diff --git "a/checkpoint-4968/trainer_state.json" "b/checkpoint-4968/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-4968/trainer_state.json"
@@ -0,0 +1,34881 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9996981587684877,
+  "eval_steps": 500,
+  "global_step": 4968,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0006036824630244491,
+      "grad_norm": 0.876740574836731,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 1.3437,
+      "step": 1
+    },
+    {
+      "epoch": 0.0012073649260488982,
+      "grad_norm": 0.8783963322639465,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 1.4745,
+      "step": 2
+    },
+    {
+      "epoch": 0.0018110473890733474,
+      "grad_norm": 0.8955765962600708,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.6589,
+      "step": 3
+    },
+    {
+      "epoch": 0.0024147298520977964,
+      "grad_norm": 1.0829285383224487,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 1.5662,
+      "step": 4
+    },
+    {
+      "epoch": 0.003018412315122246,
+      "grad_norm": 1.1975041627883911,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.4658,
+      "step": 5
+    },
+    {
+      "epoch": 0.003622094778146695,
+      "grad_norm": 1.0904583930969238,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.6691,
+      "step": 6
+    },
+    {
+      "epoch": 0.004225777241171144,
+      "grad_norm": 0.6331161260604858,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 1.5705,
+      "step": 7
+    },
+    {
+      "epoch": 0.004829459704195593,
+      "grad_norm": 0.3753370940685272,
+      "learning_rate": 1.0666666666666667e-05,
+      "loss": 1.5175,
+      "step": 8
+    },
+    {
+      "epoch": 0.005433142167220043,
+      "grad_norm": 2.936147451400757,
+      "learning_rate": 1.2e-05,
+      "loss": 1.5404,
+      "step": 9
+    },
+    {
+      "epoch": 0.006036824630244492,
+      "grad_norm": 2.635810613632202,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.5605,
+      "step": 10
+    },
+    {
+      "epoch": 0.006640507093268941,
+      "grad_norm": 1.9180281162261963,
+      "learning_rate": 1.4666666666666668e-05,
+      "loss": 1.5239,
+      "step": 11
+    },
+    {
+      "epoch": 0.00724418955629339,
+      "grad_norm": 3.021333932876587,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.5081,
+      "step": 12
+    },
+    {
+      "epoch": 0.00784787201931784,
+      "grad_norm": 1.9040541648864746,
+      "learning_rate": 1.7333333333333336e-05,
+      "loss": 1.4366,
+      "step": 13
+    },
+    {
+      "epoch": 0.008451554482342288,
+      "grad_norm": 1.0555269718170166,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 1.6606,
+      "step": 14
+    },
+    {
+      "epoch": 0.009055236945366738,
+      "grad_norm": 0.5196813344955444,
+      "learning_rate": 2e-05,
+      "loss": 1.7105,
+      "step": 15
+    },
+    {
+      "epoch": 0.009658919408391186,
+      "grad_norm": 1.1674487590789795,
+      "learning_rate": 2.1333333333333335e-05,
+      "loss": 1.3767,
+      "step": 16
+    },
+    {
+      "epoch": 0.010262601871415635,
+      "grad_norm": 1.3149781227111816,
+      "learning_rate": 2.2666666666666668e-05,
+      "loss": 1.2422,
+      "step": 17
+    },
+    {
+      "epoch": 0.010866284334440085,
+      "grad_norm": 0.5175302028656006,
+      "learning_rate": 2.4e-05,
+      "loss": 1.3791,
+      "step": 18
+    },
+    {
+      "epoch": 0.011469966797464533,
+      "grad_norm": 0.3626837432384491,
+      "learning_rate": 2.5333333333333337e-05,
+      "loss": 1.4945,
+      "step": 19
+    },
+    {
+      "epoch": 0.012073649260488983,
+      "grad_norm": 0.26292356848716736,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 1.3598,
+      "step": 20
+    },
+    {
+      "epoch": 0.012677331723513431,
+      "grad_norm": 0.2678084671497345,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 1.3896,
+      "step": 21
+    },
+    {
+      "epoch": 0.013281014186537881,
+      "grad_norm": 0.34738680720329285,
+      "learning_rate": 2.9333333333333336e-05,
+      "loss": 1.3734,
+      "step": 22
+    },
+    {
+      "epoch": 0.01388469664956233,
+      "grad_norm": 0.3626849949359894,
+      "learning_rate": 3.066666666666667e-05,
+      "loss": 1.306,
+      "step": 23
+    },
+    {
+      "epoch": 0.01448837911258678,
+      "grad_norm": 0.2588627338409424,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 1.3373,
+      "step": 24
+    },
+    {
+      "epoch": 0.01509206157561123,
+      "grad_norm": 0.6032963395118713,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 1.1394,
+      "step": 25
+    },
+    {
+      "epoch": 0.01569574403863568,
+      "grad_norm": 0.194705531001091,
+      "learning_rate": 3.466666666666667e-05,
+      "loss": 1.0727,
+      "step": 26
+    },
+    {
+      "epoch": 0.016299426501660125,
+      "grad_norm": 0.3403484523296356,
+      "learning_rate": 3.6e-05,
+      "loss": 1.045,
+      "step": 27
+    },
+    {
+      "epoch": 0.016903108964684575,
+      "grad_norm": 0.3210083544254303,
+      "learning_rate": 3.733333333333334e-05,
+      "loss": 1.285,
+      "step": 28
+    },
+    {
+      "epoch": 0.017506791427709025,
+      "grad_norm": 0.6296122074127197,
+      "learning_rate": 3.866666666666667e-05,
+      "loss": 1.15,
+      "step": 29
+    },
+    {
+      "epoch": 0.018110473890733475,
+      "grad_norm": 0.4753726124763489,
+      "learning_rate": 4e-05,
+      "loss": 1.0053,
+      "step": 30
+    },
+    {
+      "epoch": 0.018714156353757925,
+      "grad_norm": 0.19821159541606903,
+      "learning_rate": 4.133333333333333e-05,
+      "loss": 1.1965,
+      "step": 31
+    },
+    {
+      "epoch": 0.01931783881678237,
+      "grad_norm": 0.2836245000362396,
+      "learning_rate": 4.266666666666667e-05,
+      "loss": 1.0633,
+      "step": 32
+    },
+    {
+      "epoch": 0.01992152127980682,
+      "grad_norm": 0.19199256598949432,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.9728,
+      "step": 33
+    },
+    {
+      "epoch": 0.02052520374283127,
+      "grad_norm": 0.17820274829864502,
+      "learning_rate": 4.5333333333333335e-05,
+      "loss": 1.0132,
+      "step": 34
+    },
+    {
+      "epoch": 0.02112888620585572,
+      "grad_norm": 0.20217391848564148,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 1.0202,
+      "step": 35
+    },
+    {
+      "epoch": 0.02173256866888017,
+      "grad_norm": 0.18383841216564178,
+      "learning_rate": 4.8e-05,
+      "loss": 0.9666,
+      "step": 36
+    },
+    {
+      "epoch": 0.022336251131904617,
+      "grad_norm": 0.17872729897499084,
+      "learning_rate": 4.933333333333334e-05,
+      "loss": 0.9385,
+      "step": 37
+    },
+    {
+      "epoch": 0.022939933594929067,
+      "grad_norm": 0.14148443937301636,
+      "learning_rate": 5.0666666666666674e-05,
+      "loss": 0.9338,
+      "step": 38
+    },
+    {
+      "epoch": 0.023543616057953517,
+      "grad_norm": 0.14587634801864624,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 0.964,
+      "step": 39
+    },
+    {
+      "epoch": 0.024147298520977967,
+      "grad_norm": 0.17298458516597748,
+      "learning_rate": 5.333333333333333e-05,
+      "loss": 0.9934,
+      "step": 40
+    },
+    {
+      "epoch": 0.024750980984002413,
+      "grad_norm": 0.1523771435022354,
+      "learning_rate": 5.466666666666666e-05,
+      "loss": 0.9995,
+      "step": 41
+    },
+    {
+      "epoch": 0.025354663447026863,
+      "grad_norm": 0.16784684360027313,
+      "learning_rate": 5.6000000000000006e-05,
+      "loss": 0.9621,
+      "step": 42
+    },
+    {
+      "epoch": 0.025958345910051313,
+      "grad_norm": 0.18494223058223724,
+      "learning_rate": 5.7333333333333336e-05,
+      "loss": 0.9181,
+      "step": 43
+    },
+    {
+      "epoch": 0.026562028373075763,
+      "grad_norm": 0.15912078320980072,
+      "learning_rate": 5.866666666666667e-05,
+      "loss": 0.9311,
+      "step": 44
+    },
+    {
+      "epoch": 0.027165710836100213,
+      "grad_norm": 0.17092670500278473,
+      "learning_rate": 6e-05,
+      "loss": 0.8838,
+      "step": 45
+    },
+    {
+      "epoch": 0.02776939329912466,
+      "grad_norm": 0.2552148401737213,
+      "learning_rate": 6.133333333333334e-05,
+      "loss": 0.9121,
+      "step": 46
+    },
+    {
+      "epoch": 0.02837307576214911,
+      "grad_norm": 0.20719660818576813,
+      "learning_rate": 6.266666666666667e-05,
+      "loss": 0.8881,
+      "step": 47
+    },
+    {
+      "epoch": 0.02897675822517356,
+      "grad_norm": 0.25345510244369507,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 0.8592,
+      "step": 48
+    },
+    {
+      "epoch": 0.02958044068819801,
+      "grad_norm": 0.23677493631839752,
+      "learning_rate": 6.533333333333334e-05,
+      "loss": 0.8428,
+      "step": 49
+    },
+    {
+      "epoch": 0.03018412315122246,
+      "grad_norm": 0.3329518437385559,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.7946,
+      "step": 50
+    },
+    {
+      "epoch": 0.030787805614246905,
+      "grad_norm": 0.1412961632013321,
+      "learning_rate": 6.800000000000001e-05,
+      "loss": 0.9836,
+      "step": 51
+    },
+    {
+      "epoch": 0.03139148807727136,
+      "grad_norm": 0.1430910974740982,
+      "learning_rate": 6.933333333333334e-05,
+      "loss": 1.0782,
+      "step": 52
+    },
+    {
+      "epoch": 0.031995170540295804,
+      "grad_norm": 0.19594986736774445,
+      "learning_rate": 7.066666666666667e-05,
+      "loss": 0.9402,
+      "step": 53
+    },
+    {
+      "epoch": 0.03259885300332025,
+      "grad_norm": 0.15413224697113037,
+      "learning_rate": 7.2e-05,
+      "loss": 1.0387,
+      "step": 54
+    },
+    {
+      "epoch": 0.033202535466344704,
+      "grad_norm": 0.15673322975635529,
+      "learning_rate": 7.333333333333333e-05,
+      "loss": 0.9795,
+      "step": 55
+    },
+    {
+      "epoch": 0.03380621792936915,
+      "grad_norm": 0.14926938712596893,
+      "learning_rate": 7.466666666666667e-05,
+      "loss": 1.088,
+      "step": 56
+    },
+    {
+      "epoch": 0.034409900392393604,
+      "grad_norm": 0.20912237465381622,
+      "learning_rate": 7.6e-05,
+      "loss": 0.9247,
+      "step": 57
+    },
+    {
+      "epoch": 0.03501358285541805,
+      "grad_norm": 0.1995040327310562,
+      "learning_rate": 7.733333333333333e-05,
+      "loss": 0.9202,
+      "step": 58
+    },
+    {
+      "epoch": 0.0356172653184425,
+      "grad_norm": 0.20646536350250244,
+      "learning_rate": 7.866666666666666e-05,
+      "loss": 0.8604,
+      "step": 59
+    },
+    {
+      "epoch": 0.03622094778146695,
+      "grad_norm": 0.2546747624874115,
+      "learning_rate": 8e-05,
+      "loss": 0.955,
+      "step": 60
+    },
+    {
+      "epoch": 0.036824630244491396,
+      "grad_norm": 0.25772592425346375,
+      "learning_rate": 8.133333333333334e-05,
+      "loss": 0.9363,
+      "step": 61
+    },
+    {
+      "epoch": 0.03742831270751585,
+      "grad_norm": 0.4138891398906708,
+      "learning_rate": 8.266666666666667e-05,
+      "loss": 0.83,
+      "step": 62
+    },
+    {
+      "epoch": 0.038031995170540296,
+      "grad_norm": 0.2324582189321518,
+      "learning_rate": 8.4e-05,
+      "loss": 0.95,
+      "step": 63
+    },
+    {
+      "epoch": 0.03863567763356474,
+      "grad_norm": 0.22521646320819855,
+      "learning_rate": 8.533333333333334e-05,
+      "loss": 0.8359,
+      "step": 64
+    },
+    {
+      "epoch": 0.039239360096589196,
+      "grad_norm": 0.19713546335697174,
+      "learning_rate": 8.666666666666667e-05,
+      "loss": 0.9552,
+      "step": 65
+    },
+    {
+      "epoch": 0.03984304255961364,
+      "grad_norm": 0.1569695621728897,
+      "learning_rate": 8.800000000000001e-05,
+      "loss": 0.9283,
+      "step": 66
+    },
+    {
+      "epoch": 0.040446725022638096,
+      "grad_norm": 0.13675987720489502,
+      "learning_rate": 8.933333333333334e-05,
+      "loss": 0.9144,
+      "step": 67
+    },
+    {
+      "epoch": 0.04105040748566254,
+      "grad_norm": 0.08005592226982117,
+      "learning_rate": 9.066666666666667e-05,
+      "loss": 0.9661,
+      "step": 68
+    },
+    {
+      "epoch": 0.04165408994868699,
+      "grad_norm": 0.17173191905021667,
+      "learning_rate": 9.200000000000001e-05,
+      "loss": 1.1486,
+      "step": 69
+    },
+    {
+      "epoch": 0.04225777241171144,
+      "grad_norm": 0.2080589085817337,
+      "learning_rate": 9.333333333333334e-05,
+      "loss": 0.8091,
+      "step": 70
+    },
+    {
+      "epoch": 0.04286145487473589,
+      "grad_norm": 0.21545596420764923,
+      "learning_rate": 9.466666666666667e-05,
+      "loss": 0.8606,
+      "step": 71
+    },
+    {
+      "epoch": 0.04346513733776034,
+      "grad_norm": 0.21326108276844025,
+      "learning_rate": 9.6e-05,
+      "loss": 0.9271,
+      "step": 72
+    },
+    {
+      "epoch": 0.04406881980078479,
+      "grad_norm": 0.21748293936252594,
+      "learning_rate": 9.733333333333335e-05,
+      "loss": 1.1042,
+      "step": 73
+    },
+    {
+      "epoch": 0.044672502263809234,
+      "grad_norm": 0.20910663902759552,
+      "learning_rate": 9.866666666666668e-05,
+      "loss": 0.8998,
+      "step": 74
+    },
+    {
+      "epoch": 0.04527618472683369,
+      "grad_norm": 0.19514353573322296,
+      "learning_rate": 0.0001,
+      "loss": 0.8976,
+      "step": 75
+    },
+    {
+      "epoch": 0.045879867189858134,
+      "grad_norm": 0.17271868884563446,
+      "learning_rate": 0.00010133333333333335,
+      "loss": 0.8919,
+      "step": 76
+    },
+    {
+      "epoch": 0.04648354965288258,
+      "grad_norm": 0.17146329581737518,
+      "learning_rate": 0.00010266666666666666,
+      "loss": 0.8703,
+      "step": 77
+    },
+    {
+      "epoch": 0.047087232115907034,
+      "grad_norm": 0.12326353043317795,
+      "learning_rate": 0.00010400000000000001,
+      "loss": 1.0872,
+      "step": 78
+    },
+    {
+      "epoch": 0.04769091457893148,
+      "grad_norm": 0.08894680440425873,
+      "learning_rate": 0.00010533333333333332,
+      "loss": 0.9586,
+      "step": 79
+    },
+    {
+      "epoch": 0.04829459704195593,
+      "grad_norm": 0.07784762233495712,
+      "learning_rate": 0.00010666666666666667,
+      "loss": 0.9304,
+      "step": 80
+    },
+    {
+      "epoch": 0.04889827950498038,
+      "grad_norm": 0.08525023609399796,
+      "learning_rate": 0.00010800000000000001,
+      "loss": 1.1813,
+      "step": 81
+    },
+    {
+      "epoch": 0.049501961968004826,
+      "grad_norm": 0.11610118299722672,
+      "learning_rate": 0.00010933333333333333,
+      "loss": 0.8561,
+      "step": 82
+    },
+    {
+      "epoch": 0.05010564443102928,
+      "grad_norm": 0.12873214483261108,
+      "learning_rate": 0.00011066666666666667,
+      "loss": 0.8394,
+      "step": 83
+    },
+    {
+      "epoch": 0.050709326894053726,
+      "grad_norm": 0.11888518184423447,
+      "learning_rate": 0.00011200000000000001,
+      "loss": 0.873,
+      "step": 84
+    },
+    {
+      "epoch": 0.05131300935707818,
+      "grad_norm": 0.12383485585451126,
+      "learning_rate": 0.00011333333333333334,
+      "loss": 0.803,
+      "step": 85
+    },
+    {
+      "epoch": 0.051916691820102626,
+      "grad_norm": 0.13411828875541687,
+      "learning_rate": 0.00011466666666666667,
+      "loss": 0.8332,
+      "step": 86
+    },
+    {
+      "epoch": 0.05252037428312707,
+      "grad_norm": 0.12794946134090424,
+      "learning_rate": 0.000116,
+      "loss": 0.8405,
+      "step": 87
+    },
+    {
+      "epoch": 0.053124056746151525,
+      "grad_norm": 0.12056224048137665,
+      "learning_rate": 0.00011733333333333334,
+      "loss": 0.9108,
+      "step": 88
+    },
+    {
+      "epoch": 0.05372773920917597,
+      "grad_norm": 0.11550690233707428,
+      "learning_rate": 0.00011866666666666669,
+      "loss": 0.9052,
+      "step": 89
+    },
+    {
+      "epoch": 0.054331421672200425,
+      "grad_norm": 0.09730254113674164,
+      "learning_rate": 0.00012,
+      "loss": 0.9104,
+      "step": 90
+    },
+    {
+      "epoch": 0.05493510413522487,
+      "grad_norm": 0.10430070012807846,
+      "learning_rate": 0.00012133333333333335,
+      "loss": 0.8333,
+      "step": 91
+    },
+    {
+      "epoch": 0.05553878659824932,
+      "grad_norm": 0.10595888644456863,
+      "learning_rate": 0.00012266666666666668,
+      "loss": 0.7792,
+      "step": 92
+    },
+    {
+      "epoch": 0.05614246906127377,
+      "grad_norm": 0.13614441454410553,
+      "learning_rate": 0.000124,
+      "loss": 0.8317,
+      "step": 93
+    },
+    {
+      "epoch": 0.05674615152429822,
+      "grad_norm": 0.1545422077178955,
+      "learning_rate": 0.00012533333333333334,
+      "loss": 0.8139,
+      "step": 94
+    },
+    {
+      "epoch": 0.05734983398732267,
+      "grad_norm": 0.17710909247398376,
+      "learning_rate": 0.00012666666666666666,
+      "loss": 0.8969,
+      "step": 95
+    },
+    {
+      "epoch": 0.05795351645034712,
+      "grad_norm": 0.17680825293064117,
+      "learning_rate": 0.00012800000000000002,
+      "loss": 0.7812,
+      "step": 96
+    },
+    {
+      "epoch": 0.058557198913371564,
+      "grad_norm": 0.17637892067432404,
+      "learning_rate": 0.00012933333333333332,
+      "loss": 0.7518,
+      "step": 97
+    },
+    {
+      "epoch": 0.05916088137639602,
+      "grad_norm": 0.14965130388736725,
+      "learning_rate": 0.00013066666666666668,
+      "loss": 0.6467,
+      "step": 98
+    },
+    {
+      "epoch": 0.05976456383942046,
+      "grad_norm": 0.15980888903141022,
+      "learning_rate": 0.000132,
+      "loss": 0.724,
+      "step": 99
+    },
+    {
+      "epoch": 0.06036824630244492,
+      "grad_norm": 0.1575448364019394,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 0.523,
+      "step": 100
+    },
+    {
+      "epoch": 0.06097192876546936,
+      "grad_norm": 0.0918162539601326,
+      "learning_rate": 0.00013466666666666667,
+      "loss": 0.8158,
+      "step": 101
+    },
+    {
+      "epoch": 0.06157561122849381,
+      "grad_norm": 0.09651898592710495,
+      "learning_rate": 0.00013600000000000003,
+      "loss": 1.1298,
+      "step": 102
+    },
+    {
+      "epoch": 0.06217929369151826,
+      "grad_norm": 0.10406182706356049,
+      "learning_rate": 0.00013733333333333333,
+      "loss": 1.2655,
+      "step": 103
+    },
+    {
+      "epoch": 0.06278297615454272,
+      "grad_norm": 0.09533622860908508,
+      "learning_rate": 0.00013866666666666669,
+      "loss": 0.9408,
+      "step": 104
+    },
+    {
+      "epoch": 0.06338665861756716,
+      "grad_norm": 0.1048218235373497,
+      "learning_rate": 0.00014,
+      "loss": 0.9564,
+      "step": 105
+    },
+    {
+      "epoch": 0.06399034108059161,
+      "grad_norm": 0.09610721468925476,
+      "learning_rate": 0.00014133333333333334,
+      "loss": 0.8791,
+      "step": 106
+    },
+    {
+      "epoch": 0.06459402354361606,
+      "grad_norm": 0.08564590662717819,
+      "learning_rate": 0.00014266666666666667,
+      "loss": 1.0284,
+      "step": 107
+    },
+    {
+      "epoch": 0.0651977060066405,
+      "grad_norm": 0.07987428456544876,
+      "learning_rate": 0.000144,
+      "loss": 1.1595,
+      "step": 108
+    },
+    {
+      "epoch": 0.06580138846966496,
+      "grad_norm": 0.09171764552593231,
+      "learning_rate": 0.00014533333333333333,
+      "loss": 0.9098,
+      "step": 109
+    },
+    {
+      "epoch": 0.06640507093268941,
+      "grad_norm": 0.08003644645214081,
+      "learning_rate": 0.00014666666666666666,
+      "loss": 1.0118,
+      "step": 110
+    },
+    {
+      "epoch": 0.06700875339571385,
+      "grad_norm": 0.08099836111068726,
+      "learning_rate": 0.000148,
+      "loss": 1.0011,
+      "step": 111
+    },
+    {
+      "epoch": 0.0676124358587383,
+      "grad_norm": 0.08993902802467346,
+      "learning_rate": 0.00014933333333333335,
+      "loss": 1.2031,
+      "step": 112
+    },
+    {
+      "epoch": 0.06821611832176275,
+      "grad_norm": 0.0858883187174797,
+      "learning_rate": 0.00015066666666666668,
+      "loss": 0.7726,
+      "step": 113
+    },
+    {
+      "epoch": 0.06881980078478721,
+      "grad_norm": 0.07739171385765076,
+      "learning_rate": 0.000152,
+      "loss": 0.8213,
+      "step": 114
+    },
+    {
+      "epoch": 0.06942348324781165,
+      "grad_norm": 0.09992550313472748,
+      "learning_rate": 0.00015333333333333334,
+      "loss": 1.2648,
+      "step": 115
+    },
+    {
+      "epoch": 0.0700271657108361,
+      "grad_norm": 0.08734551072120667,
+      "learning_rate": 0.00015466666666666667,
+      "loss": 0.9113,
+      "step": 116
+    },
+    {
+      "epoch": 0.07063084817386055,
+      "grad_norm": 0.13743777573108673,
+      "learning_rate": 0.00015600000000000002,
+      "loss": 1.2323,
+      "step": 117
+    },
+    {
+      "epoch": 0.071234530636885,
+      "grad_norm": 0.08608172088861465,
+      "learning_rate": 0.00015733333333333333,
+      "loss": 1.0779,
+      "step": 118
+    },
+    {
+      "epoch": 0.07183821309990945,
+      "grad_norm": 0.07633006572723389,
+      "learning_rate": 0.00015866666666666668,
+      "loss": 0.9064,
+      "step": 119
+    },
+    {
+      "epoch": 0.0724418955629339,
+      "grad_norm": 0.07484789192676544,
+      "learning_rate": 0.00016,
+      "loss": 0.8816,
+      "step": 120
+    },
+    {
+      "epoch": 0.07304557802595835,
+      "grad_norm": 0.06947381794452667,
+      "learning_rate": 0.00016133333333333334,
+      "loss": 0.9609,
+      "step": 121
+    },
+    {
+      "epoch": 0.07364926048898279,
+      "grad_norm": 0.09331609308719635,
+      "learning_rate": 0.00016266666666666667,
+      "loss": 0.899,
+      "step": 122
+    },
+    {
+      "epoch": 0.07425294295200724,
+      "grad_norm": 0.0736856609582901,
+      "learning_rate": 0.000164,
+      "loss": 0.8749,
+      "step": 123
+    },
+    {
+      "epoch": 0.0748566254150317,
+      "grad_norm": 0.08171521127223969,
+      "learning_rate": 0.00016533333333333333,
+      "loss": 0.8952,
+      "step": 124
+    },
+    {
+      "epoch": 0.07546030787805615,
+      "grad_norm": 0.0707223191857338,
+      "learning_rate": 0.0001666666666666667,
+      "loss": 0.9492,
+      "step": 125
+    },
+    {
+      "epoch": 0.07606399034108059,
+      "grad_norm": 0.07196628302335739,
+      "learning_rate": 0.000168,
+      "loss": 1.0499,
+      "step": 126
+    },
+    {
+      "epoch": 0.07666767280410504,
+      "grad_norm": 0.07485999912023544,
+      "learning_rate": 0.00016933333333333335,
+      "loss": 0.8929,
+      "step": 127
+    },
+    {
+      "epoch": 0.07727135526712949,
+      "grad_norm": 0.06874241679906845,
+      "learning_rate": 0.00017066666666666668,
+      "loss": 0.8589,
+      "step": 128
+    },
+    {
+      "epoch": 0.07787503773015395,
+      "grad_norm": 0.07588055729866028,
+      "learning_rate": 0.000172,
+      "loss": 0.9973,
+      "step": 129
+    },
+    {
+      "epoch": 0.07847872019317839,
+      "grad_norm": 0.07442892342805862,
+      "learning_rate": 0.00017333333333333334,
+      "loss": 1.3857,
+      "step": 130
+    },
+    {
+      "epoch": 0.07908240265620284,
+      "grad_norm": 0.07991475611925125,
+      "learning_rate": 0.00017466666666666667,
+      "loss": 0.8986,
+      "step": 131
+    },
+    {
+      "epoch": 0.07968608511922728,
+      "grad_norm": 0.07599324733018875,
+      "learning_rate": 0.00017600000000000002,
+      "loss": 0.874,
+      "step": 132
+    },
+    {
+      "epoch": 0.08028976758225173,
+      "grad_norm": 0.08738681674003601,
+      "learning_rate": 0.00017733333333333335,
+      "loss": 0.8635,
+      "step": 133
+    },
+    {
+      "epoch": 0.08089345004527619,
+      "grad_norm": 0.08158082515001297,
+      "learning_rate": 0.00017866666666666668,
+      "loss": 0.8143,
+      "step": 134
+    },
+    {
+      "epoch": 0.08149713250830064,
+      "grad_norm": 0.08535363525152206,
+      "learning_rate": 0.00018,
+      "loss": 0.7629,
+      "step": 135
+    },
+    {
+      "epoch": 0.08210081497132508,
+      "grad_norm": 0.08319278806447983,
+      "learning_rate": 0.00018133333333333334,
+      "loss": 0.7807,
+      "step": 136
+    },
+    {
+      "epoch": 0.08270449743434953,
+      "grad_norm": 0.10717540234327316,
+      "learning_rate": 0.00018266666666666667,
+      "loss": 0.7731,
+      "step": 137
+    },
+    {
+      "epoch": 0.08330817989737398,
+      "grad_norm": 0.08758540451526642,
+      "learning_rate": 0.00018400000000000003,
+      "loss": 0.7561,
+      "step": 138
+    },
+    {
+      "epoch": 0.08391186236039844,
+      "grad_norm": 0.09366493672132492,
+      "learning_rate": 0.00018533333333333333,
+      "loss": 0.7703,
+      "step": 139
+    },
+    {
+      "epoch": 0.08451554482342288,
+      "grad_norm": 0.10316765308380127,
+      "learning_rate": 0.0001866666666666667,
+      "loss": 0.7798,
+      "step": 140
+    },
+    {
+      "epoch": 0.08511922728644733,
+      "grad_norm": 0.09655182808637619,
+      "learning_rate": 0.000188,
+      "loss": 0.8053,
+      "step": 141
+    },
+    {
+      "epoch": 0.08572290974947178,
+      "grad_norm": 0.09309862554073334,
+      "learning_rate": 0.00018933333333333335,
+      "loss": 0.7777,
+      "step": 142
+    },
+    {
+      "epoch": 0.08632659221249622,
+      "grad_norm": 0.11444847285747528,
+      "learning_rate": 0.00019066666666666668,
+      "loss": 0.8681,
+      "step": 143
+    },
+    {
+      "epoch": 0.08693027467552068,
+      "grad_norm": 0.09801818430423737,
+      "learning_rate": 0.000192,
+      "loss": 0.6858,
+      "step": 144
+    },
+    {
+      "epoch": 0.08753395713854513,
+      "grad_norm": 0.10028098523616791,
+      "learning_rate": 0.00019333333333333333,
+      "loss": 0.8063,
+      "step": 145
+    },
+    {
+      "epoch": 0.08813763960156958,
+      "grad_norm": 0.11404330283403397,
+      "learning_rate": 0.0001946666666666667,
+      "loss": 0.8138,
+      "step": 146
+    },
+    {
+      "epoch": 0.08874132206459402,
+      "grad_norm": 0.12089970707893372,
+      "learning_rate": 0.000196,
+      "loss": 0.653,
+      "step": 147
+    },
+    {
+      "epoch": 0.08934500452761847,
+      "grad_norm": 0.13996511697769165,
+      "learning_rate": 0.00019733333333333335,
+      "loss": 0.6567,
+      "step": 148
+    },
+    {
+      "epoch": 0.08994868699064293,
+      "grad_norm": 0.1570245325565338,
+      "learning_rate": 0.00019866666666666668,
+      "loss": 0.6688,
+      "step": 149
+    },
+    {
+      "epoch": 0.09055236945366738,
+      "grad_norm": 0.14232978224754333,
+      "learning_rate": 0.0002,
+      "loss": 0.5686,
+      "step": 150
+    },
+    {
+      "epoch": 0.09115605191669182,
+      "grad_norm": 0.09911559522151947,
+      "learning_rate": 0.00019995848899958488,
+      "loss": 0.9231,
+      "step": 151
+    },
+    {
+      "epoch": 0.09175973437971627,
+      "grad_norm": 0.0942869484424591,
+      "learning_rate": 0.0001999169779991698,
+      "loss": 0.9987,
+      "step": 152
+    },
+    {
+      "epoch": 0.09236341684274071,
+      "grad_norm": 0.08815930783748627,
+      "learning_rate": 0.00019987546699875468,
+      "loss": 1.0607,
+      "step": 153
+    },
+    {
+      "epoch": 0.09296709930576516,
+      "grad_norm": 0.0868469625711441,
+      "learning_rate": 0.00019983395599833956,
+      "loss": 0.9998,
+      "step": 154
+    },
+    {
+      "epoch": 0.09357078176878962,
+      "grad_norm": 0.07780348509550095,
+      "learning_rate": 0.00019979244499792446,
+      "loss": 0.911,
+      "step": 155
+    },
+    {
+      "epoch": 0.09417446423181407,
+      "grad_norm": 0.09089189767837524,
+      "learning_rate": 0.00019975093399750936,
+      "loss": 0.9344,
+      "step": 156
+    },
+    {
+      "epoch": 0.09477814669483851,
+      "grad_norm": 0.07558704167604446,
+      "learning_rate": 0.00019970942299709423,
+      "loss": 0.8435,
+      "step": 157
+    },
+    {
+      "epoch": 0.09538182915786296,
+      "grad_norm": 0.1009376272559166,
+      "learning_rate": 0.00019966791199667913,
+      "loss": 0.8571,
+      "step": 158
+    },
+    {
+      "epoch": 0.0959855116208874,
+      "grad_norm": 0.07474809885025024,
+      "learning_rate": 0.00019962640099626403,
+      "loss": 0.9764,
+      "step": 159
+    },
+    {
+      "epoch": 0.09658919408391187,
+      "grad_norm": 0.0775795429944992,
+      "learning_rate": 0.0001995848899958489,
+      "loss": 1.0164,
+      "step": 160
+    },
+    {
+      "epoch": 0.09719287654693631,
+      "grad_norm": 0.07155506312847137,
+      "learning_rate": 0.0001995433789954338,
+      "loss": 0.9886,
+      "step": 161
+    },
+    {
+      "epoch": 0.09779655900996076,
+      "grad_norm": 0.0788518413901329,
+      "learning_rate": 0.00019950186799501867,
+      "loss": 0.8248,
+      "step": 162
+    },
+    {
+      "epoch": 0.0984002414729852,
+      "grad_norm": 0.06630228459835052,
+      "learning_rate": 0.00019946035699460357,
+      "loss": 0.9792,
+      "step": 163
+    },
+    {
+      "epoch": 0.09900392393600965,
+      "grad_norm": 0.20900733768939972,
+      "learning_rate": 0.00019941884599418847,
+      "loss": 0.9877,
+      "step": 164
+    },
+    {
+      "epoch": 0.09960760639903411,
+      "grad_norm": 0.06860389560461044,
+      "learning_rate": 0.00019937733499377335,
+      "loss": 0.9082,
+      "step": 165
+    },
+    {
+      "epoch": 0.10021128886205856,
+      "grad_norm": 0.06557682156562805,
+      "learning_rate": 0.00019933582399335825,
+      "loss": 0.8677,
+      "step": 166
+    },
+    {
+      "epoch": 0.100814971325083,
+      "grad_norm": 0.07390300929546356,
+      "learning_rate": 0.00019929431299294315,
+      "loss": 1.1377,
+      "step": 167
+    },
+    {
+      "epoch": 0.10141865378810745,
+      "grad_norm": 0.07005932927131653,
+      "learning_rate": 0.00019925280199252802,
+      "loss": 1.0111,
+      "step": 168
+    },
+    {
+      "epoch": 0.1020223362511319,
+      "grad_norm": 0.06733494251966476,
+      "learning_rate": 0.00019921129099211292,
+      "loss": 0.9273,
+      "step": 169
+    },
+    {
+      "epoch": 0.10262601871415636,
+      "grad_norm": 0.4364582300186157,
+      "learning_rate": 0.00019916977999169782,
+      "loss": 0.863,
+      "step": 170
+    },
+    {
+      "epoch": 0.1032297011771808,
+      "grad_norm": 0.06318307667970657,
+      "learning_rate": 0.0001991282689912827,
+      "loss": 1.0073,
+      "step": 171
+    },
+    {
+      "epoch": 0.10383338364020525,
+      "grad_norm": 492.8619689941406,
+      "learning_rate": 0.00019908675799086757,
+      "loss": 0.9392,
+      "step": 172
+    },
+    {
+      "epoch": 0.1044370661032297,
+      "grad_norm": 14.513906478881836,
+      "learning_rate": 0.0001990452469904525,
+      "loss": 0.8524,
+      "step": 173
+    },
+    {
+      "epoch": 0.10504074856625414,
+      "grad_norm": 0.08823119848966599,
+      "learning_rate": 0.00019900373599003737,
+      "loss": 0.8712,
+      "step": 174
+    },
+    {
+      "epoch": 0.1056444310292786,
+      "grad_norm": 0.0739288404583931,
+      "learning_rate": 0.00019896222498962227,
+      "loss": 1.0265,
+      "step": 175
+    },
+    {
+      "epoch": 0.10624811349230305,
+      "grad_norm": 6.7402520179748535,
+      "learning_rate": 0.00019892071398920714,
+      "loss": 0.832,
+      "step": 176
+    },
+    {
+      "epoch": 0.1068517959553275,
+      "grad_norm": 0.20480509102344513,
+      "learning_rate": 0.00019887920298879204,
+      "loss": 0.962,
+      "step": 177
+    },
+    {
+      "epoch": 0.10745547841835194,
+      "grad_norm": 2.3439409732818604,
+      "learning_rate": 0.00019883769198837694,
+      "loss": 0.8044,
+      "step": 178
+    },
+    {
+      "epoch": 0.10805916088137639,
+      "grad_norm": 10.576101303100586,
+      "learning_rate": 0.0001987961809879618,
+      "loss": 1.1324,
+      "step": 179
+    },
+    {
+      "epoch": 0.10866284334440085,
+      "grad_norm": 5.063470840454102,
+      "learning_rate": 0.0001987546699875467,
+      "loss": 1.0371,
+      "step": 180
+    },
+    {
+      "epoch": 0.1092665258074253,
+      "grad_norm": 0.47612178325653076,
+      "learning_rate": 0.0001987131589871316,
+      "loss": 1.0325,
+      "step": 181
+    },
+    {
+      "epoch": 0.10987020827044974,
+      "grad_norm": 0.5040671229362488,
+      "learning_rate": 0.00019867164798671649,
+      "loss": 0.8587,
+      "step": 182
+    },
+    {
+      "epoch": 0.11047389073347419,
+      "grad_norm": 0.49563658237457275,
+      "learning_rate": 0.00019863013698630139,
+      "loss": 0.9701,
+      "step": 183
+    },
+    {
+      "epoch": 0.11107757319649864,
+      "grad_norm": 0.5639724731445312,
+      "learning_rate": 0.00019858862598588629,
+      "loss": 0.8009,
+      "step": 184
+    },
+    {
+      "epoch": 0.1116812556595231,
+      "grad_norm": 0.10233303904533386,
+      "learning_rate": 0.00019854711498547116,
+      "loss": 0.8951,
+      "step": 185
+    },
+    {
+      "epoch": 0.11228493812254754,
+      "grad_norm": 0.1553623080253601,
+      "learning_rate": 0.00019850560398505603,
+      "loss": 0.7539,
+      "step": 186
+    },
+    {
+      "epoch": 0.11288862058557199,
+      "grad_norm": 0.15513239800930023,
+      "learning_rate": 0.00019846409298464096,
+      "loss": 0.8451,
+      "step": 187
+    },
+    {
+      "epoch": 0.11349230304859644,
+      "grad_norm": 0.15111730992794037,
+      "learning_rate": 0.00019842258198422583,
+      "loss": 0.8565,
+      "step": 188
+    },
+    {
+      "epoch": 0.11409598551162088,
+      "grad_norm": 0.17460127174854279,
+      "learning_rate": 0.0001983810709838107,
+      "loss": 0.8017,
+      "step": 189
+    },
+    {
+      "epoch": 0.11469966797464534,
+      "grad_norm": 0.19736827909946442,
+      "learning_rate": 0.0001983395599833956,
+      "loss": 0.7927,
+      "step": 190
+    },
+    {
+      "epoch": 0.11530335043766979,
+      "grad_norm": 0.15492677688598633,
+      "learning_rate": 0.0001982980489829805,
+      "loss": 0.8179,
+      "step": 191
+    },
+    {
+      "epoch": 0.11590703290069423,
+      "grad_norm": 0.33444538712501526,
+      "learning_rate": 0.00019825653798256538,
+      "loss": 0.8264,
+      "step": 192
+    },
+    {
+      "epoch": 0.11651071536371868,
+      "grad_norm": 0.15822017192840576,
+      "learning_rate": 0.00019821502698215028,
+      "loss": 0.7953,
+      "step": 193
+    },
+    {
+      "epoch": 0.11711439782674313,
+      "grad_norm": 0.19879087805747986,
+      "learning_rate": 0.00019817351598173518,
+      "loss": 0.8023,
+      "step": 194
+    },
+    {
+      "epoch": 0.11771808028976759,
+      "grad_norm": 0.155591681599617,
+      "learning_rate": 0.00019813200498132005,
+      "loss": 0.6876,
+      "step": 195
+    },
+    {
+      "epoch": 0.11832176275279203,
+      "grad_norm": 0.19856096804141998,
+      "learning_rate": 0.00019809049398090495,
+      "loss": 0.7227,
+      "step": 196
+    },
+    {
+      "epoch": 0.11892544521581648,
+      "grad_norm": 0.14677810668945312,
+      "learning_rate": 0.00019804898298048985,
+      "loss": 0.729,
+      "step": 197
+    },
+    {
+      "epoch": 0.11952912767884093,
+      "grad_norm": 3.824247121810913,
+      "learning_rate": 0.00019800747198007472,
+      "loss": 0.7045,
+      "step": 198
+    },
+    {
+      "epoch": 0.12013281014186537,
+      "grad_norm": 1.2843711376190186,
+      "learning_rate": 0.00019796596097965962,
+      "loss": 0.681,
+      "step": 199
+    },
+    {
+      "epoch": 0.12073649260488983,
+      "grad_norm": 0.15802475810050964,
+      "learning_rate": 0.0001979244499792445,
+      "loss": 0.54,
+      "step": 200
+    },
+    {
+      "epoch": 0.12134017506791428,
+      "grad_norm": 0.08058023452758789,
+      "learning_rate": 0.0001978829389788294,
+      "loss": 1.2773,
+      "step": 201
+    },
+    {
+      "epoch": 0.12194385753093873,
+      "grad_norm": 0.09442666172981262,
+      "learning_rate": 0.0001978414279784143,
+      "loss": 1.0519,
+      "step": 202
+    },
+    {
+      "epoch": 0.12254753999396317,
+      "grad_norm": 0.11098414659500122,
+      "learning_rate": 0.00019779991697799917,
+      "loss": 0.9809,
+      "step": 203
+    },
+    {
+      "epoch": 0.12315122245698762,
+      "grad_norm": 0.08391015231609344,
+      "learning_rate": 0.00019775840597758407,
+      "loss": 0.7483,
+      "step": 204
+    },
+    {
+      "epoch": 0.12375490492001208,
+      "grad_norm": 0.09060636162757874,
+      "learning_rate": 0.00019771689497716897,
+      "loss": 0.9001,
+      "step": 205
+    },
+    {
+      "epoch": 0.12435858738303653,
+      "grad_norm": 0.08621667325496674,
+      "learning_rate": 0.00019767538397675384,
+      "loss": 1.0983,
+      "step": 206
+    },
+    {
+      "epoch": 0.12496226984606097,
+      "grad_norm": 0.07971488684415817,
+      "learning_rate": 0.00019763387297633874,
+      "loss": 0.8414,
+      "step": 207
+    },
+    {
+      "epoch": 0.12556595230908543,
+      "grad_norm": 0.07815425097942352,
+      "learning_rate": 0.00019759236197592364,
+      "loss": 0.8872,
+      "step": 208
+    },
+    {
+      "epoch": 0.12616963477210988,
+      "grad_norm": 0.0847320556640625,
+      "learning_rate": 0.00019755085097550852,
+      "loss": 0.8398,
+      "step": 209
+    },
+    {
+      "epoch": 0.12677331723513433,
+      "grad_norm": 0.08897458761930466,
+      "learning_rate": 0.0001975093399750934,
+      "loss": 0.7865,
+      "step": 210
+    },
+    {
+      "epoch": 0.12737699969815877,
+      "grad_norm": 0.0703200101852417,
+      "learning_rate": 0.00019746782897467832,
+      "loss": 0.8256,
+      "step": 211
+    },
+    {
+      "epoch": 0.12798068216118322,
+      "grad_norm": 0.07879285514354706,
+      "learning_rate": 0.0001974263179742632,
+      "loss": 0.9032,
+      "step": 212
+    },
+    {
+      "epoch": 0.12858436462420766,
+      "grad_norm": 0.07276301831007004,
+      "learning_rate": 0.00019738480697384806,
+      "loss": 0.9857,
+      "step": 213
+    },
+    {
+      "epoch": 0.1291880470872321,
+      "grad_norm": 0.09440230578184128,
+      "learning_rate": 0.00019734329597343296,
+      "loss": 0.9156,
+      "step": 214
+    },
+    {
+      "epoch": 0.12979172955025656,
+      "grad_norm": 0.1108328178524971,
+      "learning_rate": 0.00019730178497301786,
+      "loss": 0.9671,
+      "step": 215
+    },
+    {
+      "epoch": 0.130395412013281,
+      "grad_norm": 0.07690040022134781,
+      "learning_rate": 0.00019726027397260273,
+      "loss": 0.9945,
+      "step": 216
+    },
+    {
+      "epoch": 0.13099909447630545,
+      "grad_norm": 0.06875025480985641,
+      "learning_rate": 0.00019721876297218763,
+      "loss": 0.911,
+      "step": 217
+    },
+    {
+      "epoch": 0.13160277693932992,
+      "grad_norm": 0.09118712693452835,
+      "learning_rate": 0.00019717725197177253,
+      "loss": 1.1363,
+      "step": 218
+    },
+    {
+      "epoch": 0.13220645940235437,
+      "grad_norm": 0.07356920093297958,
+      "learning_rate": 0.00019713574097135743,
+      "loss": 1.14,
+      "step": 219
+    },
+    {
+      "epoch": 0.13281014186537882,
+      "grad_norm": 0.06750470399856567,
+      "learning_rate": 0.0001970942299709423,
+      "loss": 0.9591,
+      "step": 220
+    },
+    {
+      "epoch": 0.13341382432840326,
+      "grad_norm": 0.11778965592384338,
+      "learning_rate": 0.0001970527189705272,
+      "loss": 0.9048,
+      "step": 221
+    },
+    {
+      "epoch": 0.1340175067914277,
+      "grad_norm": 0.07331351190805435,
+      "learning_rate": 0.0001970112079701121,
+      "loss": 0.8792,
+      "step": 222
+    },
+    {
+      "epoch": 0.13462118925445216,
+      "grad_norm": 0.08289069682359695,
+      "learning_rate": 0.00019696969696969698,
+      "loss": 1.1292,
+      "step": 223
+    },
+    {
+      "epoch": 0.1352248717174766,
+      "grad_norm": 0.07485181093215942,
+      "learning_rate": 0.00019692818596928185,
+      "loss": 1.013,
+      "step": 224
+    },
+    {
+      "epoch": 0.13582855418050105,
+      "grad_norm": 0.08499836176633835,
+      "learning_rate": 0.00019688667496886678,
+      "loss": 0.9177,
+      "step": 225
+    },
+    {
+      "epoch": 0.1364322366435255,
+      "grad_norm": 0.06567627936601639,
+      "learning_rate": 0.00019684516396845165,
+      "loss": 0.9016,
+      "step": 226
+    },
+    {
+      "epoch": 0.13703591910654994,
+      "grad_norm": 0.07324650138616562,
+      "learning_rate": 0.00019680365296803653,
+      "loss": 0.9649,
+      "step": 227
+    },
+    {
+      "epoch": 0.13763960156957442,
+      "grad_norm": 0.06484885513782501,
+      "learning_rate": 0.00019676214196762143,
+      "loss": 0.824,
+      "step": 228
+    },
+    {
+      "epoch": 0.13824328403259886,
+      "grad_norm": 0.06534811854362488,
+      "learning_rate": 0.00019672063096720633,
+      "loss": 0.7969,
+      "step": 229
+    },
+    {
+      "epoch": 0.1388469664956233,
+      "grad_norm": 0.06635700166225433,
+      "learning_rate": 0.0001966791199667912,
+      "loss": 0.9041,
+      "step": 230
+    },
+    {
+      "epoch": 0.13945064895864775,
+      "grad_norm": 0.06364124268293381,
+      "learning_rate": 0.0001966376089663761,
+      "loss": 0.8935,
+      "step": 231
+    },
+    {
+      "epoch": 0.1400543314216722,
+      "grad_norm": 0.0640692487359047,
+      "learning_rate": 0.000196596097965961,
+      "loss": 0.9025,
+      "step": 232
+    },
+    {
+      "epoch": 0.14065801388469665,
+      "grad_norm": 0.07015056163072586,
+      "learning_rate": 0.00019655458696554587,
+      "loss": 0.8065,
+      "step": 233
+    },
+    {
+      "epoch": 0.1412616963477211,
+      "grad_norm": 0.07654477655887604,
+      "learning_rate": 0.00019651307596513077,
+      "loss": 0.8408,
+      "step": 234
+    },
+    {
+      "epoch": 0.14186537881074554,
+      "grad_norm": 0.06782008707523346,
+      "learning_rate": 0.00019647156496471567,
+      "loss": 0.7275,
+      "step": 235
+    },
+    {
+      "epoch": 0.14246906127377,
+      "grad_norm": 0.07276429980993271,
+      "learning_rate": 0.00019643005396430055,
+      "loss": 0.7955,
+      "step": 236
+    },
+    {
+      "epoch": 0.14307274373679443,
+      "grad_norm": 0.08035031706094742,
+      "learning_rate": 0.00019638854296388545,
+      "loss": 0.7074,
+      "step": 237
+    },
+    {
+      "epoch": 0.1436764261998189,
+      "grad_norm": 0.10624828189611435,
+      "learning_rate": 0.00019634703196347032,
+      "loss": 0.8799,
+      "step": 238
+    },
+    {
+      "epoch": 0.14428010866284335,
+      "grad_norm": 0.0777585506439209,
+      "learning_rate": 0.00019630552096305522,
+      "loss": 0.7528,
+      "step": 239
+    },
+    {
+      "epoch": 0.1448837911258678,
+      "grad_norm": 0.0803024098277092,
+      "learning_rate": 0.00019626400996264012,
+      "loss": 0.7702,
+      "step": 240
+    },
+    {
+      "epoch": 0.14548747358889225,
+      "grad_norm": 0.0814545750617981,
+      "learning_rate": 0.000196222498962225,
+      "loss": 0.78,
+      "step": 241
+    },
+    {
+      "epoch": 0.1460911560519167,
+      "grad_norm": 0.0940927267074585,
+      "learning_rate": 0.0001961809879618099,
+      "loss": 0.7294,
+      "step": 242
+    },
+    {
+      "epoch": 0.14669483851494114,
+      "grad_norm": 0.08877824991941452,
+      "learning_rate": 0.0001961394769613948,
+      "loss": 0.7663,
+      "step": 243
+    },
+    {
+      "epoch": 0.14729852097796559,
+      "grad_norm": 0.09321099519729614,
+      "learning_rate": 0.00019609796596097966,
+      "loss": 0.7524,
+      "step": 244
+    },
+    {
+      "epoch": 0.14790220344099003,
+      "grad_norm": 0.09710489213466644,
+      "learning_rate": 0.00019605645496056454,
+      "loss": 0.8147,
+      "step": 245
+    },
+    {
+      "epoch": 0.14850588590401448,
+      "grad_norm": 0.10399672389030457,
+      "learning_rate": 0.00019601494396014946,
+      "loss": 0.7327,
+      "step": 246
+    },
+    {
+      "epoch": 0.14910956836703892,
+      "grad_norm": 0.11608180403709412,
+      "learning_rate": 0.00019597343295973434,
+      "loss": 0.7464,
+      "step": 247
+    },
+    {
+      "epoch": 0.1497132508300634,
+      "grad_norm": 0.10949523001909256,
+      "learning_rate": 0.0001959319219593192,
+      "loss": 0.6295,
+      "step": 248
+    },
+    {
+      "epoch": 0.15031693329308785,
+      "grad_norm": 0.10902351886034012,
+      "learning_rate": 0.00019589041095890414,
+      "loss": 0.6048,
+      "step": 249
+    },
+    {
+      "epoch": 0.1509206157561123,
+      "grad_norm": 0.11830253154039383,
+      "learning_rate": 0.000195848899958489,
+      "loss": 0.4881,
+      "step": 250
+    },
+    {
+      "epoch": 0.15152429821913674,
+      "grad_norm": 0.06550572067499161,
+      "learning_rate": 0.00019580738895807388,
+      "loss": 0.8051,
+      "step": 251
+    },
+    {
+      "epoch": 0.15212798068216118,
+      "grad_norm": 0.08449156582355499,
+      "learning_rate": 0.00019576587795765878,
+      "loss": 1.0096,
+      "step": 252
+    },
+    {
+      "epoch": 0.15273166314518563,
+      "grad_norm": 0.07215554267168045,
+      "learning_rate": 0.00019572436695724368,
+      "loss": 0.7212,
+      "step": 253
+    },
+    {
+      "epoch": 0.15333534560821008,
+      "grad_norm": 0.0707031637430191,
+      "learning_rate": 0.00019568285595682856,
+      "loss": 0.8708,
+      "step": 254
+    },
+    {
+      "epoch": 0.15393902807123452,
+      "grad_norm": 0.22556833922863007,
+      "learning_rate": 0.00019564134495641346,
+      "loss": 0.9448,
+      "step": 255
+    },
+    {
+      "epoch": 0.15454271053425897,
+      "grad_norm": 0.06614769995212555,
+      "learning_rate": 0.00019559983395599836,
+      "loss": 1.2733,
+      "step": 256
+    },
+    {
+      "epoch": 0.15514639299728342,
+      "grad_norm": 0.14129576086997986,
+      "learning_rate": 0.00019555832295558323,
+      "loss": 0.9814,
+      "step": 257
+    },
+    {
+      "epoch": 0.1557500754603079,
+      "grad_norm": 0.061934180557727814,
+      "learning_rate": 0.00019551681195516813,
+      "loss": 0.8052,
+      "step": 258
+    },
+    {
+      "epoch": 0.15635375792333234,
+      "grad_norm": 0.09368140995502472,
+      "learning_rate": 0.000195475300954753,
+      "loss": 0.9107,
+      "step": 259
+    },
+    {
+      "epoch": 0.15695744038635678,
+      "grad_norm": 0.066087506711483,
+      "learning_rate": 0.0001954337899543379,
+      "loss": 0.7128,
+      "step": 260
+    },
+    {
+      "epoch": 0.15756112284938123,
+      "grad_norm": 0.06857267022132874,
+      "learning_rate": 0.0001953922789539228,
+      "loss": 0.8081,
+      "step": 261
+    },
+    {
+      "epoch": 0.15816480531240568,
+      "grad_norm": 0.06954636424779892,
+      "learning_rate": 0.00019535076795350768,
+      "loss": 0.8834,
+      "step": 262
+    },
+    {
+      "epoch": 0.15876848777543012,
+      "grad_norm": 0.07501845061779022,
+      "learning_rate": 0.0001953092569530926,
+      "loss": 0.7618,
+      "step": 263
+    },
+    {
+      "epoch": 0.15937217023845457,
+      "grad_norm": 0.06725212186574936,
+      "learning_rate": 0.00019526774595267748,
+      "loss": 0.852,
+      "step": 264
+    },
+    {
+      "epoch": 0.15997585270147902,
+      "grad_norm": 0.06601674854755402,
+      "learning_rate": 0.00019522623495226235,
+      "loss": 0.8241,
+      "step": 265
+    },
+    {
+      "epoch": 0.16057953516450346,
+      "grad_norm": 0.0759337842464447,
+      "learning_rate": 0.00019518472395184725,
+      "loss": 1.1918,
+      "step": 266
+    },
+    {
+      "epoch": 0.1611832176275279,
+      "grad_norm": 0.0703561007976532,
+      "learning_rate": 0.00019514321295143215,
+      "loss": 0.8613,
+      "step": 267
+    },
+    {
+      "epoch": 0.16178690009055238,
+      "grad_norm": 0.07018351554870605,
+      "learning_rate": 0.00019510170195101702,
+      "loss": 1.0508,
+      "step": 268
+    },
+    {
+      "epoch": 0.16239058255357683,
+      "grad_norm": 0.06850980967283249,
+      "learning_rate": 0.00019506019095060192,
+      "loss": 0.9527,
+      "step": 269
+    },
+    {
+      "epoch": 0.16299426501660128,
+      "grad_norm": 0.06927502900362015,
+      "learning_rate": 0.00019501867995018682,
+      "loss": 0.9985,
+      "step": 270
+    },
+    {
+      "epoch": 0.16359794747962572,
+      "grad_norm": 0.07486286759376526,
+      "learning_rate": 0.0001949771689497717,
+      "loss": 0.9889,
+      "step": 271
+    },
+    {
+      "epoch": 0.16420162994265017,
+      "grad_norm": 0.06840648502111435,
+      "learning_rate": 0.0001949356579493566,
+      "loss": 0.902,
+      "step": 272
+    },
+    {
+      "epoch": 0.16480531240567461,
+      "grad_norm": 0.06660640239715576,
+      "learning_rate": 0.00019489414694894147,
+      "loss": 0.9774,
+      "step": 273
+    },
+    {
+      "epoch": 0.16540899486869906,
+      "grad_norm": 0.06089378520846367,
+      "learning_rate": 0.00019485263594852637,
+      "loss": 1.0334,
+      "step": 274
+    },
+    {
+      "epoch": 0.1660126773317235,
+      "grad_norm": 0.06978226453065872,
+      "learning_rate": 0.00019481112494811127,
+      "loss": 0.8932,
+      "step": 275
+    },
+    {
+      "epoch": 0.16661635979474795,
+      "grad_norm": 0.06877858191728592,
+      "learning_rate": 0.00019476961394769614,
+      "loss": 0.91,
+      "step": 276
+    },
+    {
+      "epoch": 0.1672200422577724,
+      "grad_norm": 0.06825589388608932,
+      "learning_rate": 0.00019472810294728104,
+      "loss": 0.9188,
+      "step": 277
+    },
+    {
+      "epoch": 0.16782372472079687,
+      "grad_norm": 0.06272375583648682,
+      "learning_rate": 0.00019468659194686594,
+      "loss": 0.8665,
+      "step": 278
+    },
+    {
+      "epoch": 0.16842740718382132,
+      "grad_norm": 0.07434337586164474,
+      "learning_rate": 0.0001946450809464508,
+      "loss": 0.9386,
+      "step": 279
+    },
+    {
+      "epoch": 0.16903108964684577,
+      "grad_norm": 0.062427766621112823,
+      "learning_rate": 0.0001946035699460357,
+      "loss": 0.8038,
+      "step": 280
+    },
+    {
+      "epoch": 0.1696347721098702,
+      "grad_norm": 0.07670903950929642,
+      "learning_rate": 0.0001945620589456206,
+      "loss": 0.8494,
+      "step": 281
+    },
+    {
+      "epoch": 0.17023845457289466,
+      "grad_norm": 0.06715535372495651,
+      "learning_rate": 0.00019452054794520549,
+      "loss": 1.0171,
+      "step": 282
+    },
+    {
+      "epoch": 0.1708421370359191,
+      "grad_norm": 0.0629987046122551,
+      "learning_rate": 0.00019447903694479036,
+      "loss": 0.8715,
+      "step": 283
+    },
+    {
+      "epoch": 0.17144581949894355,
+      "grad_norm": 0.0751439705491066,
+      "learning_rate": 0.00019443752594437529,
+      "loss": 0.8137,
+      "step": 284
+    },
+    {
+      "epoch": 0.172049501961968,
+      "grad_norm": 0.07195259630680084,
+      "learning_rate": 0.00019439601494396016,
+      "loss": 0.8276,
+      "step": 285
+    },
+    {
+      "epoch": 0.17265318442499245,
+      "grad_norm": 0.06936553120613098,
+      "learning_rate": 0.00019435450394354503,
+      "loss": 0.7662,
+      "step": 286
+    },
+    {
+      "epoch": 0.1732568668880169,
+      "grad_norm": 0.07795591652393341,
+      "learning_rate": 0.00019431299294312993,
+      "loss": 0.7729,
+      "step": 287
+    },
+    {
+      "epoch": 0.17386054935104137,
+      "grad_norm": 0.07397985458374023,
+      "learning_rate": 0.00019427148194271483,
+      "loss": 0.7924,
+      "step": 288
+    },
+    {
+      "epoch": 0.1744642318140658,
+      "grad_norm": 0.08324669301509857,
+      "learning_rate": 0.0001942299709422997,
+      "loss": 0.8163,
+      "step": 289
+    },
+    {
+      "epoch": 0.17506791427709026,
+      "grad_norm": 0.08435431867837906,
+      "learning_rate": 0.0001941884599418846,
+      "loss": 0.7395,
+      "step": 290
+    },
+    {
+      "epoch": 0.1756715967401147,
+      "grad_norm": 0.08156009763479233,
+      "learning_rate": 0.0001941469489414695,
+      "loss": 0.7299,
+      "step": 291
+    },
+    {
+      "epoch": 0.17627527920313915,
+      "grad_norm": 0.08714080601930618,
+      "learning_rate": 0.00019410543794105438,
+      "loss": 0.7918,
+      "step": 292
+    },
+    {
+      "epoch": 0.1768789616661636,
+      "grad_norm": 0.09345049411058426,
+      "learning_rate": 0.00019406392694063928,
+      "loss": 0.7841,
+      "step": 293
+    },
+    {
+      "epoch": 0.17748264412918804,
+      "grad_norm": 0.09327434748411179,
+      "learning_rate": 0.00019402241594022418,
+      "loss": 0.7073,
+      "step": 294
+    },
+    {
+      "epoch": 0.1780863265922125,
+      "grad_norm": 0.09284470230340958,
+      "learning_rate": 0.00019398090493980905,
+      "loss": 0.6427,
+      "step": 295
+    },
+    {
+      "epoch": 0.17869000905523694,
+      "grad_norm": 0.09828530997037888,
+      "learning_rate": 0.00019393939393939395,
+      "loss": 0.6953,
+      "step": 296
+    },
+    {
+      "epoch": 0.17929369151826138,
+      "grad_norm": 0.10501807928085327,
+      "learning_rate": 0.00019389788293897882,
+      "loss": 0.7286,
+      "step": 297
+    },
+    {
+      "epoch": 0.17989737398128586,
+      "grad_norm": 0.14845532178878784,
+      "learning_rate": 0.00019385637193856372,
+      "loss": 0.6602,
+      "step": 298
+    },
+    {
+      "epoch": 0.1805010564443103,
+      "grad_norm": 0.1208406463265419,
+      "learning_rate": 0.00019381486093814862,
+      "loss": 0.5862,
+      "step": 299
+    },
+    {
+      "epoch": 0.18110473890733475,
+      "grad_norm": 0.127839133143425,
+      "learning_rate": 0.0001937733499377335,
+      "loss": 0.4804,
+      "step": 300
+    },
+    {
+      "epoch": 0.1817084213703592,
+      "grad_norm": 0.07252588123083115,
+      "learning_rate": 0.0001937318389373184,
+      "loss": 0.8918,
+      "step": 301
+    },
+    {
+      "epoch": 0.18231210383338364,
+      "grad_norm": 0.07854969054460526,
+      "learning_rate": 0.0001936903279369033,
+      "loss": 0.9295,
+      "step": 302
+    },
+    {
+      "epoch": 0.1829157862964081,
+      "grad_norm": 0.07926355302333832,
+      "learning_rate": 0.00019364881693648817,
+      "loss": 1.0768,
+      "step": 303
+    },
+    {
+      "epoch": 0.18351946875943254,
+      "grad_norm": 0.1117272675037384,
+      "learning_rate": 0.00019360730593607307,
+      "loss": 0.89,
+      "step": 304
+    },
+    {
+      "epoch": 0.18412315122245698,
+      "grad_norm": 0.09249380975961685,
+      "learning_rate": 0.00019356579493565797,
+      "loss": 0.9915,
+      "step": 305
+    },
+    {
+      "epoch": 0.18472683368548143,
+      "grad_norm": 0.0759844034910202,
+      "learning_rate": 0.00019352428393524284,
+      "loss": 0.9705,
+      "step": 306
+    },
+    {
+      "epoch": 0.18533051614850587,
+      "grad_norm": 0.06819566339254379,
+      "learning_rate": 0.00019348277293482774,
+      "loss": 0.9801,
+      "step": 307
+    },
+    {
+      "epoch": 0.18593419861153032,
+      "grad_norm": 0.07146774232387543,
+      "learning_rate": 0.00019344126193441264,
+      "loss": 0.8923,
+      "step": 308
+    },
+    {
+      "epoch": 0.1865378810745548,
+      "grad_norm": 0.06504259258508682,
+      "learning_rate": 0.00019339975093399752,
+      "loss": 0.8941,
+      "step": 309
+    },
+    {
+      "epoch": 0.18714156353757924,
+      "grad_norm": 0.06857950240373611,
+      "learning_rate": 0.00019335823993358242,
+      "loss": 0.9558,
+      "step": 310
+    },
+    {
+      "epoch": 0.1877452460006037,
+      "grad_norm": 0.06270836293697357,
+      "learning_rate": 0.0001933167289331673,
+      "loss": 0.9532,
+      "step": 311
+    },
+    {
+      "epoch": 0.18834892846362813,
+      "grad_norm": 0.13885007798671722,
+      "learning_rate": 0.0001932752179327522,
+      "loss": 0.8591,
+      "step": 312
+    },
+    {
+      "epoch": 0.18895261092665258,
+      "grad_norm": 0.07498139888048172,
+      "learning_rate": 0.0001932337069323371,
+      "loss": 0.9812,
+      "step": 313
+    },
+    {
+      "epoch": 0.18955629338967703,
+      "grad_norm": 0.0773036777973175,
+      "learning_rate": 0.00019319219593192196,
+      "loss": 0.8807,
+      "step": 314
+    },
+    {
+      "epoch": 0.19015997585270147,
+      "grad_norm": 0.07549306750297546,
+      "learning_rate": 0.00019315068493150686,
+      "loss": 1.0929,
+      "step": 315
+    },
+    {
+      "epoch": 0.19076365831572592,
+      "grad_norm": 0.06931062042713165,
+      "learning_rate": 0.00019310917393109176,
+      "loss": 0.9293,
+      "step": 316
+    },
+    {
+      "epoch": 0.19136734077875037,
+      "grad_norm": 0.06612774729728699,
+      "learning_rate": 0.00019306766293067664,
+      "loss": 1.1528,
+      "step": 317
+    },
+    {
+      "epoch": 0.1919710232417748,
+      "grad_norm": 0.07355131208896637,
+      "learning_rate": 0.00019302615193026154,
+      "loss": 1.4277,
+      "step": 318
+    },
+    {
+      "epoch": 0.1925747057047993,
+      "grad_norm": 0.07465571165084839,
+      "learning_rate": 0.00019298464092984644,
+      "loss": 0.8799,
+      "step": 319
+    },
+    {
+      "epoch": 0.19317838816782373,
+      "grad_norm": 0.07097172737121582,
+      "learning_rate": 0.0001929431299294313,
+      "loss": 0.7634,
+      "step": 320
+    },
+    {
+      "epoch": 0.19378207063084818,
+      "grad_norm": 0.07077208161354065,
+      "learning_rate": 0.00019290161892901618,
+      "loss": 0.9721,
+      "step": 321
+    },
+    {
+      "epoch": 0.19438575309387263,
+      "grad_norm": 0.06071571260690689,
+      "learning_rate": 0.0001928601079286011,
+      "loss": 0.8238,
+      "step": 322
+    },
+    {
+      "epoch": 0.19498943555689707,
+      "grad_norm": 0.07852339744567871,
+      "learning_rate": 0.00019281859692818598,
+      "loss": 0.7897,
+      "step": 323
+    },
+    {
+      "epoch": 0.19559311801992152,
+      "grad_norm": 0.06762294471263885,
+      "learning_rate": 0.00019277708592777085,
+      "loss": 0.9692,
+      "step": 324
+    },
+    {
+      "epoch": 0.19619680048294597,
+      "grad_norm": 0.06589538604021072,
+      "learning_rate": 0.00019273557492735575,
+      "loss": 0.8701,
+      "step": 325
+    },
+    {
+      "epoch": 0.1968004829459704,
+      "grad_norm": 0.07512197643518448,
+      "learning_rate": 0.00019269406392694065,
+      "loss": 0.8921,
+      "step": 326
+    },
+    {
+      "epoch": 0.19740416540899486,
+      "grad_norm": 0.06501419097185135,
+      "learning_rate": 0.00019265255292652553,
+      "loss": 0.8559,
+      "step": 327
+    },
+    {
+      "epoch": 0.1980078478720193,
+      "grad_norm": 0.06205839663743973,
+      "learning_rate": 0.00019261104192611043,
+      "loss": 0.9534,
+      "step": 328
+    },
+    {
+      "epoch": 0.19861153033504378,
+      "grad_norm": 0.06113965064287186,
+      "learning_rate": 0.00019256953092569533,
+      "loss": 0.7965,
+      "step": 329
+    },
+    {
+      "epoch": 0.19921521279806823,
+      "grad_norm": 0.06127781420946121,
+      "learning_rate": 0.0001925280199252802,
+      "loss": 0.8429,
+      "step": 330
+    },
+    {
+      "epoch": 0.19981889526109267,
+      "grad_norm": 0.07560814172029495,
+      "learning_rate": 0.0001924865089248651,
+      "loss": 0.8128,
+      "step": 331
+    },
+    {
+      "epoch": 0.20042257772411712,
+      "grad_norm": 0.06620073318481445,
+      "learning_rate": 0.00019244499792445,
+      "loss": 0.958,
+      "step": 332
+    },
+    {
+      "epoch": 0.20102626018714156,
+      "grad_norm": 0.0671965479850769,
+      "learning_rate": 0.00019240348692403487,
+      "loss": 0.8134,
+      "step": 333
+    },
+    {
+      "epoch": 0.201629942650166,
+      "grad_norm": 0.0731448382139206,
+      "learning_rate": 0.00019236197592361977,
+      "loss": 0.7264,
+      "step": 334
+    },
+    {
+      "epoch": 0.20223362511319046,
+      "grad_norm": 0.06810946017503738,
+      "learning_rate": 0.00019232046492320465,
+      "loss": 0.7509,
+      "step": 335
+    },
+    {
+      "epoch": 0.2028373075762149,
+      "grad_norm": 0.08012348413467407,
+      "learning_rate": 0.00019227895392278955,
+      "loss": 0.8237,
+      "step": 336
+    },
+    {
+      "epoch": 0.20344099003923935,
+      "grad_norm": 0.07328704744577408,
+      "learning_rate": 0.00019223744292237445,
+      "loss": 0.7719,
+      "step": 337
+    },
+    {
+      "epoch": 0.2040446725022638,
+      "grad_norm": 0.07838919013738632,
+      "learning_rate": 0.00019219593192195932,
+      "loss": 0.8404,
+      "step": 338
+    },
+    {
+      "epoch": 0.20464835496528827,
+      "grad_norm": 0.07977066934108734,
+      "learning_rate": 0.00019215442092154422,
+      "loss": 0.7614,
+      "step": 339
+    },
+    {
+      "epoch": 0.20525203742831272,
+      "grad_norm": 0.07425908744335175,
+      "learning_rate": 0.00019211290992112912,
+      "loss": 0.7439,
+      "step": 340
+    },
+    {
+      "epoch": 0.20585571989133716,
+      "grad_norm": 0.08555571734905243,
+      "learning_rate": 0.000192071398920714,
+      "loss": 0.7711,
+      "step": 341
+    },
+    {
+      "epoch": 0.2064594023543616,
+      "grad_norm": 0.0798811987042427,
+      "learning_rate": 0.0001920298879202989,
+      "loss": 0.7246,
+      "step": 342
+    },
+    {
+      "epoch": 0.20706308481738606,
+      "grad_norm": 0.09403866529464722,
+      "learning_rate": 0.0001919883769198838,
+      "loss": 0.7171,
+      "step": 343
+    },
+    {
+      "epoch": 0.2076667672804105,
+      "grad_norm": 0.09114021807909012,
+      "learning_rate": 0.00019194686591946866,
+      "loss": 0.7605,
+      "step": 344
+    },
+    {
+      "epoch": 0.20827044974343495,
+      "grad_norm": 0.1013076901435852,
+      "learning_rate": 0.00019190535491905354,
+      "loss": 0.7583,
+      "step": 345
+    },
+    {
+      "epoch": 0.2088741322064594,
+      "grad_norm": 0.10122162848711014,
+      "learning_rate": 0.00019186384391863847,
+      "loss": 0.7477,
+      "step": 346
+    },
+    {
+      "epoch": 0.20947781466948384,
+      "grad_norm": 0.09927545487880707,
+      "learning_rate": 0.00019182233291822334,
+      "loss": 0.6728,
+      "step": 347
+    },
+    {
+      "epoch": 0.2100814971325083,
+      "grad_norm": 0.12509341537952423,
+      "learning_rate": 0.0001917808219178082,
+      "loss": 0.6543,
+      "step": 348
+    },
+    {
+      "epoch": 0.21068517959553276,
+      "grad_norm": 0.11812290549278259,
+      "learning_rate": 0.0001917393109173931,
+      "loss": 0.6701,
+      "step": 349
+    },
+    {
+      "epoch": 0.2112888620585572,
+      "grad_norm": 0.12166016548871994,
+      "learning_rate": 0.000191697799916978,
+      "loss": 0.5261,
+      "step": 350
+    },
+    {
+      "epoch": 0.21189254452158165,
+      "grad_norm": 0.0700993686914444,
+      "learning_rate": 0.0001916562889165629,
+      "loss": 1.2022,
+      "step": 351
+    },
+    {
+      "epoch": 0.2124962269846061,
+      "grad_norm": 0.08300930261611938,
+      "learning_rate": 0.00019161477791614778,
+      "loss": 0.9249,
+      "step": 352
+    },
+    {
+      "epoch": 0.21309990944763055,
+      "grad_norm": 0.06463124603033066,
+      "learning_rate": 0.00019157326691573268,
+      "loss": 0.8147,
+      "step": 353
+    },
+    {
+      "epoch": 0.213703591910655,
+      "grad_norm": 0.06934931129217148,
+      "learning_rate": 0.00019153175591531758,
+      "loss": 0.9904,
+      "step": 354
+    },
+    {
+      "epoch": 0.21430727437367944,
+      "grad_norm": 0.06555134057998657,
+      "learning_rate": 0.00019149024491490246,
+      "loss": 0.8825,
+      "step": 355
+    },
+    {
+      "epoch": 0.2149109568367039,
+      "grad_norm": 0.07276731729507446,
+      "learning_rate": 0.00019144873391448736,
+      "loss": 1.0806,
+      "step": 356
+    },
+    {
+      "epoch": 0.21551463929972833,
+      "grad_norm": 0.06886199861764908,
+      "learning_rate": 0.00019140722291407226,
+      "loss": 0.9756,
+      "step": 357
+    },
+    {
+      "epoch": 0.21611832176275278,
+      "grad_norm": 0.07671529054641724,
+      "learning_rate": 0.00019136571191365713,
+      "loss": 0.9871,
+      "step": 358
+    },
+    {
+      "epoch": 0.21672200422577725,
+      "grad_norm": 0.08558958023786545,
+      "learning_rate": 0.000191324200913242,
+      "loss": 0.9553,
+      "step": 359
+    },
+    {
+      "epoch": 0.2173256866888017,
+      "grad_norm": 0.07093027234077454,
+      "learning_rate": 0.00019128268991282693,
+      "loss": 0.8247,
+      "step": 360
+    },
+    {
+      "epoch": 0.21792936915182615,
+      "grad_norm": 0.07704441994428635,
+      "learning_rate": 0.0001912411789124118,
+      "loss": 1.0793,
+      "step": 361
+    },
+    {
+      "epoch": 0.2185330516148506,
+      "grad_norm": 0.06594787538051605,
+      "learning_rate": 0.00019119966791199668,
+      "loss": 0.866,
+      "step": 362
+    },
+    {
+      "epoch": 0.21913673407787504,
+      "grad_norm": 0.07074205577373505,
+      "learning_rate": 0.00019115815691158158,
+      "loss": 0.9088,
+      "step": 363
+    },
+    {
+      "epoch": 0.21974041654089949,
+      "grad_norm": 0.06734868884086609,
+      "learning_rate": 0.00019111664591116648,
+      "loss": 0.8861,
+      "step": 364
+    },
+    {
+      "epoch": 0.22034409900392393,
+      "grad_norm": 0.07831276208162308,
+      "learning_rate": 0.00019107513491075135,
+      "loss": 0.939,
+      "step": 365
+    },
+    {
+      "epoch": 0.22094778146694838,
+      "grad_norm": 0.06146768108010292,
+      "learning_rate": 0.00019103362391033625,
+      "loss": 0.9476,
+      "step": 366
+    },
+    {
+      "epoch": 0.22155146392997283,
+      "grad_norm": 0.062476493418216705,
+      "learning_rate": 0.00019099211290992115,
+      "loss": 0.821,
+      "step": 367
+    },
+    {
+      "epoch": 0.22215514639299727,
+      "grad_norm": 0.07490051537752151,
+      "learning_rate": 0.00019095060190950602,
+      "loss": 0.9016,
+      "step": 368
+    },
+    {
+      "epoch": 0.22275882885602175,
+      "grad_norm": 0.07467661798000336,
+      "learning_rate": 0.00019090909090909092,
+      "loss": 0.786,
+      "step": 369
+    },
+    {
+      "epoch": 0.2233625113190462,
+      "grad_norm": 0.0667489618062973,
+      "learning_rate": 0.00019086757990867582,
+      "loss": 0.7985,
+      "step": 370
+    },
+    {
+      "epoch": 0.22396619378207064,
+      "grad_norm": 0.06847196817398071,
+      "learning_rate": 0.0001908260689082607,
+      "loss": 0.7912,
+      "step": 371
+    },
+    {
+      "epoch": 0.22456987624509508,
+      "grad_norm": 0.06785493344068527,
+      "learning_rate": 0.0001907845579078456,
+      "loss": 0.9787,
+      "step": 372
+    },
+    {
+      "epoch": 0.22517355870811953,
+      "grad_norm": 0.06378093361854553,
+      "learning_rate": 0.00019074304690743047,
+      "loss": 1.0737,
+      "step": 373
+    },
+    {
+      "epoch": 0.22577724117114398,
+      "grad_norm": 0.07684706896543503,
+      "learning_rate": 0.00019070153590701537,
+      "loss": 0.8541,
+      "step": 374
+    },
+    {
+      "epoch": 0.22638092363416842,
+      "grad_norm": 0.06430470943450928,
+      "learning_rate": 0.00019066002490660027,
+      "loss": 0.8404,
+      "step": 375
+    },
+    {
+      "epoch": 0.22698460609719287,
+      "grad_norm": 0.062282588332891464,
+      "learning_rate": 0.00019061851390618514,
+      "loss": 0.8361,
+      "step": 376
+    },
+    {
+      "epoch": 0.22758828856021732,
+      "grad_norm": 0.06882327795028687,
+      "learning_rate": 0.00019057700290577004,
+      "loss": 0.7329,
+      "step": 377
+    },
+    {
+      "epoch": 0.22819197102324176,
+      "grad_norm": 0.06568577885627747,
+      "learning_rate": 0.00019053549190535494,
+      "loss": 0.8607,
+      "step": 378
+    },
+    {
+      "epoch": 0.22879565348626624,
+      "grad_norm": 0.07336780428886414,
+      "learning_rate": 0.00019049398090493981,
+      "loss": 0.8122,
+      "step": 379
+    },
+    {
+      "epoch": 0.22939933594929068,
+      "grad_norm": 0.06682398915290833,
+      "learning_rate": 0.0001904524699045247,
+      "loss": 0.8005,
+      "step": 380
+    },
+    {
+      "epoch": 0.23000301841231513,
+      "grad_norm": 0.07804804295301437,
+      "learning_rate": 0.00019041095890410961,
+      "loss": 0.8991,
+      "step": 381
+    },
+    {
+      "epoch": 0.23060670087533958,
+      "grad_norm": 0.07914794981479645,
+      "learning_rate": 0.0001903694479036945,
+      "loss": 0.7798,
+      "step": 382
+    },
+    {
+      "epoch": 0.23121038333836402,
+      "grad_norm": 0.10398049652576447,
+      "learning_rate": 0.00019032793690327936,
+      "loss": 0.7993,
+      "step": 383
+    },
+    {
+      "epoch": 0.23181406580138847,
+      "grad_norm": 0.0717068761587143,
+      "learning_rate": 0.0001902864259028643,
+      "loss": 0.7845,
+      "step": 384
+    },
+    {
+      "epoch": 0.23241774826441292,
+      "grad_norm": 0.07129625976085663,
+      "learning_rate": 0.00019024491490244916,
+      "loss": 0.7409,
+      "step": 385
+    },
+    {
+      "epoch": 0.23302143072743736,
+      "grad_norm": 0.07239419221878052,
+      "learning_rate": 0.00019020340390203403,
+      "loss": 0.7352,
+      "step": 386
+    },
+    {
+      "epoch": 0.2336251131904618,
+      "grad_norm": 0.07583223283290863,
+      "learning_rate": 0.00019016189290161893,
+      "loss": 0.8076,
+      "step": 387
+    },
+    {
+      "epoch": 0.23422879565348625,
+      "grad_norm": 0.07693106681108475,
+      "learning_rate": 0.00019012038190120383,
+      "loss": 0.7828,
+      "step": 388
+    },
+    {
+      "epoch": 0.23483247811651073,
+      "grad_norm": 0.08271142840385437,
+      "learning_rate": 0.0001900788709007887,
+      "loss": 0.7281,
+      "step": 389
+    },
+    {
+      "epoch": 0.23543616057953518,
+      "grad_norm": 0.08440615981817245,
+      "learning_rate": 0.0001900373599003736,
+      "loss": 0.7776,
+      "step": 390
+    },
+    {
+      "epoch": 0.23603984304255962,
+      "grad_norm": 0.08483248949050903,
+      "learning_rate": 0.0001899958488999585,
+      "loss": 0.7142,
+      "step": 391
+    },
+    {
+      "epoch": 0.23664352550558407,
+      "grad_norm": 0.10398207604885101,
+      "learning_rate": 0.00018995433789954338,
+      "loss": 0.7971,
+      "step": 392
+    },
+    {
+      "epoch": 0.23724720796860851,
+      "grad_norm": 0.09310369938611984,
+      "learning_rate": 0.00018991282689912828,
+      "loss": 0.7611,
+      "step": 393
+    },
+    {
+      "epoch": 0.23785089043163296,
+      "grad_norm": 0.09080198407173157,
+      "learning_rate": 0.00018987131589871315,
+      "loss": 0.7963,
+      "step": 394
+    },
+    {
+      "epoch": 0.2384545728946574,
+      "grad_norm": 0.09882273524999619,
+      "learning_rate": 0.00018982980489829805,
+      "loss": 0.7826,
+      "step": 395
+    },
+    {
+      "epoch": 0.23905825535768185,
+      "grad_norm": 0.11651689559221268,
+      "learning_rate": 0.00018978829389788295,
+      "loss": 0.7177,
+      "step": 396
+    },
+    {
+      "epoch": 0.2396619378207063,
+      "grad_norm": 0.10827039927244186,
+      "learning_rate": 0.00018974678289746782,
+      "loss": 0.6913,
+      "step": 397
+    },
+    {
+      "epoch": 0.24026562028373075,
+      "grad_norm": 0.10637848824262619,
+      "learning_rate": 0.00018970527189705275,
+      "loss": 0.5721,
+      "step": 398
+    },
+    {
+      "epoch": 0.24086930274675522,
+      "grad_norm": 0.11688965559005737,
+      "learning_rate": 0.00018966376089663762,
+      "loss": 0.5603,
+      "step": 399
+    },
+    {
+      "epoch": 0.24147298520977967,
+      "grad_norm": 0.12776370346546173,
+      "learning_rate": 0.0001896222498962225,
+      "loss": 0.4937,
+      "step": 400
+    },
+    {
+      "epoch": 0.2420766676728041,
+      "grad_norm": 0.06769760698080063,
+      "learning_rate": 0.0001895807388958074,
+      "loss": 0.8702,
+      "step": 401
+    },
+    {
+      "epoch": 0.24268035013582856,
+      "grad_norm": 0.07302338629961014,
+      "learning_rate": 0.0001895392278953923,
+      "loss": 0.8243,
+      "step": 402
+    },
+    {
+      "epoch": 0.243284032598853,
+      "grad_norm": 0.07331572473049164,
+      "learning_rate": 0.00018949771689497717,
+      "loss": 1.1457,
+      "step": 403
+    },
+    {
+      "epoch": 0.24388771506187745,
+      "grad_norm": 0.07224598526954651,
+      "learning_rate": 0.00018945620589456207,
+      "loss": 1.0838,
+      "step": 404
+    },
+    {
+      "epoch": 0.2444913975249019,
+      "grad_norm": 0.06487125158309937,
+      "learning_rate": 0.00018941469489414697,
+      "loss": 0.8927,
+      "step": 405
+    },
+    {
+      "epoch": 0.24509507998792635,
+      "grad_norm": 0.06399485468864441,
+      "learning_rate": 0.00018937318389373184,
+      "loss": 0.7973,
+      "step": 406
+    },
+    {
+      "epoch": 0.2456987624509508,
+      "grad_norm": 0.08675524592399597,
+      "learning_rate": 0.00018933167289331674,
+      "loss": 0.9525,
+      "step": 407
+    },
+    {
+      "epoch": 0.24630244491397524,
+      "grad_norm": 0.0663706362247467,
+      "learning_rate": 0.00018929016189290162,
+      "loss": 0.7668,
+      "step": 408
+    },
+    {
+      "epoch": 0.24690612737699968,
+      "grad_norm": 0.09714970737695694,
+      "learning_rate": 0.00018924865089248652,
+      "loss": 0.783,
+      "step": 409
+    },
+    {
+      "epoch": 0.24750980984002416,
+      "grad_norm": 0.06401702761650085,
+      "learning_rate": 0.00018920713989207142,
+      "loss": 1.0896,
+      "step": 410
+    },
+    {
+      "epoch": 0.2481134923030486,
+      "grad_norm": 0.07394849509000778,
+      "learning_rate": 0.0001891656288916563,
+      "loss": 0.79,
+      "step": 411
+    },
+    {
+      "epoch": 0.24871717476607305,
+      "grad_norm": 0.1234453096985817,
+      "learning_rate": 0.0001891241178912412,
+      "loss": 0.841,
+      "step": 412
+    },
+    {
+      "epoch": 0.2493208572290975,
+      "grad_norm": 0.06933286041021347,
+      "learning_rate": 0.0001890826068908261,
+      "loss": 1.1143,
+      "step": 413
+    },
+    {
+      "epoch": 0.24992453969212194,
+      "grad_norm": 0.07103940844535828,
+      "learning_rate": 0.00018904109589041096,
+      "loss": 0.8918,
+      "step": 414
+    },
+    {
+      "epoch": 0.2505282221551464,
+      "grad_norm": 0.11104138940572739,
+      "learning_rate": 0.00018899958488999586,
+      "loss": 0.922,
+      "step": 415
+    },
+    {
+      "epoch": 0.25113190461817086,
+      "grad_norm": 0.06877300888299942,
+      "learning_rate": 0.00018895807388958076,
+      "loss": 0.9959,
+      "step": 416
+    },
+    {
+      "epoch": 0.2517355870811953,
+      "grad_norm": 0.07087550312280655,
+      "learning_rate": 0.00018891656288916564,
+      "loss": 0.8376,
+      "step": 417
+    },
+    {
+      "epoch": 0.25233926954421976,
+      "grad_norm": 0.08183194696903229,
+      "learning_rate": 0.0001888750518887505,
+      "loss": 0.8378,
+      "step": 418
+    },
+    {
+      "epoch": 0.2529429520072442,
+      "grad_norm": 0.06973780691623688,
+      "learning_rate": 0.00018883354088833544,
+      "loss": 0.8195,
+      "step": 419
+    },
+    {
+      "epoch": 0.25354663447026865,
+      "grad_norm": 0.07654432952404022,
+      "learning_rate": 0.0001887920298879203,
+      "loss": 0.8851,
+      "step": 420
+    },
+    {
+      "epoch": 0.2541503169332931,
+      "grad_norm": 0.0699959322810173,
+      "learning_rate": 0.00018875051888750518,
+      "loss": 1.0005,
+      "step": 421
+    },
+    {
+      "epoch": 0.25475399939631754,
+      "grad_norm": 0.06692960858345032,
+      "learning_rate": 0.00018870900788709008,
+      "loss": 0.8751,
+      "step": 422
+    },
+    {
+      "epoch": 0.255357681859342,
+      "grad_norm": 0.06952139735221863,
+      "learning_rate": 0.00018866749688667498,
+      "loss": 0.968,
+      "step": 423
+    },
+    {
+      "epoch": 0.25596136432236644,
+      "grad_norm": 0.0676087811589241,
+      "learning_rate": 0.00018862598588625985,
+      "loss": 0.9608,
+      "step": 424
+    },
+    {
+      "epoch": 0.2565650467853909,
+      "grad_norm": 0.07644575089216232,
+      "learning_rate": 0.00018858447488584475,
+      "loss": 0.8708,
+      "step": 425
+    },
+    {
+      "epoch": 0.25716872924841533,
+      "grad_norm": 0.0646698921918869,
+      "learning_rate": 0.00018854296388542965,
+      "loss": 0.7413,
+      "step": 426
+    },
+    {
+      "epoch": 0.2577724117114398,
+      "grad_norm": 0.06822269409894943,
+      "learning_rate": 0.00018850145288501453,
+      "loss": 0.8585,
+      "step": 427
+    },
+    {
+      "epoch": 0.2583760941744642,
+      "grad_norm": 0.23585358262062073,
+      "learning_rate": 0.00018845994188459943,
+      "loss": 0.8879,
+      "step": 428
+    },
+    {
+      "epoch": 0.25897977663748867,
+      "grad_norm": 0.07148776948451996,
+      "learning_rate": 0.00018841843088418433,
+      "loss": 0.9396,
+      "step": 429
+    },
+    {
+      "epoch": 0.2595834591005131,
+      "grad_norm": 0.07239923626184464,
+      "learning_rate": 0.0001883769198837692,
+      "loss": 0.9303,
+      "step": 430
+    },
+    {
+      "epoch": 0.26018714156353756,
+      "grad_norm": 0.06573071330785751,
+      "learning_rate": 0.0001883354088833541,
+      "loss": 0.866,
+      "step": 431
+    },
+    {
+      "epoch": 0.260790824026562,
+      "grad_norm": 0.06852304190397263,
+      "learning_rate": 0.00018829389788293897,
+      "loss": 1.0774,
+      "step": 432
+    },
+    {
+      "epoch": 0.26139450648958645,
+      "grad_norm": 0.06725315749645233,
+      "learning_rate": 0.00018825238688252387,
+      "loss": 0.9174,
+      "step": 433
+    },
+    {
+      "epoch": 0.2619981889526109,
+      "grad_norm": 0.07200594246387482,
+      "learning_rate": 0.00018821087588210877,
+      "loss": 0.7667,
+      "step": 434
+    },
+    {
+      "epoch": 0.2626018714156354,
+      "grad_norm": 0.06961148232221603,
+      "learning_rate": 0.00018816936488169365,
+      "loss": 0.7284,
+      "step": 435
+    },
+    {
+      "epoch": 0.26320555387865985,
+      "grad_norm": 0.07602632790803909,
+      "learning_rate": 0.00018812785388127855,
+      "loss": 0.8159,
+      "step": 436
+    },
+    {
+      "epoch": 0.2638092363416843,
+      "grad_norm": 0.0729806199669838,
+      "learning_rate": 0.00018808634288086345,
+      "loss": 0.7094,
+      "step": 437
+    },
+    {
+      "epoch": 0.26441291880470874,
+      "grad_norm": 0.07881097495555878,
+      "learning_rate": 0.00018804483188044832,
+      "loss": 0.7488,
+      "step": 438
+    },
+    {
+      "epoch": 0.2650166012677332,
+      "grad_norm": 0.08327119052410126,
+      "learning_rate": 0.00018800332088003322,
+      "loss": 0.7672,
+      "step": 439
+    },
+    {
+      "epoch": 0.26562028373075763,
+      "grad_norm": 0.08234728127717972,
+      "learning_rate": 0.00018796180987961812,
+      "loss": 0.7644,
+      "step": 440
+    },
+    {
+      "epoch": 0.2662239661937821,
+      "grad_norm": 0.09000684320926666,
+      "learning_rate": 0.000187920298879203,
+      "loss": 0.793,
+      "step": 441
+    },
+    {
+      "epoch": 0.2668276486568065,
+      "grad_norm": 0.07901246100664139,
+      "learning_rate": 0.0001878787878787879,
+      "loss": 0.7346,
+      "step": 442
+    },
+    {
+      "epoch": 0.267431331119831,
+      "grad_norm": 0.09163561463356018,
+      "learning_rate": 0.0001878372768783728,
+      "loss": 0.762,
+      "step": 443
+    },
+    {
+      "epoch": 0.2680350135828554,
+      "grad_norm": 0.09499283879995346,
+      "learning_rate": 0.00018779576587795767,
+      "loss": 0.7669,
+      "step": 444
+    },
+    {
+      "epoch": 0.26863869604587987,
+      "grad_norm": 0.1023356169462204,
+      "learning_rate": 0.00018775425487754257,
+      "loss": 0.7431,
+      "step": 445
+    },
+    {
+      "epoch": 0.2692423785089043,
+      "grad_norm": 0.10197634249925613,
+      "learning_rate": 0.00018771274387712744,
+      "loss": 0.7059,
+      "step": 446
+    },
+    {
+      "epoch": 0.26984606097192876,
+      "grad_norm": 0.09975861012935638,
+      "learning_rate": 0.00018767123287671234,
+      "loss": 0.658,
+      "step": 447
+    },
+    {
+      "epoch": 0.2704497434349532,
+      "grad_norm": 0.11030542105436325,
+      "learning_rate": 0.00018762972187629724,
+      "loss": 0.6281,
+      "step": 448
+    },
+    {
+      "epoch": 0.27105342589797765,
+      "grad_norm": 0.11011619865894318,
+      "learning_rate": 0.0001875882108758821,
+      "loss": 0.5196,
+      "step": 449
+    },
+    {
+      "epoch": 0.2716571083610021,
+      "grad_norm": 0.12410090863704681,
+      "learning_rate": 0.000187546699875467,
+      "loss": 0.4871,
+      "step": 450
+    },
+    {
+      "epoch": 0.27226079082402654,
+      "grad_norm": 0.06774574518203735,
+      "learning_rate": 0.0001875051888750519,
+      "loss": 0.7875,
+      "step": 451
+    },
+    {
+      "epoch": 0.272864473287051,
+      "grad_norm": 0.06862092763185501,
+      "learning_rate": 0.00018746367787463678,
+      "loss": 0.8497,
+      "step": 452
+    },
+    {
+      "epoch": 0.27346815575007544,
+      "grad_norm": 0.07752003520727158,
+      "learning_rate": 0.00018742216687422168,
+      "loss": 0.7635,
+      "step": 453
+    },
+    {
+      "epoch": 0.2740718382130999,
+      "grad_norm": 0.138156920671463,
+      "learning_rate": 0.00018738065587380658,
+      "loss": 0.8682,
+      "step": 454
+    },
+    {
+      "epoch": 0.2746755206761244,
+      "grad_norm": 0.07247649133205414,
+      "learning_rate": 0.00018733914487339146,
+      "loss": 0.9772,
+      "step": 455
+    },
+    {
+      "epoch": 0.27527920313914883,
+      "grad_norm": 0.06809130311012268,
+      "learning_rate": 0.00018729763387297633,
+      "loss": 0.9594,
+      "step": 456
+    },
+    {
+      "epoch": 0.2758828856021733,
+      "grad_norm": 0.08479005098342896,
+      "learning_rate": 0.00018725612287256126,
+      "loss": 1.0778,
+      "step": 457
+    },
+    {
+      "epoch": 0.2764865680651977,
+      "grad_norm": 0.06665640324354172,
+      "learning_rate": 0.00018721461187214613,
+      "loss": 0.9377,
+      "step": 458
+    },
+    {
+      "epoch": 0.27709025052822217,
+      "grad_norm": 0.06898784637451172,
+      "learning_rate": 0.000187173100871731,
+      "loss": 0.8999,
+      "step": 459
+    },
+    {
+      "epoch": 0.2776939329912466,
+      "grad_norm": 0.06468936055898666,
+      "learning_rate": 0.0001871315898713159,
+      "loss": 0.9559,
+      "step": 460
+    },
+    {
+      "epoch": 0.27829761545427106,
+      "grad_norm": 0.0695393905043602,
+      "learning_rate": 0.0001870900788709008,
+      "loss": 0.8692,
+      "step": 461
+    },
+    {
+      "epoch": 0.2789012979172955,
+      "grad_norm": 0.10467389971017838,
+      "learning_rate": 0.00018704856787048568,
+      "loss": 0.8858,
+      "step": 462
+    },
+    {
+      "epoch": 0.27950498038031996,
+      "grad_norm": 0.06943607330322266,
+      "learning_rate": 0.00018700705687007058,
+      "loss": 1.0079,
+      "step": 463
+    },
+    {
+      "epoch": 0.2801086628433444,
+      "grad_norm": 0.07984504848718643,
+      "learning_rate": 0.00018696554586965548,
+      "loss": 0.8257,
+      "step": 464
+    },
+    {
+      "epoch": 0.28071234530636885,
+      "grad_norm": 0.06335430592298508,
+      "learning_rate": 0.00018692403486924035,
+      "loss": 1.1628,
+      "step": 465
+    },
+    {
+      "epoch": 0.2813160277693933,
+      "grad_norm": 0.06936845183372498,
+      "learning_rate": 0.00018688252386882525,
+      "loss": 0.8205,
+      "step": 466
+    },
+    {
+      "epoch": 0.28191971023241774,
+      "grad_norm": 0.0703270360827446,
+      "learning_rate": 0.00018684101286841015,
+      "loss": 1.053,
+      "step": 467
+    },
+    {
+      "epoch": 0.2825233926954422,
+      "grad_norm": 0.06945156306028366,
+      "learning_rate": 0.00018679950186799502,
+      "loss": 0.8785,
+      "step": 468
+    },
+    {
+      "epoch": 0.28312707515846663,
+      "grad_norm": 0.08337710797786713,
+      "learning_rate": 0.00018675799086757992,
+      "loss": 1.0398,
+      "step": 469
+    },
+    {
+      "epoch": 0.2837307576214911,
+      "grad_norm": 0.07238437235355377,
+      "learning_rate": 0.0001867164798671648,
+      "loss": 0.9231,
+      "step": 470
+    },
+    {
+      "epoch": 0.2843344400845155,
+      "grad_norm": 0.0756809413433075,
+      "learning_rate": 0.0001866749688667497,
+      "loss": 1.1661,
+      "step": 471
+    },
+    {
+      "epoch": 0.28493812254754,
+      "grad_norm": 0.06537111848592758,
+      "learning_rate": 0.0001866334578663346,
+      "loss": 0.7645,
+      "step": 472
+    },
+    {
+      "epoch": 0.2855418050105644,
+      "grad_norm": 0.07265693694353104,
+      "learning_rate": 0.00018659194686591947,
+      "loss": 1.1106,
+      "step": 473
+    },
+    {
+      "epoch": 0.28614548747358887,
+      "grad_norm": 0.06632302701473236,
+      "learning_rate": 0.00018655043586550437,
+      "loss": 0.8943,
+      "step": 474
+    },
+    {
+      "epoch": 0.28674916993661337,
+      "grad_norm": 0.06948743760585785,
+      "learning_rate": 0.00018650892486508927,
+      "loss": 1.0051,
+      "step": 475
+    },
+    {
+      "epoch": 0.2873528523996378,
+      "grad_norm": 0.07059241831302643,
+      "learning_rate": 0.00018646741386467414,
+      "loss": 1.0785,
+      "step": 476
+    },
+    {
+      "epoch": 0.28795653486266226,
+      "grad_norm": 0.06578975915908813,
+      "learning_rate": 0.00018642590286425901,
+      "loss": 0.8156,
+      "step": 477
+    },
+    {
+      "epoch": 0.2885602173256867,
+      "grad_norm": 0.06178348883986473,
+      "learning_rate": 0.00018638439186384394,
+      "loss": 0.8512,
+      "step": 478
+    },
+    {
+      "epoch": 0.28916389978871115,
+      "grad_norm": 0.06867639720439911,
+      "learning_rate": 0.00018634288086342881,
+      "loss": 0.8549,
+      "step": 479
+    },
+    {
+      "epoch": 0.2897675822517356,
+      "grad_norm": 0.07352021336555481,
+      "learning_rate": 0.0001863013698630137,
+      "loss": 0.7585,
+      "step": 480
+    },
+    {
+      "epoch": 0.29037126471476005,
+      "grad_norm": 0.07471666485071182,
+      "learning_rate": 0.00018625985886259861,
+      "loss": 0.9468,
+      "step": 481
+    },
+    {
+      "epoch": 0.2909749471777845,
+      "grad_norm": 0.06977469474077225,
+      "learning_rate": 0.0001862183478621835,
+      "loss": 0.7439,
+      "step": 482
+    },
+    {
+      "epoch": 0.29157862964080894,
+      "grad_norm": 0.07040760666131973,
+      "learning_rate": 0.00018617683686176836,
+      "loss": 0.7331,
+      "step": 483
+    },
+    {
+      "epoch": 0.2921823121038334,
+      "grad_norm": 0.07527907937765121,
+      "learning_rate": 0.00018613532586135326,
+      "loss": 0.8511,
+      "step": 484
+    },
+    {
+      "epoch": 0.29278599456685783,
+      "grad_norm": 0.07468552142381668,
+      "learning_rate": 0.00018609381486093816,
+      "loss": 0.8146,
+      "step": 485
+    },
+    {
+      "epoch": 0.2933896770298823,
+      "grad_norm": 0.0728183388710022,
+      "learning_rate": 0.00018605230386052306,
+      "loss": 0.8189,
+      "step": 486
+    },
+    {
+      "epoch": 0.2939933594929067,
+      "grad_norm": 0.07946665585041046,
+      "learning_rate": 0.00018601079286010793,
+      "loss": 0.8002,
+      "step": 487
+    },
+    {
+      "epoch": 0.29459704195593117,
+      "grad_norm": 0.08301442861557007,
+      "learning_rate": 0.00018596928185969283,
+      "loss": 0.76,
+      "step": 488
+    },
+    {
+      "epoch": 0.2952007244189556,
+      "grad_norm": 0.07893984019756317,
+      "learning_rate": 0.00018592777085927773,
+      "loss": 0.7711,
+      "step": 489
+    },
+    {
+      "epoch": 0.29580440688198006,
+      "grad_norm": 0.08230159431695938,
+      "learning_rate": 0.0001858862598588626,
+      "loss": 0.7041,
+      "step": 490
+    },
+    {
+      "epoch": 0.2964080893450045,
+      "grad_norm": 0.09270324558019638,
+      "learning_rate": 0.00018584474885844748,
+      "loss": 0.7666,
+      "step": 491
+    },
+    {
+      "epoch": 0.29701177180802896,
+      "grad_norm": 0.09311135858297348,
+      "learning_rate": 0.0001858032378580324,
+      "loss": 0.7878,
+      "step": 492
+    },
+    {
+      "epoch": 0.2976154542710534,
+      "grad_norm": 0.09099919348955154,
+      "learning_rate": 0.00018576172685761728,
+      "loss": 0.7074,
+      "step": 493
+    },
+    {
+      "epoch": 0.29821913673407785,
+      "grad_norm": 0.09507730603218079,
+      "learning_rate": 0.00018572021585720215,
+      "loss": 0.6735,
+      "step": 494
+    },
+    {
+      "epoch": 0.29882281919710235,
+      "grad_norm": 0.09427333623170853,
+      "learning_rate": 0.00018567870485678708,
+      "loss": 0.7592,
+      "step": 495
+    },
+    {
+      "epoch": 0.2994265016601268,
+      "grad_norm": 0.1038491502404213,
+      "learning_rate": 0.00018563719385637195,
+      "loss": 0.7356,
+      "step": 496
+    },
+    {
+      "epoch": 0.30003018412315124,
+      "grad_norm": 0.1097760945558548,
+      "learning_rate": 0.00018559568285595683,
+      "loss": 0.6825,
+      "step": 497
+    },
+    {
+      "epoch": 0.3006338665861757,
+      "grad_norm": 0.10912280529737473,
+      "learning_rate": 0.00018555417185554173,
+      "loss": 0.6431,
+      "step": 498
+    },
+    {
+      "epoch": 0.30123754904920014,
+      "grad_norm": 0.1107010692358017,
+      "learning_rate": 0.00018551266085512663,
+      "loss": 0.5303,
+      "step": 499
+    },
+    {
+      "epoch": 0.3018412315122246,
+      "grad_norm": 0.12602561712265015,
+      "learning_rate": 0.0001854711498547115,
+      "loss": 0.4541,
+      "step": 500
+    },
+    {
+      "epoch": 0.3018412315122246,
+      "eval_loss": 0.8565130233764648,
+      "eval_runtime": 1219.0924,
+      "eval_samples_per_second": 2.289,
+      "eval_steps_per_second": 0.286,
+      "step": 500
+    },
+    {
+      "epoch": 0.30244491397524903,
+      "grad_norm": 0.11287853866815567,
+      "learning_rate": 0.0001854296388542964,
+      "loss": 0.8697,
+      "step": 501
+    },
+    {
+      "epoch": 0.3030485964382735,
+      "grad_norm": 0.08270236849784851,
+      "learning_rate": 0.0001853881278538813,
+      "loss": 0.8881,
+      "step": 502
+    },
+    {
+      "epoch": 0.3036522789012979,
+      "grad_norm": 0.07861533761024475,
+      "learning_rate": 0.00018534661685346617,
+      "loss": 0.8053,
+      "step": 503
+    },
+    {
+      "epoch": 0.30425596136432237,
+      "grad_norm": 0.07589036971330643,
+      "learning_rate": 0.00018530510585305107,
+      "loss": 1.0165,
+      "step": 504
+    },
+    {
+      "epoch": 0.3048596438273468,
+      "grad_norm": 0.07618770003318787,
+      "learning_rate": 0.00018526359485263594,
+      "loss": 1.0332,
+      "step": 505
+    },
+    {
+      "epoch": 0.30546332629037126,
+      "grad_norm": 0.07502375543117523,
+      "learning_rate": 0.00018522208385222084,
+      "loss": 0.976,
+      "step": 506
+    },
+    {
+      "epoch": 0.3060670087533957,
+      "grad_norm": 0.07355045527219772,
+      "learning_rate": 0.00018518057285180574,
+      "loss": 0.7887,
+      "step": 507
+    },
+    {
+      "epoch": 0.30667069121642015,
+      "grad_norm": 0.08850781619548798,
+      "learning_rate": 0.00018513906185139062,
+      "loss": 1.0632,
+      "step": 508
+    },
+    {
+      "epoch": 0.3072743736794446,
+      "grad_norm": 0.06971225887537003,
+      "learning_rate": 0.00018509755085097552,
+      "loss": 0.9073,
+      "step": 509
+    },
+    {
+      "epoch": 0.30787805614246905,
+      "grad_norm": 0.16970930993556976,
+      "learning_rate": 0.00018505603985056042,
+      "loss": 0.9196,
+      "step": 510
+    },
+    {
+      "epoch": 0.3084817386054935,
+      "grad_norm": 0.07109450548887253,
+      "learning_rate": 0.0001850145288501453,
+      "loss": 0.8324,
+      "step": 511
+    },
+    {
+      "epoch": 0.30908542106851794,
+      "grad_norm": 0.06761594116687775,
+      "learning_rate": 0.0001849730178497302,
+      "loss": 0.8924,
+      "step": 512
+    },
+    {
+      "epoch": 0.3096891035315424,
+      "grad_norm": 0.06660860776901245,
+      "learning_rate": 0.0001849315068493151,
+      "loss": 0.8244,
+      "step": 513
+    },
+    {
+      "epoch": 0.31029278599456683,
+      "grad_norm": 0.08837427198886871,
+      "learning_rate": 0.00018488999584889996,
+      "loss": 0.8228,
+      "step": 514
+    },
+    {
+      "epoch": 0.3108964684575913,
+      "grad_norm": 0.09773479402065277,
+      "learning_rate": 0.00018484848484848484,
+      "loss": 0.8686,
+      "step": 515
+    },
+    {
+      "epoch": 0.3115001509206158,
+      "grad_norm": 0.07643985003232956,
+      "learning_rate": 0.00018480697384806976,
+      "loss": 0.8456,
+      "step": 516
+    },
+    {
+      "epoch": 0.31210383338364023,
+      "grad_norm": 0.06785845011472702,
+      "learning_rate": 0.00018476546284765464,
+      "loss": 0.8381,
+      "step": 517
+    },
+    {
+      "epoch": 0.3127075158466647,
+      "grad_norm": 0.0758255124092102,
+      "learning_rate": 0.0001847239518472395,
+      "loss": 0.8728,
+      "step": 518
+    },
+    {
+      "epoch": 0.3133111983096891,
+      "grad_norm": 0.0750250369310379,
+      "learning_rate": 0.0001846824408468244,
+      "loss": 0.7163,
+      "step": 519
+    },
+    {
+      "epoch": 0.31391488077271357,
+      "grad_norm": 0.06817881017923355,
+      "learning_rate": 0.0001846409298464093,
+      "loss": 0.7658,
+      "step": 520
+    },
+    {
+      "epoch": 0.314518563235738,
+      "grad_norm": 0.06935089081525803,
+      "learning_rate": 0.00018459941884599418,
+      "loss": 0.8448,
+      "step": 521
+    },
+    {
+      "epoch": 0.31512224569876246,
+      "grad_norm": 0.0738990381360054,
+      "learning_rate": 0.00018455790784557908,
+      "loss": 0.8026,
+      "step": 522
+    },
+    {
+      "epoch": 0.3157259281617869,
+      "grad_norm": 0.07354303449392319,
+      "learning_rate": 0.00018451639684516398,
+      "loss": 0.8194,
+      "step": 523
+    },
+    {
+      "epoch": 0.31632961062481135,
+      "grad_norm": 0.07702893018722534,
+      "learning_rate": 0.00018447488584474886,
+      "loss": 0.8923,
+      "step": 524
+    },
+    {
+      "epoch": 0.3169332930878358,
+      "grad_norm": 0.09026607125997543,
+      "learning_rate": 0.00018443337484433376,
+      "loss": 0.9738,
+      "step": 525
+    },
+    {
+      "epoch": 0.31753697555086025,
+      "grad_norm": 0.08617381006479263,
+      "learning_rate": 0.00018439186384391866,
+      "loss": 0.9765,
+      "step": 526
+    },
+    {
+      "epoch": 0.3181406580138847,
+      "grad_norm": 0.07624173164367676,
+      "learning_rate": 0.00018435035284350353,
+      "loss": 1.0301,
+      "step": 527
+    },
+    {
+      "epoch": 0.31874434047690914,
+      "grad_norm": 0.06759200245141983,
+      "learning_rate": 0.00018430884184308843,
+      "loss": 0.8237,
+      "step": 528
+    },
+    {
+      "epoch": 0.3193480229399336,
+      "grad_norm": 0.07457785308361053,
+      "learning_rate": 0.0001842673308426733,
+      "loss": 0.7363,
+      "step": 529
+    },
+    {
+      "epoch": 0.31995170540295803,
+      "grad_norm": 0.08332669734954834,
+      "learning_rate": 0.00018422581984225823,
+      "loss": 0.7471,
+      "step": 530
+    },
+    {
+      "epoch": 0.3205553878659825,
+      "grad_norm": 0.0689413771033287,
+      "learning_rate": 0.0001841843088418431,
+      "loss": 0.8585,
+      "step": 531
+    },
+    {
+      "epoch": 0.3211590703290069,
+      "grad_norm": 0.06793367117643356,
+      "learning_rate": 0.00018414279784142797,
+      "loss": 1.033,
+      "step": 532
+    },
+    {
+      "epoch": 0.32176275279203137,
+      "grad_norm": 0.07640314847230911,
+      "learning_rate": 0.00018410128684101287,
+      "loss": 1.0033,
+      "step": 533
+    },
+    {
+      "epoch": 0.3223664352550558,
+      "grad_norm": 0.07172773033380508,
+      "learning_rate": 0.00018405977584059777,
+      "loss": 0.7355,
+      "step": 534
+    },
+    {
+      "epoch": 0.32297011771808026,
+      "grad_norm": 0.07412228733301163,
+      "learning_rate": 0.00018401826484018265,
+      "loss": 0.7548,
+      "step": 535
+    },
+    {
+      "epoch": 0.32357380018110476,
+      "grad_norm": 0.07675584405660629,
+      "learning_rate": 0.00018397675383976755,
+      "loss": 0.7105,
+      "step": 536
+    },
+    {
+      "epoch": 0.3241774826441292,
+      "grad_norm": 0.07256907969713211,
+      "learning_rate": 0.00018393524283935245,
+      "loss": 0.7383,
+      "step": 537
+    },
+    {
+      "epoch": 0.32478116510715366,
+      "grad_norm": 0.08796676248311996,
+      "learning_rate": 0.00018389373183893732,
+      "loss": 0.7606,
+      "step": 538
+    },
+    {
+      "epoch": 0.3253848475701781,
+      "grad_norm": 0.07517845928668976,
+      "learning_rate": 0.00018385222083852222,
+      "loss": 0.8343,
+      "step": 539
+    },
+    {
+      "epoch": 0.32598853003320255,
+      "grad_norm": 0.08098675310611725,
+      "learning_rate": 0.00018381070983810712,
+      "loss": 0.7752,
+      "step": 540
+    },
+    {
+      "epoch": 0.326592212496227,
+      "grad_norm": 0.08791480958461761,
+      "learning_rate": 0.000183769198837692,
+      "loss": 0.6954,
+      "step": 541
+    },
+    {
+      "epoch": 0.32719589495925144,
+      "grad_norm": 0.09234142303466797,
+      "learning_rate": 0.0001837276878372769,
+      "loss": 0.7773,
+      "step": 542
+    },
+    {
+      "epoch": 0.3277995774222759,
+      "grad_norm": 0.09258239716291428,
+      "learning_rate": 0.00018368617683686177,
+      "loss": 0.7787,
+      "step": 543
+    },
+    {
+      "epoch": 0.32840325988530034,
+      "grad_norm": 0.09329955279827118,
+      "learning_rate": 0.00018364466583644667,
+      "loss": 0.6927,
+      "step": 544
+    },
+    {
+      "epoch": 0.3290069423483248,
+      "grad_norm": 0.1046968549489975,
+      "learning_rate": 0.00018360315483603157,
+      "loss": 0.742,
+      "step": 545
+    },
+    {
+      "epoch": 0.32961062481134923,
+      "grad_norm": 0.11193925142288208,
+      "learning_rate": 0.00018356164383561644,
+      "loss": 0.7062,
+      "step": 546
+    },
+    {
+      "epoch": 0.3302143072743737,
+      "grad_norm": 0.10192117840051651,
+      "learning_rate": 0.00018352013283520134,
+      "loss": 0.6246,
+      "step": 547
+    },
+    {
+      "epoch": 0.3308179897373981,
+      "grad_norm": 0.11968918889760971,
+      "learning_rate": 0.00018347862183478624,
+      "loss": 0.6263,
+      "step": 548
+    },
+    {
+      "epoch": 0.33142167220042257,
+      "grad_norm": 0.11495087295770645,
+      "learning_rate": 0.0001834371108343711,
+      "loss": 0.5845,
+      "step": 549
+    },
+    {
+      "epoch": 0.332025354663447,
+      "grad_norm": 0.1315176784992218,
+      "learning_rate": 0.000183395599833956,
+      "loss": 0.5137,
+      "step": 550
+    },
+    {
+      "epoch": 0.33262903712647146,
+      "grad_norm": 0.07526993751525879,
+      "learning_rate": 0.0001833540888335409,
+      "loss": 0.8475,
+      "step": 551
+    },
+    {
+      "epoch": 0.3332327195894959,
+      "grad_norm": 0.08781188726425171,
+      "learning_rate": 0.00018331257783312579,
+      "loss": 1.0003,
+      "step": 552
+    },
+    {
+      "epoch": 0.33383640205252035,
+      "grad_norm": 0.07417803257703781,
+      "learning_rate": 0.00018327106683271066,
+      "loss": 0.9112,
+      "step": 553
+    },
+    {
+      "epoch": 0.3344400845155448,
+      "grad_norm": 0.08806382119655609,
+      "learning_rate": 0.00018322955583229559,
+      "loss": 0.9675,
+      "step": 554
+    },
+    {
+      "epoch": 0.33504376697856925,
+      "grad_norm": 0.08356613665819168,
+      "learning_rate": 0.00018318804483188046,
+      "loss": 0.808,
+      "step": 555
+    },
+    {
+      "epoch": 0.33564744944159375,
+      "grad_norm": 0.08373204618692398,
+      "learning_rate": 0.00018314653383146533,
+      "loss": 1.1798,
+      "step": 556
+    },
+    {
+      "epoch": 0.3362511319046182,
+      "grad_norm": 0.07093155384063721,
+      "learning_rate": 0.00018310502283105023,
+      "loss": 1.0393,
+      "step": 557
+    },
+    {
+      "epoch": 0.33685481436764264,
+      "grad_norm": 0.11354014277458191,
+      "learning_rate": 0.00018306351183063513,
+      "loss": 0.8586,
+      "step": 558
+    },
+    {
+      "epoch": 0.3374584968306671,
+      "grad_norm": 0.07332495599985123,
+      "learning_rate": 0.00018302200083022,
+      "loss": 0.6829,
+      "step": 559
+    },
+    {
+      "epoch": 0.33806217929369153,
+      "grad_norm": 0.07165679335594177,
+      "learning_rate": 0.0001829804898298049,
+      "loss": 0.8619,
+      "step": 560
+    },
+    {
+      "epoch": 0.338665861756716,
+      "grad_norm": 0.07097924500703812,
+      "learning_rate": 0.0001829389788293898,
+      "loss": 0.7529,
+      "step": 561
+    },
+    {
+      "epoch": 0.3392695442197404,
+      "grad_norm": 0.06817946583032608,
+      "learning_rate": 0.00018289746782897468,
+      "loss": 0.9555,
+      "step": 562
+    },
+    {
+      "epoch": 0.3398732266827649,
+      "grad_norm": 0.07513990998268127,
+      "learning_rate": 0.00018285595682855958,
+      "loss": 0.874,
+      "step": 563
+    },
+    {
+      "epoch": 0.3404769091457893,
+      "grad_norm": 0.07172597199678421,
+      "learning_rate": 0.00018281444582814448,
+      "loss": 0.8608,
+      "step": 564
+    },
+    {
+      "epoch": 0.34108059160881377,
+      "grad_norm": 0.08141127973794937,
+      "learning_rate": 0.00018277293482772935,
+      "loss": 1.1752,
+      "step": 565
+    },
+    {
+      "epoch": 0.3416842740718382,
+      "grad_norm": 0.06964030116796494,
+      "learning_rate": 0.00018273142382731425,
+      "loss": 0.8235,
+      "step": 566
+    },
+    {
+      "epoch": 0.34228795653486266,
+      "grad_norm": 0.07054075598716736,
+      "learning_rate": 0.00018268991282689912,
+      "loss": 0.9064,
+      "step": 567
+    },
+    {
+      "epoch": 0.3428916389978871,
+      "grad_norm": 0.08551137149333954,
+      "learning_rate": 0.00018264840182648402,
+      "loss": 1.1779,
+      "step": 568
+    },
+    {
+      "epoch": 0.34349532146091155,
+      "grad_norm": 0.06879527121782303,
+      "learning_rate": 0.00018260689082606892,
+      "loss": 0.8866,
+      "step": 569
+    },
+    {
+      "epoch": 0.344099003923936,
+      "grad_norm": 0.07618753612041473,
+      "learning_rate": 0.0001825653798256538,
+      "loss": 0.9287,
+      "step": 570
+    },
+    {
+      "epoch": 0.34470268638696044,
+      "grad_norm": 0.0731632187962532,
+      "learning_rate": 0.0001825238688252387,
+      "loss": 1.2391,
+      "step": 571
+    },
+    {
+      "epoch": 0.3453063688499849,
+      "grad_norm": 0.06929552555084229,
+      "learning_rate": 0.0001824823578248236,
+      "loss": 0.8374,
+      "step": 572
+    },
+    {
+      "epoch": 0.34591005131300934,
+      "grad_norm": 0.08287245035171509,
+      "learning_rate": 0.00018244084682440847,
+      "loss": 0.9951,
+      "step": 573
+    },
+    {
+      "epoch": 0.3465137337760338,
+      "grad_norm": 0.08636925369501114,
+      "learning_rate": 0.00018239933582399337,
+      "loss": 0.7663,
+      "step": 574
+    },
+    {
+      "epoch": 0.34711741623905823,
+      "grad_norm": 0.07229752093553543,
+      "learning_rate": 0.00018235782482357827,
+      "loss": 0.9063,
+      "step": 575
+    },
+    {
+      "epoch": 0.34772109870208273,
+      "grad_norm": 0.08006583899259567,
+      "learning_rate": 0.00018231631382316314,
+      "loss": 0.8608,
+      "step": 576
+    },
+    {
+      "epoch": 0.3483247811651072,
+      "grad_norm": 0.07500839233398438,
+      "learning_rate": 0.00018227480282274804,
+      "loss": 1.0615,
+      "step": 577
+    },
+    {
+      "epoch": 0.3489284636281316,
+      "grad_norm": 0.07585518062114716,
+      "learning_rate": 0.00018223329182233294,
+      "loss": 1.3967,
+      "step": 578
+    },
+    {
+      "epoch": 0.34953214609115607,
+      "grad_norm": 0.06830618530511856,
+      "learning_rate": 0.00018219178082191782,
+      "loss": 1.0998,
+      "step": 579
+    },
+    {
+      "epoch": 0.3501358285541805,
+      "grad_norm": 0.0734933540225029,
+      "learning_rate": 0.00018215026982150272,
+      "loss": 0.8375,
+      "step": 580
+    },
+    {
+      "epoch": 0.35073951101720496,
+      "grad_norm": 0.07250437140464783,
+      "learning_rate": 0.0001821087588210876,
+      "loss": 1.0154,
+      "step": 581
+    },
+    {
+      "epoch": 0.3513431934802294,
+      "grad_norm": 0.07496879994869232,
+      "learning_rate": 0.0001820672478206725,
+      "loss": 0.8673,
+      "step": 582
+    },
+    {
+      "epoch": 0.35194687594325386,
+      "grad_norm": 0.07836568355560303,
+      "learning_rate": 0.0001820257368202574,
+      "loss": 0.8566,
+      "step": 583
+    },
+    {
+      "epoch": 0.3525505584062783,
+      "grad_norm": 0.07284071296453476,
+      "learning_rate": 0.00018198422581984226,
+      "loss": 0.7801,
+      "step": 584
+    },
+    {
+      "epoch": 0.35315424086930275,
+      "grad_norm": 0.07430974394083023,
+      "learning_rate": 0.00018194271481942716,
+      "loss": 0.7227,
+      "step": 585
+    },
+    {
+      "epoch": 0.3537579233323272,
+      "grad_norm": 0.0871773213148117,
+      "learning_rate": 0.00018190120381901206,
+      "loss": 0.6884,
+      "step": 586
+    },
+    {
+      "epoch": 0.35436160579535164,
+      "grad_norm": 0.08565318584442139,
+      "learning_rate": 0.00018185969281859693,
+      "loss": 0.7762,
+      "step": 587
+    },
+    {
+      "epoch": 0.3549652882583761,
+      "grad_norm": 0.07946131378412247,
+      "learning_rate": 0.00018181818181818183,
+      "loss": 1.008,
+      "step": 588
+    },
+    {
+      "epoch": 0.35556897072140053,
+      "grad_norm": 0.07984606176614761,
+      "learning_rate": 0.00018177667081776673,
+      "loss": 0.8231,
+      "step": 589
+    },
+    {
+      "epoch": 0.356172653184425,
+      "grad_norm": 0.08642979711294174,
+      "learning_rate": 0.0001817351598173516,
+      "loss": 0.6995,
+      "step": 590
+    },
+    {
+      "epoch": 0.3567763356474494,
+      "grad_norm": 0.0884912982583046,
+      "learning_rate": 0.00018169364881693648,
+      "loss": 0.7622,
+      "step": 591
+    },
+    {
+      "epoch": 0.3573800181104739,
+      "grad_norm": 0.08437898755073547,
+      "learning_rate": 0.0001816521378165214,
+      "loss": 0.6997,
+      "step": 592
+    },
+    {
+      "epoch": 0.3579837005734983,
+      "grad_norm": 0.1158236563205719,
+      "learning_rate": 0.00018161062681610628,
+      "loss": 0.7354,
+      "step": 593
+    },
+    {
+      "epoch": 0.35858738303652277,
+      "grad_norm": 0.09563788026571274,
+      "learning_rate": 0.00018156911581569115,
+      "loss": 0.8278,
+      "step": 594
+    },
+    {
+      "epoch": 0.3591910654995472,
+      "grad_norm": 0.10473807901144028,
+      "learning_rate": 0.00018152760481527605,
+      "loss": 0.7105,
+      "step": 595
+    },
+    {
+      "epoch": 0.3597947479625717,
+      "grad_norm": 0.09579318761825562,
+      "learning_rate": 0.00018148609381486095,
+      "loss": 0.6555,
+      "step": 596
+    },
+    {
+      "epoch": 0.36039843042559616,
+      "grad_norm": 0.10552844405174255,
+      "learning_rate": 0.00018144458281444583,
+      "loss": 0.6662,
+      "step": 597
+    },
+    {
+      "epoch": 0.3610021128886206,
+      "grad_norm": 0.10778263211250305,
+      "learning_rate": 0.00018140307181403073,
+      "loss": 0.6501,
+      "step": 598
+    },
+    {
+      "epoch": 0.36160579535164505,
+      "grad_norm": 0.11160840839147568,
+      "learning_rate": 0.00018136156081361563,
+      "loss": 0.5741,
+      "step": 599
+    },
+    {
+      "epoch": 0.3622094778146695,
+      "grad_norm": 0.11790695786476135,
+      "learning_rate": 0.0001813200498132005,
+      "loss": 0.507,
+      "step": 600
+    },
+    {
+      "epoch": 0.36281316027769395,
+      "grad_norm": 0.09489453583955765,
+      "learning_rate": 0.0001812785388127854,
+      "loss": 0.905,
+      "step": 601
+    },
+    {
+      "epoch": 0.3634168427407184,
+      "grad_norm": 0.07241594046354294,
+      "learning_rate": 0.0001812370278123703,
+      "loss": 0.9352,
+      "step": 602
+    },
+    {
+      "epoch": 0.36402052520374284,
+      "grad_norm": 0.07990288734436035,
+      "learning_rate": 0.00018119551681195517,
+      "loss": 0.9155,
+      "step": 603
+    },
+    {
+      "epoch": 0.3646242076667673,
+      "grad_norm": 0.07297532260417938,
+      "learning_rate": 0.00018115400581154007,
+      "loss": 0.9952,
+      "step": 604
+    },
+    {
+      "epoch": 0.36522789012979173,
+      "grad_norm": 0.10625027120113373,
+      "learning_rate": 0.00018111249481112495,
+      "loss": 1.0103,
+      "step": 605
+    },
+    {
+      "epoch": 0.3658315725928162,
+      "grad_norm": 0.0755513533949852,
+      "learning_rate": 0.00018107098381070985,
+      "loss": 0.9527,
+      "step": 606
+    },
+    {
+      "epoch": 0.3664352550558406,
+      "grad_norm": 0.0824398621916771,
+      "learning_rate": 0.00018102947281029475,
+      "loss": 1.032,
+      "step": 607
