diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,16213 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.99675535366645,
+  "eval_steps": 500,
+  "global_step": 2310,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0012978585334198572,
+      "grad_norm": 51.281074849235615,
+      "learning_rate": 0.0,
+      "loss": 11.2228,
+      "step": 1
+    },
+    {
+      "epoch": 0.0025957170668397143,
+      "grad_norm": 52.288477767034706,
+      "learning_rate": 2.1645021645021646e-07,
+      "loss": 11.2142,
+      "step": 2
+    },
+    {
+      "epoch": 0.003893575600259572,
+      "grad_norm": 51.690439395204805,
+      "learning_rate": 4.329004329004329e-07,
+      "loss": 11.2982,
+      "step": 3
+    },
+    {
+      "epoch": 0.005191434133679429,
+      "grad_norm": 52.70272970065567,
+      "learning_rate": 6.493506493506494e-07,
+      "loss": 11.2202,
+      "step": 4
+    },
+    {
+      "epoch": 0.006489292667099286,
+      "grad_norm": 52.19249689221791,
+      "learning_rate": 8.658008658008658e-07,
+      "loss": 11.223,
+      "step": 5
+    },
+    {
+      "epoch": 0.007787151200519144,
+      "grad_norm": 52.51499888342824,
+      "learning_rate": 1.0822510822510822e-06,
+      "loss": 11.1764,
+      "step": 6
+    },
+    {
+      "epoch": 0.009085009733939001,
+      "grad_norm": 54.65313049626493,
+      "learning_rate": 1.2987012987012988e-06,
+      "loss": 11.1836,
+      "step": 7
+    },
+    {
+      "epoch": 0.010382868267358857,
+      "grad_norm": 56.26332260448297,
+      "learning_rate": 1.5151515151515152e-06,
+      "loss": 10.9434,
+      "step": 8
+    },
+    {
+      "epoch": 0.011680726800778715,
+      "grad_norm": 56.526613959934075,
+      "learning_rate": 1.7316017316017317e-06,
+      "loss": 10.8116,
+      "step": 9
+    },
+    {
+      "epoch": 0.012978585334198572,
+      "grad_norm": 73.22487510820876,
+      "learning_rate": 1.948051948051948e-06,
+      "loss": 10.0099,
+      "step": 10
+    },
+    {
+      "epoch": 0.01427644386761843,
+      "grad_norm": 78.76237638740871,
+      "learning_rate": 2.1645021645021643e-06,
+      "loss": 9.7022,
+      "step": 11
+    },
+    {
+      "epoch": 0.015574302401038288,
+      "grad_norm": 84.53260376952899,
+      "learning_rate": 2.3809523809523808e-06,
+      "loss": 9.2879,
+      "step": 12
+    },
+    {
+      "epoch": 0.016872160934458143,
+      "grad_norm": 93.89336284136559,
+      "learning_rate": 2.5974025974025976e-06,
+      "loss": 9.1068,
+      "step": 13
+    },
+    {
+      "epoch": 0.018170019467878003,
+      "grad_norm": 67.50935080000205,
+      "learning_rate": 2.813852813852814e-06,
+      "loss": 4.1366,
+      "step": 14
+    },
+    {
+      "epoch": 0.01946787800129786,
+      "grad_norm": 60.451555462271216,
+      "learning_rate": 3.0303030303030305e-06,
+      "loss": 3.7321,
+      "step": 15
+    },
+    {
+      "epoch": 0.020765736534717714,
+      "grad_norm": 48.941978109448335,
+      "learning_rate": 3.2467532467532465e-06,
+      "loss": 3.3276,
+      "step": 16
+    },
+    {
+      "epoch": 0.022063595068137574,
+      "grad_norm": 41.193189217962875,
+      "learning_rate": 3.4632034632034634e-06,
+      "loss": 3.0135,
+      "step": 17
+    },
+    {
+      "epoch": 0.02336145360155743,
+      "grad_norm": 19.806882723512313,
+      "learning_rate": 3.67965367965368e-06,
+      "loss": 2.1732,
+      "step": 18
+    },
+    {
+      "epoch": 0.02465931213497729,
+      "grad_norm": 6.601644234547365,
+      "learning_rate": 3.896103896103896e-06,
+      "loss": 1.5589,
+      "step": 19
+    },
+    {
+      "epoch": 0.025957170668397145,
+      "grad_norm": 5.420878318524882,
+      "learning_rate": 4.112554112554113e-06,
+      "loss": 1.4694,
+      "step": 20
+    },
+    {
+      "epoch": 0.027255029201817,
+      "grad_norm": 4.513535899746422,
+      "learning_rate": 4.329004329004329e-06,
+      "loss": 1.4127,
+      "step": 21
+    },
+    {
+      "epoch": 0.02855288773523686,
+      "grad_norm": 3.4399120070700926,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 1.308,
+      "step": 22
+    },
+    {
+      "epoch": 0.029850746268656716,
+      "grad_norm": 2.737903832715275,
+      "learning_rate": 4.7619047619047615e-06,
+      "loss": 1.2391,
+      "step": 23
+    },
+    {
+      "epoch": 0.031148604802076575,
+      "grad_norm": 2.229792362203531,
+      "learning_rate": 4.978354978354978e-06,
+      "loss": 1.207,
+      "step": 24
+    },
+    {
+      "epoch": 0.03244646333549643,
+      "grad_norm": 1.67506952925642,
+      "learning_rate": 5.194805194805195e-06,
+      "loss": 1.164,
+      "step": 25
+    },
+    {
+      "epoch": 0.03374432186891629,
+      "grad_norm": 7.810491533013112,
+      "learning_rate": 5.411255411255411e-06,
+      "loss": 1.0519,
+      "step": 26
+    },
+    {
+      "epoch": 0.03504218040233614,
+      "grad_norm": 1.8418087052814074,
+      "learning_rate": 5.627705627705628e-06,
+      "loss": 1.0332,
+      "step": 27
+    },
+    {
+      "epoch": 0.036340038935756006,
+      "grad_norm": 1.6946651958442733,
+      "learning_rate": 5.844155844155844e-06,
+      "loss": 1.0257,
+      "step": 28
+    },
+    {
+      "epoch": 0.03763789746917586,
+      "grad_norm": 1.0837092117736122,
+      "learning_rate": 6.060606060606061e-06,
+      "loss": 0.9995,
+      "step": 29
+    },
+    {
+      "epoch": 0.03893575600259572,
+      "grad_norm": 1.022247931246769,
+      "learning_rate": 6.277056277056277e-06,
+      "loss": 0.9658,
+      "step": 30
+    },
+    {
+      "epoch": 0.04023361453601557,
+      "grad_norm": 1.0439465825515253,
+      "learning_rate": 6.493506493506493e-06,
+      "loss": 0.9258,
+      "step": 31
+    },
+    {
+      "epoch": 0.04153147306943543,
+      "grad_norm": 0.8476038489892367,
+      "learning_rate": 6.709956709956711e-06,
+      "loss": 0.9183,
+      "step": 32
+    },
+    {
+      "epoch": 0.04282933160285529,
+      "grad_norm": 0.7260505313959857,
+      "learning_rate": 6.926406926406927e-06,
+      "loss": 0.8859,
+      "step": 33
+    },
+    {
+      "epoch": 0.04412719013627515,
+      "grad_norm": 0.9334322529996619,
+      "learning_rate": 7.142857142857143e-06,
+      "loss": 0.8775,
+      "step": 34
+    },
+    {
+      "epoch": 0.045425048669695,
+      "grad_norm": 0.7507371805560344,
+      "learning_rate": 7.35930735930736e-06,
+      "loss": 0.8506,
+      "step": 35
+    },
+    {
+      "epoch": 0.04672290720311486,
+      "grad_norm": 0.7037392218293158,
+      "learning_rate": 7.5757575757575764e-06,
+      "loss": 0.8373,
+      "step": 36
+    },
+    {
+      "epoch": 0.048020765736534715,
+      "grad_norm": 0.7026115757535957,
+      "learning_rate": 7.792207792207792e-06,
+      "loss": 0.8148,
+      "step": 37
+    },
+    {
+      "epoch": 0.04931862426995458,
+      "grad_norm": 0.6329073089403997,
+      "learning_rate": 8.008658008658008e-06,
+      "loss": 0.7989,
+      "step": 38
+    },
+    {
+      "epoch": 0.050616482803374434,
+      "grad_norm": 0.576557174722755,
+      "learning_rate": 8.225108225108225e-06,
+      "loss": 0.801,
+      "step": 39
+    },
+    {
+      "epoch": 0.05191434133679429,
+      "grad_norm": 0.633584321792007,
+      "learning_rate": 8.441558441558442e-06,
+      "loss": 0.8154,
+      "step": 40
+    },
+    {
+      "epoch": 0.053212199870214145,
+      "grad_norm": 0.6357768126157509,
+      "learning_rate": 8.658008658008657e-06,
+      "loss": 0.7985,
+      "step": 41
+    },
+    {
+      "epoch": 0.054510058403634,
+      "grad_norm": 0.4606140950872704,
+      "learning_rate": 8.874458874458876e-06,
+      "loss": 0.7875,
+      "step": 42
+    },
+    {
+      "epoch": 0.055807916937053864,
+      "grad_norm": 0.42579840291728105,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.7882,
+      "step": 43
+    },
+    {
+      "epoch": 0.05710577547047372,
+      "grad_norm": 0.5127047756782175,
+      "learning_rate": 9.307359307359308e-06,
+      "loss": 0.7668,
+      "step": 44
+    },
+    {
+      "epoch": 0.058403634003893576,
+      "grad_norm": 0.5275747680292829,
+      "learning_rate": 9.523809523809523e-06,
+      "loss": 0.7556,
+      "step": 45
+    },
+    {
+      "epoch": 0.05970149253731343,
+      "grad_norm": 0.4422307893352111,
+      "learning_rate": 9.740259740259742e-06,
+      "loss": 0.7469,
+      "step": 46
+    },
+    {
+      "epoch": 0.06099935107073329,
+      "grad_norm": 0.3950972183567316,
+      "learning_rate": 9.956709956709957e-06,
+      "loss": 0.7257,
+      "step": 47
+    },
+    {
+      "epoch": 0.06229720960415315,
+      "grad_norm": 0.4294144227066294,
+      "learning_rate": 1.0173160173160174e-05,
+      "loss": 0.7082,
+      "step": 48
+    },
+    {
+      "epoch": 0.063595068137573,
+      "grad_norm": 0.4261355016492852,
+      "learning_rate": 1.038961038961039e-05,
+      "loss": 0.7202,
+      "step": 49
+    },
+    {
+      "epoch": 0.06489292667099286,
+      "grad_norm": 0.40006881327817506,
+      "learning_rate": 1.0606060606060607e-05,
+      "loss": 0.7006,
+      "step": 50
+    },
+    {
+      "epoch": 0.06619078520441272,
+      "grad_norm": 0.3484479390008924,
+      "learning_rate": 1.0822510822510823e-05,
+      "loss": 0.7141,
+      "step": 51
+    },
+    {
+      "epoch": 0.06748864373783257,
+      "grad_norm": 0.3508582969164509,
+      "learning_rate": 1.103896103896104e-05,
+      "loss": 0.7203,
+      "step": 52
+    },
+    {
+      "epoch": 0.06878650227125244,
+      "grad_norm": 0.3754749738655716,
+      "learning_rate": 1.1255411255411256e-05,
+      "loss": 0.7465,
+      "step": 53
+    },
+    {
+      "epoch": 0.07008436080467229,
+      "grad_norm": 0.33275270814242425,
+      "learning_rate": 1.1471861471861473e-05,
+      "loss": 0.6844,
+      "step": 54
+    },
+    {
+      "epoch": 0.07138221933809215,
+      "grad_norm": 0.29711493887953333,
+      "learning_rate": 1.1688311688311688e-05,
+      "loss": 0.668,
+      "step": 55
+    },
+    {
+      "epoch": 0.07268007787151201,
+      "grad_norm": 0.3254569707924083,
+      "learning_rate": 1.1904761904761905e-05,
+      "loss": 0.7234,
+      "step": 56
+    },
+    {
+      "epoch": 0.07397793640493186,
+      "grad_norm": 0.2925311216603525,
+      "learning_rate": 1.2121212121212122e-05,
+      "loss": 0.7004,
+      "step": 57
+    },
+    {
+      "epoch": 0.07527579493835172,
+      "grad_norm": 0.2781203466736231,
+      "learning_rate": 1.2337662337662339e-05,
+      "loss": 0.6845,
+      "step": 58
+    },
+    {
+      "epoch": 0.07657365347177157,
+      "grad_norm": 0.27946888261667213,
+      "learning_rate": 1.2554112554112554e-05,
+      "loss": 0.6999,
+      "step": 59
+    },
+    {
+      "epoch": 0.07787151200519143,
+      "grad_norm": 0.2728571313678063,
+      "learning_rate": 1.2770562770562773e-05,
+      "loss": 0.6639,
+      "step": 60
+    },
+    {
+      "epoch": 0.0791693705386113,
+      "grad_norm": 0.3093993935391829,
+      "learning_rate": 1.2987012987012986e-05,
+      "loss": 0.701,
+      "step": 61
+    },
+    {
+      "epoch": 0.08046722907203115,
+      "grad_norm": 0.2852724472177098,
+      "learning_rate": 1.3203463203463205e-05,
+      "loss": 0.681,
+      "step": 62
+    },
+    {
+      "epoch": 0.08176508760545101,
+      "grad_norm": 0.2693071822601781,
+      "learning_rate": 1.3419913419913421e-05,
+      "loss": 0.6679,
+      "step": 63
+    },
+    {
+      "epoch": 0.08306294613887086,
+      "grad_norm": 0.2883803733655785,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 0.6871,
+      "step": 64
+    },
+    {
+      "epoch": 0.08436080467229072,
+      "grad_norm": 0.27168971444927753,
+      "learning_rate": 1.3852813852813853e-05,
+      "loss": 0.6478,
+      "step": 65
+    },
+    {
+      "epoch": 0.08565866320571058,
+      "grad_norm": 0.2780741659791045,
+      "learning_rate": 1.406926406926407e-05,
+      "loss": 0.6654,
+      "step": 66
+    },
+    {
+      "epoch": 0.08695652173913043,
+      "grad_norm": 0.2669958151004055,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.6494,
+      "step": 67
+    },
+    {
+      "epoch": 0.0882543802725503,
+      "grad_norm": 0.2645854701003351,
+      "learning_rate": 1.4502164502164502e-05,
+      "loss": 0.6343,
+      "step": 68
+    },
+    {
+      "epoch": 0.08955223880597014,
+      "grad_norm": 0.27977755521966374,
+      "learning_rate": 1.471861471861472e-05,
+      "loss": 0.6703,
+      "step": 69
+    },
+    {
+      "epoch": 0.09085009733939,
+      "grad_norm": 0.2701714280796314,
+      "learning_rate": 1.4935064935064936e-05,
+      "loss": 0.6657,
+      "step": 70
+    },
+    {
+      "epoch": 0.09214795587280987,
+      "grad_norm": 0.3340236352400633,
+      "learning_rate": 1.5151515151515153e-05,
+      "loss": 0.6654,
+      "step": 71
+    },
+    {
+      "epoch": 0.09344581440622972,
+      "grad_norm": 0.25125625871192836,
+      "learning_rate": 1.5367965367965366e-05,
+      "loss": 0.6829,
+      "step": 72
+    },
+    {
+      "epoch": 0.09474367293964958,
+      "grad_norm": 0.27623404854696354,
+      "learning_rate": 1.5584415584415583e-05,
+      "loss": 0.6864,
+      "step": 73
+    },
+    {
+      "epoch": 0.09604153147306943,
+      "grad_norm": 0.2855287411905892,
+      "learning_rate": 1.5800865800865803e-05,
+      "loss": 0.6669,
+      "step": 74
+    },
+    {
+      "epoch": 0.0973393900064893,
+      "grad_norm": 0.2544109696892319,
+      "learning_rate": 1.6017316017316017e-05,
+      "loss": 0.6441,
+      "step": 75
+    },
+    {
+      "epoch": 0.09863724853990916,
+      "grad_norm": 0.29021289781813303,
+      "learning_rate": 1.6233766233766234e-05,
+      "loss": 0.6664,
+      "step": 76
+    },
+    {
+      "epoch": 0.099935107073329,
+      "grad_norm": 0.26812240880351107,
+      "learning_rate": 1.645021645021645e-05,
+      "loss": 0.6391,
+      "step": 77
+    },
+    {
+      "epoch": 0.10123296560674887,
+      "grad_norm": 0.27576904300300786,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.6355,
+      "step": 78
+    },
+    {
+      "epoch": 0.10253082414016872,
+      "grad_norm": 0.2814054287273717,
+      "learning_rate": 1.6883116883116884e-05,
+      "loss": 0.6265,
+      "step": 79
+    },
+    {
+      "epoch": 0.10382868267358858,
+      "grad_norm": 0.3103049215962741,
+      "learning_rate": 1.70995670995671e-05,
+      "loss": 0.6497,
+      "step": 80
+    },
+    {
+      "epoch": 0.10512654120700844,
+      "grad_norm": 0.2728333867240695,
+      "learning_rate": 1.7316017316017315e-05,
+      "loss": 0.628,
+      "step": 81
+    },
+    {
+      "epoch": 0.10642439974042829,
+      "grad_norm": 0.29691347602771223,
+      "learning_rate": 1.7532467532467535e-05,
+      "loss": 0.6481,
+      "step": 82
+    },
+    {
+      "epoch": 0.10772225827384815,
+      "grad_norm": 0.29273954514595735,
+      "learning_rate": 1.7748917748917752e-05,
+      "loss": 0.6261,
+      "step": 83
+    },
+    {
+      "epoch": 0.109020116807268,
+      "grad_norm": 0.3074962736781368,
+      "learning_rate": 1.7965367965367965e-05,
+      "loss": 0.6299,
+      "step": 84
+    },
+    {
+      "epoch": 0.11031797534068787,
+      "grad_norm": 0.29602233662175786,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 0.6335,
+      "step": 85
+    },
+    {
+      "epoch": 0.11161583387410773,
+      "grad_norm": 0.2830666870801868,
+      "learning_rate": 1.83982683982684e-05,
+      "loss": 0.641,
+      "step": 86
+    },
+    {
+      "epoch": 0.11291369240752758,
+      "grad_norm": 0.3125259689124729,
+      "learning_rate": 1.8614718614718616e-05,
+      "loss": 0.6388,
+      "step": 87
+    },
+    {
+      "epoch": 0.11421155094094744,
+      "grad_norm": 0.26645549323423784,
+      "learning_rate": 1.8831168831168833e-05,
+      "loss": 0.6208,
+      "step": 88
+    },
+    {
+      "epoch": 0.11550940947436729,
+      "grad_norm": 0.28954783071217016,
+      "learning_rate": 1.9047619047619046e-05,
+      "loss": 0.6273,
+      "step": 89
+    },
+    {
+      "epoch": 0.11680726800778715,
+      "grad_norm": 0.28149679569001645,
+      "learning_rate": 1.9264069264069266e-05,
+      "loss": 0.6028,
+      "step": 90
+    },
+    {
+      "epoch": 0.11810512654120701,
+      "grad_norm": 0.2906262772721881,
+      "learning_rate": 1.9480519480519483e-05,
+      "loss": 0.6245,
+      "step": 91
+    },
+    {
+      "epoch": 0.11940298507462686,
+      "grad_norm": 0.2711185379042277,
+      "learning_rate": 1.9696969696969697e-05,
+      "loss": 0.6267,
+      "step": 92
+    },
+    {
+      "epoch": 0.12070084360804673,
+      "grad_norm": 0.3052664513793837,
+      "learning_rate": 1.9913419913419914e-05,
+      "loss": 0.6346,
+      "step": 93
+    },
+    {
+      "epoch": 0.12199870214146658,
+      "grad_norm": 0.29479074178005676,
+      "learning_rate": 2.012987012987013e-05,
+      "loss": 0.6255,
+      "step": 94
+    },
+    {
+      "epoch": 0.12329656067488644,
+      "grad_norm": 0.3687002197662538,
+      "learning_rate": 2.0346320346320347e-05,
+      "loss": 0.6269,
+      "step": 95
+    },
+    {
+      "epoch": 0.1245944192083063,
+      "grad_norm": 0.26974731920341294,
+      "learning_rate": 2.0562770562770564e-05,
+      "loss": 0.6355,
+      "step": 96
+    },
+    {
+      "epoch": 0.12589227774172615,
+      "grad_norm": 0.35521751114512884,
+      "learning_rate": 2.077922077922078e-05,
+      "loss": 0.6293,
+      "step": 97
+    },
+    {
+      "epoch": 0.127190136275146,
+      "grad_norm": 0.31122119266101045,
+      "learning_rate": 2.0995670995670998e-05,
+      "loss": 0.6548,
+      "step": 98
+    },
+    {
+      "epoch": 0.12848799480856588,
+      "grad_norm": 0.32784103974924345,
+      "learning_rate": 2.1212121212121215e-05,
+      "loss": 0.6409,
+      "step": 99
+    },
+    {
+      "epoch": 0.12978585334198572,
+      "grad_norm": 0.2862191321006967,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 0.6287,
+      "step": 100
+    },
+    {
+      "epoch": 0.13108371187540557,
+      "grad_norm": 0.2888970770108121,
+      "learning_rate": 2.1645021645021645e-05,
+      "loss": 0.5825,
+      "step": 101
+    },
+    {
+      "epoch": 0.13238157040882545,
+      "grad_norm": 0.27541204550524634,
+      "learning_rate": 2.1861471861471862e-05,
+      "loss": 0.6056,
+      "step": 102
+    },
+    {
+      "epoch": 0.1336794289422453,
+      "grad_norm": 0.2829745345550545,
+      "learning_rate": 2.207792207792208e-05,
+      "loss": 0.6388,
+      "step": 103
+    },
+    {
+      "epoch": 0.13497728747566515,
+      "grad_norm": 0.31335331278223877,
+      "learning_rate": 2.2294372294372296e-05,
+      "loss": 0.6149,
+      "step": 104
+    },
+    {
+      "epoch": 0.136275146009085,
+      "grad_norm": 0.26183513844983125,
+      "learning_rate": 2.2510822510822512e-05,
+      "loss": 0.598,
+      "step": 105
+    },
+    {
+      "epoch": 0.13757300454250487,
+      "grad_norm": 0.3166303353223508,
+      "learning_rate": 2.272727272727273e-05,
+      "loss": 0.6153,
+      "step": 106
+    },
+    {
+      "epoch": 0.13887086307592472,
+      "grad_norm": 0.2827827597423759,
+      "learning_rate": 2.2943722943722946e-05,
+      "loss": 0.5878,
+      "step": 107
+    },
+    {
+      "epoch": 0.14016872160934457,
+      "grad_norm": 0.27950978868403287,
+      "learning_rate": 2.3160173160173163e-05,
+      "loss": 0.6022,
+      "step": 108
+    },
+    {
+      "epoch": 0.14146658014276445,
+      "grad_norm": 0.31785506543954495,
+      "learning_rate": 2.3376623376623376e-05,
+      "loss": 0.6419,
+      "step": 109
+    },
+    {
+      "epoch": 0.1427644386761843,
+      "grad_norm": 0.2760724448320942,
+      "learning_rate": 2.3593073593073593e-05,
+      "loss": 0.5892,
+      "step": 110
+    },
+    {
+      "epoch": 0.14406229720960415,
+      "grad_norm": 0.31705667668464776,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.5828,
+      "step": 111
+    },
+    {
+      "epoch": 0.14536015574302402,
+      "grad_norm": 0.2786427147511611,
+      "learning_rate": 2.4025974025974027e-05,
+      "loss": 0.6189,
+      "step": 112
+    },
+    {
+      "epoch": 0.14665801427644387,
+      "grad_norm": 0.33800188191224245,
+      "learning_rate": 2.4242424242424244e-05,
+      "loss": 0.5867,
+      "step": 113
+    },
+    {
+      "epoch": 0.14795587280986372,
+      "grad_norm": 0.3183986863565769,
+      "learning_rate": 2.4458874458874457e-05,
+      "loss": 0.6244,
+      "step": 114
+    },
+    {
+      "epoch": 0.14925373134328357,
+      "grad_norm": 0.346611504802979,
+      "learning_rate": 2.4675324675324678e-05,
+      "loss": 0.6114,
+      "step": 115
+    },
+    {
+      "epoch": 0.15055158987670345,
+      "grad_norm": 0.3193746967683076,
+      "learning_rate": 2.4891774891774894e-05,
+      "loss": 0.5847,
+      "step": 116
+    },
+    {
+      "epoch": 0.1518494484101233,
+      "grad_norm": 0.329720331399979,
+      "learning_rate": 2.5108225108225108e-05,
+      "loss": 0.6104,
+      "step": 117
+    },
+    {
+      "epoch": 0.15314730694354314,
+      "grad_norm": 0.30497761214035857,
+      "learning_rate": 2.5324675324675325e-05,
+      "loss": 0.6147,
+      "step": 118
+    },
+    {
+      "epoch": 0.15444516547696302,
+      "grad_norm": 0.3065657873353463,
+      "learning_rate": 2.5541125541125545e-05,
+      "loss": 0.5891,
+      "step": 119
+    },
+    {
+      "epoch": 0.15574302401038287,
+      "grad_norm": 0.3040591111660935,
+      "learning_rate": 2.575757575757576e-05,
+      "loss": 0.5874,
+      "step": 120
+    },
+    {
+      "epoch": 0.15704088254380272,
+      "grad_norm": 0.3176140258251669,
+      "learning_rate": 2.5974025974025972e-05,
+      "loss": 0.5891,
+      "step": 121
+    },
+    {
+      "epoch": 0.1583387410772226,
+      "grad_norm": 0.33129130491628744,
+      "learning_rate": 2.6190476190476192e-05,
+      "loss": 0.5754,
+      "step": 122
+    },
+    {
+      "epoch": 0.15963659961064244,
+      "grad_norm": 0.3400250207622185,
+      "learning_rate": 2.640692640692641e-05,
+      "loss": 0.5927,
+      "step": 123
+    },
+    {
+      "epoch": 0.1609344581440623,
+      "grad_norm": 0.3294442929975534,
+      "learning_rate": 2.6623376623376623e-05,
+      "loss": 0.6016,
+      "step": 124
+    },
+    {
+      "epoch": 0.16223231667748214,
+      "grad_norm": 0.27952039743370355,
+      "learning_rate": 2.6839826839826843e-05,
+      "loss": 0.5674,
+      "step": 125
+    },
+    {
+      "epoch": 0.16353017521090202,
+      "grad_norm": 0.3263152361115472,
+      "learning_rate": 2.7056277056277056e-05,
+      "loss": 0.6185,
+      "step": 126
+    },
+    {
+      "epoch": 0.16482803374432187,
+      "grad_norm": 0.34561117525982527,
+      "learning_rate": 2.7272727272727273e-05,
+      "loss": 0.6003,
+      "step": 127
+    },
+    {
+      "epoch": 0.16612589227774172,
+      "grad_norm": 0.36330136868220264,
+      "learning_rate": 2.7489177489177493e-05,
+      "loss": 0.5796,
+      "step": 128
+    },
+    {
+      "epoch": 0.1674237508111616,
+      "grad_norm": 0.3448144052747857,
+      "learning_rate": 2.7705627705627707e-05,
+      "loss": 0.5858,
+      "step": 129
+    },
+    {
+      "epoch": 0.16872160934458144,
+      "grad_norm": 0.30841385505522906,
+      "learning_rate": 2.792207792207792e-05,
+      "loss": 0.5913,
+      "step": 130
+    },
+    {
+      "epoch": 0.1700194678780013,
+      "grad_norm": 0.3823986000835476,
+      "learning_rate": 2.813852813852814e-05,
+      "loss": 0.6089,
+      "step": 131
+    },
+    {
+      "epoch": 0.17131732641142117,
+      "grad_norm": 0.3183137204294537,
+      "learning_rate": 2.8354978354978357e-05,
+      "loss": 0.5974,
+      "step": 132
+    },
+    {
+      "epoch": 0.17261518494484102,
+      "grad_norm": 0.3375228791953999,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 0.5919,
+      "step": 133
+    },
+    {
+      "epoch": 0.17391304347826086,
+      "grad_norm": 0.34769553896113353,
+      "learning_rate": 2.878787878787879e-05,
+      "loss": 0.587,
+      "step": 134
+    },
+    {
+      "epoch": 0.1752109020116807,
+      "grad_norm": 0.34053830322587214,
+      "learning_rate": 2.9004329004329005e-05,
+      "loss": 0.5846,
+      "step": 135
+    },
+    {
+      "epoch": 0.1765087605451006,
+      "grad_norm": 0.3327693629121813,
+      "learning_rate": 2.922077922077922e-05,
+      "loss": 0.5929,
+      "step": 136
+    },
+    {
+      "epoch": 0.17780661907852044,
+      "grad_norm": 0.37595317145253215,
+      "learning_rate": 2.943722943722944e-05,
+      "loss": 0.5836,
+      "step": 137
+    },
+    {
+      "epoch": 0.1791044776119403,
+      "grad_norm": 0.31124901930305726,
+      "learning_rate": 2.9653679653679655e-05,
+      "loss": 0.5946,
+      "step": 138
+    },
+    {
+      "epoch": 0.18040233614536016,
+      "grad_norm": 0.41500685318923003,
+      "learning_rate": 2.9870129870129872e-05,
+      "loss": 0.599,
+      "step": 139
+    },
+    {
+      "epoch": 0.18170019467878,
+      "grad_norm": 0.4422225800744917,
+      "learning_rate": 3.0086580086580092e-05,
+      "loss": 0.6079,
+      "step": 140
+    },
+    {
+      "epoch": 0.18299805321219986,
+      "grad_norm": 0.3911349391427895,
+      "learning_rate": 3.0303030303030306e-05,
+      "loss": 0.5911,
+      "step": 141
+    },
+    {
+      "epoch": 0.18429591174561974,
+      "grad_norm": 0.3565760473012978,
+      "learning_rate": 3.051948051948052e-05,
+      "loss": 0.5874,
+      "step": 142
+    },
+    {
+      "epoch": 0.1855937702790396,
+      "grad_norm": 0.3316833578762426,
+      "learning_rate": 3.073593073593073e-05,
+      "loss": 0.5987,
+      "step": 143
+    },
+    {
+      "epoch": 0.18689162881245944,
+      "grad_norm": 0.4255792906025628,
+      "learning_rate": 3.095238095238095e-05,
+      "loss": 0.5674,
+      "step": 144
+    },
+    {
+      "epoch": 0.18818948734587929,
+      "grad_norm": 0.3111389344438918,
+      "learning_rate": 3.1168831168831166e-05,
+      "loss": 0.5916,
+      "step": 145
+    },
+    {
+      "epoch": 0.18948734587929916,
+      "grad_norm": 0.40391893328316164,
+      "learning_rate": 3.1385281385281387e-05,
+      "loss": 0.5862,
+      "step": 146
+    },
+    {
+      "epoch": 0.190785204412719,
+      "grad_norm": 0.3571856870514297,
+      "learning_rate": 3.160173160173161e-05,
+      "loss": 0.5783,
+      "step": 147
+    },
+    {
+      "epoch": 0.19208306294613886,
+      "grad_norm": 0.34724535128608686,
+      "learning_rate": 3.181818181818182e-05,
+      "loss": 0.593,
+      "step": 148
+    },
+    {
+      "epoch": 0.19338092147955874,
+      "grad_norm": 0.36623311715616075,
+      "learning_rate": 3.2034632034632034e-05,
+      "loss": 0.5791,
+      "step": 149
+    },
+    {
+      "epoch": 0.1946787800129786,
+      "grad_norm": 0.35421377131407383,
+      "learning_rate": 3.2251082251082254e-05,
+      "loss": 0.5869,
+      "step": 150
+    },
+    {
+      "epoch": 0.19597663854639844,
+      "grad_norm": 0.3580175565804796,
+      "learning_rate": 3.246753246753247e-05,
+      "loss": 0.5731,
+      "step": 151
+    },
+    {
+      "epoch": 0.1972744970798183,
+      "grad_norm": 0.3779107544260428,
+      "learning_rate": 3.268398268398268e-05,
+      "loss": 0.5888,
+      "step": 152
+    },
+    {
+      "epoch": 0.19857235561323816,
+      "grad_norm": 0.381401724832965,
+      "learning_rate": 3.29004329004329e-05,
+      "loss": 0.5754,
+      "step": 153
+    },
+    {
+      "epoch": 0.199870214146658,
+      "grad_norm": 0.3996699371198549,
+      "learning_rate": 3.311688311688312e-05,
+      "loss": 0.5878,
+      "step": 154
+    },
+    {
+      "epoch": 0.20116807268007786,
+      "grad_norm": 0.3498521285804811,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.5624,
+      "step": 155
+    },
+    {
+      "epoch": 0.20246593121349774,
+      "grad_norm": 0.4329402753533996,
+      "learning_rate": 3.3549783549783555e-05,
+      "loss": 0.5823,
+      "step": 156
+    },
+    {
+      "epoch": 0.20376378974691758,
+      "grad_norm": 0.4715275117713498,
+      "learning_rate": 3.376623376623377e-05,
+      "loss": 0.5679,
+      "step": 157
+    },
+    {
+      "epoch": 0.20506164828033743,
+      "grad_norm": 0.4087297995001702,
+      "learning_rate": 3.398268398268398e-05,
+      "loss": 0.5492,
+      "step": 158
+    },
+    {
+      "epoch": 0.2063595068137573,
+      "grad_norm": 0.3963104181486302,
+      "learning_rate": 3.41991341991342e-05,
+      "loss": 0.5779,
+      "step": 159
+    },
+    {
+      "epoch": 0.20765736534717716,
+      "grad_norm": 0.4626784659383467,
+      "learning_rate": 3.4415584415584416e-05,
+      "loss": 0.5791,
+      "step": 160
+    },
+    {
+      "epoch": 0.208955223880597,
+      "grad_norm": 0.4629189431646934,
+      "learning_rate": 3.463203463203463e-05,
+      "loss": 0.5709,
+      "step": 161
+    },
+    {
+      "epoch": 0.21025308241401688,
+      "grad_norm": 0.4327284524192223,
+      "learning_rate": 3.484848484848485e-05,
+      "loss": 0.5821,
+      "step": 162
+    },
+    {
+      "epoch": 0.21155094094743673,
+      "grad_norm": 0.42226923421652224,
+      "learning_rate": 3.506493506493507e-05,
+      "loss": 0.579,
+      "step": 163
+    },
+    {
+      "epoch": 0.21284879948085658,
+      "grad_norm": 0.37986989822155737,
+      "learning_rate": 3.528138528138528e-05,
+      "loss": 0.5656,
+      "step": 164
+    },
+    {
+      "epoch": 0.21414665801427643,
+      "grad_norm": 0.4629547655665463,
+      "learning_rate": 3.5497835497835503e-05,
+      "loss": 0.5703,
+      "step": 165
+    },
+    {
+      "epoch": 0.2154445165476963,
+      "grad_norm": 0.41674661311211725,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 0.5775,
+      "step": 166
+    },
+    {
+      "epoch": 0.21674237508111616,
+      "grad_norm": 0.37812170353301494,
+      "learning_rate": 3.593073593073593e-05,
+      "loss": 0.5647,
+      "step": 167
+    },
+    {
+      "epoch": 0.218040233614536,
+      "grad_norm": 0.3533683709945352,
+      "learning_rate": 3.6147186147186144e-05,
+      "loss": 0.5742,
+      "step": 168
+    },
+    {
+      "epoch": 0.21933809214795588,
+      "grad_norm": 0.3327311378407231,
+      "learning_rate": 3.6363636363636364e-05,
+      "loss": 0.5616,
+      "step": 169
+    },
+    {
+      "epoch": 0.22063595068137573,
+      "grad_norm": 0.31532044110811386,
+      "learning_rate": 3.6580086580086584e-05,
+      "loss": 0.5434,
+      "step": 170
+    },
+    {
+      "epoch": 0.22193380921479558,
+      "grad_norm": 0.3191051407668251,
+      "learning_rate": 3.67965367965368e-05,
+      "loss": 0.5719,
+      "step": 171
+    },
+    {
+      "epoch": 0.22323166774821546,
+      "grad_norm": 0.36508138364995835,
+      "learning_rate": 3.701298701298702e-05,
+      "loss": 0.5496,
+      "step": 172
+    },
+    {
+      "epoch": 0.2245295262816353,
+      "grad_norm": 0.35917301960844417,
+      "learning_rate": 3.722943722943723e-05,
+      "loss": 0.5599,
+      "step": 173
+    },
+    {
+      "epoch": 0.22582738481505515,
+      "grad_norm": 0.332146935347223,
+      "learning_rate": 3.7445887445887445e-05,
+      "loss": 0.5729,
+      "step": 174
+    },
+    {
+      "epoch": 0.227125243348475,
+      "grad_norm": 0.3761709742507644,
+      "learning_rate": 3.7662337662337665e-05,
+      "loss": 0.5395,
+      "step": 175
+    },
+    {
+      "epoch": 0.22842310188189488,
+      "grad_norm": 0.357015737631827,
+      "learning_rate": 3.787878787878788e-05,
+      "loss": 0.5661,
+      "step": 176
+    },
+    {
+      "epoch": 0.22972096041531473,
+      "grad_norm": 0.34903223053324706,
+      "learning_rate": 3.809523809523809e-05,
+      "loss": 0.5633,
+      "step": 177
+    },
+    {
+      "epoch": 0.23101881894873458,
+      "grad_norm": 0.3592680565530814,
+      "learning_rate": 3.831168831168831e-05,
+      "loss": 0.5504,
+      "step": 178
+    },
+    {
+      "epoch": 0.23231667748215445,
+      "grad_norm": 0.31763219861115954,
+      "learning_rate": 3.852813852813853e-05,
+      "loss": 0.5693,
+      "step": 179
+    },
+    {
+      "epoch": 0.2336145360155743,
+      "grad_norm": 0.37831351950166914,
+      "learning_rate": 3.8744588744588746e-05,
+      "loss": 0.5545,
+      "step": 180
+    },
+    {
+      "epoch": 0.23491239454899415,
+      "grad_norm": 0.3032029950520603,
+      "learning_rate": 3.8961038961038966e-05,
+      "loss": 0.5483,
+      "step": 181
+    },
+    {
+      "epoch": 0.23621025308241403,
+      "grad_norm": 0.45423832636818023,
+      "learning_rate": 3.917748917748918e-05,
+      "loss": 0.5646,
+      "step": 182
+    },
+    {
+      "epoch": 0.23750811161583388,
+      "grad_norm": 0.3885683241280637,
+      "learning_rate": 3.939393939393939e-05,
+      "loss": 0.5782,
+      "step": 183
+    },
+    {
+      "epoch": 0.23880597014925373,
+      "grad_norm": 0.4113001708391826,
+      "learning_rate": 3.9610389610389614e-05,
+      "loss": 0.5709,
+      "step": 184
+    },
+    {
+      "epoch": 0.24010382868267358,
+      "grad_norm": 0.43060691545097807,
+      "learning_rate": 3.982683982683983e-05,
+      "loss": 0.5357,
+      "step": 185
+    },
+    {
+      "epoch": 0.24140168721609345,
+      "grad_norm": 0.48621329563873417,
+      "learning_rate": 4.004329004329004e-05,
+      "loss": 0.5438,
+      "step": 186
+    },
+    {
+      "epoch": 0.2426995457495133,
+      "grad_norm": 0.346819520203559,
+      "learning_rate": 4.025974025974026e-05,
+      "loss": 0.5448,
+      "step": 187
+    },
+    {
+      "epoch": 0.24399740428293315,
+      "grad_norm": 0.5771040244138606,
+      "learning_rate": 4.047619047619048e-05,
+      "loss": 0.5609,
+      "step": 188
+    },
+    {
+      "epoch": 0.24529526281635303,
+      "grad_norm": 0.5691856093486398,
+      "learning_rate": 4.0692640692640695e-05,
+      "loss": 0.5509,
+      "step": 189
+    },
+    {
+      "epoch": 0.24659312134977288,
+      "grad_norm": 0.5658078548327832,
+      "learning_rate": 4.0909090909090915e-05,
+      "loss": 0.5457,
+      "step": 190
+    },
+    {
+      "epoch": 0.24789097988319272,
+      "grad_norm": 0.32565141992028185,
+      "learning_rate": 4.112554112554113e-05,
+      "loss": 0.5576,
+      "step": 191
+    },
+    {
+      "epoch": 0.2491888384166126,
+      "grad_norm": 0.693035116005457,
+      "learning_rate": 4.134199134199134e-05,
+      "loss": 0.5818,
+      "step": 192
+    },
+    {
+      "epoch": 0.25048669695003245,
+      "grad_norm": 0.5767521454545272,
+      "learning_rate": 4.155844155844156e-05,
+      "loss": 0.5651,
+      "step": 193
+    },
+    {
+      "epoch": 0.2517845554834523,
+      "grad_norm": 0.5780821207088752,
+      "learning_rate": 4.1774891774891775e-05,
+      "loss": 0.569,
+      "step": 194
+    },
+    {
+      "epoch": 0.25308241401687215,
+      "grad_norm": 0.37604239901597153,
+      "learning_rate": 4.1991341991341996e-05,
+      "loss": 0.543,
+      "step": 195
+    },
+    {
+      "epoch": 0.254380272550292,
+      "grad_norm": 0.5156588377708116,
+      "learning_rate": 4.220779220779221e-05,
+      "loss": 0.5655,
+      "step": 196
+    },
+    {
+      "epoch": 0.2556781310837119,
+      "grad_norm": 0.547020541236707,
+      "learning_rate": 4.242424242424243e-05,
+      "loss": 0.5823,
+      "step": 197
+    },
+    {
+      "epoch": 0.25697598961713175,
+      "grad_norm": 0.4902045464021542,
+      "learning_rate": 4.264069264069264e-05,
+      "loss": 0.5819,
+      "step": 198
+    },
+    {
+      "epoch": 0.2582738481505516,
+      "grad_norm": 0.43892225186858413,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.5407,
+      "step": 199
+    },
+    {
+      "epoch": 0.25957170668397145,
+      "grad_norm": 0.35311657422045256,
+      "learning_rate": 4.3073593073593077e-05,
+      "loss": 0.5333,
+      "step": 200
+    },
+    {
+      "epoch": 0.2608695652173913,
+      "grad_norm": 0.550191337628468,
+      "learning_rate": 4.329004329004329e-05,
+      "loss": 0.5691,
+      "step": 201
+    },
+    {
+      "epoch": 0.26216742375081115,
+      "grad_norm": 0.3252801360738619,
+      "learning_rate": 4.3506493506493503e-05,
+      "loss": 0.5533,
+      "step": 202
+    },
+    {
+      "epoch": 0.263465282284231,
+      "grad_norm": 0.4553304513015423,
+      "learning_rate": 4.3722943722943724e-05,
+      "loss": 0.5565,
+      "step": 203
+    },
+    {
+      "epoch": 0.2647631408176509,
+      "grad_norm": 0.3263307722581273,
+      "learning_rate": 4.3939393939393944e-05,
+      "loss": 0.5671,
+      "step": 204
+    },
+    {
+      "epoch": 0.26606099935107075,
+      "grad_norm": 0.4000844274004943,
+      "learning_rate": 4.415584415584416e-05,
+      "loss": 0.5399,
+      "step": 205
+    },
+    {
+      "epoch": 0.2673588578844906,
+      "grad_norm": 0.38431545582799964,
+      "learning_rate": 4.437229437229438e-05,
+      "loss": 0.5417,
+      "step": 206
+    },
+    {
+      "epoch": 0.26865671641791045,
+      "grad_norm": 0.40816346897613404,
+      "learning_rate": 4.458874458874459e-05,
+      "loss": 0.5592,
+      "step": 207
+    },
+    {
+      "epoch": 0.2699545749513303,
+      "grad_norm": 0.34584489381728045,
+      "learning_rate": 4.4805194805194805e-05,
+      "loss": 0.5438,
+      "step": 208
+    },
+    {
+      "epoch": 0.27125243348475014,
+      "grad_norm": 0.4537307823944973,
+      "learning_rate": 4.5021645021645025e-05,
+      "loss": 0.5399,
+      "step": 209
+    },
+    {
+      "epoch": 0.27255029201817,
+      "grad_norm": 0.5752635325535591,
+      "learning_rate": 4.523809523809524e-05,
+      "loss": 0.5672,
+      "step": 210
+    },
+    {
+      "epoch": 0.2738481505515899,
+      "grad_norm": 0.4083818210095887,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 0.5617,
+      "step": 211
+    },
+    {
+      "epoch": 0.27514600908500975,
+      "grad_norm": 0.3170399657064755,
+      "learning_rate": 4.567099567099568e-05,
+      "loss": 0.5352,
+      "step": 212
+    },
+    {
+      "epoch": 0.2764438676184296,
+      "grad_norm": 0.31917717130826856,
+      "learning_rate": 4.588744588744589e-05,
+      "loss": 0.5617,
+      "step": 213
+    },
+    {
+      "epoch": 0.27774172615184944,
+      "grad_norm": 0.3772448329589651,
+      "learning_rate": 4.6103896103896106e-05,
+      "loss": 0.5662,
+      "step": 214
+    },
+    {
+      "epoch": 0.2790395846852693,
+      "grad_norm": 0.3799776585928483,
+      "learning_rate": 4.6320346320346326e-05,
+      "loss": 0.5814,
+      "step": 215
+    },
+    {
+      "epoch": 0.28033744321868914,
+      "grad_norm": 0.376252486652199,
+      "learning_rate": 4.653679653679654e-05,
+      "loss": 0.5475,
+      "step": 216
+    },
+    {
+      "epoch": 0.28163530175210905,
+      "grad_norm": 0.3703755301826731,
+      "learning_rate": 4.675324675324675e-05,
+      "loss": 0.5488,
+      "step": 217
+    },
+    {
+      "epoch": 0.2829331602855289,
+      "grad_norm": 0.32861168704985866,
+      "learning_rate": 4.696969696969697e-05,
+      "loss": 0.554,
+      "step": 218
+    },
+    {
+      "epoch": 0.28423101881894874,
+      "grad_norm": 0.3475845025879547,
+      "learning_rate": 4.718614718614719e-05,
+      "loss": 0.5407,
+      "step": 219
+    },
+    {
+      "epoch": 0.2855288773523686,
+      "grad_norm": 0.3648655973805309,
+      "learning_rate": 4.740259740259741e-05,
+      "loss": 0.5466,
+      "step": 220
+    },
+    {
+      "epoch": 0.28682673588578844,
+      "grad_norm": 0.3350866523035428,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 0.5548,
+      "step": 221
+    },
+    {
+      "epoch": 0.2881245944192083,
+      "grad_norm": 0.43767143054287594,
+      "learning_rate": 4.783549783549784e-05,
+      "loss": 0.5684,
+      "step": 222
+    },
+    {
+      "epoch": 0.28942245295262814,
+      "grad_norm": 0.421178133286777,
+      "learning_rate": 4.8051948051948054e-05,
+      "loss": 0.5639,
+      "step": 223
+    },
+    {
+      "epoch": 0.29072031148604804,
+      "grad_norm": 0.37835877083504477,
+      "learning_rate": 4.826839826839827e-05,
+      "loss": 0.5566,
+      "step": 224
+    },
+    {
+      "epoch": 0.2920181700194679,
+      "grad_norm": 0.3417724143733512,
+      "learning_rate": 4.848484848484849e-05,
+      "loss": 0.5552,
+      "step": 225
+    },
+    {
+      "epoch": 0.29331602855288774,
+      "grad_norm": 0.3870541340632366,
+      "learning_rate": 4.87012987012987e-05,
+      "loss": 0.549,
+      "step": 226
+    },
+    {
+      "epoch": 0.2946138870863076,
+      "grad_norm": 0.4889598386044001,
+      "learning_rate": 4.8917748917748915e-05,
+      "loss": 0.5538,
+      "step": 227
+    },
+    {
+      "epoch": 0.29591174561972744,
+      "grad_norm": 0.4543222558469965,
+      "learning_rate": 4.9134199134199135e-05,
+      "loss": 0.5651,
+      "step": 228
+    },
+    {
+      "epoch": 0.2972096041531473,
+      "grad_norm": 0.38147571297168936,
+      "learning_rate": 4.9350649350649355e-05,
+      "loss": 0.5456,
+      "step": 229
+    },
+    {
+      "epoch": 0.29850746268656714,
+      "grad_norm": 0.48062886052178266,
+      "learning_rate": 4.956709956709957e-05,
+      "loss": 0.5519,
+      "step": 230
+    },
+    {
+      "epoch": 0.29980532121998704,
+      "grad_norm": 0.3436776708584428,
+      "learning_rate": 4.978354978354979e-05,
+      "loss": 0.5572,
+      "step": 231
+    },
+    {
+      "epoch": 0.3011031797534069,
+      "grad_norm": 0.48075516118965306,
+      "learning_rate": 5e-05,
+      "loss": 0.5566,
+      "step": 232
+    },
+    {
+      "epoch": 0.30240103828682674,
+      "grad_norm": 0.5764784128417795,
+      "learning_rate": 4.997594997594998e-05,
+      "loss": 0.5819,
+      "step": 233
+    },
+    {
+      "epoch": 0.3036988968202466,
+      "grad_norm": 0.396476527818061,
+      "learning_rate": 4.995189995189995e-05,
+      "loss": 0.5411,
+      "step": 234
+    },
+    {
+      "epoch": 0.30499675535366644,
+      "grad_norm": 0.46291378630567925,
+      "learning_rate": 4.992784992784993e-05,
+      "loss": 0.5552,
+      "step": 235
+    },
+    {
+      "epoch": 0.3062946138870863,
+      "grad_norm": 0.44861478710130637,
+      "learning_rate": 4.990379990379991e-05,
+      "loss": 0.544,
+      "step": 236
+    },
+    {
+      "epoch": 0.3075924724205062,
+      "grad_norm": 0.3873611746053732,
+      "learning_rate": 4.987974987974988e-05,
+      "loss": 0.5556,
+      "step": 237
+    },
+    {
+      "epoch": 0.30889033095392604,
+      "grad_norm": 0.41664468323948045,
+      "learning_rate": 4.985569985569986e-05,
+      "loss": 0.5771,
+      "step": 238
+    },
+    {
+      "epoch": 0.3101881894873459,
+      "grad_norm": 0.3859756658142421,
+      "learning_rate": 4.983164983164983e-05,
+      "loss": 0.532,
+      "step": 239
+    },
+    {
+      "epoch": 0.31148604802076574,
+      "grad_norm": 0.44937677319362224,
+      "learning_rate": 4.980759980759981e-05,
+      "loss": 0.5426,
+      "step": 240
+    },
+    {
+      "epoch": 0.3127839065541856,
+      "grad_norm": 0.4349437401518082,
+      "learning_rate": 4.978354978354979e-05,
+      "loss": 0.5259,
+      "step": 241
+    },
+    {
+      "epoch": 0.31408176508760544,
+      "grad_norm": 0.400324790012705,
+      "learning_rate": 4.9759499759499764e-05,
+      "loss": 0.5537,
+      "step": 242
+    },
+    {
+      "epoch": 0.3153796236210253,
+      "grad_norm": 0.43872297509664254,
+      "learning_rate": 4.973544973544973e-05,
+      "loss": 0.549,
+      "step": 243
+    },
+    {
+      "epoch": 0.3166774821544452,
+      "grad_norm": 0.4653708053643151,
+      "learning_rate": 4.971139971139971e-05,
+      "loss": 0.5254,
+      "step": 244
+    },
+    {
+      "epoch": 0.31797534068786504,
+      "grad_norm": 0.40941811760654495,
+      "learning_rate": 4.968734968734969e-05,
+      "loss": 0.568,
+      "step": 245
+    },
+    {
+      "epoch": 0.3192731992212849,
+      "grad_norm": 0.5368348479077355,
+      "learning_rate": 4.966329966329967e-05,
+      "loss": 0.5476,
+      "step": 246
+    },
+    {
+      "epoch": 0.32057105775470474,
+      "grad_norm": 0.4544642184839637,
+      "learning_rate": 4.963924963924964e-05,
+      "loss": 0.5566,
+      "step": 247
+    },
+    {
+      "epoch": 0.3218689162881246,
+      "grad_norm": 0.42978031279738266,
+      "learning_rate": 4.961519961519962e-05,
+      "loss": 0.548,
+      "step": 248
+    },
+    {
+      "epoch": 0.32316677482154443,
+      "grad_norm": 0.41191622365654873,
+      "learning_rate": 4.9591149591149594e-05,
+      "loss": 0.5458,
+      "step": 249
+    },
+    {
+      "epoch": 0.3244646333549643,
+      "grad_norm": 0.6074054124348204,
+      "learning_rate": 4.956709956709957e-05,
+      "loss": 0.5519,
+      "step": 250
+    },
+    {
+      "epoch": 0.3257624918883842,
+      "grad_norm": 0.4651053481351256,
+      "learning_rate": 4.9543049543049543e-05,
+      "loss": 0.5811,
+      "step": 251
+    },
+    {
+      "epoch": 0.32706035042180404,
+      "grad_norm": 0.4240240962916135,
+      "learning_rate": 4.951899951899952e-05,
+      "loss": 0.5523,
+      "step": 252
+    },
+    {
+      "epoch": 0.3283582089552239,
+      "grad_norm": 0.5066208761057746,
+      "learning_rate": 4.94949494949495e-05,
+      "loss": 0.544,
+      "step": 253
+    },
+    {
+      "epoch": 0.32965606748864373,
+      "grad_norm": 0.38109072762259544,
+      "learning_rate": 4.9470899470899475e-05,
+      "loss": 0.5538,
+      "step": 254
+    },
+    {
+      "epoch": 0.3309539260220636,
+      "grad_norm": 0.5117807003713138,
+      "learning_rate": 4.944684944684945e-05,
+      "loss": 0.5577,
+      "step": 255
+    },
+    {
+      "epoch": 0.33225178455548343,
+      "grad_norm": 0.44912086500472626,
+      "learning_rate": 4.9422799422799424e-05,
+      "loss": 0.5495,
+      "step": 256
+    },
+    {
+      "epoch": 0.33354964308890334,
+      "grad_norm": 0.3651331486905666,
+      "learning_rate": 4.93987493987494e-05,
+      "loss": 0.5631,
+      "step": 257
+    },
+    {
+      "epoch": 0.3348475016223232,
+      "grad_norm": 0.5611125950484844,
+      "learning_rate": 4.937469937469938e-05,
+      "loss": 0.5465,
+      "step": 258
+    },
+    {
+      "epoch": 0.33614536015574303,
+      "grad_norm": 0.5300284860526002,
+      "learning_rate": 4.9350649350649355e-05,
+      "loss": 0.5425,
+      "step": 259
+    },
+    {
+      "epoch": 0.3374432186891629,
+      "grad_norm": 0.42241934122178765,
+      "learning_rate": 4.932659932659932e-05,
+      "loss": 0.5613,
+      "step": 260
+    },
+    {
+      "epoch": 0.33874107722258273,
+      "grad_norm": 0.6480707951702842,
+      "learning_rate": 4.9302549302549305e-05,
+      "loss": 0.5443,
+      "step": 261
+    },
+    {
+      "epoch": 0.3400389357560026,
+      "grad_norm": 0.5458559898285835,
+      "learning_rate": 4.927849927849928e-05,
+      "loss": 0.5362,
+      "step": 262
+    },
+    {
+      "epoch": 0.34133679428942243,
+      "grad_norm": 0.4307852761395753,
+      "learning_rate": 4.925444925444926e-05,
+      "loss": 0.5405,
+      "step": 263
+    },
+    {
+      "epoch": 0.34263465282284233,
+      "grad_norm": 0.5693990395449862,
+      "learning_rate": 4.923039923039923e-05,
+      "loss": 0.5455,
+      "step": 264
+    },
+    {
+      "epoch": 0.3439325113562622,
+      "grad_norm": 0.4427765568418805,
+      "learning_rate": 4.9206349206349204e-05,
+      "loss": 0.5475,
+      "step": 265
+    },
+    {
+      "epoch": 0.34523036988968203,
+      "grad_norm": 0.4724926699957873,
+      "learning_rate": 4.9182299182299185e-05,
+      "loss": 0.5502,
+      "step": 266
+    },
+    {
+      "epoch": 0.3465282284231019,
+      "grad_norm": 0.6296467164625645,
+      "learning_rate": 4.915824915824916e-05,
+      "loss": 0.555,
+      "step": 267
+    },
+    {
+      "epoch": 0.34782608695652173,
+      "grad_norm": 0.521325771991002,
+      "learning_rate": 4.9134199134199135e-05,
+      "loss": 0.5521,
+      "step": 268
+    },
+    {
+      "epoch": 0.3491239454899416,
+      "grad_norm": 0.4920183923356473,
+      "learning_rate": 4.911014911014911e-05,
+      "loss": 0.5589,
+      "step": 269
+    },
+    {
+      "epoch": 0.3504218040233614,
+      "grad_norm": 0.6860051439883974,
+      "learning_rate": 4.908609908609909e-05,
+      "loss": 0.5396,
+      "step": 270
+    },
+    {
+      "epoch": 0.35171966255678133,
+      "grad_norm": 0.38098025544839875,
+      "learning_rate": 4.9062049062049066e-05,
+      "loss": 0.5439,
+      "step": 271
+    },
+    {
+      "epoch": 0.3530175210902012,
+      "grad_norm": 0.545523500518169,
+      "learning_rate": 4.903799903799904e-05,
+      "loss": 0.5285,
+      "step": 272
+    },
+    {
+      "epoch": 0.35431537962362103,
+      "grad_norm": 0.4773245042110645,
+      "learning_rate": 4.9013949013949016e-05,
+      "loss": 0.5506,
+      "step": 273
+    },
+    {
+      "epoch": 0.3556132381570409,
+      "grad_norm": 0.41823644467627,
+      "learning_rate": 4.898989898989899e-05,
+      "loss": 0.5382,
+      "step": 274
+    },
+    {
+      "epoch": 0.3569110966904607,
+      "grad_norm": 0.43108861210799143,
+      "learning_rate": 4.896584896584897e-05,
+      "loss": 0.5312,
+      "step": 275
+    },
+    {
+      "epoch": 0.3582089552238806,
+      "grad_norm": 0.35256122918946825,
+      "learning_rate": 4.894179894179895e-05,
+      "loss": 0.5507,
+      "step": 276
+    },
+    {
+      "epoch": 0.3595068137573005,
+      "grad_norm": 0.641603115121163,
+      "learning_rate": 4.8917748917748915e-05,
+      "loss": 0.5494,
+      "step": 277
+    },
+    {
+      "epoch": 0.36080467229072033,
+      "grad_norm": 0.42144449610145046,
+      "learning_rate": 4.8893698893698896e-05,
+      "loss": 0.5537,
+      "step": 278
+    },
+    {
+      "epoch": 0.3621025308241402,
+      "grad_norm": 0.4421221398296794,
+      "learning_rate": 4.886964886964887e-05,
+      "loss": 0.5305,
+      "step": 279
+    },
+    {
+      "epoch": 0.36340038935756,
+      "grad_norm": 0.34904354726043524,
+      "learning_rate": 4.884559884559885e-05,
+      "loss": 0.5151,
+      "step": 280
+    },
+    {
+      "epoch": 0.3646982478909799,
+      "grad_norm": 0.567323138161088,
+      "learning_rate": 4.882154882154882e-05,
+      "loss": 0.5394,
+      "step": 281
+    },
+    {
+      "epoch": 0.3659961064243997,
+      "grad_norm": 0.4275900383373202,
+      "learning_rate": 4.8797498797498795e-05,
+      "loss": 0.5793,
+      "step": 282
+    },
+    {
+      "epoch": 0.3672939649578196,
+      "grad_norm": 0.43590374764579676,
+      "learning_rate": 4.877344877344878e-05,
+      "loss": 0.5238,
+      "step": 283
+    },
+    {
+      "epoch": 0.3685918234912395,
+      "grad_norm": 0.3806185860855704,
+      "learning_rate": 4.874939874939875e-05,
+      "loss": 0.5311,
+      "step": 284
+    },
+    {
+      "epoch": 0.3698896820246593,
+      "grad_norm": 0.36597622680635733,
+      "learning_rate": 4.8725348725348726e-05,
+      "loss": 0.5483,
+      "step": 285
+    },
+    {
+      "epoch": 0.3711875405580792,
+      "grad_norm": 0.39934249219009466,
+      "learning_rate": 4.87012987012987e-05,
+      "loss": 0.5323,
+      "step": 286
+    },
+    {
+      "epoch": 0.372485399091499,
+      "grad_norm": 0.35489673738601485,
+      "learning_rate": 4.8677248677248676e-05,
+      "loss": 0.5197,
+      "step": 287
+    },
+    {
+      "epoch": 0.3737832576249189,
+      "grad_norm": 0.35597996373456253,
+      "learning_rate": 4.865319865319866e-05,
+      "loss": 0.541,
+      "step": 288
+    },
+    {
+      "epoch": 0.3750811161583387,
+      "grad_norm": 0.30995272924377104,
+      "learning_rate": 4.862914862914863e-05,
+      "loss": 0.531,
+      "step": 289
+    },
+    {
+      "epoch": 0.37637897469175857,
