad_norm": 3.897876545256366,
+      "learning_rate": 3.6631429283502185e-06,
+      "loss": 0.4726,
+      "step": 8621
+    },
+    {
+      "epoch": 0.5986668518261352,
+      "grad_norm": 2.5919789344186235,
+      "learning_rate": 3.662059413447989e-06,
+      "loss": 0.2068,
+      "step": 8622
+    },
+    {
+      "epoch": 0.5987362866268574,
+      "grad_norm": 3.5179564507626537,
+      "learning_rate": 3.6609759662172826e-06,
+      "loss": 0.4282,
+      "step": 8623
+    },
+    {
+      "epoch": 0.5988057214275795,
+      "grad_norm": 3.500731537975049,
+      "learning_rate": 3.659892586712896e-06,
+      "loss": 0.3883,
+      "step": 8624
+    },
+    {
+      "epoch": 0.5988751562283017,
+      "grad_norm": 3.8957347071692476,
+      "learning_rate": 3.658809274989624e-06,
+      "loss": 0.566,
+      "step": 8625
+    },
+    {
+      "epoch": 0.5989445910290238,
+      "grad_norm": 4.20055416943044,
+      "learning_rate": 3.657726031102264e-06,
+      "loss": 0.6276,
+      "step": 8626
+    },
+    {
+      "epoch": 0.5990140258297458,
+      "grad_norm": 4.345415490468386,
+      "learning_rate": 3.656642855105601e-06,
+      "loss": 0.2989,
+      "step": 8627
+    },
+    {
+      "epoch": 0.599083460630468,
+      "grad_norm": 4.022248682785972,
+      "learning_rate": 3.6555597470544214e-06,
+      "loss": 0.3746,
+      "step": 8628
+    },
+    {
+      "epoch": 0.5991528954311901,
+      "grad_norm": 3.284370262405294,
+      "learning_rate": 3.654476707003508e-06,
+      "loss": 0.44,
+      "step": 8629
+    },
+    {
+      "epoch": 0.5992223302319122,
+      "grad_norm": 3.346547665855775,
+      "learning_rate": 3.653393735007643e-06,
+      "loss": 0.4198,
+      "step": 8630
+    },
+    {
+      "epoch": 0.5992917650326344,
+      "grad_norm": 4.529580981964243,
+      "learning_rate": 3.652310831121598e-06,
+      "loss": 0.5439,
+      "step": 8631
+    },
+    {
+      "epoch": 0.5993611998333565,
+      "grad_norm": 4.300155370920969,
+      "learning_rate": 3.651227995400145e-06,
+      "loss": 0.4082,
+      "step": 8632
+    },
+    {
+      "epoch": 0.5994306346340786,
+      "grad_norm": 4.391527118825359,
+      "learning_rate": 3.6501452278980555e-06,
+      "loss": 0.4562,
+      "step": 8633
+    },
+    {
+      "epoch": 0.5995000694348007,
+      "grad_norm": 5.864002266378577,
+      "learning_rate": 3.649062528670092e-06,
+      "loss": 0.9552,
+      "step": 8634
+    },
+    {
+      "epoch": 0.5995695042355228,
+      "grad_norm": 3.3583079605869837,
+      "learning_rate": 3.647979897771016e-06,
+      "loss": 0.365,
+      "step": 8635
+    },
+    {
+      "epoch": 0.599638939036245,
+      "grad_norm": 4.438735818610624,
+      "learning_rate": 3.646897335255588e-06,
+      "loss": 0.4548,
+      "step": 8636
+    },
+    {
+      "epoch": 0.5997083738369671,
+      "grad_norm": 5.273928347055552,
+      "learning_rate": 3.6458148411785633e-06,
+      "loss": 0.6335,
+      "step": 8637
+    },
+    {
+      "epoch": 0.5997778086376893,
+      "grad_norm": 4.10915053409455,
+      "learning_rate": 3.6447324155946883e-06,
+      "loss": 0.4588,
+      "step": 8638
+    },
+    {
+      "epoch": 0.5998472434384113,
+      "grad_norm": 4.354382010528509,
+      "learning_rate": 3.643650058558716e-06,
+      "loss": 0.6129,
+      "step": 8639
+    },
+    {
+      "epoch": 0.5999166782391334,
+      "grad_norm": 4.345784147557371,
+      "learning_rate": 3.64256777012539e-06,
+      "loss": 0.4098,
+      "step": 8640
+    },
+    {
+      "epoch": 0.5999861130398556,
+      "grad_norm": 3.707584022554522,
+      "learning_rate": 3.6414855503494473e-06,
+      "loss": 0.3188,
+      "step": 8641
+    },
+    {
+      "epoch": 0.6000555478405777,
+      "grad_norm": 4.757621519271678,
+      "learning_rate": 3.6404033992856293e-06,
+      "loss": 0.6449,
+      "step": 8642
+    },
+    {
+      "epoch": 0.6001249826412998,
+      "grad_norm": 3.880107314296162,
+      "learning_rate": 3.6393213169886686e-06,
+      "loss": 0.5856,
+      "step": 8643
+    },
+    {
+      "epoch": 0.600194417442022,
+      "grad_norm": 4.936598863948294,
+      "learning_rate": 3.638239303513298e-06,
+      "loss": 0.5337,
+      "step": 8644
+    },
+    {
+      "epoch": 0.600263852242744,
+      "grad_norm": 4.120337172898331,
+      "learning_rate": 3.63715735891424e-06,
+      "loss": 0.4638,
+      "step": 8645
+    },
+    {
+      "epoch": 0.6003332870434662,
+      "grad_norm": 2.73339366609182,
+      "learning_rate": 3.636075483246222e-06,
+      "loss": 0.1481,
+      "step": 8646
+    },
+    {
+      "epoch": 0.6004027218441883,
+      "grad_norm": 3.627615872107336,
+      "learning_rate": 3.6349936765639616e-06,
+      "loss": 0.359,
+      "step": 8647
+    },
+    {
+      "epoch": 0.6004721566449104,
+      "grad_norm": 4.320692410097562,
+      "learning_rate": 3.6339119389221783e-06,
+      "loss": 0.4608,
+      "step": 8648
+    },
+    {
+      "epoch": 0.6005415914456326,
+      "grad_norm": 3.9988797239677916,
+      "learning_rate": 3.6328302703755825e-06,
+      "loss": 0.4037,
+      "step": 8649
+    },
+    {
+      "epoch": 0.6006110262463547,
+      "grad_norm": 3.5148687749478515,
+      "learning_rate": 3.631748670978883e-06,
+      "loss": 0.3221,
+      "step": 8650
+    },
+    {
+      "epoch": 0.6006804610470768,
+      "grad_norm": 4.612059747326045,
+      "learning_rate": 3.63066714078679e-06,
+      "loss": 0.4108,
+      "step": 8651
+    },
+    {
+      "epoch": 0.6007498958477989,
+      "grad_norm": 5.027465871338378,
+      "learning_rate": 3.629585679854001e-06,
+      "loss": 0.5507,
+      "step": 8652
+    },
+    {
+      "epoch": 0.600819330648521,
+      "grad_norm": 2.38998514833837,
+      "learning_rate": 3.6285042882352193e-06,
+      "loss": 0.1307,
+      "step": 8653
+    },
+    {
+      "epoch": 0.6008887654492432,
+      "grad_norm": 4.435243456427651,
+      "learning_rate": 3.627422965985137e-06,
+      "loss": 0.2746,
+      "step": 8654
+    },
+    {
+      "epoch": 0.6009582002499653,
+      "grad_norm": 5.100130199929668,
+      "learning_rate": 3.6263417131584506e-06,
+      "loss": 0.7242,
+      "step": 8655
+    },
+    {
+      "epoch": 0.6010276350506875,
+      "grad_norm": 3.9133888363348315,
+      "learning_rate": 3.625260529809844e-06,
+      "loss": 0.4415,
+      "step": 8656
+    },
+    {
+      "epoch": 0.6010970698514095,
+      "grad_norm": 3.3109831892797015,
+      "learning_rate": 3.624179415994004e-06,
+      "loss": 0.3364,
+      "step": 8657
+    },
+    {
+      "epoch": 0.6011665046521316,
+      "grad_norm": 3.7542900437375004,
+      "learning_rate": 3.6230983717656142e-06,
+      "loss": 0.2582,
+      "step": 8658
+    },
+    {
+      "epoch": 0.6012359394528538,
+      "grad_norm": 7.929298163740489,
+      "learning_rate": 3.62201739717935e-06,
+      "loss": 0.2628,
+      "step": 8659
+    },
+    {
+      "epoch": 0.6013053742535759,
+      "grad_norm": 3.234264522405559,
+      "learning_rate": 3.620936492289885e-06,
+      "loss": 0.3657,
+      "step": 8660
+    },
+    {
+      "epoch": 0.601374809054298,
+      "grad_norm": 3.6221428714487347,
+      "learning_rate": 3.619855657151892e-06,
+      "loss": 0.6025,
+      "step": 8661
+    },
+    {
+      "epoch": 0.6014442438550202,
+      "grad_norm": 1.8947079876273223,
+      "learning_rate": 3.6187748918200393e-06,
+      "loss": 0.1474,
+      "step": 8662
+    },
+    {
+      "epoch": 0.6015136786557422,
+      "grad_norm": 4.18935700379198,
+      "learning_rate": 3.6176941963489897e-06,
+      "loss": 0.4117,
+      "step": 8663
+    },
+    {
+      "epoch": 0.6015831134564644,
+      "grad_norm": 3.379113088835297,
+      "learning_rate": 3.616613570793402e-06,
+      "loss": 0.2874,
+      "step": 8664
+    },
+    {
+      "epoch": 0.6016525482571865,
+      "grad_norm": 4.399729440285468,
+      "learning_rate": 3.6155330152079364e-06,
+      "loss": 0.7185,
+      "step": 8665
+    },
+    {
+      "epoch": 0.6017219830579086,
+      "grad_norm": 3.6623941908405744,
+      "learning_rate": 3.6144525296472432e-06,
+      "loss": 0.3593,
+      "step": 8666
+    },
+    {
+      "epoch": 0.6017914178586308,
+      "grad_norm": 2.6767288226483736,
+      "learning_rate": 3.6133721141659726e-06,
+      "loss": 0.2048,
+      "step": 8667
+    },
+    {
+      "epoch": 0.6018608526593529,
+      "grad_norm": 3.3392668204244385,
+      "learning_rate": 3.612291768818772e-06,
+      "loss": 0.3512,
+      "step": 8668
+    },
+    {
+      "epoch": 0.601930287460075,
+      "grad_norm": 4.348289012079685,
+      "learning_rate": 3.611211493660285e-06,
+      "loss": 0.5815,
+      "step": 8669
+    },
+    {
+      "epoch": 0.6019997222607971,
+      "grad_norm": 3.4674558129043027,
+      "learning_rate": 3.6101312887451467e-06,
+      "loss": 0.4518,
+      "step": 8670
+    },
+    {
+      "epoch": 0.6020691570615192,
+      "grad_norm": 4.9541904768993685,
+      "learning_rate": 3.609051154127995e-06,
+      "loss": 0.5296,
+      "step": 8671
+    },
+    {
+      "epoch": 0.6021385918622414,
+      "grad_norm": 3.402912515886812,
+      "learning_rate": 3.607971089863464e-06,
+      "loss": 0.4476,
+      "step": 8672
+    },
+    {
+      "epoch": 0.6022080266629635,
+      "grad_norm": 5.242159546952663,
+      "learning_rate": 3.6068910960061787e-06,
+      "loss": 0.3154,
+      "step": 8673
+    },
+    {
+      "epoch": 0.6022774614636855,
+      "grad_norm": 4.719783281340203,
+      "learning_rate": 3.6058111726107647e-06,
+      "loss": 0.4873,
+      "step": 8674
+    },
+    {
+      "epoch": 0.6023468962644077,
+      "grad_norm": 4.158722427726425,
+      "learning_rate": 3.6047313197318456e-06,
+      "loss": 0.5464,
+      "step": 8675
+    },
+    {
+      "epoch": 0.6024163310651298,
+      "grad_norm": 5.3219974669796,
+      "learning_rate": 3.6036515374240376e-06,
+      "loss": 0.687,
+      "step": 8676
+    },
+    {
+      "epoch": 0.602485765865852,
+      "grad_norm": 4.008408998789982,
+      "learning_rate": 3.6025718257419532e-06,
+      "loss": 0.5832,
+      "step": 8677
+    },
+    {
+      "epoch": 0.6025552006665741,
+      "grad_norm": 3.8724585356435326,
+      "learning_rate": 3.6014921847402058e-06,
+      "loss": 0.6422,
+      "step": 8678
+    },
+    {
+      "epoch": 0.6026246354672962,
+      "grad_norm": 4.0352091085747395,
+      "learning_rate": 3.6004126144733996e-06,
+      "loss": 0.4024,
+      "step": 8679
+    },
+    {
+      "epoch": 0.6026940702680184,
+      "grad_norm": 3.991269382579248,
+      "learning_rate": 3.5993331149961417e-06,
+      "loss": 0.572,
+      "step": 8680
+    },
+    {
+      "epoch": 0.6027635050687404,
+      "grad_norm": 3.379690840860304,
+      "learning_rate": 3.598253686363028e-06,
+      "loss": 0.3761,
+      "step": 8681
+    },
+    {
+      "epoch": 0.6028329398694626,
+      "grad_norm": 3.8800209348608625,
+      "learning_rate": 3.597174328628658e-06,
+      "loss": 0.4973,
+      "step": 8682
+    },
+    {
+      "epoch": 0.6029023746701847,
+      "grad_norm": 3.8026134950243926,
+      "learning_rate": 3.5960950418476236e-06,
+      "loss": 0.2977,
+      "step": 8683
+    },
+    {
+      "epoch": 0.6029718094709068,
+      "grad_norm": 3.432440725632957,
+      "learning_rate": 3.5950158260745117e-06,
+      "loss": 0.2832,
+      "step": 8684
+    },
+    {
+      "epoch": 0.603041244271629,
+      "grad_norm": 3.0947062323770393,
+      "learning_rate": 3.5939366813639103e-06,
+      "loss": 0.3841,
+      "step": 8685
+    },
+    {
+      "epoch": 0.603110679072351,
+      "grad_norm": 10.541388090290436,
+      "learning_rate": 3.5928576077703993e-06,
+      "loss": 0.5674,
+      "step": 8686
+    },
+    {
+      "epoch": 0.6031801138730731,
+      "grad_norm": 2.572908154764333,
+      "learning_rate": 3.5917786053485603e-06,
+      "loss": 0.2591,
+      "step": 8687
+    },
+    {
+      "epoch": 0.6032495486737953,
+      "grad_norm": 4.687081412677106,
+      "learning_rate": 3.5906996741529653e-06,
+      "loss": 0.4936,
+      "step": 8688
+    },
+    {
+      "epoch": 0.6033189834745174,
+      "grad_norm": 4.144613169078189,
+      "learning_rate": 3.589620814238185e-06,
+      "loss": 0.5671,
+      "step": 8689
+    },
+    {
+      "epoch": 0.6033884182752396,
+      "grad_norm": 4.044968399254317,
+      "learning_rate": 3.5885420256587895e-06,
+      "loss": 0.4994,
+      "step": 8690
+    },
+    {
+      "epoch": 0.6034578530759617,
+      "grad_norm": 3.562266820497365,
+      "learning_rate": 3.5874633084693395e-06,
+      "loss": 0.2858,
+      "step": 8691
+    },
+    {
+      "epoch": 0.6035272878766837,
+      "grad_norm": 4.885451100321208,
+      "learning_rate": 3.586384662724398e-06,
+      "loss": 0.6234,
+      "step": 8692
+    },
+    {
+      "epoch": 0.6035967226774059,
+      "grad_norm": 3.7994971280086114,
+      "learning_rate": 3.5853060884785197e-06,
+      "loss": 0.3424,
+      "step": 8693
+    },
+    {
+      "epoch": 0.603666157478128,
+      "grad_norm": 3.8891386626011846,
+      "learning_rate": 3.5842275857862595e-06,
+      "loss": 0.5786,
+      "step": 8694
+    },
+    {
+      "epoch": 0.6037355922788502,
+      "grad_norm": 3.4482185141396053,
+      "learning_rate": 3.5831491547021664e-06,
+      "loss": 0.2276,
+      "step": 8695
+    },
+    {
+      "epoch": 0.6038050270795723,
+      "grad_norm": 4.178711851078422,
+      "learning_rate": 3.5820707952807842e-06,
+      "loss": 0.3249,
+      "step": 8696
+    },
+    {
+      "epoch": 0.6038744618802944,
+      "grad_norm": 2.9326195752460325,
+      "learning_rate": 3.580992507576659e-06,
+      "loss": 0.3305,
+      "step": 8697
+    },
+    {
+      "epoch": 0.6039438966810166,
+      "grad_norm": 2.95286702967796,
+      "learning_rate": 3.5799142916443253e-06,
+      "loss": 0.2302,
+      "step": 8698
+    },
+    {
+      "epoch": 0.6040133314817386,
+      "grad_norm": 2.2719617892512636,
+      "learning_rate": 3.5788361475383194e-06,
+      "loss": 0.2012,
+      "step": 8699
+    },
+    {
+      "epoch": 0.6040827662824607,
+      "grad_norm": 3.2873191825193633,
+      "learning_rate": 3.577758075313173e-06,
+      "loss": 0.4519,
+      "step": 8700
+    },
+    {
+      "epoch": 0.6041522010831829,
+      "grad_norm": 3.644117340517901,
+      "learning_rate": 3.5766800750234156e-06,
+      "loss": 0.378,
+      "step": 8701
+    },
+    {
+      "epoch": 0.604221635883905,
+      "grad_norm": 4.367153651821961,
+      "learning_rate": 3.5756021467235686e-06,
+      "loss": 0.5127,
+      "step": 8702
+    },
+    {
+      "epoch": 0.6042910706846272,
+      "grad_norm": 4.947311418662219,
+      "learning_rate": 3.574524290468152e-06,
+      "loss": 0.4625,
+      "step": 8703
+    },
+    {
+      "epoch": 0.6043605054853493,
+      "grad_norm": 4.090274042044305,
+      "learning_rate": 3.5734465063116847e-06,
+      "loss": 0.5356,
+      "step": 8704
+    },
+    {
+      "epoch": 0.6044299402860713,
+      "grad_norm": 3.65364948683183,
+      "learning_rate": 3.5723687943086804e-06,
+      "loss": 0.4798,
+      "step": 8705
+    },
+    {
+      "epoch": 0.6044993750867935,
+      "grad_norm": 4.72837814626655,
+      "learning_rate": 3.571291154513644e-06,
+      "loss": 0.4627,
+      "step": 8706
+    },
+    {
+      "epoch": 0.6045688098875156,
+      "grad_norm": 2.6334657432812123,
+      "learning_rate": 3.570213586981086e-06,
+      "loss": 0.2183,
+      "step": 8707
+    },
+    {
+      "epoch": 0.6046382446882378,
+      "grad_norm": 4.5491830992829545,
+      "learning_rate": 3.569136091765508e-06,
+      "loss": 0.6933,
+      "step": 8708
+    },
+    {
+      "epoch": 0.6047076794889599,
+      "grad_norm": 3.5993397738785045,
+      "learning_rate": 3.5680586689214047e-06,
+      "loss": 0.2839,
+      "step": 8709
+    },
+    {
+      "epoch": 0.604777114289682,
+      "grad_norm": 2.7082637840660566,
+      "learning_rate": 3.566981318503274e-06,
+      "loss": 0.27,
+      "step": 8710
+    },
+    {
+      "epoch": 0.6048465490904041,
+      "grad_norm": 3.705569244903075,
+      "learning_rate": 3.5659040405656074e-06,
+      "loss": 0.2535,
+      "step": 8711
+    },
+    {
+      "epoch": 0.6049159838911262,
+      "grad_norm": 6.678303518008989,
+      "learning_rate": 3.5648268351628935e-06,
+      "loss": 0.7644,
+      "step": 8712
+    },
+    {
+      "epoch": 0.6049854186918484,
+      "grad_norm": 2.821744364470262,
+      "learning_rate": 3.5637497023496115e-06,
+      "loss": 0.2779,
+      "step": 8713
+    },
+    {
+      "epoch": 0.6050548534925705,
+      "grad_norm": 3.198409396898823,
+      "learning_rate": 3.5626726421802455e-06,
+      "loss": 0.3139,
+      "step": 8714
+    },
+    {
+      "epoch": 0.6051242882932926,
+      "grad_norm": 5.676992668860264,
+      "learning_rate": 3.5615956547092715e-06,
+      "loss": 0.7564,
+      "step": 8715
+    },
+    {
+      "epoch": 0.6051937230940148,
+      "grad_norm": 4.042745478267019,
+      "learning_rate": 3.56051873999116e-06,
+      "loss": 0.3531,
+      "step": 8716
+    },
+    {
+      "epoch": 0.6052631578947368,
+      "grad_norm": 3.1634752682152327,
+      "learning_rate": 3.5594418980803825e-06,
+      "loss": 0.418,
+      "step": 8717
+    },
+    {
+      "epoch": 0.6053325926954589,
+      "grad_norm": 9.234277288418367,
+      "learning_rate": 3.558365129031403e-06,
+      "loss": 0.292,
+      "step": 8718
+    },
+    {
+      "epoch": 0.6054020274961811,
+      "grad_norm": 3.7192162728405926,
+      "learning_rate": 3.5572884328986856e-06,
+      "loss": 0.5705,
+      "step": 8719
+    },
+    {
+      "epoch": 0.6054714622969032,
+      "grad_norm": 3.6212870565342894,
+      "learning_rate": 3.5562118097366848e-06,
+      "loss": 0.4759,
+      "step": 8720
+    },
+    {
+      "epoch": 0.6055408970976254,
+      "grad_norm": 5.061682244145416,
+      "learning_rate": 3.5551352595998583e-06,
+      "loss": 0.8653,
+      "step": 8721
+    },
+    {
+      "epoch": 0.6056103318983475,
+      "grad_norm": 3.7414676949175845,
+      "learning_rate": 3.5540587825426565e-06,
+      "loss": 0.5166,
+      "step": 8722
+    },
+    {
+      "epoch": 0.6056797666990695,
+      "grad_norm": 4.648427830199838,
+      "learning_rate": 3.552982378619523e-06,
+      "loss": 0.3677,
+      "step": 8723
+    },
+    {
+      "epoch": 0.6057492014997917,
+      "grad_norm": 3.208149578923469,
+      "learning_rate": 3.5519060478849036e-06,
+      "loss": 0.3744,
+      "step": 8724
+    },
+    {
+      "epoch": 0.6058186363005138,
+      "grad_norm": 3.311356909795269,
+      "learning_rate": 3.550829790393238e-06,
+      "loss": 0.3831,
+      "step": 8725
+    },
+    {
+      "epoch": 0.605888071101236,
+      "grad_norm": 2.8624753519597723,
+      "learning_rate": 3.5497536061989625e-06,
+      "loss": 0.2523,
+      "step": 8726
+    },
+    {
+      "epoch": 0.6059575059019581,
+      "grad_norm": 4.15747624473492,
+      "learning_rate": 3.5486774953565083e-06,
+      "loss": 0.4347,
+      "step": 8727
+    },
+    {
+      "epoch": 0.6060269407026802,
+      "grad_norm": 4.402562774565508,
+      "learning_rate": 3.5476014579203032e-06,
+      "loss": 0.6639,
+      "step": 8728
+    },
+    {
+      "epoch": 0.6060963755034023,
+      "grad_norm": 4.546749278528348,
+      "learning_rate": 3.5465254939447737e-06,
+      "loss": 0.3551,
+      "step": 8729
+    },
+    {
+      "epoch": 0.6061658103041244,
+      "grad_norm": 2.51016087230274,
+      "learning_rate": 3.5454496034843407e-06,
+      "loss": 0.2819,
+      "step": 8730
+    },
+    {
+      "epoch": 0.6062352451048465,
+      "grad_norm": 4.671286233097564,
+      "learning_rate": 3.5443737865934193e-06,
+      "loss": 0.6115,
+      "step": 8731
+    },
+    {
+      "epoch": 0.6063046799055687,
+      "grad_norm": 3.913294174180678,
+      "learning_rate": 3.5432980433264243e-06,
+      "loss": 0.4216,
+      "step": 8732
+    },
+    {
+      "epoch": 0.6063741147062908,
+      "grad_norm": 4.1952545591775054,
+      "learning_rate": 3.5422223737377677e-06,
+      "loss": 0.585,
+      "step": 8733
+    },
+    {
+      "epoch": 0.606443549507013,
+      "grad_norm": 3.5357635734944624,
+      "learning_rate": 3.5411467778818532e-06,
+      "loss": 0.3554,
+      "step": 8734
+    },
+    {
+      "epoch": 0.606512984307735,
+      "grad_norm": 2.8776568312925876,
+      "learning_rate": 3.5400712558130824e-06,
+      "loss": 0.3419,
+      "step": 8735
+    },
+    {
+      "epoch": 0.6065824191084571,
+      "grad_norm": 4.035596338161299,
+      "learning_rate": 3.5389958075858567e-06,
+      "loss": 0.4722,
+      "step": 8736
+    },
+    {
+      "epoch": 0.6066518539091793,
+      "grad_norm": 3.997916688543593,
+      "learning_rate": 3.5379204332545715e-06,
+      "loss": 0.6014,
+      "step": 8737
+    },
+    {
+      "epoch": 0.6067212887099014,
+      "grad_norm": 3.3846190203961797,
+      "learning_rate": 3.5368451328736133e-06,
+      "loss": 0.4997,
+      "step": 8738
+    },
+    {
+      "epoch": 0.6067907235106236,
+      "grad_norm": 3.758923787989218,
+      "learning_rate": 3.535769906497374e-06,
+      "loss": 0.4671,
+      "step": 8739
+    },
+    {
+      "epoch": 0.6068601583113457,
+      "grad_norm": 3.063391493095799,
+      "learning_rate": 3.534694754180237e-06,
+      "loss": 0.2856,
+      "step": 8740
+    },
+    {
+      "epoch": 0.6069295931120677,
+      "grad_norm": 3.5308203911660088,
+      "learning_rate": 3.5336196759765797e-06,
+      "loss": 0.4827,
+      "step": 8741
+    },
+    {
+      "epoch": 0.6069990279127899,
+      "grad_norm": 3.6196541717798354,
+      "learning_rate": 3.5325446719407787e-06,
+      "loss": 0.2828,
+      "step": 8742
+    },
+    {
+      "epoch": 0.607068462713512,
+      "grad_norm": 3.2834500337065387,
+      "learning_rate": 3.5314697421272092e-06,
+      "loss": 0.2852,
+      "step": 8743
+    },
+    {
+      "epoch": 0.6071378975142341,
+      "grad_norm": 4.609996918158254,
+      "learning_rate": 3.5303948865902405e-06,
+      "loss": 0.6473,
+      "step": 8744
+    },
+    {
+      "epoch": 0.6072073323149563,
+      "grad_norm": 3.904154807980599,
+      "learning_rate": 3.5293201053842317e-06,
+      "loss": 0.4375,
+      "step": 8745
+    },
+    {
+      "epoch": 0.6072767671156784,
+      "grad_norm": 3.606667791509874,
+      "learning_rate": 3.52824539856355e-06,
+      "loss": 0.3763,
+      "step": 8746
+    },
+    {
+      "epoch": 0.6073462019164005,
+      "grad_norm": 3.9963688788720058,
+      "learning_rate": 3.527170766182551e-06,
+      "loss": 0.3507,
+      "step": 8747
+    },
+    {
+      "epoch": 0.6074156367171226,
+      "grad_norm": 2.8185441768025803,
+      "learning_rate": 3.5260962082955863e-06,
+      "loss": 0.3351,
+      "step": 8748
+    },
+    {
+      "epoch": 0.6074850715178447,
+      "grad_norm": 3.766094116212072,
+      "learning_rate": 3.5250217249570085e-06,
+      "loss": 0.4437,
+      "step": 8749
+    },
+    {
+      "epoch": 0.6075545063185669,
+      "grad_norm": 3.2812710033297403,
+      "learning_rate": 3.523947316221162e-06,
+      "loss": 0.2882,
+      "step": 8750
+    },
+    {
+      "epoch": 0.607623941119289,
+      "grad_norm": 3.669777838150931,
+      "learning_rate": 3.522872982142392e-06,
+      "loss": 0.316,
+      "step": 8751
+    },
+    {
+      "epoch": 0.6076933759200112,
+      "grad_norm": 3.4990757878870817,
+      "learning_rate": 3.5217987227750327e-06,
+      "loss": 0.4352,
+      "step": 8752
+    },
+    {
+      "epoch": 0.6077628107207332,
+      "grad_norm": 3.7920229325892176,
+      "learning_rate": 3.5207245381734223e-06,
+      "loss": 0.2807,
+      "step": 8753
+    },
+    {
+      "epoch": 0.6078322455214553,
+      "grad_norm": 3.1464042249534203,
+      "learning_rate": 3.5196504283918897e-06,
+      "loss": 0.2947,
+      "step": 8754
+    },
+    {
+      "epoch": 0.6079016803221775,
+      "grad_norm": 3.658905341414522,
+      "learning_rate": 3.518576393484766e-06,
+      "loss": 0.4309,
+      "step": 8755
+    },
+    {
+      "epoch": 0.6079711151228996,
+      "grad_norm": 4.820644362463891,
+      "learning_rate": 3.5175024335063707e-06,
+      "loss": 0.6304,
+      "step": 8756
+    },
+    {
+      "epoch": 0.6080405499236217,
+      "grad_norm": 2.4751253879652,
+      "learning_rate": 3.516428548511023e-06,
+      "loss": 0.1461,
+      "step": 8757
+    },
+    {
+      "epoch": 0.6081099847243439,
+      "grad_norm": 3.3926635511300534,
+      "learning_rate": 3.515354738553044e-06,
+      "loss": 0.5285,
+      "step": 8758
+    },
+    {
+      "epoch": 0.6081794195250659,
+      "grad_norm": 3.292710082121794,
+      "learning_rate": 3.51428100368674e-06,
+      "loss": 0.1882,
+      "step": 8759
+    },
+    {
+      "epoch": 0.6082488543257881,
+      "grad_norm": 3.593982526990828,
+      "learning_rate": 3.5132073439664215e-06,
+      "loss": 0.3269,
+      "step": 8760
+    },
+    {
+      "epoch": 0.6083182891265102,
+      "grad_norm": 4.994855225589099,
+      "learning_rate": 3.5121337594463923e-06,
+      "loss": 0.5468,
+      "step": 8761
+    },
+    {
+      "epoch": 0.6083877239272323,
+      "grad_norm": 7.484275729992098,
+      "learning_rate": 3.5110602501809565e-06,
+      "loss": 0.6978,
+      "step": 8762
+    },
+    {
+      "epoch": 0.6084571587279545,
+      "grad_norm": 11.848722941882333,
+      "learning_rate": 3.5099868162244077e-06,
+      "loss": 0.7132,
+      "step": 8763
+    },
+    {
+      "epoch": 0.6085265935286766,
+      "grad_norm": 3.767247583808914,
+      "learning_rate": 3.508913457631038e-06,
+      "loss": 0.3759,
+      "step": 8764
+    },
+    {
+      "epoch": 0.6085960283293987,
+      "grad_norm": 5.878782909083482,
+      "learning_rate": 3.5078401744551404e-06,
+      "loss": 0.7495,
+      "step": 8765
+    },
+    {
+      "epoch": 0.6086654631301208,
+      "grad_norm": 4.450280679700323,
+      "learning_rate": 3.5067669667509975e-06,
+      "loss": 0.4169,
+      "step": 8766
+    },
+    {
+      "epoch": 0.6087348979308429,
+      "grad_norm": 3.887879756049915,
+      "learning_rate": 3.5056938345728908e-06,
+      "loss": 0.3313,
+      "step": 8767
+    },
+    {
+      "epoch": 0.6088043327315651,
+      "grad_norm": 3.4498904786225433,
+      "learning_rate": 3.5046207779750994e-06,
+      "loss": 0.2941,
+      "step": 8768
+    },
+    {
+      "epoch": 0.6088737675322872,
+      "grad_norm": 4.121688356850154,
+      "learning_rate": 3.503547797011899e-06,
+      "loss": 0.5696,
+      "step": 8769
+    },
+    {
+      "epoch": 0.6089432023330094,
+      "grad_norm": 3.9662655076621314,
+      "learning_rate": 3.5024748917375552e-06,
+      "loss": 0.5685,
+      "step": 8770
+    },
+    {
+      "epoch": 0.6090126371337314,
+      "grad_norm": 4.542597359992858,
+      "learning_rate": 3.5014020622063365e-06,
+      "loss": 0.4078,
+      "step": 8771
+    },
+    {
+      "epoch": 0.6090820719344535,
+      "grad_norm": 3.323797452605993,
+      "learning_rate": 3.500329308472508e-06,
+      "loss": 0.3538,
+      "step": 8772
+    },
+    {
+      "epoch": 0.6091515067351757,
+      "grad_norm": 4.148309669616246,
+      "learning_rate": 3.4992566305903253e-06,
+      "loss": 0.4835,
+      "step": 8773
+    },
+    {
+      "epoch": 0.6092209415358978,
+      "grad_norm": 3.653952497111871,
+      "learning_rate": 3.4981840286140425e-06,
+      "loss": 0.4486,
+      "step": 8774
+    },
+    {
+      "epoch": 0.6092903763366199,
+      "grad_norm": 2.695493999153073,
+      "learning_rate": 3.497111502597913e-06,
+      "loss": 0.1049,
+      "step": 8775
+    },
+    {
+      "epoch": 0.6093598111373421,
+      "grad_norm": 4.07849329769582,
+      "learning_rate": 3.4960390525961853e-06,
+      "loss": 0.6116,
+      "step": 8776
+    },
+    {
+      "epoch": 0.6094292459380641,
+      "grad_norm": 4.275370125409202,
+      "learning_rate": 3.494966678663097e-06,
+      "loss": 0.5959,
+      "step": 8777
+    },
+    {
+      "epoch": 0.6094986807387863,
+      "grad_norm": 5.791210458057832,
+      "learning_rate": 3.4938943808528924e-06,
+      "loss": 0.3814,
+      "step": 8778
+    },
+    {
+      "epoch": 0.6095681155395084,
+      "grad_norm": 2.9804918785192314,
+      "learning_rate": 3.4928221592198053e-06,
+      "loss": 0.2367,
+      "step": 8779
+    },
+    {
+      "epoch": 0.6096375503402305,
+      "grad_norm": 5.398198364466904,
+      "learning_rate": 3.4917500138180695e-06,
+      "loss": 0.3856,
+      "step": 8780
+    },
+    {
+      "epoch": 0.6097069851409527,
+      "grad_norm": 3.2649536457690767,
+      "learning_rate": 3.490677944701909e-06,
+      "loss": 0.1792,
+      "step": 8781
+    },
+    {
+      "epoch": 0.6097764199416748,
+      "grad_norm": 2.819860429628112,
+      "learning_rate": 3.489605951925552e-06,
+      "loss": 0.4197,
+      "step": 8782
+    },
+    {
+      "epoch": 0.6098458547423969,
+      "grad_norm": 5.247148237561463,
+      "learning_rate": 3.4885340355432175e-06,
+      "loss": 0.7434,
+      "step": 8783
+    },
+    {
+      "epoch": 0.609915289543119,
+      "grad_norm": 4.236740290689382,
+      "learning_rate": 3.4874621956091186e-06,
+      "loss": 0.3853,
+      "step": 8784
+    },
+    {
+      "epoch": 0.6099847243438411,
+      "grad_norm": 3.6618283189949636,
+      "learning_rate": 3.486390432177472e-06,
+      "loss": 0.3223,
+      "step": 8785
+    },
+    {
+      "epoch": 0.6100541591445633,
+      "grad_norm": 3.6704054038804292,
+      "learning_rate": 3.4853187453024827e-06,
+      "loss": 0.4361,
+      "step": 8786
+    },
+    {
+      "epoch": 0.6101235939452854,
+      "grad_norm": 3.6285768004194274,
+      "learning_rate": 3.48424713503836e-06,
+      "loss": 0.2842,
+      "step": 8787
+    },
+    {
+      "epoch": 0.6101930287460074,
+      "grad_norm": 4.42477390677402,
+      "learning_rate": 3.483175601439301e-06,
+      "loss": 0.4643,
+      "step": 8788
+    },
+    {
+      "epoch": 0.6102624635467296,
+      "grad_norm": 4.036486958323342,
+      "learning_rate": 3.482104144559503e-06,
+      "loss": 0.4778,
+      "step": 8789
+    },
+    {
+      "epoch": 0.6103318983474517,
+      "grad_norm": 5.144953646562805,
+      "learning_rate": 3.4810327644531606e-06,
+      "loss": 0.468,
+      "step": 8790
+    },
+    {
+      "epoch": 0.6104013331481739,
+      "grad_norm": 3.327906577397172,
+      "learning_rate": 3.479961461174461e-06,
+      "loss": 0.347,
+      "step": 8791
+    },
+    {
+      "epoch": 0.610470767948896,
+      "grad_norm": 4.536725368663805,
+      "learning_rate": 3.478890234777591e-06,
+      "loss": 0.4807,
+      "step": 8792
+    },
+    {
+      "epoch": 0.6105402027496181,
+      "grad_norm": 3.066558974834664,
+      "learning_rate": 3.4778190853167315e-06,
+      "loss": 0.35,
+      "step": 8793
+    },
+    {
+      "epoch": 0.6106096375503403,
+      "grad_norm": 3.764874721401947,
+      "learning_rate": 3.4767480128460616e-06,
+      "loss": 0.3399,
+      "step": 8794
+    },
+    {
+      "epoch": 0.6106790723510623,
+      "grad_norm": 3.8708608296847467,
+      "learning_rate": 3.475677017419753e-06,
+      "loss": 0.3217,
+      "step": 8795
+    },
+    {
+      "epoch": 0.6107485071517845,
+      "grad_norm": 3.493059073168975,
+      "learning_rate": 3.474606099091975e-06,
+      "loss": 0.4612,
+      "step": 8796
+    },
+    {
+      "epoch": 0.6108179419525066,
+      "grad_norm": 3.7398319023151787,
+      "learning_rate": 3.4735352579168956e-06,
+      "loss": 0.4019,
+      "step": 8797
+    },
+    {
+      "epoch": 0.6108873767532287,
+      "grad_norm": 2.568160475932117,
+      "learning_rate": 3.4724644939486753e-06,
+      "loss": 0.1737,
+      "step": 8798
+    },
+    {
+      "epoch": 0.6109568115539509,
+      "grad_norm": 4.234843802130741,
+      "learning_rate": 3.471393807241471e-06,
+      "loss": 0.6381,
+      "step": 8799
+    },
+    {
+      "epoch": 0.611026246354673,
+      "grad_norm": 3.55702450378208,
+      "learning_rate": 3.4703231978494388e-06,
+      "loss": 0.4172,
+      "step": 8800
+    },
+    {
+      "epoch": 0.611095681155395,
+      "grad_norm": 3.863896878506644,
+      "learning_rate": 3.46925266582673e-06,
+      "loss": 0.4431,
+      "step": 8801
+    },
+    {
+      "epoch": 0.6111651159561172,
+      "grad_norm": 3.3630169996881905,
+      "learning_rate": 3.4681822112274877e-06,
+      "loss": 0.3636,
+      "step": 8802
+    },
+    {
+      "epoch": 0.6112345507568393,
+      "grad_norm": 4.130317869284291,
+      "learning_rate": 3.467111834105855e-06,
+      "loss": 0.44,
+      "step": 8803
+    },
+    {
+      "epoch": 0.6113039855575615,
+      "grad_norm": 3.447223413116083,
+      "learning_rate": 3.4660415345159727e-06,
+      "loss": 0.3196,
+      "step": 8804
+    },
+    {
+      "epoch": 0.6113734203582836,
+      "grad_norm": 4.445051833966361,
+      "learning_rate": 3.4649713125119742e-06,
+      "loss": 0.6113,
+      "step": 8805
+    },
+    {
+      "epoch": 0.6114428551590056,
+      "grad_norm": 3.2950981307291056,
+      "learning_rate": 3.4639011681479885e-06,
+      "loss": 0.3285,
+      "step": 8806
+    },
+    {
+      "epoch": 0.6115122899597278,
+      "grad_norm": 4.526714737940395,
+      "learning_rate": 3.4628311014781435e-06,
+      "loss": 0.6081,
+      "step": 8807
+    },
+    {
+      "epoch": 0.6115817247604499,
+      "grad_norm": 3.2820146872311375,
+      "learning_rate": 3.461761112556563e-06,
+      "loss": 0.2442,
+      "step": 8808
+    },
+    {
+      "epoch": 0.6116511595611721,
+      "grad_norm": 3.3767762409644915,
+      "learning_rate": 3.4606912014373633e-06,
+      "loss": 0.4206,
+      "step": 8809
+    },
+    {
+      "epoch": 0.6117205943618942,
+      "grad_norm": 2.828917003674686,
+      "learning_rate": 3.4596213681746615e-06,
+      "loss": 0.227,
+      "step": 8810
+    },
+    {
+      "epoch": 0.6117900291626163,
+      "grad_norm": 3.070436221140293,
+      "learning_rate": 3.4585516128225673e-06,
+      "loss": 0.3192,
+      "step": 8811
+    },
+    {
+      "epoch": 0.6118594639633385,
+      "grad_norm": 2.9965596268550847,
+      "learning_rate": 3.4574819354351896e-06,
+      "loss": 0.2736,
+      "step": 8812
+    },
+    {
+      "epoch": 0.6119288987640605,
+      "grad_norm": 4.253918162474615,
+      "learning_rate": 3.4564123360666292e-06,
+      "loss": 0.4445,
+      "step": 8813
+    },
+    {
+      "epoch": 0.6119983335647826,
+      "grad_norm": 3.4834202100898275,
+      "learning_rate": 3.455342814770986e-06,
+      "loss": 0.467,
+      "step": 8814
+    },
+    {
+      "epoch": 0.6120677683655048,
+      "grad_norm": 2.8502495553489937,
+      "learning_rate": 3.454273371602357e-06,
+      "loss": 0.2169,
+      "step": 8815
+    },
+    {
+      "epoch": 0.6121372031662269,
+      "grad_norm": 6.873085745477257,
+      "learning_rate": 3.4532040066148296e-06,
+      "loss": 0.7758,
+      "step": 8816
+    },
+    {
+      "epoch": 0.6122066379669491,
+      "grad_norm": 3.9070505450184583,
+      "learning_rate": 3.4521347198624945e-06,
+      "loss": 0.5103,
+      "step": 8817
+    },
+    {
+      "epoch": 0.6122760727676712,
+      "grad_norm": 3.5089996568781245,
+      "learning_rate": 3.4510655113994317e-06,
+      "loss": 0.3198,
+      "step": 8818
+    },
+    {
+      "epoch": 0.6123455075683932,
+      "grad_norm": 4.26302970854848,
+      "learning_rate": 3.4499963812797253e-06,
+      "loss": 0.4869,
+      "step": 8819
+    },
+    {
+      "epoch": 0.6124149423691154,
+      "grad_norm": 4.263491769328727,
+      "learning_rate": 3.4489273295574473e-06,
+      "loss": 0.4804,
+      "step": 8820
+    },
+    {
+      "epoch": 0.6124843771698375,
+      "grad_norm": 3.59272162039872,
+      "learning_rate": 3.4478583562866687e-06,
+      "loss": 0.3145,
+      "step": 8821
+    },
+    {
+      "epoch": 0.6125538119705597,
+      "grad_norm": 4.565373399713263,
+      "learning_rate": 3.4467894615214593e-06,
+      "loss": 0.6119,
+      "step": 8822
+    },
+    {
+      "epoch": 0.6126232467712818,
+      "grad_norm": 4.733431228688889,
+      "learning_rate": 3.44572064531588e-06,
+      "loss": 0.4495,
+      "step": 8823
+    },
+    {
+      "epoch": 0.6126926815720038,
+      "grad_norm": 2.499875554095829,
+      "learning_rate": 3.4446519077239925e-06,
+      "loss": 0.2889,
+      "step": 8824
+    },
+    {
+      "epoch": 0.612762116372726,
+      "grad_norm": 4.210601456668067,
+      "learning_rate": 3.4435832487998513e-06,
+      "loss": 0.6043,
+      "step": 8825
+    },
+    {
+      "epoch": 0.6128315511734481,
+      "grad_norm": 4.0874179985612615,
+      "learning_rate": 3.4425146685975084e-06,
+      "loss": 0.3603,
+      "step": 8826
+    },
+    {
+      "epoch": 0.6129009859741703,
+      "grad_norm": 7.6718983598756845,
+      "learning_rate": 3.4414461671710118e-06,
+      "loss": 0.3045,
+      "step": 8827
+    },
+    {
+      "epoch": 0.6129704207748924,
+      "grad_norm": 4.114497972604883,
+      "learning_rate": 3.440377744574403e-06,
+      "loss": 0.4455,
+      "step": 8828
+    },
+    {
+      "epoch": 0.6130398555756145,
+      "grad_norm": 7.6175148987063785,
+      "learning_rate": 3.4393094008617235e-06,
+      "loss": 0.5355,
+      "step": 8829
+    },
+    {
+      "epoch": 0.6131092903763367,
+      "grad_norm": 3.3153797679956667,
+      "learning_rate": 3.4382411360870103e-06,
+      "loss": 0.2266,
+      "step": 8830
+    },
+    {
+      "epoch": 0.6131787251770587,
+      "grad_norm": 4.658982003162793,
+      "learning_rate": 3.4371729503042905e-06,
+      "loss": 0.456,
+      "step": 8831
+    },
+    {
+      "epoch": 0.6132481599777808,
+      "grad_norm": 2.995687420310173,
+      "learning_rate": 3.4361048435675946e-06,
+      "loss": 0.2585,
+      "step": 8832
+    },
+    {
+      "epoch": 0.613317594778503,
+      "grad_norm": 4.041921263592843,
+      "learning_rate": 3.4350368159309485e-06,
+      "loss": 0.4858,
+      "step": 8833
+    },
+    {
+      "epoch": 0.6133870295792251,
+      "grad_norm": 3.6743486765491986,
+      "learning_rate": 3.433968867448368e-06,
+      "loss": 0.4095,
+      "step": 8834
+    },
+    {
+      "epoch": 0.6134564643799473,
+      "grad_norm": 4.7755375754577365,
+      "learning_rate": 3.4329009981738694e-06,
+      "loss": 0.519,
+      "step": 8835
+    },
+    {
+      "epoch": 0.6135258991806694,
+      "grad_norm": 3.8121589577816497,
+      "learning_rate": 3.4318332081614653e-06,
+      "loss": 0.4024,
+      "step": 8836
+    },
+    {
+      "epoch": 0.6135953339813914,
+      "grad_norm": 2.4652277615141935,
+      "learning_rate": 3.4307654974651646e-06,
+      "loss": 0.1867,
+      "step": 8837
+    },
+    {
+      "epoch": 0.6136647687821136,
+      "grad_norm": 3.743097312374,
+      "learning_rate": 3.429697866138967e-06,
+      "loss": 0.1834,
+      "step": 8838
+    },
+    {
+      "epoch": 0.6137342035828357,
+      "grad_norm": 3.7070803664321565,
+      "learning_rate": 3.428630314236876e-06,
+      "loss": 0.2808,
+      "step": 8839
+    },
+    {
+      "epoch": 0.6138036383835579,
+      "grad_norm": 4.470430758771801,
+      "learning_rate": 3.427562841812886e-06,
+      "loss": 0.5572,
+      "step": 8840
+    },
+    {
+      "epoch": 0.61387307318428,
+      "grad_norm": 3.9595464403271685,
+      "learning_rate": 3.4264954489209863e-06,
+      "loss": 0.4157,
+      "step": 8841
+    },
+    {
+      "epoch": 0.613942507985002,
+      "grad_norm": 4.2733463495985955,
+      "learning_rate": 3.4254281356151665e-06,
+      "loss": 0.5208,
+      "step": 8842
+    },
+    {
+      "epoch": 0.6140119427857242,
+      "grad_norm": 4.050260943499243,
+      "learning_rate": 3.42436090194941e-06,
+      "loss": 0.5675,
+      "step": 8843
+    },
+    {
+      "epoch": 0.6140813775864463,
+      "grad_norm": 3.42306606736064,
+      "learning_rate": 3.4232937479776977e-06,
+      "loss": 0.328,
+      "step": 8844
+    },
+    {
+      "epoch": 0.6141508123871684,
+      "grad_norm": 3.086523847368888,
+      "learning_rate": 3.4222266737540017e-06,
+      "loss": 0.3963,
+      "step": 8845
+    },
+    {
+      "epoch": 0.6142202471878906,
+      "grad_norm": 3.137793094328264,
+      "learning_rate": 3.421159679332296e-06,
+      "loss": 0.2813,
+      "step": 8846
+    },
+    {
+      "epoch": 0.6142896819886127,
+      "grad_norm": 2.7669615693239464,
+      "learning_rate": 3.4200927647665484e-06,
+      "loss": 0.2516,
+      "step": 8847
+    },
+    {
+      "epoch": 0.6143591167893349,
+      "grad_norm": 3.800275682901952,
+      "learning_rate": 3.4190259301107187e-06,
+      "loss": 0.5211,
+      "step": 8848
+    },
+    {
+      "epoch": 0.6144285515900569,
+      "grad_norm": 3.858443102032203,
+      "learning_rate": 3.4179591754187692e-06,
+      "loss": 0.4899,
+      "step": 8849
+    },
+    {
+      "epoch": 0.614497986390779,
+      "grad_norm": 3.7112136230199213,
+      "learning_rate": 3.4168925007446535e-06,
+      "loss": 0.5609,
+      "step": 8850
+    },
+    {
+      "epoch": 0.6145674211915012,
+      "grad_norm": 3.5705100471043894,
+      "learning_rate": 3.415825906142326e-06,
+      "loss": 0.243,
+      "step": 8851
+    },
+    {
+      "epoch": 0.6146368559922233,
+      "grad_norm": 3.5555371947258845,
+      "learning_rate": 3.4147593916657294e-06,
+      "loss": 0.4282,
+      "step": 8852
+    },
+    {
+      "epoch": 0.6147062907929455,
+      "grad_norm": 4.348133127226109,
+      "learning_rate": 3.41369295736881e-06,
+      "loss": 0.526,
+      "step": 8853
+    },
+    {
+      "epoch": 0.6147757255936676,
+      "grad_norm": 2.9015866024526384,
+      "learning_rate": 3.4126266033055042e-06,
+      "loss": 0.2459,
+      "step": 8854
+    },
+    {
+      "epoch": 0.6148451603943896,
+      "grad_norm": 3.443649522031746,
+      "learning_rate": 3.411560329529752e-06,
+      "loss": 0.3978,
+      "step": 8855
+    },
+    {
+      "epoch": 0.6149145951951118,
+      "grad_norm": 3.8251773224159353,
+      "learning_rate": 3.4104941360954785e-06,
+      "loss": 0.5023,
+      "step": 8856
+    },
+    {
+      "epoch": 0.6149840299958339,
+      "grad_norm": 4.687300855272303,
+      "learning_rate": 3.4094280230566123e-06,
+      "loss": 0.4949,
+      "step": 8857
+    },
+    {
+      "epoch": 0.615053464796556,
+      "grad_norm": 3.3203615763143164,
+      "learning_rate": 3.4083619904670796e-06,
+      "loss": 0.3176,
+      "step": 8858
+    },
+    {
+      "epoch": 0.6151228995972782,
+      "grad_norm": 4.278693301282499,
+      "learning_rate": 3.407296038380794e-06,
+      "loss": 0.3024,
+      "step": 8859
+    },
+    {
+      "epoch": 0.6151923343980003,
+      "grad_norm": 3.568714253800124,
+      "learning_rate": 3.406230166851673e-06,
+      "loss": 0.4886,
+      "step": 8860
+    },
+    {
+      "epoch": 0.6152617691987224,
+      "grad_norm": 2.1286816237504875,
+      "learning_rate": 3.4051643759336256e-06,
+      "loss": 0.1472,
+      "step": 8861
+    },
+    {
+      "epoch": 0.6153312039994445,
+      "grad_norm": 4.20814328590448,
+      "learning_rate": 3.404098665680562e-06,
+      "loss": 0.684,
+      "step": 8862
+    },
+    {
+      "epoch": 0.6154006388001666,
+      "grad_norm": 3.4462645805893715,
+      "learning_rate": 3.4030330361463803e-06,
+      "loss": 0.3764,
+      "step": 8863
+    },
+    {
+      "epoch": 0.6154700736008888,
+      "grad_norm": 3.4337351573514594,
+      "learning_rate": 3.40196748738498e-06,
+      "loss": 0.4078,
+      "step": 8864
+    },
+    {
+      "epoch": 0.6155395084016109,
+      "grad_norm": 4.411755843560975,
+      "learning_rate": 3.400902019450258e-06,
+      "loss": 0.4952,
+      "step": 8865
+    },
+    {
+      "epoch": 0.6156089432023331,
+      "grad_norm": 3.8201259387704023,
+      "learning_rate": 3.3998366323961013e-06,
+      "loss": 0.5878,
+      "step": 8866
+    },
+    {
+      "epoch": 0.6156783780030551,
+      "grad_norm": 3.998293258277663,
+      "learning_rate": 3.3987713262763957e-06,
+      "loss": 0.4708,
+      "step": 8867
+    },
+    {
+      "epoch": 0.6157478128037772,
+      "grad_norm": 3.6092836582110697,
+      "learning_rate": 3.397706101145026e-06,
+      "loss": 0.3317,
+      "step": 8868
+    },
+    {
+      "epoch": 0.6158172476044994,
+      "grad_norm": 2.456280766453252,
+      "learning_rate": 3.396640957055869e-06,
+      "loss": 0.1868,
+      "step": 8869
+    },
+    {
+      "epoch": 0.6158866824052215,
+      "grad_norm": 4.299979372384046,
+      "learning_rate": 3.395575894062797e-06,
+      "loss": 0.4995,
+      "step": 8870
+    },
+    {
+      "epoch": 0.6159561172059436,
+      "grad_norm": 3.5152924490110893,
+      "learning_rate": 3.3945109122196797e-06,
+      "loss": 0.3992,
+      "step": 8871
+    },
+    {
+      "epoch": 0.6160255520066658,
+      "grad_norm": 4.327403529507235,
+      "learning_rate": 3.3934460115803863e-06,
+      "loss": 0.6731,
+      "step": 8872
+    },
+    {
+      "epoch": 0.6160949868073878,
+      "grad_norm": 2.2403095590063744,
+      "learning_rate": 3.3923811921987748e-06,
+      "loss": 0.1671,
+      "step": 8873
+    },
+    {
+      "epoch": 0.61616442160811,
+      "grad_norm": 3.5622165988462764,
+      "learning_rate": 3.3913164541287026e-06,
+      "loss": 0.3192,
+      "step": 8874
+    },
+    {
+      "epoch": 0.6162338564088321,
+      "grad_norm": 3.8751735615082565,
+      "learning_rate": 3.390251797424024e-06,
+      "loss": 0.4521,
+      "step": 8875
+    },
+    {
+      "epoch": 0.6163032912095542,
+      "grad_norm": 3.932985385424576,
+      "learning_rate": 3.389187222138589e-06,
+      "loss": 0.2698,
+      "step": 8876
+    },
+    {
+      "epoch": 0.6163727260102764,
+      "grad_norm": 5.404326713159761,
+      "learning_rate": 3.3881227283262403e-06,
+      "loss": 0.5754,
+      "step": 8877
+    },
+    {
+      "epoch": 0.6164421608109985,
+      "grad_norm": 3.7388839658478794,
+      "learning_rate": 3.3870583160408203e-06,
+      "loss": 0.5022,
+      "step": 8878
+    },
+    {
+      "epoch": 0.6165115956117206,
+      "grad_norm": 3.0764830862051675,
+      "learning_rate": 3.3859939853361674e-06,
+      "loss": 0.1961,
+      "step": 8879
+    },
+    {
+      "epoch": 0.6165810304124427,
+      "grad_norm": 3.710339976383918,
+      "learning_rate": 3.3849297362661092e-06,
+      "loss": 0.3062,
+      "step": 8880
+    },
+    {
+      "epoch": 0.6166504652131648,
+      "grad_norm": 3.4193998625342212,
+      "learning_rate": 3.3838655688844785e-06,
+      "loss": 0.3121,
+      "step": 8881
+    },
+    {
+      "epoch": 0.616719900013887,
+      "grad_norm": 3.998249990459775,
+      "learning_rate": 3.3828014832450987e-06,
+      "loss": 0.4919,
+      "step": 8882
+    },
+    {
+      "epoch": 0.6167893348146091,
+      "grad_norm": 4.003892193030432,
+      "learning_rate": 3.3817374794017916e-06,
+      "loss": 0.586,
+      "step": 8883
+    },
+    {
+      "epoch": 0.6168587696153313,
+      "grad_norm": 2.780894147925231,
+      "learning_rate": 3.380673557408369e-06,
+      "loss": 0.1996,
+      "step": 8884
+    },
+    {
+      "epoch": 0.6169282044160533,
+      "grad_norm": 4.308226561724884,
+      "learning_rate": 3.3796097173186472e-06,
+      "loss": 0.4837,
+      "step": 8885
+    },
+    {
+      "epoch": 0.6169976392167754,
+      "grad_norm": 2.2490334509221808,
+      "learning_rate": 3.37854595918643e-06,
+      "loss": 0.2372,
+      "step": 8886
+    },
+    {
+      "epoch": 0.6170670740174976,
+      "grad_norm": 3.448596747981524,
+      "learning_rate": 3.377482283065526e-06,
+      "loss": 0.4637,
+      "step": 8887
+    },
+    {
+      "epoch": 0.6171365088182197,
+      "grad_norm": 3.580542009223081,
+      "learning_rate": 3.376418689009732e-06,
+      "loss": 0.3825,
+      "step": 8888
+    },
+    {
+      "epoch": 0.6172059436189418,
+      "grad_norm": 4.148608292931165,
+      "learning_rate": 3.3753551770728417e-06,
+      "loss": 0.5765,
+      "step": 8889
+    },
+    {
+      "epoch": 0.617275378419664,
+      "grad_norm": 3.565873083541428,
+      "learning_rate": 3.37429174730865e-06,
+      "loss": 0.2782,
+      "step": 8890
+    },
+    {
+      "epoch": 0.617344813220386,
+      "grad_norm": 5.038561254483094,
+      "learning_rate": 3.3732283997709414e-06,
+      "loss": 0.6114,
+      "step": 8891
+    },
+    {
+      "epoch": 0.6174142480211082,
+      "grad_norm": 3.4638264423546223,
+      "learning_rate": 3.3721651345134985e-06,
+      "loss": 0.3435,
+      "step": 8892
+    },
+    {
+      "epoch": 0.6174836828218303,
+      "grad_norm": 5.4677454552394495,
+      "learning_rate": 3.371101951590101e-06,
+      "loss": 0.5218,
+      "step": 8893
+    },
+    {
+      "epoch": 0.6175531176225524,
+      "grad_norm": 4.082852809269385,
+      "learning_rate": 3.3700388510545256e-06,
+      "loss": 0.4324,
+      "step": 8894
+    },
+    {
+      "epoch": 0.6176225524232746,
+      "grad_norm": 2.9785945596914787,
+      "learning_rate": 3.36897583296054e-06,
+      "loss": 0.3668,
+      "step": 8895
+    },
+    {
+      "epoch": 0.6176919872239967,
+      "grad_norm": 7.050132443067151,
+      "learning_rate": 3.3679128973619097e-06,
+      "loss": 0.7564,
+      "step": 8896
+    },
+    {
+      "epoch": 0.6177614220247188,
+      "grad_norm": 3.2394877629889463,
+      "learning_rate": 3.366850044312401e-06,
+      "loss": 0.4109,
+      "step": 8897
+    },
+    {
+      "epoch": 0.6178308568254409,
+      "grad_norm": 3.797158443456836,
+      "learning_rate": 3.365787273865767e-06,
+      "loss": 0.4654,
+      "step": 8898
+    },
+    {
+      "epoch": 0.617900291626163,
+      "grad_norm": 3.117512770790711,
+      "learning_rate": 3.3647245860757626e-06,
+      "loss": 0.2712,
+      "step": 8899
+    },
+    {
+      "epoch": 0.6179697264268852,
+      "grad_norm": 5.029807959578462,
+      "learning_rate": 3.3636619809961398e-06,
+      "loss": 0.3782,
+      "step": 8900
+    },
+    {
+      "epoch": 0.6180391612276073,
+      "grad_norm": 2.7476560484303634,
+      "learning_rate": 3.362599458680643e-06,
+      "loss": 0.2044,
+      "step": 8901
+    },
+    {
+      "epoch": 0.6181085960283293,
+      "grad_norm": 3.111577718647958,
+      "learning_rate": 3.3615370191830104e-06,
+      "loss": 0.3491,
+      "step": 8902
+    },
+    {
+      "epoch": 0.6181780308290515,
+      "grad_norm": 3.2224597351197217,
+      "learning_rate": 3.360474662556982e-06,
+      "loss": 0.2718,
+      "step": 8903
+    },
+    {
+      "epoch": 0.6182474656297736,
+      "grad_norm": 2.936160571109473,
+      "learning_rate": 3.3594123888562915e-06,
+      "loss": 0.2204,
+      "step": 8904
+    },
+    {
+      "epoch": 0.6183169004304958,
+      "grad_norm": 3.544824785258668,
+      "learning_rate": 3.3583501981346646e-06,
+      "loss": 0.4693,
+      "step": 8905
+    },
+    {
+      "epoch": 0.6183863352312179,
+      "grad_norm": 4.118583421756876,
+      "learning_rate": 3.3572880904458267e-06,
+      "loss": 0.4732,
+      "step": 8906
+    },
+    {
+      "epoch": 0.61845577003194,
+      "grad_norm": 5.748645519302652,
+      "learning_rate": 3.3562260658434984e-06,
+      "loss": 0.3019,
+      "step": 8907
+    },
+    {
+      "epoch": 0.6185252048326622,
+      "grad_norm": 4.25843478269813,
+      "learning_rate": 3.355164124381398e-06,
+      "loss": 0.4267,
+      "step": 8908
+    },
+    {
+      "epoch": 0.6185946396333842,
+      "grad_norm": 4.090064727931622,
+      "learning_rate": 3.3541022661132324e-06,
+      "loss": 0.525,
+      "step": 8909
+    },
+    {
+      "epoch": 0.6186640744341064,
+      "grad_norm": 4.482175342473842,
+      "learning_rate": 3.3530404910927124e-06,
+      "loss": 0.4554,
+      "step": 8910
+    },
+    {
+      "epoch": 0.6187335092348285,
+      "grad_norm": 4.808436071628028,
+      "learning_rate": 3.35197879937354e-06,
+      "loss": 0.2951,
+      "step": 8911
+    },
+    {
+      "epoch": 0.6188029440355506,
+      "grad_norm": 4.612688068329443,
+      "learning_rate": 3.3509171910094162e-06,
+      "loss": 0.6373,
+      "step": 8912
+    },
+    {
+      "epoch": 0.6188723788362728,
+      "grad_norm": 2.8273767538153622,
+      "learning_rate": 3.3498556660540337e-06,
+      "loss": 0.2209,
+      "step": 8913
+    },
+    {
+      "epoch": 0.6189418136369949,
+      "grad_norm": 4.23562953268146,
+      "learning_rate": 3.348794224561085e-06,
+      "loss": 0.2846,
+      "step": 8914
+    },
+    {
+      "epoch": 0.6190112484377169,
+      "grad_norm": 3.2248588096946573,
+      "learning_rate": 3.347732866584258e-06,
+      "loss": 0.2979,
+      "step": 8915
+    },
+    {
+      "epoch": 0.6190806832384391,
+      "grad_norm": 4.293269968820316,
+      "learning_rate": 3.3466715921772297e-06,
+      "loss": 0.5579,
+      "step": 8916
+    },
+    {
+      "epoch": 0.6191501180391612,
+      "grad_norm": 4.148686548597329,
+      "learning_rate": 3.345610401393683e-06,
+      "loss": 0.5158,
+      "step": 8917
+    },
+    {
+      "epoch": 0.6192195528398834,
+      "grad_norm": 3.254696254648424,
+      "learning_rate": 3.344549294287288e-06,
+      "loss": 0.2812,
+      "step": 8918
+    },
+    {
+      "epoch": 0.6192889876406055,
+      "grad_norm": 3.9736566934074675,
+      "learning_rate": 3.34348827091172e-06,
+      "loss": 0.6182,
+      "step": 8919
+    },
+    {
+      "epoch": 0.6193584224413275,
+      "grad_norm": 3.3794847061171485,
+      "learning_rate": 3.342427331320639e-06,
+      "loss": 0.4227,
+      "step": 8920
+    },
+    {
+      "epoch": 0.6194278572420497,
+      "grad_norm": 5.1063443574964955,
+      "learning_rate": 3.341366475567707e-06,
+      "loss": 0.5266,
+      "step": 8921
+    },
+    {
+      "epoch": 0.6194972920427718,
+      "grad_norm": 2.4202464062809286,
+      "learning_rate": 3.340305703706584e-06,
+      "loss": 0.2438,
+      "step": 8922
+    },
+    {
+      "epoch": 0.619566726843494,
+      "grad_norm": 3.2525305594531497,
+      "learning_rate": 3.339245015790918e-06,
+      "loss": 0.2282,
+      "step": 8923
+    },
+    {
+      "epoch": 0.6196361616442161,
+      "grad_norm": 4.358863905389245,
+      "learning_rate": 3.338184411874361e-06,
+      "loss": 0.3752,
+      "step": 8924
+    },
+    {
+      "epoch": 0.6197055964449382,
+      "grad_norm": 4.181693113430663,
+      "learning_rate": 3.3371238920105554e-06,
+      "loss": 0.4458,
+      "step": 8925
+    },
+    {
+      "epoch": 0.6197750312456604,
+      "grad_norm": 4.1391026237390625,
+      "learning_rate": 3.336063456253143e-06,
+      "loss": 0.4326,
+      "step": 8926
+    },
+    {
+      "epoch": 0.6198444660463824,
+      "grad_norm": 3.074673598875534,
+      "learning_rate": 3.3350031046557573e-06,
+      "loss": 0.2806,
+      "step": 8927
+    },
+    {
+      "epoch": 0.6199139008471045,
+      "grad_norm": 3.189230893579366,
+      "learning_rate": 3.3339428372720295e-06,
+      "loss": 0.3194,
+      "step": 8928
+    },
+    {
+      "epoch": 0.6199833356478267,
+      "grad_norm": 4.586643892909302,
+      "learning_rate": 3.3328826541555905e-06,
+      "loss": 0.4879,
+      "step": 8929
+    },
+    {
+      "epoch": 0.6200527704485488,
+      "grad_norm": 3.340480594932884,
+      "learning_rate": 3.3318225553600582e-06,
+      "loss": 0.3196,
+      "step": 8930
+    },
+    {
+      "epoch": 0.620122205249271,
+      "grad_norm": 3.3354803777115496,
+      "learning_rate": 3.3307625409390533e-06,
+      "loss": 0.3078,
+      "step": 8931
+    },
+    {
+      "epoch": 0.620191640049993,
+      "grad_norm": 6.136814589448868,
+      "learning_rate": 3.3297026109461896e-06,
+      "loss": 0.7291,
+      "step": 8932
+    },
+    {
+      "epoch": 0.6202610748507151,
+      "grad_norm": 4.177866688371145,
+      "learning_rate": 3.32864276543508e-06,
+      "loss": 0.4478,
+      "step": 8933
+    },
+    {
+      "epoch": 0.6203305096514373,
+      "grad_norm": 3.426458910265748,
+      "learning_rate": 3.3275830044593272e-06,
+      "loss": 0.3917,
+      "step": 8934
+    },
+    {
+      "epoch": 0.6203999444521594,
+      "grad_norm": 3.78701457833112,
+      "learning_rate": 3.3265233280725325e-06,
+      "loss": 0.6512,
+      "step": 8935
+    },
+    {
+      "epoch": 0.6204693792528816,
+      "grad_norm": 4.1605233376360635,
+      "learning_rate": 3.325463736328295e-06,
+      "loss": 0.5566,
+      "step": 8936
+    },
+    {
+      "epoch": 0.6205388140536037,
+      "grad_norm": 4.102994492402127,
+      "learning_rate": 3.324404229280208e-06,
+      "loss": 0.321,
+      "step": 8937
+    },
+    {
+      "epoch": 0.6206082488543258,
+      "grad_norm": 3.1760675064732777,
+      "learning_rate": 3.323344806981857e-06,
+      "loss": 0.4112,
+      "step": 8938
+    },
+    {
+      "epoch": 0.6206776836550479,
+      "grad_norm": 3.686598083025207,
+      "learning_rate": 3.322285469486829e-06,
+      "loss": 0.4862,
+      "step": 8939
+    },
+    {
+      "epoch": 0.62074711845577,
+      "grad_norm": 3.992048391250304,
+      "learning_rate": 3.3212262168487053e-06,
+      "loss": 0.4923,
+      "step": 8940
+    },
+    {
+      "epoch": 0.6208165532564921,
+      "grad_norm": 4.395638780436744,
+      "learning_rate": 3.320167049121057e-06,
+      "loss": 0.464,
+      "step": 8941
+    },
+    {
+      "epoch": 0.6208859880572143,
+      "grad_norm": 4.463266375264487,
+      "learning_rate": 3.319107966357459e-06,
+      "loss": 0.5432,
+      "step": 8942
+    },
+    {
+      "epoch": 0.6209554228579364,
+      "grad_norm": 3.835221332881704,
+      "learning_rate": 3.318048968611479e-06,
+      "loss": 0.5147,
+      "step": 8943
+    },
+    {
+      "epoch": 0.6210248576586586,
+      "grad_norm": 4.9437114297164895,
+      "learning_rate": 3.3169900559366797e-06,
+      "loss": 0.5789,
+      "step": 8944
+    },
+    {
+      "epoch": 0.6210942924593806,
+      "grad_norm": 5.446543230435596,
+      "learning_rate": 3.3159312283866173e-06,
+      "loss": 0.5016,
+      "step": 8945
+    },
+    {
+      "epoch": 0.6211637272601027,
+      "grad_norm": 3.331033647084621,
+      "learning_rate": 3.3148724860148486e-06,
+      "loss": 0.2522,
+      "step": 8946
+    },
+    {
+      "epoch": 0.6212331620608249,
+      "grad_norm": 4.446017173800562,
+      "learning_rate": 3.3138138288749233e-06,
+      "loss": 0.4903,
+      "step": 8947
+    },
+    {
+      "epoch": 0.621302596861547,
+      "grad_norm": 3.9759594698348404,
+      "learning_rate": 3.3127552570203848e-06,
+      "loss": 0.4596,
+      "step": 8948
+    },
+    {
+      "epoch": 0.6213720316622692,
+      "grad_norm": 3.4766982739504386,
+      "learning_rate": 3.3116967705047778e-06,
+      "loss": 0.5102,
+      "step": 8949
+    },
+    {
+      "epoch": 0.6214414664629913,
+      "grad_norm": 3.108054077664294,
+      "learning_rate": 3.3106383693816356e-06,
+      "loss": 0.2278,
+      "step": 8950
+    },
+    {
+      "epoch": 0.6215109012637133,
+      "grad_norm": 2.81017797846336,
+      "learning_rate": 3.3095800537044955e-06,
+      "loss": 0.2269,
+      "step": 8951
+    },
+    {
+      "epoch": 0.6215803360644355,
+      "grad_norm": 11.29337652881421,
+      "learning_rate": 3.3085218235268814e-06,
+      "loss": 0.4081,
+      "step": 8952
+    },
+    {
+      "epoch": 0.6216497708651576,
+      "grad_norm": 4.269991170430919,
+      "learning_rate": 3.3074636789023207e-06,
+      "loss": 0.6077,
+      "step": 8953
+    },
+    {
+      "epoch": 0.6217192056658798,
+      "grad_norm": 3.878249165013554,
+      "learning_rate": 3.3064056198843324e-06,
+      "loss": 0.4451,
+      "step": 8954
+    },
+    {
+      "epoch": 0.6217886404666019,
+      "grad_norm": 3.495601004928863,
+      "learning_rate": 3.30534764652643e-06,
+      "loss": 0.3896,
+      "step": 8955
+    },
+    {
+      "epoch": 0.621858075267324,
+      "grad_norm": 4.117749286960358,
+      "learning_rate": 3.3042897588821266e-06,
+      "loss": 0.3503,
+      "step": 8956
+    },
+    {
+      "epoch": 0.6219275100680461,
+      "grad_norm": 2.9455276888058735,
+      "learning_rate": 3.303231957004928e-06,
+      "loss": 0.3606,
+      "step": 8957
+    },
+    {
+      "epoch": 0.6219969448687682,
+      "grad_norm": 3.576283646783539,
+      "learning_rate": 3.3021742409483377e-06,
+      "loss": 0.2957,
+      "step": 8958
+    },
+    {
+      "epoch": 0.6220663796694903,
+      "grad_norm": 3.8510852096967017,
+      "learning_rate": 3.3011166107658533e-06,
+      "loss": 0.4388,
+      "step": 8959
+    },
+    {
+      "epoch": 0.6221358144702125,
+      "grad_norm": 3.5140867863313914,
+      "learning_rate": 3.3000590665109666e-06,
+      "loss": 0.396,
+      "step": 8960
+    },
+    {
+      "epoch": 0.6222052492709346,
+      "grad_norm": 5.053596372067923,
+      "learning_rate": 3.299001608237169e-06,
+      "loss": 0.5821,
+      "step": 8961
+    },
+    {
+      "epoch": 0.6222746840716568,
+      "grad_norm": 4.194408828184071,
+      "learning_rate": 3.297944235997947e-06,
+      "loss": 0.4454,
+      "step": 8962
+    },
+    {
+      "epoch": 0.6223441188723788,
+      "grad_norm": 3.794174361042864,
+      "learning_rate": 3.296886949846778e-06,
+      "loss": 0.3753,
+      "step": 8963
+    },
+    {
+      "epoch": 0.6224135536731009,
+      "grad_norm": 3.325305633369674,
+      "learning_rate": 3.29582974983714e-06,
+      "loss": 0.3717,
+      "step": 8964
+    },
+    {
+      "epoch": 0.6224829884738231,
+      "grad_norm": 4.097661349877509,
+      "learning_rate": 3.294772636022506e-06,
+      "loss": 0.31,
+      "step": 8965
+    },
+    {
+      "epoch": 0.6225524232745452,
+      "grad_norm": 3.7101726783391054,
+      "learning_rate": 3.2937156084563426e-06,
+      "loss": 0.434,
+      "step": 8966
+    },
+    {
+      "epoch": 0.6226218580752674,
+      "grad_norm": 4.678320279816437,
+      "learning_rate": 3.2926586671921112e-06,
+      "loss": 0.5411,
+      "step": 8967
+    },
+    {
+      "epoch": 0.6226912928759895,
+      "grad_norm": 2.767381312053914,
+      "learning_rate": 3.2916018122832734e-06,
+      "loss": 0.1098,
+      "step": 8968
+    },
+    {
+      "epoch": 0.6227607276767115,
+      "grad_norm": 3.5352557440866175,
+      "learning_rate": 3.2905450437832837e-06,
+      "loss": 0.3617,
+      "step": 8969
+    },
+    {
+      "epoch": 0.6228301624774337,
+      "grad_norm": 5.82755823278559,
+      "learning_rate": 3.2894883617455898e-06,
+      "loss": 0.6603,
+      "step": 8970
+    },
+    {
+      "epoch": 0.6228995972781558,
+      "grad_norm": 3.9041101833879077,
+      "learning_rate": 3.2884317662236383e-06,
+      "loss": 0.4414,
+      "step": 8971
+    },
+    {
+      "epoch": 0.6229690320788779,
+      "grad_norm": 4.004251511431086,
+      "learning_rate": 3.287375257270874e-06,
+      "loss": 0.5276,
+      "step": 8972
+    },
+    {
+      "epoch": 0.6230384668796001,
+      "grad_norm": 4.841416363184572,
+      "learning_rate": 3.2863188349407293e-06,
+      "loss": 0.6096,
+      "step": 8973
+    },
+    {
+      "epoch": 0.6231079016803222,
+      "grad_norm": 3.3382927605081996,
+      "learning_rate": 3.285262499286638e-06,
+      "loss": 0.3478,
+      "step": 8974
+    },
+    {
+      "epoch": 0.6231773364810443,
+      "grad_norm": 4.318534336309797,
+      "learning_rate": 3.284206250362031e-06,
+      "loss": 0.5535,
+      "step": 8975
+    },
+    {
+      "epoch": 0.6232467712817664,
+      "grad_norm": 4.280109432267604,
+      "learning_rate": 3.2831500882203306e-06,
+      "loss": 0.3308,
+      "step": 8976
+    },
+    {
+      "epoch": 0.6233162060824885,
+      "grad_norm": 4.7821423933146034,
+      "learning_rate": 3.2820940129149537e-06,
+      "loss": 0.6301,
+      "step": 8977
+    },
+    {
+      "epoch": 0.6233856408832107,
+      "grad_norm": 3.7154703456734093,
+      "learning_rate": 3.2810380244993194e-06,
+      "loss": 0.4113,
+      "step": 8978
+    },
+    {
+      "epoch": 0.6234550756839328,
+      "grad_norm": 4.156670549543116,
+      "learning_rate": 3.2799821230268374e-06,
+      "loss": 0.5935,
+      "step": 8979
+    },
+    {
+      "epoch": 0.623524510484655,
+      "grad_norm": 4.04397332128435,
+      "learning_rate": 3.2789263085509117e-06,
+      "loss": 0.4655,
+      "step": 8980
+    },
+    {
+      "epoch": 0.623593945285377,
+      "grad_norm": 3.919783901319768,
+      "learning_rate": 3.2778705811249457e-06,
+      "loss": 0.4802,
+      "step": 8981
+    },
+    {
+      "epoch": 0.6236633800860991,
+      "grad_norm": 4.240432975680613,
+      "learning_rate": 3.276814940802337e-06,
+      "loss": 0.5088,
+      "step": 8982
+    },
+    {
+      "epoch": 0.6237328148868213,
+      "grad_norm": 4.437544004404756,
+      "learning_rate": 3.2757593876364802e-06,
+      "loss": 0.5367,
+      "step": 8983
+    },
+    {
+      "epoch": 0.6238022496875434,
+      "grad_norm": 4.421928936623407,
+      "learning_rate": 3.2747039216807607e-06,
+      "loss": 0.5654,
+      "step": 8984
+    },
+    {
+      "epoch": 0.6238716844882655,
+      "grad_norm": 4.09492559396691,
+      "learning_rate": 3.273648542988566e-06,
+      "loss": 0.5398,
+      "step": 8985
+    },
+    {
+      "epoch": 0.6239411192889877,
+      "grad_norm": 3.3251721284795224,
+      "learning_rate": 3.2725932516132718e-06,
+      "loss": 0.3802,
+      "step": 8986
+    },
+    {
+      "epoch": 0.6240105540897097,
+      "grad_norm": 4.0172754100899715,
+      "learning_rate": 3.27153804760826e-06,
+      "loss": 0.4968,
+      "step": 8987
+    },
+    {
+      "epoch": 0.6240799888904319,
+      "grad_norm": 2.9831868573379277,
+      "learning_rate": 3.270482931026896e-06,
+      "loss": 0.337,
+      "step": 8988
+    },
+    {
+      "epoch": 0.624149423691154,
+      "grad_norm": 4.4904188958483635,
+      "learning_rate": 3.269427901922547e-06,
+      "loss": 0.4517,
+      "step": 8989
+    },
+    {
+      "epoch": 0.6242188584918761,
+      "grad_norm": 4.630562921268973,
+      "learning_rate": 3.268372960348579e-06,
+      "loss": 0.5792,
+      "step": 8990
+    },
+    {
+      "epoch": 0.6242882932925983,
+      "grad_norm": 4.0097492103052605,
+      "learning_rate": 3.2673181063583465e-06,
+      "loss": 0.3977,
+      "step": 8991
+    },
+    {
+      "epoch": 0.6243577280933204,
+      "grad_norm": 3.9674126128040546,
+      "learning_rate": 3.2662633400052024e-06,
+      "loss": 0.415,
+      "step": 8992
+    },
+    {
+      "epoch": 0.6244271628940425,
+      "grad_norm": 4.571299549606146,
+      "learning_rate": 3.265208661342496e-06,
+      "loss": 0.7069,
+      "step": 8993
+    },
+    {
+      "epoch": 0.6244965976947646,
+      "grad_norm": 2.682808453270179,
+      "learning_rate": 3.2641540704235754e-06,
+      "loss": 0.1882,
+      "step": 8994
+    },
+    {
+      "epoch": 0.6245660324954867,
+      "grad_norm": 3.9389360522043355,
+      "learning_rate": 3.263099567301776e-06,
+      "loss": 0.3573,
+      "step": 8995
+    },
+    {
+      "epoch": 0.6246354672962089,
+      "grad_norm": 3.36854396165017,
+      "learning_rate": 3.262045152030434e-06,
+      "loss": 0.36,
+      "step": 8996
+    },
+    {
+      "epoch": 0.624704902096931,
+      "grad_norm": 3.7646545474297173,
+      "learning_rate": 3.260990824662884e-06,
+      "loss": 0.3665,
+      "step": 8997
+    },
+    {
+      "epoch": 0.624774336897653,
+      "grad_norm": 4.319426677523966,
+      "learning_rate": 3.259936585252449e-06,
+      "loss": 0.4495,
+      "step": 8998
+    },
+    {
+      "epoch": 0.6248437716983752,
+      "grad_norm": 6.785832555900826,
+      "learning_rate": 3.258882433852452e-06,
+      "loss": 0.5352,
+      "step": 8999
+    },
+    {
+      "epoch": 0.6249132064990973,
+      "grad_norm": 4.204417482643048,
+      "learning_rate": 3.2578283705162117e-06,
+      "loss": 0.4865,
+      "step": 9000
+    },
+    {
+      "epoch": 0.6249826412998195,
+      "grad_norm": 2.6676647165279617,
+      "learning_rate": 3.256774395297042e-06,
+      "loss": 0.179,
+      "step": 9001
+    },
+    {
+      "epoch": 0.6250520761005416,
+      "grad_norm": 3.5890064653683136,
+      "learning_rate": 3.255720508248249e-06,
+      "loss": 0.3673,
+      "step": 9002
+    },
+    {
+      "epoch": 0.6251215109012637,
+      "grad_norm": 4.192944739343706,
+      "learning_rate": 3.254666709423139e-06,
+      "loss": 0.3802,
+      "step": 9003
+    },
+    {
+      "epoch": 0.6251909457019859,
+      "grad_norm": 4.441264127308771,
+      "learning_rate": 3.253612998875014e-06,
+      "loss": 0.4433,
+      "step": 9004
+    },
+    {
+      "epoch": 0.6252603805027079,
+      "grad_norm": 4.494084642904855,
+      "learning_rate": 3.2525593766571663e-06,
+      "loss": 0.4904,
+      "step": 9005
+    },
+    {
+      "epoch": 0.6253298153034301,
+      "grad_norm": 4.408331407940073,
+      "learning_rate": 3.251505842822886e-06,
+      "loss": 0.4434,
+      "step": 9006
+    },
+    {
+      "epoch": 0.6253992501041522,
+      "grad_norm": 4.041272252944142,
+      "learning_rate": 3.250452397425464e-06,
+      "loss": 0.393,
+      "step": 9007
+    },
+    {
+      "epoch": 0.6254686849048743,
+      "grad_norm": 4.011531290242249,
+      "learning_rate": 3.2493990405181797e-06,
+      "loss": 0.481,
+      "step": 9008
+    },
+    {
+      "epoch": 0.6255381197055965,
+      "grad_norm": 4.2362711664151425,
+      "learning_rate": 3.24834577215431e-06,
+      "loss": 0.6241,
+      "step": 9009
+    },
+    {
+      "epoch": 0.6256075545063186,
+      "grad_norm": 4.278327500342482,
+      "learning_rate": 3.2472925923871297e-06,
+      "loss": 0.5626,
+      "step": 9010
+    },
+    {
+      "epoch": 0.6256769893070407,
+      "grad_norm": 4.227086493343128,
+      "learning_rate": 3.2462395012699053e-06,
+      "loss": 0.3774,
+      "step": 9011
+    },
+    {
+      "epoch": 0.6257464241077628,
+      "grad_norm": 4.516249183747287,
+      "learning_rate": 3.245186498855906e-06,
+      "loss": 0.6426,
+      "step": 9012
+    },
+    {
+      "epoch": 0.6258158589084849,
+      "grad_norm": 3.097392369470254,
+      "learning_rate": 3.244133585198384e-06,
+      "loss": 0.2213,
+      "step": 9013
+    },
+    {
+      "epoch": 0.6258852937092071,
+      "grad_norm": 4.331486255058671,
+      "learning_rate": 3.2430807603505996e-06,
+      "loss": 0.557,
+      "step": 9014
+    },
+    {
+      "epoch": 0.6259547285099292,
+      "grad_norm": 3.875794829016102,
+      "learning_rate": 3.2420280243658044e-06,
+      "loss": 0.3534,
+      "step": 9015
+    },
+    {
+      "epoch": 0.6260241633106512,
+      "grad_norm": 4.148788326155634,
+      "learning_rate": 3.24097537729724e-06,
+      "loss": 0.3867,
+      "step": 9016
+    },
+    {
+      "epoch": 0.6260935981113734,
+      "grad_norm": 4.837398845783604,
+      "learning_rate": 3.2399228191981515e-06,
+      "loss": 0.57,
+      "step": 9017
+    },
+    {
+      "epoch": 0.6261630329120955,
+      "grad_norm": 3.8692866917323063,
+      "learning_rate": 3.238870350121773e-06,
+      "loss": 0.59,
+      "step": 9018
+    },
+    {
+      "epoch": 0.6262324677128177,
+      "grad_norm": 3.2516628088330526,
+      "learning_rate": 3.237817970121343e-06,
+      "loss": 0.2857,
+      "step": 9019
+    },
+    {
+      "epoch": 0.6263019025135398,
+      "grad_norm": 7.595626555877067,
+      "learning_rate": 3.236765679250084e-06,
+      "loss": 0.21,
+      "step": 9020
+    },
+    {
+      "epoch": 0.6263713373142619,
+      "grad_norm": 3.8447979977921207,
+      "learning_rate": 3.235713477561221e-06,
+      "loss": 0.4557,
+      "step": 9021
+    },
+    {
+      "epoch": 0.6264407721149841,
+      "grad_norm": 2.890888777142742,
+      "learning_rate": 3.234661365107976e-06,
+      "loss": 0.3525,
+      "step": 9022
+    },
+    {
+      "epoch": 0.6265102069157061,
+      "grad_norm": 3.815185239918401,
+      "learning_rate": 3.2336093419435597e-06,
+      "loss": 0.51,
+      "step": 9023
+    },
+    {
+      "epoch": 0.6265796417164283,
+      "grad_norm": 4.027388270659074,
+      "learning_rate": 3.232557408121185e-06,
+      "loss": 0.4839,
+      "step": 9024
+    },
+    {
+      "epoch": 0.6266490765171504,
+      "grad_norm": 3.2612492799509827,
+      "learning_rate": 3.2315055636940553e-06,
+      "loss": 0.2954,
+      "step": 9025
+    },
+    {
+      "epoch": 0.6267185113178725,
+      "grad_norm": 3.354474441164539,
+      "learning_rate": 3.2304538087153748e-06,
+      "loss": 0.3099,
+      "step": 9026
+    },
+    {
+      "epoch": 0.6267879461185947,
+      "grad_norm": 3.8525876880102587,
+      "learning_rate": 3.2294021432383375e-06,
+      "loss": 0.4962,
+      "step": 9027
+    },
+    {
+      "epoch": 0.6268573809193168,
+      "grad_norm": 3.8854912035547637,
+      "learning_rate": 3.2283505673161354e-06,
+      "loss": 0.5525,
+      "step": 9028
+    },
+    {
+      "epoch": 0.6269268157200388,
+      "grad_norm": 4.008680604361545,
+      "learning_rate": 3.2272990810019587e-06,
+      "loss": 0.4658,
+      "step": 9029
+    },
+    {
+      "epoch": 0.626996250520761,
+      "grad_norm": 3.2660675058326194,
+      "learning_rate": 3.2262476843489876e-06,
+      "loss": 0.3845,
+      "step": 9030
+    },
+    {
+      "epoch": 0.6270656853214831,
+      "grad_norm": 3.4016747154614007,
+      "learning_rate": 3.225196377410401e-06,
+      "loss": 0.3127,
+      "step": 9031
+    },
+    {
+      "epoch": 0.6271351201222053,
+      "grad_norm": 3.7010499987079646,
+      "learning_rate": 3.2241451602393725e-06,
+      "loss": 0.4047,
+      "step": 9032
+    },
+    {
+      "epoch": 0.6272045549229274,
+      "grad_norm": 4.133439477857955,
+      "learning_rate": 3.2230940328890746e-06,
+      "loss": 0.4985,
+      "step": 9033
+    },
+    {
+      "epoch": 0.6272739897236494,
+      "grad_norm": 3.633030554670607,
+      "learning_rate": 3.222042995412669e-06,
+      "loss": 0.415,
+      "step": 9034
+    },
+    {
+      "epoch": 0.6273434245243716,
+      "grad_norm": 3.8941318514447887,
+      "learning_rate": 3.2209920478633157e-06,
+      "loss": 0.4426,
+      "step": 9035
+    },
+    {
+      "epoch": 0.6274128593250937,
+      "grad_norm": 4.355914523541789,
+      "learning_rate": 3.2199411902941724e-06,
+      "loss": 0.5264,
+      "step": 9036
+    },
+    {
+      "epoch": 0.6274822941258159,
+      "grad_norm": 3.66399642635389,
+      "learning_rate": 3.2188904227583905e-06,
+      "loss": 0.439,
+      "step": 9037
+    },
+    {
+      "epoch": 0.627551728926538,
+      "grad_norm": 5.7953863089403725,
+      "learning_rate": 3.2178397453091137e-06,
+      "loss": 0.7627,
+      "step": 9038
+    },
+    {
+      "epoch": 0.6276211637272601,
+      "grad_norm": 8.762645532771407,
+      "learning_rate": 3.216789157999487e-06,
+      "loss": 0.4795,
+      "step": 9039
+    },
+    {
+      "epoch": 0.6276905985279823,
+      "grad_norm": 4.210207491433881,
+      "learning_rate": 3.2157386608826473e-06,
+      "loss": 0.5654,
+      "step": 9040
+    },
+    {
+      "epoch": 0.6277600333287043,
+      "grad_norm": 4.32036350155374,
+      "learning_rate": 3.2146882540117253e-06,
+      "loss": 0.6372,
+      "step": 9041
+    },
+    {
+      "epoch": 0.6278294681294264,
+      "grad_norm": 4.439258143489007,
+      "learning_rate": 3.2136379374398514e-06,
+      "loss": 0.5532,
+      "step": 9042
+    },
+    {
+      "epoch": 0.6278989029301486,
+      "grad_norm": 4.320433934369404,
+      "learning_rate": 3.2125877112201497e-06,
+      "loss": 0.4743,
+      "step": 9043
+    },
+    {
+      "epoch": 0.6279683377308707,
+      "grad_norm": 3.8006062649271923,
+      "learning_rate": 3.2115375754057408e-06,
+      "loss": 0.5284,
+      "step": 9044
+    },
+    {
+      "epoch": 0.6280377725315929,
+      "grad_norm": 3.8265181132884716,
+      "learning_rate": 3.210487530049735e-06,
+      "loss": 0.3664,
+      "step": 9045
+    },
+    {
+      "epoch": 0.628107207332315,
+      "grad_norm": 3.5485285900860233,
+      "learning_rate": 3.2094375752052464e-06,
+      "loss": 0.4097,
+      "step": 9046
+    },
+    {
+      "epoch": 0.628176642133037,
+      "grad_norm": 2.8807486986276047,
+      "learning_rate": 3.20838771092538e-06,
+      "loss": 0.3151,
+      "step": 9047
+    },
+    {
+      "epoch": 0.6282460769337592,
+      "grad_norm": 4.118234514357313,
+      "learning_rate": 3.207337937263233e-06,
+      "loss": 0.4802,
+      "step": 9048
+    },
+    {
+      "epoch": 0.6283155117344813,
+      "grad_norm": 4.75502354866508,
+      "learning_rate": 3.2062882542719065e-06,
+      "loss": 0.7166,
+      "step": 9049
+    },
+    {
+      "epoch": 0.6283849465352035,
+      "grad_norm": 5.477466643250541,
+      "learning_rate": 3.2052386620044894e-06,
+      "loss": 0.56,
+      "step": 9050
+    },
+    {
+      "epoch": 0.6284543813359256,
+      "grad_norm": 3.3067271582116127,
+      "learning_rate": 3.2041891605140717e-06,
+      "loss": 0.3326,
+      "step": 9051
+    },
+    {
+      "epoch": 0.6285238161366477,
+      "grad_norm": 4.131545396951763,
+      "learning_rate": 3.203139749853732e-06,
+      "loss": 0.4406,
+      "step": 9052
+    },
+    {
+      "epoch": 0.6285932509373698,
+      "grad_norm": 4.2597589680827985,
+      "learning_rate": 3.202090430076551e-06,
+      "loss": 0.4907,
+      "step": 9053
+    },
+    {
+      "epoch": 0.6286626857380919,
+      "grad_norm": 4.001488334879843,
+      "learning_rate": 3.2010412012356033e-06,
+      "loss": 0.3803,
+      "step": 9054
+    },
+    {
+      "epoch": 0.628732120538814,
+      "grad_norm": 4.383751412856482,
+      "learning_rate": 3.1999920633839537e-06,
+      "loss": 0.5641,
+      "step": 9055
+    },
+    {
+      "epoch": 0.6288015553395362,
+      "grad_norm": 3.879851740950923,
+      "learning_rate": 3.1989430165746697e-06,
+      "loss": 0.4171,
+      "step": 9056
+    },
+    {
+      "epoch": 0.6288709901402583,
+      "grad_norm": 4.048561836292851,
+      "learning_rate": 3.197894060860809e-06,
+      "loss": 0.519,
+      "step": 9057
+    },
+    {
+      "epoch": 0.6289404249409805,
+      "grad_norm": 5.145456084574193,
+      "learning_rate": 3.1968451962954295e-06,
+      "loss": 0.593,
+      "step": 9058
+    },
+    {
+      "epoch": 0.6290098597417025,
+      "grad_norm": 3.612395344716014,
+      "learning_rate": 3.1957964229315786e-06,
+      "loss": 0.3915,
+      "step": 9059
+    },
+    {
+      "epoch": 0.6290792945424246,
+      "grad_norm": 3.908662020871548,
+      "learning_rate": 3.1947477408223018e-06,
+      "loss": 0.3766,
+      "step": 9060
+    },
+    {
+      "epoch": 0.6291487293431468,
+      "grad_norm": 2.454000261762465,
+      "learning_rate": 3.1936991500206427e-06,
+      "loss": 0.2675,
+      "step": 9061
+    },
+    {
+      "epoch": 0.6292181641438689,
+      "grad_norm": 4.741723339937936,
+      "learning_rate": 3.1926506505796372e-06,
+      "loss": 0.556,
+      "step": 9062
+    },
+    {
+      "epoch": 0.6292875989445911,
+      "grad_norm": 3.70455977897339,
+      "learning_rate": 3.1916022425523154e-06,
+      "loss": 0.2425,
+      "step": 9063
+    },
+    {
+      "epoch": 0.6293570337453132,
+      "grad_norm": 3.3076529339459824,
+      "learning_rate": 3.190553925991705e-06,
+      "loss": 0.3268,
+      "step": 9064
+    },
+    {
+      "epoch": 0.6294264685460352,
+      "grad_norm": 3.2657091345156104,
+      "learning_rate": 3.1895057009508322e-06,
+      "loss": 0.464,
+      "step": 9065
+    },
+    {
+      "epoch": 0.6294959033467574,
+      "grad_norm": 3.4743753174772327,
+      "learning_rate": 3.1884575674827113e-06,
+      "loss": 0.5878,
+      "step": 9066
+    },
+    {
+      "epoch": 0.6295653381474795,
+      "grad_norm": 3.9985169685032935,
+      "learning_rate": 3.1874095256403547e-06,
+      "loss": 0.3798,
+      "step": 9067
+    },
+    {
+      "epoch": 0.6296347729482017,
+      "grad_norm": 5.74882975098507,
+      "learning_rate": 3.186361575476775e-06,
+      "loss": 0.6254,
+      "step": 9068
+    },
+    {
+      "epoch": 0.6297042077489238,
+      "grad_norm": 3.70218517709707,
+      "learning_rate": 3.185313717044976e-06,
+      "loss": 0.5777,
+      "step": 9069
+    },
+    {
+      "epoch": 0.6297736425496459,
+      "grad_norm": 3.8857262609012797,
+      "learning_rate": 3.1842659503979535e-06,
+      "loss": 0.497,
+      "step": 9070
+    },
+    {
+      "epoch": 0.629843077350368,
+      "grad_norm": 3.115277029655286,
+      "learning_rate": 3.1832182755887055e-06,
+      "loss": 0.3111,
+      "step": 9071
+    },
+    {
+      "epoch": 0.6299125121510901,
+      "grad_norm": 3.431121871853508,
+      "learning_rate": 3.182170692670222e-06,
+      "loss": 0.4478,
+      "step": 9072
+    },
+    {
+      "epoch": 0.6299819469518122,
+      "grad_norm": 3.513610449149736,
+      "learning_rate": 3.1811232016954857e-06,
+      "loss": 0.3327,
+      "step": 9073
+    },
+    {
+      "epoch": 0.6300513817525344,
+      "grad_norm": 2.3319133734152198,
+      "learning_rate": 3.18007580271748e-06,
+      "loss": 0.1969,
+      "step": 9074
+    },
+    {
+      "epoch": 0.6301208165532565,
+      "grad_norm": 3.4960578268544795,
+      "learning_rate": 3.1790284957891815e-06,
+      "loss": 0.5259,
+      "step": 9075
+    },
+    {
+      "epoch": 0.6301902513539787,
+      "grad_norm": 4.784284808198029,
+      "learning_rate": 3.1779812809635624e-06,
+      "loss": 0.3209,
+      "step": 9076
+    },
+    {
+      "epoch": 0.6302596861547007,
+      "grad_norm": 4.012218277341716,
+      "learning_rate": 3.1769341582935863e-06,
+      "loss": 0.3555,
+      "step": 9077
+    },
+    {
+      "epoch": 0.6303291209554228,
+      "grad_norm": 4.446047199406678,
+      "learning_rate": 3.175887127832219e-06,
+      "loss": 0.6318,
+      "step": 9078
+    },
+    {
+      "epoch": 0.630398555756145,
+      "grad_norm": 3.0163621753875995,
+      "learning_rate": 3.1748401896324173e-06,
+      "loss": 0.3288,
+      "step": 9079
+    },
+    {
+      "epoch": 0.6304679905568671,
+      "grad_norm": 2.742546354995392,
+      "learning_rate": 3.1737933437471314e-06,
+      "loss": 0.2384,
+      "step": 9080
+    },
+    {
+      "epoch": 0.6305374253575893,
+      "grad_norm": 3.6768064090314354,
+      "learning_rate": 3.1727465902293125e-06,
+      "loss": 0.4935,
+      "step": 9081
+    },
+    {
+      "epoch": 0.6306068601583114,
+      "grad_norm": 2.7434549600501414,
+      "learning_rate": 3.1716999291319026e-06,
+      "loss": 0.1034,
+      "step": 9082
+    },
+    {
+      "epoch": 0.6306762949590334,
+      "grad_norm": 3.800859727663394,
+      "learning_rate": 3.170653360507843e-06,
+      "loss": 0.3391,
+      "step": 9083
+    },
+    {
+      "epoch": 0.6307457297597556,
+      "grad_norm": 2.6726134812277538,
+      "learning_rate": 3.169606884410065e-06,
+      "loss": 0.22,
+      "step": 9084
+    },
+    {
+      "epoch": 0.6308151645604777,
+      "grad_norm": 4.231115437929923,
+      "learning_rate": 3.1685605008915004e-06,
+      "loss": 0.3776,
+      "step": 9085
+    },
+    {
+      "epoch": 0.6308845993611998,
+      "grad_norm": 3.826693861310103,
+      "learning_rate": 3.1675142100050744e-06,
+      "loss": 0.3184,
+      "step": 9086
+    },
+    {
+      "epoch": 0.630954034161922,
+      "grad_norm": 3.663681085753831,
+      "learning_rate": 3.166468011803704e-06,
+      "loss": 0.4207,
+      "step": 9087
+    },
+    {
+      "epoch": 0.631023468962644,
+      "grad_norm": 3.375147307648443,
+      "learning_rate": 3.1654219063403076e-06,
+      "loss": 0.3182,
+      "step": 9088
+    },
+    {
+      "epoch": 0.6310929037633662,
+      "grad_norm": 3.736604715097471,
+      "learning_rate": 3.164375893667794e-06,
+      "loss": 0.3128,
+      "step": 9089
+    },
+    {
+      "epoch": 0.6311623385640883,
+      "grad_norm": 3.73312822649804,
+      "learning_rate": 3.1633299738390726e-06,
+      "loss": 0.3098,
+      "step": 9090
+    },
+    {
+      "epoch": 0.6312317733648104,
+      "grad_norm": 2.7690171891500723,
+      "learning_rate": 3.162284146907042e-06,
+      "loss": 0.3444,
+      "step": 9091
+    },
+    {
+      "epoch": 0.6313012081655326,
+      "grad_norm": 4.18514692708291,
+      "learning_rate": 3.1612384129245986e-06,
+      "loss": 0.562,
+      "step": 9092
+    },
+    {
+      "epoch": 0.6313706429662547,
+      "grad_norm": 3.382995159227156,
+      "learning_rate": 3.1601927719446357e-06,
+      "loss": 0.367,
+      "step": 9093
+    },
+    {
+      "epoch": 0.6314400777669769,
+      "grad_norm": 3.1906500266927265,
+      "learning_rate": 3.1591472240200427e-06,
+      "loss": 0.2477,
+      "step": 9094
+    },
+    {
+      "epoch": 0.6315095125676989,
+      "grad_norm": 4.416880471462263,
+      "learning_rate": 3.1581017692036986e-06,
+      "loss": 0.5047,
+      "step": 9095
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 2.893121131467638,
+      "learning_rate": 3.1570564075484824e-06,
+      "loss": 0.3209,
+      "step": 9096
+    },
+    {
+      "epoch": 0.6316483821691432,
+      "grad_norm": 3.8434697881373787,
+      "learning_rate": 3.156011139107269e-06,
+      "loss": 0.5578,
+      "step": 9097
+    },
+    {
+      "epoch": 0.6317178169698653,
+      "grad_norm": 4.863409223416476,
+      "learning_rate": 3.1549659639329256e-06,
+      "loss": 0.6212,
+      "step": 9098
+    },
+    {
+      "epoch": 0.6317872517705874,
+      "grad_norm": 3.3468939151848414,
+      "learning_rate": 3.153920882078314e-06,
+      "loss": 0.5158,
+      "step": 9099
+    },
+    {
+      "epoch": 0.6318566865713096,
+      "grad_norm": 3.7067201692810814,
+      "learning_rate": 3.152875893596296e-06,
+      "loss": 0.4897,
+      "step": 9100
+    },
+    {
+      "epoch": 0.6319261213720316,
+      "grad_norm": 3.1456197258398535,
+      "learning_rate": 3.1518309985397265e-06,
+      "loss": 0.3053,
+      "step": 9101
+    },
+    {
+      "epoch": 0.6319955561727538,
+      "grad_norm": 3.9198227091925153,
+      "learning_rate": 3.1507861969614517e-06,
+      "loss": 0.4363,
+      "step": 9102
+    },
+    {
+      "epoch": 0.6320649909734759,
+      "grad_norm": 5.044639932319711,
+      "learning_rate": 3.149741488914319e-06,
+      "loss": 0.5066,
+      "step": 9103
+    },
+    {
+      "epoch": 0.632134425774198,
+      "grad_norm": 3.805464765361104,
+      "learning_rate": 3.14869687445117e-06,
+      "loss": 0.3658,
+      "step": 9104
+    },
+    {
+      "epoch": 0.6322038605749202,
+      "grad_norm": 4.527840748366383,
+      "learning_rate": 3.1476523536248365e-06,
+      "loss": 0.4906,
+      "step": 9105
+    },
+    {
+      "epoch": 0.6322732953756423,
+      "grad_norm": 3.700763323443944,
+      "learning_rate": 3.1466079264881504e-06,
+      "loss": 0.4415,
+      "step": 9106
+    },
+    {
+      "epoch": 0.6323427301763644,
+      "grad_norm": 3.7476152299160486,
+      "learning_rate": 3.145563593093939e-06,
+      "loss": 0.3267,
+      "step": 9107
+    },
+    {
+      "epoch": 0.6324121649770865,
+      "grad_norm": 4.405255047935883,
+      "learning_rate": 3.1445193534950238e-06,
+      "loss": 0.4081,
+      "step": 9108
+    },
+    {
+      "epoch": 0.6324815997778086,
+      "grad_norm": 3.126115439796439,
+      "learning_rate": 3.1434752077442176e-06,
+      "loss": 0.2721,
+      "step": 9109
+    },
+    {
+      "epoch": 0.6325510345785308,
+      "grad_norm": 3.9276357009718357,
+      "learning_rate": 3.142431155894336e-06,
+      "loss": 0.3816,
+      "step": 9110
+    },
+    {
+      "epoch": 0.6326204693792529,
+      "grad_norm": 3.5783861664470953,
+      "learning_rate": 3.1413871979981857e-06,
+      "loss": 0.4196,
+      "step": 9111
+    },
+    {
+      "epoch": 0.632689904179975,
+      "grad_norm": 3.6296268359062283,
+      "learning_rate": 3.1403433341085656e-06,
+      "loss": 0.422,
+      "step": 9112
+    },
+    {
+      "epoch": 0.6327593389806971,
+      "grad_norm": 3.2728292597054423,
+      "learning_rate": 3.139299564278275e-06,
+      "loss": 0.285,
+      "step": 9113
+    },
+    {
+      "epoch": 0.6328287737814192,
+      "grad_norm": 4.198386314155632,
+      "learning_rate": 3.1382558885601083e-06,
+      "loss": 0.537,
+      "step": 9114
+    },
+    {
+      "epoch": 0.6328982085821414,
+      "grad_norm": 3.8199848362621505,
+      "learning_rate": 3.1372123070068526e-06,
+      "loss": 0.3613,
+      "step": 9115
+    },
+    {
+      "epoch": 0.6329676433828635,
+      "grad_norm": 4.088662462855837,
+      "learning_rate": 3.136168819671289e-06,
+      "loss": 0.4417,
+      "step": 9116
+    },
+    {
+      "epoch": 0.6330370781835856,
+      "grad_norm": 2.890919771929855,
+      "learning_rate": 3.135125426606198e-06,
+      "loss": 0.2857,
+      "step": 9117
+    },
+    {
+      "epoch": 0.6331065129843078,
+      "grad_norm": 3.3794837237678936,
+      "learning_rate": 3.1340821278643517e-06,
+      "loss": 0.4171,
+      "step": 9118
+    },
+    {
+      "epoch": 0.6331759477850298,
+      "grad_norm": 4.197204846182557,
+      "learning_rate": 3.1330389234985227e-06,
+      "loss": 0.4376,
+      "step": 9119
+    },
+    {
+      "epoch": 0.633245382585752,
+      "grad_norm": 3.713106767913252,
+      "learning_rate": 3.131995813561471e-06,
+      "loss": 0.3739,
+      "step": 9120
+    },
+    {
+      "epoch": 0.6333148173864741,
+      "grad_norm": 3.7582068089545384,
+      "learning_rate": 3.130952798105956e-06,
+      "loss": 0.3419,
+      "step": 9121
+    },
+    {
+      "epoch": 0.6333842521871962,
+      "grad_norm": 3.687507563916093,
+      "learning_rate": 3.129909877184736e-06,
+      "loss": 0.3733,
+      "step": 9122
+    },
+    {
+      "epoch": 0.6334536869879184,
+      "grad_norm": 3.8300578761503767,
+      "learning_rate": 3.1288670508505553e-06,
+      "loss": 0.2964,
+      "step": 9123
+    },
+    {
+      "epoch": 0.6335231217886405,
+      "grad_norm": 4.511827133565723,
+      "learning_rate": 3.1278243191561634e-06,
+      "loss": 0.6327,
+      "step": 9124
+    },
+    {
+      "epoch": 0.6335925565893626,
+      "grad_norm": 4.4568938968699925,
+      "learning_rate": 3.1267816821542978e-06,
+      "loss": 0.4722,
+      "step": 9125
+    },
+    {
+      "epoch": 0.6336619913900847,
+      "grad_norm": 3.2337831406696464,
+      "learning_rate": 3.1257391398976976e-06,
+      "loss": 0.4047,
+      "step": 9126
+    },
+    {
+      "epoch": 0.6337314261908068,
+      "grad_norm": 3.3788951686237114,
+      "learning_rate": 3.1246966924390894e-06,
+      "loss": 0.3955,
+      "step": 9127
+    },
+    {
+      "epoch": 0.633800860991529,
+      "grad_norm": 3.605748428797732,
+      "learning_rate": 3.1236543398311993e-06,
+      "loss": 0.5181,
+      "step": 9128
+    },
+    {
+      "epoch": 0.6338702957922511,
+      "grad_norm": 5.3409707596416895,
+      "learning_rate": 3.1226120821267512e-06,
+      "loss": 0.5262,
+      "step": 9129
+    },
+    {
+      "epoch": 0.6339397305929731,
+      "grad_norm": 3.5995351170178984,
+      "learning_rate": 3.121569919378459e-06,
+      "loss": 0.3828,
+      "step": 9130
+    },
+    {
+      "epoch": 0.6340091653936953,
+      "grad_norm": 4.166939013846498,
+      "learning_rate": 3.1205278516390338e-06,
+      "loss": 0.5369,
+      "step": 9131
+    },
+    {
+      "epoch": 0.6340786001944174,
+      "grad_norm": 2.535761624502143,
+      "learning_rate": 3.1194858789611827e-06,
+      "loss": 0.1961,
+      "step": 9132
+    },
+    {
+      "epoch": 0.6341480349951396,
+      "grad_norm": 4.007329306605491,
+      "learning_rate": 3.1184440013976103e-06,
+      "loss": 0.2878,
+      "step": 9133
+    },
+    {
+      "epoch": 0.6342174697958617,
+      "grad_norm": 5.514203966393208,
+      "learning_rate": 3.11740221900101e-06,
+      "loss": 0.577,
+      "step": 9134
+    },
+    {
+      "epoch": 0.6342869045965838,
+      "grad_norm": 4.604426319202341,
+      "learning_rate": 3.116360531824074e-06,
+      "loss": 0.4144,
+      "step": 9135
+    },
+    {
+      "epoch": 0.634356339397306,
+      "grad_norm": 2.7595258394355575,
+      "learning_rate": 3.1153189399194935e-06,
+      "loss": 0.2687,
+      "step": 9136
+    },
+    {
+      "epoch": 0.634425774198028,
+      "grad_norm": 4.085625996009995,
+      "learning_rate": 3.114277443339947e-06,
+      "loss": 0.5095,
+      "step": 9137
+    },
+    {
+      "epoch": 0.6344952089987502,
+      "grad_norm": 3.955963767938622,
+      "learning_rate": 3.1132360421381123e-06,
+      "loss": 0.5686,
+      "step": 9138
+    },
+    {
+      "epoch": 0.6345646437994723,
+      "grad_norm": 3.3987516402706883,
+      "learning_rate": 3.112194736366665e-06,
+      "loss": 0.3147,
+      "step": 9139
+    },
+    {
+      "epoch": 0.6346340786001944,
+      "grad_norm": 3.848326083806358,
+      "learning_rate": 3.1111535260782723e-06,
+      "loss": 0.4984,
+      "step": 9140
+    },
+    {
+      "epoch": 0.6347035134009166,
+      "grad_norm": 3.4941473582728078,
+      "learning_rate": 3.1101124113255947e-06,
+      "loss": 0.2504,
+      "step": 9141
+    },
+    {
+      "epoch": 0.6347729482016387,
+      "grad_norm": 3.486891948914801,
+      "learning_rate": 3.1090713921612937e-06,
+      "loss": 0.3349,
+      "step": 9142
+    },
+    {
+      "epoch": 0.6348423830023607,
+      "grad_norm": 3.212744681813171,
+      "learning_rate": 3.1080304686380218e-06,
+      "loss": 0.3667,
+      "step": 9143
+    },
+    {
+      "epoch": 0.6349118178030829,
+      "grad_norm": 3.465894362717034,
+      "learning_rate": 3.10698964080843e-06,
+      "loss": 0.401,
+      "step": 9144
+    },
+    {
+      "epoch": 0.634981252603805,
+      "grad_norm": 4.842518361655951,
+      "learning_rate": 3.105948908725157e-06,
+      "loss": 0.5998,
+      "step": 9145
+    },
+    {
+      "epoch": 0.6350506874045272,
+      "grad_norm": 2.9034744565300685,
+      "learning_rate": 3.1049082724408464e-06,
+      "loss": 0.2476,
+      "step": 9146
+    },
+    {
+      "epoch": 0.6351201222052493,
+      "grad_norm": 3.977694882727863,
+      "learning_rate": 3.1038677320081322e-06,
+      "loss": 0.3819,
+      "step": 9147
+    },
+    {
+      "epoch": 0.6351895570059714,
+      "grad_norm": 3.237019929530829,
+      "learning_rate": 3.1028272874796406e-06,
+      "loss": 0.2966,
+      "step": 9148
+    },
+    {
+      "epoch": 0.6352589918066935,
+      "grad_norm": 3.277904586955306,
+      "learning_rate": 3.1017869389079985e-06,
+      "loss": 0.2749,
+      "step": 9149
+    },
+    {
+      "epoch": 0.6353284266074156,
+      "grad_norm": 3.026511340443682,
+      "learning_rate": 3.100746686345824e-06,
+      "loss": 0.3198,
+      "step": 9150
+    },
+    {
+      "epoch": 0.6353978614081378,
+      "grad_norm": 4.367265916985372,
+      "learning_rate": 3.0997065298457353e-06,
+      "loss": 0.6081,
+      "step": 9151
+    },
+    {
+      "epoch": 0.6354672962088599,
+      "grad_norm": 4.566468567971123,
+      "learning_rate": 3.098666469460339e-06,
+      "loss": 0.4312,
+      "step": 9152
+    },
+    {
+      "epoch": 0.635536731009582,
+      "grad_norm": 4.217257836991581,
+      "learning_rate": 3.09762650524224e-06,
+      "loss": 0.5185,
+      "step": 9153
+    },
+    {
+      "epoch": 0.6356061658103042,
+      "grad_norm": 2.3964626680034242,
+      "learning_rate": 3.096586637244042e-06,
+      "loss": 0.1905,
+      "step": 9154
+    },
+    {
+      "epoch": 0.6356756006110262,
+      "grad_norm": 3.8074629842982923,
+      "learning_rate": 3.095546865518336e-06,
+      "loss": 0.4193,
+      "step": 9155
+    },
+    {
+      "epoch": 0.6357450354117483,
+      "grad_norm": 4.827086755163089,
+      "learning_rate": 3.094507190117715e-06,
+      "loss": 0.4293,
+      "step": 9156
+    },
+    {
+      "epoch": 0.6358144702124705,
+      "grad_norm": 2.990164308735573,
+      "learning_rate": 3.0934676110947635e-06,
+      "loss": 0.2474,
+      "step": 9157
+    },
+    {
+      "epoch": 0.6358839050131926,
+      "grad_norm": 4.471027081612553,
+      "learning_rate": 3.0924281285020646e-06,
+      "loss": 0.6616,
+      "step": 9158
+    },
+    {
+      "epoch": 0.6359533398139148,
+      "grad_norm": 4.0299693091815225,
+      "learning_rate": 3.0913887423921917e-06,
+      "loss": 0.5341,
+      "step": 9159
+    },
+    {
+      "epoch": 0.6360227746146369,
+      "grad_norm": 4.348721162546524,
+      "learning_rate": 3.090349452817715e-06,
+      "loss": 0.6003,
+      "step": 9160
+    },
+    {
+      "epoch": 0.6360922094153589,
+      "grad_norm": 3.82363678430812,
+      "learning_rate": 3.0893102598312047e-06,
+      "loss": 0.5168,
+      "step": 9161
+    },
+    {
+      "epoch": 0.6361616442160811,
+      "grad_norm": 3.594729607567887,
+      "learning_rate": 3.088271163485218e-06,
+      "loss": 0.5112,
+      "step": 9162
+    },
+    {
+      "epoch": 0.6362310790168032,
+      "grad_norm": 2.835408198276321,
+      "learning_rate": 3.0872321638323117e-06,
+      "loss": 0.1965,
+      "step": 9163
+    },
+    {
+      "epoch": 0.6363005138175254,
+      "grad_norm": 3.885422020306639,
+      "learning_rate": 3.086193260925038e-06,
+      "loss": 0.3341,
+      "step": 9164
+    },
+    {
+      "epoch": 0.6363699486182475,
+      "grad_norm": 3.504304409267629,
+      "learning_rate": 3.085154454815945e-06,
+      "loss": 0.3923,
+      "step": 9165
+    },
+    {
+      "epoch": 0.6364393834189696,
+      "grad_norm": 3.645453873153388,
+      "learning_rate": 3.084115745557573e-06,
+      "loss": 0.3383,
+      "step": 9166
+    },
+    {
+      "epoch": 0.6365088182196917,
+      "grad_norm": 4.409696684021182,
+      "learning_rate": 3.0830771332024567e-06,
+      "loss": 0.594,
+      "step": 9167
+    },
+    {
+      "epoch": 0.6365782530204138,
+      "grad_norm": 2.963163384210168,
+      "learning_rate": 3.0820386178031313e-06,
+      "loss": 0.3585,
+      "step": 9168
+    },
+    {
+      "epoch": 0.6366476878211359,
+      "grad_norm": 4.117616242330566,
+      "learning_rate": 3.0810001994121237e-06,
+      "loss": 0.3832,
+      "step": 9169
+    },
+    {
+      "epoch": 0.6367171226218581,
+      "grad_norm": 2.9640720695834277,
+      "learning_rate": 3.079961878081952e-06,
+      "loss": 0.2228,
+      "step": 9170
+    },
+    {
+      "epoch": 0.6367865574225802,
+      "grad_norm": 3.386964593622585,
+      "learning_rate": 3.078923653865137e-06,
+      "loss": 0.2889,
+      "step": 9171
+    },
+    {
+      "epoch": 0.6368559922233024,
+      "grad_norm": 5.3027892092155255,
+      "learning_rate": 3.0778855268141905e-06,
+      "loss": 0.6608,
+      "step": 9172
+    },
+    {
+      "epoch": 0.6369254270240244,
+      "grad_norm": 4.036928537132345,
+      "learning_rate": 3.076847496981617e-06,
+      "loss": 0.5347,
+      "step": 9173
+    },
+    {
+      "epoch": 0.6369948618247465,
+      "grad_norm": 4.362397885665151,
+      "learning_rate": 3.0758095644199215e-06,
+      "loss": 0.5091,
+      "step": 9174
+    },
+    {
+      "epoch": 0.6370642966254687,
+      "grad_norm": 4.599330372845791,
+      "learning_rate": 3.0747717291816013e-06,
+      "loss": 0.5214,
+      "step": 9175
+    },
+    {
+      "epoch": 0.6371337314261908,
+      "grad_norm": 3.7927700716358315,
+      "learning_rate": 3.0737339913191494e-06,
+      "loss": 0.4355,
+      "step": 9176
+    },
+    {
+      "epoch": 0.637203166226913,
+      "grad_norm": 4.059265231002045,
+      "learning_rate": 3.0726963508850507e-06,
+      "loss": 0.6361,
+      "step": 9177
+    },
+    {
+      "epoch": 0.637272601027635,
+      "grad_norm": 4.205293404857485,
+      "learning_rate": 3.0716588079317912e-06,
+      "loss": 0.558,
+      "step": 9178
+    },
+    {
+      "epoch": 0.6373420358283571,
+      "grad_norm": 3.48428331349899,
+      "learning_rate": 3.0706213625118477e-06,
+      "loss": 0.4593,
+      "step": 9179
+    },
+    {
+      "epoch": 0.6374114706290793,
+      "grad_norm": 3.51581763091995,
+      "learning_rate": 3.06958401467769e-06,
+      "loss": 0.3077,
+      "step": 9180
+    },
+    {
+      "epoch": 0.6374809054298014,
+      "grad_norm": 5.095879932939914,
+      "learning_rate": 3.06854676448179e-06,
+      "loss": 0.4579,
+      "step": 9181
+    },
+    {
+      "epoch": 0.6375503402305236,
+      "grad_norm": 3.8311189047380445,
+      "learning_rate": 3.067509611976608e-06,
+      "loss": 0.444,
+      "step": 9182
+    },
+    {
+      "epoch": 0.6376197750312457,
+      "grad_norm": 3.190544033935898,
+      "learning_rate": 3.0664725572146048e-06,
+      "loss": 0.4386,
+      "step": 9183
+    },
+    {
+      "epoch": 0.6376892098319678,
+      "grad_norm": 5.109490575221364,
+      "learning_rate": 3.06543560024823e-06,
+      "loss": 0.3498,
+      "step": 9184
+    },
+    {
+      "epoch": 0.6377586446326899,
+      "grad_norm": 4.639418743248024,
+      "learning_rate": 3.064398741129934e-06,
+      "loss": 0.7355,
+      "step": 9185
+    },
+    {
+      "epoch": 0.637828079433412,
+      "grad_norm": 4.491716608473725,
+      "learning_rate": 3.0633619799121605e-06,
+      "loss": 0.4238,
+      "step": 9186
+    },
+    {
+      "epoch": 0.6378975142341341,
+      "grad_norm": 4.083749807547369,
+      "learning_rate": 3.0623253166473447e-06,
+      "loss": 0.5447,
+      "step": 9187
+    },
+    {
+      "epoch": 0.6379669490348563,
+      "grad_norm": 4.58828917576333,
+      "learning_rate": 3.061288751387923e-06,
+      "loss": 0.6826,
+      "step": 9188
+    },
+    {
+      "epoch": 0.6380363838355784,
+      "grad_norm": 2.1574993530422284,
+      "learning_rate": 3.0602522841863213e-06,
+      "loss": 0.1673,
+      "step": 9189
+    },
+    {
+      "epoch": 0.6381058186363006,
+      "grad_norm": 3.095599145691764,
+      "learning_rate": 3.0592159150949665e-06,
+      "loss": 0.3611,
+      "step": 9190
+    },
+    {
+      "epoch": 0.6381752534370226,
+      "grad_norm": 3.06301454519742,
+      "learning_rate": 3.0581796441662726e-06,
+      "loss": 0.2606,
+      "step": 9191
+    },
+    {
+      "epoch": 0.6382446882377447,
+      "grad_norm": 3.7398299833285473,
+      "learning_rate": 3.057143471452655e-06,
+      "loss": 0.5205,
+      "step": 9192
+    },
+    {
+      "epoch": 0.6383141230384669,
+      "grad_norm": 4.379182693150694,
+      "learning_rate": 3.056107397006521e-06,
+      "loss": 0.6175,
+      "step": 9193
+    },
+    {
+      "epoch": 0.638383557839189,
+      "grad_norm": 3.50329170915092,
+      "learning_rate": 3.0550714208802784e-06,
+      "loss": 0.3814,
+      "step": 9194
+    },
+    {
+      "epoch": 0.6384529926399112,
+      "grad_norm": 3.986684558682219,
+      "learning_rate": 3.054035543126321e-06,
+      "loss": 0.5117,
+      "step": 9195
+    },
+    {
+      "epoch": 0.6385224274406333,
+      "grad_norm": 4.538989947927328,
+      "learning_rate": 3.0529997637970427e-06,
+      "loss": 0.5644,
+      "step": 9196
+    },
+    {
+      "epoch": 0.6385918622413553,
+      "grad_norm": 3.85732703851041,
+      "learning_rate": 3.051964082944835e-06,
+      "loss": 0.3692,
+      "step": 9197
+    },
+    {
+      "epoch": 0.6386612970420775,
+      "grad_norm": 4.0775988933878375,
+      "learning_rate": 3.050928500622079e-06,
+      "loss": 0.3508,
+      "step": 9198
+    },
+    {
+      "epoch": 0.6387307318427996,
+      "grad_norm": 3.190283741040305,
+      "learning_rate": 3.049893016881153e-06,
+      "loss": 0.2838,
+      "step": 9199
+    },
+    {
+      "epoch": 0.6388001666435217,
+      "grad_norm": 3.581947921733446,
+      "learning_rate": 3.0488576317744324e-06,
+      "loss": 0.4977,
+      "step": 9200
+    },
+    {
+      "epoch": 0.6388696014442439,
+      "grad_norm": 3.2675839920562297,
+      "learning_rate": 3.0478223453542854e-06,
+      "loss": 0.2496,
+      "step": 9201
+    },
+    {
+      "epoch": 0.638939036244966,
+      "grad_norm": 4.085595421299331,
+      "learning_rate": 3.0467871576730732e-06,
+      "loss": 0.3809,
+      "step": 9202
+    },
+    {
+      "epoch": 0.6390084710456881,
+      "grad_norm": 4.000723923632379,
+      "learning_rate": 3.0457520687831577e-06,
+      "loss": 0.4013,
+      "step": 9203
+    },
+    {
+      "epoch": 0.6390779058464102,
+      "grad_norm": 4.265001820942161,
+      "learning_rate": 3.0447170787368917e-06,
+      "loss": 0.5814,
+      "step": 9204
+    },
+    {
+      "epoch": 0.6391473406471323,
+      "grad_norm": 5.050124054977223,
+      "learning_rate": 3.043682187586623e-06,
+      "loss": 0.7794,
+      "step": 9205
+    },
+    {
+      "epoch": 0.6392167754478545,
+      "grad_norm": 3.1133868284628488,
+      "learning_rate": 3.042647395384695e-06,
+      "loss": 0.5265,
+      "step": 9206
+    },
+    {
+      "epoch": 0.6392862102485766,
+      "grad_norm": 2.7650399210438312,
+      "learning_rate": 3.041612702183446e-06,
+      "loss": 0.212,
+      "step": 9207
+    },
+    {
+      "epoch": 0.6393556450492988,
+      "grad_norm": 3.8289477239861816,
+      "learning_rate": 3.0405781080352132e-06,
+      "loss": 0.4708,
+      "step": 9208
+    },
+    {
+      "epoch": 0.6394250798500208,
+      "grad_norm": 3.0974020931115938,
+      "learning_rate": 3.0395436129923195e-06,
+      "loss": 0.2894,
+      "step": 9209
+    },
+    {
+      "epoch": 0.6394945146507429,
+      "grad_norm": 4.167026765069896,
+      "learning_rate": 3.038509217107093e-06,
+      "loss": 0.6079,
+      "step": 9210
+    },
+    {
+      "epoch": 0.6395639494514651,
+      "grad_norm": 4.547077902789774,
+      "learning_rate": 3.037474920431852e-06,
+      "loss": 0.4101,
+      "step": 9211
+    },
+    {
+      "epoch": 0.6396333842521872,
+      "grad_norm": 3.9678593277322007,
+      "learning_rate": 3.0364407230189065e-06,
+      "loss": 0.3608,
+      "step": 9212
+    },
+    {
+      "epoch": 0.6397028190529093,
+      "grad_norm": 3.84118401822579,
+      "learning_rate": 3.0354066249205665e-06,
+      "loss": 0.4282,
+      "step": 9213
+    },
+    {
+      "epoch": 0.6397722538536315,
+      "grad_norm": 2.989267316218682,
+      "learning_rate": 3.0343726261891384e-06,
+      "loss": 0.2451,
+      "step": 9214
+    },
+    {
+      "epoch": 0.6398416886543535,
+      "grad_norm": 2.872453265565878,
+      "learning_rate": 3.0333387268769187e-06,
+      "loss": 0.3888,
+      "step": 9215
+    },
+    {
+      "epoch": 0.6399111234550757,
+      "grad_norm": 3.857653654404107,
+      "learning_rate": 3.032304927036199e-06,
+      "loss": 0.4433,
+      "step": 9216
+    },
+    {
+      "epoch": 0.6399805582557978,
+      "grad_norm": 4.0454267085239035,
+      "learning_rate": 3.0312712267192713e-06,
+      "loss": 0.5639,
+      "step": 9217
+    },
+    {
+      "epoch": 0.6400499930565199,
+      "grad_norm": 3.9770261557645985,
+      "learning_rate": 3.0302376259784156e-06,
+      "loss": 0.5282,
+      "step": 9218
+    },
+    {
+      "epoch": 0.6401194278572421,
+      "grad_norm": 4.087713146463761,
+      "learning_rate": 3.0292041248659136e-06,
+      "loss": 0.4148,
+      "step": 9219
+    },
+    {
+      "epoch": 0.6401888626579642,
+      "grad_norm": 4.526583811865199,
+      "learning_rate": 3.0281707234340365e-06,
+      "loss": 0.5451,
+      "step": 9220
+    },
+    {
+      "epoch": 0.6402582974586863,
+      "grad_norm": 3.153833275766328,
+      "learning_rate": 3.0271374217350513e-06,
+      "loss": 0.2717,
+      "step": 9221
+    },
+    {
+      "epoch": 0.6403277322594084,
+      "grad_norm": 3.0581615320885907,
+      "learning_rate": 3.0261042198212254e-06,
+      "loss": 0.2514,
+      "step": 9222
+    },
+    {
+      "epoch": 0.6403971670601305,
+      "grad_norm": 3.4754140996312417,
+      "learning_rate": 3.0250711177448124e-06,
+      "loss": 0.2466,
+      "step": 9223
+    },
+    {
+      "epoch": 0.6404666018608527,
+      "grad_norm": 2.824554805909999,
+      "learning_rate": 3.0240381155580677e-06,
+      "loss": 0.2196,
+      "step": 9224
+    },
+    {
+      "epoch": 0.6405360366615748,
+      "grad_norm": 2.759779438842844,
+      "learning_rate": 3.0230052133132394e-06,
+      "loss": 0.2793,
+      "step": 9225
+    },
+    {
+      "epoch": 0.6406054714622968,
+      "grad_norm": 4.148619318996465,
+      "learning_rate": 3.0219724110625716e-06,
+      "loss": 0.6206,
+      "step": 9226
+    },
+    {
+      "epoch": 0.640674906263019,
+      "grad_norm": 4.032413160928527,
+      "learning_rate": 3.020939708858299e-06,
+      "loss": 0.4216,
+      "step": 9227
+    },
+    {
+      "epoch": 0.6407443410637411,
+      "grad_norm": 3.525562155733383,
+      "learning_rate": 3.0199071067526568e-06,
+      "loss": 0.2415,
+      "step": 9228
+    },
+    {
+      "epoch": 0.6408137758644633,
+      "grad_norm": 3.546115546267027,
+      "learning_rate": 3.0188746047978743e-06,
+      "loss": 0.385,
+      "step": 9229
+    },
+    {
+      "epoch": 0.6408832106651854,
+      "grad_norm": 3.6978198108449374,
+      "learning_rate": 3.0178422030461714e-06,
+      "loss": 0.44,
+      "step": 9230
+    },
+    {
+      "epoch": 0.6409526454659075,
+      "grad_norm": 3.958845986852517,
+      "learning_rate": 3.0168099015497653e-06,
+      "loss": 0.4344,
+      "step": 9231
+    },
+    {
+      "epoch": 0.6410220802666297,
+      "grad_norm": 2.829624068401706,
+      "learning_rate": 3.0157777003608724e-06,
+      "loss": 0.3333,
+      "step": 9232
+    },
+    {
+      "epoch": 0.6410915150673517,
+      "grad_norm": 4.110706734502829,
+      "learning_rate": 3.0147455995316983e-06,
+      "loss": 0.4729,
+      "step": 9233
+    },
+    {
+      "epoch": 0.6411609498680739,
+      "grad_norm": 4.594615858764353,
+      "learning_rate": 3.013713599114444e-06,
+      "loss": 0.6442,
+      "step": 9234
+    },
+    {
+      "epoch": 0.641230384668796,
+      "grad_norm": 4.20527620229064,
+      "learning_rate": 3.0126816991613074e-06,
+      "loss": 0.5115,
+      "step": 9235
+    },
+    {
+      "epoch": 0.6412998194695181,
+      "grad_norm": 2.8735100264985896,
+      "learning_rate": 3.0116498997244837e-06,
+      "loss": 0.1481,
+      "step": 9236
+    },
+    {
+      "epoch": 0.6413692542702403,
+      "grad_norm": 3.419586929280404,
+      "learning_rate": 3.010618200856158e-06,
+      "loss": 0.2749,
+      "step": 9237
+    },
+    {
+      "epoch": 0.6414386890709624,
+      "grad_norm": 4.403665044999153,
+      "learning_rate": 3.009586602608511e-06,
+      "loss": 0.3478,
+      "step": 9238
+    },
+    {
+      "epoch": 0.6415081238716845,
+      "grad_norm": 5.149912788284767,
+      "learning_rate": 3.008555105033722e-06,
+      "loss": 0.5213,
+      "step": 9239
+    },
+    {
+      "epoch": 0.6415775586724066,
+      "grad_norm": 2.623899093127052,
+      "learning_rate": 3.0075237081839637e-06,
+      "loss": 0.105,
+      "step": 9240
+    },
+    {
+      "epoch": 0.6416469934731287,
+      "grad_norm": 4.4069661978926895,
+      "learning_rate": 3.006492412111399e-06,
+      "loss": 0.4818,
+      "step": 9241
+    },
+    {
+      "epoch": 0.6417164282738509,
+      "grad_norm": 4.040430397226186,
+      "learning_rate": 3.005461216868194e-06,
+      "loss": 0.2917,
+      "step": 9242
+    },
+    {
+      "epoch": 0.641785863074573,
+      "grad_norm": 2.600201974593412,
+      "learning_rate": 3.0044301225065026e-06,
+      "loss": 0.2026,
+      "step": 9243
+    },
+    {
+      "epoch": 0.641855297875295,
+      "grad_norm": 3.6272162738628633,
+      "learning_rate": 3.0033991290784793e-06,
+      "loss": 0.5295,
+      "step": 9244
+    },
+    {
+      "epoch": 0.6419247326760172,
+      "grad_norm": 3.5671470684859696,
+      "learning_rate": 3.002368236636267e-06,
+      "loss": 0.1925,
+      "step": 9245
+    },
+    {
+      "epoch": 0.6419941674767393,
+      "grad_norm": 3.9941066726524665,
+      "learning_rate": 3.0013374452320093e-06,
+      "loss": 0.3482,
+      "step": 9246
+    },
+    {
+      "epoch": 0.6420636022774615,
+      "grad_norm": 3.921454590055716,
+      "learning_rate": 3.0003067549178432e-06,
+      "loss": 0.4234,
+      "step": 9247
+    },
+    {
+      "epoch": 0.6421330370781836,
+      "grad_norm": 2.6385570502835973,
+      "learning_rate": 2.999276165745897e-06,
+      "loss": 0.1885,
+      "step": 9248
+    },
+    {
+      "epoch": 0.6422024718789057,
+      "grad_norm": 3.8665810016721767,
+      "learning_rate": 2.998245677768299e-06,
+      "loss": 0.3181,
+      "step": 9249
+    },
+    {
+      "epoch": 0.6422719066796279,
+      "grad_norm": 3.715810186000461,
+      "learning_rate": 2.9972152910371686e-06,
+      "loss": 0.4123,
+      "step": 9250
+    },
+    {
+      "epoch": 0.6423413414803499,
+      "grad_norm": 4.305073678726027,
+      "learning_rate": 2.9961850056046243e-06,
+      "loss": 0.4204,
+      "step": 9251
+    },
+    {
+      "epoch": 0.6424107762810721,
+      "grad_norm": 3.644948504333781,
+      "learning_rate": 2.9951548215227744e-06,
+      "loss": 0.4583,
+      "step": 9252
+    },
+    {
+      "epoch": 0.6424802110817942,
+      "grad_norm": 3.1164302782202307,
+      "learning_rate": 2.994124738843723e-06,
+      "loss": 0.1733,
+      "step": 9253
+    },
+    {
+      "epoch": 0.6425496458825163,
+      "grad_norm": 3.2753664899275563,
+      "learning_rate": 2.9930947576195755e-06,
+      "loss": 0.3275,
+      "step": 9254
+    },
+    {
+      "epoch": 0.6426190806832385,
+      "grad_norm": 4.629362178816517,
+      "learning_rate": 2.9920648779024218e-06,
+      "loss": 0.4944,
+      "step": 9255
+    },
+    {
+      "epoch": 0.6426885154839606,
+      "grad_norm": 4.090489019122891,
+      "learning_rate": 2.991035099744356e-06,
+      "loss": 0.4967,
+      "step": 9256
+    },
+    {
+      "epoch": 0.6427579502846826,
+      "grad_norm": 5.338297088656182,
+      "learning_rate": 2.9900054231974595e-06,
+      "loss": 0.5633,
+      "step": 9257
+    },
+    {
+      "epoch": 0.6428273850854048,
+      "grad_norm": 3.090780318212311,
+      "learning_rate": 2.988975848313817e-06,
+      "loss": 0.2025,
+      "step": 9258
+    },
+    {
+      "epoch": 0.6428968198861269,
+      "grad_norm": 2.6801072044715806,
+      "learning_rate": 2.9879463751454995e-06,
+      "loss": 0.2895,
+      "step": 9259
+    },
+    {
+      "epoch": 0.6429662546868491,
+      "grad_norm": 3.8634603642899563,
+      "learning_rate": 2.986917003744576e-06,
+      "loss": 0.767,
+      "step": 9260
+    },
+    {
+      "epoch": 0.6430356894875712,
+      "grad_norm": 4.203461797468175,
+      "learning_rate": 2.9858877341631147e-06,
+      "loss": 0.3959,
+      "step": 9261
+    },
+    {
+      "epoch": 0.6431051242882933,
+      "grad_norm": 3.242818750460865,
+      "learning_rate": 2.984858566453172e-06,
+      "loss": 0.2447,
+      "step": 9262
+    },
+    {
+      "epoch": 0.6431745590890154,
+      "grad_norm": 4.212255819282311,
+      "learning_rate": 2.9838295006668017e-06,
+      "loss": 0.5383,
+      "step": 9263
+    },
+    {
+      "epoch": 0.6432439938897375,
+      "grad_norm": 4.777169198005545,
+      "learning_rate": 2.982800536856053e-06,
+      "loss": 0.5767,
+      "step": 9264
+    },
+    {
+      "epoch": 0.6433134286904597,
+      "grad_norm": 4.440459526544951,
+      "learning_rate": 2.9817716750729734e-06,
+      "loss": 0.5095,
+      "step": 9265
+    },
+    {
+      "epoch": 0.6433828634911818,
+      "grad_norm": 4.3443924943978605,
+      "learning_rate": 2.9807429153695978e-06,
+      "loss": 0.5769,
+      "step": 9266
+    },
+    {
+      "epoch": 0.6434522982919039,
+      "grad_norm": 5.774244938021644,
+      "learning_rate": 2.979714257797959e-06,
+      "loss": 0.5874,
+      "step": 9267
+    },
+    {
+      "epoch": 0.6435217330926261,
+      "grad_norm": 4.141338818881184,
+      "learning_rate": 2.978685702410088e-06,
+      "loss": 0.3195,
+      "step": 9268
+    },
+    {
+      "epoch": 0.6435911678933481,
+      "grad_norm": 4.57017352619534,
+      "learning_rate": 2.977657249258008e-06,
+      "loss": 0.2907,
+      "step": 9269
+    },
+    {
+      "epoch": 0.6436606026940702,
+      "grad_norm": 3.3870542807694655,
+      "learning_rate": 2.976628898393734e-06,
+      "loss": 0.2796,
+      "step": 9270
+    },
+    {
+      "epoch": 0.6437300374947924,
+      "grad_norm": 4.255446700411184,
+      "learning_rate": 2.975600649869282e-06,
+      "loss": 0.5319,
+      "step": 9271
+    },
+    {
+      "epoch": 0.6437994722955145,
+      "grad_norm": 4.777211866562169,
+      "learning_rate": 2.9745725037366592e-06,
+      "loss": 0.71,
+      "step": 9272
+    },
+    {
+      "epoch": 0.6438689070962367,
+      "grad_norm": 4.063096944041227,
+      "learning_rate": 2.9735444600478645e-06,
+      "loss": 0.4951,
+      "step": 9273
+    },
+    {
+      "epoch": 0.6439383418969588,
+      "grad_norm": 3.2525037389842084,
+      "learning_rate": 2.972516518854899e-06,
+      "loss": 0.3324,
+      "step": 9274
+    },
+    {
+      "epoch": 0.6440077766976808,
+      "grad_norm": 4.056170994305566,
+      "learning_rate": 2.9714886802097544e-06,
+      "loss": 0.4801,
+      "step": 9275
+    },
+    {
+      "epoch": 0.644077211498403,
+      "grad_norm": 3.4055820485624655,
+      "learning_rate": 2.970460944164417e-06,
+      "loss": 0.3944,
+      "step": 9276
+    },
+    {
+      "epoch": 0.6441466462991251,
+      "grad_norm": 3.160960310144805,
+      "learning_rate": 2.969433310770868e-06,
+      "loss": 0.3766,
+      "step": 9277
+    },
+    {
+      "epoch": 0.6442160810998473,
+      "grad_norm": 3.645772111136888,
+      "learning_rate": 2.9684057800810844e-06,
+      "loss": 0.2753,
+      "step": 9278
+    },
+    {
+      "epoch": 0.6442855159005694,
+      "grad_norm": 4.0263276900214615,
+      "learning_rate": 2.967378352147039e-06,
+      "loss": 0.3478,
+      "step": 9279
+    },
+    {
+      "epoch": 0.6443549507012915,
+      "grad_norm": 3.752482071975173,
+      "learning_rate": 2.9663510270206943e-06,
+      "loss": 0.467,
+      "step": 9280
+    },
+    {
+      "epoch": 0.6444243855020136,
+      "grad_norm": 3.6982982992348115,
+      "learning_rate": 2.9653238047540145e-06,
+      "loss": 0.6211,
+      "step": 9281
+    },
+    {
+      "epoch": 0.6444938203027357,
+      "grad_norm": 3.0495770811725325,
+      "learning_rate": 2.964296685398953e-06,
+      "loss": 0.2402,
+      "step": 9282
+    },
+    {
+      "epoch": 0.6445632551034578,
+      "grad_norm": 4.45616646446758,
+      "learning_rate": 2.9632696690074637e-06,
+      "loss": 0.5415,
+      "step": 9283
+    },
+    {
+      "epoch": 0.64463268990418,
+      "grad_norm": 3.6830146353708937,
+      "learning_rate": 2.962242755631488e-06,
+      "loss": 0.2941,
+      "step": 9284
+    },
+    {
+      "epoch": 0.6447021247049021,
+      "grad_norm": 2.9766987603458417,
+      "learning_rate": 2.9612159453229684e-06,
+      "loss": 0.1465,
+      "step": 9285
+    },
+    {
+      "epoch": 0.6447715595056243,
+      "grad_norm": 2.8155170167967447,
+      "learning_rate": 2.9601892381338403e-06,
+      "loss": 0.1859,
+      "step": 9286
+    },
+    {
+      "epoch": 0.6448409943063463,
+      "grad_norm": 4.614336730960642,
+      "learning_rate": 2.9591626341160306e-06,
+      "loss": 0.5458,
+      "step": 9287
+    },
+    {
+      "epoch": 0.6449104291070684,
+      "grad_norm": 4.159466938555124,
+      "learning_rate": 2.9581361333214666e-06,
+      "loss": 0.221,
+      "step": 9288
+    },
+    {
+      "epoch": 0.6449798639077906,
+      "grad_norm": 2.87577471129891,
+      "learning_rate": 2.9571097358020655e-06,
+      "loss": 0.2529,
+      "step": 9289
+    },
+    {
+      "epoch": 0.6450492987085127,
+      "grad_norm": 4.691077294551126,
+      "learning_rate": 2.9560834416097438e-06,
+      "loss": 0.4253,
+      "step": 9290
+    },
+    {
+      "epoch": 0.6451187335092349,
+      "grad_norm": 4.186629536946255,
+      "learning_rate": 2.9550572507964084e-06,
+      "loss": 0.5545,
+      "step": 9291
+    },
+    {
+      "epoch": 0.645188168309957,
+      "grad_norm": 4.088181002105651,
+      "learning_rate": 2.9540311634139616e-06,
+      "loss": 0.2277,
+      "step": 9292
+    },
+    {
+      "epoch": 0.645257603110679,
+      "grad_norm": 3.345797464352814,
+      "learning_rate": 2.9530051795143046e-06,
+      "loss": 0.3451,
+      "step": 9293
+    },
+    {
+      "epoch": 0.6453270379114012,
+      "grad_norm": 3.1945937936846662,
+      "learning_rate": 2.95197929914933e-06,
+      "loss": 0.1959,
+      "step": 9294
+    },
+    {
+      "epoch": 0.6453964727121233,
+      "grad_norm": 3.763238481239324,
+      "learning_rate": 2.950953522370924e-06,
+      "loss": 0.3678,
+      "step": 9295
+    },
+    {
+      "epoch": 0.6454659075128454,
+      "grad_norm": 4.271868434021699,
+      "learning_rate": 2.9499278492309702e-06,
+      "loss": 0.6152,
+      "step": 9296
+    },
+    {
+      "epoch": 0.6455353423135676,
+      "grad_norm": 2.435005242879208,
+      "learning_rate": 2.948902279781347e-06,
+      "loss": 0.1579,
+      "step": 9297
+    },
+    {
+      "epoch": 0.6456047771142897,
+      "grad_norm": 4.3125500864892095,
+      "learning_rate": 2.947876814073925e-06,
+      "loss": 0.4421,
+      "step": 9298
+    },
+    {
+      "epoch": 0.6456742119150118,
+      "grad_norm": 4.339643505583527,
+      "learning_rate": 2.9468514521605705e-06,
+      "loss": 0.3355,
+      "step": 9299
+    },
+    {
+      "epoch": 0.6457436467157339,
+      "grad_norm": 4.0756121044204585,
+      "learning_rate": 2.945826194093147e-06,
+      "loss": 0.3912,
+      "step": 9300
+    },
+    {
+      "epoch": 0.645813081516456,
+      "grad_norm": 3.7548965988748924,
+      "learning_rate": 2.9448010399235118e-06,
+      "loss": 0.4316,
+      "step": 9301
+    },
+    {
+      "epoch": 0.6458825163171782,
+      "grad_norm": 4.196059796538449,
+      "learning_rate": 2.9437759897035114e-06,
+      "loss": 0.4771,
+      "step": 9302
+    },
+    {
+      "epoch": 0.6459519511179003,
+      "grad_norm": 3.496743051493309,
+      "learning_rate": 2.9427510434849954e-06,
+      "loss": 0.2502,
+      "step": 9303
+    },
+    {
+      "epoch": 0.6460213859186225,
+      "grad_norm": 3.770622589578973,
+      "learning_rate": 2.9417262013198055e-06,
+      "loss": 0.3055,
+      "step": 9304
+    },
+    {
+      "epoch": 0.6460908207193445,
+      "grad_norm": 2.4486059725104377,
+      "learning_rate": 2.940701463259774e-06,
+      "loss": 0.2529,
+      "step": 9305
+    },
+    {
+      "epoch": 0.6461602555200666,
+      "grad_norm": 3.5994856689681356,
+      "learning_rate": 2.939676829356732e-06,
+      "loss": 0.4097,
+      "step": 9306
+    },
+    {
+      "epoch": 0.6462296903207888,
+      "grad_norm": 4.566875924389935,
+      "learning_rate": 2.9386522996625045e-06,
+      "loss": 0.5534,
+      "step": 9307
+    },
+    {
+      "epoch": 0.6462991251215109,
+      "grad_norm": 3.7398873365120946,
+      "learning_rate": 2.9376278742289133e-06,
+      "loss": 0.3693,
+      "step": 9308
+    },
+    {
+      "epoch": 0.6463685599222331,
+      "grad_norm": 3.6462965957186624,
+      "learning_rate": 2.9366035531077676e-06,
+      "loss": 0.4176,
+      "step": 9309
+    },
+    {
+      "epoch": 0.6464379947229552,
+      "grad_norm": 3.301633085355053,
+      "learning_rate": 2.93557933635088e-06,
+      "loss": 0.4995,
+      "step": 9310
+    },
+    {
+      "epoch": 0.6465074295236772,
+      "grad_norm": 4.66987493322447,
+      "learning_rate": 2.934555224010055e-06,
+      "loss": 0.5465,
+      "step": 9311
+    },
+    {
+      "epoch": 0.6465768643243994,
+      "grad_norm": 3.3879447225887005,
+      "learning_rate": 2.9335312161370877e-06,
+      "loss": 0.3307,
+      "step": 9312
+    },
+    {
+      "epoch": 0.6466462991251215,
+      "grad_norm": 3.157931854386785,
+      "learning_rate": 2.9325073127837746e-06,
+      "loss": 0.2664,
+      "step": 9313
+    },
+    {
+      "epoch": 0.6467157339258436,
+      "grad_norm": 3.5718797988927484,
+      "learning_rate": 2.9314835140019002e-06,
+      "loss": 0.3344,
+      "step": 9314
+    },
+    {
+      "epoch": 0.6467851687265658,
+      "grad_norm": 4.07947656641248,
+      "learning_rate": 2.930459819843251e-06,
+      "loss": 0.3712,
+      "step": 9315
+    },
+    {
+      "epoch": 0.6468546035272879,
+      "grad_norm": 3.324083947054877,
+      "learning_rate": 2.929436230359601e-06,
+      "loss": 0.3227,
+      "step": 9316
+    },
+    {
+      "epoch": 0.64692403832801,
+      "grad_norm": 4.408991516546936,
+      "learning_rate": 2.928412745602724e-06,
+      "loss": 0.5036,
+      "step": 9317
+    },
+    {
+      "epoch": 0.6469934731287321,
+      "grad_norm": 3.1950205194458876,
+      "learning_rate": 2.927389365624387e-06,
+      "loss": 0.3897,
+      "step": 9318
+    },
+    {
+      "epoch": 0.6470629079294542,
+      "grad_norm": 3.699530482376584,
+      "learning_rate": 2.926366090476349e-06,
+      "loss": 0.36,
+      "step": 9319
+    },
+    {
+      "epoch": 0.6471323427301764,
+      "grad_norm": 3.1454626127292244,
+      "learning_rate": 2.9253429202103683e-06,
+      "loss": 0.398,
+      "step": 9320
+    },
+    {
+      "epoch": 0.6472017775308985,
+      "grad_norm": 3.613941857386303,
+      "learning_rate": 2.9243198548781944e-06,
+      "loss": 0.2997,
+      "step": 9321
+    },
+    {
+      "epoch": 0.6472712123316207,
+      "grad_norm": 5.476086439522293,
+      "learning_rate": 2.9232968945315754e-06,
+      "loss": 0.6228,
+      "step": 9322
+    },
+    {
+      "epoch": 0.6473406471323427,
+      "grad_norm": 5.174463313074087,
+      "learning_rate": 2.9222740392222482e-06,
+      "loss": 0.5212,
+      "step": 9323
+    },
+    {
+      "epoch": 0.6474100819330648,
+      "grad_norm": 3.899216200689656,
+      "learning_rate": 2.921251289001948e-06,
+      "loss": 0.3755,
+      "step": 9324
+    },
+    {
+      "epoch": 0.647479516733787,
+      "grad_norm": 3.6275258758170517,
+      "learning_rate": 2.920228643922405e-06,
+      "loss": 0.4094,
+      "step": 9325
+    },
+    {
+      "epoch": 0.6475489515345091,
+      "grad_norm": 3.6260691164197483,
+      "learning_rate": 2.919206104035348e-06,
+      "loss": 0.3234,
+      "step": 9326
+    },
+    {
+      "epoch": 0.6476183863352312,
+      "grad_norm": 3.881638948091992,
+      "learning_rate": 2.918183669392487e-06,
+      "loss": 0.3377,
+      "step": 9327
+    },
+    {
+      "epoch": 0.6476878211359534,
+      "grad_norm": 3.9760436370029497,
+      "learning_rate": 2.917161340045542e-06,
+      "loss": 0.4103,
+      "step": 9328
+    },
+    {
+      "epoch": 0.6477572559366754,
+      "grad_norm": 4.147155756551901,
+      "learning_rate": 2.9161391160462217e-06,
+      "loss": 0.4818,
+      "step": 9329
+    },
+    {
+      "epoch": 0.6478266907373976,
+      "grad_norm": 3.884510537932424,
+      "learning_rate": 2.915116997446223e-06,
+      "loss": 0.3239,
+      "step": 9330
+    },
+    {
+      "epoch": 0.6478961255381197,
+      "grad_norm": 5.105626892689658,
+      "learning_rate": 2.9140949842972486e-06,
+      "loss": 0.456,
+      "step": 9331
+    },
+    {
+      "epoch": 0.6479655603388418,
+      "grad_norm": 2.6109505187509248,
+      "learning_rate": 2.913073076650991e-06,
+      "loss": 0.1872,
+      "step": 9332
+    },
+    {
+      "epoch": 0.648034995139564,
+      "grad_norm": 5.67933043200784,
+      "learning_rate": 2.9120512745591346e-06,
+      "loss": 0.5114,
+      "step": 9333
+    },
+    {
+      "epoch": 0.648104429940286,
+      "grad_norm": 3.66553518854289,
+      "learning_rate": 2.911029578073362e-06,
+      "loss": 0.4519,
+      "step": 9334
+    },
+    {
+      "epoch": 0.6481738647410082,
+      "grad_norm": 4.354977181623627,
+      "learning_rate": 2.9100079872453502e-06,
+      "loss": 0.5249,
+      "step": 9335
+    },
+    {
+      "epoch": 0.6482432995417303,
+      "grad_norm": 4.56304854028624,
+      "learning_rate": 2.9089865021267695e-06,
+      "loss": 0.2835,
+      "step": 9336
+    },
+    {
+      "epoch": 0.6483127343424524,
+      "grad_norm": 3.8785521605118056,
+      "learning_rate": 2.9079651227692845e-06,
+      "loss": 0.4648,
+      "step": 9337
+    },
+    {
+      "epoch": 0.6483821691431746,
+      "grad_norm": 4.039012143331419,
+      "learning_rate": 2.9069438492245567e-06,
+      "loss": 0.3608,
+      "step": 9338
+    },
+    {
+      "epoch": 0.6484516039438967,
+      "grad_norm": 3.394670340118899,
+      "learning_rate": 2.9059226815442386e-06,
+      "loss": 0.3325,
+      "step": 9339
+    },
+    {
+      "epoch": 0.6485210387446187,
+      "grad_norm": 3.898850774297787,
+      "learning_rate": 2.9049016197799856e-06,
+      "loss": 0.2918,
+      "step": 9340
+    },
+    {
+      "epoch": 0.6485904735453409,
+      "grad_norm": 5.3741973194503085,
+      "learning_rate": 2.903880663983436e-06,
+      "loss": 0.3897,
+      "step": 9341
+    },
+    {
+      "epoch": 0.648659908346063,
+      "grad_norm": 4.331797937415311,
+      "learning_rate": 2.9028598142062283e-06,
+      "loss": 0.3659,
+      "step": 9342
+    },
+    {
+      "epoch": 0.6487293431467852,
+      "grad_norm": 5.110900150942827,
+      "learning_rate": 2.9018390705000022e-06,
+      "loss": 0.7566,
+      "step": 9343
+    },
+    {
+      "epoch": 0.6487987779475073,
+      "grad_norm": 3.458513341315482,
+      "learning_rate": 2.900818432916378e-06,
+      "loss": 0.3842,
+      "step": 9344
+    },
+    {
+      "epoch": 0.6488682127482294,
+      "grad_norm": 4.582739325769734,
+      "learning_rate": 2.899797901506984e-06,
+      "loss": 0.5856,
+      "step": 9345
+    },
+    {
+      "epoch": 0.6489376475489516,
+      "grad_norm": 4.0171285482929235,
+      "learning_rate": 2.8987774763234354e-06,
+      "loss": 0.5015,
+      "step": 9346