+    },
+    {
+      "epoch": 0.36703893751886507,
+      "grad_norm": 0.07464707642793655,
+      "learning_rate": 0.00018098796180987962,
+      "loss": 0.9073,
+      "step": 608
+    },
+    {
+      "epoch": 0.3676426199818895,
+      "grad_norm": 0.07036054134368896,
+      "learning_rate": 0.00018094645080946452,
+      "loss": 0.8517,
+      "step": 609
+    },
+    {
+      "epoch": 0.36824630244491396,
+      "grad_norm": 0.10687808692455292,
+      "learning_rate": 0.00018090493980904942,
+      "loss": 0.7676,
+      "step": 610
+    },
+    {
+      "epoch": 0.3688499849079384,
+      "grad_norm": 0.07665056735277176,
+      "learning_rate": 0.0001808634288086343,
+      "loss": 0.9801,
+      "step": 611
+    },
+    {
+      "epoch": 0.36945366737096286,
+      "grad_norm": 0.06765671819448471,
+      "learning_rate": 0.00018082191780821916,
+      "loss": 0.7543,
+      "step": 612
+    },
+    {
+      "epoch": 0.3700573498339873,
+      "grad_norm": 0.07217054814100266,
+      "learning_rate": 0.0001807804068078041,
+      "loss": 1.0022,
+      "step": 613
+    },
+    {
+      "epoch": 0.37066103229701175,
+      "grad_norm": 0.14387869834899902,
+      "learning_rate": 0.00018073889580738896,
+      "loss": 0.7746,
+      "step": 614
+    },
+    {
+      "epoch": 0.3712647147600362,
+      "grad_norm": 0.07988094538450241,
+      "learning_rate": 0.00018069738480697384,
+      "loss": 1.327,
+      "step": 615
+    },
+    {
+      "epoch": 0.37186839722306064,
+      "grad_norm": 0.06843268126249313,
+      "learning_rate": 0.00018065587380655876,
+      "loss": 0.7978,
+      "step": 616
+    },
+    {
+      "epoch": 0.37247207968608514,
+      "grad_norm": 0.0766974464058876,
+      "learning_rate": 0.00018061436280614364,
+      "loss": 0.7766,
+      "step": 617
+    },
+    {
+      "epoch": 0.3730757621491096,
+      "grad_norm": 0.08933945000171661,
+      "learning_rate": 0.0001805728518057285,
+      "loss": 0.7874,
+      "step": 618
+    },
+    {
+      "epoch": 0.37367944461213404,
+      "grad_norm": 0.07440406084060669,
+      "learning_rate": 0.0001805313408053134,
+      "loss": 0.722,
+      "step": 619
+    },
+    {
+      "epoch": 0.3742831270751585,
+      "grad_norm": 0.0785699188709259,
+      "learning_rate": 0.0001804898298048983,
+      "loss": 1.0223,
+      "step": 620
+    },
+    {
+      "epoch": 0.37488680953818293,
+      "grad_norm": 0.07815410941839218,
+      "learning_rate": 0.0001804483188044832,
+      "loss": 0.8886,
+      "step": 621
+    },
+    {
+      "epoch": 0.3754904920012074,
+      "grad_norm": 0.06365236639976501,
+      "learning_rate": 0.00018040680780406808,
+      "loss": 0.7909,
+      "step": 622
+    },
+    {
+      "epoch": 0.3760941744642318,
+      "grad_norm": 0.07768423855304718,
+      "learning_rate": 0.00018036529680365298,
+      "loss": 0.8831,
+      "step": 623
+    },
+    {
+      "epoch": 0.37669785692725627,
+      "grad_norm": 0.07100118696689606,
+      "learning_rate": 0.00018032378580323788,
+      "loss": 0.8649,
+      "step": 624
+    },
+    {
+      "epoch": 0.3773015393902807,
+      "grad_norm": 0.08485390990972519,
+      "learning_rate": 0.00018028227480282276,
+      "loss": 0.731,
+      "step": 625
+    },
+    {
+      "epoch": 0.37790522185330516,
+      "grad_norm": 0.07133664935827255,
+      "learning_rate": 0.00018024076380240763,
+      "loss": 0.7825,
+      "step": 626
+    },
+    {
+      "epoch": 0.3785089043163296,
+      "grad_norm": 0.06745673716068268,
+      "learning_rate": 0.00018019925280199256,
+      "loss": 0.7547,
+      "step": 627
+    },
+    {
+      "epoch": 0.37911258677935405,
+      "grad_norm": 0.08499856293201447,
+      "learning_rate": 0.00018015774180157743,
+      "loss": 0.7289,
+      "step": 628
+    },
+    {
+      "epoch": 0.3797162692423785,
+      "grad_norm": 0.07550381869077682,
+      "learning_rate": 0.0001801162308011623,
+      "loss": 1.0874,
+      "step": 629
+    },
+    {
+      "epoch": 0.38031995170540295,
+      "grad_norm": 0.07187377661466599,
+      "learning_rate": 0.00018007471980074723,
+      "loss": 0.7885,
+      "step": 630
+    },
+    {
+      "epoch": 0.3809236341684274,
+      "grad_norm": 0.07693067193031311,
+      "learning_rate": 0.0001800332088003321,
+      "loss": 0.9349,
+      "step": 631
+    },
+    {
+      "epoch": 0.38152731663145184,
+      "grad_norm": 0.0718834400177002,
+      "learning_rate": 0.00017999169779991698,
+      "loss": 0.8609,
+      "step": 632
+    },
+    {
+      "epoch": 0.3821309990944763,
+      "grad_norm": 0.07122069597244263,
+      "learning_rate": 0.00017995018679950188,
+      "loss": 0.8418,
+      "step": 633
+    },
+    {
+      "epoch": 0.38273468155750073,
+      "grad_norm": 0.0743737518787384,
+      "learning_rate": 0.00017990867579908678,
+      "loss": 0.7195,
+      "step": 634
+    },
+    {
+      "epoch": 0.3833383640205252,
+      "grad_norm": 0.07467867434024811,
+      "learning_rate": 0.00017986716479867165,
+      "loss": 0.7802,
+      "step": 635
+    },
+    {
+      "epoch": 0.3839420464835496,
+      "grad_norm": 0.07535380125045776,
+      "learning_rate": 0.00017982565379825655,
+      "loss": 0.7726,
+      "step": 636
+    },
+    {
+      "epoch": 0.38454572894657413,
+      "grad_norm": 0.07515112310647964,
+      "learning_rate": 0.00017978414279784145,
+      "loss": 0.7936,
+      "step": 637
+    },
+    {
+      "epoch": 0.3851494114095986,
+      "grad_norm": 0.08222845941781998,
+      "learning_rate": 0.00017974263179742632,
+      "loss": 0.7473,
+      "step": 638
+    },
+    {
+      "epoch": 0.385753093872623,
+      "grad_norm": 0.09565534442663193,
+      "learning_rate": 0.00017970112079701122,
+      "loss": 0.7899,
+      "step": 639
+    },
+    {
+      "epoch": 0.38635677633564747,
+      "grad_norm": 0.0862174928188324,
+      "learning_rate": 0.0001796596097965961,
+      "loss": 0.7633,
+      "step": 640
+    },
+    {
+      "epoch": 0.3869604587986719,
+      "grad_norm": 0.08321114629507065,
+      "learning_rate": 0.000179618098796181,
+      "loss": 0.6477,
+      "step": 641
+    },
+    {
+      "epoch": 0.38756414126169636,
+      "grad_norm": 0.09201504290103912,
+      "learning_rate": 0.0001795765877957659,
+      "loss": 0.7508,
+      "step": 642
+    },
+    {
+      "epoch": 0.3881678237247208,
+      "grad_norm": 0.09954454004764557,
+      "learning_rate": 0.00017953507679535077,
+      "loss": 0.7636,
+      "step": 643
+    },
+    {
+      "epoch": 0.38877150618774525,
+      "grad_norm": 0.10669595003128052,
+      "learning_rate": 0.00017949356579493567,
+      "loss": 0.7761,
+      "step": 644
+    },
+    {
+      "epoch": 0.3893751886507697,
+      "grad_norm": 0.09882800281047821,
+      "learning_rate": 0.00017945205479452057,
+      "loss": 0.7333,
+      "step": 645
+    },
+    {
+      "epoch": 0.38997887111379415,
+      "grad_norm": 0.10011550784111023,
+      "learning_rate": 0.00017941054379410544,
+      "loss": 0.6698,
+      "step": 646
+    },
+    {
+      "epoch": 0.3905825535768186,
+      "grad_norm": 0.1063833087682724,
+      "learning_rate": 0.00017936903279369034,
+      "loss": 0.6371,
+      "step": 647
+    },
+    {
+      "epoch": 0.39118623603984304,
+      "grad_norm": 0.11087442189455032,
+      "learning_rate": 0.00017932752179327524,
+      "loss": 0.6137,
+      "step": 648
+    },
+    {
+      "epoch": 0.3917899185028675,
+      "grad_norm": 0.1278572678565979,
+      "learning_rate": 0.0001792860107928601,
+      "loss": 0.6264,
+      "step": 649
+    },
+    {
+      "epoch": 0.39239360096589193,
+      "grad_norm": 0.13009613752365112,
+      "learning_rate": 0.00017924449979244499,
+      "loss": 0.5163,
+      "step": 650
+    },
+    {
+      "epoch": 0.3929972834289164,
+      "grad_norm": 0.08168315887451172,
+      "learning_rate": 0.0001792029887920299,
+      "loss": 0.8955,
+      "step": 651
+    },
+    {
+      "epoch": 0.3936009658919408,
+      "grad_norm": 0.08920527994632721,
+      "learning_rate": 0.00017916147779161479,
+      "loss": 1.0973,
+      "step": 652
+    },
+    {
+      "epoch": 0.39420464835496527,
+      "grad_norm": 0.0853213369846344,
+      "learning_rate": 0.00017911996679119966,
+      "loss": 0.8951,
+      "step": 653
+    },
+    {
+      "epoch": 0.3948083308179897,
+      "grad_norm": 0.07048202306032181,
+      "learning_rate": 0.00017907845579078456,
+      "loss": 0.6935,
+      "step": 654
+    },
+    {
+      "epoch": 0.39541201328101416,
+      "grad_norm": 0.08051397651433945,
+      "learning_rate": 0.00017903694479036946,
+      "loss": 1.1476,
+      "step": 655
+    },
+    {
+      "epoch": 0.3960156957440386,
+      "grad_norm": 0.07060935348272324,
+      "learning_rate": 0.00017899543378995433,
+      "loss": 0.7518,
+      "step": 656
+    },
+    {
+      "epoch": 0.3966193782070631,
+      "grad_norm": 0.0792861208319664,
+      "learning_rate": 0.00017895392278953923,
+      "loss": 0.843,
+      "step": 657
+    },
+    {
+      "epoch": 0.39722306067008756,
+      "grad_norm": 0.07856278121471405,
+      "learning_rate": 0.00017891241178912413,
+      "loss": 0.949,
+      "step": 658
+    },
+    {
+      "epoch": 0.397826743133112,
+      "grad_norm": 0.07348600775003433,
+      "learning_rate": 0.000178870900788709,
+      "loss": 0.902,
+      "step": 659
+    },
+    {
+      "epoch": 0.39843042559613645,
+      "grad_norm": 0.06807007640600204,
+      "learning_rate": 0.0001788293897882939,
+      "loss": 0.8443,
+      "step": 660
+    },
+    {
+      "epoch": 0.3990341080591609,
+      "grad_norm": 0.07370761781930923,
+      "learning_rate": 0.0001787878787878788,
+      "loss": 0.8161,
+      "step": 661
+    },
+    {
+      "epoch": 0.39963779052218534,
+      "grad_norm": 0.08829770982265472,
+      "learning_rate": 0.00017874636778746368,
+      "loss": 0.7679,
+      "step": 662
+    },
+    {
+      "epoch": 0.4002414729852098,
+      "grad_norm": 0.07680663466453552,
+      "learning_rate": 0.00017870485678704858,
+      "loss": 0.8106,
+      "step": 663
+    },
+    {
+      "epoch": 0.40084515544823424,
+      "grad_norm": 0.07080180197954178,
+      "learning_rate": 0.00017866334578663345,
+      "loss": 0.8348,
+      "step": 664
+    },
+    {
+      "epoch": 0.4014488379112587,
+      "grad_norm": 0.0829470232129097,
+      "learning_rate": 0.00017862183478621838,
+      "loss": 0.8589,
+      "step": 665
+    },
+    {
+      "epoch": 0.40205252037428313,
+      "grad_norm": 0.065013587474823,
+      "learning_rate": 0.00017858032378580325,
+      "loss": 0.736,
+      "step": 666
+    },
+    {
+      "epoch": 0.4026562028373076,
+      "grad_norm": 0.07887616008520126,
+      "learning_rate": 0.00017853881278538812,
+      "loss": 0.8547,
+      "step": 667
+    },
+    {
+      "epoch": 0.403259885300332,
+      "grad_norm": 0.07337549328804016,
+      "learning_rate": 0.00017849730178497302,
+      "loss": 0.8766,
+      "step": 668
+    },
+    {
+      "epoch": 0.40386356776335647,
+      "grad_norm": 0.06863244622945786,
+      "learning_rate": 0.00017845579078455792,
+      "loss": 0.7484,
+      "step": 669
+    },
+    {
+      "epoch": 0.4044672502263809,
+      "grad_norm": 0.07943541556596756,
+      "learning_rate": 0.0001784142797841428,
+      "loss": 0.8951,
+      "step": 670
+    },
+    {
+      "epoch": 0.40507093268940536,
+      "grad_norm": 0.07468891143798828,
+      "learning_rate": 0.0001783727687837277,
+      "loss": 0.8805,
+      "step": 671
+    },
+    {
+      "epoch": 0.4056746151524298,
+      "grad_norm": 0.06805651634931564,
+      "learning_rate": 0.0001783312577833126,
+      "loss": 0.9149,
+      "step": 672
+    },
+    {
+      "epoch": 0.40627829761545425,
+      "grad_norm": 0.08728659898042679,
+      "learning_rate": 0.00017828974678289747,
+      "loss": 1.1018,
+      "step": 673
+    },
+    {
+      "epoch": 0.4068819800784787,
+      "grad_norm": 0.07230555266141891,
+      "learning_rate": 0.00017824823578248237,
+      "loss": 0.8377,
+      "step": 674
+    },
+    {
+      "epoch": 0.40748566254150315,
+      "grad_norm": 0.07561583071947098,
+      "learning_rate": 0.00017820672478206727,
+      "loss": 0.8613,
+      "step": 675
+    },
+    {
+      "epoch": 0.4080893450045276,
+      "grad_norm": 0.07437833398580551,
+      "learning_rate": 0.00017816521378165214,
+      "loss": 0.9668,
+      "step": 676
+    },
+    {
+      "epoch": 0.4086930274675521,
+      "grad_norm": 0.09480807930231094,
+      "learning_rate": 0.00017812370278123704,
+      "loss": 0.9128,
+      "step": 677
+    },
+    {
+      "epoch": 0.40929670993057654,
+      "grad_norm": 0.07243344187736511,
+      "learning_rate": 0.00017808219178082192,
+      "loss": 1.0159,
+      "step": 678
+    },
+    {
+      "epoch": 0.409900392393601,
+      "grad_norm": 0.07010567933320999,
+      "learning_rate": 0.00017804068078040682,
+      "loss": 0.7627,
+      "step": 679
+    },
+    {
+      "epoch": 0.41050407485662543,
+      "grad_norm": 0.07201089709997177,
+      "learning_rate": 0.00017799916977999172,
+      "loss": 1.0637,
+      "step": 680
+    },
+    {
+      "epoch": 0.4111077573196499,
+      "grad_norm": 0.0743207260966301,
+      "learning_rate": 0.0001779576587795766,
+      "loss": 0.8637,
+      "step": 681
+    },
+    {
+      "epoch": 0.4117114397826743,
+      "grad_norm": 0.06971647590398788,
+      "learning_rate": 0.0001779161477791615,
+      "loss": 0.8119,
+      "step": 682
+    },
+    {
+      "epoch": 0.4123151222456988,
+      "grad_norm": 0.07709638774394989,
+      "learning_rate": 0.0001778746367787464,
+      "loss": 0.8523,
+      "step": 683
+    },
+    {
+      "epoch": 0.4129188047087232,
+      "grad_norm": 0.07961713522672653,
+      "learning_rate": 0.00017783312577833126,
+      "loss": 0.8088,
+      "step": 684
+    },
+    {
+      "epoch": 0.41352248717174767,
+      "grad_norm": 0.0781940221786499,
+      "learning_rate": 0.00017779161477791616,
+      "loss": 0.7029,
+      "step": 685
+    },
+    {
+      "epoch": 0.4141261696347721,
+      "grad_norm": 0.07617820799350739,
+      "learning_rate": 0.00017775010377750106,
+      "loss": 0.7593,
+      "step": 686
+    },
+    {
+      "epoch": 0.41472985209779656,
+      "grad_norm": 0.08119190484285355,
+      "learning_rate": 0.00017770859277708593,
+      "loss": 0.8067,
+      "step": 687
+    },
+    {
+      "epoch": 0.415333534560821,
+      "grad_norm": 0.07956133782863617,
+      "learning_rate": 0.0001776670817766708,
+      "loss": 0.7693,
+      "step": 688
+    },
+    {
+      "epoch": 0.41593721702384545,
+      "grad_norm": 0.08153844624757767,
+      "learning_rate": 0.00017762557077625574,
+      "loss": 0.7713,
+      "step": 689
+    },
+    {
+      "epoch": 0.4165408994868699,
+      "grad_norm": 0.08829324692487717,
+      "learning_rate": 0.0001775840597758406,
+      "loss": 0.7574,
+      "step": 690
+    },
+    {
+      "epoch": 0.41714458194989434,
+      "grad_norm": 0.09734578430652618,
+      "learning_rate": 0.00017754254877542548,
+      "loss": 0.7962,
+      "step": 691
+    },
+    {
+      "epoch": 0.4177482644129188,
+      "grad_norm": 0.08630529791116714,
+      "learning_rate": 0.00017750103777501038,
+      "loss": 0.7482,
+      "step": 692
+    },
+    {
+      "epoch": 0.41835194687594324,
+      "grad_norm": 0.09447003155946732,
+      "learning_rate": 0.00017745952677459528,
+      "loss": 0.716,
+      "step": 693
+    },
+    {
+      "epoch": 0.4189556293389677,
+      "grad_norm": 0.09416269510984421,
+      "learning_rate": 0.00017741801577418015,
+      "loss": 0.7481,
+      "step": 694
+    },
+    {
+      "epoch": 0.41955931180199213,
+      "grad_norm": 0.10847954452037811,
+      "learning_rate": 0.00017737650477376505,
+      "loss": 0.668,
+      "step": 695
+    },
+    {
+      "epoch": 0.4201629942650166,
+      "grad_norm": 0.10643991082906723,
+      "learning_rate": 0.00017733499377334995,
+      "loss": 0.6973,
+      "step": 696
+    },
+    {
+      "epoch": 0.4207666767280411,
+      "grad_norm": 0.10840442776679993,
+      "learning_rate": 0.00017729348277293483,
+      "loss": 0.6755,
+      "step": 697
+    },
+    {
+      "epoch": 0.4213703591910655,
+      "grad_norm": 0.10803359001874924,
+      "learning_rate": 0.00017725197177251973,
+      "loss": 0.6234,
+      "step": 698
+    },
+    {
+      "epoch": 0.42197404165408997,
+      "grad_norm": 0.12027863413095474,
+      "learning_rate": 0.00017721046077210463,
+      "loss": 0.6337,
+      "step": 699
+    },
+    {
+      "epoch": 0.4225777241171144,
+      "grad_norm": 0.1258537471294403,
+      "learning_rate": 0.0001771689497716895,
+      "loss": 0.447,
+      "step": 700
+    },
+    {
+      "epoch": 0.42318140658013886,
+      "grad_norm": 0.07856849581003189,
+      "learning_rate": 0.0001771274387712744,
+      "loss": 1.0295,
+      "step": 701
+    },
+    {
+      "epoch": 0.4237850890431633,
+      "grad_norm": 0.08135174959897995,
+      "learning_rate": 0.00017708592777085927,
+      "loss": 0.8548,
+      "step": 702
+    },
+    {
+      "epoch": 0.42438877150618776,
+      "grad_norm": 0.08222603797912598,
+      "learning_rate": 0.00017704441677044417,
+      "loss": 0.8463,
+      "step": 703
+    },
+    {
+      "epoch": 0.4249924539692122,
+      "grad_norm": 0.07562818378210068,
+      "learning_rate": 0.00017700290577002907,
+      "loss": 0.7474,
+      "step": 704
+    },
+    {
+      "epoch": 0.42559613643223665,
+      "grad_norm": 0.11416960507631302,
+      "learning_rate": 0.00017696139476961395,
+      "loss": 0.9287,
+      "step": 705
+    },
+    {
+      "epoch": 0.4261998188952611,
+      "grad_norm": 0.07070968300104141,
+      "learning_rate": 0.00017691988376919885,
+      "loss": 0.7686,
+      "step": 706
+    },
+    {
+      "epoch": 0.42680350135828554,
+      "grad_norm": 0.10440545529127121,
+      "learning_rate": 0.00017687837276878375,
+      "loss": 0.8466,
+      "step": 707
+    },
+    {
+      "epoch": 0.42740718382131,
+      "grad_norm": 0.10694804787635803,
+      "learning_rate": 0.00017683686176836862,
+      "loss": 0.7819,
+      "step": 708
+    },
+    {
+      "epoch": 0.42801086628433443,
+      "grad_norm": 0.07630432397127151,
+      "learning_rate": 0.00017679535076795352,
+      "loss": 0.8324,
+      "step": 709
+    },
+    {
+      "epoch": 0.4286145487473589,
+      "grad_norm": 0.07171212136745453,
+      "learning_rate": 0.00017675383976753842,
+      "loss": 0.819,
+      "step": 710
+    },
+    {
+      "epoch": 0.4292182312103833,
+      "grad_norm": 0.06875535845756531,
+      "learning_rate": 0.0001767123287671233,
+      "loss": 0.7489,
+      "step": 711
+    },
+    {
+      "epoch": 0.4298219136734078,
+      "grad_norm": 0.07789517194032669,
+      "learning_rate": 0.0001766708177667082,
+      "loss": 0.8193,
+      "step": 712
+    },
+    {
+      "epoch": 0.4304255961364322,
+      "grad_norm": 0.07542435824871063,
+      "learning_rate": 0.0001766293067662931,
+      "loss": 1.0486,
+      "step": 713
+    },
+    {
+      "epoch": 0.43102927859945667,
+      "grad_norm": 0.07148928195238113,
+      "learning_rate": 0.00017658779576587796,
+      "loss": 1.1104,
+      "step": 714
+    },
+    {
+      "epoch": 0.4316329610624811,
+      "grad_norm": 0.0701369047164917,
+      "learning_rate": 0.00017654628476546286,
+      "loss": 0.8409,
+      "step": 715
+    },
+    {
+      "epoch": 0.43223664352550556,
+      "grad_norm": 0.07325640320777893,
+      "learning_rate": 0.00017650477376504774,
+      "loss": 0.8249,
+      "step": 716
+    },
+    {
+      "epoch": 0.43284032598853,
+      "grad_norm": 0.07507720589637756,
+      "learning_rate": 0.00017646326276463264,
+      "loss": 0.8903,
+      "step": 717
+    },
+    {
+      "epoch": 0.4334440084515545,
+      "grad_norm": 0.07389452308416367,
+      "learning_rate": 0.00017642175176421754,
+      "loss": 0.9039,
+      "step": 718
+    },
+    {
+      "epoch": 0.43404769091457895,
+      "grad_norm": 0.07989213615655899,
+      "learning_rate": 0.0001763802407638024,
+      "loss": 0.8314,
+      "step": 719
+    },
+    {
+      "epoch": 0.4346513733776034,
+      "grad_norm": 0.0715138241648674,
+      "learning_rate": 0.0001763387297633873,
+      "loss": 0.8579,
+      "step": 720
+    },
+    {
+      "epoch": 0.43525505584062785,
+      "grad_norm": 0.09497915208339691,
+      "learning_rate": 0.0001762972187629722,
+      "loss": 0.8342,
+      "step": 721
+    },
+    {
+      "epoch": 0.4358587383036523,
+      "grad_norm": 0.07308909296989441,
+      "learning_rate": 0.00017625570776255708,
+      "loss": 0.87,
+      "step": 722
+    },
+    {
+      "epoch": 0.43646242076667674,
+      "grad_norm": 0.08746104687452316,
+      "learning_rate": 0.00017621419676214196,
+      "loss": 1.1309,
+      "step": 723
+    },
+    {
+      "epoch": 0.4370661032297012,
+      "grad_norm": 0.09257882833480835,
+      "learning_rate": 0.00017617268576172688,
+      "loss": 0.7745,
+      "step": 724
+    },
+    {
+      "epoch": 0.43766978569272563,
+      "grad_norm": 0.08195153623819351,
+      "learning_rate": 0.00017613117476131176,
+      "loss": 0.7545,
+      "step": 725
+    },
+    {
+      "epoch": 0.4382734681557501,
+      "grad_norm": 0.07948730885982513,
+      "learning_rate": 0.00017608966376089663,
+      "loss": 0.8637,
+      "step": 726
+    },
+    {
+      "epoch": 0.4388771506187745,
+      "grad_norm": 0.08171603083610535,
+      "learning_rate": 0.00017604815276048156,
+      "loss": 0.8005,
+      "step": 727
+    },
+    {
+      "epoch": 0.43948083308179897,
+      "grad_norm": 0.0735010951757431,
+      "learning_rate": 0.00017600664176006643,
+      "loss": 0.8667,
+      "step": 728
+    },
+    {
+      "epoch": 0.4400845155448234,
+      "grad_norm": 0.07493755221366882,
+      "learning_rate": 0.0001759651307596513,
+      "loss": 0.8406,
+      "step": 729
+    },
+    {
+      "epoch": 0.44068819800784786,
+      "grad_norm": 0.07382465153932571,
+      "learning_rate": 0.0001759236197592362,
+      "loss": 0.9226,
+      "step": 730
+    },
+    {
+      "epoch": 0.4412918804708723,
+      "grad_norm": 0.10297898948192596,
+      "learning_rate": 0.0001758821087588211,
+      "loss": 0.8471,
+      "step": 731
+    },
+    {
+      "epoch": 0.44189556293389676,
+      "grad_norm": 0.08343175053596497,
+      "learning_rate": 0.00017584059775840598,
+      "loss": 0.8474,
+      "step": 732
+    },
+    {
+      "epoch": 0.4424992453969212,
+      "grad_norm": 0.07784884423017502,
+      "learning_rate": 0.00017579908675799088,
+      "loss": 0.7984,
+      "step": 733
+    },
+    {
+      "epoch": 0.44310292785994565,
+      "grad_norm": 0.07451200485229492,
+      "learning_rate": 0.00017575757575757578,
+      "loss": 0.8766,
+      "step": 734
+    },
+    {
+      "epoch": 0.4437066103229701,
+      "grad_norm": 0.08384682983160019,
+      "learning_rate": 0.00017571606475716065,
+      "loss": 0.806,
+      "step": 735
+    },
+    {
+      "epoch": 0.44431029278599454,
+      "grad_norm": 0.0747980996966362,
+      "learning_rate": 0.00017567455375674555,
+      "loss": 0.6971,
+      "step": 736
+    },
+    {
+      "epoch": 0.444913975249019,
+      "grad_norm": 0.07966098934412003,
+      "learning_rate": 0.00017563304275633042,
+      "loss": 0.8149,
+      "step": 737
+    },
+    {
+      "epoch": 0.4455176577120435,
+      "grad_norm": 0.08828859031200409,
+      "learning_rate": 0.00017559153175591532,
+      "loss": 0.7579,
+      "step": 738
+    },
+    {
+      "epoch": 0.44612134017506794,
+      "grad_norm": 0.08065015077590942,
+      "learning_rate": 0.00017555002075550022,
+      "loss": 0.7766,
+      "step": 739
+    },
+    {
+      "epoch": 0.4467250226380924,
+      "grad_norm": 0.09102361649274826,
+      "learning_rate": 0.0001755085097550851,
+      "loss": 0.8112,
+      "step": 740
+    },
+    {
+      "epoch": 0.44732870510111683,
+      "grad_norm": 0.09253661334514618,
+      "learning_rate": 0.00017546699875467,
+      "loss": 0.7208,
+      "step": 741
+    },
+    {
+      "epoch": 0.4479323875641413,
+      "grad_norm": 0.11388830095529556,
+      "learning_rate": 0.0001754254877542549,
+      "loss": 0.759,
+      "step": 742
+    },
+    {
+      "epoch": 0.4485360700271657,
+      "grad_norm": 0.09276237338781357,
+      "learning_rate": 0.00017538397675383977,
+      "loss": 0.7928,
+      "step": 743
+    },
+    {
+      "epoch": 0.44913975249019017,
+      "grad_norm": 0.09758086502552032,
+      "learning_rate": 0.00017534246575342467,
+      "loss": 0.7086,
+      "step": 744
+    },
+    {
+      "epoch": 0.4497434349532146,
+      "grad_norm": 0.1056419312953949,
+      "learning_rate": 0.00017530095475300957,
+      "loss": 0.7533,
+      "step": 745
+    },
+    {
+      "epoch": 0.45034711741623906,
+      "grad_norm": 0.10619323700666428,
+      "learning_rate": 0.00017525944375259444,
+      "loss": 0.7371,
+      "step": 746
+    },
+    {
+      "epoch": 0.4509507998792635,
+      "grad_norm": 0.13212867081165314,
+      "learning_rate": 0.00017521793275217931,
+      "loss": 0.6729,
+      "step": 747
+    },
+    {
+      "epoch": 0.45155448234228796,
+      "grad_norm": 0.12233047932386398,
+      "learning_rate": 0.00017517642175176424,
+      "loss": 0.6855,
+      "step": 748
+    },
+    {
+      "epoch": 0.4521581648053124,
+      "grad_norm": 0.11795574426651001,
+      "learning_rate": 0.00017513491075134911,
+      "loss": 0.522,
+      "step": 749
+    },
+    {
+      "epoch": 0.45276184726833685,
+      "grad_norm": 0.12232359498739243,
+      "learning_rate": 0.000175093399750934,
+      "loss": 0.5003,
+      "step": 750
+    },
+    {
+      "epoch": 0.4533655297313613,
+      "grad_norm": 0.11266329884529114,
+      "learning_rate": 0.0001750518887505189,
+      "loss": 0.7975,
+      "step": 751
+    },
+    {
+      "epoch": 0.45396921219438574,
+      "grad_norm": 0.12558221817016602,
+      "learning_rate": 0.0001750103777501038,
+      "loss": 1.0336,
+      "step": 752
+    },
+    {
+      "epoch": 0.4545728946574102,
+      "grad_norm": 0.08540918678045273,
+      "learning_rate": 0.0001749688667496887,
+      "loss": 0.8133,
+      "step": 753
+    },
+    {
+      "epoch": 0.45517657712043463,
+      "grad_norm": 0.07365544140338898,
+      "learning_rate": 0.00017492735574927356,
+      "loss": 0.9823,
+      "step": 754
+    },
+    {
+      "epoch": 0.4557802595834591,
+      "grad_norm": 0.07597804814577103,
+      "learning_rate": 0.00017488584474885846,
+      "loss": 0.8333,
+      "step": 755
+    },
+    {
+      "epoch": 0.4563839420464835,
+      "grad_norm": 0.12274730950593948,
+      "learning_rate": 0.00017484433374844336,
+      "loss": 0.8916,
+      "step": 756
+    },
+    {
+      "epoch": 0.456987624509508,
+      "grad_norm": 0.07644704729318619,
+      "learning_rate": 0.00017480282274802823,
+      "loss": 1.2536,
+      "step": 757
+    },
+    {
+      "epoch": 0.4575913069725325,
+      "grad_norm": 0.0743841826915741,
+      "learning_rate": 0.00017476131174761313,
+      "loss": 0.8892,
+      "step": 758
+    },
+    {
+      "epoch": 0.4581949894355569,
+      "grad_norm": 0.07549133896827698,
+      "learning_rate": 0.00017471980074719803,
+      "loss": 0.9356,
+      "step": 759
+    },
+    {
+      "epoch": 0.45879867189858137,
+      "grad_norm": 0.10283336043357849,
+      "learning_rate": 0.0001746782897467829,
+      "loss": 0.8478,
+      "step": 760
+    },
+    {
+      "epoch": 0.4594023543616058,
+      "grad_norm": 0.0773346945643425,
+      "learning_rate": 0.00017463677874636778,
+      "loss": 0.8264,
+      "step": 761
+    },
+    {
+      "epoch": 0.46000603682463026,
+      "grad_norm": 0.08246394246816635,
+      "learning_rate": 0.0001745952677459527,
+      "loss": 0.8789,
+      "step": 762
+    },
+    {
+      "epoch": 0.4606097192876547,
+      "grad_norm": 0.0820598155260086,
+      "learning_rate": 0.00017455375674553758,
+      "loss": 0.8355,
+      "step": 763
+    },
+    {
+      "epoch": 0.46121340175067915,
+      "grad_norm": 0.08660144358873367,
+      "learning_rate": 0.00017451224574512245,
+      "loss": 0.9191,
+      "step": 764
+    },
+    {
+      "epoch": 0.4618170842137036,
+      "grad_norm": 0.07887571305036545,
+      "learning_rate": 0.00017447073474470735,
+      "loss": 0.8801,
+      "step": 765
+    },
+    {
+      "epoch": 0.46242076667672805,
+      "grad_norm": 0.07366171479225159,
+      "learning_rate": 0.00017442922374429225,
+      "loss": 0.9114,
+      "step": 766
+    },
+    {
+      "epoch": 0.4630244491397525,
+      "grad_norm": 0.07119464874267578,
+      "learning_rate": 0.00017438771274387712,
+      "loss": 0.7584,
+      "step": 767
+    },
+    {
+      "epoch": 0.46362813160277694,
+      "grad_norm": 0.07550892978906631,
+      "learning_rate": 0.00017434620174346202,
+      "loss": 0.9126,
+      "step": 768
+    },
+    {
+      "epoch": 0.4642318140658014,
+      "grad_norm": 0.07863172888755798,
+      "learning_rate": 0.00017430469074304692,
+      "loss": 0.8234,
+      "step": 769
+    },
+    {
+      "epoch": 0.46483549652882583,
+      "grad_norm": 0.08048754185438156,
+      "learning_rate": 0.0001742631797426318,
+      "loss": 0.7634,
+      "step": 770
+    },
+    {
+      "epoch": 0.4654391789918503,
+      "grad_norm": 0.0739179328083992,
+      "learning_rate": 0.0001742216687422167,
+      "loss": 0.782,
+      "step": 771
+    },
+    {
+      "epoch": 0.4660428614548747,
+      "grad_norm": 0.07369952648878098,
+      "learning_rate": 0.0001741801577418016,
+      "loss": 0.9965,
+      "step": 772
+    },
+    {
+      "epoch": 0.46664654391789917,
+      "grad_norm": 0.0785854384303093,
+      "learning_rate": 0.00017413864674138647,
+      "loss": 0.9077,
+      "step": 773
+    },
+    {
+      "epoch": 0.4672502263809236,
+      "grad_norm": 0.07083092629909515,
+      "learning_rate": 0.00017409713574097137,
+      "loss": 0.8157,
+      "step": 774
+    },
+    {
+      "epoch": 0.46785390884394806,
+      "grad_norm": 0.07597315311431885,
+      "learning_rate": 0.00017405562474055624,
+      "loss": 1.1588,
+      "step": 775
+    },
+    {
+      "epoch": 0.4684575913069725,
+      "grad_norm": 0.07124720513820648,
+      "learning_rate": 0.00017401411374014114,
+      "loss": 0.8374,
+      "step": 776
+    },
+    {
+      "epoch": 0.46906127376999696,
+      "grad_norm": 0.0762951672077179,
+      "learning_rate": 0.00017397260273972604,
+      "loss": 0.7713,
+      "step": 777
+    },
+    {
+      "epoch": 0.46966495623302146,
+      "grad_norm": 0.09306569397449493,
+      "learning_rate": 0.00017393109173931092,
+      "loss": 0.7561,
+      "step": 778
+    },
+    {
+      "epoch": 0.4702686386960459,
+      "grad_norm": 0.09182307124137878,
+      "learning_rate": 0.00017388958073889582,
+      "loss": 0.816,
+      "step": 779
+    },
+    {
+      "epoch": 0.47087232115907035,
+      "grad_norm": 0.08006258308887482,
+      "learning_rate": 0.00017384806973848072,
+      "loss": 0.7516,
+      "step": 780
+    },
+    {
+      "epoch": 0.4714760036220948,
+      "grad_norm": 0.07114402204751968,
+      "learning_rate": 0.0001738065587380656,
+      "loss": 0.8256,
+      "step": 781
+    },
+    {
+      "epoch": 0.47207968608511924,
+      "grad_norm": 0.08175129443407059,
+      "learning_rate": 0.0001737650477376505,
+      "loss": 0.9031,
+      "step": 782
+    },
+    {
+      "epoch": 0.4726833685481437,
+      "grad_norm": 0.06829746812582016,
+      "learning_rate": 0.0001737235367372354,
+      "loss": 0.7038,
+      "step": 783
+    },
+    {
+      "epoch": 0.47328705101116814,
+      "grad_norm": 0.07961759716272354,
+      "learning_rate": 0.00017368202573682026,
+      "loss": 0.8112,
+      "step": 784
+    },
+    {
+      "epoch": 0.4738907334741926,
+      "grad_norm": 0.08282366394996643,
+      "learning_rate": 0.00017364051473640514,
+      "loss": 0.7348,
+      "step": 785
+    },
+    {
+      "epoch": 0.47449441593721703,
+      "grad_norm": 0.07623188942670822,
+      "learning_rate": 0.00017359900373599006,
+      "loss": 0.7162,
+      "step": 786
+    },
+    {
+      "epoch": 0.4750980984002415,
+      "grad_norm": 0.07791578769683838,
+      "learning_rate": 0.00017355749273557494,
+      "loss": 0.7827,
+      "step": 787
+    },
+    {
+      "epoch": 0.4757017808632659,
+      "grad_norm": 0.07546679675579071,
+      "learning_rate": 0.0001735159817351598,
+      "loss": 0.733,
+      "step": 788
+    },
+    {
+      "epoch": 0.47630546332629037,
+      "grad_norm": 0.08498945832252502,
+      "learning_rate": 0.0001734744707347447,
+      "loss": 0.8845,
+      "step": 789
+    },
+    {
+      "epoch": 0.4769091457893148,
+      "grad_norm": 0.08883391320705414,
+      "learning_rate": 0.0001734329597343296,
+      "loss": 0.802,
+      "step": 790
+    },
+    {
+      "epoch": 0.47751282825233926,
+      "grad_norm": 0.08220940828323364,
+      "learning_rate": 0.00017339144873391448,
+      "loss": 0.7421,
+      "step": 791
+    },
+    {
+      "epoch": 0.4781165107153637,
+      "grad_norm": 0.0887073203921318,
+      "learning_rate": 0.00017334993773349938,
+      "loss": 0.7697,
+      "step": 792
+    },
+    {
+      "epoch": 0.47872019317838815,
+      "grad_norm": 0.09844100475311279,
+      "learning_rate": 0.00017330842673308428,
+      "loss": 0.7502,
+      "step": 793
+    },
+    {
+      "epoch": 0.4793238756414126,
+      "grad_norm": 0.10373784601688385,
+      "learning_rate": 0.00017326691573266915,
+      "loss": 0.7348,
+      "step": 794
+    },
+    {
+      "epoch": 0.47992755810443705,
+      "grad_norm": 0.09787629544734955,
+      "learning_rate": 0.00017322540473225405,
+      "loss": 0.7995,
+      "step": 795
+    },
+    {
+      "epoch": 0.4805312405674615,
+      "grad_norm": 0.10174565017223358,
+      "learning_rate": 0.00017318389373183895,
+      "loss": 0.6651,
+      "step": 796
+    },
+    {
+      "epoch": 0.48113492303048594,
+      "grad_norm": 0.10913451015949249,
+      "learning_rate": 0.00017314238273142385,
+      "loss": 0.6945,
+      "step": 797
+    },
+    {
+      "epoch": 0.48173860549351044,
+      "grad_norm": 0.109565369784832,
+      "learning_rate": 0.00017310087173100873,
+      "loss": 0.6027,
+      "step": 798
+    },
+    {
+      "epoch": 0.4823422879565349,
+      "grad_norm": 0.12080711126327515,
+      "learning_rate": 0.0001730593607305936,
+      "loss": 0.533,
+      "step": 799
+    },
+    {
+      "epoch": 0.48294597041955933,
+      "grad_norm": 0.11455202102661133,
+      "learning_rate": 0.00017301784973017853,
+      "loss": 0.4213,
+      "step": 800
+    },
+    {
+      "epoch": 0.4835496528825838,
+      "grad_norm": 0.08588794618844986,
+      "learning_rate": 0.0001729763387297634,
+      "loss": 0.8639,
+      "step": 801
+    },
+    {
+      "epoch": 0.4841533353456082,
+      "grad_norm": 0.08318472653627396,
+      "learning_rate": 0.00017293482772934827,
+      "loss": 0.9486,
+      "step": 802
+    },
+    {
+      "epoch": 0.4847570178086327,
+      "grad_norm": 0.08030420541763306,
+      "learning_rate": 0.00017289331672893317,
+      "loss": 0.8172,
+      "step": 803
+    },
+    {
+      "epoch": 0.4853607002716571,
+      "grad_norm": 0.08063428103923798,
+      "learning_rate": 0.00017285180572851807,
+      "loss": 0.7898,
+      "step": 804
+    },
+    {
+      "epoch": 0.48596438273468157,
+      "grad_norm": 0.07770851999521255,
+      "learning_rate": 0.00017281029472810295,
+      "loss": 1.0558,
+      "step": 805
+    },
+    {
+      "epoch": 0.486568065197706,
+      "grad_norm": 0.08149967342615128,
+      "learning_rate": 0.00017276878372768785,
+      "loss": 0.8185,
+      "step": 806
+    },
+    {
+      "epoch": 0.48717174766073046,
+      "grad_norm": 0.08028626441955566,
+      "learning_rate": 0.00017272727272727275,
+      "loss": 0.858,
+      "step": 807
+    },
+    {
+      "epoch": 0.4877754301237549,
+      "grad_norm": 0.07975967228412628,
+      "learning_rate": 0.00017268576172685762,
+      "loss": 0.861,
+      "step": 808
+    },
+    {
+      "epoch": 0.48837911258677935,
+      "grad_norm": 0.17568281292915344,
+      "learning_rate": 0.00017264425072644252,
+      "loss": 0.8738,
+      "step": 809
+    },
+    {
+      "epoch": 0.4889827950498038,
+      "grad_norm": 0.0747470036149025,
+      "learning_rate": 0.00017260273972602742,
+      "loss": 1.0316,
+      "step": 810
+    },
+    {
+      "epoch": 0.48958647751282824,
+      "grad_norm": 0.0755024403333664,
+      "learning_rate": 0.0001725612287256123,
+      "loss": 0.8248,
+      "step": 811
+    },
+    {
+      "epoch": 0.4901901599758527,
+      "grad_norm": 0.0713796466588974,
+      "learning_rate": 0.0001725197177251972,
+      "loss": 0.764,
+      "step": 812
+    },
+    {
+      "epoch": 0.49079384243887714,
+      "grad_norm": 0.07347027212381363,
+      "learning_rate": 0.00017247820672478207,
+      "loss": 0.8605,
+      "step": 813
+    },
+    {
+      "epoch": 0.4913975249019016,
+      "grad_norm": 0.07424753159284592,
+      "learning_rate": 0.00017243669572436697,
+      "loss": 0.7838,
+      "step": 814
+    },
+    {
+      "epoch": 0.49200120736492603,
+      "grad_norm": 0.08433883637189865,
+      "learning_rate": 0.00017239518472395187,
+      "loss": 0.8561,
+      "step": 815
+    },
+    {
+      "epoch": 0.4926048898279505,
+      "grad_norm": 0.07728437334299088,
+      "learning_rate": 0.00017235367372353674,
+      "loss": 0.8175,
+      "step": 816
+    },
+    {
+      "epoch": 0.4932085722909749,
+      "grad_norm": 0.0686701089143753,
+      "learning_rate": 0.00017231216272312164,
+      "loss": 0.9421,
+      "step": 817
+    },
+    {
+      "epoch": 0.49381225475399937,
+      "grad_norm": 0.08988066762685776,
+      "learning_rate": 0.00017227065172270654,
+      "loss": 0.8417,
+      "step": 818
+    },
+    {
+      "epoch": 0.49441593721702387,
+      "grad_norm": 0.07794896513223648,
+      "learning_rate": 0.0001722291407222914,
+      "loss": 0.9656,
+      "step": 819
+    },
+    {
+      "epoch": 0.4950196196800483,
+      "grad_norm": 0.0833602100610733,
+      "learning_rate": 0.00017218762972187628,
+      "loss": 0.8488,
+      "step": 820
+    },
+    {
+      "epoch": 0.49562330214307276,
+      "grad_norm": 0.08479736745357513,
+      "learning_rate": 0.0001721461187214612,
+      "loss": 0.8215,
+      "step": 821
+    },
+    {
+      "epoch": 0.4962269846060972,
+      "grad_norm": 0.07611726969480515,
+      "learning_rate": 0.00017210460772104608,
+      "loss": 0.7847,
+      "step": 822
+    },
+    {
+      "epoch": 0.49683066706912166,
+      "grad_norm": 0.07969488203525543,
+      "learning_rate": 0.00017206309672063096,
+      "loss": 0.7988,
+      "step": 823
+    },
+    {
+      "epoch": 0.4974343495321461,
+      "grad_norm": 0.07162702083587646,
+      "learning_rate": 0.00017202158572021588,
+      "loss": 0.8497,
+      "step": 824
+    },
+    {
+      "epoch": 0.49803803199517055,
+      "grad_norm": 0.0738188773393631,
+      "learning_rate": 0.00017198007471980076,
+      "loss": 0.9446,
+      "step": 825
+    },
+    {
+      "epoch": 0.498641714458195,
+      "grad_norm": 0.36350712180137634,
+      "learning_rate": 0.00017193856371938563,
+      "loss": 0.7566,
+      "step": 826
+    },
+    {
+      "epoch": 0.49924539692121944,
+      "grad_norm": 0.0755830928683281,
+      "learning_rate": 0.00017189705271897053,
+      "loss": 1.0117,
+      "step": 827
+    },
+    {
+      "epoch": 0.4998490793842439,
+      "grad_norm": 0.07795432209968567,
+      "learning_rate": 0.00017185554171855543,
+      "loss": 0.8203,
+      "step": 828
+    },
+    {
+      "epoch": 0.5004527618472684,
+      "grad_norm": 0.08904866874217987,
+      "learning_rate": 0.0001718140307181403,
+      "loss": 0.7696,
+      "step": 829
+    },
+    {
+      "epoch": 0.5010564443102928,
+      "grad_norm": 0.07873449474573135,
+      "learning_rate": 0.0001717725197177252,
+      "loss": 0.8633,
+      "step": 830
+    },
+    {
+      "epoch": 0.5016601267733173,
+      "grad_norm": 0.0743393748998642,
+      "learning_rate": 0.0001717310087173101,
+      "loss": 0.8497,
+      "step": 831
+    },
+    {
+      "epoch": 0.5022638092363417,
+      "grad_norm": 0.07956568151712418,
+      "learning_rate": 0.00017168949771689498,
+      "loss": 0.7725,
+      "step": 832
+    },
+    {
+      "epoch": 0.5028674916993662,
+      "grad_norm": 0.07717841863632202,
+      "learning_rate": 0.00017164798671647988,
+      "loss": 0.8644,
+      "step": 833
+    },
+    {
+      "epoch": 0.5034711741623906,
+      "grad_norm": 0.07158881425857544,
+      "learning_rate": 0.00017160647571606475,
+      "loss": 0.7757,
+      "step": 834
+    },
+    {
+      "epoch": 0.5040748566254151,
+      "grad_norm": 0.0767531618475914,
+      "learning_rate": 0.00017156496471564965,
+      "loss": 0.7311,
+      "step": 835
+    },
+    {
+      "epoch": 0.5046785390884395,
+      "grad_norm": 0.07811619341373444,
+      "learning_rate": 0.00017152345371523455,
+      "loss": 0.7659,
+      "step": 836
+    },
+    {
+      "epoch": 0.505282221551464,
+      "grad_norm": 0.07905134558677673,
+      "learning_rate": 0.00017148194271481942,
+      "loss": 0.7226,
+      "step": 837
+    },
+    {
+      "epoch": 0.5058859040144884,
+      "grad_norm": 0.08344464004039764,
+      "learning_rate": 0.00017144043171440432,
+      "loss": 0.7939,
+      "step": 838
+    },
+    {
+      "epoch": 0.5064895864775129,
+      "grad_norm": 0.07940410822629929,
+      "learning_rate": 0.00017139892071398922,
+      "loss": 0.7231,
+      "step": 839
+    },
+    {
+      "epoch": 0.5070932689405373,
+      "grad_norm": 0.08555116504430771,
+      "learning_rate": 0.0001713574097135741,
+      "loss": 0.7344,
+      "step": 840
+    },
+    {
+      "epoch": 0.5076969514035617,
+      "grad_norm": 0.09922269731760025,
+      "learning_rate": 0.000171315898713159,
+      "loss": 0.6382,
+      "step": 841
+    },
+    {
+      "epoch": 0.5083006338665862,
+      "grad_norm": 0.08947370946407318,
+      "learning_rate": 0.0001712743877127439,
+      "loss": 0.7337,
+      "step": 842
+    },
+    {
+      "epoch": 0.5089043163296106,
+      "grad_norm": 0.11347758769989014,
+      "learning_rate": 0.00017123287671232877,
+      "loss": 0.7492,
+      "step": 843
+    },
+    {
+      "epoch": 0.5095079987926351,
+      "grad_norm": 0.09824637323617935,
+      "learning_rate": 0.00017119136571191367,
+      "loss": 0.6762,
+      "step": 844
+    },
+    {
+      "epoch": 0.5101116812556595,
+      "grad_norm": 0.1099235787987709,
+      "learning_rate": 0.00017114985471149857,
+      "loss": 0.7085,
+      "step": 845
+    },
+    {
+      "epoch": 0.510715363718684,
+      "grad_norm": 0.11087027192115784,
+      "learning_rate": 0.00017110834371108344,
+      "loss": 0.7089,
+      "step": 846
+    },
+    {
+      "epoch": 0.5113190461817084,
+      "grad_norm": 0.13687917590141296,
+      "learning_rate": 0.00017106683271066834,
+      "loss": 0.6154,
+      "step": 847
+    },
+    {
+      "epoch": 0.5119227286447329,
+      "grad_norm": 0.11547640711069107,
+      "learning_rate": 0.00017102532171025321,
+      "loss": 0.5773,
+      "step": 848
+    },
+    {
+      "epoch": 0.5125264111077573,
+      "grad_norm": 0.12279754877090454,
+      "learning_rate": 0.00017098381070983811,
+      "loss": 0.5398,
+      "step": 849
+    },
+    {
+      "epoch": 0.5131300935707818,
+      "grad_norm": 0.12324254214763641,
+      "learning_rate": 0.00017094229970942301,
+      "loss": 0.4505,
+      "step": 850
+    },
+    {
+      "epoch": 0.5137337760338062,
+      "grad_norm": 0.07736624777317047,
+      "learning_rate": 0.0001709007887090079,
+      "loss": 0.9161,
+      "step": 851
+    },
+    {
+      "epoch": 0.5143374584968307,
+      "grad_norm": 0.08110132068395615,
+      "learning_rate": 0.0001708592777085928,
+      "loss": 0.7757,
+      "step": 852
+    },
+    {
+      "epoch": 0.5149411409598551,
+      "grad_norm": 0.08207719773054123,
+      "learning_rate": 0.0001708177667081777,
+      "loss": 0.9293,
+      "step": 853
+    },
+    {
+      "epoch": 0.5155448234228796,
+      "grad_norm": 0.07672082632780075,
+      "learning_rate": 0.00017077625570776256,
+      "loss": 0.8568,
+      "step": 854
+    },
+    {
+      "epoch": 0.516148505885904,
+      "grad_norm": 0.07514925301074982,
+      "learning_rate": 0.00017073474470734746,
+      "loss": 0.9212,
+      "step": 855
+    },
+    {
+      "epoch": 0.5167521883489284,
+      "grad_norm": 0.07040940970182419,
+      "learning_rate": 0.00017069323370693236,
+      "loss": 0.7008,
+      "step": 856
+    },
+    {
+      "epoch": 0.5173558708119529,
+      "grad_norm": 0.07006466388702393,
+      "learning_rate": 0.00017065172270651723,
+      "loss": 0.8403,
+      "step": 857
+    },
+    {
+      "epoch": 0.5179595532749773,
+      "grad_norm": 0.0859081968665123,
+      "learning_rate": 0.0001706102117061021,
+      "loss": 0.8658,
+      "step": 858
+    },
+    {
+      "epoch": 0.5185632357380018,
+      "grad_norm": 0.08266257494688034,
+      "learning_rate": 0.00017056870070568703,
+      "loss": 1.0393,
+      "step": 859
+    },
+    {
+      "epoch": 0.5191669182010262,
+      "grad_norm": 0.06764436513185501,
+      "learning_rate": 0.0001705271897052719,
+      "loss": 0.7812,
+      "step": 860
+    },
+    {
+      "epoch": 0.5197706006640507,
+      "grad_norm": 0.0848420113325119,
+      "learning_rate": 0.00017048567870485678,
+      "loss": 0.8968,
+      "step": 861
+    },
+    {
+      "epoch": 0.5203742831270751,
+      "grad_norm": 0.11803008615970612,
+      "learning_rate": 0.00017044416770444168,
+      "loss": 0.8015,
+      "step": 862
+    },
+    {
+      "epoch": 0.5209779655900996,
+      "grad_norm": 0.07445228844881058,
+      "learning_rate": 0.00017040265670402658,
+      "loss": 0.9567,
+      "step": 863
+    },
+    {
+      "epoch": 0.521581648053124,
+      "grad_norm": 0.0766722708940506,
+      "learning_rate": 0.00017036114570361145,
+      "loss": 1.2654,
+      "step": 864
+    },
+    {
+      "epoch": 0.5221853305161485,
+      "grad_norm": 0.0825088769197464,
+      "learning_rate": 0.00017031963470319635,
+      "loss": 1.1549,
+      "step": 865
+    },
+    {
+      "epoch": 0.5227890129791729,
+      "grad_norm": 0.0870489627122879,
+      "learning_rate": 0.00017027812370278125,
+      "loss": 0.945,
+      "step": 866
+    },
+    {
+      "epoch": 0.5233926954421974,
+      "grad_norm": 0.07320686429738998,
+      "learning_rate": 0.00017023661270236613,
+      "loss": 0.7858,
+      "step": 867
+    },
+    {
+      "epoch": 0.5239963779052218,
+      "grad_norm": 0.16626937687397003,
+      "learning_rate": 0.00017019510170195103,
+      "loss": 0.8496,
+      "step": 868
+    },
+    {
+      "epoch": 0.5246000603682464,
+      "grad_norm": 0.07288725674152374,
+      "learning_rate": 0.00017015359070153593,
+      "loss": 1.0189,
+      "step": 869
+    },
+    {
+      "epoch": 0.5252037428312708,
+      "grad_norm": 0.07230374217033386,
+      "learning_rate": 0.0001701120797011208,
+      "loss": 0.7273,
+      "step": 870
+    },
+    {
+      "epoch": 0.5258074252942952,
+      "grad_norm": 0.0818004459142685,
+      "learning_rate": 0.0001700705687007057,
+      "loss": 1.0229,
+      "step": 871
+    },
+    {
+      "epoch": 0.5264111077573197,
+      "grad_norm": 0.08718733489513397,
+      "learning_rate": 0.00017002905770029057,
+      "loss": 0.9639,
+      "step": 872
+    },
+    {
+      "epoch": 0.5270147902203441,
+      "grad_norm": 0.07159387320280075,
+      "learning_rate": 0.00016998754669987547,
+      "loss": 0.796,
+      "step": 873
+    },
+    {
+      "epoch": 0.5276184726833686,
+      "grad_norm": 0.08361268043518066,
+      "learning_rate": 0.00016994603569946037,
+      "loss": 0.8615,
+      "step": 874
+    },
+    {
+      "epoch": 0.528222155146393,
+      "grad_norm": 0.07266636192798615,
+      "learning_rate": 0.00016990452469904524,
+      "loss": 0.7398,
+      "step": 875
+    },
+    {
+      "epoch": 0.5288258376094175,
+      "grad_norm": 0.08292002975940704,
+      "learning_rate": 0.00016986301369863014,
+      "loss": 0.773,
+      "step": 876
+    },
+    {
+      "epoch": 0.5294295200724419,
+      "grad_norm": 0.08378379046916962,
+      "learning_rate": 0.00016982150269821504,
+      "loss": 0.882,
+      "step": 877
+    },
+    {
+      "epoch": 0.5300332025354664,
+      "grad_norm": 0.07400650531053543,
+      "learning_rate": 0.00016977999169779992,
+      "loss": 0.8248,
+      "step": 878
+    },
+    {
+      "epoch": 0.5306368849984908,
+      "grad_norm": 0.07494434714317322,
+      "learning_rate": 0.00016973848069738482,
+      "loss": 1.0832,
+      "step": 879
+    },
+    {
+      "epoch": 0.5312405674615153,
+      "grad_norm": 0.07507789880037308,
+      "learning_rate": 0.00016969696969696972,
+      "loss": 1.004,
+      "step": 880
+    },
+    {
+      "epoch": 0.5318442499245397,
+      "grad_norm": 0.08284460753202438,
+      "learning_rate": 0.0001696554586965546,
+      "loss": 0.7898,
+      "step": 881
+    },
+    {
+      "epoch": 0.5324479323875642,
+      "grad_norm": 0.07790551334619522,
+      "learning_rate": 0.00016961394769613946,
+      "loss": 1.0397,
+      "step": 882
+    },
+    {
+      "epoch": 0.5330516148505886,
+      "grad_norm": 0.07756227999925613,
+      "learning_rate": 0.0001695724366957244,
+      "loss": 0.7647,
+      "step": 883
+    },
+    {
+      "epoch": 0.533655297313613,
+      "grad_norm": 0.0776314064860344,
+      "learning_rate": 0.00016953092569530926,
+      "loss": 0.7354,
+      "step": 884
+    },
+    {
+      "epoch": 0.5342589797766375,
+      "grad_norm": 0.09555763006210327,
+      "learning_rate": 0.00016948941469489414,
+      "loss": 0.7396,
+      "step": 885
+    },
+    {
+      "epoch": 0.534862662239662,
+      "grad_norm": 0.08198418468236923,
+      "learning_rate": 0.00016944790369447904,
+      "loss": 0.8173,
+      "step": 886
+    },
+    {
+      "epoch": 0.5354663447026864,
+      "grad_norm": 0.0823531225323677,
+      "learning_rate": 0.00016940639269406394,
+      "loss": 0.7131,
+      "step": 887
+    },
+    {
+      "epoch": 0.5360700271657108,
+      "grad_norm": 0.08781224489212036,
+      "learning_rate": 0.00016936488169364884,
+      "loss": 0.6971,
+      "step": 888
+    },
+    {
+      "epoch": 0.5366737096287353,
+      "grad_norm": 0.0982750728726387,
+      "learning_rate": 0.0001693233706932337,
+      "loss": 0.8086,
+      "step": 889
+    },
+    {
+      "epoch": 0.5372773920917597,
+      "grad_norm": 0.08869129419326782,
+      "learning_rate": 0.0001692818596928186,
+      "loss": 0.7577,
+      "step": 890
+    },
+    {
+      "epoch": 0.5378810745547842,
+      "grad_norm": 0.08565336465835571,
+      "learning_rate": 0.0001692403486924035,
+      "loss": 0.7097,
+      "step": 891
+    },
+    {
+      "epoch": 0.5384847570178086,
+      "grad_norm": 0.08914349973201752,
+      "learning_rate": 0.00016919883769198838,
+      "loss": 0.6995,
+      "step": 892
+    },
+    {
+      "epoch": 0.5390884394808331,
+      "grad_norm": 0.10026069730520248,
+      "learning_rate": 0.00016915732669157328,
+      "loss": 0.7999,
+      "step": 893
+    },
+    {
+      "epoch": 0.5396921219438575,
+      "grad_norm": 0.1060984879732132,
+      "learning_rate": 0.00016911581569115818,
+      "loss": 0.7441,
+      "step": 894
+    },
+    {
+      "epoch": 0.540295804406882,
+      "grad_norm": 0.10654401034116745,
+      "learning_rate": 0.00016907430469074306,
+      "loss": 0.7129,
+      "step": 895
+    },
+    {
+      "epoch": 0.5408994868699064,
+      "grad_norm": 0.1091652438044548,
+      "learning_rate": 0.00016903279369032793,
+      "loss": 0.6522,
+      "step": 896
+    },
+    {
+      "epoch": 0.5415031693329309,
+      "grad_norm": 0.11495837569236755,
+      "learning_rate": 0.00016899128268991286,
+      "loss": 0.6387,
+      "step": 897
+    },
+    {
+      "epoch": 0.5421068517959553,
+      "grad_norm": 0.1222093477845192,
+      "learning_rate": 0.00016894977168949773,
+      "loss": 0.6086,
+      "step": 898
+    },
+    {
+      "epoch": 0.5427105342589797,
+      "grad_norm": 0.11351132392883301,
+      "learning_rate": 0.0001689082606890826,
+      "loss": 0.5575,
+      "step": 899
+    },
+    {
+      "epoch": 0.5433142167220042,
+      "grad_norm": 0.12081324309110641,
+      "learning_rate": 0.0001688667496886675,
+      "loss": 0.484,
+      "step": 900
+    },
+    {
+      "epoch": 0.5439178991850286,
+      "grad_norm": 0.10098535567522049,
+      "learning_rate": 0.0001688252386882524,
+      "loss": 0.8019,
+      "step": 901
+    },
+    {
+      "epoch": 0.5445215816480531,
+      "grad_norm": 0.07991302758455276,
+      "learning_rate": 0.00016878372768783727,
+      "loss": 0.8477,
+      "step": 902
+    },
+    {
+      "epoch": 0.5451252641110775,
+      "grad_norm": 0.07697251439094543,
+      "learning_rate": 0.00016874221668742217,
+      "loss": 0.7631,
+      "step": 903
+    },
+    {
+      "epoch": 0.545728946574102,
+      "grad_norm": 0.10373490303754807,
+      "learning_rate": 0.00016870070568700707,
+      "loss": 0.9223,
+      "step": 904
+    },
+    {
+      "epoch": 0.5463326290371264,
+      "grad_norm": 0.0777481198310852,
+      "learning_rate": 0.00016865919468659195,
+      "loss": 0.882,
+      "step": 905
+    },
+    {
+      "epoch": 0.5469363115001509,
+      "grad_norm": 0.08426318317651749,
+      "learning_rate": 0.00016861768368617685,
+      "loss": 1.0619,
+      "step": 906
+    },
+    {
+      "epoch": 0.5475399939631753,
+      "grad_norm": 0.07732190936803818,
+      "learning_rate": 0.00016857617268576175,
+      "loss": 0.7125,
+      "step": 907
+    },
+    {
+      "epoch": 0.5481436764261998,
+      "grad_norm": 0.07979828119277954,
+      "learning_rate": 0.00016853466168534662,
+      "loss": 0.8596,
+      "step": 908
+    },
+    {
+      "epoch": 0.5487473588892242,
+      "grad_norm": 0.10036496073007584,
+      "learning_rate": 0.00016849315068493152,
+      "loss": 0.8396,
+      "step": 909
+    },
+    {
+      "epoch": 0.5493510413522488,
+      "grad_norm": 0.07785385847091675,
+      "learning_rate": 0.0001684516396845164,
+      "loss": 0.7962,
+      "step": 910
+    },
+    {
+      "epoch": 0.5499547238152732,
+      "grad_norm": 0.07757363468408585,
+      "learning_rate": 0.0001684101286841013,
+      "loss": 0.9085,
+      "step": 911
+    },
+    {
+      "epoch": 0.5505584062782977,
+      "grad_norm": 0.07866904139518738,
+      "learning_rate": 0.0001683686176836862,
+      "loss": 0.7676,
+      "step": 912
+    },
+    {
+      "epoch": 0.5511620887413221,
+      "grad_norm": 0.08171574026346207,
+      "learning_rate": 0.00016832710668327107,
+      "loss": 0.8823,
+      "step": 913
+    },
+    {
+      "epoch": 0.5517657712043466,
+      "grad_norm": 0.09291453659534454,
+      "learning_rate": 0.00016828559568285597,
+      "loss": 0.8212,
+      "step": 914
+    },
+    {
+      "epoch": 0.552369453667371,
+      "grad_norm": 0.06771212071180344,
+      "learning_rate": 0.00016824408468244087,
+      "loss": 0.8255,
+      "step": 915
+    },
+    {
+      "epoch": 0.5529731361303954,
+      "grad_norm": 0.07624577730894089,
+      "learning_rate": 0.00016820257368202574,
+      "loss": 1.1116,
+      "step": 916
+    },
+    {
+      "epoch": 0.5535768185934199,
+      "grad_norm": 0.06806713342666626,
+      "learning_rate": 0.00016816106268161064,
+      "loss": 0.7341,
+      "step": 917
+    },
+    {
+      "epoch": 0.5541805010564443,
+      "grad_norm": 0.1294786036014557,
+      "learning_rate": 0.00016811955168119554,
+      "loss": 0.8429,
+      "step": 918
+    },
+    {
+      "epoch": 0.5547841835194688,
+      "grad_norm": 0.08023308962583542,
+      "learning_rate": 0.0001680780406807804,
+      "loss": 1.1129,
+      "step": 919
+    },
+    {
+      "epoch": 0.5553878659824932,
+      "grad_norm": 0.09215756505727768,
+      "learning_rate": 0.00016803652968036529,
+      "loss": 0.935,
+      "step": 920
+    },
+    {
+      "epoch": 0.5559915484455177,
+      "grad_norm": 0.11087724566459656,
+      "learning_rate": 0.0001679950186799502,
+      "loss": 0.764,
+      "step": 921
+    },
+    {
+      "epoch": 0.5565952309085421,
+      "grad_norm": 0.07170010358095169,
+      "learning_rate": 0.00016795350767953509,
+      "loss": 0.71,
+      "step": 922
+    },
+    {
+      "epoch": 0.5571989133715666,
+      "grad_norm": 0.07650398463010788,
+      "learning_rate": 0.00016791199667911996,
+      "loss": 0.7885,
+      "step": 923
+    },
+    {
+      "epoch": 0.557802595834591,
+      "grad_norm": 0.08096078783273697,
+      "learning_rate": 0.00016787048567870486,
+      "loss": 0.7443,
+      "step": 924
+    },
+    {
+      "epoch": 0.5584062782976155,
+      "grad_norm": 0.07324434071779251,
+      "learning_rate": 0.00016782897467828976,
+      "loss": 0.8045,
+      "step": 925
+    },
+    {
+      "epoch": 0.5590099607606399,
+      "grad_norm": 0.11291930824518204,
+      "learning_rate": 0.00016778746367787463,
+      "loss": 0.8388,
+      "step": 926
+    },
+    {
+      "epoch": 0.5596136432236644,
+      "grad_norm": 0.07567749172449112,
+      "learning_rate": 0.00016774595267745953,
+      "loss": 0.8203,
+      "step": 927
+    },
+    {
+      "epoch": 0.5602173256866888,
+      "grad_norm": 0.07367605715990067,
+      "learning_rate": 0.00016770444167704443,
+      "loss": 0.7428,
+      "step": 928
+    },
+    {
+      "epoch": 0.5608210081497133,
+      "grad_norm": 0.11504574120044708,
+      "learning_rate": 0.0001676629306766293,
+      "loss": 1.0331,
+      "step": 929
+    },
+    {
+      "epoch": 0.5614246906127377,
+      "grad_norm": 0.06944482028484344,
+      "learning_rate": 0.0001676214196762142,
+      "loss": 1.0409,
+      "step": 930
+    },
+    {
+      "epoch": 0.5620283730757621,
+      "grad_norm": 0.07487515360116959,
+      "learning_rate": 0.0001675799086757991,
+      "loss": 0.747,
+      "step": 931
+    },
+    {
+      "epoch": 0.5626320555387866,
+      "grad_norm": 0.09043581038713455,
+      "learning_rate": 0.000167538397675384,
+      "loss": 0.8308,
+      "step": 932
+    },
+    {
+      "epoch": 0.563235738001811,
+      "grad_norm": 0.07596288621425629,
+      "learning_rate": 0.00016749688667496888,
+      "loss": 0.6969,
+      "step": 933
+    },
+    {
+      "epoch": 0.5638394204648355,
+      "grad_norm": 0.08029479533433914,
+      "learning_rate": 0.00016745537567455375,
+      "loss": 0.7532,
+      "step": 934
+    },
+    {
+      "epoch": 0.5644431029278599,
+      "grad_norm": 0.08100748807191849,
+      "learning_rate": 0.00016741386467413868,
+      "loss": 0.7859,
+      "step": 935
+    },
+    {
+      "epoch": 0.5650467853908844,
+      "grad_norm": 0.08451302349567413,
+      "learning_rate": 0.00016737235367372355,
+      "loss": 0.7942,
+      "step": 936
+    },
+    {
+      "epoch": 0.5656504678539088,
+      "grad_norm": 0.0821182057261467,
+      "learning_rate": 0.00016733084267330842,
+      "loss": 0.7002,
+      "step": 937
+    },
+    {
+      "epoch": 0.5662541503169333,
+      "grad_norm": 0.08984285593032837,
+      "learning_rate": 0.00016728933167289332,
+      "loss": 0.7721,
+      "step": 938
+    },
+    {
+      "epoch": 0.5668578327799577,
+      "grad_norm": 0.09170673787593842,
+      "learning_rate": 0.00016724782067247822,
+      "loss": 0.7705,
+      "step": 939
+    },
+    {
+      "epoch": 0.5674615152429822,
+      "grad_norm": 0.09316025674343109,
+      "learning_rate": 0.0001672063096720631,
+      "loss": 0.7349,
+      "step": 940
+    },
+    {
+      "epoch": 0.5680651977060066,
+      "grad_norm": 0.09121593087911606,
+      "learning_rate": 0.000167164798671648,
+      "loss": 0.7074,
+      "step": 941
+    },
+    {
+      "epoch": 0.568668880169031,
+      "grad_norm": 0.099585622549057,
+      "learning_rate": 0.0001671232876712329,
+      "loss": 0.772,
+      "step": 942
+    },
+    {
+      "epoch": 0.5692725626320555,
+      "grad_norm": 0.09845862537622452,
+      "learning_rate": 0.00016708177667081777,
+      "loss": 0.6837,
+      "step": 943
+    },
+    {
+      "epoch": 0.56987624509508,
+      "grad_norm": 0.10467004030942917,
+      "learning_rate": 0.00016704026567040267,
+      "loss": 0.7469,
+      "step": 944
+    },
+    {
+      "epoch": 0.5704799275581044,
+      "grad_norm": 0.10466741770505905,
+      "learning_rate": 0.00016699875466998757,
+      "loss": 0.709,
+      "step": 945
+    },
+    {
+      "epoch": 0.5710836100211288,
+      "grad_norm": 0.11000916361808777,
+      "learning_rate": 0.00016695724366957244,
+      "loss": 0.6697,
+      "step": 946
+    },
+    {
+      "epoch": 0.5716872924841533,
+      "grad_norm": 0.1157151609659195,
+      "learning_rate": 0.00016691573266915734,
+      "loss": 0.6396,
+      "step": 947
+    },
+    {
+      "epoch": 0.5722909749471777,
+      "grad_norm": 0.11827955394983292,
+      "learning_rate": 0.00016687422166874222,
+      "loss": 0.6224,
+      "step": 948
+    },
+    {
+      "epoch": 0.5728946574102022,
+      "grad_norm": 0.12374337762594223,
+      "learning_rate": 0.00016683271066832712,
+      "loss": 0.5617,
+      "step": 949
+    },
+    {
+      "epoch": 0.5734983398732267,
+      "grad_norm": 0.12392137944698334,
+      "learning_rate": 0.00016679119966791202,
+      "loss": 0.4256,
+      "step": 950
+    },
+    {
+      "epoch": 0.5741020223362512,
+      "grad_norm": 0.09002061933279037,
+      "learning_rate": 0.0001667496886674969,
+      "loss": 0.8641,
+      "step": 951
+    },
+    {
+      "epoch": 0.5747057047992756,
+      "grad_norm": 0.0768359899520874,
+      "learning_rate": 0.0001667081776670818,
+      "loss": 0.6958,
+      "step": 952
+    },
+    {
+      "epoch": 0.5753093872623001,
+      "grad_norm": 0.07474057376384735,
+      "learning_rate": 0.0001666666666666667,
+      "loss": 0.9245,
+      "step": 953
+    },
+    {
+      "epoch": 0.5759130697253245,
+      "grad_norm": 0.0858532041311264,
+      "learning_rate": 0.00016662515566625156,
+      "loss": 0.801,
+      "step": 954
+    },
+    {
+      "epoch": 0.576516752188349,
+      "grad_norm": 0.08939801156520844,
+      "learning_rate": 0.00016658364466583643,
+      "loss": 0.8461,
+      "step": 955
+    },
+    {
+      "epoch": 0.5771204346513734,
+      "grad_norm": 0.07809244841337204,
+      "learning_rate": 0.00016654213366542136,
+      "loss": 0.9988,
+      "step": 956
+    },
+    {
+      "epoch": 0.5777241171143979,
+      "grad_norm": 0.08366213738918304,
+      "learning_rate": 0.00016650062266500623,
+      "loss": 0.8169,
+      "step": 957
+    },
+    {
+      "epoch": 0.5783277995774223,
+      "grad_norm": 0.08782380819320679,
+      "learning_rate": 0.0001664591116645911,
+      "loss": 0.8715,
+      "step": 958
+    },
+    {
+      "epoch": 0.5789314820404468,
+      "grad_norm": 0.0959751307964325,
+      "learning_rate": 0.00016641760066417603,
+      "loss": 0.9998,
+      "step": 959
+    },
+    {
+      "epoch": 0.5795351645034712,
+      "grad_norm": 0.07628454267978668,
+      "learning_rate": 0.0001663760896637609,
+      "loss": 0.8369,
+      "step": 960
+    },
+    {
+      "epoch": 0.5801388469664956,
+      "grad_norm": 0.0932275727391243,
+      "learning_rate": 0.00016633457866334578,
+      "loss": 1.0541,
+      "step": 961
+    },
+    {
+      "epoch": 0.5807425294295201,
+      "grad_norm": 0.09154736250638962,
+      "learning_rate": 0.00016629306766293068,
+      "loss": 0.7972,
+      "step": 962
+    },
+    {
+      "epoch": 0.5813462118925445,
+      "grad_norm": 0.09588005393743515,
+      "learning_rate": 0.00016625155666251558,
+      "loss": 0.9873,
+      "step": 963
+    },
+    {
+      "epoch": 0.581949894355569,
+      "grad_norm": 0.06718328595161438,
+      "learning_rate": 0.00016621004566210045,
+      "loss": 0.839,
+      "step": 964
+    },
+    {
+      "epoch": 0.5825535768185934,
+      "grad_norm": 0.09526249766349792,
+      "learning_rate": 0.00016616853466168535,
+      "loss": 0.8999,
+      "step": 965
+    },
+    {
+      "epoch": 0.5831572592816179,
+      "grad_norm": 0.0714835673570633,
+      "learning_rate": 0.00016612702366127025,
+      "loss": 0.7127,
+      "step": 966
+    },
+    {
+      "epoch": 0.5837609417446423,
+      "grad_norm": 0.07606557011604309,
+      "learning_rate": 0.00016608551266085513,
+      "loss": 0.8162,
+      "step": 967
+    },
+    {
+      "epoch": 0.5843646242076668,
+      "grad_norm": 0.07503029704093933,
+      "learning_rate": 0.00016604400166044003,
+      "loss": 0.9174,
+      "step": 968
+    },
+    {
+      "epoch": 0.5849683066706912,
+      "grad_norm": 0.08567387610673904,
+      "learning_rate": 0.0001660024906600249,
+      "loss": 0.9408,
+      "step": 969
+    },
+    {
+      "epoch": 0.5855719891337157,
+      "grad_norm": 0.08126292377710342,
+      "learning_rate": 0.0001659609796596098,
+      "loss": 0.9288,
+      "step": 970
+    },
+    {
+      "epoch": 0.5861756715967401,
+      "grad_norm": 0.07483746111392975,
+      "learning_rate": 0.0001659194686591947,
+      "loss": 0.8598,
+      "step": 971
+    },
+    {
+      "epoch": 0.5867793540597646,
+      "grad_norm": 0.07251130044460297,
+      "learning_rate": 0.00016587795765877957,
+      "loss": 0.9461,
+      "step": 972
+    },
+    {
+      "epoch": 0.587383036522789,
+      "grad_norm": 0.08908132463693619,
+      "learning_rate": 0.00016583644665836447,
+      "loss": 0.9473,
+      "step": 973
+    },
+    {
+      "epoch": 0.5879867189858135,
+      "grad_norm": 0.08767648786306381,
+      "learning_rate": 0.00016579493565794937,
+      "loss": 0.757,
+      "step": 974
+    },
+    {
+      "epoch": 0.5885904014488379,
+      "grad_norm": 0.07573854178190231,
+      "learning_rate": 0.00016575342465753425,
+      "loss": 0.7972,
+      "step": 975
+    },
+    {
+      "epoch": 0.5891940839118623,
+      "grad_norm": 0.074210524559021,
+      "learning_rate": 0.00016571191365711915,
+      "loss": 0.8437,
+      "step": 976
+    },
+    {
+      "epoch": 0.5897977663748868,
+      "grad_norm": 0.08098217099905014,
+      "learning_rate": 0.00016567040265670405,
+      "loss": 0.9021,
+      "step": 977
+    },
+    {
+      "epoch": 0.5904014488379112,
+      "grad_norm": 0.07945941388607025,
+      "learning_rate": 0.00016562889165628892,
+      "loss": 0.8371,
+      "step": 978
+    },
+    {
+      "epoch": 0.5910051313009357,
+      "grad_norm": 0.10039619356393814,
+      "learning_rate": 0.00016558738065587382,
+      "loss": 0.7602,
+      "step": 979
+    },
+    {
+      "epoch": 0.5916088137639601,
+      "grad_norm": 0.08295184373855591,
+      "learning_rate": 0.00016554586965545872,
+      "loss": 0.761,
+      "step": 980
+    },
+    {
+      "epoch": 0.5922124962269846,
+      "grad_norm": 0.08036410063505173,
+      "learning_rate": 0.0001655043586550436,
+      "loss": 0.9951,
+      "step": 981
+    },
+    {
+      "epoch": 0.592816178690009,
+      "grad_norm": 0.07915826141834259,
+      "learning_rate": 0.0001654628476546285,
+      "loss": 0.82,
+      "step": 982
+    },
+    {
+      "epoch": 0.5934198611530335,
+      "grad_norm": 0.07575780898332596,
+      "learning_rate": 0.00016542133665421336,
+      "loss": 0.7732,
+      "step": 983
+    },
+    {
+      "epoch": 0.5940235436160579,
+      "grad_norm": 0.08057866990566254,
+      "learning_rate": 0.00016537982565379826,
+      "loss": 0.7388,
+      "step": 984
+    },
+    {
+      "epoch": 0.5946272260790824,
+      "grad_norm": 0.07897903025150299,
+      "learning_rate": 0.00016533831465338316,
+      "loss": 0.733,
+      "step": 985
+    },
+    {
+      "epoch": 0.5952309085421068,
+      "grad_norm": 0.1108403429389,
+      "learning_rate": 0.00016529680365296804,
+      "loss": 0.7831,
+      "step": 986
+    },
+    {
+      "epoch": 0.5958345910051313,
+      "grad_norm": 0.08219526708126068,
+      "learning_rate": 0.00016525529265255294,
+      "loss": 0.754,
+      "step": 987
+    },
+    {
+      "epoch": 0.5964382734681557,
+      "grad_norm": 0.09057065844535828,
+      "learning_rate": 0.00016521378165213784,
+      "loss": 0.7478,
+      "step": 988
+    },
+    {
+      "epoch": 0.5970419559311801,
+      "grad_norm": 0.09317967295646667,
+      "learning_rate": 0.0001651722706517227,
+      "loss": 0.7583,
+      "step": 989
+    },
+    {
+      "epoch": 0.5976456383942047,
+      "grad_norm": 0.09358835965394974,
+      "learning_rate": 0.0001651307596513076,
+      "loss": 0.6958,
+      "step": 990
+    },
+    {
+      "epoch": 0.5982493208572291,
+      "grad_norm": 0.100360207259655,
+      "learning_rate": 0.0001650892486508925,
+      "loss": 0.7544,
+      "step": 991
+    },
+    {
+      "epoch": 0.5988530033202536,
+      "grad_norm": 0.09013186395168304,
+      "learning_rate": 0.00016504773765047738,
+      "loss": 0.7112,
+      "step": 992
+    },
+    {
+      "epoch": 0.599456685783278,
+      "grad_norm": 0.1008266881108284,
+      "learning_rate": 0.00016500622665006226,
+      "loss": 0.767,
+      "step": 993
+    },
+    {
+      "epoch": 0.6000603682463025,
+      "grad_norm": 0.09665394574403763,
+      "learning_rate": 0.00016496471564964718,
+      "loss": 0.6782,
+      "step": 994
+    },
+    {
+      "epoch": 0.6006640507093269,
+      "grad_norm": 0.10552255064249039,
+      "learning_rate": 0.00016492320464923206,
+      "loss": 0.6667,
+      "step": 995
+    },
+    {
+      "epoch": 0.6012677331723514,
+      "grad_norm": 0.11099070310592651,
+      "learning_rate": 0.00016488169364881693,
+      "loss": 0.7023,
+      "step": 996
+    },
+    {
+      "epoch": 0.6018714156353758,
+      "grad_norm": 0.10965953767299652,
+      "learning_rate": 0.00016484018264840183,
+      "loss": 0.6285,
+      "step": 997
+    },
+    {
+      "epoch": 0.6024750980984003,
+      "grad_norm": 0.11270109564065933,
+      "learning_rate": 0.00016479867164798673,
+      "loss": 0.612,
+      "step": 998
+    },
+    {
+      "epoch": 0.6030787805614247,
+      "grad_norm": 0.11348733305931091,
+      "learning_rate": 0.0001647571606475716,
+      "loss": 0.5591,
+      "step": 999
+    },
+    {
+      "epoch": 0.6036824630244492,
+      "grad_norm": 0.1182008683681488,
+      "learning_rate": 0.0001647156496471565,
+      "loss": 0.4379,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6036824630244492,
+      "eval_loss": 0.8275489211082458,
+      "eval_runtime": 1218.7348,
+      "eval_samples_per_second": 2.289,
+      "eval_steps_per_second": 0.286,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6042861454874736,
+      "grad_norm": 0.0789058580994606,
+      "learning_rate": 0.0001646741386467414,
+      "loss": 0.7819,
+      "step": 1001
+    },
+    {
+      "epoch": 0.6048898279504981,
+      "grad_norm": 0.0920637771487236,
+      "learning_rate": 0.00016463262764632627,
+      "loss": 1.4056,
+      "step": 1002
+    },
+    {
+      "epoch": 0.6054935104135225,
+      "grad_norm": 0.07369931042194366,
+      "learning_rate": 0.00016459111664591118,
+      "loss": 0.9731,
+      "step": 1003
+    },
+    {
+      "epoch": 0.606097192876547,
+      "grad_norm": 0.08552141487598419,
+      "learning_rate": 0.00016454960564549608,
+      "loss": 0.7628,
+      "step": 1004
+    },
+    {
+      "epoch": 0.6067008753395714,
+      "grad_norm": 0.07296184450387955,
+      "learning_rate": 0.00016450809464508095,
+      "loss": 0.7961,
+      "step": 1005
+    },
+    {
+      "epoch": 0.6073045578025958,
+      "grad_norm": 0.08106445521116257,
+      "learning_rate": 0.00016446658364466585,
+      "loss": 0.9692,
+      "step": 1006
+    },
+    {
+      "epoch": 0.6079082402656203,
+      "grad_norm": 0.08443611115217209,
+      "learning_rate": 0.00016442507264425072,
+      "loss": 1.3388,
+      "step": 1007
+    },
+    {
+      "epoch": 0.6085119227286447,
+      "grad_norm": 0.08953455090522766,
+      "learning_rate": 0.00016438356164383562,
+      "loss": 0.9006,
+      "step": 1008
+    },
+    {
+      "epoch": 0.6091156051916692,
+      "grad_norm": 0.09113611280918121,
+      "learning_rate": 0.00016434205064342052,
+      "loss": 0.9404,
+      "step": 1009
+    },
+    {
+      "epoch": 0.6097192876546936,
+      "grad_norm": 0.08092870563268661,
+      "learning_rate": 0.0001643005396430054,
+      "loss": 0.8747,
+      "step": 1010
+    },
+    {
+      "epoch": 0.6103229701177181,
+      "grad_norm": 0.07962828874588013,
+      "learning_rate": 0.0001642590286425903,
+      "loss": 1.0154,
+      "step": 1011
+    },
+    {
+      "epoch": 0.6109266525807425,
+      "grad_norm": 0.07279300689697266,
+      "learning_rate": 0.0001642175176421752,
+      "loss": 0.8086,
+      "step": 1012
+    },
+    {
+      "epoch": 0.611530335043767,
+      "grad_norm": 0.08411548286676407,
+      "learning_rate": 0.00016417600664176007,
+      "loss": 0.8582,
+      "step": 1013
+    },
+    {
+      "epoch": 0.6121340175067914,
+      "grad_norm": 0.0790749341249466,
+      "learning_rate": 0.00016413449564134497,
+      "loss": 0.7265,
+      "step": 1014
+    },
+    {
+      "epoch": 0.6127376999698159,
+      "grad_norm": 0.0777488648891449,
+      "learning_rate": 0.00016409298464092987,
+      "loss": 0.8086,
+      "step": 1015
+    },
+    {
+      "epoch": 0.6133413824328403,
+      "grad_norm": 0.09185469895601273,
+      "learning_rate": 0.00016405147364051474,
+      "loss": 0.8752,
+      "step": 1016
+    },
+    {
+      "epoch": 0.6139450648958648,
+      "grad_norm": 0.09490189701318741,
+      "learning_rate": 0.0001640099626400996,
+      "loss": 0.9276,
+      "step": 1017
+    },
+    {
+      "epoch": 0.6145487473588892,
+      "grad_norm": 0.10677061229944229,
+      "learning_rate": 0.00016396845163968454,
+      "loss": 1.1395,
+      "step": 1018
+    },
+    {
+      "epoch": 0.6151524298219136,
+      "grad_norm": 0.08552467823028564,
+      "learning_rate": 0.0001639269406392694,
+      "loss": 1.0333,
+      "step": 1019
+    },
+    {
+      "epoch": 0.6157561122849381,
+      "grad_norm": 0.10448099672794342,
+      "learning_rate": 0.0001638854296388543,
+      "loss": 1.003,
+      "step": 1020
+    },
+    {
+      "epoch": 0.6163597947479625,
+      "grad_norm": 0.08592557907104492,
+      "learning_rate": 0.00016384391863843919,
+      "loss": 0.7925,
+      "step": 1021
+    },
+    {
+      "epoch": 0.616963477210987,
+      "grad_norm": 0.082502081990242,
+      "learning_rate": 0.00016380240763802409,
+      "loss": 0.8321,
+      "step": 1022
+    },
+    {
+      "epoch": 0.6175671596740114,
+      "grad_norm": 0.08332712948322296,
+      "learning_rate": 0.00016376089663760899,
+      "loss": 0.9372,
+      "step": 1023
+    },
+    {
+      "epoch": 0.6181708421370359,
+      "grad_norm": 0.09017034620046616,
+      "learning_rate": 0.00016371938563719386,
+      "loss": 0.8795,
+      "step": 1024
+    },
+    {
+      "epoch": 0.6187745246000603,
+      "grad_norm": 0.08691825717687607,
+      "learning_rate": 0.00016367787463677876,
+      "loss": 0.7948,
+      "step": 1025
+    },
+    {
+      "epoch": 0.6193782070630848,
+      "grad_norm": 0.07839145511388779,
+      "learning_rate": 0.00016363636363636366,
+      "loss": 0.7879,
+      "step": 1026
+    },
+    {
+      "epoch": 0.6199818895261092,
+      "grad_norm": 0.09319109469652176,
+      "learning_rate": 0.00016359485263594853,
+      "loss": 0.8614,
+      "step": 1027
+    },
+    {
+      "epoch": 0.6205855719891337,
+      "grad_norm": 0.0755649283528328,
+      "learning_rate": 0.00016355334163553343,
+      "loss": 0.7723,
+      "step": 1028
+    },
+    {
+      "epoch": 0.6211892544521581,
+      "grad_norm": 0.08007936924695969,
+      "learning_rate": 0.00016351183063511833,
+      "loss": 0.7863,
+      "step": 1029
+    },
+    {
+      "epoch": 0.6217929369151826,
+      "grad_norm": 0.11180046945810318,
+      "learning_rate": 0.0001634703196347032,
+      "loss": 1.0614,
+      "step": 1030
+    },
+    {
+      "epoch": 0.6223966193782071,
+      "grad_norm": 0.09498503804206848,
+      "learning_rate": 0.00016342880863428808,
+      "loss": 0.8302,
+      "step": 1031
+    },
+    {
+      "epoch": 0.6230003018412316,
+      "grad_norm": 0.08199071139097214,
+      "learning_rate": 0.000163387297633873,
+      "loss": 1.0062,
+      "step": 1032
+    },
+    {
+      "epoch": 0.623603984304256,
+      "grad_norm": 0.08216526359319687,
+      "learning_rate": 0.00016334578663345788,
+      "loss": 0.6999,
+      "step": 1033
+    },
+    {
+      "epoch": 0.6242076667672805,
+      "grad_norm": 0.07529989629983902,
+      "learning_rate": 0.00016330427563304275,
+      "loss": 0.8036,
+      "step": 1034
+    },
+    {
+      "epoch": 0.6248113492303049,
+      "grad_norm": 0.08288833498954773,
+      "learning_rate": 0.00016326276463262765,
+      "loss": 0.7338,
+      "step": 1035
+    },
+    {
+      "epoch": 0.6254150316933293,
+      "grad_norm": 0.08039028942584991,
+      "learning_rate": 0.00016322125363221255,
+      "loss": 0.7554,
+      "step": 1036
+    },
+    {
+      "epoch": 0.6260187141563538,
+      "grad_norm": 0.08144210278987885,
+      "learning_rate": 0.00016317974263179742,
+      "loss": 0.7421,
+      "step": 1037
+    },
+    {
+      "epoch": 0.6266223966193782,
+      "grad_norm": 0.08654642850160599,
+      "learning_rate": 0.00016313823163138232,
+      "loss": 0.7317,
+      "step": 1038
+    },
+    {
+      "epoch": 0.6272260790824027,
+      "grad_norm": 0.08834250271320343,
+      "learning_rate": 0.00016309672063096722,
+      "loss": 0.7208,
+      "step": 1039
+    },
+    {
+      "epoch": 0.6278297615454271,
+      "grad_norm": 0.08896996825933456,
+      "learning_rate": 0.0001630552096305521,
+      "loss": 0.7637,
+      "step": 1040
+    },
+    {
+      "epoch": 0.6284334440084516,
+      "grad_norm": 0.09299125522375107,
+      "learning_rate": 0.000163013698630137,
+      "loss": 0.6993,
+      "step": 1041
+    },
+    {
+      "epoch": 0.629037126471476,
+      "grad_norm": 0.0911908969283104,
+      "learning_rate": 0.0001629721876297219,
+      "loss": 0.7455,
+      "step": 1042
+    },
+    {
+      "epoch": 0.6296408089345005,
+      "grad_norm": 0.0976100042462349,
+      "learning_rate": 0.00016293067662930677,
+      "loss": 0.6683,
+      "step": 1043
+    },
+    {
+      "epoch": 0.6302444913975249,
+      "grad_norm": 0.10183601081371307,
+      "learning_rate": 0.00016288916562889167,
+      "loss": 0.7513,
+      "step": 1044
+    },
+    {
+      "epoch": 0.6308481738605494,
+      "grad_norm": 0.10491038113832474,
+      "learning_rate": 0.00016284765462847654,
+      "loss": 0.7886,
+      "step": 1045
+    },
+    {
+      "epoch": 0.6314518563235738,
+      "grad_norm": 0.1056092381477356,
+      "learning_rate": 0.00016280614362806144,
+      "loss": 0.6803,
+      "step": 1046
+    },
+    {
+      "epoch": 0.6320555387865983,
+      "grad_norm": 0.1121145635843277,
+      "learning_rate": 0.00016276463262764634,
+      "loss": 0.6256,
+      "step": 1047
+    },
+    {
+      "epoch": 0.6326592212496227,
+      "grad_norm": 0.11745337396860123,
+      "learning_rate": 0.00016272312162723122,
+      "loss": 0.6546,
+      "step": 1048
+    },
+    {
+      "epoch": 0.6332629037126472,
+      "grad_norm": 0.12515030801296234,
+      "learning_rate": 0.00016268161062681612,
+      "loss": 0.5382,
+      "step": 1049
+    },
+    {
+      "epoch": 0.6338665861756716,
+      "grad_norm": 0.13032299280166626,
+      "learning_rate": 0.00016264009962640102,
+      "loss": 0.4321,
+      "step": 1050
+    },
+    {
+      "epoch": 0.634470268638696,
+      "grad_norm": 0.08480330556631088,
+      "learning_rate": 0.0001625985886259859,
+      "loss": 0.8055,
+      "step": 1051
+    },
+    {
+      "epoch": 0.6350739511017205,
+      "grad_norm": 0.08330792188644409,
+      "learning_rate": 0.00016255707762557076,
+      "loss": 1.1714,
+      "step": 1052
+    },
+    {
+      "epoch": 0.6356776335647449,
+      "grad_norm": 0.09432794153690338,
+      "learning_rate": 0.0001625155666251557,
+      "loss": 1.0071,
+      "step": 1053
+    },
+    {
+      "epoch": 0.6362813160277694,
+      "grad_norm": 0.084816575050354,
+      "learning_rate": 0.00016247405562474056,
+      "loss": 0.7843,
+      "step": 1054
+    },
+    {
+      "epoch": 0.6368849984907938,
+      "grad_norm": 0.09600014239549637,
+      "learning_rate": 0.00016243254462432543,
+      "loss": 0.8295,
+      "step": 1055
+    },
+    {
+      "epoch": 0.6374886809538183,
+      "grad_norm": 0.10093321651220322,
+      "learning_rate": 0.00016239103362391036,
+      "loss": 0.8685,
+      "step": 1056
+    },
+    {
+      "epoch": 0.6380923634168427,
+      "grad_norm": 0.07668054848909378,
+      "learning_rate": 0.00016234952262349523,
+      "loss": 1.042,
+      "step": 1057
+    },
+    {
+      "epoch": 0.6386960458798672,
+      "grad_norm": 0.12126658856868744,
+      "learning_rate": 0.0001623080116230801,
+      "loss": 0.8384,
+      "step": 1058
+    },
+    {
+      "epoch": 0.6392997283428916,
+      "grad_norm": 0.07691983133554459,
+      "learning_rate": 0.000162266500622665,
+      "loss": 0.9138,
+      "step": 1059
+    },
+    {
+      "epoch": 0.6399034108059161,
+      "grad_norm": 0.07935553044080734,
+      "learning_rate": 0.0001622249896222499,
+      "loss": 0.7028,
+      "step": 1060
+    },
+    {
+      "epoch": 0.6405070932689405,
+      "grad_norm": 0.0814066082239151,
+      "learning_rate": 0.00016218347862183478,
+      "loss": 0.8901,
+      "step": 1061
+    },
+    {
+      "epoch": 0.641110775731965,
+      "grad_norm": 0.09054214507341385,
+      "learning_rate": 0.00016214196762141968,
+      "loss": 0.9142,
+      "step": 1062
+    },
+    {
+      "epoch": 0.6417144581949894,
+      "grad_norm": 0.0810336098074913,
+      "learning_rate": 0.00016210045662100458,
+      "loss": 0.7975,
+      "step": 1063
+    },
+    {
+      "epoch": 0.6423181406580138,
+      "grad_norm": 0.07674489170312881,
+      "learning_rate": 0.00016205894562058945,
+      "loss": 0.9827,
+      "step": 1064
+    },
+    {
+      "epoch": 0.6429218231210383,
+      "grad_norm": 0.07564074546098709,
+      "learning_rate": 0.00016201743462017435,
+      "loss": 0.9164,
+      "step": 1065
+    },
+    {
+      "epoch": 0.6435255055840627,
+      "grad_norm": 0.082376629114151,
+      "learning_rate": 0.00016197592361975923,
+      "loss": 1.0141,
+      "step": 1066
+    },
+    {
+      "epoch": 0.6441291880470872,
+      "grad_norm": 0.09569506347179413,
+      "learning_rate": 0.00016193441261934415,
+      "loss": 1.031,
+      "step": 1067
+    },
+    {
+      "epoch": 0.6447328705101116,
+      "grad_norm": 0.07831252366304398,
+      "learning_rate": 0.00016189290161892903,
+      "loss": 0.7855,
+      "step": 1068
+    },
+    {
+      "epoch": 0.6453365529731361,
+      "grad_norm": 0.12247852981090546,
+      "learning_rate": 0.0001618513906185139,
+      "loss": 0.9584,
+      "step": 1069
+    },
+    {
+      "epoch": 0.6459402354361605,
+      "grad_norm": 0.07706090807914734,
+      "learning_rate": 0.00016180987961809883,
+      "loss": 0.9001,
+      "step": 1070
+    },
+    {
+      "epoch": 0.6465439178991851,
+      "grad_norm": 0.07961270213127136,
+      "learning_rate": 0.0001617683686176837,
+      "loss": 0.725,
+      "step": 1071
+    },
+    {
+      "epoch": 0.6471476003622095,
+      "grad_norm": 0.08146402984857559,
+      "learning_rate": 0.00016172685761726857,
+      "loss": 0.8213,
+      "step": 1072
+    },
+    {
+      "epoch": 0.647751282825234,
+      "grad_norm": 0.1310458779335022,
+      "learning_rate": 0.00016168534661685347,
+      "loss": 0.8039,
+      "step": 1073
+    },
+    {
+      "epoch": 0.6483549652882584,
+      "grad_norm": 0.09053052216768265,
+      "learning_rate": 0.00016164383561643837,
+      "loss": 0.8154,
+      "step": 1074
+    },
+    {
+      "epoch": 0.6489586477512829,
+      "grad_norm": 0.07922950387001038,
+      "learning_rate": 0.00016160232461602325,
+      "loss": 0.8074,
+      "step": 1075
+    },
+    {
+      "epoch": 0.6495623302143073,
+      "grad_norm": 0.07535174489021301,
+      "learning_rate": 0.00016156081361560815,
+      "loss": 0.7056,
+      "step": 1076
+    },
+    {
+      "epoch": 0.6501660126773318,
+      "grad_norm": 0.08822023123502731,
+      "learning_rate": 0.00016151930261519305,
+      "loss": 0.8269,
+      "step": 1077
+    },
+    {
+      "epoch": 0.6507696951403562,
+      "grad_norm": 0.12088391184806824,
+      "learning_rate": 0.00016147779161477792,
+      "loss": 1.0016,
+      "step": 1078
+    },
+    {
+      "epoch": 0.6513733776033807,
+      "grad_norm": 0.08213537931442261,
+      "learning_rate": 0.00016143628061436282,
+      "loss": 0.7239,
+      "step": 1079
+    },
+    {
+      "epoch": 0.6519770600664051,
+      "grad_norm": 0.08252882957458496,
+      "learning_rate": 0.0001613947696139477,
+      "loss": 0.8806,
+      "step": 1080
+    },
+    {
+      "epoch": 0.6525807425294295,
+      "grad_norm": 0.06710948050022125,
+      "learning_rate": 0.0001613532586135326,
+      "loss": 0.9193,
+      "step": 1081
+    },
+    {
+      "epoch": 0.653184424992454,
+      "grad_norm": 0.14059540629386902,
+      "learning_rate": 0.0001613117476131175,
+      "loss": 0.8201,
+      "step": 1082
+    },
+    {
+      "epoch": 0.6537881074554784,
+      "grad_norm": 0.08148171007633209,
+      "learning_rate": 0.00016127023661270236,
+      "loss": 0.7983,
+      "step": 1083
+    },
+    {
+      "epoch": 0.6543917899185029,
+      "grad_norm": 0.07753555476665497,
+      "learning_rate": 0.00016122872561228726,
+      "loss": 0.6901,
+      "step": 1084
+    },
+    {
+      "epoch": 0.6549954723815273,
+      "grad_norm": 0.08410129696130753,
+      "learning_rate": 0.00016118721461187216,
+      "loss": 0.7755,
+      "step": 1085
+    },
+    {
+      "epoch": 0.6555991548445518,
+      "grad_norm": 0.08018611371517181,
+      "learning_rate": 0.00016114570361145704,
+      "loss": 0.7569,
+      "step": 1086
+    },
+    {
+      "epoch": 0.6562028373075762,
+      "grad_norm": 0.08290990442037582,
+      "learning_rate": 0.00016110419261104194,
+      "loss": 0.7694,
+      "step": 1087
+    },
+    {
+      "epoch": 0.6568065197706007,
+      "grad_norm": 0.08168758451938629,
+      "learning_rate": 0.00016106268161062684,
+      "loss": 0.7458,
+      "step": 1088
+    },
+    {
+      "epoch": 0.6574102022336251,
+      "grad_norm": 0.09103231132030487,
+      "learning_rate": 0.0001610211706102117,
+      "loss": 0.7203,
+      "step": 1089
+    },
+    {
+      "epoch": 0.6580138846966496,
+      "grad_norm": 0.0930967926979065,
+      "learning_rate": 0.00016097965960979658,
+      "loss": 0.764,
+      "step": 1090
+    },
+    {
+      "epoch": 0.658617567159674,
+      "grad_norm": 0.0999932810664177,
+      "learning_rate": 0.0001609381486093815,
+      "loss": 0.7885,
+      "step": 1091
+    },
+    {
+      "epoch": 0.6592212496226985,
+      "grad_norm": 0.09345666319131851,
+      "learning_rate": 0.00016089663760896638,
+      "loss": 0.7567,
+      "step": 1092
+    },
+    {
+      "epoch": 0.6598249320857229,
+      "grad_norm": 0.09461873769760132,
+      "learning_rate": 0.00016085512660855126,
+      "loss": 0.6643,
+      "step": 1093
+    },
+    {
+      "epoch": 0.6604286145487474,
+      "grad_norm": 0.10697057098150253,
+      "learning_rate": 0.00016081361560813616,
+      "loss": 0.7219,
+      "step": 1094
+    },
+    {
+      "epoch": 0.6610322970117718,
+      "grad_norm": 0.11102855950593948,
+      "learning_rate": 0.00016077210460772106,
+      "loss": 0.6777,
+      "step": 1095
+    },
+    {
+      "epoch": 0.6616359794747962,
+      "grad_norm": 0.11301318556070328,
+      "learning_rate": 0.00016073059360730593,
+      "loss": 0.7065,
+      "step": 1096
+    },
+    {
+      "epoch": 0.6622396619378207,
+      "grad_norm": 0.11578387022018433,
+      "learning_rate": 0.00016068908260689083,
+      "loss": 0.731,
+      "step": 1097
+    },
+    {
+      "epoch": 0.6628433444008451,
+      "grad_norm": 0.11346578598022461,
+      "learning_rate": 0.00016064757160647573,
+      "loss": 0.579,
+      "step": 1098
+    },
+    {
+      "epoch": 0.6634470268638696,
+      "grad_norm": 0.1209607645869255,
+      "learning_rate": 0.0001606060606060606,
+      "loss": 0.5572,
+      "step": 1099
+    },
+    {
+      "epoch": 0.664050709326894,
+      "grad_norm": 0.11950193345546722,
+      "learning_rate": 0.0001605645496056455,
+      "loss": 0.3567,
+      "step": 1100
+    },
+    {
+      "epoch": 0.6646543917899185,
+      "grad_norm": 0.07838748395442963,
+      "learning_rate": 0.0001605230386052304,
+      "loss": 1.0775,
+      "step": 1101
+    },
+    {
+      "epoch": 0.6652580742529429,
+      "grad_norm": 0.1073046401143074,
+      "learning_rate": 0.00016048152760481528,
+      "loss": 0.693,
+      "step": 1102
+    },
+    {
+      "epoch": 0.6658617567159674,
+      "grad_norm": 0.0886329859495163,
+      "learning_rate": 0.00016044001660440018,
+      "loss": 0.7818,
+      "step": 1103
+    },
+    {
+      "epoch": 0.6664654391789918,
+      "grad_norm": 0.07882886379957199,
+      "learning_rate": 0.00016039850560398505,
+      "loss": 0.7407,
+      "step": 1104
+    },
+    {
+      "epoch": 0.6670691216420163,
+      "grad_norm": 0.09764014929533005,
+      "learning_rate": 0.00016035699460356995,
+      "loss": 0.8944,
+      "step": 1105
+    },
+    {
+      "epoch": 0.6676728041050407,
+      "grad_norm": 0.08643536269664764,
+      "learning_rate": 0.00016031548360315485,
+      "loss": 0.8511,
+      "step": 1106
+    },
+    {
+      "epoch": 0.6682764865680652,
+      "grad_norm": 0.08230648189783096,
+      "learning_rate": 0.00016027397260273972,
+      "loss": 0.9837,
+      "step": 1107
+    },
+    {
+      "epoch": 0.6688801690310896,
+      "grad_norm": 0.08523669838905334,
+      "learning_rate": 0.00016023246160232462,
+      "loss": 0.7902,
+      "step": 1108
+    },
+    {
+      "epoch": 0.669483851494114,
+      "grad_norm": 0.08347252011299133,
+      "learning_rate": 0.00016019095060190952,
+      "loss": 0.7791,
+      "step": 1109
+    },
+    {
+      "epoch": 0.6700875339571385,
+      "grad_norm": 0.08182717114686966,
+      "learning_rate": 0.0001601494396014944,
+      "loss": 0.8484,
+      "step": 1110
+    },
+    {
+      "epoch": 0.6706912164201629,
+      "grad_norm": 0.0879136249423027,
+      "learning_rate": 0.0001601079286010793,
+      "loss": 0.8469,
+      "step": 1111
+    },
+    {
+      "epoch": 0.6712948988831875,
+      "grad_norm": 0.08389199525117874,
+      "learning_rate": 0.0001600664176006642,
+      "loss": 0.9403,
+      "step": 1112
+    },
+    {
+      "epoch": 0.6718985813462119,
+      "grad_norm": 0.07789173722267151,
+      "learning_rate": 0.00016002490660024907,
+      "loss": 0.8443,
+      "step": 1113
+    },
+    {
+      "epoch": 0.6725022638092364,
+      "grad_norm": 0.07873908430337906,
+      "learning_rate": 0.00015998339559983397,
+      "loss": 0.8464,
+      "step": 1114
+    },
+    {
+      "epoch": 0.6731059462722608,
+      "grad_norm": 0.10451683402061462,
+      "learning_rate": 0.00015994188459941887,
+      "loss": 0.8507,
+      "step": 1115
+    },
+    {
+      "epoch": 0.6737096287352853,
+      "grad_norm": 0.10225295275449753,
+      "learning_rate": 0.00015990037359900374,
+      "loss": 1.1292,
+      "step": 1116
+    },
+    {
+      "epoch": 0.6743133111983097,
+      "grad_norm": 0.08053860068321228,
+      "learning_rate": 0.00015985886259858864,
+      "loss": 0.8511,
+      "step": 1117
+    },
+    {
+      "epoch": 0.6749169936613342,
+      "grad_norm": 0.15555590391159058,
+      "learning_rate": 0.00015981735159817351,
+      "loss": 0.6958,
+      "step": 1118
+    },
+    {
+      "epoch": 0.6755206761243586,
+      "grad_norm": 0.09447921067476273,
+      "learning_rate": 0.00015977584059775841,
+      "loss": 0.8228,
+      "step": 1119
+    },
+    {
+      "epoch": 0.6761243585873831,
+      "grad_norm": 0.11645516008138657,
+      "learning_rate": 0.00015973432959734331,
+      "loss": 0.8909,
+      "step": 1120
+    },
+    {
+      "epoch": 0.6767280410504075,
+      "grad_norm": 0.09540687501430511,
+      "learning_rate": 0.0001596928185969282,
+      "loss": 0.7483,
+      "step": 1121
+    },
+    {
+      "epoch": 0.677331723513432,
+      "grad_norm": 0.08288270980119705,
+      "learning_rate": 0.0001596513075965131,
+      "loss": 0.837,
+      "step": 1122
+    },
+    {
+      "epoch": 0.6779354059764564,
+      "grad_norm": 0.0898602157831192,
+      "learning_rate": 0.000159609796596098,
+      "loss": 0.7893,
+      "step": 1123
+    },
+    {
+      "epoch": 0.6785390884394809,
+      "grad_norm": 0.08173252642154694,
+      "learning_rate": 0.00015956828559568286,
+      "loss": 0.7965,
+      "step": 1124
+    },
+    {
+      "epoch": 0.6791427709025053,
+      "grad_norm": 0.08245470374822617,
+      "learning_rate": 0.00015952677459526776,
+      "loss": 0.9573,
+      "step": 1125
+    },
+    {
+      "epoch": 0.6797464533655297,
+      "grad_norm": 0.07909370958805084,
+      "learning_rate": 0.00015948526359485266,
+      "loss": 0.7852,
+      "step": 1126
+    },
+    {
+      "epoch": 0.6803501358285542,
+      "grad_norm": 0.07963291555643082,
+      "learning_rate": 0.00015944375259443753,
+      "loss": 0.8806,
+      "step": 1127
+    },
+    {
+      "epoch": 0.6809538182915786,
+      "grad_norm": 0.07449831813573837,
+      "learning_rate": 0.0001594022415940224,
+      "loss": 0.8094,
+      "step": 1128
+    },
+    {
+      "epoch": 0.6815575007546031,
+      "grad_norm": 0.07845521718263626,
+      "learning_rate": 0.00015936073059360733,
+      "loss": 0.8813,
+      "step": 1129
+    },
+    {
+      "epoch": 0.6821611832176275,
+      "grad_norm": 0.0760309100151062,
+      "learning_rate": 0.0001593192195931922,
+      "loss": 0.9867,
+      "step": 1130
+    },
+    {
+      "epoch": 0.682764865680652,
+      "grad_norm": 0.07355882227420807,
+      "learning_rate": 0.00015927770859277708,
+      "loss": 0.8448,
+      "step": 1131
+    },
+    {
+      "epoch": 0.6833685481436764,
+      "grad_norm": 0.07743940502405167,
+      "learning_rate": 0.00015923619759236198,
+      "loss": 1.0274,
+      "step": 1132
+    },
+    {
+      "epoch": 0.6839722306067009,
+      "grad_norm": 0.08236207067966461,
+      "learning_rate": 0.00015919468659194688,
+      "loss": 0.7691,
+      "step": 1133
+    },
+    {
+      "epoch": 0.6845759130697253,
+      "grad_norm": 0.07984744012355804,
+      "learning_rate": 0.00015915317559153175,
+      "loss": 0.7605,
+      "step": 1134
+    },
+    {
+      "epoch": 0.6851795955327498,
+      "grad_norm": 0.08415329456329346,
+      "learning_rate": 0.00015911166459111665,
+      "loss": 0.7576,
+      "step": 1135
+    },
+    {
+      "epoch": 0.6857832779957742,
+      "grad_norm": 0.08308225870132446,
+      "learning_rate": 0.00015907015359070155,
+      "loss": 0.6959,
+      "step": 1136
+    },
+    {
+      "epoch": 0.6863869604587987,
+      "grad_norm": 0.08338841050863266,
+      "learning_rate": 0.00015902864259028642,
+      "loss": 0.6998,
+      "step": 1137
+    },
+    {
+      "epoch": 0.6869906429218231,
+      "grad_norm": 0.08652734011411667,
+      "learning_rate": 0.00015898713158987132,
+      "loss": 0.7469,
+      "step": 1138
+    },
+    {
+      "epoch": 0.6875943253848475,
+      "grad_norm": 0.11346086114645004,
+      "learning_rate": 0.00015894562058945622,
+      "loss": 0.6995,
+      "step": 1139
+    },
+    {
+      "epoch": 0.688198007847872,
+      "grad_norm": 0.08581293374300003,
+      "learning_rate": 0.0001589041095890411,
+      "loss": 0.7138,
+      "step": 1140
+    },
+    {
+      "epoch": 0.6888016903108964,
+      "grad_norm": 0.0915643498301506,
+      "learning_rate": 0.000158862598588626,
+      "loss": 0.7836,
+      "step": 1141
+    },
+    {
+      "epoch": 0.6894053727739209,
+      "grad_norm": 0.09856559336185455,
+      "learning_rate": 0.00015882108758821087,
+      "loss": 0.7151,
+      "step": 1142
+    },
+    {
+      "epoch": 0.6900090552369453,
+      "grad_norm": 0.09949114173650742,
+      "learning_rate": 0.00015877957658779577,
+      "loss": 0.7292,
+      "step": 1143
+    },
+    {
+      "epoch": 0.6906127376999698,
+      "grad_norm": 0.10931304842233658,
+      "learning_rate": 0.00015873806558738067,
+      "loss": 0.777,
+      "step": 1144
+    },
+    {
+      "epoch": 0.6912164201629942,
+      "grad_norm": 0.10665369033813477,
+      "learning_rate": 0.00015869655458696554,
+      "loss": 0.6729,
+      "step": 1145
+    },
+    {
+      "epoch": 0.6918201026260187,
+      "grad_norm": 0.1097274124622345,
+      "learning_rate": 0.00015865504358655044,
+      "loss": 0.6532,
+      "step": 1146
+    },
+    {
+      "epoch": 0.6924237850890431,
+      "grad_norm": 0.12797416746616364,
+      "learning_rate": 0.00015861353258613534,
+      "loss": 0.6375,
+      "step": 1147
+    },
+    {
+      "epoch": 0.6930274675520676,
+      "grad_norm": 0.11834930628538132,
+      "learning_rate": 0.00015857202158572022,
+      "loss": 0.6374,
+      "step": 1148
+    },
+    {
+      "epoch": 0.693631150015092,
+      "grad_norm": 0.11373134702444077,
+      "learning_rate": 0.00015853051058530512,
+      "loss": 0.5455,
+      "step": 1149
+    },
+    {
+      "epoch": 0.6942348324781165,
+      "grad_norm": 0.13020096719264984,
+      "learning_rate": 0.00015848899958489002,
+      "loss": 0.4624,
+      "step": 1150
+    },
+    {
+      "epoch": 0.6948385149411409,
+      "grad_norm": 0.09528189897537231,
+      "learning_rate": 0.0001584474885844749,
+      "loss": 0.8045,
+      "step": 1151
+    },
+    {
+      "epoch": 0.6954421974041655,
+      "grad_norm": 0.09046769887208939,
+      "learning_rate": 0.00015840597758405976,
+      "loss": 0.7992,
+      "step": 1152
+    },
+    {
+      "epoch": 0.6960458798671899,
+      "grad_norm": 0.10954567044973373,
+      "learning_rate": 0.0001583644665836447,
+      "loss": 0.7242,
+      "step": 1153
+    },
+    {
+      "epoch": 0.6966495623302144,
+      "grad_norm": 0.08049467206001282,
+      "learning_rate": 0.00015832295558322956,
+      "loss": 0.9732,
+      "step": 1154
+    },
+    {
+      "epoch": 0.6972532447932388,
+      "grad_norm": 0.08102589845657349,
+      "learning_rate": 0.00015828144458281446,
+      "loss": 1.0654,
+      "step": 1155
+    },
+    {
+      "epoch": 0.6978569272562632,
+      "grad_norm": 0.07799682021141052,
+      "learning_rate": 0.00015823993358239934,
+      "loss": 0.879,
+      "step": 1156
+    },
+    {
+      "epoch": 0.6984606097192877,
+      "grad_norm": 0.08080107718706131,
+      "learning_rate": 0.00015819842258198424,
+      "loss": 0.8515,
+      "step": 1157
+    },
+    {
+      "epoch": 0.6990642921823121,
+      "grad_norm": 0.07894068211317062,
+      "learning_rate": 0.00015815691158156914,
+      "loss": 0.8024,
+      "step": 1158
+    },
+    {
+      "epoch": 0.6996679746453366,
+      "grad_norm": 0.0848923772573471,
+      "learning_rate": 0.000158115400581154,
+      "loss": 0.7698,
+      "step": 1159
+    },
+    {
+      "epoch": 0.700271657108361,
+      "grad_norm": 0.08654095232486725,
+      "learning_rate": 0.0001580738895807389,
+      "loss": 0.8809,
+      "step": 1160
+    },
+    {
+      "epoch": 0.7008753395713855,
+      "grad_norm": 0.07776283472776413,
+      "learning_rate": 0.0001580323785803238,
+      "loss": 0.7659,
+      "step": 1161
+    },
+    {
+      "epoch": 0.7014790220344099,
+      "grad_norm": 0.09321299195289612,
+      "learning_rate": 0.00015799086757990868,
+      "loss": 0.9147,
+      "step": 1162
+    },
+    {
+      "epoch": 0.7020827044974344,
+      "grad_norm": 0.10445485264062881,
+      "learning_rate": 0.00015794935657949358,
+      "loss": 0.7182,
+      "step": 1163
+    },
+    {
+      "epoch": 0.7026863869604588,
+      "grad_norm": 0.08144387602806091,
+      "learning_rate": 0.00015790784557907848,
+      "loss": 0.9075,
+      "step": 1164
+    },
+    {
+      "epoch": 0.7032900694234833,
+      "grad_norm": 0.08363913744688034,
+      "learning_rate": 0.00015786633457866335,
+      "loss": 0.7935,
+      "step": 1165
+    },
+    {
+      "epoch": 0.7038937518865077,
+      "grad_norm": 0.07295918464660645,
+      "learning_rate": 0.00015782482357824823,
+      "loss": 0.99,
+      "step": 1166
+    },
+    {
+      "epoch": 0.7044974343495322,
+      "grad_norm": 0.0843958631157875,
+      "learning_rate": 0.00015778331257783315,
+      "loss": 0.8588,
+      "step": 1167
+    },
+    {
+      "epoch": 0.7051011168125566,
+      "grad_norm": 0.08498518168926239,
+      "learning_rate": 0.00015774180157741803,
+      "loss": 0.7837,
+      "step": 1168
+    },
+    {
+      "epoch": 0.705704799275581,
+      "grad_norm": 0.08214520663022995,
+      "learning_rate": 0.0001577002905770029,
+      "loss": 0.8411,
+      "step": 1169
+    },
+    {
+      "epoch": 0.7063084817386055,
+      "grad_norm": 0.08001639693975449,
+      "learning_rate": 0.0001576587795765878,
+      "loss": 0.7665,
+      "step": 1170
+    },
+    {
+      "epoch": 0.70691216420163,
+      "grad_norm": 0.08271230757236481,
+      "learning_rate": 0.0001576172685761727,
+      "loss": 0.8341,
+      "step": 1171
+    },
+    {
+      "epoch": 0.7075158466646544,
+      "grad_norm": 0.08699033409357071,
+      "learning_rate": 0.00015757575757575757,
+      "loss": 0.7784,
+      "step": 1172
+    },
+    {
+      "epoch": 0.7081195291276788,
+      "grad_norm": 0.09423791617155075,
+      "learning_rate": 0.00015753424657534247,
+      "loss": 0.9262,
+      "step": 1173
+    },
+    {
+      "epoch": 0.7087232115907033,
+      "grad_norm": 0.07842440158128738,
+      "learning_rate": 0.00015749273557492737,
+      "loss": 0.9249,
+      "step": 1174
+    },
+    {
+      "epoch": 0.7093268940537277,
+      "grad_norm": 0.08183681219816208,
+      "learning_rate": 0.00015745122457451225,
+      "loss": 0.7941,
+      "step": 1175
+    },
+    {
+      "epoch": 0.7099305765167522,
+      "grad_norm": 0.08315775543451309,
+      "learning_rate": 0.00015740971357409715,
+      "loss": 0.8225,
+      "step": 1176
+    },
+    {
+      "epoch": 0.7105342589797766,
+      "grad_norm": 0.08027949929237366,
+      "learning_rate": 0.00015736820257368205,
+      "loss": 0.8441,
+      "step": 1177
+    },
+    {
+      "epoch": 0.7111379414428011,
+      "grad_norm": 0.08332298696041107,
+      "learning_rate": 0.00015732669157326692,
+      "loss": 0.9679,
+      "step": 1178
+    },
+    {
+      "epoch": 0.7117416239058255,
+      "grad_norm": 0.09133647382259369,
+      "learning_rate": 0.00015728518057285182,
+      "loss": 0.9468,
+      "step": 1179
+    },
+    {
+      "epoch": 0.71234530636885,
+      "grad_norm": 0.08058296889066696,
+      "learning_rate": 0.0001572436695724367,
+      "loss": 0.8374,
+      "step": 1180
+    },
+    {
+      "epoch": 0.7129489888318744,
+      "grad_norm": 0.08813148736953735,
+      "learning_rate": 0.0001572021585720216,
+      "loss": 0.9568,
+      "step": 1181
+    },
+    {
+      "epoch": 0.7135526712948989,
+      "grad_norm": 0.07533196359872818,
+      "learning_rate": 0.0001571606475716065,
+      "loss": 0.8627,
+      "step": 1182
+    },
+    {
+      "epoch": 0.7141563537579233,
+      "grad_norm": 0.08658763766288757,
+      "learning_rate": 0.00015711913657119137,
+      "loss": 0.7429,
+      "step": 1183
+    },
+    {
+      "epoch": 0.7147600362209477,
+      "grad_norm": 0.0844459980726242,
+      "learning_rate": 0.00015707762557077627,
+      "loss": 0.7966,
+      "step": 1184
+    },
+    {
+      "epoch": 0.7153637186839722,
+      "grad_norm": 0.08175187557935715,
+      "learning_rate": 0.00015703611457036117,
+      "loss": 0.7226,
+      "step": 1185
+    },
+    {
+      "epoch": 0.7159674011469966,
+      "grad_norm": 0.08582815527915955,
+      "learning_rate": 0.00015699460356994604,
+      "loss": 0.717,
+      "step": 1186
+    },
+    {
+      "epoch": 0.7165710836100211,
+      "grad_norm": 0.08503274619579315,
+      "learning_rate": 0.0001569530925695309,
+      "loss": 0.7526,
+      "step": 1187
+    },
+    {
+      "epoch": 0.7171747660730455,
+      "grad_norm": 0.08774889260530472,
+      "learning_rate": 0.00015691158156911584,
+      "loss": 0.7016,
+      "step": 1188
+    },
+    {
+      "epoch": 0.71777844853607,
+      "grad_norm": 0.10124623775482178,
+      "learning_rate": 0.0001568700705687007,
+      "loss": 0.6455,
+      "step": 1189
+    },
+    {
+      "epoch": 0.7183821309990944,
+      "grad_norm": 0.10203047096729279,
+      "learning_rate": 0.00015682855956828558,
+      "loss": 0.6779,
+      "step": 1190
+    },
+    {
+      "epoch": 0.7189858134621189,
+      "grad_norm": 0.09329473972320557,
+      "learning_rate": 0.0001567870485678705,
+      "loss": 0.7313,
+      "step": 1191
+    },
+    {
+      "epoch": 0.7195894959251434,
+      "grad_norm": 0.09363296627998352,
+      "learning_rate": 0.00015674553756745538,
+      "loss": 0.6968,
+      "step": 1192
+    },
+    {
+      "epoch": 0.7201931783881679,
+      "grad_norm": 0.10082225501537323,
+      "learning_rate": 0.00015670402656704026,
+      "loss": 0.737,
+      "step": 1193
+    },
+    {
+      "epoch": 0.7207968608511923,
+      "grad_norm": 0.10495683550834656,
+      "learning_rate": 0.00015666251556662516,
+      "loss": 0.7186,
+      "step": 1194
+    },
+    {
+      "epoch": 0.7214005433142168,
+      "grad_norm": 0.10491074621677399,
+      "learning_rate": 0.00015662100456621006,
+      "loss": 0.6949,
+      "step": 1195
+    },
+    {
+      "epoch": 0.7220042257772412,
+      "grad_norm": 0.12182420492172241,
+      "learning_rate": 0.00015657949356579493,
+      "loss": 0.667,
+      "step": 1196
+    },
+    {
+      "epoch": 0.7226079082402657,
+      "grad_norm": 0.12086670100688934,
+      "learning_rate": 0.00015653798256537983,
+      "loss": 0.5885,
+      "step": 1197
+    },
+    {
+      "epoch": 0.7232115907032901,
+      "grad_norm": 0.11433517187833786,
+      "learning_rate": 0.00015649647156496473,
+      "loss": 0.6147,
+      "step": 1198
+    },
+    {
+      "epoch": 0.7238152731663146,
+      "grad_norm": 0.12221046537160873,
+      "learning_rate": 0.00015645496056454963,
+      "loss": 0.5129,
+      "step": 1199
+    },
+    {
+      "epoch": 0.724418955629339,
+      "grad_norm": 0.12147875875234604,
+      "learning_rate": 0.0001564134495641345,
+      "loss": 0.4205,
+      "step": 1200
+    },
+    {
+      "epoch": 0.7250226380923634,
+      "grad_norm": 0.08808968216180801,
+      "learning_rate": 0.00015637193856371938,
+      "loss": 0.8195,
+      "step": 1201
+    },
+    {
+      "epoch": 0.7256263205553879,
+      "grad_norm": 0.08320089429616928,
+      "learning_rate": 0.0001563304275633043,
+      "loss": 0.8505,
+      "step": 1202
+    },
+    {
+      "epoch": 0.7262300030184123,
+      "grad_norm": 0.0852508470416069,
+      "learning_rate": 0.00015628891656288918,
+      "loss": 1.0964,
+      "step": 1203
+    },
+    {
+      "epoch": 0.7268336854814368,
+      "grad_norm": 0.09474988281726837,
+      "learning_rate": 0.00015624740556247405,
+      "loss": 0.8924,
+      "step": 1204
+    },
+    {
+      "epoch": 0.7274373679444612,
+      "grad_norm": 0.092880479991436,
+      "learning_rate": 0.00015620589456205898,
+      "loss": 0.8223,
+      "step": 1205
+    },
+    {
+      "epoch": 0.7280410504074857,
+      "grad_norm": 0.10932306200265884,
+      "learning_rate": 0.00015616438356164385,
+      "loss": 0.8786,
+      "step": 1206
+    },
+    {
+      "epoch": 0.7286447328705101,
+      "grad_norm": 0.0782761201262474,
+      "learning_rate": 0.00015612287256122872,
+      "loss": 0.7801,
+      "step": 1207
+    },
+    {
+      "epoch": 0.7292484153335346,
+      "grad_norm": 0.09566731005907059,
+      "learning_rate": 0.00015608136156081362,
+      "loss": 0.9474,
+      "step": 1208
+    },
+    {
+      "epoch": 0.729852097796559,
+      "grad_norm": 0.08776892721652985,
+      "learning_rate": 0.00015603985056039852,
+      "loss": 0.7888,
+      "step": 1209
+    },
+    {
+      "epoch": 0.7304557802595835,
+      "grad_norm": 0.07871197909116745,
+      "learning_rate": 0.0001559983395599834,
+      "loss": 0.7624,
+      "step": 1210
+    },
+    {
+      "epoch": 0.7310594627226079,
+      "grad_norm": 0.19957715272903442,
+      "learning_rate": 0.0001559568285595683,
+      "loss": 0.8708,
+      "step": 1211
+    },
+    {
+      "epoch": 0.7316631451856324,
+      "grad_norm": 0.0865522176027298,
+      "learning_rate": 0.0001559153175591532,
+      "loss": 0.9895,
+      "step": 1212
+    },
+    {
+      "epoch": 0.7322668276486568,
+      "grad_norm": 0.08585958927869797,