+      "grad_norm": 0.3041222657562842,
+      "learning_rate": 4.860509860509861e-05,
+      "loss": 0.5263,
+      "step": 290
+    },
+    {
+      "epoch": 0.3776768332251785,
+      "grad_norm": 0.27479710316885086,
+      "learning_rate": 4.858104858104858e-05,
+      "loss": 0.5179,
+      "step": 291
+    },
+    {
+      "epoch": 0.3789746917585983,
+      "grad_norm": 0.4108809131825242,
+      "learning_rate": 4.8556998556998563e-05,
+      "loss": 0.5285,
+      "step": 292
+    },
+    {
+      "epoch": 0.3802725502920182,
+      "grad_norm": 0.3283706178094482,
+      "learning_rate": 4.853294853294854e-05,
+      "loss": 0.5485,
+      "step": 293
+    },
+    {
+      "epoch": 0.381570408825438,
+      "grad_norm": 0.3628325275789365,
+      "learning_rate": 4.8508898508898506e-05,
+      "loss": 0.5342,
+      "step": 294
+    },
+    {
+      "epoch": 0.38286826735885787,
+      "grad_norm": 0.3545709020214379,
+      "learning_rate": 4.848484848484849e-05,
+      "loss": 0.5441,
+      "step": 295
+    },
+    {
+      "epoch": 0.3841661258922777,
+      "grad_norm": 0.27536849505708144,
+      "learning_rate": 4.846079846079846e-05,
+      "loss": 0.5189,
+      "step": 296
+    },
+    {
+      "epoch": 0.3854639844256976,
+      "grad_norm": 0.31314568760395595,
+      "learning_rate": 4.8436748436748444e-05,
+      "loss": 0.5165,
+      "step": 297
+    },
+    {
+      "epoch": 0.3867618429591175,
+      "grad_norm": 0.31727676668467136,
+      "learning_rate": 4.841269841269841e-05,
+      "loss": 0.5185,
+      "step": 298
+    },
+    {
+      "epoch": 0.3880597014925373,
+      "grad_norm": 0.35285183197833564,
+      "learning_rate": 4.838864838864839e-05,
+      "loss": 0.5328,
+      "step": 299
+    },
+    {
+      "epoch": 0.3893575600259572,
+      "grad_norm": 0.2990420731073892,
+      "learning_rate": 4.836459836459837e-05,
+      "loss": 0.532,
+      "step": 300
+    },
+    {
+      "epoch": 0.390655418559377,
+      "grad_norm": 0.38606448483559813,
+      "learning_rate": 4.834054834054834e-05,
+      "loss": 0.5461,
+      "step": 301
+    },
+    {
+      "epoch": 0.39195327709279687,
+      "grad_norm": 0.37652402001442803,
+      "learning_rate": 4.831649831649832e-05,
+      "loss": 0.5379,
+      "step": 302
+    },
+    {
+      "epoch": 0.3932511356262167,
+      "grad_norm": 0.34953978468725405,
+      "learning_rate": 4.829244829244829e-05,
+      "loss": 0.5359,
+      "step": 303
+    },
+    {
+      "epoch": 0.3945489941596366,
+      "grad_norm": 0.3382778166946982,
+      "learning_rate": 4.826839826839827e-05,
+      "loss": 0.5342,
+      "step": 304
+    },
+    {
+      "epoch": 0.3958468526930565,
+      "grad_norm": 0.34560665492104875,
+      "learning_rate": 4.824434824434825e-05,
+      "loss": 0.5324,
+      "step": 305
+    },
+    {
+      "epoch": 0.3971447112264763,
+      "grad_norm": 0.34496470111641636,
+      "learning_rate": 4.8220298220298224e-05,
+      "loss": 0.5339,
+      "step": 306
+    },
+    {
+      "epoch": 0.39844256975989617,
+      "grad_norm": 0.40001685434062584,
+      "learning_rate": 4.81962481962482e-05,
+      "loss": 0.5272,
+      "step": 307
+    },
+    {
+      "epoch": 0.399740428293316,
+      "grad_norm": 0.366032696592655,
+      "learning_rate": 4.8172198172198173e-05,
+      "loss": 0.5306,
+      "step": 308
+    },
+    {
+      "epoch": 0.40103828682673587,
+      "grad_norm": 0.37927598899770393,
+      "learning_rate": 4.814814814814815e-05,
+      "loss": 0.5591,
+      "step": 309
+    },
+    {
+      "epoch": 0.4023361453601557,
+      "grad_norm": 0.32812121834756386,
+      "learning_rate": 4.812409812409813e-05,
+      "loss": 0.5422,
+      "step": 310
+    },
+    {
+      "epoch": 0.4036340038935756,
+      "grad_norm": 0.35171717899329513,
+      "learning_rate": 4.81000481000481e-05,
+      "loss": 0.532,
+      "step": 311
+    },
+    {
+      "epoch": 0.40493186242699547,
+      "grad_norm": 0.3756784968486016,
+      "learning_rate": 4.807599807599808e-05,
+      "loss": 0.5392,
+      "step": 312
+    },
+    {
+      "epoch": 0.4062297209604153,
+      "grad_norm": 0.3426264703785813,
+      "learning_rate": 4.8051948051948054e-05,
+      "loss": 0.5457,
+      "step": 313
+    },
+    {
+      "epoch": 0.40752757949383517,
+      "grad_norm": 0.39836935230937326,
+      "learning_rate": 4.8027898027898036e-05,
+      "loss": 0.5268,
+      "step": 314
+    },
+    {
+      "epoch": 0.408825438027255,
+      "grad_norm": 0.33486717640072616,
+      "learning_rate": 4.8003848003848004e-05,
+      "loss": 0.5298,
+      "step": 315
+    },
+    {
+      "epoch": 0.41012329656067487,
+      "grad_norm": 0.3463640087410465,
+      "learning_rate": 4.797979797979798e-05,
+      "loss": 0.5372,
+      "step": 316
+    },
+    {
+      "epoch": 0.41142115509409477,
+      "grad_norm": 0.2981951724559669,
+      "learning_rate": 4.795574795574796e-05,
+      "loss": 0.5193,
+      "step": 317
+    },
+    {
+      "epoch": 0.4127190136275146,
+      "grad_norm": 0.37701472504733063,
+      "learning_rate": 4.7931697931697935e-05,
+      "loss": 0.5262,
+      "step": 318
+    },
+    {
+      "epoch": 0.41401687216093447,
+      "grad_norm": 0.2958251594721693,
+      "learning_rate": 4.790764790764791e-05,
+      "loss": 0.5262,
+      "step": 319
+    },
+    {
+      "epoch": 0.4153147306943543,
+      "grad_norm": 0.36512530778352836,
+      "learning_rate": 4.7883597883597884e-05,
+      "loss": 0.5451,
+      "step": 320
+    },
+    {
+      "epoch": 0.41661258922777417,
+      "grad_norm": 0.32275488837011096,
+      "learning_rate": 4.785954785954786e-05,
+      "loss": 0.5378,
+      "step": 321
+    },
+    {
+      "epoch": 0.417910447761194,
+      "grad_norm": 0.29968884353456,
+      "learning_rate": 4.783549783549784e-05,
+      "loss": 0.5308,
+      "step": 322
+    },
+    {
+      "epoch": 0.41920830629461386,
+      "grad_norm": 0.3472967912976659,
+      "learning_rate": 4.7811447811447815e-05,
+      "loss": 0.5262,
+      "step": 323
+    },
+    {
+      "epoch": 0.42050616482803377,
+      "grad_norm": 0.3256673670375662,
+      "learning_rate": 4.778739778739779e-05,
+      "loss": 0.5349,
+      "step": 324
+    },
+    {
+      "epoch": 0.4218040233614536,
+      "grad_norm": 0.3830575202323324,
+      "learning_rate": 4.7763347763347765e-05,
+      "loss": 0.5343,
+      "step": 325
+    },
+    {
+      "epoch": 0.42310188189487347,
+      "grad_norm": 0.34375094386741617,
+      "learning_rate": 4.773929773929774e-05,
+      "loss": 0.5295,
+      "step": 326
+    },
+    {
+      "epoch": 0.4243997404282933,
+      "grad_norm": 0.32100699117380493,
+      "learning_rate": 4.771524771524772e-05,
+      "loss": 0.5241,
+      "step": 327
+    },
+    {
+      "epoch": 0.42569759896171316,
+      "grad_norm": 0.3546414912790039,
+      "learning_rate": 4.769119769119769e-05,
+      "loss": 0.5292,
+      "step": 328
+    },
+    {
+      "epoch": 0.426995457495133,
+      "grad_norm": 0.367282717001635,
+      "learning_rate": 4.766714766714767e-05,
+      "loss": 0.525,
+      "step": 329
+    },
+    {
+      "epoch": 0.42829331602855286,
+      "grad_norm": 0.36332040957365974,
+      "learning_rate": 4.7643097643097646e-05,
+      "loss": 0.5395,
+      "step": 330
+    },
+    {
+      "epoch": 0.42959117456197277,
+      "grad_norm": 0.36242424332632034,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 0.5533,
+      "step": 331
+    },
+    {
+      "epoch": 0.4308890330953926,
+      "grad_norm": 0.36697609874383924,
+      "learning_rate": 4.7594997594997595e-05,
+      "loss": 0.5359,
+      "step": 332
+    },
+    {
+      "epoch": 0.43218689162881246,
+      "grad_norm": 0.33118162802731466,
+      "learning_rate": 4.757094757094757e-05,
+      "loss": 0.5317,
+      "step": 333
+    },
+    {
+      "epoch": 0.4334847501622323,
+      "grad_norm": 0.30441401984534905,
+      "learning_rate": 4.754689754689755e-05,
+      "loss": 0.5117,
+      "step": 334
+    },
+    {
+      "epoch": 0.43478260869565216,
+      "grad_norm": 0.38992931838953715,
+      "learning_rate": 4.7522847522847526e-05,
+      "loss": 0.5461,
+      "step": 335
+    },
+    {
+      "epoch": 0.436080467229072,
+      "grad_norm": 0.3335314021890073,
+      "learning_rate": 4.74987974987975e-05,
+      "loss": 0.521,
+      "step": 336
+    },
+    {
+      "epoch": 0.4373783257624919,
+      "grad_norm": 0.44568000339670255,
+      "learning_rate": 4.7474747474747476e-05,
+      "loss": 0.5351,
+      "step": 337
+    },
+    {
+      "epoch": 0.43867618429591176,
+      "grad_norm": 0.30521804239806394,
+      "learning_rate": 4.745069745069745e-05,
+      "loss": 0.5354,
+      "step": 338
+    },
+    {
+      "epoch": 0.4399740428293316,
+      "grad_norm": 0.40068879251857975,
+      "learning_rate": 4.742664742664743e-05,
+      "loss": 0.519,
+      "step": 339
+    },
+    {
+      "epoch": 0.44127190136275146,
+      "grad_norm": 0.2988879048992771,
+      "learning_rate": 4.740259740259741e-05,
+      "loss": 0.5219,
+      "step": 340
+    },
+    {
+      "epoch": 0.4425697598961713,
+      "grad_norm": 0.3568374963385924,
+      "learning_rate": 4.737854737854738e-05,
+      "loss": 0.5407,
+      "step": 341
+    },
+    {
+      "epoch": 0.44386761842959116,
+      "grad_norm": 0.3314619863009568,
+      "learning_rate": 4.7354497354497356e-05,
+      "loss": 0.5293,
+      "step": 342
+    },
+    {
+      "epoch": 0.445165476963011,
+      "grad_norm": 0.38859118585963526,
+      "learning_rate": 4.733044733044733e-05,
+      "loss": 0.5313,
+      "step": 343
+    },
+    {
+      "epoch": 0.4464633354964309,
+      "grad_norm": 0.29870660291558937,
+      "learning_rate": 4.730639730639731e-05,
+      "loss": 0.5235,
+      "step": 344
+    },
+    {
+      "epoch": 0.44776119402985076,
+      "grad_norm": 0.33559539703922564,
+      "learning_rate": 4.728234728234728e-05,
+      "loss": 0.5165,
+      "step": 345
+    },
+    {
+      "epoch": 0.4490590525632706,
+      "grad_norm": 0.29408841700456767,
+      "learning_rate": 4.725829725829726e-05,
+      "loss": 0.5385,
+      "step": 346
+    },
+    {
+      "epoch": 0.45035691109669046,
+      "grad_norm": 0.37054598216697,
+      "learning_rate": 4.723424723424724e-05,
+      "loss": 0.5107,
+      "step": 347
+    },
+    {
+      "epoch": 0.4516547696301103,
+      "grad_norm": 0.3250925011044857,
+      "learning_rate": 4.721019721019721e-05,
+      "loss": 0.5233,
+      "step": 348
+    },
+    {
+      "epoch": 0.45295262816353016,
+      "grad_norm": 0.35856402885603006,
+      "learning_rate": 4.718614718614719e-05,
+      "loss": 0.5055,
+      "step": 349
+    },
+    {
+      "epoch": 0.45425048669695,
+      "grad_norm": 0.34612856309129164,
+      "learning_rate": 4.716209716209716e-05,
+      "loss": 0.5404,
+      "step": 350
+    },
+    {
+      "epoch": 0.4555483452303699,
+      "grad_norm": 0.33303853027334285,
+      "learning_rate": 4.713804713804714e-05,
+      "loss": 0.5178,
+      "step": 351
+    },
+    {
+      "epoch": 0.45684620376378976,
+      "grad_norm": 0.32091795432054987,
+      "learning_rate": 4.711399711399712e-05,
+      "loss": 0.5222,
+      "step": 352
+    },
+    {
+      "epoch": 0.4581440622972096,
+      "grad_norm": 0.34934754787554123,
+      "learning_rate": 4.708994708994709e-05,
+      "loss": 0.5405,
+      "step": 353
+    },
+    {
+      "epoch": 0.45944192083062946,
+      "grad_norm": 0.2937202903692653,
+      "learning_rate": 4.706589706589707e-05,
+      "loss": 0.5283,
+      "step": 354
+    },
+    {
+      "epoch": 0.4607397793640493,
+      "grad_norm": 0.3464664667340698,
+      "learning_rate": 4.704184704184704e-05,
+      "loss": 0.5257,
+      "step": 355
+    },
+    {
+      "epoch": 0.46203763789746916,
+      "grad_norm": 0.3021689351056674,
+      "learning_rate": 4.7017797017797024e-05,
+      "loss": 0.5256,
+      "step": 356
+    },
+    {
+      "epoch": 0.46333549643088906,
+      "grad_norm": 0.3373492256075124,
+      "learning_rate": 4.6993746993747e-05,
+      "loss": 0.5262,
+      "step": 357
+    },
+    {
+      "epoch": 0.4646333549643089,
+      "grad_norm": 0.3279466476251607,
+      "learning_rate": 4.696969696969697e-05,
+      "loss": 0.5252,
+      "step": 358
+    },
+    {
+      "epoch": 0.46593121349772876,
+      "grad_norm": 0.3151794533478745,
+      "learning_rate": 4.694564694564695e-05,
+      "loss": 0.5355,
+      "step": 359
+    },
+    {
+      "epoch": 0.4672290720311486,
+      "grad_norm": 0.3676469350203011,
+      "learning_rate": 4.692159692159692e-05,
+      "loss": 0.5536,
+      "step": 360
+    },
+    {
+      "epoch": 0.46852693056456846,
+      "grad_norm": 0.2638129347242171,
+      "learning_rate": 4.6897546897546904e-05,
+      "loss": 0.5167,
+      "step": 361
+    },
+    {
+      "epoch": 0.4698247890979883,
+      "grad_norm": 0.32036105761534295,
+      "learning_rate": 4.687349687349687e-05,
+      "loss": 0.5105,
+      "step": 362
+    },
+    {
+      "epoch": 0.47112264763140815,
+      "grad_norm": 0.3312350329187521,
+      "learning_rate": 4.6849446849446854e-05,
+      "loss": 0.5403,
+      "step": 363
+    },
+    {
+      "epoch": 0.47242050616482806,
+      "grad_norm": 0.2691270807481844,
+      "learning_rate": 4.682539682539683e-05,
+      "loss": 0.5129,
+      "step": 364
+    },
+    {
+      "epoch": 0.4737183646982479,
+      "grad_norm": 0.31512841418214993,
+      "learning_rate": 4.68013468013468e-05,
+      "loss": 0.5188,
+      "step": 365
+    },
+    {
+      "epoch": 0.47501622323166776,
+      "grad_norm": 0.32669598998030175,
+      "learning_rate": 4.677729677729678e-05,
+      "loss": 0.5384,
+      "step": 366
+    },
+    {
+      "epoch": 0.4763140817650876,
+      "grad_norm": 0.3235775681673074,
+      "learning_rate": 4.675324675324675e-05,
+      "loss": 0.5194,
+      "step": 367
+    },
+    {
+      "epoch": 0.47761194029850745,
+      "grad_norm": 0.3095310587265592,
+      "learning_rate": 4.6729196729196734e-05,
+      "loss": 0.5323,
+      "step": 368
+    },
+    {
+      "epoch": 0.4789097988319273,
+      "grad_norm": 0.32341229991033627,
+      "learning_rate": 4.670514670514671e-05,
+      "loss": 0.5192,
+      "step": 369
+    },
+    {
+      "epoch": 0.48020765736534715,
+      "grad_norm": 0.3175500385279334,
+      "learning_rate": 4.6681096681096684e-05,
+      "loss": 0.53,
+      "step": 370
+    },
+    {
+      "epoch": 0.48150551589876706,
+      "grad_norm": 0.34206247662543693,
+      "learning_rate": 4.665704665704666e-05,
+      "loss": 0.5314,
+      "step": 371
+    },
+    {
+      "epoch": 0.4828033744321869,
+      "grad_norm": 0.3185235414836324,
+      "learning_rate": 4.6632996632996634e-05,
+      "loss": 0.5367,
+      "step": 372
+    },
+    {
+      "epoch": 0.48410123296560675,
+      "grad_norm": 0.28453878466972515,
+      "learning_rate": 4.6608946608946615e-05,
+      "loss": 0.5223,
+      "step": 373
+    },
+    {
+      "epoch": 0.4853990914990266,
+      "grad_norm": 0.2957039810409513,
+      "learning_rate": 4.658489658489659e-05,
+      "loss": 0.5005,
+      "step": 374
+    },
+    {
+      "epoch": 0.48669695003244645,
+      "grad_norm": 0.2940626178906619,
+      "learning_rate": 4.656084656084656e-05,
+      "loss": 0.5098,
+      "step": 375
+    },
+    {
+      "epoch": 0.4879948085658663,
+      "grad_norm": 0.33976965417394467,
+      "learning_rate": 4.653679653679654e-05,
+      "loss": 0.5155,
+      "step": 376
+    },
+    {
+      "epoch": 0.4892926670992862,
+      "grad_norm": 0.2993214412594064,
+      "learning_rate": 4.6512746512746514e-05,
+      "loss": 0.5183,
+      "step": 377
+    },
+    {
+      "epoch": 0.49059052563270605,
+      "grad_norm": 0.3550998192684892,
+      "learning_rate": 4.6488696488696496e-05,
+      "loss": 0.5404,
+      "step": 378
+    },
+    {
+      "epoch": 0.4918883841661259,
+      "grad_norm": 0.3961098073492471,
+      "learning_rate": 4.6464646464646464e-05,
+      "loss": 0.5411,
+      "step": 379
+    },
+    {
+      "epoch": 0.49318624269954575,
+      "grad_norm": 0.34269318810223304,
+      "learning_rate": 4.6440596440596445e-05,
+      "loss": 0.5206,
+      "step": 380
+    },
+    {
+      "epoch": 0.4944841012329656,
+      "grad_norm": 0.29509416892424534,
+      "learning_rate": 4.641654641654642e-05,
+      "loss": 0.5189,
+      "step": 381
+    },
+    {
+      "epoch": 0.49578195976638545,
+      "grad_norm": 0.32772442148274133,
+      "learning_rate": 4.6392496392496395e-05,
+      "loss": 0.5368,
+      "step": 382
+    },
+    {
+      "epoch": 0.4970798182998053,
+      "grad_norm": 0.2719733229054414,
+      "learning_rate": 4.636844636844637e-05,
+      "loss": 0.5027,
+      "step": 383
+    },
+    {
+      "epoch": 0.4983776768332252,
+      "grad_norm": 0.32827976870034653,
+      "learning_rate": 4.6344396344396344e-05,
+      "loss": 0.5225,
+      "step": 384
+    },
+    {
+      "epoch": 0.49967553536664505,
+      "grad_norm": 0.36529779846696075,
+      "learning_rate": 4.6320346320346326e-05,
+      "loss": 0.4961,
+      "step": 385
+    },
+    {
+      "epoch": 0.5009733939000649,
+      "grad_norm": 0.34737533192311987,
+      "learning_rate": 4.62962962962963e-05,
+      "loss": 0.5056,
+      "step": 386
+    },
+    {
+      "epoch": 0.5022712524334848,
+      "grad_norm": 0.32570267249669654,
+      "learning_rate": 4.6272246272246276e-05,
+      "loss": 0.5114,
+      "step": 387
+    },
+    {
+      "epoch": 0.5035691109669046,
+      "grad_norm": 0.3419484703073112,
+      "learning_rate": 4.624819624819625e-05,
+      "loss": 0.5141,
+      "step": 388
+    },
+    {
+      "epoch": 0.5048669695003245,
+      "grad_norm": 0.34141193067026915,
+      "learning_rate": 4.6224146224146225e-05,
+      "loss": 0.5296,
+      "step": 389
+    },
+    {
+      "epoch": 0.5061648280337443,
+      "grad_norm": 0.314416021500765,
+      "learning_rate": 4.620009620009621e-05,
+      "loss": 0.5202,
+      "step": 390
+    },
+    {
+      "epoch": 0.5074626865671642,
+      "grad_norm": 0.2608120507481533,
+      "learning_rate": 4.617604617604618e-05,
+      "loss": 0.507,
+      "step": 391
+    },
+    {
+      "epoch": 0.508760545100584,
+      "grad_norm": 0.3415961124944403,
+      "learning_rate": 4.615199615199615e-05,
+      "loss": 0.5513,
+      "step": 392
+    },
+    {
+      "epoch": 0.5100584036340039,
+      "grad_norm": 0.3545772824977214,
+      "learning_rate": 4.612794612794613e-05,
+      "loss": 0.5195,
+      "step": 393
+    },
+    {
+      "epoch": 0.5113562621674238,
+      "grad_norm": 0.2759048918555112,
+      "learning_rate": 4.6103896103896106e-05,
+      "loss": 0.5145,
+      "step": 394
+    },
+    {
+      "epoch": 0.5126541207008436,
+      "grad_norm": 0.29424902451922874,
+      "learning_rate": 4.607984607984609e-05,
+      "loss": 0.4965,
+      "step": 395
+    },
+    {
+      "epoch": 0.5139519792342635,
+      "grad_norm": 0.2865983981532377,
+      "learning_rate": 4.6055796055796055e-05,
+      "loss": 0.51,
+      "step": 396
+    },
+    {
+      "epoch": 0.5152498377676833,
+      "grad_norm": 0.2826395400080094,
+      "learning_rate": 4.603174603174603e-05,
+      "loss": 0.5177,
+      "step": 397
+    },
+    {
+      "epoch": 0.5165476963011032,
+      "grad_norm": 0.3154832346968727,
+      "learning_rate": 4.600769600769601e-05,
+      "loss": 0.5167,
+      "step": 398
+    },
+    {
+      "epoch": 0.517845554834523,
+      "grad_norm": 0.28657500627349963,
+      "learning_rate": 4.5983645983645986e-05,
+      "loss": 0.4915,
+      "step": 399
+    },
+    {
+      "epoch": 0.5191434133679429,
+      "grad_norm": 0.3108878669769025,
+      "learning_rate": 4.595959595959596e-05,
+      "loss": 0.5094,
+      "step": 400
+    },
+    {
+      "epoch": 0.5204412719013628,
+      "grad_norm": 0.3105653890499028,
+      "learning_rate": 4.5935545935545936e-05,
+      "loss": 0.5395,
+      "step": 401
+    },
+    {
+      "epoch": 0.5217391304347826,
+      "grad_norm": 0.3041614944176647,
+      "learning_rate": 4.591149591149592e-05,
+      "loss": 0.5015,
+      "step": 402
+    },
+    {
+      "epoch": 0.5230369889682025,
+      "grad_norm": 0.2774234169618742,
+      "learning_rate": 4.588744588744589e-05,
+      "loss": 0.4953,
+      "step": 403
+    },
+    {
+      "epoch": 0.5243348475016223,
+      "grad_norm": 0.3292325244305617,
+      "learning_rate": 4.586339586339587e-05,
+      "loss": 0.5326,
+      "step": 404
+    },
+    {
+      "epoch": 0.5256327060350422,
+      "grad_norm": 0.280230976708198,
+      "learning_rate": 4.583934583934584e-05,
+      "loss": 0.5186,
+      "step": 405
+    },
+    {
+      "epoch": 0.526930564568462,
+      "grad_norm": 0.29220060360384115,
+      "learning_rate": 4.5815295815295817e-05,
+      "loss": 0.5224,
+      "step": 406
+    },
+    {
+      "epoch": 0.5282284231018819,
+      "grad_norm": 0.2973987484773138,
+      "learning_rate": 4.57912457912458e-05,
+      "loss": 0.5198,
+      "step": 407
+    },
+    {
+      "epoch": 0.5295262816353018,
+      "grad_norm": 0.31013343434720114,
+      "learning_rate": 4.576719576719577e-05,
+      "loss": 0.5303,
+      "step": 408
+    },
+    {
+      "epoch": 0.5308241401687216,
+      "grad_norm": 0.3091402470001352,
+      "learning_rate": 4.574314574314574e-05,
+      "loss": 0.5176,
+      "step": 409
+    },
+    {
+      "epoch": 0.5321219987021415,
+      "grad_norm": 0.2974831281530903,
+      "learning_rate": 4.571909571909572e-05,
+      "loss": 0.5153,
+      "step": 410
+    },
+    {
+      "epoch": 0.5334198572355613,
+      "grad_norm": 0.30558238497100093,
+      "learning_rate": 4.56950456950457e-05,
+      "loss": 0.5171,
+      "step": 411
+    },
+    {
+      "epoch": 0.5347177157689812,
+      "grad_norm": 0.3050417584271822,
+      "learning_rate": 4.567099567099568e-05,
+      "loss": 0.51,
+      "step": 412
+    },
+    {
+      "epoch": 0.536015574302401,
+      "grad_norm": 0.3062163771011129,
+      "learning_rate": 4.564694564694565e-05,
+      "loss": 0.5182,
+      "step": 413
+    },
+    {
+      "epoch": 0.5373134328358209,
+      "grad_norm": 0.2918708356467548,
+      "learning_rate": 4.562289562289562e-05,
+      "loss": 0.5372,
+      "step": 414
+    },
+    {
+      "epoch": 0.5386112913692408,
+      "grad_norm": 0.3100802477130393,
+      "learning_rate": 4.55988455988456e-05,
+      "loss": 0.504,
+      "step": 415
+    },
+    {
+      "epoch": 0.5399091499026606,
+      "grad_norm": 0.28080959444703624,
+      "learning_rate": 4.557479557479558e-05,
+      "loss": 0.5133,
+      "step": 416
+    },
+    {
+      "epoch": 0.5412070084360805,
+      "grad_norm": 0.31510663571529784,
+      "learning_rate": 4.555074555074555e-05,
+      "loss": 0.5083,
+      "step": 417
+    },
+    {
+      "epoch": 0.5425048669695003,
+      "grad_norm": 0.2987074534694197,
+      "learning_rate": 4.552669552669553e-05,
+      "loss": 0.5233,
+      "step": 418
+    },
+    {
+      "epoch": 0.5438027255029202,
+      "grad_norm": 0.3018100646658762,
+      "learning_rate": 4.55026455026455e-05,
+      "loss": 0.5125,
+      "step": 419
+    },
+    {
+      "epoch": 0.54510058403634,
+      "grad_norm": 0.3068942632441547,
+      "learning_rate": 4.5478595478595484e-05,
+      "loss": 0.5141,
+      "step": 420
+    },
+    {
+      "epoch": 0.5463984425697599,
+      "grad_norm": 0.30401355001057445,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 0.5275,
+      "step": 421
+    },
+    {
+      "epoch": 0.5476963011031798,
+      "grad_norm": 0.3203825407640607,
+      "learning_rate": 4.543049543049543e-05,
+      "loss": 0.5374,
+      "step": 422
+    },
+    {
+      "epoch": 0.5489941596365996,
+      "grad_norm": 0.2890332355349151,
+      "learning_rate": 4.540644540644541e-05,
+      "loss": 0.4928,
+      "step": 423
+    },
+    {
+      "epoch": 0.5502920181700195,
+      "grad_norm": 0.3167375640999411,
+      "learning_rate": 4.538239538239538e-05,
+      "loss": 0.5238,
+      "step": 424
+    },
+    {
+      "epoch": 0.5515898767034393,
+      "grad_norm": 0.2838739525584143,
+      "learning_rate": 4.535834535834536e-05,
+      "loss": 0.5193,
+      "step": 425
+    },
+    {
+      "epoch": 0.5528877352368592,
+      "grad_norm": 0.31600565211015347,
+      "learning_rate": 4.533429533429533e-05,
+      "loss": 0.5408,
+      "step": 426
+    },
+    {
+      "epoch": 0.5541855937702791,
+      "grad_norm": 0.2936881048070723,
+      "learning_rate": 4.5310245310245314e-05,
+      "loss": 0.517,
+      "step": 427
+    },
+    {
+      "epoch": 0.5554834523036989,
+      "grad_norm": 0.2869325695453175,
+      "learning_rate": 4.528619528619529e-05,
+      "loss": 0.5272,
+      "step": 428
+    },
+    {
+      "epoch": 0.5567813108371188,
+      "grad_norm": 0.26396682696755225,
+      "learning_rate": 4.5262145262145264e-05,
+      "loss": 0.5092,
+      "step": 429
+    },
+    {
+      "epoch": 0.5580791693705386,
+      "grad_norm": 0.2778797627866081,
+      "learning_rate": 4.523809523809524e-05,
+      "loss": 0.5107,
+      "step": 430
+    },
+    {
+      "epoch": 0.5593770279039585,
+      "grad_norm": 0.29638166733777366,
+      "learning_rate": 4.521404521404521e-05,
+      "loss": 0.5207,
+      "step": 431
+    },
+    {
+      "epoch": 0.5606748864373783,
+      "grad_norm": 0.31086056140406293,
+      "learning_rate": 4.5189995189995195e-05,
+      "loss": 0.5213,
+      "step": 432
+    },
+    {
+      "epoch": 0.5619727449707982,
+      "grad_norm": 0.2777342628277087,
+      "learning_rate": 4.516594516594517e-05,
+      "loss": 0.5225,
+      "step": 433
+    },
+    {
+      "epoch": 0.5632706035042181,
+      "grad_norm": 0.31153041204212967,
+      "learning_rate": 4.5141895141895144e-05,
+      "loss": 0.5029,
+      "step": 434
+    },
+    {
+      "epoch": 0.5645684620376379,
+      "grad_norm": 0.2833387330391144,
+      "learning_rate": 4.511784511784512e-05,
+      "loss": 0.5269,
+      "step": 435
+    },
+    {
+      "epoch": 0.5658663205710578,
+      "grad_norm": 0.3208653719812624,
+      "learning_rate": 4.5093795093795094e-05,
+      "loss": 0.5027,
+      "step": 436
+    },
+    {
+      "epoch": 0.5671641791044776,
+      "grad_norm": 0.3024059727560176,
+      "learning_rate": 4.5069745069745075e-05,
+      "loss": 0.5219,
+      "step": 437
+    },
+    {
+      "epoch": 0.5684620376378975,
+      "grad_norm": 0.2822549725146957,
+      "learning_rate": 4.504569504569504e-05,
+      "loss": 0.5095,
+      "step": 438
+    },
+    {
+      "epoch": 0.5697598961713173,
+      "grad_norm": 0.3016137705597104,
+      "learning_rate": 4.5021645021645025e-05,
+      "loss": 0.5104,
+      "step": 439
+    },
+    {
+      "epoch": 0.5710577547047372,
+      "grad_norm": 0.3627945641240943,
+      "learning_rate": 4.4997594997595e-05,
+      "loss": 0.5264,
+      "step": 440
+    },
+    {
+      "epoch": 0.5723556132381571,
+      "grad_norm": 0.27507558856590775,
+      "learning_rate": 4.4973544973544974e-05,
+      "loss": 0.5139,
+      "step": 441
+    },
+    {
+      "epoch": 0.5736534717715769,
+      "grad_norm": 0.3593784232319199,
+      "learning_rate": 4.494949494949495e-05,
+      "loss": 0.5212,
+      "step": 442
+    },
+    {
+      "epoch": 0.5749513303049968,
+      "grad_norm": 0.29557452932516204,
+      "learning_rate": 4.4925444925444924e-05,
+      "loss": 0.502,
+      "step": 443
+    },
+    {
+      "epoch": 0.5762491888384166,
+      "grad_norm": 0.3005471775259294,
+      "learning_rate": 4.4901394901394906e-05,
+      "loss": 0.4912,
+      "step": 444
+    },
+    {
+      "epoch": 0.5775470473718365,
+      "grad_norm": 0.26919062978615377,
+      "learning_rate": 4.487734487734488e-05,
+      "loss": 0.5313,
+      "step": 445
+    },
+    {
+      "epoch": 0.5788449059052563,
+      "grad_norm": 0.2556241030912058,
+      "learning_rate": 4.4853294853294855e-05,
+      "loss": 0.5085,
+      "step": 446
+    },
+    {
+      "epoch": 0.5801427644386762,
+      "grad_norm": 0.2733433348389189,
+      "learning_rate": 4.482924482924483e-05,
+      "loss": 0.5232,
+      "step": 447
+    },
+    {
+      "epoch": 0.5814406229720961,
+      "grad_norm": 0.2699232360629045,
+      "learning_rate": 4.4805194805194805e-05,
+      "loss": 0.5049,
+      "step": 448
+    },
+    {
+      "epoch": 0.5827384815055159,
+      "grad_norm": 0.28747431650418886,
+      "learning_rate": 4.4781144781144786e-05,
+      "loss": 0.5112,
+      "step": 449
+    },
+    {
+      "epoch": 0.5840363400389358,
+      "grad_norm": 0.28804716410878617,
+      "learning_rate": 4.475709475709476e-05,
+      "loss": 0.4905,
+      "step": 450
+    },
+    {
+      "epoch": 0.5853341985723556,
+      "grad_norm": 0.2919584848714507,
+      "learning_rate": 4.4733044733044736e-05,
+      "loss": 0.5272,
+      "step": 451
+    },
+    {
+      "epoch": 0.5866320571057755,
+      "grad_norm": 0.27840735576549713,
+      "learning_rate": 4.470899470899471e-05,
+      "loss": 0.4885,
+      "step": 452
+    },
+    {
+      "epoch": 0.5879299156391953,
+      "grad_norm": 0.29490953800516345,
+      "learning_rate": 4.4684944684944685e-05,
+      "loss": 0.5164,
+      "step": 453
+    },
+    {
+      "epoch": 0.5892277741726152,
+      "grad_norm": 0.3240170627979527,
+      "learning_rate": 4.466089466089467e-05,
+      "loss": 0.5173,
+      "step": 454
+    },
+    {
+      "epoch": 0.5905256327060351,
+      "grad_norm": 0.2665880580848304,
+      "learning_rate": 4.4636844636844635e-05,
+      "loss": 0.5045,
+      "step": 455
+    },
+    {
+      "epoch": 0.5918234912394549,
+      "grad_norm": 0.36453608305554464,
+      "learning_rate": 4.4612794612794616e-05,
+      "loss": 0.5176,
+      "step": 456
+    },
+    {
+      "epoch": 0.5931213497728748,
+      "grad_norm": 0.2971475504780928,
+      "learning_rate": 4.458874458874459e-05,
+      "loss": 0.5047,
+      "step": 457
+    },
+    {
+      "epoch": 0.5944192083062946,
+      "grad_norm": 0.34100322893736984,
+      "learning_rate": 4.4564694564694566e-05,
+      "loss": 0.5111,
+      "step": 458
+    },
+    {
+      "epoch": 0.5957170668397145,
+      "grad_norm": 0.27240121549823265,
+      "learning_rate": 4.454064454064454e-05,
+      "loss": 0.4885,
+      "step": 459
+    },
+    {
+      "epoch": 0.5970149253731343,
+      "grad_norm": 0.31589767714076145,
+      "learning_rate": 4.4516594516594515e-05,
+      "loss": 0.4994,
+      "step": 460
+    },
+    {
+      "epoch": 0.5983127839065542,
+      "grad_norm": 0.2801464661937106,
+      "learning_rate": 4.44925444925445e-05,
+      "loss": 0.497,
+      "step": 461
+    },
+    {
+      "epoch": 0.5996106424399741,
+      "grad_norm": 0.33064691940201346,
+      "learning_rate": 4.446849446849447e-05,
+      "loss": 0.5336,
+      "step": 462
+    },
+    {
+      "epoch": 0.6009085009733939,
+      "grad_norm": 0.2866479133025442,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.5346,
+      "step": 463
+    },
+    {
+      "epoch": 0.6022063595068138,
+      "grad_norm": 0.33165659226246563,
+      "learning_rate": 4.442039442039442e-05,
+      "loss": 0.5145,
+      "step": 464
+    },
+    {
+      "epoch": 0.6035042180402336,
+      "grad_norm": 0.4080281603274731,
+      "learning_rate": 4.4396344396344396e-05,
+      "loss": 0.5083,
+      "step": 465
+    },
+    {
+      "epoch": 0.6048020765736535,
+      "grad_norm": 0.32997017809734314,
+      "learning_rate": 4.437229437229438e-05,
+      "loss": 0.5108,
+      "step": 466
+    },
+    {
+      "epoch": 0.6060999351070734,
+      "grad_norm": 0.4193305220535223,
+      "learning_rate": 4.434824434824435e-05,
+      "loss": 0.5333,
+      "step": 467
+    },
+    {
+      "epoch": 0.6073977936404932,
+      "grad_norm": 0.33536790460482346,
+      "learning_rate": 4.432419432419432e-05,
+      "loss": 0.5315,
+      "step": 468
+    },
+    {
+      "epoch": 0.6086956521739131,
+      "grad_norm": 0.30331774787903265,
+      "learning_rate": 4.43001443001443e-05,
+      "loss": 0.5044,
+      "step": 469
+    },
+    {
+      "epoch": 0.6099935107073329,
+      "grad_norm": 0.3332494229182582,
+      "learning_rate": 4.427609427609428e-05,
+      "loss": 0.5271,
+      "step": 470
+    },
+    {
+      "epoch": 0.6112913692407528,
+      "grad_norm": 0.27631817841293893,
+      "learning_rate": 4.425204425204426e-05,
+      "loss": 0.4869,
+      "step": 471
+    },
+    {
+      "epoch": 0.6125892277741726,
+      "grad_norm": 0.30864904652906816,
+      "learning_rate": 4.4227994227994226e-05,
+      "loss": 0.5189,
+      "step": 472
+    },
+    {
+      "epoch": 0.6138870863075925,
+      "grad_norm": 0.25286124571299967,
+      "learning_rate": 4.420394420394421e-05,
+      "loss": 0.5243,
+      "step": 473
+    },
+    {
+      "epoch": 0.6151849448410124,
+      "grad_norm": 0.32194768257527906,
+      "learning_rate": 4.417989417989418e-05,
+      "loss": 0.517,
+      "step": 474
+    },
+    {
+      "epoch": 0.6164828033744322,
+      "grad_norm": 0.25295912491765415,
+      "learning_rate": 4.415584415584416e-05,
+      "loss": 0.5119,
+      "step": 475
+    },
+    {
+      "epoch": 0.6177806619078521,
+      "grad_norm": 0.30031239529360704,
+      "learning_rate": 4.413179413179413e-05,
+      "loss": 0.5167,
+      "step": 476
+    },
+    {
+      "epoch": 0.6190785204412719,
+      "grad_norm": 0.3457458465491688,
+      "learning_rate": 4.410774410774411e-05,
+      "loss": 0.5122,
+      "step": 477
+    },
+    {
+      "epoch": 0.6203763789746918,
+      "grad_norm": 0.3091494265523315,
+      "learning_rate": 4.408369408369409e-05,
+      "loss": 0.4996,
+      "step": 478
+    },
+    {
+      "epoch": 0.6216742375081116,
+      "grad_norm": 0.34841455852307157,
+      "learning_rate": 4.405964405964406e-05,
+      "loss": 0.5187,
+      "step": 479
+    },
+    {
+      "epoch": 0.6229720960415315,
+      "grad_norm": 0.28466491288804874,
+      "learning_rate": 4.403559403559404e-05,
+      "loss": 0.5081,
+      "step": 480
+    },
+    {
+      "epoch": 0.6242699545749514,
+      "grad_norm": 0.31239695738713696,
+      "learning_rate": 4.401154401154401e-05,
+      "loss": 0.5188,
+      "step": 481
+    },
+    {
+      "epoch": 0.6255678131083712,
+      "grad_norm": 0.2906342979686575,
+      "learning_rate": 4.398749398749399e-05,
+      "loss": 0.4979,
+      "step": 482
+    },
+    {
+      "epoch": 0.6268656716417911,
+      "grad_norm": 0.32742410431546853,
+      "learning_rate": 4.396344396344397e-05,
+      "loss": 0.5096,
+      "step": 483
+    },
+    {
+      "epoch": 0.6281635301752109,
+      "grad_norm": 0.28587236180759673,
+      "learning_rate": 4.3939393939393944e-05,
+      "loss": 0.5282,
+      "step": 484
+    },
+    {
+      "epoch": 0.6294613887086308,
+      "grad_norm": 0.2939115603096443,
+      "learning_rate": 4.391534391534391e-05,
+      "loss": 0.5172,
+      "step": 485
+    },
+    {
+      "epoch": 0.6307592472420506,
+      "grad_norm": 0.2873210431560928,
+      "learning_rate": 4.3891293891293894e-05,
+      "loss": 0.5229,
+      "step": 486
+    },
+    {
+      "epoch": 0.6320571057754705,
+      "grad_norm": 0.342818284796367,
+      "learning_rate": 4.386724386724387e-05,
+      "loss": 0.5272,
+      "step": 487
+    },
+    {
+      "epoch": 0.6333549643088904,
+      "grad_norm": 0.2663532951872564,
+      "learning_rate": 4.384319384319385e-05,
+      "loss": 0.5087,
+      "step": 488
+    },
+    {
+      "epoch": 0.6346528228423102,
+      "grad_norm": 0.25849575306366,
+      "learning_rate": 4.381914381914382e-05,
+      "loss": 0.503,
+      "step": 489
+    },
+    {
+      "epoch": 0.6359506813757301,
+      "grad_norm": 0.2682428237326465,
+      "learning_rate": 4.379509379509379e-05,
+      "loss": 0.5178,
+      "step": 490
+    },
+    {
+      "epoch": 0.6372485399091499,
+      "grad_norm": 0.2899634415594277,
+      "learning_rate": 4.3771043771043774e-05,
+      "loss": 0.4964,
+      "step": 491
+    },
+    {
+      "epoch": 0.6385463984425698,
+      "grad_norm": 0.3453086828842896,
+      "learning_rate": 4.374699374699375e-05,
+      "loss": 0.53,
+      "step": 492
+    },
+    {
+      "epoch": 0.6398442569759896,
+      "grad_norm": 0.34399408107909996,
+      "learning_rate": 4.3722943722943724e-05,
+      "loss": 0.5193,
+      "step": 493
+    },
+    {
+      "epoch": 0.6411421155094095,
+      "grad_norm": 0.3610030879163129,
+      "learning_rate": 4.36988936988937e-05,
+      "loss": 0.5144,
+      "step": 494
+    },
+    {
+      "epoch": 0.6424399740428294,
+      "grad_norm": 0.23843044383570788,
+      "learning_rate": 4.367484367484368e-05,
+      "loss": 0.4999,
+      "step": 495
+    },
+    {
+      "epoch": 0.6437378325762492,
+      "grad_norm": 0.3654439591676623,
+      "learning_rate": 4.3650793650793655e-05,
+      "loss": 0.4888,
+      "step": 496
+    },
+    {
+      "epoch": 0.6450356911096691,
+      "grad_norm": 0.28776010247656836,
+      "learning_rate": 4.362674362674363e-05,
+      "loss": 0.5022,
+      "step": 497
+    },
+    {
+      "epoch": 0.6463335496430889,
+      "grad_norm": 0.3647131705869751,
+      "learning_rate": 4.3602693602693604e-05,
+      "loss": 0.5092,
+      "step": 498
+    },
+    {
+      "epoch": 0.6476314081765088,
+      "grad_norm": 0.30736812776643446,
+      "learning_rate": 4.357864357864358e-05,
+      "loss": 0.5171,
+      "step": 499
+    },
+    {
+      "epoch": 0.6489292667099286,
+      "grad_norm": 0.36290147629484104,
+      "learning_rate": 4.355459355459356e-05,
+      "loss": 0.5223,
+      "step": 500
+    },
+    {
+      "epoch": 0.6502271252433485,
+      "grad_norm": 0.32228223382695725,
+      "learning_rate": 4.3530543530543535e-05,
+      "loss": 0.5099,
+      "step": 501
+    },
+    {
+      "epoch": 0.6515249837767684,
+      "grad_norm": 0.31393689007483594,
+      "learning_rate": 4.3506493506493503e-05,
+      "loss": 0.496,
+      "step": 502
+    },
+    {
+      "epoch": 0.6528228423101882,
+      "grad_norm": 0.2966759326603879,
+      "learning_rate": 4.3482443482443485e-05,
+      "loss": 0.5173,
+      "step": 503
+    },
+    {
+      "epoch": 0.6541207008436081,
+      "grad_norm": 0.2864744517114858,
+      "learning_rate": 4.345839345839346e-05,
+      "loss": 0.503,
+      "step": 504
+    },
+    {
+      "epoch": 0.6554185593770279,
+      "grad_norm": 0.28016826596559247,
+      "learning_rate": 4.343434343434344e-05,
+      "loss": 0.5095,
+      "step": 505
+    },
+    {
+      "epoch": 0.6567164179104478,
+      "grad_norm": 0.3045274480234983,
+      "learning_rate": 4.341029341029341e-05,
+      "loss": 0.5234,
+      "step": 506
+    },
+    {
+      "epoch": 0.6580142764438677,
+      "grad_norm": 0.2865539821179636,
+      "learning_rate": 4.3386243386243384e-05,
+      "loss": 0.5057,
+      "step": 507
+    },
+    {
+      "epoch": 0.6593121349772875,
+      "grad_norm": 0.28016725527352626,
+      "learning_rate": 4.3362193362193366e-05,
+      "loss": 0.5054,
+      "step": 508
+    },
+    {
+      "epoch": 0.6606099935107074,
+      "grad_norm": 0.2779087438851858,
+      "learning_rate": 4.333814333814334e-05,
+      "loss": 0.4844,
+      "step": 509
+    },
+    {
+      "epoch": 0.6619078520441272,
+      "grad_norm": 0.29308593442315034,
+      "learning_rate": 4.3314093314093315e-05,
+      "loss": 0.5026,
+      "step": 510
+    },
+    {
+      "epoch": 0.6632057105775471,
+      "grad_norm": 0.24617150101353785,
+      "learning_rate": 4.329004329004329e-05,
+      "loss": 0.505,
+      "step": 511
+    },
+    {
+      "epoch": 0.6645035691109669,
+      "grad_norm": 0.2801536462432465,
+      "learning_rate": 4.3265993265993265e-05,
+      "loss": 0.4957,
+      "step": 512
+    },
+    {
+      "epoch": 0.6658014276443868,
+      "grad_norm": 0.2590262249669081,
+      "learning_rate": 4.3241943241943246e-05,
+      "loss": 0.4976,
+      "step": 513
+    },
+    {
+      "epoch": 0.6670992861778067,
+      "grad_norm": 0.27675213215164485,
+      "learning_rate": 4.321789321789322e-05,
+      "loss": 0.5016,
+      "step": 514
+    },
+    {
+      "epoch": 0.6683971447112265,
+      "grad_norm": 0.3211262394859621,
+      "learning_rate": 4.3193843193843196e-05,
+      "loss": 0.5285,
+      "step": 515
+    },
+    {
+      "epoch": 0.6696950032446464,
+      "grad_norm": 0.2847594895492132,
+      "learning_rate": 4.316979316979317e-05,
+      "loss": 0.5174,
+      "step": 516
+    },
+    {
+      "epoch": 0.6709928617780662,
+      "grad_norm": 0.31731208678548406,
+      "learning_rate": 4.314574314574315e-05,
+      "loss": 0.5287,
+      "step": 517
+    },
+    {
+      "epoch": 0.6722907203114861,
+      "grad_norm": 0.26600293134050695,
+      "learning_rate": 4.312169312169313e-05,
+      "loss": 0.5105,
+      "step": 518
+    },
+    {
+      "epoch": 0.6735885788449059,
+      "grad_norm": 0.29880462234281113,
+      "learning_rate": 4.3097643097643095e-05,
+      "loss": 0.5375,
+      "step": 519
+    },
+    {
+      "epoch": 0.6748864373783258,
+      "grad_norm": 0.2652094878668775,
+      "learning_rate": 4.3073593073593077e-05,
+      "loss": 0.5033,
+      "step": 520
+    },
+    {
+      "epoch": 0.6761842959117457,
+      "grad_norm": 0.315140738606816,
+      "learning_rate": 4.304954304954305e-05,
+      "loss": 0.5238,
+      "step": 521
+    },
+    {
+      "epoch": 0.6774821544451655,
+      "grad_norm": 0.2852888179467452,
+      "learning_rate": 4.302549302549303e-05,
+      "loss": 0.5125,
+      "step": 522
+    },
+    {
+      "epoch": 0.6787800129785854,
+      "grad_norm": 0.3217782609108167,
+      "learning_rate": 4.3001443001443e-05,
+      "loss": 0.5084,
+      "step": 523
+    },
+    {
+      "epoch": 0.6800778715120052,
+      "grad_norm": 0.3067930968649758,
+      "learning_rate": 4.2977392977392976e-05,
+      "loss": 0.4999,
+      "step": 524
+    },
+    {
+      "epoch": 0.6813757300454251,
+      "grad_norm": 0.2937819263154037,
+      "learning_rate": 4.295334295334296e-05,
+      "loss": 0.5256,
+      "step": 525
+    },
+    {
+      "epoch": 0.6826735885788449,
+      "grad_norm": 0.32438054578281567,
+      "learning_rate": 4.292929292929293e-05,
+      "loss": 0.4907,
+      "step": 526
+    },
+    {
+      "epoch": 0.6839714471122648,
+      "grad_norm": 0.2742147889295781,
+      "learning_rate": 4.290524290524291e-05,
+      "loss": 0.5068,
+      "step": 527
+    },
+    {
+      "epoch": 0.6852693056456847,
+      "grad_norm": 0.35488588986537717,
+      "learning_rate": 4.288119288119288e-05,
+      "loss": 0.5248,
+      "step": 528
+    },
+    {
+      "epoch": 0.6865671641791045,
+      "grad_norm": 0.26229530604678386,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.5037,
+      "step": 529
+    },
+    {
+      "epoch": 0.6878650227125244,
+      "grad_norm": 0.3461696681941986,
+      "learning_rate": 4.283309283309284e-05,
+      "loss": 0.5025,
+      "step": 530
+    },
+    {
+      "epoch": 0.6891628812459442,
+      "grad_norm": 0.266178675206237,
+      "learning_rate": 4.280904280904281e-05,
+      "loss": 0.4896,
+      "step": 531
+    },
+    {
+      "epoch": 0.6904607397793641,
+      "grad_norm": 0.34686998824653287,
+      "learning_rate": 4.278499278499279e-05,
+      "loss": 0.5034,
+      "step": 532
+    },
+    {
+      "epoch": 0.6917585983127839,
+      "grad_norm": 0.3320503579783302,
+      "learning_rate": 4.276094276094276e-05,
+      "loss": 0.5178,
+      "step": 533
+    },
+    {
+      "epoch": 0.6930564568462038,
+      "grad_norm": 0.3083799644603529,
+      "learning_rate": 4.273689273689274e-05,
+      "loss": 0.526,
+      "step": 534
+    },
+    {
+      "epoch": 0.6943543153796237,
+      "grad_norm": 0.36890093348582187,
+      "learning_rate": 4.271284271284272e-05,
+      "loss": 0.5311,
+      "step": 535
+    },
+    {
+      "epoch": 0.6956521739130435,
+      "grad_norm": 0.30993196929952377,
+      "learning_rate": 4.2688792688792686e-05,
+      "loss": 0.5246,
+      "step": 536
+    },
+    {
+      "epoch": 0.6969500324464634,
+      "grad_norm": 0.3135809737260554,
+      "learning_rate": 4.266474266474267e-05,
+      "loss": 0.5042,
+      "step": 537
+    },
+    {
+      "epoch": 0.6982478909798832,
+      "grad_norm": 0.32235641085299527,
+      "learning_rate": 4.264069264069264e-05,
+      "loss": 0.5226,
+      "step": 538
+    },
+    {
+      "epoch": 0.6995457495133031,
+      "grad_norm": 0.29955767973991637,
+      "learning_rate": 4.2616642616642624e-05,
+      "loss": 0.5003,
+      "step": 539
+    },
+    {
+      "epoch": 0.7008436080467229,
+      "grad_norm": 0.2728001277044658,
+      "learning_rate": 4.259259259259259e-05,
+      "loss": 0.5035,
+      "step": 540
+    },
+    {
+      "epoch": 0.7021414665801428,
+      "grad_norm": 0.3125500639615984,
+      "learning_rate": 4.256854256854257e-05,
+      "loss": 0.4982,
+      "step": 541
+    },
+    {
+      "epoch": 0.7034393251135627,
+      "grad_norm": 0.3214238247187388,
+      "learning_rate": 4.254449254449255e-05,
+      "loss": 0.5015,
+      "step": 542
+    },
+    {
+      "epoch": 0.7047371836469825,
+      "grad_norm": 0.34677033141949526,
+      "learning_rate": 4.2520442520442523e-05,
+      "loss": 0.5071,
+      "step": 543
+    },
+    {
+      "epoch": 0.7060350421804024,
+      "grad_norm": 0.320737794563556,
+      "learning_rate": 4.24963924963925e-05,
+      "loss": 0.5036,
+      "step": 544
+    },
+    {
+      "epoch": 0.7073329007138222,
+      "grad_norm": 0.28140631509820974,
+      "learning_rate": 4.247234247234247e-05,
+      "loss": 0.472,
+      "step": 545
+    },
+    {
+      "epoch": 0.7086307592472421,
+      "grad_norm": 0.2876262450309547,
+      "learning_rate": 4.244829244829245e-05,
+      "loss": 0.5033,
+      "step": 546
+    },
+    {
+      "epoch": 0.7099286177806619,
+      "grad_norm": 0.28203604302358143,
+      "learning_rate": 4.242424242424243e-05,
+      "loss": 0.4933,
+      "step": 547
+    },
+    {
+      "epoch": 0.7112264763140818,
+      "grad_norm": 0.3106404772330975,
+      "learning_rate": 4.2400192400192404e-05,
+      "loss": 0.4972,
+      "step": 548
+    },
+    {
+      "epoch": 0.7125243348475017,
+      "grad_norm": 0.25531650904916336,
+      "learning_rate": 4.237614237614238e-05,
+      "loss": 0.524,
+      "step": 549
+    },
+    {
+      "epoch": 0.7138221933809215,
+      "grad_norm": 0.3694832680055122,
+      "learning_rate": 4.2352092352092354e-05,
+      "loss": 0.5215,
+      "step": 550
+    },
+    {
+      "epoch": 0.7151200519143414,
+      "grad_norm": 0.29317455967258776,
+      "learning_rate": 4.232804232804233e-05,
+      "loss": 0.4978,
+      "step": 551
+    },
+    {
+      "epoch": 0.7164179104477612,
+      "grad_norm": 0.36952833950680053,
+      "learning_rate": 4.230399230399231e-05,
+      "loss": 0.5091,
+      "step": 552
+    },
+    {
+      "epoch": 0.7177157689811811,
+      "grad_norm": 0.3458300719165068,
+      "learning_rate": 4.227994227994228e-05,
+      "loss": 0.5179,
+      "step": 553
+    },
+    {
+      "epoch": 0.719013627514601,
+      "grad_norm": 0.35910338257547214,
+      "learning_rate": 4.225589225589226e-05,
+      "loss": 0.4967,
+      "step": 554
+    },
+    {
+      "epoch": 0.7203114860480208,
+      "grad_norm": 0.3832017565467235,
+      "learning_rate": 4.2231842231842234e-05,
+      "loss": 0.506,
+      "step": 555
+    },
+    {
+      "epoch": 0.7216093445814407,
+      "grad_norm": 0.3270524496685099,
+      "learning_rate": 4.220779220779221e-05,
+      "loss": 0.4956,
+      "step": 556
+    },
+    {
+      "epoch": 0.7229072031148605,
+      "grad_norm": 0.31306380662178745,
+      "learning_rate": 4.2183742183742184e-05,
+      "loss": 0.516,
+      "step": 557
+    },
+    {
+      "epoch": 0.7242050616482804,
+      "grad_norm": 0.294604631221543,
+      "learning_rate": 4.215969215969216e-05,
+      "loss": 0.5055,
+      "step": 558
+    },
+    {
+      "epoch": 0.7255029201817002,
+      "grad_norm": 0.3534780044338388,
+      "learning_rate": 4.213564213564214e-05,
+      "loss": 0.4975,
+      "step": 559
+    },
+    {
+      "epoch": 0.72680077871512,
+      "grad_norm": 0.33032987931239965,
+      "learning_rate": 4.2111592111592115e-05,
+      "loss": 0.5118,
+      "step": 560
+    },
+    {
+      "epoch": 0.72809863724854,
+      "grad_norm": 0.3196832192635056,
+      "learning_rate": 4.208754208754209e-05,
+      "loss": 0.5093,
+      "step": 561
+    },
+    {
+      "epoch": 0.7293964957819598,
+      "grad_norm": 0.36785704594666774,
+      "learning_rate": 4.2063492063492065e-05,
+      "loss": 0.5069,
+      "step": 562
+    },
+    {
+      "epoch": 0.7306943543153797,
+      "grad_norm": 0.3702503469527744,
+      "learning_rate": 4.203944203944204e-05,
+      "loss": 0.513,
+      "step": 563
+    },
+    {
+      "epoch": 0.7319922128487995,
+      "grad_norm": 0.3070674221331164,
+      "learning_rate": 4.201539201539202e-05,
+      "loss": 0.497,
+      "step": 564
+    },
+    {
+      "epoch": 0.7332900713822194,
+      "grad_norm": 0.35750959007798994,
+      "learning_rate": 4.1991341991341996e-05,
+      "loss": 0.5085,
+      "step": 565
+    },
+    {
+      "epoch": 0.7345879299156391,
+      "grad_norm": 0.2835364219292076,
+      "learning_rate": 4.196729196729197e-05,
+      "loss": 0.5091,
+      "step": 566
+    },
+    {
+      "epoch": 0.735885788449059,
+      "grad_norm": 0.2884098082465498,
+      "learning_rate": 4.1943241943241945e-05,
+      "loss": 0.4884,
+      "step": 567
+    },
+    {
+      "epoch": 0.737183646982479,
+      "grad_norm": 0.3203510406175552,
+      "learning_rate": 4.191919191919192e-05,
+      "loss": 0.4971,
+      "step": 568
+    },
+    {
+      "epoch": 0.7384815055158988,
+      "grad_norm": 0.27371373687668255,
+      "learning_rate": 4.18951418951419e-05,
+      "loss": 0.5095,
+      "step": 569
+    },
+    {
+      "epoch": 0.7397793640493187,
+      "grad_norm": 0.34717402203397457,
+      "learning_rate": 4.187109187109187e-05,
+      "loss": 0.5014,
+      "step": 570
+    },
+    {
+      "epoch": 0.7410772225827384,
+      "grad_norm": 0.30582393639621713,
+      "learning_rate": 4.184704184704185e-05,
+      "loss": 0.5181,
+      "step": 571
+    },
+    {
+      "epoch": 0.7423750811161584,
+      "grad_norm": 0.32112393843480735,
+      "learning_rate": 4.1822991822991826e-05,
+      "loss": 0.5006,
+      "step": 572
+    },
+    {
+      "epoch": 0.7436729396495781,
+      "grad_norm": 0.33979137877685406,
+      "learning_rate": 4.17989417989418e-05,
+      "loss": 0.5248,
+      "step": 573
+    },
+    {
+      "epoch": 0.744970798182998,
+      "grad_norm": 0.3209001348833202,
+      "learning_rate": 4.1774891774891775e-05,
+      "loss": 0.5013,
+      "step": 574
+    },
+    {
+      "epoch": 0.746268656716418,
+      "grad_norm": 0.3266878409508907,
+      "learning_rate": 4.175084175084175e-05,
+      "loss": 0.4915,
+      "step": 575
+    },
+    {
+      "epoch": 0.7475665152498377,
+      "grad_norm": 0.30503612561210064,
+      "learning_rate": 4.172679172679173e-05,
+      "loss": 0.5046,
+      "step": 576
+    },
+    {
+      "epoch": 0.7488643737832577,
+      "grad_norm": 0.28971405798539174,
+      "learning_rate": 4.1702741702741707e-05,
+      "loss": 0.494,
+      "step": 577
+    },
+    {
+      "epoch": 0.7501622323166774,
+      "grad_norm": 0.29780493895562776,
+      "learning_rate": 4.167869167869168e-05,
+      "loss": 0.492,
+      "step": 578
+    },
+    {
+      "epoch": 0.7514600908500974,
+      "grad_norm": 0.32685361493691256,
+      "learning_rate": 4.1654641654641656e-05,
+      "loss": 0.5101,
+      "step": 579
+    },
+    {
+      "epoch": 0.7527579493835171,