+    },
+    {
+      "epoch": 0.6490070823496736,
+      "grad_norm": 3.3826203749936776,
+      "learning_rate": 2.897757157417344e-06,
+      "loss": 0.3244,
+      "step": 9347
+    },
+    {
+      "epoch": 0.6490765171503958,
+      "grad_norm": 3.0735300449725087,
+      "learning_rate": 2.896736944840317e-06,
+      "loss": 0.2157,
+      "step": 9348
+    },
+    {
+      "epoch": 0.6491459519511179,
+      "grad_norm": 3.3004858462260085,
+      "learning_rate": 2.895716838643956e-06,
+      "loss": 0.3785,
+      "step": 9349
+    },
+    {
+      "epoch": 0.64921538675184,
+      "grad_norm": 3.8865975745398527,
+      "learning_rate": 2.8946968388798557e-06,
+      "loss": 0.4063,
+      "step": 9350
+    },
+    {
+      "epoch": 0.6492848215525622,
+      "grad_norm": 3.94926838100648,
+      "learning_rate": 2.8936769455996066e-06,
+      "loss": 0.4054,
+      "step": 9351
+    },
+    {
+      "epoch": 0.6493542563532843,
+      "grad_norm": 4.741055921822281,
+      "learning_rate": 2.892657158854795e-06,
+      "loss": 0.5535,
+      "step": 9352
+    },
+    {
+      "epoch": 0.6494236911540063,
+      "grad_norm": 3.332688508916348,
+      "learning_rate": 2.891637478696998e-06,
+      "loss": 0.3072,
+      "step": 9353
+    },
+    {
+      "epoch": 0.6494931259547285,
+      "grad_norm": 2.917056835295274,
+      "learning_rate": 2.8906179051777926e-06,
+      "loss": 0.3087,
+      "step": 9354
+    },
+    {
+      "epoch": 0.6495625607554506,
+      "grad_norm": 3.5053490858694802,
+      "learning_rate": 2.8895984383487475e-06,
+      "loss": 0.387,
+      "step": 9355
+    },
+    {
+      "epoch": 0.6496319955561728,
+      "grad_norm": 3.805396138769203,
+      "learning_rate": 2.888579078261422e-06,
+      "loss": 0.3763,
+      "step": 9356
+    },
+    {
+      "epoch": 0.6497014303568949,
+      "grad_norm": 3.955696978857148,
+      "learning_rate": 2.88755982496738e-06,
+      "loss": 0.3298,
+      "step": 9357
+    },
+    {
+      "epoch": 0.649770865157617,
+      "grad_norm": 5.121203581236098,
+      "learning_rate": 2.886540678518174e-06,
+      "loss": 0.8178,
+      "step": 9358
+    },
+    {
+      "epoch": 0.6498402999583391,
+      "grad_norm": 3.420584347673197,
+      "learning_rate": 2.885521638965344e-06,
+      "loss": 0.2548,
+      "step": 9359
+    },
+    {
+      "epoch": 0.6499097347590612,
+      "grad_norm": 4.166473809225999,
+      "learning_rate": 2.884502706360439e-06,
+      "loss": 0.4677,
+      "step": 9360
+    },
+    {
+      "epoch": 0.6499791695597834,
+      "grad_norm": 3.128657149499682,
+      "learning_rate": 2.883483880754996e-06,
+      "loss": 0.3595,
+      "step": 9361
+    },
+    {
+      "epoch": 0.6500486043605055,
+      "grad_norm": 2.308489802678098,
+      "learning_rate": 2.8824651622005397e-06,
+      "loss": 0.0821,
+      "step": 9362
+    },
+    {
+      "epoch": 0.6501180391612276,
+      "grad_norm": 4.257329271348108,
+      "learning_rate": 2.8814465507486e-06,
+      "loss": 0.4832,
+      "step": 9363
+    },
+    {
+      "epoch": 0.6501874739619498,
+      "grad_norm": 3.298318149970269,
+      "learning_rate": 2.880428046450697e-06,
+      "loss": 0.3627,
+      "step": 9364
+    },
+    {
+      "epoch": 0.6502569087626718,
+      "grad_norm": 3.051162484634395,
+      "learning_rate": 2.879409649358345e-06,
+      "loss": 0.2751,
+      "step": 9365
+    },
+    {
+      "epoch": 0.650326343563394,
+      "grad_norm": 3.86974315888777,
+      "learning_rate": 2.8783913595230534e-06,
+      "loss": 0.3285,
+      "step": 9366
+    },
+    {
+      "epoch": 0.6503957783641161,
+      "grad_norm": 5.004707192382065,
+      "learning_rate": 2.8773731769963258e-06,
+      "loss": 0.526,
+      "step": 9367
+    },
+    {
+      "epoch": 0.6504652131648382,
+      "grad_norm": 3.2439490702939455,
+      "learning_rate": 2.8763551018296613e-06,
+      "loss": 0.2709,
+      "step": 9368
+    },
+    {
+      "epoch": 0.6505346479655604,
+      "grad_norm": 3.3770258010722896,
+      "learning_rate": 2.875337134074553e-06,
+      "loss": 0.403,
+      "step": 9369
+    },
+    {
+      "epoch": 0.6506040827662825,
+      "grad_norm": 4.6734116230833385,
+      "learning_rate": 2.874319273782488e-06,
+      "loss": 0.3776,
+      "step": 9370
+    },
+    {
+      "epoch": 0.6506735175670045,
+      "grad_norm": 5.937027780975314,
+      "learning_rate": 2.8733015210049462e-06,
+      "loss": 0.743,
+      "step": 9371
+    },
+    {
+      "epoch": 0.6507429523677267,
+      "grad_norm": 3.737889984245252,
+      "learning_rate": 2.872283875793411e-06,
+      "loss": 0.4344,
+      "step": 9372
+    },
+    {
+      "epoch": 0.6508123871684488,
+      "grad_norm": 4.634185624799447,
+      "learning_rate": 2.871266338199348e-06,
+      "loss": 0.5887,
+      "step": 9373
+    },
+    {
+      "epoch": 0.650881821969171,
+      "grad_norm": 4.940839812456883,
+      "learning_rate": 2.870248908274223e-06,
+      "loss": 0.4494,
+      "step": 9374
+    },
+    {
+      "epoch": 0.6509512567698931,
+      "grad_norm": 7.047144938774491,
+      "learning_rate": 2.8692315860695e-06,
+      "loss": 0.8221,
+      "step": 9375
+    },
+    {
+      "epoch": 0.6510206915706152,
+      "grad_norm": 3.4407492783022375,
+      "learning_rate": 2.868214371636632e-06,
+      "loss": 0.3819,
+      "step": 9376
+    },
+    {
+      "epoch": 0.6510901263713373,
+      "grad_norm": 3.0803540924676724,
+      "learning_rate": 2.8671972650270695e-06,
+      "loss": 0.3629,
+      "step": 9377
+    },
+    {
+      "epoch": 0.6511595611720594,
+      "grad_norm": 4.978125860415443,
+      "learning_rate": 2.8661802662922556e-06,
+      "loss": 0.6539,
+      "step": 9378
+    },
+    {
+      "epoch": 0.6512289959727816,
+      "grad_norm": 4.502976219657025,
+      "learning_rate": 2.865163375483629e-06,
+      "loss": 0.505,
+      "step": 9379
+    },
+    {
+      "epoch": 0.6512984307735037,
+      "grad_norm": 3.9076168663370994,
+      "learning_rate": 2.8641465926526235e-06,
+      "loss": 0.2995,
+      "step": 9380
+    },
+    {
+      "epoch": 0.6513678655742258,
+      "grad_norm": 3.752686275580926,
+      "learning_rate": 2.8631299178506656e-06,
+      "loss": 0.4979,
+      "step": 9381
+    },
+    {
+      "epoch": 0.651437300374948,
+      "grad_norm": 3.2833461122910363,
+      "learning_rate": 2.8621133511291792e-06,
+      "loss": 0.2101,
+      "step": 9382
+    },
+    {
+      "epoch": 0.65150673517567,
+      "grad_norm": 4.614988181049766,
+      "learning_rate": 2.8610968925395805e-06,
+      "loss": 0.6809,
+      "step": 9383
+    },
+    {
+      "epoch": 0.6515761699763921,
+      "grad_norm": 3.271682433424328,
+      "learning_rate": 2.86008054213328e-06,
+      "loss": 0.2899,
+      "step": 9384
+    },
+    {
+      "epoch": 0.6516456047771143,
+      "grad_norm": 4.42284071177736,
+      "learning_rate": 2.8590642999616823e-06,
+      "loss": 0.5159,
+      "step": 9385
+    },
+    {
+      "epoch": 0.6517150395778364,
+      "grad_norm": 5.7427574701124735,
+      "learning_rate": 2.8580481660761944e-06,
+      "loss": 0.7401,
+      "step": 9386
+    },
+    {
+      "epoch": 0.6517844743785586,
+      "grad_norm": 4.245649908746984,
+      "learning_rate": 2.8570321405282043e-06,
+      "loss": 0.4425,
+      "step": 9387
+    },
+    {
+      "epoch": 0.6518539091792807,
+      "grad_norm": 3.2752949081075173,
+      "learning_rate": 2.856016223369101e-06,
+      "loss": 0.3203,
+      "step": 9388
+    },
+    {
+      "epoch": 0.6519233439800027,
+      "grad_norm": 3.6892365387933337,
+      "learning_rate": 2.855000414650273e-06,
+      "loss": 0.2599,
+      "step": 9389
+    },
+    {
+      "epoch": 0.6519927787807249,
+      "grad_norm": 3.3998215643764187,
+      "learning_rate": 2.8539847144230992e-06,
+      "loss": 0.4094,
+      "step": 9390
+    },
+    {
+      "epoch": 0.652062213581447,
+      "grad_norm": 4.155136311122048,
+      "learning_rate": 2.852969122738946e-06,
+      "loss": 0.5109,
+      "step": 9391
+    },
+    {
+      "epoch": 0.6521316483821692,
+      "grad_norm": 3.1716562335203293,
+      "learning_rate": 2.8519536396491877e-06,
+      "loss": 0.313,
+      "step": 9392
+    },
+    {
+      "epoch": 0.6522010831828913,
+      "grad_norm": 4.112345617261162,
+      "learning_rate": 2.8509382652051843e-06,
+      "loss": 0.3294,
+      "step": 9393
+    },
+    {
+      "epoch": 0.6522705179836134,
+      "grad_norm": 4.065197620457973,
+      "learning_rate": 2.849922999458289e-06,
+      "loss": 0.4338,
+      "step": 9394
+    },
+    {
+      "epoch": 0.6523399527843355,
+      "grad_norm": 2.8851051699457133,
+      "learning_rate": 2.8489078424598565e-06,
+      "loss": 0.4062,
+      "step": 9395
+    },
+    {
+      "epoch": 0.6524093875850576,
+      "grad_norm": 4.1469101304846445,
+      "learning_rate": 2.847892794261231e-06,
+      "loss": 0.3137,
+      "step": 9396
+    },
+    {
+      "epoch": 0.6524788223857797,
+      "grad_norm": 6.404058408765919,
+      "learning_rate": 2.846877854913753e-06,
+      "loss": 0.689,
+      "step": 9397
+    },
+    {
+      "epoch": 0.6525482571865019,
+      "grad_norm": 4.072306161301421,
+      "learning_rate": 2.845863024468757e-06,
+      "loss": 0.4959,
+      "step": 9398
+    },
+    {
+      "epoch": 0.652617691987224,
+      "grad_norm": 4.637139648833932,
+      "learning_rate": 2.844848302977572e-06,
+      "loss": 0.5147,
+      "step": 9399
+    },
+    {
+      "epoch": 0.6526871267879462,
+      "grad_norm": 3.8949063037172493,
+      "learning_rate": 2.8438336904915186e-06,
+      "loss": 0.355,
+      "step": 9400
+    },
+    {
+      "epoch": 0.6527565615886682,
+      "grad_norm": 3.157622893164029,
+      "learning_rate": 2.8428191870619214e-06,
+      "loss": 0.3536,
+      "step": 9401
+    },
+    {
+      "epoch": 0.6528259963893903,
+      "grad_norm": 3.217229612491774,
+      "learning_rate": 2.841804792740086e-06,
+      "loss": 0.3362,
+      "step": 9402
+    },
+    {
+      "epoch": 0.6528954311901125,
+      "grad_norm": 4.614250766195664,
+      "learning_rate": 2.840790507577321e-06,
+      "loss": 0.4752,
+      "step": 9403
+    },
+    {
+      "epoch": 0.6529648659908346,
+      "grad_norm": 4.654944607537033,
+      "learning_rate": 2.8397763316249326e-06,
+      "loss": 0.6321,
+      "step": 9404
+    },
+    {
+      "epoch": 0.6530343007915568,
+      "grad_norm": 9.08052888358467,
+      "learning_rate": 2.8387622649342085e-06,
+      "loss": 0.4345,
+      "step": 9405
+    },
+    {
+      "epoch": 0.6531037355922789,
+      "grad_norm": 3.7585305374610862,
+      "learning_rate": 2.837748307556446e-06,
+      "loss": 0.5195,
+      "step": 9406
+    },
+    {
+      "epoch": 0.6531731703930009,
+      "grad_norm": 4.222379292556262,
+      "learning_rate": 2.836734459542927e-06,
+      "loss": 0.3009,
+      "step": 9407
+    },
+    {
+      "epoch": 0.6532426051937231,
+      "grad_norm": 3.3073759503276037,
+      "learning_rate": 2.8357207209449313e-06,
+      "loss": 0.4324,
+      "step": 9408
+    },
+    {
+      "epoch": 0.6533120399944452,
+      "grad_norm": 4.3155187687353385,
+      "learning_rate": 2.834707091813732e-06,
+      "loss": 0.3764,
+      "step": 9409
+    },
+    {
+      "epoch": 0.6533814747951673,
+      "grad_norm": 4.510995642625837,
+      "learning_rate": 2.8336935722005978e-06,
+      "loss": 0.5204,
+      "step": 9410
+    },
+    {
+      "epoch": 0.6534509095958895,
+      "grad_norm": 4.415632176963448,
+      "learning_rate": 2.8326801621567916e-06,
+      "loss": 0.3294,
+      "step": 9411
+    },
+    {
+      "epoch": 0.6535203443966116,
+      "grad_norm": 5.107058864692759,
+      "learning_rate": 2.8316668617335707e-06,
+      "loss": 0.604,
+      "step": 9412
+    },
+    {
+      "epoch": 0.6535897791973337,
+      "grad_norm": 3.500325629384166,
+      "learning_rate": 2.8306536709821857e-06,
+      "loss": 0.2805,
+      "step": 9413
+    },
+    {
+      "epoch": 0.6536592139980558,
+      "grad_norm": 3.939608135289932,
+      "learning_rate": 2.8296405899538816e-06,
+      "loss": 0.3309,
+      "step": 9414
+    },
+    {
+      "epoch": 0.6537286487987779,
+      "grad_norm": 3.371634977330208,
+      "learning_rate": 2.8286276186999045e-06,
+      "loss": 0.3919,
+      "step": 9415
+    },
+    {
+      "epoch": 0.6537980835995001,
+      "grad_norm": 4.065187481058551,
+      "learning_rate": 2.8276147572714837e-06,
+      "loss": 0.4225,
+      "step": 9416
+    },
+    {
+      "epoch": 0.6538675184002222,
+      "grad_norm": 4.5360742353141745,
+      "learning_rate": 2.8266020057198473e-06,
+      "loss": 0.5915,
+      "step": 9417
+    },
+    {
+      "epoch": 0.6539369532009444,
+      "grad_norm": 3.961525376429098,
+      "learning_rate": 2.8255893640962272e-06,
+      "loss": 0.4723,
+      "step": 9418
+    },
+    {
+      "epoch": 0.6540063880016664,
+      "grad_norm": 4.028084661502884,
+      "learning_rate": 2.8245768324518343e-06,
+      "loss": 0.4204,
+      "step": 9419
+    },
+    {
+      "epoch": 0.6540758228023885,
+      "grad_norm": 4.390266134362424,
+      "learning_rate": 2.8235644108378814e-06,
+      "loss": 0.4997,
+      "step": 9420
+    },
+    {
+      "epoch": 0.6541452576031107,
+      "grad_norm": 3.196205460919116,
+      "learning_rate": 2.822552099305579e-06,
+      "loss": 0.3422,
+      "step": 9421
+    },
+    {
+      "epoch": 0.6542146924038328,
+      "grad_norm": 4.317644580182893,
+      "learning_rate": 2.8215398979061302e-06,
+      "loss": 0.6501,
+      "step": 9422
+    },
+    {
+      "epoch": 0.654284127204555,
+      "grad_norm": 4.746119261789729,
+      "learning_rate": 2.8205278066907242e-06,
+      "loss": 0.7082,
+      "step": 9423
+    },
+    {
+      "epoch": 0.6543535620052771,
+      "grad_norm": 4.339973239398779,
+      "learning_rate": 2.8195158257105577e-06,
+      "loss": 0.5299,
+      "step": 9424
+    },
+    {
+      "epoch": 0.6544229968059991,
+      "grad_norm": 3.9512468422616496,
+      "learning_rate": 2.818503955016813e-06,
+      "loss": 0.5035,
+      "step": 9425
+    },
+    {
+      "epoch": 0.6544924316067213,
+      "grad_norm": 3.2062485142961266,
+      "learning_rate": 2.8174921946606704e-06,
+      "loss": 0.329,
+      "step": 9426
+    },
+    {
+      "epoch": 0.6545618664074434,
+      "grad_norm": 5.023163478942884,
+      "learning_rate": 2.8164805446933018e-06,
+      "loss": 0.466,
+      "step": 9427
+    },
+    {
+      "epoch": 0.6546313012081655,
+      "grad_norm": 4.065959456524006,
+      "learning_rate": 2.815469005165877e-06,
+      "loss": 0.5718,
+      "step": 9428
+    },
+    {
+      "epoch": 0.6547007360088877,
+      "grad_norm": 4.077856629595067,
+      "learning_rate": 2.8144575761295585e-06,
+      "loss": 0.4269,
+      "step": 9429
+    },
+    {
+      "epoch": 0.6547701708096098,
+      "grad_norm": 3.8468307093769765,
+      "learning_rate": 2.813446257635502e-06,
+      "loss": 0.343,
+      "step": 9430
+    },
+    {
+      "epoch": 0.6548396056103319,
+      "grad_norm": 3.878269166011031,
+      "learning_rate": 2.8124350497348595e-06,
+      "loss": 0.4979,
+      "step": 9431
+    },
+    {
+      "epoch": 0.654909040411054,
+      "grad_norm": 3.9485188757479985,
+      "learning_rate": 2.8114239524787756e-06,
+      "loss": 0.5487,
+      "step": 9432
+    },
+    {
+      "epoch": 0.6549784752117761,
+      "grad_norm": 3.691865044168531,
+      "learning_rate": 2.8104129659183943e-06,
+      "loss": 0.2353,
+      "step": 9433
+    },
+    {
+      "epoch": 0.6550479100124983,
+      "grad_norm": 3.87956581474772,
+      "learning_rate": 2.809402090104845e-06,
+      "loss": 0.585,
+      "step": 9434
+    },
+    {
+      "epoch": 0.6551173448132204,
+      "grad_norm": 3.7261727767842414,
+      "learning_rate": 2.80839132508926e-06,
+      "loss": 0.5071,
+      "step": 9435
+    },
+    {
+      "epoch": 0.6551867796139426,
+      "grad_norm": 4.372315241923775,
+      "learning_rate": 2.807380670922765e-06,
+      "loss": 0.6274,
+      "step": 9436
+    },
+    {
+      "epoch": 0.6552562144146646,
+      "grad_norm": 3.3406027324353476,
+      "learning_rate": 2.8063701276564693e-06,
+      "loss": 0.1662,
+      "step": 9437
+    },
+    {
+      "epoch": 0.6553256492153867,
+      "grad_norm": 3.4589709672516737,
+      "learning_rate": 2.8053596953414926e-06,
+      "loss": 0.434,
+      "step": 9438
+    },
+    {
+      "epoch": 0.6553950840161089,
+      "grad_norm": 3.4633096284696436,
+      "learning_rate": 2.804349374028939e-06,
+      "loss": 0.3813,
+      "step": 9439
+    },
+    {
+      "epoch": 0.655464518816831,
+      "grad_norm": 4.795957499113184,
+      "learning_rate": 2.8033391637699104e-06,
+      "loss": 0.7162,
+      "step": 9440
+    },
+    {
+      "epoch": 0.6555339536175531,
+      "grad_norm": 3.1346465241284958,
+      "learning_rate": 2.8023290646155e-06,
+      "loss": 0.3977,
+      "step": 9441
+    },
+    {
+      "epoch": 0.6556033884182753,
+      "grad_norm": 4.0510520458471735,
+      "learning_rate": 2.8013190766167996e-06,
+      "loss": 0.5197,
+      "step": 9442
+    },
+    {
+      "epoch": 0.6556728232189973,
+      "grad_norm": 3.1486789912602884,
+      "learning_rate": 2.800309199824892e-06,
+      "loss": 0.2936,
+      "step": 9443
+    },
+    {
+      "epoch": 0.6557422580197195,
+      "grad_norm": 2.854366861333353,
+      "learning_rate": 2.7992994342908553e-06,
+      "loss": 0.3496,
+      "step": 9444
+    },
+    {
+      "epoch": 0.6558116928204416,
+      "grad_norm": 5.911400212759046,
+      "learning_rate": 2.798289780065764e-06,
+      "loss": 0.7833,
+      "step": 9445
+    },
+    {
+      "epoch": 0.6558811276211637,
+      "grad_norm": 4.92580756945385,
+      "learning_rate": 2.7972802372006817e-06,
+      "loss": 0.6177,
+      "step": 9446
+    },
+    {
+      "epoch": 0.6559505624218859,
+      "grad_norm": 2.1508113209527675,
+      "learning_rate": 2.796270805746677e-06,
+      "loss": 0.1657,
+      "step": 9447
+    },
+    {
+      "epoch": 0.656019997222608,
+      "grad_norm": 3.391681722784454,
+      "learning_rate": 2.7952614857547988e-06,
+      "loss": 0.3564,
+      "step": 9448
+    },
+    {
+      "epoch": 0.6560894320233301,
+      "grad_norm": 4.561336486198823,
+      "learning_rate": 2.7942522772760972e-06,
+      "loss": 0.5534,
+      "step": 9449
+    },
+    {
+      "epoch": 0.6561588668240522,
+      "grad_norm": 3.5379538366712118,
+      "learning_rate": 2.7932431803616222e-06,
+      "loss": 0.4115,
+      "step": 9450
+    },
+    {
+      "epoch": 0.6562283016247743,
+      "grad_norm": 4.173720774769059,
+      "learning_rate": 2.792234195062412e-06,
+      "loss": 0.591,
+      "step": 9451
+    },
+    {
+      "epoch": 0.6562977364254965,
+      "grad_norm": 4.828417939356,
+      "learning_rate": 2.791225321429494e-06,
+      "loss": 0.6814,
+      "step": 9452
+    },
+    {
+      "epoch": 0.6563671712262186,
+      "grad_norm": 4.30530759688471,
+      "learning_rate": 2.790216559513902e-06,
+      "loss": 0.4877,
+      "step": 9453
+    },
+    {
+      "epoch": 0.6564366060269406,
+      "grad_norm": 4.233169351653628,
+      "learning_rate": 2.789207909366657e-06,
+      "loss": 0.4658,
+      "step": 9454
+    },
+    {
+      "epoch": 0.6565060408276628,
+      "grad_norm": 3.4197145306518486,
+      "learning_rate": 2.788199371038771e-06,
+      "loss": 0.4003,
+      "step": 9455
+    },
+    {
+      "epoch": 0.6565754756283849,
+      "grad_norm": 3.6696890831276456,
+      "learning_rate": 2.78719094458126e-06,
+      "loss": 0.3445,
+      "step": 9456
+    },
+    {
+      "epoch": 0.6566449104291071,
+      "grad_norm": 3.6881360678497104,
+      "learning_rate": 2.7861826300451265e-06,
+      "loss": 0.3285,
+      "step": 9457
+    },
+    {
+      "epoch": 0.6567143452298292,
+      "grad_norm": 3.8447697234504146,
+      "learning_rate": 2.7851744274813715e-06,
+      "loss": 0.6384,
+      "step": 9458
+    },
+    {
+      "epoch": 0.6567837800305513,
+      "grad_norm": 4.010011938636868,
+      "learning_rate": 2.7841663369409865e-06,
+      "loss": 0.5725,
+      "step": 9459
+    },
+    {
+      "epoch": 0.6568532148312735,
+      "grad_norm": 3.2609916663868344,
+      "learning_rate": 2.7831583584749617e-06,
+      "loss": 0.4202,
+      "step": 9460
+    },
+    {
+      "epoch": 0.6569226496319955,
+      "grad_norm": 4.986409902104478,
+      "learning_rate": 2.782150492134278e-06,
+      "loss": 0.3977,
+      "step": 9461
+    },
+    {
+      "epoch": 0.6569920844327177,
+      "grad_norm": 3.5874794718612164,
+      "learning_rate": 2.781142737969913e-06,
+      "loss": 0.5873,
+      "step": 9462
+    },
+    {
+      "epoch": 0.6570615192334398,
+      "grad_norm": 5.393404136180233,
+      "learning_rate": 2.780135096032838e-06,
+      "loss": 0.5011,
+      "step": 9463
+    },
+    {
+      "epoch": 0.6571309540341619,
+      "grad_norm": 5.613559341184744,
+      "learning_rate": 2.779127566374015e-06,
+      "loss": 0.5624,
+      "step": 9464
+    },
+    {
+      "epoch": 0.6572003888348841,
+      "grad_norm": 4.324332084541916,
+      "learning_rate": 2.778120149044411e-06,
+      "loss": 0.372,
+      "step": 9465
+    },
+    {
+      "epoch": 0.6572698236356062,
+      "grad_norm": 3.82158140877189,
+      "learning_rate": 2.777112844094972e-06,
+      "loss": 0.2801,
+      "step": 9466
+    },
+    {
+      "epoch": 0.6573392584363282,
+      "grad_norm": 4.280744272105727,
+      "learning_rate": 2.7761056515766514e-06,
+      "loss": 0.535,
+      "step": 9467
+    },
+    {
+      "epoch": 0.6574086932370504,
+      "grad_norm": 3.034637235214766,
+      "learning_rate": 2.775098571540393e-06,
+      "loss": 0.2876,
+      "step": 9468
+    },
+    {
+      "epoch": 0.6574781280377725,
+      "grad_norm": 3.8984467380337797,
+      "learning_rate": 2.774091604037127e-06,
+      "loss": 0.3841,
+      "step": 9469
+    },
+    {
+      "epoch": 0.6575475628384947,
+      "grad_norm": 3.612568467326305,
+      "learning_rate": 2.7730847491177903e-06,
+      "loss": 0.3854,
+      "step": 9470
+    },
+    {
+      "epoch": 0.6576169976392168,
+      "grad_norm": 2.4300020398700846,
+      "learning_rate": 2.7720780068333074e-06,
+      "loss": 0.2166,
+      "step": 9471
+    },
+    {
+      "epoch": 0.6576864324399389,
+      "grad_norm": 3.3605197919468237,
+      "learning_rate": 2.7710713772345988e-06,
+      "loss": 0.4421,
+      "step": 9472
+    },
+    {
+      "epoch": 0.657755867240661,
+      "grad_norm": 4.574376599185595,
+      "learning_rate": 2.7700648603725767e-06,
+      "loss": 0.3837,
+      "step": 9473
+    },
+    {
+      "epoch": 0.6578253020413831,
+      "grad_norm": 3.5178147516700258,
+      "learning_rate": 2.7690584562981505e-06,
+      "loss": 0.3384,
+      "step": 9474
+    },
+    {
+      "epoch": 0.6578947368421053,
+      "grad_norm": 3.594964246235022,
+      "learning_rate": 2.7680521650622216e-06,
+      "loss": 0.4173,
+      "step": 9475
+    },
+    {
+      "epoch": 0.6579641716428274,
+      "grad_norm": 4.560789712788866,
+      "learning_rate": 2.7670459867156925e-06,
+      "loss": 0.4824,
+      "step": 9476
+    },
+    {
+      "epoch": 0.6580336064435495,
+      "grad_norm": 4.260460039277188,
+      "learning_rate": 2.766039921309448e-06,
+      "loss": 0.5662,
+      "step": 9477
+    },
+    {
+      "epoch": 0.6581030412442717,
+      "grad_norm": 4.351240083279505,
+      "learning_rate": 2.765033968894375e-06,
+      "loss": 0.4114,
+      "step": 9478
+    },
+    {
+      "epoch": 0.6581724760449937,
+      "grad_norm": 4.273458786860868,
+      "learning_rate": 2.764028129521359e-06,
+      "loss": 0.6259,
+      "step": 9479
+    },
+    {
+      "epoch": 0.6582419108457159,
+      "grad_norm": 3.1936253346742953,
+      "learning_rate": 2.7630224032412677e-06,
+      "loss": 0.2226,
+      "step": 9480
+    },
+    {
+      "epoch": 0.658311345646438,
+      "grad_norm": 4.1660025582388585,
+      "learning_rate": 2.76201679010497e-06,
+      "loss": 0.573,
+      "step": 9481
+    },
+    {
+      "epoch": 0.6583807804471601,
+      "grad_norm": 4.245996062120205,
+      "learning_rate": 2.7610112901633325e-06,
+      "loss": 0.296,
+      "step": 9482
+    },
+    {
+      "epoch": 0.6584502152478823,
+      "grad_norm": 4.139649590303785,
+      "learning_rate": 2.760005903467213e-06,
+      "loss": 0.5474,
+      "step": 9483
+    },
+    {
+      "epoch": 0.6585196500486044,
+      "grad_norm": 4.1500634222347506,
+      "learning_rate": 2.7590006300674553e-06,
+      "loss": 0.5,
+      "step": 9484
+    },
+    {
+      "epoch": 0.6585890848493264,
+      "grad_norm": 3.1030816063227964,
+      "learning_rate": 2.7579954700149124e-06,
+      "loss": 0.3219,
+      "step": 9485
+    },
+    {
+      "epoch": 0.6586585196500486,
+      "grad_norm": 2.388486828059138,
+      "learning_rate": 2.7569904233604215e-06,
+      "loss": 0.1244,
+      "step": 9486
+    },
+    {
+      "epoch": 0.6587279544507707,
+      "grad_norm": 3.2980890652964208,
+      "learning_rate": 2.7559854901548165e-06,
+      "loss": 0.3751,
+      "step": 9487
+    },
+    {
+      "epoch": 0.6587973892514929,
+      "grad_norm": 4.057768517231572,
+      "learning_rate": 2.754980670448927e-06,
+      "loss": 0.3973,
+      "step": 9488
+    },
+    {
+      "epoch": 0.658866824052215,
+      "grad_norm": 4.586690125122526,
+      "learning_rate": 2.753975964293574e-06,
+      "loss": 0.5271,
+      "step": 9489
+    },
+    {
+      "epoch": 0.658936258852937,
+      "grad_norm": 4.714494701496468,
+      "learning_rate": 2.752971371739575e-06,
+      "loss": 0.5462,
+      "step": 9490
+    },
+    {
+      "epoch": 0.6590056936536592,
+      "grad_norm": 2.64844438023939,
+      "learning_rate": 2.7519668928377417e-06,
+      "loss": 0.2088,
+      "step": 9491
+    },
+    {
+      "epoch": 0.6590751284543813,
+      "grad_norm": 4.122080466376102,
+      "learning_rate": 2.75096252763888e-06,
+      "loss": 0.4102,
+      "step": 9492
+    },
+    {
+      "epoch": 0.6591445632551035,
+      "grad_norm": 3.4759312255850365,
+      "learning_rate": 2.749958276193787e-06,
+      "loss": 0.3943,
+      "step": 9493
+    },
+    {
+      "epoch": 0.6592139980558256,
+      "grad_norm": 2.1566262098524267,
+      "learning_rate": 2.748954138553259e-06,
+      "loss": 0.1623,
+      "step": 9494
+    },
+    {
+      "epoch": 0.6592834328565477,
+      "grad_norm": 2.810574917776052,
+      "learning_rate": 2.747950114768082e-06,
+      "loss": 0.1843,
+      "step": 9495
+    },
+    {
+      "epoch": 0.6593528676572699,
+      "grad_norm": 2.1790045492095036,
+      "learning_rate": 2.7469462048890418e-06,
+      "loss": 0.1607,
+      "step": 9496
+    },
+    {
+      "epoch": 0.6594223024579919,
+      "grad_norm": 4.632378980619174,
+      "learning_rate": 2.745942408966914e-06,
+      "loss": 0.4814,
+      "step": 9497
+    },
+    {
+      "epoch": 0.659491737258714,
+      "grad_norm": 4.199321776653104,
+      "learning_rate": 2.7449387270524654e-06,
+      "loss": 0.4418,
+      "step": 9498
+    },
+    {
+      "epoch": 0.6595611720594362,
+      "grad_norm": 3.4535412185618735,
+      "learning_rate": 2.7439351591964658e-06,
+      "loss": 0.2742,
+      "step": 9499
+    },
+    {
+      "epoch": 0.6596306068601583,
+      "grad_norm": 4.484573202722018,
+      "learning_rate": 2.742931705449673e-06,
+      "loss": 0.5249,
+      "step": 9500
+    },
+    {
+      "epoch": 0.6597000416608805,
+      "grad_norm": 3.330238970247279,
+      "learning_rate": 2.741928365862841e-06,
+      "loss": 0.3292,
+      "step": 9501
+    },
+    {
+      "epoch": 0.6597694764616026,
+      "grad_norm": 3.492060577913008,
+      "learning_rate": 2.7409251404867166e-06,
+      "loss": 0.4112,
+      "step": 9502
+    },
+    {
+      "epoch": 0.6598389112623246,
+      "grad_norm": 3.926728572281711,
+      "learning_rate": 2.7399220293720423e-06,
+      "loss": 0.3863,
+      "step": 9503
+    },
+    {
+      "epoch": 0.6599083460630468,
+      "grad_norm": 5.5201527395032555,
+      "learning_rate": 2.738919032569555e-06,
+      "loss": 0.7849,
+      "step": 9504
+    },
+    {
+      "epoch": 0.6599777808637689,
+      "grad_norm": 2.529980733017903,
+      "learning_rate": 2.7379161501299836e-06,
+      "loss": 0.1904,
+      "step": 9505
+    },
+    {
+      "epoch": 0.6600472156644911,
+      "grad_norm": 3.798546529842414,
+      "learning_rate": 2.736913382104054e-06,
+      "loss": 0.4624,
+      "step": 9506
+    },
+    {
+      "epoch": 0.6601166504652132,
+      "grad_norm": 3.850280368946491,
+      "learning_rate": 2.7359107285424826e-06,
+      "loss": 0.4397,
+      "step": 9507
+    },
+    {
+      "epoch": 0.6601860852659353,
+      "grad_norm": 4.016073474375954,
+      "learning_rate": 2.734908189495988e-06,
+      "loss": 0.4539,
+      "step": 9508
+    },
+    {
+      "epoch": 0.6602555200666574,
+      "grad_norm": 3.7905226266710943,
+      "learning_rate": 2.7339057650152723e-06,
+      "loss": 0.5917,
+      "step": 9509
+    },
+    {
+      "epoch": 0.6603249548673795,
+      "grad_norm": 3.1879243721701775,
+      "learning_rate": 2.732903455151037e-06,
+      "loss": 0.3622,
+      "step": 9510
+    },
+    {
+      "epoch": 0.6603943896681016,
+      "grad_norm": 3.850743682813295,
+      "learning_rate": 2.7319012599539836e-06,
+      "loss": 0.4112,
+      "step": 9511
+    },
+    {
+      "epoch": 0.6604638244688238,
+      "grad_norm": 2.6806593864397343,
+      "learning_rate": 2.730899179474796e-06,
+      "loss": 0.2308,
+      "step": 9512
+    },
+    {
+      "epoch": 0.6605332592695459,
+      "grad_norm": 3.947650848410535,
+      "learning_rate": 2.729897213764158e-06,
+      "loss": 0.3364,
+      "step": 9513
+    },
+    {
+      "epoch": 0.6606026940702681,
+      "grad_norm": 3.2865625183857228,
+      "learning_rate": 2.728895362872752e-06,
+      "loss": 0.3076,
+      "step": 9514
+    },
+    {
+      "epoch": 0.6606721288709901,
+      "grad_norm": 3.3879047214706217,
+      "learning_rate": 2.727893626851249e-06,
+      "loss": 0.2817,
+      "step": 9515
+    },
+    {
+      "epoch": 0.6607415636717122,
+      "grad_norm": 4.020696075693928,
+      "learning_rate": 2.726892005750315e-06,
+      "loss": 0.4665,
+      "step": 9516
+    },
+    {
+      "epoch": 0.6608109984724344,
+      "grad_norm": 3.6926961145755106,
+      "learning_rate": 2.7258904996206113e-06,
+      "loss": 0.2341,
+      "step": 9517
+    },
+    {
+      "epoch": 0.6608804332731565,
+      "grad_norm": 4.913795642498241,
+      "learning_rate": 2.7248891085127927e-06,
+      "loss": 0.5222,
+      "step": 9518
+    },
+    {
+      "epoch": 0.6609498680738787,
+      "grad_norm": 3.957949082858564,
+      "learning_rate": 2.7238878324775087e-06,
+      "loss": 0.3418,
+      "step": 9519
+    },
+    {
+      "epoch": 0.6610193028746008,
+      "grad_norm": 4.545325193675648,
+      "learning_rate": 2.7228866715654024e-06,
+      "loss": 0.6244,
+      "step": 9520
+    },
+    {
+      "epoch": 0.6610887376753228,
+      "grad_norm": 4.951857946006939,
+      "learning_rate": 2.721885625827112e-06,
+      "loss": 0.5444,
+      "step": 9521
+    },
+    {
+      "epoch": 0.661158172476045,
+      "grad_norm": 2.4536647070951934,
+      "learning_rate": 2.7208846953132685e-06,
+      "loss": 0.1821,
+      "step": 9522
+    },
+    {
+      "epoch": 0.6612276072767671,
+      "grad_norm": 3.417244302680744,
+      "learning_rate": 2.7198838800744974e-06,
+      "loss": 0.2703,
+      "step": 9523
+    },
+    {
+      "epoch": 0.6612970420774892,
+      "grad_norm": 4.459790982249659,
+      "learning_rate": 2.7188831801614202e-06,
+      "loss": 0.3464,
+      "step": 9524
+    },
+    {
+      "epoch": 0.6613664768782114,
+      "grad_norm": 4.200540792354211,
+      "learning_rate": 2.7178825956246503e-06,
+      "loss": 0.7596,
+      "step": 9525
+    },
+    {
+      "epoch": 0.6614359116789335,
+      "grad_norm": 3.2995915762971153,
+      "learning_rate": 2.7168821265147965e-06,
+      "loss": 0.2297,
+      "step": 9526
+    },
+    {
+      "epoch": 0.6615053464796556,
+      "grad_norm": 2.894082688728053,
+      "learning_rate": 2.7158817728824593e-06,
+      "loss": 0.2135,
+      "step": 9527
+    },
+    {
+      "epoch": 0.6615747812803777,
+      "grad_norm": 2.6857276171046776,
+      "learning_rate": 2.7148815347782392e-06,
+      "loss": 0.256,
+      "step": 9528
+    },
+    {
+      "epoch": 0.6616442160810998,
+      "grad_norm": 3.4579157197694914,
+      "learning_rate": 2.7138814122527274e-06,
+      "loss": 0.3779,
+      "step": 9529
+    },
+    {
+      "epoch": 0.661713650881822,
+      "grad_norm": 3.1637071704333195,
+      "learning_rate": 2.7128814053565027e-06,
+      "loss": 0.1874,
+      "step": 9530
+    },
+    {
+      "epoch": 0.6617830856825441,
+      "grad_norm": 3.310116618892332,
+      "learning_rate": 2.7118815141401504e-06,
+      "loss": 0.4021,
+      "step": 9531
+    },
+    {
+      "epoch": 0.6618525204832663,
+      "grad_norm": 4.516345195196129,
+      "learning_rate": 2.7108817386542417e-06,
+      "loss": 0.5105,
+      "step": 9532
+    },
+    {
+      "epoch": 0.6619219552839883,
+      "grad_norm": 3.473607401592932,
+      "learning_rate": 2.7098820789493453e-06,
+      "loss": 0.4443,
+      "step": 9533
+    },
+    {
+      "epoch": 0.6619913900847104,
+      "grad_norm": 3.654458863424914,
+      "learning_rate": 2.708882535076021e-06,
+      "loss": 0.4112,
+      "step": 9534
+    },
+    {
+      "epoch": 0.6620608248854326,
+      "grad_norm": 3.659929473515625,
+      "learning_rate": 2.7078831070848254e-06,
+      "loss": 0.3556,
+      "step": 9535
+    },
+    {
+      "epoch": 0.6621302596861547,
+      "grad_norm": 3.68895679562386,
+      "learning_rate": 2.7068837950263093e-06,
+      "loss": 0.3124,
+      "step": 9536
+    },
+    {
+      "epoch": 0.6621996944868769,
+      "grad_norm": 6.135996576626905,
+      "learning_rate": 2.7058845989510156e-06,
+      "loss": 1.1037,
+      "step": 9537
+    },
+    {
+      "epoch": 0.662269129287599,
+      "grad_norm": 4.050658674131581,
+      "learning_rate": 2.704885518909483e-06,
+      "loss": 0.3642,
+      "step": 9538
+    },
+    {
+      "epoch": 0.662338564088321,
+      "grad_norm": 2.7595776298732235,
+      "learning_rate": 2.703886554952242e-06,
+      "loss": 0.2577,
+      "step": 9539
+    },
+    {
+      "epoch": 0.6624079988890432,
+      "grad_norm": 4.200284803503475,
+      "learning_rate": 2.7028877071298243e-06,
+      "loss": 0.4466,
+      "step": 9540
+    },
+    {
+      "epoch": 0.6624774336897653,
+      "grad_norm": 3.4055673911198623,
+      "learning_rate": 2.701888975492745e-06,
+      "loss": 0.2973,
+      "step": 9541
+    },
+    {
+      "epoch": 0.6625468684904874,
+      "grad_norm": 3.981354654422262,
+      "learning_rate": 2.7008903600915194e-06,
+      "loss": 0.5716,
+      "step": 9542
+    },
+    {
+      "epoch": 0.6626163032912096,
+      "grad_norm": 4.178806245999192,
+      "learning_rate": 2.699891860976661e-06,
+      "loss": 0.611,
+      "step": 9543
+    },
+    {
+      "epoch": 0.6626857380919317,
+      "grad_norm": 3.38498200024003,
+      "learning_rate": 2.6988934781986686e-06,
+      "loss": 0.3335,
+      "step": 9544
+    },
+    {
+      "epoch": 0.6627551728926538,
+      "grad_norm": 5.794123213895749,
+      "learning_rate": 2.697895211808037e-06,
+      "loss": 0.6271,
+      "step": 9545
+    },
+    {
+      "epoch": 0.6628246076933759,
+      "grad_norm": 3.143598899451623,
+      "learning_rate": 2.6968970618552638e-06,
+      "loss": 0.2275,
+      "step": 9546
+    },
+    {
+      "epoch": 0.662894042494098,
+      "grad_norm": 5.123682207267904,
+      "learning_rate": 2.6958990283908305e-06,
+      "loss": 0.495,
+      "step": 9547
+    },
+    {
+      "epoch": 0.6629634772948202,
+      "grad_norm": 2.5973732475025515,
+      "learning_rate": 2.694901111465217e-06,
+      "loss": 0.2782,
+      "step": 9548
+    },
+    {
+      "epoch": 0.6630329120955423,
+      "grad_norm": 4.7580558291304165,
+      "learning_rate": 2.6939033111288964e-06,
+      "loss": 0.6355,
+      "step": 9549
+    },
+    {
+      "epoch": 0.6631023468962645,
+      "grad_norm": 4.511425975517061,
+      "learning_rate": 2.6929056274323373e-06,
+      "loss": 0.6073,
+      "step": 9550
+    },
+    {
+      "epoch": 0.6631717816969865,
+      "grad_norm": 6.531573305544805,
+      "learning_rate": 2.6919080604260004e-06,
+      "loss": 0.6995,
+      "step": 9551
+    },
+    {
+      "epoch": 0.6632412164977086,
+      "grad_norm": 4.131580711063387,
+      "learning_rate": 2.6909106101603434e-06,
+      "loss": 0.4971,
+      "step": 9552
+    },
+    {
+      "epoch": 0.6633106512984308,
+      "grad_norm": 3.3712046380750076,
+      "learning_rate": 2.689913276685814e-06,
+      "loss": 0.3827,
+      "step": 9553
+    },
+    {
+      "epoch": 0.6633800860991529,
+      "grad_norm": 3.6905076080797934,
+      "learning_rate": 2.688916060052856e-06,
+      "loss": 0.5169,
+      "step": 9554
+    },
+    {
+      "epoch": 0.663449520899875,
+      "grad_norm": 3.0672731121113967,
+      "learning_rate": 2.6879189603119104e-06,
+      "loss": 0.3544,
+      "step": 9555
+    },
+    {
+      "epoch": 0.6635189557005972,
+      "grad_norm": 3.639212225449984,
+      "learning_rate": 2.686921977513404e-06,
+      "loss": 0.3049,
+      "step": 9556
+    },
+    {
+      "epoch": 0.6635883905013192,
+      "grad_norm": 3.2271750112385975,
+      "learning_rate": 2.6859251117077694e-06,
+      "loss": 0.345,
+      "step": 9557
+    },
+    {
+      "epoch": 0.6636578253020414,
+      "grad_norm": 3.2387168975810687,
+      "learning_rate": 2.684928362945426e-06,
+      "loss": 0.389,
+      "step": 9558
+    },
+    {
+      "epoch": 0.6637272601027635,
+      "grad_norm": 3.709161272280732,
+      "learning_rate": 2.6839317312767823e-06,
+      "loss": 0.4305,
+      "step": 9559
+    },
+    {
+      "epoch": 0.6637966949034856,
+      "grad_norm": 5.069231118560854,
+      "learning_rate": 2.6829352167522525e-06,
+      "loss": 0.2784,
+      "step": 9560
+    },
+    {
+      "epoch": 0.6638661297042078,
+      "grad_norm": 4.43590595847331,
+      "learning_rate": 2.6819388194222396e-06,
+      "loss": 0.4425,
+      "step": 9561
+    },
+    {
+      "epoch": 0.6639355645049299,
+      "grad_norm": 3.832172992143812,
+      "learning_rate": 2.6809425393371342e-06,
+      "loss": 0.3517,
+      "step": 9562
+    },
+    {
+      "epoch": 0.664004999305652,
+      "grad_norm": 5.6195934213869645,
+      "learning_rate": 2.6799463765473333e-06,
+      "loss": 0.6691,
+      "step": 9563
+    },
+    {
+      "epoch": 0.6640744341063741,
+      "grad_norm": 3.12562356500426,
+      "learning_rate": 2.678950331103218e-06,
+      "loss": 0.4752,
+      "step": 9564
+    },
+    {
+      "epoch": 0.6641438689070962,
+      "grad_norm": 3.3942817384914927,
+      "learning_rate": 2.67795440305517e-06,
+      "loss": 0.2632,
+      "step": 9565
+    },
+    {
+      "epoch": 0.6642133037078184,
+      "grad_norm": 3.8551789043050384,
+      "learning_rate": 2.67695859245356e-06,
+      "loss": 0.4326,
+      "step": 9566
+    },
+    {
+      "epoch": 0.6642827385085405,
+      "grad_norm": 3.7555617970050545,
+      "learning_rate": 2.675962899348756e-06,
+      "loss": 0.3402,
+      "step": 9567
+    },
+    {
+      "epoch": 0.6643521733092626,
+      "grad_norm": 4.211523738861425,
+      "learning_rate": 2.674967323791119e-06,
+      "loss": 0.3856,
+      "step": 9568
+    },
+    {
+      "epoch": 0.6644216081099847,
+      "grad_norm": 3.4331081408378763,
+      "learning_rate": 2.6739718658310033e-06,
+      "loss": 0.2676,
+      "step": 9569
+    },
+    {
+      "epoch": 0.6644910429107068,
+      "grad_norm": 4.285693483471447,
+      "learning_rate": 2.672976525518759e-06,
+      "loss": 0.5442,
+      "step": 9570
+    },
+    {
+      "epoch": 0.664560477711429,
+      "grad_norm": 3.888445124132855,
+      "learning_rate": 2.671981302904727e-06,
+      "loss": 0.5365,
+      "step": 9571
+    },
+    {
+      "epoch": 0.6646299125121511,
+      "grad_norm": 3.8224308428310985,
+      "learning_rate": 2.6709861980392494e-06,
+      "loss": 0.6311,
+      "step": 9572
+    },
+    {
+      "epoch": 0.6646993473128732,
+      "grad_norm": 4.610168520527076,
+      "learning_rate": 2.6699912109726534e-06,
+      "loss": 0.6537,
+      "step": 9573
+    },
+    {
+      "epoch": 0.6647687821135954,
+      "grad_norm": 4.1172500613350715,
+      "learning_rate": 2.668996341755263e-06,
+      "loss": 0.2709,
+      "step": 9574
+    },
+    {
+      "epoch": 0.6648382169143174,
+      "grad_norm": 4.006935433093141,
+      "learning_rate": 2.6680015904374033e-06,
+      "loss": 0.3276,
+      "step": 9575
+    },
+    {
+      "epoch": 0.6649076517150396,
+      "grad_norm": 4.459840572863572,
+      "learning_rate": 2.667006957069381e-06,
+      "loss": 0.5642,
+      "step": 9576
+    },
+    {
+      "epoch": 0.6649770865157617,
+      "grad_norm": 3.5446073566824445,
+      "learning_rate": 2.6660124417015086e-06,
+      "loss": 0.3719,
+      "step": 9577
+    },
+    {
+      "epoch": 0.6650465213164838,
+      "grad_norm": 4.509106318707326,
+      "learning_rate": 2.665018044384087e-06,
+      "loss": 0.3725,
+      "step": 9578
+    },
+    {
+      "epoch": 0.665115956117206,
+      "grad_norm": 4.013338723972772,
+      "learning_rate": 2.664023765167409e-06,
+      "loss": 0.4884,
+      "step": 9579
+    },
+    {
+      "epoch": 0.665185390917928,
+      "grad_norm": 2.4969553263938202,
+      "learning_rate": 2.6630296041017655e-06,
+      "loss": 0.1839,
+      "step": 9580
+    },
+    {
+      "epoch": 0.6652548257186501,
+      "grad_norm": 4.54527118544182,
+      "learning_rate": 2.6620355612374414e-06,
+      "loss": 0.4285,
+      "step": 9581
+    },
+    {
+      "epoch": 0.6653242605193723,
+      "grad_norm": 2.6246937454178476,
+      "learning_rate": 2.661041636624712e-06,
+      "loss": 0.3278,
+      "step": 9582
+    },
+    {
+      "epoch": 0.6653936953200944,
+      "grad_norm": 4.13639168689652,
+      "learning_rate": 2.6600478303138503e-06,
+      "loss": 0.4605,
+      "step": 9583
+    },
+    {
+      "epoch": 0.6654631301208166,
+      "grad_norm": 6.337254886899172,
+      "learning_rate": 2.659054142355122e-06,
+      "loss": 0.8315,
+      "step": 9584
+    },
+    {
+      "epoch": 0.6655325649215387,
+      "grad_norm": 3.5660072477611666,
+      "learning_rate": 2.6580605727987835e-06,
+      "loss": 0.3029,
+      "step": 9585
+    },
+    {
+      "epoch": 0.6656019997222608,
+      "grad_norm": 3.4681466279547006,
+      "learning_rate": 2.657067121695095e-06,
+      "loss": 0.5109,
+      "step": 9586
+    },
+    {
+      "epoch": 0.6656714345229829,
+      "grad_norm": 2.6789334898486,
+      "learning_rate": 2.656073789094299e-06,
+      "loss": 0.2461,
+      "step": 9587
+    },
+    {
+      "epoch": 0.665740869323705,
+      "grad_norm": 3.164142642041622,
+      "learning_rate": 2.6550805750466357e-06,
+      "loss": 0.4248,
+      "step": 9588
+    },
+    {
+      "epoch": 0.6658103041244272,
+      "grad_norm": 4.881216234072196,
+      "learning_rate": 2.6540874796023464e-06,
+      "loss": 0.3402,
+      "step": 9589
+    },
+    {
+      "epoch": 0.6658797389251493,
+      "grad_norm": 3.701304767925541,
+      "learning_rate": 2.653094502811659e-06,
+      "loss": 0.3135,
+      "step": 9590
+    },
+    {
+      "epoch": 0.6659491737258714,
+      "grad_norm": 3.773606619990969,
+      "learning_rate": 2.652101644724792e-06,
+      "loss": 0.5722,
+      "step": 9591
+    },
+    {
+      "epoch": 0.6660186085265936,
+      "grad_norm": 4.553747967699819,
+      "learning_rate": 2.651108905391969e-06,
+      "loss": 0.4253,
+      "step": 9592
+    },
+    {
+      "epoch": 0.6660880433273156,
+      "grad_norm": 2.8010505599686746,
+      "learning_rate": 2.6501162848634023e-06,
+      "loss": 0.2124,
+      "step": 9593
+    },
+    {
+      "epoch": 0.6661574781280378,
+      "grad_norm": 3.617273965230249,
+      "learning_rate": 2.64912378318929e-06,
+      "loss": 0.2613,
+      "step": 9594
+    },
+    {
+      "epoch": 0.6662269129287599,
+      "grad_norm": 3.7457612366613136,
+      "learning_rate": 2.6481314004198387e-06,
+      "loss": 0.3692,
+      "step": 9595
+    },
+    {
+      "epoch": 0.666296347729482,
+      "grad_norm": 4.360504629816085,
+      "learning_rate": 2.6471391366052403e-06,
+      "loss": 0.6264,
+      "step": 9596
+    },
+    {
+      "epoch": 0.6663657825302042,
+      "grad_norm": 2.96776975161655,
+      "learning_rate": 2.646146991795682e-06,
+      "loss": 0.2847,
+      "step": 9597
+    },
+    {
+      "epoch": 0.6664352173309263,
+      "grad_norm": 4.132995594336135,
+      "learning_rate": 2.6451549660413445e-06,
+      "loss": 0.6963,
+      "step": 9598
+    },
+    {
+      "epoch": 0.6665046521316483,
+      "grad_norm": 3.5885931131855786,
+      "learning_rate": 2.6441630593924045e-06,
+      "loss": 0.3527,
+      "step": 9599
+    },
+    {
+      "epoch": 0.6665740869323705,
+      "grad_norm": 4.471230673697358,
+      "learning_rate": 2.643171271899031e-06,
+      "loss": 0.3338,
+      "step": 9600
+    },
+    {
+      "epoch": 0.6666435217330926,
+      "grad_norm": 3.0055029956156787,
+      "learning_rate": 2.6421796036113876e-06,
+      "loss": 0.3438,
+      "step": 9601
+    },
+    {
+      "epoch": 0.6667129565338148,
+      "grad_norm": 3.0158895005580506,
+      "learning_rate": 2.641188054579632e-06,
+      "loss": 0.3821,
+      "step": 9602
+    },
+    {
+      "epoch": 0.6667823913345369,
+      "grad_norm": 3.347053470068318,
+      "learning_rate": 2.6401966248539136e-06,
+      "loss": 0.4291,
+      "step": 9603
+    },
+    {
+      "epoch": 0.666851826135259,
+      "grad_norm": 4.23308599150836,
+      "learning_rate": 2.639205314484383e-06,
+      "loss": 0.4721,
+      "step": 9604
+    },
+    {
+      "epoch": 0.6669212609359811,
+      "grad_norm": 4.118729811873483,
+      "learning_rate": 2.6382141235211735e-06,
+      "loss": 0.4391,
+      "step": 9605
+    },
+    {
+      "epoch": 0.6669906957367032,
+      "grad_norm": 4.021447662441426,
+      "learning_rate": 2.637223052014422e-06,
+      "loss": 0.5309,
+      "step": 9606
+    },
+    {
+      "epoch": 0.6670601305374254,
+      "grad_norm": 4.211485549403837,
+      "learning_rate": 2.6362321000142554e-06,
+      "loss": 0.6249,
+      "step": 9607
+    },
+    {
+      "epoch": 0.6671295653381475,
+      "grad_norm": 2.848214270460804,
+      "learning_rate": 2.635241267570794e-06,
+      "loss": 0.2839,
+      "step": 9608
+    },
+    {
+      "epoch": 0.6671990001388696,
+      "grad_norm": 2.1904887106568744,
+      "learning_rate": 2.6342505547341534e-06,
+      "loss": 0.1716,
+      "step": 9609
+    },
+    {
+      "epoch": 0.6672684349395918,
+      "grad_norm": 4.321615045387327,
+      "learning_rate": 2.633259961554443e-06,
+      "loss": 0.4088,
+      "step": 9610
+    },
+    {
+      "epoch": 0.6673378697403138,
+      "grad_norm": 4.639402687449674,
+      "learning_rate": 2.632269488081765e-06,
+      "loss": 0.4933,
+      "step": 9611
+    },
+    {
+      "epoch": 0.6674073045410359,
+      "grad_norm": 4.074304870513707,
+      "learning_rate": 2.631279134366217e-06,
+      "loss": 0.4103,
+      "step": 9612
+    },
+    {
+      "epoch": 0.6674767393417581,
+      "grad_norm": 3.1927742734862345,
+      "learning_rate": 2.6302889004578908e-06,
+      "loss": 0.3551,
+      "step": 9613
+    },
+    {
+      "epoch": 0.6675461741424802,
+      "grad_norm": 3.858039895475399,
+      "learning_rate": 2.6292987864068697e-06,
+      "loss": 0.4343,
+      "step": 9614
+    },
+    {
+      "epoch": 0.6676156089432024,
+      "grad_norm": 3.15800783113702,
+      "learning_rate": 2.6283087922632346e-06,
+      "loss": 0.3691,
+      "step": 9615
+    },
+    {
+      "epoch": 0.6676850437439245,
+      "grad_norm": 3.500584036853311,
+      "learning_rate": 2.6273189180770555e-06,
+      "loss": 0.3729,
+      "step": 9616
+    },
+    {
+      "epoch": 0.6677544785446465,
+      "grad_norm": 2.933519674551038,
+      "learning_rate": 2.6263291638983997e-06,
+      "loss": 0.2655,
+      "step": 9617
+    },
+    {
+      "epoch": 0.6678239133453687,
+      "grad_norm": 3.9866604323647725,
+      "learning_rate": 2.625339529777332e-06,
+      "loss": 0.4231,
+      "step": 9618
+    },
+    {
+      "epoch": 0.6678933481460908,
+      "grad_norm": 3.2193799869386175,
+      "learning_rate": 2.6243500157639025e-06,
+      "loss": 0.435,
+      "step": 9619
+    },
+    {
+      "epoch": 0.667962782946813,
+      "grad_norm": 4.07473587253958,
+      "learning_rate": 2.6233606219081586e-06,
+      "loss": 0.4929,
+      "step": 9620
+    },
+    {
+      "epoch": 0.6680322177475351,
+      "grad_norm": 3.1030786731354305,
+      "learning_rate": 2.6223713482601474e-06,
+      "loss": 0.2653,
+      "step": 9621
+    },
+    {
+      "epoch": 0.6681016525482572,
+      "grad_norm": 2.445388410051266,
+      "learning_rate": 2.621382194869905e-06,
+      "loss": 0.2165,
+      "step": 9622
+    },
+    {
+      "epoch": 0.6681710873489793,
+      "grad_norm": 3.8817264424970896,
+      "learning_rate": 2.620393161787455e-06,
+      "loss": 0.3792,
+      "step": 9623
+    },
+    {
+      "epoch": 0.6682405221497014,
+      "grad_norm": 4.348645404799426,
+      "learning_rate": 2.6194042490628286e-06,
+      "loss": 0.6629,
+      "step": 9624
+    },
+    {
+      "epoch": 0.6683099569504235,
+      "grad_norm": 3.5489593555584262,
+      "learning_rate": 2.618415456746043e-06,
+      "loss": 0.2525,
+      "step": 9625
+    },
+    {
+      "epoch": 0.6683793917511457,
+      "grad_norm": 4.8643024420139,
+      "learning_rate": 2.6174267848871047e-06,
+      "loss": 0.7433,
+      "step": 9626
+    },
+    {
+      "epoch": 0.6684488265518678,
+      "grad_norm": 5.788648089949168,
+      "learning_rate": 2.6164382335360263e-06,
+      "loss": 0.7028,
+      "step": 9627
+    },
+    {
+      "epoch": 0.66851826135259,
+      "grad_norm": 2.2121620930031134,
+      "learning_rate": 2.6154498027428043e-06,
+      "loss": 0.0881,
+      "step": 9628
+    },
+    {
+      "epoch": 0.668587696153312,
+      "grad_norm": 3.2832432535295633,
+      "learning_rate": 2.614461492557433e-06,
+      "loss": 0.2979,
+      "step": 9629
+    },
+    {
+      "epoch": 0.6686571309540341,
+      "grad_norm": 3.264318757681284,
+      "learning_rate": 2.6134733030299e-06,
+      "loss": 0.4192,
+      "step": 9630
+    },
+    {
+      "epoch": 0.6687265657547563,
+      "grad_norm": 3.3952948696430987,
+      "learning_rate": 2.612485234210187e-06,
+      "loss": 0.2337,
+      "step": 9631
+    },
+    {
+      "epoch": 0.6687960005554784,
+      "grad_norm": 2.7404422565995885,
+      "learning_rate": 2.611497286148267e-06,
+      "loss": 0.2651,
+      "step": 9632
+    },
+    {
+      "epoch": 0.6688654353562006,
+      "grad_norm": 3.7309460232114215,
+      "learning_rate": 2.610509458894116e-06,
+      "loss": 0.3854,
+      "step": 9633
+    },
+    {
+      "epoch": 0.6689348701569227,
+      "grad_norm": 4.146778787354469,
+      "learning_rate": 2.60952175249769e-06,
+      "loss": 0.4416,
+      "step": 9634
+    },
+    {
+      "epoch": 0.6690043049576447,
+      "grad_norm": 4.227578610180248,
+      "learning_rate": 2.608534167008947e-06,
+      "loss": 0.4939,
+      "step": 9635
+    },
+    {
+      "epoch": 0.6690737397583669,
+      "grad_norm": 3.6741687472670566,
+      "learning_rate": 2.6075467024778433e-06,
+      "loss": 0.432,
+      "step": 9636
+    },
+    {
+      "epoch": 0.669143174559089,
+      "grad_norm": 3.698801323170315,
+      "learning_rate": 2.606559358954317e-06,
+      "loss": 0.345,
+      "step": 9637
+    },
+    {
+      "epoch": 0.6692126093598111,
+      "grad_norm": 4.346247046851423,
+      "learning_rate": 2.6055721364883113e-06,
+      "loss": 0.4974,
+      "step": 9638
+    },
+    {
+      "epoch": 0.6692820441605333,
+      "grad_norm": 4.046134477381292,
+      "learning_rate": 2.604585035129758e-06,
+      "loss": 0.4396,
+      "step": 9639
+    },
+    {
+      "epoch": 0.6693514789612554,
+      "grad_norm": 3.75597114853743,
+      "learning_rate": 2.603598054928582e-06,
+      "loss": 0.4566,
+      "step": 9640
+    },
+    {
+      "epoch": 0.6694209137619775,
+      "grad_norm": 3.5317690429154056,
+      "learning_rate": 2.6026111959347054e-06,
+      "loss": 0.2216,
+      "step": 9641
+    },
+    {
+      "epoch": 0.6694903485626996,
+      "grad_norm": 4.1842138775506,
+      "learning_rate": 2.601624458198042e-06,
+      "loss": 0.477,
+      "step": 9642
+    },
+    {
+      "epoch": 0.6695597833634217,
+      "grad_norm": 3.882019313490578,
+      "learning_rate": 2.6006378417684995e-06,
+      "loss": 0.5063,
+      "step": 9643
+    },
+    {
+      "epoch": 0.6696292181641439,
+      "grad_norm": 2.213100444316833,
+      "learning_rate": 2.599651346695979e-06,
+      "loss": 0.1193,
+      "step": 9644
+    },
+    {
+      "epoch": 0.669698652964866,
+      "grad_norm": 4.156878932701719,
+      "learning_rate": 2.598664973030378e-06,
+      "loss": 0.5239,
+      "step": 9645
+    },
+    {
+      "epoch": 0.6697680877655882,
+      "grad_norm": 4.801751722639197,
+      "learning_rate": 2.597678720821584e-06,
+      "loss": 0.3587,
+      "step": 9646
+    },
+    {
+      "epoch": 0.6698375225663102,
+      "grad_norm": 3.730166942670027,
+      "learning_rate": 2.5966925901194847e-06,
+      "loss": 0.3098,
+      "step": 9647
+    },
+    {
+      "epoch": 0.6699069573670323,
+      "grad_norm": 4.364207282402331,
+      "learning_rate": 2.595706580973953e-06,
+      "loss": 0.5057,
+      "step": 9648
+    },
+    {
+      "epoch": 0.6699763921677545,
+      "grad_norm": 3.1705210217551056,
+      "learning_rate": 2.5947206934348603e-06,
+      "loss": 0.2881,
+      "step": 9649
+    },
+    {
+      "epoch": 0.6700458269684766,
+      "grad_norm": 3.037068116227046,
+      "learning_rate": 2.593734927552076e-06,
+      "loss": 0.325,
+      "step": 9650
+    },
+    {
+      "epoch": 0.6701152617691988,
+      "grad_norm": 2.2121210042965895,
+      "learning_rate": 2.5927492833754553e-06,
+      "loss": 0.2793,
+      "step": 9651
+    },
+    {
+      "epoch": 0.6701846965699209,
+      "grad_norm": 7.347952234165531,
+      "learning_rate": 2.591763760954849e-06,
+      "loss": 0.4855,
+      "step": 9652
+    },
+    {
+      "epoch": 0.6702541313706429,
+      "grad_norm": 2.7150719836913293,
+      "learning_rate": 2.5907783603401092e-06,
+      "loss": 0.2143,
+      "step": 9653
+    },
+    {
+      "epoch": 0.6703235661713651,
+      "grad_norm": 4.019797804392938,
+      "learning_rate": 2.5897930815810756e-06,
+      "loss": 0.5557,
+      "step": 9654
+    },
+    {
+      "epoch": 0.6703930009720872,
+      "grad_norm": 3.6512079219894495,
+      "learning_rate": 2.5888079247275768e-06,
+      "loss": 0.4634,
+      "step": 9655
+    },
+    {
+      "epoch": 0.6704624357728093,
+      "grad_norm": 3.0866801980047325,
+      "learning_rate": 2.587822889829446e-06,
+      "loss": 0.3596,
+      "step": 9656
+    },
+    {
+      "epoch": 0.6705318705735315,
+      "grad_norm": 4.106324050363236,
+      "learning_rate": 2.5868379769365043e-06,
+      "loss": 0.4068,
+      "step": 9657
+    },
+    {
+      "epoch": 0.6706013053742536,
+      "grad_norm": 3.697600217139591,
+      "learning_rate": 2.585853186098567e-06,
+      "loss": 0.292,
+      "step": 9658
+    },
+    {
+      "epoch": 0.6706707401749757,
+      "grad_norm": 3.0078087472050035,
+      "learning_rate": 2.5848685173654444e-06,
+      "loss": 0.3413,
+      "step": 9659
+    },
+    {
+      "epoch": 0.6707401749756978,
+      "grad_norm": 4.131844811756808,
+      "learning_rate": 2.5838839707869377e-06,
+      "loss": 0.491,
+      "step": 9660
+    },
+    {
+      "epoch": 0.6708096097764199,
+      "grad_norm": 3.421265875111743,
+      "learning_rate": 2.582899546412847e-06,
+      "loss": 0.2461,
+      "step": 9661
+    },
+    {
+      "epoch": 0.6708790445771421,
+      "grad_norm": 3.743779563111239,
+      "learning_rate": 2.5819152442929616e-06,
+      "loss": 0.5023,
+      "step": 9662
+    },
+    {
+      "epoch": 0.6709484793778642,
+      "grad_norm": 3.309660503402367,
+      "learning_rate": 2.580931064477068e-06,
+      "loss": 0.4283,
+      "step": 9663
+    },
+    {
+      "epoch": 0.6710179141785864,
+      "grad_norm": 4.458567942553772,
+      "learning_rate": 2.579947007014941e-06,
+      "loss": 0.4283,
+      "step": 9664
+    },
+    {
+      "epoch": 0.6710873489793084,
+      "grad_norm": 3.0541000429995977,
+      "learning_rate": 2.5789630719563597e-06,
+      "loss": 0.3452,
+      "step": 9665
+    },
+    {
+      "epoch": 0.6711567837800305,
+      "grad_norm": 3.3621359357103686,
+      "learning_rate": 2.577979259351082e-06,
+      "loss": 0.2917,
+      "step": 9666
+    },
+    {
+      "epoch": 0.6712262185807527,
+      "grad_norm": 4.879415383627998,
+      "learning_rate": 2.5769955692488756e-06,
+      "loss": 0.3457,
+      "step": 9667
+    },
+    {
+      "epoch": 0.6712956533814748,
+      "grad_norm": 3.010817151475008,
+      "learning_rate": 2.576012001699494e-06,
+      "loss": 0.256,
+      "step": 9668
+    },
+    {
+      "epoch": 0.6713650881821969,
+      "grad_norm": 3.851151968221527,
+      "learning_rate": 2.575028556752678e-06,
+      "loss": 0.4603,
+      "step": 9669
+    },
+    {
+      "epoch": 0.6714345229829191,
+      "grad_norm": 3.096567400433019,
+      "learning_rate": 2.5740452344581748e-06,
+      "loss": 0.3588,
+      "step": 9670
+    },
+    {
+      "epoch": 0.6715039577836411,
+      "grad_norm": 6.727280938943735,
+      "learning_rate": 2.5730620348657186e-06,
+      "loss": 0.422,
+      "step": 9671
+    },
+    {
+      "epoch": 0.6715733925843633,
+      "grad_norm": 3.9113635387073034,
+      "learning_rate": 2.5720789580250396e-06,
+      "loss": 0.4502,
+      "step": 9672
+    },
+    {
+      "epoch": 0.6716428273850854,
+      "grad_norm": 3.5090095411586364,
+      "learning_rate": 2.5710960039858585e-06,
+      "loss": 0.5032,
+      "step": 9673
+    },
+    {
+      "epoch": 0.6717122621858075,
+      "grad_norm": 3.4556708167806307,
+      "learning_rate": 2.5701131727978933e-06,
+      "loss": 0.3501,
+      "step": 9674
+    },
+    {
+      "epoch": 0.6717816969865297,
+      "grad_norm": 4.582318093833808,
+      "learning_rate": 2.569130464510854e-06,
+      "loss": 0.6041,
+      "step": 9675
+    },
+    {
+      "epoch": 0.6718511317872518,
+      "grad_norm": 3.163866374720068,
+      "learning_rate": 2.5681478791744453e-06,
+      "loss": 0.3645,
+      "step": 9676
+    },
+    {
+      "epoch": 0.6719205665879739,
+      "grad_norm": 4.16681102675984,
+      "learning_rate": 2.5671654168383653e-06,
+      "loss": 0.4192,
+      "step": 9677
+    },
+    {
+      "epoch": 0.671990001388696,
+      "grad_norm": 4.017742423323768,
+      "learning_rate": 2.566183077552303e-06,
+      "loss": 0.3997,
+      "step": 9678
+    },
+    {
+      "epoch": 0.6720594361894181,
+      "grad_norm": 4.039512586415069,
+      "learning_rate": 2.5652008613659505e-06,
+      "loss": 0.3319,
+      "step": 9679
+    },
+    {
+      "epoch": 0.6721288709901403,
+      "grad_norm": 3.7546754633128443,
+      "learning_rate": 2.564218768328982e-06,
+      "loss": 0.4189,
+      "step": 9680
+    },
+    {
+      "epoch": 0.6721983057908624,
+      "grad_norm": 3.6258382057881313,
+      "learning_rate": 2.56323679849107e-06,
+      "loss": 0.3748,
+      "step": 9681
+    },
+    {
+      "epoch": 0.6722677405915845,
+      "grad_norm": 3.248460218715005,
+      "learning_rate": 2.562254951901885e-06,
+      "loss": 0.4327,
+      "step": 9682
+    },
+    {
+      "epoch": 0.6723371753923066,
+      "grad_norm": 4.6279859414738365,
+      "learning_rate": 2.5612732286110875e-06,
+      "loss": 0.4936,
+      "step": 9683
+    },
+    {
+      "epoch": 0.6724066101930287,
+      "grad_norm": 4.1435774187448535,
+      "learning_rate": 2.560291628668327e-06,
+      "loss": 0.4733,
+      "step": 9684
+    },
+    {
+      "epoch": 0.6724760449937509,
+      "grad_norm": 4.262201692772094,
+      "learning_rate": 2.559310152123257e-06,
+      "loss": 0.5389,
+      "step": 9685
+    },
+    {
+      "epoch": 0.672545479794473,
+      "grad_norm": 4.465493299089475,
+      "learning_rate": 2.558328799025518e-06,
+      "loss": 0.371,
+      "step": 9686
+    },
+    {
+      "epoch": 0.6726149145951951,
+      "grad_norm": 3.9016478429978343,
+      "learning_rate": 2.5573475694247452e-06,
+      "loss": 0.3965,
+      "step": 9687
+    },
+    {
+      "epoch": 0.6726843493959173,
+      "grad_norm": 3.580978263940241,
+      "learning_rate": 2.5563664633705687e-06,
+      "loss": 0.4331,
+      "step": 9688
+    },
+    {
+      "epoch": 0.6727537841966393,
+      "grad_norm": 3.639827556388489,
+      "learning_rate": 2.555385480912611e-06,
+      "loss": 0.5104,
+      "step": 9689
+    },
+    {
+      "epoch": 0.6728232189973615,
+      "grad_norm": 5.487090621809878,
+      "learning_rate": 2.554404622100489e-06,
+      "loss": 0.6626,
+      "step": 9690
+    },
+    {
+      "epoch": 0.6728926537980836,
+      "grad_norm": 4.166366805571317,
+      "learning_rate": 2.553423886983815e-06,
+      "loss": 0.3542,
+      "step": 9691
+    },
+    {
+      "epoch": 0.6729620885988057,
+      "grad_norm": 4.739370512927739,
+      "learning_rate": 2.552443275612191e-06,
+      "loss": 0.6167,
+      "step": 9692
+    },
+    {
+      "epoch": 0.6730315233995279,
+      "grad_norm": 3.9958821012261847,
+      "learning_rate": 2.5514627880352173e-06,
+      "loss": 0.3444,
+      "step": 9693
+    },
+    {
+      "epoch": 0.67310095820025,
+      "grad_norm": 4.169661328641242,
+      "learning_rate": 2.550482424302484e-06,
+      "loss": 0.2941,
+      "step": 9694
+    },
+    {
+      "epoch": 0.673170393000972,
+      "grad_norm": 4.284982421587112,
+      "learning_rate": 2.5495021844635792e-06,
+      "loss": 0.6146,
+      "step": 9695
+    },
+    {
+      "epoch": 0.6732398278016942,
+      "grad_norm": 3.816213151288866,
+      "learning_rate": 2.5485220685680786e-06,
+      "loss": 0.474,
+      "step": 9696
+    },
+    {
+      "epoch": 0.6733092626024163,
+      "grad_norm": 3.664107032515996,
+      "learning_rate": 2.547542076665561e-06,
+      "loss": 0.5443,
+      "step": 9697
+    },
+    {
+      "epoch": 0.6733786974031385,
+      "grad_norm": 3.7187744550337976,
+      "learning_rate": 2.546562208805587e-06,
+      "loss": 0.3318,
+      "step": 9698
+    },
+    {
+      "epoch": 0.6734481322038606,
+      "grad_norm": 4.812640658707071,
+      "learning_rate": 2.5455824650377214e-06,
+      "loss": 0.5023,
+      "step": 9699
+    },
+    {
+      "epoch": 0.6735175670045827,
+      "grad_norm": 4.244860495278399,
+      "learning_rate": 2.54460284541152e-06,
+      "loss": 0.6481,
+      "step": 9700
+    },
+    {
+      "epoch": 0.6735870018053048,
+      "grad_norm": 3.5020361806040787,
+      "learning_rate": 2.543623349976524e-06,
+      "loss": 0.3104,
+      "step": 9701
+    },
+    {
+      "epoch": 0.6736564366060269,
+      "grad_norm": 3.8387648081446444,
+      "learning_rate": 2.5426439787822816e-06,
+      "loss": 0.4668,
+      "step": 9702
+    },
+    {
+      "epoch": 0.6737258714067491,
+      "grad_norm": 1.8760198728090534,
+      "learning_rate": 2.541664731878326e-06,
+      "loss": 0.2283,
+      "step": 9703
+    },
+    {
+      "epoch": 0.6737953062074712,
+      "grad_norm": 4.425266195572725,
+      "learning_rate": 2.5406856093141864e-06,
+      "loss": 0.4582,
+      "step": 9704
+    },
+    {
+      "epoch": 0.6738647410081933,
+      "grad_norm": 3.4715635025041403,
+      "learning_rate": 2.539706611139385e-06,
+      "loss": 0.3162,
+      "step": 9705
+    },
+    {
+      "epoch": 0.6739341758089155,
+      "grad_norm": 4.194945771853427,
+      "learning_rate": 2.53872773740344e-06,
+      "loss": 0.295,
+      "step": 9706
+    },
+    {
+      "epoch": 0.6740036106096375,
+      "grad_norm": 5.5276811629922955,
+      "learning_rate": 2.5377489881558593e-06,
+      "loss": 0.5347,
+      "step": 9707
+    },
+    {
+      "epoch": 0.6740730454103596,
+      "grad_norm": 3.241952589415352,
+      "learning_rate": 2.5367703634461515e-06,
+      "loss": 0.3898,
+      "step": 9708
+    },
+    {
+      "epoch": 0.6741424802110818,
+      "grad_norm": 4.124360047425286,
+      "learning_rate": 2.53579186332381e-06,
+      "loss": 0.4553,
+      "step": 9709
+    },
+    {
+      "epoch": 0.6742119150118039,
+      "grad_norm": 4.988125358972835,
+      "learning_rate": 2.5348134878383256e-06,
+      "loss": 0.3358,
+      "step": 9710
+    },
+    {
+      "epoch": 0.6742813498125261,
+      "grad_norm": 7.130663917767488,
+      "learning_rate": 2.5338352370391893e-06,
+      "loss": 0.7001,
+      "step": 9711
+    },
+    {
+      "epoch": 0.6743507846132482,
+      "grad_norm": 2.794576112703203,
+      "learning_rate": 2.5328571109758747e-06,
+      "loss": 0.3411,
+      "step": 9712
+    },
+    {
+      "epoch": 0.6744202194139702,
+      "grad_norm": 4.10559869220069,
+      "learning_rate": 2.531879109697853e-06,
+      "loss": 0.3211,
+      "step": 9713
+    },
+    {
+      "epoch": 0.6744896542146924,
+      "grad_norm": 3.61868125005041,
+      "learning_rate": 2.530901233254595e-06,
+      "loss": 0.3552,
+      "step": 9714
+    },
+    {
+      "epoch": 0.6745590890154145,
+      "grad_norm": 3.9379609389537102,
+      "learning_rate": 2.529923481695561e-06,
+      "loss": 0.3296,
+      "step": 9715
+    },
+    {
+      "epoch": 0.6746285238161367,
+      "grad_norm": 3.136717494851782,
+      "learning_rate": 2.528945855070198e-06,
+      "loss": 0.286,
+      "step": 9716
+    },
+    {
+      "epoch": 0.6746979586168588,
+      "grad_norm": 4.947895257725833,
+      "learning_rate": 2.5279683534279596e-06,
+      "loss": 0.5271,
+      "step": 9717
+    },
+    {
+      "epoch": 0.6747673934175809,
+      "grad_norm": 3.620309611794032,
+      "learning_rate": 2.5269909768182842e-06,
+      "loss": 0.3837,
+      "step": 9718
+    },
+    {
+      "epoch": 0.674836828218303,
+      "grad_norm": 3.3405238655066642,
+      "learning_rate": 2.526013725290607e-06,
+      "loss": 0.3687,
+      "step": 9719
+    },
+    {
+      "epoch": 0.6749062630190251,
+      "grad_norm": 4.240929915472618,
+      "learning_rate": 2.5250365988943566e-06,
+      "loss": 0.5275,
+      "step": 9720
+    },
+    {
+      "epoch": 0.6749756978197473,
+      "grad_norm": 3.2041513441473266,
+      "learning_rate": 2.524059597678954e-06,
+      "loss": 0.3013,
+      "step": 9721
+    },
+    {
+      "epoch": 0.6750451326204694,
+      "grad_norm": 4.019398798219384,
+      "learning_rate": 2.5230827216938148e-06,
+      "loss": 0.5949,
+      "step": 9722
+    },
+    {
+      "epoch": 0.6751145674211915,
+      "grad_norm": 3.7713699646517314,
+      "learning_rate": 2.5221059709883496e-06,
+      "loss": 0.4183,
+      "step": 9723
+    },
+    {
+      "epoch": 0.6751840022219137,
+      "grad_norm": 5.395947669685341,
+      "learning_rate": 2.52112934561196e-06,
+      "loss": 0.7064,
+      "step": 9724
+    },
+    {
+      "epoch": 0.6752534370226357,
+      "grad_norm": 3.6830206742384166,
+      "learning_rate": 2.520152845614043e-06,
+      "loss": 0.5228,
+      "step": 9725
+    },
+    {
+      "epoch": 0.6753228718233578,
+      "grad_norm": 3.076269817256479,
+      "learning_rate": 2.5191764710439892e-06,
+      "loss": 0.323,
+      "step": 9726
+    },
+    {
+      "epoch": 0.67539230662408,
+      "grad_norm": 4.7961961343879596,
+      "learning_rate": 2.5182002219511796e-06,
+      "loss": 0.4945,
+      "step": 9727
+    },
+    {
+      "epoch": 0.6754617414248021,
+      "grad_norm": 3.9008483363731608,
+      "learning_rate": 2.517224098384997e-06,
+      "loss": 0.5505,
+      "step": 9728
+    },
+    {
+      "epoch": 0.6755311762255243,
+      "grad_norm": 4.448066835598669,
+      "learning_rate": 2.516248100394812e-06,
+      "loss": 0.4391,
+      "step": 9729
+    },
+    {
+      "epoch": 0.6756006110262464,
+      "grad_norm": 3.890368288206,
+      "learning_rate": 2.5152722280299836e-06,
+      "loss": 0.3821,
+      "step": 9730
+    },
+    {
+      "epoch": 0.6756700458269684,
+      "grad_norm": 3.538181630039485,
+      "learning_rate": 2.5142964813398763e-06,
+      "loss": 0.4109,
+      "step": 9731
+    },
+    {
+      "epoch": 0.6757394806276906,
+      "grad_norm": 3.670843140725206,
+      "learning_rate": 2.5133208603738414e-06,
+      "loss": 0.4277,
+      "step": 9732
+    },
+    {
+      "epoch": 0.6758089154284127,
+      "grad_norm": 3.6245782401617923,
+      "learning_rate": 2.5123453651812204e-06,
+      "loss": 0.4623,
+      "step": 9733
+    },
+    {
+      "epoch": 0.6758783502291349,
+      "grad_norm": 2.9457726935829682,
+      "learning_rate": 2.5113699958113575e-06,
+      "loss": 0.1603,
+      "step": 9734
+    },
+    {
+      "epoch": 0.675947785029857,
+      "grad_norm": 3.94014017342817,
+      "learning_rate": 2.510394752313584e-06,
+      "loss": 0.6213,
+      "step": 9735
+    },
+    {
+      "epoch": 0.676017219830579,
+      "grad_norm": 3.677023398941641,
+      "learning_rate": 2.509419634737227e-06,
+      "loss": 0.3307,
+      "step": 9736
+    },
+    {
+      "epoch": 0.6760866546313012,
+      "grad_norm": 2.336695177585533,
+      "learning_rate": 2.5084446431316067e-06,
+      "loss": 0.2083,
+      "step": 9737
+    },
+    {
+      "epoch": 0.6761560894320233,
+      "grad_norm": 7.145143989209525,
+      "learning_rate": 2.5074697775460365e-06,
+      "loss": 0.6122,
+      "step": 9738
+    },
+    {
+      "epoch": 0.6762255242327454,
+      "grad_norm": 4.451275925271477,
+      "learning_rate": 2.5064950380298232e-06,
+      "loss": 0.6391,
+      "step": 9739
+    },
+    {
+      "epoch": 0.6762949590334676,
+      "grad_norm": 4.137005283171907,
+      "learning_rate": 2.5055204246322728e-06,
+      "loss": 0.4332,
+      "step": 9740
+    },
+    {
+      "epoch": 0.6763643938341897,
+      "grad_norm": 4.700908113805155,
+      "learning_rate": 2.504545937402675e-06,
+      "loss": 0.6117,
+      "step": 9741
+    },
+    {
+      "epoch": 0.6764338286349119,
+      "grad_norm": 3.904947765494735,
+      "learning_rate": 2.5035715763903173e-06,
+      "loss": 0.3427,
+      "step": 9742
+    },
+    {
+      "epoch": 0.6765032634356339,
+      "grad_norm": 4.580013987464403,
+      "learning_rate": 2.5025973416444883e-06,
+      "loss": 0.5832,
+      "step": 9743
+    },
+    {
+      "epoch": 0.676572698236356,
+      "grad_norm": 3.269092805447098,
+      "learning_rate": 2.5016232332144584e-06,
+      "loss": 0.1876,
+      "step": 9744
+    },
+    {
+      "epoch": 0.6766421330370782,
+      "grad_norm": 3.3620760735665667,
+      "learning_rate": 2.500649251149496e-06,
+      "loss": 0.3505,
+      "step": 9745
+    },
+    {
+      "epoch": 0.6767115678378003,
+      "grad_norm": 3.502455504892584,
+      "learning_rate": 2.4996753954988683e-06,
+      "loss": 0.3747,
+      "step": 9746
+    },
+    {
+      "epoch": 0.6767810026385225,
+      "grad_norm": 2.622000938289736,
+      "learning_rate": 2.4987016663118295e-06,
+      "loss": 0.3391,
+      "step": 9747
+    },
+    {
+      "epoch": 0.6768504374392446,
+      "grad_norm": 5.026809252514797,
+      "learning_rate": 2.4977280636376304e-06,
+      "loss": 0.6436,
+      "step": 9748
+    },
+    {
+      "epoch": 0.6769198722399666,
+      "grad_norm": 4.302572484864318,
+      "learning_rate": 2.4967545875255137e-06,
+      "loss": 0.503,
+      "step": 9749
+    },
+    {
+      "epoch": 0.6769893070406888,
+      "grad_norm": 4.0992726521276674,
+      "learning_rate": 2.4957812380247183e-06,
+      "loss": 0.4639,
+      "step": 9750
+    },
+    {
+      "epoch": 0.6770587418414109,
+      "grad_norm": 3.447739014194639,
+      "learning_rate": 2.494808015184474e-06,
+      "loss": 0.4517,
+      "step": 9751
+    },
+    {
+      "epoch": 0.677128176642133,
+      "grad_norm": 4.457185391753953,
+      "learning_rate": 2.493834919054005e-06,
+      "loss": 0.3666,
+      "step": 9752
+    },
+    {
+      "epoch": 0.6771976114428552,
+      "grad_norm": 3.6152678178821627,
+      "learning_rate": 2.4928619496825296e-06,
+      "loss": 0.2474,
+      "step": 9753
+    },
+    {
+      "epoch": 0.6772670462435773,
+      "grad_norm": 3.749536840507778,
+      "learning_rate": 2.4918891071192603e-06,
+      "loss": 0.3497,
+      "step": 9754
+    },
+    {
+      "epoch": 0.6773364810442994,
+      "grad_norm": 4.070206381880925,
+      "learning_rate": 2.490916391413402e-06,
+      "loss": 0.4138,
+      "step": 9755
+    },
+    {
+      "epoch": 0.6774059158450215,
+      "grad_norm": 4.370123486426939,
+      "learning_rate": 2.489943802614151e-06,
+      "loss": 0.5095,
+      "step": 9756
+    },
+    {
+      "epoch": 0.6774753506457436,
+      "grad_norm": 3.8354352757635852,
+      "learning_rate": 2.488971340770706e-06,
+      "loss": 0.3033,
+      "step": 9757
+    },
+    {
+      "epoch": 0.6775447854464658,
+      "grad_norm": 3.1775994526404134,
+      "learning_rate": 2.4879990059322477e-06,
+      "loss": 0.2755,
+      "step": 9758
+    },
+    {
+      "epoch": 0.6776142202471879,
+      "grad_norm": 4.1702007371367795,
+      "learning_rate": 2.487026798147955e-06,
+      "loss": 0.4457,
+      "step": 9759
+    },
+    {
+      "epoch": 0.6776836550479101,
+      "grad_norm": 3.6304675652533693,
+      "learning_rate": 2.486054717467005e-06,
+      "loss": 0.291,
+      "step": 9760
+    },
+    {
+      "epoch": 0.6777530898486321,
+      "grad_norm": 2.583086833181414,
+      "learning_rate": 2.485082763938565e-06,
+      "loss": 0.1789,
+      "step": 9761
+    },
+    {
+      "epoch": 0.6778225246493542,
+      "grad_norm": 4.089371429553828,
+      "learning_rate": 2.484110937611789e-06,
+      "loss": 0.4765,
+      "step": 9762
+    },
+    {
+      "epoch": 0.6778919594500764,
+      "grad_norm": 3.518959333092312,
+      "learning_rate": 2.4831392385358367e-06,
+      "loss": 0.4935,
+      "step": 9763
+    },
+    {
+      "epoch": 0.6779613942507985,
+      "grad_norm": 3.7920916130237976,
+      "learning_rate": 2.4821676667598533e-06,
+      "loss": 0.272,
+      "step": 9764
+    },
+    {
+      "epoch": 0.6780308290515206,
+      "grad_norm": 5.675743168002094,
+      "learning_rate": 2.4811962223329806e-06,
+      "loss": 0.611,
+      "step": 9765
+    },
+    {
+      "epoch": 0.6781002638522428,
+      "grad_norm": 3.992265908866341,
+      "learning_rate": 2.4802249053043525e-06,
+      "loss": 0.4023,
+      "step": 9766
+    },
+    {
+      "epoch": 0.6781696986529648,
+      "grad_norm": 3.193885330440984,
+      "learning_rate": 2.4792537157230977e-06,
+      "loss": 0.2608,
+      "step": 9767
+    },
+    {
+      "epoch": 0.678239133453687,
+      "grad_norm": 3.473120120301395,
+      "learning_rate": 2.4782826536383377e-06,
+      "loss": 0.3732,
+      "step": 9768
+    },
+    {
+      "epoch": 0.6783085682544091,
+      "grad_norm": 4.249550672481268,
+      "learning_rate": 2.477311719099187e-06,
+      "loss": 0.497,
+      "step": 9769
+    },
+    {
+      "epoch": 0.6783780030551312,
+      "grad_norm": 5.083217894018776,
+      "learning_rate": 2.476340912154755e-06,
+      "loss": 0.7283,
+      "step": 9770
+    },
+    {
+      "epoch": 0.6784474378558534,
+      "grad_norm": 3.8054011520094146,
+      "learning_rate": 2.475370232854142e-06,
+      "loss": 0.4105,
+      "step": 9771
+    },
+    {
+      "epoch": 0.6785168726565755,
+      "grad_norm": 4.3101436049052975,
+      "learning_rate": 2.4743996812464497e-06,
+      "loss": 0.6002,
+      "step": 9772
+    },
+    {
+      "epoch": 0.6785863074572976,
+      "grad_norm": 4.183488453020607,
+      "learning_rate": 2.473429257380762e-06,
+      "loss": 0.5859,
+      "step": 9773
+    },
+    {
+      "epoch": 0.6786557422580197,
+      "grad_norm": 1.8007433083212114,
+      "learning_rate": 2.472458961306162e-06,
+      "loss": 0.0895,
+      "step": 9774
+    },
+    {
+      "epoch": 0.6787251770587418,
+      "grad_norm": 3.7364372105373915,
+      "learning_rate": 2.471488793071731e-06,
+      "loss": 0.3044,
+      "step": 9775
+    },
+    {
+      "epoch": 0.678794611859464,
+      "grad_norm": 3.227121480066868,
+      "learning_rate": 2.470518752726533e-06,
+      "loss": 0.3059,
+      "step": 9776
+    },
+    {
+      "epoch": 0.6788640466601861,
+      "grad_norm": 4.610378546644763,
+      "learning_rate": 2.469548840319633e-06,
+      "loss": 0.4545,
+      "step": 9777
+    },
+    {
+      "epoch": 0.6789334814609083,
+      "grad_norm": 2.6498224193838222,
+      "learning_rate": 2.468579055900091e-06,
+      "loss": 0.2713,
+      "step": 9778
+    },
+    {
+      "epoch": 0.6790029162616303,
+      "grad_norm": 4.08439232512555,
+      "learning_rate": 2.467609399516956e-06,
+      "loss": 0.511,
+      "step": 9779
+    },
+    {
+      "epoch": 0.6790723510623524,
+      "grad_norm": 3.9045889907063382,
+      "learning_rate": 2.466639871219272e-06,
+      "loss": 0.3825,
+      "step": 9780
+    },
+    {
+      "epoch": 0.6791417858630746,
+      "grad_norm": 4.048589629516293,
+      "learning_rate": 2.4656704710560762e-06,
+      "loss": 0.3581,
+      "step": 9781
+    },
+    {
+      "epoch": 0.6792112206637967,
+      "grad_norm": 3.906987869773705,
+      "learning_rate": 2.4647011990764004e-06,
+      "loss": 0.3823,
+      "step": 9782
+    },
+    {
+      "epoch": 0.6792806554645188,
+      "grad_norm": 3.1283142429687123,
+      "learning_rate": 2.4637320553292694e-06,
+      "loss": 0.2515,
+      "step": 9783
+    },
+    {
+      "epoch": 0.679350090265241,
+      "grad_norm": 4.673842667752665,
+      "learning_rate": 2.4627630398637003e-06,
+      "loss": 0.6072,
+      "step": 9784
+    },
+    {
+      "epoch": 0.679419525065963,
+      "grad_norm": 3.422247096702537,
+      "learning_rate": 2.4617941527287065e-06,
+      "loss": 0.3338,
+      "step": 9785
+    },
+    {
+      "epoch": 0.6794889598666852,
+      "grad_norm": 3.5404415532155498,
+      "learning_rate": 2.4608253939732916e-06,
+      "loss": 0.3626,
+      "step": 9786
+    },
+    {
+      "epoch": 0.6795583946674073,
+      "grad_norm": 4.230133652992493,
+      "learning_rate": 2.4598567636464555e-06,
+      "loss": 0.3833,
+      "step": 9787
+    },
+    {
+      "epoch": 0.6796278294681294,
+      "grad_norm": 4.2977484399390535,
+      "learning_rate": 2.4588882617971876e-06,
+      "loss": 0.5147,
+      "step": 9788
+    },
+    {
+      "epoch": 0.6796972642688516,
+      "grad_norm": 4.248234780759611,
+      "learning_rate": 2.457919888474478e-06,
+      "loss": 0.5749,
+      "step": 9789
+    },
+    {
+      "epoch": 0.6797666990695737,
+      "grad_norm": 4.120775296393542,
+      "learning_rate": 2.4569516437273057e-06,
+      "loss": 0.4236,
+      "step": 9790
+    },
+    {
+      "epoch": 0.6798361338702958,
+      "grad_norm": 3.967674741087389,
+      "learning_rate": 2.455983527604638e-06,
+      "loss": 0.5141,
+      "step": 9791
+    },
+    {
+      "epoch": 0.6799055686710179,
+      "grad_norm": 4.814784082828865,
+      "learning_rate": 2.4550155401554465e-06,
+      "loss": 0.4685,
+      "step": 9792
+    },
+    {
+      "epoch": 0.67997500347174,
+      "grad_norm": 3.4887386894046966,
+      "learning_rate": 2.454047681428691e-06,
+      "loss": 0.2946,
+      "step": 9793
+    },
+    {
+      "epoch": 0.6800444382724622,
+      "grad_norm": 3.9472123323822412,
+      "learning_rate": 2.4530799514733196e-06,
+      "loss": 0.4068,
+      "step": 9794
+    },
+    {
+      "epoch": 0.6801138730731843,
+      "grad_norm": 3.3454137791367344,
+      "learning_rate": 2.4521123503382833e-06,
+      "loss": 0.2648,
+      "step": 9795
+    },
+    {
+      "epoch": 0.6801833078739064,
+      "grad_norm": 4.723927272266753,
+      "learning_rate": 2.4511448780725218e-06,
+      "loss": 0.5581,
+      "step": 9796
+    },
+    {
+      "epoch": 0.6802527426746285,
+      "grad_norm": 3.2935673405432273,
+      "learning_rate": 2.4501775347249673e-06,
+      "loss": 0.327,
+      "step": 9797
+    },
+    {
+      "epoch": 0.6803221774753506,
+      "grad_norm": 3.851935796947837,
+      "learning_rate": 2.4492103203445484e-06,
+      "loss": 0.5686,
+      "step": 9798
+    },
+    {
+      "epoch": 0.6803916122760728,
+      "grad_norm": 5.14645595182481,
+      "learning_rate": 2.448243234980185e-06,
+      "loss": 0.6463,
+      "step": 9799
+    },
+    {
+      "epoch": 0.6804610470767949,
+      "grad_norm": 4.671569036324234,
+      "learning_rate": 2.4472762786807913e-06,
+      "loss": 0.6234,
+      "step": 9800
+    },
+    {
+      "epoch": 0.680530481877517,
+      "grad_norm": 3.3705166050616406,
+      "learning_rate": 2.4463094514952753e-06,
+      "loss": 0.3874,
+      "step": 9801
+    },
+    {
+      "epoch": 0.6805999166782392,
+      "grad_norm": 4.366585635441514,
+      "learning_rate": 2.4453427534725365e-06,
+      "loss": 0.5305,
+      "step": 9802
+    },
+    {
+      "epoch": 0.6806693514789612,
+      "grad_norm": 3.512261448837231,
+      "learning_rate": 2.4443761846614695e-06,
+      "loss": 0.2289,
+      "step": 9803
+    },
+    {
+      "epoch": 0.6807387862796834,
+      "grad_norm": 3.1531138672918604,
+      "learning_rate": 2.4434097451109663e-06,
+      "loss": 0.2733,
+      "step": 9804
+    },
+    {
+      "epoch": 0.6808082210804055,
+      "grad_norm": 6.024589483375098,
+      "learning_rate": 2.442443434869904e-06,
+      "loss": 0.5628,
+      "step": 9805
+    },
+    {
+      "epoch": 0.6808776558811276,
+      "grad_norm": 2.6348623262979634,
+      "learning_rate": 2.441477253987156e-06,
+      "loss": 0.2849,
+      "step": 9806
+    },
+    {
+      "epoch": 0.6809470906818498,
+      "grad_norm": 4.6013348558501885,
+      "learning_rate": 2.4405112025115973e-06,
+      "loss": 0.5005,
+      "step": 9807
+    },
+    {
+      "epoch": 0.6810165254825719,
+      "grad_norm": 3.968140317015296,
+      "learning_rate": 2.439545280492082e-06,
+      "loss": 0.563,
+      "step": 9808
+    },
+    {
+      "epoch": 0.6810859602832939,
+      "grad_norm": 4.692451579638296,
+      "learning_rate": 2.4385794879774714e-06,
+      "loss": 0.4515,
+      "step": 9809
+    },
+    {
+      "epoch": 0.6811553950840161,
+      "grad_norm": 3.381600686299743,
+      "learning_rate": 2.4376138250166114e-06,
+      "loss": 0.2589,
+      "step": 9810
+    },
+    {
+      "epoch": 0.6812248298847382,
+      "grad_norm": 3.454506973994623,
+      "learning_rate": 2.436648291658345e-06,
+      "loss": 0.2894,
+      "step": 9811
+    },
+    {
+      "epoch": 0.6812942646854604,
+      "grad_norm": 3.3844751861396825,
+      "learning_rate": 2.435682887951507e-06,
+      "loss": 0.2106,
+      "step": 9812
+    },
+    {
+      "epoch": 0.6813636994861825,
+      "grad_norm": 3.0477782983983737,
+      "learning_rate": 2.434717613944927e-06,
+      "loss": 0.2554,
+      "step": 9813
+    },
+    {
+      "epoch": 0.6814331342869046,
+      "grad_norm": 5.336881185391047,
+      "learning_rate": 2.433752469687427e-06,
+      "loss": 0.6978,
+      "step": 9814
+    },
+    {
+      "epoch": 0.6815025690876267,
+      "grad_norm": 4.726502487413906,
+      "learning_rate": 2.4327874552278236e-06,
+      "loss": 0.6078,
+      "step": 9815
+    },
+    {
+      "epoch": 0.6815720038883488,
+      "grad_norm": 6.004848463000493,
+      "learning_rate": 2.4318225706149257e-06,
+      "loss": 0.525,
+      "step": 9816
+    },
+    {
+      "epoch": 0.681641438689071,
+      "grad_norm": 3.541127098000716,
+      "learning_rate": 2.4308578158975338e-06,
+      "loss": 0.3951,
+      "step": 9817
+    },
+    {
+      "epoch": 0.6817108734897931,
+      "grad_norm": 3.926064963166198,
+      "learning_rate": 2.4298931911244505e-06,
+      "loss": 0.522,
+      "step": 9818
+    },
+    {
+      "epoch": 0.6817803082905152,
+      "grad_norm": 4.162048099655802,
+      "learning_rate": 2.428928696344459e-06,
+      "loss": 0.1981,
+      "step": 9819
+    },
+    {
+      "epoch": 0.6818497430912374,
+      "grad_norm": 4.457082748464746,
+      "learning_rate": 2.4279643316063432e-06,
+      "loss": 0.4376,
+      "step": 9820
+    },
+    {
+      "epoch": 0.6819191778919594,
+      "grad_norm": 4.883138669728883,
+      "learning_rate": 2.427000096958883e-06,
+      "loss": 0.6762,
+      "step": 9821
+    },
+    {
+      "epoch": 0.6819886126926815,
+      "grad_norm": 3.505156444799803,
+      "learning_rate": 2.426035992450848e-06,
+      "loss": 0.3529,
+      "step": 9822
+    },
+    {
+      "epoch": 0.6820580474934037,
+      "grad_norm": 2.2955569415383437,
+      "learning_rate": 2.425072018130996e-06,
+      "loss": 0.1911,
+      "step": 9823
+    },
+    {
+      "epoch": 0.6821274822941258,
+      "grad_norm": 3.9554722787225183,
+      "learning_rate": 2.42410817404809e-06,
+      "loss": 0.3052,
+      "step": 9824
+    },
+    {
+      "epoch": 0.682196917094848,
+      "grad_norm": 4.423032289292352,
+      "learning_rate": 2.423144460250879e-06,
+      "loss": 0.767,
+      "step": 9825
+    },
+    {
+      "epoch": 0.6822663518955701,
+      "grad_norm": 3.919395822501866,
+      "learning_rate": 2.422180876788102e-06,
+      "loss": 0.399,
+      "step": 9826
+    },
+    {
+      "epoch": 0.6823357866962921,
+      "grad_norm": 3.759330949911373,
+      "learning_rate": 2.4212174237085007e-06,
+      "loss": 0.4432,
+      "step": 9827
+    },
+    {
+      "epoch": 0.6824052214970143,
+      "grad_norm": 3.812652143984822,
+      "learning_rate": 2.4202541010608045e-06,
+      "loss": 0.4258,
+      "step": 9828
+    },
+    {
+      "epoch": 0.6824746562977364,
+      "grad_norm": 3.290250345342252,
+      "learning_rate": 2.419290908893736e-06,
+      "loss": 0.4326,
+      "step": 9829
+    },
+    {
+      "epoch": 0.6825440910984586,
+      "grad_norm": 4.388961955153422,
+      "learning_rate": 2.418327847256014e-06,
+      "loss": 0.225,
+      "step": 9830
+    },
+    {
+      "epoch": 0.6826135258991807,
+      "grad_norm": 4.777111873850835,
+      "learning_rate": 2.4173649161963475e-06,
+      "loss": 0.5539,
+      "step": 9831
+    },
+    {
+      "epoch": 0.6826829606999028,
+      "grad_norm": 3.4778579198757895,
+      "learning_rate": 2.416402115763441e-06,
+      "loss": 0.4531,
+      "step": 9832
+    },
+    {
+      "epoch": 0.6827523955006249,
+      "grad_norm": 3.321419067205405,
+      "learning_rate": 2.4154394460059923e-06,
+      "loss": 0.3097,
+      "step": 9833
+    },
+    {
+      "epoch": 0.682821830301347,
+      "grad_norm": 2.6870853837166777,
+      "learning_rate": 2.4144769069726913e-06,
+      "loss": 0.1043,
+      "step": 9834
+    },
+    {
+      "epoch": 0.6828912651020692,
+      "grad_norm": 4.485782733413446,
+      "learning_rate": 2.413514498712221e-06,
+      "loss": 0.5505,
+      "step": 9835
+    },
+    {
+      "epoch": 0.6829606999027913,
+      "grad_norm": 3.7519678654551405,
+      "learning_rate": 2.412552221273265e-06,
+      "loss": 0.4472,
+      "step": 9836
+    },
+    {
+      "epoch": 0.6830301347035134,
+      "grad_norm": 3.0660247520721917,
+      "learning_rate": 2.411590074704485e-06,
+      "loss": 0.2066,
+      "step": 9837
+    },
+    {
+      "epoch": 0.6830995695042356,
+      "grad_norm": 3.883769911046286,
+      "learning_rate": 2.4106280590545524e-06,
+      "loss": 0.4629,
+      "step": 9838
+    },
+    {
+      "epoch": 0.6831690043049576,
+      "grad_norm": 3.17446447432747,
+      "learning_rate": 2.4096661743721222e-06,
+      "loss": 0.4595,
+      "step": 9839
+    },
+    {
+      "epoch": 0.6832384391056797,
+      "grad_norm": 2.9699594447814897,
+      "learning_rate": 2.4087044207058464e-06,
+      "loss": 0.3838,
+      "step": 9840
+    },
+    {
+      "epoch": 0.6833078739064019,
+      "grad_norm": 4.467682145984409,
+      "learning_rate": 2.4077427981043683e-06,
+      "loss": 0.3624,
+      "step": 9841
+    },
+    {
+      "epoch": 0.683377308707124,
+      "grad_norm": 3.871110591826709,
+      "learning_rate": 2.406781306616326e-06,
+      "loss": 0.5333,
+      "step": 9842
+    },
+    {
+      "epoch": 0.6834467435078462,
+      "grad_norm": 4.420531011626496,
+      "learning_rate": 2.4058199462903507e-06,
+      "loss": 0.4566,
+      "step": 9843
+    },
+    {
+      "epoch": 0.6835161783085683,
+      "grad_norm": 4.329720520729106,
+      "learning_rate": 2.4048587171750674e-06,
+      "loss": 0.4845,
+      "step": 9844
+    },
+    {
+      "epoch": 0.6835856131092903,
+      "grad_norm": 2.8571727935793394,
+      "learning_rate": 2.4038976193190933e-06,
+      "loss": 0.179,
+      "step": 9845
+    },
+    {
+      "epoch": 0.6836550479100125,
+      "grad_norm": 5.2119036760788005,
+      "learning_rate": 2.402936652771038e-06,
+      "loss": 0.5882,
+      "step": 9846
+    },
+    {
+      "epoch": 0.6837244827107346,
+      "grad_norm": 4.0971113864352064,
+      "learning_rate": 2.4019758175795123e-06,
+      "loss": 0.4917,
+      "step": 9847
+    },
+    {
+      "epoch": 0.6837939175114568,
+      "grad_norm": 2.7377028159682433,
+      "learning_rate": 2.4010151137931075e-06,
+      "loss": 0.2143,
+      "step": 9848
+    },
+    {
+      "epoch": 0.6838633523121789,
+      "grad_norm": 3.355829205627902,
+      "learning_rate": 2.400054541460415e-06,
+      "loss": 0.3104,
+      "step": 9849
+    },
+    {
+      "epoch": 0.683932787112901,
+      "grad_norm": 4.575762143658527,
+      "learning_rate": 2.3990941006300267e-06,
+      "loss": 0.5211,
+      "step": 9850
+    },
+    {
+      "epoch": 0.6840022219136231,
+      "grad_norm": 3.6274356905656204,
+      "learning_rate": 2.398133791350513e-06,
+      "loss": 0.479,
+      "step": 9851
+    },
+    {
+      "epoch": 0.6840716567143452,
+      "grad_norm": 3.213791766494711,
+      "learning_rate": 2.397173613670446e-06,
+      "loss": 0.373,
+      "step": 9852
+    },
+    {
+      "epoch": 0.6841410915150673,
+      "grad_norm": 5.5625155185211534,
+      "learning_rate": 2.3962135676383946e-06,
+      "loss": 0.5589,
+      "step": 9853
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 3.5479415820895257,
+      "learning_rate": 2.3952536533029164e-06,
+      "loss": 0.3961,
+      "step": 9854
+    },
+    {
+      "epoch": 0.6842799611165116,
+      "grad_norm": 4.304659394842034,
+      "learning_rate": 2.394293870712557e-06,
+      "loss": 0.3954,
+      "step": 9855
+    },
+    {
+      "epoch": 0.6843493959172338,
+      "grad_norm": 2.9469344723330217,
+      "learning_rate": 2.3933342199158672e-06,
+      "loss": 0.2235,
+      "step": 9856
+    },
+    {
+      "epoch": 0.6844188307179558,
+      "grad_norm": 3.72194463101503,
+      "learning_rate": 2.3923747009613823e-06,
+      "loss": 0.388,
+      "step": 9857
+    },
+    {
+      "epoch": 0.6844882655186779,
+      "grad_norm": 4.458577840190922,
+      "learning_rate": 2.3914153138976347e-06,
+      "loss": 0.5241,
+      "step": 9858
+    },
+    {
+      "epoch": 0.6845577003194001,
+      "grad_norm": 4.563625355555192,
+      "learning_rate": 2.390456058773149e-06,
+      "loss": 0.4951,
+      "step": 9859
+    },
+    {
+      "epoch": 0.6846271351201222,
+      "grad_norm": 3.1891317695617456,
+      "learning_rate": 2.389496935636442e-06,
+      "loss": 0.4283,
+      "step": 9860
+    },
+    {
+      "epoch": 0.6846965699208444,
+      "grad_norm": 7.761057046491241,
+      "learning_rate": 2.3885379445360267e-06,
+      "loss": 0.5915,
+      "step": 9861
+    },
+    {
+      "epoch": 0.6847660047215665,
+      "grad_norm": 4.725632435289322,
+      "learning_rate": 2.3875790855204074e-06,
+      "loss": 0.2836,
+      "step": 9862
+    },
+    {
+      "epoch": 0.6848354395222885,
+      "grad_norm": 3.550226777742127,
+      "learning_rate": 2.3866203586380815e-06,
+      "loss": 0.3609,
+      "step": 9863
+    },
+    {
+      "epoch": 0.6849048743230107,
+      "grad_norm": 5.243164559437213,
+      "learning_rate": 2.3856617639375386e-06,
+      "loss": 0.6243,
+      "step": 9864
+    },
+    {
+      "epoch": 0.6849743091237328,
+      "grad_norm": 3.139078133660744,
+      "learning_rate": 2.3847033014672693e-06,
+      "loss": 0.2132,
+      "step": 9865
+    },
+    {
+      "epoch": 0.6850437439244549,
+      "grad_norm": 4.280098854404842,
+      "learning_rate": 2.3837449712757456e-06,
+      "loss": 0.424,
+      "step": 9866
+    },
+    {
+      "epoch": 0.6851131787251771,
+      "grad_norm": 3.9676959371392377,
+      "learning_rate": 2.382786773411439e-06,
+      "loss": 0.551,
+      "step": 9867
+    },
+    {
+      "epoch": 0.6851826135258992,
+      "grad_norm": 3.8895289132689386,
+      "learning_rate": 2.38182870792282e-06,
+      "loss": 0.3518,
+      "step": 9868
+    },
+    {
+      "epoch": 0.6852520483266213,
+      "grad_norm": 3.386574184599235,
+      "learning_rate": 2.3808707748583375e-06,
+      "loss": 0.4454,
+      "step": 9869
+    },
+    {
+      "epoch": 0.6853214831273434,
+      "grad_norm": 3.0772652248404055,
+      "learning_rate": 2.37991297426645e-06,
+      "loss": 0.3227,
+      "step": 9870
+    },
+    {
+      "epoch": 0.6853909179280655,
+      "grad_norm": 3.9581502261761505,
+      "learning_rate": 2.378955306195599e-06,
+      "loss": 0.4379,
+      "step": 9871
+    },
+    {
+      "epoch": 0.6854603527287877,
+      "grad_norm": 3.745209164161974,
+      "learning_rate": 2.3779977706942232e-06,
+      "loss": 0.3932,
+      "step": 9872
+    },
+    {
+      "epoch": 0.6855297875295098,
+      "grad_norm": 7.362292670325728,
+      "learning_rate": 2.3770403678107527e-06,
+      "loss": 0.5089,
+      "step": 9873
+    },
+    {
+      "epoch": 0.685599222330232,
+      "grad_norm": 3.4551543865411674,
+      "learning_rate": 2.376083097593613e-06,
+      "loss": 0.3409,
+      "step": 9874
+    },
+    {
+      "epoch": 0.685668657130954,
+      "grad_norm": 3.6950473301357785,
+      "learning_rate": 2.37512596009122e-06,
+      "loss": 0.4172,
+      "step": 9875
+    },
+    {
+      "epoch": 0.6857380919316761,
+      "grad_norm": 4.683621242369369,
+      "learning_rate": 2.374168955351987e-06,
+      "loss": 0.5417,
+      "step": 9876
+    },
+    {
+      "epoch": 0.6858075267323983,
+      "grad_norm": 2.845316893945616,
+      "learning_rate": 2.373212083424316e-06,
+      "loss": 0.2425,
+      "step": 9877
+    },
+    {
+      "epoch": 0.6858769615331204,
+      "grad_norm": 4.146781290959154,
+      "learning_rate": 2.3722553443566038e-06,
+      "loss": 0.5117,
+      "step": 9878
+    },
+    {
+      "epoch": 0.6859463963338425,
+      "grad_norm": 3.643348728144497,
+      "learning_rate": 2.3712987381972464e-06,
+      "loss": 0.2761,
+      "step": 9879
+    },
+    {
+      "epoch": 0.6860158311345647,
+      "grad_norm": 3.9017239229434586,
+      "learning_rate": 2.370342264994623e-06,
+      "loss": 0.5139,
+      "step": 9880
+    },
+    {
+      "epoch": 0.6860852659352867,
+      "grad_norm": 4.34224936926708,
+      "learning_rate": 2.3693859247971096e-06,
+      "loss": 0.4932,
+      "step": 9881
+    },
+    {
+      "epoch": 0.6861547007360089,
+      "grad_norm": 3.4022923221584795,
+      "learning_rate": 2.368429717653083e-06,
+      "loss": 0.4389,
+      "step": 9882
+    },
+    {
+      "epoch": 0.686224135536731,
+      "grad_norm": 3.7071649879454442,
+      "learning_rate": 2.3674736436109018e-06,
+      "loss": 0.2884,
+      "step": 9883
+    },
+    {
+      "epoch": 0.6862935703374531,
+      "grad_norm": 3.190885950397369,
+      "learning_rate": 2.3665177027189236e-06,
+      "loss": 0.4317,
+      "step": 9884
+    },
+    {
+      "epoch": 0.6863630051381753,
+      "grad_norm": 3.602591788453721,
+      "learning_rate": 2.365561895025501e-06,
+      "loss": 0.4219,
+      "step": 9885
+    },
+    {
+      "epoch": 0.6864324399388974,
+      "grad_norm": 3.790112195645559,
+      "learning_rate": 2.3646062205789787e-06,
+      "loss": 0.2881,
+      "step": 9886
+    },
+    {
+      "epoch": 0.6865018747396195,
+      "grad_norm": 4.0798318215741896,
+      "learning_rate": 2.363650679427688e-06,
+      "loss": 0.5128,
+      "step": 9887
+    },
+    {
+      "epoch": 0.6865713095403416,
+      "grad_norm": 6.378104590589276,
+      "learning_rate": 2.3626952716199647e-06,
+      "loss": 0.8599,
+      "step": 9888
+    },
+    {
+      "epoch": 0.6866407443410637,
+      "grad_norm": 3.7217274911588625,
+      "learning_rate": 2.3617399972041306e-06,
+      "loss": 0.534,
+      "step": 9889
+    },
+    {
+      "epoch": 0.6867101791417859,
+      "grad_norm": 2.66682747115505,
+      "learning_rate": 2.360784856228501e-06,
+      "loss": 0.231,
+      "step": 9890
+    },
+    {
+      "epoch": 0.686779613942508,
+      "grad_norm": 4.052588606123804,
+      "learning_rate": 2.3598298487413877e-06,
+      "loss": 0.4671,
+      "step": 9891
+    },
+    {
+      "epoch": 0.6868490487432302,
+      "grad_norm": 3.634110962963185,
+      "learning_rate": 2.358874974791093e-06,
+      "loss": 0.3105,
+      "step": 9892
+    },
+    {
+      "epoch": 0.6869184835439522,
+      "grad_norm": 5.7255852888577925,
+      "learning_rate": 2.3579202344259134e-06,
+      "loss": 0.4778,
+      "step": 9893
+    },
+    {
+      "epoch": 0.6869879183446743,
+      "grad_norm": 3.674345076350443,
+      "learning_rate": 2.3569656276941384e-06,
+      "loss": 0.3107,
+      "step": 9894
+    },
+    {
+      "epoch": 0.6870573531453965,
+      "grad_norm": 2.869175328462186,
+      "learning_rate": 2.3560111546440512e-06,
+      "loss": 0.1981,
+      "step": 9895
+    },
+    {
+      "epoch": 0.6871267879461186,
+      "grad_norm": 4.084003961052353,
+      "learning_rate": 2.355056815323926e-06,
+      "loss": 0.4533,
+      "step": 9896
+    },
+    {
+      "epoch": 0.6871962227468407,
+      "grad_norm": 3.1662151278136825,
+      "learning_rate": 2.3541026097820385e-06,
+      "loss": 0.2483,
+      "step": 9897
+    },
+    {
+      "epoch": 0.6872656575475629,
+      "grad_norm": 4.362849235929402,
+      "learning_rate": 2.3531485380666425e-06,
+      "loss": 0.2923,
+      "step": 9898
+    },
+    {
+      "epoch": 0.6873350923482849,
+      "grad_norm": 3.6564910269434883,
+      "learning_rate": 2.3521946002260003e-06,
+      "loss": 0.3972,
+      "step": 9899
+    },
+    {
+      "epoch": 0.6874045271490071,
+      "grad_norm": 4.220993109871395,
+      "learning_rate": 2.3512407963083617e-06,
+      "loss": 0.5648,
+      "step": 9900
+    },
+    {
+      "epoch": 0.6874739619497292,
+      "grad_norm": 4.581219059763115,
+      "learning_rate": 2.3502871263619616e-06,
+      "loss": 0.4656,
+      "step": 9901
+    },
+    {
+      "epoch": 0.6875433967504513,
+      "grad_norm": 6.187364063775447,
+      "learning_rate": 2.3493335904350427e-06,
+      "loss": 0.5475,
+      "step": 9902
+    },
+    {
+      "epoch": 0.6876128315511735,
+      "grad_norm": 4.018151480429149,
+      "learning_rate": 2.348380188575831e-06,
+      "loss": 0.5931,
+      "step": 9903
+    },
+    {
+      "epoch": 0.6876822663518956,
+      "grad_norm": 3.5922712318921097,
+      "learning_rate": 2.347426920832549e-06,
+      "loss": 0.3366,
+      "step": 9904
+    },
+    {
+      "epoch": 0.6877517011526177,
+      "grad_norm": 4.5581271579348765,
+      "learning_rate": 2.346473787253412e-06,
+      "loss": 0.4588,
+      "step": 9905
+    },
+    {
+      "epoch": 0.6878211359533398,
+      "grad_norm": 3.8155163476561356,
+      "learning_rate": 2.3455207878866283e-06,
+      "loss": 0.4549,
+      "step": 9906
+    },
+    {
+      "epoch": 0.6878905707540619,
+      "grad_norm": 4.926024292497031,
+      "learning_rate": 2.3445679227803984e-06,
+      "loss": 0.6584,
+      "step": 9907
+    },
+    {
+      "epoch": 0.6879600055547841,
+      "grad_norm": 5.476888498565792,
+      "learning_rate": 2.343615191982919e-06,
+      "loss": 0.391,
+      "step": 9908
+    },
+    {
+      "epoch": 0.6880294403555062,
+      "grad_norm": 4.086756105009742,
+      "learning_rate": 2.342662595542377e-06,
+      "loss": 0.5883,
+      "step": 9909
+    },
+    {
+      "epoch": 0.6880988751562283,
+      "grad_norm": 3.963603336463058,
+      "learning_rate": 2.3417101335069524e-06,
+      "loss": 0.4375,
+      "step": 9910
+    },
+    {
+      "epoch": 0.6881683099569504,
+      "grad_norm": 4.097531133120274,
+      "learning_rate": 2.3407578059248247e-06,
+      "loss": 0.5242,
+      "step": 9911
+    },
+    {
+      "epoch": 0.6882377447576725,
+      "grad_norm": 3.217336821741175,
+      "learning_rate": 2.339805612844156e-06,
+      "loss": 0.2713,
+      "step": 9912
+    },
+    {
+      "epoch": 0.6883071795583947,
+      "grad_norm": 5.435268303591026,
+      "learning_rate": 2.3388535543131075e-06,
+      "loss": 0.3886,
+      "step": 9913
+    },
+    {
+      "epoch": 0.6883766143591168,
+      "grad_norm": 4.814671256785052,
+      "learning_rate": 2.337901630379837e-06,
+      "loss": 0.4295,
+      "step": 9914
+    },
+    {
+      "epoch": 0.6884460491598389,
+      "grad_norm": 3.7031426441761615,
+      "learning_rate": 2.336949841092491e-06,
+      "loss": 0.374,
+      "step": 9915
+    },
+    {
+      "epoch": 0.6885154839605611,
+      "grad_norm": 3.632389814733023,
+      "learning_rate": 2.3359981864992054e-06,
+      "loss": 0.377,
+      "step": 9916
+    },
+    {
+      "epoch": 0.6885849187612831,
+      "grad_norm": 4.3169679519059505,
+      "learning_rate": 2.3350466666481184e-06,
+      "loss": 0.3984,
+      "step": 9917
+    },
+    {
+      "epoch": 0.6886543535620053,
+      "grad_norm": 4.363839981326813,
+      "learning_rate": 2.334095281587356e-06,
+      "loss": 0.5476,
+      "step": 9918
+    },
+    {
+      "epoch": 0.6887237883627274,
+      "grad_norm": 4.679343375249857,
+      "learning_rate": 2.3331440313650372e-06,
+      "loss": 0.5443,
+      "step": 9919
+    },
+    {
+      "epoch": 0.6887932231634495,
+      "grad_norm": 3.6524224076625575,
+      "learning_rate": 2.3321929160292756e-06,
+      "loss": 0.4618,
+      "step": 9920
+    },
+    {
+      "epoch": 0.6888626579641717,
+      "grad_norm": 3.527208751985019,
+      "learning_rate": 2.3312419356281774e-06,
+      "loss": 0.3985,
+      "step": 9921
+    },
+    {
+      "epoch": 0.6889320927648938,
+      "grad_norm": 3.1291605396034528,
+      "learning_rate": 2.3302910902098423e-06,
+      "loss": 0.3014,
+      "step": 9922
+    },
+    {
+      "epoch": 0.6890015275656158,
+      "grad_norm": 4.058476030252018,
+      "learning_rate": 2.329340379822363e-06,
+      "loss": 0.4411,
+      "step": 9923
+    },
+    {
+      "epoch": 0.689070962366338,
+      "grad_norm": 4.22663055423953,
+      "learning_rate": 2.328389804513825e-06,
+      "loss": 0.4878,
+      "step": 9924
+    },
+    {
+      "epoch": 0.6891403971670601,
+      "grad_norm": 4.171863469028785,
+      "learning_rate": 2.327439364332307e-06,
+      "loss": 0.4672,
+      "step": 9925
+    },
+    {
+      "epoch": 0.6892098319677823,
+      "grad_norm": 2.8447947524819495,
+      "learning_rate": 2.3264890593258827e-06,
+      "loss": 0.2973,
+      "step": 9926
+    },
+    {
+      "epoch": 0.6892792667685044,
+      "grad_norm": 4.250807345218405,
+      "learning_rate": 2.3255388895426143e-06,
+      "loss": 0.5103,
+      "step": 9927
+    },
+    {
+      "epoch": 0.6893487015692265,
+      "grad_norm": 4.560028909629896,
+      "learning_rate": 2.3245888550305644e-06,
+      "loss": 0.3779,
+      "step": 9928
+    },
+    {
+      "epoch": 0.6894181363699486,
+      "grad_norm": 3.7738900747012445,
+      "learning_rate": 2.3236389558377847e-06,
+      "loss": 0.3704,
+      "step": 9929
+    },
+    {
+      "epoch": 0.6894875711706707,
+      "grad_norm": 2.602156630841338,
+      "learning_rate": 2.3226891920123146e-06,
+      "loss": 0.3247,
+      "step": 9930
+    },
+    {
+      "epoch": 0.6895570059713929,
+      "grad_norm": 3.821689861329341,
+      "learning_rate": 2.3217395636021966e-06,
+      "loss": 0.3454,
+      "step": 9931
+    },
+    {
+      "epoch": 0.689626440772115,
+      "grad_norm": 2.871325671516494,
+      "learning_rate": 2.320790070655464e-06,
+      "loss": 0.2711,
+      "step": 9932
+    },
+    {
+      "epoch": 0.6896958755728371,
+      "grad_norm": 2.782848497728131,
+      "learning_rate": 2.319840713220133e-06,
+      "loss": 0.2436,
+      "step": 9933
+    },
+    {
+      "epoch": 0.6897653103735593,
+      "grad_norm": 3.3090362754466724,
+      "learning_rate": 2.318891491344228e-06,
+      "loss": 0.1628,
+      "step": 9934
+    },
+    {
+      "epoch": 0.6898347451742813,
+      "grad_norm": 3.8832291473238048,
+      "learning_rate": 2.3179424050757576e-06,
+      "loss": 0.3062,
+      "step": 9935
+    },
+    {
+      "epoch": 0.6899041799750034,
+      "grad_norm": 3.2256047648077892,
+      "learning_rate": 2.3169934544627254e-06,
+      "loss": 0.2507,
+      "step": 9936
+    },
+    {
+      "epoch": 0.6899736147757256,
+      "grad_norm": 3.700991186835935,
+      "learning_rate": 2.3160446395531284e-06,
+      "loss": 0.3939,
+      "step": 9937
+    },
+    {
+      "epoch": 0.6900430495764477,
+      "grad_norm": 3.6825527103557105,
+      "learning_rate": 2.3150959603949565e-06,
+      "loss": 0.3839,
+      "step": 9938
+    },
+    {
+      "epoch": 0.6901124843771699,
+      "grad_norm": 4.705895078883613,
+      "learning_rate": 2.314147417036193e-06,
+      "loss": 0.5134,
+      "step": 9939
+    },
+    {
+      "epoch": 0.690181919177892,
+      "grad_norm": 4.1956008113146925,
+      "learning_rate": 2.313199009524813e-06,
+      "loss": 0.5509,
+      "step": 9940
+    },
+    {
+      "epoch": 0.690251353978614,
+      "grad_norm": 2.9589955686634224,
+      "learning_rate": 2.3122507379087874e-06,
+      "loss": 0.1923,
+      "step": 9941
+    },
+    {
+      "epoch": 0.6903207887793362,
+      "grad_norm": 3.6815847327890583,
+      "learning_rate": 2.311302602236076e-06,
+      "loss": 0.3182,
+      "step": 9942
+    },
+    {
+      "epoch": 0.6903902235800583,
+      "grad_norm": 4.4253312886036,
+      "learning_rate": 2.3103546025546404e-06,
+      "loss": 0.4737,
+      "step": 9943
+    },
+    {
+      "epoch": 0.6904596583807805,
+      "grad_norm": 4.119312455531526,
+      "learning_rate": 2.3094067389124234e-06,
+      "loss": 0.3395,
+      "step": 9944
+    },
+    {
+      "epoch": 0.6905290931815026,
+      "grad_norm": 3.3771347621407197,
+      "learning_rate": 2.3084590113573675e-06,
+      "loss": 0.3735,
+      "step": 9945
+    },
+    {
+      "epoch": 0.6905985279822247,
+      "grad_norm": 4.2040133796116885,
+      "learning_rate": 2.3075114199374105e-06,
+      "loss": 0.3394,
+      "step": 9946
+    },
+    {
+      "epoch": 0.6906679627829468,
+      "grad_norm": 2.2468502031652426,
+      "learning_rate": 2.3065639647004812e-06,
+      "loss": 0.2361,
+      "step": 9947
+    },
+    {
+      "epoch": 0.6907373975836689,
+      "grad_norm": 3.310621361342557,
+      "learning_rate": 2.305616645694495e-06,
+      "loss": 0.3389,
+      "step": 9948
+    },
+    {
+      "epoch": 0.6908068323843911,
+      "grad_norm": 3.771569545352427,
+      "learning_rate": 2.3046694629673715e-06,
+      "loss": 0.4583,
+      "step": 9949
+    },
+    {
+      "epoch": 0.6908762671851132,
+      "grad_norm": 2.4293178451401922,
+      "learning_rate": 2.303722416567017e-06,
+      "loss": 0.147,
+      "step": 9950
+    },
+    {
+      "epoch": 0.6909457019858353,
+      "grad_norm": 3.912046223810117,
+      "learning_rate": 2.3027755065413316e-06,
+      "loss": 0.4233,
+      "step": 9951
+    },
+    {
+      "epoch": 0.6910151367865575,
+      "grad_norm": 3.2334436837175886,
+      "learning_rate": 2.301828732938209e-06,
+      "loss": 0.3161,
+      "step": 9952
+    },
+    {
+      "epoch": 0.6910845715872795,
+      "grad_norm": 4.447255467970093,
+      "learning_rate": 2.3008820958055358e-06,
+      "loss": 0.5692,
+      "step": 9953
+    },
+    {
+      "epoch": 0.6911540063880016,
+      "grad_norm": 4.796928012056432,
+      "learning_rate": 2.2999355951911917e-06,
+      "loss": 0.5366,
+      "step": 9954
+    },
+    {
+      "epoch": 0.6912234411887238,
+      "grad_norm": 3.494245152203853,
+      "learning_rate": 2.2989892311430502e-06,
+      "loss": 0.2756,
+      "step": 9955
+    },
+    {
+      "epoch": 0.6912928759894459,
+      "grad_norm": 3.586523140757075,
+      "learning_rate": 2.298043003708977e-06,
+      "loss": 0.5022,
+      "step": 9956
+    },
+    {
+      "epoch": 0.6913623107901681,
+      "grad_norm": 2.943459209026951,
+      "learning_rate": 2.297096912936831e-06,
+      "loss": 0.3286,
+      "step": 9957
+    },
+    {
+      "epoch": 0.6914317455908902,
+      "grad_norm": 4.11711616031827,
+      "learning_rate": 2.2961509588744644e-06,
+      "loss": 0.4856,
+      "step": 9958
+    },
+    {
+      "epoch": 0.6915011803916122,
+      "grad_norm": 4.3386763272695354,
+      "learning_rate": 2.295205141569721e-06,
+      "loss": 0.4595,
+      "step": 9959
+    },
+    {
+      "epoch": 0.6915706151923344,
+      "grad_norm": 3.628194814896253,
+      "learning_rate": 2.2942594610704427e-06,
+      "loss": 0.3946,
+      "step": 9960
+    },
+    {
+      "epoch": 0.6916400499930565,
+      "grad_norm": 5.28469705531795,
+      "learning_rate": 2.293313917424461e-06,
+      "loss": 0.6866,
+      "step": 9961
+    },
+    {
+      "epoch": 0.6917094847937787,
+      "grad_norm": 7.377312066440847,
+      "learning_rate": 2.292368510679595e-06,
+      "loss": 0.6803,
+      "step": 9962
+    },
+    {
+      "epoch": 0.6917789195945008,
+      "grad_norm": 3.4551424772539834,
+      "learning_rate": 2.2914232408836678e-06,
+      "loss": 0.4117,
+      "step": 9963
+    },
+    {
+      "epoch": 0.6918483543952229,
+      "grad_norm": 2.8658780482426103,
+      "learning_rate": 2.2904781080844896e-06,
+      "loss": 0.2443,
+      "step": 9964
+    },
+    {
+      "epoch": 0.691917789195945,
+      "grad_norm": 3.5650724855171196,
+      "learning_rate": 2.2895331123298593e-06,
+      "loss": 0.301,
+      "step": 9965
+    },
+    {
+      "epoch": 0.6919872239966671,
+      "grad_norm": 4.208389580726364,
+      "learning_rate": 2.2885882536675785e-06,
+      "loss": 0.4016,
+      "step": 9966
+    },
+    {
+      "epoch": 0.6920566587973892,
+      "grad_norm": 2.545987157531226,
+      "learning_rate": 2.287643532145436e-06,
+      "loss": 0.2877,
+      "step": 9967
+    },
+    {
+      "epoch": 0.6921260935981114,
+      "grad_norm": 3.980263278627177,
+      "learning_rate": 2.2866989478112146e-06,
+      "loss": 0.4073,
+      "step": 9968
+    },
+    {
+      "epoch": 0.6921955283988335,
+      "grad_norm": 4.888326135473543,
+      "learning_rate": 2.2857545007126896e-06,
+      "loss": 0.5063,
+      "step": 9969
+    },
+    {
+      "epoch": 0.6922649631995557,
+      "grad_norm": 3.699699563637695,
+      "learning_rate": 2.2848101908976305e-06,
+      "loss": 0.491,
+      "step": 9970
+    },
+    {
+      "epoch": 0.6923343980002777,
+      "grad_norm": 5.288522628410597,
+      "learning_rate": 2.2838660184137977e-06,
+      "loss": 0.6455,
+      "step": 9971
+    },
+    {
+      "epoch": 0.6924038328009998,
+      "grad_norm": 4.5159676061653355,
+      "learning_rate": 2.2829219833089527e-06,
+      "loss": 0.4156,
+      "step": 9972
+    },
+    {
+      "epoch": 0.692473267601722,
+      "grad_norm": 3.7817293702763526,
+      "learning_rate": 2.2819780856308366e-06,
+      "loss": 0.3614,
+      "step": 9973
+    },
+    {
+      "epoch": 0.6925427024024441,
+      "grad_norm": 4.371606431665333,
+      "learning_rate": 2.2810343254271915e-06,
+      "loss": 0.5014,
+      "step": 9974
+    },
+    {
+      "epoch": 0.6926121372031663,
+      "grad_norm": 3.1271499527580047,
+      "learning_rate": 2.2800907027457574e-06,
+      "loss": 0.3256,
+      "step": 9975
+    },
+    {
+      "epoch": 0.6926815720038884,
+      "grad_norm": 4.610960416169119,
+      "learning_rate": 2.279147217634256e-06,
+      "loss": 0.5309,
+      "step": 9976
+    },
+    {
+      "epoch": 0.6927510068046104,
+      "grad_norm": 2.7230391754338634,
+      "learning_rate": 2.2782038701404075e-06,
+      "loss": 0.2263,
+      "step": 9977
+    },
+    {
+      "epoch": 0.6928204416053326,
+      "grad_norm": 4.013955946987216,
+      "learning_rate": 2.277260660311929e-06,
+      "loss": 0.4409,
+      "step": 9978
+    },
+    {
+      "epoch": 0.6928898764060547,
+      "grad_norm": 2.6577442987258335,
+      "learning_rate": 2.2763175881965256e-06,
+      "loss": 0.2713,
+      "step": 9979
+    },
+    {
+      "epoch": 0.6929593112067768,
+      "grad_norm": 3.882094715395453,
+      "learning_rate": 2.275374653841896e-06,
+      "loss": 0.4543,
+      "step": 9980
+    },
+    {
+      "epoch": 0.693028746007499,
+      "grad_norm": 4.49746276689424,
+      "learning_rate": 2.2744318572957335e-06,
+      "loss": 0.59,
+      "step": 9981
+    },
+    {
+      "epoch": 0.693098180808221,
+      "grad_norm": 2.8683359796632426,
+      "learning_rate": 2.2734891986057238e-06,
+      "loss": 0.3016,
+      "step": 9982
+    },
+    {
+      "epoch": 0.6931676156089432,
+      "grad_norm": 4.248412387745265,
+      "learning_rate": 2.2725466778195447e-06,
+      "loss": 0.4497,
+      "step": 9983
+    },
+    {
+      "epoch": 0.6932370504096653,
+      "grad_norm": 3.4125098081177767,
+      "learning_rate": 2.2716042949848686e-06,
+      "loss": 0.358,
+      "step": 9984
+    },
+    {
+      "epoch": 0.6933064852103874,
+      "grad_norm": 4.646363050802294,
+      "learning_rate": 2.27066205014936e-06,
+      "loss": 0.5883,
+      "step": 9985
+    },
+    {
+      "epoch": 0.6933759200111096,
+      "grad_norm": 3.611587277186948,
+      "learning_rate": 2.2697199433606765e-06,
+      "loss": 0.4002,
+      "step": 9986
+    },
+    {
+      "epoch": 0.6934453548118317,
+      "grad_norm": 9.103611317580159,
+      "learning_rate": 2.2687779746664685e-06,
+      "loss": 0.7162,
+      "step": 9987
+    },
+    {
+      "epoch": 0.6935147896125539,
+      "grad_norm": 3.1084333985097232,
+      "learning_rate": 2.267836144114378e-06,
+      "loss": 0.3405,
+      "step": 9988
+    },
+    {
+      "epoch": 0.6935842244132759,
+      "grad_norm": 3.704921884320625,
+      "learning_rate": 2.266894451752048e-06,
+      "loss": 0.3252,
+      "step": 9989
+    },
+    {
+      "epoch": 0.693653659213998,
+      "grad_norm": 3.2011819661170473,
+      "learning_rate": 2.265952897627102e-06,
+      "loss": 0.3137,
+      "step": 9990
+    },
+    {
+      "epoch": 0.6937230940147202,
+      "grad_norm": 2.4864967599411343,
+      "learning_rate": 2.265011481787162e-06,
+      "loss": 0.2109,
+      "step": 9991
+    },
+    {
+      "epoch": 0.6937925288154423,
+      "grad_norm": 1.7591570786156565,
+      "learning_rate": 2.2640702042798486e-06,
+      "loss": 0.1036,
+      "step": 9992
+    },
+    {
+      "epoch": 0.6938619636161644,
+      "grad_norm": 4.319459278633239,
+      "learning_rate": 2.2631290651527703e-06,
+      "loss": 0.5381,
+      "step": 9993
+    },
+    {
+      "epoch": 0.6939313984168866,
+      "grad_norm": 2.9664689230413224,
+      "learning_rate": 2.262188064453523e-06,
+      "loss": 0.2189,
+      "step": 9994
+    },
+    {
+      "epoch": 0.6940008332176086,
+      "grad_norm": 3.8439451646627623,
+      "learning_rate": 2.2612472022297077e-06,
+      "loss": 0.4578,
+      "step": 9995
+    },
+    {
+      "epoch": 0.6940702680183308,
+      "grad_norm": 3.577245564406531,
+      "learning_rate": 2.260306478528909e-06,
+      "loss": 0.2448,
+      "step": 9996
+    },
+    {
+      "epoch": 0.6941397028190529,
+      "grad_norm": 4.829873328913069,
+      "learning_rate": 2.2593658933987096e-06,
+      "loss": 0.6455,
+      "step": 9997
+    },
+    {
+      "epoch": 0.694209137619775,
+      "grad_norm": 4.112852943719454,
+      "learning_rate": 2.2584254468866813e-06,
+      "loss": 0.5688,
+      "step": 9998
+    },
+    {
+      "epoch": 0.6942785724204972,
+      "grad_norm": 3.9224049659659572,
+      "learning_rate": 2.2574851390403916e-06,
+      "loss": 0.4956,
+      "step": 9999
+    },
+    {
+      "epoch": 0.6943480072212193,
+      "grad_norm": 7.080493875312562,
+      "learning_rate": 2.2565449699074004e-06,
+      "loss": 0.3561,
+      "step": 10000
+    },
+    {
+      "epoch": 0.6944174420219414,
+      "grad_norm": 2.149681706162477,
+      "learning_rate": 2.2556049395352603e-06,
+      "loss": 0.1182,
+      "step": 10001
+    },
+    {
+      "epoch": 0.6944868768226635,
+      "grad_norm": 2.683896947329033,
+      "learning_rate": 2.2546650479715164e-06,
+      "loss": 0.2008,
+      "step": 10002
+    },
+    {
+      "epoch": 0.6945563116233856,
+      "grad_norm": 3.2367838822909887,
+      "learning_rate": 2.253725295263706e-06,
+      "loss": 0.3089,
+      "step": 10003
+    },
+    {
+      "epoch": 0.6946257464241078,
+      "grad_norm": 4.399103027215495,
+      "learning_rate": 2.252785681459367e-06,
+      "loss": 0.405,
+      "step": 10004
+    },
+    {
+      "epoch": 0.6946951812248299,
+      "grad_norm": 3.9502071807486603,
+      "learning_rate": 2.2518462066060172e-06,
+      "loss": 0.4586,
+      "step": 10005
+    },
+    {
+      "epoch": 0.6947646160255521,
+      "grad_norm": 3.197231795733676,
+      "learning_rate": 2.2509068707511743e-06,
+      "loss": 0.2336,
+      "step": 10006
+    },
+    {
+      "epoch": 0.6948340508262741,
+      "grad_norm": 3.79934857895278,
+      "learning_rate": 2.2499676739423555e-06,
+      "loss": 0.4209,
+      "step": 10007
+    },
+    {
+      "epoch": 0.6949034856269962,
+      "grad_norm": 3.6134450302738053,
+      "learning_rate": 2.2490286162270553e-06,
+      "loss": 0.3404,
+      "step": 10008
+    },
+    {
+      "epoch": 0.6949729204277184,
+      "grad_norm": 3.5711417777188394,
+      "learning_rate": 2.2480896976527773e-06,
+      "loss": 0.4149,
+      "step": 10009
+    },
+    {
+      "epoch": 0.6950423552284405,
+      "grad_norm": 3.974676545954251,
+      "learning_rate": 2.247150918267008e-06,
+      "loss": 0.3772,
+      "step": 10010
+    },
+    {
+      "epoch": 0.6951117900291626,
+      "grad_norm": 2.9628168753507675,
+      "learning_rate": 2.2462122781172307e-06,
+      "loss": 0.2978,
+      "step": 10011
+    },
+    {
+      "epoch": 0.6951812248298848,
+      "grad_norm": 4.883904634814912,
+      "learning_rate": 2.24527377725092e-06,
+      "loss": 0.6139,
+      "step": 10012
+    },
+    {
+      "epoch": 0.6952506596306068,
+      "grad_norm": 4.692081150410176,
+      "learning_rate": 2.2443354157155443e-06,
+      "loss": 0.5453,
+      "step": 10013
+    },
+    {
+      "epoch": 0.695320094431329,
+      "grad_norm": 4.546128891364852,
+      "learning_rate": 2.243397193558565e-06,
+      "loss": 0.5357,
+      "step": 10014
+    },
+    {
+      "epoch": 0.6953895292320511,
+      "grad_norm": 3.9268931144631782,
+      "learning_rate": 2.2424591108274367e-06,
+      "loss": 0.5201,
+      "step": 10015
+    },
+    {
+      "epoch": 0.6954589640327732,
+      "grad_norm": 5.848504035181688,
+      "learning_rate": 2.241521167569606e-06,
+      "loss": 0.7109,
+      "step": 10016
+    },
+    {
+      "epoch": 0.6955283988334954,
+      "grad_norm": 3.703538044922385,
+      "learning_rate": 2.2405833638325115e-06,
+      "loss": 0.4152,
+      "step": 10017
+    },
+    {
+      "epoch": 0.6955978336342175,
+      "grad_norm": 4.104278584820147,
+      "learning_rate": 2.2396456996635914e-06,
+      "loss": 0.3604,
+      "step": 10018
+    },
+    {
+      "epoch": 0.6956672684349396,
+      "grad_norm": 4.30717670849888,
+      "learning_rate": 2.2387081751102663e-06,
+      "loss": 0.5724,
+      "step": 10019
+    },
+    {
+      "epoch": 0.6957367032356617,
+      "grad_norm": 3.4510952098220775,
+      "learning_rate": 2.237770790219956e-06,
+      "loss": 0.424,
+      "step": 10020
+    },
+    {
+      "epoch": 0.6958061380363838,
+      "grad_norm": 3.997596727567857,
+      "learning_rate": 2.236833545040074e-06,
+      "loss": 0.5152,
+      "step": 10021
+    },
+    {
+      "epoch": 0.695875572837106,
+      "grad_norm": 3.6238806879575987,
+      "learning_rate": 2.235896439618028e-06,
+      "loss": 0.3756,
+      "step": 10022
+    },
+    {
+      "epoch": 0.6959450076378281,
+      "grad_norm": 3.5828294717131772,
+      "learning_rate": 2.234959474001207e-06,
+      "loss": 0.2976,
+      "step": 10023
+    },
+    {
+      "epoch": 0.6960144424385502,
+      "grad_norm": 3.927218849538247,
+      "learning_rate": 2.2340226482370097e-06,
+      "loss": 0.482,
+      "step": 10024
+    },
+    {
+      "epoch": 0.6960838772392723,
+      "grad_norm": 4.010839985444321,
+      "learning_rate": 2.233085962372819e-06,
+      "loss": 0.4852,
+      "step": 10025
+    },
+    {
+      "epoch": 0.6961533120399944,
+      "grad_norm": 4.045678644749858,
+      "learning_rate": 2.232149416456005e-06,
+      "loss": 0.4157,
+      "step": 10026
+    },
+    {
+      "epoch": 0.6962227468407166,
+      "grad_norm": 6.947279273016797,
+      "learning_rate": 2.2312130105339436e-06,
+      "loss": 0.5106,
+      "step": 10027
+    },
+    {
+      "epoch": 0.6962921816414387,
+      "grad_norm": 3.2125717214595615,
+      "learning_rate": 2.2302767446539954e-06,
+      "loss": 0.3465,
+      "step": 10028
+    },
+    {
+      "epoch": 0.6963616164421608,
+      "grad_norm": 3.9171374111579405,
+      "learning_rate": 2.2293406188635147e-06,
+      "loss": 0.4067,
+      "step": 10029
+    },
+    {
+      "epoch": 0.696431051242883,
+      "grad_norm": 3.089140665280616,
+      "learning_rate": 2.2284046332098506e-06,
+      "loss": 0.2562,
+      "step": 10030
+    },
+    {
+      "epoch": 0.696500486043605,
+      "grad_norm": 3.0919552022676218,
+      "learning_rate": 2.227468787740344e-06,
+      "loss": 0.3265,
+      "step": 10031
+    },
+    {
+      "epoch": 0.6965699208443272,
+      "grad_norm": 3.6911070740216667,
+      "learning_rate": 2.2265330825023284e-06,
+      "loss": 0.4252,
+      "step": 10032
+    },
+    {
+      "epoch": 0.6966393556450493,
+      "grad_norm": 3.050989252972308,
+      "learning_rate": 2.2255975175431316e-06,
+      "loss": 0.3978,
+      "step": 10033
+    },
+    {
+      "epoch": 0.6967087904457714,
+      "grad_norm": 3.7061832018046426,
+      "learning_rate": 2.2246620929100727e-06,
+      "loss": 0.4131,
+      "step": 10034
+    },
+    {
+      "epoch": 0.6967782252464936,
+      "grad_norm": 3.416150136335977,
+      "learning_rate": 2.2237268086504627e-06,
+      "loss": 0.2413,
+      "step": 10035
+    },
+    {
+      "epoch": 0.6968476600472157,
+      "grad_norm": 4.339064799852144,
+      "learning_rate": 2.2227916648116125e-06,
+      "loss": 0.435,
+      "step": 10036
+    },
+    {
+      "epoch": 0.6969170948479377,
+      "grad_norm": 5.139021932495289,
+      "learning_rate": 2.221856661440816e-06,
+      "loss": 0.4812,
+      "step": 10037
+    },
+    {
+      "epoch": 0.6969865296486599,
+      "grad_norm": 3.8684957708251866,
+      "learning_rate": 2.2209217985853636e-06,
+      "loss": 0.1889,
+      "step": 10038
+    },
+    {
+      "epoch": 0.697055964449382,
+      "grad_norm": 4.68522212726937,
+      "learning_rate": 2.2199870762925457e-06,
+      "loss": 0.6465,
+      "step": 10039
+    },
+    {
+      "epoch": 0.6971253992501042,
+      "grad_norm": 2.8143657562480313,
+      "learning_rate": 2.2190524946096324e-06,
+      "loss": 0.2094,
+      "step": 10040
+    },
+    {
+      "epoch": 0.6971948340508263,
+      "grad_norm": 3.7923248249336146,
+      "learning_rate": 2.2181180535838985e-06,
+      "loss": 0.4773,
+      "step": 10041
+    },
+    {
+      "epoch": 0.6972642688515484,
+      "grad_norm": 4.682983807369046,
+      "learning_rate": 2.2171837532626056e-06,
+      "loss": 0.5874,
+      "step": 10042
+    },
+    {
+      "epoch": 0.6973337036522705,
+      "grad_norm": 3.37847923647,
+      "learning_rate": 2.2162495936930096e-06,
+      "loss": 0.3527,
+      "step": 10043
+    },
+    {
+      "epoch": 0.6974031384529926,
+      "grad_norm": 4.067124633327287,
+      "learning_rate": 2.2153155749223593e-06,
+      "loss": 0.5589,
+      "step": 10044
+    },
+    {
+      "epoch": 0.6974725732537148,
+      "grad_norm": 7.267528293294326,
+      "learning_rate": 2.2143816969978955e-06,
+      "loss": 0.4738,
+      "step": 10045
+    },
+    {
+      "epoch": 0.6975420080544369,
+      "grad_norm": 3.196400952828588,
+      "learning_rate": 2.213447959966854e-06,
+      "loss": 0.3386,
+      "step": 10046
+    },
+    {
+      "epoch": 0.697611442855159,
+      "grad_norm": 4.706714981899952,
+      "learning_rate": 2.212514363876461e-06,
+      "loss": 0.5165,
+      "step": 10047
+    },
+    {
+      "epoch": 0.6976808776558812,
+      "grad_norm": 3.1959109057736574,
+      "learning_rate": 2.2115809087739377e-06,
+      "loss": 0.2976,
+      "step": 10048
+    },
+    {
+      "epoch": 0.6977503124566032,
+      "grad_norm": 3.3241351387592126,
+      "learning_rate": 2.2106475947064942e-06,
+      "loss": 0.4036,
+      "step": 10049
+    },
+    {
+      "epoch": 0.6978197472573253,
+      "grad_norm": 4.0335666970866475,
+      "learning_rate": 2.2097144217213433e-06,
+      "loss": 0.5312,
+      "step": 10050
+    },
+    {
+      "epoch": 0.6978891820580475,
+      "grad_norm": 4.272895188798744,
+      "learning_rate": 2.2087813898656775e-06,
+      "loss": 0.5471,
+      "step": 10051
+    },
+    {
+      "epoch": 0.6979586168587696,
+      "grad_norm": 3.20167676167326,
+      "learning_rate": 2.2078484991866887e-06,
+      "loss": 0.3573,
+      "step": 10052
+    },
+    {
+      "epoch": 0.6980280516594918,
+      "grad_norm": 5.179061016554086,
+      "learning_rate": 2.2069157497315653e-06,
+      "loss": 0.6509,
+      "step": 10053
+    },
+    {
+      "epoch": 0.6980974864602139,
+      "grad_norm": 3.481199931831641,
+      "learning_rate": 2.2059831415474843e-06,
+      "loss": 0.4277,
+      "step": 10054
+    },
+    {
+      "epoch": 0.6981669212609359,
+      "grad_norm": 4.2869130508936495,
+      "learning_rate": 2.2050506746816102e-06,
+      "loss": 0.4134,
+      "step": 10055
+    },
+    {
+      "epoch": 0.6982363560616581,
+      "grad_norm": 4.687752346560159,
+      "learning_rate": 2.204118349181113e-06,
+      "loss": 0.5337,
+      "step": 10056
+    },
+    {
+      "epoch": 0.6983057908623802,
+      "grad_norm": 3.710086330418404,
+      "learning_rate": 2.2031861650931474e-06,
+      "loss": 0.447,
+      "step": 10057
+    },
+    {
+      "epoch": 0.6983752256631024,
+      "grad_norm": 3.6122840292723577,
+      "learning_rate": 2.202254122464857e-06,
+      "loss": 0.449,
+      "step": 10058
+    },
+    {
+      "epoch": 0.6984446604638245,
+      "grad_norm": 3.5791070010525057,
+      "learning_rate": 2.201322221343389e-06,
+      "loss": 0.3884,
+      "step": 10059
+    },
+    {
+      "epoch": 0.6985140952645466,
+      "grad_norm": 3.363298056010146,
+      "learning_rate": 2.2003904617758764e-06,
+      "loss": 0.3556,
+      "step": 10060
+    },
+    {
+      "epoch": 0.6985835300652687,
+      "grad_norm": 2.8658020079568582,
+      "learning_rate": 2.199458843809446e-06,
+      "loss": 0.3605,
+      "step": 10061
+    },
+    {
+      "epoch": 0.6986529648659908,
+      "grad_norm": 3.733506339657755,
+      "learning_rate": 2.198527367491218e-06,
+      "loss": 0.3846,
+      "step": 10062
+    },
+    {
+      "epoch": 0.6987223996667129,
+      "grad_norm": 3.499459273869188,
+      "learning_rate": 2.1975960328683065e-06,
+      "loss": 0.5001,
+      "step": 10063
+    },
+    {
+      "epoch": 0.6987918344674351,
+      "grad_norm": 3.2245211192724277,
+      "learning_rate": 2.196664839987816e-06,
+      "loss": 0.2981,
+      "step": 10064
+    },
+    {
+      "epoch": 0.6988612692681572,
+      "grad_norm": 3.8877544224514193,
+      "learning_rate": 2.1957337888968454e-06,
+      "loss": 0.54,
+      "step": 10065
+    },
+    {
+      "epoch": 0.6989307040688794,
+      "grad_norm": 2.956624330306984,
+      "learning_rate": 2.1948028796424873e-06,
+      "loss": 0.2034,
+      "step": 10066
+    },
+    {
+      "epoch": 0.6990001388696014,
+      "grad_norm": 4.124842478150322,
+      "learning_rate": 2.1938721122718234e-06,
+      "loss": 0.5741,
+      "step": 10067
+    },
+    {
+      "epoch": 0.6990695736703235,
+      "grad_norm": 3.3447640594206094,
+      "learning_rate": 2.1929414868319356e-06,
+      "loss": 0.2739,
+      "step": 10068
+    },
+    {
+      "epoch": 0.6991390084710457,
+      "grad_norm": 3.6518967157681943,
+      "learning_rate": 2.1920110033698876e-06,
+      "loss": 0.4074,
+      "step": 10069
+    },
+    {
+      "epoch": 0.6992084432717678,
+      "grad_norm": 3.5314287576611534,
+      "learning_rate": 2.1910806619327473e-06,
+      "loss": 0.3233,
+      "step": 10070
+    },
+    {
+      "epoch": 0.69927787807249,
+      "grad_norm": 2.9737161884832304,
+      "learning_rate": 2.190150462567569e-06,
+      "loss": 0.3309,
+      "step": 10071
+    },
+    {
+      "epoch": 0.6993473128732121,