+      "learning_rate": 0.00015587380655873807,
+      "loss": 0.8575,
+      "step": 1213
+    },
+    {
+      "epoch": 0.7328705101116813,
+      "grad_norm": 0.07868848741054535,
+      "learning_rate": 0.00015583229555832297,
+      "loss": 0.768,
+      "step": 1214
+    },
+    {
+      "epoch": 0.7334741925747057,
+      "grad_norm": 0.08140038698911667,
+      "learning_rate": 0.00015579078455790784,
+      "loss": 0.7544,
+      "step": 1215
+    },
+    {
+      "epoch": 0.7340778750377301,
+      "grad_norm": 0.08099745959043503,
+      "learning_rate": 0.00015574927355749274,
+      "loss": 0.7234,
+      "step": 1216
+    },
+    {
+      "epoch": 0.7346815575007546,
+      "grad_norm": 0.08431069552898407,
+      "learning_rate": 0.00015570776255707764,
+      "loss": 0.8079,
+      "step": 1217
+    },
+    {
+      "epoch": 0.735285239963779,
+      "grad_norm": 0.08134233951568604,
+      "learning_rate": 0.00015566625155666251,
+      "loss": 1.0639,
+      "step": 1218
+    },
+    {
+      "epoch": 0.7358889224268035,
+      "grad_norm": 0.08087371289730072,
+      "learning_rate": 0.00015562474055624741,
+      "loss": 0.7534,
+      "step": 1219
+    },
+    {
+      "epoch": 0.7364926048898279,
+      "grad_norm": 0.08062789589166641,
+      "learning_rate": 0.00015558322955583231,
+      "loss": 0.8352,
+      "step": 1220
+    },
+    {
+      "epoch": 0.7370962873528524,
+      "grad_norm": 0.08649458736181259,
+      "learning_rate": 0.0001555417185554172,
+      "loss": 0.7991,
+      "step": 1221
+    },
+    {
+      "epoch": 0.7376999698158768,
+      "grad_norm": 0.0848165899515152,
+      "learning_rate": 0.0001555002075550021,
+      "loss": 0.8912,
+      "step": 1222
+    },
+    {
+      "epoch": 0.7383036522789013,
+      "grad_norm": 0.07689525932073593,
+      "learning_rate": 0.000155458696554587,
+      "loss": 0.8181,
+      "step": 1223
+    },
+    {
+      "epoch": 0.7389073347419257,
+      "grad_norm": 0.07876717299222946,
+      "learning_rate": 0.00015541718555417186,
+      "loss": 0.7244,
+      "step": 1224
+    },
+    {
+      "epoch": 0.7395110172049502,
+      "grad_norm": 0.07784243673086166,
+      "learning_rate": 0.00015537567455375673,
+      "loss": 1.539,
+      "step": 1225
+    },
+    {
+      "epoch": 0.7401146996679746,
+      "grad_norm": 0.08357193320989609,
+      "learning_rate": 0.00015533416355334166,
+      "loss": 0.8915,
+      "step": 1226
+    },
+    {
+      "epoch": 0.740718382130999,
+      "grad_norm": 0.07806842029094696,
+      "learning_rate": 0.00015529265255292653,
+      "loss": 0.801,
+      "step": 1227
+    },
+    {
+      "epoch": 0.7413220645940235,
+      "grad_norm": 0.08752121776342392,
+      "learning_rate": 0.0001552511415525114,
+      "loss": 0.7526,
+      "step": 1228
+    },
+    {
+      "epoch": 0.741925747057048,
+      "grad_norm": 0.0805983915925026,
+      "learning_rate": 0.0001552096305520963,
+      "loss": 0.9307,
+      "step": 1229
+    },
+    {
+      "epoch": 0.7425294295200724,
+      "grad_norm": 0.08306834846735,
+      "learning_rate": 0.0001551681195516812,
+      "loss": 0.7313,
+      "step": 1230
+    },
+    {
+      "epoch": 0.7431331119830968,
+      "grad_norm": 0.08837762475013733,
+      "learning_rate": 0.00015512660855126608,
+      "loss": 0.9053,
+      "step": 1231
+    },
+    {
+      "epoch": 0.7437367944461213,
+      "grad_norm": 0.07785294204950333,
+      "learning_rate": 0.00015508509755085098,
+      "loss": 0.7965,
+      "step": 1232
+    },
+    {
+      "epoch": 0.7443404769091458,
+      "grad_norm": 0.0776657834649086,
+      "learning_rate": 0.00015504358655043588,
+      "loss": 0.8275,
+      "step": 1233
+    },
+    {
+      "epoch": 0.7449441593721703,
+      "grad_norm": 0.08299911767244339,
+      "learning_rate": 0.00015500207555002075,
+      "loss": 0.7684,
+      "step": 1234
+    },
+    {
+      "epoch": 0.7455478418351947,
+      "grad_norm": 0.09403832256793976,
+      "learning_rate": 0.00015496056454960565,
+      "loss": 0.8317,
+      "step": 1235
+    },
+    {
+      "epoch": 0.7461515242982192,
+      "grad_norm": 0.08008915185928345,
+      "learning_rate": 0.00015491905354919055,
+      "loss": 0.6594,
+      "step": 1236
+    },
+    {
+      "epoch": 0.7467552067612436,
+      "grad_norm": 0.08133239299058914,
+      "learning_rate": 0.00015487754254877543,
+      "loss": 0.6942,
+      "step": 1237
+    },
+    {
+      "epoch": 0.7473588892242681,
+      "grad_norm": 0.08288057893514633,
+      "learning_rate": 0.00015483603154836033,
+      "loss": 0.7323,
+      "step": 1238
+    },
+    {
+      "epoch": 0.7479625716872925,
+      "grad_norm": 0.08818881958723068,
+      "learning_rate": 0.0001547945205479452,
+      "loss": 0.6966,
+      "step": 1239
+    },
+    {
+      "epoch": 0.748566254150317,
+      "grad_norm": 0.09948495030403137,
+      "learning_rate": 0.0001547530095475301,
+      "loss": 0.7126,
+      "step": 1240
+    },
+    {
+      "epoch": 0.7491699366133414,
+      "grad_norm": 0.12626221776008606,
+      "learning_rate": 0.000154711498547115,
+      "loss": 0.7647,
+      "step": 1241
+    },
+    {
+      "epoch": 0.7497736190763659,
+      "grad_norm": 0.10528393089771271,
+      "learning_rate": 0.00015466998754669987,
+      "loss": 0.8261,
+      "step": 1242
+    },
+    {
+      "epoch": 0.7503773015393903,
+      "grad_norm": 0.10515029728412628,
+      "learning_rate": 0.00015462847654628477,
+      "loss": 0.6662,
+      "step": 1243
+    },
+    {
+      "epoch": 0.7509809840024148,
+      "grad_norm": 0.10484447330236435,
+      "learning_rate": 0.00015458696554586967,
+      "loss": 0.7143,
+      "step": 1244
+    },
+    {
+      "epoch": 0.7515846664654392,
+      "grad_norm": 0.10932087153196335,
+      "learning_rate": 0.00015454545454545454,
+      "loss": 0.6561,
+      "step": 1245
+    },
+    {
+      "epoch": 0.7521883489284636,
+      "grad_norm": 0.10715696215629578,
+      "learning_rate": 0.00015450394354503944,
+      "loss": 0.6272,
+      "step": 1246
+    },
+    {
+      "epoch": 0.7527920313914881,
+      "grad_norm": 0.11597730964422226,
+      "learning_rate": 0.00015446243254462434,
+      "loss": 0.6093,
+      "step": 1247
+    },
+    {
+      "epoch": 0.7533957138545125,
+      "grad_norm": 0.11862548440694809,
+      "learning_rate": 0.00015442092154420922,
+      "loss": 0.6208,
+      "step": 1248
+    },
+    {
+      "epoch": 0.753999396317537,
+      "grad_norm": 0.12388130277395248,
+      "learning_rate": 0.00015437941054379412,
+      "loss": 0.5587,
+      "step": 1249
+    },
+    {
+      "epoch": 0.7546030787805614,
+      "grad_norm": 0.12081338465213776,
+      "learning_rate": 0.00015433789954337902,
+      "loss": 0.432,
+      "step": 1250
+    },
+    {
+      "epoch": 0.7552067612435859,
+      "grad_norm": 0.08639662712812424,
+      "learning_rate": 0.0001542963885429639,
+      "loss": 0.9072,
+      "step": 1251
+    },
+    {
+      "epoch": 0.7558104437066103,
+      "grad_norm": 0.0837317556142807,
+      "learning_rate": 0.0001542548775425488,
+      "loss": 1.1732,
+      "step": 1252
+    },
+    {
+      "epoch": 0.7564141261696348,
+      "grad_norm": 0.09014665335416794,
+      "learning_rate": 0.00015421336654213366,
+      "loss": 0.8012,
+      "step": 1253
+    },
+    {
+      "epoch": 0.7570178086326592,
+      "grad_norm": 0.09888771176338196,
+      "learning_rate": 0.00015417185554171856,
+      "loss": 0.8288,
+      "step": 1254
+    },
+    {
+      "epoch": 0.7576214910956837,
+      "grad_norm": 0.09830440580844879,
+      "learning_rate": 0.00015413034454130346,
+      "loss": 0.9115,
+      "step": 1255
+    },
+    {
+      "epoch": 0.7582251735587081,
+      "grad_norm": 0.08318249881267548,
+      "learning_rate": 0.00015408883354088834,
+      "loss": 0.8073,
+      "step": 1256
+    },
+    {
+      "epoch": 0.7588288560217326,
+      "grad_norm": 0.09623551368713379,
+      "learning_rate": 0.00015404732254047324,
+      "loss": 0.8798,
+      "step": 1257
+    },
+    {
+      "epoch": 0.759432538484757,
+      "grad_norm": 0.18773970007896423,
+      "learning_rate": 0.00015400581154005814,
+      "loss": 0.8237,
+      "step": 1258
+    },
+    {
+      "epoch": 0.7600362209477814,
+      "grad_norm": 0.07493336498737335,
+      "learning_rate": 0.000153964300539643,
+      "loss": 0.9224,
+      "step": 1259
+    },
+    {
+      "epoch": 0.7606399034108059,
+      "grad_norm": 0.08891238272190094,
+      "learning_rate": 0.0001539227895392279,
+      "loss": 0.8604,
+      "step": 1260
+    },
+    {
+      "epoch": 0.7612435858738303,
+      "grad_norm": 0.0810379758477211,
+      "learning_rate": 0.0001538812785388128,
+      "loss": 0.7966,
+      "step": 1261
+    },
+    {
+      "epoch": 0.7618472683368548,
+      "grad_norm": 0.08036971092224121,
+      "learning_rate": 0.00015383976753839768,
+      "loss": 0.7795,
+      "step": 1262
+    },
+    {
+      "epoch": 0.7624509507998792,
+      "grad_norm": 0.08294712007045746,
+      "learning_rate": 0.00015379825653798256,
+      "loss": 1.0023,
+      "step": 1263
+    },
+    {
+      "epoch": 0.7630546332629037,
+      "grad_norm": 0.0813121646642685,
+      "learning_rate": 0.00015375674553756748,
+      "loss": 1.0195,
+      "step": 1264
+    },
+    {
+      "epoch": 0.7636583157259281,
+      "grad_norm": 0.08313670009374619,
+      "learning_rate": 0.00015371523453715236,
+      "loss": 0.9885,
+      "step": 1265
+    },
+    {
+      "epoch": 0.7642619981889526,
+      "grad_norm": 0.07355551421642303,
+      "learning_rate": 0.00015367372353673723,
+      "loss": 0.8474,
+      "step": 1266
+    },
+    {
+      "epoch": 0.764865680651977,
+      "grad_norm": 0.13657139241695404,
+      "learning_rate": 0.00015363221253632213,
+      "loss": 0.9142,
+      "step": 1267
+    },
+    {
+      "epoch": 0.7654693631150015,
+      "grad_norm": 0.09950345009565353,
+      "learning_rate": 0.00015359070153590703,
+      "loss": 1.0635,
+      "step": 1268
+    },
+    {
+      "epoch": 0.7660730455780259,
+      "grad_norm": 0.07820272445678711,
+      "learning_rate": 0.0001535491905354919,
+      "loss": 1.1326,
+      "step": 1269
+    },
+    {
+      "epoch": 0.7666767280410504,
+      "grad_norm": 0.07532312721014023,
+      "learning_rate": 0.0001535076795350768,
+      "loss": 0.9895,
+      "step": 1270
+    },
+    {
+      "epoch": 0.7672804105040748,
+      "grad_norm": 0.09391959011554718,
+      "learning_rate": 0.0001534661685346617,
+      "loss": 0.7384,
+      "step": 1271
+    },
+    {
+      "epoch": 0.7678840929670993,
+      "grad_norm": 0.07820689678192139,
+      "learning_rate": 0.00015342465753424657,
+      "loss": 0.734,
+      "step": 1272
+    },
+    {
+      "epoch": 0.7684877754301238,
+      "grad_norm": 0.08966294676065445,
+      "learning_rate": 0.00015338314653383147,
+      "loss": 0.7925,
+      "step": 1273
+    },
+    {
+      "epoch": 0.7690914578931483,
+      "grad_norm": 0.08040351420640945,
+      "learning_rate": 0.00015334163553341637,
+      "loss": 0.7811,
+      "step": 1274
+    },
+    {
+      "epoch": 0.7696951403561727,
+      "grad_norm": 0.08416478335857391,
+      "learning_rate": 0.00015330012453300125,
+      "loss": 0.9939,
+      "step": 1275
+    },
+    {
+      "epoch": 0.7702988228191971,
+      "grad_norm": 0.08706527948379517,
+      "learning_rate": 0.00015325861353258615,
+      "loss": 0.8099,
+      "step": 1276
+    },
+    {
+      "epoch": 0.7709025052822216,
+      "grad_norm": 0.10363662242889404,
+      "learning_rate": 0.00015321710253217102,
+      "loss": 0.7835,
+      "step": 1277
+    },
+    {
+      "epoch": 0.771506187745246,
+      "grad_norm": 0.080405592918396,
+      "learning_rate": 0.00015317559153175592,
+      "loss": 0.7816,
+      "step": 1278
+    },
+    {
+      "epoch": 0.7721098702082705,
+      "grad_norm": 0.0954577773809433,
+      "learning_rate": 0.00015313408053134082,
+      "loss": 1.0492,
+      "step": 1279
+    },
+    {
+      "epoch": 0.7727135526712949,
+      "grad_norm": 0.08522694557905197,
+      "learning_rate": 0.0001530925695309257,
+      "loss": 0.7938,
+      "step": 1280
+    },
+    {
+      "epoch": 0.7733172351343194,
+      "grad_norm": 0.0870218351483345,
+      "learning_rate": 0.0001530510585305106,
+      "loss": 1.0509,
+      "step": 1281
+    },
+    {
+      "epoch": 0.7739209175973438,
+      "grad_norm": 0.07454058527946472,
+      "learning_rate": 0.0001530095475300955,
+      "loss": 0.8381,
+      "step": 1282
+    },
+    {
+      "epoch": 0.7745246000603683,
+      "grad_norm": 0.0701836496591568,
+      "learning_rate": 0.00015296803652968037,
+      "loss": 0.6133,
+      "step": 1283
+    },
+    {
+      "epoch": 0.7751282825233927,
+      "grad_norm": 0.08319991081953049,
+      "learning_rate": 0.00015292652552926524,
+      "loss": 0.7434,
+      "step": 1284
+    },
+    {
+      "epoch": 0.7757319649864172,
+      "grad_norm": 0.07945888489484787,
+      "learning_rate": 0.00015288501452885017,
+      "loss": 0.7266,
+      "step": 1285
+    },
+    {
+      "epoch": 0.7763356474494416,
+      "grad_norm": 0.0800202265381813,
+      "learning_rate": 0.00015284350352843504,
+      "loss": 0.7203,
+      "step": 1286
+    },
+    {
+      "epoch": 0.7769393299124661,
+      "grad_norm": 0.08976299315690994,
+      "learning_rate": 0.0001528019925280199,
+      "loss": 0.7075,
+      "step": 1287
+    },
+    {
+      "epoch": 0.7775430123754905,
+      "grad_norm": 0.08427184820175171,
+      "learning_rate": 0.00015276048152760484,
+      "loss": 0.6855,
+      "step": 1288
+    },
+    {
+      "epoch": 0.778146694838515,
+      "grad_norm": 0.09182146191596985,
+      "learning_rate": 0.0001527189705271897,
+      "loss": 0.7375,
+      "step": 1289
+    },
+    {
+      "epoch": 0.7787503773015394,
+      "grad_norm": 0.0932517945766449,
+      "learning_rate": 0.0001526774595267746,
+      "loss": 0.7168,
+      "step": 1290
+    },
+    {
+      "epoch": 0.7793540597645638,
+      "grad_norm": 0.11614131927490234,
+      "learning_rate": 0.00015263594852635949,
+      "loss": 0.7134,
+      "step": 1291
+    },
+    {
+      "epoch": 0.7799577422275883,
+      "grad_norm": 0.10372986644506454,
+      "learning_rate": 0.00015259443752594439,
+      "loss": 0.7539,
+      "step": 1292
+    },
+    {
+      "epoch": 0.7805614246906127,
+      "grad_norm": 0.10486335307359695,
+      "learning_rate": 0.00015255292652552929,
+      "loss": 0.7198,
+      "step": 1293
+    },
+    {
+      "epoch": 0.7811651071536372,
+      "grad_norm": 0.11475211381912231,
+      "learning_rate": 0.00015251141552511416,
+      "loss": 0.7086,
+      "step": 1294
+    },
+    {
+      "epoch": 0.7817687896166616,
+      "grad_norm": 0.10555735975503922,
+      "learning_rate": 0.00015246990452469906,
+      "loss": 0.6697,
+      "step": 1295
+    },
+    {
+      "epoch": 0.7823724720796861,
+      "grad_norm": 0.10920953750610352,
+      "learning_rate": 0.00015242839352428396,
+      "loss": 0.6179,
+      "step": 1296
+    },
+    {
+      "epoch": 0.7829761545427105,
+      "grad_norm": 0.13240636885166168,
+      "learning_rate": 0.00015238688252386883,
+      "loss": 0.6766,
+      "step": 1297
+    },
+    {
+      "epoch": 0.783579837005735,
+      "grad_norm": 0.1218361034989357,
+      "learning_rate": 0.0001523453715234537,
+      "loss": 0.6364,
+      "step": 1298
+    },
+    {
+      "epoch": 0.7841835194687594,
+      "grad_norm": 0.1165798008441925,
+      "learning_rate": 0.00015230386052303863,
+      "loss": 0.5109,
+      "step": 1299
+    },
+    {
+      "epoch": 0.7847872019317839,
+      "grad_norm": 0.12739436328411102,
+      "learning_rate": 0.0001522623495226235,
+      "loss": 0.4305,
+      "step": 1300
+    },
+    {
+      "epoch": 0.7853908843948083,
+      "grad_norm": 0.10032903403043747,
+      "learning_rate": 0.00015222083852220838,
+      "loss": 0.8067,
+      "step": 1301
+    },
+    {
+      "epoch": 0.7859945668578328,
+      "grad_norm": 0.08724936097860336,
+      "learning_rate": 0.0001521793275217933,
+      "loss": 0.9813,
+      "step": 1302
+    },
+    {
+      "epoch": 0.7865982493208572,
+      "grad_norm": 0.08223757147789001,
+      "learning_rate": 0.00015213781652137818,
+      "loss": 0.756,
+      "step": 1303
+    },
+    {
+      "epoch": 0.7872019317838816,
+      "grad_norm": 0.09227702766656876,
+      "learning_rate": 0.00015209630552096305,
+      "loss": 0.8074,
+      "step": 1304
+    },
+    {
+      "epoch": 0.7878056142469061,
+      "grad_norm": 0.08399201929569244,
+      "learning_rate": 0.00015205479452054795,
+      "loss": 1.0582,
+      "step": 1305
+    },
+    {
+      "epoch": 0.7884092967099305,
+      "grad_norm": 0.0827503651380539,
+      "learning_rate": 0.00015201328352013285,
+      "loss": 0.7878,
+      "step": 1306
+    },
+    {
+      "epoch": 0.789012979172955,
+      "grad_norm": 0.08233575522899628,
+      "learning_rate": 0.00015197177251971772,
+      "loss": 0.8512,
+      "step": 1307
+    },
+    {
+      "epoch": 0.7896166616359794,
+      "grad_norm": 0.09783722460269928,
+      "learning_rate": 0.00015193026151930262,
+      "loss": 0.8324,
+      "step": 1308
+    },
+    {
+      "epoch": 0.7902203440990039,
+      "grad_norm": 0.09064453095197678,
+      "learning_rate": 0.00015188875051888752,
+      "loss": 0.8361,
+      "step": 1309
+    },
+    {
+      "epoch": 0.7908240265620283,
+      "grad_norm": 0.09642300009727478,
+      "learning_rate": 0.0001518472395184724,
+      "loss": 0.7578,
+      "step": 1310
+    },
+    {
+      "epoch": 0.7914277090250528,
+      "grad_norm": 0.08258376270532608,
+      "learning_rate": 0.0001518057285180573,
+      "loss": 0.9612,
+      "step": 1311
+    },
+    {
+      "epoch": 0.7920313914880772,
+      "grad_norm": 0.08849532902240753,
+      "learning_rate": 0.00015176421751764217,
+      "loss": 0.798,
+      "step": 1312
+    },
+    {
+      "epoch": 0.7926350739511017,
+      "grad_norm": 0.08029762655496597,
+      "learning_rate": 0.00015172270651722707,
+      "loss": 0.8082,
+      "step": 1313
+    },
+    {
+      "epoch": 0.7932387564141262,
+      "grad_norm": 0.0844390019774437,
+      "learning_rate": 0.00015168119551681197,
+      "loss": 0.7453,
+      "step": 1314
+    },
+    {
+      "epoch": 0.7938424388771507,
+      "grad_norm": 0.09358351677656174,
+      "learning_rate": 0.00015163968451639684,
+      "loss": 0.8036,
+      "step": 1315
+    },
+    {
+      "epoch": 0.7944461213401751,
+      "grad_norm": 0.08466824889183044,
+      "learning_rate": 0.00015159817351598174,
+      "loss": 0.777,
+      "step": 1316
+    },
+    {
+      "epoch": 0.7950498038031996,
+      "grad_norm": 0.08159632235765457,
+      "learning_rate": 0.00015155666251556664,
+      "loss": 0.8004,
+      "step": 1317
+    },
+    {
+      "epoch": 0.795653486266224,
+      "grad_norm": 0.1605248749256134,
+      "learning_rate": 0.00015151515151515152,
+      "loss": 0.8841,
+      "step": 1318
+    },
+    {
+      "epoch": 0.7962571687292485,
+      "grad_norm": 0.08063754439353943,
+      "learning_rate": 0.00015147364051473642,
+      "loss": 0.9292,
+      "step": 1319
+    },
+    {
+      "epoch": 0.7968608511922729,
+      "grad_norm": 0.08406137675046921,
+      "learning_rate": 0.00015143212951432132,
+      "loss": 0.7878,
+      "step": 1320
+    },
+    {
+      "epoch": 0.7974645336552973,
+      "grad_norm": 0.07940443605184555,
+      "learning_rate": 0.0001513906185139062,
+      "loss": 0.8981,
+      "step": 1321
+    },
+    {
+      "epoch": 0.7980682161183218,
+      "grad_norm": 0.07914227992296219,
+      "learning_rate": 0.00015134910751349106,
+      "loss": 1.1571,
+      "step": 1322
+    },
+    {
+      "epoch": 0.7986718985813462,
+      "grad_norm": 0.07836727797985077,
+      "learning_rate": 0.000151307596513076,
+      "loss": 0.8384,
+      "step": 1323
+    },
+    {
+      "epoch": 0.7992755810443707,
+      "grad_norm": 0.08515416085720062,
+      "learning_rate": 0.00015126608551266086,
+      "loss": 0.7468,
+      "step": 1324
+    },
+    {
+      "epoch": 0.7998792635073951,
+      "grad_norm": 0.0839102566242218,
+      "learning_rate": 0.00015122457451224573,
+      "loss": 0.9835,
+      "step": 1325
+    },
+    {
+      "epoch": 0.8004829459704196,
+      "grad_norm": 0.2878398299217224,
+      "learning_rate": 0.00015118306351183063,
+      "loss": 1.232,
+      "step": 1326
+    },
+    {
+      "epoch": 0.801086628433444,
+      "grad_norm": 0.0905466228723526,
+      "learning_rate": 0.00015114155251141553,
+      "loss": 0.8908,
+      "step": 1327
+    },
+    {
+      "epoch": 0.8016903108964685,
+      "grad_norm": 0.08527706563472748,
+      "learning_rate": 0.0001511000415110004,
+      "loss": 0.8004,
+      "step": 1328
+    },
+    {
+      "epoch": 0.8022939933594929,
+      "grad_norm": 0.08341808617115021,
+      "learning_rate": 0.0001510585305105853,
+      "loss": 0.7974,
+      "step": 1329
+    },
+    {
+      "epoch": 0.8028976758225174,
+      "grad_norm": 0.07585755735635757,
+      "learning_rate": 0.0001510170195101702,
+      "loss": 0.7528,
+      "step": 1330
+    },
+    {
+      "epoch": 0.8035013582855418,
+      "grad_norm": 0.08576681464910507,
+      "learning_rate": 0.00015097550850975508,
+      "loss": 0.7634,
+      "step": 1331
+    },
+    {
+      "epoch": 0.8041050407485663,
+      "grad_norm": 0.10057312995195389,
+      "learning_rate": 0.00015093399750933998,
+      "loss": 0.6505,
+      "step": 1332
+    },
+    {
+      "epoch": 0.8047087232115907,
+      "grad_norm": 0.08735567331314087,
+      "learning_rate": 0.00015089248650892488,
+      "loss": 0.6856,
+      "step": 1333
+    },
+    {
+      "epoch": 0.8053124056746152,
+      "grad_norm": 0.09076007455587387,
+      "learning_rate": 0.00015085097550850978,
+      "loss": 0.7697,
+      "step": 1334
+    },
+    {
+      "epoch": 0.8059160881376396,
+      "grad_norm": 0.09162318706512451,
+      "learning_rate": 0.00015080946450809465,
+      "loss": 0.7425,
+      "step": 1335
+    },
+    {
+      "epoch": 0.806519770600664,
+      "grad_norm": 0.08863736689090729,
+      "learning_rate": 0.00015076795350767953,
+      "loss": 0.7748,
+      "step": 1336
+    },
+    {
+      "epoch": 0.8071234530636885,
+      "grad_norm": 0.10001536458730698,
+      "learning_rate": 0.00015072644250726445,
+      "loss": 0.7162,
+      "step": 1337
+    },
+    {
+      "epoch": 0.8077271355267129,
+      "grad_norm": 0.09326426684856415,
+      "learning_rate": 0.00015068493150684933,
+      "loss": 0.82,
+      "step": 1338
+    },
+    {
+      "epoch": 0.8083308179897374,
+      "grad_norm": 0.09727218747138977,
+      "learning_rate": 0.0001506434205064342,
+      "loss": 0.7534,
+      "step": 1339
+    },
+    {
+      "epoch": 0.8089345004527618,
+      "grad_norm": 0.09535179287195206,
+      "learning_rate": 0.0001506019095060191,
+      "loss": 0.7623,
+      "step": 1340
+    },
+    {
+      "epoch": 0.8095381829157863,
+      "grad_norm": 0.09758217632770538,
+      "learning_rate": 0.000150560398505604,
+      "loss": 0.7533,
+      "step": 1341
+    },
+    {
+      "epoch": 0.8101418653788107,
+      "grad_norm": 0.11768268793821335,
+      "learning_rate": 0.00015051888750518887,
+      "loss": 0.7481,
+      "step": 1342
+    },
+    {
+      "epoch": 0.8107455478418352,
+      "grad_norm": 0.09633111953735352,
+      "learning_rate": 0.00015047737650477377,
+      "loss": 0.7191,
+      "step": 1343
+    },
+    {
+      "epoch": 0.8113492303048596,
+      "grad_norm": 0.10011658817529678,
+      "learning_rate": 0.00015043586550435867,
+      "loss": 0.6665,
+      "step": 1344
+    },
+    {
+      "epoch": 0.8119529127678841,
+      "grad_norm": 0.11359147727489471,
+      "learning_rate": 0.00015039435450394354,
+      "loss": 0.7549,
+      "step": 1345
+    },
+    {
+      "epoch": 0.8125565952309085,
+      "grad_norm": 0.11092590540647507,
+      "learning_rate": 0.00015035284350352844,
+      "loss": 0.8078,
+      "step": 1346
+    },
+    {
+      "epoch": 0.813160277693933,
+      "grad_norm": 0.1197749450802803,
+      "learning_rate": 0.00015031133250311335,
+      "loss": 0.6463,
+      "step": 1347
+    },
+    {
+      "epoch": 0.8137639601569574,
+      "grad_norm": 0.12124069780111313,
+      "learning_rate": 0.00015026982150269822,
+      "loss": 0.6384,
+      "step": 1348
+    },
+    {
+      "epoch": 0.8143676426199818,
+      "grad_norm": 0.14109613001346588,
+      "learning_rate": 0.00015022831050228312,
+      "loss": 0.5564,
+      "step": 1349
+    },
+    {
+      "epoch": 0.8149713250830063,
+      "grad_norm": 0.12870782613754272,
+      "learning_rate": 0.000150186799501868,
+      "loss": 0.4514,
+      "step": 1350
+    },
+    {
+      "epoch": 0.8155750075460307,
+      "grad_norm": 0.14718082547187805,
+      "learning_rate": 0.0001501452885014529,
+      "loss": 0.8223,
+      "step": 1351
+    },
+    {
+      "epoch": 0.8161786900090552,
+      "grad_norm": 0.0827641561627388,
+      "learning_rate": 0.0001501037775010378,
+      "loss": 0.7125,
+      "step": 1352
+    },
+    {
+      "epoch": 0.8167823724720796,
+      "grad_norm": 0.0898633524775505,
+      "learning_rate": 0.00015006226650062266,
+      "loss": 1.0915,
+      "step": 1353
+    },
+    {
+      "epoch": 0.8173860549351042,
+      "grad_norm": 0.09281725436449051,
+      "learning_rate": 0.00015002075550020756,
+      "loss": 0.854,
+      "step": 1354
+    },
+    {
+      "epoch": 0.8179897373981286,
+      "grad_norm": 0.08282145857810974,
+      "learning_rate": 0.00014997924449979246,
+      "loss": 1.0761,
+      "step": 1355
+    },
+    {
+      "epoch": 0.8185934198611531,
+      "grad_norm": 0.08490285277366638,
+      "learning_rate": 0.00014993773349937734,
+      "loss": 1.1083,
+      "step": 1356
+    },
+    {
+      "epoch": 0.8191971023241775,
+      "grad_norm": 0.09798530489206314,
+      "learning_rate": 0.00014989622249896224,
+      "loss": 1.0771,
+      "step": 1357
+    },
+    {
+      "epoch": 0.819800784787202,
+      "grad_norm": 0.08119674026966095,
+      "learning_rate": 0.00014985471149854714,
+      "loss": 0.8048,
+      "step": 1358
+    },
+    {
+      "epoch": 0.8204044672502264,
+      "grad_norm": 0.08817379921674728,
+      "learning_rate": 0.000149813200498132,
+      "loss": 0.8062,
+      "step": 1359
+    },
+    {
+      "epoch": 0.8210081497132509,
+      "grad_norm": 0.08717963844537735,
+      "learning_rate": 0.00014977168949771688,
+      "loss": 0.8923,
+      "step": 1360
+    },
+    {
+      "epoch": 0.8216118321762753,
+      "grad_norm": 0.08484237641096115,
+      "learning_rate": 0.0001497301784973018,
+      "loss": 0.882,
+      "step": 1361
+    },
+    {
+      "epoch": 0.8222155146392998,
+      "grad_norm": 0.0800343006849289,
+      "learning_rate": 0.00014968866749688668,
+      "loss": 0.8224,
+      "step": 1362
+    },
+    {
+      "epoch": 0.8228191971023242,
+      "grad_norm": 0.08844833076000214,
+      "learning_rate": 0.00014964715649647156,
+      "loss": 0.7582,
+      "step": 1363
+    },
+    {
+      "epoch": 0.8234228795653487,
+      "grad_norm": 0.08437283337116241,
+      "learning_rate": 0.00014960564549605646,
+      "loss": 0.847,
+      "step": 1364
+    },
+    {
+      "epoch": 0.8240265620283731,
+      "grad_norm": 0.08416473865509033,
+      "learning_rate": 0.00014956413449564136,
+      "loss": 0.8066,
+      "step": 1365
+    },
+    {
+      "epoch": 0.8246302444913975,
+      "grad_norm": 0.0806473046541214,
+      "learning_rate": 0.00014952262349522623,
+      "loss": 0.797,
+      "step": 1366
+    },
+    {
+      "epoch": 0.825233926954422,
+      "grad_norm": 0.08505918830633163,
+      "learning_rate": 0.00014948111249481113,
+      "loss": 0.826,
+      "step": 1367
+    },
+    {
+      "epoch": 0.8258376094174464,
+      "grad_norm": 0.08299072831869125,
+      "learning_rate": 0.00014943960149439603,
+      "loss": 0.7974,
+      "step": 1368
+    },
+    {
+      "epoch": 0.8264412918804709,
+      "grad_norm": 0.0950680747628212,
+      "learning_rate": 0.0001493980904939809,
+      "loss": 0.8961,
+      "step": 1369
+    },
+    {
+      "epoch": 0.8270449743434953,
+      "grad_norm": 0.08349784463644028,
+      "learning_rate": 0.0001493565794935658,
+      "loss": 0.8791,
+      "step": 1370
+    },
+    {
+      "epoch": 0.8276486568065198,
+      "grad_norm": 0.08619910478591919,
+      "learning_rate": 0.0001493150684931507,
+      "loss": 0.8367,
+      "step": 1371
+    },
+    {
+      "epoch": 0.8282523392695442,
+      "grad_norm": 0.09480497241020203,
+      "learning_rate": 0.00014927355749273557,
+      "loss": 0.8101,
+      "step": 1372
+    },
+    {
+      "epoch": 0.8288560217325687,
+      "grad_norm": 0.0833049863576889,
+      "learning_rate": 0.00014923204649232047,
+      "loss": 0.8616,
+      "step": 1373
+    },
+    {
+      "epoch": 0.8294597041955931,
+      "grad_norm": 0.08111268281936646,
+      "learning_rate": 0.00014919053549190535,
+      "loss": 0.7388,
+      "step": 1374
+    },
+    {
+      "epoch": 0.8300633866586176,
+      "grad_norm": 0.08618942648172379,
+      "learning_rate": 0.00014914902449149025,
+      "loss": 0.8678,
+      "step": 1375
+    },
+    {
+      "epoch": 0.830667069121642,
+      "grad_norm": 0.08400508016347885,
+      "learning_rate": 0.00014910751349107515,
+      "loss": 0.7766,
+      "step": 1376
+    },
+    {
+      "epoch": 0.8312707515846665,
+      "grad_norm": 0.0798826590180397,
+      "learning_rate": 0.00014906600249066002,
+      "loss": 0.8156,
+      "step": 1377
+    },
+    {
+      "epoch": 0.8318744340476909,
+      "grad_norm": 0.08481590449810028,
+      "learning_rate": 0.00014902449149024492,
+      "loss": 1.117,
+      "step": 1378
+    },
+    {
+      "epoch": 0.8324781165107153,
+      "grad_norm": 0.09439216554164886,
+      "learning_rate": 0.00014898298048982982,
+      "loss": 0.8822,
+      "step": 1379
+    },
+    {
+      "epoch": 0.8330817989737398,
+      "grad_norm": 0.09862873703241348,
+      "learning_rate": 0.0001489414694894147,
+      "loss": 0.8087,
+      "step": 1380
+    },
+    {
+      "epoch": 0.8336854814367642,
+      "grad_norm": 0.08466385304927826,
+      "learning_rate": 0.0001488999584889996,
+      "loss": 0.8568,
+      "step": 1381
+    },
+    {
+      "epoch": 0.8342891638997887,
+      "grad_norm": 0.08554793149232864,
+      "learning_rate": 0.0001488584474885845,
+      "loss": 0.95,
+      "step": 1382
+    },
+    {
+      "epoch": 0.8348928463628131,
+      "grad_norm": 0.07983773201704025,
+      "learning_rate": 0.00014881693648816937,
+      "loss": 0.7137,
+      "step": 1383
+    },
+    {
+      "epoch": 0.8354965288258376,
+      "grad_norm": 0.07740309834480286,
+      "learning_rate": 0.00014877542548775427,
+      "loss": 0.7139,
+      "step": 1384
+    },
+    {
+      "epoch": 0.836100211288862,
+      "grad_norm": 0.08271116763353348,
+      "learning_rate": 0.00014873391448733917,
+      "loss": 0.7001,
+      "step": 1385
+    },
+    {
+      "epoch": 0.8367038937518865,
+      "grad_norm": 0.08564220368862152,
+      "learning_rate": 0.00014869240348692404,
+      "loss": 0.7624,
+      "step": 1386
+    },
+    {
+      "epoch": 0.8373075762149109,
+      "grad_norm": 0.08788985759019852,
+      "learning_rate": 0.00014865089248650894,
+      "loss": 0.7563,
+      "step": 1387
+    },
+    {
+      "epoch": 0.8379112586779354,
+      "grad_norm": 0.09162076562643051,
+      "learning_rate": 0.0001486093814860938,
+      "loss": 0.7008,
+      "step": 1388
+    },
+    {
+      "epoch": 0.8385149411409598,
+      "grad_norm": 0.08919322490692139,
+      "learning_rate": 0.0001485678704856787,
+      "loss": 0.7408,
+      "step": 1389
+    },
+    {
+      "epoch": 0.8391186236039843,
+      "grad_norm": 0.4382922053337097,
+      "learning_rate": 0.0001485263594852636,
+      "loss": 0.748,
+      "step": 1390
+    },
+    {
+      "epoch": 0.8397223060670087,
+      "grad_norm": 0.09150999784469604,
+      "learning_rate": 0.00014848484848484849,
+      "loss": 0.7801,
+      "step": 1391
+    },
+    {
+      "epoch": 0.8403259885300332,
+      "grad_norm": 0.09508758038282394,
+      "learning_rate": 0.00014844333748443339,
+      "loss": 0.6773,
+      "step": 1392
+    },
+    {
+      "epoch": 0.8409296709930576,
+      "grad_norm": 0.10179366171360016,
+      "learning_rate": 0.00014840182648401829,
+      "loss": 0.7492,
+      "step": 1393
+    },
+    {
+      "epoch": 0.8415333534560822,
+      "grad_norm": 0.10912247747182846,
+      "learning_rate": 0.00014836031548360316,
+      "loss": 0.7271,
+      "step": 1394
+    },
+    {
+      "epoch": 0.8421370359191066,
+      "grad_norm": 0.10175374895334244,
+      "learning_rate": 0.00014831880448318803,
+      "loss": 0.6465,
+      "step": 1395
+    },
+    {
+      "epoch": 0.842740718382131,
+      "grad_norm": 0.10953323543071747,
+      "learning_rate": 0.00014827729348277296,
+      "loss": 0.6639,
+      "step": 1396
+    },
+    {
+      "epoch": 0.8433444008451555,
+      "grad_norm": 0.11571014672517776,
+      "learning_rate": 0.00014823578248235783,
+      "loss": 0.6389,
+      "step": 1397
+    },
+    {
+      "epoch": 0.8439480833081799,
+      "grad_norm": 0.1150619387626648,
+      "learning_rate": 0.0001481942714819427,
+      "loss": 0.5918,
+      "step": 1398
+    },
+    {
+      "epoch": 0.8445517657712044,
+      "grad_norm": 0.12481655180454254,
+      "learning_rate": 0.00014815276048152763,
+      "loss": 0.5285,
+      "step": 1399
+    },
+    {
+      "epoch": 0.8451554482342288,
+      "grad_norm": 0.12378332763910294,
+      "learning_rate": 0.0001481112494811125,
+      "loss": 0.4092,
+      "step": 1400
+    },
+    {
+      "epoch": 0.8457591306972533,
+      "grad_norm": 0.0824832022190094,
+      "learning_rate": 0.00014806973848069738,
+      "loss": 1.0004,
+      "step": 1401
+    },
+    {
+      "epoch": 0.8463628131602777,
+      "grad_norm": 0.0863024890422821,
+      "learning_rate": 0.00014802822748028228,
+      "loss": 0.8013,
+      "step": 1402
+    },
+    {
+      "epoch": 0.8469664956233022,
+      "grad_norm": 0.08444255590438843,
+      "learning_rate": 0.00014798671647986718,
+      "loss": 0.8442,
+      "step": 1403
+    },
+    {
+      "epoch": 0.8475701780863266,
+      "grad_norm": 0.10533567517995834,
+      "learning_rate": 0.00014794520547945205,
+      "loss": 0.987,
+      "step": 1404
+    },
+    {
+      "epoch": 0.8481738605493511,
+      "grad_norm": 0.09298814088106155,
+      "learning_rate": 0.00014790369447903695,
+      "loss": 0.9405,
+      "step": 1405
+    },
+    {
+      "epoch": 0.8487775430123755,
+      "grad_norm": 0.09649625420570374,
+      "learning_rate": 0.00014786218347862185,
+      "loss": 0.9325,
+      "step": 1406
+    },
+    {
+      "epoch": 0.8493812254754,
+      "grad_norm": 0.10233578085899353,
+      "learning_rate": 0.00014782067247820672,
+      "loss": 0.9722,
+      "step": 1407
+    },
+    {
+      "epoch": 0.8499849079384244,
+      "grad_norm": 0.07888230681419373,
+      "learning_rate": 0.00014777916147779162,
+      "loss": 0.6847,
+      "step": 1408
+    },
+    {
+      "epoch": 0.8505885904014489,
+      "grad_norm": 0.1061205044388771,
+      "learning_rate": 0.0001477376504773765,
+      "loss": 0.7159,
+      "step": 1409
+    },
+    {
+      "epoch": 0.8511922728644733,
+      "grad_norm": 0.09067028015851974,
+      "learning_rate": 0.0001476961394769614,
+      "loss": 1.0638,
+      "step": 1410
+    },
+    {
+      "epoch": 0.8517959553274977,
+      "grad_norm": 0.07997080683708191,
+      "learning_rate": 0.0001476546284765463,
+      "loss": 1.0731,
+      "step": 1411
+    },
+    {
+      "epoch": 0.8523996377905222,
+      "grad_norm": 0.08588366955518723,
+      "learning_rate": 0.00014761311747613117,
+      "loss": 0.8226,
+      "step": 1412
+    },
+    {
+      "epoch": 0.8530033202535466,
+      "grad_norm": 0.08147002011537552,
+      "learning_rate": 0.00014757160647571607,
+      "loss": 0.8404,
+      "step": 1413
+    },
+    {
+      "epoch": 0.8536070027165711,
+      "grad_norm": 0.08258868753910065,
+      "learning_rate": 0.00014753009547530097,
+      "loss": 0.8363,
+      "step": 1414
+    },
+    {
+      "epoch": 0.8542106851795955,
+      "grad_norm": 0.11236279457807541,
+      "learning_rate": 0.00014748858447488584,
+      "loss": 0.8376,
+      "step": 1415
+    },
+    {
+      "epoch": 0.85481436764262,
+      "grad_norm": 0.07888762652873993,
+      "learning_rate": 0.00014744707347447074,
+      "loss": 0.8395,
+      "step": 1416
+    },
+    {
+      "epoch": 0.8554180501056444,
+      "grad_norm": 0.08469484001398087,
+      "learning_rate": 0.00014740556247405564,
+      "loss": 1.0579,
+      "step": 1417
+    },
+    {
+      "epoch": 0.8560217325686689,
+      "grad_norm": 0.07748138904571533,
+      "learning_rate": 0.00014736405147364052,
+      "loss": 0.7358,
+      "step": 1418
+    },
+    {
+      "epoch": 0.8566254150316933,
+      "grad_norm": 0.08709147572517395,
+      "learning_rate": 0.0001473225404732254,
+      "loss": 0.9408,
+      "step": 1419
+    },
+    {
+      "epoch": 0.8572290974947178,
+      "grad_norm": 0.08583850413560867,
+      "learning_rate": 0.00014728102947281032,
+      "loss": 0.822,
+      "step": 1420
+    },
+    {
+      "epoch": 0.8578327799577422,
+      "grad_norm": 0.09203623980283737,
+      "learning_rate": 0.0001472395184723952,
+      "loss": 0.8535,
+      "step": 1421
+    },
+    {
+      "epoch": 0.8584364624207667,
+      "grad_norm": 0.08235791325569153,
+      "learning_rate": 0.0001471980074719801,
+      "loss": 0.8012,
+      "step": 1422
+    },
+    {
+      "epoch": 0.8590401448837911,
+      "grad_norm": 0.09391656517982483,
+      "learning_rate": 0.00014715649647156496,
+      "loss": 0.888,
+      "step": 1423
+    },
+    {
+      "epoch": 0.8596438273468155,
+      "grad_norm": 0.0862961858510971,
+      "learning_rate": 0.00014711498547114986,
+      "loss": 0.8323,
+      "step": 1424
+    },
+    {
+      "epoch": 0.86024750980984,
+      "grad_norm": 0.08743558079004288,
+      "learning_rate": 0.00014707347447073476,
+      "loss": 0.7275,
+      "step": 1425
+    },
+    {
+      "epoch": 0.8608511922728644,
+      "grad_norm": 0.07890239357948303,
+      "learning_rate": 0.00014703196347031963,
+      "loss": 0.6607,
+      "step": 1426
+    },
+    {
+      "epoch": 0.8614548747358889,
+      "grad_norm": 0.07826490700244904,
+      "learning_rate": 0.00014699045246990453,
+      "loss": 0.7179,
+      "step": 1427
+    },
+    {
+      "epoch": 0.8620585571989133,
+      "grad_norm": 0.08540347218513489,
+      "learning_rate": 0.00014694894146948943,
+      "loss": 0.7166,
+      "step": 1428
+    },
+    {
+      "epoch": 0.8626622396619378,
+      "grad_norm": 0.08672580122947693,
+      "learning_rate": 0.0001469074304690743,
+      "loss": 1.0812,
+      "step": 1429
+    },
+    {
+      "epoch": 0.8632659221249622,
+      "grad_norm": 0.08532918244600296,
+      "learning_rate": 0.0001468659194686592,
+      "loss": 0.7676,
+      "step": 1430
+    },
+    {
+      "epoch": 0.8638696045879867,
+      "grad_norm": 0.08678078651428223,
+      "learning_rate": 0.0001468244084682441,
+      "loss": 0.8672,
+      "step": 1431
+    },
+    {
+      "epoch": 0.8644732870510111,
+      "grad_norm": 0.08257097750902176,
+      "learning_rate": 0.00014678289746782898,
+      "loss": 0.7108,
+      "step": 1432
+    },
+    {
+      "epoch": 0.8650769695140356,
+      "grad_norm": 0.09253834187984467,
+      "learning_rate": 0.00014674138646741385,
+      "loss": 0.8333,
+      "step": 1433
+    },
+    {
+      "epoch": 0.86568065197706,
+      "grad_norm": 0.08253934979438782,
+      "learning_rate": 0.00014669987546699878,
+      "loss": 0.7099,
+      "step": 1434
+    },
+    {
+      "epoch": 0.8662843344400846,
+      "grad_norm": 0.08953447639942169,
+      "learning_rate": 0.00014665836446658365,
+      "loss": 0.7593,
+      "step": 1435
+    },
+    {
+      "epoch": 0.866888016903109,
+      "grad_norm": 0.09555850178003311,
+      "learning_rate": 0.00014661685346616853,
+      "loss": 0.6711,
+      "step": 1436
+    },
+    {
+      "epoch": 0.8674916993661335,
+      "grad_norm": 0.08465403318405151,
+      "learning_rate": 0.00014657534246575343,
+      "loss": 0.7276,
+      "step": 1437
+    },
+    {
+      "epoch": 0.8680953818291579,
+      "grad_norm": 0.08792877197265625,
+      "learning_rate": 0.00014653383146533833,
+      "loss": 0.7075,
+      "step": 1438
+    },
+    {
+      "epoch": 0.8686990642921824,
+      "grad_norm": 0.09290292859077454,
+      "learning_rate": 0.0001464923204649232,
+      "loss": 0.7923,
+      "step": 1439
+    },
+    {
+      "epoch": 0.8693027467552068,
+      "grad_norm": 0.09828725457191467,
+      "learning_rate": 0.0001464508094645081,
+      "loss": 0.7282,
+      "step": 1440
+    },
+    {
+      "epoch": 0.8699064292182312,
+      "grad_norm": 0.1370926946401596,
+      "learning_rate": 0.000146409298464093,
+      "loss": 0.7228,
+      "step": 1441
+    },
+    {
+      "epoch": 0.8705101116812557,
+      "grad_norm": 0.09725096821784973,
+      "learning_rate": 0.00014636778746367787,
+      "loss": 0.7643,
+      "step": 1442
+    },
+    {
+      "epoch": 0.8711137941442801,
+      "grad_norm": 0.09183010458946228,
+      "learning_rate": 0.00014632627646326277,
+      "loss": 0.7243,
+      "step": 1443
+    },
+    {
+      "epoch": 0.8717174766073046,
+      "grad_norm": 0.0972151830792427,
+      "learning_rate": 0.00014628476546284767,
+      "loss": 0.6837,
+      "step": 1444
+    },
+    {
+      "epoch": 0.872321159070329,
+      "grad_norm": 0.11110269278287888,
+      "learning_rate": 0.00014624325446243255,
+      "loss": 0.7487,
+      "step": 1445
+    },
+    {
+      "epoch": 0.8729248415333535,
+      "grad_norm": 0.10543739795684814,
+      "learning_rate": 0.00014620174346201745,
+      "loss": 0.6177,
+      "step": 1446
+    },
+    {
+      "epoch": 0.8735285239963779,
+      "grad_norm": 0.13775797188282013,
+      "learning_rate": 0.00014616023246160232,
+      "loss": 0.6354,
+      "step": 1447
+    },
+    {
+      "epoch": 0.8741322064594024,
+      "grad_norm": 0.11157894879579544,
+      "learning_rate": 0.00014611872146118722,
+      "loss": 0.5814,
+      "step": 1448
+    },
+    {
+      "epoch": 0.8747358889224268,
+      "grad_norm": 0.11887330561876297,
+      "learning_rate": 0.00014607721046077212,
+      "loss": 0.5033,
+      "step": 1449
+    },
+    {
+      "epoch": 0.8753395713854513,
+      "grad_norm": 0.1345004439353943,
+      "learning_rate": 0.000146035699460357,
+      "loss": 0.4091,
+      "step": 1450
+    },
+    {
+      "epoch": 0.8759432538484757,
+      "grad_norm": 0.1368480920791626,
+      "learning_rate": 0.0001459941884599419,
+      "loss": 0.7745,
+      "step": 1451
+    },
+    {
+      "epoch": 0.8765469363115002,
+      "grad_norm": 0.09147506207227707,
+      "learning_rate": 0.0001459526774595268,
+      "loss": 1.1362,
+      "step": 1452
+    },
+    {
+      "epoch": 0.8771506187745246,
+      "grad_norm": 0.07865995168685913,
+      "learning_rate": 0.00014591116645911166,
+      "loss": 0.7771,
+      "step": 1453
+    },
+    {
+      "epoch": 0.877754301237549,
+      "grad_norm": 0.0899992287158966,
+      "learning_rate": 0.00014586965545869656,
+      "loss": 0.779,
+      "step": 1454
+    },
+    {
+      "epoch": 0.8783579837005735,
+      "grad_norm": 0.0916723981499672,
+      "learning_rate": 0.00014582814445828146,
+      "loss": 1.2033,
+      "step": 1455
+    },
+    {
+      "epoch": 0.8789616661635979,
+      "grad_norm": 0.09812314063310623,
+      "learning_rate": 0.00014578663345786634,
+      "loss": 0.9668,
+      "step": 1456
+    },
+    {
+      "epoch": 0.8795653486266224,
+      "grad_norm": 0.09160758554935455,
+      "learning_rate": 0.0001457451224574512,
+      "loss": 0.8744,
+      "step": 1457
+    },
+    {
+      "epoch": 0.8801690310896468,
+      "grad_norm": 0.0805237665772438,
+      "learning_rate": 0.00014570361145703614,
+      "loss": 0.9906,
+      "step": 1458
+    },
+    {
+      "epoch": 0.8807727135526713,
+      "grad_norm": 0.09051381051540375,
+      "learning_rate": 0.000145662100456621,
+      "loss": 1.0932,
+      "step": 1459
+    },
+    {
+      "epoch": 0.8813763960156957,
+      "grad_norm": 0.08794981986284256,
+      "learning_rate": 0.00014562058945620588,
+      "loss": 0.7806,
+      "step": 1460
+    },
+    {
+      "epoch": 0.8819800784787202,
+      "grad_norm": 0.0870475247502327,
+      "learning_rate": 0.00014557907845579078,
+      "loss": 1.0775,
+      "step": 1461
+    },
+    {
+      "epoch": 0.8825837609417446,
+      "grad_norm": 0.0890730544924736,
+      "learning_rate": 0.00014553756745537568,
+      "loss": 0.7489,
+      "step": 1462
+    },
+    {
+      "epoch": 0.8831874434047691,
+      "grad_norm": 0.08005908131599426,
+      "learning_rate": 0.00014549605645496056,
+      "loss": 1.0614,
+      "step": 1463
+    },
+    {
+      "epoch": 0.8837911258677935,
+      "grad_norm": 0.10012607276439667,
+      "learning_rate": 0.00014545454545454546,
+      "loss": 0.8,
+      "step": 1464
+    },
+    {
+      "epoch": 0.884394808330818,
+      "grad_norm": 0.09003414958715439,
+      "learning_rate": 0.00014541303445413036,
+      "loss": 0.7296,
+      "step": 1465
+    },
+    {
+      "epoch": 0.8849984907938424,
+      "grad_norm": 0.09428632259368896,
+      "learning_rate": 0.00014537152345371523,
+      "loss": 0.7256,
+      "step": 1466
+    },
+    {
+      "epoch": 0.8856021732568669,
+      "grad_norm": 0.08243349939584732,
+      "learning_rate": 0.00014533001245330013,
+      "loss": 0.7458,
+      "step": 1467
+    },
+    {
+      "epoch": 0.8862058557198913,
+      "grad_norm": 0.07649943977594376,
+      "learning_rate": 0.00014528850145288503,
+      "loss": 0.7372,
+      "step": 1468
+    },
+    {
+      "epoch": 0.8868095381829157,
+      "grad_norm": 0.08396594226360321,
+      "learning_rate": 0.00014524699045246993,
+      "loss": 0.8017,
+      "step": 1469
+    },
+    {
+      "epoch": 0.8874132206459402,
+      "grad_norm": 0.10132227838039398,
+      "learning_rate": 0.0001452054794520548,
+      "loss": 1.2413,
+      "step": 1470
+    },
+    {
+      "epoch": 0.8880169031089646,
+      "grad_norm": 0.0780363529920578,
+      "learning_rate": 0.00014516396845163968,
+      "loss": 0.8076,
+      "step": 1471
+    },
+    {
+      "epoch": 0.8886205855719891,
+      "grad_norm": 0.09692296385765076,
+      "learning_rate": 0.0001451224574512246,
+      "loss": 0.9112,
+      "step": 1472
+    },
+    {
+      "epoch": 0.8892242680350135,
+      "grad_norm": 0.08603715151548386,
+      "learning_rate": 0.00014508094645080948,
+      "loss": 0.8524,
+      "step": 1473
+    },
+    {
+      "epoch": 0.889827950498038,
+      "grad_norm": 0.0896432101726532,
+      "learning_rate": 0.00014503943545039435,
+      "loss": 1.0479,
+      "step": 1474
+    },
+    {
+      "epoch": 0.8904316329610625,
+      "grad_norm": 0.0839807540178299,
+      "learning_rate": 0.00014499792444997925,
+      "loss": 0.794,
+      "step": 1475
+    },
+    {
+      "epoch": 0.891035315424087,
+      "grad_norm": 0.08278023451566696,
+      "learning_rate": 0.00014495641344956415,
+      "loss": 0.8375,
+      "step": 1476
+    },
+    {
+      "epoch": 0.8916389978871114,
+      "grad_norm": 0.07943416386842728,
+      "learning_rate": 0.00014491490244914902,
+      "loss": 1.0645,
+      "step": 1477
+    },
+    {
+      "epoch": 0.8922426803501359,
+      "grad_norm": 0.08287323266267776,
+      "learning_rate": 0.00014487339144873392,
+      "loss": 0.7824,
+      "step": 1478
+    },
+    {
+      "epoch": 0.8928463628131603,
+      "grad_norm": 0.09486839175224304,
+      "learning_rate": 0.00014483188044831882,
+      "loss": 0.9189,
+      "step": 1479
+    },
+    {
+      "epoch": 0.8934500452761848,
+      "grad_norm": 0.09007449448108673,
+      "learning_rate": 0.0001447903694479037,
+      "loss": 0.7276,
+      "step": 1480
+    },
+    {
+      "epoch": 0.8940537277392092,
+      "grad_norm": 0.08944438397884369,
+      "learning_rate": 0.0001447488584474886,
+      "loss": 0.8821,
+      "step": 1481
+    },
+    {
+      "epoch": 0.8946574102022337,
+      "grad_norm": 0.08388552069664001,
+      "learning_rate": 0.0001447073474470735,
+      "loss": 0.7537,
+      "step": 1482
+    },
+    {
+      "epoch": 0.8952610926652581,
+      "grad_norm": 0.09146512299776077,
+      "learning_rate": 0.00014466583644665837,
+      "loss": 0.7041,
+      "step": 1483
+    },
+    {
+      "epoch": 0.8958647751282826,
+      "grad_norm": 0.09139760583639145,
+      "learning_rate": 0.00014462432544624327,
+      "loss": 0.7686,
+      "step": 1484
+    },
+    {
+      "epoch": 0.896468457591307,
+      "grad_norm": 0.08465917408466339,
+      "learning_rate": 0.00014458281444582814,
+      "loss": 0.7442,
+      "step": 1485
+    },
+    {
+      "epoch": 0.8970721400543314,
+      "grad_norm": 0.08978710323572159,
+      "learning_rate": 0.00014454130344541304,
+      "loss": 0.7162,
+      "step": 1486
+    },
+    {
+      "epoch": 0.8976758225173559,
+      "grad_norm": 0.09368009865283966,
+      "learning_rate": 0.00014449979244499794,
+      "loss": 0.6723,
+      "step": 1487
+    },
+    {
+      "epoch": 0.8982795049803803,
+      "grad_norm": 0.09958908706903458,
+      "learning_rate": 0.0001444582814445828,
+      "loss": 0.7469,
+      "step": 1488
+    },
+    {
+      "epoch": 0.8988831874434048,
+      "grad_norm": 0.09646685421466827,
+      "learning_rate": 0.0001444167704441677,
+      "loss": 0.7735,
+      "step": 1489
+    },
+    {
+      "epoch": 0.8994868699064292,
+      "grad_norm": 0.09356389194726944,
+      "learning_rate": 0.00014437525944375261,
+      "loss": 0.7194,
+      "step": 1490
+    },
+    {
+      "epoch": 0.9000905523694537,
+      "grad_norm": 0.09738507866859436,
+      "learning_rate": 0.0001443337484433375,
+      "loss": 0.6941,
+      "step": 1491
+    },
+    {
+      "epoch": 0.9006942348324781,
+      "grad_norm": 0.10621945559978485,
+      "learning_rate": 0.0001442922374429224,
+      "loss": 0.636,
+      "step": 1492
+    },
+    {
+      "epoch": 0.9012979172955026,
+      "grad_norm": 0.10333646088838577,
+      "learning_rate": 0.0001442507264425073,
+      "loss": 0.6574,
+      "step": 1493
+    },
+    {
+      "epoch": 0.901901599758527,
+      "grad_norm": 0.1097961962223053,
+      "learning_rate": 0.00014420921544209216,
+      "loss": 0.7928,
+      "step": 1494
+    },
+    {
+      "epoch": 0.9025052822215515,
+      "grad_norm": 0.11894567310810089,
+      "learning_rate": 0.00014416770444167703,
+      "loss": 0.6904,
+      "step": 1495
+    },
+    {
+      "epoch": 0.9031089646845759,
+      "grad_norm": 0.12034012377262115,
+      "learning_rate": 0.00014412619344126196,
+      "loss": 0.6091,
+      "step": 1496
+    },
+    {
+      "epoch": 0.9037126471476004,
+      "grad_norm": 0.11718066781759262,
+      "learning_rate": 0.00014408468244084683,
+      "loss": 0.6336,
+      "step": 1497
+    },
+    {
+      "epoch": 0.9043163296106248,
+      "grad_norm": 0.130398690700531,
+      "learning_rate": 0.0001440431714404317,
+      "loss": 0.5771,
+      "step": 1498
+    },
+    {
+      "epoch": 0.9049200120736492,
+      "grad_norm": 0.12476742267608643,
+      "learning_rate": 0.0001440016604400166,
+      "loss": 0.5312,
+      "step": 1499
+    },
+    {
+      "epoch": 0.9055236945366737,
+      "grad_norm": 0.12364601343870163,
+      "learning_rate": 0.0001439601494396015,
+      "loss": 0.3942,
+      "step": 1500
+    },
+    {
+      "epoch": 0.9055236945366737,
+      "eval_loss": 0.8090887069702148,
+      "eval_runtime": 1219.063,
+      "eval_samples_per_second": 2.289,
+      "eval_steps_per_second": 0.286,
+      "step": 1500
+    },
+    {
+      "epoch": 0.9061273769996981,
+      "grad_norm": 0.0810943990945816,
+      "learning_rate": 0.00014391863843918638,
+      "loss": 1.0772,
+      "step": 1501
+    },
+    {
+      "epoch": 0.9067310594627226,
+      "grad_norm": 0.08426682651042938,
+      "learning_rate": 0.00014387712743877128,
+      "loss": 0.7147,
+      "step": 1502
+    },
+    {
+      "epoch": 0.907334741925747,
+      "grad_norm": 0.09331586956977844,
+      "learning_rate": 0.00014383561643835618,
+      "loss": 0.8483,
+      "step": 1503
+    },
+    {
+      "epoch": 0.9079384243887715,
+      "grad_norm": 0.09169352054595947,
+      "learning_rate": 0.00014379410543794105,
+      "loss": 0.763,
+      "step": 1504
+    },
+    {
+      "epoch": 0.9085421068517959,
+      "grad_norm": 0.08313553780317307,
+      "learning_rate": 0.00014375259443752595,
+      "loss": 0.7842,
+      "step": 1505
+    },
+    {
+      "epoch": 0.9091457893148204,
+      "grad_norm": 0.08828964829444885,
+      "learning_rate": 0.00014371108343711085,
+      "loss": 0.9434,
+      "step": 1506
+    },
+    {
+      "epoch": 0.9097494717778448,
+      "grad_norm": 0.08548730611801147,
+      "learning_rate": 0.00014366957243669572,
+      "loss": 0.7584,
+      "step": 1507
+    },
+    {
+      "epoch": 0.9103531542408693,
+      "grad_norm": 0.09586931765079498,
+      "learning_rate": 0.00014362806143628062,
+      "loss": 1.016,
+      "step": 1508
+    },
+    {
+      "epoch": 0.9109568367038937,
+      "grad_norm": 0.08847955614328384,
+      "learning_rate": 0.0001435865504358655,
+      "loss": 0.886,
+      "step": 1509
+    },
+    {
+      "epoch": 0.9115605191669182,
+      "grad_norm": 0.14398765563964844,
+      "learning_rate": 0.0001435450394354504,
+      "loss": 0.7924,
+      "step": 1510
+    },
+    {
+      "epoch": 0.9121642016299426,
+      "grad_norm": 0.08921834826469421,
+      "learning_rate": 0.0001435035284350353,
+      "loss": 0.8978,
+      "step": 1511
+    },
+    {
+      "epoch": 0.912767884092967,
+      "grad_norm": 0.07638692855834961,
+      "learning_rate": 0.00014346201743462017,
+      "loss": 0.6995,
+      "step": 1512
+    },
+    {
+      "epoch": 0.9133715665559915,
+      "grad_norm": 0.08616790175437927,
+      "learning_rate": 0.00014342050643420507,
+      "loss": 0.9325,
+      "step": 1513
+    },
+    {
+      "epoch": 0.913975249019016,
+      "grad_norm": 0.08327416330575943,
+      "learning_rate": 0.00014337899543378997,
+      "loss": 1.0005,
+      "step": 1514
+    },
+    {
+      "epoch": 0.9145789314820404,
+      "grad_norm": 0.07625047117471695,
+      "learning_rate": 0.00014333748443337484,
+      "loss": 0.9611,
+      "step": 1515
+    },
+    {
+      "epoch": 0.915182613945065,
+      "grad_norm": 0.08822344988584518,
+      "learning_rate": 0.00014329597343295974,
+      "loss": 0.8797,
+      "step": 1516
+    },
+    {
+      "epoch": 0.9157862964080894,
+      "grad_norm": 0.08142781257629395,
+      "learning_rate": 0.00014325446243254464,
+      "loss": 0.8249,
+      "step": 1517
+    },
+    {
+      "epoch": 0.9163899788711138,
+      "grad_norm": 0.07951053231954575,
+      "learning_rate": 0.00014321295143212952,
+      "loss": 0.8624,
+      "step": 1518
+    },
+    {
+      "epoch": 0.9169936613341383,
+      "grad_norm": 0.08235177397727966,
+      "learning_rate": 0.00014317144043171442,
+      "loss": 0.8208,
+      "step": 1519
+    },
+    {
+      "epoch": 0.9175973437971627,
+      "grad_norm": 0.09271861612796783,
+      "learning_rate": 0.00014312992943129932,
+      "loss": 0.789,
+      "step": 1520
+    },
+    {
+      "epoch": 0.9182010262601872,
+      "grad_norm": 0.09429887682199478,
+      "learning_rate": 0.0001430884184308842,
+      "loss": 1.1317,
+      "step": 1521
+    },
+    {
+      "epoch": 0.9188047087232116,
+      "grad_norm": 0.08084212988615036,
+      "learning_rate": 0.0001430469074304691,
+      "loss": 0.7235,
+      "step": 1522
+    },
+    {
+      "epoch": 0.9194083911862361,
+      "grad_norm": 0.08684766292572021,
+      "learning_rate": 0.00014300539643005396,
+      "loss": 0.9076,
+      "step": 1523
+    },
+    {
+      "epoch": 0.9200120736492605,
+      "grad_norm": 0.07625840604305267,
+      "learning_rate": 0.00014296388542963886,
+      "loss": 0.743,
+      "step": 1524
+    },
+    {
+      "epoch": 0.920615756112285,
+      "grad_norm": 0.0968519002199173,
+      "learning_rate": 0.00014292237442922376,
+      "loss": 0.8984,
+      "step": 1525
+    },
+    {
+      "epoch": 0.9212194385753094,
+      "grad_norm": 0.08472032099962234,
+      "learning_rate": 0.00014288086342880864,
+      "loss": 1.0295,
+      "step": 1526
+    },
+    {
+      "epoch": 0.9218231210383339,
+      "grad_norm": 0.0939970538020134,
+      "learning_rate": 0.00014283935242839354,
+      "loss": 0.9342,
+      "step": 1527
+    },
+    {
+      "epoch": 0.9224268035013583,
+      "grad_norm": 0.14263185858726501,
+      "learning_rate": 0.00014279784142797844,
+      "loss": 0.8136,
+      "step": 1528
+    },
+    {
+      "epoch": 0.9230304859643828,
+      "grad_norm": 0.08136036992073059,
+      "learning_rate": 0.0001427563304275633,
+      "loss": 0.8533,
+      "step": 1529
+    },
+    {
+      "epoch": 0.9236341684274072,
+      "grad_norm": 0.0846790298819542,
+      "learning_rate": 0.00014271481942714818,
+      "loss": 0.857,
+      "step": 1530
+    },
+    {
+      "epoch": 0.9242378508904316,
+      "grad_norm": 0.08513128757476807,
+      "learning_rate": 0.0001426733084267331,
+      "loss": 0.8139,
+      "step": 1531
+    },
+    {
+      "epoch": 0.9248415333534561,
+      "grad_norm": 0.07980025559663773,
+      "learning_rate": 0.00014263179742631798,
+      "loss": 0.745,
+      "step": 1532
+    },
+    {
+      "epoch": 0.9254452158164805,
+      "grad_norm": 0.0855269581079483,
+      "learning_rate": 0.00014259028642590285,
+      "loss": 0.7256,
+      "step": 1533
+    },
+    {
+      "epoch": 0.926048898279505,
+      "grad_norm": 0.08665366470813751,
+      "learning_rate": 0.00014254877542548778,
+      "loss": 0.7378,
+      "step": 1534
+    },
+    {
+      "epoch": 0.9266525807425294,
+      "grad_norm": 0.098544642329216,
+      "learning_rate": 0.00014250726442507265,
+      "loss": 0.7067,
+      "step": 1535
+    },
+    {
+      "epoch": 0.9272562632055539,
+      "grad_norm": 0.08651740849018097,
+      "learning_rate": 0.00014246575342465753,
+      "loss": 0.7731,
+      "step": 1536
+    },
+    {
+      "epoch": 0.9278599456685783,
+      "grad_norm": 0.09180111438035965,
+      "learning_rate": 0.00014242424242424243,
+      "loss": 0.7735,
+      "step": 1537
+    },
+    {
+      "epoch": 0.9284636281316028,
+      "grad_norm": 0.09831391274929047,
+      "learning_rate": 0.00014238273142382733,
+      "loss": 0.7889,
+      "step": 1538
+    },
+    {
+      "epoch": 0.9290673105946272,
+      "grad_norm": 0.09092969447374344,
+      "learning_rate": 0.0001423412204234122,
+      "loss": 0.8703,
+      "step": 1539
+    },
+    {
+      "epoch": 0.9296709930576517,
+      "grad_norm": 0.08774839341640472,
+      "learning_rate": 0.0001422997094229971,
+      "loss": 0.6438,
+      "step": 1540
+    },
+    {
+      "epoch": 0.9302746755206761,
+      "grad_norm": 0.10328055173158646,
+      "learning_rate": 0.000142258198422582,
+      "loss": 0.7747,
+      "step": 1541
+    },
+    {
+      "epoch": 0.9308783579837006,
+      "grad_norm": 0.0955677479505539,
+      "learning_rate": 0.00014221668742216687,
+      "loss": 0.7044,
+      "step": 1542
+    },
+    {
+      "epoch": 0.931482040446725,
+      "grad_norm": 0.12321915477514267,
+      "learning_rate": 0.00014217517642175177,
+      "loss": 0.7503,
+      "step": 1543
+    },
+    {
+      "epoch": 0.9320857229097494,
+      "grad_norm": 0.10311946272850037,
+      "learning_rate": 0.00014213366542133665,
+      "loss": 0.7091,
+      "step": 1544
+    },
+    {
+      "epoch": 0.9326894053727739,
+      "grad_norm": 0.10908188670873642,
+      "learning_rate": 0.00014209215442092155,
+      "loss": 0.6301,
+      "step": 1545
+    },
+    {
+      "epoch": 0.9332930878357983,
+      "grad_norm": 0.10754991322755814,
+      "learning_rate": 0.00014205064342050645,
+      "loss": 0.6588,
+      "step": 1546
+    },
+    {
+      "epoch": 0.9338967702988228,
+      "grad_norm": 0.11673349887132645,
+      "learning_rate": 0.00014200913242009132,
+      "loss": 0.6268,
+      "step": 1547
+    },
+    {
+      "epoch": 0.9345004527618472,
+      "grad_norm": 0.12704375386238098,
+      "learning_rate": 0.00014196762141967622,
+      "loss": 0.6035,
+      "step": 1548
+    },
+    {
+      "epoch": 0.9351041352248717,
+      "grad_norm": 0.1293480098247528,
+      "learning_rate": 0.00014192611041926112,
+      "loss": 0.5336,
+      "step": 1549
+    },
+    {
+      "epoch": 0.9357078176878961,
+      "grad_norm": 0.13309147953987122,
+      "learning_rate": 0.000141884599418846,
+      "loss": 0.4451,
+      "step": 1550
+    },
+    {
+      "epoch": 0.9363115001509206,
+      "grad_norm": 0.1289544254541397,
+      "learning_rate": 0.0001418430884184309,
+      "loss": 0.818,
+      "step": 1551
+    },
+    {
+      "epoch": 0.936915182613945,
+      "grad_norm": 0.10981032997369766,
+      "learning_rate": 0.0001418015774180158,
+      "loss": 0.8971,
+      "step": 1552
+    },
+    {
+      "epoch": 0.9375188650769695,
+      "grad_norm": 0.07926082611083984,
+      "learning_rate": 0.00014176006641760067,
+      "loss": 0.7756,
+      "step": 1553
+    },
+    {
+      "epoch": 0.9381225475399939,
+      "grad_norm": 0.08216461539268494,
+      "learning_rate": 0.00014171855541718554,
+      "loss": 0.7566,
+      "step": 1554
+    },
+    {
+      "epoch": 0.9387262300030184,
+      "grad_norm": 0.08312779664993286,
+      "learning_rate": 0.00014167704441677047,
+      "loss": 0.7627,
+      "step": 1555
+    },
+    {
+      "epoch": 0.9393299124660429,
+      "grad_norm": 0.08494763821363449,
+      "learning_rate": 0.00014163553341635534,
+      "loss": 1.0068,
+      "step": 1556
+    },
+    {
+      "epoch": 0.9399335949290674,
+      "grad_norm": 0.10769182443618774,
+      "learning_rate": 0.00014159402241594024,
+      "loss": 0.8372,
+      "step": 1557
+    },
+    {
+      "epoch": 0.9405372773920918,
+      "grad_norm": 0.08497872948646545,
+      "learning_rate": 0.0001415525114155251,
+      "loss": 0.8532,
+      "step": 1558
+    },
+    {
+      "epoch": 0.9411409598551163,
+      "grad_norm": 0.07787720859050751,
+      "learning_rate": 0.00014151100041511,
+      "loss": 0.7097,
+      "step": 1559
+    },
+    {
+      "epoch": 0.9417446423181407,
+      "grad_norm": 0.10002875328063965,
+      "learning_rate": 0.0001414694894146949,
+      "loss": 0.8718,
+      "step": 1560
+    },
+    {
+      "epoch": 0.9423483247811651,
+      "grad_norm": 0.09802395850419998,
+      "learning_rate": 0.00014142797841427978,
+      "loss": 0.8319,
+      "step": 1561
+    },
+    {
+      "epoch": 0.9429520072441896,
+      "grad_norm": 0.08751332759857178,
+      "learning_rate": 0.00014138646741386468,
+      "loss": 0.7035,
+      "step": 1562
+    },
+    {
+      "epoch": 0.943555689707214,
+      "grad_norm": 0.08736349642276764,
+      "learning_rate": 0.00014134495641344958,
+      "loss": 0.8215,
+      "step": 1563
+    },
+    {
+      "epoch": 0.9441593721702385,
+      "grad_norm": 0.09416454285383224,
+      "learning_rate": 0.00014130344541303446,
+      "loss": 0.8634,
+      "step": 1564
+    },
+    {
+      "epoch": 0.9447630546332629,
+      "grad_norm": 0.0884031280875206,
+      "learning_rate": 0.00014126193441261936,
+      "loss": 0.848,
+      "step": 1565
+    },
+    {
+      "epoch": 0.9453667370962874,
+      "grad_norm": 0.08755529671907425,
+      "learning_rate": 0.00014122042341220426,
+      "loss": 0.8167,
+      "step": 1566
+    },
+    {
+      "epoch": 0.9459704195593118,
+      "grad_norm": 0.10254927724599838,
+      "learning_rate": 0.00014117891241178913,
+      "loss": 0.8119,
+      "step": 1567
+    },
+    {
+      "epoch": 0.9465741020223363,
+      "grad_norm": 0.0907716229557991,
+      "learning_rate": 0.000141137401411374,
+      "loss": 1.0695,
+      "step": 1568
+    },
+    {
+      "epoch": 0.9471777844853607,
+      "grad_norm": 0.08859268575906754,
+      "learning_rate": 0.00014109589041095893,
+      "loss": 0.8692,
+      "step": 1569
+    },
+    {
+      "epoch": 0.9477814669483852,
+      "grad_norm": 0.0878458023071289,
+      "learning_rate": 0.0001410543794105438,
+      "loss": 0.758,
+      "step": 1570
+    },
+    {
+      "epoch": 0.9483851494114096,
+      "grad_norm": 0.10177980363368988,
+      "learning_rate": 0.00014101286841012868,
+      "loss": 0.7831,
+      "step": 1571
+    },
+    {
+      "epoch": 0.9489888318744341,
+      "grad_norm": 0.08606535941362381,
+      "learning_rate": 0.00014097135740971358,
+      "loss": 1.0743,
+      "step": 1572
+    },
+    {
+      "epoch": 0.9495925143374585,
+      "grad_norm": 0.09055022895336151,
+      "learning_rate": 0.00014092984640929848,
+      "loss": 0.9108,
+      "step": 1573
+    },
+    {
+      "epoch": 0.950196196800483,
+      "grad_norm": 0.08120472729206085,
+      "learning_rate": 0.00014088833540888335,
+      "loss": 0.7569,
+      "step": 1574
+    },
+    {
+      "epoch": 0.9507998792635074,
+      "grad_norm": 0.08704983443021774,
+      "learning_rate": 0.00014084682440846825,
+      "loss": 0.8356,
+      "step": 1575
+    },
+    {
+      "epoch": 0.9514035617265318,
+      "grad_norm": 0.08406449109315872,
+      "learning_rate": 0.00014080531340805315,
+      "loss": 0.7055,
+      "step": 1576
+    },
+    {
+      "epoch": 0.9520072441895563,
+      "grad_norm": 0.08247263729572296,
+      "learning_rate": 0.00014076380240763802,
+      "loss": 0.8572,
+      "step": 1577
+    },
+    {
+      "epoch": 0.9526109266525807,
+      "grad_norm": 0.13809391856193542,
+      "learning_rate": 0.00014072229140722292,
+      "loss": 0.7313,
+      "step": 1578
+    },
+    {
+      "epoch": 0.9532146091156052,
+      "grad_norm": 0.09925238788127899,
+      "learning_rate": 0.00014068078040680782,
+      "loss": 0.7794,
+      "step": 1579
+    },
+    {
+      "epoch": 0.9538182915786296,
+      "grad_norm": 0.08268136531114578,
+      "learning_rate": 0.0001406392694063927,
+      "loss": 0.7268,
+      "step": 1580
+    },
+    {
+      "epoch": 0.9544219740416541,
+      "grad_norm": 0.08466971665620804,
+      "learning_rate": 0.0001405977584059776,
+      "loss": 0.8175,
+      "step": 1581
+    },
+    {
+      "epoch": 0.9550256565046785,
+      "grad_norm": 0.08671456575393677,
+      "learning_rate": 0.00014055624740556247,
+      "loss": 0.7747,
+      "step": 1582
+    },
+    {
+      "epoch": 0.955629338967703,
+      "grad_norm": 0.2081962525844574,
+      "learning_rate": 0.00014051473640514737,
+      "loss": 0.8782,
+      "step": 1583
+    },
+    {
+      "epoch": 0.9562330214307274,
+      "grad_norm": 0.11799836158752441,
+      "learning_rate": 0.00014047322540473227,
+      "loss": 0.8098,
+      "step": 1584
+    },
+    {
+      "epoch": 0.9568367038937519,
+      "grad_norm": 0.08406732231378555,
+      "learning_rate": 0.00014043171440431714,
+      "loss": 0.7716,
+      "step": 1585
+    },
+    {
+      "epoch": 0.9574403863567763,
+      "grad_norm": 0.08422478288412094,
+      "learning_rate": 0.00014039020340390204,
+      "loss": 0.7237,
+      "step": 1586
+    },
+    {
+      "epoch": 0.9580440688198008,
+      "grad_norm": 0.09907133132219315,
+      "learning_rate": 0.00014034869240348694,
+      "loss": 0.7298,
+      "step": 1587
+    },
+    {
+      "epoch": 0.9586477512828252,
+      "grad_norm": 0.08633650094270706,
+      "learning_rate": 0.00014030718140307181,
+      "loss": 0.7262,
+      "step": 1588
+    },
+    {
+      "epoch": 0.9592514337458496,
+      "grad_norm": 0.0880611315369606,
+      "learning_rate": 0.00014026567040265671,
+      "loss": 0.6691,
+      "step": 1589
+    },
+    {
+      "epoch": 0.9598551162088741,
+      "grad_norm": 0.0872715413570404,
+      "learning_rate": 0.00014022415940224161,
+      "loss": 0.6593,
+      "step": 1590
+    },
+    {
+      "epoch": 0.9604587986718985,
+      "grad_norm": 0.09152337163686752,
+      "learning_rate": 0.0001401826484018265,
+      "loss": 0.7065,
+      "step": 1591
+    },
+    {
+      "epoch": 0.961062481134923,
+      "grad_norm": 0.1341097503900528,
+      "learning_rate": 0.00014014113740141136,
+      "loss": 0.723,
+      "step": 1592
+    },
+    {
+      "epoch": 0.9616661635979474,
+      "grad_norm": 0.09733890742063522,
+      "learning_rate": 0.0001400996264009963,
+      "loss": 0.7194,
+      "step": 1593
+    },
+    {
+      "epoch": 0.9622698460609719,
+      "grad_norm": 0.10491717606782913,
+      "learning_rate": 0.00014005811540058116,
+      "loss": 0.6986,
+      "step": 1594
+    },
+    {
+      "epoch": 0.9628735285239963,
+      "grad_norm": 0.10709907114505768,
+      "learning_rate": 0.00014001660440016603,
+      "loss": 0.7316,
+      "step": 1595
+    },
+    {
+      "epoch": 0.9634772109870209,
+      "grad_norm": 0.1096893846988678,
+      "learning_rate": 0.00013997509339975093,
+      "loss": 0.6352,
+      "step": 1596
+    },
+    {
+      "epoch": 0.9640808934500453,
+      "grad_norm": 0.11475121229887009,
+      "learning_rate": 0.00013993358239933583,
+      "loss": 0.6632,
+      "step": 1597
+    },
+    {
+      "epoch": 0.9646845759130698,
+      "grad_norm": 0.12020900100469589,
+      "learning_rate": 0.0001398920713989207,
+      "loss": 0.6824,
+      "step": 1598
+    },
+    {
+      "epoch": 0.9652882583760942,
+      "grad_norm": 0.12162651866674423,
+      "learning_rate": 0.0001398505603985056,
+      "loss": 0.5713,
+      "step": 1599
+    },
+    {
+      "epoch": 0.9658919408391187,
+      "grad_norm": 0.13266035914421082,
+      "learning_rate": 0.0001398090493980905,
+      "loss": 0.4436,
+      "step": 1600
+    },
+    {
+      "epoch": 0.9664956233021431,
+      "grad_norm": 0.0847954973578453,
+      "learning_rate": 0.0001397675383976754,
+      "loss": 0.8099,
+      "step": 1601
+    },
+    {
+      "epoch": 0.9670993057651676,
+      "grad_norm": 0.08815553784370422,
+      "learning_rate": 0.00013972602739726028,
+      "loss": 0.8847,
+      "step": 1602
+    },
+    {
+      "epoch": 0.967702988228192,
+      "grad_norm": 0.0948781669139862,
+      "learning_rate": 0.00013968451639684518,
+      "loss": 1.0048,
+      "step": 1603
+    },
+    {
+      "epoch": 0.9683066706912165,
+      "grad_norm": 0.08522479981184006,
+      "learning_rate": 0.00013964300539643008,
+      "loss": 1.0174,
+      "step": 1604
+    },
+    {
+      "epoch": 0.9689103531542409,
+      "grad_norm": 0.07847335189580917,
+      "learning_rate": 0.00013960149439601495,
+      "loss": 0.7368,
+      "step": 1605
+    },
+    {
+      "epoch": 0.9695140356172653,
+      "grad_norm": 0.10585256665945053,
+      "learning_rate": 0.00013955998339559983,
+      "loss": 0.9218,
+      "step": 1606
+    },
+    {
+      "epoch": 0.9701177180802898,
+      "grad_norm": 0.08924368768930435,
+      "learning_rate": 0.00013951847239518475,
+      "loss": 0.8461,
+      "step": 1607
+    },
+    {
+      "epoch": 0.9707214005433142,
+      "grad_norm": 0.07704899460077286,
+      "learning_rate": 0.00013947696139476963,
+      "loss": 0.7675,
+      "step": 1608
+    },
+    {
+      "epoch": 0.9713250830063387,
+      "grad_norm": 0.08050254732370377,
+      "learning_rate": 0.0001394354503943545,
+      "loss": 0.845,
+      "step": 1609
+    },
+    {
+      "epoch": 0.9719287654693631,
+      "grad_norm": 0.08915068209171295,
+      "learning_rate": 0.0001393939393939394,
+      "loss": 0.8523,
+      "step": 1610
+    },
+    {
+      "epoch": 0.9725324479323876,
+      "grad_norm": 0.07807064801454544,
+      "learning_rate": 0.0001393524283935243,
+      "loss": 0.6999,
+      "step": 1611
+    },
+    {
+      "epoch": 0.973136130395412,
+      "grad_norm": 0.1043185442686081,
+      "learning_rate": 0.00013931091739310917,
+      "loss": 0.8859,
+      "step": 1612
+    },
+    {
+      "epoch": 0.9737398128584365,
+      "grad_norm": 0.1474182903766632,
+      "learning_rate": 0.00013926940639269407,
+      "loss": 1.1711,
+      "step": 1613
+    },
+    {
+      "epoch": 0.9743434953214609,
+      "grad_norm": 0.08791965246200562,
+      "learning_rate": 0.00013922789539227897,
+      "loss": 0.9619,
+      "step": 1614
+    },
+    {
+      "epoch": 0.9749471777844854,
+      "grad_norm": 0.08670973777770996,
+      "learning_rate": 0.00013918638439186384,
+      "loss": 1.1479,
+      "step": 1615
+    },
+    {
+      "epoch": 0.9755508602475098,
+      "grad_norm": 0.09029529243707657,
+      "learning_rate": 0.00013914487339144874,
+      "loss": 0.7685,
+      "step": 1616
+    },
+    {
+      "epoch": 0.9761545427105343,
+      "grad_norm": 0.15448486804962158,
+      "learning_rate": 0.00013910336239103364,
+      "loss": 0.9595,
+      "step": 1617
+    },
+    {
+      "epoch": 0.9767582251735587,
+      "grad_norm": 0.08685984462499619,
+      "learning_rate": 0.00013906185139061852,
+      "loss": 0.7231,
+      "step": 1618
+    },
+    {
+      "epoch": 0.9773619076365831,
+      "grad_norm": 0.09811729937791824,
+      "learning_rate": 0.00013902034039020342,
+      "loss": 0.9077,
+      "step": 1619
+    },
+    {
+      "epoch": 0.9779655900996076,
+      "grad_norm": 0.10184766352176666,
+      "learning_rate": 0.0001389788293897883,
+      "loss": 0.8866,
+      "step": 1620
+    },
+    {
+      "epoch": 0.978569272562632,
+      "grad_norm": 0.13190968334674835,
+      "learning_rate": 0.0001389373183893732,
+      "loss": 0.8538,
+      "step": 1621
+    },
+    {
+      "epoch": 0.9791729550256565,
+      "grad_norm": 0.0916040688753128,
+      "learning_rate": 0.0001388958073889581,
+      "loss": 0.8823,
+      "step": 1622
+    },
+    {
+      "epoch": 0.9797766374886809,
+      "grad_norm": 0.08365904539823532,
+      "learning_rate": 0.00013885429638854296,
+      "loss": 0.7825,
+      "step": 1623
+    },
+    {
+      "epoch": 0.9803803199517054,
+      "grad_norm": 0.09449176490306854,
+      "learning_rate": 0.00013881278538812786,
+      "loss": 0.9603,
+      "step": 1624
+    },
+    {
+      "epoch": 0.9809840024147298,
+      "grad_norm": 0.08852092921733856,
+      "learning_rate": 0.00013877127438771276,
+      "loss": 0.7159,
+      "step": 1625
+    },
+    {
+      "epoch": 0.9815876848777543,
+      "grad_norm": 0.10536504536867142,
+      "learning_rate": 0.00013872976338729764,
+      "loss": 1.0687,
+      "step": 1626
+    },
+    {
+      "epoch": 0.9821913673407787,
+      "grad_norm": 0.07926955074071884,
+      "learning_rate": 0.0001386882523868825,
+      "loss": 1.2306,
+      "step": 1627
+    },
+    {
+      "epoch": 0.9827950498038032,
+      "grad_norm": 0.08607269078493118,
+      "learning_rate": 0.00013864674138646744,
+      "loss": 0.8071,
+      "step": 1628
+    },
+    {
+      "epoch": 0.9833987322668276,
+      "grad_norm": 0.08337133377790451,
+      "learning_rate": 0.0001386052303860523,
+      "loss": 0.9241,
+      "step": 1629
+    },
+    {
+      "epoch": 0.9840024147298521,
+      "grad_norm": 0.0874747484922409,
+      "learning_rate": 0.00013856371938563718,
+      "loss": 0.7326,
+      "step": 1630
+    },
+    {
+      "epoch": 0.9846060971928765,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0001385222083852221,
+      "loss": 0.7567,
+      "step": 1631
+    },
+    {
+      "epoch": 0.985209779655901,
+      "grad_norm": 0.09537570923566818,
+      "learning_rate": 0.00013848069738480698,
+      "loss": 1.0592,
+      "step": 1632
+    },
+    {
+      "epoch": 0.9858134621189254,
+      "grad_norm": 0.0916895717382431,
+      "learning_rate": 0.00013843918638439186,
+      "loss": 0.7127,
+      "step": 1633
+    },
+    {
+      "epoch": 0.9864171445819498,
+      "grad_norm": 0.08885890990495682,
+      "learning_rate": 0.00013839767538397676,
+      "loss": 0.7207,
+      "step": 1634
+    },
+    {
+      "epoch": 0.9870208270449743,
+      "grad_norm": 0.08471481502056122,
+      "learning_rate": 0.00013835616438356166,
+      "loss": 0.7205,
+      "step": 1635
+    },
+    {
+      "epoch": 0.9876245095079987,
+      "grad_norm": 0.0972660556435585,
+      "learning_rate": 0.00013831465338314653,
+      "loss": 0.793,
+      "step": 1636
+    },
+    {
+      "epoch": 0.9882281919710233,
+      "grad_norm": 0.09207335114479065,
+      "learning_rate": 0.00013827314238273143,
+      "loss": 0.7134,
+      "step": 1637
+    },
+    {
+      "epoch": 0.9888318744340477,
+      "grad_norm": 0.0968717560172081,
+      "learning_rate": 0.00013823163138231633,
+      "loss": 0.6872,
+      "step": 1638
+    },
+    {
+      "epoch": 0.9894355568970722,
+      "grad_norm": 0.09645961970090866,
+      "learning_rate": 0.0001381901203819012,
+      "loss": 0.7423,
+      "step": 1639
+    },
+    {
+      "epoch": 0.9900392393600966,
+      "grad_norm": 0.09619560837745667,
+      "learning_rate": 0.0001381486093814861,
+      "loss": 0.7063,
+      "step": 1640
+    },
+    {
+      "epoch": 0.9906429218231211,
+      "grad_norm": 0.09428079426288605,
+      "learning_rate": 0.00013810709838107097,
+      "loss": 0.7305,
+      "step": 1641
+    },
+    {
+      "epoch": 0.9912466042861455,
+      "grad_norm": 0.10160063952207565,
+      "learning_rate": 0.00013806558738065587,
+      "loss": 0.7281,
+      "step": 1642
+    },
+    {
+      "epoch": 0.99185028674917,
+      "grad_norm": 0.10067980736494064,
+      "learning_rate": 0.00013802407638024077,
+      "loss": 0.7407,
+      "step": 1643
+    },
+    {
+      "epoch": 0.9924539692121944,
+      "grad_norm": 0.18993158638477325,
+      "learning_rate": 0.00013798256537982565,
+      "loss": 0.6791,
+      "step": 1644
+    },
+    {
+      "epoch": 0.9930576516752189,
+      "grad_norm": 0.10360643267631531,
+      "learning_rate": 0.00013794105437941057,
+      "loss": 0.7026,
+      "step": 1645
+    },
+    {
+      "epoch": 0.9936613341382433,
+      "grad_norm": 0.11464804410934448,
+      "learning_rate": 0.00013789954337899545,
+      "loss": 0.6614,
+      "step": 1646
+    },
+    {
+      "epoch": 0.9942650166012678,
+      "grad_norm": 0.11702211946249008,
+      "learning_rate": 0.00013785803237858032,
+      "loss": 0.6386,
+      "step": 1647
+    },
+    {
+      "epoch": 0.9948686990642922,
+      "grad_norm": 0.12298808991909027,
+      "learning_rate": 0.00013781652137816522,
+      "loss": 0.574,
+      "step": 1648
+    },
+    {
+      "epoch": 0.9954723815273167,
+      "grad_norm": 0.12374621629714966,
+      "learning_rate": 0.00013777501037775012,
+      "loss": 0.5813,
+      "step": 1649
+    },
+    {
+      "epoch": 0.9960760639903411,
+      "grad_norm": 0.12134044617414474,
+      "learning_rate": 0.000137733499377335,
+      "loss": 0.4442,
+      "step": 1650
+    },
+    {
+      "epoch": 0.9966797464533655,
+      "grad_norm": 0.08615315705537796,
+      "learning_rate": 0.0001376919883769199,
+      "loss": 0.839,
+      "step": 1651
+    },
+    {
+      "epoch": 0.99728342891639,
+      "grad_norm": 0.08225715905427933,
+      "learning_rate": 0.0001376504773765048,
+      "loss": 0.8809,
+      "step": 1652
+    },
+    {
+      "epoch": 0.9978871113794144,
+      "grad_norm": 0.0893816277384758,
+      "learning_rate": 0.00013760896637608967,
+      "loss": 0.8767,
+      "step": 1653
+    },
+    {
+      "epoch": 0.9984907938424389,
+      "grad_norm": 0.09268541634082794,
+      "learning_rate": 0.00013756745537567457,
+      "loss": 0.8083,
+      "step": 1654
+    },
+    {
+      "epoch": 0.9990944763054633,
+      "grad_norm": 0.10186032205820084,
+      "learning_rate": 0.00013752594437525944,
+      "loss": 0.783,
+      "step": 1655
+    },
+    {
+      "epoch": 0.9996981587684878,
+      "grad_norm": 0.11257860064506531,
+      "learning_rate": 0.00013748443337484434,
+      "loss": 0.6633,
+      "step": 1656
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1917642205953598,
+      "learning_rate": 0.00013744292237442924,
+      "loss": 0.476,
+      "step": 1657
+    },
+    {
+      "epoch": 1.0006036824630244,
+      "grad_norm": 0.09141898155212402,
+      "learning_rate": 0.0001374014113740141,
+      "loss": 0.8386,
+      "step": 1658
+    },
+    {
+      "epoch": 1.001207364926049,
+      "grad_norm": 0.09631163626909256,
+      "learning_rate": 0.000137359900373599,
+      "loss": 0.6534,
+      "step": 1659
+    },
+    {
+      "epoch": 1.0018110473890733,
+      "grad_norm": 0.09060285240411758,
+      "learning_rate": 0.0001373183893731839,
+      "loss": 0.9631,
+      "step": 1660
+    },
+    {
+      "epoch": 1.0024147298520978,
+      "grad_norm": 0.10698094218969345,
+      "learning_rate": 0.00013727687837276878,
+      "loss": 0.7317,
+      "step": 1661
+    },
+    {
+      "epoch": 1.0030184123151222,
+      "grad_norm": 0.08738933503627777,
+      "learning_rate": 0.00013723536737235369,
+      "loss": 0.689,
+      "step": 1662
+    },
+    {
+      "epoch": 1.0036220947781467,
+      "grad_norm": 0.09262979030609131,
+      "learning_rate": 0.00013719385637193859,
+      "loss": 0.7741,
+      "step": 1663
+    },
+    {
+      "epoch": 1.0042257772411711,
+      "grad_norm": 0.08418719470500946,
+      "learning_rate": 0.00013715234537152346,
+      "loss": 0.7623,
+      "step": 1664
+    },
+    {
+      "epoch": 1.0048294597041956,
+      "grad_norm": 0.09748540818691254,
+      "learning_rate": 0.00013711083437110833,
+      "loss": 0.8195,
+      "step": 1665
+    },
+    {
+      "epoch": 1.00543314216722,
+      "grad_norm": 0.08836773782968521,
+      "learning_rate": 0.00013706932337069326,
+      "loss": 0.6536,
+      "step": 1666
+    },
+    {
+      "epoch": 1.0060368246302445,
+      "grad_norm": 0.0891902819275856,
+      "learning_rate": 0.00013702781237027813,
+      "loss": 0.647,
+      "step": 1667
+    },
+    {
+      "epoch": 1.006640507093269,
+      "grad_norm": 0.08402203768491745,
+      "learning_rate": 0.000136986301369863,
+      "loss": 0.9145,
+      "step": 1668
+    },
+    {
+      "epoch": 1.0072441895562934,
+      "grad_norm": 0.08421420305967331,
+      "learning_rate": 0.0001369447903694479,
+      "loss": 0.6114,
+      "step": 1669
+    },
+    {
+      "epoch": 1.0078478720193178,
+      "grad_norm": 0.09319291263818741,
+      "learning_rate": 0.0001369032793690328,
+      "loss": 0.6923,
+      "step": 1670
+    },
+    {
+      "epoch": 1.0084515544823422,
+      "grad_norm": 0.0897219106554985,
+      "learning_rate": 0.00013686176836861768,
+      "loss": 0.7721,
+      "step": 1671
+    },
+    {
+      "epoch": 1.0090552369453667,
+      "grad_norm": 0.08546704798936844,
+      "learning_rate": 0.00013682025736820258,
+      "loss": 0.7558,
+      "step": 1672
+    },
+    {
+      "epoch": 1.0096589194083911,
+      "grad_norm": 0.10542399436235428,
+      "learning_rate": 0.00013677874636778748,
+      "loss": 0.7204,
+      "step": 1673
+    },
+    {
+      "epoch": 1.0102626018714156,
+      "grad_norm": 0.0866287350654602,
+      "learning_rate": 0.00013673723536737235,
+      "loss": 0.6816,
+      "step": 1674
+    },
+    {
+      "epoch": 1.01086628433444,
+      "grad_norm": 0.0957920253276825,
+      "learning_rate": 0.00013669572436695725,
+      "loss": 0.7554,
+      "step": 1675
+    },
+    {
+      "epoch": 1.0114699667974645,
+      "grad_norm": 0.07835828512907028,
+      "learning_rate": 0.00013665421336654215,
+      "loss": 0.8906,
+      "step": 1676
+    },
+    {
+      "epoch": 1.012073649260489,
+      "grad_norm": 0.09102047979831696,
+      "learning_rate": 0.00013661270236612702,
+      "loss": 0.707,
+      "step": 1677
+    },
+    {
+      "epoch": 1.0126773317235134,
+      "grad_norm": 0.10025649517774582,
+      "learning_rate": 0.00013657119136571192,
+      "loss": 1.0865,
+      "step": 1678
+    },
+    {
+      "epoch": 1.0132810141865378,
+      "grad_norm": 0.08916833996772766,
+      "learning_rate": 0.0001365296803652968,
+      "loss": 0.7058,
+      "step": 1679
+    },
+    {
+      "epoch": 1.0138846966495623,
+      "grad_norm": 0.0850253701210022,
+      "learning_rate": 0.0001364881693648817,
+      "loss": 0.9324,
+      "step": 1680
+    },
+    {
+      "epoch": 1.0144883791125867,
+      "grad_norm": 0.09609053283929825,
+      "learning_rate": 0.0001364466583644666,
+      "loss": 0.7521,
+      "step": 1681
+    },
+    {
+      "epoch": 1.0150920615756112,
+      "grad_norm": 0.0949166864156723,
+      "learning_rate": 0.00013640514736405147,
+      "loss": 0.7548,
+      "step": 1682
+    },
+    {
+      "epoch": 1.0156957440386356,
+      "grad_norm": 0.08850863575935364,
+      "learning_rate": 0.00013636363636363637,
+      "loss": 0.7067,
+      "step": 1683
+    },
+    {
+      "epoch": 1.01629942650166,
+      "grad_norm": 0.09187474101781845,
+      "learning_rate": 0.00013632212536322127,
+      "loss": 0.7852,
+      "step": 1684
+    },
+    {
+      "epoch": 1.0169031089646845,
+      "grad_norm": 0.09210950136184692,
+      "learning_rate": 0.00013628061436280614,
+      "loss": 0.8244,
+      "step": 1685
+    },
+    {
+      "epoch": 1.017506791427709,
+      "grad_norm": 0.08923452347517014,
+      "learning_rate": 0.00013623910336239104,
+      "loss": 0.9229,
+      "step": 1686
+    },
+    {
+      "epoch": 1.0181104738907334,
+      "grad_norm": 0.08911499381065369,
+      "learning_rate": 0.00013619759236197594,
+      "loss": 0.7053,
+      "step": 1687
+    },
+    {
+      "epoch": 1.0187141563537578,
+      "grad_norm": 0.09007196873426437,
+      "learning_rate": 0.00013615608136156081,
+      "loss": 0.8376,
+      "step": 1688
+    },
+    {
+      "epoch": 1.0193178388167823,
+      "grad_norm": 0.08651362359523773,
+      "learning_rate": 0.0001361145703611457,
+      "loss": 0.6759,
+      "step": 1689
+    },
+    {
+      "epoch": 1.0199215212798067,
+      "grad_norm": 0.08241482824087143,
+      "learning_rate": 0.00013607305936073061,
+      "loss": 0.6885,
+      "step": 1690
+    },
+    {
+      "epoch": 1.0205252037428312,
+      "grad_norm": 0.10755082219839096,
+      "learning_rate": 0.0001360315483603155,
+      "loss": 0.6961,
+      "step": 1691
+    },
+    {
+      "epoch": 1.0211288862058556,
+      "grad_norm": 0.09004219621419907,
+      "learning_rate": 0.0001359900373599004,
+      "loss": 0.6243,
+      "step": 1692
+    },
+    {
+      "epoch": 1.02173256866888,
+      "grad_norm": 0.0957816019654274,
+      "learning_rate": 0.00013594852635948526,
+      "loss": 0.6462,
+      "step": 1693
+    },
+    {
+      "epoch": 1.0223362511319045,
+      "grad_norm": 0.1017783135175705,
+      "learning_rate": 0.00013590701535907016,
+      "loss": 0.6332,
+      "step": 1694
+    },
+    {
+      "epoch": 1.022939933594929,
+      "grad_norm": 0.09232212603092194,
+      "learning_rate": 0.00013586550435865506,
+      "loss": 0.6461,
+      "step": 1695
+    },
+    {
+      "epoch": 1.0235436160579534,
+      "grad_norm": 0.09460336714982986,
+      "learning_rate": 0.00013582399335823993,
+      "loss": 0.6641,
+      "step": 1696
+    },
+    {
+      "epoch": 1.0241472985209779,
+      "grad_norm": 0.09913098812103271,
+      "learning_rate": 0.00013578248235782483,
+      "loss": 0.6559,
+      "step": 1697
+    },
+    {
+      "epoch": 1.0247509809840025,
+      "grad_norm": 0.10010208189487457,
+      "learning_rate": 0.00013574097135740973,
+      "loss": 0.5921,
+      "step": 1698
+    },
+    {
+      "epoch": 1.025354663447027,
+      "grad_norm": 0.10247639566659927,
+      "learning_rate": 0.0001356994603569946,
+      "loss": 0.6158,
+      "step": 1699
+    },
+    {
+      "epoch": 1.0259583459100514,
+      "grad_norm": 0.10905825346708298,
+      "learning_rate": 0.0001356579493565795,
+      "loss": 0.66,
+      "step": 1700
+    },
+    {
+      "epoch": 1.0265620283730759,
+      "grad_norm": 0.11569482833147049,
+      "learning_rate": 0.0001356164383561644,
+      "loss": 0.6431,
+      "step": 1701
+    },
+    {
+      "epoch": 1.0271657108361003,
+      "grad_norm": 0.14849194884300232,
+      "learning_rate": 0.00013557492735574928,
+      "loss": 0.6045,
+      "step": 1702
+    },
+    {
+      "epoch": 1.0277693932991248,
+      "grad_norm": 0.11826399713754654,
+      "learning_rate": 0.00013553341635533415,
+      "loss": 0.6559,
+      "step": 1703
+    },
+    {
+      "epoch": 1.0283730757621492,
+      "grad_norm": 0.1232730820775032,
+      "learning_rate": 0.00013549190535491908,
+      "loss": 0.5124,
+      "step": 1704
+    },
+    {
+      "epoch": 1.0289767582251736,
+      "grad_norm": 0.121260866522789,
+      "learning_rate": 0.00013545039435450395,
+      "loss": 0.4503,
+      "step": 1705
+    },
+    {
+      "epoch": 1.029580440688198,
+      "grad_norm": 0.1328078657388687,
+      "learning_rate": 0.00013540888335408883,
+      "loss": 0.4254,
+      "step": 1706
+    },
+    {
+      "epoch": 1.0301841231512225,
+      "grad_norm": 0.13635770976543427,
+      "learning_rate": 0.00013536737235367373,
+      "loss": 0.3007,
+      "step": 1707
+    },
+    {
+      "epoch": 1.030787805614247,
+      "grad_norm": 0.10616832226514816,
+      "learning_rate": 0.00013532586135325863,
+      "loss": 0.7256,
+      "step": 1708
+    },
+    {
+      "epoch": 1.0313914880772714,
+      "grad_norm": 0.10195182263851166,
+      "learning_rate": 0.0001352843503528435,
+      "loss": 0.7402,
+      "step": 1709
+    },
+    {
+      "epoch": 1.0319951705402959,
+      "grad_norm": 0.11944833397865295,
+      "learning_rate": 0.0001352428393524284,
+      "loss": 0.757,
+      "step": 1710
+    },
+    {
+      "epoch": 1.0325988530033203,
+      "grad_norm": 0.12061698734760284,
+      "learning_rate": 0.0001352013283520133,
+      "loss": 1.0738,
+      "step": 1711
+    },
+    {
+      "epoch": 1.0332025354663448,
+      "grad_norm": 0.10394060611724854,
+      "learning_rate": 0.00013515981735159817,
+      "loss": 0.7147,
+      "step": 1712
+    },
+    {
+      "epoch": 1.0338062179293692,
+      "grad_norm": 0.10730050504207611,
+      "learning_rate": 0.00013511830635118307,
+      "loss": 0.9586,
+      "step": 1713
+    },
+    {
+      "epoch": 1.0344099003923937,
+      "grad_norm": 0.09942850470542908,
+      "learning_rate": 0.00013507679535076797,
+      "loss": 1.0808,
+      "step": 1714
+    },
+    {
+      "epoch": 1.0350135828554181,
+      "grad_norm": 0.10033493489027023,
+      "learning_rate": 0.00013503528435035284,
+      "loss": 0.7759,
+      "step": 1715
+    },
+    {
+      "epoch": 1.0356172653184426,
+      "grad_norm": 0.08915812522172928,
+      "learning_rate": 0.00013499377334993774,
+      "loss": 0.7716,
+      "step": 1716
+    },
+    {
+      "epoch": 1.036220947781467,
+      "grad_norm": 0.09855514019727707,
+      "learning_rate": 0.00013495226234952262,
+      "loss": 0.8139,
+      "step": 1717
+    },
+    {
+      "epoch": 1.0368246302444915,
+      "grad_norm": 0.0998401790857315,
+      "learning_rate": 0.00013491075134910752,
+      "loss": 0.8308,
+      "step": 1718
+    },
+    {
+      "epoch": 1.037428312707516,
+      "grad_norm": 0.09024068713188171,
+      "learning_rate": 0.00013486924034869242,
+      "loss": 0.7503,
+      "step": 1719
+    },
+    {
+      "epoch": 1.0380319951705403,
+      "grad_norm": 0.09240993112325668,
+      "learning_rate": 0.0001348277293482773,
+      "loss": 0.9344,
+      "step": 1720
+    },
+    {
+      "epoch": 1.0386356776335648,
+      "grad_norm": 0.09110064059495926,
+      "learning_rate": 0.0001347862183478622,
+      "loss": 0.7452,
+      "step": 1721
+    },
+    {
+      "epoch": 1.0392393600965892,
+      "grad_norm": 0.09389975666999817,
+      "learning_rate": 0.0001347447073474471,
+      "loss": 1.061,
+      "step": 1722
+    },
+    {
+      "epoch": 1.0398430425596137,
+      "grad_norm": 0.10916675627231598,
+      "learning_rate": 0.00013470319634703196,
+      "loss": 0.7749,
+      "step": 1723
+    },
+    {
+      "epoch": 1.0404467250226381,
+      "grad_norm": 0.08076412975788116,
+      "learning_rate": 0.00013466168534661686,
+      "loss": 0.5778,
+      "step": 1724
+    },
+    {
+      "epoch": 1.0410504074856626,
+      "grad_norm": 0.09830320626497269,
+      "learning_rate": 0.00013462017434620176,
+      "loss": 0.7115,
+      "step": 1725
+    },
+    {
+      "epoch": 1.041654089948687,
+      "grad_norm": 0.09751685708761215,
+      "learning_rate": 0.00013457866334578664,
+      "loss": 0.7439,
+      "step": 1726
+    },
+    {
+      "epoch": 1.0422577724117115,
+      "grad_norm": 0.09098626673221588,
+      "learning_rate": 0.0001345371523453715,
+      "loss": 0.7062,
+      "step": 1727
+    },
+    {
+      "epoch": 1.042861454874736,
+      "grad_norm": 0.10907137393951416,
+      "learning_rate": 0.00013449564134495644,
+      "loss": 0.888,
+      "step": 1728
+    },
+    {
+      "epoch": 1.0434651373377604,
+      "grad_norm": 0.09241855889558792,
+      "learning_rate": 0.0001344541303445413,
+      "loss": 0.6564,
+      "step": 1729
+    },
+    {
+      "epoch": 1.0440688198007848,
+      "grad_norm": 0.092192642390728,
+      "learning_rate": 0.00013441261934412618,
+      "loss": 0.6394,
+      "step": 1730
+    },
+    {
+      "epoch": 1.0446725022638093,
+      "grad_norm": 0.08987250179052353,
+      "learning_rate": 0.00013437110834371108,
+      "loss": 0.8383,
+      "step": 1731
+    },
+    {
+      "epoch": 1.0452761847268337,
+      "grad_norm": 0.09562289714813232,
+      "learning_rate": 0.00013432959734329598,
+      "loss": 0.6214,
+      "step": 1732
+    },
+    {
+      "epoch": 1.0458798671898581,
+      "grad_norm": 0.0972490981221199,
+      "learning_rate": 0.00013428808634288086,
+      "loss": 0.7911,
+      "step": 1733
+    },
+    {
+      "epoch": 1.0464835496528826,
+      "grad_norm": 0.10719991475343704,
+      "learning_rate": 0.00013424657534246576,
+      "loss": 0.8041,
+      "step": 1734
+    },
+    {
+      "epoch": 1.047087232115907,
+      "grad_norm": 0.09392597526311874,
+      "learning_rate": 0.00013420506434205066,
+      "loss": 0.9184,
+      "step": 1735
+    },
+    {
+      "epoch": 1.0476909145789315,
+      "grad_norm": 0.09040381014347076,
+      "learning_rate": 0.00013416355334163556,
+      "loss": 0.8159,
+      "step": 1736
+    },
+    {
+      "epoch": 1.048294597041956,
+      "grad_norm": 0.09064770489931107,
+      "learning_rate": 0.00013412204234122043,
+      "loss": 0.768,
+      "step": 1737
+    },
+    {
+      "epoch": 1.0488982795049804,
+      "grad_norm": 0.0895996019244194,
+      "learning_rate": 0.00013408053134080533,
+      "loss": 0.6926,
+      "step": 1738
+    },
+    {
+      "epoch": 1.0495019619680048,
+      "grad_norm": 0.10096043348312378,
+      "learning_rate": 0.00013403902034039023,
+      "loss": 0.8543,
+      "step": 1739
+    },
+    {
+      "epoch": 1.0501056444310293,
+      "grad_norm": 0.09393858164548874,
+      "learning_rate": 0.0001339975093399751,
+      "loss": 0.6274,
+      "step": 1740
+    },
+    {
+      "epoch": 1.0507093268940537,
+      "grad_norm": 0.0904972106218338,
+      "learning_rate": 0.00013395599833955997,
+      "loss": 0.636,
+      "step": 1741
+    },
+    {
+      "epoch": 1.0513130093570782,
+      "grad_norm": 0.09535270184278488,
+      "learning_rate": 0.0001339144873391449,
+      "loss": 0.6623,
+      "step": 1742
+    },
+    {
+      "epoch": 1.0519166918201026,
+      "grad_norm": 0.09259801357984543,
+      "learning_rate": 0.00013387297633872977,
+      "loss": 0.659,
+      "step": 1743
+    },
+    {
+      "epoch": 1.052520374283127,
+      "grad_norm": 0.09762239456176758,
+      "learning_rate": 0.00013383146533831465,
+      "loss": 0.6494,
+      "step": 1744
+    },
+    {
+      "epoch": 1.0531240567461515,
+      "grad_norm": 0.104673370718956,
+      "learning_rate": 0.00013378995433789955,
+      "loss": 0.6844,
+      "step": 1745
+    },
+    {
+      "epoch": 1.053727739209176,
+      "grad_norm": 0.108694888651371,
+      "learning_rate": 0.00013374844333748445,
+      "loss": 0.6817,
+      "step": 1746
+    },
+    {
+      "epoch": 1.0543314216722004,
+      "grad_norm": 0.10462528467178345,
+      "learning_rate": 0.00013370693233706932,
+      "loss": 0.6151,
+      "step": 1747
+    },
+    {
+      "epoch": 1.0549351041352248,
+      "grad_norm": 0.11397890001535416,
+      "learning_rate": 0.00013366542133665422,
+      "loss": 0.6114,
+      "step": 1748
+    },
+    {
+      "epoch": 1.0555387865982493,
+      "grad_norm": 0.10102736949920654,
+      "learning_rate": 0.00013362391033623912,
+      "loss": 0.6314,
+      "step": 1749
+    },
+    {
+      "epoch": 1.0561424690612737,
+      "grad_norm": 0.12406179308891296,
+      "learning_rate": 0.000133582399335824,
+      "loss": 0.6278,
+      "step": 1750
+    },
+    {
+      "epoch": 1.0567461515242982,
+      "grad_norm": 0.11435554176568985,
+      "learning_rate": 0.0001335408883354089,
+      "loss": 0.5789,
+      "step": 1751
+    },
+    {
+      "epoch": 1.0573498339873226,
+      "grad_norm": 0.12788087129592896,
+      "learning_rate": 0.0001334993773349938,
+      "loss": 0.5205,
+      "step": 1752
+    },
+    {
+      "epoch": 1.057953516450347,
+      "grad_norm": 0.1364794224500656,
+      "learning_rate": 0.00013345786633457867,
+      "loss": 0.5034,
+      "step": 1753
+    },
+    {
+      "epoch": 1.0585571989133715,
+      "grad_norm": 0.13836418092250824,
+      "learning_rate": 0.00013341635533416357,
+      "loss": 0.5269,
+      "step": 1754
+    },
+    {
+      "epoch": 1.059160881376396,
+      "grad_norm": 0.14710794389247894,
+      "learning_rate": 0.00013337484433374844,
+      "loss": 0.4156,
+      "step": 1755
+    },
+    {
+      "epoch": 1.0597645638394204,
+      "grad_norm": 0.13123385608196259,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 0.32,
+      "step": 1756
+    },
+    {
+      "epoch": 1.0603682463024449,
+      "grad_norm": 0.15206588804721832,
+      "learning_rate": 0.00013329182233291824,
+      "loss": 0.2739,
+      "step": 1757
+    },
+    {
+      "epoch": 1.0609719287654693,
+      "grad_norm": 0.13307802379131317,
+      "learning_rate": 0.0001332503113325031,
+      "loss": 0.8787,
+      "step": 1758
+    },
+    {
+      "epoch": 1.0615756112284938,
+      "grad_norm": 0.11052402853965759,
+      "learning_rate": 0.000133208800332088,
+      "loss": 0.7224,
+      "step": 1759
+    },
+    {
+      "epoch": 1.0621792936915182,
+      "grad_norm": 0.17755720019340515,
+      "learning_rate": 0.0001331672893316729,
+      "loss": 0.9626,
+      "step": 1760
+    },
+    {
+      "epoch": 1.0627829761545426,
+      "grad_norm": 0.10982788354158401,
+      "learning_rate": 0.00013312577833125779,
+      "loss": 0.789,
+      "step": 1761
+    },
+    {
+      "epoch": 1.063386658617567,
+      "grad_norm": 0.10073138773441315,
+      "learning_rate": 0.00013308426733084266,
+      "loss": 0.7448,
+      "step": 1762
+    },
+    {
+      "epoch": 1.0639903410805915,
+      "grad_norm": 0.09374914318323135,
+      "learning_rate": 0.00013304275633042759,
+      "loss": 0.7442,
+      "step": 1763
+    },
+    {
+      "epoch": 1.064594023543616,
+      "grad_norm": 0.10434511303901672,
+      "learning_rate": 0.00013300124533001246,
+      "loss": 0.8159,
+      "step": 1764
+    },
+    {
+      "epoch": 1.0651977060066404,
+      "grad_norm": 0.11345966905355453,
+      "learning_rate": 0.00013295973432959733,
+      "loss": 0.804,
+      "step": 1765
+    },
+    {
+      "epoch": 1.0658013884696649,
+      "grad_norm": 0.10115978866815567,
+      "learning_rate": 0.00013291822332918226,
+      "loss": 0.802,
+      "step": 1766
+    },
+    {
+      "epoch": 1.0664050709326893,
+      "grad_norm": 0.10045889765024185,
+      "learning_rate": 0.00013287671232876713,
+      "loss": 0.7314,
+      "step": 1767
+    },
+    {
+      "epoch": 1.0670087533957138,
+      "grad_norm": 0.11507212370634079,
+      "learning_rate": 0.000132835201328352,
+      "loss": 1.0092,
+      "step": 1768
+    },
+    {
+      "epoch": 1.0676124358587382,
+      "grad_norm": 0.09701373428106308,
+      "learning_rate": 0.0001327936903279369,
+      "loss": 0.9924,
+      "step": 1769
+    },
+    {
+      "epoch": 1.0682161183217627,
+      "grad_norm": 0.10776803642511368,
+      "learning_rate": 0.0001327521793275218,
+      "loss": 0.7889,
+      "step": 1770
+    },
+    {
+      "epoch": 1.068819800784787,
+      "grad_norm": 0.09671414643526077,
+      "learning_rate": 0.00013271066832710668,
+      "loss": 0.7996,
+      "step": 1771
+    },
+    {
+      "epoch": 1.0694234832478116,
+      "grad_norm": 0.10002996027469635,
+      "learning_rate": 0.00013266915732669158,
+      "loss": 0.7149,
+      "step": 1772
+    },
+    {
+      "epoch": 1.070027165710836,
+      "grad_norm": 0.09639476984739304,
+      "learning_rate": 0.00013262764632627648,
+      "loss": 0.9136,
+      "step": 1773
+    },
+    {
+      "epoch": 1.0706308481738604,
+      "grad_norm": 0.11277900636196136,
+      "learning_rate": 0.00013258613532586135,
+      "loss": 0.7176,
+      "step": 1774
+    },
+    {
+      "epoch": 1.071234530636885,
+      "grad_norm": 0.09514438360929489,
+      "learning_rate": 0.00013254462432544625,
+      "loss": 0.7988,
+      "step": 1775
+    },
+    {
+      "epoch": 1.0718382130999093,
+      "grad_norm": 0.09252556413412094,
+      "learning_rate": 0.00013250311332503112,
+      "loss": 0.7631,
+      "step": 1776
+    },
+    {
+      "epoch": 1.072441895562934,
+      "grad_norm": 0.10669627785682678,
+      "learning_rate": 0.00013246160232461602,
+      "loss": 0.8116,
+      "step": 1777
+    },
+    {
+      "epoch": 1.0730455780259582,
+      "grad_norm": 0.10841196030378342,
+      "learning_rate": 0.00013242009132420092,
+      "loss": 0.7423,
+      "step": 1778
+    },
+    {
+      "epoch": 1.073649260488983,
+      "grad_norm": 0.10168831795454025,
+      "learning_rate": 0.0001323785803237858,
+      "loss": 0.7702,
+      "step": 1779
+    },
+    {
+      "epoch": 1.0742529429520071,
+      "grad_norm": 0.10742917656898499,
+      "learning_rate": 0.00013233706932337072,
+      "loss": 0.8964,
+      "step": 1780
+    },
+    {
+      "epoch": 1.0748566254150318,
+      "grad_norm": 0.09869187325239182,
+      "learning_rate": 0.0001322955583229556,
+      "loss": 0.7048,
+      "step": 1781
+    },
+    {
+      "epoch": 1.0754603078780562,
+      "grad_norm": 0.1356370896100998,
+      "learning_rate": 0.00013225404732254047,
+      "loss": 0.9056,
+      "step": 1782
+    },
+    {
+      "epoch": 1.0760639903410807,
+      "grad_norm": 0.0955684557557106,
+      "learning_rate": 0.00013221253632212537,
+      "loss": 0.715,
+      "step": 1783
+    },
+    {
+      "epoch": 1.0766676728041051,
+      "grad_norm": 0.09964524954557419,
+      "learning_rate": 0.00013217102532171027,
+      "loss": 0.7039,
+      "step": 1784
+    },
+    {
+      "epoch": 1.0772713552671296,
+      "grad_norm": 0.10292885452508926,
+      "learning_rate": 0.00013212951432129514,
+      "loss": 0.9838,
+      "step": 1785
+    },
+    {
+      "epoch": 1.077875037730154,
+      "grad_norm": 0.101532481610775,
+      "learning_rate": 0.00013208800332088004,
+      "loss": 0.6629,
+      "step": 1786
+    },
+    {
+      "epoch": 1.0784787201931785,
+      "grad_norm": 0.09781339764595032,
+      "learning_rate": 0.00013204649232046494,
+      "loss": 0.789,
+      "step": 1787
+    },
+    {
+      "epoch": 1.079082402656203,
+      "grad_norm": 0.10302070528268814,
+      "learning_rate": 0.00013200498132004982,
+      "loss": 0.7772,
+      "step": 1788
+    },
+    {
+      "epoch": 1.0796860851192274,
+      "grad_norm": 0.10221746563911438,
+      "learning_rate": 0.00013196347031963472,
+      "loss": 0.8008,
+      "step": 1789
+    },
+    {
+      "epoch": 1.0802897675822518,
+      "grad_norm": 0.09660208970308304,
+      "learning_rate": 0.0001319219593192196,
+      "loss": 0.8582,
+      "step": 1790
+    },
+    {
+      "epoch": 1.0808934500452763,
+      "grad_norm": 0.09637465327978134,
+      "learning_rate": 0.0001318804483188045,
+      "loss": 0.7061,
+      "step": 1791
+    },
+    {
+      "epoch": 1.0814971325083007,
+      "grad_norm": 0.09803172200918198,
+      "learning_rate": 0.0001318389373183894,
+      "loss": 0.6043,
+      "step": 1792
+    },
+    {
+      "epoch": 1.0821008149713252,
+      "grad_norm": 0.10259698331356049,
+      "learning_rate": 0.00013179742631797426,
+      "loss": 0.6007,
+      "step": 1793
+    },
+    {
+      "epoch": 1.0827044974343496,
+      "grad_norm": 0.10633745789527893,
+      "learning_rate": 0.00013175591531755916,
+      "loss": 0.6679,
+      "step": 1794
+    },
+    {
+      "epoch": 1.083308179897374,
+      "grad_norm": 0.10431138426065445,
+      "learning_rate": 0.00013171440431714406,
+      "loss": 0.6999,
+      "step": 1795
+    },
+    {
+      "epoch": 1.0839118623603985,
+      "grad_norm": 0.10492059588432312,
+      "learning_rate": 0.00013167289331672893,
+      "loss": 0.7314,
+      "step": 1796
+    },
+    {
+      "epoch": 1.084515544823423,
+      "grad_norm": 0.1053004339337349,
+      "learning_rate": 0.00013163138231631383,
+      "loss": 0.5873,
+      "step": 1797
+    },
+    {
+      "epoch": 1.0851192272864474,
+      "grad_norm": 0.10796685516834259,
+      "learning_rate": 0.00013158987131589873,
+      "loss": 0.6122,
+      "step": 1798
+    },
+    {
+      "epoch": 1.0857229097494718,
+      "grad_norm": 0.11339765042066574,
+      "learning_rate": 0.0001315483603154836,
+      "loss": 0.6486,
+      "step": 1799
+    },
+    {
+      "epoch": 1.0863265922124963,
+      "grad_norm": 0.11134713143110275,
+      "learning_rate": 0.00013150684931506848,
+      "loss": 0.5925,
+      "step": 1800
+    },
+    {
+      "epoch": 1.0869302746755207,
+      "grad_norm": 0.12589313089847565,
+      "learning_rate": 0.0001314653383146534,
+      "loss": 0.6238,
+      "step": 1801
+    },
+    {
+      "epoch": 1.0875339571385452,
+      "grad_norm": 0.1222606673836708,
+      "learning_rate": 0.00013142382731423828,
+      "loss": 0.5478,
+      "step": 1802
+    },
+    {
+      "epoch": 1.0881376396015696,
+      "grad_norm": 0.1365615576505661,
+      "learning_rate": 0.00013138231631382315,
+      "loss": 0.5323,
+      "step": 1803
+    },
+    {
+      "epoch": 1.088741322064594,
+      "grad_norm": 0.1305444985628128,
+      "learning_rate": 0.00013134080531340805,
+      "loss": 0.5246,
+      "step": 1804
+    },
+    {
+      "epoch": 1.0893450045276185,
+      "grad_norm": 0.1343401074409485,
+      "learning_rate": 0.00013129929431299295,
+      "loss": 0.4212,
+      "step": 1805
+    },
+    {
+      "epoch": 1.089948686990643,
+      "grad_norm": 0.14103184640407562,
+      "learning_rate": 0.00013125778331257783,
+      "loss": 0.4255,
+      "step": 1806
+    },
+    {
+      "epoch": 1.0905523694536674,
+      "grad_norm": 0.15704935789108276,
+      "learning_rate": 0.00013121627231216273,
+      "loss": 0.3559,
+      "step": 1807
+    },
+    {
+      "epoch": 1.0911560519166918,
+      "grad_norm": 0.4550192058086395,
+      "learning_rate": 0.00013117476131174763,
+      "loss": 0.7013,
+      "step": 1808
+    },
+    {
+      "epoch": 1.0917597343797163,
+      "grad_norm": 0.12131811678409576,
+      "learning_rate": 0.0001311332503113325,
+      "loss": 0.7005,
+      "step": 1809
+    },
+    {
+      "epoch": 1.0923634168427407,
+      "grad_norm": 0.11735139787197113,
+      "learning_rate": 0.0001310917393109174,
+      "loss": 0.6871,
+      "step": 1810
+    },
+    {
+      "epoch": 1.0929670993057652,
+      "grad_norm": 0.11844879388809204,
+      "learning_rate": 0.0001310502283105023,
+      "loss": 1.2869,
+      "step": 1811
+    },
+    {
+      "epoch": 1.0935707817687896,
+      "grad_norm": 0.11029795557260513,
+      "learning_rate": 0.00013100871731008717,
+      "loss": 0.7609,
+      "step": 1812
+    },
+    {
+      "epoch": 1.094174464231814,
+      "grad_norm": 0.11960664391517639,
+      "learning_rate": 0.00013096720630967207,
+      "loss": 0.6971,
+      "step": 1813
+    },
+    {
+      "epoch": 1.0947781466948385,
+      "grad_norm": 0.09530165046453476,
+      "learning_rate": 0.00013092569530925695,
+      "loss": 0.7224,
+      "step": 1814
+    },
+    {
+      "epoch": 1.095381829157863,
+      "grad_norm": 0.1267603486776352,
+      "learning_rate": 0.00013088418430884185,
+      "loss": 0.815,
+      "step": 1815
+    },
+    {
+      "epoch": 1.0959855116208874,
+      "grad_norm": 0.11272140592336655,
+      "learning_rate": 0.00013084267330842675,
+      "loss": 0.78,
+      "step": 1816
+    },
+    {
+      "epoch": 1.0965891940839119,
+      "grad_norm": 0.10290253907442093,
+      "learning_rate": 0.00013080116230801162,
