diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,63159 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.7210058322975428,
+  "eval_steps": 500,
+  "global_step": 9000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00019122287025528253,
+      "grad_norm": 17.15780371351428,
+      "learning_rate": 0.0,
+      "loss": 0.8136,
+      "step": 1
+    },
+    {
+      "epoch": 0.00038244574051056506,
+      "grad_norm": 10.661322049436311,
+      "learning_rate": 8.80855856868319e-08,
+      "loss": 1.5096,
+      "step": 2
+    },
+    {
+      "epoch": 0.0005736686107658476,
+      "grad_norm": 5.3643277696887885,
+      "learning_rate": 1.396123501676888e-07,
+      "loss": 0.4811,
+      "step": 3
+    },
+    {
+      "epoch": 0.0007648914810211301,
+      "grad_norm": 9.87808565628534,
+      "learning_rate": 1.761711713736638e-07,
+      "loss": 0.9424,
+      "step": 4
+    },
+    {
+      "epoch": 0.0009561143512764127,
+      "grad_norm": 5.225935871324979,
+      "learning_rate": 2.0452839616086312e-07,
+      "loss": 0.8183,
+      "step": 5
+    },
+    {
+      "epoch": 0.0011473372215316951,
+      "grad_norm": 9.04058039127108,
+      "learning_rate": 2.2769793585452067e-07,
+      "loss": 0.6391,
+      "step": 6
+    },
+    {
+      "epoch": 0.0013385600917869777,
+      "grad_norm": 14.522107864262992,
+      "learning_rate": 2.4728750254025437e-07,
+      "loss": 0.8569,
+      "step": 7
+    },
+    {
+      "epoch": 0.0015297829620422602,
+      "grad_norm": 17.46333963482277,
+      "learning_rate": 2.6425675706049573e-07,
+      "loss": 1.0278,
+      "step": 8
+    },
+    {
+      "epoch": 0.0017210058322975428,
+      "grad_norm": 11.55143117253395,
+      "learning_rate": 2.792247003353776e-07,
+      "loss": 0.6788,
+      "step": 9
+    },
+    {
+      "epoch": 0.0019122287025528254,
+      "grad_norm": 9.8083215065763,
+      "learning_rate": 2.92613981847695e-07,
+      "loss": 0.4014,
+      "step": 10
+    },
+    {
+      "epoch": 0.002103451572808108,
+      "grad_norm": 13.132143238449551,
+      "learning_rate": 3.0472606027121127e-07,
+      "loss": 0.553,
+      "step": 11
+    },
+    {
+      "epoch": 0.0022946744430633902,
+      "grad_norm": 13.03315673344246,
+      "learning_rate": 3.157835215413526e-07,
+      "loss": 0.5281,
+      "step": 12
+    },
+    {
+      "epoch": 0.002485897313318673,
+      "grad_norm": 13.821671008873675,
+      "learning_rate": 3.259553998712733e-07,
+      "loss": 0.7281,
+      "step": 13
+    },
+    {
+      "epoch": 0.0026771201835739554,
+      "grad_norm": 12.271257290643208,
+      "learning_rate": 3.3537308822708625e-07,
+      "loss": 0.9487,
+      "step": 14
+    },
+    {
+      "epoch": 0.002868343053829238,
+      "grad_norm": 10.749341013027518,
+      "learning_rate": 3.4414074632855187e-07,
+      "loss": 0.8007,
+      "step": 15
+    },
+    {
+      "epoch": 0.0030595659240845205,
+      "grad_norm": 6.62160188868718,
+      "learning_rate": 3.523423427473276e-07,
+      "loss": 0.5471,
+      "step": 16
+    },
+    {
+      "epoch": 0.0032507887943398033,
+      "grad_norm": 6.226866672797083,
+      "learning_rate": 3.6004655834469816e-07,
+      "loss": 0.5128,
+      "step": 17
+    },
+    {
+      "epoch": 0.0034420116645950856,
+      "grad_norm": 6.150707867830387,
+      "learning_rate": 3.673102860222094e-07,
+      "loss": 1.0285,
+      "step": 18
+    },
+    {
+      "epoch": 0.003633234534850368,
+      "grad_norm": 11.544970612156051,
+      "learning_rate": 3.741811829768857e-07,
+      "loss": 0.7353,
+      "step": 19
+    },
+    {
+      "epoch": 0.0038244574051056507,
+      "grad_norm": 8.347400338262599,
+      "learning_rate": 3.806995675345269e-07,
+      "loss": 0.669,
+      "step": 20
+    },
+    {
+      "epoch": 0.0040156802753609335,
+      "grad_norm": 13.91830586965418,
+      "learning_rate": 3.868998527079432e-07,
+      "loss": 0.8399,
+      "step": 21
+    },
+    {
+      "epoch": 0.004206903145616216,
+      "grad_norm": 15.449430604584085,
+      "learning_rate": 3.9281164595804315e-07,
+      "loss": 0.6436,
+      "step": 22
+    },
+    {
+      "epoch": 0.004398126015871498,
+      "grad_norm": 10.550571907054042,
+      "learning_rate": 3.9846060428995293e-07,
+      "loss": 0.7047,
+      "step": 23
+    },
+    {
+      "epoch": 0.0045893488861267805,
+      "grad_norm": 16.988800847854062,
+      "learning_rate": 4.038691072281845e-07,
+      "loss": 0.858,
+      "step": 24
+    },
+    {
+      "epoch": 0.004780571756382064,
+      "grad_norm": 14.741495040374657,
+      "learning_rate": 4.0905679232172624e-07,
+      "loss": 0.9656,
+      "step": 25
+    },
+    {
+      "epoch": 0.004971794626637346,
+      "grad_norm": 12.110148664801509,
+      "learning_rate": 4.140409855581052e-07,
+      "loss": 0.9949,
+      "step": 26
+    },
+    {
+      "epoch": 0.005163017496892628,
+      "grad_norm": 7.694057985861227,
+      "learning_rate": 4.188370505030663e-07,
+      "loss": 0.7379,
+      "step": 27
+    },
+    {
+      "epoch": 0.005354240367147911,
+      "grad_norm": 4.7312152251524,
+      "learning_rate": 4.2345867391391823e-07,
+      "loss": 0.456,
+      "step": 28
+    },
+    {
+      "epoch": 0.005545463237403193,
+      "grad_norm": 11.459542801343355,
+      "learning_rate": 4.279181012113107e-07,
+      "loss": 1.0985,
+      "step": 29
+    },
+    {
+      "epoch": 0.005736686107658476,
+      "grad_norm": 10.442674279415069,
+      "learning_rate": 4.3222633201538385e-07,
+      "loss": 0.7029,
+      "step": 30
+    },
+    {
+      "epoch": 0.005927908977913759,
+      "grad_norm": 6.920215416350231,
+      "learning_rate": 4.3639328360796955e-07,
+      "loss": 0.4725,
+      "step": 31
+    },
+    {
+      "epoch": 0.006119131848169041,
+      "grad_norm": 8.47817524980501,
+      "learning_rate": 4.4042792843415953e-07,
+      "loss": 0.4567,
+      "step": 32
+    },
+    {
+      "epoch": 0.006310354718424323,
+      "grad_norm": 5.817956803587339,
+      "learning_rate": 4.4433841043890004e-07,
+      "loss": 0.8878,
+      "step": 33
+    },
+    {
+      "epoch": 0.0065015775886796065,
+      "grad_norm": 7.852224202527448,
+      "learning_rate": 4.481321440315301e-07,
+      "loss": 0.2936,
+      "step": 34
+    },
+    {
+      "epoch": 0.006692800458934889,
+      "grad_norm": 11.856995757301757,
+      "learning_rate": 4.5181589870111747e-07,
+      "loss": 0.5478,
+      "step": 35
+    },
+    {
+      "epoch": 0.006884023329190171,
+      "grad_norm": 11.203925493869011,
+      "learning_rate": 4.5539587170904135e-07,
+      "loss": 0.4584,
+      "step": 36
+    },
+    {
+      "epoch": 0.0070752461994454535,
+      "grad_norm": 20.06007517959336,
+      "learning_rate": 4.5887775081966365e-07,
+      "loss": 0.5565,
+      "step": 37
+    },
+    {
+      "epoch": 0.007266469069700736,
+      "grad_norm": 9.932639414809586,
+      "learning_rate": 4.622667686637176e-07,
+      "loss": 0.7379,
+      "step": 38
+    },
+    {
+      "epoch": 0.007457691939956019,
+      "grad_norm": 10.881380861953625,
+      "learning_rate": 4.6556775003896203e-07,
+      "loss": 1.1213,
+      "step": 39
+    },
+    {
+      "epoch": 0.007648914810211301,
+      "grad_norm": 9.533536242835448,
+      "learning_rate": 4.687851532213588e-07,
+      "loss": 1.1104,
+      "step": 40
+    },
+    {
+      "epoch": 0.007840137680466585,
+      "grad_norm": 5.567654225509079,
+      "learning_rate": 4.7192310617444426e-07,
+      "loss": 0.6172,
+      "step": 41
+    },
+    {
+      "epoch": 0.008031360550721867,
+      "grad_norm": 7.5000282287066415,
+      "learning_rate": 4.7498543839477507e-07,
+      "loss": 0.8246,
+      "step": 42
+    },
+    {
+      "epoch": 0.00822258342097715,
+      "grad_norm": 5.7544385944762375,
+      "learning_rate": 4.779757090097475e-07,
+      "loss": 0.5471,
+      "step": 43
+    },
+    {
+      "epoch": 0.008413806291232432,
+      "grad_norm": 8.455779443295121,
+      "learning_rate": 4.80897231644875e-07,
+      "loss": 0.4524,
+      "step": 44
+    },
+    {
+      "epoch": 0.008605029161487714,
+      "grad_norm": 5.316679981816917,
+      "learning_rate": 4.837530964962406e-07,
+      "loss": 0.8138,
+      "step": 45
+    },
+    {
+      "epoch": 0.008796252031742996,
+      "grad_norm": 5.998670112728145,
+      "learning_rate": 4.865461899767849e-07,
+      "loss": 0.6946,
+      "step": 46
+    },
+    {
+      "epoch": 0.008987474901998279,
+      "grad_norm": 6.821890981518299,
+      "learning_rate": 4.892792122495718e-07,
+      "loss": 0.4556,
+      "step": 47
+    },
+    {
+      "epoch": 0.009178697772253561,
+      "grad_norm": 4.798524454931962,
+      "learning_rate": 4.919546929150164e-07,
+      "loss": 0.5977,
+      "step": 48
+    },
+    {
+      "epoch": 0.009369920642508843,
+      "grad_norm": 7.1260835593284515,
+      "learning_rate": 4.945750050805087e-07,
+      "loss": 0.4846,
+      "step": 49
+    },
+    {
+      "epoch": 0.009561143512764127,
+      "grad_norm": 7.741588088157945,
+      "learning_rate": 4.971423780085582e-07,
+      "loss": 0.3271,
+      "step": 50
+    },
+    {
+      "epoch": 0.00975236638301941,
+      "grad_norm": 13.496307998480315,
+      "learning_rate": 4.996589085123869e-07,
+      "loss": 0.8473,
+      "step": 51
+    },
+    {
+      "epoch": 0.009943589253274692,
+      "grad_norm": 5.956360425856135,
+      "learning_rate": 5.021265712449371e-07,
+      "loss": 0.5307,
+      "step": 52
+    },
+    {
+      "epoch": 0.010134812123529974,
+      "grad_norm": 3.9755450372880774,
+      "learning_rate": 5.045472280077838e-07,
+      "loss": 0.555,
+      "step": 53
+    },
+    {
+      "epoch": 0.010326034993785257,
+      "grad_norm": 3.907655020754835,
+      "learning_rate": 5.069226361898982e-07,
+      "loss": 0.2512,
+      "step": 54
+    },
+    {
+      "epoch": 0.010517257864040539,
+      "grad_norm": 3.9692048540560254,
+      "learning_rate": 5.092544564320745e-07,
+      "loss": 0.6572,
+      "step": 55
+    },
+    {
+      "epoch": 0.010708480734295821,
+      "grad_norm": 4.818195292288783,
+      "learning_rate": 5.115442596007502e-07,
+      "loss": 0.5755,
+      "step": 56
+    },
+    {
+      "epoch": 0.010899703604551104,
+      "grad_norm": 5.5092934206993665,
+      "learning_rate": 5.137935331445745e-07,
+      "loss": 0.5655,
+      "step": 57
+    },
+    {
+      "epoch": 0.011090926474806386,
+      "grad_norm": 5.419112587383769,
+      "learning_rate": 5.160036868981425e-07,
+      "loss": 0.6466,
+      "step": 58
+    },
+    {
+      "epoch": 0.01128214934506167,
+      "grad_norm": 4.188125335916501,
+      "learning_rate": 5.181760583896086e-07,
+      "loss": 0.5067,
+      "step": 59
+    },
+    {
+      "epoch": 0.011473372215316953,
+      "grad_norm": 7.111750547481919,
+      "learning_rate": 5.203119177022157e-07,
+      "loss": 0.4598,
+      "step": 60
+    },
+    {
+      "epoch": 0.011664595085572235,
+      "grad_norm": 4.459002126757809,
+      "learning_rate": 5.224124719339889e-07,
+      "loss": 0.2168,
+      "step": 61
+    },
+    {
+      "epoch": 0.011855817955827517,
+      "grad_norm": 5.214808454644133,
+      "learning_rate": 5.244788692948015e-07,
+      "loss": 0.3364,
+      "step": 62
+    },
+    {
+      "epoch": 0.0120470408260828,
+      "grad_norm": 7.2183203631278445,
+      "learning_rate": 5.265122028756319e-07,
+      "loss": 0.3654,
+      "step": 63
+    },
+    {
+      "epoch": 0.012238263696338082,
+      "grad_norm": 7.602486063258585,
+      "learning_rate": 5.285135141209915e-07,
+      "loss": 1.2169,
+      "step": 64
+    },
+    {
+      "epoch": 0.012429486566593364,
+      "grad_norm": 7.335481704673498,
+      "learning_rate": 5.304837960321364e-07,
+      "loss": 0.7286,
+      "step": 65
+    },
+    {
+      "epoch": 0.012620709436848647,
+      "grad_norm": 7.795860698522115,
+      "learning_rate": 5.324239961257319e-07,
+      "loss": 0.4335,
+      "step": 66
+    },
+    {
+      "epoch": 0.012811932307103929,
+      "grad_norm": 4.727430834493708,
+      "learning_rate": 5.343350191700329e-07,
+      "loss": 0.2253,
+      "step": 67
+    },
+    {
+      "epoch": 0.013003155177359213,
+      "grad_norm": 7.86403000929298,
+      "learning_rate": 5.36217729718362e-07,
+      "loss": 0.892,
+      "step": 68
+    },
+    {
+      "epoch": 0.013194378047614495,
+      "grad_norm": 4.696419126883423,
+      "learning_rate": 5.380729544576418e-07,
+      "loss": 0.5052,
+      "step": 69
+    },
+    {
+      "epoch": 0.013385600917869778,
+      "grad_norm": 5.378138845493509,
+      "learning_rate": 5.399014843879494e-07,
+      "loss": 0.4812,
+      "step": 70
+    },
+    {
+      "epoch": 0.01357682378812506,
+      "grad_norm": 3.4386363145401573,
+      "learning_rate": 5.417040768474773e-07,
+      "loss": 0.4428,
+      "step": 71
+    },
+    {
+      "epoch": 0.013768046658380342,
+      "grad_norm": 6.232078798630819,
+      "learning_rate": 5.434814573958733e-07,
+      "loss": 0.226,
+      "step": 72
+    },
+    {
+      "epoch": 0.013959269528635625,
+      "grad_norm": 4.525132462896826,
+      "learning_rate": 5.452343215676822e-07,
+      "loss": 0.5107,
+      "step": 73
+    },
+    {
+      "epoch": 0.014150492398890907,
+      "grad_norm": 7.723447289326565,
+      "learning_rate": 5.469633365064957e-07,
+      "loss": 0.3176,
+      "step": 74
+    },
+    {
+      "epoch": 0.01434171526914619,
+      "grad_norm": 8.169643352666359,
+      "learning_rate": 5.48669142489415e-07,
+      "loss": 0.2915,
+      "step": 75
+    },
+    {
+      "epoch": 0.014532938139401472,
+      "grad_norm": 5.381727931545296,
+      "learning_rate": 5.503523543505495e-07,
+      "loss": 0.5128,
+      "step": 76
+    },
+    {
+      "epoch": 0.014724161009656756,
+      "grad_norm": 6.676343506895845,
+      "learning_rate": 5.520135628114658e-07,
+      "loss": 0.9979,
+      "step": 77
+    },
+    {
+      "epoch": 0.014915383879912038,
+      "grad_norm": 5.527595304015183,
+      "learning_rate": 5.53653335725794e-07,
+      "loss": 0.4359,
+      "step": 78
+    },
+    {
+      "epoch": 0.01510660675016732,
+      "grad_norm": 5.945889939927779,
+      "learning_rate": 5.552722192445556e-07,
+      "loss": 0.5422,
+      "step": 79
+    },
+    {
+      "epoch": 0.015297829620422603,
+      "grad_norm": 2.907335960372418,
+      "learning_rate": 5.568707389081907e-07,
+      "loss": 0.3555,
+      "step": 80
+    },
+    {
+      "epoch": 0.015489052490677885,
+      "grad_norm": 5.813996809332267,
+      "learning_rate": 5.584494006707552e-07,
+      "loss": 0.5434,
+      "step": 81
+    },
+    {
+      "epoch": 0.01568027536093317,
+      "grad_norm": 5.064981676465826,
+      "learning_rate": 5.600086918612762e-07,
+      "loss": 0.5502,
+      "step": 82
+    },
+    {
+      "epoch": 0.01587149823118845,
+      "grad_norm": 3.4994830703467557,
+      "learning_rate": 5.615490820868417e-07,
+      "loss": 0.1646,
+      "step": 83
+    },
+    {
+      "epoch": 0.016062721101443734,
+      "grad_norm": 3.3621432941224265,
+      "learning_rate": 5.63071024081607e-07,
+      "loss": 0.1819,
+      "step": 84
+    },
+    {
+      "epoch": 0.016253943971699016,
+      "grad_norm": 4.4670979040180026,
+      "learning_rate": 5.645749545055614e-07,
+      "loss": 0.3665,
+      "step": 85
+    },
+    {
+      "epoch": 0.0164451668419543,
+      "grad_norm": 6.135858369826533,
+      "learning_rate": 5.660612946965794e-07,
+      "loss": 0.4984,
+      "step": 86
+    },
+    {
+      "epoch": 0.01663638971220958,
+      "grad_norm": 6.113486543078536,
+      "learning_rate": 5.675304513789995e-07,
+      "loss": 0.2955,
+      "step": 87
+    },
+    {
+      "epoch": 0.016827612582464863,
+      "grad_norm": 8.028468500126255,
+      "learning_rate": 5.689828173317071e-07,
+      "loss": 0.6361,
+      "step": 88
+    },
+    {
+      "epoch": 0.017018835452720146,
+      "grad_norm": 5.73067510918489,
+      "learning_rate": 5.704187720184725e-07,
+      "loss": 0.6421,
+      "step": 89
+    },
+    {
+      "epoch": 0.017210058322975428,
+      "grad_norm": 5.530448208902912,
+      "learning_rate": 5.718386821830726e-07,
+      "loss": 0.8147,
+      "step": 90
+    },
+    {
+      "epoch": 0.01740128119323071,
+      "grad_norm": 2.971386832688174,
+      "learning_rate": 5.732429024115276e-07,
+      "loss": 0.1586,
+      "step": 91
+    },
+    {
+      "epoch": 0.017592504063485993,
+      "grad_norm": 4.7086398719768034,
+      "learning_rate": 5.746317756636168e-07,
+      "loss": 0.7898,
+      "step": 92
+    },
+    {
+      "epoch": 0.017783726933741275,
+      "grad_norm": 4.768253189545859,
+      "learning_rate": 5.760056337756583e-07,
+      "loss": 0.773,
+      "step": 93
+    },
+    {
+      "epoch": 0.017974949803996557,
+      "grad_norm": 3.4881188503280205,
+      "learning_rate": 5.773647979364037e-07,
+      "loss": 0.3182,
+      "step": 94
+    },
+    {
+      "epoch": 0.01816617267425184,
+      "grad_norm": 5.07209584242463,
+      "learning_rate": 5.787095791377488e-07,
+      "loss": 0.4557,
+      "step": 95
+    },
+    {
+      "epoch": 0.018357395544507122,
+      "grad_norm": 6.442548531411296,
+      "learning_rate": 5.800402786018483e-07,
+      "loss": 0.3598,
+      "step": 96
+    },
+    {
+      "epoch": 0.018548618414762404,
+      "grad_norm": 2.9032253810581454,
+      "learning_rate": 5.813571881860966e-07,
+      "loss": 0.1164,
+      "step": 97
+    },
+    {
+      "epoch": 0.018739841285017687,
+      "grad_norm": 4.909342186209263,
+      "learning_rate": 5.826605907673407e-07,
+      "loss": 0.4375,
+      "step": 98
+    },
+    {
+      "epoch": 0.01893106415527297,
+      "grad_norm": 6.12772573428026,
+      "learning_rate": 5.839507606065888e-07,
+      "loss": 0.2106,
+      "step": 99
+    },
+    {
+      "epoch": 0.019122287025528255,
+      "grad_norm": 11.516655843447188,
+      "learning_rate": 5.8522796369539e-07,
+      "loss": 0.8053,
+      "step": 100
+    },
+    {
+      "epoch": 0.019313509895783537,
+      "grad_norm": 10.005602221039174,
+      "learning_rate": 5.864924580849814e-07,
+      "loss": 0.7651,
+      "step": 101
+    },
+    {
+      "epoch": 0.01950473276603882,
+      "grad_norm": 5.261664737914131,
+      "learning_rate": 5.877444941992188e-07,
+      "loss": 0.5622,
+      "step": 102
+    },
+    {
+      "epoch": 0.019695955636294102,
+      "grad_norm": 13.299189400888428,
+      "learning_rate": 5.889843151322441e-07,
+      "loss": 0.7012,
+      "step": 103
+    },
+    {
+      "epoch": 0.019887178506549384,
+      "grad_norm": 2.9295965155142745,
+      "learning_rate": 5.90212156931769e-07,
+      "loss": 0.561,
+      "step": 104
+    },
+    {
+      "epoch": 0.020078401376804667,
+      "grad_norm": 4.335384079265192,
+      "learning_rate": 5.914282488688063e-07,
+      "loss": 0.3964,
+      "step": 105
+    },
+    {
+      "epoch": 0.02026962424705995,
+      "grad_norm": 4.32237726852537,
+      "learning_rate": 5.926328136946157e-07,
+      "loss": 0.5624,
+      "step": 106
+    },
+    {
+      "epoch": 0.02046084711731523,
+      "grad_norm": 4.104008292355256,
+      "learning_rate": 5.938260678855866e-07,
+      "loss": 0.334,
+      "step": 107
+    },
+    {
+      "epoch": 0.020652069987570514,
+      "grad_norm": 4.562026116831141,
+      "learning_rate": 5.950082218767301e-07,
+      "loss": 0.4256,
+      "step": 108
+    },
+    {
+      "epoch": 0.020843292857825796,
+      "grad_norm": 4.353701382589275,
+      "learning_rate": 5.961794802844105e-07,
+      "loss": 0.4306,
+      "step": 109
+    },
+    {
+      "epoch": 0.021034515728081078,
+      "grad_norm": 5.343087316994396,
+      "learning_rate": 5.973400421189063e-07,
+      "loss": 0.4865,
+      "step": 110
+    },
+    {
+      "epoch": 0.02122573859833636,
+      "grad_norm": 5.222869891502607,
+      "learning_rate": 5.984901009873524e-07,
+      "loss": 0.22,
+      "step": 111
+    },
+    {
+      "epoch": 0.021416961468591643,
+      "grad_norm": 5.3957670151257044,
+      "learning_rate": 5.99629845287582e-07,
+      "loss": 0.3001,
+      "step": 112
+    },
+    {
+      "epoch": 0.021608184338846925,
+      "grad_norm": 5.622995824870297,
+      "learning_rate": 6.007594583933514e-07,
+      "loss": 0.3788,
+      "step": 113
+    },
+    {
+      "epoch": 0.021799407209102208,
+      "grad_norm": 9.064800102104474,
+      "learning_rate": 6.018791188314064e-07,
+      "loss": 0.7641,
+      "step": 114
+    },
+    {
+      "epoch": 0.02199063007935749,
+      "grad_norm": 4.936238574403362,
+      "learning_rate": 6.029890004508161e-07,
+      "loss": 0.8081,
+      "step": 115
+    },
+    {
+      "epoch": 0.022181852949612772,
+      "grad_norm": 3.9558903969141337,
+      "learning_rate": 6.040892725849744e-07,
+      "loss": 0.8974,
+      "step": 116
+    },
+    {
+      "epoch": 0.022373075819868055,
+      "grad_norm": 3.3649091690750716,
+      "learning_rate": 6.051801002066509e-07,
+      "loss": 0.4396,
+      "step": 117
+    },
+    {
+      "epoch": 0.02256429869012334,
+      "grad_norm": 4.49195948809922,
+      "learning_rate": 6.062616440764405e-07,
+      "loss": 0.368,
+      "step": 118
+    },
+    {
+      "epoch": 0.022755521560378623,
+      "grad_norm": 4.643589703962347,
+      "learning_rate": 6.073340608849526e-07,
+      "loss": 0.2087,
+      "step": 119
+    },
+    {
+      "epoch": 0.022946744430633905,
+      "grad_norm": 3.499456908460912,
+      "learning_rate": 6.083975033890476e-07,
+      "loss": 0.4289,
+      "step": 120
+    },
+    {
+      "epoch": 0.023137967300889187,
+      "grad_norm": 4.095948306100202,
+      "learning_rate": 6.094521205424225e-07,
+      "loss": 0.5533,
+      "step": 121
+    },
+    {
+      "epoch": 0.02332919017114447,
+      "grad_norm": 6.479183609951553,
+      "learning_rate": 6.104980576208209e-07,
+      "loss": 0.2811,
+      "step": 122
+    },
+    {
+      "epoch": 0.023520413041399752,
+      "grad_norm": 3.787872064526689,
+      "learning_rate": 6.11535456342133e-07,
+      "loss": 0.257,
+      "step": 123
+    },
+    {
+      "epoch": 0.023711635911655034,
+      "grad_norm": 4.533567901227362,
+      "learning_rate": 6.125644549816335e-07,
+      "loss": 0.1675,
+      "step": 124
+    },
+    {
+      "epoch": 0.023902858781910317,
+      "grad_norm": 3.502907635141068,
+      "learning_rate": 6.135851884825895e-07,
+      "loss": 0.1512,
+      "step": 125
+    },
+    {
+      "epoch": 0.0240940816521656,
+      "grad_norm": 5.480565674537168,
+      "learning_rate": 6.145977885624638e-07,
+      "loss": 0.5728,
+      "step": 126
+    },
+    {
+      "epoch": 0.02428530452242088,
+      "grad_norm": 7.214219822778287,
+      "learning_rate": 6.156023838149195e-07,
+      "loss": 0.9082,
+      "step": 127
+    },
+    {
+      "epoch": 0.024476527392676164,
+      "grad_norm": 2.968448181872049,
+      "learning_rate": 6.165990998078233e-07,
+      "loss": 0.3586,
+      "step": 128
+    },
+    {
+      "epoch": 0.024667750262931446,
+      "grad_norm": 5.150490243040204,
+      "learning_rate": 6.175880591774363e-07,
+      "loss": 0.3006,
+      "step": 129
+    },
+    {
+      "epoch": 0.02485897313318673,
+      "grad_norm": 2.8727784903430402,
+      "learning_rate": 6.185693817189682e-07,
+      "loss": 0.4156,
+      "step": 130
+    },
+    {
+      "epoch": 0.02505019600344201,
+      "grad_norm": 3.3548120071339986,
+      "learning_rate": 6.195431844736615e-07,
+      "loss": 0.2978,
+      "step": 131
+    },
+    {
+      "epoch": 0.025241418873697293,
+      "grad_norm": 4.170378112848519,
+      "learning_rate": 6.205095818125637e-07,
+      "loss": 0.4204,
+      "step": 132
+    },
+    {
+      "epoch": 0.025432641743952576,
+      "grad_norm": 3.259364501702142,
+      "learning_rate": 6.214686855171402e-07,
+      "loss": 0.3252,
+      "step": 133
+    },
+    {
+      "epoch": 0.025623864614207858,
+      "grad_norm": 2.964526253689361,
+      "learning_rate": 6.224206048568647e-07,
+      "loss": 0.1165,
+      "step": 134
+    },
+    {
+      "epoch": 0.02581508748446314,
+      "grad_norm": 3.6816482665288732,
+      "learning_rate": 6.233654466639295e-07,
+      "loss": 0.1539,
+      "step": 135
+    },
+    {
+      "epoch": 0.026006310354718426,
+      "grad_norm": 6.571318886119907,
+      "learning_rate": 6.243033154051938e-07,
+      "loss": 0.1594,
+      "step": 136
+    },
+    {
+      "epoch": 0.02619753322497371,
+      "grad_norm": 4.394239573653678,
+      "learning_rate": 6.252343132515015e-07,
+      "loss": 0.2078,
+      "step": 137
+    },
+    {
+      "epoch": 0.02638875609522899,
+      "grad_norm": 4.631845511632262,
+      "learning_rate": 6.261585401444736e-07,
+      "loss": 0.5005,
+      "step": 138
+    },
+    {
+      "epoch": 0.026579978965484273,
+      "grad_norm": 7.2815266175996705,
+      "learning_rate": 6.270760938608936e-07,
+      "loss": 0.7544,
+      "step": 139
+    },
+    {
+      "epoch": 0.026771201835739555,
+      "grad_norm": 8.139810063726005,
+      "learning_rate": 6.279870700747812e-07,
+      "loss": 1.0577,
+      "step": 140
+    },
+    {
+      "epoch": 0.026962424705994838,
+      "grad_norm": 3.959787777502588,
+      "learning_rate": 6.288915624172606e-07,
+      "loss": 0.4815,
+      "step": 141
+    },
+    {
+      "epoch": 0.02715364757625012,
+      "grad_norm": 4.977633422412205,
+      "learning_rate": 6.297896625343093e-07,
+      "loss": 0.5626,
+      "step": 142
+    },
+    {
+      "epoch": 0.027344870446505402,
+      "grad_norm": 4.787229501438016,
+      "learning_rate": 6.306814601424844e-07,
+      "loss": 0.4885,
+      "step": 143
+    },
+    {
+      "epoch": 0.027536093316760685,
+      "grad_norm": 5.6227765245229655,
+      "learning_rate": 6.315670430827051e-07,
+      "loss": 0.5215,
+      "step": 144
+    },
+    {
+      "epoch": 0.027727316187015967,
+      "grad_norm": 2.9041402398708147,
+      "learning_rate": 6.324464973721739e-07,
+      "loss": 0.5722,
+      "step": 145
+    },
+    {
+      "epoch": 0.02791853905727125,
+      "grad_norm": 5.326043670532558,
+      "learning_rate": 6.333199072545141e-07,
+      "loss": 0.4261,
+      "step": 146
+    },
+    {
+      "epoch": 0.028109761927526532,
+      "grad_norm": 2.6571916594030296,
+      "learning_rate": 6.341873552481976e-07,
+      "loss": 0.1734,
+      "step": 147
+    },
+    {
+      "epoch": 0.028300984797781814,
+      "grad_norm": 4.829662288129763,
+      "learning_rate": 6.350489221933275e-07,
+      "loss": 0.3828,
+      "step": 148
+    },
+    {
+      "epoch": 0.028492207668037096,
+      "grad_norm": 3.654054553154509,
+      "learning_rate": 6.359046872968492e-07,
+      "loss": 0.196,
+      "step": 149
+    },
+    {
+      "epoch": 0.02868343053829238,
+      "grad_norm": 10.604885314292506,
+      "learning_rate": 6.367547281762469e-07,
+      "loss": 0.3811,
+      "step": 150
+    },
+    {
+      "epoch": 0.02887465340854766,
+      "grad_norm": 7.450070825022508,
+      "learning_rate": 6.375991209017893e-07,
+      "loss": 0.6186,
+      "step": 151
+    },
+    {
+      "epoch": 0.029065876278802943,
+      "grad_norm": 4.540021942690619,
+      "learning_rate": 6.384379400373815e-07,
+      "loss": 0.5988,
+      "step": 152
+    },
+    {
+      "epoch": 0.029257099149058226,
+      "grad_norm": 2.9331288382023284,
+      "learning_rate": 6.392712586800756e-07,
+      "loss": 0.4118,
+      "step": 153
+    },
+    {
+      "epoch": 0.02944832201931351,
+      "grad_norm": 5.494926887282672,
+      "learning_rate": 6.400991484982976e-07,
+      "loss": 0.8861,
+      "step": 154
+    },
+    {
+      "epoch": 0.029639544889568794,
+      "grad_norm": 4.567479173167016,
+      "learning_rate": 6.409216797688327e-07,
+      "loss": 0.4507,
+      "step": 155
+    },
+    {
+      "epoch": 0.029830767759824076,