+      "grad_norm": 5.363532773615369,
+      "learning_rate": 2.1892204053214007e-06,
+      "loss": 0.5382,
+      "step": 10072
+    },
+    {
+      "epoch": 0.6994167476739341,
+      "grad_norm": 2.816608238515145,
+      "learning_rate": 2.188290490241284e-06,
+      "loss": 0.3418,
+      "step": 10073
+    },
+    {
+      "epoch": 0.6994861824746563,
+      "grad_norm": 3.2159791791376935,
+      "learning_rate": 2.187360717374252e-06,
+      "loss": 0.3258,
+      "step": 10074
+    },
+    {
+      "epoch": 0.6995556172753784,
+      "grad_norm": 4.838792065796856,
+      "learning_rate": 2.186431086767332e-06,
+      "loss": 0.449,
+      "step": 10075
+    },
+    {
+      "epoch": 0.6996250520761006,
+      "grad_norm": 3.596881700694623,
+      "learning_rate": 2.185501598467545e-06,
+      "loss": 0.2655,
+      "step": 10076
+    },
+    {
+      "epoch": 0.6996944868768227,
+      "grad_norm": 3.0302052317278707,
+      "learning_rate": 2.184572252521901e-06,
+      "loss": 0.2253,
+      "step": 10077
+    },
+    {
+      "epoch": 0.6997639216775448,
+      "grad_norm": 5.662021755560674,
+      "learning_rate": 2.1836430489774057e-06,
+      "loss": 0.7511,
+      "step": 10078
+    },
+    {
+      "epoch": 0.6998333564782669,
+      "grad_norm": 4.05518579895283,
+      "learning_rate": 2.182713987881061e-06,
+      "loss": 0.5213,
+      "step": 10079
+    },
+    {
+      "epoch": 0.699902791278989,
+      "grad_norm": 3.4533161515607746,
+      "learning_rate": 2.1817850692798537e-06,
+      "loss": 0.359,
+      "step": 10080
+    },
+    {
+      "epoch": 0.6999722260797111,
+      "grad_norm": 4.006006469657829,
+      "learning_rate": 2.1808562932207668e-06,
+      "loss": 0.4392,
+      "step": 10081
+    },
+    {
+      "epoch": 0.7000416608804333,
+      "grad_norm": 3.6394153217505987,
+      "learning_rate": 2.1799276597507814e-06,
+      "loss": 0.3765,
+      "step": 10082
+    },
+    {
+      "epoch": 0.7001110956811554,
+      "grad_norm": 5.083913145981617,
+      "learning_rate": 2.178999168916862e-06,
+      "loss": 0.4633,
+      "step": 10083
+    },
+    {
+      "epoch": 0.7001805304818776,
+      "grad_norm": 3.497667866638374,
+      "learning_rate": 2.178070820765971e-06,
+      "loss": 0.2848,
+      "step": 10084
+    },
+    {
+      "epoch": 0.7002499652825996,
+      "grad_norm": 3.763865628904409,
+      "learning_rate": 2.177142615345067e-06,
+      "loss": 0.4864,
+      "step": 10085
+    },
+    {
+      "epoch": 0.7003194000833217,
+      "grad_norm": 5.09560591032557,
+      "learning_rate": 2.1762145527010962e-06,
+      "loss": 0.2536,
+      "step": 10086
+    },
+    {
+      "epoch": 0.7003888348840439,
+      "grad_norm": 3.8848795232393862,
+      "learning_rate": 2.175286632880995e-06,
+      "loss": 0.2773,
+      "step": 10087
+    },
+    {
+      "epoch": 0.700458269684766,
+      "grad_norm": 3.5734470866731005,
+      "learning_rate": 2.1743588559317013e-06,
+      "loss": 0.375,
+      "step": 10088
+    },
+    {
+      "epoch": 0.7005277044854882,
+      "grad_norm": 3.031106340519871,
+      "learning_rate": 2.1734312219001392e-06,
+      "loss": 0.3704,
+      "step": 10089
+    },
+    {
+      "epoch": 0.7005971392862103,
+      "grad_norm": 3.1666146648399383,
+      "learning_rate": 2.172503730833227e-06,
+      "loss": 0.3584,
+      "step": 10090
+    },
+    {
+      "epoch": 0.7006665740869323,
+      "grad_norm": 5.310853558570267,
+      "learning_rate": 2.1715763827778767e-06,
+      "loss": 0.7335,
+      "step": 10091
+    },
+    {
+      "epoch": 0.7007360088876545,
+      "grad_norm": 3.3999933984337485,
+      "learning_rate": 2.1706491777809923e-06,
+      "loss": 0.3928,
+      "step": 10092
+    },
+    {
+      "epoch": 0.7008054436883766,
+      "grad_norm": 3.9601944015225734,
+      "learning_rate": 2.169722115889471e-06,
+      "loss": 0.4921,
+      "step": 10093
+    },
+    {
+      "epoch": 0.7008748784890987,
+      "grad_norm": 3.5809445318515634,
+      "learning_rate": 2.1687951971502024e-06,
+      "loss": 0.3948,
+      "step": 10094
+    },
+    {
+      "epoch": 0.7009443132898209,
+      "grad_norm": 3.6800108860372216,
+      "learning_rate": 2.1678684216100693e-06,
+      "loss": 0.3082,
+      "step": 10095
+    },
+    {
+      "epoch": 0.701013748090543,
+      "grad_norm": 1.8879790660417708,
+      "learning_rate": 2.1669417893159444e-06,
+      "loss": 0.1529,
+      "step": 10096
+    },
+    {
+      "epoch": 0.7010831828912651,
+      "grad_norm": 4.283057485069001,
+      "learning_rate": 2.166015300314702e-06,
+      "loss": 0.4475,
+      "step": 10097
+    },
+    {
+      "epoch": 0.7011526176919872,
+      "grad_norm": 3.0596570729907016,
+      "learning_rate": 2.165088954653195e-06,
+      "loss": 0.1993,
+      "step": 10098
+    },
+    {
+      "epoch": 0.7012220524927093,
+      "grad_norm": 3.1148095735171686,
+      "learning_rate": 2.1641627523782827e-06,
+      "loss": 0.2377,
+      "step": 10099
+    },
+    {
+      "epoch": 0.7012914872934315,
+      "grad_norm": 4.536049232565741,
+      "learning_rate": 2.163236693536811e-06,
+      "loss": 0.5971,
+      "step": 10100
+    },
+    {
+      "epoch": 0.7013609220941536,
+      "grad_norm": 4.033601922306188,
+      "learning_rate": 2.162310778175614e-06,
+      "loss": 0.5141,
+      "step": 10101
+    },
+    {
+      "epoch": 0.7014303568948758,
+      "grad_norm": 3.1660411232714454,
+      "learning_rate": 2.161385006341528e-06,
+      "loss": 0.3322,
+      "step": 10102
+    },
+    {
+      "epoch": 0.7014997916955978,
+      "grad_norm": 4.437380457607839,
+      "learning_rate": 2.160459378081377e-06,
+      "loss": 0.441,
+      "step": 10103
+    },
+    {
+      "epoch": 0.7015692264963199,
+      "grad_norm": 2.5691066837552667,
+      "learning_rate": 2.1595338934419774e-06,
+      "loss": 0.2863,
+      "step": 10104
+    },
+    {
+      "epoch": 0.7016386612970421,
+      "grad_norm": 4.297599010734527,
+      "learning_rate": 2.158608552470139e-06,
+      "loss": 0.3735,
+      "step": 10105
+    },
+    {
+      "epoch": 0.7017080960977642,
+      "grad_norm": 3.4297195893863437,
+      "learning_rate": 2.1576833552126645e-06,
+      "loss": 0.265,
+      "step": 10106
+    },
+    {
+      "epoch": 0.7017775308984863,
+      "grad_norm": 3.0915197359133093,
+      "learning_rate": 2.1567583017163502e-06,
+      "loss": 0.3039,
+      "step": 10107
+    },
+    {
+      "epoch": 0.7018469656992085,
+      "grad_norm": 3.8238107892238777,
+      "learning_rate": 2.155833392027983e-06,
+      "loss": 0.3724,
+      "step": 10108
+    },
+    {
+      "epoch": 0.7019164004999305,
+      "grad_norm": 4.537413695776585,
+      "learning_rate": 2.1549086261943446e-06,
+      "loss": 0.6079,
+      "step": 10109
+    },
+    {
+      "epoch": 0.7019858353006527,
+      "grad_norm": 3.8011039643377877,
+      "learning_rate": 2.153984004262207e-06,
+      "loss": 0.4688,
+      "step": 10110
+    },
+    {
+      "epoch": 0.7020552701013748,
+      "grad_norm": 3.0175582453140004,
+      "learning_rate": 2.1530595262783415e-06,
+      "loss": 0.2592,
+      "step": 10111
+    },
+    {
+      "epoch": 0.7021247049020969,
+      "grad_norm": 4.192938878072801,
+      "learning_rate": 2.1521351922895018e-06,
+      "loss": 0.5948,
+      "step": 10112
+    },
+    {
+      "epoch": 0.7021941397028191,
+      "grad_norm": 3.9076778633930167,
+      "learning_rate": 2.15121100234244e-06,
+      "loss": 0.2829,
+      "step": 10113
+    },
+    {
+      "epoch": 0.7022635745035412,
+      "grad_norm": 4.579189177232965,
+      "learning_rate": 2.1502869564839054e-06,
+      "loss": 0.5742,
+      "step": 10114
+    },
+    {
+      "epoch": 0.7023330093042633,
+      "grad_norm": 4.358471117786607,
+      "learning_rate": 2.14936305476063e-06,
+      "loss": 0.441,
+      "step": 10115
+    },
+    {
+      "epoch": 0.7024024441049854,
+      "grad_norm": 3.3581020972103284,
+      "learning_rate": 2.1484392972193436e-06,
+      "loss": 0.4672,
+      "step": 10116
+    },
+    {
+      "epoch": 0.7024718789057075,
+      "grad_norm": 4.295773929686663,
+      "learning_rate": 2.1475156839067725e-06,
+      "loss": 0.4964,
+      "step": 10117
+    },
+    {
+      "epoch": 0.7025413137064297,
+      "grad_norm": 2.8786730977589405,
+      "learning_rate": 2.146592214869632e-06,
+      "loss": 0.2112,
+      "step": 10118
+    },
+    {
+      "epoch": 0.7026107485071518,
+      "grad_norm": 4.401165107551852,
+      "learning_rate": 2.145668890154626e-06,
+      "loss": 0.4498,
+      "step": 10119
+    },
+    {
+      "epoch": 0.7026801833078739,
+      "grad_norm": 3.582131523080712,
+      "learning_rate": 2.1447457098084583e-06,
+      "loss": 0.3764,
+      "step": 10120
+    },
+    {
+      "epoch": 0.702749618108596,
+      "grad_norm": 3.564158365752833,
+      "learning_rate": 2.143822673877822e-06,
+      "loss": 0.2471,
+      "step": 10121
+    },
+    {
+      "epoch": 0.7028190529093181,
+      "grad_norm": 2.8385295597486095,
+      "learning_rate": 2.142899782409403e-06,
+      "loss": 0.3147,
+      "step": 10122
+    },
+    {
+      "epoch": 0.7028884877100403,
+      "grad_norm": 3.683953361570253,
+      "learning_rate": 2.1419770354498798e-06,
+      "loss": 0.4972,
+      "step": 10123
+    },
+    {
+      "epoch": 0.7029579225107624,
+      "grad_norm": 3.614069427398909,
+      "learning_rate": 2.141054433045924e-06,
+      "loss": 0.4029,
+      "step": 10124
+    },
+    {
+      "epoch": 0.7030273573114845,
+      "grad_norm": 3.542946864349542,
+      "learning_rate": 2.1401319752441997e-06,
+      "loss": 0.3783,
+      "step": 10125
+    },
+    {
+      "epoch": 0.7030967921122067,
+      "grad_norm": 8.033213625077124,
+      "learning_rate": 2.1392096620913643e-06,
+      "loss": 0.518,
+      "step": 10126
+    },
+    {
+      "epoch": 0.7031662269129287,
+      "grad_norm": 3.551196811249551,
+      "learning_rate": 2.138287493634067e-06,
+      "loss": 0.3438,
+      "step": 10127
+    },
+    {
+      "epoch": 0.7032356617136509,
+      "grad_norm": 3.4525765117952294,
+      "learning_rate": 2.137365469918948e-06,
+      "loss": 0.3609,
+      "step": 10128
+    },
+    {
+      "epoch": 0.703305096514373,
+      "grad_norm": 4.4059092285601,
+      "learning_rate": 2.136443590992648e-06,
+      "loss": 0.4914,
+      "step": 10129
+    },
+    {
+      "epoch": 0.7033745313150951,
+      "grad_norm": 3.923718719552914,
+      "learning_rate": 2.135521856901787e-06,
+      "loss": 0.4075,
+      "step": 10130
+    },
+    {
+      "epoch": 0.7034439661158173,
+      "grad_norm": 3.3042187581259883,
+      "learning_rate": 2.134600267692991e-06,
+      "loss": 0.4363,
+      "step": 10131
+    },
+    {
+      "epoch": 0.7035134009165394,
+      "grad_norm": 4.969205513647791,
+      "learning_rate": 2.133678823412873e-06,
+      "loss": 0.383,
+      "step": 10132
+    },
+    {
+      "epoch": 0.7035828357172615,
+      "grad_norm": 4.977250202750475,
+      "learning_rate": 2.1327575241080333e-06,
+      "loss": 0.8724,
+      "step": 10133
+    },
+    {
+      "epoch": 0.7036522705179836,
+      "grad_norm": 3.048437017004731,
+      "learning_rate": 2.131836369825075e-06,
+      "loss": 0.2658,
+      "step": 10134
+    },
+    {
+      "epoch": 0.7037217053187057,
+      "grad_norm": 3.1852701075664,
+      "learning_rate": 2.1309153606105875e-06,
+      "loss": 0.339,
+      "step": 10135
+    },
+    {
+      "epoch": 0.7037911401194279,
+      "grad_norm": 3.7120933755632817,
+      "learning_rate": 2.1299944965111546e-06,
+      "loss": 0.3654,
+      "step": 10136
+    },
+    {
+      "epoch": 0.70386057492015,
+      "grad_norm": 4.1932691282580965,
+      "learning_rate": 2.129073777573352e-06,
+      "loss": 0.4264,
+      "step": 10137
+    },
+    {
+      "epoch": 0.703930009720872,
+      "grad_norm": 4.028992983185161,
+      "learning_rate": 2.12815320384375e-06,
+      "loss": 0.5207,
+      "step": 10138
+    },
+    {
+      "epoch": 0.7039994445215942,
+      "grad_norm": 3.9215330292588253,
+      "learning_rate": 2.127232775368908e-06,
+      "loss": 0.4332,
+      "step": 10139
+    },
+    {
+      "epoch": 0.7040688793223163,
+      "grad_norm": 2.667881894955853,
+      "learning_rate": 2.1263124921953824e-06,
+      "loss": 0.2091,
+      "step": 10140
+    },
+    {
+      "epoch": 0.7041383141230385,
+      "grad_norm": 4.288862069188275,
+      "learning_rate": 2.1253923543697193e-06,
+      "loss": 0.745,
+      "step": 10141
+    },
+    {
+      "epoch": 0.7042077489237606,
+      "grad_norm": 4.056514171117148,
+      "learning_rate": 2.1244723619384555e-06,
+      "loss": 0.3935,
+      "step": 10142
+    },
+    {
+      "epoch": 0.7042771837244827,
+      "grad_norm": 4.14125095936072,
+      "learning_rate": 2.1235525149481295e-06,
+      "loss": 0.4191,
+      "step": 10143
+    },
+    {
+      "epoch": 0.7043466185252049,
+      "grad_norm": 5.300532006437179,
+      "learning_rate": 2.1226328134452612e-06,
+      "loss": 0.7302,
+      "step": 10144
+    },
+    {
+      "epoch": 0.7044160533259269,
+      "grad_norm": 3.8349528683035974,
+      "learning_rate": 2.121713257476367e-06,
+      "loss": 0.4243,
+      "step": 10145
+    },
+    {
+      "epoch": 0.7044854881266491,
+      "grad_norm": 3.451271500956862,
+      "learning_rate": 2.120793847087963e-06,
+      "loss": 0.2921,
+      "step": 10146
+    },
+    {
+      "epoch": 0.7045549229273712,
+      "grad_norm": 4.122798255141095,
+      "learning_rate": 2.119874582326546e-06,
+      "loss": 0.3171,
+      "step": 10147
+    },
+    {
+      "epoch": 0.7046243577280933,
+      "grad_norm": 4.318609853013091,
+      "learning_rate": 2.1189554632386116e-06,
+      "loss": 0.4093,
+      "step": 10148
+    },
+    {
+      "epoch": 0.7046937925288155,
+      "grad_norm": 4.536505869543018,
+      "learning_rate": 2.1180364898706525e-06,
+      "loss": 0.5372,
+      "step": 10149
+    },
+    {
+      "epoch": 0.7047632273295376,
+      "grad_norm": 3.1117815259459114,
+      "learning_rate": 2.117117662269146e-06,
+      "loss": 0.3515,
+      "step": 10150
+    },
+    {
+      "epoch": 0.7048326621302596,
+      "grad_norm": 3.7225651864418423,
+      "learning_rate": 2.1161989804805657e-06,
+      "loss": 0.46,
+      "step": 10151
+    },
+    {
+      "epoch": 0.7049020969309818,
+      "grad_norm": 4.761140005836736,
+      "learning_rate": 2.115280444551378e-06,
+      "loss": 0.6187,
+      "step": 10152
+    },
+    {
+      "epoch": 0.7049715317317039,
+      "grad_norm": 3.822788202363711,
+      "learning_rate": 2.1143620545280414e-06,
+      "loss": 0.2576,
+      "step": 10153
+    },
+    {
+      "epoch": 0.7050409665324261,
+      "grad_norm": 3.5175776023152103,
+      "learning_rate": 2.113443810457006e-06,
+      "loss": 0.3512,
+      "step": 10154
+    },
+    {
+      "epoch": 0.7051104013331482,
+      "grad_norm": 3.581592892711257,
+      "learning_rate": 2.112525712384717e-06,
+      "loss": 0.3565,
+      "step": 10155
+    },
+    {
+      "epoch": 0.7051798361338703,
+      "grad_norm": 4.831493848511022,
+      "learning_rate": 2.1116077603576097e-06,
+      "loss": 0.3864,
+      "step": 10156
+    },
+    {
+      "epoch": 0.7052492709345924,
+      "grad_norm": 3.6280485595893075,
+      "learning_rate": 2.110689954422114e-06,
+      "loss": 0.3548,
+      "step": 10157
+    },
+    {
+      "epoch": 0.7053187057353145,
+      "grad_norm": 3.5343620364872383,
+      "learning_rate": 2.10977229462465e-06,
+      "loss": 0.4109,
+      "step": 10158
+    },
+    {
+      "epoch": 0.7053881405360367,
+      "grad_norm": 4.035579577403022,
+      "learning_rate": 2.1088547810116315e-06,
+      "loss": 0.5433,
+      "step": 10159
+    },
+    {
+      "epoch": 0.7054575753367588,
+      "grad_norm": 4.617696581174488,
+      "learning_rate": 2.107937413629469e-06,
+      "loss": 0.693,
+      "step": 10160
+    },
+    {
+      "epoch": 0.7055270101374809,
+      "grad_norm": 4.129150431219783,
+      "learning_rate": 2.10702019252456e-06,
+      "loss": 0.5244,
+      "step": 10161
+    },
+    {
+      "epoch": 0.7055964449382031,
+      "grad_norm": 4.369682756553619,
+      "learning_rate": 2.106103117743294e-06,
+      "loss": 0.5638,
+      "step": 10162
+    },
+    {
+      "epoch": 0.7056658797389251,
+      "grad_norm": 2.0971993730631397,
+      "learning_rate": 2.1051861893320585e-06,
+      "loss": 0.0466,
+      "step": 10163
+    },
+    {
+      "epoch": 0.7057353145396472,
+      "grad_norm": 4.1585937092775165,
+      "learning_rate": 2.1042694073372318e-06,
+      "loss": 0.4475,
+      "step": 10164
+    },
+    {
+      "epoch": 0.7058047493403694,
+      "grad_norm": 4.1947423425254895,
+      "learning_rate": 2.1033527718051776e-06,
+      "loss": 0.6706,
+      "step": 10165
+    },
+    {
+      "epoch": 0.7058741841410915,
+      "grad_norm": 4.105399399382044,
+      "learning_rate": 2.1024362827822655e-06,
+      "loss": 0.4999,
+      "step": 10166
+    },
+    {
+      "epoch": 0.7059436189418137,
+      "grad_norm": 3.5046972621438783,
+      "learning_rate": 2.101519940314846e-06,
+      "loss": 0.288,
+      "step": 10167
+    },
+    {
+      "epoch": 0.7060130537425358,
+      "grad_norm": 3.8731571912563667,
+      "learning_rate": 2.1006037444492687e-06,
+      "loss": 0.3247,
+      "step": 10168
+    },
+    {
+      "epoch": 0.7060824885432578,
+      "grad_norm": 3.4874687198632968,
+      "learning_rate": 2.0996876952318735e-06,
+      "loss": 0.3886,
+      "step": 10169
+    },
+    {
+      "epoch": 0.70615192334398,
+      "grad_norm": 2.883904383180147,
+      "learning_rate": 2.0987717927089923e-06,
+      "loss": 0.1806,
+      "step": 10170
+    },
+    {
+      "epoch": 0.7062213581447021,
+      "grad_norm": 3.0748646131919237,
+      "learning_rate": 2.0978560369269514e-06,
+      "loss": 0.2195,
+      "step": 10171
+    },
+    {
+      "epoch": 0.7062907929454243,
+      "grad_norm": 4.563377073606305,
+      "learning_rate": 2.096940427932067e-06,
+      "loss": 0.4869,
+      "step": 10172
+    },
+    {
+      "epoch": 0.7063602277461464,
+      "grad_norm": 4.367218883282367,
+      "learning_rate": 2.0960249657706515e-06,
+      "loss": 0.5821,
+      "step": 10173
+    },
+    {
+      "epoch": 0.7064296625468685,
+      "grad_norm": 3.6092434657124803,
+      "learning_rate": 2.095109650489006e-06,
+      "loss": 0.3408,
+      "step": 10174
+    },
+    {
+      "epoch": 0.7064990973475906,
+      "grad_norm": 3.8476093198889325,
+      "learning_rate": 2.0941944821334296e-06,
+      "loss": 0.4668,
+      "step": 10175
+    },
+    {
+      "epoch": 0.7065685321483127,
+      "grad_norm": 3.9182401058194114,
+      "learning_rate": 2.0932794607502073e-06,
+      "loss": 0.4615,
+      "step": 10176
+    },
+    {
+      "epoch": 0.7066379669490348,
+      "grad_norm": 4.088550420666286,
+      "learning_rate": 2.0923645863856183e-06,
+      "loss": 0.3528,
+      "step": 10177
+    },
+    {
+      "epoch": 0.706707401749757,
+      "grad_norm": 3.5557788914237904,
+      "learning_rate": 2.0914498590859405e-06,
+      "loss": 0.3379,
+      "step": 10178
+    },
+    {
+      "epoch": 0.7067768365504791,
+      "grad_norm": 4.135823979388475,
+      "learning_rate": 2.090535278897437e-06,
+      "loss": 0.5393,
+      "step": 10179
+    },
+    {
+      "epoch": 0.7068462713512013,
+      "grad_norm": 3.1370120564698594,
+      "learning_rate": 2.0896208458663673e-06,
+      "loss": 0.292,
+      "step": 10180
+    },
+    {
+      "epoch": 0.7069157061519233,
+      "grad_norm": 2.9838560398437695,
+      "learning_rate": 2.0887065600389822e-06,
+      "loss": 0.1911,
+      "step": 10181
+    },
+    {
+      "epoch": 0.7069851409526454,
+      "grad_norm": 3.6585182101325877,
+      "learning_rate": 2.087792421461524e-06,
+      "loss": 0.4162,
+      "step": 10182
+    },
+    {
+      "epoch": 0.7070545757533676,
+      "grad_norm": 4.229777348484727,
+      "learning_rate": 2.086878430180231e-06,
+      "loss": 0.465,
+      "step": 10183
+    },
+    {
+      "epoch": 0.7071240105540897,
+      "grad_norm": 2.8780421479827756,
+      "learning_rate": 2.0859645862413295e-06,
+      "loss": 0.2349,
+      "step": 10184
+    },
+    {
+      "epoch": 0.7071934453548119,
+      "grad_norm": 4.204842581902981,
+      "learning_rate": 2.085050889691042e-06,
+      "loss": 0.57,
+      "step": 10185
+    },
+    {
+      "epoch": 0.707262880155534,
+      "grad_norm": 3.7791702911539065,
+      "learning_rate": 2.084137340575582e-06,
+      "loss": 0.3876,
+      "step": 10186
+    },
+    {
+      "epoch": 0.707332314956256,
+      "grad_norm": 3.9936188099122996,
+      "learning_rate": 2.083223938941156e-06,
+      "loss": 0.5581,
+      "step": 10187
+    },
+    {
+      "epoch": 0.7074017497569782,
+      "grad_norm": 5.230191131615469,
+      "learning_rate": 2.082310684833961e-06,
+      "loss": 0.6194,
+      "step": 10188
+    },
+    {
+      "epoch": 0.7074711845577003,
+      "grad_norm": 3.497735854878422,
+      "learning_rate": 2.081397578300193e-06,
+      "loss": 0.3726,
+      "step": 10189
+    },
+    {
+      "epoch": 0.7075406193584225,
+      "grad_norm": 4.258537619725338,
+      "learning_rate": 2.080484619386032e-06,
+      "loss": 0.5504,
+      "step": 10190
+    },
+    {
+      "epoch": 0.7076100541591446,
+      "grad_norm": 3.8274090734964106,
+      "learning_rate": 2.0795718081376528e-06,
+      "loss": 0.3236,
+      "step": 10191
+    },
+    {
+      "epoch": 0.7076794889598667,
+      "grad_norm": 4.107262496634714,
+      "learning_rate": 2.0786591446012288e-06,
+      "loss": 0.4319,
+      "step": 10192
+    },
+    {
+      "epoch": 0.7077489237605888,
+      "grad_norm": 3.757180025022487,
+      "learning_rate": 2.077746628822921e-06,
+      "loss": 0.5161,
+      "step": 10193
+    },
+    {
+      "epoch": 0.7078183585613109,
+      "grad_norm": 3.579391599389343,
+      "learning_rate": 2.076834260848879e-06,
+      "loss": 0.3334,
+      "step": 10194
+    },
+    {
+      "epoch": 0.707887793362033,
+      "grad_norm": 3.5046715833804116,
+      "learning_rate": 2.0759220407252535e-06,
+      "loss": 0.3576,
+      "step": 10195
+    },
+    {
+      "epoch": 0.7079572281627552,
+      "grad_norm": 4.061333519979376,
+      "learning_rate": 2.0750099684981843e-06,
+      "loss": 0.4029,
+      "step": 10196
+    },
+    {
+      "epoch": 0.7080266629634773,
+      "grad_norm": 3.5300832731125382,
+      "learning_rate": 2.0740980442137978e-06,
+      "loss": 0.3067,
+      "step": 10197
+    },
+    {
+      "epoch": 0.7080960977641995,
+      "grad_norm": 3.7904861626658466,
+      "learning_rate": 2.073186267918223e-06,
+      "loss": 0.4404,
+      "step": 10198
+    },
+    {
+      "epoch": 0.7081655325649215,
+      "grad_norm": 4.418968872586018,
+      "learning_rate": 2.072274639657574e-06,
+      "loss": 0.5312,
+      "step": 10199
+    },
+    {
+      "epoch": 0.7082349673656436,
+      "grad_norm": 4.347363669681173,
+      "learning_rate": 2.071363159477962e-06,
+      "loss": 0.6069,
+      "step": 10200
+    },
+    {
+      "epoch": 0.7083044021663658,
+      "grad_norm": 3.165630552170293,
+      "learning_rate": 2.070451827425487e-06,
+      "loss": 0.2555,
+      "step": 10201
+    },
+    {
+      "epoch": 0.7083738369670879,
+      "grad_norm": 2.301867843485286,
+      "learning_rate": 2.0695406435462435e-06,
+      "loss": 0.1002,
+      "step": 10202
+    },
+    {
+      "epoch": 0.7084432717678101,
+      "grad_norm": 3.4974477091403586,
+      "learning_rate": 2.0686296078863166e-06,
+      "loss": 0.344,
+      "step": 10203
+    },
+    {
+      "epoch": 0.7085127065685322,
+      "grad_norm": 3.366741202929514,
+      "learning_rate": 2.06771872049179e-06,
+      "loss": 0.3095,
+      "step": 10204
+    },
+    {
+      "epoch": 0.7085821413692542,
+      "grad_norm": 2.750444408087107,
+      "learning_rate": 2.066807981408732e-06,
+      "loss": 0.3316,
+      "step": 10205
+    },
+    {
+      "epoch": 0.7086515761699764,
+      "grad_norm": 5.343489581647093,
+      "learning_rate": 2.065897390683204e-06,
+      "loss": 0.4287,
+      "step": 10206
+    },
+    {
+      "epoch": 0.7087210109706985,
+      "grad_norm": 3.0973831663186746,
+      "learning_rate": 2.06498694836127e-06,
+      "loss": 0.3348,
+      "step": 10207
+    },
+    {
+      "epoch": 0.7087904457714206,
+      "grad_norm": 4.2003750317444615,
+      "learning_rate": 2.0640766544889727e-06,
+      "loss": 0.4096,
+      "step": 10208
+    },
+    {
+      "epoch": 0.7088598805721428,
+      "grad_norm": 4.422291907239041,
+      "learning_rate": 2.0631665091123546e-06,
+      "loss": 0.5262,
+      "step": 10209
+    },
+    {
+      "epoch": 0.7089293153728649,
+      "grad_norm": 3.9135457466310735,
+      "learning_rate": 2.0622565122774536e-06,
+      "loss": 0.4723,
+      "step": 10210
+    },
+    {
+      "epoch": 0.708998750173587,
+      "grad_norm": 2.94048536337275,
+      "learning_rate": 2.0613466640302932e-06,
+      "loss": 0.2456,
+      "step": 10211
+    },
+    {
+      "epoch": 0.7090681849743091,
+      "grad_norm": 3.966559899025331,
+      "learning_rate": 2.060436964416893e-06,
+      "loss": 0.4796,
+      "step": 10212
+    },
+    {
+      "epoch": 0.7091376197750312,
+      "grad_norm": 6.766207460268734,
+      "learning_rate": 2.0595274134832655e-06,
+      "loss": 0.4539,
+      "step": 10213
+    },
+    {
+      "epoch": 0.7092070545757534,
+      "grad_norm": 3.176146008308433,
+      "learning_rate": 2.0586180112754135e-06,
+      "loss": 0.3057,
+      "step": 10214
+    },
+    {
+      "epoch": 0.7092764893764755,
+      "grad_norm": 3.9709084696850527,
+      "learning_rate": 2.057708757839334e-06,
+      "loss": 0.3702,
+      "step": 10215
+    },
+    {
+      "epoch": 0.7093459241771977,
+      "grad_norm": 2.9293878659088657,
+      "learning_rate": 2.0567996532210166e-06,
+      "loss": 0.1611,
+      "step": 10216
+    },
+    {
+      "epoch": 0.7094153589779197,
+      "grad_norm": 4.053415852493665,
+      "learning_rate": 2.055890697466442e-06,
+      "loss": 0.3693,
+      "step": 10217
+    },
+    {
+      "epoch": 0.7094847937786418,
+      "grad_norm": 4.362724943874189,
+      "learning_rate": 2.0549818906215845e-06,
+      "loss": 0.502,
+      "step": 10218
+    },
+    {
+      "epoch": 0.709554228579364,
+      "grad_norm": 3.3652144744456907,
+      "learning_rate": 2.05407323273241e-06,
+      "loss": 0.3656,
+      "step": 10219
+    },
+    {
+      "epoch": 0.7096236633800861,
+      "grad_norm": 3.6154586595779827,
+      "learning_rate": 2.053164723844877e-06,
+      "loss": 0.3367,
+      "step": 10220
+    },
+    {
+      "epoch": 0.7096930981808082,
+      "grad_norm": 4.465145511071633,
+      "learning_rate": 2.052256364004941e-06,
+      "loss": 0.5165,
+      "step": 10221
+    },
+    {
+      "epoch": 0.7097625329815304,
+      "grad_norm": 3.7311401559263917,
+      "learning_rate": 2.05134815325854e-06,
+      "loss": 0.383,
+      "step": 10222
+    },
+    {
+      "epoch": 0.7098319677822524,
+      "grad_norm": 2.908552013973653,
+      "learning_rate": 2.0504400916516116e-06,
+      "loss": 0.1719,
+      "step": 10223
+    },
+    {
+      "epoch": 0.7099014025829746,
+      "grad_norm": 3.781793566243704,
+      "learning_rate": 2.0495321792300876e-06,
+      "loss": 0.499,
+      "step": 10224
+    },
+    {
+      "epoch": 0.7099708373836967,
+      "grad_norm": 4.477355037442522,
+      "learning_rate": 2.0486244160398887e-06,
+      "loss": 0.4837,
+      "step": 10225
+    },
+    {
+      "epoch": 0.7100402721844188,
+      "grad_norm": 3.968750202975932,
+      "learning_rate": 2.0477168021269237e-06,
+      "loss": 0.576,
+      "step": 10226
+    },
+    {
+      "epoch": 0.710109706985141,
+      "grad_norm": 3.237730071763434,
+      "learning_rate": 2.046809337537104e-06,
+      "loss": 0.2904,
+      "step": 10227
+    },
+    {
+      "epoch": 0.710179141785863,
+      "grad_norm": 3.5355339282178093,
+      "learning_rate": 2.0459020223163263e-06,
+      "loss": 0.4037,
+      "step": 10228
+    },
+    {
+      "epoch": 0.7102485765865852,
+      "grad_norm": 2.88035120259962,
+      "learning_rate": 2.0449948565104816e-06,
+      "loss": 0.2002,
+      "step": 10229
+    },
+    {
+      "epoch": 0.7103180113873073,
+      "grad_norm": 2.915186761549717,
+      "learning_rate": 2.044087840165453e-06,
+      "loss": 0.2778,
+      "step": 10230
+    },
+    {
+      "epoch": 0.7103874461880294,
+      "grad_norm": 3.522298557445293,
+      "learning_rate": 2.043180973327116e-06,
+      "loss": 0.4344,
+      "step": 10231
+    },
+    {
+      "epoch": 0.7104568809887516,
+      "grad_norm": 3.8339276925061228,
+      "learning_rate": 2.0422742560413404e-06,
+      "loss": 0.4759,
+      "step": 10232
+    },
+    {
+      "epoch": 0.7105263157894737,
+      "grad_norm": 3.7996320393161365,
+      "learning_rate": 2.041367688353986e-06,
+      "loss": 0.2358,
+      "step": 10233
+    },
+    {
+      "epoch": 0.7105957505901958,
+      "grad_norm": 3.8626462939037314,
+      "learning_rate": 2.0404612703109057e-06,
+      "loss": 0.3665,
+      "step": 10234
+    },
+    {
+      "epoch": 0.7106651853909179,
+      "grad_norm": 4.274676271973296,
+      "learning_rate": 2.0395550019579435e-06,
+      "loss": 0.4061,
+      "step": 10235
+    },
+    {
+      "epoch": 0.71073462019164,
+      "grad_norm": 5.221887225666951,
+      "learning_rate": 2.0386488833409433e-06,
+      "loss": 0.4785,
+      "step": 10236
+    },
+    {
+      "epoch": 0.7108040549923622,
+      "grad_norm": 3.5509909492170117,
+      "learning_rate": 2.03774291450573e-06,
+      "loss": 0.4291,
+      "step": 10237
+    },
+    {
+      "epoch": 0.7108734897930843,
+      "grad_norm": 4.032676964383473,
+      "learning_rate": 2.0368370954981254e-06,
+      "loss": 0.4853,
+      "step": 10238
+    },
+    {
+      "epoch": 0.7109429245938064,
+      "grad_norm": 2.7217984738598564,
+      "learning_rate": 2.035931426363952e-06,
+      "loss": 0.1494,
+      "step": 10239
+    },
+    {
+      "epoch": 0.7110123593945286,
+      "grad_norm": 5.049383456701229,
+      "learning_rate": 2.0350259071490084e-06,
+      "loss": 0.5735,
+      "step": 10240
+    },
+    {
+      "epoch": 0.7110817941952506,
+      "grad_norm": 3.5940910615726005,
+      "learning_rate": 2.0341205378991018e-06,
+      "loss": 0.2347,
+      "step": 10241
+    },
+    {
+      "epoch": 0.7111512289959728,
+      "grad_norm": 3.003081098279493,
+      "learning_rate": 2.033215318660022e-06,
+      "loss": 0.1955,
+      "step": 10242
+    },
+    {
+      "epoch": 0.7112206637966949,
+      "grad_norm": 3.1885132088039008,
+      "learning_rate": 2.0323102494775544e-06,
+      "loss": 0.3374,
+      "step": 10243
+    },
+    {
+      "epoch": 0.711290098597417,
+      "grad_norm": 4.1267214201917435,
+      "learning_rate": 2.0314053303974764e-06,
+      "loss": 0.3987,
+      "step": 10244
+    },
+    {
+      "epoch": 0.7113595333981392,
+      "grad_norm": 3.9929828175615714,
+      "learning_rate": 2.030500561465557e-06,
+      "loss": 0.4196,
+      "step": 10245
+    },
+    {
+      "epoch": 0.7114289681988613,
+      "grad_norm": 3.500043143998452,
+      "learning_rate": 2.0295959427275597e-06,
+      "loss": 0.2078,
+      "step": 10246
+    },
+    {
+      "epoch": 0.7114984029995834,
+      "grad_norm": 3.0619426436234227,
+      "learning_rate": 2.028691474229238e-06,
+      "loss": 0.2922,
+      "step": 10247
+    },
+    {
+      "epoch": 0.7115678378003055,
+      "grad_norm": 4.459717625918739,
+      "learning_rate": 2.027787156016339e-06,
+      "loss": 0.6904,
+      "step": 10248
+    },
+    {
+      "epoch": 0.7116372726010276,
+      "grad_norm": 6.091099985782494,
+      "learning_rate": 2.026882988134601e-06,
+      "loss": 0.7108,
+      "step": 10249
+    },
+    {
+      "epoch": 0.7117067074017498,
+      "grad_norm": 3.8221776818251136,
+      "learning_rate": 2.0259789706297604e-06,
+      "loss": 0.4027,
+      "step": 10250
+    },
+    {
+      "epoch": 0.7117761422024719,
+      "grad_norm": 4.1754689254674275,
+      "learning_rate": 2.025075103547537e-06,
+      "loss": 0.4369,
+      "step": 10251
+    },
+    {
+      "epoch": 0.711845577003194,
+      "grad_norm": 4.243266467977239,
+      "learning_rate": 2.024171386933646e-06,
+      "loss": 0.6651,
+      "step": 10252
+    },
+    {
+      "epoch": 0.7119150118039161,
+      "grad_norm": 3.808983578454301,
+      "learning_rate": 2.023267820833801e-06,
+      "loss": 0.5844,
+      "step": 10253
+    },
+    {
+      "epoch": 0.7119844466046382,
+      "grad_norm": 3.3413043601578205,
+      "learning_rate": 2.022364405293703e-06,
+      "loss": 0.3171,
+      "step": 10254
+    },
+    {
+      "epoch": 0.7120538814053604,
+      "grad_norm": 4.35088920397091,
+      "learning_rate": 2.02146114035904e-06,
+      "loss": 0.5152,
+      "step": 10255
+    },
+    {
+      "epoch": 0.7121233162060825,
+      "grad_norm": 4.460693919941862,
+      "learning_rate": 2.0205580260755043e-06,
+      "loss": 0.4889,
+      "step": 10256
+    },
+    {
+      "epoch": 0.7121927510068046,
+      "grad_norm": 4.279693120785385,
+      "learning_rate": 2.019655062488774e-06,
+      "loss": 0.072,
+      "step": 10257
+    },
+    {
+      "epoch": 0.7122621858075268,
+      "grad_norm": 3.5454998742295807,
+      "learning_rate": 2.0187522496445145e-06,
+      "loss": 0.2851,
+      "step": 10258
+    },
+    {
+      "epoch": 0.7123316206082488,
+      "grad_norm": 3.469199511681457,
+      "learning_rate": 2.017849587588394e-06,
+      "loss": 0.3623,
+      "step": 10259
+    },
+    {
+      "epoch": 0.712401055408971,
+      "grad_norm": 3.1131116609492233,
+      "learning_rate": 2.0169470763660675e-06,
+      "loss": 0.4296,
+      "step": 10260
+    },
+    {
+      "epoch": 0.7124704902096931,
+      "grad_norm": 3.8794660348925687,
+      "learning_rate": 2.0160447160231823e-06,
+      "loss": 0.3428,
+      "step": 10261
+    },
+    {
+      "epoch": 0.7125399250104152,
+      "grad_norm": 4.4240191962832345,
+      "learning_rate": 2.015142506605378e-06,
+      "loss": 0.3967,
+      "step": 10262
+    },
+    {
+      "epoch": 0.7126093598111374,
+      "grad_norm": 4.742019580380956,
+      "learning_rate": 2.0142404481582888e-06,
+      "loss": 0.5509,
+      "step": 10263
+    },
+    {
+      "epoch": 0.7126787946118595,
+      "grad_norm": 3.5427807075817763,
+      "learning_rate": 2.013338540727539e-06,
+      "loss": 0.3121,
+      "step": 10264
+    },
+    {
+      "epoch": 0.7127482294125815,
+      "grad_norm": 4.857330896660007,
+      "learning_rate": 2.012436784358746e-06,
+      "loss": 0.5318,
+      "step": 10265
+    },
+    {
+      "epoch": 0.7128176642133037,
+      "grad_norm": 4.410608768044099,
+      "learning_rate": 2.01153517909752e-06,
+      "loss": 0.5933,
+      "step": 10266
+    },
+    {
+      "epoch": 0.7128870990140258,
+      "grad_norm": 4.734544478533617,
+      "learning_rate": 2.0106337249894613e-06,
+      "loss": 0.5929,
+      "step": 10267
+    },
+    {
+      "epoch": 0.712956533814748,
+      "grad_norm": 4.245343027103758,
+      "learning_rate": 2.0097324220801696e-06,
+      "loss": 0.5292,
+      "step": 10268
+    },
+    {
+      "epoch": 0.7130259686154701,
+      "grad_norm": 4.633856383149074,
+      "learning_rate": 2.0088312704152247e-06,
+      "loss": 0.6787,
+      "step": 10269
+    },
+    {
+      "epoch": 0.7130954034161922,
+      "grad_norm": 6.2002036660909,
+      "learning_rate": 2.007930270040211e-06,
+      "loss": 0.5811,
+      "step": 10270
+    },
+    {
+      "epoch": 0.7131648382169143,
+      "grad_norm": 3.1013742741004777,
+      "learning_rate": 2.0070294210007004e-06,
+      "loss": 0.3194,
+      "step": 10271
+    },
+    {
+      "epoch": 0.7132342730176364,
+      "grad_norm": 5.202356837403619,
+      "learning_rate": 2.006128723342251e-06,
+      "loss": 0.423,
+      "step": 10272
+    },
+    {
+      "epoch": 0.7133037078183586,
+      "grad_norm": 3.9610527922688403,
+      "learning_rate": 2.0052281771104255e-06,
+      "loss": 0.4671,
+      "step": 10273
+    },
+    {
+      "epoch": 0.7133731426190807,
+      "grad_norm": 3.6238533350804554,
+      "learning_rate": 2.0043277823507697e-06,
+      "loss": 0.3581,
+      "step": 10274
+    },
+    {
+      "epoch": 0.7134425774198028,
+      "grad_norm": 3.9927396517254974,
+      "learning_rate": 2.0034275391088245e-06,
+      "loss": 0.4482,
+      "step": 10275
+    },
+    {
+      "epoch": 0.713512012220525,
+      "grad_norm": 3.9096826367983772,
+      "learning_rate": 2.0025274474301238e-06,
+      "loss": 0.4933,
+      "step": 10276
+    },
+    {
+      "epoch": 0.713581447021247,
+      "grad_norm": 4.267205255289849,
+      "learning_rate": 2.0016275073601927e-06,
+      "loss": 0.4924,
+      "step": 10277
+    },
+    {
+      "epoch": 0.7136508818219691,
+      "grad_norm": 5.846700274558715,
+      "learning_rate": 2.0007277189445478e-06,
+      "loss": 0.7268,
+      "step": 10278
+    },
+    {
+      "epoch": 0.7137203166226913,
+      "grad_norm": 3.197860546306613,
+      "learning_rate": 1.9998280822287043e-06,
+      "loss": 0.2401,
+      "step": 10279
+    },
+    {
+      "epoch": 0.7137897514234134,
+      "grad_norm": 4.130063578033967,
+      "learning_rate": 1.9989285972581595e-06,
+      "loss": 0.2663,
+      "step": 10280
+    },
+    {
+      "epoch": 0.7138591862241356,
+      "grad_norm": 4.175152336158564,
+      "learning_rate": 1.9980292640784095e-06,
+      "loss": 0.3002,
+      "step": 10281
+    },
+    {
+      "epoch": 0.7139286210248577,
+      "grad_norm": 3.388710457670331,
+      "learning_rate": 1.997130082734945e-06,
+      "loss": 0.2875,
+      "step": 10282
+    },
+    {
+      "epoch": 0.7139980558255797,
+      "grad_norm": 3.5778496133832505,
+      "learning_rate": 1.9962310532732413e-06,
+      "loss": 0.5172,
+      "step": 10283
+    },
+    {
+      "epoch": 0.7140674906263019,
+      "grad_norm": 4.323732719302313,
+      "learning_rate": 1.9953321757387704e-06,
+      "loss": 0.3105,
+      "step": 10284
+    },
+    {
+      "epoch": 0.714136925427024,
+      "grad_norm": 4.052305678389807,
+      "learning_rate": 1.994433450177e-06,
+      "loss": 0.3866,
+      "step": 10285
+    },
+    {
+      "epoch": 0.7142063602277462,
+      "grad_norm": 3.749262124733676,
+      "learning_rate": 1.993534876633386e-06,
+      "loss": 0.3014,
+      "step": 10286
+    },
+    {
+      "epoch": 0.7142757950284683,
+      "grad_norm": 2.7221677046336588,
+      "learning_rate": 1.9926364551533727e-06,
+      "loss": 0.3221,
+      "step": 10287
+    },
+    {
+      "epoch": 0.7143452298291904,
+      "grad_norm": 4.280799263476717,
+      "learning_rate": 1.991738185782406e-06,
+      "loss": 0.5113,
+      "step": 10288
+    },
+    {
+      "epoch": 0.7144146646299125,
+      "grad_norm": 4.022547668120578,
+      "learning_rate": 1.99084006856592e-06,
+      "loss": 0.4923,
+      "step": 10289
+    },
+    {
+      "epoch": 0.7144840994306346,
+      "grad_norm": 3.617828295975291,
+      "learning_rate": 1.9899421035493336e-06,
+      "loss": 0.453,
+      "step": 10290
+    },
+    {
+      "epoch": 0.7145535342313567,
+      "grad_norm": 4.2778688453422,
+      "learning_rate": 1.989044290778072e-06,
+      "loss": 0.5161,
+      "step": 10291
+    },
+    {
+      "epoch": 0.7146229690320789,
+      "grad_norm": 3.9448701862028983,
+      "learning_rate": 1.9881466302975432e-06,
+      "loss": 0.5158,
+      "step": 10292
+    },
+    {
+      "epoch": 0.714692403832801,
+      "grad_norm": 3.664423732695156,
+      "learning_rate": 1.98724912215315e-06,
+      "loss": 0.4505,
+      "step": 10293
+    },
+    {
+      "epoch": 0.7147618386335232,
+      "grad_norm": 3.2678375703983082,
+      "learning_rate": 1.986351766390286e-06,
+      "loss": 0.2866,
+      "step": 10294
+    },
+    {
+      "epoch": 0.7148312734342452,
+      "grad_norm": 4.158709177797214,
+      "learning_rate": 1.9854545630543405e-06,
+      "loss": 0.4678,
+      "step": 10295
+    },
+    {
+      "epoch": 0.7149007082349673,
+      "grad_norm": 4.429980432296422,
+      "learning_rate": 1.9845575121906913e-06,
+      "loss": 0.5944,
+      "step": 10296
+    },
+    {
+      "epoch": 0.7149701430356895,
+      "grad_norm": 4.349010131972708,
+      "learning_rate": 1.983660613844711e-06,
+      "loss": 0.7745,
+      "step": 10297
+    },
+    {
+      "epoch": 0.7150395778364116,
+      "grad_norm": 3.756604965833278,
+      "learning_rate": 1.982763868061764e-06,
+      "loss": 0.4949,
+      "step": 10298
+    },
+    {
+      "epoch": 0.7151090126371338,
+      "grad_norm": 2.8291109422806344,
+      "learning_rate": 1.981867274887204e-06,
+      "loss": 0.2355,
+      "step": 10299
+    },
+    {
+      "epoch": 0.7151784474378559,
+      "grad_norm": 4.247133914553282,
+      "learning_rate": 1.9809708343663863e-06,
+      "loss": 0.4082,
+      "step": 10300
+    },
+    {
+      "epoch": 0.7152478822385779,
+      "grad_norm": 3.9475217159480938,
+      "learning_rate": 1.9800745465446436e-06,
+      "loss": 0.5585,
+      "step": 10301
+    },
+    {
+      "epoch": 0.7153173170393001,
+      "grad_norm": 4.4970577170303505,
+      "learning_rate": 1.9791784114673146e-06,
+      "loss": 0.4855,
+      "step": 10302
+    },
+    {
+      "epoch": 0.7153867518400222,
+      "grad_norm": 3.2251235796382947,
+      "learning_rate": 1.9782824291797237e-06,
+      "loss": 0.3114,
+      "step": 10303
+    },
+    {
+      "epoch": 0.7154561866407444,
+      "grad_norm": 3.733810189294256,
+      "learning_rate": 1.977386599727188e-06,
+      "loss": 0.4752,
+      "step": 10304
+    },
+    {
+      "epoch": 0.7155256214414665,
+      "grad_norm": 4.505326185519829,
+      "learning_rate": 1.976490923155018e-06,
+      "loss": 0.375,
+      "step": 10305
+    },
+    {
+      "epoch": 0.7155950562421886,
+      "grad_norm": 3.2617868126369585,
+      "learning_rate": 1.9755953995085155e-06,
+      "loss": 0.4362,
+      "step": 10306
+    },
+    {
+      "epoch": 0.7156644910429107,
+      "grad_norm": 4.765167638524704,
+      "learning_rate": 1.9747000288329748e-06,
+      "loss": 0.4187,
+      "step": 10307
+    },
+    {
+      "epoch": 0.7157339258436328,
+      "grad_norm": 3.6545828698252905,
+      "learning_rate": 1.973804811173684e-06,
+      "loss": 0.3378,
+      "step": 10308
+    },
+    {
+      "epoch": 0.7158033606443549,
+      "grad_norm": 4.310249983091472,
+      "learning_rate": 1.972909746575921e-06,
+      "loss": 0.6332,
+      "step": 10309
+    },
+    {
+      "epoch": 0.7158727954450771,
+      "grad_norm": 2.0469125333809837,
+      "learning_rate": 1.9720148350849566e-06,
+      "loss": 0.1441,
+      "step": 10310
+    },
+    {
+      "epoch": 0.7159422302457992,
+      "grad_norm": 4.451796016636683,
+      "learning_rate": 1.9711200767460586e-06,
+      "loss": 0.5256,
+      "step": 10311
+    },
+    {
+      "epoch": 0.7160116650465214,
+      "grad_norm": 4.006588309250673,
+      "learning_rate": 1.970225471604478e-06,
+      "loss": 0.2637,
+      "step": 10312
+    },
+    {
+      "epoch": 0.7160810998472434,
+      "grad_norm": 3.349476034569757,
+      "learning_rate": 1.9693310197054627e-06,
+      "loss": 0.3325,
+      "step": 10313
+    },
+    {
+      "epoch": 0.7161505346479655,
+      "grad_norm": 3.8461471350715266,
+      "learning_rate": 1.9684367210942584e-06,
+      "loss": 0.4986,
+      "step": 10314
+    },
+    {
+      "epoch": 0.7162199694486877,
+      "grad_norm": 3.8739108903679154,
+      "learning_rate": 1.9675425758160927e-06,
+      "loss": 0.4683,
+      "step": 10315
+    },
+    {
+      "epoch": 0.7162894042494098,
+      "grad_norm": 3.6367773515914577,
+      "learning_rate": 1.9666485839161903e-06,
+      "loss": 0.4329,
+      "step": 10316
+    },
+    {
+      "epoch": 0.716358839050132,
+      "grad_norm": 4.378257841755187,
+      "learning_rate": 1.965754745439772e-06,
+      "loss": 0.5452,
+      "step": 10317
+    },
+    {
+      "epoch": 0.7164282738508541,
+      "grad_norm": 3.884110485774903,
+      "learning_rate": 1.9648610604320467e-06,
+      "loss": 0.3565,
+      "step": 10318
+    },
+    {
+      "epoch": 0.7164977086515761,
+      "grad_norm": 4.623689031783223,
+      "learning_rate": 1.9639675289382115e-06,
+      "loss": 0.6918,
+      "step": 10319
+    },
+    {
+      "epoch": 0.7165671434522983,
+      "grad_norm": 4.225074253888768,
+      "learning_rate": 1.9630741510034646e-06,
+      "loss": 0.5476,
+      "step": 10320
+    },
+    {
+      "epoch": 0.7166365782530204,
+      "grad_norm": 3.2039931692167536,
+      "learning_rate": 1.96218092667299e-06,
+      "loss": 0.2619,
+      "step": 10321
+    },
+    {
+      "epoch": 0.7167060130537425,
+      "grad_norm": 3.4436096467232367,
+      "learning_rate": 1.9612878559919673e-06,
+      "loss": 0.3397,
+      "step": 10322
+    },
+    {
+      "epoch": 0.7167754478544647,
+      "grad_norm": 3.818075751303728,
+      "learning_rate": 1.960394939005566e-06,
+      "loss": 0.3356,
+      "step": 10323
+    },
+    {
+      "epoch": 0.7168448826551868,
+      "grad_norm": 3.4247548880612206,
+      "learning_rate": 1.9595021757589496e-06,
+      "loss": 0.3518,
+      "step": 10324
+    },
+    {
+      "epoch": 0.716914317455909,
+      "grad_norm": 3.642101640670237,
+      "learning_rate": 1.958609566297272e-06,
+      "loss": 0.4255,
+      "step": 10325
+    },
+    {
+      "epoch": 0.716983752256631,
+      "grad_norm": 4.011159203124636,
+      "learning_rate": 1.957717110665681e-06,
+      "loss": 0.4696,
+      "step": 10326
+    },
+    {
+      "epoch": 0.7170531870573531,
+      "grad_norm": 3.9967840929984484,
+      "learning_rate": 1.956824808909317e-06,
+      "loss": 0.4702,
+      "step": 10327
+    },
+    {
+      "epoch": 0.7171226218580753,
+      "grad_norm": 3.3649957770874948,
+      "learning_rate": 1.9559326610733077e-06,
+      "loss": 0.4584,
+      "step": 10328
+    },
+    {
+      "epoch": 0.7171920566587974,
+      "grad_norm": 3.103648931140334,
+      "learning_rate": 1.9550406672027837e-06,
+      "loss": 0.2119,
+      "step": 10329
+    },
+    {
+      "epoch": 0.7172614914595196,
+      "grad_norm": 5.567371371952293,
+      "learning_rate": 1.9541488273428534e-06,
+      "loss": 0.6729,
+      "step": 10330
+    },
+    {
+      "epoch": 0.7173309262602416,
+      "grad_norm": 2.8639494269657684,
+      "learning_rate": 1.953257141538631e-06,
+      "loss": 0.2318,
+      "step": 10331
+    },
+    {
+      "epoch": 0.7174003610609637,
+      "grad_norm": 3.193118514037793,
+      "learning_rate": 1.9523656098352162e-06,
+      "loss": 0.4453,
+      "step": 10332
+    },
+    {
+      "epoch": 0.7174697958616859,
+      "grad_norm": 4.278928955471596,
+      "learning_rate": 1.9514742322776968e-06,
+      "loss": 0.5416,
+      "step": 10333
+    },
+    {
+      "epoch": 0.717539230662408,
+      "grad_norm": 3.017751094277211,
+      "learning_rate": 1.950583008911163e-06,
+      "loss": 0.2469,
+      "step": 10334
+    },
+    {
+      "epoch": 0.7176086654631301,
+      "grad_norm": 4.8235734492926055,
+      "learning_rate": 1.9496919397806898e-06,
+      "loss": 0.5053,
+      "step": 10335
+    },
+    {
+      "epoch": 0.7176781002638523,
+      "grad_norm": 4.492921336757589,
+      "learning_rate": 1.948801024931346e-06,
+      "loss": 0.4533,
+      "step": 10336
+    },
+    {
+      "epoch": 0.7177475350645743,
+      "grad_norm": 6.33195854033049,
+      "learning_rate": 1.947910264408195e-06,
+      "loss": 0.4194,
+      "step": 10337
+    },
+    {
+      "epoch": 0.7178169698652965,
+      "grad_norm": 4.139448565051755,
+      "learning_rate": 1.9470196582562877e-06,
+      "loss": 0.5734,
+      "step": 10338
+    },
+    {
+      "epoch": 0.7178864046660186,
+      "grad_norm": 3.6245010704987792,
+      "learning_rate": 1.9461292065206726e-06,
+      "loss": 0.4707,
+      "step": 10339
+    },
+    {
+      "epoch": 0.7179558394667407,
+      "grad_norm": 4.135308336798016,
+      "learning_rate": 1.945238909246386e-06,
+      "loss": 0.5861,
+      "step": 10340
+    },
+    {
+      "epoch": 0.7180252742674629,
+      "grad_norm": 3.8152999821739098,
+      "learning_rate": 1.9443487664784587e-06,
+      "loss": 0.5539,
+      "step": 10341
+    },
+    {
+      "epoch": 0.718094709068185,
+      "grad_norm": 2.3691593332426297,
+      "learning_rate": 1.9434587782619113e-06,
+      "loss": 0.2617,
+      "step": 10342
+    },
+    {
+      "epoch": 0.7181641438689071,
+      "grad_norm": 3.291795561724959,
+      "learning_rate": 1.942568944641764e-06,
+      "loss": 0.3892,
+      "step": 10343
+    },
+    {
+      "epoch": 0.7182335786696292,
+      "grad_norm": 4.047865186084089,
+      "learning_rate": 1.941679265663018e-06,
+      "loss": 0.6025,
+      "step": 10344
+    },
+    {
+      "epoch": 0.7183030134703513,
+      "grad_norm": 5.744633300685295,
+      "learning_rate": 1.940789741370672e-06,
+      "loss": 0.4259,
+      "step": 10345
+    },
+    {
+      "epoch": 0.7183724482710735,
+      "grad_norm": 5.9308705829550625,
+      "learning_rate": 1.9399003718097224e-06,
+      "loss": 0.5873,
+      "step": 10346
+    },
+    {
+      "epoch": 0.7184418830717956,
+      "grad_norm": 4.377756597756969,
+      "learning_rate": 1.939011157025148e-06,
+      "loss": 0.5717,
+      "step": 10347
+    },
+    {
+      "epoch": 0.7185113178725177,
+      "grad_norm": 3.4066513824482523,
+      "learning_rate": 1.9381220970619237e-06,
+      "loss": 0.4097,
+      "step": 10348
+    },
+    {
+      "epoch": 0.7185807526732398,
+      "grad_norm": 3.1109932005999164,
+      "learning_rate": 1.9372331919650205e-06,
+      "loss": 0.3406,
+      "step": 10349
+    },
+    {
+      "epoch": 0.7186501874739619,
+      "grad_norm": 3.060398903516486,
+      "learning_rate": 1.936344441779397e-06,
+      "loss": 0.2693,
+      "step": 10350
+    },
+    {
+      "epoch": 0.7187196222746841,
+      "grad_norm": 3.465404605835533,
+      "learning_rate": 1.9354558465500044e-06,
+      "loss": 0.3913,
+      "step": 10351
+    },
+    {
+      "epoch": 0.7187890570754062,
+      "grad_norm": 4.1049677910661035,
+      "learning_rate": 1.934567406321788e-06,
+      "loss": 0.6235,
+      "step": 10352
+    },
+    {
+      "epoch": 0.7188584918761283,
+      "grad_norm": 4.881389486145153,
+      "learning_rate": 1.933679121139683e-06,
+      "loss": 0.5979,
+      "step": 10353
+    },
+    {
+      "epoch": 0.7189279266768505,
+      "grad_norm": 3.648992695445106,
+      "learning_rate": 1.932790991048619e-06,
+      "loss": 0.3266,
+      "step": 10354
+    },
+    {
+      "epoch": 0.7189973614775725,
+      "grad_norm": 4.326416716160098,
+      "learning_rate": 1.931903016093515e-06,
+      "loss": 0.7641,
+      "step": 10355
+    },
+    {
+      "epoch": 0.7190667962782947,
+      "grad_norm": 3.8862646799721725,
+      "learning_rate": 1.9310151963192847e-06,
+      "loss": 0.3119,
+      "step": 10356
+    },
+    {
+      "epoch": 0.7191362310790168,
+      "grad_norm": 4.497789632752804,
+      "learning_rate": 1.930127531770833e-06,
+      "loss": 0.6797,
+      "step": 10357
+    },
+    {
+      "epoch": 0.7192056658797389,
+      "grad_norm": 3.6669789494683913,
+      "learning_rate": 1.929240022493057e-06,
+      "loss": 0.3102,
+      "step": 10358
+    },
+    {
+      "epoch": 0.7192751006804611,
+      "grad_norm": 3.1917938954653975,
+      "learning_rate": 1.928352668530844e-06,
+      "loss": 0.3264,
+      "step": 10359
+    },
+    {
+      "epoch": 0.7193445354811832,
+      "grad_norm": 3.254455571459327,
+      "learning_rate": 1.9274654699290797e-06,
+      "loss": 0.2302,
+      "step": 10360
+    },
+    {
+      "epoch": 0.7194139702819053,
+      "grad_norm": 3.0650925235826634,
+      "learning_rate": 1.926578426732636e-06,
+      "loss": 0.4025,
+      "step": 10361
+    },
+    {
+      "epoch": 0.7194834050826274,
+      "grad_norm": 4.61136272395699,
+      "learning_rate": 1.9256915389863746e-06,
+      "loss": 0.3383,
+      "step": 10362
+    },
+    {
+      "epoch": 0.7195528398833495,
+      "grad_norm": 4.196656349664857,
+      "learning_rate": 1.924804806735158e-06,
+      "loss": 0.4692,
+      "step": 10363
+    },
+    {
+      "epoch": 0.7196222746840717,
+      "grad_norm": 3.782286021574904,
+      "learning_rate": 1.923918230023836e-06,
+      "loss": 0.4036,
+      "step": 10364
+    },
+    {
+      "epoch": 0.7196917094847938,
+      "grad_norm": 3.3212379188802954,
+      "learning_rate": 1.923031808897246e-06,
+      "loss": 0.3843,
+      "step": 10365
+    },
+    {
+      "epoch": 0.7197611442855159,
+      "grad_norm": 4.845230799994959,
+      "learning_rate": 1.9221455434002272e-06,
+      "loss": 0.6732,
+      "step": 10366
+    },
+    {
+      "epoch": 0.719830579086238,
+      "grad_norm": 3.9720150571105086,
+      "learning_rate": 1.9212594335776035e-06,
+      "loss": 0.4554,
+      "step": 10367
+    },
+    {
+      "epoch": 0.7199000138869601,
+      "grad_norm": 4.163420696442243,
+      "learning_rate": 1.920373479474194e-06,
+      "loss": 0.5003,
+      "step": 10368
+    },
+    {
+      "epoch": 0.7199694486876823,
+      "grad_norm": 3.0662536630299932,
+      "learning_rate": 1.9194876811348088e-06,
+      "loss": 0.2985,
+      "step": 10369
+    },
+    {
+      "epoch": 0.7200388834884044,
+      "grad_norm": 2.5763261011504937,
+      "learning_rate": 1.9186020386042514e-06,
+      "loss": 0.218,
+      "step": 10370
+    },
+    {
+      "epoch": 0.7201083182891265,
+      "grad_norm": 5.625630249552127,
+      "learning_rate": 1.9177165519273154e-06,
+      "loss": 0.2906,
+      "step": 10371
+    },
+    {
+      "epoch": 0.7201777530898487,
+      "grad_norm": 3.706872863994817,
+      "learning_rate": 1.916831221148789e-06,
+      "loss": 0.4384,
+      "step": 10372
+    },
+    {
+      "epoch": 0.7202471878905707,
+      "grad_norm": 3.526486731367472,
+      "learning_rate": 1.9159460463134506e-06,
+      "loss": 0.3581,
+      "step": 10373
+    },
+    {
+      "epoch": 0.7203166226912929,
+      "grad_norm": 3.634394140501347,
+      "learning_rate": 1.915061027466069e-06,
+      "loss": 0.3842,
+      "step": 10374
+    },
+    {
+      "epoch": 0.720386057492015,
+      "grad_norm": 10.021057348217356,
+      "learning_rate": 1.9141761646514134e-06,
+      "loss": 0.3748,
+      "step": 10375
+    },
+    {
+      "epoch": 0.7204554922927371,
+      "grad_norm": 3.8866344344343107,
+      "learning_rate": 1.913291457914234e-06,
+      "loss": 0.4212,
+      "step": 10376
+    },
+    {
+      "epoch": 0.7205249270934593,
+      "grad_norm": 3.410931107051103,
+      "learning_rate": 1.9124069072992778e-06,
+      "loss": 0.3757,
+      "step": 10377
+    },
+    {
+      "epoch": 0.7205943618941814,
+      "grad_norm": 3.0257560921732836,
+      "learning_rate": 1.9115225128512903e-06,
+      "loss": 0.2282,
+      "step": 10378
+    },
+    {
+      "epoch": 0.7206637966949034,
+      "grad_norm": 3.7541917046800606,
+      "learning_rate": 1.9106382746149977e-06,
+      "loss": 0.3513,
+      "step": 10379
+    },
+    {
+      "epoch": 0.7207332314956256,
+      "grad_norm": 4.058968343059793,
+      "learning_rate": 1.9097541926351233e-06,
+      "loss": 0.4742,
+      "step": 10380
+    },
+    {
+      "epoch": 0.7208026662963477,
+      "grad_norm": 4.309107107285307,
+      "learning_rate": 1.908870266956386e-06,
+      "loss": 0.4641,
+      "step": 10381
+    },
+    {
+      "epoch": 0.7208721010970699,
+      "grad_norm": 5.021108926161804,
+      "learning_rate": 1.9079864976234936e-06,
+      "loss": 0.5119,
+      "step": 10382
+    },
+    {
+      "epoch": 0.720941535897792,
+      "grad_norm": 3.5055533002914663,
+      "learning_rate": 1.9071028846811451e-06,
+      "loss": 0.4275,
+      "step": 10383
+    },
+    {
+      "epoch": 0.721010970698514,
+      "grad_norm": 3.394838762949089,
+      "learning_rate": 1.9062194281740327e-06,
+      "loss": 0.2748,
+      "step": 10384
+    },
+    {
+      "epoch": 0.7210804054992362,
+      "grad_norm": 4.801379524664417,
+      "learning_rate": 1.9053361281468408e-06,
+      "loss": 0.4321,
+      "step": 10385
+    },
+    {
+      "epoch": 0.7211498402999583,
+      "grad_norm": 4.529608501261774,
+      "learning_rate": 1.9044529846442461e-06,
+      "loss": 0.6396,
+      "step": 10386
+    },
+    {
+      "epoch": 0.7212192751006805,
+      "grad_norm": 4.639755963630356,
+      "learning_rate": 1.9035699977109162e-06,
+      "loss": 0.2645,
+      "step": 10387
+    },
+    {
+      "epoch": 0.7212887099014026,
+      "grad_norm": 3.382497242739017,
+      "learning_rate": 1.9026871673915121e-06,
+      "loss": 0.5789,
+      "step": 10388
+    },
+    {
+      "epoch": 0.7213581447021247,
+      "grad_norm": 4.285060359994898,
+      "learning_rate": 1.9018044937306862e-06,
+      "loss": 0.4603,
+      "step": 10389
+    },
+    {
+      "epoch": 0.7214275795028469,
+      "grad_norm": 4.722807281501047,
+      "learning_rate": 1.900921976773083e-06,
+      "loss": 0.4153,
+      "step": 10390
+    },
+    {
+      "epoch": 0.7214970143035689,
+      "grad_norm": 3.7248109974500054,
+      "learning_rate": 1.9000396165633378e-06,
+      "loss": 0.3958,
+      "step": 10391
+    },
+    {
+      "epoch": 0.721566449104291,
+      "grad_norm": 4.576365644423474,
+      "learning_rate": 1.8991574131460826e-06,
+      "loss": 0.6004,
+      "step": 10392
+    },
+    {
+      "epoch": 0.7216358839050132,
+      "grad_norm": 4.98528050561496,
+      "learning_rate": 1.8982753665659382e-06,
+      "loss": 0.4566,
+      "step": 10393
+    },
+    {
+      "epoch": 0.7217053187057353,
+      "grad_norm": 4.037760117883098,
+      "learning_rate": 1.8973934768675128e-06,
+      "loss": 0.5679,
+      "step": 10394
+    },
+    {
+      "epoch": 0.7217747535064575,
+      "grad_norm": 2.8735158838748167,
+      "learning_rate": 1.8965117440954162e-06,
+      "loss": 0.2303,
+      "step": 10395
+    },
+    {
+      "epoch": 0.7218441883071796,
+      "grad_norm": 3.443778259938719,
+      "learning_rate": 1.8956301682942451e-06,
+      "loss": 0.1924,
+      "step": 10396
+    },
+    {
+      "epoch": 0.7219136231079016,
+      "grad_norm": 3.8886432294215147,
+      "learning_rate": 1.8947487495085842e-06,
+      "loss": 0.2988,
+      "step": 10397
+    },
+    {
+      "epoch": 0.7219830579086238,
+      "grad_norm": 3.5679895726899478,
+      "learning_rate": 1.8938674877830194e-06,
+      "loss": 0.471,
+      "step": 10398
+    },
+    {
+      "epoch": 0.7220524927093459,
+      "grad_norm": 2.844415399820664,
+      "learning_rate": 1.892986383162122e-06,
+      "loss": 0.3274,
+      "step": 10399
+    },
+    {
+      "epoch": 0.7221219275100681,
+      "grad_norm": 3.424525664420587,
+      "learning_rate": 1.8921054356904572e-06,
+      "loss": 0.4557,
+      "step": 10400
+    },
+    {
+      "epoch": 0.7221913623107902,
+      "grad_norm": 3.3451587368890903,
+      "learning_rate": 1.8912246454125831e-06,
+      "loss": 0.3493,
+      "step": 10401
+    },
+    {
+      "epoch": 0.7222607971115123,
+      "grad_norm": 4.131787150672724,
+      "learning_rate": 1.8903440123730487e-06,
+      "loss": 0.524,
+      "step": 10402
+    },
+    {
+      "epoch": 0.7223302319122344,
+      "grad_norm": 5.135661372963556,
+      "learning_rate": 1.8894635366163956e-06,
+      "loss": 0.6289,
+      "step": 10403
+    },
+    {
+      "epoch": 0.7223996667129565,
+      "grad_norm": 3.8328552848654214,
+      "learning_rate": 1.8885832181871567e-06,
+      "loss": 0.5213,
+      "step": 10404
+    },
+    {
+      "epoch": 0.7224691015136786,
+      "grad_norm": 5.912377009610363,
+      "learning_rate": 1.8877030571298577e-06,
+      "loss": 0.6889,
+      "step": 10405
+    },
+    {
+      "epoch": 0.7225385363144008,
+      "grad_norm": 4.595121205975614,
+      "learning_rate": 1.886823053489014e-06,
+      "loss": 0.4055,
+      "step": 10406
+    },
+    {
+      "epoch": 0.7226079711151229,
+      "grad_norm": 4.577352965878118,
+      "learning_rate": 1.8859432073091422e-06,
+      "loss": 0.65,
+      "step": 10407
+    },
+    {
+      "epoch": 0.7226774059158451,
+      "grad_norm": 4.45659078189218,
+      "learning_rate": 1.8850635186347366e-06,
+      "loss": 0.5359,
+      "step": 10408
+    },
+    {
+      "epoch": 0.7227468407165671,
+      "grad_norm": 4.353325853863383,
+      "learning_rate": 1.8841839875102918e-06,
+      "loss": 0.5194,
+      "step": 10409
+    },
+    {
+      "epoch": 0.7228162755172892,
+      "grad_norm": 5.458297866441244,
+      "learning_rate": 1.8833046139802968e-06,
+      "loss": 0.265,
+      "step": 10410
+    },
+    {
+      "epoch": 0.7228857103180114,
+      "grad_norm": 4.633364774875235,
+      "learning_rate": 1.8824253980892281e-06,
+      "loss": 0.585,
+      "step": 10411
+    },
+    {
+      "epoch": 0.7229551451187335,
+      "grad_norm": 3.3341461652373843,
+      "learning_rate": 1.8815463398815548e-06,
+      "loss": 0.3202,
+      "step": 10412
+    },
+    {
+      "epoch": 0.7230245799194557,
+      "grad_norm": 5.014911007763133,
+      "learning_rate": 1.8806674394017382e-06,
+      "loss": 0.5599,
+      "step": 10413
+    },
+    {
+      "epoch": 0.7230940147201778,
+      "grad_norm": 3.9145199624704516,
+      "learning_rate": 1.8797886966942336e-06,
+      "loss": 0.3198,
+      "step": 10414
+    },
+    {
+      "epoch": 0.7231634495208998,
+      "grad_norm": 3.552068864398843,
+      "learning_rate": 1.8789101118034858e-06,
+      "loss": 0.3215,
+      "step": 10415
+    },
+    {
+      "epoch": 0.723232884321622,
+      "grad_norm": 3.0524214986926332,
+      "learning_rate": 1.8780316847739327e-06,
+      "loss": 0.2782,
+      "step": 10416
+    },
+    {
+      "epoch": 0.7233023191223441,
+      "grad_norm": 3.3194884128421127,
+      "learning_rate": 1.8771534156500037e-06,
+      "loss": 0.2133,
+      "step": 10417
+    },
+    {
+      "epoch": 0.7233717539230663,
+      "grad_norm": 6.643639390820683,
+      "learning_rate": 1.8762753044761218e-06,
+      "loss": 0.7841,
+      "step": 10418
+    },
+    {
+      "epoch": 0.7234411887237884,
+      "grad_norm": 3.491015914474742,
+      "learning_rate": 1.8753973512967005e-06,
+      "loss": 0.2749,
+      "step": 10419
+    },
+    {
+      "epoch": 0.7235106235245105,
+      "grad_norm": 3.5955664844133963,
+      "learning_rate": 1.874519556156143e-06,
+      "loss": 0.3446,
+      "step": 10420
+    },
+    {
+      "epoch": 0.7235800583252326,
+      "grad_norm": 2.9217095498533414,
+      "learning_rate": 1.8736419190988537e-06,
+      "loss": 0.1956,
+      "step": 10421
+    },
+    {
+      "epoch": 0.7236494931259547,
+      "grad_norm": 4.247071466552116,
+      "learning_rate": 1.8727644401692163e-06,
+      "loss": 0.4185,
+      "step": 10422
+    },
+    {
+      "epoch": 0.7237189279266768,
+      "grad_norm": 3.5017975927804788,
+      "learning_rate": 1.8718871194116133e-06,
+      "loss": 0.4481,
+      "step": 10423
+    },
+    {
+      "epoch": 0.723788362727399,
+      "grad_norm": 4.893756368947829,
+      "learning_rate": 1.8710099568704215e-06,
+      "loss": 0.4342,
+      "step": 10424
+    },
+    {
+      "epoch": 0.7238577975281211,
+      "grad_norm": 3.6936595140201933,
+      "learning_rate": 1.8701329525900076e-06,
+      "loss": 0.3599,
+      "step": 10425
+    },
+    {
+      "epoch": 0.7239272323288433,
+      "grad_norm": 3.8201770033232307,
+      "learning_rate": 1.8692561066147236e-06,
+      "loss": 0.2703,
+      "step": 10426
+    },
+    {
+      "epoch": 0.7239966671295653,
+      "grad_norm": 3.256825535518447,
+      "learning_rate": 1.8683794189889248e-06,
+      "loss": 0.3155,
+      "step": 10427
+    },
+    {
+      "epoch": 0.7240661019302874,
+      "grad_norm": 5.6849929945389786,
+      "learning_rate": 1.867502889756953e-06,
+      "loss": 0.6592,
+      "step": 10428
+    },
+    {
+      "epoch": 0.7241355367310096,
+      "grad_norm": 2.9809670425017054,
+      "learning_rate": 1.8666265189631367e-06,
+      "loss": 0.3314,
+      "step": 10429
+    },
+    {
+      "epoch": 0.7242049715317317,
+      "grad_norm": 4.212355767852974,
+      "learning_rate": 1.8657503066518072e-06,
+      "loss": 0.3658,
+      "step": 10430
+    },
+    {
+      "epoch": 0.7242744063324539,
+      "grad_norm": 4.323903712676294,
+      "learning_rate": 1.8648742528672808e-06,
+      "loss": 0.5472,
+      "step": 10431
+    },
+    {
+      "epoch": 0.724343841133176,
+      "grad_norm": 2.617350105935429,
+      "learning_rate": 1.8639983576538672e-06,
+      "loss": 0.1862,
+      "step": 10432
+    },
+    {
+      "epoch": 0.724413275933898,
+      "grad_norm": 3.248252473679322,
+      "learning_rate": 1.8631226210558678e-06,
+      "loss": 0.4001,
+      "step": 10433
+    },
+    {
+      "epoch": 0.7244827107346202,
+      "grad_norm": 4.128433096137738,
+      "learning_rate": 1.8622470431175764e-06,
+      "loss": 0.4915,
+      "step": 10434
+    },
+    {
+      "epoch": 0.7245521455353423,
+      "grad_norm": 3.863701648651562,
+      "learning_rate": 1.8613716238832774e-06,
+      "loss": 0.2347,
+      "step": 10435
+    },
+    {
+      "epoch": 0.7246215803360644,
+      "grad_norm": 4.1399887874890515,
+      "learning_rate": 1.8604963633972539e-06,
+      "loss": 0.5786,
+      "step": 10436
+    },
+    {
+      "epoch": 0.7246910151367866,
+      "grad_norm": 3.455954382110435,
+      "learning_rate": 1.8596212617037695e-06,
+      "loss": 0.3837,
+      "step": 10437
+    },
+    {
+      "epoch": 0.7247604499375087,
+      "grad_norm": 3.4697099484394953,
+      "learning_rate": 1.8587463188470867e-06,
+      "loss": 0.3358,
+      "step": 10438
+    },
+    {
+      "epoch": 0.7248298847382308,
+      "grad_norm": 3.762731027466558,
+      "learning_rate": 1.8578715348714638e-06,
+      "loss": 0.2935,
+      "step": 10439
+    },
+    {
+      "epoch": 0.7248993195389529,
+      "grad_norm": 3.925431339962249,
+      "learning_rate": 1.8569969098211399e-06,
+      "loss": 0.5109,
+      "step": 10440
+    },
+    {
+      "epoch": 0.724968754339675,
+      "grad_norm": 3.931060937458799,
+      "learning_rate": 1.8561224437403569e-06,
+      "loss": 0.3199,
+      "step": 10441
+    },
+    {
+      "epoch": 0.7250381891403972,
+      "grad_norm": 3.7276782860789353,
+      "learning_rate": 1.8552481366733437e-06,
+      "loss": 0.4382,
+      "step": 10442
+    },
+    {
+      "epoch": 0.7251076239411193,
+      "grad_norm": 3.851752227364194,
+      "learning_rate": 1.8543739886643214e-06,
+      "loss": 0.4471,
+      "step": 10443
+    },
+    {
+      "epoch": 0.7251770587418415,
+      "grad_norm": 4.473183690953418,
+      "learning_rate": 1.8534999997575026e-06,
+      "loss": 0.5164,
+      "step": 10444
+    },
+    {
+      "epoch": 0.7252464935425635,
+      "grad_norm": 2.947366210469233,
+      "learning_rate": 1.8526261699970937e-06,
+      "loss": 0.3323,
+      "step": 10445
+    },
+    {
+      "epoch": 0.7253159283432856,
+      "grad_norm": 4.842107933900169,
+      "learning_rate": 1.8517524994272917e-06,
+      "loss": 0.4656,
+      "step": 10446
+    },
+    {
+      "epoch": 0.7253853631440078,
+      "grad_norm": 5.162199496891453,
+      "learning_rate": 1.8508789880922857e-06,
+      "loss": 0.402,
+      "step": 10447
+    },
+    {
+      "epoch": 0.7254547979447299,
+      "grad_norm": 3.1272347272857486,
+      "learning_rate": 1.8500056360362568e-06,
+      "loss": 0.2671,
+      "step": 10448
+    },
+    {
+      "epoch": 0.725524232745452,
+      "grad_norm": 5.482311826227409,
+      "learning_rate": 1.8491324433033787e-06,
+      "loss": 0.6265,
+      "step": 10449
+    },
+    {
+      "epoch": 0.7255936675461742,
+      "grad_norm": 3.21324538880097,
+      "learning_rate": 1.8482594099378154e-06,
+      "loss": 0.3105,
+      "step": 10450
+    },
+    {
+      "epoch": 0.7256631023468962,
+      "grad_norm": 4.208348695972901,
+      "learning_rate": 1.8473865359837257e-06,
+      "loss": 0.3104,
+      "step": 10451
+    },
+    {
+      "epoch": 0.7257325371476184,
+      "grad_norm": 3.0272283436760996,
+      "learning_rate": 1.846513821485255e-06,
+      "loss": 0.256,
+      "step": 10452
+    },
+    {
+      "epoch": 0.7258019719483405,
+      "grad_norm": 5.75661854706887,
+      "learning_rate": 1.845641266486551e-06,
+      "loss": 0.7772,
+      "step": 10453
+    },
+    {
+      "epoch": 0.7258714067490626,
+      "grad_norm": 3.300814890445595,
+      "learning_rate": 1.8447688710317402e-06,
+      "loss": 0.3599,
+      "step": 10454
+    },
+    {
+      "epoch": 0.7259408415497848,
+      "grad_norm": 4.1604795104864,
+      "learning_rate": 1.8438966351649478e-06,
+      "loss": 0.4031,
+      "step": 10455
+    },
+    {
+      "epoch": 0.7260102763505069,
+      "grad_norm": 4.11808694476157,
+      "learning_rate": 1.8430245589302943e-06,
+      "loss": 0.457,
+      "step": 10456
+    },
+    {
+      "epoch": 0.726079711151229,
+      "grad_norm": 4.210490363365273,
+      "learning_rate": 1.8421526423718878e-06,
+      "loss": 0.4652,
+      "step": 10457
+    },
+    {
+      "epoch": 0.7261491459519511,
+      "grad_norm": 3.42724931718003,
+      "learning_rate": 1.8412808855338237e-06,
+      "loss": 0.3925,
+      "step": 10458
+    },
+    {
+      "epoch": 0.7262185807526732,
+      "grad_norm": 2.657240049410536,
+      "learning_rate": 1.8404092884601998e-06,
+      "loss": 0.2128,
+      "step": 10459
+    },
+    {
+      "epoch": 0.7262880155533954,
+      "grad_norm": 4.249968656747497,
+      "learning_rate": 1.839537851195099e-06,
+      "loss": 0.562,
+      "step": 10460
+    },
+    {
+      "epoch": 0.7263574503541175,
+      "grad_norm": 3.8245258436166836,
+      "learning_rate": 1.838666573782597e-06,
+      "loss": 0.4739,
+      "step": 10461
+    },
+    {
+      "epoch": 0.7264268851548396,
+      "grad_norm": 2.59958998655832,
+      "learning_rate": 1.8377954562667627e-06,
+      "loss": 0.3104,
+      "step": 10462
+    },
+    {
+      "epoch": 0.7264963199555617,
+      "grad_norm": 4.866218067793073,
+      "learning_rate": 1.8369244986916556e-06,
+      "loss": 0.3223,
+      "step": 10463
+    },
+    {
+      "epoch": 0.7265657547562838,
+      "grad_norm": 4.892854945361474,
+      "learning_rate": 1.8360537011013286e-06,
+      "loss": 0.4744,
+      "step": 10464
+    },
+    {
+      "epoch": 0.726635189557006,
+      "grad_norm": 5.476254607352019,
+      "learning_rate": 1.835183063539825e-06,
+      "loss": 0.5793,
+      "step": 10465
+    },
+    {
+      "epoch": 0.7267046243577281,
+      "grad_norm": 2.921998900734984,
+      "learning_rate": 1.8343125860511808e-06,
+      "loss": 0.2855,
+      "step": 10466
+    },
+    {
+      "epoch": 0.7267740591584502,
+      "grad_norm": 4.410731545322353,
+      "learning_rate": 1.8334422686794218e-06,
+      "loss": 0.4924,
+      "step": 10467
+    },
+    {
+      "epoch": 0.7268434939591724,
+      "grad_norm": 4.085849210258286,
+      "learning_rate": 1.8325721114685735e-06,
+      "loss": 0.359,
+      "step": 10468
+    },
+    {
+      "epoch": 0.7269129287598944,
+      "grad_norm": 3.5644106349719977,
+      "learning_rate": 1.8317021144626418e-06,
+      "loss": 0.3354,
+      "step": 10469
+    },
+    {
+      "epoch": 0.7269823635606166,
+      "grad_norm": 3.158365048687018,
+      "learning_rate": 1.8308322777056297e-06,
+      "loss": 0.4888,
+      "step": 10470
+    },
+    {
+      "epoch": 0.7270517983613387,
+      "grad_norm": 3.9303843434828445,
+      "learning_rate": 1.8299626012415389e-06,
+      "loss": 0.3609,
+      "step": 10471
+    },
+    {
+      "epoch": 0.7271212331620608,
+      "grad_norm": 6.599042062038926,
+      "learning_rate": 1.829093085114349e-06,
+      "loss": 0.7258,
+      "step": 10472
+    },
+    {
+      "epoch": 0.727190667962783,
+      "grad_norm": 4.184879342939031,
+      "learning_rate": 1.8282237293680444e-06,
+      "loss": 0.4301,
+      "step": 10473
+    },
+    {
+      "epoch": 0.7272601027635051,
+      "grad_norm": 4.451119699734514,
+      "learning_rate": 1.8273545340465948e-06,
+      "loss": 0.5293,
+      "step": 10474
+    },
+    {
+      "epoch": 0.7273295375642271,
+      "grad_norm": 4.398997749649687,
+      "learning_rate": 1.8264854991939623e-06,
+      "loss": 0.5675,
+      "step": 10475
+    },
+    {
+      "epoch": 0.7273989723649493,
+      "grad_norm": 4.448067733852272,
+      "learning_rate": 1.8256166248541024e-06,
+      "loss": 0.6692,
+      "step": 10476
+    },
+    {
+      "epoch": 0.7274684071656714,
+      "grad_norm": 3.708971740325353,
+      "learning_rate": 1.8247479110709616e-06,
+      "loss": 0.4476,
+      "step": 10477
+    },
+    {
+      "epoch": 0.7275378419663936,
+      "grad_norm": 2.4855686031326303,
+      "learning_rate": 1.8238793578884778e-06,
+      "loss": 0.1343,
+      "step": 10478
+    },
+    {
+      "epoch": 0.7276072767671157,
+      "grad_norm": 3.773123443437252,
+      "learning_rate": 1.823010965350583e-06,
+      "loss": 0.4621,
+      "step": 10479
+    },
+    {
+      "epoch": 0.7276767115678378,
+      "grad_norm": 3.887977735083403,
+      "learning_rate": 1.8221427335011983e-06,
+      "loss": 0.3585,
+      "step": 10480
+    },
+    {
+      "epoch": 0.7277461463685599,
+      "grad_norm": 1.9983578420190555,
+      "learning_rate": 1.821274662384236e-06,
+      "loss": 0.1022,
+      "step": 10481
+    },
+    {
+      "epoch": 0.727815581169282,
+      "grad_norm": 3.324723701368467,
+      "learning_rate": 1.8204067520436081e-06,
+      "loss": 0.2719,
+      "step": 10482
+    },
+    {
+      "epoch": 0.7278850159700042,
+      "grad_norm": 3.5660804570084084,
+      "learning_rate": 1.8195390025232069e-06,
+      "loss": 0.517,
+      "step": 10483
+    },
+    {
+      "epoch": 0.7279544507707263,
+      "grad_norm": 4.489318101460402,
+      "learning_rate": 1.818671413866922e-06,
+      "loss": 0.3725,
+      "step": 10484
+    },
+    {
+      "epoch": 0.7280238855714484,
+      "grad_norm": 3.906122543091545,
+      "learning_rate": 1.817803986118639e-06,
+      "loss": 0.3255,
+      "step": 10485
+    },
+    {
+      "epoch": 0.7280933203721706,
+      "grad_norm": 4.317719500880463,
+      "learning_rate": 1.8169367193222315e-06,
+      "loss": 0.4528,
+      "step": 10486
+    },
+    {
+      "epoch": 0.7281627551728926,
+      "grad_norm": 3.838412397300347,
+      "learning_rate": 1.8160696135215593e-06,
+      "loss": 0.4493,
+      "step": 10487
+    },
+    {
+      "epoch": 0.7282321899736148,
+      "grad_norm": 3.2759970803867797,
+      "learning_rate": 1.8152026687604845e-06,
+      "loss": 0.3043,
+      "step": 10488
+    },
+    {
+      "epoch": 0.7283016247743369,
+      "grad_norm": 3.4089725995355806,
+      "learning_rate": 1.8143358850828574e-06,
+      "loss": 0.37,
+      "step": 10489
+    },
+    {
+      "epoch": 0.728371059575059,
+      "grad_norm": 3.649035721723917,
+      "learning_rate": 1.8134692625325122e-06,
+      "loss": 0.4728,
+      "step": 10490
+    },
+    {
+      "epoch": 0.7284404943757812,
+      "grad_norm": 4.183406835687212,
+      "learning_rate": 1.812602801153288e-06,
+      "loss": 0.4875,
+      "step": 10491
+    },
+    {
+      "epoch": 0.7285099291765033,
+      "grad_norm": 3.1038897017837757,
+      "learning_rate": 1.811736500989007e-06,
+      "loss": 0.1829,
+      "step": 10492
+    },
+    {
+      "epoch": 0.7285793639772253,
+      "grad_norm": 3.907787400133084,
+      "learning_rate": 1.8108703620834867e-06,
+      "loss": 0.3847,
+      "step": 10493
+    },
+    {
+      "epoch": 0.7286487987779475,
+      "grad_norm": 5.073873055142729,
+      "learning_rate": 1.8100043844805343e-06,
+      "loss": 0.5125,
+      "step": 10494
+    },
+    {
+      "epoch": 0.7287182335786696,
+      "grad_norm": 4.105170910212948,
+      "learning_rate": 1.8091385682239503e-06,
+      "loss": 0.4146,
+      "step": 10495
+    },
+    {
+      "epoch": 0.7287876683793918,
+      "grad_norm": 3.3586930671701536,
+      "learning_rate": 1.8082729133575267e-06,
+      "loss": 0.2951,
+      "step": 10496
+    },
+    {
+      "epoch": 0.7288571031801139,
+      "grad_norm": 3.910943558899524,
+      "learning_rate": 1.807407419925048e-06,
+      "loss": 0.4517,
+      "step": 10497
+    },
+    {
+      "epoch": 0.728926537980836,
+      "grad_norm": 4.43391871594695,
+      "learning_rate": 1.8065420879702888e-06,
+      "loss": 0.483,
+      "step": 10498
+    },
+    {
+      "epoch": 0.7289959727815581,
+      "grad_norm": 3.722803566859759,
+      "learning_rate": 1.8056769175370159e-06,
+      "loss": 0.3976,
+      "step": 10499
+    },
+    {
+      "epoch": 0.7290654075822802,
+      "grad_norm": 3.622510456360244,
+      "learning_rate": 1.804811908668993e-06,
+      "loss": 0.3786,
+      "step": 10500
+    },
+    {
+      "epoch": 0.7291348423830024,
+      "grad_norm": 2.448157517500828,
+      "learning_rate": 1.8039470614099647e-06,
+      "loss": 0.2437,
+      "step": 10501
+    },
+    {
+      "epoch": 0.7292042771837245,
+      "grad_norm": 3.961519510373321,
+      "learning_rate": 1.8030823758036791e-06,
+      "loss": 0.4784,
+      "step": 10502
+    },
+    {
+      "epoch": 0.7292737119844466,
+      "grad_norm": 4.29226445555635,
+      "learning_rate": 1.8022178518938715e-06,
+      "loss": 0.5028,
+      "step": 10503
+    },
+    {
+      "epoch": 0.7293431467851688,
+      "grad_norm": 3.33734522452691,
+      "learning_rate": 1.8013534897242625e-06,
+      "loss": 0.2753,
+      "step": 10504
+    },
+    {
+      "epoch": 0.7294125815858908,
+      "grad_norm": 3.9729155806300933,
+      "learning_rate": 1.8004892893385766e-06,
+      "loss": 0.3048,
+      "step": 10505
+    },
+    {
+      "epoch": 0.7294820163866129,
+      "grad_norm": 3.7052280366624397,
+      "learning_rate": 1.7996252507805218e-06,
+      "loss": 0.4737,
+      "step": 10506
+    },
+    {
+      "epoch": 0.7295514511873351,
+      "grad_norm": 3.9964376681114735,
+      "learning_rate": 1.7987613740938003e-06,
+      "loss": 0.2655,
+      "step": 10507
+    },
+    {
+      "epoch": 0.7296208859880572,
+      "grad_norm": 4.267274680989353,
+      "learning_rate": 1.797897659322106e-06,
+      "loss": 0.4288,
+      "step": 10508
+    },
+    {
+      "epoch": 0.7296903207887794,
+      "grad_norm": 4.558896270599049,
+      "learning_rate": 1.7970341065091246e-06,
+      "loss": 0.536,
+      "step": 10509
+    },
+    {
+      "epoch": 0.7297597555895015,
+      "grad_norm": 3.6820911703646173,
+      "learning_rate": 1.7961707156985324e-06,
+      "loss": 0.4562,
+      "step": 10510
+    },
+    {
+      "epoch": 0.7298291903902235,
+      "grad_norm": 3.5131344427311095,
+      "learning_rate": 1.7953074869340032e-06,
+      "loss": 0.3073,
+      "step": 10511
+    },
+    {
+      "epoch": 0.7298986251909457,
+      "grad_norm": 3.690547985525075,
+      "learning_rate": 1.7944444202591933e-06,
+      "loss": 0.3631,
+      "step": 10512
+    },
+    {
+      "epoch": 0.7299680599916678,
+      "grad_norm": 3.719140078552416,
+      "learning_rate": 1.7935815157177554e-06,
+      "loss": 0.3906,
+      "step": 10513
+    },
+    {
+      "epoch": 0.73003749479239,
+      "grad_norm": 3.879398143880803,
+      "learning_rate": 1.7927187733533396e-06,
+      "loss": 0.3386,
+      "step": 10514
+    },
+    {
+      "epoch": 0.7301069295931121,
+      "grad_norm": 4.218767558784246,
+      "learning_rate": 1.7918561932095773e-06,
+      "loss": 0.5616,
+      "step": 10515
+    },
+    {
+      "epoch": 0.7301763643938342,
+      "grad_norm": 4.715215060930386,
+      "learning_rate": 1.7909937753300966e-06,
+      "loss": 0.7104,
+      "step": 10516
+    },
+    {
+      "epoch": 0.7302457991945563,
+      "grad_norm": 3.7780169990909824,
+      "learning_rate": 1.7901315197585217e-06,
+      "loss": 0.4348,
+      "step": 10517
+    },
+    {
+      "epoch": 0.7303152339952784,
+      "grad_norm": 4.042296457584013,
+      "learning_rate": 1.7892694265384635e-06,
+      "loss": 0.4143,
+      "step": 10518
+    },
+    {
+      "epoch": 0.7303846687960005,
+      "grad_norm": 4.264413124108612,
+      "learning_rate": 1.7884074957135216e-06,
+      "loss": 0.4555,
+      "step": 10519
+    },
+    {
+      "epoch": 0.7304541035967227,
+      "grad_norm": 2.469162485752799,
+      "learning_rate": 1.7875457273272956e-06,
+      "loss": 0.1771,
+      "step": 10520
+    },
+    {
+      "epoch": 0.7305235383974448,
+      "grad_norm": 4.167556826555442,
+      "learning_rate": 1.7866841214233721e-06,
+      "loss": 0.6714,
+      "step": 10521
+    },
+    {
+      "epoch": 0.730592973198167,
+      "grad_norm": 3.3310053007104377,
+      "learning_rate": 1.7858226780453292e-06,
+      "loss": 0.4217,
+      "step": 10522
+    },
+    {
+      "epoch": 0.730662407998889,
+      "grad_norm": 3.4434611942084885,
+      "learning_rate": 1.784961397236738e-06,
+      "loss": 0.2951,
+      "step": 10523
+    },
+    {
+      "epoch": 0.7307318427996111,
+      "grad_norm": 3.5352861901437485,
+      "learning_rate": 1.7841002790411616e-06,
+      "loss": 0.3887,
+      "step": 10524
+    },
+    {
+      "epoch": 0.7308012776003333,
+      "grad_norm": 3.4971151295247487,
+      "learning_rate": 1.7832393235021545e-06,
+      "loss": 0.2548,
+      "step": 10525
+    },
+    {
+      "epoch": 0.7308707124010554,
+      "grad_norm": 4.238930623100809,
+      "learning_rate": 1.7823785306632619e-06,
+      "loss": 0.5031,
+      "step": 10526
+    },
+    {
+      "epoch": 0.7309401472017776,
+      "grad_norm": 3.455251116520109,
+      "learning_rate": 1.7815179005680221e-06,
+      "loss": 0.3743,
+      "step": 10527
+    },
+    {
+      "epoch": 0.7310095820024997,
+      "grad_norm": 3.8518685882978354,
+      "learning_rate": 1.7806574332599652e-06,
+      "loss": 0.4775,
+      "step": 10528
+    },
+    {
+      "epoch": 0.7310790168032217,
+      "grad_norm": 3.783315387842042,
+      "learning_rate": 1.7797971287826121e-06,
+      "loss": 0.5054,
+      "step": 10529
+    },
+    {
+      "epoch": 0.7311484516039439,
+      "grad_norm": 3.231229144150515,
+      "learning_rate": 1.778936987179475e-06,
+      "loss": 0.3327,
+      "step": 10530
+    },
+    {
+      "epoch": 0.731217886404666,
+      "grad_norm": 3.5790989952365373,
+      "learning_rate": 1.7780770084940612e-06,
+      "loss": 0.3378,
+      "step": 10531
+    },
+    {
+      "epoch": 0.7312873212053881,
+      "grad_norm": 3.6676136803723147,
+      "learning_rate": 1.7772171927698689e-06,
+      "loss": 0.3853,
+      "step": 10532
+    },
+    {
+      "epoch": 0.7313567560061103,
+      "grad_norm": 4.443682839813009,
+      "learning_rate": 1.7763575400503806e-06,
+      "loss": 0.6299,
+      "step": 10533
+    },
+    {
+      "epoch": 0.7314261908068324,
+      "grad_norm": 3.1081066690934853,
+      "learning_rate": 1.7754980503790814e-06,
+      "loss": 0.3724,
+      "step": 10534
+    },
+    {
+      "epoch": 0.7314956256075545,
+      "grad_norm": 3.76067526479934,
+      "learning_rate": 1.7746387237994427e-06,
+      "loss": 0.3632,
+      "step": 10535
+    },
+    {
+      "epoch": 0.7315650604082766,
+      "grad_norm": 2.9821728953082407,
+      "learning_rate": 1.7737795603549274e-06,
+      "loss": 0.2617,
+      "step": 10536
+    },
+    {
+      "epoch": 0.7316344952089987,
+      "grad_norm": 3.9188570127809252,
+      "learning_rate": 1.772920560088991e-06,
+      "loss": 0.4416,
+      "step": 10537
+    },
+    {
+      "epoch": 0.7317039300097209,
+      "grad_norm": 2.9213538657891087,
+      "learning_rate": 1.772061723045081e-06,
+      "loss": 0.356,
+      "step": 10538
+    },
+    {
+      "epoch": 0.731773364810443,
+      "grad_norm": 3.441116726955416,
+      "learning_rate": 1.7712030492666365e-06,
+      "loss": 0.3216,
+      "step": 10539
+    },
+    {
+      "epoch": 0.7318427996111652,
+      "grad_norm": 4.085639653745112,
+      "learning_rate": 1.770344538797088e-06,
+      "loss": 0.624,
+      "step": 10540
+    },
+    {
+      "epoch": 0.7319122344118872,
+      "grad_norm": 2.529913742600644,
+      "learning_rate": 1.7694861916798584e-06,
+      "loss": 0.2951,
+      "step": 10541
+    },
+    {
+      "epoch": 0.7319816692126093,
+      "grad_norm": 2.4643820132652734,
+      "learning_rate": 1.7686280079583595e-06,
+      "loss": 0.1795,
+      "step": 10542
+    },
+    {
+      "epoch": 0.7320511040133315,
+      "grad_norm": 2.338269229588886,
+      "learning_rate": 1.7677699876760036e-06,
+      "loss": 0.211,
+      "step": 10543
+    },
+    {
+      "epoch": 0.7321205388140536,
+      "grad_norm": 3.579323118881092,
+      "learning_rate": 1.7669121308761816e-06,
+      "loss": 0.3066,
+      "step": 10544
+    },
+    {
+      "epoch": 0.7321899736147758,
+      "grad_norm": 4.077552183746781,
+      "learning_rate": 1.7660544376022842e-06,
+      "loss": 0.3834,
+      "step": 10545
+    },
+    {
+      "epoch": 0.7322594084154979,
+      "grad_norm": 3.9770409146871257,
+      "learning_rate": 1.7651969078976978e-06,
+      "loss": 0.4588,
+      "step": 10546
+    },
+    {
+      "epoch": 0.7323288432162199,
+      "grad_norm": 4.503823996316076,
+      "learning_rate": 1.764339541805789e-06,
+      "loss": 0.6985,
+      "step": 10547
+    },
+    {
+      "epoch": 0.7323982780169421,
+      "grad_norm": 3.6587060572075263,
+      "learning_rate": 1.7634823393699236e-06,
+      "loss": 0.3497,
+      "step": 10548
+    },
+    {
+      "epoch": 0.7324677128176642,
+      "grad_norm": 2.7901669987809767,
+      "learning_rate": 1.7626253006334603e-06,
+      "loss": 0.24,
+      "step": 10549
+    },
+    {
+      "epoch": 0.7325371476183863,
+      "grad_norm": 3.4708876180741597,
+      "learning_rate": 1.7617684256397476e-06,
+      "loss": 0.2834,
+      "step": 10550
+    },
+    {
+      "epoch": 0.7326065824191085,
+      "grad_norm": 3.138755476637452,
+      "learning_rate": 1.760911714432121e-06,
+      "loss": 0.2244,
+      "step": 10551
+    },
+    {
+      "epoch": 0.7326760172198306,
+      "grad_norm": 3.692187376985266,
+      "learning_rate": 1.7600551670539158e-06,
+      "loss": 0.243,
+      "step": 10552
+    },
+    {
+      "epoch": 0.7327454520205527,
+      "grad_norm": 3.152687812929138,
+      "learning_rate": 1.7591987835484536e-06,
+      "loss": 0.2951,
+      "step": 10553
+    },
+    {
+      "epoch": 0.7328148868212748,
+      "grad_norm": 3.7674976979512858,
+      "learning_rate": 1.7583425639590502e-06,
+      "loss": 0.3472,
+      "step": 10554
+    },
+    {
+      "epoch": 0.7328843216219969,
+      "grad_norm": 3.4068277812819523,
+      "learning_rate": 1.7574865083290116e-06,
+      "loss": 0.3432,
+      "step": 10555
+    },
+    {
+      "epoch": 0.7329537564227191,
+      "grad_norm": 4.309598547239222,
+      "learning_rate": 1.756630616701636e-06,
+      "loss": 0.3401,
+      "step": 10556
+    },
+    {
+      "epoch": 0.7330231912234412,
+      "grad_norm": 2.6643862516545753,
+      "learning_rate": 1.7557748891202142e-06,
+      "loss": 0.2417,
+      "step": 10557
+    },
+    {
+      "epoch": 0.7330926260241634,
+      "grad_norm": 3.777351395000426,
+      "learning_rate": 1.7549193256280273e-06,
+      "loss": 0.4583,
+      "step": 10558
+    },
+    {
+      "epoch": 0.7331620608248854,
+      "grad_norm": 4.107006602791873,
+      "learning_rate": 1.754063926268349e-06,
+      "loss": 0.6083,
+      "step": 10559
+    },
+    {
+      "epoch": 0.7332314956256075,
+      "grad_norm": 4.003110049455225,
+      "learning_rate": 1.7532086910844438e-06,
+      "loss": 0.5299,
+      "step": 10560
+    },
+    {
+      "epoch": 0.7333009304263297,
+      "grad_norm": 5.149417127303963,
+      "learning_rate": 1.752353620119569e-06,
+      "loss": 0.5884,
+      "step": 10561
+    },
+    {
+      "epoch": 0.7333703652270518,
+      "grad_norm": 3.7482200794838496,
+      "learning_rate": 1.751498713416972e-06,
+      "loss": 0.4493,
+      "step": 10562
+    },
+    {
+      "epoch": 0.7334398000277739,
+      "grad_norm": 3.3738013442176826,
+      "learning_rate": 1.7506439710198953e-06,
+      "loss": 0.5243,
+      "step": 10563
+    },
+    {
+      "epoch": 0.7335092348284961,
+      "grad_norm": 4.387644485450916,
+      "learning_rate": 1.7497893929715714e-06,
+      "loss": 0.481,
+      "step": 10564
+    },
+    {
+      "epoch": 0.7335786696292181,
+      "grad_norm": 4.491377061087058,
+      "learning_rate": 1.7489349793152193e-06,
+      "loss": 0.4243,
+      "step": 10565
+    },
+    {
+      "epoch": 0.7336481044299403,
+      "grad_norm": 3.3066742651551126,
+      "learning_rate": 1.7480807300940584e-06,
+      "loss": 0.3714,
+      "step": 10566
+    },
+    {
+      "epoch": 0.7337175392306624,
+      "grad_norm": 3.4646704190588937,
+      "learning_rate": 1.7472266453512942e-06,
+      "loss": 0.2724,
+      "step": 10567
+    },
+    {
+      "epoch": 0.7337869740313845,
+      "grad_norm": 3.7862121221384784,
+      "learning_rate": 1.746372725130126e-06,
+      "loss": 0.5116,
+      "step": 10568
+    },
+    {
+      "epoch": 0.7338564088321067,
+      "grad_norm": 3.930365913342115,
+      "learning_rate": 1.7455189694737434e-06,
+      "loss": 0.566,
+      "step": 10569
+    },
+    {
+      "epoch": 0.7339258436328288,
+      "grad_norm": 2.8121578799514646,
+      "learning_rate": 1.7446653784253287e-06,
+      "loss": 0.3351,
+      "step": 10570
+    },
+    {
+      "epoch": 0.733995278433551,
+      "grad_norm": 4.0249084310036185,
+      "learning_rate": 1.7438119520280549e-06,
+      "loss": 0.4863,
+      "step": 10571
+    },
+    {
+      "epoch": 0.734064713234273,
+      "grad_norm": 4.732715308458906,
+      "learning_rate": 1.7429586903250889e-06,
+      "loss": 0.81,
+      "step": 10572
+    },
+    {
+      "epoch": 0.7341341480349951,
+      "grad_norm": 4.6556108558713065,
+      "learning_rate": 1.742105593359586e-06,
+      "loss": 0.5392,
+      "step": 10573
+    },
+    {
+      "epoch": 0.7342035828357173,
+      "grad_norm": 3.7063553562532614,
+      "learning_rate": 1.7412526611746945e-06,
+      "loss": 0.4437,
+      "step": 10574
+    },
+    {
+      "epoch": 0.7342730176364394,
+      "grad_norm": 5.626877407794784,
+      "learning_rate": 1.7403998938135587e-06,
+      "loss": 0.5388,
+      "step": 10575
+    },
+    {
+      "epoch": 0.7343424524371615,
+      "grad_norm": 7.55607016230924,
+      "learning_rate": 1.7395472913193062e-06,
+      "loss": 0.7203,
+      "step": 10576
+    },
+    {
+      "epoch": 0.7344118872378836,
+      "grad_norm": 3.833093680957534,
+      "learning_rate": 1.7386948537350612e-06,
+      "loss": 0.3414,
+      "step": 10577
+    },
+    {
+      "epoch": 0.7344813220386057,
+      "grad_norm": 3.9882920370627164,
+      "learning_rate": 1.737842581103943e-06,
+      "loss": 0.5102,
+      "step": 10578
+    },
+    {
+      "epoch": 0.7345507568393279,
+      "grad_norm": 3.564528985960407,
+      "learning_rate": 1.736990473469054e-06,
+      "loss": 0.1944,
+      "step": 10579
+    },
+    {
+      "epoch": 0.73462019164005,
+      "grad_norm": 5.0481737893539815,
+      "learning_rate": 1.7361385308734923e-06,
+      "loss": 0.8724,
+      "step": 10580
+    },
+    {
+      "epoch": 0.7346896264407721,
+      "grad_norm": 4.112235154830738,
+      "learning_rate": 1.7352867533603529e-06,
+      "loss": 0.4618,
+      "step": 10581
+    },
+    {
+      "epoch": 0.7347590612414943,
+      "grad_norm": 5.00657933892249,
+      "learning_rate": 1.7344351409727151e-06,
+      "loss": 0.5871,
+      "step": 10582
+    },
+    {
+      "epoch": 0.7348284960422163,
+      "grad_norm": 3.9659554676904727,
+      "learning_rate": 1.733583693753652e-06,
+      "loss": 0.1648,
+      "step": 10583
+    },
+    {
+      "epoch": 0.7348979308429385,
+      "grad_norm": 4.45003692163835,
+      "learning_rate": 1.7327324117462302e-06,
+      "loss": 0.3574,
+      "step": 10584
+    },
+    {
+      "epoch": 0.7349673656436606,
+      "grad_norm": 3.062951870285048,
+      "learning_rate": 1.7318812949935055e-06,
+      "loss": 0.321,
+      "step": 10585
+    },
+    {
+      "epoch": 0.7350368004443827,
+      "grad_norm": 4.090718713734945,
+      "learning_rate": 1.7310303435385267e-06,
+      "loss": 0.3697,
+      "step": 10586
+    },
+    {
+      "epoch": 0.7351062352451049,
+      "grad_norm": 4.921462382847311,
+      "learning_rate": 1.7301795574243345e-06,
+      "loss": 0.6726,
+      "step": 10587
+    },
+    {
+      "epoch": 0.735175670045827,
+      "grad_norm": 3.594562587801793,
+      "learning_rate": 1.7293289366939597e-06,
+      "loss": 0.3201,
+      "step": 10588
+    },
+    {
+      "epoch": 0.735245104846549,
+      "grad_norm": 3.606940662180231,
+      "learning_rate": 1.7284784813904265e-06,
+      "loss": 0.4745,
+      "step": 10589
+    },
+    {
+      "epoch": 0.7353145396472712,
+      "grad_norm": 4.845083882722045,
+      "learning_rate": 1.72762819155675e-06,
+      "loss": 0.5281,
+      "step": 10590
+    },
+    {
+      "epoch": 0.7353839744479933,
+      "grad_norm": 4.010441406547519,
+      "learning_rate": 1.7267780672359348e-06,
+      "loss": 0.5498,
+      "step": 10591
+    },
+    {
+      "epoch": 0.7354534092487155,
+      "grad_norm": 4.414236996827545,
+      "learning_rate": 1.7259281084709823e-06,
+      "loss": 0.468,
+      "step": 10592
+    },
+    {
+      "epoch": 0.7355228440494376,
+      "grad_norm": 3.730984742364967,
+      "learning_rate": 1.7250783153048834e-06,
+      "loss": 0.4136,
+      "step": 10593
+    },
+    {
+      "epoch": 0.7355922788501597,
+      "grad_norm": 3.059830910129794,
+      "learning_rate": 1.7242286877806136e-06,
+      "loss": 0.3391,
+      "step": 10594
+    },
+    {
+      "epoch": 0.7356617136508818,
+      "grad_norm": 4.054309637032811,
+      "learning_rate": 1.7233792259411519e-06,
+      "loss": 0.4948,
+      "step": 10595
+    },
+    {
+      "epoch": 0.7357311484516039,
+      "grad_norm": 6.324899559916543,
+      "learning_rate": 1.7225299298294634e-06,
+      "loss": 0.7274,
+      "step": 10596
+    },
+    {
+      "epoch": 0.7358005832523261,
+      "grad_norm": 4.212299812710117,
+      "learning_rate": 1.7216807994884982e-06,
+      "loss": 0.591,
+      "step": 10597
+    },
+    {
+      "epoch": 0.7358700180530482,
+      "grad_norm": 3.689736656394049,
+      "learning_rate": 1.7208318349612108e-06,
+      "loss": 0.4432,
+      "step": 10598
+    },
+    {
+      "epoch": 0.7359394528537703,
+      "grad_norm": 4.616659342138726,
+      "learning_rate": 1.7199830362905384e-06,
+      "loss": 0.6354,
+      "step": 10599
+    },
+    {
+      "epoch": 0.7360088876544925,
+      "grad_norm": 5.02419656416883,
+      "learning_rate": 1.7191344035194118e-06,
+      "loss": 0.6663,
+      "step": 10600
+    },
+    {
+      "epoch": 0.7360783224552145,
+      "grad_norm": 2.912271065716018,
+      "learning_rate": 1.7182859366907555e-06,
+      "loss": 0.2638,
+      "step": 10601
+    },
+    {
+      "epoch": 0.7361477572559367,
+      "grad_norm": 3.8427746515479337,
+      "learning_rate": 1.7174376358474826e-06,
+      "loss": 0.4737,
+      "step": 10602
+    },
+    {
+      "epoch": 0.7362171920566588,
+      "grad_norm": 4.402436089286478,
+      "learning_rate": 1.7165895010324995e-06,
+      "loss": 0.4816,
+      "step": 10603
+    },
+    {
+      "epoch": 0.7362866268573809,
+      "grad_norm": 2.003931027552049,
+      "learning_rate": 1.7157415322887038e-06,
+      "loss": 0.0697,
+      "step": 10604
+    },
+    {
+      "epoch": 0.7363560616581031,
+      "grad_norm": 3.909509254591714,
+      "learning_rate": 1.7148937296589852e-06,
+      "loss": 0.4498,
+      "step": 10605
+    },
+    {
+      "epoch": 0.7364254964588252,
+      "grad_norm": 4.78734186696786,
+      "learning_rate": 1.714046093186223e-06,
+      "loss": 0.43,
+      "step": 10606
+    },
+    {
+      "epoch": 0.7364949312595472,
+      "grad_norm": 4.668730637176559,
+      "learning_rate": 1.713198622913294e-06,
+      "loss": 0.5842,
+      "step": 10607
+    },
+    {
+      "epoch": 0.7365643660602694,
+      "grad_norm": 4.842395930405256,
+      "learning_rate": 1.7123513188830582e-06,
+      "loss": 0.477,
+      "step": 10608
+    },
+    {
+      "epoch": 0.7366338008609915,
+      "grad_norm": 3.9693036500587016,
+      "learning_rate": 1.7115041811383703e-06,
+      "loss": 0.3928,
+      "step": 10609
+    },
+    {
+      "epoch": 0.7367032356617137,
+      "grad_norm": 3.699737742266566,
+      "learning_rate": 1.7106572097220836e-06,
+      "loss": 0.4692,
+      "step": 10610
+    },
+    {
+      "epoch": 0.7367726704624358,
+      "grad_norm": 3.571920398156959,
+      "learning_rate": 1.7098104046770298e-06,
+      "loss": 0.2492,
+      "step": 10611
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 5.110931579329591,
+      "learning_rate": 1.7089637660460434e-06,
+      "loss": 0.7028,
+      "step": 10612
+    },
+    {
+      "epoch": 0.73691154006388,
+      "grad_norm": 4.229279096110702,
+      "learning_rate": 1.7081172938719465e-06,
+      "loss": 0.474,
+      "step": 10613
+    },
+    {
+      "epoch": 0.7369809748646021,
+      "grad_norm": 2.954081228509903,
+      "learning_rate": 1.7072709881975513e-06,
+      "loss": 0.2551,
+      "step": 10614
+    },
+    {
+      "epoch": 0.7370504096653243,
+      "grad_norm": 4.263877901998473,
+      "learning_rate": 1.7064248490656642e-06,
+      "loss": 0.4503,
+      "step": 10615
+    },
+    {
+      "epoch": 0.7371198444660464,
+      "grad_norm": 3.40520953064715,
+      "learning_rate": 1.7055788765190807e-06,
+      "loss": 0.2357,
+      "step": 10616
+    },
+    {
+      "epoch": 0.7371892792667685,
+      "grad_norm": 2.539062774060217,
+      "learning_rate": 1.7047330706005898e-06,
+      "loss": 0.1736,
+      "step": 10617
+    },
+    {
+      "epoch": 0.7372587140674907,
+      "grad_norm": 4.692035551478609,
+      "learning_rate": 1.703887431352972e-06,
+      "loss": 0.6744,
+      "step": 10618
+    },
+    {
+      "epoch": 0.7373281488682127,
+      "grad_norm": 3.598012866552002,
+      "learning_rate": 1.7030419588189972e-06,
+      "loss": 0.4952,
+      "step": 10619
+    },
+    {
+      "epoch": 0.7373975836689348,
+      "grad_norm": 4.386760215409834,
+      "learning_rate": 1.7021966530414303e-06,
+      "loss": 0.6006,
+      "step": 10620
+    },
+    {
+      "epoch": 0.737467018469657,
+      "grad_norm": 3.994348005953582,
+      "learning_rate": 1.7013515140630243e-06,
+      "loss": 0.5086,
+      "step": 10621
+    },
+    {
+      "epoch": 0.7375364532703791,
+      "grad_norm": 3.5903462485368225,
+      "learning_rate": 1.700506541926526e-06,
+      "loss": 0.3866,
+      "step": 10622
+    },
+    {
+      "epoch": 0.7376058880711013,
+      "grad_norm": 3.9257774142925728,
+      "learning_rate": 1.699661736674672e-06,
+      "loss": 0.6985,
+      "step": 10623
+    },
+    {
+      "epoch": 0.7376753228718234,
+      "grad_norm": 2.846875914430184,
+      "learning_rate": 1.6988170983501938e-06,
+      "loss": 0.2356,
+      "step": 10624
+    },
+    {
+      "epoch": 0.7377447576725454,
+      "grad_norm": 3.788747775143197,
+      "learning_rate": 1.6979726269958135e-06,
+      "loss": 0.3149,
+      "step": 10625
+    },
+    {
+      "epoch": 0.7378141924732676,
+      "grad_norm": 4.332525506170356,
+      "learning_rate": 1.697128322654238e-06,
+      "loss": 0.5235,
+      "step": 10626
+    },
+    {
+      "epoch": 0.7378836272739897,
+      "grad_norm": 3.125292498000971,
+      "learning_rate": 1.6962841853681755e-06,
+      "loss": 0.177,
+      "step": 10627
+    },
+    {
+      "epoch": 0.7379530620747119,
+      "grad_norm": 3.6056274709317604,
+      "learning_rate": 1.6954402151803223e-06,
+      "loss": 0.436,
+      "step": 10628
+    },
+    {
+      "epoch": 0.738022496875434,
+      "grad_norm": 3.0866583293624474,
+      "learning_rate": 1.6945964121333608e-06,
+      "loss": 0.1496,
+      "step": 10629
+    },
+    {
+      "epoch": 0.738091931676156,
+      "grad_norm": 3.014486342078335,
+      "learning_rate": 1.6937527762699735e-06,
+      "loss": 0.3299,
+      "step": 10630
+    },
+    {
+      "epoch": 0.7381613664768782,
+      "grad_norm": 3.4669886901950244,
+      "learning_rate": 1.69290930763283e-06,
+      "loss": 0.358,
+      "step": 10631
+    },
+    {
+      "epoch": 0.7382308012776003,
+      "grad_norm": 4.584519170127436,
+      "learning_rate": 1.692066006264591e-06,
+      "loss": 0.3636,
+      "step": 10632
+    },
+    {
+      "epoch": 0.7383002360783224,
+      "grad_norm": 3.9200094881594505,
+      "learning_rate": 1.691222872207911e-06,
+      "loss": 0.5111,
+      "step": 10633
+    },
+    {
+      "epoch": 0.7383696708790446,
+      "grad_norm": 3.5949064119449985,
+      "learning_rate": 1.6903799055054332e-06,
+      "loss": 0.327,
+      "step": 10634
+    },
+    {
+      "epoch": 0.7384391056797667,
+      "grad_norm": 3.994188747945362,
+      "learning_rate": 1.689537106199795e-06,
+      "loss": 0.5309,
+      "step": 10635
+    },
+    {
+      "epoch": 0.7385085404804889,
+      "grad_norm": 3.9450100318848413,
+      "learning_rate": 1.6886944743336241e-06,
+      "loss": 0.2352,
+      "step": 10636
+    },
+    {
+      "epoch": 0.7385779752812109,
+      "grad_norm": 5.011949239872855,
+      "learning_rate": 1.6878520099495394e-06,
+      "loss": 0.8115,
+      "step": 10637
+    },
+    {
+      "epoch": 0.738647410081933,
+      "grad_norm": 2.59902254334525,
+      "learning_rate": 1.6870097130901509e-06,
+      "loss": 0.1226,
+      "step": 10638
+    },
+    {
+      "epoch": 0.7387168448826552,
+      "grad_norm": 3.779459647632509,
+      "learning_rate": 1.6861675837980652e-06,
+      "loss": 0.4686,
+      "step": 10639
+    },
+    {
+      "epoch": 0.7387862796833773,
+      "grad_norm": 5.194121947771429,
+      "learning_rate": 1.6853256221158715e-06,
+      "loss": 0.6672,
+      "step": 10640
+    },
+    {
+      "epoch": 0.7388557144840995,
+      "grad_norm": 3.1394744060762427,
+      "learning_rate": 1.6844838280861552e-06,
+      "loss": 0.2295,
+      "step": 10641
+    },
+    {
+      "epoch": 0.7389251492848216,
+      "grad_norm": 3.624280284024086,
+      "learning_rate": 1.6836422017514969e-06,
+      "loss": 0.2599,
+      "step": 10642
+    },
+    {
+      "epoch": 0.7389945840855436,
+      "grad_norm": 3.33501032330394,
+      "learning_rate": 1.6828007431544635e-06,
+      "loss": 0.337,
+      "step": 10643
+    },
+    {
+      "epoch": 0.7390640188862658,
+      "grad_norm": 4.587733642153555,
+      "learning_rate": 1.6819594523376149e-06,
+      "loss": 0.3987,
+      "step": 10644
+    },
+    {
+      "epoch": 0.7391334536869879,
+      "grad_norm": 4.318189897701996,
+      "learning_rate": 1.6811183293435023e-06,
+      "loss": 0.4891,
+      "step": 10645
+    },
+    {
+      "epoch": 0.73920288848771,
+      "grad_norm": 4.372005282893266,
+      "learning_rate": 1.6802773742146694e-06,
+      "loss": 0.3233,
+      "step": 10646
+    },
+    {
+      "epoch": 0.7392723232884322,
+      "grad_norm": 3.7103020237333224,
+      "learning_rate": 1.6794365869936502e-06,
+      "loss": 0.2844,
+      "step": 10647
+    },
+    {
+      "epoch": 0.7393417580891543,
+      "grad_norm": 3.87504184454104,
+      "learning_rate": 1.6785959677229708e-06,
+      "loss": 0.5376,
+      "step": 10648
+    },
+    {
+      "epoch": 0.7394111928898764,
+      "grad_norm": 5.361744027738653,
+      "learning_rate": 1.6777555164451493e-06,
+      "loss": 0.6037,
+      "step": 10649
+    },
+    {
+      "epoch": 0.7394806276905985,
+      "grad_norm": 3.958438418182807,
+      "learning_rate": 1.6769152332026938e-06,
+      "loss": 0.4888,
+      "step": 10650
+    },
+    {
+      "epoch": 0.7395500624913206,
+      "grad_norm": 3.95243140380165,
+      "learning_rate": 1.6760751180381062e-06,
+      "loss": 0.4886,
+      "step": 10651
+    },
+    {
+      "epoch": 0.7396194972920428,
+      "grad_norm": 2.878847248770449,
+      "learning_rate": 1.6752351709938758e-06,
+      "loss": 0.2575,
+      "step": 10652
+    },
+    {
+      "epoch": 0.7396889320927649,
+      "grad_norm": 4.183764920761019,
+      "learning_rate": 1.6743953921124917e-06,
+      "loss": 0.5084,
+      "step": 10653
+    },
+    {
+      "epoch": 0.7397583668934871,
+      "grad_norm": 4.275057056813458,
+      "learning_rate": 1.673555781436424e-06,
+      "loss": 0.6561,
+      "step": 10654
+    },
+    {
+      "epoch": 0.7398278016942091,
+      "grad_norm": 3.997526638019358,
+      "learning_rate": 1.6727163390081385e-06,
+      "loss": 0.5602,
+      "step": 10655
+    },
+    {
+      "epoch": 0.7398972364949312,
+      "grad_norm": 3.586168525124893,
+      "learning_rate": 1.6718770648700977e-06,
+      "loss": 0.3754,
+      "step": 10656
+    },
+    {
+      "epoch": 0.7399666712956534,
+      "grad_norm": 5.108306301142851,
+      "learning_rate": 1.6710379590647503e-06,
+      "loss": 0.6142,
+      "step": 10657
+    },
+    {
+      "epoch": 0.7400361060963755,
+      "grad_norm": 3.524018542146671,
+      "learning_rate": 1.6701990216345326e-06,
+      "loss": 0.3744,
+      "step": 10658
+    },
+    {
+      "epoch": 0.7401055408970977,
+      "grad_norm": 4.33551059011242,
+      "learning_rate": 1.6693602526218821e-06,
+      "loss": 0.4488,
+      "step": 10659
+    },
+    {
+      "epoch": 0.7401749756978198,
+      "grad_norm": 4.740040974930598,
+      "learning_rate": 1.6685216520692233e-06,
+      "loss": 0.4963,
+      "step": 10660
+    },
+    {
+      "epoch": 0.7402444104985418,
+      "grad_norm": 3.9379839510375927,
+      "learning_rate": 1.6676832200189658e-06,
+      "loss": 0.4384,
+      "step": 10661
+    },
+    {
+      "epoch": 0.740313845299264,
+      "grad_norm": 5.2101380785618385,
+      "learning_rate": 1.6668449565135219e-06,
+      "loss": 0.6737,
+      "step": 10662
+    },
+    {
+      "epoch": 0.7403832800999861,
+      "grad_norm": 3.105409845484737,
+      "learning_rate": 1.6660068615952884e-06,
+      "loss": 0.1342,
+      "step": 10663
+    },
+    {
+      "epoch": 0.7404527149007082,
+      "grad_norm": 5.632949540306084,
+      "learning_rate": 1.6651689353066552e-06,
+      "loss": 0.6824,
+      "step": 10664
+    },
+    {
+      "epoch": 0.7405221497014304,
+      "grad_norm": 4.067908020251523,
+      "learning_rate": 1.6643311776900034e-06,
+      "loss": 0.4641,
+      "step": 10665
+    },
+    {
+      "epoch": 0.7405915845021525,
+      "grad_norm": 3.4590098718425866,
+      "learning_rate": 1.6634935887877068e-06,
+      "loss": 0.3261,
+      "step": 10666
+    },
+    {
+      "epoch": 0.7406610193028746,
+      "grad_norm": 4.066589112270933,
+      "learning_rate": 1.6626561686421266e-06,
+      "loss": 0.2903,
+      "step": 10667
+    },
+    {
+      "epoch": 0.7407304541035967,
+      "grad_norm": 3.8238287116007115,
+      "learning_rate": 1.6618189172956245e-06,
+      "loss": 0.3464,
+      "step": 10668
+    },
+    {
+      "epoch": 0.7407998889043188,
+      "grad_norm": 3.4091300163487515,
+      "learning_rate": 1.6609818347905426e-06,
+      "loss": 0.5223,
+      "step": 10669
+    },
+    {
+      "epoch": 0.740869323705041,
+      "grad_norm": 3.4834387727021308,
+      "learning_rate": 1.6601449211692194e-06,
+      "loss": 0.2462,
+      "step": 10670
+    },
+    {
+      "epoch": 0.7409387585057631,
+      "grad_norm": 5.288438458344361,
+      "learning_rate": 1.6593081764739898e-06,
+      "loss": 0.5196,
+      "step": 10671
+    },
+    {
+      "epoch": 0.7410081933064853,
+      "grad_norm": 3.4207708687792984,
+      "learning_rate": 1.6584716007471696e-06,
+      "loss": 0.2983,
+      "step": 10672
+    },
+    {
+      "epoch": 0.7410776281072073,
+      "grad_norm": 4.385583102051702,
+      "learning_rate": 1.657635194031076e-06,
+      "loss": 0.5508,
+      "step": 10673
+    },
+    {
+      "epoch": 0.7411470629079294,
+      "grad_norm": 3.9426851805668544,
+      "learning_rate": 1.656798956368012e-06,
+      "loss": 0.707,
+      "step": 10674
+    },
+    {
+      "epoch": 0.7412164977086516,
+      "grad_norm": 4.5392998150025115,
+      "learning_rate": 1.6559628878002736e-06,
+      "loss": 0.5385,
+      "step": 10675
+    },
+    {
+      "epoch": 0.7412859325093737,
+      "grad_norm": 3.9916476892254815,
+      "learning_rate": 1.6551269883701486e-06,
+      "loss": 0.5663,
+      "step": 10676
+    },
+    {
+      "epoch": 0.7413553673100958,
+      "grad_norm": 3.6908051614373605,
+      "learning_rate": 1.6542912581199155e-06,
+      "loss": 0.2546,
+      "step": 10677
+    },
+    {
+      "epoch": 0.741424802110818,
+      "grad_norm": 3.849045138272675,
+      "learning_rate": 1.6534556970918447e-06,
+      "loss": 0.5313,
+      "step": 10678
+    },
+    {
+      "epoch": 0.74149423691154,
+      "grad_norm": 3.828181398266914,
+      "learning_rate": 1.652620305328198e-06,
+      "loss": 0.3945,
+      "step": 10679
+    },
+    {
+      "epoch": 0.7415636717122622,
+      "grad_norm": 4.298421912022778,
+      "learning_rate": 1.6517850828712283e-06,
+      "loss": 0.4906,
+      "step": 10680
+    },
+    {
+      "epoch": 0.7416331065129843,
+      "grad_norm": 3.916682994162593,
+      "learning_rate": 1.6509500297631786e-06,
+      "loss": 0.3398,
+      "step": 10681
+    },
+    {
+      "epoch": 0.7417025413137064,
+      "grad_norm": 4.504665215498936,
+      "learning_rate": 1.65011514604629e-06,
+      "loss": 0.4661,
+      "step": 10682
+    },
+    {
+      "epoch": 0.7417719761144286,
+      "grad_norm": 3.4210735869302065,
+      "learning_rate": 1.6492804317627852e-06,
+      "loss": 0.2425,
+      "step": 10683
+    },
+    {
+      "epoch": 0.7418414109151507,
+      "grad_norm": 4.736972380110522,
+      "learning_rate": 1.6484458869548824e-06,
+      "loss": 0.5308,
+      "step": 10684
+    },
+    {
+      "epoch": 0.7419108457158728,
+      "grad_norm": 3.3866915493995235,
+      "learning_rate": 1.6476115116647979e-06,
+      "loss": 0.3424,
+      "step": 10685
+    },
+    {
+      "epoch": 0.7419802805165949,
+      "grad_norm": 4.176760613307982,
+      "learning_rate": 1.646777305934728e-06,
+      "loss": 0.5837,
+      "step": 10686
+    },
+    {
+      "epoch": 0.742049715317317,
+      "grad_norm": 3.323665279313503,
+      "learning_rate": 1.6459432698068655e-06,
+      "loss": 0.2203,
+      "step": 10687
+    },
+    {
+      "epoch": 0.7421191501180392,
+      "grad_norm": 4.367397662764991,
+      "learning_rate": 1.6451094033233993e-06,
+      "loss": 0.4499,
+      "step": 10688
+    },
+    {
+      "epoch": 0.7421885849187613,
+      "grad_norm": 3.5143825290078916,
+      "learning_rate": 1.644275706526504e-06,
+      "loss": 0.3207,
+      "step": 10689
+    },
+    {
+      "epoch": 0.7422580197194834,
+      "grad_norm": 3.561132172175201,
+      "learning_rate": 1.6434421794583437e-06,
+      "loss": 0.5917,
+      "step": 10690
+    },
+    {
+      "epoch": 0.7423274545202055,
+      "grad_norm": 3.152377054179345,
+      "learning_rate": 1.6426088221610815e-06,
+      "loss": 0.2704,
+      "step": 10691
+    },
+    {
+      "epoch": 0.7423968893209276,
+      "grad_norm": 3.369850231829523,
+      "learning_rate": 1.6417756346768653e-06,
+      "loss": 0.4817,
+      "step": 10692
+    },
+    {
+      "epoch": 0.7424663241216498,
+      "grad_norm": 3.510689412744198,
+      "learning_rate": 1.6409426170478376e-06,
+      "loss": 0.4003,
+      "step": 10693
+    },
+    {
+      "epoch": 0.7425357589223719,
+      "grad_norm": 3.2593195733276215,
+      "learning_rate": 1.640109769316131e-06,
+      "loss": 0.3343,
+      "step": 10694
+    },
+    {
+      "epoch": 0.742605193723094,
+      "grad_norm": 3.7921478869028435,
+      "learning_rate": 1.639277091523871e-06,
+      "loss": 0.3897,
+      "step": 10695
+    },
+    {
+      "epoch": 0.7426746285238162,
+      "grad_norm": 12.452257479201311,
+      "learning_rate": 1.6384445837131724e-06,
+      "loss": 0.477,
+      "step": 10696
+    },
+    {
+      "epoch": 0.7427440633245382,
+      "grad_norm": 3.49881076665471,
+      "learning_rate": 1.637612245926143e-06,
+      "loss": 0.31,
+      "step": 10697
+    },
+    {
+      "epoch": 0.7428134981252604,
+      "grad_norm": 3.139905925812879,
+      "learning_rate": 1.6367800782048808e-06,
+      "loss": 0.3202,
+      "step": 10698
+    },
+    {
+      "epoch": 0.7428829329259825,
+      "grad_norm": 4.205274354879697,
+      "learning_rate": 1.6359480805914757e-06,
+      "loss": 0.327,
+      "step": 10699
+    },
+    {
+      "epoch": 0.7429523677267046,
+      "grad_norm": 4.986827690555049,
+      "learning_rate": 1.635116253128013e-06,
+      "loss": 0.7973,
+      "step": 10700
+    },
+    {
+      "epoch": 0.7430218025274268,
+      "grad_norm": 3.73900556902471,
+      "learning_rate": 1.6342845958565607e-06,
+      "loss": 0.4383,
+      "step": 10701
+    },
+    {
+      "epoch": 0.7430912373281489,
+      "grad_norm": 4.6272361630638335,
+      "learning_rate": 1.6334531088191834e-06,
+      "loss": 0.4296,
+      "step": 10702
+    },
+    {
+      "epoch": 0.7431606721288709,
+      "grad_norm": 3.3939648847837485,
+      "learning_rate": 1.6326217920579418e-06,
+      "loss": 0.488,
+      "step": 10703
+    },
+    {
+      "epoch": 0.7432301069295931,
+      "grad_norm": 6.25345925262801,
+      "learning_rate": 1.6317906456148763e-06,
+      "loss": 0.5421,
+      "step": 10704
+    },
+    {
+      "epoch": 0.7432995417303152,
+      "grad_norm": 4.502575771352725,
+      "learning_rate": 1.6309596695320301e-06,
+      "loss": 0.6098,
+      "step": 10705
+    },
+    {
+      "epoch": 0.7433689765310374,
+      "grad_norm": 3.0639678756382627,
+      "learning_rate": 1.6301288638514319e-06,
+      "loss": 0.2767,
+      "step": 10706
+    },
+    {
+      "epoch": 0.7434384113317595,
+      "grad_norm": 2.5940522320351818,
+      "learning_rate": 1.6292982286151015e-06,
+      "loss": 0.1974,
+      "step": 10707
+    },
+    {
+      "epoch": 0.7435078461324816,
+      "grad_norm": 3.36531594061589,
+      "learning_rate": 1.6284677638650536e-06,
+      "loss": 0.2959,
+      "step": 10708
+    },
+    {
+      "epoch": 0.7435772809332037,
+      "grad_norm": 4.1209477184486545,
+      "learning_rate": 1.6276374696432906e-06,
+      "loss": 0.2717,
+      "step": 10709
+    },
+    {
+      "epoch": 0.7436467157339258,
+      "grad_norm": 5.094487672716471,
+      "learning_rate": 1.6268073459918081e-06,
+      "loss": 0.3275,
+      "step": 10710
+    },
+    {
+      "epoch": 0.743716150534648,
+      "grad_norm": 4.617978640641326,
+      "learning_rate": 1.6259773929525929e-06,
+      "loss": 0.516,
+      "step": 10711
+    },
+    {
+      "epoch": 0.7437855853353701,
+      "grad_norm": 4.3007004207057244,
+      "learning_rate": 1.6251476105676233e-06,
+      "loss": 0.4942,
+      "step": 10712
+    },
+    {
+      "epoch": 0.7438550201360922,
+      "grad_norm": 4.507369502120782,
+      "learning_rate": 1.6243179988788677e-06,
+      "loss": 0.5102,
+      "step": 10713
+    },
+    {
+      "epoch": 0.7439244549368144,
+      "grad_norm": 3.909731558896055,
+      "learning_rate": 1.6234885579282906e-06,
+      "loss": 0.5227,
+      "step": 10714
+    },
+    {
+      "epoch": 0.7439938897375364,
+      "grad_norm": 4.402916143364794,
+      "learning_rate": 1.6226592877578395e-06,
+      "loss": 0.6036,
+      "step": 10715
+    },
+    {
+      "epoch": 0.7440633245382586,
+      "grad_norm": 3.443645077991829,
+      "learning_rate": 1.621830188409459e-06,
+      "loss": 0.3471,
+      "step": 10716
+    },
+    {
+      "epoch": 0.7441327593389807,
+      "grad_norm": 2.7998720548341387,
+      "learning_rate": 1.6210012599250858e-06,
+      "loss": 0.3916,
+      "step": 10717
+    },
+    {
+      "epoch": 0.7442021941397028,
+      "grad_norm": 3.732907849999485,
+      "learning_rate": 1.6201725023466476e-06,
+      "loss": 0.4536,
+      "step": 10718
+    },
+    {
+      "epoch": 0.744271628940425,
+      "grad_norm": 4.587264814300033,
+      "learning_rate": 1.6193439157160557e-06,
+      "loss": 0.512,
+      "step": 10719
+    },
+    {
+      "epoch": 0.7443410637411471,
+      "grad_norm": 4.359301508876848,
+      "learning_rate": 1.6185155000752252e-06,
+      "loss": 0.6733,
+      "step": 10720
+    },
+    {
+      "epoch": 0.7444104985418691,
+      "grad_norm": 5.1990413694372934,
+      "learning_rate": 1.6176872554660562e-06,
+      "loss": 0.3316,
+      "step": 10721
+    },
+    {
+      "epoch": 0.7444799333425913,
+      "grad_norm": 3.8093599209303846,
+      "learning_rate": 1.6168591819304348e-06,
+      "loss": 0.4291,
+      "step": 10722
+    },
+    {
+      "epoch": 0.7445493681433134,
+      "grad_norm": 4.242857888634728,
+      "learning_rate": 1.616031279510249e-06,
+      "loss": 0.3842,
+      "step": 10723
+    },
+    {
+      "epoch": 0.7446188029440356,
+      "grad_norm": 4.827956027437733,
+      "learning_rate": 1.6152035482473721e-06,
+      "loss": 0.4372,
+      "step": 10724
+    },
+    {
+      "epoch": 0.7446882377447577,
+      "grad_norm": 3.37151837933274,
+      "learning_rate": 1.61437598818367e-06,
+      "loss": 0.3953,
+      "step": 10725
+    },
+    {
+      "epoch": 0.7447576725454798,
+      "grad_norm": 5.025975528285516,
+      "learning_rate": 1.613548599360999e-06,
+      "loss": 0.6674,
+      "step": 10726
+    },
+    {
+      "epoch": 0.7448271073462019,
+      "grad_norm": 4.6569875934445335,
+      "learning_rate": 1.612721381821208e-06,
+      "loss": 0.3867,
+      "step": 10727
+    },
+    {
+      "epoch": 0.744896542146924,
+      "grad_norm": 3.840588181174105,
+      "learning_rate": 1.6118943356061362e-06,
+      "loss": 0.3752,
+      "step": 10728
+    },
+    {
+      "epoch": 0.7449659769476462,
+      "grad_norm": 5.239786664718549,
+      "learning_rate": 1.6110674607576155e-06,
+      "loss": 0.7656,
+      "step": 10729
+    },
+    {
+      "epoch": 0.7450354117483683,
+      "grad_norm": 4.091971181305727,
+      "learning_rate": 1.6102407573174673e-06,
+      "loss": 0.3595,
+      "step": 10730
+    },
+    {
+      "epoch": 0.7451048465490904,
+      "grad_norm": 6.996525987195954,
+      "learning_rate": 1.609414225327504e-06,
+      "loss": 0.73,
+      "step": 10731
+    },
+    {
+      "epoch": 0.7451742813498126,
+      "grad_norm": 3.5023602291871208,
+      "learning_rate": 1.6085878648295362e-06,
+      "loss": 0.4316,
+      "step": 10732
+    },
+    {
+      "epoch": 0.7452437161505346,
+      "grad_norm": 8.369436038659725,
+      "learning_rate": 1.6077616758653535e-06,
+      "loss": 0.6986,
+      "step": 10733
+    },
+    {
+      "epoch": 0.7453131509512567,
+      "grad_norm": 3.2503108024914606,
+      "learning_rate": 1.6069356584767476e-06,
+      "loss": 0.2635,
+      "step": 10734
+    },
+    {
+      "epoch": 0.7453825857519789,
+      "grad_norm": 3.417845314701299,
+      "learning_rate": 1.6061098127054986e-06,
+      "loss": 0.3492,
+      "step": 10735
+    },
+    {
+      "epoch": 0.745452020552701,
+      "grad_norm": 3.963536802357365,
+      "learning_rate": 1.6052841385933722e-06,
+      "loss": 0.5561,
+      "step": 10736
+    },
+    {
+      "epoch": 0.7455214553534232,
+      "grad_norm": 2.899756155959895,
+      "learning_rate": 1.6044586361821334e-06,
+      "loss": 0.2992,
+      "step": 10737
+    },
+    {
+      "epoch": 0.7455908901541453,
+      "grad_norm": 4.056068132552119,
+      "learning_rate": 1.6036333055135345e-06,
+      "loss": 0.346,
+      "step": 10738
+    },
+    {
+      "epoch": 0.7456603249548673,
+      "grad_norm": 5.024307644332964,
+      "learning_rate": 1.6028081466293205e-06,
+      "loss": 0.6109,
+      "step": 10739
+    },
+    {
+      "epoch": 0.7457297597555895,
+      "grad_norm": 4.288763680268224,
+      "learning_rate": 1.6019831595712253e-06,
+      "loss": 0.4385,
+      "step": 10740
+    },
+    {
+      "epoch": 0.7457991945563116,
+      "grad_norm": 5.253143407926876,
+      "learning_rate": 1.6011583443809775e-06,
+      "loss": 0.5774,
+      "step": 10741
+    },
+    {
+      "epoch": 0.7458686293570338,
+      "grad_norm": 6.125331191291605,
+      "learning_rate": 1.6003337011002928e-06,
+      "loss": 0.7968,
+      "step": 10742
+    },
+    {
+      "epoch": 0.7459380641577559,
+      "grad_norm": 4.491532595821953,
+      "learning_rate": 1.5995092297708853e-06,
+      "loss": 0.6202,
+      "step": 10743
+    },
+    {
+      "epoch": 0.746007498958478,
+      "grad_norm": 4.227695370365082,
+      "learning_rate": 1.598684930434452e-06,
+      "loss": 0.5645,
+      "step": 10744
+    },
+    {
+      "epoch": 0.7460769337592001,
+      "grad_norm": 4.610301110265456,
+      "learning_rate": 1.5978608031326843e-06,
+      "loss": 0.5898,
+      "step": 10745
+    },
+    {
+      "epoch": 0.7461463685599222,
+      "grad_norm": 3.3250132519725963,
+      "learning_rate": 1.5970368479072706e-06,
+      "loss": 0.2467,
+      "step": 10746
+    },
+    {
+      "epoch": 0.7462158033606443,
+      "grad_norm": 4.475805380480244,
+      "learning_rate": 1.5962130647998808e-06,
+      "loss": 0.5252,
+      "step": 10747
+    },
+    {
+      "epoch": 0.7462852381613665,
+      "grad_norm": 3.7565357315095937,
+      "learning_rate": 1.5953894538521808e-06,
+      "loss": 0.3475,
+      "step": 10748
+    },
+    {
+      "epoch": 0.7463546729620886,
+      "grad_norm": 3.9483797873444706,
+      "learning_rate": 1.5945660151058313e-06,
+      "loss": 0.4317,
+      "step": 10749
+    },
+    {
+      "epoch": 0.7464241077628108,
+      "grad_norm": 3.5784515376103876,
+      "learning_rate": 1.593742748602481e-06,
+      "loss": 0.3668,
+      "step": 10750
+    },
+    {
+      "epoch": 0.7464935425635328,
+      "grad_norm": 4.52426849579044,
+      "learning_rate": 1.592919654383765e-06,
+      "loss": 0.5252,
+      "step": 10751
+    },
+    {
+      "epoch": 0.7465629773642549,
+      "grad_norm": 4.645301720287652,
+      "learning_rate": 1.5920967324913189e-06,
+      "loss": 0.4686,
+      "step": 10752
+    },
+    {
+      "epoch": 0.7466324121649771,
+      "grad_norm": 3.630575073213047,
+      "learning_rate": 1.5912739829667634e-06,
+      "loss": 0.4839,
+      "step": 10753
+    },
+    {
+      "epoch": 0.7467018469656992,
+      "grad_norm": 5.103885647452111,
+      "learning_rate": 1.5904514058517129e-06,
+      "loss": 0.7937,
+      "step": 10754
+    },
+    {
+      "epoch": 0.7467712817664214,
+      "grad_norm": 3.5805977973637946,
+      "learning_rate": 1.5896290011877724e-06,
+      "loss": 0.3834,
+      "step": 10755
+    },
+    {
+      "epoch": 0.7468407165671435,
+      "grad_norm": 3.085330222388638,
+      "learning_rate": 1.5888067690165377e-06,
+      "loss": 0.2817,
+      "step": 10756
+    },
+    {
+      "epoch": 0.7469101513678655,
+      "grad_norm": 5.371420411151042,
+      "learning_rate": 1.5879847093795963e-06,
+      "loss": 0.6069,
+      "step": 10757
+    },
+    {
+      "epoch": 0.7469795861685877,
+      "grad_norm": 4.63678851263773,
+      "learning_rate": 1.5871628223185276e-06,
+      "loss": 0.438,
+      "step": 10758
+    },
+    {
+      "epoch": 0.7470490209693098,
+      "grad_norm": 2.451520096724459,
+      "learning_rate": 1.586341107874902e-06,
+      "loss": 0.1333,
+      "step": 10759
+    },
+    {
+      "epoch": 0.7471184557700319,
+      "grad_norm": 2.9624286283012875,
+      "learning_rate": 1.5855195660902795e-06,
+      "loss": 0.3513,
+      "step": 10760
+    },
+    {
+      "epoch": 0.7471878905707541,
+      "grad_norm": 3.9496752528056156,
+      "learning_rate": 1.5846981970062142e-06,
+      "loss": 0.4849,
+      "step": 10761
+    },
+    {
+      "epoch": 0.7472573253714762,
+      "grad_norm": 3.6247196333969063,
+      "learning_rate": 1.5838770006642472e-06,
+      "loss": 0.1976,
+      "step": 10762
+    },
+    {
+      "epoch": 0.7473267601721983,
+      "grad_norm": 3.7226403352849684,
+      "learning_rate": 1.5830559771059179e-06,
+      "loss": 0.4268,
+      "step": 10763
+    },
+    {
+      "epoch": 0.7473961949729204,
+      "grad_norm": 3.3911655258994102,
+      "learning_rate": 1.5822351263727525e-06,
+      "loss": 0.4843,
+      "step": 10764
+    },
+    {
+      "epoch": 0.7474656297736425,
+      "grad_norm": 4.272109474991247,
+      "learning_rate": 1.5814144485062632e-06,
+      "loss": 0.543,
+      "step": 10765
+    },
+    {
+      "epoch": 0.7475350645743647,
+      "grad_norm": 3.73506786183235,
+      "learning_rate": 1.580593943547965e-06,
+      "loss": 0.4154,
+      "step": 10766
+    },
+    {
+      "epoch": 0.7476044993750868,
+      "grad_norm": 2.504551809721038,
+      "learning_rate": 1.579773611539357e-06,
+      "loss": 0.2156,
+      "step": 10767
+    },
+    {
+      "epoch": 0.747673934175809,
+      "grad_norm": 4.502962883950316,
+      "learning_rate": 1.5789534525219264e-06,
+      "loss": 0.3648,
+      "step": 10768
+    },
+    {
+      "epoch": 0.747743368976531,
+      "grad_norm": 2.8362106367606397,
+      "learning_rate": 1.5781334665371606e-06,
+      "loss": 0.2606,
+      "step": 10769
+    },
+    {
+      "epoch": 0.7478128037772531,
+      "grad_norm": 3.0329793469758384,
+      "learning_rate": 1.5773136536265316e-06,
+      "loss": 0.2942,
+      "step": 10770
+    },
+    {
+      "epoch": 0.7478822385779753,
+      "grad_norm": 3.2096632581618776,
+      "learning_rate": 1.5764940138315054e-06,
+      "loss": 0.2565,
+      "step": 10771
+    },
+    {
+      "epoch": 0.7479516733786974,
+      "grad_norm": 3.614935601039839,
+      "learning_rate": 1.575674547193538e-06,
+      "loss": 0.5519,
+      "step": 10772
+    },
+    {
+      "epoch": 0.7480211081794196,
+      "grad_norm": 4.31374214616722,
+      "learning_rate": 1.5748552537540767e-06,
+      "loss": 0.4851,
+      "step": 10773
+    },
+    {
+      "epoch": 0.7480905429801417,
+      "grad_norm": 4.198776640616115,
+      "learning_rate": 1.5740361335545594e-06,
+      "loss": 0.7891,
+      "step": 10774
+    },
+    {
+      "epoch": 0.7481599777808637,
+      "grad_norm": 4.374388127008037,
+      "learning_rate": 1.5732171866364215e-06,
+      "loss": 0.424,
+      "step": 10775
+    },
+    {
+      "epoch": 0.7482294125815859,
+      "grad_norm": 3.7762069518537484,
+      "learning_rate": 1.5723984130410786e-06,
+      "loss": 0.3127,
+      "step": 10776
+    },
+    {
+      "epoch": 0.748298847382308,
+      "grad_norm": 4.2095153584519425,
+      "learning_rate": 1.5715798128099434e-06,
+      "loss": 0.4966,
+      "step": 10777
+    },
+    {
+      "epoch": 0.7483682821830301,
+      "grad_norm": 3.4465296106155674,
+      "learning_rate": 1.5707613859844256e-06,
+      "loss": 0.4021,
+      "step": 10778
+    },
+    {
+      "epoch": 0.7484377169837523,
+      "grad_norm": 4.433322252930832,
+      "learning_rate": 1.5699431326059144e-06,
+      "loss": 0.5788,
+      "step": 10779
+    },
+    {
+      "epoch": 0.7485071517844744,
+      "grad_norm": 3.8884119864142237,
+      "learning_rate": 1.569125052715797e-06,
+      "loss": 0.5514,
+      "step": 10780
+    },
+    {
+      "epoch": 0.7485765865851965,
+      "grad_norm": 4.24004776531264,
+      "learning_rate": 1.5683071463554533e-06,
+      "loss": 0.379,
+      "step": 10781
+    },
+    {
+      "epoch": 0.7486460213859186,
+      "grad_norm": 3.778303820482812,
+      "learning_rate": 1.5674894135662511e-06,
+      "loss": 0.3796,
+      "step": 10782
+    },
+    {
+      "epoch": 0.7487154561866407,
+      "grad_norm": 3.2977740412209724,
+      "learning_rate": 1.5666718543895504e-06,
+      "loss": 0.3489,
+      "step": 10783
+    },
+    {
+      "epoch": 0.7487848909873629,
+      "grad_norm": 3.2309793608249584,
+      "learning_rate": 1.565854468866702e-06,
+      "loss": 0.2639,
+      "step": 10784
+    },
+    {
+      "epoch": 0.748854325788085,
+      "grad_norm": 3.686978238339916,
+      "learning_rate": 1.5650372570390486e-06,
+      "loss": 0.3476,
+      "step": 10785
+    },
+    {
+      "epoch": 0.7489237605888072,
+      "grad_norm": 4.615793781438089,
+      "learning_rate": 1.564220218947924e-06,
+      "loss": 0.6413,
+      "step": 10786
+    },
+    {
+      "epoch": 0.7489931953895292,
+      "grad_norm": 3.8349704554233814,
+      "learning_rate": 1.5634033546346528e-06,
+      "loss": 0.3406,
+      "step": 10787
+    },
+    {
+      "epoch": 0.7490626301902513,
+      "grad_norm": 3.668357356361762,
+      "learning_rate": 1.5625866641405507e-06,
+      "loss": 0.3792,
+      "step": 10788
+    },
+    {
+      "epoch": 0.7491320649909735,
+      "grad_norm": 3.1241498140073625,
+      "learning_rate": 1.5617701475069258e-06,
+      "loss": 0.2701,
+      "step": 10789
+    },
+    {
+      "epoch": 0.7492014997916956,
+      "grad_norm": 4.667344470913812,
+      "learning_rate": 1.5609538047750765e-06,
+      "loss": 0.5592,
+      "step": 10790
+    },
+    {
+      "epoch": 0.7492709345924177,
+      "grad_norm": 3.9011715709285157,
+      "learning_rate": 1.5601376359862919e-06,
+      "loss": 0.5367,
+      "step": 10791
+    },
+    {
+      "epoch": 0.7493403693931399,
+      "grad_norm": 4.798538971638306,
+      "learning_rate": 1.5593216411818534e-06,
+      "loss": 0.5888,
+      "step": 10792
+    },
+    {
+      "epoch": 0.7494098041938619,
+      "grad_norm": 4.759249304416267,