+      "loss": 0.7218,
+      "step": 1817
+    },
+    {
+      "epoch": 1.0971928765469363,
+      "grad_norm": 0.09555505961179733,
+      "learning_rate": 0.00013075965130759652,
+      "loss": 0.7541,
+      "step": 1818
+    },
+    {
+      "epoch": 1.0977965590099608,
+      "grad_norm": 0.10414687544107437,
+      "learning_rate": 0.00013071814030718142,
+      "loss": 0.9603,
+      "step": 1819
+    },
+    {
+      "epoch": 1.0984002414729852,
+      "grad_norm": 0.09657555818557739,
+      "learning_rate": 0.0001306766293067663,
+      "loss": 0.7862,
+      "step": 1820
+    },
+    {
+      "epoch": 1.0990039239360097,
+      "grad_norm": 0.09514836966991425,
+      "learning_rate": 0.0001306351183063512,
+      "loss": 0.7408,
+      "step": 1821
+    },
+    {
+      "epoch": 1.099607606399034,
+      "grad_norm": 0.09766734391450882,
+      "learning_rate": 0.0001305936073059361,
+      "loss": 0.7459,
+      "step": 1822
+    },
+    {
+      "epoch": 1.1002112888620585,
+      "grad_norm": 0.08910214155912399,
+      "learning_rate": 0.00013055209630552096,
+      "loss": 1.0229,
+      "step": 1823
+    },
+    {
+      "epoch": 1.100814971325083,
+      "grad_norm": 0.10597441345453262,
+      "learning_rate": 0.00013051058530510586,
+      "loss": 0.7877,
+      "step": 1824
+    },
+    {
+      "epoch": 1.1014186537881074,
+      "grad_norm": 0.10066650807857513,
+      "learning_rate": 0.00013046907430469076,
+      "loss": 0.6925,
+      "step": 1825
+    },
+    {
+      "epoch": 1.1020223362511319,
+      "grad_norm": 0.11692957580089569,
+      "learning_rate": 0.00013042756330427564,
+      "loss": 0.8301,
+      "step": 1826
+    },
+    {
+      "epoch": 1.1026260187141563,
+      "grad_norm": 0.10520220547914505,
+      "learning_rate": 0.00013038605230386054,
+      "loss": 0.8868,
+      "step": 1827
+    },
+    {
+      "epoch": 1.1032297011771808,
+      "grad_norm": 0.10756828635931015,
+      "learning_rate": 0.0001303445413034454,
+      "loss": 0.7164,
+      "step": 1828
+    },
+    {
+      "epoch": 1.1038333836402052,
+      "grad_norm": 0.10687938332557678,
+      "learning_rate": 0.0001303030303030303,
+      "loss": 0.8075,
+      "step": 1829
+    },
+    {
+      "epoch": 1.1044370661032297,
+      "grad_norm": 0.09681583940982819,
+      "learning_rate": 0.0001302615193026152,
+      "loss": 0.7073,
+      "step": 1830
+    },
+    {
+      "epoch": 1.1050407485662541,
+      "grad_norm": 0.0960078164935112,
+      "learning_rate": 0.00013022000830220008,
+      "loss": 0.7665,
+      "step": 1831
+    },
+    {
+      "epoch": 1.1056444310292786,
+      "grad_norm": 0.10860790312290192,
+      "learning_rate": 0.00013017849730178498,
+      "loss": 0.7155,
+      "step": 1832
+    },
+    {
+      "epoch": 1.106248113492303,
+      "grad_norm": 0.10167111456394196,
+      "learning_rate": 0.00013013698630136988,
+      "loss": 0.7887,
+      "step": 1833
+    },
+    {
+      "epoch": 1.1068517959553275,
+      "grad_norm": 0.09937427192926407,
+      "learning_rate": 0.00013009547530095476,
+      "loss": 0.7305,
+      "step": 1834
+    },
+    {
+      "epoch": 1.107455478418352,
+      "grad_norm": 0.09677271544933319,
+      "learning_rate": 0.00013005396430053966,
+      "loss": 0.7064,
+      "step": 1835
+    },
+    {
+      "epoch": 1.1080591608813763,
+      "grad_norm": 0.10012287646532059,
+      "learning_rate": 0.00013001245330012456,
+      "loss": 0.6711,
+      "step": 1836
+    },
+    {
+      "epoch": 1.1086628433444008,
+      "grad_norm": 0.10501682013273239,
+      "learning_rate": 0.00012997094229970943,
+      "loss": 0.8093,
+      "step": 1837
+    },
+    {
+      "epoch": 1.1092665258074252,
+      "grad_norm": 0.09496523439884186,
+      "learning_rate": 0.0001299294312992943,
+      "loss": 0.7329,
+      "step": 1838
+    },
+    {
+      "epoch": 1.1098702082704497,
+      "grad_norm": 0.09581634402275085,
+      "learning_rate": 0.00012988792029887923,
+      "loss": 0.6305,
+      "step": 1839
+    },
+    {
+      "epoch": 1.1104738907334741,
+      "grad_norm": 0.09801549464464188,
+      "learning_rate": 0.0001298464092984641,
+      "loss": 0.5904,
+      "step": 1840
+    },
+    {
+      "epoch": 1.1110775731964986,
+      "grad_norm": 0.0966857448220253,
+      "learning_rate": 0.00012980489829804898,
+      "loss": 0.7596,
+      "step": 1841
+    },
+    {
+      "epoch": 1.111681255659523,
+      "grad_norm": 0.10019635409116745,
+      "learning_rate": 0.00012976338729763388,
+      "loss": 0.6856,
+      "step": 1842
+    },
+    {
+      "epoch": 1.1122849381225475,
+      "grad_norm": 0.09953883290290833,
+      "learning_rate": 0.00012972187629721878,
+      "loss": 0.7132,
+      "step": 1843
+    },
+    {
+      "epoch": 1.112888620585572,
+      "grad_norm": 0.09991202503442764,
+      "learning_rate": 0.00012968036529680365,
+      "loss": 0.613,
+      "step": 1844
+    },
+    {
+      "epoch": 1.1134923030485964,
+      "grad_norm": 0.09988027811050415,
+      "learning_rate": 0.00012963885429638855,
+      "loss": 0.6142,
+      "step": 1845
+    },
+    {
+      "epoch": 1.1140959855116208,
+      "grad_norm": 0.10957928746938705,
+      "learning_rate": 0.00012959734329597345,
+      "loss": 0.6573,
+      "step": 1846
+    },
+    {
+      "epoch": 1.1146996679746453,
+      "grad_norm": 0.11322573572397232,
+      "learning_rate": 0.00012955583229555832,
+      "loss": 0.6891,
+      "step": 1847
+    },
+    {
+      "epoch": 1.1153033504376697,
+      "grad_norm": 0.11455134302377701,
+      "learning_rate": 0.00012951432129514322,
+      "loss": 0.6676,
+      "step": 1848
+    },
+    {
+      "epoch": 1.1159070329006942,
+      "grad_norm": 0.14914073050022125,
+      "learning_rate": 0.00012947281029472812,
+      "loss": 0.6206,
+      "step": 1849
+    },
+    {
+      "epoch": 1.1165107153637186,
+      "grad_norm": 0.12033107876777649,
+      "learning_rate": 0.000129431299294313,
+      "loss": 0.5269,
+      "step": 1850
+    },
+    {
+      "epoch": 1.117114397826743,
+      "grad_norm": 0.13541612029075623,
+      "learning_rate": 0.0001293897882938979,
+      "loss": 0.6263,
+      "step": 1851
+    },
+    {
+      "epoch": 1.1177180802897675,
+      "grad_norm": 0.12948794662952423,
+      "learning_rate": 0.00012934827729348277,
+      "loss": 0.6229,
+      "step": 1852
+    },
+    {
+      "epoch": 1.118321762752792,
+      "grad_norm": 0.13281860947608948,
+      "learning_rate": 0.00012930676629306767,
+      "loss": 0.6001,
+      "step": 1853
+    },
+    {
+      "epoch": 1.1189254452158164,
+      "grad_norm": 0.14206023514270782,
+      "learning_rate": 0.00012926525529265257,
+      "loss": 0.5866,
+      "step": 1854
+    },
+    {
+      "epoch": 1.1195291276788408,
+      "grad_norm": 0.15054769814014435,
+      "learning_rate": 0.00012922374429223744,
+      "loss": 0.4866,
+      "step": 1855
+    },
+    {
+      "epoch": 1.1201328101418653,
+      "grad_norm": 0.15093187987804413,
+      "learning_rate": 0.00012918223329182234,
+      "loss": 0.4194,
+      "step": 1856
+    },
+    {
+      "epoch": 1.1207364926048897,
+      "grad_norm": 0.14579585194587708,
+      "learning_rate": 0.00012914072229140724,
+      "loss": 0.3209,
+      "step": 1857
+    },
+    {
+      "epoch": 1.1213401750679144,
+      "grad_norm": 0.11571165919303894,
+      "learning_rate": 0.0001290992112909921,
+      "loss": 0.8016,
+      "step": 1858
+    },
+    {
+      "epoch": 1.1219438575309386,
+      "grad_norm": 0.13167442381381989,
+      "learning_rate": 0.00012905770029057699,
+      "loss": 0.7073,
+      "step": 1859
+    },
+    {
+      "epoch": 1.1225475399939633,
+      "grad_norm": 0.09710317105054855,
+      "learning_rate": 0.0001290161892901619,
+      "loss": 0.6618,
+      "step": 1860
+    },
+    {
+      "epoch": 1.1231512224569875,
+      "grad_norm": 0.1127190887928009,
+      "learning_rate": 0.00012897467828974679,
+      "loss": 0.7702,
+      "step": 1861
+    },
+    {
+      "epoch": 1.1237549049200122,
+      "grad_norm": 0.09531667083501816,
+      "learning_rate": 0.00012893316728933166,
+      "loss": 0.7331,
+      "step": 1862
+    },
+    {
+      "epoch": 1.1243585873830366,
+      "grad_norm": 0.10616014897823334,
+      "learning_rate": 0.0001288916562889166,
+      "loss": 0.6404,
+      "step": 1863
+    },
+    {
+      "epoch": 1.124962269846061,
+      "grad_norm": 0.09906172007322311,
+      "learning_rate": 0.00012885014528850146,
+      "loss": 0.7295,
+      "step": 1864
+    },
+    {
+      "epoch": 1.1255659523090855,
+      "grad_norm": 0.10826185345649719,
+      "learning_rate": 0.00012880863428808633,
+      "loss": 0.7036,
+      "step": 1865
+    },
+    {
+      "epoch": 1.12616963477211,
+      "grad_norm": 0.09856102615594864,
+      "learning_rate": 0.00012876712328767123,
+      "loss": 0.6587,
+      "step": 1866
+    },
+    {
+      "epoch": 1.1267733172351344,
+      "grad_norm": 0.11088103801012039,
+      "learning_rate": 0.00012872561228725613,
+      "loss": 0.8501,
+      "step": 1867
+    },
+    {
+      "epoch": 1.1273769996981589,
+      "grad_norm": 0.10369934886693954,
+      "learning_rate": 0.00012868410128684103,
+      "loss": 0.9292,
+      "step": 1868
+    },
+    {
+      "epoch": 1.1279806821611833,
+      "grad_norm": 0.10459262132644653,
+      "learning_rate": 0.0001286425902864259,
+      "loss": 0.8489,
+      "step": 1869
+    },
+    {
+      "epoch": 1.1285843646242077,
+      "grad_norm": 0.10528787970542908,
+      "learning_rate": 0.0001286010792860108,
+      "loss": 0.7793,
+      "step": 1870
+    },
+    {
+      "epoch": 1.1291880470872322,
+      "grad_norm": 0.10074903815984726,
+      "learning_rate": 0.0001285595682855957,
+      "loss": 0.7929,
+      "step": 1871
+    },
+    {
+      "epoch": 1.1297917295502566,
+      "grad_norm": 0.09653479605913162,
+      "learning_rate": 0.00012851805728518058,
+      "loss": 0.8025,
+      "step": 1872
+    },
+    {
+      "epoch": 1.130395412013281,
+      "grad_norm": 0.10775607079267502,
+      "learning_rate": 0.00012847654628476545,
+      "loss": 0.9679,
+      "step": 1873
+    },
+    {
+      "epoch": 1.1309990944763055,
+      "grad_norm": 0.10731203109025955,
+      "learning_rate": 0.00012843503528435038,
+      "loss": 0.7715,
+      "step": 1874
+    },
+    {
+      "epoch": 1.13160277693933,
+      "grad_norm": 0.1124483272433281,
+      "learning_rate": 0.00012839352428393525,
+      "loss": 0.8295,
+      "step": 1875
+    },
+    {
+      "epoch": 1.1322064594023544,
+      "grad_norm": 0.21262244880199432,
+      "learning_rate": 0.00012835201328352012,
+      "loss": 1.0585,
+      "step": 1876
+    },
+    {
+      "epoch": 1.1328101418653789,
+      "grad_norm": 0.10024654865264893,
+      "learning_rate": 0.00012831050228310505,
+      "loss": 0.7567,
+      "step": 1877
+    },
+    {
+      "epoch": 1.1334138243284033,
+      "grad_norm": 0.11772340536117554,
+      "learning_rate": 0.00012826899128268992,
+      "loss": 0.8871,
+      "step": 1878
+    },
+    {
+      "epoch": 1.1340175067914278,
+      "grad_norm": 0.10074224323034286,
+      "learning_rate": 0.0001282274802822748,
+      "loss": 0.9096,
+      "step": 1879
+    },
+    {
+      "epoch": 1.1346211892544522,
+      "grad_norm": 0.1045253723859787,
+      "learning_rate": 0.0001281859692818597,
+      "loss": 0.6523,
+      "step": 1880
+    },
+    {
+      "epoch": 1.1352248717174767,
+      "grad_norm": 0.10628263652324677,
+      "learning_rate": 0.0001281444582814446,
+      "loss": 0.6945,
+      "step": 1881
+    },
+    {
+      "epoch": 1.135828554180501,
+      "grad_norm": 0.09413371235132217,
+      "learning_rate": 0.00012810294728102947,
+      "loss": 0.6471,
+      "step": 1882
+    },
+    {
+      "epoch": 1.1364322366435256,
+      "grad_norm": 0.10187766700983047,
+      "learning_rate": 0.00012806143628061437,
+      "loss": 0.7088,
+      "step": 1883
+    },
+    {
+      "epoch": 1.13703591910655,
+      "grad_norm": 0.10155726969242096,
+      "learning_rate": 0.00012801992528019927,
+      "loss": 0.8189,
+      "step": 1884
+    },
+    {
+      "epoch": 1.1376396015695744,
+      "grad_norm": 0.101055808365345,
+      "learning_rate": 0.00012797841427978414,
+      "loss": 0.753,
+      "step": 1885
+    },
+    {
+      "epoch": 1.138243284032599,
+      "grad_norm": 0.10111107677221298,
+      "learning_rate": 0.00012793690327936904,
+      "loss": 0.8097,
+      "step": 1886
+    },
+    {
+      "epoch": 1.1388469664956233,
+      "grad_norm": 0.10023274272680283,
+      "learning_rate": 0.00012789539227895392,
+      "loss": 0.8272,
+      "step": 1887
+    },
+    {
+      "epoch": 1.1394506489586478,
+      "grad_norm": 0.09850174188613892,
+      "learning_rate": 0.00012785388127853882,
+      "loss": 0.9205,
+      "step": 1888
+    },
+    {
+      "epoch": 1.1400543314216722,
+      "grad_norm": 0.10670360922813416,
+      "learning_rate": 0.00012781237027812372,
+      "loss": 0.7138,
+      "step": 1889
+    },
+    {
+      "epoch": 1.1406580138846967,
+      "grad_norm": 0.10679405182600021,
+      "learning_rate": 0.0001277708592777086,
+      "loss": 0.8036,
+      "step": 1890
+    },
+    {
+      "epoch": 1.1412616963477211,
+      "grad_norm": 0.10706860572099686,
+      "learning_rate": 0.0001277293482772935,
+      "loss": 0.6126,
+      "step": 1891
+    },
+    {
+      "epoch": 1.1418653788107456,
+      "grad_norm": 0.10065101832151413,
+      "learning_rate": 0.0001276878372768784,
+      "loss": 0.6491,
+      "step": 1892
+    },
+    {
+      "epoch": 1.14246906127377,
+      "grad_norm": 0.10702386498451233,
+      "learning_rate": 0.00012764632627646326,
+      "loss": 0.7233,
+      "step": 1893
+    },
+    {
+      "epoch": 1.1430727437367945,
+      "grad_norm": 0.09997487813234329,
+      "learning_rate": 0.00012760481527604816,
+      "loss": 0.6635,
+      "step": 1894
+    },
+    {
+      "epoch": 1.143676426199819,
+      "grad_norm": 0.09972809255123138,
+      "learning_rate": 0.00012756330427563306,
+      "loss": 0.5934,
+      "step": 1895
+    },
+    {
+      "epoch": 1.1442801086628434,
+      "grad_norm": 0.11297734826803207,
+      "learning_rate": 0.00012752179327521794,
+      "loss": 0.6151,
+      "step": 1896
+    },
+    {
+      "epoch": 1.1448837911258678,
+      "grad_norm": 0.15789483487606049,
+      "learning_rate": 0.0001274802822748028,
+      "loss": 0.6286,
+      "step": 1897
+    },
+    {
+      "epoch": 1.1454874735888922,
+      "grad_norm": 0.10323326289653778,
+      "learning_rate": 0.00012743877127438774,
+      "loss": 0.5922,
+      "step": 1898
+    },
+    {
+      "epoch": 1.1460911560519167,
+      "grad_norm": 0.11414547264575958,
+      "learning_rate": 0.0001273972602739726,
+      "loss": 0.6048,
+      "step": 1899
+    },
+    {
+      "epoch": 1.1466948385149411,
+      "grad_norm": 0.11642369627952576,
+      "learning_rate": 0.00012735574927355748,
+      "loss": 0.6391,
+      "step": 1900
+    },
+    {
+      "epoch": 1.1472985209779656,
+      "grad_norm": 0.12589019536972046,
+      "learning_rate": 0.00012731423827314238,
+      "loss": 0.6153,
+      "step": 1901
+    },
+    {
+      "epoch": 1.14790220344099,
+      "grad_norm": 0.12789148092269897,
+      "learning_rate": 0.00012727272727272728,
+      "loss": 0.6229,
+      "step": 1902
+    },
+    {
+      "epoch": 1.1485058859040145,
+      "grad_norm": 0.14213663339614868,
+      "learning_rate": 0.00012723121627231215,
+      "loss": 0.579,
+      "step": 1903
+    },
+    {
+      "epoch": 1.149109568367039,
+      "grad_norm": 0.14148949086666107,
+      "learning_rate": 0.00012718970527189705,
+      "loss": 0.5596,
+      "step": 1904
+    },
+    {
+      "epoch": 1.1497132508300634,
+      "grad_norm": 0.14953409135341644,
+      "learning_rate": 0.00012714819427148195,
+      "loss": 0.4723,
+      "step": 1905
+    },
+    {
+      "epoch": 1.1503169332930878,
+      "grad_norm": 0.1479395627975464,
+      "learning_rate": 0.00012710668327106683,
+      "loss": 0.4351,
+      "step": 1906
+    },
+    {
+      "epoch": 1.1509206157561123,
+      "grad_norm": 0.13544803857803345,
+      "learning_rate": 0.00012706517227065173,
+      "loss": 0.3325,
+      "step": 1907
+    },
+    {
+      "epoch": 1.1515242982191367,
+      "grad_norm": 0.1175408810377121,
+      "learning_rate": 0.00012702366127023663,
+      "loss": 0.6767,
+      "step": 1908
+    },
+    {
+      "epoch": 1.1521279806821612,
+      "grad_norm": 0.209543377161026,
+      "learning_rate": 0.0001269821502698215,
+      "loss": 0.9705,
+      "step": 1909
+    },
+    {
+      "epoch": 1.1527316631451856,
+      "grad_norm": 0.10693584382534027,
+      "learning_rate": 0.0001269406392694064,
+      "loss": 0.7096,
+      "step": 1910
+    },
+    {
+      "epoch": 1.15333534560821,
+      "grad_norm": 0.10303391516208649,
+      "learning_rate": 0.00012689912826899127,
+      "loss": 0.7388,
+      "step": 1911
+    },
+    {
+      "epoch": 1.1539390280712345,
+      "grad_norm": 0.1123664528131485,
+      "learning_rate": 0.00012685761726857617,
+      "loss": 0.7372,
+      "step": 1912
+    },
+    {
+      "epoch": 1.154542710534259,
+      "grad_norm": 0.10753045976161957,
+      "learning_rate": 0.00012681610626816107,
+      "loss": 0.7714,
+      "step": 1913
+    },
+    {
+      "epoch": 1.1551463929972834,
+      "grad_norm": 0.10717937350273132,
+      "learning_rate": 0.00012677459526774595,
+      "loss": 0.6893,
+      "step": 1914
+    },
+    {
+      "epoch": 1.1557500754603078,
+      "grad_norm": 0.09875158965587616,
+      "learning_rate": 0.00012673308426733085,
+      "loss": 0.7437,
+      "step": 1915
+    },
+    {
+      "epoch": 1.1563537579233323,
+      "grad_norm": 0.12408756464719772,
+      "learning_rate": 0.00012669157326691575,
+      "loss": 0.7512,
+      "step": 1916
+    },
+    {
+      "epoch": 1.1569574403863567,
+      "grad_norm": 0.10806753486394882,
+      "learning_rate": 0.00012665006226650062,
+      "loss": 0.6687,
+      "step": 1917
+    },
+    {
+      "epoch": 1.1575611228493812,
+      "grad_norm": 0.10482674092054367,
+      "learning_rate": 0.00012660855126608552,
+      "loss": 0.6417,
+      "step": 1918
+    },
+    {
+      "epoch": 1.1581648053124056,
+      "grad_norm": 0.10658420622348785,
+      "learning_rate": 0.00012656704026567042,
+      "loss": 0.876,
+      "step": 1919
+    },
+    {
+      "epoch": 1.15876848777543,
+      "grad_norm": 0.10012143105268478,
+      "learning_rate": 0.0001265255292652553,
+      "loss": 0.6473,
+      "step": 1920
+    },
+    {
+      "epoch": 1.1593721702384545,
+      "grad_norm": 0.12289892137050629,
+      "learning_rate": 0.0001264840182648402,
+      "loss": 0.7602,
+      "step": 1921
+    },
+    {
+      "epoch": 1.159975852701479,
+      "grad_norm": 0.0990784764289856,
+      "learning_rate": 0.0001264425072644251,
+      "loss": 0.6372,
+      "step": 1922
+    },
+    {
+      "epoch": 1.1605795351645034,
+      "grad_norm": 0.0976581797003746,
+      "learning_rate": 0.00012640099626400997,
+      "loss": 0.7036,
+      "step": 1923
+    },
+    {
+      "epoch": 1.1611832176275279,
+      "grad_norm": 0.11021608114242554,
+      "learning_rate": 0.00012635948526359487,
+      "loss": 0.7666,
+      "step": 1924
+    },
+    {
+      "epoch": 1.1617869000905523,
+      "grad_norm": 0.2441258430480957,
+      "learning_rate": 0.00012631797426317974,
+      "loss": 0.7026,
+      "step": 1925
+    },
+    {
+      "epoch": 1.1623905825535767,
+      "grad_norm": 0.10685604065656662,
+      "learning_rate": 0.00012627646326276464,
+      "loss": 0.8768,
+      "step": 1926
+    },
+    {
+      "epoch": 1.1629942650166012,
+      "grad_norm": 0.10796701908111572,
+      "learning_rate": 0.00012623495226234954,
+      "loss": 0.9038,
+      "step": 1927
+    },
+    {
+      "epoch": 1.1635979474796256,
+      "grad_norm": 0.1107090562582016,
+      "learning_rate": 0.0001261934412619344,
+      "loss": 0.7874,
+      "step": 1928
+    },
+    {
+      "epoch": 1.16420162994265,
+      "grad_norm": 0.10373709350824356,
+      "learning_rate": 0.0001261519302615193,
+      "loss": 0.7946,
+      "step": 1929
+    },
+    {
+      "epoch": 1.1648053124056745,
+      "grad_norm": 0.09496215730905533,
+      "learning_rate": 0.0001261104192611042,
+      "loss": 0.688,
+      "step": 1930
+    },
+    {
+      "epoch": 1.165408994868699,
+      "grad_norm": 0.10575946420431137,
+      "learning_rate": 0.00012606890826068908,
+      "loss": 0.7762,
+      "step": 1931
+    },
+    {
+      "epoch": 1.1660126773317234,
+      "grad_norm": 0.0949690192937851,
+      "learning_rate": 0.00012602739726027398,
+      "loss": 0.6066,
+      "step": 1932
+    },
+    {
+      "epoch": 1.1666163597947479,
+      "grad_norm": 0.12045732140541077,
+      "learning_rate": 0.00012598588625985888,
+      "loss": 0.9106,
+      "step": 1933
+    },
+    {
+      "epoch": 1.1672200422577723,
+      "grad_norm": 0.10367399454116821,
+      "learning_rate": 0.00012594437525944376,
+      "loss": 0.7499,
+      "step": 1934
+    },
+    {
+      "epoch": 1.167823724720797,
+      "grad_norm": 0.09977395832538605,
+      "learning_rate": 0.00012590286425902863,
+      "loss": 0.6043,
+      "step": 1935
+    },
+    {
+      "epoch": 1.1684274071838212,
+      "grad_norm": 0.10659980773925781,
+      "learning_rate": 0.00012586135325861356,
+      "loss": 0.8781,
+      "step": 1936
+    },
+    {
+      "epoch": 1.1690310896468459,
+      "grad_norm": 0.0970248430967331,
+      "learning_rate": 0.00012581984225819843,
+      "loss": 0.6159,
+      "step": 1937
+    },
+    {
+      "epoch": 1.16963477210987,
+      "grad_norm": 0.10996770858764648,
+      "learning_rate": 0.0001257783312577833,
+      "loss": 0.7535,
+      "step": 1938
+    },
+    {
+      "epoch": 1.1702384545728948,
+      "grad_norm": 0.10588741302490234,
+      "learning_rate": 0.0001257368202573682,
+      "loss": 0.6802,
+      "step": 1939
+    },
+    {
+      "epoch": 1.170842137035919,
+      "grad_norm": 0.1041964441537857,
+      "learning_rate": 0.0001256953092569531,
+      "loss": 0.731,
+      "step": 1940
+    },
+    {
+      "epoch": 1.1714458194989437,
+      "grad_norm": 0.10040424019098282,
+      "learning_rate": 0.00012565379825653798,
+      "loss": 0.645,
+      "step": 1941
+    },
+    {
+      "epoch": 1.1720495019619679,
+      "grad_norm": 0.11066184937953949,
+      "learning_rate": 0.00012561228725612288,
+      "loss": 0.6855,
+      "step": 1942
+    },
+    {
+      "epoch": 1.1726531844249926,
+      "grad_norm": 0.10693058371543884,
+      "learning_rate": 0.00012557077625570778,
+      "loss": 0.6097,
+      "step": 1943
+    },
+    {
+      "epoch": 1.1732568668880168,
+      "grad_norm": 0.10633829981088638,
+      "learning_rate": 0.00012552926525529265,
+      "loss": 0.6503,
+      "step": 1944
+    },
+    {
+      "epoch": 1.1738605493510414,
+      "grad_norm": 0.10232945531606674,
+      "learning_rate": 0.00012548775425487755,
+      "loss": 0.6123,
+      "step": 1945
+    },
+    {
+      "epoch": 1.174464231814066,
+      "grad_norm": 0.11123668402433395,
+      "learning_rate": 0.00012544624325446245,
+      "loss": 0.574,
+      "step": 1946
+    },
+    {
+      "epoch": 1.1750679142770903,
+      "grad_norm": 0.11907467991113663,
+      "learning_rate": 0.00012540473225404732,
+      "loss": 0.6328,
+      "step": 1947
+    },
+    {
+      "epoch": 1.1756715967401148,
+      "grad_norm": 0.11772100627422333,
+      "learning_rate": 0.00012536322125363222,
+      "loss": 0.5881,
+      "step": 1948
+    },
+    {
+      "epoch": 1.1762752792031392,
+      "grad_norm": 0.11829902976751328,
+      "learning_rate": 0.0001253217102532171,
+      "loss": 0.6201,
+      "step": 1949
+    },
+    {
+      "epoch": 1.1768789616661637,
+      "grad_norm": 0.13115333020687103,
+      "learning_rate": 0.000125280199252802,
+      "loss": 0.5381,
+      "step": 1950
+    },
+    {
+      "epoch": 1.1774826441291881,
+      "grad_norm": 0.1208442747592926,
+      "learning_rate": 0.0001252386882523869,
+      "loss": 0.5937,
+      "step": 1951
+    },
+    {
+      "epoch": 1.1780863265922126,
+      "grad_norm": 0.13914504647254944,
+      "learning_rate": 0.00012519717725197177,
+      "loss": 0.5468,
+      "step": 1952
+    },
+    {
+      "epoch": 1.178690009055237,
+      "grad_norm": 0.14542445540428162,
+      "learning_rate": 0.00012515566625155667,
+      "loss": 0.5401,
+      "step": 1953
+    },
+    {
+      "epoch": 1.1792936915182615,
+      "grad_norm": 0.14165444672107697,
+      "learning_rate": 0.00012511415525114157,
+      "loss": 0.4952,
+      "step": 1954
+    },
+    {
+      "epoch": 1.179897373981286,
+      "grad_norm": 0.14939822256565094,
+      "learning_rate": 0.00012507264425072644,
+      "loss": 0.4343,
+      "step": 1955
+    },
+    {
+      "epoch": 1.1805010564443104,
+      "grad_norm": 0.14716513454914093,
+      "learning_rate": 0.00012503113325031131,
+      "loss": 0.3985,
+      "step": 1956
+    },
+    {
+      "epoch": 1.1811047389073348,
+      "grad_norm": 0.157403826713562,
+      "learning_rate": 0.00012498962224989624,
+      "loss": 0.3312,
+      "step": 1957
+    },
+    {
+      "epoch": 1.1817084213703593,
+      "grad_norm": 0.12419522553682327,
+      "learning_rate": 0.00012494811124948111,
+      "loss": 0.9177,
+      "step": 1958
+    },
+    {
+      "epoch": 1.1823121038333837,
+      "grad_norm": 0.10016956180334091,
+      "learning_rate": 0.00012490660024906601,
+      "loss": 0.8362,
+      "step": 1959
+    },
+    {
+      "epoch": 1.1829157862964081,
+      "grad_norm": 0.12086569517850876,
+      "learning_rate": 0.00012486508924865091,
+      "loss": 0.7991,
+      "step": 1960
+    },
+    {
+      "epoch": 1.1835194687594326,
+      "grad_norm": 0.11325091868638992,
+      "learning_rate": 0.0001248235782482358,
+      "loss": 0.7633,
+      "step": 1961
+    },
+    {
+      "epoch": 1.184123151222457,
+      "grad_norm": 0.11552873998880386,
+      "learning_rate": 0.0001247820672478207,
+      "loss": 0.7515,
+      "step": 1962
+    },
+    {
+      "epoch": 1.1847268336854815,
+      "grad_norm": 0.09600497037172318,
+      "learning_rate": 0.00012474055624740556,
+      "loss": 0.698,
+      "step": 1963
+    },
+    {
+      "epoch": 1.185330516148506,
+      "grad_norm": 0.0974835678935051,
+      "learning_rate": 0.00012469904524699046,
+      "loss": 0.6602,
+      "step": 1964
+    },
+    {
+      "epoch": 1.1859341986115304,
+      "grad_norm": 0.23812483251094818,
+      "learning_rate": 0.00012465753424657536,
+      "loss": 0.7079,
+      "step": 1965
+    },
+    {
+      "epoch": 1.1865378810745548,
+      "grad_norm": 0.09787070006132126,
+      "learning_rate": 0.00012461602324616023,
+      "loss": 0.752,
+      "step": 1966
+    },
+    {
+      "epoch": 1.1871415635375793,
+      "grad_norm": 0.10244850814342499,
+      "learning_rate": 0.00012457451224574513,
+      "loss": 0.8414,
+      "step": 1967
+    },
+    {
+      "epoch": 1.1877452460006037,
+      "grad_norm": 0.11027634143829346,
+      "learning_rate": 0.00012453300124533003,
+      "loss": 0.8229,
+      "step": 1968
+    },
+    {
+      "epoch": 1.1883489284636282,
+      "grad_norm": 0.1009831577539444,
+      "learning_rate": 0.0001244914902449149,
+      "loss": 0.9911,
+      "step": 1969
+    },
+    {
+      "epoch": 1.1889526109266526,
+      "grad_norm": 0.10704313963651657,
+      "learning_rate": 0.00012444997924449978,
+      "loss": 0.7212,
+      "step": 1970
+    },
+    {
+      "epoch": 1.189556293389677,
+      "grad_norm": 0.10770639777183533,
+      "learning_rate": 0.0001244084682440847,
+      "loss": 0.7613,
+      "step": 1971
+    },
+    {
+      "epoch": 1.1901599758527015,
+      "grad_norm": 0.10859589278697968,
+      "learning_rate": 0.00012436695724366958,
+      "loss": 0.8025,
+      "step": 1972
+    },
+    {
+      "epoch": 1.190763658315726,
+      "grad_norm": 0.09854567050933838,
+      "learning_rate": 0.00012432544624325445,
+      "loss": 0.6613,
+      "step": 1973
+    },
+    {
+      "epoch": 1.1913673407787504,
+      "grad_norm": 0.10922684520483017,
+      "learning_rate": 0.00012428393524283938,
+      "loss": 0.6675,
+      "step": 1974
+    },
+    {
+      "epoch": 1.1919710232417748,
+      "grad_norm": 0.11279135942459106,
+      "learning_rate": 0.00012424242424242425,
+      "loss": 0.7562,
+      "step": 1975
+    },
+    {
+      "epoch": 1.1925747057047993,
+      "grad_norm": 0.10514727234840393,
+      "learning_rate": 0.00012420091324200913,
+      "loss": 0.6954,
+      "step": 1976
+    },
+    {
+      "epoch": 1.1931783881678237,
+      "grad_norm": 0.0970945954322815,
+      "learning_rate": 0.00012415940224159403,
+      "loss": 0.8071,
+      "step": 1977
+    },
+    {
+      "epoch": 1.1937820706308482,
+      "grad_norm": 0.11609134823083878,
+      "learning_rate": 0.00012411789124117893,
+      "loss": 0.8732,
+      "step": 1978
+    },
+    {
+      "epoch": 1.1943857530938726,
+      "grad_norm": 0.10586526989936829,
+      "learning_rate": 0.0001240763802407638,
+      "loss": 0.6324,
+      "step": 1979
+    },
+    {
+      "epoch": 1.194989435556897,
+      "grad_norm": 0.11197761446237564,
+      "learning_rate": 0.0001240348692403487,
+      "loss": 0.7699,
+      "step": 1980
+    },
+    {
+      "epoch": 1.1955931180199215,
+      "grad_norm": 0.11370430141687393,
+      "learning_rate": 0.0001239933582399336,
+      "loss": 0.8104,
+      "step": 1981
+    },
+    {
+      "epoch": 1.196196800482946,
+      "grad_norm": 0.0997292697429657,
+      "learning_rate": 0.00012395184723951847,
+      "loss": 0.7494,
+      "step": 1982
+    },
+    {
+      "epoch": 1.1968004829459704,
+      "grad_norm": 0.10209706425666809,
+      "learning_rate": 0.00012391033623910337,
+      "loss": 0.6899,
+      "step": 1983
+    },
+    {
+      "epoch": 1.1974041654089949,
+      "grad_norm": 0.09956377744674683,
+      "learning_rate": 0.00012386882523868824,
+      "loss": 0.7469,
+      "step": 1984
+    },
+    {
+      "epoch": 1.1980078478720193,
+      "grad_norm": 0.12185216695070267,
+      "learning_rate": 0.00012382731423827314,
+      "loss": 1.0278,
+      "step": 1985
+    },
+    {
+      "epoch": 1.1986115303350438,
+      "grad_norm": 0.16837285459041595,
+      "learning_rate": 0.00012378580323785804,
+      "loss": 0.8619,
+      "step": 1986
+    },
+    {
+      "epoch": 1.1992152127980682,
+      "grad_norm": 0.10831692814826965,
+      "learning_rate": 0.00012374429223744292,
+      "loss": 0.6662,
+      "step": 1987
+    },
+    {
+      "epoch": 1.1998188952610926,
+      "grad_norm": 0.09968318045139313,
+      "learning_rate": 0.00012370278123702782,
+      "loss": 0.7561,
+      "step": 1988
+    },
+    {
+      "epoch": 1.200422577724117,
+      "grad_norm": 0.09389069676399231,
+      "learning_rate": 0.00012366127023661272,
+      "loss": 0.6967,
+      "step": 1989
+    },
+    {
+      "epoch": 1.2010262601871415,
+      "grad_norm": 0.10120224207639694,
+      "learning_rate": 0.0001236197592361976,
+      "loss": 0.7008,
+      "step": 1990
+    },
+    {
+      "epoch": 1.201629942650166,
+      "grad_norm": 0.1032261773943901,
+      "learning_rate": 0.0001235782482357825,
+      "loss": 0.6327,
+      "step": 1991
+    },
+    {
+      "epoch": 1.2022336251131904,
+      "grad_norm": 0.10337409377098083,
+      "learning_rate": 0.0001235367372353674,
+      "loss": 0.6639,
+      "step": 1992
+    },
+    {
+      "epoch": 1.2028373075762149,
+      "grad_norm": 0.1101367250084877,
+      "learning_rate": 0.00012349522623495226,
+      "loss": 0.6232,
+      "step": 1993
+    },
+    {
+      "epoch": 1.2034409900392393,
+      "grad_norm": 0.1051281988620758,
+      "learning_rate": 0.00012345371523453714,
+      "loss": 0.6032,
+      "step": 1994
+    },
+    {
+      "epoch": 1.2040446725022638,
+      "grad_norm": 0.10944084078073502,
+      "learning_rate": 0.00012341220423412206,
+      "loss": 0.635,
+      "step": 1995
+    },
+    {
+      "epoch": 1.2046483549652882,
+      "grad_norm": 0.10979338735342026,
+      "learning_rate": 0.00012337069323370694,
+      "loss": 0.5944,
+      "step": 1996
+    },
+    {
+      "epoch": 1.2052520374283127,
+      "grad_norm": 0.11819171160459518,
+      "learning_rate": 0.0001233291822332918,
+      "loss": 0.6785,
+      "step": 1997
+    },
+    {
+      "epoch": 1.205855719891337,
+      "grad_norm": 0.12030988931655884,
+      "learning_rate": 0.0001232876712328767,
+      "loss": 0.6699,
+      "step": 1998
+    },
+    {
+      "epoch": 1.2064594023543616,
+      "grad_norm": 0.12456244975328445,
+      "learning_rate": 0.0001232461602324616,
+      "loss": 0.6065,
+      "step": 1999
+    },
+    {
+      "epoch": 1.207063084817386,
+      "grad_norm": 0.12571309506893158,
+      "learning_rate": 0.00012320464923204648,
+      "loss": 0.5504,
+      "step": 2000
+    },
+    {
+      "epoch": 1.207063084817386,
+      "eval_loss": 0.8049178123474121,
+      "eval_runtime": 1222.2521,
+      "eval_samples_per_second": 2.283,
+      "eval_steps_per_second": 0.286,
+      "step": 2000
+    },
+    {
+      "epoch": 1.2076667672804104,
+      "grad_norm": 0.12835955619812012,
+      "learning_rate": 0.00012316313823163138,
+      "loss": 0.6229,
+      "step": 2001
+    },
+    {
+      "epoch": 1.208270449743435,
+      "grad_norm": 0.12921744585037231,
+      "learning_rate": 0.00012312162723121628,
+      "loss": 0.5465,
+      "step": 2002
+    },
+    {
+      "epoch": 1.2088741322064593,
+      "grad_norm": 0.139405757188797,
+      "learning_rate": 0.00012308011623080118,
+      "loss": 0.5361,
+      "step": 2003
+    },
+    {
+      "epoch": 1.2094778146694838,
+      "grad_norm": 0.14658528566360474,
+      "learning_rate": 0.00012303860523038605,
+      "loss": 0.4847,
+      "step": 2004
+    },
+    {
+      "epoch": 1.2100814971325082,
+      "grad_norm": 0.15193478763103485,
+      "learning_rate": 0.00012299709422997095,
+      "loss": 0.4933,
+      "step": 2005
+    },
+    {
+      "epoch": 1.2106851795955327,
+      "grad_norm": 0.15982936322689056,
+      "learning_rate": 0.00012295558322955586,
+      "loss": 0.4008,
+      "step": 2006
+    },
+    {
+      "epoch": 1.2112888620585571,
+      "grad_norm": 0.15444207191467285,
+      "learning_rate": 0.00012291407222914073,
+      "loss": 0.3094,
+      "step": 2007
+    },
+    {
+      "epoch": 1.2118925445215816,
+      "grad_norm": 0.11019429564476013,
+      "learning_rate": 0.0001228725612287256,
+      "loss": 0.808,
+      "step": 2008
+    },
+    {
+      "epoch": 1.212496226984606,
+      "grad_norm": 0.11978921294212341,
+      "learning_rate": 0.00012283105022831053,
+      "loss": 0.7044,
+      "step": 2009
+    },
+    {
+      "epoch": 1.2130999094476305,
+      "grad_norm": 0.10817541182041168,
+      "learning_rate": 0.0001227895392278954,
+      "loss": 0.7133,
+      "step": 2010
+    },
+    {
+      "epoch": 1.213703591910655,
+      "grad_norm": 0.10823164880275726,
+      "learning_rate": 0.00012274802822748027,
+      "loss": 0.7497,
+      "step": 2011
+    },
+    {
+      "epoch": 1.2143072743736794,
+      "grad_norm": 0.10769373923540115,
+      "learning_rate": 0.0001227065172270652,
+      "loss": 0.9813,
+      "step": 2012
+    },
+    {
+      "epoch": 1.2149109568367038,
+      "grad_norm": 0.11639753729104996,
+      "learning_rate": 0.00012266500622665007,
+      "loss": 0.8259,
+      "step": 2013
+    },
+    {
+      "epoch": 1.2155146392997283,
+      "grad_norm": 0.10166247934103012,
+      "learning_rate": 0.00012262349522623495,
+      "loss": 0.6783,
+      "step": 2014
+    },
+    {
+      "epoch": 1.2161183217627527,
+      "grad_norm": 0.10975832492113113,
+      "learning_rate": 0.00012258198422581985,
+      "loss": 0.8338,
+      "step": 2015
+    },
+    {
+      "epoch": 1.2167220042257774,
+      "grad_norm": 0.12426348030567169,
+      "learning_rate": 0.00012254047322540475,
+      "loss": 0.7086,
+      "step": 2016
+    },
+    {
+      "epoch": 1.2173256866888016,
+      "grad_norm": 0.09932038187980652,
+      "learning_rate": 0.00012249896222498962,
+      "loss": 0.7193,
+      "step": 2017
+    },
+    {
+      "epoch": 1.2179293691518263,
+      "grad_norm": 0.11152757704257965,
+      "learning_rate": 0.00012245745122457452,
+      "loss": 0.7668,
+      "step": 2018
+    },
+    {
+      "epoch": 1.2185330516148505,
+      "grad_norm": 0.1197136715054512,
+      "learning_rate": 0.00012241594022415942,
+      "loss": 1.1513,
+      "step": 2019
+    },
+    {
+      "epoch": 1.2191367340778752,
+      "grad_norm": 0.12218881398439407,
+      "learning_rate": 0.0001223744292237443,
+      "loss": 0.5988,
+      "step": 2020
+    },
+    {
+      "epoch": 1.2197404165408994,
+      "grad_norm": 0.10389691591262817,
+      "learning_rate": 0.0001223329182233292,
+      "loss": 0.7079,
+      "step": 2021
+    },
+    {
+      "epoch": 1.220344099003924,
+      "grad_norm": 0.10646260529756546,
+      "learning_rate": 0.00012229140722291407,
+      "loss": 0.6731,
+      "step": 2022
+    },
+    {
+      "epoch": 1.2209477814669483,
+      "grad_norm": 0.11292260885238647,
+      "learning_rate": 0.00012224989622249897,
+      "loss": 0.7529,
+      "step": 2023
+    },
+    {
+      "epoch": 1.221551463929973,
+      "grad_norm": 0.11751693487167358,
+      "learning_rate": 0.00012220838522208387,
+      "loss": 0.873,
+      "step": 2024
+    },
+    {
+      "epoch": 1.2221551463929972,
+      "grad_norm": 0.10794904083013535,
+      "learning_rate": 0.00012216687422166874,
+      "loss": 0.8017,
+      "step": 2025
+    },
+    {
+      "epoch": 1.2227588288560218,
+      "grad_norm": 0.10329126566648483,
+      "learning_rate": 0.00012212536322125364,
+      "loss": 0.8863,
+      "step": 2026
+    },
+    {
+      "epoch": 1.2233625113190463,
+      "grad_norm": 0.10396264493465424,
+      "learning_rate": 0.00012208385222083854,
+      "loss": 0.6399,
+      "step": 2027
+    },
+    {
+      "epoch": 1.2239661937820707,
+      "grad_norm": 0.12621258199214935,
+      "learning_rate": 0.00012204234122042341,
+      "loss": 0.7813,
+      "step": 2028
+    },
+    {
+      "epoch": 1.2245698762450952,
+      "grad_norm": 0.11045660823583603,
+      "learning_rate": 0.0001220008302200083,
+      "loss": 0.7743,
+      "step": 2029
+    },
+    {
+      "epoch": 1.2251735587081196,
+      "grad_norm": 0.11149439960718155,
+      "learning_rate": 0.0001219593192195932,
+      "loss": 0.788,
+      "step": 2030
+    },
+    {
+      "epoch": 1.225777241171144,
+      "grad_norm": 0.11239415407180786,
+      "learning_rate": 0.00012191780821917808,
+      "loss": 1.0028,
+      "step": 2031
+    },
+    {
+      "epoch": 1.2263809236341685,
+      "grad_norm": 0.11321971565485,
+      "learning_rate": 0.00012187629721876297,
+      "loss": 0.7841,
+      "step": 2032
+    },
+    {
+      "epoch": 1.226984606097193,
+      "grad_norm": 0.10027159005403519,
+      "learning_rate": 0.00012183478621834787,
+      "loss": 0.728,
+      "step": 2033
+    },
+    {
+      "epoch": 1.2275882885602174,
+      "grad_norm": 0.09943227469921112,
+      "learning_rate": 0.00012179327521793276,
+      "loss": 0.7397,
+      "step": 2034
+    },
+    {
+      "epoch": 1.2281919710232418,
+      "grad_norm": 0.09892190247774124,
+      "learning_rate": 0.00012175176421751764,
+      "loss": 0.806,
+      "step": 2035
+    },
+    {
+      "epoch": 1.2287956534862663,
+      "grad_norm": 0.10850216448307037,
+      "learning_rate": 0.00012171025321710254,
+      "loss": 0.8292,
+      "step": 2036
+    },
+    {
+      "epoch": 1.2293993359492907,
+      "grad_norm": 0.102396160364151,
+      "learning_rate": 0.00012166874221668743,
+      "loss": 0.7506,
+      "step": 2037
+    },
+    {
+      "epoch": 1.2300030184123152,
+      "grad_norm": 0.12359751015901566,
+      "learning_rate": 0.0001216272312162723,
+      "loss": 0.6433,
+      "step": 2038
+    },
+    {
+      "epoch": 1.2306067008753396,
+      "grad_norm": 0.10474561899900436,
+      "learning_rate": 0.00012158572021585722,
+      "loss": 0.7758,
+      "step": 2039
+    },
+    {
+      "epoch": 1.231210383338364,
+      "grad_norm": 0.10111495852470398,
+      "learning_rate": 0.00012154420921544209,
+      "loss": 0.6707,
+      "step": 2040
+    },
+    {
+      "epoch": 1.2318140658013885,
+      "grad_norm": 0.10861996561288834,
+      "learning_rate": 0.00012150269821502698,
+      "loss": 0.6847,
+      "step": 2041
+    },
+    {
+      "epoch": 1.232417748264413,
+      "grad_norm": 0.10541880130767822,
+      "learning_rate": 0.00012146118721461188,
+      "loss": 0.6607,
+      "step": 2042
+    },
+    {
+      "epoch": 1.2330214307274374,
+      "grad_norm": 0.11882983893156052,
+      "learning_rate": 0.00012141967621419676,
+      "loss": 0.6679,
+      "step": 2043
+    },
+    {
+      "epoch": 1.2336251131904619,
+      "grad_norm": 0.10306091606616974,
+      "learning_rate": 0.00012137816521378165,
+      "loss": 0.6325,
+      "step": 2044
+    },
+    {
+      "epoch": 1.2342287956534863,
+      "grad_norm": 0.11227353662252426,
+      "learning_rate": 0.00012133665421336655,
+      "loss": 0.6578,
+      "step": 2045
+    },
+    {
+      "epoch": 1.2348324781165108,
+      "grad_norm": 0.11392517387866974,
+      "learning_rate": 0.00012129514321295144,
+      "loss": 0.6137,
+      "step": 2046
+    },
+    {
+      "epoch": 1.2354361605795352,
+      "grad_norm": 0.117860347032547,
+      "learning_rate": 0.00012125363221253634,
+      "loss": 0.6464,
+      "step": 2047
+    },
+    {
+      "epoch": 1.2360398430425596,
+      "grad_norm": 0.11536695808172226,
+      "learning_rate": 0.00012121212121212122,
+      "loss": 0.5613,
+      "step": 2048
+    },
+    {
+      "epoch": 1.236643525505584,
+      "grad_norm": 0.12926429510116577,
+      "learning_rate": 0.00012117061021170611,
+      "loss": 0.588,
+      "step": 2049
+    },
+    {
+      "epoch": 1.2372472079686085,
+      "grad_norm": 0.12854550778865814,
+      "learning_rate": 0.00012112909921129101,
+      "loss": 0.643,
+      "step": 2050
+    },
+    {
+      "epoch": 1.237850890431633,
+      "grad_norm": 0.12910380959510803,
+      "learning_rate": 0.0001210875882108759,
+      "loss": 0.6016,
+      "step": 2051
+    },
+    {
+      "epoch": 1.2384545728946574,
+      "grad_norm": 0.1356167048215866,
+      "learning_rate": 0.00012104607721046077,
+      "loss": 0.5593,
+      "step": 2052
+    },
+    {
+      "epoch": 1.2390582553576819,
+      "grad_norm": 0.13904690742492676,
+      "learning_rate": 0.00012100456621004568,
+      "loss": 0.5238,
+      "step": 2053
+    },
+    {
+      "epoch": 1.2396619378207063,
+      "grad_norm": 0.14844262599945068,
+      "learning_rate": 0.00012096305520963056,
+      "loss": 0.5047,
+      "step": 2054
+    },
+    {
+      "epoch": 1.2402656202837308,
+      "grad_norm": 0.14652326703071594,
+      "learning_rate": 0.00012092154420921544,
+      "loss": 0.4741,
+      "step": 2055
+    },
+    {
+      "epoch": 1.2408693027467552,
+      "grad_norm": 0.15764220058918,
+      "learning_rate": 0.00012088003320880034,
+      "loss": 0.3982,
+      "step": 2056
+    },
+    {
+      "epoch": 1.2414729852097797,
+      "grad_norm": 0.146280899643898,
+      "learning_rate": 0.00012083852220838523,
+      "loss": 0.3012,
+      "step": 2057
+    },
+    {
+      "epoch": 1.2420766676728041,
+      "grad_norm": 0.13532495498657227,
+      "learning_rate": 0.00012079701120797011,
+      "loss": 0.7918,
+      "step": 2058
+    },
+    {
+      "epoch": 1.2426803501358286,
+      "grad_norm": 0.11222409456968307,
+      "learning_rate": 0.00012075550020755501,
+      "loss": 0.7279,
+      "step": 2059
+    },
+    {
+      "epoch": 1.243284032598853,
+      "grad_norm": 0.12232893705368042,
+      "learning_rate": 0.0001207139892071399,
+      "loss": 0.8694,
+      "step": 2060
+    },
+    {
+      "epoch": 1.2438877150618775,
+      "grad_norm": 0.11951932311058044,
+      "learning_rate": 0.00012067247820672479,
+      "loss": 0.6634,
+      "step": 2061
+    },
+    {
+      "epoch": 1.244491397524902,
+      "grad_norm": 0.10638052970170975,
+      "learning_rate": 0.00012063096720630969,
+      "loss": 0.7437,
+      "step": 2062
+    },
+    {
+      "epoch": 1.2450950799879263,
+      "grad_norm": 0.10910008102655411,
+      "learning_rate": 0.00012058945620589457,
+      "loss": 0.8246,
+      "step": 2063
+    },
+    {
+      "epoch": 1.2456987624509508,
+      "grad_norm": 0.12002308666706085,
+      "learning_rate": 0.00012054794520547945,
+      "loss": 0.8136,
+      "step": 2064
+    },
+    {
+      "epoch": 1.2463024449139752,
+      "grad_norm": 0.10318660736083984,
+      "learning_rate": 0.00012050643420506436,
+      "loss": 0.9517,
+      "step": 2065
+    },
+    {
+      "epoch": 1.2469061273769997,
+      "grad_norm": 0.10524383932352066,
+      "learning_rate": 0.00012046492320464923,
+      "loss": 0.8019,
+      "step": 2066
+    },
+    {
+      "epoch": 1.2475098098400241,
+      "grad_norm": 0.10615126043558121,
+      "learning_rate": 0.00012042341220423412,
+      "loss": 0.732,
+      "step": 2067
+    },
+    {
+      "epoch": 1.2481134923030486,
+      "grad_norm": 0.10820569843053818,
+      "learning_rate": 0.00012038190120381902,
+      "loss": 0.9747,
+      "step": 2068
+    },
+    {
+      "epoch": 1.248717174766073,
+      "grad_norm": 0.10168500244617462,
+      "learning_rate": 0.0001203403902034039,
+      "loss": 0.7427,
+      "step": 2069
+    },
+    {
+      "epoch": 1.2493208572290975,
+      "grad_norm": 0.11007564514875412,
+      "learning_rate": 0.0001202988792029888,
+      "loss": 0.6818,
+      "step": 2070
+    },
+    {
+      "epoch": 1.249924539692122,
+      "grad_norm": 0.10033000260591507,
+      "learning_rate": 0.0001202573682025737,
+      "loss": 0.7134,
+      "step": 2071
+    },
+    {
+      "epoch": 1.2505282221551464,
+      "grad_norm": 0.11808007955551147,
+      "learning_rate": 0.00012021585720215858,
+      "loss": 0.7718,
+      "step": 2072
+    },
+    {
+      "epoch": 1.2511319046181708,
+      "grad_norm": 0.11681666970252991,
+      "learning_rate": 0.00012017434620174347,
+      "loss": 0.7327,
+      "step": 2073
+    },
+    {
+      "epoch": 1.2517355870811953,
+      "grad_norm": 0.11521414667367935,
+      "learning_rate": 0.00012013283520132837,
+      "loss": 0.754,
+      "step": 2074
+    },
+    {
+      "epoch": 1.2523392695442197,
+      "grad_norm": 0.09546779096126556,
+      "learning_rate": 0.00012009132420091325,
+      "loss": 0.7399,
+      "step": 2075
+    },
+    {
+      "epoch": 1.2529429520072441,
+      "grad_norm": 0.10306226462125778,
+      "learning_rate": 0.00012004981320049813,
+      "loss": 0.682,
+      "step": 2076
+    },
+    {
+      "epoch": 1.2535466344702686,
+      "grad_norm": 0.11089253425598145,
+      "learning_rate": 0.00012000830220008304,
+      "loss": 0.868,
+      "step": 2077
+    },
+    {
+      "epoch": 1.254150316933293,
+      "grad_norm": 0.12641939520835876,
+      "learning_rate": 0.00011996679119966791,
+      "loss": 0.8043,
+      "step": 2078
+    },
+    {
+      "epoch": 1.2547539993963175,
+      "grad_norm": 0.11239389330148697,
+      "learning_rate": 0.0001199252801992528,
+      "loss": 0.6766,
+      "step": 2079
+    },
+    {
+      "epoch": 1.255357681859342,
+      "grad_norm": 0.11025764048099518,
+      "learning_rate": 0.0001198837691988377,
+      "loss": 0.7516,
+      "step": 2080
+    },
+    {
+      "epoch": 1.2559613643223664,
+      "grad_norm": 0.10319259762763977,
+      "learning_rate": 0.00011984225819842259,
+      "loss": 0.791,
+      "step": 2081
+    },
+    {
+      "epoch": 1.2565650467853908,
+      "grad_norm": 0.1244928389787674,
+      "learning_rate": 0.00011980074719800747,
+      "loss": 0.8266,
+      "step": 2082
+    },
+    {
+      "epoch": 1.2571687292484153,
+      "grad_norm": 0.09494119137525558,
+      "learning_rate": 0.00011975923619759237,
+      "loss": 0.8277,
+      "step": 2083
+    },
+    {
+      "epoch": 1.2577724117114397,
+      "grad_norm": 0.11088497191667557,
+      "learning_rate": 0.00011971772519717726,
+      "loss": 0.7471,
+      "step": 2084
+    },
+    {
+      "epoch": 1.2583760941744642,
+      "grad_norm": 0.10349318385124207,
+      "learning_rate": 0.00011967621419676213,
+      "loss": 0.8844,
+      "step": 2085
+    },
+    {
+      "epoch": 1.2589797766374886,
+      "grad_norm": 0.09668376296758652,
+      "learning_rate": 0.00011963470319634704,
+      "loss": 0.7249,
+      "step": 2086
+    },
+    {
+      "epoch": 1.259583459100513,
+      "grad_norm": 0.11646587401628494,
+      "learning_rate": 0.00011959319219593193,
+      "loss": 0.9458,
+      "step": 2087
+    },
+    {
+      "epoch": 1.2601871415635375,
+      "grad_norm": 0.11103519797325134,
+      "learning_rate": 0.0001195516811955168,
+      "loss": 0.8437,
+      "step": 2088
+    },
+    {
+      "epoch": 1.260790824026562,
+      "grad_norm": 0.12291695177555084,
+      "learning_rate": 0.00011951017019510172,
+      "loss": 0.8271,
+      "step": 2089
+    },
+    {
+      "epoch": 1.2613945064895864,
+      "grad_norm": 0.10509349405765533,
+      "learning_rate": 0.00011946865919468659,
+      "loss": 0.7225,
+      "step": 2090
+    },
+    {
+      "epoch": 1.2619981889526108,
+      "grad_norm": 0.09915387630462646,
+      "learning_rate": 0.0001194271481942715,
+      "loss": 0.5739,
+      "step": 2091
+    },
+    {
+      "epoch": 1.2626018714156353,
+      "grad_norm": 0.10708332806825638,
+      "learning_rate": 0.00011938563719385638,
+      "loss": 0.6239,
+      "step": 2092
+    },
+    {
+      "epoch": 1.26320555387866,
+      "grad_norm": 0.10963302105665207,
+      "learning_rate": 0.00011934412619344126,
+      "loss": 0.6454,
+      "step": 2093
+    },
+    {
+      "epoch": 1.2638092363416842,
+      "grad_norm": 0.10599739104509354,
+      "learning_rate": 0.00011930261519302616,
+      "loss": 0.6203,
+      "step": 2094
+    },
+    {
+      "epoch": 1.2644129188047089,
+      "grad_norm": 0.11157579720020294,
+      "learning_rate": 0.00011926110419261105,
+      "loss": 0.689,
+      "step": 2095
+    },
+    {
+      "epoch": 1.265016601267733,
+      "grad_norm": 0.11258841305971146,
+      "learning_rate": 0.00011921959319219594,
+      "loss": 0.5814,
+      "step": 2096
+    },
+    {
+      "epoch": 1.2656202837307577,
+      "grad_norm": 0.11933767050504684,
+      "learning_rate": 0.00011917808219178084,
+      "loss": 0.6181,
+      "step": 2097
+    },
+    {
+      "epoch": 1.266223966193782,
+      "grad_norm": 0.1172049269080162,
+      "learning_rate": 0.00011913657119136572,
+      "loss": 0.6099,
+      "step": 2098
+    },
+    {
+      "epoch": 1.2668276486568066,
+      "grad_norm": 0.1134132593870163,
+      "learning_rate": 0.0001190950601909506,
+      "loss": 0.6055,
+      "step": 2099
+    },
+    {
+      "epoch": 1.2674313311198309,
+      "grad_norm": 0.12405924499034882,
+      "learning_rate": 0.00011905354919053551,
+      "loss": 0.5902,
+      "step": 2100
+    },
+    {
+      "epoch": 1.2680350135828555,
+      "grad_norm": 0.13423797488212585,
+      "learning_rate": 0.0001190120381901204,
+      "loss": 0.6814,
+      "step": 2101
+    },
+    {
+      "epoch": 1.2686386960458798,
+      "grad_norm": 0.13356977701187134,
+      "learning_rate": 0.00011897052718970527,
+      "loss": 0.6092,
+      "step": 2102
+    },
+    {
+      "epoch": 1.2692423785089044,
+      "grad_norm": 0.13142277300357819,
+      "learning_rate": 0.00011892901618929018,
+      "loss": 0.4849,
+      "step": 2103
+    },
+    {
+      "epoch": 1.2698460609719286,
+      "grad_norm": 0.1493036448955536,
+      "learning_rate": 0.00011888750518887506,
+      "loss": 0.5734,
+      "step": 2104
+    },
+    {
+      "epoch": 1.2704497434349533,
+      "grad_norm": 0.14654318988323212,
+      "learning_rate": 0.00011884599418845994,
+      "loss": 0.4648,
+      "step": 2105
+    },
+    {
+      "epoch": 1.2710534258979775,
+      "grad_norm": 0.1532881259918213,
+      "learning_rate": 0.00011880448318804484,
+      "loss": 0.3987,
+      "step": 2106
+    },
+    {
+      "epoch": 1.2716571083610022,
+      "grad_norm": 0.16462919116020203,
+      "learning_rate": 0.00011876297218762973,
+      "loss": 0.3358,
+      "step": 2107
+    },
+    {
+      "epoch": 1.2722607908240264,
+      "grad_norm": 0.11268685758113861,
+      "learning_rate": 0.00011872146118721462,
+      "loss": 0.6965,
+      "step": 2108
+    },
+    {
+      "epoch": 1.272864473287051,
+      "grad_norm": 0.11173847317695618,
+      "learning_rate": 0.00011867995018679952,
+      "loss": 0.9126,
+      "step": 2109
+    },
+    {
+      "epoch": 1.2734681557500753,
+      "grad_norm": 0.10636541247367859,
+      "learning_rate": 0.0001186384391863844,
+      "loss": 0.8093,
+      "step": 2110
+    },
+    {
+      "epoch": 1.2740718382131,
+      "grad_norm": 0.11559230089187622,
+      "learning_rate": 0.00011859692818596927,
+      "loss": 0.7025,
+      "step": 2111
+    },
+    {
+      "epoch": 1.2746755206761244,
+      "grad_norm": 0.11961853504180908,
+      "learning_rate": 0.00011855541718555419,
+      "loss": 0.7638,
+      "step": 2112
+    },
+    {
+      "epoch": 1.2752792031391489,
+      "grad_norm": 0.11241693049669266,
+      "learning_rate": 0.00011851390618513906,
+      "loss": 0.7422,
+      "step": 2113
+    },
+    {
+      "epoch": 1.2758828856021733,
+      "grad_norm": 0.11459380388259888,
+      "learning_rate": 0.00011847239518472395,
+      "loss": 1.0325,
+      "step": 2114
+    },
+    {
+      "epoch": 1.2764865680651978,
+      "grad_norm": 0.1107303723692894,
+      "learning_rate": 0.00011843088418430886,
+      "loss": 0.6566,
+      "step": 2115
+    },
+    {
+      "epoch": 1.2770902505282222,
+      "grad_norm": 0.11201652884483337,
+      "learning_rate": 0.00011838937318389373,
+      "loss": 0.7305,
+      "step": 2116
+    },
+    {
+      "epoch": 1.2776939329912467,
+      "grad_norm": 0.1064920499920845,
+      "learning_rate": 0.00011834786218347862,
+      "loss": 0.6963,
+      "step": 2117
+    },
+    {
+      "epoch": 1.2782976154542711,
+      "grad_norm": 0.11608701199293137,
+      "learning_rate": 0.00011830635118306352,
+      "loss": 0.7162,
+      "step": 2118
+    },
+    {
+      "epoch": 1.2789012979172956,
+      "grad_norm": 0.10472101718187332,
+      "learning_rate": 0.00011826484018264841,
+      "loss": 0.683,
+      "step": 2119
+    },
+    {
+      "epoch": 1.27950498038032,
+      "grad_norm": 0.12958820164203644,
+      "learning_rate": 0.0001182233291822333,
+      "loss": 0.8274,
+      "step": 2120
+    },
+    {
+      "epoch": 1.2801086628433445,
+      "grad_norm": 0.1009947806596756,
+      "learning_rate": 0.0001181818181818182,
+      "loss": 0.6998,
+      "step": 2121
+    },
+    {
+      "epoch": 1.280712345306369,
+      "grad_norm": 0.11905679106712341,
+      "learning_rate": 0.00011814030718140308,
+      "loss": 0.7364,
+      "step": 2122
+    },
+    {
+      "epoch": 1.2813160277693934,
+      "grad_norm": 0.10903461277484894,
+      "learning_rate": 0.00011809879618098795,
+      "loss": 0.7088,
+      "step": 2123
+    },
+    {
+      "epoch": 1.2819197102324178,
+      "grad_norm": 0.11340314149856567,
+      "learning_rate": 0.00011805728518057287,
+      "loss": 0.8662,
+      "step": 2124
+    },
+    {
+      "epoch": 1.2825233926954422,
+      "grad_norm": 0.10403577238321304,
+      "learning_rate": 0.00011801577418015774,
+      "loss": 0.8047,
+      "step": 2125
+    },
+    {
+      "epoch": 1.2831270751584667,
+      "grad_norm": 0.12345468252897263,
+      "learning_rate": 0.00011797426317974263,
+      "loss": 0.9449,
+      "step": 2126
+    },
+    {
+      "epoch": 1.2837307576214911,
+      "grad_norm": 0.11308707296848297,
+      "learning_rate": 0.00011793275217932753,
+      "loss": 0.7631,
+      "step": 2127
+    },
+    {
+      "epoch": 1.2843344400845156,
+      "grad_norm": 0.10067742317914963,
+      "learning_rate": 0.00011789124117891241,
+      "loss": 0.7624,
+      "step": 2128
+    },
+    {
+      "epoch": 1.28493812254754,
+      "grad_norm": 0.11228195577859879,
+      "learning_rate": 0.0001178497301784973,
+      "loss": 1.1616,
+      "step": 2129
+    },
+    {
+      "epoch": 1.2855418050105645,
+      "grad_norm": 0.10720618814229965,
+      "learning_rate": 0.0001178082191780822,
+      "loss": 1.3342,
+      "step": 2130
+    },
+    {
+      "epoch": 1.286145487473589,
+      "grad_norm": 0.10839608311653137,
+      "learning_rate": 0.00011776670817766709,
+      "loss": 0.9353,
+      "step": 2131
+    },
+    {
+      "epoch": 1.2867491699366134,
+      "grad_norm": 0.10751800239086151,
+      "learning_rate": 0.00011772519717725197,
+      "loss": 0.7667,
+      "step": 2132
+    },
+    {
+      "epoch": 1.2873528523996378,
+      "grad_norm": 0.10154768079519272,
+      "learning_rate": 0.00011768368617683687,
+      "loss": 0.794,
+      "step": 2133
+    },
+    {
+      "epoch": 1.2879565348626623,
+      "grad_norm": 0.10651742666959763,
+      "learning_rate": 0.00011764217517642176,
+      "loss": 0.7428,
+      "step": 2134
+    },
+    {
+      "epoch": 1.2885602173256867,
+      "grad_norm": 0.1084497794508934,
+      "learning_rate": 0.00011760066417600663,
+      "loss": 0.7594,
+      "step": 2135
+    },
+    {
+      "epoch": 1.2891638997887112,
+      "grad_norm": 0.11117968708276749,
+      "learning_rate": 0.00011755915317559155,
+      "loss": 0.7812,
+      "step": 2136
+    },
+    {
+      "epoch": 1.2897675822517356,
+      "grad_norm": 0.112984798848629,
+      "learning_rate": 0.00011751764217517642,
+      "loss": 0.6982,
+      "step": 2137
+    },
+    {
+      "epoch": 1.29037126471476,
+      "grad_norm": 0.10472138971090317,
+      "learning_rate": 0.00011747613117476133,
+      "loss": 0.6542,
+      "step": 2138
+    },
+    {
+      "epoch": 1.2909749471777845,
+      "grad_norm": 0.12380527704954147,
+      "learning_rate": 0.0001174346201743462,
+      "loss": 0.7346,
+      "step": 2139
+    },
+    {
+      "epoch": 1.291578629640809,
+      "grad_norm": 0.10141772031784058,
+      "learning_rate": 0.00011739310917393109,
+      "loss": 0.6896,
+      "step": 2140
+    },
+    {
+      "epoch": 1.2921823121038334,
+      "grad_norm": 0.10354811698198318,
+      "learning_rate": 0.00011735159817351599,
+      "loss": 0.6743,
+      "step": 2141
+    },
+    {
+      "epoch": 1.2927859945668578,
+      "grad_norm": 0.10275600850582123,
+      "learning_rate": 0.00011731008717310088,
+      "loss": 0.6862,
+      "step": 2142
+    },
+    {
+      "epoch": 1.2933896770298823,
+      "grad_norm": 0.10765067487955093,
+      "learning_rate": 0.00011726857617268576,
+      "loss": 0.6372,
+      "step": 2143
+    },
+    {
+      "epoch": 1.2939933594929067,
+      "grad_norm": 0.10741368681192398,
+      "learning_rate": 0.00011722706517227066,
+      "loss": 0.6173,
+      "step": 2144
+    },
+    {
+      "epoch": 1.2945970419559312,
+      "grad_norm": 0.11568807065486908,
+      "learning_rate": 0.00011718555417185555,
+      "loss": 0.7006,
+      "step": 2145
+    },
+    {
+      "epoch": 1.2952007244189556,
+      "grad_norm": 0.12116066366434097,
+      "learning_rate": 0.00011714404317144044,
+      "loss": 0.678,
+      "step": 2146
+    },
+    {
+      "epoch": 1.29580440688198,
+      "grad_norm": 0.1318727433681488,
+      "learning_rate": 0.00011710253217102534,
+      "loss": 0.6605,
+      "step": 2147
+    },
+    {
+      "epoch": 1.2964080893450045,
+      "grad_norm": 0.1201576441526413,
+      "learning_rate": 0.00011706102117061022,
+      "loss": 0.6411,
+      "step": 2148
+    },
+    {
+      "epoch": 1.297011771808029,
+      "grad_norm": 0.12639600038528442,
+      "learning_rate": 0.0001170195101701951,
+      "loss": 0.6679,
+      "step": 2149
+    },
+    {
+      "epoch": 1.2976154542710534,
+      "grad_norm": 0.12824396789073944,
+      "learning_rate": 0.00011697799916978001,
+      "loss": 0.5963,
+      "step": 2150
+    },
+    {
+      "epoch": 1.2982191367340778,
+      "grad_norm": 0.1317749321460724,
+      "learning_rate": 0.00011693648816936488,
+      "loss": 0.6115,
+      "step": 2151
+    },
+    {
+      "epoch": 1.2988228191971023,
+      "grad_norm": 0.14059890806674957,
+      "learning_rate": 0.00011689497716894977,
+      "loss": 0.5661,
+      "step": 2152
+    },
+    {
+      "epoch": 1.2994265016601267,
+      "grad_norm": 0.14957548677921295,
+      "learning_rate": 0.00011685346616853467,
+      "loss": 0.616,
+      "step": 2153
+    },
+    {
+      "epoch": 1.3000301841231512,
+      "grad_norm": 0.15705609321594238,
+      "learning_rate": 0.00011681195516811956,
+      "loss": 0.5566,
+      "step": 2154
+    },
+    {
+      "epoch": 1.3006338665861756,
+      "grad_norm": 0.14872965216636658,
+      "learning_rate": 0.00011677044416770444,
+      "loss": 0.4887,
+      "step": 2155
+    },
+    {
+      "epoch": 1.3012375490492,
+      "grad_norm": 0.1517983227968216,
+      "learning_rate": 0.00011672893316728934,
+      "loss": 0.4465,
+      "step": 2156
+    },
+    {
+      "epoch": 1.3018412315122245,
+      "grad_norm": 0.15609890222549438,
+      "learning_rate": 0.00011668742216687423,
+      "loss": 0.4065,
+      "step": 2157
+    },
+    {
+      "epoch": 1.302444913975249,
+      "grad_norm": 0.11183691769838333,
+      "learning_rate": 0.00011664591116645912,
+      "loss": 0.6092,
+      "step": 2158
+    },
+    {
+      "epoch": 1.3030485964382734,
+      "grad_norm": 0.10590233653783798,
+      "learning_rate": 0.00011660440016604402,
+      "loss": 0.9179,
+      "step": 2159
+    },
+    {
+      "epoch": 1.3036522789012979,
+      "grad_norm": 0.11257438361644745,
+      "learning_rate": 0.0001165628891656289,
+      "loss": 0.7511,
+      "step": 2160
+    },
+    {
+      "epoch": 1.3042559613643223,
+      "grad_norm": 0.11735875904560089,
+      "learning_rate": 0.00011652137816521377,
+      "loss": 0.8342,
+      "step": 2161
+    },
+    {
+      "epoch": 1.3048596438273468,
+      "grad_norm": 0.12115509063005447,
+      "learning_rate": 0.00011647986716479869,
+      "loss": 0.6921,
+      "step": 2162
+    },
+    {
+      "epoch": 1.3054633262903712,
+      "grad_norm": 0.10376877337694168,
+      "learning_rate": 0.00011643835616438356,
+      "loss": 0.5839,
+      "step": 2163
+    },
+    {
+      "epoch": 1.3060670087533957,
+      "grad_norm": 0.10715942084789276,
+      "learning_rate": 0.00011639684516396845,
+      "loss": 0.7511,
+      "step": 2164
+    },
+    {
+      "epoch": 1.30667069121642,
+      "grad_norm": 0.1192198321223259,
+      "learning_rate": 0.00011635533416355335,
+      "loss": 0.7023,
+      "step": 2165
+    },
+    {
+      "epoch": 1.3072743736794445,
+      "grad_norm": 0.10451126098632812,
+      "learning_rate": 0.00011631382316313823,
+      "loss": 0.642,
+      "step": 2166
+    },
+    {
+      "epoch": 1.307878056142469,
+      "grad_norm": 0.10327422618865967,
+      "learning_rate": 0.00011627231216272312,
+      "loss": 0.7792,
+      "step": 2167
+    },
+    {
+      "epoch": 1.3084817386054934,
+      "grad_norm": 0.10710068792104721,
+      "learning_rate": 0.00011623080116230802,
+      "loss": 0.705,
+      "step": 2168
+    },
+    {
+      "epoch": 1.3090854210685179,
+      "grad_norm": 0.10589960962533951,
+      "learning_rate": 0.00011618929016189291,
+      "loss": 0.6591,
+      "step": 2169
+    },
+    {
+      "epoch": 1.3096891035315423,
+      "grad_norm": 0.10254230350255966,
+      "learning_rate": 0.0001161477791614778,
+      "loss": 0.723,
+      "step": 2170
+    },
+    {
+      "epoch": 1.3102927859945668,
+      "grad_norm": 0.10744060575962067,
+      "learning_rate": 0.0001161062681610627,
+      "loss": 0.7218,
+      "step": 2171
+    },
+    {
+      "epoch": 1.3108964684575912,
+      "grad_norm": 0.11566371470689774,
+      "learning_rate": 0.00011606475716064758,
+      "loss": 0.6746,
+      "step": 2172
+    },
+    {
+      "epoch": 1.3115001509206157,
+      "grad_norm": 0.10865778475999832,
+      "learning_rate": 0.00011602324616023245,
+      "loss": 0.706,
+      "step": 2173
+    },
+    {
+      "epoch": 1.3121038333836403,
+      "grad_norm": 0.11026459187269211,
+      "learning_rate": 0.00011598173515981737,
+      "loss": 0.7467,
+      "step": 2174
+    },
+    {
+      "epoch": 1.3127075158466646,
+      "grad_norm": 0.1640511453151703,
+      "learning_rate": 0.00011594022415940224,
+      "loss": 0.754,
+      "step": 2175
+    },
+    {
+      "epoch": 1.3133111983096892,
+      "grad_norm": 0.11436708271503448,
+      "learning_rate": 0.00011589871315898713,
+      "loss": 0.8731,
+      "step": 2176
+    },
+    {
+      "epoch": 1.3139148807727135,
+      "grad_norm": 0.11194545775651932,
+      "learning_rate": 0.00011585720215857203,
+      "loss": 0.8425,
+      "step": 2177
+    },
+    {
+      "epoch": 1.3145185632357381,
+      "grad_norm": 0.10926983505487442,
+      "learning_rate": 0.00011581569115815691,
+      "loss": 0.695,
+      "step": 2178
+    },
+    {
+      "epoch": 1.3151222456987623,
+      "grad_norm": 0.11056378483772278,
+      "learning_rate": 0.0001157741801577418,
+      "loss": 0.7545,
+      "step": 2179
+    },
+    {
+      "epoch": 1.315725928161787,
+      "grad_norm": 0.11646270751953125,
+      "learning_rate": 0.0001157326691573267,
+      "loss": 0.697,
+      "step": 2180
+    },
+    {
+      "epoch": 1.3163296106248112,
+      "grad_norm": 0.096307173371315,
+      "learning_rate": 0.00011569115815691159,
+      "loss": 0.7044,
+      "step": 2181
+    },
+    {
+      "epoch": 1.316933293087836,
+      "grad_norm": 0.10772639513015747,
+      "learning_rate": 0.00011564964715649649,
+      "loss": 0.6897,
+      "step": 2182
+    },
+    {
+      "epoch": 1.3175369755508601,
+      "grad_norm": 0.10433688759803772,
+      "learning_rate": 0.00011560813615608137,
+      "loss": 0.7438,
+      "step": 2183
+    },
+    {
+      "epoch": 1.3181406580138848,
+      "grad_norm": 0.10850182920694351,
+      "learning_rate": 0.00011556662515566626,
+      "loss": 0.712,
+      "step": 2184
+    },
+    {
+      "epoch": 1.318744340476909,
+      "grad_norm": 0.11556683480739594,
+      "learning_rate": 0.00011552511415525116,
+      "loss": 0.7897,
+      "step": 2185
+    },
+    {
+      "epoch": 1.3193480229399337,
+      "grad_norm": 0.11677438765764236,
+      "learning_rate": 0.00011548360315483605,
+      "loss": 0.9773,
+      "step": 2186
+    },
+    {
+      "epoch": 1.319951705402958,
+      "grad_norm": 0.10189380496740341,
+      "learning_rate": 0.00011544209215442092,
+      "loss": 0.7935,
+      "step": 2187
+    },
+    {
+      "epoch": 1.3205553878659826,
+      "grad_norm": 0.10953009128570557,
+      "learning_rate": 0.00011540058115400583,
+      "loss": 0.7774,
+      "step": 2188
+    },
+    {
+      "epoch": 1.3211590703290068,
+      "grad_norm": 0.10907643288373947,
+      "learning_rate": 0.0001153590701535907,
+      "loss": 0.6565,
+      "step": 2189
+    },
+    {
+      "epoch": 1.3217627527920315,
+      "grad_norm": 0.10476800054311752,
+      "learning_rate": 0.00011531755915317559,
+      "loss": 0.6466,
+      "step": 2190
+    },
+    {
+      "epoch": 1.3223664352550557,
+      "grad_norm": 0.11148995906114578,
+      "learning_rate": 0.00011527604815276049,
+      "loss": 0.5893,
+      "step": 2191
+    },
+    {
+      "epoch": 1.3229701177180804,
+      "grad_norm": 0.10962202399969101,
+      "learning_rate": 0.00011523453715234538,
+      "loss": 0.6037,
+      "step": 2192
+    },
+    {
+      "epoch": 1.3235738001811048,
+      "grad_norm": 0.11205828934907913,
+      "learning_rate": 0.00011519302615193026,
+      "loss": 0.6462,
+      "step": 2193
+    },
+    {
+      "epoch": 1.3241774826441293,
+      "grad_norm": 0.11916244029998779,
+      "learning_rate": 0.00011515151515151516,
+      "loss": 0.6124,
+      "step": 2194
+    },
+    {
+      "epoch": 1.3247811651071537,
+      "grad_norm": 0.11426917463541031,
+      "learning_rate": 0.00011511000415110005,
+      "loss": 0.6288,
+      "step": 2195
+    },
+    {
+      "epoch": 1.3253848475701782,
+      "grad_norm": 0.11942315101623535,
+      "learning_rate": 0.00011506849315068494,
+      "loss": 0.6146,
+      "step": 2196
+    },
+    {
+      "epoch": 1.3259885300332026,
+      "grad_norm": 0.1271522343158722,
+      "learning_rate": 0.00011502698215026984,
+      "loss": 0.6174,
+      "step": 2197
+    },
+    {
+      "epoch": 1.326592212496227,
+      "grad_norm": 0.12273116409778595,
+      "learning_rate": 0.00011498547114985472,
+      "loss": 0.6165,
+      "step": 2198
+    },
+    {
+      "epoch": 1.3271958949592515,
+      "grad_norm": 0.12854193150997162,
+      "learning_rate": 0.0001149439601494396,
+      "loss": 0.6234,
+      "step": 2199
+    },
+    {
+      "epoch": 1.327799577422276,
+      "grad_norm": 0.12989364564418793,
+      "learning_rate": 0.00011490244914902451,
+      "loss": 0.5936,
+      "step": 2200
+    },
+    {
+      "epoch": 1.3284032598853004,
+      "grad_norm": 0.13210608065128326,
+      "learning_rate": 0.00011486093814860938,
+      "loss": 0.5591,
+      "step": 2201
+    },
+    {
+      "epoch": 1.3290069423483248,
+      "grad_norm": 0.1452115774154663,
+      "learning_rate": 0.00011481942714819427,
+      "loss": 0.5945,
+      "step": 2202
+    },
+    {
+      "epoch": 1.3296106248113493,
+      "grad_norm": 0.14136159420013428,
+      "learning_rate": 0.00011477791614777917,
+      "loss": 0.5191,
+      "step": 2203
+    },
+    {
+      "epoch": 1.3302143072743737,
+      "grad_norm": 0.1478489339351654,
+      "learning_rate": 0.00011473640514736406,
+      "loss": 0.4752,
+      "step": 2204
+    },
+    {
+      "epoch": 1.3308179897373982,
+      "grad_norm": 0.15822367370128632,
+      "learning_rate": 0.00011469489414694894,
+      "loss": 0.4601,
+      "step": 2205
+    },
+    {
+      "epoch": 1.3314216722004226,
+      "grad_norm": 0.1557387262582779,
+      "learning_rate": 0.00011465338314653384,
+      "loss": 0.4091,
+      "step": 2206
+    },
+    {
+      "epoch": 1.332025354663447,
+      "grad_norm": 0.15302599966526031,
+      "learning_rate": 0.00011461187214611873,
+      "loss": 0.3174,
+      "step": 2207
+    },
+    {
+      "epoch": 1.3326290371264715,
+      "grad_norm": 0.10820896923542023,
+      "learning_rate": 0.0001145703611457036,
+      "loss": 0.852,
+      "step": 2208
+    },
+    {
+      "epoch": 1.333232719589496,
+      "grad_norm": 0.10538190603256226,
+      "learning_rate": 0.00011452885014528852,
+      "loss": 0.6777,
+      "step": 2209
+    },
+    {
+      "epoch": 1.3338364020525204,
+      "grad_norm": 0.11455769091844559,
+      "learning_rate": 0.0001144873391448734,
+      "loss": 0.7658,
+      "step": 2210
+    },
+    {
+      "epoch": 1.3344400845155449,
+      "grad_norm": 0.12761631608009338,
+      "learning_rate": 0.00011444582814445828,
+      "loss": 0.7603,
+      "step": 2211
+    },
+    {
+      "epoch": 1.3350437669785693,
+      "grad_norm": 0.12457510828971863,
+      "learning_rate": 0.00011440431714404319,
+      "loss": 0.8002,
+      "step": 2212
+    },
+    {
+      "epoch": 1.3356474494415937,
+      "grad_norm": 0.11467002332210541,
+      "learning_rate": 0.00011436280614362806,
+      "loss": 0.7013,
+      "step": 2213
+    },
+    {
+      "epoch": 1.3362511319046182,
+      "grad_norm": 0.10846070200204849,
+      "learning_rate": 0.00011432129514321295,
+      "loss": 0.8422,
+      "step": 2214
+    },
+    {
+      "epoch": 1.3368548143676426,
+      "grad_norm": 0.11062701046466827,
+      "learning_rate": 0.00011427978414279785,
+      "loss": 0.691,
+      "step": 2215
+    },
+    {
+      "epoch": 1.337458496830667,
+      "grad_norm": 0.10322251170873642,
+      "learning_rate": 0.00011423827314238273,
+      "loss": 0.7036,
+      "step": 2216
+    },
+    {
+      "epoch": 1.3380621792936915,
+      "grad_norm": 0.09952748566865921,
+      "learning_rate": 0.00011419676214196762,
+      "loss": 0.6875,
+      "step": 2217
+    },
+    {
+      "epoch": 1.338665861756716,
+      "grad_norm": 0.11490236967802048,
+      "learning_rate": 0.00011415525114155252,
+      "loss": 0.6921,
+      "step": 2218
+    },
+    {
+      "epoch": 1.3392695442197404,
+      "grad_norm": 0.1203780472278595,
+      "learning_rate": 0.00011411374014113741,
+      "loss": 0.7415,
+      "step": 2219
+    },
+    {
+      "epoch": 1.3398732266827649,
+      "grad_norm": 0.11478865891695023,
+      "learning_rate": 0.00011407222914072228,
+      "loss": 0.7476,
+      "step": 2220
+    },
+    {
+      "epoch": 1.3404769091457893,
+      "grad_norm": 0.10916067659854889,
+      "learning_rate": 0.0001140307181403072,
+      "loss": 0.7539,
+      "step": 2221
+    },
+    {
+      "epoch": 1.3410805916088138,
+      "grad_norm": 0.11967265605926514,
+      "learning_rate": 0.00011398920713989207,
+      "loss": 0.834,
+      "step": 2222
+    },
+    {
+      "epoch": 1.3416842740718382,
+      "grad_norm": 0.12789706885814667,
+      "learning_rate": 0.00011394769613947695,
+      "loss": 0.7649,
+      "step": 2223
+    },
+    {
+      "epoch": 1.3422879565348627,
+      "grad_norm": 0.11024732142686844,
+      "learning_rate": 0.00011390618513906187,
+      "loss": 0.8246,
+      "step": 2224
+    },
+    {
+      "epoch": 1.342891638997887,
+      "grad_norm": 0.10574118793010712,
+      "learning_rate": 0.00011386467413864674,
+      "loss": 0.6421,
+      "step": 2225
+    },
+    {
+      "epoch": 1.3434953214609116,
+      "grad_norm": 0.1084161251783371,
+      "learning_rate": 0.00011382316313823165,
+      "loss": 0.8431,
+      "step": 2226
+    },
+    {
+      "epoch": 1.344099003923936,
+      "grad_norm": 0.112635038793087,
+      "learning_rate": 0.00011378165213781653,
+      "loss": 0.7057,
+      "step": 2227
+    },
+    {
+      "epoch": 1.3447026863869604,
+      "grad_norm": 0.11063949763774872,
+      "learning_rate": 0.00011374014113740141,
+      "loss": 0.7764,
+      "step": 2228
+    },
+    {
+      "epoch": 1.345306368849985,
+      "grad_norm": 0.1124119758605957,
+      "learning_rate": 0.00011369863013698631,
+      "loss": 0.9151,
+      "step": 2229
+    },
+    {
+      "epoch": 1.3459100513130093,
+      "grad_norm": 0.1197538748383522,
+      "learning_rate": 0.0001136571191365712,
+      "loss": 0.6366,
+      "step": 2230
+    },
+    {
+      "epoch": 1.3465137337760338,
+      "grad_norm": 0.1143086701631546,
+      "learning_rate": 0.00011361560813615609,
+      "loss": 1.0755,
+      "step": 2231
+    },
+    {
+      "epoch": 1.3471174162390582,
+      "grad_norm": 0.1160961166024208,
+      "learning_rate": 0.00011357409713574099,
+      "loss": 0.7346,
+      "step": 2232
+    },
+    {
+      "epoch": 1.3477210987020827,
+      "grad_norm": 0.1389804482460022,
+      "learning_rate": 0.00011353258613532587,
+      "loss": 0.7935,
+      "step": 2233
+    },
+    {
+      "epoch": 1.3483247811651071,
+      "grad_norm": 0.10732017457485199,
+      "learning_rate": 0.00011349107513491075,
+      "loss": 1.1968,
+      "step": 2234
+    },
+    {
+      "epoch": 1.3489284636281316,
+      "grad_norm": 0.10339702665805817,
+      "learning_rate": 0.00011344956413449566,
+      "loss": 0.674,
+      "step": 2235
+    },
+    {
+      "epoch": 1.349532146091156,
+      "grad_norm": 0.10818582028150558,
+      "learning_rate": 0.00011340805313408053,
+      "loss": 0.7626,
+      "step": 2236
+    },
+    {
+      "epoch": 1.3501358285541805,
+      "grad_norm": 0.1044488251209259,
+      "learning_rate": 0.00011336654213366542,
+      "loss": 0.719,
+      "step": 2237
+    },
+    {
+      "epoch": 1.350739511017205,
+      "grad_norm": 0.20112541317939758,
+      "learning_rate": 0.00011332503113325033,
+      "loss": 0.5892,
+      "step": 2238
+    },
+    {
+      "epoch": 1.3513431934802294,
+      "grad_norm": 0.10670458525419235,
+      "learning_rate": 0.0001132835201328352,
+      "loss": 0.6436,
+      "step": 2239
+    },
+    {
+      "epoch": 1.3519468759432538,
+      "grad_norm": 0.11629335582256317,
+      "learning_rate": 0.00011324200913242009,
+      "loss": 1.0747,
+      "step": 2240
+    },
+    {
+      "epoch": 1.3525505584062782,
+      "grad_norm": 0.14042681455612183,
+      "learning_rate": 0.00011320049813200499,
+      "loss": 0.6851,
+      "step": 2241
+    },
+    {
+      "epoch": 1.3531542408693027,
+      "grad_norm": 0.10726059973239899,
+      "learning_rate": 0.00011315898713158988,
+      "loss": 0.6776,
+      "step": 2242
+    },
+    {
+      "epoch": 1.3537579233323271,
+      "grad_norm": 0.10431487113237381,
+      "learning_rate": 0.00011311747613117476,
+      "loss": 0.5565,
+      "step": 2243
+    },
+    {
+      "epoch": 1.3543616057953516,
+      "grad_norm": 0.11104051023721695,
+      "learning_rate": 0.00011307596513075966,
+      "loss": 0.6523,
+      "step": 2244
+    },
+    {
+      "epoch": 1.354965288258376,
+      "grad_norm": 0.1104440838098526,
+      "learning_rate": 0.00011303445413034455,
+      "loss": 0.6657,
+      "step": 2245
+    },
+    {
+      "epoch": 1.3555689707214005,
+      "grad_norm": 0.11780980974435806,
+      "learning_rate": 0.00011299294312992942,
+      "loss": 0.6437,
+      "step": 2246
+    },
+    {
+      "epoch": 1.356172653184425,
+      "grad_norm": 0.11362163722515106,
+      "learning_rate": 0.00011295143212951434,
+      "loss": 0.6607,
+      "step": 2247
+    },
+    {
+      "epoch": 1.3567763356474494,
+      "grad_norm": 0.12305210530757904,
+      "learning_rate": 0.00011290992112909921,
+      "loss": 0.6784,
+      "step": 2248
+    },
+    {
+      "epoch": 1.3573800181104738,
+      "grad_norm": 0.12743736803531647,
+      "learning_rate": 0.0001128684101286841,
+      "loss": 0.6501,
+      "step": 2249
+    },
+    {
+      "epoch": 1.3579837005734983,
+      "grad_norm": 0.12665079534053802,
+      "learning_rate": 0.000112826899128269,
+      "loss": 0.6922,
+      "step": 2250
+    },
+    {
+      "epoch": 1.3585873830365227,
+      "grad_norm": 0.12639720737934113,
+      "learning_rate": 0.00011278538812785388,
+      "loss": 0.5699,
+      "step": 2251
+    },
+    {
+      "epoch": 1.3591910654995472,
+      "grad_norm": 0.12889716029167175,
+      "learning_rate": 0.00011274387712743877,
+      "loss": 0.5657,
+      "step": 2252
+    },
+    {
+      "epoch": 1.3597947479625718,
+      "grad_norm": 0.14307141304016113,
+      "learning_rate": 0.00011270236612702367,
+      "loss": 0.5307,
+      "step": 2253
+    },
+    {
+      "epoch": 1.360398430425596,
+      "grad_norm": 0.13598784804344177,
+      "learning_rate": 0.00011266085512660856,
+      "loss": 0.4857,
+      "step": 2254
+    },
+    {
+      "epoch": 1.3610021128886207,
+      "grad_norm": 0.150955468416214,
+      "learning_rate": 0.00011261934412619344,
+      "loss": 0.4596,
+      "step": 2255
+    },
+    {
+      "epoch": 1.361605795351645,
+      "grad_norm": 0.14729629456996918,
+      "learning_rate": 0.00011257783312577834,
+      "loss": 0.3831,
+      "step": 2256
+    },
+    {
+      "epoch": 1.3622094778146696,
+      "grad_norm": 0.15139052271842957,
+      "learning_rate": 0.00011253632212536323,
+      "loss": 0.3392,
+      "step": 2257
+    },
+    {
+      "epoch": 1.3628131602776938,
+      "grad_norm": 0.10566865652799606,
+      "learning_rate": 0.0001124948111249481,
+      "loss": 0.6181,
+      "step": 2258
+    },
+    {
+      "epoch": 1.3634168427407185,
+      "grad_norm": 0.1218709796667099,
+      "learning_rate": 0.00011245330012453302,
+      "loss": 0.7976,
+      "step": 2259
+    },
+    {
+      "epoch": 1.3640205252037427,
+      "grad_norm": 0.11592400819063187,
+      "learning_rate": 0.00011241178912411789,
+      "loss": 0.8621,
+      "step": 2260
+    },
+    {
+      "epoch": 1.3646242076667674,
+      "grad_norm": 0.1122564896941185,
+      "learning_rate": 0.00011237027812370278,
+      "loss": 0.6828,
+      "step": 2261
+    },
+    {
+      "epoch": 1.3652278901297916,
+      "grad_norm": 0.1253998577594757,
+      "learning_rate": 0.00011232876712328768,
+      "loss": 0.7367,
+      "step": 2262
+    },
+    {
+      "epoch": 1.3658315725928163,
+      "grad_norm": 0.11441956460475922,
+      "learning_rate": 0.00011228725612287256,
+      "loss": 0.7129,
+      "step": 2263
+    },
+    {
+      "epoch": 1.3664352550558405,
+      "grad_norm": 0.18277932703495026,
+      "learning_rate": 0.00011224574512245745,
+      "loss": 0.7095,
+      "step": 2264
+    },
+    {
+      "epoch": 1.3670389375188652,
+      "grad_norm": 0.10515805333852768,
+      "learning_rate": 0.00011220423412204235,
+      "loss": 0.6439,
+      "step": 2265
+    },
+    {
+      "epoch": 1.3676426199818894,
+      "grad_norm": 0.10004261136054993,
+      "learning_rate": 0.00011216272312162724,
+      "loss": 0.7665,
+      "step": 2266
+    },
+    {
+      "epoch": 1.368246302444914,
+      "grad_norm": 0.1079486608505249,
+      "learning_rate": 0.00011212121212121212,
+      "loss": 0.6854,
+      "step": 2267
+    },
+    {
+      "epoch": 1.3688499849079383,
+      "grad_norm": 0.11932551115751266,
+      "learning_rate": 0.00011207970112079702,
+      "loss": 0.7683,
+      "step": 2268
+    },
+    {
+      "epoch": 1.369453667370963,
+      "grad_norm": 0.13196608424186707,
+      "learning_rate": 0.00011203819012038191,
+      "loss": 0.8149,
+      "step": 2269
+    },
+    {
+      "epoch": 1.3700573498339872,
+      "grad_norm": 0.1305970698595047,
+      "learning_rate": 0.00011199667911996681,
+      "loss": 0.8266,
+      "step": 2270
+    },
+    {
+      "epoch": 1.3706610322970119,
+      "grad_norm": 0.10973062366247177,
+      "learning_rate": 0.0001119551681195517,
+      "loss": 0.6498,
+      "step": 2271
+    },
+    {
+      "epoch": 1.371264714760036,
+      "grad_norm": 0.11326909810304642,
+      "learning_rate": 0.00011191365711913657,
+      "loss": 0.7068,
+      "step": 2272
+    },
+    {
+      "epoch": 1.3718683972230608,
+      "grad_norm": 0.1138983964920044,
+      "learning_rate": 0.00011187214611872148,
+      "loss": 0.8917,
+      "step": 2273
+    },
+    {
+      "epoch": 1.3724720796860852,
+      "grad_norm": 0.10338523238897324,
+      "learning_rate": 0.00011183063511830635,
+      "loss": 0.6048,
+      "step": 2274
+    },
+    {
+      "epoch": 1.3730757621491096,
+      "grad_norm": 0.13740071654319763,
+      "learning_rate": 0.00011178912411789124,
+      "loss": 0.7794,
+      "step": 2275
+    },
+    {
+      "epoch": 1.373679444612134,
+      "grad_norm": 0.11318357288837433,
+      "learning_rate": 0.00011174761311747614,
+      "loss": 0.7224,
+      "step": 2276
+    },
+    {
+      "epoch": 1.3742831270751585,
+      "grad_norm": 0.10989907383918762,
+      "learning_rate": 0.00011170610211706103,
+      "loss": 0.7604,
+      "step": 2277
+    },
+    {
+      "epoch": 1.374886809538183,
+      "grad_norm": 0.11369361728429794,
+      "learning_rate": 0.00011166459111664591,
+      "loss": 0.6629,
+      "step": 2278
+    },
+    {
+      "epoch": 1.3754904920012074,
+      "grad_norm": 0.11444233357906342,
+      "learning_rate": 0.00011162308011623081,
+      "loss": 0.6644,
+      "step": 2279
+    },
+    {
+      "epoch": 1.3760941744642319,
+      "grad_norm": 0.11095111072063446,
+      "learning_rate": 0.0001115815691158157,
+      "loss": 0.8084,
+      "step": 2280
+    },
+    {
+      "epoch": 1.3766978569272563,
+      "grad_norm": 0.11327888071537018,
+      "learning_rate": 0.00011154005811540059,
+      "loss": 0.9121,
+      "step": 2281
+    },
+    {
+      "epoch": 1.3773015393902808,
+      "grad_norm": 0.11659125983715057,
+      "learning_rate": 0.00011149854711498549,
+      "loss": 0.7475,
+      "step": 2282
+    },
+    {
+      "epoch": 1.3779052218533052,
+      "grad_norm": 0.11424364149570465,
+      "learning_rate": 0.00011145703611457037,
+      "loss": 0.9011,
+      "step": 2283
+    },
+    {
+      "epoch": 1.3785089043163297,
+      "grad_norm": 0.11273805052042007,
+      "learning_rate": 0.00011141552511415525,
+      "loss": 0.6689,
+      "step": 2284
+    },
+    {
+      "epoch": 1.379112586779354,
+      "grad_norm": 0.1099931076169014,
+      "learning_rate": 0.00011137401411374016,
+      "loss": 0.7201,
+      "step": 2285
+    },
+    {
+      "epoch": 1.3797162692423786,
+      "grad_norm": 0.11520043015480042,
+      "learning_rate": 0.00011133250311332503,
+      "loss": 0.6714,
+      "step": 2286
+    },
+    {
+      "epoch": 1.380319951705403,
+      "grad_norm": 0.10492314398288727,
+      "learning_rate": 0.00011129099211290992,
+      "loss": 0.8564,
+      "step": 2287
+    },
+    {
+      "epoch": 1.3809236341684274,
+      "grad_norm": 0.10921221226453781,
+      "learning_rate": 0.00011124948111249482,
+      "loss": 0.7088,
+      "step": 2288
+    },
+    {
+      "epoch": 1.381527316631452,
+      "grad_norm": 0.11280515789985657,
+      "learning_rate": 0.0001112079701120797,
+      "loss": 1.1316,
+      "step": 2289
+    },
+    {
+      "epoch": 1.3821309990944763,
+      "grad_norm": 0.10194700956344604,
+      "learning_rate": 0.00011116645911166459,
+      "loss": 1.0434,
+      "step": 2290
+    },
+    {
+      "epoch": 1.3827346815575008,
+      "grad_norm": 0.11088079959154129,
+      "learning_rate": 0.00011112494811124949,
+      "loss": 0.663,
+      "step": 2291
+    },
+    {
+      "epoch": 1.3833383640205252,
+      "grad_norm": 0.10625745356082916,
+      "learning_rate": 0.00011108343711083438,
+      "loss": 0.6609,
+      "step": 2292
+    },
+    {
+      "epoch": 1.3839420464835497,
+      "grad_norm": 0.11164417862892151,
+      "learning_rate": 0.00011104192611041927,
+      "loss": 0.646,
+      "step": 2293
+    },
+    {
+      "epoch": 1.3845457289465741,
+      "grad_norm": 0.11619970202445984,
+      "learning_rate": 0.00011100041511000417,
+      "loss": 0.6225,
+      "step": 2294
+    },
+    {
+      "epoch": 1.3851494114095986,
+      "grad_norm": 0.11362091451883316,
+      "learning_rate": 0.00011095890410958905,
+      "loss": 0.6535,
+      "step": 2295
+    },
+    {
+      "epoch": 1.385753093872623,
+      "grad_norm": 0.11212702840566635,
+      "learning_rate": 0.00011091739310917392,
+      "loss": 0.585,
+      "step": 2296
+    },
+    {
+      "epoch": 1.3863567763356475,
+      "grad_norm": 0.12629151344299316,
+      "learning_rate": 0.00011087588210875884,
+      "loss": 0.6642,
+      "step": 2297
+    },
+    {
+      "epoch": 1.386960458798672,
+      "grad_norm": 0.12351097911596298,
+      "learning_rate": 0.00011083437110834371,
+      "loss": 0.6866,
+      "step": 2298
+    },
+    {
+      "epoch": 1.3875641412616964,
+      "grad_norm": 0.11870943009853363,
+      "learning_rate": 0.0001107928601079286,
+      "loss": 0.6017,
+      "step": 2299
+    },
+    {
+      "epoch": 1.3881678237247208,
+      "grad_norm": 0.12148743122816086,
+      "learning_rate": 0.0001107513491075135,
+      "loss": 0.574,
+      "step": 2300
+    },
+    {
+      "epoch": 1.3887715061877453,
+      "grad_norm": 0.1272771656513214,
+      "learning_rate": 0.00011070983810709838,
+      "loss": 0.5676,
+      "step": 2301
+    },
+    {
+      "epoch": 1.3893751886507697,
+      "grad_norm": 0.13474443554878235,
+      "learning_rate": 0.00011066832710668327,
+      "loss": 0.5647,
+      "step": 2302
+    },
+    {
+      "epoch": 1.3899788711137941,
+      "grad_norm": 0.16455814242362976,
+      "learning_rate": 0.00011062681610626817,
+      "loss": 0.513,
+      "step": 2303
+    },
+    {
+      "epoch": 1.3905825535768186,
+      "grad_norm": 0.15098370611667633,
+      "learning_rate": 0.00011058530510585306,
+      "loss": 0.5227,
+      "step": 2304
+    },
+    {
+      "epoch": 1.391186236039843,
+      "grad_norm": 0.14636565744876862,
+      "learning_rate": 0.00011054379410543793,
+      "loss": 0.4745,
+      "step": 2305
+    },
+    {
+      "epoch": 1.3917899185028675,
+      "grad_norm": 0.15192949771881104,
+      "learning_rate": 0.00011050228310502284,
+      "loss": 0.4152,
+      "step": 2306
+    },
+    {
+      "epoch": 1.392393600965892,
+      "grad_norm": 0.15223491191864014,
+      "learning_rate": 0.00011046077210460773,
+      "loss": 0.3253,
+      "step": 2307
+    },
+    {
+      "epoch": 1.3929972834289164,
+      "grad_norm": 0.10637323558330536,
+      "learning_rate": 0.0001104192611041926,
+      "loss": 0.6033,
+      "step": 2308
+    },
+    {
+      "epoch": 1.3936009658919408,
+      "grad_norm": 0.11009783297777176,
+      "learning_rate": 0.00011037775010377752,
+      "loss": 0.6934,
+      "step": 2309
+    },
+    {
+      "epoch": 1.3942046483549653,
+      "grad_norm": 0.10806847363710403,
+      "learning_rate": 0.00011033623910336239,
+      "loss": 0.6502,
+      "step": 2310
+    },
+    {
+      "epoch": 1.3948083308179897,
+      "grad_norm": 0.11237577348947525,
+      "learning_rate": 0.00011029472810294728,
+      "loss": 0.684,
+      "step": 2311
+    },
+    {
+      "epoch": 1.3954120132810142,
+      "grad_norm": 0.15474088490009308,
+      "learning_rate": 0.00011025321710253218,
+      "loss": 0.7796,
+      "step": 2312
+    },
+    {
+      "epoch": 1.3960156957440386,
+      "grad_norm": 0.10955344140529633,
+      "learning_rate": 0.00011021170610211706,
+      "loss": 0.6851,
+      "step": 2313
+    },
+    {
+      "epoch": 1.396619378207063,
+      "grad_norm": 0.12131255120038986,
+      "learning_rate": 0.00011017019510170196,
+      "loss": 0.7527,
+      "step": 2314
+    },
+    {
+      "epoch": 1.3972230606700875,
+      "grad_norm": 0.10648848861455917,
+      "learning_rate": 0.00011012868410128685,
+      "loss": 0.7382,
+      "step": 2315
+    },
+    {
+      "epoch": 1.397826743133112,
+      "grad_norm": 0.10780826210975647,
+      "learning_rate": 0.00011008717310087174,
+      "loss": 0.873,
+      "step": 2316
+    },
+    {
+      "epoch": 1.3984304255961364,
+      "grad_norm": 0.11691659688949585,
+      "learning_rate": 0.00011004566210045664,
+      "loss": 1.0688,
+      "step": 2317
+    },
+    {
+      "epoch": 1.3990341080591608,
+      "grad_norm": 0.1263553500175476,
+      "learning_rate": 0.00011000415110004152,
+      "loss": 0.7616,
+      "step": 2318
+    },
+    {
+      "epoch": 1.3996377905221853,
+      "grad_norm": 0.1272163987159729,
+      "learning_rate": 0.0001099626400996264,
+      "loss": 0.8121,
+      "step": 2319
+    },
+    {
+      "epoch": 1.4002414729852097,
+      "grad_norm": 0.267022967338562,
+      "learning_rate": 0.00010992112909921131,
+      "loss": 0.8719,
+      "step": 2320
+    },
+    {
+      "epoch": 1.4008451554482342,
+      "grad_norm": 0.10211081057786942,
+      "learning_rate": 0.0001098796180987962,
+      "loss": 1.0426,
+      "step": 2321
+    },
+    {
+      "epoch": 1.4014488379112586,
+      "grad_norm": 0.1085173562169075,
+      "learning_rate": 0.00010983810709838107,
+      "loss": 0.6862,
+      "step": 2322
+    },
+    {
+      "epoch": 1.402052520374283,
+      "grad_norm": 0.1208515390753746,
+      "learning_rate": 0.00010979659609796598,
+      "loss": 0.7391,
+      "step": 2323
+    },
+    {
+      "epoch": 1.4026562028373075,
+      "grad_norm": 0.10681469738483429,
+      "learning_rate": 0.00010975508509755085,
+      "loss": 0.7787,
+      "step": 2324
+    },
+    {
+      "epoch": 1.403259885300332,
+      "grad_norm": 0.1286618411540985,
+      "learning_rate": 0.00010971357409713574,
+      "loss": 0.7521,
+      "step": 2325
+    },
+    {
+      "epoch": 1.4038635677633564,
+      "grad_norm": 0.10427875816822052,
+      "learning_rate": 0.00010967206309672064,
+      "loss": 0.7142,
+      "step": 2326
+    },
+    {
+      "epoch": 1.4044672502263809,
+      "grad_norm": 0.10928861796855927,
+      "learning_rate": 0.00010963055209630553,
+      "loss": 0.8745,
+      "step": 2327
+    },
+    {
+      "epoch": 1.4050709326894053,
+      "grad_norm": 0.09785252064466476,
+      "learning_rate": 0.00010958904109589041,
+      "loss": 0.6134,
+      "step": 2328
+    },
+    {
+      "epoch": 1.4056746151524298,
+      "grad_norm": 0.10797867923974991,
+      "learning_rate": 0.00010954753009547531,
+      "loss": 1.0501,
+      "step": 2329
+    },
+    {
+      "epoch": 1.4062782976154542,
+      "grad_norm": 0.10527419298887253,
+      "learning_rate": 0.0001095060190950602,
+      "loss": 0.7023,
+      "step": 2330
+    },
+    {
+      "epoch": 1.4068819800784786,
+      "grad_norm": 0.10159947723150253,
+      "learning_rate": 0.00010946450809464507,
+      "loss": 0.733,
+      "step": 2331
+    },
+    {
+      "epoch": 1.407485662541503,
+      "grad_norm": 0.1160796582698822,
+      "learning_rate": 0.00010942299709422999,
+      "loss": 0.6167,
+      "step": 2332
+    },
+    {
+      "epoch": 1.4080893450045275,
+      "grad_norm": 0.10533631592988968,
+      "learning_rate": 0.00010938148609381486,
+      "loss": 0.9244,
+      "step": 2333
+    },
+    {
+      "epoch": 1.4086930274675522,
+      "grad_norm": 0.1092139482498169,
+      "learning_rate": 0.00010933997509339975,
+      "loss": 0.7907,
+      "step": 2334
+    },
+    {
+      "epoch": 1.4092967099305764,
+      "grad_norm": 0.12767547369003296,
+      "learning_rate": 0.00010929846409298466,
+      "loss": 0.6244,
+      "step": 2335
+    },
+    {
+      "epoch": 1.409900392393601,
+      "grad_norm": 0.1770038902759552,
+      "learning_rate": 0.00010925695309256953,
+      "loss": 0.8401,
+      "step": 2336
+    },
+    {
+      "epoch": 1.4105040748566253,
+      "grad_norm": 0.10718080401420593,
+      "learning_rate": 0.00010921544209215442,
+      "loss": 0.6228,
+      "step": 2337
+    },
+    {
+      "epoch": 1.41110775731965,
+      "grad_norm": 0.1158752515912056,
+      "learning_rate": 0.00010917393109173932,
+      "loss": 0.7027,
+      "step": 2338
+    },
+    {
+      "epoch": 1.4117114397826742,
+      "grad_norm": 0.11484018713235855,
+      "learning_rate": 0.0001091324200913242,
+      "loss": 0.7482,
+      "step": 2339
+    },
+    {
+      "epoch": 1.4123151222456989,
+      "grad_norm": 0.10645020753145218,
+      "learning_rate": 0.00010909090909090909,
+      "loss": 0.7191,
+      "step": 2340
+    },
+    {
+      "epoch": 1.412918804708723,
+      "grad_norm": 0.11423823237419128,
+      "learning_rate": 0.00010904939809049399,
+      "loss": 0.6348,
+      "step": 2341
+    },
+    {
+      "epoch": 1.4135224871717478,
+      "grad_norm": 0.10967587679624557,
+      "learning_rate": 0.00010900788709007888,
+      "loss": 0.6462,
+      "step": 2342
+    },
+    {
+      "epoch": 1.414126169634772,
+      "grad_norm": 0.1130829006433487,
+      "learning_rate": 0.00010896637608966375,
+      "loss": 0.6599,
+      "step": 2343
+    },
+    {
+      "epoch": 1.4147298520977967,
+      "grad_norm": 0.10537155717611313,
+      "learning_rate": 0.00010892486508924867,
+      "loss": 0.6092,
+      "step": 2344
+    },
+    {
+      "epoch": 1.415333534560821,
+      "grad_norm": 0.11640435457229614,
+      "learning_rate": 0.00010888335408883354,
+      "loss": 0.6485,
+      "step": 2345
+    },
+    {
+      "epoch": 1.4159372170238456,
+      "grad_norm": 0.13120651245117188,
+      "learning_rate": 0.00010884184308841842,
+      "loss": 0.8094,
+      "step": 2346
+    },
+    {
+      "epoch": 1.4165408994868698,
+      "grad_norm": 0.12212494760751724,
+      "learning_rate": 0.00010880033208800332,
+      "loss": 0.642,
+      "step": 2347
+    },
+    {
+      "epoch": 1.4171445819498945,
+      "grad_norm": 0.12393233925104141,
+      "learning_rate": 0.00010875882108758821,
+      "loss": 0.6366,
+      "step": 2348
+    },
+    {
+      "epoch": 1.4177482644129187,
+      "grad_norm": 0.13876676559448242,
+      "learning_rate": 0.0001087173100871731,
+      "loss": 0.6859,
+      "step": 2349
+    },
+    {
+      "epoch": 1.4183519468759433,
+      "grad_norm": 0.13311347365379333,
+      "learning_rate": 0.000108675799086758,
+      "loss": 0.6349,
+      "step": 2350
+    },
+    {
+      "epoch": 1.4189556293389676,
+      "grad_norm": 0.13839563727378845,
+      "learning_rate": 0.00010863428808634288,
+      "loss": 0.5701,
+      "step": 2351
+    },
+    {
+      "epoch": 1.4195593118019922,
+      "grad_norm": 0.14265376329421997,
+      "learning_rate": 0.00010859277708592777,
+      "loss": 0.5708,
+      "step": 2352
+    },
+    {
+      "epoch": 1.4201629942650165,
+      "grad_norm": 0.1401066929101944,
+      "learning_rate": 0.00010855126608551267,
+      "loss": 0.5612,
+      "step": 2353
+    },
+    {
+      "epoch": 1.4207666767280411,
+      "grad_norm": 0.14841662347316742,
+      "learning_rate": 0.00010850975508509756,
+      "loss": 0.5377,
+      "step": 2354
+    },
+    {
+      "epoch": 1.4213703591910656,
+      "grad_norm": 0.15106201171875,
+      "learning_rate": 0.00010846824408468243,
+      "loss": 0.5121,
+      "step": 2355
+    },
+    {
+      "epoch": 1.42197404165409,
+      "grad_norm": 0.15050731599330902,
+      "learning_rate": 0.00010842673308426734,
+      "loss": 0.3922,
+      "step": 2356
+    },
+    {
+      "epoch": 1.4225777241171145,
+      "grad_norm": 0.1529148668050766,
+      "learning_rate": 0.00010838522208385222,
+      "loss": 0.3001,
+      "step": 2357
+    },
+    {
+      "epoch": 1.423181406580139,
+      "grad_norm": 0.10441195219755173,
+      "learning_rate": 0.0001083437110834371,
+      "loss": 0.694,
+      "step": 2358
+    },
+    {
+      "epoch": 1.4237850890431634,
+      "grad_norm": 0.11137279868125916,
+      "learning_rate": 0.000108302200083022,
+      "loss": 0.9643,
+      "step": 2359
+    },
+    {
+      "epoch": 1.4243887715061878,
+      "grad_norm": 0.12313207983970642,
+      "learning_rate": 0.00010826068908260689,
+      "loss": 0.6247,
+      "step": 2360
+    },
+    {
+      "epoch": 1.4249924539692123,
+      "grad_norm": 0.11363215744495392,
+      "learning_rate": 0.00010821917808219179,
+      "loss": 0.7187,
+      "step": 2361
+    },
+    {
+      "epoch": 1.4255961364322367,
+      "grad_norm": 0.11889949440956116,
+      "learning_rate": 0.00010817766708177668,
+      "loss": 0.9015,
+      "step": 2362
+    },
+    {
+      "epoch": 1.4261998188952612,
+      "grad_norm": 0.1457277536392212,
+      "learning_rate": 0.00010813615608136156,
+      "loss": 0.7921,
+      "step": 2363
+    },
+    {
+      "epoch": 1.4268035013582856,
+      "grad_norm": 0.11033974587917328,
+      "learning_rate": 0.00010809464508094646,
+      "loss": 0.7434,
+      "step": 2364
+    },
+    {
+      "epoch": 1.42740718382131,
+      "grad_norm": 0.1310395896434784,
+      "learning_rate": 0.00010805313408053135,
+      "loss": 0.8429,
+      "step": 2365
+    },
+    {
+      "epoch": 1.4280108662843345,
+      "grad_norm": 0.11207132786512375,
+      "learning_rate": 0.00010801162308011624,
+      "loss": 1.0315,
+      "step": 2366
+    },
+    {
+      "epoch": 1.428614548747359,
+      "grad_norm": 0.10667724162340164,
+      "learning_rate": 0.00010797011207970114,
+      "loss": 0.5591,
+      "step": 2367
+    },
+    {
+      "epoch": 1.4292182312103834,
+      "grad_norm": 0.11251533776521683,
+      "learning_rate": 0.00010792860107928602,
+      "loss": 0.7458,
+      "step": 2368
+    },
+    {
+      "epoch": 1.4298219136734078,
+      "grad_norm": 0.11157943308353424,
+      "learning_rate": 0.0001078870900788709,
+      "loss": 0.7979,
+      "step": 2369
+    },
+    {
+      "epoch": 1.4304255961364323,
+      "grad_norm": 0.10930392146110535,
+      "learning_rate": 0.00010784557907845581,
+      "loss": 0.7159,
+      "step": 2370
+    },
+    {
+      "epoch": 1.4310292785994567,
+      "grad_norm": 0.11821290105581284,
+      "learning_rate": 0.00010780406807804068,
+      "loss": 0.8176,
+      "step": 2371
+    },
+    {
+      "epoch": 1.4316329610624812,
+      "grad_norm": 0.11287893354892731,
+      "learning_rate": 0.00010776255707762557,
+      "loss": 0.7326,
+      "step": 2372
+    },
+    {
+      "epoch": 1.4322366435255056,
+      "grad_norm": 0.11231239885091782,
+      "learning_rate": 0.00010772104607721047,
+      "loss": 0.7556,
+      "step": 2373
+    },
+    {
+      "epoch": 1.43284032598853,
+      "grad_norm": 0.10399778932332993,
+      "learning_rate": 0.00010767953507679535,
+      "loss": 0.7313,
+      "step": 2374
+    },
+    {
+      "epoch": 1.4334440084515545,
+      "grad_norm": 0.112164206802845,
+      "learning_rate": 0.00010763802407638024,
+      "loss": 0.7218,
+      "step": 2375
+    },
+    {
+      "epoch": 1.434047690914579,
+      "grad_norm": 0.10718587040901184,
+      "learning_rate": 0.00010759651307596514,
+      "loss": 0.7168,
+      "step": 2376
+    },
+    {
+      "epoch": 1.4346513733776034,
+      "grad_norm": 0.1262284815311432,
+      "learning_rate": 0.00010755500207555003,
+      "loss": 0.7782,
+      "step": 2377
+    },
+    {
+      "epoch": 1.4352550558406278,
+      "grad_norm": 0.11654514819383621,
+      "learning_rate": 0.00010751349107513491,
+      "loss": 0.7666,
+      "step": 2378
+    },
+    {
+      "epoch": 1.4358587383036523,
+      "grad_norm": 0.10238191485404968,
+      "learning_rate": 0.00010747198007471981,
+      "loss": 0.7126,
+      "step": 2379
+    },
+    {
+      "epoch": 1.4364624207666767,
+      "grad_norm": 0.11885840445756912,
+      "learning_rate": 0.0001074304690743047,
+      "loss": 0.9197,
+      "step": 2380
+    },
+    {
+      "epoch": 1.4370661032297012,
+      "grad_norm": 0.11124838888645172,
+      "learning_rate": 0.00010738895807388957,
+      "loss": 0.6942,
+      "step": 2381
+    },
+    {
+      "epoch": 1.4376697856927256,
+      "grad_norm": 0.10520321130752563,
+      "learning_rate": 0.00010734744707347449,
+      "loss": 1.0973,
+      "step": 2382
+    },
+    {
+      "epoch": 1.43827346815575,
+      "grad_norm": 0.10521334409713745,
+      "learning_rate": 0.00010730593607305936,
+      "loss": 0.6247,
+      "step": 2383
+    },
+    {
+      "epoch": 1.4388771506187745,
+      "grad_norm": 0.1028476133942604,
+      "learning_rate": 0.00010726442507264425,
+      "loss": 0.9452,
+      "step": 2384
+    },
+    {
+      "epoch": 1.439480833081799,
+      "grad_norm": 0.10759977251291275,
+      "learning_rate": 0.00010722291407222915,
+      "loss": 0.7051,
+      "step": 2385
+    },
+    {
+      "epoch": 1.4400845155448234,
+      "grad_norm": 0.13091117143630981,
+      "learning_rate": 0.00010718140307181403,
+      "loss": 0.8791,
+      "step": 2386
+    },
+    {
+      "epoch": 1.4406881980078479,
+      "grad_norm": 0.11425816267728806,
+      "learning_rate": 0.00010713989207139892,
+      "loss": 0.7791,
+      "step": 2387
+    },
+    {
+      "epoch": 1.4412918804708723,
+      "grad_norm": 0.10497016459703445,
+      "learning_rate": 0.00010709838107098382,
+      "loss": 0.6761,
+      "step": 2388
+    },
+    {
+      "epoch": 1.4418955629338968,
+      "grad_norm": 0.12967948615550995,
+      "learning_rate": 0.0001070568700705687,
+      "loss": 0.9618,
+      "step": 2389
+    },
+    {
+      "epoch": 1.4424992453969212,
+      "grad_norm": 0.11426857113838196,
+      "learning_rate": 0.00010701535907015359,
+      "loss": 0.761,
+      "step": 2390
+    },
+    {
+      "epoch": 1.4431029278599457,
+      "grad_norm": 0.36786845326423645,
+      "learning_rate": 0.00010697384806973849,
+      "loss": 0.722,
+      "step": 2391
+    },
+    {
+      "epoch": 1.44370661032297,
+      "grad_norm": 0.11083753407001495,
+      "learning_rate": 0.00010693233706932338,
+      "loss": 0.6985,
+      "step": 2392
+    },
+    {
+      "epoch": 1.4443102927859945,
+      "grad_norm": 0.11332450807094574,
+      "learning_rate": 0.00010689082606890825,
+      "loss": 0.6547,
+      "step": 2393
+    },
+    {
+      "epoch": 1.444913975249019,
+      "grad_norm": 0.11285028606653214,
+      "learning_rate": 0.00010684931506849317,
+      "loss": 0.6623,
+      "step": 2394
+    },
+    {
+      "epoch": 1.4455176577120434,
+      "grad_norm": 0.12340085208415985,
+      "learning_rate": 0.00010680780406807804,
+      "loss": 0.7428,
+      "step": 2395
+    },
+    {
+      "epoch": 1.4461213401750679,
+      "grad_norm": 0.11893676221370697,
+      "learning_rate": 0.00010676629306766293,
+      "loss": 0.6858,
+      "step": 2396
+    },
+    {
+      "epoch": 1.4467250226380923,
+      "grad_norm": 0.11241084337234497,
+      "learning_rate": 0.00010672478206724783,
+      "loss": 0.6083,
+      "step": 2397
+    },
+    {
+      "epoch": 1.4473287051011168,
+      "grad_norm": 0.12452124804258347,
+      "learning_rate": 0.00010668327106683271,
+      "loss": 0.6005,
+      "step": 2398
+    },
+    {
+      "epoch": 1.4479323875641412,
+      "grad_norm": 0.3717198073863983,
+      "learning_rate": 0.0001066417600664176,
+      "loss": 0.651,
+      "step": 2399
+    },
+    {
+      "epoch": 1.4485360700271657,
+      "grad_norm": 0.13178178668022156,
+      "learning_rate": 0.0001066002490660025,
+      "loss": 0.5641,
+      "step": 2400
+    },
+    {
+      "epoch": 1.4491397524901901,
+      "grad_norm": 0.13347181677818298,
+      "learning_rate": 0.00010655873806558738,
+      "loss": 0.565,
+      "step": 2401
+    },
+    {
+      "epoch": 1.4497434349532146,
+      "grad_norm": 0.13921667635440826,
+      "learning_rate": 0.00010651722706517227,
+      "loss": 0.5777,
+      "step": 2402
+    },
+    {
+      "epoch": 1.450347117416239,
+      "grad_norm": 0.14336366951465607,
+      "learning_rate": 0.00010647571606475717,
+      "loss": 0.484,
+      "step": 2403
+    },
+    {
+      "epoch": 1.4509507998792635,
+      "grad_norm": 0.16313937306404114,
+      "learning_rate": 0.00010643420506434206,
+      "loss": 0.5522,
+      "step": 2404
+    },
+    {
+      "epoch": 1.451554482342288,
+      "grad_norm": 0.16395455598831177,
+      "learning_rate": 0.00010639269406392696,
+      "loss": 0.5376,
+      "step": 2405
+    },
+    {
+      "epoch": 1.4521581648053123,
+      "grad_norm": 0.15675240755081177,
+      "learning_rate": 0.00010635118306351184,
+      "loss": 0.4248,
+      "step": 2406
+    },
+    {
+      "epoch": 1.4527618472683368,
+      "grad_norm": 0.16776950657367706,
+      "learning_rate": 0.00010630967206309672,
+      "loss": 0.3375,
+      "step": 2407
+    },
+    {
+      "epoch": 1.4533655297313612,
+      "grad_norm": 0.10852830857038498,
+      "learning_rate": 0.00010626816106268163,
+      "loss": 0.6665,
+      "step": 2408
+    },
+    {
+      "epoch": 1.4539692121943857,
+      "grad_norm": 0.115415558218956,
+      "learning_rate": 0.0001062266500622665,
+      "loss": 0.7669,
+      "step": 2409
+    },
+    {
+      "epoch": 1.4545728946574101,
+      "grad_norm": 0.12048669904470444,
+      "learning_rate": 0.00010618513906185139,
+      "loss": 0.7165,
+      "step": 2410
+    },
+    {
+      "epoch": 1.4551765771204346,
+      "grad_norm": 0.15835049748420715,
+      "learning_rate": 0.00010614362806143629,
+      "loss": 0.7294,
+      "step": 2411
+    },
+    {
+      "epoch": 1.455780259583459,
+      "grad_norm": 0.11197509616613388,
+      "learning_rate": 0.00010610211706102118,
+      "loss": 0.9129,
+      "step": 2412
+    },
+    {
+      "epoch": 1.4563839420464835,
+      "grad_norm": 0.11068537831306458,
+      "learning_rate": 0.00010606060606060606,
+      "loss": 0.7544,
+      "step": 2413
+    },
+    {
+      "epoch": 1.456987624509508,
+      "grad_norm": 0.1250140517950058,
+      "learning_rate": 0.00010601909506019096,
+      "loss": 0.8565,
+      "step": 2414
+    },
+    {
+      "epoch": 1.4575913069725326,
+      "grad_norm": 0.11533838510513306,
+      "learning_rate": 0.00010597758405977585,
+      "loss": 0.7451,
+      "step": 2415
+    },
+    {
+      "epoch": 1.4581949894355568,
+      "grad_norm": 0.11751358211040497,
+      "learning_rate": 0.00010593607305936074,
+      "loss": 0.852,
+      "step": 2416
+    },
+    {
+      "epoch": 1.4587986718985815,
+      "grad_norm": 0.1017642691731453,
+      "learning_rate": 0.00010589456205894564,
+      "loss": 0.5441,
+      "step": 2417
+    },
+    {
+      "epoch": 1.4594023543616057,
+      "grad_norm": 0.10722334682941437,
+      "learning_rate": 0.00010585305105853052,
+      "loss": 0.8963,
+      "step": 2418
+    },
+    {
+      "epoch": 1.4600060368246304,
+      "grad_norm": 0.11673782765865326,
+      "learning_rate": 0.0001058115400581154,
+      "loss": 0.8146,
+      "step": 2419
+    },
+    {
+      "epoch": 1.4606097192876546,
+      "grad_norm": 0.12002434581518173,
+      "learning_rate": 0.00010577002905770031,
+      "loss": 0.741,
+      "step": 2420
+    },
+    {
+      "epoch": 1.4612134017506793,
+      "grad_norm": 0.11690894514322281,
+      "learning_rate": 0.00010572851805728518,
+      "loss": 0.7929,