+      "grad_norm": 6.408398970262428,
+      "learning_rate": 6.417389214126258e-07,
+      "loss": 0.512,
+      "step": 156
+    },
+    {
+      "epoch": 0.03002199063007936,
+      "grad_norm": 2.7135450152024436,
+      "learning_rate": 6.425509410294353e-07,
+      "loss": 0.1068,
+      "step": 157
+    },
+    {
+      "epoch": 0.03021321350033464,
+      "grad_norm": 4.676477111179603,
+      "learning_rate": 6.433578049313875e-07,
+      "loss": 0.5994,
+      "step": 158
+    },
+    {
+      "epoch": 0.030404436370589923,
+      "grad_norm": 2.664107604777627,
+      "learning_rate": 6.441595781754726e-07,
+      "loss": 0.0989,
+      "step": 159
+    },
+    {
+      "epoch": 0.030595659240845206,
+      "grad_norm": 5.023386241597931,
+      "learning_rate": 6.449563245950226e-07,
+      "loss": 0.3389,
+      "step": 160
+    },
+    {
+      "epoch": 0.030786882111100488,
+      "grad_norm": 3.0037810022958427,
+      "learning_rate": 6.457481068302073e-07,
+      "loss": 0.1202,
+      "step": 161
+    },
+    {
+      "epoch": 0.03097810498135577,
+      "grad_norm": 3.982112466523883,
+      "learning_rate": 6.46534986357587e-07,
+      "loss": 0.1542,
+      "step": 162
+    },
+    {
+      "epoch": 0.031169327851611053,
+      "grad_norm": 6.675431815477984,
+      "learning_rate": 6.473170235187556e-07,
+      "loss": 0.468,
+      "step": 163
+    },
+    {
+      "epoch": 0.03136055072186634,
+      "grad_norm": 6.050610396433867,
+      "learning_rate": 6.480942775481081e-07,
+      "loss": 0.8679,
+      "step": 164
+    },
+    {
+      "epoch": 0.03155177359212162,
+      "grad_norm": 6.723173167045605,
+      "learning_rate": 6.488668065997631e-07,
+      "loss": 0.612,
+      "step": 165
+    },
+    {
+      "epoch": 0.0317429964623769,
+      "grad_norm": 3.2010990341678425,
+      "learning_rate": 6.496346677736736e-07,
+      "loss": 0.7017,
+      "step": 166
+    },
+    {
+      "epoch": 0.031934219332632185,
+      "grad_norm": 5.257466955294226,
+      "learning_rate": 6.503979171409517e-07,
+      "loss": 0.3419,
+      "step": 167
+    },
+    {
+      "epoch": 0.03212544220288747,
+      "grad_norm": 5.506552520997499,
+      "learning_rate": 6.511566097684389e-07,
+      "loss": 0.4366,
+      "step": 168
+    },
+    {
+      "epoch": 0.03231666507314275,
+      "grad_norm": 5.0694803222129865,
+      "learning_rate": 6.519107997425466e-07,
+      "loss": 0.7164,
+      "step": 169
+    },
+    {
+      "epoch": 0.03250788794339803,
+      "grad_norm": 3.9770409196950327,
+      "learning_rate": 6.526605401923933e-07,
+      "loss": 0.4917,
+      "step": 170
+    },
+    {
+      "epoch": 0.032699110813653315,
+      "grad_norm": 4.750403939187616,
+      "learning_rate": 6.534058833122633e-07,
+      "loss": 0.2505,
+      "step": 171
+    },
+    {
+      "epoch": 0.0328903336839086,
+      "grad_norm": 3.888000583617241,
+      "learning_rate": 6.541468803834114e-07,
+      "loss": 0.2133,
+      "step": 172
+    },
+    {
+      "epoch": 0.03308155655416388,
+      "grad_norm": 7.5115064728846175,
+      "learning_rate": 6.548835817952339e-07,
+      "loss": 0.4257,
+      "step": 173
+    },
+    {
+      "epoch": 0.03327277942441916,
+      "grad_norm": 5.131375882536377,
+      "learning_rate": 6.556160370658314e-07,
+      "loss": 0.3669,
+      "step": 174
+    },
+    {
+      "epoch": 0.033464002294674444,
+      "grad_norm": 5.925333829722368,
+      "learning_rate": 6.563442948619808e-07,
+      "loss": 0.2119,
+      "step": 175
+    },
+    {
+      "epoch": 0.03365522516492973,
+      "grad_norm": 6.109607584787086,
+      "learning_rate": 6.570684030185389e-07,
+      "loss": 0.9612,
+      "step": 176
+    },
+    {
+      "epoch": 0.03384644803518501,
+      "grad_norm": 7.592810804268985,
+      "learning_rate": 6.577884085572973e-07,
+      "loss": 1.2779,
+      "step": 177
+    },
+    {
+      "epoch": 0.03403767090544029,
+      "grad_norm": 4.818424689844101,
+      "learning_rate": 6.585043577053045e-07,
+      "loss": 0.5594,
+      "step": 178
+    },
+    {
+      "epoch": 0.034228893775695574,
+      "grad_norm": 2.140917792796799,
+      "learning_rate": 6.592162959126751e-07,
+      "loss": 0.429,
+      "step": 179
+    },
+    {
+      "epoch": 0.034420116645950856,
+      "grad_norm": 3.4232742026873435,
+      "learning_rate": 6.599242678699045e-07,
+      "loss": 0.5855,
+      "step": 180
+    },
+    {
+      "epoch": 0.03461133951620614,
+      "grad_norm": 5.064149187635461,
+      "learning_rate": 6.606283175247016e-07,
+      "loss": 0.4072,
+      "step": 181
+    },
+    {
+      "epoch": 0.03480256238646142,
+      "grad_norm": 4.094918819456553,
+      "learning_rate": 6.613284880983595e-07,
+      "loss": 0.1943,
+      "step": 182
+    },
+    {
+      "epoch": 0.0349937852567167,
+      "grad_norm": 4.406188098662041,
+      "learning_rate": 6.620248221016777e-07,
+      "loss": 0.4841,
+      "step": 183
+    },
+    {
+      "epoch": 0.035185008126971985,
+      "grad_norm": 5.195480642250907,
+      "learning_rate": 6.627173613504487e-07,
+      "loss": 0.2962,
+      "step": 184
+    },
+    {
+      "epoch": 0.03537623099722727,
+      "grad_norm": 3.9543254974483966,
+      "learning_rate": 6.634061469805267e-07,
+      "loss": 0.1621,
+      "step": 185
+    },
+    {
+      "epoch": 0.03556745386748255,
+      "grad_norm": 3.5392007358660633,
+      "learning_rate": 6.640912194624904e-07,
+      "loss": 0.1637,
+      "step": 186
+    },
+    {
+      "epoch": 0.03575867673773783,
+      "grad_norm": 5.560534697969192,
+      "learning_rate": 6.647726186159094e-07,
+      "loss": 0.2361,
+      "step": 187
+    },
+    {
+      "epoch": 0.035949899607993115,
+      "grad_norm": 5.907825078297573,
+      "learning_rate": 6.654503836232355e-07,
+      "loss": 0.3953,
+      "step": 188
+    },
+    {
+      "epoch": 0.0361411224782484,
+      "grad_norm": 9.866067830376688,
+      "learning_rate": 6.661245530433207e-07,
+      "loss": 0.6147,
+      "step": 189
+    },
+    {
+      "epoch": 0.03633234534850368,
+      "grad_norm": 2.876182810410167,
+      "learning_rate": 6.667951648245808e-07,
+      "loss": 0.3441,
+      "step": 190
+    },
+    {
+      "epoch": 0.03652356821875896,
+      "grad_norm": 3.898483000893034,
+      "learning_rate": 6.674622563178132e-07,
+      "loss": 0.4829,
+      "step": 191
+    },
+    {
+      "epoch": 0.036714791089014244,
+      "grad_norm": 3.7940933650327384,
+      "learning_rate": 6.681258642886802e-07,
+      "loss": 0.4171,
+      "step": 192
+    },
+    {
+      "epoch": 0.036906013959269526,
+      "grad_norm": 5.557360042400451,
+      "learning_rate": 6.687860249298673e-07,
+      "loss": 0.686,
+      "step": 193
+    },
+    {
+      "epoch": 0.03709723682952481,
+      "grad_norm": 4.582950067390046,
+      "learning_rate": 6.694427738729285e-07,
+      "loss": 0.1294,
+      "step": 194
+    },
+    {
+      "epoch": 0.03728845969978009,
+      "grad_norm": 3.9915170125177046,
+      "learning_rate": 6.700961461998252e-07,
+      "loss": 0.2922,
+      "step": 195
+    },
+    {
+      "epoch": 0.03747968257003537,
+      "grad_norm": 2.879231034777087,
+      "learning_rate": 6.707461764541725e-07,
+      "loss": 0.0986,
+      "step": 196
+    },
+    {
+      "epoch": 0.037670905440290656,
+      "grad_norm": 5.607774256386905,
+      "learning_rate": 6.713928986521976e-07,
+      "loss": 0.2724,
+      "step": 197
+    },
+    {
+      "epoch": 0.03786212831054594,
+      "grad_norm": 5.180862224503005,
+      "learning_rate": 6.720363462934207e-07,
+      "loss": 0.2313,
+      "step": 198
+    },
+    {
+      "epoch": 0.03805335118080122,
+      "grad_norm": 3.8117356081255744,
+      "learning_rate": 6.726765523710678e-07,
+      "loss": 0.1609,
+      "step": 199
+    },
+    {
+      "epoch": 0.03824457405105651,
+      "grad_norm": 7.210009018129166,
+      "learning_rate": 6.733135493822219e-07,
+      "loss": 0.3358,
+      "step": 200
+    },
+    {
+      "epoch": 0.03843579692131179,
+      "grad_norm": 8.492666334533965,
+      "learning_rate": 6.739473693377217e-07,
+      "loss": 0.3859,
+      "step": 201
+    },
+    {
+      "epoch": 0.038627019791567074,
+      "grad_norm": 6.657593783994018,
+      "learning_rate": 6.745780437718133e-07,
+      "loss": 0.9877,
+      "step": 202
+    },
+    {
+      "epoch": 0.03881824266182236,
+      "grad_norm": 5.794901139055418,
+      "learning_rate": 6.752056037515651e-07,
+      "loss": 0.7135,
+      "step": 203
+    },
+    {
+      "epoch": 0.03900946553207764,
+      "grad_norm": 3.466788759893671,
+      "learning_rate": 6.758300798860507e-07,
+      "loss": 0.5724,
+      "step": 204
+    },
+    {
+      "epoch": 0.03920068840233292,
+      "grad_norm": 3.936866224842069,
+      "learning_rate": 6.764515023353075e-07,
+      "loss": 0.3336,
+      "step": 205
+    },
+    {
+      "epoch": 0.039391911272588204,
+      "grad_norm": 6.126907985668118,
+      "learning_rate": 6.77069900819076e-07,
+      "loss": 0.86,
+      "step": 206
+    },
+    {
+      "epoch": 0.039583134142843486,
+      "grad_norm": 4.746166690218831,
+      "learning_rate": 6.776853046253305e-07,
+      "loss": 0.2579,
+      "step": 207
+    },
+    {
+      "epoch": 0.03977435701309877,
+      "grad_norm": 4.346696691194836,
+      "learning_rate": 6.782977426186009e-07,
+      "loss": 0.5898,
+      "step": 208
+    },
+    {
+      "epoch": 0.03996557988335405,
+      "grad_norm": 3.961325240860613,
+      "learning_rate": 6.78907243248097e-07,
+      "loss": 0.1533,
+      "step": 209
+    },
+    {
+      "epoch": 0.04015680275360933,
+      "grad_norm": 3.265877891860168,
+      "learning_rate": 6.795138345556382e-07,
+      "loss": 0.4664,
+      "step": 210
+    },
+    {
+      "epoch": 0.040348025623864615,
+      "grad_norm": 5.271756459493638,
+      "learning_rate": 6.801175441833951e-07,
+      "loss": 0.4625,
+      "step": 211
+    },
+    {
+      "epoch": 0.0405392484941199,
+      "grad_norm": 6.008857229445331,
+      "learning_rate": 6.807183993814476e-07,
+      "loss": 0.1622,
+      "step": 212
+    },
+    {
+      "epoch": 0.04073047136437518,
+      "grad_norm": 5.633182423543458,
+      "learning_rate": 6.813164270151663e-07,
+      "loss": 0.496,
+      "step": 213
+    },
+    {
+      "epoch": 0.04092169423463046,
+      "grad_norm": 9.634327517976168,
+      "learning_rate": 6.819116535724185e-07,
+      "loss": 0.6407,
+      "step": 214
+    },
+    {
+      "epoch": 0.041112917104885745,
+      "grad_norm": 6.0732358278277125,
+      "learning_rate": 6.825041051706106e-07,
+      "loss": 1.0272,
+      "step": 215
+    },
+    {
+      "epoch": 0.04130413997514103,
+      "grad_norm": 5.078875207505407,
+      "learning_rate": 6.830938075635621e-07,
+      "loss": 0.7949,
+      "step": 216
+    },
+    {
+      "epoch": 0.04149536284539631,
+      "grad_norm": 2.3651475835322473,
+      "learning_rate": 6.83680786148224e-07,
+      "loss": 0.5617,
+      "step": 217
+    },
+    {
+      "epoch": 0.04168658571565159,
+      "grad_norm": 4.632863345196851,
+      "learning_rate": 6.842650659712424e-07,
+      "loss": 0.2878,
+      "step": 218
+    },
+    {
+      "epoch": 0.041877808585906874,
+      "grad_norm": 4.920309344545372,
+      "learning_rate": 6.84846671735371e-07,
+      "loss": 0.2266,
+      "step": 219
+    },
+    {
+      "epoch": 0.042069031456162156,
+      "grad_norm": 4.391651834442359,
+      "learning_rate": 6.854256278057382e-07,
+      "loss": 0.5281,
+      "step": 220
+    },
+    {
+      "epoch": 0.04226025432641744,
+      "grad_norm": 4.699626039277765,
+      "learning_rate": 6.860019582159714e-07,
+      "loss": 0.1359,
+      "step": 221
+    },
+    {
+      "epoch": 0.04245147719667272,
+      "grad_norm": 5.035540344882702,
+      "learning_rate": 6.865756866741845e-07,
+      "loss": 0.1848,
+      "step": 222
+    },
+    {
+      "epoch": 0.042642700066928,
+      "grad_norm": 3.4953834877769183,
+      "learning_rate": 6.871468365688285e-07,
+      "loss": 0.2822,
+      "step": 223
+    },
+    {
+      "epoch": 0.042833922937183286,
+      "grad_norm": 4.135574888698569,
+      "learning_rate": 6.877154309744139e-07,
+      "loss": 0.2248,
+      "step": 224
+    },
+    {
+      "epoch": 0.04302514580743857,
+      "grad_norm": 6.821521909612992,
+      "learning_rate": 6.882814926571037e-07,
+      "loss": 0.2834,
+      "step": 225
+    },
+    {
+      "epoch": 0.04321636867769385,
+      "grad_norm": 8.975310474859194,
+      "learning_rate": 6.888450440801833e-07,
+      "loss": 0.8572,
+      "step": 226
+    },
+    {
+      "epoch": 0.04340759154794913,
+      "grad_norm": 5.722096615660651,
+      "learning_rate": 6.894061074094085e-07,
+      "loss": 0.8428,
+      "step": 227
+    },
+    {
+      "epoch": 0.043598814418204415,
+      "grad_norm": 5.096848931821844,
+      "learning_rate": 6.899647045182384e-07,
+      "loss": 0.3697,
+      "step": 228
+    },
+    {
+      "epoch": 0.0437900372884597,
+      "grad_norm": 5.052490511606434,
+      "learning_rate": 6.905208569929507e-07,
+      "loss": 0.6941,
+      "step": 229
+    },
+    {
+      "epoch": 0.04398126015871498,
+      "grad_norm": 3.5997837001670367,
+      "learning_rate": 6.91074586137648e-07,
+      "loss": 0.4985,
+      "step": 230
+    },
+    {
+      "epoch": 0.04417248302897026,
+      "grad_norm": 6.238471728756815,
+      "learning_rate": 6.916259129791544e-07,
+      "loss": 0.7925,
+      "step": 231
+    },
+    {
+      "epoch": 0.044363705899225545,
+      "grad_norm": 3.8454411205743613,
+      "learning_rate": 6.921748582718064e-07,
+      "loss": 0.3559,
+      "step": 232
+    },
+    {
+      "epoch": 0.04455492876948083,
+      "grad_norm": 4.177147437513031,
+      "learning_rate": 6.927214425021408e-07,
+      "loss": 0.3671,
+      "step": 233
+    },
+    {
+      "epoch": 0.04474615163973611,
+      "grad_norm": 4.776688537831536,
+      "learning_rate": 6.932656858934828e-07,
+      "loss": 0.1996,
+      "step": 234
+    },
+    {
+      "epoch": 0.04493737450999139,
+      "grad_norm": 4.407815857905935,
+      "learning_rate": 6.938076084104349e-07,
+      "loss": 0.3762,
+      "step": 235
+    },
+    {
+      "epoch": 0.04512859738024668,
+      "grad_norm": 5.083136799666235,
+      "learning_rate": 6.943472297632724e-07,
+      "loss": 0.3482,
+      "step": 236
+    },
+    {
+      "epoch": 0.04531982025050196,
+      "grad_norm": 5.358915882074689,
+      "learning_rate": 6.948845694122444e-07,
+      "loss": 0.2289,
+      "step": 237
+    },
+    {
+      "epoch": 0.045511043120757246,
+      "grad_norm": 5.152327461592904,
+      "learning_rate": 6.954196465717846e-07,
+      "loss": 0.6089,
+      "step": 238
+    },
+    {
+      "epoch": 0.04570226599101253,
+      "grad_norm": 6.328483062786576,
+      "learning_rate": 6.959524802146344e-07,
+      "loss": 0.6642,
+      "step": 239
+    },
+    {
+      "epoch": 0.04589348886126781,
+      "grad_norm": 7.100733071750409,
+      "learning_rate": 6.964830890758795e-07,
+      "loss": 0.8345,
+      "step": 240
+    },
+    {
+      "epoch": 0.04608471173152309,
+      "grad_norm": 3.0577110683048336,
+      "learning_rate": 6.970114916569027e-07,
+      "loss": 0.4561,
+      "step": 241
+    },
+    {
+      "epoch": 0.046275934601778375,
+      "grad_norm": 3.7177399618025984,
+      "learning_rate": 6.975377062292544e-07,
+      "loss": 0.5181,
+      "step": 242
+    },
+    {
+      "epoch": 0.04646715747203366,
+      "grad_norm": 5.277155445729136,
+      "learning_rate": 6.980617508384439e-07,
+      "loss": 0.7705,
+      "step": 243
+    },
+    {
+      "epoch": 0.04665838034228894,
+      "grad_norm": 5.4515787900122055,
+      "learning_rate": 6.985836433076528e-07,
+      "loss": 0.5063,
+      "step": 244
+    },
+    {
+      "epoch": 0.04684960321254422,
+      "grad_norm": 3.84069690559436,
+      "learning_rate": 6.991034012413719e-07,
+      "loss": 0.3251,
+      "step": 245
+    },
+    {
+      "epoch": 0.047040826082799504,
+      "grad_norm": 4.295105503904747,
+      "learning_rate": 6.996210420289649e-07,
+      "loss": 0.2984,
+      "step": 246
+    },
+    {
+      "epoch": 0.04723204895305479,
+      "grad_norm": 4.336486451736445,
+      "learning_rate": 7.00136582848159e-07,
+      "loss": 0.1566,
+      "step": 247
+    },
+    {
+      "epoch": 0.04742327182331007,
+      "grad_norm": 3.5838690216875695,
+      "learning_rate": 7.006500406684653e-07,
+      "loss": 0.2546,
+      "step": 248
+    },
+    {
+      "epoch": 0.04761449469356535,
+      "grad_norm": 4.561684731580212,
+      "learning_rate": 7.011614322545303e-07,
+      "loss": 0.2171,
+      "step": 249
+    },
+    {
+      "epoch": 0.047805717563820634,
+      "grad_norm": 6.3801461186911395,
+      "learning_rate": 7.016707741694213e-07,
+      "loss": 0.3324,
+      "step": 250
+    },
+    {
+      "epoch": 0.047996940434075916,
+      "grad_norm": 8.795516727398859,
+      "learning_rate": 7.021780827778432e-07,
+      "loss": 0.3648,
+      "step": 251
+    },
+    {
+      "epoch": 0.0481881633043312,
+      "grad_norm": 4.665740511589171,
+      "learning_rate": 7.026833742492958e-07,
+      "loss": 0.5456,
+      "step": 252
+    },
+    {
+      "epoch": 0.04837938617458648,
+      "grad_norm": 3.646178630644333,
+      "learning_rate": 7.031866645611642e-07,
+      "loss": 0.3365,
+      "step": 253
+    },
+    {
+      "epoch": 0.04857060904484176,
+      "grad_norm": 3.5785823346513776,
+      "learning_rate": 7.036879695017514e-07,
+      "loss": 0.3914,
+      "step": 254
+    },
+    {
+      "epoch": 0.048761831915097045,
+      "grad_norm": 3.2650026157245993,
+      "learning_rate": 7.0418730467325e-07,
+      "loss": 0.3578,
+      "step": 255
+    },
+    {
+      "epoch": 0.04895305478535233,
+      "grad_norm": 2.5366879671917686,
+      "learning_rate": 7.046846854946552e-07,
+      "loss": 0.2873,
+      "step": 256
+    },
+    {
+      "epoch": 0.04914427765560761,
+      "grad_norm": 5.022836414879023,
+      "learning_rate": 7.051801272046224e-07,
+      "loss": 0.6164,
+      "step": 257
+    },
+    {
+      "epoch": 0.04933550052586289,
+      "grad_norm": 3.3696152858572175,
+      "learning_rate": 7.056736448642682e-07,
+      "loss": 0.2286,
+      "step": 258
+    },
+    {
+      "epoch": 0.049526723396118175,
+      "grad_norm": 7.402289124138813,
+      "learning_rate": 7.06165253359918e-07,
+      "loss": 0.4108,
+      "step": 259
+    },
+    {
+      "epoch": 0.04971794626637346,
+      "grad_norm": 4.233452615534232,
+      "learning_rate": 7.066549674058002e-07,
+      "loss": 0.1775,
+      "step": 260
+    },
+    {
+      "epoch": 0.04990916913662874,
+      "grad_norm": 3.0549213290659756,
+      "learning_rate": 7.071428015466883e-07,
+      "loss": 0.242,
+      "step": 261
+    },
+    {
+      "epoch": 0.05010039200688402,
+      "grad_norm": 3.7388523984894655,
+      "learning_rate": 7.076287701604933e-07,
+      "loss": 0.1273,
+      "step": 262
+    },
+    {
+      "epoch": 0.050291614877139304,
+      "grad_norm": 5.547141828969504,
+      "learning_rate": 7.081128874608072e-07,
+      "loss": 0.2766,
+      "step": 263
+    },
+    {
+      "epoch": 0.050482837747394586,
+      "grad_norm": 5.193538598120362,
+      "learning_rate": 7.085951674993957e-07,
+      "loss": 0.5723,
+      "step": 264
+    },
+    {
+      "epoch": 0.05067406061764987,
+      "grad_norm": 4.977483786877945,
+      "learning_rate": 7.09075624168647e-07,
+      "loss": 0.3168,
+      "step": 265
+    },
+    {
+      "epoch": 0.05086528348790515,
+      "grad_norm": 4.445842798444021,
+      "learning_rate": 7.095542712039721e-07,
+      "loss": 0.7764,
+      "step": 266
+    },
+    {
+      "epoch": 0.05105650635816043,
+      "grad_norm": 4.001336589664032,
+      "learning_rate": 7.100311221861613e-07,
+      "loss": 0.4774,
+      "step": 267
+    },
+    {
+      "epoch": 0.051247729228415716,
+      "grad_norm": 3.733741993903512,
+      "learning_rate": 7.105061905436966e-07,
+      "loss": 0.1737,
+      "step": 268
+    },
+    {
+      "epoch": 0.051438952098671,
+      "grad_norm": 5.158823746564726,
+      "learning_rate": 7.109794895550202e-07,
+      "loss": 0.4786,
+      "step": 269
+    },
+    {
+      "epoch": 0.05163017496892628,
+      "grad_norm": 3.287217154863667,
+      "learning_rate": 7.114510323507614e-07,
+      "loss": 0.3361,
+      "step": 270
+    },
+    {
+      "epoch": 0.05182139783918156,
+      "grad_norm": 2.6198389088268623,
+      "learning_rate": 7.119208319159228e-07,
+      "loss": 0.2647,
+      "step": 271
+    },
+    {
+      "epoch": 0.05201262070943685,
+      "grad_norm": 3.3784436212374307,
+      "learning_rate": 7.123889010920259e-07,
+      "loss": 0.1998,
+      "step": 272
+    },
+    {
+      "epoch": 0.052203843579692134,
+      "grad_norm": 3.377927711223764,
+      "learning_rate": 7.128552525792165e-07,
+      "loss": 0.0999,
+      "step": 273
+    },
+    {
+      "epoch": 0.05239506644994742,
+      "grad_norm": 2.8466125166281797,
+      "learning_rate": 7.133198989383333e-07,
+      "loss": 0.1114,
+      "step": 274
+    },
+    {
+      "epoch": 0.0525862893202027,
+      "grad_norm": 6.700983163238354,
+      "learning_rate": 7.137828525929376e-07,
+      "loss": 0.1937,
+      "step": 275
+    },
+    {
+      "epoch": 0.05277751219045798,
+      "grad_norm": 8.842308331311262,
+      "learning_rate": 7.142441258313054e-07,
+      "loss": 0.4153,
+      "step": 276
+    },
+    {
+      "epoch": 0.052968735060713264,
+      "grad_norm": 3.9799241282069913,
+      "learning_rate": 7.147037308083868e-07,
+      "loss": 0.4791,
+      "step": 277
+    },
+    {
+      "epoch": 0.053159957930968546,
+      "grad_norm": 3.2030054581828185,
+      "learning_rate": 7.151616795477254e-07,
+      "loss": 0.3231,
+      "step": 278
+    },
+    {
+      "epoch": 0.05335118080122383,
+      "grad_norm": 4.137896405272443,
+      "learning_rate": 7.15617983943347e-07,
+      "loss": 0.4528,
+      "step": 279
+    },
+    {
+      "epoch": 0.05354240367147911,
+      "grad_norm": 3.9535548764230746,
+      "learning_rate": 7.160726557616131e-07,
+      "loss": 0.2657,
+      "step": 280
+    },
+    {
+      "epoch": 0.05373362654173439,
+      "grad_norm": 4.420138068480847,
+      "learning_rate": 7.165257066430412e-07,
+      "loss": 0.4264,
+      "step": 281
+    },
+    {
+      "epoch": 0.053924849411989675,
+      "grad_norm": 2.41510935186329,
+      "learning_rate": 7.169771481040925e-07,
+      "loss": 0.3967,
+      "step": 282
+    },
+    {
+      "epoch": 0.05411607228224496,
+      "grad_norm": 3.930736849093163,
+      "learning_rate": 7.174269915389296e-07,
+      "loss": 0.249,
+      "step": 283
+    },
+    {
+      "epoch": 0.05430729515250024,
+      "grad_norm": 3.546594818745103,
+      "learning_rate": 7.178752482211412e-07,
+      "loss": 0.2049,
+      "step": 284
+    },
+    {
+      "epoch": 0.05449851802275552,
+      "grad_norm": 5.6417728851702105,
+      "learning_rate": 7.183219293054376e-07,
+      "loss": 0.472,
+      "step": 285
+    },
+    {
+      "epoch": 0.054689740893010805,
+      "grad_norm": 6.32047364886494,
+      "learning_rate": 7.187670458293164e-07,
+      "loss": 0.4477,
+      "step": 286
+    },
+    {
+      "epoch": 0.05488096376326609,
+      "grad_norm": 4.7992015253969145,
+      "learning_rate": 7.192106087146986e-07,
+      "loss": 0.231,
+      "step": 287
+    },
+    {
+      "epoch": 0.05507218663352137,
+      "grad_norm": 30.856213247838657,
+      "learning_rate": 7.196526287695371e-07,
+      "loss": 0.4011,
+      "step": 288
+    },
+    {
+      "epoch": 0.05526340950377665,
+      "grad_norm": 5.0208120174140145,
+      "learning_rate": 7.200931166893963e-07,
+      "loss": 0.4237,
+      "step": 289
+    },
+    {
+      "epoch": 0.055454632374031934,
+      "grad_norm": 3.434157671313009,
+      "learning_rate": 7.205320830590057e-07,
+      "loss": 0.2537,
+      "step": 290
+    },
+    {
+      "epoch": 0.055645855244287216,
+      "grad_norm": 2.283736872225542,
+      "learning_rate": 7.209695383537853e-07,
+      "loss": 0.3596,
+      "step": 291
+    },
+    {
+      "epoch": 0.0558370781145425,
+      "grad_norm": 4.651564320598191,
+      "learning_rate": 7.214054929413462e-07,
+      "loss": 0.3323,
+      "step": 292
+    },
+    {
+      "epoch": 0.05602830098479778,
+      "grad_norm": 4.440733899815404,
+      "learning_rate": 7.21839957082964e-07,
+      "loss": 0.2647,
+      "step": 293
+    },
+    {
+      "epoch": 0.056219523855053063,
+      "grad_norm": 4.48700065703586,
+      "learning_rate": 7.222729409350294e-07,
+      "loss": 0.3825,
+      "step": 294
+    },
+    {
+      "epoch": 0.056410746725308346,
+      "grad_norm": 6.15977508778365,
+      "learning_rate": 7.227044545504718e-07,
+      "loss": 0.5826,
+      "step": 295
+    },
+    {
+      "epoch": 0.05660196959556363,
+      "grad_norm": 4.171149830174135,
+      "learning_rate": 7.231345078801594e-07,
+      "loss": 0.2288,
+      "step": 296
+    },
+    {
+      "epoch": 0.05679319246581891,
+      "grad_norm": 3.0452734234701895,
+      "learning_rate": 7.235631107742775e-07,
+      "loss": 0.2099,
+      "step": 297
+    },
+    {
+      "epoch": 0.05698441533607419,
+      "grad_norm": 5.943806401094454,
+      "learning_rate": 7.239902729836812e-07,
+      "loss": 0.402,
+      "step": 298
+    },
+    {
+      "epoch": 0.057175638206329475,
+      "grad_norm": 3.2526033685019953,
+      "learning_rate": 7.244160041612263e-07,
+      "loss": 0.0948,
+      "step": 299
+    },
+    {
+      "epoch": 0.05736686107658476,
+      "grad_norm": 7.674983061234362,
+      "learning_rate": 7.248403138630788e-07,
+      "loss": 0.2311,
+      "step": 300
+    },
+    {
+      "epoch": 0.05755808394684004,
+      "grad_norm": 8.946314593160956,
+      "learning_rate": 7.252632115500019e-07,
+      "loss": 0.3702,
+      "step": 301
+    },
+    {
+      "epoch": 0.05774930681709532,
+      "grad_norm": 5.485635900505078,
+      "learning_rate": 7.256847065886214e-07,
+      "loss": 0.2843,
+      "step": 302
+    },
+    {
+      "epoch": 0.057940529687350605,
+      "grad_norm": 2.181503141300826,
+      "learning_rate": 7.2610480825267e-07,
+      "loss": 0.4001,
+      "step": 303
+    },
+    {
+      "epoch": 0.05813175255760589,
+      "grad_norm": 2.532821920708016,
+      "learning_rate": 7.265235257242134e-07,
+      "loss": 0.5317,
+      "step": 304
+    },
+    {
+      "epoch": 0.05832297542786117,
+      "grad_norm": 4.44877738873464,
+      "learning_rate": 7.269408680948521e-07,
+      "loss": 0.3738,
+      "step": 305
+    },
+    {
+      "epoch": 0.05851419829811645,
+      "grad_norm": 3.669403543051019,
+      "learning_rate": 7.273568443669076e-07,
+      "loss": 0.2896,
+      "step": 306
+    },
+    {
+      "epoch": 0.058705421168371734,
+      "grad_norm": 2.8053316104591115,
+      "learning_rate": 7.277714634545871e-07,
+      "loss": 0.2209,
+      "step": 307
+    },
+    {
+      "epoch": 0.05889664403862702,
+      "grad_norm": 4.280136297676111,
+      "learning_rate": 7.281847341851294e-07,
+      "loss": 0.3417,
+      "step": 308
+    },
+    {
+      "epoch": 0.059087866908882306,
+      "grad_norm": 3.8998798156217513,
+      "learning_rate": 7.28596665299933e-07,
+      "loss": 0.1973,
+      "step": 309
+    },
+    {
+      "epoch": 0.05927908977913759,
+      "grad_norm": 6.375902897343062,
+      "learning_rate": 7.290072654556646e-07,
+      "loss": 0.359,
+      "step": 310
+    },
+    {
+      "epoch": 0.05947031264939287,
+      "grad_norm": 5.696797578906003,
+      "learning_rate": 7.294165432253522e-07,
+      "loss": 0.3979,
+      "step": 311
+    },
+    {
+      "epoch": 0.05966153551964815,
+      "grad_norm": 2.3430100353708236,
+      "learning_rate": 7.298245070994577e-07,