+      "grad_norm": 0.34614664346128227,
+      "learning_rate": 4.163059163059163e-05,
+      "loss": 0.4953,
+      "step": 580
+    },
+    {
+      "epoch": 0.754055807916937,
+      "grad_norm": 0.24964382812915295,
+      "learning_rate": 4.160654160654161e-05,
+      "loss": 0.4974,
+      "step": 581
+    },
+    {
+      "epoch": 0.755353666450357,
+      "grad_norm": 0.3266391465805975,
+      "learning_rate": 4.158249158249159e-05,
+      "loss": 0.5039,
+      "step": 582
+    },
+    {
+      "epoch": 0.7566515249837767,
+      "grad_norm": 0.34857873431761155,
+      "learning_rate": 4.155844155844156e-05,
+      "loss": 0.5056,
+      "step": 583
+    },
+    {
+      "epoch": 0.7579493835171967,
+      "grad_norm": 0.2921239047290261,
+      "learning_rate": 4.153439153439154e-05,
+      "loss": 0.4967,
+      "step": 584
+    },
+    {
+      "epoch": 0.7592472420506164,
+      "grad_norm": 0.33268372260105683,
+      "learning_rate": 4.151034151034151e-05,
+      "loss": 0.5066,
+      "step": 585
+    },
+    {
+      "epoch": 0.7605451005840363,
+      "grad_norm": 0.318064660501317,
+      "learning_rate": 4.148629148629149e-05,
+      "loss": 0.5014,
+      "step": 586
+    },
+    {
+      "epoch": 0.7618429591174561,
+      "grad_norm": 0.40789097570888044,
+      "learning_rate": 4.146224146224146e-05,
+      "loss": 0.5236,
+      "step": 587
+    },
+    {
+      "epoch": 0.763140817650876,
+      "grad_norm": 0.36460994866717067,
+      "learning_rate": 4.143819143819144e-05,
+      "loss": 0.5028,
+      "step": 588
+    },
+    {
+      "epoch": 0.764438676184296,
+      "grad_norm": 0.32621095461004207,
+      "learning_rate": 4.141414141414142e-05,
+      "loss": 0.5,
+      "step": 589
+    },
+    {
+      "epoch": 0.7657365347177157,
+      "grad_norm": 0.35949920016397585,
+      "learning_rate": 4.139009139009139e-05,
+      "loss": 0.5028,
+      "step": 590
+    },
+    {
+      "epoch": 0.7670343932511356,
+      "grad_norm": 0.26064162839844207,
+      "learning_rate": 4.136604136604137e-05,
+      "loss": 0.4926,
+      "step": 591
+    },
+    {
+      "epoch": 0.7683322517845554,
+      "grad_norm": 0.33955569046497985,
+      "learning_rate": 4.134199134199134e-05,
+      "loss": 0.491,
+      "step": 592
+    },
+    {
+      "epoch": 0.7696301103179753,
+      "grad_norm": 0.302141896432662,
+      "learning_rate": 4.131794131794132e-05,
+      "loss": 0.4884,
+      "step": 593
+    },
+    {
+      "epoch": 0.7709279688513953,
+      "grad_norm": 0.30640851435457384,
+      "learning_rate": 4.12938912938913e-05,
+      "loss": 0.4993,
+      "step": 594
+    },
+    {
+      "epoch": 0.772225827384815,
+      "grad_norm": 0.3040997259974209,
+      "learning_rate": 4.126984126984127e-05,
+      "loss": 0.5207,
+      "step": 595
+    },
+    {
+      "epoch": 0.773523685918235,
+      "grad_norm": 0.3146837697580934,
+      "learning_rate": 4.124579124579125e-05,
+      "loss": 0.5041,
+      "step": 596
+    },
+    {
+      "epoch": 0.7748215444516547,
+      "grad_norm": 0.2837662015282775,
+      "learning_rate": 4.122174122174122e-05,
+      "loss": 0.4958,
+      "step": 597
+    },
+    {
+      "epoch": 0.7761194029850746,
+      "grad_norm": 0.3005344336762094,
+      "learning_rate": 4.1197691197691204e-05,
+      "loss": 0.4995,
+      "step": 598
+    },
+    {
+      "epoch": 0.7774172615184944,
+      "grad_norm": 0.2593994091267606,
+      "learning_rate": 4.117364117364118e-05,
+      "loss": 0.4949,
+      "step": 599
+    },
+    {
+      "epoch": 0.7787151200519143,
+      "grad_norm": 0.31127336265884026,
+      "learning_rate": 4.114959114959115e-05,
+      "loss": 0.5026,
+      "step": 600
+    },
+    {
+      "epoch": 0.7800129785853342,
+      "grad_norm": 0.24332809263192706,
+      "learning_rate": 4.112554112554113e-05,
+      "loss": 0.506,
+      "step": 601
+    },
+    {
+      "epoch": 0.781310837118754,
+      "grad_norm": 0.268091615023721,
+      "learning_rate": 4.11014911014911e-05,
+      "loss": 0.5148,
+      "step": 602
+    },
+    {
+      "epoch": 0.782608695652174,
+      "grad_norm": 0.27606994914293354,
+      "learning_rate": 4.1077441077441085e-05,
+      "loss": 0.5058,
+      "step": 603
+    },
+    {
+      "epoch": 0.7839065541855937,
+      "grad_norm": 0.25091695629276883,
+      "learning_rate": 4.105339105339105e-05,
+      "loss": 0.4939,
+      "step": 604
+    },
+    {
+      "epoch": 0.7852044127190136,
+      "grad_norm": 0.25969299604058604,
+      "learning_rate": 4.1029341029341034e-05,
+      "loss": 0.5045,
+      "step": 605
+    },
+    {
+      "epoch": 0.7865022712524334,
+      "grad_norm": 0.28645937532292653,
+      "learning_rate": 4.100529100529101e-05,
+      "loss": 0.51,
+      "step": 606
+    },
+    {
+      "epoch": 0.7878001297858533,
+      "grad_norm": 0.2867588479387004,
+      "learning_rate": 4.0981240981240984e-05,
+      "loss": 0.4849,
+      "step": 607
+    },
+    {
+      "epoch": 0.7890979883192732,
+      "grad_norm": 0.2646936431763157,
+      "learning_rate": 4.095719095719096e-05,
+      "loss": 0.5139,
+      "step": 608
+    },
+    {
+      "epoch": 0.790395846852693,
+      "grad_norm": 0.3083485975619241,
+      "learning_rate": 4.093314093314093e-05,
+      "loss": 0.52,
+      "step": 609
+    },
+    {
+      "epoch": 0.791693705386113,
+      "grad_norm": 0.29615253606758346,
+      "learning_rate": 4.0909090909090915e-05,
+      "loss": 0.491,
+      "step": 610
+    },
+    {
+      "epoch": 0.7929915639195327,
+      "grad_norm": 0.29803496441037525,
+      "learning_rate": 4.088504088504089e-05,
+      "loss": 0.5054,
+      "step": 611
+    },
+    {
+      "epoch": 0.7942894224529526,
+      "grad_norm": 0.2854912264040868,
+      "learning_rate": 4.0860990860990864e-05,
+      "loss": 0.4838,
+      "step": 612
+    },
+    {
+      "epoch": 0.7955872809863724,
+      "grad_norm": 0.2861142997625756,
+      "learning_rate": 4.083694083694084e-05,
+      "loss": 0.4777,
+      "step": 613
+    },
+    {
+      "epoch": 0.7968851395197923,
+      "grad_norm": 0.28878974958497106,
+      "learning_rate": 4.0812890812890814e-05,
+      "loss": 0.5043,
+      "step": 614
+    },
+    {
+      "epoch": 0.7981829980532122,
+      "grad_norm": 0.28908126130671624,
+      "learning_rate": 4.0788840788840795e-05,
+      "loss": 0.5077,
+      "step": 615
+    },
+    {
+      "epoch": 0.799480856586632,
+      "grad_norm": 0.31059225731020423,
+      "learning_rate": 4.0764790764790763e-05,
+      "loss": 0.4846,
+      "step": 616
+    },
+    {
+      "epoch": 0.8007787151200519,
+      "grad_norm": 0.2990367658353648,
+      "learning_rate": 4.074074074074074e-05,
+      "loss": 0.5075,
+      "step": 617
+    },
+    {
+      "epoch": 0.8020765736534717,
+      "grad_norm": 0.29529728033655306,
+      "learning_rate": 4.071669071669072e-05,
+      "loss": 0.4876,
+      "step": 618
+    },
+    {
+      "epoch": 0.8033744321868916,
+      "grad_norm": 0.299331177429087,
+      "learning_rate": 4.0692640692640695e-05,
+      "loss": 0.5032,
+      "step": 619
+    },
+    {
+      "epoch": 0.8046722907203114,
+      "grad_norm": 0.2842657684401892,
+      "learning_rate": 4.066859066859067e-05,
+      "loss": 0.4889,
+      "step": 620
+    },
+    {
+      "epoch": 0.8059701492537313,
+      "grad_norm": 0.2692733717815561,
+      "learning_rate": 4.0644540644540644e-05,
+      "loss": 0.5009,
+      "step": 621
+    },
+    {
+      "epoch": 0.8072680077871512,
+      "grad_norm": 0.3436883319030681,
+      "learning_rate": 4.062049062049062e-05,
+      "loss": 0.5185,
+      "step": 622
+    },
+    {
+      "epoch": 0.808565866320571,
+      "grad_norm": 0.2653175278056993,
+      "learning_rate": 4.05964405964406e-05,
+      "loss": 0.4739,
+      "step": 623
+    },
+    {
+      "epoch": 0.8098637248539909,
+      "grad_norm": 0.31209657077735303,
+      "learning_rate": 4.0572390572390575e-05,
+      "loss": 0.4944,
+      "step": 624
+    },
+    {
+      "epoch": 0.8111615833874107,
+      "grad_norm": 0.3396825057908641,
+      "learning_rate": 4.054834054834055e-05,
+      "loss": 0.5203,
+      "step": 625
+    },
+    {
+      "epoch": 0.8124594419208306,
+      "grad_norm": 0.31652376647001546,
+      "learning_rate": 4.0524290524290525e-05,
+      "loss": 0.4973,
+      "step": 626
+    },
+    {
+      "epoch": 0.8137573004542504,
+      "grad_norm": 0.41987696956302806,
+      "learning_rate": 4.05002405002405e-05,
+      "loss": 0.5008,
+      "step": 627
+    },
+    {
+      "epoch": 0.8150551589876703,
+      "grad_norm": 0.32125784926625567,
+      "learning_rate": 4.047619047619048e-05,
+      "loss": 0.5107,
+      "step": 628
+    },
+    {
+      "epoch": 0.8163530175210902,
+      "grad_norm": 0.43302794720660975,
+      "learning_rate": 4.045214045214045e-05,
+      "loss": 0.5174,
+      "step": 629
+    },
+    {
+      "epoch": 0.81765087605451,
+      "grad_norm": 0.29529734876987174,
+      "learning_rate": 4.042809042809043e-05,
+      "loss": 0.4881,
+      "step": 630
+    },
+    {
+      "epoch": 0.8189487345879299,
+      "grad_norm": 0.4076264173563411,
+      "learning_rate": 4.0404040404040405e-05,
+      "loss": 0.5034,
+      "step": 631
+    },
+    {
+      "epoch": 0.8202465931213497,
+      "grad_norm": 0.30337707563686833,
+      "learning_rate": 4.037999037999039e-05,
+      "loss": 0.5119,
+      "step": 632
+    },
+    {
+      "epoch": 0.8215444516547696,
+      "grad_norm": 0.39849923453663594,
+      "learning_rate": 4.0355940355940355e-05,
+      "loss": 0.5216,
+      "step": 633
+    },
+    {
+      "epoch": 0.8228423101881895,
+      "grad_norm": 0.26447226558452136,
+      "learning_rate": 4.033189033189033e-05,
+      "loss": 0.5122,
+      "step": 634
+    },
+    {
+      "epoch": 0.8241401687216093,
+      "grad_norm": 0.36530243282807756,
+      "learning_rate": 4.030784030784031e-05,
+      "loss": 0.4918,
+      "step": 635
+    },
+    {
+      "epoch": 0.8254380272550292,
+      "grad_norm": 0.3160155549438362,
+      "learning_rate": 4.0283790283790286e-05,
+      "loss": 0.5024,
+      "step": 636
+    },
+    {
+      "epoch": 0.826735885788449,
+      "grad_norm": 0.33636766065888035,
+      "learning_rate": 4.025974025974026e-05,
+      "loss": 0.5075,
+      "step": 637
+    },
+    {
+      "epoch": 0.8280337443218689,
+      "grad_norm": 0.29456591212102723,
+      "learning_rate": 4.0235690235690236e-05,
+      "loss": 0.5241,
+      "step": 638
+    },
+    {
+      "epoch": 0.8293316028552887,
+      "grad_norm": 0.3220137418817016,
+      "learning_rate": 4.021164021164021e-05,
+      "loss": 0.5028,
+      "step": 639
+    },
+    {
+      "epoch": 0.8306294613887086,
+      "grad_norm": 0.279849046005973,
+      "learning_rate": 4.018759018759019e-05,
+      "loss": 0.4937,
+      "step": 640
+    },
+    {
+      "epoch": 0.8319273199221285,
+      "grad_norm": 0.34243863539028374,
+      "learning_rate": 4.016354016354017e-05,
+      "loss": 0.4992,
+      "step": 641
+    },
+    {
+      "epoch": 0.8332251784555483,
+      "grad_norm": 0.3077281111260478,
+      "learning_rate": 4.013949013949014e-05,
+      "loss": 0.5013,
+      "step": 642
+    },
+    {
+      "epoch": 0.8345230369889682,
+      "grad_norm": 0.2917135387110751,
+      "learning_rate": 4.0115440115440116e-05,
+      "loss": 0.5084,
+      "step": 643
+    },
+    {
+      "epoch": 0.835820895522388,
+      "grad_norm": 0.4035461806364624,
+      "learning_rate": 4.009139009139009e-05,
+      "loss": 0.5075,
+      "step": 644
+    },
+    {
+      "epoch": 0.8371187540558079,
+      "grad_norm": 0.28209498726622767,
+      "learning_rate": 4.006734006734007e-05,
+      "loss": 0.5018,
+      "step": 645
+    },
+    {
+      "epoch": 0.8384166125892277,
+      "grad_norm": 0.32365984928312647,
+      "learning_rate": 4.004329004329004e-05,
+      "loss": 0.5059,
+      "step": 646
+    },
+    {
+      "epoch": 0.8397144711226476,
+      "grad_norm": 0.27577530319773297,
+      "learning_rate": 4.001924001924002e-05,
+      "loss": 0.4833,
+      "step": 647
+    },
+    {
+      "epoch": 0.8410123296560675,
+      "grad_norm": 0.28322226416250573,
+      "learning_rate": 3.999518999519e-05,
+      "loss": 0.4918,
+      "step": 648
+    },
+    {
+      "epoch": 0.8423101881894873,
+      "grad_norm": 0.2855631382527533,
+      "learning_rate": 3.997113997113997e-05,
+      "loss": 0.4865,
+      "step": 649
+    },
+    {
+      "epoch": 0.8436080467229072,
+      "grad_norm": 0.32968743054146016,
+      "learning_rate": 3.9947089947089946e-05,
+      "loss": 0.4995,
+      "step": 650
+    },
+    {
+      "epoch": 0.844905905256327,
+      "grad_norm": 0.2546632505302031,
+      "learning_rate": 3.992303992303992e-05,
+      "loss": 0.5007,
+      "step": 651
+    },
+    {
+      "epoch": 0.8462037637897469,
+      "grad_norm": 0.3379202280608477,
+      "learning_rate": 3.98989898989899e-05,
+      "loss": 0.5191,
+      "step": 652
+    },
+    {
+      "epoch": 0.8475016223231667,
+      "grad_norm": 0.2828398301217406,
+      "learning_rate": 3.987493987493988e-05,
+      "loss": 0.511,
+      "step": 653
+    },
+    {
+      "epoch": 0.8487994808565866,
+      "grad_norm": 0.2759621170821239,
+      "learning_rate": 3.985088985088985e-05,
+      "loss": 0.4833,
+      "step": 654
+    },
+    {
+      "epoch": 0.8500973393900065,
+      "grad_norm": 0.30549235671106895,
+      "learning_rate": 3.982683982683983e-05,
+      "loss": 0.5041,
+      "step": 655
+    },
+    {
+      "epoch": 0.8513951979234263,
+      "grad_norm": 0.27394626074529427,
+      "learning_rate": 3.98027898027898e-05,
+      "loss": 0.4765,
+      "step": 656
+    },
+    {
+      "epoch": 0.8526930564568462,
+      "grad_norm": 0.28778374402990897,
+      "learning_rate": 3.9778739778739783e-05,
+      "loss": 0.5204,
+      "step": 657
+    },
+    {
+      "epoch": 0.853990914990266,
+      "grad_norm": 0.34038234379097493,
+      "learning_rate": 3.975468975468976e-05,
+      "loss": 0.4974,
+      "step": 658
+    },
+    {
+      "epoch": 0.8552887735236859,
+      "grad_norm": 0.2866059277868264,
+      "learning_rate": 3.973063973063973e-05,
+      "loss": 0.5054,
+      "step": 659
+    },
+    {
+      "epoch": 0.8565866320571057,
+      "grad_norm": 0.2927739560082091,
+      "learning_rate": 3.970658970658971e-05,
+      "loss": 0.4974,
+      "step": 660
+    },
+    {
+      "epoch": 0.8578844905905256,
+      "grad_norm": 0.3262341662863849,
+      "learning_rate": 3.968253968253968e-05,
+      "loss": 0.4989,
+      "step": 661
+    },
+    {
+      "epoch": 0.8591823491239455,
+      "grad_norm": 0.2845679103896212,
+      "learning_rate": 3.9658489658489664e-05,
+      "loss": 0.4874,
+      "step": 662
+    },
+    {
+      "epoch": 0.8604802076573653,
+      "grad_norm": 0.2655596774616536,
+      "learning_rate": 3.963443963443963e-05,
+      "loss": 0.4843,
+      "step": 663
+    },
+    {
+      "epoch": 0.8617780661907852,
+      "grad_norm": 0.2820305028454277,
+      "learning_rate": 3.9610389610389614e-05,
+      "loss": 0.5164,
+      "step": 664
+    },
+    {
+      "epoch": 0.863075924724205,
+      "grad_norm": 0.2940016391705861,
+      "learning_rate": 3.958633958633959e-05,
+      "loss": 0.4886,
+      "step": 665
+    },
+    {
+      "epoch": 0.8643737832576249,
+      "grad_norm": 0.2547644694514986,
+      "learning_rate": 3.956228956228956e-05,
+      "loss": 0.5051,
+      "step": 666
+    },
+    {
+      "epoch": 0.8656716417910447,
+      "grad_norm": 0.26023827863988136,
+      "learning_rate": 3.953823953823954e-05,
+      "loss": 0.5217,
+      "step": 667
+    },
+    {
+      "epoch": 0.8669695003244646,
+      "grad_norm": 0.27927530276749113,
+      "learning_rate": 3.951418951418951e-05,
+      "loss": 0.5092,
+      "step": 668
+    },
+    {
+      "epoch": 0.8682673588578845,
+      "grad_norm": 0.26218361367284654,
+      "learning_rate": 3.9490139490139494e-05,
+      "loss": 0.4889,
+      "step": 669
+    },
+    {
+      "epoch": 0.8695652173913043,
+      "grad_norm": 0.2574083587232944,
+      "learning_rate": 3.946608946608947e-05,
+      "loss": 0.498,
+      "step": 670
+    },
+    {
+      "epoch": 0.8708630759247242,
+      "grad_norm": 0.28970396117550484,
+      "learning_rate": 3.9442039442039444e-05,
+      "loss": 0.5005,
+      "step": 671
+    },
+    {
+      "epoch": 0.872160934458144,
+      "grad_norm": 0.253672603517478,
+      "learning_rate": 3.941798941798942e-05,
+      "loss": 0.4802,
+      "step": 672
+    },
+    {
+      "epoch": 0.8734587929915639,
+      "grad_norm": 0.28176527595198714,
+      "learning_rate": 3.939393939393939e-05,
+      "loss": 0.5009,
+      "step": 673
+    },
+    {
+      "epoch": 0.8747566515249838,
+      "grad_norm": 0.29970103183262975,
+      "learning_rate": 3.9369889369889375e-05,
+      "loss": 0.494,
+      "step": 674
+    },
+    {
+      "epoch": 0.8760545100584036,
+      "grad_norm": 0.28208623298136165,
+      "learning_rate": 3.934583934583935e-05,
+      "loss": 0.4784,
+      "step": 675
+    },
+    {
+      "epoch": 0.8773523685918235,
+      "grad_norm": 0.31291964058774646,
+      "learning_rate": 3.9321789321789324e-05,
+      "loss": 0.4869,
+      "step": 676
+    },
+    {
+      "epoch": 0.8786502271252433,
+      "grad_norm": 0.2979349183006068,
+      "learning_rate": 3.92977392977393e-05,
+      "loss": 0.5046,
+      "step": 677
+    },
+    {
+      "epoch": 0.8799480856586632,
+      "grad_norm": 0.2908014815865184,
+      "learning_rate": 3.9273689273689274e-05,
+      "loss": 0.5265,
+      "step": 678
+    },
+    {
+      "epoch": 0.881245944192083,
+      "grad_norm": 0.26416143914819024,
+      "learning_rate": 3.9249639249639256e-05,
+      "loss": 0.4818,
+      "step": 679
+    },
+    {
+      "epoch": 0.8825438027255029,
+      "grad_norm": 0.2933751247833131,
+      "learning_rate": 3.9225589225589224e-05,
+      "loss": 0.5012,
+      "step": 680
+    },
+    {
+      "epoch": 0.8838416612589228,
+      "grad_norm": 0.3126465117694497,
+      "learning_rate": 3.9201539201539205e-05,
+      "loss": 0.5146,
+      "step": 681
+    },
+    {
+      "epoch": 0.8851395197923426,
+      "grad_norm": 0.2524418661547154,
+      "learning_rate": 3.917748917748918e-05,
+      "loss": 0.5016,
+      "step": 682
+    },
+    {
+      "epoch": 0.8864373783257625,
+      "grad_norm": 0.25838371662824994,
+      "learning_rate": 3.9153439153439155e-05,
+      "loss": 0.4788,
+      "step": 683
+    },
+    {
+      "epoch": 0.8877352368591823,
+      "grad_norm": 0.26415640627712256,
+      "learning_rate": 3.912938912938913e-05,
+      "loss": 0.5026,
+      "step": 684
+    },
+    {
+      "epoch": 0.8890330953926022,
+      "grad_norm": 0.24247698201137077,
+      "learning_rate": 3.9105339105339104e-05,
+      "loss": 0.4785,
+      "step": 685
+    },
+    {
+      "epoch": 0.890330953926022,
+      "grad_norm": 0.2592072694481307,
+      "learning_rate": 3.9081289081289086e-05,
+      "loss": 0.4977,
+      "step": 686
+    },
+    {
+      "epoch": 0.8916288124594419,
+      "grad_norm": 0.2542832037438216,
+      "learning_rate": 3.905723905723906e-05,
+      "loss": 0.4965,
+      "step": 687
+    },
+    {
+      "epoch": 0.8929266709928618,
+      "grad_norm": 0.262249098843731,
+      "learning_rate": 3.9033189033189035e-05,
+      "loss": 0.4973,
+      "step": 688
+    },
+    {
+      "epoch": 0.8942245295262816,
+      "grad_norm": 0.31610300163969496,
+      "learning_rate": 3.900913900913901e-05,
+      "loss": 0.4825,
+      "step": 689
+    },
+    {
+      "epoch": 0.8955223880597015,
+      "grad_norm": 0.2740619461513036,
+      "learning_rate": 3.8985088985088985e-05,
+      "loss": 0.4975,
+      "step": 690
+    },
+    {
+      "epoch": 0.8968202465931213,
+      "grad_norm": 0.2741811791339238,
+      "learning_rate": 3.8961038961038966e-05,
+      "loss": 0.4959,
+      "step": 691
+    },
+    {
+      "epoch": 0.8981181051265412,
+      "grad_norm": 0.271574424484243,
+      "learning_rate": 3.893698893698894e-05,
+      "loss": 0.4871,
+      "step": 692
+    },
+    {
+      "epoch": 0.899415963659961,
+      "grad_norm": 0.23758978275848563,
+      "learning_rate": 3.891293891293891e-05,
+      "loss": 0.4862,
+      "step": 693
+    },
+    {
+      "epoch": 0.9007138221933809,
+      "grad_norm": 0.27263432443106744,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 0.5041,
+      "step": 694
+    },
+    {
+      "epoch": 0.9020116807268008,
+      "grad_norm": 0.2707424886921677,
+      "learning_rate": 3.8864838864838866e-05,
+      "loss": 0.4669,
+      "step": 695
+    },
+    {
+      "epoch": 0.9033095392602206,
+      "grad_norm": 0.25600324849109557,
+      "learning_rate": 3.884078884078885e-05,
+      "loss": 0.4915,
+      "step": 696
+    },
+    {
+      "epoch": 0.9046073977936405,
+      "grad_norm": 0.28010923150865535,
+      "learning_rate": 3.8816738816738815e-05,
+      "loss": 0.5038,
+      "step": 697
+    },
+    {
+      "epoch": 0.9059052563270603,
+      "grad_norm": 0.28506679888273495,
+      "learning_rate": 3.87926887926888e-05,
+      "loss": 0.4965,
+      "step": 698
+    },
+    {
+      "epoch": 0.9072031148604802,
+      "grad_norm": 0.26956128070889057,
+      "learning_rate": 3.876863876863877e-05,
+      "loss": 0.5141,
+      "step": 699
+    },
+    {
+      "epoch": 0.9085009733939,
+      "grad_norm": 0.25947152657252537,
+      "learning_rate": 3.8744588744588746e-05,
+      "loss": 0.477,
+      "step": 700
+    },
+    {
+      "epoch": 0.9097988319273199,
+      "grad_norm": 0.29443853949197607,
+      "learning_rate": 3.872053872053872e-05,
+      "loss": 0.5059,
+      "step": 701
+    },
+    {
+      "epoch": 0.9110966904607398,
+      "grad_norm": 0.23371391184316104,
+      "learning_rate": 3.8696488696488696e-05,
+      "loss": 0.4783,
+      "step": 702
+    },
+    {
+      "epoch": 0.9123945489941596,
+      "grad_norm": 0.2960003000238748,
+      "learning_rate": 3.867243867243868e-05,
+      "loss": 0.4975,
+      "step": 703
+    },
+    {
+      "epoch": 0.9136924075275795,
+      "grad_norm": 0.2447202791191107,
+      "learning_rate": 3.864838864838865e-05,
+      "loss": 0.4922,
+      "step": 704
+    },
+    {
+      "epoch": 0.9149902660609993,
+      "grad_norm": 0.26642870018861,
+      "learning_rate": 3.862433862433863e-05,
+      "loss": 0.5252,
+      "step": 705
+    },
+    {
+      "epoch": 0.9162881245944192,
+      "grad_norm": 0.2560976171455726,
+      "learning_rate": 3.86002886002886e-05,
+      "loss": 0.5012,
+      "step": 706
+    },
+    {
+      "epoch": 0.917585983127839,
+      "grad_norm": 0.28883925682227224,
+      "learning_rate": 3.8576238576238576e-05,
+      "loss": 0.509,
+      "step": 707
+    },
+    {
+      "epoch": 0.9188838416612589,
+      "grad_norm": 0.2513723680500846,
+      "learning_rate": 3.855218855218856e-05,
+      "loss": 0.5064,
+      "step": 708
+    },
+    {
+      "epoch": 0.9201817001946788,
+      "grad_norm": 0.26385183299541554,
+      "learning_rate": 3.852813852813853e-05,
+      "loss": 0.4844,
+      "step": 709
+    },
+    {
+      "epoch": 0.9214795587280986,
+      "grad_norm": 0.2739471730680778,
+      "learning_rate": 3.85040885040885e-05,
+      "loss": 0.5062,
+      "step": 710
+    },
+    {
+      "epoch": 0.9227774172615185,
+      "grad_norm": 0.29349547383156327,
+      "learning_rate": 3.848003848003848e-05,
+      "loss": 0.4663,
+      "step": 711
+    },
+    {
+      "epoch": 0.9240752757949383,
+      "grad_norm": 0.24335538681766872,
+      "learning_rate": 3.845598845598846e-05,
+      "loss": 0.4802,
+      "step": 712
+    },
+    {
+      "epoch": 0.9253731343283582,
+      "grad_norm": 0.26992374528269225,
+      "learning_rate": 3.843193843193844e-05,
+      "loss": 0.4737,
+      "step": 713
+    },
+    {
+      "epoch": 0.9266709928617781,
+      "grad_norm": 0.2601068353604271,
+      "learning_rate": 3.8407888407888407e-05,
+      "loss": 0.4871,
+      "step": 714
+    },
+    {
+      "epoch": 0.9279688513951979,
+      "grad_norm": 0.2758032793334676,
+      "learning_rate": 3.838383838383838e-05,
+      "loss": 0.497,
+      "step": 715
+    },
+    {
+      "epoch": 0.9292667099286178,
+      "grad_norm": 0.28995105804236226,
+      "learning_rate": 3.835978835978836e-05,
+      "loss": 0.4954,
+      "step": 716
+    },
+    {
+      "epoch": 0.9305645684620376,
+      "grad_norm": 0.2711126929498793,
+      "learning_rate": 3.833573833573834e-05,
+      "loss": 0.4899,
+      "step": 717
+    },
+    {
+      "epoch": 0.9318624269954575,
+      "grad_norm": 0.27659146601067053,
+      "learning_rate": 3.831168831168831e-05,
+      "loss": 0.4724,
+      "step": 718
+    },
+    {
+      "epoch": 0.9331602855288773,
+      "grad_norm": 0.27572149711082233,
+      "learning_rate": 3.828763828763829e-05,
+      "loss": 0.4771,
+      "step": 719
+    },
+    {
+      "epoch": 0.9344581440622972,
+      "grad_norm": 0.26747547869691946,
+      "learning_rate": 3.826358826358827e-05,
+      "loss": 0.508,
+      "step": 720
+    },
+    {
+      "epoch": 0.9357560025957171,
+      "grad_norm": 0.25233671196063495,
+      "learning_rate": 3.8239538239538244e-05,
+      "loss": 0.5112,
+      "step": 721
+    },
+    {
+      "epoch": 0.9370538611291369,
+      "grad_norm": 0.25520294945614,
+      "learning_rate": 3.821548821548822e-05,
+      "loss": 0.5015,
+      "step": 722
+    },
+    {
+      "epoch": 0.9383517196625568,
+      "grad_norm": 0.2638285175394357,
+      "learning_rate": 3.819143819143819e-05,
+      "loss": 0.5159,
+      "step": 723
+    },
+    {
+      "epoch": 0.9396495781959766,
+      "grad_norm": 0.2424275513941752,
+      "learning_rate": 3.816738816738817e-05,
+      "loss": 0.4934,
+      "step": 724
+    },
+    {
+      "epoch": 0.9409474367293965,
+      "grad_norm": 0.25978321355889966,
+      "learning_rate": 3.814333814333815e-05,
+      "loss": 0.493,
+      "step": 725
+    },
+    {
+      "epoch": 0.9422452952628163,
+      "grad_norm": 0.2717237370055125,
+      "learning_rate": 3.8119288119288124e-05,
+      "loss": 0.4848,
+      "step": 726
+    },
+    {
+      "epoch": 0.9435431537962362,
+      "grad_norm": 0.2649262510228781,
+      "learning_rate": 3.809523809523809e-05,
+      "loss": 0.5106,
+      "step": 727
+    },
+    {
+      "epoch": 0.9448410123296561,
+      "grad_norm": 0.2630223525297261,
+      "learning_rate": 3.8071188071188074e-05,
+      "loss": 0.4819,
+      "step": 728
+    },
+    {
+      "epoch": 0.9461388708630759,
+      "grad_norm": 0.2552060740231092,
+      "learning_rate": 3.804713804713805e-05,
+      "loss": 0.4846,
+      "step": 729
+    },
+    {
+      "epoch": 0.9474367293964958,
+      "grad_norm": 0.27269809006023926,
+      "learning_rate": 3.802308802308803e-05,
+      "loss": 0.4646,
+      "step": 730
+    },
+    {
+      "epoch": 0.9487345879299156,
+      "grad_norm": 0.26086007457601984,
+      "learning_rate": 3.7999037999038e-05,
+      "loss": 0.4881,
+      "step": 731
+    },
+    {
+      "epoch": 0.9500324464633355,
+      "grad_norm": 0.261121106269547,
+      "learning_rate": 3.797498797498797e-05,
+      "loss": 0.4676,
+      "step": 732
+    },
+    {
+      "epoch": 0.9513303049967553,
+      "grad_norm": 0.24120860308268785,
+      "learning_rate": 3.7950937950937954e-05,
+      "loss": 0.475,
+      "step": 733
+    },
+    {
+      "epoch": 0.9526281635301752,
+      "grad_norm": 0.23338097940225563,
+      "learning_rate": 3.792688792688793e-05,
+      "loss": 0.4913,
+      "step": 734
+    },
+    {
+      "epoch": 0.9539260220635951,
+      "grad_norm": 0.27856238721903515,
+      "learning_rate": 3.7902837902837904e-05,
+      "loss": 0.4817,
+      "step": 735
+    },
+    {
+      "epoch": 0.9552238805970149,
+      "grad_norm": 0.23489180306981414,
+      "learning_rate": 3.787878787878788e-05,
+      "loss": 0.4876,
+      "step": 736
+    },
+    {
+      "epoch": 0.9565217391304348,
+      "grad_norm": 0.24747777349253483,
+      "learning_rate": 3.7854737854737854e-05,
+      "loss": 0.4926,
+      "step": 737
+    },
+    {
+      "epoch": 0.9578195976638546,
+      "grad_norm": 0.2464671646367973,
+      "learning_rate": 3.7830687830687835e-05,
+      "loss": 0.4988,
+      "step": 738
+    },
+    {
+      "epoch": 0.9591174561972745,
+      "grad_norm": 0.2444290486715431,
+      "learning_rate": 3.780663780663781e-05,
+      "loss": 0.5102,
+      "step": 739
+    },
+    {
+      "epoch": 0.9604153147306943,
+      "grad_norm": 0.23262037897628499,
+      "learning_rate": 3.7782587782587785e-05,
+      "loss": 0.4928,
+      "step": 740
+    },
+    {
+      "epoch": 0.9617131732641142,
+      "grad_norm": 0.2320844900099328,
+      "learning_rate": 3.775853775853776e-05,
+      "loss": 0.4858,
+      "step": 741
+    },
+    {
+      "epoch": 0.9630110317975341,
+      "grad_norm": 0.25468927720147666,
+      "learning_rate": 3.773448773448774e-05,
+      "loss": 0.4978,
+      "step": 742
+    },
+    {
+      "epoch": 0.9643088903309539,
+      "grad_norm": 0.2686348081196404,
+      "learning_rate": 3.7710437710437716e-05,
+      "loss": 0.4774,
+      "step": 743
+    },
+    {
+      "epoch": 0.9656067488643738,
+      "grad_norm": 0.25462323553264127,
+      "learning_rate": 3.7686387686387684e-05,
+      "loss": 0.4967,
+      "step": 744
+    },
+    {
+      "epoch": 0.9669046073977936,
+      "grad_norm": 0.23826397601414423,
+      "learning_rate": 3.7662337662337665e-05,
+      "loss": 0.4803,
+      "step": 745
+    },
+    {
+      "epoch": 0.9682024659312135,
+      "grad_norm": 0.28828391446426194,
+      "learning_rate": 3.763828763828764e-05,
+      "loss": 0.4853,
+      "step": 746
+    },
+    {
+      "epoch": 0.9695003244646333,
+      "grad_norm": 0.28433469305996517,
+      "learning_rate": 3.761423761423762e-05,
+      "loss": 0.4952,
+      "step": 747
+    },
+    {
+      "epoch": 0.9707981829980532,
+      "grad_norm": 0.23492438563324666,
+      "learning_rate": 3.759018759018759e-05,
+      "loss": 0.4707,
+      "step": 748
+    },
+    {
+      "epoch": 0.9720960415314731,
+      "grad_norm": 0.24099143399264922,
+      "learning_rate": 3.7566137566137564e-05,
+      "loss": 0.5025,
+      "step": 749
+    },
+    {
+      "epoch": 0.9733939000648929,
+      "grad_norm": 0.26521862280904784,
+      "learning_rate": 3.7542087542087546e-05,
+      "loss": 0.4801,
+      "step": 750
+    },
+    {
+      "epoch": 0.9746917585983128,
+      "grad_norm": 0.25301181696938563,
+      "learning_rate": 3.751803751803752e-05,
+      "loss": 0.4812,
+      "step": 751
+    },
+    {
+      "epoch": 0.9759896171317326,
+      "grad_norm": 0.27340749938479236,
+      "learning_rate": 3.7493987493987495e-05,
+      "loss": 0.5218,
+      "step": 752
+    },
+    {
+      "epoch": 0.9772874756651525,
+      "grad_norm": 0.2758375543775843,
+      "learning_rate": 3.746993746993747e-05,
+      "loss": 0.5003,
+      "step": 753
+    },
+    {
+      "epoch": 0.9785853341985724,
+      "grad_norm": 0.2894871095962033,
+      "learning_rate": 3.7445887445887445e-05,
+      "loss": 0.4898,
+      "step": 754
+    },
+    {
+      "epoch": 0.9798831927319922,
+      "grad_norm": 0.24603375955455123,
+      "learning_rate": 3.7421837421837427e-05,
+      "loss": 0.4975,
+      "step": 755
+    },
+    {
+      "epoch": 0.9811810512654121,
+      "grad_norm": 0.3430084532330243,
+      "learning_rate": 3.73977873977874e-05,
+      "loss": 0.4986,
+      "step": 756
+    },
+    {
+      "epoch": 0.9824789097988319,
+      "grad_norm": 0.26260792839174424,
+      "learning_rate": 3.7373737373737376e-05,
+      "loss": 0.4948,
+      "step": 757
+    },
+    {
+      "epoch": 0.9837767683322518,
+      "grad_norm": 0.3417705691604155,
+      "learning_rate": 3.734968734968735e-05,
+      "loss": 0.4892,
+      "step": 758
+    },
+    {
+      "epoch": 0.9850746268656716,
+      "grad_norm": 0.2367132738219532,
+      "learning_rate": 3.7325637325637326e-05,
+      "loss": 0.4837,
+      "step": 759
+    },
+    {
+      "epoch": 0.9863724853990915,
+      "grad_norm": 0.35539307871553294,
+      "learning_rate": 3.730158730158731e-05,
+      "loss": 0.5081,
+      "step": 760
+    },
+    {
+      "epoch": 0.9876703439325114,
+      "grad_norm": 0.22735379818962181,
+      "learning_rate": 3.7277537277537275e-05,
+      "loss": 0.4791,
+      "step": 761
+    },
+    {
+      "epoch": 0.9889682024659312,
+      "grad_norm": 0.30959128828024085,
+      "learning_rate": 3.725348725348726e-05,
+      "loss": 0.4977,
+      "step": 762
+    },
+    {
+      "epoch": 0.9902660609993511,
+      "grad_norm": 0.24407830637335193,
+      "learning_rate": 3.722943722943723e-05,
+      "loss": 0.4889,
+      "step": 763
+    },
+    {
+      "epoch": 0.9915639195327709,
+      "grad_norm": 0.37047588851002045,
+      "learning_rate": 3.720538720538721e-05,
+      "loss": 0.4887,
+      "step": 764
+    },
+    {
+      "epoch": 0.9928617780661908,
+      "grad_norm": 0.3132029888056198,
+      "learning_rate": 3.718133718133718e-05,
+      "loss": 0.4808,
+      "step": 765
+    },
+    {
+      "epoch": 0.9941596365996106,
+      "grad_norm": 0.336725063578875,
+      "learning_rate": 3.7157287157287156e-05,
+      "loss": 0.4944,
+      "step": 766
+    },
+    {
+      "epoch": 0.9954574951330305,
+      "grad_norm": 0.29559283526551605,
+      "learning_rate": 3.713323713323714e-05,
+      "loss": 0.485,
+      "step": 767
+    },
+    {
+      "epoch": 0.9967553536664504,
+      "grad_norm": 0.3181505432986373,
+      "learning_rate": 3.710918710918711e-05,
+      "loss": 0.5109,
+      "step": 768
+    },
+    {
+      "epoch": 0.9980532121998702,
+      "grad_norm": 0.25792486030536,
+      "learning_rate": 3.708513708513709e-05,
+      "loss": 0.4837,
+      "step": 769
+    },
+    {
+      "epoch": 0.9993510707332901,
+      "grad_norm": 0.2662559802060241,
+      "learning_rate": 3.706108706108706e-05,
+      "loss": 0.4768,
+      "step": 770
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2662559802060241,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 0.4749,
+      "step": 771
+    },
+    {
+      "epoch": 1.0012978585334198,
+      "grad_norm": 0.45710604360462226,
+      "learning_rate": 3.701298701298702e-05,
+      "loss": 0.4295,
+      "step": 772
+    },
+    {
+      "epoch": 1.0025957170668398,
+      "grad_norm": 0.3984763478659025,
+      "learning_rate": 3.698893698893699e-05,
+      "loss": 0.4211,
+      "step": 773
+    },
+    {
+      "epoch": 1.0038935756002596,
+      "grad_norm": 0.33604362226653633,
+      "learning_rate": 3.696488696488697e-05,
+      "loss": 0.4472,
+      "step": 774
+    },
+    {
+      "epoch": 1.0051914341336794,
+      "grad_norm": 0.35492193650921217,
+      "learning_rate": 3.694083694083694e-05,
+      "loss": 0.4144,
+      "step": 775
+    },
+    {
+      "epoch": 1.0064892926670992,
+      "grad_norm": 0.34395756575343545,
+      "learning_rate": 3.691678691678692e-05,
+      "loss": 0.4371,
+      "step": 776
+    },
+    {
+      "epoch": 1.0077871512005192,
+      "grad_norm": 0.3027073670465442,
+      "learning_rate": 3.68927368927369e-05,
+      "loss": 0.4418,
+      "step": 777
+    },
+    {
+      "epoch": 1.009085009733939,
+      "grad_norm": 0.33164997403681556,
+      "learning_rate": 3.686868686868687e-05,
+      "loss": 0.4236,
+      "step": 778
+    },
+    {
+      "epoch": 1.0103828682673588,
+      "grad_norm": 0.29836741552546153,
+      "learning_rate": 3.684463684463685e-05,
+      "loss": 0.4291,
+      "step": 779
+    },
+    {
+      "epoch": 1.0116807268007788,
+      "grad_norm": 0.31908724983675907,
+      "learning_rate": 3.682058682058682e-05,
+      "loss": 0.4444,
+      "step": 780
+    },
+    {
+      "epoch": 1.0129785853341986,
+      "grad_norm": 0.3063522863393575,
+      "learning_rate": 3.67965367965368e-05,
+      "loss": 0.4235,
+      "step": 781
+    },
+    {
+      "epoch": 1.0142764438676184,
+      "grad_norm": 0.286969078111983,
+      "learning_rate": 3.677248677248677e-05,
+      "loss": 0.4273,
+      "step": 782
+    },
+    {
+      "epoch": 1.0155743024010382,
+      "grad_norm": 0.30969412667904017,
+      "learning_rate": 3.674843674843675e-05,
+      "loss": 0.4179,
+      "step": 783
+    },
+    {
+      "epoch": 1.0168721609344582,
+      "grad_norm": 0.28846243749065087,
+      "learning_rate": 3.672438672438673e-05,
+      "loss": 0.445,
+      "step": 784
+    },
+    {
+      "epoch": 1.018170019467878,
+      "grad_norm": 0.3052289303187025,
+      "learning_rate": 3.6700336700336704e-05,
+      "loss": 0.4282,
+      "step": 785
+    },
+    {
+      "epoch": 1.0194678780012978,
+      "grad_norm": 0.28776824445687055,
+      "learning_rate": 3.667628667628668e-05,
+      "loss": 0.4361,
+      "step": 786
+    },
+    {
+      "epoch": 1.0207657365347178,
+      "grad_norm": 0.25471106959244577,
+      "learning_rate": 3.665223665223665e-05,
+      "loss": 0.4177,
+      "step": 787
+    },
+    {
+      "epoch": 1.0220635950681376,
+      "grad_norm": 0.2740291049864792,
+      "learning_rate": 3.662818662818663e-05,
+      "loss": 0.4394,
+      "step": 788
+    },
+    {
+      "epoch": 1.0233614536015574,
+      "grad_norm": 0.3017972640732574,
+      "learning_rate": 3.660413660413661e-05,
+      "loss": 0.4386,
+      "step": 789
+    },
+    {
+      "epoch": 1.0246593121349772,
+      "grad_norm": 0.23597430678446688,
+      "learning_rate": 3.6580086580086584e-05,
+      "loss": 0.4499,
+      "step": 790
+    },
+    {
+      "epoch": 1.0259571706683972,
+      "grad_norm": 0.2879673662024183,
+      "learning_rate": 3.655603655603656e-05,
+      "loss": 0.4293,
+      "step": 791
+    },
+    {
+      "epoch": 1.027255029201817,
+      "grad_norm": 0.2700731107191033,
+      "learning_rate": 3.6531986531986534e-05,
+      "loss": 0.4258,
+      "step": 792
+    },
+    {
+      "epoch": 1.0285528877352368,
+      "grad_norm": 0.25488494703667464,
+      "learning_rate": 3.650793650793651e-05,
+      "loss": 0.4074,
+      "step": 793
+    },
+    {
+      "epoch": 1.0298507462686568,
+      "grad_norm": 0.26394634723070287,
+      "learning_rate": 3.648388648388649e-05,
+      "loss": 0.4508,
+      "step": 794
+    },
+    {
+      "epoch": 1.0311486048020766,
+      "grad_norm": 0.27582489513269015,
+      "learning_rate": 3.645983645983646e-05,
+      "loss": 0.4412,
+      "step": 795
+    },
+    {
+      "epoch": 1.0324464633354964,
+      "grad_norm": 0.24726232362927064,
+      "learning_rate": 3.643578643578644e-05,
+      "loss": 0.4168,
+      "step": 796
+    },
+    {
+      "epoch": 1.0337443218689162,
+      "grad_norm": 0.28259272497475907,
+      "learning_rate": 3.6411736411736415e-05,
+      "loss": 0.4154,
+      "step": 797
+    },
+    {
+      "epoch": 1.0350421804023362,
+      "grad_norm": 0.2663149191606831,
+      "learning_rate": 3.638768638768639e-05,
+      "loss": 0.4189,
+      "step": 798
+    },
+    {
+      "epoch": 1.036340038935756,
+      "grad_norm": 0.23665837923986385,
+      "learning_rate": 3.6363636363636364e-05,
+      "loss": 0.4098,
+      "step": 799
+    },
+    {
+      "epoch": 1.0376378974691758,
+      "grad_norm": 0.26291163598038214,
+      "learning_rate": 3.633958633958634e-05,
+      "loss": 0.4311,
+      "step": 800
+    },
+    {
+      "epoch": 1.0389357560025958,
+      "grad_norm": 0.27970192834528934,
+      "learning_rate": 3.631553631553632e-05,
+      "loss": 0.4145,
+      "step": 801
+    },
+    {
+      "epoch": 1.0402336145360156,
+      "grad_norm": 0.2710942408225925,
+      "learning_rate": 3.6291486291486295e-05,
+      "loss": 0.4277,
+      "step": 802
+    },
+    {
+      "epoch": 1.0415314730694354,
+      "grad_norm": 0.2727015101923209,
+      "learning_rate": 3.626743626743627e-05,
+      "loss": 0.4232,
+      "step": 803
+    },
+    {
+      "epoch": 1.0428293316028552,
+      "grad_norm": 0.3128392157246963,
+      "learning_rate": 3.6243386243386245e-05,
+      "loss": 0.4335,
+      "step": 804
+    },
+    {
+      "epoch": 1.0441271901362752,
+      "grad_norm": 0.2349817412715045,
+      "learning_rate": 3.621933621933622e-05,
+      "loss": 0.4253,
+      "step": 805
+    },
+    {
+      "epoch": 1.045425048669695,
+      "grad_norm": 0.28497577605120694,
+      "learning_rate": 3.61952861952862e-05,
+      "loss": 0.4168,
+      "step": 806
+    },
+    {
+      "epoch": 1.0467229072031148,
+      "grad_norm": 0.2587789310002607,
+      "learning_rate": 3.617123617123617e-05,
+      "loss": 0.4217,
+      "step": 807
+    },
+    {
+      "epoch": 1.0480207657365348,
+      "grad_norm": 0.26828544455125314,
+      "learning_rate": 3.6147186147186144e-05,
+      "loss": 0.4074,
+      "step": 808
+    },
+    {
+      "epoch": 1.0493186242699546,
+      "grad_norm": 0.2598287776154231,
+      "learning_rate": 3.6123136123136125e-05,
+      "loss": 0.4184,
+      "step": 809
+    },
+    {
+      "epoch": 1.0506164828033744,
+      "grad_norm": 0.2703287966299121,
+      "learning_rate": 3.60990860990861e-05,
+      "loss": 0.4323,
+      "step": 810
+    },
+    {
+      "epoch": 1.0519143413367944,
+      "grad_norm": 0.318496639053465,
+      "learning_rate": 3.6075036075036075e-05,
+      "loss": 0.4209,
+      "step": 811
+    },
+    {
+      "epoch": 1.0532121998702142,
+      "grad_norm": 0.2697909678170961,
+      "learning_rate": 3.605098605098605e-05,
+      "loss": 0.4244,
+      "step": 812
+    },
+    {
+      "epoch": 1.054510058403634,
+      "grad_norm": 0.2969724315201258,
+      "learning_rate": 3.602693602693603e-05,
+      "loss": 0.4172,
+      "step": 813
+    },
+    {
+      "epoch": 1.0558079169370538,
+      "grad_norm": 0.30303096771353466,
+      "learning_rate": 3.6002886002886006e-05,
+      "loss": 0.404,
+      "step": 814
+    },
+    {
+      "epoch": 1.0571057754704738,
+      "grad_norm": 0.240870306849622,
+      "learning_rate": 3.597883597883598e-05,
+      "loss": 0.4226,
+      "step": 815
+    },
+    {
+      "epoch": 1.0584036340038936,
+      "grad_norm": 0.31791173210740425,
+      "learning_rate": 3.5954785954785956e-05,
+      "loss": 0.411,
+      "step": 816
+    },
+    {
+      "epoch": 1.0597014925373134,
+      "grad_norm": 0.2804167030463435,
+      "learning_rate": 3.593073593073593e-05,
+      "loss": 0.4297,
+      "step": 817
+    },
+    {
+      "epoch": 1.0609993510707332,
+      "grad_norm": 0.24716383688085644,
+      "learning_rate": 3.590668590668591e-05,
+      "loss": 0.434,
+      "step": 818
+    },
+    {
+      "epoch": 1.0622972096041532,
+      "grad_norm": 0.3039180273884032,
+      "learning_rate": 3.588263588263589e-05,
+      "loss": 0.4212,
+      "step": 819
+    },
+    {
+      "epoch": 1.063595068137573,
+      "grad_norm": 0.319488059258111,
+      "learning_rate": 3.5858585858585855e-05,
+      "loss": 0.4068,
+      "step": 820
+    },
+    {
+      "epoch": 1.0648929266709928,
+      "grad_norm": 0.22619788557964504,
+      "learning_rate": 3.5834535834535836e-05,
+      "loss": 0.4311,
+      "step": 821
+    },
+    {
+      "epoch": 1.0661907852044128,
+      "grad_norm": 0.34410773893661634,
+      "learning_rate": 3.581048581048581e-05,
+      "loss": 0.4349,
+      "step": 822
+    },
+    {
+      "epoch": 1.0674886437378326,
+      "grad_norm": 0.3100368353729728,
+      "learning_rate": 3.578643578643579e-05,
+      "loss": 0.4352,
+      "step": 823
+    },
+    {
+      "epoch": 1.0687865022712524,
+      "grad_norm": 0.2901826811884039,
+      "learning_rate": 3.576238576238576e-05,
+      "loss": 0.4157,
+      "step": 824
+    },
+    {
+      "epoch": 1.0700843608046724,
+      "grad_norm": 0.31611344846131356,
+      "learning_rate": 3.5738335738335735e-05,
+      "loss": 0.4693,
+      "step": 825
+    },
+    {
+      "epoch": 1.0713822193380922,
+      "grad_norm": 0.35227684302990314,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 0.4356,
+      "step": 826
+    },
+    {
+      "epoch": 1.072680077871512,
+      "grad_norm": 0.25990758753916315,
+      "learning_rate": 3.569023569023569e-05,
+      "loss": 0.4149,
+      "step": 827
+    },
+    {
+      "epoch": 1.0739779364049318,
+      "grad_norm": 0.33795998379210196,
+      "learning_rate": 3.5666185666185667e-05,
+      "loss": 0.4231,
+      "step": 828
+    },
+    {
+      "epoch": 1.0752757949383518,
+      "grad_norm": 0.260416289520159,
+      "learning_rate": 3.564213564213564e-05,
+      "loss": 0.4439,
+      "step": 829
+    },
+    {
+      "epoch": 1.0765736534717716,
+      "grad_norm": 0.2745403629951124,
+      "learning_rate": 3.5618085618085616e-05,
+      "loss": 0.4363,
+      "step": 830
+    },
+    {
+      "epoch": 1.0778715120051914,
+      "grad_norm": 0.30247544618833483,
+      "learning_rate": 3.55940355940356e-05,
+      "loss": 0.4195,
+      "step": 831
+    },
+    {
+      "epoch": 1.0791693705386114,
+      "grad_norm": 0.32082708667036386,
+      "learning_rate": 3.556998556998557e-05,
+      "loss": 0.4207,
+      "step": 832
+    },
+    {
+      "epoch": 1.0804672290720312,
+      "grad_norm": 0.2897448897920795,
+      "learning_rate": 3.554593554593555e-05,
+      "loss": 0.4256,
+      "step": 833
+    },
+    {
+      "epoch": 1.081765087605451,
+      "grad_norm": 0.2799359981594651,
+      "learning_rate": 3.552188552188552e-05,
+      "loss": 0.4566,
+      "step": 834
+    },
+    {
+      "epoch": 1.0830629461388708,
+      "grad_norm": 0.3069382837366587,
+      "learning_rate": 3.5497835497835503e-05,
+      "loss": 0.4422,
+      "step": 835
+    },
+    {
+      "epoch": 1.0843608046722908,
+      "grad_norm": 0.2432398771659819,
+      "learning_rate": 3.547378547378548e-05,
+      "loss": 0.4058,
+      "step": 836
+    },
+    {
+      "epoch": 1.0856586632057106,
+      "grad_norm": 0.2946884230374921,
+      "learning_rate": 3.5449735449735446e-05,
+      "loss": 0.4212,
+      "step": 837
+    },
+    {
+      "epoch": 1.0869565217391304,
+      "grad_norm": 0.2833055873065403,
+      "learning_rate": 3.542568542568543e-05,
+      "loss": 0.4236,
+      "step": 838
+    },
+    {
+      "epoch": 1.0882543802725504,
+      "grad_norm": 0.24633104520706475,
+      "learning_rate": 3.54016354016354e-05,
+      "loss": 0.4296,
+      "step": 839
+    },
+    {
+      "epoch": 1.0895522388059702,
+      "grad_norm": 0.30305714876494183,
+      "learning_rate": 3.5377585377585384e-05,
+      "loss": 0.4125,
+      "step": 840
+    },
+    {
+      "epoch": 1.09085009733939,
+      "grad_norm": 0.27515539449115956,
+      "learning_rate": 3.535353535353535e-05,
+      "loss": 0.4343,
+      "step": 841
+    },
+    {
+      "epoch": 1.0921479558728098,
+      "grad_norm": 0.2566570554342314,
+      "learning_rate": 3.532948532948533e-05,
+      "loss": 0.4259,
+      "step": 842
+    },
+    {
+      "epoch": 1.0934458144062298,
+      "grad_norm": 0.27580070959828384,
+      "learning_rate": 3.530543530543531e-05,
+      "loss": 0.4429,
+      "step": 843
+    },
+    {
+      "epoch": 1.0947436729396496,
+      "grad_norm": 0.40316368454416424,
+      "learning_rate": 3.528138528138528e-05,
+      "loss": 0.4471,
+      "step": 844
+    },
+    {
+      "epoch": 1.0960415314730694,
+      "grad_norm": 0.2693638650733015,
+      "learning_rate": 3.525733525733526e-05,
+      "loss": 0.4259,
+      "step": 845
+    },
+    {
+      "epoch": 1.0973393900064894,
+      "grad_norm": 0.3044405084823207,
+      "learning_rate": 3.523328523328523e-05,
+      "loss": 0.4187,
+      "step": 846
+    },
+    {
+      "epoch": 1.0986372485399092,
+      "grad_norm": 0.24933029657871267,
+      "learning_rate": 3.520923520923521e-05,
+      "loss": 0.4328,
+      "step": 847
+    },
+    {
+      "epoch": 1.099935107073329,
+      "grad_norm": 0.26906838606368316,
+      "learning_rate": 3.518518518518519e-05,
+      "loss": 0.4368,
+      "step": 848
+    },
+    {
+      "epoch": 1.1012329656067488,
+      "grad_norm": 0.26971278689485273,
+      "learning_rate": 3.5161135161135164e-05,
+      "loss": 0.4297,
+      "step": 849
+    },
+    {
+      "epoch": 1.1025308241401688,
+      "grad_norm": 0.2747922938561738,
+      "learning_rate": 3.513708513708514e-05,
+      "loss": 0.4373,
+      "step": 850
+    },
+    {
+      "epoch": 1.1038286826735886,
+      "grad_norm": 0.2761210141034917,
+      "learning_rate": 3.5113035113035113e-05,
+      "loss": 0.4382,
+      "step": 851
+    },
+    {
+      "epoch": 1.1051265412070084,
+      "grad_norm": 0.2964790930440498,
+      "learning_rate": 3.508898508898509e-05,
+      "loss": 0.4231,
+      "step": 852
+    },
+    {
+      "epoch": 1.1064243997404284,
+      "grad_norm": 0.25620765135533435,
+      "learning_rate": 3.506493506493507e-05,
+      "loss": 0.4408,
+      "step": 853
+    },
+    {
+      "epoch": 1.1077222582738482,
+      "grad_norm": 0.21064152286484897,
+      "learning_rate": 3.504088504088504e-05,
+      "loss": 0.417,
+      "step": 854
+    },
+    {
+      "epoch": 1.109020116807268,
+      "grad_norm": 0.2958430063839789,
+      "learning_rate": 3.501683501683502e-05,
+      "loss": 0.4339,
+      "step": 855
+    },
+    {
+      "epoch": 1.1103179753406878,
+      "grad_norm": 0.22482067437635359,
+      "learning_rate": 3.4992784992784994e-05,
+      "loss": 0.4295,
+      "step": 856
+    },
+    {
+      "epoch": 1.1116158338741078,
+      "grad_norm": 0.2955967782295721,
+      "learning_rate": 3.4968734968734976e-05,
+      "loss": 0.4015,
+      "step": 857
+    },
+    {
+      "epoch": 1.1129136924075276,
+      "grad_norm": 0.2843718668047481,
+      "learning_rate": 3.4944684944684944e-05,
+      "loss": 0.4218,
+      "step": 858
+    },
+    {
+      "epoch": 1.1142115509409474,
+      "grad_norm": 0.2764727160970055,
+      "learning_rate": 3.492063492063492e-05,
+      "loss": 0.4246,
+      "step": 859
+    },
+    {
+      "epoch": 1.1155094094743674,
+      "grad_norm": 0.2624405170772858,
+      "learning_rate": 3.48965848965849e-05,
+      "loss": 0.407,
+      "step": 860
+    },
+    {
+      "epoch": 1.1168072680077872,
+      "grad_norm": 0.2666157900698976,
+      "learning_rate": 3.4872534872534875e-05,
+      "loss": 0.4402,
+      "step": 861
+    },
+    {
+      "epoch": 1.118105126541207,
+      "grad_norm": 1.2301966564667228,
+      "learning_rate": 3.484848484848485e-05,
+      "loss": 0.4452,
+      "step": 862
+    },
+    {
+      "epoch": 1.1194029850746268,
+      "grad_norm": 0.3054038468605886,
+      "learning_rate": 3.4824434824434824e-05,
+      "loss": 0.4354,
+      "step": 863
+    },
+    {
+      "epoch": 1.1207008436080468,
+      "grad_norm": 0.2582864922377692,
+      "learning_rate": 3.48003848003848e-05,
+      "loss": 0.4495,
+      "step": 864
+    },
+    {
+      "epoch": 1.1219987021414666,
+      "grad_norm": 0.2500252146157672,
+      "learning_rate": 3.477633477633478e-05,
+      "loss": 0.4395,
+      "step": 865
+    },
+    {
+      "epoch": 1.1232965606748864,
+      "grad_norm": 0.29631159663934065,
+      "learning_rate": 3.4752284752284755e-05,
+      "loss": 0.4514,
+      "step": 866
+    },
+    {
+      "epoch": 1.1245944192083064,
+      "grad_norm": 0.28152136468600736,
+      "learning_rate": 3.472823472823473e-05,
+      "loss": 0.4397,
+      "step": 867
+    },
+    {
+      "epoch": 1.1258922777417262,
+      "grad_norm": 0.21131772623693806,
+      "learning_rate": 3.4704184704184705e-05,
+      "loss": 0.4285,
+      "step": 868
+    },
+    {
+      "epoch": 1.127190136275146,
+      "grad_norm": 0.25615102365106246,
+      "learning_rate": 3.468013468013468e-05,
+      "loss": 0.4133,
+      "step": 869
+    },
+    {
+      "epoch": 1.128487994808566,
+      "grad_norm": 0.23917269321305726,
+      "learning_rate": 3.465608465608466e-05,