+      "learning_rate": 1.5585058204030324e-06,
+      "loss": 0.4821,
+      "step": 10793
+    },
+    {
+      "epoch": 0.7494792389945841,
+      "grad_norm": 4.632998171494326,
+      "learning_rate": 1.5576901736910916e-06,
+      "loss": 0.6155,
+      "step": 10794
+    },
+    {
+      "epoch": 0.7495486737953062,
+      "grad_norm": 3.7465869284027975,
+      "learning_rate": 1.5568747010872881e-06,
+      "loss": 0.4414,
+      "step": 10795
+    },
+    {
+      "epoch": 0.7496181085960283,
+      "grad_norm": 3.544095792551927,
+      "learning_rate": 1.5560594026328674e-06,
+      "loss": 0.2607,
+      "step": 10796
+    },
+    {
+      "epoch": 0.7496875433967505,
+      "grad_norm": 3.5155628327023924,
+      "learning_rate": 1.5552442783690618e-06,
+      "loss": 0.2695,
+      "step": 10797
+    },
+    {
+      "epoch": 0.7497569781974726,
+      "grad_norm": 2.964366086313482,
+      "learning_rate": 1.554429328337105e-06,
+      "loss": 0.3318,
+      "step": 10798
+    },
+    {
+      "epoch": 0.7498264129981947,
+      "grad_norm": 3.9308731033453093,
+      "learning_rate": 1.5536145525782125e-06,
+      "loss": 0.4363,
+      "step": 10799
+    },
+    {
+      "epoch": 0.7498958477989168,
+      "grad_norm": 3.9051659731151678,
+      "learning_rate": 1.5527999511335973e-06,
+      "loss": 0.4176,
+      "step": 10800
+    },
+    {
+      "epoch": 0.7499652825996389,
+      "grad_norm": 3.729238828892474,
+      "learning_rate": 1.5519855240444592e-06,
+      "loss": 0.3564,
+      "step": 10801
+    },
+    {
+      "epoch": 0.7500347174003611,
+      "grad_norm": 5.684337033058436,
+      "learning_rate": 1.5511712713519917e-06,
+      "loss": 0.4475,
+      "step": 10802
+    },
+    {
+      "epoch": 0.7501041522010832,
+      "grad_norm": 4.963222512340322,
+      "learning_rate": 1.5503571930973788e-06,
+      "loss": 0.3704,
+      "step": 10803
+    },
+    {
+      "epoch": 0.7501735870018053,
+      "grad_norm": 3.3679724509121725,
+      "learning_rate": 1.5495432893217954e-06,
+      "loss": 0.3256,
+      "step": 10804
+    },
+    {
+      "epoch": 0.7502430218025274,
+      "grad_norm": 3.551451503744839,
+      "learning_rate": 1.5487295600664082e-06,
+      "loss": 0.2862,
+      "step": 10805
+    },
+    {
+      "epoch": 0.7503124566032495,
+      "grad_norm": 4.118530458548856,
+      "learning_rate": 1.5479160053723724e-06,
+      "loss": 0.1842,
+      "step": 10806
+    },
+    {
+      "epoch": 0.7503818914039717,
+      "grad_norm": 4.475244658256488,
+      "learning_rate": 1.5471026252808425e-06,
+      "loss": 0.3785,
+      "step": 10807
+    },
+    {
+      "epoch": 0.7504513262046938,
+      "grad_norm": 3.6188302680077125,
+      "learning_rate": 1.5462894198329525e-06,
+      "loss": 0.4149,
+      "step": 10808
+    },
+    {
+      "epoch": 0.7505207610054159,
+      "grad_norm": 2.8734628134609985,
+      "learning_rate": 1.5454763890698343e-06,
+      "loss": 0.2985,
+      "step": 10809
+    },
+    {
+      "epoch": 0.7505901958061381,
+      "grad_norm": 4.09254119199055,
+      "learning_rate": 1.5446635330326143e-06,
+      "loss": 0.4912,
+      "step": 10810
+    },
+    {
+      "epoch": 0.7506596306068601,
+      "grad_norm": 3.8011498164116686,
+      "learning_rate": 1.5438508517624013e-06,
+      "loss": 0.461,
+      "step": 10811
+    },
+    {
+      "epoch": 0.7507290654075823,
+      "grad_norm": 3.7660988716721966,
+      "learning_rate": 1.5430383453002994e-06,
+      "loss": 0.289,
+      "step": 10812
+    },
+    {
+      "epoch": 0.7507985002083044,
+      "grad_norm": 3.871700841257737,
+      "learning_rate": 1.542226013687408e-06,
+      "loss": 0.4064,
+      "step": 10813
+    },
+    {
+      "epoch": 0.7508679350090265,
+      "grad_norm": 3.262608480930866,
+      "learning_rate": 1.5414138569648123e-06,
+      "loss": 0.2609,
+      "step": 10814
+    },
+    {
+      "epoch": 0.7509373698097487,
+      "grad_norm": 4.040779300852778,
+      "learning_rate": 1.5406018751735897e-06,
+      "loss": 0.5068,
+      "step": 10815
+    },
+    {
+      "epoch": 0.7510068046104708,
+      "grad_norm": 3.774810794282543,
+      "learning_rate": 1.53979006835481e-06,
+      "loss": 0.3397,
+      "step": 10816
+    },
+    {
+      "epoch": 0.7510762394111928,
+      "grad_norm": 3.936055734514942,
+      "learning_rate": 1.538978436549533e-06,
+      "loss": 0.4956,
+      "step": 10817
+    },
+    {
+      "epoch": 0.751145674211915,
+      "grad_norm": 3.499377305583845,
+      "learning_rate": 1.5381669797988102e-06,
+      "loss": 0.2968,
+      "step": 10818
+    },
+    {
+      "epoch": 0.7512151090126371,
+      "grad_norm": 5.9183578317186845,
+      "learning_rate": 1.5373556981436844e-06,
+      "loss": 0.4002,
+      "step": 10819
+    },
+    {
+      "epoch": 0.7512845438133593,
+      "grad_norm": 3.367113603604195,
+      "learning_rate": 1.5365445916251892e-06,
+      "loss": 0.2772,
+      "step": 10820
+    },
+    {
+      "epoch": 0.7513539786140814,
+      "grad_norm": 3.6204883671421926,
+      "learning_rate": 1.5357336602843493e-06,
+      "loss": 0.3106,
+      "step": 10821
+    },
+    {
+      "epoch": 0.7514234134148035,
+      "grad_norm": 4.1995443239462915,
+      "learning_rate": 1.5349229041621805e-06,
+      "loss": 0.5288,
+      "step": 10822
+    },
+    {
+      "epoch": 0.7514928482155256,
+      "grad_norm": 2.3651253268659755,
+      "learning_rate": 1.5341123232996885e-06,
+      "loss": 0.2201,
+      "step": 10823
+    },
+    {
+      "epoch": 0.7515622830162477,
+      "grad_norm": 3.8934704116646848,
+      "learning_rate": 1.5333019177378755e-06,
+      "loss": 0.2655,
+      "step": 10824
+    },
+    {
+      "epoch": 0.7516317178169699,
+      "grad_norm": 2.4630705927357233,
+      "learning_rate": 1.5324916875177298e-06,
+      "loss": 0.203,
+      "step": 10825
+    },
+    {
+      "epoch": 0.751701152617692,
+      "grad_norm": 3.672495719394316,
+      "learning_rate": 1.531681632680228e-06,
+      "loss": 0.3714,
+      "step": 10826
+    },
+    {
+      "epoch": 0.7517705874184141,
+      "grad_norm": 4.115347602760836,
+      "learning_rate": 1.5308717532663459e-06,
+      "loss": 0.3668,
+      "step": 10827
+    },
+    {
+      "epoch": 0.7518400222191363,
+      "grad_norm": 4.114003218512139,
+      "learning_rate": 1.5300620493170464e-06,
+      "loss": 0.5001,
+      "step": 10828
+    },
+    {
+      "epoch": 0.7519094570198583,
+      "grad_norm": 3.883999831962567,
+      "learning_rate": 1.529252520873279e-06,
+      "loss": 0.4119,
+      "step": 10829
+    },
+    {
+      "epoch": 0.7519788918205804,
+      "grad_norm": 4.099435219746042,
+      "learning_rate": 1.5284431679759931e-06,
+      "loss": 0.3746,
+      "step": 10830
+    },
+    {
+      "epoch": 0.7520483266213026,
+      "grad_norm": 4.261140404968297,
+      "learning_rate": 1.5276339906661237e-06,
+      "loss": 0.5792,
+      "step": 10831
+    },
+    {
+      "epoch": 0.7521177614220247,
+      "grad_norm": 3.989681417245947,
+      "learning_rate": 1.5268249889845976e-06,
+      "loss": 0.5014,
+      "step": 10832
+    },
+    {
+      "epoch": 0.7521871962227469,
+      "grad_norm": 4.1525628306265,
+      "learning_rate": 1.5260161629723335e-06,
+      "loss": 0.3542,
+      "step": 10833
+    },
+    {
+      "epoch": 0.752256631023469,
+      "grad_norm": 4.403864980357464,
+      "learning_rate": 1.5252075126702404e-06,
+      "loss": 0.4907,
+      "step": 10834
+    },
+    {
+      "epoch": 0.752326065824191,
+      "grad_norm": 4.667618177595061,
+      "learning_rate": 1.5243990381192198e-06,
+      "loss": 0.4436,
+      "step": 10835
+    },
+    {
+      "epoch": 0.7523955006249132,
+      "grad_norm": 3.635707599117771,
+      "learning_rate": 1.523590739360163e-06,
+      "loss": 0.4036,
+      "step": 10836
+    },
+    {
+      "epoch": 0.7524649354256353,
+      "grad_norm": 3.937004148282174,
+      "learning_rate": 1.522782616433952e-06,
+      "loss": 0.4583,
+      "step": 10837
+    },
+    {
+      "epoch": 0.7525343702263575,
+      "grad_norm": 3.404220925640154,
+      "learning_rate": 1.5219746693814608e-06,
+      "loss": 0.2288,
+      "step": 10838
+    },
+    {
+      "epoch": 0.7526038050270796,
+      "grad_norm": 3.6370156474716353,
+      "learning_rate": 1.521166898243558e-06,
+      "loss": 0.4149,
+      "step": 10839
+    },
+    {
+      "epoch": 0.7526732398278017,
+      "grad_norm": 4.001346093756767,
+      "learning_rate": 1.5203593030610952e-06,
+      "loss": 0.3355,
+      "step": 10840
+    },
+    {
+      "epoch": 0.7527426746285238,
+      "grad_norm": 3.388106660978092,
+      "learning_rate": 1.51955188387492e-06,
+      "loss": 0.4198,
+      "step": 10841
+    },
+    {
+      "epoch": 0.7528121094292459,
+      "grad_norm": 4.365037132100963,
+      "learning_rate": 1.5187446407258754e-06,
+      "loss": 0.3926,
+      "step": 10842
+    },
+    {
+      "epoch": 0.7528815442299681,
+      "grad_norm": 4.557933514482715,
+      "learning_rate": 1.5179375736547841e-06,
+      "loss": 0.3109,
+      "step": 10843
+    },
+    {
+      "epoch": 0.7529509790306902,
+      "grad_norm": 4.04629911341105,
+      "learning_rate": 1.5171306827024717e-06,
+      "loss": 0.503,
+      "step": 10844
+    },
+    {
+      "epoch": 0.7530204138314123,
+      "grad_norm": 3.930145655538194,
+      "learning_rate": 1.5163239679097485e-06,
+      "loss": 0.4098,
+      "step": 10845
+    },
+    {
+      "epoch": 0.7530898486321345,
+      "grad_norm": 4.780393180299837,
+      "learning_rate": 1.5155174293174169e-06,
+      "loss": 0.3488,
+      "step": 10846
+    },
+    {
+      "epoch": 0.7531592834328565,
+      "grad_norm": 3.712062323488566,
+      "learning_rate": 1.5147110669662706e-06,
+      "loss": 0.4577,
+      "step": 10847
+    },
+    {
+      "epoch": 0.7532287182335786,
+      "grad_norm": 4.574773515264511,
+      "learning_rate": 1.5139048808970946e-06,
+      "loss": 0.2985,
+      "step": 10848
+    },
+    {
+      "epoch": 0.7532981530343008,
+      "grad_norm": 3.9969745331329376,
+      "learning_rate": 1.513098871150665e-06,
+      "loss": 0.3586,
+      "step": 10849
+    },
+    {
+      "epoch": 0.7533675878350229,
+      "grad_norm": 4.703913016772175,
+      "learning_rate": 1.5122930377677491e-06,
+      "loss": 0.49,
+      "step": 10850
+    },
+    {
+      "epoch": 0.7534370226357451,
+      "grad_norm": 4.17241359021055,
+      "learning_rate": 1.5114873807891046e-06,
+      "loss": 0.5335,
+      "step": 10851
+    },
+    {
+      "epoch": 0.7535064574364672,
+      "grad_norm": 3.4750077882147274,
+      "learning_rate": 1.510681900255479e-06,
+      "loss": 0.3889,
+      "step": 10852
+    },
+    {
+      "epoch": 0.7535758922371892,
+      "grad_norm": 4.337672957674126,
+      "learning_rate": 1.5098765962076179e-06,
+      "loss": 0.3779,
+      "step": 10853
+    },
+    {
+      "epoch": 0.7536453270379114,
+      "grad_norm": 4.963217297367776,
+      "learning_rate": 1.5090714686862474e-06,
+      "loss": 0.4656,
+      "step": 10854
+    },
+    {
+      "epoch": 0.7537147618386335,
+      "grad_norm": 3.3649512188678554,
+      "learning_rate": 1.5082665177320904e-06,
+      "loss": 0.2619,
+      "step": 10855
+    },
+    {
+      "epoch": 0.7537841966393557,
+      "grad_norm": 3.7016798046377044,
+      "learning_rate": 1.5074617433858635e-06,
+      "loss": 0.475,
+      "step": 10856
+    },
+    {
+      "epoch": 0.7538536314400778,
+      "grad_norm": 3.497121932521108,
+      "learning_rate": 1.5066571456882706e-06,
+      "loss": 0.5004,
+      "step": 10857
+    },
+    {
+      "epoch": 0.7539230662407999,
+      "grad_norm": 3.513544525193985,
+      "learning_rate": 1.5058527246800036e-06,
+      "loss": 0.2602,
+      "step": 10858
+    },
+    {
+      "epoch": 0.753992501041522,
+      "grad_norm": 2.7748400544612246,
+      "learning_rate": 1.5050484804017535e-06,
+      "loss": 0.1948,
+      "step": 10859
+    },
+    {
+      "epoch": 0.7540619358422441,
+      "grad_norm": 4.180236064269588,
+      "learning_rate": 1.5042444128941973e-06,
+      "loss": 0.4174,
+      "step": 10860
+    },
+    {
+      "epoch": 0.7541313706429662,
+      "grad_norm": 3.7099289523615657,
+      "learning_rate": 1.5034405221980008e-06,
+      "loss": 0.3756,
+      "step": 10861
+    },
+    {
+      "epoch": 0.7542008054436884,
+      "grad_norm": 3.6075002210497438,
+      "learning_rate": 1.5026368083538274e-06,
+      "loss": 0.3312,
+      "step": 10862
+    },
+    {
+      "epoch": 0.7542702402444105,
+      "grad_norm": 5.8062302508348305,
+      "learning_rate": 1.5018332714023259e-06,
+      "loss": 0.5987,
+      "step": 10863
+    },
+    {
+      "epoch": 0.7543396750451327,
+      "grad_norm": 3.772019394556794,
+      "learning_rate": 1.5010299113841397e-06,
+      "loss": 0.4437,
+      "step": 10864
+    },
+    {
+      "epoch": 0.7544091098458547,
+      "grad_norm": 4.177867038414421,
+      "learning_rate": 1.5002267283399014e-06,
+      "loss": 0.2815,
+      "step": 10865
+    },
+    {
+      "epoch": 0.7544785446465768,
+      "grad_norm": 5.761422937283608,
+      "learning_rate": 1.4994237223102348e-06,
+      "loss": 0.508,
+      "step": 10866
+    },
+    {
+      "epoch": 0.754547979447299,
+      "grad_norm": 3.780957287474016,
+      "learning_rate": 1.498620893335755e-06,
+      "loss": 0.518,
+      "step": 10867
+    },
+    {
+      "epoch": 0.7546174142480211,
+      "grad_norm": 2.368172189958371,
+      "learning_rate": 1.4978182414570686e-06,
+      "loss": 0.0944,
+      "step": 10868
+    },
+    {
+      "epoch": 0.7546868490487433,
+      "grad_norm": 5.407462471123704,
+      "learning_rate": 1.4970157667147722e-06,
+      "loss": 0.6807,
+      "step": 10869
+    },
+    {
+      "epoch": 0.7547562838494654,
+      "grad_norm": 3.3425220849430484,
+      "learning_rate": 1.4962134691494535e-06,
+      "loss": 0.3829,
+      "step": 10870
+    },
+    {
+      "epoch": 0.7548257186501874,
+      "grad_norm": 4.972395060596145,
+      "learning_rate": 1.4954113488016953e-06,
+      "loss": 0.6794,
+      "step": 10871
+    },
+    {
+      "epoch": 0.7548951534509096,
+      "grad_norm": 4.161418687028618,
+      "learning_rate": 1.4946094057120635e-06,
+      "loss": 0.4752,
+      "step": 10872
+    },
+    {
+      "epoch": 0.7549645882516317,
+      "grad_norm": 4.208006096791668,
+      "learning_rate": 1.4938076399211204e-06,
+      "loss": 0.5209,
+      "step": 10873
+    },
+    {
+      "epoch": 0.7550340230523538,
+      "grad_norm": 4.628660155705154,
+      "learning_rate": 1.4930060514694205e-06,
+      "loss": 0.6538,
+      "step": 10874
+    },
+    {
+      "epoch": 0.755103457853076,
+      "grad_norm": 4.490862294122593,
+      "learning_rate": 1.4922046403975066e-06,
+      "loss": 0.3604,
+      "step": 10875
+    },
+    {
+      "epoch": 0.7551728926537981,
+      "grad_norm": 4.751099851937854,
+      "learning_rate": 1.4914034067459121e-06,
+      "loss": 0.6423,
+      "step": 10876
+    },
+    {
+      "epoch": 0.7552423274545202,
+      "grad_norm": 4.352000952317273,
+      "learning_rate": 1.4906023505551636e-06,
+      "loss": 0.4182,
+      "step": 10877
+    },
+    {
+      "epoch": 0.7553117622552423,
+      "grad_norm": 4.22780731862895,
+      "learning_rate": 1.4898014718657766e-06,
+      "loss": 0.2967,
+      "step": 10878
+    },
+    {
+      "epoch": 0.7553811970559644,
+      "grad_norm": 3.5822158384145,
+      "learning_rate": 1.489000770718259e-06,
+      "loss": 0.4534,
+      "step": 10879
+    },
+    {
+      "epoch": 0.7554506318566866,
+      "grad_norm": 4.1824914233985195,
+      "learning_rate": 1.4882002471531098e-06,
+      "loss": 0.5078,
+      "step": 10880
+    },
+    {
+      "epoch": 0.7555200666574087,
+      "grad_norm": 5.142505104374632,
+      "learning_rate": 1.4873999012108181e-06,
+      "loss": 0.5922,
+      "step": 10881
+    },
+    {
+      "epoch": 0.7555895014581309,
+      "grad_norm": 3.932518587642267,
+      "learning_rate": 1.486599732931865e-06,
+      "loss": 0.543,
+      "step": 10882
+    },
+    {
+      "epoch": 0.7556589362588529,
+      "grad_norm": 3.4152065987041818,
+      "learning_rate": 1.485799742356721e-06,
+      "loss": 0.3439,
+      "step": 10883
+    },
+    {
+      "epoch": 0.755728371059575,
+      "grad_norm": 4.046163460223747,
+      "learning_rate": 1.4849999295258482e-06,
+      "loss": 0.5141,
+      "step": 10884
+    },
+    {
+      "epoch": 0.7557978058602972,
+      "grad_norm": 4.122519009710842,
+      "learning_rate": 1.4842002944797045e-06,
+      "loss": 0.3613,
+      "step": 10885
+    },
+    {
+      "epoch": 0.7558672406610193,
+      "grad_norm": 2.644082516938727,
+      "learning_rate": 1.4834008372587305e-06,
+      "loss": 0.2647,
+      "step": 10886
+    },
+    {
+      "epoch": 0.7559366754617414,
+      "grad_norm": 4.669648632098493,
+      "learning_rate": 1.4826015579033609e-06,
+      "loss": 0.5353,
+      "step": 10887
+    },
+    {
+      "epoch": 0.7560061102624636,
+      "grad_norm": 4.019301436356018,
+      "learning_rate": 1.4818024564540257e-06,
+      "loss": 0.4886,
+      "step": 10888
+    },
+    {
+      "epoch": 0.7560755450631856,
+      "grad_norm": 4.680988454273236,
+      "learning_rate": 1.481003532951143e-06,
+      "loss": 0.4747,
+      "step": 10889
+    },
+    {
+      "epoch": 0.7561449798639078,
+      "grad_norm": 4.164668628694312,
+      "learning_rate": 1.480204787435116e-06,
+      "loss": 0.5531,
+      "step": 10890
+    },
+    {
+      "epoch": 0.7562144146646299,
+      "grad_norm": 3.685863543818351,
+      "learning_rate": 1.4794062199463499e-06,
+      "loss": 0.3992,
+      "step": 10891
+    },
+    {
+      "epoch": 0.756283849465352,
+      "grad_norm": 3.717078368859052,
+      "learning_rate": 1.478607830525235e-06,
+      "loss": 0.3209,
+      "step": 10892
+    },
+    {
+      "epoch": 0.7563532842660742,
+      "grad_norm": 3.9986374885378204,
+      "learning_rate": 1.4778096192121478e-06,
+      "loss": 0.4633,
+      "step": 10893
+    },
+    {
+      "epoch": 0.7564227190667963,
+      "grad_norm": 3.5311867353538404,
+      "learning_rate": 1.477011586047466e-06,
+      "loss": 0.3858,
+      "step": 10894
+    },
+    {
+      "epoch": 0.7564921538675184,
+      "grad_norm": 3.485576507525538,
+      "learning_rate": 1.476213731071552e-06,
+      "loss": 0.2748,
+      "step": 10895
+    },
+    {
+      "epoch": 0.7565615886682405,
+      "grad_norm": 5.073771462552815,
+      "learning_rate": 1.4754160543247597e-06,
+      "loss": 0.8224,
+      "step": 10896
+    },
+    {
+      "epoch": 0.7566310234689626,
+      "grad_norm": 5.422545770992055,
+      "learning_rate": 1.4746185558474358e-06,
+      "loss": 0.4369,
+      "step": 10897
+    },
+    {
+      "epoch": 0.7567004582696848,
+      "grad_norm": 3.958965071722932,
+      "learning_rate": 1.4738212356799153e-06,
+      "loss": 0.3833,
+      "step": 10898
+    },
+    {
+      "epoch": 0.7567698930704069,
+      "grad_norm": 4.88867367738557,
+      "learning_rate": 1.4730240938625251e-06,
+      "loss": 0.6082,
+      "step": 10899
+    },
+    {
+      "epoch": 0.7568393278711291,
+      "grad_norm": 4.467608551966727,
+      "learning_rate": 1.4722271304355884e-06,
+      "loss": 0.3547,
+      "step": 10900
+    },
+    {
+      "epoch": 0.7569087626718511,
+      "grad_norm": 3.9899162358376965,
+      "learning_rate": 1.4714303454394103e-06,
+      "loss": 0.583,
+      "step": 10901
+    },
+    {
+      "epoch": 0.7569781974725732,
+      "grad_norm": 3.354183586815086,
+      "learning_rate": 1.4706337389142905e-06,
+      "loss": 0.2078,
+      "step": 10902
+    },
+    {
+      "epoch": 0.7570476322732954,
+      "grad_norm": 3.9184933082372346,
+      "learning_rate": 1.4698373109005253e-06,
+      "loss": 0.4165,
+      "step": 10903
+    },
+    {
+      "epoch": 0.7571170670740175,
+      "grad_norm": 3.8744722239441667,
+      "learning_rate": 1.4690410614383914e-06,
+      "loss": 0.4644,
+      "step": 10904
+    },
+    {
+      "epoch": 0.7571865018747396,
+      "grad_norm": 4.426706589796809,
+      "learning_rate": 1.4682449905681672e-06,
+      "loss": 0.4417,
+      "step": 10905
+    },
+    {
+      "epoch": 0.7572559366754618,
+      "grad_norm": 8.27906077644034,
+      "learning_rate": 1.4674490983301143e-06,
+      "loss": 0.8201,
+      "step": 10906
+    },
+    {
+      "epoch": 0.7573253714761838,
+      "grad_norm": 4.676584687506646,
+      "learning_rate": 1.4666533847644887e-06,
+      "loss": 0.5956,
+      "step": 10907
+    },
+    {
+      "epoch": 0.757394806276906,
+      "grad_norm": 3.078449690980452,
+      "learning_rate": 1.4658578499115372e-06,
+      "loss": 0.3931,
+      "step": 10908
+    },
+    {
+      "epoch": 0.7574642410776281,
+      "grad_norm": 3.2204840088878997,
+      "learning_rate": 1.465062493811496e-06,
+      "loss": 0.4068,
+      "step": 10909
+    },
+    {
+      "epoch": 0.7575336758783502,
+      "grad_norm": 4.318542502859666,
+      "learning_rate": 1.464267316504594e-06,
+      "loss": 0.4775,
+      "step": 10910
+    },
+    {
+      "epoch": 0.7576031106790724,
+      "grad_norm": 3.256763616788462,
+      "learning_rate": 1.4634723180310495e-06,
+      "loss": 0.2183,
+      "step": 10911
+    },
+    {
+      "epoch": 0.7576725454797945,
+      "grad_norm": 3.5248678420744097,
+      "learning_rate": 1.4626774984310742e-06,
+      "loss": 0.3286,
+      "step": 10912
+    },
+    {
+      "epoch": 0.7577419802805166,
+      "grad_norm": 3.7711773436452263,
+      "learning_rate": 1.4618828577448663e-06,
+      "loss": 0.4209,
+      "step": 10913
+    },
+    {
+      "epoch": 0.7578114150812387,
+      "grad_norm": 4.72696928367763,
+      "learning_rate": 1.4610883960126232e-06,
+      "loss": 0.3767,
+      "step": 10914
+    },
+    {
+      "epoch": 0.7578808498819608,
+      "grad_norm": 4.808793829319325,
+      "learning_rate": 1.4602941132745225e-06,
+      "loss": 0.5401,
+      "step": 10915
+    },
+    {
+      "epoch": 0.757950284682683,
+      "grad_norm": 3.615677660030094,
+      "learning_rate": 1.459500009570739e-06,
+      "loss": 0.3157,
+      "step": 10916
+    },
+    {
+      "epoch": 0.7580197194834051,
+      "grad_norm": 3.7737435773844057,
+      "learning_rate": 1.4587060849414415e-06,
+      "loss": 0.3845,
+      "step": 10917
+    },
+    {
+      "epoch": 0.7580891542841272,
+      "grad_norm": 3.766402120036325,
+      "learning_rate": 1.4579123394267814e-06,
+      "loss": 0.5296,
+      "step": 10918
+    },
+    {
+      "epoch": 0.7581585890848493,
+      "grad_norm": 3.1311428156071432,
+      "learning_rate": 1.4571187730669057e-06,
+      "loss": 0.2308,
+      "step": 10919
+    },
+    {
+      "epoch": 0.7582280238855714,
+      "grad_norm": 3.835725771422906,
+      "learning_rate": 1.4563253859019544e-06,
+      "loss": 0.505,
+      "step": 10920
+    },
+    {
+      "epoch": 0.7582974586862936,
+      "grad_norm": 3.146939907260934,
+      "learning_rate": 1.4555321779720571e-06,
+      "loss": 0.3194,
+      "step": 10921
+    },
+    {
+      "epoch": 0.7583668934870157,
+      "grad_norm": 3.9043515776473408,
+      "learning_rate": 1.4547391493173285e-06,
+      "loss": 0.268,
+      "step": 10922
+    },
+    {
+      "epoch": 0.7584363282877378,
+      "grad_norm": 3.8527789628750364,
+      "learning_rate": 1.4539462999778831e-06,
+      "loss": 0.4096,
+      "step": 10923
+    },
+    {
+      "epoch": 0.75850576308846,
+      "grad_norm": 3.414074510647886,
+      "learning_rate": 1.4531536299938216e-06,
+      "loss": 0.4134,
+      "step": 10924
+    },
+    {
+      "epoch": 0.758575197889182,
+      "grad_norm": 3.1377960132999143,
+      "learning_rate": 1.4523611394052356e-06,
+      "loss": 0.3202,
+      "step": 10925
+    },
+    {
+      "epoch": 0.7586446326899042,
+      "grad_norm": 4.069910720808549,
+      "learning_rate": 1.4515688282522094e-06,
+      "loss": 0.347,
+      "step": 10926
+    },
+    {
+      "epoch": 0.7587140674906263,
+      "grad_norm": 3.8571936785460585,
+      "learning_rate": 1.450776696574816e-06,
+      "loss": 0.4382,
+      "step": 10927
+    },
+    {
+      "epoch": 0.7587835022913484,
+      "grad_norm": 4.643614880634198,
+      "learning_rate": 1.449984744413121e-06,
+      "loss": 0.6005,
+      "step": 10928
+    },
+    {
+      "epoch": 0.7588529370920706,
+      "grad_norm": 3.5938137935970578,
+      "learning_rate": 1.4491929718071807e-06,
+      "loss": 0.1946,
+      "step": 10929
+    },
+    {
+      "epoch": 0.7589223718927927,
+      "grad_norm": 4.00247687648354,
+      "learning_rate": 1.4484013787970418e-06,
+      "loss": 0.5941,
+      "step": 10930
+    },
+    {
+      "epoch": 0.7589918066935147,
+      "grad_norm": 3.8574954815912497,
+      "learning_rate": 1.447609965422741e-06,
+      "loss": 0.3641,
+      "step": 10931
+    },
+    {
+      "epoch": 0.7590612414942369,
+      "grad_norm": 5.037720016603001,
+      "learning_rate": 1.446818731724311e-06,
+      "loss": 0.7039,
+      "step": 10932
+    },
+    {
+      "epoch": 0.759130676294959,
+      "grad_norm": 4.240567145618297,
+      "learning_rate": 1.446027677741766e-06,
+      "loss": 0.3219,
+      "step": 10933
+    },
+    {
+      "epoch": 0.7592001110956812,
+      "grad_norm": 3.4015416361454283,
+      "learning_rate": 1.445236803515121e-06,
+      "loss": 0.2688,
+      "step": 10934
+    },
+    {
+      "epoch": 0.7592695458964033,
+      "grad_norm": 5.384306442141966,
+      "learning_rate": 1.4444461090843776e-06,
+      "loss": 0.5274,
+      "step": 10935
+    },
+    {
+      "epoch": 0.7593389806971254,
+      "grad_norm": 4.4560760267245225,
+      "learning_rate": 1.4436555944895237e-06,
+      "loss": 0.3908,
+      "step": 10936
+    },
+    {
+      "epoch": 0.7594084154978475,
+      "grad_norm": 5.138149924464828,
+      "learning_rate": 1.4428652597705468e-06,
+      "loss": 0.735,
+      "step": 10937
+    },
+    {
+      "epoch": 0.7594778502985696,
+      "grad_norm": 4.339409895962606,
+      "learning_rate": 1.4420751049674197e-06,
+      "loss": 0.5518,
+      "step": 10938
+    },
+    {
+      "epoch": 0.7595472850992918,
+      "grad_norm": 3.645186962249192,
+      "learning_rate": 1.4412851301201081e-06,
+      "loss": 0.4549,
+      "step": 10939
+    },
+    {
+      "epoch": 0.7596167199000139,
+      "grad_norm": 3.33934283139182,
+      "learning_rate": 1.4404953352685674e-06,
+      "loss": 0.28,
+      "step": 10940
+    },
+    {
+      "epoch": 0.759686154700736,
+      "grad_norm": 3.1084539357089267,
+      "learning_rate": 1.4397057204527454e-06,
+      "loss": 0.307,
+      "step": 10941
+    },
+    {
+      "epoch": 0.7597555895014582,
+      "grad_norm": 3.5363094914532818,
+      "learning_rate": 1.4389162857125787e-06,
+      "loss": 0.3696,
+      "step": 10942
+    },
+    {
+      "epoch": 0.7598250243021802,
+      "grad_norm": 4.544872836449702,
+      "learning_rate": 1.438127031087997e-06,
+      "loss": 0.6233,
+      "step": 10943
+    },
+    {
+      "epoch": 0.7598944591029023,
+      "grad_norm": 4.309297769849216,
+      "learning_rate": 1.4373379566189187e-06,
+      "loss": 0.3705,
+      "step": 10944
+    },
+    {
+      "epoch": 0.7599638939036245,
+      "grad_norm": 4.94856867090616,
+      "learning_rate": 1.4365490623452545e-06,
+      "loss": 0.6321,
+      "step": 10945
+    },
+    {
+      "epoch": 0.7600333287043466,
+      "grad_norm": 4.76191566096668,
+      "learning_rate": 1.4357603483069093e-06,
+      "loss": 0.6405,
+      "step": 10946
+    },
+    {
+      "epoch": 0.7601027635050688,
+      "grad_norm": 4.019769199876363,
+      "learning_rate": 1.4349718145437709e-06,
+      "loss": 0.5488,
+      "step": 10947
+    },
+    {
+      "epoch": 0.7601721983057909,
+      "grad_norm": 2.8543690064401614,
+      "learning_rate": 1.4341834610957223e-06,
+      "loss": 0.2474,
+      "step": 10948
+    },
+    {
+      "epoch": 0.7602416331065129,
+      "grad_norm": 3.620728182205127,
+      "learning_rate": 1.4333952880026408e-06,
+      "loss": 0.3312,
+      "step": 10949
+    },
+    {
+      "epoch": 0.7603110679072351,
+      "grad_norm": 3.700413527920763,
+      "learning_rate": 1.4326072953043912e-06,
+      "loss": 0.2188,
+      "step": 10950
+    },
+    {
+      "epoch": 0.7603805027079572,
+      "grad_norm": 3.5941979097540813,
+      "learning_rate": 1.431819483040825e-06,
+      "loss": 0.5416,
+      "step": 10951
+    },
+    {
+      "epoch": 0.7604499375086794,
+      "grad_norm": 4.388212391367575,
+      "learning_rate": 1.4310318512517928e-06,
+      "loss": 0.6015,
+      "step": 10952
+    },
+    {
+      "epoch": 0.7605193723094015,
+      "grad_norm": 4.9117577966896,
+      "learning_rate": 1.430244399977133e-06,
+      "loss": 0.6144,
+      "step": 10953
+    },
+    {
+      "epoch": 0.7605888071101236,
+      "grad_norm": 4.01610718792258,
+      "learning_rate": 1.4294571292566688e-06,
+      "loss": 0.5142,
+      "step": 10954
+    },
+    {
+      "epoch": 0.7606582419108457,
+      "grad_norm": 3.8762706093362485,
+      "learning_rate": 1.4286700391302244e-06,
+      "loss": 0.6493,
+      "step": 10955
+    },
+    {
+      "epoch": 0.7607276767115678,
+      "grad_norm": 3.423684601803941,
+      "learning_rate": 1.4278831296376079e-06,
+      "loss": 0.2988,
+      "step": 10956
+    },
+    {
+      "epoch": 0.76079711151229,
+      "grad_norm": 3.8209232494697862,
+      "learning_rate": 1.4270964008186206e-06,
+      "loss": 0.3774,
+      "step": 10957
+    },
+    {
+      "epoch": 0.7608665463130121,
+      "grad_norm": 3.3411175676014473,
+      "learning_rate": 1.4263098527130547e-06,
+      "loss": 0.391,
+      "step": 10958
+    },
+    {
+      "epoch": 0.7609359811137342,
+      "grad_norm": 3.727050047535712,
+      "learning_rate": 1.425523485360692e-06,
+      "loss": 0.3562,
+      "step": 10959
+    },
+    {
+      "epoch": 0.7610054159144564,
+      "grad_norm": 2.9857723148221487,
+      "learning_rate": 1.424737298801307e-06,
+      "loss": 0.3153,
+      "step": 10960
+    },
+    {
+      "epoch": 0.7610748507151784,
+      "grad_norm": 3.5180359478289884,
+      "learning_rate": 1.4239512930746635e-06,
+      "loss": 0.4533,
+      "step": 10961
+    },
+    {
+      "epoch": 0.7611442855159005,
+      "grad_norm": 3.7867690947130237,
+      "learning_rate": 1.423165468220517e-06,
+      "loss": 0.4351,
+      "step": 10962
+    },
+    {
+      "epoch": 0.7612137203166227,
+      "grad_norm": 3.8978401723867875,
+      "learning_rate": 1.4223798242786124e-06,
+      "loss": 0.604,
+      "step": 10963
+    },
+    {
+      "epoch": 0.7612831551173448,
+      "grad_norm": 4.558177344395122,
+      "learning_rate": 1.4215943612886907e-06,
+      "loss": 0.5644,
+      "step": 10964
+    },
+    {
+      "epoch": 0.761352589918067,
+      "grad_norm": 3.8463845879042147,
+      "learning_rate": 1.4208090792904743e-06,
+      "loss": 0.2956,
+      "step": 10965
+    },
+    {
+      "epoch": 0.7614220247187891,
+      "grad_norm": 4.608890113536634,
+      "learning_rate": 1.4200239783236857e-06,
+      "loss": 0.3729,
+      "step": 10966
+    },
+    {
+      "epoch": 0.7614914595195111,
+      "grad_norm": 4.031550935031122,
+      "learning_rate": 1.4192390584280347e-06,
+      "loss": 0.4503,
+      "step": 10967
+    },
+    {
+      "epoch": 0.7615608943202333,
+      "grad_norm": 3.464828254085752,
+      "learning_rate": 1.4184543196432176e-06,
+      "loss": 0.2297,
+      "step": 10968
+    },
+    {
+      "epoch": 0.7616303291209554,
+      "grad_norm": 2.9174019812399616,
+      "learning_rate": 1.4176697620089296e-06,
+      "loss": 0.1686,
+      "step": 10969
+    },
+    {
+      "epoch": 0.7616997639216776,
+      "grad_norm": 3.6217188116511307,
+      "learning_rate": 1.4168853855648513e-06,
+      "loss": 0.4203,
+      "step": 10970
+    },
+    {
+      "epoch": 0.7617691987223997,
+      "grad_norm": 3.5868076564151035,
+      "learning_rate": 1.4161011903506555e-06,
+      "loss": 0.2856,
+      "step": 10971
+    },
+    {
+      "epoch": 0.7618386335231218,
+      "grad_norm": 2.9579709330508863,
+      "learning_rate": 1.4153171764060058e-06,
+      "loss": 0.3851,
+      "step": 10972
+    },
+    {
+      "epoch": 0.761908068323844,
+      "grad_norm": 4.272188936418521,
+      "learning_rate": 1.4145333437705576e-06,
+      "loss": 0.384,
+      "step": 10973
+    },
+    {
+      "epoch": 0.761977503124566,
+      "grad_norm": 3.993314028996715,
+      "learning_rate": 1.4137496924839556e-06,
+      "loss": 0.3477,
+      "step": 10974
+    },
+    {
+      "epoch": 0.7620469379252881,
+      "grad_norm": 4.901878263879728,
+      "learning_rate": 1.4129662225858354e-06,
+      "loss": 0.3346,
+      "step": 10975
+    },
+    {
+      "epoch": 0.7621163727260103,
+      "grad_norm": 9.646807348821858,
+      "learning_rate": 1.412182934115825e-06,
+      "loss": 0.4908,
+      "step": 10976
+    },
+    {
+      "epoch": 0.7621858075267324,
+      "grad_norm": 4.6339626309225235,
+      "learning_rate": 1.4113998271135404e-06,
+      "loss": 0.5413,
+      "step": 10977
+    },
+    {
+      "epoch": 0.7622552423274546,
+      "grad_norm": 3.7782081073902036,
+      "learning_rate": 1.410616901618595e-06,
+      "loss": 0.2719,
+      "step": 10978
+    },
+    {
+      "epoch": 0.7623246771281766,
+      "grad_norm": 3.51054444337811,
+      "learning_rate": 1.4098341576705832e-06,
+      "loss": 0.2677,
+      "step": 10979
+    },
+    {
+      "epoch": 0.7623941119288987,
+      "grad_norm": 3.9557926383295006,
+      "learning_rate": 1.4090515953090961e-06,
+      "loss": 0.4163,
+      "step": 10980
+    },
+    {
+      "epoch": 0.7624635467296209,
+      "grad_norm": 3.7247359613431055,
+      "learning_rate": 1.408269214573717e-06,
+      "loss": 0.4653,
+      "step": 10981
+    },
+    {
+      "epoch": 0.762532981530343,
+      "grad_norm": 3.3624363950277107,
+      "learning_rate": 1.4074870155040182e-06,
+      "loss": 0.3879,
+      "step": 10982
+    },
+    {
+      "epoch": 0.7626024163310652,
+      "grad_norm": 4.19338304593881,
+      "learning_rate": 1.4067049981395581e-06,
+      "loss": 0.5461,
+      "step": 10983
+    },
+    {
+      "epoch": 0.7626718511317873,
+      "grad_norm": 4.935492317679381,
+      "learning_rate": 1.405923162519895e-06,
+      "loss": 0.5417,
+      "step": 10984
+    },
+    {
+      "epoch": 0.7627412859325093,
+      "grad_norm": 4.6064147495218615,
+      "learning_rate": 1.4051415086845705e-06,
+      "loss": 0.5944,
+      "step": 10985
+    },
+    {
+      "epoch": 0.7628107207332315,
+      "grad_norm": 4.2376599290643595,
+      "learning_rate": 1.4043600366731213e-06,
+      "loss": 0.4908,
+      "step": 10986
+    },
+    {
+      "epoch": 0.7628801555339536,
+      "grad_norm": 4.751898471858316,
+      "learning_rate": 1.403578746525072e-06,
+      "loss": 0.5594,
+      "step": 10987
+    },
+    {
+      "epoch": 0.7629495903346757,
+      "grad_norm": 6.6423669602073465,
+      "learning_rate": 1.4027976382799408e-06,
+      "loss": 0.5553,
+      "step": 10988
+    },
+    {
+      "epoch": 0.7630190251353979,
+      "grad_norm": 4.8269199077628295,
+      "learning_rate": 1.4020167119772342e-06,
+      "loss": 0.7402,
+      "step": 10989
+    },
+    {
+      "epoch": 0.76308845993612,
+      "grad_norm": 4.152973773517244,
+      "learning_rate": 1.4012359676564513e-06,
+      "loss": 0.5146,
+      "step": 10990
+    },
+    {
+      "epoch": 0.7631578947368421,
+      "grad_norm": 3.8811602090113273,
+      "learning_rate": 1.4004554053570806e-06,
+      "loss": 0.3829,
+      "step": 10991
+    },
+    {
+      "epoch": 0.7632273295375642,
+      "grad_norm": 3.2270909585437995,
+      "learning_rate": 1.3996750251186025e-06,
+      "loss": 0.3147,
+      "step": 10992
+    },
+    {
+      "epoch": 0.7632967643382863,
+      "grad_norm": 3.709802856407118,
+      "learning_rate": 1.398894826980487e-06,
+      "loss": 0.3527,
+      "step": 10993
+    },
+    {
+      "epoch": 0.7633661991390085,
+      "grad_norm": 4.958090947198897,
+      "learning_rate": 1.3981148109821952e-06,
+      "loss": 0.4774,
+      "step": 10994
+    },
+    {
+      "epoch": 0.7634356339397306,
+      "grad_norm": 3.6044344804891026,
+      "learning_rate": 1.3973349771631823e-06,
+      "loss": 0.428,
+      "step": 10995
+    },
+    {
+      "epoch": 0.7635050687404528,
+      "grad_norm": 4.240270993438797,
+      "learning_rate": 1.3965553255628906e-06,
+      "loss": 0.5392,
+      "step": 10996
+    },
+    {
+      "epoch": 0.7635745035411748,
+      "grad_norm": 4.684313099285424,
+      "learning_rate": 1.3957758562207496e-06,
+      "loss": 0.6308,
+      "step": 10997
+    },
+    {
+      "epoch": 0.7636439383418969,
+      "grad_norm": 3.805792597752898,
+      "learning_rate": 1.3949965691761896e-06,
+      "loss": 0.4469,
+      "step": 10998
+    },
+    {
+      "epoch": 0.7637133731426191,
+      "grad_norm": 3.405494520488374,
+      "learning_rate": 1.3942174644686245e-06,
+      "loss": 0.2651,
+      "step": 10999
+    },
+    {
+      "epoch": 0.7637828079433412,
+      "grad_norm": 3.4067240180800984,
+      "learning_rate": 1.3934385421374575e-06,
+      "loss": 0.3343,
+      "step": 11000
+    },
+    {
+      "epoch": 0.7638522427440633,
+      "grad_norm": 3.3445573379671716,
+      "learning_rate": 1.3926598022220894e-06,
+      "loss": 0.3754,
+      "step": 11001
+    },
+    {
+      "epoch": 0.7639216775447855,
+      "grad_norm": 4.245815220116447,
+      "learning_rate": 1.3918812447619062e-06,
+      "loss": 0.4401,
+      "step": 11002
+    },
+    {
+      "epoch": 0.7639911123455075,
+      "grad_norm": 6.6184195002970485,
+      "learning_rate": 1.391102869796287e-06,
+      "loss": 0.5715,
+      "step": 11003
+    },
+    {
+      "epoch": 0.7640605471462297,
+      "grad_norm": 3.911845541428749,
+      "learning_rate": 1.3903246773646006e-06,
+      "loss": 0.5033,
+      "step": 11004
+    },
+    {
+      "epoch": 0.7641299819469518,
+      "grad_norm": 3.713713104821458,
+      "learning_rate": 1.3895466675062074e-06,
+      "loss": 0.447,
+      "step": 11005
+    },
+    {
+      "epoch": 0.7641994167476739,
+      "grad_norm": 2.5811308716934405,
+      "learning_rate": 1.3887688402604571e-06,
+      "loss": 0.2323,
+      "step": 11006
+    },
+    {
+      "epoch": 0.7642688515483961,
+      "grad_norm": 4.733080297858727,
+      "learning_rate": 1.3879911956666952e-06,
+      "loss": 0.3949,
+      "step": 11007
+    },
+    {
+      "epoch": 0.7643382863491182,
+      "grad_norm": 4.176857060839163,
+      "learning_rate": 1.3872137337642505e-06,
+      "loss": 0.4814,
+      "step": 11008
+    },
+    {
+      "epoch": 0.7644077211498403,
+      "grad_norm": 3.8485496966395796,
+      "learning_rate": 1.3864364545924448e-06,
+      "loss": 0.3188,
+      "step": 11009
+    },
+    {
+      "epoch": 0.7644771559505624,
+      "grad_norm": 2.275475647298644,
+      "learning_rate": 1.3856593581905975e-06,
+      "loss": 0.1935,
+      "step": 11010
+    },
+    {
+      "epoch": 0.7645465907512845,
+      "grad_norm": 4.148206732542511,
+      "learning_rate": 1.384882444598008e-06,
+      "loss": 0.6678,
+      "step": 11011
+    },
+    {
+      "epoch": 0.7646160255520067,
+      "grad_norm": 3.11227029777124,
+      "learning_rate": 1.3841057138539726e-06,
+      "loss": 0.3046,
+      "step": 11012
+    },
+    {
+      "epoch": 0.7646854603527288,
+      "grad_norm": 5.434531004613774,
+      "learning_rate": 1.3833291659977798e-06,
+      "loss": 0.3342,
+      "step": 11013
+    },
+    {
+      "epoch": 0.764754895153451,
+      "grad_norm": 3.356218109948482,
+      "learning_rate": 1.3825528010687057e-06,
+      "loss": 0.3896,
+      "step": 11014
+    },
+    {
+      "epoch": 0.764824329954173,
+      "grad_norm": 3.528432579504851,
+      "learning_rate": 1.3817766191060172e-06,
+      "loss": 0.2837,
+      "step": 11015
+    },
+    {
+      "epoch": 0.7648937647548951,
+      "grad_norm": 5.743410417382064,
+      "learning_rate": 1.3810006201489729e-06,
+      "loss": 0.3698,
+      "step": 11016
+    },
+    {
+      "epoch": 0.7649631995556173,
+      "grad_norm": 3.8026511819010755,
+      "learning_rate": 1.380224804236822e-06,
+      "loss": 0.484,
+      "step": 11017
+    },
+    {
+      "epoch": 0.7650326343563394,
+      "grad_norm": 4.279163249706843,
+      "learning_rate": 1.3794491714088048e-06,
+      "loss": 0.5687,
+      "step": 11018
+    },
+    {
+      "epoch": 0.7651020691570615,
+      "grad_norm": 6.633459168956002,
+      "learning_rate": 1.3786737217041513e-06,
+      "loss": 0.961,
+      "step": 11019
+    },
+    {
+      "epoch": 0.7651715039577837,
+      "grad_norm": 3.2612433089063164,
+      "learning_rate": 1.3778984551620833e-06,
+      "loss": 0.3707,
+      "step": 11020
+    },
+    {
+      "epoch": 0.7652409387585057,
+      "grad_norm": 3.4735747204569027,
+      "learning_rate": 1.377123371821813e-06,
+      "loss": 0.3748,
+      "step": 11021
+    },
+    {
+      "epoch": 0.7653103735592279,
+      "grad_norm": 4.027200753699064,
+      "learning_rate": 1.3763484717225422e-06,
+      "loss": 0.4372,
+      "step": 11022
+    },
+    {
+      "epoch": 0.76537980835995,
+      "grad_norm": 3.1978618087549866,
+      "learning_rate": 1.3755737549034647e-06,
+      "loss": 0.352,
+      "step": 11023
+    },
+    {
+      "epoch": 0.7654492431606721,
+      "grad_norm": 3.6215600884610044,
+      "learning_rate": 1.3747992214037675e-06,
+      "loss": 0.4578,
+      "step": 11024
+    },
+    {
+      "epoch": 0.7655186779613943,
+      "grad_norm": 4.708609023443965,
+      "learning_rate": 1.3740248712626226e-06,
+      "loss": 0.5726,
+      "step": 11025
+    },
+    {
+      "epoch": 0.7655881127621164,
+      "grad_norm": 3.422983561562488,
+      "learning_rate": 1.3732507045191951e-06,
+      "loss": 0.2419,
+      "step": 11026
+    },
+    {
+      "epoch": 0.7656575475628385,
+      "grad_norm": 3.1951605708713138,
+      "learning_rate": 1.3724767212126444e-06,
+      "loss": 0.2084,
+      "step": 11027
+    },
+    {
+      "epoch": 0.7657269823635606,
+      "grad_norm": 3.674804002846155,
+      "learning_rate": 1.3717029213821176e-06,
+      "loss": 0.3337,
+      "step": 11028
+    },
+    {
+      "epoch": 0.7657964171642827,
+      "grad_norm": 3.500491709275867,
+      "learning_rate": 1.3709293050667487e-06,
+      "loss": 0.5162,
+      "step": 11029
+    },
+    {
+      "epoch": 0.7658658519650049,
+      "grad_norm": 3.8004001702432078,
+      "learning_rate": 1.37015587230567e-06,
+      "loss": 0.1409,
+      "step": 11030
+    },
+    {
+      "epoch": 0.765935286765727,
+      "grad_norm": 3.100742469697651,
+      "learning_rate": 1.3693826231380002e-06,
+      "loss": 0.2833,
+      "step": 11031
+    },
+    {
+      "epoch": 0.766004721566449,
+      "grad_norm": 3.8323918660038174,
+      "learning_rate": 1.3686095576028492e-06,
+      "loss": 0.3172,
+      "step": 11032
+    },
+    {
+      "epoch": 0.7660741563671712,
+      "grad_norm": 4.170177234340749,
+      "learning_rate": 1.3678366757393175e-06,
+      "loss": 0.592,
+      "step": 11033
+    },
+    {
+      "epoch": 0.7661435911678933,
+      "grad_norm": 5.603019246987314,
+      "learning_rate": 1.3670639775864963e-06,
+      "loss": 0.7365,
+      "step": 11034
+    },
+    {
+      "epoch": 0.7662130259686155,
+      "grad_norm": 2.3899431905811177,
+      "learning_rate": 1.3662914631834684e-06,
+      "loss": 0.1378,
+      "step": 11035
+    },
+    {
+      "epoch": 0.7662824607693376,
+      "grad_norm": 4.210255953075779,
+      "learning_rate": 1.3655191325693061e-06,
+      "loss": 0.5173,
+      "step": 11036
+    },
+    {
+      "epoch": 0.7663518955700597,
+      "grad_norm": 3.7388554398164286,
+      "learning_rate": 1.3647469857830736e-06,
+      "loss": 0.2854,
+      "step": 11037
+    },
+    {
+      "epoch": 0.7664213303707819,
+      "grad_norm": 3.541205578750305,
+      "learning_rate": 1.3639750228638237e-06,
+      "loss": 0.3633,
+      "step": 11038
+    },
+    {
+      "epoch": 0.7664907651715039,
+      "grad_norm": 2.769618483680573,
+      "learning_rate": 1.3632032438506059e-06,
+      "loss": 0.2469,
+      "step": 11039
+    },
+    {
+      "epoch": 0.7665601999722261,
+      "grad_norm": 4.095283652497651,
+      "learning_rate": 1.3624316487824508e-06,
+      "loss": 0.3301,
+      "step": 11040
+    },
+    {
+      "epoch": 0.7666296347729482,
+      "grad_norm": 3.566725008994478,
+      "learning_rate": 1.3616602376983851e-06,
+      "loss": 0.29,
+      "step": 11041
+    },
+    {
+      "epoch": 0.7666990695736703,
+      "grad_norm": 3.884235960206201,
+      "learning_rate": 1.3608890106374312e-06,
+      "loss": 0.3209,
+      "step": 11042
+    },
+    {
+      "epoch": 0.7667685043743925,
+      "grad_norm": 4.252279050155112,
+      "learning_rate": 1.3601179676385912e-06,
+      "loss": 0.4251,
+      "step": 11043
+    },
+    {
+      "epoch": 0.7668379391751146,
+      "grad_norm": 3.5623754911128684,
+      "learning_rate": 1.3593471087408638e-06,
+      "loss": 0.5348,
+      "step": 11044
+    },
+    {
+      "epoch": 0.7669073739758366,
+      "grad_norm": 3.2175435326464457,
+      "learning_rate": 1.358576433983242e-06,
+      "loss": 0.285,
+      "step": 11045
+    },
+    {
+      "epoch": 0.7669768087765588,
+      "grad_norm": 2.4555681251078836,
+      "learning_rate": 1.3578059434047037e-06,
+      "loss": 0.3022,
+      "step": 11046
+    },
+    {
+      "epoch": 0.7670462435772809,
+      "grad_norm": 3.659576030922715,
+      "learning_rate": 1.357035637044219e-06,
+      "loss": 0.3639,
+      "step": 11047
+    },
+    {
+      "epoch": 0.7671156783780031,
+      "grad_norm": 3.8291064089381917,
+      "learning_rate": 1.3562655149407494e-06,
+      "loss": 0.3501,
+      "step": 11048
+    },
+    {
+      "epoch": 0.7671851131787252,
+      "grad_norm": 3.5446508387418603,
+      "learning_rate": 1.3554955771332473e-06,
+      "loss": 0.3788,
+      "step": 11049
+    },
+    {
+      "epoch": 0.7672545479794473,
+      "grad_norm": 3.460828445742911,
+      "learning_rate": 1.3547258236606542e-06,
+      "loss": 0.2884,
+      "step": 11050
+    },
+    {
+      "epoch": 0.7673239827801694,
+      "grad_norm": 4.003112816465432,
+      "learning_rate": 1.3539562545619045e-06,
+      "loss": 0.5391,
+      "step": 11051
+    },
+    {
+      "epoch": 0.7673934175808915,
+      "grad_norm": 4.526404952861863,
+      "learning_rate": 1.353186869875922e-06,
+      "loss": 0.5574,
+      "step": 11052
+    },
+    {
+      "epoch": 0.7674628523816137,
+      "grad_norm": 3.872098341102574,
+      "learning_rate": 1.3524176696416203e-06,
+      "loss": 0.3986,
+      "step": 11053
+    },
+    {
+      "epoch": 0.7675322871823358,
+      "grad_norm": 5.311546058713373,
+      "learning_rate": 1.3516486538979056e-06,
+      "loss": 0.7617,
+      "step": 11054
+    },
+    {
+      "epoch": 0.7676017219830579,
+      "grad_norm": 6.528486560143228,
+      "learning_rate": 1.3508798226836722e-06,
+      "loss": 0.4809,
+      "step": 11055
+    },
+    {
+      "epoch": 0.7676711567837801,
+      "grad_norm": 5.197957867697868,
+      "learning_rate": 1.35011117603781e-06,
+      "loss": 0.6618,
+      "step": 11056
+    },
+    {
+      "epoch": 0.7677405915845021,
+      "grad_norm": 3.8128046904813013,
+      "learning_rate": 1.3493427139991954e-06,
+      "loss": 0.3211,
+      "step": 11057
+    },
+    {
+      "epoch": 0.7678100263852242,
+      "grad_norm": 3.6100193703705634,
+      "learning_rate": 1.3485744366066921e-06,
+      "loss": 0.3594,
+      "step": 11058
+    },
+    {
+      "epoch": 0.7678794611859464,
+      "grad_norm": 3.466134751811614,
+      "learning_rate": 1.347806343899164e-06,
+      "loss": 0.2964,
+      "step": 11059
+    },
+    {
+      "epoch": 0.7679488959866685,
+      "grad_norm": 3.8921848312473837,
+      "learning_rate": 1.3470384359154592e-06,
+      "loss": 0.2056,
+      "step": 11060
+    },
+    {
+      "epoch": 0.7680183307873907,
+      "grad_norm": 3.4611479870315858,
+      "learning_rate": 1.3462707126944141e-06,
+      "loss": 0.3876,
+      "step": 11061
+    },
+    {
+      "epoch": 0.7680877655881128,
+      "grad_norm": 4.452158673833338,
+      "learning_rate": 1.3455031742748632e-06,
+      "loss": 0.4647,
+      "step": 11062
+    },
+    {
+      "epoch": 0.7681572003888348,
+      "grad_norm": 4.448376828207619,
+      "learning_rate": 1.3447358206956268e-06,
+      "loss": 0.4656,
+      "step": 11063
+    },
+    {
+      "epoch": 0.768226635189557,
+      "grad_norm": 3.8963970791603826,
+      "learning_rate": 1.3439686519955158e-06,
+      "loss": 0.4096,
+      "step": 11064
+    },
+    {
+      "epoch": 0.7682960699902791,
+      "grad_norm": 4.054021194060909,
+      "learning_rate": 1.3432016682133337e-06,
+      "loss": 0.3679,
+      "step": 11065
+    },
+    {
+      "epoch": 0.7683655047910013,
+      "grad_norm": 4.239365275354775,
+      "learning_rate": 1.3424348693878731e-06,
+      "loss": 0.5044,
+      "step": 11066
+    },
+    {
+      "epoch": 0.7684349395917234,
+      "grad_norm": 3.2482667266524343,
+      "learning_rate": 1.341668255557918e-06,
+      "loss": 0.3654,
+      "step": 11067
+    },
+    {
+      "epoch": 0.7685043743924455,
+      "grad_norm": 3.2766740786919955,
+      "learning_rate": 1.340901826762243e-06,
+      "loss": 0.218,
+      "step": 11068
+    },
+    {
+      "epoch": 0.7685738091931676,
+      "grad_norm": 4.228232124209404,
+      "learning_rate": 1.3401355830396129e-06,
+      "loss": 0.4182,
+      "step": 11069
+    },
+    {
+      "epoch": 0.7686432439938897,
+      "grad_norm": 3.75177060764092,
+      "learning_rate": 1.3393695244287819e-06,
+      "loss": 0.3856,
+      "step": 11070
+    },
+    {
+      "epoch": 0.7687126787946119,
+      "grad_norm": 4.640480166162068,
+      "learning_rate": 1.3386036509685014e-06,
+      "loss": 0.5131,
+      "step": 11071
+    },
+    {
+      "epoch": 0.768782113595334,
+      "grad_norm": 4.601072141059501,
+      "learning_rate": 1.3378379626975035e-06,
+      "loss": 0.4706,
+      "step": 11072
+    },
+    {
+      "epoch": 0.7688515483960561,
+      "grad_norm": 3.734949047524425,
+      "learning_rate": 1.337072459654516e-06,
+      "loss": 0.3289,
+      "step": 11073
+    },
+    {
+      "epoch": 0.7689209831967783,
+      "grad_norm": 2.981265161101862,
+      "learning_rate": 1.3363071418782613e-06,
+      "loss": 0.2877,
+      "step": 11074
+    },
+    {
+      "epoch": 0.7689904179975003,
+      "grad_norm": 3.8250752708045734,
+      "learning_rate": 1.3355420094074427e-06,
+      "loss": 0.3943,
+      "step": 11075
+    },
+    {
+      "epoch": 0.7690598527982224,
+      "grad_norm": 2.926501179191031,
+      "learning_rate": 1.3347770622807649e-06,
+      "loss": 0.2121,
+      "step": 11076
+    },
+    {
+      "epoch": 0.7691292875989446,
+      "grad_norm": 4.045149367083221,
+      "learning_rate": 1.334012300536915e-06,
+      "loss": 0.4724,
+      "step": 11077
+    },
+    {
+      "epoch": 0.7691987223996667,
+      "grad_norm": 4.478289337564379,
+      "learning_rate": 1.3332477242145747e-06,
+      "loss": 0.568,
+      "step": 11078
+    },
+    {
+      "epoch": 0.7692681572003889,
+      "grad_norm": 3.742266215722157,
+      "learning_rate": 1.3324833333524156e-06,
+      "loss": 0.4501,
+      "step": 11079
+    },
+    {
+      "epoch": 0.769337592001111,
+      "grad_norm": 4.239586890367387,
+      "learning_rate": 1.331719127989099e-06,
+      "loss": 0.5153,
+      "step": 11080
+    },
+    {
+      "epoch": 0.769407026801833,
+      "grad_norm": 3.0390562303454,
+      "learning_rate": 1.3309551081632781e-06,
+      "loss": 0.1832,
+      "step": 11081
+    },
+    {
+      "epoch": 0.7694764616025552,
+      "grad_norm": 4.395028641237015,
+      "learning_rate": 1.3301912739135964e-06,
+      "loss": 0.4599,
+      "step": 11082
+    },
+    {
+      "epoch": 0.7695458964032773,
+      "grad_norm": 3.8206407649689247,
+      "learning_rate": 1.329427625278687e-06,
+      "loss": 0.3166,
+      "step": 11083
+    },
+    {
+      "epoch": 0.7696153312039995,
+      "grad_norm": 2.7696648377365842,
+      "learning_rate": 1.328664162297173e-06,
+      "loss": 0.2558,
+      "step": 11084
+    },
+    {
+      "epoch": 0.7696847660047216,
+      "grad_norm": 2.975794473150173,
+      "learning_rate": 1.3279008850076747e-06,
+      "loss": 0.321,
+      "step": 11085
+    },
+    {
+      "epoch": 0.7697542008054437,
+      "grad_norm": 3.705988267959164,
+      "learning_rate": 1.3271377934487927e-06,
+      "loss": 0.426,
+      "step": 11086
+    },
+    {
+      "epoch": 0.7698236356061658,
+      "grad_norm": 3.004881050812452,
+      "learning_rate": 1.3263748876591237e-06,
+      "loss": 0.2529,
+      "step": 11087
+    },
+    {
+      "epoch": 0.7698930704068879,
+      "grad_norm": 6.690827262559391,
+      "learning_rate": 1.3256121676772577e-06,
+      "loss": 0.5149,
+      "step": 11088
+    },
+    {
+      "epoch": 0.76996250520761,
+      "grad_norm": 3.387881168443318,
+      "learning_rate": 1.3248496335417715e-06,
+      "loss": 0.3394,
+      "step": 11089
+    },
+    {
+      "epoch": 0.7700319400083322,
+      "grad_norm": 11.182716191377102,
+      "learning_rate": 1.324087285291229e-06,
+      "loss": 0.4477,
+      "step": 11090
+    },
+    {
+      "epoch": 0.7701013748090543,
+      "grad_norm": 4.6528302205606265,
+      "learning_rate": 1.3233251229641937e-06,
+      "loss": 0.782,
+      "step": 11091
+    },
+    {
+      "epoch": 0.7701708096097765,
+      "grad_norm": 3.234439676702835,
+      "learning_rate": 1.3225631465992144e-06,
+      "loss": 0.3717,
+      "step": 11092
+    },
+    {
+      "epoch": 0.7702402444104985,
+      "grad_norm": 3.9225629523835486,
+      "learning_rate": 1.3218013562348274e-06,
+      "loss": 0.3742,
+      "step": 11093
+    },
+    {
+      "epoch": 0.7703096792112206,
+      "grad_norm": 3.719603057221862,
+      "learning_rate": 1.3210397519095669e-06,
+      "loss": 0.3364,
+      "step": 11094
+    },
+    {
+      "epoch": 0.7703791140119428,
+      "grad_norm": 3.7118041286228487,
+      "learning_rate": 1.3202783336619528e-06,
+      "loss": 0.2746,
+      "step": 11095
+    },
+    {
+      "epoch": 0.7704485488126649,
+      "grad_norm": 4.699558722200026,
+      "learning_rate": 1.3195171015304964e-06,
+      "loss": 0.3835,
+      "step": 11096
+    },
+    {
+      "epoch": 0.7705179836133871,
+      "grad_norm": 4.531947981971148,
+      "learning_rate": 1.3187560555537005e-06,
+      "loss": 0.4333,
+      "step": 11097
+    },
+    {
+      "epoch": 0.7705874184141092,
+      "grad_norm": 3.748795210373505,
+      "learning_rate": 1.3179951957700577e-06,
+      "loss": 0.391,
+      "step": 11098
+    },
+    {
+      "epoch": 0.7706568532148312,
+      "grad_norm": 3.34970461137328,
+      "learning_rate": 1.3172345222180516e-06,
+      "loss": 0.3171,
+      "step": 11099
+    },
+    {
+      "epoch": 0.7707262880155534,
+      "grad_norm": 4.543096327741528,
+      "learning_rate": 1.3164740349361555e-06,
+      "loss": 0.6598,
+      "step": 11100
+    },
+    {
+      "epoch": 0.7707957228162755,
+      "grad_norm": 3.999562499357687,
+      "learning_rate": 1.3157137339628352e-06,
+      "loss": 0.4614,
+      "step": 11101
+    },
+    {
+      "epoch": 0.7708651576169976,
+      "grad_norm": 3.908686521988311,
+      "learning_rate": 1.314953619336543e-06,
+      "loss": 0.4525,
+      "step": 11102
+    },
+    {
+      "epoch": 0.7709345924177198,
+      "grad_norm": 3.963515152662578,
+      "learning_rate": 1.3141936910957303e-06,
+      "loss": 0.3669,
+      "step": 11103
+    },
+    {
+      "epoch": 0.7710040272184419,
+      "grad_norm": 4.739756173807416,
+      "learning_rate": 1.3134339492788267e-06,
+      "loss": 0.5854,
+      "step": 11104
+    },
+    {
+      "epoch": 0.771073462019164,
+      "grad_norm": 4.49243072899559,
+      "learning_rate": 1.3126743939242643e-06,
+      "loss": 0.5688,
+      "step": 11105
+    },
+    {
+      "epoch": 0.7711428968198861,
+      "grad_norm": 2.708406004214198,
+      "learning_rate": 1.3119150250704578e-06,
+      "loss": 0.2601,
+      "step": 11106
+    },
+    {
+      "epoch": 0.7712123316206082,
+      "grad_norm": 5.026015686256312,
+      "learning_rate": 1.3111558427558163e-06,
+      "loss": 0.4407,
+      "step": 11107
+    },
+    {
+      "epoch": 0.7712817664213304,
+      "grad_norm": 5.774790291668631,
+      "learning_rate": 1.3103968470187384e-06,
+      "loss": 0.656,
+      "step": 11108
+    },
+    {
+      "epoch": 0.7713512012220525,
+      "grad_norm": 5.732531113390396,
+      "learning_rate": 1.3096380378976126e-06,
+      "loss": 0.5403,
+      "step": 11109
+    },
+    {
+      "epoch": 0.7714206360227747,
+      "grad_norm": 3.6236871925607743,
+      "learning_rate": 1.3088794154308188e-06,
+      "loss": 0.4528,
+      "step": 11110
+    },
+    {
+      "epoch": 0.7714900708234967,
+      "grad_norm": 3.326469555615002,
+      "learning_rate": 1.3081209796567273e-06,
+      "loss": 0.4517,
+      "step": 11111
+    },
+    {
+      "epoch": 0.7715595056242188,
+      "grad_norm": 3.8946023878743663,
+      "learning_rate": 1.3073627306136994e-06,
+      "loss": 0.2974,
+      "step": 11112
+    },
+    {
+      "epoch": 0.771628940424941,
+      "grad_norm": 4.604630629939227,
+      "learning_rate": 1.3066046683400847e-06,
+      "loss": 0.4481,
+      "step": 11113
+    },
+    {
+      "epoch": 0.7716983752256631,
+      "grad_norm": 4.3383117078592,
+      "learning_rate": 1.305846792874229e-06,
+      "loss": 0.5713,
+      "step": 11114
+    },
+    {
+      "epoch": 0.7717678100263852,
+      "grad_norm": 3.4441635510815214,
+      "learning_rate": 1.305089104254461e-06,
+      "loss": 0.3985,
+      "step": 11115
+    },
+    {
+      "epoch": 0.7718372448271074,
+      "grad_norm": 2.3300318488355924,
+      "learning_rate": 1.3043316025191033e-06,
+      "loss": 0.272,
+      "step": 11116
+    },
+    {
+      "epoch": 0.7719066796278294,
+      "grad_norm": 4.58129962100579,
+      "learning_rate": 1.3035742877064745e-06,
+      "loss": 0.3659,
+      "step": 11117
+    },
+    {
+      "epoch": 0.7719761144285516,
+      "grad_norm": 3.577670938560837,
+      "learning_rate": 1.3028171598548739e-06,
+      "loss": 0.2649,
+      "step": 11118
+    },
+    {
+      "epoch": 0.7720455492292737,
+      "grad_norm": 4.51070262258794,
+      "learning_rate": 1.302060219002596e-06,
+      "loss": 0.6092,
+      "step": 11119
+    },
+    {
+      "epoch": 0.7721149840299958,
+      "grad_norm": 3.4713305699704615,
+      "learning_rate": 1.3013034651879296e-06,
+      "loss": 0.3951,
+      "step": 11120
+    },
+    {
+      "epoch": 0.772184418830718,
+      "grad_norm": 4.506783336053114,
+      "learning_rate": 1.3005468984491498e-06,
+      "loss": 0.6209,
+      "step": 11121
+    },
+    {
+      "epoch": 0.7722538536314401,
+      "grad_norm": 4.480311284300879,
+      "learning_rate": 1.2997905188245192e-06,
+      "loss": 0.4549,
+      "step": 11122
+    },
+    {
+      "epoch": 0.7723232884321622,
+      "grad_norm": 4.9291650661949244,
+      "learning_rate": 1.2990343263522981e-06,
+      "loss": 0.6107,
+      "step": 11123
+    },
+    {
+      "epoch": 0.7723927232328843,
+      "grad_norm": 3.925324749050907,
+      "learning_rate": 1.2982783210707344e-06,
+      "loss": 0.5315,
+      "step": 11124
+    },
+    {
+      "epoch": 0.7724621580336064,
+      "grad_norm": 4.2775014290038955,
+      "learning_rate": 1.2975225030180611e-06,
+      "loss": 0.3596,
+      "step": 11125
+    },
+    {
+      "epoch": 0.7725315928343286,
+      "grad_norm": 5.451064868441418,
+      "learning_rate": 1.2967668722325122e-06,
+      "loss": 0.4843,
+      "step": 11126
+    },
+    {
+      "epoch": 0.7726010276350507,
+      "grad_norm": 3.768753251155597,
+      "learning_rate": 1.296011428752304e-06,
+      "loss": 0.3946,
+      "step": 11127
+    },
+    {
+      "epoch": 0.7726704624357729,
+      "grad_norm": 4.2742716557016776,
+      "learning_rate": 1.295256172615646e-06,
+      "loss": 0.6154,
+      "step": 11128
+    },
+    {
+      "epoch": 0.7727398972364949,
+      "grad_norm": 4.989508233636311,
+      "learning_rate": 1.2945011038607391e-06,
+      "loss": 0.4086,
+      "step": 11129
+    },
+    {
+      "epoch": 0.772809332037217,
+      "grad_norm": 4.290905212256956,
+      "learning_rate": 1.2937462225257736e-06,
+      "loss": 0.583,
+      "step": 11130
+    },
+    {
+      "epoch": 0.7728787668379392,
+      "grad_norm": 3.342178181321543,
+      "learning_rate": 1.2929915286489282e-06,
+      "loss": 0.4671,
+      "step": 11131
+    },
+    {
+      "epoch": 0.7729482016386613,
+      "grad_norm": 4.239194978571748,
+      "learning_rate": 1.2922370222683794e-06,
+      "loss": 0.5314,
+      "step": 11132
+    },
+    {
+      "epoch": 0.7730176364393834,
+      "grad_norm": 4.373494955072838,
+      "learning_rate": 1.2914827034222854e-06,
+      "loss": 0.5807,
+      "step": 11133
+    },
+    {
+      "epoch": 0.7730870712401056,
+      "grad_norm": 3.5426233860507566,
+      "learning_rate": 1.2907285721487977e-06,
+      "loss": 0.5046,
+      "step": 11134
+    },
+    {
+      "epoch": 0.7731565060408276,
+      "grad_norm": 6.912792922237515,
+      "learning_rate": 1.2899746284860648e-06,
+      "loss": 0.3068,
+      "step": 11135
+    },
+    {
+      "epoch": 0.7732259408415498,
+      "grad_norm": 3.4953234177342196,
+      "learning_rate": 1.2892208724722143e-06,
+      "loss": 0.2164,
+      "step": 11136
+    },
+    {
+      "epoch": 0.7732953756422719,
+      "grad_norm": 4.3429618460675465,
+      "learning_rate": 1.2884673041453744e-06,
+      "loss": 0.509,
+      "step": 11137
+    },
+    {
+      "epoch": 0.773364810442994,
+      "grad_norm": 4.1691942238732205,
+      "learning_rate": 1.2877139235436581e-06,
+      "loss": 0.702,
+      "step": 11138
+    },
+    {
+      "epoch": 0.7734342452437162,
+      "grad_norm": 2.1622376248559547,
+      "learning_rate": 1.2869607307051712e-06,
+      "loss": 0.0621,
+      "step": 11139
+    },
+    {
+      "epoch": 0.7735036800444383,
+      "grad_norm": 3.370398681781712,
+      "learning_rate": 1.286207725668009e-06,
+      "loss": 0.3279,
+      "step": 11140
+    },
+    {
+      "epoch": 0.7735731148451604,
+      "grad_norm": 4.418365996213047,
+      "learning_rate": 1.2854549084702572e-06,
+      "loss": 0.5992,
+      "step": 11141
+    },
+    {
+      "epoch": 0.7736425496458825,
+      "grad_norm": 4.039835781910002,
+      "learning_rate": 1.2847022791499925e-06,
+      "loss": 0.45,
+      "step": 11142
+    },
+    {
+      "epoch": 0.7737119844466046,
+      "grad_norm": 4.598566850025614,
+      "learning_rate": 1.283949837745283e-06,
+      "loss": 0.4429,
+      "step": 11143
+    },
+    {
+      "epoch": 0.7737814192473268,
+      "grad_norm": 4.047718471545714,
+      "learning_rate": 1.2831975842941857e-06,
+      "loss": 0.5609,
+      "step": 11144
+    },
+    {
+      "epoch": 0.7738508540480489,
+      "grad_norm": 4.87478660792395,
+      "learning_rate": 1.2824455188347467e-06,
+      "loss": 0.5243,
+      "step": 11145
+    },
+    {
+      "epoch": 0.773920288848771,
+      "grad_norm": 4.5626222464537145,
+      "learning_rate": 1.2816936414050101e-06,
+      "loss": 0.4664,
+      "step": 11146
+    },
+    {
+      "epoch": 0.7739897236494931,
+      "grad_norm": 4.449152530240054,
+      "learning_rate": 1.2809419520430005e-06,
+      "loss": 0.4394,
+      "step": 11147
+    },
+    {
+      "epoch": 0.7740591584502152,
+      "grad_norm": 3.9104753064122777,
+      "learning_rate": 1.2801904507867364e-06,
+      "loss": 0.4877,
+      "step": 11148
+    },
+    {
+      "epoch": 0.7741285932509374,
+      "grad_norm": 3.965173988542395,
+      "learning_rate": 1.2794391376742333e-06,
+      "loss": 0.3623,
+      "step": 11149
+    },
+    {
+      "epoch": 0.7741980280516595,
+      "grad_norm": 4.1383813612732,
+      "learning_rate": 1.278688012743487e-06,
+      "loss": 0.5335,
+      "step": 11150
+    },
+    {
+      "epoch": 0.7742674628523816,
+      "grad_norm": 6.293982406771598,
+      "learning_rate": 1.2779370760324883e-06,
+      "loss": 0.2873,
+      "step": 11151
+    },
+    {
+      "epoch": 0.7743368976531038,
+      "grad_norm": 3.7902786235906336,
+      "learning_rate": 1.2771863275792224e-06,
+      "loss": 0.4438,
+      "step": 11152
+    },
+    {
+      "epoch": 0.7744063324538258,
+      "grad_norm": 3.5599776231892606,
+      "learning_rate": 1.2764357674216604e-06,
+      "loss": 0.3033,
+      "step": 11153
+    },
+    {
+      "epoch": 0.774475767254548,
+      "grad_norm": 4.160360799135336,
+      "learning_rate": 1.2756853955977615e-06,
+      "loss": 0.4496,
+      "step": 11154
+    },
+    {
+      "epoch": 0.7745452020552701,
+      "grad_norm": 3.6985132761543933,
+      "learning_rate": 1.2749352121454823e-06,
+      "loss": 0.421,
+      "step": 11155
+    },
+    {
+      "epoch": 0.7746146368559922,
+      "grad_norm": 3.5901490194287913,
+      "learning_rate": 1.274185217102764e-06,
+      "loss": 0.2036,
+      "step": 11156
+    },
+    {
+      "epoch": 0.7746840716567144,
+      "grad_norm": 6.492080285726098,
+      "learning_rate": 1.2734354105075424e-06,
+      "loss": 0.5901,
+      "step": 11157
+    },
+    {
+      "epoch": 0.7747535064574365,
+      "grad_norm": 3.843998910953194,
+      "learning_rate": 1.2726857923977404e-06,
+      "loss": 0.3586,
+      "step": 11158
+    },
+    {
+      "epoch": 0.7748229412581585,
+      "grad_norm": 4.634498578618431,
+      "learning_rate": 1.271936362811273e-06,
+      "loss": 0.6976,
+      "step": 11159
+    },
+    {
+      "epoch": 0.7748923760588807,
+      "grad_norm": 4.356909338343021,
+      "learning_rate": 1.2711871217860461e-06,
+      "loss": 0.551,
+      "step": 11160
+    },
+    {
+      "epoch": 0.7749618108596028,
+      "grad_norm": 2.1160183251843185,
+      "learning_rate": 1.2704380693599554e-06,
+      "loss": 0.1749,
+      "step": 11161
+    },
+    {
+      "epoch": 0.775031245660325,
+      "grad_norm": 3.7708734675145603,
+      "learning_rate": 1.2696892055708865e-06,
+      "loss": 0.4576,
+      "step": 11162
+    },
+    {
+      "epoch": 0.7751006804610471,
+      "grad_norm": 4.189641816182364,
+      "learning_rate": 1.2689405304567154e-06,
+      "loss": 0.3033,
+      "step": 11163
+    },
+    {
+      "epoch": 0.7751701152617692,
+      "grad_norm": 3.2469457590471174,
+      "learning_rate": 1.2681920440553124e-06,
+      "loss": 0.3359,
+      "step": 11164
+    },
+    {
+      "epoch": 0.7752395500624913,
+      "grad_norm": 3.159495107922084,
+      "learning_rate": 1.2674437464045309e-06,
+      "loss": 0.3316,
+      "step": 11165
+    },
+    {
+      "epoch": 0.7753089848632134,
+      "grad_norm": 4.187543324894305,
+      "learning_rate": 1.2666956375422225e-06,
+      "loss": 0.4446,
+      "step": 11166
+    },
+    {
+      "epoch": 0.7753784196639356,
+      "grad_norm": 4.501309253207573,
+      "learning_rate": 1.2659477175062252e-06,
+      "loss": 0.5944,
+      "step": 11167
+    },
+    {
+      "epoch": 0.7754478544646577,
+      "grad_norm": 4.1310521787448975,
+      "learning_rate": 1.2651999863343638e-06,
+      "loss": 0.5855,
+      "step": 11168
+    },
+    {
+      "epoch": 0.7755172892653798,
+      "grad_norm": 4.103732258574949,
+      "learning_rate": 1.2644524440644628e-06,
+      "loss": 0.4594,
+      "step": 11169
+    },
+    {
+      "epoch": 0.775586724066102,
+      "grad_norm": 3.1038501954351827,
+      "learning_rate": 1.26370509073433e-06,
+      "loss": 0.1778,
+      "step": 11170
+    },
+    {
+      "epoch": 0.775656158866824,
+      "grad_norm": 3.315473051285473,
+      "learning_rate": 1.2629579263817659e-06,
+      "loss": 0.3184,
+      "step": 11171
+    },
+    {
+      "epoch": 0.7757255936675461,
+      "grad_norm": 3.3907964650048306,
+      "learning_rate": 1.2622109510445613e-06,
+      "loss": 0.4255,
+      "step": 11172
+    },
+    {
+      "epoch": 0.7757950284682683,
+      "grad_norm": 3.5330528223402475,
+      "learning_rate": 1.261464164760497e-06,
+      "loss": 0.4794,
+      "step": 11173
+    },
+    {
+      "epoch": 0.7758644632689904,
+      "grad_norm": 4.482184282501224,
+      "learning_rate": 1.260717567567345e-06,
+      "loss": 0.404,
+      "step": 11174
+    },
+    {
+      "epoch": 0.7759338980697126,
+      "grad_norm": 3.7916923243713367,
+      "learning_rate": 1.2599711595028668e-06,
+      "loss": 0.3922,
+      "step": 11175
+    },
+    {
+      "epoch": 0.7760033328704347,
+      "grad_norm": 3.9800725797856944,
+      "learning_rate": 1.2592249406048156e-06,
+      "loss": 0.3623,
+      "step": 11176
+    },
+    {
+      "epoch": 0.7760727676711567,
+      "grad_norm": 4.530939846911843,
+      "learning_rate": 1.2584789109109325e-06,
+      "loss": 0.5085,
+      "step": 11177
+    },
+    {
+      "epoch": 0.7761422024718789,
+      "grad_norm": 4.525872438064664,
+      "learning_rate": 1.2577330704589546e-06,
+      "loss": 0.6661,
+      "step": 11178
+    },
+    {
+      "epoch": 0.776211637272601,
+      "grad_norm": 4.550337782203634,
+      "learning_rate": 1.2569874192866023e-06,
+      "loss": 0.4555,
+      "step": 11179
+    },
+    {
+      "epoch": 0.7762810720733232,
+      "grad_norm": 3.1989676489406493,
+      "learning_rate": 1.2562419574315893e-06,
+      "loss": 0.3755,
+      "step": 11180
+    },
+    {
+      "epoch": 0.7763505068740453,
+      "grad_norm": 3.69667624966104,
+      "learning_rate": 1.2554966849316248e-06,
+      "loss": 0.4358,
+      "step": 11181
+    },
+    {
+      "epoch": 0.7764199416747674,
+      "grad_norm": 2.812789794811207,
+      "learning_rate": 1.2547516018243989e-06,
+      "loss": 0.2851,
+      "step": 11182
+    },
+    {
+      "epoch": 0.7764893764754895,
+      "grad_norm": 3.696093825882281,
+      "learning_rate": 1.2540067081475982e-06,
+      "loss": 0.592,
+      "step": 11183
+    },
+    {
+      "epoch": 0.7765588112762116,
+      "grad_norm": 3.0240310655375127,
+      "learning_rate": 1.2532620039388999e-06,
+      "loss": 0.3427,
+      "step": 11184
+    },
+    {
+      "epoch": 0.7766282460769337,
+      "grad_norm": 4.426728765442772,
+      "learning_rate": 1.2525174892359703e-06,
+      "loss": 0.3118,
+      "step": 11185
+    },
+    {
+      "epoch": 0.7766976808776559,
+      "grad_norm": 3.1907989879479324,
+      "learning_rate": 1.2517731640764658e-06,
+      "loss": 0.2592,
+      "step": 11186
+    },
+    {
+      "epoch": 0.776767115678378,
+      "grad_norm": 3.57740323825533,
+      "learning_rate": 1.2510290284980325e-06,
+      "loss": 0.4283,
+      "step": 11187
+    },
+    {
+      "epoch": 0.7768365504791002,
+      "grad_norm": 3.299562039862095,
+      "learning_rate": 1.250285082538309e-06,
+      "loss": 0.3577,
+      "step": 11188
+    },
+    {
+      "epoch": 0.7769059852798222,
+      "grad_norm": 9.413314086670317,
+      "learning_rate": 1.2495413262349232e-06,
+      "loss": 0.3637,
+      "step": 11189
+    },
+    {
+      "epoch": 0.7769754200805443,
+      "grad_norm": 3.6647091312422937,
+      "learning_rate": 1.248797759625493e-06,
+      "loss": 0.4019,
+      "step": 11190
+    },
+    {
+      "epoch": 0.7770448548812665,
+      "grad_norm": 4.454577746079493,
+      "learning_rate": 1.2480543827476271e-06,
+      "loss": 0.3812,
+      "step": 11191
+    },
+    {
+      "epoch": 0.7771142896819886,
+      "grad_norm": 4.204303993916557,
+      "learning_rate": 1.2473111956389256e-06,
+      "loss": 0.4174,
+      "step": 11192
+    },
+    {
+      "epoch": 0.7771837244827108,
+      "grad_norm": 3.733056030932026,
+      "learning_rate": 1.2465681983369765e-06,
+      "loss": 0.3973,
+      "step": 11193
+    },
+    {
+      "epoch": 0.7772531592834329,
+      "grad_norm": 6.573629691502059,
+      "learning_rate": 1.2458253908793593e-06,
+      "loss": 0.6169,
+      "step": 11194
+    },
+    {
+      "epoch": 0.7773225940841549,
+      "grad_norm": 4.425803197350088,
+      "learning_rate": 1.245082773303648e-06,
+      "loss": 0.5293,
+      "step": 11195
+    },
+    {
+      "epoch": 0.7773920288848771,
+      "grad_norm": 3.442383394349508,
+      "learning_rate": 1.2443403456474017e-06,
+      "loss": 0.4475,
+      "step": 11196
+    },
+    {
+      "epoch": 0.7774614636855992,
+      "grad_norm": 4.418300191948008,
+      "learning_rate": 1.2435981079481685e-06,
+      "loss": 0.4828,
+      "step": 11197
+    },
+    {
+      "epoch": 0.7775308984863214,
+      "grad_norm": 4.122954938482227,
+      "learning_rate": 1.242856060243493e-06,
+      "loss": 0.3046,
+      "step": 11198
+    },
+    {
+      "epoch": 0.7776003332870435,
+      "grad_norm": 3.458910747379604,
+      "learning_rate": 1.2421142025709087e-06,
+      "loss": 0.3116,
+      "step": 11199
+    },
+    {
+      "epoch": 0.7776697680877656,
+      "grad_norm": 5.177356225642038,
+      "learning_rate": 1.2413725349679323e-06,
+      "loss": 0.4959,
+      "step": 11200
+    },
+    {
+      "epoch": 0.7777392028884877,
+      "grad_norm": 3.9561133792476832,
+      "learning_rate": 1.240631057472082e-06,
+      "loss": 0.4218,
+      "step": 11201
+    },
+    {
+      "epoch": 0.7778086376892098,
+      "grad_norm": 4.600400626689768,
+      "learning_rate": 1.2398897701208584e-06,
+      "loss": 0.6321,
+      "step": 11202
+    },
+    {
+      "epoch": 0.7778780724899319,
+      "grad_norm": 4.684915422183934,
+      "learning_rate": 1.2391486729517555e-06,
+      "loss": 0.665,
+      "step": 11203
+    },
+    {
+      "epoch": 0.7779475072906541,
+      "grad_norm": 4.468353799148326,
+      "learning_rate": 1.2384077660022564e-06,
+      "loss": 0.7129,
+      "step": 11204
+    },
+    {
+      "epoch": 0.7780169420913762,
+      "grad_norm": 3.9434280793964187,
+      "learning_rate": 1.2376670493098369e-06,
+      "loss": 0.5862,
+      "step": 11205
+    },
+    {
+      "epoch": 0.7780863768920984,
+      "grad_norm": 4.407253756433314,
+      "learning_rate": 1.2369265229119604e-06,
+      "loss": 0.6763,
+      "step": 11206
+    },
+    {
+      "epoch": 0.7781558116928204,
+      "grad_norm": 4.348571153194414,
+      "learning_rate": 1.2361861868460817e-06,
+      "loss": 0.276,
+      "step": 11207
+    },
+    {
+      "epoch": 0.7782252464935425,
+      "grad_norm": 4.781600539493049,
+      "learning_rate": 1.2354460411496471e-06,
+      "loss": 0.4435,
+      "step": 11208
+    },
+    {
+      "epoch": 0.7782946812942647,
+      "grad_norm": 4.73954866057255,
+      "learning_rate": 1.2347060858600901e-06,
+      "loss": 0.5868,
+      "step": 11209
+    },
+    {
+      "epoch": 0.7783641160949868,
+      "grad_norm": 3.667012292074982,
+      "learning_rate": 1.2339663210148418e-06,
+      "loss": 0.3197,
+      "step": 11210
+    },
+    {
+      "epoch": 0.778433550895709,
+      "grad_norm": 3.381520937209768,
+      "learning_rate": 1.2332267466513136e-06,
+      "loss": 0.4305,
+      "step": 11211
+    },
+    {
+      "epoch": 0.7785029856964311,
+      "grad_norm": 4.3554728188783125,
+      "learning_rate": 1.2324873628069128e-06,
+      "loss": 0.3372,
+      "step": 11212
+    },
+    {
+      "epoch": 0.7785724204971531,
+      "grad_norm": 2.8944360516061134,
+      "learning_rate": 1.2317481695190397e-06,
+      "loss": 0.2599,
+      "step": 11213
+    },
+    {
+      "epoch": 0.7786418552978753,
+      "grad_norm": 2.890490663249299,
+      "learning_rate": 1.2310091668250812e-06,
+      "loss": 0.2777,
+      "step": 11214
+    },
+    {
+      "epoch": 0.7787112900985974,
+      "grad_norm": 3.3529103723832026,
+      "learning_rate": 1.2302703547624117e-06,
+      "loss": 0.4323,
+      "step": 11215
+    },
+    {
+      "epoch": 0.7787807248993195,
+      "grad_norm": 4.315742041441327,
+      "learning_rate": 1.229531733368403e-06,
+      "loss": 0.6468,
+      "step": 11216
+    },
+    {
+      "epoch": 0.7788501597000417,
+      "grad_norm": 3.012007804277922,
+      "learning_rate": 1.2287933026804127e-06,
+      "loss": 0.332,
+      "step": 11217
+    },
+    {
+      "epoch": 0.7789195945007638,
+      "grad_norm": 3.113154367919129,
+      "learning_rate": 1.2280550627357895e-06,
+      "loss": 0.2818,
+      "step": 11218
+    },
+    {
+      "epoch": 0.778989029301486,
+      "grad_norm": 3.18966482272493,
+      "learning_rate": 1.2273170135718737e-06,
+      "loss": 0.3165,
+      "step": 11219
+    },
+    {
+      "epoch": 0.779058464102208,
+      "grad_norm": 3.8347605356456924,
+      "learning_rate": 1.2265791552259936e-06,
+      "loss": 0.5446,
+      "step": 11220
+    },
+    {
+      "epoch": 0.7791278989029301,
+      "grad_norm": 4.240622930131518,
+      "learning_rate": 1.2258414877354703e-06,
+      "loss": 0.4574,
+      "step": 11221
+    },
+    {
+      "epoch": 0.7791973337036523,
+      "grad_norm": 5.407028027413241,
+      "learning_rate": 1.2251040111376134e-06,
+      "loss": 0.6103,
+      "step": 11222
+    },
+    {
+      "epoch": 0.7792667685043744,
+      "grad_norm": 3.9794865762751694,
+      "learning_rate": 1.2243667254697244e-06,
+      "loss": 0.3972,
+      "step": 11223
+    },
+    {
+      "epoch": 0.7793362033050966,
+      "grad_norm": 3.3105021397290617,
+      "learning_rate": 1.2236296307690938e-06,
+      "loss": 0.2735,
+      "step": 11224
+    },
+    {
+      "epoch": 0.7794056381058186,
+      "grad_norm": 7.101593547162669,
+      "learning_rate": 1.2228927270730034e-06,
+      "loss": 0.3886,
+      "step": 11225
+    },
+    {
+      "epoch": 0.7794750729065407,
+      "grad_norm": 3.6320232204462144,
+      "learning_rate": 1.2221560144187234e-06,
+      "loss": 0.4237,
+      "step": 11226
+    },
+    {
+      "epoch": 0.7795445077072629,
+      "grad_norm": 3.414145462836716,
+      "learning_rate": 1.2214194928435187e-06,
+      "loss": 0.3518,
+      "step": 11227
+    },
+    {
+      "epoch": 0.779613942507985,
+      "grad_norm": 2.181605485923959,
+      "learning_rate": 1.2206831623846416e-06,
+      "loss": 0.1132,
+      "step": 11228
+    },
+    {
+      "epoch": 0.7796833773087071,
+      "grad_norm": 4.021769952031729,
+      "learning_rate": 1.2199470230793303e-06,
+      "loss": 0.428,
+      "step": 11229
+    },
+    {
+      "epoch": 0.7797528121094293,
+      "grad_norm": 4.024510015530156,
+      "learning_rate": 1.2192110749648233e-06,
+      "loss": 0.4476,
+      "step": 11230
+    },
+    {
+      "epoch": 0.7798222469101513,
+      "grad_norm": 3.8428545033013193,
+      "learning_rate": 1.2184753180783426e-06,
+      "loss": 0.6037,
+      "step": 11231
+    },
+    {
+      "epoch": 0.7798916817108735,
+      "grad_norm": 2.665232030006567,
+      "learning_rate": 1.2177397524570988e-06,
+      "loss": 0.3054,
+      "step": 11232
+    },
+    {
+      "epoch": 0.7799611165115956,
+      "grad_norm": 5.267989077484462,
+      "learning_rate": 1.2170043781383e-06,
+      "loss": 0.6951,
+      "step": 11233
+    },
+    {
+      "epoch": 0.7800305513123177,
+      "grad_norm": 5.576370981818137,
+      "learning_rate": 1.2162691951591387e-06,
+      "loss": 0.4987,
+      "step": 11234
+    },
+    {
+      "epoch": 0.7800999861130399,
+      "grad_norm": 2.5268382964252236,
+      "learning_rate": 1.2155342035568002e-06,
+      "loss": 0.2081,
+      "step": 11235
+    },
+    {
+      "epoch": 0.780169420913762,
+      "grad_norm": 3.0816699803578915,
+      "learning_rate": 1.214799403368459e-06,
+      "loss": 0.2639,
+      "step": 11236
+    },
+    {
+      "epoch": 0.7802388557144841,
+      "grad_norm": 3.6181086210979063,
+      "learning_rate": 1.2140647946312807e-06,
+      "loss": 0.3882,
+      "step": 11237
+    },
+    {
+      "epoch": 0.7803082905152062,
+      "grad_norm": 3.4696028305666453,
+      "learning_rate": 1.2133303773824196e-06,
+      "loss": 0.2354,
+      "step": 11238
+    },
+    {
+      "epoch": 0.7803777253159283,
+      "grad_norm": 4.407893344862321,
+      "learning_rate": 1.2125961516590263e-06,
+      "loss": 0.6045,
+      "step": 11239
+    },
+    {
+      "epoch": 0.7804471601166505,
+      "grad_norm": 4.293439763404174,
+      "learning_rate": 1.211862117498232e-06,
+      "loss": 0.4276,
+      "step": 11240
+    },
+    {
+      "epoch": 0.7805165949173726,
+      "grad_norm": 4.379460585466756,
+      "learning_rate": 1.2111282749371645e-06,
+      "loss": 0.457,
+      "step": 11241
+    },
+    {
+      "epoch": 0.7805860297180947,
+      "grad_norm": 4.066907193858863,
+      "learning_rate": 1.2103946240129438e-06,
+      "loss": 0.1741,
+      "step": 11242
+    },
+    {
+      "epoch": 0.7806554645188168,
+      "grad_norm": 4.085401848180031,
+      "learning_rate": 1.2096611647626733e-06,
+      "loss": 0.3117,
+      "step": 11243
+    },
+    {
+      "epoch": 0.7807248993195389,
+      "grad_norm": 3.3251191478432895,
+      "learning_rate": 1.2089278972234507e-06,
+      "loss": 0.4288,
+      "step": 11244
+    },
+    {
+      "epoch": 0.7807943341202611,
+      "grad_norm": 3.7193105088589844,
+      "learning_rate": 1.2081948214323664e-06,
+      "loss": 0.4398,
+      "step": 11245
+    },
+    {
+      "epoch": 0.7808637689209832,
+      "grad_norm": 3.073414544856734,
+      "learning_rate": 1.2074619374264978e-06,
+      "loss": 0.2748,
+      "step": 11246
+    },
+    {
+      "epoch": 0.7809332037217053,
+      "grad_norm": 4.601508182394447,
+      "learning_rate": 1.2067292452429124e-06,
+      "loss": 0.5329,
+      "step": 11247
+    },
+    {
+      "epoch": 0.7810026385224275,
+      "grad_norm": 4.728242211643195,
+      "learning_rate": 1.20599674491867e-06,
+      "loss": 0.6275,
+      "step": 11248
+    },
+    {
+      "epoch": 0.7810720733231495,
+      "grad_norm": 4.759074461582454,
+      "learning_rate": 1.2052644364908184e-06,
+      "loss": 0.6808,
+      "step": 11249
+    },
+    {
+      "epoch": 0.7811415081238717,
+      "grad_norm": 4.164736966673322,
+      "learning_rate": 1.2045323199963976e-06,
+      "loss": 0.4705,
+      "step": 11250
+    },
+    {
+      "epoch": 0.7812109429245938,
+      "grad_norm": 4.578369622572767,
+      "learning_rate": 1.2038003954724376e-06,
+      "loss": 0.6658,
+      "step": 11251
+    },
+    {
+      "epoch": 0.7812803777253159,
+      "grad_norm": 3.4822720150675006,
+      "learning_rate": 1.2030686629559574e-06,
+      "loss": 0.3235,
+      "step": 11252
+    },
+    {
+      "epoch": 0.7813498125260381,
+      "grad_norm": 3.624964442548062,
+      "learning_rate": 1.202337122483968e-06,
+      "loss": 0.385,
+      "step": 11253
+    },
+    {
+      "epoch": 0.7814192473267602,
+      "grad_norm": 3.97712023280229,
+      "learning_rate": 1.20160577409347e-06,
+      "loss": 0.4751,
+      "step": 11254
+    },
+    {
+      "epoch": 0.7814886821274823,
+      "grad_norm": 5.879873981579039,
+      "learning_rate": 1.2008746178214514e-06,
+      "loss": 0.4977,
+      "step": 11255
+    },
+    {
+      "epoch": 0.7815581169282044,
+      "grad_norm": 2.8985257073468507,
+      "learning_rate": 1.2001436537048994e-06,
+      "loss": 0.3399,
+      "step": 11256
+    },
+    {
+      "epoch": 0.7816275517289265,
+      "grad_norm": 3.0558038455456162,
+      "learning_rate": 1.1994128817807799e-06,
+      "loss": 0.2203,
+      "step": 11257
+    },
+    {
+      "epoch": 0.7816969865296487,
+      "grad_norm": 4.807277853865202,
+      "learning_rate": 1.1986823020860543e-06,
+      "loss": 0.659,
+      "step": 11258
+    },
+    {
+      "epoch": 0.7817664213303708,
+      "grad_norm": 4.658283278105699,
+      "learning_rate": 1.1979519146576784e-06,
+      "loss": 0.5582,
+      "step": 11259
+    },
+    {
+      "epoch": 0.7818358561310929,
+      "grad_norm": 4.045257622114772,
+      "learning_rate": 1.1972217195325936e-06,
+      "loss": 0.4471,
+      "step": 11260
+    },
+    {
+      "epoch": 0.781905290931815,
+      "grad_norm": 4.115604195055608,
+      "learning_rate": 1.196491716747728e-06,
+      "loss": 0.2566,
+      "step": 11261
+    },
+    {
+      "epoch": 0.7819747257325371,
+      "grad_norm": 4.609078265124921,
+      "learning_rate": 1.195761906340009e-06,
+      "loss": 0.4705,
+      "step": 11262
+    },
+    {
+      "epoch": 0.7820441605332593,
+      "grad_norm": 5.723452325117509,
+      "learning_rate": 1.1950322883463478e-06,
+      "loss": 0.4996,
+      "step": 11263
+    },
+    {
+      "epoch": 0.7821135953339814,
+      "grad_norm": 3.881851879338284,
+      "learning_rate": 1.1943028628036479e-06,
+      "loss": 0.3259,
+      "step": 11264
+    },
+    {
+      "epoch": 0.7821830301347035,
+      "grad_norm": 3.6404610830946975,
+      "learning_rate": 1.1935736297488032e-06,
+      "loss": 0.3962,
+      "step": 11265
+    },
+    {
+      "epoch": 0.7822524649354257,
+      "grad_norm": 3.852032070702059,
+      "learning_rate": 1.1928445892186968e-06,
+      "loss": 0.434,
+      "step": 11266
+    },
+    {
+      "epoch": 0.7823218997361477,
+      "grad_norm": 4.141008831667328,
+      "learning_rate": 1.1921157412502028e-06,
+      "loss": 0.4678,
+      "step": 11267
+    },
+    {
+      "epoch": 0.7823913345368699,
+      "grad_norm": 4.139166213952968,
+      "learning_rate": 1.1913870858801856e-06,
+      "loss": 0.27,
+      "step": 11268
+    },
+    {
+      "epoch": 0.782460769337592,
+      "grad_norm": 5.126137652570993,
+      "learning_rate": 1.1906586231455004e-06,
+      "loss": 0.598,
+      "step": 11269
+    },
+    {
+      "epoch": 0.7825302041383141,
+      "grad_norm": 4.3770950308468475,
+      "learning_rate": 1.1899303530829892e-06,
+      "loss": 0.5311,
+      "step": 11270
+    },
+    {
+      "epoch": 0.7825996389390363,
+      "grad_norm": 4.197732898960514,
+      "learning_rate": 1.189202275729493e-06,
+      "loss": 0.4435,
+      "step": 11271
+    },
+    {
+      "epoch": 0.7826690737397584,
+      "grad_norm": 3.364514654689977,
+      "learning_rate": 1.1884743911218316e-06,
+      "loss": 0.3003,
+      "step": 11272
+    },
+    {
+      "epoch": 0.7827385085404804,
+      "grad_norm": 3.4575674605396487,
+      "learning_rate": 1.1877466992968213e-06,
+      "loss": 0.3273,
+      "step": 11273
+    },
+    {
+      "epoch": 0.7828079433412026,
+      "grad_norm": 4.469433419188783,
+      "learning_rate": 1.1870192002912718e-06,
+      "loss": 0.4423,
+      "step": 11274
+    },
+    {
+      "epoch": 0.7828773781419247,
+      "grad_norm": 3.9076299214456585,
+      "learning_rate": 1.1862918941419737e-06,
+      "loss": 0.4181,
+      "step": 11275
+    },
+    {
+      "epoch": 0.7829468129426469,
+      "grad_norm": 4.389406482550695,
+      "learning_rate": 1.1855647808857174e-06,
+      "loss": 0.487,
+      "step": 11276
+    },
+    {
+      "epoch": 0.783016247743369,
+      "grad_norm": 3.6328097332679685,
+      "learning_rate": 1.1848378605592787e-06,
+      "loss": 0.3856,
+      "step": 11277
+    },
+    {
+      "epoch": 0.783085682544091,
+      "grad_norm": 4.2236849045680405,
+      "learning_rate": 1.1841111331994232e-06,
+      "loss": 0.3211,
+      "step": 11278
+    },
+    {
+      "epoch": 0.7831551173448132,
+      "grad_norm": 4.0957971864797855,
+      "learning_rate": 1.1833845988429087e-06,
+      "loss": 0.3555,
+      "step": 11279
+    },
+    {
+      "epoch": 0.7832245521455353,
+      "grad_norm": 3.3012225863481226,
+      "learning_rate": 1.1826582575264829e-06,
+      "loss": 0.2574,
+      "step": 11280
+    },
+    {
+      "epoch": 0.7832939869462575,
+      "grad_norm": 4.049365111882,
+      "learning_rate": 1.1819321092868824e-06,
+      "loss": 0.4068,
+      "step": 11281
+    },
+    {
+      "epoch": 0.7833634217469796,
+      "grad_norm": 4.329727190282177,
+      "learning_rate": 1.1812061541608356e-06,
+      "loss": 0.5488,
+      "step": 11282
+    },
+    {
+      "epoch": 0.7834328565477017,
+      "grad_norm": 3.062418882354128,
+      "learning_rate": 1.18048039218506e-06,
+      "loss": 0.2609,
+      "step": 11283
+    },
+    {
+      "epoch": 0.7835022913484239,
+      "grad_norm": 2.50708345205818,
+      "learning_rate": 1.1797548233962624e-06,
+      "loss": 0.225,
+      "step": 11284
+    },
+    {
+      "epoch": 0.7835717261491459,
+      "grad_norm": 4.166137544973621,
+      "learning_rate": 1.1790294478311465e-06,
+      "loss": 0.4001,
+      "step": 11285
+    },
+    {
+      "epoch": 0.783641160949868,
+      "grad_norm": 3.041044855762488,
+      "learning_rate": 1.1783042655263955e-06,
+      "loss": 0.2213,
+      "step": 11286
+    },
+    {
+      "epoch": 0.7837105957505902,
+      "grad_norm": 4.8354025596662105,
+      "learning_rate": 1.1775792765186888e-06,
+      "loss": 0.455,
+      "step": 11287
+    },
+    {
+      "epoch": 0.7837800305513123,
+      "grad_norm": 5.318005937227194,
+      "learning_rate": 1.1768544808446985e-06,
+      "loss": 0.6227,
+      "step": 11288
+    },
+    {
+      "epoch": 0.7838494653520345,
+      "grad_norm": 4.203548368096632,
+      "learning_rate": 1.1761298785410836e-06,
+      "loss": 0.4818,
+      "step": 11289
+    },
+    {
+      "epoch": 0.7839189001527566,
+      "grad_norm": 3.7041946831282875,
+      "learning_rate": 1.17540546964449e-06,
+      "loss": 0.4006,
+      "step": 11290
+    },
+    {
+      "epoch": 0.7839883349534786,
+      "grad_norm": 3.525087539162239,
+      "learning_rate": 1.1746812541915609e-06,
+      "loss": 0.2785,
+      "step": 11291
+    },
+    {
+      "epoch": 0.7840577697542008,
+      "grad_norm": 3.6946546337740385,
+      "learning_rate": 1.1739572322189269e-06,
+      "loss": 0.2869,
+      "step": 11292
+    },
+    {
+      "epoch": 0.7841272045549229,
+      "grad_norm": 2.9358199681807786,
+      "learning_rate": 1.1732334037632038e-06,
+      "loss": 0.1875,
+      "step": 11293
+    },
+    {
+      "epoch": 0.7841966393556451,
+      "grad_norm": 4.1319814146350495,
+      "learning_rate": 1.1725097688610065e-06,
+      "loss": 0.5681,
+      "step": 11294
+    },
+    {
+      "epoch": 0.7842660741563672,
+      "grad_norm": 5.23825998443839,
+      "learning_rate": 1.1717863275489339e-06,
+      "loss": 0.6425,
+      "step": 11295
+    },
+    {
+      "epoch": 0.7843355089570893,
+      "grad_norm": 3.7781263827885163,
+      "learning_rate": 1.1710630798635765e-06,
+      "loss": 0.3648,
+      "step": 11296
+    },
+    {
+      "epoch": 0.7844049437578114,
+      "grad_norm": 5.083380049675135,
+      "learning_rate": 1.1703400258415159e-06,
+      "loss": 0.5372,
+      "step": 11297
+    },
+    {
+      "epoch": 0.7844743785585335,
+      "grad_norm": 4.759762034732927,
+      "learning_rate": 1.1696171655193238e-06,
+      "loss": 0.5125,
+      "step": 11298
+    },
+    {
+      "epoch": 0.7845438133592556,
+      "grad_norm": 2.4277255079905204,
+      "learning_rate": 1.168894498933561e-06,
+      "loss": 0.1963,
+      "step": 11299
+    },
+    {
+      "epoch": 0.7846132481599778,
+      "grad_norm": 3.7366891157357665,
+      "learning_rate": 1.1681720261207785e-06,
+      "loss": 0.4515,
+      "step": 11300
+    },
+    {
+      "epoch": 0.7846826829606999,
+      "grad_norm": 4.694677896883966,
+      "learning_rate": 1.1674497471175194e-06,
+      "loss": 0.4774,
+      "step": 11301
+    },
+    {
+      "epoch": 0.7847521177614221,
+      "grad_norm": 3.6119023036919145,
+      "learning_rate": 1.1667276619603141e-06,
+      "loss": 0.2631,
+      "step": 11302
+    },
+    {
+      "epoch": 0.7848215525621441,
+      "grad_norm": 3.5973027115586804,
+      "learning_rate": 1.1660057706856887e-06,
+      "loss": 0.3789,
+      "step": 11303
+    },
+    {
+      "epoch": 0.7848909873628662,
+      "grad_norm": 3.296962565143426,
+      "learning_rate": 1.1652840733301512e-06,
+      "loss": 0.2093,
+      "step": 11304
+    },
+    {
+      "epoch": 0.7849604221635884,
+      "grad_norm": 3.294077115621472,
+      "learning_rate": 1.164562569930205e-06,
+      "loss": 0.3254,
+      "step": 11305
+    },
+    {
+      "epoch": 0.7850298569643105,
+      "grad_norm": 2.866217880312782,
+      "learning_rate": 1.1638412605223465e-06,
+      "loss": 0.2866,
+      "step": 11306
+    },
+    {
+      "epoch": 0.7850992917650327,
+      "grad_norm": 3.5844885014074737,
+      "learning_rate": 1.1631201451430535e-06,
+      "loss": 0.4526,
+      "step": 11307
+    },
+    {
+      "epoch": 0.7851687265657548,
+      "grad_norm": 4.475695810219736,
+      "learning_rate": 1.1623992238288034e-06,
+      "loss": 0.4257,
+      "step": 11308
+    },
+    {
+      "epoch": 0.7852381613664768,
+      "grad_norm": 4.246499966671158,
+      "learning_rate": 1.1616784966160576e-06,
+      "loss": 0.4862,
+      "step": 11309
+    },
+    {
+      "epoch": 0.785307596167199,
+      "grad_norm": 2.7874084214166017,
+      "learning_rate": 1.1609579635412704e-06,
+      "loss": 0.3554,
+      "step": 11310
+    },
+    {
+      "epoch": 0.7853770309679211,
+      "grad_norm": 4.248810921379555,
+      "learning_rate": 1.1602376246408853e-06,
+      "loss": 0.4654,
+      "step": 11311
+    },
+    {
+      "epoch": 0.7854464657686433,
+      "grad_norm": 3.507245997760363,
+      "learning_rate": 1.1595174799513358e-06,
+      "loss": 0.338,
+      "step": 11312
+    },
+    {
+      "epoch": 0.7855159005693654,
+      "grad_norm": 3.0467589220948055,
+      "learning_rate": 1.1587975295090465e-06,
+      "loss": 0.3589,
+      "step": 11313
+    },
+    {
+      "epoch": 0.7855853353700875,
+      "grad_norm": 3.3511849244859837,
+      "learning_rate": 1.1580777733504316e-06,
+      "loss": 0.282,
+      "step": 11314
+    },
+    {
+      "epoch": 0.7856547701708096,
+      "grad_norm": 3.8506623318471362,
+      "learning_rate": 1.1573582115118953e-06,
+      "loss": 0.5069,
+      "step": 11315
+    },
+    {
+      "epoch": 0.7857242049715317,
+      "grad_norm": 4.257211457454584,
+      "learning_rate": 1.156638844029831e-06,
+      "loss": 0.5351,
+      "step": 11316
+    },
+    {
+      "epoch": 0.7857936397722538,
+      "grad_norm": 3.688456548373987,
+      "learning_rate": 1.1559196709406284e-06,
+      "loss": 0.3937,
+      "step": 11317
+    },
+    {
+      "epoch": 0.785863074572976,
+      "grad_norm": 3.3632369695703925,
+      "learning_rate": 1.1552006922806569e-06,
+      "loss": 0.2828,
+      "step": 11318
+    },
+    {
+      "epoch": 0.7859325093736981,
+      "grad_norm": 4.239737360815985,
+      "learning_rate": 1.1544819080862823e-06,
+      "loss": 0.4121,
+      "step": 11319
+    },
+    {
+      "epoch": 0.7860019441744203,
+      "grad_norm": 2.461831572729932,
+      "learning_rate": 1.153763318393863e-06,
+      "loss": 0.2209,
+      "step": 11320
+    },
+    {
+      "epoch": 0.7860713789751423,
+      "grad_norm": 3.215012840104009,
+      "learning_rate": 1.153044923239744e-06,
+      "loss": 0.317,
+      "step": 11321
+    },
+    {
+      "epoch": 0.7861408137758644,
+      "grad_norm": 2.696100821982559,
+      "learning_rate": 1.152326722660257e-06,
+      "loss": 0.2525,
+      "step": 11322
+    },
+    {
+      "epoch": 0.7862102485765866,
+      "grad_norm": 4.336999945366079,
+      "learning_rate": 1.1516087166917318e-06,
+      "loss": 0.3804,
+      "step": 11323
+    },
+    {
+      "epoch": 0.7862796833773087,
+      "grad_norm": 4.640526743614021,
+      "learning_rate": 1.1508909053704837e-06,
+      "loss": 0.4588,
+      "step": 11324
+    },
+    {
+      "epoch": 0.7863491181780309,
+      "grad_norm": 3.610609363223274,
+      "learning_rate": 1.150173288732816e-06,
+      "loss": 0.4553,
+      "step": 11325
+    },
+    {
+      "epoch": 0.786418552978753,
+      "grad_norm": 5.645895054468631,
+      "learning_rate": 1.149455866815028e-06,
+      "loss": 0.5591,
+      "step": 11326
+    },
+    {
+      "epoch": 0.786487987779475,
+      "grad_norm": 4.992208129795625,
+      "learning_rate": 1.1487386396534057e-06,
+      "loss": 0.5609,
+      "step": 11327
+    },
+    {
+      "epoch": 0.7865574225801972,
+      "grad_norm": 5.7692978063073435,
+      "learning_rate": 1.1480216072842242e-06,
+      "loss": 0.4148,
+      "step": 11328
+    },
+    {
+      "epoch": 0.7866268573809193,
+      "grad_norm": 3.863420327676239,
+      "learning_rate": 1.147304769743751e-06,
+      "loss": 0.4899,
+      "step": 11329
+    },
+    {
+      "epoch": 0.7866962921816414,
+      "grad_norm": 3.383752960351298,
+      "learning_rate": 1.146588127068244e-06,
+      "loss": 0.3332,
+      "step": 11330
+    },
+    {
+      "epoch": 0.7867657269823636,
+      "grad_norm": 4.739458440692642,
+      "learning_rate": 1.1458716792939477e-06,
+      "loss": 0.5221,
+      "step": 11331
+    },
+    {
+      "epoch": 0.7868351617830857,
+      "grad_norm": 3.6766396762620577,
+      "learning_rate": 1.145155426457102e-06,
+      "loss": 0.3528,
+      "step": 11332
+    },
+    {
+      "epoch": 0.7869045965838078,
+      "grad_norm": 3.3785907412123586,
+      "learning_rate": 1.1444393685939315e-06,
+      "loss": 0.3226,
+      "step": 11333
+    },
+    {
+      "epoch": 0.7869740313845299,
+      "grad_norm": 4.114867066639711,
+      "learning_rate": 1.1437235057406543e-06,
+      "loss": 0.5705,
+      "step": 11334
+    },
+    {
+      "epoch": 0.787043466185252,
+      "grad_norm": 4.188706201760283,
+      "learning_rate": 1.1430078379334808e-06,
+      "loss": 0.6575,
+      "step": 11335
+    },
+    {
+      "epoch": 0.7871129009859742,
+      "grad_norm": 2.8431987406723107,
+      "learning_rate": 1.1422923652086043e-06,
+      "loss": 0.3519,
+      "step": 11336
+    },
+    {
+      "epoch": 0.7871823357866963,
+      "grad_norm": 3.4195178006511835,
+      "learning_rate": 1.1415770876022152e-06,
+      "loss": 0.3554,
+      "step": 11337
+    },
+    {
+      "epoch": 0.7872517705874185,
+      "grad_norm": 3.5678878208596907,
+      "learning_rate": 1.1408620051504916e-06,
+      "loss": 0.3064,
+      "step": 11338
+    },
+    {
+      "epoch": 0.7873212053881405,
+      "grad_norm": 3.4653970757440975,
+      "learning_rate": 1.1401471178896006e-06,
+      "loss": 0.329,
+      "step": 11339
+    },
+    {
+      "epoch": 0.7873906401888626,
+      "grad_norm": 4.0258921715810265,
+      "learning_rate": 1.1394324258557005e-06,
+      "loss": 0.4223,
+      "step": 11340
+    },
+    {
+      "epoch": 0.7874600749895848,
+      "grad_norm": 4.6370837855410105,
+      "learning_rate": 1.1387179290849397e-06,
+      "loss": 0.5369,
+      "step": 11341
+    },
+    {
+      "epoch": 0.7875295097903069,
+      "grad_norm": 4.634735910484719,
+      "learning_rate": 1.1380036276134571e-06,
+      "loss": 0.6078,
+      "step": 11342
+    },
+    {
+      "epoch": 0.787598944591029,
+      "grad_norm": 4.795402388224027,
+      "learning_rate": 1.1372895214773805e-06,
+      "loss": 0.5842,
+      "step": 11343
+    },
+    {
+      "epoch": 0.7876683793917512,
+      "grad_norm": 3.9914281449789253,
+      "learning_rate": 1.1365756107128296e-06,
+      "loss": 0.4008,
+      "step": 11344
+    },
+    {
+      "epoch": 0.7877378141924732,
+      "grad_norm": 2.978236126373611,
+      "learning_rate": 1.1358618953559103e-06,
+      "loss": 0.4343,
+      "step": 11345
+    },
+    {
+      "epoch": 0.7878072489931954,
+      "grad_norm": 3.14465639920799,
+      "learning_rate": 1.1351483754427274e-06,
+      "loss": 0.3717,
+      "step": 11346
+    },
+    {
+      "epoch": 0.7878766837939175,
+      "grad_norm": 4.0801536124238345,
+      "learning_rate": 1.1344350510093642e-06,
+      "loss": 0.582,
+      "step": 11347
+    },
+    {
+      "epoch": 0.7879461185946396,
+      "grad_norm": 2.1096445705145035,
+      "learning_rate": 1.1337219220919011e-06,
+      "loss": 0.1528,
+      "step": 11348
+    },
+    {
+      "epoch": 0.7880155533953618,
+      "grad_norm": 4.069408588990413,
+      "learning_rate": 1.1330089887264106e-06,
+      "loss": 0.2747,
+      "step": 11349
+    },
+    {
+      "epoch": 0.7880849881960839,
+      "grad_norm": 3.135354718168621,
+      "learning_rate": 1.1322962509489483e-06,
+      "loss": 0.292,
+      "step": 11350
+    },
+    {
+      "epoch": 0.788154422996806,
+      "grad_norm": 3.818235705967781,
+      "learning_rate": 1.1315837087955633e-06,
+      "loss": 0.2741,
+      "step": 11351
+    },
+    {
+      "epoch": 0.7882238577975281,
+      "grad_norm": 3.9448328600431477,
+      "learning_rate": 1.1308713623022988e-06,
+      "loss": 0.417,
+      "step": 11352
+    },
+    {
+      "epoch": 0.7882932925982502,
+      "grad_norm": 4.2721714069369465,
+      "learning_rate": 1.130159211505184e-06,
+      "loss": 0.463,
+      "step": 11353
+    },
+    {
+      "epoch": 0.7883627273989724,
+      "grad_norm": 4.563973400391536,
+      "learning_rate": 1.1294472564402342e-06,
+      "loss": 0.404,
+      "step": 11354
+    },
+    {
+      "epoch": 0.7884321621996945,
+      "grad_norm": 5.477863687049568,
+      "learning_rate": 1.1287354971434639e-06,
+      "loss": 0.5807,
+      "step": 11355
+    },
+    {
+      "epoch": 0.7885015970004166,
+      "grad_norm": 3.9311900698928564,
+      "learning_rate": 1.1280239336508713e-06,
+      "loss": 0.4953,
+      "step": 11356
+    },
+    {
+      "epoch": 0.7885710318011387,
+      "grad_norm": 5.308933016813782,
+      "learning_rate": 1.1273125659984468e-06,
+      "loss": 0.5735,
+      "step": 11357
+    },
+    {
+      "epoch": 0.7886404666018608,
+      "grad_norm": 14.351149690961492,
+      "learning_rate": 1.126601394222171e-06,
+      "loss": 0.4252,
+      "step": 11358
+    },
+    {
+      "epoch": 0.788709901402583,
+      "grad_norm": 5.846547295406732,
+      "learning_rate": 1.1258904183580127e-06,
+      "loss": 0.4391,
+      "step": 11359
+    },
+    {
+      "epoch": 0.7887793362033051,
+      "grad_norm": 4.9560155367976835,
+      "learning_rate": 1.1251796384419338e-06,
+      "loss": 0.7574,
+      "step": 11360
+    },
+    {
+      "epoch": 0.7888487710040272,
+      "grad_norm": 3.9002003616817498,
+      "learning_rate": 1.1244690545098842e-06,
+      "loss": 0.3342,
+      "step": 11361
+    },
+    {
+      "epoch": 0.7889182058047494,
+      "grad_norm": 3.464082850667971,
+      "learning_rate": 1.123758666597804e-06,
+      "loss": 0.3157,
+      "step": 11362
+    },
+    {
+      "epoch": 0.7889876406054714,
+      "grad_norm": 3.075223352038389,
+      "learning_rate": 1.1230484747416233e-06,
+      "loss": 0.1811,
+      "step": 11363
+    },
+    {
+      "epoch": 0.7890570754061936,
+      "grad_norm": 2.5720711961771316,
+      "learning_rate": 1.1223384789772656e-06,
+      "loss": 0.2618,
+      "step": 11364
+    },
+    {
+      "epoch": 0.7891265102069157,
+      "grad_norm": 3.402149815148726,
+      "learning_rate": 1.1216286793406378e-06,
+      "loss": 0.2784,
+      "step": 11365
+    },
+    {
+      "epoch": 0.7891959450076378,
+      "grad_norm": 5.136218681265127,
+      "learning_rate": 1.1209190758676436e-06,
+      "loss": 0.6989,
+      "step": 11366
+    },
+    {
+      "epoch": 0.78926537980836,
+      "grad_norm": 3.4413046765409376,
+      "learning_rate": 1.120209668594175e-06,
+      "loss": 0.2925,
+      "step": 11367
+    },
+    {
+      "epoch": 0.7893348146090821,
+      "grad_norm": 4.403634189919041,
+      "learning_rate": 1.1195004575561075e-06,
+      "loss": 0.4813,
+      "step": 11368
+    },
+    {
+      "epoch": 0.7894042494098042,
+      "grad_norm": 4.732150795275055,
+      "learning_rate": 1.1187914427893177e-06,
+      "loss": 0.5689,
+      "step": 11369
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 2.9846277896872633,
+      "learning_rate": 1.118082624329665e-06,
+      "loss": 0.3565,
+      "step": 11370
+    },
+    {
+      "epoch": 0.7895431190112484,
+      "grad_norm": 2.8662150701869322,
+      "learning_rate": 1.1173740022129998e-06,
+      "loss": 0.3794,
+      "step": 11371
+    },
+    {
+      "epoch": 0.7896125538119706,
+      "grad_norm": 4.214248067582184,
+      "learning_rate": 1.1166655764751648e-06,
+      "loss": 0.3747,
+      "step": 11372
+    },
+    {
+      "epoch": 0.7896819886126927,
+      "grad_norm": 3.7382274766875625,
+      "learning_rate": 1.1159573471519903e-06,
+      "loss": 0.3726,
+      "step": 11373
+    },
+    {
+      "epoch": 0.7897514234134148,
+      "grad_norm": 3.7965082916418997,
+      "learning_rate": 1.115249314279298e-06,
+      "loss": 0.3797,
+      "step": 11374
+    },
+    {
+      "epoch": 0.789820858214137,
+      "grad_norm": 3.3573902100911286,
+      "learning_rate": 1.1145414778928999e-06,
+      "loss": 0.3044,
+      "step": 11375
+    },
+    {
+      "epoch": 0.789890293014859,
+      "grad_norm": 3.485567226242866,
+      "learning_rate": 1.1138338380285969e-06,
+      "loss": 0.4248,
+      "step": 11376
+    },
+    {
+      "epoch": 0.7899597278155812,
+      "grad_norm": 3.709787352114896,
+      "learning_rate": 1.1131263947221792e-06,
+      "loss": 0.4199,
+      "step": 11377
+    },
+    {
+      "epoch": 0.7900291626163033,
+      "grad_norm": 2.241648028198475,
+      "learning_rate": 1.1124191480094331e-06,
+      "loss": 0.174,
+      "step": 11378
+    },
+    {
+      "epoch": 0.7900985974170254,
+      "grad_norm": 5.6160220069789775,
+      "learning_rate": 1.1117120979261264e-06,
+      "loss": 0.5943,
+      "step": 11379
+    },
+    {
+      "epoch": 0.7901680322177476,
+      "grad_norm": 3.0222569566143624,
+      "learning_rate": 1.1110052445080205e-06,
+      "loss": 0.336,
+      "step": 11380
+    },
+    {
+      "epoch": 0.7902374670184696,
+      "grad_norm": 5.799603260903549,
+      "learning_rate": 1.1102985877908707e-06,
+      "loss": 0.6521,
+      "step": 11381
+    },
+    {
+      "epoch": 0.7903069018191918,
+      "grad_norm": 5.239861649730828,
+      "learning_rate": 1.109592127810416e-06,
+      "loss": 0.6127,
+      "step": 11382
+    },
+    {
+      "epoch": 0.7903763366199139,
+      "grad_norm": 4.202794597479939,
+      "learning_rate": 1.1088858646023877e-06,
+      "loss": 0.4555,
+      "step": 11383
+    },
+    {
+      "epoch": 0.790445771420636,
+      "grad_norm": 5.2336135889429904,
+      "learning_rate": 1.1081797982025106e-06,
+      "loss": 0.639,
+      "step": 11384
+    },
+    {
+      "epoch": 0.7905152062213582,
+      "grad_norm": 5.447836132448847,
+      "learning_rate": 1.107473928646497e-06,
+      "loss": 0.3724,
+      "step": 11385
+    },
+    {
+      "epoch": 0.7905846410220803,
+      "grad_norm": 3.770882545042695,
+      "learning_rate": 1.1067682559700442e-06,
+      "loss": 0.393,
+      "step": 11386
+    },
+    {
+      "epoch": 0.7906540758228023,
+      "grad_norm": 5.588764071768843,
+      "learning_rate": 1.1060627802088497e-06,
+      "loss": 0.5672,
+      "step": 11387
+    },
+    {
+      "epoch": 0.7907235106235245,
+      "grad_norm": 5.192479853214447,
+      "learning_rate": 1.1053575013985929e-06,
+      "loss": 0.6306,
+      "step": 11388
+    },
+    {
+      "epoch": 0.7907929454242466,
+      "grad_norm": 3.221832961800813,
+      "learning_rate": 1.1046524195749463e-06,
+      "loss": 0.2934,
+      "step": 11389
+    },
+    {
+      "epoch": 0.7908623802249688,
+      "grad_norm": 4.050030552807425,
+      "learning_rate": 1.103947534773573e-06,
+      "loss": 0.479,
+      "step": 11390
+    },
+    {
+      "epoch": 0.7909318150256909,
+      "grad_norm": 3.056292858572746,
+      "learning_rate": 1.1032428470301243e-06,
+      "loss": 0.2265,
+      "step": 11391
+    },
+    {
+      "epoch": 0.791001249826413,
+      "grad_norm": 5.587868980434448,
+      "learning_rate": 1.1025383563802433e-06,
+      "loss": 0.3016,
+      "step": 11392
+    },
+    {
+      "epoch": 0.7910706846271351,
+      "grad_norm": 3.8111147807601755,
+      "learning_rate": 1.101834062859562e-06,
+      "loss": 0.3732,
+      "step": 11393
+    },
+    {
+      "epoch": 0.7911401194278572,
+      "grad_norm": 4.111932189847655,
+      "learning_rate": 1.1011299665037023e-06,
+      "loss": 0.4646,
+      "step": 11394
+    },
+    {
+      "epoch": 0.7912095542285794,
+      "grad_norm": 4.655663501246254,
+      "learning_rate": 1.100426067348276e-06,
+      "loss": 0.4728,
+      "step": 11395
+    },
+    {
+      "epoch": 0.7912789890293015,
+      "grad_norm": 4.342257963490471,
+      "learning_rate": 1.0997223654288892e-06,
+      "loss": 0.4918,
+      "step": 11396
+    },
+    {
+      "epoch": 0.7913484238300236,
+      "grad_norm": 4.022621557854533,
+      "learning_rate": 1.0990188607811287e-06,
+      "loss": 0.4864,
+      "step": 11397
+    },
+    {
+      "epoch": 0.7914178586307458,
+      "grad_norm": 4.223223567335567,
+      "learning_rate": 1.0983155534405816e-06,
+      "loss": 0.4457,
+      "step": 11398
+    },
+    {
+      "epoch": 0.7914872934314678,
+      "grad_norm": 4.853320475448637,
+      "learning_rate": 1.0976124434428204e-06,
+      "loss": 0.5384,
+      "step": 11399
+    },
+    {
+      "epoch": 0.7915567282321899,
+      "grad_norm": 3.062076961081114,
+      "learning_rate": 1.0969095308234023e-06,
+      "loss": 0.202,
+      "step": 11400
+    },
+    {
+      "epoch": 0.7916261630329121,
+      "grad_norm": 3.951845544263266,
+      "learning_rate": 1.0962068156178856e-06,
+      "loss": 0.5427,
+      "step": 11401
+    },
+    {
+      "epoch": 0.7916955978336342,
+      "grad_norm": 4.708205552598833,
+      "learning_rate": 1.0955042978618109e-06,
+      "loss": 0.7011,
+      "step": 11402
+    },
+    {
+      "epoch": 0.7917650326343564,
+      "grad_norm": 2.093864122244701,
+      "learning_rate": 1.09480197759071e-06,
+      "loss": 0.2048,
+      "step": 11403
+    },
+    {
+      "epoch": 0.7918344674350785,
+      "grad_norm": 11.81141771187345,
+      "learning_rate": 1.094099854840106e-06,
+      "loss": 0.3428,
+      "step": 11404
+    },
+    {
+      "epoch": 0.7919039022358005,
+      "grad_norm": 5.09072913830011,
+      "learning_rate": 1.0933979296455122e-06,
+      "loss": 0.4202,
+      "step": 11405
+    },
+    {
+      "epoch": 0.7919733370365227,
+      "grad_norm": 2.478239817528434,
+      "learning_rate": 1.0926962020424297e-06,
+      "loss": 0.2247,
+      "step": 11406
+    },
+    {
+      "epoch": 0.7920427718372448,
+      "grad_norm": 4.63469623575284,
+      "learning_rate": 1.0919946720663528e-06,
+      "loss": 0.4647,
+      "step": 11407
+    },
+    {
+      "epoch": 0.792112206637967,
+      "grad_norm": 3.532954161080818,
+      "learning_rate": 1.0912933397527625e-06,
+      "loss": 0.3081,
+      "step": 11408
+    },
+    {
+      "epoch": 0.7921816414386891,
+      "grad_norm": 3.9391903919124376,
+      "learning_rate": 1.090592205137131e-06,
+      "loss": 0.6705,
+      "step": 11409
+    },
+    {
+      "epoch": 0.7922510762394112,
+      "grad_norm": 4.012845833437784,
+      "learning_rate": 1.0898912682549245e-06,
+      "loss": 0.5821,
+      "step": 11410
+    },
+    {
+      "epoch": 0.7923205110401333,
+      "grad_norm": 3.5590552808185882,
+      "learning_rate": 1.0891905291415917e-06,
+      "loss": 0.2403,
+      "step": 11411
+    },
+    {
+      "epoch": 0.7923899458408554,
+      "grad_norm": 4.323847562142719,
+      "learning_rate": 1.088489987832575e-06,
+      "loss": 0.4675,
+      "step": 11412
+    },
+    {
+      "epoch": 0.7924593806415775,
+      "grad_norm": 4.749864319496748,
+      "learning_rate": 1.0877896443633118e-06,
+      "loss": 0.4445,
+      "step": 11413
+    },
+    {
+      "epoch": 0.7925288154422997,
+      "grad_norm": 4.231018638367464,
+      "learning_rate": 1.0870894987692198e-06,
+      "loss": 0.5556,
+      "step": 11414
+    },
+    {
+      "epoch": 0.7925982502430218,
+      "grad_norm": 3.434825712017066,
+      "learning_rate": 1.0863895510857119e-06,
+      "loss": 0.3795,
+      "step": 11415
+    },
+    {
+      "epoch": 0.792667685043744,
+      "grad_norm": 3.82233143463189,
+      "learning_rate": 1.0856898013481932e-06,
+      "loss": 0.3502,
+      "step": 11416
+    },
+    {
+      "epoch": 0.792737119844466,
+      "grad_norm": 3.3538319026804517,
+      "learning_rate": 1.084990249592055e-06,
+      "loss": 0.2732,
+      "step": 11417
+    },
+    {
+      "epoch": 0.7928065546451881,
+      "grad_norm": 4.087743133779592,
+      "learning_rate": 1.08429089585268e-06,
+      "loss": 0.3376,
+      "step": 11418
+    },
+    {
+      "epoch": 0.7928759894459103,
+      "grad_norm": 3.074319084527011,
+      "learning_rate": 1.0835917401654406e-06,
+      "loss": 0.4395,
+      "step": 11419
+    },
+    {
+      "epoch": 0.7929454242466324,
+      "grad_norm": 3.8185262962715427,
+      "learning_rate": 1.0828927825656987e-06,
+      "loss": 0.311,
+      "step": 11420
+    },
+    {
+      "epoch": 0.7930148590473546,
+      "grad_norm": 5.014484069464639,
+      "learning_rate": 1.0821940230888078e-06,
+      "loss": 0.5815,
+      "step": 11421
+    },
+    {
+      "epoch": 0.7930842938480767,
+      "grad_norm": 4.698885415166073,
+      "learning_rate": 1.0814954617701102e-06,
+      "loss": 0.5449,
+      "step": 11422
+    },
+    {
+      "epoch": 0.7931537286487987,
+      "grad_norm": 4.7217518995994485,
+      "learning_rate": 1.0807970986449378e-06,
+      "loss": 0.4832,
+      "step": 11423
+    },
+    {
+      "epoch": 0.7932231634495209,
+      "grad_norm": 3.672152802213499,
+      "learning_rate": 1.0800989337486129e-06,
+      "loss": 0.3005,
+      "step": 11424
+    },
+    {
+      "epoch": 0.793292598250243,
+      "grad_norm": 4.390104747309561,
+      "learning_rate": 1.0794009671164484e-06,
+      "loss": 0.5253,
+      "step": 11425
+    },
+    {
+      "epoch": 0.7933620330509652,
+      "grad_norm": 2.9158674915322536,
+      "learning_rate": 1.0787031987837455e-06,
+      "loss": 0.3402,
+      "step": 11426
+    },
+    {
+      "epoch": 0.7934314678516873,
+      "grad_norm": 3.8840060934394676,
+      "learning_rate": 1.0780056287857988e-06,
+      "loss": 0.3336,
+      "step": 11427
+    },
+    {
+      "epoch": 0.7935009026524094,
+      "grad_norm": 4.038687325193779,
+      "learning_rate": 1.0773082571578908e-06,
+      "loss": 0.4804,
+      "step": 11428
+    },
+    {
+      "epoch": 0.7935703374531315,
+      "grad_norm": 4.5654430627550004,
+      "learning_rate": 1.0766110839352894e-06,
+      "loss": 0.4577,
+      "step": 11429
+    },
+    {
+      "epoch": 0.7936397722538536,
+      "grad_norm": 2.203770933775792,
+      "learning_rate": 1.0759141091532615e-06,
+      "loss": 0.1397,
+      "step": 11430
+    },
+    {
+      "epoch": 0.7937092070545757,
+      "grad_norm": 4.95148022480219,
+      "learning_rate": 1.0752173328470582e-06,
+      "loss": 0.6316,
+      "step": 11431
+    },
+    {
+      "epoch": 0.7937786418552979,
+      "grad_norm": 3.2225166551481745,
+      "learning_rate": 1.0745207550519187e-06,
+      "loss": 0.461,
+      "step": 11432
+    },
+    {
+      "epoch": 0.79384807665602,
+      "grad_norm": 4.1472609778762415,
+      "learning_rate": 1.0738243758030791e-06,
+      "loss": 0.4408,
+      "step": 11433
+    },
+    {
+      "epoch": 0.7939175114567422,
+      "grad_norm": 2.3630805842852416,
+      "learning_rate": 1.0731281951357597e-06,
+      "loss": 0.2223,
+      "step": 11434
+    },
+    {
+      "epoch": 0.7939869462574642,
+      "grad_norm": 4.739536938339032,
+      "learning_rate": 1.0724322130851723e-06,
+      "loss": 0.407,
+      "step": 11435
+    },
+    {
+      "epoch": 0.7940563810581863,
+      "grad_norm": 4.022950296477599,
+      "learning_rate": 1.071736429686519e-06,
+      "loss": 0.3868,
+      "step": 11436
+    },
+    {
+      "epoch": 0.7941258158589085,
+      "grad_norm": 3.9784636895685215,
+      "learning_rate": 1.0710408449749927e-06,
+      "loss": 0.3614,
+      "step": 11437
+    },
+    {
+      "epoch": 0.7941952506596306,
+      "grad_norm": 4.914240697310668,
+      "learning_rate": 1.0703454589857743e-06,
+      "loss": 0.5893,
+      "step": 11438
+    },
+    {
+      "epoch": 0.7942646854603528,
+      "grad_norm": 3.047037942776805,
+      "learning_rate": 1.0696502717540353e-06,
+      "loss": 0.3215,
+      "step": 11439
+    },
+    {
+      "epoch": 0.7943341202610749,
+      "grad_norm": 3.604482979723412,
+      "learning_rate": 1.0689552833149387e-06,
+      "loss": 0.274,
+      "step": 11440
+    },
+    {
+      "epoch": 0.7944035550617969,
+      "grad_norm": 4.36668924882485,
+      "learning_rate": 1.068260493703634e-06,
+      "loss": 0.4154,
+      "step": 11441
+    },
+    {
+      "epoch": 0.7944729898625191,
+      "grad_norm": 2.8960732288298026,
+      "learning_rate": 1.0675659029552672e-06,
+      "loss": 0.2865,
+      "step": 11442
+    },
+    {
+      "epoch": 0.7945424246632412,
+      "grad_norm": 4.29806870705308,
+      "learning_rate": 1.066871511104966e-06,
+      "loss": 0.4564,
+      "step": 11443
+    },
+    {
+      "epoch": 0.7946118594639633,
+      "grad_norm": 4.257517142878325,
+      "learning_rate": 1.0661773181878515e-06,
+      "loss": 0.3435,
+      "step": 11444
+    },
+    {
+      "epoch": 0.7946812942646855,
+      "grad_norm": 3.5325034300728864,
+      "learning_rate": 1.0654833242390383e-06,
+      "loss": 0.3564,
+      "step": 11445
+    },
+    {
+      "epoch": 0.7947507290654076,
+      "grad_norm": 2.9374752219689126,
+      "learning_rate": 1.064789529293626e-06,
+      "loss": 0.2635,
+      "step": 11446
+    },
+    {
+      "epoch": 0.7948201638661297,
+      "grad_norm": 3.7010546671834534,
+      "learning_rate": 1.0640959333867068e-06,
+      "loss": 0.3089,
+      "step": 11447
+    },
+    {
+      "epoch": 0.7948895986668518,
+      "grad_norm": 3.4616129455471465,
+      "learning_rate": 1.0634025365533617e-06,
+      "loss": 0.3069,
+      "step": 11448
+    },
+    {
+      "epoch": 0.7949590334675739,
+      "grad_norm": 4.411756599817269,
+      "learning_rate": 1.0627093388286614e-06,
+      "loss": 0.3523,
+      "step": 11449
+    },
+    {
+      "epoch": 0.7950284682682961,
+      "grad_norm": 7.70337993809111,
+      "learning_rate": 1.0620163402476675e-06,
+      "loss": 0.6194,
+      "step": 11450
+    },
+    {
+      "epoch": 0.7950979030690182,
+      "grad_norm": 4.783363352291953,
+      "learning_rate": 1.061323540845431e-06,
+      "loss": 0.6487,
+      "step": 11451
+    },
+    {
+      "epoch": 0.7951673378697404,
+      "grad_norm": 4.784237460139205,
+      "learning_rate": 1.0606309406569925e-06,
+      "loss": 0.4299,
+      "step": 11452
+    },
+    {
+      "epoch": 0.7952367726704624,
+      "grad_norm": 3.9141149635222896,
+      "learning_rate": 1.0599385397173833e-06,
+      "loss": 0.3801,
+      "step": 11453
+    },
+    {
+      "epoch": 0.7953062074711845,
+      "grad_norm": 4.249914157108824,
+      "learning_rate": 1.0592463380616247e-06,
+      "loss": 0.5519,
+      "step": 11454
+    },
+    {
+      "epoch": 0.7953756422719067,
+      "grad_norm": 3.2260450781861127,
+      "learning_rate": 1.0585543357247258e-06,
+      "loss": 0.3482,
+      "step": 11455
+    },
+    {
+      "epoch": 0.7954450770726288,
+      "grad_norm": 3.8245202742903097,
+      "learning_rate": 1.0578625327416907e-06,
+      "loss": 0.4916,
+      "step": 11456
+    },
+    {
+      "epoch": 0.7955145118733509,
+      "grad_norm": 3.276290151244739,
+      "learning_rate": 1.0571709291475073e-06,
+      "loss": 0.1266,
+      "step": 11457
+    },
+    {
+      "epoch": 0.7955839466740731,
+      "grad_norm": 3.9858262208845963,
+      "learning_rate": 1.0564795249771548e-06,
+      "loss": 0.4636,
+      "step": 11458
+    },
+    {
+      "epoch": 0.7956533814747951,
+      "grad_norm": 3.2241374148736157,
+      "learning_rate": 1.0557883202656072e-06,
+      "loss": 0.4098,
+      "step": 11459
+    },
+    {
+      "epoch": 0.7957228162755173,
+      "grad_norm": 3.721026150061444,
+      "learning_rate": 1.0550973150478239e-06,
+      "loss": 0.3027,
+      "step": 11460
+    },
+    {
+      "epoch": 0.7957922510762394,
+      "grad_norm": 4.027021454197483,
+      "learning_rate": 1.0544065093587524e-06,
+      "loss": 0.5003,
+      "step": 11461
+    },
+    {
+      "epoch": 0.7958616858769615,
+      "grad_norm": 3.0545596024797854,
+      "learning_rate": 1.0537159032333361e-06,
+      "loss": 0.3103,
+      "step": 11462
+    },
+    {
+      "epoch": 0.7959311206776837,
+      "grad_norm": 3.287583059696945,
+      "learning_rate": 1.0530254967065052e-06,
+      "loss": 0.2832,
+      "step": 11463
+    },
+    {
+      "epoch": 0.7960005554784058,
+      "grad_norm": 2.809480189484644,
+      "learning_rate": 1.0523352898131756e-06,
+      "loss": 0.2993,
+      "step": 11464
+    },
+    {
+      "epoch": 0.796069990279128,
+      "grad_norm": 5.294044879190656,
+      "learning_rate": 1.0516452825882612e-06,
+      "loss": 0.5479,
+      "step": 11465
+    },
+    {
+      "epoch": 0.79613942507985,
+      "grad_norm": 3.2006299585355977,
+      "learning_rate": 1.0509554750666606e-06,
+      "loss": 0.2543,
+      "step": 11466
+    },
+    {
+      "epoch": 0.7962088598805721,
+      "grad_norm": 4.325927957459826,
+      "learning_rate": 1.0502658672832634e-06,
+      "loss": 0.4231,
+      "step": 11467
+    },
+    {
+      "epoch": 0.7962782946812943,
+      "grad_norm": 3.9752954683958612,
+      "learning_rate": 1.0495764592729495e-06,
+      "loss": 0.4619,
+      "step": 11468
+    },
+    {
+      "epoch": 0.7963477294820164,
+      "grad_norm": 3.118831831158478,
+      "learning_rate": 1.0488872510705878e-06,
+      "loss": 0.1781,
+      "step": 11469
+    },
+    {
+      "epoch": 0.7964171642827385,
+      "grad_norm": 3.448943107949009,
+      "learning_rate": 1.0481982427110365e-06,
+      "loss": 0.3753,
+      "step": 11470
+    },
+    {
+      "epoch": 0.7964865990834606,
+      "grad_norm": 3.8826892168531795,
+      "learning_rate": 1.0475094342291491e-06,
+      "loss": 0.2379,
+      "step": 11471
+    },
+    {
+      "epoch": 0.7965560338841827,
+      "grad_norm": 4.744012162350741,
+      "learning_rate": 1.046820825659761e-06,
+      "loss": 0.6411,
+      "step": 11472
+    },
+    {
+      "epoch": 0.7966254686849049,
+      "grad_norm": 6.792380990464082,
+      "learning_rate": 1.0461324170377008e-06,
+      "loss": 0.4343,
+      "step": 11473
+    },
+    {
+      "epoch": 0.796694903485627,
+      "grad_norm": 2.177860755370552,
+      "learning_rate": 1.045444208397791e-06,
+      "loss": 0.1457,
+      "step": 11474
+    },
+    {
+      "epoch": 0.7967643382863491,
+      "grad_norm": 3.594035678903114,
+      "learning_rate": 1.0447561997748373e-06,
+      "loss": 0.3707,
+      "step": 11475
+    },
+    {
+      "epoch": 0.7968337730870713,
+      "grad_norm": 3.721335180375672,
+      "learning_rate": 1.044068391203638e-06,
+      "loss": 0.3936,
+      "step": 11476
+    },
+    {
+      "epoch": 0.7969032078877933,
+      "grad_norm": 3.8675002859292134,
+      "learning_rate": 1.0433807827189846e-06,
+      "loss": 0.4478,
+      "step": 11477
+    },
+    {
+      "epoch": 0.7969726426885155,
+      "grad_norm": 3.5134327960748286,
+      "learning_rate": 1.042693374355654e-06,
+      "loss": 0.2533,
+      "step": 11478
+    },
+    {
+      "epoch": 0.7970420774892376,
+      "grad_norm": 3.8628064983866897,
+      "learning_rate": 1.0420061661484138e-06,
+      "loss": 0.5374,
+      "step": 11479
+    },
+    {
+      "epoch": 0.7971115122899597,
+      "grad_norm": 3.78714073570003,
+      "learning_rate": 1.0413191581320238e-06,
+      "loss": 0.4324,
+      "step": 11480
+    },
+    {
+      "epoch": 0.7971809470906819,
+      "grad_norm": 4.172901718676249,
+      "learning_rate": 1.0406323503412307e-06,
+      "loss": 0.4519,
+      "step": 11481
+    },
+    {
+      "epoch": 0.797250381891404,
+      "grad_norm": 3.8530761067003554,
+      "learning_rate": 1.0399457428107728e-06,
+      "loss": 0.4426,
+      "step": 11482
+    },
+    {
+      "epoch": 0.7973198166921261,
+      "grad_norm": 3.0127014281714994,
+      "learning_rate": 1.0392593355753788e-06,
+      "loss": 0.1788,
+      "step": 11483
+    },
+    {
+      "epoch": 0.7973892514928482,
+      "grad_norm": 4.691397657438658,
+      "learning_rate": 1.0385731286697653e-06,
+      "loss": 0.4566,
+      "step": 11484
+    },
+    {
+      "epoch": 0.7974586862935703,
+      "grad_norm": 4.711840683300377,
+      "learning_rate": 1.03788712212864e-06,
+      "loss": 0.6133,
+      "step": 11485
+    },
+    {
+      "epoch": 0.7975281210942925,
+      "grad_norm": 2.3018658962822935,
+      "learning_rate": 1.037201315986701e-06,
+      "loss": 0.1805,
+      "step": 11486
+    },
+    {
+      "epoch": 0.7975975558950146,
+      "grad_norm": 4.584858143115192,
+      "learning_rate": 1.0365157102786338e-06,
+      "loss": 0.6429,
+      "step": 11487
+    },
+    {
+      "epoch": 0.7976669906957367,
+      "grad_norm": 3.9788503299160687,
+      "learning_rate": 1.03583030503912e-06,
+      "loss": 0.597,
+      "step": 11488
+    },
+    {
+      "epoch": 0.7977364254964588,
+      "grad_norm": 4.3635188043688276,
+      "learning_rate": 1.0351451003028223e-06,
+      "loss": 0.5098,
+      "step": 11489
+    },
+    {
+      "epoch": 0.7978058602971809,
+      "grad_norm": 5.0342324313252655,
+      "learning_rate": 1.0344600961043971e-06,
+      "loss": 0.3083,
+      "step": 11490
+    },
+    {
+      "epoch": 0.7978752950979031,
+      "grad_norm": 4.044708962535251,
+      "learning_rate": 1.0337752924784943e-06,
+      "loss": 0.4044,
+      "step": 11491
+    },
+    {
+      "epoch": 0.7979447298986252,
+      "grad_norm": 4.3344586660248865,
+      "learning_rate": 1.033090689459751e-06,
+      "loss": 0.4512,
+      "step": 11492
+    },
+    {
+      "epoch": 0.7980141646993473,
+      "grad_norm": 3.5979046655383016,
+      "learning_rate": 1.0324062870827883e-06,
+      "loss": 0.3249,
+      "step": 11493
+    },
+    {
+      "epoch": 0.7980835995000695,
+      "grad_norm": 3.399477252365134,
+      "learning_rate": 1.031722085382228e-06,
+      "loss": 0.2209,
+      "step": 11494
+    },
+    {
+      "epoch": 0.7981530343007915,
+      "grad_norm": 4.676724002479445,
+      "learning_rate": 1.0310380843926737e-06,
+      "loss": 0.3036,
+      "step": 11495
+    },
+    {
+      "epoch": 0.7982224691015137,
+      "grad_norm": 4.392197971796721,
+      "learning_rate": 1.030354284148722e-06,
+      "loss": 0.4234,
+      "step": 11496
+    },
+    {
+      "epoch": 0.7982919039022358,
+      "grad_norm": 5.141818290355714,
+      "learning_rate": 1.0296706846849591e-06,
+      "loss": 0.7346,
+      "step": 11497
+    },
+    {
+      "epoch": 0.7983613387029579,
+      "grad_norm": 4.822164739445092,
+      "learning_rate": 1.0289872860359596e-06,
+      "loss": 0.3313,
+      "step": 11498
+    },
+    {
+      "epoch": 0.7984307735036801,
+      "grad_norm": 4.368802795018902,
+      "learning_rate": 1.02830408823629e-06,
+      "loss": 0.3877,
+      "step": 11499
+    },
+    {
+      "epoch": 0.7985002083044022,
+      "grad_norm": 3.3784158894313605,
+      "learning_rate": 1.0276210913205053e-06,
+      "loss": 0.3393,
+      "step": 11500
+    },
+    {
+      "epoch": 0.7985696431051242,
+      "grad_norm": 4.072258572738818,
+      "learning_rate": 1.0269382953231504e-06,
+      "loss": 0.3492,
+      "step": 11501
+    },
+    {
+      "epoch": 0.7986390779058464,
+      "grad_norm": 15.887596741049208,
+      "learning_rate": 1.0262557002787599e-06,
+      "loss": 0.5611,
+      "step": 11502
+    },
+    {
+      "epoch": 0.7987085127065685,
+      "grad_norm": 5.861608965646581,
+      "learning_rate": 1.025573306221862e-06,
+      "loss": 0.7034,
+      "step": 11503
+    },
+    {
+      "epoch": 0.7987779475072907,
+      "grad_norm": 4.4587152840849225,
+      "learning_rate": 1.0248911131869677e-06,
+      "loss": 0.6109,
+      "step": 11504
+    },
+    {
+      "epoch": 0.7988473823080128,
+      "grad_norm": 2.9232580904068626,
+      "learning_rate": 1.0242091212085815e-06,
+      "loss": 0.204,
+      "step": 11505
+    },
+    {
+      "epoch": 0.7989168171087349,
+      "grad_norm": 3.325370405536152,
+      "learning_rate": 1.0235273303212013e-06,
+      "loss": 0.3579,
+      "step": 11506
+    },
+    {
+      "epoch": 0.798986251909457,
+      "grad_norm": 2.556114828205371,
+      "learning_rate": 1.0228457405593061e-06,
+      "loss": 0.2057,
+      "step": 11507
+    },
+    {
+      "epoch": 0.7990556867101791,
+      "grad_norm": 3.768179106847722,
+      "learning_rate": 1.0221643519573748e-06,
+      "loss": 0.4156,
+      "step": 11508
+    },
+    {
+      "epoch": 0.7991251215109013,
+      "grad_norm": 4.034683191295604,
+      "learning_rate": 1.0214831645498697e-06,
+      "loss": 0.3954,
+      "step": 11509
+    },
+    {
+      "epoch": 0.7991945563116234,
+      "grad_norm": 4.328890217216435,
+      "learning_rate": 1.0208021783712434e-06,
+      "loss": 0.697,
+      "step": 11510
+    },
+    {
+      "epoch": 0.7992639911123455,
+      "grad_norm": 3.7213986630121294,
+      "learning_rate": 1.02012139345594e-06,
+      "loss": 0.3362,
+      "step": 11511
+    },
+    {
+      "epoch": 0.7993334259130677,
+      "grad_norm": 4.239830904993366,
+      "learning_rate": 1.019440809838394e-06,
+      "loss": 0.4021,
+      "step": 11512
+    },
+    {
+      "epoch": 0.7994028607137897,
+      "grad_norm": 2.9098737932753584,
+      "learning_rate": 1.0187604275530266e-06,
+      "loss": 0.1892,
+      "step": 11513
+    },
+    {