+      "loss": 0.1702,
+      "step": 312
+    },
+    {
+      "epoch": 0.059852758389903435,
+      "grad_norm": 9.13389550959369,
+      "learning_rate": 7.302311654869346e-07,
+      "loss": 0.4498,
+      "step": 313
+    },
+    {
+      "epoch": 0.06004398126015872,
+      "grad_norm": 11.119631725772075,
+      "learning_rate": 7.306365267162671e-07,
+      "loss": 0.8528,
+      "step": 314
+    },
+    {
+      "epoch": 0.060235204130414,
+      "grad_norm": 5.4380112330947465,
+      "learning_rate": 7.31040599036495e-07,
+      "loss": 0.9397,
+      "step": 315
+    },
+    {
+      "epoch": 0.06042642700066928,
+      "grad_norm": 6.641813714837794,
+      "learning_rate": 7.314433906182193e-07,
+      "loss": 0.4006,
+      "step": 316
+    },
+    {
+      "epoch": 0.060617649870924564,
+      "grad_norm": 5.747726654455302,
+      "learning_rate": 7.318449095545947e-07,
+      "loss": 0.3518,
+      "step": 317
+    },
+    {
+      "epoch": 0.06080887274117985,
+      "grad_norm": 5.899927701345459,
+      "learning_rate": 7.322451638623045e-07,
+      "loss": 0.8449,
+      "step": 318
+    },
+    {
+      "epoch": 0.06100009561143513,
+      "grad_norm": 1.728176139640518,
+      "learning_rate": 7.326441614825219e-07,
+      "loss": 0.3345,
+      "step": 319
+    },
+    {
+      "epoch": 0.06119131848169041,
+      "grad_norm": 2.312683510587128,
+      "learning_rate": 7.330419102818544e-07,
+      "loss": 0.188,
+      "step": 320
+    },
+    {
+      "epoch": 0.061382541351945694,
+      "grad_norm": 4.619871440601518,
+      "learning_rate": 7.334384180532755e-07,
+      "loss": 0.2106,
+      "step": 321
+    },
+    {
+      "epoch": 0.061573764222200976,
+      "grad_norm": 3.929484974551012,
+      "learning_rate": 7.338336925170393e-07,
+      "loss": 0.4164,
+      "step": 322
+    },
+    {
+      "epoch": 0.06176498709245626,
+      "grad_norm": 4.805025712181628,
+      "learning_rate": 7.342277413215839e-07,
+      "loss": 0.358,
+      "step": 323
+    },
+    {
+      "epoch": 0.06195620996271154,
+      "grad_norm": 3.261415387511229,
+      "learning_rate": 7.346205720444188e-07,
+      "loss": 0.2095,
+      "step": 324
+    },
+    {
+      "epoch": 0.06214743283296682,
+      "grad_norm": 5.720588096850544,
+      "learning_rate": 7.350121921929994e-07,
+      "loss": 0.1996,
+      "step": 325
+    },
+    {
+      "epoch": 0.062338655703222105,
+      "grad_norm": 7.27416542745734,
+      "learning_rate": 7.354026092055876e-07,
+      "loss": 1.1044,
+      "step": 326
+    },
+    {
+      "epoch": 0.06252987857347739,
+      "grad_norm": 5.878037012669659,
+      "learning_rate": 7.357918304520992e-07,
+      "loss": 0.6082,
+      "step": 327
+    },
+    {
+      "epoch": 0.06272110144373268,
+      "grad_norm": 3.625093130033491,
+      "learning_rate": 7.3617986323494e-07,
+      "loss": 0.3839,
+      "step": 328
+    },
+    {
+      "epoch": 0.06291232431398795,
+      "grad_norm": 2.8132845314216492,
+      "learning_rate": 7.365667147898262e-07,
+      "loss": 0.45,
+      "step": 329
+    },
+    {
+      "epoch": 0.06310354718424324,
+      "grad_norm": 3.4344706278183788,
+      "learning_rate": 7.36952392286595e-07,
+      "loss": 0.2434,
+      "step": 330
+    },
+    {
+      "epoch": 0.06329477005449852,
+      "grad_norm": 4.6482832169774655,
+      "learning_rate": 7.373369028300019e-07,
+      "loss": 0.4288,
+      "step": 331
+    },
+    {
+      "epoch": 0.0634859929247538,
+      "grad_norm": 3.5629153093223827,
+      "learning_rate": 7.377202534605055e-07,
+      "loss": 0.208,
+      "step": 332
+    },
+    {
+      "epoch": 0.06367721579500908,
+      "grad_norm": 3.9291296500400352,
+      "learning_rate": 7.381024511550413e-07,
+      "loss": 0.3078,
+      "step": 333
+    },
+    {
+      "epoch": 0.06386843866526437,
+      "grad_norm": 3.92617124831324,
+      "learning_rate": 7.384835028277836e-07,
+      "loss": 0.1257,
+      "step": 334
+    },
+    {
+      "epoch": 0.06405966153551965,
+      "grad_norm": 4.557324962675987,
+      "learning_rate": 7.388634153308961e-07,
+      "loss": 0.1199,
+      "step": 335
+    },
+    {
+      "epoch": 0.06425088440577494,
+      "grad_norm": 3.9758671900253066,
+      "learning_rate": 7.392421954552708e-07,
+      "loss": 0.1225,
+      "step": 336
+    },
+    {
+      "epoch": 0.06444210727603021,
+      "grad_norm": 2.6434606190838585,
+      "learning_rate": 7.39619849931257e-07,
+      "loss": 0.0612,
+      "step": 337
+    },
+    {
+      "epoch": 0.0646333301462855,
+      "grad_norm": 11.83123037071635,
+      "learning_rate": 7.399963854293785e-07,
+      "loss": 0.4875,
+      "step": 338
+    },
+    {
+      "epoch": 0.06482455301654078,
+      "grad_norm": 5.58317751690762,
+      "learning_rate": 7.403718085610401e-07,
+      "loss": 0.561,
+      "step": 339
+    },
+    {
+      "epoch": 0.06501577588679607,
+      "grad_norm": 4.807786887779114,
+      "learning_rate": 7.407461258792252e-07,
+      "loss": 0.3798,
+      "step": 340
+    },
+    {
+      "epoch": 0.06520699875705134,
+      "grad_norm": 4.1694541508031255,
+      "learning_rate": 7.411193438791809e-07,
+      "loss": 0.4548,
+      "step": 341
+    },
+    {
+      "epoch": 0.06539822162730663,
+      "grad_norm": 3.1375439462205676,
+      "learning_rate": 7.414914689990952e-07,
+      "loss": 0.4959,
+      "step": 342
+    },
+    {
+      "epoch": 0.0655894444975619,
+      "grad_norm": 3.2111035232131635,
+      "learning_rate": 7.418625076207631e-07,
+      "loss": 0.2257,
+      "step": 343
+    },
+    {
+      "epoch": 0.0657806673678172,
+      "grad_norm": 3.2660217044200235,
+      "learning_rate": 7.422324660702433e-07,
+      "loss": 0.3689,
+      "step": 344
+    },
+    {
+      "epoch": 0.06597189023807247,
+      "grad_norm": 11.192866737708686,
+      "learning_rate": 7.42601350618505e-07,
+      "loss": 0.8265,
+      "step": 345
+    },
+    {
+      "epoch": 0.06616311310832776,
+      "grad_norm": 3.795682240957007,
+      "learning_rate": 7.429691674820658e-07,
+      "loss": 0.2014,
+      "step": 346
+    },
+    {
+      "epoch": 0.06635433597858303,
+      "grad_norm": 4.8090964828461535,
+      "learning_rate": 7.433359228236214e-07,
+      "loss": 0.1405,
+      "step": 347
+    },
+    {
+      "epoch": 0.06654555884883832,
+      "grad_norm": 2.814028091686286,
+      "learning_rate": 7.437016227526632e-07,
+      "loss": 0.123,
+      "step": 348
+    },
+    {
+      "epoch": 0.0667367817190936,
+      "grad_norm": 3.7780906151909908,
+      "learning_rate": 7.440662733260908e-07,
+      "loss": 0.1654,
+      "step": 349
+    },
+    {
+      "epoch": 0.06692800458934889,
+      "grad_norm": 9.229957517978693,
+      "learning_rate": 7.444298805488125e-07,
+      "loss": 0.2545,
+      "step": 350
+    },
+    {
+      "epoch": 0.06711922745960416,
+      "grad_norm": 6.38703381055845,
+      "learning_rate": 7.447924503743396e-07,
+      "loss": 0.2295,
+      "step": 351
+    },
+    {
+      "epoch": 0.06731045032985945,
+      "grad_norm": 5.6030371944698265,
+      "learning_rate": 7.451539887053707e-07,
+      "loss": 0.7123,
+      "step": 352
+    },
+    {
+      "epoch": 0.06750167320011473,
+      "grad_norm": 5.637547549028186,
+      "learning_rate": 7.455145013943688e-07,
+      "loss": 0.7833,
+      "step": 353
+    },
+    {
+      "epoch": 0.06769289607037002,
+      "grad_norm": 5.038165345280633,
+      "learning_rate": 7.458739942441292e-07,
+      "loss": 0.2688,
+      "step": 354
+    },
+    {
+      "epoch": 0.0678841189406253,
+      "grad_norm": 4.113092749642668,
+      "learning_rate": 7.462324730083404e-07,
+      "loss": 0.681,
+      "step": 355
+    },
+    {
+      "epoch": 0.06807534181088058,
+      "grad_norm": 4.734615911516989,
+      "learning_rate": 7.465899433921365e-07,
+      "loss": 0.4157,
+      "step": 356
+    },
+    {
+      "epoch": 0.06826656468113586,
+      "grad_norm": 3.668744250005476,
+      "learning_rate": 7.469464110526413e-07,
+      "loss": 0.2831,
+      "step": 357
+    },
+    {
+      "epoch": 0.06845778755139115,
+      "grad_norm": 3.255172135215404,
+      "learning_rate": 7.473018815995071e-07,
+      "loss": 0.3168,
+      "step": 358
+    },
+    {
+      "epoch": 0.06864901042164642,
+      "grad_norm": 4.983738203836052,
+      "learning_rate": 7.476563605954426e-07,
+      "loss": 0.1816,
+      "step": 359
+    },
+    {
+      "epoch": 0.06884023329190171,
+      "grad_norm": 3.3671346779771416,
+      "learning_rate": 7.480098535567364e-07,
+      "loss": 0.2102,
+      "step": 360
+    },
+    {
+      "epoch": 0.06903145616215699,
+      "grad_norm": 6.1235851288540335,
+      "learning_rate": 7.483623659537714e-07,
+      "loss": 0.1974,
+      "step": 361
+    },
+    {
+      "epoch": 0.06922267903241228,
+      "grad_norm": 6.489894934980831,
+      "learning_rate": 7.487139032115335e-07,
+      "loss": 0.2639,
+      "step": 362
+    },
+    {
+      "epoch": 0.06941390190266757,
+      "grad_norm": 8.073531772601328,
+      "learning_rate": 7.490644707101113e-07,
+      "loss": 0.4728,
+      "step": 363
+    },
+    {
+      "epoch": 0.06960512477292284,
+      "grad_norm": 4.905531132023918,
+      "learning_rate": 7.494140737851915e-07,
+      "loss": 0.4812,
+      "step": 364
+    },
+    {
+      "epoch": 0.06979634764317813,
+      "grad_norm": 8.345002612552953,
+      "learning_rate": 7.497627177285454e-07,
+      "loss": 0.6866,
+      "step": 365
+    },
+    {
+      "epoch": 0.0699875705134334,
+      "grad_norm": 2.6232142959530917,
+      "learning_rate": 7.501104077885095e-07,
+      "loss": 0.1334,
+      "step": 366
+    },
+    {
+      "epoch": 0.0701787933836887,
+      "grad_norm": 4.248065283724481,
+      "learning_rate": 7.504571491704603e-07,
+      "loss": 0.2996,
+      "step": 367
+    },
+    {
+      "epoch": 0.07037001625394397,
+      "grad_norm": 5.698535316073347,
+      "learning_rate": 7.508029470372806e-07,
+      "loss": 0.6003,
+      "step": 368
+    },
+    {
+      "epoch": 0.07056123912419926,
+      "grad_norm": 4.366781744763125,
+      "learning_rate": 7.511478065098218e-07,
+      "loss": 0.3602,
+      "step": 369
+    },
+    {
+      "epoch": 0.07075246199445454,
+      "grad_norm": 4.596289523123091,
+      "learning_rate": 7.514917326673586e-07,
+      "loss": 0.1619,
+      "step": 370
+    },
+    {
+      "epoch": 0.07094368486470982,
+      "grad_norm": 3.5409178596714765,
+      "learning_rate": 7.518347305480383e-07,
+      "loss": 0.1819,
+      "step": 371
+    },
+    {
+      "epoch": 0.0711349077349651,
+      "grad_norm": 4.913747513565913,
+      "learning_rate": 7.521768051493222e-07,
+      "loss": 0.1535,
+      "step": 372
+    },
+    {
+      "epoch": 0.07132613060522039,
+      "grad_norm": 5.883575673646047,
+      "learning_rate": 7.525179614284244e-07,
+      "loss": 0.2136,
+      "step": 373
+    },
+    {
+      "epoch": 0.07151735347547566,
+      "grad_norm": 5.018748609036719,
+      "learning_rate": 7.528582043027414e-07,
+      "loss": 0.417,
+      "step": 374
+    },
+    {
+      "epoch": 0.07170857634573095,
+      "grad_norm": 7.886445704396162,
+      "learning_rate": 7.531975386502781e-07,
+      "loss": 0.1928,
+      "step": 375
+    },
+    {
+      "epoch": 0.07189979921598623,
+      "grad_norm": 5.224732219642244,
+      "learning_rate": 7.535359693100674e-07,
+      "loss": 0.5428,
+      "step": 376
+    },
+    {
+      "epoch": 0.07209102208624152,
+      "grad_norm": 6.541555059497363,
+      "learning_rate": 7.538735010825839e-07,
+      "loss": 0.96,
+      "step": 377
+    },
+    {
+      "epoch": 0.0722822449564968,
+      "grad_norm": 4.199542819980597,
+      "learning_rate": 7.542101387301527e-07,
+      "loss": 0.5851,
+      "step": 378
+    },
+    {
+      "epoch": 0.07247346782675208,
+      "grad_norm": 2.396258330313262,
+      "learning_rate": 7.545458869773521e-07,
+      "loss": 0.4793,
+      "step": 379
+    },
+    {
+      "epoch": 0.07266469069700736,
+      "grad_norm": 3.3019675024918413,
+      "learning_rate": 7.548807505114127e-07,
+      "loss": 0.3589,
+      "step": 380
+    },
+    {
+      "epoch": 0.07285591356726265,
+      "grad_norm": 5.612420830454983,
+      "learning_rate": 7.552147339826084e-07,
+      "loss": 0.4216,
+      "step": 381
+    },
+    {
+      "epoch": 0.07304713643751792,
+      "grad_norm": 1.8776598184479762,
+      "learning_rate": 7.555478420046451e-07,
+      "loss": 0.4278,
+      "step": 382
+    },
+    {
+      "epoch": 0.07323835930777321,
+      "grad_norm": 4.6938907038378295,
+      "learning_rate": 7.558800791550426e-07,
+      "loss": 0.4983,
+      "step": 383
+    },
+    {
+      "epoch": 0.07342958217802849,
+      "grad_norm": 3.740575901305211,
+      "learning_rate": 7.562114499755122e-07,
+      "loss": 0.3672,
+      "step": 384
+    },
+    {
+      "epoch": 0.07362080504828378,
+      "grad_norm": 3.6443226908941817,
+      "learning_rate": 7.565419589723287e-07,
+      "loss": 0.2288,
+      "step": 385
+    },
+    {
+      "epoch": 0.07381202791853905,
+      "grad_norm": 3.513123703287724,
+      "learning_rate": 7.568716106166993e-07,
+      "loss": 0.1297,
+      "step": 386
+    },
+    {
+      "epoch": 0.07400325078879434,
+      "grad_norm": 2.9090987159342485,
+      "learning_rate": 7.572004093451251e-07,
+      "loss": 0.1442,
+      "step": 387
+    },
+    {
+      "epoch": 0.07419447365904962,
+      "grad_norm": 7.47474002449837,
+      "learning_rate": 7.575283595597604e-07,
+      "loss": 0.2136,
+      "step": 388
+    },
+    {
+      "epoch": 0.0743856965293049,
+      "grad_norm": 3.5905933614454115,
+      "learning_rate": 7.578554656287656e-07,
+      "loss": 0.3118,
+      "step": 389
+    },
+    {
+      "epoch": 0.07457691939956018,
+      "grad_norm": 5.411915010550415,
+      "learning_rate": 7.581817318866572e-07,
+      "loss": 0.4124,
+      "step": 390
+    },
+    {
+      "epoch": 0.07476814226981547,
+      "grad_norm": 4.897597926953739,
+      "learning_rate": 7.58507162634651e-07,
+      "loss": 0.8647,
+      "step": 391
+    },
+    {
+      "epoch": 0.07495936514007075,
+      "grad_norm": 2.647200844264652,
+      "learning_rate": 7.588317621410044e-07,
+      "loss": 0.3633,
+      "step": 392
+    },
+    {
+      "epoch": 0.07515058801032604,
+      "grad_norm": 5.611345630356875,
+      "learning_rate": 7.591555346413503e-07,
+      "loss": 0.4683,
+      "step": 393
+    },
+    {
+      "epoch": 0.07534181088058131,
+      "grad_norm": 2.0200460283304156,
+      "learning_rate": 7.594784843390296e-07,
+      "loss": 0.1971,
+      "step": 394
+    },
+    {
+      "epoch": 0.0755330337508366,
+      "grad_norm": 3.792864027060685,
+      "learning_rate": 7.598006154054186e-07,
+      "loss": 0.2007,
+      "step": 395
+    },
+    {
+      "epoch": 0.07572425662109188,
+      "grad_norm": 4.363457849492404,
+      "learning_rate": 7.601219319802527e-07,
+      "loss": 0.3879,
+      "step": 396
+    },
+    {
+      "epoch": 0.07591547949134717,
+      "grad_norm": 2.8346293981882784,
+      "learning_rate": 7.604424381719444e-07,
+      "loss": 0.238,
+      "step": 397
+    },
+    {
+      "epoch": 0.07610670236160244,
+      "grad_norm": 4.116161935317753,
+      "learning_rate": 7.607621380578996e-07,
+      "loss": 0.3237,
+      "step": 398
+    },
+    {
+      "epoch": 0.07629792523185773,
+      "grad_norm": 6.426375174394978,
+      "learning_rate": 7.610810356848289e-07,
+      "loss": 0.2809,
+      "step": 399
+    },
+    {
+      "epoch": 0.07648914810211302,
+      "grad_norm": 8.37363616314761,
+      "learning_rate": 7.613991350690538e-07,
+      "loss": 0.289,
+      "step": 400
+    },
+    {
+      "epoch": 0.0766803709723683,
+      "grad_norm": 4.485865116100099,
+      "learning_rate": 7.617164401968115e-07,
+      "loss": 0.3947,
+      "step": 401
+    },
+    {
+      "epoch": 0.07687159384262358,
+      "grad_norm": 6.8298348876298,
+      "learning_rate": 7.620329550245536e-07,
+      "loss": 0.2965,
+      "step": 402
+    },
+    {
+      "epoch": 0.07706281671287886,
+      "grad_norm": 5.255161654438637,
+      "learning_rate": 7.623486834792428e-07,
+      "loss": 0.4888,
+      "step": 403
+    },
+    {
+      "epoch": 0.07725403958313415,
+      "grad_norm": 4.643500570629714,
+      "learning_rate": 7.626636294586452e-07,
+      "loss": 0.3973,
+      "step": 404
+    },
+    {
+      "epoch": 0.07744526245338942,
+      "grad_norm": 5.372019629004456,
+      "learning_rate": 7.629777968316182e-07,
+      "loss": 0.3156,
+      "step": 405
+    },
+    {
+      "epoch": 0.07763648532364471,
+      "grad_norm": 4.261301161799379,
+      "learning_rate": 7.63291189438397e-07,
+      "loss": 0.3789,
+      "step": 406
+    },
+    {
+      "epoch": 0.07782770819389999,
+      "grad_norm": 6.9893699041111095,
+      "learning_rate": 7.636038110908749e-07,
+      "loss": 0.5212,
+      "step": 407
+    },
+    {
+      "epoch": 0.07801893106415528,
+      "grad_norm": 2.215146536877434,
+      "learning_rate": 7.639156655728827e-07,
+      "loss": 0.3076,
+      "step": 408
+    },
+    {
+      "epoch": 0.07821015393441055,
+      "grad_norm": 3.5562869319774677,
+      "learning_rate": 7.642267566404624e-07,
+      "loss": 0.1734,
+      "step": 409
+    },
+    {
+      "epoch": 0.07840137680466584,
+      "grad_norm": 4.548905486206163,
+      "learning_rate": 7.645370880221393e-07,
+      "loss": 0.5211,
+      "step": 410
+    },
+    {
+      "epoch": 0.07859259967492112,
+      "grad_norm": 4.778489855750839,
+      "learning_rate": 7.648466634191901e-07,
+      "loss": 0.3694,
+      "step": 411
+    },
+    {
+      "epoch": 0.07878382254517641,
+      "grad_norm": 5.394996529083974,
+      "learning_rate": 7.651554865059079e-07,
+      "loss": 0.24,
+      "step": 412
+    },
+    {
+      "epoch": 0.07897504541543168,
+      "grad_norm": 5.755540914692825,
+      "learning_rate": 7.654635609298631e-07,
+      "loss": 0.3064,
+      "step": 413
+    },
+    {
+      "epoch": 0.07916626828568697,
+      "grad_norm": 7.651020615256381,
+      "learning_rate": 7.657708903121624e-07,
+      "loss": 0.4465,
+      "step": 414
+    },
+    {
+      "epoch": 0.07935749115594225,
+      "grad_norm": 3.820500219058772,
+      "learning_rate": 7.660774782477047e-07,
+      "loss": 0.3692,
+      "step": 415
+    },
+    {
+      "epoch": 0.07954871402619754,
+      "grad_norm": 3.412090089652855,
+      "learning_rate": 7.663833283054328e-07,
+      "loss": 0.4175,
+      "step": 416
+    },
+    {
+      "epoch": 0.07973993689645281,
+      "grad_norm": 2.9811932444418234,
+      "learning_rate": 7.666884440285824e-07,
+      "loss": 0.4025,
+      "step": 417
+    },
+    {
+      "epoch": 0.0799311597667081,
+      "grad_norm": 4.141846904760086,
+      "learning_rate": 7.669928289349289e-07,
+      "loss": 0.4129,
+      "step": 418
+    },
+    {
+      "epoch": 0.08012238263696338,
+      "grad_norm": 5.406463971616901,
+      "learning_rate": 7.672964865170308e-07,
+      "loss": 0.6731,
+      "step": 419
+    },
+    {
+      "epoch": 0.08031360550721867,
+      "grad_norm": 2.4838135282765315,
+      "learning_rate": 7.6759942024247e-07,
+      "loss": 0.3217,
+      "step": 420
+    },
+    {
+      "epoch": 0.08050482837747394,
+      "grad_norm": 3.90403220733271,
+      "learning_rate": 7.679016335540892e-07,
+      "loss": 0.2467,
+      "step": 421
+    },
+    {
+      "epoch": 0.08069605124772923,
+      "grad_norm": 3.6730492844695917,
+      "learning_rate": 7.68203129870227e-07,
+      "loss": 0.2495,
+      "step": 422
+    },
+    {
+      "epoch": 0.0808872741179845,
+      "grad_norm": 3.235810039815306,
+      "learning_rate": 7.685039125849492e-07,
+      "loss": 0.1483,
+      "step": 423
+    },
+    {
+      "epoch": 0.0810784969882398,
+      "grad_norm": 3.828974789089873,
+      "learning_rate": 7.688039850682795e-07,
+      "loss": 0.1435,
+      "step": 424
+    },
+    {
+      "epoch": 0.08126971985849507,
+      "grad_norm": 6.232138478749533,
+      "learning_rate": 7.691033506664244e-07,
+      "loss": 0.1926,
+      "step": 425
+    },
+    {
+      "epoch": 0.08146094272875036,
+      "grad_norm": 9.923215952691447,
+      "learning_rate": 7.694020127019981e-07,
+      "loss": 0.3646,
+      "step": 426
+    },
+    {
+      "epoch": 0.08165216559900564,
+      "grad_norm": 5.604376731064853,
+      "learning_rate": 7.696999744742434e-07,
+      "loss": 0.548,
+      "step": 427
+    },
+    {
+      "epoch": 0.08184338846926092,
+      "grad_norm": 4.192573548355953,
+      "learning_rate": 7.699972392592504e-07,
+      "loss": 0.3186,
+      "step": 428
+    },
+    {
+      "epoch": 0.0820346113395162,
+      "grad_norm": 3.0578997568849013,
+      "learning_rate": 7.702938103101732e-07,
+      "loss": 0.2445,
+      "step": 429
+    },
+    {
+      "epoch": 0.08222583420977149,
+      "grad_norm": 4.10699997647464,
+      "learning_rate": 7.705896908574425e-07,
+      "loss": 0.2969,
+      "step": 430
+    },
+    {
+      "epoch": 0.08241705708002676,
+      "grad_norm": 4.125889075440807,
+      "learning_rate": 7.708848841089775e-07,
+      "loss": 0.3968,
+      "step": 431
+    },
+    {
+      "epoch": 0.08260827995028205,
+      "grad_norm": 2.90822148877681,
+      "learning_rate": 7.711793932503941e-07,
+      "loss": 0.2111,
+      "step": 432
+    },
+    {
+      "epoch": 0.08279950282053733,
+      "grad_norm": 3.9054785615202947,
+      "learning_rate": 7.714732214452113e-07,
+      "loss": 0.4128,
+      "step": 433
+    },
+    {
+      "epoch": 0.08299072569079262,
+      "grad_norm": 3.17068353834655,
+      "learning_rate": 7.717663718350558e-07,
+      "loss": 0.2045,
+      "step": 434
+    },
+    {
+      "epoch": 0.08318194856104791,
+      "grad_norm": 3.542050479212576,
+      "learning_rate": 7.720588475398626e-07,
+      "loss": 0.1352,
+      "step": 435
+    },
+    {
+      "epoch": 0.08337317143130318,
+      "grad_norm": 2.8415807477951835,
+      "learning_rate": 7.723506516580743e-07,
+      "loss": 0.1545,
+      "step": 436
+    },
+    {
+      "epoch": 0.08356439430155847,
+      "grad_norm": 5.441360506302252,
+      "learning_rate": 7.726417872668387e-07,
+      "loss": 0.1694,
+      "step": 437
+    },
+    {
+      "epoch": 0.08375561717181375,
+      "grad_norm": 6.306513601035744,
+      "learning_rate": 7.729322574222029e-07,
+      "loss": 0.1663,
+      "step": 438
+    },
+    {
+      "epoch": 0.08394684004206904,
+      "grad_norm": 5.936832430604896,
+      "learning_rate": 7.732220651593062e-07,
+      "loss": 0.8375,
+      "step": 439
+    },
+    {
+      "epoch": 0.08413806291232431,
+      "grad_norm": 6.7207999007671955,
+      "learning_rate": 7.7351121349257e-07,
+      "loss": 0.3487,
+      "step": 440
+    },
+    {
+      "epoch": 0.0843292857825796,
+      "grad_norm": 3.1762029876892863,
+      "learning_rate": 7.737997054158864e-07,
+      "loss": 0.3224,
+      "step": 441
+    },
+    {
+      "epoch": 0.08452050865283488,
+      "grad_norm": 4.679960978418675,
+      "learning_rate": 7.740875439028032e-07,
+      "loss": 0.1545,
+      "step": 442
+    },
+    {
+      "epoch": 0.08471173152309017,
+      "grad_norm": 3.955703075625086,
+      "learning_rate": 7.743747319067098e-07,
+      "loss": 0.3782,
+      "step": 443
+    },
+    {
+      "epoch": 0.08490295439334544,
+      "grad_norm": 4.093599505970657,
+      "learning_rate": 7.746612723610163e-07,
+      "loss": 0.1855,
+      "step": 444
+    },
+    {
+      "epoch": 0.08509417726360073,
+      "grad_norm": 5.84261306881808,
+      "learning_rate": 7.749471681793357e-07,
+      "loss": 0.7162,
+      "step": 445
+    },
+    {
+      "epoch": 0.085285400133856,
+      "grad_norm": 3.7729355377506097,
+      "learning_rate": 7.752324222556604e-07,
+      "loss": 0.2404,
+      "step": 446
+    },
+    {
+      "epoch": 0.0854766230041113,
+      "grad_norm": 3.730637727275707,
+      "learning_rate": 7.755170374645381e-07,
+      "loss": 0.3878,
+      "step": 447
+    },
+    {
+      "epoch": 0.08566784587436657,
+      "grad_norm": 3.012812592682779,
+      "learning_rate": 7.758010166612457e-07,
+      "loss": 0.0996,
+      "step": 448
+    },
+    {
+      "epoch": 0.08585906874462186,
+      "grad_norm": 3.258966986920607,
+      "learning_rate": 7.76084362681962e-07,
+      "loss": 0.0732,
+      "step": 449
+    },
+    {
+      "epoch": 0.08605029161487714,
+      "grad_norm": 11.974850048806717,
+      "learning_rate": 7.763670783439357e-07,
+      "loss": 0.5527,
+      "step": 450
+    },
+    {
+      "epoch": 0.08624151448513243,
+      "grad_norm": 8.637861309968685,
+      "learning_rate": 7.766491664456554e-07,
+      "loss": 0.4977,
+      "step": 451
+    },
+    {
+      "epoch": 0.0864327373553877,
+      "grad_norm": 7.9023905746573675,
+      "learning_rate": 7.769306297670152e-07,
+      "loss": 0.8213,
+      "step": 452
+    },
+    {
+      "epoch": 0.08662396022564299,
+      "grad_norm": 4.732723741275257,
+      "learning_rate": 7.772114710694782e-07,
+      "loss": 0.4748,
+      "step": 453
+    },
+    {
+      "epoch": 0.08681518309589827,
+      "grad_norm": 3.8365967722032592,
+      "learning_rate": 7.774916930962405e-07,
+      "loss": 0.5766,
+      "step": 454
+    },
+    {
+      "epoch": 0.08700640596615355,
+      "grad_norm": 4.13945203757535,
+      "learning_rate": 7.777712985723907e-07,
+      "loss": 0.4281,
+      "step": 455
+    },
+    {
+      "epoch": 0.08719762883640883,
+      "grad_norm": 2.8477448291371616,
+      "learning_rate": 7.780502902050703e-07,
+      "loss": 0.2963,
+      "step": 456
+    },
+    {
+      "epoch": 0.08738885170666412,
+      "grad_norm": 2.683800479749976,
+      "learning_rate": 7.783286706836291e-07,
+      "loss": 0.2942,
+      "step": 457
+    },
+    {
+      "epoch": 0.0875800745769194,
+      "grad_norm": 4.60684310427989,
+      "learning_rate": 7.786064426797826e-07,
+      "loss": 0.2257,
+      "step": 458
+    },
+    {
+      "epoch": 0.08777129744717468,
+      "grad_norm": 5.347651434952635,
+      "learning_rate": 7.788836088477645e-07,
+      "loss": 0.3008,
+      "step": 459
+    },
+    {
+      "epoch": 0.08796252031742996,
+      "grad_norm": 4.295276025289044,
+      "learning_rate": 7.791601718244799e-07,
+      "loss": 0.297,
+      "step": 460
+    },
+    {
+      "epoch": 0.08815374318768525,
+      "grad_norm": 4.736638802632775,
+      "learning_rate": 7.794361342296549e-07,
+      "loss": 0.2527,
+      "step": 461
+    },
+    {
+      "epoch": 0.08834496605794052,
+      "grad_norm": 5.470716547924001,
+      "learning_rate": 7.797114986659862e-07,
+      "loss": 0.1585,
+      "step": 462
+    },
+    {
+      "epoch": 0.08853618892819581,
+      "grad_norm": 5.682493584457649,
+      "learning_rate": 7.799862677192882e-07,
+      "loss": 0.3893,
+      "step": 463
+    },
+    {
+      "epoch": 0.08872741179845109,
+      "grad_norm": 5.754359707182758,
+      "learning_rate": 7.802604439586384e-07,
+      "loss": 0.6574,
+      "step": 464
+    },
+    {
+      "epoch": 0.08891863466870638,
+      "grad_norm": 6.776328274067726,
+      "learning_rate": 7.805340299365214e-07,
+      "loss": 0.756,
+      "step": 465
+    },
+    {
+      "epoch": 0.08910985753896165,
+      "grad_norm": 3.5061668834932522,
+      "learning_rate": 7.808070281889727e-07,
+      "loss": 0.4458,
+      "step": 466
+    },
+    {
+      "epoch": 0.08930108040921694,
+      "grad_norm": 4.042529980719936,
+      "learning_rate": 7.810794412357182e-07,
+      "loss": 0.3944,
+      "step": 467
+    },
+    {
+      "epoch": 0.08949230327947222,
+      "grad_norm": 6.766379933162366,
+      "learning_rate": 7.813512715803147e-07,
+      "loss": 1.1971,
+      "step": 468
+    },
+    {
+      "epoch": 0.08968352614972751,
+      "grad_norm": 3.047827156162293,