+      "loss": 0.4435,
+      "step": 870
+    },
+    {
+      "epoch": 1.1297858533419858,
+      "grad_norm": 0.27161646725256705,
+      "learning_rate": 3.463203463203463e-05,
+      "loss": 0.4233,
+      "step": 871
+    },
+    {
+      "epoch": 1.1310837118754056,
+      "grad_norm": 0.2412152469942244,
+      "learning_rate": 3.460798460798461e-05,
+      "loss": 0.4461,
+      "step": 872
+    },
+    {
+      "epoch": 1.1323815704088254,
+      "grad_norm": 0.3179450028674149,
+      "learning_rate": 3.4583934583934586e-05,
+      "loss": 0.4119,
+      "step": 873
+    },
+    {
+      "epoch": 1.1336794289422454,
+      "grad_norm": 0.23384786069007624,
+      "learning_rate": 3.455988455988456e-05,
+      "loss": 0.4103,
+      "step": 874
+    },
+    {
+      "epoch": 1.1349772874756652,
+      "grad_norm": 0.2576089230129144,
+      "learning_rate": 3.4535834535834535e-05,
+      "loss": 0.4256,
+      "step": 875
+    },
+    {
+      "epoch": 1.136275146009085,
+      "grad_norm": 0.25522949186712174,
+      "learning_rate": 3.451178451178451e-05,
+      "loss": 0.4306,
+      "step": 876
+    },
+    {
+      "epoch": 1.1375730045425048,
+      "grad_norm": 0.30146067353056377,
+      "learning_rate": 3.448773448773449e-05,
+      "loss": 0.4373,
+      "step": 877
+    },
+    {
+      "epoch": 1.1388708630759248,
+      "grad_norm": 0.2821696115546317,
+      "learning_rate": 3.4463684463684466e-05,
+      "loss": 0.4262,
+      "step": 878
+    },
+    {
+      "epoch": 1.1401687216093446,
+      "grad_norm": 0.23559687187392536,
+      "learning_rate": 3.443963443963444e-05,
+      "loss": 0.4128,
+      "step": 879
+    },
+    {
+      "epoch": 1.1414665801427644,
+      "grad_norm": 0.3096443772818808,
+      "learning_rate": 3.4415584415584416e-05,
+      "loss": 0.424,
+      "step": 880
+    },
+    {
+      "epoch": 1.1427644386761844,
+      "grad_norm": 0.27131628167000316,
+      "learning_rate": 3.439153439153439e-05,
+      "loss": 0.4243,
+      "step": 881
+    },
+    {
+      "epoch": 1.1440622972096042,
+      "grad_norm": 0.3321877999366987,
+      "learning_rate": 3.436748436748437e-05,
+      "loss": 0.3962,
+      "step": 882
+    },
+    {
+      "epoch": 1.145360155743024,
+      "grad_norm": 0.252836911633108,
+      "learning_rate": 3.434343434343435e-05,
+      "loss": 0.4364,
+      "step": 883
+    },
+    {
+      "epoch": 1.146658014276444,
+      "grad_norm": 0.28472809342059574,
+      "learning_rate": 3.431938431938432e-05,
+      "loss": 0.4265,
+      "step": 884
+    },
+    {
+      "epoch": 1.1479558728098638,
+      "grad_norm": 0.25893718977038643,
+      "learning_rate": 3.4295334295334296e-05,
+      "loss": 0.4239,
+      "step": 885
+    },
+    {
+      "epoch": 1.1492537313432836,
+      "grad_norm": 0.29888823830438355,
+      "learning_rate": 3.427128427128427e-05,
+      "loss": 0.4641,
+      "step": 886
+    },
+    {
+      "epoch": 1.1505515898767034,
+      "grad_norm": 0.2336271423650757,
+      "learning_rate": 3.424723424723425e-05,
+      "loss": 0.4363,
+      "step": 887
+    },
+    {
+      "epoch": 1.1518494484101234,
+      "grad_norm": 0.28355868807379,
+      "learning_rate": 3.422318422318422e-05,
+      "loss": 0.4446,
+      "step": 888
+    },
+    {
+      "epoch": 1.1531473069435432,
+      "grad_norm": 0.25482834362533163,
+      "learning_rate": 3.41991341991342e-05,
+      "loss": 0.432,
+      "step": 889
+    },
+    {
+      "epoch": 1.154445165476963,
+      "grad_norm": 0.257058818612092,
+      "learning_rate": 3.417508417508418e-05,
+      "loss": 0.4238,
+      "step": 890
+    },
+    {
+      "epoch": 1.1557430240103828,
+      "grad_norm": 0.2964878759739716,
+      "learning_rate": 3.415103415103415e-05,
+      "loss": 0.4341,
+      "step": 891
+    },
+    {
+      "epoch": 1.1570408825438028,
+      "grad_norm": 0.24581976687613294,
+      "learning_rate": 3.412698412698413e-05,
+      "loss": 0.445,
+      "step": 892
+    },
+    {
+      "epoch": 1.1583387410772226,
+      "grad_norm": 0.2672951398900844,
+      "learning_rate": 3.41029341029341e-05,
+      "loss": 0.4146,
+      "step": 893
+    },
+    {
+      "epoch": 1.1596365996106424,
+      "grad_norm": 0.29744313702383335,
+      "learning_rate": 3.407888407888408e-05,
+      "loss": 0.4291,
+      "step": 894
+    },
+    {
+      "epoch": 1.1609344581440624,
+      "grad_norm": 0.2594025370257348,
+      "learning_rate": 3.405483405483406e-05,
+      "loss": 0.4393,
+      "step": 895
+    },
+    {
+      "epoch": 1.1622323166774822,
+      "grad_norm": 0.27322558327059043,
+      "learning_rate": 3.403078403078403e-05,
+      "loss": 0.413,
+      "step": 896
+    },
+    {
+      "epoch": 1.163530175210902,
+      "grad_norm": 0.27895427053368943,
+      "learning_rate": 3.400673400673401e-05,
+      "loss": 0.4282,
+      "step": 897
+    },
+    {
+      "epoch": 1.164828033744322,
+      "grad_norm": 0.3278143440045291,
+      "learning_rate": 3.398268398268398e-05,
+      "loss": 0.4503,
+      "step": 898
+    },
+    {
+      "epoch": 1.1661258922777418,
+      "grad_norm": 0.2878645741851875,
+      "learning_rate": 3.3958633958633964e-05,
+      "loss": 0.418,
+      "step": 899
+    },
+    {
+      "epoch": 1.1674237508111616,
+      "grad_norm": 0.279091054078343,
+      "learning_rate": 3.393458393458394e-05,
+      "loss": 0.4351,
+      "step": 900
+    },
+    {
+      "epoch": 1.1687216093445814,
+      "grad_norm": 0.300972554323965,
+      "learning_rate": 3.391053391053391e-05,
+      "loss": 0.4177,
+      "step": 901
+    },
+    {
+      "epoch": 1.1700194678780014,
+      "grad_norm": 0.2912604255538886,
+      "learning_rate": 3.388648388648389e-05,
+      "loss": 0.4239,
+      "step": 902
+    },
+    {
+      "epoch": 1.1713173264114212,
+      "grad_norm": 0.28729371845984225,
+      "learning_rate": 3.386243386243386e-05,
+      "loss": 0.4498,
+      "step": 903
+    },
+    {
+      "epoch": 1.172615184944841,
+      "grad_norm": 0.2983707424965568,
+      "learning_rate": 3.3838383838383844e-05,
+      "loss": 0.4093,
+      "step": 904
+    },
+    {
+      "epoch": 1.1739130434782608,
+      "grad_norm": 0.2396146655592429,
+      "learning_rate": 3.381433381433381e-05,
+      "loss": 0.4134,
+      "step": 905
+    },
+    {
+      "epoch": 1.1752109020116808,
+      "grad_norm": 0.25743340902304185,
+      "learning_rate": 3.3790283790283794e-05,
+      "loss": 0.4024,
+      "step": 906
+    },
+    {
+      "epoch": 1.1765087605451006,
+      "grad_norm": 0.27027531302973373,
+      "learning_rate": 3.376623376623377e-05,
+      "loss": 0.4518,
+      "step": 907
+    },
+    {
+      "epoch": 1.1778066190785204,
+      "grad_norm": 0.25280300819232365,
+      "learning_rate": 3.3742183742183743e-05,
+      "loss": 0.4185,
+      "step": 908
+    },
+    {
+      "epoch": 1.1791044776119404,
+      "grad_norm": 0.22682160703006224,
+      "learning_rate": 3.371813371813372e-05,
+      "loss": 0.4174,
+      "step": 909
+    },
+    {
+      "epoch": 1.1804023361453602,
+      "grad_norm": 0.23204503630025836,
+      "learning_rate": 3.369408369408369e-05,
+      "loss": 0.4177,
+      "step": 910
+    },
+    {
+      "epoch": 1.18170019467878,
+      "grad_norm": 0.25880635574030375,
+      "learning_rate": 3.3670033670033675e-05,
+      "loss": 0.4179,
+      "step": 911
+    },
+    {
+      "epoch": 1.1829980532122,
+      "grad_norm": 0.2522597708371833,
+      "learning_rate": 3.364598364598365e-05,
+      "loss": 0.4283,
+      "step": 912
+    },
+    {
+      "epoch": 1.1842959117456198,
+      "grad_norm": 0.2883869624140782,
+      "learning_rate": 3.3621933621933624e-05,
+      "loss": 0.4372,
+      "step": 913
+    },
+    {
+      "epoch": 1.1855937702790396,
+      "grad_norm": 0.25106486957221746,
+      "learning_rate": 3.35978835978836e-05,
+      "loss": 0.4281,
+      "step": 914
+    },
+    {
+      "epoch": 1.1868916288124594,
+      "grad_norm": 0.292526125260076,
+      "learning_rate": 3.3573833573833574e-05,
+      "loss": 0.4365,
+      "step": 915
+    },
+    {
+      "epoch": 1.1881894873458794,
+      "grad_norm": 0.2676690874911841,
+      "learning_rate": 3.3549783549783555e-05,
+      "loss": 0.4326,
+      "step": 916
+    },
+    {
+      "epoch": 1.1894873458792992,
+      "grad_norm": 0.26481922535161423,
+      "learning_rate": 3.352573352573353e-05,
+      "loss": 0.4274,
+      "step": 917
+    },
+    {
+      "epoch": 1.190785204412719,
+      "grad_norm": 0.2798203995195467,
+      "learning_rate": 3.35016835016835e-05,
+      "loss": 0.4368,
+      "step": 918
+    },
+    {
+      "epoch": 1.1920830629461387,
+      "grad_norm": 0.2861488916957064,
+      "learning_rate": 3.347763347763348e-05,
+      "loss": 0.4137,
+      "step": 919
+    },
+    {
+      "epoch": 1.1933809214795588,
+      "grad_norm": 0.29631412123985384,
+      "learning_rate": 3.3453583453583454e-05,
+      "loss": 0.4569,
+      "step": 920
+    },
+    {
+      "epoch": 1.1946787800129786,
+      "grad_norm": 0.2342734368682864,
+      "learning_rate": 3.3429533429533436e-05,
+      "loss": 0.4161,
+      "step": 921
+    },
+    {
+      "epoch": 1.1959766385463984,
+      "grad_norm": 0.2701444926397203,
+      "learning_rate": 3.3405483405483404e-05,
+      "loss": 0.4436,
+      "step": 922
+    },
+    {
+      "epoch": 1.1972744970798184,
+      "grad_norm": 0.24533441345926324,
+      "learning_rate": 3.3381433381433385e-05,
+      "loss": 0.4169,
+      "step": 923
+    },
+    {
+      "epoch": 1.1985723556132382,
+      "grad_norm": 0.2540671392711727,
+      "learning_rate": 3.335738335738336e-05,
+      "loss": 0.4157,
+      "step": 924
+    },
+    {
+      "epoch": 1.199870214146658,
+      "grad_norm": 0.24799812272450378,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.4185,
+      "step": 925
+    },
+    {
+      "epoch": 1.201168072680078,
+      "grad_norm": 0.22598144682861446,
+      "learning_rate": 3.330928330928331e-05,
+      "loss": 0.4316,
+      "step": 926
+    },
+    {
+      "epoch": 1.2024659312134978,
+      "grad_norm": 0.2448180172195478,
+      "learning_rate": 3.3285233285233284e-05,
+      "loss": 0.4224,
+      "step": 927
+    },
+    {
+      "epoch": 1.2037637897469176,
+      "grad_norm": 0.2765575264990623,
+      "learning_rate": 3.3261183261183266e-05,
+      "loss": 0.4152,
+      "step": 928
+    },
+    {
+      "epoch": 1.2050616482803373,
+      "grad_norm": 0.2804413742079038,
+      "learning_rate": 3.323713323713324e-05,
+      "loss": 0.4268,
+      "step": 929
+    },
+    {
+      "epoch": 1.2063595068137574,
+      "grad_norm": 0.241273559139525,
+      "learning_rate": 3.3213083213083216e-05,
+      "loss": 0.44,
+      "step": 930
+    },
+    {
+      "epoch": 1.2076573653471772,
+      "grad_norm": 0.26424527656796193,
+      "learning_rate": 3.318903318903319e-05,
+      "loss": 0.423,
+      "step": 931
+    },
+    {
+      "epoch": 1.208955223880597,
+      "grad_norm": 0.2618836109205494,
+      "learning_rate": 3.3164983164983165e-05,
+      "loss": 0.4263,
+      "step": 932
+    },
+    {
+      "epoch": 1.210253082414017,
+      "grad_norm": 0.232277628692469,
+      "learning_rate": 3.314093314093315e-05,
+      "loss": 0.4285,
+      "step": 933
+    },
+    {
+      "epoch": 1.2115509409474368,
+      "grad_norm": 0.22920838397957313,
+      "learning_rate": 3.311688311688312e-05,
+      "loss": 0.4222,
+      "step": 934
+    },
+    {
+      "epoch": 1.2128487994808566,
+      "grad_norm": 0.22768841564535697,
+      "learning_rate": 3.309283309283309e-05,
+      "loss": 0.4343,
+      "step": 935
+    },
+    {
+      "epoch": 1.2141466580142763,
+      "grad_norm": 0.2503532795435805,
+      "learning_rate": 3.306878306878307e-05,
+      "loss": 0.439,
+      "step": 936
+    },
+    {
+      "epoch": 1.2154445165476964,
+      "grad_norm": 0.22461109640165802,
+      "learning_rate": 3.3044733044733046e-05,
+      "loss": 0.4188,
+      "step": 937
+    },
+    {
+      "epoch": 1.2167423750811162,
+      "grad_norm": 0.209011116297864,
+      "learning_rate": 3.302068302068303e-05,
+      "loss": 0.4095,
+      "step": 938
+    },
+    {
+      "epoch": 1.218040233614536,
+      "grad_norm": 0.2266971578892572,
+      "learning_rate": 3.2996632996632995e-05,
+      "loss": 0.4354,
+      "step": 939
+    },
+    {
+      "epoch": 1.219338092147956,
+      "grad_norm": 0.22300168989786548,
+      "learning_rate": 3.297258297258297e-05,
+      "loss": 0.4225,
+      "step": 940
+    },
+    {
+      "epoch": 1.2206359506813758,
+      "grad_norm": 0.24047207294507855,
+      "learning_rate": 3.294853294853295e-05,
+      "loss": 0.4485,
+      "step": 941
+    },
+    {
+      "epoch": 1.2219338092147956,
+      "grad_norm": 0.26188839036093997,
+      "learning_rate": 3.2924482924482926e-05,
+      "loss": 0.4338,
+      "step": 942
+    },
+    {
+      "epoch": 1.2232316677482156,
+      "grad_norm": 0.2235845694258825,
+      "learning_rate": 3.29004329004329e-05,
+      "loss": 0.4242,
+      "step": 943
+    },
+    {
+      "epoch": 1.2245295262816354,
+      "grad_norm": 0.21723162921446287,
+      "learning_rate": 3.2876382876382876e-05,
+      "loss": 0.4241,
+      "step": 944
+    },
+    {
+      "epoch": 1.2258273848150552,
+      "grad_norm": 0.25526775092171644,
+      "learning_rate": 3.285233285233286e-05,
+      "loss": 0.4253,
+      "step": 945
+    },
+    {
+      "epoch": 1.227125243348475,
+      "grad_norm": 0.20573746450142508,
+      "learning_rate": 3.282828282828283e-05,
+      "loss": 0.4479,
+      "step": 946
+    },
+    {
+      "epoch": 1.228423101881895,
+      "grad_norm": 0.2510082428750361,
+      "learning_rate": 3.280423280423281e-05,
+      "loss": 0.4347,
+      "step": 947
+    },
+    {
+      "epoch": 1.2297209604153148,
+      "grad_norm": 0.2669964257731318,
+      "learning_rate": 3.278018278018278e-05,
+      "loss": 0.4291,
+      "step": 948
+    },
+    {
+      "epoch": 1.2310188189487346,
+      "grad_norm": 0.23768286255343224,
+      "learning_rate": 3.275613275613276e-05,
+      "loss": 0.4348,
+      "step": 949
+    },
+    {
+      "epoch": 1.2323166774821543,
+      "grad_norm": 0.253301068721141,
+      "learning_rate": 3.273208273208274e-05,
+      "loss": 0.4165,
+      "step": 950
+    },
+    {
+      "epoch": 1.2336145360155744,
+      "grad_norm": 0.24211175208943894,
+      "learning_rate": 3.270803270803271e-05,
+      "loss": 0.4525,
+      "step": 951
+    },
+    {
+      "epoch": 1.2349123945489942,
+      "grad_norm": 0.20694813799256812,
+      "learning_rate": 3.268398268398268e-05,
+      "loss": 0.4353,
+      "step": 952
+    },
+    {
+      "epoch": 1.236210253082414,
+      "grad_norm": 0.24139016385768045,
+      "learning_rate": 3.265993265993266e-05,
+      "loss": 0.4255,
+      "step": 953
+    },
+    {
+      "epoch": 1.237508111615834,
+      "grad_norm": 0.23298523425827472,
+      "learning_rate": 3.263588263588264e-05,
+      "loss": 0.4453,
+      "step": 954
+    },
+    {
+      "epoch": 1.2388059701492538,
+      "grad_norm": 0.22089226803029272,
+      "learning_rate": 3.261183261183262e-05,
+      "loss": 0.4142,
+      "step": 955
+    },
+    {
+      "epoch": 1.2401038286826735,
+      "grad_norm": 0.21279087400923866,
+      "learning_rate": 3.258778258778259e-05,
+      "loss": 0.3967,
+      "step": 956
+    },
+    {
+      "epoch": 1.2414016872160936,
+      "grad_norm": 0.21365588686190673,
+      "learning_rate": 3.256373256373256e-05,
+      "loss": 0.4265,
+      "step": 957
+    },
+    {
+      "epoch": 1.2426995457495134,
+      "grad_norm": 0.2297103141691502,
+      "learning_rate": 3.253968253968254e-05,
+      "loss": 0.4247,
+      "step": 958
+    },
+    {
+      "epoch": 1.2439974042829332,
+      "grad_norm": 0.21237733020443594,
+      "learning_rate": 3.251563251563252e-05,
+      "loss": 0.4187,
+      "step": 959
+    },
+    {
+      "epoch": 1.245295262816353,
+      "grad_norm": 0.25205287449171737,
+      "learning_rate": 3.249158249158249e-05,
+      "loss": 0.4195,
+      "step": 960
+    },
+    {
+      "epoch": 1.246593121349773,
+      "grad_norm": 0.2275565335826296,
+      "learning_rate": 3.246753246753247e-05,
+      "loss": 0.4222,
+      "step": 961
+    },
+    {
+      "epoch": 1.2478909798831928,
+      "grad_norm": 0.24497964692242122,
+      "learning_rate": 3.244348244348244e-05,
+      "loss": 0.4282,
+      "step": 962
+    },
+    {
+      "epoch": 1.2491888384166125,
+      "grad_norm": 0.2154307123634933,
+      "learning_rate": 3.2419432419432424e-05,
+      "loss": 0.4354,
+      "step": 963
+    },
+    {
+      "epoch": 1.2504866969500323,
+      "grad_norm": 0.21744389583650917,
+      "learning_rate": 3.23953823953824e-05,
+      "loss": 0.4138,
+      "step": 964
+    },
+    {
+      "epoch": 1.2517845554834524,
+      "grad_norm": 0.2430698493626977,
+      "learning_rate": 3.237133237133237e-05,
+      "loss": 0.4342,
+      "step": 965
+    },
+    {
+      "epoch": 1.2530824140168721,
+      "grad_norm": 0.2426482535744958,
+      "learning_rate": 3.234728234728235e-05,
+      "loss": 0.4253,
+      "step": 966
+    },
+    {
+      "epoch": 1.254380272550292,
+      "grad_norm": 0.22495203456758703,
+      "learning_rate": 3.232323232323233e-05,
+      "loss": 0.4308,
+      "step": 967
+    },
+    {
+      "epoch": 1.255678131083712,
+      "grad_norm": 0.24355897774937213,
+      "learning_rate": 3.2299182299182304e-05,
+      "loss": 0.4284,
+      "step": 968
+    },
+    {
+      "epoch": 1.2569759896171318,
+      "grad_norm": 0.2417579603003613,
+      "learning_rate": 3.227513227513227e-05,
+      "loss": 0.4236,
+      "step": 969
+    },
+    {
+      "epoch": 1.2582738481505515,
+      "grad_norm": 0.21950714489690643,
+      "learning_rate": 3.2251082251082254e-05,
+      "loss": 0.4319,
+      "step": 970
+    },
+    {
+      "epoch": 1.2595717066839716,
+      "grad_norm": 0.248967863409087,
+      "learning_rate": 3.222703222703223e-05,
+      "loss": 0.414,
+      "step": 971
+    },
+    {
+      "epoch": 1.2608695652173914,
+      "grad_norm": 0.2320971622059916,
+      "learning_rate": 3.220298220298221e-05,
+      "loss": 0.4396,
+      "step": 972
+    },
+    {
+      "epoch": 1.2621674237508111,
+      "grad_norm": 0.23309901348515835,
+      "learning_rate": 3.217893217893218e-05,
+      "loss": 0.423,
+      "step": 973
+    },
+    {
+      "epoch": 1.263465282284231,
+      "grad_norm": 0.2227973619365033,
+      "learning_rate": 3.215488215488215e-05,
+      "loss": 0.4256,
+      "step": 974
+    },
+    {
+      "epoch": 1.264763140817651,
+      "grad_norm": 0.245760384681702,
+      "learning_rate": 3.2130832130832135e-05,
+      "loss": 0.4326,
+      "step": 975
+    },
+    {
+      "epoch": 1.2660609993510707,
+      "grad_norm": 0.2789130551480554,
+      "learning_rate": 3.210678210678211e-05,
+      "loss": 0.4158,
+      "step": 976
+    },
+    {
+      "epoch": 1.2673588578844905,
+      "grad_norm": 0.24466337361794802,
+      "learning_rate": 3.2082732082732084e-05,
+      "loss": 0.4132,
+      "step": 977
+    },
+    {
+      "epoch": 1.2686567164179103,
+      "grad_norm": 0.26994667632692604,
+      "learning_rate": 3.205868205868206e-05,
+      "loss": 0.4193,
+      "step": 978
+    },
+    {
+      "epoch": 1.2699545749513304,
+      "grad_norm": 0.27229427206268936,
+      "learning_rate": 3.2034632034632034e-05,
+      "loss": 0.4438,
+      "step": 979
+    },
+    {
+      "epoch": 1.2712524334847501,
+      "grad_norm": 0.2562325137645907,
+      "learning_rate": 3.2010582010582015e-05,
+      "loss": 0.4356,
+      "step": 980
+    },
+    {
+      "epoch": 1.27255029201817,
+      "grad_norm": 0.25248121832020193,
+      "learning_rate": 3.198653198653199e-05,
+      "loss": 0.4143,
+      "step": 981
+    },
+    {
+      "epoch": 1.27384815055159,
+      "grad_norm": 0.2656639346583922,
+      "learning_rate": 3.1962481962481965e-05,
+      "loss": 0.4116,
+      "step": 982
+    },
+    {
+      "epoch": 1.2751460090850097,
+      "grad_norm": 0.22210045965124164,
+      "learning_rate": 3.193843193843194e-05,
+      "loss": 0.4194,
+      "step": 983
+    },
+    {
+      "epoch": 1.2764438676184295,
+      "grad_norm": 0.29306826782606415,
+      "learning_rate": 3.1914381914381914e-05,
+      "loss": 0.4148,
+      "step": 984
+    },
+    {
+      "epoch": 1.2777417261518496,
+      "grad_norm": 0.24612787413957143,
+      "learning_rate": 3.1890331890331896e-05,
+      "loss": 0.419,
+      "step": 985
+    },
+    {
+      "epoch": 1.2790395846852693,
+      "grad_norm": 0.24144328983707405,
+      "learning_rate": 3.1866281866281864e-05,
+      "loss": 0.4412,
+      "step": 986
+    },
+    {
+      "epoch": 1.2803374432186891,
+      "grad_norm": 0.24433831385926233,
+      "learning_rate": 3.1842231842231846e-05,
+      "loss": 0.4237,
+      "step": 987
+    },
+    {
+      "epoch": 1.2816353017521092,
+      "grad_norm": 0.29652784983616687,
+      "learning_rate": 3.181818181818182e-05,
+      "loss": 0.4052,
+      "step": 988
+    },
+    {
+      "epoch": 1.282933160285529,
+      "grad_norm": 0.22517076903481237,
+      "learning_rate": 3.1794131794131795e-05,
+      "loss": 0.437,
+      "step": 989
+    },
+    {
+      "epoch": 1.2842310188189487,
+      "grad_norm": 0.2695004176632199,
+      "learning_rate": 3.177008177008177e-05,
+      "loss": 0.4051,
+      "step": 990
+    },
+    {
+      "epoch": 1.2855288773523685,
+      "grad_norm": 0.2831795101586642,
+      "learning_rate": 3.1746031746031745e-05,
+      "loss": 0.4142,
+      "step": 991
+    },
+    {
+      "epoch": 1.2868267358857883,
+      "grad_norm": 0.2305605000016992,
+      "learning_rate": 3.1721981721981726e-05,
+      "loss": 0.4318,
+      "step": 992
+    },
+    {
+      "epoch": 1.2881245944192083,
+      "grad_norm": 0.27421133904277795,
+      "learning_rate": 3.16979316979317e-05,
+      "loss": 0.4303,
+      "step": 993
+    },
+    {
+      "epoch": 1.2894224529526281,
+      "grad_norm": 0.28015539417195207,
+      "learning_rate": 3.1673881673881676e-05,
+      "loss": 0.4332,
+      "step": 994
+    },
+    {
+      "epoch": 1.290720311486048,
+      "grad_norm": 0.2456153256375182,
+      "learning_rate": 3.164983164983165e-05,
+      "loss": 0.4377,
+      "step": 995
+    },
+    {
+      "epoch": 1.292018170019468,
+      "grad_norm": 0.2875091277813538,
+      "learning_rate": 3.1625781625781625e-05,
+      "loss": 0.4306,
+      "step": 996
+    },
+    {
+      "epoch": 1.2933160285528877,
+      "grad_norm": 0.24534524749679693,
+      "learning_rate": 3.160173160173161e-05,
+      "loss": 0.4389,
+      "step": 997
+    },
+    {
+      "epoch": 1.2946138870863075,
+      "grad_norm": 0.2604745981282834,
+      "learning_rate": 3.1577681577681575e-05,
+      "loss": 0.4221,
+      "step": 998
+    },
+    {
+      "epoch": 1.2959117456197276,
+      "grad_norm": 0.2519193028255322,
+      "learning_rate": 3.1553631553631556e-05,
+      "loss": 0.4613,
+      "step": 999
+    },
+    {
+      "epoch": 1.2972096041531473,
+      "grad_norm": 0.28305211286597437,
+      "learning_rate": 3.152958152958153e-05,
+      "loss": 0.4364,
+      "step": 1000
+    },
+    {
+      "epoch": 1.2985074626865671,
+      "grad_norm": 0.31302239991262554,
+      "learning_rate": 3.1505531505531506e-05,
+      "loss": 0.4315,
+      "step": 1001
+    },
+    {
+      "epoch": 1.2998053212199872,
+      "grad_norm": 0.21929034373943385,
+      "learning_rate": 3.148148148148148e-05,
+      "loss": 0.4389,
+      "step": 1002
+    },
+    {
+      "epoch": 1.301103179753407,
+      "grad_norm": 0.33967578917657276,
+      "learning_rate": 3.1457431457431456e-05,
+      "loss": 0.4245,
+      "step": 1003
+    },
+    {
+      "epoch": 1.3024010382868267,
+      "grad_norm": 0.26773599652804664,
+      "learning_rate": 3.143338143338144e-05,
+      "loss": 0.4322,
+      "step": 1004
+    },
+    {
+      "epoch": 1.3036988968202465,
+      "grad_norm": 0.3377829113222901,
+      "learning_rate": 3.140933140933141e-05,
+      "loss": 0.4156,
+      "step": 1005
+    },
+    {
+      "epoch": 1.3049967553536663,
+      "grad_norm": 0.28733563773070486,
+      "learning_rate": 3.1385281385281387e-05,
+      "loss": 0.4153,
+      "step": 1006
+    },
+    {
+      "epoch": 1.3062946138870863,
+      "grad_norm": 0.29064269118004,
+      "learning_rate": 3.136123136123136e-05,
+      "loss": 0.4248,
+      "step": 1007
+    },
+    {
+      "epoch": 1.3075924724205061,
+      "grad_norm": 0.3578706488624722,
+      "learning_rate": 3.1337181337181336e-05,
+      "loss": 0.4103,
+      "step": 1008
+    },
+    {
+      "epoch": 1.308890330953926,
+      "grad_norm": 0.2506804986945498,
+      "learning_rate": 3.131313131313132e-05,
+      "loss": 0.4455,
+      "step": 1009
+    },
+    {
+      "epoch": 1.310188189487346,
+      "grad_norm": 0.26163932223113945,
+      "learning_rate": 3.128908128908129e-05,
+      "loss": 0.418,
+      "step": 1010
+    },
+    {
+      "epoch": 1.3114860480207657,
+      "grad_norm": 0.2968145670132288,
+      "learning_rate": 3.126503126503126e-05,
+      "loss": 0.4419,
+      "step": 1011
+    },
+    {
+      "epoch": 1.3127839065541855,
+      "grad_norm": 0.29985902672925774,
+      "learning_rate": 3.124098124098124e-05,
+      "loss": 0.4397,
+      "step": 1012
+    },
+    {
+      "epoch": 1.3140817650876055,
+      "grad_norm": 0.2693824680991638,
+      "learning_rate": 3.121693121693122e-05,
+      "loss": 0.4477,
+      "step": 1013
+    },
+    {
+      "epoch": 1.3153796236210253,
+      "grad_norm": 0.2703405621428287,
+      "learning_rate": 3.11928811928812e-05,
+      "loss": 0.4325,
+      "step": 1014
+    },
+    {
+      "epoch": 1.3166774821544451,
+      "grad_norm": 0.2721038339798775,
+      "learning_rate": 3.1168831168831166e-05,
+      "loss": 0.4373,
+      "step": 1015
+    },
+    {
+      "epoch": 1.3179753406878651,
+      "grad_norm": 0.26849320227585655,
+      "learning_rate": 3.114478114478115e-05,
+      "loss": 0.4246,
+      "step": 1016
+    },
+    {
+      "epoch": 1.319273199221285,
+      "grad_norm": 0.28294666170474586,
+      "learning_rate": 3.112073112073112e-05,
+      "loss": 0.4405,
+      "step": 1017
+    },
+    {
+      "epoch": 1.3205710577547047,
+      "grad_norm": 0.7257885246786743,
+      "learning_rate": 3.10966810966811e-05,
+      "loss": 0.419,
+      "step": 1018
+    },
+    {
+      "epoch": 1.3218689162881245,
+      "grad_norm": 0.26474834284364107,
+      "learning_rate": 3.107263107263107e-05,
+      "loss": 0.4561,
+      "step": 1019
+    },
+    {
+      "epoch": 1.3231667748215443,
+      "grad_norm": 0.2836196696373746,
+      "learning_rate": 3.104858104858105e-05,
+      "loss": 0.4187,
+      "step": 1020
+    },
+    {
+      "epoch": 1.3244646333549643,
+      "grad_norm": 0.2978444678442113,
+      "learning_rate": 3.102453102453103e-05,
+      "loss": 0.4245,
+      "step": 1021
+    },
+    {
+      "epoch": 1.3257624918883841,
+      "grad_norm": 0.27039863643097406,
+      "learning_rate": 3.1000481000481e-05,
+      "loss": 0.4492,
+      "step": 1022
+    },
+    {
+      "epoch": 1.327060350421804,
+      "grad_norm": 0.28245910629768817,
+      "learning_rate": 3.097643097643098e-05,
+      "loss": 0.4205,
+      "step": 1023
+    },
+    {
+      "epoch": 1.328358208955224,
+      "grad_norm": 0.235926922135542,
+      "learning_rate": 3.095238095238095e-05,
+      "loss": 0.4176,
+      "step": 1024
+    },
+    {
+      "epoch": 1.3296560674886437,
+      "grad_norm": 0.24582624312732296,
+      "learning_rate": 3.092833092833093e-05,
+      "loss": 0.4409,
+      "step": 1025
+    },
+    {
+      "epoch": 1.3309539260220635,
+      "grad_norm": 0.25618143979663144,
+      "learning_rate": 3.090428090428091e-05,
+      "loss": 0.4411,
+      "step": 1026
+    },
+    {
+      "epoch": 1.3322517845554835,
+      "grad_norm": 0.3044844628215498,
+      "learning_rate": 3.0880230880230884e-05,
+      "loss": 0.4628,
+      "step": 1027
+    },
+    {
+      "epoch": 1.3335496430889033,
+      "grad_norm": 0.25846935424140755,
+      "learning_rate": 3.085618085618085e-05,
+      "loss": 0.431,
+      "step": 1028
+    },
+    {
+      "epoch": 1.3348475016223231,
+      "grad_norm": 0.2299843153442688,
+      "learning_rate": 3.0832130832130834e-05,
+      "loss": 0.4091,
+      "step": 1029
+    },
+    {
+      "epoch": 1.3361453601557431,
+      "grad_norm": 0.2461059374323843,
+      "learning_rate": 3.080808080808081e-05,
+      "loss": 0.4095,
+      "step": 1030
+    },
+    {
+      "epoch": 1.337443218689163,
+      "grad_norm": 0.2591709406665044,
+      "learning_rate": 3.078403078403079e-05,
+      "loss": 0.4311,
+      "step": 1031
+    },
+    {
+      "epoch": 1.3387410772225827,
+      "grad_norm": 0.22916233734032224,
+      "learning_rate": 3.075998075998076e-05,
+      "loss": 0.4485,
+      "step": 1032
+    },
+    {
+      "epoch": 1.3400389357560025,
+      "grad_norm": 0.2600469770283129,
+      "learning_rate": 3.073593073593073e-05,
+      "loss": 0.4356,
+      "step": 1033
+    },
+    {
+      "epoch": 1.3413367942894223,
+      "grad_norm": 0.24066487258683386,
+      "learning_rate": 3.0711880711880714e-05,
+      "loss": 0.4185,
+      "step": 1034
+    },
+    {
+      "epoch": 1.3426346528228423,
+      "grad_norm": 0.27145997155071,
+      "learning_rate": 3.068783068783069e-05,
+      "loss": 0.4145,
+      "step": 1035
+    },
+    {
+      "epoch": 1.3439325113562621,
+      "grad_norm": 0.22303833825329822,
+      "learning_rate": 3.0663780663780664e-05,
+      "loss": 0.4544,
+      "step": 1036
+    },
+    {
+      "epoch": 1.345230369889682,
+      "grad_norm": 0.2631308244918525,
+      "learning_rate": 3.063973063973064e-05,
+      "loss": 0.4171,
+      "step": 1037
+    },
+    {
+      "epoch": 1.346528228423102,
+      "grad_norm": 0.24013765031743725,
+      "learning_rate": 3.061568061568062e-05,
+      "loss": 0.4162,
+      "step": 1038
+    },
+    {
+      "epoch": 1.3478260869565217,
+      "grad_norm": 0.2563444998106366,
+      "learning_rate": 3.0591630591630595e-05,
+      "loss": 0.4228,
+      "step": 1039
+    },
+    {
+      "epoch": 1.3491239454899415,
+      "grad_norm": 0.26755087682195894,
+      "learning_rate": 3.056758056758057e-05,
+      "loss": 0.4393,
+      "step": 1040
+    },
+    {
+      "epoch": 1.3504218040233615,
+      "grad_norm": 0.2651622032346258,
+      "learning_rate": 3.0543530543530544e-05,
+      "loss": 0.4149,
+      "step": 1041
+    },
+    {
+      "epoch": 1.3517196625567813,
+      "grad_norm": 0.25334754508930496,
+      "learning_rate": 3.051948051948052e-05,
+      "loss": 0.3989,
+      "step": 1042
+    },
+    {
+      "epoch": 1.3530175210902011,
+      "grad_norm": 0.23721003539310276,
+      "learning_rate": 3.04954304954305e-05,
+      "loss": 0.4221,
+      "step": 1043
+    },
+    {
+      "epoch": 1.3543153796236211,
+      "grad_norm": 0.23037112505555338,
+      "learning_rate": 3.0471380471380472e-05,
+      "loss": 0.4182,
+      "step": 1044
+    },
+    {
+      "epoch": 1.355613238157041,
+      "grad_norm": 0.2436111206549788,
+      "learning_rate": 3.0447330447330447e-05,
+      "loss": 0.4109,
+      "step": 1045
+    },
+    {
+      "epoch": 1.3569110966904607,
+      "grad_norm": 0.24241659874410296,
+      "learning_rate": 3.0423280423280425e-05,
+      "loss": 0.4223,
+      "step": 1046
+    },
+    {
+      "epoch": 1.3582089552238805,
+      "grad_norm": 0.25749535429929715,
+      "learning_rate": 3.03992303992304e-05,
+      "loss": 0.4182,
+      "step": 1047
+    },
+    {
+      "epoch": 1.3595068137573005,
+      "grad_norm": 0.22344908087958898,
+      "learning_rate": 3.0375180375180378e-05,
+      "loss": 0.4321,
+      "step": 1048
+    },
+    {
+      "epoch": 1.3608046722907203,
+      "grad_norm": 0.23197019372372432,
+      "learning_rate": 3.0351130351130353e-05,
+      "loss": 0.4399,
+      "step": 1049
+    },
+    {
+      "epoch": 1.3621025308241401,
+      "grad_norm": 0.237479688071686,
+      "learning_rate": 3.0327080327080328e-05,
+      "loss": 0.4395,
+      "step": 1050
+    },
+    {
+      "epoch": 1.36340038935756,
+      "grad_norm": 0.2607112703183096,
+      "learning_rate": 3.0303030303030306e-05,
+      "loss": 0.4379,
+      "step": 1051
+    },
+    {
+      "epoch": 1.36469824789098,
+      "grad_norm": 0.2367521283120222,
+      "learning_rate": 3.027898027898028e-05,
+      "loss": 0.4102,
+      "step": 1052
+    },
+    {
+      "epoch": 1.3659961064243997,
+      "grad_norm": 0.29194250880671824,
+      "learning_rate": 3.025493025493026e-05,
+      "loss": 0.443,
+      "step": 1053
+    },
+    {
+      "epoch": 1.3672939649578195,
+      "grad_norm": 0.26648481874523033,
+      "learning_rate": 3.0230880230880233e-05,
+      "loss": 0.434,
+      "step": 1054
+    },
+    {
+      "epoch": 1.3685918234912395,
+      "grad_norm": 0.26285311775511155,
+      "learning_rate": 3.0206830206830205e-05,
+      "loss": 0.4094,
+      "step": 1055
+    },
+    {
+      "epoch": 1.3698896820246593,
+      "grad_norm": 0.2564406034453145,
+      "learning_rate": 3.0182780182780186e-05,
+      "loss": 0.4398,
+      "step": 1056
+    },
+    {
+      "epoch": 1.3711875405580791,
+      "grad_norm": 0.2796151000660689,
+      "learning_rate": 3.0158730158730158e-05,
+      "loss": 0.4276,
+      "step": 1057
+    },
+    {
+      "epoch": 1.3724853990914991,
+      "grad_norm": 0.26631625570118417,
+      "learning_rate": 3.013468013468014e-05,
+      "loss": 0.4343,
+      "step": 1058
+    },
+    {
+      "epoch": 1.373783257624919,
+      "grad_norm": 0.23806771079205116,
+      "learning_rate": 3.011063011063011e-05,
+      "loss": 0.4334,
+      "step": 1059
+    },
+    {
+      "epoch": 1.3750811161583387,
+      "grad_norm": 0.27717084828254007,
+      "learning_rate": 3.0086580086580092e-05,
+      "loss": 0.4445,
+      "step": 1060
+    },
+    {
+      "epoch": 1.3763789746917585,
+      "grad_norm": 0.25955745890460635,
+      "learning_rate": 3.0062530062530064e-05,
+      "loss": 0.4339,
+      "step": 1061
+    },
+    {
+      "epoch": 1.3776768332251785,
+      "grad_norm": 0.27833631860695396,
+      "learning_rate": 3.003848003848004e-05,
+      "loss": 0.4236,
+      "step": 1062
+    },
+    {
+      "epoch": 1.3789746917585983,
+      "grad_norm": 0.23547677910993067,
+      "learning_rate": 3.0014430014430017e-05,
+      "loss": 0.4451,
+      "step": 1063
+    },
+    {
+      "epoch": 1.3802725502920181,
+      "grad_norm": 0.2818218077693615,
+      "learning_rate": 2.999037999037999e-05,
+      "loss": 0.444,
+      "step": 1064
+    },
+    {
+      "epoch": 1.381570408825438,
+      "grad_norm": 0.22221888861761538,
+      "learning_rate": 2.996632996632997e-05,
+      "loss": 0.403,
+      "step": 1065
+    },
+    {
+      "epoch": 1.382868267358858,
+      "grad_norm": 0.2445641007844883,
+      "learning_rate": 2.9942279942279944e-05,
+      "loss": 0.4163,
+      "step": 1066
+    },
+    {
+      "epoch": 1.3841661258922777,
+      "grad_norm": 0.24596043479534688,
+      "learning_rate": 2.991822991822992e-05,
+      "loss": 0.4227,
+      "step": 1067
+    },
+    {
+      "epoch": 1.3854639844256975,
+      "grad_norm": 0.2278867553746751,
+      "learning_rate": 2.9894179894179897e-05,
+      "loss": 0.4252,
+      "step": 1068
+    },
+    {
+      "epoch": 1.3867618429591175,
+      "grad_norm": 0.2275831703154012,
+      "learning_rate": 2.9870129870129872e-05,
+      "loss": 0.4242,
+      "step": 1069
+    },
+    {
+      "epoch": 1.3880597014925373,
+      "grad_norm": 0.2792734033461531,
+      "learning_rate": 2.984607984607985e-05,
+      "loss": 0.4117,
+      "step": 1070
+    },
+    {
+      "epoch": 1.3893575600259571,
+      "grad_norm": 0.22328171573584032,
+      "learning_rate": 2.9822029822029825e-05,
+      "loss": 0.4115,
+      "step": 1071
+    },
+    {
+      "epoch": 1.3906554185593771,
+      "grad_norm": 0.2554385321095662,
+      "learning_rate": 2.9797979797979796e-05,
+      "loss": 0.4193,
+      "step": 1072
+    },
+    {
+      "epoch": 1.391953277092797,
+      "grad_norm": 0.23411526096024632,
+      "learning_rate": 2.9773929773929778e-05,
+      "loss": 0.401,
+      "step": 1073
+    },
+    {
+      "epoch": 1.3932511356262167,
+      "grad_norm": 0.2417797864860212,
+      "learning_rate": 2.974987974987975e-05,
+      "loss": 0.4175,
+      "step": 1074
+    },
+    {
+      "epoch": 1.3945489941596367,
+      "grad_norm": 0.22909915502634573,
+      "learning_rate": 2.972582972582973e-05,
+      "loss": 0.4145,
+      "step": 1075
+    },
+    {
+      "epoch": 1.3958468526930565,
+      "grad_norm": 0.24654124988803922,
+      "learning_rate": 2.9701779701779702e-05,
+      "loss": 0.4363,
+      "step": 1076
+    },
+    {
+      "epoch": 1.3971447112264763,
+      "grad_norm": 0.23841085503876774,
+      "learning_rate": 2.9677729677729677e-05,
+      "loss": 0.4263,
+      "step": 1077
+    },
+    {
+      "epoch": 1.3984425697598961,
+      "grad_norm": 0.2372620155067687,
+      "learning_rate": 2.9653679653679655e-05,
+      "loss": 0.4164,
+      "step": 1078
+    },
+    {
+      "epoch": 1.399740428293316,
+      "grad_norm": 0.2174655800712752,
+      "learning_rate": 2.962962962962963e-05,
+      "loss": 0.4131,
+      "step": 1079
+    },
+    {
+      "epoch": 1.401038286826736,
+      "grad_norm": 0.25800767555211196,
+      "learning_rate": 2.9605579605579608e-05,
+      "loss": 0.4353,
+      "step": 1080
+    },
+    {
+      "epoch": 1.4023361453601557,
+      "grad_norm": 0.22921820584725752,
+      "learning_rate": 2.9581529581529583e-05,
+      "loss": 0.4048,
+      "step": 1081
+    },
+    {
+      "epoch": 1.4036340038935755,
+      "grad_norm": 0.23402586035822698,
+      "learning_rate": 2.955747955747956e-05,
+      "loss": 0.4424,
+      "step": 1082
+    },
+    {
+      "epoch": 1.4049318624269955,
+      "grad_norm": 0.2352437377010652,
+      "learning_rate": 2.9533429533429536e-05,
+      "loss": 0.4026,
+      "step": 1083
+    },
+    {
+      "epoch": 1.4062297209604153,
+      "grad_norm": 0.2247412452603025,
+      "learning_rate": 2.950937950937951e-05,
+      "loss": 0.4336,
+      "step": 1084
+    },
+    {
+      "epoch": 1.4075275794938351,
+      "grad_norm": 0.22277806225908384,
+      "learning_rate": 2.948532948532949e-05,
+      "loss": 0.4114,
+      "step": 1085
+    },
+    {
+      "epoch": 1.4088254380272551,
+      "grad_norm": 0.241538295583898,
+      "learning_rate": 2.946127946127946e-05,
+      "loss": 0.4394,
+      "step": 1086
+    },
+    {
+      "epoch": 1.410123296560675,
+      "grad_norm": 0.2610774457770794,
+      "learning_rate": 2.943722943722944e-05,
+      "loss": 0.4078,
+      "step": 1087
+    },
+    {
+      "epoch": 1.4114211550940947,
+      "grad_norm": 0.28762989292521274,
+      "learning_rate": 2.9413179413179413e-05,
+      "loss": 0.4109,
+      "step": 1088
+    },
+    {
+      "epoch": 1.4127190136275147,
+      "grad_norm": 0.22764926899208376,
+      "learning_rate": 2.9389129389129388e-05,
+      "loss": 0.4113,
+      "step": 1089
+    },
+    {
+      "epoch": 1.4140168721609345,
+      "grad_norm": 0.2951748901233817,
+      "learning_rate": 2.9365079365079366e-05,
+      "loss": 0.4112,
+      "step": 1090
+    },
+    {
+      "epoch": 1.4153147306943543,
+      "grad_norm": 0.2491523290558,
+      "learning_rate": 2.934102934102934e-05,
+      "loss": 0.4061,
+      "step": 1091
+    },
+    {
+      "epoch": 1.416612589227774,
+      "grad_norm": 0.24774585578620387,
+      "learning_rate": 2.931697931697932e-05,
+      "loss": 0.4208,
+      "step": 1092
+    },
+    {
+      "epoch": 1.417910447761194,
+      "grad_norm": 0.2383954293588994,
+      "learning_rate": 2.9292929292929294e-05,
+      "loss": 0.4243,
+      "step": 1093
+    },
+    {
+      "epoch": 1.419208306294614,
+      "grad_norm": 0.2551439406874695,
+      "learning_rate": 2.926887926887927e-05,
+      "loss": 0.4254,
+      "step": 1094
+    },
+    {
+      "epoch": 1.4205061648280337,
+      "grad_norm": 0.24117464674091157,
+      "learning_rate": 2.9244829244829247e-05,
+      "loss": 0.4082,
+      "step": 1095
+    },
+    {
+      "epoch": 1.4218040233614535,
+      "grad_norm": 0.216503714519975,
+      "learning_rate": 2.922077922077922e-05,
+      "loss": 0.4242,
+      "step": 1096
+    },
+    {
+      "epoch": 1.4231018818948735,
+      "grad_norm": 0.2422351302114647,
+      "learning_rate": 2.91967291967292e-05,
+      "loss": 0.4181,
+      "step": 1097
+    },
+    {
+      "epoch": 1.4243997404282933,
+      "grad_norm": 0.22788810318626124,
+      "learning_rate": 2.9172679172679174e-05,
+      "loss": 0.4203,
+      "step": 1098
+    },
+    {
+      "epoch": 1.425697598961713,
+      "grad_norm": 0.24325054021695863,
+      "learning_rate": 2.9148629148629146e-05,
+      "loss": 0.423,
+      "step": 1099
+    },
+    {
+      "epoch": 1.4269954574951331,
+      "grad_norm": 0.23344217723128893,
+      "learning_rate": 2.9124579124579127e-05,
+      "loss": 0.4162,
+      "step": 1100
+    },
+    {
+      "epoch": 1.428293316028553,
+      "grad_norm": 0.2543879197815251,
+      "learning_rate": 2.91005291005291e-05,
+      "loss": 0.4376,
+      "step": 1101
+    },
+    {
+      "epoch": 1.4295911745619727,
+      "grad_norm": 0.2501682851968916,
+      "learning_rate": 2.907647907647908e-05,
+      "loss": 0.4164,
+      "step": 1102
+    },
+    {
+      "epoch": 1.4308890330953927,
+      "grad_norm": 0.21492688667239696,
+      "learning_rate": 2.905242905242905e-05,
+      "loss": 0.4113,
+      "step": 1103
+    },
+    {
+      "epoch": 1.4321868916288125,
+      "grad_norm": 0.2658354623358409,
+      "learning_rate": 2.9028379028379033e-05,
+      "loss": 0.4315,
+      "step": 1104
+    },
+    {
+      "epoch": 1.4334847501622323,
+      "grad_norm": 0.2831492165913101,
+      "learning_rate": 2.9004329004329005e-05,
+      "loss": 0.4338,
+      "step": 1105
+    },
+    {
+      "epoch": 1.434782608695652,
+      "grad_norm": 0.22471574523786844,
+      "learning_rate": 2.898027898027898e-05,
+      "loss": 0.4157,
+      "step": 1106
+    },
+    {
+      "epoch": 1.436080467229072,
+      "grad_norm": 0.24385415494263882,
+      "learning_rate": 2.8956228956228958e-05,
+      "loss": 0.4172,
+      "step": 1107
+    },
+    {
+      "epoch": 1.437378325762492,
+      "grad_norm": 0.2722198360023313,
+      "learning_rate": 2.8932178932178932e-05,
+      "loss": 0.407,
+      "step": 1108
+    },
+    {
+      "epoch": 1.4386761842959117,
+      "grad_norm": 0.21019395837235147,
+      "learning_rate": 2.890812890812891e-05,
+      "loss": 0.4189,
+      "step": 1109
+    },
+    {
+      "epoch": 1.4399740428293315,
+      "grad_norm": 0.23984115871051997,
+      "learning_rate": 2.8884078884078885e-05,
+      "loss": 0.4261,
+      "step": 1110
+    },
+    {
+      "epoch": 1.4412719013627515,
+      "grad_norm": 0.25738974208155574,
+      "learning_rate": 2.886002886002886e-05,
+      "loss": 0.4694,
+      "step": 1111
+    },
+    {
+      "epoch": 1.4425697598961713,
+      "grad_norm": 0.2739742760545878,
+      "learning_rate": 2.8835978835978838e-05,
+      "loss": 0.4284,
+      "step": 1112
+    },
+    {
+      "epoch": 1.443867618429591,
+      "grad_norm": 0.2563615388623274,
+      "learning_rate": 2.8811928811928813e-05,
+      "loss": 0.4187,
+      "step": 1113
+    },
+    {
+      "epoch": 1.4451654769630111,
+      "grad_norm": 0.2355873474417628,
+      "learning_rate": 2.878787878787879e-05,
+      "loss": 0.4201,
+      "step": 1114
+    },
+    {
+      "epoch": 1.446463335496431,
+      "grad_norm": 0.3037078461896459,
+      "learning_rate": 2.8763828763828766e-05,
+      "loss": 0.4164,
+      "step": 1115
+    },
+    {
+      "epoch": 1.4477611940298507,
+      "grad_norm": 0.2778889853166693,
+      "learning_rate": 2.8739778739778737e-05,
+      "loss": 0.4263,
+      "step": 1116
+    },
+    {
+      "epoch": 1.4490590525632707,
+      "grad_norm": 0.25304453875189337,
+      "learning_rate": 2.871572871572872e-05,
+      "loss": 0.4238,
+      "step": 1117
+    },
+    {
+      "epoch": 1.4503569110966905,
+      "grad_norm": 0.2617845594600046,
+      "learning_rate": 2.869167869167869e-05,
+      "loss": 0.4325,
+      "step": 1118
+    },
+    {
+      "epoch": 1.4516547696301103,
+      "grad_norm": 0.28455794197858575,
+      "learning_rate": 2.8667628667628672e-05,
+      "loss": 0.4156,
+      "step": 1119
+    },
+    {
+      "epoch": 1.45295262816353,
+      "grad_norm": 0.23740191563596869,
+      "learning_rate": 2.8643578643578643e-05,
+      "loss": 0.4261,
+      "step": 1120
+    },
+    {
+      "epoch": 1.45425048669695,
+      "grad_norm": 0.24072843737266889,
+      "learning_rate": 2.8619528619528618e-05,
+      "loss": 0.4151,
+      "step": 1121
+    },
+    {
+      "epoch": 1.45554834523037,
+      "grad_norm": 0.2700405419220064,
+      "learning_rate": 2.8595478595478596e-05,
+      "loss": 0.4252,
+      "step": 1122
+    },
+    {
+      "epoch": 1.4568462037637897,
+      "grad_norm": 0.2488359814096937,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 0.4428,
+      "step": 1123
+    },
+    {
+      "epoch": 1.4581440622972095,
+      "grad_norm": 0.23487079361910798,
+      "learning_rate": 2.854737854737855e-05,
+      "loss": 0.4216,
+      "step": 1124
+    },
+    {
+      "epoch": 1.4594419208306295,
+      "grad_norm": 0.2466576788103327,
+      "learning_rate": 2.8523328523328524e-05,
+      "loss": 0.4226,
+      "step": 1125
+    },
+    {
+      "epoch": 1.4607397793640493,
+      "grad_norm": 0.2391996649480345,
+      "learning_rate": 2.8499278499278502e-05,
+      "loss": 0.4096,
+      "step": 1126
+    },
+    {
+      "epoch": 1.462037637897469,
+      "grad_norm": 0.24258796808928063,
+      "learning_rate": 2.8475228475228477e-05,
+      "loss": 0.4145,
+      "step": 1127
+    },
+    {
+      "epoch": 1.4633354964308891,
+      "grad_norm": 0.2560406117346898,
+      "learning_rate": 2.845117845117845e-05,
+      "loss": 0.4312,
+      "step": 1128
+    },
+    {
+      "epoch": 1.464633354964309,
+      "grad_norm": 0.2861667925744788,
+      "learning_rate": 2.842712842712843e-05,
+      "loss": 0.4328,
+      "step": 1129
+    },
+    {
+      "epoch": 1.4659312134977287,
+      "grad_norm": 0.25402106108095945,
+      "learning_rate": 2.8403078403078404e-05,
+      "loss": 0.4258,
+      "step": 1130
+    },
+    {
+      "epoch": 1.4672290720311487,
+      "grad_norm": 0.24075563965908323,
+      "learning_rate": 2.8379028379028383e-05,
+      "loss": 0.4397,
+      "step": 1131
+    },
+    {
+      "epoch": 1.4685269305645685,
+      "grad_norm": 0.28522807177447185,
+      "learning_rate": 2.8354978354978357e-05,
+      "loss": 0.4303,
+      "step": 1132
+    },
+    {
+      "epoch": 1.4698247890979883,
+      "grad_norm": 0.26004049344709895,
+      "learning_rate": 2.833092833092833e-05,
+      "loss": 0.4294,
+      "step": 1133
+    },
+    {
+      "epoch": 1.471122647631408,
+      "grad_norm": 0.29853418709486346,
+      "learning_rate": 2.830687830687831e-05,
+      "loss": 0.4323,
+      "step": 1134
+    },
+    {
+      "epoch": 1.4724205061648281,
+      "grad_norm": 0.2633187635030568,
+      "learning_rate": 2.8282828282828282e-05,
+      "loss": 0.439,
+      "step": 1135
+    },
+    {
+      "epoch": 1.473718364698248,
+      "grad_norm": 0.3157910451013907,
+      "learning_rate": 2.8258778258778263e-05,
+      "loss": 0.4337,
+      "step": 1136
+    },
+    {
+      "epoch": 1.4750162232316677,
+      "grad_norm": 0.27203466331005977,
+      "learning_rate": 2.8234728234728235e-05,
+      "loss": 0.4198,
+      "step": 1137
+    },
+    {
+      "epoch": 1.4763140817650875,
+      "grad_norm": 0.2508023724498563,
+      "learning_rate": 2.821067821067821e-05,
+      "loss": 0.4267,
+      "step": 1138
+    },
+    {
+      "epoch": 1.4776119402985075,
+      "grad_norm": 0.2808852939089359,
+      "learning_rate": 2.8186628186628188e-05,
+      "loss": 0.4281,
+      "step": 1139
+    },
+    {
+      "epoch": 1.4789097988319273,
+      "grad_norm": 0.24576060851892864,
+      "learning_rate": 2.8162578162578162e-05,
+      "loss": 0.4206,
+      "step": 1140
+    },
+    {
+      "epoch": 1.480207657365347,
+      "grad_norm": 0.29052064234787545,
+      "learning_rate": 2.813852813852814e-05,
+      "loss": 0.413,
+      "step": 1141
+    },
+    {
+      "epoch": 1.4815055158987671,
+      "grad_norm": 0.23398439998245094,
+      "learning_rate": 2.8114478114478115e-05,
+      "loss": 0.4352,
+      "step": 1142
+    },
+    {
+      "epoch": 1.482803374432187,
+      "grad_norm": 0.25678491256047153,
+      "learning_rate": 2.809042809042809e-05,
+      "loss": 0.4346,
+      "step": 1143
+    },
+    {
+      "epoch": 1.4841012329656067,
+      "grad_norm": 0.2826048101734635,
+      "learning_rate": 2.8066378066378068e-05,
+      "loss": 0.4262,
+      "step": 1144
+    },
+    {
+      "epoch": 1.4853990914990267,
+      "grad_norm": 0.25015775061708817,
+      "learning_rate": 2.8042328042328043e-05,
+      "loss": 0.4129,
+      "step": 1145
+    },
+    {
+      "epoch": 1.4866969500324465,
+      "grad_norm": 0.2561977557181458,
+      "learning_rate": 2.801827801827802e-05,
+      "loss": 0.417,
+      "step": 1146
+    },
+    {
+      "epoch": 1.4879948085658663,
+      "grad_norm": 0.25036370521793533,
+      "learning_rate": 2.7994227994227996e-05,
+      "loss": 0.4079,
+      "step": 1147
+    },
+    {
+      "epoch": 1.4892926670992863,
+      "grad_norm": 0.28901223175805674,
+      "learning_rate": 2.7970177970177974e-05,
+      "loss": 0.4094,
+      "step": 1148
+    },
+    {
+      "epoch": 1.490590525632706,
+      "grad_norm": 0.23134484811007663,
+      "learning_rate": 2.794612794612795e-05,
+      "loss": 0.4476,
+      "step": 1149
+    },
+    {
+      "epoch": 1.491888384166126,
+      "grad_norm": 0.25137689970727467,
+      "learning_rate": 2.792207792207792e-05,
+      "loss": 0.4416,
+      "step": 1150
+    },
+    {
+      "epoch": 1.4931862426995457,
+      "grad_norm": 0.2524284266331274,
+      "learning_rate": 2.7898027898027902e-05,
+      "loss": 0.4295,
+      "step": 1151
+    },
+    {
+      "epoch": 1.4944841012329655,
+      "grad_norm": 0.22266682751444122,
+      "learning_rate": 2.7873977873977873e-05,
+      "loss": 0.4115,
+      "step": 1152
+    },
+    {
+      "epoch": 1.4957819597663855,
+      "grad_norm": 0.2085505106465029,
+      "learning_rate": 2.7849927849927855e-05,
+      "loss": 0.4271,
+      "step": 1153
+    },
+    {
+      "epoch": 1.4970798182998053,
+      "grad_norm": 0.2352572065506912,
+      "learning_rate": 2.7825877825877826e-05,
+      "loss": 0.4129,
+      "step": 1154
+    },
+    {
+      "epoch": 1.498377676833225,
+      "grad_norm": 0.2322270923460416,
+      "learning_rate": 2.78018278018278e-05,
+      "loss": 0.4404,
+      "step": 1155
+    },
+    {
+      "epoch": 1.499675535366645,
+      "grad_norm": 0.20327079840186968,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.3992,
+      "step": 1156
+    },
+    {
+      "epoch": 1.500973393900065,
+      "grad_norm": 0.22409767153079405,
+      "learning_rate": 2.7753727753727754e-05,
+      "loss": 0.4351,
+      "step": 1157
+    },
+    {
+      "epoch": 1.5022712524334847,
+      "grad_norm": 0.21789564363803948,
+      "learning_rate": 2.7729677729677732e-05,
+      "loss": 0.4263,
+      "step": 1158
+    },
+    {
+      "epoch": 1.5035691109669047,
+      "grad_norm": 0.23289144485137522,
+      "learning_rate": 2.7705627705627707e-05,
+      "loss": 0.4296,
+      "step": 1159
+    },
+    {