+      "epoch": 0.7994722955145118,
+      "grad_norm": 3.6436368058370414,
+      "learning_rate": 1.018080246634252e-06,
+      "loss": 0.3348,
+      "step": 11514
+    },
+    {
+      "epoch": 0.799541730315234,
+      "grad_norm": 4.380265032978523,
+      "learning_rate": 1.017400267116473e-06,
+      "loss": 0.4809,
+      "step": 11515
+    },
+    {
+      "epoch": 0.7996111651159561,
+      "grad_norm": 4.453844726811093,
+      "learning_rate": 1.0167204890340798e-06,
+      "loss": 0.4727,
+      "step": 11516
+    },
+    {
+      "epoch": 0.7996805999166783,
+      "grad_norm": 3.3128742035198178,
+      "learning_rate": 1.0160409124214604e-06,
+      "loss": 0.2246,
+      "step": 11517
+    },
+    {
+      "epoch": 0.7997500347174004,
+      "grad_norm": 3.4579487771258606,
+      "learning_rate": 1.0153615373129816e-06,
+      "loss": 0.4841,
+      "step": 11518
+    },
+    {
+      "epoch": 0.7998194695181224,
+      "grad_norm": 4.023629012495005,
+      "learning_rate": 1.014682363743006e-06,
+      "loss": 0.3353,
+      "step": 11519
+    },
+    {
+      "epoch": 0.7998889043188446,
+      "grad_norm": 3.0430128863745183,
+      "learning_rate": 1.014003391745888e-06,
+      "loss": 0.2903,
+      "step": 11520
+    },
+    {
+      "epoch": 0.7999583391195667,
+      "grad_norm": 3.2775009879398103,
+      "learning_rate": 1.01332462135597e-06,
+      "loss": 0.3423,
+      "step": 11521
+    },
+    {
+      "epoch": 0.8000277739202889,
+      "grad_norm": 4.5654926420445845,
+      "learning_rate": 1.0126460526075782e-06,
+      "loss": 0.4997,
+      "step": 11522
+    },
+    {
+      "epoch": 0.800097208721011,
+      "grad_norm": 3.821257671084501,
+      "learning_rate": 1.0119676855350385e-06,
+      "loss": 0.342,
+      "step": 11523
+    },
+    {
+      "epoch": 0.8001666435217331,
+      "grad_norm": 3.0939520133261023,
+      "learning_rate": 1.011289520172662e-06,
+      "loss": 0.2798,
+      "step": 11524
+    },
+    {
+      "epoch": 0.8002360783224552,
+      "grad_norm": 4.775305551151815,
+      "learning_rate": 1.0106115565547446e-06,
+      "loss": 0.6606,
+      "step": 11525
+    },
+    {
+      "epoch": 0.8003055131231773,
+      "grad_norm": 3.3817870372304797,
+      "learning_rate": 1.0099337947155824e-06,
+      "loss": 0.4102,
+      "step": 11526
+    },
+    {
+      "epoch": 0.8003749479238994,
+      "grad_norm": 5.530806245459377,
+      "learning_rate": 1.0092562346894536e-06,
+      "loss": 0.5367,
+      "step": 11527
+    },
+    {
+      "epoch": 0.8004443827246216,
+      "grad_norm": 5.0106588370089575,
+      "learning_rate": 1.0085788765106291e-06,
+      "loss": 0.4985,
+      "step": 11528
+    },
+    {
+      "epoch": 0.8005138175253437,
+      "grad_norm": 3.1802391231871656,
+      "learning_rate": 1.0079017202133678e-06,
+      "loss": 0.2264,
+      "step": 11529
+    },
+    {
+      "epoch": 0.8005832523260659,
+      "grad_norm": 3.941094362183221,
+      "learning_rate": 1.0072247658319206e-06,
+      "loss": 0.441,
+      "step": 11530
+    },
+    {
+      "epoch": 0.8006526871267879,
+      "grad_norm": 3.5782637410800726,
+      "learning_rate": 1.0065480134005261e-06,
+      "loss": 0.3526,
+      "step": 11531
+    },
+    {
+      "epoch": 0.80072212192751,
+      "grad_norm": 4.660680256942138,
+      "learning_rate": 1.0058714629534144e-06,
+      "loss": 0.4544,
+      "step": 11532
+    },
+    {
+      "epoch": 0.8007915567282322,
+      "grad_norm": 4.0151450694063,
+      "learning_rate": 1.0051951145248047e-06,
+      "loss": 0.484,
+      "step": 11533
+    },
+    {
+      "epoch": 0.8008609915289543,
+      "grad_norm": 2.6722349607481624,
+      "learning_rate": 1.0045189681489042e-06,
+      "loss": 0.2137,
+      "step": 11534
+    },
+    {
+      "epoch": 0.8009304263296765,
+      "grad_norm": 2.5746218248864623,
+      "learning_rate": 1.0038430238599156e-06,
+      "loss": 0.169,
+      "step": 11535
+    },
+    {
+      "epoch": 0.8009998611303986,
+      "grad_norm": 3.912904184170569,
+      "learning_rate": 1.0031672816920223e-06,
+      "loss": 0.4779,
+      "step": 11536
+    },
+    {
+      "epoch": 0.8010692959311206,
+      "grad_norm": 3.8829982191520056,
+      "learning_rate": 1.0024917416794068e-06,
+      "loss": 0.4297,
+      "step": 11537
+    },
+    {
+      "epoch": 0.8011387307318428,
+      "grad_norm": 4.113685857081385,
+      "learning_rate": 1.0018164038562362e-06,
+      "loss": 0.401,
+      "step": 11538
+    },
+    {
+      "epoch": 0.8012081655325649,
+      "grad_norm": 3.3450687529916854,
+      "learning_rate": 1.0011412682566656e-06,
+      "loss": 0.348,
+      "step": 11539
+    },
+    {
+      "epoch": 0.8012776003332871,
+      "grad_norm": 4.914813644910723,
+      "learning_rate": 1.0004663349148458e-06,
+      "loss": 0.5854,
+      "step": 11540
+    },
+    {
+      "epoch": 0.8013470351340092,
+      "grad_norm": 3.8401678616989856,
+      "learning_rate": 9.99791603864913e-07,
+      "loss": 0.3447,
+      "step": 11541
+    },
+    {
+      "epoch": 0.8014164699347313,
+      "grad_norm": 6.41928258122755,
+      "learning_rate": 9.99117075140994e-07,
+      "loss": 0.8809,
+      "step": 11542
+    },
+    {
+      "epoch": 0.8014859047354534,
+      "grad_norm": 5.16749519424555,
+      "learning_rate": 9.984427487772064e-07,
+      "loss": 0.4863,
+      "step": 11543
+    },
+    {
+      "epoch": 0.8015553395361755,
+      "grad_norm": 3.166091891179892,
+      "learning_rate": 9.977686248076562e-07,
+      "loss": 0.3965,
+      "step": 11544
+    },
+    {
+      "epoch": 0.8016247743368976,
+      "grad_norm": 10.046057963649357,
+      "learning_rate": 9.970947032664407e-07,
+      "loss": 0.6321,
+      "step": 11545
+    },
+    {
+      "epoch": 0.8016942091376198,
+      "grad_norm": 4.103893759892576,
+      "learning_rate": 9.96420984187645e-07,
+      "loss": 0.4439,
+      "step": 11546
+    },
+    {
+      "epoch": 0.8017636439383419,
+      "grad_norm": 3.477348990739748,
+      "learning_rate": 9.957474676053458e-07,
+      "loss": 0.4188,
+      "step": 11547
+    },
+    {
+      "epoch": 0.8018330787390641,
+      "grad_norm": 3.191902572693076,
+      "learning_rate": 9.950741535536073e-07,
+      "loss": 0.2284,
+      "step": 11548
+    },
+    {
+      "epoch": 0.8019025135397861,
+      "grad_norm": 2.91173398693346,
+      "learning_rate": 9.944010420664884e-07,
+      "loss": 0.2442,
+      "step": 11549
+    },
+    {
+      "epoch": 0.8019719483405082,
+      "grad_norm": 3.7648332715260118,
+      "learning_rate": 9.937281331780313e-07,
+      "loss": 0.267,
+      "step": 11550
+    },
+    {
+      "epoch": 0.8020413831412304,
+      "grad_norm": 3.135443969655885,
+      "learning_rate": 9.930554269222704e-07,
+      "loss": 0.3643,
+      "step": 11551
+    },
+    {
+      "epoch": 0.8021108179419525,
+      "grad_norm": 3.3438386090964114,
+      "learning_rate": 9.923829233332334e-07,
+      "loss": 0.2683,
+      "step": 11552
+    },
+    {
+      "epoch": 0.8021802527426747,
+      "grad_norm": 4.194513367440136,
+      "learning_rate": 9.91710622444934e-07,
+      "loss": 0.3672,
+      "step": 11553
+    },
+    {
+      "epoch": 0.8022496875433968,
+      "grad_norm": 7.253803188072973,
+      "learning_rate": 9.910385242913733e-07,
+      "loss": 0.5058,
+      "step": 11554
+    },
+    {
+      "epoch": 0.8023191223441188,
+      "grad_norm": 3.1316357152417953,
+      "learning_rate": 9.903666289065488e-07,
+      "loss": 0.3237,
+      "step": 11555
+    },
+    {
+      "epoch": 0.802388557144841,
+      "grad_norm": 3.615172873264802,
+      "learning_rate": 9.896949363244446e-07,
+      "loss": 0.3937,
+      "step": 11556
+    },
+    {
+      "epoch": 0.8024579919455631,
+      "grad_norm": 4.791242084283195,
+      "learning_rate": 9.890234465790293e-07,
+      "loss": 0.6589,
+      "step": 11557
+    },
+    {
+      "epoch": 0.8025274267462852,
+      "grad_norm": 3.2256680379862277,
+      "learning_rate": 9.883521597042712e-07,
+      "loss": 0.3,
+      "step": 11558
+    },
+    {
+      "epoch": 0.8025968615470074,
+      "grad_norm": 4.313426235128009,
+      "learning_rate": 9.876810757341215e-07,
+      "loss": 0.6012,
+      "step": 11559
+    },
+    {
+      "epoch": 0.8026662963477295,
+      "grad_norm": 3.8489306954683267,
+      "learning_rate": 9.870101947025223e-07,
+      "loss": 0.3325,
+      "step": 11560
+    },
+    {
+      "epoch": 0.8027357311484516,
+      "grad_norm": 4.371860127734536,
+      "learning_rate": 9.863395166434064e-07,
+      "loss": 0.4007,
+      "step": 11561
+    },
+    {
+      "epoch": 0.8028051659491737,
+      "grad_norm": 3.82734081900594,
+      "learning_rate": 9.856690415906966e-07,
+      "loss": 0.403,
+      "step": 11562
+    },
+    {
+      "epoch": 0.8028746007498958,
+      "grad_norm": 3.147724113317814,
+      "learning_rate": 9.849987695783037e-07,
+      "loss": 0.1802,
+      "step": 11563
+    },
+    {
+      "epoch": 0.802944035550618,
+      "grad_norm": 4.6390092822246425,
+      "learning_rate": 9.843287006401297e-07,
+      "loss": 0.5143,
+      "step": 11564
+    },
+    {
+      "epoch": 0.8030134703513401,
+      "grad_norm": 2.8054001100371453,
+      "learning_rate": 9.836588348100663e-07,
+      "loss": 0.1695,
+      "step": 11565
+    },
+    {
+      "epoch": 0.8030829051520623,
+      "grad_norm": 3.754858071446223,
+      "learning_rate": 9.829891721219925e-07,
+      "loss": 0.2738,
+      "step": 11566
+    },
+    {
+      "epoch": 0.8031523399527843,
+      "grad_norm": 3.9757196636384453,
+      "learning_rate": 9.82319712609784e-07,
+      "loss": 0.5841,
+      "step": 11567
+    },
+    {
+      "epoch": 0.8032217747535064,
+      "grad_norm": 4.225815402513942,
+      "learning_rate": 9.816504563072948e-07,
+      "loss": 0.4895,
+      "step": 11568
+    },
+    {
+      "epoch": 0.8032912095542286,
+      "grad_norm": 4.015230639118562,
+      "learning_rate": 9.8098140324838e-07,
+      "loss": 0.648,
+      "step": 11569
+    },
+    {
+      "epoch": 0.8033606443549507,
+      "grad_norm": 3.353840486006697,
+      "learning_rate": 9.803125534668783e-07,
+      "loss": 0.2915,
+      "step": 11570
+    },
+    {
+      "epoch": 0.8034300791556728,
+      "grad_norm": 6.491814703848902,
+      "learning_rate": 9.796439069966191e-07,
+      "loss": 0.5784,
+      "step": 11571
+    },
+    {
+      "epoch": 0.803499513956395,
+      "grad_norm": 4.137385459773879,
+      "learning_rate": 9.78975463871421e-07,
+      "loss": 0.5446,
+      "step": 11572
+    },
+    {
+      "epoch": 0.803568948757117,
+      "grad_norm": 3.8500082169425127,
+      "learning_rate": 9.783072241250947e-07,
+      "loss": 0.3533,
+      "step": 11573
+    },
+    {
+      "epoch": 0.8036383835578392,
+      "grad_norm": 3.4543491846657464,
+      "learning_rate": 9.776391877914376e-07,
+      "loss": 0.4153,
+      "step": 11574
+    },
+    {
+      "epoch": 0.8037078183585613,
+      "grad_norm": 5.328030650704951,
+      "learning_rate": 9.769713549042392e-07,
+      "loss": 0.5232,
+      "step": 11575
+    },
+    {
+      "epoch": 0.8037772531592834,
+      "grad_norm": 4.078535209694497,
+      "learning_rate": 9.763037254972773e-07,
+      "loss": 0.6775,
+      "step": 11576
+    },
+    {
+      "epoch": 0.8038466879600056,
+      "grad_norm": 5.0549122117999294,
+      "learning_rate": 9.756362996043183e-07,
+      "loss": 0.6467,
+      "step": 11577
+    },
+    {
+      "epoch": 0.8039161227607277,
+      "grad_norm": 3.251974425842952,
+      "learning_rate": 9.749690772591243e-07,
+      "loss": 0.4418,
+      "step": 11578
+    },
+    {
+      "epoch": 0.8039855575614498,
+      "grad_norm": 3.1682921672171322,
+      "learning_rate": 9.743020584954376e-07,
+      "loss": 0.2992,
+      "step": 11579
+    },
+    {
+      "epoch": 0.8040549923621719,
+      "grad_norm": 4.30418766235949,
+      "learning_rate": 9.73635243346997e-07,
+      "loss": 0.4772,
+      "step": 11580
+    },
+    {
+      "epoch": 0.804124427162894,
+      "grad_norm": 3.6000691224263774,
+      "learning_rate": 9.729686318475317e-07,
+      "loss": 0.3605,
+      "step": 11581
+    },
+    {
+      "epoch": 0.8041938619636162,
+      "grad_norm": 4.560802425802625,
+      "learning_rate": 9.72302224030755e-07,
+      "loss": 0.4419,
+      "step": 11582
+    },
+    {
+      "epoch": 0.8042632967643383,
+      "grad_norm": 3.786641925045908,
+      "learning_rate": 9.716360199303732e-07,
+      "loss": 0.517,
+      "step": 11583
+    },
+    {
+      "epoch": 0.8043327315650604,
+      "grad_norm": 4.234162751089818,
+      "learning_rate": 9.709700195800836e-07,
+      "loss": 0.2147,
+      "step": 11584
+    },
+    {
+      "epoch": 0.8044021663657825,
+      "grad_norm": 3.7024765671675364,
+      "learning_rate": 9.703042230135735e-07,
+      "loss": 0.4591,
+      "step": 11585
+    },
+    {
+      "epoch": 0.8044716011665046,
+      "grad_norm": 3.603791999534332,
+      "learning_rate": 9.696386302645127e-07,
+      "loss": 0.4421,
+      "step": 11586
+    },
+    {
+      "epoch": 0.8045410359672268,
+      "grad_norm": 5.369838582317884,
+      "learning_rate": 9.689732413665709e-07,
+      "loss": 0.607,
+      "step": 11587
+    },
+    {
+      "epoch": 0.8046104707679489,
+      "grad_norm": 4.3393935682961935,
+      "learning_rate": 9.68308056353401e-07,
+      "loss": 0.5234,
+      "step": 11588
+    },
+    {
+      "epoch": 0.804679905568671,
+      "grad_norm": 3.0287443440575013,
+      "learning_rate": 9.676430752586474e-07,
+      "loss": 0.2969,
+      "step": 11589
+    },
+    {
+      "epoch": 0.8047493403693932,
+      "grad_norm": 1.7074565822072896,
+      "learning_rate": 9.669782981159443e-07,
+      "loss": 0.102,
+      "step": 11590
+    },
+    {
+      "epoch": 0.8048187751701152,
+      "grad_norm": 3.9643302225915886,
+      "learning_rate": 9.66313724958915e-07,
+      "loss": 0.5384,
+      "step": 11591
+    },
+    {
+      "epoch": 0.8048882099708374,
+      "grad_norm": 3.6072803223760084,
+      "learning_rate": 9.656493558211731e-07,
+      "loss": 0.3078,
+      "step": 11592
+    },
+    {
+      "epoch": 0.8049576447715595,
+      "grad_norm": 3.7011261097383743,
+      "learning_rate": 9.649851907363216e-07,
+      "loss": 0.3425,
+      "step": 11593
+    },
+    {
+      "epoch": 0.8050270795722816,
+      "grad_norm": 4.8053510885454,
+      "learning_rate": 9.643212297379534e-07,
+      "loss": 0.5175,
+      "step": 11594
+    },
+    {
+      "epoch": 0.8050965143730038,
+      "grad_norm": 5.496978489531954,
+      "learning_rate": 9.636574728596504e-07,
+      "loss": 0.8954,
+      "step": 11595
+    },
+    {
+      "epoch": 0.8051659491737259,
+      "grad_norm": 3.6784338389001436,
+      "learning_rate": 9.629939201349852e-07,
+      "loss": 0.2611,
+      "step": 11596
+    },
+    {
+      "epoch": 0.8052353839744479,
+      "grad_norm": 3.6954787107595557,
+      "learning_rate": 9.62330571597518e-07,
+      "loss": 0.3939,
+      "step": 11597
+    },
+    {
+      "epoch": 0.8053048187751701,
+      "grad_norm": 3.718111944899777,
+      "learning_rate": 9.61667427280803e-07,
+      "loss": 0.3948,
+      "step": 11598
+    },
+    {
+      "epoch": 0.8053742535758922,
+      "grad_norm": 3.864399976950323,
+      "learning_rate": 9.61004487218381e-07,
+      "loss": 0.5318,
+      "step": 11599
+    },
+    {
+      "epoch": 0.8054436883766144,
+      "grad_norm": 5.548366395004784,
+      "learning_rate": 9.60341751443779e-07,
+      "loss": 0.5521,
+      "step": 11600
+    },
+    {
+      "epoch": 0.8055131231773365,
+      "grad_norm": 3.8723173766577177,
+      "learning_rate": 9.596792199905214e-07,
+      "loss": 0.3937,
+      "step": 11601
+    },
+    {
+      "epoch": 0.8055825579780586,
+      "grad_norm": 5.539681092755038,
+      "learning_rate": 9.590168928921168e-07,
+      "loss": 0.3535,
+      "step": 11602
+    },
+    {
+      "epoch": 0.8056519927787807,
+      "grad_norm": 4.624906126614333,
+      "learning_rate": 9.583547701820651e-07,
+      "loss": 0.6208,
+      "step": 11603
+    },
+    {
+      "epoch": 0.8057214275795028,
+      "grad_norm": 3.9509392130189913,
+      "learning_rate": 9.57692851893856e-07,
+      "loss": 0.1438,
+      "step": 11604
+    },
+    {
+      "epoch": 0.805790862380225,
+      "grad_norm": 5.576258915275627,
+      "learning_rate": 9.570311380609675e-07,
+      "loss": 0.932,
+      "step": 11605
+    },
+    {
+      "epoch": 0.8058602971809471,
+      "grad_norm": 3.673354287698509,
+      "learning_rate": 9.5636962871687e-07,
+      "loss": 0.464,
+      "step": 11606
+    },
+    {
+      "epoch": 0.8059297319816692,
+      "grad_norm": 3.065061282223472,
+      "learning_rate": 9.557083238950204e-07,
+      "loss": 0.2461,
+      "step": 11607
+    },
+    {
+      "epoch": 0.8059991667823914,
+      "grad_norm": 3.8798516491045025,
+      "learning_rate": 9.550472236288676e-07,
+      "loss": 0.4881,
+      "step": 11608
+    },
+    {
+      "epoch": 0.8060686015831134,
+      "grad_norm": 3.6059036895179437,
+      "learning_rate": 9.543863279518477e-07,
+      "loss": 0.4019,
+      "step": 11609
+    },
+    {
+      "epoch": 0.8061380363838356,
+      "grad_norm": 4.561426022636995,
+      "learning_rate": 9.537256368973923e-07,
+      "loss": 0.5414,
+      "step": 11610
+    },
+    {
+      "epoch": 0.8062074711845577,
+      "grad_norm": 3.2153300802351703,
+      "learning_rate": 9.530651504989141e-07,
+      "loss": 0.3647,
+      "step": 11611
+    },
+    {
+      "epoch": 0.8062769059852798,
+      "grad_norm": 3.9877563144941486,
+      "learning_rate": 9.524048687898196e-07,
+      "loss": 0.5099,
+      "step": 11612
+    },
+    {
+      "epoch": 0.806346340786002,
+      "grad_norm": 3.938540017924112,
+      "learning_rate": 9.517447918035094e-07,
+      "loss": 0.3856,
+      "step": 11613
+    },
+    {
+      "epoch": 0.8064157755867241,
+      "grad_norm": 3.7479109513828486,
+      "learning_rate": 9.510849195733657e-07,
+      "loss": 0.4977,
+      "step": 11614
+    },
+    {
+      "epoch": 0.8064852103874461,
+      "grad_norm": 2.694959370332333,
+      "learning_rate": 9.504252521327634e-07,
+      "loss": 0.2551,
+      "step": 11615
+    },
+    {
+      "epoch": 0.8065546451881683,
+      "grad_norm": 4.302011083425289,
+      "learning_rate": 9.497657895150708e-07,
+      "loss": 0.5392,
+      "step": 11616
+    },
+    {
+      "epoch": 0.8066240799888904,
+      "grad_norm": 4.670664911306058,
+      "learning_rate": 9.491065317536413e-07,
+      "loss": 0.5469,
+      "step": 11617
+    },
+    {
+      "epoch": 0.8066935147896126,
+      "grad_norm": 7.801537519026751,
+      "learning_rate": 9.484474788818199e-07,
+      "loss": 0.376,
+      "step": 11618
+    },
+    {
+      "epoch": 0.8067629495903347,
+      "grad_norm": 4.555914483239768,
+      "learning_rate": 9.477886309329404e-07,
+      "loss": 0.3173,
+      "step": 11619
+    },
+    {
+      "epoch": 0.8068323843910568,
+      "grad_norm": 3.9229716964260084,
+      "learning_rate": 9.471299879403262e-07,
+      "loss": 0.2979,
+      "step": 11620
+    },
+    {
+      "epoch": 0.806901819191779,
+      "grad_norm": 3.507326065038648,
+      "learning_rate": 9.464715499372912e-07,
+      "loss": 0.3474,
+      "step": 11621
+    },
+    {
+      "epoch": 0.806971253992501,
+      "grad_norm": 3.3112362329966025,
+      "learning_rate": 9.458133169571382e-07,
+      "loss": 0.2483,
+      "step": 11622
+    },
+    {
+      "epoch": 0.8070406887932232,
+      "grad_norm": 2.6122711959537,
+      "learning_rate": 9.451552890331606e-07,
+      "loss": 0.2489,
+      "step": 11623
+    },
+    {
+      "epoch": 0.8071101235939453,
+      "grad_norm": 3.764484858243425,
+      "learning_rate": 9.444974661986394e-07,
+      "loss": 0.4488,
+      "step": 11624
+    },
+    {
+      "epoch": 0.8071795583946674,
+      "grad_norm": 4.1026634358773055,
+      "learning_rate": 9.438398484868477e-07,
+      "loss": 0.4714,
+      "step": 11625
+    },
+    {
+      "epoch": 0.8072489931953896,
+      "grad_norm": 3.6540160689500034,
+      "learning_rate": 9.431824359310465e-07,
+      "loss": 0.4947,
+      "step": 11626
+    },
+    {
+      "epoch": 0.8073184279961116,
+      "grad_norm": 4.289930305180008,
+      "learning_rate": 9.425252285644853e-07,
+      "loss": 0.428,
+      "step": 11627
+    },
+    {
+      "epoch": 0.8073878627968337,
+      "grad_norm": 4.535471518619672,
+      "learning_rate": 9.418682264204098e-07,
+      "loss": 0.4267,
+      "step": 11628
+    },
+    {
+      "epoch": 0.8074572975975559,
+      "grad_norm": 5.2702215849549185,
+      "learning_rate": 9.412114295320446e-07,
+      "loss": 0.3765,
+      "step": 11629
+    },
+    {
+      "epoch": 0.807526732398278,
+      "grad_norm": 3.4126044570094027,
+      "learning_rate": 9.40554837932614e-07,
+      "loss": 0.382,
+      "step": 11630
+    },
+    {
+      "epoch": 0.8075961671990002,
+      "grad_norm": 5.012263533042071,
+      "learning_rate": 9.398984516553272e-07,
+      "loss": 0.5361,
+      "step": 11631
+    },
+    {
+      "epoch": 0.8076656019997223,
+      "grad_norm": 4.153983153121576,
+      "learning_rate": 9.392422707333798e-07,
+      "loss": 0.4327,
+      "step": 11632
+    },
+    {
+      "epoch": 0.8077350368004443,
+      "grad_norm": 3.7945920474430346,
+      "learning_rate": 9.385862951999647e-07,
+      "loss": 0.4802,
+      "step": 11633
+    },
+    {
+      "epoch": 0.8078044716011665,
+      "grad_norm": 3.6051615407851236,
+      "learning_rate": 9.379305250882586e-07,
+      "loss": 0.4119,
+      "step": 11634
+    },
+    {
+      "epoch": 0.8078739064018886,
+      "grad_norm": 2.859589108470037,
+      "learning_rate": 9.372749604314307e-07,
+      "loss": 0.2675,
+      "step": 11635
+    },
+    {
+      "epoch": 0.8079433412026108,
+      "grad_norm": 4.204546198260028,
+      "learning_rate": 9.366196012626377e-07,
+      "loss": 0.4848,
+      "step": 11636
+    },
+    {
+      "epoch": 0.8080127760033329,
+      "grad_norm": 3.7144024992422415,
+      "learning_rate": 9.35964447615027e-07,
+      "loss": 0.2744,
+      "step": 11637
+    },
+    {
+      "epoch": 0.808082210804055,
+      "grad_norm": 4.732323960755227,
+      "learning_rate": 9.35309499521736e-07,
+      "loss": 0.3986,
+      "step": 11638
+    },
+    {
+      "epoch": 0.8081516456047771,
+      "grad_norm": 3.5317461726339543,
+      "learning_rate": 9.346547570158909e-07,
+      "loss": 0.2648,
+      "step": 11639
+    },
+    {
+      "epoch": 0.8082210804054992,
+      "grad_norm": 3.115813496848785,
+      "learning_rate": 9.340002201306081e-07,
+      "loss": 0.3997,
+      "step": 11640
+    },
+    {
+      "epoch": 0.8082905152062213,
+      "grad_norm": 3.783953906214396,
+      "learning_rate": 9.333458888989916e-07,
+      "loss": 0.3858,
+      "step": 11641
+    },
+    {
+      "epoch": 0.8083599500069435,
+      "grad_norm": 2.922295604064217,
+      "learning_rate": 9.32691763354141e-07,
+      "loss": 0.2345,
+      "step": 11642
+    },
+    {
+      "epoch": 0.8084293848076656,
+      "grad_norm": 4.876545781173021,
+      "learning_rate": 9.320378435291371e-07,
+      "loss": 0.4578,
+      "step": 11643
+    },
+    {
+      "epoch": 0.8084988196083878,
+      "grad_norm": 3.9619938530368732,
+      "learning_rate": 9.313841294570553e-07,
+      "loss": 0.2778,
+      "step": 11644
+    },
+    {
+      "epoch": 0.8085682544091098,
+      "grad_norm": 6.291322396612963,
+      "learning_rate": 9.307306211709621e-07,
+      "loss": 0.6204,
+      "step": 11645
+    },
+    {
+      "epoch": 0.8086376892098319,
+      "grad_norm": 3.5499873703287417,
+      "learning_rate": 9.300773187039086e-07,
+      "loss": 0.3361,
+      "step": 11646
+    },
+    {
+      "epoch": 0.8087071240105541,
+      "grad_norm": 3.2739319710833152,
+      "learning_rate": 9.294242220889371e-07,
+      "loss": 0.4185,
+      "step": 11647
+    },
+    {
+      "epoch": 0.8087765588112762,
+      "grad_norm": 3.9672692755125394,
+      "learning_rate": 9.28771331359084e-07,
+      "loss": 0.5162,
+      "step": 11648
+    },
+    {
+      "epoch": 0.8088459936119984,
+      "grad_norm": 4.149445842740863,
+      "learning_rate": 9.281186465473696e-07,
+      "loss": 0.361,
+      "step": 11649
+    },
+    {
+      "epoch": 0.8089154284127205,
+      "grad_norm": 3.9513397070141014,
+      "learning_rate": 9.274661676868068e-07,
+      "loss": 0.5013,
+      "step": 11650
+    },
+    {
+      "epoch": 0.8089848632134425,
+      "grad_norm": 3.12624910204514,
+      "learning_rate": 9.268138948103966e-07,
+      "loss": 0.2656,
+      "step": 11651
+    },
+    {
+      "epoch": 0.8090542980141647,
+      "grad_norm": 3.4371604692683912,
+      "learning_rate": 9.261618279511309e-07,
+      "loss": 0.2517,
+      "step": 11652
+    },
+    {
+      "epoch": 0.8091237328148868,
+      "grad_norm": 3.3888503661696228,
+      "learning_rate": 9.255099671419898e-07,
+      "loss": 0.318,
+      "step": 11653
+    },
+    {
+      "epoch": 0.8091931676156089,
+      "grad_norm": 4.31244501796927,
+      "learning_rate": 9.248583124159438e-07,
+      "loss": 0.4893,
+      "step": 11654
+    },
+    {
+      "epoch": 0.8092626024163311,
+      "grad_norm": 4.231233989460859,
+      "learning_rate": 9.242068638059531e-07,
+      "loss": 0.3424,
+      "step": 11655
+    },
+    {
+      "epoch": 0.8093320372170532,
+      "grad_norm": 4.126074615227126,
+      "learning_rate": 9.235556213449676e-07,
+      "loss": 0.3782,
+      "step": 11656
+    },
+    {
+      "epoch": 0.8094014720177753,
+      "grad_norm": 4.369641489300976,
+      "learning_rate": 9.229045850659252e-07,
+      "loss": 0.658,
+      "step": 11657
+    },
+    {
+      "epoch": 0.8094709068184974,
+      "grad_norm": 5.539542730973939,
+      "learning_rate": 9.222537550017546e-07,
+      "loss": 0.6715,
+      "step": 11658
+    },
+    {
+      "epoch": 0.8095403416192195,
+      "grad_norm": 3.4370301747427825,
+      "learning_rate": 9.216031311853757e-07,
+      "loss": 0.3314,
+      "step": 11659
+    },
+    {
+      "epoch": 0.8096097764199417,
+      "grad_norm": 3.822453272235385,
+      "learning_rate": 9.209527136496971e-07,
+      "loss": 0.4535,
+      "step": 11660
+    },
+    {
+      "epoch": 0.8096792112206638,
+      "grad_norm": 3.963960243724221,
+      "learning_rate": 9.203025024276119e-07,
+      "loss": 0.3644,
+      "step": 11661
+    },
+    {
+      "epoch": 0.809748646021386,
+      "grad_norm": 5.4443674334287415,
+      "learning_rate": 9.196524975520104e-07,
+      "loss": 0.4643,
+      "step": 11662
+    },
+    {
+      "epoch": 0.809818080822108,
+      "grad_norm": 5.6946014127009645,
+      "learning_rate": 9.190026990557704e-07,
+      "loss": 0.558,
+      "step": 11663
+    },
+    {
+      "epoch": 0.8098875156228301,
+      "grad_norm": 4.1444481753318145,
+      "learning_rate": 9.183531069717527e-07,
+      "loss": 0.4244,
+      "step": 11664
+    },
+    {
+      "epoch": 0.8099569504235523,
+      "grad_norm": 4.59506281635092,
+      "learning_rate": 9.177037213328177e-07,
+      "loss": 0.5666,
+      "step": 11665
+    },
+    {
+      "epoch": 0.8100263852242744,
+      "grad_norm": 4.57369915159194,
+      "learning_rate": 9.170545421718091e-07,
+      "loss": 0.5455,
+      "step": 11666
+    },
+    {
+      "epoch": 0.8100958200249966,
+      "grad_norm": 3.964826243467328,
+      "learning_rate": 9.164055695215618e-07,
+      "loss": 0.3261,
+      "step": 11667
+    },
+    {
+      "epoch": 0.8101652548257187,
+      "grad_norm": 4.232016453162892,
+      "learning_rate": 9.157568034148995e-07,
+      "loss": 0.5799,
+      "step": 11668
+    },
+    {
+      "epoch": 0.8102346896264407,
+      "grad_norm": 7.29902461723007,
+      "learning_rate": 9.151082438846365e-07,
+      "loss": 0.5576,
+      "step": 11669
+    },
+    {
+      "epoch": 0.8103041244271629,
+      "grad_norm": 4.1579209502392995,
+      "learning_rate": 9.14459890963576e-07,
+      "loss": 0.4852,
+      "step": 11670
+    },
+    {
+      "epoch": 0.810373559227885,
+      "grad_norm": 6.591272821066115,
+      "learning_rate": 9.138117446845113e-07,
+      "loss": 0.8385,
+      "step": 11671
+    },
+    {
+      "epoch": 0.8104429940286071,
+      "grad_norm": 4.4263810584227015,
+      "learning_rate": 9.131638050802249e-07,
+      "loss": 0.4848,
+      "step": 11672
+    },
+    {
+      "epoch": 0.8105124288293293,
+      "grad_norm": 3.7437909334607844,
+      "learning_rate": 9.125160721834875e-07,
+      "loss": 0.2791,
+      "step": 11673
+    },
+    {
+      "epoch": 0.8105818636300514,
+      "grad_norm": 3.3582595772695902,
+      "learning_rate": 9.118685460270638e-07,
+      "loss": 0.3894,
+      "step": 11674
+    },
+    {
+      "epoch": 0.8106512984307735,
+      "grad_norm": 2.548255292175957,
+      "learning_rate": 9.11221226643702e-07,
+      "loss": 0.2183,
+      "step": 11675
+    },
+    {
+      "epoch": 0.8107207332314956,
+      "grad_norm": 3.2391437950357416,
+      "learning_rate": 9.105741140661428e-07,
+      "loss": 0.3967,
+      "step": 11676
+    },
+    {
+      "epoch": 0.8107901680322177,
+      "grad_norm": 3.846310177296689,
+      "learning_rate": 9.099272083271188e-07,
+      "loss": 0.3998,
+      "step": 11677
+    },
+    {
+      "epoch": 0.8108596028329399,
+      "grad_norm": 4.330698741297082,
+      "learning_rate": 9.092805094593476e-07,
+      "loss": 0.5589,
+      "step": 11678
+    },
+    {
+      "epoch": 0.810929037633662,
+      "grad_norm": 3.5107798499723555,
+      "learning_rate": 9.086340174955399e-07,
+      "loss": 0.2514,
+      "step": 11679
+    },
+    {
+      "epoch": 0.8109984724343842,
+      "grad_norm": 5.354326672537009,
+      "learning_rate": 9.079877324683939e-07,
+      "loss": 0.655,
+      "step": 11680
+    },
+    {
+      "epoch": 0.8110679072351062,
+      "grad_norm": 2.853083942661186,
+      "learning_rate": 9.073416544105978e-07,
+      "loss": 0.1361,
+      "step": 11681
+    },
+    {
+      "epoch": 0.8111373420358283,
+      "grad_norm": 4.161348763136716,
+      "learning_rate": 9.066957833548295e-07,
+      "loss": 0.3647,
+      "step": 11682
+    },
+    {
+      "epoch": 0.8112067768365505,
+      "grad_norm": 5.792675936872284,
+      "learning_rate": 9.060501193337567e-07,
+      "loss": 0.4758,
+      "step": 11683
+    },
+    {
+      "epoch": 0.8112762116372726,
+      "grad_norm": 5.546001126073153,
+      "learning_rate": 9.054046623800367e-07,
+      "loss": 0.6815,
+      "step": 11684
+    },
+    {
+      "epoch": 0.8113456464379947,
+      "grad_norm": 3.0948799274444045,
+      "learning_rate": 9.047594125263149e-07,
+      "loss": 0.3737,
+      "step": 11685
+    },
+    {
+      "epoch": 0.8114150812387169,
+      "grad_norm": 4.243022266742495,
+      "learning_rate": 9.041143698052285e-07,
+      "loss": 0.5286,
+      "step": 11686
+    },
+    {
+      "epoch": 0.8114845160394389,
+      "grad_norm": 4.226221300703433,
+      "learning_rate": 9.034695342494004e-07,
+      "loss": 0.3947,
+      "step": 11687
+    },
+    {
+      "epoch": 0.8115539508401611,
+      "grad_norm": 3.742213983672117,
+      "learning_rate": 9.028249058914512e-07,
+      "loss": 0.4412,
+      "step": 11688
+    },
+    {
+      "epoch": 0.8116233856408832,
+      "grad_norm": 3.7812790224035866,
+      "learning_rate": 9.021804847639798e-07,
+      "loss": 0.3201,
+      "step": 11689
+    },
+    {
+      "epoch": 0.8116928204416053,
+      "grad_norm": 5.10757928568081,
+      "learning_rate": 9.015362708995817e-07,
+      "loss": 0.4761,
+      "step": 11690
+    },
+    {
+      "epoch": 0.8117622552423275,
+      "grad_norm": 4.657862089135266,
+      "learning_rate": 9.008922643308426e-07,
+      "loss": 0.5464,
+      "step": 11691
+    },
+    {
+      "epoch": 0.8118316900430496,
+      "grad_norm": 4.796201443415369,
+      "learning_rate": 9.002484650903359e-07,
+      "loss": 0.6323,
+      "step": 11692
+    },
+    {
+      "epoch": 0.8119011248437717,
+      "grad_norm": 5.767071331560206,
+      "learning_rate": 8.996048732106194e-07,
+      "loss": 0.7317,
+      "step": 11693
+    },
+    {
+      "epoch": 0.8119705596444938,
+      "grad_norm": 3.1347832920735503,
+      "learning_rate": 8.989614887242504e-07,
+      "loss": 0.2817,
+      "step": 11694
+    },
+    {
+      "epoch": 0.8120399944452159,
+      "grad_norm": 3.649657096754373,
+      "learning_rate": 8.983183116637695e-07,
+      "loss": 0.4242,
+      "step": 11695
+    },
+    {
+      "epoch": 0.8121094292459381,
+      "grad_norm": 2.702552672857004,
+      "learning_rate": 8.976753420617046e-07,
+      "loss": 0.2099,
+      "step": 11696
+    },
+    {
+      "epoch": 0.8121788640466602,
+      "grad_norm": 4.143593427395381,
+      "learning_rate": 8.970325799505803e-07,
+      "loss": 0.3733,
+      "step": 11697
+    },
+    {
+      "epoch": 0.8122482988473823,
+      "grad_norm": 3.6186425495596732,
+      "learning_rate": 8.963900253629048e-07,
+      "loss": 0.384,
+      "step": 11698
+    },
+    {
+      "epoch": 0.8123177336481044,
+      "grad_norm": 2.130662018489372,
+      "learning_rate": 8.957476783311786e-07,
+      "loss": 0.1195,
+      "step": 11699
+    },
+    {
+      "epoch": 0.8123871684488265,
+      "grad_norm": 4.156612297779111,
+      "learning_rate": 8.951055388878904e-07,
+      "loss": 0.4369,
+      "step": 11700
+    },
+    {
+      "epoch": 0.8124566032495487,
+      "grad_norm": 2.4985340899941497,
+      "learning_rate": 8.944636070655194e-07,
+      "loss": 0.1438,
+      "step": 11701
+    },
+    {
+      "epoch": 0.8125260380502708,
+      "grad_norm": 4.016665790870501,
+      "learning_rate": 8.938218828965317e-07,
+      "loss": 0.4047,
+      "step": 11702
+    },
+    {
+      "epoch": 0.8125954728509929,
+      "grad_norm": 3.481404079123336,
+      "learning_rate": 8.931803664133893e-07,
+      "loss": 0.3394,
+      "step": 11703
+    },
+    {
+      "epoch": 0.8126649076517151,
+      "grad_norm": 5.889106192001581,
+      "learning_rate": 8.92539057648536e-07,
+      "loss": 0.4699,
+      "step": 11704
+    },
+    {
+      "epoch": 0.8127343424524371,
+      "grad_norm": 4.599518317706325,
+      "learning_rate": 8.918979566344078e-07,
+      "loss": 0.5283,
+      "step": 11705
+    },
+    {
+      "epoch": 0.8128037772531593,
+      "grad_norm": 5.07934536581827,
+      "learning_rate": 8.912570634034345e-07,
+      "loss": 0.4626,
+      "step": 11706
+    },
+    {
+      "epoch": 0.8128732120538814,
+      "grad_norm": 3.9313910684253366,
+      "learning_rate": 8.906163779880272e-07,
+      "loss": 0.5278,
+      "step": 11707
+    },
+    {
+      "epoch": 0.8129426468546035,
+      "grad_norm": 5.135765536363077,
+      "learning_rate": 8.899759004205949e-07,
+      "loss": 0.3373,
+      "step": 11708
+    },
+    {
+      "epoch": 0.8130120816553257,
+      "grad_norm": 4.380650904258193,
+      "learning_rate": 8.893356307335305e-07,
+      "loss": 0.5902,
+      "step": 11709
+    },
+    {
+      "epoch": 0.8130815164560478,
+      "grad_norm": 3.7704484005734846,
+      "learning_rate": 8.886955689592186e-07,
+      "loss": 0.2092,
+      "step": 11710
+    },
+    {
+      "epoch": 0.8131509512567698,
+      "grad_norm": 4.794211265638571,
+      "learning_rate": 8.880557151300323e-07,
+      "loss": 0.5208,
+      "step": 11711
+    },
+    {
+      "epoch": 0.813220386057492,
+      "grad_norm": 5.463236989130208,
+      "learning_rate": 8.874160692783352e-07,
+      "loss": 0.5852,
+      "step": 11712
+    },
+    {
+      "epoch": 0.8132898208582141,
+      "grad_norm": 4.614266297920103,
+      "learning_rate": 8.86776631436479e-07,
+      "loss": 0.5403,
+      "step": 11713
+    },
+    {
+      "epoch": 0.8133592556589363,
+      "grad_norm": 3.2234902159625483,
+      "learning_rate": 8.861374016368068e-07,
+      "loss": 0.2709,
+      "step": 11714
+    },
+    {
+      "epoch": 0.8134286904596584,
+      "grad_norm": 3.450784819277149,
+      "learning_rate": 8.854983799116501e-07,
+      "loss": 0.2111,
+      "step": 11715
+    },
+    {
+      "epoch": 0.8134981252603805,
+      "grad_norm": 3.2234271434619175,
+      "learning_rate": 8.848595662933291e-07,
+      "loss": 0.3071,
+      "step": 11716
+    },
+    {
+      "epoch": 0.8135675600611026,
+      "grad_norm": 3.6747911633187385,
+      "learning_rate": 8.842209608141545e-07,
+      "loss": 0.3265,
+      "step": 11717
+    },
+    {
+      "epoch": 0.8136369948618247,
+      "grad_norm": 3.2690111484913045,
+      "learning_rate": 8.835825635064266e-07,
+      "loss": 0.3828,
+      "step": 11718
+    },
+    {
+      "epoch": 0.8137064296625469,
+      "grad_norm": 2.5871667795837987,
+      "learning_rate": 8.82944374402433e-07,
+      "loss": 0.2795,
+      "step": 11719
+    },
+    {
+      "epoch": 0.813775864463269,
+      "grad_norm": 4.412982197488445,
+      "learning_rate": 8.823063935344573e-07,
+      "loss": 0.3728,
+      "step": 11720
+    },
+    {
+      "epoch": 0.8138452992639911,
+      "grad_norm": 3.248760315976615,
+      "learning_rate": 8.816686209347631e-07,
+      "loss": 0.327,
+      "step": 11721
+    },
+    {
+      "epoch": 0.8139147340647133,
+      "grad_norm": 3.9131349035505814,
+      "learning_rate": 8.810310566356083e-07,
+      "loss": 0.4029,
+      "step": 11722
+    },
+    {
+      "epoch": 0.8139841688654353,
+      "grad_norm": 4.798613554271064,
+      "learning_rate": 8.803937006692432e-07,
+      "loss": 0.4879,
+      "step": 11723
+    },
+    {
+      "epoch": 0.8140536036661575,
+      "grad_norm": 3.216667739670734,
+      "learning_rate": 8.797565530679047e-07,
+      "loss": 0.306,
+      "step": 11724
+    },
+    {
+      "epoch": 0.8141230384668796,
+      "grad_norm": 3.717894063012203,
+      "learning_rate": 8.791196138638141e-07,
+      "loss": 0.3432,
+      "step": 11725
+    },
+    {
+      "epoch": 0.8141924732676017,
+      "grad_norm": 3.826254653617694,
+      "learning_rate": 8.784828830891922e-07,
+      "loss": 0.3952,
+      "step": 11726
+    },
+    {
+      "epoch": 0.8142619080683239,
+      "grad_norm": 4.591420065925006,
+      "learning_rate": 8.778463607762422e-07,
+      "loss": 0.5335,
+      "step": 11727
+    },
+    {
+      "epoch": 0.814331342869046,
+      "grad_norm": 3.617734547356876,
+      "learning_rate": 8.772100469571587e-07,
+      "loss": 0.4482,
+      "step": 11728
+    },
+    {
+      "epoch": 0.814400777669768,
+      "grad_norm": 3.6780400753749842,
+      "learning_rate": 8.765739416641255e-07,
+      "loss": 0.4905,
+      "step": 11729
+    },
+    {
+      "epoch": 0.8144702124704902,
+      "grad_norm": 4.395861752561041,
+      "learning_rate": 8.75938044929317e-07,
+      "loss": 0.4487,
+      "step": 11730
+    },
+    {
+      "epoch": 0.8145396472712123,
+      "grad_norm": 3.805288703143475,
+      "learning_rate": 8.75302356784895e-07,
+      "loss": 0.4543,
+      "step": 11731
+    },
+    {
+      "epoch": 0.8146090820719345,
+      "grad_norm": 4.329994251527006,
+      "learning_rate": 8.746668772630124e-07,
+      "loss": 0.233,
+      "step": 11732
+    },
+    {
+      "epoch": 0.8146785168726566,
+      "grad_norm": 3.2389701564617392,
+      "learning_rate": 8.74031606395811e-07,
+      "loss": 0.4294,
+      "step": 11733
+    },
+    {
+      "epoch": 0.8147479516733787,
+      "grad_norm": 4.588858961221016,
+      "learning_rate": 8.733965442154213e-07,
+      "loss": 0.5455,
+      "step": 11734
+    },
+    {
+      "epoch": 0.8148173864741008,
+      "grad_norm": 3.660455164788703,
+      "learning_rate": 8.727616907539666e-07,
+      "loss": 0.4666,
+      "step": 11735
+    },
+    {
+      "epoch": 0.8148868212748229,
+      "grad_norm": 3.804055281698499,
+      "learning_rate": 8.721270460435544e-07,
+      "loss": 0.427,
+      "step": 11736
+    },
+    {
+      "epoch": 0.8149562560755451,
+      "grad_norm": 2.9656015720603843,
+      "learning_rate": 8.714926101162841e-07,
+      "loss": 0.2853,
+      "step": 11737
+    },
+    {
+      "epoch": 0.8150256908762672,
+      "grad_norm": 4.790025399187547,
+      "learning_rate": 8.708583830042483e-07,
+      "loss": 0.5091,
+      "step": 11738
+    },
+    {
+      "epoch": 0.8150951256769893,
+      "grad_norm": 3.9322738743468792,
+      "learning_rate": 8.702243647395203e-07,
+      "loss": 0.4801,
+      "step": 11739
+    },
+    {
+      "epoch": 0.8151645604777115,
+      "grad_norm": 7.778572672042966,
+      "learning_rate": 8.695905553541723e-07,
+      "loss": 0.5444,
+      "step": 11740
+    },
+    {
+      "epoch": 0.8152339952784335,
+      "grad_norm": 3.4139733094545552,
+      "learning_rate": 8.689569548802601e-07,
+      "loss": 0.392,
+      "step": 11741
+    },
+    {
+      "epoch": 0.8153034300791556,
+      "grad_norm": 4.350806781618852,
+      "learning_rate": 8.683235633498305e-07,
+      "loss": 0.4755,
+      "step": 11742
+    },
+    {
+      "epoch": 0.8153728648798778,
+      "grad_norm": 3.7306487742360024,
+      "learning_rate": 8.676903807949194e-07,
+      "loss": 0.4495,
+      "step": 11743
+    },
+    {
+      "epoch": 0.8154422996805999,
+      "grad_norm": 4.9268292351051075,
+      "learning_rate": 8.670574072475535e-07,
+      "loss": 0.5075,
+      "step": 11744
+    },
+    {
+      "epoch": 0.8155117344813221,
+      "grad_norm": 3.558726384200234,
+      "learning_rate": 8.664246427397466e-07,
+      "loss": 0.3177,
+      "step": 11745
+    },
+    {
+      "epoch": 0.8155811692820442,
+      "grad_norm": 2.860706756865757,
+      "learning_rate": 8.657920873035042e-07,
+      "loss": 0.346,
+      "step": 11746
+    },
+    {
+      "epoch": 0.8156506040827662,
+      "grad_norm": 4.202468146138646,
+      "learning_rate": 8.651597409708196e-07,
+      "loss": 0.5484,
+      "step": 11747
+    },
+    {
+      "epoch": 0.8157200388834884,
+      "grad_norm": 3.633624389576488,
+      "learning_rate": 8.645276037736755e-07,
+      "loss": 0.3786,
+      "step": 11748
+    },
+    {
+      "epoch": 0.8157894736842105,
+      "grad_norm": 5.198307541591239,
+      "learning_rate": 8.638956757440475e-07,
+      "loss": 0.4677,
+      "step": 11749
+    },
+    {
+      "epoch": 0.8158589084849327,
+      "grad_norm": 4.035914901780297,
+      "learning_rate": 8.63263956913895e-07,
+      "loss": 0.4651,
+      "step": 11750
+    },
+    {
+      "epoch": 0.8159283432856548,
+      "grad_norm": 4.3773828251529325,
+      "learning_rate": 8.626324473151693e-07,
+      "loss": 0.4234,
+      "step": 11751
+    },
+    {
+      "epoch": 0.8159977780863769,
+      "grad_norm": 5.1169331903104975,
+      "learning_rate": 8.620011469798139e-07,
+      "loss": 0.6327,
+      "step": 11752
+    },
+    {
+      "epoch": 0.816067212887099,
+      "grad_norm": 4.117637788918677,
+      "learning_rate": 8.613700559397598e-07,
+      "loss": 0.2584,
+      "step": 11753
+    },
+    {
+      "epoch": 0.8161366476878211,
+      "grad_norm": 4.188071517494151,
+      "learning_rate": 8.607391742269217e-07,
+      "loss": 0.4541,
+      "step": 11754
+    },
+    {
+      "epoch": 0.8162060824885432,
+      "grad_norm": 4.28867547548891,
+      "learning_rate": 8.601085018732142e-07,
+      "loss": 0.4626,
+      "step": 11755
+    },
+    {
+      "epoch": 0.8162755172892654,
+      "grad_norm": 4.1678650775571136,
+      "learning_rate": 8.594780389105351e-07,
+      "loss": 0.4996,
+      "step": 11756
+    },
+    {
+      "epoch": 0.8163449520899875,
+      "grad_norm": 3.9255288671579684,
+      "learning_rate": 8.588477853707688e-07,
+      "loss": 0.417,
+      "step": 11757
+    },
+    {
+      "epoch": 0.8164143868907097,
+      "grad_norm": 5.625644907410497,
+      "learning_rate": 8.582177412857967e-07,
+      "loss": 0.6356,
+      "step": 11758
+    },
+    {
+      "epoch": 0.8164838216914317,
+      "grad_norm": 3.0959458209092765,
+      "learning_rate": 8.575879066874848e-07,
+      "loss": 0.2577,
+      "step": 11759
+    },
+    {
+      "epoch": 0.8165532564921538,
+      "grad_norm": 4.258004912583366,
+      "learning_rate": 8.569582816076888e-07,
+      "loss": 0.4924,
+      "step": 11760
+    },
+    {
+      "epoch": 0.816622691292876,
+      "grad_norm": 3.6276148097787058,
+      "learning_rate": 8.563288660782548e-07,
+      "loss": 0.4298,
+      "step": 11761
+    },
+    {
+      "epoch": 0.8166921260935981,
+      "grad_norm": 5.264889757442817,
+      "learning_rate": 8.556996601310181e-07,
+      "loss": 0.5962,
+      "step": 11762
+    },
+    {
+      "epoch": 0.8167615608943203,
+      "grad_norm": 3.85259224254862,
+      "learning_rate": 8.550706637978029e-07,
+      "loss": 0.3875,
+      "step": 11763
+    },
+    {
+      "epoch": 0.8168309956950424,
+      "grad_norm": 5.183780422405194,
+      "learning_rate": 8.544418771104229e-07,
+      "loss": 0.6088,
+      "step": 11764
+    },
+    {
+      "epoch": 0.8169004304957644,
+      "grad_norm": 4.940234839356221,
+      "learning_rate": 8.538133001006821e-07,
+      "loss": 0.6547,
+      "step": 11765
+    },
+    {
+      "epoch": 0.8169698652964866,
+      "grad_norm": 2.6284836402872855,
+      "learning_rate": 8.531849328003716e-07,
+      "loss": 0.238,
+      "step": 11766
+    },
+    {
+      "epoch": 0.8170393000972087,
+      "grad_norm": 3.8350718188674677,
+      "learning_rate": 8.525567752412772e-07,
+      "loss": 0.3925,
+      "step": 11767
+    },
+    {
+      "epoch": 0.8171087348979308,
+      "grad_norm": 3.565168271351903,
+      "learning_rate": 8.519288274551657e-07,
+      "loss": 0.2826,
+      "step": 11768
+    },
+    {
+      "epoch": 0.817178169698653,
+      "grad_norm": 3.7217926658146303,
+      "learning_rate": 8.51301089473801e-07,
+      "loss": 0.3608,
+      "step": 11769
+    },
+    {
+      "epoch": 0.8172476044993751,
+      "grad_norm": 4.377627527759458,
+      "learning_rate": 8.506735613289341e-07,
+      "loss": 0.6863,
+      "step": 11770
+    },
+    {
+      "epoch": 0.8173170393000972,
+      "grad_norm": 3.254936429335114,
+      "learning_rate": 8.500462430523009e-07,
+      "loss": 0.3478,
+      "step": 11771
+    },
+    {
+      "epoch": 0.8173864741008193,
+      "grad_norm": 3.242908566457988,
+      "learning_rate": 8.494191346756336e-07,
+      "loss": 0.2212,
+      "step": 11772
+    },
+    {
+      "epoch": 0.8174559089015414,
+      "grad_norm": 4.830566281632919,
+      "learning_rate": 8.4879223623065e-07,
+      "loss": 0.5504,
+      "step": 11773
+    },
+    {
+      "epoch": 0.8175253437022636,
+      "grad_norm": 2.825502310342076,
+      "learning_rate": 8.481655477490574e-07,
+      "loss": 0.1777,
+      "step": 11774
+    },
+    {
+      "epoch": 0.8175947785029857,
+      "grad_norm": 4.203805270489655,
+      "learning_rate": 8.47539069262554e-07,
+      "loss": 0.2708,
+      "step": 11775
+    },
+    {
+      "epoch": 0.8176642133037079,
+      "grad_norm": 4.14694991191907,
+      "learning_rate": 8.469128008028249e-07,
+      "loss": 0.3793,
+      "step": 11776
+    },
+    {
+      "epoch": 0.8177336481044299,
+      "grad_norm": 3.4125189605766435,
+      "learning_rate": 8.462867424015458e-07,
+      "loss": 0.347,
+      "step": 11777
+    },
+    {
+      "epoch": 0.817803082905152,
+      "grad_norm": 3.5849173350360677,
+      "learning_rate": 8.456608940903854e-07,
+      "loss": 0.4366,
+      "step": 11778
+    },
+    {
+      "epoch": 0.8178725177058742,
+      "grad_norm": 4.126955692901241,
+      "learning_rate": 8.45035255900995e-07,
+      "loss": 0.4323,
+      "step": 11779
+    },
+    {
+      "epoch": 0.8179419525065963,
+      "grad_norm": 3.7335233058599573,
+      "learning_rate": 8.444098278650187e-07,
+      "loss": 0.5796,
+      "step": 11780
+    },
+    {
+      "epoch": 0.8180113873073185,
+      "grad_norm": 3.641989588834993,
+      "learning_rate": 8.437846100140929e-07,
+      "loss": 0.4593,
+      "step": 11781
+    },
+    {
+      "epoch": 0.8180808221080406,
+      "grad_norm": 4.147041786640367,
+      "learning_rate": 8.431596023798372e-07,
+      "loss": 0.5802,
+      "step": 11782
+    },
+    {
+      "epoch": 0.8181502569087626,
+      "grad_norm": 4.547038681032045,
+      "learning_rate": 8.425348049938642e-07,
+      "loss": 0.5166,
+      "step": 11783
+    },
+    {
+      "epoch": 0.8182196917094848,
+      "grad_norm": 4.1194730750005935,
+      "learning_rate": 8.419102178877775e-07,
+      "loss": 0.6038,
+      "step": 11784
+    },
+    {
+      "epoch": 0.8182891265102069,
+      "grad_norm": 3.791710422696346,
+      "learning_rate": 8.412858410931679e-07,
+      "loss": 0.4025,
+      "step": 11785
+    },
+    {
+      "epoch": 0.818358561310929,
+      "grad_norm": 3.501015018509845,
+      "learning_rate": 8.406616746416118e-07,
+      "loss": 0.4012,
+      "step": 11786
+    },
+    {
+      "epoch": 0.8184279961116512,
+      "grad_norm": 3.721183783266032,
+      "learning_rate": 8.40037718564683e-07,
+      "loss": 0.4119,
+      "step": 11787
+    },
+    {
+      "epoch": 0.8184974309123733,
+      "grad_norm": 2.271695502413639,
+      "learning_rate": 8.394139728939393e-07,
+      "loss": 0.0965,
+      "step": 11788
+    },
+    {
+      "epoch": 0.8185668657130954,
+      "grad_norm": 5.629761681595073,
+      "learning_rate": 8.387904376609285e-07,
+      "loss": 0.8634,
+      "step": 11789
+    },
+    {
+      "epoch": 0.8186363005138175,
+      "grad_norm": 3.482439444902559,
+      "learning_rate": 8.381671128971885e-07,
+      "loss": 0.245,
+      "step": 11790
+    },
+    {
+      "epoch": 0.8187057353145396,
+      "grad_norm": 3.817969049406683,
+      "learning_rate": 8.375439986342471e-07,
+      "loss": 0.3042,
+      "step": 11791
+    },
+    {
+      "epoch": 0.8187751701152618,
+      "grad_norm": 3.3459168135874386,
+      "learning_rate": 8.369210949036194e-07,
+      "loss": 0.267,
+      "step": 11792
+    },
+    {
+      "epoch": 0.8188446049159839,
+      "grad_norm": 3.404588956661382,
+      "learning_rate": 8.362984017368125e-07,
+      "loss": 0.4823,
+      "step": 11793
+    },
+    {
+      "epoch": 0.8189140397167061,
+      "grad_norm": 3.369483287613501,
+      "learning_rate": 8.356759191653207e-07,
+      "loss": 0.1967,
+      "step": 11794
+    },
+    {
+      "epoch": 0.8189834745174281,
+      "grad_norm": 3.216671663186988,
+      "learning_rate": 8.350536472206283e-07,
+      "loss": 0.2273,
+      "step": 11795
+    },
+    {
+      "epoch": 0.8190529093181502,
+      "grad_norm": 3.7274559812521004,
+      "learning_rate": 8.3443158593421e-07,
+      "loss": 0.4235,
+      "step": 11796
+    },
+    {
+      "epoch": 0.8191223441188724,
+      "grad_norm": 3.77603843400379,
+      "learning_rate": 8.338097353375286e-07,
+      "loss": 0.6066,
+      "step": 11797
+    },
+    {
+      "epoch": 0.8191917789195945,
+      "grad_norm": 4.428119762696871,
+      "learning_rate": 8.331880954620342e-07,
+      "loss": 0.4267,
+      "step": 11798
+    },
+    {
+      "epoch": 0.8192612137203166,
+      "grad_norm": 3.794264823905194,
+      "learning_rate": 8.325666663391741e-07,
+      "loss": 0.486,
+      "step": 11799
+    },
+    {
+      "epoch": 0.8193306485210388,
+      "grad_norm": 3.963416776674324,
+      "learning_rate": 8.319454480003736e-07,
+      "loss": 0.596,
+      "step": 11800
+    },
+    {
+      "epoch": 0.8194000833217608,
+      "grad_norm": 3.076051387970303,
+      "learning_rate": 8.31324440477057e-07,
+      "loss": 0.2821,
+      "step": 11801
+    },
+    {
+      "epoch": 0.819469518122483,
+      "grad_norm": 4.003870600294956,
+      "learning_rate": 8.307036438006344e-07,
+      "loss": 0.3204,
+      "step": 11802
+    },
+    {
+      "epoch": 0.8195389529232051,
+      "grad_norm": 3.999558028238334,
+      "learning_rate": 8.30083058002501e-07,
+      "loss": 0.411,
+      "step": 11803
+    },
+    {
+      "epoch": 0.8196083877239272,
+      "grad_norm": 3.6979591738081194,
+      "learning_rate": 8.294626831140495e-07,
+      "loss": 0.3266,
+      "step": 11804
+    },
+    {
+      "epoch": 0.8196778225246494,
+      "grad_norm": 3.5547863631405243,
+      "learning_rate": 8.28842519166656e-07,
+      "loss": 0.3286,
+      "step": 11805
+    },
+    {
+      "epoch": 0.8197472573253715,
+      "grad_norm": 3.3767280639410266,
+      "learning_rate": 8.282225661916882e-07,
+      "loss": 0.3916,
+      "step": 11806
+    },
+    {
+      "epoch": 0.8198166921260936,
+      "grad_norm": 3.7518527532892056,
+      "learning_rate": 8.276028242205019e-07,
+      "loss": 0.0968,
+      "step": 11807
+    },
+    {
+      "epoch": 0.8198861269268157,
+      "grad_norm": 3.749327690954258,
+      "learning_rate": 8.269832932844434e-07,
+      "loss": 0.2939,
+      "step": 11808
+    },
+    {
+      "epoch": 0.8199555617275378,
+      "grad_norm": 3.8283094404721054,
+      "learning_rate": 8.263639734148466e-07,
+      "loss": 0.4071,
+      "step": 11809
+    },
+    {
+      "epoch": 0.82002499652826,
+      "grad_norm": 3.3381380630929325,
+      "learning_rate": 8.257448646430399e-07,
+      "loss": 0.4336,
+      "step": 11810
+    },
+    {
+      "epoch": 0.8200944313289821,
+      "grad_norm": 3.8973953681706615,
+      "learning_rate": 8.251259670003331e-07,
+      "loss": 0.4016,
+      "step": 11811
+    },
+    {
+      "epoch": 0.8201638661297042,
+      "grad_norm": 3.475250815912152,
+      "learning_rate": 8.245072805180293e-07,
+      "loss": 0.3295,
+      "step": 11812
+    },
+    {
+      "epoch": 0.8202333009304263,
+      "grad_norm": 3.326826779105959,
+      "learning_rate": 8.238888052274252e-07,
+      "loss": 0.3611,
+      "step": 11813
+    },
+    {
+      "epoch": 0.8203027357311484,
+      "grad_norm": 3.853830712016339,
+      "learning_rate": 8.232705411597979e-07,
+      "loss": 0.4108,
+      "step": 11814
+    },
+    {
+      "epoch": 0.8203721705318706,
+      "grad_norm": 4.435964056258887,
+      "learning_rate": 8.226524883464198e-07,
+      "loss": 0.3734,
+      "step": 11815
+    },
+    {
+      "epoch": 0.8204416053325927,
+      "grad_norm": 4.9857630993007795,
+      "learning_rate": 8.220346468185524e-07,
+      "loss": 0.5537,
+      "step": 11816
+    },
+    {
+      "epoch": 0.8205110401333148,
+      "grad_norm": 3.6638136795905876,
+      "learning_rate": 8.21417016607447e-07,
+      "loss": 0.2572,
+      "step": 11817
+    },
+    {
+      "epoch": 0.820580474934037,
+      "grad_norm": 4.189176414677505,
+      "learning_rate": 8.207995977443373e-07,
+      "loss": 0.5802,
+      "step": 11818
+    },
+    {
+      "epoch": 0.820649909734759,
+      "grad_norm": 3.7362507666752736,
+      "learning_rate": 8.201823902604567e-07,
+      "loss": 0.4085,
+      "step": 11819
+    },
+    {
+      "epoch": 0.8207193445354812,
+      "grad_norm": 3.5357866655912953,
+      "learning_rate": 8.195653941870202e-07,
+      "loss": 0.3259,
+      "step": 11820
+    },
+    {
+      "epoch": 0.8207887793362033,
+      "grad_norm": 3.5338976934981265,
+      "learning_rate": 8.189486095552363e-07,
+      "loss": 0.4002,
+      "step": 11821
+    },
+    {
+      "epoch": 0.8208582141369254,
+      "grad_norm": 4.2675884025539235,
+      "learning_rate": 8.183320363963005e-07,
+      "loss": 0.4096,
+      "step": 11822
+    },
+    {
+      "epoch": 0.8209276489376476,
+      "grad_norm": 4.883084461692116,
+      "learning_rate": 8.177156747413978e-07,
+      "loss": 0.5743,
+      "step": 11823
+    },
+    {
+      "epoch": 0.8209970837383697,
+      "grad_norm": 3.803487902807504,
+      "learning_rate": 8.170995246217044e-07,
+      "loss": 0.3872,
+      "step": 11824
+    },
+    {
+      "epoch": 0.8210665185390917,
+      "grad_norm": 3.3177195501312537,
+      "learning_rate": 8.164835860683834e-07,
+      "loss": 0.3996,
+      "step": 11825
+    },
+    {
+      "epoch": 0.8211359533398139,
+      "grad_norm": 3.9694433216938836,
+      "learning_rate": 8.158678591125884e-07,
+      "loss": 0.4356,
+      "step": 11826
+    },
+    {
+      "epoch": 0.821205388140536,
+      "grad_norm": 3.478551937667529,
+      "learning_rate": 8.152523437854632e-07,
+      "loss": 0.3031,
+      "step": 11827
+    },
+    {
+      "epoch": 0.8212748229412582,
+      "grad_norm": 3.6500682771072537,
+      "learning_rate": 8.146370401181385e-07,
+      "loss": 0.241,
+      "step": 11828
+    },
+    {
+      "epoch": 0.8213442577419803,
+      "grad_norm": 3.6980903290299,
+      "learning_rate": 8.140219481417349e-07,
+      "loss": 0.4854,
+      "step": 11829
+    },
+    {
+      "epoch": 0.8214136925427024,
+      "grad_norm": 3.4815588583037536,
+      "learning_rate": 8.134070678873663e-07,
+      "loss": 0.4046,
+      "step": 11830
+    },
+    {
+      "epoch": 0.8214831273434245,
+      "grad_norm": 2.6631870670184865,
+      "learning_rate": 8.127923993861314e-07,
+      "loss": 0.1252,
+      "step": 11831
+    },
+    {
+      "epoch": 0.8215525621441466,
+      "grad_norm": 2.0472134750238955,
+      "learning_rate": 8.121779426691168e-07,
+      "loss": 0.112,
+      "step": 11832
+    },
+    {
+      "epoch": 0.8216219969448688,
+      "grad_norm": 3.747711253133913,
+      "learning_rate": 8.115636977674035e-07,
+      "loss": 0.436,
+      "step": 11833
+    },
+    {
+      "epoch": 0.8216914317455909,
+      "grad_norm": 3.5338476745777614,
+      "learning_rate": 8.109496647120596e-07,
+      "loss": 0.4273,
+      "step": 11834
+    },
+    {
+      "epoch": 0.821760866546313,
+      "grad_norm": 3.4009277546233685,
+      "learning_rate": 8.103358435341408e-07,
+      "loss": 0.2471,
+      "step": 11835
+    },
+    {
+      "epoch": 0.8218303013470352,
+      "grad_norm": 4.278668523702549,
+      "learning_rate": 8.097222342646949e-07,
+      "loss": 0.4842,
+      "step": 11836
+    },
+    {
+      "epoch": 0.8218997361477572,
+      "grad_norm": 3.410975248926277,
+      "learning_rate": 8.091088369347561e-07,
+      "loss": 0.4192,
+      "step": 11837
+    },
+    {
+      "epoch": 0.8219691709484794,
+      "grad_norm": 3.898112778836242,
+      "learning_rate": 8.084956515753501e-07,
+      "loss": 0.4611,
+      "step": 11838
+    },
+    {
+      "epoch": 0.8220386057492015,
+      "grad_norm": 3.691503202432564,
+      "learning_rate": 8.078826782174909e-07,
+      "loss": 0.3731,
+      "step": 11839
+    },
+    {
+      "epoch": 0.8221080405499236,
+      "grad_norm": 4.094032440388885,
+      "learning_rate": 8.072699168921827e-07,
+      "loss": 0.4603,
+      "step": 11840
+    },
+    {
+      "epoch": 0.8221774753506458,
+      "grad_norm": 3.0401355856406806,
+      "learning_rate": 8.066573676304162e-07,
+      "loss": 0.2802,
+      "step": 11841
+    },
+    {
+      "epoch": 0.8222469101513679,
+      "grad_norm": 6.4513016656549045,
+      "learning_rate": 8.060450304631773e-07,
+      "loss": 0.6185,
+      "step": 11842
+    },
+    {
+      "epoch": 0.8223163449520899,
+      "grad_norm": 3.660261474391956,
+      "learning_rate": 8.054329054214338e-07,
+      "loss": 0.4485,
+      "step": 11843
+    },
+    {
+      "epoch": 0.8223857797528121,
+      "grad_norm": 4.74406900424621,
+      "learning_rate": 8.04820992536146e-07,
+      "loss": 0.5641,
+      "step": 11844
+    },
+    {
+      "epoch": 0.8224552145535342,
+      "grad_norm": 3.0524751519418114,
+      "learning_rate": 8.042092918382683e-07,
+      "loss": 0.1614,
+      "step": 11845
+    },
+    {
+      "epoch": 0.8225246493542564,
+      "grad_norm": 4.070790026598544,
+      "learning_rate": 8.03597803358735e-07,
+      "loss": 0.3232,
+      "step": 11846
+    },
+    {
+      "epoch": 0.8225940841549785,
+      "grad_norm": 4.991697955829069,
+      "learning_rate": 8.029865271284753e-07,
+      "loss": 0.5925,
+      "step": 11847
+    },
+    {
+      "epoch": 0.8226635189557006,
+      "grad_norm": 3.9421899127522084,
+      "learning_rate": 8.023754631784092e-07,
+      "loss": 0.5229,
+      "step": 11848
+    },
+    {
+      "epoch": 0.8227329537564227,
+      "grad_norm": 5.1946706657279496,
+      "learning_rate": 8.017646115394418e-07,
+      "loss": 0.4825,
+      "step": 11849
+    },
+    {
+      "epoch": 0.8228023885571448,
+      "grad_norm": 5.384412710101801,
+      "learning_rate": 8.011539722424699e-07,
+      "loss": 0.5415,
+      "step": 11850
+    },
+    {
+      "epoch": 0.822871823357867,
+      "grad_norm": 4.687429768196469,
+      "learning_rate": 8.00543545318379e-07,
+      "loss": 0.4706,
+      "step": 11851
+    },
+    {
+      "epoch": 0.8229412581585891,
+      "grad_norm": 5.785711712375405,
+      "learning_rate": 7.999333307980434e-07,
+      "loss": 0.3475,
+      "step": 11852
+    },
+    {
+      "epoch": 0.8230106929593112,
+      "grad_norm": 4.817299707876041,
+      "learning_rate": 7.99323328712327e-07,
+      "loss": 0.4534,
+      "step": 11853
+    },
+    {
+      "epoch": 0.8230801277600334,
+      "grad_norm": 4.5749468654289105,
+      "learning_rate": 7.987135390920836e-07,
+      "loss": 0.2885,
+      "step": 11854
+    },
+    {
+      "epoch": 0.8231495625607554,
+      "grad_norm": 4.621172065712622,
+      "learning_rate": 7.981039619681552e-07,
+      "loss": 0.5033,
+      "step": 11855
+    },
+    {
+      "epoch": 0.8232189973614775,
+      "grad_norm": 4.291780027732288,
+      "learning_rate": 7.974945973713732e-07,
+      "loss": 0.3877,
+      "step": 11856
+    },
+    {
+      "epoch": 0.8232884321621997,
+      "grad_norm": 3.152468469304622,
+      "learning_rate": 7.968854453325597e-07,
+      "loss": 0.269,
+      "step": 11857
+    },
+    {
+      "epoch": 0.8233578669629218,
+      "grad_norm": 3.2433076866166526,
+      "learning_rate": 7.962765058825228e-07,
+      "loss": 0.3356,
+      "step": 11858
+    },
+    {
+      "epoch": 0.823427301763644,
+      "grad_norm": 3.9719669097725627,
+      "learning_rate": 7.956677790520645e-07,
+      "loss": 0.3124,
+      "step": 11859
+    },
+    {
+      "epoch": 0.8234967365643661,
+      "grad_norm": 4.39531555296524,
+      "learning_rate": 7.950592648719735e-07,
+      "loss": 0.316,
+      "step": 11860
+    },
+    {
+      "epoch": 0.8235661713650881,
+      "grad_norm": 3.6406819993029456,
+      "learning_rate": 7.944509633730241e-07,
+      "loss": 0.4637,
+      "step": 11861
+    },
+    {
+      "epoch": 0.8236356061658103,
+      "grad_norm": 4.212858426748226,
+      "learning_rate": 7.938428745859878e-07,
+      "loss": 0.3658,
+      "step": 11862
+    },
+    {
+      "epoch": 0.8237050409665324,
+      "grad_norm": 3.1715331844289465,
+      "learning_rate": 7.932349985416205e-07,
+      "loss": 0.199,
+      "step": 11863
+    },
+    {
+      "epoch": 0.8237744757672546,
+      "grad_norm": 5.138640461995357,
+      "learning_rate": 7.926273352706638e-07,
+      "loss": 0.5809,
+      "step": 11864
+    },
+    {
+      "epoch": 0.8238439105679767,
+      "grad_norm": 3.901967466217089,
+      "learning_rate": 7.920198848038574e-07,
+      "loss": 0.2423,
+      "step": 11865
+    },
+    {
+      "epoch": 0.8239133453686988,
+      "grad_norm": 3.986085144850681,
+      "learning_rate": 7.91412647171923e-07,
+      "loss": 0.4077,
+      "step": 11866
+    },
+    {
+      "epoch": 0.823982780169421,
+      "grad_norm": 5.419288766397098,
+      "learning_rate": 7.908056224055749e-07,
+      "loss": 0.6957,
+      "step": 11867
+    },
+    {
+      "epoch": 0.824052214970143,
+      "grad_norm": 4.727425625420789,
+      "learning_rate": 7.901988105355152e-07,
+      "loss": 0.4867,
+      "step": 11868
+    },
+    {
+      "epoch": 0.8241216497708651,
+      "grad_norm": 4.762372471184213,
+      "learning_rate": 7.89592211592436e-07,
+      "loss": 0.6424,
+      "step": 11869
+    },
+    {
+      "epoch": 0.8241910845715873,
+      "grad_norm": 5.53892895260762,
+      "learning_rate": 7.889858256070182e-07,
+      "loss": 0.811,
+      "step": 11870
+    },
+    {
+      "epoch": 0.8242605193723094,
+      "grad_norm": 4.008489719786029,
+      "learning_rate": 7.883796526099325e-07,
+      "loss": 0.4265,
+      "step": 11871
+    },
+    {
+      "epoch": 0.8243299541730316,
+      "grad_norm": 4.463427046956273,
+      "learning_rate": 7.87773692631838e-07,
+      "loss": 0.6022,
+      "step": 11872
+    },
+    {
+      "epoch": 0.8243993889737536,
+      "grad_norm": 4.263002089042544,
+      "learning_rate": 7.871679457033827e-07,
+      "loss": 0.3317,
+      "step": 11873
+    },
+    {
+      "epoch": 0.8244688237744757,
+      "grad_norm": 4.7440664640268935,
+      "learning_rate": 7.865624118552073e-07,
+      "loss": 0.5285,
+      "step": 11874
+    },
+    {
+      "epoch": 0.8245382585751979,
+      "grad_norm": 3.437832291635814,
+      "learning_rate": 7.85957091117937e-07,
+      "loss": 0.3027,
+      "step": 11875
+    },
+    {
+      "epoch": 0.82460769337592,
+      "grad_norm": 3.6765544157904677,
+      "learning_rate": 7.853519835221867e-07,
+      "loss": 0.4509,
+      "step": 11876
+    },
+    {
+      "epoch": 0.8246771281766422,
+      "grad_norm": 3.2310627234699534,
+      "learning_rate": 7.847470890985665e-07,
+      "loss": 0.3506,
+      "step": 11877
+    },
+    {
+      "epoch": 0.8247465629773643,
+      "grad_norm": 4.3003553906755405,
+      "learning_rate": 7.841424078776677e-07,
+      "loss": 0.1406,
+      "step": 11878
+    },
+    {
+      "epoch": 0.8248159977780863,
+      "grad_norm": 2.620507495163355,
+      "learning_rate": 7.835379398900738e-07,
+      "loss": 0.2078,
+      "step": 11879
+    },
+    {
+      "epoch": 0.8248854325788085,
+      "grad_norm": 2.8452927584059515,
+      "learning_rate": 7.82933685166361e-07,
+      "loss": 0.2781,
+      "step": 11880
+    },
+    {
+      "epoch": 0.8249548673795306,
+      "grad_norm": 2.982041109309288,
+      "learning_rate": 7.823296437370903e-07,
+      "loss": 0.2129,
+      "step": 11881
+    },
+    {
+      "epoch": 0.8250243021802527,
+      "grad_norm": 3.3393221261893817,
+      "learning_rate": 7.817258156328139e-07,
+      "loss": 0.3882,
+      "step": 11882
+    },
+    {
+      "epoch": 0.8250937369809749,
+      "grad_norm": 4.081932172624764,
+      "learning_rate": 7.811222008840719e-07,
+      "loss": 0.5612,
+      "step": 11883
+    },
+    {
+      "epoch": 0.825163171781697,
+      "grad_norm": 2.35604466801223,
+      "learning_rate": 7.805187995213954e-07,
+      "loss": 0.1732,
+      "step": 11884
+    },
+    {
+      "epoch": 0.8252326065824191,
+      "grad_norm": 3.815559953798593,
+      "learning_rate": 7.799156115753037e-07,
+      "loss": 0.3573,
+      "step": 11885
+    },
+    {
+      "epoch": 0.8253020413831412,
+      "grad_norm": 3.5903812441811076,
+      "learning_rate": 7.793126370763043e-07,
+      "loss": 0.3931,
+      "step": 11886
+    },
+    {
+      "epoch": 0.8253714761838633,
+      "grad_norm": 4.065662319771238,
+      "learning_rate": 7.787098760548961e-07,
+      "loss": 0.5873,
+      "step": 11887
+    },
+    {
+      "epoch": 0.8254409109845855,
+      "grad_norm": 3.5040751204565828,
+      "learning_rate": 7.781073285415658e-07,
+      "loss": 0.3467,
+      "step": 11888
+    },
+    {
+      "epoch": 0.8255103457853076,
+      "grad_norm": 4.774674273623775,
+      "learning_rate": 7.77504994566789e-07,
+      "loss": 0.4328,
+      "step": 11889
+    },
+    {
+      "epoch": 0.8255797805860298,
+      "grad_norm": 4.489129003728178,
+      "learning_rate": 7.769028741610307e-07,
+      "loss": 0.3566,
+      "step": 11890
+    },
+    {
+      "epoch": 0.8256492153867518,
+      "grad_norm": 4.224640561041638,
+      "learning_rate": 7.763009673547472e-07,
+      "loss": 0.3797,
+      "step": 11891
+    },
+    {
+      "epoch": 0.8257186501874739,
+      "grad_norm": 4.121939762447984,
+      "learning_rate": 7.756992741783825e-07,
+      "loss": 0.3783,
+      "step": 11892
+    },
+    {
+      "epoch": 0.8257880849881961,
+      "grad_norm": 4.416317616891731,
+      "learning_rate": 7.750977946623661e-07,
+      "loss": 0.5136,
+      "step": 11893
+    },
+    {
+      "epoch": 0.8258575197889182,
+      "grad_norm": 4.1350225537671985,
+      "learning_rate": 7.744965288371237e-07,
+      "loss": 0.2733,
+      "step": 11894
+    },
+    {
+      "epoch": 0.8259269545896404,
+      "grad_norm": 3.601991375956312,
+      "learning_rate": 7.738954767330664e-07,
+      "loss": 0.3348,
+      "step": 11895
+    },
+    {
+      "epoch": 0.8259963893903625,
+      "grad_norm": 3.5039716349987025,
+      "learning_rate": 7.732946383805912e-07,
+      "loss": 0.3457,
+      "step": 11896
+    },
+    {
+      "epoch": 0.8260658241910845,
+      "grad_norm": 2.433782199734208,
+      "learning_rate": 7.72694013810092e-07,
+      "loss": 0.1587,
+      "step": 11897
+    },
+    {
+      "epoch": 0.8261352589918067,
+      "grad_norm": 3.8184836264571516,
+      "learning_rate": 7.720936030519461e-07,
+      "loss": 0.2757,
+      "step": 11898
+    },
+    {
+      "epoch": 0.8262046937925288,
+      "grad_norm": 5.5016917841091715,
+      "learning_rate": 7.714934061365209e-07,
+      "loss": 0.6712,
+      "step": 11899
+    },
+    {
+      "epoch": 0.8262741285932509,
+      "grad_norm": 3.9243131350400398,
+      "learning_rate": 7.70893423094175e-07,
+      "loss": 0.521,
+      "step": 11900
+    },
+    {
+      "epoch": 0.8263435633939731,
+      "grad_norm": 6.1892890218332175,
+      "learning_rate": 7.702936539552541e-07,
+      "loss": 0.3458,
+      "step": 11901
+    },
+    {
+      "epoch": 0.8264129981946952,
+      "grad_norm": 3.071936997800823,
+      "learning_rate": 7.696940987500934e-07,
+      "loss": 0.3008,
+      "step": 11902
+    },
+    {
+      "epoch": 0.8264824329954173,
+      "grad_norm": 3.2645292949173297,
+      "learning_rate": 7.690947575090191e-07,
+      "loss": 0.429,
+      "step": 11903
+    },
+    {
+      "epoch": 0.8265518677961394,
+      "grad_norm": 4.089806173810195,
+      "learning_rate": 7.684956302623436e-07,
+      "loss": 0.4765,
+      "step": 11904
+    },
+    {
+      "epoch": 0.8266213025968615,
+      "grad_norm": 3.732425270761966,
+      "learning_rate": 7.678967170403701e-07,
+      "loss": 0.2622,
+      "step": 11905
+    },
+    {
+      "epoch": 0.8266907373975837,
+      "grad_norm": 3.276817016099543,
+      "learning_rate": 7.67298017873394e-07,
+      "loss": 0.3337,
+      "step": 11906
+    },
+    {
+      "epoch": 0.8267601721983058,
+      "grad_norm": 2.916021849146458,
+      "learning_rate": 7.666995327916932e-07,
+      "loss": 0.2954,
+      "step": 11907
+    },
+    {
+      "epoch": 0.826829606999028,
+      "grad_norm": 3.3933070556538345,
+      "learning_rate": 7.661012618255387e-07,
+      "loss": 0.3108,
+      "step": 11908
+    },
+    {
+      "epoch": 0.82689904179975,
+      "grad_norm": 3.013613117365627,
+      "learning_rate": 7.655032050051925e-07,
+      "loss": 0.4297,
+      "step": 11909
+    },
+    {
+      "epoch": 0.8269684766004721,
+      "grad_norm": 5.195818249474201,
+      "learning_rate": 7.649053623609026e-07,
+      "loss": 0.4789,
+      "step": 11910
+    },
+    {
+      "epoch": 0.8270379114011943,
+      "grad_norm": 4.604618119993308,
+      "learning_rate": 7.643077339229071e-07,
+      "loss": 0.6189,
+      "step": 11911
+    },
+    {
+      "epoch": 0.8271073462019164,
+      "grad_norm": 3.6200793514598284,
+      "learning_rate": 7.637103197214335e-07,
+      "loss": 0.509,
+      "step": 11912
+    },
+    {
+      "epoch": 0.8271767810026385,
+      "grad_norm": 4.29780312299174,
+      "learning_rate": 7.631131197866981e-07,
+      "loss": 0.4304,
+      "step": 11913
+    },
+    {
+      "epoch": 0.8272462158033607,
+      "grad_norm": 3.524860323583806,
+      "learning_rate": 7.625161341489068e-07,
+      "loss": 0.2911,
+      "step": 11914
+    },
+    {
+      "epoch": 0.8273156506040827,
+      "grad_norm": 3.8311439978811106,
+      "learning_rate": 7.619193628382548e-07,
+      "loss": 0.3845,
+      "step": 11915
+    },
+    {
+      "epoch": 0.8273850854048049,
+      "grad_norm": 4.880759236739138,
+      "learning_rate": 7.613228058849253e-07,
+      "loss": 0.3677,
+      "step": 11916
+    },
+    {
+      "epoch": 0.827454520205527,
+      "grad_norm": 5.989081087003398,
+      "learning_rate": 7.60726463319092e-07,
+      "loss": 0.6684,
+      "step": 11917
+    },
+    {
+      "epoch": 0.8275239550062491,
+      "grad_norm": 3.768389690851867,
+      "learning_rate": 7.601303351709166e-07,
+      "loss": 0.3685,
+      "step": 11918
+    },
+    {
+      "epoch": 0.8275933898069713,
+      "grad_norm": 4.082518659235821,
+      "learning_rate": 7.595344214705508e-07,
+      "loss": 0.6686,
+      "step": 11919
+    },
+    {
+      "epoch": 0.8276628246076934,
+      "grad_norm": 3.777879173562721,
+      "learning_rate": 7.58938722248137e-07,
+      "loss": 0.3157,
+      "step": 11920
+    },
+    {
+      "epoch": 0.8277322594084155,
+      "grad_norm": 4.589371680838984,
+      "learning_rate": 7.583432375338029e-07,
+      "loss": 0.4287,
+      "step": 11921
+    },
+    {
+      "epoch": 0.8278016942091376,
+      "grad_norm": 3.0669351793944624,
+      "learning_rate": 7.577479673576671e-07,
+      "loss": 0.2335,
+      "step": 11922
+    },
+    {
+      "epoch": 0.8278711290098597,
+      "grad_norm": 4.693895679343266,
+      "learning_rate": 7.571529117498394e-07,
+      "loss": 0.4043,
+      "step": 11923
+    },
+    {
+      "epoch": 0.8279405638105819,
+      "grad_norm": 4.2592931833743695,
+      "learning_rate": 7.56558070740418e-07,
+      "loss": 0.2718,
+      "step": 11924
+    },
+    {
+      "epoch": 0.828009998611304,
+      "grad_norm": 4.561344249311811,
+      "learning_rate": 7.559634443594849e-07,
+      "loss": 0.4164,
+      "step": 11925
+    },
+    {
+      "epoch": 0.8280794334120261,
+      "grad_norm": 3.1449838994737314,
+      "learning_rate": 7.553690326371193e-07,
+      "loss": 0.4223,
+      "step": 11926
+    },
+    {
+      "epoch": 0.8281488682127482,
+      "grad_norm": 5.8271672426196,
+      "learning_rate": 7.547748356033863e-07,
+      "loss": 0.5208,
+      "step": 11927
+    },
+    {
+      "epoch": 0.8282183030134703,
+      "grad_norm": 4.809114409487456,
+      "learning_rate": 7.54180853288336e-07,
+      "loss": 0.6948,
+      "step": 11928
+    },
+    {
+      "epoch": 0.8282877378141925,
+      "grad_norm": 4.085525023129078,
+      "learning_rate": 7.535870857220151e-07,
+      "loss": 0.3885,
+      "step": 11929
+    },
+    {
+      "epoch": 0.8283571726149146,
+      "grad_norm": 3.9246528759566734,
+      "learning_rate": 7.52993532934454e-07,
+      "loss": 0.4707,
+      "step": 11930
+    },
+    {
+      "epoch": 0.8284266074156367,
+      "grad_norm": 4.401499298750683,
+      "learning_rate": 7.524001949556742e-07,
+      "loss": 0.3778,
+      "step": 11931
+    },
+    {
+      "epoch": 0.8284960422163589,
+      "grad_norm": 4.051871691495512,
+      "learning_rate": 7.518070718156861e-07,
+      "loss": 0.4648,
+      "step": 11932
+    },
+    {
+      "epoch": 0.8285654770170809,
+      "grad_norm": 3.095422349533252,
+      "learning_rate": 7.512141635444892e-07,
+      "loss": 0.2205,
+      "step": 11933
+    },
+    {
+      "epoch": 0.8286349118178031,
+      "grad_norm": 3.9478884565266537,
+      "learning_rate": 7.506214701720704e-07,
+      "loss": 0.3097,
+      "step": 11934
+    },
+    {
+      "epoch": 0.8287043466185252,
+      "grad_norm": 3.6783246836342762,
+      "learning_rate": 7.500289917284121e-07,
+      "loss": 0.3038,
+      "step": 11935
+    },
+    {
+      "epoch": 0.8287737814192473,
+      "grad_norm": 3.494896439113337,
+      "learning_rate": 7.494367282434772e-07,
+      "loss": 0.3933,
+      "step": 11936
+    },
+    {
+      "epoch": 0.8288432162199695,
+      "grad_norm": 2.8360790990135065,
+      "learning_rate": 7.488446797472209e-07,
+      "loss": 0.2689,
+      "step": 11937
+    },
+    {
+      "epoch": 0.8289126510206916,
+      "grad_norm": 4.349090313301468,
+      "learning_rate": 7.482528462695932e-07,
+      "loss": 0.3252,
+      "step": 11938
+    },
+    {
+      "epoch": 0.8289820858214136,
+      "grad_norm": 3.688598116107836,
+      "learning_rate": 7.476612278405227e-07,
+      "loss": 0.4917,
+      "step": 11939
+    },
+    {
+      "epoch": 0.8290515206221358,
+      "grad_norm": 3.257715343891594,
+      "learning_rate": 7.470698244899366e-07,
+      "loss": 0.3806,
+      "step": 11940
+    },
+    {
+      "epoch": 0.8291209554228579,
+      "grad_norm": 5.129776088178313,
+      "learning_rate": 7.464786362477461e-07,
+      "loss": 0.492,
+      "step": 11941
+    },
+    {
+      "epoch": 0.8291903902235801,
+      "grad_norm": 3.4836783095926602,
+      "learning_rate": 7.458876631438533e-07,
+      "loss": 0.3956,
+      "step": 11942
+    },
+    {
+      "epoch": 0.8292598250243022,
+      "grad_norm": 3.190906658018773,
+      "learning_rate": 7.452969052081483e-07,
+      "loss": 0.3648,
+      "step": 11943
+    },
+    {
+      "epoch": 0.8293292598250243,
+      "grad_norm": 5.803172367343852,
+      "learning_rate": 7.447063624705114e-07,
+      "loss": 0.4155,
+      "step": 11944
+    },
+    {
+      "epoch": 0.8293986946257464,
+      "grad_norm": 3.4388316524600624,
+      "learning_rate": 7.441160349608117e-07,
+      "loss": 0.2793,
+      "step": 11945
+    },
+    {
+      "epoch": 0.8294681294264685,
+      "grad_norm": 2.8148218300455605,
+      "learning_rate": 7.43525922708907e-07,
+      "loss": 0.1677,
+      "step": 11946
+    },
+    {
+      "epoch": 0.8295375642271907,
+      "grad_norm": 4.175678198955694,
+      "learning_rate": 7.429360257446449e-07,
+      "loss": 0.4921,
+      "step": 11947
+    },
+    {
+      "epoch": 0.8296069990279128,
+      "grad_norm": 4.272701937033411,
+      "learning_rate": 7.423463440978596e-07,
+      "loss": 0.4954,
+      "step": 11948
+    },
+    {
+      "epoch": 0.8296764338286349,
+      "grad_norm": 4.353684967633905,
+      "learning_rate": 7.417568777983814e-07,
+      "loss": 0.4126,
+      "step": 11949
+    },
+    {
+      "epoch": 0.8297458686293571,
+      "grad_norm": 5.548368276650892,
+      "learning_rate": 7.411676268760199e-07,
+      "loss": 0.4468,
+      "step": 11950
+    },
+    {
+      "epoch": 0.8298153034300791,
+      "grad_norm": 6.362592342629756,
+      "learning_rate": 7.405785913605801e-07,
+      "loss": 0.6409,
+      "step": 11951
+    },
+    {
+      "epoch": 0.8298847382308012,
+      "grad_norm": 4.321804119384618,
+      "learning_rate": 7.399897712818572e-07,
+      "loss": 0.3932,
+      "step": 11952
+    },
+    {
+      "epoch": 0.8299541730315234,
+      "grad_norm": 3.982081394784734,
+      "learning_rate": 7.394011666696304e-07,
+      "loss": 0.4919,
+      "step": 11953
+    },
+    {
+      "epoch": 0.8300236078322455,
+      "grad_norm": 4.108935734060896,
+      "learning_rate": 7.388127775536696e-07,
+      "loss": 0.4741,
+      "step": 11954
+    },
+    {
+      "epoch": 0.8300930426329677,
+      "grad_norm": 3.110887300167369,
+      "learning_rate": 7.382246039637386e-07,
+      "loss": 0.2945,
+      "step": 11955
+    },
+    {
+      "epoch": 0.8301624774336898,
+      "grad_norm": 4.128288538408571,
+      "learning_rate": 7.376366459295853e-07,
+      "loss": 0.3703,
+      "step": 11956
+    },
+    {
+      "epoch": 0.8302319122344118,
+      "grad_norm": 1.7492305890525575,
+      "learning_rate": 7.370489034809447e-07,
+      "loss": 0.1178,
+      "step": 11957
+    },
+    {
+      "epoch": 0.830301347035134,
+      "grad_norm": 2.9614505927711776,
+      "learning_rate": 7.364613766475481e-07,
+      "loss": 0.2383,
+      "step": 11958
+    },
+    {
+      "epoch": 0.8303707818358561,
+      "grad_norm": 3.2815376919879524,
+      "learning_rate": 7.358740654591107e-07,
+      "loss": 0.3329,
+      "step": 11959
+    },
+    {
+      "epoch": 0.8304402166365783,
+      "grad_norm": 4.762145916929719,
+      "learning_rate": 7.35286969945338e-07,
+      "loss": 0.6134,
+      "step": 11960
+    },
+    {
+      "epoch": 0.8305096514373004,
+      "grad_norm": 4.656724449283461,
+      "learning_rate": 7.347000901359242e-07,
+      "loss": 0.3983,
+      "step": 11961
+    },
+    {
+      "epoch": 0.8305790862380225,
+      "grad_norm": 3.87558184903679,
+      "learning_rate": 7.341134260605537e-07,
+      "loss": 0.4133,
+      "step": 11962
+    },
+    {
+      "epoch": 0.8306485210387446,
+      "grad_norm": 3.2858645494932803,
+      "learning_rate": 7.335269777488984e-07,
+      "loss": 0.3108,
+      "step": 11963
+    },
+    {
+      "epoch": 0.8307179558394667,
+      "grad_norm": 5.081743884747921,
+      "learning_rate": 7.329407452306208e-07,
+      "loss": 0.4964,
+      "step": 11964
+    },
+    {
+      "epoch": 0.8307873906401889,
+      "grad_norm": 4.398726677619599,
+      "learning_rate": 7.323547285353721e-07,
+      "loss": 0.3621,
+      "step": 11965
+    },
+    {
+      "epoch": 0.830856825440911,
+      "grad_norm": 3.428798148131502,
+      "learning_rate": 7.317689276927909e-07,
+      "loss": 0.2505,
+      "step": 11966
+    },
+    {
+      "epoch": 0.8309262602416331,
+      "grad_norm": 3.0020140630651446,
+      "learning_rate": 7.311833427325099e-07,
+      "loss": 0.3836,
+      "step": 11967
+    },
+    {
+      "epoch": 0.8309956950423553,
+      "grad_norm": 3.6621068917816593,
+      "learning_rate": 7.305979736841434e-07,
+      "loss": 0.3016,
+      "step": 11968
+    },
+    {
+      "epoch": 0.8310651298430773,
+      "grad_norm": 4.967144712143544,
+      "learning_rate": 7.300128205772994e-07,
+      "loss": 0.469,
+      "step": 11969
+    },
+    {
+      "epoch": 0.8311345646437994,
+      "grad_norm": 4.42088218802386,
+      "learning_rate": 7.294278834415774e-07,
+      "loss": 0.4362,
+      "step": 11970
+    },
+    {
+      "epoch": 0.8312039994445216,
+      "grad_norm": 3.3350235969385755,
+      "learning_rate": 7.288431623065584e-07,
+      "loss": 0.3347,
+      "step": 11971
+    },
+    {
+      "epoch": 0.8312734342452437,
+      "grad_norm": 2.8695791482879205,
+      "learning_rate": 7.282586572018196e-07,
+      "loss": 0.1674,
+      "step": 11972
+    },
+    {
+      "epoch": 0.8313428690459659,
+      "grad_norm": 3.12251724080186,
+      "learning_rate": 7.276743681569248e-07,
+      "loss": 0.182,
+      "step": 11973
+    },
+    {
+      "epoch": 0.831412303846688,
+      "grad_norm": 4.155450737602543,
+      "learning_rate": 7.270902952014259e-07,
+      "loss": 0.4364,
+      "step": 11974
+    },
+    {
+      "epoch": 0.83148173864741,
+      "grad_norm": 4.346994334907897,
+      "learning_rate": 7.265064383648645e-07,
+      "loss": 0.3412,
+      "step": 11975
+    },
+    {
+      "epoch": 0.8315511734481322,
+      "grad_norm": 3.8013519798443585,
+      "learning_rate": 7.259227976767713e-07,
+      "loss": 0.3263,
+      "step": 11976
+    },
+    {
+      "epoch": 0.8316206082488543,
+      "grad_norm": 3.3698851913103627,
+      "learning_rate": 7.253393731666675e-07,
+      "loss": 0.1348,
+      "step": 11977
+    },
+    {
+      "epoch": 0.8316900430495765,
+      "grad_norm": 4.250324475309351,
+      "learning_rate": 7.247561648640606e-07,
+      "loss": 0.2879,
+      "step": 11978
+    },
+    {
+      "epoch": 0.8317594778502986,
+      "grad_norm": 4.313575659488107,
+      "learning_rate": 7.241731727984491e-07,
+      "loss": 0.4974,
+      "step": 11979
+    },
+    {
+      "epoch": 0.8318289126510207,
+      "grad_norm": 4.134644287477897,
+      "learning_rate": 7.235903969993185e-07,
+      "loss": 0.6356,
+      "step": 11980
+    },
+    {
+      "epoch": 0.8318983474517428,
+      "grad_norm": 3.200239001761598,
+      "learning_rate": 7.230078374961497e-07,
+      "loss": 0.2647,
+      "step": 11981
+    },
+    {
+      "epoch": 0.8319677822524649,
+      "grad_norm": 4.4029324686175775,
+      "learning_rate": 7.224254943184028e-07,
+      "loss": 0.4543,
+      "step": 11982
+    },
+    {
+      "epoch": 0.832037217053187,
+      "grad_norm": 4.264739819551894,
+      "learning_rate": 7.218433674955333e-07,
+      "loss": 0.6348,
+      "step": 11983
+    },
+    {
+      "epoch": 0.8321066518539092,
+      "grad_norm": 4.976450552361448,
+      "learning_rate": 7.212614570569865e-07,
+      "loss": 0.5577,
+      "step": 11984
+    },
+    {
+      "epoch": 0.8321760866546313,
+      "grad_norm": 3.0568343875206105,
+      "learning_rate": 7.206797630321944e-07,
+      "loss": 0.2992,
+      "step": 11985
+    },
+    {
+      "epoch": 0.8322455214553535,
+      "grad_norm": 7.8053516655279225,
+      "learning_rate": 7.200982854505751e-07,
+      "loss": 0.372,
+      "step": 11986
+    },
+    {
+      "epoch": 0.8323149562560755,
+      "grad_norm": 3.5378539685153485,
+      "learning_rate": 7.195170243415428e-07,
+      "loss": 0.4187,
+      "step": 11987
+    },
+    {
+      "epoch": 0.8323843910567976,
+      "grad_norm": 4.8658630496222885,
+      "learning_rate": 7.189359797344969e-07,
+      "loss": 0.5902,
+      "step": 11988
+    },
+    {
+      "epoch": 0.8324538258575198,
+      "grad_norm": 4.757984693157151,
+      "learning_rate": 7.183551516588221e-07,
+      "loss": 0.4108,
+      "step": 11989
+    },
+    {
+      "epoch": 0.8325232606582419,
+      "grad_norm": 3.613457068767353,
+      "learning_rate": 7.177745401439007e-07,
+      "loss": 0.3119,
+      "step": 11990
+    },
+    {
+      "epoch": 0.8325926954589641,
+      "grad_norm": 8.20288177400798,
+      "learning_rate": 7.171941452190966e-07,
+      "loss": 0.5144,
+      "step": 11991
+    },
+    {
+      "epoch": 0.8326621302596862,
+      "grad_norm": 3.1490152901619535,
+      "learning_rate": 7.166139669137668e-07,
+      "loss": 0.2896,
+      "step": 11992
+    },
+    {
+      "epoch": 0.8327315650604082,
+      "grad_norm": 4.608907674267553,
+      "learning_rate": 7.16034005257255e-07,
+      "loss": 0.6668,
+      "step": 11993
+    },
+    {
+      "epoch": 0.8328009998611304,
+      "grad_norm": 4.015424949268355,
+      "learning_rate": 7.15454260278896e-07,
+      "loss": 0.424,
+      "step": 11994
+    },
+    {
+      "epoch": 0.8328704346618525,
+      "grad_norm": 3.9733086682153957,
+      "learning_rate": 7.148747320080124e-07,
+      "loss": 0.4513,
+      "step": 11995
+    },
+    {
+      "epoch": 0.8329398694625746,
+      "grad_norm": 3.991423289105037,
+      "learning_rate": 7.142954204739156e-07,
+      "loss": 0.3679,
+      "step": 11996
+    },
+    {
+      "epoch": 0.8330093042632968,
+      "grad_norm": 5.986676936072969,
+      "learning_rate": 7.137163257059071e-07,
+      "loss": 0.5887,
+      "step": 11997
+    },
+    {
+      "epoch": 0.8330787390640189,
+      "grad_norm": 3.00824041479492,
+      "learning_rate": 7.131374477332754e-07,
+      "loss": 0.2305,
+      "step": 11998
+    },
+    {
+      "epoch": 0.833148173864741,
+      "grad_norm": 4.05830346813593,
+      "learning_rate": 7.125587865853023e-07,
+      "loss": 0.5108,
+      "step": 11999
+    },
+    {
+      "epoch": 0.8332176086654631,
+      "grad_norm": 3.545792156264135,
+      "learning_rate": 7.119803422912525e-07,
+      "loss": 0.3328,
+      "step": 12000
+    },
+    {
+      "epoch": 0.8332870434661852,
+      "grad_norm": 4.541569151605179,
+      "learning_rate": 7.114021148803863e-07,
+      "loss": 0.598,
+      "step": 12001
+    },
+    {
+      "epoch": 0.8333564782669074,
+      "grad_norm": 3.513191341475214,
+      "learning_rate": 7.108241043819486e-07,
+      "loss": 0.3435,
+      "step": 12002
+    },
+    {
+      "epoch": 0.8334259130676295,
+      "grad_norm": 4.498911108838659,
+      "learning_rate": 7.102463108251722e-07,
+      "loss": 0.4271,
+      "step": 12003
+    },
+    {
+      "epoch": 0.8334953478683517,
+      "grad_norm": 14.088428310116448,
+      "learning_rate": 7.096687342392844e-07,
+      "loss": 0.3764,
+      "step": 12004
+    },
+    {
+      "epoch": 0.8335647826690737,
+      "grad_norm": 5.390795534364898,
+      "learning_rate": 7.090913746534972e-07,
+      "loss": 0.7113,
+      "step": 12005
+    },
+    {
+      "epoch": 0.8336342174697958,
+      "grad_norm": 5.646516611894107,
+      "learning_rate": 7.085142320970128e-07,
+      "loss": 0.521,
+      "step": 12006
+    },
+    {
+      "epoch": 0.833703652270518,
+      "grad_norm": 3.6743847949003214,
+      "learning_rate": 7.079373065990225e-07,
+      "loss": 0.2251,
+      "step": 12007
+    },
+    {
+      "epoch": 0.8337730870712401,
+      "grad_norm": 4.879345695098777,
+      "learning_rate": 7.07360598188706e-07,
+      "loss": 0.5546,
+      "step": 12008
+    },
+    {
+      "epoch": 0.8338425218719622,
+      "grad_norm": 4.000258554098824,
+      "learning_rate": 7.067841068952336e-07,
+      "loss": 0.2963,
+      "step": 12009
+    },
+    {
+      "epoch": 0.8339119566726844,
+      "grad_norm": 3.888764917184493,
+      "learning_rate": 7.062078327477623e-07,
+      "loss": 0.3177,
+      "step": 12010
+    },
+    {
+      "epoch": 0.8339813914734064,
+      "grad_norm": 4.372826770993004,
+      "learning_rate": 7.056317757754405e-07,
+      "loss": 0.5725,
+      "step": 12011
+    },
+    {
+      "epoch": 0.8340508262741286,
+      "grad_norm": 3.35593725724384,
+      "learning_rate": 7.050559360074027e-07,
+      "loss": 0.356,
+      "step": 12012
+    },
+    {
+      "epoch": 0.8341202610748507,
+      "grad_norm": 3.7674666013388283,
+      "learning_rate": 7.044803134727779e-07,
+      "loss": 0.5072,
+      "step": 12013
+    },
+    {
+      "epoch": 0.8341896958755728,
+      "grad_norm": 4.083482736926625,
+      "learning_rate": 7.039049082006771e-07,
+      "loss": 0.4023,
+      "step": 12014
+    },
+    {
+      "epoch": 0.834259130676295,
+      "grad_norm": 4.3551582958513055,
+      "learning_rate": 7.033297202202038e-07,
+      "loss": 0.494,
+      "step": 12015
+    },
+    {
+      "epoch": 0.8343285654770171,
+      "grad_norm": 3.4467801773633706,
+      "learning_rate": 7.02754749560452e-07,
+      "loss": 0.4117,
+      "step": 12016
+    },
+    {
+      "epoch": 0.8343980002777392,
+      "grad_norm": 3.472441700537884,
+      "learning_rate": 7.021799962505038e-07,
+      "loss": 0.1223,
+      "step": 12017
+    },
+    {
+      "epoch": 0.8344674350784613,
+      "grad_norm": 4.722562449055913,
+      "learning_rate": 7.016054603194255e-07,
+      "loss": 0.5183,
+      "step": 12018
+    },
+    {
+      "epoch": 0.8345368698791834,
+      "grad_norm": 3.5937583623762093,
+      "learning_rate": 7.0103114179628e-07,
+      "loss": 0.3315,
+      "step": 12019
+    },
+    {
+      "epoch": 0.8346063046799056,
+      "grad_norm": 3.728916454801873,
+      "learning_rate": 7.004570407101152e-07,
+      "loss": 0.3651,
+      "step": 12020
+    },
+    {
+      "epoch": 0.8346757394806277,
+      "grad_norm": 8.527737199382713,
+      "learning_rate": 6.998831570899678e-07,
+      "loss": 0.4799,
+      "step": 12021
+    },
+    {
+      "epoch": 0.8347451742813499,
+      "grad_norm": 3.3521806168014106,
+      "learning_rate": 6.993094909648646e-07,
+      "loss": 0.3113,
+      "step": 12022
+    },
+    {
+      "epoch": 0.834814609082072,
+      "grad_norm": 4.873815299303339,
+      "learning_rate": 6.987360423638206e-07,
+      "loss": 0.4844,
+      "step": 12023
+    },
+    {
+      "epoch": 0.834884043882794,
+      "grad_norm": 4.94581898302854,
+      "learning_rate": 6.981628113158401e-07,
+      "loss": 0.3704,
+      "step": 12024
+    },
+    {
+      "epoch": 0.8349534786835162,
+      "grad_norm": 3.83743323659239,
+      "learning_rate": 6.975897978499175e-07,
+      "loss": 0.4026,
+      "step": 12025
+    },
+    {
+      "epoch": 0.8350229134842383,
+      "grad_norm": 3.6895229356698414,
+      "learning_rate": 6.970170019950334e-07,
+      "loss": 0.4636,
+      "step": 12026
+    },
+    {
+      "epoch": 0.8350923482849604,
+      "grad_norm": 3.0738458732614626,
+      "learning_rate": 6.96444423780161e-07,
+      "loss": 0.2402,
+      "step": 12027
+    },
+    {
+      "epoch": 0.8351617830856826,
+      "grad_norm": 2.9012529978130512,
+      "learning_rate": 6.958720632342597e-07,
+      "loss": 0.1474,
+      "step": 12028
+    },
+    {
+      "epoch": 0.8352312178864046,
+      "grad_norm": 5.788937707862736,
+      "learning_rate": 6.952999203862776e-07,
+      "loss": 0.5261,
+      "step": 12029
+    },
+    {
+      "epoch": 0.8353006526871268,
+      "grad_norm": 4.800132340904821,
+      "learning_rate": 6.947279952651558e-07,
+      "loss": 0.7018,
+      "step": 12030
+    },
+    {
+      "epoch": 0.8353700874878489,
+      "grad_norm": 3.80114961470897,
+      "learning_rate": 6.941562878998209e-07,
+      "loss": 0.4123,
+      "step": 12031
+    },
+    {
+      "epoch": 0.835439522288571,
+      "grad_norm": 3.713213629955207,
+      "learning_rate": 6.935847983191868e-07,
+      "loss": 0.346,
+      "step": 12032
+    },
+    {
+      "epoch": 0.8355089570892932,
+      "grad_norm": 3.9080649893585906,
+      "learning_rate": 6.930135265521615e-07,
+      "loss": 0.4942,
+      "step": 12033
+    },
+    {
+      "epoch": 0.8355783918900153,
+      "grad_norm": 3.7789738502497014,
+      "learning_rate": 6.924424726276396e-07,
+      "loss": 0.5496,
+      "step": 12034
+    },
+    {
+      "epoch": 0.8356478266907374,
+      "grad_norm": 4.099219345610852,
+      "learning_rate": 6.91871636574501e-07,
+      "loss": 0.4458,
+      "step": 12035
+    },
+    {
+      "epoch": 0.8357172614914595,
+      "grad_norm": 3.7707243361809195,
+      "learning_rate": 6.913010184216206e-07,
+      "loss": 0.4939,
+      "step": 12036
+    },
+    {
+      "epoch": 0.8357866962921816,
+      "grad_norm": 4.047240036981536,
+      "learning_rate": 6.907306181978596e-07,
+      "loss": 0.518,
+      "step": 12037
+    },
+    {
+      "epoch": 0.8358561310929038,
+      "grad_norm": 4.3985812035276,
+      "learning_rate": 6.901604359320673e-07,
+      "loss": 0.5076,
+      "step": 12038
+    },
+    {
+      "epoch": 0.8359255658936259,
+      "grad_norm": 3.277824006902283,
+      "learning_rate": 6.895904716530838e-07,
+      "loss": 0.3197,
+      "step": 12039
+    },
+    {
+      "epoch": 0.835995000694348,
+      "grad_norm": 4.757673428785765,
+      "learning_rate": 6.890207253897363e-07,
+      "loss": 0.6074,
+      "step": 12040
+    },
+    {
+      "epoch": 0.8360644354950701,
+      "grad_norm": 4.280781521027781,
+      "learning_rate": 6.884511971708413e-07,
+      "loss": 0.3922,
+      "step": 12041
+    },
+    {
+      "epoch": 0.8361338702957922,
+      "grad_norm": 4.174368929298065,
+      "learning_rate": 6.87881887025208e-07,
+      "loss": 0.4153,
+      "step": 12042
+    },
+    {
+      "epoch": 0.8362033050965144,
+      "grad_norm": 3.9650001451393253,
+      "learning_rate": 6.873127949816283e-07,
+      "loss": 0.56,
+      "step": 12043
+    },
+    {
+      "epoch": 0.8362727398972365,
+      "grad_norm": 3.9241558193612325,
+      "learning_rate": 6.867439210688859e-07,
+      "loss": 0.4091,
+      "step": 12044
+    },
+    {
+      "epoch": 0.8363421746979586,
+      "grad_norm": 3.3787642787508023,
+      "learning_rate": 6.861752653157577e-07,
+      "loss": 0.177,
+      "step": 12045
+    },
+    {
+      "epoch": 0.8364116094986808,
+      "grad_norm": 3.5483844566645333,
+      "learning_rate": 6.85606827751002e-07,
+      "loss": 0.4131,
+      "step": 12046
+    },
+    {
+      "epoch": 0.8364810442994028,
+      "grad_norm": 3.9525324911297957,
+      "learning_rate": 6.8503860840337e-07,
+      "loss": 0.2742,
+      "step": 12047
+    },
+    {
+      "epoch": 0.836550479100125,
+      "grad_norm": 3.010945775150134,
+      "learning_rate": 6.844706073016033e-07,
+      "loss": 0.2514,
+      "step": 12048
+    },
+    {
+      "epoch": 0.8366199139008471,
+      "grad_norm": 4.072115198668518,
+      "learning_rate": 6.839028244744311e-07,
+      "loss": 0.419,
+      "step": 12049
+    },
+    {
+      "epoch": 0.8366893487015692,
+      "grad_norm": 4.5846470554537495,
+      "learning_rate": 6.833352599505677e-07,
+      "loss": 0.6139,
+      "step": 12050
+    },
+    {
+      "epoch": 0.8367587835022914,
+      "grad_norm": 5.346941741372405,
+      "learning_rate": 6.827679137587234e-07,
+      "loss": 0.5601,
+      "step": 12051
+    },
+    {
+      "epoch": 0.8368282183030135,
+      "grad_norm": 2.637682265777028,
+      "learning_rate": 6.822007859275931e-07,
+      "loss": 0.2626,
+      "step": 12052
+    },
+    {
+      "epoch": 0.8368976531037355,
+      "grad_norm": 4.690070071955338,
+      "learning_rate": 6.81633876485861e-07,
+      "loss": 0.3318,
+      "step": 12053
+    },
+    {
+      "epoch": 0.8369670879044577,
+      "grad_norm": 3.0284677842712475,
+      "learning_rate": 6.810671854622003e-07,
+      "loss": 0.2484,
+      "step": 12054
+    },
+    {
+      "epoch": 0.8370365227051798,
+      "grad_norm": 4.358762279547056,
+      "learning_rate": 6.805007128852747e-07,
+      "loss": 0.5125,
+      "step": 12055
+    },
+    {
+      "epoch": 0.837105957505902,
+      "grad_norm": 4.171724072101502,
+      "learning_rate": 6.799344587837353e-07,
+      "loss": 0.3824,
+      "step": 12056
+    },
+    {
+      "epoch": 0.8371753923066241,
+      "grad_norm": 3.791692917117011,
+      "learning_rate": 6.793684231862219e-07,
+      "loss": 0.356,
+      "step": 12057
+    },
+    {
+      "epoch": 0.8372448271073462,
+      "grad_norm": 5.611619927942887,
+      "learning_rate": 6.788026061213649e-07,
+      "loss": 0.4324,
+      "step": 12058
+    },
+    {
+      "epoch": 0.8373142619080683,
+      "grad_norm": 3.1735309680238206,
+      "learning_rate": 6.782370076177824e-07,
+      "loss": 0.382,
+      "step": 12059
+    },
+    {
+      "epoch": 0.8373836967087904,
+      "grad_norm": 3.2478330976721796,
+      "learning_rate": 6.776716277040818e-07,
+      "loss": 0.3389,
+      "step": 12060
+    },
+    {
+      "epoch": 0.8374531315095126,
+      "grad_norm": 4.812072231550191,
+      "learning_rate": 6.771064664088578e-07,
+      "loss": 0.3887,
+      "step": 12061
+    },
+    {
+      "epoch": 0.8375225663102347,
+      "grad_norm": 5.325633168809928,
+      "learning_rate": 6.765415237606981e-07,
+      "loss": 0.4463,
+      "step": 12062
+    },
+    {
+      "epoch": 0.8375920011109568,
+      "grad_norm": 3.9349652489470284,
+      "learning_rate": 6.759767997881772e-07,
+      "loss": 0.4983,
+      "step": 12063
+    },
+    {
+      "epoch": 0.837661435911679,
+      "grad_norm": 3.058103105730613,
+      "learning_rate": 6.754122945198543e-07,
+      "loss": 0.2411,
+      "step": 12064
+    },
+    {
+      "epoch": 0.837730870712401,
+      "grad_norm": 14.521711583019288,
+      "learning_rate": 6.748480079842851e-07,
+      "loss": 0.3483,
+      "step": 12065
+    },
+    {
+      "epoch": 0.8378003055131231,
+      "grad_norm": 3.2568466962367726,
+      "learning_rate": 6.742839402100094e-07,
+      "loss": 0.2923,
+      "step": 12066
+    },
+    {
+      "epoch": 0.8378697403138453,
+      "grad_norm": 3.2489512572250243,
+      "learning_rate": 6.737200912255576e-07,
+      "loss": 0.1745,
+      "step": 12067
+    },
+    {
+      "epoch": 0.8379391751145674,
+      "grad_norm": 3.739548472000005,
+      "learning_rate": 6.731564610594476e-07,
+      "loss": 0.3317,
+      "step": 12068
+    },
+    {
+      "epoch": 0.8380086099152896,
+      "grad_norm": 4.291314283431788,
+      "learning_rate": 6.725930497401878e-07,
+      "loss": 0.3978,
+      "step": 12069
+    },
+    {
+      "epoch": 0.8380780447160117,
+      "grad_norm": 5.287451332434072,
+      "learning_rate": 6.720298572962746e-07,
+      "loss": 0.5458,
+      "step": 12070
+    },
+    {
+      "epoch": 0.8381474795167337,
+      "grad_norm": 4.259716934834898,
+      "learning_rate": 6.714668837561938e-07,
+      "loss": 0.2897,
+      "step": 12071
+    },
+    {
+      "epoch": 0.8382169143174559,
+      "grad_norm": 2.350690319276187,
+      "learning_rate": 6.709041291484197e-07,
+      "loss": 0.2656,
+      "step": 12072
+    },
+    {
+      "epoch": 0.838286349118178,
+      "grad_norm": 5.432258551318331,
+      "learning_rate": 6.703415935014157e-07,
+      "loss": 0.4469,
+      "step": 12073
+    },
+    {
+      "epoch": 0.8383557839189002,
+      "grad_norm": 4.86093019079665,
+      "learning_rate": 6.697792768436362e-07,
+      "loss": 0.6218,
+      "step": 12074
+    },
+    {
+      "epoch": 0.8384252187196223,
+      "grad_norm": 4.1713014482094515,
+      "learning_rate": 6.692171792035196e-07,
+      "loss": 0.5244,
+      "step": 12075
+    },
+    {
+      "epoch": 0.8384946535203444,
+      "grad_norm": 4.289203952001535,
+      "learning_rate": 6.686553006094971e-07,
+      "loss": 0.4485,
+      "step": 12076
+    },
+    {
+      "epoch": 0.8385640883210665,
+      "grad_norm": 4.690420407153732,
+      "learning_rate": 6.680936410899896e-07,
+      "loss": 0.5174,
+      "step": 12077
+    },
+    {
+      "epoch": 0.8386335231217886,
+      "grad_norm": 5.159691522135387,
+      "learning_rate": 6.675322006734037e-07,
+      "loss": 0.3372,
+      "step": 12078
+    },
+    {
+      "epoch": 0.8387029579225108,
+      "grad_norm": 4.098626387962295,
+      "learning_rate": 6.669709793881346e-07,
+      "loss": 0.4957,
+      "step": 12079
+    },
+    {
+      "epoch": 0.8387723927232329,
+      "grad_norm": 3.835449078123772,
+      "learning_rate": 6.664099772625715e-07,
+      "loss": 0.5504,
+      "step": 12080
+    },
+    {
+      "epoch": 0.838841827523955,
+      "grad_norm": 4.308911008419336,
+      "learning_rate": 6.658491943250884e-07,
+      "loss": 0.3094,
+      "step": 12081
+    },
+    {
+      "epoch": 0.8389112623246772,
+      "grad_norm": 4.277234384550386,
+      "learning_rate": 6.652886306040485e-07,
+      "loss": 0.3482,
+      "step": 12082
+    },
+    {
+      "epoch": 0.8389806971253992,
+      "grad_norm": 4.8260125451735725,
+      "learning_rate": 6.647282861278042e-07,
+      "loss": 0.4134,
+      "step": 12083
+    },
+    {
+      "epoch": 0.8390501319261213,
+      "grad_norm": 6.014046019775549,
+      "learning_rate": 6.641681609246981e-07,
+      "loss": 0.6019,
+      "step": 12084
+    },
+    {
+      "epoch": 0.8391195667268435,
+      "grad_norm": 5.041483322642799,
+      "learning_rate": 6.636082550230594e-07,
+      "loss": 0.6887,
+      "step": 12085
+    },
+    {
+      "epoch": 0.8391890015275656,
+      "grad_norm": 4.0426056686033025,
+      "learning_rate": 6.630485684512084e-07,
+      "loss": 0.4724,
+      "step": 12086
+    },
+    {
+      "epoch": 0.8392584363282878,
+      "grad_norm": 4.038158959735304,
+      "learning_rate": 6.624891012374535e-07,
+      "loss": 0.4577,
+      "step": 12087
+    },
+    {
+      "epoch": 0.8393278711290099,
+      "grad_norm": 5.7976071143839425,
+      "learning_rate": 6.619298534100916e-07,
+      "loss": 0.4808,
+      "step": 12088
+    },
+    {
+      "epoch": 0.8393973059297319,
+      "grad_norm": 3.7330327426738696,
+      "learning_rate": 6.613708249974088e-07,
+      "loss": 0.2623,
+      "step": 12089
+    },
+    {
+      "epoch": 0.8394667407304541,
+      "grad_norm": 3.752746122838492,
+      "learning_rate": 6.608120160276788e-07,
+      "loss": 0.3843,
+      "step": 12090
+    },
+    {
+      "epoch": 0.8395361755311762,
+      "grad_norm": 2.9563235041244447,
+      "learning_rate": 6.602534265291683e-07,
+      "loss": 0.2192,
+      "step": 12091
+    },
+    {
+      "epoch": 0.8396056103318984,
+      "grad_norm": 4.546821378767146,
+      "learning_rate": 6.596950565301296e-07,
+      "loss": 0.5758,
+      "step": 12092
+    },
+    {
+      "epoch": 0.8396750451326205,
+      "grad_norm": 4.978151533355817,
+      "learning_rate": 6.591369060588016e-07,
+      "loss": 0.6097,
+      "step": 12093
+    },
+    {
+      "epoch": 0.8397444799333426,
+      "grad_norm": 3.635551933708019,
+      "learning_rate": 6.585789751434179e-07,
+      "loss": 0.3664,
+      "step": 12094
+    },
+    {
+      "epoch": 0.8398139147340647,
+      "grad_norm": 3.6984216758539907,
+      "learning_rate": 6.58021263812198e-07,
+      "loss": 0.5292,
+      "step": 12095
+    },
+    {
+      "epoch": 0.8398833495347868,
+      "grad_norm": 4.394357191090792,
+      "learning_rate": 6.574637720933474e-07,
+      "loss": 0.57,
+      "step": 12096
+    },
+    {
+      "epoch": 0.8399527843355089,
+      "grad_norm": 4.777789026229585,
+      "learning_rate": 6.569065000150659e-07,
+      "loss": 0.6232,
+      "step": 12097
+    },
+    {
+      "epoch": 0.8400222191362311,
+      "grad_norm": 3.310080267310657,
+      "learning_rate": 6.563494476055399e-07,
+      "loss": 0.2869,
+      "step": 12098
+    },
+    {
+      "epoch": 0.8400916539369532,
+      "grad_norm": 7.084461522475243,
+      "learning_rate": 6.557926148929428e-07,
+      "loss": 0.3753,
+      "step": 12099
+    },
+    {
+      "epoch": 0.8401610887376754,
+      "grad_norm": 5.432044312904549,
+      "learning_rate": 6.552360019054399e-07,
+      "loss": 0.5885,
+      "step": 12100
+    },
+    {
+      "epoch": 0.8402305235383974,
+      "grad_norm": 3.045909829754521,
+      "learning_rate": 6.546796086711832e-07,
+      "loss": 0.1963,
+      "step": 12101
+    },
+    {
+      "epoch": 0.8402999583391195,
+      "grad_norm": 4.602216628706258,
+      "learning_rate": 6.541234352183151e-07,
+      "loss": 0.6552,
+      "step": 12102
+    },
+    {
+      "epoch": 0.8403693931398417,
+      "grad_norm": 3.9567074885591826,
+      "learning_rate": 6.535674815749659e-07,
+      "loss": 0.3036,
+      "step": 12103
+    },
+    {
+      "epoch": 0.8404388279405638,
+      "grad_norm": 3.804471313268846,
+      "learning_rate": 6.530117477692555e-07,
+      "loss": 0.3998,
+      "step": 12104
+    },
+    {
+      "epoch": 0.840508262741286,
+      "grad_norm": 3.338994418270334,
+      "learning_rate": 6.524562338292906e-07,
+      "loss": 0.3007,
+      "step": 12105
+    },
+    {
+      "epoch": 0.8405776975420081,
+      "grad_norm": 2.8705948591495902,
+      "learning_rate": 6.519009397831716e-07,
+      "loss": 0.0975,
+      "step": 12106
+    },
+    {
+      "epoch": 0.8406471323427301,
+      "grad_norm": 3.9358282297905403,
+      "learning_rate": 6.513458656589822e-07,
+      "loss": 0.4511,
+      "step": 12107
+    },
+    {
+      "epoch": 0.8407165671434523,
+      "grad_norm": 3.9324186438534334,
+      "learning_rate": 6.507910114847971e-07,
+      "loss": 0.4765,
+      "step": 12108
+    },
+    {
+      "epoch": 0.8407860019441744,
+      "grad_norm": 5.115345520558615,
+      "learning_rate": 6.502363772886833e-07,
+      "loss": 0.5629,
+      "step": 12109
+    },
+    {
+      "epoch": 0.8408554367448965,
+      "grad_norm": 2.408524077313909,
+      "learning_rate": 6.496819630986889e-07,
+      "loss": 0.215,
+      "step": 12110
+    },
+    {
+      "epoch": 0.8409248715456187,
+      "grad_norm": 3.685838082085246,
+      "learning_rate": 6.491277689428594e-07,
+      "loss": 0.3062,
+      "step": 12111
+    },
+    {
+      "epoch": 0.8409943063463408,
+      "grad_norm": 5.2676979876396945,
+      "learning_rate": 6.485737948492237e-07,
+      "loss": 0.5152,
+      "step": 12112
+    },
+    {
+      "epoch": 0.841063741147063,
+      "grad_norm": 2.6822704984663566,
+      "learning_rate": 6.48020040845802e-07,
+      "loss": 0.1888,
+      "step": 12113
+    },
+    {
+      "epoch": 0.841133175947785,
+      "grad_norm": 4.240257167608544,
+      "learning_rate": 6.474665069606017e-07,
+      "loss": 0.3633,
+      "step": 12114
+    },
+    {
+      "epoch": 0.8412026107485071,
+      "grad_norm": 4.451522345957834,
+      "learning_rate": 6.469131932216205e-07,
+      "loss": 0.381,
+      "step": 12115
+    },
+    {
+      "epoch": 0.8412720455492293,
+      "grad_norm": 4.7147672489306895,
+      "learning_rate": 6.463600996568443e-07,
+      "loss": 0.5011,
+      "step": 12116
+    },
+    {
+      "epoch": 0.8413414803499514,
+      "grad_norm": 3.009393104262819,
+      "learning_rate": 6.458072262942472e-07,
+      "loss": 0.198,
+      "step": 12117
+    },
+    {
+      "epoch": 0.8414109151506736,
+      "grad_norm": 4.4908408696271,
+      "learning_rate": 6.452545731617937e-07,
+      "loss": 0.6326,
+      "step": 12118
+    },
+    {
+      "epoch": 0.8414803499513956,
+      "grad_norm": 3.4053293878096182,
+      "learning_rate": 6.447021402874354e-07,
+      "loss": 0.2542,
+      "step": 12119
+    },
+    {
+      "epoch": 0.8415497847521177,
+      "grad_norm": 3.1337166419601177,
+      "learning_rate": 6.441499276991165e-07,
+      "loss": 0.4481,
+      "step": 12120
+    },
+    {
+      "epoch": 0.8416192195528399,
+      "grad_norm": 3.4176291546192443,
+      "learning_rate": 6.43597935424764e-07,
+      "loss": 0.3149,
+      "step": 12121
+    },
+    {
+      "epoch": 0.841688654353562,
+      "grad_norm": 3.4015739928887854,
+      "learning_rate": 6.430461634922974e-07,
+      "loss": 0.2852,
+      "step": 12122
+    },
+    {
+      "epoch": 0.8417580891542841,
+      "grad_norm": 1.684408820382857,
+      "learning_rate": 6.424946119296266e-07,
+      "loss": 0.133,
+      "step": 12123
+    },
+    {
+      "epoch": 0.8418275239550063,
+      "grad_norm": 4.188034154445266,
+      "learning_rate": 6.419432807646486e-07,
+      "loss": 0.5555,
+      "step": 12124
+    },
+    {
+      "epoch": 0.8418969587557283,
+      "grad_norm": 3.8707605178227267,
+      "learning_rate": 6.413921700252457e-07,
+      "loss": 0.3914,
+      "step": 12125
+    },
+    {
+      "epoch": 0.8419663935564505,
+      "grad_norm": 3.7306760827278183,
+      "learning_rate": 6.40841279739296e-07,
+      "loss": 0.3757,
+      "step": 12126
+    },
+    {
+      "epoch": 0.8420358283571726,
+      "grad_norm": 4.445559439800578,
+      "learning_rate": 6.402906099346628e-07,
+      "loss": 0.5274,
+      "step": 12127
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 3.1581581874723645,
+      "learning_rate": 6.397401606391951e-07,
+      "loss": 0.2841,
+      "step": 12128
+    },
+    {
+      "epoch": 0.8421746979586169,
+      "grad_norm": 3.3018705008585476,
+      "learning_rate": 6.391899318807371e-07,
+      "loss": 0.4344,
+      "step": 12129
+    },
+    {
+      "epoch": 0.842244132759339,
+      "grad_norm": 10.301867014811508,
+      "learning_rate": 6.386399236871177e-07,
+      "loss": 0.6764,
+      "step": 12130
+    },
+    {
+      "epoch": 0.8423135675600611,
+      "grad_norm": 3.310240825426331,
+      "learning_rate": 6.380901360861557e-07,
+      "loss": 0.31,
+      "step": 12131
+    },
+    {
+      "epoch": 0.8423830023607832,
+      "grad_norm": 3.7061237611320568,
+      "learning_rate": 6.375405691056585e-07,
+      "loss": 0.3972,
+      "step": 12132
+    },
+    {
+      "epoch": 0.8424524371615053,
+      "grad_norm": 4.834204590850926,
+      "learning_rate": 6.369912227734232e-07,
+      "loss": 0.6174,
+      "step": 12133
+    },
+    {
+      "epoch": 0.8425218719622275,
+      "grad_norm": 3.4793328933191403,
+      "learning_rate": 6.364420971172347e-07,
+      "loss": 0.3118,
+      "step": 12134
+    },
+    {
+      "epoch": 0.8425913067629496,
+      "grad_norm": 3.695846156112402,
+      "learning_rate": 6.358931921648675e-07,
+      "loss": 0.3756,
+      "step": 12135
+    },
+    {
+      "epoch": 0.8426607415636718,
+      "grad_norm": 4.584199756503634,
+      "learning_rate": 6.353445079440839e-07,
+      "loss": 0.4908,
+      "step": 12136
+    },
+    {
+      "epoch": 0.8427301763643938,
+      "grad_norm": 4.018206814414804,
+      "learning_rate": 6.347960444826357e-07,
+      "loss": 0.3748,
+      "step": 12137
+    },
+    {
+      "epoch": 0.8427996111651159,
+      "grad_norm": 3.8202765408581594,
+      "learning_rate": 6.342478018082654e-07,
+      "loss": 0.3246,
+      "step": 12138
+    },
+    {
+      "epoch": 0.8428690459658381,
+      "grad_norm": 4.150335146622992,
+      "learning_rate": 6.336997799487004e-07,
+      "loss": 0.3827,
+      "step": 12139
+    },
+    {
+      "epoch": 0.8429384807665602,
+      "grad_norm": 3.3344818365189304,
+      "learning_rate": 6.331519789316587e-07,
+      "loss": 0.3205,
+      "step": 12140
+    },
+    {
+      "epoch": 0.8430079155672823,
+      "grad_norm": 4.195926009165619,
+      "learning_rate": 6.326043987848496e-07,
+      "loss": 0.6112,
+      "step": 12141
+    },
+    {
+      "epoch": 0.8430773503680045,
+      "grad_norm": 3.199569917804556,
+      "learning_rate": 6.320570395359677e-07,
+      "loss": 0.4363,
+      "step": 12142
+    },
+    {
+      "epoch": 0.8431467851687265,
+      "grad_norm": 1.6302477172729717,
+      "learning_rate": 6.315099012126979e-07,
+      "loss": 0.1146,
+      "step": 12143
+    },
+    {
+      "epoch": 0.8432162199694487,
+      "grad_norm": 4.333952192288873,
+      "learning_rate": 6.309629838427145e-07,
+      "loss": 0.3937,
+      "step": 12144
+    },
+    {
+      "epoch": 0.8432856547701708,
+      "grad_norm": 5.078167469995292,
+      "learning_rate": 6.304162874536796e-07,
+      "loss": 0.5036,
+      "step": 12145
+    },
+    {
+      "epoch": 0.8433550895708929,
+      "grad_norm": 3.838706986443937,
+      "learning_rate": 6.298698120732439e-07,
+      "loss": 0.4605,
+      "step": 12146
+    },
+    {
+      "epoch": 0.8434245243716151,
+      "grad_norm": 4.1165877646389495,
+      "learning_rate": 6.29323557729048e-07,
+      "loss": 0.4262,
+      "step": 12147
+    },
+    {
+      "epoch": 0.8434939591723372,
+      "grad_norm": 4.440287480076317,
+      "learning_rate": 6.28777524448721e-07,
+      "loss": 0.403,
+      "step": 12148
+    },
+    {
+      "epoch": 0.8435633939730594,
+      "grad_norm": 4.840428058227124,
+      "learning_rate": 6.282317122598802e-07,
+      "loss": 0.3905,
+      "step": 12149
+    },
+    {
+      "epoch": 0.8436328287737814,
+      "grad_norm": 4.525489943046081,
+      "learning_rate": 6.276861211901325e-07,
+      "loss": 0.4227,
+      "step": 12150
+    },
+    {
+      "epoch": 0.8437022635745035,
+      "grad_norm": 3.056840340009979,
+      "learning_rate": 6.271407512670724e-07,
+      "loss": 0.2631,
+      "step": 12151
+    },
+    {
+      "epoch": 0.8437716983752257,
+      "grad_norm": 4.112520535246105,
+      "learning_rate": 6.265956025182867e-07,
+      "loss": 0.469,
+      "step": 12152
+    },
+    {
+      "epoch": 0.8438411331759478,
+      "grad_norm": 3.348017375423266,
+      "learning_rate": 6.260506749713458e-07,
+      "loss": 0.355,
+      "step": 12153
+    },
+    {
+      "epoch": 0.8439105679766699,
+      "grad_norm": 4.5898083309624935,
+      "learning_rate": 6.255059686538107e-07,
+      "loss": 0.5462,
+      "step": 12154
+    },
+    {
+      "epoch": 0.843980002777392,
+      "grad_norm": 4.084106845146247,
+      "learning_rate": 6.24961483593235e-07,
+      "loss": 0.4341,
+      "step": 12155
+    },
+    {
+      "epoch": 0.8440494375781141,
+      "grad_norm": 2.7116054805180663,
+      "learning_rate": 6.244172198171583e-07,
+      "loss": 0.1943,
+      "step": 12156
+    },
+    {
+      "epoch": 0.8441188723788363,
+      "grad_norm": 4.046027273392892,
+      "learning_rate": 6.238731773531043e-07,
+      "loss": 0.3558,
+      "step": 12157
+    },
+    {
+      "epoch": 0.8441883071795584,
+      "grad_norm": 4.456152148524055,
+      "learning_rate": 6.233293562285947e-07,
+      "loss": 0.6335,
+      "step": 12158
+    },
+    {
+      "epoch": 0.8442577419802805,
+      "grad_norm": 4.1314444734569555,
+      "learning_rate": 6.227857564711349e-07,
+      "loss": 0.3647,
+      "step": 12159
+    },
+    {
+      "epoch": 0.8443271767810027,
+      "grad_norm": 4.923585305556519,
+      "learning_rate": 6.22242378108216e-07,
+      "loss": 0.5856,
+      "step": 12160
+    },
+    {
+      "epoch": 0.8443966115817247,
+      "grad_norm": 3.1566700945344124,
+      "learning_rate": 6.216992211673251e-07,
+      "loss": 0.1857,
+      "step": 12161
+    },
+    {
+      "epoch": 0.8444660463824469,
+      "grad_norm": 4.453711800775669,
+      "learning_rate": 6.211562856759329e-07,
+      "loss": 0.4809,
+      "step": 12162
+    },
+    {
+      "epoch": 0.844535481183169,
+      "grad_norm": 3.564672612322839,
+      "learning_rate": 6.206135716615008e-07,
+      "loss": 0.4273,
+      "step": 12163
+    },
+    {
+      "epoch": 0.8446049159838911,
+      "grad_norm": 4.762640826053146,
+      "learning_rate": 6.200710791514786e-07,
+      "loss": 0.4764,
+      "step": 12164
+    },
+    {
+      "epoch": 0.8446743507846133,
+      "grad_norm": 3.9651077330421347,
+      "learning_rate": 6.195288081733053e-07,
+      "loss": 0.3398,
+      "step": 12165
+    },
+    {
+      "epoch": 0.8447437855853354,
+      "grad_norm": 4.306728738360438,
+      "learning_rate": 6.189867587544069e-07,
+      "loss": 0.4535,
+      "step": 12166
+    },
+    {
+      "epoch": 0.8448132203860574,
+      "grad_norm": 4.408322192206743,
+      "learning_rate": 6.184449309222029e-07,
+      "loss": 0.6838,
+      "step": 12167
+    },
+    {
+      "epoch": 0.8448826551867796,
+      "grad_norm": 3.91013212953302,
+      "learning_rate": 6.17903324704095e-07,
+      "loss": 0.2568,
+      "step": 12168
+    },
+    {
+      "epoch": 0.8449520899875017,
+      "grad_norm": 3.7460925152929634,
+      "learning_rate": 6.173619401274772e-07,
+      "loss": 0.2804,
+      "step": 12169
+    },
+    {
+      "epoch": 0.8450215247882239,
+      "grad_norm": 3.915106233738582,
+      "learning_rate": 6.168207772197354e-07,
+      "loss": 0.4011,
+      "step": 12170
+    },
+    {
+      "epoch": 0.845090959588946,
+      "grad_norm": 3.9746618336392054,
+      "learning_rate": 6.162798360082367e-07,
+      "loss": 0.2284,
+      "step": 12171
+    },
+    {
+      "epoch": 0.8451603943896681,
+      "grad_norm": 4.499429806141879,
+      "learning_rate": 6.157391165203447e-07,
+      "loss": 0.4028,
+      "step": 12172
+    },
+    {
+      "epoch": 0.8452298291903902,
+      "grad_norm": 4.564079409016079,
+      "learning_rate": 6.151986187834064e-07,
+      "loss": 0.3923,
+      "step": 12173
+    },
+    {
+      "epoch": 0.8452992639911123,
+      "grad_norm": 3.1485269575232784,
+      "learning_rate": 6.14658342824761e-07,
+      "loss": 0.2613,
+      "step": 12174
+    },
+    {
+      "epoch": 0.8453686987918345,
+      "grad_norm": 4.917737945703657,
+      "learning_rate": 6.14118288671734e-07,
+      "loss": 0.3409,
+      "step": 12175
+    },
+    {
+      "epoch": 0.8454381335925566,
+      "grad_norm": 4.312264302844288,
+      "learning_rate": 6.135784563516406e-07,
+      "loss": 0.4679,
+      "step": 12176
+    },
+    {
+      "epoch": 0.8455075683932787,
+      "grad_norm": 5.401662846096275,
+      "learning_rate": 6.130388458917857e-07,
+      "loss": 0.5964,
+      "step": 12177
+    },
+    {
+      "epoch": 0.8455770031940009,
+      "grad_norm": 3.6706491459442687,
+      "learning_rate": 6.124994573194614e-07,
+      "loss": 0.5256,
+      "step": 12178
+    },
+    {
+      "epoch": 0.8456464379947229,
+      "grad_norm": 5.132920500794997,
+      "learning_rate": 6.119602906619504e-07,
+      "loss": 0.6762,
+      "step": 12179
+    },
+    {
+      "epoch": 0.845715872795445,
+      "grad_norm": 3.9090582687781508,
+      "learning_rate": 6.114213459465207e-07,
+      "loss": 0.3513,
+      "step": 12180
+    },
+    {
+      "epoch": 0.8457853075961672,
+      "grad_norm": 4.71666456538568,
+      "learning_rate": 6.108826232004362e-07,
+      "loss": 0.5316,
+      "step": 12181
+    },
+    {
+      "epoch": 0.8458547423968893,
+      "grad_norm": 3.163182654861291,
+      "learning_rate": 6.103441224509404e-07,
+      "loss": 0.2611,
+      "step": 12182
+    },
+    {
+      "epoch": 0.8459241771976115,
+      "grad_norm": 3.498578759119988,
+      "learning_rate": 6.098058437252708e-07,
+      "loss": 0.2379,
+      "step": 12183
+    },
+    {
+      "epoch": 0.8459936119983336,
+      "grad_norm": 4.715978149635673,
+      "learning_rate": 6.092677870506553e-07,
+      "loss": 0.5607,
+      "step": 12184
+    },
+    {
+      "epoch": 0.8460630467990556,
+      "grad_norm": 2.811158546134452,
+      "learning_rate": 6.087299524543061e-07,
+      "loss": 0.2497,
+      "step": 12185
+    },
+    {
+      "epoch": 0.8461324815997778,
+      "grad_norm": 3.967570890139553,
+      "learning_rate": 6.08192339963426e-07,
+      "loss": 0.4508,
+      "step": 12186
+    },
+    {
+      "epoch": 0.8462019164004999,
+      "grad_norm": 4.445040087713319,
+      "learning_rate": 6.076549496052086e-07,
+      "loss": 0.3412,
+      "step": 12187
+    },
+    {
+      "epoch": 0.8462713512012221,
+      "grad_norm": 4.841178762659162,
+      "learning_rate": 6.071177814068347e-07,
+      "loss": 0.555,
+      "step": 12188
+    },
+    {
+      "epoch": 0.8463407860019442,
+      "grad_norm": 3.9330242301644835,
+      "learning_rate": 6.065808353954706e-07,
+      "loss": 0.3238,
+      "step": 12189
+    },
+    {
+      "epoch": 0.8464102208026663,
+      "grad_norm": 4.060568700525655,
+      "learning_rate": 6.060441115982768e-07,
+      "loss": 0.4425,
+      "step": 12190
+    },
+    {
+      "epoch": 0.8464796556033884,
+      "grad_norm": 3.0285159488702336,
+      "learning_rate": 6.055076100424001e-07,
+      "loss": 0.4352,
+      "step": 12191
+    },
+    {
+      "epoch": 0.8465490904041105,
+      "grad_norm": 5.030166507950769,
+      "learning_rate": 6.049713307549759e-07,
+      "loss": 0.6109,
+      "step": 12192
+    },
+    {
+      "epoch": 0.8466185252048327,
+      "grad_norm": 3.7222376207847687,
+      "learning_rate": 6.044352737631281e-07,
+      "loss": 0.3208,
+      "step": 12193
+    },
+    {
+      "epoch": 0.8466879600055548,
+      "grad_norm": 3.5891667223535757,
+      "learning_rate": 6.038994390939706e-07,
+      "loss": 0.3951,
+      "step": 12194
+    },
+    {
+      "epoch": 0.8467573948062769,
+      "grad_norm": 4.92483633963376,
+      "learning_rate": 6.033638267746045e-07,
+      "loss": 0.3356,
+      "step": 12195
+    },
+    {
+      "epoch": 0.8468268296069991,
+      "grad_norm": 2.943420523708218,
+      "learning_rate": 6.028284368321214e-07,
+      "loss": 0.2848,
+      "step": 12196
+    },
+    {
+      "epoch": 0.8468962644077211,
+      "grad_norm": 4.739292868027685,
+      "learning_rate": 6.022932692936001e-07,
+      "loss": 0.5303,
+      "step": 12197
+    },
+    {
+      "epoch": 0.8469656992084432,
+      "grad_norm": 4.140369089546648,
+      "learning_rate": 6.01758324186108e-07,
+      "loss": 0.3831,
+      "step": 12198
+    },
+    {
+      "epoch": 0.8470351340091654,
+      "grad_norm": 4.6058156256297575,
+      "learning_rate": 6.012236015367046e-07,
+      "loss": 0.6212,
+      "step": 12199
+    },
+    {
+      "epoch": 0.8471045688098875,
+      "grad_norm": 3.605251457716223,
+      "learning_rate": 6.00689101372432e-07,
+      "loss": 0.213,
+      "step": 12200
+    },
+    {
+      "epoch": 0.8471740036106097,
+      "grad_norm": 4.8381112772977435,
+      "learning_rate": 6.001548237203275e-07,
+      "loss": 0.6777,
+      "step": 12201
+    },
+    {
+      "epoch": 0.8472434384113318,
+      "grad_norm": 4.121756155822516,
+      "learning_rate": 5.996207686074146e-07,
+      "loss": 0.4533,
+      "step": 12202
+    },
+    {
+      "epoch": 0.8473128732120538,
+      "grad_norm": 4.573705556978793,
+      "learning_rate": 5.99086936060701e-07,
+      "loss": 0.5476,
+      "step": 12203
+    },
+    {
+      "epoch": 0.847382308012776,
+      "grad_norm": 3.4718319727237623,
+      "learning_rate": 5.985533261071919e-07,
+      "loss": 0.3684,
+      "step": 12204
+    },
+    {
+      "epoch": 0.8474517428134981,
+      "grad_norm": 5.669360120154309,
+      "learning_rate": 5.98019938773875e-07,
+      "loss": 0.2917,
+      "step": 12205
+    },
+    {
+      "epoch": 0.8475211776142203,
+      "grad_norm": 5.523199317575706,
+      "learning_rate": 5.974867740877282e-07,
+      "loss": 0.9849,
+      "step": 12206
+    },
+    {
+      "epoch": 0.8475906124149424,
+      "grad_norm": 3.4667141399365393,
+      "learning_rate": 5.969538320757185e-07,
+      "loss": 0.2452,
+      "step": 12207
+    },
+    {
+      "epoch": 0.8476600472156645,
+      "grad_norm": 4.070831459619358,
+      "learning_rate": 5.964211127648018e-07,
+      "loss": 0.634,
+      "step": 12208
+    },
+    {
+      "epoch": 0.8477294820163866,
+      "grad_norm": 3.798694521408924,
+      "learning_rate": 5.958886161819222e-07,
+      "loss": 0.5122,
+      "step": 12209
+    },
+    {
+      "epoch": 0.8477989168171087,
+      "grad_norm": 4.021971530101663,
+      "learning_rate": 5.953563423540126e-07,
+      "loss": 0.5093,
+      "step": 12210
+    },
+    {
+      "epoch": 0.8478683516178308,
+      "grad_norm": 3.712045113595136,
+      "learning_rate": 5.948242913079955e-07,
+      "loss": 0.3583,
+      "step": 12211
+    },
+    {
+      "epoch": 0.847937786418553,
+      "grad_norm": 3.2154494100360065,
+      "learning_rate": 5.942924630707797e-07,
+      "loss": 0.3067,
+      "step": 12212
+    },
+    {
+      "epoch": 0.8480072212192751,
+      "grad_norm": 4.096350515960314,
+      "learning_rate": 5.937608576692678e-07,
+      "loss": 0.2565,
+      "step": 12213
+    },
+    {
+      "epoch": 0.8480766560199973,
+      "grad_norm": 3.6223500079978774,
+      "learning_rate": 5.932294751303442e-07,
+      "loss": 0.4931,
+      "step": 12214
+    },
+    {
+      "epoch": 0.8481460908207193,
+      "grad_norm": 4.028436579117262,
+      "learning_rate": 5.926983154808863e-07,
+      "loss": 0.4036,
+      "step": 12215
+    },
+    {
+      "epoch": 0.8482155256214414,
+      "grad_norm": 1.8809456631074808,
+      "learning_rate": 5.921673787477628e-07,
+      "loss": 0.0989,
+      "step": 12216
+    },
+    {
+      "epoch": 0.8482849604221636,
+      "grad_norm": 3.6153824689662235,
+      "learning_rate": 5.916366649578242e-07,
+      "loss": 0.2135,
+      "step": 12217
+    },
+    {
+      "epoch": 0.8483543952228857,
+      "grad_norm": 4.028119784124289,
+      "learning_rate": 5.911061741379137e-07,
+      "loss": 0.3683,
+      "step": 12218
+    },
+    {
+      "epoch": 0.8484238300236079,
+      "grad_norm": 4.0729897327189555,
+      "learning_rate": 5.905759063148653e-07,
+      "loss": 0.3428,
+      "step": 12219
+    },
+    {
+      "epoch": 0.84849326482433,
+      "grad_norm": 2.3999245601205366,
+      "learning_rate": 5.900458615154986e-07,
+      "loss": 0.1935,
+      "step": 12220
+    },
+    {
+      "epoch": 0.848562699625052,
+      "grad_norm": 4.071949364961603,
+      "learning_rate": 5.8951603976662e-07,
+      "loss": 0.647,
+      "step": 12221
+    },
+    {
+      "epoch": 0.8486321344257742,
+      "grad_norm": 3.6461806022547014,
+      "learning_rate": 5.889864410950308e-07,
+      "loss": 0.3233,
+      "step": 12222
+    },
+    {
+      "epoch": 0.8487015692264963,
+      "grad_norm": 3.9312077660752838,
+      "learning_rate": 5.884570655275151e-07,
+      "loss": 0.4189,
+      "step": 12223
+    },
+    {
+      "epoch": 0.8487710040272184,
+      "grad_norm": 3.752492892736189,
+      "learning_rate": 5.879279130908499e-07,
+      "loss": 0.4394,
+      "step": 12224
+    },
+    {
+      "epoch": 0.8488404388279406,
+      "grad_norm": 4.192219970280358,
+      "learning_rate": 5.873989838117977e-07,
+      "loss": 0.5021,
+      "step": 12225
+    },
+    {
+      "epoch": 0.8489098736286627,
+      "grad_norm": 4.6115288604517675,
+      "learning_rate": 5.86870277717112e-07,
+      "loss": 0.5447,
+      "step": 12226
+    },
+    {
+      "epoch": 0.8489793084293848,
+      "grad_norm": 4.565363770737357,
+      "learning_rate": 5.863417948335337e-07,
+      "loss": 0.4384,
+      "step": 12227
+    },
+    {
+      "epoch": 0.8490487432301069,
+      "grad_norm": 3.8487643089359826,
+      "learning_rate": 5.858135351877931e-07,
+      "loss": 0.3198,
+      "step": 12228
+    },
+    {
+      "epoch": 0.849118178030829,
+      "grad_norm": 3.3054897375667784,
+      "learning_rate": 5.852854988066087e-07,
+      "loss": 0.5101,
+      "step": 12229
+    },
+    {
+      "epoch": 0.8491876128315512,
+      "grad_norm": 3.6064656371773283,
+      "learning_rate": 5.847576857166875e-07,
+      "loss": 0.323,
+      "step": 12230
+    },
+    {
+      "epoch": 0.8492570476322733,
+      "grad_norm": 4.279862648788139,
+      "learning_rate": 5.842300959447283e-07,
+      "loss": 0.5092,
+      "step": 12231
+    },
+    {
+      "epoch": 0.8493264824329955,
+      "grad_norm": 4.780175623324486,
+      "learning_rate": 5.83702729517412e-07,
+      "loss": 0.4791,
+      "step": 12232
+    },
+    {
+      "epoch": 0.8493959172337175,
+      "grad_norm": 3.239389872357473,
+      "learning_rate": 5.831755864614152e-07,
+      "loss": 0.2071,
+      "step": 12233
+    },
+    {
+      "epoch": 0.8494653520344396,
+      "grad_norm": 4.5088513698841455,
+      "learning_rate": 5.826486668034003e-07,
+      "loss": 0.5925,
+      "step": 12234
+    },
+    {
+      "epoch": 0.8495347868351618,
+      "grad_norm": 4.720196593058774,
+      "learning_rate": 5.821219705700155e-07,
+      "loss": 0.5545,
+      "step": 12235
+    },
+    {
+      "epoch": 0.8496042216358839,
+      "grad_norm": 4.2691571442845735,
+      "learning_rate": 5.815954977879029e-07,
+      "loss": 0.5947,
+      "step": 12236
+    },
+    {
+      "epoch": 0.849673656436606,
+      "grad_norm": 5.1415768518600276,
+      "learning_rate": 5.810692484836905e-07,
+      "loss": 0.5419,
+      "step": 12237
+    },
+    {
+      "epoch": 0.8497430912373282,