+      "learning_rate": 7.816225217102873e-07,
+      "loss": 0.1773,
+      "step": 469
+    },
+    {
+      "epoch": 0.08987474901998278,
+      "grad_norm": 2.8459668583820137,
+      "learning_rate": 7.818931940972669e-07,
+      "loss": 0.2153,
+      "step": 470
+    },
+    {
+      "epoch": 0.09006597189023807,
+      "grad_norm": 3.7900637218181297,
+      "learning_rate": 7.821632911971241e-07,
+      "loss": 0.389,
+      "step": 471
+    },
+    {
+      "epoch": 0.09025719476049336,
+      "grad_norm": 4.200216542065913,
+      "learning_rate": 7.824328154501044e-07,
+      "loss": 0.1439,
+      "step": 472
+    },
+    {
+      "epoch": 0.09044841763074864,
+      "grad_norm": 5.718902752613272,
+      "learning_rate": 7.827017692809588e-07,
+      "loss": 0.3899,
+      "step": 473
+    },
+    {
+      "epoch": 0.09063964050100393,
+      "grad_norm": 6.503441413099917,
+      "learning_rate": 7.829701550990763e-07,
+      "loss": 0.1259,
+      "step": 474
+    },
+    {
+      "epoch": 0.0908308633712592,
+      "grad_norm": 7.343396766770726,
+      "learning_rate": 7.83237975298612e-07,
+      "loss": 0.2091,
+      "step": 475
+    },
+    {
+      "epoch": 0.09102208624151449,
+      "grad_norm": 6.216227724398619,
+      "learning_rate": 7.835052322586164e-07,
+      "loss": 0.2625,
+      "step": 476
+    },
+    {
+      "epoch": 0.09121330911176977,
+      "grad_norm": 4.7431309875378105,
+      "learning_rate": 7.837719283431615e-07,
+      "loss": 0.3367,
+      "step": 477
+    },
+    {
+      "epoch": 0.09140453198202506,
+      "grad_norm": 3.168447813413641,
+      "learning_rate": 7.840380659014663e-07,
+      "loss": 0.5213,
+      "step": 478
+    },
+    {
+      "epoch": 0.09159575485228033,
+      "grad_norm": 3.320466161425953,
+      "learning_rate": 7.843036472680214e-07,
+      "loss": 0.382,
+      "step": 479
+    },
+    {
+      "epoch": 0.09178697772253562,
+      "grad_norm": 3.381283033283649,
+      "learning_rate": 7.845686747627114e-07,
+      "loss": 0.3496,
+      "step": 480
+    },
+    {
+      "epoch": 0.0919782005927909,
+      "grad_norm": 3.9570728235800416,
+      "learning_rate": 7.84833150690937e-07,
+      "loss": 0.333,
+      "step": 481
+    },
+    {
+      "epoch": 0.09216942346304619,
+      "grad_norm": 3.483990019327402,
+      "learning_rate": 7.850970773437346e-07,
+      "loss": 0.1467,
+      "step": 482
+    },
+    {
+      "epoch": 0.09236064633330146,
+      "grad_norm": 1.9306400434575404,
+      "learning_rate": 7.853604569978961e-07,
+      "loss": 0.4316,
+      "step": 483
+    },
+    {
+      "epoch": 0.09255186920355675,
+      "grad_norm": 3.4210064626196126,
+      "learning_rate": 7.856232919160863e-07,
+      "loss": 0.3498,
+      "step": 484
+    },
+    {
+      "epoch": 0.09274309207381203,
+      "grad_norm": 4.940081730216688,
+      "learning_rate": 7.858855843469596e-07,
+      "loss": 0.1567,
+      "step": 485
+    },
+    {
+      "epoch": 0.09293431494406731,
+      "grad_norm": 6.02297516667622,
+      "learning_rate": 7.861473365252758e-07,
+      "loss": 0.2491,
+      "step": 486
+    },
+    {
+      "epoch": 0.09312553781432259,
+      "grad_norm": 4.319103106183702,
+      "learning_rate": 7.864085506720136e-07,
+      "loss": 0.1863,
+      "step": 487
+    },
+    {
+      "epoch": 0.09331676068457788,
+      "grad_norm": 13.212575205261368,
+      "learning_rate": 7.866692289944848e-07,
+      "loss": 0.5057,
+      "step": 488
+    },
+    {
+      "epoch": 0.09350798355483315,
+      "grad_norm": 5.751421545364282,
+      "learning_rate": 7.869293736864445e-07,
+      "loss": 0.8092,
+      "step": 489
+    },
+    {
+      "epoch": 0.09369920642508844,
+      "grad_norm": 4.868983542015918,
+      "learning_rate": 7.871889869282038e-07,
+      "loss": 0.4129,
+      "step": 490
+    },
+    {
+      "epoch": 0.09389042929534372,
+      "grad_norm": 6.097820155148169,
+      "learning_rate": 7.874480708867382e-07,
+      "loss": 0.541,
+      "step": 491
+    },
+    {
+      "epoch": 0.09408165216559901,
+      "grad_norm": 2.295208904091658,
+      "learning_rate": 7.877066277157967e-07,
+      "loss": 0.227,
+      "step": 492
+    },
+    {
+      "epoch": 0.09427287503585428,
+      "grad_norm": 5.018284363006749,
+      "learning_rate": 7.879646595560088e-07,
+      "loss": 0.742,
+      "step": 493
+    },
+    {
+      "epoch": 0.09446409790610957,
+      "grad_norm": 2.157413334702678,
+      "learning_rate": 7.882221685349911e-07,
+      "loss": 0.3535,
+      "step": 494
+    },
+    {
+      "epoch": 0.09465532077636485,
+      "grad_norm": 4.82185366335961,
+      "learning_rate": 7.884791567674519e-07,
+      "loss": 0.3119,
+      "step": 495
+    },
+    {
+      "epoch": 0.09484654364662014,
+      "grad_norm": 4.222036932064059,
+      "learning_rate": 7.887356263552972e-07,
+      "loss": 0.1369,
+      "step": 496
+    },
+    {
+      "epoch": 0.09503776651687541,
+      "grad_norm": 4.611837472862485,
+      "learning_rate": 7.889915793877318e-07,
+      "loss": 0.1877,
+      "step": 497
+    },
+    {
+      "epoch": 0.0952289893871307,
+      "grad_norm": 2.5078574678661623,
+      "learning_rate": 7.892470179413624e-07,
+      "loss": 0.0832,
+      "step": 498
+    },
+    {
+      "epoch": 0.09542021225738598,
+      "grad_norm": 4.245970891808717,
+      "learning_rate": 7.895019440802987e-07,
+      "loss": 0.1062,
+      "step": 499
+    },
+    {
+      "epoch": 0.09561143512764127,
+      "grad_norm": 3.843643931344249,
+      "learning_rate": 7.897563598562531e-07,
+      "loss": 0.2302,
+      "step": 500
+    },
+    {
+      "epoch": 0.09561143512764127,
+      "eval_runtime": 755.1977,
+      "eval_samples_per_second": 2.031,
+      "eval_steps_per_second": 0.508,
+      "step": 500
+    },
+    {
+      "epoch": 0.09580265799789654,
+      "grad_norm": 5.653740594937457,
+      "learning_rate": 7.900102673086404e-07,
+      "loss": 0.4629,
+      "step": 501
+    },
+    {
+      "epoch": 0.09599388086815183,
+      "grad_norm": 7.824256778801741,
+      "learning_rate": 7.902636684646751e-07,
+      "loss": 0.4342,
+      "step": 502
+    },
+    {
+      "epoch": 0.09618510373840711,
+      "grad_norm": 5.118038194138945,
+      "learning_rate": 7.905165653394691e-07,
+      "loss": 0.356,
+      "step": 503
+    },
+    {
+      "epoch": 0.0963763266086624,
+      "grad_norm": 4.880970746403434,
+      "learning_rate": 7.907689599361276e-07,
+      "loss": 0.5353,
+      "step": 504
+    },
+    {
+      "epoch": 0.09656754947891767,
+      "grad_norm": 3.593124866017797,
+      "learning_rate": 7.910208542458445e-07,
+      "loss": 0.2515,
+      "step": 505
+    },
+    {
+      "epoch": 0.09675877234917296,
+      "grad_norm": 4.436342760085264,
+      "learning_rate": 7.912722502479962e-07,
+      "loss": 0.37,
+      "step": 506
+    },
+    {
+      "epoch": 0.09694999521942825,
+      "grad_norm": 3.3071574186377024,
+      "learning_rate": 7.915231499102353e-07,
+      "loss": 0.2209,
+      "step": 507
+    },
+    {
+      "epoch": 0.09714121808968353,
+      "grad_norm": 3.178284731651527,
+      "learning_rate": 7.917735551885835e-07,
+      "loss": 0.2107,
+      "step": 508
+    },
+    {
+      "epoch": 0.09733244095993882,
+      "grad_norm": 3.273829033863673,
+      "learning_rate": 7.920234680275216e-07,
+      "loss": 0.1064,
+      "step": 509
+    },
+    {
+      "epoch": 0.09752366383019409,
+      "grad_norm": 2.065659155397341,
+      "learning_rate": 7.922728903600819e-07,
+      "loss": 0.0827,
+      "step": 510
+    },
+    {
+      "epoch": 0.09771488670044938,
+      "grad_norm": 3.209562058090843,
+      "learning_rate": 7.925218241079367e-07,
+      "loss": 0.2181,
+      "step": 511
+    },
+    {
+      "epoch": 0.09790610957070466,
+      "grad_norm": 4.3142002458145505,
+      "learning_rate": 7.927702711814871e-07,
+      "loss": 0.1097,
+      "step": 512
+    },
+    {
+      "epoch": 0.09809733244095994,
+      "grad_norm": 10.230165597240969,
+      "learning_rate": 7.93018233479952e-07,
+      "loss": 0.7163,
+      "step": 513
+    },
+    {
+      "epoch": 0.09828855531121522,
+      "grad_norm": 4.216071741004056,
+      "learning_rate": 7.932657128914543e-07,
+      "loss": 0.2871,
+      "step": 514
+    },
+    {
+      "epoch": 0.09847977818147051,
+      "grad_norm": 4.562408864404427,
+      "learning_rate": 7.935127112931072e-07,
+      "loss": 0.5078,
+      "step": 515
+    },
+    {
+      "epoch": 0.09867100105172578,
+      "grad_norm": 5.543574509106259,
+      "learning_rate": 7.937592305511001e-07,
+      "loss": 0.249,
+      "step": 516
+    },
+    {
+      "epoch": 0.09886222392198107,
+      "grad_norm": 4.239063552873188,
+      "learning_rate": 7.94005272520783e-07,
+      "loss": 0.4272,
+      "step": 517
+    },
+    {
+      "epoch": 0.09905344679223635,
+      "grad_norm": 4.3337931511301075,
+      "learning_rate": 7.942508390467499e-07,
+      "loss": 0.4786,
+      "step": 518
+    },
+    {
+      "epoch": 0.09924466966249164,
+      "grad_norm": 4.0484585407023905,
+      "learning_rate": 7.944959319629227e-07,
+      "loss": 0.4542,
+      "step": 519
+    },
+    {
+      "epoch": 0.09943589253274691,
+      "grad_norm": 4.03155374491579,
+      "learning_rate": 7.947405530926321e-07,
+      "loss": 0.3287,
+      "step": 520
+    },
+    {
+      "epoch": 0.0996271154030022,
+      "grad_norm": 4.7388300817631395,
+      "learning_rate": 7.949847042487001e-07,
+      "loss": 0.326,
+      "step": 521
+    },
+    {
+      "epoch": 0.09981833827325748,
+      "grad_norm": 3.086781987397187,
+      "learning_rate": 7.952283872335202e-07,
+      "loss": 0.2463,
+      "step": 522
+    },
+    {
+      "epoch": 0.10000956114351277,
+      "grad_norm": 5.739099536405474,
+      "learning_rate": 7.95471603839137e-07,
+      "loss": 0.586,
+      "step": 523
+    },
+    {
+      "epoch": 0.10020078401376804,
+      "grad_norm": 2.63463721729466,
+      "learning_rate": 7.957143558473252e-07,
+      "loss": 0.0849,
+      "step": 524
+    },
+    {
+      "epoch": 0.10039200688402333,
+      "grad_norm": 5.0364686426046035,
+      "learning_rate": 7.959566450296694e-07,
+      "loss": 0.2298,
+      "step": 525
+    },
+    {
+      "epoch": 0.10058322975427861,
+      "grad_norm": 7.013321054300406,
+      "learning_rate": 7.961984731476392e-07,
+      "loss": 0.5049,
+      "step": 526
+    },
+    {
+      "epoch": 0.1007744526245339,
+      "grad_norm": 5.730148878576742,
+      "learning_rate": 7.964398419526678e-07,
+      "loss": 0.6402,
+      "step": 527
+    },
+    {
+      "epoch": 0.10096567549478917,
+      "grad_norm": 3.0873170497297093,
+      "learning_rate": 7.966807531862276e-07,
+      "loss": 0.1561,
+      "step": 528
+    },
+    {
+      "epoch": 0.10115689836504446,
+      "grad_norm": 3.462903155601624,
+      "learning_rate": 7.969212085799059e-07,
+      "loss": 0.3022,
+      "step": 529
+    },
+    {
+      "epoch": 0.10134812123529974,
+      "grad_norm": 6.396985309583366,
+      "learning_rate": 7.971612098554789e-07,
+      "loss": 0.6695,
+      "step": 530
+    },
+    {
+      "epoch": 0.10153934410555503,
+      "grad_norm": 3.048224673523643,
+      "learning_rate": 7.974007587249862e-07,
+      "loss": 0.2126,
+      "step": 531
+    },
+    {
+      "epoch": 0.1017305669758103,
+      "grad_norm": 3.697954519731805,
+      "learning_rate": 7.97639856890804e-07,
+      "loss": 0.1377,
+      "step": 532
+    },
+    {
+      "epoch": 0.10192178984606559,
+      "grad_norm": 3.808149388139879,
+      "learning_rate": 7.978785060457175e-07,
+      "loss": 0.2116,
+      "step": 533
+    },
+    {
+      "epoch": 0.10211301271632087,
+      "grad_norm": 4.71166480477378,
+      "learning_rate": 7.981167078729933e-07,
+      "loss": 0.2151,
+      "step": 534
+    },
+    {
+      "epoch": 0.10230423558657616,
+      "grad_norm": 2.8999429170975617,
+      "learning_rate": 7.983544640464497e-07,
+      "loss": 0.1502,
+      "step": 535
+    },
+    {
+      "epoch": 0.10249545845683143,
+      "grad_norm": 2.6596672964331582,
+      "learning_rate": 7.985917762305287e-07,
+      "loss": 0.128,
+      "step": 536
+    },
+    {
+      "epoch": 0.10268668132708672,
+      "grad_norm": 6.270691761264854,
+      "learning_rate": 7.98828646080364e-07,
+      "loss": 0.2027,
+      "step": 537
+    },
+    {
+      "epoch": 0.102877904197342,
+      "grad_norm": 8.532771048299335,
+      "learning_rate": 7.990650752418521e-07,
+      "loss": 0.4717,
+      "step": 538
+    },
+    {
+      "epoch": 0.10306912706759729,
+      "grad_norm": 4.467029587155152,
+      "learning_rate": 7.9930106535172e-07,
+      "loss": 0.769,
+      "step": 539
+    },
+    {
+      "epoch": 0.10326034993785256,
+      "grad_norm": 4.806238435847901,
+      "learning_rate": 7.995366180375933e-07,
+      "loss": 0.3743,
+      "step": 540
+    },
+    {
+      "epoch": 0.10345157280810785,
+      "grad_norm": 4.851871499066832,
+      "learning_rate": 7.997717349180633e-07,
+      "loss": 0.6985,
+      "step": 541
+    },
+    {
+      "epoch": 0.10364279567836313,
+      "grad_norm": 2.7156067139741884,
+      "learning_rate": 8.000064176027547e-07,
+      "loss": 0.4957,
+      "step": 542
+    },
+    {
+      "epoch": 0.10383401854861841,
+      "grad_norm": 5.119633539913556,
+      "learning_rate": 8.002406676923903e-07,
+      "loss": 0.4362,
+      "step": 543
+    },
+    {
+      "epoch": 0.1040252414188737,
+      "grad_norm": 5.505636101749524,
+      "learning_rate": 8.004744867788577e-07,
+      "loss": 0.3416,
+      "step": 544
+    },
+    {
+      "epoch": 0.10421646428912898,
+      "grad_norm": 4.969243630894303,
+      "learning_rate": 8.007078764452736e-07,
+      "loss": 0.3852,
+      "step": 545
+    },
+    {
+      "epoch": 0.10440768715938427,
+      "grad_norm": 4.0403295186183215,
+      "learning_rate": 8.009408382660484e-07,
+      "loss": 0.2373,
+      "step": 546
+    },
+    {
+      "epoch": 0.10459891002963954,
+      "grad_norm": 3.8552208910640235,
+      "learning_rate": 8.011733738069495e-07,
+      "loss": 0.1986,
+      "step": 547
+    },
+    {
+      "epoch": 0.10479013289989483,
+      "grad_norm": 4.031327594335589,
+      "learning_rate": 8.014054846251653e-07,
+      "loss": 0.146,
+      "step": 548
+    },
+    {
+      "epoch": 0.10498135577015011,
+      "grad_norm": 3.4131105261095693,
+      "learning_rate": 8.016371722693665e-07,
+      "loss": 0.1681,
+      "step": 549
+    },
+    {
+      "epoch": 0.1051725786404054,
+      "grad_norm": 4.485377608343439,
+      "learning_rate": 8.018684382797693e-07,
+      "loss": 0.161,
+      "step": 550
+    },
+    {
+      "epoch": 0.10536380151066067,
+      "grad_norm": 6.835084768242637,
+      "learning_rate": 8.020992841881965e-07,
+      "loss": 0.6216,
+      "step": 551
+    },
+    {
+      "epoch": 0.10555502438091596,
+      "grad_norm": 4.740469154607604,
+      "learning_rate": 8.023297115181375e-07,
+      "loss": 0.4184,
+      "step": 552
+    },
+    {
+      "epoch": 0.10574624725117124,
+      "grad_norm": 3.0695139693436166,
+      "learning_rate": 8.025597217848099e-07,
+      "loss": 0.4095,
+      "step": 553
+    },
+    {
+      "epoch": 0.10593747012142653,
+      "grad_norm": 3.8218960902582695,
+      "learning_rate": 8.027893164952188e-07,
+      "loss": 0.3347,
+      "step": 554
+    },
+    {
+      "epoch": 0.1061286929916818,
+      "grad_norm": 3.6051653468800464,
+      "learning_rate": 8.030184971482156e-07,
+      "loss": 0.3035,
+      "step": 555
+    },
+    {
+      "epoch": 0.10631991586193709,
+      "grad_norm": 4.154649239801855,
+      "learning_rate": 8.032472652345573e-07,
+      "loss": 0.2762,
+      "step": 556
+    },
+    {
+      "epoch": 0.10651113873219237,
+      "grad_norm": 3.274505368218861,
+      "learning_rate": 8.034756222369647e-07,
+      "loss": 0.3015,
+      "step": 557
+    },
+    {
+      "epoch": 0.10670236160244766,
+      "grad_norm": 4.38925247188663,
+      "learning_rate": 8.03703569630179e-07,
+      "loss": 0.3877,
+      "step": 558
+    },
+    {
+      "epoch": 0.10689358447270293,
+      "grad_norm": 2.794659637581865,
+      "learning_rate": 8.039311088810208e-07,
+      "loss": 0.2079,
+      "step": 559
+    },
+    {
+      "epoch": 0.10708480734295822,
+      "grad_norm": 2.5601330502029014,
+      "learning_rate": 8.041582414484452e-07,
+      "loss": 0.1072,
+      "step": 560
+    },
+    {
+      "epoch": 0.1072760302132135,
+      "grad_norm": 3.5993938412564828,
+      "learning_rate": 8.043849687835982e-07,
+      "loss": 0.14,
+      "step": 561
+    },
+    {
+      "epoch": 0.10746725308346879,
+      "grad_norm": 3.6070758052566854,
+      "learning_rate": 8.04611292329873e-07,
+      "loss": 0.1136,
+      "step": 562
+    },
+    {
+      "epoch": 0.10765847595372406,
+      "grad_norm": 5.9881777322062035,
+      "learning_rate": 8.048372135229646e-07,
+      "loss": 0.3441,
+      "step": 563
+    },
+    {
+      "epoch": 0.10784969882397935,
+      "grad_norm": 5.738655509195743,
+      "learning_rate": 8.050627337909244e-07,
+      "loss": 0.6943,
+      "step": 564
+    },
+    {
+      "epoch": 0.10804092169423463,
+      "grad_norm": 4.515498281634632,
+      "learning_rate": 8.052878545542145e-07,
+      "loss": 0.9675,
+      "step": 565
+    },
+    {
+      "epoch": 0.10823214456448992,
+      "grad_norm": 4.877198334969864,
+      "learning_rate": 8.055125772257615e-07,
+      "loss": 0.4901,
+      "step": 566
+    },
+    {
+      "epoch": 0.10842336743474519,
+      "grad_norm": 2.933676158494233,
+      "learning_rate": 8.057369032110095e-07,
+      "loss": 0.3891,
+      "step": 567
+    },
+    {
+      "epoch": 0.10861459030500048,
+      "grad_norm": 2.7862235187215068,
+      "learning_rate": 8.059608339079731e-07,
+      "loss": 0.1449,
+      "step": 568
+    },
+    {
+      "epoch": 0.10880581317525576,
+      "grad_norm": 3.1382315706058197,
+      "learning_rate": 8.061843707072893e-07,
+      "loss": 0.2559,
+      "step": 569
+    },
+    {
+      "epoch": 0.10899703604551104,
+      "grad_norm": 2.5718591552839594,
+      "learning_rate": 8.064075149922695e-07,
+      "loss": 0.3188,
+      "step": 570
+    },
+    {
+      "epoch": 0.10918825891576632,
+      "grad_norm": 3.1564101376841682,
+      "learning_rate": 8.066302681389513e-07,
+      "loss": 0.2762,
+      "step": 571
+    },
+    {
+      "epoch": 0.10937948178602161,
+      "grad_norm": 3.635383860850854,
+      "learning_rate": 8.068526315161484e-07,
+      "loss": 0.1493,
+      "step": 572
+    },
+    {
+      "epoch": 0.10957070465627688,
+      "grad_norm": 2.8502560483789523,
+      "learning_rate": 8.070746064855021e-07,
+      "loss": 0.0805,
+      "step": 573
+    },
+    {
+      "epoch": 0.10976192752653217,
+      "grad_norm": 1.8233717722679499,
+      "learning_rate": 8.072961944015306e-07,
+      "loss": 0.0584,
+      "step": 574
+    },
+    {
+      "epoch": 0.10995315039678745,
+      "grad_norm": 4.886350085701411,
+      "learning_rate": 8.075173966116791e-07,
+      "loss": 0.1381,
+      "step": 575
+    },
+    {
+      "epoch": 0.11014437326704274,
+      "grad_norm": 8.700060210622512,
+      "learning_rate": 8.07738214456369e-07,
+      "loss": 0.4786,
+      "step": 576
+    },
+    {
+      "epoch": 0.11033559613729801,
+      "grad_norm": 4.481765466303016,
+      "learning_rate": 8.079586492690458e-07,
+      "loss": 0.5252,
+      "step": 577
+    },
+    {
+      "epoch": 0.1105268190075533,
+      "grad_norm": 3.315711857469695,
+      "learning_rate": 8.081787023762283e-07,
+      "loss": 0.4944,
+      "step": 578
+    },
+    {
+      "epoch": 0.11071804187780858,
+      "grad_norm": 2.9046234173038847,
+      "learning_rate": 8.083983750975563e-07,
+      "loss": 0.5715,
+      "step": 579
+    },
+    {
+      "epoch": 0.11090926474806387,
+      "grad_norm": 3.188070507978576,
+      "learning_rate": 8.086176687458376e-07,
+      "loss": 0.5067,
+      "step": 580
+    },
+    {
+      "epoch": 0.11110048761831916,
+      "grad_norm": 5.432681524734754,
+      "learning_rate": 8.08836584627096e-07,
+      "loss": 0.6418,
+      "step": 581
+    },
+    {
+      "epoch": 0.11129171048857443,
+      "grad_norm": 3.4298332026994904,
+      "learning_rate": 8.090551240406173e-07,
+      "loss": 0.1308,
+      "step": 582
+    },
+    {
+      "epoch": 0.11148293335882972,
+      "grad_norm": 4.029625853783993,
+      "learning_rate": 8.092732882789951e-07,
+      "loss": 0.1294,
+      "step": 583
+    },
+    {
+      "epoch": 0.111674156229085,
+      "grad_norm": 3.437147642629907,
+      "learning_rate": 8.09491078628178e-07,
+      "loss": 0.1446,
+      "step": 584
+    },
+    {
+      "epoch": 0.11186537909934029,
+      "grad_norm": 3.6108785032687893,
+      "learning_rate": 8.09708496367514e-07,
+      "loss": 0.3056,
+      "step": 585
+    },
+    {
+      "epoch": 0.11205660196959556,
+      "grad_norm": 4.707258974003857,
+      "learning_rate": 8.099255427697959e-07,
+      "loss": 0.1659,
+      "step": 586
+    },
+    {
+      "epoch": 0.11224782483985085,
+      "grad_norm": 4.134083082024011,
+      "learning_rate": 8.101422191013063e-07,
+      "loss": 0.1397,
+      "step": 587
+    },
+    {
+      "epoch": 0.11243904771010613,
+      "grad_norm": 9.580797827515136,
+      "learning_rate": 8.103585266218613e-07,
+      "loss": 0.5202,
+      "step": 588
+    },
+    {
+      "epoch": 0.11263027058036142,
+      "grad_norm": 5.250949410468902,
+      "learning_rate": 8.105744665848553e-07,
+      "loss": 0.4793,
+      "step": 589
+    },
+    {
+      "epoch": 0.11282149345061669,
+      "grad_norm": 4.407129943857002,
+      "learning_rate": 8.107900402373036e-07,
+      "loss": 0.3449,
+      "step": 590
+    },
+    {
+      "epoch": 0.11301271632087198,
+      "grad_norm": 2.78095372368568,
+      "learning_rate": 8.110052488198865e-07,
+      "loss": 0.431,
+      "step": 591
+    },
+    {
+      "epoch": 0.11320393919112726,
+      "grad_norm": 2.5845881972655875,
+      "learning_rate": 8.112200935669913e-07,
+      "loss": 0.2633,
+      "step": 592
+    },
+    {
+      "epoch": 0.11339516206138255,
+      "grad_norm": 6.183215266468905,
+      "learning_rate": 8.114345757067559e-07,
+      "loss": 0.5887,
+      "step": 593
+    },
+    {
+      "epoch": 0.11358638493163782,
+      "grad_norm": 2.106835765490282,
+      "learning_rate": 8.116486964611095e-07,
+      "loss": 0.1857,
+      "step": 594
+    },
+    {
+      "epoch": 0.11377760780189311,
+      "grad_norm": 2.515087757437562,
+      "learning_rate": 8.118624570458157e-07,
+      "loss": 0.3126,
+      "step": 595
+    },
+    {
+      "epoch": 0.11396883067214839,
+      "grad_norm": 2.744510286350471,
+      "learning_rate": 8.120758586705131e-07,
+      "loss": 0.1696,
+      "step": 596
+    },
+    {
+      "epoch": 0.11416005354240367,
+      "grad_norm": 4.386143905735015,
+      "learning_rate": 8.122889025387566e-07,
+      "loss": 0.1026,
+      "step": 597
+    },
+    {
+      "epoch": 0.11435127641265895,
+      "grad_norm": 5.257725979913597,
+      "learning_rate": 8.125015898480581e-07,
+      "loss": 0.2496,
+      "step": 598
+    },
+    {
+      "epoch": 0.11454249928291424,
+      "grad_norm": 4.968121758816384,
+      "learning_rate": 8.127139217899272e-07,
+      "loss": 0.1875,
+      "step": 599
+    },
+    {
+      "epoch": 0.11473372215316952,
+      "grad_norm": 7.700946987427834,
+      "learning_rate": 8.129258995499107e-07,
+      "loss": 0.3312,
+      "step": 600
+    },
+    {
+      "epoch": 0.1149249450234248,
+      "grad_norm": 7.295317979958854,
+      "learning_rate": 8.131375243076327e-07,
+      "loss": 0.2643,
+      "step": 601
+    },
+    {
+      "epoch": 0.11511616789368008,
+      "grad_norm": 7.265867036459166,
+      "learning_rate": 8.133487972368338e-07,
+      "loss": 0.5976,
+      "step": 602
+    },
+    {
+      "epoch": 0.11530739076393537,
+      "grad_norm": 4.518168752598018,
+      "learning_rate": 8.135597195054105e-07,
+      "loss": 0.4805,
+      "step": 603
+    },
+    {
+      "epoch": 0.11549861363419064,
+      "grad_norm": 4.76702139986107,
+      "learning_rate": 8.137702922754533e-07,
+      "loss": 0.6433,
+      "step": 604
+    },
+    {
+      "epoch": 0.11568983650444593,
+      "grad_norm": 4.0819240136183454,
+      "learning_rate": 8.139805167032856e-07,
+      "loss": 0.377,
+      "step": 605
+    },
+    {
+      "epoch": 0.11588105937470121,
+      "grad_norm": 4.649922417434422,
+      "learning_rate": 8.141903939395019e-07,
+      "loss": 0.2709,
+      "step": 606
+    },
+    {
+      "epoch": 0.1160722822449565,
+      "grad_norm": 3.502943844496762,
+      "learning_rate": 8.143999251290055e-07,
+      "loss": 0.3221,
+      "step": 607
+    },
+    {
+      "epoch": 0.11626350511521177,
+      "grad_norm": 3.164733363486234,
+      "learning_rate": 8.146091114110453e-07,
+      "loss": 0.1107,
+      "step": 608
+    },
+    {
+      "epoch": 0.11645472798546706,
+      "grad_norm": 3.4947237026164735,
+      "learning_rate": 8.148179539192539e-07,
+      "loss": 0.1291,
+      "step": 609
+    },
+    {
+      "epoch": 0.11664595085572234,
+      "grad_norm": 2.914151515061352,
+      "learning_rate": 8.15026453781684e-07,
+      "loss": 0.1249,
+      "step": 610
+    },
+    {
+      "epoch": 0.11683717372597763,
+      "grad_norm": 6.662763820098391,
+      "learning_rate": 8.152346121208451e-07,
+      "loss": 0.2405,
+      "step": 611
+    },
+    {
+      "epoch": 0.1170283965962329,
+      "grad_norm": 8.258267046382976,
+      "learning_rate": 8.154424300537396e-07,
+      "loss": 0.1728,
+      "step": 612
+    },
+    {
+      "epoch": 0.11721961946648819,
+      "grad_norm": 6.658842772791223,
+      "learning_rate": 8.156499086918988e-07,
+      "loss": 0.2684,
+      "step": 613
+    },
+    {
+      "epoch": 0.11741084233674347,
+      "grad_norm": 4.902189870904176,
+      "learning_rate": 8.158570491414191e-07,
+      "loss": 0.4786,
+      "step": 614
+    },
+    {
+      "epoch": 0.11760206520699876,
+      "grad_norm": 5.229987922202291,
+      "learning_rate": 8.160638525029962e-07,
+      "loss": 0.8841,
+      "step": 615
+    },
+    {
+      "epoch": 0.11779328807725405,
+      "grad_norm": 3.0004175213508533,
+      "learning_rate": 8.162703198719613e-07,
+      "loss": 0.3155,
+      "step": 616
+    },
+    {
+      "epoch": 0.11798451094750932,
+      "grad_norm": 3.5188192140537495,
+      "learning_rate": 8.164764523383157e-07,
+      "loss": 0.4966,
+      "step": 617
+    },
+    {
+      "epoch": 0.11817573381776461,
+      "grad_norm": 3.7359152454589277,
+      "learning_rate": 8.166822509867647e-07,
+      "loss": 0.3786,
+      "step": 618
+    },
+    {
+      "epoch": 0.11836695668801989,
+      "grad_norm": 4.418634204345447,
+      "learning_rate": 8.168877168967527e-07,
+      "loss": 0.3448,
+      "step": 619
+    },
+    {
+      "epoch": 0.11855817955827518,
+      "grad_norm": 2.635781626001718,
+      "learning_rate": 8.170928511424965e-07,
+      "loss": 0.1973,
+      "step": 620
+    },
+    {
+      "epoch": 0.11874940242853045,
+      "grad_norm": 3.41628111237848,
+      "learning_rate": 8.172976547930193e-07,
+      "loss": 0.1898,
+      "step": 621
+    },
+    {
+      "epoch": 0.11894062529878574,
+      "grad_norm": 3.9843295225652167,
+      "learning_rate": 8.175021289121841e-07,
+      "loss": 0.1923,
+      "step": 622
+    },
+    {
+      "epoch": 0.11913184816904102,
+      "grad_norm": 2.961238634402491,
+      "learning_rate": 8.17706274558727e-07,
+      "loss": 0.1735,
+      "step": 623
+    },
+    {
+      "epoch": 0.1193230710392963,
+      "grad_norm": 3.893233684762568,
+      "learning_rate": 8.179100927862898e-07,
+      "loss": 0.269,
+      "step": 624
+    },
+    {
+      "epoch": 0.11951429390955158,