+      "epoch": 1.5048669695003245,
+      "grad_norm": 0.22790992420912343,
+      "learning_rate": 2.768157768157768e-05,
+      "loss": 0.4275,
+      "step": 1160
+    },
+    {
+      "epoch": 1.5061648280337443,
+      "grad_norm": 0.2180550660231808,
+      "learning_rate": 2.765752765752766e-05,
+      "loss": 0.4242,
+      "step": 1161
+    },
+    {
+      "epoch": 1.5074626865671643,
+      "grad_norm": 0.23490836544769822,
+      "learning_rate": 2.7633477633477635e-05,
+      "loss": 0.43,
+      "step": 1162
+    },
+    {
+      "epoch": 1.5087605451005839,
+      "grad_norm": 0.22192430223007012,
+      "learning_rate": 2.7609427609427613e-05,
+      "loss": 0.429,
+      "step": 1163
+    },
+    {
+      "epoch": 1.510058403634004,
+      "grad_norm": 0.2311428363766677,
+      "learning_rate": 2.7585377585377587e-05,
+      "loss": 0.4414,
+      "step": 1164
+    },
+    {
+      "epoch": 1.511356262167424,
+      "grad_norm": 0.2001408824188442,
+      "learning_rate": 2.756132756132756e-05,
+      "loss": 0.4225,
+      "step": 1165
+    },
+    {
+      "epoch": 1.5126541207008435,
+      "grad_norm": 0.2279619365636095,
+      "learning_rate": 2.753727753727754e-05,
+      "loss": 0.4149,
+      "step": 1166
+    },
+    {
+      "epoch": 1.5139519792342635,
+      "grad_norm": 0.24302952170040615,
+      "learning_rate": 2.7513227513227512e-05,
+      "loss": 0.4153,
+      "step": 1167
+    },
+    {
+      "epoch": 1.5152498377676833,
+      "grad_norm": 0.2252876935922963,
+      "learning_rate": 2.7489177489177493e-05,
+      "loss": 0.4135,
+      "step": 1168
+    },
+    {
+      "epoch": 1.516547696301103,
+      "grad_norm": 0.22162900859128726,
+      "learning_rate": 2.7465127465127465e-05,
+      "loss": 0.4283,
+      "step": 1169
+    },
+    {
+      "epoch": 1.517845554834523,
+      "grad_norm": 0.23728269459202284,
+      "learning_rate": 2.7441077441077446e-05,
+      "loss": 0.4284,
+      "step": 1170
+    },
+    {
+      "epoch": 1.519143413367943,
+      "grad_norm": 0.2073068726020532,
+      "learning_rate": 2.7417027417027418e-05,
+      "loss": 0.419,
+      "step": 1171
+    },
+    {
+      "epoch": 1.5204412719013627,
+      "grad_norm": 0.2308870482056988,
+      "learning_rate": 2.7392977392977392e-05,
+      "loss": 0.4409,
+      "step": 1172
+    },
+    {
+      "epoch": 1.5217391304347827,
+      "grad_norm": 0.21898533880032697,
+      "learning_rate": 2.736892736892737e-05,
+      "loss": 0.4171,
+      "step": 1173
+    },
+    {
+      "epoch": 1.5230369889682025,
+      "grad_norm": 0.21000995819843474,
+      "learning_rate": 2.7344877344877345e-05,
+      "loss": 0.417,
+      "step": 1174
+    },
+    {
+      "epoch": 1.5243348475016223,
+      "grad_norm": 0.2150245170655777,
+      "learning_rate": 2.7320827320827324e-05,
+      "loss": 0.4365,
+      "step": 1175
+    },
+    {
+      "epoch": 1.5256327060350423,
+      "grad_norm": 0.24290565598308295,
+      "learning_rate": 2.72967772967773e-05,
+      "loss": 0.4201,
+      "step": 1176
+    },
+    {
+      "epoch": 1.5269305645684619,
+      "grad_norm": 0.2304464719146474,
+      "learning_rate": 2.7272727272727273e-05,
+      "loss": 0.4149,
+      "step": 1177
+    },
+    {
+      "epoch": 1.528228423101882,
+      "grad_norm": 0.23523933221515506,
+      "learning_rate": 2.724867724867725e-05,
+      "loss": 0.453,
+      "step": 1178
+    },
+    {
+      "epoch": 1.529526281635302,
+      "grad_norm": 0.2253825255615944,
+      "learning_rate": 2.7224627224627226e-05,
+      "loss": 0.4209,
+      "step": 1179
+    },
+    {
+      "epoch": 1.5308241401687215,
+      "grad_norm": 0.2742775834013937,
+      "learning_rate": 2.7200577200577204e-05,
+      "loss": 0.442,
+      "step": 1180
+    },
+    {
+      "epoch": 1.5321219987021415,
+      "grad_norm": 0.2176528388600847,
+      "learning_rate": 2.717652717652718e-05,
+      "loss": 0.4329,
+      "step": 1181
+    },
+    {
+      "epoch": 1.5334198572355613,
+      "grad_norm": 0.23818415433225926,
+      "learning_rate": 2.715247715247715e-05,
+      "loss": 0.4187,
+      "step": 1182
+    },
+    {
+      "epoch": 1.534717715768981,
+      "grad_norm": 0.26109881547859903,
+      "learning_rate": 2.7128427128427132e-05,
+      "loss": 0.4251,
+      "step": 1183
+    },
+    {
+      "epoch": 1.536015574302401,
+      "grad_norm": 0.2196942384869763,
+      "learning_rate": 2.7104377104377103e-05,
+      "loss": 0.418,
+      "step": 1184
+    },
+    {
+      "epoch": 1.537313432835821,
+      "grad_norm": 0.2400322015222156,
+      "learning_rate": 2.7080327080327085e-05,
+      "loss": 0.4109,
+      "step": 1185
+    },
+    {
+      "epoch": 1.5386112913692407,
+      "grad_norm": 0.23150552647711828,
+      "learning_rate": 2.7056277056277056e-05,
+      "loss": 0.4264,
+      "step": 1186
+    },
+    {
+      "epoch": 1.5399091499026607,
+      "grad_norm": 0.22005403208488783,
+      "learning_rate": 2.703222703222703e-05,
+      "loss": 0.4039,
+      "step": 1187
+    },
+    {
+      "epoch": 1.5412070084360805,
+      "grad_norm": 0.22581597634393283,
+      "learning_rate": 2.700817700817701e-05,
+      "loss": 0.425,
+      "step": 1188
+    },
+    {
+      "epoch": 1.5425048669695003,
+      "grad_norm": 0.2382695341310579,
+      "learning_rate": 2.6984126984126984e-05,
+      "loss": 0.4496,
+      "step": 1189
+    },
+    {
+      "epoch": 1.5438027255029203,
+      "grad_norm": 0.2203961917107305,
+      "learning_rate": 2.6960076960076962e-05,
+      "loss": 0.4155,
+      "step": 1190
+    },
+    {
+      "epoch": 1.5451005840363399,
+      "grad_norm": 0.25210372953982285,
+      "learning_rate": 2.6936026936026937e-05,
+      "loss": 0.4342,
+      "step": 1191
+    },
+    {
+      "epoch": 1.54639844256976,
+      "grad_norm": 0.238604523146027,
+      "learning_rate": 2.691197691197691e-05,
+      "loss": 0.4323,
+      "step": 1192
+    },
+    {
+      "epoch": 1.54769630110318,
+      "grad_norm": 0.23138471132633792,
+      "learning_rate": 2.688792688792689e-05,
+      "loss": 0.4242,
+      "step": 1193
+    },
+    {
+      "epoch": 1.5489941596365995,
+      "grad_norm": 0.2320529813667351,
+      "learning_rate": 2.6863876863876865e-05,
+      "loss": 0.4363,
+      "step": 1194
+    },
+    {
+      "epoch": 1.5502920181700195,
+      "grad_norm": 0.22679612862184145,
+      "learning_rate": 2.6839826839826843e-05,
+      "loss": 0.4253,
+      "step": 1195
+    },
+    {
+      "epoch": 1.5515898767034393,
+      "grad_norm": 0.2665688161045152,
+      "learning_rate": 2.6815776815776818e-05,
+      "loss": 0.4222,
+      "step": 1196
+    },
+    {
+      "epoch": 1.552887735236859,
+      "grad_norm": 0.21178913986030537,
+      "learning_rate": 2.6791726791726796e-05,
+      "loss": 0.422,
+      "step": 1197
+    },
+    {
+      "epoch": 1.554185593770279,
+      "grad_norm": 0.24464931528999015,
+      "learning_rate": 2.676767676767677e-05,
+      "loss": 0.4241,
+      "step": 1198
+    },
+    {
+      "epoch": 1.5554834523036989,
+      "grad_norm": 0.22319718290311183,
+      "learning_rate": 2.6743626743626742e-05,
+      "loss": 0.4168,
+      "step": 1199
+    },
+    {
+      "epoch": 1.5567813108371187,
+      "grad_norm": 0.2302808693777694,
+      "learning_rate": 2.6719576719576723e-05,
+      "loss": 0.41,
+      "step": 1200
+    },
+    {
+      "epoch": 1.5580791693705387,
+      "grad_norm": 0.2317544115600513,
+      "learning_rate": 2.6695526695526695e-05,
+      "loss": 0.4555,
+      "step": 1201
+    },
+    {
+      "epoch": 1.5593770279039585,
+      "grad_norm": 0.2554067046842974,
+      "learning_rate": 2.6671476671476676e-05,
+      "loss": 0.4075,
+      "step": 1202
+    },
+    {
+      "epoch": 1.5606748864373783,
+      "grad_norm": 0.23832375199078534,
+      "learning_rate": 2.6647426647426648e-05,
+      "loss": 0.4149,
+      "step": 1203
+    },
+    {
+      "epoch": 1.5619727449707983,
+      "grad_norm": 0.2387421511606967,
+      "learning_rate": 2.6623376623376623e-05,
+      "loss": 0.4349,
+      "step": 1204
+    },
+    {
+      "epoch": 1.563270603504218,
+      "grad_norm": 0.24466721743899011,
+      "learning_rate": 2.65993265993266e-05,
+      "loss": 0.4158,
+      "step": 1205
+    },
+    {
+      "epoch": 1.5645684620376379,
+      "grad_norm": 0.23174513616055498,
+      "learning_rate": 2.6575276575276575e-05,
+      "loss": 0.4443,
+      "step": 1206
+    },
+    {
+      "epoch": 1.565866320571058,
+      "grad_norm": 0.218582346579111,
+      "learning_rate": 2.6551226551226554e-05,
+      "loss": 0.4228,
+      "step": 1207
+    },
+    {
+      "epoch": 1.5671641791044775,
+      "grad_norm": 0.23236180907143378,
+      "learning_rate": 2.652717652717653e-05,
+      "loss": 0.4198,
+      "step": 1208
+    },
+    {
+      "epoch": 1.5684620376378975,
+      "grad_norm": 0.2461597550351122,
+      "learning_rate": 2.6503126503126503e-05,
+      "loss": 0.4388,
+      "step": 1209
+    },
+    {
+      "epoch": 1.5697598961713173,
+      "grad_norm": 0.24135274528584466,
+      "learning_rate": 2.647907647907648e-05,
+      "loss": 0.4182,
+      "step": 1210
+    },
+    {
+      "epoch": 1.571057754704737,
+      "grad_norm": 0.23011430180334824,
+      "learning_rate": 2.6455026455026456e-05,
+      "loss": 0.4345,
+      "step": 1211
+    },
+    {
+      "epoch": 1.572355613238157,
+      "grad_norm": 0.25813925411615873,
+      "learning_rate": 2.6430976430976434e-05,
+      "loss": 0.4152,
+      "step": 1212
+    },
+    {
+      "epoch": 1.5736534717715769,
+      "grad_norm": 0.2361569395941438,
+      "learning_rate": 2.640692640692641e-05,
+      "loss": 0.4107,
+      "step": 1213
+    },
+    {
+      "epoch": 1.5749513303049967,
+      "grad_norm": 0.26363884372789825,
+      "learning_rate": 2.638287638287638e-05,
+      "loss": 0.4392,
+      "step": 1214
+    },
+    {
+      "epoch": 1.5762491888384167,
+      "grad_norm": 0.24244610329485705,
+      "learning_rate": 2.6358826358826362e-05,
+      "loss": 0.4164,
+      "step": 1215
+    },
+    {
+      "epoch": 1.5775470473718365,
+      "grad_norm": 0.2552987758465308,
+      "learning_rate": 2.6334776334776333e-05,
+      "loss": 0.4339,
+      "step": 1216
+    },
+    {
+      "epoch": 1.5788449059052563,
+      "grad_norm": 0.2622601300659554,
+      "learning_rate": 2.6310726310726315e-05,
+      "loss": 0.4081,
+      "step": 1217
+    },
+    {
+      "epoch": 1.5801427644386763,
+      "grad_norm": 0.23435950487013313,
+      "learning_rate": 2.6286676286676286e-05,
+      "loss": 0.4266,
+      "step": 1218
+    },
+    {
+      "epoch": 1.581440622972096,
+      "grad_norm": 0.31150362868262865,
+      "learning_rate": 2.6262626262626268e-05,
+      "loss": 0.4205,
+      "step": 1219
+    },
+    {
+      "epoch": 1.5827384815055159,
+      "grad_norm": 0.2356568945579236,
+      "learning_rate": 2.623857623857624e-05,
+      "loss": 0.4235,
+      "step": 1220
+    },
+    {
+      "epoch": 1.584036340038936,
+      "grad_norm": 0.2636851026847217,
+      "learning_rate": 2.6214526214526214e-05,
+      "loss": 0.4194,
+      "step": 1221
+    },
+    {
+      "epoch": 1.5853341985723555,
+      "grad_norm": 0.2609824789762705,
+      "learning_rate": 2.6190476190476192e-05,
+      "loss": 0.4386,
+      "step": 1222
+    },
+    {
+      "epoch": 1.5866320571057755,
+      "grad_norm": 0.2503475112982072,
+      "learning_rate": 2.6166426166426167e-05,
+      "loss": 0.4295,
+      "step": 1223
+    },
+    {
+      "epoch": 1.5879299156391953,
+      "grad_norm": 0.2748789264904923,
+      "learning_rate": 2.6142376142376145e-05,
+      "loss": 0.431,
+      "step": 1224
+    },
+    {
+      "epoch": 1.589227774172615,
+      "grad_norm": 0.2122856536086439,
+      "learning_rate": 2.611832611832612e-05,
+      "loss": 0.4151,
+      "step": 1225
+    },
+    {
+      "epoch": 1.590525632706035,
+      "grad_norm": 0.2882371321327433,
+      "learning_rate": 2.6094276094276095e-05,
+      "loss": 0.4242,
+      "step": 1226
+    },
+    {
+      "epoch": 1.5918234912394549,
+      "grad_norm": 0.22024360438567706,
+      "learning_rate": 2.6070226070226073e-05,
+      "loss": 0.4173,
+      "step": 1227
+    },
+    {
+      "epoch": 1.5931213497728747,
+      "grad_norm": 0.23708353175014626,
+      "learning_rate": 2.6046176046176048e-05,
+      "loss": 0.4251,
+      "step": 1228
+    },
+    {
+      "epoch": 1.5944192083062947,
+      "grad_norm": 0.2658200863217972,
+      "learning_rate": 2.6022126022126026e-05,
+      "loss": 0.4408,
+      "step": 1229
+    },
+    {
+      "epoch": 1.5957170668397145,
+      "grad_norm": 0.21583066555363375,
+      "learning_rate": 2.5998075998076e-05,
+      "loss": 0.4191,
+      "step": 1230
+    },
+    {
+      "epoch": 1.5970149253731343,
+      "grad_norm": 0.2777242614566809,
+      "learning_rate": 2.5974025974025972e-05,
+      "loss": 0.4393,
+      "step": 1231
+    },
+    {
+      "epoch": 1.5983127839065543,
+      "grad_norm": 0.23219187892619703,
+      "learning_rate": 2.5949975949975954e-05,
+      "loss": 0.4265,
+      "step": 1232
+    },
+    {
+      "epoch": 1.599610642439974,
+      "grad_norm": 0.29387387169794943,
+      "learning_rate": 2.5925925925925925e-05,
+      "loss": 0.4244,
+      "step": 1233
+    },
+    {
+      "epoch": 1.6009085009733939,
+      "grad_norm": 0.22151955032464254,
+      "learning_rate": 2.5901875901875906e-05,
+      "loss": 0.4085,
+      "step": 1234
+    },
+    {
+      "epoch": 1.602206359506814,
+      "grad_norm": 0.24242248455059523,
+      "learning_rate": 2.5877825877825878e-05,
+      "loss": 0.4285,
+      "step": 1235
+    },
+    {
+      "epoch": 1.6035042180402335,
+      "grad_norm": 0.2621217435997206,
+      "learning_rate": 2.5853775853775853e-05,
+      "loss": 0.4379,
+      "step": 1236
+    },
+    {
+      "epoch": 1.6048020765736535,
+      "grad_norm": 0.22823804591889496,
+      "learning_rate": 2.582972582972583e-05,
+      "loss": 0.4211,
+      "step": 1237
+    },
+    {
+      "epoch": 1.6060999351070735,
+      "grad_norm": 0.2353758128022499,
+      "learning_rate": 2.5805675805675806e-05,
+      "loss": 0.4127,
+      "step": 1238
+    },
+    {
+      "epoch": 1.607397793640493,
+      "grad_norm": 0.22811130965496038,
+      "learning_rate": 2.5781625781625784e-05,
+      "loss": 0.4246,
+      "step": 1239
+    },
+    {
+      "epoch": 1.608695652173913,
+      "grad_norm": 0.2366434232412805,
+      "learning_rate": 2.575757575757576e-05,
+      "loss": 0.4109,
+      "step": 1240
+    },
+    {
+      "epoch": 1.6099935107073329,
+      "grad_norm": 0.20375567441674386,
+      "learning_rate": 2.5733525733525737e-05,
+      "loss": 0.4146,
+      "step": 1241
+    },
+    {
+      "epoch": 1.6112913692407527,
+      "grad_norm": 0.2436449942466404,
+      "learning_rate": 2.570947570947571e-05,
+      "loss": 0.4162,
+      "step": 1242
+    },
+    {
+      "epoch": 1.6125892277741727,
+      "grad_norm": 0.22023021001348508,
+      "learning_rate": 2.5685425685425686e-05,
+      "loss": 0.4136,
+      "step": 1243
+    },
+    {
+      "epoch": 1.6138870863075925,
+      "grad_norm": 0.2069116265186359,
+      "learning_rate": 2.5661375661375664e-05,
+      "loss": 0.3999,
+      "step": 1244
+    },
+    {
+      "epoch": 1.6151849448410123,
+      "grad_norm": 0.24450308671714907,
+      "learning_rate": 2.563732563732564e-05,
+      "loss": 0.4352,
+      "step": 1245
+    },
+    {
+      "epoch": 1.6164828033744323,
+      "grad_norm": 0.2361666753423955,
+      "learning_rate": 2.5613275613275617e-05,
+      "loss": 0.4215,
+      "step": 1246
+    },
+    {
+      "epoch": 1.617780661907852,
+      "grad_norm": 0.24709753794900446,
+      "learning_rate": 2.5589225589225592e-05,
+      "loss": 0.4095,
+      "step": 1247
+    },
+    {
+      "epoch": 1.6190785204412719,
+      "grad_norm": 0.22770211218246428,
+      "learning_rate": 2.5565175565175563e-05,
+      "loss": 0.4208,
+      "step": 1248
+    },
+    {
+      "epoch": 1.6203763789746919,
+      "grad_norm": 0.24442985342584414,
+      "learning_rate": 2.5541125541125545e-05,
+      "loss": 0.4048,
+      "step": 1249
+    },
+    {
+      "epoch": 1.6216742375081115,
+      "grad_norm": 0.2449341182023967,
+      "learning_rate": 2.5517075517075516e-05,
+      "loss": 0.429,
+      "step": 1250
+    },
+    {
+      "epoch": 1.6229720960415315,
+      "grad_norm": 0.22314422338157636,
+      "learning_rate": 2.5493025493025498e-05,
+      "loss": 0.4161,
+      "step": 1251
+    },
+    {
+      "epoch": 1.6242699545749515,
+      "grad_norm": 0.22271710889727703,
+      "learning_rate": 2.546897546897547e-05,
+      "loss": 0.4223,
+      "step": 1252
+    },
+    {
+      "epoch": 1.625567813108371,
+      "grad_norm": 0.23943855813232637,
+      "learning_rate": 2.5444925444925444e-05,
+      "loss": 0.4263,
+      "step": 1253
+    },
+    {
+      "epoch": 1.626865671641791,
+      "grad_norm": 0.22346829290932305,
+      "learning_rate": 2.5420875420875422e-05,
+      "loss": 0.4002,
+      "step": 1254
+    },
+    {
+      "epoch": 1.6281635301752109,
+      "grad_norm": 0.21819410830608127,
+      "learning_rate": 2.5396825396825397e-05,
+      "loss": 0.4228,
+      "step": 1255
+    },
+    {
+      "epoch": 1.6294613887086307,
+      "grad_norm": 0.2487542450136884,
+      "learning_rate": 2.5372775372775375e-05,
+      "loss": 0.431,
+      "step": 1256
+    },
+    {
+      "epoch": 1.6307592472420507,
+      "grad_norm": 0.22276858066653343,
+      "learning_rate": 2.534872534872535e-05,
+      "loss": 0.3975,
+      "step": 1257
+    },
+    {
+      "epoch": 1.6320571057754705,
+      "grad_norm": 0.20406534653386582,
+      "learning_rate": 2.5324675324675325e-05,
+      "loss": 0.4308,
+      "step": 1258
+    },
+    {
+      "epoch": 1.6333549643088903,
+      "grad_norm": 0.2369459882014465,
+      "learning_rate": 2.5300625300625303e-05,
+      "loss": 0.4434,
+      "step": 1259
+    },
+    {
+      "epoch": 1.6346528228423103,
+      "grad_norm": 0.23054872564198348,
+      "learning_rate": 2.5276575276575278e-05,
+      "loss": 0.4296,
+      "step": 1260
+    },
+    {
+      "epoch": 1.63595068137573,
+      "grad_norm": 0.21314688817002478,
+      "learning_rate": 2.5252525252525256e-05,
+      "loss": 0.4234,
+      "step": 1261
+    },
+    {
+      "epoch": 1.6372485399091499,
+      "grad_norm": 0.22937591574682323,
+      "learning_rate": 2.522847522847523e-05,
+      "loss": 0.425,
+      "step": 1262
+    },
+    {
+      "epoch": 1.6385463984425699,
+      "grad_norm": 0.23974213218799267,
+      "learning_rate": 2.520442520442521e-05,
+      "loss": 0.4393,
+      "step": 1263
+    },
+    {
+      "epoch": 1.6398442569759895,
+      "grad_norm": 0.23441342590653153,
+      "learning_rate": 2.5180375180375184e-05,
+      "loss": 0.4474,
+      "step": 1264
+    },
+    {
+      "epoch": 1.6411421155094095,
+      "grad_norm": 0.22460634450789943,
+      "learning_rate": 2.5156325156325155e-05,
+      "loss": 0.4405,
+      "step": 1265
+    },
+    {
+      "epoch": 1.6424399740428295,
+      "grad_norm": 0.21099257965853147,
+      "learning_rate": 2.5132275132275137e-05,
+      "loss": 0.4147,
+      "step": 1266
+    },
+    {
+      "epoch": 1.643737832576249,
+      "grad_norm": 0.24160346011701583,
+      "learning_rate": 2.5108225108225108e-05,
+      "loss": 0.4397,
+      "step": 1267
+    },
+    {
+      "epoch": 1.645035691109669,
+      "grad_norm": 0.21504387068528427,
+      "learning_rate": 2.5084175084175086e-05,
+      "loss": 0.4134,
+      "step": 1268
+    },
+    {
+      "epoch": 1.6463335496430889,
+      "grad_norm": 0.20136235310740322,
+      "learning_rate": 2.506012506012506e-05,
+      "loss": 0.4352,
+      "step": 1269
+    },
+    {
+      "epoch": 1.6476314081765087,
+      "grad_norm": 0.20297036044525715,
+      "learning_rate": 2.5036075036075036e-05,
+      "loss": 0.4181,
+      "step": 1270
+    },
+    {
+      "epoch": 1.6489292667099287,
+      "grad_norm": 0.22303019507601843,
+      "learning_rate": 2.5012025012025014e-05,
+      "loss": 0.4028,
+      "step": 1271
+    },
+    {
+      "epoch": 1.6502271252433485,
+      "grad_norm": 0.22166881997259968,
+      "learning_rate": 2.498797498797499e-05,
+      "loss": 0.4347,
+      "step": 1272
+    },
+    {
+      "epoch": 1.6515249837767683,
+      "grad_norm": 0.20648838786480744,
+      "learning_rate": 2.4963924963924963e-05,
+      "loss": 0.4236,
+      "step": 1273
+    },
+    {
+      "epoch": 1.6528228423101883,
+      "grad_norm": 0.23349839066379247,
+      "learning_rate": 2.493987493987494e-05,
+      "loss": 0.419,
+      "step": 1274
+    },
+    {
+      "epoch": 1.654120700843608,
+      "grad_norm": 0.23063394385414213,
+      "learning_rate": 2.4915824915824916e-05,
+      "loss": 0.4152,
+      "step": 1275
+    },
+    {
+      "epoch": 1.6554185593770279,
+      "grad_norm": 0.2190005315364852,
+      "learning_rate": 2.4891774891774894e-05,
+      "loss": 0.4115,
+      "step": 1276
+    },
+    {
+      "epoch": 1.6567164179104479,
+      "grad_norm": 0.20078160624348626,
+      "learning_rate": 2.4867724867724866e-05,
+      "loss": 0.4245,
+      "step": 1277
+    },
+    {
+      "epoch": 1.6580142764438677,
+      "grad_norm": 0.24133729159661466,
+      "learning_rate": 2.4843674843674844e-05,
+      "loss": 0.4293,
+      "step": 1278
+    },
+    {
+      "epoch": 1.6593121349772875,
+      "grad_norm": 0.23794916923086656,
+      "learning_rate": 2.481962481962482e-05,
+      "loss": 0.4271,
+      "step": 1279
+    },
+    {
+      "epoch": 1.6606099935107075,
+      "grad_norm": 0.2574981267536903,
+      "learning_rate": 2.4795574795574797e-05,
+      "loss": 0.447,
+      "step": 1280
+    },
+    {
+      "epoch": 1.661907852044127,
+      "grad_norm": 0.23168835516119046,
+      "learning_rate": 2.4771524771524772e-05,
+      "loss": 0.4193,
+      "step": 1281
+    },
+    {
+      "epoch": 1.663205710577547,
+      "grad_norm": 0.23742732882318857,
+      "learning_rate": 2.474747474747475e-05,
+      "loss": 0.4476,
+      "step": 1282
+    },
+    {
+      "epoch": 1.6645035691109669,
+      "grad_norm": 0.2514178670168895,
+      "learning_rate": 2.4723424723424725e-05,
+      "loss": 0.4204,
+      "step": 1283
+    },
+    {
+      "epoch": 1.6658014276443867,
+      "grad_norm": 0.24624362103838568,
+      "learning_rate": 2.46993746993747e-05,
+      "loss": 0.4212,
+      "step": 1284
+    },
+    {
+      "epoch": 1.6670992861778067,
+      "grad_norm": 0.23275179943867672,
+      "learning_rate": 2.4675324675324678e-05,
+      "loss": 0.4312,
+      "step": 1285
+    },
+    {
+      "epoch": 1.6683971447112265,
+      "grad_norm": 0.24951766353093746,
+      "learning_rate": 2.4651274651274652e-05,
+      "loss": 0.4068,
+      "step": 1286
+    },
+    {
+      "epoch": 1.6696950032446463,
+      "grad_norm": 0.2052390086188538,
+      "learning_rate": 2.462722462722463e-05,
+      "loss": 0.4173,
+      "step": 1287
+    },
+    {
+      "epoch": 1.6709928617780663,
+      "grad_norm": 0.23198703435419166,
+      "learning_rate": 2.4603174603174602e-05,
+      "loss": 0.4177,
+      "step": 1288
+    },
+    {
+      "epoch": 1.672290720311486,
+      "grad_norm": 0.212107274947473,
+      "learning_rate": 2.457912457912458e-05,
+      "loss": 0.4166,
+      "step": 1289
+    },
+    {
+      "epoch": 1.6735885788449059,
+      "grad_norm": 0.2322378778891487,
+      "learning_rate": 2.4555074555074555e-05,
+      "loss": 0.434,
+      "step": 1290
+    },
+    {
+      "epoch": 1.6748864373783259,
+      "grad_norm": 0.21435317963286998,
+      "learning_rate": 2.4531024531024533e-05,
+      "loss": 0.4133,
+      "step": 1291
+    },
+    {
+      "epoch": 1.6761842959117457,
+      "grad_norm": 0.2156907157084962,
+      "learning_rate": 2.4506974506974508e-05,
+      "loss": 0.4285,
+      "step": 1292
+    },
+    {
+      "epoch": 1.6774821544451655,
+      "grad_norm": 0.2568679217265247,
+      "learning_rate": 2.4482924482924486e-05,
+      "loss": 0.4134,
+      "step": 1293
+    },
+    {
+      "epoch": 1.6787800129785855,
+      "grad_norm": 0.23974841708540973,
+      "learning_rate": 2.4458874458874457e-05,
+      "loss": 0.432,
+      "step": 1294
+    },
+    {
+      "epoch": 1.680077871512005,
+      "grad_norm": 0.24031934038845462,
+      "learning_rate": 2.4434824434824436e-05,
+      "loss": 0.4444,
+      "step": 1295
+    },
+    {
+      "epoch": 1.681375730045425,
+      "grad_norm": 0.2672955058745279,
+      "learning_rate": 2.441077441077441e-05,
+      "loss": 0.4369,
+      "step": 1296
+    },
+    {
+      "epoch": 1.6826735885788449,
+      "grad_norm": 0.250150232180256,
+      "learning_rate": 2.438672438672439e-05,
+      "loss": 0.4273,
+      "step": 1297
+    },
+    {
+      "epoch": 1.6839714471122647,
+      "grad_norm": 0.23626717443698833,
+      "learning_rate": 2.4362674362674363e-05,
+      "loss": 0.4236,
+      "step": 1298
+    },
+    {
+      "epoch": 1.6852693056456847,
+      "grad_norm": 0.22422589742898313,
+      "learning_rate": 2.4338624338624338e-05,
+      "loss": 0.4338,
+      "step": 1299
+    },
+    {
+      "epoch": 1.6865671641791045,
+      "grad_norm": 0.2308927071463409,
+      "learning_rate": 2.4314574314574316e-05,
+      "loss": 0.4386,
+      "step": 1300
+    },
+    {
+      "epoch": 1.6878650227125243,
+      "grad_norm": 0.2564472802977678,
+      "learning_rate": 2.429052429052429e-05,
+      "loss": 0.4117,
+      "step": 1301
+    },
+    {
+      "epoch": 1.6891628812459443,
+      "grad_norm": 0.23392174813654176,
+      "learning_rate": 2.426647426647427e-05,
+      "loss": 0.4154,
+      "step": 1302
+    },
+    {
+      "epoch": 1.690460739779364,
+      "grad_norm": 0.2397393509201778,
+      "learning_rate": 2.4242424242424244e-05,
+      "loss": 0.4117,
+      "step": 1303
+    },
+    {
+      "epoch": 1.6917585983127839,
+      "grad_norm": 0.22912904331451653,
+      "learning_rate": 2.4218374218374222e-05,
+      "loss": 0.4273,
+      "step": 1304
+    },
+    {
+      "epoch": 1.6930564568462039,
+      "grad_norm": 0.23255973129828944,
+      "learning_rate": 2.4194324194324193e-05,
+      "loss": 0.4199,
+      "step": 1305
+    },
+    {
+      "epoch": 1.6943543153796237,
+      "grad_norm": 0.2459474867528304,
+      "learning_rate": 2.417027417027417e-05,
+      "loss": 0.3879,
+      "step": 1306
+    },
+    {
+      "epoch": 1.6956521739130435,
+      "grad_norm": 0.21055785182005404,
+      "learning_rate": 2.4146224146224146e-05,
+      "loss": 0.4284,
+      "step": 1307
+    },
+    {
+      "epoch": 1.6969500324464635,
+      "grad_norm": 0.23246957442423627,
+      "learning_rate": 2.4122174122174125e-05,
+      "loss": 0.4131,
+      "step": 1308
+    },
+    {
+      "epoch": 1.698247890979883,
+      "grad_norm": 0.23994403266599254,
+      "learning_rate": 2.40981240981241e-05,
+      "loss": 0.4333,
+      "step": 1309
+    },
+    {
+      "epoch": 1.699545749513303,
+      "grad_norm": 0.2316301617751929,
+      "learning_rate": 2.4074074074074074e-05,
+      "loss": 0.4486,
+      "step": 1310
+    },
+    {
+      "epoch": 1.7008436080467229,
+      "grad_norm": 0.22426960542300423,
+      "learning_rate": 2.405002405002405e-05,
+      "loss": 0.4087,
+      "step": 1311
+    },
+    {
+      "epoch": 1.7021414665801426,
+      "grad_norm": 0.2150721322470821,
+      "learning_rate": 2.4025974025974027e-05,
+      "loss": 0.4073,
+      "step": 1312
+    },
+    {
+      "epoch": 1.7034393251135627,
+      "grad_norm": 0.2519319561232589,
+      "learning_rate": 2.4001924001924002e-05,
+      "loss": 0.4312,
+      "step": 1313
+    },
+    {
+      "epoch": 1.7047371836469825,
+      "grad_norm": 0.21097482449181498,
+      "learning_rate": 2.397787397787398e-05,
+      "loss": 0.4319,
+      "step": 1314
+    },
+    {
+      "epoch": 1.7060350421804023,
+      "grad_norm": 0.24220719812817199,
+      "learning_rate": 2.3953823953823955e-05,
+      "loss": 0.4203,
+      "step": 1315
+    },
+    {
+      "epoch": 1.7073329007138223,
+      "grad_norm": 0.22812751540222995,
+      "learning_rate": 2.392977392977393e-05,
+      "loss": 0.4325,
+      "step": 1316
+    },
+    {
+      "epoch": 1.708630759247242,
+      "grad_norm": 0.21991172303538978,
+      "learning_rate": 2.3905723905723908e-05,
+      "loss": 0.4307,
+      "step": 1317
+    },
+    {
+      "epoch": 1.7099286177806619,
+      "grad_norm": 0.24924033452167085,
+      "learning_rate": 2.3881673881673882e-05,
+      "loss": 0.4049,
+      "step": 1318
+    },
+    {
+      "epoch": 1.7112264763140819,
+      "grad_norm": 0.20641915282626047,
+      "learning_rate": 2.385762385762386e-05,
+      "loss": 0.4171,
+      "step": 1319
+    },
+    {
+      "epoch": 1.7125243348475017,
+      "grad_norm": 0.24635070927670322,
+      "learning_rate": 2.3833573833573835e-05,
+      "loss": 0.4273,
+      "step": 1320
+    },
+    {
+      "epoch": 1.7138221933809215,
+      "grad_norm": 0.22793016083957607,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.4028,
+      "step": 1321
+    },
+    {
+      "epoch": 1.7151200519143415,
+      "grad_norm": 0.21686313598641418,
+      "learning_rate": 2.3785473785473785e-05,
+      "loss": 0.4216,
+      "step": 1322
+    },
+    {
+      "epoch": 1.716417910447761,
+      "grad_norm": 0.20935343473455395,
+      "learning_rate": 2.3761423761423763e-05,
+      "loss": 0.4002,
+      "step": 1323
+    },
+    {
+      "epoch": 1.717715768981181,
+      "grad_norm": 0.22891020801302248,
+      "learning_rate": 2.3737373737373738e-05,
+      "loss": 0.411,
+      "step": 1324
+    },
+    {
+      "epoch": 1.719013627514601,
+      "grad_norm": 0.20360937762091913,
+      "learning_rate": 2.3713323713323716e-05,
+      "loss": 0.4445,
+      "step": 1325
+    },
+    {
+      "epoch": 1.7203114860480206,
+      "grad_norm": 0.21662567258205914,
+      "learning_rate": 2.368927368927369e-05,
+      "loss": 0.4182,
+      "step": 1326
+    },
+    {
+      "epoch": 1.7216093445814407,
+      "grad_norm": 0.2112365052652544,
+      "learning_rate": 2.3665223665223666e-05,
+      "loss": 0.4119,
+      "step": 1327
+    },
+    {
+      "epoch": 1.7229072031148605,
+      "grad_norm": 0.24045539800451038,
+      "learning_rate": 2.364117364117364e-05,
+      "loss": 0.4294,
+      "step": 1328
+    },
+    {
+      "epoch": 1.7242050616482802,
+      "grad_norm": 0.22344351793204972,
+      "learning_rate": 2.361712361712362e-05,
+      "loss": 0.4159,
+      "step": 1329
+    },
+    {
+      "epoch": 1.7255029201817003,
+      "grad_norm": 0.21385701142195507,
+      "learning_rate": 2.3593073593073593e-05,
+      "loss": 0.4208,
+      "step": 1330
+    },
+    {
+      "epoch": 1.72680077871512,
+      "grad_norm": 0.22306282993754703,
+      "learning_rate": 2.356902356902357e-05,
+      "loss": 0.417,
+      "step": 1331
+    },
+    {
+      "epoch": 1.7280986372485398,
+      "grad_norm": 0.23450328976859844,
+      "learning_rate": 2.3544973544973546e-05,
+      "loss": 0.397,
+      "step": 1332
+    },
+    {
+      "epoch": 1.7293964957819599,
+      "grad_norm": 0.23314049600175976,
+      "learning_rate": 2.352092352092352e-05,
+      "loss": 0.4297,
+      "step": 1333
+    },
+    {
+      "epoch": 1.7306943543153797,
+      "grad_norm": 0.21399972133644776,
+      "learning_rate": 2.34968734968735e-05,
+      "loss": 0.4366,
+      "step": 1334
+    },
+    {
+      "epoch": 1.7319922128487995,
+      "grad_norm": 0.22426899700350952,
+      "learning_rate": 2.3472823472823474e-05,
+      "loss": 0.433,
+      "step": 1335
+    },
+    {
+      "epoch": 1.7332900713822195,
+      "grad_norm": 0.21720443031252623,
+      "learning_rate": 2.3448773448773452e-05,
+      "loss": 0.4135,
+      "step": 1336
+    },
+    {
+      "epoch": 1.734587929915639,
+      "grad_norm": 0.22022369229968872,
+      "learning_rate": 2.3424723424723427e-05,
+      "loss": 0.4035,
+      "step": 1337
+    },
+    {
+      "epoch": 1.735885788449059,
+      "grad_norm": 0.2432882987842844,
+      "learning_rate": 2.34006734006734e-05,
+      "loss": 0.4274,
+      "step": 1338
+    },
+    {
+      "epoch": 1.737183646982479,
+      "grad_norm": 0.22954645223280482,
+      "learning_rate": 2.3376623376623376e-05,
+      "loss": 0.4265,
+      "step": 1339
+    },
+    {
+      "epoch": 1.7384815055158986,
+      "grad_norm": 0.23456332298959323,
+      "learning_rate": 2.3352573352573355e-05,
+      "loss": 0.4261,
+      "step": 1340
+    },
+    {
+      "epoch": 1.7397793640493187,
+      "grad_norm": 0.23090513352220413,
+      "learning_rate": 2.332852332852333e-05,
+      "loss": 0.4343,
+      "step": 1341
+    },
+    {
+      "epoch": 1.7410772225827384,
+      "grad_norm": 0.22635007188997747,
+      "learning_rate": 2.3304473304473308e-05,
+      "loss": 0.434,
+      "step": 1342
+    },
+    {
+      "epoch": 1.7423750811161582,
+      "grad_norm": 0.24328716551223983,
+      "learning_rate": 2.328042328042328e-05,
+      "loss": 0.4329,
+      "step": 1343
+    },
+    {
+      "epoch": 1.7436729396495783,
+      "grad_norm": 0.2245296632717372,
+      "learning_rate": 2.3256373256373257e-05,
+      "loss": 0.4135,
+      "step": 1344
+    },
+    {
+      "epoch": 1.744970798182998,
+      "grad_norm": 0.23430249945874695,
+      "learning_rate": 2.3232323232323232e-05,
+      "loss": 0.4178,
+      "step": 1345
+    },
+    {
+      "epoch": 1.7462686567164178,
+      "grad_norm": 0.21397181948116892,
+      "learning_rate": 2.320827320827321e-05,
+      "loss": 0.4236,
+      "step": 1346
+    },
+    {
+      "epoch": 1.7475665152498379,
+      "grad_norm": 0.21776737560072357,
+      "learning_rate": 2.3184223184223185e-05,
+      "loss": 0.425,
+      "step": 1347
+    },
+    {
+      "epoch": 1.7488643737832577,
+      "grad_norm": 0.23739059292954565,
+      "learning_rate": 2.3160173160173163e-05,
+      "loss": 0.4262,
+      "step": 1348
+    },
+    {
+      "epoch": 1.7501622323166774,
+      "grad_norm": 0.2207747605074272,
+      "learning_rate": 2.3136123136123138e-05,
+      "loss": 0.4097,
+      "step": 1349
+    },
+    {
+      "epoch": 1.7514600908500975,
+      "grad_norm": 0.19291564676436485,
+      "learning_rate": 2.3112073112073113e-05,
+      "loss": 0.4222,
+      "step": 1350
+    },
+    {
+      "epoch": 1.752757949383517,
+      "grad_norm": 0.20208738954938904,
+      "learning_rate": 2.308802308802309e-05,
+      "loss": 0.4199,
+      "step": 1351
+    },
+    {
+      "epoch": 1.754055807916937,
+      "grad_norm": 0.22066527169458836,
+      "learning_rate": 2.3063973063973065e-05,
+      "loss": 0.426,
+      "step": 1352
+    },
+    {
+      "epoch": 1.755353666450357,
+      "grad_norm": 0.22615489279435733,
+      "learning_rate": 2.3039923039923044e-05,
+      "loss": 0.4103,
+      "step": 1353
+    },
+    {
+      "epoch": 1.7566515249837766,
+      "grad_norm": 0.24657435823356594,
+      "learning_rate": 2.3015873015873015e-05,
+      "loss": 0.4006,
+      "step": 1354
+    },
+    {
+      "epoch": 1.7579493835171967,
+      "grad_norm": 0.2287984884377898,
+      "learning_rate": 2.2991822991822993e-05,
+      "loss": 0.4481,
+      "step": 1355
+    },
+    {
+      "epoch": 1.7592472420506164,
+      "grad_norm": 0.21060281438071618,
+      "learning_rate": 2.2967772967772968e-05,
+      "loss": 0.399,
+      "step": 1356
+    },
+    {
+      "epoch": 1.7605451005840362,
+      "grad_norm": 0.2265028463512503,
+      "learning_rate": 2.2943722943722946e-05,
+      "loss": 0.4258,
+      "step": 1357
+    },
+    {
+      "epoch": 1.7618429591174563,
+      "grad_norm": 0.21362689950493072,
+      "learning_rate": 2.291967291967292e-05,
+      "loss": 0.4178,
+      "step": 1358
+    },
+    {
+      "epoch": 1.763140817650876,
+      "grad_norm": 0.21905319629937445,
+      "learning_rate": 2.28956228956229e-05,
+      "loss": 0.435,
+      "step": 1359
+    },
+    {
+      "epoch": 1.7644386761842958,
+      "grad_norm": 0.2390674190592371,
+      "learning_rate": 2.287157287157287e-05,
+      "loss": 0.4374,
+      "step": 1360
+    },
+    {
+      "epoch": 1.7657365347177159,
+      "grad_norm": 0.262047575243414,
+      "learning_rate": 2.284752284752285e-05,
+      "loss": 0.4249,
+      "step": 1361
+    },
+    {
+      "epoch": 1.7670343932511356,
+      "grad_norm": 0.215263905999125,
+      "learning_rate": 2.2823472823472823e-05,
+      "loss": 0.4358,
+      "step": 1362
+    },
+    {
+      "epoch": 1.7683322517845554,
+      "grad_norm": 0.2566335364538136,
+      "learning_rate": 2.27994227994228e-05,
+      "loss": 0.4358,
+      "step": 1363
+    },
+    {
+      "epoch": 1.7696301103179755,
+      "grad_norm": 0.23396394290955702,
+      "learning_rate": 2.2775372775372776e-05,
+      "loss": 0.4345,
+      "step": 1364
+    },
+    {
+      "epoch": 1.7709279688513953,
+      "grad_norm": 0.2997951835182788,
+      "learning_rate": 2.275132275132275e-05,
+      "loss": 0.417,
+      "step": 1365
+    },
+    {
+      "epoch": 1.772225827384815,
+      "grad_norm": 0.25908281100055225,
+      "learning_rate": 2.272727272727273e-05,
+      "loss": 0.4283,
+      "step": 1366
+    },
+    {
+      "epoch": 1.773523685918235,
+      "grad_norm": 0.2710663340672724,
+      "learning_rate": 2.2703222703222704e-05,
+      "loss": 0.4405,
+      "step": 1367
+    },
+    {
+      "epoch": 1.7748215444516546,
+      "grad_norm": 0.25367420601807966,
+      "learning_rate": 2.267917267917268e-05,
+      "loss": 0.4149,
+      "step": 1368
+    },
+    {
+      "epoch": 1.7761194029850746,
+      "grad_norm": 0.2872738552914004,
+      "learning_rate": 2.2655122655122657e-05,
+      "loss": 0.4339,
+      "step": 1369
+    },
+    {
+      "epoch": 1.7774172615184944,
+      "grad_norm": 0.21919850447737751,
+      "learning_rate": 2.2631072631072632e-05,
+      "loss": 0.4445,
+      "step": 1370
+    },
+    {
+      "epoch": 1.7787151200519142,
+      "grad_norm": 0.27892242525176375,
+      "learning_rate": 2.2607022607022607e-05,
+      "loss": 0.4131,
+      "step": 1371
+    },
+    {
+      "epoch": 1.7800129785853342,
+      "grad_norm": 0.24026730555070555,
+      "learning_rate": 2.2582972582972585e-05,
+      "loss": 0.427,
+      "step": 1372
+    },
+    {
+      "epoch": 1.781310837118754,
+      "grad_norm": 0.2145688542497997,
+      "learning_rate": 2.255892255892256e-05,
+      "loss": 0.4391,
+      "step": 1373
+    },
+    {
+      "epoch": 1.7826086956521738,
+      "grad_norm": 0.23661442477067585,
+      "learning_rate": 2.2534872534872538e-05,
+      "loss": 0.41,
+      "step": 1374
+    },
+    {
+      "epoch": 1.7839065541855939,
+      "grad_norm": 0.2610547392581578,
+      "learning_rate": 2.2510822510822512e-05,
+      "loss": 0.4188,
+      "step": 1375
+    },
+    {
+      "epoch": 1.7852044127190136,
+      "grad_norm": 0.2493506467256105,
+      "learning_rate": 2.2486772486772487e-05,
+      "loss": 0.4206,
+      "step": 1376
+    },
+    {
+      "epoch": 1.7865022712524334,
+      "grad_norm": 0.2649167628997299,
+      "learning_rate": 2.2462722462722462e-05,
+      "loss": 0.4237,
+      "step": 1377
+    },
+    {
+      "epoch": 1.7878001297858535,
+      "grad_norm": 0.2534895267108062,
+      "learning_rate": 2.243867243867244e-05,
+      "loss": 0.4283,
+      "step": 1378
+    },
+    {
+      "epoch": 1.7890979883192732,
+      "grad_norm": 0.2709616156112994,
+      "learning_rate": 2.2414622414622415e-05,
+      "loss": 0.4113,
+      "step": 1379
+    },
+    {
+      "epoch": 1.790395846852693,
+      "grad_norm": 0.22792963880042075,
+      "learning_rate": 2.2390572390572393e-05,
+      "loss": 0.4267,
+      "step": 1380
+    },
+    {
+      "epoch": 1.791693705386113,
+      "grad_norm": 0.24622268955355062,
+      "learning_rate": 2.2366522366522368e-05,
+      "loss": 0.4051,
+      "step": 1381
+    },
+    {
+      "epoch": 1.7929915639195326,
+      "grad_norm": 0.27590035842972194,
+      "learning_rate": 2.2342472342472343e-05,
+      "loss": 0.4378,
+      "step": 1382
+    },
+    {
+      "epoch": 1.7942894224529526,
+      "grad_norm": 0.2264860712514965,
+      "learning_rate": 2.2318422318422317e-05,
+      "loss": 0.4171,
+      "step": 1383
+    },
+    {
+      "epoch": 1.7955872809863724,
+      "grad_norm": 0.27527712703496315,
+      "learning_rate": 2.2294372294372296e-05,
+      "loss": 0.4136,
+      "step": 1384
+    },
+    {
+      "epoch": 1.7968851395197922,
+      "grad_norm": 0.27052531643386396,
+      "learning_rate": 2.227032227032227e-05,
+      "loss": 0.3935,
+      "step": 1385
+    },
+    {
+      "epoch": 1.7981829980532122,
+      "grad_norm": 0.2479444281803134,
+      "learning_rate": 2.224627224627225e-05,
+      "loss": 0.4331,
+      "step": 1386
+    },
+    {
+      "epoch": 1.799480856586632,
+      "grad_norm": 0.2373284631481721,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.417,
+      "step": 1387
+    },
+    {
+      "epoch": 1.8007787151200518,
+      "grad_norm": 0.2575638652052547,
+      "learning_rate": 2.2198172198172198e-05,
+      "loss": 0.4323,
+      "step": 1388
+    },
+    {
+      "epoch": 1.8020765736534718,
+      "grad_norm": 0.2407980171885747,
+      "learning_rate": 2.2174122174122176e-05,
+      "loss": 0.4127,
+      "step": 1389
+    },
+    {
+      "epoch": 1.8033744321868916,
+      "grad_norm": 0.21117036443387086,
+      "learning_rate": 2.215007215007215e-05,
+      "loss": 0.4191,
+      "step": 1390
+    },
+    {
+      "epoch": 1.8046722907203114,
+      "grad_norm": 0.20129164818193102,
+      "learning_rate": 2.212602212602213e-05,
+      "loss": 0.4115,
+      "step": 1391
+    },
+    {
+      "epoch": 1.8059701492537314,
+      "grad_norm": 0.2540001501490721,
+      "learning_rate": 2.2101972101972104e-05,
+      "loss": 0.431,
+      "step": 1392
+    },
+    {
+      "epoch": 1.8072680077871512,
+      "grad_norm": 0.24804686013462887,
+      "learning_rate": 2.207792207792208e-05,
+      "loss": 0.4228,
+      "step": 1393
+    },
+    {
+      "epoch": 1.808565866320571,
+      "grad_norm": 0.2022520818624456,
+      "learning_rate": 2.2053872053872053e-05,
+      "loss": 0.4203,
+      "step": 1394
+    },
+    {
+      "epoch": 1.809863724853991,
+      "grad_norm": 0.23238075295532062,
+      "learning_rate": 2.202982202982203e-05,
+      "loss": 0.425,
+      "step": 1395
+    },
+    {
+      "epoch": 1.8111615833874106,
+      "grad_norm": 0.21664815140422355,
+      "learning_rate": 2.2005772005772006e-05,
+      "loss": 0.4299,
+      "step": 1396
+    },
+    {
+      "epoch": 1.8124594419208306,
+      "grad_norm": 0.21744531033366538,
+      "learning_rate": 2.1981721981721985e-05,
+      "loss": 0.4019,
+      "step": 1397
+    },
+    {
+      "epoch": 1.8137573004542504,
+      "grad_norm": 0.2138946808987489,
+      "learning_rate": 2.1957671957671956e-05,
+      "loss": 0.4256,
+      "step": 1398
+    },
+    {
+      "epoch": 1.8150551589876702,
+      "grad_norm": 0.2224576706896047,
+      "learning_rate": 2.1933621933621934e-05,
+      "loss": 0.4147,
+      "step": 1399
+    },
+    {
+      "epoch": 1.8163530175210902,
+      "grad_norm": 0.2002966133486591,
+      "learning_rate": 2.190957190957191e-05,
+      "loss": 0.426,
+      "step": 1400
+    },
+    {
+      "epoch": 1.81765087605451,
+      "grad_norm": 0.202531126088113,
+      "learning_rate": 2.1885521885521887e-05,
+      "loss": 0.4429,
+      "step": 1401
+    },
+    {
+      "epoch": 1.8189487345879298,
+      "grad_norm": 0.20516498421820234,
+      "learning_rate": 2.1861471861471862e-05,
+      "loss": 0.4208,
+      "step": 1402
+    },
+    {
+      "epoch": 1.8202465931213498,
+      "grad_norm": 0.23389135616856488,
+      "learning_rate": 2.183742183742184e-05,
+      "loss": 0.4261,
+      "step": 1403
+    },
+    {
+      "epoch": 1.8215444516547696,
+      "grad_norm": 0.21459274849252136,
+      "learning_rate": 2.1813371813371815e-05,
+      "loss": 0.4346,
+      "step": 1404
+    },
+    {
+      "epoch": 1.8228423101881894,
+      "grad_norm": 0.2078147272516738,
+      "learning_rate": 2.178932178932179e-05,
+      "loss": 0.4092,
+      "step": 1405
+    },
+    {
+      "epoch": 1.8241401687216094,
+      "grad_norm": 0.24580373752808737,
+      "learning_rate": 2.1765271765271768e-05,
+      "loss": 0.4233,
+      "step": 1406
+    },
+    {
+      "epoch": 1.8254380272550292,
+      "grad_norm": 0.22454106978014404,
+      "learning_rate": 2.1741221741221743e-05,
+      "loss": 0.4147,
+      "step": 1407
+    },
+    {
+      "epoch": 1.826735885788449,
+      "grad_norm": 0.21599295500558674,
+      "learning_rate": 2.171717171717172e-05,
+      "loss": 0.4349,
+      "step": 1408
+    },
+    {
+      "epoch": 1.828033744321869,
+      "grad_norm": 0.2127033614651673,
+      "learning_rate": 2.1693121693121692e-05,
+      "loss": 0.4206,
+      "step": 1409
+    },
+    {
+      "epoch": 1.8293316028552886,
+      "grad_norm": 0.21925596786696352,
+      "learning_rate": 2.166907166907167e-05,
+      "loss": 0.4399,
+      "step": 1410
+    },
+    {
+      "epoch": 1.8306294613887086,
+      "grad_norm": 0.21016324905145667,
+      "learning_rate": 2.1645021645021645e-05,
+      "loss": 0.4222,
+      "step": 1411
+    },
+    {
+      "epoch": 1.8319273199221286,
+      "grad_norm": 0.22520381391920555,
+      "learning_rate": 2.1620971620971623e-05,
+      "loss": 0.4258,
+      "step": 1412
+    },
+    {
+      "epoch": 1.8332251784555482,
+      "grad_norm": 0.22141690462102792,
+      "learning_rate": 2.1596921596921598e-05,
+      "loss": 0.412,
+      "step": 1413
+    },
+    {
+      "epoch": 1.8345230369889682,
+      "grad_norm": 0.2429839281627191,
+      "learning_rate": 2.1572871572871576e-05,
+      "loss": 0.4269,
+      "step": 1414
+    },
+    {
+      "epoch": 1.835820895522388,
+      "grad_norm": 0.2160140354835784,
+      "learning_rate": 2.1548821548821547e-05,
+      "loss": 0.4205,
+      "step": 1415
+    },
+    {
+      "epoch": 1.8371187540558078,
+      "grad_norm": 0.2402260672982623,
+      "learning_rate": 2.1524771524771526e-05,
+      "loss": 0.4193,
+      "step": 1416
+    },
+    {
+      "epoch": 1.8384166125892278,
+      "grad_norm": 0.29744843810112265,
+      "learning_rate": 2.15007215007215e-05,
+      "loss": 0.4325,
+      "step": 1417
+    },
+    {
+      "epoch": 1.8397144711226476,
+      "grad_norm": 0.22530015703559994,
+      "learning_rate": 2.147667147667148e-05,
+      "loss": 0.4318,
+      "step": 1418
+    },
+    {
+      "epoch": 1.8410123296560674,
+      "grad_norm": 0.29390956909610316,
+      "learning_rate": 2.1452621452621453e-05,
+      "loss": 0.4288,
+      "step": 1419
+    },
+    {
+      "epoch": 1.8423101881894874,
+      "grad_norm": 0.23358318964698258,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 0.4084,
+      "step": 1420
+    },
+    {
+      "epoch": 1.8436080467229072,
+      "grad_norm": 0.21167664114993395,
+      "learning_rate": 2.1404521404521406e-05,
+      "loss": 0.4313,
+      "step": 1421
+    },
+    {
+      "epoch": 1.844905905256327,
+      "grad_norm": 0.2460698519801602,
+      "learning_rate": 2.138047138047138e-05,
+      "loss": 0.4155,
+      "step": 1422
+    },
+    {
+      "epoch": 1.846203763789747,
+      "grad_norm": 0.23025941782631765,
+      "learning_rate": 2.135642135642136e-05,
+      "loss": 0.4057,
+      "step": 1423
+    },
+    {
+      "epoch": 1.8475016223231666,
+      "grad_norm": 0.19865359681586736,
+      "learning_rate": 2.1332371332371334e-05,
+      "loss": 0.4075,
+      "step": 1424
+    },
+    {
+      "epoch": 1.8487994808565866,
+      "grad_norm": 0.22150167838157933,
+      "learning_rate": 2.1308321308321312e-05,
+      "loss": 0.4338,
+      "step": 1425
+    },
+    {
+      "epoch": 1.8500973393900066,
+      "grad_norm": 0.27381218064289997,
+      "learning_rate": 2.1284271284271284e-05,
+      "loss": 0.4385,
+      "step": 1426
+    },
+    {
+      "epoch": 1.8513951979234262,
+      "grad_norm": 0.2386126810899565,
+      "learning_rate": 2.1260221260221262e-05,
+      "loss": 0.4138,
+      "step": 1427
+    },
+    {
+      "epoch": 1.8526930564568462,
+      "grad_norm": 0.23844253499070778,
+      "learning_rate": 2.1236171236171237e-05,
+      "loss": 0.417,
+      "step": 1428
+    },
+    {
+      "epoch": 1.853990914990266,
+      "grad_norm": 0.24734871284649604,
+      "learning_rate": 2.1212121212121215e-05,
+      "loss": 0.4266,
+      "step": 1429
+    },
+    {
+      "epoch": 1.8552887735236858,
+      "grad_norm": 0.2581372866509555,
+      "learning_rate": 2.118807118807119e-05,
+      "loss": 0.4073,
+      "step": 1430
+    },
+    {
+      "epoch": 1.8565866320571058,
+      "grad_norm": 0.20591243236055737,
+      "learning_rate": 2.1164021164021164e-05,
+      "loss": 0.3973,
+      "step": 1431
+    },
+    {
+      "epoch": 1.8578844905905256,
+      "grad_norm": 0.25393718244850216,
+      "learning_rate": 2.113997113997114e-05,
+      "loss": 0.4237,
+      "step": 1432
+    },
+    {
+      "epoch": 1.8591823491239454,
+      "grad_norm": 0.256757051595813,
+      "learning_rate": 2.1115921115921117e-05,
+      "loss": 0.4276,
+      "step": 1433
+    },
+    {
+      "epoch": 1.8604802076573654,
+      "grad_norm": 0.2199746107316156,
+      "learning_rate": 2.1091871091871092e-05,
+      "loss": 0.4027,
+      "step": 1434
+    },
+    {
+      "epoch": 1.8617780661907852,
+      "grad_norm": 0.22993418151409517,
+      "learning_rate": 2.106782106782107e-05,
+      "loss": 0.4258,
+      "step": 1435
+    },
+    {
+      "epoch": 1.863075924724205,
+      "grad_norm": 0.23986794245337564,
+      "learning_rate": 2.1043771043771045e-05,
+      "loss": 0.4092,
+      "step": 1436
+    },
+    {
+      "epoch": 1.864373783257625,
+      "grad_norm": 0.2503767269878855,
+      "learning_rate": 2.101972101972102e-05,
+      "loss": 0.4337,
+      "step": 1437
+    },
+    {
+      "epoch": 1.8656716417910446,
+      "grad_norm": 0.19966379931345576,
+      "learning_rate": 2.0995670995670998e-05,
+      "loss": 0.4083,
+      "step": 1438
+    },
+    {
+      "epoch": 1.8669695003244646,
+      "grad_norm": 0.22975695557758422,
+      "learning_rate": 2.0971620971620973e-05,
+      "loss": 0.4155,
+      "step": 1439
+    },
+    {
+      "epoch": 1.8682673588578846,
+      "grad_norm": 0.26927614268096606,
+      "learning_rate": 2.094757094757095e-05,
+      "loss": 0.3885,
+      "step": 1440
+    },
+    {
+      "epoch": 1.8695652173913042,
+      "grad_norm": 0.20373734252329936,
+      "learning_rate": 2.0923520923520926e-05,
+      "loss": 0.4291,
+      "step": 1441
+    },
+    {
+      "epoch": 1.8708630759247242,
+      "grad_norm": 0.2683958306016899,
+      "learning_rate": 2.08994708994709e-05,
+      "loss": 0.433,
+      "step": 1442
+    },
+    {
+      "epoch": 1.872160934458144,
+      "grad_norm": 0.26568578858883407,
+      "learning_rate": 2.0875420875420875e-05,
+      "loss": 0.4379,
+      "step": 1443
+    },
+    {
+      "epoch": 1.8734587929915638,
+      "grad_norm": 0.2365065973899857,
+      "learning_rate": 2.0851370851370853e-05,
+      "loss": 0.4372,
+      "step": 1444
+    },
+    {
+      "epoch": 1.8747566515249838,
+      "grad_norm": 0.2160536365282337,
+      "learning_rate": 2.0827320827320828e-05,
+      "loss": 0.4249,
+      "step": 1445
+    },
+    {
+      "epoch": 1.8760545100584036,
+      "grad_norm": 0.2698594967367338,
+      "learning_rate": 2.0803270803270806e-05,
+      "loss": 0.4251,
+      "step": 1446
+    },
+    {
+      "epoch": 1.8773523685918234,
+      "grad_norm": 0.240476141319818,
+      "learning_rate": 2.077922077922078e-05,
+      "loss": 0.4051,
+      "step": 1447
+    },
+    {
+      "epoch": 1.8786502271252434,