+      "step": 2421
+    },
+    {
+      "epoch": 1.4618170842137035,
+      "grad_norm": 0.12483794242143631,
+      "learning_rate": 0.00010568700705687007,
+      "loss": 0.8771,
+      "step": 2422
+    },
+    {
+      "epoch": 1.4624207666767282,
+      "grad_norm": 0.11409945040941238,
+      "learning_rate": 0.00010564549605645497,
+      "loss": 0.6703,
+      "step": 2423
+    },
+    {
+      "epoch": 1.4630244491397524,
+      "grad_norm": 0.11765747517347336,
+      "learning_rate": 0.00010560398505603986,
+      "loss": 0.6562,
+      "step": 2424
+    },
+    {
+      "epoch": 1.463628131602777,
+      "grad_norm": 0.10898533463478088,
+      "learning_rate": 0.00010556247405562474,
+      "loss": 0.7042,
+      "step": 2425
+    },
+    {
+      "epoch": 1.4642318140658013,
+      "grad_norm": 0.12223546206951141,
+      "learning_rate": 0.00010552096305520964,
+      "loss": 0.6604,
+      "step": 2426
+    },
+    {
+      "epoch": 1.464835496528826,
+      "grad_norm": 0.10916266590356827,
+      "learning_rate": 0.00010547945205479453,
+      "loss": 0.7925,
+      "step": 2427
+    },
+    {
+      "epoch": 1.4654391789918502,
+      "grad_norm": 0.1135077029466629,
+      "learning_rate": 0.0001054379410543794,
+      "loss": 0.7005,
+      "step": 2428
+    },
+    {
+      "epoch": 1.4660428614548748,
+      "grad_norm": 0.11547960340976715,
+      "learning_rate": 0.00010539643005396431,
+      "loss": 0.7297,
+      "step": 2429
+    },
+    {
+      "epoch": 1.466646543917899,
+      "grad_norm": 0.11045881360769272,
+      "learning_rate": 0.0001053549190535492,
+      "loss": 0.7663,
+      "step": 2430
+    },
+    {
+      "epoch": 1.4672502263809237,
+      "grad_norm": 0.12327514588832855,
+      "learning_rate": 0.00010531340805313407,
+      "loss": 0.6968,
+      "step": 2431
+    },
+    {
+      "epoch": 1.467853908843948,
+      "grad_norm": 0.11502881348133087,
+      "learning_rate": 0.00010527189705271899,
+      "loss": 0.7197,
+      "step": 2432
+    },
+    {
+      "epoch": 1.4684575913069726,
+      "grad_norm": 0.11779715120792389,
+      "learning_rate": 0.00010523038605230386,
+      "loss": 0.7248,
+      "step": 2433
+    },
+    {
+      "epoch": 1.4690612737699968,
+      "grad_norm": 0.11766118556261063,
+      "learning_rate": 0.00010518887505188875,
+      "loss": 0.7546,
+      "step": 2434
+    },
+    {
+      "epoch": 1.4696649562330215,
+      "grad_norm": 0.11458957195281982,
+      "learning_rate": 0.00010514736405147365,
+      "loss": 0.7825,
+      "step": 2435
+    },
+    {
+      "epoch": 1.470268638696046,
+      "grad_norm": 0.11159491539001465,
+      "learning_rate": 0.00010510585305105853,
+      "loss": 0.8465,
+      "step": 2436
+    },
+    {
+      "epoch": 1.4708723211590704,
+      "grad_norm": 0.10881160199642181,
+      "learning_rate": 0.00010506434205064342,
+      "loss": 0.7381,
+      "step": 2437
+    },
+    {
+      "epoch": 1.4714760036220949,
+      "grad_norm": 0.11549453437328339,
+      "learning_rate": 0.00010502283105022832,
+      "loss": 0.7784,
+      "step": 2438
+    },
+    {
+      "epoch": 1.4720796860851193,
+      "grad_norm": 0.11238040775060654,
+      "learning_rate": 0.0001049813200498132,
+      "loss": 0.8611,
+      "step": 2439
+    },
+    {
+      "epoch": 1.4726833685481437,
+      "grad_norm": 0.1089930310845375,
+      "learning_rate": 0.00010493980904939808,
+      "loss": 0.7186,
+      "step": 2440
+    },
+    {
+      "epoch": 1.4732870510111682,
+      "grad_norm": 0.10445185750722885,
+      "learning_rate": 0.00010489829804898299,
+      "loss": 0.6156,
+      "step": 2441
+    },
+    {
+      "epoch": 1.4738907334741926,
+      "grad_norm": 0.10923328250646591,
+      "learning_rate": 0.00010485678704856787,
+      "loss": 0.6738,
+      "step": 2442
+    },
+    {
+      "epoch": 1.474494415937217,
+      "grad_norm": 0.11018381267786026,
+      "learning_rate": 0.00010481527604815275,
+      "loss": 0.6475,
+      "step": 2443
+    },
+    {
+      "epoch": 1.4750980984002415,
+      "grad_norm": 0.11723464727401733,
+      "learning_rate": 0.00010477376504773767,
+      "loss": 0.7307,
+      "step": 2444
+    },
+    {
+      "epoch": 1.475701780863266,
+      "grad_norm": 0.11130370199680328,
+      "learning_rate": 0.00010473225404732254,
+      "loss": 0.6028,
+      "step": 2445
+    },
+    {
+      "epoch": 1.4763054633262904,
+      "grad_norm": 0.11799994856119156,
+      "learning_rate": 0.00010469074304690743,
+      "loss": 0.5884,
+      "step": 2446
+    },
+    {
+      "epoch": 1.4769091457893149,
+      "grad_norm": 0.12066012620925903,
+      "learning_rate": 0.00010464923204649233,
+      "loss": 0.6483,
+      "step": 2447
+    },
+    {
+      "epoch": 1.4775128282523393,
+      "grad_norm": 0.12716256082057953,
+      "learning_rate": 0.00010460772104607721,
+      "loss": 0.6412,
+      "step": 2448
+    },
+    {
+      "epoch": 1.4781165107153638,
+      "grad_norm": 0.1431380957365036,
+      "learning_rate": 0.00010456621004566211,
+      "loss": 0.6544,
+      "step": 2449
+    },
+    {
+      "epoch": 1.4787201931783882,
+      "grad_norm": 0.1343405693769455,
+      "learning_rate": 0.000104524699045247,
+      "loss": 0.5788,
+      "step": 2450
+    },
+    {
+      "epoch": 1.4793238756414127,
+      "grad_norm": 0.1583533138036728,
+      "learning_rate": 0.00010448318804483189,
+      "loss": 0.6123,
+      "step": 2451
+    },
+    {
+      "epoch": 1.479927558104437,
+      "grad_norm": 0.1386100798845291,
+      "learning_rate": 0.00010444167704441679,
+      "loss": 0.6458,
+      "step": 2452
+    },
+    {
+      "epoch": 1.4805312405674615,
+      "grad_norm": 0.1477377563714981,
+      "learning_rate": 0.00010440016604400167,
+      "loss": 0.5432,
+      "step": 2453
+    },
+    {
+      "epoch": 1.481134923030486,
+      "grad_norm": 0.14655126631259918,
+      "learning_rate": 0.00010435865504358654,
+      "loss": 0.4965,
+      "step": 2454
+    },
+    {
+      "epoch": 1.4817386054935104,
+      "grad_norm": 0.15142911672592163,
+      "learning_rate": 0.00010431714404317146,
+      "loss": 0.4564,
+      "step": 2455
+    },
+    {
+      "epoch": 1.4823422879565349,
+      "grad_norm": 0.1687132865190506,
+      "learning_rate": 0.00010427563304275633,
+      "loss": 0.4261,
+      "step": 2456
+    },
+    {
+      "epoch": 1.4829459704195593,
+      "grad_norm": 0.15374094247817993,
+      "learning_rate": 0.00010423412204234122,
+      "loss": 0.2757,
+      "step": 2457
+    },
+    {
+      "epoch": 1.4835496528825838,
+      "grad_norm": 0.11385196447372437,
+      "learning_rate": 0.00010419261104192613,
+      "loss": 0.6181,
+      "step": 2458
+    },
+    {
+      "epoch": 1.4841533353456082,
+      "grad_norm": 0.12344599515199661,
+      "learning_rate": 0.000104151100041511,
+      "loss": 0.5809,
+      "step": 2459
+    },
+    {
+      "epoch": 1.4847570178086327,
+      "grad_norm": 0.12225256860256195,
+      "learning_rate": 0.00010410958904109589,
+      "loss": 0.8097,
+      "step": 2460
+    },
+    {
+      "epoch": 1.4853607002716571,
+      "grad_norm": 0.1283671259880066,
+      "learning_rate": 0.00010406807804068079,
+      "loss": 0.8238,
+      "step": 2461
+    },
+    {
+      "epoch": 1.4859643827346816,
+      "grad_norm": 0.12050528824329376,
+      "learning_rate": 0.00010402656704026568,
+      "loss": 0.7068,
+      "step": 2462
+    },
+    {
+      "epoch": 1.486568065197706,
+      "grad_norm": 0.12876033782958984,
+      "learning_rate": 0.00010398505603985056,
+      "loss": 0.7713,
+      "step": 2463
+    },
+    {
+      "epoch": 1.4871717476607305,
+      "grad_norm": 0.12785589694976807,
+      "learning_rate": 0.00010394354503943546,
+      "loss": 0.9149,
+      "step": 2464
+    },
+    {
+      "epoch": 1.487775430123755,
+      "grad_norm": 0.10582014918327332,
+      "learning_rate": 0.00010390203403902035,
+      "loss": 0.71,
+      "step": 2465
+    },
+    {
+      "epoch": 1.4883791125867794,
+      "grad_norm": 0.10887051373720169,
+      "learning_rate": 0.00010386052303860522,
+      "loss": 1.0357,
+      "step": 2466
+    },
+    {
+      "epoch": 1.4889827950498038,
+      "grad_norm": 0.10681470483541489,
+      "learning_rate": 0.00010381901203819014,
+      "loss": 0.7339,
+      "step": 2467
+    },
+    {
+      "epoch": 1.4895864775128282,
+      "grad_norm": 0.13924428820610046,
+      "learning_rate": 0.00010377750103777501,
+      "loss": 0.6848,
+      "step": 2468
+    },
+    {
+      "epoch": 1.4901901599758527,
+      "grad_norm": 0.10944285243749619,
+      "learning_rate": 0.0001037359900373599,
+      "loss": 0.7028,
+      "step": 2469
+    },
+    {
+      "epoch": 1.4907938424388771,
+      "grad_norm": 0.11315060406923294,
+      "learning_rate": 0.0001036944790369448,
+      "loss": 0.7433,
+      "step": 2470
+    },
+    {
+      "epoch": 1.4913975249019016,
+      "grad_norm": 0.11983177065849304,
+      "learning_rate": 0.00010365296803652968,
+      "loss": 0.7025,
+      "step": 2471
+    },
+    {
+      "epoch": 1.492001207364926,
+      "grad_norm": 0.11219801008701324,
+      "learning_rate": 0.00010361145703611457,
+      "loss": 0.7046,
+      "step": 2472
+    },
+    {
+      "epoch": 1.4926048898279505,
+      "grad_norm": 0.12625454366207123,
+      "learning_rate": 0.00010356994603569947,
+      "loss": 0.7736,
+      "step": 2473
+    },
+    {
+      "epoch": 1.493208572290975,
+      "grad_norm": 0.12454172223806381,
+      "learning_rate": 0.00010352843503528436,
+      "loss": 0.7124,
+      "step": 2474
+    },
+    {
+      "epoch": 1.4938122547539994,
+      "grad_norm": 0.11191127449274063,
+      "learning_rate": 0.00010348692403486924,
+      "loss": 0.8456,
+      "step": 2475
+    },
+    {
+      "epoch": 1.4944159372170238,
+      "grad_norm": 0.10700973868370056,
+      "learning_rate": 0.00010344541303445414,
+      "loss": 0.8491,
+      "step": 2476
+    },
+    {
+      "epoch": 1.4950196196800483,
+      "grad_norm": 0.7383321523666382,
+      "learning_rate": 0.00010340390203403903,
+      "loss": 0.8734,
+      "step": 2477
+    },
+    {
+      "epoch": 1.4956233021430727,
+      "grad_norm": 0.11059489101171494,
+      "learning_rate": 0.0001033623910336239,
+      "loss": 0.6143,
+      "step": 2478
+    },
+    {
+      "epoch": 1.4962269846060972,
+      "grad_norm": 0.11313536763191223,
+      "learning_rate": 0.00010332088003320882,
+      "loss": 0.6963,
+      "step": 2479
+    },
+    {
+      "epoch": 1.4968306670691216,
+      "grad_norm": 0.11421024799346924,
+      "learning_rate": 0.00010327936903279369,
+      "loss": 0.7297,
+      "step": 2480
+    },
+    {
+      "epoch": 1.497434349532146,
+      "grad_norm": 0.11083009093999863,
+      "learning_rate": 0.00010323785803237857,
+      "loss": 0.6467,
+      "step": 2481
+    },
+    {
+      "epoch": 1.4980380319951705,
+      "grad_norm": 0.110516257584095,
+      "learning_rate": 0.00010319634703196347,
+      "loss": 1.0344,
+      "step": 2482
+    },
+    {
+      "epoch": 1.498641714458195,
+      "grad_norm": 0.12053235620260239,
+      "learning_rate": 0.00010315483603154836,
+      "loss": 0.7158,
+      "step": 2483
+    },
+    {
+      "epoch": 1.4992453969212194,
+      "grad_norm": 0.11220195889472961,
+      "learning_rate": 0.00010311332503113325,
+      "loss": 0.9486,
+      "step": 2484
+    },
+    {
+      "epoch": 1.4998490793842438,
+      "grad_norm": 0.121260866522789,
+      "learning_rate": 0.00010307181403071815,
+      "loss": 0.6877,
+      "step": 2485
+    },
+    {
+      "epoch": 1.5004527618472685,
+      "grad_norm": 0.11453468352556229,
+      "learning_rate": 0.00010303030303030303,
+      "loss": 0.6868,
+      "step": 2486
+    },
+    {
+      "epoch": 1.5010564443102927,
+      "grad_norm": 0.11923350393772125,
+      "learning_rate": 0.00010298879202988792,
+      "loss": 0.7115,
+      "step": 2487
+    },
+    {
+      "epoch": 1.5016601267733174,
+      "grad_norm": 0.11905788630247116,
+      "learning_rate": 0.00010294728102947282,
+      "loss": 0.894,
+      "step": 2488
+    },
+    {
+      "epoch": 1.5022638092363416,
+      "grad_norm": 0.12174365669488907,
+      "learning_rate": 0.00010290577002905771,
+      "loss": 0.8226,
+      "step": 2489
+    },
+    {
+      "epoch": 1.5028674916993663,
+      "grad_norm": 0.1140153631567955,
+      "learning_rate": 0.00010286425902864258,
+      "loss": 0.744,
+      "step": 2490
+    },
+    {
+      "epoch": 1.5034711741623905,
+      "grad_norm": 0.11370176821947098,
+      "learning_rate": 0.0001028227480282275,
+      "loss": 0.6642,
+      "step": 2491
+    },
+    {
+      "epoch": 1.5040748566254152,
+      "grad_norm": 0.11257719993591309,
+      "learning_rate": 0.00010278123702781237,
+      "loss": 0.6352,
+      "step": 2492
+    },
+    {
+      "epoch": 1.5046785390884394,
+      "grad_norm": 0.11130896955728531,
+      "learning_rate": 0.00010273972602739728,
+      "loss": 0.6231,
+      "step": 2493
+    },
+    {
+      "epoch": 1.505282221551464,
+      "grad_norm": 0.11869868636131287,
+      "learning_rate": 0.00010269821502698215,
+      "loss": 0.695,
+      "step": 2494
+    },
+    {
+      "epoch": 1.5058859040144883,
+      "grad_norm": 0.12181151658296585,
+      "learning_rate": 0.00010265670402656704,
+      "loss": 0.6434,
+      "step": 2495
+    },
+    {
+      "epoch": 1.506489586477513,
+      "grad_norm": 0.12335172295570374,
+      "learning_rate": 0.00010261519302615194,
+      "loss": 0.6727,
+      "step": 2496
+    },
+    {
+      "epoch": 1.5070932689405372,
+      "grad_norm": 0.12331590801477432,
+      "learning_rate": 0.00010257368202573683,
+      "loss": 0.6231,
+      "step": 2497
+    },
+    {
+      "epoch": 1.5076969514035619,
+      "grad_norm": 0.13102783262729645,
+      "learning_rate": 0.00010253217102532171,
+      "loss": 0.5846,
+      "step": 2498
+    },
+    {
+      "epoch": 1.508300633866586,
+      "grad_norm": 0.1286454200744629,
+      "learning_rate": 0.00010249066002490661,
+      "loss": 0.6141,
+      "step": 2499
+    },
+    {
+      "epoch": 1.5089043163296108,
+      "grad_norm": 0.1343620866537094,
+      "learning_rate": 0.0001024491490244915,
+      "loss": 0.6316,
+      "step": 2500
+    },
+    {
+      "epoch": 1.5089043163296108,
+      "eval_loss": 0.795173168182373,
+      "eval_runtime": 1219.0487,
+      "eval_samples_per_second": 2.289,
+      "eval_steps_per_second": 0.286,
+      "step": 2500
+    },
+    {
+      "epoch": 1.509507998792635,
+      "grad_norm": 0.1346864253282547,
+      "learning_rate": 0.00010240763802407639,
+      "loss": 0.6343,
+      "step": 2501
+    },
+    {
+      "epoch": 1.5101116812556596,
+      "grad_norm": 0.1306980401277542,
+      "learning_rate": 0.00010236612702366129,
+      "loss": 0.5615,
+      "step": 2502
+    },
+    {
+      "epoch": 1.5107153637186839,
+      "grad_norm": 0.1505928337574005,
+      "learning_rate": 0.00010232461602324617,
+      "loss": 0.5591,
+      "step": 2503
+    },
+    {
+      "epoch": 1.5113190461817085,
+      "grad_norm": 0.1430431306362152,
+      "learning_rate": 0.00010228310502283104,
+      "loss": 0.496,
+      "step": 2504
+    },
+    {
+      "epoch": 1.5119227286447328,
+      "grad_norm": 0.1660575270652771,
+      "learning_rate": 0.00010224159402241596,
+      "loss": 0.4632,
+      "step": 2505
+    },
+    {
+      "epoch": 1.5125264111077574,
+      "grad_norm": 0.17529623210430145,
+      "learning_rate": 0.00010220008302200083,
+      "loss": 0.4824,
+      "step": 2506
+    },
+    {
+      "epoch": 1.5131300935707817,
+      "grad_norm": 0.15798205137252808,
+      "learning_rate": 0.00010215857202158572,
+      "loss": 0.3551,
+      "step": 2507
+    },
+    {
+      "epoch": 1.5137337760338063,
+      "grad_norm": 0.12132739275693893,
+      "learning_rate": 0.00010211706102117062,
+      "loss": 0.9828,
+      "step": 2508
+    },
+    {
+      "epoch": 1.5143374584968305,
+      "grad_norm": 0.12069849669933319,
+      "learning_rate": 0.0001020755500207555,
+      "loss": 0.7057,
+      "step": 2509
+    },
+    {
+      "epoch": 1.5149411409598552,
+      "grad_norm": 0.1271534413099289,
+      "learning_rate": 0.00010203403902034039,
+      "loss": 0.7686,
+      "step": 2510
+    },
+    {
+      "epoch": 1.5155448234228794,
+      "grad_norm": 0.10612853616476059,
+      "learning_rate": 0.00010199252801992529,
+      "loss": 0.7055,
+      "step": 2511
+    },
+    {
+      "epoch": 1.516148505885904,
+      "grad_norm": 0.11403614282608032,
+      "learning_rate": 0.00010195101701951018,
+      "loss": 0.673,
+      "step": 2512
+    },
+    {
+      "epoch": 1.5167521883489283,
+      "grad_norm": 0.11351923644542694,
+      "learning_rate": 0.00010190950601909506,
+      "loss": 0.6534,
+      "step": 2513
+    },
+    {
+      "epoch": 1.517355870811953,
+      "grad_norm": 0.12284346669912338,
+      "learning_rate": 0.00010186799501867996,
+      "loss": 0.7367,
+      "step": 2514
+    },
+    {
+      "epoch": 1.5179595532749772,
+      "grad_norm": 0.11047890037298203,
+      "learning_rate": 0.00010182648401826485,
+      "loss": 0.7392,
+      "step": 2515
+    },
+    {
+      "epoch": 1.518563235738002,
+      "grad_norm": 0.13320279121398926,
+      "learning_rate": 0.00010178497301784972,
+      "loss": 0.7075,
+      "step": 2516
+    },
+    {
+      "epoch": 1.5191669182010261,
+      "grad_norm": 0.1211395412683487,
+      "learning_rate": 0.00010174346201743464,
+      "loss": 0.7255,
+      "step": 2517
+    },
+    {
+      "epoch": 1.5197706006640508,
+      "grad_norm": 0.10632450878620148,
+      "learning_rate": 0.00010170195101701951,
+      "loss": 0.6936,
+      "step": 2518
+    },
+    {
+      "epoch": 1.520374283127075,
+      "grad_norm": 0.10908479243516922,
+      "learning_rate": 0.0001016604400166044,
+      "loss": 0.783,
+      "step": 2519
+    },
+    {
+      "epoch": 1.5209779655900997,
+      "grad_norm": 0.11272008717060089,
+      "learning_rate": 0.0001016189290161893,
+      "loss": 0.8508,
+      "step": 2520
+    },
+    {
+      "epoch": 1.521581648053124,
+      "grad_norm": 0.11356555670499802,
+      "learning_rate": 0.00010157741801577418,
+      "loss": 0.7808,
+      "step": 2521
+    },
+    {
+      "epoch": 1.5221853305161486,
+      "grad_norm": 0.1155843734741211,
+      "learning_rate": 0.00010153590701535907,
+      "loss": 0.9635,
+      "step": 2522
+    },
+    {
+      "epoch": 1.5227890129791728,
+      "grad_norm": 0.10973259806632996,
+      "learning_rate": 0.00010149439601494397,
+      "loss": 0.7365,
+      "step": 2523
+    },
+    {
+      "epoch": 1.5233926954421975,
+      "grad_norm": 0.1167822778224945,
+      "learning_rate": 0.00010145288501452886,
+      "loss": 0.7307,
+      "step": 2524
+    },
+    {
+      "epoch": 1.5239963779052217,
+      "grad_norm": 0.11712843924760818,
+      "learning_rate": 0.00010141137401411374,
+      "loss": 0.6832,
+      "step": 2525
+    },
+    {
+      "epoch": 1.5246000603682464,
+      "grad_norm": 0.11597023159265518,
+      "learning_rate": 0.00010136986301369864,
+      "loss": 0.7971,
+      "step": 2526
+    },
+    {
+      "epoch": 1.5252037428312708,
+      "grad_norm": 0.12328234314918518,
+      "learning_rate": 0.00010132835201328353,
+      "loss": 0.9113,
+      "step": 2527
+    },
+    {
+      "epoch": 1.5258074252942952,
+      "grad_norm": 0.1276886761188507,
+      "learning_rate": 0.0001012868410128684,
+      "loss": 0.659,
+      "step": 2528
+    },
+    {
+      "epoch": 1.5264111077573197,
+      "grad_norm": 0.11608514934778214,
+      "learning_rate": 0.00010124533001245332,
+      "loss": 0.6754,
+      "step": 2529
+    },
+    {
+      "epoch": 1.5270147902203441,
+      "grad_norm": 0.10984505712985992,
+      "learning_rate": 0.00010120381901203819,
+      "loss": 0.9727,
+      "step": 2530
+    },
+    {
+      "epoch": 1.5276184726833686,
+      "grad_norm": 0.1020493283867836,
+      "learning_rate": 0.00010116230801162307,
+      "loss": 0.6945,
+      "step": 2531
+    },
+    {
+      "epoch": 1.528222155146393,
+      "grad_norm": 0.11108864098787308,
+      "learning_rate": 0.00010112079701120797,
+      "loss": 0.6623,
+      "step": 2532
+    },
+    {
+      "epoch": 1.5288258376094175,
+      "grad_norm": 0.11104828864336014,
+      "learning_rate": 0.00010107928601079286,
+      "loss": 0.7971,
+      "step": 2533
+    },
+    {
+      "epoch": 1.529429520072442,
+      "grad_norm": 0.11642879247665405,
+      "learning_rate": 0.00010103777501037775,
+      "loss": 0.7904,
+      "step": 2534
+    },
+    {
+      "epoch": 1.5300332025354664,
+      "grad_norm": 0.11537434905767441,
+      "learning_rate": 0.00010099626400996265,
+      "loss": 0.7779,
+      "step": 2535
+    },
+    {
+      "epoch": 1.5306368849984908,
+      "grad_norm": 0.11406029760837555,
+      "learning_rate": 0.00010095475300954753,
+      "loss": 0.6764,
+      "step": 2536
+    },
+    {
+      "epoch": 1.5312405674615153,
+      "grad_norm": 0.12164659798145294,
+      "learning_rate": 0.00010091324200913243,
+      "loss": 0.6798,
+      "step": 2537
+    },
+    {
+      "epoch": 1.5318442499245397,
+      "grad_norm": 0.11121895164251328,
+      "learning_rate": 0.00010087173100871732,
+      "loss": 0.9903,
+      "step": 2538
+    },
+    {
+      "epoch": 1.5324479323875642,
+      "grad_norm": 0.13991357386112213,
+      "learning_rate": 0.00010083022000830221,
+      "loss": 0.6855,
+      "step": 2539
+    },
+    {
+      "epoch": 1.5330516148505886,
+      "grad_norm": 0.11784554272890091,
+      "learning_rate": 0.00010078870900788711,
+      "loss": 0.6766,
+      "step": 2540
+    },
+    {
+      "epoch": 1.533655297313613,
+      "grad_norm": 0.12097886949777603,
+      "learning_rate": 0.000100747198007472,
+      "loss": 0.7029,
+      "step": 2541
+    },
+    {
+      "epoch": 1.5342589797766375,
+      "grad_norm": 0.12290269136428833,
+      "learning_rate": 0.00010070568700705687,
+      "loss": 0.6882,
+      "step": 2542
+    },
+    {
+      "epoch": 1.534862662239662,
+      "grad_norm": 0.11816665530204773,
+      "learning_rate": 0.00010066417600664178,
+      "loss": 0.6425,
+      "step": 2543
+    },
+    {
+      "epoch": 1.5354663447026864,
+      "grad_norm": 0.11926699429750443,
+      "learning_rate": 0.00010062266500622665,
+      "loss": 0.6564,
+      "step": 2544
+    },
+    {
+      "epoch": 1.5360700271657108,
+      "grad_norm": 0.13197208940982819,
+      "learning_rate": 0.00010058115400581154,
+      "loss": 0.6045,
+      "step": 2545
+    },
+    {
+      "epoch": 1.5366737096287353,
+      "grad_norm": 0.12189357727766037,
+      "learning_rate": 0.00010053964300539644,
+      "loss": 0.6137,
+      "step": 2546
+    },
+    {
+      "epoch": 1.5372773920917597,
+      "grad_norm": 0.12850208580493927,
+      "learning_rate": 0.00010049813200498133,
+      "loss": 0.6533,
+      "step": 2547
+    },
+    {
+      "epoch": 1.5378810745547842,
+      "grad_norm": 0.13015426695346832,
+      "learning_rate": 0.00010045662100456621,
+      "loss": 0.597,
+      "step": 2548
+    },
+    {
+      "epoch": 1.5384847570178086,
+      "grad_norm": 0.13129258155822754,
+      "learning_rate": 0.00010041511000415111,
+      "loss": 0.5412,
+      "step": 2549
+    },
+    {
+      "epoch": 1.539088439480833,
+      "grad_norm": 0.1495341658592224,
+      "learning_rate": 0.000100373599003736,
+      "loss": 0.6265,
+      "step": 2550
+    },
+    {
+      "epoch": 1.5396921219438575,
+      "grad_norm": 0.1443036049604416,
+      "learning_rate": 0.00010033208800332087,
+      "loss": 0.5974,
+      "step": 2551
+    },
+    {
+      "epoch": 1.540295804406882,
+      "grad_norm": 0.13055108487606049,
+      "learning_rate": 0.00010029057700290579,
+      "loss": 0.5123,
+      "step": 2552
+    },
+    {
+      "epoch": 1.5408994868699064,
+      "grad_norm": 0.15532496571540833,
+      "learning_rate": 0.00010024906600249067,
+      "loss": 0.5273,
+      "step": 2553
+    },
+    {
+      "epoch": 1.5415031693329309,
+      "grad_norm": 0.15928137302398682,
+      "learning_rate": 0.00010020755500207555,
+      "loss": 0.5119,
+      "step": 2554
+    },
+    {
+      "epoch": 1.5421068517959553,
+      "grad_norm": 0.1558419167995453,
+      "learning_rate": 0.00010016604400166046,
+      "loss": 0.4332,
+      "step": 2555
+    },
+    {
+      "epoch": 1.5427105342589797,
+      "grad_norm": 0.1683838963508606,
+      "learning_rate": 0.00010012453300124533,
+      "loss": 0.3724,
+      "step": 2556
+    },
+    {
+      "epoch": 1.5433142167220042,
+      "grad_norm": 0.15534067153930664,
+      "learning_rate": 0.00010008302200083022,
+      "loss": 0.3148,
+      "step": 2557
+    },
+    {
+      "epoch": 1.5439178991850286,
+      "grad_norm": 0.10773934423923492,
+      "learning_rate": 0.00010004151100041512,
+      "loss": 0.7099,
+      "step": 2558
+    },
+    {
+      "epoch": 1.544521581648053,
+      "grad_norm": 0.1566229909658432,
+      "learning_rate": 0.0001,
+      "loss": 0.677,
+      "step": 2559
+    },
+    {
+      "epoch": 1.5451252641110775,
+      "grad_norm": 0.12028037011623383,
+      "learning_rate": 9.99584889995849e-05,
+      "loss": 0.8106,
+      "step": 2560
+    },
+    {
+      "epoch": 1.545728946574102,
+      "grad_norm": 0.1171046793460846,
+      "learning_rate": 9.991697799916978e-05,
+      "loss": 0.7519,
+      "step": 2561
+    },
+    {
+      "epoch": 1.5463326290371264,
+      "grad_norm": 0.11625587195158005,
+      "learning_rate": 9.987546699875468e-05,
+      "loss": 0.7559,
+      "step": 2562
+    },
+    {
+      "epoch": 1.5469363115001509,
+      "grad_norm": 0.11146189272403717,
+      "learning_rate": 9.983395599833956e-05,
+      "loss": 0.6246,
+      "step": 2563
+    },
+    {
+      "epoch": 1.5475399939631753,
+      "grad_norm": 0.1186312735080719,
+      "learning_rate": 9.979244499792445e-05,
+      "loss": 0.8725,
+      "step": 2564
+    },
+    {
+      "epoch": 1.5481436764261998,
+      "grad_norm": 0.11686369776725769,
+      "learning_rate": 9.975093399750934e-05,
+      "loss": 0.869,
+      "step": 2565
+    },
+    {
+      "epoch": 1.5487473588892242,
+      "grad_norm": 0.11561167985200882,
+      "learning_rate": 9.970942299709424e-05,
+      "loss": 0.7023,
+      "step": 2566
+    },
+    {
+      "epoch": 1.5493510413522489,
+      "grad_norm": 0.12024765461683273,
+      "learning_rate": 9.966791199667912e-05,
+      "loss": 0.9882,
+      "step": 2567
+    },
+    {
+      "epoch": 1.549954723815273,
+      "grad_norm": 0.10977562516927719,
+      "learning_rate": 9.962640099626401e-05,
+      "loss": 0.7127,
+      "step": 2568
+    },
+    {
+      "epoch": 1.5505584062782978,
+      "grad_norm": 0.10764128714799881,
+      "learning_rate": 9.958488999584891e-05,
+      "loss": 0.617,
+      "step": 2569
+    },
+    {
+      "epoch": 1.551162088741322,
+      "grad_norm": 0.10783983021974564,
+      "learning_rate": 9.954337899543378e-05,
+      "loss": 0.6401,
+      "step": 2570
+    },
+    {
+      "epoch": 1.5517657712043467,
+      "grad_norm": 0.12389583140611649,
+      "learning_rate": 9.950186799501868e-05,
+      "loss": 0.6535,
+      "step": 2571
+    },
+    {
+      "epoch": 1.552369453667371,
+      "grad_norm": 0.11414086818695068,
+      "learning_rate": 9.946035699460357e-05,
+      "loss": 0.7015,
+      "step": 2572
+    },
+    {
+      "epoch": 1.5529731361303956,
+      "grad_norm": 0.11838074028491974,
+      "learning_rate": 9.941884599418847e-05,
+      "loss": 0.9455,
+      "step": 2573
+    },
+    {
+      "epoch": 1.5535768185934198,
+      "grad_norm": 0.11571443825960159,
+      "learning_rate": 9.937733499377336e-05,
+      "loss": 1.1223,
+      "step": 2574
+    },
+    {
+      "epoch": 1.5541805010564445,
+      "grad_norm": 0.11583063751459122,
+      "learning_rate": 9.933582399335824e-05,
+      "loss": 0.7297,
+      "step": 2575
+    },
+    {
+      "epoch": 1.5547841835194687,
+      "grad_norm": 0.11269959062337875,
+      "learning_rate": 9.929431299294314e-05,
+      "loss": 0.7613,
+      "step": 2576
+    },
+    {
+      "epoch": 1.5553878659824933,
+      "grad_norm": 0.1328345686197281,
+      "learning_rate": 9.925280199252802e-05,
+      "loss": 0.7758,
+      "step": 2577
+    },
+    {
+      "epoch": 1.5559915484455176,
+      "grad_norm": 0.12824542820453644,
+      "learning_rate": 9.921129099211292e-05,
+      "loss": 0.7867,
+      "step": 2578
+    },
+    {
+      "epoch": 1.5565952309085422,
+      "grad_norm": 0.10252919793128967,
+      "learning_rate": 9.91697799916978e-05,
+      "loss": 0.6038,
+      "step": 2579
+    },
+    {
+      "epoch": 1.5571989133715665,
+      "grad_norm": 0.11735723912715912,
+      "learning_rate": 9.912826899128269e-05,
+      "loss": 0.718,
+      "step": 2580
+    },
+    {
+      "epoch": 1.5578025958345911,
+      "grad_norm": 0.12741461396217346,
+      "learning_rate": 9.908675799086759e-05,
+      "loss": 0.62,
+      "step": 2581
+    },
+    {
+      "epoch": 1.5584062782976154,
+      "grad_norm": 0.11475570499897003,
+      "learning_rate": 9.904524699045248e-05,
+      "loss": 0.7747,
+      "step": 2582
+    },
+    {
+      "epoch": 1.55900996076064,
+      "grad_norm": 0.11514543741941452,
+      "learning_rate": 9.900373599003736e-05,
+      "loss": 0.7556,
+      "step": 2583
+    },
+    {
+      "epoch": 1.5596136432236642,
+      "grad_norm": 0.10353957116603851,
+      "learning_rate": 9.896222498962225e-05,
+      "loss": 0.6466,
+      "step": 2584
+    },
+    {
+      "epoch": 1.560217325686689,
+      "grad_norm": 0.10758772492408752,
+      "learning_rate": 9.892071398920715e-05,
+      "loss": 1.2809,
+      "step": 2585
+    },
+    {
+      "epoch": 1.5608210081497131,
+      "grad_norm": 0.11749143153429031,
+      "learning_rate": 9.887920298879203e-05,
+      "loss": 0.6888,
+      "step": 2586
+    },
+    {
+      "epoch": 1.5614246906127378,
+      "grad_norm": 0.1256055384874344,
+      "learning_rate": 9.883769198837692e-05,
+      "loss": 0.7231,
+      "step": 2587
+    },
+    {
+      "epoch": 1.562028373075762,
+      "grad_norm": 0.11417461931705475,
+      "learning_rate": 9.879618098796182e-05,
+      "loss": 0.7842,
+      "step": 2588
+    },
+    {
+      "epoch": 1.5626320555387867,
+      "grad_norm": 0.10440021008253098,
+      "learning_rate": 9.87546699875467e-05,
+      "loss": 0.6953,
+      "step": 2589
+    },
+    {
+      "epoch": 1.563235738001811,
+      "grad_norm": 0.10997917503118515,
+      "learning_rate": 9.87131589871316e-05,
+      "loss": 0.6638,
+      "step": 2590
+    },
+    {
+      "epoch": 1.5638394204648356,
+      "grad_norm": 0.1075371578335762,
+      "learning_rate": 9.867164798671648e-05,
+      "loss": 0.5918,
+      "step": 2591
+    },
+    {
+      "epoch": 1.5644431029278598,
+      "grad_norm": 0.12016437947750092,
+      "learning_rate": 9.863013698630137e-05,
+      "loss": 0.7686,
+      "step": 2592
+    },
+    {
+      "epoch": 1.5650467853908845,
+      "grad_norm": 0.119369737803936,
+      "learning_rate": 9.858862598588627e-05,
+      "loss": 0.6284,
+      "step": 2593
+    },
+    {
+      "epoch": 1.5656504678539087,
+      "grad_norm": 0.11695661395788193,
+      "learning_rate": 9.854711498547115e-05,
+      "loss": 0.6282,
+      "step": 2594
+    },
+    {
+      "epoch": 1.5662541503169334,
+      "grad_norm": 0.11359156668186188,
+      "learning_rate": 9.850560398505605e-05,
+      "loss": 0.648,
+      "step": 2595
+    },
+    {
+      "epoch": 1.5668578327799576,
+      "grad_norm": 0.11783526092767715,
+      "learning_rate": 9.846409298464093e-05,
+      "loss": 0.5769,
+      "step": 2596
+    },
+    {
+      "epoch": 1.5674615152429823,
+      "grad_norm": 0.12867143750190735,
+      "learning_rate": 9.842258198422583e-05,
+      "loss": 0.7181,
+      "step": 2597
+    },
+    {
+      "epoch": 1.5680651977060065,
+      "grad_norm": 0.12808844447135925,
+      "learning_rate": 9.838107098381071e-05,
+      "loss": 0.6346,
+      "step": 2598
+    },
+    {
+      "epoch": 1.5686688801690312,
+      "grad_norm": 0.12455364316701889,
+      "learning_rate": 9.83395599833956e-05,
+      "loss": 0.6019,
+      "step": 2599
+    },
+    {
+      "epoch": 1.5692725626320554,
+      "grad_norm": 0.1380644142627716,
+      "learning_rate": 9.82980489829805e-05,
+      "loss": 0.6872,
+      "step": 2600
+    },
+    {
+      "epoch": 1.56987624509508,
+      "grad_norm": 0.14129561185836792,
+      "learning_rate": 9.825653798256539e-05,
+      "loss": 0.5949,
+      "step": 2601
+    },
+    {
+      "epoch": 1.5704799275581043,
+      "grad_norm": 0.13899260759353638,
+      "learning_rate": 9.821502698215027e-05,
+      "loss": 0.5492,
+      "step": 2602
+    },
+    {
+      "epoch": 1.571083610021129,
+      "grad_norm": 0.13930819928646088,
+      "learning_rate": 9.817351598173516e-05,
+      "loss": 0.5476,
+      "step": 2603
+    },
+    {
+      "epoch": 1.5716872924841532,
+      "grad_norm": 0.14664794504642487,
+      "learning_rate": 9.813200498132006e-05,
+      "loss": 0.4718,
+      "step": 2604
+    },
+    {
+      "epoch": 1.5722909749471778,
+      "grad_norm": 0.1534009575843811,
+      "learning_rate": 9.809049398090495e-05,
+      "loss": 0.4487,
+      "step": 2605
+    },
+    {
+      "epoch": 1.572894657410202,
+      "grad_norm": 0.16464664041996002,
+      "learning_rate": 9.804898298048983e-05,
+      "loss": 0.4443,
+      "step": 2606
+    },
+    {
+      "epoch": 1.5734983398732267,
+      "grad_norm": 0.15683351457118988,
+      "learning_rate": 9.800747198007473e-05,
+      "loss": 0.3545,
+      "step": 2607
+    },
+    {
+      "epoch": 1.5741020223362512,
+      "grad_norm": 0.142044797539711,
+      "learning_rate": 9.79659609796596e-05,
+      "loss": 0.7322,
+      "step": 2608
+    },
+    {
+      "epoch": 1.5747057047992756,
+      "grad_norm": 0.13781249523162842,
+      "learning_rate": 9.79244499792445e-05,
+      "loss": 0.7452,
+      "step": 2609
+    },
+    {
+      "epoch": 1.5753093872623,
+      "grad_norm": 0.12717604637145996,
+      "learning_rate": 9.788293897882939e-05,
+      "loss": 0.749,
+      "step": 2610
+    },
+    {
+      "epoch": 1.5759130697253245,
+      "grad_norm": 0.11919858306646347,
+      "learning_rate": 9.784142797841428e-05,
+      "loss": 0.7633,
+      "step": 2611
+    },
+    {
+      "epoch": 1.576516752188349,
+      "grad_norm": 0.12019964307546616,
+      "learning_rate": 9.779991697799918e-05,
+      "loss": 0.732,
+      "step": 2612
+    },
+    {
+      "epoch": 1.5771204346513734,
+      "grad_norm": 0.1292140632867813,
+      "learning_rate": 9.775840597758406e-05,
+      "loss": 0.7751,
+      "step": 2613
+    },
+    {
+      "epoch": 1.5777241171143979,
+      "grad_norm": 0.11317221075296402,
+      "learning_rate": 9.771689497716895e-05,
+      "loss": 0.7186,
+      "step": 2614
+    },
+    {
+      "epoch": 1.5783277995774223,
+      "grad_norm": 0.10719773173332214,
+      "learning_rate": 9.767538397675384e-05,
+      "loss": 0.7511,
+      "step": 2615
+    },
+    {
+      "epoch": 1.5789314820404468,
+      "grad_norm": 0.11181782931089401,
+      "learning_rate": 9.763387297633874e-05,
+      "loss": 0.7537,
+      "step": 2616
+    },
+    {
+      "epoch": 1.5795351645034712,
+      "grad_norm": 0.11379014700651169,
+      "learning_rate": 9.759236197592362e-05,
+      "loss": 0.7367,
+      "step": 2617
+    },
+    {
+      "epoch": 1.5801388469664956,
+      "grad_norm": 0.10861673206090927,
+      "learning_rate": 9.755085097550851e-05,
+      "loss": 0.6676,
+      "step": 2618
+    },
+    {
+      "epoch": 1.58074252942952,
+      "grad_norm": 0.11700250953435898,
+      "learning_rate": 9.750933997509341e-05,
+      "loss": 0.6992,
+      "step": 2619
+    },
+    {
+      "epoch": 1.5813462118925445,
+      "grad_norm": 0.10913542658090591,
+      "learning_rate": 9.74678289746783e-05,
+      "loss": 0.7337,
+      "step": 2620
+    },
+    {
+      "epoch": 1.581949894355569,
+      "grad_norm": 0.10216758400201797,
+      "learning_rate": 9.742631797426318e-05,
+      "loss": 0.658,
+      "step": 2621
+    },
+    {
+      "epoch": 1.5825535768185934,
+      "grad_norm": 0.10746050626039505,
+      "learning_rate": 9.738480697384807e-05,
+      "loss": 0.673,
+      "step": 2622
+    },
+    {
+      "epoch": 1.5831572592816179,
+      "grad_norm": 0.10291888564825058,
+      "learning_rate": 9.734329597343297e-05,
+      "loss": 0.7131,
+      "step": 2623
+    },
+    {
+      "epoch": 1.5837609417446423,
+      "grad_norm": 0.11497275531291962,
+      "learning_rate": 9.730178497301786e-05,
+      "loss": 0.7278,
+      "step": 2624
+    },
+    {
+      "epoch": 1.5843646242076668,
+      "grad_norm": 0.11286524683237076,
+      "learning_rate": 9.726027397260274e-05,
+      "loss": 0.7349,
+      "step": 2625
+    },
+    {
+      "epoch": 1.5849683066706912,
+      "grad_norm": 0.12478066235780716,
+      "learning_rate": 9.721876297218764e-05,
+      "loss": 0.686,
+      "step": 2626
+    },
+    {
+      "epoch": 1.5855719891337157,
+      "grad_norm": 0.1130586788058281,
+      "learning_rate": 9.717725197177252e-05,
+      "loss": 0.7609,
+      "step": 2627
+    },
+    {
+      "epoch": 1.58617567159674,
+      "grad_norm": 0.11195287853479385,
+      "learning_rate": 9.713574097135742e-05,
+      "loss": 0.6909,
+      "step": 2628
+    },
+    {
+      "epoch": 1.5867793540597646,
+      "grad_norm": 0.127894327044487,
+      "learning_rate": 9.70942299709423e-05,
+      "loss": 0.7867,
+      "step": 2629
+    },
+    {
+      "epoch": 1.587383036522789,
+      "grad_norm": 0.10859086364507675,
+      "learning_rate": 9.705271897052719e-05,
+      "loss": 0.7013,
+      "step": 2630
+    },
+    {
+      "epoch": 1.5879867189858135,
+      "grad_norm": 0.10940494388341904,
+      "learning_rate": 9.701120797011209e-05,
+      "loss": 0.9159,
+      "step": 2631
+    },
+    {
+      "epoch": 1.588590401448838,
+      "grad_norm": 0.1171516478061676,
+      "learning_rate": 9.696969696969698e-05,
+      "loss": 0.6883,
+      "step": 2632
+    },
+    {
+      "epoch": 1.5891940839118623,
+      "grad_norm": 0.14016135036945343,
+      "learning_rate": 9.692818596928186e-05,
+      "loss": 0.6465,
+      "step": 2633
+    },
+    {
+      "epoch": 1.5897977663748868,
+      "grad_norm": 0.10493374615907669,
+      "learning_rate": 9.688667496886675e-05,
+      "loss": 0.6847,
+      "step": 2634
+    },
+    {
+      "epoch": 1.5904014488379112,
+      "grad_norm": 0.12011600285768509,
+      "learning_rate": 9.684516396845165e-05,
+      "loss": 0.7248,
+      "step": 2635
+    },
+    {
+      "epoch": 1.5910051313009357,
+      "grad_norm": 0.1263171285390854,
+      "learning_rate": 9.680365296803654e-05,
+      "loss": 0.646,
+      "step": 2636
+    },
+    {
+      "epoch": 1.5916088137639601,
+      "grad_norm": 0.1184513196349144,
+      "learning_rate": 9.676214196762142e-05,
+      "loss": 0.6105,
+      "step": 2637
+    },
+    {
+      "epoch": 1.5922124962269846,
+      "grad_norm": 0.13153387606143951,
+      "learning_rate": 9.672063096720632e-05,
+      "loss": 0.7279,
+      "step": 2638
+    },
+    {
+      "epoch": 1.592816178690009,
+      "grad_norm": 0.12811937928199768,
+      "learning_rate": 9.667911996679121e-05,
+      "loss": 0.7132,
+      "step": 2639
+    },
+    {
+      "epoch": 1.5934198611530335,
+      "grad_norm": 0.10892543196678162,
+      "learning_rate": 9.66376089663761e-05,
+      "loss": 0.8649,
+      "step": 2640
+    },
+    {
+      "epoch": 1.594023543616058,
+      "grad_norm": 0.1102840006351471,
+      "learning_rate": 9.659609796596098e-05,
+      "loss": 0.685,
+      "step": 2641
+    },
+    {
+      "epoch": 1.5946272260790824,
+      "grad_norm": 0.11115845292806625,
+      "learning_rate": 9.655458696554588e-05,
+      "loss": 0.5798,
+      "step": 2642
+    },
+    {
+      "epoch": 1.5952309085421068,
+      "grad_norm": 0.10934832692146301,
+      "learning_rate": 9.651307596513077e-05,
+      "loss": 0.6023,
+      "step": 2643
+    },
+    {
+      "epoch": 1.5958345910051313,
+      "grad_norm": 0.12175562977790833,
+      "learning_rate": 9.647156496471565e-05,
+      "loss": 0.7045,
+      "step": 2644
+    },
+    {
+      "epoch": 1.5964382734681557,
+      "grad_norm": 0.12231041491031647,
+      "learning_rate": 9.643005396430055e-05,
+      "loss": 0.6613,
+      "step": 2645
+    },
+    {
+      "epoch": 1.5970419559311801,
+      "grad_norm": 0.11286479979753494,
+      "learning_rate": 9.638854296388543e-05,
+      "loss": 0.6132,
+      "step": 2646
+    },
+    {
+      "epoch": 1.5976456383942046,
+      "grad_norm": 0.13253097236156464,
+      "learning_rate": 9.634703196347033e-05,
+      "loss": 0.65,
+      "step": 2647
+    },
+    {
+      "epoch": 1.5982493208572293,
+      "grad_norm": 0.1287652999162674,
+      "learning_rate": 9.630552096305521e-05,
+      "loss": 0.6567,
+      "step": 2648
+    },
+    {
+      "epoch": 1.5988530033202535,
+      "grad_norm": 0.15067052841186523,
+      "learning_rate": 9.62640099626401e-05,
+      "loss": 0.6427,
+      "step": 2649
+    },
+    {
+      "epoch": 1.5994566857832782,
+      "grad_norm": 0.13497310876846313,
+      "learning_rate": 9.6222498962225e-05,
+      "loss": 0.6178,
+      "step": 2650
+    },
+    {
+      "epoch": 1.6000603682463024,
+      "grad_norm": 0.13561968505382538,
+      "learning_rate": 9.618098796180989e-05,
+      "loss": 0.5794,
+      "step": 2651
+    },
+    {
+      "epoch": 1.600664050709327,
+      "grad_norm": 0.1425342857837677,
+      "learning_rate": 9.613947696139477e-05,
+      "loss": 0.5064,
+      "step": 2652
+    },
+    {
+      "epoch": 1.6012677331723513,
+      "grad_norm": 0.13813789188861847,
+      "learning_rate": 9.609796596097966e-05,
+      "loss": 0.4899,
+      "step": 2653
+    },
+    {
+      "epoch": 1.601871415635376,
+      "grad_norm": 0.14464135468006134,
+      "learning_rate": 9.605645496056456e-05,
+      "loss": 0.4724,
+      "step": 2654
+    },
+    {
+      "epoch": 1.6024750980984002,
+      "grad_norm": 0.15810330212116241,
+      "learning_rate": 9.601494396014945e-05,
+      "loss": 0.4372,
+      "step": 2655
+    },
+    {
+      "epoch": 1.6030787805614248,
+      "grad_norm": 0.17366820573806763,
+      "learning_rate": 9.597343295973433e-05,
+      "loss": 0.4398,
+      "step": 2656
+    },
+    {
+      "epoch": 1.603682463024449,
+      "grad_norm": 0.159001424908638,
+      "learning_rate": 9.593192195931923e-05,
+      "loss": 0.2789,
+      "step": 2657
+    },
+    {
+      "epoch": 1.6042861454874737,
+      "grad_norm": 0.2292790710926056,
+      "learning_rate": 9.58904109589041e-05,
+      "loss": 0.7525,
+      "step": 2658
+    },
+    {
+      "epoch": 1.604889827950498,
+      "grad_norm": 0.10119739919900894,
+      "learning_rate": 9.5848899958489e-05,
+      "loss": 0.717,
+      "step": 2659
+    },
+    {
+      "epoch": 1.6054935104135226,
+      "grad_norm": 0.10911940038204193,
+      "learning_rate": 9.580738895807389e-05,
+      "loss": 0.6715,
+      "step": 2660
+    },
+    {
+      "epoch": 1.6060971928765468,
+      "grad_norm": 0.12329068779945374,
+      "learning_rate": 9.576587795765879e-05,
+      "loss": 0.7417,
+      "step": 2661
+    },
+    {
+      "epoch": 1.6067008753395715,
+      "grad_norm": 0.12208274751901627,
+      "learning_rate": 9.572436695724368e-05,
+      "loss": 0.7459,
+      "step": 2662
+    },
+    {
+      "epoch": 1.6073045578025957,
+      "grad_norm": 0.10721976310014725,
+      "learning_rate": 9.568285595682856e-05,
+      "loss": 0.6844,
+      "step": 2663
+    },
+    {
+      "epoch": 1.6079082402656204,
+      "grad_norm": 0.1042000949382782,
+      "learning_rate": 9.564134495641347e-05,
+      "loss": 0.7077,
+      "step": 2664
+    },
+    {
+      "epoch": 1.6085119227286446,
+      "grad_norm": 0.1249273270368576,
+      "learning_rate": 9.559983395599834e-05,
+      "loss": 0.7202,
+      "step": 2665
+    },
+    {
+      "epoch": 1.6091156051916693,
+      "grad_norm": 0.11742489039897919,
+      "learning_rate": 9.555832295558324e-05,
+      "loss": 0.6975,
+      "step": 2666
+    },
+    {
+      "epoch": 1.6097192876546935,
+      "grad_norm": 0.1132553368806839,
+      "learning_rate": 9.551681195516812e-05,
+      "loss": 0.9097,
+      "step": 2667
+    },
+    {
+      "epoch": 1.6103229701177182,
+      "grad_norm": 0.1129176989197731,
+      "learning_rate": 9.547530095475301e-05,
+      "loss": 0.6419,
+      "step": 2668
+    },
+    {
+      "epoch": 1.6109266525807424,
+      "grad_norm": 0.1597258597612381,
+      "learning_rate": 9.543378995433791e-05,
+      "loss": 0.7293,
+      "step": 2669
+    },
+    {
+      "epoch": 1.611530335043767,
+      "grad_norm": 0.11292175203561783,
+      "learning_rate": 9.53922789539228e-05,
+      "loss": 0.8514,
+      "step": 2670
+    },
+    {
+      "epoch": 1.6121340175067913,
+      "grad_norm": 0.11544401198625565,
+      "learning_rate": 9.535076795350768e-05,
+      "loss": 0.6806,
+      "step": 2671
+    },
+    {
+      "epoch": 1.612737699969816,
+      "grad_norm": 0.10773265361785889,
+      "learning_rate": 9.530925695309257e-05,
+      "loss": 0.7818,
+      "step": 2672
+    },
+    {
+      "epoch": 1.6133413824328402,
+      "grad_norm": 0.11048367619514465,
+      "learning_rate": 9.526774595267747e-05,
+      "loss": 0.6819,
+      "step": 2673
+    },
+    {
+      "epoch": 1.6139450648958649,
+      "grad_norm": 0.12018372118473053,
+      "learning_rate": 9.522623495226234e-05,
+      "loss": 0.6918,
+      "step": 2674
+    },
+    {
+      "epoch": 1.614548747358889,
+      "grad_norm": 0.1126914694905281,
+      "learning_rate": 9.518472395184724e-05,
+      "loss": 1.157,
+      "step": 2675
+    },
+    {
+      "epoch": 1.6151524298219138,
+      "grad_norm": 0.11121730506420135,
+      "learning_rate": 9.514321295143214e-05,
+      "loss": 0.7578,
+      "step": 2676
+    },
+    {
+      "epoch": 1.615756112284938,
+      "grad_norm": 0.1760278046131134,
+      "learning_rate": 9.510170195101702e-05,
+      "loss": 0.7201,
+      "step": 2677
+    },
+    {
+      "epoch": 1.6163597947479627,
+      "grad_norm": 0.1170695498585701,
+      "learning_rate": 9.506019095060192e-05,
+      "loss": 0.7593,
+      "step": 2678
+    },
+    {
+      "epoch": 1.6169634772109869,
+      "grad_norm": 0.12168256938457489,
+      "learning_rate": 9.50186799501868e-05,
+      "loss": 0.6685,
+      "step": 2679
+    },
+    {
+      "epoch": 1.6175671596740115,
+      "grad_norm": 0.1391884833574295,
+      "learning_rate": 9.497716894977169e-05,
+      "loss": 0.8251,
+      "step": 2680
+    },
+    {
+      "epoch": 1.6181708421370358,
+      "grad_norm": 0.11195441335439682,
+      "learning_rate": 9.493565794935658e-05,
+      "loss": 0.6997,
+      "step": 2681
+    },
+    {
+      "epoch": 1.6187745246000604,
+      "grad_norm": 0.11107617616653442,
+      "learning_rate": 9.489414694894148e-05,
+      "loss": 0.6234,
+      "step": 2682
+    },
+    {
+      "epoch": 1.6193782070630847,
+      "grad_norm": 0.1288166344165802,
+      "learning_rate": 9.485263594852638e-05,
+      "loss": 1.0108,
+      "step": 2683
+    },
+    {
+      "epoch": 1.6199818895261093,
+      "grad_norm": 0.11174750328063965,
+      "learning_rate": 9.481112494811125e-05,
+      "loss": 0.633,
+      "step": 2684
+    },
+    {
+      "epoch": 1.6205855719891336,
+      "grad_norm": 0.11930475383996964,
+      "learning_rate": 9.476961394769615e-05,
+      "loss": 0.7875,
+      "step": 2685
+    },
+    {
+      "epoch": 1.6211892544521582,
+      "grad_norm": 0.1079435870051384,
+      "learning_rate": 9.472810294728104e-05,
+      "loss": 0.9182,
+      "step": 2686
+    },
+    {
+      "epoch": 1.6217929369151824,
+      "grad_norm": 0.10946819186210632,
+      "learning_rate": 9.468659194686592e-05,
+      "loss": 0.6537,
+      "step": 2687
+    },
+    {
+      "epoch": 1.6223966193782071,
+      "grad_norm": 0.1106157973408699,
+      "learning_rate": 9.464508094645081e-05,
+      "loss": 0.5829,
+      "step": 2688
+    },
+    {
+      "epoch": 1.6230003018412316,
+      "grad_norm": 0.1129320040345192,
+      "learning_rate": 9.460356994603571e-05,
+      "loss": 0.7042,
+      "step": 2689
+    },
+    {
+      "epoch": 1.623603984304256,
+      "grad_norm": 0.10891727358102798,
+      "learning_rate": 9.45620589456206e-05,
+      "loss": 0.6194,
+      "step": 2690
+    },
+    {
+      "epoch": 1.6242076667672805,
+      "grad_norm": 0.1271827220916748,
+      "learning_rate": 9.452054794520548e-05,
+      "loss": 0.6827,
+      "step": 2691
+    },
+    {
+      "epoch": 1.624811349230305,
+      "grad_norm": 0.10997018963098526,
+      "learning_rate": 9.447903694479038e-05,
+      "loss": 0.6144,
+      "step": 2692
+    },
+    {
+      "epoch": 1.6254150316933293,
+      "grad_norm": 0.11678522080183029,
+      "learning_rate": 9.443752594437525e-05,
+      "loss": 0.6143,
+      "step": 2693
+    },
+    {
+      "epoch": 1.6260187141563538,
+      "grad_norm": 0.13937309384346008,
+      "learning_rate": 9.439601494396015e-05,
+      "loss": 0.7645,
+      "step": 2694
+    },
+    {
+      "epoch": 1.6266223966193782,
+      "grad_norm": 0.12499664723873138,
+      "learning_rate": 9.435450394354504e-05,
+      "loss": 0.6144,
+      "step": 2695
+    },
+    {
+      "epoch": 1.6272260790824027,
+      "grad_norm": 0.12473144382238388,
+      "learning_rate": 9.431299294312993e-05,
+      "loss": 0.5597,
+      "step": 2696
+    },
+    {
+      "epoch": 1.6278297615454271,
+      "grad_norm": 0.13701920211315155,
+      "learning_rate": 9.427148194271483e-05,
+      "loss": 0.6018,
+      "step": 2697
+    },
+    {
+      "epoch": 1.6284334440084516,
+      "grad_norm": 0.12637366354465485,
+      "learning_rate": 9.422997094229971e-05,
+      "loss": 0.6128,
+      "step": 2698
+    },
+    {
+      "epoch": 1.629037126471476,
+      "grad_norm": 0.12743206322193146,
+      "learning_rate": 9.41884599418846e-05,
+      "loss": 0.6142,
+      "step": 2699
+    },
+    {
+      "epoch": 1.6296408089345005,
+      "grad_norm": 0.1346701979637146,
+      "learning_rate": 9.414694894146949e-05,
+      "loss": 0.6122,
+      "step": 2700
+    },
+    {
+      "epoch": 1.630244491397525,
+      "grad_norm": 0.13191986083984375,
+      "learning_rate": 9.410543794105439e-05,
+      "loss": 0.566,
+      "step": 2701
+    },
+    {
+      "epoch": 1.6308481738605494,
+      "grad_norm": 0.14156706631183624,
+      "learning_rate": 9.406392694063927e-05,
+      "loss": 0.5651,
+      "step": 2702
+    },
+    {
+      "epoch": 1.6314518563235738,
+      "grad_norm": 0.1478833556175232,
+      "learning_rate": 9.402241594022416e-05,
+      "loss": 0.5533,
+      "step": 2703
+    },
+    {
+      "epoch": 1.6320555387865983,
+      "grad_norm": 0.1488913893699646,
+      "learning_rate": 9.398090493980906e-05,
+      "loss": 0.5218,
+      "step": 2704
+    },
+    {
+      "epoch": 1.6326592212496227,
+      "grad_norm": 0.1752762794494629,
+      "learning_rate": 9.393939393939395e-05,
+      "loss": 0.4899,
+      "step": 2705
+    },
+    {
+      "epoch": 1.6332629037126472,
+      "grad_norm": 0.17121976613998413,
+      "learning_rate": 9.389788293897883e-05,
+      "loss": 0.4118,
+      "step": 2706
+    },
+    {
+      "epoch": 1.6338665861756716,
+      "grad_norm": 0.16428853571414948,
+      "learning_rate": 9.385637193856372e-05,
+      "loss": 0.3176,
+      "step": 2707
+    },
+    {
+      "epoch": 1.634470268638696,
+      "grad_norm": 0.11714405566453934,
+      "learning_rate": 9.381486093814862e-05,
+      "loss": 0.8348,
+      "step": 2708
+    },
+    {
+      "epoch": 1.6350739511017205,
+      "grad_norm": 0.11836113035678864,
+      "learning_rate": 9.37733499377335e-05,
+      "loss": 0.7564,
+      "step": 2709
+    },
+    {
+      "epoch": 1.635677633564745,
+      "grad_norm": 0.11199770867824554,
+      "learning_rate": 9.373183893731839e-05,
+      "loss": 0.6204,
+      "step": 2710
+    },
+    {
+      "epoch": 1.6362813160277694,
+      "grad_norm": 0.12327376753091812,
+      "learning_rate": 9.369032793690329e-05,
+      "loss": 0.7127,
+      "step": 2711
+    },
+    {
+      "epoch": 1.6368849984907938,
+      "grad_norm": 0.12553828954696655,
+      "learning_rate": 9.364881693648817e-05,
+      "loss": 0.7299,
+      "step": 2712
+    },
+    {
+      "epoch": 1.6374886809538183,
+      "grad_norm": 0.10863874852657318,
+      "learning_rate": 9.360730593607307e-05,
+      "loss": 0.669,
+      "step": 2713
+    },
+    {
+      "epoch": 1.6380923634168427,
+      "grad_norm": 0.12236540764570236,
+      "learning_rate": 9.356579493565795e-05,
+      "loss": 0.977,
+      "step": 2714
+    },
+    {
+      "epoch": 1.6386960458798672,
+      "grad_norm": 0.1284298151731491,
+      "learning_rate": 9.352428393524284e-05,
+      "loss": 0.9,
+      "step": 2715
+    },
+    {
+      "epoch": 1.6392997283428916,
+      "grad_norm": 0.12449759244918823,
+      "learning_rate": 9.348277293482774e-05,
+      "loss": 0.5975,
+      "step": 2716
+    },
+    {
+      "epoch": 1.639903410805916,
+      "grad_norm": 0.11096858978271484,
+      "learning_rate": 9.344126193441262e-05,
+      "loss": 0.7423,
+      "step": 2717
+    },
+    {
+      "epoch": 1.6405070932689405,
+      "grad_norm": 0.11665850877761841,
+      "learning_rate": 9.339975093399751e-05,
+      "loss": 0.7947,
+      "step": 2718
+    },
+    {
+      "epoch": 1.641110775731965,
+      "grad_norm": 0.1213374212384224,
+      "learning_rate": 9.33582399335824e-05,
+      "loss": 0.7891,
+      "step": 2719
+    },
+    {
+      "epoch": 1.6417144581949894,
+      "grad_norm": 0.11723334342241287,
+      "learning_rate": 9.33167289331673e-05,
+      "loss": 0.7008,
+      "step": 2720
+    },
+    {
+      "epoch": 1.6423181406580138,
+      "grad_norm": 0.39810943603515625,
+      "learning_rate": 9.327521793275218e-05,
+      "loss": 0.7724,
+      "step": 2721
+    },
+    {
+      "epoch": 1.6429218231210383,
+      "grad_norm": 0.1107415184378624,
+      "learning_rate": 9.323370693233707e-05,
+      "loss": 0.6663,
+      "step": 2722
+    },
+    {
+      "epoch": 1.6435255055840627,
+      "grad_norm": 0.14163663983345032,
+      "learning_rate": 9.319219593192197e-05,
+      "loss": 0.8732,
+      "step": 2723
+    },
+    {
+      "epoch": 1.6441291880470872,
+      "grad_norm": 0.11508408188819885,
+      "learning_rate": 9.315068493150684e-05,
+      "loss": 0.7499,
+      "step": 2724
+    },
+    {
+      "epoch": 1.6447328705101116,
+      "grad_norm": 0.11651081591844559,
+      "learning_rate": 9.310917393109174e-05,
+      "loss": 0.6448,
+      "step": 2725
+    },
+    {
+      "epoch": 1.645336552973136,
+      "grad_norm": 0.11762251704931259,
+      "learning_rate": 9.306766293067663e-05,
+      "loss": 0.6679,
+      "step": 2726
+    },
+    {
+      "epoch": 1.6459402354361605,
+      "grad_norm": 0.11654434353113174,
+      "learning_rate": 9.302615193026153e-05,
+      "loss": 0.9322,
+      "step": 2727
+    },
+    {
+      "epoch": 1.646543917899185,
+      "grad_norm": 0.11348237842321396,
+      "learning_rate": 9.298464092984642e-05,
+      "loss": 0.7319,
+      "step": 2728
+    },
+    {
+      "epoch": 1.6471476003622096,
+      "grad_norm": 0.19426052272319794,
+      "learning_rate": 9.29431299294313e-05,
+      "loss": 0.7589,
+      "step": 2729
+    },
+    {
+      "epoch": 1.6477512828252339,
+      "grad_norm": 0.1138731986284256,
+      "learning_rate": 9.29016189290162e-05,
+      "loss": 0.6751,
+      "step": 2730
+    },
+    {
+      "epoch": 1.6483549652882585,
+      "grad_norm": 0.11389704793691635,
+      "learning_rate": 9.286010792860108e-05,
+      "loss": 0.8278,
+      "step": 2731
+    },
+    {
+      "epoch": 1.6489586477512828,
+      "grad_norm": 0.13111789524555206,
+      "learning_rate": 9.281859692818598e-05,
+      "loss": 0.8243,
+      "step": 2732
+    },
+    {
+      "epoch": 1.6495623302143074,
+      "grad_norm": 0.10878365486860275,
+      "learning_rate": 9.277708592777086e-05,
+      "loss": 0.9464,
+      "step": 2733
+    },
+    {
+      "epoch": 1.6501660126773317,
+      "grad_norm": 0.11654222756624222,
+      "learning_rate": 9.273557492735575e-05,
+      "loss": 0.7815,
+      "step": 2734
+    },
+    {
+      "epoch": 1.6507696951403563,
+      "grad_norm": 0.10811632871627808,
+      "learning_rate": 9.269406392694065e-05,
+      "loss": 0.7112,
+      "step": 2735
+    },
+    {
+      "epoch": 1.6513733776033805,
+      "grad_norm": 0.1144120842218399,
+      "learning_rate": 9.265255292652554e-05,
+      "loss": 0.6915,
+      "step": 2736
+    },
+    {
+      "epoch": 1.6519770600664052,
+      "grad_norm": 0.12281805276870728,
+      "learning_rate": 9.261104192611042e-05,
+      "loss": 0.7846,
+      "step": 2737
+    },
+    {
+      "epoch": 1.6525807425294294,
+      "grad_norm": 0.11404890567064285,
+      "learning_rate": 9.256953092569531e-05,
+      "loss": 0.6841,
+      "step": 2738
+    },
+    {
+      "epoch": 1.653184424992454,
+      "grad_norm": 0.10799200087785721,
+      "learning_rate": 9.252801992528021e-05,
+      "loss": 0.7854,
+      "step": 2739
+    },
+    {
+      "epoch": 1.6537881074554783,
+      "grad_norm": 0.11798325926065445,
+      "learning_rate": 9.24865089248651e-05,
+      "loss": 0.6409,
+      "step": 2740
+    },
+    {
+      "epoch": 1.654391789918503,
+      "grad_norm": 0.11283228546380997,
+      "learning_rate": 9.244499792444998e-05,
+      "loss": 0.7441,
+      "step": 2741
+    },
+    {
+      "epoch": 1.6549954723815272,
+      "grad_norm": 0.11868225038051605,
+      "learning_rate": 9.240348692403488e-05,
+      "loss": 0.7046,
+      "step": 2742
+    },
+    {
+      "epoch": 1.655599154844552,
+      "grad_norm": 0.11451592296361923,
+      "learning_rate": 9.236197592361975e-05,
+      "loss": 0.677,
+      "step": 2743
+    },
+    {
+      "epoch": 1.6562028373075761,
+      "grad_norm": 0.11685509234666824,
+      "learning_rate": 9.232046492320465e-05,
+      "loss": 0.6602,
+      "step": 2744
+    },
+    {
+      "epoch": 1.6568065197706008,
+      "grad_norm": 0.11831925809383392,
+      "learning_rate": 9.227895392278954e-05,
+      "loss": 0.7313,
+      "step": 2745
+    },
+    {
+      "epoch": 1.657410202233625,
+      "grad_norm": 0.11828596144914627,
+      "learning_rate": 9.223744292237443e-05,
+      "loss": 0.5914,
+      "step": 2746
+    },
+    {
+      "epoch": 1.6580138846966497,
+      "grad_norm": 0.12427350878715515,
+      "learning_rate": 9.219593192195933e-05,
+      "loss": 0.6278,
+      "step": 2747
+    },
+    {
+      "epoch": 1.658617567159674,
+      "grad_norm": 0.12681810557842255,
+      "learning_rate": 9.215442092154421e-05,
+      "loss": 0.6496,
+      "step": 2748
+    },
+    {
+      "epoch": 1.6592212496226986,
+      "grad_norm": 0.1341797262430191,
+      "learning_rate": 9.211290992112911e-05,
+      "loss": 0.6372,
+      "step": 2749
+    },
+    {
+      "epoch": 1.6598249320857228,
+      "grad_norm": 0.1370631456375122,
+      "learning_rate": 9.207139892071399e-05,
+      "loss": 0.6669,
+      "step": 2750
+    },
+    {
+      "epoch": 1.6604286145487475,
+      "grad_norm": 0.14159443974494934,
+      "learning_rate": 9.202988792029889e-05,
+      "loss": 0.5644,
+      "step": 2751
+    },
+    {
+      "epoch": 1.6610322970117717,
+      "grad_norm": 0.13964007794857025,
+      "learning_rate": 9.198837691988377e-05,
+      "loss": 0.6035,
+      "step": 2752
+    },
+    {
+      "epoch": 1.6616359794747964,
+      "grad_norm": 0.154313325881958,
+      "learning_rate": 9.194686591946866e-05,
+      "loss": 0.5195,
+      "step": 2753
+    },
+    {
+      "epoch": 1.6622396619378206,
+      "grad_norm": 0.14789967238903046,
+      "learning_rate": 9.190535491905356e-05,
+      "loss": 0.5039,
+      "step": 2754
+    },
+    {
+      "epoch": 1.6628433444008452,
+      "grad_norm": 0.1544695347547531,
+      "learning_rate": 9.186384391863845e-05,
+      "loss": 0.4487,
+      "step": 2755
+    },
+    {
+      "epoch": 1.6634470268638695,
+      "grad_norm": 0.15225286781787872,
+      "learning_rate": 9.182233291822333e-05,
+      "loss": 0.4129,
+      "step": 2756
+    },
+    {
+      "epoch": 1.6640507093268941,
+      "grad_norm": 0.16880232095718384,
+      "learning_rate": 9.178082191780822e-05,
+      "loss": 0.3468,
+      "step": 2757
+    },
+    {
+      "epoch": 1.6646543917899184,
+      "grad_norm": 0.10654531419277191,
+      "learning_rate": 9.173931091739312e-05,
+      "loss": 0.6731,
+      "step": 2758
+    },
+    {
+      "epoch": 1.665258074252943,
+      "grad_norm": 0.12524224817752838,
+      "learning_rate": 9.1697799916978e-05,
+      "loss": 0.7284,
+      "step": 2759
+    },
+    {
+      "epoch": 1.6658617567159673,
+      "grad_norm": 0.11452171951532364,
+      "learning_rate": 9.165628891656289e-05,
+      "loss": 0.7422,
+      "step": 2760
+    },
+    {
+      "epoch": 1.6670691216420161,
+      "grad_norm": 0.11638887971639633,
+      "learning_rate": 9.161477791614779e-05,
+      "loss": 0.7024,
+      "step": 2761
+    },
+    {
+      "epoch": 1.6676728041050408,
+      "grad_norm": 0.1171191856265068,
+      "learning_rate": 9.157326691573267e-05,
+      "loss": 0.6618,
+      "step": 2762
+    },
+    {
+      "epoch": 1.668276486568065,
+      "grad_norm": 0.11434746533632278,
+      "learning_rate": 9.153175591531757e-05,
+      "loss": 0.876,
+      "step": 2763
+    },
+    {
+      "epoch": 1.6688801690310897,
+      "grad_norm": 0.11305181682109833,
+      "learning_rate": 9.149024491490245e-05,
+      "loss": 0.8247,
+      "step": 2764
+    },
+    {
+      "epoch": 1.669483851494114,
+      "grad_norm": 0.10942018032073975,
+      "learning_rate": 9.144873391448734e-05,
+      "loss": 0.6116,
+      "step": 2765
+    },
+    {
+      "epoch": 1.6700875339571386,
+      "grad_norm": 0.14481772482395172,
+      "learning_rate": 9.140722291407224e-05,
+      "loss": 0.7061,
+      "step": 2766
+    },
+    {
+      "epoch": 1.6706912164201628,
+      "grad_norm": 0.12722566723823547,
+      "learning_rate": 9.136571191365713e-05,
+      "loss": 0.6226,
+      "step": 2767
+    },
+    {
+      "epoch": 1.6712948988831875,
+      "grad_norm": 0.12751980125904083,
+      "learning_rate": 9.132420091324201e-05,
+      "loss": 0.7537,
+      "step": 2768
+    },
+    {
+      "epoch": 1.671898581346212,
+      "grad_norm": 0.1445055454969406,
+      "learning_rate": 9.12826899128269e-05,
+      "loss": 0.7353,
+      "step": 2769
+    },
+    {
+      "epoch": 1.6725022638092364,
+      "grad_norm": 0.15946505963802338,
+      "learning_rate": 9.12411789124118e-05,
+      "loss": 0.6124,
+      "step": 2770
+    },
+    {
+      "epoch": 1.6731059462722608,
+      "grad_norm": 0.11702273786067963,
+      "learning_rate": 9.119966791199668e-05,
+      "loss": 0.7526,
+      "step": 2771
+    },
+    {
+      "epoch": 1.6737096287352853,
+      "grad_norm": 0.12233686447143555,
+      "learning_rate": 9.115815691158157e-05,
+      "loss": 0.6447,
+      "step": 2772
+    },
+    {
+      "epoch": 1.6743133111983097,
+      "grad_norm": 0.12842793762683868,
+      "learning_rate": 9.111664591116647e-05,
+      "loss": 0.8035,
+      "step": 2773
+    },
+    {
+      "epoch": 1.6749169936613342,
+      "grad_norm": 0.27307209372520447,
+      "learning_rate": 9.107513491075136e-05,
+      "loss": 0.8121,
+      "step": 2774
+    },
+    {
+      "epoch": 1.6755206761243586,
+      "grad_norm": 0.12133491784334183,
+      "learning_rate": 9.103362391033624e-05,
+      "loss": 0.6252,
+      "step": 2775
+    },
+    {
+      "epoch": 1.676124358587383,
+      "grad_norm": 0.12983252108097076,
+      "learning_rate": 9.099211290992113e-05,
+      "loss": 0.6484,
+      "step": 2776
+    },
+    {
+      "epoch": 1.6767280410504075,
+      "grad_norm": 0.12295020371675491,
+      "learning_rate": 9.095060190950603e-05,
+      "loss": 0.6794,
+      "step": 2777
+    },
+    {
+      "epoch": 1.677331723513432,
+      "grad_norm": 0.12450043857097626,
+      "learning_rate": 9.090909090909092e-05,
+      "loss": 0.7732,
+      "step": 2778
+    },
+    {
+      "epoch": 1.6779354059764564,
+      "grad_norm": 0.15557287633419037,
+      "learning_rate": 9.08675799086758e-05,
+      "loss": 0.601,
+      "step": 2779
+    },
+    {
+      "epoch": 1.6785390884394809,
+      "grad_norm": 0.1215049996972084,
+      "learning_rate": 9.08260689082607e-05,
+      "loss": 0.6817,
+      "step": 2780
+    },
+    {
+      "epoch": 1.6791427709025053,
+      "grad_norm": 0.11101005971431732,
+      "learning_rate": 9.078455790784558e-05,
+      "loss": 0.5621,
+      "step": 2781
+    },
+    {
+      "epoch": 1.6797464533655297,
+      "grad_norm": 0.11614114046096802,
+      "learning_rate": 9.074304690743048e-05,
+      "loss": 0.6261,
+      "step": 2782
+    },
+    {
+      "epoch": 1.6803501358285542,
+      "grad_norm": 0.11843936145305634,
+      "learning_rate": 9.070153590701536e-05,
+      "loss": 0.8192,
+      "step": 2783
+    },
+    {
+      "epoch": 1.6809538182915786,
+      "grad_norm": 0.12000729143619537,
+      "learning_rate": 9.066002490660025e-05,
+      "loss": 0.6353,
+      "step": 2784
+    },
+    {
+      "epoch": 1.681557500754603,
+      "grad_norm": 0.11501980572938919,
+      "learning_rate": 9.061851390618515e-05,
+      "loss": 0.6151,
+      "step": 2785
+    },
+    {
+      "epoch": 1.6821611832176275,
+      "grad_norm": 0.13076992332935333,
+      "learning_rate": 9.057700290577004e-05,
+      "loss": 0.7825,
+      "step": 2786
+    },
+    {
+      "epoch": 1.682764865680652,
+      "grad_norm": 0.13332267105579376,
+      "learning_rate": 9.053549190535492e-05,
+      "loss": 0.8015,
+      "step": 2787
+    },
+    {
+      "epoch": 1.6833685481436764,
+      "grad_norm": 0.12166225165128708,
+      "learning_rate": 9.049398090493981e-05,
+      "loss": 0.6739,
+      "step": 2788
+    },
+    {
+      "epoch": 1.6839722306067009,
+      "grad_norm": 0.13338027894496918,
+      "learning_rate": 9.045246990452471e-05,
+      "loss": 0.6793,
+      "step": 2789
+    },
+    {
+      "epoch": 1.6845759130697253,
+      "grad_norm": 0.11463173478841782,
+      "learning_rate": 9.041095890410958e-05,
+      "loss": 0.535,
+      "step": 2790
+    },
+    {
+      "epoch": 1.6851795955327498,
+      "grad_norm": 0.11622385680675507,
+      "learning_rate": 9.036944790369448e-05,
+      "loss": 0.6049,
+      "step": 2791
+    },
+    {
+      "epoch": 1.6857832779957742,
+      "grad_norm": 0.1293102502822876,
+      "learning_rate": 9.032793690327938e-05,
+      "loss": 0.6065,
+      "step": 2792
+    },
+    {
+      "epoch": 1.6863869604587987,
+      "grad_norm": 0.12283488363027573,
+      "learning_rate": 9.028642590286426e-05,
+      "loss": 0.5595,
+      "step": 2793
+    },
+    {
+      "epoch": 1.686990642921823,
+      "grad_norm": 0.13364657759666443,
+      "learning_rate": 9.024491490244916e-05,
+      "loss": 0.5544,
+      "step": 2794
+    },
+    {
+      "epoch": 1.6875943253848475,
+      "grad_norm": 0.1329110860824585,
+      "learning_rate": 9.020340390203404e-05,
+      "loss": 0.584,
+      "step": 2795
+    },
+    {
+      "epoch": 1.688198007847872,
+      "grad_norm": 0.13042087852954865,
+      "learning_rate": 9.016189290161894e-05,
+      "loss": 0.5376,
+      "step": 2796
+    },
+    {
+      "epoch": 1.6888016903108964,
+      "grad_norm": 0.13576170802116394,
+      "learning_rate": 9.012038190120381e-05,
+      "loss": 0.4989,
+      "step": 2797
+    },
+    {
+      "epoch": 1.6894053727739209,
+      "grad_norm": 0.14754711091518402,
+      "learning_rate": 9.007887090078871e-05,
+      "loss": 0.5445,
+      "step": 2798
+    },
+    {
+      "epoch": 1.6900090552369453,
+      "grad_norm": 0.14908315241336823,
+      "learning_rate": 9.003735990037361e-05,
+      "loss": 0.5666,
+      "step": 2799
+    },
+    {
+      "epoch": 1.6906127376999698,
+      "grad_norm": 0.1546728014945984,
+      "learning_rate": 8.999584889995849e-05,
+      "loss": 0.472,
+      "step": 2800
+    },
+    {
+      "epoch": 1.6912164201629942,
+      "grad_norm": 0.16218960285186768,
+      "learning_rate": 8.995433789954339e-05,
+      "loss": 0.3952,
+      "step": 2801
+    },
+    {
+      "epoch": 1.6918201026260187,
+      "grad_norm": 0.1644655466079712,
+      "learning_rate": 8.991282689912827e-05,
+      "loss": 0.4062,
+      "step": 2802
+    },
+    {
+      "epoch": 1.6924237850890431,
+      "grad_norm": 0.1662607192993164,
+      "learning_rate": 8.987131589871316e-05,
+      "loss": 0.3727,
+      "step": 2803
+    },
+    {
+      "epoch": 1.6930274675520676,
+      "grad_norm": 0.1667158454656601,
+      "learning_rate": 8.982980489829805e-05,
+      "loss": 0.3083,
+      "step": 2804
+    },
+    {
+      "epoch": 1.693631150015092,
+      "grad_norm": 0.16382873058319092,
+      "learning_rate": 8.978829389788295e-05,
+      "loss": 0.2529,
+      "step": 2805
+    },
+    {
+      "epoch": 1.6942348324781165,
+      "grad_norm": 0.1675097644329071,
+      "learning_rate": 8.974678289746783e-05,
+      "loss": 0.2192,
+      "step": 2806
+    },
+    {
+      "epoch": 1.694838514941141,
+      "grad_norm": 0.13541610538959503,
+      "learning_rate": 8.970527189705272e-05,
+      "loss": 0.6514,
+      "step": 2807
+    },
+    {
+      "epoch": 1.6954421974041654,
+      "grad_norm": 0.13940273225307465,
+      "learning_rate": 8.966376089663762e-05,
+      "loss": 0.6512,
+      "step": 2808
+    },
+    {
+      "epoch": 1.69604587986719,
+      "grad_norm": 0.12449961155653,
+      "learning_rate": 8.962224989622249e-05,
+      "loss": 0.6299,
+      "step": 2809
+    },
+    {
+      "epoch": 1.6966495623302142,
+      "grad_norm": 0.14902494847774506,
+      "learning_rate": 8.958073889580739e-05,
+      "loss": 0.8532,
+      "step": 2810
+    },
+    {
+      "epoch": 1.697253244793239,
+      "grad_norm": 0.14396223425865173,
+      "learning_rate": 8.953922789539228e-05,
+      "loss": 0.782,
+      "step": 2811
+    },
+    {
+      "epoch": 1.6978569272562631,
+      "grad_norm": 0.1220279112458229,
+      "learning_rate": 8.949771689497717e-05,
+      "loss": 0.6884,
+      "step": 2812
+    },
+    {
+      "epoch": 1.6984606097192878,
+      "grad_norm": 0.12583515048027039,
+      "learning_rate": 8.945620589456207e-05,
+      "loss": 0.7678,
+      "step": 2813
+    },
+    {
+      "epoch": 1.699064292182312,
+      "grad_norm": 0.12447942793369293,
+      "learning_rate": 8.941469489414695e-05,
+      "loss": 0.6479,
+      "step": 2814
+    },
+    {
+      "epoch": 1.6996679746453367,
+      "grad_norm": 0.1457148939371109,
+      "learning_rate": 8.937318389373184e-05,
+      "loss": 0.6261,
+      "step": 2815
+    },
+    {
+      "epoch": 1.700271657108361,
+      "grad_norm": 0.11962111294269562,
+      "learning_rate": 8.933167289331673e-05,
+      "loss": 0.6798,
+      "step": 2816
+    },
+    {
+      "epoch": 1.7008753395713856,
+      "grad_norm": 0.13384521007537842,
+      "learning_rate": 8.929016189290163e-05,
+      "loss": 0.7494,
+      "step": 2817
+    },
+    {
+      "epoch": 1.7014790220344098,
+      "grad_norm": 0.12299605458974838,
+      "learning_rate": 8.924865089248651e-05,
+      "loss": 0.6161,
+      "step": 2818
+    },
+    {
+      "epoch": 1.7020827044974345,
+      "grad_norm": 0.13254275918006897,
+      "learning_rate": 8.92071398920714e-05,
+      "loss": 0.693,
+      "step": 2819
+    },
+    {
+      "epoch": 1.7026863869604587,
+      "grad_norm": 0.13070107996463776,
+      "learning_rate": 8.91656288916563e-05,
+      "loss": 0.5099,
+      "step": 2820
+    },
+    {
+      "epoch": 1.7032900694234834,
+      "grad_norm": 0.13564686477184296,
+      "learning_rate": 8.912411789124119e-05,
+      "loss": 0.6848,
+      "step": 2821
+    },
+    {
+      "epoch": 1.7038937518865076,
+      "grad_norm": 0.12641029059886932,
+      "learning_rate": 8.908260689082607e-05,
+      "loss": 0.655,
+      "step": 2822
+    },
+    {
+      "epoch": 1.7044974343495323,
+      "grad_norm": 0.11833371967077255,
+      "learning_rate": 8.904109589041096e-05,
+      "loss": 0.6897,
+      "step": 2823
+    },
+    {
+      "epoch": 1.7051011168125565,
+      "grad_norm": 0.1325460523366928,
+      "learning_rate": 8.899958488999586e-05,
+      "loss": 0.677,
+      "step": 2824
+    },
+    {
+      "epoch": 1.7057047992755812,
+      "grad_norm": 0.13542653620243073,
+      "learning_rate": 8.895807388958074e-05,
+      "loss": 0.636,
+      "step": 2825
+    },
+    {
+      "epoch": 1.7063084817386054,
+      "grad_norm": 0.1281089037656784,
+      "learning_rate": 8.891656288916563e-05,
+      "loss": 0.6886,
+      "step": 2826
+    },
+    {
+      "epoch": 1.70691216420163,
+      "grad_norm": 0.11729513853788376,
+      "learning_rate": 8.887505188875053e-05,
+      "loss": 0.6064,
+      "step": 2827
+    },
+    {
+      "epoch": 1.7075158466646543,
+      "grad_norm": 0.12129988521337509,
+      "learning_rate": 8.88335408883354e-05,
+      "loss": 0.6609,
+      "step": 2828
+    },
+    {
+      "epoch": 1.708119529127679,
+      "grad_norm": 0.12838177382946014,
+      "learning_rate": 8.87920298879203e-05,
+      "loss": 0.615,
+      "step": 2829
+    },
+    {
+      "epoch": 1.7087232115907032,
+      "grad_norm": 0.12746667861938477,
+      "learning_rate": 8.875051888750519e-05,
+      "loss": 0.6424,
+      "step": 2830
+    },
+    {
+      "epoch": 1.7093268940537278,
+      "grad_norm": 0.12252197414636612,
+      "learning_rate": 8.870900788709008e-05,
+      "loss": 0.588,
+      "step": 2831
+    },
+    {
+      "epoch": 1.709930576516752,
+      "grad_norm": 0.11774613708257675,
+      "learning_rate": 8.866749688667498e-05,
+      "loss": 0.8513,
+      "step": 2832
+    },
+    {
+      "epoch": 1.7105342589797767,
+      "grad_norm": 0.13314837217330933,
+      "learning_rate": 8.862598588625986e-05,
+      "loss": 0.7945,
+      "step": 2833
+    },
+    {
+      "epoch": 1.711137941442801,
+      "grad_norm": 0.14223603904247284,
+      "learning_rate": 8.858447488584475e-05,
+      "loss": 0.6475,
+      "step": 2834
+    },
+    {
+      "epoch": 1.7117416239058256,
+      "grad_norm": 0.1323416829109192,
+      "learning_rate": 8.854296388542964e-05,
+      "loss": 0.9064,
+      "step": 2835
+    },
+    {
+      "epoch": 1.7123453063688499,
+      "grad_norm": 0.13644298911094666,
+      "learning_rate": 8.850145288501454e-05,
+      "loss": 0.6499,
+      "step": 2836
+    },
+    {
+      "epoch": 1.7129489888318745,
+      "grad_norm": 0.12705856561660767,
+      "learning_rate": 8.845994188459942e-05,
+      "loss": 0.846,
+      "step": 2837
+    },
+    {
+      "epoch": 1.7135526712948987,
+      "grad_norm": 0.13383372128009796,
+      "learning_rate": 8.841843088418431e-05,
+      "loss": 0.671,
+      "step": 2838
+    },
+    {
+      "epoch": 1.7141563537579234,
+      "grad_norm": 0.12026913464069366,
+      "learning_rate": 8.837691988376921e-05,
+      "loss": 0.6495,
+      "step": 2839
+    },
+    {
+      "epoch": 1.7147600362209476,
+      "grad_norm": 0.12273270636796951,
+      "learning_rate": 8.83354088833541e-05,
+      "loss": 0.679,
+      "step": 2840
+    },
+    {
+      "epoch": 1.7153637186839723,
+      "grad_norm": 0.1200748011469841,
+      "learning_rate": 8.829389788293898e-05,
+      "loss": 0.6017,
+      "step": 2841
+    },
+    {
+      "epoch": 1.7159674011469965,
+      "grad_norm": 0.11923687905073166,
+      "learning_rate": 8.825238688252387e-05,
+      "loss": 0.5635,
+      "step": 2842
+    },
+    {
+      "epoch": 1.7165710836100212,
+      "grad_norm": 0.1298258900642395,
+      "learning_rate": 8.821087588210877e-05,
+      "loss": 0.5879,
+      "step": 2843
+    },
+    {
+      "epoch": 1.7171747660730454,
+      "grad_norm": 0.13525626063346863,
+      "learning_rate": 8.816936488169366e-05,
+      "loss": 0.5725,
+      "step": 2844
+    },
+    {
+      "epoch": 1.71777844853607,
+      "grad_norm": 0.13261830806732178,
+      "learning_rate": 8.812785388127854e-05,
+      "loss": 0.511,
+      "step": 2845
+    },
+    {
+      "epoch": 1.7183821309990943,
+      "grad_norm": 0.1352899670600891,
+      "learning_rate": 8.808634288086344e-05,
+      "loss": 0.5725,
+      "step": 2846
+    },
+    {
+      "epoch": 1.718985813462119,
+      "grad_norm": 0.12940122187137604,
+      "learning_rate": 8.804483188044831e-05,
+      "loss": 0.5499,
+      "step": 2847
+    },
+    {
+      "epoch": 1.7195894959251434,
+      "grad_norm": 0.16278207302093506,
+      "learning_rate": 8.800332088003321e-05,
+      "loss": 0.6085,
+      "step": 2848
+    },
+    {
+      "epoch": 1.7201931783881679,
+      "grad_norm": 0.16731347143650055,
+      "learning_rate": 8.79618098796181e-05,
+      "loss": 0.57,
+      "step": 2849
+    },
+    {
+      "epoch": 1.7207968608511923,
+      "grad_norm": 0.15214033424854279,
+      "learning_rate": 8.792029887920299e-05,
+      "loss": 0.5269,
+      "step": 2850
+    },
+    {
+      "epoch": 1.7214005433142168,
+      "grad_norm": 0.1611013114452362,
+      "learning_rate": 8.787878787878789e-05,
+      "loss": 0.5705,
+      "step": 2851
+    },
+    {
+      "epoch": 1.7220042257772412,
+      "grad_norm": 0.16054251790046692,
+      "learning_rate": 8.783727687837277e-05,
+      "loss": 0.4106,
+      "step": 2852
+    },
+    {
+      "epoch": 1.7226079082402657,
+      "grad_norm": 0.16081520915031433,
+      "learning_rate": 8.779576587795766e-05,
+      "loss": 0.3376,
+      "step": 2853
+    },
+    {
+      "epoch": 1.72321159070329,
+      "grad_norm": 0.18261933326721191,
+      "learning_rate": 8.775425487754255e-05,
+      "loss": 0.3801,
+      "step": 2854
+    },
+    {
+      "epoch": 1.7238152731663146,
+      "grad_norm": 0.1675270050764084,
+      "learning_rate": 8.771274387712745e-05,
+      "loss": 0.3022,
+      "step": 2855
+    },
+    {
+      "epoch": 1.724418955629339,
+      "grad_norm": 0.17773354053497314,
+      "learning_rate": 8.767123287671233e-05,
+      "loss": 0.2563,
+      "step": 2856
+    },
+    {
+      "epoch": 1.7250226380923634,
+      "grad_norm": 0.12368903309106827,
+      "learning_rate": 8.762972187629722e-05,
+      "loss": 0.6228,
+      "step": 2857
+    },
+    {
+      "epoch": 1.725626320555388,
+      "grad_norm": 0.12369142472743988,
+      "learning_rate": 8.758821087588212e-05,
+      "loss": 0.5617,
+      "step": 2858
+    },
+    {
+      "epoch": 1.7262300030184123,
+      "grad_norm": 0.1579550802707672,
+      "learning_rate": 8.7546699875467e-05,
+      "loss": 0.6571,
+      "step": 2859
+    },
+    {
+      "epoch": 1.7268336854814368,
+      "grad_norm": 0.13555946946144104,
+      "learning_rate": 8.75051888750519e-05,
+      "loss": 0.7208,
+      "step": 2860
+    },
+    {
+      "epoch": 1.7274373679444612,
+      "grad_norm": 0.14086709916591644,
+      "learning_rate": 8.746367787463678e-05,
+      "loss": 0.801,
+      "step": 2861
+    },
+    {
+      "epoch": 1.7280410504074857,
+      "grad_norm": 0.16081978380680084,
+      "learning_rate": 8.742216687422168e-05,
+      "loss": 0.6317,
+      "step": 2862
+    },
+    {
+      "epoch": 1.7286447328705101,
+      "grad_norm": 0.1376749724149704,
+      "learning_rate": 8.738065587380657e-05,
+      "loss": 0.6912,
+      "step": 2863
+    },
+    {
+      "epoch": 1.7292484153335346,
+      "grad_norm": 0.1608782410621643,
+      "learning_rate": 8.733914487339145e-05,
+      "loss": 0.6217,
+      "step": 2864
+    },
+    {
+      "epoch": 1.729852097796559,
+      "grad_norm": 0.1334768831729889,
+      "learning_rate": 8.729763387297635e-05,
+      "loss": 0.9018,
+      "step": 2865
+    },
+    {
+      "epoch": 1.7304557802595835,
+      "grad_norm": 0.1475462019443512,
+      "learning_rate": 8.725612287256123e-05,
+      "loss": 0.5775,
+      "step": 2866
+    },
+    {
+      "epoch": 1.731059462722608,
+      "grad_norm": 0.26409459114074707,
+      "learning_rate": 8.721461187214613e-05,
+      "loss": 0.6337,
+      "step": 2867
+    },
+    {
+      "epoch": 1.7316631451856324,
+      "grad_norm": 0.1443539261817932,
+      "learning_rate": 8.717310087173101e-05,
+      "loss": 0.619,
+      "step": 2868
+    },
+    {
+      "epoch": 1.7322668276486568,
+      "grad_norm": 0.1351659595966339,
+      "learning_rate": 8.71315898713159e-05,
+      "loss": 0.6511,
+      "step": 2869
+    },
+    {
+      "epoch": 1.7328705101116813,
+      "grad_norm": 0.1312374323606491,
+      "learning_rate": 8.70900788709008e-05,
+      "loss": 0.6636,
+      "step": 2870
+    },
+    {
+      "epoch": 1.7334741925747057,
+      "grad_norm": 0.13947537541389465,
+      "learning_rate": 8.704856787048569e-05,
+      "loss": 0.682,
+      "step": 2871
+    },
+    {
+      "epoch": 1.7340778750377301,
+      "grad_norm": 0.1194947212934494,
+      "learning_rate": 8.700705687007057e-05,
+      "loss": 0.6474,
+      "step": 2872
+    },
+    {
+      "epoch": 1.7346815575007546,
+      "grad_norm": 0.11692965775728226,
+      "learning_rate": 8.696554586965546e-05,
+      "loss": 0.6318,
+      "step": 2873
+    },
+    {
+      "epoch": 1.735285239963779,
+      "grad_norm": 0.13894155621528625,
+      "learning_rate": 8.692403486924036e-05,
+      "loss": 0.642,
+      "step": 2874
+    },
+    {
+      "epoch": 1.7358889224268035,
+      "grad_norm": 0.1313973218202591,
+      "learning_rate": 8.688252386882524e-05,
+      "loss": 0.8355,
+      "step": 2875
+    },
+    {
+      "epoch": 1.736492604889828,
+      "grad_norm": 0.15785734355449677,
+      "learning_rate": 8.684101286841013e-05,
+      "loss": 0.7546,
+      "step": 2876
+    },
+    {
+      "epoch": 1.7370962873528524,
+      "grad_norm": 0.12201385200023651,
+      "learning_rate": 8.679950186799503e-05,
+      "loss": 0.5731,
+      "step": 2877
+    },
+    {
+      "epoch": 1.7376999698158768,
+      "grad_norm": 0.13196420669555664,
+      "learning_rate": 8.67579908675799e-05,
+      "loss": 0.6438,
+      "step": 2878
+    },
+    {
+      "epoch": 1.7383036522789013,
+      "grad_norm": 0.1272536814212799,
+      "learning_rate": 8.67164798671648e-05,
+      "loss": 0.5866,
+      "step": 2879
+    },
+    {
+      "epoch": 1.7389073347419257,
+      "grad_norm": 0.1259777694940567,
+      "learning_rate": 8.667496886674969e-05,
+      "loss": 0.677,
+      "step": 2880
+    },
+    {
+      "epoch": 1.7395110172049502,
+      "grad_norm": 0.13071627914905548,
+      "learning_rate": 8.663345786633458e-05,
+      "loss": 0.7244,
+      "step": 2881
+    },
+    {
+      "epoch": 1.7401146996679746,
+      "grad_norm": 0.12744002044200897,
+      "learning_rate": 8.659194686591948e-05,
+      "loss": 0.9371,
+      "step": 2882
+    },
+    {
+      "epoch": 1.740718382130999,
+      "grad_norm": 0.137882798910141,
+      "learning_rate": 8.655043586550436e-05,
+      "loss": 0.6717,
+      "step": 2883
+    },
+    {
+      "epoch": 1.7413220645940235,
+      "grad_norm": 0.13243485987186432,
+      "learning_rate": 8.650892486508926e-05,
+      "loss": 0.7217,
+      "step": 2884
+    },
+    {
+      "epoch": 1.741925747057048,
+      "grad_norm": 0.13238345086574554,
+      "learning_rate": 8.646741386467414e-05,
+      "loss": 0.7141,
+      "step": 2885
+    },
+    {
+      "epoch": 1.7425294295200724,
+      "grad_norm": 0.13204292953014374,
+      "learning_rate": 8.642590286425904e-05,
+      "loss": 0.7224,
+      "step": 2886
+    },
+    {
+      "epoch": 1.7431331119830968,
+      "grad_norm": 0.12200792878866196,
+      "learning_rate": 8.638439186384392e-05,
+      "loss": 0.585,
+      "step": 2887
+    },
+    {
+      "epoch": 1.7437367944461213,
+      "grad_norm": 0.12141503393650055,
+      "learning_rate": 8.634288086342881e-05,
+      "loss": 0.9708,
+      "step": 2888
+    },
+    {
+      "epoch": 1.7443404769091457,
+      "grad_norm": 0.14114977419376373,
+      "learning_rate": 8.630136986301371e-05,
+      "loss": 0.5942,
+      "step": 2889
+    },
+    {
+      "epoch": 1.7449441593721704,
+      "grad_norm": 0.12927395105361938,
+      "learning_rate": 8.62598588625986e-05,
+      "loss": 0.6257,
+      "step": 2890
+    },
+    {
+      "epoch": 1.7455478418351946,
+      "grad_norm": 0.1255059540271759,
+      "learning_rate": 8.621834786218348e-05,
+      "loss": 0.5921,
+      "step": 2891
+    },
+    {
+      "epoch": 1.7461515242982193,
+      "grad_norm": 0.1200440302491188,
+      "learning_rate": 8.617683686176837e-05,
+      "loss": 0.5587,
+      "step": 2892
+    },
+    {
+      "epoch": 1.7467552067612435,
+      "grad_norm": 0.12957656383514404,
+      "learning_rate": 8.613532586135327e-05,
+      "loss": 0.5773,
+      "step": 2893
+    },
+    {
+      "epoch": 1.7473588892242682,
+      "grad_norm": 0.13232208788394928,
+      "learning_rate": 8.609381486093814e-05,
+      "loss": 0.681,
+      "step": 2894
+    },
+    {
+      "epoch": 1.7479625716872924,
+      "grad_norm": 0.12879258394241333,
+      "learning_rate": 8.605230386052304e-05,
+      "loss": 0.5747,
+      "step": 2895
+    },
+    {
+      "epoch": 1.748566254150317,
+      "grad_norm": 0.1347392499446869,
+      "learning_rate": 8.601079286010794e-05,
+      "loss": 0.5334,
+      "step": 2896
+    },
+    {
+      "epoch": 1.7491699366133413,
+      "grad_norm": 0.16375549137592316,
+      "learning_rate": 8.596928185969282e-05,
+      "loss": 0.5834,
+      "step": 2897
+    },
+    {
+      "epoch": 1.749773619076366,
+      "grad_norm": 0.14500340819358826,
+      "learning_rate": 8.592777085927772e-05,
+      "loss": 0.539,
+      "step": 2898
+    },
+    {
+      "epoch": 1.7503773015393902,
+      "grad_norm": 0.16076421737670898,
+      "learning_rate": 8.58862598588626e-05,
+      "loss": 0.5071,
+      "step": 2899
+    },
+    {
+      "epoch": 1.7509809840024149,
+      "grad_norm": 0.14887742698192596,
+      "learning_rate": 8.584474885844749e-05,
+      "loss": 0.4643,
+      "step": 2900
+    },
+    {
+      "epoch": 1.751584666465439,
+      "grad_norm": 0.16193953156471252,
+      "learning_rate": 8.580323785803237e-05,
+      "loss": 0.494,
+      "step": 2901
+    },
+    {
+      "epoch": 1.7521883489284638,
+      "grad_norm": 0.1675998717546463,
+      "learning_rate": 8.576172685761727e-05,
+      "loss": 0.3624,
+      "step": 2902
+    },
+    {
+      "epoch": 1.752792031391488,
+      "grad_norm": 0.18320401012897491,
+      "learning_rate": 8.572021585720216e-05,
+      "loss": 0.4361,
+      "step": 2903
+    },
+    {
+      "epoch": 1.7533957138545126,
+      "grad_norm": 0.1881968080997467,
+      "learning_rate": 8.567870485678705e-05,
+      "loss": 0.3647,
+      "step": 2904
+    },
+    {
+      "epoch": 1.7539993963175369,
+      "grad_norm": 0.2042665183544159,
+      "learning_rate": 8.563719385637195e-05,
+      "loss": 0.3493,
+      "step": 2905
+    },
+    {
+      "epoch": 1.7546030787805615,
+      "grad_norm": 0.18004922568798065,
+      "learning_rate": 8.559568285595683e-05,
+      "loss": 0.2261,
+      "step": 2906
+    },
+    {
+      "epoch": 1.7552067612435858,
+      "grad_norm": 0.15333791077136993,
+      "learning_rate": 8.555417185554172e-05,
+      "loss": 1.0165,
+      "step": 2907
+    },
+    {
+      "epoch": 1.7558104437066104,
+      "grad_norm": 0.1424827128648758,
+      "learning_rate": 8.551266085512661e-05,
+      "loss": 0.6105,
+      "step": 2908
+    },
+    {
+      "epoch": 1.7564141261696347,
+      "grad_norm": 0.13473393023014069,
+      "learning_rate": 8.547114985471151e-05,
+      "loss": 0.6586,
+      "step": 2909
+    },
+    {
+      "epoch": 1.7570178086326593,
+      "grad_norm": 0.1452847421169281,
+      "learning_rate": 8.54296388542964e-05,
+      "loss": 0.7158,
+      "step": 2910
+    },
+    {
+      "epoch": 1.7576214910956836,
+      "grad_norm": 0.1423470824956894,
+      "learning_rate": 8.538812785388128e-05,
+      "loss": 0.6961,
+      "step": 2911
+    },
+    {
+      "epoch": 1.7582251735587082,
+      "grad_norm": 0.12046578526496887,
+      "learning_rate": 8.534661685346618e-05,
+      "loss": 0.8319,
+      "step": 2912
+    },
+    {
+      "epoch": 1.7588288560217324,
+      "grad_norm": 0.13641320168972015,
+      "learning_rate": 8.530510585305105e-05,
+      "loss": 0.6154,
+      "step": 2913
+    },
+    {
+      "epoch": 1.7594325384847571,
+      "grad_norm": 0.1376158744096756,
+      "learning_rate": 8.526359485263595e-05,
+      "loss": 0.6259,
+      "step": 2914
+    },
+    {
+      "epoch": 1.7600362209477813,
+      "grad_norm": 0.1257615089416504,
+      "learning_rate": 8.522208385222084e-05,
+      "loss": 0.6727,
+      "step": 2915
+    },
+    {
+      "epoch": 1.760639903410806,
+      "grad_norm": 0.12842705845832825,
+      "learning_rate": 8.518057285180573e-05,
+      "loss": 0.8247,
+      "step": 2916
+    },
+    {
+      "epoch": 1.7612435858738302,
+      "grad_norm": 0.12923744320869446,
+      "learning_rate": 8.513906185139063e-05,
+      "loss": 0.7409,
+      "step": 2917
+    },
+    {
+      "epoch": 1.761847268336855,
+      "grad_norm": 0.17179526388645172,
+      "learning_rate": 8.509755085097551e-05,
+      "loss": 0.6223,
+      "step": 2918
+    },
+    {
+      "epoch": 1.7624509507998791,
+      "grad_norm": 0.1274694800376892,
+      "learning_rate": 8.50560398505604e-05,
+      "loss": 0.6175,
+      "step": 2919
+    },
+    {
+      "epoch": 1.7630546332629038,
+      "grad_norm": 0.1250637173652649,
+      "learning_rate": 8.501452885014529e-05,
+      "loss": 0.8064,
+      "step": 2920
+    },
+    {
+      "epoch": 1.763658315725928,
+      "grad_norm": 0.12304326891899109,
+      "learning_rate": 8.497301784973019e-05,
+      "loss": 0.6699,
+      "step": 2921
+    },
+    {
+      "epoch": 1.7642619981889527,
+      "grad_norm": 0.1374034285545349,
+      "learning_rate": 8.493150684931507e-05,
+      "loss": 0.6121,
+      "step": 2922
+    },
+    {
+      "epoch": 1.764865680651977,
+      "grad_norm": 0.13162750005722046,
+      "learning_rate": 8.488999584889996e-05,
+      "loss": 0.7042,
+      "step": 2923
+    },
+    {
+      "epoch": 1.7654693631150016,
+      "grad_norm": 0.13185490667819977,
+      "learning_rate": 8.484848484848486e-05,
+      "loss": 0.6263,
+      "step": 2924
+    },
+    {
+      "epoch": 1.7660730455780258,
+      "grad_norm": 0.1368507295846939,
+      "learning_rate": 8.480697384806973e-05,
+      "loss": 0.695,
+      "step": 2925
+    },
+    {
+      "epoch": 1.7666767280410505,
+      "grad_norm": 0.12848018109798431,
+      "learning_rate": 8.476546284765463e-05,
+      "loss": 0.5866,
+      "step": 2926
+    },
+    {
+      "epoch": 1.7672804105040747,
+      "grad_norm": 0.12568457424640656,
+      "learning_rate": 8.472395184723952e-05,
+      "loss": 0.7202,
+      "step": 2927
+    },
+    {
+      "epoch": 1.7678840929670994,
+      "grad_norm": 0.14278379082679749,
+      "learning_rate": 8.468244084682442e-05,
+      "loss": 0.8685,
+      "step": 2928
+    },
+    {
+      "epoch": 1.7684877754301238,
+      "grad_norm": 0.1360083371400833,
+      "learning_rate": 8.46409298464093e-05,
+      "loss": 0.6535,
+      "step": 2929
+    },
+    {
+      "epoch": 1.7690914578931483,
+      "grad_norm": 0.12538328766822815,
+      "learning_rate": 8.459941884599419e-05,
+      "loss": 0.614,
+      "step": 2930
+    },
+    {
+      "epoch": 1.7696951403561727,
+      "grad_norm": 0.13750702142715454,
+      "learning_rate": 8.455790784557909e-05,
+      "loss": 0.6894,
+      "step": 2931
+    },
+    {
+      "epoch": 1.7702988228191971,
+      "grad_norm": 0.12563706934452057,
+      "learning_rate": 8.451639684516396e-05,
+      "loss": 0.6747,
+      "step": 2932
+    },
+    {
+      "epoch": 1.7709025052822216,
+      "grad_norm": 0.14879825711250305,
+      "learning_rate": 8.447488584474886e-05,
+      "loss": 0.747,
+      "step": 2933
+    },
+    {
+      "epoch": 1.771506187745246,
+      "grad_norm": 0.12372935563325882,
+      "learning_rate": 8.443337484433375e-05,
+      "loss": 0.5759,
+      "step": 2934
+    },
+    {
+      "epoch": 1.7721098702082705,
+      "grad_norm": 0.13434575498104095,
+      "learning_rate": 8.439186384391864e-05,
+      "loss": 0.8362,
+      "step": 2935
+    },
+    {
+      "epoch": 1.772713552671295,
+      "grad_norm": 0.12838785350322723,
+      "learning_rate": 8.435035284350354e-05,
+      "loss": 0.8738,
+      "step": 2936
+    },
+    {
+      "epoch": 1.7733172351343194,
+      "grad_norm": 0.12814517319202423,
+      "learning_rate": 8.430884184308842e-05,
+      "loss": 0.7118,
+      "step": 2937
+    },
+    {
+      "epoch": 1.7739209175973438,
+      "grad_norm": 0.11639917641878128,
+      "learning_rate": 8.426733084267331e-05,
+      "loss": 0.7066,
+      "step": 2938
+    },
+    {
+      "epoch": 1.7745246000603683,
+      "grad_norm": 0.1350032538175583,
+      "learning_rate": 8.42258198422582e-05,
+      "loss": 0.7287,
+      "step": 2939
+    },
+    {
+      "epoch": 1.7751282825233927,
+      "grad_norm": 0.12561342120170593,
+      "learning_rate": 8.41843088418431e-05,
+      "loss": 0.6189,
+      "step": 2940
+    },
+    {
+      "epoch": 1.7757319649864172,
+      "grad_norm": 0.13215911388397217,
+      "learning_rate": 8.414279784142798e-05,
+      "loss": 0.5686,
+      "step": 2941
+    },
+    {
+      "epoch": 1.7763356474494416,
+      "grad_norm": 0.12100309878587723,
+      "learning_rate": 8.410128684101287e-05,
+      "loss": 0.5502,
+      "step": 2942
+    },
+    {
+      "epoch": 1.776939329912466,
+      "grad_norm": 0.1172775849699974,
+      "learning_rate": 8.405977584059777e-05,
+      "loss": 0.5211,
+      "step": 2943
+    },
+    {
+      "epoch": 1.7775430123754905,
+      "grad_norm": 0.13135451078414917,
+      "learning_rate": 8.401826484018264e-05,
+      "loss": 0.6114,
+      "step": 2944
+    },
+    {
+      "epoch": 1.778146694838515,
+      "grad_norm": 0.16509322822093964,
+      "learning_rate": 8.397675383976754e-05,
+      "loss": 0.5613,
+      "step": 2945
+    },
+    {
+      "epoch": 1.7787503773015394,
+      "grad_norm": 0.14752458035945892,
+      "learning_rate": 8.393524283935243e-05,
+      "loss": 0.5444,
+      "step": 2946
+    },
+    {
+      "epoch": 1.7793540597645638,
+      "grad_norm": 0.15376845002174377,
+      "learning_rate": 8.389373183893732e-05,
+      "loss": 0.5722,
+      "step": 2947
+    },
+    {
+      "epoch": 1.7799577422275883,
+      "grad_norm": 0.14886072278022766,
+      "learning_rate": 8.385222083852222e-05,
+      "loss": 0.5796,
+      "step": 2948
+    },
+    {
+      "epoch": 1.7805614246906127,
+      "grad_norm": 0.14707441627979279,
+      "learning_rate": 8.38107098381071e-05,
+      "loss": 0.5235,
+      "step": 2949
+    },
+    {
+      "epoch": 1.7811651071536372,
+      "grad_norm": 0.16511821746826172,
+      "learning_rate": 8.3769198837692e-05,
+      "loss": 0.6322,
+      "step": 2950
+    },
+    {
+      "epoch": 1.7817687896166616,
+      "grad_norm": 0.1711074262857437,
+      "learning_rate": 8.372768783727688e-05,
+      "loss": 0.4643,
+      "step": 2951
+    },
+    {
+      "epoch": 1.782372472079686,
+      "grad_norm": 0.1574665904045105,
+      "learning_rate": 8.368617683686178e-05,
+      "loss": 0.3831,
+      "step": 2952
+    },
+    {
+      "epoch": 1.7829761545427105,
+      "grad_norm": 0.17709819972515106,
+      "learning_rate": 8.364466583644666e-05,
+      "loss": 0.3542,
+      "step": 2953
+    },
+    {
+      "epoch": 1.783579837005735,
+      "grad_norm": 0.1887846142053604,
+      "learning_rate": 8.360315483603155e-05,
+      "loss": 0.3804,
+      "step": 2954
+    },
+    {
+      "epoch": 1.7841835194687594,
+      "grad_norm": 0.19664205610752106,
+      "learning_rate": 8.356164383561645e-05,
+      "loss": 0.3055,
+      "step": 2955
+    },
+    {
+      "epoch": 1.7847872019317839,
+      "grad_norm": 0.1803441047668457,
+      "learning_rate": 8.352013283520133e-05,
+      "loss": 0.2447,
+      "step": 2956
+    },
+    {
+      "epoch": 1.7853908843948083,
+      "grad_norm": 0.13248445093631744,
+      "learning_rate": 8.347862183478622e-05,
+      "loss": 0.76,
+      "step": 2957
+    },
+    {
+      "epoch": 1.7859945668578328,
+      "grad_norm": 0.12137053906917572,
+      "learning_rate": 8.343711083437111e-05,
+      "loss": 0.6176,
+      "step": 2958
+    },
+    {
+      "epoch": 1.7865982493208572,
+      "grad_norm": 0.1513027399778366,
+      "learning_rate": 8.339559983395601e-05,
+      "loss": 0.611,
+      "step": 2959
+    },
+    {
+      "epoch": 1.7872019317838816,
+      "grad_norm": 0.15577901899814606,
+      "learning_rate": 8.33540888335409e-05,
+      "loss": 0.6373,
+      "step": 2960
+    },
+    {
+      "epoch": 1.787805614246906,
+      "grad_norm": 0.14755643904209137,
+      "learning_rate": 8.331257783312578e-05,
+      "loss": 1.0106,
+      "step": 2961
+    },
+    {
+      "epoch": 1.7884092967099305,
+      "grad_norm": 0.127186581492424,
+      "learning_rate": 8.327106683271068e-05,
+      "loss": 0.7419,
+      "step": 2962
+    },
+    {
+      "epoch": 1.789012979172955,
+      "grad_norm": 0.14396396279335022,
+      "learning_rate": 8.322955583229555e-05,
+      "loss": 0.6834,
+      "step": 2963
+    },
+    {
+      "epoch": 1.7896166616359794,
+      "grad_norm": 0.12821994721889496,
+      "learning_rate": 8.318804483188045e-05,
+      "loss": 0.9047,
+      "step": 2964
+    },
+    {
+      "epoch": 1.7902203440990039,
+      "grad_norm": 0.12685029208660126,
+      "learning_rate": 8.314653383146534e-05,
+      "loss": 0.5295,
+      "step": 2965
+    },
+    {
+      "epoch": 1.7908240265620283,
+      "grad_norm": 0.16082331538200378,
+      "learning_rate": 8.310502283105023e-05,
+      "loss": 0.6977,
+      "step": 2966
+    },
+    {
+      "epoch": 1.7914277090250528,
+      "grad_norm": 0.13325533270835876,
+      "learning_rate": 8.306351183063513e-05,
+      "loss": 1.0607,
+      "step": 2967
+    },
+    {
+      "epoch": 1.7920313914880772,
+      "grad_norm": 0.12702691555023193,
+      "learning_rate": 8.302200083022001e-05,
+      "loss": 0.6663,
+      "step": 2968
+    },
+    {
+      "epoch": 1.7926350739511017,
+      "grad_norm": 0.13634531199932098,
+      "learning_rate": 8.29804898298049e-05,
+      "loss": 0.6418,
+      "step": 2969
+    },
+    {
+      "epoch": 1.7932387564141261,
+      "grad_norm": 0.11988352239131927,
+      "learning_rate": 8.293897882938979e-05,
+      "loss": 0.7847,
+      "step": 2970
+    },
+    {
+      "epoch": 1.7938424388771508,
+      "grad_norm": 0.13620884716510773,
+      "learning_rate": 8.289746782897469e-05,
+      "loss": 0.6376,
+      "step": 2971
+    },
+    {
+      "epoch": 1.794446121340175,
+      "grad_norm": 0.13789290189743042,
+      "learning_rate": 8.285595682855957e-05,
+      "loss": 1.144,
+      "step": 2972
+    },
+    {
+      "epoch": 1.7950498038031997,
+      "grad_norm": 0.12296643108129501,
+      "learning_rate": 8.281444582814446e-05,
+      "loss": 0.6537,
+      "step": 2973
+    },
+    {
+      "epoch": 1.795653486266224,
+      "grad_norm": 0.13792656362056732,
+      "learning_rate": 8.277293482772936e-05,
+      "loss": 0.628,
+      "step": 2974
+    },
+    {
+      "epoch": 1.7962571687292486,
+      "grad_norm": 0.13332706689834595,
+      "learning_rate": 8.273142382731425e-05,
+      "loss": 0.584,
+      "step": 2975
+    },
+    {
+      "epoch": 1.7968608511922728,
+      "grad_norm": 0.11899926513433456,
+      "learning_rate": 8.268991282689913e-05,
+      "loss": 0.6707,
+      "step": 2976
+    },
+    {
+      "epoch": 1.7974645336552975,
+      "grad_norm": 0.3628057539463043,
+      "learning_rate": 8.264840182648402e-05,
+      "loss": 0.6138,
+      "step": 2977
+    },
+    {
+      "epoch": 1.7980682161183217,
+      "grad_norm": 0.1424243003129959,
+      "learning_rate": 8.260689082606892e-05,
+      "loss": 0.708,
+      "step": 2978
+    },
+    {
+      "epoch": 1.7986718985813464,
+      "grad_norm": 0.13742388784885406,
+      "learning_rate": 8.25653798256538e-05,
+      "loss": 0.7981,
+      "step": 2979
+    },
+    {
+      "epoch": 1.7992755810443706,
+      "grad_norm": 0.12499826401472092,
+      "learning_rate": 8.252386882523869e-05,
+      "loss": 0.7833,
+      "step": 2980
+    },
+    {
+      "epoch": 1.7998792635073952,
+      "grad_norm": 0.13351675868034363,
+      "learning_rate": 8.248235782482359e-05,
+      "loss": 0.9222,
+      "step": 2981
+    },
+    {
+      "epoch": 1.8004829459704195,
+      "grad_norm": 0.13743843138217926,
+      "learning_rate": 8.244084682440846e-05,
+      "loss": 0.6614,
+      "step": 2982
+    },
+    {
+      "epoch": 1.8010866284334441,
+      "grad_norm": 0.14949648082256317,
+      "learning_rate": 8.239933582399336e-05,
+      "loss": 0.5954,
+      "step": 2983
+    },
+    {
+      "epoch": 1.8016903108964684,
+      "grad_norm": 0.1315966248512268,
+      "learning_rate": 8.235782482357825e-05,
+      "loss": 0.7275,
+      "step": 2984
+    },
+    {
+      "epoch": 1.802293993359493,
+      "grad_norm": 0.16671594977378845,
+      "learning_rate": 8.231631382316314e-05,
+      "loss": 0.7025,
+      "step": 2985
+    },
+    {
+      "epoch": 1.8028976758225173,
+      "grad_norm": 0.12334276735782623,
+      "learning_rate": 8.227480282274804e-05,
+      "loss": 0.6606,
+      "step": 2986
+    },
+    {
+      "epoch": 1.803501358285542,
+      "grad_norm": 0.11940109729766846,
+      "learning_rate": 8.223329182233292e-05,
+      "loss": 0.6042,
+      "step": 2987
+    },
+    {
+      "epoch": 1.8041050407485661,
+      "grad_norm": 0.1273908019065857,
+      "learning_rate": 8.219178082191781e-05,
+      "loss": 0.6492,
+      "step": 2988
+    },
+    {
+      "epoch": 1.8047087232115908,
+      "grad_norm": 0.12909753620624542,
+      "learning_rate": 8.21502698215027e-05,
+      "loss": 0.5704,
+      "step": 2989
+    },
+    {
+      "epoch": 1.805312405674615,
+      "grad_norm": 0.1314334124326706,
+      "learning_rate": 8.21087588210876e-05,
+      "loss": 0.548,
+      "step": 2990
+    },
+    {
+      "epoch": 1.8059160881376397,
+      "grad_norm": 0.14163191616535187,
+      "learning_rate": 8.206724782067248e-05,
+      "loss": 0.6069,
+      "step": 2991
+    },
+    {
+      "epoch": 1.806519770600664,
+      "grad_norm": 0.14366085827350616,
+      "learning_rate": 8.202573682025737e-05,
+      "loss": 0.5977,
+      "step": 2992
+    },
+    {
+      "epoch": 1.8071234530636886,
+      "grad_norm": 0.1461668461561203,
+      "learning_rate": 8.198422581984227e-05,
+      "loss": 0.6007,
+      "step": 2993
+    },
+    {
+      "epoch": 1.8077271355267128,
+      "grad_norm": 0.14245660603046417,
+      "learning_rate": 8.194271481942716e-05,
+      "loss": 0.5456,
+      "step": 2994
+    },
+    {
+      "epoch": 1.8083308179897375,
+      "grad_norm": 0.15712513029575348,
+      "learning_rate": 8.190120381901204e-05,
+      "loss": 0.522,
+      "step": 2995
+    },
+    {
+      "epoch": 1.8089345004527617,
+      "grad_norm": 0.14800165593624115,
+      "learning_rate": 8.185969281859693e-05,
+      "loss": 0.541,
+      "step": 2996
+    },
+    {
+      "epoch": 1.8095381829157864,
+      "grad_norm": 0.1499776840209961,
+      "learning_rate": 8.181818181818183e-05,
+      "loss": 0.5566,
+      "step": 2997
+    },
+    {
+      "epoch": 1.8101418653788106,
+      "grad_norm": 0.15904314815998077,
+      "learning_rate": 8.177667081776672e-05,
+      "loss": 0.533,
+      "step": 2998
+    },
+    {
+      "epoch": 1.8107455478418353,
+      "grad_norm": 0.17498090863227844,
+      "learning_rate": 8.17351598173516e-05,
+      "loss": 0.535,
+      "step": 2999
+    },
+    {
+      "epoch": 1.8113492303048595,
+      "grad_norm": 0.16436955332756042,
+      "learning_rate": 8.16936488169365e-05,
+      "loss": 0.5246,
+      "step": 3000
+    },
+    {
+      "epoch": 1.8113492303048595,
+      "eval_loss": 0.6211307644844055,
+      "eval_runtime": 1059.4646,
+      "eval_samples_per_second": 2.633,
+      "eval_steps_per_second": 0.329,
+      "step": 3000
+    },
+    {
+      "epoch": 1.8119529127678842,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.165213781652138e-05,
+      "loss": 0.4371,
+      "step": 3001
+    },
+    {
+      "epoch": 1.8125565952309084,
+      "grad_norm": 0.203125,
+      "learning_rate": 8.161062681610628e-05,
+      "loss": 0.4749,
+      "step": 3002
+    },
+    {
+      "epoch": 1.813160277693933,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.156911581569116e-05,
+      "loss": 0.3594,
+      "step": 3003
+    },
+    {
+      "epoch": 1.8137639601569573,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 8.152760481527605e-05,
+      "loss": 0.3817,
+      "step": 3004
+    },
+    {
+      "epoch": 1.814367642619982,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.148609381486095e-05,
+      "loss": 0.3289,
+      "step": 3005
+    },
+    {
+      "epoch": 1.8149713250830062,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.144458281444583e-05,
+      "loss": 0.276,
+      "step": 3006
+    },
+    {
+      "epoch": 1.8155750075460309,
+      "grad_norm": 0.154296875,
+      "learning_rate": 8.140307181403072e-05,
+      "loss": 0.6628,
+      "step": 3007
+    },
+    {
+      "epoch": 1.816178690009055,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 8.136156081361561e-05,
+      "loss": 0.6329,
+      "step": 3008
+    },
+    {
+      "epoch": 1.8167823724720797,
+      "grad_norm": 0.134765625,
+      "learning_rate": 8.132004981320051e-05,
+      "loss": 0.6207,
+      "step": 3009
+    },
+    {
+      "epoch": 1.8173860549351042,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 8.127853881278538e-05,
+      "loss": 0.7047,
+      "step": 3010
+    },
+    {
+      "epoch": 1.8179897373981286,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.123702781237028e-05,
+      "loss": 0.717,
+      "step": 3011
+    },
+    {
+      "epoch": 1.818593419861153,
+      "grad_norm": 0.134765625,
+      "learning_rate": 8.119551681195518e-05,
+      "loss": 0.5587,
+      "step": 3012
+    },
+    {
+      "epoch": 1.8191971023241775,
+      "grad_norm": 0.14453125,
+      "learning_rate": 8.115400581154005e-05,
+      "loss": 0.7154,
+      "step": 3013
+    },
+    {
+      "epoch": 1.819800784787202,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 8.111249481112495e-05,
+      "loss": 0.6538,
+      "step": 3014
+    },
+    {
+      "epoch": 1.8204044672502264,
+      "grad_norm": 0.1328125,
+      "learning_rate": 8.107098381070984e-05,
+      "loss": 0.6695,
+      "step": 3015
+    },
+    {
+      "epoch": 1.8210081497132509,
+      "grad_norm": 0.134765625,
+      "learning_rate": 8.102947281029473e-05,
+      "loss": 0.8783,
+      "step": 3016
+    },
+    {
+      "epoch": 1.8216118321762753,
+      "grad_norm": 0.13671875,
+      "learning_rate": 8.098796180987961e-05,
+      "loss": 0.6434,
+      "step": 3017
+    },
+    {
+      "epoch": 1.8222155146392998,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 8.094645080946451e-05,
+      "loss": 0.6594,
+      "step": 3018
+    },
+    {
+      "epoch": 1.8228191971023242,
+      "grad_norm": 0.142578125,
+      "learning_rate": 8.090493980904941e-05,
+      "loss": 0.6878,
+      "step": 3019
+    },
+    {
+      "epoch": 1.8234228795653487,
+      "grad_norm": 0.125,
+      "learning_rate": 8.086342880863429e-05,
+      "loss": 0.7494,
+      "step": 3020
+    },
+    {
+      "epoch": 1.824026562028373,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 8.082191780821919e-05,
+      "loss": 0.9943,
+      "step": 3021
+    },
+    {
+      "epoch": 1.8246302444913975,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 8.078040680780407e-05,
+      "loss": 0.636,
+      "step": 3022
+    },
+    {
+      "epoch": 1.825233926954422,
+      "grad_norm": 0.140625,
+      "learning_rate": 8.073889580738896e-05,
+      "loss": 0.7553,
+      "step": 3023
+    },
+    {
+      "epoch": 1.8258376094174464,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 8.069738480697385e-05,
+      "loss": 0.6998,
+      "step": 3024
+    },
+    {
+      "epoch": 1.8264412918804709,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 8.065587380655875e-05,
+      "loss": 0.6978,
+      "step": 3025
+    },
+    {
+      "epoch": 1.8270449743434953,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 8.061436280614363e-05,
+      "loss": 0.6634,
+      "step": 3026
+    },
+    {
+      "epoch": 1.8276486568065198,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 8.057285180572852e-05,
+      "loss": 0.6892,
+      "step": 3027
+    },
+    {
+      "epoch": 1.8282523392695442,
+      "grad_norm": 0.138671875,
+      "learning_rate": 8.053134080531342e-05,
+      "loss": 0.7321,
+      "step": 3028
+    },
+    {