+      "grad_norm": 5.502294408622213,
+      "learning_rate": 8.181135846434525e-07,
+      "loss": 0.2465,
+      "step": 625
+    },
+    {
+      "epoch": 0.11970551677980687,
+      "grad_norm": 5.556588793145781,
+      "learning_rate": 8.183167511737664e-07,
+      "loss": 0.5243,
+      "step": 626
+    },
+    {
+      "epoch": 0.11989673965006215,
+      "grad_norm": 3.9331254424636377,
+      "learning_rate": 8.185195934157858e-07,
+      "loss": 0.4858,
+      "step": 627
+    },
+    {
+      "epoch": 0.12008796252031743,
+      "grad_norm": 5.2342562078404615,
+      "learning_rate": 8.187221124030991e-07,
+      "loss": 0.4925,
+      "step": 628
+    },
+    {
+      "epoch": 0.12027918539057271,
+      "grad_norm": 3.851279350844941,
+      "learning_rate": 8.18924309164362e-07,
+      "loss": 0.6046,
+      "step": 629
+    },
+    {
+      "epoch": 0.120470408260828,
+      "grad_norm": 5.333326915896052,
+      "learning_rate": 8.19126184723327e-07,
+      "loss": 0.4016,
+      "step": 630
+    },
+    {
+      "epoch": 0.12066163113108327,
+      "grad_norm": 3.9134669659839543,
+      "learning_rate": 8.193277400988761e-07,
+      "loss": 0.4016,
+      "step": 631
+    },
+    {
+      "epoch": 0.12085285400133856,
+      "grad_norm": 4.521816773688918,
+      "learning_rate": 8.195289763050512e-07,
+      "loss": 0.5702,
+      "step": 632
+    },
+    {
+      "epoch": 0.12104407687159384,
+      "grad_norm": 2.518136328820607,
+      "learning_rate": 8.197298943510837e-07,
+      "loss": 0.1794,
+      "step": 633
+    },
+    {
+      "epoch": 0.12123529974184913,
+      "grad_norm": 6.174542355277769,
+      "learning_rate": 8.199304952414266e-07,
+      "loss": 0.3751,
+      "step": 634
+    },
+    {
+      "epoch": 0.1214265226121044,
+      "grad_norm": 4.133720428019578,
+      "learning_rate": 8.201307799757827e-07,
+      "loss": 0.1916,
+      "step": 635
+    },
+    {
+      "epoch": 0.1216177454823597,
+      "grad_norm": 5.256740602510739,
+      "learning_rate": 8.203307495491365e-07,
+      "loss": 0.2093,
+      "step": 636
+    },
+    {
+      "epoch": 0.12180896835261497,
+      "grad_norm": 7.05314597421214,
+      "learning_rate": 8.20530404951782e-07,
+      "loss": 0.1383,
+      "step": 637
+    },
+    {
+      "epoch": 0.12200019122287026,
+      "grad_norm": 4.983678308664488,
+      "learning_rate": 8.207297471693538e-07,
+      "loss": 0.2274,
+      "step": 638
+    },
+    {
+      "epoch": 0.12219141409312553,
+      "grad_norm": 4.37326298699751,
+      "learning_rate": 8.209287771828549e-07,
+      "loss": 0.3113,
+      "step": 639
+    },
+    {
+      "epoch": 0.12238263696338082,
+      "grad_norm": 6.55650302627809,
+      "learning_rate": 8.211274959686864e-07,
+      "loss": 0.3547,
+      "step": 640
+    },
+    {
+      "epoch": 0.1225738598336361,
+      "grad_norm": 2.9295766580594504,
+      "learning_rate": 8.213259044986763e-07,
+      "loss": 0.1771,
+      "step": 641
+    },
+    {
+      "epoch": 0.12276508270389139,
+      "grad_norm": 5.807637467516509,
+      "learning_rate": 8.215240037401073e-07,
+      "loss": 0.201,
+      "step": 642
+    },
+    {
+      "epoch": 0.12295630557414666,
+      "grad_norm": 4.244203316883387,
+      "learning_rate": 8.217217946557463e-07,
+      "loss": 0.2093,
+      "step": 643
+    },
+    {
+      "epoch": 0.12314752844440195,
+      "grad_norm": 2.827863565237441,
+      "learning_rate": 8.219192782038712e-07,
+      "loss": 0.236,
+      "step": 644
+    },
+    {
+      "epoch": 0.12333875131465723,
+      "grad_norm": 3.3916712961626128,
+      "learning_rate": 8.221164553382995e-07,
+      "loss": 0.2092,
+      "step": 645
+    },
+    {
+      "epoch": 0.12352997418491252,
+      "grad_norm": 5.307233825707824,
+      "learning_rate": 8.223133270084158e-07,
+      "loss": 0.3207,
+      "step": 646
+    },
+    {
+      "epoch": 0.12372119705516779,
+      "grad_norm": 3.795224808378335,
+      "learning_rate": 8.225098941591994e-07,
+      "loss": 0.3294,
+      "step": 647
+    },
+    {
+      "epoch": 0.12391241992542308,
+      "grad_norm": 2.9521952785317995,
+      "learning_rate": 8.227061577312508e-07,
+      "loss": 0.0687,
+      "step": 648
+    },
+    {
+      "epoch": 0.12410364279567836,
+      "grad_norm": 4.740087707287526,
+      "learning_rate": 8.229021186608199e-07,
+      "loss": 0.0755,
+      "step": 649
+    },
+    {
+      "epoch": 0.12429486566593365,
+      "grad_norm": 4.584303712713143,
+      "learning_rate": 8.230977778798313e-07,
+      "loss": 0.3394,
+      "step": 650
+    },
+    {
+      "epoch": 0.12448608853618892,
+      "grad_norm": 4.942314215310116,
+      "learning_rate": 8.232931363159127e-07,
+      "loss": 0.3568,
+      "step": 651
+    },
+    {
+      "epoch": 0.12467731140644421,
+      "grad_norm": 5.047247810884417,
+      "learning_rate": 8.234881948924195e-07,
+      "loss": 0.923,
+      "step": 652
+    },
+    {
+      "epoch": 0.1248685342766995,
+      "grad_norm": 4.6980547267930355,
+      "learning_rate": 8.236829545284619e-07,
+      "loss": 0.5192,
+      "step": 653
+    },
+    {
+      "epoch": 0.12505975714695478,
+      "grad_norm": 7.0128502742983585,
+      "learning_rate": 8.238774161389313e-07,
+      "loss": 0.3228,
+      "step": 654
+    },
+    {
+      "epoch": 0.12525098001721005,
+      "grad_norm": 3.4990318866769323,
+      "learning_rate": 8.240715806345247e-07,
+      "loss": 0.0948,
+      "step": 655
+    },
+    {
+      "epoch": 0.12544220288746535,
+      "grad_norm": 4.419577065510713,
+      "learning_rate": 8.242654489217719e-07,
+      "loss": 0.1963,
+      "step": 656
+    },
+    {
+      "epoch": 0.12563342575772063,
+      "grad_norm": 2.7416500423433687,
+      "learning_rate": 8.244590219030597e-07,
+      "loss": 0.1906,
+      "step": 657
+    },
+    {
+      "epoch": 0.1258246486279759,
+      "grad_norm": 4.664915687314292,
+      "learning_rate": 8.246523004766582e-07,
+      "loss": 0.2968,
+      "step": 658
+    },
+    {
+      "epoch": 0.12601587149823118,
+      "grad_norm": 4.264042263744809,
+      "learning_rate": 8.248452855367438e-07,
+      "loss": 0.2182,
+      "step": 659
+    },
+    {
+      "epoch": 0.12620709436848648,
+      "grad_norm": 2.7907360195789463,
+      "learning_rate": 8.250379779734269e-07,
+      "loss": 0.108,
+      "step": 660
+    },
+    {
+      "epoch": 0.12639831723874176,
+      "grad_norm": 2.795994113746745,
+      "learning_rate": 8.252303786727742e-07,
+      "loss": 0.0919,
+      "step": 661
+    },
+    {
+      "epoch": 0.12658954010899703,
+      "grad_norm": 3.7787298205887696,
+      "learning_rate": 8.254224885168338e-07,
+      "loss": 0.1062,
+      "step": 662
+    },
+    {
+      "epoch": 0.1267807629792523,
+      "grad_norm": 8.874490642706231,
+      "learning_rate": 8.256143083836602e-07,
+      "loss": 0.3567,
+      "step": 663
+    },
+    {
+      "epoch": 0.1269719858495076,
+      "grad_norm": 6.788956302211692,
+      "learning_rate": 8.258058391473374e-07,
+      "loss": 0.7042,
+      "step": 664
+    },
+    {
+      "epoch": 0.1271632087197629,
+      "grad_norm": 3.8324075216643356,
+      "learning_rate": 8.259970816780033e-07,
+      "loss": 0.3641,
+      "step": 665
+    },
+    {
+      "epoch": 0.12735443159001816,
+      "grad_norm": 3.684577818538604,
+      "learning_rate": 8.261880368418732e-07,
+      "loss": 0.3928,
+      "step": 666
+    },
+    {
+      "epoch": 0.12754565446027344,
+      "grad_norm": 3.5713207065097246,
+      "learning_rate": 8.263787055012635e-07,
+      "loss": 0.3934,
+      "step": 667
+    },
+    {
+      "epoch": 0.12773687733052874,
+      "grad_norm": 3.340852997638635,
+      "learning_rate": 8.265690885146154e-07,
+      "loss": 0.2038,
+      "step": 668
+    },
+    {
+      "epoch": 0.12792810020078402,
+      "grad_norm": 4.411209021334993,
+      "learning_rate": 8.267591867365172e-07,
+      "loss": 0.2798,
+      "step": 669
+    },
+    {
+      "epoch": 0.1281193230710393,
+      "grad_norm": 5.2980185205547015,
+      "learning_rate": 8.269490010177279e-07,
+      "loss": 0.615,
+      "step": 670
+    },
+    {
+      "epoch": 0.12831054594129457,
+      "grad_norm": 5.319467978872236,
+      "learning_rate": 8.271385322052003e-07,
+      "loss": 0.2417,
+      "step": 671
+    },
+    {
+      "epoch": 0.12850176881154987,
+      "grad_norm": 4.490099506252596,
+      "learning_rate": 8.273277811421027e-07,
+      "loss": 0.2012,
+      "step": 672
+    },
+    {
+      "epoch": 0.12869299168180515,
+      "grad_norm": 4.0632976042439095,
+      "learning_rate": 8.275167486678428e-07,
+      "loss": 0.1524,
+      "step": 673
+    },
+    {
+      "epoch": 0.12888421455206042,
+      "grad_norm": 4.17977694121592,
+      "learning_rate": 8.277054356180889e-07,
+      "loss": 0.1321,
+      "step": 674
+    },
+    {
+      "epoch": 0.1290754374223157,
+      "grad_norm": 5.6852741287421935,
+      "learning_rate": 8.278938428247926e-07,
+      "loss": 0.2382,
+      "step": 675
+    },
+    {
+      "epoch": 0.129266660292571,
+      "grad_norm": 8.95486023283171,
+      "learning_rate": 8.280819711162104e-07,
+      "loss": 0.6329,
+      "step": 676
+    },
+    {
+      "epoch": 0.12945788316282628,
+      "grad_norm": 7.7367753436755455,
+      "learning_rate": 8.28269821316926e-07,
+      "loss": 0.5651,
+      "step": 677
+    },
+    {
+      "epoch": 0.12964910603308155,
+      "grad_norm": 4.545684134147084,
+      "learning_rate": 8.284573942478719e-07,
+      "loss": 0.467,
+      "step": 678
+    },
+    {
+      "epoch": 0.12984032890333683,
+      "grad_norm": 2.8243853043741565,
+      "learning_rate": 8.286446907263509e-07,
+      "loss": 0.2342,
+      "step": 679
+    },
+    {
+      "epoch": 0.13003155177359213,
+      "grad_norm": 5.014933981189918,
+      "learning_rate": 8.288317115660569e-07,
+      "loss": 0.5885,
+      "step": 680
+    },
+    {
+      "epoch": 0.1302227746438474,
+      "grad_norm": 2.964313927153435,
+      "learning_rate": 8.290184575770972e-07,
+      "loss": 0.2872,
+      "step": 681
+    },
+    {
+      "epoch": 0.13041399751410268,
+      "grad_norm": 2.759375110595977,
+      "learning_rate": 8.292049295660127e-07,
+      "loss": 0.2327,
+      "step": 682
+    },
+    {
+      "epoch": 0.13060522038435796,
+      "grad_norm": 3.926329252796712,
+      "learning_rate": 8.29391128335799e-07,
+      "loss": 0.1479,
+      "step": 683
+    },
+    {
+      "epoch": 0.13079644325461326,
+      "grad_norm": 5.709048186972315,
+      "learning_rate": 8.295770546859271e-07,
+      "loss": 0.3726,
+      "step": 684
+    },
+    {
+      "epoch": 0.13098766612486853,
+      "grad_norm": 7.155759303165108,
+      "learning_rate": 8.297627094123646e-07,
+      "loss": 0.2276,
+      "step": 685
+    },
+    {
+      "epoch": 0.1311788889951238,
+      "grad_norm": 6.24878711375787,
+      "learning_rate": 8.299480933075951e-07,
+      "loss": 0.4387,
+      "step": 686
+    },
+    {
+      "epoch": 0.13137011186537909,
+      "grad_norm": 4.94436554482667,
+      "learning_rate": 8.301332071606393e-07,
+      "loss": 0.1672,
+      "step": 687
+    },
+    {
+      "epoch": 0.1315613347356344,
+      "grad_norm": 5.649337844532316,
+      "learning_rate": 8.303180517570751e-07,
+      "loss": 0.331,
+      "step": 688
+    },
+    {
+      "epoch": 0.13175255760588966,
+      "grad_norm": 7.207277106663537,
+      "learning_rate": 8.305026278790571e-07,
+      "loss": 0.3126,
+      "step": 689
+    },
+    {
+      "epoch": 0.13194378047614494,
+      "grad_norm": 3.8297865395286057,
+      "learning_rate": 8.306869363053367e-07,
+      "loss": 0.2574,
+      "step": 690
+    },
+    {
+      "epoch": 0.13213500334640024,
+      "grad_norm": 2.6142431293655175,
+      "learning_rate": 8.308709778112823e-07,
+      "loss": 0.2838,
+      "step": 691
+    },
+    {
+      "epoch": 0.13232622621665552,
+      "grad_norm": 2.510547133309657,
+      "learning_rate": 8.310547531688977e-07,
+      "loss": 0.1928,
+      "step": 692
+    },
+    {
+      "epoch": 0.1325174490869108,
+      "grad_norm": 5.92891675038601,
+      "learning_rate": 8.312382631468432e-07,
+      "loss": 1.058,
+      "step": 693
+    },
+    {
+      "epoch": 0.13270867195716607,
+      "grad_norm": 4.1545693577628375,
+      "learning_rate": 8.314215085104534e-07,
+      "loss": 0.3845,
+      "step": 694
+    },
+    {
+      "epoch": 0.13289989482742137,
+      "grad_norm": 2.7066733824203437,
+      "learning_rate": 8.316044900217567e-07,
+      "loss": 0.0842,
+      "step": 695
+    },
+    {
+      "epoch": 0.13309111769767665,
+      "grad_norm": 3.281034190029134,
+      "learning_rate": 8.317872084394952e-07,
+      "loss": 0.1626,
+      "step": 696
+    },
+    {
+      "epoch": 0.13328234056793192,
+      "grad_norm": 4.176459490201122,
+      "learning_rate": 8.319696645191424e-07,
+      "loss": 0.1429,
+      "step": 697
+    },
+    {
+      "epoch": 0.1334735634381872,
+      "grad_norm": 3.000368731408535,
+      "learning_rate": 8.321518590129228e-07,
+      "loss": 0.1879,
+      "step": 698
+    },
+    {
+      "epoch": 0.1336647863084425,
+      "grad_norm": 2.738063743757689,
+      "learning_rate": 8.323337926698297e-07,
+      "loss": 0.1068,
+      "step": 699
+    },
+    {
+      "epoch": 0.13385600917869778,
+      "grad_norm": 4.932749527247412,
+      "learning_rate": 8.325154662356444e-07,
+      "loss": 0.2229,
+      "step": 700
+    },
+    {
+      "epoch": 0.13404723204895305,
+      "grad_norm": 5.995893026465237,
+      "learning_rate": 8.326968804529545e-07,
+      "loss": 0.2936,
+      "step": 701
+    },
+    {
+      "epoch": 0.13423845491920833,
+      "grad_norm": 9.43334667443005,
+      "learning_rate": 8.328780360611715e-07,
+      "loss": 0.6052,
+      "step": 702
+    },
+    {
+      "epoch": 0.13442967778946363,
+      "grad_norm": 6.595575595172921,
+      "learning_rate": 8.330589337965495e-07,
+      "loss": 0.4899,
+      "step": 703
+    },
+    {
+      "epoch": 0.1346209006597189,
+      "grad_norm": 2.2543870753511372,
+      "learning_rate": 8.332395743922026e-07,
+      "loss": 0.3329,
+      "step": 704
+    },
+    {
+      "epoch": 0.13481212352997418,
+      "grad_norm": 2.724851579736371,
+      "learning_rate": 8.334199585781237e-07,
+      "loss": 0.2025,
+      "step": 705
+    },
+    {
+      "epoch": 0.13500334640022946,
+      "grad_norm": 4.850717172984337,
+      "learning_rate": 8.336000870812008e-07,
+      "loss": 0.52,
+      "step": 706
+    },
+    {
+      "epoch": 0.13519456927048476,
+      "grad_norm": 3.588862670065184,
+      "learning_rate": 8.337799606252358e-07,
+      "loss": 0.4943,
+      "step": 707
+    },
+    {
+      "epoch": 0.13538579214074004,
+      "grad_norm": 3.935125937088554,
+      "learning_rate": 8.339595799309612e-07,
+      "loss": 0.2554,
+      "step": 708
+    },
+    {
+      "epoch": 0.1355770150109953,
+      "grad_norm": 3.155900690370416,
+      "learning_rate": 8.341389457160581e-07,
+      "loss": 0.2196,
+      "step": 709
+    },
+    {
+      "epoch": 0.1357682378812506,
+      "grad_norm": 3.735936688239992,
+      "learning_rate": 8.343180586951725e-07,
+      "loss": 0.2065,
+      "step": 710
+    },
+    {
+      "epoch": 0.1359594607515059,
+      "grad_norm": 3.6565687252572316,
+      "learning_rate": 8.34496919579933e-07,
+      "loss": 0.1044,
+      "step": 711
+    },
+    {
+      "epoch": 0.13615068362176116,
+      "grad_norm": 2.8475117378562294,
+      "learning_rate": 8.346755290789683e-07,
+      "loss": 0.1121,
+      "step": 712
+    },
+    {
+      "epoch": 0.13634190649201644,
+      "grad_norm": 9.143981004510007,
+      "learning_rate": 8.348538878979225e-07,
+      "loss": 0.3217,
+      "step": 713
+    },
+    {
+      "epoch": 0.13653312936227172,
+      "grad_norm": 5.160650791262229,
+      "learning_rate": 8.350319967394732e-07,
+      "loss": 0.4366,
+      "step": 714
+    },
+    {
+      "epoch": 0.13672435223252702,
+      "grad_norm": 3.7289959461766062,
+      "learning_rate": 8.352098563033476e-07,
+      "loss": 0.4575,
+      "step": 715
+    },
+    {
+      "epoch": 0.1369155751027823,
+      "grad_norm": 2.988858515015109,
+      "learning_rate": 8.35387467286339e-07,
+      "loss": 0.3337,
+      "step": 716
+    },
+    {
+      "epoch": 0.13710679797303757,
+      "grad_norm": 2.2041754416547685,
+      "learning_rate": 8.35564830382323e-07,
+      "loss": 0.2148,
+      "step": 717
+    },
+    {
+      "epoch": 0.13729802084329285,
+      "grad_norm": 3.9817012417360473,
+      "learning_rate": 8.357419462822745e-07,
+      "loss": 0.4471,
+      "step": 718
+    },
+    {
+      "epoch": 0.13748924371354815,
+      "grad_norm": 3.5326357122810714,
+      "learning_rate": 8.359188156742828e-07,
+      "loss": 0.3017,
+      "step": 719
+    },
+    {
+      "epoch": 0.13768046658380342,
+      "grad_norm": 5.499627013997478,
+      "learning_rate": 8.360954392435682e-07,
+      "loss": 0.3438,
+      "step": 720
+    },
+    {
+      "epoch": 0.1378716894540587,
+      "grad_norm": 4.015446402886375,
+      "learning_rate": 8.362718176724985e-07,
+      "loss": 0.2512,
+      "step": 721
+    },
+    {
+      "epoch": 0.13806291232431397,
+      "grad_norm": 4.4256489423350125,
+      "learning_rate": 8.364479516406033e-07,
+      "loss": 0.1951,
+      "step": 722
+    },
+    {
+      "epoch": 0.13825413519456928,
+      "grad_norm": 2.6989982724716013,
+      "learning_rate": 8.366238418245916e-07,
+      "loss": 0.3233,
+      "step": 723
+    },
+    {
+      "epoch": 0.13844535806482455,
+      "grad_norm": 4.220749896706954,
+      "learning_rate": 8.367994888983653e-07,
+      "loss": 0.0799,
+      "step": 724
+    },
+    {
+      "epoch": 0.13863658093507983,
+      "grad_norm": 4.6922128001524035,
+      "learning_rate": 8.36974893533037e-07,
+      "loss": 0.2609,
+      "step": 725
+    },
+    {
+      "epoch": 0.13882780380533513,
+      "grad_norm": 5.239585355703926,
+      "learning_rate": 8.371500563969431e-07,
+      "loss": 0.3794,
+      "step": 726
+    },
+    {
+      "epoch": 0.1390190266755904,
+      "grad_norm": 5.6444271724833825,
+      "learning_rate": 8.373249781556611e-07,
+      "loss": 0.4265,
+      "step": 727
+    },
+    {
+      "epoch": 0.13921024954584568,
+      "grad_norm": 12.874825189385158,
+      "learning_rate": 8.374996594720234e-07,
+      "loss": 0.1557,
+      "step": 728
+    },
+    {
+      "epoch": 0.13940147241610096,
+      "grad_norm": 6.246589035045759,
+      "learning_rate": 8.376741010061326e-07,
+      "loss": 0.4656,
+      "step": 729
+    },
+    {
+      "epoch": 0.13959269528635626,
+      "grad_norm": 3.553782320576692,
+      "learning_rate": 8.378483034153773e-07,
+      "loss": 0.4438,
+      "step": 730
+    },
+    {
+      "epoch": 0.13978391815661154,
+      "grad_norm": 8.414150000270425,
+      "learning_rate": 8.380222673544456e-07,
+      "loss": 0.3461,
+      "step": 731
+    },
+    {
+      "epoch": 0.1399751410268668,
+      "grad_norm": 3.992427333488187,
+      "learning_rate": 8.381959934753415e-07,
+      "loss": 0.1596,
+      "step": 732
+    },
+    {
+      "epoch": 0.1401663638971221,
+      "grad_norm": 3.253622823509377,
+      "learning_rate": 8.383694824273978e-07,
+      "loss": 0.2173,
+      "step": 733
+    },
+    {
+      "epoch": 0.1403575867673774,
+      "grad_norm": 6.650463185635727,
+      "learning_rate": 8.385427348572922e-07,
+      "loss": 0.2942,
+      "step": 734
+    },
+    {
+      "epoch": 0.14054880963763267,
+      "grad_norm": 6.750220683693567,
+      "learning_rate": 8.387157514090607e-07,
+      "loss": 0.2192,
+      "step": 735
+    },
+    {
+      "epoch": 0.14074003250788794,
+      "grad_norm": 2.4290687288287622,
+      "learning_rate": 8.388885327241125e-07,
+      "loss": 0.0661,
+      "step": 736
+    },
+    {
+      "epoch": 0.14093125537814322,
+      "grad_norm": 4.998038288569705,
+      "learning_rate": 8.390610794412441e-07,
+      "loss": 0.2057,
+      "step": 737
+    },
+    {
+      "epoch": 0.14112247824839852,
+      "grad_norm": 6.666571266763296,
+      "learning_rate": 8.392333921966537e-07,
+      "loss": 0.2478,
+      "step": 738
+    },
+    {
+      "epoch": 0.1413137011186538,
+      "grad_norm": 7.802970017561948,
+      "learning_rate": 8.394054716239548e-07,
+      "loss": 0.614,
+      "step": 739
+    },
+    {
+      "epoch": 0.14150492398890907,
+      "grad_norm": 4.682713417122535,
+      "learning_rate": 8.395773183541907e-07,
+      "loss": 0.5138,
+      "step": 740
+    },
+    {
+      "epoch": 0.14169614685916435,
+      "grad_norm": 5.387413063144449,
+      "learning_rate": 8.397489330158477e-07,
+      "loss": 0.2729,
+      "step": 741
+    },
+    {
+      "epoch": 0.14188736972941965,
+      "grad_norm": 3.479057188462596,
+      "learning_rate": 8.399203162348702e-07,
+      "loss": 0.3687,
+      "step": 742
+    },
+    {
+      "epoch": 0.14207859259967492,
+      "grad_norm": 4.551272857700814,
+      "learning_rate": 8.400914686346726e-07,
+      "loss": 0.3785,
+      "step": 743
+    },
+    {
+      "epoch": 0.1422698154699302,
+      "grad_norm": 7.804578503524439,
+      "learning_rate": 8.40262390836154e-07,
+      "loss": 0.6874,
+      "step": 744
+    },
+    {
+      "epoch": 0.14246103834018548,
+      "grad_norm": 3.8687682957131884,
+      "learning_rate": 8.404330834577123e-07,
+      "loss": 0.4866,
+      "step": 745
+    },
+    {
+      "epoch": 0.14265226121044078,
+      "grad_norm": 5.863589304761385,
+      "learning_rate": 8.406035471152562e-07,
+      "loss": 0.2371,
+      "step": 746
+    },
+    {
+      "epoch": 0.14284348408069605,
+      "grad_norm": 4.707764830349534,
+      "learning_rate": 8.407737824222193e-07,
+      "loss": 0.2128,
+      "step": 747
+    },
+    {
+      "epoch": 0.14303470695095133,
+      "grad_norm": 4.251489490469575,
+      "learning_rate": 8.409437899895732e-07,
+      "loss": 0.1703,
+      "step": 748
+    },
+    {
+      "epoch": 0.1432259298212066,
+      "grad_norm": 2.966800250214094,
+      "learning_rate": 8.411135704258411e-07,
+      "loss": 0.2383,
+      "step": 749
+    },
+    {
+      "epoch": 0.1434171526914619,
+      "grad_norm": 7.344924638022668,
+      "learning_rate": 8.4128312433711e-07,
+      "loss": 0.2943,
+      "step": 750
+    },
+    {
+      "epoch": 0.14360837556171718,
+      "grad_norm": 7.862119767982026,
+      "learning_rate": 8.414524523270445e-07,
+      "loss": 0.3308,
+      "step": 751
+    },
+    {
+      "epoch": 0.14379959843197246,
+      "grad_norm": 3.4859638782496623,
+      "learning_rate": 8.416215549968995e-07,
+      "loss": 0.4663,
+      "step": 752
+    },
+    {
+      "epoch": 0.14399082130222773,
+      "grad_norm": 4.676784627618103,
+      "learning_rate": 8.41790432945532e-07,
+      "loss": 0.3519,
+      "step": 753
+    },
+    {
+      "epoch": 0.14418204417248304,
+      "grad_norm": 2.237665700903838,
+      "learning_rate": 8.419590867694159e-07,
+      "loss": 0.2823,
+      "step": 754
+    },
+    {
+      "epoch": 0.1443732670427383,
+      "grad_norm": 2.1779656941678223,
+      "learning_rate": 8.421275170626525e-07,
+      "loss": 0.0762,
+      "step": 755
+    },
+    {
+      "epoch": 0.1445644899129936,
+      "grad_norm": 2.686916997793837,
+      "learning_rate": 8.422957244169846e-07,
+      "loss": 0.1363,
+      "step": 756
+    },
+    {
+      "epoch": 0.14475571278324886,
+      "grad_norm": 5.085788988327088,
+      "learning_rate": 8.424637094218077e-07,
+      "loss": 0.3886,
+      "step": 757
+    },
+    {
+      "epoch": 0.14494693565350417,
+      "grad_norm": 2.6631666002692724,
+      "learning_rate": 8.426314726641841e-07,
+      "loss": 0.1785,
+      "step": 758
+    },
+    {
+      "epoch": 0.14513815852375944,
+      "grad_norm": 2.795637059227297,
+      "learning_rate": 8.42799014728853e-07,
+      "loss": 0.2409,
+      "step": 759
+    },
+    {
+      "epoch": 0.14532938139401472,
+      "grad_norm": 2.6549739689817033,
+      "learning_rate": 8.429663361982447e-07,
+      "loss": 0.2129,
+      "step": 760
+    },
+    {
+      "epoch": 0.14552060426427002,
+      "grad_norm": 2.478423661734591,
+      "learning_rate": 8.431334376524914e-07,
+      "loss": 0.0661,
+      "step": 761
+    },
+    {
+      "epoch": 0.1457118271345253,
+      "grad_norm": 3.361353943374842,
+      "learning_rate": 8.433003196694403e-07,
+      "loss": 0.1218,
+      "step": 762
+    },
+    {
+      "epoch": 0.14590305000478057,
+      "grad_norm": 6.336697270645386,
+      "learning_rate": 8.434669828246649e-07,
+      "loss": 0.2231,
+      "step": 763
+    },
+    {
+      "epoch": 0.14609427287503585,
+      "grad_norm": 6.181755713284897,
+      "learning_rate": 8.43633427691477e-07,
+      "loss": 0.662,
+      "step": 764
+    },
+    {
+      "epoch": 0.14628549574529115,
+      "grad_norm": 5.270431543448297,
+      "learning_rate": 8.437996548409389e-07,
+      "loss": 0.7659,
+      "step": 765
+    },
+    {
+      "epoch": 0.14647671861554643,
+      "grad_norm": 2.204202591351508,
+      "learning_rate": 8.439656648418745e-07,
+      "loss": 0.4799,
+      "step": 766
+    },
+    {
+      "epoch": 0.1466679414858017,
+      "grad_norm": 2.582740582378152,
+      "learning_rate": 8.441314582608819e-07,
+      "loss": 0.1577,
+      "step": 767
+    },
+    {
+      "epoch": 0.14685916435605698,
+      "grad_norm": 4.9528057585162575,
+      "learning_rate": 8.44297035662344e-07,
+      "loss": 0.4503,
+      "step": 768
+    },
+    {
+      "epoch": 0.14705038722631228,
+      "grad_norm": 3.974457848962078,
+      "learning_rate": 8.444623976084409e-07,
+      "loss": 0.4331,
+      "step": 769
+    },
+    {
+      "epoch": 0.14724161009656755,
+      "grad_norm": 2.1764887921189113,
+      "learning_rate": 8.446275446591606e-07,
+      "loss": 0.2566,
+      "step": 770
+    },
+    {
+      "epoch": 0.14743283296682283,
+      "grad_norm": 2.9169944760620417,
+      "learning_rate": 8.447924773723112e-07,
+      "loss": 0.2271,
+      "step": 771
+    },
+    {
+      "epoch": 0.1476240558370781,
+      "grad_norm": 2.786660065285683,
+      "learning_rate": 8.449571963035312e-07,
+      "loss": 0.1284,
+      "step": 772
+    },
+    {
+      "epoch": 0.1478152787073334,
+      "grad_norm": 4.215472269083977,
+      "learning_rate": 8.451217020063017e-07,
+      "loss": 0.1672,
+      "step": 773
+    },
+    {
+      "epoch": 0.14800650157758868,
+      "grad_norm": 5.427977543806206,
+      "learning_rate": 8.452859950319571e-07,
+      "loss": 0.1873,
+      "step": 774
+    },
+    {
+      "epoch": 0.14819772444784396,
+      "grad_norm": 5.9109914004283155,
+      "learning_rate": 8.454500759296959e-07,
+      "loss": 0.3141,
+      "step": 775
+    },
+    {
+      "epoch": 0.14838894731809923,
+      "grad_norm": 6.438865146514563,
+      "learning_rate": 8.456139452465922e-07,
+      "loss": 0.6282,
+      "step": 776
+    },
+    {
+      "epoch": 0.14858017018835454,
+      "grad_norm": 6.59273633351746,
+      "learning_rate": 8.457776035276069e-07,
+      "loss": 0.9413,
+      "step": 777
+    },
+    {
+      "epoch": 0.1487713930586098,
+      "grad_norm": 3.8685225214968844,
+      "learning_rate": 8.459410513155975e-07,
+      "loss": 0.4192,
+      "step": 778
+    },
+    {
+      "epoch": 0.1489626159288651,
+      "grad_norm": 9.415903606274666,
+      "learning_rate": 8.4610428915133e-07,
+      "loss": 0.4883,
+      "step": 779
+    },
+    {
+      "epoch": 0.14915383879912036,
+      "grad_norm": 2.3006965867739515,
+      "learning_rate": 8.46267317573489e-07,
+      "loss": 0.1907,
+      "step": 780
+    },
+    {
+      "epoch": 0.14934506166937567,
+      "grad_norm": 5.578287074194811,
+      "learning_rate": 8.464301371186886e-07,
+      "loss": 0.2639,
+      "step": 781
+    },
+    {