+      "grad_norm": 0.20313150197250998,
+      "learning_rate": 2.0755170755170756e-05,
+      "loss": 0.4142,
+      "step": 1448
+    },
+    {
+      "epoch": 1.8799480856586632,
+      "grad_norm": 0.2513888218859537,
+      "learning_rate": 2.073112073112073e-05,
+      "loss": 0.4235,
+      "step": 1449
+    },
+    {
+      "epoch": 1.881245944192083,
+      "grad_norm": 0.263020254508393,
+      "learning_rate": 2.070707070707071e-05,
+      "loss": 0.4133,
+      "step": 1450
+    },
+    {
+      "epoch": 1.882543802725503,
+      "grad_norm": 0.19807928758761542,
+      "learning_rate": 2.0683020683020683e-05,
+      "loss": 0.4211,
+      "step": 1451
+    },
+    {
+      "epoch": 1.8838416612589228,
+      "grad_norm": 0.28553479995616016,
+      "learning_rate": 2.065897065897066e-05,
+      "loss": 0.4295,
+      "step": 1452
+    },
+    {
+      "epoch": 1.8851395197923426,
+      "grad_norm": 0.22454512768715873,
+      "learning_rate": 2.0634920634920636e-05,
+      "loss": 0.4164,
+      "step": 1453
+    },
+    {
+      "epoch": 1.8864373783257626,
+      "grad_norm": 0.2212283630425153,
+      "learning_rate": 2.061087061087061e-05,
+      "loss": 0.4309,
+      "step": 1454
+    },
+    {
+      "epoch": 1.8877352368591822,
+      "grad_norm": 0.21880750363041376,
+      "learning_rate": 2.058682058682059e-05,
+      "loss": 0.4148,
+      "step": 1455
+    },
+    {
+      "epoch": 1.8890330953926022,
+      "grad_norm": 0.2698709703952382,
+      "learning_rate": 2.0562770562770564e-05,
+      "loss": 0.4339,
+      "step": 1456
+    },
+    {
+      "epoch": 1.890330953926022,
+      "grad_norm": 0.21631366892137663,
+      "learning_rate": 2.0538720538720542e-05,
+      "loss": 0.4209,
+      "step": 1457
+    },
+    {
+      "epoch": 1.8916288124594418,
+      "grad_norm": 0.22312561756649457,
+      "learning_rate": 2.0514670514670517e-05,
+      "loss": 0.4205,
+      "step": 1458
+    },
+    {
+      "epoch": 1.8929266709928618,
+      "grad_norm": 0.22982817420831553,
+      "learning_rate": 2.0490620490620492e-05,
+      "loss": 0.4127,
+      "step": 1459
+    },
+    {
+      "epoch": 1.8942245295262816,
+      "grad_norm": 0.23011803773822845,
+      "learning_rate": 2.0466570466570467e-05,
+      "loss": 0.4018,
+      "step": 1460
+    },
+    {
+      "epoch": 1.8955223880597014,
+      "grad_norm": 0.2156774757909124,
+      "learning_rate": 2.0442520442520445e-05,
+      "loss": 0.448,
+      "step": 1461
+    },
+    {
+      "epoch": 1.8968202465931214,
+      "grad_norm": 0.2465313564522942,
+      "learning_rate": 2.041847041847042e-05,
+      "loss": 0.4433,
+      "step": 1462
+    },
+    {
+      "epoch": 1.8981181051265412,
+      "grad_norm": 0.21906089100592563,
+      "learning_rate": 2.0394420394420398e-05,
+      "loss": 0.4103,
+      "step": 1463
+    },
+    {
+      "epoch": 1.899415963659961,
+      "grad_norm": 0.21024875722994074,
+      "learning_rate": 2.037037037037037e-05,
+      "loss": 0.3951,
+      "step": 1464
+    },
+    {
+      "epoch": 1.900713822193381,
+      "grad_norm": 0.18955824822069273,
+      "learning_rate": 2.0346320346320347e-05,
+      "loss": 0.4174,
+      "step": 1465
+    },
+    {
+      "epoch": 1.9020116807268008,
+      "grad_norm": 0.23144049792472646,
+      "learning_rate": 2.0322270322270322e-05,
+      "loss": 0.3998,
+      "step": 1466
+    },
+    {
+      "epoch": 1.9033095392602206,
+      "grad_norm": 0.21081067517865779,
+      "learning_rate": 2.02982202982203e-05,
+      "loss": 0.4103,
+      "step": 1467
+    },
+    {
+      "epoch": 1.9046073977936406,
+      "grad_norm": 0.2194010873045385,
+      "learning_rate": 2.0274170274170275e-05,
+      "loss": 0.4123,
+      "step": 1468
+    },
+    {
+      "epoch": 1.9059052563270602,
+      "grad_norm": 0.20757690294910305,
+      "learning_rate": 2.025012025012025e-05,
+      "loss": 0.4012,
+      "step": 1469
+    },
+    {
+      "epoch": 1.9072031148604802,
+      "grad_norm": 0.2628488845364361,
+      "learning_rate": 2.0226070226070225e-05,
+      "loss": 0.4118,
+      "step": 1470
+    },
+    {
+      "epoch": 1.9085009733939,
+      "grad_norm": 0.2503984380267546,
+      "learning_rate": 2.0202020202020203e-05,
+      "loss": 0.4277,
+      "step": 1471
+    },
+    {
+      "epoch": 1.9097988319273198,
+      "grad_norm": 0.20731051341055765,
+      "learning_rate": 2.0177970177970177e-05,
+      "loss": 0.4223,
+      "step": 1472
+    },
+    {
+      "epoch": 1.9110966904607398,
+      "grad_norm": 0.2469892137078129,
+      "learning_rate": 2.0153920153920156e-05,
+      "loss": 0.4278,
+      "step": 1473
+    },
+    {
+      "epoch": 1.9123945489941596,
+      "grad_norm": 0.2302509090632293,
+      "learning_rate": 2.012987012987013e-05,
+      "loss": 0.4286,
+      "step": 1474
+    },
+    {
+      "epoch": 1.9136924075275794,
+      "grad_norm": 0.21560820581873713,
+      "learning_rate": 2.0105820105820105e-05,
+      "loss": 0.4036,
+      "step": 1475
+    },
+    {
+      "epoch": 1.9149902660609994,
+      "grad_norm": 0.21761526673837062,
+      "learning_rate": 2.0081770081770083e-05,
+      "loss": 0.4383,
+      "step": 1476
+    },
+    {
+      "epoch": 1.9162881245944192,
+      "grad_norm": 0.25419859148323953,
+      "learning_rate": 2.0057720057720058e-05,
+      "loss": 0.4168,
+      "step": 1477
+    },
+    {
+      "epoch": 1.917585983127839,
+      "grad_norm": 0.21447148417291215,
+      "learning_rate": 2.0033670033670036e-05,
+      "loss": 0.4233,
+      "step": 1478
+    },
+    {
+      "epoch": 1.918883841661259,
+      "grad_norm": 0.22177181102304355,
+      "learning_rate": 2.000962000962001e-05,
+      "loss": 0.3973,
+      "step": 1479
+    },
+    {
+      "epoch": 1.9201817001946788,
+      "grad_norm": 0.2361964777550035,
+      "learning_rate": 1.9985569985569986e-05,
+      "loss": 0.4062,
+      "step": 1480
+    },
+    {
+      "epoch": 1.9214795587280986,
+      "grad_norm": 0.2268625180335479,
+      "learning_rate": 1.996151996151996e-05,
+      "loss": 0.4288,
+      "step": 1481
+    },
+    {
+      "epoch": 1.9227774172615186,
+      "grad_norm": 0.22109891172640594,
+      "learning_rate": 1.993746993746994e-05,
+      "loss": 0.4168,
+      "step": 1482
+    },
+    {
+      "epoch": 1.9240752757949382,
+      "grad_norm": 0.24199558362942594,
+      "learning_rate": 1.9913419913419914e-05,
+      "loss": 0.4368,
+      "step": 1483
+    },
+    {
+      "epoch": 1.9253731343283582,
+      "grad_norm": 0.23386196057480746,
+      "learning_rate": 1.9889369889369892e-05,
+      "loss": 0.4929,
+      "step": 1484
+    },
+    {
+      "epoch": 1.9266709928617782,
+      "grad_norm": 2.3863223206635507,
+      "learning_rate": 1.9865319865319866e-05,
+      "loss": 0.4321,
+      "step": 1485
+    },
+    {
+      "epoch": 1.9279688513951978,
+      "grad_norm": 0.2117744722116347,
+      "learning_rate": 1.984126984126984e-05,
+      "loss": 0.4046,
+      "step": 1486
+    },
+    {
+      "epoch": 1.9292667099286178,
+      "grad_norm": 0.23753639694866985,
+      "learning_rate": 1.9817219817219816e-05,
+      "loss": 0.4088,
+      "step": 1487
+    },
+    {
+      "epoch": 1.9305645684620376,
+      "grad_norm": 0.2029549567060751,
+      "learning_rate": 1.9793169793169794e-05,
+      "loss": 0.4129,
+      "step": 1488
+    },
+    {
+      "epoch": 1.9318624269954574,
+      "grad_norm": 0.20999056789664505,
+      "learning_rate": 1.976911976911977e-05,
+      "loss": 0.4149,
+      "step": 1489
+    },
+    {
+      "epoch": 1.9331602855288774,
+      "grad_norm": 0.25609740431868805,
+      "learning_rate": 1.9745069745069747e-05,
+      "loss": 0.4422,
+      "step": 1490
+    },
+    {
+      "epoch": 1.9344581440622972,
+      "grad_norm": 0.23240223312760538,
+      "learning_rate": 1.9721019721019722e-05,
+      "loss": 0.4376,
+      "step": 1491
+    },
+    {
+      "epoch": 1.935756002595717,
+      "grad_norm": 0.22288941915151747,
+      "learning_rate": 1.9696969696969697e-05,
+      "loss": 0.4031,
+      "step": 1492
+    },
+    {
+      "epoch": 1.937053861129137,
+      "grad_norm": 0.20823811668022293,
+      "learning_rate": 1.9672919672919675e-05,
+      "loss": 0.4141,
+      "step": 1493
+    },
+    {
+      "epoch": 1.9383517196625568,
+      "grad_norm": 0.22958571482808876,
+      "learning_rate": 1.964886964886965e-05,
+      "loss": 0.4205,
+      "step": 1494
+    },
+    {
+      "epoch": 1.9396495781959766,
+      "grad_norm": 0.24269527070284858,
+      "learning_rate": 1.9624819624819628e-05,
+      "loss": 0.414,
+      "step": 1495
+    },
+    {
+      "epoch": 1.9409474367293966,
+      "grad_norm": 0.2070554761070819,
+      "learning_rate": 1.9600769600769603e-05,
+      "loss": 0.4045,
+      "step": 1496
+    },
+    {
+      "epoch": 1.9422452952628162,
+      "grad_norm": 0.24376293095622897,
+      "learning_rate": 1.9576719576719577e-05,
+      "loss": 0.4257,
+      "step": 1497
+    },
+    {
+      "epoch": 1.9435431537962362,
+      "grad_norm": 0.2254498669697948,
+      "learning_rate": 1.9552669552669552e-05,
+      "loss": 0.421,
+      "step": 1498
+    },
+    {
+      "epoch": 1.9448410123296562,
+      "grad_norm": 0.21748513808130843,
+      "learning_rate": 1.952861952861953e-05,
+      "loss": 0.4062,
+      "step": 1499
+    },
+    {
+      "epoch": 1.9461388708630758,
+      "grad_norm": 0.2148376810996354,
+      "learning_rate": 1.9504569504569505e-05,
+      "loss": 0.4203,
+      "step": 1500
+    },
+    {
+      "epoch": 1.9474367293964958,
+      "grad_norm": 0.25871259714383205,
+      "learning_rate": 1.9480519480519483e-05,
+      "loss": 0.4229,
+      "step": 1501
+    },
+    {
+      "epoch": 1.9487345879299156,
+      "grad_norm": 0.19582347887373358,
+      "learning_rate": 1.9456469456469455e-05,
+      "loss": 0.4081,
+      "step": 1502
+    },
+    {
+      "epoch": 1.9500324464633354,
+      "grad_norm": 0.22789399470009464,
+      "learning_rate": 1.9432419432419433e-05,
+      "loss": 0.4245,
+      "step": 1503
+    },
+    {
+      "epoch": 1.9513303049967554,
+      "grad_norm": 0.23018173092515049,
+      "learning_rate": 1.9408369408369408e-05,
+      "loss": 0.4216,
+      "step": 1504
+    },
+    {
+      "epoch": 1.9526281635301752,
+      "grad_norm": 0.21444832133823605,
+      "learning_rate": 1.9384319384319386e-05,
+      "loss": 0.4147,
+      "step": 1505
+    },
+    {
+      "epoch": 1.953926022063595,
+      "grad_norm": 0.20562647584839736,
+      "learning_rate": 1.936026936026936e-05,
+      "loss": 0.4008,
+      "step": 1506
+    },
+    {
+      "epoch": 1.955223880597015,
+      "grad_norm": 0.23110467483063488,
+      "learning_rate": 1.933621933621934e-05,
+      "loss": 0.3972,
+      "step": 1507
+    },
+    {
+      "epoch": 1.9565217391304348,
+      "grad_norm": 0.21834150518213843,
+      "learning_rate": 1.9312169312169313e-05,
+      "loss": 0.4256,
+      "step": 1508
+    },
+    {
+      "epoch": 1.9578195976638546,
+      "grad_norm": 0.21958450668275112,
+      "learning_rate": 1.9288119288119288e-05,
+      "loss": 0.4088,
+      "step": 1509
+    },
+    {
+      "epoch": 1.9591174561972746,
+      "grad_norm": 0.20052094185224426,
+      "learning_rate": 1.9264069264069266e-05,
+      "loss": 0.4121,
+      "step": 1510
+    },
+    {
+      "epoch": 1.9604153147306942,
+      "grad_norm": 0.24326880305407378,
+      "learning_rate": 1.924001924001924e-05,
+      "loss": 0.422,
+      "step": 1511
+    },
+    {
+      "epoch": 1.9617131732641142,
+      "grad_norm": 0.2410106190975958,
+      "learning_rate": 1.921596921596922e-05,
+      "loss": 0.4085,
+      "step": 1512
+    },
+    {
+      "epoch": 1.9630110317975342,
+      "grad_norm": 0.20377491892233185,
+      "learning_rate": 1.919191919191919e-05,
+      "loss": 0.4312,
+      "step": 1513
+    },
+    {
+      "epoch": 1.9643088903309538,
+      "grad_norm": 0.22992091739225845,
+      "learning_rate": 1.916786916786917e-05,
+      "loss": 0.4283,
+      "step": 1514
+    },
+    {
+      "epoch": 1.9656067488643738,
+      "grad_norm": 0.23320180740415136,
+      "learning_rate": 1.9143819143819144e-05,
+      "loss": 0.4167,
+      "step": 1515
+    },
+    {
+      "epoch": 1.9669046073977936,
+      "grad_norm": 0.21478096347520134,
+      "learning_rate": 1.9119769119769122e-05,
+      "loss": 0.4373,
+      "step": 1516
+    },
+    {
+      "epoch": 1.9682024659312134,
+      "grad_norm": 0.24312143244424492,
+      "learning_rate": 1.9095719095719097e-05,
+      "loss": 0.4384,
+      "step": 1517
+    },
+    {
+      "epoch": 1.9695003244646334,
+      "grad_norm": 0.22013684407762923,
+      "learning_rate": 1.9071669071669075e-05,
+      "loss": 0.4244,
+      "step": 1518
+    },
+    {
+      "epoch": 1.9707981829980532,
+      "grad_norm": 0.22394887240003014,
+      "learning_rate": 1.9047619047619046e-05,
+      "loss": 0.419,
+      "step": 1519
+    },
+    {
+      "epoch": 1.972096041531473,
+      "grad_norm": 0.2319362635386066,
+      "learning_rate": 1.9023569023569024e-05,
+      "loss": 0.4059,
+      "step": 1520
+    },
+    {
+      "epoch": 1.973393900064893,
+      "grad_norm": 0.1980072863895625,
+      "learning_rate": 1.8999518999519e-05,
+      "loss": 0.4012,
+      "step": 1521
+    },
+    {
+      "epoch": 1.9746917585983128,
+      "grad_norm": 0.21340412052310542,
+      "learning_rate": 1.8975468975468977e-05,
+      "loss": 0.4252,
+      "step": 1522
+    },
+    {
+      "epoch": 1.9759896171317326,
+      "grad_norm": 0.20523875562201788,
+      "learning_rate": 1.8951418951418952e-05,
+      "loss": 0.3954,
+      "step": 1523
+    },
+    {
+      "epoch": 1.9772874756651526,
+      "grad_norm": 0.21593958433489607,
+      "learning_rate": 1.8927368927368927e-05,
+      "loss": 0.4204,
+      "step": 1524
+    },
+    {
+      "epoch": 1.9785853341985724,
+      "grad_norm": 0.1982941991349422,
+      "learning_rate": 1.8903318903318905e-05,
+      "loss": 0.3982,
+      "step": 1525
+    },
+    {
+      "epoch": 1.9798831927319922,
+      "grad_norm": 0.20466190375196575,
+      "learning_rate": 1.887926887926888e-05,
+      "loss": 0.4134,
+      "step": 1526
+    },
+    {
+      "epoch": 1.9811810512654122,
+      "grad_norm": 0.21442210205444864,
+      "learning_rate": 1.8855218855218858e-05,
+      "loss": 0.4221,
+      "step": 1527
+    },
+    {
+      "epoch": 1.9824789097988318,
+      "grad_norm": 0.2077434816627499,
+      "learning_rate": 1.8831168831168833e-05,
+      "loss": 0.4168,
+      "step": 1528
+    },
+    {
+      "epoch": 1.9837767683322518,
+      "grad_norm": 0.20932779718622976,
+      "learning_rate": 1.880711880711881e-05,
+      "loss": 0.4129,
+      "step": 1529
+    },
+    {
+      "epoch": 1.9850746268656716,
+      "grad_norm": 0.20073142812922465,
+      "learning_rate": 1.8783068783068782e-05,
+      "loss": 0.4282,
+      "step": 1530
+    },
+    {
+      "epoch": 1.9863724853990914,
+      "grad_norm": 0.21070295646641607,
+      "learning_rate": 1.875901875901876e-05,
+      "loss": 0.414,
+      "step": 1531
+    },
+    {
+      "epoch": 1.9876703439325114,
+      "grad_norm": 0.1983254407503139,
+      "learning_rate": 1.8734968734968735e-05,
+      "loss": 0.4246,
+      "step": 1532
+    },
+    {
+      "epoch": 1.9889682024659312,
+      "grad_norm": 0.2063440750783136,
+      "learning_rate": 1.8710918710918713e-05,
+      "loss": 0.4127,
+      "step": 1533
+    },
+    {
+      "epoch": 1.990266060999351,
+      "grad_norm": 0.20062950578015543,
+      "learning_rate": 1.8686868686868688e-05,
+      "loss": 0.404,
+      "step": 1534
+    },
+    {
+      "epoch": 1.991563919532771,
+      "grad_norm": 0.22191712136507424,
+      "learning_rate": 1.8662818662818663e-05,
+      "loss": 0.4119,
+      "step": 1535
+    },
+    {
+      "epoch": 1.9928617780661908,
+      "grad_norm": 0.2215336165604822,
+      "learning_rate": 1.8638768638768638e-05,
+      "loss": 0.4121,
+      "step": 1536
+    },
+    {
+      "epoch": 1.9941596365996106,
+      "grad_norm": 0.20271253230410582,
+      "learning_rate": 1.8614718614718616e-05,
+      "loss": 0.4023,
+      "step": 1537
+    },
+    {
+      "epoch": 1.9954574951330306,
+      "grad_norm": 0.26159702346568764,
+      "learning_rate": 1.859066859066859e-05,
+      "loss": 0.4102,
+      "step": 1538
+    },
+    {
+      "epoch": 1.9967553536664504,
+      "grad_norm": 0.21830457585192162,
+      "learning_rate": 1.856661856661857e-05,
+      "loss": 0.4129,
+      "step": 1539
+    },
+    {
+      "epoch": 1.9980532121998702,
+      "grad_norm": 0.21634003518886286,
+      "learning_rate": 1.8542568542568544e-05,
+      "loss": 0.3979,
+      "step": 1540
+    },
+    {
+      "epoch": 1.9993510707332902,
+      "grad_norm": 0.21520818627840688,
+      "learning_rate": 1.8518518518518518e-05,
+      "loss": 0.4072,
+      "step": 1541
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3162004070666462,
+      "learning_rate": 1.8494468494468496e-05,
+      "loss": 0.3731,
+      "step": 1542
+    },
+    {
+      "epoch": 2.00129785853342,
+      "grad_norm": 0.3002746540509754,
+      "learning_rate": 1.847041847041847e-05,
+      "loss": 0.351,
+      "step": 1543
+    },
+    {
+      "epoch": 2.0025957170668396,
+      "grad_norm": 0.22179389161987628,
+      "learning_rate": 1.844636844636845e-05,
+      "loss": 0.3591,
+      "step": 1544
+    },
+    {
+      "epoch": 2.0038935756002596,
+      "grad_norm": 0.29853309659990307,
+      "learning_rate": 1.8422318422318424e-05,
+      "loss": 0.3502,
+      "step": 1545
+    },
+    {
+      "epoch": 2.0051914341336796,
+      "grad_norm": 0.27887815631774426,
+      "learning_rate": 1.83982683982684e-05,
+      "loss": 0.3397,
+      "step": 1546
+    },
+    {
+      "epoch": 2.006489292667099,
+      "grad_norm": 0.21973812049478386,
+      "learning_rate": 1.8374218374218374e-05,
+      "loss": 0.3429,
+      "step": 1547
+    },
+    {
+      "epoch": 2.007787151200519,
+      "grad_norm": 0.26933885808676494,
+      "learning_rate": 1.8350168350168352e-05,
+      "loss": 0.3436,
+      "step": 1548
+    },
+    {
+      "epoch": 2.009085009733939,
+      "grad_norm": 0.2996171251030684,
+      "learning_rate": 1.8326118326118327e-05,
+      "loss": 0.3409,
+      "step": 1549
+    },
+    {
+      "epoch": 2.010382868267359,
+      "grad_norm": 0.24083124258461439,
+      "learning_rate": 1.8302068302068305e-05,
+      "loss": 0.3386,
+      "step": 1550
+    },
+    {
+      "epoch": 2.011680726800779,
+      "grad_norm": 0.30787012297971555,
+      "learning_rate": 1.827801827801828e-05,
+      "loss": 0.3478,
+      "step": 1551
+    },
+    {
+      "epoch": 2.0129785853341984,
+      "grad_norm": 0.2857849577396285,
+      "learning_rate": 1.8253968253968254e-05,
+      "loss": 0.347,
+      "step": 1552
+    },
+    {
+      "epoch": 2.0142764438676184,
+      "grad_norm": 0.2458814691129703,
+      "learning_rate": 1.822991822991823e-05,
+      "loss": 0.338,
+      "step": 1553
+    },
+    {
+      "epoch": 2.0155743024010384,
+      "grad_norm": 0.24217238914022393,
+      "learning_rate": 1.8205868205868207e-05,
+      "loss": 0.3527,
+      "step": 1554
+    },
+    {
+      "epoch": 2.016872160934458,
+      "grad_norm": 0.2708034381508514,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 0.343,
+      "step": 1555
+    },
+    {
+      "epoch": 2.018170019467878,
+      "grad_norm": 0.2771300303365467,
+      "learning_rate": 1.815776815776816e-05,
+      "loss": 0.3578,
+      "step": 1556
+    },
+    {
+      "epoch": 2.019467878001298,
+      "grad_norm": 0.24314761286723605,
+      "learning_rate": 1.8133718133718135e-05,
+      "loss": 0.3628,
+      "step": 1557
+    },
+    {
+      "epoch": 2.0207657365347176,
+      "grad_norm": 0.24528635873262158,
+      "learning_rate": 1.810966810966811e-05,
+      "loss": 0.3478,
+      "step": 1558
+    },
+    {
+      "epoch": 2.0220635950681376,
+      "grad_norm": 0.2441718287388452,
+      "learning_rate": 1.8085618085618085e-05,
+      "loss": 0.3463,
+      "step": 1559
+    },
+    {
+      "epoch": 2.0233614536015576,
+      "grad_norm": 0.2429476875932917,
+      "learning_rate": 1.8061568061568063e-05,
+      "loss": 0.3533,
+      "step": 1560
+    },
+    {
+      "epoch": 2.024659312134977,
+      "grad_norm": 0.2584288721917747,
+      "learning_rate": 1.8037518037518038e-05,
+      "loss": 0.3443,
+      "step": 1561
+    },
+    {
+      "epoch": 2.025957170668397,
+      "grad_norm": 0.21619552093380776,
+      "learning_rate": 1.8013468013468016e-05,
+      "loss": 0.3313,
+      "step": 1562
+    },
+    {
+      "epoch": 2.027255029201817,
+      "grad_norm": 0.22416123032351348,
+      "learning_rate": 1.798941798941799e-05,
+      "loss": 0.3396,
+      "step": 1563
+    },
+    {
+      "epoch": 2.028552887735237,
+      "grad_norm": 0.23762867221262482,
+      "learning_rate": 1.7965367965367965e-05,
+      "loss": 0.3328,
+      "step": 1564
+    },
+    {
+      "epoch": 2.029850746268657,
+      "grad_norm": 0.23910726203342522,
+      "learning_rate": 1.7941317941317943e-05,
+      "loss": 0.3373,
+      "step": 1565
+    },
+    {
+      "epoch": 2.0311486048020764,
+      "grad_norm": 0.2270224344033579,
+      "learning_rate": 1.7917267917267918e-05,
+      "loss": 0.3414,
+      "step": 1566
+    },
+    {
+      "epoch": 2.0324464633354964,
+      "grad_norm": 0.22685492291746387,
+      "learning_rate": 1.7893217893217896e-05,
+      "loss": 0.334,
+      "step": 1567
+    },
+    {
+      "epoch": 2.0337443218689164,
+      "grad_norm": 0.23079013731214076,
+      "learning_rate": 1.7869167869167868e-05,
+      "loss": 0.3373,
+      "step": 1568
+    },
+    {
+      "epoch": 2.035042180402336,
+      "grad_norm": 0.23832074407215584,
+      "learning_rate": 1.7845117845117846e-05,
+      "loss": 0.3402,
+      "step": 1569
+    },
+    {
+      "epoch": 2.036340038935756,
+      "grad_norm": 0.2413146931238051,
+      "learning_rate": 1.782106782106782e-05,
+      "loss": 0.3476,
+      "step": 1570
+    },
+    {
+      "epoch": 2.037637897469176,
+      "grad_norm": 0.23685355784727574,
+      "learning_rate": 1.77970177970178e-05,
+      "loss": 0.3397,
+      "step": 1571
+    },
+    {
+      "epoch": 2.0389357560025956,
+      "grad_norm": 0.24437850977020956,
+      "learning_rate": 1.7772967772967774e-05,
+      "loss": 0.3445,
+      "step": 1572
+    },
+    {
+      "epoch": 2.0402336145360156,
+      "grad_norm": 0.22724458516557208,
+      "learning_rate": 1.7748917748917752e-05,
+      "loss": 0.3443,
+      "step": 1573
+    },
+    {
+      "epoch": 2.0415314730694356,
+      "grad_norm": 0.23475541449011794,
+      "learning_rate": 1.7724867724867723e-05,
+      "loss": 0.3432,
+      "step": 1574
+    },
+    {
+      "epoch": 2.042829331602855,
+      "grad_norm": 0.21469511225658197,
+      "learning_rate": 1.77008177008177e-05,
+      "loss": 0.3473,
+      "step": 1575
+    },
+    {
+      "epoch": 2.044127190136275,
+      "grad_norm": 0.22486022557380209,
+      "learning_rate": 1.7676767676767676e-05,
+      "loss": 0.3566,
+      "step": 1576
+    },
+    {
+      "epoch": 2.045425048669695,
+      "grad_norm": 0.22895350371242218,
+      "learning_rate": 1.7652717652717654e-05,
+      "loss": 0.3478,
+      "step": 1577
+    },
+    {
+      "epoch": 2.046722907203115,
+      "grad_norm": 0.24538812579393815,
+      "learning_rate": 1.762866762866763e-05,
+      "loss": 0.3332,
+      "step": 1578
+    },
+    {
+      "epoch": 2.048020765736535,
+      "grad_norm": 0.25912360209705504,
+      "learning_rate": 1.7604617604617604e-05,
+      "loss": 0.344,
+      "step": 1579
+    },
+    {
+      "epoch": 2.0493186242699544,
+      "grad_norm": 0.19959244601082998,
+      "learning_rate": 1.7580567580567582e-05,
+      "loss": 0.3337,
+      "step": 1580
+    },
+    {
+      "epoch": 2.0506164828033744,
+      "grad_norm": 0.22265382752478494,
+      "learning_rate": 1.7556517556517557e-05,
+      "loss": 0.3385,
+      "step": 1581
+    },
+    {
+      "epoch": 2.0519143413367944,
+      "grad_norm": 0.2165757161328648,
+      "learning_rate": 1.7532467532467535e-05,
+      "loss": 0.3339,
+      "step": 1582
+    },
+    {
+      "epoch": 2.053212199870214,
+      "grad_norm": 0.21372021503164076,
+      "learning_rate": 1.750841750841751e-05,
+      "loss": 0.3507,
+      "step": 1583
+    },
+    {
+      "epoch": 2.054510058403634,
+      "grad_norm": 0.2336377004408556,
+      "learning_rate": 1.7484367484367488e-05,
+      "loss": 0.3491,
+      "step": 1584
+    },
+    {
+      "epoch": 2.055807916937054,
+      "grad_norm": 0.2117993328839063,
+      "learning_rate": 1.746031746031746e-05,
+      "loss": 0.3407,
+      "step": 1585
+    },
+    {
+      "epoch": 2.0571057754704736,
+      "grad_norm": 0.21231266244658922,
+      "learning_rate": 1.7436267436267437e-05,
+      "loss": 0.3444,
+      "step": 1586
+    },
+    {
+      "epoch": 2.0584036340038936,
+      "grad_norm": 0.21032023819015722,
+      "learning_rate": 1.7412217412217412e-05,
+      "loss": 0.3398,
+      "step": 1587
+    },
+    {
+      "epoch": 2.0597014925373136,
+      "grad_norm": 0.2371048058409055,
+      "learning_rate": 1.738816738816739e-05,
+      "loss": 0.3445,
+      "step": 1588
+    },
+    {
+      "epoch": 2.060999351070733,
+      "grad_norm": 0.2059222075267882,
+      "learning_rate": 1.7364117364117365e-05,
+      "loss": 0.3443,
+      "step": 1589
+    },
+    {
+      "epoch": 2.062297209604153,
+      "grad_norm": 0.22719406397240552,
+      "learning_rate": 1.734006734006734e-05,
+      "loss": 0.3495,
+      "step": 1590
+    },
+    {
+      "epoch": 2.063595068137573,
+      "grad_norm": 0.2222707506963988,
+      "learning_rate": 1.7316017316017315e-05,
+      "loss": 0.3399,
+      "step": 1591
+    },
+    {
+      "epoch": 2.064892926670993,
+      "grad_norm": 0.22555682797470167,
+      "learning_rate": 1.7291967291967293e-05,
+      "loss": 0.3383,
+      "step": 1592
+    },
+    {
+      "epoch": 2.066190785204413,
+      "grad_norm": 0.22889368520998704,
+      "learning_rate": 1.7267917267917268e-05,
+      "loss": 0.3418,
+      "step": 1593
+    },
+    {
+      "epoch": 2.0674886437378324,
+      "grad_norm": 0.22102058324621057,
+      "learning_rate": 1.7243867243867246e-05,
+      "loss": 0.3325,
+      "step": 1594
+    },
+    {
+      "epoch": 2.0687865022712524,
+      "grad_norm": 0.23774221641545448,
+      "learning_rate": 1.721981721981722e-05,
+      "loss": 0.3369,
+      "step": 1595
+    },
+    {
+      "epoch": 2.0700843608046724,
+      "grad_norm": 0.24890061582412498,
+      "learning_rate": 1.7195767195767195e-05,
+      "loss": 0.3441,
+      "step": 1596
+    },
+    {
+      "epoch": 2.071382219338092,
+      "grad_norm": 0.2100873376295878,
+      "learning_rate": 1.7171717171717173e-05,
+      "loss": 0.3303,
+      "step": 1597
+    },
+    {
+      "epoch": 2.072680077871512,
+      "grad_norm": 0.22680230775268373,
+      "learning_rate": 1.7147667147667148e-05,
+      "loss": 0.3371,
+      "step": 1598
+    },
+    {
+      "epoch": 2.073977936404932,
+      "grad_norm": 0.20382351560763964,
+      "learning_rate": 1.7123617123617126e-05,
+      "loss": 0.3337,
+      "step": 1599
+    },
+    {
+      "epoch": 2.0752757949383516,
+      "grad_norm": 0.21534525112062658,
+      "learning_rate": 1.70995670995671e-05,
+      "loss": 0.3371,
+      "step": 1600
+    },
+    {
+      "epoch": 2.0765736534717716,
+      "grad_norm": 0.21827103094501965,
+      "learning_rate": 1.7075517075517076e-05,
+      "loss": 0.3357,
+      "step": 1601
+    },
+    {
+      "epoch": 2.0778715120051916,
+      "grad_norm": 0.21047818536264323,
+      "learning_rate": 1.705146705146705e-05,
+      "loss": 0.3358,
+      "step": 1602
+    },
+    {
+      "epoch": 2.079169370538611,
+      "grad_norm": 0.24359393587806577,
+      "learning_rate": 1.702741702741703e-05,
+      "loss": 0.3409,
+      "step": 1603
+    },
+    {
+      "epoch": 2.080467229072031,
+      "grad_norm": 0.32771668536616283,
+      "learning_rate": 1.7003367003367004e-05,
+      "loss": 0.3296,
+      "step": 1604
+    },
+    {
+      "epoch": 2.081765087605451,
+      "grad_norm": 0.2105738510166506,
+      "learning_rate": 1.6979316979316982e-05,
+      "loss": 0.3471,
+      "step": 1605
+    },
+    {
+      "epoch": 2.0830629461388708,
+      "grad_norm": 0.2609805732511619,
+      "learning_rate": 1.6955266955266957e-05,
+      "loss": 0.3395,
+      "step": 1606
+    },
+    {
+      "epoch": 2.084360804672291,
+      "grad_norm": 0.20813077643429093,
+      "learning_rate": 1.693121693121693e-05,
+      "loss": 0.3323,
+      "step": 1607
+    },
+    {
+      "epoch": 2.0856586632057104,
+      "grad_norm": 0.20588845509767667,
+      "learning_rate": 1.6907166907166906e-05,
+      "loss": 0.3395,
+      "step": 1608
+    },
+    {
+      "epoch": 2.0869565217391304,
+      "grad_norm": 0.2231300777445713,
+      "learning_rate": 1.6883116883116884e-05,
+      "loss": 0.3448,
+      "step": 1609
+    },
+    {
+      "epoch": 2.0882543802725504,
+      "grad_norm": 0.2262598516285643,
+      "learning_rate": 1.685906685906686e-05,
+      "loss": 0.3524,
+      "step": 1610
+    },
+    {
+      "epoch": 2.08955223880597,
+      "grad_norm": 0.2112479505042923,
+      "learning_rate": 1.6835016835016837e-05,
+      "loss": 0.335,
+      "step": 1611
+    },
+    {
+      "epoch": 2.09085009733939,
+      "grad_norm": 0.22189847243097133,
+      "learning_rate": 1.6810966810966812e-05,
+      "loss": 0.3312,
+      "step": 1612
+    },
+    {
+      "epoch": 2.09214795587281,
+      "grad_norm": 0.21750458672162346,
+      "learning_rate": 1.6786916786916787e-05,
+      "loss": 0.3382,
+      "step": 1613
+    },
+    {
+      "epoch": 2.0934458144062296,
+      "grad_norm": 0.22791119863516698,
+      "learning_rate": 1.6762866762866765e-05,
+      "loss": 0.3346,
+      "step": 1614
+    },
+    {
+      "epoch": 2.0947436729396496,
+      "grad_norm": 0.2253801306495037,
+      "learning_rate": 1.673881673881674e-05,
+      "loss": 0.3587,
+      "step": 1615
+    },
+    {
+      "epoch": 2.0960415314730696,
+      "grad_norm": 0.21941342953368462,
+      "learning_rate": 1.6714766714766718e-05,
+      "loss": 0.322,
+      "step": 1616
+    },
+    {
+      "epoch": 2.097339390006489,
+      "grad_norm": 0.24442422654973892,
+      "learning_rate": 1.6690716690716693e-05,
+      "loss": 0.3613,
+      "step": 1617
+    },
+    {
+      "epoch": 2.098637248539909,
+      "grad_norm": 0.20892381899190737,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.3372,
+      "step": 1618
+    },
+    {
+      "epoch": 2.099935107073329,
+      "grad_norm": 0.2204239619365003,
+      "learning_rate": 1.6642616642616642e-05,
+      "loss": 0.345,
+      "step": 1619
+    },
+    {
+      "epoch": 2.1012329656067488,
+      "grad_norm": 0.2433539902885864,
+      "learning_rate": 1.661856661856662e-05,
+      "loss": 0.3399,
+      "step": 1620
+    },
+    {
+      "epoch": 2.102530824140169,
+      "grad_norm": 0.21756713439856754,
+      "learning_rate": 1.6594516594516595e-05,
+      "loss": 0.3258,
+      "step": 1621
+    },
+    {
+      "epoch": 2.103828682673589,
+      "grad_norm": 0.21185754465198486,
+      "learning_rate": 1.6570466570466573e-05,
+      "loss": 0.3338,
+      "step": 1622
+    },
+    {
+      "epoch": 2.1051265412070084,
+      "grad_norm": 0.19886634840357484,
+      "learning_rate": 1.6546416546416545e-05,
+      "loss": 0.3379,
+      "step": 1623
+    },
+    {
+      "epoch": 2.1064243997404284,
+      "grad_norm": 0.23378633487014983,
+      "learning_rate": 1.6522366522366523e-05,
+      "loss": 0.3518,
+      "step": 1624
+    },
+    {
+      "epoch": 2.107722258273848,
+      "grad_norm": 0.22367396102680723,
+      "learning_rate": 1.6498316498316498e-05,
+      "loss": 0.3341,
+      "step": 1625
+    },
+    {
+      "epoch": 2.109020116807268,
+      "grad_norm": 0.2063911182342616,
+      "learning_rate": 1.6474266474266476e-05,
+      "loss": 0.3446,
+      "step": 1626
+    },
+    {
+      "epoch": 2.110317975340688,
+      "grad_norm": 0.22050227302851286,
+      "learning_rate": 1.645021645021645e-05,
+      "loss": 0.3359,
+      "step": 1627
+    },
+    {
+      "epoch": 2.1116158338741076,
+      "grad_norm": 0.24896315784423037,
+      "learning_rate": 1.642616642616643e-05,
+      "loss": 0.3311,
+      "step": 1628
+    },
+    {
+      "epoch": 2.1129136924075276,
+      "grad_norm": 0.24014540903069778,
+      "learning_rate": 1.6402116402116404e-05,
+      "loss": 0.3455,
+      "step": 1629
+    },
+    {
+      "epoch": 2.1142115509409476,
+      "grad_norm": 0.22208241595868494,
+      "learning_rate": 1.637806637806638e-05,
+      "loss": 0.3609,
+      "step": 1630
+    },
+    {
+      "epoch": 2.115509409474367,
+      "grad_norm": 0.25185975208100675,
+      "learning_rate": 1.6354016354016356e-05,
+      "loss": 0.3355,
+      "step": 1631
+    },
+    {
+      "epoch": 2.116807268007787,
+      "grad_norm": 0.25068577932779473,
+      "learning_rate": 1.632996632996633e-05,
+      "loss": 0.3378,
+      "step": 1632
+    },
+    {
+      "epoch": 2.118105126541207,
+      "grad_norm": 0.23743465262529753,
+      "learning_rate": 1.630591630591631e-05,
+      "loss": 0.3453,
+      "step": 1633
+    },
+    {
+      "epoch": 2.1194029850746268,
+      "grad_norm": 0.2873478730260964,
+      "learning_rate": 1.628186628186628e-05,
+      "loss": 0.3469,
+      "step": 1634
+    },
+    {
+      "epoch": 2.120700843608047,
+      "grad_norm": 0.23190098964145786,
+      "learning_rate": 1.625781625781626e-05,
+      "loss": 0.3299,
+      "step": 1635
+    },
+    {
+      "epoch": 2.1219987021414664,
+      "grad_norm": 0.24692649408648704,
+      "learning_rate": 1.6233766233766234e-05,
+      "loss": 0.3413,
+      "step": 1636
+    },
+    {
+      "epoch": 2.1232965606748864,
+      "grad_norm": 0.2601882707422891,
+      "learning_rate": 1.6209716209716212e-05,
+      "loss": 0.3237,
+      "step": 1637
+    },
+    {
+      "epoch": 2.1245944192083064,
+      "grad_norm": 0.2307540235499652,
+      "learning_rate": 1.6185666185666187e-05,
+      "loss": 0.3479,
+      "step": 1638
+    },
+    {
+      "epoch": 2.125892277741726,
+      "grad_norm": 0.24207343026907496,
+      "learning_rate": 1.6161616161616165e-05,
+      "loss": 0.342,
+      "step": 1639
+    },
+    {
+      "epoch": 2.127190136275146,
+      "grad_norm": 0.25704183477977455,
+      "learning_rate": 1.6137566137566136e-05,
+      "loss": 0.347,
+      "step": 1640
+    },
+    {
+      "epoch": 2.128487994808566,
+      "grad_norm": 0.2204667533876293,
+      "learning_rate": 1.6113516113516114e-05,
+      "loss": 0.3421,
+      "step": 1641
+    },
+    {
+      "epoch": 2.1297858533419856,
+      "grad_norm": 0.209970904431735,
+      "learning_rate": 1.608946608946609e-05,
+      "loss": 0.3434,
+      "step": 1642
+    },
+    {
+      "epoch": 2.1310837118754056,
+      "grad_norm": 0.2322772777553896,
+      "learning_rate": 1.6065416065416067e-05,
+      "loss": 0.3462,
+      "step": 1643
+    },
+    {
+      "epoch": 2.1323815704088256,
+      "grad_norm": 0.24539006715088627,
+      "learning_rate": 1.6041366041366042e-05,
+      "loss": 0.3474,
+      "step": 1644
+    },
+    {
+      "epoch": 2.133679428942245,
+      "grad_norm": 0.2071657985362404,
+      "learning_rate": 1.6017316017316017e-05,
+      "loss": 0.315,
+      "step": 1645
+    },
+    {
+      "epoch": 2.134977287475665,
+      "grad_norm": 0.22559015132345064,
+      "learning_rate": 1.5993265993265995e-05,
+      "loss": 0.3435,
+      "step": 1646
+    },
+    {
+      "epoch": 2.136275146009085,
+      "grad_norm": 0.21866568313609377,
+      "learning_rate": 1.596921596921597e-05,
+      "loss": 0.3214,
+      "step": 1647
+    },
+    {
+      "epoch": 2.1375730045425048,
+      "grad_norm": 0.22054227732700918,
+      "learning_rate": 1.5945165945165948e-05,
+      "loss": 0.3366,
+      "step": 1648
+    },
+    {
+      "epoch": 2.1388708630759248,
+      "grad_norm": 0.2513837156975892,
+      "learning_rate": 1.5921115921115923e-05,
+      "loss": 0.3661,
+      "step": 1649
+    },
+    {
+      "epoch": 2.140168721609345,
+      "grad_norm": 0.22217224038750824,
+      "learning_rate": 1.5897065897065898e-05,
+      "loss": 0.3377,
+      "step": 1650
+    },
+    {
+      "epoch": 2.1414665801427644,
+      "grad_norm": 0.21936038896279167,
+      "learning_rate": 1.5873015873015872e-05,
+      "loss": 0.3326,
+      "step": 1651
+    },
+    {
+      "epoch": 2.1427644386761844,
+      "grad_norm": 0.2064910202725993,
+      "learning_rate": 1.584896584896585e-05,
+      "loss": 0.3365,
+      "step": 1652
+    },
+    {
+      "epoch": 2.144062297209604,
+      "grad_norm": 0.20100161814145587,
+      "learning_rate": 1.5824915824915825e-05,
+      "loss": 0.3278,
+      "step": 1653
+    },
+    {
+      "epoch": 2.145360155743024,
+      "grad_norm": 0.21879340248910953,
+      "learning_rate": 1.5800865800865803e-05,
+      "loss": 0.3431,
+      "step": 1654
+    },
+    {
+      "epoch": 2.146658014276444,
+      "grad_norm": 0.2129168285291938,
+      "learning_rate": 1.5776815776815778e-05,
+      "loss": 0.3503,
+      "step": 1655
+    },
+    {
+      "epoch": 2.1479558728098636,
+      "grad_norm": 0.2091423056576906,
+      "learning_rate": 1.5752765752765753e-05,
+      "loss": 0.3492,
+      "step": 1656
+    },
+    {
+      "epoch": 2.1492537313432836,
+      "grad_norm": 0.2828239816378975,
+      "learning_rate": 1.5728715728715728e-05,
+      "loss": 0.3564,
+      "step": 1657
+    },
+    {
+      "epoch": 2.1505515898767036,
+      "grad_norm": 0.21413983501492062,
+      "learning_rate": 1.5704665704665706e-05,
+      "loss": 0.3544,
+      "step": 1658
+    },
+    {
+      "epoch": 2.151849448410123,
+      "grad_norm": 0.21542714936607865,
+      "learning_rate": 1.568061568061568e-05,
+      "loss": 0.3399,
+      "step": 1659
+    },
+    {
+      "epoch": 2.153147306943543,
+      "grad_norm": 0.23220715953725188,
+      "learning_rate": 1.565656565656566e-05,
+      "loss": 0.3322,
+      "step": 1660
+    },
+    {
+      "epoch": 2.154445165476963,
+      "grad_norm": 0.20221618002166372,
+      "learning_rate": 1.563251563251563e-05,
+      "loss": 0.3297,
+      "step": 1661
+    },
+    {
+      "epoch": 2.1557430240103828,
+      "grad_norm": 0.23075242395941706,
+      "learning_rate": 1.560846560846561e-05,
+      "loss": 0.3372,
+      "step": 1662
+    },
+    {
+      "epoch": 2.1570408825438028,
+      "grad_norm": 0.2065260094429175,
+      "learning_rate": 1.5584415584415583e-05,
+      "loss": 0.3396,
+      "step": 1663
+    },
+    {
+      "epoch": 2.158338741077223,
+      "grad_norm": 0.2048397557697161,
+      "learning_rate": 1.556036556036556e-05,
+      "loss": 0.338,
+      "step": 1664
+    },
+    {
+      "epoch": 2.1596365996106424,
+      "grad_norm": 0.21058822127057414,
+      "learning_rate": 1.5536315536315536e-05,
+      "loss": 0.3475,
+      "step": 1665
+    },
+    {
+      "epoch": 2.1609344581440624,
+      "grad_norm": 0.20856117486455514,
+      "learning_rate": 1.5512265512265514e-05,
+      "loss": 0.3565,
+      "step": 1666
+    },
+    {
+      "epoch": 2.162232316677482,
+      "grad_norm": 0.2101578604301147,
+      "learning_rate": 1.548821548821549e-05,
+      "loss": 0.3341,
+      "step": 1667
+    },
+    {
+      "epoch": 2.163530175210902,
+      "grad_norm": 0.1948843188896066,
+      "learning_rate": 1.5464165464165464e-05,
+      "loss": 0.3307,
+      "step": 1668
+    },
+    {
+      "epoch": 2.164828033744322,
+      "grad_norm": 0.22409902293345668,
+      "learning_rate": 1.5440115440115442e-05,
+      "loss": 0.3484,
+      "step": 1669
+    },
+    {
+      "epoch": 2.1661258922777415,
+      "grad_norm": 0.22671547696900604,
+      "learning_rate": 1.5416065416065417e-05,
+      "loss": 0.3523,
+      "step": 1670
+    },
+    {
+      "epoch": 2.1674237508111616,
+      "grad_norm": 0.21140422198147263,
+      "learning_rate": 1.5392015392015395e-05,
+      "loss": 0.3455,
+      "step": 1671
+    },
+    {
+      "epoch": 2.1687216093445816,
+      "grad_norm": 0.21245427765806485,
+      "learning_rate": 1.5367965367965366e-05,
+      "loss": 0.3429,
+      "step": 1672
+    },
+    {
+      "epoch": 2.170019467878001,
+      "grad_norm": 0.20475755840127652,
+      "learning_rate": 1.5343915343915344e-05,
+      "loss": 0.334,
+      "step": 1673
+    },
+    {
+      "epoch": 2.171317326411421,
+      "grad_norm": 0.22161947726300274,
+      "learning_rate": 1.531986531986532e-05,
+      "loss": 0.3445,
+      "step": 1674
+    },
+    {
+      "epoch": 2.172615184944841,
+      "grad_norm": 0.20686901638420155,
+      "learning_rate": 1.5295815295815297e-05,
+      "loss": 0.3423,
+      "step": 1675
+    },
+    {
+      "epoch": 2.1739130434782608,
+      "grad_norm": 0.21241410102955396,
+      "learning_rate": 1.5271765271765272e-05,
+      "loss": 0.3272,
+      "step": 1676
+    },
+    {
+      "epoch": 2.1752109020116808,
+      "grad_norm": 0.20178219626778365,
+      "learning_rate": 1.524771524771525e-05,
+      "loss": 0.3429,
+      "step": 1677
+    },
+    {
+      "epoch": 2.176508760545101,
+      "grad_norm": 0.24450118577688046,
+      "learning_rate": 1.5223665223665223e-05,
+      "loss": 0.3481,
+      "step": 1678
+    },
+    {
+      "epoch": 2.1778066190785204,
+      "grad_norm": 0.20454351303651527,
+      "learning_rate": 1.51996151996152e-05,
+      "loss": 0.3581,
+      "step": 1679
+    },
+    {
+      "epoch": 2.1791044776119404,
+      "grad_norm": 0.21613091983231933,
+      "learning_rate": 1.5175565175565176e-05,
+      "loss": 0.3299,
+      "step": 1680
+    },
+    {
+      "epoch": 2.1804023361453604,
+      "grad_norm": 0.23413467617760164,
+      "learning_rate": 1.5151515151515153e-05,
+      "loss": 0.3622,
+      "step": 1681
+    },
+    {
+      "epoch": 2.18170019467878,
+      "grad_norm": 0.21133908102093862,
+      "learning_rate": 1.512746512746513e-05,
+      "loss": 0.3323,
+      "step": 1682
+    },
+    {
+      "epoch": 2.1829980532122,
+      "grad_norm": 0.2117083520838165,
+      "learning_rate": 1.5103415103415102e-05,
+      "loss": 0.3394,
+      "step": 1683
+    },
+    {
+      "epoch": 2.1842959117456195,
+      "grad_norm": 0.21732760719194733,
+      "learning_rate": 1.5079365079365079e-05,
+      "loss": 0.3419,
+      "step": 1684
+    },
+    {
+      "epoch": 2.1855937702790396,
+      "grad_norm": 0.20428681560868892,
+      "learning_rate": 1.5055315055315055e-05,
+      "loss": 0.3346,
+      "step": 1685
+    },
+    {
+      "epoch": 2.1868916288124596,
+      "grad_norm": 0.21590103314227366,
+      "learning_rate": 1.5031265031265032e-05,
+      "loss": 0.3662,
+      "step": 1686
+    },
+    {
+      "epoch": 2.188189487345879,
+      "grad_norm": 0.21412757670479102,
+      "learning_rate": 1.5007215007215008e-05,
+      "loss": 0.3334,
+      "step": 1687
+    },
+    {
+      "epoch": 2.189487345879299,
+      "grad_norm": 0.2161794768756431,
+      "learning_rate": 1.4983164983164985e-05,
+      "loss": 0.334,
+      "step": 1688
+    },
+    {
+      "epoch": 2.190785204412719,
+      "grad_norm": 0.2121354087724828,
+      "learning_rate": 1.495911495911496e-05,
+      "loss": 0.344,
+      "step": 1689
+    },
+    {
+      "epoch": 2.1920830629461387,
+      "grad_norm": 0.20921645686294296,
+      "learning_rate": 1.4935064935064936e-05,
+      "loss": 0.3397,
+      "step": 1690
+    },
+    {
+      "epoch": 2.1933809214795588,
+      "grad_norm": 0.20966308146376983,
+      "learning_rate": 1.4911014911014912e-05,
+      "loss": 0.3464,
+      "step": 1691
+    },
+    {
+      "epoch": 2.194678780012979,
+      "grad_norm": 0.21908775100267822,
+      "learning_rate": 1.4886964886964889e-05,
+      "loss": 0.3317,
+      "step": 1692
+    },
+    {
+      "epoch": 2.1959766385463984,
+      "grad_norm": 0.22468058324603687,
+      "learning_rate": 1.4862914862914865e-05,
+      "loss": 0.3407,
+      "step": 1693
+    },
+    {
+      "epoch": 2.1972744970798184,
+      "grad_norm": 0.22075013149066078,
+      "learning_rate": 1.4838864838864838e-05,
+      "loss": 0.3283,
+      "step": 1694
+    },
+    {
+      "epoch": 2.198572355613238,
+      "grad_norm": 0.21551594408964253,
+      "learning_rate": 1.4814814814814815e-05,
+      "loss": 0.3199,
+      "step": 1695
+    },
+    {
+      "epoch": 2.199870214146658,
+      "grad_norm": 0.21485802609507748,
+      "learning_rate": 1.4790764790764791e-05,
+      "loss": 0.3287,
+      "step": 1696
+    },
+    {
+      "epoch": 2.201168072680078,
+      "grad_norm": 0.2151673200316436,
+      "learning_rate": 1.4766714766714768e-05,
+      "loss": 0.3426,
+      "step": 1697
+    },
+    {
+      "epoch": 2.2024659312134975,
+      "grad_norm": 0.22612559865377715,
+      "learning_rate": 1.4742664742664744e-05,
+      "loss": 0.3573,
+      "step": 1698
+    },
+    {
+      "epoch": 2.2037637897469176,
+      "grad_norm": 0.21837527709726925,
+      "learning_rate": 1.471861471861472e-05,
+      "loss": 0.3325,
+      "step": 1699
+    },
+    {
+      "epoch": 2.2050616482803376,
+      "grad_norm": 0.21236356177018392,
+      "learning_rate": 1.4694564694564694e-05,
+      "loss": 0.3553,
+      "step": 1700
+    },
+    {
+      "epoch": 2.206359506813757,
+      "grad_norm": 0.19590709294592334,
+      "learning_rate": 1.467051467051467e-05,
+      "loss": 0.3215,
+      "step": 1701
+    },
+    {
+      "epoch": 2.207657365347177,
+      "grad_norm": 0.22771176110161379,
+      "learning_rate": 1.4646464646464647e-05,
+      "loss": 0.3465,
+      "step": 1702
+    },
+    {
+      "epoch": 2.208955223880597,
+      "grad_norm": 0.22229793471063694,
+      "learning_rate": 1.4622414622414623e-05,
+      "loss": 0.3432,
+      "step": 1703
+    },
+    {
+      "epoch": 2.2102530824140167,
+      "grad_norm": 0.21252084935676113,
+      "learning_rate": 1.45983645983646e-05,
+      "loss": 0.3508,
+      "step": 1704
+    },
+    {
+      "epoch": 2.2115509409474368,
+      "grad_norm": 0.2202632087669158,
+      "learning_rate": 1.4574314574314573e-05,
+      "loss": 0.3369,
+      "step": 1705
+    },
+    {
+      "epoch": 2.2128487994808568,
+      "grad_norm": 0.21520394343961766,
+      "learning_rate": 1.455026455026455e-05,
+      "loss": 0.3327,
+      "step": 1706
+    },
+    {
+      "epoch": 2.2141466580142763,
+      "grad_norm": 0.23322877877504564,
+      "learning_rate": 1.4526214526214526e-05,
+      "loss": 0.3491,
+      "step": 1707
+    },
+    {
+      "epoch": 2.2154445165476964,
+      "grad_norm": 0.22519715240573737,
+      "learning_rate": 1.4502164502164502e-05,
+      "loss": 0.3403,
+      "step": 1708
+    },
+    {
+      "epoch": 2.2167423750811164,
+      "grad_norm": 0.22210131676178194,
+      "learning_rate": 1.4478114478114479e-05,
+      "loss": 0.344,
+      "step": 1709
+    },
+    {
+      "epoch": 2.218040233614536,
+      "grad_norm": 0.2119128486555464,
+      "learning_rate": 1.4454064454064455e-05,
+      "loss": 0.3562,
+      "step": 1710
+    },
+    {
+      "epoch": 2.219338092147956,
+      "grad_norm": 0.26673890205097306,
+      "learning_rate": 1.443001443001443e-05,
+      "loss": 0.3284,
+      "step": 1711
+    },
+    {
+      "epoch": 2.2206359506813755,
+      "grad_norm": 0.22633092379431619,
+      "learning_rate": 1.4405964405964406e-05,
+      "loss": 0.3383,
+      "step": 1712
+    },
+    {
+      "epoch": 2.2219338092147956,
+      "grad_norm": 0.21248117786007845,
+      "learning_rate": 1.4381914381914383e-05,
+      "loss": 0.3458,
+      "step": 1713
+    },
+    {
+      "epoch": 2.2232316677482156,
+      "grad_norm": 0.23708025124779677,
+      "learning_rate": 1.435786435786436e-05,
+      "loss": 0.3521,
+      "step": 1714
+    },
+    {
+      "epoch": 2.224529526281635,
+      "grad_norm": 0.2216080470200345,
+      "learning_rate": 1.4333814333814336e-05,
+      "loss": 0.338,
+      "step": 1715
+    },
+    {
+      "epoch": 2.225827384815055,
+      "grad_norm": 0.20064012115244553,
+      "learning_rate": 1.4309764309764309e-05,
+      "loss": 0.3323,
+      "step": 1716
+    },
+    {
+      "epoch": 2.227125243348475,
+      "grad_norm": 0.22946959688466415,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.3533,
+      "step": 1717
+    },
+    {
+      "epoch": 2.2284231018818947,
+      "grad_norm": 0.23129226313271936,
+      "learning_rate": 1.4261664261664262e-05,
+      "loss": 0.3477,
+      "step": 1718
+    },
+    {
+      "epoch": 2.2297209604153148,
+      "grad_norm": 0.23801346245682223,
+      "learning_rate": 1.4237614237614238e-05,
+      "loss": 0.3525,
+      "step": 1719
+    },
+    {
+      "epoch": 2.2310188189487348,
+      "grad_norm": 0.20874849017749872,
+      "learning_rate": 1.4213564213564215e-05,
+      "loss": 0.3414,
+      "step": 1720
+    },
+    {
+      "epoch": 2.2323166774821543,
+      "grad_norm": 0.21629318124043379,
+      "learning_rate": 1.4189514189514191e-05,
+      "loss": 0.3379,
+      "step": 1721
+    },
+    {
+      "epoch": 2.2336145360155744,
+      "grad_norm": 0.22905806827791111,
+      "learning_rate": 1.4165464165464164e-05,
+      "loss": 0.337,
+      "step": 1722
+    },
+    {
+      "epoch": 2.234912394548994,
+      "grad_norm": 0.2279391640803175,
+      "learning_rate": 1.4141414141414141e-05,
+      "loss": 0.3536,
+      "step": 1723
+    },
+    {
+      "epoch": 2.236210253082414,
+      "grad_norm": 0.20743141809266266,
+      "learning_rate": 1.4117364117364117e-05,
+      "loss": 0.3331,
+      "step": 1724
+    },
+    {
+      "epoch": 2.237508111615834,
+      "grad_norm": 0.21930312729137855,
+      "learning_rate": 1.4093314093314094e-05,
+      "loss": 0.3426,
+      "step": 1725
+    },
+    {
+      "epoch": 2.2388059701492535,
+      "grad_norm": 0.21584060808323735,
+      "learning_rate": 1.406926406926407e-05,
+      "loss": 0.3326,
+      "step": 1726
+    },
+    {
+      "epoch": 2.2401038286826735,
+      "grad_norm": 0.20452984796446286,
+      "learning_rate": 1.4045214045214045e-05,
+      "loss": 0.3407,
+      "step": 1727
+    },
+    {
+      "epoch": 2.2414016872160936,
+      "grad_norm": 0.21278894400954876,
+      "learning_rate": 1.4021164021164022e-05,
+      "loss": 0.3272,
+      "step": 1728
+    },
+    {
+      "epoch": 2.242699545749513,
+      "grad_norm": 0.19989940211340443,
+      "learning_rate": 1.3997113997113998e-05,
+      "loss": 0.3268,
+      "step": 1729
+    },
+    {
+      "epoch": 2.243997404282933,
+      "grad_norm": 0.20756531216735427,
+      "learning_rate": 1.3973063973063974e-05,
+      "loss": 0.3428,
+      "step": 1730
+    },
+    {
+      "epoch": 2.245295262816353,
+      "grad_norm": 0.2289163023889471,
+      "learning_rate": 1.3949013949013951e-05,
+      "loss": 0.3531,
+      "step": 1731
+    },
+    {
+      "epoch": 2.2465931213497727,
+      "grad_norm": 0.19474714109733912,
+      "learning_rate": 1.3924963924963927e-05,
+      "loss": 0.3232,
+      "step": 1732
+    },
+    {
+      "epoch": 2.2478909798831928,
+      "grad_norm": 0.20474938955793603,
+      "learning_rate": 1.39009139009139e-05,
+      "loss": 0.3368,
+      "step": 1733
+    },
+    {
+      "epoch": 2.2491888384166128,
+      "grad_norm": 0.2134278581088506,
+      "learning_rate": 1.3876863876863877e-05,
+      "loss": 0.3332,
+      "step": 1734
+    },
+    {
+      "epoch": 2.2504866969500323,
+      "grad_norm": 0.20557863182364947,
+      "learning_rate": 1.3852813852813853e-05,
+      "loss": 0.3189,
+      "step": 1735
+    },
+    {
+      "epoch": 2.2517845554834524,