+      "epoch": 1.8288560217325687,
+      "grad_norm": 0.123046875,
+      "learning_rate": 8.048982980489829e-05,
+      "loss": 0.6714,
+      "step": 3029
+    },
+    {
+      "epoch": 1.8294597041955931,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 8.044831880448319e-05,
+      "loss": 0.7105,
+      "step": 3030
+    },
+    {
+      "epoch": 1.8300633866586176,
+      "grad_norm": 0.12890625,
+      "learning_rate": 8.040680780406808e-05,
+      "loss": 0.7032,
+      "step": 3031
+    },
+    {
+      "epoch": 1.830667069121642,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 8.036529680365296e-05,
+      "loss": 0.6389,
+      "step": 3032
+    },
+    {
+      "epoch": 1.8312707515846665,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 8.032378580323786e-05,
+      "loss": 0.8036,
+      "step": 3033
+    },
+    {
+      "epoch": 1.831874434047691,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 8.028227480282275e-05,
+      "loss": 0.6143,
+      "step": 3034
+    },
+    {
+      "epoch": 1.8324781165107153,
+      "grad_norm": 0.134765625,
+      "learning_rate": 8.024076380240764e-05,
+      "loss": 0.6233,
+      "step": 3035
+    },
+    {
+      "epoch": 1.8330817989737398,
+      "grad_norm": 0.138671875,
+      "learning_rate": 8.019925280199252e-05,
+      "loss": 0.5732,
+      "step": 3036
+    },
+    {
+      "epoch": 1.8336854814367642,
+      "grad_norm": 0.125,
+      "learning_rate": 8.015774180157742e-05,
+      "loss": 0.7028,
+      "step": 3037
+    },
+    {
+      "epoch": 1.8342891638997887,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 8.011623080116231e-05,
+      "loss": 0.7947,
+      "step": 3038
+    },
+    {
+      "epoch": 1.8348928463628131,
+      "grad_norm": 0.12890625,
+      "learning_rate": 8.00747198007472e-05,
+      "loss": 0.6311,
+      "step": 3039
+    },
+    {
+      "epoch": 1.8354965288258376,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 8.00332088003321e-05,
+      "loss": 0.6732,
+      "step": 3040
+    },
+    {
+      "epoch": 1.836100211288862,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 7.999169779991698e-05,
+      "loss": 0.5363,
+      "step": 3041
+    },
+    {
+      "epoch": 1.8367038937518865,
+      "grad_norm": 0.126953125,
+      "learning_rate": 7.995018679950187e-05,
+      "loss": 0.5187,
+      "step": 3042
+    },
+    {
+      "epoch": 1.837307576214911,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 7.990867579908676e-05,
+      "loss": 0.6697,
+      "step": 3043
+    },
+    {
+      "epoch": 1.8379112586779354,
+      "grad_norm": 0.12890625,
+      "learning_rate": 7.986716479867166e-05,
+      "loss": 0.5733,
+      "step": 3044
+    },
+    {
+      "epoch": 1.8385149411409598,
+      "grad_norm": 0.142578125,
+      "learning_rate": 7.982565379825654e-05,
+      "loss": 0.5313,
+      "step": 3045
+    },
+    {
+      "epoch": 1.8391186236039843,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 7.978414279784143e-05,
+      "loss": 0.5485,
+      "step": 3046
+    },
+    {
+      "epoch": 1.8397223060670087,
+      "grad_norm": 0.15625,
+      "learning_rate": 7.974263179742633e-05,
+      "loss": 0.5707,
+      "step": 3047
+    },
+    {
+      "epoch": 1.8403259885300332,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 7.97011207970112e-05,
+      "loss": 0.5065,
+      "step": 3048
+    },
+    {
+      "epoch": 1.8409296709930576,
+      "grad_norm": 0.158203125,
+      "learning_rate": 7.96596097965961e-05,
+      "loss": 0.5392,
+      "step": 3049
+    },
+    {
+      "epoch": 1.8415333534560823,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 7.961809879618099e-05,
+      "loss": 0.4614,
+      "step": 3050
+    },
+    {
+      "epoch": 1.8421370359191065,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.957658779576588e-05,
+      "loss": 0.4686,
+      "step": 3051
+    },
+    {
+      "epoch": 1.8427407183821312,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.953507679535078e-05,
+      "loss": 0.429,
+      "step": 3052
+    },
+    {
+      "epoch": 1.8433444008451554,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.949356579493566e-05,
+      "loss": 0.3595,
+      "step": 3053
+    },
+    {
+      "epoch": 1.84394808330818,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.945205479452055e-05,
+      "loss": 0.3526,
+      "step": 3054
+    },
+    {
+      "epoch": 1.8445517657712043,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.941054379410544e-05,
+      "loss": 0.3008,
+      "step": 3055
+    },
+    {
+      "epoch": 1.845155448234229,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.936903279369034e-05,
+      "loss": 0.222,
+      "step": 3056
+    },
+    {
+      "epoch": 1.8457591306972532,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 7.932752179327522e-05,
+      "loss": 0.99,
+      "step": 3057
+    },
+    {
+      "epoch": 1.8463628131602778,
+      "grad_norm": 0.1328125,
+      "learning_rate": 7.928601079286011e-05,
+      "loss": 0.6689,
+      "step": 3058
+    },
+    {
+      "epoch": 1.846966495623302,
+      "grad_norm": 0.12890625,
+      "learning_rate": 7.924449979244501e-05,
+      "loss": 0.5613,
+      "step": 3059
+    },
+    {
+      "epoch": 1.8475701780863267,
+      "grad_norm": 0.125,
+      "learning_rate": 7.920298879202988e-05,
+      "loss": 0.6002,
+      "step": 3060
+    },
+    {
+      "epoch": 1.848173860549351,
+      "grad_norm": 0.15234375,
+      "learning_rate": 7.916147779161478e-05,
+      "loss": 0.6666,
+      "step": 3061
+    },
+    {
+      "epoch": 1.8487775430123756,
+      "grad_norm": 0.13671875,
+      "learning_rate": 7.911996679119967e-05,
+      "loss": 0.6661,
+      "step": 3062
+    },
+    {
+      "epoch": 1.8493812254753998,
+      "grad_norm": 0.1328125,
+      "learning_rate": 7.907845579078457e-05,
+      "loss": 0.588,
+      "step": 3063
+    },
+    {
+      "epoch": 1.8499849079384245,
+      "grad_norm": 0.14453125,
+      "learning_rate": 7.903694479036945e-05,
+      "loss": 0.9441,
+      "step": 3064
+    },
+    {
+      "epoch": 1.8505885904014487,
+      "grad_norm": 0.134765625,
+      "learning_rate": 7.899543378995434e-05,
+      "loss": 0.9012,
+      "step": 3065
+    },
+    {
+      "epoch": 1.8511922728644734,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 7.895392278953924e-05,
+      "loss": 0.8596,
+      "step": 3066
+    },
+    {
+      "epoch": 1.8517959553274976,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 7.891241178912411e-05,
+      "loss": 0.6868,
+      "step": 3067
+    },
+    {
+      "epoch": 1.8523996377905223,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 7.887090078870901e-05,
+      "loss": 0.6323,
+      "step": 3068
+    },
+    {
+      "epoch": 1.8530033202535465,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 7.88293897882939e-05,
+      "loss": 0.5647,
+      "step": 3069
+    },
+    {
+      "epoch": 1.8536070027165712,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 7.878787878787879e-05,
+      "loss": 0.6628,
+      "step": 3070
+    },
+    {
+      "epoch": 1.8542106851795954,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 7.874636778746369e-05,
+      "loss": 0.6885,
+      "step": 3071
+    },
+    {
+      "epoch": 1.85481436764262,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 7.870485678704857e-05,
+      "loss": 0.5712,
+      "step": 3072
+    },
+    {
+      "epoch": 1.8554180501056443,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 7.866334578663346e-05,
+      "loss": 0.6282,
+      "step": 3073
+    },
+    {
+      "epoch": 1.856021732568669,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 7.862183478621835e-05,
+      "loss": 0.6369,
+      "step": 3074
+    },
+    {
+      "epoch": 1.8566254150316932,
+      "grad_norm": 0.134765625,
+      "learning_rate": 7.858032378580325e-05,
+      "loss": 0.6647,
+      "step": 3075
+    },
+    {
+      "epoch": 1.8572290974947179,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 7.853881278538813e-05,
+      "loss": 0.9152,
+      "step": 3076
+    },
+    {
+      "epoch": 1.857832779957742,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 7.849730178497302e-05,
+      "loss": 0.6018,
+      "step": 3077
+    },
+    {
+      "epoch": 1.8584364624207668,
+      "grad_norm": 0.14453125,
+      "learning_rate": 7.845579078455792e-05,
+      "loss": 0.9967,
+      "step": 3078
+    },
+    {
+      "epoch": 1.859040144883791,
+      "grad_norm": 0.11328125,
+      "learning_rate": 7.841427978414279e-05,
+      "loss": 0.6408,
+      "step": 3079
+    },
+    {
+      "epoch": 1.8596438273468157,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 7.837276878372769e-05,
+      "loss": 0.7113,
+      "step": 3080
+    },
+    {
+      "epoch": 1.8602475098098399,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 7.833125778331258e-05,
+      "loss": 0.6914,
+      "step": 3081
+    },
+    {
+      "epoch": 1.8608511922728646,
+      "grad_norm": 0.125,
+      "learning_rate": 7.828974678289747e-05,
+      "loss": 0.8815,
+      "step": 3082
+    },
+    {
+      "epoch": 1.8614548747358888,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 7.824823578248237e-05,
+      "loss": 0.9352,
+      "step": 3083
+    },
+    {
+      "epoch": 1.8620585571989134,
+      "grad_norm": 0.140625,
+      "learning_rate": 7.820672478206725e-05,
+      "loss": 0.6636,
+      "step": 3084
+    },
+    {
+      "epoch": 1.8626622396619377,
+      "grad_norm": 0.13671875,
+      "learning_rate": 7.816521378165215e-05,
+      "loss": 0.6623,
+      "step": 3085
+    },
+    {
+      "epoch": 1.8632659221249623,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 7.812370278123702e-05,
+      "loss": 0.8932,
+      "step": 3086
+    },
+    {
+      "epoch": 1.8638696045879866,
+      "grad_norm": 0.130859375,
+      "learning_rate": 7.808219178082192e-05,
+      "loss": 0.5604,
+      "step": 3087
+    },
+    {
+      "epoch": 1.8644732870510112,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 7.804068078040681e-05,
+      "loss": 0.6196,
+      "step": 3088
+    },
+    {
+      "epoch": 1.8650769695140355,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 7.79991697799917e-05,
+      "loss": 0.6475,
+      "step": 3089
+    },
+    {
+      "epoch": 1.8656806519770601,
+      "grad_norm": 0.134765625,
+      "learning_rate": 7.79576587795766e-05,
+      "loss": 0.6318,
+      "step": 3090
+    },
+    {
+      "epoch": 1.8662843344400846,
+      "grad_norm": 0.134765625,
+      "learning_rate": 7.791614777916148e-05,
+      "loss": 0.5979,
+      "step": 3091
+    },
+    {
+      "epoch": 1.866888016903109,
+      "grad_norm": 0.130859375,
+      "learning_rate": 7.787463677874637e-05,
+      "loss": 0.5466,
+      "step": 3092
+    },
+    {
+      "epoch": 1.8674916993661335,
+      "grad_norm": 0.13671875,
+      "learning_rate": 7.783312577833126e-05,
+      "loss": 0.6918,
+      "step": 3093
+    },
+    {
+      "epoch": 1.868095381829158,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 7.779161477791616e-05,
+      "loss": 0.8258,
+      "step": 3094
+    },
+    {
+      "epoch": 1.8686990642921824,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.775010377750104e-05,
+      "loss": 0.5167,
+      "step": 3095
+    },
+    {
+      "epoch": 1.8693027467552068,
+      "grad_norm": 0.13671875,
+      "learning_rate": 7.770859277708593e-05,
+      "loss": 0.5426,
+      "step": 3096
+    },
+    {
+      "epoch": 1.8699064292182312,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 7.766708177667083e-05,
+      "loss": 0.5094,
+      "step": 3097
+    },
+    {
+      "epoch": 1.8705101116812557,
+      "grad_norm": 0.14453125,
+      "learning_rate": 7.76255707762557e-05,
+      "loss": 0.5342,
+      "step": 3098
+    },
+    {
+      "epoch": 1.8711137941442801,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 7.75840597758406e-05,
+      "loss": 0.5153,
+      "step": 3099
+    },
+    {
+      "epoch": 1.8717174766073046,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 7.754254877542549e-05,
+      "loss": 0.5314,
+      "step": 3100
+    },
+    {
+      "epoch": 1.872321159070329,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 7.750103777501038e-05,
+      "loss": 0.4814,
+      "step": 3101
+    },
+    {
+      "epoch": 1.8729248415333535,
+      "grad_norm": 0.162109375,
+      "learning_rate": 7.745952677459528e-05,
+      "loss": 0.4209,
+      "step": 3102
+    },
+    {
+      "epoch": 1.873528523996378,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.741801577418016e-05,
+      "loss": 0.4853,
+      "step": 3103
+    },
+    {
+      "epoch": 1.8741322064594024,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.737650477376505e-05,
+      "loss": 0.427,
+      "step": 3104
+    },
+    {
+      "epoch": 1.8747358889224268,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.733499377334994e-05,
+      "loss": 0.2704,
+      "step": 3105
+    },
+    {
+      "epoch": 1.8753395713854513,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.729348277293484e-05,
+      "loss": 0.2182,
+      "step": 3106
+    },
+    {
+      "epoch": 1.8759432538484757,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 7.725197177251972e-05,
+      "loss": 0.6389,
+      "step": 3107
+    },
+    {
+      "epoch": 1.8765469363115002,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 7.721046077210461e-05,
+      "loss": 0.7304,
+      "step": 3108
+    },
+    {
+      "epoch": 1.8771506187745246,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 7.716894977168951e-05,
+      "loss": 0.622,
+      "step": 3109
+    },
+    {
+      "epoch": 1.877754301237549,
+      "grad_norm": 0.140625,
+      "learning_rate": 7.71274387712744e-05,
+      "loss": 0.6608,
+      "step": 3110
+    },
+    {
+      "epoch": 1.8783579837005735,
+      "grad_norm": 0.13671875,
+      "learning_rate": 7.708592777085928e-05,
+      "loss": 0.6979,
+      "step": 3111
+    },
+    {
+      "epoch": 1.878961666163598,
+      "grad_norm": 0.14453125,
+      "learning_rate": 7.704441677044417e-05,
+      "loss": 0.64,
+      "step": 3112
+    },
+    {
+      "epoch": 1.8795653486266224,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 7.700290577002907e-05,
+      "loss": 0.846,
+      "step": 3113
+    },
+    {
+      "epoch": 1.8801690310896468,
+      "grad_norm": 0.134765625,
+      "learning_rate": 7.696139476961395e-05,
+      "loss": 0.6574,
+      "step": 3114
+    },
+    {
+      "epoch": 1.8807727135526713,
+      "grad_norm": 0.12890625,
+      "learning_rate": 7.691988376919884e-05,
+      "loss": 0.6364,
+      "step": 3115
+    },
+    {
+      "epoch": 1.8813763960156957,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 7.687837276878374e-05,
+      "loss": 0.916,
+      "step": 3116
+    },
+    {
+      "epoch": 1.8819800784787202,
+      "grad_norm": 0.134765625,
+      "learning_rate": 7.683686176836861e-05,
+      "loss": 0.7091,
+      "step": 3117
+    },
+    {
+      "epoch": 1.8825837609417446,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 7.679535076795351e-05,
+      "loss": 0.72,
+      "step": 3118
+    },
+    {
+      "epoch": 1.883187443404769,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 7.67538397675384e-05,
+      "loss": 0.6347,
+      "step": 3119
+    },
+    {
+      "epoch": 1.8837911258677935,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 7.671232876712329e-05,
+      "loss": 0.6204,
+      "step": 3120
+    },
+    {
+      "epoch": 1.884394808330818,
+      "grad_norm": 0.126953125,
+      "learning_rate": 7.667081776670819e-05,
+      "loss": 0.5928,
+      "step": 3121
+    },
+    {
+      "epoch": 1.8849984907938424,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 7.662930676629307e-05,
+      "loss": 0.6527,
+      "step": 3122
+    },
+    {
+      "epoch": 1.8856021732568669,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 7.658779576587796e-05,
+      "loss": 0.6877,
+      "step": 3123
+    },
+    {
+      "epoch": 1.8862058557198913,
+      "grad_norm": 0.146484375,
+      "learning_rate": 7.654628476546285e-05,
+      "loss": 0.792,
+      "step": 3124
+    },
+    {
+      "epoch": 1.8868095381829157,
+      "grad_norm": 0.1328125,
+      "learning_rate": 7.650477376504775e-05,
+      "loss": 0.6725,
+      "step": 3125
+    },
+    {
+      "epoch": 1.8874132206459402,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 7.646326276463262e-05,
+      "loss": 0.7279,
+      "step": 3126
+    },
+    {
+      "epoch": 1.8880169031089646,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 7.642175176421752e-05,
+      "loss": 0.6495,
+      "step": 3127
+    },
+    {
+      "epoch": 1.888620585571989,
+      "grad_norm": 0.130859375,
+      "learning_rate": 7.638024076380242e-05,
+      "loss": 0.516,
+      "step": 3128
+    },
+    {
+      "epoch": 1.8892242680350135,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 7.63387297633873e-05,
+      "loss": 0.9977,
+      "step": 3129
+    },
+    {
+      "epoch": 1.889827950498038,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 7.629721876297219e-05,
+      "loss": 0.5642,
+      "step": 3130
+    },
+    {
+      "epoch": 1.8904316329610626,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 7.625570776255708e-05,
+      "loss": 0.9672,
+      "step": 3131
+    },
+    {
+      "epoch": 1.8910353154240869,
+      "grad_norm": 0.126953125,
+      "learning_rate": 7.621419676214198e-05,
+      "loss": 0.6098,
+      "step": 3132
+    },
+    {
+      "epoch": 1.8916389978871115,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 7.617268576172685e-05,
+      "loss": 0.5991,
+      "step": 3133
+    },
+    {
+      "epoch": 1.8922426803501358,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 7.613117476131175e-05,
+      "loss": 0.585,
+      "step": 3134
+    },
+    {
+      "epoch": 1.8928463628131604,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 7.608966376089665e-05,
+      "loss": 0.6032,
+      "step": 3135
+    },
+    {
+      "epoch": 1.8934500452761847,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 7.604815276048153e-05,
+      "loss": 0.5951,
+      "step": 3136
+    },
+    {
+      "epoch": 1.8940537277392093,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 7.600664176006643e-05,
+      "loss": 0.5891,
+      "step": 3137
+    },
+    {
+      "epoch": 1.8946574102022335,
+      "grad_norm": 0.134765625,
+      "learning_rate": 7.596513075965131e-05,
+      "loss": 0.6856,
+      "step": 3138
+    },
+    {
+      "epoch": 1.8952610926652582,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 7.59236197592362e-05,
+      "loss": 0.5943,
+      "step": 3139
+    },
+    {
+      "epoch": 1.8958647751282824,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 7.588210875882108e-05,
+      "loss": 0.5702,
+      "step": 3140
+    },
+    {
+      "epoch": 1.896468457591307,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 7.584059775840598e-05,
+      "loss": 0.6605,
+      "step": 3141
+    },
+    {
+      "epoch": 1.8970721400543313,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 7.579908675799087e-05,
+      "loss": 0.5738,
+      "step": 3142
+    },
+    {
+      "epoch": 1.897675822517356,
+      "grad_norm": 0.146484375,
+      "learning_rate": 7.575757575757576e-05,
+      "loss": 0.6339,
+      "step": 3143
+    },
+    {
+      "epoch": 1.8982795049803802,
+      "grad_norm": 0.13671875,
+      "learning_rate": 7.571606475716066e-05,
+      "loss": 0.6243,
+      "step": 3144
+    },
+    {
+      "epoch": 1.898883187443405,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.567455375674553e-05,
+      "loss": 0.5782,
+      "step": 3145
+    },
+    {
+      "epoch": 1.8994868699064291,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 7.563304275633043e-05,
+      "loss": 0.4687,
+      "step": 3146
+    },
+    {
+      "epoch": 1.9000905523694538,
+      "grad_norm": 0.14453125,
+      "learning_rate": 7.559153175591532e-05,
+      "loss": 0.4899,
+      "step": 3147
+    },
+    {
+      "epoch": 1.900694234832478,
+      "grad_norm": 0.150390625,
+      "learning_rate": 7.55500207555002e-05,
+      "loss": 0.4926,
+      "step": 3148
+    },
+    {
+      "epoch": 1.9012979172955027,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 7.55085097550851e-05,
+      "loss": 0.511,
+      "step": 3149
+    },
+    {
+      "epoch": 1.901901599758527,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.546699875466999e-05,
+      "loss": 0.556,
+      "step": 3150
+    },
+    {
+      "epoch": 1.9025052822215516,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.542548775425489e-05,
+      "loss": 0.4537,
+      "step": 3151
+    },
+    {
+      "epoch": 1.9031089646845758,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.538397675383976e-05,
+      "loss": 0.4078,
+      "step": 3152
+    },
+    {
+      "epoch": 1.9037126471476005,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.534246575342466e-05,
+      "loss": 0.4259,
+      "step": 3153
+    },
+    {
+      "epoch": 1.9043163296106247,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.530095475300955e-05,
+      "loss": 0.3296,
+      "step": 3154
+    },
+    {
+      "epoch": 1.9049200120736494,
+      "grad_norm": 0.193359375,
+      "learning_rate": 7.525944375259444e-05,
+      "loss": 0.3254,
+      "step": 3155
+    },
+    {
+      "epoch": 1.9055236945366736,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.521793275217934e-05,
+      "loss": 0.21,
+      "step": 3156
+    },
+    {
+      "epoch": 1.9061273769996983,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 7.517642175176422e-05,
+      "loss": 0.604,
+      "step": 3157
+    },
+    {
+      "epoch": 1.9067310594627225,
+      "grad_norm": 0.134765625,
+      "learning_rate": 7.513491075134911e-05,
+      "loss": 0.6113,
+      "step": 3158
+    },
+    {
+      "epoch": 1.9073347419257471,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 7.5093399750934e-05,
+      "loss": 0.6336,
+      "step": 3159
+    },
+    {
+      "epoch": 1.9079384243887714,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 7.50518887505189e-05,
+      "loss": 0.6376,
+      "step": 3160
+    },
+    {
+      "epoch": 1.908542106851796,
+      "grad_norm": 0.13671875,
+      "learning_rate": 7.501037775010378e-05,
+      "loss": 0.6476,
+      "step": 3161
+    },
+    {
+      "epoch": 1.9091457893148203,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 7.496886674968867e-05,
+      "loss": 0.6033,
+      "step": 3162
+    },
+    {
+      "epoch": 1.909749471777845,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 7.492735574927357e-05,
+      "loss": 0.6158,
+      "step": 3163
+    },
+    {
+      "epoch": 1.9103531542408692,
+      "grad_norm": 0.134765625,
+      "learning_rate": 7.488584474885844e-05,
+      "loss": 0.6094,
+      "step": 3164
+    },
+    {
+      "epoch": 1.9109568367038938,
+      "grad_norm": 0.140625,
+      "learning_rate": 7.484433374844334e-05,
+      "loss": 0.6744,
+      "step": 3165
+    },
+    {
+      "epoch": 1.911560519166918,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.480282274802823e-05,
+      "loss": 0.6104,
+      "step": 3166
+    },
+    {
+      "epoch": 1.9121642016299427,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 7.476131174761311e-05,
+      "loss": 0.6194,
+      "step": 3167
+    },
+    {
+      "epoch": 1.912767884092967,
+      "grad_norm": 0.126953125,
+      "learning_rate": 7.471980074719801e-05,
+      "loss": 0.8527,
+      "step": 3168
+    },
+    {
+      "epoch": 1.9133715665559916,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 7.46782897467829e-05,
+      "loss": 0.6211,
+      "step": 3169
+    },
+    {
+      "epoch": 1.9139752490190158,
+      "grad_norm": 0.146484375,
+      "learning_rate": 7.463677874636779e-05,
+      "loss": 0.6325,
+      "step": 3170
+    },
+    {
+      "epoch": 1.9145789314820405,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 7.459526774595267e-05,
+      "loss": 0.7137,
+      "step": 3171
+    },
+    {
+      "epoch": 1.915182613945065,
+      "grad_norm": 0.15234375,
+      "learning_rate": 7.455375674553757e-05,
+      "loss": 0.6791,
+      "step": 3172
+    },
+    {
+      "epoch": 1.9157862964080894,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 7.451224574512246e-05,
+      "loss": 0.6181,
+      "step": 3173
+    },
+    {
+      "epoch": 1.9163899788711138,
+      "grad_norm": 0.15234375,
+      "learning_rate": 7.447073474470735e-05,
+      "loss": 0.6748,
+      "step": 3174
+    },
+    {
+      "epoch": 1.9169936613341383,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 7.442922374429225e-05,
+      "loss": 0.7586,
+      "step": 3175
+    },
+    {
+      "epoch": 1.9175973437971627,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 7.438771274387713e-05,
+      "loss": 0.6438,
+      "step": 3176
+    },
+    {
+      "epoch": 1.9182010262601872,
+      "grad_norm": 0.130859375,
+      "learning_rate": 7.434620174346202e-05,
+      "loss": 0.8398,
+      "step": 3177
+    },
+    {
+      "epoch": 1.9188047087232116,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 7.43046907430469e-05,
+      "loss": 0.6768,
+      "step": 3178
+    },
+    {
+      "epoch": 1.919408391186236,
+      "grad_norm": 0.1484375,
+      "learning_rate": 7.42631797426318e-05,
+      "loss": 0.707,
+      "step": 3179
+    },
+    {
+      "epoch": 1.9200120736492605,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 7.422166874221669e-05,
+      "loss": 0.9903,
+      "step": 3180
+    },
+    {
+      "epoch": 1.920615756112285,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 7.418015774180158e-05,
+      "loss": 0.8726,
+      "step": 3181
+    },
+    {
+      "epoch": 1.9212194385753094,
+      "grad_norm": 0.125,
+      "learning_rate": 7.413864674138648e-05,
+      "loss": 0.6792,
+      "step": 3182
+    },
+    {
+      "epoch": 1.9218231210383339,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 7.409713574097135e-05,
+      "loss": 0.6856,
+      "step": 3183
+    },
+    {
+      "epoch": 1.9224268035013583,
+      "grad_norm": 0.134765625,
+      "learning_rate": 7.405562474055625e-05,
+      "loss": 0.6322,
+      "step": 3184
+    },
+    {
+      "epoch": 1.9230304859643828,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 7.401411374014114e-05,
+      "loss": 0.7408,
+      "step": 3185
+    },
+    {
+      "epoch": 1.9236341684274072,
+      "grad_norm": 0.1328125,
+      "learning_rate": 7.397260273972603e-05,
+      "loss": 0.6326,
+      "step": 3186
+    },
+    {
+      "epoch": 1.9242378508904316,
+      "grad_norm": 0.1328125,
+      "learning_rate": 7.393109173931093e-05,
+      "loss": 1.1231,
+      "step": 3187
+    },
+    {
+      "epoch": 1.924841533353456,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 7.388958073889581e-05,
+      "loss": 0.5966,
+      "step": 3188
+    },
+    {
+      "epoch": 1.9254452158164805,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 7.38480697384807e-05,
+      "loss": 0.6252,
+      "step": 3189
+    },
+    {
+      "epoch": 1.926048898279505,
+      "grad_norm": 0.14453125,
+      "learning_rate": 7.380655873806558e-05,
+      "loss": 0.6251,
+      "step": 3190
+    },
+    {
+      "epoch": 1.9266525807425294,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 7.376504773765048e-05,
+      "loss": 0.6284,
+      "step": 3191
+    },
+    {
+      "epoch": 1.9272562632055539,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 7.372353673723537e-05,
+      "loss": 0.5263,
+      "step": 3192
+    },
+    {
+      "epoch": 1.9278599456685783,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 7.368202573682026e-05,
+      "loss": 0.5516,
+      "step": 3193
+    },
+    {
+      "epoch": 1.9284636281316028,
+      "grad_norm": 0.138671875,
+      "learning_rate": 7.364051473640516e-05,
+      "loss": 0.5675,
+      "step": 3194
+    },
+    {
+      "epoch": 1.9290673105946272,
+      "grad_norm": 0.146484375,
+      "learning_rate": 7.359900373599004e-05,
+      "loss": 0.5492,
+      "step": 3195
+    },
+    {
+      "epoch": 1.9296709930576517,
+      "grad_norm": 0.14453125,
+      "learning_rate": 7.355749273557493e-05,
+      "loss": 0.5312,
+      "step": 3196
+    },
+    {
+      "epoch": 1.930274675520676,
+      "grad_norm": 0.150390625,
+      "learning_rate": 7.351598173515982e-05,
+      "loss": 0.5808,
+      "step": 3197
+    },
+    {
+      "epoch": 1.9308783579837006,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 7.347447073474472e-05,
+      "loss": 0.5105,
+      "step": 3198
+    },
+    {
+      "epoch": 1.931482040446725,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 7.34329597343296e-05,
+      "loss": 0.514,
+      "step": 3199
+    },
+    {
+      "epoch": 1.9320857229097494,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.339144873391449e-05,
+      "loss": 0.5566,
+      "step": 3200
+    },
+    {
+      "epoch": 1.932689405372774,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.334993773349939e-05,
+      "loss": 0.4912,
+      "step": 3201
+    },
+    {
+      "epoch": 1.9332930878357983,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.330842673308426e-05,
+      "loss": 0.4308,
+      "step": 3202
+    },
+    {
+      "epoch": 1.9338967702988228,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.326691573266916e-05,
+      "loss": 0.3551,
+      "step": 3203
+    },
+    {
+      "epoch": 1.9345004527618472,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.322540473225405e-05,
+      "loss": 0.3277,
+      "step": 3204
+    },
+    {
+      "epoch": 1.9351041352248717,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.318389373183894e-05,
+      "loss": 0.2927,
+      "step": 3205
+    },
+    {
+      "epoch": 1.9357078176878961,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.314238273142384e-05,
+      "loss": 0.2479,
+      "step": 3206
+    },
+    {
+      "epoch": 1.9363115001509206,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.310087173100872e-05,
+      "loss": 0.5836,
+      "step": 3207
+    },
+    {
+      "epoch": 1.936915182613945,
+      "grad_norm": 0.140625,
+      "learning_rate": 7.305936073059361e-05,
+      "loss": 0.5894,
+      "step": 3208
+    },
+    {
+      "epoch": 1.9375188650769695,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 7.30178497301785e-05,
+      "loss": 0.6468,
+      "step": 3209
+    },
+    {
+      "epoch": 1.938122547539994,
+      "grad_norm": 0.14453125,
+      "learning_rate": 7.29763387297634e-05,
+      "loss": 0.6436,
+      "step": 3210
+    },
+    {
+      "epoch": 1.9387262300030184,
+      "grad_norm": 0.140625,
+      "learning_rate": 7.293482772934828e-05,
+      "loss": 0.6471,
+      "step": 3211
+    },
+    {
+      "epoch": 1.939329912466043,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 7.289331672893317e-05,
+      "loss": 0.6474,
+      "step": 3212
+    },
+    {
+      "epoch": 1.9399335949290673,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 7.285180572851807e-05,
+      "loss": 0.8865,
+      "step": 3213
+    },
+    {
+      "epoch": 1.940537277392092,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 7.281029472810294e-05,
+      "loss": 0.9,
+      "step": 3214
+    },
+    {
+      "epoch": 1.9411409598551161,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 7.276878372768784e-05,
+      "loss": 0.6733,
+      "step": 3215
+    },
+    {
+      "epoch": 1.9417446423181408,
+      "grad_norm": 0.13671875,
+      "learning_rate": 7.272727272727273e-05,
+      "loss": 0.6217,
+      "step": 3216
+    },
+    {
+      "epoch": 1.942348324781165,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 7.268576172685761e-05,
+      "loss": 0.5774,
+      "step": 3217
+    },
+    {
+      "epoch": 1.9429520072441897,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 7.264425072644251e-05,
+      "loss": 0.7204,
+      "step": 3218
+    },
+    {
+      "epoch": 1.943555689707214,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 7.26027397260274e-05,
+      "loss": 0.709,
+      "step": 3219
+    },
+    {
+      "epoch": 1.9441593721702386,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 7.25612287256123e-05,
+      "loss": 0.685,
+      "step": 3220
+    },
+    {
+      "epoch": 1.9447630546332628,
+      "grad_norm": 0.14453125,
+      "learning_rate": 7.251971772519717e-05,
+      "loss": 0.6819,
+      "step": 3221
+    },
+    {
+      "epoch": 1.9453667370962875,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 7.247820672478207e-05,
+      "loss": 0.6286,
+      "step": 3222
+    },
+    {
+      "epoch": 1.9459704195593117,
+      "grad_norm": 0.126953125,
+      "learning_rate": 7.243669572436696e-05,
+      "loss": 0.6284,
+      "step": 3223
+    },
+    {
+      "epoch": 1.9465741020223364,
+      "grad_norm": 0.138671875,
+      "learning_rate": 7.239518472395185e-05,
+      "loss": 0.7257,
+      "step": 3224
+    },
+    {
+      "epoch": 1.9471777844853606,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 7.235367372353675e-05,
+      "loss": 0.6584,
+      "step": 3225
+    },
+    {
+      "epoch": 1.9477814669483853,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 7.231216272312163e-05,
+      "loss": 0.6025,
+      "step": 3226
+    },
+    {
+      "epoch": 1.9483851494114095,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.227065172270652e-05,
+      "loss": 0.6027,
+      "step": 3227
+    },
+    {
+      "epoch": 1.9489888318744342,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.22291407222914e-05,
+      "loss": 0.7539,
+      "step": 3228
+    },
+    {
+      "epoch": 1.9495925143374584,
+      "grad_norm": 0.150390625,
+      "learning_rate": 7.218762972187631e-05,
+      "loss": 0.6194,
+      "step": 3229
+    },
+    {
+      "epoch": 1.950196196800483,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 7.21461187214612e-05,
+      "loss": 0.7173,
+      "step": 3230
+    },
+    {
+      "epoch": 1.9507998792635073,
+      "grad_norm": 0.154296875,
+      "learning_rate": 7.210460772104608e-05,
+      "loss": 0.7522,
+      "step": 3231
+    },
+    {
+      "epoch": 1.951403561726532,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.206309672063098e-05,
+      "loss": 0.7973,
+      "step": 3232
+    },
+    {
+      "epoch": 1.9520072441895562,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 7.202158572021585e-05,
+      "loss": 0.7033,
+      "step": 3233
+    },
+    {
+      "epoch": 1.9526109266525808,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 7.198007471980075e-05,
+      "loss": 0.6033,
+      "step": 3234
+    },
+    {
+      "epoch": 1.953214609115605,
+      "grad_norm": 0.140625,
+      "learning_rate": 7.193856371938564e-05,
+      "loss": 0.5346,
+      "step": 3235
+    },
+    {
+      "epoch": 1.9538182915786297,
+      "grad_norm": 0.146484375,
+      "learning_rate": 7.189705271897053e-05,
+      "loss": 0.666,
+      "step": 3236
+    },
+    {
+      "epoch": 1.954421974041654,
+      "grad_norm": 0.12890625,
+      "learning_rate": 7.185554171855543e-05,
+      "loss": 0.6379,
+      "step": 3237
+    },
+    {
+      "epoch": 1.9550256565046786,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 7.181403071814031e-05,
+      "loss": 0.5682,
+      "step": 3238
+    },
+    {
+      "epoch": 1.9556293389677029,
+      "grad_norm": 0.134765625,
+      "learning_rate": 7.17725197177252e-05,
+      "loss": 0.697,
+      "step": 3239
+    },
+    {
+      "epoch": 1.9562330214307275,
+      "grad_norm": 0.12890625,
+      "learning_rate": 7.173100871731009e-05,
+      "loss": 0.5934,
+      "step": 3240
+    },
+    {
+      "epoch": 1.9568367038937517,
+      "grad_norm": 0.134765625,
+      "learning_rate": 7.168949771689499e-05,
+      "loss": 0.5425,
+      "step": 3241
+    },
+    {
+      "epoch": 1.9574403863567764,
+      "grad_norm": 0.134765625,
+      "learning_rate": 7.164798671647987e-05,
+      "loss": 0.5661,
+      "step": 3242
+    },
+    {
+      "epoch": 1.9580440688198006,
+      "grad_norm": 0.138671875,
+      "learning_rate": 7.160647571606476e-05,
+      "loss": 0.598,
+      "step": 3243
+    },
+    {
+      "epoch": 1.9586477512828253,
+      "grad_norm": 0.140625,
+      "learning_rate": 7.156496471564966e-05,
+      "loss": 0.5434,
+      "step": 3244
+    },
+    {
+      "epoch": 1.9592514337458495,
+      "grad_norm": 0.14453125,
+      "learning_rate": 7.152345371523454e-05,
+      "loss": 0.5406,
+      "step": 3245
+    },
+    {
+      "epoch": 1.9598551162088742,
+      "grad_norm": 0.15625,
+      "learning_rate": 7.148194271481943e-05,
+      "loss": 0.5284,
+      "step": 3246
+    },
+    {
+      "epoch": 1.9604587986718984,
+      "grad_norm": 0.154296875,
+      "learning_rate": 7.144043171440432e-05,
+      "loss": 0.5847,
+      "step": 3247
+    },
+    {
+      "epoch": 1.961062481134923,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 7.139892071398922e-05,
+      "loss": 0.5552,
+      "step": 3248
+    },
+    {
+      "epoch": 1.9616661635979473,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.135740971357409e-05,
+      "loss": 0.4877,
+      "step": 3249
+    },
+    {
+      "epoch": 1.962269846060972,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.131589871315899e-05,
+      "loss": 0.5421,
+      "step": 3250
+    },
+    {
+      "epoch": 1.9628735285239962,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.127438771274389e-05,
+      "loss": 0.4781,
+      "step": 3251
+    },
+    {
+      "epoch": 1.9634772109870209,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.123287671232876e-05,
+      "loss": 0.4282,
+      "step": 3252
+    },
+    {
+      "epoch": 1.9640808934500453,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.119136571191366e-05,
+      "loss": 0.4592,
+      "step": 3253
+    },
+    {
+      "epoch": 1.9646845759130698,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.114985471149855e-05,
+      "loss": 0.3757,
+      "step": 3254
+    },
+    {
+      "epoch": 1.9652882583760942,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 7.110834371108344e-05,
+      "loss": 0.3974,
+      "step": 3255
+    },
+    {
+      "epoch": 1.9658919408391187,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.106683271066832e-05,
+      "loss": 0.249,
+      "step": 3256
+    },
+    {
+      "epoch": 1.9664956233021431,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 7.102532171025322e-05,
+      "loss": 0.643,
+      "step": 3257
+    },
+    {
+      "epoch": 1.9670993057651676,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 7.098381070983811e-05,
+      "loss": 0.5703,
+      "step": 3258
+    },
+    {
+      "epoch": 1.967702988228192,
+      "grad_norm": 0.138671875,
+      "learning_rate": 7.0942299709423e-05,
+      "loss": 0.6712,
+      "step": 3259
+    },
+    {
+      "epoch": 1.9683066706912165,
+      "grad_norm": 0.138671875,
+      "learning_rate": 7.09007887090079e-05,
+      "loss": 0.6473,
+      "step": 3260
+    },
+    {
+      "epoch": 1.968910353154241,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 7.085927770859277e-05,
+      "loss": 0.7151,
+      "step": 3261
+    },
+    {
+      "epoch": 1.9695140356172653,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 7.081776670817767e-05,
+      "loss": 0.6657,
+      "step": 3262
+    },
+    {
+      "epoch": 1.9701177180802898,
+      "grad_norm": 0.15625,
+      "learning_rate": 7.077625570776256e-05,
+      "loss": 0.7846,
+      "step": 3263
+    },
+    {
+      "epoch": 1.9707214005433142,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.073474470734746e-05,
+      "loss": 0.6625,
+      "step": 3264
+    },
+    {
+      "epoch": 1.9713250830063387,
+      "grad_norm": 0.14453125,
+      "learning_rate": 7.069323370693234e-05,
+      "loss": 0.5718,
+      "step": 3265
+    },
+    {
+      "epoch": 1.9719287654693631,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 7.065172270651723e-05,
+      "loss": 0.855,
+      "step": 3266
+    },
+    {
+      "epoch": 1.9725324479323876,
+      "grad_norm": 0.140625,
+      "learning_rate": 7.061021170610213e-05,
+      "loss": 0.8964,
+      "step": 3267
+    },
+    {
+      "epoch": 1.973136130395412,
+      "grad_norm": 0.146484375,
+      "learning_rate": 7.0568700705687e-05,
+      "loss": 0.9168,
+      "step": 3268
+    },
+    {
+      "epoch": 1.9737398128584365,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 7.05271897052719e-05,
+      "loss": 0.6553,
+      "step": 3269
+    },
+    {
+      "epoch": 1.974343495321461,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 7.048567870485679e-05,
+      "loss": 0.6188,
+      "step": 3270
+    },
+    {
+      "epoch": 1.9749471777844854,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 7.044416770444167e-05,
+      "loss": 0.5445,
+      "step": 3271
+    },
+    {
+      "epoch": 1.9755508602475098,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 7.040265670402657e-05,
+      "loss": 0.6776,
+      "step": 3272
+    },
+    {
+      "epoch": 1.9761545427105343,
+      "grad_norm": 0.140625,
+      "learning_rate": 7.036114570361146e-05,
+      "loss": 0.6244,
+      "step": 3273
+    },
+    {
+      "epoch": 1.9767582251735587,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 7.031963470319635e-05,
+      "loss": 0.6687,
+      "step": 3274
+    },
+    {
+      "epoch": 1.9773619076365831,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 7.027812370278123e-05,
+      "loss": 0.5678,
+      "step": 3275
+    },
+    {
+      "epoch": 1.9779655900996076,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 7.023661270236613e-05,
+      "loss": 0.7597,
+      "step": 3276
+    },
+    {
+      "epoch": 1.978569272562632,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 7.019510170195102e-05,
+      "loss": 0.6088,
+      "step": 3277
+    },
+    {
+      "epoch": 1.9791729550256565,
+      "grad_norm": 0.12109375,
+      "learning_rate": 7.015359070153591e-05,
+      "loss": 0.8278,
+      "step": 3278
+    },
+    {
+      "epoch": 1.979776637488681,
+      "grad_norm": 0.138671875,
+      "learning_rate": 7.011207970112081e-05,
+      "loss": 1.011,
+      "step": 3279
+    },
+    {
+      "epoch": 1.9803803199517054,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 7.007056870070568e-05,
+      "loss": 0.5667,
+      "step": 3280
+    },
+    {
+      "epoch": 1.9809840024147298,
+      "grad_norm": 0.126953125,
+      "learning_rate": 7.002905770029058e-05,
+      "loss": 0.7015,
+      "step": 3281
+    },
+    {
+      "epoch": 1.9815876848777543,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 6.998754669987547e-05,
+      "loss": 0.614,
+      "step": 3282
+    },
+    {
+      "epoch": 1.9821913673407787,
+      "grad_norm": 0.146484375,
+      "learning_rate": 6.994603569946035e-05,
+      "loss": 0.6907,
+      "step": 3283
+    },
+    {
+      "epoch": 1.9827950498038032,
+      "grad_norm": 0.12890625,
+      "learning_rate": 6.990452469904525e-05,
+      "loss": 0.5913,
+      "step": 3284
+    },
+    {
+      "epoch": 1.9833987322668276,
+      "grad_norm": 0.134765625,
+      "learning_rate": 6.986301369863014e-05,
+      "loss": 0.6824,
+      "step": 3285
+    },
+    {
+      "epoch": 1.984002414729852,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 6.982150269821504e-05,
+      "loss": 0.6013,
+      "step": 3286
+    },
+    {
+      "epoch": 1.9846060971928765,
+      "grad_norm": 0.14453125,
+      "learning_rate": 6.977999169779991e-05,
+      "loss": 0.8288,
+      "step": 3287
+    },
+    {
+      "epoch": 1.985209779655901,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 6.973848069738481e-05,
+      "loss": 0.6383,
+      "step": 3288
+    },
+    {
+      "epoch": 1.9858134621189254,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 6.96969696969697e-05,
+      "loss": 0.7101,
+      "step": 3289
+    },
+    {
+      "epoch": 1.9864171445819498,
+      "grad_norm": 0.123046875,
+      "learning_rate": 6.965545869655459e-05,
+      "loss": 0.7906,
+      "step": 3290
+    },
+    {
+      "epoch": 1.9870208270449743,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 6.961394769613949e-05,
+      "loss": 0.5729,
+      "step": 3291
+    },
+    {
+      "epoch": 1.9876245095079987,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 6.957243669572437e-05,
+      "loss": 0.6291,
+      "step": 3292
+    },
+    {
+      "epoch": 1.9882281919710234,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 6.953092569530926e-05,
+      "loss": 0.6279,
+      "step": 3293
+    },
+    {
+      "epoch": 1.9888318744340476,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 6.948941469489415e-05,
+      "loss": 0.5872,
+      "step": 3294
+    },
+    {
+      "epoch": 1.9894355568970723,
+      "grad_norm": 0.150390625,
+      "learning_rate": 6.944790369447905e-05,
+      "loss": 0.5892,
+      "step": 3295
+    },
+    {
+      "epoch": 1.9900392393600965,
+      "grad_norm": 0.140625,
+      "learning_rate": 6.940639269406393e-05,
+      "loss": 0.5677,
+      "step": 3296
+    },
+    {
+      "epoch": 1.9906429218231212,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 6.936488169364882e-05,
+      "loss": 0.5897,
+      "step": 3297
+    },
+    {
+      "epoch": 1.9912466042861454,
+      "grad_norm": 0.14453125,
+      "learning_rate": 6.932337069323372e-05,
+      "loss": 0.5832,
+      "step": 3298
+    },
+    {
+      "epoch": 1.99185028674917,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 6.928185969281859e-05,
+      "loss": 0.5267,
+      "step": 3299
+    },
+    {
+      "epoch": 1.9924539692121943,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 6.924034869240349e-05,
+      "loss": 0.491,
+      "step": 3300
+    },
+    {
+      "epoch": 1.993057651675219,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.919883769198838e-05,
+      "loss": 0.4929,
+      "step": 3301
+    },
+    {
+      "epoch": 1.9936613341382432,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.915732669157326e-05,
+      "loss": 0.4367,
+      "step": 3302
+    },
+    {
+      "epoch": 1.9942650166012679,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.911581569115816e-05,
+      "loss": 0.3382,
+      "step": 3303
+    },
+    {
+      "epoch": 1.994868699064292,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.907430469074305e-05,
+      "loss": 0.3819,
+      "step": 3304
+    },
+    {
+      "epoch": 1.9954723815273168,
+      "grad_norm": 0.1875,
+      "learning_rate": 6.903279369032794e-05,
+      "loss": 0.3094,
+      "step": 3305
+    },
+    {
+      "epoch": 1.996076063990341,
+      "grad_norm": 0.220703125,
+      "learning_rate": 6.899128268991282e-05,
+      "loss": 0.2772,
+      "step": 3306
+    },
+    {
+      "epoch": 1.9966797464533657,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 6.894977168949772e-05,
+      "loss": 0.8034,
+      "step": 3307
+    },
+    {
+      "epoch": 1.9972834289163899,
+      "grad_norm": 0.1484375,
+      "learning_rate": 6.890826068908261e-05,
+      "loss": 0.6767,
+      "step": 3308
+    },
+    {
+      "epoch": 1.9978871113794145,
+      "grad_norm": 0.142578125,
+      "learning_rate": 6.88667496886675e-05,
+      "loss": 0.7063,
+      "step": 3309
+    },
+    {
+      "epoch": 1.9984907938424388,
+      "grad_norm": 0.23828125,
+      "learning_rate": 6.88252386882524e-05,
+      "loss": 0.5946,
+      "step": 3310
+    },
+    {
+      "epoch": 1.9990944763054634,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 6.878372768783728e-05,
+      "loss": 0.7957,
+      "step": 3311
+    },
+    {
+      "epoch": 1.9996981587684877,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 6.874221668742217e-05,
+      "loss": 0.4724,
+      "step": 3312
+    },
+    {
+      "epoch": 2.0006036824630247,
+      "grad_norm": 0.279296875,
+      "learning_rate": 6.870070568700706e-05,
+      "loss": 0.9675,
+      "step": 3313
+    },
+    {
+      "epoch": 2.001207364926049,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 6.865919468659196e-05,
+      "loss": 0.6052,
+      "step": 3314
+    },
+    {
+      "epoch": 2.0018110473890736,
+      "grad_norm": 0.140625,
+      "learning_rate": 6.861768368617684e-05,
+      "loss": 0.675,
+      "step": 3315
+    },
+    {
+      "epoch": 2.002414729852098,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 6.857617268576173e-05,
+      "loss": 0.7636,
+      "step": 3316
+    },
+    {
+      "epoch": 2.0030184123151225,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 6.853466168534663e-05,
+      "loss": 0.5467,
+      "step": 3317
+    },
+    {
+      "epoch": 2.0036220947781467,
+      "grad_norm": 0.1328125,
+      "learning_rate": 6.84931506849315e-05,
+      "loss": 0.6109,
+      "step": 3318
+    },
+    {
+      "epoch": 2.0042257772411713,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 6.84516396845164e-05,
+      "loss": 0.7048,
+      "step": 3319
+    },
+    {
+      "epoch": 2.0048294597041956,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 6.841012868410129e-05,
+      "loss": 1.0679,
+      "step": 3320
+    },
+    {
+      "epoch": 2.0054331421672202,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 6.836861768368617e-05,
+      "loss": 0.5886,
+      "step": 3321
+    },
+    {
+      "epoch": 2.0060368246302445,
+      "grad_norm": 0.142578125,
+      "learning_rate": 6.832710668327108e-05,
+      "loss": 0.6399,
+      "step": 3322
+    },
+    {
+      "epoch": 2.006640507093269,
+      "grad_norm": 0.134765625,
+      "learning_rate": 6.828559568285596e-05,
+      "loss": 0.6309,
+      "step": 3323
+    },
+    {
+      "epoch": 2.0072441895562934,
+      "grad_norm": 0.130859375,
+      "learning_rate": 6.824408468244085e-05,
+      "loss": 0.5685,
+      "step": 3324
+    },
+    {
+      "epoch": 2.007847872019318,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 6.820257368202573e-05,
+      "loss": 0.5241,
+      "step": 3325
+    },
+    {
+      "epoch": 2.0084515544823422,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 6.816106268161063e-05,
+      "loss": 0.6275,
+      "step": 3326
+    },
+    {
+      "epoch": 2.009055236945367,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 6.811955168119552e-05,
+      "loss": 0.853,
+      "step": 3327
+    },
+    {
+      "epoch": 2.009658919408391,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 6.807804068078041e-05,
+      "loss": 0.5848,
+      "step": 3328
+    },
+    {
+      "epoch": 2.010262601871416,
+      "grad_norm": 0.134765625,
+      "learning_rate": 6.803652968036531e-05,
+      "loss": 0.8565,
+      "step": 3329
+    },
+    {
+      "epoch": 2.01086628433444,
+      "grad_norm": 0.1484375,
+      "learning_rate": 6.79950186799502e-05,
+      "loss": 0.7739,
+      "step": 3330
+    },
+    {
+      "epoch": 2.0114699667974647,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 6.795350767953508e-05,
+      "loss": 0.6635,
+      "step": 3331
+    },
+    {
+      "epoch": 2.012073649260489,
+      "grad_norm": 0.12890625,
+      "learning_rate": 6.791199667911997e-05,
+      "loss": 0.5807,
+      "step": 3332
+    },
+    {
+      "epoch": 2.0126773317235136,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 6.787048567870487e-05,
+      "loss": 0.6374,
+      "step": 3333
+    },
+    {
+      "epoch": 2.013281014186538,
+      "grad_norm": 0.125,
+      "learning_rate": 6.782897467828975e-05,
+      "loss": 0.5274,
+      "step": 3334
+    },
+    {
+      "epoch": 2.0138846966495625,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.778746367787464e-05,
+      "loss": 0.6502,
+      "step": 3335
+    },
+    {
+      "epoch": 2.0144883791125867,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 6.774595267745954e-05,
+      "loss": 0.4959,
+      "step": 3336
+    },
+    {
+      "epoch": 2.0150920615756114,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 6.770444167704441e-05,
+      "loss": 0.6071,
+      "step": 3337
+    },
+    {
+      "epoch": 2.0156957440386356,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 6.766293067662931e-05,
+      "loss": 0.5929,
+      "step": 3338
+    },
+    {
+      "epoch": 2.0162994265016603,
+      "grad_norm": 0.15234375,
+      "learning_rate": 6.76214196762142e-05,
+      "loss": 0.6494,
+      "step": 3339
+    },
+    {
+      "epoch": 2.0169031089646845,
+      "grad_norm": 0.150390625,
+      "learning_rate": 6.757990867579909e-05,
+      "loss": 1.0521,
+      "step": 3340
+    },
+    {
+      "epoch": 2.017506791427709,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 6.753839767538399e-05,
+      "loss": 0.6239,
+      "step": 3341
+    },
+    {
+      "epoch": 2.0181104738907334,
+      "grad_norm": 0.134765625,
+      "learning_rate": 6.749688667496887e-05,
+      "loss": 0.6032,
+      "step": 3342
+    },
+    {
+      "epoch": 2.018714156353758,
+      "grad_norm": 0.130859375,
+      "learning_rate": 6.745537567455376e-05,
+      "loss": 0.5677,
+      "step": 3343
+    },
+    {
+      "epoch": 2.0193178388167823,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 6.741386467413865e-05,
+      "loss": 0.5771,
+      "step": 3344
+    },
+    {
+      "epoch": 2.019921521279807,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 6.737235367372355e-05,
+      "loss": 0.5857,
+      "step": 3345
+    },
+    {
+      "epoch": 2.020525203742831,
+      "grad_norm": 0.138671875,
+      "learning_rate": 6.733084267330843e-05,
+      "loss": 0.548,
+      "step": 3346
+    },
+    {
+      "epoch": 2.021128886205856,
+      "grad_norm": 0.13671875,
+      "learning_rate": 6.728933167289332e-05,
+      "loss": 0.5641,
+      "step": 3347
+    },
+    {
+      "epoch": 2.02173256866888,
+      "grad_norm": 0.126953125,
+      "learning_rate": 6.724782067247822e-05,
+      "loss": 0.5014,
+      "step": 3348
+    },
+    {
+      "epoch": 2.0223362511319047,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 6.720630967206309e-05,
+      "loss": 0.5817,
+      "step": 3349
+    },
+    {
+      "epoch": 2.022939933594929,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 6.716479867164799e-05,
+      "loss": 0.5907,
+      "step": 3350
+    },
+    {
+      "epoch": 2.0235436160579536,
+      "grad_norm": 0.1484375,
+      "learning_rate": 6.712328767123288e-05,
+      "loss": 0.5136,
+      "step": 3351
+    },
+    {
+      "epoch": 2.024147298520978,
+      "grad_norm": 0.150390625,
+      "learning_rate": 6.708177667081778e-05,
+      "loss": 0.5174,
+      "step": 3352
+    },
+    {
+      "epoch": 2.0247509809840025,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 6.704026567040266e-05,
+      "loss": 0.4647,
+      "step": 3353
+    },
+    {
+      "epoch": 2.0253546634470267,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 6.699875466998755e-05,
+      "loss": 0.4701,
+      "step": 3354
+    },
+    {
+      "epoch": 2.0259583459100514,
+      "grad_norm": 0.1640625,
+      "learning_rate": 6.695724366957245e-05,
+      "loss": 0.4639,
+      "step": 3355
+    },
+    {
+      "epoch": 2.0265620283730756,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.691573266915732e-05,
+      "loss": 0.4349,
+      "step": 3356
+    },
+    {
+      "epoch": 2.0271657108361003,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.687422166874222e-05,
+      "loss": 0.4562,
+      "step": 3357
+    },
+    {
+      "epoch": 2.0277693932991245,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.683271066832711e-05,
+      "loss": 0.3564,
+      "step": 3358
+    },
+    {
+      "epoch": 2.028373075762149,
+      "grad_norm": 0.201171875,
+      "learning_rate": 6.6791199667912e-05,
+      "loss": 0.3151,
+      "step": 3359
+    },
+    {
+      "epoch": 2.0289767582251734,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 6.67496886674969e-05,
+      "loss": 0.2686,
+      "step": 3360
+    },
+    {
+      "epoch": 2.029580440688198,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 6.670817766708178e-05,
+      "loss": 0.2815,
+      "step": 3361
+    },
+    {
+      "epoch": 2.0301841231512223,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.2146,
+      "step": 3362
+    },
+    {
+      "epoch": 2.030787805614247,
+      "grad_norm": 0.154296875,
+      "learning_rate": 6.662515566625156e-05,
+      "loss": 0.6592,
+      "step": 3363
+    },
+    {
+      "epoch": 2.031391488077271,
+      "grad_norm": 0.158203125,
+      "learning_rate": 6.658364466583646e-05,
+      "loss": 0.6675,
+      "step": 3364
+    },
+    {
+      "epoch": 2.031995170540296,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 6.654213366542133e-05,
+      "loss": 0.6109,
+      "step": 3365
+    },
+    {
+      "epoch": 2.03259885300332,
+      "grad_norm": 0.14453125,
+      "learning_rate": 6.650062266500623e-05,
+      "loss": 0.5674,
+      "step": 3366
+    },
+    {
+      "epoch": 2.0332025354663448,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 6.645911166459113e-05,
+      "loss": 0.5863,
+      "step": 3367
+    },
+    {
+      "epoch": 2.033806217929369,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 6.6417600664176e-05,
+      "loss": 0.7262,
+      "step": 3368
+    },
+    {
+      "epoch": 2.0344099003923937,
+      "grad_norm": 0.142578125,
+      "learning_rate": 6.63760896637609e-05,
+      "loss": 0.8392,
+      "step": 3369
+    },
+    {
+      "epoch": 2.035013582855418,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 6.633457866334579e-05,
+      "loss": 0.5431,
+      "step": 3370
+    },
+    {
+      "epoch": 2.0356172653184426,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 6.629306766293068e-05,
+      "loss": 0.4897,
+      "step": 3371
+    },
+    {
+      "epoch": 2.036220947781467,
+      "grad_norm": 0.1484375,
+      "learning_rate": 6.625155666251556e-05,
+      "loss": 0.7426,
+      "step": 3372
+    },
+    {
+      "epoch": 2.0368246302444915,
+      "grad_norm": 0.138671875,
+      "learning_rate": 6.621004566210046e-05,
+      "loss": 0.8515,
+      "step": 3373
+    },
+    {
+      "epoch": 2.0374283127075157,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 6.616853466168536e-05,
+      "loss": 0.4825,
+      "step": 3374
+    },
+    {
+      "epoch": 2.0380319951705403,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 6.612702366127023e-05,
+      "loss": 0.6163,
+      "step": 3375
+    },
+    {
+      "epoch": 2.0386356776335646,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 6.608551266085513e-05,
+      "loss": 0.4817,
+      "step": 3376
+    },
+    {
+      "epoch": 2.0392393600965892,
+      "grad_norm": 0.142578125,
+      "learning_rate": 6.604400166044002e-05,
+      "loss": 0.7172,
+      "step": 3377
+    },
+    {
+      "epoch": 2.0398430425596135,
+      "grad_norm": 0.140625,
+      "learning_rate": 6.600249066002491e-05,
+      "loss": 0.6068,
+      "step": 3378
+    },
+    {
+      "epoch": 2.040446725022638,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 6.59609796596098e-05,
+      "loss": 0.8924,
+      "step": 3379
+    },
+    {
+      "epoch": 2.0410504074856624,
+      "grad_norm": 0.134765625,
+      "learning_rate": 6.59194686591947e-05,
+      "loss": 0.8522,
+      "step": 3380
+    },
+    {
+      "epoch": 2.041654089948687,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 6.587795765877958e-05,
+      "loss": 0.7332,
+      "step": 3381
+    },
+    {
+      "epoch": 2.0422577724117112,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 6.583644665836447e-05,
+      "loss": 0.5677,
+      "step": 3382
+    },
+    {
+      "epoch": 2.042861454874736,
+      "grad_norm": 0.15234375,
+      "learning_rate": 6.579493565794937e-05,
+      "loss": 0.8008,
+      "step": 3383
+    },
+    {
+      "epoch": 2.04346513733776,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 6.575342465753424e-05,
+      "loss": 0.6705,
+      "step": 3384
+    },
+    {
+      "epoch": 2.044068819800785,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 6.571191365711914e-05,
+      "loss": 0.5933,
+      "step": 3385
+    },
+    {
+      "epoch": 2.044672502263809,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.567040265670403e-05,
+      "loss": 0.8533,
+      "step": 3386
+    },
+    {
+      "epoch": 2.0452761847268337,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.562889165628891e-05,
+      "loss": 0.7038,
+      "step": 3387
+    },
+    {
+      "epoch": 2.045879867189858,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 6.558738065587381e-05,
+      "loss": 0.584,
+      "step": 3388
+    },
+    {
+      "epoch": 2.0464835496528826,
+      "grad_norm": 0.14453125,
+      "learning_rate": 6.55458696554587e-05,
+      "loss": 0.9452,
+      "step": 3389
+    },
+    {
+      "epoch": 2.047087232115907,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 6.550435865504359e-05,
+      "loss": 0.5957,
+      "step": 3390
+    },
+    {
+      "epoch": 2.0476909145789315,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 6.546284765462847e-05,
+      "loss": 0.6208,
+      "step": 3391
+    },
+    {
+      "epoch": 2.0482945970419557,
+      "grad_norm": 0.15234375,
+      "learning_rate": 6.542133665421337e-05,
+      "loss": 0.7178,
+      "step": 3392
+    },
+    {
+      "epoch": 2.0488982795049804,
+      "grad_norm": 0.255859375,
+      "learning_rate": 6.537982565379826e-05,
+      "loss": 0.7035,
+      "step": 3393
+    },
+    {
+      "epoch": 2.049501961968005,
+      "grad_norm": 0.154296875,
+      "learning_rate": 6.533831465338315e-05,
+      "loss": 0.7374,
+      "step": 3394
+    },
+    {
+      "epoch": 2.0501056444310293,
+      "grad_norm": 0.130859375,
+      "learning_rate": 6.529680365296805e-05,
+      "loss": 0.5717,
+      "step": 3395
+    },
+    {
+      "epoch": 2.050709326894054,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 6.525529265255293e-05,
+      "loss": 0.5723,
+      "step": 3396
+    },
+    {
+      "epoch": 2.051313009357078,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 6.521378165213782e-05,
+      "loss": 0.6278,
+      "step": 3397
+    },
+    {
+      "epoch": 2.051916691820103,
+      "grad_norm": 0.140625,
+      "learning_rate": 6.51722706517227e-05,
+      "loss": 0.5791,
+      "step": 3398
+    },
+    {
+      "epoch": 2.052520374283127,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 6.51307596513076e-05,
+      "loss": 0.5202,
+      "step": 3399
+    },
+    {
+      "epoch": 2.0531240567461517,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 6.508924865089249e-05,
+      "loss": 0.4888,
+      "step": 3400
+    },
+    {
+      "epoch": 2.053727739209176,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 6.504773765047738e-05,
+      "loss": 0.5308,
+      "step": 3401
+    },
+    {
+      "epoch": 2.0543314216722006,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 6.500622665006228e-05,
+      "loss": 0.4807,
+      "step": 3402
+    },
+    {
+      "epoch": 2.054935104135225,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.496471564964715e-05,
+      "loss": 0.5628,
+      "step": 3403
+    },
+    {
+      "epoch": 2.0555387865982495,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.492320464923205e-05,
+      "loss": 0.5691,
+      "step": 3404
+    },
+    {
+      "epoch": 2.0561424690612737,
+      "grad_norm": 0.15625,
+      "learning_rate": 6.488169364881694e-05,
+      "loss": 0.4539,
+      "step": 3405
+    },
+    {
+      "epoch": 2.0567461515242984,
+      "grad_norm": 0.162109375,
+      "learning_rate": 6.484018264840182e-05,
+      "loss": 0.598,
+      "step": 3406
+    },
+    {
+      "epoch": 2.0573498339873226,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.479867164798672e-05,
+      "loss": 0.417,
+      "step": 3407
+    },
+    {
+      "epoch": 2.0579535164503473,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.475716064757161e-05,
+      "loss": 0.3549,
+      "step": 3408
+    },
+    {
+      "epoch": 2.0585571989133715,
+      "grad_norm": 0.19921875,
+      "learning_rate": 6.47156496471565e-05,
+      "loss": 0.4513,
+      "step": 3409
+    },
+    {
+      "epoch": 2.059160881376396,
+      "grad_norm": 0.18359375,
+      "learning_rate": 6.467413864674138e-05,
+      "loss": 0.3228,
+      "step": 3410
+    },
+    {
+      "epoch": 2.0597645638394204,
+      "grad_norm": 0.2109375,
+      "learning_rate": 6.463262764632628e-05,
+      "loss": 0.3144,
+      "step": 3411
+    },
+    {
+      "epoch": 2.060368246302445,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.459111664591117e-05,
+      "loss": 0.2016,
+      "step": 3412
+    },
+    {
+      "epoch": 2.0609719287654693,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 6.454960564549606e-05,
+      "loss": 0.5913,
+      "step": 3413
+    },
+    {
+      "epoch": 2.061575611228494,
+      "grad_norm": 0.14453125,
+      "learning_rate": 6.450809464508096e-05,
+      "loss": 1.0476,
+      "step": 3414
+    },
+    {
+      "epoch": 2.062179293691518,
+      "grad_norm": 0.14453125,
+      "learning_rate": 6.446658364466583e-05,
+      "loss": 0.6198,
+      "step": 3415
+    },
+    {
+      "epoch": 2.062782976154543,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 6.442507264425073e-05,
+      "loss": 0.5562,
+      "step": 3416
+    },
+    {
+      "epoch": 2.063386658617567,
+      "grad_norm": 0.150390625,
+      "learning_rate": 6.438356164383562e-05,
+      "loss": 0.646,
+      "step": 3417
+    },
+    {
+      "epoch": 2.0639903410805918,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 6.434205064342052e-05,
+      "loss": 0.6686,
+      "step": 3418
+    },
+    {
+      "epoch": 2.064594023543616,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 6.43005396430054e-05,
+      "loss": 0.8418,
+      "step": 3419
+    },
+    {
+      "epoch": 2.0651977060066407,
+      "grad_norm": 0.142578125,
+      "learning_rate": 6.425902864259029e-05,
+      "loss": 1.04,
+      "step": 3420
+    },
+    {
+      "epoch": 2.065801388469665,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.421751764217519e-05,
+      "loss": 0.6007,
+      "step": 3421
+    },
+    {
+      "epoch": 2.0664050709326895,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 6.417600664176006e-05,
+      "loss": 0.6781,
+      "step": 3422
+    },
+    {
+      "epoch": 2.0670087533957138,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 6.413449564134496e-05,
+      "loss": 0.6264,
+      "step": 3423
+    },
+    {
+      "epoch": 2.0676124358587384,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.409298464092985e-05,
+      "loss": 0.7689,
+      "step": 3424
+    },
+    {
+      "epoch": 2.0682161183217627,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 6.405147364051474e-05,
+      "loss": 0.5546,
+      "step": 3425
+    },
+    {
+      "epoch": 2.0688198007847873,
+      "grad_norm": 0.130859375,
+      "learning_rate": 6.400996264009964e-05,
+      "loss": 0.7563,
+      "step": 3426
+    },
+    {
+      "epoch": 2.0694234832478116,
+      "grad_norm": 0.13671875,
+      "learning_rate": 6.396845163968452e-05,
+      "loss": 0.6584,
+      "step": 3427
+    },
+    {
+      "epoch": 2.0700271657108362,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 6.392694063926941e-05,
+      "loss": 0.6751,
+      "step": 3428
+    },
+    {
+      "epoch": 2.0706308481738604,
+      "grad_norm": 0.130859375,
+      "learning_rate": 6.38854296388543e-05,
+      "loss": 0.607,
+      "step": 3429
+    },
+    {
+      "epoch": 2.071234530636885,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.38439186384392e-05,
+      "loss": 0.6411,
+      "step": 3430
+    },
+    {
+      "epoch": 2.0718382130999093,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 6.380240763802408e-05,
+      "loss": 0.6765,
+      "step": 3431
+    },
+    {
+      "epoch": 2.072441895562934,
+      "grad_norm": 0.14453125,
+      "learning_rate": 6.376089663760897e-05,
+      "loss": 1.0625,
+      "step": 3432
+    },
+    {
+      "epoch": 2.0730455780259582,
+      "grad_norm": 0.13671875,
+      "learning_rate": 6.371938563719387e-05,
+      "loss": 0.8033,
+      "step": 3433
+    },
+    {
+      "epoch": 2.073649260488983,
+      "grad_norm": 0.134765625,
+      "learning_rate": 6.367787463677874e-05,
+      "loss": 0.5758,
+      "step": 3434
+    },
+    {
+      "epoch": 2.074252942952007,
+      "grad_norm": 0.154296875,
+      "learning_rate": 6.363636363636364e-05,
+      "loss": 0.8254,
+      "step": 3435
+    },
+    {
+      "epoch": 2.074856625415032,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 6.359485263594853e-05,
+      "loss": 0.5519,
+      "step": 3436
+    },
+    {
+      "epoch": 2.075460307878056,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 6.355334163553341e-05,
+      "loss": 0.5636,
+      "step": 3437
+    },
+    {
+      "epoch": 2.0760639903410807,
+      "grad_norm": 0.1640625,
+      "learning_rate": 6.351183063511831e-05,
+      "loss": 0.7238,
+      "step": 3438
+    },
+    {
+      "epoch": 2.076667672804105,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 6.34703196347032e-05,
+      "loss": 0.6205,
+      "step": 3439
+    },
+    {
+      "epoch": 2.0772713552671296,
+      "grad_norm": 0.138671875,
+      "learning_rate": 6.342880863428809e-05,
+      "loss": 0.5848,
+      "step": 3440
+    },
+    {
+      "epoch": 2.077875037730154,
+      "grad_norm": 0.140625,
+      "learning_rate": 6.338729763387297e-05,
+      "loss": 0.7892,
+      "step": 3441
+    },
+    {
+      "epoch": 2.0784787201931785,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 6.334578663345787e-05,
+      "loss": 0.6707,
+      "step": 3442
+    },
+    {
+      "epoch": 2.0790824026562027,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 6.330427563304276e-05,
+      "loss": 0.7932,
+      "step": 3443
+    },
+    {
+      "epoch": 2.0796860851192274,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 6.326276463262765e-05,
+      "loss": 0.662,
+      "step": 3444
+    },
+    {
+      "epoch": 2.0802897675822516,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 6.322125363221255e-05,
+      "loss": 0.7459,
+      "step": 3445
+    },
+    {
+      "epoch": 2.0808934500452763,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 6.317974263179743e-05,
+      "loss": 0.5336,
+      "step": 3446
+    },
+    {
+      "epoch": 2.0814971325083005,
+      "grad_norm": 0.146484375,
+      "learning_rate": 6.313823163138232e-05,
+      "loss": 0.619,
+      "step": 3447
+    },
+    {
+      "epoch": 2.082100814971325,
+      "grad_norm": 0.146484375,
+      "learning_rate": 6.30967206309672e-05,
+      "loss": 0.5999,
+      "step": 3448
+    },
+    {
+      "epoch": 2.0827044974343494,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 6.30552096305521e-05,
+      "loss": 0.5155,
+      "step": 3449
+    },
+    {
+      "epoch": 2.083308179897374,
+      "grad_norm": 0.162109375,
+      "learning_rate": 6.301369863013699e-05,
+      "loss": 0.5144,
+      "step": 3450
+    },
+    {
+      "epoch": 2.0839118623603983,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 6.297218762972188e-05,
+      "loss": 0.5631,
+      "step": 3451
+    },
+    {
+      "epoch": 2.084515544823423,
+      "grad_norm": 0.15234375,
+      "learning_rate": 6.293067662930678e-05,
+      "loss": 0.6323,
+      "step": 3452
+    },
+    {
+      "epoch": 2.085119227286447,
+      "grad_norm": 0.162109375,
+      "learning_rate": 6.288916562889165e-05,
+      "loss": 0.447,
+      "step": 3453
+    },
+    {
+      "epoch": 2.085722909749472,
+      "grad_norm": 0.16015625,
+      "learning_rate": 6.284765462847655e-05,
+      "loss": 0.5228,
+      "step": 3454
+    },
+    {
+      "epoch": 2.086326592212496,
+      "grad_norm": 0.166015625,
+      "learning_rate": 6.280614362806144e-05,
+      "loss": 0.4278,
+      "step": 3455
+    },
+    {
+      "epoch": 2.0869302746755207,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.276463262764632e-05,
+      "loss": 0.4298,
+      "step": 3456
+    },
+    {
+      "epoch": 2.087533957138545,
+      "grad_norm": 0.18359375,
+      "learning_rate": 6.272312162723122e-05,
+      "loss": 0.4683,
+      "step": 3457
+    },
+    {
+      "epoch": 2.0881376396015696,
+      "grad_norm": 0.18359375,
+      "learning_rate": 6.268161062681611e-05,
+      "loss": 0.4114,
+      "step": 3458
+    },
+    {
+      "epoch": 2.088741322064594,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 6.2640099626401e-05,
+      "loss": 0.4019,
+      "step": 3459
+    },
+    {
+      "epoch": 2.0893450045276185,
+      "grad_norm": 0.203125,
+      "learning_rate": 6.259858862598588e-05,
+      "loss": 0.3359,
+      "step": 3460
+    },
+    {
+      "epoch": 2.0899486869906427,
+      "grad_norm": 0.205078125,
+      "learning_rate": 6.255707762557078e-05,
+      "loss": 0.2717,
+      "step": 3461
+    },
+    {
+      "epoch": 2.0905523694536674,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 6.251556662515566e-05,
+      "loss": 0.1975,
+      "step": 3462
+    },
+    {
+      "epoch": 2.0911560519166916,
+      "grad_norm": 0.142578125,
+      "learning_rate": 6.247405562474056e-05,
+      "loss": 0.9222,
+      "step": 3463
+    },
+    {
+      "epoch": 2.0917597343797163,
+      "grad_norm": 0.142578125,
+      "learning_rate": 6.243254462432546e-05,
+      "loss": 0.5585,
+      "step": 3464
+    },
+    {
+      "epoch": 2.0923634168427405,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 6.239103362391034e-05,
+      "loss": 0.653,
+      "step": 3465
+    },
+    {
+      "epoch": 2.092967099305765,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 6.234952262349523e-05,
+      "loss": 0.6304,
+      "step": 3466
+    },
+    {
+      "epoch": 2.0935707817687894,
+      "grad_norm": 0.154296875,
+      "learning_rate": 6.230801162308012e-05,
+      "loss": 0.6062,
+      "step": 3467
+    },
+    {
+      "epoch": 2.094174464231814,
+      "grad_norm": 0.15625,
+      "learning_rate": 6.226650062266502e-05,
+      "loss": 0.5847,
+      "step": 3468
+    },
+    {
+      "epoch": 2.0947781466948383,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 6.222498962224989e-05,
+      "loss": 0.6474,
+      "step": 3469
+    },
+    {
+      "epoch": 2.095381829157863,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 6.218347862183479e-05,
+      "loss": 0.6167,
+      "step": 3470
+    },
+    {
+      "epoch": 2.095985511620887,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 6.214196762141969e-05,
+      "loss": 0.6131,
+      "step": 3471
+    },
+    {
+      "epoch": 2.096589194083912,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.210045662100456e-05,
+      "loss": 0.7248,
+      "step": 3472
+    },
+    {
+      "epoch": 2.097192876546936,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.205894562058946e-05,
+      "loss": 0.5809,
+      "step": 3473
+    },
+    {
+      "epoch": 2.0977965590099608,
+      "grad_norm": 0.150390625,
+      "learning_rate": 6.201743462017435e-05,
+      "loss": 0.6752,
+      "step": 3474
+    },
+    {
+      "epoch": 2.0984002414729854,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 6.197592361975924e-05,
+      "loss": 0.6793,
+      "step": 3475
+    },
+    {
+      "epoch": 2.0990039239360097,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 6.193441261934412e-05,
+      "loss": 0.6169,
+      "step": 3476
+    },
+    {
+      "epoch": 2.0996076063990343,
+      "grad_norm": 0.1484375,
+      "learning_rate": 6.189290161892902e-05,
+      "loss": 0.5645,
+      "step": 3477
+    },
+    {
+      "epoch": 2.1002112888620585,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 6.185139061851391e-05,
+      "loss": 0.5911,
+      "step": 3478
+    },
+    {
+      "epoch": 2.100814971325083,
+      "grad_norm": 0.146484375,
+      "learning_rate": 6.18098796180988e-05,
+      "loss": 0.5954,
+      "step": 3479
+    },
+    {
+      "epoch": 2.1014186537881074,
+      "grad_norm": 0.134765625,
+      "learning_rate": 6.17683686176837e-05,
+      "loss": 0.5374,
+      "step": 3480
+    },
+    {
+      "epoch": 2.102022336251132,
+      "grad_norm": 0.14453125,
+      "learning_rate": 6.172685761726857e-05,
+      "loss": 0.5509,
+      "step": 3481
+    },
+    {
+      "epoch": 2.1026260187141563,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 6.168534661685347e-05,
+      "loss": 0.6542,
+      "step": 3482
+    },
+    {
+      "epoch": 2.103229701177181,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 6.164383561643835e-05,
+      "loss": 0.7288,
+      "step": 3483
+    },
+    {
+      "epoch": 2.1038333836402052,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 6.160232461602324e-05,
+      "loss": 0.6775,
+      "step": 3484
+    },
+    {
+      "epoch": 2.10443706610323,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 6.156081361560814e-05,
+      "loss": 0.6655,
+      "step": 3485
+    },
+    {
+      "epoch": 2.105040748566254,
+      "grad_norm": 0.142578125,
+      "learning_rate": 6.151930261519303e-05,
+      "loss": 0.5989,
+      "step": 3486
+    },
+    {
+      "epoch": 2.105644431029279,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 6.147779161477793e-05,
+      "loss": 0.6795,
+      "step": 3487
+    },
+    {
+      "epoch": 2.106248113492303,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 6.14362806143628e-05,
+      "loss": 0.6076,
+      "step": 3488
+    },
+    {
+      "epoch": 2.1068517959553277,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 6.13947696139477e-05,
+      "loss": 0.5628,
+      "step": 3489
+    },
+    {
+      "epoch": 2.107455478418352,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 6.13532586135326e-05,
+      "loss": 0.6294,
+      "step": 3490
+    },
+    {
+      "epoch": 2.1080591608813766,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 6.131174761311747e-05,
+      "loss": 0.6435,
+      "step": 3491
+    },
+    {
+      "epoch": 2.108662843344401,
+      "grad_norm": 0.13671875,
+      "learning_rate": 6.127023661270237e-05,
+      "loss": 0.6564,
+      "step": 3492
+    },
+    {
+      "epoch": 2.1092665258074255,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 6.122872561228726e-05,
+      "loss": 0.6102,
+      "step": 3493
+    },
+    {
+      "epoch": 2.1098702082704497,
+      "grad_norm": 0.1328125,
+      "learning_rate": 6.118721461187215e-05,
+      "loss": 0.5999,
+      "step": 3494
+    },
+    {
+      "epoch": 2.1104738907334744,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 6.114570361145703e-05,
+      "loss": 0.6853,
+      "step": 3495
+    },
+    {
+      "epoch": 2.1110775731964986,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 6.110419261104193e-05,
+      "loss": 0.5913,
+      "step": 3496
+    },
+    {
+      "epoch": 2.1116812556595232,
+      "grad_norm": 0.1328125,
+      "learning_rate": 6.106268161062682e-05,
+      "loss": 0.5238,
+      "step": 3497
+    },
+    {
+      "epoch": 2.1122849381225475,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 6.1021170610211706e-05,
+      "loss": 0.5639,
+      "step": 3498
+    },
+    {
+      "epoch": 2.112888620585572,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 6.09796596097966e-05,
+      "loss": 0.5235,
+      "step": 3499
+    },
+    {
+      "epoch": 2.1134923030485964,
+      "grad_norm": 0.1484375,
+      "learning_rate": 6.0938148609381486e-05,
+      "loss": 0.5606,
+      "step": 3500
+    },
+    {
+      "epoch": 2.1134923030485964,
+      "eval_loss": 0.617886483669281,
+      "eval_runtime": 1059.7393,
+      "eval_samples_per_second": 2.633,
+      "eval_steps_per_second": 0.329,
+      "step": 3500
+    },
+    {
+      "epoch": 2.114095985511621,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 6.089663760896638e-05,
+      "loss": 0.5319,
+      "step": 3501
+    },
+    {
+      "epoch": 2.1146996679746453,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 6.085512660855127e-05,
+      "loss": 0.4808,
+      "step": 3502
+    },
+    {
+      "epoch": 2.11530335043767,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 6.081361560813615e-05,
+      "loss": 0.5266,
+      "step": 3503
+    },
+    {
+      "epoch": 2.115907032900694,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.0772104607721045e-05,
+      "loss": 0.5523,
+      "step": 3504
+    },
+    {
+      "epoch": 2.116510715363719,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.073059360730594e-05,
+      "loss": 0.4785,
+      "step": 3505
+    },
+    {
+      "epoch": 2.117114397826743,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.0689082606890825e-05,
+      "loss": 0.4426,
+      "step": 3506
+    },
+    {
+      "epoch": 2.1177180802897677,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.064757160647572e-05,
+      "loss": 0.4489,
+      "step": 3507
+    },
+    {
+      "epoch": 2.118321762752792,
+      "grad_norm": 0.19921875,
+      "learning_rate": 6.060606060606061e-05,
+      "loss": 0.4412,
+      "step": 3508
+    },
+    {
+      "epoch": 2.1189254452158166,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.0564549605645505e-05,
+      "loss": 0.3421,
+      "step": 3509
+    },
+    {
+      "epoch": 2.119529127678841,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 6.0523038605230384e-05,
+      "loss": 0.3627,
+      "step": 3510
+    },
+    {
+      "epoch": 2.1201328101418655,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.048152760481528e-05,
+      "loss": 0.265,
+      "step": 3511
+    },
+    {
+      "epoch": 2.1207364926048897,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 6.044001660440017e-05,
+      "loss": 0.2138,
+      "step": 3512
+    },
+    {
+      "epoch": 2.1213401750679144,
+      "grad_norm": 0.138671875,
+      "learning_rate": 6.039850560398506e-05,
+      "loss": 0.574,
+      "step": 3513
+    },
+    {
+      "epoch": 2.1219438575309386,
+      "grad_norm": 0.158203125,
+      "learning_rate": 6.035699460356995e-05,
+      "loss": 0.754,
+      "step": 3514
+    },
+    {
+      "epoch": 2.1225475399939633,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 6.0315483603154844e-05,
+      "loss": 0.6438,
+      "step": 3515
+    },
+    {
+      "epoch": 2.1231512224569875,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 6.0273972602739724e-05,
+      "loss": 0.6454,
+      "step": 3516
+    },
+    {
+      "epoch": 2.123754904920012,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 6.023246160232462e-05,
+      "loss": 0.5961,
+      "step": 3517
+    },
+    {
+      "epoch": 2.1243585873830364,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 6.019095060190951e-05,
+      "loss": 0.6337,
+      "step": 3518
+    },
+    {
+      "epoch": 2.124962269846061,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 6.01494396014944e-05,
+      "loss": 0.58,
+      "step": 3519
+    },
+    {
+      "epoch": 2.1255659523090853,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.010792860107929e-05,
+      "loss": 0.8145,
+      "step": 3520
+    },
+    {
+      "epoch": 2.12616963477211,
+      "grad_norm": 0.12890625,
+      "learning_rate": 6.006641760066418e-05,
+      "loss": 0.5888,
+      "step": 3521
+    },
+    {
+      "epoch": 2.126773317235134,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 6.002490660024906e-05,
+      "loss": 0.5517,
+      "step": 3522
+    },
+    {
+      "epoch": 2.127376999698159,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.9983395599833956e-05,
+      "loss": 0.7016,
+      "step": 3523
+    },
+    {
+      "epoch": 2.127980682161183,
+      "grad_norm": 0.154296875,
+      "learning_rate": 5.994188459941885e-05,
+      "loss": 0.6212,
+      "step": 3524
+    },
+    {
+      "epoch": 2.1285843646242077,
+      "grad_norm": 0.140625,
+      "learning_rate": 5.9900373599003736e-05,
+      "loss": 0.8909,
+      "step": 3525
+    },
+    {
+      "epoch": 2.129188047087232,
+      "grad_norm": 0.150390625,
+      "learning_rate": 5.985886259858863e-05,
+      "loss": 0.8154,
+      "step": 3526
+    },
+    {
+      "epoch": 2.1297917295502566,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 5.981735159817352e-05,
+      "loss": 0.6854,
+      "step": 3527
+    },
+    {
+      "epoch": 2.130395412013281,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 5.97758405977584e-05,
+      "loss": 0.5829,
+      "step": 3528
+    },
+    {
+      "epoch": 2.1309990944763055,
+      "grad_norm": 0.134765625,
+      "learning_rate": 5.9734329597343295e-05,
+      "loss": 0.7917,
+      "step": 3529
+    },
+    {
+      "epoch": 2.1316027769393298,
+      "grad_norm": 0.138671875,
+      "learning_rate": 5.969281859692819e-05,
+      "loss": 0.6034,
+      "step": 3530
+    },
+    {
+      "epoch": 2.1322064594023544,
+      "grad_norm": 0.140625,
+      "learning_rate": 5.965130759651308e-05,
+      "loss": 0.6514,
+      "step": 3531
+    },
+    {
+      "epoch": 2.1328101418653787,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 5.960979659609797e-05,
+      "loss": 0.5277,
+      "step": 3532
+    },
+    {
+      "epoch": 2.1334138243284033,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 5.956828559568286e-05,
+      "loss": 0.6354,
+      "step": 3533
+    },
+    {
+      "epoch": 2.1340175067914275,
+      "grad_norm": 0.140625,
+      "learning_rate": 5.9526774595267755e-05,
+      "loss": 0.7139,
+      "step": 3534
+    },
+    {
+      "epoch": 2.134621189254452,
+      "grad_norm": 0.1484375,
+      "learning_rate": 5.9485263594852635e-05,
+      "loss": 0.6376,
+      "step": 3535
+    },
+    {
+      "epoch": 2.1352248717174764,
+      "grad_norm": 0.1640625,
+      "learning_rate": 5.944375259443753e-05,
+      "loss": 0.6315,
+      "step": 3536
+    },
+    {
+      "epoch": 2.135828554180501,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 5.940224159402242e-05,
+      "loss": 0.5425,
+      "step": 3537
+    },
+    {
+      "epoch": 2.1364322366435253,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 5.936073059360731e-05,
+      "loss": 0.6503,
+      "step": 3538
+    },
+    {
+      "epoch": 2.13703591910655,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 5.93192195931922e-05,
+      "loss": 0.6238,
+      "step": 3539
+    },
+    {
+      "epoch": 2.137639601569574,
+      "grad_norm": 0.185546875,
+      "learning_rate": 5.9277708592777094e-05,
+      "loss": 0.6117,
+      "step": 3540
+    },
+    {
+      "epoch": 2.138243284032599,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 5.9236197592361974e-05,
+      "loss": 0.7165,
+      "step": 3541
+    },
+    {
+      "epoch": 2.138846966495623,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 5.919468659194687e-05,
+      "loss": 0.7887,
+      "step": 3542
+    },
+    {
+      "epoch": 2.139450648958648,
+      "grad_norm": 0.15625,
+      "learning_rate": 5.915317559153176e-05,
+      "loss": 0.7581,
+      "step": 3543
+    },
+    {
+      "epoch": 2.140054331421672,
+      "grad_norm": 0.142578125,
+      "learning_rate": 5.911166459111665e-05,
+      "loss": 0.6213,
+      "step": 3544
+    },
+    {
+      "epoch": 2.1406580138846967,
+      "grad_norm": 0.142578125,
+      "learning_rate": 5.907015359070154e-05,
+      "loss": 0.7873,
+      "step": 3545
+    },
+    {
+      "epoch": 2.141261696347721,
+      "grad_norm": 0.142578125,
+      "learning_rate": 5.902864259028643e-05,
+      "loss": 0.6151,
+      "step": 3546
+    },
+    {
+      "epoch": 2.1418653788107456,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 5.898713158987131e-05,
+      "loss": 0.5618,
+      "step": 3547
+    },
+    {
+      "epoch": 2.14246906127377,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 5.8945620589456206e-05,
+      "loss": 0.5792,
+      "step": 3548
+    },
+    {
+      "epoch": 2.1430727437367945,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 5.89041095890411e-05,
+      "loss": 0.5894,
+      "step": 3549
+    },
+    {
+      "epoch": 2.1436764261998187,
+      "grad_norm": 0.15234375,
+      "learning_rate": 5.8862598588625986e-05,
+      "loss": 0.5326,
+      "step": 3550
+    },
+    {
+      "epoch": 2.1442801086628434,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 5.882108758821088e-05,
+      "loss": 0.5008,
+      "step": 3551
+    },
+    {
+      "epoch": 2.144883791125868,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 5.877957658779577e-05,
+      "loss": 0.4708,
+      "step": 3552
+    },
+    {
+      "epoch": 2.1454874735888922,
+      "grad_norm": 0.158203125,
+      "learning_rate": 5.8738065587380666e-05,
+      "loss": 0.4923,
+      "step": 3553
+    },
+    {
+      "epoch": 2.1460911560519165,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 5.8696554586965546e-05,
+      "loss": 0.4978,
+      "step": 3554
+    },
+    {
+      "epoch": 2.146694838514941,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 5.865504358655044e-05,
+      "loss": 0.4812,
+      "step": 3555
+    },
+    {
+      "epoch": 2.147298520977966,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 5.861353258613533e-05,
+      "loss": 0.4216,
+      "step": 3556
+    },
+    {
+      "epoch": 2.14790220344099,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.857202158572022e-05,
+      "loss": 0.4385,
+      "step": 3557
+    },
+    {
+      "epoch": 2.1485058859040143,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.853051058530511e-05,
+      "loss": 0.3954,
+      "step": 3558
+    },
+    {
+      "epoch": 2.149109568367039,
+      "grad_norm": 0.189453125,
+      "learning_rate": 5.8488999584890005e-05,
+      "loss": 0.3699,
+      "step": 3559
+    },
+    {
+      "epoch": 2.1497132508300636,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 5.8447488584474885e-05,
+      "loss": 0.332,
+      "step": 3560
+    },
+    {
+      "epoch": 2.150316933293088,
+      "grad_norm": 0.208984375,
+      "learning_rate": 5.840597758405978e-05,
+      "loss": 0.2789,
+      "step": 3561
+    },
+    {
+      "epoch": 2.1509206157561125,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 5.836446658364467e-05,
+      "loss": 0.2135,
+      "step": 3562
+    },
+    {
+      "epoch": 2.1515242982191367,
+      "grad_norm": 0.15234375,
+      "learning_rate": 5.832295558322956e-05,
+      "loss": 0.6707,
+      "step": 3563
+    },
+    {
+      "epoch": 2.1521279806821614,
+      "grad_norm": 0.138671875,
+      "learning_rate": 5.828144458281445e-05,
+      "loss": 0.5971,
+      "step": 3564
+    },
+    {
+      "epoch": 2.1527316631451856,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 5.8239933582399344e-05,
+      "loss": 0.7411,
+      "step": 3565
+    },
+    {
+      "epoch": 2.1533353456082103,
+      "grad_norm": 0.150390625,
+      "learning_rate": 5.8198422581984224e-05,
+      "loss": 0.7833,
+      "step": 3566
+    },
+    {
+      "epoch": 2.1539390280712345,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 5.815691158156912e-05,
+      "loss": 0.519,
+      "step": 3567
+    },
+    {
+      "epoch": 2.154542710534259,
+      "grad_norm": 0.138671875,
+      "learning_rate": 5.811540058115401e-05,
+      "loss": 0.5645,
+      "step": 3568
+    },
+    {
+      "epoch": 2.1551463929972834,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.80738895807389e-05,
+      "loss": 0.5951,
+      "step": 3569
+    },
+    {
+      "epoch": 2.155750075460308,
+      "grad_norm": 0.150390625,
+      "learning_rate": 5.803237858032379e-05,
+      "loss": 0.7003,
+      "step": 3570
+    },
+    {
+      "epoch": 2.1563537579233323,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.7990867579908683e-05,
+      "loss": 0.7031,
+      "step": 3571
+    },
+    {
+      "epoch": 2.156957440386357,
+      "grad_norm": 0.158203125,
+      "learning_rate": 5.794935657949356e-05,
+      "loss": 0.8678,
+      "step": 3572
+    },
+    {
+      "epoch": 2.157561122849381,
+      "grad_norm": 0.16015625,
+      "learning_rate": 5.7907845579078456e-05,
+      "loss": 0.6421,
+      "step": 3573
+    },
+    {
+      "epoch": 2.158164805312406,
+      "grad_norm": 0.15234375,
+      "learning_rate": 5.786633457866335e-05,
+      "loss": 0.631,
+      "step": 3574
+    },
+    {
+      "epoch": 2.15876848777543,
+      "grad_norm": 0.15234375,
+      "learning_rate": 5.782482357824824e-05,
+      "loss": 0.5944,
+      "step": 3575
+    },
+    {
+      "epoch": 2.1593721702384547,
+      "grad_norm": 0.1484375,
+      "learning_rate": 5.778331257783313e-05,
+      "loss": 0.6048,
+      "step": 3576
+    },
+    {
+      "epoch": 2.159975852701479,
+      "grad_norm": 0.146484375,
+      "learning_rate": 5.774180157741802e-05,
+      "loss": 0.6322,
+      "step": 3577
+    },
+    {
+      "epoch": 2.1605795351645036,
+      "grad_norm": 0.134765625,
+      "learning_rate": 5.7700290577002916e-05,
+      "loss": 0.8564,
+      "step": 3578
+    },
+    {
+      "epoch": 2.161183217627528,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 5.7658779576587796e-05,
+      "loss": 0.6573,
+      "step": 3579
+    },
+    {
+      "epoch": 2.1617869000905525,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 5.761726857617269e-05,
+      "loss": 0.6145,
+      "step": 3580
+    },
+    {
+      "epoch": 2.1623905825535767,
+      "grad_norm": 0.15234375,
+      "learning_rate": 5.757575757575758e-05,
+      "loss": 0.6561,
+      "step": 3581
+    },
+    {
+      "epoch": 2.1629942650166014,
+      "grad_norm": 0.150390625,
+      "learning_rate": 5.753424657534247e-05,
+      "loss": 0.5926,
+      "step": 3582
+    },
+    {
+      "epoch": 2.1635979474796256,
+      "grad_norm": 0.15234375,
+      "learning_rate": 5.749273557492736e-05,
+      "loss": 0.5751,
+      "step": 3583
+    },
+    {
+      "epoch": 2.1642016299426503,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 5.7451224574512255e-05,
+      "loss": 0.6293,
+      "step": 3584
+    },
+    {
+      "epoch": 2.1648053124056745,
+      "grad_norm": 0.13671875,
+      "learning_rate": 5.7409713574097135e-05,
+      "loss": 0.6692,
+      "step": 3585
+    },
+    {
+      "epoch": 2.165408994868699,
+      "grad_norm": 0.12890625,
+      "learning_rate": 5.736820257368203e-05,
+      "loss": 0.7354,
+      "step": 3586
+    },
+    {
+      "epoch": 2.1660126773317234,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 5.732669157326692e-05,
+      "loss": 0.6266,
+      "step": 3587
+    },
+    {
+      "epoch": 2.166616359794748,
+      "grad_norm": 0.14453125,
+      "learning_rate": 5.72851805728518e-05,
+      "loss": 0.6857,
+      "step": 3588
+    },
+    {
+      "epoch": 2.1672200422577723,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 5.72436695724367e-05,
+      "loss": 0.7397,
+      "step": 3589
+    },
+    {
+      "epoch": 2.167823724720797,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 5.7202158572021594e-05,
+      "loss": 0.5876,
+      "step": 3590
+    },
+    {
+      "epoch": 2.168427407183821,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 5.7160647571606474e-05,
+      "loss": 0.5402,
+      "step": 3591
+    },
+    {
+      "epoch": 2.169031089646846,
+      "grad_norm": 0.142578125,
+      "learning_rate": 5.711913657119137e-05,
+      "loss": 0.6663,
+      "step": 3592
+    },
+    {
+      "epoch": 2.16963477210987,
+      "grad_norm": 0.166015625,
+      "learning_rate": 5.707762557077626e-05,
+      "loss": 0.8411,
+      "step": 3593
+    },
+    {
+      "epoch": 2.1702384545728948,
+      "grad_norm": 0.14453125,
+      "learning_rate": 5.703611457036114e-05,
+      "loss": 0.5573,
+      "step": 3594
+    },
+    {
+      "epoch": 2.170842137035919,
+      "grad_norm": 0.150390625,
+      "learning_rate": 5.6994603569946034e-05,
+      "loss": 0.5112,
+      "step": 3595
+    },
+    {
+      "epoch": 2.1714458194989437,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 5.6953092569530934e-05,
+      "loss": 0.5262,
+      "step": 3596
+    },
+    {
+      "epoch": 2.172049501961968,
+      "grad_norm": 0.140625,
+      "learning_rate": 5.691158156911583e-05,
+      "loss": 0.5988,
+      "step": 3597
+    },
+    {
+      "epoch": 2.1726531844249926,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 5.687007056870071e-05,
+      "loss": 0.5237,
+      "step": 3598
+    },
+    {
+      "epoch": 2.173256866888017,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 5.68285595682856e-05,
+      "loss": 0.7144,
+      "step": 3599
+    },
+    {
+      "epoch": 2.1738605493510414,
+      "grad_norm": 0.158203125,
+      "learning_rate": 5.678704856787049e-05,
+      "loss": 0.5637,
+      "step": 3600
+    },
+    {
+      "epoch": 2.1744642318140657,
+      "grad_norm": 0.15625,
+      "learning_rate": 5.674553756745537e-05,
+      "loss": 0.5446,
+      "step": 3601
+    },
+    {
+      "epoch": 2.1750679142770903,
+      "grad_norm": 0.162109375,
+      "learning_rate": 5.6704026567040266e-05,
+      "loss": 0.5356,
+      "step": 3602
+    },
+    {
+      "epoch": 2.1756715967401146,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 5.6662515566625166e-05,
+      "loss": 0.4565,
+      "step": 3603
+    },
+    {
+      "epoch": 2.1762752792031392,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 5.6621004566210046e-05,
+      "loss": 0.4962,
+      "step": 3604
+    },
+    {
+      "epoch": 2.1768789616661635,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 5.657949356579494e-05,
+      "loss": 0.4551,
+      "step": 3605
+    },
+    {
+      "epoch": 2.177482644129188,
+      "grad_norm": 0.189453125,
+      "learning_rate": 5.653798256537983e-05,
+      "loss": 0.5525,
+      "step": 3606
+    },
+    {
+      "epoch": 2.1780863265922124,
+      "grad_norm": 0.185546875,
+      "learning_rate": 5.649647156496471e-05,
+      "loss": 0.4038,
+      "step": 3607
+    },
+    {
+      "epoch": 2.178690009055237,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 5.6454960564549605e-05,
+      "loss": 0.3714,
+      "step": 3608
+    },
+    {
+      "epoch": 2.1792936915182612,
+      "grad_norm": 0.1875,
+      "learning_rate": 5.64134495641345e-05,
+      "loss": 0.3403,
+      "step": 3609
+    },
+    {
+      "epoch": 2.179897373981286,
+      "grad_norm": 0.220703125,
+      "learning_rate": 5.6371938563719385e-05,
+      "loss": 0.3037,
+      "step": 3610
+    },
+    {
+      "epoch": 2.18050105644431,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.633042756330428e-05,
+      "loss": 0.23,
+      "step": 3611
+    },
+    {
+      "epoch": 2.181104738907335,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.628891656288917e-05,
+      "loss": 0.2189,
+      "step": 3612
+    },
+    {
+      "epoch": 2.181708421370359,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 5.624740556247405e-05,
+      "loss": 0.85,
+      "step": 3613
+    },
+    {
+      "epoch": 2.1823121038333837,
+      "grad_norm": 0.150390625,
+      "learning_rate": 5.6205894562058945e-05,
+      "loss": 0.6516,
+      "step": 3614
+    },
+    {
+      "epoch": 2.182915786296408,
+      "grad_norm": 0.142578125,
+      "learning_rate": 5.616438356164384e-05,
+      "loss": 0.6577,
+      "step": 3615
+    },
+    {
+      "epoch": 2.1835194687594326,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 5.6122872561228724e-05,
+      "loss": 0.5393,
+      "step": 3616
+    },
+    {
+      "epoch": 2.184123151222457,
+      "grad_norm": 0.134765625,
+      "learning_rate": 5.608136156081362e-05,
+      "loss": 0.5632,
+      "step": 3617
+    },
+    {
+      "epoch": 2.1847268336854815,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.603985056039851e-05,
+      "loss": 0.6989,
+      "step": 3618
+    },
+    {
+      "epoch": 2.1853305161485057,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 5.5998339559983404e-05,
+      "loss": 0.5195,
+      "step": 3619
+    },
+    {
+      "epoch": 2.1859341986115304,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 5.5956828559568284e-05,
+      "loss": 0.6129,
+      "step": 3620
+    },
+    {
+      "epoch": 2.1865378810745546,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 5.591531755915318e-05,
+      "loss": 0.6626,
+      "step": 3621
+    },
+    {
+      "epoch": 2.1871415635375793,
+      "grad_norm": 0.14453125,
+      "learning_rate": 5.587380655873807e-05,
+      "loss": 0.8414,
+      "step": 3622
+    },
+    {
+      "epoch": 2.1877452460006035,
+      "grad_norm": 0.140625,
+      "learning_rate": 5.583229555832296e-05,
+      "loss": 0.5994,
+      "step": 3623
+    },
+    {
+      "epoch": 2.188348928463628,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 5.579078455790785e-05,
+      "loss": 0.5107,
+      "step": 3624
+    },
+    {
+      "epoch": 2.1889526109266524,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 5.574927355749274e-05,
+      "loss": 0.612,
+      "step": 3625
+    },
+    {
+      "epoch": 2.189556293389677,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 5.570776255707762e-05,
+      "loss": 0.6054,
+      "step": 3626
+    },
+    {
+      "epoch": 2.1901599758527013,
+      "grad_norm": 0.154296875,
+      "learning_rate": 5.5666251556662516e-05,
+      "loss": 0.6852,
+      "step": 3627
+    },
+    {
+      "epoch": 2.190763658315726,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 5.562474055624741e-05,
+      "loss": 0.6717,
+      "step": 3628
+    },
+    {
+      "epoch": 2.19136734077875,
+      "grad_norm": 0.125,
+      "learning_rate": 5.5583229555832296e-05,
+      "loss": 0.5286,
+      "step": 3629
+    },
+    {
+      "epoch": 2.191971023241775,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 5.554171855541719e-05,
+      "loss": 0.627,
+      "step": 3630
+    },
+    {
+      "epoch": 2.192574705704799,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 5.550020755500208e-05,
+      "loss": 0.5631,
+      "step": 3631
+    },
+    {
+      "epoch": 2.1931783881678237,
+      "grad_norm": 0.13671875,
+      "learning_rate": 5.545869655458696e-05,
+      "loss": 0.644,
+      "step": 3632
+    },
+    {
+      "epoch": 2.1937820706308484,
+      "grad_norm": 0.142578125,
+      "learning_rate": 5.5417185554171856e-05,
+      "loss": 0.6144,
+      "step": 3633
+    },
+    {
+      "epoch": 2.1943857530938726,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 5.537567455375675e-05,
+      "loss": 0.5612,
+      "step": 3634
+    },
+    {
+      "epoch": 2.194989435556897,
+      "grad_norm": 0.130859375,
+      "learning_rate": 5.5334163553341635e-05,
+      "loss": 0.8231,
+      "step": 3635
+    },
+    {
+      "epoch": 2.1955931180199215,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 5.529265255292653e-05,
+      "loss": 0.6544,
+      "step": 3636
+    },
+    {
+      "epoch": 2.196196800482946,
+      "grad_norm": 0.13671875,
+      "learning_rate": 5.525114155251142e-05,
+      "loss": 0.5665,
+      "step": 3637
+    },
+    {
+      "epoch": 2.1968004829459704,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 5.52096305520963e-05,
+      "loss": 0.6914,
+      "step": 3638
+    },
+    {
+      "epoch": 2.1974041654089946,
+      "grad_norm": 0.1484375,
+      "learning_rate": 5.5168119551681195e-05,
+      "loss": 0.6857,
+      "step": 3639
+    },
+    {
+      "epoch": 2.1980078478720193,
+      "grad_norm": 0.140625,
+      "learning_rate": 5.512660855126609e-05,
+      "loss": 0.5609,
+      "step": 3640
+    },
+    {
+      "epoch": 2.198611530335044,
+      "grad_norm": 0.15234375,
+      "learning_rate": 5.508509755085098e-05,
+      "loss": 0.5702,
+      "step": 3641
+    },
+    {
+      "epoch": 2.199215212798068,
+      "grad_norm": 0.14453125,
+      "learning_rate": 5.504358655043587e-05,
+      "loss": 0.6845,
+      "step": 3642
+    },
+    {
+      "epoch": 2.199818895261093,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 5.500207555002076e-05,
+      "loss": 0.8029,
+      "step": 3643
+    },
+    {
+      "epoch": 2.200422577724117,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 5.4960564549605654e-05,
+      "loss": 0.582,
+      "step": 3644
+    },
+    {
+      "epoch": 2.2010262601871418,
+      "grad_norm": 0.216796875,
+      "learning_rate": 5.4919053549190534e-05,
+      "loss": 0.6569,
+      "step": 3645
+    },
+    {
+      "epoch": 2.201629942650166,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 5.487754254877543e-05,
+      "loss": 0.5649,
+      "step": 3646
+    },
+    {
+      "epoch": 2.2022336251131907,
+      "grad_norm": 0.142578125,
+      "learning_rate": 5.483603154836032e-05,
+      "loss": 0.5068,
+      "step": 3647
+    },
+    {
+      "epoch": 2.202837307576215,
+      "grad_norm": 0.146484375,
+      "learning_rate": 5.479452054794521e-05,
+      "loss": 0.5033,
+      "step": 3648
+    },
+    {
+      "epoch": 2.2034409900392395,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 5.47530095475301e-05,
+      "loss": 0.5824,
+      "step": 3649
+    },
+    {
+      "epoch": 2.2040446725022638,
+      "grad_norm": 0.14453125,
+      "learning_rate": 5.4711498547114994e-05,
+      "loss": 0.5361,
+      "step": 3650
+    },
+    {
+      "epoch": 2.2046483549652884,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.466998754669987e-05,
+      "loss": 0.5451,
+      "step": 3651
+    },
+    {
+      "epoch": 2.2052520374283127,
+      "grad_norm": 0.1484375,
+      "learning_rate": 5.4628476546284766e-05,
+      "loss": 0.4929,
+      "step": 3652
+    },
+    {
+      "epoch": 2.2058557198913373,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.458696554586966e-05,
+      "loss": 0.5349,
+      "step": 3653
+    },
+    {
+      "epoch": 2.2064594023543616,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.4545454545454546e-05,
+      "loss": 0.5587,
+      "step": 3654
+    },
+    {
+      "epoch": 2.2070630848173862,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 5.450394354503944e-05,
+      "loss": 0.4678,
+      "step": 3655
+    },
+    {
+      "epoch": 2.2076667672804104,
+      "grad_norm": 0.181640625,
+      "learning_rate": 5.446243254462433e-05,
+      "loss": 0.4529,
+      "step": 3656
+    },
+    {
+      "epoch": 2.208270449743435,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 5.442092154420921e-05,
+      "loss": 0.4281,
+      "step": 3657
+    },
+    {
+      "epoch": 2.2088741322064593,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.4379410543794106e-05,
+      "loss": 0.3754,
+      "step": 3658
+    },
+    {
+      "epoch": 2.209477814669484,
+      "grad_norm": 0.3203125,
+      "learning_rate": 5.4337899543379e-05,
+      "loss": 0.3509,
+      "step": 3659
+    },
+    {
+      "epoch": 2.2100814971325082,
+      "grad_norm": 0.201171875,
+      "learning_rate": 5.4296388542963885e-05,
+      "loss": 0.368,
+      "step": 3660
+    },
+    {
+      "epoch": 2.210685179595533,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 5.425487754254878e-05,
+      "loss": 0.2797,
+      "step": 3661
+    },
+    {
+      "epoch": 2.211288862058557,
+      "grad_norm": 0.189453125,
+      "learning_rate": 5.421336654213367e-05,
+      "loss": 0.19,
+      "step": 3662
+    },
+    {
+      "epoch": 2.211892544521582,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 5.417185554171855e-05,
+      "loss": 1.0725,
+      "step": 3663
+    },
+    {
+      "epoch": 2.212496226984606,
+      "grad_norm": 0.234375,
+      "learning_rate": 5.4130344541303445e-05,
+      "loss": 0.655,
+      "step": 3664
+    },
+    {
+      "epoch": 2.2130999094476307,
+      "grad_norm": 0.154296875,
+      "learning_rate": 5.408883354088834e-05,
+      "loss": 0.693,
+      "step": 3665
+    },
+    {
+      "epoch": 2.213703591910655,
+      "grad_norm": 0.146484375,
+      "learning_rate": 5.404732254047323e-05,
+      "loss": 0.8072,
+      "step": 3666
+    },
+    {
+      "epoch": 2.2143072743736796,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 5.400581154005812e-05,
+      "loss": 0.686,
+      "step": 3667
+    },
+    {
+      "epoch": 2.214910956836704,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 5.396430053964301e-05,
+      "loss": 0.9237,
+      "step": 3668
+    },
+    {
+      "epoch": 2.2155146392997285,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 5.3922789539227904e-05,
+      "loss": 0.5754,
+      "step": 3669
+    },
+    {
+      "epoch": 2.2161183217627527,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 5.3881278538812784e-05,
+      "loss": 0.6025,
+      "step": 3670
+    },
+    {
+      "epoch": 2.2167220042257774,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 5.383976753839768e-05,
+      "loss": 0.7161,
+      "step": 3671
+    },
+    {
+      "epoch": 2.2173256866888016,
+      "grad_norm": 0.134765625,
+      "learning_rate": 5.379825653798257e-05,
+      "loss": 0.8643,
+      "step": 3672
+    },
+    {
+      "epoch": 2.2179293691518263,
+      "grad_norm": 0.142578125,
+      "learning_rate": 5.375674553756746e-05,
+      "loss": 0.7475,
+      "step": 3673
+    },
+    {
+      "epoch": 2.2185330516148505,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 5.371523453715235e-05,
+      "loss": 0.5257,