+      "epoch": 0.14953628453963094,
+      "grad_norm": 2.245084479426817,
+      "learning_rate": 8.465927483214831e-07,
+      "loss": 0.2254,
+      "step": 782
+    },
+    {
+      "epoch": 0.14972750740988622,
+      "grad_norm": 3.533408121601068,
+      "learning_rate": 8.467551517143769e-07,
+      "loss": 0.1634,
+      "step": 783
+    },
+    {
+      "epoch": 0.1499187302801415,
+      "grad_norm": 5.9258176215306,
+      "learning_rate": 8.469173478278365e-07,
+      "loss": 0.3014,
+      "step": 784
+    },
+    {
+      "epoch": 0.1501099531503968,
+      "grad_norm": 4.7154958253513755,
+      "learning_rate": 8.470793371902985e-07,
+      "loss": 0.3888,
+      "step": 785
+    },
+    {
+      "epoch": 0.15030117602065207,
+      "grad_norm": 2.590217298722648,
+      "learning_rate": 8.472411203281822e-07,
+      "loss": 0.0627,
+      "step": 786
+    },
+    {
+      "epoch": 0.15049239889090735,
+      "grad_norm": 2.698745443137961,
+      "learning_rate": 8.474026977658988e-07,
+      "loss": 0.0479,
+      "step": 787
+    },
+    {
+      "epoch": 0.15068362176116262,
+      "grad_norm": 8.63756806585541,
+      "learning_rate": 8.475640700258616e-07,
+      "loss": 0.5746,
+      "step": 788
+    },
+    {
+      "epoch": 0.15087484463141793,
+      "grad_norm": 4.776890980435375,
+      "learning_rate": 8.47725237628496e-07,
+      "loss": 0.3227,
+      "step": 789
+    },
+    {
+      "epoch": 0.1510660675016732,
+      "grad_norm": 4.527539858713693,
+      "learning_rate": 8.478862010922507e-07,
+      "loss": 0.4555,
+      "step": 790
+    },
+    {
+      "epoch": 0.15125729037192848,
+      "grad_norm": 5.119781256115698,
+      "learning_rate": 8.480469609336057e-07,
+      "loss": 0.6263,
+      "step": 791
+    },
+    {
+      "epoch": 0.15144851324218375,
+      "grad_norm": 3.421054410740934,
+      "learning_rate": 8.482075176670845e-07,
+      "loss": 0.2978,
+      "step": 792
+    },
+    {
+      "epoch": 0.15163973611243906,
+      "grad_norm": 3.3644524106908853,
+      "learning_rate": 8.483678718052622e-07,
+      "loss": 0.308,
+      "step": 793
+    },
+    {
+      "epoch": 0.15183095898269433,
+      "grad_norm": 4.23900820918457,
+      "learning_rate": 8.485280238587763e-07,
+      "loss": 0.4162,
+      "step": 794
+    },
+    {
+      "epoch": 0.1520221818529496,
+      "grad_norm": 2.716413436586113,
+      "learning_rate": 8.486879743363357e-07,
+      "loss": 0.3288,
+      "step": 795
+    },
+    {
+      "epoch": 0.15221340472320488,
+      "grad_norm": 3.3800164135415893,
+      "learning_rate": 8.488477237447315e-07,
+      "loss": 0.1414,
+      "step": 796
+    },
+    {
+      "epoch": 0.15240462759346018,
+      "grad_norm": 5.802930933224303,
+      "learning_rate": 8.490072725888457e-07,
+      "loss": 0.2537,
+      "step": 797
+    },
+    {
+      "epoch": 0.15259585046371546,
+      "grad_norm": 1.6207372802018707,
+      "learning_rate": 8.491666213716608e-07,
+      "loss": 0.2309,
+      "step": 798
+    },
+    {
+      "epoch": 0.15278707333397074,
+      "grad_norm": 3.0359499960036995,
+      "learning_rate": 8.493257705942699e-07,
+      "loss": 0.2256,
+      "step": 799
+    },
+    {
+      "epoch": 0.15297829620422604,
+      "grad_norm": 8.57973021894528,
+      "learning_rate": 8.494847207558858e-07,
+      "loss": 0.2328,
+      "step": 800
+    },
+    {
+      "epoch": 0.15316951907448131,
+      "grad_norm": 10.154686252812857,
+      "learning_rate": 8.496434723538503e-07,
+      "loss": 0.5614,
+      "step": 801
+    },
+    {
+      "epoch": 0.1533607419447366,
+      "grad_norm": 4.832465433706562,
+      "learning_rate": 8.498020258836433e-07,
+      "loss": 0.5281,
+      "step": 802
+    },
+    {
+      "epoch": 0.15355196481499186,
+      "grad_norm": 3.554132640417855,
+      "learning_rate": 8.499603818388935e-07,
+      "loss": 0.4612,
+      "step": 803
+    },
+    {
+      "epoch": 0.15374318768524717,
+      "grad_norm": 7.049916357531723,
+      "learning_rate": 8.501185407113855e-07,
+      "loss": 0.2447,
+      "step": 804
+    },
+    {
+      "epoch": 0.15393441055550244,
+      "grad_norm": 3.6513502105783453,
+      "learning_rate": 8.502765029910705e-07,
+      "loss": 0.2633,
+      "step": 805
+    },
+    {
+      "epoch": 0.15412563342575772,
+      "grad_norm": 3.139699011791346,
+      "learning_rate": 8.504342691660747e-07,
+      "loss": 0.2731,
+      "step": 806
+    },
+    {
+      "epoch": 0.154316856296013,
+      "grad_norm": 2.854961052677372,
+      "learning_rate": 8.505918397227089e-07,
+      "loss": 0.1513,
+      "step": 807
+    },
+    {
+      "epoch": 0.1545080791662683,
+      "grad_norm": 3.105818665182121,
+      "learning_rate": 8.50749215145477e-07,
+      "loss": 0.1081,
+      "step": 808
+    },
+    {
+      "epoch": 0.15469930203652357,
+      "grad_norm": 3.2122917196833463,
+      "learning_rate": 8.50906395917085e-07,
+      "loss": 0.259,
+      "step": 809
+    },
+    {
+      "epoch": 0.15489052490677885,
+      "grad_norm": 2.95563622327139,
+      "learning_rate": 8.510633825184502e-07,
+      "loss": 0.1621,
+      "step": 810
+    },
+    {
+      "epoch": 0.15508174777703412,
+      "grad_norm": 3.8574504300762706,
+      "learning_rate": 8.512201754287095e-07,
+      "loss": 0.0681,
+      "step": 811
+    },
+    {
+      "epoch": 0.15527297064728943,
+      "grad_norm": 3.9711001188807624,
+      "learning_rate": 8.513767751252289e-07,
+      "loss": 0.1387,
+      "step": 812
+    },
+    {
+      "epoch": 0.1554641935175447,
+      "grad_norm": 6.603162943942684,
+      "learning_rate": 8.515331820836115e-07,
+      "loss": 0.184,
+      "step": 813
+    },
+    {
+      "epoch": 0.15565541638779998,
+      "grad_norm": 5.21175715507089,
+      "learning_rate": 8.516893967777068e-07,
+      "loss": 0.4824,
+      "step": 814
+    },
+    {
+      "epoch": 0.15584663925805525,
+      "grad_norm": 4.890216651906937,
+      "learning_rate": 8.518454196796187e-07,
+      "loss": 0.4987,
+      "step": 815
+    },
+    {
+      "epoch": 0.15603786212831056,
+      "grad_norm": 3.2596031043312794,
+      "learning_rate": 8.520012512597145e-07,
+      "loss": 0.2056,
+      "step": 816
+    },
+    {
+      "epoch": 0.15622908499856583,
+      "grad_norm": 2.9706435739942583,
+      "learning_rate": 8.521568919866333e-07,
+      "loss": 0.5511,
+      "step": 817
+    },
+    {
+      "epoch": 0.1564203078688211,
+      "grad_norm": 8.90956840934075,
+      "learning_rate": 8.523123423272942e-07,
+      "loss": 0.4872,
+      "step": 818
+    },
+    {
+      "epoch": 0.15661153073907638,
+      "grad_norm": 5.191979553852468,
+      "learning_rate": 8.524676027469053e-07,
+      "loss": 0.2948,
+      "step": 819
+    },
+    {
+      "epoch": 0.15680275360933169,
+      "grad_norm": 2.696109554362726,
+      "learning_rate": 8.526226737089711e-07,
+      "loss": 0.3816,
+      "step": 820
+    },
+    {
+      "epoch": 0.15699397647958696,
+      "grad_norm": 2.424401325378912,
+      "learning_rate": 8.527775556753022e-07,
+      "loss": 0.1623,
+      "step": 821
+    },
+    {
+      "epoch": 0.15718519934984224,
+      "grad_norm": 4.090134516511551,
+      "learning_rate": 8.529322491060221e-07,
+      "loss": 0.1906,
+      "step": 822
+    },
+    {
+      "epoch": 0.1573764222200975,
+      "grad_norm": 2.598109797987741,
+      "learning_rate": 8.530867544595762e-07,
+      "loss": 0.1172,
+      "step": 823
+    },
+    {
+      "epoch": 0.15756764509035281,
+      "grad_norm": 2.953014068942041,
+      "learning_rate": 8.532410721927399e-07,
+      "loss": 0.0881,
+      "step": 824
+    },
+    {
+      "epoch": 0.1577588679606081,
+      "grad_norm": 7.692262988327414,
+      "learning_rate": 8.533952027606263e-07,
+      "loss": 0.1837,
+      "step": 825
+    },
+    {
+      "epoch": 0.15795009083086337,
+      "grad_norm": 5.400890008556135,
+      "learning_rate": 8.535491466166949e-07,
+      "loss": 0.3562,
+      "step": 826
+    },
+    {
+      "epoch": 0.15814131370111864,
+      "grad_norm": 5.347469708328804,
+      "learning_rate": 8.537029042127591e-07,
+      "loss": 0.4186,
+      "step": 827
+    },
+    {
+      "epoch": 0.15833253657137394,
+      "grad_norm": 4.358401610287609,
+      "learning_rate": 8.538564759989943e-07,
+      "loss": 0.1305,
+      "step": 828
+    },
+    {
+      "epoch": 0.15852375944162922,
+      "grad_norm": 4.39547580994654,
+      "learning_rate": 8.540098624239458e-07,
+      "loss": 0.232,
+      "step": 829
+    },
+    {
+      "epoch": 0.1587149823118845,
+      "grad_norm": 4.6987503338342025,
+      "learning_rate": 8.541630639345366e-07,
+      "loss": 0.4468,
+      "step": 830
+    },
+    {
+      "epoch": 0.15890620518213977,
+      "grad_norm": 5.077608146774177,
+      "learning_rate": 8.543160809760757e-07,
+      "loss": 0.3146,
+      "step": 831
+    },
+    {
+      "epoch": 0.15909742805239507,
+      "grad_norm": 2.220026428134082,
+      "learning_rate": 8.544689139922647e-07,
+      "loss": 0.085,
+      "step": 832
+    },
+    {
+      "epoch": 0.15928865092265035,
+      "grad_norm": 3.2938826300828814,
+      "learning_rate": 8.54621563425207e-07,
+      "loss": 0.0992,
+      "step": 833
+    },
+    {
+      "epoch": 0.15947987379290562,
+      "grad_norm": 6.029102951307179,
+      "learning_rate": 8.547740297154143e-07,
+      "loss": 0.3019,
+      "step": 834
+    },
+    {
+      "epoch": 0.15967109666316093,
+      "grad_norm": 3.9129991743095185,
+      "learning_rate": 8.549263133018148e-07,
+      "loss": 0.2443,
+      "step": 835
+    },
+    {
+      "epoch": 0.1598623195334162,
+      "grad_norm": 2.9749374960095505,
+      "learning_rate": 8.550784146217607e-07,
+      "loss": 0.0561,
+      "step": 836
+    },
+    {
+      "epoch": 0.16005354240367148,
+      "grad_norm": 5.051089956669373,
+      "learning_rate": 8.55230334111036e-07,
+      "loss": 0.212,
+      "step": 837
+    },
+    {
+      "epoch": 0.16024476527392675,
+      "grad_norm": 5.493552503627086,
+      "learning_rate": 8.553820722038627e-07,
+      "loss": 0.2737,
+      "step": 838
+    },
+    {
+      "epoch": 0.16043598814418206,
+      "grad_norm": 5.308297503403851,
+      "learning_rate": 8.555336293329105e-07,
+      "loss": 0.583,
+      "step": 839
+    },
+    {
+      "epoch": 0.16062721101443733,
+      "grad_norm": 8.546015331823236,
+      "learning_rate": 8.55685005929302e-07,
+      "loss": 0.4959,
+      "step": 840
+    },
+    {
+      "epoch": 0.1608184338846926,
+      "grad_norm": 1.7159967738834814,
+      "learning_rate": 8.558362024226214e-07,
+      "loss": 0.3358,
+      "step": 841
+    },
+    {
+      "epoch": 0.16100965675494788,
+      "grad_norm": 2.5903861051878336,
+      "learning_rate": 8.559872192409212e-07,
+      "loss": 0.1672,
+      "step": 842
+    },
+    {
+      "epoch": 0.1612008796252032,
+      "grad_norm": 2.2480160125284314,
+      "learning_rate": 8.561380568107299e-07,
+      "loss": 0.1578,
+      "step": 843
+    },
+    {
+      "epoch": 0.16139210249545846,
+      "grad_norm": 4.911456219000848,
+      "learning_rate": 8.562887155570589e-07,
+      "loss": 0.5101,
+      "step": 844
+    },
+    {
+      "epoch": 0.16158332536571374,
+      "grad_norm": 3.5549023909118986,
+      "learning_rate": 8.564391959034097e-07,
+      "loss": 0.1332,
+      "step": 845
+    },
+    {
+      "epoch": 0.161774548235969,
+      "grad_norm": 2.3392654239186412,
+      "learning_rate": 8.565894982717812e-07,
+      "loss": 0.0635,
+      "step": 846
+    },
+    {
+      "epoch": 0.16196577110622432,
+      "grad_norm": 2.7080241882731513,
+      "learning_rate": 8.567396230826769e-07,
+      "loss": 0.0729,
+      "step": 847
+    },
+    {
+      "epoch": 0.1621569939764796,
+      "grad_norm": 3.1328293949727573,
+      "learning_rate": 8.568895707551115e-07,
+      "loss": 0.1141,
+      "step": 848
+    },
+    {
+      "epoch": 0.16234821684673487,
+      "grad_norm": 3.415539454778832,
+      "learning_rate": 8.570393417066184e-07,
+      "loss": 0.1999,
+      "step": 849
+    },
+    {
+      "epoch": 0.16253943971699014,
+      "grad_norm": 2.794329801128633,
+      "learning_rate": 8.571889363532564e-07,
+      "loss": 0.1363,
+      "step": 850
+    },
+    {
+      "epoch": 0.16273066258724544,
+      "grad_norm": 10.962795456283429,
+      "learning_rate": 8.573383551096167e-07,
+      "loss": 0.513,
+      "step": 851
+    },
+    {
+      "epoch": 0.16292188545750072,
+      "grad_norm": 4.101862433229884,
+      "learning_rate": 8.574875983888299e-07,
+      "loss": 0.4412,
+      "step": 852
+    },
+    {
+      "epoch": 0.163113108327756,
+      "grad_norm": 5.153904976930645,
+      "learning_rate": 8.57636666602573e-07,
+      "loss": 0.201,
+      "step": 853
+    },
+    {
+      "epoch": 0.16330433119801127,
+      "grad_norm": 7.858417065829325,
+      "learning_rate": 8.577855601610753e-07,
+      "loss": 0.3753,
+      "step": 854
+    },
+    {
+      "epoch": 0.16349555406826657,
+      "grad_norm": 2.454115163611335,
+      "learning_rate": 8.579342794731264e-07,
+      "loss": 0.3711,
+      "step": 855
+    },
+    {
+      "epoch": 0.16368677693852185,
+      "grad_norm": 5.382765726846858,
+      "learning_rate": 8.580828249460824e-07,
+      "loss": 0.8317,
+      "step": 856
+    },
+    {
+      "epoch": 0.16387799980877713,
+      "grad_norm": 3.1680753151389216,
+      "learning_rate": 8.582311969858723e-07,
+      "loss": 0.1414,
+      "step": 857
+    },
+    {
+      "epoch": 0.1640692226790324,
+      "grad_norm": 3.137869769796017,
+      "learning_rate": 8.583793959970052e-07,
+      "loss": 0.2476,
+      "step": 858
+    },
+    {
+      "epoch": 0.1642604455492877,
+      "grad_norm": 3.818246731474669,
+      "learning_rate": 8.585274223825764e-07,
+      "loss": 0.2584,
+      "step": 859
+    },
+    {
+      "epoch": 0.16445166841954298,
+      "grad_norm": 3.5065220963540553,
+      "learning_rate": 8.586752765442745e-07,
+      "loss": 0.135,
+      "step": 860
+    },
+    {
+      "epoch": 0.16464289128979825,
+      "grad_norm": 4.220255943238565,
+      "learning_rate": 8.588229588823875e-07,
+      "loss": 0.2696,
+      "step": 861
+    },
+    {
+      "epoch": 0.16483411416005353,
+      "grad_norm": 3.050430336282281,
+      "learning_rate": 8.589704697958093e-07,
+      "loss": 0.0866,
+      "step": 862
+    },
+    {
+      "epoch": 0.16502533703030883,
+      "grad_norm": 11.02762014520334,
+      "learning_rate": 8.59117809682047e-07,
+      "loss": 0.2588,
+      "step": 863
+    },
+    {
+      "epoch": 0.1652165599005641,
+      "grad_norm": 5.0723242860970625,
+      "learning_rate": 8.592649789372258e-07,
+      "loss": 0.4958,
+      "step": 864
+    },
+    {
+      "epoch": 0.16540778277081938,
+      "grad_norm": 4.150964344537788,
+      "learning_rate": 8.59411977956097e-07,
+      "loss": 0.4932,
+      "step": 865
+    },
+    {
+      "epoch": 0.16559900564107466,
+      "grad_norm": 2.7723346075425273,
+      "learning_rate": 8.595588071320432e-07,
+      "loss": 0.3292,
+      "step": 866
+    },
+    {
+      "epoch": 0.16579022851132996,
+      "grad_norm": 3.196810569374933,
+      "learning_rate": 8.597054668570851e-07,
+      "loss": 0.3981,
+      "step": 867
+    },
+    {
+      "epoch": 0.16598145138158524,
+      "grad_norm": 4.164603650856296,
+      "learning_rate": 8.598519575218877e-07,
+      "loss": 0.4997,
+      "step": 868
+    },
+    {
+      "epoch": 0.1661726742518405,
+      "grad_norm": 3.05637010833697,
+      "learning_rate": 8.599982795157668e-07,
+      "loss": 0.1928,
+      "step": 869
+    },
+    {
+      "epoch": 0.16636389712209582,
+      "grad_norm": 4.52589299784899,
+      "learning_rate": 8.601444332266945e-07,
+      "loss": 0.3583,
+      "step": 870
+    },
+    {
+      "epoch": 0.1665551199923511,
+      "grad_norm": 2.393127926306381,
+      "learning_rate": 8.602904190413062e-07,
+      "loss": 0.1238,
+      "step": 871
+    },
+    {
+      "epoch": 0.16674634286260637,
+      "grad_norm": 3.06564605425683,
+      "learning_rate": 8.604362373449062e-07,
+      "loss": 0.1125,
+      "step": 872
+    },
+    {
+      "epoch": 0.16693756573286164,
+      "grad_norm": 2.3688826833317256,
+      "learning_rate": 8.60581888521474e-07,
+      "loss": 0.1393,
+      "step": 873
+    },
+    {
+      "epoch": 0.16712878860311695,
+      "grad_norm": 5.0085498189658155,
+      "learning_rate": 8.607273729536706e-07,
+      "loss": 0.1918,
+      "step": 874
+    },
+    {
+      "epoch": 0.16732001147337222,
+      "grad_norm": 11.067727098636052,
+      "learning_rate": 8.608726910228437e-07,
+      "loss": 0.1988,
+      "step": 875
+    },
+    {
+      "epoch": 0.1675112343436275,
+      "grad_norm": 5.15125212771384,
+      "learning_rate": 8.610178431090349e-07,
+      "loss": 0.4859,
+      "step": 876
+    },
+    {
+      "epoch": 0.16770245721388277,
+      "grad_norm": 4.626741854545013,
+      "learning_rate": 8.611628295909844e-07,
+      "loss": 0.553,
+      "step": 877
+    },
+    {
+      "epoch": 0.16789368008413807,
+      "grad_norm": 4.7568007522500295,
+      "learning_rate": 8.613076508461382e-07,
+      "loss": 0.6153,
+      "step": 878
+    },
+    {
+      "epoch": 0.16808490295439335,
+      "grad_norm": 2.514907545054999,
+      "learning_rate": 8.614523072506528e-07,
+      "loss": 0.1246,
+      "step": 879
+    },
+    {
+      "epoch": 0.16827612582464863,
+      "grad_norm": 2.7940698120581278,
+      "learning_rate": 8.61596799179402e-07,
+      "loss": 0.3403,
+      "step": 880
+    },
+    {
+      "epoch": 0.1684673486949039,
+      "grad_norm": 5.206032979959668,
+      "learning_rate": 8.617411270059822e-07,
+      "loss": 0.337,
+      "step": 881
+    },
+    {
+      "epoch": 0.1686585715651592,
+      "grad_norm": 3.615518864794115,
+      "learning_rate": 8.618852911027183e-07,
+      "loss": 0.2661,
+      "step": 882
+    },
+    {
+      "epoch": 0.16884979443541448,
+      "grad_norm": 3.111652249624822,
+      "learning_rate": 8.620292918406695e-07,
+      "loss": 0.1383,
+      "step": 883
+    },
+    {
+      "epoch": 0.16904101730566976,
+      "grad_norm": 5.633404535298809,
+      "learning_rate": 8.621731295896353e-07,
+      "loss": 0.4742,
+      "step": 884
+    },
+    {
+      "epoch": 0.16923224017592503,
+      "grad_norm": 3.7223115423186455,
+      "learning_rate": 8.623168047181604e-07,
+      "loss": 0.2966,
+      "step": 885
+    },
+    {
+      "epoch": 0.16942346304618033,
+      "grad_norm": 3.563206786598968,
+      "learning_rate": 8.624603175935417e-07,
+      "loss": 0.1045,
+      "step": 886
+    },
+    {
+      "epoch": 0.1696146859164356,
+      "grad_norm": 3.974607455466367,
+      "learning_rate": 8.626036685818322e-07,
+      "loss": 0.068,
+      "step": 887
+    },
+    {
+      "epoch": 0.16980590878669088,
+      "grad_norm": 6.138882532974719,
+      "learning_rate": 8.627468580478482e-07,
+      "loss": 0.1926,
+      "step": 888
+    },
+    {
+      "epoch": 0.16999713165694616,
+      "grad_norm": 10.19268554518616,
+      "learning_rate": 8.62889886355174e-07,
+      "loss": 0.7554,
+      "step": 889
+    },
+    {
+      "epoch": 0.17018835452720146,
+      "grad_norm": 3.5214933119166965,
+      "learning_rate": 8.630327538661676e-07,
+      "loss": 0.3336,
+      "step": 890
+    },
+    {
+      "epoch": 0.17037957739745674,
+      "grad_norm": 5.5042585012478815,
+      "learning_rate": 8.631754609419664e-07,
+      "loss": 0.49,
+      "step": 891
+    },
+    {
+      "epoch": 0.170570800267712,
+      "grad_norm": 3.1293652372853633,
+      "learning_rate": 8.633180079424923e-07,
+      "loss": 0.5471,
+      "step": 892
+    },
+    {
+      "epoch": 0.1707620231379673,
+      "grad_norm": 3.593783701863033,
+      "learning_rate": 8.634603952264574e-07,
+      "loss": 0.5304,
+      "step": 893
+    },
+    {
+      "epoch": 0.1709532460082226,
+      "grad_norm": 2.8376402563981653,
+      "learning_rate": 8.636026231513699e-07,
+      "loss": 0.3289,
+      "step": 894
+    },
+    {
+      "epoch": 0.17114446887847787,
+      "grad_norm": 3.461260984988155,
+      "learning_rate": 8.637446920735382e-07,
+      "loss": 0.1412,
+      "step": 895
+    },
+    {
+      "epoch": 0.17133569174873314,
+      "grad_norm": 5.446192491046945,
+      "learning_rate": 8.638866023480777e-07,
+      "loss": 0.3277,
+      "step": 896
+    },
+    {
+      "epoch": 0.17152691461898842,
+      "grad_norm": 4.5984792434607185,
+      "learning_rate": 8.64028354328915e-07,
+      "loss": 0.1942,
+      "step": 897
+    },
+    {
+      "epoch": 0.17171813748924372,
+      "grad_norm": 6.007134963623791,
+      "learning_rate": 8.641699483687939e-07,
+      "loss": 0.1413,
+      "step": 898
+    },
+    {
+      "epoch": 0.171909360359499,
+      "grad_norm": 4.073286321049267,
+      "learning_rate": 8.643113848192803e-07,
+      "loss": 0.0917,
+      "step": 899
+    },
+    {
+      "epoch": 0.17210058322975427,
+      "grad_norm": 8.764753274529689,
+      "learning_rate": 8.644526640307677e-07,
+      "loss": 0.1761,
+      "step": 900
+    },
+    {
+      "epoch": 0.17229180610000955,
+      "grad_norm": 3.912446379370298,
+      "learning_rate": 8.64593786352482e-07,
+      "loss": 0.3151,
+      "step": 901
+    },
+    {
+      "epoch": 0.17248302897026485,
+      "grad_norm": 6.502794398676239,
+      "learning_rate": 8.647347521324875e-07,
+      "loss": 0.3713,
+      "step": 902
+    },
+    {
+      "epoch": 0.17267425184052013,
+      "grad_norm": 3.7803300731597673,
+      "learning_rate": 8.648755617176907e-07,
+      "loss": 0.3129,
+      "step": 903
+    },
+    {
+      "epoch": 0.1728654747107754,
+      "grad_norm": 2.139721310731296,
+      "learning_rate": 8.65016215453847e-07,
+      "loss": 0.1674,
+      "step": 904
+    },
+    {
+      "epoch": 0.1730566975810307,
+      "grad_norm": 3.0769328795790414,
+      "learning_rate": 8.651567136855646e-07,
+      "loss": 0.2738,
+      "step": 905
+    },
+    {
+      "epoch": 0.17324792045128598,
+      "grad_norm": 4.385421765008108,
+      "learning_rate": 8.652970567563101e-07,
+      "loss": 0.4744,
+      "step": 906
+    },
+    {
+      "epoch": 0.17343914332154126,
+      "grad_norm": 4.172124162298477,
+      "learning_rate": 8.654372450084132e-07,
+      "loss": 0.2156,
+      "step": 907
+    },
+    {
+      "epoch": 0.17363036619179653,
+      "grad_norm": 4.336836987922702,
+      "learning_rate": 8.655772787830724e-07,
+      "loss": 0.3622,
+      "step": 908
+    },
+    {
+      "epoch": 0.17382158906205183,
+      "grad_norm": 3.2502467722005655,
+      "learning_rate": 8.657171584203589e-07,
+      "loss": 0.4732,
+      "step": 909
+    },
+    {
+      "epoch": 0.1740128119323071,
+      "grad_norm": 5.2469064590029015,
+      "learning_rate": 8.658568842592227e-07,
+      "loss": 0.3476,
+      "step": 910
+    },
+    {
+      "epoch": 0.17420403480256239,
+      "grad_norm": 4.406998374639167,
+      "learning_rate": 8.659964566374968e-07,
+      "loss": 0.1819,
+      "step": 911
+    },
+    {
+      "epoch": 0.17439525767281766,
+      "grad_norm": 2.1039016355286146,
+      "learning_rate": 8.66135875891902e-07,
+      "loss": 0.0801,
+      "step": 912
+    },
+    {
+      "epoch": 0.17458648054307296,
+      "grad_norm": 9.558584970112436,
+      "learning_rate": 8.662751423580529e-07,
+      "loss": 0.3169,
+      "step": 913
+    },
+    {
+      "epoch": 0.17477770341332824,
+      "grad_norm": 4.950853278087762,
+      "learning_rate": 8.66414256370461e-07,
+      "loss": 0.3492,
+      "step": 914
+    },
+    {
+      "epoch": 0.17496892628358351,
+      "grad_norm": 4.497566836901101,
+      "learning_rate": 8.665532182625408e-07,
+      "loss": 0.569,
+      "step": 915
+    },
+    {
+      "epoch": 0.1751601491538388,
+      "grad_norm": 3.727158984544305,
+      "learning_rate": 8.666920283666144e-07,
+      "loss": 0.1821,
+      "step": 916
+    },
+    {
+      "epoch": 0.1753513720240941,
+      "grad_norm": 3.8381347578602067,
+      "learning_rate": 8.66830687013916e-07,
+      "loss": 0.4188,
+      "step": 917
+    },
+    {
+      "epoch": 0.17554259489434937,
+      "grad_norm": 4.508995178680313,
+      "learning_rate": 8.669691945345965e-07,
+      "loss": 0.2885,
+      "step": 918
+    },
+    {
+      "epoch": 0.17573381776460464,
+      "grad_norm": 3.333869382353446,
+      "learning_rate": 8.671075512577286e-07,
+      "loss": 0.1343,
+      "step": 919
+    },
+    {
+      "epoch": 0.17592504063485992,
+      "grad_norm": 3.292820466708831,
+      "learning_rate": 8.672457575113118e-07,
+      "loss": 0.2396,
+      "step": 920
+    },
+    {
+      "epoch": 0.17611626350511522,
+      "grad_norm": 4.650634861748401,
+      "learning_rate": 8.673838136222758e-07,
+      "loss": 0.4561,
+      "step": 921
+    },
+    {
+      "epoch": 0.1763074863753705,
+      "grad_norm": 5.982677567585065,
+      "learning_rate": 8.675217199164869e-07,
+      "loss": 0.1878,
+      "step": 922
+    },
+    {
+      "epoch": 0.17649870924562577,
+      "grad_norm": 3.4464544017171086,
+      "learning_rate": 8.676594767187506e-07,
+      "loss": 0.2956,
+      "step": 923
+    },
+    {
+      "epoch": 0.17668993211588105,
+      "grad_norm": 3.892185005668046,
+      "learning_rate": 8.677970843528182e-07,
+      "loss": 0.0893,
+      "step": 924
+    },
+    {
+      "epoch": 0.17688115498613635,
+      "grad_norm": 6.750848787418212,
+      "learning_rate": 8.6793454314139e-07,
+      "loss": 0.3371,
+      "step": 925
+    },
+    {
+      "epoch": 0.17707237785639163,
+      "grad_norm": 6.759533119556762,
+      "learning_rate": 8.680718534061202e-07,
+      "loss": 0.6884,
+      "step": 926
+    },
+    {
+      "epoch": 0.1772636007266469,
+      "grad_norm": 4.900944459162629,
+      "learning_rate": 8.682090154676216e-07,
+      "loss": 0.5866,
+      "step": 927
+    },
+    {
+      "epoch": 0.17745482359690218,
+      "grad_norm": 2.6848808655690077,
+      "learning_rate": 8.683460296454702e-07,
+      "loss": 0.3614,
+      "step": 928
+    },
+    {
+      "epoch": 0.17764604646715748,
+      "grad_norm": 1.9337029320303187,
+      "learning_rate": 8.68482896258209e-07,
+      "loss": 0.2472,
+      "step": 929
+    },
+    {
+      "epoch": 0.17783726933741276,
+      "grad_norm": 3.1188126918936154,
+      "learning_rate": 8.686196156233533e-07,
+      "loss": 0.1616,
+      "step": 930
+    },
+    {
+      "epoch": 0.17802849220766803,
+      "grad_norm": 3.4788392573429108,
+      "learning_rate": 8.687561880573946e-07,
+      "loss": 0.1446,
+      "step": 931
+    },
+    {
+      "epoch": 0.1782197150779233,
+      "grad_norm": 4.349587010915789,
+      "learning_rate": 8.688926138758046e-07,
+      "loss": 0.5091,
+      "step": 932
+    },
+    {
+      "epoch": 0.1784109379481786,
+      "grad_norm": 3.8068763656000004,
+      "learning_rate": 8.69028893393041e-07,
+      "loss": 0.304,
+      "step": 933
+    },
+    {
+      "epoch": 0.17860216081843389,
+      "grad_norm": 2.8758983037951764,
+      "learning_rate": 8.691650269225502e-07,
+      "loss": 0.3466,
+      "step": 934
+    },
+    {
+      "epoch": 0.17879338368868916,
+      "grad_norm": 3.559196546734751,
+      "learning_rate": 8.693010147767726e-07,
+      "loss": 0.2711,
+      "step": 935
+    },
+    {
+      "epoch": 0.17898460655894444,
+      "grad_norm": 3.533067353751901,
+      "learning_rate": 8.694368572671465e-07,
+      "loss": 0.1283,
+      "step": 936
+    },
+    {
+      "epoch": 0.17917582942919974,
+      "grad_norm": 3.088570005513082,
+      "learning_rate": 8.695725547041129e-07,
+      "loss": 0.1104,
+      "step": 937
+    },
+    {
+      "epoch": 0.17936705229945502,
+      "grad_norm": 9.976229071366344,
+      "learning_rate": 8.697081073971192e-07,
+      "loss": 0.5461,
+      "step": 938