+      "grad_norm": 0.20536064005062202,
+      "learning_rate": 1.382876382876383e-05,
+      "loss": 0.3432,
+      "step": 1736
+    },
+    {
+      "epoch": 2.2530824140168724,
+      "grad_norm": 0.20167055126244063,
+      "learning_rate": 1.3804713804713806e-05,
+      "loss": 0.3333,
+      "step": 1737
+    },
+    {
+      "epoch": 2.254380272550292,
+      "grad_norm": 0.21070060115934763,
+      "learning_rate": 1.378066378066378e-05,
+      "loss": 0.3447,
+      "step": 1738
+    },
+    {
+      "epoch": 2.255678131083712,
+      "grad_norm": 0.2207628970020607,
+      "learning_rate": 1.3756613756613756e-05,
+      "loss": 0.3583,
+      "step": 1739
+    },
+    {
+      "epoch": 2.256975989617132,
+      "grad_norm": 0.2127435360633616,
+      "learning_rate": 1.3732563732563732e-05,
+      "loss": 0.3348,
+      "step": 1740
+    },
+    {
+      "epoch": 2.2582738481505515,
+      "grad_norm": 0.21284122659552568,
+      "learning_rate": 1.3708513708513709e-05,
+      "loss": 0.3625,
+      "step": 1741
+    },
+    {
+      "epoch": 2.2595717066839716,
+      "grad_norm": 0.1950008896432417,
+      "learning_rate": 1.3684463684463685e-05,
+      "loss": 0.3324,
+      "step": 1742
+    },
+    {
+      "epoch": 2.260869565217391,
+      "grad_norm": 0.2020508328430615,
+      "learning_rate": 1.3660413660413662e-05,
+      "loss": 0.3479,
+      "step": 1743
+    },
+    {
+      "epoch": 2.262167423750811,
+      "grad_norm": 0.20145545550530372,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 0.3499,
+      "step": 1744
+    },
+    {
+      "epoch": 2.263465282284231,
+      "grad_norm": 0.20947296790103498,
+      "learning_rate": 1.3612313612313613e-05,
+      "loss": 0.3499,
+      "step": 1745
+    },
+    {
+      "epoch": 2.2647631408176507,
+      "grad_norm": 0.20527034801018748,
+      "learning_rate": 1.358826358826359e-05,
+      "loss": 0.3311,
+      "step": 1746
+    },
+    {
+      "epoch": 2.2660609993510707,
+      "grad_norm": 0.21571814965895064,
+      "learning_rate": 1.3564213564213566e-05,
+      "loss": 0.3445,
+      "step": 1747
+    },
+    {
+      "epoch": 2.2673588578844908,
+      "grad_norm": 0.1951070244580519,
+      "learning_rate": 1.3540163540163542e-05,
+      "loss": 0.3317,
+      "step": 1748
+    },
+    {
+      "epoch": 2.2686567164179103,
+      "grad_norm": 0.20164098805440273,
+      "learning_rate": 1.3516113516113516e-05,
+      "loss": 0.3489,
+      "step": 1749
+    },
+    {
+      "epoch": 2.2699545749513304,
+      "grad_norm": 0.2037235219687979,
+      "learning_rate": 1.3492063492063492e-05,
+      "loss": 0.3447,
+      "step": 1750
+    },
+    {
+      "epoch": 2.27125243348475,
+      "grad_norm": 0.21220292515247122,
+      "learning_rate": 1.3468013468013468e-05,
+      "loss": 0.3395,
+      "step": 1751
+    },
+    {
+      "epoch": 2.27255029201817,
+      "grad_norm": 0.2037471406983462,
+      "learning_rate": 1.3443963443963445e-05,
+      "loss": 0.3415,
+      "step": 1752
+    },
+    {
+      "epoch": 2.27384815055159,
+      "grad_norm": 0.20298389579886292,
+      "learning_rate": 1.3419913419913421e-05,
+      "loss": 0.3374,
+      "step": 1753
+    },
+    {
+      "epoch": 2.2751460090850095,
+      "grad_norm": 0.20691493521870025,
+      "learning_rate": 1.3395863395863398e-05,
+      "loss": 0.3519,
+      "step": 1754
+    },
+    {
+      "epoch": 2.2764438676184295,
+      "grad_norm": 0.2013802963124298,
+      "learning_rate": 1.3371813371813371e-05,
+      "loss": 0.3411,
+      "step": 1755
+    },
+    {
+      "epoch": 2.2777417261518496,
+      "grad_norm": 0.20436433898344716,
+      "learning_rate": 1.3347763347763347e-05,
+      "loss": 0.3447,
+      "step": 1756
+    },
+    {
+      "epoch": 2.279039584685269,
+      "grad_norm": 0.1926683227358606,
+      "learning_rate": 1.3323713323713324e-05,
+      "loss": 0.3326,
+      "step": 1757
+    },
+    {
+      "epoch": 2.280337443218689,
+      "grad_norm": 0.22362998880849946,
+      "learning_rate": 1.32996632996633e-05,
+      "loss": 0.338,
+      "step": 1758
+    },
+    {
+      "epoch": 2.281635301752109,
+      "grad_norm": 0.19696154548860742,
+      "learning_rate": 1.3275613275613277e-05,
+      "loss": 0.3302,
+      "step": 1759
+    },
+    {
+      "epoch": 2.2829331602855287,
+      "grad_norm": 0.21290597699073446,
+      "learning_rate": 1.3251563251563252e-05,
+      "loss": 0.3341,
+      "step": 1760
+    },
+    {
+      "epoch": 2.2842310188189487,
+      "grad_norm": 0.2025599154856036,
+      "learning_rate": 1.3227513227513228e-05,
+      "loss": 0.3292,
+      "step": 1761
+    },
+    {
+      "epoch": 2.2855288773523688,
+      "grad_norm": 0.20827197895684288,
+      "learning_rate": 1.3203463203463205e-05,
+      "loss": 0.348,
+      "step": 1762
+    },
+    {
+      "epoch": 2.2868267358857883,
+      "grad_norm": 0.2031367197949487,
+      "learning_rate": 1.3179413179413181e-05,
+      "loss": 0.3353,
+      "step": 1763
+    },
+    {
+      "epoch": 2.2881245944192083,
+      "grad_norm": 0.20119905187117781,
+      "learning_rate": 1.3155363155363157e-05,
+      "loss": 0.3291,
+      "step": 1764
+    },
+    {
+      "epoch": 2.2894224529526284,
+      "grad_norm": 0.19920483051203258,
+      "learning_rate": 1.3131313131313134e-05,
+      "loss": 0.3475,
+      "step": 1765
+    },
+    {
+      "epoch": 2.290720311486048,
+      "grad_norm": 0.3011556934952649,
+      "learning_rate": 1.3107263107263107e-05,
+      "loss": 0.3535,
+      "step": 1766
+    },
+    {
+      "epoch": 2.292018170019468,
+      "grad_norm": 0.21100649854372194,
+      "learning_rate": 1.3083213083213083e-05,
+      "loss": 0.3469,
+      "step": 1767
+    },
+    {
+      "epoch": 2.293316028552888,
+      "grad_norm": 0.1984331828091875,
+      "learning_rate": 1.305916305916306e-05,
+      "loss": 0.346,
+      "step": 1768
+    },
+    {
+      "epoch": 2.2946138870863075,
+      "grad_norm": 0.2128841448679342,
+      "learning_rate": 1.3035113035113036e-05,
+      "loss": 0.3426,
+      "step": 1769
+    },
+    {
+      "epoch": 2.2959117456197276,
+      "grad_norm": 0.20032623160831234,
+      "learning_rate": 1.3011063011063013e-05,
+      "loss": 0.3323,
+      "step": 1770
+    },
+    {
+      "epoch": 2.297209604153147,
+      "grad_norm": 0.21020072070413953,
+      "learning_rate": 1.2987012987012986e-05,
+      "loss": 0.345,
+      "step": 1771
+    },
+    {
+      "epoch": 2.298507462686567,
+      "grad_norm": 0.220140017454239,
+      "learning_rate": 1.2962962962962962e-05,
+      "loss": 0.338,
+      "step": 1772
+    },
+    {
+      "epoch": 2.299805321219987,
+      "grad_norm": 0.20602192883284073,
+      "learning_rate": 1.2938912938912939e-05,
+      "loss": 0.3295,
+      "step": 1773
+    },
+    {
+      "epoch": 2.3011031797534067,
+      "grad_norm": 0.19782214494971023,
+      "learning_rate": 1.2914862914862915e-05,
+      "loss": 0.3443,
+      "step": 1774
+    },
+    {
+      "epoch": 2.3024010382868267,
+      "grad_norm": 0.2212349266901331,
+      "learning_rate": 1.2890812890812892e-05,
+      "loss": 0.3372,
+      "step": 1775
+    },
+    {
+      "epoch": 2.3036988968202468,
+      "grad_norm": 0.20379287350155195,
+      "learning_rate": 1.2866762866762868e-05,
+      "loss": 0.3298,
+      "step": 1776
+    },
+    {
+      "epoch": 2.3049967553536663,
+      "grad_norm": 0.22633645234017427,
+      "learning_rate": 1.2842712842712843e-05,
+      "loss": 0.3358,
+      "step": 1777
+    },
+    {
+      "epoch": 2.3062946138870863,
+      "grad_norm": 0.22571485246122885,
+      "learning_rate": 1.281866281866282e-05,
+      "loss": 0.3535,
+      "step": 1778
+    },
+    {
+      "epoch": 2.3075924724205064,
+      "grad_norm": 0.1971894196957368,
+      "learning_rate": 1.2794612794612796e-05,
+      "loss": 0.3437,
+      "step": 1779
+    },
+    {
+      "epoch": 2.308890330953926,
+      "grad_norm": 0.21039001676553398,
+      "learning_rate": 1.2770562770562773e-05,
+      "loss": 0.3436,
+      "step": 1780
+    },
+    {
+      "epoch": 2.310188189487346,
+      "grad_norm": 0.21586183237415976,
+      "learning_rate": 1.2746512746512749e-05,
+      "loss": 0.3371,
+      "step": 1781
+    },
+    {
+      "epoch": 2.3114860480207655,
+      "grad_norm": 0.22449090420078008,
+      "learning_rate": 1.2722462722462722e-05,
+      "loss": 0.3426,
+      "step": 1782
+    },
+    {
+      "epoch": 2.3127839065541855,
+      "grad_norm": 0.20000822387430392,
+      "learning_rate": 1.2698412698412699e-05,
+      "loss": 0.3273,
+      "step": 1783
+    },
+    {
+      "epoch": 2.3140817650876055,
+      "grad_norm": 0.19796366453881245,
+      "learning_rate": 1.2674362674362675e-05,
+      "loss": 0.341,
+      "step": 1784
+    },
+    {
+      "epoch": 2.315379623621025,
+      "grad_norm": 0.20625398962095803,
+      "learning_rate": 1.2650312650312651e-05,
+      "loss": 0.3327,
+      "step": 1785
+    },
+    {
+      "epoch": 2.316677482154445,
+      "grad_norm": 0.2018595792474862,
+      "learning_rate": 1.2626262626262628e-05,
+      "loss": 0.3425,
+      "step": 1786
+    },
+    {
+      "epoch": 2.317975340687865,
+      "grad_norm": 0.21810379177370443,
+      "learning_rate": 1.2602212602212604e-05,
+      "loss": 0.3411,
+      "step": 1787
+    },
+    {
+      "epoch": 2.3192731992212847,
+      "grad_norm": 0.21122946645750976,
+      "learning_rate": 1.2578162578162577e-05,
+      "loss": 0.3519,
+      "step": 1788
+    },
+    {
+      "epoch": 2.3205710577547047,
+      "grad_norm": 0.21682608106667023,
+      "learning_rate": 1.2554112554112554e-05,
+      "loss": 0.3391,
+      "step": 1789
+    },
+    {
+      "epoch": 2.3218689162881248,
+      "grad_norm": 0.21596334841986267,
+      "learning_rate": 1.253006253006253e-05,
+      "loss": 0.3478,
+      "step": 1790
+    },
+    {
+      "epoch": 2.3231667748215443,
+      "grad_norm": 0.21752697824305056,
+      "learning_rate": 1.2506012506012507e-05,
+      "loss": 0.3327,
+      "step": 1791
+    },
+    {
+      "epoch": 2.3244646333549643,
+      "grad_norm": 0.19911928426673434,
+      "learning_rate": 1.2481962481962482e-05,
+      "loss": 0.3337,
+      "step": 1792
+    },
+    {
+      "epoch": 2.3257624918883844,
+      "grad_norm": 0.21067483324446745,
+      "learning_rate": 1.2457912457912458e-05,
+      "loss": 0.3383,
+      "step": 1793
+    },
+    {
+      "epoch": 2.327060350421804,
+      "grad_norm": 0.21855693528672904,
+      "learning_rate": 1.2433862433862433e-05,
+      "loss": 0.3392,
+      "step": 1794
+    },
+    {
+      "epoch": 2.328358208955224,
+      "grad_norm": 0.2313881551522217,
+      "learning_rate": 1.240981240981241e-05,
+      "loss": 0.344,
+      "step": 1795
+    },
+    {
+      "epoch": 2.329656067488644,
+      "grad_norm": 0.21647217328647403,
+      "learning_rate": 1.2385762385762386e-05,
+      "loss": 0.3498,
+      "step": 1796
+    },
+    {
+      "epoch": 2.3309539260220635,
+      "grad_norm": 0.23563572241047098,
+      "learning_rate": 1.2361712361712362e-05,
+      "loss": 0.3492,
+      "step": 1797
+    },
+    {
+      "epoch": 2.3322517845554835,
+      "grad_norm": 0.22879253600129817,
+      "learning_rate": 1.2337662337662339e-05,
+      "loss": 0.3476,
+      "step": 1798
+    },
+    {
+      "epoch": 2.3335496430889036,
+      "grad_norm": 0.2194770239864021,
+      "learning_rate": 1.2313612313612315e-05,
+      "loss": 0.3499,
+      "step": 1799
+    },
+    {
+      "epoch": 2.334847501622323,
+      "grad_norm": 0.2323432486506921,
+      "learning_rate": 1.228956228956229e-05,
+      "loss": 0.3498,
+      "step": 1800
+    },
+    {
+      "epoch": 2.336145360155743,
+      "grad_norm": 0.2379829541757471,
+      "learning_rate": 1.2265512265512267e-05,
+      "loss": 0.3455,
+      "step": 1801
+    },
+    {
+      "epoch": 2.3374432186891627,
+      "grad_norm": 0.22305188673020696,
+      "learning_rate": 1.2241462241462243e-05,
+      "loss": 0.3464,
+      "step": 1802
+    },
+    {
+      "epoch": 2.3387410772225827,
+      "grad_norm": 0.20626742501221118,
+      "learning_rate": 1.2217412217412218e-05,
+      "loss": 0.3383,
+      "step": 1803
+    },
+    {
+      "epoch": 2.3400389357560027,
+      "grad_norm": 0.20476597434634727,
+      "learning_rate": 1.2193362193362194e-05,
+      "loss": 0.3271,
+      "step": 1804
+    },
+    {
+      "epoch": 2.3413367942894223,
+      "grad_norm": 0.2361285638480301,
+      "learning_rate": 1.2169312169312169e-05,
+      "loss": 0.363,
+      "step": 1805
+    },
+    {
+      "epoch": 2.3426346528228423,
+      "grad_norm": 0.21885176640576573,
+      "learning_rate": 1.2145262145262145e-05,
+      "loss": 0.3322,
+      "step": 1806
+    },
+    {
+      "epoch": 2.3439325113562623,
+      "grad_norm": 0.2186287737946408,
+      "learning_rate": 1.2121212121212122e-05,
+      "loss": 0.3517,
+      "step": 1807
+    },
+    {
+      "epoch": 2.345230369889682,
+      "grad_norm": 0.22242925263530489,
+      "learning_rate": 1.2097162097162097e-05,
+      "loss": 0.3556,
+      "step": 1808
+    },
+    {
+      "epoch": 2.346528228423102,
+      "grad_norm": 0.21839105455863506,
+      "learning_rate": 1.2073112073112073e-05,
+      "loss": 0.3537,
+      "step": 1809
+    },
+    {
+      "epoch": 2.3478260869565215,
+      "grad_norm": 0.19288692731778082,
+      "learning_rate": 1.204906204906205e-05,
+      "loss": 0.3267,
+      "step": 1810
+    },
+    {
+      "epoch": 2.3491239454899415,
+      "grad_norm": 0.21299009964604498,
+      "learning_rate": 1.2025012025012024e-05,
+      "loss": 0.3343,
+      "step": 1811
+    },
+    {
+      "epoch": 2.3504218040233615,
+      "grad_norm": 0.20073387684513502,
+      "learning_rate": 1.2000962000962001e-05,
+      "loss": 0.3366,
+      "step": 1812
+    },
+    {
+      "epoch": 2.351719662556781,
+      "grad_norm": 0.20924433976474296,
+      "learning_rate": 1.1976911976911977e-05,
+      "loss": 0.3503,
+      "step": 1813
+    },
+    {
+      "epoch": 2.353017521090201,
+      "grad_norm": 0.2024959718616962,
+      "learning_rate": 1.1952861952861954e-05,
+      "loss": 0.3398,
+      "step": 1814
+    },
+    {
+      "epoch": 2.354315379623621,
+      "grad_norm": 0.20136147992617448,
+      "learning_rate": 1.192881192881193e-05,
+      "loss": 0.3329,
+      "step": 1815
+    },
+    {
+      "epoch": 2.3556132381570407,
+      "grad_norm": 0.2023856480954257,
+      "learning_rate": 1.1904761904761905e-05,
+      "loss": 0.3431,
+      "step": 1816
+    },
+    {
+      "epoch": 2.3569110966904607,
+      "grad_norm": 0.19571106019001464,
+      "learning_rate": 1.1880711880711882e-05,
+      "loss": 0.3409,
+      "step": 1817
+    },
+    {
+      "epoch": 2.3582089552238807,
+      "grad_norm": 0.20043918045056716,
+      "learning_rate": 1.1856661856661858e-05,
+      "loss": 0.3457,
+      "step": 1818
+    },
+    {
+      "epoch": 2.3595068137573003,
+      "grad_norm": 0.20301799957042976,
+      "learning_rate": 1.1832611832611833e-05,
+      "loss": 0.358,
+      "step": 1819
+    },
+    {
+      "epoch": 2.3608046722907203,
+      "grad_norm": 0.1985053306211951,
+      "learning_rate": 1.180856180856181e-05,
+      "loss": 0.3476,
+      "step": 1820
+    },
+    {
+      "epoch": 2.3621025308241403,
+      "grad_norm": 0.1946833217729552,
+      "learning_rate": 1.1784511784511786e-05,
+      "loss": 0.3381,
+      "step": 1821
+    },
+    {
+      "epoch": 2.36340038935756,
+      "grad_norm": 0.20694128635780762,
+      "learning_rate": 1.176046176046176e-05,
+      "loss": 0.3484,
+      "step": 1822
+    },
+    {
+      "epoch": 2.36469824789098,
+      "grad_norm": 0.19680244677531286,
+      "learning_rate": 1.1736411736411737e-05,
+      "loss": 0.3412,
+      "step": 1823
+    },
+    {
+      "epoch": 2.3659961064244,
+      "grad_norm": 0.19475058414044913,
+      "learning_rate": 1.1712361712361713e-05,
+      "loss": 0.3338,
+      "step": 1824
+    },
+    {
+      "epoch": 2.3672939649578195,
+      "grad_norm": 0.20017845696117334,
+      "learning_rate": 1.1688311688311688e-05,
+      "loss": 0.3292,
+      "step": 1825
+    },
+    {
+      "epoch": 2.3685918234912395,
+      "grad_norm": 0.21405181485690658,
+      "learning_rate": 1.1664261664261665e-05,
+      "loss": 0.3319,
+      "step": 1826
+    },
+    {
+      "epoch": 2.3698896820246595,
+      "grad_norm": 0.2255906580166369,
+      "learning_rate": 1.164021164021164e-05,
+      "loss": 0.3583,
+      "step": 1827
+    },
+    {
+      "epoch": 2.371187540558079,
+      "grad_norm": 0.19990250671178067,
+      "learning_rate": 1.1616161616161616e-05,
+      "loss": 0.337,
+      "step": 1828
+    },
+    {
+      "epoch": 2.372485399091499,
+      "grad_norm": 0.19827952220037648,
+      "learning_rate": 1.1592111592111592e-05,
+      "loss": 0.3432,
+      "step": 1829
+    },
+    {
+      "epoch": 2.3737832576249187,
+      "grad_norm": 0.19939279257051523,
+      "learning_rate": 1.1568061568061569e-05,
+      "loss": 0.3374,
+      "step": 1830
+    },
+    {
+      "epoch": 2.3750811161583387,
+      "grad_norm": 0.1906800360211246,
+      "learning_rate": 1.1544011544011545e-05,
+      "loss": 0.3423,
+      "step": 1831
+    },
+    {
+      "epoch": 2.3763789746917587,
+      "grad_norm": 0.2027119176012166,
+      "learning_rate": 1.1519961519961522e-05,
+      "loss": 0.3431,
+      "step": 1832
+    },
+    {
+      "epoch": 2.3776768332251783,
+      "grad_norm": 0.20771653103434248,
+      "learning_rate": 1.1495911495911497e-05,
+      "loss": 0.3621,
+      "step": 1833
+    },
+    {
+      "epoch": 2.3789746917585983,
+      "grad_norm": 0.18554441127828813,
+      "learning_rate": 1.1471861471861473e-05,
+      "loss": 0.326,
+      "step": 1834
+    },
+    {
+      "epoch": 2.3802725502920183,
+      "grad_norm": 0.19747340923777565,
+      "learning_rate": 1.144781144781145e-05,
+      "loss": 0.3247,
+      "step": 1835
+    },
+    {
+      "epoch": 2.381570408825438,
+      "grad_norm": 0.20647886448091093,
+      "learning_rate": 1.1423761423761424e-05,
+      "loss": 0.3341,
+      "step": 1836
+    },
+    {
+      "epoch": 2.382868267358858,
+      "grad_norm": 0.1957627449624196,
+      "learning_rate": 1.13997113997114e-05,
+      "loss": 0.3328,
+      "step": 1837
+    },
+    {
+      "epoch": 2.3841661258922775,
+      "grad_norm": 0.19525704689585352,
+      "learning_rate": 1.1375661375661376e-05,
+      "loss": 0.347,
+      "step": 1838
+    },
+    {
+      "epoch": 2.3854639844256975,
+      "grad_norm": 0.20804623050610585,
+      "learning_rate": 1.1351611351611352e-05,
+      "loss": 0.3313,
+      "step": 1839
+    },
+    {
+      "epoch": 2.3867618429591175,
+      "grad_norm": 0.21139171271078994,
+      "learning_rate": 1.1327561327561329e-05,
+      "loss": 0.3295,
+      "step": 1840
+    },
+    {
+      "epoch": 2.388059701492537,
+      "grad_norm": 0.19695078516317913,
+      "learning_rate": 1.1303511303511303e-05,
+      "loss": 0.3244,
+      "step": 1841
+    },
+    {
+      "epoch": 2.389357560025957,
+      "grad_norm": 0.21329704668287522,
+      "learning_rate": 1.127946127946128e-05,
+      "loss": 0.3496,
+      "step": 1842
+    },
+    {
+      "epoch": 2.390655418559377,
+      "grad_norm": 0.19692253824786773,
+      "learning_rate": 1.1255411255411256e-05,
+      "loss": 0.3414,
+      "step": 1843
+    },
+    {
+      "epoch": 2.3919532770927967,
+      "grad_norm": 0.20145116731281673,
+      "learning_rate": 1.1231361231361231e-05,
+      "loss": 0.3411,
+      "step": 1844
+    },
+    {
+      "epoch": 2.3932511356262167,
+      "grad_norm": 0.21094040457540925,
+      "learning_rate": 1.1207311207311207e-05,
+      "loss": 0.3425,
+      "step": 1845
+    },
+    {
+      "epoch": 2.3945489941596367,
+      "grad_norm": 0.20933806797553264,
+      "learning_rate": 1.1183261183261184e-05,
+      "loss": 0.3447,
+      "step": 1846
+    },
+    {
+      "epoch": 2.3958468526930563,
+      "grad_norm": 0.22025133979918343,
+      "learning_rate": 1.1159211159211159e-05,
+      "loss": 0.3276,
+      "step": 1847
+    },
+    {
+      "epoch": 2.3971447112264763,
+      "grad_norm": 0.2033688921914023,
+      "learning_rate": 1.1135161135161135e-05,
+      "loss": 0.3427,
+      "step": 1848
+    },
+    {
+      "epoch": 2.3984425697598963,
+      "grad_norm": 0.20001119671379927,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.3291,
+      "step": 1849
+    },
+    {
+      "epoch": 2.399740428293316,
+      "grad_norm": 0.2096895946679153,
+      "learning_rate": 1.1087061087061088e-05,
+      "loss": 0.3529,
+      "step": 1850
+    },
+    {
+      "epoch": 2.401038286826736,
+      "grad_norm": 0.20086557980683176,
+      "learning_rate": 1.1063011063011065e-05,
+      "loss": 0.3386,
+      "step": 1851
+    },
+    {
+      "epoch": 2.402336145360156,
+      "grad_norm": 0.19654561939004062,
+      "learning_rate": 1.103896103896104e-05,
+      "loss": 0.347,
+      "step": 1852
+    },
+    {
+      "epoch": 2.4036340038935755,
+      "grad_norm": 0.20190747910646842,
+      "learning_rate": 1.1014911014911016e-05,
+      "loss": 0.3355,
+      "step": 1853
+    },
+    {
+      "epoch": 2.4049318624269955,
+      "grad_norm": 0.20073723518377382,
+      "learning_rate": 1.0990860990860992e-05,
+      "loss": 0.3506,
+      "step": 1854
+    },
+    {
+      "epoch": 2.4062297209604155,
+      "grad_norm": 0.19812022485550956,
+      "learning_rate": 1.0966810966810967e-05,
+      "loss": 0.3326,
+      "step": 1855
+    },
+    {
+      "epoch": 2.407527579493835,
+      "grad_norm": 0.20293093391898026,
+      "learning_rate": 1.0942760942760944e-05,
+      "loss": 0.3437,
+      "step": 1856
+    },
+    {
+      "epoch": 2.408825438027255,
+      "grad_norm": 0.18445064925773152,
+      "learning_rate": 1.091871091871092e-05,
+      "loss": 0.3412,
+      "step": 1857
+    },
+    {
+      "epoch": 2.4101232965606747,
+      "grad_norm": 0.18982196010225733,
+      "learning_rate": 1.0894660894660895e-05,
+      "loss": 0.3334,
+      "step": 1858
+    },
+    {
+      "epoch": 2.4114211550940947,
+      "grad_norm": 0.19790842980140105,
+      "learning_rate": 1.0870610870610871e-05,
+      "loss": 0.341,
+      "step": 1859
+    },
+    {
+      "epoch": 2.4127190136275147,
+      "grad_norm": 0.21825470803362326,
+      "learning_rate": 1.0846560846560846e-05,
+      "loss": 0.3463,
+      "step": 1860
+    },
+    {
+      "epoch": 2.4140168721609343,
+      "grad_norm": 0.19742393864907667,
+      "learning_rate": 1.0822510822510823e-05,
+      "loss": 0.3275,
+      "step": 1861
+    },
+    {
+      "epoch": 2.4153147306943543,
+      "grad_norm": 0.19224523357142126,
+      "learning_rate": 1.0798460798460799e-05,
+      "loss": 0.3325,
+      "step": 1862
+    },
+    {
+      "epoch": 2.4166125892277743,
+      "grad_norm": 0.19938049624693138,
+      "learning_rate": 1.0774410774410774e-05,
+      "loss": 0.3262,
+      "step": 1863
+    },
+    {
+      "epoch": 2.417910447761194,
+      "grad_norm": 0.19485806663699845,
+      "learning_rate": 1.075036075036075e-05,
+      "loss": 0.3357,
+      "step": 1864
+    },
+    {
+      "epoch": 2.419208306294614,
+      "grad_norm": 0.20583844260408463,
+      "learning_rate": 1.0726310726310727e-05,
+      "loss": 0.3408,
+      "step": 1865
+    },
+    {
+      "epoch": 2.420506164828034,
+      "grad_norm": 0.19685425374116253,
+      "learning_rate": 1.0702260702260703e-05,
+      "loss": 0.34,
+      "step": 1866
+    },
+    {
+      "epoch": 2.4218040233614535,
+      "grad_norm": 0.19475698760728785,
+      "learning_rate": 1.067821067821068e-05,
+      "loss": 0.3415,
+      "step": 1867
+    },
+    {
+      "epoch": 2.4231018818948735,
+      "grad_norm": 0.197170239994665,
+      "learning_rate": 1.0654160654160656e-05,
+      "loss": 0.3328,
+      "step": 1868
+    },
+    {
+      "epoch": 2.424399740428293,
+      "grad_norm": 0.2312526818007591,
+      "learning_rate": 1.0630110630110631e-05,
+      "loss": 0.3517,
+      "step": 1869
+    },
+    {
+      "epoch": 2.425697598961713,
+      "grad_norm": 0.21684364271507464,
+      "learning_rate": 1.0606060606060607e-05,
+      "loss": 0.3383,
+      "step": 1870
+    },
+    {
+      "epoch": 2.426995457495133,
+      "grad_norm": 0.1985082584320083,
+      "learning_rate": 1.0582010582010582e-05,
+      "loss": 0.3246,
+      "step": 1871
+    },
+    {
+      "epoch": 2.4282933160285527,
+      "grad_norm": 0.21714639429646007,
+      "learning_rate": 1.0557960557960559e-05,
+      "loss": 0.3497,
+      "step": 1872
+    },
+    {
+      "epoch": 2.4295911745619727,
+      "grad_norm": 0.22321524923011035,
+      "learning_rate": 1.0533910533910535e-05,
+      "loss": 0.3553,
+      "step": 1873
+    },
+    {
+      "epoch": 2.4308890330953927,
+      "grad_norm": 0.23372261304917366,
+      "learning_rate": 1.050986050986051e-05,
+      "loss": 0.3686,
+      "step": 1874
+    },
+    {
+      "epoch": 2.4321868916288123,
+      "grad_norm": 0.20001574330722693,
+      "learning_rate": 1.0485810485810486e-05,
+      "loss": 0.3351,
+      "step": 1875
+    },
+    {
+      "epoch": 2.4334847501622323,
+      "grad_norm": 0.20673565072923414,
+      "learning_rate": 1.0461760461760463e-05,
+      "loss": 0.3596,
+      "step": 1876
+    },
+    {
+      "epoch": 2.4347826086956523,
+      "grad_norm": 0.20095320206649409,
+      "learning_rate": 1.0437710437710438e-05,
+      "loss": 0.3328,
+      "step": 1877
+    },
+    {
+      "epoch": 2.436080467229072,
+      "grad_norm": 0.2789190666720269,
+      "learning_rate": 1.0413660413660414e-05,
+      "loss": 0.3509,
+      "step": 1878
+    },
+    {
+      "epoch": 2.437378325762492,
+      "grad_norm": 0.2079166722723446,
+      "learning_rate": 1.038961038961039e-05,
+      "loss": 0.3492,
+      "step": 1879
+    },
+    {
+      "epoch": 2.438676184295912,
+      "grad_norm": 0.21591855675091434,
+      "learning_rate": 1.0365560365560365e-05,
+      "loss": 0.3375,
+      "step": 1880
+    },
+    {
+      "epoch": 2.4399740428293315,
+      "grad_norm": 0.21397550791993689,
+      "learning_rate": 1.0341510341510342e-05,
+      "loss": 0.3495,
+      "step": 1881
+    },
+    {
+      "epoch": 2.4412719013627515,
+      "grad_norm": 0.3807729201699182,
+      "learning_rate": 1.0317460317460318e-05,
+      "loss": 0.3449,
+      "step": 1882
+    },
+    {
+      "epoch": 2.4425697598961715,
+      "grad_norm": 0.19473145374740133,
+      "learning_rate": 1.0293410293410295e-05,
+      "loss": 0.3412,
+      "step": 1883
+    },
+    {
+      "epoch": 2.443867618429591,
+      "grad_norm": 0.20643794857809838,
+      "learning_rate": 1.0269360269360271e-05,
+      "loss": 0.3548,
+      "step": 1884
+    },
+    {
+      "epoch": 2.445165476963011,
+      "grad_norm": 0.19967818102155932,
+      "learning_rate": 1.0245310245310246e-05,
+      "loss": 0.3251,
+      "step": 1885
+    },
+    {
+      "epoch": 2.446463335496431,
+      "grad_norm": 0.2068701964008534,
+      "learning_rate": 1.0221260221260222e-05,
+      "loss": 0.3503,
+      "step": 1886
+    },
+    {
+      "epoch": 2.4477611940298507,
+      "grad_norm": 0.19485423370421984,
+      "learning_rate": 1.0197210197210199e-05,
+      "loss": 0.3316,
+      "step": 1887
+    },
+    {
+      "epoch": 2.4490590525632707,
+      "grad_norm": 0.20945089504608722,
+      "learning_rate": 1.0173160173160174e-05,
+      "loss": 0.3723,
+      "step": 1888
+    },
+    {
+      "epoch": 2.4503569110966903,
+      "grad_norm": 0.21961631414794303,
+      "learning_rate": 1.014911014911015e-05,
+      "loss": 0.3555,
+      "step": 1889
+    },
+    {
+      "epoch": 2.4516547696301103,
+      "grad_norm": 0.20913009484803424,
+      "learning_rate": 1.0125060125060125e-05,
+      "loss": 0.3433,
+      "step": 1890
+    },
+    {
+      "epoch": 2.4529526281635303,
+      "grad_norm": 0.1975805461078145,
+      "learning_rate": 1.0101010101010101e-05,
+      "loss": 0.3522,
+      "step": 1891
+    },
+    {
+      "epoch": 2.45425048669695,
+      "grad_norm": 0.18824749115573988,
+      "learning_rate": 1.0076960076960078e-05,
+      "loss": 0.3262,
+      "step": 1892
+    },
+    {
+      "epoch": 2.45554834523037,
+      "grad_norm": 0.19363390712933798,
+      "learning_rate": 1.0052910052910053e-05,
+      "loss": 0.3304,
+      "step": 1893
+    },
+    {
+      "epoch": 2.45684620376379,
+      "grad_norm": 0.20877531448498393,
+      "learning_rate": 1.0028860028860029e-05,
+      "loss": 0.3502,
+      "step": 1894
+    },
+    {
+      "epoch": 2.4581440622972095,
+      "grad_norm": 0.2061242277033731,
+      "learning_rate": 1.0004810004810006e-05,
+      "loss": 0.3382,
+      "step": 1895
+    },
+    {
+      "epoch": 2.4594419208306295,
+      "grad_norm": 0.20527048632536887,
+      "learning_rate": 9.98075998075998e-06,
+      "loss": 0.3391,
+      "step": 1896
+    },
+    {
+      "epoch": 2.460739779364049,
+      "grad_norm": 0.20055534262640298,
+      "learning_rate": 9.956709956709957e-06,
+      "loss": 0.3564,
+      "step": 1897
+    },
+    {
+      "epoch": 2.462037637897469,
+      "grad_norm": 0.19785197665929594,
+      "learning_rate": 9.932659932659933e-06,
+      "loss": 0.3443,
+      "step": 1898
+    },
+    {
+      "epoch": 2.463335496430889,
+      "grad_norm": 0.2037702638037453,
+      "learning_rate": 9.908609908609908e-06,
+      "loss": 0.348,
+      "step": 1899
+    },
+    {
+      "epoch": 2.4646333549643087,
+      "grad_norm": 0.21220856247877268,
+      "learning_rate": 9.884559884559884e-06,
+      "loss": 0.3428,
+      "step": 1900
+    },
+    {
+      "epoch": 2.4659312134977287,
+      "grad_norm": 0.2219575251336399,
+      "learning_rate": 9.860509860509861e-06,
+      "loss": 0.3418,
+      "step": 1901
+    },
+    {
+      "epoch": 2.4672290720311487,
+      "grad_norm": 0.19647517582803717,
+      "learning_rate": 9.836459836459837e-06,
+      "loss": 0.3345,
+      "step": 1902
+    },
+    {
+      "epoch": 2.4685269305645683,
+      "grad_norm": 0.19654210291881336,
+      "learning_rate": 9.812409812409814e-06,
+      "loss": 0.3439,
+      "step": 1903
+    },
+    {
+      "epoch": 2.4698247890979883,
+      "grad_norm": 0.20215632118085225,
+      "learning_rate": 9.788359788359789e-06,
+      "loss": 0.343,
+      "step": 1904
+    },
+    {
+      "epoch": 2.4711226476314083,
+      "grad_norm": 0.21700717313986337,
+      "learning_rate": 9.764309764309765e-06,
+      "loss": 0.3422,
+      "step": 1905
+    },
+    {
+      "epoch": 2.472420506164828,
+      "grad_norm": 0.22639821333763582,
+      "learning_rate": 9.740259740259742e-06,
+      "loss": 0.3545,
+      "step": 1906
+    },
+    {
+      "epoch": 2.473718364698248,
+      "grad_norm": 0.20194895913978017,
+      "learning_rate": 9.716209716209716e-06,
+      "loss": 0.3389,
+      "step": 1907
+    },
+    {
+      "epoch": 2.475016223231668,
+      "grad_norm": 0.20577729584744323,
+      "learning_rate": 9.692159692159693e-06,
+      "loss": 0.3399,
+      "step": 1908
+    },
+    {
+      "epoch": 2.4763140817650875,
+      "grad_norm": 0.19912299573383097,
+      "learning_rate": 9.66810966810967e-06,
+      "loss": 0.3373,
+      "step": 1909
+    },
+    {
+      "epoch": 2.4776119402985075,
+      "grad_norm": 0.21269657354433175,
+      "learning_rate": 9.644059644059644e-06,
+      "loss": 0.3266,
+      "step": 1910
+    },
+    {
+      "epoch": 2.4789097988319275,
+      "grad_norm": 0.20736564955023581,
+      "learning_rate": 9.62000962000962e-06,
+      "loss": 0.3778,
+      "step": 1911
+    },
+    {
+      "epoch": 2.480207657365347,
+      "grad_norm": 0.20499397029892405,
+      "learning_rate": 9.595959595959595e-06,
+      "loss": 0.3476,
+      "step": 1912
+    },
+    {
+      "epoch": 2.481505515898767,
+      "grad_norm": 0.21222717155575171,
+      "learning_rate": 9.571909571909572e-06,
+      "loss": 0.3614,
+      "step": 1913
+    },
+    {
+      "epoch": 2.482803374432187,
+      "grad_norm": 0.19863505520154515,
+      "learning_rate": 9.547859547859548e-06,
+      "loss": 0.3405,
+      "step": 1914
+    },
+    {
+      "epoch": 2.4841012329656067,
+      "grad_norm": 0.18907377927162114,
+      "learning_rate": 9.523809523809523e-06,
+      "loss": 0.3286,
+      "step": 1915
+    },
+    {
+      "epoch": 2.4853990914990267,
+      "grad_norm": 0.21676854818994787,
+      "learning_rate": 9.4997594997595e-06,
+      "loss": 0.3464,
+      "step": 1916
+    },
+    {
+      "epoch": 2.4866969500324463,
+      "grad_norm": 0.20682249845168457,
+      "learning_rate": 9.475709475709476e-06,
+      "loss": 0.3482,
+      "step": 1917
+    },
+    {
+      "epoch": 2.4879948085658663,
+      "grad_norm": 0.20359288013276863,
+      "learning_rate": 9.451659451659452e-06,
+      "loss": 0.3361,
+      "step": 1918
+    },
+    {
+      "epoch": 2.4892926670992863,
+      "grad_norm": 0.20782718092266766,
+      "learning_rate": 9.427609427609429e-06,
+      "loss": 0.3437,
+      "step": 1919
+    },
+    {
+      "epoch": 2.490590525632706,
+      "grad_norm": 0.19751302818390967,
+      "learning_rate": 9.403559403559405e-06,
+      "loss": 0.3291,
+      "step": 1920
+    },
+    {
+      "epoch": 2.491888384166126,
+      "grad_norm": 0.21156577333606844,
+      "learning_rate": 9.37950937950938e-06,
+      "loss": 0.3402,
+      "step": 1921
+    },
+    {
+      "epoch": 2.493186242699546,
+      "grad_norm": 0.20413132600430678,
+      "learning_rate": 9.355459355459357e-06,
+      "loss": 0.3483,
+      "step": 1922
+    },
+    {
+      "epoch": 2.4944841012329655,
+      "grad_norm": 0.1966838578810161,
+      "learning_rate": 9.331409331409331e-06,
+      "loss": 0.3412,
+      "step": 1923
+    },
+    {
+      "epoch": 2.4957819597663855,
+      "grad_norm": 0.19973328891467468,
+      "learning_rate": 9.307359307359308e-06,
+      "loss": 0.3458,
+      "step": 1924
+    },
+    {
+      "epoch": 2.497079818299805,
+      "grad_norm": 0.2020638046963019,
+      "learning_rate": 9.283309283309284e-06,
+      "loss": 0.3393,
+      "step": 1925
+    },
+    {
+      "epoch": 2.498377676833225,
+      "grad_norm": 0.22766804942928376,
+      "learning_rate": 9.259259259259259e-06,
+      "loss": 0.3255,
+      "step": 1926
+    },
+    {
+      "epoch": 2.499675535366645,
+      "grad_norm": 0.1997890343589566,
+      "learning_rate": 9.235209235209236e-06,
+      "loss": 0.3379,
+      "step": 1927
+    },
+    {
+      "epoch": 2.5009733939000647,
+      "grad_norm": 0.19417057565689014,
+      "learning_rate": 9.211159211159212e-06,
+      "loss": 0.3362,
+      "step": 1928
+    },
+    {
+      "epoch": 2.5022712524334847,
+      "grad_norm": 0.2159121005299441,
+      "learning_rate": 9.187109187109187e-06,
+      "loss": 0.333,
+      "step": 1929
+    },
+    {
+      "epoch": 2.5035691109669047,
+      "grad_norm": 0.2152179356087276,
+      "learning_rate": 9.163059163059163e-06,
+      "loss": 0.3299,
+      "step": 1930
+    },
+    {
+      "epoch": 2.5048669695003243,
+      "grad_norm": 0.19905359149185228,
+      "learning_rate": 9.13900913900914e-06,
+      "loss": 0.3402,
+      "step": 1931
+    },
+    {
+      "epoch": 2.5061648280337443,
+      "grad_norm": 0.20691948697032309,
+      "learning_rate": 9.114959114959115e-06,
+      "loss": 0.3513,
+      "step": 1932
+    },
+    {
+      "epoch": 2.5074626865671643,
+      "grad_norm": 0.19894235448528472,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.3288,
+      "step": 1933
+    },
+    {
+      "epoch": 2.508760545100584,
+      "grad_norm": 0.21082966995035995,
+      "learning_rate": 9.066859066859068e-06,
+      "loss": 0.3332,
+      "step": 1934
+    },
+    {
+      "epoch": 2.510058403634004,
+      "grad_norm": 0.20169901454411296,
+      "learning_rate": 9.042809042809042e-06,
+      "loss": 0.3223,
+      "step": 1935
+    },
+    {
+      "epoch": 2.511356262167424,
+      "grad_norm": 0.21866454354506568,
+      "learning_rate": 9.018759018759019e-06,
+      "loss": 0.3392,
+      "step": 1936
+    },
+    {
+      "epoch": 2.5126541207008435,
+      "grad_norm": 0.21257413698744998,
+      "learning_rate": 8.994708994708995e-06,
+      "loss": 0.3452,
+      "step": 1937
+    },
+    {
+      "epoch": 2.5139519792342635,
+      "grad_norm": 0.19946752360820091,
+      "learning_rate": 8.970658970658972e-06,
+      "loss": 0.3426,
+      "step": 1938
+    },
+    {
+      "epoch": 2.5152498377676835,
+      "grad_norm": 0.2048047114149695,
+      "learning_rate": 8.946608946608948e-06,
+      "loss": 0.3496,
+      "step": 1939
+    },
+    {
+      "epoch": 2.516547696301103,
+      "grad_norm": 0.2060713206544726,
+      "learning_rate": 8.922558922558923e-06,
+      "loss": 0.3466,
+      "step": 1940
+    },
+    {
+      "epoch": 2.517845554834523,
+      "grad_norm": 0.21499517524206285,
+      "learning_rate": 8.8985088985089e-06,
+      "loss": 0.3391,
+      "step": 1941
+    },
+    {
+      "epoch": 2.519143413367943,
+      "grad_norm": 0.2043639715896453,
+      "learning_rate": 8.874458874458876e-06,
+      "loss": 0.3463,
+      "step": 1942
+    },
+    {
+      "epoch": 2.5204412719013627,
+      "grad_norm": 0.2183414837407522,
+      "learning_rate": 8.85040885040885e-06,
+      "loss": 0.3427,
+      "step": 1943
+    },
+    {
+      "epoch": 2.5217391304347827,
+      "grad_norm": 0.20285948319637043,
+      "learning_rate": 8.826358826358827e-06,
+      "loss": 0.3377,
+      "step": 1944
+    },
+    {
+      "epoch": 2.5230369889682027,
+      "grad_norm": 0.19391653715006987,
+      "learning_rate": 8.802308802308802e-06,
+      "loss": 0.3431,
+      "step": 1945
+    },
+    {
+      "epoch": 2.5243348475016223,
+      "grad_norm": 0.19725005677345353,
+      "learning_rate": 8.778258778258778e-06,
+      "loss": 0.3402,
+      "step": 1946
+    },
+    {
+      "epoch": 2.5256327060350423,
+      "grad_norm": 0.20776675755315593,
+      "learning_rate": 8.754208754208755e-06,
+      "loss": 0.3432,
+      "step": 1947
+    },
+    {
+      "epoch": 2.526930564568462,
+      "grad_norm": 0.20770257485418178,
+      "learning_rate": 8.73015873015873e-06,
+      "loss": 0.3309,
+      "step": 1948
+    },
+    {
+      "epoch": 2.528228423101882,
+      "grad_norm": 0.19473645944952384,
+      "learning_rate": 8.706108706108706e-06,
+      "loss": 0.3335,
+      "step": 1949
+    },
+    {
+      "epoch": 2.529526281635302,
+      "grad_norm": 0.20760911826948453,
+      "learning_rate": 8.682058682058683e-06,
+      "loss": 0.3418,
+      "step": 1950
+    },
+    {
+      "epoch": 2.5308241401687215,
+      "grad_norm": 0.2094780858154728,
+      "learning_rate": 8.658008658008657e-06,
+      "loss": 0.3425,
+      "step": 1951
+    },
+    {
+      "epoch": 2.5321219987021415,
+      "grad_norm": 0.19840367931370975,
+      "learning_rate": 8.633958633958634e-06,
+      "loss": 0.3417,
+      "step": 1952
+    },
+    {
+      "epoch": 2.533419857235561,
+      "grad_norm": 0.19624410854082408,
+      "learning_rate": 8.60990860990861e-06,
+      "loss": 0.3456,
+      "step": 1953
+    },
+    {
+      "epoch": 2.534717715768981,
+      "grad_norm": 0.19653675310930965,
+      "learning_rate": 8.585858585858587e-06,
+      "loss": 0.3287,
+      "step": 1954
+    },
+    {
+      "epoch": 2.536015574302401,
+      "grad_norm": 0.2033148019089131,
+      "learning_rate": 8.561808561808563e-06,
+      "loss": 0.3416,
+      "step": 1955
+    },
+    {
+      "epoch": 2.5373134328358207,
+      "grad_norm": 0.19360893151616687,
+      "learning_rate": 8.537758537758538e-06,
+      "loss": 0.3554,
+      "step": 1956
+    },
+    {
+      "epoch": 2.5386112913692407,
+      "grad_norm": 0.18776659365965748,
+      "learning_rate": 8.513708513708514e-06,
+      "loss": 0.3394,
+      "step": 1957
+    },
+    {
+      "epoch": 2.5399091499026607,
+      "grad_norm": 0.21687377756718906,
+      "learning_rate": 8.489658489658491e-06,
+      "loss": 0.3447,
+      "step": 1958
+    },
+    {
+      "epoch": 2.5412070084360803,
+      "grad_norm": 0.19788163493461905,
+      "learning_rate": 8.465608465608466e-06,
+      "loss": 0.3383,
+      "step": 1959
+    },
+    {
+      "epoch": 2.5425048669695003,
+      "grad_norm": 0.20991331629937682,
+      "learning_rate": 8.441558441558442e-06,
+      "loss": 0.3405,
+      "step": 1960
+    },
+    {
+      "epoch": 2.5438027255029203,
+      "grad_norm": 0.20315647768031084,
+      "learning_rate": 8.417508417508419e-06,
+      "loss": 0.3454,
+      "step": 1961
+    },
+    {
+      "epoch": 2.54510058403634,
+      "grad_norm": 0.23523075305367147,
+      "learning_rate": 8.393458393458393e-06,
+      "loss": 0.3618,
+      "step": 1962
+    },
+    {
+      "epoch": 2.54639844256976,
+      "grad_norm": 0.2157971364103013,
+      "learning_rate": 8.36940836940837e-06,
+      "loss": 0.3497,
+      "step": 1963
+    },
+    {
+      "epoch": 2.54769630110318,
+      "grad_norm": 0.20499915960426765,
+      "learning_rate": 8.345358345358346e-06,
+      "loss": 0.3309,
+      "step": 1964
+    },
+    {
+      "epoch": 2.5489941596365995,
+      "grad_norm": 0.20698252650858404,
+      "learning_rate": 8.321308321308321e-06,
+      "loss": 0.3331,
+      "step": 1965
+    },
+    {
+      "epoch": 2.5502920181700195,
+      "grad_norm": 0.20188066664017376,
+      "learning_rate": 8.297258297258298e-06,
+      "loss": 0.346,
+      "step": 1966
+    },
+    {
+      "epoch": 2.5515898767034395,
+      "grad_norm": 0.22092684369504248,
+      "learning_rate": 8.273208273208272e-06,
+      "loss": 0.3487,
+      "step": 1967
+    },
+    {
+      "epoch": 2.552887735236859,
+      "grad_norm": 0.20340970119145418,
+      "learning_rate": 8.249158249158249e-06,
+      "loss": 0.3341,
+      "step": 1968
+    },
+    {
+      "epoch": 2.554185593770279,
+      "grad_norm": 0.20252572395415291,
+      "learning_rate": 8.225108225108225e-06,
+      "loss": 0.3332,
+      "step": 1969
+    },
+    {
+      "epoch": 2.555483452303699,
+      "grad_norm": 0.1879834295873596,
+      "learning_rate": 8.201058201058202e-06,
+      "loss": 0.3212,
+      "step": 1970
+    },
+    {
+      "epoch": 2.5567813108371187,
+      "grad_norm": 0.19299674074621231,
+      "learning_rate": 8.177008177008178e-06,
+      "loss": 0.3404,
+      "step": 1971
+    },
+    {
+      "epoch": 2.5580791693705387,
+      "grad_norm": 0.20796973098772337,
+      "learning_rate": 8.152958152958155e-06,
+      "loss": 0.3389,
+      "step": 1972
+    },
+    {
+      "epoch": 2.5593770279039587,
+      "grad_norm": 0.21058263937992205,
+      "learning_rate": 8.12890812890813e-06,
+      "loss": 0.3382,
+      "step": 1973
+    },
+    {
+      "epoch": 2.5606748864373783,
+      "grad_norm": 0.19698041890440957,
+      "learning_rate": 8.104858104858106e-06,
+      "loss": 0.3406,
+      "step": 1974
+    },
+    {
+      "epoch": 2.5619727449707983,
+      "grad_norm": 0.20441051012340986,
+      "learning_rate": 8.080808080808082e-06,
+      "loss": 0.3568,
+      "step": 1975
+    },
+    {
+      "epoch": 2.5632706035042183,
+      "grad_norm": 0.20023706231769592,
+      "learning_rate": 8.056758056758057e-06,
+      "loss": 0.3453,
+      "step": 1976
+    },
+    {
+      "epoch": 2.564568462037638,
+      "grad_norm": 0.19384483347770198,
+      "learning_rate": 8.032708032708034e-06,
+      "loss": 0.3541,
+      "step": 1977
+    },
+    {
+      "epoch": 2.565866320571058,
+      "grad_norm": 0.1902457140768143,
+      "learning_rate": 8.008658008658008e-06,
+      "loss": 0.3259,
+      "step": 1978
+    },
+    {
+      "epoch": 2.5671641791044775,
+      "grad_norm": 0.2065028347871094,
+      "learning_rate": 7.984607984607985e-06,
+      "loss": 0.3543,
+      "step": 1979
+    },
+    {
+      "epoch": 2.5684620376378975,
+      "grad_norm": 0.19772657046385608,
+      "learning_rate": 7.960557960557961e-06,
+      "loss": 0.3353,
+      "step": 1980
+    },
+    {
+      "epoch": 2.569759896171317,
+      "grad_norm": 0.18849021172503813,
+      "learning_rate": 7.936507936507936e-06,
+      "loss": 0.3223,
+      "step": 1981
+    },
+    {
+      "epoch": 2.571057754704737,
+      "grad_norm": 0.208888741548615,
+      "learning_rate": 7.912457912457913e-06,
+      "loss": 0.3444,
+      "step": 1982
+    },
+    {
+      "epoch": 2.572355613238157,
+      "grad_norm": 0.22525608656131163,
+      "learning_rate": 7.888407888407889e-06,
+      "loss": 0.3361,
+      "step": 1983
+    },
+    {
+      "epoch": 2.5736534717715767,
+      "grad_norm": 0.207515663353734,
+      "learning_rate": 7.864357864357864e-06,
+      "loss": 0.3368,
+      "step": 1984
+    },
+    {
+      "epoch": 2.5749513303049967,
+      "grad_norm": 0.2154517901626009,
+      "learning_rate": 7.84030784030784e-06,
+      "loss": 0.3345,
+      "step": 1985
+    },
+    {
+      "epoch": 2.5762491888384167,
+      "grad_norm": 0.2059272342950417,
+      "learning_rate": 7.816257816257815e-06,
+      "loss": 0.3657,
+      "step": 1986
+    },
+    {
+      "epoch": 2.5775470473718363,
+      "grad_norm": 0.20247815650907755,
+      "learning_rate": 7.792207792207792e-06,
+      "loss": 0.3248,
+      "step": 1987
+    },
+    {
+      "epoch": 2.5788449059052563,
+      "grad_norm": 0.18967243424535923,
+      "learning_rate": 7.768157768157768e-06,
+      "loss": 0.3345,
+      "step": 1988
+    },
+    {
+      "epoch": 2.5801427644386763,
+      "grad_norm": 0.20514510031749197,
+      "learning_rate": 7.744107744107745e-06,
+      "loss": 0.3407,
+      "step": 1989
+    },
+    {
+      "epoch": 2.581440622972096,
+      "grad_norm": 0.2047332936097657,
+      "learning_rate": 7.720057720057721e-06,
+      "loss": 0.3404,
+      "step": 1990
+    },
+    {
+      "epoch": 2.582738481505516,
+      "grad_norm": 0.19001160907023656,
+      "learning_rate": 7.696007696007697e-06,
+      "loss": 0.3363,
+      "step": 1991
+    },
+    {
+      "epoch": 2.584036340038936,
+      "grad_norm": 0.2007791373982576,
+      "learning_rate": 7.671957671957672e-06,
+      "loss": 0.345,
+      "step": 1992
+    },
+    {
+      "epoch": 2.5853341985723555,
+      "grad_norm": 0.1995014007677766,
+      "learning_rate": 7.647907647907649e-06,
+      "loss": 0.3373,
+      "step": 1993
+    },
+    {
+      "epoch": 2.5866320571057755,
+      "grad_norm": 0.22960201823541218,
+      "learning_rate": 7.623857623857625e-06,
+      "loss": 0.35,
+      "step": 1994
+    },
+    {
+      "epoch": 2.5879299156391955,
+      "grad_norm": 0.2041708033798508,
+      "learning_rate": 7.5998075998076e-06,
+      "loss": 0.3299,
+      "step": 1995
+    },
+    {
+      "epoch": 2.589227774172615,
+      "grad_norm": 0.19021132164070584,
+      "learning_rate": 7.5757575757575764e-06,
+      "loss": 0.3434,
+      "step": 1996
+    },
+    {
+      "epoch": 2.590525632706035,
+      "grad_norm": 0.19160256744066273,
+      "learning_rate": 7.551707551707551e-06,
+      "loss": 0.3376,
+      "step": 1997
+    },
+    {
+      "epoch": 2.591823491239455,
+      "grad_norm": 0.1955416018816942,
+      "learning_rate": 7.527657527657528e-06,
+      "loss": 0.3407,
+      "step": 1998
+    },
+    {
+      "epoch": 2.5931213497728747,
+      "grad_norm": 0.1994363091424892,
+      "learning_rate": 7.503607503607504e-06,
+      "loss": 0.3271,
+      "step": 1999
+    },
+    {
+      "epoch": 2.5944192083062947,
+      "grad_norm": 0.19719206828799127,
+      "learning_rate": 7.47955747955748e-06,
+      "loss": 0.3491,
+      "step": 2000
+    },
+    {
+      "epoch": 2.5957170668397147,
+      "grad_norm": 0.21854928408023744,
+      "learning_rate": 7.455507455507456e-06,
+      "loss": 0.3559,
+      "step": 2001
+    },
+    {
+      "epoch": 2.5970149253731343,
+      "grad_norm": 0.2004433126098414,
+      "learning_rate": 7.431457431457433e-06,
+      "loss": 0.353,
+      "step": 2002
+    },
+    {
+      "epoch": 2.5983127839065543,
+      "grad_norm": 0.21521520240395595,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.3533,
+      "step": 2003
+    },
+    {
+      "epoch": 2.5996106424399743,
+      "grad_norm": 0.1904741360498635,
+      "learning_rate": 7.383357383357384e-06,
+      "loss": 0.3375,
+      "step": 2004
+    },
+    {
+      "epoch": 2.600908500973394,
+      "grad_norm": 0.20148594815991572,
+      "learning_rate": 7.35930735930736e-06,
+      "loss": 0.3404,
+      "step": 2005
+    },
+    {
+      "epoch": 2.602206359506814,
+      "grad_norm": 0.20222565117203797,
+      "learning_rate": 7.335257335257335e-06,
+      "loss": 0.3325,
+      "step": 2006
+    },
+    {
+      "epoch": 2.6035042180402335,
+      "grad_norm": 0.2086022602926794,
+      "learning_rate": 7.311207311207312e-06,
+      "loss": 0.3323,
+      "step": 2007
+    },
+    {
+      "epoch": 2.6048020765736535,
+      "grad_norm": 0.19695846789935645,
+      "learning_rate": 7.2871572871572864e-06,
+      "loss": 0.3419,
+      "step": 2008
+    },
+    {
+      "epoch": 2.6060999351070735,
+      "grad_norm": 0.2003548511519371,
+      "learning_rate": 7.263107263107263e-06,
+      "loss": 0.3356,
+      "step": 2009
+    },
+    {
+      "epoch": 2.607397793640493,
+      "grad_norm": 0.20970980037255266,
+      "learning_rate": 7.239057239057239e-06,
+      "loss": 0.3385,
+      "step": 2010
+    },
+    {
+      "epoch": 2.608695652173913,
+      "grad_norm": 0.19898685255635887,
+      "learning_rate": 7.215007215007215e-06,
+      "loss": 0.3427,
+      "step": 2011
+    },
+    {
+      "epoch": 2.6099935107073327,
+      "grad_norm": 0.20226511821516943,
+      "learning_rate": 7.1909571909571915e-06,
+      "loss": 0.3272,
+      "step": 2012
+    },
+    {
+      "epoch": 2.6112913692407527,
+      "grad_norm": 0.18832236422091425,
+      "learning_rate": 7.166907166907168e-06,
+      "loss": 0.3354,
+      "step": 2013
+    },
+    {
+      "epoch": 2.6125892277741727,
+      "grad_norm": 0.18327010064211902,
+      "learning_rate": 7.142857142857143e-06,
+      "loss": 0.3286,
+      "step": 2014
+    },
+    {
+      "epoch": 2.6138870863075923,
+      "grad_norm": 0.19895899041097867,
+      "learning_rate": 7.118807118807119e-06,
+      "loss": 0.3297,
+      "step": 2015
+    },
+    {
+      "epoch": 2.6151849448410123,
+      "grad_norm": 0.2069075643758518,
+      "learning_rate": 7.094757094757096e-06,
+      "loss": 0.3302,
+      "step": 2016
+    },
+    {
+      "epoch": 2.6164828033744323,
+      "grad_norm": 0.19524313780033972,
+      "learning_rate": 7.0707070707070704e-06,
+      "loss": 0.3365,
+      "step": 2017
+    },
+    {
+      "epoch": 2.617780661907852,
+      "grad_norm": 0.19938059659766813,
+      "learning_rate": 7.046657046657047e-06,
+      "loss": 0.3338,
+      "step": 2018
+    },
+    {
+      "epoch": 2.619078520441272,
+      "grad_norm": 0.19493636421814658,
+      "learning_rate": 7.0226070226070225e-06,
+      "loss": 0.3315,
+      "step": 2019
+    },
+    {
+      "epoch": 2.620376378974692,
+      "grad_norm": 0.19839871511366805,
+      "learning_rate": 6.998556998556999e-06,
+      "loss": 0.3328,
+      "step": 2020
+    },
+    {
+      "epoch": 2.6216742375081115,
+      "grad_norm": 0.20131069199570611,
+      "learning_rate": 6.9745069745069755e-06,
+      "loss": 0.3509,
+      "step": 2021
+    },
+    {
+      "epoch": 2.6229720960415315,
+      "grad_norm": 0.19462988449384028,
+      "learning_rate": 6.95045695045695e-06,
+      "loss": 0.3491,
+      "step": 2022
+    },
+    {
+      "epoch": 2.6242699545749515,
+      "grad_norm": 0.21282789287292148,
+      "learning_rate": 6.926406926406927e-06,
+      "loss": 0.34,
+      "step": 2023
+    },
+    {
+      "epoch": 2.625567813108371,
+      "grad_norm": 0.1993074561130325,
+      "learning_rate": 6.902356902356903e-06,
+      "loss": 0.3466,