+      "step": 3674
+    },
+    {
+      "epoch": 2.219136734077875,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 5.3673723536737244e-05,
+      "loss": 0.6046,
+      "step": 3675
+    },
+    {
+      "epoch": 2.2197404165408994,
+      "grad_norm": 0.146484375,
+      "learning_rate": 5.3632212536322123e-05,
+      "loss": 0.5928,
+      "step": 3676
+    },
+    {
+      "epoch": 2.220344099003924,
+      "grad_norm": 0.154296875,
+      "learning_rate": 5.359070153590702e-05,
+      "loss": 0.7284,
+      "step": 3677
+    },
+    {
+      "epoch": 2.2209477814669483,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 5.354919053549191e-05,
+      "loss": 0.6325,
+      "step": 3678
+    },
+    {
+      "epoch": 2.221551463929973,
+      "grad_norm": 0.162109375,
+      "learning_rate": 5.3507679535076796e-05,
+      "loss": 0.7032,
+      "step": 3679
+    },
+    {
+      "epoch": 2.222155146392997,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 5.346616853466169e-05,
+      "loss": 0.5552,
+      "step": 3680
+    },
+    {
+      "epoch": 2.222758828856022,
+      "grad_norm": 0.154296875,
+      "learning_rate": 5.342465753424658e-05,
+      "loss": 0.6397,
+      "step": 3681
+    },
+    {
+      "epoch": 2.223362511319046,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 5.338314653383146e-05,
+      "loss": 0.6766,
+      "step": 3682
+    },
+    {
+      "epoch": 2.2239661937820707,
+      "grad_norm": 0.15234375,
+      "learning_rate": 5.3341635533416356e-05,
+      "loss": 0.6068,
+      "step": 3683
+    },
+    {
+      "epoch": 2.224569876245095,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 5.330012453300125e-05,
+      "loss": 0.7731,
+      "step": 3684
+    },
+    {
+      "epoch": 2.2251735587081196,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 5.3258613532586136e-05,
+      "loss": 0.6333,
+      "step": 3685
+    },
+    {
+      "epoch": 2.225777241171144,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 5.321710253217103e-05,
+      "loss": 0.6319,
+      "step": 3686
+    },
+    {
+      "epoch": 2.2263809236341685,
+      "grad_norm": 0.142578125,
+      "learning_rate": 5.317559153175592e-05,
+      "loss": 0.6213,
+      "step": 3687
+    },
+    {
+      "epoch": 2.2269846060971927,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 5.3134080531340815e-05,
+      "loss": 0.6248,
+      "step": 3688
+    },
+    {
+      "epoch": 2.2275882885602174,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 5.3092569530925695e-05,
+      "loss": 0.6109,
+      "step": 3689
+    },
+    {
+      "epoch": 2.2281919710232416,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 5.305105853051059e-05,
+      "loss": 0.8635,
+      "step": 3690
+    },
+    {
+      "epoch": 2.2287956534862663,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 5.300954753009548e-05,
+      "loss": 0.5466,
+      "step": 3691
+    },
+    {
+      "epoch": 2.2293993359492905,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.296803652968037e-05,
+      "loss": 0.7101,
+      "step": 3692
+    },
+    {
+      "epoch": 2.230003018412315,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 5.292652552926526e-05,
+      "loss": 0.6589,
+      "step": 3693
+    },
+    {
+      "epoch": 2.2306067008753394,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 5.2885014528850155e-05,
+      "loss": 0.6498,
+      "step": 3694
+    },
+    {
+      "epoch": 2.231210383338364,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 5.2843503528435034e-05,
+      "loss": 0.5665,
+      "step": 3695
+    },
+    {
+      "epoch": 2.2318140658013883,
+      "grad_norm": 0.146484375,
+      "learning_rate": 5.280199252801993e-05,
+      "loss": 0.5706,
+      "step": 3696
+    },
+    {
+      "epoch": 2.232417748264413,
+      "grad_norm": 0.140625,
+      "learning_rate": 5.276048152760482e-05,
+      "loss": 0.55,
+      "step": 3697
+    },
+    {
+      "epoch": 2.233021430727437,
+      "grad_norm": 0.162109375,
+      "learning_rate": 5.27189705271897e-05,
+      "loss": 0.5897,
+      "step": 3698
+    },
+    {
+      "epoch": 2.233625113190462,
+      "grad_norm": 0.193359375,
+      "learning_rate": 5.26774595267746e-05,
+      "loss": 0.6028,
+      "step": 3699
+    },
+    {
+      "epoch": 2.234228795653486,
+      "grad_norm": 0.162109375,
+      "learning_rate": 5.2635948526359494e-05,
+      "loss": 0.5925,
+      "step": 3700
+    },
+    {
+      "epoch": 2.2348324781165108,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 5.2594437525944374e-05,
+      "loss": 0.5198,
+      "step": 3701
+    },
+    {
+      "epoch": 2.235436160579535,
+      "grad_norm": 0.158203125,
+      "learning_rate": 5.255292652552927e-05,
+      "loss": 0.5255,
+      "step": 3702
+    },
+    {
+      "epoch": 2.2360398430425596,
+      "grad_norm": 0.16015625,
+      "learning_rate": 5.251141552511416e-05,
+      "loss": 0.5007,
+      "step": 3703
+    },
+    {
+      "epoch": 2.236643525505584,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.246990452469904e-05,
+      "loss": 0.5518,
+      "step": 3704
+    },
+    {
+      "epoch": 2.2372472079686085,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.242839352428393e-05,
+      "loss": 0.487,
+      "step": 3705
+    },
+    {
+      "epoch": 2.2378508904316328,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 5.238688252386883e-05,
+      "loss": 0.4804,
+      "step": 3706
+    },
+    {
+      "epoch": 2.2384545728946574,
+      "grad_norm": 0.181640625,
+      "learning_rate": 5.234537152345371e-05,
+      "loss": 0.4207,
+      "step": 3707
+    },
+    {
+      "epoch": 2.2390582553576817,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.2303860523038606e-05,
+      "loss": 0.391,
+      "step": 3708
+    },
+    {
+      "epoch": 2.2396619378207063,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 5.22623495226235e-05,
+      "loss": 0.3689,
+      "step": 3709
+    },
+    {
+      "epoch": 2.2402656202837306,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 5.222083852220839e-05,
+      "loss": 0.34,
+      "step": 3710
+    },
+    {
+      "epoch": 2.240869302746755,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.217932752179327e-05,
+      "loss": 0.2817,
+      "step": 3711
+    },
+    {
+      "epoch": 2.2414729852097794,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 5.2137816521378166e-05,
+      "loss": 0.2022,
+      "step": 3712
+    },
+    {
+      "epoch": 2.242076667672804,
+      "grad_norm": 0.140625,
+      "learning_rate": 5.2096305520963066e-05,
+      "loss": 0.5713,
+      "step": 3713
+    },
+    {
+      "epoch": 2.242680350135829,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 5.2054794520547945e-05,
+      "loss": 0.5972,
+      "step": 3714
+    },
+    {
+      "epoch": 2.243284032598853,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 5.201328352013284e-05,
+      "loss": 1.3534,
+      "step": 3715
+    },
+    {
+      "epoch": 2.2438877150618772,
+      "grad_norm": 0.13671875,
+      "learning_rate": 5.197177251971773e-05,
+      "loss": 0.544,
+      "step": 3716
+    },
+    {
+      "epoch": 2.244491397524902,
+      "grad_norm": 0.14453125,
+      "learning_rate": 5.193026151930261e-05,
+      "loss": 0.7116,
+      "step": 3717
+    },
+    {
+      "epoch": 2.2450950799879266,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 5.1888750518887505e-05,
+      "loss": 0.4888,
+      "step": 3718
+    },
+    {
+      "epoch": 2.245698762450951,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 5.18472395184724e-05,
+      "loss": 0.6472,
+      "step": 3719
+    },
+    {
+      "epoch": 2.246302444913975,
+      "grad_norm": 0.181640625,
+      "learning_rate": 5.1805728518057285e-05,
+      "loss": 0.7131,
+      "step": 3720
+    },
+    {
+      "epoch": 2.2469061273769997,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 5.176421751764218e-05,
+      "loss": 0.6801,
+      "step": 3721
+    },
+    {
+      "epoch": 2.2475098098400244,
+      "grad_norm": 0.146484375,
+      "learning_rate": 5.172270651722707e-05,
+      "loss": 0.5962,
+      "step": 3722
+    },
+    {
+      "epoch": 2.2481134923030486,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 5.168119551681195e-05,
+      "loss": 0.6611,
+      "step": 3723
+    },
+    {
+      "epoch": 2.2487171747660732,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 5.1639684516396844e-05,
+      "loss": 0.6505,
+      "step": 3724
+    },
+    {
+      "epoch": 2.2493208572290975,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 5.159817351598174e-05,
+      "loss": 0.5929,
+      "step": 3725
+    },
+    {
+      "epoch": 2.249924539692122,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 5.1556662515566624e-05,
+      "loss": 0.6366,
+      "step": 3726
+    },
+    {
+      "epoch": 2.2505282221551464,
+      "grad_norm": 0.158203125,
+      "learning_rate": 5.151515151515152e-05,
+      "loss": 0.7009,
+      "step": 3727
+    },
+    {
+      "epoch": 2.251131904618171,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 5.147364051473641e-05,
+      "loss": 0.6801,
+      "step": 3728
+    },
+    {
+      "epoch": 2.2517355870811953,
+      "grad_norm": 0.146484375,
+      "learning_rate": 5.143212951432129e-05,
+      "loss": 0.7714,
+      "step": 3729
+    },
+    {
+      "epoch": 2.25233926954422,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 5.139061851390618e-05,
+      "loss": 0.5927,
+      "step": 3730
+    },
+    {
+      "epoch": 2.252942952007244,
+      "grad_norm": 0.1484375,
+      "learning_rate": 5.1349107513491076e-05,
+      "loss": 0.581,
+      "step": 3731
+    },
+    {
+      "epoch": 2.253546634470269,
+      "grad_norm": 0.1484375,
+      "learning_rate": 5.130759651307597e-05,
+      "loss": 0.6646,
+      "step": 3732
+    },
+    {
+      "epoch": 2.254150316933293,
+      "grad_norm": 0.138671875,
+      "learning_rate": 5.1266085512660856e-05,
+      "loss": 0.6749,
+      "step": 3733
+    },
+    {
+      "epoch": 2.2547539993963177,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 5.122457451224575e-05,
+      "loss": 0.7835,
+      "step": 3734
+    },
+    {
+      "epoch": 2.255357681859342,
+      "grad_norm": 0.13671875,
+      "learning_rate": 5.118306351183064e-05,
+      "loss": 0.5993,
+      "step": 3735
+    },
+    {
+      "epoch": 2.2559613643223666,
+      "grad_norm": 0.146484375,
+      "learning_rate": 5.114155251141552e-05,
+      "loss": 0.8175,
+      "step": 3736
+    },
+    {
+      "epoch": 2.256565046785391,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.1100041511000416e-05,
+      "loss": 0.5801,
+      "step": 3737
+    },
+    {
+      "epoch": 2.2571687292484155,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 5.105853051058531e-05,
+      "loss": 0.6434,
+      "step": 3738
+    },
+    {
+      "epoch": 2.2577724117114397,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.1017019510170195e-05,
+      "loss": 0.6335,
+      "step": 3739
+    },
+    {
+      "epoch": 2.2583760941744644,
+      "grad_norm": 0.16015625,
+      "learning_rate": 5.097550850975509e-05,
+      "loss": 0.8821,
+      "step": 3740
+    },
+    {
+      "epoch": 2.2589797766374886,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 5.093399750933998e-05,
+      "loss": 0.6041,
+      "step": 3741
+    },
+    {
+      "epoch": 2.2595834591005133,
+      "grad_norm": 0.19140625,
+      "learning_rate": 5.089248650892486e-05,
+      "loss": 0.8298,
+      "step": 3742
+    },
+    {
+      "epoch": 2.2601871415635375,
+      "grad_norm": 0.142578125,
+      "learning_rate": 5.0850975508509755e-05,
+      "loss": 0.7614,
+      "step": 3743
+    },
+    {
+      "epoch": 2.260790824026562,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.080946450809465e-05,
+      "loss": 0.508,
+      "step": 3744
+    },
+    {
+      "epoch": 2.2613945064895864,
+      "grad_norm": 0.275390625,
+      "learning_rate": 5.0767953507679535e-05,
+      "loss": 0.7969,
+      "step": 3745
+    },
+    {
+      "epoch": 2.261998188952611,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 5.072644250726443e-05,
+      "loss": 0.4903,
+      "step": 3746
+    },
+    {
+      "epoch": 2.2626018714156353,
+      "grad_norm": 0.1484375,
+      "learning_rate": 5.068493150684932e-05,
+      "loss": 0.5706,
+      "step": 3747
+    },
+    {
+      "epoch": 2.26320555387866,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 5.06434205064342e-05,
+      "loss": 0.5735,
+      "step": 3748
+    },
+    {
+      "epoch": 2.263809236341684,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 5.0601909506019094e-05,
+      "loss": 0.568,
+      "step": 3749
+    },
+    {
+      "epoch": 2.264412918804709,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.056039850560399e-05,
+      "loss": 0.5466,
+      "step": 3750
+    },
+    {
+      "epoch": 2.265016601267733,
+      "grad_norm": 0.154296875,
+      "learning_rate": 5.0518887505188874e-05,
+      "loss": 0.5017,
+      "step": 3751
+    },
+    {
+      "epoch": 2.2656202837307577,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 5.047737650477377e-05,
+      "loss": 0.4916,
+      "step": 3752
+    },
+    {
+      "epoch": 2.266223966193782,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 5.043586550435866e-05,
+      "loss": 0.5218,
+      "step": 3753
+    },
+    {
+      "epoch": 2.2668276486568066,
+      "grad_norm": 0.1640625,
+      "learning_rate": 5.0394354503943554e-05,
+      "loss": 0.4766,
+      "step": 3754
+    },
+    {
+      "epoch": 2.267431331119831,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 5.0352843503528433e-05,
+      "loss": 0.5308,
+      "step": 3755
+    },
+    {
+      "epoch": 2.2680350135828555,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.031133250311333e-05,
+      "loss": 0.4462,
+      "step": 3756
+    },
+    {
+      "epoch": 2.2686386960458798,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 5.026982150269822e-05,
+      "loss": 0.3928,
+      "step": 3757
+    },
+    {
+      "epoch": 2.2692423785089044,
+      "grad_norm": 0.193359375,
+      "learning_rate": 5.0228310502283106e-05,
+      "loss": 0.3917,
+      "step": 3758
+    },
+    {
+      "epoch": 2.2698460609719286,
+      "grad_norm": 0.1875,
+      "learning_rate": 5.0186799501868e-05,
+      "loss": 0.349,
+      "step": 3759
+    },
+    {
+      "epoch": 2.2704497434349533,
+      "grad_norm": 0.203125,
+      "learning_rate": 5.014528850145289e-05,
+      "loss": 0.3521,
+      "step": 3760
+    },
+    {
+      "epoch": 2.2710534258979775,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 5.010377750103777e-05,
+      "loss": 0.2588,
+      "step": 3761
+    },
+    {
+      "epoch": 2.271657108361002,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.0062266500622666e-05,
+      "loss": 0.1769,
+      "step": 3762
+    },
+    {
+      "epoch": 2.2722607908240264,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.002075550020756e-05,
+      "loss": 0.6932,
+      "step": 3763
+    },
+    {
+      "epoch": 2.272864473287051,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 4.997924449979245e-05,
+      "loss": 0.6689,
+      "step": 3764
+    },
+    {
+      "epoch": 2.2734681557500753,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 4.993773349937734e-05,
+      "loss": 1.0518,
+      "step": 3765
+    },
+    {
+      "epoch": 2.2740718382131,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 4.9896222498962225e-05,
+      "loss": 0.6322,
+      "step": 3766
+    },
+    {
+      "epoch": 2.274675520676124,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 4.985471149854712e-05,
+      "loss": 0.6422,
+      "step": 3767
+    },
+    {
+      "epoch": 2.275279203139149,
+      "grad_norm": 0.1484375,
+      "learning_rate": 4.9813200498132005e-05,
+      "loss": 0.6127,
+      "step": 3768
+    },
+    {
+      "epoch": 2.275882885602173,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 4.977168949771689e-05,
+      "loss": 0.6145,
+      "step": 3769
+    },
+    {
+      "epoch": 2.276486568065198,
+      "grad_norm": 0.146484375,
+      "learning_rate": 4.9730178497301785e-05,
+      "loss": 0.5599,
+      "step": 3770
+    },
+    {
+      "epoch": 2.277090250528222,
+      "grad_norm": 0.142578125,
+      "learning_rate": 4.968866749688668e-05,
+      "loss": 0.8418,
+      "step": 3771
+    },
+    {
+      "epoch": 2.2776939329912467,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 4.964715649647157e-05,
+      "loss": 0.562,
+      "step": 3772
+    },
+    {
+      "epoch": 2.278297615454271,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 4.960564549605646e-05,
+      "loss": 0.6149,
+      "step": 3773
+    },
+    {
+      "epoch": 2.2789012979172956,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 4.9564134495641344e-05,
+      "loss": 0.6435,
+      "step": 3774
+    },
+    {
+      "epoch": 2.27950498038032,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 4.952262349522624e-05,
+      "loss": 0.7169,
+      "step": 3775
+    },
+    {
+      "epoch": 2.2801086628433445,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 4.9481112494811124e-05,
+      "loss": 0.9838,
+      "step": 3776
+    },
+    {
+      "epoch": 2.2807123453063687,
+      "grad_norm": 0.150390625,
+      "learning_rate": 4.943960149439602e-05,
+      "loss": 0.6183,
+      "step": 3777
+    },
+    {
+      "epoch": 2.2813160277693934,
+      "grad_norm": 0.138671875,
+      "learning_rate": 4.939809049398091e-05,
+      "loss": 0.5653,
+      "step": 3778
+    },
+    {
+      "epoch": 2.2819197102324176,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 4.93565794935658e-05,
+      "loss": 0.8278,
+      "step": 3779
+    },
+    {
+      "epoch": 2.2825233926954422,
+      "grad_norm": 0.142578125,
+      "learning_rate": 4.9315068493150684e-05,
+      "loss": 0.6817,
+      "step": 3780
+    },
+    {
+      "epoch": 2.2831270751584665,
+      "grad_norm": 0.154296875,
+      "learning_rate": 4.927355749273558e-05,
+      "loss": 0.548,
+      "step": 3781
+    },
+    {
+      "epoch": 2.283730757621491,
+      "grad_norm": 0.142578125,
+      "learning_rate": 4.923204649232046e-05,
+      "loss": 0.6265,
+      "step": 3782
+    },
+    {
+      "epoch": 2.2843344400845154,
+      "grad_norm": 0.1484375,
+      "learning_rate": 4.9190535491905357e-05,
+      "loss": 0.5528,
+      "step": 3783
+    },
+    {
+      "epoch": 2.28493812254754,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 4.914902449149025e-05,
+      "loss": 0.9304,
+      "step": 3784
+    },
+    {
+      "epoch": 2.2855418050105643,
+      "grad_norm": 0.154296875,
+      "learning_rate": 4.9107513491075136e-05,
+      "loss": 0.6049,
+      "step": 3785
+    },
+    {
+      "epoch": 2.286145487473589,
+      "grad_norm": 0.140625,
+      "learning_rate": 4.906600249066003e-05,
+      "loss": 0.6067,
+      "step": 3786
+    },
+    {
+      "epoch": 2.2867491699366136,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 4.9024491490244916e-05,
+      "loss": 0.5666,
+      "step": 3787
+    },
+    {
+      "epoch": 2.287352852399638,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 4.89829804898298e-05,
+      "loss": 0.6094,
+      "step": 3788
+    },
+    {
+      "epoch": 2.287956534862662,
+      "grad_norm": 0.140625,
+      "learning_rate": 4.8941469489414696e-05,
+      "loss": 0.4883,
+      "step": 3789
+    },
+    {
+      "epoch": 2.2885602173256867,
+      "grad_norm": 0.140625,
+      "learning_rate": 4.889995848899959e-05,
+      "loss": 0.5744,
+      "step": 3790
+    },
+    {
+      "epoch": 2.2891638997887114,
+      "grad_norm": 0.134765625,
+      "learning_rate": 4.8858447488584476e-05,
+      "loss": 0.6248,
+      "step": 3791
+    },
+    {
+      "epoch": 2.2897675822517356,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 4.881693648816937e-05,
+      "loss": 0.8144,
+      "step": 3792
+    },
+    {
+      "epoch": 2.29037126471476,
+      "grad_norm": 0.14453125,
+      "learning_rate": 4.8775425487754255e-05,
+      "loss": 0.6317,
+      "step": 3793
+    },
+    {
+      "epoch": 2.2909749471777845,
+      "grad_norm": 0.15234375,
+      "learning_rate": 4.873391448733915e-05,
+      "loss": 0.6998,
+      "step": 3794
+    },
+    {
+      "epoch": 2.291578629640809,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 4.8692403486924035e-05,
+      "loss": 0.5846,
+      "step": 3795
+    },
+    {
+      "epoch": 2.2921823121038334,
+      "grad_norm": 0.13671875,
+      "learning_rate": 4.865089248650893e-05,
+      "loss": 0.5276,
+      "step": 3796
+    },
+    {
+      "epoch": 2.2927859945668576,
+      "grad_norm": 0.16015625,
+      "learning_rate": 4.860938148609382e-05,
+      "loss": 0.5752,
+      "step": 3797
+    },
+    {
+      "epoch": 2.2933896770298823,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 4.856787048567871e-05,
+      "loss": 0.5444,
+      "step": 3798
+    },
+    {
+      "epoch": 2.293993359492907,
+      "grad_norm": 0.140625,
+      "learning_rate": 4.8526359485263595e-05,
+      "loss": 0.5088,
+      "step": 3799
+    },
+    {
+      "epoch": 2.294597041955931,
+      "grad_norm": 0.16015625,
+      "learning_rate": 4.848484848484849e-05,
+      "loss": 0.5237,
+      "step": 3800
+    },
+    {
+      "epoch": 2.2952007244189554,
+      "grad_norm": 0.162109375,
+      "learning_rate": 4.8443337484433374e-05,
+      "loss": 0.5165,
+      "step": 3801
+    },
+    {
+      "epoch": 2.29580440688198,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.840182648401827e-05,
+      "loss": 0.5137,
+      "step": 3802
+    },
+    {
+      "epoch": 2.2964080893450047,
+      "grad_norm": 0.16015625,
+      "learning_rate": 4.836031548360316e-05,
+      "loss": 0.5314,
+      "step": 3803
+    },
+    {
+      "epoch": 2.297011771808029,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.831880448318805e-05,
+      "loss": 0.5597,
+      "step": 3804
+    },
+    {
+      "epoch": 2.297615454271053,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.827729348277294e-05,
+      "loss": 0.492,
+      "step": 3805
+    },
+    {
+      "epoch": 2.298219136734078,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 4.823578248235783e-05,
+      "loss": 0.4438,
+      "step": 3806
+    },
+    {
+      "epoch": 2.2988228191971025,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.8194271481942714e-05,
+      "loss": 0.3985,
+      "step": 3807
+    },
+    {
+      "epoch": 2.2994265016601267,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.815276048152761e-05,
+      "loss": 0.513,
+      "step": 3808
+    },
+    {
+      "epoch": 2.3000301841231514,
+      "grad_norm": 0.2109375,
+      "learning_rate": 4.81112494811125e-05,
+      "loss": 0.3765,
+      "step": 3809
+    },
+    {
+      "epoch": 2.3006338665861756,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 4.8069738480697387e-05,
+      "loss": 0.2697,
+      "step": 3810
+    },
+    {
+      "epoch": 2.3012375490492003,
+      "grad_norm": 0.185546875,
+      "learning_rate": 4.802822748028228e-05,
+      "loss": 0.2245,
+      "step": 3811
+    },
+    {
+      "epoch": 2.3018412315122245,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 4.7986716479867166e-05,
+      "loss": 0.1918,
+      "step": 3812
+    },
+    {
+      "epoch": 2.302444913975249,
+      "grad_norm": 0.1484375,
+      "learning_rate": 4.794520547945205e-05,
+      "loss": 0.7845,
+      "step": 3813
+    },
+    {
+      "epoch": 2.3030485964382734,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 4.7903694479036946e-05,
+      "loss": 0.5505,
+      "step": 3814
+    },
+    {
+      "epoch": 2.303652278901298,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 4.786218347862184e-05,
+      "loss": 0.601,
+      "step": 3815
+    },
+    {
+      "epoch": 2.3042559613643223,
+      "grad_norm": 0.15234375,
+      "learning_rate": 4.782067247820673e-05,
+      "loss": 0.5139,
+      "step": 3816
+    },
+    {
+      "epoch": 2.304859643827347,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 4.777916147779162e-05,
+      "loss": 0.6804,
+      "step": 3817
+    },
+    {
+      "epoch": 2.305463326290371,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 4.7737650477376505e-05,
+      "loss": 0.5875,
+      "step": 3818
+    },
+    {
+      "epoch": 2.306067008753396,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.76961394769614e-05,
+      "loss": 0.6261,
+      "step": 3819
+    },
+    {
+      "epoch": 2.30667069121642,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 4.7654628476546285e-05,
+      "loss": 0.597,
+      "step": 3820
+    },
+    {
+      "epoch": 2.3072743736794448,
+      "grad_norm": 0.14453125,
+      "learning_rate": 4.761311747613117e-05,
+      "loss": 0.4979,
+      "step": 3821
+    },
+    {
+      "epoch": 2.307878056142469,
+      "grad_norm": 0.13671875,
+      "learning_rate": 4.757160647571607e-05,
+      "loss": 0.5344,
+      "step": 3822
+    },
+    {
+      "epoch": 2.3084817386054937,
+      "grad_norm": 0.154296875,
+      "learning_rate": 4.753009547530096e-05,
+      "loss": 0.575,
+      "step": 3823
+    },
+    {
+      "epoch": 2.309085421068518,
+      "grad_norm": 0.154296875,
+      "learning_rate": 4.7488584474885845e-05,
+      "loss": 0.6303,
+      "step": 3824
+    },
+    {
+      "epoch": 2.3096891035315426,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 4.744707347447074e-05,
+      "loss": 0.65,
+      "step": 3825
+    },
+    {
+      "epoch": 2.3102927859945668,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 4.7405562474055624e-05,
+      "loss": 0.6018,
+      "step": 3826
+    },
+    {
+      "epoch": 2.3108964684575914,
+      "grad_norm": 0.16015625,
+      "learning_rate": 4.736405147364052e-05,
+      "loss": 0.6526,
+      "step": 3827
+    },
+    {
+      "epoch": 2.3115001509206157,
+      "grad_norm": 0.140625,
+      "learning_rate": 4.7322540473225404e-05,
+      "loss": 0.5563,
+      "step": 3828
+    },
+    {
+      "epoch": 2.3121038333836403,
+      "grad_norm": 0.13671875,
+      "learning_rate": 4.72810294728103e-05,
+      "loss": 0.6763,
+      "step": 3829
+    },
+    {
+      "epoch": 2.3127075158466646,
+      "grad_norm": 0.1484375,
+      "learning_rate": 4.723951847239519e-05,
+      "loss": 0.5513,
+      "step": 3830
+    },
+    {
+      "epoch": 2.3133111983096892,
+      "grad_norm": 0.146484375,
+      "learning_rate": 4.719800747198008e-05,
+      "loss": 0.7,
+      "step": 3831
+    },
+    {
+      "epoch": 2.3139148807727135,
+      "grad_norm": 0.140625,
+      "learning_rate": 4.7156496471564964e-05,
+      "loss": 0.6488,
+      "step": 3832
+    },
+    {
+      "epoch": 2.314518563235738,
+      "grad_norm": 0.15625,
+      "learning_rate": 4.711498547114986e-05,
+      "loss": 0.6514,
+      "step": 3833
+    },
+    {
+      "epoch": 2.3151222456987623,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 4.7073474470734743e-05,
+      "loss": 0.5488,
+      "step": 3834
+    },
+    {
+      "epoch": 2.315725928161787,
+      "grad_norm": 0.146484375,
+      "learning_rate": 4.703196347031964e-05,
+      "loss": 0.6489,
+      "step": 3835
+    },
+    {
+      "epoch": 2.3163296106248112,
+      "grad_norm": 0.146484375,
+      "learning_rate": 4.699045246990453e-05,
+      "loss": 0.5646,
+      "step": 3836
+    },
+    {
+      "epoch": 2.316933293087836,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 4.6948941469489416e-05,
+      "loss": 0.6441,
+      "step": 3837
+    },
+    {
+      "epoch": 2.31753697555086,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 4.690743046907431e-05,
+      "loss": 0.665,
+      "step": 3838
+    },
+    {
+      "epoch": 2.318140658013885,
+      "grad_norm": 0.12890625,
+      "learning_rate": 4.6865919468659196e-05,
+      "loss": 0.5684,
+      "step": 3839
+    },
+    {
+      "epoch": 2.318744340476909,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 4.682440846824408e-05,
+      "loss": 0.5394,
+      "step": 3840
+    },
+    {
+      "epoch": 2.3193480229399337,
+      "grad_norm": 0.154296875,
+      "learning_rate": 4.6782897467828976e-05,
+      "loss": 0.5669,
+      "step": 3841
+    },
+    {
+      "epoch": 2.319951705402958,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 4.674138646741387e-05,
+      "loss": 0.6004,
+      "step": 3842
+    },
+    {
+      "epoch": 2.3205553878659826,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 4.6699875466998756e-05,
+      "loss": 0.6637,
+      "step": 3843
+    },
+    {
+      "epoch": 2.321159070329007,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 4.665836446658365e-05,
+      "loss": 0.5828,
+      "step": 3844
+    },
+    {
+      "epoch": 2.3217627527920315,
+      "grad_norm": 0.13671875,
+      "learning_rate": 4.6616853466168535e-05,
+      "loss": 0.5728,
+      "step": 3845
+    },
+    {
+      "epoch": 2.3223664352550557,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 4.657534246575342e-05,
+      "loss": 0.5247,
+      "step": 3846
+    },
+    {
+      "epoch": 2.3229701177180804,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 4.6533831465338315e-05,
+      "loss": 0.5778,
+      "step": 3847
+    },
+    {
+      "epoch": 2.3235738001811046,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 4.649232046492321e-05,
+      "loss": 0.4809,
+      "step": 3848
+    },
+    {
+      "epoch": 2.3241774826441293,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 4.64508094645081e-05,
+      "loss": 0.5791,
+      "step": 3849
+    },
+    {
+      "epoch": 2.3247811651071535,
+      "grad_norm": 0.154296875,
+      "learning_rate": 4.640929846409299e-05,
+      "loss": 0.4888,
+      "step": 3850
+    },
+    {
+      "epoch": 2.325384847570178,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 4.6367787463677875e-05,
+      "loss": 0.5856,
+      "step": 3851
+    },
+    {
+      "epoch": 2.3259885300332024,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 4.632627646326277e-05,
+      "loss": 0.5226,
+      "step": 3852
+    },
+    {
+      "epoch": 2.326592212496227,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.6284765462847654e-05,
+      "loss": 0.494,
+      "step": 3853
+    },
+    {
+      "epoch": 2.3271958949592513,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.624325446243255e-05,
+      "loss": 0.4667,
+      "step": 3854
+    },
+    {
+      "epoch": 2.327799577422276,
+      "grad_norm": 0.177734375,
+      "learning_rate": 4.620174346201744e-05,
+      "loss": 0.4776,
+      "step": 3855
+    },
+    {
+      "epoch": 2.3284032598853,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.616023246160233e-05,
+      "loss": 0.4472,
+      "step": 3856
+    },
+    {
+      "epoch": 2.329006942348325,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.6118721461187214e-05,
+      "loss": 0.3986,
+      "step": 3857
+    },
+    {
+      "epoch": 2.329610624811349,
+      "grad_norm": 0.19140625,
+      "learning_rate": 4.607721046077211e-05,
+      "loss": 0.3597,
+      "step": 3858
+    },
+    {
+      "epoch": 2.3302143072743737,
+      "grad_norm": 0.201171875,
+      "learning_rate": 4.6035699460356994e-05,
+      "loss": 0.3309,
+      "step": 3859
+    },
+    {
+      "epoch": 2.330817989737398,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 4.599418845994189e-05,
+      "loss": 0.2997,
+      "step": 3860
+    },
+    {
+      "epoch": 2.3314216722004226,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 4.595267745952678e-05,
+      "loss": 0.2452,
+      "step": 3861
+    },
+    {
+      "epoch": 2.332025354663447,
+      "grad_norm": 0.21484375,
+      "learning_rate": 4.5911166459111667e-05,
+      "loss": 0.1978,
+      "step": 3862
+    },
+    {
+      "epoch": 2.3326290371264715,
+      "grad_norm": 0.14453125,
+      "learning_rate": 4.586965545869656e-05,
+      "loss": 0.6661,
+      "step": 3863
+    },
+    {
+      "epoch": 2.3332327195894957,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 4.5828144458281446e-05,
+      "loss": 0.5825,
+      "step": 3864
+    },
+    {
+      "epoch": 2.3338364020525204,
+      "grad_norm": 0.158203125,
+      "learning_rate": 4.578663345786633e-05,
+      "loss": 0.6307,
+      "step": 3865
+    },
+    {
+      "epoch": 2.3344400845155446,
+      "grad_norm": 0.36328125,
+      "learning_rate": 4.5745122457451226e-05,
+      "loss": 0.555,
+      "step": 3866
+    },
+    {
+      "epoch": 2.3350437669785693,
+      "grad_norm": 0.150390625,
+      "learning_rate": 4.570361145703612e-05,
+      "loss": 0.5548,
+      "step": 3867
+    },
+    {
+      "epoch": 2.335647449441594,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 4.5662100456621006e-05,
+      "loss": 0.7686,
+      "step": 3868
+    },
+    {
+      "epoch": 2.336251131904618,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 4.56205894562059e-05,
+      "loss": 0.8087,
+      "step": 3869
+    },
+    {
+      "epoch": 2.3368548143676424,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 4.5579078455790786e-05,
+      "loss": 0.7414,
+      "step": 3870
+    },
+    {
+      "epoch": 2.337458496830667,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 4.553756745537568e-05,
+      "loss": 0.5163,
+      "step": 3871
+    },
+    {
+      "epoch": 2.3380621792936918,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 4.5496056454960565e-05,
+      "loss": 0.6429,
+      "step": 3872
+    },
+    {
+      "epoch": 2.338665861756716,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 0.6131,
+      "step": 3873
+    },
+    {
+      "epoch": 2.33926954421974,
+      "grad_norm": 0.13671875,
+      "learning_rate": 4.541303445413035e-05,
+      "loss": 0.546,
+      "step": 3874
+    },
+    {
+      "epoch": 2.339873226682765,
+      "grad_norm": 0.1484375,
+      "learning_rate": 4.537152345371524e-05,
+      "loss": 0.7007,
+      "step": 3875
+    },
+    {
+      "epoch": 2.3404769091457895,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 4.5330012453300125e-05,
+      "loss": 0.503,
+      "step": 3876
+    },
+    {
+      "epoch": 2.3410805916088138,
+      "grad_norm": 0.15625,
+      "learning_rate": 4.528850145288502e-05,
+      "loss": 0.6473,
+      "step": 3877
+    },
+    {
+      "epoch": 2.341684274071838,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 4.5246990452469905e-05,
+      "loss": 0.6571,
+      "step": 3878
+    },
+    {
+      "epoch": 2.3422879565348627,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 4.520547945205479e-05,
+      "loss": 0.5842,
+      "step": 3879
+    },
+    {
+      "epoch": 2.3428916389978873,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 4.516396845163969e-05,
+      "loss": 0.5991,
+      "step": 3880
+    },
+    {
+      "epoch": 2.3434953214609116,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 4.512245745122458e-05,
+      "loss": 0.6484,
+      "step": 3881
+    },
+    {
+      "epoch": 2.3440990039239358,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 4.508094645080947e-05,
+      "loss": 0.603,
+      "step": 3882
+    },
+    {
+      "epoch": 2.3447026863869604,
+      "grad_norm": 0.146484375,
+      "learning_rate": 4.503943545039436e-05,
+      "loss": 0.6013,
+      "step": 3883
+    },
+    {
+      "epoch": 2.345306368849985,
+      "grad_norm": 0.150390625,
+      "learning_rate": 4.4997924449979244e-05,
+      "loss": 0.6968,
+      "step": 3884
+    },
+    {
+      "epoch": 2.3459100513130093,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 4.495641344956414e-05,
+      "loss": 0.7278,
+      "step": 3885
+    },
+    {
+      "epoch": 2.3465137337760336,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 4.4914902449149024e-05,
+      "loss": 0.5987,
+      "step": 3886
+    },
+    {
+      "epoch": 2.3471174162390582,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 4.487339144873392e-05,
+      "loss": 0.6939,
+      "step": 3887
+    },
+    {
+      "epoch": 2.347721098702083,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 4.483188044831881e-05,
+      "loss": 0.6199,
+      "step": 3888
+    },
+    {
+      "epoch": 2.348324781165107,
+      "grad_norm": 0.16015625,
+      "learning_rate": 4.4790369447903697e-05,
+      "loss": 0.653,
+      "step": 3889
+    },
+    {
+      "epoch": 2.348928463628132,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 4.474885844748858e-05,
+      "loss": 0.5988,
+      "step": 3890
+    },
+    {
+      "epoch": 2.349532146091156,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 4.4707347447073476e-05,
+      "loss": 0.603,
+      "step": 3891
+    },
+    {
+      "epoch": 2.3501358285541807,
+      "grad_norm": 0.142578125,
+      "learning_rate": 4.466583644665836e-05,
+      "loss": 0.5959,
+      "step": 3892
+    },
+    {
+      "epoch": 2.350739511017205,
+      "grad_norm": 0.13671875,
+      "learning_rate": 4.4624325446243256e-05,
+      "loss": 0.6052,
+      "step": 3893
+    },
+    {
+      "epoch": 2.3513431934802296,
+      "grad_norm": 0.1484375,
+      "learning_rate": 4.458281444582815e-05,
+      "loss": 1.0572,
+      "step": 3894
+    },
+    {
+      "epoch": 2.351946875943254,
+      "grad_norm": 0.1640625,
+      "learning_rate": 4.4541303445413036e-05,
+      "loss": 0.6273,
+      "step": 3895
+    },
+    {
+      "epoch": 2.3525505584062785,
+      "grad_norm": 0.15625,
+      "learning_rate": 4.449979244499793e-05,
+      "loss": 0.6219,
+      "step": 3896
+    },
+    {
+      "epoch": 2.3531542408693027,
+      "grad_norm": 0.142578125,
+      "learning_rate": 4.4458281444582815e-05,
+      "loss": 0.5061,
+      "step": 3897
+    },
+    {
+      "epoch": 2.3537579233323274,
+      "grad_norm": 0.150390625,
+      "learning_rate": 4.44167704441677e-05,
+      "loss": 0.5157,
+      "step": 3898
+    },
+    {
+      "epoch": 2.3543616057953516,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 4.4375259443752595e-05,
+      "loss": 0.5173,
+      "step": 3899
+    },
+    {
+      "epoch": 2.3549652882583763,
+      "grad_norm": 0.15234375,
+      "learning_rate": 4.433374844333749e-05,
+      "loss": 0.5796,
+      "step": 3900
+    },
+    {
+      "epoch": 2.3555689707214005,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 4.4292237442922375e-05,
+      "loss": 0.5505,
+      "step": 3901
+    },
+    {
+      "epoch": 2.356172653184425,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.425072644250727e-05,
+      "loss": 0.5425,
+      "step": 3902
+    },
+    {
+      "epoch": 2.3567763356474494,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 4.4209215442092155e-05,
+      "loss": 0.4759,
+      "step": 3903
+    },
+    {
+      "epoch": 2.357380018110474,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.416770444167705e-05,
+      "loss": 0.5116,
+      "step": 3904
+    },
+    {
+      "epoch": 2.3579837005734983,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.4126193441261934e-05,
+      "loss": 0.3866,
+      "step": 3905
+    },
+    {
+      "epoch": 2.358587383036523,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.408468244084683e-05,
+      "loss": 0.4082,
+      "step": 3906
+    },
+    {
+      "epoch": 2.359191065499547,
+      "grad_norm": 0.19140625,
+      "learning_rate": 4.404317144043172e-05,
+      "loss": 0.4422,
+      "step": 3907
+    },
+    {
+      "epoch": 2.359794747962572,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 4.400166044001661e-05,
+      "loss": 0.4258,
+      "step": 3908
+    },
+    {
+      "epoch": 2.360398430425596,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 4.3960149439601494e-05,
+      "loss": 0.3006,
+      "step": 3909
+    },
+    {
+      "epoch": 2.3610021128886207,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 4.391863843918639e-05,
+      "loss": 0.2906,
+      "step": 3910
+    },
+    {
+      "epoch": 2.361605795351645,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 4.3877127438771274e-05,
+      "loss": 0.2704,
+      "step": 3911
+    },
+    {
+      "epoch": 2.3622094778146696,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.383561643835617e-05,
+      "loss": 0.191,
+      "step": 3912
+    },
+    {
+      "epoch": 2.362813160277694,
+      "grad_norm": 0.146484375,
+      "learning_rate": 4.379410543794106e-05,
+      "loss": 0.6088,
+      "step": 3913
+    },
+    {
+      "epoch": 2.3634168427407185,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 4.375259443752595e-05,
+      "loss": 0.7435,
+      "step": 3914
+    },
+    {
+      "epoch": 2.3640205252037427,
+      "grad_norm": 0.15234375,
+      "learning_rate": 4.371108343711084e-05,
+      "loss": 0.6249,
+      "step": 3915
+    },
+    {
+      "epoch": 2.3646242076667674,
+      "grad_norm": 0.15234375,
+      "learning_rate": 4.3669572436695726e-05,
+      "loss": 0.6178,
+      "step": 3916
+    },
+    {
+      "epoch": 2.3652278901297916,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 4.362806143628061e-05,
+      "loss": 0.6073,
+      "step": 3917
+    },
+    {
+      "epoch": 2.3658315725928163,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 4.3586550435865506e-05,
+      "loss": 0.647,
+      "step": 3918
+    },
+    {
+      "epoch": 2.3664352550558405,
+      "grad_norm": 0.14453125,
+      "learning_rate": 4.35450394354504e-05,
+      "loss": 0.6807,
+      "step": 3919
+    },
+    {
+      "epoch": 2.367038937518865,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 4.3503528435035286e-05,
+      "loss": 0.5945,
+      "step": 3920
+    },
+    {
+      "epoch": 2.3676426199818894,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 4.346201743462018e-05,
+      "loss": 0.6119,
+      "step": 3921
+    },
+    {
+      "epoch": 2.368246302444914,
+      "grad_norm": 0.1640625,
+      "learning_rate": 4.3420506434205066e-05,
+      "loss": 0.6251,
+      "step": 3922
+    },
+    {
+      "epoch": 2.3688499849079383,
+      "grad_norm": 0.162109375,
+      "learning_rate": 4.337899543378995e-05,
+      "loss": 0.5501,
+      "step": 3923
+    },
+    {
+      "epoch": 2.369453667370963,
+      "grad_norm": 0.14453125,
+      "learning_rate": 4.3337484433374845e-05,
+      "loss": 0.6416,
+      "step": 3924
+    },
+    {
+      "epoch": 2.370057349833987,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 4.329597343295974e-05,
+      "loss": 0.6151,
+      "step": 3925
+    },
+    {
+      "epoch": 2.370661032297012,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 4.325446243254463e-05,
+      "loss": 0.7269,
+      "step": 3926
+    },
+    {
+      "epoch": 2.371264714760036,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 4.321295143212952e-05,
+      "loss": 0.5769,
+      "step": 3927
+    },
+    {
+      "epoch": 2.3718683972230608,
+      "grad_norm": 0.140625,
+      "learning_rate": 4.3171440431714405e-05,
+      "loss": 0.6409,
+      "step": 3928
+    },
+    {
+      "epoch": 2.372472079686085,
+      "grad_norm": 0.14453125,
+      "learning_rate": 4.31299294312993e-05,
+      "loss": 0.6153,
+      "step": 3929
+    },
+    {
+      "epoch": 2.3730757621491096,
+      "grad_norm": 0.14453125,
+      "learning_rate": 4.3088418430884185e-05,
+      "loss": 0.5554,
+      "step": 3930
+    },
+    {
+      "epoch": 2.373679444612134,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.304690743046907e-05,
+      "loss": 0.5879,
+      "step": 3931
+    },
+    {
+      "epoch": 2.3742831270751585,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 4.300539643005397e-05,
+      "loss": 0.5892,
+      "step": 3932
+    },
+    {
+      "epoch": 2.3748868095381828,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.296388542963886e-05,
+      "loss": 0.621,
+      "step": 3933
+    },
+    {
+      "epoch": 2.3754904920012074,
+      "grad_norm": 0.15234375,
+      "learning_rate": 4.2922374429223744e-05,
+      "loss": 1.0716,
+      "step": 3934
+    },
+    {
+      "epoch": 2.3760941744642317,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 4.288086342880864e-05,
+      "loss": 0.6236,
+      "step": 3935
+    },
+    {
+      "epoch": 2.3766978569272563,
+      "grad_norm": 0.142578125,
+      "learning_rate": 4.2839352428393524e-05,
+      "loss": 0.6244,
+      "step": 3936
+    },
+    {
+      "epoch": 2.3773015393902805,
+      "grad_norm": 0.150390625,
+      "learning_rate": 4.279784142797842e-05,
+      "loss": 0.6411,
+      "step": 3937
+    },
+    {
+      "epoch": 2.377905221853305,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 4.2756330427563304e-05,
+      "loss": 0.517,
+      "step": 3938
+    },
+    {
+      "epoch": 2.3785089043163294,
+      "grad_norm": 0.14453125,
+      "learning_rate": 4.27148194271482e-05,
+      "loss": 0.8545,
+      "step": 3939
+    },
+    {
+      "epoch": 2.379112586779354,
+      "grad_norm": 0.154296875,
+      "learning_rate": 4.267330842673309e-05,
+      "loss": 0.9618,
+      "step": 3940
+    },
+    {
+      "epoch": 2.3797162692423783,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 4.2631797426317977e-05,
+      "loss": 0.6273,
+      "step": 3941
+    },
+    {
+      "epoch": 2.380319951705403,
+      "grad_norm": 0.13671875,
+      "learning_rate": 4.259028642590286e-05,
+      "loss": 0.9013,
+      "step": 3942
+    },
+    {
+      "epoch": 2.3809236341684272,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 4.2548775425487756e-05,
+      "loss": 0.8099,
+      "step": 3943
+    },
+    {
+      "epoch": 2.381527316631452,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 4.250726442507264e-05,
+      "loss": 0.633,
+      "step": 3944
+    },
+    {
+      "epoch": 2.382130999094476,
+      "grad_norm": 0.1640625,
+      "learning_rate": 4.2465753424657536e-05,
+      "loss": 0.5849,
+      "step": 3945
+    },
+    {
+      "epoch": 2.382734681557501,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 4.242424242424243e-05,
+      "loss": 0.5141,
+      "step": 3946
+    },
+    {
+      "epoch": 2.383338364020525,
+      "grad_norm": 0.142578125,
+      "learning_rate": 4.2382731423827316e-05,
+      "loss": 0.5301,
+      "step": 3947
+    },
+    {
+      "epoch": 2.3839420464835497,
+      "grad_norm": 0.142578125,
+      "learning_rate": 4.234122042341221e-05,
+      "loss": 0.5945,
+      "step": 3948
+    },
+    {
+      "epoch": 2.3845457289465744,
+      "grad_norm": 0.146484375,
+      "learning_rate": 4.2299709422997096e-05,
+      "loss": 0.5041,
+      "step": 3949
+    },
+    {
+      "epoch": 2.3851494114095986,
+      "grad_norm": 0.15625,
+      "learning_rate": 4.225819842258198e-05,
+      "loss": 0.5563,
+      "step": 3950
+    },
+    {
+      "epoch": 2.385753093872623,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 4.2216687422166875e-05,
+      "loss": 0.5933,
+      "step": 3951
+    },
+    {
+      "epoch": 2.3863567763356475,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 4.217517642175177e-05,
+      "loss": 0.5283,
+      "step": 3952
+    },
+    {
+      "epoch": 2.386960458798672,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.2133665421336655e-05,
+      "loss": 0.5161,
+      "step": 3953
+    },
+    {
+      "epoch": 2.3875641412616964,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 4.209215442092155e-05,
+      "loss": 0.5154,
+      "step": 3954
+    },
+    {
+      "epoch": 2.3881678237247206,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 4.2050643420506435e-05,
+      "loss": 0.4278,
+      "step": 3955
+    },
+    {
+      "epoch": 2.3887715061877453,
+      "grad_norm": 0.177734375,
+      "learning_rate": 4.200913242009132e-05,
+      "loss": 0.5358,
+      "step": 3956
+    },
+    {
+      "epoch": 2.38937518865077,
+      "grad_norm": 0.1796875,
+      "learning_rate": 4.1967621419676215e-05,
+      "loss": 0.4306,
+      "step": 3957
+    },
+    {
+      "epoch": 2.389978871113794,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 4.192611041926111e-05,
+      "loss": 0.4228,
+      "step": 3958
+    },
+    {
+      "epoch": 2.3905825535768184,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 4.1884599418846e-05,
+      "loss": 0.3486,
+      "step": 3959
+    },
+    {
+      "epoch": 2.391186236039843,
+      "grad_norm": 0.205078125,
+      "learning_rate": 4.184308841843089e-05,
+      "loss": 0.3173,
+      "step": 3960
+    },
+    {
+      "epoch": 2.3917899185028677,
+      "grad_norm": 0.212890625,
+      "learning_rate": 4.1801577418015774e-05,
+      "loss": 0.328,
+      "step": 3961
+    },
+    {
+      "epoch": 2.392393600965892,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 4.176006641760067e-05,
+      "loss": 0.2109,
+      "step": 3962
+    },
+    {
+      "epoch": 2.392997283428916,
+      "grad_norm": 0.14453125,
+      "learning_rate": 4.1718555417185554e-05,
+      "loss": 0.7375,
+      "step": 3963
+    },
+    {
+      "epoch": 2.393600965891941,
+      "grad_norm": 0.146484375,
+      "learning_rate": 4.167704441677045e-05,
+      "loss": 0.515,
+      "step": 3964
+    },
+    {
+      "epoch": 2.3942046483549655,
+      "grad_norm": 0.150390625,
+      "learning_rate": 4.163553341635534e-05,
+      "loss": 0.6026,
+      "step": 3965
+    },
+    {
+      "epoch": 2.3948083308179897,
+      "grad_norm": 0.154296875,
+      "learning_rate": 4.159402241594023e-05,
+      "loss": 0.5353,
+      "step": 3966
+    },
+    {
+      "epoch": 2.395412013281014,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 4.155251141552511e-05,
+      "loss": 0.6007,
+      "step": 3967
+    },
+    {
+      "epoch": 2.3960156957440386,
+      "grad_norm": 0.134765625,
+      "learning_rate": 4.1511000415110007e-05,
+      "loss": 0.5639,
+      "step": 3968
+    },
+    {
+      "epoch": 2.3966193782070633,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 4.146948941469489e-05,
+      "loss": 0.5642,
+      "step": 3969
+    },
+    {
+      "epoch": 2.3972230606700875,
+      "grad_norm": 0.146484375,
+      "learning_rate": 4.1427978414279786e-05,
+      "loss": 0.584,
+      "step": 3970
+    },
+    {
+      "epoch": 2.397826743133112,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 4.138646741386468e-05,
+      "loss": 0.5897,
+      "step": 3971
+    },
+    {
+      "epoch": 2.3984304255961364,
+      "grad_norm": 0.1484375,
+      "learning_rate": 4.1344956413449566e-05,
+      "loss": 0.7358,
+      "step": 3972
+    },
+    {
+      "epoch": 2.399034108059161,
+      "grad_norm": 0.177734375,
+      "learning_rate": 4.130344541303446e-05,
+      "loss": 0.5912,
+      "step": 3973
+    },
+    {
+      "epoch": 2.3996377905221853,
+      "grad_norm": 0.158203125,
+      "learning_rate": 4.1261934412619346e-05,
+      "loss": 0.8172,
+      "step": 3974
+    },
+    {
+      "epoch": 2.40024147298521,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 4.122042341220423e-05,
+      "loss": 0.8762,
+      "step": 3975
+    },
+    {
+      "epoch": 2.400845155448234,
+      "grad_norm": 0.1484375,
+      "learning_rate": 4.1178912411789126e-05,
+      "loss": 0.6193,
+      "step": 3976
+    },
+    {
+      "epoch": 2.401448837911259,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 4.113740141137402e-05,
+      "loss": 0.6507,
+      "step": 3977
+    },
+    {
+      "epoch": 2.402052520374283,
+      "grad_norm": 0.158203125,
+      "learning_rate": 4.1095890410958905e-05,
+      "loss": 0.6294,
+      "step": 3978
+    },
+    {
+      "epoch": 2.4026562028373077,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.10543794105438e-05,
+      "loss": 0.6852,
+      "step": 3979
+    },
+    {
+      "epoch": 2.403259885300332,
+      "grad_norm": 0.189453125,
+      "learning_rate": 4.1012868410128685e-05,
+      "loss": 0.6219,
+      "step": 3980
+    },
+    {
+      "epoch": 2.4038635677633566,
+      "grad_norm": 0.138671875,
+      "learning_rate": 4.097135740971358e-05,
+      "loss": 0.5864,
+      "step": 3981
+    },
+    {
+      "epoch": 2.404467250226381,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 4.0929846409298465e-05,
+      "loss": 0.5816,
+      "step": 3982
+    },
+    {
+      "epoch": 2.4050709326894055,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.088833540888336e-05,
+      "loss": 0.7003,
+      "step": 3983
+    },
+    {
+      "epoch": 2.4056746151524298,
+      "grad_norm": 0.140625,
+      "learning_rate": 4.084682440846825e-05,
+      "loss": 0.6031,
+      "step": 3984
+    },
+    {
+      "epoch": 2.4062782976154544,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 4.080531340805314e-05,
+      "loss": 0.9704,
+      "step": 3985
+    },
+    {
+      "epoch": 2.4068819800784786,
+      "grad_norm": 0.142578125,
+      "learning_rate": 4.0763802407638024e-05,
+      "loss": 0.5736,
+      "step": 3986
+    },
+    {
+      "epoch": 2.4074856625415033,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 4.072229140722292e-05,
+      "loss": 0.8293,
+      "step": 3987
+    },
+    {
+      "epoch": 2.4080893450045275,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 4.0680780406807804e-05,
+      "loss": 0.5689,
+      "step": 3988
+    },
+    {
+      "epoch": 2.408693027467552,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 4.063926940639269e-05,
+      "loss": 0.6113,
+      "step": 3989
+    },
+    {
+      "epoch": 2.4092967099305764,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 4.059775840597759e-05,
+      "loss": 0.5141,
+      "step": 3990
+    },
+    {
+      "epoch": 2.409900392393601,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 4.055624740556248e-05,
+      "loss": 0.8989,
+      "step": 3991
+    },
+    {
+      "epoch": 2.4105040748566253,
+      "grad_norm": 0.140625,
+      "learning_rate": 4.0514736405147363e-05,
+      "loss": 0.5546,
+      "step": 3992
+    },
+    {
+      "epoch": 2.41110775731965,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 4.047322540473226e-05,
+      "loss": 0.5768,
+      "step": 3993
+    },
+    {
+      "epoch": 2.411711439782674,
+      "grad_norm": 0.373046875,
+      "learning_rate": 4.043171440431714e-05,
+      "loss": 0.5518,
+      "step": 3994
+    },
+    {
+      "epoch": 2.412315122245699,
+      "grad_norm": 0.1328125,
+      "learning_rate": 4.0390203403902036e-05,
+      "loss": 0.5835,
+      "step": 3995
+    },
+    {
+      "epoch": 2.412918804708723,
+      "grad_norm": 0.14453125,
+      "learning_rate": 4.034869240348692e-05,
+      "loss": 0.8638,
+      "step": 3996
+    },
+    {
+      "epoch": 2.4135224871717478,
+      "grad_norm": 0.138671875,
+      "learning_rate": 4.0307181403071816e-05,
+      "loss": 0.5474,
+      "step": 3997
+    },
+    {
+      "epoch": 2.414126169634772,
+      "grad_norm": 0.13671875,
+      "learning_rate": 4.026567040265671e-05,
+      "loss": 0.5179,
+      "step": 3998
+    },
+    {
+      "epoch": 2.4147298520977967,
+      "grad_norm": 0.150390625,
+      "learning_rate": 4.0224159402241596e-05,
+      "loss": 0.5755,
+      "step": 3999
+    },
+    {
+      "epoch": 2.415333534560821,
+      "grad_norm": 0.15234375,
+      "learning_rate": 4.018264840182648e-05,
+      "loss": 0.4952,
+      "step": 4000
+    },
+    {
+      "epoch": 2.415333534560821,
+      "eval_loss": 0.6154825687408447,
+      "eval_runtime": 1059.7325,
+      "eval_samples_per_second": 2.633,
+      "eval_steps_per_second": 0.329,
+      "step": 4000
+    },
+    {
+      "epoch": 2.4159372170238456,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 4.0141137401411376e-05,
+      "loss": 0.5103,
+      "step": 4001
+    },
+    {
+      "epoch": 2.41654089948687,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 4.009962640099626e-05,
+      "loss": 0.5184,
+      "step": 4002
+    },
+    {
+      "epoch": 2.4171445819498945,
+      "grad_norm": 0.1640625,
+      "learning_rate": 4.0058115400581155e-05,
+      "loss": 0.4408,
+      "step": 4003
+    },
+    {
+      "epoch": 2.4177482644129187,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 4.001660440016605e-05,
+      "loss": 0.531,
+      "step": 4004
+    },
+    {
+      "epoch": 2.4183519468759433,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 3.9975093399750935e-05,
+      "loss": 0.5268,
+      "step": 4005
+    },
+    {
+      "epoch": 2.4189556293389676,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.993358239933583e-05,
+      "loss": 0.4319,
+      "step": 4006
+    },
+    {
+      "epoch": 2.4195593118019922,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.9892071398920715e-05,
+      "loss": 0.3856,
+      "step": 4007
+    },
+    {
+      "epoch": 2.4201629942650165,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 3.98505603985056e-05,
+      "loss": 0.4184,
+      "step": 4008
+    },
+    {
+      "epoch": 2.420766676728041,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 3.9809049398090495e-05,
+      "loss": 0.4034,
+      "step": 4009
+    },
+    {
+      "epoch": 2.4213703591910654,
+      "grad_norm": 0.193359375,
+      "learning_rate": 3.976753839767539e-05,
+      "loss": 0.2712,
+      "step": 4010
+    },
+    {
+      "epoch": 2.42197404165409,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.9726027397260274e-05,
+      "loss": 0.2546,
+      "step": 4011
+    },
+    {
+      "epoch": 2.4225777241171143,
+      "grad_norm": 0.185546875,
+      "learning_rate": 3.968451639684517e-05,
+      "loss": 0.2069,
+      "step": 4012
+    },
+    {
+      "epoch": 2.423181406580139,
+      "grad_norm": 0.16015625,
+      "learning_rate": 3.9643005396430054e-05,
+      "loss": 0.6653,
+      "step": 4013
+    },
+    {
+      "epoch": 2.423785089043163,
+      "grad_norm": 0.1484375,
+      "learning_rate": 3.960149439601494e-05,
+      "loss": 0.5395,
+      "step": 4014
+    },
+    {
+      "epoch": 2.424388771506188,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 3.9559983395599834e-05,
+      "loss": 0.6241,
+      "step": 4015
+    },
+    {
+      "epoch": 2.424992453969212,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 3.951847239518473e-05,
+      "loss": 0.628,
+      "step": 4016
+    },
+    {
+      "epoch": 2.4255961364322367,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.947696139476962e-05,
+      "loss": 0.6188,
+      "step": 4017
+    },
+    {
+      "epoch": 2.426199818895261,
+      "grad_norm": 0.15234375,
+      "learning_rate": 3.943545039435451e-05,
+      "loss": 0.7127,
+      "step": 4018
+    },
+    {
+      "epoch": 2.4268035013582856,
+      "grad_norm": 0.154296875,
+      "learning_rate": 3.939393939393939e-05,
+      "loss": 0.8119,
+      "step": 4019
+    },
+    {
+      "epoch": 2.42740718382131,
+      "grad_norm": 0.1484375,
+      "learning_rate": 3.935242839352429e-05,
+      "loss": 0.5477,
+      "step": 4020
+    },
+    {
+      "epoch": 2.4280108662843345,
+      "grad_norm": 0.138671875,
+      "learning_rate": 3.931091739310917e-05,
+      "loss": 0.557,
+      "step": 4021
+    },
+    {
+      "epoch": 2.4286145487473587,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 3.9269406392694066e-05,
+      "loss": 0.8675,
+      "step": 4022
+    },
+    {
+      "epoch": 2.4292182312103834,
+      "grad_norm": 0.1328125,
+      "learning_rate": 3.922789539227896e-05,
+      "loss": 0.7214,
+      "step": 4023
+    },
+    {
+      "epoch": 2.4298219136734076,
+      "grad_norm": 0.14453125,
+      "learning_rate": 3.9186384391863846e-05,
+      "loss": 0.6073,
+      "step": 4024
+    },
+    {
+      "epoch": 2.4304255961364323,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 3.914487339144873e-05,
+      "loss": 0.6403,
+      "step": 4025
+    },
+    {
+      "epoch": 2.4310292785994565,
+      "grad_norm": 0.146484375,
+      "learning_rate": 3.9103362391033626e-05,
+      "loss": 0.7011,
+      "step": 4026
+    },
+    {
+      "epoch": 2.431632961062481,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 3.906185139061851e-05,
+      "loss": 0.6148,
+      "step": 4027
+    },
+    {
+      "epoch": 2.4322366435255054,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 3.9020340390203406e-05,
+      "loss": 0.743,
+      "step": 4028
+    },
+    {
+      "epoch": 2.43284032598853,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 3.89788293897883e-05,
+      "loss": 0.6212,
+      "step": 4029
+    },
+    {
+      "epoch": 2.4334440084515547,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 3.8937318389373185e-05,
+      "loss": 0.5285,
+      "step": 4030
+    },
+    {
+      "epoch": 2.434047690914579,
+      "grad_norm": 0.140625,
+      "learning_rate": 3.889580738895808e-05,
+      "loss": 0.7145,
+      "step": 4031
+    },
+    {
+      "epoch": 2.434651373377603,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 3.8854296388542965e-05,
+      "loss": 0.6085,
+      "step": 4032
+    },
+    {
+      "epoch": 2.435255055840628,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 3.881278538812785e-05,
+      "loss": 0.8197,
+      "step": 4033
+    },
+    {
+      "epoch": 2.4358587383036525,
+      "grad_norm": 0.142578125,
+      "learning_rate": 3.8771274387712745e-05,
+      "loss": 0.5681,
+      "step": 4034
+    },
+    {
+      "epoch": 2.4364624207666767,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.872976338729764e-05,
+      "loss": 0.6784,
+      "step": 4035
+    },
+    {
+      "epoch": 2.437066103229701,
+      "grad_norm": 0.15234375,
+      "learning_rate": 3.8688252386882525e-05,
+      "loss": 1.1645,
+      "step": 4036
+    },
+    {
+      "epoch": 2.4376697856927256,
+      "grad_norm": 0.15234375,
+      "learning_rate": 3.864674138646742e-05,
+      "loss": 0.8407,
+      "step": 4037
+    },
+    {
+      "epoch": 2.4382734681557503,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.8605230386052304e-05,
+      "loss": 0.5816,
+      "step": 4038
+    },
+    {
+      "epoch": 2.4388771506187745,
+      "grad_norm": 0.208984375,
+      "learning_rate": 3.85637193856372e-05,
+      "loss": 0.9043,
+      "step": 4039
+    },
+    {
+      "epoch": 2.4394808330817987,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 3.8522208385222084e-05,
+      "loss": 0.5051,
+      "step": 4040
+    },
+    {
+      "epoch": 2.4400845155448234,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 3.848069738480698e-05,
+      "loss": 0.5917,
+      "step": 4041
+    },
+    {
+      "epoch": 2.440688198007848,
+      "grad_norm": 0.15625,
+      "learning_rate": 3.843918638439187e-05,
+      "loss": 0.6743,
+      "step": 4042
+    },
+    {
+      "epoch": 2.4412918804708723,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 3.839767538397676e-05,
+      "loss": 0.6768,
+      "step": 4043
+    },
+    {
+      "epoch": 2.4418955629338965,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 3.8356164383561644e-05,
+      "loss": 0.7161,
+      "step": 4044
+    },
+    {
+      "epoch": 2.442499245396921,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 3.831465338314654e-05,
+      "loss": 0.5666,
+      "step": 4045
+    },
+    {
+      "epoch": 2.443102927859946,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 3.827314238273142e-05,
+      "loss": 0.5713,
+      "step": 4046
+    },
+    {
+      "epoch": 2.44370661032297,
+      "grad_norm": 0.142578125,
+      "learning_rate": 3.823163138231631e-05,
+      "loss": 0.5426,
+      "step": 4047
+    },
+    {
+      "epoch": 2.4443102927859943,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 3.819012038190121e-05,
+      "loss": 0.5295,
+      "step": 4048
+    },
+    {
+      "epoch": 2.444913975249019,
+      "grad_norm": 0.158203125,
+      "learning_rate": 3.8148609381486096e-05,
+      "loss": 0.5478,
+      "step": 4049
+    },
+    {
+      "epoch": 2.4455176577120437,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.810709838107099e-05,
+      "loss": 0.5445,
+      "step": 4050
+    },
+    {
+      "epoch": 2.446121340175068,
+      "grad_norm": 0.185546875,
+      "learning_rate": 3.8065587380655876e-05,
+      "loss": 0.5635,
+      "step": 4051
+    },
+    {
+      "epoch": 2.4467250226380926,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.802407638024076e-05,
+      "loss": 0.5435,
+      "step": 4052
+    },
+    {
+      "epoch": 2.4473287051011168,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 3.7982565379825656e-05,
+      "loss": 0.4776,
+      "step": 4053
+    },
+    {
+      "epoch": 2.4479323875641414,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.794105437941054e-05,
+      "loss": 0.4868,
+      "step": 4054
+    },
+    {
+      "epoch": 2.4485360700271657,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 3.7899543378995436e-05,
+      "loss": 0.4523,
+      "step": 4055
+    },
+    {
+      "epoch": 2.4491397524901903,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.785803237858033e-05,
+      "loss": 0.3794,
+      "step": 4056
+    },
+    {
+      "epoch": 2.4497434349532146,
+      "grad_norm": 0.193359375,
+      "learning_rate": 3.7816521378165215e-05,
+      "loss": 0.4109,
+      "step": 4057
+    },
+    {
+      "epoch": 2.4503471174162392,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 3.77750103777501e-05,
+      "loss": 0.3978,
+      "step": 4058
+    },
+    {
+      "epoch": 2.4509507998792635,
+      "grad_norm": 0.189453125,
+      "learning_rate": 3.7733499377334995e-05,
+      "loss": 0.3167,
+      "step": 4059
+    },
+    {
+      "epoch": 2.451554482342288,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 3.769198837691988e-05,
+      "loss": 0.3251,
+      "step": 4060
+    },
+    {
+      "epoch": 2.4521581648053123,
+      "grad_norm": 0.189453125,
+      "learning_rate": 3.7650477376504775e-05,
+      "loss": 0.241,
+      "step": 4061
+    },
+    {
+      "epoch": 2.452761847268337,
+      "grad_norm": 0.20703125,
+      "learning_rate": 3.760896637608967e-05,
+      "loss": 0.2344,
+      "step": 4062
+    },
+    {
+      "epoch": 2.4533655297313612,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 3.7567455375674554e-05,
+      "loss": 0.6431,
+      "step": 4063
+    },
+    {
+      "epoch": 2.453969212194386,
+      "grad_norm": 0.14453125,
+      "learning_rate": 3.752594437525945e-05,
+      "loss": 0.5608,
+      "step": 4064
+    },
+    {
+      "epoch": 2.45457289465741,
+      "grad_norm": 0.158203125,
+      "learning_rate": 3.7484433374844334e-05,
+      "loss": 0.6186,
+      "step": 4065
+    },
+    {
+      "epoch": 2.455176577120435,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 3.744292237442922e-05,
+      "loss": 0.5959,
+      "step": 4066
+    },
+    {
+      "epoch": 2.455780259583459,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 3.7401411374014114e-05,
+      "loss": 0.5077,
+      "step": 4067
+    },
+    {
+      "epoch": 2.4563839420464837,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 3.735990037359901e-05,
+      "loss": 0.578,
+      "step": 4068
+    },
+    {
+      "epoch": 2.456987624509508,
+      "grad_norm": 0.154296875,
+      "learning_rate": 3.7318389373183894e-05,
+      "loss": 0.6015,
+      "step": 4069
+    },
+    {
+      "epoch": 2.4575913069725326,
+      "grad_norm": 0.146484375,
+      "learning_rate": 3.727687837276879e-05,
+      "loss": 0.6267,
+      "step": 4070
+    },
+    {
+      "epoch": 2.458194989435557,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 3.7235367372353673e-05,
+      "loss": 0.6992,
+      "step": 4071
+    },
+    {
+      "epoch": 2.4587986718985815,
+      "grad_norm": 0.15234375,
+      "learning_rate": 3.719385637193857e-05,
+      "loss": 0.6082,
+      "step": 4072
+    },
+    {
+      "epoch": 2.4594023543616057,
+      "grad_norm": 0.150390625,
+      "learning_rate": 3.715234537152345e-05,
+      "loss": 0.6808,
+      "step": 4073
+    },
+    {
+      "epoch": 2.4600060368246304,
+      "grad_norm": 0.154296875,
+      "learning_rate": 3.7110834371108346e-05,
+      "loss": 0.6117,
+      "step": 4074
+    },
+    {
+      "epoch": 2.4606097192876546,
+      "grad_norm": 0.146484375,
+      "learning_rate": 3.706932337069324e-05,
+      "loss": 0.9315,
+      "step": 4075
+    },
+    {
+      "epoch": 2.4612134017506793,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 3.7027812370278126e-05,
+      "loss": 0.647,
+      "step": 4076
+    },
+    {
+      "epoch": 2.4618170842137035,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 3.698630136986301e-05,
+      "loss": 0.7295,
+      "step": 4077
+    },
+    {
+      "epoch": 2.462420766676728,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 3.6944790369447906e-05,
+      "loss": 0.6171,
+      "step": 4078
+    },
+    {
+      "epoch": 2.4630244491397524,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 3.690327936903279e-05,
+      "loss": 0.698,
+      "step": 4079
+    },
+    {
+      "epoch": 2.463628131602777,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 3.6861768368617686e-05,
+      "loss": 0.5728,
+      "step": 4080
+    },
+    {
+      "epoch": 2.4642318140658013,
+      "grad_norm": 0.13671875,
+      "learning_rate": 3.682025736820258e-05,
+      "loss": 0.6312,
+      "step": 4081
+    },
+    {
+      "epoch": 2.464835496528826,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 3.6778746367787465e-05,
+      "loss": 0.6361,
+      "step": 4082
+    },
+    {
+      "epoch": 2.46543917899185,
+      "grad_norm": 0.138671875,
+      "learning_rate": 3.673723536737236e-05,
+      "loss": 0.6073,
+      "step": 4083
+    },
+    {
+      "epoch": 2.466042861454875,
+      "grad_norm": 0.16796875,
+      "learning_rate": 3.6695724366957245e-05,
+      "loss": 0.5762,
+      "step": 4084
+    },
+    {
+      "epoch": 2.466646543917899,
+      "grad_norm": 0.146484375,
+      "learning_rate": 3.665421336654213e-05,
+      "loss": 0.704,
+      "step": 4085
+    },
+    {
+      "epoch": 2.4672502263809237,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 3.6612702366127025e-05,
+      "loss": 0.5739,
+      "step": 4086
+    },
+    {
+      "epoch": 2.467853908843948,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.657119136571192e-05,
+      "loss": 0.6475,
+      "step": 4087
+    },
+    {
+      "epoch": 2.4684575913069726,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 3.6529680365296805e-05,
+      "loss": 0.6407,
+      "step": 4088
+    },
+    {
+      "epoch": 2.469061273769997,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 3.64881693648817e-05,
+      "loss": 0.6102,
+      "step": 4089
+    },
+    {
+      "epoch": 2.4696649562330215,
+      "grad_norm": 0.138671875,
+      "learning_rate": 3.6446658364466584e-05,
+      "loss": 0.5091,
+      "step": 4090
+    },
+    {
+      "epoch": 2.4702686386960457,
+      "grad_norm": 0.146484375,
+      "learning_rate": 3.640514736405147e-05,
+      "loss": 0.5409,
+      "step": 4091
+    },
+    {
+      "epoch": 2.4708723211590704,
+      "grad_norm": 0.142578125,
+      "learning_rate": 3.6363636363636364e-05,
+      "loss": 0.7817,
+      "step": 4092
+    },
+    {
+      "epoch": 2.4714760036220946,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 3.632212536322126e-05,
+      "loss": 0.604,
+      "step": 4093
+    },
+    {
+      "epoch": 2.4720796860851193,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 3.628061436280615e-05,
+      "loss": 0.6486,
+      "step": 4094
+    },
+    {
+      "epoch": 2.4726833685481435,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 3.623910336239104e-05,
+      "loss": 0.5922,
+      "step": 4095
+    },
+    {
+      "epoch": 2.473287051011168,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 3.6197592361975924e-05,
+      "loss": 0.5334,
+      "step": 4096
+    },
+    {
+      "epoch": 2.4738907334741924,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.615608136156082e-05,
+      "loss": 0.5537,
+      "step": 4097
+    },
+    {
+      "epoch": 2.474494415937217,
+      "grad_norm": 0.146484375,
+      "learning_rate": 3.61145703611457e-05,
+      "loss": 0.5546,
+      "step": 4098
+    },
+    {
+      "epoch": 2.4750980984002413,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 3.60730593607306e-05,
+      "loss": 0.4821,
+      "step": 4099
+    },
+    {
+      "epoch": 2.475701780863266,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 3.603154836031549e-05,
+      "loss": 0.5078,
+      "step": 4100
+    },
+    {
+      "epoch": 2.47630546332629,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 3.5990037359900376e-05,
+      "loss": 0.4714,
+      "step": 4101
+    },
+    {
+      "epoch": 2.476909145789315,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.594852635948526e-05,
+      "loss": 0.5351,
+      "step": 4102
+    },
+    {
+      "epoch": 2.477512828252339,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.5907015359070156e-05,
+      "loss": 0.5389,
+      "step": 4103
+    },
+    {
+      "epoch": 2.4781165107153638,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.586550435865504e-05,
+      "loss": 0.5224,
+      "step": 4104
+    },
+    {
+      "epoch": 2.478720193178388,
+      "grad_norm": 0.18359375,
+      "learning_rate": 3.5823993358239936e-05,
+      "loss": 0.4301,
+      "step": 4105
+    },
+    {
+      "epoch": 2.4793238756414127,
+      "grad_norm": 0.181640625,
+      "learning_rate": 3.578248235782483e-05,
+      "loss": 0.4746,
+      "step": 4106
+    },
+    {
+      "epoch": 2.479927558104437,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.5740971357409716e-05,
+      "loss": 0.3572,
+      "step": 4107
+    },
+    {
+      "epoch": 2.4805312405674615,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 3.569946035699461e-05,
+      "loss": 0.4036,
+      "step": 4108
+    },
+    {
+      "epoch": 2.4811349230304858,
+      "grad_norm": 0.21484375,
+      "learning_rate": 3.5657949356579495e-05,
+      "loss": 0.3364,
+      "step": 4109
+    },
+    {
+      "epoch": 2.4817386054935104,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.561643835616438e-05,
+      "loss": 0.3264,
+      "step": 4110
+    },
+    {
+      "epoch": 2.482342287956535,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.5574927355749275e-05,
+      "loss": 0.2525,
+      "step": 4111
+    },
+    {
+      "epoch": 2.4829459704195593,
+      "grad_norm": 0.353515625,
+      "learning_rate": 3.553341635533416e-05,
+      "loss": 0.2355,
+      "step": 4112
+    },
+    {
+      "epoch": 2.4835496528825836,
+      "grad_norm": 0.13671875,
+      "learning_rate": 3.5491905354919055e-05,
+      "loss": 0.6404,
+      "step": 4113
+    },
+    {
+      "epoch": 2.4841533353456082,
+      "grad_norm": 0.15625,
+      "learning_rate": 3.545039435450395e-05,
+      "loss": 0.6651,
+      "step": 4114
+    },
+    {
+      "epoch": 2.484757017808633,
+      "grad_norm": 0.146484375,
+      "learning_rate": 3.5408883354088835e-05,
+      "loss": 0.685,
+      "step": 4115
+    },
+    {
+      "epoch": 2.485360700271657,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 3.536737235367373e-05,
+      "loss": 0.5109,
+      "step": 4116
+    },
+    {
+      "epoch": 2.4859643827346813,
+      "grad_norm": 0.140625,
+      "learning_rate": 3.5325861353258614e-05,
+      "loss": 0.7178,
+      "step": 4117
+    },
+    {
+      "epoch": 2.486568065197706,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 3.52843503528435e-05,
+      "loss": 0.5863,
+      "step": 4118
+    },
+    {
+      "epoch": 2.4871717476607307,
+      "grad_norm": 0.146484375,
+      "learning_rate": 3.5242839352428394e-05,
+      "loss": 0.6693,
+      "step": 4119
+    },
+    {
+      "epoch": 2.487775430123755,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 3.520132835201329e-05,
+      "loss": 0.6018,
+      "step": 4120
+    },
+    {
+      "epoch": 2.488379112586779,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 3.5159817351598174e-05,
+      "loss": 0.6141,
+      "step": 4121
+    },
+    {
+      "epoch": 2.488982795049804,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 3.511830635118307e-05,
+      "loss": 0.7566,
+      "step": 4122
+    },
+    {
+      "epoch": 2.4895864775128285,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 3.5076795350767954e-05,
+      "loss": 0.6085,
+      "step": 4123
+    },
+    {
+      "epoch": 2.4901901599758527,
+      "grad_norm": 0.14453125,
+      "learning_rate": 3.503528435035284e-05,
+      "loss": 0.6169,
+      "step": 4124
+    },
+    {
+      "epoch": 2.490793842438877,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 3.499377334993773e-05,
+      "loss": 1.1005,
+      "step": 4125
+    },
+    {
+      "epoch": 2.4913975249019016,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 3.4952262349522627e-05,
+      "loss": 0.6369,
+      "step": 4126
+    },
+    {
+      "epoch": 2.4920012073649263,
+      "grad_norm": 0.140625,
+      "learning_rate": 3.491075134910752e-05,
+      "loss": 0.7739,
+      "step": 4127
+    },
+    {
+      "epoch": 2.4926048898279505,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 3.4869240348692406e-05,
+      "loss": 0.6243,
+      "step": 4128
+    },
+    {
+      "epoch": 2.4932085722909747,
+      "grad_norm": 0.158203125,
+      "learning_rate": 3.482772934827729e-05,
+      "loss": 0.6468,
+      "step": 4129
+    },
+    {
+      "epoch": 2.4938122547539994,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 3.4786218347862186e-05,
+      "loss": 0.6467,
+      "step": 4130
+    },
+    {
+      "epoch": 2.494415937217024,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 3.474470734744707e-05,
+      "loss": 0.6936,
+      "step": 4131
+    },
+    {
+      "epoch": 2.4950196196800483,
+      "grad_norm": 0.138671875,
+      "learning_rate": 3.4703196347031966e-05,
+      "loss": 0.6583,
+      "step": 4132
+    },
+    {
+      "epoch": 2.495623302143073,
+      "grad_norm": 0.13671875,
+      "learning_rate": 3.466168534661686e-05,
+      "loss": 0.55,
+      "step": 4133
+    },
+    {
+      "epoch": 2.496226984606097,
+      "grad_norm": 0.14453125,
+      "learning_rate": 3.4620174346201746e-05,
+      "loss": 0.6066,
+      "step": 4134
+    },
+    {
+      "epoch": 2.496830667069122,
+      "grad_norm": 0.1328125,
+      "learning_rate": 3.457866334578663e-05,
+      "loss": 0.6148,
+      "step": 4135
+    },
+    {
+      "epoch": 2.497434349532146,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 3.4537152345371525e-05,
+      "loss": 0.6258,
+      "step": 4136
+    },
+    {
+      "epoch": 2.4980380319951707,
+      "grad_norm": 0.142578125,
+      "learning_rate": 3.449564134495641e-05,
+      "loss": 0.6906,
+      "step": 4137
+    },
+    {
+      "epoch": 2.498641714458195,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 3.4454130344541305e-05,
+      "loss": 0.601,
+      "step": 4138
+    },
+    {
+      "epoch": 2.4992453969212196,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 3.44126193441262e-05,
+      "loss": 0.6109,
+      "step": 4139
+    },
+    {
+      "epoch": 2.499849079384244,
+      "grad_norm": 0.138671875,
+      "learning_rate": 3.4371108343711085e-05,
+      "loss": 0.6139,
+      "step": 4140
+    },
+    {
+      "epoch": 2.5004527618472685,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 3.432959734329598e-05,
+      "loss": 0.4505,
+      "step": 4141
+    },
+    {
+      "epoch": 2.5010564443102927,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 3.4288086342880864e-05,
+      "loss": 0.5743,
+      "step": 4142
+    },
+    {
+      "epoch": 2.5016601267733174,
+      "grad_norm": 0.138671875,
+      "learning_rate": 3.424657534246575e-05,
+      "loss": 0.6745,
+      "step": 4143
+    },
+    {
+      "epoch": 2.5022638092363416,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 3.4205064342050644e-05,
+      "loss": 0.7918,
+      "step": 4144
+    },
+    {
+      "epoch": 2.5028674916993663,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 3.416355334163554e-05,
+      "loss": 0.5862,
+      "step": 4145
+    },
+    {
+      "epoch": 2.5034711741623905,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 3.4122042341220424e-05,
+      "loss": 0.6077,
+      "step": 4146
+    },
+    {
+      "epoch": 2.504074856625415,
+      "grad_norm": 0.142578125,
+      "learning_rate": 3.408053134080532e-05,
+      "loss": 0.5726,
+      "step": 4147
+    },
+    {
+      "epoch": 2.5046785390884394,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 3.4039020340390204e-05,
+      "loss": 0.5069,
+      "step": 4148
+    },
+    {
+      "epoch": 2.505282221551464,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 3.39975093399751e-05,
+      "loss": 0.5979,
+      "step": 4149
+    },
+    {
+      "epoch": 2.5058859040144883,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 3.3955998339559983e-05,
+      "loss": 0.5018,
+      "step": 4150
+    },
+    {
+      "epoch": 2.506489586477513,
+      "grad_norm": 0.158203125,
+      "learning_rate": 3.391448733914488e-05,
+      "loss": 0.4882,
+      "step": 4151
+    },
+    {
+      "epoch": 2.507093268940537,
+      "grad_norm": 0.158203125,
+      "learning_rate": 3.387297633872977e-05,
+      "loss": 0.5086,
+      "step": 4152
+    },
+    {
+      "epoch": 2.507696951403562,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.3831465338314656e-05,
+      "loss": 0.523,
+      "step": 4153
+    },
+    {
+      "epoch": 2.508300633866586,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.378995433789954e-05,
+      "loss": 0.4776,
+      "step": 4154
+    },
+    {
+      "epoch": 2.5089043163296108,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.3748443337484436e-05,
+      "loss": 0.5318,
+      "step": 4155
+    },
+    {
+      "epoch": 2.509507998792635,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.370693233706932e-05,
+      "loss": 0.4694,
+      "step": 4156
+    },
+    {
+      "epoch": 2.5101116812556596,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 3.3665421336654216e-05,
+      "loss": 0.4135,
+      "step": 4157
+    },
+    {
+      "epoch": 2.510715363718684,
+      "grad_norm": 0.18359375,
+      "learning_rate": 3.362391033623911e-05,
+      "loss": 0.3618,
+      "step": 4158
+    },
+    {
+      "epoch": 2.5113190461817085,
+      "grad_norm": 0.20703125,
+      "learning_rate": 3.3582399335823996e-05,
+      "loss": 0.412,
+      "step": 4159
+    },
+    {
+      "epoch": 2.5119227286447328,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 3.354088833540889e-05,
+      "loss": 0.3636,
+      "step": 4160
+    },
+    {
+      "epoch": 2.5125264111077574,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.3499377334993775e-05,
+      "loss": 0.2725,
+      "step": 4161
+    },
+    {
+      "epoch": 2.5131300935707817,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 3.345786633457866e-05,
+      "loss": 0.2317,
+      "step": 4162
+    },
+    {
+      "epoch": 2.5137337760338063,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 3.3416355334163555e-05,
+      "loss": 0.5467,
+      "step": 4163
+    },
+    {
+      "epoch": 2.5143374584968305,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 3.337484433374845e-05,
+      "loss": 0.6671,
+      "step": 4164
+    },
+    {
+      "epoch": 2.514941140959855,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.653,
+      "step": 4165
+    },
+    {
+      "epoch": 2.5155448234228794,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 3.329182233291823e-05,
+      "loss": 0.6325,
+      "step": 4166
+    },
+    {
+      "epoch": 2.516148505885904,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 3.3250311332503115e-05,
+      "loss": 0.593,
+      "step": 4167
+    },
+    {
+      "epoch": 2.5167521883489283,
+      "grad_norm": 0.162109375,
+      "learning_rate": 3.3208800332088e-05,
+      "loss": 0.6592,
+      "step": 4168
+    },
+    {
+      "epoch": 2.517355870811953,
+      "grad_norm": 0.15234375,
+      "learning_rate": 3.3167289331672894e-05,
+      "loss": 0.6527,
+      "step": 4169
+    },
+    {
+      "epoch": 2.5179595532749772,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 3.312577833125778e-05,
+      "loss": 0.6054,
+      "step": 4170
+    },
+    {
+      "epoch": 2.518563235738002,
+      "grad_norm": 0.150390625,
+      "learning_rate": 3.308426733084268e-05,
+      "loss": 0.6671,
+      "step": 4171
+    },
+    {
+      "epoch": 2.519166918201026,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 3.304275633042757e-05,
+      "loss": 0.527,
+      "step": 4172
+    },
+    {
+      "epoch": 2.519770600664051,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 3.3001245330012454e-05,
+      "loss": 0.5836,
+      "step": 4173
+    },
+    {
+      "epoch": 2.520374283127075,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 3.295973432959735e-05,
+      "loss": 0.5882,
+      "step": 4174
+    },
+    {
+      "epoch": 2.5209779655900997,
+      "grad_norm": 0.14453125,
+      "learning_rate": 3.2918223329182234e-05,
+      "loss": 0.6148,
+      "step": 4175
+    },
+    {
+      "epoch": 2.521581648053124,
+      "grad_norm": 0.16796875,
+      "learning_rate": 3.287671232876712e-05,
+      "loss": 0.6586,
+      "step": 4176
+    },
+    {
+      "epoch": 2.5221853305161486,
+      "grad_norm": 0.13671875,
+      "learning_rate": 3.2835201328352013e-05,
+      "loss": 0.496,
+      "step": 4177
+    },
+    {
+      "epoch": 2.522789012979173,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 3.279369032793691e-05,
+      "loss": 0.6182,
+      "step": 4178
+    },
+    {
+      "epoch": 2.5233926954421975,
+      "grad_norm": 0.15625,
+      "learning_rate": 3.275217932752179e-05,
+      "loss": 0.5696,
+      "step": 4179
+    },
+    {
+      "epoch": 2.5239963779052217,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 3.2710668327106686e-05,
+      "loss": 0.643,
+      "step": 4180
+    },
+    {
+      "epoch": 2.5246000603682464,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 3.266915732669157e-05,
+      "loss": 0.6102,
+      "step": 4181
+    },
+    {
+      "epoch": 2.5252037428312706,
+      "grad_norm": 0.15234375,
+      "learning_rate": 3.2627646326276466e-05,
+      "loss": 0.5995,
+      "step": 4182
+    },
+    {
+      "epoch": 2.5258074252942952,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 3.258613532586135e-05,
+      "loss": 0.5864,
+      "step": 4183
+    },
+    {
+      "epoch": 2.52641110775732,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 3.2544624325446246e-05,
+      "loss": 0.5564,
+      "step": 4184
+    },
+    {
+      "epoch": 2.527014790220344,
+      "grad_norm": 0.15234375,
+      "learning_rate": 3.250311332503114e-05,
+      "loss": 0.5808,
+      "step": 4185
+    },
+    {
+      "epoch": 2.5276184726833684,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.2461602324616026e-05,
+      "loss": 0.5563,
+      "step": 4186
+    },
+    {
+      "epoch": 2.528222155146393,
+      "grad_norm": 0.140625,
+      "learning_rate": 3.242009132420091e-05,
+      "loss": 0.5779,
+      "step": 4187
+    },
+    {
+      "epoch": 2.5288258376094177,
+      "grad_norm": 0.14453125,
+      "learning_rate": 3.2378580323785805e-05,
+      "loss": 0.6378,
+      "step": 4188
+    },
+    {
+      "epoch": 2.529429520072442,
+      "grad_norm": 0.1484375,
+      "learning_rate": 3.233706932337069e-05,
+      "loss": 0.5168,
+      "step": 4189
+    },
+    {
+      "epoch": 2.530033202535466,
+      "grad_norm": 0.146484375,
+      "learning_rate": 3.2295558322955585e-05,
+      "loss": 0.6191,
+      "step": 4190
+    },
+    {
+      "epoch": 2.530636884998491,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 3.225404732254048e-05,
+      "loss": 0.6251,
+      "step": 4191
+    },
+    {
+      "epoch": 2.5312405674615155,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 3.2212536322125365e-05,
+      "loss": 0.592,
+      "step": 4192
+    },
+    {
+      "epoch": 2.5318442499245397,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 3.217102532171026e-05,
+      "loss": 0.5686,
+      "step": 4193
+    },
+    {
+      "epoch": 2.532447932387564,
+      "grad_norm": 0.138671875,
+      "learning_rate": 3.2129514321295145e-05,
+      "loss": 0.7272,
+      "step": 4194
+    },
+    {
+      "epoch": 2.5330516148505886,
+      "grad_norm": 0.1328125,
+      "learning_rate": 3.208800332088003e-05,
+      "loss": 0.684,
+      "step": 4195
+    },
+    {
+      "epoch": 2.5336552973136133,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 3.2046492320464924e-05,
+      "loss": 0.5861,
+      "step": 4196
+    },
+    {
+      "epoch": 2.5342589797766375,
+      "grad_norm": 0.1484375,
+      "learning_rate": 3.200498132004982e-05,
+      "loss": 0.555,
+      "step": 4197
+    },
+    {
+      "epoch": 2.5348626622396617,
+      "grad_norm": 0.154296875,
+      "learning_rate": 3.1963470319634704e-05,
+      "loss": 0.5677,
+      "step": 4198
+    },
+    {
+      "epoch": 2.5354663447026864,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 3.19219593192196e-05,
+      "loss": 0.5287,
+      "step": 4199
+    },
+    {
+      "epoch": 2.536070027165711,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.1880448318804484e-05,
+      "loss": 0.5441,
+      "step": 4200
+    },
+    {
+      "epoch": 2.5366737096287353,
+      "grad_norm": 0.16015625,
+      "learning_rate": 3.183893731838937e-05,
+      "loss": 0.4927,
+      "step": 4201
+    },
+    {
+      "epoch": 2.5372773920917595,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 3.1797426317974264e-05,
+      "loss": 0.4615,
+      "step": 4202
+    },
+    {
+      "epoch": 2.537881074554784,
+      "grad_norm": 0.16015625,
+      "learning_rate": 3.175591531755916e-05,
+      "loss": 0.4678,
+      "step": 4203
+    },
+    {
+      "epoch": 2.538484757017809,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 3.171440431714404e-05,
+      "loss": 0.48,
+      "step": 4204
+    },
+    {
+      "epoch": 2.539088439480833,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.1672893316728937e-05,
+      "loss": 0.5278,
+      "step": 4205
+    },
+    {
+      "epoch": 2.5396921219438573,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.163138231631382e-05,
+      "loss": 0.4566,
+      "step": 4206
+    },
+    {
+      "epoch": 2.540295804406882,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.1589871315898716e-05,
+      "loss": 0.4041,
+      "step": 4207
+    },
+    {
+      "epoch": 2.5408994868699066,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.15483603154836e-05,
+      "loss": 0.4396,
+      "step": 4208
+    },
+    {
+      "epoch": 2.541503169332931,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.1506849315068496e-05,
+      "loss": 0.4315,
+      "step": 4209
+    },
+    {
+      "epoch": 2.542106851795955,
+      "grad_norm": 0.201171875,
+      "learning_rate": 3.146533831465339e-05,
+      "loss": 0.318,
+      "step": 4210
+    },
+    {
+      "epoch": 2.5427105342589797,
+      "grad_norm": 0.201171875,
+      "learning_rate": 3.1423827314238276e-05,
+      "loss": 0.2812,
+      "step": 4211
+    },
+    {
+      "epoch": 2.5433142167220044,
+      "grad_norm": 0.1953125,
+      "learning_rate": 3.138231631382316e-05,
+      "loss": 0.2069,
+      "step": 4212
+    },
+    {
+      "epoch": 2.5439178991850286,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 3.1340805313408056e-05,
+      "loss": 0.5955,
+      "step": 4213
+    },
+    {
+      "epoch": 2.544521581648053,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 3.129929431299294e-05,
+      "loss": 0.5779,
+      "step": 4214
+    },
+    {
+      "epoch": 2.5451252641110775,
+      "grad_norm": 0.142578125,
+      "learning_rate": 3.125778331257783e-05,
+      "loss": 0.6802,
+      "step": 4215
+    },
+    {
+      "epoch": 2.545728946574102,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 3.121627231216273e-05,
+      "loss": 0.8038,
+      "step": 4216
+    },
+    {
+      "epoch": 2.5463326290371264,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 3.1174761311747615e-05,
+      "loss": 0.8096,
+      "step": 4217
+    },
+    {
+      "epoch": 2.5469363115001507,
+      "grad_norm": 0.140625,
+      "learning_rate": 3.113325031133251e-05,
+      "loss": 1.203,
+      "step": 4218
+    },
+    {
+      "epoch": 2.5475399939631753,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 3.1091739310917395e-05,
+      "loss": 0.7647,
+      "step": 4219
+    },
+    {
+      "epoch": 2.5481436764262,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 3.105022831050228e-05,
+      "loss": 0.5275,
+      "step": 4220
+    },
+    {
+      "epoch": 2.548747358889224,
+      "grad_norm": 0.13671875,
+      "learning_rate": 3.1008717310087175e-05,
+      "loss": 0.658,
+      "step": 4221
+    },
+    {
+      "epoch": 2.549351041352249,
+      "grad_norm": 0.150390625,
+      "learning_rate": 3.096720630967206e-05,
+      "loss": 0.6599,
+      "step": 4222
+    },
+    {
+      "epoch": 2.549954723815273,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 3.0925695309256954e-05,
+      "loss": 0.6281,
+      "step": 4223
+    },
+    {
+      "epoch": 2.5505584062782978,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 3.088418430884185e-05,
+      "loss": 0.6384,
+      "step": 4224
+    },
+    {
+      "epoch": 2.551162088741322,
+      "grad_norm": 0.14453125,
+      "learning_rate": 3.0842673308426734e-05,
+      "loss": 0.6109,
+      "step": 4225
+    },
+    {
+      "epoch": 2.5517657712043467,
+      "grad_norm": 0.154296875,
+      "learning_rate": 3.080116230801162e-05,
+      "loss": 0.7078,
+      "step": 4226
+    },
+    {
+      "epoch": 2.552369453667371,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 3.0759651307596514e-05,
+      "loss": 0.5598,
+      "step": 4227
+    },
+    {
+      "epoch": 2.5529731361303956,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 3.07181403071814e-05,
+      "loss": 0.8109,
+      "step": 4228
+    },
+    {
+      "epoch": 2.55357681859342,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.06766293067663e-05,
+      "loss": 0.6628,
+      "step": 4229
+    },
+    {
+      "epoch": 2.5541805010564445,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 3.063511830635119e-05,
+      "loss": 0.6804,
+      "step": 4230
+    },
+    {
+      "epoch": 2.5547841835194687,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 3.059360730593607e-05,
+      "loss": 0.5877,
+      "step": 4231
+    },
+    {
+      "epoch": 2.5553878659824933,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 3.0552096305520966e-05,
+      "loss": 0.656,
+      "step": 4232
+    },
+    {
+      "epoch": 2.5559915484455176,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 3.0510585305105853e-05,
+      "loss": 0.676,
+      "step": 4233
+    },
+    {
+      "epoch": 2.5565952309085422,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 3.0469074304690743e-05,
+      "loss": 0.5534,
+      "step": 4234
+    },
+    {
+      "epoch": 2.5571989133715665,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 3.0427563304275636e-05,
+      "loss": 0.6103,
+      "step": 4235
+    },
+    {
+      "epoch": 2.557802595834591,
+      "grad_norm": 0.150390625,
+      "learning_rate": 3.0386052303860523e-05,
+      "loss": 0.6122,
+      "step": 4236
+    },
+    {
+      "epoch": 2.5584062782976154,
+      "grad_norm": 0.14453125,
+      "learning_rate": 3.0344541303445412e-05,
+      "loss": 0.6039,
+      "step": 4237
+    },
+    {
+      "epoch": 2.55900996076064,
+      "grad_norm": 0.185546875,
+      "learning_rate": 3.0303030303030306e-05,
+      "loss": 0.5834,
+      "step": 4238
+    },
+    {
+      "epoch": 2.5596136432236642,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 3.0261519302615192e-05,
+      "loss": 0.6298,
+      "step": 4239
+    },
+    {
+      "epoch": 2.560217325686689,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 3.0220008302200085e-05,
+      "loss": 0.5479,
+      "step": 4240
+    },
+    {
+      "epoch": 2.560821008149713,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 3.0178497301784975e-05,
+      "loss": 0.5281,
+      "step": 4241
+    },
+    {
+      "epoch": 2.561424690612738,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.0136986301369862e-05,
+      "loss": 0.7109,
+      "step": 4242
+    },
+    {
+      "epoch": 2.562028373075762,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 3.0095475300954755e-05,
+      "loss": 0.6514,
+      "step": 4243
+    },
+    {
+      "epoch": 2.5626320555387867,
+      "grad_norm": 0.162109375,
+      "learning_rate": 3.0053964300539645e-05,
+      "loss": 1.1654,
+      "step": 4244
+    },
+    {
+      "epoch": 2.563235738001811,
+      "grad_norm": 0.158203125,
+      "learning_rate": 3.001245330012453e-05,
+      "loss": 0.5638,
+      "step": 4245
+    },
+    {
+      "epoch": 2.5638394204648356,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 2.9970942299709425e-05,
+      "loss": 0.5556,
+      "step": 4246
+    },
+    {
+      "epoch": 2.56444310292786,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 2.9929431299294315e-05,
+      "loss": 0.6126,
+      "step": 4247
+    },
+    {
+      "epoch": 2.5650467853908845,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.98879202988792e-05,
+      "loss": 0.5597,
+      "step": 4248
+    },
+    {
+      "epoch": 2.5656504678539087,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 2.9846409298464094e-05,
+      "loss": 0.5701,
+      "step": 4249
+    },
+    {
+      "epoch": 2.5662541503169334,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.9804898298048984e-05,
+      "loss": 0.5262,
+      "step": 4250
+    },
+    {
+      "epoch": 2.5668578327799576,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.9763387297633877e-05,
+      "loss": 0.6605,
+      "step": 4251
+    },
+    {
+      "epoch": 2.5674615152429823,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.9721876297218764e-05,
+      "loss": 0.5195,
+      "step": 4252
+    },
+    {
+      "epoch": 2.5680651977060065,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.9680365296803654e-05,
+      "loss": 0.5135,
+      "step": 4253
+    },
+    {
+      "epoch": 2.568668880169031,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.9638854296388547e-05,
+      "loss": 0.55,
+      "step": 4254
+    },
+    {
+      "epoch": 2.5692725626320554,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.9597343295973434e-05,
+      "loss": 0.4785,
+      "step": 4255
+    },
+    {
+      "epoch": 2.56987624509508,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.9555832295558323e-05,
+      "loss": 0.4183,
+      "step": 4256
+    },
+    {
+      "epoch": 2.5704799275581043,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.9514321295143217e-05,
+      "loss": 0.4437,
+      "step": 4257
+    },
+    {
+      "epoch": 2.571083610021129,
+      "grad_norm": 0.19140625,
+      "learning_rate": 2.9472810294728103e-05,
+      "loss": 0.4089,
+      "step": 4258
+    },
+    {
+      "epoch": 2.571687292484153,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.9431299294312993e-05,
+      "loss": 0.355,
+      "step": 4259
+    },
+    {
+      "epoch": 2.572290974947178,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.9389788293897886e-05,
+      "loss": 0.4149,
+      "step": 4260
+    },
+    {
+      "epoch": 2.572894657410202,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.9348277293482773e-05,
+      "loss": 0.2763,
+      "step": 4261
+    },
+    {
+      "epoch": 2.5734983398732267,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.9306766293067666e-05,
+      "loss": 0.2371,
+      "step": 4262
+    },
+    {
+      "epoch": 2.574102022336251,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.9265255292652556e-05,
+      "loss": 0.7012,
+      "step": 4263
+    },
+    {
+      "epoch": 2.5747057047992756,
+      "grad_norm": 0.302734375,
+      "learning_rate": 2.9223744292237442e-05,
+      "loss": 0.7176,
+      "step": 4264
+    },
+    {
+      "epoch": 2.5753093872623003,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 2.9182233291822336e-05,
+      "loss": 0.6191,
+      "step": 4265
+    },
+    {
+      "epoch": 2.5759130697253245,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.9140722291407226e-05,
+      "loss": 0.7192,
+      "step": 4266
+    },
+    {
+      "epoch": 2.5765167521883487,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.9099211290992112e-05,
+      "loss": 0.6505,
+      "step": 4267
+    },
+    {
+      "epoch": 2.5771204346513734,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 2.9057700290577005e-05,
+      "loss": 0.6055,
+      "step": 4268
+    },
+    {
+      "epoch": 2.577724117114398,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 2.9016189290161895e-05,
+      "loss": 0.7176,
+      "step": 4269
+    },
+    {
+      "epoch": 2.5783277995774223,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.897467828974678e-05,
+      "loss": 0.7773,
+      "step": 4270
+    },
+    {
+      "epoch": 2.5789314820404465,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 2.8933167289331675e-05,
+      "loss": 0.5553,
+      "step": 4271
+    },
+    {
+      "epoch": 2.579535164503471,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.8891656288916565e-05,
+      "loss": 0.7196,
+      "step": 4272
+    },
+    {
+      "epoch": 2.580138846966496,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.8850145288501458e-05,
+      "loss": 0.5476,
+      "step": 4273
+    },
+    {
+      "epoch": 2.58074252942952,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.8808634288086344e-05,
+      "loss": 0.6434,
+      "step": 4274
+    },
+    {
+      "epoch": 2.5813462118925443,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.8767123287671234e-05,
+      "loss": 0.8742,
+      "step": 4275
+    },
+    {
+      "epoch": 2.581949894355569,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.8725612287256128e-05,
+      "loss": 0.6174,
+      "step": 4276
+    },
+    {
+      "epoch": 2.5825535768185937,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.8684101286841014e-05,
+      "loss": 0.6064,
+      "step": 4277
+    },
+    {
+      "epoch": 2.583157259281618,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.86425902864259e-05,
+      "loss": 0.5538,
+      "step": 4278
+    },
+    {
+      "epoch": 2.583760941744642,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.8601079286010797e-05,
+      "loss": 0.5924,
+      "step": 4279
+    },
+    {
+      "epoch": 2.5843646242076668,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.8559568285595684e-05,
+      "loss": 0.5754,
+      "step": 4280
+    },
+    {
+      "epoch": 2.5849683066706914,
+      "grad_norm": 0.142578125,
+      "learning_rate": 2.851805728518057e-05,
+      "loss": 0.5529,
+      "step": 4281
+    },
+    {
+      "epoch": 2.5855719891337157,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.8476546284765467e-05,
+      "loss": 0.7275,
+      "step": 4282
+    },
+    {
+      "epoch": 2.58617567159674,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.8435035284350353e-05,
+      "loss": 0.6515,
+      "step": 4283
+    },
+    {
+      "epoch": 2.5867793540597646,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.8393524283935247e-05,
+      "loss": 0.6439,
+      "step": 4284
+    },
+    {
+      "epoch": 2.5873830365227892,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.8352013283520133e-05,
+      "loss": 0.5381,
+      "step": 4285
+    },
+    {
+      "epoch": 2.5879867189858135,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 2.8310502283105023e-05,
+      "loss": 0.865,
+      "step": 4286
+    },
+    {
+      "epoch": 2.5885904014488377,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.8268991282689916e-05,
+      "loss": 0.7643,
+      "step": 4287
+    },
+    {
+      "epoch": 2.5891940839118623,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.8227480282274803e-05,
+      "loss": 0.6419,
+      "step": 4288
+    },
+    {
+      "epoch": 2.589797766374887,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.8185969281859693e-05,
+      "loss": 0.6687,
+      "step": 4289
+    },
+    {
+      "epoch": 2.5904014488379112,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.8144458281444586e-05,
+      "loss": 0.7835,
+      "step": 4290
+    },
+    {
+      "epoch": 2.5910051313009355,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.8102947281029472e-05,
+      "loss": 0.6415,
+      "step": 4291
+    },
+    {
+      "epoch": 2.59160881376396,
+      "grad_norm": 0.154296875,
+      "learning_rate": 2.8061436280614362e-05,
+      "loss": 0.7106,
+      "step": 4292
+    },
+    {
+      "epoch": 2.592212496226985,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.8019925280199255e-05,
+      "loss": 0.617,
+      "step": 4293
+    },
+    {
+      "epoch": 2.592816178690009,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 2.7978414279784142e-05,
+      "loss": 0.595,
+      "step": 4294
+    },
+    {
+      "epoch": 2.5934198611530332,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 2.7936903279369035e-05,
+      "loss": 0.5337,
+      "step": 4295
+    },
+    {
+      "epoch": 2.594023543616058,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.7895392278953925e-05,
+      "loss": 0.5098,
+      "step": 4296
+    },
+    {
+      "epoch": 2.5946272260790826,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.785388127853881e-05,
+      "loss": 0.519,
+      "step": 4297
+    },
+    {
+      "epoch": 2.595230908542107,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.7812370278123705e-05,
+      "loss": 0.5765,
+      "step": 4298
+    },
+    {
+      "epoch": 2.595834591005131,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 2.7770859277708595e-05,
+      "loss": 0.5621,
+      "step": 4299
+    },
+    {
+      "epoch": 2.5964382734681557,
+      "grad_norm": 0.154296875,
+      "learning_rate": 2.772934827729348e-05,
+      "loss": 0.5247,
+      "step": 4300
+    },
+    {
+      "epoch": 2.5970419559311804,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 2.7687837276878374e-05,
+      "loss": 0.455,
+      "step": 4301
+    },
+    {
+      "epoch": 2.5976456383942046,
+      "grad_norm": 0.154296875,
+      "learning_rate": 2.7646326276463264e-05,
+      "loss": 0.4566,
+      "step": 4302
+    },
+    {
+      "epoch": 2.5982493208572293,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.760481527604815e-05,
+      "loss": 0.4368,
+      "step": 4303
+    },
+    {
+      "epoch": 2.5988530033202535,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.7563304275633044e-05,
+      "loss": 0.5235,
+      "step": 4304
+    },
+    {
+      "epoch": 2.599456685783278,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.7521793275217934e-05,
+      "loss": 0.4767,
+      "step": 4305
+    },
+    {
+      "epoch": 2.6000603682463024,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 2.7480282274802827e-05,
+      "loss": 0.3947,
+      "step": 4306
+    },
+    {
+      "epoch": 2.600664050709327,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 2.7438771274387714e-05,
+      "loss": 0.3915,
+      "step": 4307
+    },
+    {
+      "epoch": 2.6012677331723513,
+      "grad_norm": 0.1953125,
+      "learning_rate": 2.7397260273972603e-05,
+      "loss": 0.3913,
+      "step": 4308
+    },
+    {
+      "epoch": 2.601871415635376,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.7355749273557497e-05,
+      "loss": 0.3407,
+      "step": 4309
+    },
+    {
+      "epoch": 2.6024750980984,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.7314238273142383e-05,
+      "loss": 0.3446,
+      "step": 4310
+    },
+    {
+      "epoch": 2.603078780561425,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.7272727272727273e-05,
+      "loss": 0.3252,
+      "step": 4311
+    },
+    {
+      "epoch": 2.603682463024449,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.7231216272312166e-05,
+      "loss": 0.1949,
+      "step": 4312
+    },
+    {
+      "epoch": 2.6042861454874737,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.7189705271897053e-05,
+      "loss": 0.6236,
+      "step": 4313
+    },
+    {
+      "epoch": 2.604889827950498,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.7148194271481943e-05,
+      "loss": 0.5858,
+      "step": 4314
+    },
+    {
+      "epoch": 2.6054935104135226,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 2.7106683271066836e-05,
+      "loss": 0.6003,
+      "step": 4315
+    },
+    {
+      "epoch": 2.606097192876547,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 2.7065172270651722e-05,
+      "loss": 0.5897,
+      "step": 4316
+    },
+    {
+      "epoch": 2.6067008753395715,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.7023661270236616e-05,
+      "loss": 0.6271,
+      "step": 4317
+    },
+    {
+      "epoch": 2.6073045578025957,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 2.6982150269821506e-05,
+      "loss": 0.5715,
+      "step": 4318
+    },
+    {
+      "epoch": 2.6079082402656204,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.6940639269406392e-05,
+      "loss": 0.5873,
+      "step": 4319
+    },
+    {