+    },
+    {
+      "epoch": 0.1795582751697103,
+      "grad_norm": 4.50069422135309,
+      "learning_rate": 8.698435156546233e-07,
+      "loss": 0.5042,
+      "step": 939
+    },
+    {
+      "epoch": 0.17974949803996557,
+      "grad_norm": 2.9778178614042603,
+      "learning_rate": 8.699787797840987e-07,
+      "loss": 0.3605,
+      "step": 940
+    },
+    {
+      "epoch": 0.17994072091022087,
+      "grad_norm": 2.370100388312168,
+      "learning_rate": 8.701139000920376e-07,
+      "loss": 0.3331,
+      "step": 941
+    },
+    {
+      "epoch": 0.18013194378047614,
+      "grad_norm": 4.056836214892448,
+      "learning_rate": 8.70248876883956e-07,
+      "loss": 0.4944,
+      "step": 942
+    },
+    {
+      "epoch": 0.18032316665073142,
+      "grad_norm": 2.548068837456131,
+      "learning_rate": 8.703837104643972e-07,
+      "loss": 0.1389,
+      "step": 943
+    },
+    {
+      "epoch": 0.18051438952098672,
+      "grad_norm": 2.314267282227545,
+      "learning_rate": 8.705184011369362e-07,
+      "loss": 0.2098,
+      "step": 944
+    },
+    {
+      "epoch": 0.180705612391242,
+      "grad_norm": 2.3574830610651754,
+      "learning_rate": 8.706529492041838e-07,
+      "loss": 0.1666,
+      "step": 945
+    },
+    {
+      "epoch": 0.18089683526149727,
+      "grad_norm": 3.5939357543705435,
+      "learning_rate": 8.707873549677907e-07,
+      "loss": 0.2681,
+      "step": 946
+    },
+    {
+      "epoch": 0.18108805813175255,
+      "grad_norm": 3.459588677096927,
+      "learning_rate": 8.709216187284513e-07,
+      "loss": 0.1073,
+      "step": 947
+    },
+    {
+      "epoch": 0.18127928100200785,
+      "grad_norm": 5.0977716776323465,
+      "learning_rate": 8.710557407859081e-07,
+      "loss": 0.384,
+      "step": 948
+    },
+    {
+      "epoch": 0.18147050387226313,
+      "grad_norm": 2.7978743999940545,
+      "learning_rate": 8.711897214389556e-07,
+      "loss": 0.0943,
+      "step": 949
+    },
+    {
+      "epoch": 0.1816617267425184,
+      "grad_norm": 8.431668449180984,
+      "learning_rate": 8.713235609854439e-07,
+      "loss": 0.2383,
+      "step": 950
+    },
+    {
+      "epoch": 0.18185294961277368,
+      "grad_norm": 8.218872069405979,
+      "learning_rate": 8.714572597222834e-07,
+      "loss": 0.4696,
+      "step": 951
+    },
+    {
+      "epoch": 0.18204417248302898,
+      "grad_norm": 4.846683918022786,
+      "learning_rate": 8.715908179454482e-07,
+      "loss": 0.6473,
+      "step": 952
+    },
+    {
+      "epoch": 0.18223539535328426,
+      "grad_norm": 5.239105364598264,
+      "learning_rate": 8.717242359499804e-07,
+      "loss": 0.7599,
+      "step": 953
+    },
+    {
+      "epoch": 0.18242661822353953,
+      "grad_norm": 3.9584950999617137,
+      "learning_rate": 8.718575140299933e-07,
+      "loss": 0.2287,
+      "step": 954
+    },
+    {
+      "epoch": 0.1826178410937948,
+      "grad_norm": 2.70536490892935,
+      "learning_rate": 8.719906524786763e-07,
+      "loss": 0.1838,
+      "step": 955
+    },
+    {
+      "epoch": 0.1828090639640501,
+      "grad_norm": 4.23751568186511,
+      "learning_rate": 8.721236515882981e-07,
+      "loss": 0.3244,
+      "step": 956
+    },
+    {
+      "epoch": 0.1830002868343054,
+      "grad_norm": 3.423377556321745,
+      "learning_rate": 8.722565116502108e-07,
+      "loss": 0.2515,
+      "step": 957
+    },
+    {
+      "epoch": 0.18319150970456066,
+      "grad_norm": 3.1198393362200054,
+      "learning_rate": 8.723892329548533e-07,
+      "loss": 0.1676,
+      "step": 958
+    },
+    {
+      "epoch": 0.18338273257481594,
+      "grad_norm": 3.206615797446979,
+      "learning_rate": 8.725218157917558e-07,
+      "loss": 0.2686,
+      "step": 959
+    },
+    {
+      "epoch": 0.18357395544507124,
+      "grad_norm": 3.4238576501507905,
+      "learning_rate": 8.726542604495434e-07,
+      "loss": 0.0789,
+      "step": 960
+    },
+    {
+      "epoch": 0.18376517831532652,
+      "grad_norm": 2.729403701467094,
+      "learning_rate": 8.727865672159391e-07,
+      "loss": 0.0808,
+      "step": 961
+    },
+    {
+      "epoch": 0.1839564011855818,
+      "grad_norm": 3.8832766833013928,
+      "learning_rate": 8.729187363777689e-07,
+      "loss": 0.0871,
+      "step": 962
+    },
+    {
+      "epoch": 0.18414762405583707,
+      "grad_norm": 8.328049879468894,
+      "learning_rate": 8.730507682209642e-07,
+      "loss": 0.4171,
+      "step": 963
+    },
+    {
+      "epoch": 0.18433884692609237,
+      "grad_norm": 4.445938897540548,
+      "learning_rate": 8.731826630305665e-07,
+      "loss": 0.5968,
+      "step": 964
+    },
+    {
+      "epoch": 0.18453006979634765,
+      "grad_norm": 4.604229036256835,
+      "learning_rate": 8.733144210907305e-07,
+      "loss": 0.6361,
+      "step": 965
+    },
+    {
+      "epoch": 0.18472129266660292,
+      "grad_norm": 3.0408001714981863,
+      "learning_rate": 8.73446042684728e-07,
+      "loss": 0.2536,
+      "step": 966
+    },
+    {
+      "epoch": 0.1849125155368582,
+      "grad_norm": 4.207877383066533,
+      "learning_rate": 8.735775280949515e-07,
+      "loss": 0.2919,
+      "step": 967
+    },
+    {
+      "epoch": 0.1851037384071135,
+      "grad_norm": 4.389437586183228,
+      "learning_rate": 8.737088776029182e-07,
+      "loss": 0.5364,
+      "step": 968
+    },
+    {
+      "epoch": 0.18529496127736877,
+      "grad_norm": 3.7031842319564046,
+      "learning_rate": 8.738400914892728e-07,
+      "loss": 0.2656,
+      "step": 969
+    },
+    {
+      "epoch": 0.18548618414762405,
+      "grad_norm": 2.972020806908454,
+      "learning_rate": 8.739711700337916e-07,
+      "loss": 0.1613,
+      "step": 970
+    },
+    {
+      "epoch": 0.18567740701787933,
+      "grad_norm": 3.0615605061551734,
+      "learning_rate": 8.741021135153865e-07,
+      "loss": 0.1762,
+      "step": 971
+    },
+    {
+      "epoch": 0.18586862988813463,
+      "grad_norm": 3.6049941919203277,
+      "learning_rate": 8.742329222121077e-07,
+      "loss": 0.2217,
+      "step": 972
+    },
+    {
+      "epoch": 0.1860598527583899,
+      "grad_norm": 6.2146497848424405,
+      "learning_rate": 8.743635964011479e-07,
+      "loss": 0.1865,
+      "step": 973
+    },
+    {
+      "epoch": 0.18625107562864518,
+      "grad_norm": 2.8138890544978756,
+      "learning_rate": 8.744941363588456e-07,
+      "loss": 0.0728,
+      "step": 974
+    },
+    {
+      "epoch": 0.18644229849890046,
+      "grad_norm": 5.394559182019467,
+      "learning_rate": 8.746245423606883e-07,
+      "loss": 0.1599,
+      "step": 975
+    },
+    {
+      "epoch": 0.18663352136915576,
+      "grad_norm": 6.13252387855879,
+      "learning_rate": 8.747548146813165e-07,
+      "loss": 0.4884,
+      "step": 976
+    },
+    {
+      "epoch": 0.18682474423941103,
+      "grad_norm": 4.620267766814013,
+      "learning_rate": 8.748849535945271e-07,
+      "loss": 0.7369,
+      "step": 977
+    },
+    {
+      "epoch": 0.1870159671096663,
+      "grad_norm": 3.879098754900235,
+      "learning_rate": 8.750149593732764e-07,
+      "loss": 0.407,
+      "step": 978
+    },
+    {
+      "epoch": 0.1872071899799216,
+      "grad_norm": 6.506588164674151,
+      "learning_rate": 8.751448322896839e-07,
+      "loss": 0.3573,
+      "step": 979
+    },
+    {
+      "epoch": 0.1873984128501769,
+      "grad_norm": 4.600190714324167,
+      "learning_rate": 8.752745726150357e-07,
+      "loss": 0.3872,
+      "step": 980
+    },
+    {
+      "epoch": 0.18758963572043216,
+      "grad_norm": 5.596592697615045,
+      "learning_rate": 8.75404180619788e-07,
+      "loss": 0.6989,
+      "step": 981
+    },
+    {
+      "epoch": 0.18778085859068744,
+      "grad_norm": 4.495631110469463,
+      "learning_rate": 8.755336565735703e-07,
+      "loss": 0.1475,
+      "step": 982
+    },
+    {
+      "epoch": 0.18797208146094274,
+      "grad_norm": 3.1147194176652975,
+      "learning_rate": 8.756630007451884e-07,
+      "loss": 0.1308,
+      "step": 983
+    },
+    {
+      "epoch": 0.18816330433119802,
+      "grad_norm": 3.4663137867640526,
+      "learning_rate": 8.757922134026288e-07,
+      "loss": 0.094,
+      "step": 984
+    },
+    {
+      "epoch": 0.1883545272014533,
+      "grad_norm": 4.292013185208876,
+      "learning_rate": 8.759212948130608e-07,
+      "loss": 0.2054,
+      "step": 985
+    },
+    {
+      "epoch": 0.18854575007170857,
+      "grad_norm": 3.48697500344964,
+      "learning_rate": 8.760502452428407e-07,
+      "loss": 0.0754,
+      "step": 986
+    },
+    {
+      "epoch": 0.18873697294196387,
+      "grad_norm": 4.591277410735394,
+      "learning_rate": 8.761790649575149e-07,
+      "loss": 0.0662,
+      "step": 987
+    },
+    {
+      "epoch": 0.18892819581221915,
+      "grad_norm": 7.420465275245418,
+      "learning_rate": 8.763077542218229e-07,
+      "loss": 0.5108,
+      "step": 988
+    },
+    {
+      "epoch": 0.18911941868247442,
+      "grad_norm": 3.8565898529113904,
+      "learning_rate": 8.764363132997006e-07,
+      "loss": 0.2265,
+      "step": 989
+    },
+    {
+      "epoch": 0.1893106415527297,
+      "grad_norm": 2.6777321490633983,
+      "learning_rate": 8.765647424542838e-07,
+      "loss": 0.19,
+      "step": 990
+    },
+    {
+      "epoch": 0.189501864422985,
+      "grad_norm": 4.018349521076428,
+      "learning_rate": 8.766930419479117e-07,
+      "loss": 0.3072,
+      "step": 991
+    },
+    {
+      "epoch": 0.18969308729324028,
+      "grad_norm": 5.216905279257907,
+      "learning_rate": 8.76821212042129e-07,
+      "loss": 0.5055,
+      "step": 992
+    },
+    {
+      "epoch": 0.18988431016349555,
+      "grad_norm": 3.3496290200772316,
+      "learning_rate": 8.769492529976907e-07,
+      "loss": 0.3047,
+      "step": 993
+    },
+    {
+      "epoch": 0.19007553303375083,
+      "grad_norm": 2.018688269459688,
+      "learning_rate": 8.770771650745637e-07,
+      "loss": 0.1683,
+      "step": 994
+    },
+    {
+      "epoch": 0.19026675590400613,
+      "grad_norm": 2.1715382925614715,
+      "learning_rate": 8.772049485319309e-07,
+      "loss": 0.1396,
+      "step": 995
+    },
+    {
+      "epoch": 0.1904579787742614,
+      "grad_norm": 4.314919939944978,
+      "learning_rate": 8.773326036281943e-07,
+      "loss": 0.204,
+      "step": 996
+    },
+    {
+      "epoch": 0.19064920164451668,
+      "grad_norm": 3.7547036236096534,
+      "learning_rate": 8.774601306209778e-07,
+      "loss": 0.1489,
+      "step": 997
+    },
+    {
+      "epoch": 0.19084042451477196,
+      "grad_norm": 7.651842740586098,
+      "learning_rate": 8.775875297671306e-07,
+      "loss": 0.2831,
+      "step": 998
+    },
+    {
+      "epoch": 0.19103164738502726,
+      "grad_norm": 5.2199446173174415,
+      "learning_rate": 8.777148013227301e-07,
+      "loss": 0.1095,
+      "step": 999
+    },
+    {
+      "epoch": 0.19122287025528253,
+      "grad_norm": 5.304097430708015,
+      "learning_rate": 8.77841945543085e-07,
+      "loss": 0.2341,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19122287025528253,
+      "eval_runtime": 728.573,
+      "eval_samples_per_second": 2.105,
+      "eval_steps_per_second": 0.527,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1914140931255378,
+      "grad_norm": 6.070674458320611,
+      "learning_rate": 8.779689626827389e-07,
+      "loss": 0.6226,
+      "step": 1001
+    },
+    {
+      "epoch": 0.19160531599579309,
+      "grad_norm": 4.523990893731704,
+      "learning_rate": 8.780958529954723e-07,
+      "loss": 0.3278,
+      "step": 1002
+    },
+    {
+      "epoch": 0.1917965388660484,
+      "grad_norm": 3.6723818165438535,
+      "learning_rate": 8.782226167343068e-07,
+      "loss": 0.4181,
+      "step": 1003
+    },
+    {
+      "epoch": 0.19198776173630366,
+      "grad_norm": 5.1827120458685165,
+      "learning_rate": 8.783492541515071e-07,
+      "loss": 0.1841,
+      "step": 1004
+    },
+    {
+      "epoch": 0.19217898460655894,
+      "grad_norm": 4.566315192366287,
+      "learning_rate": 8.784757654985848e-07,
+      "loss": 0.4202,
+      "step": 1005
+    },
+    {
+      "epoch": 0.19237020747681421,
+      "grad_norm": 4.24230237492657,
+      "learning_rate": 8.786021510263011e-07,
+      "loss": 0.3539,
+      "step": 1006
+    },
+    {
+      "epoch": 0.19256143034706952,
+      "grad_norm": 4.197025908187728,
+      "learning_rate": 8.787284109846696e-07,
+      "loss": 0.4031,
+      "step": 1007
+    },
+    {
+      "epoch": 0.1927526532173248,
+      "grad_norm": 4.288963108266074,
+      "learning_rate": 8.788545456229596e-07,
+      "loss": 0.0682,
+      "step": 1008
+    },
+    {
+      "epoch": 0.19294387608758007,
+      "grad_norm": 4.160887232214315,
+      "learning_rate": 8.789805551896987e-07,
+      "loss": 0.1929,
+      "step": 1009
+    },
+    {
+      "epoch": 0.19313509895783534,
+      "grad_norm": 6.6911579716234995,
+      "learning_rate": 8.791064399326763e-07,
+      "loss": 0.1993,
+      "step": 1010
+    },
+    {
+      "epoch": 0.19332632182809065,
+      "grad_norm": 3.503521237861696,
+      "learning_rate": 8.792322000989458e-07,
+      "loss": 0.1249,
+      "step": 1011
+    },
+    {
+      "epoch": 0.19351754469834592,
+      "grad_norm": 5.945923301409967,
+      "learning_rate": 8.793578359348281e-07,
+      "loss": 0.1048,
+      "step": 1012
+    },
+    {
+      "epoch": 0.1937087675686012,
+      "grad_norm": 5.847742032917058,
+      "learning_rate": 8.794833476859138e-07,
+      "loss": 0.1712,
+      "step": 1013
+    },
+    {
+      "epoch": 0.1938999904388565,
+      "grad_norm": 4.972813126575913,
+      "learning_rate": 8.796087355970673e-07,
+      "loss": 0.2406,
+      "step": 1014
+    },
+    {
+      "epoch": 0.19409121330911178,
+      "grad_norm": 4.6427067155887265,
+      "learning_rate": 8.797339999124281e-07,
+      "loss": 0.3925,
+      "step": 1015
+    },
+    {
+      "epoch": 0.19428243617936705,
+      "grad_norm": 3.7361652759531774,
+      "learning_rate": 8.798591408754154e-07,
+      "loss": 0.4331,
+      "step": 1016
+    },
+    {
+      "epoch": 0.19447365904962233,
+      "grad_norm": 3.8954291040478566,
+      "learning_rate": 8.799841587287289e-07,
+      "loss": 0.1918,
+      "step": 1017
+    },
+    {
+      "epoch": 0.19466488191987763,
+      "grad_norm": 2.813674342503777,
+      "learning_rate": 8.801090537143535e-07,
+      "loss": 0.2559,
+      "step": 1018
+    },
+    {
+      "epoch": 0.1948561047901329,
+      "grad_norm": 4.450463500154893,
+      "learning_rate": 8.802338260735613e-07,
+      "loss": 0.5023,
+      "step": 1019
+    },
+    {
+      "epoch": 0.19504732766038818,
+      "grad_norm": 2.955048595616257,
+      "learning_rate": 8.803584760469138e-07,
+      "loss": 0.2139,
+      "step": 1020
+    },
+    {
+      "epoch": 0.19523855053064346,
+      "grad_norm": 3.1691648433685744,
+      "learning_rate": 8.804830038742663e-07,
+      "loss": 0.1678,
+      "step": 1021
+    },
+    {
+      "epoch": 0.19542977340089876,
+      "grad_norm": 6.319356688098561,
+      "learning_rate": 8.806074097947685e-07,
+      "loss": 0.4917,
+      "step": 1022
+    },
+    {
+      "epoch": 0.19562099627115404,
+      "grad_norm": 3.919161279441986,
+      "learning_rate": 8.807316940468695e-07,
+      "loss": 0.0764,
+      "step": 1023
+    },
+    {
+      "epoch": 0.1958122191414093,
+      "grad_norm": 3.3072708890167135,
+      "learning_rate": 8.808558568683191e-07,
+      "loss": 0.1027,
+      "step": 1024
+    },
+    {
+      "epoch": 0.19600344201166459,
+      "grad_norm": 11.138667251610034,
+      "learning_rate": 8.809798984961705e-07,
+      "loss": 0.3142,
+      "step": 1025
+    },
+    {
+      "epoch": 0.1961946648819199,
+      "grad_norm": 8.385173937836557,
+      "learning_rate": 8.81103819166784e-07,
+      "loss": 0.2826,
+      "step": 1026
+    },
+    {
+      "epoch": 0.19638588775217516,
+      "grad_norm": 4.391379294985433,
+      "learning_rate": 8.812276191158287e-07,
+      "loss": 0.6458,
+      "step": 1027
+    },
+    {
+      "epoch": 0.19657711062243044,
+      "grad_norm": 3.746364484278066,
+      "learning_rate": 8.813512985782862e-07,
+      "loss": 0.3799,
+      "step": 1028
+    },
+    {
+      "epoch": 0.19676833349268572,
+      "grad_norm": 3.9210891240212926,
+      "learning_rate": 8.814748577884519e-07,
+      "loss": 0.4056,
+      "step": 1029
+    },
+    {
+      "epoch": 0.19695955636294102,
+      "grad_norm": 7.130110397019224,
+      "learning_rate": 8.815982969799391e-07,
+      "loss": 0.3978,
+      "step": 1030
+    },
+    {
+      "epoch": 0.1971507792331963,
+      "grad_norm": 3.867043355702179,
+      "learning_rate": 8.817216163856807e-07,
+      "loss": 0.1473,
+      "step": 1031
+    },
+    {
+      "epoch": 0.19734200210345157,
+      "grad_norm": 4.0540081770766925,
+      "learning_rate": 8.818448162379321e-07,
+      "loss": 0.3711,
+      "step": 1032
+    },
+    {
+      "epoch": 0.19753322497370684,
+      "grad_norm": 3.9877535987499604,
+      "learning_rate": 8.81967896768274e-07,
+      "loss": 0.3568,
+      "step": 1033
+    },
+    {
+      "epoch": 0.19772444784396215,
+      "grad_norm": 3.3125374270070167,
+      "learning_rate": 8.820908582076149e-07,
+      "loss": 0.1745,
+      "step": 1034
+    },
+    {
+      "epoch": 0.19791567071421742,
+      "grad_norm": 3.5043357152150305,
+      "learning_rate": 8.822137007861936e-07,
+      "loss": 0.1367,
+      "step": 1035
+    },
+    {
+      "epoch": 0.1981068935844727,
+      "grad_norm": 5.265965662953312,
+      "learning_rate": 8.82336424733582e-07,
+      "loss": 0.1142,
+      "step": 1036
+    },
+    {
+      "epoch": 0.19829811645472797,
+      "grad_norm": 4.664573449752898,
+      "learning_rate": 8.824590302786871e-07,
+      "loss": 0.1458,
+      "step": 1037
+    },
+    {
+      "epoch": 0.19848933932498328,
+      "grad_norm": 6.3788886084607554,
+      "learning_rate": 8.825815176497546e-07,
+      "loss": 0.2975,
+      "step": 1038
+    },
+    {
+      "epoch": 0.19868056219523855,
+      "grad_norm": 4.967813463539116,
+      "learning_rate": 8.827038870743705e-07,
+      "loss": 0.7499,
+      "step": 1039
+    },
+    {
+      "epoch": 0.19887178506549383,
+      "grad_norm": 4.436237424229102,
+      "learning_rate": 8.82826138779464e-07,
+      "loss": 0.3742,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1990630079357491,
+      "grad_norm": 3.3467504996429525,
+      "learning_rate": 8.829482729913101e-07,
+      "loss": 0.4704,
+      "step": 1041
+    },
+    {
+      "epoch": 0.1992542308060044,
+      "grad_norm": 2.3387767670739406,
+      "learning_rate": 8.83070289935532e-07,
+      "loss": 0.3258,
+      "step": 1042
+    },
+    {
+      "epoch": 0.19944545367625968,
+      "grad_norm": 2.4167257389434424,
+      "learning_rate": 8.831921898371036e-07,
+      "loss": 0.2708,
+      "step": 1043
+    },
+    {
+      "epoch": 0.19963667654651496,
+      "grad_norm": 2.9507140960202562,
+      "learning_rate": 8.833139729203521e-07,
+      "loss": 0.3461,
+      "step": 1044
+    },
+    {
+      "epoch": 0.19982789941677023,
+      "grad_norm": 1.795543143892563,
+      "learning_rate": 8.834356394089602e-07,
+      "loss": 0.1353,
+      "step": 1045
+    },
+    {
+      "epoch": 0.20001912228702554,
+      "grad_norm": 5.529078170347962,
+      "learning_rate": 8.835571895259688e-07,
+      "loss": 0.3759,
+      "step": 1046
+    },
+    {
+      "epoch": 0.2002103451572808,
+      "grad_norm": 5.013583614581529,
+      "learning_rate": 8.836786234937796e-07,
+      "loss": 0.385,
+      "step": 1047
+    },
+    {
+      "epoch": 0.2004015680275361,
+      "grad_norm": 4.312896489147245,
+      "learning_rate": 8.837999415341573e-07,
+      "loss": 0.0952,
+      "step": 1048
+    },
+    {
+      "epoch": 0.20059279089779136,
+      "grad_norm": 3.58815974107252,
+      "learning_rate": 8.839211438682318e-07,
+      "loss": 0.1084,
+      "step": 1049
+    },
+    {
+      "epoch": 0.20078401376804667,
+      "grad_norm": 7.549251079677617,
+      "learning_rate": 8.840422307165013e-07,
+      "loss": 0.232,
+      "step": 1050
+    },
+    {
+      "epoch": 0.20097523663830194,
+      "grad_norm": 7.938972636905925,
+      "learning_rate": 8.841632022988341e-07,
+      "loss": 0.2051,
+      "step": 1051
+    },
+    {
+      "epoch": 0.20116645950855722,
+      "grad_norm": 5.494197038292407,
+      "learning_rate": 8.842840588344711e-07,
+      "loss": 0.8366,
+      "step": 1052
+    },
+    {
+      "epoch": 0.20135768237881252,
+      "grad_norm": 3.7256069277750767,
+      "learning_rate": 8.844048005420283e-07,
+      "loss": 0.2259,
+      "step": 1053
+    },
+    {
+      "epoch": 0.2015489052490678,
+      "grad_norm": 3.4580204986882515,
+      "learning_rate": 8.845254276394997e-07,
+      "loss": 0.5657,
+      "step": 1054
+    },
+    {
+      "epoch": 0.20174012811932307,
+      "grad_norm": 3.373345004603082,
+      "learning_rate": 8.846459403442582e-07,
+      "loss": 0.3884,
+      "step": 1055
+    },
+    {
+      "epoch": 0.20193135098957835,
+      "grad_norm": 4.10588917997697,
+      "learning_rate": 8.847663388730595e-07,
+      "loss": 0.7206,
+      "step": 1056
+    },
+    {
+      "epoch": 0.20212257385983365,
+      "grad_norm": 4.148706229893604,
+      "learning_rate": 8.848866234420438e-07,
+      "loss": 0.373,
+      "step": 1057
+    },
+    {
+      "epoch": 0.20231379673008892,
+      "grad_norm": 3.7656168957381673,
+      "learning_rate": 8.850067942667378e-07,
+      "loss": 0.2079,
+      "step": 1058
+    },
+    {
+      "epoch": 0.2025050196003442,
+      "grad_norm": 3.600820071805055,
+      "learning_rate": 8.851268515620577e-07,
+      "loss": 0.3324,
+      "step": 1059
+    },
+    {
+      "epoch": 0.20269624247059947,
+      "grad_norm": 3.9332360083182585,
+      "learning_rate": 8.852467955423108e-07,
+      "loss": 0.1122,
+      "step": 1060
+    },
+    {
+      "epoch": 0.20288746534085478,
+      "grad_norm": 3.283503312747191,
+      "learning_rate": 8.853666264211984e-07,
+      "loss": 0.1124,
+      "step": 1061
+    },
+    {
+      "epoch": 0.20307868821111005,
+      "grad_norm": 4.108672687579544,
+      "learning_rate": 8.85486344411818e-07,
+      "loss": 0.0831,
+      "step": 1062
+    },
+    {
+      "epoch": 0.20326991108136533,
+      "grad_norm": 90.75142582385413,
+      "learning_rate": 8.856059497266651e-07,
+      "loss": 0.2714,
+      "step": 1063
+    },
+    {
+      "epoch": 0.2034611339516206,
+      "grad_norm": 5.609425494038489,
+      "learning_rate": 8.857254425776359e-07,
+      "loss": 0.516,
+      "step": 1064
+    },
+    {
+      "epoch": 0.2036523568218759,
+      "grad_norm": 5.555297474694639,
+      "learning_rate": 8.858448231760291e-07,
+      "loss": 0.6784,
+      "step": 1065
+    },
+    {
+      "epoch": 0.20384357969213118,
+      "grad_norm": 8.21572367716629,
+      "learning_rate": 8.859640917325494e-07,
+      "loss": 0.4609,
+      "step": 1066
+    },
+    {
+      "epoch": 0.20403480256238646,
+      "grad_norm": 3.7620268920310425,
+      "learning_rate": 8.860832484573078e-07,
+      "loss": 0.2183,
+      "step": 1067
+    },
+    {
+      "epoch": 0.20422602543264173,
+      "grad_norm": 3.4241438359415204,
+      "learning_rate": 8.862022935598252e-07,
+      "loss": 0.3097,
+      "step": 1068
+    },
+    {
+      "epoch": 0.20441724830289704,
+      "grad_norm": 3.642487798394141,
+      "learning_rate": 8.863212272490343e-07,
+      "loss": 0.0989,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2046084711731523,
+      "grad_norm": 3.161833253318326,
+      "learning_rate": 8.864400497332817e-07,
+      "loss": 0.1486,
+      "step": 1070
+    },
+    {
+      "epoch": 0.2047996940434076,
+      "grad_norm": 2.7842553007039945,
+      "learning_rate": 8.865587612203301e-07,
+      "loss": 0.0562,
+      "step": 1071
+    },
+    {
+      "epoch": 0.20499091691366286,
+      "grad_norm": 3.506448119306931,
+      "learning_rate": 8.866773619173606e-07,
+      "loss": 0.2518,
+      "step": 1072
+    },
+    {
+      "epoch": 0.20518213978391817,
+      "grad_norm": 4.95386949593831,
+      "learning_rate": 8.867958520309744e-07,
+      "loss": 0.2624,
+      "step": 1073
+    },
+    {
+      "epoch": 0.20537336265417344,
+      "grad_norm": 3.4203327483322385,
+      "learning_rate": 8.86914231767196e-07,
+      "loss": 0.1462,
+      "step": 1074
+    },
+    {
+      "epoch": 0.20556458552442872,
+      "grad_norm": 16.99265983538922,
+      "learning_rate": 8.870325013314737e-07,
+      "loss": 0.1843,
+      "step": 1075
+    },
+    {
+      "epoch": 0.205755808394684,
+      "grad_norm": 8.714635828565168,
+      "learning_rate": 8.871506609286841e-07,
+      "loss": 0.4877,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2059470312649393,
+      "grad_norm": 3.9263983549880956,
+      "learning_rate": 8.872687107631313e-07,
+      "loss": 0.5715,
+      "step": 1077
+    },
+    {
+      "epoch": 0.20613825413519457,
+      "grad_norm": 5.843103383238807,
+      "learning_rate": 8.873866510385519e-07,
+      "loss": 0.3729,
+      "step": 1078
+    },
+    {
+      "epoch": 0.20632947700544985,
+      "grad_norm": 6.248246824422532,
+      "learning_rate": 8.87504481958115e-07,
+      "loss": 0.388,
+      "step": 1079
+    },
+    {
+      "epoch": 0.20652069987570512,
+      "grad_norm": 3.5701602427531944,
+      "learning_rate": 8.876222037244252e-07,
+      "loss": 0.2412,
+      "step": 1080
+    },
+    {
+      "epoch": 0.20671192274596042,
+      "grad_norm": 3.410788635835878,
+      "learning_rate": 8.877398165395246e-07,
+      "loss": 0.1217,
+      "step": 1081
+    },
+    {
+      "epoch": 0.2069031456162157,
+      "grad_norm": 4.8734159585929815,
+      "learning_rate": 8.878573206048952e-07,
+      "loss": 0.2123,
+      "step": 1082
+    },
+    {
+      "epoch": 0.20709436848647098,
+      "grad_norm": 2.8634899110211687,
+      "learning_rate": 8.879747161214603e-07,
+      "loss": 0.4339,
+      "step": 1083
+    },
+    {
+      "epoch": 0.20728559135672625,
+      "grad_norm": 2.6176710906126366,
+      "learning_rate": 8.880920032895867e-07,
+      "loss": 0.1754,
+      "step": 1084
+    },
+    {
+      "epoch": 0.20747681422698155,
+      "grad_norm": 4.693272406081448,
+      "learning_rate": 8.882091823090872e-07,
+      "loss": 0.2597,
+      "step": 1085
+    },
+    {
+      "epoch": 0.20766803709723683,
+      "grad_norm": 4.645378375620822,
+      "learning_rate": 8.883262533792221e-07,
+      "loss": 0.1247,
+      "step": 1086
+    },
+    {
+      "epoch": 0.2078592599674921,
+      "grad_norm": 3.1194512505785896,
+      "learning_rate": 8.884432166987023e-07,
+      "loss": 0.0891,
+      "step": 1087
+    },
+    {
+      "epoch": 0.2080504828377474,
+      "grad_norm": 7.674519318181218,
+      "learning_rate": 8.885600724656896e-07,
+      "loss": 0.1428,
+      "step": 1088
+    },
+    {
+      "epoch": 0.20824170570800268,
+      "grad_norm": 7.389611721136649,
+      "learning_rate": 8.886768208778001e-07,
+      "loss": 0.2913,
+      "step": 1089
+    },
+    {
+      "epoch": 0.20843292857825796,
+      "grad_norm": 4.640892766041201,
+      "learning_rate": 8.887934621321054e-07,
+      "loss": 0.2151,
+      "step": 1090
+    },
+    {
+      "epoch": 0.20862415144851323,
+      "grad_norm": 4.231818183829642,
+      "learning_rate": 8.889099964251357e-07,
+      "loss": 0.3849,
+      "step": 1091
+    },
+    {
+      "epoch": 0.20881537431876854,
+      "grad_norm": 2.438618916674797,
+      "learning_rate": 8.890264239528803e-07,
+      "loss": 0.3755,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2090065971890238,
+      "grad_norm": 3.3018311766287383,
+      "learning_rate": 8.891427449107904e-07,
+      "loss": 0.1179,
+      "step": 1093
+    },
+    {
+      "epoch": 0.2091978200592791,