+      "grad_norm": 5.198268022727231,
+      "learning_rate": 5.805432226839952e-07,
+      "loss": 0.446,
+      "step": 12238
+    },
+    {
+      "epoch": 0.8498125260380502,
+      "grad_norm": 4.61904349401658,
+      "learning_rate": 5.800174204154224e-07,
+      "loss": 0.4455,
+      "step": 12239
+    },
+    {
+      "epoch": 0.8498819608387724,
+      "grad_norm": 4.1301285351195265,
+      "learning_rate": 5.794918417045669e-07,
+      "loss": 0.4342,
+      "step": 12240
+    },
+    {
+      "epoch": 0.8499513956394945,
+      "grad_norm": 3.3955045439556173,
+      "learning_rate": 5.789664865780126e-07,
+      "loss": 0.3304,
+      "step": 12241
+    },
+    {
+      "epoch": 0.8500208304402166,
+      "grad_norm": 2.987779453931659,
+      "learning_rate": 5.784413550623297e-07,
+      "loss": 0.2258,
+      "step": 12242
+    },
+    {
+      "epoch": 0.8500902652409388,
+      "grad_norm": 4.349686047304896,
+      "learning_rate": 5.779164471840803e-07,
+      "loss": 0.3548,
+      "step": 12243
+    },
+    {
+      "epoch": 0.8501597000416609,
+      "grad_norm": 3.327286912739435,
+      "learning_rate": 5.773917629698122e-07,
+      "loss": 0.3871,
+      "step": 12244
+    },
+    {
+      "epoch": 0.850229134842383,
+      "grad_norm": 3.948618879192863,
+      "learning_rate": 5.768673024460658e-07,
+      "loss": 0.5174,
+      "step": 12245
+    },
+    {
+      "epoch": 0.8502985696431051,
+      "grad_norm": 3.862347198483988,
+      "learning_rate": 5.763430656393654e-07,
+      "loss": 0.5158,
+      "step": 12246
+    },
+    {
+      "epoch": 0.8503680044438272,
+      "grad_norm": 4.317379985292529,
+      "learning_rate": 5.758190525762259e-07,
+      "loss": 0.5908,
+      "step": 12247
+    },
+    {
+      "epoch": 0.8504374392445494,
+      "grad_norm": 5.090662349349204,
+      "learning_rate": 5.752952632831538e-07,
+      "loss": 0.5552,
+      "step": 12248
+    },
+    {
+      "epoch": 0.8505068740452715,
+      "grad_norm": 3.975057580438571,
+      "learning_rate": 5.747716977866408e-07,
+      "loss": 0.3974,
+      "step": 12249
+    },
+    {
+      "epoch": 0.8505763088459937,
+      "grad_norm": 4.071979506423305,
+      "learning_rate": 5.742483561131662e-07,
+      "loss": 0.4792,
+      "step": 12250
+    },
+    {
+      "epoch": 0.8506457436467157,
+      "grad_norm": 4.314378412000532,
+      "learning_rate": 5.737252382892028e-07,
+      "loss": 0.5487,
+      "step": 12251
+    },
+    {
+      "epoch": 0.8507151784474378,
+      "grad_norm": 2.984807290709036,
+      "learning_rate": 5.732023443412077e-07,
+      "loss": 0.3208,
+      "step": 12252
+    },
+    {
+      "epoch": 0.85078461324816,
+      "grad_norm": 4.8416319603725,
+      "learning_rate": 5.726796742956292e-07,
+      "loss": 0.5465,
+      "step": 12253
+    },
+    {
+      "epoch": 0.8508540480488821,
+      "grad_norm": 4.860150716542382,
+      "learning_rate": 5.721572281789028e-07,
+      "loss": 0.4569,
+      "step": 12254
+    },
+    {
+      "epoch": 0.8509234828496042,
+      "grad_norm": 3.361643017469316,
+      "learning_rate": 5.71635006017453e-07,
+      "loss": 0.3839,
+      "step": 12255
+    },
+    {
+      "epoch": 0.8509929176503264,
+      "grad_norm": 4.177977318689177,
+      "learning_rate": 5.711130078376942e-07,
+      "loss": 0.387,
+      "step": 12256
+    },
+    {
+      "epoch": 0.8510623524510484,
+      "grad_norm": 3.5248143272627908,
+      "learning_rate": 5.705912336660275e-07,
+      "loss": 0.43,
+      "step": 12257
+    },
+    {
+      "epoch": 0.8511317872517706,
+      "grad_norm": 6.064225970056006,
+      "learning_rate": 5.700696835288438e-07,
+      "loss": 0.5049,
+      "step": 12258
+    },
+    {
+      "epoch": 0.8512012220524927,
+      "grad_norm": 4.565825394973027,
+      "learning_rate": 5.695483574525229e-07,
+      "loss": 0.3958,
+      "step": 12259
+    },
+    {
+      "epoch": 0.8512706568532148,
+      "grad_norm": 5.298029660504252,
+      "learning_rate": 5.69027255463433e-07,
+      "loss": 0.5505,
+      "step": 12260
+    },
+    {
+      "epoch": 0.851340091653937,
+      "grad_norm": 5.633766162176873,
+      "learning_rate": 5.68506377587929e-07,
+      "loss": 0.7292,
+      "step": 12261
+    },
+    {
+      "epoch": 0.8514095264546591,
+      "grad_norm": 5.929971081247732,
+      "learning_rate": 5.67985723852359e-07,
+      "loss": 0.3402,
+      "step": 12262
+    },
+    {
+      "epoch": 0.8514789612553813,
+      "grad_norm": 3.344784161895592,
+      "learning_rate": 5.674652942830572e-07,
+      "loss": 0.4162,
+      "step": 12263
+    },
+    {
+      "epoch": 0.8515483960561033,
+      "grad_norm": 3.6895928471744974,
+      "learning_rate": 5.669450889063427e-07,
+      "loss": 0.3938,
+      "step": 12264
+    },
+    {
+      "epoch": 0.8516178308568254,
+      "grad_norm": 3.983689101468244,
+      "learning_rate": 5.664251077485305e-07,
+      "loss": 0.2958,
+      "step": 12265
+    },
+    {
+      "epoch": 0.8516872656575476,
+      "grad_norm": 3.5481353379054132,
+      "learning_rate": 5.659053508359203e-07,
+      "loss": 0.398,
+      "step": 12266
+    },
+    {
+      "epoch": 0.8517567004582697,
+      "grad_norm": 2.6459195208619652,
+      "learning_rate": 5.65385818194798e-07,
+      "loss": 0.1637,
+      "step": 12267
+    },
+    {
+      "epoch": 0.8518261352589918,
+      "grad_norm": 3.788530726971908,
+      "learning_rate": 5.648665098514439e-07,
+      "loss": 0.4592,
+      "step": 12268
+    },
+    {
+      "epoch": 0.851895570059714,
+      "grad_norm": 4.23138568261878,
+      "learning_rate": 5.64347425832123e-07,
+      "loss": 0.6579,
+      "step": 12269
+    },
+    {
+      "epoch": 0.851965004860436,
+      "grad_norm": 3.839308367714743,
+      "learning_rate": 5.638285661630899e-07,
+      "loss": 0.4547,
+      "step": 12270
+    },
+    {
+      "epoch": 0.8520344396611582,
+      "grad_norm": 4.086932317361261,
+      "learning_rate": 5.633099308705886e-07,
+      "loss": 0.4562,
+      "step": 12271
+    },
+    {
+      "epoch": 0.8521038744618803,
+      "grad_norm": 4.421305315132611,
+      "learning_rate": 5.627915199808498e-07,
+      "loss": 0.4116,
+      "step": 12272
+    },
+    {
+      "epoch": 0.8521733092626024,
+      "grad_norm": 3.8091753897441585,
+      "learning_rate": 5.622733335200941e-07,
+      "loss": 0.6617,
+      "step": 12273
+    },
+    {
+      "epoch": 0.8522427440633246,
+      "grad_norm": 3.519101651242471,
+      "learning_rate": 5.61755371514534e-07,
+      "loss": 0.4311,
+      "step": 12274
+    },
+    {
+      "epoch": 0.8523121788640466,
+      "grad_norm": 3.6878332049447606,
+      "learning_rate": 5.612376339903636e-07,
+      "loss": 0.2974,
+      "step": 12275
+    },
+    {
+      "epoch": 0.8523816136647688,
+      "grad_norm": 4.630776964236321,
+      "learning_rate": 5.607201209737701e-07,
+      "loss": 0.4619,
+      "step": 12276
+    },
+    {
+      "epoch": 0.8524510484654909,
+      "grad_norm": 3.9828606842456584,
+      "learning_rate": 5.602028324909315e-07,
+      "loss": 0.5292,
+      "step": 12277
+    },
+    {
+      "epoch": 0.852520483266213,
+      "grad_norm": 4.284576995416832,
+      "learning_rate": 5.596857685680085e-07,
+      "loss": 0.5218,
+      "step": 12278
+    },
+    {
+      "epoch": 0.8525899180669352,
+      "grad_norm": 15.363578501773345,
+      "learning_rate": 5.591689292311536e-07,
+      "loss": 0.6177,
+      "step": 12279
+    },
+    {
+      "epoch": 0.8526593528676573,
+      "grad_norm": 3.227742148113535,
+      "learning_rate": 5.586523145065103e-07,
+      "loss": 0.1368,
+      "step": 12280
+    },
+    {
+      "epoch": 0.8527287876683793,
+      "grad_norm": 3.3265515992038344,
+      "learning_rate": 5.581359244202073e-07,
+      "loss": 0.3771,
+      "step": 12281
+    },
+    {
+      "epoch": 0.8527982224691015,
+      "grad_norm": 4.2286324474483346,
+      "learning_rate": 5.576197589983629e-07,
+      "loss": 0.4629,
+      "step": 12282
+    },
+    {
+      "epoch": 0.8528676572698236,
+      "grad_norm": 3.598579892785284,
+      "learning_rate": 5.57103818267084e-07,
+      "loss": 0.3721,
+      "step": 12283
+    },
+    {
+      "epoch": 0.8529370920705458,
+      "grad_norm": 5.170346212615741,
+      "learning_rate": 5.565881022524667e-07,
+      "loss": 0.5754,
+      "step": 12284
+    },
+    {
+      "epoch": 0.8530065268712679,
+      "grad_norm": 3.3997444710626423,
+      "learning_rate": 5.560726109805947e-07,
+      "loss": 0.3155,
+      "step": 12285
+    },
+    {
+      "epoch": 0.85307596167199,
+      "grad_norm": 4.791155592801918,
+      "learning_rate": 5.555573444775419e-07,
+      "loss": 0.6648,
+      "step": 12286
+    },
+    {
+      "epoch": 0.8531453964727121,
+      "grad_norm": 4.327778893074458,
+      "learning_rate": 5.550423027693697e-07,
+      "loss": 0.4721,
+      "step": 12287
+    },
+    {
+      "epoch": 0.8532148312734342,
+      "grad_norm": 3.6147258471487946,
+      "learning_rate": 5.54527485882127e-07,
+      "loss": 0.3952,
+      "step": 12288
+    },
+    {
+      "epoch": 0.8532842660741564,
+      "grad_norm": 3.3165743663401894,
+      "learning_rate": 5.540128938418543e-07,
+      "loss": 0.2843,
+      "step": 12289
+    },
+    {
+      "epoch": 0.8533537008748785,
+      "grad_norm": 3.368433799948595,
+      "learning_rate": 5.534985266745774e-07,
+      "loss": 0.3802,
+      "step": 12290
+    },
+    {
+      "epoch": 0.8534231356756006,
+      "grad_norm": 3.933978091796547,
+      "learning_rate": 5.529843844063154e-07,
+      "loss": 0.5325,
+      "step": 12291
+    },
+    {
+      "epoch": 0.8534925704763228,
+      "grad_norm": 3.43378183232829,
+      "learning_rate": 5.524704670630704e-07,
+      "loss": 0.3722,
+      "step": 12292
+    },
+    {
+      "epoch": 0.8535620052770448,
+      "grad_norm": 3.738411061960536,
+      "learning_rate": 5.519567746708355e-07,
+      "loss": 0.3663,
+      "step": 12293
+    },
+    {
+      "epoch": 0.8536314400777669,
+      "grad_norm": 3.9691906990800363,
+      "learning_rate": 5.514433072555941e-07,
+      "loss": 0.4228,
+      "step": 12294
+    },
+    {
+      "epoch": 0.8537008748784891,
+      "grad_norm": 5.46245184185958,
+      "learning_rate": 5.509300648433174e-07,
+      "loss": 0.51,
+      "step": 12295
+    },
+    {
+      "epoch": 0.8537703096792112,
+      "grad_norm": 3.277630844259243,
+      "learning_rate": 5.504170474599618e-07,
+      "loss": 0.3504,
+      "step": 12296
+    },
+    {
+      "epoch": 0.8538397444799334,
+      "grad_norm": 3.4549482049710694,
+      "learning_rate": 5.499042551314776e-07,
+      "loss": 0.4014,
+      "step": 12297
+    },
+    {
+      "epoch": 0.8539091792806555,
+      "grad_norm": 4.019008467868914,
+      "learning_rate": 5.493916878838013e-07,
+      "loss": 0.5574,
+      "step": 12298
+    },
+    {
+      "epoch": 0.8539786140813775,
+      "grad_norm": 4.21839809196305,
+      "learning_rate": 5.488793457428566e-07,
+      "loss": 0.4218,
+      "step": 12299
+    },
+    {
+      "epoch": 0.8540480488820997,
+      "grad_norm": 2.4801310627169664,
+      "learning_rate": 5.483672287345576e-07,
+      "loss": 0.1504,
+      "step": 12300
+    },
+    {
+      "epoch": 0.8541174836828218,
+      "grad_norm": 3.6559023850592203,
+      "learning_rate": 5.478553368848072e-07,
+      "loss": 0.473,
+      "step": 12301
+    },
+    {
+      "epoch": 0.854186918483544,
+      "grad_norm": 5.42300207878597,
+      "learning_rate": 5.473436702194956e-07,
+      "loss": 0.6136,
+      "step": 12302
+    },
+    {
+      "epoch": 0.8542563532842661,
+      "grad_norm": 3.3119062527166507,
+      "learning_rate": 5.468322287645034e-07,
+      "loss": 0.3005,
+      "step": 12303
+    },
+    {
+      "epoch": 0.8543257880849882,
+      "grad_norm": 3.388901200465658,
+      "learning_rate": 5.463210125456969e-07,
+      "loss": 0.1815,
+      "step": 12304
+    },
+    {
+      "epoch": 0.8543952228857103,
+      "grad_norm": 3.9925382658988697,
+      "learning_rate": 5.458100215889339e-07,
+      "loss": 0.4848,
+      "step": 12305
+    },
+    {
+      "epoch": 0.8544646576864324,
+      "grad_norm": 4.103495620680537,
+      "learning_rate": 5.45299255920061e-07,
+      "loss": 0.4879,
+      "step": 12306
+    },
+    {
+      "epoch": 0.8545340924871546,
+      "grad_norm": 6.961395361230468,
+      "learning_rate": 5.447887155649101e-07,
+      "loss": 0.7065,
+      "step": 12307
+    },
+    {
+      "epoch": 0.8546035272878767,
+      "grad_norm": 3.5705730178129156,
+      "learning_rate": 5.442784005493041e-07,
+      "loss": 0.3166,
+      "step": 12308
+    },
+    {
+      "epoch": 0.8546729620885988,
+      "grad_norm": 4.007083496805182,
+      "learning_rate": 5.43768310899056e-07,
+      "loss": 0.4543,
+      "step": 12309
+    },
+    {
+      "epoch": 0.854742396889321,
+      "grad_norm": 3.7626892869195108,
+      "learning_rate": 5.432584466399632e-07,
+      "loss": 0.4297,
+      "step": 12310
+    },
+    {
+      "epoch": 0.854811831690043,
+      "grad_norm": 5.48392820069255,
+      "learning_rate": 5.427488077978143e-07,
+      "loss": 0.4668,
+      "step": 12311
+    },
+    {
+      "epoch": 0.8548812664907651,
+      "grad_norm": 4.754016039017512,
+      "learning_rate": 5.422393943983878e-07,
+      "loss": 0.4999,
+      "step": 12312
+    },
+    {
+      "epoch": 0.8549507012914873,
+      "grad_norm": 3.5486684278149117,
+      "learning_rate": 5.417302064674485e-07,
+      "loss": 0.2995,
+      "step": 12313
+    },
+    {
+      "epoch": 0.8550201360922094,
+      "grad_norm": 4.5036436618144435,
+      "learning_rate": 5.412212440307501e-07,
+      "loss": 0.3581,
+      "step": 12314
+    },
+    {
+      "epoch": 0.8550895708929316,
+      "grad_norm": 4.509403789310186,
+      "learning_rate": 5.407125071140363e-07,
+      "loss": 0.4214,
+      "step": 12315
+    },
+    {
+      "epoch": 0.8551590056936537,
+      "grad_norm": 2.6874951927372557,
+      "learning_rate": 5.402039957430378e-07,
+      "loss": 0.282,
+      "step": 12316
+    },
+    {
+      "epoch": 0.8552284404943757,
+      "grad_norm": 4.583547984979454,
+      "learning_rate": 5.396957099434741e-07,
+      "loss": 0.5758,
+      "step": 12317
+    },
+    {
+      "epoch": 0.8552978752950979,
+      "grad_norm": 4.261571638624774,
+      "learning_rate": 5.391876497410548e-07,
+      "loss": 0.4011,
+      "step": 12318
+    },
+    {
+      "epoch": 0.85536731009582,
+      "grad_norm": 3.932238622598854,
+      "learning_rate": 5.386798151614764e-07,
+      "loss": 0.4097,
+      "step": 12319
+    },
+    {
+      "epoch": 0.8554367448965422,
+      "grad_norm": 3.895268006392734,
+      "learning_rate": 5.38172206230424e-07,
+      "loss": 0.5534,
+      "step": 12320
+    },
+    {
+      "epoch": 0.8555061796972643,
+      "grad_norm": 3.509724360885379,
+      "learning_rate": 5.376648229735731e-07,
+      "loss": 0.3486,
+      "step": 12321
+    },
+    {
+      "epoch": 0.8555756144979864,
+      "grad_norm": 5.046432003675272,
+      "learning_rate": 5.37157665416585e-07,
+      "loss": 0.6183,
+      "step": 12322
+    },
+    {
+      "epoch": 0.8556450492987085,
+      "grad_norm": 3.8420358184632035,
+      "learning_rate": 5.366507335851129e-07,
+      "loss": 0.476,
+      "step": 12323
+    },
+    {
+      "epoch": 0.8557144840994306,
+      "grad_norm": 3.2741549365089324,
+      "learning_rate": 5.361440275047969e-07,
+      "loss": 0.2644,
+      "step": 12324
+    },
+    {
+      "epoch": 0.8557839189001527,
+      "grad_norm": 4.079362184383126,
+      "learning_rate": 5.356375472012631e-07,
+      "loss": 0.464,
+      "step": 12325
+    },
+    {
+      "epoch": 0.8558533537008749,
+      "grad_norm": 4.165102903331528,
+      "learning_rate": 5.351312927001312e-07,
+      "loss": 0.4577,
+      "step": 12326
+    },
+    {
+      "epoch": 0.855922788501597,
+      "grad_norm": 5.622002796731427,
+      "learning_rate": 5.346252640270072e-07,
+      "loss": 0.8521,
+      "step": 12327
+    },
+    {
+      "epoch": 0.8559922233023192,
+      "grad_norm": 4.623472022132399,
+      "learning_rate": 5.341194612074824e-07,
+      "loss": 0.6321,
+      "step": 12328
+    },
+    {
+      "epoch": 0.8560616581030412,
+      "grad_norm": 3.345821449163678,
+      "learning_rate": 5.336138842671429e-07,
+      "loss": 0.3314,
+      "step": 12329
+    },
+    {
+      "epoch": 0.8561310929037633,
+      "grad_norm": 4.068165998365277,
+      "learning_rate": 5.331085332315594e-07,
+      "loss": 0.3998,
+      "step": 12330
+    },
+    {
+      "epoch": 0.8562005277044855,
+      "grad_norm": 3.5183321601005453,
+      "learning_rate": 5.32603408126291e-07,
+      "loss": 0.3596,
+      "step": 12331
+    },
+    {
+      "epoch": 0.8562699625052076,
+      "grad_norm": 2.6483731422259247,
+      "learning_rate": 5.320985089768876e-07,
+      "loss": 0.2351,
+      "step": 12332
+    },
+    {
+      "epoch": 0.8563393973059298,
+      "grad_norm": 4.7137919415894345,
+      "learning_rate": 5.315938358088851e-07,
+      "loss": 0.5836,
+      "step": 12333
+    },
+    {
+      "epoch": 0.8564088321066519,
+      "grad_norm": 3.2532818523197617,
+      "learning_rate": 5.310893886478108e-07,
+      "loss": 0.3109,
+      "step": 12334
+    },
+    {
+      "epoch": 0.8564782669073739,
+      "grad_norm": 3.5252464512522494,
+      "learning_rate": 5.305851675191781e-07,
+      "loss": 0.322,
+      "step": 12335
+    },
+    {
+      "epoch": 0.8565477017080961,
+      "grad_norm": 7.649791104076804,
+      "learning_rate": 5.3008117244849e-07,
+      "loss": 0.6806,
+      "step": 12336
+    },
+    {
+      "epoch": 0.8566171365088182,
+      "grad_norm": 4.239958439249266,
+      "learning_rate": 5.295774034612372e-07,
+      "loss": 0.4313,
+      "step": 12337
+    },
+    {
+      "epoch": 0.8566865713095403,
+      "grad_norm": 3.39858900405438,
+      "learning_rate": 5.290738605829026e-07,
+      "loss": 0.2401,
+      "step": 12338
+    },
+    {
+      "epoch": 0.8567560061102625,
+      "grad_norm": 4.1558241732338725,
+      "learning_rate": 5.285705438389521e-07,
+      "loss": 0.4439,
+      "step": 12339
+    },
+    {
+      "epoch": 0.8568254409109846,
+      "grad_norm": 3.8964469967319126,
+      "learning_rate": 5.280674532548424e-07,
+      "loss": 0.3994,
+      "step": 12340
+    },
+    {
+      "epoch": 0.8568948757117067,
+      "grad_norm": 4.267217188862538,
+      "learning_rate": 5.275645888560233e-07,
+      "loss": 0.5703,
+      "step": 12341
+    },
+    {
+      "epoch": 0.8569643105124288,
+      "grad_norm": 2.9280935191690065,
+      "learning_rate": 5.270619506679236e-07,
+      "loss": 0.229,
+      "step": 12342
+    },
+    {
+      "epoch": 0.8570337453131509,
+      "grad_norm": 3.745986984152243,
+      "learning_rate": 5.265595387159705e-07,
+      "loss": 0.2808,
+      "step": 12343
+    },
+    {
+      "epoch": 0.8571031801138731,
+      "grad_norm": 3.3326795133892304,
+      "learning_rate": 5.260573530255736e-07,
+      "loss": 0.259,
+      "step": 12344
+    },
+    {
+      "epoch": 0.8571726149145952,
+      "grad_norm": 3.431093524695176,
+      "learning_rate": 5.255553936221331e-07,
+      "loss": 0.2873,
+      "step": 12345
+    },
+    {
+      "epoch": 0.8572420497153174,
+      "grad_norm": 4.132601325597009,
+      "learning_rate": 5.25053660531038e-07,
+      "loss": 0.5389,
+      "step": 12346
+    },
+    {
+      "epoch": 0.8573114845160394,
+      "grad_norm": 4.376452809498118,
+      "learning_rate": 5.245521537776649e-07,
+      "loss": 0.4308,
+      "step": 12347
+    },
+    {
+      "epoch": 0.8573809193167615,
+      "grad_norm": 4.249522349769682,
+      "learning_rate": 5.24050873387379e-07,
+      "loss": 0.4767,
+      "step": 12348
+    },
+    {
+      "epoch": 0.8574503541174837,
+      "grad_norm": 3.9774814651689336,
+      "learning_rate": 5.235498193855354e-07,
+      "loss": 0.4055,
+      "step": 12349
+    },
+    {
+      "epoch": 0.8575197889182058,
+      "grad_norm": 2.921955233444921,
+      "learning_rate": 5.230489917974768e-07,
+      "loss": 0.2861,
+      "step": 12350
+    },
+    {
+      "epoch": 0.8575892237189279,
+      "grad_norm": 4.44052867020784,
+      "learning_rate": 5.22548390648533e-07,
+      "loss": 0.5184,
+      "step": 12351
+    },
+    {
+      "epoch": 0.8576586585196501,
+      "grad_norm": 4.765860968933322,
+      "learning_rate": 5.220480159640268e-07,
+      "loss": 0.4761,
+      "step": 12352
+    },
+    {
+      "epoch": 0.8577280933203721,
+      "grad_norm": 3.3747488324476898,
+      "learning_rate": 5.215478677692642e-07,
+      "loss": 0.3322,
+      "step": 12353
+    },
+    {
+      "epoch": 0.8577975281210943,
+      "grad_norm": 3.8031731152570005,
+      "learning_rate": 5.210479460895413e-07,
+      "loss": 0.3972,
+      "step": 12354
+    },
+    {
+      "epoch": 0.8578669629218164,
+      "grad_norm": 3.5934029623017376,
+      "learning_rate": 5.205482509501463e-07,
+      "loss": 0.2899,
+      "step": 12355
+    },
+    {
+      "epoch": 0.8579363977225385,
+      "grad_norm": 3.177147790060027,
+      "learning_rate": 5.200487823763529e-07,
+      "loss": 0.4276,
+      "step": 12356
+    },
+    {
+      "epoch": 0.8580058325232607,
+      "grad_norm": 4.858997766589289,
+      "learning_rate": 5.195495403934209e-07,
+      "loss": 0.42,
+      "step": 12357
+    },
+    {
+      "epoch": 0.8580752673239828,
+      "grad_norm": 11.479131480367396,
+      "learning_rate": 5.190505250266036e-07,
+      "loss": 0.2795,
+      "step": 12358
+    },
+    {
+      "epoch": 0.858144702124705,
+      "grad_norm": 2.794703419707781,
+      "learning_rate": 5.185517363011416e-07,
+      "loss": 0.1971,
+      "step": 12359
+    },
+    {
+      "epoch": 0.858214136925427,
+      "grad_norm": 3.6063450472942837,
+      "learning_rate": 5.180531742422595e-07,
+      "loss": 0.4005,
+      "step": 12360
+    },
+    {
+      "epoch": 0.8582835717261491,
+      "grad_norm": 4.17125988060682,
+      "learning_rate": 5.175548388751772e-07,
+      "loss": 0.3932,
+      "step": 12361
+    },
+    {
+      "epoch": 0.8583530065268713,
+      "grad_norm": 3.7825077353753382,
+      "learning_rate": 5.17056730225099e-07,
+      "loss": 0.4931,
+      "step": 12362
+    },
+    {
+      "epoch": 0.8584224413275934,
+      "grad_norm": 3.145042493724635,
+      "learning_rate": 5.165588483172185e-07,
+      "loss": 0.3043,
+      "step": 12363
+    },
+    {
+      "epoch": 0.8584918761283155,
+      "grad_norm": 5.188653605490194,
+      "learning_rate": 5.160611931767185e-07,
+      "loss": 0.5134,
+      "step": 12364
+    },
+    {
+      "epoch": 0.8585613109290376,
+      "grad_norm": 4.10439367951557,
+      "learning_rate": 5.155637648287687e-07,
+      "loss": 0.4534,
+      "step": 12365
+    },
+    {
+      "epoch": 0.8586307457297597,
+      "grad_norm": 3.0912102861616235,
+      "learning_rate": 5.150665632985297e-07,
+      "loss": 0.243,
+      "step": 12366
+    },
+    {
+      "epoch": 0.8587001805304819,
+      "grad_norm": 3.1320756946183597,
+      "learning_rate": 5.14569588611149e-07,
+      "loss": 0.2977,
+      "step": 12367
+    },
+    {
+      "epoch": 0.858769615331204,
+      "grad_norm": 3.426097063836646,
+      "learning_rate": 5.140728407917628e-07,
+      "loss": 0.2931,
+      "step": 12368
+    },
+    {
+      "epoch": 0.8588390501319261,
+      "grad_norm": 3.020723946399208,
+      "learning_rate": 5.135763198654947e-07,
+      "loss": 0.2949,
+      "step": 12369
+    },
+    {
+      "epoch": 0.8589084849326483,
+      "grad_norm": 3.8349109421803442,
+      "learning_rate": 5.13080025857462e-07,
+      "loss": 0.3828,
+      "step": 12370
+    },
+    {
+      "epoch": 0.8589779197333703,
+      "grad_norm": 4.238156740132818,
+      "learning_rate": 5.125839587927623e-07,
+      "loss": 0.4172,
+      "step": 12371
+    },
+    {
+      "epoch": 0.8590473545340925,
+      "grad_norm": 4.0165695742907594,
+      "learning_rate": 5.120881186964888e-07,
+      "loss": 0.4794,
+      "step": 12372
+    },
+    {
+      "epoch": 0.8591167893348146,
+      "grad_norm": 2.903841450424251,
+      "learning_rate": 5.115925055937205e-07,
+      "loss": 0.2705,
+      "step": 12373
+    },
+    {
+      "epoch": 0.8591862241355367,
+      "grad_norm": 4.469243559537574,
+      "learning_rate": 5.110971195095238e-07,
+      "loss": 0.3706,
+      "step": 12374
+    },
+    {
+      "epoch": 0.8592556589362589,
+      "grad_norm": 3.0368659349288496,
+      "learning_rate": 5.10601960468955e-07,
+      "loss": 0.4166,
+      "step": 12375
+    },
+    {
+      "epoch": 0.859325093736981,
+      "grad_norm": 4.022607807565288,
+      "learning_rate": 5.101070284970594e-07,
+      "loss": 0.2505,
+      "step": 12376
+    },
+    {
+      "epoch": 0.8593945285377032,
+      "grad_norm": 3.8295013150507273,
+      "learning_rate": 5.096123236188694e-07,
+      "loss": 0.2476,
+      "step": 12377
+    },
+    {
+      "epoch": 0.8594639633384252,
+      "grad_norm": 3.2789729830279484,
+      "learning_rate": 5.09117845859407e-07,
+      "loss": 0.3142,
+      "step": 12378
+    },
+    {
+      "epoch": 0.8595333981391473,
+      "grad_norm": 3.230404566931052,
+      "learning_rate": 5.086235952436819e-07,
+      "loss": 0.2772,
+      "step": 12379
+    },
+    {
+      "epoch": 0.8596028329398695,
+      "grad_norm": 5.186344873873336,
+      "learning_rate": 5.081295717966928e-07,
+      "loss": 0.7393,
+      "step": 12380
+    },
+    {
+      "epoch": 0.8596722677405916,
+      "grad_norm": 3.8174422166804316,
+      "learning_rate": 5.076357755434281e-07,
+      "loss": 0.395,
+      "step": 12381
+    },
+    {
+      "epoch": 0.8597417025413137,
+      "grad_norm": 4.666192286727472,
+      "learning_rate": 5.071422065088621e-07,
+      "loss": 0.4968,
+      "step": 12382
+    },
+    {
+      "epoch": 0.8598111373420358,
+      "grad_norm": 3.9651775673686482,
+      "learning_rate": 5.066488647179579e-07,
+      "loss": 0.4934,
+      "step": 12383
+    },
+    {
+      "epoch": 0.8598805721427579,
+      "grad_norm": 2.9459854804671823,
+      "learning_rate": 5.061557501956715e-07,
+      "loss": 0.213,
+      "step": 12384
+    },
+    {
+      "epoch": 0.8599500069434801,
+      "grad_norm": 1.7298313904286367,
+      "learning_rate": 5.056628629669413e-07,
+      "loss": 0.0741,
+      "step": 12385
+    },
+    {
+      "epoch": 0.8600194417442022,
+      "grad_norm": 3.9243441663304703,
+      "learning_rate": 5.051702030566969e-07,
+      "loss": 0.4282,
+      "step": 12386
+    },
+    {
+      "epoch": 0.8600888765449243,
+      "grad_norm": 3.1915614812310444,
+      "learning_rate": 5.046777704898582e-07,
+      "loss": 0.2351,
+      "step": 12387
+    },
+    {
+      "epoch": 0.8601583113456465,
+      "grad_norm": 5.8815364778123165,
+      "learning_rate": 5.041855652913325e-07,
+      "loss": 0.6513,
+      "step": 12388
+    },
+    {
+      "epoch": 0.8602277461463685,
+      "grad_norm": 4.935711368278944,
+      "learning_rate": 5.036935874860111e-07,
+      "loss": 0.5023,
+      "step": 12389
+    },
+    {
+      "epoch": 0.8602971809470907,
+      "grad_norm": 6.034346714885967,
+      "learning_rate": 5.032018370987812e-07,
+      "loss": 0.47,
+      "step": 12390
+    },
+    {
+      "epoch": 0.8603666157478128,
+      "grad_norm": 3.807652106262269,
+      "learning_rate": 5.027103141545148e-07,
+      "loss": 0.2236,
+      "step": 12391
+    },
+    {
+      "epoch": 0.8604360505485349,
+      "grad_norm": 4.1081781937123525,
+      "learning_rate": 5.022190186780701e-07,
+      "loss": 0.5549,
+      "step": 12392
+    },
+    {
+      "epoch": 0.8605054853492571,
+      "grad_norm": 3.998779442757059,
+      "learning_rate": 5.017279506942985e-07,
+      "loss": 0.4686,
+      "step": 12393
+    },
+    {
+      "epoch": 0.8605749201499792,
+      "grad_norm": 4.109943023970411,
+      "learning_rate": 5.012371102280366e-07,
+      "loss": 0.5381,
+      "step": 12394
+    },
+    {
+      "epoch": 0.8606443549507012,
+      "grad_norm": 3.7218651383309678,
+      "learning_rate": 5.007464973041115e-07,
+      "loss": 0.3322,
+      "step": 12395
+    },
+    {
+      "epoch": 0.8607137897514234,
+      "grad_norm": 4.121022869251024,
+      "learning_rate": 5.002561119473376e-07,
+      "loss": 0.355,
+      "step": 12396
+    },
+    {
+      "epoch": 0.8607832245521455,
+      "grad_norm": 7.288388066759092,
+      "learning_rate": 4.997659541825173e-07,
+      "loss": 0.5127,
+      "step": 12397
+    },
+    {
+      "epoch": 0.8608526593528677,
+      "grad_norm": 4.057317307326759,
+      "learning_rate": 4.992760240344418e-07,
+      "loss": 0.3623,
+      "step": 12398
+    },
+    {
+      "epoch": 0.8609220941535898,
+      "grad_norm": 4.225378725501834,
+      "learning_rate": 4.987863215278943e-07,
+      "loss": 0.3855,
+      "step": 12399
+    },
+    {
+      "epoch": 0.8609915289543119,
+      "grad_norm": 4.187455606729623,
+      "learning_rate": 4.982968466876398e-07,
+      "loss": 0.4348,
+      "step": 12400
+    },
+    {
+      "epoch": 0.861060963755034,
+      "grad_norm": 5.83181126038767,
+      "learning_rate": 4.978075995384362e-07,
+      "loss": 0.4329,
+      "step": 12401
+    },
+    {
+      "epoch": 0.8611303985557561,
+      "grad_norm": 3.160199892085061,
+      "learning_rate": 4.973185801050312e-07,
+      "loss": 0.2364,
+      "step": 12402
+    },
+    {
+      "epoch": 0.8611998333564783,
+      "grad_norm": 2.5076371229371746,
+      "learning_rate": 4.968297884121553e-07,
+      "loss": 0.1836,
+      "step": 12403
+    },
+    {
+      "epoch": 0.8612692681572004,
+      "grad_norm": 3.4632920805048877,
+      "learning_rate": 4.963412244845345e-07,
+      "loss": 0.3068,
+      "step": 12404
+    },
+    {
+      "epoch": 0.8613387029579225,
+      "grad_norm": 5.383389480348123,
+      "learning_rate": 4.958528883468778e-07,
+      "loss": 0.4755,
+      "step": 12405
+    },
+    {
+      "epoch": 0.8614081377586447,
+      "grad_norm": 2.1833797444863663,
+      "learning_rate": 4.953647800238859e-07,
+      "loss": 0.0984,
+      "step": 12406
+    },
+    {
+      "epoch": 0.8614775725593667,
+      "grad_norm": 4.71025148140748,
+      "learning_rate": 4.948768995402453e-07,
+      "loss": 0.5459,
+      "step": 12407
+    },
+    {
+      "epoch": 0.8615470073600888,
+      "grad_norm": 3.5218711238696527,
+      "learning_rate": 4.943892469206335e-07,
+      "loss": 0.3206,
+      "step": 12408
+    },
+    {
+      "epoch": 0.861616442160811,
+      "grad_norm": 3.8864440981656543,
+      "learning_rate": 4.939018221897151e-07,
+      "loss": 0.4047,
+      "step": 12409
+    },
+    {
+      "epoch": 0.8616858769615331,
+      "grad_norm": 4.17485527182572,
+      "learning_rate": 4.934146253721439e-07,
+      "loss": 0.5595,
+      "step": 12410
+    },
+    {
+      "epoch": 0.8617553117622553,
+      "grad_norm": 3.5720838167485836,
+      "learning_rate": 4.929276564925611e-07,
+      "loss": 0.36,
+      "step": 12411
+    },
+    {
+      "epoch": 0.8618247465629774,
+      "grad_norm": 4.5706519225115425,
+      "learning_rate": 4.924409155755966e-07,
+      "loss": 0.5903,
+      "step": 12412
+    },
+    {
+      "epoch": 0.8618941813636994,
+      "grad_norm": 2.469652730633405,
+      "learning_rate": 4.919544026458717e-07,
+      "loss": 0.2693,
+      "step": 12413
+    },
+    {
+      "epoch": 0.8619636161644216,
+      "grad_norm": 4.17909519642469,
+      "learning_rate": 4.914681177279907e-07,
+      "loss": 0.3521,
+      "step": 12414
+    },
+    {
+      "epoch": 0.8620330509651437,
+      "grad_norm": 4.197170879164432,
+      "learning_rate": 4.909820608465499e-07,
+      "loss": 0.5699,
+      "step": 12415
+    },
+    {
+      "epoch": 0.8621024857658659,
+      "grad_norm": 5.222314019123935,
+      "learning_rate": 4.904962320261364e-07,
+      "loss": 0.6539,
+      "step": 12416
+    },
+    {
+      "epoch": 0.862171920566588,
+      "grad_norm": 4.030675470527228,
+      "learning_rate": 4.900106312913189e-07,
+      "loss": 0.4918,
+      "step": 12417
+    },
+    {
+      "epoch": 0.8622413553673101,
+      "grad_norm": 4.8105376801118656,
+      "learning_rate": 4.895252586666599e-07,
+      "loss": 0.7096,
+      "step": 12418
+    },
+    {
+      "epoch": 0.8623107901680322,
+      "grad_norm": 3.6265574196845813,
+      "learning_rate": 4.890401141767104e-07,
+      "loss": 0.1999,
+      "step": 12419
+    },
+    {
+      "epoch": 0.8623802249687543,
+      "grad_norm": 3.479167984834636,
+      "learning_rate": 4.885551978460079e-07,
+      "loss": 0.289,
+      "step": 12420
+    },
+    {
+      "epoch": 0.8624496597694764,
+      "grad_norm": 4.688369845052894,
+      "learning_rate": 4.880705096990767e-07,
+      "loss": 0.3519,
+      "step": 12421
+    },
+    {
+      "epoch": 0.8625190945701986,
+      "grad_norm": 6.750551182495961,
+      "learning_rate": 4.875860497604345e-07,
+      "loss": 0.5872,
+      "step": 12422
+    },
+    {
+      "epoch": 0.8625885293709207,
+      "grad_norm": 4.765882874911608,
+      "learning_rate": 4.871018180545839e-07,
+      "loss": 0.6733,
+      "step": 12423
+    },
+    {
+      "epoch": 0.8626579641716429,
+      "grad_norm": 2.9256252136539174,
+      "learning_rate": 4.866178146060163e-07,
+      "loss": 0.2462,
+      "step": 12424
+    },
+    {
+      "epoch": 0.862727398972365,
+      "grad_norm": 3.8824082516153564,
+      "learning_rate": 4.861340394392127e-07,
+      "loss": 0.3964,
+      "step": 12425
+    },
+    {
+      "epoch": 0.862796833773087,
+      "grad_norm": 4.060090453614927,
+      "learning_rate": 4.856504925786415e-07,
+      "loss": 0.3939,
+      "step": 12426
+    },
+    {
+      "epoch": 0.8628662685738092,
+      "grad_norm": 4.542620672818572,
+      "learning_rate": 4.851671740487602e-07,
+      "loss": 0.5503,
+      "step": 12427
+    },
+    {
+      "epoch": 0.8629357033745313,
+      "grad_norm": 3.7797245520819813,
+      "learning_rate": 4.846840838740147e-07,
+      "loss": 0.4229,
+      "step": 12428
+    },
+    {
+      "epoch": 0.8630051381752535,
+      "grad_norm": 3.364276192680082,
+      "learning_rate": 4.842012220788384e-07,
+      "loss": 0.3261,
+      "step": 12429
+    },
+    {
+      "epoch": 0.8630745729759756,
+      "grad_norm": 3.2884530187518424,
+      "learning_rate": 4.837185886876533e-07,
+      "loss": 0.2802,
+      "step": 12430
+    },
+    {
+      "epoch": 0.8631440077766976,
+      "grad_norm": 4.942494844380097,
+      "learning_rate": 4.832361837248734e-07,
+      "loss": 0.5706,
+      "step": 12431
+    },
+    {
+      "epoch": 0.8632134425774198,
+      "grad_norm": 4.178505893884873,
+      "learning_rate": 4.827540072148939e-07,
+      "loss": 0.3445,
+      "step": 12432
+    },
+    {
+      "epoch": 0.8632828773781419,
+      "grad_norm": 4.052459940084662,
+      "learning_rate": 4.822720591821068e-07,
+      "loss": 0.112,
+      "step": 12433
+    },
+    {
+      "epoch": 0.8633523121788641,
+      "grad_norm": 2.9573999684721346,
+      "learning_rate": 4.817903396508877e-07,
+      "loss": 0.2076,
+      "step": 12434
+    },
+    {
+      "epoch": 0.8634217469795862,
+      "grad_norm": 4.380866050295134,
+      "learning_rate": 4.81308848645598e-07,
+      "loss": 0.4278,
+      "step": 12435
+    },
+    {
+      "epoch": 0.8634911817803083,
+      "grad_norm": 3.683129217684303,
+      "learning_rate": 4.808275861905947e-07,
+      "loss": 0.3046,
+      "step": 12436
+    },
+    {
+      "epoch": 0.8635606165810304,
+      "grad_norm": 4.339321749435191,
+      "learning_rate": 4.803465523102186e-07,
+      "loss": 0.4309,
+      "step": 12437
+    },
+    {
+      "epoch": 0.8636300513817525,
+      "grad_norm": 4.318028991326213,
+      "learning_rate": 4.798657470287993e-07,
+      "loss": 0.5358,
+      "step": 12438
+    },
+    {
+      "epoch": 0.8636994861824746,
+      "grad_norm": 3.264117965413702,
+      "learning_rate": 4.793851703706559e-07,
+      "loss": 0.2319,
+      "step": 12439
+    },
+    {
+      "epoch": 0.8637689209831968,
+      "grad_norm": 3.0686639832495675,
+      "learning_rate": 4.789048223600951e-07,
+      "loss": 0.2197,
+      "step": 12440
+    },
+    {
+      "epoch": 0.8638383557839189,
+      "grad_norm": 3.7864772848224995,
+      "learning_rate": 4.784247030214123e-07,
+      "loss": 0.2281,
+      "step": 12441
+    },
+    {
+      "epoch": 0.8639077905846411,
+      "grad_norm": 3.6430810006259873,
+      "learning_rate": 4.779448123788915e-07,
+      "loss": 0.422,
+      "step": 12442
+    },
+    {
+      "epoch": 0.8639772253853631,
+      "grad_norm": 3.9524174067852686,
+      "learning_rate": 4.774651504568051e-07,
+      "loss": 0.3518,
+      "step": 12443
+    },
+    {
+      "epoch": 0.8640466601860852,
+      "grad_norm": 5.260713653314153,
+      "learning_rate": 4.769857172794129e-07,
+      "loss": 0.5136,
+      "step": 12444
+    },
+    {
+      "epoch": 0.8641160949868074,
+      "grad_norm": 4.115444790185732,
+      "learning_rate": 4.7650651287096673e-07,
+      "loss": 0.5037,
+      "step": 12445
+    },
+    {
+      "epoch": 0.8641855297875295,
+      "grad_norm": 3.6819106555640495,
+      "learning_rate": 4.760275372557016e-07,
+      "loss": 0.251,
+      "step": 12446
+    },
+    {
+      "epoch": 0.8642549645882517,
+      "grad_norm": 4.060274832509004,
+      "learning_rate": 4.7554879045784317e-07,
+      "loss": 0.2655,
+      "step": 12447
+    },
+    {
+      "epoch": 0.8643243993889738,
+      "grad_norm": 5.4386910520289975,
+      "learning_rate": 4.750702725016093e-07,
+      "loss": 0.4917,
+      "step": 12448
+    },
+    {
+      "epoch": 0.8643938341896958,
+      "grad_norm": 4.450027048685496,
+      "learning_rate": 4.7459198341120004e-07,
+      "loss": 0.502,
+      "step": 12449
+    },
+    {
+      "epoch": 0.864463268990418,
+      "grad_norm": 3.7291228170358353,
+      "learning_rate": 4.7411392321080606e-07,
+      "loss": 0.5111,
+      "step": 12450
+    },
+    {
+      "epoch": 0.8645327037911401,
+      "grad_norm": 3.4805922263223525,
+      "learning_rate": 4.736360919246097e-07,
+      "loss": 0.2967,
+      "step": 12451
+    },
+    {
+      "epoch": 0.8646021385918622,
+      "grad_norm": 4.450087662709071,
+      "learning_rate": 4.7315848957677767e-07,
+      "loss": 0.5248,
+      "step": 12452
+    },
+    {
+      "epoch": 0.8646715733925844,
+      "grad_norm": 3.6429657597617817,
+      "learning_rate": 4.726811161914668e-07,
+      "loss": 0.3782,
+      "step": 12453
+    },
+    {
+      "epoch": 0.8647410081933065,
+      "grad_norm": 5.480448534070716,
+      "learning_rate": 4.7220397179282227e-07,
+      "loss": 0.4102,
+      "step": 12454
+    },
+    {
+      "epoch": 0.8648104429940286,
+      "grad_norm": 4.137790261032268,
+      "learning_rate": 4.7172705640497695e-07,
+      "loss": 0.3362,
+      "step": 12455
+    },
+    {
+      "epoch": 0.8648798777947507,
+      "grad_norm": 3.863180907123696,
+      "learning_rate": 4.712503700520532e-07,
+      "loss": 0.4368,
+      "step": 12456
+    },
+    {
+      "epoch": 0.8649493125954728,
+      "grad_norm": 4.140162498694205,
+      "learning_rate": 4.707739127581612e-07,
+      "loss": 0.5447,
+      "step": 12457
+    },
+    {
+      "epoch": 0.865018747396195,
+      "grad_norm": 3.91639652390322,
+      "learning_rate": 4.702976845473994e-07,
+      "loss": 0.3337,
+      "step": 12458
+    },
+    {
+      "epoch": 0.8650881821969171,
+      "grad_norm": 3.356567131408803,
+      "learning_rate": 4.698216854438553e-07,
+      "loss": 0.2813,
+      "step": 12459
+    },
+    {
+      "epoch": 0.8651576169976393,
+      "grad_norm": 4.006134601892631,
+      "learning_rate": 4.69345915471604e-07,
+      "loss": 0.377,
+      "step": 12460
+    },
+    {
+      "epoch": 0.8652270517983613,
+      "grad_norm": 2.9570029558467223,
+      "learning_rate": 4.6887037465470853e-07,
+      "loss": 0.2651,
+      "step": 12461
+    },
+    {
+      "epoch": 0.8652964865990834,
+      "grad_norm": 4.920846169559993,
+      "learning_rate": 4.683950630172229e-07,
+      "loss": 0.5727,
+      "step": 12462
+    },
+    {
+      "epoch": 0.8653659213998056,
+      "grad_norm": 10.789596103442541,
+      "learning_rate": 4.67919980583188e-07,
+      "loss": 0.5007,
+      "step": 12463
+    },
+    {
+      "epoch": 0.8654353562005277,
+      "grad_norm": 3.3626453613818645,
+      "learning_rate": 4.6744512737663063e-07,
+      "loss": 0.2618,
+      "step": 12464
+    },
+    {
+      "epoch": 0.8655047910012498,
+      "grad_norm": 3.7391183458011246,
+      "learning_rate": 4.6697050342157045e-07,
+      "loss": 0.4561,
+      "step": 12465
+    },
+    {
+      "epoch": 0.865574225801972,
+      "grad_norm": 3.098009759773729,
+      "learning_rate": 4.664961087420139e-07,
+      "loss": 0.2445,
+      "step": 12466
+    },
+    {
+      "epoch": 0.865643660602694,
+      "grad_norm": 3.2282784774103237,
+      "learning_rate": 4.6602194336195173e-07,
+      "loss": 0.3593,
+      "step": 12467
+    },
+    {
+      "epoch": 0.8657130954034162,
+      "grad_norm": 3.01019199546965,
+      "learning_rate": 4.6554800730537084e-07,
+      "loss": 0.3448,
+      "step": 12468
+    },
+    {
+      "epoch": 0.8657825302041383,
+      "grad_norm": 4.356681361639288,
+      "learning_rate": 4.6507430059623994e-07,
+      "loss": 0.5009,
+      "step": 12469
+    },
+    {
+      "epoch": 0.8658519650048604,
+      "grad_norm": 5.212131992363794,
+      "learning_rate": 4.6460082325851974e-07,
+      "loss": 0.7968,
+      "step": 12470
+    },
+    {
+      "epoch": 0.8659213998055826,
+      "grad_norm": 3.6984155809395487,
+      "learning_rate": 4.641275753161578e-07,
+      "loss": 0.453,
+      "step": 12471
+    },
+    {
+      "epoch": 0.8659908346063047,
+      "grad_norm": 4.483860320415954,
+      "learning_rate": 4.6365455679309003e-07,
+      "loss": 0.6148,
+      "step": 12472
+    },
+    {
+      "epoch": 0.8660602694070269,
+      "grad_norm": 4.589579815444638,
+      "learning_rate": 4.631817677132422e-07,
+      "loss": 0.364,
+      "step": 12473
+    },
+    {
+      "epoch": 0.8661297042077489,
+      "grad_norm": 4.318332057572095,
+      "learning_rate": 4.627092081005263e-07,
+      "loss": 0.4526,
+      "step": 12474
+    },
+    {
+      "epoch": 0.866199139008471,
+      "grad_norm": 3.854282690739607,
+      "learning_rate": 4.6223687797884444e-07,
+      "loss": 0.4851,
+      "step": 12475
+    },
+    {
+      "epoch": 0.8662685738091932,
+      "grad_norm": 4.1743104714804975,
+      "learning_rate": 4.6176477737208515e-07,
+      "loss": 0.4572,
+      "step": 12476
+    },
+    {
+      "epoch": 0.8663380086099153,
+      "grad_norm": 5.014095408385739,
+      "learning_rate": 4.612929063041305e-07,
+      "loss": 0.444,
+      "step": 12477
+    },
+    {
+      "epoch": 0.8664074434106374,
+      "grad_norm": 5.770329459678691,
+      "learning_rate": 4.608212647988436e-07,
+      "loss": 0.3561,
+      "step": 12478
+    },
+    {
+      "epoch": 0.8664768782113595,
+      "grad_norm": 6.773986142408096,
+      "learning_rate": 4.6034985288007984e-07,
+      "loss": 0.4905,
+      "step": 12479
+    },
+    {
+      "epoch": 0.8665463130120816,
+      "grad_norm": 4.338688123289783,
+      "learning_rate": 4.5987867057168456e-07,
+      "loss": 0.3147,
+      "step": 12480
+    },
+    {
+      "epoch": 0.8666157478128038,
+      "grad_norm": 3.4344264612397404,
+      "learning_rate": 4.5940771789748926e-07,
+      "loss": 0.3809,
+      "step": 12481
+    },
+    {
+      "epoch": 0.8666851826135259,
+      "grad_norm": 3.6231626662898755,
+      "learning_rate": 4.5893699488131217e-07,
+      "loss": 0.2755,
+      "step": 12482
+    },
+    {
+      "epoch": 0.866754617414248,
+      "grad_norm": 2.6935962508894793,
+      "learning_rate": 4.584665015469636e-07,
+      "loss": 0.1877,
+      "step": 12483
+    },
+    {
+      "epoch": 0.8668240522149702,
+      "grad_norm": 3.2297964365348486,
+      "learning_rate": 4.579962379182407e-07,
+      "loss": 0.2753,
+      "step": 12484
+    },
+    {
+      "epoch": 0.8668934870156922,
+      "grad_norm": 4.486183041832954,
+      "learning_rate": 4.575262040189288e-07,
+      "loss": 0.334,
+      "step": 12485
+    },
+    {
+      "epoch": 0.8669629218164144,
+      "grad_norm": 4.510899188820433,
+      "learning_rate": 4.570563998728006e-07,
+      "loss": 0.5014,
+      "step": 12486
+    },
+    {
+      "epoch": 0.8670323566171365,
+      "grad_norm": 3.9863505996368884,
+      "learning_rate": 4.5658682550361933e-07,
+      "loss": 0.4206,
+      "step": 12487
+    },
+    {
+      "epoch": 0.8671017914178586,
+      "grad_norm": 3.4766609511786193,
+      "learning_rate": 4.5611748093513544e-07,
+      "loss": 0.3441,
+      "step": 12488
+    },
+    {
+      "epoch": 0.8671712262185808,
+      "grad_norm": 4.4642026444751,
+      "learning_rate": 4.5564836619108765e-07,
+      "loss": 0.5091,
+      "step": 12489
+    },
+    {
+      "epoch": 0.8672406610193029,
+      "grad_norm": 4.281646795955753,
+      "learning_rate": 4.5517948129520263e-07,
+      "loss": 0.5544,
+      "step": 12490
+    },
+    {
+      "epoch": 0.867310095820025,
+      "grad_norm": 4.850790554228908,
+      "learning_rate": 4.5471082627119746e-07,
+      "loss": 0.4631,
+      "step": 12491
+    },
+    {
+      "epoch": 0.8673795306207471,
+      "grad_norm": 4.591486419774582,
+      "learning_rate": 4.5424240114277487e-07,
+      "loss": 0.6294,
+      "step": 12492
+    },
+    {
+      "epoch": 0.8674489654214692,
+      "grad_norm": 5.197386119862946,
+      "learning_rate": 4.537742059336264e-07,
+      "loss": 0.4583,
+      "step": 12493
+    },
+    {
+      "epoch": 0.8675184002221914,
+      "grad_norm": 3.992774245045591,
+      "learning_rate": 4.53306240667436e-07,
+      "loss": 0.4455,
+      "step": 12494
+    },
+    {
+      "epoch": 0.8675878350229135,
+      "grad_norm": 6.7074798651977625,
+      "learning_rate": 4.5283850536787133e-07,
+      "loss": 0.6659,
+      "step": 12495
+    },
+    {
+      "epoch": 0.8676572698236356,
+      "grad_norm": 4.735310948675853,
+      "learning_rate": 4.5237100005858794e-07,
+      "loss": 0.6564,
+      "step": 12496
+    },
+    {
+      "epoch": 0.8677267046243577,
+      "grad_norm": 4.561215131830352,
+      "learning_rate": 4.519037247632341e-07,
+      "loss": 0.443,
+      "step": 12497
+    },
+    {
+      "epoch": 0.8677961394250798,
+      "grad_norm": 4.564225470633781,
+      "learning_rate": 4.514366795054448e-07,
+      "loss": 0.4683,
+      "step": 12498
+    },
+    {
+      "epoch": 0.867865574225802,
+      "grad_norm": 4.496684442810846,
+      "learning_rate": 4.509698643088389e-07,
+      "loss": 0.3268,
+      "step": 12499
+    },
+    {
+      "epoch": 0.8679350090265241,
+      "grad_norm": 4.352535399613586,
+      "learning_rate": 4.505032791970304e-07,
+      "loss": 0.4339,
+      "step": 12500
+    },
+    {
+      "epoch": 0.8680044438272462,
+      "grad_norm": 3.8794268647331878,
+      "learning_rate": 4.5003692419361864e-07,
+      "loss": 0.2622,
+      "step": 12501
+    },
+    {
+      "epoch": 0.8680738786279684,
+      "grad_norm": 6.103129700542274,
+      "learning_rate": 4.4957079932218985e-07,
+      "loss": 0.4832,
+      "step": 12502
+    },
+    {
+      "epoch": 0.8681433134286904,
+      "grad_norm": 6.538270262006157,
+      "learning_rate": 4.491049046063212e-07,
+      "loss": 0.5364,
+      "step": 12503
+    },
+    {
+      "epoch": 0.8682127482294126,
+      "grad_norm": 4.443892140212067,
+      "learning_rate": 4.4863924006957716e-07,
+      "loss": 0.547,
+      "step": 12504
+    },
+    {
+      "epoch": 0.8682821830301347,
+      "grad_norm": 4.089286278982273,
+      "learning_rate": 4.4817380573550894e-07,
+      "loss": 0.3442,
+      "step": 12505
+    },
+    {
+      "epoch": 0.8683516178308568,
+      "grad_norm": 3.4936444104181885,
+      "learning_rate": 4.477086016276605e-07,
+      "loss": 0.3197,
+      "step": 12506
+    },
+    {
+      "epoch": 0.868421052631579,
+      "grad_norm": 5.0186433731745,
+      "learning_rate": 4.472436277695591e-07,
+      "loss": 0.5511,
+      "step": 12507
+    },
+    {
+      "epoch": 0.8684904874323011,
+      "grad_norm": 5.080693544770023,
+      "learning_rate": 4.46778884184722e-07,
+      "loss": 0.395,
+      "step": 12508
+    },
+    {
+      "epoch": 0.8685599222330231,
+      "grad_norm": 5.247110691810856,
+      "learning_rate": 4.4631437089665886e-07,
+      "loss": 0.5719,
+      "step": 12509
+    },
+    {
+      "epoch": 0.8686293570337453,
+      "grad_norm": 4.01006728928897,
+      "learning_rate": 4.4585008792886076e-07,
+      "loss": 0.4548,
+      "step": 12510
+    },
+    {
+      "epoch": 0.8686987918344674,
+      "grad_norm": 3.8618763527587423,
+      "learning_rate": 4.4538603530481117e-07,
+      "loss": 0.3607,
+      "step": 12511
+    },
+    {
+      "epoch": 0.8687682266351896,
+      "grad_norm": 3.4525475638776015,
+      "learning_rate": 4.4492221304798243e-07,
+      "loss": 0.1727,
+      "step": 12512
+    },
+    {
+      "epoch": 0.8688376614359117,
+      "grad_norm": 3.7831203826512136,
+      "learning_rate": 4.444586211818341e-07,
+      "loss": 0.4719,
+      "step": 12513
+    },
+    {
+      "epoch": 0.8689070962366338,
+      "grad_norm": 3.653530545564179,
+      "learning_rate": 4.439952597298142e-07,
+      "loss": 0.287,
+      "step": 12514
+    },
+    {
+      "epoch": 0.868976531037356,
+      "grad_norm": 2.9658327036025947,
+      "learning_rate": 4.435321287153582e-07,
+      "loss": 0.1481,
+      "step": 12515
+    },
+    {
+      "epoch": 0.869045965838078,
+      "grad_norm": 4.244686345336976,
+      "learning_rate": 4.430692281618909e-07,
+      "loss": 0.4227,
+      "step": 12516
+    },
+    {
+      "epoch": 0.8691154006388002,
+      "grad_norm": 4.38252372517127,
+      "learning_rate": 4.4260655809282573e-07,
+      "loss": 0.4785,
+      "step": 12517
+    },
+    {
+      "epoch": 0.8691848354395223,
+      "grad_norm": 5.540574411192276,
+      "learning_rate": 4.421441185315639e-07,
+      "loss": 0.6755,
+      "step": 12518
+    },
+    {
+      "epoch": 0.8692542702402444,
+      "grad_norm": 3.627995825690956,
+      "learning_rate": 4.416819095014946e-07,
+      "loss": 0.3358,
+      "step": 12519
+    },
+    {
+      "epoch": 0.8693237050409666,
+      "grad_norm": 3.855529907227254,
+      "learning_rate": 4.412199310259968e-07,
+      "loss": 0.4421,
+      "step": 12520
+    },
+    {
+      "epoch": 0.8693931398416886,
+      "grad_norm": 3.1824778165987633,
+      "learning_rate": 4.407581831284358e-07,
+      "loss": 0.2853,
+      "step": 12521
+    },
+    {
+      "epoch": 0.8694625746424107,
+      "grad_norm": 3.5800943230357443,
+      "learning_rate": 4.402966658321661e-07,
+      "loss": 0.4394,
+      "step": 12522
+    },
+    {
+      "epoch": 0.8695320094431329,
+      "grad_norm": 4.407088957199432,
+      "learning_rate": 4.398353791605331e-07,
+      "loss": 0.4707,
+      "step": 12523
+    },
+    {
+      "epoch": 0.869601444243855,
+      "grad_norm": 3.5179232612766613,
+      "learning_rate": 4.3937432313686576e-07,
+      "loss": 0.5042,
+      "step": 12524
+    },
+    {
+      "epoch": 0.8696708790445772,
+      "grad_norm": 5.998355318549039,
+      "learning_rate": 4.3891349778448336e-07,
+      "loss": 0.3789,
+      "step": 12525
+    },
+    {
+      "epoch": 0.8697403138452993,
+      "grad_norm": 4.923394931674917,
+      "learning_rate": 4.3845290312669597e-07,
+      "loss": 0.6041,
+      "step": 12526
+    },
+    {
+      "epoch": 0.8698097486460213,
+      "grad_norm": 3.4369502722219396,
+      "learning_rate": 4.3799253918680006e-07,
+      "loss": 0.3404,
+      "step": 12527
+    },
+    {
+      "epoch": 0.8698791834467435,
+      "grad_norm": 7.322313594197815,
+      "learning_rate": 4.37532405988077e-07,
+      "loss": 0.5699,
+      "step": 12528
+    },
+    {
+      "epoch": 0.8699486182474656,
+      "grad_norm": 5.319938113101198,
+      "learning_rate": 4.370725035538037e-07,
+      "loss": 0.6454,
+      "step": 12529
+    },
+    {
+      "epoch": 0.8700180530481878,
+      "grad_norm": 3.550949985934782,
+      "learning_rate": 4.3661283190723935e-07,
+      "loss": 0.2547,
+      "step": 12530
+    },
+    {
+      "epoch": 0.8700874878489099,
+      "grad_norm": 2.854765564841774,
+      "learning_rate": 4.3615339107163423e-07,
+      "loss": 0.3221,
+      "step": 12531
+    },
+    {
+      "epoch": 0.870156922649632,
+      "grad_norm": 4.307725314920983,
+      "learning_rate": 4.356941810702259e-07,
+      "loss": 0.5485,
+      "step": 12532
+    },
+    {
+      "epoch": 0.8702263574503541,
+      "grad_norm": 4.149558358556161,
+      "learning_rate": 4.352352019262418e-07,
+      "loss": 0.4346,
+      "step": 12533
+    },
+    {
+      "epoch": 0.8702957922510762,
+      "grad_norm": 4.11456254311931,
+      "learning_rate": 4.347764536628951e-07,
+      "loss": 0.5049,
+      "step": 12534
+    },
+    {
+      "epoch": 0.8703652270517983,
+      "grad_norm": 3.804645696309221,
+      "learning_rate": 4.3431793630339e-07,
+      "loss": 0.3573,
+      "step": 12535
+    },
+    {
+      "epoch": 0.8704346618525205,
+      "grad_norm": 3.5007707140653612,
+      "learning_rate": 4.338596498709169e-07,
+      "loss": 0.2918,
+      "step": 12536
+    },
+    {
+      "epoch": 0.8705040966532426,
+      "grad_norm": 3.7795506218693524,
+      "learning_rate": 4.3340159438865494e-07,
+      "loss": 0.2067,
+      "step": 12537
+    },
+    {
+      "epoch": 0.8705735314539648,
+      "grad_norm": 5.124772632332957,
+      "learning_rate": 4.329437698797745e-07,
+      "loss": 0.523,
+      "step": 12538
+    },
+    {
+      "epoch": 0.8706429662546868,
+      "grad_norm": 3.816774844654177,
+      "learning_rate": 4.3248617636742985e-07,
+      "loss": 0.4392,
+      "step": 12539
+    },
+    {
+      "epoch": 0.8707124010554089,
+      "grad_norm": 2.265441093974705,
+      "learning_rate": 4.320288138747647e-07,
+      "loss": 0.2182,
+      "step": 12540
+    },
+    {
+      "epoch": 0.8707818358561311,
+      "grad_norm": 3.584893490289672,
+      "learning_rate": 4.31571682424915e-07,
+      "loss": 0.3536,
+      "step": 12541
+    },
+    {
+      "epoch": 0.8708512706568532,
+      "grad_norm": 3.6226738848571096,
+      "learning_rate": 4.3111478204099775e-07,
+      "loss": 0.4136,
+      "step": 12542
+    },
+    {
+      "epoch": 0.8709207054575754,
+      "grad_norm": 3.9325129333898783,
+      "learning_rate": 4.3065811274612623e-07,
+      "loss": 0.3918,
+      "step": 12543
+    },
+    {
+      "epoch": 0.8709901402582975,
+      "grad_norm": 3.967945241782063,
+      "learning_rate": 4.302016745633969e-07,
+      "loss": 0.468,
+      "step": 12544
+    },
+    {
+      "epoch": 0.8710595750590195,
+      "grad_norm": 5.232783261925022,
+      "learning_rate": 4.2974546751589575e-07,
+      "loss": 0.3449,
+      "step": 12545
+    },
+    {
+      "epoch": 0.8711290098597417,
+      "grad_norm": 3.931726964374569,
+      "learning_rate": 4.292894916266971e-07,
+      "loss": 0.424,
+      "step": 12546
+    },
+    {
+      "epoch": 0.8711984446604638,
+      "grad_norm": 4.782269357824777,
+      "learning_rate": 4.288337469188636e-07,
+      "loss": 0.4208,
+      "step": 12547
+    },
+    {
+      "epoch": 0.871267879461186,
+      "grad_norm": 2.8372192078647345,
+      "learning_rate": 4.283782334154468e-07,
+      "loss": 0.173,
+      "step": 12548
+    },
+    {
+      "epoch": 0.8713373142619081,
+      "grad_norm": 3.8865195951874396,
+      "learning_rate": 4.279229511394861e-07,
+      "loss": 0.3726,
+      "step": 12549
+    },
+    {
+      "epoch": 0.8714067490626302,
+      "grad_norm": 3.8927440489906497,
+      "learning_rate": 4.2746790011400863e-07,
+      "loss": 0.3687,
+      "step": 12550
+    },
+    {
+      "epoch": 0.8714761838633523,
+      "grad_norm": 4.110865473141025,
+      "learning_rate": 4.2701308036203036e-07,
+      "loss": 0.3274,
+      "step": 12551
+    },
+    {
+      "epoch": 0.8715456186640744,
+      "grad_norm": 4.206186029917439,
+      "learning_rate": 4.2655849190655576e-07,
+      "loss": 0.5328,
+      "step": 12552
+    },
+    {
+      "epoch": 0.8716150534647965,
+      "grad_norm": 5.259134229543114,
+      "learning_rate": 4.26104134770578e-07,
+      "loss": 0.478,
+      "step": 12553
+    },
+    {
+      "epoch": 0.8716844882655187,
+      "grad_norm": 4.504765983175128,
+      "learning_rate": 4.25650008977076e-07,
+      "loss": 0.3844,
+      "step": 12554
+    },
+    {
+      "epoch": 0.8717539230662408,
+      "grad_norm": 4.404772942720361,
+      "learning_rate": 4.251961145490213e-07,
+      "loss": 0.5807,
+      "step": 12555
+    },
+    {
+      "epoch": 0.871823357866963,
+      "grad_norm": 3.7667045113396864,
+      "learning_rate": 4.2474245150937176e-07,
+      "loss": 0.3746,
+      "step": 12556
+    },
+    {
+      "epoch": 0.871892792667685,
+      "grad_norm": 4.096397600370992,
+      "learning_rate": 4.242890198810695e-07,
+      "loss": 0.4211,
+      "step": 12557
+    },
+    {
+      "epoch": 0.8719622274684071,
+      "grad_norm": 3.3064893160057625,
+      "learning_rate": 4.238358196870518e-07,
+      "loss": 0.3423,
+      "step": 12558
+    },
+    {
+      "epoch": 0.8720316622691293,
+      "grad_norm": 3.9563497368911023,
+      "learning_rate": 4.233828509502408e-07,
+      "loss": 0.4513,
+      "step": 12559
+    },
+    {
+      "epoch": 0.8721010970698514,
+      "grad_norm": 3.7327506745668257,
+      "learning_rate": 4.229301136935449e-07,
+      "loss": 0.416,
+      "step": 12560
+    },
+    {
+      "epoch": 0.8721705318705736,
+      "grad_norm": 3.371553052611191,
+      "learning_rate": 4.224776079398657e-07,
+      "loss": 0.2004,
+      "step": 12561
+    },
+    {
+      "epoch": 0.8722399666712957,
+      "grad_norm": 2.671650901547622,
+      "learning_rate": 4.220253337120889e-07,
+      "loss": 0.1814,
+      "step": 12562
+    },
+    {
+      "epoch": 0.8723094014720177,
+      "grad_norm": 3.3513482429027097,
+      "learning_rate": 4.215732910330905e-07,
+      "loss": 0.4341,
+      "step": 12563
+    },
+    {
+      "epoch": 0.8723788362727399,
+      "grad_norm": 2.5346461212512774,
+      "learning_rate": 4.2112147992573403e-07,
+      "loss": 0.2473,
+      "step": 12564
+    },
+    {
+      "epoch": 0.872448271073462,
+      "grad_norm": 3.90611545733432,
+      "learning_rate": 4.206699004128717e-07,
+      "loss": 0.4764,
+      "step": 12565
+    },
+    {
+      "epoch": 0.8725177058741841,
+      "grad_norm": 3.5711973218110162,
+      "learning_rate": 4.202185525173441e-07,
+      "loss": 0.3163,
+      "step": 12566
+    },
+    {
+      "epoch": 0.8725871406749063,
+      "grad_norm": 3.1402997016783654,
+      "learning_rate": 4.197674362619797e-07,
+      "loss": 0.326,
+      "step": 12567
+    },
+    {
+      "epoch": 0.8726565754756284,
+      "grad_norm": 4.67574544630349,
+      "learning_rate": 4.1931655166959575e-07,
+      "loss": 0.4871,
+      "step": 12568
+    },
+    {
+      "epoch": 0.8727260102763506,
+      "grad_norm": 3.0708134403279392,
+      "learning_rate": 4.1886589876299567e-07,
+      "loss": 0.2945,
+      "step": 12569
+    },
+    {
+      "epoch": 0.8727954450770726,
+      "grad_norm": 4.737306363422412,
+      "learning_rate": 4.184154775649768e-07,
+      "loss": 0.5643,
+      "step": 12570
+    },
+    {
+      "epoch": 0.8728648798777947,
+      "grad_norm": 3.6394318316785785,
+      "learning_rate": 4.179652880983176e-07,
+      "loss": 0.3624,
+      "step": 12571
+    },
+    {
+      "epoch": 0.8729343146785169,
+      "grad_norm": 3.8847758302824,
+      "learning_rate": 4.175153303857887e-07,
+      "loss": 0.4923,
+      "step": 12572
+    },
+    {
+      "epoch": 0.873003749479239,
+      "grad_norm": 3.1678836473096217,
+      "learning_rate": 4.170656044501503e-07,
+      "loss": 0.2273,
+      "step": 12573
+    },
+    {
+      "epoch": 0.8730731842799612,
+      "grad_norm": 4.035057885110074,
+      "learning_rate": 4.166161103141464e-07,
+      "loss": 0.4978,
+      "step": 12574
+    },
+    {
+      "epoch": 0.8731426190806832,
+      "grad_norm": 4.236121573758128,
+      "learning_rate": 4.161668480005138e-07,
+      "loss": 0.2953,
+      "step": 12575
+    },
+    {
+      "epoch": 0.8732120538814053,
+      "grad_norm": 4.304436140713968,
+      "learning_rate": 4.157178175319748e-07,
+      "loss": 0.4418,
+      "step": 12576
+    },
+    {
+      "epoch": 0.8732814886821275,
+      "grad_norm": 4.667126598531289,
+      "learning_rate": 4.152690189312414e-07,
+      "loss": 0.5538,
+      "step": 12577
+    },
+    {
+      "epoch": 0.8733509234828496,
+      "grad_norm": 3.679895405624628,
+      "learning_rate": 4.148204522210131e-07,
+      "loss": 0.3302,
+      "step": 12578
+    },
+    {
+      "epoch": 0.8734203582835717,
+      "grad_norm": 4.320447610936786,
+      "learning_rate": 4.1437211742397734e-07,
+      "loss": 0.2611,
+      "step": 12579
+    },
+    {
+      "epoch": 0.8734897930842939,
+      "grad_norm": 5.253916669706793,
+      "learning_rate": 4.139240145628115e-07,
+      "loss": 0.5923,
+      "step": 12580
+    },
+    {
+      "epoch": 0.8735592278850159,
+      "grad_norm": 4.124862893288386,
+      "learning_rate": 4.134761436601792e-07,
+      "loss": 0.43,
+      "step": 12581
+    },
+    {
+      "epoch": 0.8736286626857381,
+      "grad_norm": 3.6500129818949674,
+      "learning_rate": 4.130285047387339e-07,
+      "loss": 0.3821,
+      "step": 12582
+    },
+    {
+      "epoch": 0.8736980974864602,
+      "grad_norm": 4.110696272907427,
+      "learning_rate": 4.1258109782111476e-07,
+      "loss": 0.4587,
+      "step": 12583
+    },
+    {
+      "epoch": 0.8737675322871823,
+      "grad_norm": 4.424237306625331,
+      "learning_rate": 4.1213392292995537e-07,
+      "loss": 0.714,
+      "step": 12584
+    },
+    {
+      "epoch": 0.8738369670879045,
+      "grad_norm": 3.7320199626994612,
+      "learning_rate": 4.1168698008786925e-07,
+      "loss": 0.5332,
+      "step": 12585
+    },
+    {
+      "epoch": 0.8739064018886266,
+      "grad_norm": 3.7254062421172622,
+      "learning_rate": 4.1124026931746275e-07,
+      "loss": 0.2751,
+      "step": 12586
+    },
+    {
+      "epoch": 0.8739758366893488,
+      "grad_norm": 4.172097300752744,
+      "learning_rate": 4.1079379064133176e-07,
+      "loss": 0.3854,
+      "step": 12587
+    },
+    {
+      "epoch": 0.8740452714900708,
+      "grad_norm": 3.6899557749230816,
+      "learning_rate": 4.103475440820587e-07,
+      "loss": 0.4824,
+      "step": 12588
+    },
+    {
+      "epoch": 0.8741147062907929,
+      "grad_norm": 4.499574103922014,
+      "learning_rate": 4.0990152966221165e-07,
+      "loss": 0.6648,
+      "step": 12589
+    },
+    {
+      "epoch": 0.8741841410915151,
+      "grad_norm": 3.4986139799483964,
+      "learning_rate": 4.0945574740435203e-07,
+      "loss": 0.3635,
+      "step": 12590
+    },
+    {
+      "epoch": 0.8742535758922372,
+      "grad_norm": 3.796033793640912,
+      "learning_rate": 4.090101973310268e-07,
+      "loss": 0.5684,
+      "step": 12591
+    },
+    {
+      "epoch": 0.8743230106929593,
+      "grad_norm": 3.670243120610566,
+      "learning_rate": 4.0856487946476895e-07,
+      "loss": 0.4573,
+      "step": 12592
+    },
+    {
+      "epoch": 0.8743924454936814,
+      "grad_norm": 3.9657178196162186,
+      "learning_rate": 4.0811979382810506e-07,
+      "loss": 0.4062,
+      "step": 12593
+    },
+    {
+      "epoch": 0.8744618802944035,
+      "grad_norm": 3.031784508853248,
+      "learning_rate": 4.0767494044354597e-07,
+      "loss": 0.3889,
+      "step": 12594
+    },
+    {
+      "epoch": 0.8745313150951257,
+      "grad_norm": 4.315318649879098,
+      "learning_rate": 4.0723031933359193e-07,
+      "loss": 0.2987,
+      "step": 12595
+    },
+    {
+      "epoch": 0.8746007498958478,
+      "grad_norm": 5.025409254513043,
+      "learning_rate": 4.0678593052073114e-07,
+      "loss": 0.543,
+      "step": 12596
+    },
+    {
+      "epoch": 0.8746701846965699,
+      "grad_norm": 2.618939459496006,
+      "learning_rate": 4.0634177402744e-07,
+      "loss": 0.2107,
+      "step": 12597
+    },
+    {
+      "epoch": 0.8747396194972921,
+      "grad_norm": 4.019225632921441,
+      "learning_rate": 4.058978498761845e-07,
+      "loss": 0.4605,
+      "step": 12598
+    },
+    {
+      "epoch": 0.8748090542980141,
+      "grad_norm": 3.0606394788415257,
+      "learning_rate": 4.054541580894167e-07,
+      "loss": 0.3346,
+      "step": 12599
+    },
+    {
+      "epoch": 0.8748784890987363,
+      "grad_norm": 4.00611315859264,
+      "learning_rate": 4.0501069868957856e-07,
+      "loss": 0.4352,
+      "step": 12600
+    },
+    {
+      "epoch": 0.8749479238994584,
+      "grad_norm": 2.396993521572427,
+      "learning_rate": 4.045674716990994e-07,
+      "loss": 0.1533,
+      "step": 12601
+    },
+    {
+      "epoch": 0.8750173587001805,
+      "grad_norm": 5.571146997033519,
+      "learning_rate": 4.041244771403985e-07,
+      "loss": 0.6474,
+      "step": 12602
+    },
+    {
+      "epoch": 0.8750867935009027,
+      "grad_norm": 4.73467024757072,
+      "learning_rate": 4.0368171503587963e-07,
+      "loss": 0.3417,
+      "step": 12603
+    },
+    {
+      "epoch": 0.8751562283016248,
+      "grad_norm": 3.7075292726850697,
+      "learning_rate": 4.032391854079393e-07,
+      "loss": 0.2482,
+      "step": 12604
+    },
+    {
+      "epoch": 0.875225663102347,
+      "grad_norm": 5.130587694141415,
+      "learning_rate": 4.0279688827895905e-07,
+      "loss": 0.5931,
+      "step": 12605
+    },
+    {
+      "epoch": 0.875295097903069,
+      "grad_norm": 3.140700657148172,
+      "learning_rate": 4.023548236713104e-07,
+      "loss": 0.3465,
+      "step": 12606
+    },
+    {
+      "epoch": 0.8753645327037911,
+      "grad_norm": 6.77453630533239,
+      "learning_rate": 4.019129916073522e-07,
+      "loss": 0.2749,
+      "step": 12607
+    },
+    {
+      "epoch": 0.8754339675045133,
+      "grad_norm": 3.9859463077890065,
+      "learning_rate": 4.0147139210943155e-07,
+      "loss": 0.3835,
+      "step": 12608
+    },
+    {
+      "epoch": 0.8755034023052354,
+      "grad_norm": 3.0789113122434175,
+      "learning_rate": 4.0103002519988445e-07,
+      "loss": 0.3287,
+      "step": 12609
+    },
+    {
+      "epoch": 0.8755728371059575,
+      "grad_norm": 3.445806902792383,
+      "learning_rate": 4.005888909010347e-07,
+      "loss": 0.3654,
+      "step": 12610
+    },
+    {
+      "epoch": 0.8756422719066796,
+      "grad_norm": 5.110273170312192,
+      "learning_rate": 4.001479892351945e-07,
+      "loss": 0.5662,
+      "step": 12611
+    },
+    {
+      "epoch": 0.8757117067074017,
+      "grad_norm": 3.384073851585577,
+      "learning_rate": 3.997073202246626e-07,
+      "loss": 0.3497,
+      "step": 12612
+    },
+    {
+      "epoch": 0.8757811415081239,
+      "grad_norm": 6.105989910662532,
+      "learning_rate": 3.9926688389173074e-07,
+      "loss": 0.7357,
+      "step": 12613
+    },
+    {
+      "epoch": 0.875850576308846,
+      "grad_norm": 3.676308334668443,
+      "learning_rate": 3.9882668025867264e-07,
+      "loss": 0.4337,
+      "step": 12614
+    },
+    {
+      "epoch": 0.8759200111095681,
+      "grad_norm": 4.074124533577655,
+      "learning_rate": 3.983867093477539e-07,
+      "loss": 0.4398,
+      "step": 12615
+    },
+    {
+      "epoch": 0.8759894459102903,
+      "grad_norm": 4.239186644871816,
+      "learning_rate": 3.9794697118123007e-07,
+      "loss": 0.4549,
+      "step": 12616
+    },
+    {
+      "epoch": 0.8760588807110123,
+      "grad_norm": 5.016385487611267,
+      "learning_rate": 3.975074657813405e-07,
+      "loss": 0.3726,
+      "step": 12617
+    },
+    {
+      "epoch": 0.8761283155117345,
+      "grad_norm": 3.3678270290961336,
+      "learning_rate": 3.9706819317031355e-07,
+      "loss": 0.3761,
+      "step": 12618
+    },
+    {
+      "epoch": 0.8761977503124566,
+      "grad_norm": 3.6639048286158786,
+      "learning_rate": 3.966291533703698e-07,
+      "loss": 0.3069,
+      "step": 12619
+    },
+    {
+      "epoch": 0.8762671851131787,
+      "grad_norm": 4.195334065068262,
+      "learning_rate": 3.961903464037159e-07,
+      "loss": 0.3566,
+      "step": 12620
+    },
+    {
+      "epoch": 0.8763366199139009,
+      "grad_norm": 4.5602322676139195,
+      "learning_rate": 3.9575177229254247e-07,
+      "loss": 0.4911,
+      "step": 12621
+    },
+    {
+      "epoch": 0.876406054714623,
+      "grad_norm": 4.517805898530028,
+      "learning_rate": 3.953134310590351e-07,
+      "loss": 0.4461,
+      "step": 12622
+    },
+    {
+      "epoch": 0.876475489515345,
+      "grad_norm": 2.822823593010967,
+      "learning_rate": 3.948753227253643e-07,
+      "loss": 0.3225,
+      "step": 12623
+    },
+    {
+      "epoch": 0.8765449243160672,
+      "grad_norm": 5.112648586758141,
+      "learning_rate": 3.9443744731368804e-07,
+      "loss": 0.4956,
+      "step": 12624
+    },
+    {
+      "epoch": 0.8766143591167893,
+      "grad_norm": 4.662548127363206,
+      "learning_rate": 3.939998048461546e-07,
+      "loss": 0.3771,
+      "step": 12625
+    },
+    {
+      "epoch": 0.8766837939175115,
+      "grad_norm": 8.915333391766282,
+      "learning_rate": 3.9356239534489857e-07,
+      "loss": 0.5583,
+      "step": 12626
+    },
+    {
+      "epoch": 0.8767532287182336,
+      "grad_norm": 3.9092344329258304,
+      "learning_rate": 3.931252188320445e-07,
+      "loss": 0.2941,
+      "step": 12627
+    },
+    {
+      "epoch": 0.8768226635189557,
+      "grad_norm": 4.19105871778343,
+      "learning_rate": 3.9268827532970344e-07,
+      "loss": 0.4441,
+      "step": 12628
+    },
+    {
+      "epoch": 0.8768920983196778,
+      "grad_norm": 4.6513537056612355,
+      "learning_rate": 3.922515648599762e-07,
+      "loss": 0.6096,
+      "step": 12629
+    },
+    {
+      "epoch": 0.8769615331203999,
+      "grad_norm": 4.291674837798719,
+      "learning_rate": 3.918150874449511e-07,
+      "loss": 0.5898,
+      "step": 12630
+    },
+    {
+      "epoch": 0.8770309679211221,
+      "grad_norm": 3.0278689231914866,
+      "learning_rate": 3.9137884310670396e-07,
+      "loss": 0.3127,
+      "step": 12631
+    },
+    {
+      "epoch": 0.8771004027218442,
+      "grad_norm": 4.1605490414467985,
+      "learning_rate": 3.9094283186729866e-07,
+      "loss": 0.3284,
+      "step": 12632
+    },
+    {
+      "epoch": 0.8771698375225663,
+      "grad_norm": 3.3417952172712493,
+      "learning_rate": 3.9050705374879097e-07,
+      "loss": 0.2936,
+      "step": 12633
+    },
+    {
+      "epoch": 0.8772392723232885,
+      "grad_norm": 4.786249632664648,
+      "learning_rate": 3.90071508773221e-07,
+      "loss": 0.3526,
+      "step": 12634
+    },
+    {
+      "epoch": 0.8773087071240105,
+      "grad_norm": 3.972142077091876,
+      "learning_rate": 3.896361969626161e-07,
+      "loss": 0.4675,
+      "step": 12635
+    },
+    {
+      "epoch": 0.8773781419247326,
+      "grad_norm": 3.6012966916971805,
+      "learning_rate": 3.89201118338996e-07,
+      "loss": 0.269,
+      "step": 12636
+    },
+    {
+      "epoch": 0.8774475767254548,
+      "grad_norm": 4.6629417124770605,
+      "learning_rate": 3.8876627292436577e-07,
+      "loss": 0.5505,
+      "step": 12637
+    },
+    {
+      "epoch": 0.8775170115261769,
+      "grad_norm": 3.6903982074673003,
+      "learning_rate": 3.8833166074072007e-07,
+      "loss": 0.3713,
+      "step": 12638
+    },
+    {
+      "epoch": 0.8775864463268991,
+      "grad_norm": 4.6796801634560135,
+      "learning_rate": 3.8789728181004017e-07,
+      "loss": 0.4503,
+      "step": 12639
+    },
+    {
+      "epoch": 0.8776558811276212,
+      "grad_norm": 2.8912570578768984,
+      "learning_rate": 3.874631361542963e-07,
+      "loss": 0.3085,
+      "step": 12640
+    },
+    {
+      "epoch": 0.8777253159283432,
+      "grad_norm": 4.166656775077804,
+      "learning_rate": 3.8702922379544814e-07,
+      "loss": 0.4316,
+      "step": 12641
+    },
+    {
+      "epoch": 0.8777947507290654,
+      "grad_norm": 3.820620341473406,
+      "learning_rate": 3.86595544755442e-07,
+      "loss": 0.3623,
+      "step": 12642
+    },
+    {
+      "epoch": 0.8778641855297875,
+      "grad_norm": 3.8015599326835776,
+      "learning_rate": 3.861620990562126e-07,
+      "loss": 0.4097,
+      "step": 12643
+    },
+    {
+      "epoch": 0.8779336203305097,
+      "grad_norm": 4.132296166827703,
+      "learning_rate": 3.857288867196823e-07,
+      "loss": 0.4902,
+      "step": 12644
+    },
+    {
+      "epoch": 0.8780030551312318,
+      "grad_norm": 3.992867200307658,
+      "learning_rate": 3.852959077677654e-07,
+      "loss": 0.4596,
+      "step": 12645
+    },
+    {
+      "epoch": 0.8780724899319539,
+      "grad_norm": 3.2711590741923997,
+      "learning_rate": 3.8486316222235864e-07,
+      "loss": 0.1807,
+      "step": 12646
+    },
+    {
+      "epoch": 0.878141924732676,
+      "grad_norm": 3.7169861304398797,
+      "learning_rate": 3.844306501053502e-07,
+      "loss": 0.2897,
+      "step": 12647
+    },
+    {
+      "epoch": 0.8782113595333981,
+      "grad_norm": 3.7936713249323404,
+      "learning_rate": 3.839983714386181e-07,
+      "loss": 0.3987,
+      "step": 12648
+    },
+    {
+      "epoch": 0.8782807943341202,
+      "grad_norm": 3.942630945232322,
+      "learning_rate": 3.835663262440242e-07,
+      "loss": 0.4574,
+      "step": 12649
+    },
+    {
+      "epoch": 0.8783502291348424,
+      "grad_norm": 4.355991869468211,
+      "learning_rate": 3.8313451454342064e-07,
+      "loss": 0.442,
+      "step": 12650
+    },
+    {
+      "epoch": 0.8784196639355645,
+      "grad_norm": 4.579467624392515,
+      "learning_rate": 3.827029363586504e-07,
+      "loss": 0.5073,
+      "step": 12651
+    },
+    {
+      "epoch": 0.8784890987362867,
+      "grad_norm": 3.2550135907896482,
+      "learning_rate": 3.82271591711541e-07,
+      "loss": 0.2771,
+      "step": 12652
+    },
+    {
+      "epoch": 0.8785585335370087,
+      "grad_norm": 2.9855263354187516,
+      "learning_rate": 3.818404806239079e-07,
+      "loss": 0.3344,
+      "step": 12653
+    },
+    {
+      "epoch": 0.8786279683377308,
+      "grad_norm": 3.04457199173619,
+      "learning_rate": 3.814096031175579e-07,
+      "loss": 0.1929,
+      "step": 12654
+    },
+    {
+      "epoch": 0.878697403138453,
+      "grad_norm": 3.4069836926640664,
+      "learning_rate": 3.809789592142843e-07,
+      "loss": 0.4072,
+      "step": 12655
+    },
+    {
+      "epoch": 0.8787668379391751,
+      "grad_norm": 3.844267439635705,
+      "learning_rate": 3.805485489358679e-07,
+      "loss": 0.4193,
+      "step": 12656
+    },
+    {
+      "epoch": 0.8788362727398973,
+      "grad_norm": 3.9221023368185057,
+      "learning_rate": 3.801183723040791e-07,
+      "loss": 0.4036,
+      "step": 12657
+    },
+    {
+      "epoch": 0.8789057075406194,
+      "grad_norm": 4.426114254252452,
+      "learning_rate": 3.7968842934067496e-07,
+      "loss": 0.615,
+      "step": 12658
+    },
+    {
+      "epoch": 0.8789751423413414,
+      "grad_norm": 3.4735238091240794,
+      "learning_rate": 3.7925872006740136e-07,
+      "loss": 0.2957,
+      "step": 12659
+    },
+    {
+      "epoch": 0.8790445771420636,
+      "grad_norm": 3.6216257281109825,
+      "learning_rate": 3.788292445059938e-07,
+      "loss": 0.6276,
+      "step": 12660
+    },
+    {
+      "epoch": 0.8791140119427857,
+      "grad_norm": 4.6578165274889605,
+      "learning_rate": 3.784000026781731e-07,
+      "loss": 0.4337,
+      "step": 12661
+    },
+    {
+      "epoch": 0.8791834467435079,
+      "grad_norm": 9.18987774923812,
+      "learning_rate": 3.7797099460565036e-07,
+      "loss": 0.8011,
+      "step": 12662
+    },
+    {
+      "epoch": 0.87925288154423,
+      "grad_norm": 3.3000471517590375,
+      "learning_rate": 3.775422203101259e-07,
+      "loss": 0.1853,
+      "step": 12663
+    },
+    {
+      "epoch": 0.8793223163449521,
+      "grad_norm": 3.772326122213835,
+      "learning_rate": 3.771136798132835e-07,
+      "loss": 0.267,
+      "step": 12664
+    },
+    {
+      "epoch": 0.8793917511456742,
+      "grad_norm": 3.3859419195871685,
+      "learning_rate": 3.7668537313680087e-07,
+      "loss": 0.3804,
+      "step": 12665
+    },
+    {
+      "epoch": 0.8794611859463963,
+      "grad_norm": 4.685378955076384,
+      "learning_rate": 3.7625730030234175e-07,
+      "loss": 0.5829,
+      "step": 12666
+    },
+    {
+      "epoch": 0.8795306207471184,
+      "grad_norm": 4.708020617201583,
+      "learning_rate": 3.758294613315544e-07,
+      "loss": 0.5483,
+      "step": 12667
+    },
+    {
+      "epoch": 0.8796000555478406,
+      "grad_norm": 4.190510249854428,
+      "learning_rate": 3.7540185624608096e-07,
+      "loss": 0.3454,
+      "step": 12668
+    },
+    {
+      "epoch": 0.8796694903485627,
+      "grad_norm": 4.108965470514319,
+      "learning_rate": 3.7497448506754854e-07,
+      "loss": 0.3344,
+      "step": 12669
+    },
+    {
+      "epoch": 0.8797389251492849,
+      "grad_norm": 4.996142393352747,
+      "learning_rate": 3.7454734781757265e-07,
+      "loss": 0.4025,
+      "step": 12670
+    },
+    {
+      "epoch": 0.879808359950007,
+      "grad_norm": 3.841471433459928,
+      "learning_rate": 3.741204445177588e-07,
+      "loss": 0.3632,
+      "step": 12671
+    },
+    {
+      "epoch": 0.879877794750729,
+      "grad_norm": 4.999474765744073,
+      "learning_rate": 3.736937751896974e-07,
+      "loss": 0.6572,
+      "step": 12672
+    },
+    {
+      "epoch": 0.8799472295514512,
+      "grad_norm": 3.5004973759737226,
+      "learning_rate": 3.732673398549702e-07,
+      "loss": 0.3181,
+      "step": 12673
+    },
+    {
+      "epoch": 0.8800166643521733,
+      "grad_norm": 2.841794115592065,
+      "learning_rate": 3.7284113853514604e-07,
+      "loss": 0.2392,
+      "step": 12674
+    },
+    {
+      "epoch": 0.8800860991528955,
+      "grad_norm": 3.3162983773781294,
+      "learning_rate": 3.724151712517804e-07,
+      "loss": 0.1652,
+      "step": 12675
+    },
+    {
+      "epoch": 0.8801555339536176,
+      "grad_norm": 4.210404759501814,
+      "learning_rate": 3.719894380264183e-07,
+      "loss": 0.4902,
+      "step": 12676
+    },
+    {
+      "epoch": 0.8802249687543396,
+      "grad_norm": 3.859281851974725,
+      "learning_rate": 3.7156393888059527e-07,
+      "loss": 0.3392,
+      "step": 12677
+    },
+    {
+      "epoch": 0.8802944035550618,
+      "grad_norm": 4.835021900836062,
+      "learning_rate": 3.711386738358297e-07,
+      "loss": 0.5415,
+      "step": 12678
+    },
+    {
+      "epoch": 0.8803638383557839,
+      "grad_norm": 4.134633742220323,
+      "learning_rate": 3.70713642913631e-07,
+      "loss": 0.3743,
+      "step": 12679
+    },
+    {
+      "epoch": 0.880433273156506,
+      "grad_norm": 3.576183698128719,
+      "learning_rate": 3.702888461355003e-07,
+      "loss": 0.4207,
+      "step": 12680
+    },
+    {
+      "epoch": 0.8805027079572282,
+      "grad_norm": 4.398132638846183,
+      "learning_rate": 3.6986428352291937e-07,
+      "loss": 0.6311,
+      "step": 12681
+    },
+    {
+      "epoch": 0.8805721427579503,
+      "grad_norm": 3.9058294176657196,
+      "learning_rate": 3.694399550973632e-07,
+      "loss": 0.3683,
+      "step": 12682
+    },
+    {
+      "epoch": 0.8806415775586725,
+      "grad_norm": 3.8231625104467635,
+      "learning_rate": 3.6901586088029516e-07,
+      "loss": 0.484,
+      "step": 12683
+    },
+    {
+      "epoch": 0.8807110123593945,
+      "grad_norm": 3.8898835086243873,
+      "learning_rate": 3.6859200089316426e-07,
+      "loss": 0.3479,
+      "step": 12684
+    },
+    {
+      "epoch": 0.8807804471601166,
+      "grad_norm": 3.9659673186307276,
+      "learning_rate": 3.6816837515740946e-07,
+      "loss": 0.4872,
+      "step": 12685
+    },
+    {
+      "epoch": 0.8808498819608388,
+      "grad_norm": 4.158553391528972,
+      "learning_rate": 3.677449836944563e-07,
+      "loss": 0.3587,
+      "step": 12686
+    },
+    {
+      "epoch": 0.8809193167615609,
+      "grad_norm": 3.7778478260029256,
+      "learning_rate": 3.6732182652572057e-07,
+      "loss": 0.2871,
+      "step": 12687
+    },
+    {
+      "epoch": 0.8809887515622831,
+      "grad_norm": 3.1938660342687517,
+      "learning_rate": 3.6689890367260507e-07,
+      "loss": 0.2493,
+      "step": 12688
+    },
+    {
+      "epoch": 0.8810581863630051,
+      "grad_norm": 3.7941734408743226,
+      "learning_rate": 3.664762151564999e-07,
+      "loss": 0.4458,
+      "step": 12689
+    },
+    {
+      "epoch": 0.8811276211637272,
+      "grad_norm": 4.098733049693217,
+      "learning_rate": 3.660537609987846e-07,
+      "loss": 0.4641,
+      "step": 12690
+    },
+    {
+      "epoch": 0.8811970559644494,
+      "grad_norm": 4.028860519529612,
+      "learning_rate": 3.656315412208261e-07,
+      "loss": 0.4478,
+      "step": 12691
+    },
+    {
+      "epoch": 0.8812664907651715,
+      "grad_norm": 2.918185055922554,
+      "learning_rate": 3.6520955584398056e-07,
+      "loss": 0.2254,
+      "step": 12692
+    },
+    {
+      "epoch": 0.8813359255658936,
+      "grad_norm": 3.34426464416976,
+      "learning_rate": 3.6478780488958974e-07,
+      "loss": 0.3035,
+      "step": 12693
+    },
+    {
+      "epoch": 0.8814053603666158,
+      "grad_norm": 4.429481927535252,
+      "learning_rate": 3.643662883789878e-07,
+      "loss": 0.5489,
+      "step": 12694
+    },
+    {
+      "epoch": 0.8814747951673378,
+      "grad_norm": 3.8273504833703362,
+      "learning_rate": 3.6394500633349425e-07,
+      "loss": 0.3708,
+      "step": 12695
+    },
+    {
+      "epoch": 0.88154422996806,
+      "grad_norm": 4.352768491029765,
+      "learning_rate": 3.6352395877441493e-07,
+      "loss": 0.4468,
+      "step": 12696
+    },
+    {
+      "epoch": 0.8816136647687821,
+      "grad_norm": 3.97796817359362,
+      "learning_rate": 3.6310314572304775e-07,
+      "loss": 0.5135,
+      "step": 12697
+    },
+    {
+      "epoch": 0.8816830995695042,
+      "grad_norm": 4.544509006029131,
+      "learning_rate": 3.626825672006773e-07,
+      "loss": 0.354,
+      "step": 12698
+    },
+    {
+      "epoch": 0.8817525343702264,
+      "grad_norm": 3.738559685056856,
+      "learning_rate": 3.622622232285733e-07,
+      "loss": 0.3289,
+      "step": 12699
+    },
+    {
+      "epoch": 0.8818219691709485,
+      "grad_norm": 3.623546972480734,
+      "learning_rate": 3.6184211382799875e-07,
+      "loss": 0.3173,
+      "step": 12700
+    },
+    {
+      "epoch": 0.8818914039716707,
+      "grad_norm": 2.6672287336018643,
+      "learning_rate": 3.6142223902020214e-07,
+      "loss": 0.2758,
+      "step": 12701
+    },
+    {
+      "epoch": 0.8819608387723927,
+      "grad_norm": 3.8079443357632723,
+      "learning_rate": 3.610025988264193e-07,
+      "loss": 0.4385,
+      "step": 12702
+    },
+    {
+      "epoch": 0.8820302735731148,
+      "grad_norm": 3.580002911235336,
+      "learning_rate": 3.6058319326787603e-07,
+      "loss": 0.1842,
+      "step": 12703
+    },
+    {
+      "epoch": 0.882099708373837,
+      "grad_norm": 4.889151974094751,
+      "learning_rate": 3.601640223657849e-07,
+      "loss": 0.5459,
+      "step": 12704
+    },
+    {
+      "epoch": 0.8821691431745591,
+      "grad_norm": 3.111699489361301,
+      "learning_rate": 3.597450861413471e-07,
+      "loss": 0.2637,
+      "step": 12705
+    },
+    {
+      "epoch": 0.8822385779752812,
+      "grad_norm": 3.9026267811988604,
+      "learning_rate": 3.5932638461575244e-07,
+      "loss": 0.4011,
+      "step": 12706
+    },
+    {
+      "epoch": 0.8823080127760033,
+      "grad_norm": 3.0997107176118854,
+      "learning_rate": 3.589079178101773e-07,
+      "loss": 0.2146,
+      "step": 12707
+    },
+    {
+      "epoch": 0.8823774475767254,
+      "grad_norm": 4.415997193583695,
+      "learning_rate": 3.584896857457876e-07,
+      "loss": 0.5051,
+      "step": 12708
+    },
+    {
+      "epoch": 0.8824468823774476,
+      "grad_norm": 3.8402917681412188,
+      "learning_rate": 3.580716884437385e-07,
+      "loss": 0.3979,
+      "step": 12709
+    },
+    {
+      "epoch": 0.8825163171781697,
+      "grad_norm": 4.801482224365796,
+      "learning_rate": 3.5765392592517044e-07,
+      "loss": 0.4159,
+      "step": 12710
+    },
+    {
+      "epoch": 0.8825857519788918,
+      "grad_norm": 4.442724235905857,
+      "learning_rate": 3.572363982112126e-07,
+      "loss": 0.6103,
+      "step": 12711
+    },
+    {
+      "epoch": 0.882655186779614,
+      "grad_norm": 5.22368719083455,
+      "learning_rate": 3.568191053229847e-07,
+      "loss": 0.4527,
+      "step": 12712
+    },
+    {
+      "epoch": 0.882724621580336,
+      "grad_norm": 3.9302027455103805,
+      "learning_rate": 3.5640204728159266e-07,
+      "loss": 0.4312,
+      "step": 12713
+    },
+    {
+      "epoch": 0.8827940563810582,
+      "grad_norm": 3.972405226974514,
+      "learning_rate": 3.559852241081302e-07,
+      "loss": 0.4222,
+      "step": 12714
+    },
+    {
+      "epoch": 0.8828634911817803,
+      "grad_norm": 4.917674485548975,
+      "learning_rate": 3.555686358236804e-07,
+      "loss": 0.5951,
+      "step": 12715
+    },
+    {
+      "epoch": 0.8829329259825024,
+      "grad_norm": 4.029427733456624,
+      "learning_rate": 3.551522824493131e-07,
+      "loss": 0.357,
+      "step": 12716
+    },
+    {
+      "epoch": 0.8830023607832246,
+      "grad_norm": 2.7753817190898076,
+      "learning_rate": 3.547361640060876e-07,
+      "loss": 0.2712,
+      "step": 12717
+    },
+    {
+      "epoch": 0.8830717955839467,
+      "grad_norm": 4.633762114409212,
+      "learning_rate": 3.543202805150503e-07,
+      "loss": 0.6012,
+      "step": 12718
+    },
+    {
+      "epoch": 0.8831412303846687,
+      "grad_norm": 4.964980065719515,
+      "learning_rate": 3.539046319972367e-07,
+      "loss": 0.4519,
+      "step": 12719
+    },
+    {
+      "epoch": 0.8832106651853909,
+      "grad_norm": 4.464788278461588,
+      "learning_rate": 3.534892184736688e-07,
+      "loss": 0.3419,
+      "step": 12720
+    },
+    {
+      "epoch": 0.883280099986113,
+      "grad_norm": 3.8196769046504486,
+      "learning_rate": 3.530740399653587e-07,
+      "loss": 0.3093,
+      "step": 12721
+    },
+    {
+      "epoch": 0.8833495347868352,
+      "grad_norm": 4.199211137671611,
+      "learning_rate": 3.526590964933052e-07,
+      "loss": 0.5223,
+      "step": 12722
+    },
+    {
+      "epoch": 0.8834189695875573,
+      "grad_norm": 3.4651955952471485,
+      "learning_rate": 3.5224438807849594e-07,
+      "loss": 0.2419,
+      "step": 12723
+    },
+    {
+      "epoch": 0.8834884043882794,
+      "grad_norm": 3.988985114580155,
+      "learning_rate": 3.5182991474190573e-07,
+      "loss": 0.8336,
+      "step": 12724
+    },
+    {
+      "epoch": 0.8835578391890015,
+      "grad_norm": 3.0199584854856383,
+      "learning_rate": 3.514156765044985e-07,
+      "loss": 0.3336,
+      "step": 12725
+    },
+    {
+      "epoch": 0.8836272739897236,
+      "grad_norm": 3.4218371562388743,
+      "learning_rate": 3.510016733872268e-07,
+      "loss": 0.2836,
+      "step": 12726
+    },
+    {
+      "epoch": 0.8836967087904458,
+      "grad_norm": 4.95277231711674,
+      "learning_rate": 3.5058790541103003e-07,
+      "loss": 0.5291,
+      "step": 12727
+    },
+    {
+      "epoch": 0.8837661435911679,
+      "grad_norm": 3.9885738526162293,
+      "learning_rate": 3.5017437259683487e-07,
+      "loss": 0.2873,
+      "step": 12728
+    },
+    {
+      "epoch": 0.88383557839189,
+      "grad_norm": 3.976576150825886,
+      "learning_rate": 3.497610749655589e-07,
+      "loss": 0.5622,
+      "step": 12729
+    },
+    {
+      "epoch": 0.8839050131926122,
+      "grad_norm": 4.7203247654926495,
+      "learning_rate": 3.4934801253810656e-07,
+      "loss": 0.4644,
+      "step": 12730
+    },
+    {
+      "epoch": 0.8839744479933342,
+      "grad_norm": 5.587966168831514,
+      "learning_rate": 3.489351853353673e-07,
+      "loss": 0.6989,
+      "step": 12731
+    },
+    {
+      "epoch": 0.8840438827940564,
+      "grad_norm": 4.387133364631243,
+      "learning_rate": 3.485225933782244e-07,
+      "loss": 0.3713,
+      "step": 12732
+    },
+    {
+      "epoch": 0.8841133175947785,
+      "grad_norm": 3.789527249945736,
+      "learning_rate": 3.48110236687545e-07,
+      "loss": 0.3252,
+      "step": 12733
+    },
+    {
+      "epoch": 0.8841827523955006,
+      "grad_norm": 4.1147341436474125,
+      "learning_rate": 3.476981152841857e-07,
+      "loss": 0.5674,
+      "step": 12734
+    },
+    {
+      "epoch": 0.8842521871962228,
+      "grad_norm": 3.545641475231648,
+      "learning_rate": 3.472862291889917e-07,
+      "loss": 0.3125,
+      "step": 12735
+    },
+    {
+      "epoch": 0.8843216219969449,
+      "grad_norm": 3.609643425963449,
+      "learning_rate": 3.4687457842279503e-07,
+      "loss": 0.5486,
+      "step": 12736
+    },
+    {
+      "epoch": 0.8843910567976669,
+      "grad_norm": 3.453174685742406,
+      "learning_rate": 3.4646316300641636e-07,
+      "loss": 0.3823,
+      "step": 12737
+    },
+    {
+      "epoch": 0.8844604915983891,
+      "grad_norm": 3.5744564382108304,
+      "learning_rate": 3.460519829606662e-07,
+      "loss": 0.4,
+      "step": 12738
+    },
+    {
+      "epoch": 0.8845299263991112,
+      "grad_norm": 3.981195410336905,
+      "learning_rate": 3.4564103830634024e-07,
+      "loss": 0.487,
+      "step": 12739
+    },
+    {
+      "epoch": 0.8845993611998334,
+      "grad_norm": 5.900618063011281,
+      "learning_rate": 3.4523032906422225e-07,
+      "loss": 0.435,
+      "step": 12740
+    },
+    {
+      "epoch": 0.8846687960005555,
+      "grad_norm": 4.198322082859053,
+      "learning_rate": 3.448198552550891e-07,
+      "loss": 0.5029,
+      "step": 12741
+    },
+    {
+      "epoch": 0.8847382308012776,
+      "grad_norm": 3.4530794755925327,
+      "learning_rate": 3.444096168996991e-07,
+      "loss": 0.4522,
+      "step": 12742
+    },
+    {
+      "epoch": 0.8848076656019997,
+      "grad_norm": 5.243828308957343,
+      "learning_rate": 3.439996140188018e-07,
+      "loss": 0.408,
+      "step": 12743
+    },
+    {
+      "epoch": 0.8848771004027218,
+      "grad_norm": 4.758751932550911,
+      "learning_rate": 3.435898466331361e-07,
+      "loss": 0.5295,
+      "step": 12744
+    },
+    {
+      "epoch": 0.884946535203444,
+      "grad_norm": 3.9425006777772187,
+      "learning_rate": 3.4318031476342715e-07,
+      "loss": 0.4525,
+      "step": 12745
+    },
+    {
+      "epoch": 0.8850159700041661,
+      "grad_norm": 2.9813214304713913,
+      "learning_rate": 3.4277101843038785e-07,
+      "loss": 0.2667,
+      "step": 12746
+    },
+    {
+      "epoch": 0.8850854048048882,
+      "grad_norm": 3.1216116126948,
+      "learning_rate": 3.423619576547205e-07,
+      "loss": 0.2429,
+      "step": 12747
+    },
+    {
+      "epoch": 0.8851548396056104,
+      "grad_norm": 4.994077530071743,
+      "learning_rate": 3.4195313245711514e-07,
+      "loss": 0.4746,
+      "step": 12748
+    },
+    {
+      "epoch": 0.8852242744063324,
+      "grad_norm": 3.7331412703198437,
+      "learning_rate": 3.4154454285824924e-07,
+      "loss": 0.398,
+      "step": 12749
+    },
+    {
+      "epoch": 0.8852937092070545,
+      "grad_norm": 3.5133878355916925,
+      "learning_rate": 3.4113618887878894e-07,
+      "loss": 0.2985,
+      "step": 12750
+    },
+    {
+      "epoch": 0.8853631440077767,
+      "grad_norm": 4.408846392387566,
+      "learning_rate": 3.407280705393884e-07,
+      "loss": 0.4437,
+      "step": 12751
+    },
+    {
+      "epoch": 0.8854325788084988,
+      "grad_norm": 6.394271493485068,
+      "learning_rate": 3.403201878606899e-07,
+      "loss": 0.5735,
+      "step": 12752
+    },
+    {
+      "epoch": 0.885502013609221,
+      "grad_norm": 4.424049907089565,
+      "learning_rate": 3.39912540863323e-07,
+      "loss": 0.6709,
+      "step": 12753
+    },
+    {
+      "epoch": 0.8855714484099431,
+      "grad_norm": 4.5274157315501355,
+      "learning_rate": 3.395051295679064e-07,
+      "loss": 0.3046,
+      "step": 12754
+    },
+    {
+      "epoch": 0.8856408832106651,
+      "grad_norm": 4.853856296491952,
+      "learning_rate": 3.390979539950479e-07,
+      "loss": 0.3848,
+      "step": 12755
+    },
+    {
+      "epoch": 0.8857103180113873,
+      "grad_norm": 5.593923492862155,
+      "learning_rate": 3.386910141653399e-07,
+      "loss": 0.5193,
+      "step": 12756
+    },
+    {
+      "epoch": 0.8857797528121094,
+      "grad_norm": 4.565479140200587,
+      "learning_rate": 3.382843100993649e-07,
+      "loss": 0.4083,
+      "step": 12757
+    },
+    {
+      "epoch": 0.8858491876128316,
+      "grad_norm": 3.968843219780575,
+      "learning_rate": 3.3787784181769533e-07,
+      "loss": 0.417,
+      "step": 12758
+    },
+    {
+      "epoch": 0.8859186224135537,
+      "grad_norm": 3.893853526762602,
+      "learning_rate": 3.3747160934088976e-07,
+      "loss": 0.3735,
+      "step": 12759
+    },
+    {
+      "epoch": 0.8859880572142758,
+      "grad_norm": 3.991488499747273,
+      "learning_rate": 3.3706561268949276e-07,
+      "loss": 0.4928,
+      "step": 12760
+    },
+    {
+      "epoch": 0.886057492014998,
+      "grad_norm": 3.4161515450748974,
+      "learning_rate": 3.366598518840408e-07,
+      "loss": 0.2873,
+      "step": 12761
+    },
+    {
+      "epoch": 0.88612692681572,
+      "grad_norm": 6.454785446147054,
+      "learning_rate": 3.3625432694505746e-07,
+      "loss": 0.5785,
+      "step": 12762
+    },
+    {
+      "epoch": 0.8861963616164421,
+      "grad_norm": 4.879053916398975,
+      "learning_rate": 3.3584903789305233e-07,
+      "loss": 0.4228,
+      "step": 12763
+    },
+    {
+      "epoch": 0.8862657964171643,
+      "grad_norm": 4.351524129746283,
+      "learning_rate": 3.354439847485247e-07,
+      "loss": 0.6223,
+      "step": 12764
+    },
+    {
+      "epoch": 0.8863352312178864,
+      "grad_norm": 2.5065225290514372,
+      "learning_rate": 3.3503916753196254e-07,
+      "loss": 0.239,
+      "step": 12765
+    },
+    {
+      "epoch": 0.8864046660186086,
+      "grad_norm": 2.456231418847352,
+      "learning_rate": 3.346345862638406e-07,
+      "loss": 0.0807,
+      "step": 12766
+    },
+    {
+      "epoch": 0.8864741008193306,
+      "grad_norm": 4.592162900792178,
+      "learning_rate": 3.34230240964622e-07,
+      "loss": 0.5394,
+      "step": 12767
+    },
+    {
+      "epoch": 0.8865435356200527,
+      "grad_norm": 4.065203923117166,
+      "learning_rate": 3.3382613165475807e-07,
+      "loss": 0.3497,
+      "step": 12768
+    },
+    {
+      "epoch": 0.8866129704207749,
+      "grad_norm": 3.186644246087373,
+      "learning_rate": 3.3342225835468754e-07,
+      "loss": 0.2981,
+      "step": 12769
+    },
+    {
+      "epoch": 0.886682405221497,
+      "grad_norm": 5.420063628362271,
+      "learning_rate": 3.3301862108483964e-07,
+      "loss": 0.5912,
+      "step": 12770
+    },
+    {
+      "epoch": 0.8867518400222192,
+      "grad_norm": 4.927465996201278,
+      "learning_rate": 3.326152198656285e-07,
+      "loss": 0.4938,
+      "step": 12771
+    },
+    {
+      "epoch": 0.8868212748229413,
+      "grad_norm": 4.005408831020627,
+      "learning_rate": 3.322120547174573e-07,
+      "loss": 0.4848,
+      "step": 12772
+    },
+    {
+      "epoch": 0.8868907096236633,
+      "grad_norm": 2.627080537802142,
+      "learning_rate": 3.3180912566071975e-07,
+      "loss": 0.1652,
+      "step": 12773
+    },
+    {
+      "epoch": 0.8869601444243855,
+      "grad_norm": 4.164698275224688,
+      "learning_rate": 3.3140643271579287e-07,
+      "loss": 0.4219,
+      "step": 12774
+    },
+    {
+      "epoch": 0.8870295792251076,
+      "grad_norm": 3.9968159405571932,
+      "learning_rate": 3.3100397590304645e-07,
+      "loss": 0.4772,
+      "step": 12775
+    },
+    {
+      "epoch": 0.8870990140258297,
+      "grad_norm": 3.241536811905565,
+      "learning_rate": 3.306017552428359e-07,
+      "loss": 0.1911,
+      "step": 12776
+    },
+    {
+      "epoch": 0.8871684488265519,
+      "grad_norm": 3.5437850709659604,
+      "learning_rate": 3.3019977075550436e-07,
+      "loss": 0.3443,
+      "step": 12777
+    },
+    {
+      "epoch": 0.887237883627274,
+      "grad_norm": 4.063452972664163,
+      "learning_rate": 3.297980224613845e-07,
+      "loss": 0.4281,
+      "step": 12778
+    },
+    {
+      "epoch": 0.8873073184279962,
+      "grad_norm": 4.362934216573782,
+      "learning_rate": 3.293965103807956e-07,
+      "loss": 0.4766,
+      "step": 12779
+    },
+    {
+      "epoch": 0.8873767532287182,
+      "grad_norm": 3.8439030057020283,
+      "learning_rate": 3.2899523453404635e-07,
+      "loss": 0.3446,
+      "step": 12780
+    },
+    {
+      "epoch": 0.8874461880294403,
+      "grad_norm": 3.01740956779014,
+      "learning_rate": 3.285941949414323e-07,
+      "loss": 0.3692,
+      "step": 12781
+    },
+    {
+      "epoch": 0.8875156228301625,
+      "grad_norm": 3.565528847331846,
+      "learning_rate": 3.2819339162323826e-07,
+      "loss": 0.3305,
+      "step": 12782
+    },
+    {
+      "epoch": 0.8875850576308846,
+      "grad_norm": 5.072536479030013,
+      "learning_rate": 3.277928245997347e-07,
+      "loss": 0.5515,
+      "step": 12783
+    },
+    {
+      "epoch": 0.8876544924316068,
+      "grad_norm": 12.128032611159652,
+      "learning_rate": 3.2739249389118487e-07,
+      "loss": 0.3927,
+      "step": 12784
+    },
+    {
+      "epoch": 0.8877239272323288,
+      "grad_norm": 4.199790191276391,
+      "learning_rate": 3.2699239951783477e-07,
+      "loss": 0.2579,
+      "step": 12785
+    },
+    {
+      "epoch": 0.8877933620330509,
+      "grad_norm": 3.401943405285184,
+      "learning_rate": 3.265925414999199e-07,
+      "loss": 0.2318,
+      "step": 12786
+    },
+    {
+      "epoch": 0.8878627968337731,
+      "grad_norm": 4.453269277090434,
+      "learning_rate": 3.261929198576669e-07,
+      "loss": 0.4031,
+      "step": 12787
+    },
+    {
+      "epoch": 0.8879322316344952,
+      "grad_norm": 3.7857765360679965,
+      "learning_rate": 3.257935346112884e-07,
+      "loss": 0.2769,
+      "step": 12788
+    },
+    {
+      "epoch": 0.8880016664352174,
+      "grad_norm": 3.980913851082153,
+      "learning_rate": 3.253943857809816e-07,
+      "loss": 0.4005,
+      "step": 12789
+    },
+    {
+      "epoch": 0.8880711012359395,
+      "grad_norm": 3.9704615029819013,
+      "learning_rate": 3.2499547338693815e-07,
+      "loss": 0.5446,
+      "step": 12790
+    },
+    {
+      "epoch": 0.8881405360366615,
+      "grad_norm": 4.580995929496314,
+      "learning_rate": 3.245967974493347e-07,
+      "loss": 0.3834,
+      "step": 12791
+    },
+    {
+      "epoch": 0.8882099708373837,
+      "grad_norm": 3.911201017527928,
+      "learning_rate": 3.241983579883329e-07,
+      "loss": 0.433,
+      "step": 12792
+    },
+    {
+      "epoch": 0.8882794056381058,
+      "grad_norm": 4.146645668179906,
+      "learning_rate": 3.2380015502408767e-07,
+      "loss": 0.4232,
+      "step": 12793
+    },
+    {
+      "epoch": 0.8883488404388279,
+      "grad_norm": 2.217702922049972,
+      "learning_rate": 3.234021885767391e-07,
+      "loss": 0.1392,
+      "step": 12794
+    },
+    {
+      "epoch": 0.8884182752395501,
+      "grad_norm": 4.267387497850099,
+      "learning_rate": 3.2300445866641603e-07,
+      "loss": 0.5891,
+      "step": 12795
+    },
+    {
+      "epoch": 0.8884877100402722,
+      "grad_norm": 4.481287548659737,
+      "learning_rate": 3.2260696531323464e-07,
+      "loss": 0.5603,
+      "step": 12796
+    },
+    {
+      "epoch": 0.8885571448409944,
+      "grad_norm": 4.8593917066126116,
+      "learning_rate": 3.222097085373005e-07,
+      "loss": 0.5428,
+      "step": 12797
+    },
+    {
+      "epoch": 0.8886265796417164,
+      "grad_norm": 3.254290156850447,
+      "learning_rate": 3.218126883587064e-07,
+      "loss": 0.3581,
+      "step": 12798
+    },
+    {
+      "epoch": 0.8886960144424385,
+      "grad_norm": 3.23905754878414,
+      "learning_rate": 3.214159047975324e-07,
+      "loss": 0.3206,
+      "step": 12799
+    },
+    {
+      "epoch": 0.8887654492431607,
+      "grad_norm": 3.4720483769098336,
+      "learning_rate": 3.210193578738474e-07,
+      "loss": 0.2885,
+      "step": 12800
+    },
+    {
+      "epoch": 0.8888348840438828,
+      "grad_norm": 4.441164111727076,
+      "learning_rate": 3.2062304760770825e-07,
+      "loss": 0.4466,
+      "step": 12801
+    },
+    {
+      "epoch": 0.888904318844605,
+      "grad_norm": 3.472163700677182,
+      "learning_rate": 3.2022697401916215e-07,
+      "loss": 0.3727,
+      "step": 12802
+    },
+    {
+      "epoch": 0.888973753645327,
+      "grad_norm": 4.150602688090228,
+      "learning_rate": 3.198311371282381e-07,
+      "loss": 0.4964,
+      "step": 12803
+    },
+    {
+      "epoch": 0.8890431884460491,
+      "grad_norm": 2.876602990657226,
+      "learning_rate": 3.194355369549601e-07,
+      "loss": 0.2766,
+      "step": 12804
+    },
+    {
+      "epoch": 0.8891126232467713,
+      "grad_norm": 3.6635571070717043,
+      "learning_rate": 3.1904017351933715e-07,
+      "loss": 0.298,
+      "step": 12805
+    },
+    {
+      "epoch": 0.8891820580474934,
+      "grad_norm": 3.683589839225522,
+      "learning_rate": 3.186450468413638e-07,
+      "loss": 0.3893,
+      "step": 12806
+    },
+    {
+      "epoch": 0.8892514928482155,
+      "grad_norm": 5.032389855644175,
+      "learning_rate": 3.1825015694102736e-07,
+      "loss": 0.264,
+      "step": 12807
+    },
+    {
+      "epoch": 0.8893209276489377,
+      "grad_norm": 4.625058846296304,
+      "learning_rate": 3.178555038383002e-07,
+      "loss": 0.6085,
+      "step": 12808
+    },
+    {
+      "epoch": 0.8893903624496597,
+      "grad_norm": 3.994245846464518,
+      "learning_rate": 3.1746108755314307e-07,
+      "loss": 0.4114,
+      "step": 12809
+    },
+    {
+      "epoch": 0.8894597972503819,
+      "grad_norm": 3.2871479565223165,
+      "learning_rate": 3.1706690810550613e-07,
+      "loss": 0.3126,
+      "step": 12810
+    },
+    {
+      "epoch": 0.889529232051104,
+      "grad_norm": 2.6788534825148336,
+      "learning_rate": 3.166729655153256e-07,
+      "loss": 0.1957,
+      "step": 12811
+    },
+    {
+      "epoch": 0.8895986668518261,
+      "grad_norm": 4.054121480576292,
+      "learning_rate": 3.1627925980252717e-07,
+      "loss": 0.3945,
+      "step": 12812
+    },
+    {
+      "epoch": 0.8896681016525483,
+      "grad_norm": 2.7179163563916684,
+      "learning_rate": 3.1588579098702333e-07,
+      "loss": 0.2858,
+      "step": 12813
+    },
+    {
+      "epoch": 0.8897375364532704,
+      "grad_norm": 4.539162753457772,
+      "learning_rate": 3.154925590887159e-07,
+      "loss": 0.6347,
+      "step": 12814
+    },
+    {
+      "epoch": 0.8898069712539926,
+      "grad_norm": 3.6091242386609497,
+      "learning_rate": 3.150995641274934e-07,
+      "loss": 0.4113,
+      "step": 12815
+    },
+    {
+      "epoch": 0.8898764060547146,
+      "grad_norm": 3.70707630163381,
+      "learning_rate": 3.1470680612323503e-07,
+      "loss": 0.3912,
+      "step": 12816
+    },
+    {
+      "epoch": 0.8899458408554367,
+      "grad_norm": 3.7386078346168583,
+      "learning_rate": 3.1431428509580365e-07,
+      "loss": 0.3863,
+      "step": 12817
+    },
+    {
+      "epoch": 0.8900152756561589,
+      "grad_norm": 3.9583681600677894,
+      "learning_rate": 3.1392200106505287e-07,
+      "loss": 0.4671,
+      "step": 12818
+    },
+    {
+      "epoch": 0.890084710456881,
+      "grad_norm": 2.757319084506268,
+      "learning_rate": 3.1352995405082576e-07,
+      "loss": 0.1399,
+      "step": 12819
+    },
+    {
+      "epoch": 0.8901541452576031,
+      "grad_norm": 3.4486305106287825,
+      "learning_rate": 3.131381440729508e-07,
+      "loss": 0.2294,
+      "step": 12820
+    },
+    {
+      "epoch": 0.8902235800583252,
+      "grad_norm": 4.542063529139695,
+      "learning_rate": 3.1274657115124397e-07,
+      "loss": 0.4704,
+      "step": 12821
+    },
+    {
+      "epoch": 0.8902930148590473,
+      "grad_norm": 3.7980928417680637,
+      "learning_rate": 3.12355235305512e-07,
+      "loss": 0.4042,
+      "step": 12822
+    },
+    {
+      "epoch": 0.8903624496597695,
+      "grad_norm": 3.1687854317295976,
+      "learning_rate": 3.119641365555481e-07,
+      "loss": 0.2905,
+      "step": 12823
+    },
+    {
+      "epoch": 0.8904318844604916,
+      "grad_norm": 3.1578138576732737,
+      "learning_rate": 3.1157327492113255e-07,
+      "loss": 0.3377,
+      "step": 12824
+    },
+    {
+      "epoch": 0.8905013192612137,
+      "grad_norm": 4.9479643719424695,
+      "learning_rate": 3.1118265042203553e-07,
+      "loss": 0.4317,
+      "step": 12825
+    },
+    {
+      "epoch": 0.8905707540619359,
+      "grad_norm": 4.001335759330411,
+      "learning_rate": 3.107922630780147e-07,
+      "loss": 0.3943,
+      "step": 12826
+    },
+    {
+      "epoch": 0.8906401888626579,
+      "grad_norm": 3.9556990379016277,
+      "learning_rate": 3.104021129088153e-07,
+      "loss": 0.4359,
+      "step": 12827
+    },
+    {
+      "epoch": 0.8907096236633801,
+      "grad_norm": 4.159421379416483,
+      "learning_rate": 3.1001219993416986e-07,
+      "loss": 0.5689,
+      "step": 12828
+    },
+    {
+      "epoch": 0.8907790584641022,
+      "grad_norm": 3.7036053809797167,
+      "learning_rate": 3.0962252417380045e-07,
+      "loss": 0.3535,
+      "step": 12829
+    },
+    {
+      "epoch": 0.8908484932648243,
+      "grad_norm": 4.4094530755562085,
+      "learning_rate": 3.092330856474163e-07,
+      "loss": 0.4673,
+      "step": 12830
+    },
+    {
+      "epoch": 0.8909179280655465,
+      "grad_norm": 4.395263331199589,
+      "learning_rate": 3.088438843747149e-07,
+      "loss": 0.4345,
+      "step": 12831
+    },
+    {
+      "epoch": 0.8909873628662686,
+      "grad_norm": 3.863118442902577,
+      "learning_rate": 3.0845492037538117e-07,
+      "loss": 0.4144,
+      "step": 12832
+    },
+    {
+      "epoch": 0.8910567976669906,
+      "grad_norm": 4.625055215497941,
+      "learning_rate": 3.080661936690876e-07,
+      "loss": 0.5333,
+      "step": 12833
+    },
+    {
+      "epoch": 0.8911262324677128,
+      "grad_norm": 3.459298215402622,
+      "learning_rate": 3.0767770427549857e-07,
+      "loss": 0.2689,
+      "step": 12834
+    },
+    {
+      "epoch": 0.8911956672684349,
+      "grad_norm": 3.420341231284176,
+      "learning_rate": 3.0728945221425997e-07,
+      "loss": 0.3997,
+      "step": 12835
+    },
+    {
+      "epoch": 0.8912651020691571,
+      "grad_norm": 4.064553612764126,
+      "learning_rate": 3.069014375050117e-07,
+      "loss": 0.3529,
+      "step": 12836
+    },
+    {
+      "epoch": 0.8913345368698792,
+      "grad_norm": 3.507685745807081,
+      "learning_rate": 3.0651366016737794e-07,
+      "loss": 0.2878,
+      "step": 12837
+    },
+    {
+      "epoch": 0.8914039716706013,
+      "grad_norm": 3.2822661061839464,
+      "learning_rate": 3.06126120220972e-07,
+      "loss": 0.4179,
+      "step": 12838
+    },
+    {
+      "epoch": 0.8914734064713234,
+      "grad_norm": 4.270972169712055,
+      "learning_rate": 3.057388176853959e-07,
+      "loss": 0.3712,
+      "step": 12839
+    },
+    {
+      "epoch": 0.8915428412720455,
+      "grad_norm": 3.415788309261789,
+      "learning_rate": 3.0535175258023797e-07,
+      "loss": 0.3393,
+      "step": 12840
+    },
+    {
+      "epoch": 0.8916122760727677,
+      "grad_norm": 3.706596251951023,
+      "learning_rate": 3.0496492492507636e-07,
+      "loss": 0.3564,
+      "step": 12841
+    },
+    {
+      "epoch": 0.8916817108734898,
+      "grad_norm": 5.195888331262543,
+      "learning_rate": 3.0457833473947595e-07,
+      "loss": 0.4879,
+      "step": 12842
+    },
+    {
+      "epoch": 0.8917511456742119,
+      "grad_norm": 4.226074969457352,
+      "learning_rate": 3.041919820429906e-07,
+      "loss": 0.4914,
+      "step": 12843
+    },
+    {
+      "epoch": 0.8918205804749341,
+      "grad_norm": 2.7297929457870556,
+      "learning_rate": 3.0380586685515965e-07,
+      "loss": 0.2075,
+      "step": 12844
+    },
+    {
+      "epoch": 0.8918900152756561,
+      "grad_norm": 4.111806190454208,
+      "learning_rate": 3.0341998919551586e-07,
+      "loss": 0.3406,
+      "step": 12845
+    },
+    {
+      "epoch": 0.8919594500763783,
+      "grad_norm": 4.301417007083595,
+      "learning_rate": 3.0303434908357355e-07,
+      "loss": 0.473,
+      "step": 12846
+    },
+    {
+      "epoch": 0.8920288848771004,
+      "grad_norm": 2.8294405638990114,
+      "learning_rate": 3.026489465388377e-07,
+      "loss": 0.1983,
+      "step": 12847
+    },
+    {
+      "epoch": 0.8920983196778225,
+      "grad_norm": 4.520265092031402,
+      "learning_rate": 3.0226378158080495e-07,
+      "loss": 0.5387,
+      "step": 12848
+    },
+    {
+      "epoch": 0.8921677544785447,
+      "grad_norm": 4.159576014079366,
+      "learning_rate": 3.0187885422895303e-07,
+      "loss": 0.6115,
+      "step": 12849
+    },
+    {
+      "epoch": 0.8922371892792668,
+      "grad_norm": 3.568281876383245,
+      "learning_rate": 3.014941645027519e-07,
+      "loss": 0.3014,
+      "step": 12850
+    },
+    {
+      "epoch": 0.8923066240799888,
+      "grad_norm": 3.599825103200741,
+      "learning_rate": 3.0110971242166e-07,
+      "loss": 0.3437,
+      "step": 12851
+    },
+    {
+      "epoch": 0.892376058880711,
+      "grad_norm": 5.068268914030686,
+      "learning_rate": 3.0072549800512216e-07,
+      "loss": 0.4464,
+      "step": 12852
+    },
+    {
+      "epoch": 0.8924454936814331,
+      "grad_norm": 3.8789950123010613,
+      "learning_rate": 3.0034152127256966e-07,
+      "loss": 0.4476,
+      "step": 12853
+    },
+    {
+      "epoch": 0.8925149284821553,
+      "grad_norm": 2.939982041177012,
+      "learning_rate": 2.999577822434258e-07,
+      "loss": 0.1564,
+      "step": 12854
+    },
+    {
+      "epoch": 0.8925843632828774,
+      "grad_norm": 4.99509754427649,
+      "learning_rate": 2.9957428093709885e-07,
+      "loss": 0.5681,
+      "step": 12855
+    },
+    {
+      "epoch": 0.8926537980835995,
+      "grad_norm": 4.770607016159064,
+      "learning_rate": 2.9919101737298615e-07,
+      "loss": 0.4218,
+      "step": 12856
+    },
+    {
+      "epoch": 0.8927232328843216,
+      "grad_norm": 2.845691271695139,
+      "learning_rate": 2.988079915704717e-07,
+      "loss": 0.2832,
+      "step": 12857
+    },
+    {
+      "epoch": 0.8927926676850437,
+      "grad_norm": 2.286775951779598,
+      "learning_rate": 2.984252035489299e-07,
+      "loss": 0.0966,
+      "step": 12858
+    },
+    {
+      "epoch": 0.8928621024857659,
+      "grad_norm": 4.312400199508682,
+      "learning_rate": 2.980426533277209e-07,
+      "loss": 0.3796,
+      "step": 12859
+    },
+    {
+      "epoch": 0.892931537286488,
+      "grad_norm": 4.508689968288141,
+      "learning_rate": 2.976603409261941e-07,
+      "loss": 0.4633,
+      "step": 12860
+    },
+    {
+      "epoch": 0.8930009720872101,
+      "grad_norm": 3.470159154909685,
+      "learning_rate": 2.9727826636368586e-07,
+      "loss": 0.3905,
+      "step": 12861
+    },
+    {
+      "epoch": 0.8930704068879323,
+      "grad_norm": 3.39971287568434,
+      "learning_rate": 2.968964296595217e-07,
+      "loss": 0.3133,
+      "step": 12862
+    },
+    {
+      "epoch": 0.8931398416886543,
+      "grad_norm": 3.951582877566643,
+      "learning_rate": 2.96514830833014e-07,
+      "loss": 0.4118,
+      "step": 12863
+    },
+    {
+      "epoch": 0.8932092764893764,
+      "grad_norm": 2.662907995410107,
+      "learning_rate": 2.961334699034635e-07,
+      "loss": 0.254,
+      "step": 12864
+    },
+    {
+      "epoch": 0.8932787112900986,
+      "grad_norm": 3.7258462642264556,
+      "learning_rate": 2.9575234689015965e-07,
+      "loss": 0.2729,
+      "step": 12865
+    },
+    {
+      "epoch": 0.8933481460908207,
+      "grad_norm": 3.8908226603081246,
+      "learning_rate": 2.953714618123793e-07,
+      "loss": 0.3516,
+      "step": 12866
+    },
+    {
+      "epoch": 0.8934175808915429,
+      "grad_norm": 4.214120431869059,
+      "learning_rate": 2.9499081468938594e-07,
+      "loss": 0.4436,
+      "step": 12867
+    },
+    {
+      "epoch": 0.893487015692265,
+      "grad_norm": 3.8448962919748646,
+      "learning_rate": 2.9461040554043365e-07,
+      "loss": 0.3761,
+      "step": 12868
+    },
+    {
+      "epoch": 0.893556450492987,
+      "grad_norm": 4.447035620151592,
+      "learning_rate": 2.942302343847625e-07,
+      "loss": 0.4246,
+      "step": 12869
+    },
+    {
+      "epoch": 0.8936258852937092,
+      "grad_norm": 3.169314913408982,
+      "learning_rate": 2.938503012416011e-07,
+      "loss": 0.2155,
+      "step": 12870
+    },
+    {
+      "epoch": 0.8936953200944313,
+      "grad_norm": 4.623362617989716,
+      "learning_rate": 2.9347060613016564e-07,
+      "loss": 0.5392,
+      "step": 12871
+    },
+    {
+      "epoch": 0.8937647548951535,
+      "grad_norm": 4.072746544466012,
+      "learning_rate": 2.9309114906966197e-07,
+      "loss": 0.5797,
+      "step": 12872
+    },
+    {
+      "epoch": 0.8938341896958756,
+      "grad_norm": 3.0597334939781846,
+      "learning_rate": 2.927119300792813e-07,
+      "loss": 0.2887,
+      "step": 12873
+    },
+    {
+      "epoch": 0.8939036244965977,
+      "grad_norm": 3.455371105441647,
+      "learning_rate": 2.9233294917820454e-07,
+      "loss": 0.3669,
+      "step": 12874
+    },
+    {
+      "epoch": 0.8939730592973198,
+      "grad_norm": 3.836268660844611,
+      "learning_rate": 2.9195420638560013e-07,
+      "loss": 0.3296,
+      "step": 12875
+    },
+    {
+      "epoch": 0.8940424940980419,
+      "grad_norm": 2.5184324375340266,
+      "learning_rate": 2.9157570172062386e-07,
+      "loss": 0.1885,
+      "step": 12876
+    },
+    {
+      "epoch": 0.894111928898764,
+      "grad_norm": 3.468973558208001,
+      "learning_rate": 2.9119743520242216e-07,
+      "loss": 0.4362,
+      "step": 12877
+    },
+    {
+      "epoch": 0.8941813636994862,
+      "grad_norm": 2.744911371048733,
+      "learning_rate": 2.9081940685012464e-07,
+      "loss": 0.1756,
+      "step": 12878
+    },
+    {
+      "epoch": 0.8942507985002083,
+      "grad_norm": 4.498927499720553,
+      "learning_rate": 2.9044161668285275e-07,
+      "loss": 0.3641,
+      "step": 12879
+    },
+    {
+      "epoch": 0.8943202333009305,
+      "grad_norm": 3.013664754266281,
+      "learning_rate": 2.900640647197156e-07,
+      "loss": 0.2386,
+      "step": 12880
+    },
+    {
+      "epoch": 0.8943896681016525,
+      "grad_norm": 4.380782575717267,
+      "learning_rate": 2.896867509798074e-07,
+      "loss": 0.5476,
+      "step": 12881
+    },
+    {
+      "epoch": 0.8944591029023746,
+      "grad_norm": 3.194992856690282,
+      "learning_rate": 2.8930967548221334e-07,
+      "loss": 0.3822,
+      "step": 12882
+    },
+    {
+      "epoch": 0.8945285377030968,
+      "grad_norm": 3.954452200291028,
+      "learning_rate": 2.889328382460055e-07,
+      "loss": 0.3742,
+      "step": 12883
+    },
+    {
+      "epoch": 0.8945979725038189,
+      "grad_norm": 4.115262132264761,
+      "learning_rate": 2.8855623929024357e-07,
+      "loss": 0.2841,
+      "step": 12884
+    },
+    {
+      "epoch": 0.8946674073045411,
+      "grad_norm": 5.052765348126075,
+      "learning_rate": 2.881798786339762e-07,
+      "loss": 0.3741,
+      "step": 12885
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 3.617217884568912,
+      "learning_rate": 2.878037562962388e-07,
+      "loss": 0.2994,
+      "step": 12886
+    },
+    {
+      "epoch": 0.8948062769059852,
+      "grad_norm": 3.987513437086178,
+      "learning_rate": 2.8742787229605494e-07,
+      "loss": 0.375,
+      "step": 12887
+    },
+    {
+      "epoch": 0.8948757117067074,
+      "grad_norm": 5.1224525053250085,
+      "learning_rate": 2.870522266524367e-07,
+      "loss": 0.573,
+      "step": 12888
+    },
+    {
+      "epoch": 0.8949451465074295,
+      "grad_norm": 2.6481383068233626,
+      "learning_rate": 2.8667681938438384e-07,
+      "loss": 0.2097,
+      "step": 12889
+    },
+    {
+      "epoch": 0.8950145813081516,
+      "grad_norm": 4.302001559100502,
+      "learning_rate": 2.8630165051088397e-07,
+      "loss": 0.3005,
+      "step": 12890
+    },
+    {
+      "epoch": 0.8950840161088738,
+      "grad_norm": 3.5844565984767454,
+      "learning_rate": 2.8592672005091307e-07,
+      "loss": 0.4488,
+      "step": 12891
+    },
+    {
+      "epoch": 0.8951534509095959,
+      "grad_norm": 2.4295574838960032,
+      "learning_rate": 2.855520280234336e-07,
+      "loss": 0.174,
+      "step": 12892
+    },
+    {
+      "epoch": 0.895222885710318,
+      "grad_norm": 4.274389114395069,
+      "learning_rate": 2.8517757444739836e-07,
+      "loss": 0.4211,
+      "step": 12893
+    },
+    {
+      "epoch": 0.8952923205110401,
+      "grad_norm": 3.5620174423235063,
+      "learning_rate": 2.848033593417454e-07,
+      "loss": 0.3719,
+      "step": 12894
+    },
+    {
+      "epoch": 0.8953617553117622,
+      "grad_norm": 3.357212606211454,
+      "learning_rate": 2.844293827254041e-07,
+      "loss": 0.233,
+      "step": 12895
+    },
+    {
+      "epoch": 0.8954311901124844,
+      "grad_norm": 3.553012340361745,
+      "learning_rate": 2.8405564461728706e-07,
+      "loss": 0.3876,
+      "step": 12896
+    },
+    {
+      "epoch": 0.8955006249132065,
+      "grad_norm": 4.705628512769024,
+      "learning_rate": 2.836821450363003e-07,
+      "loss": 0.6429,
+      "step": 12897
+    },
+    {
+      "epoch": 0.8955700597139287,
+      "grad_norm": 4.165164560393895,
+      "learning_rate": 2.833088840013343e-07,
+      "loss": 0.3633,
+      "step": 12898
+    },
+    {
+      "epoch": 0.8956394945146507,
+      "grad_norm": 3.1987366408535545,
+      "learning_rate": 2.829358615312661e-07,
+      "loss": 0.3947,
+      "step": 12899
+    },
+    {
+      "epoch": 0.8957089293153728,
+      "grad_norm": 4.530685887602706,
+      "learning_rate": 2.8256307764496503e-07,
+      "loss": 0.3641,
+      "step": 12900
+    },
+    {
+      "epoch": 0.895778364116095,
+      "grad_norm": 3.957455735579709,
+      "learning_rate": 2.82190532361285e-07,
+      "loss": 0.3392,
+      "step": 12901
+    },
+    {
+      "epoch": 0.8958477989168171,
+      "grad_norm": 6.111207436450076,
+      "learning_rate": 2.8181822569906983e-07,
+      "loss": 0.5198,
+      "step": 12902
+    },
+    {
+      "epoch": 0.8959172337175393,
+      "grad_norm": 3.501150484132737,
+      "learning_rate": 2.814461576771499e-07,
+      "loss": 0.3344,
+      "step": 12903
+    },
+    {
+      "epoch": 0.8959866685182614,
+      "grad_norm": 4.130746907424348,
+      "learning_rate": 2.8107432831434355e-07,
+      "loss": 0.4193,
+      "step": 12904
+    },
+    {
+      "epoch": 0.8960561033189834,
+      "grad_norm": 5.2765705484883,
+      "learning_rate": 2.8070273762945797e-07,
+      "loss": 0.6294,
+      "step": 12905
+    },
+    {
+      "epoch": 0.8961255381197056,
+      "grad_norm": 4.065809948273628,
+      "learning_rate": 2.803313856412876e-07,
+      "loss": 0.45,
+      "step": 12906
+    },
+    {
+      "epoch": 0.8961949729204277,
+      "grad_norm": 2.1505868102158785,
+      "learning_rate": 2.7996027236861514e-07,
+      "loss": 0.1173,
+      "step": 12907
+    },
+    {
+      "epoch": 0.8962644077211498,
+      "grad_norm": 3.6164522357719338,
+      "learning_rate": 2.7958939783021057e-07,
+      "loss": 0.3747,
+      "step": 12908
+    },
+    {
+      "epoch": 0.896333842521872,
+      "grad_norm": 3.2789791885258444,
+      "learning_rate": 2.7921876204483387e-07,
+      "loss": 0.3375,
+      "step": 12909
+    },
+    {
+      "epoch": 0.8964032773225941,
+      "grad_norm": 4.302051014171993,
+      "learning_rate": 2.7884836503122956e-07,
+      "loss": 0.4435,
+      "step": 12910
+    },
+    {
+      "epoch": 0.8964727121233163,
+      "grad_norm": 3.1925006951741755,
+      "learning_rate": 2.784782068081315e-07,
+      "loss": 0.3418,
+      "step": 12911
+    },
+    {
+      "epoch": 0.8965421469240383,
+      "grad_norm": 3.565139274210202,
+      "learning_rate": 2.7810828739426533e-07,
+      "loss": 0.349,
+      "step": 12912
+    },
+    {
+      "epoch": 0.8966115817247604,
+      "grad_norm": 4.853294259501187,
+      "learning_rate": 2.7773860680833706e-07,
+      "loss": 0.5076,
+      "step": 12913
+    },
+    {
+      "epoch": 0.8966810165254826,
+      "grad_norm": 3.257888405696199,
+      "learning_rate": 2.773691650690463e-07,
+      "loss": 0.2372,
+      "step": 12914
+    },
+    {
+      "epoch": 0.8967504513262047,
+      "grad_norm": 2.3821591604489787,
+      "learning_rate": 2.769999621950803e-07,
+      "loss": 0.2213,
+      "step": 12915
+    },
+    {
+      "epoch": 0.8968198861269269,
+      "grad_norm": 2.9701936148645904,
+      "learning_rate": 2.766309982051113e-07,
+      "loss": 0.303,
+      "step": 12916
+    },
+    {
+      "epoch": 0.896889320927649,
+      "grad_norm": 4.258624028878511,
+      "learning_rate": 2.7626227311780105e-07,
+      "loss": 0.3848,
+      "step": 12917
+    },
+    {
+      "epoch": 0.896958755728371,
+      "grad_norm": 3.1360053188952812,
+      "learning_rate": 2.7589378695180027e-07,
+      "loss": 0.386,
+      "step": 12918
+    },
+    {
+      "epoch": 0.8970281905290932,
+      "grad_norm": 3.504461249644197,
+      "learning_rate": 2.7552553972574614e-07,
+      "loss": 0.3022,
+      "step": 12919
+    },
+    {
+      "epoch": 0.8970976253298153,
+      "grad_norm": 5.9034992727899205,
+      "learning_rate": 2.751575314582644e-07,
+      "loss": 0.4118,
+      "step": 12920
+    },
+    {
+      "epoch": 0.8971670601305374,
+      "grad_norm": 3.4579194707666634,
+      "learning_rate": 2.747897621679674e-07,
+      "loss": 0.2884,
+      "step": 12921
+    },
+    {
+      "epoch": 0.8972364949312596,
+      "grad_norm": 4.253551168506638,
+      "learning_rate": 2.74422231873458e-07,
+      "loss": 0.3824,
+      "step": 12922
+    },
+    {
+      "epoch": 0.8973059297319816,
+      "grad_norm": 3.8287897330114347,
+      "learning_rate": 2.740549405933246e-07,
+      "loss": 0.4417,
+      "step": 12923
+    },
+    {
+      "epoch": 0.8973753645327038,
+      "grad_norm": 3.3737153011687955,
+      "learning_rate": 2.736878883461441e-07,
+      "loss": 0.3343,
+      "step": 12924
+    },
+    {
+      "epoch": 0.8974447993334259,
+      "grad_norm": 3.141404413860597,
+      "learning_rate": 2.733210751504817e-07,
+      "loss": 0.228,
+      "step": 12925
+    },
+    {
+      "epoch": 0.897514234134148,
+      "grad_norm": 4.328134240873491,
+      "learning_rate": 2.7295450102489127e-07,
+      "loss": 0.5278,
+      "step": 12926
+    },
+    {
+      "epoch": 0.8975836689348702,
+      "grad_norm": 4.372730357774936,
+      "learning_rate": 2.7258816598791427e-07,
+      "loss": 0.3937,
+      "step": 12927
+    },
+    {
+      "epoch": 0.8976531037355923,
+      "grad_norm": 4.25056635375421,
+      "learning_rate": 2.722220700580763e-07,
+      "loss": 0.4751,
+      "step": 12928
+    },
+    {
+      "epoch": 0.8977225385363145,
+      "grad_norm": 4.049791838382141,
+      "learning_rate": 2.718562132538971e-07,
+      "loss": 0.4032,
+      "step": 12929
+    },
+    {
+      "epoch": 0.8977919733370365,
+      "grad_norm": 4.4959586251954535,
+      "learning_rate": 2.7149059559388126e-07,
+      "loss": 0.4552,
+      "step": 12930
+    },
+    {
+      "epoch": 0.8978614081377586,
+      "grad_norm": 4.12449014113677,
+      "learning_rate": 2.7112521709651896e-07,
+      "loss": 0.349,
+      "step": 12931
+    },
+    {
+      "epoch": 0.8979308429384808,
+      "grad_norm": 4.303215565335407,
+      "learning_rate": 2.7076007778029325e-07,
+      "loss": 0.3358,
+      "step": 12932
+    },
+    {
+      "epoch": 0.8980002777392029,
+      "grad_norm": 5.221679613651867,
+      "learning_rate": 2.70395177663671e-07,
+      "loss": 0.4333,
+      "step": 12933
+    },
+    {
+      "epoch": 0.898069712539925,
+      "grad_norm": 4.718617696215007,
+      "learning_rate": 2.700305167651085e-07,
+      "loss": 0.3898,
+      "step": 12934
+    },
+    {
+      "epoch": 0.8981391473406471,
+      "grad_norm": 4.018147620291798,
+      "learning_rate": 2.6966609510304995e-07,
+      "loss": 0.4427,
+      "step": 12935
+    },
+    {
+      "epoch": 0.8982085821413692,
+      "grad_norm": 3.3984143802994415,
+      "learning_rate": 2.693019126959284e-07,
+      "loss": 0.2871,
+      "step": 12936
+    },
+    {
+      "epoch": 0.8982780169420914,
+      "grad_norm": 4.354403276095537,
+      "learning_rate": 2.6893796956216244e-07,
+      "loss": 0.5377,
+      "step": 12937
+    },
+    {
+      "epoch": 0.8983474517428135,
+      "grad_norm": 3.5342686773579954,
+      "learning_rate": 2.685742657201601e-07,
+      "loss": 0.4195,
+      "step": 12938
+    },
+    {
+      "epoch": 0.8984168865435356,
+      "grad_norm": 5.135105535455678,
+      "learning_rate": 2.6821080118831775e-07,
+      "loss": 0.6695,
+      "step": 12939
+    },
+    {
+      "epoch": 0.8984863213442578,
+      "grad_norm": 3.7698273401362274,
+      "learning_rate": 2.67847575985018e-07,
+      "loss": 0.3073,
+      "step": 12940
+    },
+    {
+      "epoch": 0.8985557561449798,
+      "grad_norm": 3.3237398968772,
+      "learning_rate": 2.674845901286349e-07,
+      "loss": 0.3527,
+      "step": 12941
+    },
+    {
+      "epoch": 0.898625190945702,
+      "grad_norm": 3.659822417326642,
+      "learning_rate": 2.6712184363752447e-07,
+      "loss": 0.2923,
+      "step": 12942
+    },
+    {
+      "epoch": 0.8986946257464241,
+      "grad_norm": 2.7203776694679758,
+      "learning_rate": 2.667593365300353e-07,
+      "loss": 0.093,
+      "step": 12943
+    },
+    {
+      "epoch": 0.8987640605471462,
+      "grad_norm": 4.360891489953443,
+      "learning_rate": 2.663970688245038e-07,
+      "loss": 0.397,
+      "step": 12944
+    },
+    {
+      "epoch": 0.8988334953478684,
+      "grad_norm": 3.55398659231266,
+      "learning_rate": 2.66035040539252e-07,
+      "loss": 0.2696,
+      "step": 12945
+    },
+    {
+      "epoch": 0.8989029301485905,
+      "grad_norm": 3.938602362393521,
+      "learning_rate": 2.6567325169259085e-07,
+      "loss": 0.2985,
+      "step": 12946
+    },
+    {
+      "epoch": 0.8989723649493125,
+      "grad_norm": 3.7470886996640274,
+      "learning_rate": 2.6531170230281954e-07,
+      "loss": 0.4329,
+      "step": 12947
+    },
+    {
+      "epoch": 0.8990417997500347,
+      "grad_norm": 3.7273255572022754,
+      "learning_rate": 2.6495039238822463e-07,
+      "loss": 0.5164,
+      "step": 12948
+    },
+    {
+      "epoch": 0.8991112345507568,
+      "grad_norm": 3.236643231517121,
+      "learning_rate": 2.645893219670803e-07,
+      "loss": 0.2181,
+      "step": 12949
+    },
+    {
+      "epoch": 0.899180669351479,
+      "grad_norm": 4.069185383680679,
+      "learning_rate": 2.6422849105765034e-07,
+      "loss": 0.348,
+      "step": 12950
+    },
+    {
+      "epoch": 0.8992501041522011,
+      "grad_norm": 4.28488822157351,
+      "learning_rate": 2.6386789967818407e-07,
+      "loss": 0.57,
+      "step": 12951
+    },
+    {
+      "epoch": 0.8993195389529232,
+      "grad_norm": 4.91881961176447,
+      "learning_rate": 2.6350754784692014e-07,
+      "loss": 0.4561,
+      "step": 12952
+    },
+    {
+      "epoch": 0.8993889737536453,
+      "grad_norm": 4.533544274943324,
+      "learning_rate": 2.6314743558208457e-07,
+      "loss": 0.5344,
+      "step": 12953
+    },
+    {
+      "epoch": 0.8994584085543674,
+      "grad_norm": 4.886559031694449,
+      "learning_rate": 2.627875629018911e-07,
+      "loss": 0.8121,
+      "step": 12954
+    },
+    {
+      "epoch": 0.8995278433550896,
+      "grad_norm": 3.5113065457779356,
+      "learning_rate": 2.624279298245436e-07,
+      "loss": 0.3544,
+      "step": 12955
+    },
+    {
+      "epoch": 0.8995972781558117,
+      "grad_norm": 4.001815023685714,
+      "learning_rate": 2.6206853636823016e-07,
+      "loss": 0.4938,
+      "step": 12956
+    },
+    {
+      "epoch": 0.8996667129565338,
+      "grad_norm": 3.5534610699023053,
+      "learning_rate": 2.6170938255112745e-07,
+      "loss": 0.3296,
+      "step": 12957
+    },
+    {
+      "epoch": 0.899736147757256,
+      "grad_norm": 4.60547226414963,
+      "learning_rate": 2.613504683914031e-07,
+      "loss": 0.5194,
+      "step": 12958
+    },
+    {
+      "epoch": 0.899805582557978,
+      "grad_norm": 4.4658279320436804,
+      "learning_rate": 2.60991793907211e-07,
+      "loss": 0.4448,
+      "step": 12959
+    },
+    {
+      "epoch": 0.8998750173587002,
+      "grad_norm": 2.98950975394793,
+      "learning_rate": 2.6063335911668983e-07,
+      "loss": 0.2301,
+      "step": 12960
+    },
+    {