+      "epoch": 2.6085119227286446,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 2.6899128268991285e-05,
+      "loss": 0.5097,
+      "step": 4320
+    },
+    {
+      "epoch": 2.6091156051916693,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 2.6857617268576175e-05,
+      "loss": 0.8605,
+      "step": 4321
+    },
+    {
+      "epoch": 2.6097192876546935,
+      "grad_norm": 0.154296875,
+      "learning_rate": 2.6816106268161062e-05,
+      "loss": 0.6028,
+      "step": 4322
+    },
+    {
+      "epoch": 2.610322970117718,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.6774595267745955e-05,
+      "loss": 0.6827,
+      "step": 4323
+    },
+    {
+      "epoch": 2.6109266525807424,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.6733084267330845e-05,
+      "loss": 0.6564,
+      "step": 4324
+    },
+    {
+      "epoch": 2.611530335043767,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 2.669157326691573e-05,
+      "loss": 0.5687,
+      "step": 4325
+    },
+    {
+      "epoch": 2.6121340175067913,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.6650062266500625e-05,
+      "loss": 0.856,
+      "step": 4326
+    },
+    {
+      "epoch": 2.612737699969816,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.6608551266085514e-05,
+      "loss": 0.6517,
+      "step": 4327
+    },
+    {
+      "epoch": 2.61334138243284,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.6567040265670408e-05,
+      "loss": 0.6049,
+      "step": 4328
+    },
+    {
+      "epoch": 2.613945064895865,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.6525529265255294e-05,
+      "loss": 0.6425,
+      "step": 4329
+    },
+    {
+      "epoch": 2.614548747358889,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.6484018264840184e-05,
+      "loss": 0.6713,
+      "step": 4330
+    },
+    {
+      "epoch": 2.6151524298219138,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.6442507264425077e-05,
+      "loss": 0.7973,
+      "step": 4331
+    },
+    {
+      "epoch": 2.615756112284938,
+      "grad_norm": 0.158203125,
+      "learning_rate": 2.6400996264009964e-05,
+      "loss": 0.8084,
+      "step": 4332
+    },
+    {
+      "epoch": 2.6163597947479627,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.635948526359485e-05,
+      "loss": 0.7568,
+      "step": 4333
+    },
+    {
+      "epoch": 2.616963477210987,
+      "grad_norm": 0.142578125,
+      "learning_rate": 2.6317974263179747e-05,
+      "loss": 0.6364,
+      "step": 4334
+    },
+    {
+      "epoch": 2.6175671596740115,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.6276463262764633e-05,
+      "loss": 0.7264,
+      "step": 4335
+    },
+    {
+      "epoch": 2.6181708421370358,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 2.623495226234952e-05,
+      "loss": 0.6437,
+      "step": 4336
+    },
+    {
+      "epoch": 2.6187745246000604,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 2.6193441261934417e-05,
+      "loss": 0.9275,
+      "step": 4337
+    },
+    {
+      "epoch": 2.6193782070630847,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.6151930261519303e-05,
+      "loss": 0.6317,
+      "step": 4338
+    },
+    {
+      "epoch": 2.6199818895261093,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.6110419261104196e-05,
+      "loss": 0.5622,
+      "step": 4339
+    },
+    {
+      "epoch": 2.6205855719891336,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.6068908260689083e-05,
+      "loss": 0.6879,
+      "step": 4340
+    },
+    {
+      "epoch": 2.6211892544521582,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 2.6027397260273973e-05,
+      "loss": 0.8795,
+      "step": 4341
+    },
+    {
+      "epoch": 2.6217929369151824,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 2.5985886259858866e-05,
+      "loss": 0.5918,
+      "step": 4342
+    },
+    {
+      "epoch": 2.622396619378207,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.5944375259443752e-05,
+      "loss": 0.5913,
+      "step": 4343
+    },
+    {
+      "epoch": 2.6230003018412313,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.5902864259028642e-05,
+      "loss": 0.7423,
+      "step": 4344
+    },
+    {
+      "epoch": 2.623603984304256,
+      "grad_norm": 0.14453125,
+      "learning_rate": 2.5861353258613536e-05,
+      "loss": 0.6315,
+      "step": 4345
+    },
+    {
+      "epoch": 2.6242076667672807,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.5819842258198422e-05,
+      "loss": 0.5559,
+      "step": 4346
+    },
+    {
+      "epoch": 2.624811349230305,
+      "grad_norm": 0.142578125,
+      "learning_rate": 2.5778331257783312e-05,
+      "loss": 0.4504,
+      "step": 4347
+    },
+    {
+      "epoch": 2.625415031693329,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.5736820257368205e-05,
+      "loss": 0.5707,
+      "step": 4348
+    },
+    {
+      "epoch": 2.626018714156354,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.569530925695309e-05,
+      "loss": 0.5686,
+      "step": 4349
+    },
+    {
+      "epoch": 2.6266223966193785,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.5653798256537985e-05,
+      "loss": 0.5619,
+      "step": 4350
+    },
+    {
+      "epoch": 2.6272260790824027,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.5612287256122875e-05,
+      "loss": 0.467,
+      "step": 4351
+    },
+    {
+      "epoch": 2.627829761545427,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.557077625570776e-05,
+      "loss": 0.5095,
+      "step": 4352
+    },
+    {
+      "epoch": 2.6284334440084516,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.5529265255292654e-05,
+      "loss": 0.5222,
+      "step": 4353
+    },
+    {
+      "epoch": 2.6290371264714762,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.5487754254877544e-05,
+      "loss": 0.5631,
+      "step": 4354
+    },
+    {
+      "epoch": 2.6296408089345005,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.544624325446243e-05,
+      "loss": 0.4773,
+      "step": 4355
+    },
+    {
+      "epoch": 2.6302444913975247,
+      "grad_norm": 0.1953125,
+      "learning_rate": 2.5404732254047324e-05,
+      "loss": 0.5103,
+      "step": 4356
+    },
+    {
+      "epoch": 2.6308481738605494,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.5363221253632214e-05,
+      "loss": 0.4466,
+      "step": 4357
+    },
+    {
+      "epoch": 2.631451856323574,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.53217102532171e-05,
+      "loss": 0.41,
+      "step": 4358
+    },
+    {
+      "epoch": 2.6320555387865983,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.5280199252801994e-05,
+      "loss": 0.3226,
+      "step": 4359
+    },
+    {
+      "epoch": 2.6326592212496225,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 2.5238688252386884e-05,
+      "loss": 0.2631,
+      "step": 4360
+    },
+    {
+      "epoch": 2.633262903712647,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.5197177251971777e-05,
+      "loss": 0.2883,
+      "step": 4361
+    },
+    {
+      "epoch": 2.633866586175672,
+      "grad_norm": 0.224609375,
+      "learning_rate": 2.5155666251556663e-05,
+      "loss": 0.2396,
+      "step": 4362
+    },
+    {
+      "epoch": 2.634470268638696,
+      "grad_norm": 0.14453125,
+      "learning_rate": 2.5114155251141553e-05,
+      "loss": 0.6584,
+      "step": 4363
+    },
+    {
+      "epoch": 2.6350739511017203,
+      "grad_norm": 0.130859375,
+      "learning_rate": 2.5072644250726446e-05,
+      "loss": 0.5349,
+      "step": 4364
+    },
+    {
+      "epoch": 2.635677633564745,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.5031133250311333e-05,
+      "loss": 0.6576,
+      "step": 4365
+    },
+    {
+      "epoch": 2.6362813160277696,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.4989622249896226e-05,
+      "loss": 0.61,
+      "step": 4366
+    },
+    {
+      "epoch": 2.636884998490794,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.4948111249481113e-05,
+      "loss": 0.7938,
+      "step": 4367
+    },
+    {
+      "epoch": 2.637488680953818,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.4906600249066003e-05,
+      "loss": 0.5602,
+      "step": 4368
+    },
+    {
+      "epoch": 2.6380923634168427,
+      "grad_norm": 0.158203125,
+      "learning_rate": 2.4865089248650892e-05,
+      "loss": 0.6619,
+      "step": 4369
+    },
+    {
+      "epoch": 2.6386960458798674,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 2.4823578248235786e-05,
+      "loss": 0.6877,
+      "step": 4370
+    },
+    {
+      "epoch": 2.6392997283428916,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 2.4782067247820672e-05,
+      "loss": 0.5548,
+      "step": 4371
+    },
+    {
+      "epoch": 2.639903410805916,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.4740556247405562e-05,
+      "loss": 0.5933,
+      "step": 4372
+    },
+    {
+      "epoch": 2.6405070932689405,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.4699045246990455e-05,
+      "loss": 0.6843,
+      "step": 4373
+    },
+    {
+      "epoch": 2.641110775731965,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.4657534246575342e-05,
+      "loss": 0.6951,
+      "step": 4374
+    },
+    {
+      "epoch": 2.6417144581949894,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.461602324616023e-05,
+      "loss": 0.6361,
+      "step": 4375
+    },
+    {
+      "epoch": 2.6423181406580136,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.4574512245745125e-05,
+      "loss": 0.6971,
+      "step": 4376
+    },
+    {
+      "epoch": 2.6429218231210383,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 2.4533001245330015e-05,
+      "loss": 0.6643,
+      "step": 4377
+    },
+    {
+      "epoch": 2.643525505584063,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 2.44914902449149e-05,
+      "loss": 0.5958,
+      "step": 4378
+    },
+    {
+      "epoch": 2.644129188047087,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 2.4449979244499795e-05,
+      "loss": 0.5614,
+      "step": 4379
+    },
+    {
+      "epoch": 2.6447328705101114,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.4408468244084684e-05,
+      "loss": 0.6388,
+      "step": 4380
+    },
+    {
+      "epoch": 2.645336552973136,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.4366957243669574e-05,
+      "loss": 0.6615,
+      "step": 4381
+    },
+    {
+      "epoch": 2.6459402354361607,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.4325446243254464e-05,
+      "loss": 0.5931,
+      "step": 4382
+    },
+    {
+      "epoch": 2.646543917899185,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.4283935242839354e-05,
+      "loss": 0.618,
+      "step": 4383
+    },
+    {
+      "epoch": 2.6471476003622096,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.4242424242424244e-05,
+      "loss": 0.7889,
+      "step": 4384
+    },
+    {
+      "epoch": 2.647751282825234,
+      "grad_norm": 0.158203125,
+      "learning_rate": 2.4200913242009134e-05,
+      "loss": 0.5992,
+      "step": 4385
+    },
+    {
+      "epoch": 2.6483549652882585,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 2.4159402241594024e-05,
+      "loss": 0.6226,
+      "step": 4386
+    },
+    {
+      "epoch": 2.6489586477512828,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.4117891241178914e-05,
+      "loss": 0.6265,
+      "step": 4387
+    },
+    {
+      "epoch": 2.6495623302143074,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.4076380240763803e-05,
+      "loss": 0.5859,
+      "step": 4388
+    },
+    {
+      "epoch": 2.6501660126773317,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.4034869240348693e-05,
+      "loss": 0.6647,
+      "step": 4389
+    },
+    {
+      "epoch": 2.6507696951403563,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.3993358239933583e-05,
+      "loss": 0.6158,
+      "step": 4390
+    },
+    {
+      "epoch": 2.6513733776033805,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.3951847239518473e-05,
+      "loss": 0.7374,
+      "step": 4391
+    },
+    {
+      "epoch": 2.651977060066405,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.3910336239103366e-05,
+      "loss": 0.7737,
+      "step": 4392
+    },
+    {
+      "epoch": 2.6525807425294294,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.3868825238688253e-05,
+      "loss": 0.5964,
+      "step": 4393
+    },
+    {
+      "epoch": 2.653184424992454,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.3827314238273143e-05,
+      "loss": 0.7454,
+      "step": 4394
+    },
+    {
+      "epoch": 2.6537881074554783,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.3785803237858036e-05,
+      "loss": 0.5554,
+      "step": 4395
+    },
+    {
+      "epoch": 2.654391789918503,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 2.3744292237442922e-05,
+      "loss": 0.6501,
+      "step": 4396
+    },
+    {
+      "epoch": 2.654995472381527,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.3702781237027812e-05,
+      "loss": 0.5483,
+      "step": 4397
+    },
+    {
+      "epoch": 2.655599154844552,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.3661270236612702e-05,
+      "loss": 0.5345,
+      "step": 4398
+    },
+    {
+      "epoch": 2.656202837307576,
+      "grad_norm": 0.142578125,
+      "learning_rate": 2.3619759236197595e-05,
+      "loss": 0.4837,
+      "step": 4399
+    },
+    {
+      "epoch": 2.656806519770601,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.3578248235782482e-05,
+      "loss": 0.5075,
+      "step": 4400
+    },
+    {
+      "epoch": 2.657410202233625,
+      "grad_norm": 0.158203125,
+      "learning_rate": 2.3536737235367372e-05,
+      "loss": 0.5213,
+      "step": 4401
+    },
+    {
+      "epoch": 2.6580138846966497,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 2.3495226234952265e-05,
+      "loss": 0.488,
+      "step": 4402
+    },
+    {
+      "epoch": 2.658617567159674,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.3453715234537155e-05,
+      "loss": 0.5311,
+      "step": 4403
+    },
+    {
+      "epoch": 2.6592212496226986,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.341220423412204e-05,
+      "loss": 0.4742,
+      "step": 4404
+    },
+    {
+      "epoch": 2.659824932085723,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.3370693233706935e-05,
+      "loss": 0.498,
+      "step": 4405
+    },
+    {
+      "epoch": 2.6604286145487475,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.3329182233291824e-05,
+      "loss": 0.4055,
+      "step": 4406
+    },
+    {
+      "epoch": 2.6610322970117717,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.328767123287671e-05,
+      "loss": 0.3957,
+      "step": 4407
+    },
+    {
+      "epoch": 2.6616359794747964,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.3246160232461604e-05,
+      "loss": 0.3804,
+      "step": 4408
+    },
+    {
+      "epoch": 2.6622396619378206,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.3204649232046494e-05,
+      "loss": 0.337,
+      "step": 4409
+    },
+    {
+      "epoch": 2.6628433444008452,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.3163138231631384e-05,
+      "loss": 0.3131,
+      "step": 4410
+    },
+    {
+      "epoch": 2.6634470268638695,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.3121627231216274e-05,
+      "loss": 0.2671,
+      "step": 4411
+    },
+    {
+      "epoch": 2.664050709326894,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.3080116230801164e-05,
+      "loss": 0.2049,
+      "step": 4412
+    },
+    {
+      "epoch": 2.6646543917899184,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.3038605230386054e-05,
+      "loss": 0.7446,
+      "step": 4413
+    },
+    {
+      "epoch": 2.665258074252943,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.2997094229970943e-05,
+      "loss": 0.6451,
+      "step": 4414
+    },
+    {
+      "epoch": 2.6658617567159673,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.2955583229555833e-05,
+      "loss": 0.6584,
+      "step": 4415
+    },
+    {
+      "epoch": 2.666465439178992,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.2914072229140723e-05,
+      "loss": 0.595,
+      "step": 4416
+    },
+    {
+      "epoch": 2.667069121642016,
+      "grad_norm": 0.158203125,
+      "learning_rate": 2.2872561228725613e-05,
+      "loss": 0.7,
+      "step": 4417
+    },
+    {
+      "epoch": 2.667672804105041,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.2831050228310503e-05,
+      "loss": 0.6295,
+      "step": 4418
+    },
+    {
+      "epoch": 2.668276486568065,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 2.2789539227895393e-05,
+      "loss": 0.5823,
+      "step": 4419
+    },
+    {
+      "epoch": 2.6688801690310897,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.2748028227480283e-05,
+      "loss": 0.6256,
+      "step": 4420
+    },
+    {
+      "epoch": 2.669483851494114,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.2706517227065176e-05,
+      "loss": 0.6366,
+      "step": 4421
+    },
+    {
+      "epoch": 2.6700875339571386,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 2.2665006226650062e-05,
+      "loss": 0.6766,
+      "step": 4422
+    },
+    {
+      "epoch": 2.670691216420163,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.2623495226234952e-05,
+      "loss": 0.6968,
+      "step": 4423
+    },
+    {
+      "epoch": 2.6712948988831875,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.2581984225819846e-05,
+      "loss": 0.6291,
+      "step": 4424
+    },
+    {
+      "epoch": 2.6718985813462117,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 2.2540473225404735e-05,
+      "loss": 0.649,
+      "step": 4425
+    },
+    {
+      "epoch": 2.6725022638092364,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.2498962224989622e-05,
+      "loss": 0.6302,
+      "step": 4426
+    },
+    {
+      "epoch": 2.673105946272261,
+      "grad_norm": 0.154296875,
+      "learning_rate": 2.2457451224574512e-05,
+      "loss": 0.6207,
+      "step": 4427
+    },
+    {
+      "epoch": 2.6737096287352853,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.2415940224159405e-05,
+      "loss": 0.6337,
+      "step": 4428
+    },
+    {
+      "epoch": 2.6743133111983095,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.237442922374429e-05,
+      "loss": 0.655,
+      "step": 4429
+    },
+    {
+      "epoch": 2.674916993661334,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 2.233291822332918e-05,
+      "loss": 0.6017,
+      "step": 4430
+    },
+    {
+      "epoch": 2.675520676124359,
+      "grad_norm": 0.15625,
+      "learning_rate": 2.2291407222914075e-05,
+      "loss": 0.7195,
+      "step": 4431
+    },
+    {
+      "epoch": 2.676124358587383,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.2249896222498965e-05,
+      "loss": 0.4655,
+      "step": 4432
+    },
+    {
+      "epoch": 2.6767280410504073,
+      "grad_norm": 0.154296875,
+      "learning_rate": 2.220838522208385e-05,
+      "loss": 0.6356,
+      "step": 4433
+    },
+    {
+      "epoch": 2.677331723513432,
+      "grad_norm": 0.298828125,
+      "learning_rate": 2.2166874221668744e-05,
+      "loss": 0.6569,
+      "step": 4434
+    },
+    {
+      "epoch": 2.6779354059764566,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.2125363221253634e-05,
+      "loss": 0.654,
+      "step": 4435
+    },
+    {
+      "epoch": 2.678539088439481,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 2.2083852220838524e-05,
+      "loss": 0.5911,
+      "step": 4436
+    },
+    {
+      "epoch": 2.679142770902505,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.2042341220423414e-05,
+      "loss": 0.6751,
+      "step": 4437
+    },
+    {
+      "epoch": 2.6797464533655297,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.2000830220008304e-05,
+      "loss": 0.6674,
+      "step": 4438
+    },
+    {
+      "epoch": 2.6803501358285544,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 2.1959319219593194e-05,
+      "loss": 0.8804,
+      "step": 4439
+    },
+    {
+      "epoch": 2.6809538182915786,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.1917808219178083e-05,
+      "loss": 0.5718,
+      "step": 4440
+    },
+    {
+      "epoch": 2.681557500754603,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.1876297218762973e-05,
+      "loss": 0.6883,
+      "step": 4441
+    },
+    {
+      "epoch": 2.6821611832176275,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 2.1834786218347863e-05,
+      "loss": 0.7524,
+      "step": 4442
+    },
+    {
+      "epoch": 2.682764865680652,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 2.1793275217932753e-05,
+      "loss": 0.5331,
+      "step": 4443
+    },
+    {
+      "epoch": 2.6833685481436764,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 2.1751764217517643e-05,
+      "loss": 0.786,
+      "step": 4444
+    },
+    {
+      "epoch": 2.6839722306067006,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.1710253217102533e-05,
+      "loss": 0.6069,
+      "step": 4445
+    },
+    {
+      "epoch": 2.6845759130697253,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.1668742216687423e-05,
+      "loss": 0.6164,
+      "step": 4446
+    },
+    {
+      "epoch": 2.68517959553275,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 2.1627231216272316e-05,
+      "loss": 0.5618,
+      "step": 4447
+    },
+    {
+      "epoch": 2.685783277995774,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.1585720215857202e-05,
+      "loss": 0.535,
+      "step": 4448
+    },
+    {
+      "epoch": 2.6863869604587984,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.1544209215442092e-05,
+      "loss": 0.6048,
+      "step": 4449
+    },
+    {
+      "epoch": 2.686990642921823,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.1502698215026986e-05,
+      "loss": 0.5397,
+      "step": 4450
+    },
+    {
+      "epoch": 2.6875943253848478,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.1461187214611872e-05,
+      "loss": 0.5,
+      "step": 4451
+    },
+    {
+      "epoch": 2.688198007847872,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.1419676214196762e-05,
+      "loss": 0.514,
+      "step": 4452
+    },
+    {
+      "epoch": 2.688801690310896,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.1378165213781652e-05,
+      "loss": 0.4886,
+      "step": 4453
+    },
+    {
+      "epoch": 2.689405372773921,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.1336654213366545e-05,
+      "loss": 0.4477,
+      "step": 4454
+    },
+    {
+      "epoch": 2.6900090552369456,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.129514321295143e-05,
+      "loss": 0.5355,
+      "step": 4455
+    },
+    {
+      "epoch": 2.69061273769997,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.125363221253632e-05,
+      "loss": 0.414,
+      "step": 4456
+    },
+    {
+      "epoch": 2.691216420162994,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.1212121212121215e-05,
+      "loss": 0.4451,
+      "step": 4457
+    },
+    {
+      "epoch": 2.6918201026260187,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.1170610211706105e-05,
+      "loss": 0.4357,
+      "step": 4458
+    },
+    {
+      "epoch": 2.6924237850890433,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.112909921129099e-05,
+      "loss": 0.3608,
+      "step": 4459
+    },
+    {
+      "epoch": 2.6930274675520676,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.1087588210875884e-05,
+      "loss": 0.2876,
+      "step": 4460
+    },
+    {
+      "epoch": 2.693631150015092,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.1046077210460774e-05,
+      "loss": 0.2317,
+      "step": 4461
+    },
+    {
+      "epoch": 2.6942348324781165,
+      "grad_norm": 0.1953125,
+      "learning_rate": 2.100456621004566e-05,
+      "loss": 0.1918,
+      "step": 4462
+    },
+    {
+      "epoch": 2.694838514941141,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.0963055209630554e-05,
+      "loss": 0.6023,
+      "step": 4463
+    },
+    {
+      "epoch": 2.6954421974041654,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.0921544209215444e-05,
+      "loss": 0.6662,
+      "step": 4464
+    },
+    {
+      "epoch": 2.69604587986719,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 2.0880033208800334e-05,
+      "loss": 0.5831,
+      "step": 4465
+    },
+    {
+      "epoch": 2.6966495623302142,
+      "grad_norm": 0.158203125,
+      "learning_rate": 2.0838522208385224e-05,
+      "loss": 0.6397,
+      "step": 4466
+    },
+    {
+      "epoch": 2.697253244793239,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.0797011207970113e-05,
+      "loss": 0.6301,
+      "step": 4467
+    },
+    {
+      "epoch": 2.697856927256263,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.0755500207555003e-05,
+      "loss": 0.7089,
+      "step": 4468
+    },
+    {
+      "epoch": 2.698460609719288,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.0713989207139893e-05,
+      "loss": 0.519,
+      "step": 4469
+    },
+    {
+      "epoch": 2.699064292182312,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 2.0672478206724783e-05,
+      "loss": 0.6105,
+      "step": 4470
+    },
+    {
+      "epoch": 2.6996679746453367,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.0630967206309673e-05,
+      "loss": 0.7206,
+      "step": 4471
+    },
+    {
+      "epoch": 2.700271657108361,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.0589456205894563e-05,
+      "loss": 0.6736,
+      "step": 4472
+    },
+    {
+      "epoch": 2.7008753395713856,
+      "grad_norm": 0.14453125,
+      "learning_rate": 2.0547945205479453e-05,
+      "loss": 0.6061,
+      "step": 4473
+    },
+    {
+      "epoch": 2.70147902203441,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.0506434205064342e-05,
+      "loss": 0.5616,
+      "step": 4474
+    },
+    {
+      "epoch": 2.7020827044974345,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 2.0464923204649232e-05,
+      "loss": 0.6123,
+      "step": 4475
+    },
+    {
+      "epoch": 2.7026863869604587,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.0423412204234126e-05,
+      "loss": 0.8862,
+      "step": 4476
+    },
+    {
+      "epoch": 2.7032900694234834,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 2.0381901203819012e-05,
+      "loss": 0.7349,
+      "step": 4477
+    },
+    {
+      "epoch": 2.7038937518865076,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.0340390203403902e-05,
+      "loss": 0.6979,
+      "step": 4478
+    },
+    {
+      "epoch": 2.7044974343495323,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.0298879202988795e-05,
+      "loss": 0.6383,
+      "step": 4479
+    },
+    {
+      "epoch": 2.7051011168125565,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.0257368202573682e-05,
+      "loss": 0.6267,
+      "step": 4480
+    },
+    {
+      "epoch": 2.705704799275581,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.021585720215857e-05,
+      "loss": 0.6785,
+      "step": 4481
+    },
+    {
+      "epoch": 2.7063084817386054,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.017434620174346e-05,
+      "loss": 0.753,
+      "step": 4482
+    },
+    {
+      "epoch": 2.70691216420163,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.0132835201328355e-05,
+      "loss": 0.5688,
+      "step": 4483
+    },
+    {
+      "epoch": 2.7075158466646543,
+      "grad_norm": 0.142578125,
+      "learning_rate": 2.009132420091324e-05,
+      "loss": 0.7494,
+      "step": 4484
+    },
+    {
+      "epoch": 2.708119529127679,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.004981320049813e-05,
+      "loss": 0.74,
+      "step": 4485
+    },
+    {
+      "epoch": 2.708723211590703,
+      "grad_norm": 0.142578125,
+      "learning_rate": 2.0008302200083024e-05,
+      "loss": 0.5672,
+      "step": 4486
+    },
+    {
+      "epoch": 2.709326894053728,
+      "grad_norm": 0.15234375,
+      "learning_rate": 1.9966791199667914e-05,
+      "loss": 0.8192,
+      "step": 4487
+    },
+    {
+      "epoch": 2.709930576516752,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.99252801992528e-05,
+      "loss": 0.6851,
+      "step": 4488
+    },
+    {
+      "epoch": 2.7105342589797767,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.9883769198837694e-05,
+      "loss": 0.6071,
+      "step": 4489
+    },
+    {
+      "epoch": 2.711137941442801,
+      "grad_norm": 0.154296875,
+      "learning_rate": 1.9842258198422584e-05,
+      "loss": 0.6756,
+      "step": 4490
+    },
+    {
+      "epoch": 2.7117416239058256,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.980074719800747e-05,
+      "loss": 0.6658,
+      "step": 4491
+    },
+    {
+      "epoch": 2.71234530636885,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.9759236197592364e-05,
+      "loss": 0.5517,
+      "step": 4492
+    },
+    {
+      "epoch": 2.7129489888318745,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.9717725197177253e-05,
+      "loss": 0.6189,
+      "step": 4493
+    },
+    {
+      "epoch": 2.7135526712948987,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.9676214196762143e-05,
+      "loss": 0.5697,
+      "step": 4494
+    },
+    {
+      "epoch": 2.7141563537579234,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.9634703196347033e-05,
+      "loss": 0.6535,
+      "step": 4495
+    },
+    {
+      "epoch": 2.7147600362209476,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.9593192195931923e-05,
+      "loss": 0.5543,
+      "step": 4496
+    },
+    {
+      "epoch": 2.7153637186839723,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 1.9551681195516813e-05,
+      "loss": 0.5723,
+      "step": 4497
+    },
+    {
+      "epoch": 2.7159674011469965,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.9510170195101703e-05,
+      "loss": 0.5723,
+      "step": 4498
+    },
+    {
+      "epoch": 2.716571083610021,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.9468659194686593e-05,
+      "loss": 0.5244,
+      "step": 4499
+    },
+    {
+      "epoch": 2.7171747660730454,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.9427148194271483e-05,
+      "loss": 0.5962,
+      "step": 4500
+    },
+    {
+      "epoch": 2.7171747660730454,
+      "eval_loss": 0.6135643124580383,
+      "eval_runtime": 1059.4817,
+      "eval_samples_per_second": 2.633,
+      "eval_steps_per_second": 0.329,
+      "step": 4500
+    },
+    {
+      "epoch": 2.71777844853607,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.9385637193856372e-05,
+      "loss": 0.5302,
+      "step": 4501
+    },
+    {
+      "epoch": 2.7183821309990943,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.9344126193441262e-05,
+      "loss": 0.5348,
+      "step": 4502
+    },
+    {
+      "epoch": 2.718985813462119,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.9302615193026152e-05,
+      "loss": 0.5221,
+      "step": 4503
+    },
+    {
+      "epoch": 2.7195894959251437,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.9261104192611042e-05,
+      "loss": 0.4312,
+      "step": 4504
+    },
+    {
+      "epoch": 2.720193178388168,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.9219593192195935e-05,
+      "loss": 0.4838,
+      "step": 4505
+    },
+    {
+      "epoch": 2.720796860851192,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.9178082191780822e-05,
+      "loss": 0.471,
+      "step": 4506
+    },
+    {
+      "epoch": 2.7214005433142168,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.913657119136571e-05,
+      "loss": 0.4915,
+      "step": 4507
+    },
+    {
+      "epoch": 2.7220042257772414,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.9095060190950605e-05,
+      "loss": 0.3882,
+      "step": 4508
+    },
+    {
+      "epoch": 2.7226079082402657,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.9053549190535495e-05,
+      "loss": 0.3019,
+      "step": 4509
+    },
+    {
+      "epoch": 2.72321159070329,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 1.901203819012038e-05,
+      "loss": 0.3022,
+      "step": 4510
+    },
+    {
+      "epoch": 2.7238152731663146,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.897052718970527e-05,
+      "loss": 0.3075,
+      "step": 4511
+    },
+    {
+      "epoch": 2.7244189556293392,
+      "grad_norm": 0.205078125,
+      "learning_rate": 1.8929016189290164e-05,
+      "loss": 0.2109,
+      "step": 4512
+    },
+    {
+      "epoch": 2.7250226380923634,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.888750518887505e-05,
+      "loss": 0.6597,
+      "step": 4513
+    },
+    {
+      "epoch": 2.7256263205553877,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.884599418845994e-05,
+      "loss": 0.5063,
+      "step": 4514
+    },
+    {
+      "epoch": 2.7262300030184123,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.8804483188044834e-05,
+      "loss": 0.722,
+      "step": 4515
+    },
+    {
+      "epoch": 2.726833685481437,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.8762972187629724e-05,
+      "loss": 0.733,
+      "step": 4516
+    },
+    {
+      "epoch": 2.7274373679444612,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 1.872146118721461e-05,
+      "loss": 0.5692,
+      "step": 4517
+    },
+    {
+      "epoch": 2.7280410504074855,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 1.8679950186799504e-05,
+      "loss": 0.5854,
+      "step": 4518
+    },
+    {
+      "epoch": 2.72864473287051,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 1.8638439186384393e-05,
+      "loss": 0.4641,
+      "step": 4519
+    },
+    {
+      "epoch": 2.729248415333535,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.8596928185969283e-05,
+      "loss": 0.5903,
+      "step": 4520
+    },
+    {
+      "epoch": 2.729852097796559,
+      "grad_norm": 0.1484375,
+      "learning_rate": 1.8555417185554173e-05,
+      "loss": 0.5939,
+      "step": 4521
+    },
+    {
+      "epoch": 2.7304557802595832,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.8513906185139063e-05,
+      "loss": 0.6747,
+      "step": 4522
+    },
+    {
+      "epoch": 2.731059462722608,
+      "grad_norm": 0.142578125,
+      "learning_rate": 1.8472395184723953e-05,
+      "loss": 0.6233,
+      "step": 4523
+    },
+    {
+      "epoch": 2.7316631451856326,
+      "grad_norm": 0.142578125,
+      "learning_rate": 1.8430884184308843e-05,
+      "loss": 0.5499,
+      "step": 4524
+    },
+    {
+      "epoch": 2.732266827648657,
+      "grad_norm": 0.15234375,
+      "learning_rate": 1.8389373183893733e-05,
+      "loss": 0.5983,
+      "step": 4525
+    },
+    {
+      "epoch": 2.732870510111681,
+      "grad_norm": 0.154296875,
+      "learning_rate": 1.8347862183478623e-05,
+      "loss": 0.7181,
+      "step": 4526
+    },
+    {
+      "epoch": 2.7334741925747057,
+      "grad_norm": 0.15625,
+      "learning_rate": 1.8306351183063512e-05,
+      "loss": 0.5598,
+      "step": 4527
+    },
+    {
+      "epoch": 2.7340778750377304,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.8264840182648402e-05,
+      "loss": 0.6007,
+      "step": 4528
+    },
+    {
+      "epoch": 2.7346815575007546,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.8223329182233292e-05,
+      "loss": 0.5836,
+      "step": 4529
+    },
+    {
+      "epoch": 2.735285239963779,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 0.5936,
+      "step": 4530
+    },
+    {
+      "epoch": 2.7358889224268035,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.8140307181403075e-05,
+      "loss": 0.8953,
+      "step": 4531
+    },
+    {
+      "epoch": 2.736492604889828,
+      "grad_norm": 0.142578125,
+      "learning_rate": 1.8098796180987962e-05,
+      "loss": 0.5644,
+      "step": 4532
+    },
+    {
+      "epoch": 2.7370962873528524,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.805728518057285e-05,
+      "loss": 0.7689,
+      "step": 4533
+    },
+    {
+      "epoch": 2.7376999698158766,
+      "grad_norm": 0.15625,
+      "learning_rate": 1.8015774180157745e-05,
+      "loss": 0.7984,
+      "step": 4534
+    },
+    {
+      "epoch": 2.7383036522789013,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.797426317974263e-05,
+      "loss": 0.7707,
+      "step": 4535
+    },
+    {
+      "epoch": 2.738907334741926,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.793275217932752e-05,
+      "loss": 0.7229,
+      "step": 4536
+    },
+    {
+      "epoch": 2.73951101720495,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.7891241178912415e-05,
+      "loss": 0.9094,
+      "step": 4537
+    },
+    {
+      "epoch": 2.7401146996679744,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.7849730178497304e-05,
+      "loss": 0.6551,
+      "step": 4538
+    },
+    {
+      "epoch": 2.740718382130999,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.780821917808219e-05,
+      "loss": 0.696,
+      "step": 4539
+    },
+    {
+      "epoch": 2.7413220645940237,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.776670817766708e-05,
+      "loss": 0.7026,
+      "step": 4540
+    },
+    {
+      "epoch": 2.741925747057048,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.7725197177251974e-05,
+      "loss": 0.6059,
+      "step": 4541
+    },
+    {
+      "epoch": 2.742529429520072,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.7683686176836864e-05,
+      "loss": 0.6566,
+      "step": 4542
+    },
+    {
+      "epoch": 2.743133111983097,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.764217517642175e-05,
+      "loss": 0.6337,
+      "step": 4543
+    },
+    {
+      "epoch": 2.7437367944461215,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.7600664176006644e-05,
+      "loss": 0.512,
+      "step": 4544
+    },
+    {
+      "epoch": 2.7443404769091457,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.7559153175591534e-05,
+      "loss": 0.4839,
+      "step": 4545
+    },
+    {
+      "epoch": 2.7449441593721704,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.751764217517642e-05,
+      "loss": 0.5621,
+      "step": 4546
+    },
+    {
+      "epoch": 2.7455478418351946,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.7476131174761313e-05,
+      "loss": 0.5687,
+      "step": 4547
+    },
+    {
+      "epoch": 2.7461515242982193,
+      "grad_norm": 0.142578125,
+      "learning_rate": 1.7434620174346203e-05,
+      "loss": 0.5606,
+      "step": 4548
+    },
+    {
+      "epoch": 2.7467552067612435,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.7393109173931093e-05,
+      "loss": 0.5483,
+      "step": 4549
+    },
+    {
+      "epoch": 2.747358889224268,
+      "grad_norm": 0.150390625,
+      "learning_rate": 1.7351598173515983e-05,
+      "loss": 0.4612,
+      "step": 4550
+    },
+    {
+      "epoch": 2.7479625716872924,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.7310087173100873e-05,
+      "loss": 0.5156,
+      "step": 4551
+    },
+    {
+      "epoch": 2.748566254150317,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.7268576172685763e-05,
+      "loss": 0.5005,
+      "step": 4552
+    },
+    {
+      "epoch": 2.7491699366133413,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.7227065172270653e-05,
+      "loss": 0.454,
+      "step": 4553
+    },
+    {
+      "epoch": 2.749773619076366,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.7185554171855542e-05,
+      "loss": 0.5291,
+      "step": 4554
+    },
+    {
+      "epoch": 2.75037730153939,
+      "grad_norm": 0.18359375,
+      "learning_rate": 1.7144043171440432e-05,
+      "loss": 0.5026,
+      "step": 4555
+    },
+    {
+      "epoch": 2.750980984002415,
+      "grad_norm": 0.189453125,
+      "learning_rate": 1.7102532171025322e-05,
+      "loss": 0.4815,
+      "step": 4556
+    },
+    {
+      "epoch": 2.751584666465439,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 1.7061021170610212e-05,
+      "loss": 0.4753,
+      "step": 4557
+    },
+    {
+      "epoch": 2.7521883489284638,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.7019510170195102e-05,
+      "loss": 0.4151,
+      "step": 4558
+    },
+    {
+      "epoch": 2.752792031391488,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.6977999169779992e-05,
+      "loss": 0.3104,
+      "step": 4559
+    },
+    {
+      "epoch": 2.7533957138545126,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.6936488169364885e-05,
+      "loss": 0.3303,
+      "step": 4560
+    },
+    {
+      "epoch": 2.753999396317537,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.689497716894977e-05,
+      "loss": 0.308,
+      "step": 4561
+    },
+    {
+      "epoch": 2.7546030787805615,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.685346616853466e-05,
+      "loss": 0.2067,
+      "step": 4562
+    },
+    {
+      "epoch": 2.7552067612435858,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.6811955168119555e-05,
+      "loss": 1.0449,
+      "step": 4563
+    },
+    {
+      "epoch": 2.7558104437066104,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.6770444167704444e-05,
+      "loss": 0.6562,
+      "step": 4564
+    },
+    {
+      "epoch": 2.7564141261696347,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.672893316728933e-05,
+      "loss": 0.5127,
+      "step": 4565
+    },
+    {
+      "epoch": 2.7570178086326593,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.6687422166874224e-05,
+      "loss": 0.6374,
+      "step": 4566
+    },
+    {
+      "epoch": 2.7576214910956836,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.6645911166459114e-05,
+      "loss": 0.7662,
+      "step": 4567
+    },
+    {
+      "epoch": 2.758225173558708,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.6604400166044e-05,
+      "loss": 0.7541,
+      "step": 4568
+    },
+    {
+      "epoch": 2.7588288560217324,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.656288916562889e-05,
+      "loss": 0.735,
+      "step": 4569
+    },
+    {
+      "epoch": 2.759432538484757,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 1.6521378165213784e-05,
+      "loss": 0.59,
+      "step": 4570
+    },
+    {
+      "epoch": 2.7600362209477813,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.6479867164798674e-05,
+      "loss": 0.5174,
+      "step": 4571
+    },
+    {
+      "epoch": 2.760639903410806,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.643835616438356e-05,
+      "loss": 0.5364,
+      "step": 4572
+    },
+    {
+      "epoch": 2.7612435858738302,
+      "grad_norm": 0.150390625,
+      "learning_rate": 1.6396845163968453e-05,
+      "loss": 0.6259,
+      "step": 4573
+    },
+    {
+      "epoch": 2.761847268336855,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.6355334163553343e-05,
+      "loss": 0.8076,
+      "step": 4574
+    },
+    {
+      "epoch": 2.762450950799879,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.6313823163138233e-05,
+      "loss": 0.5766,
+      "step": 4575
+    },
+    {
+      "epoch": 2.763054633262904,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.6272312162723123e-05,
+      "loss": 0.6713,
+      "step": 4576
+    },
+    {
+      "epoch": 2.763658315725928,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.6230801162308013e-05,
+      "loss": 0.6865,
+      "step": 4577
+    },
+    {
+      "epoch": 2.7642619981889527,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.6189290161892903e-05,
+      "loss": 0.8193,
+      "step": 4578
+    },
+    {
+      "epoch": 2.764865680651977,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.6147779161477793e-05,
+      "loss": 0.5764,
+      "step": 4579
+    },
+    {
+      "epoch": 2.7654693631150016,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.6106268161062682e-05,
+      "loss": 0.7132,
+      "step": 4580
+    },
+    {
+      "epoch": 2.766073045578026,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.6064757160647572e-05,
+      "loss": 0.7006,
+      "step": 4581
+    },
+    {
+      "epoch": 2.7666767280410505,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.6023246160232462e-05,
+      "loss": 0.62,
+      "step": 4582
+    },
+    {
+      "epoch": 2.7672804105040747,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.5981735159817352e-05,
+      "loss": 0.5348,
+      "step": 4583
+    },
+    {
+      "epoch": 2.7678840929670994,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.5940224159402242e-05,
+      "loss": 0.6188,
+      "step": 4584
+    },
+    {
+      "epoch": 2.768487775430124,
+      "grad_norm": 0.15625,
+      "learning_rate": 1.5898713158987132e-05,
+      "loss": 0.6714,
+      "step": 4585
+    },
+    {
+      "epoch": 2.7690914578931483,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.585720215857202e-05,
+      "loss": 0.7338,
+      "step": 4586
+    },
+    {
+      "epoch": 2.7696951403561725,
+      "grad_norm": 0.15625,
+      "learning_rate": 1.581569115815691e-05,
+      "loss": 0.5708,
+      "step": 4587
+    },
+    {
+      "epoch": 2.770298822819197,
+      "grad_norm": 0.15234375,
+      "learning_rate": 1.57741801577418e-05,
+      "loss": 0.6864,
+      "step": 4588
+    },
+    {
+      "epoch": 2.770902505282222,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.5732669157326695e-05,
+      "loss": 0.57,
+      "step": 4589
+    },
+    {
+      "epoch": 2.771506187745246,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.569115815691158e-05,
+      "loss": 0.5914,
+      "step": 4590
+    },
+    {
+      "epoch": 2.7721098702082703,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 1.564964715649647e-05,
+      "loss": 0.9134,
+      "step": 4591
+    },
+    {
+      "epoch": 2.772713552671295,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.5608136156081364e-05,
+      "loss": 0.6086,
+      "step": 4592
+    },
+    {
+      "epoch": 2.7733172351343196,
+      "grad_norm": 0.154296875,
+      "learning_rate": 1.5566625155666254e-05,
+      "loss": 0.7351,
+      "step": 4593
+    },
+    {
+      "epoch": 2.773920917597344,
+      "grad_norm": 0.1484375,
+      "learning_rate": 1.552511415525114e-05,
+      "loss": 0.8879,
+      "step": 4594
+    },
+    {
+      "epoch": 2.774524600060368,
+      "grad_norm": 0.150390625,
+      "learning_rate": 1.548360315483603e-05,
+      "loss": 0.8547,
+      "step": 4595
+    },
+    {
+      "epoch": 2.7751282825233927,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.5442092154420924e-05,
+      "loss": 0.5326,
+      "step": 4596
+    },
+    {
+      "epoch": 2.7757319649864174,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.540058115400581e-05,
+      "loss": 0.4954,
+      "step": 4597
+    },
+    {
+      "epoch": 2.7763356474494416,
+      "grad_norm": 0.16015625,
+      "learning_rate": 1.53590701535907e-05,
+      "loss": 0.6014,
+      "step": 4598
+    },
+    {
+      "epoch": 2.776939329912466,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.5317559153175593e-05,
+      "loss": 0.5491,
+      "step": 4599
+    },
+    {
+      "epoch": 2.7775430123754905,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.5276048152760483e-05,
+      "loss": 0.5466,
+      "step": 4600
+    },
+    {
+      "epoch": 2.778146694838515,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.5234537152345371e-05,
+      "loss": 0.508,
+      "step": 4601
+    },
+    {
+      "epoch": 2.7787503773015394,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.5193026151930261e-05,
+      "loss": 0.5131,
+      "step": 4602
+    },
+    {
+      "epoch": 2.7793540597645636,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.5151515151515153e-05,
+      "loss": 0.4846,
+      "step": 4603
+    },
+    {
+      "epoch": 2.7799577422275883,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.5110004151100043e-05,
+      "loss": 0.5484,
+      "step": 4604
+    },
+    {
+      "epoch": 2.780561424690613,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.5068493150684931e-05,
+      "loss": 0.4347,
+      "step": 4605
+    },
+    {
+      "epoch": 2.781165107153637,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.5026982150269822e-05,
+      "loss": 0.4263,
+      "step": 4606
+    },
+    {
+      "epoch": 2.7817687896166614,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.4985471149854712e-05,
+      "loss": 0.4077,
+      "step": 4607
+    },
+    {
+      "epoch": 2.782372472079686,
+      "grad_norm": 0.201171875,
+      "learning_rate": 1.49439601494396e-05,
+      "loss": 0.4031,
+      "step": 4608
+    },
+    {
+      "epoch": 2.7829761545427107,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.4902449149024492e-05,
+      "loss": 0.3571,
+      "step": 4609
+    },
+    {
+      "epoch": 2.783579837005735,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.4860938148609382e-05,
+      "loss": 0.2875,
+      "step": 4610
+    },
+    {
+      "epoch": 2.784183519468759,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.4819427148194274e-05,
+      "loss": 0.257,
+      "step": 4611
+    },
+    {
+      "epoch": 2.784787201931784,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 1.4777916147779162e-05,
+      "loss": 0.2157,
+      "step": 4612
+    },
+    {
+      "epoch": 2.7853908843948085,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.4736405147364052e-05,
+      "loss": 0.7389,
+      "step": 4613
+    },
+    {
+      "epoch": 2.7859945668578328,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.4694894146948943e-05,
+      "loss": 0.7929,
+      "step": 4614
+    },
+    {
+      "epoch": 2.786598249320857,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.4653383146533833e-05,
+      "loss": 0.6031,
+      "step": 4615
+    },
+    {
+      "epoch": 2.7872019317838816,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.4611872146118721e-05,
+      "loss": 0.5555,
+      "step": 4616
+    },
+    {
+      "epoch": 2.7878056142469063,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.4570361145703613e-05,
+      "loss": 0.6512,
+      "step": 4617
+    },
+    {
+      "epoch": 2.7884092967099305,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.4528850145288503e-05,
+      "loss": 0.6039,
+      "step": 4618
+    },
+    {
+      "epoch": 2.7890129791729548,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.448733914487339e-05,
+      "loss": 0.6583,
+      "step": 4619
+    },
+    {
+      "epoch": 2.7896166616359794,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.4445828144458282e-05,
+      "loss": 0.6774,
+      "step": 4620
+    },
+    {
+      "epoch": 2.790220344099004,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.4404317144043172e-05,
+      "loss": 0.5934,
+      "step": 4621
+    },
+    {
+      "epoch": 2.7908240265620283,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.4362806143628064e-05,
+      "loss": 0.6648,
+      "step": 4622
+    },
+    {
+      "epoch": 2.7914277090250526,
+      "grad_norm": 0.130859375,
+      "learning_rate": 1.432129514321295e-05,
+      "loss": 0.5695,
+      "step": 4623
+    },
+    {
+      "epoch": 2.792031391488077,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.4279784142797842e-05,
+      "loss": 0.9932,
+      "step": 4624
+    },
+    {
+      "epoch": 2.792635073951102,
+      "grad_norm": 0.154296875,
+      "learning_rate": 1.4238273142382733e-05,
+      "loss": 0.6688,
+      "step": 4625
+    },
+    {
+      "epoch": 2.793238756414126,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.4196762141967623e-05,
+      "loss": 0.8921,
+      "step": 4626
+    },
+    {
+      "epoch": 2.793842438877151,
+      "grad_norm": 0.12890625,
+      "learning_rate": 1.4155251141552511e-05,
+      "loss": 0.8431,
+      "step": 4627
+    },
+    {
+      "epoch": 2.794446121340175,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.4113740141137401e-05,
+      "loss": 0.6138,
+      "step": 4628
+    },
+    {
+      "epoch": 2.7950498038031997,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.4072229140722293e-05,
+      "loss": 0.7836,
+      "step": 4629
+    },
+    {
+      "epoch": 2.795653486266224,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.4030718140307181e-05,
+      "loss": 0.8925,
+      "step": 4630
+    },
+    {
+      "epoch": 2.7962571687292486,
+      "grad_norm": 0.15234375,
+      "learning_rate": 1.3989207139892071e-05,
+      "loss": 0.6221,
+      "step": 4631
+    },
+    {
+      "epoch": 2.796860851192273,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.3947696139476963e-05,
+      "loss": 0.6567,
+      "step": 4632
+    },
+    {
+      "epoch": 2.7974645336552975,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.3906185139061852e-05,
+      "loss": 0.5648,
+      "step": 4633
+    },
+    {
+      "epoch": 2.7980682161183217,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.386467413864674e-05,
+      "loss": 0.6673,
+      "step": 4634
+    },
+    {
+      "epoch": 2.7986718985813464,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.3823163138231632e-05,
+      "loss": 0.605,
+      "step": 4635
+    },
+    {
+      "epoch": 2.7992755810443706,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.3781652137816522e-05,
+      "loss": 0.6171,
+      "step": 4636
+    },
+    {
+      "epoch": 2.7998792635073952,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.3740141137401414e-05,
+      "loss": 0.5335,
+      "step": 4637
+    },
+    {
+      "epoch": 2.8004829459704195,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.3698630136986302e-05,
+      "loss": 0.7067,
+      "step": 4638
+    },
+    {
+      "epoch": 2.801086628433444,
+      "grad_norm": 0.1484375,
+      "learning_rate": 1.3657119136571192e-05,
+      "loss": 0.655,
+      "step": 4639
+    },
+    {
+      "epoch": 2.8016903108964684,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.3615608136156083e-05,
+      "loss": 0.5433,
+      "step": 4640
+    },
+    {
+      "epoch": 2.802293993359493,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.3574097135740971e-05,
+      "loss": 0.7166,
+      "step": 4641
+    },
+    {
+      "epoch": 2.8028976758225173,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.3532586135325861e-05,
+      "loss": 0.6537,
+      "step": 4642
+    },
+    {
+      "epoch": 2.803501358285542,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.3491075134910753e-05,
+      "loss": 0.6526,
+      "step": 4643
+    },
+    {
+      "epoch": 2.804105040748566,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.3449564134495643e-05,
+      "loss": 0.6646,
+      "step": 4644
+    },
+    {
+      "epoch": 2.804708723211591,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.3408053134080531e-05,
+      "loss": 0.5491,
+      "step": 4645
+    },
+    {
+      "epoch": 2.805312405674615,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.3366542133665422e-05,
+      "loss": 0.5573,
+      "step": 4646
+    },
+    {
+      "epoch": 2.8059160881376397,
+      "grad_norm": 0.154296875,
+      "learning_rate": 1.3325031133250312e-05,
+      "loss": 0.5577,
+      "step": 4647
+    },
+    {
+      "epoch": 2.806519770600664,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.3283520132835204e-05,
+      "loss": 0.5257,
+      "step": 4648
+    },
+    {
+      "epoch": 2.8071234530636886,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.3242009132420092e-05,
+      "loss": 0.6006,
+      "step": 4649
+    },
+    {
+      "epoch": 2.807727135526713,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.3200498132004982e-05,
+      "loss": 0.5615,
+      "step": 4650
+    },
+    {
+      "epoch": 2.8083308179897375,
+      "grad_norm": 0.15625,
+      "learning_rate": 1.3158987131589873e-05,
+      "loss": 0.5161,
+      "step": 4651
+    },
+    {
+      "epoch": 2.8089345004527617,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.311747613117476e-05,
+      "loss": 0.5035,
+      "step": 4652
+    },
+    {
+      "epoch": 2.8095381829157864,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.3075965130759652e-05,
+      "loss": 0.5236,
+      "step": 4653
+    },
+    {
+      "epoch": 2.8101418653788106,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.3034454130344541e-05,
+      "loss": 0.5802,
+      "step": 4654
+    },
+    {
+      "epoch": 2.8107455478418353,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.2992943129929433e-05,
+      "loss": 0.4602,
+      "step": 4655
+    },
+    {
+      "epoch": 2.8113492303048595,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.2951432129514321e-05,
+      "loss": 0.4859,
+      "step": 4656
+    },
+    {
+      "epoch": 2.811952912767884,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.2909921129099211e-05,
+      "loss": 0.4449,
+      "step": 4657
+    },
+    {
+      "epoch": 2.8125565952309084,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.2868410128684103e-05,
+      "loss": 0.3615,
+      "step": 4658
+    },
+    {
+      "epoch": 2.813160277693933,
+      "grad_norm": 0.19921875,
+      "learning_rate": 1.2826899128268992e-05,
+      "loss": 0.325,
+      "step": 4659
+    },
+    {
+      "epoch": 2.8137639601569573,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.278538812785388e-05,
+      "loss": 0.3904,
+      "step": 4660
+    },
+    {
+      "epoch": 2.814367642619982,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.2743877127438772e-05,
+      "loss": 0.271,
+      "step": 4661
+    },
+    {
+      "epoch": 2.814971325083006,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.2702366127023662e-05,
+      "loss": 0.2315,
+      "step": 4662
+    },
+    {
+      "epoch": 2.815575007546031,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.266085512660855e-05,
+      "loss": 0.7124,
+      "step": 4663
+    },
+    {
+      "epoch": 2.816178690009055,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.2619344126193442e-05,
+      "loss": 0.6489,
+      "step": 4664
+    },
+    {
+      "epoch": 2.8167823724720797,
+      "grad_norm": 0.15625,
+      "learning_rate": 1.2577833125778332e-05,
+      "loss": 0.966,
+      "step": 4665
+    },
+    {
+      "epoch": 2.8173860549351044,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.2536322125363223e-05,
+      "loss": 0.7182,
+      "step": 4666
+    },
+    {
+      "epoch": 2.8179897373981286,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.2494811124948113e-05,
+      "loss": 0.6668,
+      "step": 4667
+    },
+    {
+      "epoch": 2.818593419861153,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 1.2453300124533001e-05,
+      "loss": 0.6513,
+      "step": 4668
+    },
+    {
+      "epoch": 2.8191971023241775,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.2411789124117893e-05,
+      "loss": 0.5954,
+      "step": 4669
+    },
+    {
+      "epoch": 2.819800784787202,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.2370278123702781e-05,
+      "loss": 0.7074,
+      "step": 4670
+    },
+    {
+      "epoch": 2.8204044672502264,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.2328767123287671e-05,
+      "loss": 0.5661,
+      "step": 4671
+    },
+    {
+      "epoch": 2.8210081497132506,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.2287256122872562e-05,
+      "loss": 0.6607,
+      "step": 4672
+    },
+    {
+      "epoch": 2.8216118321762753,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.224574512245745e-05,
+      "loss": 0.7987,
+      "step": 4673
+    },
+    {
+      "epoch": 2.8222155146393,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.2204234122042342e-05,
+      "loss": 0.8645,
+      "step": 4674
+    },
+    {
+      "epoch": 2.822819197102324,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.2162723121627232e-05,
+      "loss": 0.4833,
+      "step": 4675
+    },
+    {
+      "epoch": 2.8234228795653484,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.2121212121212122e-05,
+      "loss": 0.6758,
+      "step": 4676
+    },
+    {
+      "epoch": 2.824026562028373,
+      "grad_norm": 0.1484375,
+      "learning_rate": 1.2079701120797012e-05,
+      "loss": 0.6682,
+      "step": 4677
+    },
+    {
+      "epoch": 2.8246302444913978,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.2038190120381902e-05,
+      "loss": 0.753,
+      "step": 4678
+    },
+    {
+      "epoch": 2.825233926954422,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 1.1996679119966792e-05,
+      "loss": 0.4793,
+      "step": 4679
+    },
+    {
+      "epoch": 2.825837609417446,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.1955168119551683e-05,
+      "loss": 0.6239,
+      "step": 4680
+    },
+    {
+      "epoch": 2.826441291880471,
+      "grad_norm": 0.15625,
+      "learning_rate": 1.1913657119136571e-05,
+      "loss": 0.607,
+      "step": 4681
+    },
+    {
+      "epoch": 2.8270449743434956,
+      "grad_norm": 0.150390625,
+      "learning_rate": 1.1872146118721461e-05,
+      "loss": 0.7043,
+      "step": 4682
+    },
+    {
+      "epoch": 2.82764865680652,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 1.1830635118306351e-05,
+      "loss": 0.7399,
+      "step": 4683
+    },
+    {
+      "epoch": 2.828252339269544,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.1789124117891241e-05,
+      "loss": 0.5987,
+      "step": 4684
+    },
+    {
+      "epoch": 2.8288560217325687,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.1747613117476132e-05,
+      "loss": 0.6033,
+      "step": 4685
+    },
+    {
+      "epoch": 2.8294597041955933,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.170610211706102e-05,
+      "loss": 0.5987,
+      "step": 4686
+    },
+    {
+      "epoch": 2.8300633866586176,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.1664591116645912e-05,
+      "loss": 0.6025,
+      "step": 4687
+    },
+    {
+      "epoch": 2.830667069121642,
+      "grad_norm": 0.142578125,
+      "learning_rate": 1.1623080116230802e-05,
+      "loss": 0.588,
+      "step": 4688
+    },
+    {
+      "epoch": 2.8312707515846665,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 1.1581569115815692e-05,
+      "loss": 0.6295,
+      "step": 4689
+    },
+    {
+      "epoch": 2.831874434047691,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.1540058115400582e-05,
+      "loss": 0.7745,
+      "step": 4690
+    },
+    {
+      "epoch": 2.8324781165107153,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.1498547114985472e-05,
+      "loss": 0.6095,
+      "step": 4691
+    },
+    {
+      "epoch": 2.8330817989737396,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.1457036114570362e-05,
+      "loss": 0.8522,
+      "step": 4692
+    },
+    {
+      "epoch": 2.8336854814367642,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.1415525114155251e-05,
+      "loss": 0.7047,
+      "step": 4693
+    },
+    {
+      "epoch": 2.834289163899789,
+      "grad_norm": 0.150390625,
+      "learning_rate": 1.1374014113740141e-05,
+      "loss": 0.561,
+      "step": 4694
+    },
+    {
+      "epoch": 2.834892846362813,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.1332503113325031e-05,
+      "loss": 0.8301,
+      "step": 4695
+    },
+    {
+      "epoch": 2.8354965288258374,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.1290992112909923e-05,
+      "loss": 0.5409,
+      "step": 4696
+    },
+    {
+      "epoch": 2.836100211288862,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.1249481112494811e-05,
+      "loss": 0.5462,
+      "step": 4697
+    },
+    {
+      "epoch": 2.8367038937518867,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.1207970112079703e-05,
+      "loss": 0.5022,
+      "step": 4698
+    },
+    {
+      "epoch": 2.837307576214911,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 1.116645911166459e-05,
+      "loss": 0.6117,
+      "step": 4699
+    },
+    {
+      "epoch": 2.837911258677935,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.1124948111249482e-05,
+      "loss": 0.4875,
+      "step": 4700
+    },
+    {
+      "epoch": 2.83851494114096,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.1083437110834372e-05,
+      "loss": 0.5104,
+      "step": 4701
+    },
+    {
+      "epoch": 2.8391186236039845,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.1041926110419262e-05,
+      "loss": 0.5757,
+      "step": 4702
+    },
+    {
+      "epoch": 2.8397223060670087,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.1000415110004152e-05,
+      "loss": 0.4598,
+      "step": 4703
+    },
+    {
+      "epoch": 2.840325988530033,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.0958904109589042e-05,
+      "loss": 0.494,
+      "step": 4704
+    },
+    {
+      "epoch": 2.8409296709930576,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.0917393109173932e-05,
+      "loss": 0.4241,
+      "step": 4705
+    },
+    {
+      "epoch": 2.8415333534560823,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.0875882108758821e-05,
+      "loss": 0.4557,
+      "step": 4706
+    },
+    {
+      "epoch": 2.8421370359191065,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.0834371108343711e-05,
+      "loss": 0.4101,
+      "step": 4707
+    },
+    {
+      "epoch": 2.842740718382131,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.0792860107928601e-05,
+      "loss": 0.3463,
+      "step": 4708
+    },
+    {
+      "epoch": 2.8433444008451554,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.0751349107513493e-05,
+      "loss": 0.3673,
+      "step": 4709
+    },
+    {
+      "epoch": 2.84394808330818,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.0709838107098381e-05,
+      "loss": 0.2754,
+      "step": 4710
+    },
+    {
+      "epoch": 2.8445517657712043,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.0668327106683273e-05,
+      "loss": 0.2248,
+      "step": 4711
+    },
+    {
+      "epoch": 2.845155448234229,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.062681610626816e-05,
+      "loss": 0.2027,
+      "step": 4712
+    },
+    {
+      "epoch": 2.845759130697253,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.0585305105853052e-05,
+      "loss": 0.674,
+      "step": 4713
+    },
+    {
+      "epoch": 2.846362813160278,
+      "grad_norm": 0.15234375,
+      "learning_rate": 1.0543794105437942e-05,
+      "loss": 0.8905,
+      "step": 4714
+    },
+    {
+      "epoch": 2.846966495623302,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.050228310502283e-05,
+      "loss": 0.7134,
+      "step": 4715
+    },
+    {
+      "epoch": 2.8475701780863267,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.0460772104607722e-05,
+      "loss": 0.5395,
+      "step": 4716
+    },
+    {
+      "epoch": 2.848173860549351,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.0419261104192612e-05,
+      "loss": 0.4918,
+      "step": 4717
+    },
+    {
+      "epoch": 2.8487775430123756,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.0377750103777502e-05,
+      "loss": 0.5988,
+      "step": 4718
+    },
+    {
+      "epoch": 2.8493812254754,
+      "grad_norm": 0.142578125,
+      "learning_rate": 1.0336239103362392e-05,
+      "loss": 0.5771,
+      "step": 4719
+    },
+    {
+      "epoch": 2.8499849079384245,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.0294728102947281e-05,
+      "loss": 0.6012,
+      "step": 4720
+    },
+    {
+      "epoch": 2.8505885904014487,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.0253217102532171e-05,
+      "loss": 0.6447,
+      "step": 4721
+    },
+    {
+      "epoch": 2.8511922728644734,
+      "grad_norm": 0.15625,
+      "learning_rate": 1.0211706102117063e-05,
+      "loss": 0.6304,
+      "step": 4722
+    },
+    {
+      "epoch": 2.8517959553274976,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.0170195101701951e-05,
+      "loss": 1.1191,
+      "step": 4723
+    },
+    {
+      "epoch": 2.8523996377905223,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.0128684101286841e-05,
+      "loss": 0.8335,
+      "step": 4724
+    },
+    {
+      "epoch": 2.8530033202535465,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.008717310087173e-05,
+      "loss": 0.5978,
+      "step": 4725
+    },
+    {
+      "epoch": 2.853607002716571,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.004566210045662e-05,
+      "loss": 0.6225,
+      "step": 4726
+    },
+    {
+      "epoch": 2.8542106851795954,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.0004151100041512e-05,
+      "loss": 0.7899,
+      "step": 4727
+    },
+    {
+      "epoch": 2.85481436764262,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 9.9626400996264e-06,
+      "loss": 0.6002,
+      "step": 4728
+    },
+    {
+      "epoch": 2.8554180501056443,
+      "grad_norm": 0.150390625,
+      "learning_rate": 9.921129099211292e-06,
+      "loss": 0.6239,
+      "step": 4729
+    },
+    {
+      "epoch": 2.856021732568669,
+      "grad_norm": 0.138671875,
+      "learning_rate": 9.879618098796182e-06,
+      "loss": 0.6067,
+      "step": 4730
+    },
+    {
+      "epoch": 2.856625415031693,
+      "grad_norm": 0.140625,
+      "learning_rate": 9.838107098381072e-06,
+      "loss": 0.5572,
+      "step": 4731
+    },
+    {
+      "epoch": 2.857229097494718,
+      "grad_norm": 0.154296875,
+      "learning_rate": 9.796596097965962e-06,
+      "loss": 0.5798,
+      "step": 4732
+    },
+    {
+      "epoch": 2.857832779957742,
+      "grad_norm": 0.140625,
+      "learning_rate": 9.755085097550851e-06,
+      "loss": 0.6557,
+      "step": 4733
+    },
+    {
+      "epoch": 2.8584364624207668,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 9.713574097135741e-06,
+      "loss": 0.8296,
+      "step": 4734
+    },
+    {
+      "epoch": 2.859040144883791,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 9.672063096720631e-06,
+      "loss": 0.9093,
+      "step": 4735
+    },
+    {
+      "epoch": 2.8596438273468157,
+      "grad_norm": 0.13671875,
+      "learning_rate": 9.630552096305521e-06,
+      "loss": 0.6259,
+      "step": 4736
+    },
+    {
+      "epoch": 2.86024750980984,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 9.589041095890411e-06,
+      "loss": 0.6026,
+      "step": 4737
+    },
+    {
+      "epoch": 2.8608511922728646,
+      "grad_norm": 0.16015625,
+      "learning_rate": 9.547530095475302e-06,
+      "loss": 0.67,
+      "step": 4738
+    },
+    {
+      "epoch": 2.8614548747358888,
+      "grad_norm": 0.1484375,
+      "learning_rate": 9.50601909506019e-06,
+      "loss": 0.5923,
+      "step": 4739
+    },
+    {
+      "epoch": 2.8620585571989134,
+      "grad_norm": 0.14453125,
+      "learning_rate": 9.464508094645082e-06,
+      "loss": 0.7869,
+      "step": 4740
+    },
+    {
+      "epoch": 2.8626622396619377,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 9.42299709422997e-06,
+      "loss": 0.5751,
+      "step": 4741
+    },
+    {
+      "epoch": 2.8632659221249623,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.381486093814862e-06,
+      "loss": 0.7413,
+      "step": 4742
+    },
+    {
+      "epoch": 2.8638696045879866,
+      "grad_norm": 0.14453125,
+      "learning_rate": 9.339975093399752e-06,
+      "loss": 0.6217,
+      "step": 4743
+    },
+    {
+      "epoch": 2.8644732870510112,
+      "grad_norm": 0.142578125,
+      "learning_rate": 9.298464092984642e-06,
+      "loss": 0.5413,
+      "step": 4744
+    },
+    {
+      "epoch": 2.8650769695140355,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 9.256953092569532e-06,
+      "loss": 0.5922,
+      "step": 4745
+    },
+    {
+      "epoch": 2.86568065197706,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 9.215442092154421e-06,
+      "loss": 0.5451,
+      "step": 4746
+    },
+    {
+      "epoch": 2.866284334440085,
+      "grad_norm": 0.154296875,
+      "learning_rate": 9.173931091739311e-06,
+      "loss": 0.6009,
+      "step": 4747
+    },
+    {
+      "epoch": 2.866888016903109,
+      "grad_norm": 0.16015625,
+      "learning_rate": 9.132420091324201e-06,
+      "loss": 0.6489,
+      "step": 4748
+    },
+    {
+      "epoch": 2.8674916993661332,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.5733,
+      "step": 4749
+    },
+    {
+      "epoch": 2.868095381829158,
+      "grad_norm": 0.150390625,
+      "learning_rate": 9.049398090493981e-06,
+      "loss": 0.5809,
+      "step": 4750
+    },
+    {
+      "epoch": 2.8686990642921826,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.007887090078872e-06,
+      "loss": 0.6779,
+      "step": 4751
+    },
+    {
+      "epoch": 2.869302746755207,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 8.96637608966376e-06,
+      "loss": 0.5155,
+      "step": 4752
+    },
+    {
+      "epoch": 2.869906429218231,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.924865089248652e-06,
+      "loss": 0.5757,
+      "step": 4753
+    },
+    {
+      "epoch": 2.8705101116812557,
+      "grad_norm": 0.16015625,
+      "learning_rate": 8.88335408883354e-06,
+      "loss": 0.4536,
+      "step": 4754
+    },
+    {
+      "epoch": 2.8711137941442804,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.841843088418432e-06,
+      "loss": 0.4535,
+      "step": 4755
+    },
+    {
+      "epoch": 2.8717174766073046,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 8.800332088003322e-06,
+      "loss": 0.463,
+      "step": 4756
+    },
+    {
+      "epoch": 2.872321159070329,
+      "grad_norm": 0.203125,
+      "learning_rate": 8.75882108758821e-06,
+      "loss": 0.452,
+      "step": 4757
+    },
+    {
+      "epoch": 2.8729248415333535,
+      "grad_norm": 0.208984375,
+      "learning_rate": 8.717310087173102e-06,
+      "loss": 0.387,
+      "step": 4758
+    },
+    {
+      "epoch": 2.873528523996378,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 8.675799086757991e-06,
+      "loss": 0.3369,
+      "step": 4759
+    },
+    {
+      "epoch": 2.8741322064594024,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 8.634288086342881e-06,
+      "loss": 0.2758,
+      "step": 4760
+    },
+    {
+      "epoch": 2.8747358889224266,
+      "grad_norm": 0.19921875,
+      "learning_rate": 8.592777085927771e-06,
+      "loss": 0.2384,
+      "step": 4761
+    },
+    {
+      "epoch": 2.8753395713854513,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.551266085512661e-06,
+      "loss": 0.2201,
+      "step": 4762
+    },
+    {
+      "epoch": 2.875943253848476,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 8.509755085097551e-06,
+      "loss": 0.5102,
+      "step": 4763
+    },
+    {
+      "epoch": 2.8765469363115,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.468244084682442e-06,
+      "loss": 0.7652,
+      "step": 4764
+    },
+    {
+      "epoch": 2.8771506187745244,
+      "grad_norm": 0.142578125,
+      "learning_rate": 8.42673308426733e-06,
+      "loss": 0.8522,
+      "step": 4765
+    },
+    {
+      "epoch": 2.877754301237549,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 8.385222083852222e-06,
+      "loss": 1.0595,
+      "step": 4766
+    },
+    {
+      "epoch": 2.8783579837005737,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 8.343711083437112e-06,
+      "loss": 0.5743,
+      "step": 4767
+    },
+    {
+      "epoch": 2.878961666163598,
+      "grad_norm": 0.154296875,
+      "learning_rate": 8.302200083022e-06,
+      "loss": 0.6199,
+      "step": 4768
+    },
+    {
+      "epoch": 2.879565348626622,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 8.260689082606892e-06,
+      "loss": 0.6884,
+      "step": 4769
+    },
+    {
+      "epoch": 2.880169031089647,
+      "grad_norm": 0.140625,
+      "learning_rate": 8.21917808219178e-06,
+      "loss": 0.5411,
+      "step": 4770
+    },
+    {
+      "epoch": 2.8807727135526715,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 8.177667081776672e-06,
+      "loss": 0.5833,
+      "step": 4771
+    },
+    {
+      "epoch": 2.8813763960156957,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 8.136156081361561e-06,
+      "loss": 0.8275,
+      "step": 4772
+    },
+    {
+      "epoch": 2.88198007847872,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 8.094645080946451e-06,
+      "loss": 0.8618,
+      "step": 4773
+    },
+    {
+      "epoch": 2.8825837609417446,
+      "grad_norm": 0.146484375,
+      "learning_rate": 8.053134080531341e-06,
+      "loss": 0.9263,
+      "step": 4774
+    },
+    {
+      "epoch": 2.8831874434047693,
+      "grad_norm": 0.1484375,
+      "learning_rate": 8.011623080116231e-06,
+      "loss": 0.5723,
+      "step": 4775
+    },
+    {
+      "epoch": 2.8837911258677935,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 7.970112079701121e-06,
+      "loss": 0.5347,
+      "step": 4776
+    },
+    {
+      "epoch": 2.8843948083308177,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 7.92860107928601e-06,
+      "loss": 0.8219,
+      "step": 4777
+    },
+    {
+      "epoch": 2.8849984907938424,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 7.8870900788709e-06,
+      "loss": 0.6338,
+      "step": 4778
+    },
+    {
+      "epoch": 2.885602173256867,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 7.84557907845579e-06,
+      "loss": 0.696,
+      "step": 4779
+    },
+    {
+      "epoch": 2.8862058557198913,
+      "grad_norm": 0.130859375,
+      "learning_rate": 7.804068078040682e-06,
+      "loss": 0.7539,
+      "step": 4780
+    },
+    {
+      "epoch": 2.8868095381829155,
+      "grad_norm": 0.13671875,
+      "learning_rate": 7.76255707762557e-06,
+      "loss": 0.5847,
+      "step": 4781
+    },
+    {
+      "epoch": 2.88741322064594,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 7.721046077210462e-06,
+      "loss": 0.6444,
+      "step": 4782
+    },
+    {
+      "epoch": 2.888016903108965,
+      "grad_norm": 0.154296875,
+      "learning_rate": 7.67953507679535e-06,
+      "loss": 0.4944,
+      "step": 4783
+    },
+    {
+      "epoch": 2.888620585571989,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 7.638024076380242e-06,
+      "loss": 0.6233,
+      "step": 4784
+    },
+    {
+      "epoch": 2.8892242680350133,
+      "grad_norm": 0.150390625,
+      "learning_rate": 7.596513075965131e-06,
+      "loss": 0.6144,
+      "step": 4785
+    },
+    {
+      "epoch": 2.889827950498038,
+      "grad_norm": 0.16015625,
+      "learning_rate": 7.555002075550021e-06,
+      "loss": 0.5874,
+      "step": 4786
+    },
+    {
+      "epoch": 2.8904316329610626,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 7.513491075134911e-06,
+      "loss": 0.7338,
+      "step": 4787
+    },
+    {
+      "epoch": 2.891035315424087,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 7.4719800747198e-06,
+      "loss": 0.6417,
+      "step": 4788
+    },
+    {
+      "epoch": 2.8916389978871115,
+      "grad_norm": 0.15625,
+      "learning_rate": 7.430469074304691e-06,
+      "loss": 0.8307,
+      "step": 4789
+    },
+    {
+      "epoch": 2.8922426803501358,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.388958073889581e-06,
+      "loss": 0.637,
+      "step": 4790
+    },
+    {
+      "epoch": 2.8928463628131604,
+      "grad_norm": 0.146484375,
+      "learning_rate": 7.3474470734744716e-06,
+      "loss": 0.6723,
+      "step": 4791
+    },
+    {
+      "epoch": 2.8934500452761847,
+      "grad_norm": 0.142578125,
+      "learning_rate": 7.305936073059361e-06,
+      "loss": 0.5976,
+      "step": 4792
+    },
+    {
+      "epoch": 2.8940537277392093,
+      "grad_norm": 0.140625,
+      "learning_rate": 7.264425072644251e-06,
+      "loss": 0.5413,
+      "step": 4793
+    },
+    {
+      "epoch": 2.8946574102022335,
+      "grad_norm": 0.150390625,
+      "learning_rate": 7.222914072229141e-06,
+      "loss": 0.6038,
+      "step": 4794
+    },
+    {
+      "epoch": 2.895261092665258,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 7.181403071814032e-06,
+      "loss": 0.5861,
+      "step": 4795
+    },
+    {
+      "epoch": 2.8958647751282824,
+      "grad_norm": 0.14453125,
+      "learning_rate": 7.139892071398921e-06,
+      "loss": 0.5549,
+      "step": 4796
+    },
+    {
+      "epoch": 2.896468457591307,
+      "grad_norm": 0.146484375,
+      "learning_rate": 7.098381070983812e-06,
+      "loss": 0.5177,
+      "step": 4797
+    },
+    {
+      "epoch": 2.8970721400543313,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 7.056870070568701e-06,
+      "loss": 0.5494,
+      "step": 4798
+    },
+    {
+      "epoch": 2.897675822517356,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.0153590701535905e-06,
+      "loss": 0.5592,
+      "step": 4799
+    },
+    {
+      "epoch": 2.8982795049803802,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 6.973848069738481e-06,
+      "loss": 0.5537,
+      "step": 4800
+    },
+    {
+      "epoch": 2.898883187443405,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.93233706932337e-06,
+      "loss": 0.4819,
+      "step": 4801
+    },
+    {
+      "epoch": 2.899486869906429,
+      "grad_norm": 0.15234375,
+      "learning_rate": 6.890826068908261e-06,
+      "loss": 0.45,
+      "step": 4802
+    },
+    {
+      "epoch": 2.900090552369454,
+      "grad_norm": 0.166015625,
+      "learning_rate": 6.849315068493151e-06,
+      "loss": 0.4768,
+      "step": 4803
+    },
+    {
+      "epoch": 2.900694234832478,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.807804068078042e-06,
+      "loss": 0.5442,
+      "step": 4804
+    },
+    {
+      "epoch": 2.9012979172955027,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.766293067662931e-06,
+      "loss": 0.4122,
+      "step": 4805
+    },
+    {
+      "epoch": 2.901901599758527,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 6.724782067247821e-06,
+      "loss": 0.5047,
+      "step": 4806
+    },
+    {
+      "epoch": 2.9025052822215516,
+      "grad_norm": 0.197265625,
+      "learning_rate": 6.683271066832711e-06,
+      "loss": 0.4761,
+      "step": 4807
+    },
+    {
+      "epoch": 2.903108964684576,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 6.641760066417602e-06,
+      "loss": 0.3971,
+      "step": 4808
+    },
+    {
+      "epoch": 2.9037126471476005,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 6.600249066002491e-06,
+      "loss": 0.3409,
+      "step": 4809
+    },
+    {
+      "epoch": 2.9043163296106247,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 6.55873806558738e-06,
+      "loss": 0.2721,
+      "step": 4810
+    },
+    {
+      "epoch": 2.9049200120736494,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 6.517227065172271e-06,
+      "loss": 0.2655,
+      "step": 4811
+    },
+    {
+      "epoch": 2.9055236945366736,
+      "grad_norm": 0.208984375,
+      "learning_rate": 6.4757160647571606e-06,
+      "loss": 0.2052,
+      "step": 4812
+    },
+    {
+      "epoch": 2.9061273769996983,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 6.434205064342051e-06,
+      "loss": 0.7075,
+      "step": 4813
+    },
+    {
+      "epoch": 2.9067310594627225,
+      "grad_norm": 0.1640625,
+      "learning_rate": 6.39269406392694e-06,
+      "loss": 0.6261,
+      "step": 4814
+    },
+    {
+      "epoch": 2.907334741925747,
+      "grad_norm": 0.14453125,
+      "learning_rate": 6.351183063511831e-06,
+      "loss": 0.5754,
+      "step": 4815
+    },
+    {
+      "epoch": 2.9079384243887714,
+      "grad_norm": 0.310546875,
+      "learning_rate": 6.309672063096721e-06,
+      "loss": 0.6827,
+      "step": 4816
+    },
+    {
+      "epoch": 2.908542106851796,
+      "grad_norm": 0.1484375,
+      "learning_rate": 6.268161062681612e-06,
+      "loss": 0.6161,
+      "step": 4817
+    },
+    {
+      "epoch": 2.9091457893148203,
+      "grad_norm": 0.16015625,
+      "learning_rate": 6.226650062266501e-06,
+      "loss": 0.65,
+      "step": 4818
+    },
+    {
+      "epoch": 2.909749471777845,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 6.1851390618513905e-06,
+      "loss": 0.8536,
+      "step": 4819
+    },
+    {
+      "epoch": 2.910353154240869,
+      "grad_norm": 0.162109375,
+      "learning_rate": 6.143628061436281e-06,
+      "loss": 0.7328,
+      "step": 4820
+    },
+    {
+      "epoch": 2.910956836703894,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 6.102117061021171e-06,
+      "loss": 0.6667,
+      "step": 4821
+    },
+    {
+      "epoch": 2.911560519166918,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 6.060606060606061e-06,
+      "loss": 0.7117,
+      "step": 4822
+    },
+    {
+      "epoch": 2.9121642016299427,
+      "grad_norm": 0.138671875,
+      "learning_rate": 6.019095060190951e-06,
+      "loss": 0.5267,
+      "step": 4823
+    },
+    {
+      "epoch": 2.912767884092967,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 5.9775840597758416e-06,
+      "loss": 0.6946,
+      "step": 4824
+    },
+    {
+      "epoch": 2.9133715665559916,
+      "grad_norm": 0.146484375,
+      "learning_rate": 5.936073059360731e-06,
+      "loss": 0.6895,
+      "step": 4825
+    },
+    {
+      "epoch": 2.913975249019016,
+      "grad_norm": 0.15234375,
+      "learning_rate": 5.8945620589456205e-06,
+      "loss": 0.6591,
+      "step": 4826
+    },
+    {
+      "epoch": 2.9145789314820405,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 5.85305105853051e-06,
+      "loss": 0.599,
+      "step": 4827
+    },
+    {
+      "epoch": 2.915182613945065,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 5.811540058115401e-06,
+      "loss": 0.6378,
+      "step": 4828
+    },
+    {
+      "epoch": 2.9157862964080894,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 5.770029057700291e-06,
+      "loss": 0.7807,
+      "step": 4829
+    },
+    {
+      "epoch": 2.9163899788711136,
+      "grad_norm": 0.146484375,
+      "learning_rate": 5.728518057285181e-06,
+      "loss": 0.6659,
+      "step": 4830
+    },
+    {
+      "epoch": 2.9169936613341383,
+      "grad_norm": 0.162109375,
+      "learning_rate": 5.687007056870071e-06,
+      "loss": 0.7076,
+      "step": 4831
+    },
+    {
+      "epoch": 2.917597343797163,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 5.645496056454961e-06,
+      "loss": 0.6585,
+      "step": 4832
+    },
+    {
+      "epoch": 2.918201026260187,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 5.603985056039851e-06,
+      "loss": 0.7424,
+      "step": 4833
+    },
+    {
+      "epoch": 2.9188047087232114,
+      "grad_norm": 0.130859375,
+      "learning_rate": 5.562474055624741e-06,
+      "loss": 0.5734,
+      "step": 4834
+    },
+    {
+      "epoch": 2.919408391186236,
+      "grad_norm": 0.154296875,
+      "learning_rate": 5.520963055209631e-06,
+      "loss": 0.5937,
+      "step": 4835
+    },
+    {
+      "epoch": 2.9200120736492607,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 5.479452054794521e-06,
+      "loss": 0.5906,
+      "step": 4836
+    },
+    {
+      "epoch": 2.920615756112285,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 5.437941054379411e-06,
+      "loss": 0.5796,
+      "step": 4837
+    },
+    {
+      "epoch": 2.921219438575309,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 5.396430053964301e-06,
+      "loss": 0.4951,
+      "step": 4838
+    },
+    {
+      "epoch": 2.921823121038334,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 5.3549190535491905e-06,
+      "loss": 0.6764,
+      "step": 4839
+    },
+    {
+      "epoch": 2.9224268035013585,
+      "grad_norm": 0.140625,
+      "learning_rate": 5.31340805313408e-06,
+      "loss": 0.7883,
+      "step": 4840
+    },
+    {
+      "epoch": 2.9230304859643828,
+      "grad_norm": 0.15625,
+      "learning_rate": 5.271897052718971e-06,
+      "loss": 0.6394,
+      "step": 4841
+    },
+    {
+      "epoch": 2.923634168427407,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 5.230386052303861e-06,
+      "loss": 0.5879,
+      "step": 4842
+    },
+    {
+      "epoch": 2.9242378508904316,
+      "grad_norm": 0.134765625,
+      "learning_rate": 5.188875051888751e-06,
+      "loss": 0.5605,
+      "step": 4843
+    },
+    {
+      "epoch": 2.9248415333534563,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 5.147364051473641e-06,
+      "loss": 0.6246,
+      "step": 4844
+    },
+    {
+      "epoch": 2.9254452158164805,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 5.105853051058531e-06,
+      "loss": 0.6152,
+      "step": 4845
+    },
+    {
+      "epoch": 2.9260488982795048,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 5.0643420506434204e-06,
+      "loss": 0.5924,
+      "step": 4846
+    },
+    {
+      "epoch": 2.9266525807425294,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 5.02283105022831e-06,
+      "loss": 0.6363,
+      "step": 4847
+    },
+    {
+      "epoch": 2.927256263205554,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 4.9813200498132e-06,
+      "loss": 0.5652,
+      "step": 4848
+    },
+    {
+      "epoch": 2.9278599456685783,
+      "grad_norm": 0.146484375,
+      "learning_rate": 4.939809049398091e-06,
+      "loss": 0.4982,
+      "step": 4849
+    },
+    {
+      "epoch": 2.9284636281316025,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 4.898298048982981e-06,
+      "loss": 0.5055,
+      "step": 4850
+    },
+    {
+      "epoch": 2.929067310594627,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 4.856787048567871e-06,
+      "loss": 0.4912,
+      "step": 4851
+    },
+    {
+      "epoch": 2.929670993057652,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.8152760481527605e-06,
+      "loss": 0.5231,
+      "step": 4852
+    },
+    {
+      "epoch": 2.930274675520676,
+      "grad_norm": 0.162109375,
+      "learning_rate": 4.773765047737651e-06,
+      "loss": 0.4411,
+      "step": 4853
+    },
+    {
+      "epoch": 2.9308783579837003,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 4.732254047322541e-06,
+      "loss": 0.5186,
+      "step": 4854
+    },
+    {
+      "epoch": 2.931482040446725,
+      "grad_norm": 0.1640625,
+      "learning_rate": 4.690743046907431e-06,
+      "loss": 0.4872,
+      "step": 4855
+    },
+    {
+      "epoch": 2.9320857229097497,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.649232046492321e-06,
+      "loss": 0.4443,
+      "step": 4856
+    },
+    {
+      "epoch": 2.932689405372774,
+      "grad_norm": 0.193359375,
+      "learning_rate": 4.607721046077211e-06,
+      "loss": 0.596,
+      "step": 4857
+    },
+    {
+      "epoch": 2.933293087835798,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 4.566210045662101e-06,
+      "loss": 0.3698,
+      "step": 4858
+    },
+    {
+      "epoch": 2.933896770298823,
+      "grad_norm": 0.1953125,
+      "learning_rate": 4.5246990452469905e-06,
+      "loss": 0.3508,
+      "step": 4859
+    },
+    {
+      "epoch": 2.9345004527618475,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 4.48318804483188e-06,
+      "loss": 0.3613,
+      "step": 4860
+    },
+    {
+      "epoch": 2.9351041352248717,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 4.44167704441677e-06,
+      "loss": 0.3906,
+      "step": 4861
+    },
+    {
+      "epoch": 2.935707817687896,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 4.400166044001661e-06,
+      "loss": 0.2025,
+      "step": 4862
+    },
+    {
+      "epoch": 2.9363115001509206,
+      "grad_norm": 0.14453125,
+      "learning_rate": 4.358655043586551e-06,
+      "loss": 0.7123,
+      "step": 4863
+    },
+    {
+      "epoch": 2.9369151826139452,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 4.317144043171441e-06,
+      "loss": 0.6143,
+      "step": 4864
+    },
+    {
+      "epoch": 2.9375188650769695,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 4.2756330427563305e-06,
+      "loss": 0.6854,
+      "step": 4865
+    },
+    {
+      "epoch": 2.9381225475399937,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 4.234122042341221e-06,
+      "loss": 0.5651,
+      "step": 4866
+    },
+    {
+      "epoch": 2.9387262300030184,
+      "grad_norm": 0.138671875,
+      "learning_rate": 4.192611041926111e-06,
+      "loss": 0.5303,
+      "step": 4867
+    },
+    {
+      "epoch": 2.939329912466043,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 4.151100041511e-06,
+      "loss": 0.6131,
+      "step": 4868
+    },
+    {
+      "epoch": 2.9399335949290673,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.10958904109589e-06,
+      "loss": 0.6331,
+      "step": 4869
+    },
+    {
+      "epoch": 2.940537277392092,
+      "grad_norm": 0.142578125,
+      "learning_rate": 4.068078040680781e-06,
+      "loss": 0.567,
+      "step": 4870
+    },
+    {
+      "epoch": 2.941140959855116,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 4.026567040265671e-06,
+      "loss": 0.6356,
+      "step": 4871
+    },
+    {
+      "epoch": 2.941744642318141,
+      "grad_norm": 0.15625,
+      "learning_rate": 3.9850560398505605e-06,
+      "loss": 0.5873,
+      "step": 4872
+    },
+    {
+      "epoch": 2.942348324781165,
+      "grad_norm": 0.13671875,
+      "learning_rate": 3.94354503943545e-06,
+      "loss": 0.7104,
+      "step": 4873
+    },
+    {
+      "epoch": 2.9429520072441897,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 3.902034039020341e-06,
+      "loss": 0.6211,
+      "step": 4874
+    },
+    {
+      "epoch": 2.943555689707214,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.860523038605231e-06,
+      "loss": 0.613,
+      "step": 4875
+    },
+    {
+      "epoch": 2.9441593721702386,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 3.819012038190121e-06,
+      "loss": 0.7952,
+      "step": 4876
+    },
+    {
+      "epoch": 2.944763054633263,
+      "grad_norm": 0.1484375,
+      "learning_rate": 3.7775010377750107e-06,
+      "loss": 0.5848,
+      "step": 4877
+    },
+    {
+      "epoch": 2.9453667370962875,
+      "grad_norm": 0.146484375,
+      "learning_rate": 3.7359900373599e-06,
+      "loss": 0.5493,
+      "step": 4878
+    },
+    {
+      "epoch": 2.9459704195593117,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 3.6944790369447904e-06,
+      "loss": 0.6622,
+      "step": 4879
+    },
+    {
+      "epoch": 2.9465741020223364,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 3.6529680365296803e-06,
+      "loss": 0.6939,
+      "step": 4880
+    },
+    {
+      "epoch": 2.9471777844853606,
+      "grad_norm": 0.14453125,
+      "learning_rate": 3.6114570361145706e-06,
+      "loss": 0.6388,
+      "step": 4881
+    },
+    {
+      "epoch": 2.9477814669483853,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 3.5699460356994605e-06,
+      "loss": 0.5472,
+      "step": 4882
+    },
+    {
+      "epoch": 2.9483851494114095,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 3.5284350352843503e-06,
+      "loss": 0.5682,
+      "step": 4883
+    },
+    {
+      "epoch": 2.948988831874434,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 3.4869240348692406e-06,
+      "loss": 0.5742,
+      "step": 4884
+    },
+    {
+      "epoch": 2.9495925143374584,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.4454130344541305e-06,
+      "loss": 0.7286,
+      "step": 4885
+    },
+    {
+      "epoch": 2.950196196800483,
+      "grad_norm": 0.14453125,
+      "learning_rate": 3.403902034039021e-06,
+      "loss": 0.7722,
+      "step": 4886
+    },
+    {
+      "epoch": 2.9507998792635073,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 3.3623910336239107e-06,
+      "loss": 0.6654,
+      "step": 4887
+    },
+    {
+      "epoch": 2.951403561726532,
+      "grad_norm": 0.146484375,
+      "learning_rate": 3.320880033208801e-06,
+      "loss": 0.6901,
+      "step": 4888
+    },
+    {
+      "epoch": 2.952007244189556,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 3.27936903279369e-06,
+      "loss": 0.637,
+      "step": 4889
+    },
+    {
+      "epoch": 2.952610926652581,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 3.2378580323785803e-06,
+      "loss": 0.7953,
+      "step": 4890
+    },
+    {
+      "epoch": 2.953214609115605,
+      "grad_norm": 0.1328125,
+      "learning_rate": 3.19634703196347e-06,
+      "loss": 0.5706,
+      "step": 4891
+    },
+    {
+      "epoch": 2.9538182915786297,
+      "grad_norm": 0.296875,
+      "learning_rate": 3.1548360315483604e-06,
+      "loss": 0.6715,
+      "step": 4892
+    },
+    {
+      "epoch": 2.954421974041654,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 3.1133250311332503e-06,
+      "loss": 0.5795,
+      "step": 4893
+    },
+    {
+      "epoch": 2.9550256565046786,
+      "grad_norm": 0.142578125,
+      "learning_rate": 3.0718140307181406e-06,
+      "loss": 0.5772,
+      "step": 4894
+    },
+    {
+      "epoch": 2.955629338967703,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 3.0303030303030305e-06,
+      "loss": 0.7201,
+      "step": 4895
+    },
+    {
+      "epoch": 2.9562330214307275,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.9887920298879208e-06,
+      "loss": 0.6037,
+      "step": 4896
+    },
+    {
+      "epoch": 2.9568367038937517,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.9472810294728102e-06,
+      "loss": 0.5457,
+      "step": 4897
+    },
+    {
+      "epoch": 2.9574403863567764,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.9057700290577005e-06,
+      "loss": 0.6012,
+      "step": 4898
+    },
+    {
+      "epoch": 2.9580440688198006,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 2.8642590286425904e-06,
+      "loss": 0.4818,
+      "step": 4899
+    },
+    {
+      "epoch": 2.9586477512828253,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.8227480282274807e-06,
+      "loss": 0.4889,
+      "step": 4900
+    },
+    {
+      "epoch": 2.9592514337458495,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.7812370278123706e-06,
+      "loss": 0.6557,
+      "step": 4901
+    },
+    {
+      "epoch": 2.959855116208874,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.7397260273972604e-06,
+      "loss": 0.4602,
+      "step": 4902
+    },
+    {
+      "epoch": 2.9604587986718984,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.6982150269821503e-06,
+      "loss": 0.514,
+      "step": 4903
+    },
+    {
+      "epoch": 2.961062481134923,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.65670402656704e-06,
+      "loss": 0.4437,
+      "step": 4904
+    },
+    {
+      "epoch": 2.9616661635979473,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.6151930261519305e-06,
+      "loss": 0.4955,
+      "step": 4905
+    },
+    {
+      "epoch": 2.962269846060972,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.5736820257368203e-06,
+      "loss": 0.4478,
+      "step": 4906
+    },
+    {
+      "epoch": 2.962873528523996,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 2.5321710253217102e-06,
+      "loss": 0.3539,
+      "step": 4907
+    },
+    {
+      "epoch": 2.963477210987021,
+      "grad_norm": 0.232421875,
+      "learning_rate": 2.4906600249066e-06,
+      "loss": 0.3473,
+      "step": 4908
+    },
+    {
+      "epoch": 2.9640808934500456,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 2.4491490244914904e-06,
+      "loss": 0.3788,
+      "step": 4909
+    },
+    {
+      "epoch": 2.9646845759130698,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.4076380240763803e-06,
+      "loss": 0.3048,
+      "step": 4910
+    },
+    {
+      "epoch": 2.965288258376094,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.3661270236612705e-06,
+      "loss": 0.2741,
+      "step": 4911
+    },
+    {
+      "epoch": 2.9658919408391187,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.3246160232461604e-06,
+      "loss": 0.2138,
+      "step": 4912
+    },
+    {
+      "epoch": 2.9664956233021433,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 2.2831050228310503e-06,
+      "loss": 0.8374,
+      "step": 4913
+    },
+    {
+      "epoch": 2.9670993057651676,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.24159402241594e-06,
+      "loss": 0.5438,
+      "step": 4914
+    },
+    {
+      "epoch": 2.967702988228192,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.2000830220008305e-06,
+      "loss": 0.5951,
+      "step": 4915
+    },
+    {
+      "epoch": 2.9683066706912165,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.1585720215857203e-06,
+      "loss": 0.9442,
+      "step": 4916
+    },
+    {
+      "epoch": 2.968910353154241,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.1170610211706106e-06,
+      "loss": 0.5268,
+      "step": 4917
+    },
+    {
+      "epoch": 2.9695140356172653,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.0755500207555e-06,
+      "loss": 0.6036,
+      "step": 4918
+    },
+    {
+      "epoch": 2.9701177180802896,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 2.0340390203403904e-06,
+      "loss": 0.7207,
+      "step": 4919
+    },
+    {
+      "epoch": 2.9707214005433142,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 1.9925280199252802e-06,
+      "loss": 0.8401,
+      "step": 4920
+    },
+    {
+      "epoch": 2.971325083006339,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.9510170195101705e-06,
+      "loss": 0.6123,
+      "step": 4921
+    },
+    {
+      "epoch": 2.971928765469363,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.9095060190950604e-06,
+      "loss": 0.6068,
+      "step": 4922
+    },
+    {
+      "epoch": 2.9725324479323874,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.86799501867995e-06,
+      "loss": 0.8205,
+      "step": 4923
+    },
+    {
+      "epoch": 2.973136130395412,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.8264840182648401e-06,
+      "loss": 0.6167,
+      "step": 4924
+    },
+    {
+      "epoch": 2.9737398128584367,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.7849730178497302e-06,
+      "loss": 0.5753,
+      "step": 4925
+    },
+    {
+      "epoch": 2.974343495321461,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.7434620174346203e-06,
+      "loss": 0.558,
+      "step": 4926
+    },
+    {
+      "epoch": 2.974947177784485,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.7019510170195104e-06,
+      "loss": 0.8501,
+      "step": 4927
+    },
+    {
+      "epoch": 2.97555086024751,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.6604400166044005e-06,
+      "loss": 0.7634,
+      "step": 4928
+    },
+    {
+      "epoch": 2.9761545427105345,
+      "grad_norm": 0.154296875,
+      "learning_rate": 1.6189290161892901e-06,
+      "loss": 0.8527,
+      "step": 4929
+    },
+    {
+      "epoch": 2.9767582251735587,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.5774180157741802e-06,
+      "loss": 0.7618,
+      "step": 4930
+    },
+    {
+      "epoch": 2.977361907636583,
+      "grad_norm": 0.142578125,
+      "learning_rate": 1.5359070153590703e-06,
+      "loss": 0.5931,
+      "step": 4931
+    },
+    {
+      "epoch": 2.9779655900996076,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.4943960149439604e-06,
+      "loss": 0.9053,
+      "step": 4932
+    },
+    {
+      "epoch": 2.9785692725626323,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.4528850145288503e-06,
+      "loss": 0.6993,
+      "step": 4933
+    },
+    {
+      "epoch": 2.9791729550256565,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.4113740141137403e-06,
+      "loss": 0.5533,
+      "step": 4934
+    },
+    {
+      "epoch": 2.9797766374886807,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.3698630136986302e-06,
+      "loss": 0.8972,
+      "step": 4935
+    },
+    {
+      "epoch": 2.9803803199517054,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.32835201328352e-06,
+      "loss": 0.5559,
+      "step": 4936
+    },
+    {
+      "epoch": 2.98098400241473,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.2868410128684102e-06,
+      "loss": 0.6695,
+      "step": 4937
+    },
+    {
+      "epoch": 2.9815876848777543,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.2453300124533e-06,
+      "loss": 0.6848,
+      "step": 4938
+    },
+    {
+      "epoch": 2.9821913673407785,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.2038190120381901e-06,
+      "loss": 0.6172,
+      "step": 4939
+    },
+    {
+      "epoch": 2.982795049803803,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.1623080116230802e-06,
+      "loss": 0.5114,
+      "step": 4940
+    },
+    {
+      "epoch": 2.983398732266828,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.12079701120797e-06,
+      "loss": 0.6146,
+      "step": 4941
+    },
+    {
+      "epoch": 2.984002414729852,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.0792860107928602e-06,
+      "loss": 0.5792,
+      "step": 4942
+    },
+    {
+      "epoch": 2.9846060971928763,
+      "grad_norm": 0.15625,
+      "learning_rate": 1.03777501037775e-06,
+      "loss": 0.7686,
+      "step": 4943
+    },
+    {
+      "epoch": 2.985209779655901,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 9.962640099626401e-07,
+      "loss": 0.6975,
+      "step": 4944
+    },
+    {
+      "epoch": 2.9858134621189256,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 9.547530095475302e-07,
+      "loss": 0.5667,
+      "step": 4945
+    },
+    {
+      "epoch": 2.98641714458195,
+      "grad_norm": 0.158203125,
+      "learning_rate": 9.132420091324201e-07,
+      "loss": 0.5319,
+      "step": 4946
+    },
+    {
+      "epoch": 2.987020827044974,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.717310087173102e-07,
+      "loss": 0.573,
+      "step": 4947
+    },
+    {
+      "epoch": 2.9876245095079987,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 8.302200083022002e-07,
+      "loss": 0.5194,
+      "step": 4948
+    },
+    {
+      "epoch": 2.9882281919710234,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 7.887090078870901e-07,
+      "loss": 0.541,
+      "step": 4949
+    },
+    {
+      "epoch": 2.9888318744340476,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 7.471980074719802e-07,
+      "loss": 0.5331,
+      "step": 4950
+    },
+    {
+      "epoch": 2.9894355568970723,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.056870070568702e-07,
+      "loss": 0.501,
+      "step": 4951
+    },
+    {
+      "epoch": 2.9900392393600965,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.6417600664176e-07,
+      "loss": 0.5099,
+      "step": 4952
+    },
+    {
+      "epoch": 2.990642921823121,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 6.2266500622665e-07,
+      "loss": 0.4793,
+      "step": 4953
+    },
+    {
+      "epoch": 2.9912466042861454,
+      "grad_norm": 0.18359375,
+      "learning_rate": 5.811540058115401e-07,
+      "loss": 0.5053,
+      "step": 4954
+    },
+    {
+      "epoch": 2.99185028674917,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 5.396430053964301e-07,
+      "loss": 0.4264,
+      "step": 4955
+    },
+    {
+      "epoch": 2.9924539692121943,
+      "grad_norm": 0.19140625,
+      "learning_rate": 4.981320049813201e-07,
+      "loss": 0.5337,
+      "step": 4956
+    },
+    {
+      "epoch": 2.993057651675219,
+      "grad_norm": 0.189453125,
+      "learning_rate": 4.5662100456621004e-07,
+      "loss": 0.4105,
+      "step": 4957
+    },
+    {
+      "epoch": 2.993661334138243,
+      "grad_norm": 0.212890625,
+      "learning_rate": 4.151100041511001e-07,
+      "loss": 0.4347,
+      "step": 4958
+    },
+    {
+      "epoch": 2.994265016601268,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.735990037359901e-07,
+      "loss": 0.3364,
+      "step": 4959
+    },
+    {
+      "epoch": 2.994868699064292,
+      "grad_norm": 0.22265625,
+      "learning_rate": 3.3208800332088e-07,
+      "loss": 0.3349,
+      "step": 4960
+    },
+    {
+      "epoch": 2.9954723815273168,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.9057700290577005e-07,
+      "loss": 0.2545,
+      "step": 4961
+    },
+    {
+      "epoch": 2.996076063990341,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.4906600249066003e-07,
+      "loss": 0.184,
+      "step": 4962
+    },
+    {
+      "epoch": 2.9966797464533657,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.0755500207555006e-07,
+      "loss": 0.6758,
+      "step": 4963
+    },
+    {
+      "epoch": 2.99728342891639,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.6604400166044e-07,
+      "loss": 0.5998,
+      "step": 4964
+    },
+    {
+      "epoch": 2.9978871113794145,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.2453300124533001e-07,
+      "loss": 0.7291,
+      "step": 4965
+    },
+    {
+      "epoch": 2.9984907938424388,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 8.302200083022e-08,
+      "loss": 0.6966,
+      "step": 4966
+    },
+    {
+      "epoch": 2.9990944763054634,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.151100041511e-08,
+      "loss": 0.5435,
+      "step": 4967
+    },
+    {
+      "epoch": 2.9996981587684877,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0,
+      "loss": 0.4412,
+      "step": 4968
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 4968,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 276,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.988555585450428e+19,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}