+      "step": 2024
+    },
+    {
+      "epoch": 2.626865671641791,
+      "grad_norm": 0.20351150350698377,
+      "learning_rate": 6.878306878306878e-06,
+      "loss": 0.3598,
+      "step": 2025
+    },
+    {
+      "epoch": 2.628163530175211,
+      "grad_norm": 0.19771094464637703,
+      "learning_rate": 6.854256854256854e-06,
+      "loss": 0.336,
+      "step": 2026
+    },
+    {
+      "epoch": 2.6294613887086307,
+      "grad_norm": 0.21482701918153888,
+      "learning_rate": 6.830206830206831e-06,
+      "loss": 0.34,
+      "step": 2027
+    },
+    {
+      "epoch": 2.6307592472420507,
+      "grad_norm": 0.2003506016053636,
+      "learning_rate": 6.8061568061568065e-06,
+      "loss": 0.3337,
+      "step": 2028
+    },
+    {
+      "epoch": 2.6320571057754707,
+      "grad_norm": 0.19105711872310324,
+      "learning_rate": 6.782106782106783e-06,
+      "loss": 0.3384,
+      "step": 2029
+    },
+    {
+      "epoch": 2.6333549643088903,
+      "grad_norm": 0.20159758006212594,
+      "learning_rate": 6.758056758056758e-06,
+      "loss": 0.3375,
+      "step": 2030
+    },
+    {
+      "epoch": 2.6346528228423103,
+      "grad_norm": 0.20143060805909305,
+      "learning_rate": 6.734006734006734e-06,
+      "loss": 0.3491,
+      "step": 2031
+    },
+    {
+      "epoch": 2.6359506813757303,
+      "grad_norm": 0.19866829942887484,
+      "learning_rate": 6.709956709956711e-06,
+      "loss": 0.3376,
+      "step": 2032
+    },
+    {
+      "epoch": 2.63724853990915,
+      "grad_norm": 0.2037924476485615,
+      "learning_rate": 6.6859066859066855e-06,
+      "loss": 0.3353,
+      "step": 2033
+    },
+    {
+      "epoch": 2.63854639844257,
+      "grad_norm": 0.19840737231971423,
+      "learning_rate": 6.661856661856662e-06,
+      "loss": 0.3419,
+      "step": 2034
+    },
+    {
+      "epoch": 2.6398442569759895,
+      "grad_norm": 0.2012069800711804,
+      "learning_rate": 6.637806637806638e-06,
+      "loss": 0.3374,
+      "step": 2035
+    },
+    {
+      "epoch": 2.6411421155094095,
+      "grad_norm": 0.2083568986903432,
+      "learning_rate": 6.613756613756614e-06,
+      "loss": 0.3535,
+      "step": 2036
+    },
+    {
+      "epoch": 2.6424399740428295,
+      "grad_norm": 0.19697725314574688,
+      "learning_rate": 6.5897065897065905e-06,
+      "loss": 0.3449,
+      "step": 2037
+    },
+    {
+      "epoch": 2.643737832576249,
+      "grad_norm": 0.19818150221871475,
+      "learning_rate": 6.565656565656567e-06,
+      "loss": 0.3409,
+      "step": 2038
+    },
+    {
+      "epoch": 2.645035691109669,
+      "grad_norm": 0.1978527267292534,
+      "learning_rate": 6.541606541606542e-06,
+      "loss": 0.3373,
+      "step": 2039
+    },
+    {
+      "epoch": 2.6463335496430886,
+      "grad_norm": 0.18635435314872623,
+      "learning_rate": 6.517556517556518e-06,
+      "loss": 0.3342,
+      "step": 2040
+    },
+    {
+      "epoch": 2.6476314081765087,
+      "grad_norm": 0.19790356410543516,
+      "learning_rate": 6.493506493506493e-06,
+      "loss": 0.3436,
+      "step": 2041
+    },
+    {
+      "epoch": 2.6489292667099287,
+      "grad_norm": 0.20316872420087156,
+      "learning_rate": 6.4694564694564695e-06,
+      "loss": 0.3448,
+      "step": 2042
+    },
+    {
+      "epoch": 2.6502271252433482,
+      "grad_norm": 0.19838963723019196,
+      "learning_rate": 6.445406445406446e-06,
+      "loss": 0.3467,
+      "step": 2043
+    },
+    {
+      "epoch": 2.6515249837767683,
+      "grad_norm": 0.19768166770591664,
+      "learning_rate": 6.4213564213564216e-06,
+      "loss": 0.3412,
+      "step": 2044
+    },
+    {
+      "epoch": 2.6528228423101883,
+      "grad_norm": 0.18870632230633216,
+      "learning_rate": 6.397306397306398e-06,
+      "loss": 0.3451,
+      "step": 2045
+    },
+    {
+      "epoch": 2.654120700843608,
+      "grad_norm": 0.19129582515190627,
+      "learning_rate": 6.3732563732563745e-06,
+      "loss": 0.3454,
+      "step": 2046
+    },
+    {
+      "epoch": 2.655418559377028,
+      "grad_norm": 0.20596889818566627,
+      "learning_rate": 6.349206349206349e-06,
+      "loss": 0.3534,
+      "step": 2047
+    },
+    {
+      "epoch": 2.656716417910448,
+      "grad_norm": 0.1951677844464869,
+      "learning_rate": 6.325156325156326e-06,
+      "loss": 0.3465,
+      "step": 2048
+    },
+    {
+      "epoch": 2.6580142764438675,
+      "grad_norm": 0.1839023960232542,
+      "learning_rate": 6.301106301106302e-06,
+      "loss": 0.3237,
+      "step": 2049
+    },
+    {
+      "epoch": 2.6593121349772875,
+      "grad_norm": 0.18411994909099402,
+      "learning_rate": 6.277056277056277e-06,
+      "loss": 0.3402,
+      "step": 2050
+    },
+    {
+      "epoch": 2.6606099935107075,
+      "grad_norm": 0.19031776842501638,
+      "learning_rate": 6.2530062530062535e-06,
+      "loss": 0.3354,
+      "step": 2051
+    },
+    {
+      "epoch": 2.661907852044127,
+      "grad_norm": 0.21384183187184516,
+      "learning_rate": 6.228956228956229e-06,
+      "loss": 0.342,
+      "step": 2052
+    },
+    {
+      "epoch": 2.663205710577547,
+      "grad_norm": 0.18478906620074495,
+      "learning_rate": 6.204906204906205e-06,
+      "loss": 0.3348,
+      "step": 2053
+    },
+    {
+      "epoch": 2.664503569110967,
+      "grad_norm": 0.20269895193238666,
+      "learning_rate": 6.180856180856181e-06,
+      "loss": 0.3351,
+      "step": 2054
+    },
+    {
+      "epoch": 2.6658014276443867,
+      "grad_norm": 0.20051102629513898,
+      "learning_rate": 6.156806156806158e-06,
+      "loss": 0.3385,
+      "step": 2055
+    },
+    {
+      "epoch": 2.6670992861778067,
+      "grad_norm": 0.18318606924324213,
+      "learning_rate": 6.132756132756133e-06,
+      "loss": 0.3364,
+      "step": 2056
+    },
+    {
+      "epoch": 2.6683971447112267,
+      "grad_norm": 0.19493171826087835,
+      "learning_rate": 6.108706108706109e-06,
+      "loss": 0.3348,
+      "step": 2057
+    },
+    {
+      "epoch": 2.6696950032446463,
+      "grad_norm": 0.18393836124851373,
+      "learning_rate": 6.0846560846560845e-06,
+      "loss": 0.3364,
+      "step": 2058
+    },
+    {
+      "epoch": 2.6709928617780663,
+      "grad_norm": 0.19573395789299228,
+      "learning_rate": 6.060606060606061e-06,
+      "loss": 0.3501,
+      "step": 2059
+    },
+    {
+      "epoch": 2.6722907203114863,
+      "grad_norm": 0.19289300424261566,
+      "learning_rate": 6.036556036556037e-06,
+      "loss": 0.3331,
+      "step": 2060
+    },
+    {
+      "epoch": 2.673588578844906,
+      "grad_norm": 0.20391450990957627,
+      "learning_rate": 6.012506012506012e-06,
+      "loss": 0.3523,
+      "step": 2061
+    },
+    {
+      "epoch": 2.674886437378326,
+      "grad_norm": 0.20029237980281211,
+      "learning_rate": 5.988455988455989e-06,
+      "loss": 0.3333,
+      "step": 2062
+    },
+    {
+      "epoch": 2.676184295911746,
+      "grad_norm": 0.1866620677693766,
+      "learning_rate": 5.964405964405965e-06,
+      "loss": 0.3311,
+      "step": 2063
+    },
+    {
+      "epoch": 2.6774821544451655,
+      "grad_norm": 0.19121945839733376,
+      "learning_rate": 5.940355940355941e-06,
+      "loss": 0.3411,
+      "step": 2064
+    },
+    {
+      "epoch": 2.6787800129785855,
+      "grad_norm": 0.19768175445567718,
+      "learning_rate": 5.916305916305916e-06,
+      "loss": 0.3331,
+      "step": 2065
+    },
+    {
+      "epoch": 2.680077871512005,
+      "grad_norm": 0.19564286501355985,
+      "learning_rate": 5.892255892255893e-06,
+      "loss": 0.3494,
+      "step": 2066
+    },
+    {
+      "epoch": 2.681375730045425,
+      "grad_norm": 0.18680263316102796,
+      "learning_rate": 5.8682058682058685e-06,
+      "loss": 0.3261,
+      "step": 2067
+    },
+    {
+      "epoch": 2.6826735885788446,
+      "grad_norm": 0.18888920820250896,
+      "learning_rate": 5.844155844155844e-06,
+      "loss": 0.3319,
+      "step": 2068
+    },
+    {
+      "epoch": 2.6839714471122647,
+      "grad_norm": 0.19397052933336428,
+      "learning_rate": 5.82010582010582e-06,
+      "loss": 0.3361,
+      "step": 2069
+    },
+    {
+      "epoch": 2.6852693056456847,
+      "grad_norm": 0.19077138187186174,
+      "learning_rate": 5.796055796055796e-06,
+      "loss": 0.3299,
+      "step": 2070
+    },
+    {
+      "epoch": 2.6865671641791042,
+      "grad_norm": 0.21127811092361148,
+      "learning_rate": 5.772005772005773e-06,
+      "loss": 0.3351,
+      "step": 2071
+    },
+    {
+      "epoch": 2.6878650227125243,
+      "grad_norm": 0.19439235797880994,
+      "learning_rate": 5.747955747955748e-06,
+      "loss": 0.3311,
+      "step": 2072
+    },
+    {
+      "epoch": 2.6891628812459443,
+      "grad_norm": 0.18768581482039637,
+      "learning_rate": 5.723905723905725e-06,
+      "loss": 0.3408,
+      "step": 2073
+    },
+    {
+      "epoch": 2.690460739779364,
+      "grad_norm": 0.20150498398104075,
+      "learning_rate": 5.6998556998557e-06,
+      "loss": 0.3297,
+      "step": 2074
+    },
+    {
+      "epoch": 2.691758598312784,
+      "grad_norm": 0.2030708714806736,
+      "learning_rate": 5.675805675805676e-06,
+      "loss": 0.3464,
+      "step": 2075
+    },
+    {
+      "epoch": 2.693056456846204,
+      "grad_norm": 0.19331515757587614,
+      "learning_rate": 5.651755651755652e-06,
+      "loss": 0.341,
+      "step": 2076
+    },
+    {
+      "epoch": 2.6943543153796234,
+      "grad_norm": 0.20425020368275043,
+      "learning_rate": 5.627705627705628e-06,
+      "loss": 0.3533,
+      "step": 2077
+    },
+    {
+      "epoch": 2.6956521739130435,
+      "grad_norm": 0.2332987934303732,
+      "learning_rate": 5.603655603655604e-06,
+      "loss": 0.3395,
+      "step": 2078
+    },
+    {
+      "epoch": 2.6969500324464635,
+      "grad_norm": 0.19691484544610982,
+      "learning_rate": 5.579605579605579e-06,
+      "loss": 0.3493,
+      "step": 2079
+    },
+    {
+      "epoch": 2.698247890979883,
+      "grad_norm": 0.2052051327790077,
+      "learning_rate": 5.555555555555556e-06,
+      "loss": 0.3622,
+      "step": 2080
+    },
+    {
+      "epoch": 2.699545749513303,
+      "grad_norm": 0.19298325427314583,
+      "learning_rate": 5.531505531505532e-06,
+      "loss": 0.3275,
+      "step": 2081
+    },
+    {
+      "epoch": 2.700843608046723,
+      "grad_norm": 0.19205634748728384,
+      "learning_rate": 5.507455507455508e-06,
+      "loss": 0.3364,
+      "step": 2082
+    },
+    {
+      "epoch": 2.7021414665801426,
+      "grad_norm": 0.199120001609843,
+      "learning_rate": 5.4834054834054835e-06,
+      "loss": 0.3604,
+      "step": 2083
+    },
+    {
+      "epoch": 2.7034393251135627,
+      "grad_norm": 0.19279039644707233,
+      "learning_rate": 5.45935545935546e-06,
+      "loss": 0.3531,
+      "step": 2084
+    },
+    {
+      "epoch": 2.7047371836469827,
+      "grad_norm": 0.19816454924229257,
+      "learning_rate": 5.435305435305436e-06,
+      "loss": 0.3386,
+      "step": 2085
+    },
+    {
+      "epoch": 2.7060350421804023,
+      "grad_norm": 0.1978192250026057,
+      "learning_rate": 5.411255411255411e-06,
+      "loss": 0.3353,
+      "step": 2086
+    },
+    {
+      "epoch": 2.7073329007138223,
+      "grad_norm": 0.1866947813546459,
+      "learning_rate": 5.387205387205387e-06,
+      "loss": 0.3361,
+      "step": 2087
+    },
+    {
+      "epoch": 2.7086307592472423,
+      "grad_norm": 0.19130243354442364,
+      "learning_rate": 5.363155363155363e-06,
+      "loss": 0.3339,
+      "step": 2088
+    },
+    {
+      "epoch": 2.709928617780662,
+      "grad_norm": 0.20123966523314857,
+      "learning_rate": 5.33910533910534e-06,
+      "loss": 0.352,
+      "step": 2089
+    },
+    {
+      "epoch": 2.711226476314082,
+      "grad_norm": 0.204082189254048,
+      "learning_rate": 5.3150553150553154e-06,
+      "loss": 0.3465,
+      "step": 2090
+    },
+    {
+      "epoch": 2.712524334847502,
+      "grad_norm": 0.18978137246025184,
+      "learning_rate": 5.291005291005291e-06,
+      "loss": 0.3436,
+      "step": 2091
+    },
+    {
+      "epoch": 2.7138221933809215,
+      "grad_norm": 0.18714999050678918,
+      "learning_rate": 5.2669552669552675e-06,
+      "loss": 0.3191,
+      "step": 2092
+    },
+    {
+      "epoch": 2.7151200519143415,
+      "grad_norm": 0.19155333104247205,
+      "learning_rate": 5.242905242905243e-06,
+      "loss": 0.3298,
+      "step": 2093
+    },
+    {
+      "epoch": 2.716417910447761,
+      "grad_norm": 0.19864886304787258,
+      "learning_rate": 5.218855218855219e-06,
+      "loss": 0.337,
+      "step": 2094
+    },
+    {
+      "epoch": 2.717715768981181,
+      "grad_norm": 0.20900011549166994,
+      "learning_rate": 5.194805194805195e-06,
+      "loss": 0.3351,
+      "step": 2095
+    },
+    {
+      "epoch": 2.719013627514601,
+      "grad_norm": 0.19283552005981475,
+      "learning_rate": 5.170755170755171e-06,
+      "loss": 0.3415,
+      "step": 2096
+    },
+    {
+      "epoch": 2.7203114860480206,
+      "grad_norm": 0.19735002707288396,
+      "learning_rate": 5.146705146705147e-06,
+      "loss": 0.3511,
+      "step": 2097
+    },
+    {
+      "epoch": 2.7216093445814407,
+      "grad_norm": 0.19007560174944538,
+      "learning_rate": 5.122655122655123e-06,
+      "loss": 0.3337,
+      "step": 2098
+    },
+    {
+      "epoch": 2.7229072031148602,
+      "grad_norm": 0.1945528470127734,
+      "learning_rate": 5.0986050986050994e-06,
+      "loss": 0.3333,
+      "step": 2099
+    },
+    {
+      "epoch": 2.7242050616482802,
+      "grad_norm": 0.19541816607187268,
+      "learning_rate": 5.074555074555075e-06,
+      "loss": 0.3375,
+      "step": 2100
+    },
+    {
+      "epoch": 2.7255029201817003,
+      "grad_norm": 0.1965347736989204,
+      "learning_rate": 5.050505050505051e-06,
+      "loss": 0.3374,
+      "step": 2101
+    },
+    {
+      "epoch": 2.72680077871512,
+      "grad_norm": 0.20176319781405136,
+      "learning_rate": 5.026455026455026e-06,
+      "loss": 0.3292,
+      "step": 2102
+    },
+    {
+      "epoch": 2.72809863724854,
+      "grad_norm": 0.20334136636920452,
+      "learning_rate": 5.002405002405003e-06,
+      "loss": 0.3546,
+      "step": 2103
+    },
+    {
+      "epoch": 2.72939649578196,
+      "grad_norm": 0.20179646454271286,
+      "learning_rate": 4.978354978354978e-06,
+      "loss": 0.3401,
+      "step": 2104
+    },
+    {
+      "epoch": 2.7306943543153794,
+      "grad_norm": 0.20239489126680496,
+      "learning_rate": 4.954304954304954e-06,
+      "loss": 0.362,
+      "step": 2105
+    },
+    {
+      "epoch": 2.7319922128487995,
+      "grad_norm": 0.1870990703654753,
+      "learning_rate": 4.9302549302549305e-06,
+      "loss": 0.3443,
+      "step": 2106
+    },
+    {
+      "epoch": 2.7332900713822195,
+      "grad_norm": 0.18557759206233268,
+      "learning_rate": 4.906204906204907e-06,
+      "loss": 0.3313,
+      "step": 2107
+    },
+    {
+      "epoch": 2.734587929915639,
+      "grad_norm": 0.19525861207810466,
+      "learning_rate": 4.8821548821548826e-06,
+      "loss": 0.3471,
+      "step": 2108
+    },
+    {
+      "epoch": 2.735885788449059,
+      "grad_norm": 0.1863895368451734,
+      "learning_rate": 4.858104858104858e-06,
+      "loss": 0.3302,
+      "step": 2109
+    },
+    {
+      "epoch": 2.737183646982479,
+      "grad_norm": 0.19520260521522104,
+      "learning_rate": 4.834054834054835e-06,
+      "loss": 0.3285,
+      "step": 2110
+    },
+    {
+      "epoch": 2.7384815055158986,
+      "grad_norm": 0.19296947043696586,
+      "learning_rate": 4.81000481000481e-06,
+      "loss": 0.3493,
+      "step": 2111
+    },
+    {
+      "epoch": 2.7397793640493187,
+      "grad_norm": 0.19207052807282113,
+      "learning_rate": 4.785954785954786e-06,
+      "loss": 0.3557,
+      "step": 2112
+    },
+    {
+      "epoch": 2.7410772225827387,
+      "grad_norm": 0.19274032451154754,
+      "learning_rate": 4.7619047619047615e-06,
+      "loss": 0.3363,
+      "step": 2113
+    },
+    {
+      "epoch": 2.7423750811161582,
+      "grad_norm": 0.1991025079179398,
+      "learning_rate": 4.737854737854738e-06,
+      "loss": 0.3383,
+      "step": 2114
+    },
+    {
+      "epoch": 2.7436729396495783,
+      "grad_norm": 0.20322726939044952,
+      "learning_rate": 4.7138047138047145e-06,
+      "loss": 0.3309,
+      "step": 2115
+    },
+    {
+      "epoch": 2.7449707981829983,
+      "grad_norm": 0.1953261545822402,
+      "learning_rate": 4.68975468975469e-06,
+      "loss": 0.3324,
+      "step": 2116
+    },
+    {
+      "epoch": 2.746268656716418,
+      "grad_norm": 0.20572870381079225,
+      "learning_rate": 4.665704665704666e-06,
+      "loss": 0.3519,
+      "step": 2117
+    },
+    {
+      "epoch": 2.747566515249838,
+      "grad_norm": 0.1896514715456808,
+      "learning_rate": 4.641654641654642e-06,
+      "loss": 0.3269,
+      "step": 2118
+    },
+    {
+      "epoch": 2.748864373783258,
+      "grad_norm": 0.20395631514698148,
+      "learning_rate": 4.617604617604618e-06,
+      "loss": 0.3378,
+      "step": 2119
+    },
+    {
+      "epoch": 2.7501622323166774,
+      "grad_norm": 0.18572879939359394,
+      "learning_rate": 4.5935545935545934e-06,
+      "loss": 0.3577,
+      "step": 2120
+    },
+    {
+      "epoch": 2.7514600908500975,
+      "grad_norm": 0.1888314312446457,
+      "learning_rate": 4.56950456950457e-06,
+      "loss": 0.3323,
+      "step": 2121
+    },
+    {
+      "epoch": 2.752757949383517,
+      "grad_norm": 0.19430952155026918,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.3439,
+      "step": 2122
+    },
+    {
+      "epoch": 2.754055807916937,
+      "grad_norm": 0.42636074237494337,
+      "learning_rate": 4.521404521404521e-06,
+      "loss": 0.3473,
+      "step": 2123
+    },
+    {
+      "epoch": 2.755353666450357,
+      "grad_norm": 0.2086969030917497,
+      "learning_rate": 4.497354497354498e-06,
+      "loss": 0.3651,
+      "step": 2124
+    },
+    {
+      "epoch": 2.7566515249837766,
+      "grad_norm": 0.18713017378956393,
+      "learning_rate": 4.473304473304474e-06,
+      "loss": 0.3391,
+      "step": 2125
+    },
+    {
+      "epoch": 2.7579493835171967,
+      "grad_norm": 0.1997677484748378,
+      "learning_rate": 4.44925444925445e-06,
+      "loss": 0.3296,
+      "step": 2126
+    },
+    {
+      "epoch": 2.7592472420506162,
+      "grad_norm": 0.19085510775569323,
+      "learning_rate": 4.425204425204425e-06,
+      "loss": 0.3306,
+      "step": 2127
+    },
+    {
+      "epoch": 2.7605451005840362,
+      "grad_norm": 0.1897332096553123,
+      "learning_rate": 4.401154401154401e-06,
+      "loss": 0.3319,
+      "step": 2128
+    },
+    {
+      "epoch": 2.7618429591174563,
+      "grad_norm": 0.18447234964642742,
+      "learning_rate": 4.377104377104377e-06,
+      "loss": 0.3381,
+      "step": 2129
+    },
+    {
+      "epoch": 2.763140817650876,
+      "grad_norm": 0.19659685588771536,
+      "learning_rate": 4.353054353054353e-06,
+      "loss": 0.3411,
+      "step": 2130
+    },
+    {
+      "epoch": 2.764438676184296,
+      "grad_norm": 0.19285636988233915,
+      "learning_rate": 4.329004329004329e-06,
+      "loss": 0.3363,
+      "step": 2131
+    },
+    {
+      "epoch": 2.765736534717716,
+      "grad_norm": 0.1873294345390938,
+      "learning_rate": 4.304954304954305e-06,
+      "loss": 0.3479,
+      "step": 2132
+    },
+    {
+      "epoch": 2.7670343932511354,
+      "grad_norm": 0.20641352605961297,
+      "learning_rate": 4.280904280904282e-06,
+      "loss": 0.3479,
+      "step": 2133
+    },
+    {
+      "epoch": 2.7683322517845554,
+      "grad_norm": 0.19611830886976564,
+      "learning_rate": 4.256854256854257e-06,
+      "loss": 0.3598,
+      "step": 2134
+    },
+    {
+      "epoch": 2.7696301103179755,
+      "grad_norm": 0.19362363231720492,
+      "learning_rate": 4.232804232804233e-06,
+      "loss": 0.3587,
+      "step": 2135
+    },
+    {
+      "epoch": 2.770927968851395,
+      "grad_norm": 0.18330247717729053,
+      "learning_rate": 4.208754208754209e-06,
+      "loss": 0.3395,
+      "step": 2136
+    },
+    {
+      "epoch": 2.772225827384815,
+      "grad_norm": 0.194933200111678,
+      "learning_rate": 4.184704184704185e-06,
+      "loss": 0.3508,
+      "step": 2137
+    },
+    {
+      "epoch": 2.773523685918235,
+      "grad_norm": 0.21351784132569623,
+      "learning_rate": 4.1606541606541606e-06,
+      "loss": 0.3569,
+      "step": 2138
+    },
+    {
+      "epoch": 2.7748215444516546,
+      "grad_norm": 0.2022298125861802,
+      "learning_rate": 4.136604136604136e-06,
+      "loss": 0.3452,
+      "step": 2139
+    },
+    {
+      "epoch": 2.7761194029850746,
+      "grad_norm": 0.19008703004655647,
+      "learning_rate": 4.112554112554113e-06,
+      "loss": 0.3261,
+      "step": 2140
+    },
+    {
+      "epoch": 2.7774172615184947,
+      "grad_norm": 0.1996334755494925,
+      "learning_rate": 4.088504088504089e-06,
+      "loss": 0.3348,
+      "step": 2141
+    },
+    {
+      "epoch": 2.7787151200519142,
+      "grad_norm": 0.2016162252782046,
+      "learning_rate": 4.064454064454065e-06,
+      "loss": 0.339,
+      "step": 2142
+    },
+    {
+      "epoch": 2.7800129785853342,
+      "grad_norm": 0.20145152124370821,
+      "learning_rate": 4.040404040404041e-06,
+      "loss": 0.3384,
+      "step": 2143
+    },
+    {
+      "epoch": 2.7813108371187543,
+      "grad_norm": 0.20921251143803626,
+      "learning_rate": 4.016354016354017e-06,
+      "loss": 0.3449,
+      "step": 2144
+    },
+    {
+      "epoch": 2.782608695652174,
+      "grad_norm": 0.18838640289015987,
+      "learning_rate": 3.9923039923039925e-06,
+      "loss": 0.3249,
+      "step": 2145
+    },
+    {
+      "epoch": 2.783906554185594,
+      "grad_norm": 0.203141862957814,
+      "learning_rate": 3.968253968253968e-06,
+      "loss": 0.3474,
+      "step": 2146
+    },
+    {
+      "epoch": 2.785204412719014,
+      "grad_norm": 0.18959922111635913,
+      "learning_rate": 3.9442039442039446e-06,
+      "loss": 0.3358,
+      "step": 2147
+    },
+    {
+      "epoch": 2.7865022712524334,
+      "grad_norm": 0.2034563222680623,
+      "learning_rate": 3.92015392015392e-06,
+      "loss": 0.3399,
+      "step": 2148
+    },
+    {
+      "epoch": 2.7878001297858535,
+      "grad_norm": 0.197654010004629,
+      "learning_rate": 3.896103896103896e-06,
+      "loss": 0.3384,
+      "step": 2149
+    },
+    {
+      "epoch": 2.7890979883192735,
+      "grad_norm": 0.20680720078235312,
+      "learning_rate": 3.872053872053872e-06,
+      "loss": 0.3326,
+      "step": 2150
+    },
+    {
+      "epoch": 2.790395846852693,
+      "grad_norm": 0.1927827514450044,
+      "learning_rate": 3.848003848003849e-06,
+      "loss": 0.3276,
+      "step": 2151
+    },
+    {
+      "epoch": 2.791693705386113,
+      "grad_norm": 0.1899459712119537,
+      "learning_rate": 3.823953823953824e-06,
+      "loss": 0.3368,
+      "step": 2152
+    },
+    {
+      "epoch": 2.7929915639195326,
+      "grad_norm": 0.18276333170806777,
+      "learning_rate": 3.7999037999038e-06,
+      "loss": 0.3252,
+      "step": 2153
+    },
+    {
+      "epoch": 2.7942894224529526,
+      "grad_norm": 0.19256844716061763,
+      "learning_rate": 3.7758537758537756e-06,
+      "loss": 0.3483,
+      "step": 2154
+    },
+    {
+      "epoch": 2.795587280986372,
+      "grad_norm": 0.18473354170932832,
+      "learning_rate": 3.751803751803752e-06,
+      "loss": 0.3329,
+      "step": 2155
+    },
+    {
+      "epoch": 2.7968851395197922,
+      "grad_norm": 0.20272211241013205,
+      "learning_rate": 3.727753727753728e-06,
+      "loss": 0.363,
+      "step": 2156
+    },
+    {
+      "epoch": 2.7981829980532122,
+      "grad_norm": 0.210319586938317,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 0.3547,
+      "step": 2157
+    },
+    {
+      "epoch": 2.799480856586632,
+      "grad_norm": 0.18619320502174647,
+      "learning_rate": 3.67965367965368e-06,
+      "loss": 0.3282,
+      "step": 2158
+    },
+    {
+      "epoch": 2.800778715120052,
+      "grad_norm": 0.1772781571540208,
+      "learning_rate": 3.655603655603656e-06,
+      "loss": 0.33,
+      "step": 2159
+    },
+    {
+      "epoch": 2.802076573653472,
+      "grad_norm": 0.19401337158226317,
+      "learning_rate": 3.6315536315536315e-06,
+      "loss": 0.3418,
+      "step": 2160
+    },
+    {
+      "epoch": 2.8033744321868914,
+      "grad_norm": 0.20093342296511638,
+      "learning_rate": 3.6075036075036075e-06,
+      "loss": 0.3324,
+      "step": 2161
+    },
+    {
+      "epoch": 2.8046722907203114,
+      "grad_norm": 0.18887641527097687,
+      "learning_rate": 3.583453583453584e-06,
+      "loss": 0.3256,
+      "step": 2162
+    },
+    {
+      "epoch": 2.8059701492537314,
+      "grad_norm": 0.17961558157640115,
+      "learning_rate": 3.5594035594035596e-06,
+      "loss": 0.3328,
+      "step": 2163
+    },
+    {
+      "epoch": 2.807268007787151,
+      "grad_norm": 0.20347532828593798,
+      "learning_rate": 3.5353535353535352e-06,
+      "loss": 0.3421,
+      "step": 2164
+    },
+    {
+      "epoch": 2.808565866320571,
+      "grad_norm": 0.18010386527024905,
+      "learning_rate": 3.5113035113035113e-06,
+      "loss": 0.326,
+      "step": 2165
+    },
+    {
+      "epoch": 2.809863724853991,
+      "grad_norm": 0.18682472145471216,
+      "learning_rate": 3.4872534872534877e-06,
+      "loss": 0.3277,
+      "step": 2166
+    },
+    {
+      "epoch": 2.8111615833874106,
+      "grad_norm": 0.18668496331694528,
+      "learning_rate": 3.4632034632034634e-06,
+      "loss": 0.3441,
+      "step": 2167
+    },
+    {
+      "epoch": 2.8124594419208306,
+      "grad_norm": 0.18876447344150002,
+      "learning_rate": 3.439153439153439e-06,
+      "loss": 0.3543,
+      "step": 2168
+    },
+    {
+      "epoch": 2.8137573004542507,
+      "grad_norm": 0.1880026989268264,
+      "learning_rate": 3.4151034151034154e-06,
+      "loss": 0.3419,
+      "step": 2169
+    },
+    {
+      "epoch": 2.8150551589876702,
+      "grad_norm": 0.19326058199934312,
+      "learning_rate": 3.3910533910533915e-06,
+      "loss": 0.3332,
+      "step": 2170
+    },
+    {
+      "epoch": 2.8163530175210902,
+      "grad_norm": 0.18329023377490067,
+      "learning_rate": 3.367003367003367e-06,
+      "loss": 0.3632,
+      "step": 2171
+    },
+    {
+      "epoch": 2.8176508760545103,
+      "grad_norm": 0.19371890235019304,
+      "learning_rate": 3.3429533429533427e-06,
+      "loss": 0.3377,
+      "step": 2172
+    },
+    {
+      "epoch": 2.81894873458793,
+      "grad_norm": 0.18600523979644987,
+      "learning_rate": 3.318903318903319e-06,
+      "loss": 0.3469,
+      "step": 2173
+    },
+    {
+      "epoch": 2.82024659312135,
+      "grad_norm": 0.19389227471137455,
+      "learning_rate": 3.2948532948532953e-06,
+      "loss": 0.3387,
+      "step": 2174
+    },
+    {
+      "epoch": 2.82154445165477,
+      "grad_norm": 0.18954299093028096,
+      "learning_rate": 3.270803270803271e-06,
+      "loss": 0.3279,
+      "step": 2175
+    },
+    {
+      "epoch": 2.8228423101881894,
+      "grad_norm": 0.19152410986871543,
+      "learning_rate": 3.2467532467532465e-06,
+      "loss": 0.3374,
+      "step": 2176
+    },
+    {
+      "epoch": 2.8241401687216094,
+      "grad_norm": 0.196063104407719,
+      "learning_rate": 3.222703222703223e-06,
+      "loss": 0.3385,
+      "step": 2177
+    },
+    {
+      "epoch": 2.8254380272550295,
+      "grad_norm": 0.18850542953314792,
+      "learning_rate": 3.198653198653199e-06,
+      "loss": 0.349,
+      "step": 2178
+    },
+    {
+      "epoch": 2.826735885788449,
+      "grad_norm": 0.20124590955928826,
+      "learning_rate": 3.1746031746031746e-06,
+      "loss": 0.3415,
+      "step": 2179
+    },
+    {
+      "epoch": 2.828033744321869,
+      "grad_norm": 0.20248550914830157,
+      "learning_rate": 3.150553150553151e-06,
+      "loss": 0.3461,
+      "step": 2180
+    },
+    {
+      "epoch": 2.8293316028552886,
+      "grad_norm": 0.20035006058516966,
+      "learning_rate": 3.1265031265031267e-06,
+      "loss": 0.3432,
+      "step": 2181
+    },
+    {
+      "epoch": 2.8306294613887086,
+      "grad_norm": 0.1855009910859687,
+      "learning_rate": 3.1024531024531023e-06,
+      "loss": 0.3187,
+      "step": 2182
+    },
+    {
+      "epoch": 2.8319273199221286,
+      "grad_norm": 0.2143582348750643,
+      "learning_rate": 3.078403078403079e-06,
+      "loss": 0.3405,
+      "step": 2183
+    },
+    {
+      "epoch": 2.833225178455548,
+      "grad_norm": 0.19882736391050926,
+      "learning_rate": 3.0543530543530544e-06,
+      "loss": 0.3437,
+      "step": 2184
+    },
+    {
+      "epoch": 2.8345230369889682,
+      "grad_norm": 0.19373263047996803,
+      "learning_rate": 3.0303030303030305e-06,
+      "loss": 0.3391,
+      "step": 2185
+    },
+    {
+      "epoch": 2.835820895522388,
+      "grad_norm": 0.19378870740750723,
+      "learning_rate": 3.006253006253006e-06,
+      "loss": 0.3478,
+      "step": 2186
+    },
+    {
+      "epoch": 2.837118754055808,
+      "grad_norm": 0.24107367256249349,
+      "learning_rate": 2.9822029822029826e-06,
+      "loss": 0.3458,
+      "step": 2187
+    },
+    {
+      "epoch": 2.838416612589228,
+      "grad_norm": 0.18521053064833032,
+      "learning_rate": 2.958152958152958e-06,
+      "loss": 0.3383,
+      "step": 2188
+    },
+    {
+      "epoch": 2.8397144711226474,
+      "grad_norm": 0.1881021394765785,
+      "learning_rate": 2.9341029341029342e-06,
+      "loss": 0.3517,
+      "step": 2189
+    },
+    {
+      "epoch": 2.8410123296560674,
+      "grad_norm": 0.20171700206005888,
+      "learning_rate": 2.91005291005291e-06,
+      "loss": 0.3689,
+      "step": 2190
+    },
+    {
+      "epoch": 2.8423101881894874,
+      "grad_norm": 0.18467275960567497,
+      "learning_rate": 2.8860028860028863e-06,
+      "loss": 0.3271,
+      "step": 2191
+    },
+    {
+      "epoch": 2.843608046722907,
+      "grad_norm": 0.1935758058556294,
+      "learning_rate": 2.8619528619528624e-06,
+      "loss": 0.3457,
+      "step": 2192
+    },
+    {
+      "epoch": 2.844905905256327,
+      "grad_norm": 0.20313128506718564,
+      "learning_rate": 2.837902837902838e-06,
+      "loss": 0.3315,
+      "step": 2193
+    },
+    {
+      "epoch": 2.846203763789747,
+      "grad_norm": 0.2694894225052093,
+      "learning_rate": 2.813852813852814e-06,
+      "loss": 0.3651,
+      "step": 2194
+    },
+    {
+      "epoch": 2.8475016223231666,
+      "grad_norm": 0.191890934429557,
+      "learning_rate": 2.7898027898027897e-06,
+      "loss": 0.3457,
+      "step": 2195
+    },
+    {
+      "epoch": 2.8487994808565866,
+      "grad_norm": 0.18753053839949832,
+      "learning_rate": 2.765752765752766e-06,
+      "loss": 0.3438,
+      "step": 2196
+    },
+    {
+      "epoch": 2.8500973393900066,
+      "grad_norm": 0.18148036317355992,
+      "learning_rate": 2.7417027417027418e-06,
+      "loss": 0.3271,
+      "step": 2197
+    },
+    {
+      "epoch": 2.851395197923426,
+      "grad_norm": 0.20235355112701617,
+      "learning_rate": 2.717652717652718e-06,
+      "loss": 0.3468,
+      "step": 2198
+    },
+    {
+      "epoch": 2.8526930564568462,
+      "grad_norm": 0.19882040438487505,
+      "learning_rate": 2.6936026936026934e-06,
+      "loss": 0.3649,
+      "step": 2199
+    },
+    {
+      "epoch": 2.8539909149902662,
+      "grad_norm": 0.1919921579971501,
+      "learning_rate": 2.66955266955267e-06,
+      "loss": 0.3373,
+      "step": 2200
+    },
+    {
+      "epoch": 2.855288773523686,
+      "grad_norm": 0.19166967508036267,
+      "learning_rate": 2.6455026455026455e-06,
+      "loss": 0.3407,
+      "step": 2201
+    },
+    {
+      "epoch": 2.856586632057106,
+      "grad_norm": 0.18413982998209266,
+      "learning_rate": 2.6214526214526216e-06,
+      "loss": 0.3281,
+      "step": 2202
+    },
+    {
+      "epoch": 2.857884490590526,
+      "grad_norm": 0.1963183914870018,
+      "learning_rate": 2.5974025974025976e-06,
+      "loss": 0.3519,
+      "step": 2203
+    },
+    {
+      "epoch": 2.8591823491239454,
+      "grad_norm": 0.19124327719338702,
+      "learning_rate": 2.5733525733525737e-06,
+      "loss": 0.3546,
+      "step": 2204
+    },
+    {
+      "epoch": 2.8604802076573654,
+      "grad_norm": 0.1908455405574935,
+      "learning_rate": 2.5493025493025497e-06,
+      "loss": 0.3441,
+      "step": 2205
+    },
+    {
+      "epoch": 2.8617780661907855,
+      "grad_norm": 0.20074510401322135,
+      "learning_rate": 2.5252525252525253e-06,
+      "loss": 0.3623,
+      "step": 2206
+    },
+    {
+      "epoch": 2.863075924724205,
+      "grad_norm": 0.19215908226190542,
+      "learning_rate": 2.5012025012025014e-06,
+      "loss": 0.3323,
+      "step": 2207
+    },
+    {
+      "epoch": 2.864373783257625,
+      "grad_norm": 0.18434791337783116,
+      "learning_rate": 2.477152477152477e-06,
+      "loss": 0.3319,
+      "step": 2208
+    },
+    {
+      "epoch": 2.8656716417910446,
+      "grad_norm": 0.19538722948283108,
+      "learning_rate": 2.4531024531024535e-06,
+      "loss": 0.3275,
+      "step": 2209
+    },
+    {
+      "epoch": 2.8669695003244646,
+      "grad_norm": 0.1882905413491712,
+      "learning_rate": 2.429052429052429e-06,
+      "loss": 0.3301,
+      "step": 2210
+    },
+    {
+      "epoch": 2.8682673588578846,
+      "grad_norm": 0.18413079624889964,
+      "learning_rate": 2.405002405002405e-06,
+      "loss": 0.3316,
+      "step": 2211
+    },
+    {
+      "epoch": 2.869565217391304,
+      "grad_norm": 0.18114340544426985,
+      "learning_rate": 2.3809523809523808e-06,
+      "loss": 0.3293,
+      "step": 2212
+    },
+    {
+      "epoch": 2.8708630759247242,
+      "grad_norm": 0.18339952929530665,
+      "learning_rate": 2.3569023569023572e-06,
+      "loss": 0.3385,
+      "step": 2213
+    },
+    {
+      "epoch": 2.872160934458144,
+      "grad_norm": 0.19149046899206099,
+      "learning_rate": 2.332852332852333e-06,
+      "loss": 0.3362,
+      "step": 2214
+    },
+    {
+      "epoch": 2.873458792991564,
+      "grad_norm": 0.21144971266274312,
+      "learning_rate": 2.308802308802309e-06,
+      "loss": 0.3426,
+      "step": 2215
+    },
+    {
+      "epoch": 2.874756651524984,
+      "grad_norm": 0.18252025886810355,
+      "learning_rate": 2.284752284752285e-06,
+      "loss": 0.3461,
+      "step": 2216
+    },
+    {
+      "epoch": 2.8760545100584034,
+      "grad_norm": 0.18661014168815393,
+      "learning_rate": 2.2607022607022606e-06,
+      "loss": 0.3532,
+      "step": 2217
+    },
+    {
+      "epoch": 2.8773523685918234,
+      "grad_norm": 0.19921497008286657,
+      "learning_rate": 2.236652236652237e-06,
+      "loss": 0.3456,
+      "step": 2218
+    },
+    {
+      "epoch": 2.8786502271252434,
+      "grad_norm": 0.19199558097836697,
+      "learning_rate": 2.2126022126022127e-06,
+      "loss": 0.336,
+      "step": 2219
+    },
+    {
+      "epoch": 2.879948085658663,
+      "grad_norm": 0.18697397277980365,
+      "learning_rate": 2.1885521885521887e-06,
+      "loss": 0.3274,
+      "step": 2220
+    },
+    {
+      "epoch": 2.881245944192083,
+      "grad_norm": 0.18651465996633548,
+      "learning_rate": 2.1645021645021643e-06,
+      "loss": 0.3362,
+      "step": 2221
+    },
+    {
+      "epoch": 2.882543802725503,
+      "grad_norm": 0.18811253057615676,
+      "learning_rate": 2.140452140452141e-06,
+      "loss": 0.3441,
+      "step": 2222
+    },
+    {
+      "epoch": 2.8838416612589226,
+      "grad_norm": 0.1807114251046355,
+      "learning_rate": 2.1164021164021164e-06,
+      "loss": 0.3237,
+      "step": 2223
+    },
+    {
+      "epoch": 2.8851395197923426,
+      "grad_norm": 0.18668302514006135,
+      "learning_rate": 2.0923520923520925e-06,
+      "loss": 0.3556,
+      "step": 2224
+    },
+    {
+      "epoch": 2.8864373783257626,
+      "grad_norm": 0.1951670359448047,
+      "learning_rate": 2.068302068302068e-06,
+      "loss": 0.3181,
+      "step": 2225
+    },
+    {
+      "epoch": 2.887735236859182,
+      "grad_norm": 0.1874121175894903,
+      "learning_rate": 2.0442520442520446e-06,
+      "loss": 0.3447,
+      "step": 2226
+    },
+    {
+      "epoch": 2.8890330953926022,
+      "grad_norm": 0.18533011275342226,
+      "learning_rate": 2.0202020202020206e-06,
+      "loss": 0.3363,
+      "step": 2227
+    },
+    {
+      "epoch": 2.8903309539260222,
+      "grad_norm": 0.18793474012414535,
+      "learning_rate": 1.9961519961519962e-06,
+      "loss": 0.345,
+      "step": 2228
+    },
+    {
+      "epoch": 2.891628812459442,
+      "grad_norm": 0.18922319032385707,
+      "learning_rate": 1.9721019721019723e-06,
+      "loss": 0.3386,
+      "step": 2229
+    },
+    {
+      "epoch": 2.892926670992862,
+      "grad_norm": 0.19294630111421893,
+      "learning_rate": 1.948051948051948e-06,
+      "loss": 0.3513,
+      "step": 2230
+    },
+    {
+      "epoch": 2.894224529526282,
+      "grad_norm": 0.1831585914628477,
+      "learning_rate": 1.9240019240019244e-06,
+      "loss": 0.3185,
+      "step": 2231
+    },
+    {
+      "epoch": 2.8955223880597014,
+      "grad_norm": 0.18959540248425852,
+      "learning_rate": 1.8999518999519e-06,
+      "loss": 0.3448,
+      "step": 2232
+    },
+    {
+      "epoch": 2.8968202465931214,
+      "grad_norm": 0.18655567267427287,
+      "learning_rate": 1.875901875901876e-06,
+      "loss": 0.3262,
+      "step": 2233
+    },
+    {
+      "epoch": 2.8981181051265414,
+      "grad_norm": 0.18952603893507794,
+      "learning_rate": 1.8518518518518519e-06,
+      "loss": 0.3282,
+      "step": 2234
+    },
+    {
+      "epoch": 2.899415963659961,
+      "grad_norm": 0.1876885696047965,
+      "learning_rate": 1.827801827801828e-06,
+      "loss": 0.3435,
+      "step": 2235
+    },
+    {
+      "epoch": 2.900713822193381,
+      "grad_norm": 0.1865624408496654,
+      "learning_rate": 1.8037518037518038e-06,
+      "loss": 0.3331,
+      "step": 2236
+    },
+    {
+      "epoch": 2.902011680726801,
+      "grad_norm": 0.19122160432352084,
+      "learning_rate": 1.7797017797017798e-06,
+      "loss": 0.3371,
+      "step": 2237
+    },
+    {
+      "epoch": 2.9033095392602206,
+      "grad_norm": 0.19352943277773518,
+      "learning_rate": 1.7556517556517556e-06,
+      "loss": 0.3405,
+      "step": 2238
+    },
+    {
+      "epoch": 2.9046073977936406,
+      "grad_norm": 0.19319000412284978,
+      "learning_rate": 1.7316017316017317e-06,
+      "loss": 0.3399,
+      "step": 2239
+    },
+    {
+      "epoch": 2.90590525632706,
+      "grad_norm": 0.19327464804923486,
+      "learning_rate": 1.7075517075517077e-06,
+      "loss": 0.3319,
+      "step": 2240
+    },
+    {
+      "epoch": 2.90720311486048,
+      "grad_norm": 0.19976992892290674,
+      "learning_rate": 1.6835016835016836e-06,
+      "loss": 0.3432,
+      "step": 2241
+    },
+    {
+      "epoch": 2.9085009733939,
+      "grad_norm": 0.1892168913000648,
+      "learning_rate": 1.6594516594516596e-06,
+      "loss": 0.3463,
+      "step": 2242
+    },
+    {
+      "epoch": 2.90979883192732,
+      "grad_norm": 0.19443589296751324,
+      "learning_rate": 1.6354016354016354e-06,
+      "loss": 0.357,
+      "step": 2243
+    },
+    {
+      "epoch": 2.91109669046074,
+      "grad_norm": 0.18449321307823713,
+      "learning_rate": 1.6113516113516115e-06,
+      "loss": 0.345,
+      "step": 2244
+    },
+    {
+      "epoch": 2.9123945489941594,
+      "grad_norm": 0.1858254182171351,
+      "learning_rate": 1.5873015873015873e-06,
+      "loss": 0.3447,
+      "step": 2245
+    },
+    {
+      "epoch": 2.9136924075275794,
+      "grad_norm": 0.2004597208671706,
+      "learning_rate": 1.5632515632515634e-06,
+      "loss": 0.3366,
+      "step": 2246
+    },
+    {
+      "epoch": 2.9149902660609994,
+      "grad_norm": 0.18322062527491037,
+      "learning_rate": 1.5392015392015394e-06,
+      "loss": 0.3375,
+      "step": 2247
+    },
+    {
+      "epoch": 2.916288124594419,
+      "grad_norm": 0.1905983562403602,
+      "learning_rate": 1.5151515151515152e-06,
+      "loss": 0.3411,
+      "step": 2248
+    },
+    {
+      "epoch": 2.917585983127839,
+      "grad_norm": 0.1867573407190652,
+      "learning_rate": 1.4911014911014913e-06,
+      "loss": 0.336,
+      "step": 2249
+    },
+    {
+      "epoch": 2.918883841661259,
+      "grad_norm": 0.19058206255135468,
+      "learning_rate": 1.4670514670514671e-06,
+      "loss": 0.3474,
+      "step": 2250
+    },
+    {
+      "epoch": 2.9201817001946786,
+      "grad_norm": 0.18685316734046567,
+      "learning_rate": 1.4430014430014432e-06,
+      "loss": 0.3502,
+      "step": 2251
+    },
+    {
+      "epoch": 2.9214795587280986,
+      "grad_norm": 0.18640666171276632,
+      "learning_rate": 1.418951418951419e-06,
+      "loss": 0.3317,
+      "step": 2252
+    },
+    {
+      "epoch": 2.9227774172615186,
+      "grad_norm": 0.19155464358225485,
+      "learning_rate": 1.3949013949013948e-06,
+      "loss": 0.3352,
+      "step": 2253
+    },
+    {
+      "epoch": 2.924075275794938,
+      "grad_norm": 0.1868855983928718,
+      "learning_rate": 1.3708513708513709e-06,
+      "loss": 0.328,
+      "step": 2254
+    },
+    {
+      "epoch": 2.925373134328358,
+      "grad_norm": 0.18433362766279043,
+      "learning_rate": 1.3468013468013467e-06,
+      "loss": 0.3452,
+      "step": 2255
+    },
+    {
+      "epoch": 2.9266709928617782,
+      "grad_norm": 0.19259127006608057,
+      "learning_rate": 1.3227513227513228e-06,
+      "loss": 0.3404,
+      "step": 2256
+    },
+    {
+      "epoch": 2.927968851395198,
+      "grad_norm": 0.18080624245350022,
+      "learning_rate": 1.2987012987012988e-06,
+      "loss": 0.3266,
+      "step": 2257
+    },
+    {
+      "epoch": 2.929266709928618,
+      "grad_norm": 0.17871124072334996,
+      "learning_rate": 1.2746512746512749e-06,
+      "loss": 0.3395,
+      "step": 2258
+    },
+    {
+      "epoch": 2.930564568462038,
+      "grad_norm": 0.18709418548907147,
+      "learning_rate": 1.2506012506012507e-06,
+      "loss": 0.346,
+      "step": 2259
+    },
+    {
+      "epoch": 2.9318624269954574,
+      "grad_norm": 0.18683092960850883,
+      "learning_rate": 1.2265512265512267e-06,
+      "loss": 0.336,
+      "step": 2260
+    },
+    {
+      "epoch": 2.9331602855288774,
+      "grad_norm": 0.18777575130149565,
+      "learning_rate": 1.2025012025012026e-06,
+      "loss": 0.3368,
+      "step": 2261
+    },
+    {
+      "epoch": 2.9344581440622974,
+      "grad_norm": 0.18324636658598714,
+      "learning_rate": 1.1784511784511786e-06,
+      "loss": 0.3292,
+      "step": 2262
+    },
+    {
+      "epoch": 2.935756002595717,
+      "grad_norm": 0.1851227917603969,
+      "learning_rate": 1.1544011544011545e-06,
+      "loss": 0.3378,
+      "step": 2263
+    },
+    {
+      "epoch": 2.937053861129137,
+      "grad_norm": 0.19847788748302606,
+      "learning_rate": 1.1303511303511303e-06,
+      "loss": 0.3754,
+      "step": 2264
+    },
+    {
+      "epoch": 2.938351719662557,
+      "grad_norm": 0.177806134860686,
+      "learning_rate": 1.1063011063011063e-06,
+      "loss": 0.327,
+      "step": 2265
+    },
+    {
+      "epoch": 2.9396495781959766,
+      "grad_norm": 0.197095005642012,
+      "learning_rate": 1.0822510822510822e-06,
+      "loss": 0.344,
+      "step": 2266
+    },
+    {
+      "epoch": 2.9409474367293966,
+      "grad_norm": 0.18894739122645604,
+      "learning_rate": 1.0582010582010582e-06,
+      "loss": 0.3437,
+      "step": 2267
+    },
+    {
+      "epoch": 2.942245295262816,
+      "grad_norm": 0.1763401949490533,
+      "learning_rate": 1.034151034151034e-06,
+      "loss": 0.3285,
+      "step": 2268
+    },
+    {
+      "epoch": 2.943543153796236,
+      "grad_norm": 0.1786777654803748,
+      "learning_rate": 1.0101010101010103e-06,
+      "loss": 0.3264,
+      "step": 2269
+    },
+    {
+      "epoch": 2.9448410123296562,
+      "grad_norm": 0.18511698934832105,
+      "learning_rate": 9.860509860509861e-07,
+      "loss": 0.3403,
+      "step": 2270
+    },
+    {
+      "epoch": 2.946138870863076,
+      "grad_norm": 0.18872890425471747,
+      "learning_rate": 9.620009620009622e-07,
+      "loss": 0.3288,
+      "step": 2271
+    },
+    {
+      "epoch": 2.947436729396496,
+      "grad_norm": 0.18279278752067737,
+      "learning_rate": 9.37950937950938e-07,
+      "loss": 0.3422,
+      "step": 2272
+    },
+    {
+      "epoch": 2.9487345879299154,
+      "grad_norm": 0.18006141885171842,
+      "learning_rate": 9.13900913900914e-07,
+      "loss": 0.3446,
+      "step": 2273
+    },
+    {
+      "epoch": 2.9500324464633354,
+      "grad_norm": 0.19260565452121156,
+      "learning_rate": 8.898508898508899e-07,
+      "loss": 0.3398,
+      "step": 2274
+    },
+    {
+      "epoch": 2.9513303049967554,
+      "grad_norm": 0.21921401490874187,
+      "learning_rate": 8.658008658008658e-07,
+      "loss": 0.3568,
+      "step": 2275
+    },
+    {
+      "epoch": 2.952628163530175,
+      "grad_norm": 0.18842161857636638,
+      "learning_rate": 8.417508417508418e-07,
+      "loss": 0.3498,
+      "step": 2276
+    },
+    {
+      "epoch": 2.953926022063595,
+      "grad_norm": 0.17891619649785445,
+      "learning_rate": 8.177008177008177e-07,
+      "loss": 0.3272,
+      "step": 2277
+    },
+    {
+      "epoch": 2.955223880597015,
+      "grad_norm": 0.17755022636746284,
+      "learning_rate": 7.936507936507937e-07,
+      "loss": 0.3231,
+      "step": 2278
+    },
+    {
+      "epoch": 2.9565217391304346,
+      "grad_norm": 0.1863595372909174,
+      "learning_rate": 7.696007696007697e-07,
+      "loss": 0.3408,
+      "step": 2279
+    },
+    {
+      "epoch": 2.9578195976638546,
+      "grad_norm": 0.17943740925178142,
+      "learning_rate": 7.455507455507456e-07,
+      "loss": 0.3222,
+      "step": 2280
+    },
+    {
+      "epoch": 2.9591174561972746,
+      "grad_norm": 0.1875857112899972,
+      "learning_rate": 7.215007215007216e-07,
+      "loss": 0.33,
+      "step": 2281
+    },
+    {
+      "epoch": 2.960415314730694,
+      "grad_norm": 0.18672707365536773,
+      "learning_rate": 6.974506974506974e-07,
+      "loss": 0.3377,
+      "step": 2282
+    },
+    {
+      "epoch": 2.961713173264114,
+      "grad_norm": 0.18854762948864245,
+      "learning_rate": 6.734006734006734e-07,
+      "loss": 0.3511,
+      "step": 2283
+    },
+    {
+      "epoch": 2.9630110317975342,
+      "grad_norm": 0.1846633145163194,
+      "learning_rate": 6.493506493506494e-07,
+      "loss": 0.3301,
+      "step": 2284
+    },
+    {
+      "epoch": 2.964308890330954,
+      "grad_norm": 0.1782259793697707,
+      "learning_rate": 6.253006253006253e-07,
+      "loss": 0.3327,
+      "step": 2285
+    },
+    {
+      "epoch": 2.965606748864374,
+      "grad_norm": 0.19406554087810526,
+      "learning_rate": 6.012506012506013e-07,
+      "loss": 0.3374,
+      "step": 2286
+    },
+    {
+      "epoch": 2.966904607397794,
+      "grad_norm": 0.18974085077422986,
+      "learning_rate": 5.772005772005772e-07,
+      "loss": 0.3193,
+      "step": 2287
+    },
+    {
+      "epoch": 2.9682024659312134,
+      "grad_norm": 0.19068994562305627,
+      "learning_rate": 5.531505531505532e-07,
+      "loss": 0.3387,
+      "step": 2288
+    },
+    {
+      "epoch": 2.9695003244646334,
+      "grad_norm": 0.17821215477258306,
+      "learning_rate": 5.291005291005291e-07,
+      "loss": 0.3328,
+      "step": 2289
+    },
+    {
+      "epoch": 2.9707981829980534,
+      "grad_norm": 0.18413236462451124,
+      "learning_rate": 5.050505050505052e-07,
+      "loss": 0.3362,
+      "step": 2290
+    },
+    {
+      "epoch": 2.972096041531473,
+      "grad_norm": 0.18085396718066815,
+      "learning_rate": 4.810004810004811e-07,
+      "loss": 0.3302,
+      "step": 2291
+    },
+    {
+      "epoch": 2.973393900064893,
+      "grad_norm": 0.18231587065998014,
+      "learning_rate": 4.56950456950457e-07,
+      "loss": 0.3338,
+      "step": 2292
+    },
+    {
+      "epoch": 2.974691758598313,
+      "grad_norm": 0.18433591352078926,
+      "learning_rate": 4.329004329004329e-07,
+      "loss": 0.3373,
+      "step": 2293
+    },
+    {
+      "epoch": 2.9759896171317326,
+      "grad_norm": 0.17897899773682865,
+      "learning_rate": 4.0885040885040886e-07,
+      "loss": 0.3303,
+      "step": 2294
+    },
+    {
+      "epoch": 2.9772874756651526,
+      "grad_norm": 0.18401504997308174,
+      "learning_rate": 3.8480038480038485e-07,
+      "loss": 0.3295,
+      "step": 2295
+    },
+    {
+      "epoch": 2.9785853341985726,
+      "grad_norm": 0.18601416230069387,
+      "learning_rate": 3.607503607503608e-07,
+      "loss": 0.3384,
+      "step": 2296
+    },
+    {
+      "epoch": 2.979883192731992,
+      "grad_norm": 0.18294283749703283,
+      "learning_rate": 3.367003367003367e-07,
+      "loss": 0.3414,
+      "step": 2297
+    },
+    {
+      "epoch": 2.981181051265412,
+      "grad_norm": 0.18248918456066632,
+      "learning_rate": 3.1265031265031267e-07,
+      "loss": 0.3522,
+      "step": 2298
+    },
+    {
+      "epoch": 2.982478909798832,
+      "grad_norm": 0.18259396477447506,
+      "learning_rate": 2.886002886002886e-07,
+      "loss": 0.3194,
+      "step": 2299
+    },
+    {
+      "epoch": 2.983776768332252,
+      "grad_norm": 0.1885739190189894,
+      "learning_rate": 2.6455026455026455e-07,
+      "loss": 0.3407,
+      "step": 2300
+    },
+    {
+      "epoch": 2.9850746268656714,
+      "grad_norm": 0.18564276265962515,
+      "learning_rate": 2.4050024050024055e-07,
+      "loss": 0.3431,
+      "step": 2301
+    },
+    {
+      "epoch": 2.9863724853990914,
+      "grad_norm": 0.17794361853046303,
+      "learning_rate": 2.1645021645021646e-07,
+      "loss": 0.3228,
+      "step": 2302
+    },
+    {
+      "epoch": 2.9876703439325114,
+      "grad_norm": 0.18130600626680995,
+      "learning_rate": 1.9240019240019243e-07,
+      "loss": 0.3336,
+      "step": 2303
+    },
+    {
+      "epoch": 2.988968202465931,
+      "grad_norm": 0.18435788490293198,
+      "learning_rate": 1.6835016835016834e-07,
+      "loss": 0.3308,
+      "step": 2304
+    },
+    {
+      "epoch": 2.990266060999351,
+      "grad_norm": 0.18127043211913135,
+      "learning_rate": 1.443001443001443e-07,
+      "loss": 0.3449,
+      "step": 2305
+    },
+    {
+      "epoch": 2.991563919532771,
+      "grad_norm": 0.1799885551759602,
+      "learning_rate": 1.2025012025012027e-07,
+      "loss": 0.3267,
+      "step": 2306
+    },
+    {
+      "epoch": 2.9928617780661906,
+      "grad_norm": 0.18020390891152432,
+      "learning_rate": 9.620009620009621e-08,
+      "loss": 0.3376,
+      "step": 2307
+    },
+    {
+      "epoch": 2.9941596365996106,
+      "grad_norm": 0.18299772015112434,
+      "learning_rate": 7.215007215007215e-08,
+      "loss": 0.3404,
+      "step": 2308
+    },
+    {
+      "epoch": 2.9954574951330306,
+      "grad_norm": 0.17991344814248053,
+      "learning_rate": 4.8100048100048107e-08,
+      "loss": 0.3293,
+      "step": 2309
+    },
+    {
+      "epoch": 2.99675535366645,
+      "grad_norm": 0.18545369820383786,
+      "learning_rate": 2.4050024050024053e-08,
+      "loss": 0.3376,
+      "step": 2310
+    },
+    {
+      "epoch": 2.99675535366645,
+      "step": 2310,
+      "total_flos": 2.5679880641918796e+19,
+      "train_loss": 0.5017085661361744,
+      "train_runtime": 66070.1568,
+      "train_samples_per_second": 0.56,
+      "train_steps_per_second": 0.035
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2310,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.5679880641918796e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}