+      "epoch": 0.8999444521594223,
+      "grad_norm": 3.2029875421650003,
+      "learning_rate": 2.6027516403797135e-07,
+      "loss": 0.275,
+      "step": 12961
+    },
+    {
+      "epoch": 0.9000138869601444,
+      "grad_norm": 3.194784992993854,
+      "learning_rate": 2.5991720868917216e-07,
+      "loss": 0.3154,
+      "step": 12962
+    },
+    {
+      "epoch": 0.9000833217608666,
+      "grad_norm": 4.982827886146601,
+      "learning_rate": 2.5955949308839545e-07,
+      "loss": 0.8683,
+      "step": 12963
+    },
+    {
+      "epoch": 0.9001527565615887,
+      "grad_norm": 3.9277956766939903,
+      "learning_rate": 2.592020172537363e-07,
+      "loss": 0.4277,
+      "step": 12964
+    },
+    {
+      "epoch": 0.9002221913623107,
+      "grad_norm": 4.372360976809634,
+      "learning_rate": 2.588447812032746e-07,
+      "loss": 0.4201,
+      "step": 12965
+    },
+    {
+      "epoch": 0.9002916261630329,
+      "grad_norm": 4.372097880989456,
+      "learning_rate": 2.5848778495507876e-07,
+      "loss": 0.5957,
+      "step": 12966
+    },
+    {
+      "epoch": 0.900361060963755,
+      "grad_norm": 4.85031711931894,
+      "learning_rate": 2.5813102852720487e-07,
+      "loss": 0.3726,
+      "step": 12967
+    },
+    {
+      "epoch": 0.9004304957644772,
+      "grad_norm": 3.6557133161209654,
+      "learning_rate": 2.577745119376984e-07,
+      "loss": 0.3604,
+      "step": 12968
+    },
+    {
+      "epoch": 0.9004999305651993,
+      "grad_norm": 5.812470646942752,
+      "learning_rate": 2.574182352045901e-07,
+      "loss": 0.6442,
+      "step": 12969
+    },
+    {
+      "epoch": 0.9005693653659214,
+      "grad_norm": 3.064031630593118,
+      "learning_rate": 2.570621983459021e-07,
+      "loss": 0.1917,
+      "step": 12970
+    },
+    {
+      "epoch": 0.9006388001666435,
+      "grad_norm": 3.3897776916762683,
+      "learning_rate": 2.5670640137964053e-07,
+      "loss": 0.3184,
+      "step": 12971
+    },
+    {
+      "epoch": 0.9007082349673656,
+      "grad_norm": 4.163334129236111,
+      "learning_rate": 2.563508443238011e-07,
+      "loss": 0.428,
+      "step": 12972
+    },
+    {
+      "epoch": 0.9007776697680878,
+      "grad_norm": 3.3823411763199918,
+      "learning_rate": 2.5599552719636933e-07,
+      "loss": 0.3516,
+      "step": 12973
+    },
+    {
+      "epoch": 0.9008471045688099,
+      "grad_norm": 3.6700561492783916,
+      "learning_rate": 2.556404500153148e-07,
+      "loss": 0.3557,
+      "step": 12974
+    },
+    {
+      "epoch": 0.900916539369532,
+      "grad_norm": 3.995218927239919,
+      "learning_rate": 2.5528561279859756e-07,
+      "loss": 0.2665,
+      "step": 12975
+    },
+    {
+      "epoch": 0.9009859741702542,
+      "grad_norm": 3.0094200666856783,
+      "learning_rate": 2.5493101556416544e-07,
+      "loss": 0.2654,
+      "step": 12976
+    },
+    {
+      "epoch": 0.9010554089709762,
+      "grad_norm": 3.2953052408947006,
+      "learning_rate": 2.5457665832995305e-07,
+      "loss": 0.3069,
+      "step": 12977
+    },
+    {
+      "epoch": 0.9011248437716983,
+      "grad_norm": 3.4710947342011225,
+      "learning_rate": 2.542225411138832e-07,
+      "loss": 0.2857,
+      "step": 12978
+    },
+    {
+      "epoch": 0.9011942785724205,
+      "grad_norm": 3.6863493458590737,
+      "learning_rate": 2.5386866393386775e-07,
+      "loss": 0.2976,
+      "step": 12979
+    },
+    {
+      "epoch": 0.9012637133731426,
+      "grad_norm": 3.451147237073787,
+      "learning_rate": 2.5351502680780406e-07,
+      "loss": 0.3346,
+      "step": 12980
+    },
+    {
+      "epoch": 0.9013331481738648,
+      "grad_norm": 3.1789780825732876,
+      "learning_rate": 2.531616297535794e-07,
+      "loss": 0.3592,
+      "step": 12981
+    },
+    {
+      "epoch": 0.9014025829745869,
+      "grad_norm": 6.056553169980843,
+      "learning_rate": 2.528084727890684e-07,
+      "loss": 0.5311,
+      "step": 12982
+    },
+    {
+      "epoch": 0.9014720177753089,
+      "grad_norm": 6.049113295295283,
+      "learning_rate": 2.524555559321329e-07,
+      "loss": 0.572,
+      "step": 12983
+    },
+    {
+      "epoch": 0.9015414525760311,
+      "grad_norm": 4.20719668247249,
+      "learning_rate": 2.5210287920062304e-07,
+      "loss": 0.4302,
+      "step": 12984
+    },
+    {
+      "epoch": 0.9016108873767532,
+      "grad_norm": 4.578346458285832,
+      "learning_rate": 2.517504426123768e-07,
+      "loss": 0.6333,
+      "step": 12985
+    },
+    {
+      "epoch": 0.9016803221774754,
+      "grad_norm": 4.600342020563578,
+      "learning_rate": 2.5139824618521993e-07,
+      "loss": 0.5494,
+      "step": 12986
+    },
+    {
+      "epoch": 0.9017497569781975,
+      "grad_norm": 4.85189613511165,
+      "learning_rate": 2.5104628993696757e-07,
+      "loss": 0.4686,
+      "step": 12987
+    },
+    {
+      "epoch": 0.9018191917789196,
+      "grad_norm": 5.5573796484820015,
+      "learning_rate": 2.5069457388541883e-07,
+      "loss": 0.4881,
+      "step": 12988
+    },
+    {
+      "epoch": 0.9018886265796418,
+      "grad_norm": 2.868008636470154,
+      "learning_rate": 2.5034309804836454e-07,
+      "loss": 0.2633,
+      "step": 12989
+    },
+    {
+      "epoch": 0.9019580613803638,
+      "grad_norm": 2.755381922612412,
+      "learning_rate": 2.4999186244358153e-07,
+      "loss": 0.1875,
+      "step": 12990
+    },
+    {
+      "epoch": 0.9020274961810859,
+      "grad_norm": 6.203060277131062,
+      "learning_rate": 2.4964086708883674e-07,
+      "loss": 0.4863,
+      "step": 12991
+    },
+    {
+      "epoch": 0.9020969309818081,
+      "grad_norm": 5.574116650970451,
+      "learning_rate": 2.4929011200187926e-07,
+      "loss": 0.8191,
+      "step": 12992
+    },
+    {
+      "epoch": 0.9021663657825302,
+      "grad_norm": 4.696648482417634,
+      "learning_rate": 2.4893959720045323e-07,
+      "loss": 0.5019,
+      "step": 12993
+    },
+    {
+      "epoch": 0.9022358005832524,
+      "grad_norm": 3.4360410094352685,
+      "learning_rate": 2.485893227022862e-07,
+      "loss": 0.3141,
+      "step": 12994
+    },
+    {
+      "epoch": 0.9023052353839744,
+      "grad_norm": 4.0729095214234405,
+      "learning_rate": 2.4823928852509506e-07,
+      "loss": 0.4624,
+      "step": 12995
+    },
+    {
+      "epoch": 0.9023746701846965,
+      "grad_norm": 3.207610360015554,
+      "learning_rate": 2.478894946865829e-07,
+      "loss": 0.3684,
+      "step": 12996
+    },
+    {
+      "epoch": 0.9024441049854187,
+      "grad_norm": 6.801288089257314,
+      "learning_rate": 2.4753994120444326e-07,
+      "loss": 0.492,
+      "step": 12997
+    },
+    {
+      "epoch": 0.9025135397861408,
+      "grad_norm": 4.551896710699016,
+      "learning_rate": 2.4719062809635543e-07,
+      "loss": 0.5121,
+      "step": 12998
+    },
+    {
+      "epoch": 0.902582974586863,
+      "grad_norm": 2.589783132566977,
+      "learning_rate": 2.4684155537998743e-07,
+      "loss": 0.1387,
+      "step": 12999
+    },
+    {
+      "epoch": 0.9026524093875851,
+      "grad_norm": 3.5143753968419453,
+      "learning_rate": 2.464927230729952e-07,
+      "loss": 0.3518,
+      "step": 13000
+    },
+    {
+      "epoch": 0.9027218441883071,
+      "grad_norm": 4.704342507227891,
+      "learning_rate": 2.461441311930213e-07,
+      "loss": 0.4438,
+      "step": 13001
+    },
+    {
+      "epoch": 0.9027912789890293,
+      "grad_norm": 3.601530773408349,
+      "learning_rate": 2.457957797576993e-07,
+      "loss": 0.3508,
+      "step": 13002
+    },
+    {
+      "epoch": 0.9028607137897514,
+      "grad_norm": 4.9375105836349915,
+      "learning_rate": 2.454476687846463e-07,
+      "loss": 0.5646,
+      "step": 13003
+    },
+    {
+      "epoch": 0.9029301485904735,
+      "grad_norm": 3.7277551838373664,
+      "learning_rate": 2.4509979829146933e-07,
+      "loss": 0.4147,
+      "step": 13004
+    },
+    {
+      "epoch": 0.9029995833911957,
+      "grad_norm": 5.155712573765384,
+      "learning_rate": 2.447521682957654e-07,
+      "loss": 0.6069,
+      "step": 13005
+    },
+    {
+      "epoch": 0.9030690181919178,
+      "grad_norm": 3.0750698859186616,
+      "learning_rate": 2.444047788151144e-07,
+      "loss": 0.2836,
+      "step": 13006
+    },
+    {
+      "epoch": 0.90313845299264,
+      "grad_norm": 3.2028164443671314,
+      "learning_rate": 2.4405762986708937e-07,
+      "loss": 0.2498,
+      "step": 13007
+    },
+    {
+      "epoch": 0.903207887793362,
+      "grad_norm": 3.4606134672905893,
+      "learning_rate": 2.4371072146924744e-07,
+      "loss": 0.4066,
+      "step": 13008
+    },
+    {
+      "epoch": 0.9032773225940841,
+      "grad_norm": 3.6699471677955393,
+      "learning_rate": 2.433640536391352e-07,
+      "loss": 0.4276,
+      "step": 13009
+    },
+    {
+      "epoch": 0.9033467573948063,
+      "grad_norm": 4.730326885375009,
+      "learning_rate": 2.430176263942863e-07,
+      "loss": 0.4253,
+      "step": 13010
+    },
+    {
+      "epoch": 0.9034161921955284,
+      "grad_norm": 4.0011606825698145,
+      "learning_rate": 2.426714397522234e-07,
+      "loss": 0.4129,
+      "step": 13011
+    },
+    {
+      "epoch": 0.9034856269962506,
+      "grad_norm": 5.757822786117658,
+      "learning_rate": 2.423254937304553e-07,
+      "loss": 0.469,
+      "step": 13012
+    },
+    {
+      "epoch": 0.9035550617969726,
+      "grad_norm": 3.7449379065922175,
+      "learning_rate": 2.419797883464808e-07,
+      "loss": 0.3974,
+      "step": 13013
+    },
+    {
+      "epoch": 0.9036244965976947,
+      "grad_norm": 3.6885168625379645,
+      "learning_rate": 2.4163432361778426e-07,
+      "loss": 0.3684,
+      "step": 13014
+    },
+    {
+      "epoch": 0.9036939313984169,
+      "grad_norm": 4.506636005222762,
+      "learning_rate": 2.412890995618383e-07,
+      "loss": 0.533,
+      "step": 13015
+    },
+    {
+      "epoch": 0.903763366199139,
+      "grad_norm": 4.506640079825281,
+      "learning_rate": 2.4094411619610627e-07,
+      "loss": 0.6252,
+      "step": 13016
+    },
+    {
+      "epoch": 0.9038328009998612,
+      "grad_norm": 4.143794407553553,
+      "learning_rate": 2.4059937353803467e-07,
+      "loss": 0.678,
+      "step": 13017
+    },
+    {
+      "epoch": 0.9039022358005833,
+      "grad_norm": 3.115126398988296,
+      "learning_rate": 2.4025487160506066e-07,
+      "loss": 0.3779,
+      "step": 13018
+    },
+    {
+      "epoch": 0.9039716706013053,
+      "grad_norm": 4.01327067181937,
+      "learning_rate": 2.3991061041461037e-07,
+      "loss": 0.4553,
+      "step": 13019
+    },
+    {
+      "epoch": 0.9040411054020275,
+      "grad_norm": 4.360936113192454,
+      "learning_rate": 2.395665899840949e-07,
+      "loss": 0.2577,
+      "step": 13020
+    },
+    {
+      "epoch": 0.9041105402027496,
+      "grad_norm": 4.462828322160175,
+      "learning_rate": 2.3922281033091356e-07,
+      "loss": 0.527,
+      "step": 13021
+    },
+    {
+      "epoch": 0.9041799750034717,
+      "grad_norm": 3.036306935145937,
+      "learning_rate": 2.388792714724564e-07,
+      "loss": 0.2916,
+      "step": 13022
+    },
+    {
+      "epoch": 0.9042494098041939,
+      "grad_norm": 4.919271859926876,
+      "learning_rate": 2.3853597342609836e-07,
+      "loss": 0.5928,
+      "step": 13023
+    },
+    {
+      "epoch": 0.904318844604916,
+      "grad_norm": 3.669787228989885,
+      "learning_rate": 2.3819291620920116e-07,
+      "loss": 0.393,
+      "step": 13024
+    },
+    {
+      "epoch": 0.9043882794056382,
+      "grad_norm": 4.323014778250983,
+      "learning_rate": 2.3785009983911866e-07,
+      "loss": 0.5831,
+      "step": 13025
+    },
+    {
+      "epoch": 0.9044577142063602,
+      "grad_norm": 2.6593313664283715,
+      "learning_rate": 2.3750752433318979e-07,
+      "loss": 0.2034,
+      "step": 13026
+    },
+    {
+      "epoch": 0.9045271490070823,
+      "grad_norm": 3.111678345988155,
+      "learning_rate": 2.3716518970874126e-07,
+      "loss": 0.2414,
+      "step": 13027
+    },
+    {
+      "epoch": 0.9045965838078045,
+      "grad_norm": 4.1727674296138595,
+      "learning_rate": 2.368230959830875e-07,
+      "loss": 0.425,
+      "step": 13028
+    },
+    {
+      "epoch": 0.9046660186085266,
+      "grad_norm": 3.1757675516671635,
+      "learning_rate": 2.3648124317353195e-07,
+      "loss": 0.4538,
+      "step": 13029
+    },
+    {
+      "epoch": 0.9047354534092488,
+      "grad_norm": 4.800423717454337,
+      "learning_rate": 2.361396312973646e-07,
+      "loss": 0.3738,
+      "step": 13030
+    },
+    {
+      "epoch": 0.9048048882099708,
+      "grad_norm": 3.7757961157067164,
+      "learning_rate": 2.3579826037186392e-07,
+      "loss": 0.4477,
+      "step": 13031
+    },
+    {
+      "epoch": 0.9048743230106929,
+      "grad_norm": 4.90570741745172,
+      "learning_rate": 2.3545713041429664e-07,
+      "loss": 0.4406,
+      "step": 13032
+    },
+    {
+      "epoch": 0.9049437578114151,
+      "grad_norm": 3.106769926054841,
+      "learning_rate": 2.3511624144191503e-07,
+      "loss": 0.2469,
+      "step": 13033
+    },
+    {
+      "epoch": 0.9050131926121372,
+      "grad_norm": 5.747091596601599,
+      "learning_rate": 2.347755934719642e-07,
+      "loss": 0.4732,
+      "step": 13034
+    },
+    {
+      "epoch": 0.9050826274128593,
+      "grad_norm": 3.407454672069645,
+      "learning_rate": 2.3443518652166986e-07,
+      "loss": 0.2804,
+      "step": 13035
+    },
+    {
+      "epoch": 0.9051520622135815,
+      "grad_norm": 4.37999867356424,
+      "learning_rate": 2.340950206082515e-07,
+      "loss": 0.4581,
+      "step": 13036
+    },
+    {
+      "epoch": 0.9052214970143035,
+      "grad_norm": 3.8709224997868046,
+      "learning_rate": 2.3375509574891487e-07,
+      "loss": 0.5326,
+      "step": 13037
+    },
+    {
+      "epoch": 0.9052909318150257,
+      "grad_norm": 3.3948751929757144,
+      "learning_rate": 2.3341541196085116e-07,
+      "loss": 0.351,
+      "step": 13038
+    },
+    {
+      "epoch": 0.9053603666157478,
+      "grad_norm": 3.492627929989953,
+      "learning_rate": 2.3307596926124276e-07,
+      "loss": 0.2733,
+      "step": 13039
+    },
+    {
+      "epoch": 0.9054298014164699,
+      "grad_norm": 4.040892140245462,
+      "learning_rate": 2.3273676766725762e-07,
+      "loss": 0.5133,
+      "step": 13040
+    },
+    {
+      "epoch": 0.9054992362171921,
+      "grad_norm": 4.600885641962837,
+      "learning_rate": 2.3239780719605197e-07,
+      "loss": 0.6244,
+      "step": 13041
+    },
+    {
+      "epoch": 0.9055686710179142,
+      "grad_norm": 4.556268243693032,
+      "learning_rate": 2.320590878647705e-07,
+      "loss": 0.6481,
+      "step": 13042
+    },
+    {
+      "epoch": 0.9056381058186364,
+      "grad_norm": 4.781475339631384,
+      "learning_rate": 2.3172060969054499e-07,
+      "loss": 0.437,
+      "step": 13043
+    },
+    {
+      "epoch": 0.9057075406193584,
+      "grad_norm": 3.626838532706603,
+      "learning_rate": 2.3138237269049456e-07,
+      "loss": 0.3012,
+      "step": 13044
+    },
+    {
+      "epoch": 0.9057769754200805,
+      "grad_norm": 4.613560299179202,
+      "learning_rate": 2.3104437688172943e-07,
+      "loss": 0.4228,
+      "step": 13045
+    },
+    {
+      "epoch": 0.9058464102208027,
+      "grad_norm": 4.107885776825883,
+      "learning_rate": 2.3070662228134256e-07,
+      "loss": 0.1829,
+      "step": 13046
+    },
+    {
+      "epoch": 0.9059158450215248,
+      "grad_norm": 3.929738844200927,
+      "learning_rate": 2.3036910890641696e-07,
+      "loss": 0.492,
+      "step": 13047
+    },
+    {
+      "epoch": 0.9059852798222469,
+      "grad_norm": 4.126676063574219,
+      "learning_rate": 2.300318367740262e-07,
+      "loss": 0.3411,
+      "step": 13048
+    },
+    {
+      "epoch": 0.906054714622969,
+      "grad_norm": 3.480877673696189,
+      "learning_rate": 2.2969480590122662e-07,
+      "loss": 0.2988,
+      "step": 13049
+    },
+    {
+      "epoch": 0.9061241494236911,
+      "grad_norm": 4.680954651550427,
+      "learning_rate": 2.293580163050657e-07,
+      "loss": 0.4577,
+      "step": 13050
+    },
+    {
+      "epoch": 0.9061935842244133,
+      "grad_norm": 4.964504499439824,
+      "learning_rate": 2.2902146800257818e-07,
+      "loss": 0.3604,
+      "step": 13051
+    },
+    {
+      "epoch": 0.9062630190251354,
+      "grad_norm": 5.677240729048436,
+      "learning_rate": 2.2868516101078653e-07,
+      "loss": 0.3269,
+      "step": 13052
+    },
+    {
+      "epoch": 0.9063324538258575,
+      "grad_norm": 4.475095451962496,
+      "learning_rate": 2.2834909534669935e-07,
+      "loss": 0.1831,
+      "step": 13053
+    },
+    {
+      "epoch": 0.9064018886265797,
+      "grad_norm": 3.378598009062433,
+      "learning_rate": 2.2801327102731586e-07,
+      "loss": 0.2614,
+      "step": 13054
+    },
+    {
+      "epoch": 0.9064713234273017,
+      "grad_norm": 3.9248998180960855,
+      "learning_rate": 2.276776880696213e-07,
+      "loss": 0.2751,
+      "step": 13055
+    },
+    {
+      "epoch": 0.9065407582280239,
+      "grad_norm": 5.064390245017869,
+      "learning_rate": 2.273423464905894e-07,
+      "loss": 0.6056,
+      "step": 13056
+    },
+    {
+      "epoch": 0.906610193028746,
+      "grad_norm": 4.180137343380059,
+      "learning_rate": 2.2700724630718095e-07,
+      "loss": 0.2763,
+      "step": 13057
+    },
+    {
+      "epoch": 0.9066796278294681,
+      "grad_norm": 2.8921803899519465,
+      "learning_rate": 2.2667238753634468e-07,
+      "loss": 0.2479,
+      "step": 13058
+    },
+    {
+      "epoch": 0.9067490626301903,
+      "grad_norm": 3.846601768127386,
+      "learning_rate": 2.2633777019501757e-07,
+      "loss": 0.3874,
+      "step": 13059
+    },
+    {
+      "epoch": 0.9068184974309124,
+      "grad_norm": 4.36868718087222,
+      "learning_rate": 2.260033943001244e-07,
+      "loss": 0.3805,
+      "step": 13060
+    },
+    {
+      "epoch": 0.9068879322316344,
+      "grad_norm": 3.8000645180588597,
+      "learning_rate": 2.2566925986857725e-07,
+      "loss": 0.3017,
+      "step": 13061
+    },
+    {
+      "epoch": 0.9069573670323566,
+      "grad_norm": 3.397392851789259,
+      "learning_rate": 2.2533536691727697e-07,
+      "loss": 0.3284,
+      "step": 13062
+    },
+    {
+      "epoch": 0.9070268018330787,
+      "grad_norm": 3.881401928406966,
+      "learning_rate": 2.2500171546311068e-07,
+      "loss": 0.3942,
+      "step": 13063
+    },
+    {
+      "epoch": 0.9070962366338009,
+      "grad_norm": 3.1282542593337404,
+      "learning_rate": 2.2466830552295426e-07,
+      "loss": 0.2871,
+      "step": 13064
+    },
+    {
+      "epoch": 0.907165671434523,
+      "grad_norm": 2.878897714053635,
+      "learning_rate": 2.243351371136704e-07,
+      "loss": 0.2043,
+      "step": 13065
+    },
+    {
+      "epoch": 0.9072351062352451,
+      "grad_norm": 4.720691268587579,
+      "learning_rate": 2.2400221025211278e-07,
+      "loss": 0.4706,
+      "step": 13066
+    },
+    {
+      "epoch": 0.9073045410359672,
+      "grad_norm": 3.4750226630922,
+      "learning_rate": 2.2366952495511796e-07,
+      "loss": 0.3156,
+      "step": 13067
+    },
+    {
+      "epoch": 0.9073739758366893,
+      "grad_norm": 2.057385681840549,
+      "learning_rate": 2.2333708123951414e-07,
+      "loss": 0.2007,
+      "step": 13068
+    },
+    {
+      "epoch": 0.9074434106374115,
+      "grad_norm": 4.938059011505567,
+      "learning_rate": 2.2300487912211623e-07,
+      "loss": 0.557,
+      "step": 13069
+    },
+    {
+      "epoch": 0.9075128454381336,
+      "grad_norm": 4.547617072396145,
+      "learning_rate": 2.2267291861972463e-07,
+      "loss": 0.5346,
+      "step": 13070
+    },
+    {
+      "epoch": 0.9075822802388557,
+      "grad_norm": 5.1877502980555406,
+      "learning_rate": 2.2234119974913148e-07,
+      "loss": 0.7655,
+      "step": 13071
+    },
+    {
+      "epoch": 0.9076517150395779,
+      "grad_norm": 9.408937863179526,
+      "learning_rate": 2.220097225271145e-07,
+      "loss": 0.3667,
+      "step": 13072
+    },
+    {
+      "epoch": 0.9077211498403,
+      "grad_norm": 3.5700303734613437,
+      "learning_rate": 2.2167848697043858e-07,
+      "loss": 0.3974,
+      "step": 13073
+    },
+    {
+      "epoch": 0.9077905846410221,
+      "grad_norm": 3.1885770264765627,
+      "learning_rate": 2.2134749309585813e-07,
+      "loss": 0.2297,
+      "step": 13074
+    },
+    {
+      "epoch": 0.9078600194417442,
+      "grad_norm": 4.034696146134198,
+      "learning_rate": 2.2101674092011417e-07,
+      "loss": 0.4953,
+      "step": 13075
+    },
+    {
+      "epoch": 0.9079294542424663,
+      "grad_norm": 4.090098010251932,
+      "learning_rate": 2.2068623045993498e-07,
+      "loss": 0.4153,
+      "step": 13076
+    },
+    {
+      "epoch": 0.9079988890431885,
+      "grad_norm": 3.391655356363241,
+      "learning_rate": 2.203559617320389e-07,
+      "loss": 0.3102,
+      "step": 13077
+    },
+    {
+      "epoch": 0.9080683238439106,
+      "grad_norm": 3.079239633772833,
+      "learning_rate": 2.2002593475312972e-07,
+      "loss": 0.2552,
+      "step": 13078
+    },
+    {
+      "epoch": 0.9081377586446326,
+      "grad_norm": 2.6520299463217714,
+      "learning_rate": 2.1969614953989914e-07,
+      "loss": 0.2322,
+      "step": 13079
+    },
+    {
+      "epoch": 0.9082071934453548,
+      "grad_norm": 4.470167912877859,
+      "learning_rate": 2.1936660610902882e-07,
+      "loss": 0.4205,
+      "step": 13080
+    },
+    {
+      "epoch": 0.9082766282460769,
+      "grad_norm": 3.903469663391334,
+      "learning_rate": 2.1903730447718596e-07,
+      "loss": 0.3695,
+      "step": 13081
+    },
+    {
+      "epoch": 0.9083460630467991,
+      "grad_norm": 4.3259627401575536,
+      "learning_rate": 2.1870824466102503e-07,
+      "loss": 0.49,
+      "step": 13082
+    },
+    {
+      "epoch": 0.9084154978475212,
+      "grad_norm": 4.824051317464986,
+      "learning_rate": 2.1837942667719157e-07,
+      "loss": 0.5019,
+      "step": 13083
+    },
+    {
+      "epoch": 0.9084849326482433,
+      "grad_norm": 4.041752634741632,
+      "learning_rate": 2.1805085054231622e-07,
+      "loss": 0.4854,
+      "step": 13084
+    },
+    {
+      "epoch": 0.9085543674489654,
+      "grad_norm": 6.5575539073310205,
+      "learning_rate": 2.1772251627301676e-07,
+      "loss": 0.4013,
+      "step": 13085
+    },
+    {
+      "epoch": 0.9086238022496875,
+      "grad_norm": 4.611684692420382,
+      "learning_rate": 2.173944238859016e-07,
+      "loss": 0.5457,
+      "step": 13086
+    },
+    {
+      "epoch": 0.9086932370504097,
+      "grad_norm": 3.72698208210969,
+      "learning_rate": 2.1706657339756466e-07,
+      "loss": 0.3076,
+      "step": 13087
+    },
+    {
+      "epoch": 0.9087626718511318,
+      "grad_norm": 3.5867654864515948,
+      "learning_rate": 2.1673896482458768e-07,
+      "loss": 0.4468,
+      "step": 13088
+    },
+    {
+      "epoch": 0.9088321066518539,
+      "grad_norm": 4.879969154658898,
+      "learning_rate": 2.1641159818354184e-07,
+      "loss": 0.5419,
+      "step": 13089
+    },
+    {
+      "epoch": 0.9089015414525761,
+      "grad_norm": 3.1424057618503256,
+      "learning_rate": 2.1608447349098393e-07,
+      "loss": 0.2672,
+      "step": 13090
+    },
+    {
+      "epoch": 0.9089709762532981,
+      "grad_norm": 3.3077327260424587,
+      "learning_rate": 2.1575759076346014e-07,
+      "loss": 0.3627,
+      "step": 13091
+    },
+    {
+      "epoch": 0.9090404110540202,
+      "grad_norm": 3.2595704214147245,
+      "learning_rate": 2.154309500175039e-07,
+      "loss": 0.3339,
+      "step": 13092
+    },
+    {
+      "epoch": 0.9091098458547424,
+      "grad_norm": 3.1818829550765457,
+      "learning_rate": 2.1510455126963592e-07,
+      "loss": 0.3083,
+      "step": 13093
+    },
+    {
+      "epoch": 0.9091792806554645,
+      "grad_norm": 3.6306790361797634,
+      "learning_rate": 2.1477839453636518e-07,
+      "loss": 0.4544,
+      "step": 13094
+    },
+    {
+      "epoch": 0.9092487154561867,
+      "grad_norm": 4.090606769185803,
+      "learning_rate": 2.144524798341885e-07,
+      "loss": 0.4622,
+      "step": 13095
+    },
+    {
+      "epoch": 0.9093181502569088,
+      "grad_norm": 4.656064699023622,
+      "learning_rate": 2.1412680717958934e-07,
+      "loss": 0.5533,
+      "step": 13096
+    },
+    {
+      "epoch": 0.9093875850576308,
+      "grad_norm": 4.181148433151871,
+      "learning_rate": 2.1380137658904176e-07,
+      "loss": 0.2659,
+      "step": 13097
+    },
+    {
+      "epoch": 0.909457019858353,
+      "grad_norm": 3.0973463079305166,
+      "learning_rate": 2.1347618807900482e-07,
+      "loss": 0.3757,
+      "step": 13098
+    },
+    {
+      "epoch": 0.9095264546590751,
+      "grad_norm": 4.124002180933459,
+      "learning_rate": 2.1315124166592484e-07,
+      "loss": 0.5821,
+      "step": 13099
+    },
+    {
+      "epoch": 0.9095958894597973,
+      "grad_norm": 5.183879174023916,
+      "learning_rate": 2.1282653736623914e-07,
+      "loss": 0.5093,
+      "step": 13100
+    },
+    {
+      "epoch": 0.9096653242605194,
+      "grad_norm": 5.2218200376180866,
+      "learning_rate": 2.1250207519636967e-07,
+      "loss": 0.5617,
+      "step": 13101
+    },
+    {
+      "epoch": 0.9097347590612415,
+      "grad_norm": 4.970940100298773,
+      "learning_rate": 2.1217785517272827e-07,
+      "loss": 0.5363,
+      "step": 13102
+    },
+    {
+      "epoch": 0.9098041938619637,
+      "grad_norm": 3.6929497502478412,
+      "learning_rate": 2.1185387731171293e-07,
+      "loss": 0.4188,
+      "step": 13103
+    },
+    {
+      "epoch": 0.9098736286626857,
+      "grad_norm": 3.987627841430215,
+      "learning_rate": 2.1153014162971052e-07,
+      "loss": 0.2798,
+      "step": 13104
+    },
+    {
+      "epoch": 0.9099430634634078,
+      "grad_norm": 4.979498710588045,
+      "learning_rate": 2.1120664814309522e-07,
+      "loss": 0.5792,
+      "step": 13105
+    },
+    {
+      "epoch": 0.91001249826413,
+      "grad_norm": 3.6260793420066806,
+      "learning_rate": 2.1088339686822834e-07,
+      "loss": 0.4317,
+      "step": 13106
+    },
+    {
+      "epoch": 0.9100819330648521,
+      "grad_norm": 3.912139358259877,
+      "learning_rate": 2.1056038782146015e-07,
+      "loss": 0.2036,
+      "step": 13107
+    },
+    {
+      "epoch": 0.9101513678655743,
+      "grad_norm": 4.587462412676917,
+      "learning_rate": 2.1023762101912704e-07,
+      "loss": 0.3209,
+      "step": 13108
+    },
+    {
+      "epoch": 0.9102208026662963,
+      "grad_norm": 3.7109577291065663,
+      "learning_rate": 2.0991509647755703e-07,
+      "loss": 0.368,
+      "step": 13109
+    },
+    {
+      "epoch": 0.9102902374670184,
+      "grad_norm": 3.0770212384633875,
+      "learning_rate": 2.095928142130599e-07,
+      "loss": 0.2124,
+      "step": 13110
+    },
+    {
+      "epoch": 0.9103596722677406,
+      "grad_norm": 4.305743383728951,
+      "learning_rate": 2.0927077424193698e-07,
+      "loss": 0.418,
+      "step": 13111
+    },
+    {
+      "epoch": 0.9104291070684627,
+      "grad_norm": 3.619542197417392,
+      "learning_rate": 2.0894897658047863e-07,
+      "loss": 0.4248,
+      "step": 13112
+    },
+    {
+      "epoch": 0.9104985418691849,
+      "grad_norm": 3.12456480609724,
+      "learning_rate": 2.0862742124495849e-07,
+      "loss": 0.3744,
+      "step": 13113
+    },
+    {
+      "epoch": 0.910567976669907,
+      "grad_norm": 2.979345355795465,
+      "learning_rate": 2.0830610825164132e-07,
+      "loss": 0.2418,
+      "step": 13114
+    },
+    {
+      "epoch": 0.910637411470629,
+      "grad_norm": 4.92449446593579,
+      "learning_rate": 2.079850376167797e-07,
+      "loss": 0.5774,
+      "step": 13115
+    },
+    {
+      "epoch": 0.9107068462713512,
+      "grad_norm": 4.43583426950556,
+      "learning_rate": 2.0766420935661224e-07,
+      "loss": 0.6745,
+      "step": 13116
+    },
+    {
+      "epoch": 0.9107762810720733,
+      "grad_norm": 4.486465249955106,
+      "learning_rate": 2.0734362348736602e-07,
+      "loss": 0.5872,
+      "step": 13117
+    },
+    {
+      "epoch": 0.9108457158727954,
+      "grad_norm": 5.384097997897376,
+      "learning_rate": 2.070232800252564e-07,
+      "loss": 0.5412,
+      "step": 13118
+    },
+    {
+      "epoch": 0.9109151506735176,
+      "grad_norm": 4.35103971016548,
+      "learning_rate": 2.0670317898648539e-07,
+      "loss": 0.3021,
+      "step": 13119
+    },
+    {
+      "epoch": 0.9109845854742397,
+      "grad_norm": 4.31781133386301,
+      "learning_rate": 2.0638332038724395e-07,
+      "loss": 0.4234,
+      "step": 13120
+    },
+    {
+      "epoch": 0.9110540202749619,
+      "grad_norm": 4.731078052258188,
+      "learning_rate": 2.060637042437097e-07,
+      "loss": 0.5315,
+      "step": 13121
+    },
+    {
+      "epoch": 0.9111234550756839,
+      "grad_norm": 5.107630053721914,
+      "learning_rate": 2.0574433057204857e-07,
+      "loss": 0.448,
+      "step": 13122
+    },
+    {
+      "epoch": 0.911192889876406,
+      "grad_norm": 3.2288047178390507,
+      "learning_rate": 2.0542519938841377e-07,
+      "loss": 0.3396,
+      "step": 13123
+    },
+    {
+      "epoch": 0.9112623246771282,
+      "grad_norm": 2.904088071258934,
+      "learning_rate": 2.0510631070894738e-07,
+      "loss": 0.1951,
+      "step": 13124
+    },
+    {
+      "epoch": 0.9113317594778503,
+      "grad_norm": 4.206015643693509,
+      "learning_rate": 2.0478766454977705e-07,
+      "loss": 0.2977,
+      "step": 13125
+    },
+    {
+      "epoch": 0.9114011942785725,
+      "grad_norm": 4.458746333437605,
+      "learning_rate": 2.0446926092702157e-07,
+      "loss": 0.5085,
+      "step": 13126
+    },
+    {
+      "epoch": 0.9114706290792945,
+      "grad_norm": 4.213553729978643,
+      "learning_rate": 2.041510998567847e-07,
+      "loss": 0.5041,
+      "step": 13127
+    },
+    {
+      "epoch": 0.9115400638800166,
+      "grad_norm": 3.7938070538951405,
+      "learning_rate": 2.0383318135515696e-07,
+      "loss": 0.3449,
+      "step": 13128
+    },
+    {
+      "epoch": 0.9116094986807388,
+      "grad_norm": 4.281740109494131,
+      "learning_rate": 2.035155054382204e-07,
+      "loss": 0.5241,
+      "step": 13129
+    },
+    {
+      "epoch": 0.9116789334814609,
+      "grad_norm": 5.480033043775263,
+      "learning_rate": 2.031980721220428e-07,
+      "loss": 0.3525,
+      "step": 13130
+    },
+    {
+      "epoch": 0.911748368282183,
+      "grad_norm": 4.235983615570748,
+      "learning_rate": 2.0288088142267738e-07,
+      "loss": 0.4991,
+      "step": 13131
+    },
+    {
+      "epoch": 0.9118178030829052,
+      "grad_norm": 4.896283107934131,
+      "learning_rate": 2.0256393335616908e-07,
+      "loss": 0.4829,
+      "step": 13132
+    },
+    {
+      "epoch": 0.9118872378836272,
+      "grad_norm": 4.217037722044566,
+      "learning_rate": 2.02247227938549e-07,
+      "loss": 0.4453,
+      "step": 13133
+    },
+    {
+      "epoch": 0.9119566726843494,
+      "grad_norm": 3.6659673916493296,
+      "learning_rate": 2.0193076518583432e-07,
+      "loss": 0.3288,
+      "step": 13134
+    },
+    {
+      "epoch": 0.9120261074850715,
+      "grad_norm": 3.6400615594539003,
+      "learning_rate": 2.0161454511403278e-07,
+      "loss": 0.2917,
+      "step": 13135
+    },
+    {
+      "epoch": 0.9120955422857936,
+      "grad_norm": 5.996333766272976,
+      "learning_rate": 2.012985677391377e-07,
+      "loss": 0.7983,
+      "step": 13136
+    },
+    {
+      "epoch": 0.9121649770865158,
+      "grad_norm": 5.320652772708727,
+      "learning_rate": 2.0098283307713074e-07,
+      "loss": 0.7372,
+      "step": 13137
+    },
+    {
+      "epoch": 0.9122344118872379,
+      "grad_norm": 4.41517487059579,
+      "learning_rate": 2.0066734114398133e-07,
+      "loss": 0.3472,
+      "step": 13138
+    },
+    {
+      "epoch": 0.91230384668796,
+      "grad_norm": 3.077609143357197,
+      "learning_rate": 2.003520919556473e-07,
+      "loss": 0.2265,
+      "step": 13139
+    },
+    {
+      "epoch": 0.9123732814886821,
+      "grad_norm": 4.421108047239686,
+      "learning_rate": 2.000370855280731e-07,
+      "loss": 0.3491,
+      "step": 13140
+    },
+    {
+      "epoch": 0.9124427162894042,
+      "grad_norm": 3.770967885855413,
+      "learning_rate": 1.997223218771921e-07,
+      "loss": 0.429,
+      "step": 13141
+    },
+    {
+      "epoch": 0.9125121510901264,
+      "grad_norm": 3.767351199194019,
+      "learning_rate": 1.9940780101892433e-07,
+      "loss": 0.288,
+      "step": 13142
+    },
+    {
+      "epoch": 0.9125815858908485,
+      "grad_norm": 4.04283038612596,
+      "learning_rate": 1.990935229691765e-07,
+      "loss": 0.3414,
+      "step": 13143
+    },
+    {
+      "epoch": 0.9126510206915707,
+      "grad_norm": 5.234372040102413,
+      "learning_rate": 1.9877948774384703e-07,
+      "loss": 0.456,
+      "step": 13144
+    },
+    {
+      "epoch": 0.9127204554922927,
+      "grad_norm": 4.301840014487488,
+      "learning_rate": 1.9846569535881767e-07,
+      "loss": 0.3411,
+      "step": 13145
+    },
+    {
+      "epoch": 0.9127898902930148,
+      "grad_norm": 4.345692905338997,
+      "learning_rate": 1.9815214582995955e-07,
+      "loss": 0.4133,
+      "step": 13146
+    },
+    {
+      "epoch": 0.912859325093737,
+      "grad_norm": 4.620465126924886,
+      "learning_rate": 1.9783883917313285e-07,
+      "loss": 0.6071,
+      "step": 13147
+    },
+    {
+      "epoch": 0.9129287598944591,
+      "grad_norm": 2.7182267876185304,
+      "learning_rate": 1.9752577540418315e-07,
+      "loss": 0.169,
+      "step": 13148
+    },
+    {
+      "epoch": 0.9129981946951812,
+      "grad_norm": 3.59505303118981,
+      "learning_rate": 1.9721295453894618e-07,
+      "loss": 0.3005,
+      "step": 13149
+    },
+    {
+      "epoch": 0.9130676294959034,
+      "grad_norm": 4.129588110346204,
+      "learning_rate": 1.9690037659324256e-07,
+      "loss": 0.4218,
+      "step": 13150
+    },
+    {
+      "epoch": 0.9131370642966254,
+      "grad_norm": 3.874565154033344,
+      "learning_rate": 1.9658804158288303e-07,
+      "loss": 0.5208,
+      "step": 13151
+    },
+    {
+      "epoch": 0.9132064990973476,
+      "grad_norm": 4.232318982492772,
+      "learning_rate": 1.9627594952366491e-07,
+      "loss": 0.5539,
+      "step": 13152
+    },
+    {
+      "epoch": 0.9132759338980697,
+      "grad_norm": 3.786113284056205,
+      "learning_rate": 1.9596410043137393e-07,
+      "loss": 0.4187,
+      "step": 13153
+    },
+    {
+      "epoch": 0.9133453686987918,
+      "grad_norm": 4.288814734476795,
+      "learning_rate": 1.9565249432178192e-07,
+      "loss": 0.5484,
+      "step": 13154
+    },
+    {
+      "epoch": 0.913414803499514,
+      "grad_norm": 4.6077161340572905,
+      "learning_rate": 1.953411312106507e-07,
+      "loss": 0.581,
+      "step": 13155
+    },
+    {
+      "epoch": 0.9134842383002361,
+      "grad_norm": 3.7002858088458694,
+      "learning_rate": 1.950300111137282e-07,
+      "loss": 0.2884,
+      "step": 13156
+    },
+    {
+      "epoch": 0.9135536731009583,
+      "grad_norm": 3.4228897042453195,
+      "learning_rate": 1.947191340467497e-07,
+      "loss": 0.2982,
+      "step": 13157
+    },
+    {
+      "epoch": 0.9136231079016803,
+      "grad_norm": 4.4247415412960045,
+      "learning_rate": 1.9440850002544086e-07,
+      "loss": 0.5913,
+      "step": 13158
+    },
+    {
+      "epoch": 0.9136925427024024,
+      "grad_norm": 5.161644179865504,
+      "learning_rate": 1.940981090655125e-07,
+      "loss": 0.5296,
+      "step": 13159
+    },
+    {
+      "epoch": 0.9137619775031246,
+      "grad_norm": 4.240518231752646,
+      "learning_rate": 1.9378796118266208e-07,
+      "loss": 0.3994,
+      "step": 13160
+    },
+    {
+      "epoch": 0.9138314123038467,
+      "grad_norm": 4.530987075162484,
+      "learning_rate": 1.9347805639257923e-07,
+      "loss": 0.4318,
+      "step": 13161
+    },
+    {
+      "epoch": 0.9139008471045688,
+      "grad_norm": 4.516957955514719,
+      "learning_rate": 1.9316839471093752e-07,
+      "loss": 0.2844,
+      "step": 13162
+    },
+    {
+      "epoch": 0.913970281905291,
+      "grad_norm": 3.207712536734829,
+      "learning_rate": 1.9285897615339776e-07,
+      "loss": 0.1931,
+      "step": 13163
+    },
+    {
+      "epoch": 0.914039716706013,
+      "grad_norm": 4.03011072165392,
+      "learning_rate": 1.9254980073561246e-07,
+      "loss": 0.5403,
+      "step": 13164
+    },
+    {
+      "epoch": 0.9141091515067352,
+      "grad_norm": 3.709004275165941,
+      "learning_rate": 1.92240868473218e-07,
+      "loss": 0.2588,
+      "step": 13165
+    },
+    {
+      "epoch": 0.9141785863074573,
+      "grad_norm": 4.849142845541423,
+      "learning_rate": 1.919321793818396e-07,
+      "loss": 0.34,
+      "step": 13166
+    },
+    {
+      "epoch": 0.9142480211081794,
+      "grad_norm": 4.836180380826499,
+      "learning_rate": 1.916237334770915e-07,
+      "loss": 0.5642,
+      "step": 13167
+    },
+    {
+      "epoch": 0.9143174559089016,
+      "grad_norm": 3.4192675363392953,
+      "learning_rate": 1.91315530774574e-07,
+      "loss": 0.272,
+      "step": 13168
+    },
+    {
+      "epoch": 0.9143868907096236,
+      "grad_norm": 4.461030118373413,
+      "learning_rate": 1.9100757128987513e-07,
+      "loss": 0.5749,
+      "step": 13169
+    },
+    {
+      "epoch": 0.9144563255103458,
+      "grad_norm": 3.5118586422396962,
+      "learning_rate": 1.9069985503857135e-07,
+      "loss": 0.3602,
+      "step": 13170
+    },
+    {
+      "epoch": 0.9145257603110679,
+      "grad_norm": 4.182774019121627,
+      "learning_rate": 1.9039238203622745e-07,
+      "loss": 0.349,
+      "step": 13171
+    },
+    {
+      "epoch": 0.91459519511179,
+      "grad_norm": 4.905202071340501,
+      "learning_rate": 1.9008515229839318e-07,
+      "loss": 0.3973,
+      "step": 13172
+    },
+    {
+      "epoch": 0.9146646299125122,
+      "grad_norm": 4.413039607281084,
+      "learning_rate": 1.8977816584061115e-07,
+      "loss": 0.6284,
+      "step": 13173
+    },
+    {
+      "epoch": 0.9147340647132343,
+      "grad_norm": 4.085802079557506,
+      "learning_rate": 1.89471422678405e-07,
+      "loss": 0.459,
+      "step": 13174
+    },
+    {
+      "epoch": 0.9148034995139563,
+      "grad_norm": 2.3108712034100463,
+      "learning_rate": 1.8916492282729072e-07,
+      "loss": 0.2271,
+      "step": 13175
+    },
+    {
+      "epoch": 0.9148729343146785,
+      "grad_norm": 4.3812902883413525,
+      "learning_rate": 1.8885866630277083e-07,
+      "loss": 0.4005,
+      "step": 13176
+    },
+    {
+      "epoch": 0.9149423691154006,
+      "grad_norm": 3.862582454130768,
+      "learning_rate": 1.8855265312033577e-07,
+      "loss": 0.483,
+      "step": 13177
+    },
+    {
+      "epoch": 0.9150118039161228,
+      "grad_norm": 3.084282224808126,
+      "learning_rate": 1.8824688329546315e-07,
+      "loss": 0.3058,
+      "step": 13178
+    },
+    {
+      "epoch": 0.9150812387168449,
+      "grad_norm": 2.5293551785501323,
+      "learning_rate": 1.8794135684361837e-07,
+      "loss": 0.1241,
+      "step": 13179
+    },
+    {
+      "epoch": 0.915150673517567,
+      "grad_norm": 4.268197955635075,
+      "learning_rate": 1.8763607378025406e-07,
+      "loss": 0.4241,
+      "step": 13180
+    },
+    {
+      "epoch": 0.9152201083182891,
+      "grad_norm": 3.707227929041444,
+      "learning_rate": 1.8733103412081177e-07,
+      "loss": 0.3503,
+      "step": 13181
+    },
+    {
+      "epoch": 0.9152895431190112,
+      "grad_norm": 5.877329062520276,
+      "learning_rate": 1.8702623788072028e-07,
+      "loss": 0.7088,
+      "step": 13182
+    },
+    {
+      "epoch": 0.9153589779197334,
+      "grad_norm": 3.5167450667373106,
+      "learning_rate": 1.86721685075395e-07,
+      "loss": 0.3157,
+      "step": 13183
+    },
+    {
+      "epoch": 0.9154284127204555,
+      "grad_norm": 4.546791752627194,
+      "learning_rate": 1.8641737572024033e-07,
+      "loss": 0.5028,
+      "step": 13184
+    },
+    {
+      "epoch": 0.9154978475211776,
+      "grad_norm": 4.144392595469462,
+      "learning_rate": 1.8611330983064779e-07,
+      "loss": 0.2806,
+      "step": 13185
+    },
+    {
+      "epoch": 0.9155672823218998,
+      "grad_norm": 4.256470108514353,
+      "learning_rate": 1.8580948742199622e-07,
+      "loss": 0.4336,
+      "step": 13186
+    },
+    {
+      "epoch": 0.9156367171226218,
+      "grad_norm": 4.261859307341919,
+      "learning_rate": 1.8550590850965443e-07,
+      "loss": 0.4324,
+      "step": 13187
+    },
+    {
+      "epoch": 0.9157061519233439,
+      "grad_norm": 3.52455772933577,
+      "learning_rate": 1.8520257310897516e-07,
+      "loss": 0.3491,
+      "step": 13188
+    },
+    {
+      "epoch": 0.9157755867240661,
+      "grad_norm": 4.672633881327756,
+      "learning_rate": 1.848994812353011e-07,
+      "loss": 0.52,
+      "step": 13189
+    },
+    {
+      "epoch": 0.9158450215247882,
+      "grad_norm": 5.053456379977315,
+      "learning_rate": 1.8459663290396278e-07,
+      "loss": 0.2968,
+      "step": 13190
+    },
+    {
+      "epoch": 0.9159144563255104,
+      "grad_norm": 4.199483525488225,
+      "learning_rate": 1.8429402813027853e-07,
+      "loss": 0.5053,
+      "step": 13191
+    },
+    {
+      "epoch": 0.9159838911262325,
+      "grad_norm": 5.206677005417869,
+      "learning_rate": 1.8399166692955217e-07,
+      "loss": 0.5118,
+      "step": 13192
+    },
+    {
+      "epoch": 0.9160533259269545,
+      "grad_norm": 5.094949363906195,
+      "learning_rate": 1.836895493170776e-07,
+      "loss": 0.4364,
+      "step": 13193
+    },
+    {
+      "epoch": 0.9161227607276767,
+      "grad_norm": 4.410648550347962,
+      "learning_rate": 1.833876753081365e-07,
+      "loss": 0.3204,
+      "step": 13194
+    },
+    {
+      "epoch": 0.9161921955283988,
+      "grad_norm": 4.071826634437693,
+      "learning_rate": 1.8308604491799554e-07,
+      "loss": 0.4599,
+      "step": 13195
+    },
+    {
+      "epoch": 0.916261630329121,
+      "grad_norm": 3.327936807243054,
+      "learning_rate": 1.8278465816191194e-07,
+      "loss": 0.2187,
+      "step": 13196
+    },
+    {
+      "epoch": 0.9163310651298431,
+      "grad_norm": 3.4793058211951178,
+      "learning_rate": 1.8248351505512906e-07,
+      "loss": 0.2691,
+      "step": 13197
+    },
+    {
+      "epoch": 0.9164004999305652,
+      "grad_norm": 4.279305454266996,
+      "learning_rate": 1.821826156128792e-07,
+      "loss": 0.444,
+      "step": 13198
+    },
+    {
+      "epoch": 0.9164699347312874,
+      "grad_norm": 5.630757092699069,
+      "learning_rate": 1.818819598503807e-07,
+      "loss": 0.4323,
+      "step": 13199
+    },
+    {
+      "epoch": 0.9165393695320094,
+      "grad_norm": 4.524862603023,
+      "learning_rate": 1.815815477828403e-07,
+      "loss": 0.5411,
+      "step": 13200
+    },
+    {
+      "epoch": 0.9166088043327316,
+      "grad_norm": 3.7393453977510585,
+      "learning_rate": 1.8128137942545254e-07,
+      "loss": 0.3696,
+      "step": 13201
+    },
+    {
+      "epoch": 0.9166782391334537,
+      "grad_norm": 3.694405157551287,
+      "learning_rate": 1.8098145479340078e-07,
+      "loss": 0.3588,
+      "step": 13202
+    },
+    {
+      "epoch": 0.9167476739341758,
+      "grad_norm": 3.2889094398678904,
+      "learning_rate": 1.8068177390185348e-07,
+      "loss": 0.3362,
+      "step": 13203
+    },
+    {
+      "epoch": 0.916817108734898,
+      "grad_norm": 3.7868491226048087,
+      "learning_rate": 1.8038233676596796e-07,
+      "loss": 0.4876,
+      "step": 13204
+    },
+    {
+      "epoch": 0.91688654353562,
+      "grad_norm": 4.347107936584964,
+      "learning_rate": 1.800831434008915e-07,
+      "loss": 0.3629,
+      "step": 13205
+    },
+    {
+      "epoch": 0.9169559783363421,
+      "grad_norm": 3.3288660794735896,
+      "learning_rate": 1.7978419382175428e-07,
+      "loss": 0.2376,
+      "step": 13206
+    },
+    {
+      "epoch": 0.9170254131370643,
+      "grad_norm": 4.357758878373558,
+      "learning_rate": 1.794854880436786e-07,
+      "loss": 0.3279,
+      "step": 13207
+    },
+    {
+      "epoch": 0.9170948479377864,
+      "grad_norm": 3.7916209749486818,
+      "learning_rate": 1.7918702608177186e-07,
+      "loss": 0.3335,
+      "step": 13208
+    },
+    {
+      "epoch": 0.9171642827385086,
+      "grad_norm": 4.068544470916159,
+      "learning_rate": 1.7888880795113083e-07,
+      "loss": 0.6633,
+      "step": 13209
+    },
+    {
+      "epoch": 0.9172337175392307,
+      "grad_norm": 4.598300631864856,
+      "learning_rate": 1.7859083366683793e-07,
+      "loss": 0.5846,
+      "step": 13210
+    },
+    {
+      "epoch": 0.9173031523399527,
+      "grad_norm": 4.877864533407342,
+      "learning_rate": 1.78293103243965e-07,
+      "loss": 0.5308,
+      "step": 13211
+    },
+    {
+      "epoch": 0.9173725871406749,
+      "grad_norm": 4.012021340204592,
+      "learning_rate": 1.7799561669757103e-07,
+      "loss": 0.4508,
+      "step": 13212
+    },
+    {
+      "epoch": 0.917442021941397,
+      "grad_norm": 4.056358268684323,
+      "learning_rate": 1.776983740427024e-07,
+      "loss": 0.456,
+      "step": 13213
+    },
+    {
+      "epoch": 0.9175114567421192,
+      "grad_norm": 4.065072105354813,
+      "learning_rate": 1.774013752943926e-07,
+      "loss": 0.5014,
+      "step": 13214
+    },
+    {
+      "epoch": 0.9175808915428413,
+      "grad_norm": 3.8114492903877513,
+      "learning_rate": 1.771046204676641e-07,
+      "loss": 0.2906,
+      "step": 13215
+    },
+    {
+      "epoch": 0.9176503263435634,
+      "grad_norm": 5.640629438637088,
+      "learning_rate": 1.7680810957752702e-07,
+      "loss": 0.38,
+      "step": 13216
+    },
+    {
+      "epoch": 0.9177197611442856,
+      "grad_norm": 3.9378314589925454,
+      "learning_rate": 1.765118426389778e-07,
+      "loss": 0.3722,
+      "step": 13217
+    },
+    {
+      "epoch": 0.9177891959450076,
+      "grad_norm": 4.866989601091155,
+      "learning_rate": 1.7621581966700051e-07,
+      "loss": 0.4743,
+      "step": 13218
+    },
+    {
+      "epoch": 0.9178586307457297,
+      "grad_norm": 5.976386976635638,
+      "learning_rate": 1.7592004067657041e-07,
+      "loss": 0.7048,
+      "step": 13219
+    },
+    {
+      "epoch": 0.9179280655464519,
+      "grad_norm": 3.2847469562493203,
+      "learning_rate": 1.7562450568264444e-07,
+      "loss": 0.2608,
+      "step": 13220
+    },
+    {
+      "epoch": 0.917997500347174,
+      "grad_norm": 3.9599363592648653,
+      "learning_rate": 1.7532921470017173e-07,
+      "loss": 0.4972,
+      "step": 13221
+    },
+    {
+      "epoch": 0.9180669351478962,
+      "grad_norm": 3.76859785500059,
+      "learning_rate": 1.750341677440881e-07,
+      "loss": 0.3654,
+      "step": 13222
+    },
+    {
+      "epoch": 0.9181363699486182,
+      "grad_norm": 3.584613913306775,
+      "learning_rate": 1.7473936482931718e-07,
+      "loss": 0.3657,
+      "step": 13223
+    },
+    {
+      "epoch": 0.9182058047493403,
+      "grad_norm": 4.819573222505809,
+      "learning_rate": 1.744448059707682e-07,
+      "loss": 0.6046,
+      "step": 13224
+    },
+    {
+      "epoch": 0.9182752395500625,
+      "grad_norm": 3.405697693611392,
+      "learning_rate": 1.7415049118334083e-07,
+      "loss": 0.296,
+      "step": 13225
+    },
+    {
+      "epoch": 0.9183446743507846,
+      "grad_norm": 4.185277550514758,
+      "learning_rate": 1.738564204819204e-07,
+      "loss": 0.4308,
+      "step": 13226
+    },
+    {
+      "epoch": 0.9184141091515068,
+      "grad_norm": 5.09992880037529,
+      "learning_rate": 1.7356259388138176e-07,
+      "loss": 0.4437,
+      "step": 13227
+    },
+    {
+      "epoch": 0.9184835439522289,
+      "grad_norm": 4.694179696395701,
+      "learning_rate": 1.7326901139658514e-07,
+      "loss": 0.5961,
+      "step": 13228
+    },
+    {
+      "epoch": 0.9185529787529509,
+      "grad_norm": 3.5838688843420106,
+      "learning_rate": 1.729756730423804e-07,
+      "loss": 0.3195,
+      "step": 13229
+    },
+    {
+      "epoch": 0.9186224135536731,
+      "grad_norm": 3.6760958948571845,
+      "learning_rate": 1.7268257883360394e-07,
+      "loss": 0.4393,
+      "step": 13230
+    },
+    {
+      "epoch": 0.9186918483543952,
+      "grad_norm": 2.9584001868479812,
+      "learning_rate": 1.723897287850801e-07,
+      "loss": 0.3919,
+      "step": 13231
+    },
+    {
+      "epoch": 0.9187612831551173,
+      "grad_norm": 4.722996632878373,
+      "learning_rate": 1.7209712291162085e-07,
+      "loss": 0.6724,
+      "step": 13232
+    },
+    {
+      "epoch": 0.9188307179558395,
+      "grad_norm": 4.25715981972975,
+      "learning_rate": 1.7180476122802548e-07,
+      "loss": 0.4906,
+      "step": 13233
+    },
+    {
+      "epoch": 0.9189001527565616,
+      "grad_norm": 3.810037760473995,
+      "learning_rate": 1.7151264374908273e-07,
+      "loss": 0.3251,
+      "step": 13234
+    },
+    {
+      "epoch": 0.9189695875572838,
+      "grad_norm": 5.391328664969378,
+      "learning_rate": 1.712207704895663e-07,
+      "loss": 0.7413,
+      "step": 13235
+    },
+    {
+      "epoch": 0.9190390223580058,
+      "grad_norm": 2.7894993101818946,
+      "learning_rate": 1.7092914146423834e-07,
+      "loss": 0.2573,
+      "step": 13236
+    },
+    {
+      "epoch": 0.9191084571587279,
+      "grad_norm": 3.5334720819866305,
+      "learning_rate": 1.7063775668785143e-07,
+      "loss": 0.4482,
+      "step": 13237
+    },
+    {
+      "epoch": 0.9191778919594501,
+      "grad_norm": 3.2130815993870376,
+      "learning_rate": 1.7034661617514102e-07,
+      "loss": 0.3157,
+      "step": 13238
+    },
+    {
+      "epoch": 0.9192473267601722,
+      "grad_norm": 2.9479394920341764,
+      "learning_rate": 1.700557199408337e-07,
+      "loss": 0.2149,
+      "step": 13239
+    },
+    {
+      "epoch": 0.9193167615608944,
+      "grad_norm": 4.382621519179842,
+      "learning_rate": 1.697650679996432e-07,
+      "loss": 0.502,
+      "step": 13240
+    },
+    {
+      "epoch": 0.9193861963616164,
+      "grad_norm": 3.9385369364031715,
+      "learning_rate": 1.6947466036626947e-07,
+      "loss": 0.4312,
+      "step": 13241
+    },
+    {
+      "epoch": 0.9194556311623385,
+      "grad_norm": 5.21086882412259,
+      "learning_rate": 1.691844970554013e-07,
+      "loss": 0.615,
+      "step": 13242
+    },
+    {
+      "epoch": 0.9195250659630607,
+      "grad_norm": 5.038349045925452,
+      "learning_rate": 1.6889457808171473e-07,
+      "loss": 0.6033,
+      "step": 13243
+    },
+    {
+      "epoch": 0.9195945007637828,
+      "grad_norm": 4.33985111383103,
+      "learning_rate": 1.6860490345987358e-07,
+      "loss": 0.3521,
+      "step": 13244
+    },
+    {
+      "epoch": 0.9196639355645049,
+      "grad_norm": 3.6348218633322915,
+      "learning_rate": 1.6831547320452945e-07,
+      "loss": 0.3688,
+      "step": 13245
+    },
+    {
+      "epoch": 0.9197333703652271,
+      "grad_norm": 2.2391799227611715,
+      "learning_rate": 1.6802628733032123e-07,
+      "loss": 0.1203,
+      "step": 13246
+    },
+    {
+      "epoch": 0.9198028051659491,
+      "grad_norm": 4.156520408728646,
+      "learning_rate": 1.6773734585187495e-07,
+      "loss": 0.497,
+      "step": 13247
+    },
+    {
+      "epoch": 0.9198722399666713,
+      "grad_norm": 5.057616130028476,
+      "learning_rate": 1.6744864878380728e-07,
+      "loss": 0.4434,
+      "step": 13248
+    },
+    {
+      "epoch": 0.9199416747673934,
+      "grad_norm": 4.116897760876217,
+      "learning_rate": 1.6716019614071766e-07,
+      "loss": 0.4944,
+      "step": 13249
+    },
+    {
+      "epoch": 0.9200111095681155,
+      "grad_norm": 3.288221731762155,
+      "learning_rate": 1.6687198793719605e-07,
+      "loss": 0.2487,
+      "step": 13250
+    },
+    {
+      "epoch": 0.9200805443688377,
+      "grad_norm": 3.3360130806993156,
+      "learning_rate": 1.6658402418782083e-07,
+      "loss": 0.2485,
+      "step": 13251
+    },
+    {
+      "epoch": 0.9201499791695598,
+      "grad_norm": 4.1688821330911665,
+      "learning_rate": 1.6629630490715642e-07,
+      "loss": 0.3957,
+      "step": 13252
+    },
+    {
+      "epoch": 0.920219413970282,
+      "grad_norm": 4.602391138795815,
+      "learning_rate": 1.6600883010975454e-07,
+      "loss": 0.5724,
+      "step": 13253
+    },
+    {
+      "epoch": 0.920288848771004,
+      "grad_norm": 3.65147619981418,
+      "learning_rate": 1.6572159981015634e-07,
+      "loss": 0.3903,
+      "step": 13254
+    },
+    {
+      "epoch": 0.9203582835717261,
+      "grad_norm": 5.170488632408698,
+      "learning_rate": 1.654346140228902e-07,
+      "loss": 0.6156,
+      "step": 13255
+    },
+    {
+      "epoch": 0.9204277183724483,
+      "grad_norm": 6.4843181972837804,
+      "learning_rate": 1.651478727624689e-07,
+      "loss": 0.4088,
+      "step": 13256
+    },
+    {
+      "epoch": 0.9204971531731704,
+      "grad_norm": 3.4931291236595743,
+      "learning_rate": 1.6486137604339813e-07,
+      "loss": 0.4011,
+      "step": 13257
+    },
+    {
+      "epoch": 0.9205665879738926,
+      "grad_norm": 3.7680024008452033,
+      "learning_rate": 1.6457512388016795e-07,
+      "loss": 0.4589,
+      "step": 13258
+    },
+    {
+      "epoch": 0.9206360227746146,
+      "grad_norm": 3.1342241125727073,
+      "learning_rate": 1.6428911628725563e-07,
+      "loss": 0.3699,
+      "step": 13259
+    },
+    {
+      "epoch": 0.9207054575753367,
+      "grad_norm": 4.226559482047216,
+      "learning_rate": 1.6400335327912854e-07,
+      "loss": 0.3166,
+      "step": 13260
+    },
+    {
+      "epoch": 0.9207748923760589,
+      "grad_norm": 2.955892670788817,
+      "learning_rate": 1.6371783487023897e-07,
+      "loss": 0.2298,
+      "step": 13261
+    },
+    {
+      "epoch": 0.920844327176781,
+      "grad_norm": 5.034508761736175,
+      "learning_rate": 1.6343256107502925e-07,
+      "loss": 0.6193,
+      "step": 13262
+    },
+    {
+      "epoch": 0.9209137619775031,
+      "grad_norm": 4.803019608120732,
+      "learning_rate": 1.631475319079273e-07,
+      "loss": 0.3488,
+      "step": 13263
+    },
+    {
+      "epoch": 0.9209831967782253,
+      "grad_norm": 4.285440672482229,
+      "learning_rate": 1.628627473833494e-07,
+      "loss": 0.5497,
+      "step": 13264
+    },
+    {
+      "epoch": 0.9210526315789473,
+      "grad_norm": 4.84450915866471,
+      "learning_rate": 1.6257820751570063e-07,
+      "loss": 0.2294,
+      "step": 13265
+    },
+    {
+      "epoch": 0.9211220663796695,
+      "grad_norm": 3.6426051437246545,
+      "learning_rate": 1.622939123193723e-07,
+      "loss": 0.3159,
+      "step": 13266
+    },
+    {
+      "epoch": 0.9211915011803916,
+      "grad_norm": 4.312335916224246,
+      "learning_rate": 1.6200986180874346e-07,
+      "loss": 0.5334,
+      "step": 13267
+    },
+    {
+      "epoch": 0.9212609359811137,
+      "grad_norm": 5.3598060493380055,
+      "learning_rate": 1.6172605599818147e-07,
+      "loss": 0.6245,
+      "step": 13268
+    },
+    {
+      "epoch": 0.9213303707818359,
+      "grad_norm": 3.9457633467402125,
+      "learning_rate": 1.61442494902041e-07,
+      "loss": 0.3488,
+      "step": 13269
+    },
+    {
+      "epoch": 0.921399805582558,
+      "grad_norm": 3.5351825237952696,
+      "learning_rate": 1.611591785346628e-07,
+      "loss": 0.442,
+      "step": 13270
+    },
+    {
+      "epoch": 0.9214692403832802,
+      "grad_norm": 4.384392902487356,
+      "learning_rate": 1.6087610691037868e-07,
+      "loss": 0.6413,
+      "step": 13271
+    },
+    {
+      "epoch": 0.9215386751840022,
+      "grad_norm": 2.852414864167171,
+      "learning_rate": 1.6059328004350504e-07,
+      "loss": 0.191,
+      "step": 13272
+    },
+    {
+      "epoch": 0.9216081099847243,
+      "grad_norm": 3.9789472758633657,
+      "learning_rate": 1.6031069794834709e-07,
+      "loss": 0.245,
+      "step": 13273
+    },
+    {
+      "epoch": 0.9216775447854465,
+      "grad_norm": 3.9672704089694544,
+      "learning_rate": 1.600283606391978e-07,
+      "loss": 0.5368,
+      "step": 13274
+    },
+    {
+      "epoch": 0.9217469795861686,
+      "grad_norm": 3.295467421759712,
+      "learning_rate": 1.5974626813033743e-07,
+      "loss": 0.3576,
+      "step": 13275
+    },
+    {
+      "epoch": 0.9218164143868907,
+      "grad_norm": 3.9857068558116313,
+      "learning_rate": 1.594644204360335e-07,
+      "loss": 0.4698,
+      "step": 13276
+    },
+    {
+      "epoch": 0.9218858491876128,
+      "grad_norm": 5.939251652722061,
+      "learning_rate": 1.5918281757054122e-07,
+      "loss": 0.3042,
+      "step": 13277
+    },
+    {
+      "epoch": 0.9219552839883349,
+      "grad_norm": 3.4526225540413256,
+      "learning_rate": 1.5890145954810478e-07,
+      "loss": 0.2741,
+      "step": 13278
+    },
+    {
+      "epoch": 0.9220247187890571,
+      "grad_norm": 4.034337080556549,
+      "learning_rate": 1.5862034638295388e-07,
+      "loss": 0.4025,
+      "step": 13279
+    },
+    {
+      "epoch": 0.9220941535897792,
+      "grad_norm": 4.026573781583536,
+      "learning_rate": 1.5833947808930828e-07,
+      "loss": 0.5206,
+      "step": 13280
+    },
+    {
+      "epoch": 0.9221635883905013,
+      "grad_norm": 2.6585815587371022,
+      "learning_rate": 1.5805885468137329e-07,
+      "loss": 0.1908,
+      "step": 13281
+    },
+    {
+      "epoch": 0.9222330231912235,
+      "grad_norm": 4.11864478898156,
+      "learning_rate": 1.577784761733414e-07,
+      "loss": 0.4444,
+      "step": 13282
+    },
+    {
+      "epoch": 0.9223024579919455,
+      "grad_norm": 3.553709735341548,
+      "learning_rate": 1.5749834257939523e-07,
+      "loss": 0.2517,
+      "step": 13283
+    },
+    {
+      "epoch": 0.9223718927926677,
+      "grad_norm": 11.038280186854696,
+      "learning_rate": 1.5721845391370394e-07,
+      "loss": 0.5494,
+      "step": 13284
+    },
+    {
+      "epoch": 0.9224413275933898,
+      "grad_norm": 7.965257479315802,
+      "learning_rate": 1.5693881019042235e-07,
+      "loss": 0.3387,
+      "step": 13285
+    },
+    {
+      "epoch": 0.9225107623941119,
+      "grad_norm": 4.222757466805156,
+      "learning_rate": 1.566594114236958e-07,
+      "loss": 0.383,
+      "step": 13286
+    },
+    {
+      "epoch": 0.9225801971948341,
+      "grad_norm": 3.6959150346619967,
+      "learning_rate": 1.5638025762765518e-07,
+      "loss": 0.3647,
+      "step": 13287
+    },
+    {
+      "epoch": 0.9226496319955562,
+      "grad_norm": 4.193364213624366,
+      "learning_rate": 1.561013488164209e-07,
+      "loss": 0.4148,
+      "step": 13288
+    },
+    {
+      "epoch": 0.9227190667962782,
+      "grad_norm": 4.912501551946821,
+      "learning_rate": 1.5582268500409893e-07,
+      "loss": 0.5047,
+      "step": 13289
+    },
+    {
+      "epoch": 0.9227885015970004,
+      "grad_norm": 3.964488784649308,
+      "learning_rate": 1.5554426620478402e-07,
+      "loss": 0.4374,
+      "step": 13290
+    },
+    {
+      "epoch": 0.9228579363977225,
+      "grad_norm": 3.3647218424063112,
+      "learning_rate": 1.552660924325583e-07,
+      "loss": 0.4019,
+      "step": 13291
+    },
+    {
+      "epoch": 0.9229273711984447,
+      "grad_norm": 4.651062096247276,
+      "learning_rate": 1.5498816370149105e-07,
+      "loss": 0.5114,
+      "step": 13292
+    },
+    {
+      "epoch": 0.9229968059991668,
+      "grad_norm": 11.359765893872696,
+      "learning_rate": 1.547104800256405e-07,
+      "loss": 0.3849,
+      "step": 13293
+    },
+    {
+      "epoch": 0.9230662407998889,
+      "grad_norm": 4.297665333684869,
+      "learning_rate": 1.5443304141905092e-07,
+      "loss": 0.5264,
+      "step": 13294
+    },
+    {
+      "epoch": 0.923135675600611,
+      "grad_norm": 3.2419756232215753,
+      "learning_rate": 1.5415584789575444e-07,
+      "loss": 0.2964,
+      "step": 13295
+    },
+    {
+      "epoch": 0.9232051104013331,
+      "grad_norm": 3.8380765077570347,
+      "learning_rate": 1.5387889946977207e-07,
+      "loss": 0.6351,
+      "step": 13296
+    },
+    {
+      "epoch": 0.9232745452020553,
+      "grad_norm": 4.2556913476755565,
+      "learning_rate": 1.5360219615511097e-07,
+      "loss": 0.5869,
+      "step": 13297
+    },
+    {
+      "epoch": 0.9233439800027774,
+      "grad_norm": 3.2453627362161837,
+      "learning_rate": 1.5332573796576767e-07,
+      "loss": 0.2551,
+      "step": 13298
+    },
+    {
+      "epoch": 0.9234134148034995,
+      "grad_norm": 4.567420446415648,
+      "learning_rate": 1.5304952491572322e-07,
+      "loss": 0.4406,
+      "step": 13299
+    },
+    {
+      "epoch": 0.9234828496042217,
+      "grad_norm": 3.7246895345484963,
+      "learning_rate": 1.5277355701894924e-07,
+      "loss": 0.314,
+      "step": 13300
+    },
+    {
+      "epoch": 0.9235522844049437,
+      "grad_norm": 4.084053511806754,
+      "learning_rate": 1.5249783428940456e-07,
+      "loss": 0.3632,
+      "step": 13301
+    },
+    {
+      "epoch": 0.9236217192056658,
+      "grad_norm": 2.898323971987378,
+      "learning_rate": 1.5222235674103304e-07,
+      "loss": 0.2198,
+      "step": 13302
+    },
+    {
+      "epoch": 0.923691154006388,
+      "grad_norm": 3.710216119772604,
+      "learning_rate": 1.5194712438776904e-07,
+      "loss": 0.4989,
+      "step": 13303
+    },
+    {
+      "epoch": 0.9237605888071101,
+      "grad_norm": 3.64956909979816,
+      "learning_rate": 1.5167213724353426e-07,
+      "loss": 0.2736,
+      "step": 13304
+    },
+    {
+      "epoch": 0.9238300236078323,
+      "grad_norm": 3.9450867521006434,
+      "learning_rate": 1.5139739532223642e-07,
+      "loss": 0.4004,
+      "step": 13305
+    },
+    {
+      "epoch": 0.9238994584085544,
+      "grad_norm": 3.7613422047895995,
+      "learning_rate": 1.5112289863777163e-07,
+      "loss": 0.3616,
+      "step": 13306
+    },
+    {
+      "epoch": 0.9239688932092764,
+      "grad_norm": 5.211063469598256,
+      "learning_rate": 1.508486472040238e-07,
+      "loss": 0.4765,
+      "step": 13307
+    },
+    {
+      "epoch": 0.9240383280099986,
+      "grad_norm": 4.061588904009952,
+      "learning_rate": 1.505746410348635e-07,
+      "loss": 0.3618,
+      "step": 13308
+    },
+    {
+      "epoch": 0.9241077628107207,
+      "grad_norm": 4.531682732207523,
+      "learning_rate": 1.5030088014415133e-07,
+      "loss": 0.5235,
+      "step": 13309
+    },
+    {
+      "epoch": 0.9241771976114429,
+      "grad_norm": 3.1638927643588213,
+      "learning_rate": 1.5002736454573287e-07,
+      "loss": 0.2853,
+      "step": 13310
+    },
+    {
+      "epoch": 0.924246632412165,
+      "grad_norm": 3.8604299931400887,
+      "learning_rate": 1.497540942534409e-07,
+      "loss": 0.3547,
+      "step": 13311
+    },
+    {
+      "epoch": 0.9243160672128871,
+      "grad_norm": 3.348296392298224,
+      "learning_rate": 1.4948106928109995e-07,
+      "loss": 0.248,
+      "step": 13312
+    },
+    {
+      "epoch": 0.9243855020136093,
+      "grad_norm": 3.8796676490964517,
+      "learning_rate": 1.4920828964251733e-07,
+      "loss": 0.3375,
+      "step": 13313
+    },
+    {
+      "epoch": 0.9244549368143313,
+      "grad_norm": 3.632342457620517,
+      "learning_rate": 1.4893575535148973e-07,
+      "loss": 0.3273,
+      "step": 13314
+    },
+    {
+      "epoch": 0.9245243716150535,
+      "grad_norm": 3.4475054608039644,
+      "learning_rate": 1.4866346642180286e-07,
+      "loss": 0.4601,
+      "step": 13315
+    },
+    {
+      "epoch": 0.9245938064157756,
+      "grad_norm": 4.380433506883066,
+      "learning_rate": 1.483914228672284e-07,
+      "loss": 0.4999,
+      "step": 13316
+    },
+    {
+      "epoch": 0.9246632412164977,
+      "grad_norm": 2.5281999448329056,
+      "learning_rate": 1.4811962470152485e-07,
+      "loss": 0.1931,
+      "step": 13317
+    },
+    {
+      "epoch": 0.9247326760172199,
+      "grad_norm": 2.913557164250303,
+      "learning_rate": 1.478480719384412e-07,
+      "loss": 0.2598,
+      "step": 13318
+    },
+    {
+      "epoch": 0.924802110817942,
+      "grad_norm": 4.6090760493367915,
+      "learning_rate": 1.4757676459171145e-07,
+      "loss": 0.492,
+      "step": 13319
+    },
+    {
+      "epoch": 0.924871545618664,
+      "grad_norm": 3.0279305150821787,
+      "learning_rate": 1.4730570267505796e-07,
+      "loss": 0.2288,
+      "step": 13320
+    },
+    {
+      "epoch": 0.9249409804193862,
+      "grad_norm": 4.103128710965601,
+      "learning_rate": 1.4703488620219087e-07,
+      "loss": 0.3181,
+      "step": 13321
+    },
+    {
+      "epoch": 0.9250104152201083,
+      "grad_norm": 3.7922170834422384,
+      "learning_rate": 1.4676431518680757e-07,
+      "loss": 0.4872,
+      "step": 13322
+    },
+    {
+      "epoch": 0.9250798500208305,
+      "grad_norm": 4.4807413462046295,
+      "learning_rate": 1.4649398964259376e-07,
+      "loss": 0.5223,
+      "step": 13323
+    },
+    {
+      "epoch": 0.9251492848215526,
+      "grad_norm": 3.448960407105857,
+      "learning_rate": 1.4622390958322185e-07,
+      "loss": 0.3404,
+      "step": 13324
+    },
+    {
+      "epoch": 0.9252187196222746,
+      "grad_norm": 3.8986114075130383,
+      "learning_rate": 1.4595407502235203e-07,
+      "loss": 0.4343,
+      "step": 13325
+    },
+    {
+      "epoch": 0.9252881544229968,
+      "grad_norm": 3.6105242861733755,
+      "learning_rate": 1.4568448597363227e-07,
+      "loss": 0.3114,
+      "step": 13326
+    },
+    {
+      "epoch": 0.9253575892237189,
+      "grad_norm": 3.6050439825245832,
+      "learning_rate": 1.4541514245069776e-07,
+      "loss": 0.419,
+      "step": 13327
+    },
+    {
+      "epoch": 0.9254270240244411,
+      "grad_norm": 4.664839831309387,
+      "learning_rate": 1.4514604446717207e-07,
+      "loss": 0.4244,
+      "step": 13328
+    },
+    {
+      "epoch": 0.9254964588251632,
+      "grad_norm": 3.8118731179601166,
+      "learning_rate": 1.448771920366665e-07,
+      "loss": 0.2723,
+      "step": 13329
+    },
+    {
+      "epoch": 0.9255658936258853,
+      "grad_norm": 4.338576541188535,
+      "learning_rate": 1.4460858517277855e-07,
+      "loss": 0.3205,
+      "step": 13330
+    },
+    {
+      "epoch": 0.9256353284266075,
+      "grad_norm": 3.4887750455650575,
+      "learning_rate": 1.4434022388909342e-07,
+      "loss": 0.3753,
+      "step": 13331
+    },
+    {
+      "epoch": 0.9257047632273295,
+      "grad_norm": 3.9332030672595106,
+      "learning_rate": 1.4407210819918582e-07,
+      "loss": 0.3537,
+      "step": 13332
+    },
+    {
+      "epoch": 0.9257741980280516,
+      "grad_norm": 2.361709121622135,
+      "learning_rate": 1.4380423811661603e-07,
+      "loss": 0.2156,
+      "step": 13333
+    },
+    {
+      "epoch": 0.9258436328287738,
+      "grad_norm": 4.921957634251678,
+      "learning_rate": 1.4353661365493265e-07,
+      "loss": 0.3616,
+      "step": 13334
+    },
+    {
+      "epoch": 0.9259130676294959,
+      "grad_norm": 4.662194249980464,
+      "learning_rate": 1.4326923482767153e-07,
+      "loss": 0.6619,
+      "step": 13335
+    },
+    {
+      "epoch": 0.9259825024302181,
+      "grad_norm": 3.514248205787015,
+      "learning_rate": 1.4300210164835737e-07,
+      "loss": 0.4691,
+      "step": 13336
+    },
+    {
+      "epoch": 0.9260519372309401,
+      "grad_norm": 4.317618879115888,
+      "learning_rate": 1.427352141305005e-07,
+      "loss": 0.3852,
+      "step": 13337
+    },
+    {
+      "epoch": 0.9261213720316622,
+      "grad_norm": 3.3425766621548973,
+      "learning_rate": 1.4246857228760015e-07,
+      "loss": 0.4199,
+      "step": 13338
+    },
+    {
+      "epoch": 0.9261908068323844,
+      "grad_norm": 2.3232312131200716,
+      "learning_rate": 1.4220217613314325e-07,
+      "loss": 0.1737,
+      "step": 13339
+    },
+    {
+      "epoch": 0.9262602416331065,
+      "grad_norm": 8.01487637849613,
+      "learning_rate": 1.4193602568060182e-07,
+      "loss": 0.3074,
+      "step": 13340
+    },
+    {
+      "epoch": 0.9263296764338287,
+      "grad_norm": 3.449634575308138,
+      "learning_rate": 1.4167012094344068e-07,
+      "loss": 0.3382,
+      "step": 13341
+    },
+    {
+      "epoch": 0.9263991112345508,
+      "grad_norm": 4.410114519094716,
+      "learning_rate": 1.4140446193510682e-07,
+      "loss": 0.322,
+      "step": 13342
+    },
+    {
+      "epoch": 0.9264685460352728,
+      "grad_norm": 3.4394082796744736,
+      "learning_rate": 1.4113904866903672e-07,
+      "loss": 0.379,
+      "step": 13343
+    },
+    {
+      "epoch": 0.926537980835995,
+      "grad_norm": 3.6781233059270493,
+      "learning_rate": 1.408738811586563e-07,
+      "loss": 0.3576,
+      "step": 13344
+    },
+    {
+      "epoch": 0.9266074156367171,
+      "grad_norm": 4.3628366633387365,
+      "learning_rate": 1.4060895941737596e-07,
+      "loss": 0.6982,
+      "step": 13345
+    },
+    {
+      "epoch": 0.9266768504374392,
+      "grad_norm": 3.9871849731194287,
+      "learning_rate": 1.4034428345859496e-07,
+      "loss": 0.2542,
+      "step": 13346
+    },
+    {
+      "epoch": 0.9267462852381614,
+      "grad_norm": 5.942378573209138,
+      "learning_rate": 1.4007985329570205e-07,
+      "loss": 0.6101,
+      "step": 13347
+    },
+    {
+      "epoch": 0.9268157200388835,
+      "grad_norm": 4.346096340257077,
+      "learning_rate": 1.3981566894207044e-07,
+      "loss": 0.5666,
+      "step": 13348
+    },
+    {
+      "epoch": 0.9268851548396057,
+      "grad_norm": 4.280100559377925,
+      "learning_rate": 1.3955173041106274e-07,
+      "loss": 0.5099,
+      "step": 13349
+    },
+    {
+      "epoch": 0.9269545896403277,
+      "grad_norm": 3.328453778222543,
+      "learning_rate": 1.392880377160283e-07,
+      "loss": 0.2706,
+      "step": 13350
+    },
+    {
+      "epoch": 0.9270240244410498,
+      "grad_norm": 3.8451592399104157,
+      "learning_rate": 1.390245908703053e-07,
+      "loss": 0.4853,
+      "step": 13351
+    },
+    {
+      "epoch": 0.927093459241772,
+      "grad_norm": 4.248739999242768,
+      "learning_rate": 1.38761389887217e-07,
+      "loss": 0.5379,
+      "step": 13352
+    },
+    {
+      "epoch": 0.9271628940424941,
+      "grad_norm": 4.279608584608455,
+      "learning_rate": 1.3849843478007773e-07,
+      "loss": 0.6773,
+      "step": 13353
+    },
+    {
+      "epoch": 0.9272323288432163,
+      "grad_norm": 8.97334313142825,
+      "learning_rate": 1.3823572556218578e-07,
+      "loss": 0.1931,
+      "step": 13354
+    },
+    {
+      "epoch": 0.9273017636439383,
+      "grad_norm": 3.333794703994239,
+      "learning_rate": 1.3797326224682938e-07,
+      "loss": 0.3352,
+      "step": 13355
+    },
+    {
+      "epoch": 0.9273711984446604,
+      "grad_norm": 3.0496757173581077,
+      "learning_rate": 1.3771104484728405e-07,
+      "loss": 0.3036,
+      "step": 13356
+    },
+    {
+      "epoch": 0.9274406332453826,
+      "grad_norm": 4.705592693072218,
+      "learning_rate": 1.3744907337681134e-07,
+      "loss": 0.4031,
+      "step": 13357
+    },
+    {
+      "epoch": 0.9275100680461047,
+      "grad_norm": 3.5077775476340896,
+      "learning_rate": 1.3718734784866238e-07,
+      "loss": 0.3633,
+      "step": 13358
+    },
+    {
+      "epoch": 0.9275795028468268,
+      "grad_norm": 3.7177010346117574,
+      "learning_rate": 1.3692586827607545e-07,
+      "loss": 0.3666,
+      "step": 13359
+    },
+    {
+      "epoch": 0.927648937647549,
+      "grad_norm": 4.018276430880449,
+      "learning_rate": 1.3666463467227386e-07,
+      "loss": 0.3882,
+      "step": 13360
+    },
+    {
+      "epoch": 0.927718372448271,
+      "grad_norm": 3.4497912868772227,
+      "learning_rate": 1.364036470504726e-07,
+      "loss": 0.3745,
+      "step": 13361
+    },
+    {
+      "epoch": 0.9277878072489932,
+      "grad_norm": 3.667429090884349,
+      "learning_rate": 1.361429054238722e-07,
+      "loss": 0.3754,
+      "step": 13362
+    },
+    {
+      "epoch": 0.9278572420497153,
+      "grad_norm": 4.0263715034177805,
+      "learning_rate": 1.3588240980565825e-07,
+      "loss": 0.3888,
+      "step": 13363
+    },
+    {
+      "epoch": 0.9279266768504374,
+      "grad_norm": 3.0923393133320016,
+      "learning_rate": 1.3562216020900852e-07,
+      "loss": 0.2734,
+      "step": 13364
+    },
+    {
+      "epoch": 0.9279961116511596,
+      "grad_norm": 4.676160870650495,
+      "learning_rate": 1.3536215664708585e-07,
+      "loss": 0.4859,
+      "step": 13365
+    },
+    {
+      "epoch": 0.9280655464518817,
+      "grad_norm": 3.1269741153218304,
+      "learning_rate": 1.3510239913304024e-07,
+      "loss": 0.1929,
+      "step": 13366
+    },
+    {
+      "epoch": 0.9281349812526039,
+      "grad_norm": 4.743728173270907,
+      "learning_rate": 1.348428876800101e-07,
+      "loss": 0.4814,
+      "step": 13367
+    },
+    {
+      "epoch": 0.9282044160533259,
+      "grad_norm": 3.918670705184777,
+      "learning_rate": 1.345836223011221e-07,
+      "loss": 0.2752,
+      "step": 13368
+    },
+    {
+      "epoch": 0.928273850854048,
+      "grad_norm": 4.015663948749614,
+      "learning_rate": 1.34324603009488e-07,
+      "loss": 0.4704,
+      "step": 13369
+    },
+    {
+      "epoch": 0.9283432856547702,
+      "grad_norm": 3.9021087013061253,
+      "learning_rate": 1.3406582981821016e-07,
+      "loss": 0.4257,
+      "step": 13370
+    },
+    {
+      "epoch": 0.9284127204554923,
+      "grad_norm": 4.300624593773659,
+      "learning_rate": 1.3380730274037634e-07,
+      "loss": 0.4127,
+      "step": 13371
+    },
+    {
+      "epoch": 0.9284821552562145,
+      "grad_norm": 4.149234118757186,
+      "learning_rate": 1.3354902178906227e-07,
+      "loss": 0.5341,
+      "step": 13372
+    },
+    {
+      "epoch": 0.9285515900569365,
+      "grad_norm": 4.153791252283487,
+      "learning_rate": 1.33290986977333e-07,
+      "loss": 0.6317,
+      "step": 13373
+    },
+    {
+      "epoch": 0.9286210248576586,
+      "grad_norm": 4.423569173936873,
+      "learning_rate": 1.3303319831823759e-07,
+      "loss": 0.4995,
+      "step": 13374
+    },
+    {
+      "epoch": 0.9286904596583808,
+      "grad_norm": 4.210045115366354,
+      "learning_rate": 1.3277565582481555e-07,
+      "loss": 0.4809,
+      "step": 13375
+    },
+    {
+      "epoch": 0.9287598944591029,
+      "grad_norm": 6.279177272831838,
+      "learning_rate": 1.3251835951009428e-07,
+      "loss": 0.9155,
+      "step": 13376
+    },
+    {
+      "epoch": 0.928829329259825,
+      "grad_norm": 4.056269349944527,
+      "learning_rate": 1.3226130938708504e-07,
+      "loss": 0.3655,
+      "step": 13377
+    },
+    {
+      "epoch": 0.9288987640605472,
+      "grad_norm": 4.256605815422928,
+      "learning_rate": 1.3200450546879129e-07,
+      "loss": 0.4373,
+      "step": 13378
+    },
+    {
+      "epoch": 0.9289681988612692,
+      "grad_norm": 3.4137491845716905,
+      "learning_rate": 1.3174794776820154e-07,
+      "loss": 0.4286,
+      "step": 13379
+    },
+    {
+      "epoch": 0.9290376336619914,
+      "grad_norm": 3.3266658383358365,
+      "learning_rate": 1.31491636298291e-07,
+      "loss": 0.3604,
+      "step": 13380
+    },
+    {
+      "epoch": 0.9291070684627135,
+      "grad_norm": 3.6086989314861846,
+      "learning_rate": 1.3123557107202478e-07,
+      "loss": 0.3309,
+      "step": 13381
+    },
+    {
+      "epoch": 0.9291765032634356,
+      "grad_norm": 3.306958706280662,
+      "learning_rate": 1.3097975210235424e-07,
+      "loss": 0.2856,
+      "step": 13382
+    },
+    {
+      "epoch": 0.9292459380641578,
+      "grad_norm": 5.032928133386874,
+      "learning_rate": 1.3072417940221792e-07,
+      "loss": 0.5646,
+      "step": 13383
+    },
+    {
+      "epoch": 0.9293153728648799,
+      "grad_norm": 4.713520359021732,
+      "learning_rate": 1.304688529845427e-07,
+      "loss": 0.58,
+      "step": 13384
+    },
+    {
+      "epoch": 0.929384807665602,
+      "grad_norm": 4.509716528398813,
+      "learning_rate": 1.302137728622427e-07,
+      "loss": 0.4278,
+      "step": 13385
+    },
+    {
+      "epoch": 0.9294542424663241,
+      "grad_norm": 3.986201257290607,
+      "learning_rate": 1.299589390482192e-07,
+      "loss": 0.4246,
+      "step": 13386
+    },
+    {
+      "epoch": 0.9295236772670462,
+      "grad_norm": 4.151068229343168,
+      "learning_rate": 1.2970435155536198e-07,
+      "loss": 0.3676,
+      "step": 13387
+    },
+    {
+      "epoch": 0.9295931120677684,
+      "grad_norm": 4.544722756172872,
+      "learning_rate": 1.2945001039654792e-07,
+      "loss": 0.539,
+      "step": 13388
+    },
+    {
+      "epoch": 0.9296625468684905,
+      "grad_norm": 5.739027037458772,
+      "learning_rate": 1.2919591558464006e-07,
+      "loss": 0.6711,
+      "step": 13389
+    },
+    {
+      "epoch": 0.9297319816692126,
+      "grad_norm": 3.6361227206766826,
+      "learning_rate": 1.2894206713249146e-07,
+      "loss": 0.4703,
+      "step": 13390
+    },
+    {
+      "epoch": 0.9298014164699347,
+      "grad_norm": 2.4565688423590992,
+      "learning_rate": 1.2868846505294185e-07,
+      "loss": 0.2294,
+      "step": 13391
+    },
+    {
+      "epoch": 0.9298708512706568,
+      "grad_norm": 3.286327270538481,
+      "learning_rate": 1.2843510935881655e-07,
+      "loss": 0.2578,
+      "step": 13392
+    },
+    {
+      "epoch": 0.929940286071379,
+      "grad_norm": 3.128018898931536,
+      "learning_rate": 1.281820000629308e-07,
+      "loss": 0.2529,
+      "step": 13393
+    },
+    {
+      "epoch": 0.9300097208721011,
+      "grad_norm": 4.21390290092579,
+      "learning_rate": 1.2792913717808775e-07,
+      "loss": 0.3177,
+      "step": 13394
+    },
+    {
+      "epoch": 0.9300791556728232,
+      "grad_norm": 5.478445347797341,
+      "learning_rate": 1.2767652071707437e-07,
+      "loss": 0.4805,
+      "step": 13395
+    },
+    {
+      "epoch": 0.9301485904735454,
+      "grad_norm": 3.7802128615196713,
+      "learning_rate": 1.2742415069267044e-07,
+      "loss": 0.4558,
+      "step": 13396
+    },
+    {
+      "epoch": 0.9302180252742674,
+      "grad_norm": 4.064468842600926,
+      "learning_rate": 1.2717202711763854e-07,
+      "loss": 0.4456,
+      "step": 13397
+    },
+    {
+      "epoch": 0.9302874600749896,
+      "grad_norm": 3.531028681754328,
+      "learning_rate": 1.2692015000473178e-07,
+      "loss": 0.2084,
+      "step": 13398
+    },
+    {
+      "epoch": 0.9303568948757117,
+      "grad_norm": 4.249615000533118,
+      "learning_rate": 1.2666851936668944e-07,
+      "loss": 0.4942,
+      "step": 13399
+    },
+    {
+      "epoch": 0.9304263296764338,
+      "grad_norm": 3.578678679234296,
+      "learning_rate": 1.264171352162391e-07,
+      "loss": 0.3868,
+      "step": 13400
+    },
+    {
+      "epoch": 0.930495764477156,
+      "grad_norm": 5.720178640214364,
+      "learning_rate": 1.2616599756609505e-07,
+      "loss": 0.4775,
+      "step": 13401
+    },
+    {
+      "epoch": 0.9305651992778781,
+      "grad_norm": 4.845708447135187,
+      "learning_rate": 1.259151064289593e-07,
+      "loss": 0.52,
+      "step": 13402
+    },
+    {
+      "epoch": 0.9306346340786001,
+      "grad_norm": 4.212503304641778,
+      "learning_rate": 1.2566446181752289e-07,
+      "loss": 0.5581,
+      "step": 13403
+    },
+    {
+      "epoch": 0.9307040688793223,
+      "grad_norm": 2.386083942465921,
+      "learning_rate": 1.2541406374446118e-07,
+      "loss": 0.1476,
+      "step": 13404
+    },
+    {
+      "epoch": 0.9307735036800444,
+      "grad_norm": 4.18141908470409,
+      "learning_rate": 1.2516391222244127e-07,
+      "loss": 0.4586,
+      "step": 13405
+    },
+    {
+      "epoch": 0.9308429384807666,
+      "grad_norm": 3.5163799426976374,
+      "learning_rate": 1.249140072641142e-07,
+      "loss": 0.3246,
+      "step": 13406
+    },
+    {
+      "epoch": 0.9309123732814887,
+      "grad_norm": 4.178001038043437,
+      "learning_rate": 1.246643488821192e-07,
+      "loss": 0.468,
+      "step": 13407
+    },
+    {
+      "epoch": 0.9309818080822108,
+      "grad_norm": 5.642762921846356,
+      "learning_rate": 1.244149370890857e-07,
+      "loss": 0.563,
+      "step": 13408
+    },
+    {
+      "epoch": 0.931051242882933,
+      "grad_norm": 3.048448747338096,
+      "learning_rate": 1.2416577189762746e-07,
+      "loss": 0.2191,
+      "step": 13409
+    },
+    {
+      "epoch": 0.931120677683655,
+      "grad_norm": 3.2086226890240916,
+      "learning_rate": 1.2391685332034665e-07,
+      "loss": 0.2427,
+      "step": 13410
+    },
+    {
+      "epoch": 0.9311901124843772,
+      "grad_norm": 3.4641191975310344,
+      "learning_rate": 1.2366818136983427e-07,
+      "loss": 0.3698,
+      "step": 13411
+    },
+    {
+      "epoch": 0.9312595472850993,
+      "grad_norm": 3.545007994079871,
+      "learning_rate": 1.2341975605866695e-07,
+      "loss": 0.2668,
+      "step": 13412
+    },
+    {
+      "epoch": 0.9313289820858214,
+      "grad_norm": 5.804644124989576,
+      "learning_rate": 1.2317157739941022e-07,
+      "loss": 0.4835,
+      "step": 13413
+    },
+    {
+      "epoch": 0.9313984168865436,
+      "grad_norm": 4.515875968365035,
+      "learning_rate": 1.2292364540461677e-07,
+      "loss": 0.5697,
+      "step": 13414
+    },
+    {
+      "epoch": 0.9314678516872656,
+      "grad_norm": 3.1315653553394527,
+      "learning_rate": 1.2267596008682603e-07,
+      "loss": 0.2375,
+      "step": 13415
+    },
+    {
+      "epoch": 0.9315372864879877,
+      "grad_norm": 3.3184663759057296,
+      "learning_rate": 1.2242852145856688e-07,
+      "loss": 0.2525,
+      "step": 13416
+    },
+    {
+      "epoch": 0.9316067212887099,
+      "grad_norm": 3.424081520205238,
+      "learning_rate": 1.2218132953235317e-07,
+      "loss": 0.3291,
+      "step": 13417
+    },
+    {
+      "epoch": 0.931676156089432,
+      "grad_norm": 3.813937944151166,
+      "learning_rate": 1.219343843206877e-07,
+      "loss": 0.3237,
+      "step": 13418
+    },
+    {
+      "epoch": 0.9317455908901542,
+      "grad_norm": 3.661951809017704,
+      "learning_rate": 1.2168768583606273e-07,
+      "loss": 0.4094,
+      "step": 13419
+    },
+    {
+      "epoch": 0.9318150256908763,
+      "grad_norm": 3.929789939462007,
+      "learning_rate": 1.2144123409095377e-07,
+      "loss": 0.2826,
+      "step": 13420
+    },
+    {
+      "epoch": 0.9318844604915983,
+      "grad_norm": 10.832490776522803,
+      "learning_rate": 1.2119502909782587e-07,
+      "loss": 0.5107,
+      "step": 13421
+    },
+    {
+      "epoch": 0.9319538952923205,
+      "grad_norm": 5.956124979801653,
+      "learning_rate": 1.2094907086913354e-07,
+      "loss": 0.4399,
+      "step": 13422
+    },
+    {
+      "epoch": 0.9320233300930426,
+      "grad_norm": 2.8859685636733388,
+      "learning_rate": 1.2070335941731627e-07,
+      "loss": 0.2168,
+      "step": 13423
+    },
+    {
+      "epoch": 0.9320927648937648,
+      "grad_norm": 3.554188485274006,
+      "learning_rate": 1.204578947548013e-07,
+      "loss": 0.3635,
+      "step": 13424
+    },
+    {
+      "epoch": 0.9321621996944869,
+      "grad_norm": 8.690797818111259,
+      "learning_rate": 1.202126768940043e-07,
+      "loss": 0.3039,
+      "step": 13425
+    },
+    {
+      "epoch": 0.932231634495209,
+      "grad_norm": 3.701417700911383,
+      "learning_rate": 1.199677058473292e-07,
+      "loss": 0.3535,
+      "step": 13426
+    },
+    {
+      "epoch": 0.9323010692959312,
+      "grad_norm": 4.498592734422018,
+      "learning_rate": 1.1972298162716445e-07,
+      "loss": 0.4338,
+      "step": 13427
+    },
+    {
+      "epoch": 0.9323705040966532,
+      "grad_norm": 3.516900149016874,
+      "learning_rate": 1.1947850424588901e-07,
+      "loss": 0.3017,
+      "step": 13428
+    },
+    {
+      "epoch": 0.9324399388973754,
+      "grad_norm": 4.979195014796428,
+      "learning_rate": 1.192342737158686e-07,
+      "loss": 0.5719,
+      "step": 13429
+    },
+    {
+      "epoch": 0.9325093736980975,
+      "grad_norm": 3.591473416861948,
+      "learning_rate": 1.189902900494555e-07,
+      "loss": 0.435,
+      "step": 13430
+    },
+    {
+      "epoch": 0.9325788084988196,
+      "grad_norm": 3.8272588083386583,
+      "learning_rate": 1.1874655325898987e-07,
+      "loss": 0.545,
+      "step": 13431
+    },
+    {
+      "epoch": 0.9326482432995418,
+      "grad_norm": 3.195703194198509,
+      "learning_rate": 1.1850306335680073e-07,
+      "loss": 0.2745,
+      "step": 13432
+    },
+    {
+      "epoch": 0.9327176781002638,
+      "grad_norm": 3.9146850941400584,
+      "learning_rate": 1.1825982035520211e-07,
+      "loss": 0.3814,
+      "step": 13433
+    },
+    {
+      "epoch": 0.9327871129009859,
+      "grad_norm": 6.145934325374261,
+      "learning_rate": 1.1801682426649863e-07,
+      "loss": 0.3428,
+      "step": 13434
+    },
+    {
+      "epoch": 0.9328565477017081,
+      "grad_norm": 3.058108617903866,
+      "learning_rate": 1.1777407510297934e-07,
+      "loss": 0.2451,
+      "step": 13435
+    },
+    {
+      "epoch": 0.9329259825024302,
+      "grad_norm": 3.8314845893129066,
+      "learning_rate": 1.1753157287692218e-07,
+      "loss": 0.4413,
+      "step": 13436
+    },
+    {
+      "epoch": 0.9329954173031524,
+      "grad_norm": 4.860471434691263,
+      "learning_rate": 1.1728931760059403e-07,
+      "loss": 0.5077,
+      "step": 13437
+    },
+    {
+      "epoch": 0.9330648521038745,
+      "grad_norm": 4.360127232377627,
+      "learning_rate": 1.170473092862462e-07,
+      "loss": 0.4182,
+      "step": 13438
+    },
+    {
+      "epoch": 0.9331342869045965,
+      "grad_norm": 4.01883914287503,
+      "learning_rate": 1.1680554794612053e-07,
+      "loss": 0.4122,
+      "step": 13439
+    },
+    {
+      "epoch": 0.9332037217053187,
+      "grad_norm": 3.5192261733944483,
+      "learning_rate": 1.1656403359244506e-07,
+      "loss": 0.2713,
+      "step": 13440
+    },
+    {
+      "epoch": 0.9332731565060408,
+      "grad_norm": 4.759885201869912,
+      "learning_rate": 1.1632276623743388e-07,
+      "loss": 0.5096,
+      "step": 13441
+    },
+    {
+      "epoch": 0.933342591306763,
+      "grad_norm": 4.484439959724089,
+      "learning_rate": 1.1608174589329169e-07,
+      "loss": 0.4654,
+      "step": 13442
+    },
+    {
+      "epoch": 0.9334120261074851,
+      "grad_norm": 5.359800097867683,
+      "learning_rate": 1.158409725722076e-07,
+      "loss": 0.5306,
+      "step": 13443
+    },
+    {
+      "epoch": 0.9334814609082072,
+      "grad_norm": 2.883298355208876,
+      "learning_rate": 1.156004462863608e-07,
+      "loss": 0.2709,
+      "step": 13444
+    },
+    {
+      "epoch": 0.9335508957089294,
+      "grad_norm": 3.672999199793784,
+      "learning_rate": 1.1536016704791653e-07,
+      "loss": 0.2555,
+      "step": 13445
+    },
+    {
+      "epoch": 0.9336203305096514,
+      "grad_norm": 4.696010194572382,
+      "learning_rate": 1.151201348690273e-07,
+      "loss": 0.4557,
+      "step": 13446
+    },
+    {
+      "epoch": 0.9336897653103735,
+      "grad_norm": 4.325958261413924,
+      "learning_rate": 1.1488034976183337e-07,
+      "loss": 0.4916,
+      "step": 13447
+    },
+    {
+      "epoch": 0.9337592001110957,
+      "grad_norm": 4.000456079808372,
+      "learning_rate": 1.1464081173846453e-07,
+      "loss": 0.4493,
+      "step": 13448
+    },
+    {
+      "epoch": 0.9338286349118178,
+      "grad_norm": 5.860291211382361,
+      "learning_rate": 1.1440152081103495e-07,
+      "loss": 0.6867,
+      "step": 13449
+    },
+    {
+      "epoch": 0.93389806971254,
+      "grad_norm": 3.4304650129905254,
+      "learning_rate": 1.1416247699164774e-07,
+      "loss": 0.262,
+      "step": 13450
+    },
+    {
+      "epoch": 0.933967504513262,
+      "grad_norm": 4.324414033478601,
+      "learning_rate": 1.1392368029239432e-07,
+      "loss": 0.5765,
+      "step": 13451
+    },
+    {
+      "epoch": 0.9340369393139841,
+      "grad_norm": 4.1090117819007945,
+      "learning_rate": 1.136851307253517e-07,
+      "loss": 0.6206,
+      "step": 13452
+    },
+    {
+      "epoch": 0.9341063741147063,
+      "grad_norm": 22.70521511503073,
+      "learning_rate": 1.134468283025858e-07,
+      "loss": 0.446,
+      "step": 13453
+    },
+    {
+      "epoch": 0.9341758089154284,
+      "grad_norm": 3.1375896847645017,
+      "learning_rate": 1.1320877303614974e-07,
+      "loss": 0.1845,
+      "step": 13454
+    },
+    {
+      "epoch": 0.9342452437161506,
+      "grad_norm": 3.9366941394620176,
+      "learning_rate": 1.1297096493808502e-07,
+      "loss": 0.5268,
+      "step": 13455
+    },
+    {
+      "epoch": 0.9343146785168727,
+      "grad_norm": 4.727456453335802,
+      "learning_rate": 1.127334040204181e-07,
+      "loss": 0.4459,
+      "step": 13456
+    },
+    {
+      "epoch": 0.9343841133175947,
+      "grad_norm": 3.5436875273224477,
+      "learning_rate": 1.124960902951655e-07,
+      "loss": 0.2426,
+      "step": 13457
+    },
+    {
+      "epoch": 0.9344535481183169,
+      "grad_norm": 4.060824085443516,
+      "learning_rate": 1.122590237743304e-07,
+      "loss": 0.5865,
+      "step": 13458
+    },
+    {
+      "epoch": 0.934522982919039,
+      "grad_norm": 4.592407023884917,
+      "learning_rate": 1.1202220446990264e-07,
+      "loss": 0.4623,
+      "step": 13459
+    },
+    {
+      "epoch": 0.9345924177197611,
+      "grad_norm": 4.654510570030044,
+      "learning_rate": 1.1178563239386042e-07,
+      "loss": 0.4249,
+      "step": 13460
+    },
+    {
+      "epoch": 0.9346618525204833,
+      "grad_norm": 3.3427191398895735,
+      "learning_rate": 1.1154930755816973e-07,
+      "loss": 0.4212,
+      "step": 13461
+    },
+    {
+      "epoch": 0.9347312873212054,
+      "grad_norm": 3.9900870445539063,
+      "learning_rate": 1.1131322997478378e-07,
+      "loss": 0.4292,
+      "step": 13462
+    },
+    {
+      "epoch": 0.9348007221219276,
+      "grad_norm": 3.6701892258213076,
+      "learning_rate": 1.1107739965564247e-07,
+      "loss": 0.4247,
+      "step": 13463
+    },
+    {
+      "epoch": 0.9348701569226496,
+      "grad_norm": 4.091622478104981,
+      "learning_rate": 1.1084181661267401e-07,
+      "loss": 0.3738,
+      "step": 13464
+    },
+    {
+      "epoch": 0.9349395917233717,
+      "grad_norm": 3.8968206805966217,
+      "learning_rate": 1.106064808577939e-07,
+      "loss": 0.3929,
+      "step": 13465
+    },
+    {
+      "epoch": 0.9350090265240939,
+      "grad_norm": 3.8833799006397878,
+      "learning_rate": 1.1037139240290595e-07,
+      "loss": 0.3367,
+      "step": 13466
+    },
+    {
+      "epoch": 0.935078461324816,
+      "grad_norm": 2.8563722383506667,
+      "learning_rate": 1.1013655125989897e-07,
+      "loss": 0.2624,
+      "step": 13467
+    },
+    {
+      "epoch": 0.9351478961255382,
+      "grad_norm": 4.840690267749723,
+      "learning_rate": 1.0990195744065235e-07,
+      "loss": 0.519,
+      "step": 13468
+    },
+    {
+      "epoch": 0.9352173309262602,
+      "grad_norm": 4.625550267150267,
+      "learning_rate": 1.096676109570316e-07,
+      "loss": 0.6056,
+      "step": 13469
+    },
+    {
+      "epoch": 0.9352867657269823,
+      "grad_norm": 4.352917068098696,
+      "learning_rate": 1.094335118208889e-07,
+      "loss": 0.5337,
+      "step": 13470
+    },
+    {
+      "epoch": 0.9353562005277045,
+      "grad_norm": 3.818749752352249,
+      "learning_rate": 1.0919966004406535e-07,
+      "loss": 0.49,
+      "step": 13471
+    },
+    {
+      "epoch": 0.9354256353284266,
+      "grad_norm": 7.667444718670821,
+      "learning_rate": 1.0896605563838814e-07,
+      "loss": 0.6346,
+      "step": 13472
+    },
+    {
+      "epoch": 0.9354950701291487,
+      "grad_norm": 2.9323667467715713,
+      "learning_rate": 1.0873269861567393e-07,
+      "loss": 0.19,
+      "step": 13473
+    },
+    {
+      "epoch": 0.9355645049298709,
+      "grad_norm": 4.248786566800276,
+      "learning_rate": 1.0849958898772495e-07,
+      "loss": 0.4926,
+      "step": 13474
+    },
+    {
+      "epoch": 0.935633939730593,
+      "grad_norm": 3.7745792531890037,
+      "learning_rate": 1.082667267663312e-07,
+      "loss": 0.2563,
+      "step": 13475
+    },
+    {
+      "epoch": 0.9357033745313151,
+      "grad_norm": 3.976548065217878,
+      "learning_rate": 1.0803411196327163e-07,
+      "loss": 0.4498,
+      "step": 13476
+    },
+    {
+      "epoch": 0.9357728093320372,
+      "grad_norm": 4.872525356625967,
+      "learning_rate": 1.0780174459031068e-07,
+      "loss": 0.4766,
+      "step": 13477
+    },
+    {
+      "epoch": 0.9358422441327593,
+      "grad_norm": 3.833434856429584,
+      "learning_rate": 1.0756962465920173e-07,
+      "loss": 0.3923,
+      "step": 13478
+    },
+    {
+      "epoch": 0.9359116789334815,
+      "grad_norm": 2.759833034624004,
+      "learning_rate": 1.073377521816843e-07,
+      "loss": 0.2756,
+      "step": 13479
+    },
+    {
+      "epoch": 0.9359811137342036,
+      "grad_norm": 4.562036650345542,
+      "learning_rate": 1.0710612716948787e-07,
+      "loss": 0.4028,
+      "step": 13480
+    },
+    {
+      "epoch": 0.9360505485349258,
+      "grad_norm": 4.444843515264536,
+      "learning_rate": 1.0687474963432643e-07,
+      "loss": 0.5697,
+      "step": 13481
+    },
+    {
+      "epoch": 0.9361199833356478,
+      "grad_norm": 5.229619455561418,
+      "learning_rate": 1.0664361958790281e-07,
+      "loss": 0.2508,
+      "step": 13482
+    },
+    {
+      "epoch": 0.9361894181363699,
+      "grad_norm": 6.434617031694959,
+      "learning_rate": 1.064127370419088e-07,
+      "loss": 0.8769,
+      "step": 13483
+    },
+    {
+      "epoch": 0.9362588529370921,
+      "grad_norm": 4.577895472317831,
+      "learning_rate": 1.061821020080206e-07,
+      "loss": 0.5083,
+      "step": 13484
+    },
+    {
+      "epoch": 0.9363282877378142,
+      "grad_norm": 4.066058276183629,
+      "learning_rate": 1.0595171449790331e-07,
+      "loss": 0.3382,
+      "step": 13485
+    },
+    {
+      "epoch": 0.9363977225385363,
+      "grad_norm": 3.137949752217085,
+      "learning_rate": 1.0572157452321097e-07,
+      "loss": 0.4231,
+      "step": 13486
+    },
+    {
+      "epoch": 0.9364671573392584,
+      "grad_norm": 2.750542677603122,
+      "learning_rate": 1.0549168209558314e-07,
+      "loss": 0.237,
+      "step": 13487
+    },
+    {
+      "epoch": 0.9365365921399805,
+      "grad_norm": 2.7685388448366863,
+      "learning_rate": 1.0526203722664718e-07,
+      "loss": 0.1481,
+      "step": 13488
+    },
+    {
+      "epoch": 0.9366060269407027,
+      "grad_norm": 4.0841931261152435,
+      "learning_rate": 1.0503263992801883e-07,
+      "loss": 0.4172,
+      "step": 13489
+    },
+    {
+      "epoch": 0.9366754617414248,
+      "grad_norm": 4.029558062267512,
+      "learning_rate": 1.0480349021130043e-07,
+      "loss": 0.4543,
+      "step": 13490
+    },
+    {
+      "epoch": 0.9367448965421469,
+      "grad_norm": 5.334640452208785,
+      "learning_rate": 1.0457458808808218e-07,
+      "loss": 0.5332,
+      "step": 13491
+    },
+    {
+      "epoch": 0.9368143313428691,
+      "grad_norm": 5.748058963015037,
+      "learning_rate": 1.0434593356994205e-07,
+      "loss": 0.6989,
+      "step": 13492
+    },
+    {
+      "epoch": 0.9368837661435911,
+      "grad_norm": 4.104095274790957,
+      "learning_rate": 1.041175266684441e-07,
+      "loss": 0.5037,
+      "step": 13493
+    },
+    {
+      "epoch": 0.9369532009443133,
+      "grad_norm": 6.506968511149294,
+      "learning_rate": 1.0388936739514244e-07,
+      "loss": 0.4864,
+      "step": 13494
+    },
+    {
+      "epoch": 0.9370226357450354,
+      "grad_norm": 3.9416774335824765,
+      "learning_rate": 1.0366145576157561e-07,
+      "loss": 0.4411,
+      "step": 13495
+    },
+    {
+      "epoch": 0.9370920705457575,
+      "grad_norm": 5.431260756563639,
+      "learning_rate": 1.0343379177927216e-07,
+      "loss": 0.6711,
+      "step": 13496
+    },
+    {
+      "epoch": 0.9371615053464797,
+      "grad_norm": 5.361559508726321,
+      "learning_rate": 1.0320637545974565e-07,
+      "loss": 0.5056,
+      "step": 13497
+    },
+    {
+      "epoch": 0.9372309401472018,
+      "grad_norm": 4.041141580723797,
+      "learning_rate": 1.0297920681450135e-07,
+      "loss": 0.3833,
+      "step": 13498
+    },
+    {
+      "epoch": 0.937300374947924,
+      "grad_norm": 3.6765910103796267,
+      "learning_rate": 1.0275228585502561e-07,
+      "loss": 0.4067,
+      "step": 13499
+    },
+    {
+      "epoch": 0.937369809748646,
+      "grad_norm": 4.064952658681537,
+      "learning_rate": 1.0252561259279869e-07,
+      "loss": 0.4555,
+      "step": 13500
+    },
+    {
+      "epoch": 0.9374392445493681,
+      "grad_norm": 3.5411392097851335,
+      "learning_rate": 1.0229918703928476e-07,
+      "loss": 0.4584,
+      "step": 13501
+    },
+    {
+      "epoch": 0.9375086793500903,
+      "grad_norm": 4.4878025421574455,
+      "learning_rate": 1.0207300920593521e-07,
+      "loss": 0.5027,
+      "step": 13502
+    },
+    {
+      "epoch": 0.9375781141508124,
+      "grad_norm": 4.813316529581607,
+      "learning_rate": 1.0184707910419034e-07,
+      "loss": 0.361,
+      "step": 13503
+    },
+    {
+      "epoch": 0.9376475489515345,
+      "grad_norm": 3.8300531472896817,
+      "learning_rate": 1.0162139674547767e-07,
+      "loss": 0.4098,
+      "step": 13504
+    },
+    {
+      "epoch": 0.9377169837522566,
+      "grad_norm": 3.84495921170896,
+      "learning_rate": 1.0139596214121194e-07,
+      "loss": 0.3125,
+      "step": 13505
+    },
+    {
+      "epoch": 0.9377864185529787,
+      "grad_norm": 3.7938559796447957,
+      "learning_rate": 1.0117077530279573e-07,
+      "loss": 0.3513,
+      "step": 13506
+    },
+    {
+      "epoch": 0.9378558533537009,
+      "grad_norm": 3.921435496293258,
+      "learning_rate": 1.009458362416177e-07,
+      "loss": 0.3348,
+      "step": 13507
+    },
+    {
+      "epoch": 0.937925288154423,
+      "grad_norm": 4.391843412748748,
+      "learning_rate": 1.0072114496905594e-07,
+      "loss": 0.5073,
+      "step": 13508
+    },
+    {
+      "epoch": 0.9379947229551451,
+      "grad_norm": 2.8221074311109873,
+      "learning_rate": 1.0049670149647472e-07,
+      "loss": 0.2055,
+      "step": 13509
+    },
+    {
+      "epoch": 0.9380641577558673,
+      "grad_norm": 3.783487249674332,
+      "learning_rate": 1.0027250583522608e-07,
+      "loss": 0.4454,
+      "step": 13510
+    },
+    {
+      "epoch": 0.9381335925565893,
+      "grad_norm": 3.781246958037558,
+      "learning_rate": 1.0004855799664925e-07,
+      "loss": 0.3285,
+      "step": 13511
+    },
+    {
+      "epoch": 0.9382030273573115,
+      "grad_norm": 4.1040878131815015,
+      "learning_rate": 9.982485799207297e-08,
+      "loss": 0.3678,
+      "step": 13512
+    },
+    {
+      "epoch": 0.9382724621580336,
+      "grad_norm": 4.162139424537777,
+      "learning_rate": 9.960140583280931e-08,
+      "loss": 0.3741,
+      "step": 13513
+    },
+    {
+      "epoch": 0.9383418969587557,
+      "grad_norm": 3.469408287752279,
+      "learning_rate": 9.937820153016142e-08,
+      "loss": 0.4232,
+      "step": 13514
+    },
+    {
+      "epoch": 0.9384113317594779,
+      "grad_norm": 4.556789795129862,
+      "learning_rate": 9.915524509541918e-08,
+      "loss": 0.2777,
+      "step": 13515
+    },
+    {
+      "epoch": 0.9384807665602,
+      "grad_norm": 3.489778945188901,
+      "learning_rate": 9.893253653985912e-08,
+      "loss": 0.2998,
+      "step": 13516
+    },
+    {
+      "epoch": 0.938550201360922,
+      "grad_norm": 4.285413433780867,
+      "learning_rate": 9.871007587474446e-08,
+      "loss": 0.3929,
+      "step": 13517
+    },
+    {
+      "epoch": 0.9386196361616442,
+      "grad_norm": 3.9591833749435175,
+      "learning_rate": 9.84878631113284e-08,
+      "loss": 0.3959,
+      "step": 13518
+    },
+    {
+      "epoch": 0.9386890709623663,
+      "grad_norm": 5.387392372611148,
+      "learning_rate": 9.826589826085031e-08,
+      "loss": 0.4413,
+      "step": 13519
+    },
+    {
+      "epoch": 0.9387585057630885,
+      "grad_norm": 2.8635993681352456,
+      "learning_rate": 9.804418133453619e-08,
+      "loss": 0.1934,
+      "step": 13520
+    },
+    {
+      "epoch": 0.9388279405638106,
+      "grad_norm": 4.866964687212404,
+      "learning_rate": 9.782271234359986e-08,
+      "loss": 0.5989,
+      "step": 13521
+    },
+    {
+      "epoch": 0.9388973753645327,
+      "grad_norm": 4.643351610670739,
+      "learning_rate": 9.760149129924402e-08,
+      "loss": 0.338,
+      "step": 13522
+    },
+    {
+      "epoch": 0.9389668101652549,
+      "grad_norm": 2.906174147518404,
+      "learning_rate": 9.738051821265748e-08,
+      "loss": 0.2335,
+      "step": 13523
+    },
+    {
+      "epoch": 0.9390362449659769,
+      "grad_norm": 3.5292706143830177,
+      "learning_rate": 9.715979309501633e-08,
+      "loss": 0.3101,
+      "step": 13524
+    },
+    {
+      "epoch": 0.9391056797666991,
+      "grad_norm": 4.745984302493065,
+      "learning_rate": 9.693931595748495e-08,
+      "loss": 0.4334,
+      "step": 13525
+    },
+    {
+      "epoch": 0.9391751145674212,
+      "grad_norm": 5.195262732164282,
+      "learning_rate": 9.6719086811215e-08,
+      "loss": 0.7283,
+      "step": 13526
+    },
+    {
+      "epoch": 0.9392445493681433,
+      "grad_norm": 2.9315489056034716,
+      "learning_rate": 9.649910566734533e-08,
+      "loss": 0.1775,
+      "step": 13527
+    },
+    {
+      "epoch": 0.9393139841688655,
+      "grad_norm": 3.4572051785343763,
+      "learning_rate": 9.627937253700148e-08,
+      "loss": 0.5215,
+      "step": 13528
+    },
+    {
+      "epoch": 0.9393834189695875,
+      "grad_norm": 4.312134973301884,
+      "learning_rate": 9.605988743129846e-08,
+      "loss": 0.3665,
+      "step": 13529
+    },
+    {
+      "epoch": 0.9394528537703096,
+      "grad_norm": 3.9660749592026545,
+      "learning_rate": 9.584065036133738e-08,
+      "loss": 0.434,
+      "step": 13530
+    },
+    {
+      "epoch": 0.9395222885710318,
+      "grad_norm": 4.941213570084439,
+      "learning_rate": 9.562166133820605e-08,
+      "loss": 0.5513,
+      "step": 13531
+    },
+    {
+      "epoch": 0.9395917233717539,
+      "grad_norm": 2.726472287545314,
+      "learning_rate": 9.54029203729817e-08,
+      "loss": 0.3021,
+      "step": 13532
+    },
+    {
+      "epoch": 0.9396611581724761,
+      "grad_norm": 3.0720416731053777,
+      "learning_rate": 9.518442747672884e-08,
+      "loss": 0.3117,
+      "step": 13533
+    },
+    {
+      "epoch": 0.9397305929731982,
+      "grad_norm": 3.187475375582225,
+      "learning_rate": 9.496618266049585e-08,
+      "loss": 0.3082,
+      "step": 13534
+    },
+    {
+      "epoch": 0.9398000277739202,
+      "grad_norm": 4.366750879342117,
+      "learning_rate": 9.474818593532387e-08,
+      "loss": 0.2929,
+      "step": 13535
+    },
+    {
+      "epoch": 0.9398694625746424,
+      "grad_norm": 3.7075092609496796,
+      "learning_rate": 9.453043731223744e-08,
+      "loss": 0.3566,
+      "step": 13536
+    },
+    {
+      "epoch": 0.9399388973753645,
+      "grad_norm": 4.5483399104013,
+      "learning_rate": 9.431293680225107e-08,
+      "loss": 0.301,
+      "step": 13537
+    },
+    {
+      "epoch": 0.9400083321760867,
+      "grad_norm": 2.677531624056893,
+      "learning_rate": 9.409568441636486e-08,
+      "loss": 0.1611,
+      "step": 13538
+    },
+    {
+      "epoch": 0.9400777669768088,
+      "grad_norm": 4.087009267678147,
+      "learning_rate": 9.387868016556778e-08,
+      "loss": 0.4068,
+      "step": 13539
+    },
+    {
+      "epoch": 0.9401472017775309,
+      "grad_norm": 5.262953592620084,
+      "learning_rate": 9.366192406083496e-08,
+      "loss": 0.6833,
+      "step": 13540
+    },
+    {
+      "epoch": 0.940216636578253,
+      "grad_norm": 3.8583115109245805,
+      "learning_rate": 9.344541611313097e-08,
+      "loss": 0.3615,
+      "step": 13541
+    },
+    {
+      "epoch": 0.9402860713789751,
+      "grad_norm": 3.554101766266527,
+      "learning_rate": 9.322915633340478e-08,
+      "loss": 0.3903,
+      "step": 13542
+    },
+    {
+      "epoch": 0.9403555061796972,
+      "grad_norm": 3.349736085075466,
+      "learning_rate": 9.301314473259548e-08,
+      "loss": 0.3232,
+      "step": 13543
+    },
+    {
+      "epoch": 0.9404249409804194,
+      "grad_norm": 3.43454475157128,
+      "learning_rate": 9.279738132162986e-08,
+      "loss": 0.3465,
+      "step": 13544
+    },
+    {
+      "epoch": 0.9404943757811415,
+      "grad_norm": 4.656818229672937,
+      "learning_rate": 9.258186611141861e-08,
+      "loss": 0.6034,
+      "step": 13545
+    },
+    {
+      "epoch": 0.9405638105818637,
+      "grad_norm": 4.311806484004886,
+      "learning_rate": 9.23665991128636e-08,
+      "loss": 0.4458,
+      "step": 13546
+    },
+    {
+      "epoch": 0.9406332453825857,
+      "grad_norm": 3.46112641615352,
+      "learning_rate": 9.215158033685279e-08,
+      "loss": 0.4616,
+      "step": 13547
+    },
+    {
+      "epoch": 0.9407026801833078,
+      "grad_norm": 6.5717914993356565,
+      "learning_rate": 9.193680979426189e-08,
+      "loss": 0.4539,
+      "step": 13548
+    },
+    {
+      "epoch": 0.94077211498403,
+      "grad_norm": 4.4687906266355055,
+      "learning_rate": 9.17222874959528e-08,
+      "loss": 0.3101,
+      "step": 13549
+    },
+    {
+      "epoch": 0.9408415497847521,
+      "grad_norm": 4.970795925362148,
+      "learning_rate": 9.150801345277627e-08,
+      "loss": 0.7092,
+      "step": 13550
+    },
+    {
+      "epoch": 0.9409109845854743,
+      "grad_norm": 4.488446301791286,
+      "learning_rate": 9.12939876755703e-08,
+      "loss": 0.6194,
+      "step": 13551
+    },
+    {
+      "epoch": 0.9409804193861964,
+      "grad_norm": 4.310892199117376,
+      "learning_rate": 9.108021017516011e-08,
+      "loss": 0.4281,
+      "step": 13552
+    },
+    {
+      "epoch": 0.9410498541869184,
+      "grad_norm": 4.119755353454573,
+      "learning_rate": 9.086668096235763e-08,
+      "loss": 0.3699,
+      "step": 13553
+    },
+    {
+      "epoch": 0.9411192889876406,
+      "grad_norm": 3.4131732832625707,
+      "learning_rate": 9.065340004796363e-08,
+      "loss": 0.3681,
+      "step": 13554
+    },
+    {
+      "epoch": 0.9411887237883627,
+      "grad_norm": 4.158828961030812,
+      "learning_rate": 9.044036744276563e-08,
+      "loss": 0.4205,
+      "step": 13555
+    },
+    {
+      "epoch": 0.9412581585890849,
+      "grad_norm": 4.227836134912753,
+      "learning_rate": 9.022758315753777e-08,
+      "loss": 0.4319,
+      "step": 13556
+    },
+    {
+      "epoch": 0.941327593389807,
+      "grad_norm": 3.5729394484009966,
+      "learning_rate": 9.001504720304255e-08,
+      "loss": 0.3436,
+      "step": 13557
+    },
+    {
+      "epoch": 0.9413970281905291,
+      "grad_norm": 3.606500028263425,
+      "learning_rate": 8.980275959003137e-08,
+      "loss": 0.4628,
+      "step": 13558
+    },
+    {
+      "epoch": 0.9414664629912513,
+      "grad_norm": 4.87273642752597,
+      "learning_rate": 8.959072032923955e-08,
+      "loss": 0.4831,
+      "step": 13559
+    },
+    {
+      "epoch": 0.9415358977919733,
+      "grad_norm": 3.2680077011066597,
+      "learning_rate": 8.937892943139293e-08,
+      "loss": 0.2821,
+      "step": 13560
+    },
+    {
+      "epoch": 0.9416053325926954,
+      "grad_norm": 5.912828145509751,
+      "learning_rate": 8.916738690720294e-08,
+      "loss": 0.4134,
+      "step": 13561
+    },
+    {
+      "epoch": 0.9416747673934176,
+      "grad_norm": 3.1993500918691784,
+      "learning_rate": 8.89560927673705e-08,
+      "loss": 0.2994,
+      "step": 13562
+    },
+    {
+      "epoch": 0.9417442021941397,
+      "grad_norm": 3.788071428517056,
+      "learning_rate": 8.874504702258035e-08,
+      "loss": 0.3698,
+      "step": 13563
+    },
+    {
+      "epoch": 0.9418136369948619,
+      "grad_norm": 4.688298209887736,
+      "learning_rate": 8.853424968350898e-08,
+      "loss": 0.4245,
+      "step": 13564
+    },
+    {
+      "epoch": 0.941883071795584,
+      "grad_norm": 4.308820410261503,
+      "learning_rate": 8.832370076081732e-08,
+      "loss": 0.4014,
+      "step": 13565
+    },
+    {
+      "epoch": 0.941952506596306,
+      "grad_norm": 4.5693004033117885,
+      "learning_rate": 8.811340026515514e-08,
+      "loss": 0.6363,
+      "step": 13566
+    },
+    {
+      "epoch": 0.9420219413970282,
+      "grad_norm": 3.283505656763116,
+      "learning_rate": 8.790334820715896e-08,
+      "loss": 0.3479,
+      "step": 13567
+    },
+    {
+      "epoch": 0.9420913761977503,
+      "grad_norm": 5.464350971884272,
+      "learning_rate": 8.769354459745305e-08,
+      "loss": 0.9185,
+      "step": 13568
+    },
+    {
+      "epoch": 0.9421608109984725,
+      "grad_norm": 3.9863662020789072,
+      "learning_rate": 8.748398944664893e-08,
+      "loss": 0.5141,
+      "step": 13569
+    },
+    {
+      "epoch": 0.9422302457991946,
+      "grad_norm": 4.719082102141583,
+      "learning_rate": 8.72746827653459e-08,
+      "loss": 0.4558,
+      "step": 13570
+    },
+    {
+      "epoch": 0.9422996805999166,
+      "grad_norm": 4.599825972209393,
+      "learning_rate": 8.706562456412993e-08,
+      "loss": 0.4936,
+      "step": 13571
+    },
+    {
+      "epoch": 0.9423691154006388,
+      "grad_norm": 5.569131630842515,
+      "learning_rate": 8.685681485357533e-08,
+      "loss": 0.4829,
+      "step": 13572
+    },
+    {
+      "epoch": 0.9424385502013609,
+      "grad_norm": 4.767376147783786,
+      "learning_rate": 8.664825364424423e-08,
+      "loss": 0.526,
+      "step": 13573
+    },
+    {
+      "epoch": 0.942507985002083,
+      "grad_norm": 2.891391760687909,
+      "learning_rate": 8.643994094668373e-08,
+      "loss": 0.1826,
+      "step": 13574
+    },
+    {
+      "epoch": 0.9425774198028052,
+      "grad_norm": 4.463103628812413,
+      "learning_rate": 8.623187677143097e-08,
+      "loss": 0.4217,
+      "step": 13575
+    },
+    {
+      "epoch": 0.9426468546035273,
+      "grad_norm": 4.412614767505137,
+      "learning_rate": 8.602406112901029e-08,
+      "loss": 0.3742,
+      "step": 13576
+    },
+    {
+      "epoch": 0.9427162894042495,
+      "grad_norm": 4.148445632782792,
+      "learning_rate": 8.581649402993164e-08,
+      "loss": 0.6073,
+      "step": 13577
+    },
+    {
+      "epoch": 0.9427857242049715,
+      "grad_norm": 3.3867396065274824,
+      "learning_rate": 8.560917548469328e-08,
+      "loss": 0.4722,
+      "step": 13578
+    },
+    {
+      "epoch": 0.9428551590056936,
+      "grad_norm": 3.4053379611300403,
+      "learning_rate": 8.540210550378291e-08,
+      "loss": 0.3951,
+      "step": 13579
+    },
+    {
+      "epoch": 0.9429245938064158,
+      "grad_norm": 4.337626772882801,
+      "learning_rate": 8.519528409767219e-08,
+      "loss": 0.4722,
+      "step": 13580
+    },
+    {
+      "epoch": 0.9429940286071379,
+      "grad_norm": 5.557160679922473,
+      "learning_rate": 8.498871127682273e-08,
+      "loss": 0.5584,
+      "step": 13581
+    },
+    {
+      "epoch": 0.9430634634078601,
+      "grad_norm": 4.287362822264059,
+      "learning_rate": 8.47823870516823e-08,
+      "loss": 0.4428,
+      "step": 13582
+    },
+    {
+      "epoch": 0.9431328982085821,
+      "grad_norm": 3.9072783643594406,
+      "learning_rate": 8.457631143268696e-08,
+      "loss": 0.3192,
+      "step": 13583
+    },
+    {
+      "epoch": 0.9432023330093042,
+      "grad_norm": 3.2883531684664113,
+      "learning_rate": 8.437048443026008e-08,
+      "loss": 0.3507,
+      "step": 13584
+    },
+    {
+      "epoch": 0.9432717678100264,
+      "grad_norm": 5.314703373564217,
+      "learning_rate": 8.41649060548111e-08,
+      "loss": 0.7017,
+      "step": 13585
+    },
+    {
+      "epoch": 0.9433412026107485,
+      "grad_norm": 3.2570621722187645,
+      "learning_rate": 8.39595763167389e-08,
+      "loss": 0.2558,
+      "step": 13586
+    },
+    {
+      "epoch": 0.9434106374114706,
+      "grad_norm": 3.3661333885657276,
+      "learning_rate": 8.375449522642798e-08,
+      "loss": 0.3303,
+      "step": 13587
+    },
+    {
+      "epoch": 0.9434800722121928,
+      "grad_norm": 3.470174114466503,
+      "learning_rate": 8.35496627942517e-08,
+      "loss": 0.2422,
+      "step": 13588
+    },
+    {
+      "epoch": 0.9435495070129148,
+      "grad_norm": 3.878313362439,
+      "learning_rate": 8.33450790305701e-08,
+      "loss": 0.5227,
+      "step": 13589
+    },
+    {
+      "epoch": 0.943618941813637,
+      "grad_norm": 3.5809938880428884,
+      "learning_rate": 8.314074394573102e-08,
+      "loss": 0.441,
+      "step": 13590
+    },
+    {
+      "epoch": 0.9436883766143591,
+      "grad_norm": 4.175595098868105,
+      "learning_rate": 8.293665755006952e-08,
+      "loss": 0.3558,
+      "step": 13591
+    },
+    {
+      "epoch": 0.9437578114150812,
+      "grad_norm": 4.350352064433606,
+      "learning_rate": 8.273281985390736e-08,
+      "loss": 0.4766,
+      "step": 13592
+    },
+    {
+      "epoch": 0.9438272462158034,
+      "grad_norm": 4.693164219525063,
+      "learning_rate": 8.252923086755515e-08,
+      "loss": 0.3884,
+      "step": 13593
+    },
+    {
+      "epoch": 0.9438966810165255,
+      "grad_norm": 4.43000429430502,
+      "learning_rate": 8.232589060131024e-08,
+      "loss": 0.4014,
+      "step": 13594
+    },
+    {
+      "epoch": 0.9439661158172477,
+      "grad_norm": 4.18384862514878,
+      "learning_rate": 8.212279906545661e-08,
+      "loss": 0.3991,
+      "step": 13595
+    },
+    {
+      "epoch": 0.9440355506179697,
+      "grad_norm": 3.8866349778943845,
+      "learning_rate": 8.191995627026716e-08,
+      "loss": 0.2645,
+      "step": 13596
+    },
+    {
+      "epoch": 0.9441049854186918,
+      "grad_norm": 3.835499506881386,
+      "learning_rate": 8.17173622260009e-08,
+      "loss": 0.2673,
+      "step": 13597
+    },
+    {
+      "epoch": 0.944174420219414,
+      "grad_norm": 2.7804260265165714,
+      "learning_rate": 8.15150169429052e-08,
+      "loss": 0.1906,
+      "step": 13598
+    },
+    {
+      "epoch": 0.9442438550201361,
+      "grad_norm": 3.726927363715068,
+      "learning_rate": 8.131292043121408e-08,
+      "loss": 0.3155,
+      "step": 13599
+    },
+    {
+      "epoch": 0.9443132898208582,
+      "grad_norm": 2.636932593239316,
+      "learning_rate": 8.111107270114993e-08,
+      "loss": 0.29,
+      "step": 13600
+    },
+    {
+      "epoch": 0.9443827246215803,
+      "grad_norm": 5.075182793939531,
+      "learning_rate": 8.090947376292125e-08,
+      "loss": 0.4154,
+      "step": 13601
+    },
+    {
+      "epoch": 0.9444521594223024,
+      "grad_norm": 4.516224788902047,
+      "learning_rate": 8.070812362672541e-08,
+      "loss": 0.5251,
+      "step": 13602
+    },
+    {
+      "epoch": 0.9445215942230246,
+      "grad_norm": 3.01062600771442,
+      "learning_rate": 8.050702230274598e-08,
+      "loss": 0.2955,
+      "step": 13603
+    },
+    {
+      "epoch": 0.9445910290237467,
+      "grad_norm": 4.732334170640067,
+      "learning_rate": 8.03061698011537e-08,
+      "loss": 0.6322,
+      "step": 13604
+    },
+    {
+      "epoch": 0.9446604638244688,
+      "grad_norm": 7.58345074903841,
+      "learning_rate": 8.010556613210985e-08,
+      "loss": 0.5992,
+      "step": 13605
+    },
+    {
+      "epoch": 0.944729898625191,
+      "grad_norm": 5.151881953073426,
+      "learning_rate": 7.990521130575856e-08,
+      "loss": 0.5578,
+      "step": 13606
+    },
+    {
+      "epoch": 0.944799333425913,
+      "grad_norm": 4.599824654385063,
+      "learning_rate": 7.970510533223396e-08,
+      "loss": 0.3198,
+      "step": 13607
+    },
+    {
+      "epoch": 0.9448687682266352,
+      "grad_norm": 4.76039793827116,
+      "learning_rate": 7.950524822165794e-08,
+      "loss": 0.4448,
+      "step": 13608
+    },
+    {
+      "epoch": 0.9449382030273573,
+      "grad_norm": 3.1831429451178495,
+      "learning_rate": 7.930563998413798e-08,
+      "loss": 0.3076,
+      "step": 13609
+    },
+    {
+      "epoch": 0.9450076378280794,
+      "grad_norm": 3.331479866424242,
+      "learning_rate": 7.9106280629771e-08,
+      "loss": 0.3245,
+      "step": 13610
+    },
+    {
+      "epoch": 0.9450770726288016,
+      "grad_norm": 3.4857400326875596,
+      "learning_rate": 7.890717016864003e-08,
+      "loss": 0.2541,
+      "step": 13611
+    },
+    {
+      "epoch": 0.9451465074295237,
+      "grad_norm": 4.380912193402173,
+      "learning_rate": 7.870830861081535e-08,
+      "loss": 0.3585,
+      "step": 13612
+    },
+    {
+      "epoch": 0.9452159422302459,
+      "grad_norm": 4.0616374755369336,
+      "learning_rate": 7.850969596635616e-08,
+      "loss": 0.3385,
+      "step": 13613
+    },
+    {
+      "epoch": 0.9452853770309679,
+      "grad_norm": 4.30601052713788,
+      "learning_rate": 7.831133224530718e-08,
+      "loss": 0.5041,
+      "step": 13614
+    },
+    {
+      "epoch": 0.94535481183169,
+      "grad_norm": 4.180230425362507,
+      "learning_rate": 7.811321745770206e-08,
+      "loss": 0.568,
+      "step": 13615
+    },
+    {
+      "epoch": 0.9454242466324122,
+      "grad_norm": 3.3222430442037307,
+      "learning_rate": 7.791535161356057e-08,
+      "loss": 0.3584,
+      "step": 13616
+    },
+    {
+      "epoch": 0.9454936814331343,
+      "grad_norm": 4.2963917001234035,
+      "learning_rate": 7.771773472289079e-08,
+      "loss": 0.3711,
+      "step": 13617
+    },
+    {
+      "epoch": 0.9455631162338564,
+      "grad_norm": 3.3858535642324203,
+      "learning_rate": 7.752036679568809e-08,
+      "loss": 0.261,
+      "step": 13618
+    },
+    {
+      "epoch": 0.9456325510345786,
+      "grad_norm": 5.180334421663684,
+      "learning_rate": 7.732324784193556e-08,
+      "loss": 0.4812,
+      "step": 13619
+    },
+    {
+      "epoch": 0.9457019858353006,
+      "grad_norm": 3.7477264181699623,
+      "learning_rate": 7.712637787160249e-08,
+      "loss": 0.3107,
+      "step": 13620
+    },
+    {
+      "epoch": 0.9457714206360228,
+      "grad_norm": 3.5983490600353543,
+      "learning_rate": 7.692975689464643e-08,
+      "loss": 0.3771,
+      "step": 13621
+    },
+    {
+      "epoch": 0.9458408554367449,
+      "grad_norm": 4.630320842483007,
+      "learning_rate": 7.673338492101278e-08,
+      "loss": 0.5198,
+      "step": 13622
+    },
+    {
+      "epoch": 0.945910290237467,
+      "grad_norm": 3.7944668473495917,
+      "learning_rate": 7.65372619606336e-08,
+      "loss": 0.3566,
+      "step": 13623
+    },
+    {
+      "epoch": 0.9459797250381892,
+      "grad_norm": 4.344666272104659,
+      "learning_rate": 7.634138802342761e-08,
+      "loss": 0.4359,
+      "step": 13624
+    },
+    {
+      "epoch": 0.9460491598389112,
+      "grad_norm": 3.8892797514625874,
+      "learning_rate": 7.6145763119303e-08,
+      "loss": 0.5293,
+      "step": 13625
+    },
+    {
+      "epoch": 0.9461185946396334,
+      "grad_norm": 2.868281995099764,
+      "learning_rate": 7.595038725815463e-08,
+      "loss": 0.3005,
+      "step": 13626
+    },
+    {
+      "epoch": 0.9461880294403555,
+      "grad_norm": 4.6190484295314596,
+      "learning_rate": 7.575526044986293e-08,
+      "loss": 0.616,
+      "step": 13627
+    },
+    {
+      "epoch": 0.9462574642410776,
+      "grad_norm": 4.131773625339716,
+      "learning_rate": 7.556038270429833e-08,
+      "loss": 0.5011,
+      "step": 13628
+    },
+    {
+      "epoch": 0.9463268990417998,
+      "grad_norm": 3.551755689602933,
+      "learning_rate": 7.536575403131741e-08,
+      "loss": 0.3168,
+      "step": 13629
+    },
+    {
+      "epoch": 0.9463963338425219,
+      "grad_norm": 4.26427803506579,
+      "learning_rate": 7.517137444076394e-08,
+      "loss": 0.3911,
+      "step": 13630
+    },
+    {
+      "epoch": 0.9464657686432439,
+      "grad_norm": 3.67287463096618,
+      "learning_rate": 7.497724394246952e-08,
+      "loss": 0.4951,
+      "step": 13631
+    },
+    {
+      "epoch": 0.9465352034439661,
+      "grad_norm": 3.9445375479516085,
+      "learning_rate": 7.478336254625296e-08,
+      "loss": 0.4539,
+      "step": 13632
+    },
+    {
+      "epoch": 0.9466046382446882,
+      "grad_norm": 3.113842843150875,
+      "learning_rate": 7.458973026192085e-08,
+      "loss": 0.2598,
+      "step": 13633
+    },
+    {
+      "epoch": 0.9466740730454104,
+      "grad_norm": 3.8740852515980846,
+      "learning_rate": 7.439634709926647e-08,
+      "loss": 0.3621,
+      "step": 13634
+    },
+    {
+      "epoch": 0.9467435078461325,
+      "grad_norm": 5.805722662976317,
+      "learning_rate": 7.420321306807143e-08,
+      "loss": 0.5896,
+      "step": 13635
+    },
+    {
+      "epoch": 0.9468129426468546,
+      "grad_norm": 3.461579509130838,
+      "learning_rate": 7.401032817810294e-08,
+      "loss": 0.3879,
+      "step": 13636
+    },
+    {
+      "epoch": 0.9468823774475768,
+      "grad_norm": 3.724007203079931,
+      "learning_rate": 7.38176924391193e-08,
+      "loss": 0.5304,
+      "step": 13637
+    },
+    {
+      "epoch": 0.9469518122482988,
+      "grad_norm": 4.182345340163832,
+      "learning_rate": 7.362530586086159e-08,
+      "loss": 0.5385,
+      "step": 13638
+    },
+    {
+      "epoch": 0.947021247049021,
+      "grad_norm": 4.294224826126223,
+      "learning_rate": 7.343316845306148e-08,
+      "loss": 0.5147,
+      "step": 13639
+    },
+    {
+      "epoch": 0.9470906818497431,
+      "grad_norm": 4.439940147242975,
+      "learning_rate": 7.32412802254373e-08,
+      "loss": 0.598,
+      "step": 13640
+    },
+    {
+      "epoch": 0.9471601166504652,
+      "grad_norm": 4.51005899672905,
+      "learning_rate": 7.304964118769353e-08,
+      "loss": 0.4493,
+      "step": 13641
+    },
+    {
+      "epoch": 0.9472295514511874,
+      "grad_norm": 3.928415110307402,
+      "learning_rate": 7.285825134952463e-08,
+      "loss": 0.3015,
+      "step": 13642
+    },
+    {
+      "epoch": 0.9472989862519094,
+      "grad_norm": 3.4409527267050555,
+      "learning_rate": 7.266711072060895e-08,
+      "loss": 0.334,
+      "step": 13643
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 5.13573259161139,
+      "learning_rate": 7.247621931061599e-08,
+      "loss": 0.7405,
+      "step": 13644
+    },
+    {
+      "epoch": 0.9474378558533537,
+      "grad_norm": 4.480573953211173,
+      "learning_rate": 7.228557712919914e-08,
+      "loss": 0.4192,
+      "step": 13645
+    },
+    {
+      "epoch": 0.9475072906540758,
+      "grad_norm": 3.8821490364372315,
+      "learning_rate": 7.209518418600236e-08,
+      "loss": 0.4955,
+      "step": 13646
+    },
+    {
+      "epoch": 0.947576725454798,
+      "grad_norm": 3.355088879117939,
+      "learning_rate": 7.190504049065462e-08,
+      "loss": 0.2372,
+      "step": 13647
+    },
+    {
+      "epoch": 0.9476461602555201,
+      "grad_norm": 5.5295241750359025,
+      "learning_rate": 7.171514605277374e-08,
+      "loss": 0.6387,
+      "step": 13648
+    },
+    {
+      "epoch": 0.9477155950562421,
+      "grad_norm": 3.714547535016453,
+      "learning_rate": 7.152550088196375e-08,
+      "loss": 0.4036,
+      "step": 13649
+    },
+    {
+      "epoch": 0.9477850298569643,
+      "grad_norm": 4.46452770464274,
+      "learning_rate": 7.13361049878164e-08,
+      "loss": 0.5211,
+      "step": 13650
+    },
+    {
+      "epoch": 0.9478544646576864,
+      "grad_norm": 4.312987851264458,
+      "learning_rate": 7.114695837991293e-08,
+      "loss": 0.4387,
+      "step": 13651
+    },
+    {
+      "epoch": 0.9479238994584086,
+      "grad_norm": 4.731264961559913,
+      "learning_rate": 7.095806106781788e-08,
+      "loss": 0.5521,
+      "step": 13652
+    },
+    {
+      "epoch": 0.9479933342591307,
+      "grad_norm": 4.058083029633025,
+      "learning_rate": 7.076941306108698e-08,
+      "loss": 0.3503,
+      "step": 13653
+    },
+    {
+      "epoch": 0.9480627690598528,
+      "grad_norm": 4.75617517251498,
+      "learning_rate": 7.05810143692609e-08,
+      "loss": 0.5376,
+      "step": 13654
+    },
+    {
+      "epoch": 0.948132203860575,
+      "grad_norm": 3.3126038554675783,
+      "learning_rate": 7.039286500187037e-08,
+      "loss": 0.2923,
+      "step": 13655
+    },
+    {
+      "epoch": 0.948201638661297,
+      "grad_norm": 4.1791718164097045,
+      "learning_rate": 7.020496496842888e-08,
+      "loss": 0.4386,
+      "step": 13656
+    },
+    {
+      "epoch": 0.9482710734620191,
+      "grad_norm": 4.218528832283942,
+      "learning_rate": 7.001731427844272e-08,
+      "loss": 0.3796,
+      "step": 13657
+    },
+    {
+      "epoch": 0.9483405082627413,
+      "grad_norm": 3.6979818182633526,
+      "learning_rate": 6.982991294140262e-08,
+      "loss": 0.2869,
+      "step": 13658
+    },
+    {
+      "epoch": 0.9484099430634634,
+      "grad_norm": 2.35607874179691,
+      "learning_rate": 6.964276096678546e-08,
+      "loss": 0.1814,
+      "step": 13659
+    },
+    {
+      "epoch": 0.9484793778641856,
+      "grad_norm": 4.123098524745145,
+      "learning_rate": 6.945585836405922e-08,
+      "loss": 0.435,
+      "step": 13660
+    },
+    {
+      "epoch": 0.9485488126649076,
+      "grad_norm": 4.655868099093231,
+      "learning_rate": 6.926920514267576e-08,
+      "loss": 0.5642,
+      "step": 13661
+    },
+    {
+      "epoch": 0.9486182474656297,
+      "grad_norm": 2.598059996934651,
+      "learning_rate": 6.908280131207701e-08,
+      "loss": 0.1911,
+      "step": 13662
+    },
+    {
+      "epoch": 0.9486876822663519,
+      "grad_norm": 4.457444072110447,
+      "learning_rate": 6.889664688168984e-08,
+      "loss": 0.6129,
+      "step": 13663
+    },
+    {
+      "epoch": 0.948757117067074,
+      "grad_norm": 3.315774232290024,
+      "learning_rate": 6.871074186093118e-08,
+      "loss": 0.2844,
+      "step": 13664
+    },
+    {
+      "epoch": 0.9488265518677962,
+      "grad_norm": 3.927788001597242,
+      "learning_rate": 6.852508625920184e-08,
+      "loss": 0.3313,
+      "step": 13665
+    },
+    {
+      "epoch": 0.9488959866685183,
+      "grad_norm": 3.8207391627188154,
+      "learning_rate": 6.833968008589486e-08,
+      "loss": 0.346,
+      "step": 13666
+    },
+    {
+      "epoch": 0.9489654214692403,
+      "grad_norm": 3.276000249567987,
+      "learning_rate": 6.815452335038552e-08,
+      "loss": 0.2086,
+      "step": 13667
+    },
+    {
+      "epoch": 0.9490348562699625,
+      "grad_norm": 3.372665226802382,
+      "learning_rate": 6.796961606203912e-08,
+      "loss": 0.6035,
+      "step": 13668
+    },
+    {
+      "epoch": 0.9491042910706846,
+      "grad_norm": 3.5885233580748275,
+      "learning_rate": 6.778495823020925e-08,
+      "loss": 0.4418,
+      "step": 13669
+    },
+    {
+      "epoch": 0.9491737258714068,
+      "grad_norm": 3.704607490842046,
+      "learning_rate": 6.760054986423459e-08,
+      "loss": 0.4242,
+      "step": 13670
+    },
+    {
+      "epoch": 0.9492431606721289,
+      "grad_norm": 5.301991881890821,
+      "learning_rate": 6.741639097344322e-08,
+      "loss": 0.5544,
+      "step": 13671
+    },
+    {
+      "epoch": 0.949312595472851,
+      "grad_norm": 3.1617610184585563,
+      "learning_rate": 6.723248156714879e-08,
+      "loss": 0.2743,
+      "step": 13672
+    },
+    {
+      "epoch": 0.9493820302735732,
+      "grad_norm": 2.9273173446321006,
+      "learning_rate": 6.704882165465388e-08,
+      "loss": 0.2733,
+      "step": 13673
+    },
+    {
+      "epoch": 0.9494514650742952,
+      "grad_norm": 4.0490038541021915,
+      "learning_rate": 6.686541124524715e-08,
+      "loss": 0.3737,
+      "step": 13674
+    },
+    {
+      "epoch": 0.9495208998750173,
+      "grad_norm": 3.5924407858174585,
+      "learning_rate": 6.668225034820619e-08,
+      "loss": 0.4399,
+      "step": 13675
+    },
+    {
+      "epoch": 0.9495903346757395,
+      "grad_norm": 3.43117871428001,
+      "learning_rate": 6.649933897279471e-08,
+      "loss": 0.3219,
+      "step": 13676
+    },
+    {
+      "epoch": 0.9496597694764616,
+      "grad_norm": 3.7217145665656006,
+      "learning_rate": 6.631667712826362e-08,
+      "loss": 0.2544,
+      "step": 13677
+    },
+    {
+      "epoch": 0.9497292042771838,
+      "grad_norm": 3.758525541437729,
+      "learning_rate": 6.613426482385222e-08,
+      "loss": 0.3291,
+      "step": 13678
+    },
+    {
+      "epoch": 0.9497986390779058,
+      "grad_norm": 4.258746501467238,
+      "learning_rate": 6.595210206878644e-08,
+      "loss": 0.3361,
+      "step": 13679
+    },
+    {
+      "epoch": 0.9498680738786279,
+      "grad_norm": 3.554279115930025,
+      "learning_rate": 6.57701888722806e-08,
+      "loss": 0.2929,
+      "step": 13680
+    },
+    {
+      "epoch": 0.9499375086793501,
+      "grad_norm": 3.7673448291774734,
+      "learning_rate": 6.558852524353509e-08,
+      "loss": 0.4341,
+      "step": 13681
+    },
+    {
+      "epoch": 0.9500069434800722,
+      "grad_norm": 3.351238846744773,
+      "learning_rate": 6.54071111917376e-08,
+      "loss": 0.3523,
+      "step": 13682
+    },
+    {
+      "epoch": 0.9500763782807944,
+      "grad_norm": 3.1161316831082817,
+      "learning_rate": 6.522594672606519e-08,
+      "loss": 0.3187,
+      "step": 13683
+    },
+    {
+      "epoch": 0.9501458130815165,
+      "grad_norm": 5.358799128246562,
+      "learning_rate": 6.504503185568056e-08,
+      "loss": 0.6084,
+      "step": 13684
+    },
+    {
+      "epoch": 0.9502152478822385,
+      "grad_norm": 5.625555475746649,
+      "learning_rate": 6.486436658973305e-08,
+      "loss": 0.4486,
+      "step": 13685
+    },
+    {
+      "epoch": 0.9502846826829607,
+      "grad_norm": 3.8169861363144126,
+      "learning_rate": 6.468395093736201e-08,
+      "loss": 0.4327,
+      "step": 13686
+    },
+    {
+      "epoch": 0.9503541174836828,
+      "grad_norm": 3.6956254941635147,
+      "learning_rate": 6.450378490769238e-08,
+      "loss": 0.4737,
+      "step": 13687
+    },
+    {
+      "epoch": 0.9504235522844049,
+      "grad_norm": 3.5095411747770098,
+      "learning_rate": 6.432386850983575e-08,
+      "loss": 0.2572,
+      "step": 13688
+    },
+    {
+      "epoch": 0.9504929870851271,
+      "grad_norm": 4.259990413136725,
+      "learning_rate": 6.414420175289316e-08,
+      "loss": 0.4709,
+      "step": 13689
+    },
+    {
+      "epoch": 0.9505624218858492,
+      "grad_norm": 3.8901611509865326,
+      "learning_rate": 6.396478464595124e-08,
+      "loss": 0.4531,
+      "step": 13690
+    },
+    {
+      "epoch": 0.9506318566865714,
+      "grad_norm": 4.616724500012755,
+      "learning_rate": 6.378561719808495e-08,
+      "loss": 0.655,
+      "step": 13691
+    },
+    {
+      "epoch": 0.9507012914872934,
+      "grad_norm": 3.4537477401329206,
+      "learning_rate": 6.360669941835707e-08,
+      "loss": 0.275,
+      "step": 13692
+    },
+    {
+      "epoch": 0.9507707262880155,
+      "grad_norm": 3.512772786078367,
+      "learning_rate": 6.342803131581587e-08,
+      "loss": 0.3024,
+      "step": 13693
+    },
+    {
+      "epoch": 0.9508401610887377,
+      "grad_norm": 4.8958741741450815,
+      "learning_rate": 6.32496128994986e-08,
+      "loss": 0.5526,
+      "step": 13694
+    },
+    {
+      "epoch": 0.9509095958894598,
+      "grad_norm": 3.865406298166846,
+      "learning_rate": 6.307144417843025e-08,
+      "loss": 0.41,
+      "step": 13695
+    },
+    {
+      "epoch": 0.950979030690182,
+      "grad_norm": 4.008626378788147,
+      "learning_rate": 6.28935251616214e-08,
+      "loss": 0.3548,
+      "step": 13696
+    },
+    {
+      "epoch": 0.951048465490904,
+      "grad_norm": 3.234938443853972,
+      "learning_rate": 6.271585585807095e-08,
+      "loss": 0.2858,
+      "step": 13697
+    },
+    {
+      "epoch": 0.9511179002916261,
+      "grad_norm": 3.2906952625792756,
+      "learning_rate": 6.253843627676614e-08,
+      "loss": 0.2802,
+      "step": 13698
+    },
+    {
+      "epoch": 0.9511873350923483,
+      "grad_norm": 2.9383512680773967,
+      "learning_rate": 6.236126642667983e-08,
+      "loss": 0.2325,
+      "step": 13699
+    },
+    {
+      "epoch": 0.9512567698930704,
+      "grad_norm": 3.6388071059489047,
+      "learning_rate": 6.218434631677316e-08,
+      "loss": 0.4709,
+      "step": 13700
+    },
+    {
+      "epoch": 0.9513262046937925,
+      "grad_norm": 3.513599857057769,
+      "learning_rate": 6.200767595599566e-08,
+      "loss": 0.3443,
+      "step": 13701
+    },
+    {
+      "epoch": 0.9513956394945147,
+      "grad_norm": 2.98873291672003,
+      "learning_rate": 6.183125535328072e-08,
+      "loss": 0.2342,
+      "step": 13702
+    },
+    {
+      "epoch": 0.9514650742952367,
+      "grad_norm": 3.0775042067561578,
+      "learning_rate": 6.165508451755397e-08,
+      "loss": 0.2925,
+      "step": 13703
+    },
+    {
+      "epoch": 0.9515345090959589,
+      "grad_norm": 3.4704343870757812,
+      "learning_rate": 6.147916345772497e-08,
+      "loss": 0.206,
+      "step": 13704
+    },
+    {
+      "epoch": 0.951603943896681,
+      "grad_norm": 3.0964683795669963,
+      "learning_rate": 6.130349218269104e-08,
+      "loss": 0.3312,
+      "step": 13705
+    },
+    {
+      "epoch": 0.9516733786974031,
+      "grad_norm": 3.5393836860593177,
+      "learning_rate": 6.112807070133842e-08,
+      "loss": 0.2131,
+      "step": 13706
+    },
+    {
+      "epoch": 0.9517428134981253,
+      "grad_norm": 4.569400179364535,
+      "learning_rate": 6.095289902253887e-08,
+      "loss": 0.4264,
+      "step": 13707
+    },
+    {
+      "epoch": 0.9518122482988474,
+      "grad_norm": 5.161394653168402,
+      "learning_rate": 6.07779771551531e-08,
+      "loss": 0.5866,
+      "step": 13708
+    },
+    {
+      "epoch": 0.9518816830995696,
+      "grad_norm": 5.0977324811951945,
+      "learning_rate": 6.060330510802848e-08,
+      "loss": 0.3997,
+      "step": 13709
+    },
+    {
+      "epoch": 0.9519511179002916,
+      "grad_norm": 4.759533807947581,
+      "learning_rate": 6.042888288999904e-08,
+      "loss": 0.4045,
+      "step": 13710
+    },
+    {
+      "epoch": 0.9520205527010137,
+      "grad_norm": 5.505368708477018,
+      "learning_rate": 6.025471050988662e-08,
+      "loss": 0.5899,
+      "step": 13711
+    },
+    {
+      "epoch": 0.9520899875017359,
+      "grad_norm": 5.442442112067745,
+      "learning_rate": 6.008078797650196e-08,
+      "loss": 0.5671,
+      "step": 13712
+    },
+    {
+      "epoch": 0.952159422302458,
+      "grad_norm": 4.057394585951996,
+      "learning_rate": 5.990711529864079e-08,
+      "loss": 0.5129,
+      "step": 13713
+    },
+    {
+      "epoch": 0.9522288571031801,
+      "grad_norm": 2.659618891571195,
+      "learning_rate": 5.973369248508775e-08,
+      "loss": 0.2432,
+      "step": 13714
+    },
+    {
+      "epoch": 0.9522982919039022,
+      "grad_norm": 4.104167267102356,
+      "learning_rate": 5.9560519544614725e-08,
+      "loss": 0.6073,
+      "step": 13715
+    },
+    {
+      "epoch": 0.9523677267046243,
+      "grad_norm": 4.607507065467092,
+      "learning_rate": 5.938759648597969e-08,
+      "loss": 0.3594,
+      "step": 13716
+    },
+    {
+      "epoch": 0.9524371615053465,
+      "grad_norm": 4.846589279139945,
+      "learning_rate": 5.9214923317928996e-08,
+      "loss": 0.4522,
+      "step": 13717
+    },
+    {
+      "epoch": 0.9525065963060686,
+      "grad_norm": 4.119954192191566,
+      "learning_rate": 5.904250004919676e-08,
+      "loss": 0.4353,
+      "step": 13718
+    },
+    {
+      "epoch": 0.9525760311067907,
+      "grad_norm": 4.648236196200457,
+      "learning_rate": 5.8870326688503785e-08,
+      "loss": 0.7015,
+      "step": 13719
+    },
+    {
+      "epoch": 0.9526454659075129,
+      "grad_norm": 4.080707223352292,
+      "learning_rate": 5.869840324455867e-08,
+      "loss": 0.3911,
+      "step": 13720
+    },
+    {
+      "epoch": 0.952714900708235,
+      "grad_norm": 2.8789349070511805,
+      "learning_rate": 5.852672972605722e-08,
+      "loss": 0.2019,
+      "step": 13721
+    },
+    {
+      "epoch": 0.9527843355089571,
+      "grad_norm": 3.410282571025948,
+      "learning_rate": 5.835530614168139e-08,
+      "loss": 0.2992,
+      "step": 13722
+    },
+    {
+      "epoch": 0.9528537703096792,
+      "grad_norm": 4.894677996671319,
+      "learning_rate": 5.818413250010313e-08,
+      "loss": 0.239,
+      "step": 13723
+    },
+    {
+      "epoch": 0.9529232051104013,
+      "grad_norm": 3.3017609860449375,
+      "learning_rate": 5.801320880997885e-08,
+      "loss": 0.4211,
+      "step": 13724
+    },
+    {
+      "epoch": 0.9529926399111235,
+      "grad_norm": 3.6449909829302447,
+      "learning_rate": 5.784253507995441e-08,
+      "loss": 0.4747,
+      "step": 13725
+    },
+    {
+      "epoch": 0.9530620747118456,
+      "grad_norm": 2.844534890946508,
+      "learning_rate": 5.7672111318661796e-08,
+      "loss": 0.2872,
+      "step": 13726
+    },
+    {
+      "epoch": 0.9531315095125678,
+      "grad_norm": 4.253589585625809,
+      "learning_rate": 5.750193753472133e-08,
+      "loss": 0.433,
+      "step": 13727
+    },
+    {
+      "epoch": 0.9532009443132898,
+      "grad_norm": 4.573107826333037,
+      "learning_rate": 5.733201373674002e-08,
+      "loss": 0.4319,
+      "step": 13728
+    },
+    {
+      "epoch": 0.9532703791140119,
+      "grad_norm": 3.8093677377233943,
+      "learning_rate": 5.716233993331266e-08,
+      "loss": 0.3048,
+      "step": 13729
+    },
+    {
+      "epoch": 0.9533398139147341,
+      "grad_norm": 3.9826577582816105,
+      "learning_rate": 5.6992916133020713e-08,
+      "loss": 0.5305,
+      "step": 13730
+    },
+    {
+      "epoch": 0.9534092487154562,
+      "grad_norm": 5.936500169456354,
+      "learning_rate": 5.682374234443344e-08,
+      "loss": 0.4476,
+      "step": 13731
+    },
+    {
+      "epoch": 0.9534786835161783,
+      "grad_norm": 3.5337222091478004,
+      "learning_rate": 5.665481857610844e-08,
+      "loss": 0.3033,
+      "step": 13732
+    },
+    {
+      "epoch": 0.9535481183169005,
+      "grad_norm": 4.474230508273001,
+      "learning_rate": 5.648614483658887e-08,
+      "loss": 0.3867,
+      "step": 13733
+    },
+    {
+      "epoch": 0.9536175531176225,
+      "grad_norm": 3.4395490258977115,
+      "learning_rate": 5.6317721134405145e-08,
+      "loss": 0.403,
+      "step": 13734
+    },
+    {
+      "epoch": 0.9536869879183447,
+      "grad_norm": 4.449479349855479,
+      "learning_rate": 5.614954747807766e-08,
+      "loss": 0.5633,
+      "step": 13735
+    },
+    {
+      "epoch": 0.9537564227190668,
+      "grad_norm": 3.602797052026786,
+      "learning_rate": 5.5981623876111833e-08,
+      "loss": 0.4116,
+      "step": 13736
+    },
+    {
+      "epoch": 0.9538258575197889,
+      "grad_norm": 3.8764715732710227,
+      "learning_rate": 5.5813950337000876e-08,
+      "loss": 0.2844,
+      "step": 13737
+    },
+    {
+      "epoch": 0.9538952923205111,
+      "grad_norm": 3.8746349156932447,
+      "learning_rate": 5.564652686922523e-08,
+      "loss": 0.4043,
+      "step": 13738
+    },
+    {
+      "epoch": 0.9539647271212331,
+      "grad_norm": 2.831249944933847,
+      "learning_rate": 5.547935348125366e-08,
+      "loss": 0.1951,
+      "step": 13739
+    },
+    {
+      "epoch": 0.9540341619219553,
+      "grad_norm": 5.002261056652214,
+      "learning_rate": 5.531243018154109e-08,
+      "loss": 0.6011,
+      "step": 13740
+    },
+    {
+      "epoch": 0.9541035967226774,
+      "grad_norm": 3.5291833462110653,
+      "learning_rate": 5.5145756978530754e-08,
+      "loss": 0.3471,
+      "step": 13741
+    },
+    {
+      "epoch": 0.9541730315233995,
+      "grad_norm": 4.264188643167261,
+      "learning_rate": 5.4979333880652574e-08,
+      "loss": 0.5141,
+      "step": 13742
+    },
+    {
+      "epoch": 0.9542424663241217,
+      "grad_norm": 3.9364891155309274,
+      "learning_rate": 5.4813160896323714e-08,
+      "loss": 0.3907,
+      "step": 13743
+    },
+    {
+      "epoch": 0.9543119011248438,
+      "grad_norm": 4.387793071077238,
+      "learning_rate": 5.4647238033949666e-08,
+      "loss": 0.3864,
+      "step": 13744
+    },
+    {
+      "epoch": 0.9543813359255658,
+      "grad_norm": 4.822898455030683,
+      "learning_rate": 5.4481565301922057e-08,
+      "loss": 0.5582,
+      "step": 13745
+    },
+    {
+      "epoch": 0.954450770726288,
+      "grad_norm": 4.938522930095148,
+      "learning_rate": 5.431614270862029e-08,
+      "loss": 0.7861,
+      "step": 13746
+    },
+    {
+      "epoch": 0.9545202055270101,
+      "grad_norm": 4.066376041766102,
+      "learning_rate": 5.4150970262412114e-08,
+      "loss": 0.2856,
+      "step": 13747
+    },
+    {
+      "epoch": 0.9545896403277323,
+      "grad_norm": 3.6130713492136137,
+      "learning_rate": 5.398604797165141e-08,
+      "loss": 0.3425,
+      "step": 13748
+    },
+    {
+      "epoch": 0.9546590751284544,
+      "grad_norm": 4.881464304116736,
+      "learning_rate": 5.3821375844678725e-08,
+      "loss": 0.4625,
+      "step": 13749
+    },
+    {
+      "epoch": 0.9547285099291765,
+      "grad_norm": 4.86315785564698,
+      "learning_rate": 5.365695388982461e-08,
+      "loss": 0.668,
+      "step": 13750
+    },
+    {
+      "epoch": 0.9547979447298987,
+      "grad_norm": 4.414774526005412,
+      "learning_rate": 5.349278211540465e-08,
+      "loss": 0.4378,
+      "step": 13751
+    },
+    {
+      "epoch": 0.9548673795306207,
+      "grad_norm": 5.212315794946315,
+      "learning_rate": 5.332886052972219e-08,
+      "loss": 0.4562,
+      "step": 13752
+    },
+    {
+      "epoch": 0.9549368143313429,
+      "grad_norm": 4.244604861683294,
+      "learning_rate": 5.3165189141068385e-08,
+      "loss": 0.481,
+      "step": 13753
+    },
+    {
+      "epoch": 0.955006249132065,
+      "grad_norm": 2.623848249585915,
+      "learning_rate": 5.300176795772161e-08,
+      "loss": 0.0998,
+      "step": 13754
+    },
+    {
+      "epoch": 0.9550756839327871,
+      "grad_norm": 3.1011437769689216,
+      "learning_rate": 5.2838596987947464e-08,
+      "loss": 0.2928,
+      "step": 13755
+    },
+    {
+      "epoch": 0.9551451187335093,
+      "grad_norm": 4.816876318480773,
+      "learning_rate": 5.2675676239999365e-08,
+      "loss": 0.4824,
+      "step": 13756
+    },
+    {
+      "epoch": 0.9552145535342313,
+      "grad_norm": 3.3565758188776256,
+      "learning_rate": 5.2513005722116816e-08,
+      "loss": 0.3917,
+      "step": 13757
+    },
+    {
+      "epoch": 0.9552839883349534,
+      "grad_norm": 3.345690512458578,
+      "learning_rate": 5.235058544252825e-08,
+      "loss": 0.2665,
+      "step": 13758
+    },
+    {
+      "epoch": 0.9553534231356756,
+      "grad_norm": 3.701444267394486,
+      "learning_rate": 5.218841540944819e-08,
+      "loss": 0.4149,
+      "step": 13759
+    },
+    {
+      "epoch": 0.9554228579363977,
+      "grad_norm": 3.583136724246739,
+      "learning_rate": 5.202649563107898e-08,
+      "loss": 0.3153,
+      "step": 13760
+    },
+    {
+      "epoch": 0.9554922927371199,
+      "grad_norm": 6.221288795909009,
+      "learning_rate": 5.1864826115611275e-08,
+      "loss": 0.3226,
+      "step": 13761
+    },
+    {
+      "epoch": 0.955561727537842,
+      "grad_norm": 4.410783839641024,
+      "learning_rate": 5.170340687122133e-08,
+      "loss": 0.3513,
+      "step": 13762
+    },
+    {
+      "epoch": 0.955631162338564,
+      "grad_norm": 4.9048981793991215,
+      "learning_rate": 5.1542237906073154e-08,
+      "loss": 0.4872,
+      "step": 13763
+    },
+    {
+      "epoch": 0.9557005971392862,
+      "grad_norm": 4.0147064224730995,
+      "learning_rate": 5.1381319228319684e-08,
+      "loss": 0.4485,
+      "step": 13764
+    },
+    {
+      "epoch": 0.9557700319400083,
+      "grad_norm": 4.065462247455766,
+      "learning_rate": 5.12206508460994e-08,
+      "loss": 0.3438,
+      "step": 13765
+    },
+    {
+      "epoch": 0.9558394667407305,
+      "grad_norm": 5.005870930483714,
+      "learning_rate": 5.106023276753802e-08,
+      "loss": 0.5467,
+      "step": 13766
+    },
+    {
+      "epoch": 0.9559089015414526,
+      "grad_norm": 4.1956769404460195,
+      "learning_rate": 5.090006500075018e-08,
+      "loss": 0.4811,
+      "step": 13767
+    },
+    {
+      "epoch": 0.9559783363421747,
+      "grad_norm": 3.164243683731383,
+      "learning_rate": 5.074014755383716e-08,
+      "loss": 0.206,
+      "step": 13768
+    },
+    {
+      "epoch": 0.9560477711428969,
+      "grad_norm": 4.034796986554963,
+      "learning_rate": 5.0580480434886946e-08,
+      "loss": 0.3752,
+      "step": 13769
+    },
+    {
+      "epoch": 0.9561172059436189,
+      "grad_norm": 3.1577514924323142,
+      "learning_rate": 5.04210636519753e-08,
+      "loss": 0.3422,
+      "step": 13770
+    },
+    {
+      "epoch": 0.956186640744341,
+      "grad_norm": 4.621330442701489,
+      "learning_rate": 5.026189721316521e-08,
+      "loss": 0.4951,
+      "step": 13771
+    },
+    {
+      "epoch": 0.9562560755450632,
+      "grad_norm": 3.645406553491685,
+      "learning_rate": 5.010298112650691e-08,
+      "loss": 0.3262,
+      "step": 13772
+    },
+    {
+      "epoch": 0.9563255103457853,
+      "grad_norm": 2.738805959666506,
+      "learning_rate": 4.9944315400039524e-08,
+      "loss": 0.165,
+      "step": 13773
+    },
+    {
+      "epoch": 0.9563949451465075,
+      "grad_norm": 5.5379366339991165,
+      "learning_rate": 4.978590004178718e-08,
+      "loss": 0.6481,
+      "step": 13774
+    },
+    {
+      "epoch": 0.9564643799472295,
+      "grad_norm": 2.812040631679995,
+      "learning_rate": 4.962773505976182e-08,
+      "loss": 0.2133,
+      "step": 13775
+    },
+    {
+      "epoch": 0.9565338147479516,
+      "grad_norm": 4.185047982000624,
+      "learning_rate": 4.94698204619648e-08,
+      "loss": 0.4786,
+      "step": 13776
+    },
+    {
+      "epoch": 0.9566032495486738,
+      "grad_norm": 4.22031613308588,
+      "learning_rate": 4.931215625638197e-08,
+      "loss": 0.4203,
+      "step": 13777
+    },
+    {
+      "epoch": 0.9566726843493959,
+      "grad_norm": 3.2502871450722446,
+      "learning_rate": 4.91547424509875e-08,
+      "loss": 0.3303,
+      "step": 13778
+    },
+    {
+      "epoch": 0.9567421191501181,
+      "grad_norm": 2.840915325002215,
+      "learning_rate": 4.8997579053744473e-08,
+      "loss": 0.1366,
+      "step": 13779
+    },
+    {
+      "epoch": 0.9568115539508402,
+      "grad_norm": 3.7913872967781597,
+      "learning_rate": 4.8840666072601516e-08,
+      "loss": 0.3388,
+      "step": 13780
+    },
+    {
+      "epoch": 0.9568809887515622,
+      "grad_norm": 3.46239936309188,
+      "learning_rate": 4.868400351549507e-08,
+      "loss": 0.2844,
+      "step": 13781
+    },
+    {
+      "epoch": 0.9569504235522844,
+      "grad_norm": 6.718760040575691,
+      "learning_rate": 4.852759139034935e-08,
+      "loss": 0.4516,
+      "step": 13782
+    },
+    {
+      "epoch": 0.9570198583530065,
+      "grad_norm": 3.662048534926939,
+      "learning_rate": 4.837142970507469e-08,
+      "loss": 0.3151,
+      "step": 13783
+    },
+    {
+      "epoch": 0.9570892931537287,
+      "grad_norm": 4.143910616518008,
+      "learning_rate": 4.8215518467569775e-08,
+      "loss": 0.419,
+      "step": 13784
+    },
+    {
+      "epoch": 0.9571587279544508,
+      "grad_norm": 2.8260291756886793,
+      "learning_rate": 4.8059857685720524e-08,
+      "loss": 0.3153,
+      "step": 13785
+    },
+    {
+      "epoch": 0.9572281627551729,
+      "grad_norm": 5.632561323738521,
+      "learning_rate": 4.790444736740063e-08,
+      "loss": 0.4623,
+      "step": 13786
+    },
+    {
+      "epoch": 0.957297597555895,
+      "grad_norm": 4.533362266438861,
+      "learning_rate": 4.7749287520469925e-08,
+      "loss": 0.3911,
+      "step": 13787
+    },
+    {
+      "epoch": 0.9573670323566171,
+      "grad_norm": 4.030234423717611,
+      "learning_rate": 4.759437815277601e-08,
+      "loss": 0.4466,
+      "step": 13788
+    },
+    {
+      "epoch": 0.9574364671573392,
+      "grad_norm": 3.760078342574025,
+      "learning_rate": 4.743971927215485e-08,
+      "loss": 0.4153,
+      "step": 13789
+    },
+    {
+      "epoch": 0.9575059019580614,
+      "grad_norm": 5.438482204493829,
+      "learning_rate": 4.728531088642851e-08,
+      "loss": 0.5471,
+      "step": 13790
+    },
+    {
+      "epoch": 0.9575753367587835,
+      "grad_norm": 3.5605754782880865,
+      "learning_rate": 4.713115300340687e-08,
+      "loss": 0.4443,
+      "step": 13791
+    },
+    {
+      "epoch": 0.9576447715595057,
+      "grad_norm": 3.2605874775568604,
+      "learning_rate": 4.697724563088646e-08,
+      "loss": 0.23,
+      "step": 13792
+    },
+    {
+      "epoch": 0.9577142063602277,
+      "grad_norm": 3.8953526992560596,
+      "learning_rate": 4.682358877665272e-08,
+      "loss": 0.4293,
+      "step": 13793
+    },
+    {
+      "epoch": 0.9577836411609498,
+      "grad_norm": 3.688126146802298,
+      "learning_rate": 4.6670182448476654e-08,
+      "loss": 0.5331,
+      "step": 13794
+    },
+    {
+      "epoch": 0.957853075961672,
+      "grad_norm": 3.6621761384077947,
+      "learning_rate": 4.6517026654117614e-08,
+      "loss": 0.3177,
+      "step": 13795
+    },
+    {
+      "epoch": 0.9579225107623941,
+      "grad_norm": 2.9829489682129067,
+      "learning_rate": 4.6364121401322185e-08,
+      "loss": 0.2041,
+      "step": 13796
+    },
+    {
+      "epoch": 0.9579919455631163,
+      "grad_norm": 3.8755040473465487,
+      "learning_rate": 4.6211466697824724e-08,
+      "loss": 0.3712,
+      "step": 13797
+    },
+    {
+      "epoch": 0.9580613803638384,
+      "grad_norm": 3.7293845240685735,
+      "learning_rate": 4.605906255134518e-08,
+      "loss": 0.4221,
+      "step": 13798
+    },
+    {
+      "epoch": 0.9581308151645604,
+      "grad_norm": 5.200251316713918,
+      "learning_rate": 4.590690896959238e-08,
+      "loss": 0.5624,
+      "step": 13799
+    },
+    {
+      "epoch": 0.9582002499652826,
+      "grad_norm": 3.5477653655164607,
+      "learning_rate": 4.5755005960262386e-08,
+      "loss": 0.2367,
+      "step": 13800
+    },
+    {
+      "epoch": 0.9582696847660047,
+      "grad_norm": 3.55678544387724,
+      "learning_rate": 4.5603353531037954e-08,
+      "loss": 0.3706,
+      "step": 13801
+    },
+    {
+      "epoch": 0.9583391195667268,
+      "grad_norm": 4.019138298153231,
+      "learning_rate": 4.54519516895896e-08,
+      "loss": 0.3819,
+      "step": 13802
+    },
+    {
+      "epoch": 0.958408554367449,
+      "grad_norm": 4.1423853727256965,
+      "learning_rate": 4.53008004435751e-08,
+      "loss": 0.4493,
+      "step": 13803
+    },
+    {
+      "epoch": 0.9584779891681711,
+      "grad_norm": 3.439216853021881,
+      "learning_rate": 4.514989980063944e-08,
+      "loss": 0.3431,
+      "step": 13804
+    },
+    {
+      "epoch": 0.9585474239688933,
+      "grad_norm": 4.119036257431424,
+      "learning_rate": 4.499924976841541e-08,
+      "loss": 0.3071,
+      "step": 13805
+    },
+    {
+      "epoch": 0.9586168587696153,
+      "grad_norm": 4.3364542957782115,
+      "learning_rate": 4.484885035452191e-08,
+      "loss": 0.4829,
+      "step": 13806
+    },
+    {
+      "epoch": 0.9586862935703374,
+      "grad_norm": 4.154648087154131,
+      "learning_rate": 4.469870156656619e-08,
+      "loss": 0.4828,
+      "step": 13807
+    },
+    {
+      "epoch": 0.9587557283710596,
+      "grad_norm": 4.386339857521736,
+      "learning_rate": 4.454880341214385e-08,
+      "loss": 0.414,
+      "step": 13808
+    },
+    {
+      "epoch": 0.9588251631717817,
+      "grad_norm": 4.2006746850425865,
+      "learning_rate": 4.4399155898834365e-08,
+      "loss": 0.2814,
+      "step": 13809
+    },
+    {
+      "epoch": 0.9588945979725039,
+      "grad_norm": 4.4862546746366885,
+      "learning_rate": 4.4249759034208916e-08,
+      "loss": 0.5541,
+      "step": 13810
+    },
+    {
+      "epoch": 0.958964032773226,
+      "grad_norm": 5.247518739724254,
+      "learning_rate": 4.410061282582256e-08,
+      "loss": 0.4731,
+      "step": 13811
+    },
+    {
+      "epoch": 0.959033467573948,
+      "grad_norm": 3.427119471248524,
+      "learning_rate": 4.395171728121872e-08,
+      "loss": 0.2677,
+      "step": 13812
+    },
+    {
+      "epoch": 0.9591029023746702,
+      "grad_norm": 5.207259037309656,
+      "learning_rate": 4.380307240792969e-08,
+      "loss": 0.5715,
+      "step": 13813
+    },
+    {
+      "epoch": 0.9591723371753923,
+      "grad_norm": 3.723407690108263,
+      "learning_rate": 4.365467821347225e-08,
+      "loss": 0.3877,
+      "step": 13814
+    },
+    {
+      "epoch": 0.9592417719761144,
+      "grad_norm": 4.655179146418831,
+      "learning_rate": 4.350653470535315e-08,
+      "loss": 0.4317,
+      "step": 13815
+    },
+    {
+      "epoch": 0.9593112067768366,
+      "grad_norm": 4.75184632912496,
+      "learning_rate": 4.3358641891064757e-08,
+      "loss": 0.4672,
+      "step": 13816
+    },
+    {
+      "epoch": 0.9593806415775586,
+      "grad_norm": 4.343633391023645,
+      "learning_rate": 4.3210999778087184e-08,
+      "loss": 0.4242,
+      "step": 13817
+    },
+    {
+      "epoch": 0.9594500763782808,
+      "grad_norm": 4.742145894209304,
+      "learning_rate": 4.30636083738889e-08,
+      "loss": 0.5523,
+      "step": 13818
+    },
+    {
+      "epoch": 0.9595195111790029,
+      "grad_norm": 3.63360962971277,
+      "learning_rate": 4.291646768592339e-08,
+      "loss": 0.1969,
+      "step": 13819
+    },
+    {
+      "epoch": 0.959588945979725,
+      "grad_norm": 4.658256364052358,
+      "learning_rate": 4.2769577721634146e-08,
+      "loss": 0.5623,
+      "step": 13820
+    },
+    {
+      "epoch": 0.9596583807804472,
+      "grad_norm": 3.1681681870570597,
+      "learning_rate": 4.262293848844967e-08,
+      "loss": 0.3458,
+      "step": 13821
+    },
+    {
+      "epoch": 0.9597278155811693,
+      "grad_norm": 3.1429083666928124,
+      "learning_rate": 4.2476549993787363e-08,
+      "loss": 0.3386,
+      "step": 13822
+    },
+    {
+      "epoch": 0.9597972503818915,
+      "grad_norm": 3.758413772178014,
+      "learning_rate": 4.233041224505241e-08,
+      "loss": 0.32,
+      "step": 13823
+    },
+    {
+      "epoch": 0.9598666851826135,
+      "grad_norm": 4.712591199974532,
+      "learning_rate": 4.21845252496339e-08,
+      "loss": 0.3828,
+      "step": 13824
+    },
+    {
+      "epoch": 0.9599361199833356,
+      "grad_norm": 3.370755783479801,
+      "learning_rate": 4.2038889014912595e-08,
+      "loss": 0.4166,
+      "step": 13825
+    },
+    {
+      "epoch": 0.9600055547840578,
+      "grad_norm": 4.8701764259809215,
+      "learning_rate": 4.189350354825428e-08,
+      "loss": 0.4534,
+      "step": 13826
+    },
+    {
+      "epoch": 0.9600749895847799,
+      "grad_norm": 3.9778007454806343,
+      "learning_rate": 4.174836885701139e-08,
+      "loss": 0.2872,
+      "step": 13827
+    },
+    {
+      "epoch": 0.960144424385502,
+      "grad_norm": 4.249046231927962,
+      "learning_rate": 4.160348494852528e-08,
+      "loss": 0.5902,
+      "step": 13828
+    },
+    {
+      "epoch": 0.9602138591862242,
+      "grad_norm": 4.7314813108457,
+      "learning_rate": 4.1458851830124545e-08,
+      "loss": 0.4503,
+      "step": 13829
+    },
+    {
+      "epoch": 0.9602832939869462,
+      "grad_norm": 4.299908866757219,
+      "learning_rate": 4.131446950912388e-08,
+      "loss": 0.4853,
+      "step": 13830
+    },
+    {
+      "epoch": 0.9603527287876684,
+      "grad_norm": 4.087174207564177,
+      "learning_rate": 4.1170337992826325e-08,
+      "loss": 0.3416,
+      "step": 13831
+    },
+    {
+      "epoch": 0.9604221635883905,
+      "grad_norm": 3.2249131898160246,
+      "learning_rate": 4.102645728852217e-08,
+      "loss": 0.2984,
+      "step": 13832
+    },
+    {
+      "epoch": 0.9604915983891126,
+      "grad_norm": 4.891650256840407,
+      "learning_rate": 4.0882827403487814e-08,
+      "loss": 0.4193,
+      "step": 13833
+    },
+    {
+      "epoch": 0.9605610331898348,
+      "grad_norm": 4.020020612618311,
+      "learning_rate": 4.0739448344988554e-08,
+      "loss": 0.3364,
+      "step": 13834
+    },
+    {
+      "epoch": 0.9606304679905568,
+      "grad_norm": 3.6139678591424387,
+      "learning_rate": 4.059632012027637e-08,
+      "loss": 0.4618,
+      "step": 13835
+    },
+    {
+      "epoch": 0.960699902791279,
+      "grad_norm": 3.303165549296177,
+      "learning_rate": 4.045344273658991e-08,
+      "loss": 0.3134,
+      "step": 13836
+    },
+    {
+      "epoch": 0.9607693375920011,
+      "grad_norm": 3.758252980498182,
+      "learning_rate": 4.031081620115673e-08,
+      "loss": 0.3988,
+      "step": 13837
+    },
+    {
+      "epoch": 0.9608387723927232,
+      "grad_norm": 4.530037871823786,
+      "learning_rate": 4.0168440521189954e-08,
+      "loss": 0.3457,
+      "step": 13838
+    },
+    {
+      "epoch": 0.9609082071934454,
+      "grad_norm": 3.8646392516747667,
+      "learning_rate": 4.002631570389104e-08,
+      "loss": 0.188,
+      "step": 13839
+    },
+    {
+      "epoch": 0.9609776419941675,
+      "grad_norm": 4.329407701100087,
+      "learning_rate": 3.9884441756448676e-08,
+      "loss": 0.4367,
+      "step": 13840
+    },
+    {
+      "epoch": 0.9610470767948895,
+      "grad_norm": 4.600340799932416,
+      "learning_rate": 3.9742818686038244e-08,
+      "loss": 0.5344,
+      "step": 13841
+    },
+    {
+      "epoch": 0.9611165115956117,
+      "grad_norm": 3.702597773537803,
+      "learning_rate": 3.9601446499822894e-08,
+      "loss": 0.5135,
+      "step": 13842
+    },
+    {
+      "epoch": 0.9611859463963338,
+      "grad_norm": 4.5623747942811255,
+      "learning_rate": 3.946032520495413e-08,
+      "loss": 0.5533,
+      "step": 13843
+    },
+    {
+      "epoch": 0.961255381197056,
+      "grad_norm": 3.7726541753239977,
+      "learning_rate": 3.931945480856847e-08,
+      "loss": 0.4268,
+      "step": 13844
+    },
+    {
+      "epoch": 0.9613248159977781,
+      "grad_norm": 3.6886432017973907,
+      "learning_rate": 3.9178835317791317e-08,
+      "loss": 0.2752,
+      "step": 13845
+    },
+    {
+      "epoch": 0.9613942507985002,
+      "grad_norm": 4.245120722768961,
+      "learning_rate": 3.903846673973477e-08,
+      "loss": 0.4737,
+      "step": 13846
+    },
+    {
+      "epoch": 0.9614636855992224,
+      "grad_norm": 4.339663662356036,
+      "learning_rate": 3.889834908149925e-08,
+      "loss": 0.4936,
+      "step": 13847
+    },
+    {
+      "epoch": 0.9615331203999444,
+      "grad_norm": 3.128509208211385,
+      "learning_rate": 3.8758482350171325e-08,
+      "loss": 0.2537,
+      "step": 13848
+    },
+    {
+      "epoch": 0.9616025552006666,
+      "grad_norm": 4.929401318062302,
+      "learning_rate": 3.8618866552825874e-08,
+      "loss": 0.475,
+      "step": 13849
+    },
+    {
+      "epoch": 0.9616719900013887,
+      "grad_norm": 3.4409667197554143,
+      "learning_rate": 3.847950169652282e-08,
+      "loss": 0.3655,
+      "step": 13850
+    },
+    {
+      "epoch": 0.9617414248021108,
+      "grad_norm": 4.465699661608561,
+      "learning_rate": 3.8340387788313174e-08,
+      "loss": 0.5647,
+      "step": 13851
+    },
+    {
+      "epoch": 0.961810859602833,
+      "grad_norm": 3.7656554880294215,
+      "learning_rate": 3.820152483523187e-08,
+      "loss": 0.3255,
+      "step": 13852
+    },
+    {
+      "epoch": 0.961880294403555,
+      "grad_norm": 6.9087627607817215,
+      "learning_rate": 3.806291284430275e-08,
+      "loss": 0.603,
+      "step": 13853
+    },
+    {
+      "epoch": 0.9619497292042772,
+      "grad_norm": 3.7872487150111525,
+      "learning_rate": 3.792455182253685e-08,
+      "loss": 0.4127,
+      "step": 13854
+    },
+    {
+      "epoch": 0.9620191640049993,
+      "grad_norm": 7.69238565634778,
+      "learning_rate": 3.778644177693247e-08,
+      "loss": 0.6673,
+      "step": 13855
+    },
+    {
+      "epoch": 0.9620885988057214,
+      "grad_norm": 3.8270105228991014,
+      "learning_rate": 3.7648582714474026e-08,
+      "loss": 0.3714,
+      "step": 13856
+    },
+    {
+      "epoch": 0.9621580336064436,
+      "grad_norm": 3.283479712176172,
+      "learning_rate": 3.7510974642135375e-08,
+      "loss": 0.227,
+      "step": 13857
+    },
+    {
+      "epoch": 0.9622274684071657,
+      "grad_norm": 2.9699510666847657,
+      "learning_rate": 3.737361756687596e-08,
+      "loss": 0.3364,
+      "step": 13858
+    },
+    {
+      "epoch": 0.9622969032078877,
+      "grad_norm": 4.708008899051446,
+      "learning_rate": 3.7236511495643557e-08,
+      "loss": 0.513,
+      "step": 13859
+    },
+    {
+      "epoch": 0.9623663380086099,
+      "grad_norm": 3.935371065598654,
+      "learning_rate": 3.709965643537261e-08,
+      "loss": 0.3193,
+      "step": 13860
+    },
+    {
+      "epoch": 0.962435772809332,
+      "grad_norm": 4.575474964215627,
+      "learning_rate": 3.6963052392985364e-08,
+      "loss": 0.5336,
+      "step": 13861
+    },
+    {
+      "epoch": 0.9625052076100542,
+      "grad_norm": 3.7754618598643965,
+      "learning_rate": 3.682669937539074e-08,
+      "loss": 0.4019,
+      "step": 13862
+    },
+    {
+      "epoch": 0.9625746424107763,
+      "grad_norm": 3.0647322849142227,
+      "learning_rate": 3.669059738948544e-08,
+      "loss": 0.2647,
+      "step": 13863
+    },
+    {
+      "epoch": 0.9626440772114984,
+      "grad_norm": 4.9964011116632605,
+      "learning_rate": 3.6554746442152843e-08,
+      "loss": 0.5095,
+      "step": 13864
+    },
+    {
+      "epoch": 0.9627135120122206,
+      "grad_norm": 6.0119350438537795,
+      "learning_rate": 3.641914654026524e-08,
+      "loss": 0.5313,
+      "step": 13865
+    },
+    {
+      "epoch": 0.9627829468129426,
+      "grad_norm": 4.191329280499595,
+      "learning_rate": 3.6283797690680466e-08,
+      "loss": 0.3224,
+      "step": 13866
+    },
+    {
+      "epoch": 0.9628523816136648,
+      "grad_norm": 3.802544345205221,
+      "learning_rate": 3.614869990024417e-08,
+      "loss": 0.4551,
+      "step": 13867
+    },
+    {
+      "epoch": 0.9629218164143869,
+      "grad_norm": 4.182819548930081,
+      "learning_rate": 3.601385317578976e-08,
+      "loss": 0.2626,
+      "step": 13868
+    },
+    {
+      "epoch": 0.962991251215109,
+      "grad_norm": 5.381595996103115,
+      "learning_rate": 3.58792575241379e-08,
+      "loss": 0.4892,
+      "step": 13869
+    },
+    {
+      "epoch": 0.9630606860158312,
+      "grad_norm": 4.179363144423296,
+      "learning_rate": 3.5744912952095366e-08,
+      "loss": 0.441,
+      "step": 13870
+    },
+    {
+      "epoch": 0.9631301208165532,
+      "grad_norm": 3.09670662235687,
+      "learning_rate": 3.5610819466457834e-08,
+      "loss": 0.2507,
+      "step": 13871
+    },
+    {
+      "epoch": 0.9631995556172753,
+      "grad_norm": 4.087113320364595,
+      "learning_rate": 3.5476977074007654e-08,
+      "loss": 0.378,
+      "step": 13872
+    },
+    {
+      "epoch": 0.9632689904179975,
+      "grad_norm": 4.511752164745252,
+      "learning_rate": 3.5343385781513865e-08,
+      "loss": 0.5731,
+      "step": 13873
+    },
+    {
+      "epoch": 0.9633384252187196,
+      "grad_norm": 3.216509604128841,
+      "learning_rate": 3.521004559573438e-08,
+      "loss": 0.3191,
+      "step": 13874
+    },
+    {
+      "epoch": 0.9634078600194418,
+      "grad_norm": 3.3513061953819943,
+      "learning_rate": 3.507695652341214e-08,
+      "loss": 0.2203,
+      "step": 13875
+    },
+    {
+      "epoch": 0.9634772948201639,
+      "grad_norm": 4.513362211202487,
+      "learning_rate": 3.494411857127955e-08,
+      "loss": 0.6145,
+      "step": 13876
+    },
+    {
+      "epoch": 0.9635467296208859,
+      "grad_norm": 4.5578955948467215,
+      "learning_rate": 3.481153174605512e-08,
+      "loss": 0.5258,
+      "step": 13877
+    },
+    {
+      "epoch": 0.9636161644216081,
+      "grad_norm": 3.202541192684505,
+      "learning_rate": 3.467919605444458e-08,
+      "loss": 0.2442,
+      "step": 13878
+    },
+    {
+      "epoch": 0.9636855992223302,
+      "grad_norm": 3.1658711881299553,
+      "learning_rate": 3.454711150314205e-08,
+      "loss": 0.2255,
+      "step": 13879
+    },
+    {
+      "epoch": 0.9637550340230524,
+      "grad_norm": 3.515793754764216,
+      "learning_rate": 3.441527809882772e-08,
+      "loss": 0.1941,
+      "step": 13880
+    },
+    {
+      "epoch": 0.9638244688237745,
+      "grad_norm": 3.697013486145506,
+      "learning_rate": 3.4283695848169596e-08,
+      "loss": 0.3346,
+      "step": 13881
+    },
+    {
+      "epoch": 0.9638939036244966,
+      "grad_norm": 4.799943640311567,
+      "learning_rate": 3.415236475782291e-08,
+      "loss": 0.263,
+      "step": 13882
+    },
+    {
+      "epoch": 0.9639633384252188,
+      "grad_norm": 3.291851007490705,
+      "learning_rate": 3.402128483443068e-08,
+      "loss": 0.2072,
+      "step": 13883
+    },
+    {
+      "epoch": 0.9640327732259408,
+      "grad_norm": 2.839818722319311,
+      "learning_rate": 3.38904560846226e-08,
+      "loss": 0.2364,
+      "step": 13884
+    },
+    {
+      "epoch": 0.9641022080266629,
+      "grad_norm": 3.9703807204503305,
+      "learning_rate": 3.375987851501506e-08,
+      "loss": 0.3083,
+      "step": 13885
+    },
+    {
+      "epoch": 0.9641716428273851,
+      "grad_norm": 4.624207629727348,
+      "learning_rate": 3.362955213221386e-08,
+      "loss": 0.5569,
+      "step": 13886
+    },
+    {
+      "epoch": 0.9642410776281072,
+      "grad_norm": 3.4817969193369405,
+      "learning_rate": 3.349947694280986e-08,
+      "loss": 0.2156,
+      "step": 13887
+    },
+    {
+      "epoch": 0.9643105124288294,
+      "grad_norm": 3.691874359185989,
+      "learning_rate": 3.336965295338224e-08,
+      "loss": 0.4888,
+      "step": 13888
+    },
+    {
+      "epoch": 0.9643799472295514,
+      "grad_norm": 2.9049687006134506,
+      "learning_rate": 3.32400801704974e-08,
+      "loss": 0.2363,
+      "step": 13889
+    },
+    {
+      "epoch": 0.9644493820302735,
+      "grad_norm": 3.363611821759633,
+      "learning_rate": 3.311075860070956e-08,
+      "loss": 0.393,
+      "step": 13890
+    },
+    {
+      "epoch": 0.9645188168309957,
+      "grad_norm": 3.7991262378174597,
+      "learning_rate": 3.298168825055848e-08,
+      "loss": 0.3684,
+      "step": 13891
+    },
+    {
+      "epoch": 0.9645882516317178,
+      "grad_norm": 5.125278126407595,
+      "learning_rate": 3.2852869126573375e-08,
+      "loss": 0.7835,
+      "step": 13892
+    },
+    {
+      "epoch": 0.96465768643244,
+      "grad_norm": 3.940684579465366,
+      "learning_rate": 3.27243012352696e-08,
+      "loss": 0.4326,
+      "step": 13893
+    },
+    {
+      "epoch": 0.9647271212331621,
+      "grad_norm": 3.5985701664257577,
+      "learning_rate": 3.259598458314917e-08,
+      "loss": 0.4241,
+      "step": 13894
+    },
+    {
+      "epoch": 0.9647965560338841,
+      "grad_norm": 3.2520534912901007,
+      "learning_rate": 3.2467919176703554e-08,
+      "loss": 0.35,
+      "step": 13895
+    },
+    {
+      "epoch": 0.9648659908346063,
+      "grad_norm": 3.478673188989901,
+      "learning_rate": 3.234010502240925e-08,
+      "loss": 0.2868,
+      "step": 13896
+    },
+    {
+      "epoch": 0.9649354256353284,
+      "grad_norm": 3.625263396999371,
+      "learning_rate": 3.221254212673108e-08,
+      "loss": 0.3143,
+      "step": 13897
+    },
+    {
+      "epoch": 0.9650048604360505,
+      "grad_norm": 4.651318088961176,
+      "learning_rate": 3.2085230496121113e-08,
+      "loss": 0.5551,
+      "step": 13898
+    },
+    {
+      "epoch": 0.9650742952367727,
+      "grad_norm": 3.9267170889251823,
+      "learning_rate": 3.195817013701863e-08,
+      "loss": 0.2931,
+      "step": 13899
+    },
+    {
+      "epoch": 0.9651437300374948,
+      "grad_norm": 3.9273372507117843,
+      "learning_rate": 3.183136105584961e-08,
+      "loss": 0.4554,
+      "step": 13900
+    },
+    {
+      "epoch": 0.965213164838217,
+      "grad_norm": 3.5188110258859866,
+      "learning_rate": 3.170480325902947e-08,
+      "loss": 0.3658,
+      "step": 13901
+    },
+    {
+      "epoch": 0.965282599638939,
+      "grad_norm": 3.746477851652635,
+      "learning_rate": 3.157849675295754e-08,
+      "loss": 0.3493,
+      "step": 13902
+    },
+    {
+      "epoch": 0.9653520344396611,
+      "grad_norm": 3.937142812994007,
+      "learning_rate": 3.145244154402316e-08,
+      "loss": 0.3912,
+      "step": 13903
+    },
+    {
+      "epoch": 0.9654214692403833,
+      "grad_norm": 4.259630467268845,
+      "learning_rate": 3.132663763860233e-08,
+      "loss": 0.5141,
+      "step": 13904
+    },
+    {
+      "epoch": 0.9654909040411054,
+      "grad_norm": 3.824007358663462,
+      "learning_rate": 3.120108504305774e-08,
+      "loss": 0.4208,
+      "step": 13905
+    },
+    {
+      "epoch": 0.9655603388418276,
+      "grad_norm": 7.103374291264321,
+      "learning_rate": 3.1075783763739876e-08,
+      "loss": 0.6728,
+      "step": 13906
+    },
+    {
+      "epoch": 0.9656297736425496,
+      "grad_norm": 3.716147070598669,
+      "learning_rate": 3.095073380698588e-08,
+      "loss": 0.4646,
+      "step": 13907
+    },
+    {
+      "epoch": 0.9656992084432717,
+      "grad_norm": 4.21908422581669,
+      "learning_rate": 3.082593517912069e-08,
+      "loss": 0.4748,
+      "step": 13908
+    },
+    {
+      "epoch": 0.9657686432439939,
+      "grad_norm": 3.5916909202712373,
+      "learning_rate": 3.070138788645704e-08,
+      "loss": 0.4576,
+      "step": 13909
+    },
+    {
+      "epoch": 0.965838078044716,
+      "grad_norm": 4.029572919934993,
+      "learning_rate": 3.0577091935294345e-08,
+      "loss": 0.6416,
+      "step": 13910
+    },
+    {
+      "epoch": 0.9659075128454382,
+      "grad_norm": 5.220624757156848,
+      "learning_rate": 3.045304733191812e-08,
+      "loss": 0.5375,
+      "step": 13911
+    },
+    {
+      "epoch": 0.9659769476461603,
+      "grad_norm": 2.1084965171159125,
+      "learning_rate": 3.0329254082604454e-08,
+      "loss": 0.084,
+      "step": 13912
+    },
+    {
+      "epoch": 0.9660463824468823,
+      "grad_norm": 3.4264187075226173,
+      "learning_rate": 3.020571219361279e-08,
+      "loss": 0.3971,
+      "step": 13913
+    },
+    {
+      "epoch": 0.9661158172476045,
+      "grad_norm": 2.8629674011372024,
+      "learning_rate": 3.0082421671192576e-08,
+      "loss": 0.2313,
+      "step": 13914
+    },
+    {
+      "epoch": 0.9661852520483266,
+      "grad_norm": 4.098420077202031,
+      "learning_rate": 2.9959382521579926e-08,
+      "loss": 0.4725,
+      "step": 13915
+    },
+    {
+      "epoch": 0.9662546868490487,
+      "grad_norm": 3.526280752932261,
+      "learning_rate": 2.983659475099765e-08,
+      "loss": 0.4076,
+      "step": 13916
+    },
+    {
+      "epoch": 0.9663241216497709,
+      "grad_norm": 4.028203215288937,
+      "learning_rate": 2.9714058365655775e-08,
+      "loss": 0.5523,
+      "step": 13917
+    },
+    {
+      "epoch": 0.966393556450493,
+      "grad_norm": 4.108010214760812,
+      "learning_rate": 2.959177337175323e-08,
+      "loss": 0.4114,
+      "step": 13918
+    },
+    {
+      "epoch": 0.9664629912512152,
+      "grad_norm": 4.635602410599053,
+      "learning_rate": 2.946973977547396e-08,
+      "loss": 0.3416,
+      "step": 13919
+    },
+    {
+      "epoch": 0.9665324260519372,
+      "grad_norm": 3.6684954895781363,
+      "learning_rate": 2.9347957582990806e-08,
+      "loss": 0.3796,
+      "step": 13920
+    },
+    {
+      "epoch": 0.9666018608526593,
+      "grad_norm": 3.499867262010852,
+      "learning_rate": 2.922642680046328e-08,
+      "loss": 0.2188,
+      "step": 13921
+    },
+    {
+      "epoch": 0.9666712956533815,
+      "grad_norm": 5.50023340497982,
+      "learning_rate": 2.9105147434038693e-08,
+      "loss": 0.5761,
+      "step": 13922
+    },
+    {
+      "epoch": 0.9667407304541036,
+      "grad_norm": 3.5674508432278267,
+      "learning_rate": 2.898411948985047e-08,
+      "loss": 0.4676,
+      "step": 13923
+    },
+    {
+      "epoch": 0.9668101652548258,
+      "grad_norm": 2.87132609284704,
+      "learning_rate": 2.886334297402038e-08,
+      "loss": 0.2831,
+      "step": 13924
+    },
+    {
+      "epoch": 0.9668796000555478,
+      "grad_norm": 3.5372852542334465,
+      "learning_rate": 2.8742817892657427e-08,
+      "loss": 0.3986,
+      "step": 13925
+    },
+    {
+      "epoch": 0.9669490348562699,
+      "grad_norm": 3.869300358259012,
+      "learning_rate": 2.862254425185673e-08,
+      "loss": 0.5334,
+      "step": 13926
+    },
+    {
+      "epoch": 0.9670184696569921,
+      "grad_norm": 3.475380526051635,
+      "learning_rate": 2.8502522057702876e-08,
+      "loss": 0.3608,
+      "step": 13927
+    },
+    {
+      "epoch": 0.9670879044577142,
+      "grad_norm": 4.050096232006329,
+      "learning_rate": 2.8382751316266e-08,
+      "loss": 0.3324,
+      "step": 13928
+    },
+    {
+      "epoch": 0.9671573392584363,
+      "grad_norm": 3.790407215323487,
+      "learning_rate": 2.8263232033602926e-08,
+      "loss": 0.2709,
+      "step": 13929
+    },
+    {
+      "epoch": 0.9672267740591585,
+      "grad_norm": 4.753498772660416,
+      "learning_rate": 2.814396421576049e-08,
+      "loss": 0.5267,
+      "step": 13930
+    },
+    {
+      "epoch": 0.9672962088598805,
+      "grad_norm": 4.945551648944414,
+      "learning_rate": 2.802494786876997e-08,
+      "loss": 0.5693,
+      "step": 13931
+    },
+    {
+      "epoch": 0.9673656436606027,
+      "grad_norm": 3.178679104435819,
+      "learning_rate": 2.7906182998650998e-08,
+      "loss": 0.2418,
+      "step": 13932
+    },
+    {
+      "epoch": 0.9674350784613248,
+      "grad_norm": 4.115920893131331,
+      "learning_rate": 2.7787669611412106e-08,
+      "loss": 0.2757,
+      "step": 13933
+    },
+    {
+      "epoch": 0.9675045132620469,
+      "grad_norm": 4.068440622896221,
+      "learning_rate": 2.766940771304516e-08,
+      "loss": 0.2915,
+      "step": 13934
+    },
+    {
+      "epoch": 0.9675739480627691,
+      "grad_norm": 4.514547329071096,
+      "learning_rate": 2.7551397309533712e-08,
+      "loss": 0.6216,
+      "step": 13935
+    },
+    {
+      "epoch": 0.9676433828634912,
+      "grad_norm": 2.714369859630633,
+      "learning_rate": 2.7433638406845763e-08,
+      "loss": 0.1784,
+      "step": 13936
+    },
+    {
+      "epoch": 0.9677128176642134,
+      "grad_norm": 5.4820121104677035,
+      "learning_rate": 2.7316131010937664e-08,
+      "loss": 0.5198,
+      "step": 13937
+    },
+    {
+      "epoch": 0.9677822524649354,
+      "grad_norm": 3.3829991763733127,
+      "learning_rate": 2.7198875127752987e-08,
+      "loss": 0.3666,
+      "step": 13938
+    },
+    {
+      "epoch": 0.9678516872656575,
+      "grad_norm": 4.45692373172024,
+      "learning_rate": 2.708187076322144e-08,
+      "loss": 0.6309,
+      "step": 13939
+    },
+    {
+      "epoch": 0.9679211220663797,
+      "grad_norm": 4.379398574176032,
+      "learning_rate": 2.6965117923262175e-08,
+      "loss": 0.4919,
+      "step": 13940
+    },
+    {
+      "epoch": 0.9679905568671018,
+      "grad_norm": 5.470372126905691,
+      "learning_rate": 2.6848616613779355e-08,
+      "loss": 0.5843,
+      "step": 13941
+    },
+    {
+      "epoch": 0.9680599916678239,
+      "grad_norm": 4.03698712234339,
+      "learning_rate": 2.6732366840666047e-08,
+      "loss": 0.4028,
+      "step": 13942
+    },
+    {
+      "epoch": 0.968129426468546,
+      "grad_norm": 3.3348693931461693,
+      "learning_rate": 2.661636860980199e-08,
+      "loss": 0.3513,
+      "step": 13943
+    },
+    {
+      "epoch": 0.9681988612692681,
+      "grad_norm": 3.972175910683446,
+      "learning_rate": 2.6500621927054716e-08,
+      "loss": 0.3687,
+      "step": 13944
+    },
+    {
+      "epoch": 0.9682682960699903,
+      "grad_norm": 4.860776908972846,
+      "learning_rate": 2.638512679827787e-08,
+      "loss": 0.5181,
+      "step": 13945
+    },
+    {
+      "epoch": 0.9683377308707124,
+      "grad_norm": 3.4581590627111676,
+      "learning_rate": 2.6269883229312897e-08,
+      "loss": 0.295,
+      "step": 13946
+    },
+    {
+      "epoch": 0.9684071656714345,
+      "grad_norm": 3.0561549584548975,
+      "learning_rate": 2.615489122599013e-08,
+      "loss": 0.2337,
+      "step": 13947
+    },
+    {
+      "epoch": 0.9684766004721567,
+      "grad_norm": 3.299673438627948,
+      "learning_rate": 2.604015079412381e-08,
+      "loss": 0.3189,
+      "step": 13948
+    },
+    {
+      "epoch": 0.9685460352728787,
+      "grad_norm": 3.648966254590199,
+      "learning_rate": 2.5925661939518176e-08,
+      "loss": 0.3434,
+      "step": 13949
+    },
+    {
+      "epoch": 0.9686154700736009,
+      "grad_norm": 3.384671444658221,
+      "learning_rate": 2.581142466796416e-08,
+      "loss": 0.3467,
+      "step": 13950
+    },
+    {
+      "epoch": 0.968684904874323,
+      "grad_norm": 4.618378282876279,
+      "learning_rate": 2.5697438985239908e-08,
+      "loss": 0.5661,
+      "step": 13951
+    },
+    {
+      "epoch": 0.9687543396750451,
+      "grad_norm": 3.399563428698885,
+      "learning_rate": 2.558370489711026e-08,
+      "loss": 0.3341,
+      "step": 13952
+    },
+    {
+      "epoch": 0.9688237744757673,
+      "grad_norm": 5.039923006690745,
+      "learning_rate": 2.5470222409327282e-08,
+      "loss": 0.5195,
+      "step": 13953
+    },
+    {
+      "epoch": 0.9688932092764894,
+      "grad_norm": 5.5412437171272835,
+      "learning_rate": 2.5356991527631934e-08,
+      "loss": 0.5078,
+      "step": 13954
+    },
+    {
+      "epoch": 0.9689626440772114,
+      "grad_norm": 4.684621506584292,
+      "learning_rate": 2.524401225775075e-08,
+      "loss": 0.5066,
+      "step": 13955
+    },
+    {
+      "epoch": 0.9690320788779336,
+      "grad_norm": 4.354608301912683,
+      "learning_rate": 2.5131284605398597e-08,
+      "loss": 0.5612,
+      "step": 13956
+    },
+    {
+      "epoch": 0.9691015136786557,
+      "grad_norm": 2.878837585698226,
+      "learning_rate": 2.501880857627592e-08,
+      "loss": 0.1896,
+      "step": 13957
+    },
+    {
+      "epoch": 0.9691709484793779,
+      "grad_norm": 4.692310733848334,
+      "learning_rate": 2.4906584176072612e-08,
+      "loss": 0.6131,
+      "step": 13958
+    },
+    {
+      "epoch": 0.9692403832801,
+      "grad_norm": 4.542866085970759,
+      "learning_rate": 2.4794611410464132e-08,
+      "loss": 0.3039,
+      "step": 13959
+    },
+    {
+      "epoch": 0.9693098180808221,
+      "grad_norm": 4.400934802127237,
+      "learning_rate": 2.4682890285114835e-08,
+      "loss": 0.5373,
+      "step": 13960
+    },
+    {
+      "epoch": 0.9693792528815443,
+      "grad_norm": 4.139093443720174,
+      "learning_rate": 2.457142080567465e-08,
+      "loss": 0.4823,
+      "step": 13961
+    },
+    {
+      "epoch": 0.9694486876822663,
+      "grad_norm": 3.096100549334194,
+      "learning_rate": 2.44602029777824e-08,
+      "loss": 0.2249,
+      "step": 13962
+    },
+    {
+      "epoch": 0.9695181224829885,
+      "grad_norm": 4.129648384323139,
+      "learning_rate": 2.434923680706247e-08,
+      "loss": 0.3959,
+      "step": 13963
+    },
+    {
+      "epoch": 0.9695875572837106,
+      "grad_norm": 3.5392475179846046,
+      "learning_rate": 2.4238522299128153e-08,
+      "loss": 0.3351,
+      "step": 13964
+    },
+    {
+      "epoch": 0.9696569920844327,
+      "grad_norm": 4.5226004228687735,
+      "learning_rate": 2.4128059459578857e-08,
+      "loss": 0.6346,
+      "step": 13965
+    },
+    {
+      "epoch": 0.9697264268851549,
+      "grad_norm": 4.413413076664162,
+      "learning_rate": 2.401784829400178e-08,
+      "loss": 0.3915,
+      "step": 13966
+    },
+    {
+      "epoch": 0.969795861685877,
+      "grad_norm": 3.383218116125762,
+      "learning_rate": 2.390788880797079e-08,
+      "loss": 0.4067,
+      "step": 13967
+    },
+    {
+      "epoch": 0.9698652964865991,
+      "grad_norm": 2.589696996819408,
+      "learning_rate": 2.379818100704867e-08,
+      "loss": 0.1912,
+      "step": 13968
+    },
+    {
+      "epoch": 0.9699347312873212,
+      "grad_norm": 4.407675251681019,
+      "learning_rate": 2.36887248967832e-08,
+      "loss": 0.406,
+      "step": 13969
+    },
+    {
+      "epoch": 0.9700041660880433,
+      "grad_norm": 4.516638036766161,
+      "learning_rate": 2.3579520482710504e-08,
+      "loss": 0.5152,
+      "step": 13970
+    },
+    {
+      "epoch": 0.9700736008887655,
+      "grad_norm": 4.604787964476955,
+      "learning_rate": 2.3470567770355058e-08,
+      "loss": 0.7047,
+      "step": 13971
+    },
+    {
+      "epoch": 0.9701430356894876,
+      "grad_norm": 3.9292022770929997,
+      "learning_rate": 2.3361866765226894e-08,
+      "loss": 0.4295,
+      "step": 13972
+    },
+    {
+      "epoch": 0.9702124704902096,
+      "grad_norm": 4.628537408208618,
+      "learning_rate": 2.3253417472823837e-08,
+      "loss": 0.4073,
+      "step": 13973
+    },
+    {
+      "epoch": 0.9702819052909318,
+      "grad_norm": 3.7994650360175917,
+      "learning_rate": 2.3145219898631498e-08,
+      "loss": 0.4849,
+      "step": 13974
+    },
+    {
+      "epoch": 0.9703513400916539,
+      "grad_norm": 5.082416493744106,
+      "learning_rate": 2.3037274048122173e-08,
+      "loss": 0.6425,
+      "step": 13975
+    },
+    {
+      "epoch": 0.9704207748923761,
+      "grad_norm": 4.4076748883366355,
+      "learning_rate": 2.292957992675593e-08,
+      "loss": 0.4873,
+      "step": 13976
+    },
+    {
+      "epoch": 0.9704902096930982,
+      "grad_norm": 3.7700442046890354,
+      "learning_rate": 2.2822137539979528e-08,
+      "loss": 0.2949,
+      "step": 13977
+    },
+    {
+      "epoch": 0.9705596444938203,
+      "grad_norm": 4.593228253565943,
+      "learning_rate": 2.2714946893226952e-08,
+      "loss": 0.3469,
+      "step": 13978
+    },
+    {
+      "epoch": 0.9706290792945425,
+      "grad_norm": 4.036183114120982,
+      "learning_rate": 2.2608007991920532e-08,
+      "loss": 0.402,
+      "step": 13979
+    },
+    {
+      "epoch": 0.9706985140952645,
+      "grad_norm": 5.216959280318881,
+      "learning_rate": 2.2501320841468722e-08,
+      "loss": 0.4756,
+      "step": 13980
+    },
+    {
+      "epoch": 0.9707679488959867,
+      "grad_norm": 3.5811915559320036,
+      "learning_rate": 2.2394885447267757e-08,
+      "loss": 0.4073,
+      "step": 13981
+    },
+    {
+      "epoch": 0.9708373836967088,
+      "grad_norm": 3.962072468992664,
+      "learning_rate": 2.2288701814701107e-08,
+      "loss": 0.3681,
+      "step": 13982
+    },
+    {
+      "epoch": 0.9709068184974309,
+      "grad_norm": 3.857228661279531,
+      "learning_rate": 2.2182769949138928e-08,
+      "loss": 0.4605,
+      "step": 13983
+    },
+    {
+      "epoch": 0.9709762532981531,
+      "grad_norm": 5.4152422344940705,
+      "learning_rate": 2.2077089855939705e-08,
+      "loss": 0.388,
+      "step": 13984
+    },
+    {
+      "epoch": 0.9710456880988751,
+      "grad_norm": 4.665761789217447,
+      "learning_rate": 2.1971661540448053e-08,
+      "loss": 0.5015,
+      "step": 13985
+    },
+    {
+      "epoch": 0.9711151228995972,
+      "grad_norm": 3.4001502341896264,
+      "learning_rate": 2.186648500799693e-08,
+      "loss": 0.2066,
+      "step": 13986
+    },
+    {
+      "epoch": 0.9711845577003194,
+      "grad_norm": 3.597338897756042,
+      "learning_rate": 2.1761560263905968e-08,
+      "loss": 0.4317,
+      "step": 13987
+    },
+    {
+      "epoch": 0.9712539925010415,
+      "grad_norm": 4.142210765077964,
+      "learning_rate": 2.1656887313482034e-08,
+      "loss": 0.498,
+      "step": 13988
+    },
+    {
+      "epoch": 0.9713234273017637,
+      "grad_norm": 4.25314399287038,
+      "learning_rate": 2.1552466162019224e-08,
+      "loss": 0.5422,
+      "step": 13989
+    },
+    {
+      "epoch": 0.9713928621024858,
+      "grad_norm": 3.9778184826208363,
+      "learning_rate": 2.1448296814798873e-08,
+      "loss": 0.4588,
+      "step": 13990
+    },
+    {
+      "epoch": 0.9714622969032078,
+      "grad_norm": 4.1594816750134305,
+      "learning_rate": 2.1344379277090653e-08,
+      "loss": 0.4741,
+      "step": 13991
+    },
+    {
+      "epoch": 0.97153173170393,
+      "grad_norm": 3.7406261776150664,
+      "learning_rate": 2.124071355414925e-08,
+      "loss": 0.2361,
+      "step": 13992
+    },
+    {
+      "epoch": 0.9716011665046521,
+      "grad_norm": 5.712416804111109,
+      "learning_rate": 2.1137299651219357e-08,
+      "loss": 0.3263,
+      "step": 13993
+    },
+    {
+      "epoch": 0.9716706013053743,
+      "grad_norm": 3.2608341873698468,
+      "learning_rate": 2.1034137573531232e-08,
+      "loss": 0.1761,
+      "step": 13994
+    },
+    {
+      "epoch": 0.9717400361060964,
+      "grad_norm": 3.99311135850991,
+      "learning_rate": 2.0931227326301262e-08,
+      "loss": 0.4674,
+      "step": 13995
+    },
+    {
+      "epoch": 0.9718094709068185,
+      "grad_norm": 4.505905708922432,
+      "learning_rate": 2.0828568914736392e-08,
+      "loss": 0.5078,
+      "step": 13996
+    },
+    {
+      "epoch": 0.9718789057075407,
+      "grad_norm": 2.8261682407864397,
+      "learning_rate": 2.0726162344028022e-08,
+      "loss": 0.1952,
+      "step": 13997
+    },
+    {
+      "epoch": 0.9719483405082627,
+      "grad_norm": 4.0033332747957315,
+      "learning_rate": 2.0624007619355903e-08,
+      "loss": 0.2589,
+      "step": 13998
+    },
+    {
+      "epoch": 0.9720177753089848,
+      "grad_norm": 9.477326408079183,
+      "learning_rate": 2.0522104745887006e-08,
+      "loss": 0.4496,
+      "step": 13999
+    },
+    {
+      "epoch": 0.972087210109707,
+      "grad_norm": 5.021280389494722,
+      "learning_rate": 2.042045372877499e-08,
+      "loss": 0.3323,
+      "step": 14000
+    },
+    {
+      "epoch": 0.9721566449104291,
+      "grad_norm": 4.302617182703042,
+      "learning_rate": 2.0319054573162412e-08,
+      "loss": 0.2489,
+      "step": 14001
+    },
+    {
+      "epoch": 0.9722260797111513,
+      "grad_norm": 3.610580336626151,
+      "learning_rate": 2.0217907284176274e-08,
+      "loss": 0.219,
+      "step": 14002
+    },
+    {
+      "epoch": 0.9722955145118733,
+      "grad_norm": 4.345118384187324,
+      "learning_rate": 2.0117011866934156e-08,
+      "loss": 0.5438,
+      "step": 14003
+    },
+    {
+      "epoch": 0.9723649493125954,
+      "grad_norm": 3.8227164733982075,
+      "learning_rate": 2.0016368326538084e-08,
+      "loss": 0.3896,
+      "step": 14004
+    },
+    {
+      "epoch": 0.9724343841133176,
+      "grad_norm": 1.7397228874314912,
+      "learning_rate": 1.9915976668078983e-08,
+      "loss": 0.0844,
+      "step": 14005
+    },
+    {
+      "epoch": 0.9725038189140397,
+      "grad_norm": 4.553455310483831,
+      "learning_rate": 1.9815836896634466e-08,
+      "loss": 0.6173,
+      "step": 14006
+    },
+    {
+      "epoch": 0.9725732537147619,
+      "grad_norm": 4.1421629355474785,
+      "learning_rate": 1.971594901726881e-08,
+      "loss": 0.4996,
+      "step": 14007
+    },
+    {
+      "epoch": 0.972642688515484,
+      "grad_norm": 4.355582354489459,
+      "learning_rate": 1.961631303503575e-08,
+      "loss": 0.7239,
+      "step": 14008
+    },
+    {
+      "epoch": 0.972712123316206,
+      "grad_norm": 5.153257124962559,
+      "learning_rate": 1.951692895497348e-08,
+      "loss": 0.7,
+      "step": 14009
+    },
+    {
+      "epoch": 0.9727815581169282,
+      "grad_norm": 3.996943419966421,
+      "learning_rate": 1.941779678210909e-08,
+      "loss": 0.4551,
+      "step": 14010
+    },
+    {
+      "epoch": 0.9728509929176503,
+      "grad_norm": 4.153744690478692,
+      "learning_rate": 1.9318916521456897e-08,
+      "loss": 0.4887,
+      "step": 14011
+    },
+    {
+      "epoch": 0.9729204277183724,
+      "grad_norm": 3.962604828600932,
+      "learning_rate": 1.922028817801791e-08,
+      "loss": 0.4199,
+      "step": 14012
+    },
+    {
+      "epoch": 0.9729898625190946,
+      "grad_norm": 4.870594754406823,
+      "learning_rate": 1.9121911756780355e-08,
+      "loss": 0.4947,
+      "step": 14013
+    },
+    {
+      "epoch": 0.9730592973198167,
+      "grad_norm": 4.281700257933232,
+      "learning_rate": 1.902378726272025e-08,
+      "loss": 0.3289,
+      "step": 14014
+    },
+    {
+      "epoch": 0.9731287321205389,
+      "grad_norm": 3.558708207213513,
+      "learning_rate": 1.892591470080085e-08,
+      "loss": 0.3555,
+      "step": 14015
+    },
+    {
+      "epoch": 0.9731981669212609,
+      "grad_norm": 3.4708083130096505,
+      "learning_rate": 1.8828294075972085e-08,
+      "loss": 0.2805,
+      "step": 14016
+    },
+    {
+      "epoch": 0.973267601721983,
+      "grad_norm": 3.6734092229500144,
+      "learning_rate": 1.8730925393171674e-08,
+      "loss": 0.2815,
+      "step": 14017
+    },
+    {
+      "epoch": 0.9733370365227052,
+      "grad_norm": 3.0859388753168786,
+      "learning_rate": 1.8633808657324558e-08,
+      "loss": 0.2198,
+      "step": 14018
+    },
+    {
+      "epoch": 0.9734064713234273,
+      "grad_norm": 3.924874347953349,
+      "learning_rate": 1.8536943873342374e-08,
+      "loss": 0.5066,
+      "step": 14019
+    },
+    {
+      "epoch": 0.9734759061241495,
+      "grad_norm": 2.799261342492207,
+      "learning_rate": 1.8440331046124525e-08,
+      "loss": 0.1969,
+      "step": 14020
+    },
+    {
+      "epoch": 0.9735453409248715,
+      "grad_norm": 5.383938386457754,
+      "learning_rate": 1.8343970180557668e-08,
+      "loss": 0.5267,
+      "step": 14021
+    },
+    {
+      "epoch": 0.9736147757255936,
+      "grad_norm": 2.6840559973776013,
+      "learning_rate": 1.8247861281516232e-08,
+      "loss": 0.2126,
+      "step": 14022
+    },
+    {
+      "epoch": 0.9736842105263158,
+      "grad_norm": 4.4850777318586585,
+      "learning_rate": 1.815200435386022e-08,
+      "loss": 0.5508,
+      "step": 14023
+    },
+    {
+      "epoch": 0.9737536453270379,
+      "grad_norm": 4.594805184212075,
+      "learning_rate": 1.805639940243853e-08,
+      "loss": 0.6097,
+      "step": 14024
+    },
+    {
+      "epoch": 0.9738230801277601,
+      "grad_norm": 2.7276943312260005,
+      "learning_rate": 1.7961046432086736e-08,
+      "loss": 0.172,
+      "step": 14025
+    },
+    {
+      "epoch": 0.9738925149284822,
+      "grad_norm": 3.6331541129600162,
+      "learning_rate": 1.7865945447628208e-08,
+      "loss": 0.3603,
+      "step": 14026
+    },
+    {
+      "epoch": 0.9739619497292042,
+      "grad_norm": 4.995939059703874,
+      "learning_rate": 1.7771096453871873e-08,
+      "loss": 0.5302,
+      "step": 14027
+    },
+    {
+      "epoch": 0.9740313845299264,
+      "grad_norm": 4.931120298441444,
+      "learning_rate": 1.7676499455615557e-08,
+      "loss": 0.6647,
+      "step": 14028
+    },
+    {
+      "epoch": 0.9741008193306485,
+      "grad_norm": 3.702582140068964,
+      "learning_rate": 1.758215445764433e-08,
+      "loss": 0.314,
+      "step": 14029
+    },
+    {
+      "epoch": 0.9741702541313706,
+      "grad_norm": 3.7634410155971256,
+      "learning_rate": 1.748806146472992e-08,
+      "loss": 0.3839,
+      "step": 14030
+    },
+    {
+      "epoch": 0.9742396889320928,
+      "grad_norm": 3.2031003032610417,
+      "learning_rate": 1.739422048163131e-08,
+      "loss": 0.3455,
+      "step": 14031
+    },
+    {
+      "epoch": 0.9743091237328149,
+      "grad_norm": 4.242435519955967,
+      "learning_rate": 1.730063151309469e-08,
+      "loss": 0.3657,
+      "step": 14032
+    },
+    {
+      "epoch": 0.974378558533537,
+      "grad_norm": 2.776620079247058,
+      "learning_rate": 1.7207294563853505e-08,
+      "loss": 0.2832,
+      "step": 14033
+    },
+    {
+      "epoch": 0.9744479933342591,
+      "grad_norm": 4.650814779014532,
+      "learning_rate": 1.7114209638628975e-08,
+      "loss": 0.339,
+      "step": 14034
+    },
+    {
+      "epoch": 0.9745174281349812,
+      "grad_norm": 2.997803483146635,
+      "learning_rate": 1.702137674212956e-08,
+      "loss": 0.2789,
+      "step": 14035
+    },
+    {
+      "epoch": 0.9745868629357034,
+      "grad_norm": 2.9351497361518826,
+      "learning_rate": 1.692879587904983e-08,
+      "loss": 0.2489,
+      "step": 14036
+    },
+    {
+      "epoch": 0.9746562977364255,
+      "grad_norm": 3.8783029495589414,
+      "learning_rate": 1.6836467054073825e-08,
+      "loss": 0.3264,
+      "step": 14037
+    },
+    {
+      "epoch": 0.9747257325371477,
+      "grad_norm": 5.293899694224254,
+      "learning_rate": 1.674439027186947e-08,
+      "loss": 0.6804,
+      "step": 14038
+    },
+    {
+      "epoch": 0.9747951673378698,
+      "grad_norm": 4.468216425530285,
+      "learning_rate": 1.6652565537095268e-08,
+      "loss": 0.6283,
+      "step": 14039
+    },
+    {
+      "epoch": 0.9748646021385918,
+      "grad_norm": 5.453208909883312,
+      "learning_rate": 1.6560992854395274e-08,
+      "loss": 0.7144,
+      "step": 14040
+    },
+    {
+      "epoch": 0.974934036939314,
+      "grad_norm": 3.8511193064651463,
+      "learning_rate": 1.6469672228401345e-08,
+      "loss": 0.4038,
+      "step": 14041
+    },
+    {
+      "epoch": 0.9750034717400361,
+      "grad_norm": 4.650049111742977,
+      "learning_rate": 1.6378603663732006e-08,
+      "loss": 0.4701,
+      "step": 14042
+    },
+    {
+      "epoch": 0.9750729065407582,
+      "grad_norm": 3.366645068576717,
+      "learning_rate": 1.628778716499302e-08,
+      "loss": 0.2936,
+      "step": 14043
+    },
+    {
+      "epoch": 0.9751423413414804,
+      "grad_norm": 3.4702266533829884,
+      "learning_rate": 1.619722273677904e-08,
+      "loss": 0.2644,
+      "step": 14044
+    },
+    {
+      "epoch": 0.9752117761422024,
+      "grad_norm": 5.199502611004693,
+      "learning_rate": 1.610691038366974e-08,
+      "loss": 0.6135,
+      "step": 14045
+    },
+    {
+      "epoch": 0.9752812109429246,
+      "grad_norm": 5.093024797791528,
+      "learning_rate": 1.6016850110233128e-08,
+      "loss": 0.6644,
+      "step": 14046
+    },
+    {
+      "epoch": 0.9753506457436467,
+      "grad_norm": 4.637626062284038,
+      "learning_rate": 1.5927041921024456e-08,
+      "loss": 0.5238,
+      "step": 14047
+    },
+    {
+      "epoch": 0.9754200805443688,
+      "grad_norm": 4.245040285105386,
+      "learning_rate": 1.5837485820586197e-08,
+      "loss": 0.3344,
+      "step": 14048
+    },
+    {
+      "epoch": 0.975489515345091,
+      "grad_norm": 3.806489248533676,
+      "learning_rate": 1.574818181344806e-08,
+      "loss": 0.3949,
+      "step": 14049
+    },
+    {
+      "epoch": 0.9755589501458131,
+      "grad_norm": 4.715285661589339,
+      "learning_rate": 1.565912990412699e-08,
+      "loss": 0.5764,
+      "step": 14050
+    },
+    {
+      "epoch": 0.9756283849465353,
+      "grad_norm": 4.132531433645996,
+      "learning_rate": 1.5570330097126607e-08,
+      "loss": 0.4733,
+      "step": 14051
+    },
+    {
+      "epoch": 0.9756978197472573,
+      "grad_norm": 3.2754803373674473,
+      "learning_rate": 1.5481782396938872e-08,
+      "loss": 0.3494,
+      "step": 14052
+    },
+    {
+      "epoch": 0.9757672545479794,
+      "grad_norm": 3.3347359764122073,
+      "learning_rate": 1.5393486808041867e-08,
+      "loss": 0.3196,
+      "step": 14053
+    },
+    {
+      "epoch": 0.9758366893487016,
+      "grad_norm": 4.447169383680507,
+      "learning_rate": 1.5305443334902583e-08,
+      "loss": 0.4422,
+      "step": 14054
+    },
+    {
+      "epoch": 0.9759061241494237,
+      "grad_norm": 3.969648705020858,
+      "learning_rate": 1.5217651981973004e-08,
+      "loss": 0.3482,
+      "step": 14055
+    },
+    {
+      "epoch": 0.9759755589501458,
+      "grad_norm": 2.1921582542511713,
+      "learning_rate": 1.513011275369347e-08,
+      "loss": 0.1054,
+      "step": 14056
+    },
+    {
+      "epoch": 0.976044993750868,
+      "grad_norm": 3.8832505255716,
+      "learning_rate": 1.5042825654492665e-08,
+      "loss": 0.3924,
+      "step": 14057
+    },
+    {
+      "epoch": 0.97611442855159,
+      "grad_norm": 3.289537179477533,
+      "learning_rate": 1.4955790688784833e-08,
+      "loss": 0.3595,
+      "step": 14058
+    },
+    {
+      "epoch": 0.9761838633523122,
+      "grad_norm": 3.0644739477825396,
+      "learning_rate": 1.4869007860972007e-08,
+      "loss": 0.2068,
+      "step": 14059
+    },
+    {
+      "epoch": 0.9762532981530343,
+      "grad_norm": 4.096612005375542,
+      "learning_rate": 1.4782477175443456e-08,
+      "loss": 0.3175,
+      "step": 14060
+    },
+    {
+      "epoch": 0.9763227329537564,
+      "grad_norm": 3.667531207697449,
+      "learning_rate": 1.4696198636576231e-08,
+      "loss": 0.4892,
+      "step": 14061
+    },
+    {
+      "epoch": 0.9763921677544786,
+      "grad_norm": 3.9123484108005058,
+      "learning_rate": 1.4610172248734067e-08,
+      "loss": 0.575,
+      "step": 14062
+    },
+    {
+      "epoch": 0.9764616025552006,
+      "grad_norm": 4.419122039267187,
+      "learning_rate": 1.4524398016267927e-08,
+      "loss": 0.5252,
+      "step": 14063
+    },
+    {
+      "epoch": 0.9765310373559228,
+      "grad_norm": 2.884299421029459,
+      "learning_rate": 1.4438875943516562e-08,
+      "loss": 0.3354,
+      "step": 14064
+    },
+    {
+      "epoch": 0.9766004721566449,
+      "grad_norm": 4.4642855114783275,
+      "learning_rate": 1.43536060348054e-08,
+      "loss": 0.4586,
+      "step": 14065
+    },
+    {
+      "epoch": 0.976669906957367,
+      "grad_norm": 3.725707585551653,
+      "learning_rate": 1.4268588294447105e-08,
+      "loss": 0.4864,
+      "step": 14066
+    },
+    {
+      "epoch": 0.9767393417580892,
+      "grad_norm": 4.453230055797431,
+      "learning_rate": 1.418382272674157e-08,
+      "loss": 0.5683,
+      "step": 14067
+    },
+    {
+      "epoch": 0.9768087765588113,
+      "grad_norm": 3.72750020637448,
+      "learning_rate": 1.4099309335976474e-08,
+      "loss": 0.2964,
+      "step": 14068
+    },
+    {
+      "epoch": 0.9768782113595333,
+      "grad_norm": 3.645604244971247,
+      "learning_rate": 1.4015048126426734e-08,
+      "loss": 0.4134,
+      "step": 14069
+    },
+    {
+      "epoch": 0.9769476461602555,
+      "grad_norm": 5.784177350338976,
+      "learning_rate": 1.3931039102353383e-08,
+      "loss": 0.6415,
+      "step": 14070
+    },
+    {
+      "epoch": 0.9770170809609776,
+      "grad_norm": 4.146574278201863,
+      "learning_rate": 1.3847282268006357e-08,
+      "loss": 0.3822,
+      "step": 14071
+    },
+    {
+      "epoch": 0.9770865157616998,
+      "grad_norm": 3.8973869437181103,
+      "learning_rate": 1.376377762762171e-08,
+      "loss": 0.3145,
+      "step": 14072
+    },
+    {
+      "epoch": 0.9771559505624219,
+      "grad_norm": 3.7636944714959606,
+      "learning_rate": 1.3680525185422177e-08,
+      "loss": 0.2118,
+      "step": 14073
+    },
+    {
+      "epoch": 0.977225385363144,
+      "grad_norm": 4.758499981745516,
+      "learning_rate": 1.3597524945619945e-08,
+      "loss": 0.4027,
+      "step": 14074
+    },
+    {
+      "epoch": 0.9772948201638662,
+      "grad_norm": 3.0733841253417107,
+      "learning_rate": 1.3514776912412208e-08,
+      "loss": 0.2956,
+      "step": 14075
+    },
+    {
+      "epoch": 0.9773642549645882,
+      "grad_norm": 4.140393052224029,
+      "learning_rate": 1.3432281089983956e-08,
+      "loss": 0.4269,
+      "step": 14076
+    },
+    {
+      "epoch": 0.9774336897653104,
+      "grad_norm": 3.5635142593657636,
+      "learning_rate": 1.335003748250907e-08,
+      "loss": 0.3161,
+      "step": 14077
+    },
+    {
+      "epoch": 0.9775031245660325,
+      "grad_norm": 2.8875187195242598,
+      "learning_rate": 1.3268046094145892e-08,
+      "loss": 0.2647,
+      "step": 14078
+    },
+    {
+      "epoch": 0.9775725593667546,
+      "grad_norm": 4.12734233227536,
+      "learning_rate": 1.3186306929042214e-08,
+      "loss": 0.4996,
+      "step": 14079
+    },
+    {
+      "epoch": 0.9776419941674768,
+      "grad_norm": 3.6233855783744864,
+      "learning_rate": 1.3104819991332506e-08,
+      "loss": 0.4636,
+      "step": 14080
+    },
+    {
+      "epoch": 0.9777114289681988,
+      "grad_norm": 4.7655919438223515,
+      "learning_rate": 1.3023585285137918e-08,
+      "loss": 0.4999,
+      "step": 14081
+    },
+    {
+      "epoch": 0.977780863768921,
+      "grad_norm": 4.338123120796176,
+      "learning_rate": 1.2942602814566829e-08,
+      "loss": 0.3928,
+      "step": 14082
+    },
+    {
+      "epoch": 0.9778502985696431,
+      "grad_norm": 3.474490758938959,
+      "learning_rate": 1.2861872583715962e-08,
+      "loss": 0.3457,
+      "step": 14083
+    },
+    {
+      "epoch": 0.9779197333703652,
+      "grad_norm": 4.189072065583572,
+      "learning_rate": 1.2781394596668162e-08,
+      "loss": 0.5384,
+      "step": 14084
+    },
+    {
+      "epoch": 0.9779891681710874,
+      "grad_norm": 10.838476821743498,
+      "learning_rate": 1.2701168857494062e-08,
+      "loss": 0.3271,
+      "step": 14085
+    },
+    {
+      "epoch": 0.9780586029718095,
+      "grad_norm": 4.401879839841229,
+      "learning_rate": 1.262119537025097e-08,
+      "loss": 0.3156,
+      "step": 14086
+    },
+    {
+      "epoch": 0.9781280377725315,
+      "grad_norm": 3.7527818008339753,
+      "learning_rate": 1.2541474138985099e-08,
+      "loss": 0.3326,
+      "step": 14087
+    },
+    {
+      "epoch": 0.9781974725732537,
+      "grad_norm": 3.366481565929875,
+      "learning_rate": 1.2462005167726553e-08,
+      "loss": 0.3272,
+      "step": 14088
+    },
+    {
+      "epoch": 0.9782669073739758,
+      "grad_norm": 3.968869740589805,
+      "learning_rate": 1.238278846049712e-08,
+      "loss": 0.5455,
+      "step": 14089
+    },
+    {
+      "epoch": 0.978336342174698,
+      "grad_norm": 4.4403806593110975,
+      "learning_rate": 1.2303824021301925e-08,
+      "loss": 0.3175,
+      "step": 14090
+    },
+    {
+      "epoch": 0.9784057769754201,
+      "grad_norm": 4.492118851394086,
+      "learning_rate": 1.2225111854134997e-08,
+      "loss": 0.547,
+      "step": 14091
+    },
+    {
+      "epoch": 0.9784752117761422,
+      "grad_norm": 4.1103904018939055,
+      "learning_rate": 1.2146651962978151e-08,
+      "loss": 0.42,
+      "step": 14092
+    },
+    {
+      "epoch": 0.9785446465768644,
+      "grad_norm": 4.096429101482497,
+      "learning_rate": 1.206844435179988e-08,
+      "loss": 0.4513,
+      "step": 14093
+    },
+    {
+      "epoch": 0.9786140813775864,
+      "grad_norm": 4.000859251593539,
+      "learning_rate": 1.1990489024554797e-08,
+      "loss": 0.4535,
+      "step": 14094
+    },
+    {
+      "epoch": 0.9786835161783086,
+      "grad_norm": 3.9083888157693965,
+      "learning_rate": 1.191278598518697e-08,
+      "loss": 0.3842,
+      "step": 14095
+    },
+    {
+      "epoch": 0.9787529509790307,
+      "grad_norm": 3.001478569978702,
+      "learning_rate": 1.1835335237626033e-08,
+      "loss": 0.2125,
+      "step": 14096
+    },
+    {
+      "epoch": 0.9788223857797528,
+      "grad_norm": 3.6682027844070113,
+      "learning_rate": 1.1758136785788854e-08,
+      "loss": 0.3298,
+      "step": 14097
+    },
+    {
+      "epoch": 0.978891820580475,
+      "grad_norm": 4.57049687022491,
+      "learning_rate": 1.1681190633581197e-08,
+      "loss": 0.5791,
+      "step": 14098
+    },
+    {
+      "epoch": 0.978961255381197,
+      "grad_norm": 3.7324466620370873,
+      "learning_rate": 1.1604496784893838e-08,
+      "loss": 0.3683,
+      "step": 14099
+    },
+    {
+      "epoch": 0.9790306901819191,
+      "grad_norm": 4.135751380637167,
+      "learning_rate": 1.1528055243605896e-08,
+      "loss": 0.5112,
+      "step": 14100
+    },
+    {
+      "epoch": 0.9791001249826413,
+      "grad_norm": 4.929288847928859,
+      "learning_rate": 1.1451866013584834e-08,
+      "loss": 0.7065,
+      "step": 14101
+    },
+    {
+      "epoch": 0.9791695597833634,
+      "grad_norm": 3.628106871153336,
+      "learning_rate": 1.1375929098683124e-08,
+      "loss": 0.425,
+      "step": 14102
+    },
+    {
+      "epoch": 0.9792389945840856,
+      "grad_norm": 5.046142894497169,
+      "learning_rate": 1.1300244502742142e-08,
+      "loss": 0.6785,
+      "step": 14103
+    },
+    {
+      "epoch": 0.9793084293848077,
+      "grad_norm": 4.1546328291859576,
+      "learning_rate": 1.1224812229589376e-08,
+      "loss": 0.4606,
+      "step": 14104
+    },
+    {
+      "epoch": 0.9793778641855297,
+      "grad_norm": 3.470277117568092,
+      "learning_rate": 1.1149632283040667e-08,
+      "loss": 0.342,
+      "step": 14105
+    },
+    {
+      "epoch": 0.9794472989862519,
+      "grad_norm": 4.589791903990124,
+      "learning_rate": 1.1074704666897972e-08,
+      "loss": 0.5509,
+      "step": 14106
+    },
+    {
+      "epoch": 0.979516733786974,
+      "grad_norm": 3.882001590830077,
+      "learning_rate": 1.1000029384951593e-08,
+      "loss": 0.4499,
+      "step": 14107
+    },
+    {
+      "epoch": 0.9795861685876962,
+      "grad_norm": 3.8949986906195138,
+      "learning_rate": 1.092560644097851e-08,
+      "loss": 0.3785,
+      "step": 14108
+    },
+    {
+      "epoch": 0.9796556033884183,
+      "grad_norm": 3.8848782925952796,
+      "learning_rate": 1.0851435838742374e-08,
+      "loss": 0.2903,
+      "step": 14109
+    },
+    {
+      "epoch": 0.9797250381891404,
+      "grad_norm": 2.693951429835676,
+      "learning_rate": 1.0777517581995189e-08,
+      "loss": 0.164,
+      "step": 14110
+    },
+    {
+      "epoch": 0.9797944729898626,
+      "grad_norm": 2.790320800779656,
+      "learning_rate": 1.0703851674475074e-08,
+      "loss": 0.259,
+      "step": 14111
+    },
+    {
+      "epoch": 0.9798639077905846,
+      "grad_norm": 5.045466073428521,
+      "learning_rate": 1.0630438119909048e-08,
+      "loss": 0.4457,
+      "step": 14112
+    },
+    {
+      "epoch": 0.9799333425913067,
+      "grad_norm": 3.918648184439637,
+      "learning_rate": 1.055727692200914e-08,
+      "loss": 0.4784,
+      "step": 14113
+    },
+    {
+      "epoch": 0.9800027773920289,
+      "grad_norm": 4.769809544447168,
+      "learning_rate": 1.0484368084476281e-08,
+      "loss": 0.376,
+      "step": 14114
+    },
+    {
+      "epoch": 0.980072212192751,
+      "grad_norm": 3.871387398081835,
+      "learning_rate": 1.0411711610998632e-08,
+      "loss": 0.326,
+      "step": 14115
+    },
+    {
+      "epoch": 0.9801416469934732,
+      "grad_norm": 5.198775243204175,
+      "learning_rate": 1.0339307505249918e-08,
+      "loss": 0.6159,
+      "step": 14116
+    },
+    {
+      "epoch": 0.9802110817941952,
+      "grad_norm": 4.189067507587056,
+      "learning_rate": 1.0267155770892767e-08,
+      "loss": 0.6066,
+      "step": 14117
+    },
+    {
+      "epoch": 0.9802805165949173,
+      "grad_norm": 5.505401133345937,
+      "learning_rate": 1.0195256411577037e-08,
+      "loss": 0.535,
+      "step": 14118
+    },
+    {
+      "epoch": 0.9803499513956395,
+      "grad_norm": 5.734626077939506,
+      "learning_rate": 1.0123609430938709e-08,
+      "loss": 0.7098,
+      "step": 14119
+    },
+    {
+      "epoch": 0.9804193861963616,
+      "grad_norm": 3.9781117663079337,
+      "learning_rate": 1.0052214832601547e-08,
+      "loss": 0.3901,
+      "step": 14120
+    },
+    {
+      "epoch": 0.9804888209970838,
+      "grad_norm": 3.367742073536449,
+      "learning_rate": 9.981072620177112e-09,
+      "loss": 0.1947,
+      "step": 14121
+    },
+    {
+      "epoch": 0.9805582557978059,
+      "grad_norm": 4.082077112135187,
+      "learning_rate": 9.910182797263079e-09,
+      "loss": 0.3973,
+      "step": 14122
+    },
+    {
+      "epoch": 0.980627690598528,
+      "grad_norm": 6.02665435647441,
+      "learning_rate": 9.839545367446024e-09,
+      "loss": 0.4334,
+      "step": 14123
+    },
+    {
+      "epoch": 0.9806971253992501,
+      "grad_norm": 3.826928592624585,
+      "learning_rate": 9.769160334297533e-09,
+      "loss": 0.5478,
+      "step": 14124
+    },
+    {
+      "epoch": 0.9807665601999722,
+      "grad_norm": 4.3772798889108495,
+      "learning_rate": 9.699027701378095e-09,
+      "loss": 0.4715,
+      "step": 14125
+    },
+    {
+      "epoch": 0.9808359950006943,
+      "grad_norm": 4.607695835737244,
+      "learning_rate": 9.629147472234868e-09,
+      "loss": 0.3438,
+      "step": 14126
+    },
+    {
+      "epoch": 0.9809054298014165,
+      "grad_norm": 4.212014137988611,
+      "learning_rate": 9.559519650402805e-09,
+      "loss": 0.3875,
+      "step": 14127
+    },
+    {
+      "epoch": 0.9809748646021386,
+      "grad_norm": 4.754714971375352,
+      "learning_rate": 9.490144239402976e-09,
+      "loss": 0.5744,
+      "step": 14128
+    },
+    {
+      "epoch": 0.9810442994028608,
+      "grad_norm": 4.943645687968767,
+      "learning_rate": 9.421021242744244e-09,
+      "loss": 0.4683,
+      "step": 14129
+    },
+    {
+      "epoch": 0.9811137342035828,
+      "grad_norm": 4.671313135898658,
+      "learning_rate": 9.352150663923254e-09,
+      "loss": 0.7006,
+      "step": 14130
+    },
+    {
+      "epoch": 0.9811831690043049,
+      "grad_norm": 3.6089221112469816,
+      "learning_rate": 9.283532506422777e-09,
+      "loss": 0.3423,
+      "step": 14131
+    },
+    {
+      "epoch": 0.9812526038050271,
+      "grad_norm": 4.696693145201252,
+      "learning_rate": 9.215166773714479e-09,
+      "loss": 0.5602,
+      "step": 14132
+    },
+    {
+      "epoch": 0.9813220386057492,
+      "grad_norm": 8.366195893118412,
+      "learning_rate": 9.147053469255041e-09,
+      "loss": 0.6616,
+      "step": 14133
+    },
+    {
+      "epoch": 0.9813914734064714,
+      "grad_norm": 4.371286666423065,
+      "learning_rate": 9.079192596490594e-09,
+      "loss": 0.4674,
+      "step": 14134
+    },
+    {
+      "epoch": 0.9814609082071934,
+      "grad_norm": 2.5539821177237454,
+      "learning_rate": 9.011584158852838e-09,
+      "loss": 0.2391,
+      "step": 14135
+    },
+    {
+      "epoch": 0.9815303430079155,
+      "grad_norm": 2.4957108760730704,
+      "learning_rate": 8.944228159761259e-09,
+      "loss": 0.1908,
+      "step": 14136
+    },
+    {
+      "epoch": 0.9815997778086377,
+      "grad_norm": 5.590215268791873,
+      "learning_rate": 8.877124602623133e-09,
+      "loss": 0.4717,
+      "step": 14137
+    },
+    {
+      "epoch": 0.9816692126093598,
+      "grad_norm": 4.554911578679251,
+      "learning_rate": 8.810273490831855e-09,
+      "loss": 0.4343,
+      "step": 14138
+    },
+    {
+      "epoch": 0.981738647410082,
+      "grad_norm": 3.609519446723891,
+      "learning_rate": 8.743674827769166e-09,
+      "loss": 0.3441,
+      "step": 14139
+    },
+    {
+      "epoch": 0.9818080822108041,
+      "grad_norm": 4.682400225371072,
+      "learning_rate": 8.677328616803482e-09,
+      "loss": 0.5968,
+      "step": 14140
+    },
+    {
+      "epoch": 0.9818775170115261,
+      "grad_norm": 5.29781204664176,
+      "learning_rate": 8.611234861290451e-09,
+      "loss": 0.559,
+      "step": 14141
+    },
+    {
+      "epoch": 0.9819469518122483,
+      "grad_norm": 4.065611354346616,
+      "learning_rate": 8.545393564572957e-09,
+      "loss": 0.4812,
+      "step": 14142
+    },
+    {
+      "epoch": 0.9820163866129704,
+      "grad_norm": 5.389734326414243,
+      "learning_rate": 8.479804729981667e-09,
+      "loss": 0.659,
+      "step": 14143
+    },
+    {
+      "epoch": 0.9820858214136925,
+      "grad_norm": 3.01990955899393,
+      "learning_rate": 8.414468360832817e-09,
+      "loss": 0.3784,
+      "step": 14144
+    },
+    {
+      "epoch": 0.9821552562144147,
+      "grad_norm": 4.176493850689642,
+      "learning_rate": 8.349384460432652e-09,
+      "loss": 0.4268,
+      "step": 14145
+    },
+    {
+      "epoch": 0.9822246910151368,
+      "grad_norm": 4.074199684074437,
+      "learning_rate": 8.284553032071319e-09,
+      "loss": 0.4792,
+      "step": 14146
+    },
+    {
+      "epoch": 0.982294125815859,
+      "grad_norm": 4.131093091531945,
+      "learning_rate": 8.219974079029524e-09,
+      "loss": 0.5293,
+      "step": 14147
+    },
+    {
+      "epoch": 0.982363560616581,
+      "grad_norm": 3.8749695924108445,
+      "learning_rate": 8.155647604572437e-09,
+      "loss": 0.4293,
+      "step": 14148
+    },
+    {
+      "epoch": 0.9824329954173031,
+      "grad_norm": 3.8739585367146914,
+      "learning_rate": 8.091573611954118e-09,
+      "loss": 0.3286,
+      "step": 14149
+    },
+    {
+      "epoch": 0.9825024302180253,
+      "grad_norm": 7.368468135141709,
+      "learning_rate": 8.02775210441531e-09,
+      "loss": 0.3853,
+      "step": 14150
+    },
+    {
+      "epoch": 0.9825718650187474,
+      "grad_norm": 3.826584228956766,
+      "learning_rate": 7.964183085183986e-09,
+      "loss": 0.497,
+      "step": 14151
+    },
+    {
+      "epoch": 0.9826412998194696,
+      "grad_norm": 4.300948281869529,
+      "learning_rate": 7.90086655747535e-09,
+      "loss": 0.2242,
+      "step": 14152
+    },
+    {
+      "epoch": 0.9827107346201917,
+      "grad_norm": 4.749027840000823,
+      "learning_rate": 7.837802524491844e-09,
+      "loss": 0.4977,
+      "step": 14153
+    },
+    {
+      "epoch": 0.9827801694209137,
+      "grad_norm": 5.04991920666717,
+      "learning_rate": 7.774990989423692e-09,
+      "loss": 0.5446,
+      "step": 14154
+    },
+    {
+      "epoch": 0.9828496042216359,
+      "grad_norm": 3.7447001871968,
+      "learning_rate": 7.712431955447242e-09,
+      "loss": 0.3505,
+      "step": 14155
+    },
+    {
+      "epoch": 0.982919039022358,
+      "grad_norm": 4.657464663673959,
+      "learning_rate": 7.650125425726628e-09,
+      "loss": 0.562,
+      "step": 14156
+    },
+    {
+      "epoch": 0.9829884738230801,
+      "grad_norm": 2.992367973552169,
+      "learning_rate": 7.588071403413221e-09,
+      "loss": 0.353,
+      "step": 14157
+    },
+    {
+      "epoch": 0.9830579086238023,
+      "grad_norm": 3.825488736538764,
+      "learning_rate": 7.526269891646176e-09,
+      "loss": 0.4779,
+      "step": 14158
+    },
+    {
+      "epoch": 0.9831273434245243,
+      "grad_norm": 3.597152256526084,
+      "learning_rate": 7.464720893551324e-09,
+      "loss": 0.372,
+      "step": 14159
+    },
+    {
+      "epoch": 0.9831967782252465,
+      "grad_norm": 4.876675183059779,
+      "learning_rate": 7.403424412241178e-09,
+      "loss": 0.5298,
+      "step": 14160
+    },
+    {
+      "epoch": 0.9832662130259686,
+      "grad_norm": 3.755935809743802,
+      "learning_rate": 7.342380450816034e-09,
+      "loss": 0.3097,
+      "step": 14161
+    },
+    {
+      "epoch": 0.9833356478266907,
+      "grad_norm": 4.376632789534533,
+      "learning_rate": 7.281589012363976e-09,
+      "loss": 0.4964,
+      "step": 14162
+    },
+    {
+      "epoch": 0.9834050826274129,
+      "grad_norm": 3.912126786450365,
+      "learning_rate": 7.22105009995977e-09,
+      "loss": 0.3354,
+      "step": 14163
+    },
+    {
+      "epoch": 0.983474517428135,
+      "grad_norm": 3.6394292092041023,
+      "learning_rate": 7.160763716664853e-09,
+      "loss": 0.3885,
+      "step": 14164
+    },
+    {
+      "epoch": 0.9835439522288572,
+      "grad_norm": 2.919243442843382,
+      "learning_rate": 7.10072986552901e-09,
+      "loss": 0.1903,
+      "step": 14165
+    },
+    {
+      "epoch": 0.9836133870295792,
+      "grad_norm": 4.173854438218789,
+      "learning_rate": 7.0409485495881444e-09,
+      "loss": 0.431,
+      "step": 14166
+    },
+    {
+      "epoch": 0.9836828218303013,
+      "grad_norm": 4.686502651890082,
+      "learning_rate": 6.981419771865949e-09,
+      "loss": 0.671,
+      "step": 14167
+    },
+    {
+      "epoch": 0.9837522566310235,
+      "grad_norm": 4.120819515339244,
+      "learning_rate": 6.922143535373904e-09,
+      "loss": 0.454,
+      "step": 14168
+    },
+    {
+      "epoch": 0.9838216914317456,
+      "grad_norm": 4.331144582877351,
+      "learning_rate": 6.863119843109611e-09,
+      "loss": 0.4561,
+      "step": 14169
+    },
+    {
+      "epoch": 0.9838911262324677,
+      "grad_norm": 5.596195732068436,
+      "learning_rate": 6.80434869805846e-09,
+      "loss": 0.3442,
+      "step": 14170
+    },
+    {
+      "epoch": 0.9839605610331899,
+      "grad_norm": 4.89207446982786,
+      "learning_rate": 6.7458301031936295e-09,
+      "loss": 0.5232,
+      "step": 14171
+    },
+    {
+      "epoch": 0.9840299958339119,
+      "grad_norm": 3.4273103989318034,
+      "learning_rate": 6.687564061474416e-09,
+      "loss": 0.3381,
+      "step": 14172
+    },
+    {
+      "epoch": 0.9840994306346341,
+      "grad_norm": 3.229415934682325,
+      "learning_rate": 6.629550575847355e-09,
+      "loss": 0.2252,
+      "step": 14173
+    },
+    {
+      "epoch": 0.9841688654353562,
+      "grad_norm": 3.1772466200707856,
+      "learning_rate": 6.571789649247873e-09,
+      "loss": 0.2441,
+      "step": 14174
+    },
+    {
+      "epoch": 0.9842383002360783,
+      "grad_norm": 3.5274997732370528,
+      "learning_rate": 6.514281284596413e-09,
+      "loss": 0.3999,
+      "step": 14175
+    },
+    {
+      "epoch": 0.9843077350368005,
+      "grad_norm": 4.4079541734597845,
+      "learning_rate": 6.457025484802315e-09,
+      "loss": 0.3907,
+      "step": 14176
+    },
+    {
+      "epoch": 0.9843771698375225,
+      "grad_norm": 4.122472300530622,
+      "learning_rate": 6.40002225276104e-09,
+      "loss": 0.5338,
+      "step": 14177
+    },
+    {
+      "epoch": 0.9844466046382447,
+      "grad_norm": 3.589075220721958,
+      "learning_rate": 6.3432715913563926e-09,
+      "loss": 0.3137,
+      "step": 14178
+    },
+    {
+      "epoch": 0.9845160394389668,
+      "grad_norm": 4.209194871194281,
+      "learning_rate": 6.286773503458299e-09,
+      "loss": 0.34,
+      "step": 14179
+    },
+    {
+      "epoch": 0.9845854742396889,
+      "grad_norm": 4.317748281712373,
+      "learning_rate": 6.230527991923918e-09,
+      "loss": 0.4693,
+      "step": 14180
+    },
+    {
+      "epoch": 0.9846549090404111,
+      "grad_norm": 4.504707847332406,
+      "learning_rate": 6.174535059598752e-09,
+      "loss": 0.4781,
+      "step": 14181
+    },
+    {
+      "epoch": 0.9847243438411332,
+      "grad_norm": 4.086500518127666,
+      "learning_rate": 6.118794709314979e-09,
+      "loss": 0.3124,
+      "step": 14182
+    },
+    {
+      "epoch": 0.9847937786418552,
+      "grad_norm": 4.488603063683037,
+      "learning_rate": 6.063306943891456e-09,
+      "loss": 0.5561,
+      "step": 14183
+    },
+    {
+      "epoch": 0.9848632134425774,
+      "grad_norm": 3.57468711045499,
+      "learning_rate": 6.008071766135382e-09,
+      "loss": 0.4904,
+      "step": 14184
+    },
+    {
+      "epoch": 0.9849326482432995,
+      "grad_norm": 5.985968591119745,
+      "learning_rate": 5.953089178839522e-09,
+      "loss": 0.3403,
+      "step": 14185
+    },
+    {
+      "epoch": 0.9850020830440217,
+      "grad_norm": 3.71301667238091,
+      "learning_rate": 5.898359184784985e-09,
+      "loss": 0.518,
+      "step": 14186
+    },
+    {
+      "epoch": 0.9850715178447438,
+      "grad_norm": 4.557541910643263,
+      "learning_rate": 5.843881786740668e-09,
+      "loss": 0.3161,
+      "step": 14187
+    },
+    {
+      "epoch": 0.9851409526454659,
+      "grad_norm": 3.526151989708653,
+      "learning_rate": 5.7896569874610345e-09,
+      "loss": 0.3959,
+      "step": 14188
+    },
+    {
+      "epoch": 0.985210387446188,
+      "grad_norm": 2.8416666859683675,
+      "learning_rate": 5.735684789689444e-09,
+      "loss": 0.2019,
+      "step": 14189
+    },
+    {
+      "epoch": 0.9852798222469101,
+      "grad_norm": 3.4584920144762927,
+      "learning_rate": 5.681965196155937e-09,
+      "loss": 0.3093,
+      "step": 14190
+    },
+    {
+      "epoch": 0.9853492570476323,
+      "grad_norm": 4.14636616090418,
+      "learning_rate": 5.628498209576672e-09,
+      "loss": 0.4154,
+      "step": 14191
+    },
+    {
+      "epoch": 0.9854186918483544,
+      "grad_norm": 3.9049732087820788,
+      "learning_rate": 5.575283832656708e-09,
+      "loss": 0.3858,
+      "step": 14192
+    },
+    {
+      "epoch": 0.9854881266490765,
+      "grad_norm": 3.8332280371123515,
+      "learning_rate": 5.522322068087227e-09,
+      "loss": 0.2814,
+      "step": 14193
+    },
+    {
+      "epoch": 0.9855575614497987,
+      "grad_norm": 5.171097059368011,
+      "learning_rate": 5.469612918547196e-09,
+      "loss": 0.352,
+      "step": 14194
+    },
+    {
+      "epoch": 0.9856269962505207,
+      "grad_norm": 3.3162668708479477,
+      "learning_rate": 5.41715638670226e-09,
+      "loss": 0.2855,
+      "step": 14195
+    },
+    {
+      "epoch": 0.9856964310512429,
+      "grad_norm": 3.629310011488377,
+      "learning_rate": 5.3649524752058535e-09,
+      "loss": 0.2959,
+      "step": 14196
+    },
+    {
+      "epoch": 0.985765865851965,
+      "grad_norm": 5.003806680354501,
+      "learning_rate": 5.313001186698641e-09,
+      "loss": 0.3164,
+      "step": 14197
+    },
+    {
+      "epoch": 0.9858353006526871,
+      "grad_norm": 4.344733664439827,
+      "learning_rate": 5.26130252380741e-09,
+      "loss": 0.4822,
+      "step": 14198
+    },
+    {
+      "epoch": 0.9859047354534093,
+      "grad_norm": 3.864044038068947,
+      "learning_rate": 5.209856489147847e-09,
+      "loss": 0.4809,
+      "step": 14199
+    },
+    {
+      "epoch": 0.9859741702541314,
+      "grad_norm": 3.149019775061917,
+      "learning_rate": 5.158663085321758e-09,
+      "loss": 0.3169,
+      "step": 14200
+    },
+    {
+      "epoch": 0.9860436050548534,
+      "grad_norm": 4.4814503851221055,
+      "learning_rate": 5.107722314918184e-09,
+      "loss": 0.5874,
+      "step": 14201
+    },
+    {
+      "epoch": 0.9861130398555756,
+      "grad_norm": 3.0312936100976198,
+      "learning_rate": 5.057034180514509e-09,
+      "loss": 0.3351,
+      "step": 14202
+    },
+    {
+      "epoch": 0.9861824746562977,
+      "grad_norm": 2.906331784651308,
+      "learning_rate": 5.00659868467368e-09,
+      "loss": 0.3824,
+      "step": 14203
+    },
+    {
+      "epoch": 0.9862519094570199,
+      "grad_norm": 3.7906194678540004,
+      "learning_rate": 4.956415829946437e-09,
+      "loss": 0.312,
+      "step": 14204
+    },
+    {
+      "epoch": 0.986321344257742,
+      "grad_norm": 4.8510019580673704,
+      "learning_rate": 4.906485618871859e-09,
+      "loss": 0.6608,
+      "step": 14205
+    },
+    {
+      "epoch": 0.9863907790584641,
+      "grad_norm": 3.7813065156221333,
+      "learning_rate": 4.856808053974593e-09,
+      "loss": 0.3299,
+      "step": 14206
+    },
+    {
+      "epoch": 0.9864602138591863,
+      "grad_norm": 4.133393283046333,
+      "learning_rate": 4.807383137767629e-09,
+      "loss": 0.3333,
+      "step": 14207
+    },
+    {
+      "epoch": 0.9865296486599083,
+      "grad_norm": 4.143413824971348,
+      "learning_rate": 4.75821087275119e-09,
+      "loss": 0.3314,
+      "step": 14208
+    },
+    {
+      "epoch": 0.9865990834606305,
+      "grad_norm": 3.924968519938344,
+      "learning_rate": 4.7092912614116195e-09,
+      "loss": 0.3509,
+      "step": 14209
+    },
+    {
+      "epoch": 0.9866685182613526,
+      "grad_norm": 4.458364063164606,
+      "learning_rate": 4.660624306223605e-09,
+      "loss": 0.3226,
+      "step": 14210
+    },
+    {
+      "epoch": 0.9867379530620747,
+      "grad_norm": 3.1188221630282777,
+      "learning_rate": 4.61221000964851e-09,
+      "loss": 0.335,
+      "step": 14211
+    },
+    {
+      "epoch": 0.9868073878627969,
+      "grad_norm": 2.6807442275385607,
+      "learning_rate": 4.5640483741349325e-09,
+      "loss": 0.2218,
+      "step": 14212
+    },
+    {
+      "epoch": 0.986876822663519,
+      "grad_norm": 4.610058498184213,
+      "learning_rate": 4.516139402119813e-09,
+      "loss": 0.6192,
+      "step": 14213
+    },
+    {
+      "epoch": 0.986946257464241,
+      "grad_norm": 4.009593942032591,
+      "learning_rate": 4.4684830960251e-09,
+      "loss": 0.5193,
+      "step": 14214
+    },
+    {
+      "epoch": 0.9870156922649632,
+      "grad_norm": 3.680663467870674,
+      "learning_rate": 4.421079458261646e-09,
+      "loss": 0.3762,
+      "step": 14215
+    },
+    {
+      "epoch": 0.9870851270656853,
+      "grad_norm": 3.1512938151262007,
+      "learning_rate": 4.373928491226975e-09,
+      "loss": 0.235,
+      "step": 14216
+    },
+    {
+      "epoch": 0.9871545618664075,
+      "grad_norm": 3.2395615532317663,
+      "learning_rate": 4.327030197306403e-09,
+      "loss": 0.3058,
+      "step": 14217
+    },
+    {
+      "epoch": 0.9872239966671296,
+      "grad_norm": 4.980506581599429,
+      "learning_rate": 4.2803845788713664e-09,
+      "loss": 0.4203,
+      "step": 14218
+    },
+    {
+      "epoch": 0.9872934314678516,
+      "grad_norm": 3.744771345696337,
+      "learning_rate": 4.233991638281642e-09,
+      "loss": 0.4522,
+      "step": 14219
+    },
+    {
+      "epoch": 0.9873628662685738,
+      "grad_norm": 4.119476329631578,
+      "learning_rate": 4.187851377883689e-09,
+      "loss": 0.4914,
+      "step": 14220
+    },
+    {
+      "epoch": 0.9874323010692959,
+      "grad_norm": 3.317723696709438,
+      "learning_rate": 4.14196380001064e-09,
+      "loss": 0.2931,
+      "step": 14221
+    },
+    {
+      "epoch": 0.9875017358700181,
+      "grad_norm": 5.218260272938277,
+      "learning_rate": 4.096328906984526e-09,
+      "loss": 0.585,
+      "step": 14222
+    },
+    {
+      "epoch": 0.9875711706707402,
+      "grad_norm": 3.230147301402645,
+      "learning_rate": 4.0509467011123906e-09,
+      "loss": 0.3427,
+      "step": 14223
+    },
+    {
+      "epoch": 0.9876406054714623,
+      "grad_norm": 3.109538729330601,
+      "learning_rate": 4.005817184690175e-09,
+      "loss": 0.3414,
+      "step": 14224
+    },
+    {
+      "epoch": 0.9877100402721845,
+      "grad_norm": 3.7601146698927734,
+      "learning_rate": 3.960940359999943e-09,
+      "loss": 0.3362,
+      "step": 14225
+    },
+    {
+      "epoch": 0.9877794750729065,
+      "grad_norm": 4.439727874924048,
+      "learning_rate": 3.9163162293126554e-09,
+      "loss": 0.2895,
+      "step": 14226
+    },
+    {
+      "epoch": 0.9878489098736286,
+      "grad_norm": 2.508634621559771,
+      "learning_rate": 3.871944794884286e-09,
+      "loss": 0.1936,
+      "step": 14227
+    },
+    {
+      "epoch": 0.9879183446743508,
+      "grad_norm": 4.20587110479154,
+      "learning_rate": 3.82782605895915e-09,
+      "loss": 0.2941,
+      "step": 14228
+    },
+    {
+      "epoch": 0.9879877794750729,
+      "grad_norm": 4.122638343125635,
+      "learning_rate": 3.783960023769906e-09,
+      "loss": 0.538,
+      "step": 14229
+    },
+    {
+      "epoch": 0.9880572142757951,
+      "grad_norm": 4.213129670831205,
+      "learning_rate": 3.7403466915336695e-09,
+      "loss": 0.4974,
+      "step": 14230
+    },
+    {
+      "epoch": 0.9881266490765171,
+      "grad_norm": 3.754216109402717,
+      "learning_rate": 3.6969860644570087e-09,
+      "loss": 0.3353,
+      "step": 14231
+    },
+    {
+      "epoch": 0.9881960838772392,
+      "grad_norm": 4.953921450927265,
+      "learning_rate": 3.6538781447331694e-09,
+      "loss": 0.5899,
+      "step": 14232
+    },
+    {
+      "epoch": 0.9882655186779614,
+      "grad_norm": 3.4467166992592237,
+      "learning_rate": 3.6110229345426297e-09,
+      "loss": 0.3421,
+      "step": 14233
+    },
+    {
+      "epoch": 0.9883349534786835,
+      "grad_norm": 3.506998569487072,
+      "learning_rate": 3.5684204360525443e-09,
+      "loss": 0.4057,
+      "step": 14234
+    },
+    {
+      "epoch": 0.9884043882794057,
+      "grad_norm": 3.0907568675238353,
+      "learning_rate": 3.526070651417857e-09,
+      "loss": 0.2455,
+      "step": 14235
+    },
+    {
+      "epoch": 0.9884738230801278,
+      "grad_norm": 3.362668306929536,
+      "learning_rate": 3.483973582780742e-09,
+      "loss": 0.2925,
+      "step": 14236
+    },
+    {
+      "epoch": 0.9885432578808498,
+      "grad_norm": 3.4578164349244984,
+      "learning_rate": 3.442129232270608e-09,
+      "loss": 0.2931,
+      "step": 14237
+    },
+    {
+      "epoch": 0.988612692681572,
+      "grad_norm": 3.85679652992569,
+      "learning_rate": 3.4005376020029848e-09,
+      "loss": 0.4281,
+      "step": 14238
+    },
+    {
+      "epoch": 0.9886821274822941,
+      "grad_norm": 3.0586036118853728,
+      "learning_rate": 3.3591986940828546e-09,
+      "loss": 0.2463,
+      "step": 14239
+    },
+    {
+      "epoch": 0.9887515622830162,
+      "grad_norm": 3.803732510223765,
+      "learning_rate": 3.318112510599658e-09,
+      "loss": 0.2636,
+      "step": 14240
+    },
+    {
+      "epoch": 0.9888209970837384,
+      "grad_norm": 2.8853049483691926,
+      "learning_rate": 3.2772790536328426e-09,
+      "loss": 0.1564,
+      "step": 14241
+    },
+    {
+      "epoch": 0.9888904318844605,
+      "grad_norm": 3.7314160271238386,
+      "learning_rate": 3.2366983252463126e-09,
+      "loss": 0.3955,
+      "step": 14242
+    },
+    {
+      "epoch": 0.9889598666851827,
+      "grad_norm": 4.545927575892611,
+      "learning_rate": 3.1963703274939805e-09,
+      "loss": 0.3805,
+      "step": 14243
+    },
+    {
+      "epoch": 0.9890293014859047,
+      "grad_norm": 4.248724984063371,
+      "learning_rate": 3.1562950624147713e-09,
+      "loss": 0.5861,
+      "step": 14244
+    },
+    {
+      "epoch": 0.9890987362866268,
+      "grad_norm": 4.253931839935822,
+      "learning_rate": 3.1164725320359525e-09,
+      "loss": 0.5171,
+      "step": 14245
+    },
+    {
+      "epoch": 0.989168171087349,
+      "grad_norm": 5.316945959766639,
+      "learning_rate": 3.0769027383714677e-09,
+      "loss": 0.4917,
+      "step": 14246
+    },
+    {
+      "epoch": 0.9892376058880711,
+      "grad_norm": 4.058011368721931,
+      "learning_rate": 3.03758568342305e-09,
+      "loss": 0.5262,
+      "step": 14247
+    },
+    {
+      "epoch": 0.9893070406887933,
+      "grad_norm": 3.4919473846590545,
+      "learning_rate": 2.9985213691791082e-09,
+      "loss": 0.2666,
+      "step": 14248
+    },
+    {
+      "epoch": 0.9893764754895154,
+      "grad_norm": 3.318359125052282,
+      "learning_rate": 2.9597097976152845e-09,
+      "loss": 0.3349,
+      "step": 14249
+    },
+    {
+      "epoch": 0.9894459102902374,
+      "grad_norm": 3.1524749485761334,
+      "learning_rate": 2.921150970694453e-09,
+      "loss": 0.2583,
+      "step": 14250
+    },
+    {
+      "epoch": 0.9895153450909596,
+      "grad_norm": 5.137345018754172,
+      "learning_rate": 2.8828448903678307e-09,
+      "loss": 0.4136,
+      "step": 14251
+    },
+    {
+      "epoch": 0.9895847798916817,
+      "grad_norm": 4.78156723056081,
+      "learning_rate": 2.844791558572202e-09,
+      "loss": 0.4089,
+      "step": 14252
+    },
+    {
+      "epoch": 0.9896542146924038,
+      "grad_norm": 3.8922416130471724,
+      "learning_rate": 2.806990977232138e-09,
+      "loss": 0.4361,
+      "step": 14253
+    },
+    {
+      "epoch": 0.989723649493126,
+      "grad_norm": 5.229318373526022,
+      "learning_rate": 2.769443148259443e-09,
+      "loss": 0.5801,
+      "step": 14254
+    },
+    {
+      "epoch": 0.989793084293848,
+      "grad_norm": 2.4235266948588223,
+      "learning_rate": 2.7321480735542638e-09,
+      "loss": 0.1813,
+      "step": 14255
+    },
+    {
+      "epoch": 0.9898625190945702,
+      "grad_norm": 2.8685022915012044,
+      "learning_rate": 2.695105755001759e-09,
+      "loss": 0.2734,
+      "step": 14256
+    },
+    {
+      "epoch": 0.9899319538952923,
+      "grad_norm": 3.2518068959560114,
+      "learning_rate": 2.658316194475985e-09,
+      "loss": 0.3173,
+      "step": 14257
+    },
+    {
+      "epoch": 0.9900013886960144,
+      "grad_norm": 3.1318949610475864,
+      "learning_rate": 2.621779393838231e-09,
+      "loss": 0.294,
+      "step": 14258
+    },
+    {
+      "epoch": 0.9900708234967366,
+      "grad_norm": 3.538063291172598,
+      "learning_rate": 2.5854953549353524e-09,
+      "loss": 0.3819,
+      "step": 14259
+    },
+    {
+      "epoch": 0.9901402582974587,
+      "grad_norm": 3.3415004386966904,
+      "learning_rate": 2.5494640796036585e-09,
+      "loss": 0.3756,
+      "step": 14260
+    },
+    {
+      "epoch": 0.9902096930981809,
+      "grad_norm": 3.5622495862132495,
+      "learning_rate": 2.51368556966447e-09,
+      "loss": 0.3242,
+      "step": 14261
+    },
+    {
+      "epoch": 0.9902791278989029,
+      "grad_norm": 3.7482023050587836,
+      "learning_rate": 2.4781598269285613e-09,
+      "loss": 0.4508,
+      "step": 14262
+    },
+    {
+      "epoch": 0.990348562699625,
+      "grad_norm": 3.9950671345953728,
+      "learning_rate": 2.4428868531922722e-09,
+      "loss": 0.321,
+      "step": 14263
+    },
+    {
+      "epoch": 0.9904179975003472,
+      "grad_norm": 3.2183729135146186,
+      "learning_rate": 2.407866650239177e-09,
+      "loss": 0.3491,
+      "step": 14264
+    },
+    {
+      "epoch": 0.9904874323010693,
+      "grad_norm": 1.6428671812482867,
+      "learning_rate": 2.3730992198411907e-09,
+      "loss": 0.1066,
+      "step": 14265
+    },
+    {
+      "epoch": 0.9905568671017915,
+      "grad_norm": 4.01500593309831,
+      "learning_rate": 2.338584563756907e-09,
+      "loss": 0.4169,
+      "step": 14266
+    },
+    {
+      "epoch": 0.9906263019025136,
+      "grad_norm": 4.395543049611679,
+      "learning_rate": 2.3043226837315967e-09,
+      "loss": 0.5331,
+      "step": 14267
+    },
+    {
+      "epoch": 0.9906957367032356,
+      "grad_norm": 4.7131437188914855,
+      "learning_rate": 2.2703135814983178e-09,
+      "loss": 0.6364,
+      "step": 14268
+    },
+    {
+      "epoch": 0.9907651715039578,
+      "grad_norm": 4.4089483284487025,
+      "learning_rate": 2.236557258777361e-09,
+      "loss": 0.5121,
+      "step": 14269
+    },
+    {
+      "epoch": 0.9908346063046799,
+      "grad_norm": 2.8187029126115704,
+      "learning_rate": 2.203053717276249e-09,
+      "loss": 0.1886,
+      "step": 14270
+    },
+    {
+      "epoch": 0.990904041105402,
+      "grad_norm": 3.68986898732914,
+      "learning_rate": 2.169802958688627e-09,
+      "loss": 0.2882,
+      "step": 14271
+    },
+    {
+      "epoch": 0.9909734759061242,
+      "grad_norm": 3.5070658445179195,
+      "learning_rate": 2.1368049846975936e-09,
+      "loss": 0.3263,
+      "step": 14272
+    },
+    {
+      "epoch": 0.9910429107068462,
+      "grad_norm": 3.4575929077687424,
+      "learning_rate": 2.1040597969712586e-09,
+      "loss": 0.1993,
+      "step": 14273
+    },
+    {
+      "epoch": 0.9911123455075684,
+      "grad_norm": 4.921026638718083,
+      "learning_rate": 2.0715673971660744e-09,
+      "loss": 0.7173,
+      "step": 14274
+    },
+    {
+      "epoch": 0.9911817803082905,
+      "grad_norm": 5.689504801078911,
+      "learning_rate": 2.0393277869251717e-09,
+      "loss": 0.59,
+      "step": 14275
+    },
+    {
+      "epoch": 0.9912512151090126,
+      "grad_norm": 3.233557456374596,
+      "learning_rate": 2.0073409678800227e-09,
+      "loss": 0.2097,
+      "step": 14276
+    },
+    {
+      "epoch": 0.9913206499097348,
+      "grad_norm": 4.570685483882463,
+      "learning_rate": 1.9756069416476677e-09,
+      "loss": 0.5052,
+      "step": 14277
+    },
+    {
+      "epoch": 0.9913900847104569,
+      "grad_norm": 3.970399761257439,
+      "learning_rate": 1.9441257098334886e-09,
+      "loss": 0.3548,
+      "step": 14278
+    },
+    {
+      "epoch": 0.991459519511179,
+      "grad_norm": 4.3469033650536195,
+      "learning_rate": 1.9128972740295458e-09,
+      "loss": 0.4343,
+      "step": 14279
+    },
+    {
+      "epoch": 0.9915289543119011,
+      "grad_norm": 3.274271664665922,
+      "learning_rate": 1.8819216358156865e-09,
+      "loss": 0.4066,
+      "step": 14280
+    },
+    {
+      "epoch": 0.9915983891126232,
+      "grad_norm": 3.621070949269692,
+      "learning_rate": 1.8511987967589906e-09,
+      "loss": 0.3817,
+      "step": 14281
+    },
+    {
+      "epoch": 0.9916678239133454,
+      "grad_norm": 3.8249458126028766,
+      "learning_rate": 1.8207287584121046e-09,
+      "loss": 0.3736,
+      "step": 14282
+    },
+    {
+      "epoch": 0.9917372587140675,
+      "grad_norm": 4.146319164937174,
+      "learning_rate": 1.7905115223171287e-09,
+      "loss": 0.6416,
+      "step": 14283
+    },
+    {
+      "epoch": 0.9918066935147896,
+      "grad_norm": 3.9439980721220382,
+      "learning_rate": 1.7605470900028399e-09,
+      "loss": 0.459,
+      "step": 14284
+    },
+    {
+      "epoch": 0.9918761283155118,
+      "grad_norm": 4.110899005641816,
+      "learning_rate": 1.7308354629835822e-09,
+      "loss": 0.3902,
+      "step": 14285
+    },
+    {
+      "epoch": 0.9919455631162338,
+      "grad_norm": 3.81131684874795,
+      "learning_rate": 1.7013766427625976e-09,
+      "loss": 0.4883,
+      "step": 14286
+    },
+    {
+      "epoch": 0.992014997916956,
+      "grad_norm": 4.327292220682511,
+      "learning_rate": 1.6721706308303609e-09,
+      "loss": 0.3972,
+      "step": 14287
+    },
+    {
+      "epoch": 0.9920844327176781,
+      "grad_norm": 4.619631209062738,
+      "learning_rate": 1.6432174286634684e-09,
+      "loss": 0.4488,
+      "step": 14288
+    },
+    {
+      "epoch": 0.9921538675184002,
+      "grad_norm": 3.716669698779563,
+      "learning_rate": 1.6145170377268593e-09,
+      "loss": 0.1797,
+      "step": 14289
+    },
+    {
+      "epoch": 0.9922233023191224,
+      "grad_norm": 2.8123973836239937,
+      "learning_rate": 1.5860694594721504e-09,
+      "loss": 0.2102,
+      "step": 14290
+    },
+    {
+      "epoch": 0.9922927371198444,
+      "grad_norm": 2.4821299991790053,
+      "learning_rate": 1.5578746953376355e-09,
+      "loss": 0.1956,
+      "step": 14291
+    },
+    {
+      "epoch": 0.9923621719205666,
+      "grad_norm": 2.9432849741502443,
+      "learning_rate": 1.5299327467499515e-09,
+      "loss": 0.1909,
+      "step": 14292
+    },
+    {
+      "epoch": 0.9924316067212887,
+      "grad_norm": 3.3279888053593556,
+      "learning_rate": 1.5022436151218566e-09,
+      "loss": 0.3511,
+      "step": 14293
+    },
+    {
+      "epoch": 0.9925010415220108,
+      "grad_norm": 4.1882467731649555,
+      "learning_rate": 1.4748073018538978e-09,
+      "loss": 0.4041,
+      "step": 14294
+    },
+    {
+      "epoch": 0.992570476322733,
+      "grad_norm": 4.999210832809188,
+      "learning_rate": 1.4476238083344086e-09,
+      "loss": 0.5945,
+      "step": 14295
+    },
+    {
+      "epoch": 0.9926399111234551,
+      "grad_norm": 3.676634550792359,
+      "learning_rate": 1.4206931359378452e-09,
+      "loss": 0.341,
+      "step": 14296
+    },
+    {
+      "epoch": 0.9927093459241771,
+      "grad_norm": 5.204906464761587,
+      "learning_rate": 1.394015286025896e-09,
+      "loss": 0.6376,
+      "step": 14297
+    },
+    {
+      "epoch": 0.9927787807248993,
+      "grad_norm": 2.9809916611844356,
+      "learning_rate": 1.3675902599485923e-09,
+      "loss": 0.2389,
+      "step": 14298
+    },
+    {
+      "epoch": 0.9928482155256214,
+      "grad_norm": 4.788205271186599,
+      "learning_rate": 1.3414180590426428e-09,
+      "loss": 0.4541,
+      "step": 14299
+    },
+    {
+      "epoch": 0.9929176503263436,
+      "grad_norm": 4.396971422959542,
+      "learning_rate": 1.3154986846308782e-09,
+      "loss": 0.4539,
+      "step": 14300
+    },
+    {
+      "epoch": 0.9929870851270657,
+      "grad_norm": 3.4993828652186894,
+      "learning_rate": 1.2898321380250268e-09,
+      "loss": 0.2373,
+      "step": 14301
+    },
+    {
+      "epoch": 0.9930565199277878,
+      "grad_norm": 4.241135310146371,
+      "learning_rate": 1.2644184205229393e-09,
+      "loss": 0.3309,
+      "step": 14302
+    },
+    {
+      "epoch": 0.99312595472851,
+      "grad_norm": 4.163670563864669,
+      "learning_rate": 1.2392575334102542e-09,
+      "loss": 0.3438,
+      "step": 14303
+    },
+    {
+      "epoch": 0.993195389529232,
+      "grad_norm": 4.928342356441496,
+      "learning_rate": 1.214349477959287e-09,
+      "loss": 0.4909,
+      "step": 14304
+    },
+    {
+      "epoch": 0.9932648243299542,
+      "grad_norm": 3.6313620635274764,
+      "learning_rate": 1.1896942554301404e-09,
+      "loss": 0.477,
+      "step": 14305
+    },
+    {
+      "epoch": 0.9933342591306763,
+      "grad_norm": 3.471134556966968,
+      "learning_rate": 1.1652918670695957e-09,
+      "loss": 0.3222,
+      "step": 14306
+    },
+    {
+      "epoch": 0.9934036939313984,
+      "grad_norm": 3.5291741071732075,
+      "learning_rate": 1.141142314111665e-09,
+      "loss": 0.4845,
+      "step": 14307
+    },
+    {
+      "epoch": 0.9934731287321206,
+      "grad_norm": 3.765400422710313,
+      "learning_rate": 1.1172455977787044e-09,
+      "loss": 0.4228,
+      "step": 14308
+    },
+    {
+      "epoch": 0.9935425635328426,
+      "grad_norm": 3.804016674577753,
+      "learning_rate": 1.0936017192786363e-09,
+      "loss": 0.5,
+      "step": 14309
+    },
+    {
+      "epoch": 0.9936119983335647,
+      "grad_norm": 3.1696330186674877,
+      "learning_rate": 1.0702106798077261e-09,
+      "loss": 0.2484,
+      "step": 14310
+    },
+    {
+      "epoch": 0.9936814331342869,
+      "grad_norm": 3.057451940789772,
+      "learning_rate": 1.0470724805489163e-09,
+      "loss": 0.3217,
+      "step": 14311
+    },
+    {
+      "epoch": 0.993750867935009,
+      "grad_norm": 4.276196679236634,
+      "learning_rate": 1.024187122672382e-09,
+      "loss": 0.4818,
+      "step": 14312
+    },
+    {
+      "epoch": 0.9938203027357312,
+      "grad_norm": 3.498487589320478,
+      "learning_rate": 1.0015546073355309e-09,
+      "loss": 0.2358,
+      "step": 14313
+    },
+    {
+      "epoch": 0.9938897375364533,
+      "grad_norm": 3.803493460886441,
+      "learning_rate": 9.791749356835579e-10,
+      "loss": 0.3194,
+      "step": 14314
+    },
+    {
+      "epoch": 0.9939591723371753,
+      "grad_norm": 3.811656074957991,
+      "learning_rate": 9.570481088483352e-10,
+      "loss": 0.269,
+      "step": 14315
+    },
+    {
+      "epoch": 0.9940286071378975,
+      "grad_norm": 3.935123208256896,
+      "learning_rate": 9.351741279484128e-10,
+      "loss": 0.3469,
+      "step": 14316
+    },
+    {
+      "epoch": 0.9940980419386196,
+      "grad_norm": 2.70180021577124,
+      "learning_rate": 9.135529940906829e-10,
+      "loss": 0.1904,
+      "step": 14317
+    },
+    {
+      "epoch": 0.9941674767393418,
+      "grad_norm": 3.241460830434318,
+      "learning_rate": 8.921847083687152e-10,
+      "loss": 0.2558,
+      "step": 14318
+    },
+    {
+      "epoch": 0.9942369115400639,
+      "grad_norm": 3.612127990033217,
+      "learning_rate": 8.710692718627567e-10,
+      "loss": 0.4021,
+      "step": 14319
+    },
+    {
+      "epoch": 0.994306346340786,
+      "grad_norm": 5.32856884521121,
+      "learning_rate": 8.502066856413971e-10,
+      "loss": 0.2568,
+      "step": 14320
+    },
+    {
+      "epoch": 0.9943757811415082,
+      "grad_norm": 4.136154519001808,
+      "learning_rate": 8.295969507599033e-10,
+      "loss": 0.4283,
+      "step": 14321
+    },
+    {
+      "epoch": 0.9944452159422302,
+      "grad_norm": 6.247594261274532,
+      "learning_rate": 8.092400682602198e-10,
+      "loss": 0.8369,
+      "step": 14322
+    },
+    {
+      "epoch": 0.9945146507429524,
+      "grad_norm": 6.899591566016916,
+      "learning_rate": 7.891360391726332e-10,
+      "loss": 0.5912,
+      "step": 14323
+    },
+    {
+      "epoch": 0.9945840855436745,
+      "grad_norm": 3.6852629169544016,
+      "learning_rate": 7.692848645129979e-10,
+      "loss": 0.2857,
+      "step": 14324
+    },
+    {
+      "epoch": 0.9946535203443966,
+      "grad_norm": 4.620357253190392,
+      "learning_rate": 7.496865452866209e-10,
+      "loss": 0.3683,
+      "step": 14325
+    },
+    {
+      "epoch": 0.9947229551451188,
+      "grad_norm": 4.259646133206421,
+      "learning_rate": 7.303410824832657e-10,
+      "loss": 0.356,
+      "step": 14326
+    },
+    {
+      "epoch": 0.9947923899458408,
+      "grad_norm": 3.3375291263487274,
+      "learning_rate": 7.112484770827044e-10,
+      "loss": 0.3905,
+      "step": 14327
+    },
+    {
+      "epoch": 0.9948618247465629,
+      "grad_norm": 2.3152696550521927,
+      "learning_rate": 6.924087300497206e-10,
+      "loss": 0.1936,
+      "step": 14328
+    },
+    {
+      "epoch": 0.9949312595472851,
+      "grad_norm": 4.3917848806331925,
+      "learning_rate": 6.738218423379961e-10,
+      "loss": 0.4615,
+      "step": 14329
+    },
+    {
+      "epoch": 0.9950006943480072,
+      "grad_norm": 1.8163046293100398,
+      "learning_rate": 6.554878148873344e-10,
+      "loss": 0.1138,
+      "step": 14330
+    },
+    {
+      "epoch": 0.9950701291487294,
+      "grad_norm": 3.399325711960607,
+      "learning_rate": 6.37406648624217e-10,
+      "loss": 0.3499,
+      "step": 14331
+    },
+    {
+      "epoch": 0.9951395639494515,
+      "grad_norm": 4.652243487534354,
+      "learning_rate": 6.195783444645776e-10,
+      "loss": 0.496,
+      "step": 14332
+    },
+    {
+      "epoch": 0.9952089987501735,
+      "grad_norm": 3.7719703148456194,
+      "learning_rate": 6.020029033093622e-10,
+      "loss": 0.3249,
+      "step": 14333
+    },
+    {
+      "epoch": 0.9952784335508957,
+      "grad_norm": 3.41673245698817,
+      "learning_rate": 5.846803260478595e-10,
+      "loss": 0.2825,
+      "step": 14334
+    },
+    {
+      "epoch": 0.9953478683516178,
+      "grad_norm": 4.828105014118911,
+      "learning_rate": 5.676106135554805e-10,
+      "loss": 0.5024,
+      "step": 14335
+    },
+    {
+      "epoch": 0.99541730315234,
+      "grad_norm": 4.223585567026278,
+      "learning_rate": 5.507937666965335e-10,
+      "loss": 0.5431,
+      "step": 14336
+    },
+    {
+      "epoch": 0.9954867379530621,
+      "grad_norm": 4.876472785475875,
+      "learning_rate": 5.342297863208945e-10,
+      "loss": 0.3773,
+      "step": 14337
+    },
+    {
+      "epoch": 0.9955561727537842,
+      "grad_norm": 4.052987243835515,
+      "learning_rate": 5.179186732667818e-10,
+      "loss": 0.5475,
+      "step": 14338
+    },
+    {
+      "epoch": 0.9956256075545064,
+      "grad_norm": 4.292254185584279,
+      "learning_rate": 5.018604283590911e-10,
+      "loss": 0.4824,
+      "step": 14339
+    },
+    {
+      "epoch": 0.9956950423552284,
+      "grad_norm": 4.452921171595044,
+      "learning_rate": 4.860550524099505e-10,
+      "loss": 0.5335,
+      "step": 14340
+    },
+    {
+      "epoch": 0.9957644771559505,
+      "grad_norm": 4.378897176409008,
+      "learning_rate": 4.705025462187207e-10,
+      "loss": 0.5605,
+      "step": 14341
+    },
+    {
+      "epoch": 0.9958339119566727,
+      "grad_norm": 4.4018985299093,
+      "learning_rate": 4.552029105719946e-10,
+      "loss": 0.5458,
+      "step": 14342
+    },
+    {
+      "epoch": 0.9959033467573948,
+      "grad_norm": 4.274582997510982,
+      "learning_rate": 4.401561462441528e-10,
+      "loss": 0.5652,
+      "step": 14343
+    },
+    {
+      "epoch": 0.995972781558117,
+      "grad_norm": 3.526885518426328,
+      "learning_rate": 4.25362253995143e-10,
+      "loss": 0.1884,
+      "step": 14344
+    },
+    {
+      "epoch": 0.996042216358839,
+      "grad_norm": 3.064406321902503,
+      "learning_rate": 4.1082123457436564e-10,
+      "loss": 0.2163,
+      "step": 14345
+    },
+    {
+      "epoch": 0.9961116511595611,
+      "grad_norm": 3.439231271991117,
+      "learning_rate": 3.965330887167884e-10,
+      "loss": 0.3964,
+      "step": 14346
+    },
+    {
+      "epoch": 0.9961810859602833,
+      "grad_norm": 3.9051098654732215,
+      "learning_rate": 3.8249781714572167e-10,
+      "loss": 0.5116,
+      "step": 14347
+    },
+    {
+      "epoch": 0.9962505207610054,
+      "grad_norm": 4.59841724202004,
+      "learning_rate": 3.687154205694876e-10,
+      "loss": 0.5343,
+      "step": 14348
+    },
+    {
+      "epoch": 0.9963199555617276,
+      "grad_norm": 3.7779883918339126,
+      "learning_rate": 3.551858996869717e-10,
+      "loss": 0.4635,
+      "step": 14349
+    },
+    {
+      "epoch": 0.9963893903624497,
+      "grad_norm": 4.713106235984339,
+      "learning_rate": 3.4190925518096106e-10,
+      "loss": 0.546,
+      "step": 14350
+    },
+    {
+      "epoch": 0.9964588251631717,
+      "grad_norm": 4.271927391880247,
+      "learning_rate": 3.288854877242509e-10,
+      "loss": 0.4137,
+      "step": 14351
+    },
+    {
+      "epoch": 0.9965282599638939,
+      "grad_norm": 4.631151367705543,
+      "learning_rate": 3.161145979746483e-10,
+      "loss": 0.5257,
+      "step": 14352
+    },
+    {
+      "epoch": 0.996597694764616,
+      "grad_norm": 3.8280930635546975,
+      "learning_rate": 3.0359658657830306e-10,
+      "loss": 0.4239,
+      "step": 14353
+    },
+    {
+      "epoch": 0.9966671295653381,
+      "grad_norm": 3.6349747874549223,
+      "learning_rate": 2.9133145416915253e-10,
+      "loss": 0.374,
+      "step": 14354
+    },
+    {
+      "epoch": 0.9967365643660603,
+      "grad_norm": 3.7011679805033055,
+      "learning_rate": 2.793192013661461e-10,
+      "loss": 0.3401,
+      "step": 14355
+    },
+    {
+      "epoch": 0.9968059991667824,
+      "grad_norm": 4.464926053303731,
+      "learning_rate": 2.67559828778241e-10,
+      "loss": 0.6501,
+      "step": 14356
+    },
+    {
+      "epoch": 0.9968754339675046,
+      "grad_norm": 4.7481025144504665,
+      "learning_rate": 2.560533369994067e-10,
+      "loss": 0.3476,
+      "step": 14357
+    },
+    {
+      "epoch": 0.9969448687682266,
+      "grad_norm": 4.131734328997915,
+      "learning_rate": 2.4479972661139993e-10,
+      "loss": 0.4138,
+      "step": 14358
+    },
+    {
+      "epoch": 0.9970143035689487,
+      "grad_norm": 4.881259808307799,
+      "learning_rate": 2.3379899818432025e-10,
+      "loss": 0.5463,
+      "step": 14359
+    },
+    {
+      "epoch": 0.9970837383696709,
+      "grad_norm": 6.08629152175895,
+      "learning_rate": 2.2305115227383433e-10,
+      "loss": 0.285,
+      "step": 14360
+    },
+    {
+      "epoch": 0.997153173170393,
+      "grad_norm": 3.9147647844031073,
+      "learning_rate": 2.125561894239514e-10,
+      "loss": 0.4627,
+      "step": 14361
+    },
+    {
+      "epoch": 0.9972226079711152,
+      "grad_norm": 4.296743034553536,
+      "learning_rate": 2.0231411016480295e-10,
+      "loss": 0.4637,
+      "step": 14362
+    },
+    {
+      "epoch": 0.9972920427718373,
+      "grad_norm": 4.869943486579398,
+      "learning_rate": 1.9232491501541827e-10,
+      "loss": 0.498,
+      "step": 14363
+    },
+    {
+      "epoch": 0.9973614775725593,
+      "grad_norm": 2.913578423169678,
+      "learning_rate": 1.8258860448094885e-10,
+      "loss": 0.2188,
+      "step": 14364
+    },
+    {
+      "epoch": 0.9974309123732815,
+      "grad_norm": 4.394009231080312,
+      "learning_rate": 1.7310517905266832e-10,
+      "loss": 0.3591,
+      "step": 14365
+    },
+    {
+      "epoch": 0.9975003471740036,
+      "grad_norm": 4.473613556361954,
+      "learning_rate": 1.638746392113033e-10,
+      "loss": 0.351,
+      "step": 14366
+    },
+    {
+      "epoch": 0.9975697819747257,
+      "grad_norm": 4.527002859289881,
+      "learning_rate": 1.5489698542370258e-10,
+      "loss": 0.298,
+      "step": 14367
+    },
+    {
+      "epoch": 0.9976392167754479,
+      "grad_norm": 4.931518392819714,
+      "learning_rate": 1.4617221814339222e-10,
+      "loss": 0.3902,
+      "step": 14368
+    },
+    {
+      "epoch": 0.99770865157617,
+      "grad_norm": 3.295643777382922,
+      "learning_rate": 1.3770033781224101e-10,
+      "loss": 0.2652,
+      "step": 14369
+    },
+    {
+      "epoch": 0.9977780863768921,
+      "grad_norm": 4.336957428345614,
+      "learning_rate": 1.2948134485879504e-10,
+      "loss": 0.4922,
+      "step": 14370
+    },
+    {
+      "epoch": 0.9978475211776142,
+      "grad_norm": 3.800642716750994,
+      "learning_rate": 1.2151523969772262e-10,
+      "loss": 0.4345,
+      "step": 14371
+    },
+    {
+      "epoch": 0.9979169559783363,
+      "grad_norm": 5.637153835001134,
+      "learning_rate": 1.1380202273314489e-10,
+      "loss": 0.4615,
+      "step": 14372
+    },
+    {
+      "epoch": 0.9979863907790585,
+      "grad_norm": 4.372711692886307,
+      "learning_rate": 1.0634169435419505e-10,
+      "loss": 0.3265,
+      "step": 14373
+    },
+    {
+      "epoch": 0.9980558255797806,
+      "grad_norm": 4.03590999986349,
+      "learning_rate": 9.913425493945916e-11,
+      "loss": 0.2412,
+      "step": 14374
+    },
+    {
+      "epoch": 0.9981252603805028,
+      "grad_norm": 3.0456800452687065,
+      "learning_rate": 9.217970485198013e-11,
+      "loss": 0.3345,
+      "step": 14375
+    },
+    {
+      "epoch": 0.9981946951812248,
+      "grad_norm": 4.965162411580839,
+      "learning_rate": 8.54780444448089e-11,
+      "loss": 0.5296,
+      "step": 14376
+    },
+    {
+      "epoch": 0.9982641299819469,
+      "grad_norm": 4.744323212513148,
+      "learning_rate": 7.902927405600835e-11,
+      "loss": 0.5469,
+      "step": 14377
+    },
+    {
+      "epoch": 0.9983335647826691,
+      "grad_norm": 4.124086971903773,
+      "learning_rate": 7.28333940125392e-11,
+      "loss": 0.3947,
+      "step": 14378
+    },
+    {
+      "epoch": 0.9984029995833912,
+      "grad_norm": 3.4519323720795585,
+      "learning_rate": 6.68904046269292e-11,
+      "loss": 0.2258,
+      "step": 14379
+    },
+    {
+      "epoch": 0.9984724343841134,
+      "grad_norm": 5.76855628145136,
+      "learning_rate": 6.120030620060391e-11,
+      "loss": 0.6673,
+      "step": 14380
+    },
+    {
+      "epoch": 0.9985418691848355,
+      "grad_norm": 3.891626573084236,
+      "learning_rate": 5.576309902055599e-11,
+      "loss": 0.4636,
+      "step": 14381
+    },
+    {
+      "epoch": 0.9986113039855575,
+      "grad_norm": 3.159311806430729,
+      "learning_rate": 5.0578783362675856e-11,
+      "loss": 0.3372,
+      "step": 14382
+    },
+    {
+      "epoch": 0.9986807387862797,
+      "grad_norm": 4.208190358657429,
+      "learning_rate": 4.5647359488421025e-11,
+      "loss": 0.473,
+      "step": 14383
+    },
+    {
+      "epoch": 0.9987501735870018,
+      "grad_norm": 3.1290034236584243,
+      "learning_rate": 4.096882764759169e-11,
+      "loss": 0.2771,
+      "step": 14384
+    },
+    {
+      "epoch": 0.9988196083877239,
+      "grad_norm": 3.589393401645181,
+      "learning_rate": 3.6543188076110235e-11,
+      "loss": 0.2638,
+      "step": 14385
+    },
+    {
+      "epoch": 0.9988890431884461,
+      "grad_norm": 4.011885412095503,
+      "learning_rate": 3.237044099935194e-11,
+      "loss": 0.2933,
+      "step": 14386
+    },
+    {
+      "epoch": 0.9989584779891681,
+      "grad_norm": 2.367969525572387,
+      "learning_rate": 2.8450586626593836e-11,
+      "loss": 0.2107,
+      "step": 14387
+    },
+    {
+      "epoch": 0.9990279127898903,
+      "grad_norm": 3.986704134123939,
+      "learning_rate": 2.4783625157676072e-11,
+      "loss": 0.3447,
+      "step": 14388
+    },
+    {
+      "epoch": 0.9990973475906124,
+      "grad_norm": 3.0388981460566695,
+      "learning_rate": 2.136955677689567e-11,
+      "loss": 0.1861,
+      "step": 14389
+    },
+    {
+      "epoch": 0.9991667823913345,
+      "grad_norm": 3.0442833650455903,
+      "learning_rate": 1.8208381657447426e-11,
+      "loss": 0.3338,
+      "step": 14390
+    },
+    {
+      "epoch": 0.9992362171920567,
+      "grad_norm": 4.509848356790682,
+      "learning_rate": 1.5300099959203453e-11,
+      "loss": 0.4787,
+      "step": 14391
+    },
+    {
+      "epoch": 0.9993056519927788,
+      "grad_norm": 4.103838688282994,
+      "learning_rate": 1.2644711829268297e-11,
+      "loss": 0.31,
+      "step": 14392
+    },
+    {
+      "epoch": 0.999375086793501,
+      "grad_norm": 3.262261419453084,
+      "learning_rate": 1.024221740197895e-11,
+      "loss": 0.2631,
+      "step": 14393
+    },
+    {
+      "epoch": 0.999444521594223,
+      "grad_norm": 2.661518672983248,
+      "learning_rate": 8.092616798904829e-12,
+      "loss": 0.1516,
+      "step": 14394
+    },
+    {
+      "epoch": 0.9995139563949451,
+      "grad_norm": 3.9376351127357148,
+      "learning_rate": 6.195910128292682e-12,
+      "loss": 0.4019,
+      "step": 14395
+    },
+    {
+      "epoch": 0.9995833911956673,
+      "grad_norm": 4.374059515767892,
+      "learning_rate": 4.5520974867319104e-12,
+      "loss": 0.5041,
+      "step": 14396
+    },
+    {
+      "epoch": 0.9996528259963894,
+      "grad_norm": 2.606162531344671,
+      "learning_rate": 3.16117895693413e-12,
+      "loss": 0.3171,
+      "step": 14397
+    },
+    {
+      "epoch": 0.9997222607971115,
+      "grad_norm": 3.9646854163658993,
+      "learning_rate": 2.0231546088433917e-12,
+      "loss": 0.5151,
+      "step": 14398
+    },
+    {
+      "epoch": 0.9997916955978337,
+      "grad_norm": 5.148206642292722,
+      "learning_rate": 1.1380245013015156e-12,
+      "loss": 0.3505,
+      "step": 14399
+    },
+    {
+      "epoch": 0.9998611303985557,
+      "grad_norm": 6.692968332845101,
+      "learning_rate": 5.057886776071996e-13,
+      "loss": 0.3662,
+      "step": 14400
+    },
+    {
+      "epoch": 0.9999305651992779,
+      "grad_norm": 3.977562708427656,
+      "learning_rate": 1.2644717106713445e-13,
+      "loss": 0.2606,
+      "step": 14401
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.821303598316035,
+      "learning_rate": 0.0,
+      "loss": 0.4289,
+      "step": 14402
+    },
+    {
+      "epoch": 1.0,
+      "step": 14402,
+      "total_flos": 153613769664512.0,
+      "train_loss": 0.47091111083759557,
+      "train_runtime": 42172.3476,
+      "train_samples_per_second": 2.732,
+      "train_steps_per_second": 0.342
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 14402,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 5000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 153613769664512.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}