+      "grad_norm": 4.6029490015758805,
+      "learning_rate": 8.892589594937814e-07,
+      "loss": 0.5971,
+      "step": 1094
+    },
+    {
+      "epoch": 0.20938904292953436,
+      "grad_norm": 3.2025670007491036,
+      "learning_rate": 8.893750678962342e-07,
+      "loss": 0.2324,
+      "step": 1095
+    },
+    {
+      "epoch": 0.20958026579978967,
+      "grad_norm": 4.832590155631314,
+      "learning_rate": 8.894910703119972e-07,
+      "loss": 0.1853,
+      "step": 1096
+    },
+    {
+      "epoch": 0.20977148867004494,
+      "grad_norm": 2.3415035417242023,
+      "learning_rate": 8.896069669343886e-07,
+      "loss": 0.0929,
+      "step": 1097
+    },
+    {
+      "epoch": 0.20996271154030022,
+      "grad_norm": 3.931084992989459,
+      "learning_rate": 8.897227579561984e-07,
+      "loss": 0.1103,
+      "step": 1098
+    },
+    {
+      "epoch": 0.2101539344105555,
+      "grad_norm": 2.9864577769844556,
+      "learning_rate": 8.898384435696897e-07,
+      "loss": 0.1156,
+      "step": 1099
+    },
+    {
+      "epoch": 0.2103451572808108,
+      "grad_norm": 9.880432566080854,
+      "learning_rate": 8.899540239666012e-07,
+      "loss": 0.2588,
+      "step": 1100
+    },
+    {
+      "epoch": 0.21053638015106607,
+      "grad_norm": 4.985469112075217,
+      "learning_rate": 8.90069499338149e-07,
+      "loss": 0.4929,
+      "step": 1101
+    },
+    {
+      "epoch": 0.21072760302132135,
+      "grad_norm": 5.134771011134965,
+      "learning_rate": 8.901848698750283e-07,
+      "loss": 0.3576,
+      "step": 1102
+    },
+    {
+      "epoch": 0.21091882589157662,
+      "grad_norm": 10.009543633209118,
+      "learning_rate": 8.903001357674154e-07,
+      "loss": 0.4792,
+      "step": 1103
+    },
+    {
+      "epoch": 0.21111004876183193,
+      "grad_norm": 2.317476125221768,
+      "learning_rate": 8.904152972049694e-07,
+      "loss": 0.3401,
+      "step": 1104
+    },
+    {
+      "epoch": 0.2113012716320872,
+      "grad_norm": 3.8900003812735795,
+      "learning_rate": 8.905303543768346e-07,
+      "loss": 0.5154,
+      "step": 1105
+    },
+    {
+      "epoch": 0.21149249450234248,
+      "grad_norm": 7.733162093184951,
+      "learning_rate": 8.906453074716419e-07,
+      "loss": 0.4956,
+      "step": 1106
+    },
+    {
+      "epoch": 0.21168371737259775,
+      "grad_norm": 3.2679843684258025,
+      "learning_rate": 8.907601566775106e-07,
+      "loss": 0.1877,
+      "step": 1107
+    },
+    {
+      "epoch": 0.21187494024285305,
+      "grad_norm": 3.318358369123229,
+      "learning_rate": 8.908749021820507e-07,
+      "loss": 0.2905,
+      "step": 1108
+    },
+    {
+      "epoch": 0.21206616311310833,
+      "grad_norm": 4.0094935291874085,
+      "learning_rate": 8.909895441723643e-07,
+      "loss": 0.3586,
+      "step": 1109
+    },
+    {
+      "epoch": 0.2122573859833636,
+      "grad_norm": 3.6482598794507846,
+      "learning_rate": 8.911040828350476e-07,
+      "loss": 0.1529,
+      "step": 1110
+    },
+    {
+      "epoch": 0.21244860885361888,
+      "grad_norm": 4.479003666713787,
+      "learning_rate": 8.912185183561927e-07,
+      "loss": 0.1047,
+      "step": 1111
+    },
+    {
+      "epoch": 0.21263983172387418,
+      "grad_norm": 3.6554753713757755,
+      "learning_rate": 8.913328509213893e-07,
+      "loss": 0.0939,
+      "step": 1112
+    },
+    {
+      "epoch": 0.21283105459412946,
+      "grad_norm": 5.668497388230929,
+      "learning_rate": 8.91447080715727e-07,
+      "loss": 0.225,
+      "step": 1113
+    },
+    {
+      "epoch": 0.21302227746438473,
+      "grad_norm": 9.539787895995694,
+      "learning_rate": 8.915612079237965e-07,
+      "loss": 0.4509,
+      "step": 1114
+    },
+    {
+      "epoch": 0.21321350033464,
+      "grad_norm": 5.675562813028229,
+      "learning_rate": 8.916752327296916e-07,
+      "loss": 0.7546,
+      "step": 1115
+    },
+    {
+      "epoch": 0.2134047232048953,
+      "grad_norm": 8.806147890060732,
+      "learning_rate": 8.917891553170109e-07,
+      "loss": 0.4637,
+      "step": 1116
+    },
+    {
+      "epoch": 0.2135959460751506,
+      "grad_norm": 3.821156663987021,
+      "learning_rate": 8.919029758688601e-07,
+      "loss": 0.1248,
+      "step": 1117
+    },
+    {
+      "epoch": 0.21378716894540586,
+      "grad_norm": 4.904438243121308,
+      "learning_rate": 8.920166945678528e-07,
+      "loss": 0.3521,
+      "step": 1118
+    },
+    {
+      "epoch": 0.21397839181566114,
+      "grad_norm": 3.3831660061755464,
+      "learning_rate": 8.92130311596113e-07,
+      "loss": 0.3353,
+      "step": 1119
+    },
+    {
+      "epoch": 0.21416961468591644,
+      "grad_norm": 3.1721667827327655,
+      "learning_rate": 8.92243827135277e-07,
+      "loss": 0.068,
+      "step": 1120
+    },
+    {
+      "epoch": 0.21436083755617172,
+      "grad_norm": 2.936854839495238,
+      "learning_rate": 8.923572413664943e-07,
+      "loss": 0.1043,
+      "step": 1121
+    },
+    {
+      "epoch": 0.214552060426427,
+      "grad_norm": 2.14166913706517,
+      "learning_rate": 8.924705544704301e-07,
+      "loss": 0.0474,
+      "step": 1122
+    },
+    {
+      "epoch": 0.2147432832966823,
+      "grad_norm": 5.041676965287884,
+      "learning_rate": 8.925837666272666e-07,
+      "loss": 0.2002,
+      "step": 1123
+    },
+    {
+      "epoch": 0.21493450616693757,
+      "grad_norm": 4.189509806216216,
+      "learning_rate": 8.926968780167049e-07,
+      "loss": 0.111,
+      "step": 1124
+    },
+    {
+      "epoch": 0.21512572903719285,
+      "grad_norm": 5.456566366875169,
+      "learning_rate": 8.928098888179669e-07,
+      "loss": 0.0698,
+      "step": 1125
+    },
+    {
+      "epoch": 0.21531695190744812,
+      "grad_norm": 14.849172015499667,
+      "learning_rate": 8.929227992097965e-07,
+      "loss": 0.3835,
+      "step": 1126
+    },
+    {
+      "epoch": 0.21550817477770343,
+      "grad_norm": 3.578771008141054,
+      "learning_rate": 8.930356093704618e-07,
+      "loss": 0.2301,
+      "step": 1127
+    },
+    {
+      "epoch": 0.2156993976479587,
+      "grad_norm": 5.229304986874638,
+      "learning_rate": 8.931483194777562e-07,
+      "loss": 0.5588,
+      "step": 1128
+    },
+    {
+      "epoch": 0.21589062051821398,
+      "grad_norm": 3.9334013663620606,
+      "learning_rate": 8.932609297090011e-07,
+      "loss": 0.3984,
+      "step": 1129
+    },
+    {
+      "epoch": 0.21608184338846925,
+      "grad_norm": 4.193069853113685,
+      "learning_rate": 8.933734402410463e-07,
+      "loss": 0.138,
+      "step": 1130
+    },
+    {
+      "epoch": 0.21627306625872456,
+      "grad_norm": 4.131724800376933,
+      "learning_rate": 8.934858512502727e-07,
+      "loss": 0.3218,
+      "step": 1131
+    },
+    {
+      "epoch": 0.21646428912897983,
+      "grad_norm": 2.9504181904908493,
+      "learning_rate": 8.935981629125934e-07,
+      "loss": 0.2052,
+      "step": 1132
+    },
+    {
+      "epoch": 0.2166555119992351,
+      "grad_norm": 3.1808270186322516,
+      "learning_rate": 8.937103754034554e-07,
+      "loss": 0.2906,
+      "step": 1133
+    },
+    {
+      "epoch": 0.21684673486949038,
+      "grad_norm": 3.4836174522016177,
+      "learning_rate": 8.938224888978414e-07,
+      "loss": 0.1637,
+      "step": 1134
+    },
+    {
+      "epoch": 0.21703795773974568,
+      "grad_norm": 3.02109517683652,
+      "learning_rate": 8.939345035702716e-07,
+      "loss": 0.075,
+      "step": 1135
+    },
+    {
+      "epoch": 0.21722918061000096,
+      "grad_norm": 3.2859183390263236,
+      "learning_rate": 8.940464195948049e-07,
+      "loss": 0.1073,
+      "step": 1136
+    },
+    {
+      "epoch": 0.21742040348025624,
+      "grad_norm": 3.378466909434232,
+      "learning_rate": 8.94158237145041e-07,
+      "loss": 0.0884,
+      "step": 1137
+    },
+    {
+      "epoch": 0.2176116263505115,
+      "grad_norm": 4.338645295748355,
+      "learning_rate": 8.942699563941212e-07,
+      "loss": 0.3969,
+      "step": 1138
+    },
+    {
+      "epoch": 0.21780284922076681,
+      "grad_norm": 4.155814929648839,
+      "learning_rate": 8.943815775147311e-07,
+      "loss": 0.2843,
+      "step": 1139
+    },
+    {
+      "epoch": 0.2179940720910221,
+      "grad_norm": 4.210479088568596,
+      "learning_rate": 8.944931006791016e-07,
+      "loss": 0.5032,
+      "step": 1140
+    },
+    {
+      "epoch": 0.21818529496127737,
+      "grad_norm": 4.254674024276883,
+      "learning_rate": 8.9460452605901e-07,
+      "loss": 0.6387,
+      "step": 1141
+    },
+    {
+      "epoch": 0.21837651783153264,
+      "grad_norm": 2.9067489031727853,
+      "learning_rate": 8.947158538257832e-07,
+      "loss": 0.5248,
+      "step": 1142
+    },
+    {
+      "epoch": 0.21856774070178794,
+      "grad_norm": 4.314747763606791,
+      "learning_rate": 8.948270841502972e-07,
+      "loss": 0.3452,
+      "step": 1143
+    },
+    {
+      "epoch": 0.21875896357204322,
+      "grad_norm": 4.5433755628526855,
+      "learning_rate": 8.949382172029803e-07,
+      "loss": 0.6194,
+      "step": 1144
+    },
+    {
+      "epoch": 0.2189501864422985,
+      "grad_norm": 4.833425038662555,
+      "learning_rate": 8.950492531538137e-07,
+      "loss": 0.3998,
+      "step": 1145
+    },
+    {
+      "epoch": 0.21914140931255377,
+      "grad_norm": 2.3834550835338577,
+      "learning_rate": 8.95160192172334e-07,
+      "loss": 0.2437,
+      "step": 1146
+    },
+    {
+      "epoch": 0.21933263218280907,
+      "grad_norm": 3.4900149557673092,
+      "learning_rate": 8.952710344276333e-07,
+      "loss": 0.113,
+      "step": 1147
+    },
+    {
+      "epoch": 0.21952385505306435,
+      "grad_norm": 2.1465642357115446,
+      "learning_rate": 8.953817800883624e-07,
+      "loss": 0.257,
+      "step": 1148
+    },
+    {
+      "epoch": 0.21971507792331962,
+      "grad_norm": 5.011160601653226,
+      "learning_rate": 8.954924293227314e-07,
+      "loss": 0.1288,
+      "step": 1149
+    },
+    {
+      "epoch": 0.2199063007935749,
+      "grad_norm": 5.991773369713984,
+      "learning_rate": 8.956029822985112e-07,
+      "loss": 0.2317,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2200975236638302,
+      "grad_norm": 6.525271366951226,
+      "learning_rate": 8.957134391830351e-07,
+      "loss": 0.3748,
+      "step": 1151
+    },
+    {
+      "epoch": 0.22028874653408548,
+      "grad_norm": 3.79501598671422,
+      "learning_rate": 8.958238001432009e-07,
+      "loss": 0.3391,
+      "step": 1152
+    },
+    {
+      "epoch": 0.22047996940434075,
+      "grad_norm": 2.9699434792634576,
+      "learning_rate": 8.959340653454717e-07,
+      "loss": 0.1757,
+      "step": 1153
+    },
+    {
+      "epoch": 0.22067119227459603,
+      "grad_norm": 2.4640567448076514,
+      "learning_rate": 8.960442349558777e-07,
+      "loss": 0.2912,
+      "step": 1154
+    },
+    {
+      "epoch": 0.22086241514485133,
+      "grad_norm": 1.8927202830744823,
+      "learning_rate": 8.961543091400176e-07,
+      "loss": 0.1672,
+      "step": 1155
+    },
+    {
+      "epoch": 0.2210536380151066,
+      "grad_norm": 4.716811836803393,
+      "learning_rate": 8.962642880630602e-07,
+      "loss": 0.4733,
+      "step": 1156
+    },
+    {
+      "epoch": 0.22124486088536188,
+      "grad_norm": 2.527900834866829,
+      "learning_rate": 8.96374171889746e-07,
+      "loss": 0.1183,
+      "step": 1157
+    },
+    {
+      "epoch": 0.22143608375561716,
+      "grad_norm": 3.541453246623829,
+      "learning_rate": 8.96483960784388e-07,
+      "loss": 0.1905,
+      "step": 1158
+    },
+    {
+      "epoch": 0.22162730662587246,
+      "grad_norm": 3.9625599564790694,
+      "learning_rate": 8.965936549108746e-07,
+      "loss": 0.1549,
+      "step": 1159
+    },
+    {
+      "epoch": 0.22181852949612774,
+      "grad_norm": 4.139108977699483,
+      "learning_rate": 8.967032544326695e-07,
+      "loss": 0.2615,
+      "step": 1160
+    },
+    {
+      "epoch": 0.222009752366383,
+      "grad_norm": 4.238400505954687,
+      "learning_rate": 8.968127595128139e-07,
+      "loss": 0.1328,
+      "step": 1161
+    },
+    {
+      "epoch": 0.22220097523663832,
+      "grad_norm": 4.315934071671014,
+      "learning_rate": 8.969221703139279e-07,
+      "loss": 0.1719,
+      "step": 1162
+    },
+    {
+      "epoch": 0.2223921981068936,
+      "grad_norm": 8.68934710509328,
+      "learning_rate": 8.970314869982123e-07,
+      "loss": 0.2882,
+      "step": 1163
+    },
+    {
+      "epoch": 0.22258342097714887,
+      "grad_norm": 3.7049019328741153,
+      "learning_rate": 8.971407097274491e-07,
+      "loss": 0.3623,
+      "step": 1164
+    },
+    {
+      "epoch": 0.22277464384740414,
+      "grad_norm": 5.729877256773093,
+      "learning_rate": 8.972498386630039e-07,
+      "loss": 0.479,
+      "step": 1165
+    },
+    {
+      "epoch": 0.22296586671765944,
+      "grad_norm": 2.781888567022161,
+      "learning_rate": 8.97358873965827e-07,
+      "loss": 0.3906,
+      "step": 1166
+    },
+    {
+      "epoch": 0.22315708958791472,
+      "grad_norm": 4.048511071311434,
+      "learning_rate": 8.974678157964544e-07,
+      "loss": 0.561,
+      "step": 1167
+    },
+    {
+      "epoch": 0.22334831245817,
+      "grad_norm": 2.738923393412176,
+      "learning_rate": 8.975766643150099e-07,
+      "loss": 0.2944,
+      "step": 1168
+    },
+    {
+      "epoch": 0.22353953532842527,
+      "grad_norm": 2.560108836959882,
+      "learning_rate": 8.976854196812061e-07,
+      "loss": 0.247,
+      "step": 1169
+    },
+    {
+      "epoch": 0.22373075819868057,
+      "grad_norm": 3.532748419253456,
+      "learning_rate": 8.977940820543459e-07,
+      "loss": 0.4399,
+      "step": 1170
+    },
+    {
+      "epoch": 0.22392198106893585,
+      "grad_norm": 3.3635893181081298,
+      "learning_rate": 8.979026515933239e-07,
+      "loss": 0.091,
+      "step": 1171
+    },
+    {
+      "epoch": 0.22411320393919112,
+      "grad_norm": 2.9577572937824597,
+      "learning_rate": 8.980111284566278e-07,
+      "loss": 0.0838,
+      "step": 1172
+    },
+    {
+      "epoch": 0.2243044268094464,
+      "grad_norm": 5.518204645015665,
+      "learning_rate": 8.9811951280234e-07,
+      "loss": 0.1394,
+      "step": 1173
+    },
+    {
+      "epoch": 0.2244956496797017,
+      "grad_norm": 3.5702076568902643,
+      "learning_rate": 8.982278047881382e-07,
+      "loss": 0.1315,
+      "step": 1174
+    },
+    {
+      "epoch": 0.22468687254995698,
+      "grad_norm": 6.091871079521605,
+      "learning_rate": 8.98336004571298e-07,
+      "loss": 0.1774,
+      "step": 1175
+    },
+    {
+      "epoch": 0.22487809542021225,
+      "grad_norm": 10.857827143508977,
+      "learning_rate": 8.984441123086933e-07,
+      "loss": 0.413,
+      "step": 1176
+    },
+    {
+      "epoch": 0.22506931829046753,
+      "grad_norm": 5.373212827714807,
+      "learning_rate": 8.985521281567979e-07,
+      "loss": 0.6765,
+      "step": 1177
+    },
+    {
+      "epoch": 0.22526054116072283,
+      "grad_norm": 4.516109664936991,
+      "learning_rate": 8.986600522716872e-07,
+      "loss": 0.4992,
+      "step": 1178
+    },
+    {
+      "epoch": 0.2254517640309781,
+      "grad_norm": 5.063637499702517,
+      "learning_rate": 8.987678848090391e-07,
+      "loss": 0.6182,
+      "step": 1179
+    },
+    {
+      "epoch": 0.22564298690123338,
+      "grad_norm": 4.343158050380614,
+      "learning_rate": 8.988756259241355e-07,
+      "loss": 0.5289,
+      "step": 1180
+    },
+    {
+      "epoch": 0.22583420977148866,
+      "grad_norm": 2.662633070508226,
+      "learning_rate": 8.98983275771864e-07,
+      "loss": 0.1306,
+      "step": 1181
+    },
+    {
+      "epoch": 0.22602543264174396,
+      "grad_norm": 9.6678914138321,
+      "learning_rate": 8.990908345067184e-07,
+      "loss": 0.302,
+      "step": 1182
+    },
+    {
+      "epoch": 0.22621665551199924,
+      "grad_norm": 2.8602308649637207,
+      "learning_rate": 8.991983022828009e-07,
+      "loss": 0.1483,
+      "step": 1183
+    },
+    {
+      "epoch": 0.2264078783822545,
+      "grad_norm": 1.648759710993305,
+      "learning_rate": 8.993056792538232e-07,
+      "loss": 0.1182,
+      "step": 1184
+    },
+    {
+      "epoch": 0.2265991012525098,
+      "grad_norm": 2.731719291401996,
+      "learning_rate": 8.994129655731074e-07,
+      "loss": 0.2208,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2267903241227651,
+      "grad_norm": 3.814013821589326,
+      "learning_rate": 8.995201613935877e-07,
+      "loss": 0.1311,
+      "step": 1186
+    },
+    {
+      "epoch": 0.22698154699302037,
+      "grad_norm": 4.82015639783451,
+      "learning_rate": 8.996272668678117e-07,
+      "loss": 0.0847,
+      "step": 1187
+    },
+    {
+      "epoch": 0.22717276986327564,
+      "grad_norm": 8.190574942169858,
+      "learning_rate": 8.997342821479415e-07,
+      "loss": 0.3363,
+      "step": 1188
+    },
+    {
+      "epoch": 0.22736399273353092,
+      "grad_norm": 5.15593575184688,
+      "learning_rate": 8.998412073857548e-07,
+      "loss": 0.2721,
+      "step": 1189
+    },
+    {
+      "epoch": 0.22755521560378622,
+      "grad_norm": 4.261881208950923,
+      "learning_rate": 8.999480427326476e-07,
+      "loss": 0.5089,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2277464384740415,
+      "grad_norm": 2.857090271738332,
+      "learning_rate": 9.000547883396331e-07,
+      "loss": 0.3051,
+      "step": 1191
+    },
+    {
+      "epoch": 0.22793766134429677,
+      "grad_norm": 4.311429761687626,
+      "learning_rate": 9.00161444357345e-07,
+      "loss": 0.5416,
+      "step": 1192
+    },
+    {
+      "epoch": 0.22812888421455205,
+      "grad_norm": 3.5786604168359313,
+      "learning_rate": 9.002680109360378e-07,
+      "loss": 0.3437,
+      "step": 1193
+    },
+    {
+      "epoch": 0.22832010708480735,
+      "grad_norm": 5.026039221043231,
+      "learning_rate": 9.003744882255885e-07,
+      "loss": 0.3367,
+      "step": 1194
+    },
+    {
+      "epoch": 0.22851132995506263,
+      "grad_norm": 2.3748695940054256,
+      "learning_rate": 9.004808763754974e-07,
+      "loss": 0.3038,
+      "step": 1195
+    },
+    {
+      "epoch": 0.2287025528253179,
+      "grad_norm": 2.7057474458215918,
+      "learning_rate": 9.0058717553489e-07,
+      "loss": 0.1171,
+      "step": 1196
+    },
+    {
+      "epoch": 0.2288937756955732,
+      "grad_norm": 2.6507140870921537,
+      "learning_rate": 9.006933858525176e-07,
+      "loss": 0.4792,
+      "step": 1197
+    },
+    {
+      "epoch": 0.22908499856582848,
+      "grad_norm": 4.300191591119739,
+      "learning_rate": 9.007995074767592e-07,
+      "loss": 0.2245,
+      "step": 1198
+    },
+    {
+      "epoch": 0.22927622143608375,
+      "grad_norm": 4.242376333849675,
+      "learning_rate": 9.009055405556217e-07,
+      "loss": 0.1308,
+      "step": 1199
+    },
+    {
+      "epoch": 0.22946744430633903,
+      "grad_norm": 9.103402275160558,
+      "learning_rate": 9.010114852367426e-07,
+      "loss": 0.2023,
+      "step": 1200
+    },
+    {
+      "epoch": 0.22965866717659433,
+      "grad_norm": 5.454410783954768,
+      "learning_rate": 9.0111734166739e-07,
+      "loss": 0.6782,
+      "step": 1201
+    },
+    {
+      "epoch": 0.2298498900468496,
+      "grad_norm": 3.76390713964513,
+      "learning_rate": 9.012231099944646e-07,
+      "loss": 0.0975,
+      "step": 1202
+    },
+    {
+      "epoch": 0.23004111291710488,
+      "grad_norm": 3.49618472188729,
+      "learning_rate": 9.013287903645002e-07,
+      "loss": 0.3123,
+      "step": 1203
+    },
+    {
+      "epoch": 0.23023233578736016,
+      "grad_norm": 2.283063925096685,
+      "learning_rate": 9.014343829236657e-07,
+      "loss": 0.1078,
+      "step": 1204
+    },
+    {
+      "epoch": 0.23042355865761546,
+      "grad_norm": 3.025846719070477,
+      "learning_rate": 9.015398878177658e-07,
+      "loss": 0.2737,
+      "step": 1205
+    },
+    {
+      "epoch": 0.23061478152787074,
+      "grad_norm": 4.352639960372761,
+      "learning_rate": 9.016453051922424e-07,
+      "loss": 0.5128,
+      "step": 1206
+    },
+    {
+      "epoch": 0.230806004398126,
+      "grad_norm": 4.22489786419632,
+      "learning_rate": 9.017506351921755e-07,
+      "loss": 0.3248,
+      "step": 1207
+    },
+    {
+      "epoch": 0.2309972272683813,
+      "grad_norm": 2.181821373373645,
+      "learning_rate": 9.018558779622851e-07,
+      "loss": 0.228,
+      "step": 1208
+    },
+    {
+      "epoch": 0.2311884501386366,
+      "grad_norm": 2.660958817419565,
+      "learning_rate": 9.019610336469317e-07,
+      "loss": 0.2373,
+      "step": 1209
+    },
+    {
+      "epoch": 0.23137967300889187,
+      "grad_norm": 3.884969499780042,
+      "learning_rate": 9.020661023901175e-07,
+      "loss": 0.1842,
+      "step": 1210
+    },
+    {
+      "epoch": 0.23157089587914714,
+      "grad_norm": 3.0935160230533305,
+      "learning_rate": 9.021710843354883e-07,
+      "loss": 0.1357,
+      "step": 1211
+    },
+    {
+      "epoch": 0.23176211874940242,
+      "grad_norm": 6.765779489070462,
+      "learning_rate": 9.02275979626334e-07,
+      "loss": 0.1072,
+      "step": 1212
+    },
+    {
+      "epoch": 0.23195334161965772,
+      "grad_norm": 4.527486356212294,
+      "learning_rate": 9.023807884055897e-07,
+      "loss": 0.2136,
+      "step": 1213
+    },
+    {
+      "epoch": 0.232144564489913,
+      "grad_norm": 4.0154209901515205,
+      "learning_rate": 9.024855108158375e-07,
+      "loss": 0.5395,
+      "step": 1214
+    },
+    {
+      "epoch": 0.23233578736016827,
+      "grad_norm": 3.819451670013683,
+      "learning_rate": 9.02590146999307e-07,
+      "loss": 0.3241,
+      "step": 1215
+    },
+    {
+      "epoch": 0.23252701023042355,
+      "grad_norm": 5.034060054804271,
+      "learning_rate": 9.026946970978773e-07,
+      "loss": 0.2662,
+      "step": 1216
+    },
+    {
+      "epoch": 0.23271823310067885,
+      "grad_norm": 3.2247393539729705,
+      "learning_rate": 9.027991612530766e-07,
+      "loss": 0.5528,
+      "step": 1217
+    },
+    {
+      "epoch": 0.23290945597093413,
+      "grad_norm": 2.6861017715079587,
+      "learning_rate": 9.029035396060857e-07,
+      "loss": 0.184,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2331006788411894,
+      "grad_norm": 7.167739639986295,
+      "learning_rate": 9.030078322977366e-07,
+      "loss": 0.4245,
+      "step": 1219
+    },
+    {
+      "epoch": 0.23329190171144468,
+      "grad_norm": 3.534036802643193,
+      "learning_rate": 9.031120394685159e-07,
+      "loss": 0.2124,
+      "step": 1220
+    },
+    {
+      "epoch": 0.23348312458169998,
+      "grad_norm": 3.168267364162992,
+      "learning_rate": 9.032161612585637e-07,
+      "loss": 0.3235,
+      "step": 1221
+    },
+    {
+      "epoch": 0.23367434745195526,
+      "grad_norm": 3.8465309317676635,
+      "learning_rate": 9.03320197807677e-07,
+      "loss": 0.1941,
+      "step": 1222
+    },
+    {
+      "epoch": 0.23386557032221053,
+      "grad_norm": 3.3958789533737503,
+      "learning_rate": 9.03424149255309e-07,
+      "loss": 0.2437,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2340567931924658,
+      "grad_norm": 3.270155542856501,
+      "learning_rate": 9.035280157405715e-07,
+      "loss": 0.0987,
+      "step": 1224
+    },
+    {
+      "epoch": 0.2342480160627211,
+      "grad_norm": 6.91340618574602,
+      "learning_rate": 9.036317974022349e-07,
+      "loss": 0.355,
+      "step": 1225
+    },
+    {
+      "epoch": 0.23443923893297638,
+      "grad_norm": 4.303592272551415,
+      "learning_rate": 9.037354943787308e-07,
+      "loss": 0.4324,
+      "step": 1226
+    },
+    {
+      "epoch": 0.23463046180323166,
+      "grad_norm": 3.4072982155155667,
+      "learning_rate": 9.038391068081511e-07,
+      "loss": 0.4405,
+      "step": 1227
+    },
+    {
+      "epoch": 0.23482168467348694,
+      "grad_norm": 3.138281408019818,
+      "learning_rate": 9.03942634828251e-07,
+      "loss": 0.4101,
+      "step": 1228
+    },
+    {
+      "epoch": 0.23501290754374224,
+      "grad_norm": 2.817661212144018,
+      "learning_rate": 9.040460785764488e-07,
+      "loss": 0.113,
+      "step": 1229
+    },
+    {
+      "epoch": 0.23520413041399751,
+      "grad_norm": 3.5727863047018276,
+      "learning_rate": 9.041494381898281e-07,
+      "loss": 0.3263,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2353953532842528,
+      "grad_norm": 4.573331751772178,
+      "learning_rate": 9.042527138051376e-07,
+      "loss": 0.2399,
+      "step": 1231
+    },
+    {
+      "epoch": 0.2355865761545081,
+      "grad_norm": 1.505337913065843,
+      "learning_rate": 9.043559055587932e-07,
+      "loss": 0.1341,
+      "step": 1232
+    },
+    {
+      "epoch": 0.23577779902476337,
+      "grad_norm": 4.043935053982098,
+      "learning_rate": 9.044590135868789e-07,
+      "loss": 0.2172,
+      "step": 1233
+    },
+    {
+      "epoch": 0.23596902189501864,
+      "grad_norm": 2.604758192909713,
+      "learning_rate": 9.045620380251476e-07,
+      "loss": 0.1185,
+      "step": 1234
+    },
+    {
+      "epoch": 0.23616024476527392,
+      "grad_norm": 4.59186042796832,
+      "learning_rate": 9.046649790090222e-07,
+      "loss": 0.4579,
+      "step": 1235
+    },
+    {
+      "epoch": 0.23635146763552922,
+      "grad_norm": 3.7649779970861026,
+      "learning_rate": 9.047678366735967e-07,
+      "loss": 0.2194,
+      "step": 1236
+    },
+    {
+      "epoch": 0.2365426905057845,
+      "grad_norm": 2.841287070358524,
+      "learning_rate": 9.048706111536376e-07,
+      "loss": 0.0679,
+      "step": 1237
+    },
+    {
+      "epoch": 0.23673391337603977,
+      "grad_norm": 8.3383433222884,
+      "learning_rate": 9.049733025835846e-07,
+      "loss": 0.2038,
+      "step": 1238
+    },
+    {
+      "epoch": 0.23692513624629505,
+      "grad_norm": 4.122120169866984,
+      "learning_rate": 9.050759110975517e-07,
+      "loss": 0.3571,
+      "step": 1239
+    },
+    {
+      "epoch": 0.23711635911655035,
+      "grad_norm": 4.355002558780798,
+      "learning_rate": 9.051784368293285e-07,
+      "loss": 0.3959,
+      "step": 1240
+    },
+    {
+      "epoch": 0.23730758198680563,
+      "grad_norm": 4.077350645247691,
+      "learning_rate": 9.052808799123804e-07,
+      "loss": 0.7273,
+      "step": 1241
+    },
+    {
+      "epoch": 0.2374988048570609,
+      "grad_norm": 3.477250140474325,
+      "learning_rate": 9.053832404798512e-07,
+      "loss": 0.3513,
+      "step": 1242
+    },
+    {
+      "epoch": 0.23769002772731618,
+      "grad_norm": 4.562312292130794,
+      "learning_rate": 9.054855186645626e-07,
+      "loss": 0.7533,
+      "step": 1243
+    },
+    {
+      "epoch": 0.23788125059757148,
+      "grad_norm": 2.281265363249584,
+      "learning_rate": 9.05587714599016e-07,
+      "loss": 0.3358,
+      "step": 1244
+    },
+    {
+      "epoch": 0.23807247346782676,
+      "grad_norm": 3.223632220823181,
+      "learning_rate": 9.056898284153936e-07,
+      "loss": 0.378,
+      "step": 1245
+    },
+    {
+      "epoch": 0.23826369633808203,
+      "grad_norm": 4.002408256360905,
+      "learning_rate": 9.057918602455589e-07,
+      "loss": 0.1863,
+      "step": 1246
+    },
+    {
+      "epoch": 0.2384549192083373,
+      "grad_norm": 4.967390918196361,
+      "learning_rate": 9.058938102210583e-07,
+      "loss": 0.3103,
+      "step": 1247
+    },
+    {
+      "epoch": 0.2386461420785926,
+      "grad_norm": 2.3657850889110916,
+      "learning_rate": 9.059956784731217e-07,
+      "loss": 0.0726,
+      "step": 1248
+    },
+    {
+      "epoch": 0.23883736494884789,
+      "grad_norm": 2.7429330361175426,
+      "learning_rate": 9.060974651326635e-07,
+      "loss": 0.0661,
+      "step": 1249
+    },
+    {
+      "epoch": 0.23902858781910316,
+      "grad_norm": 3.144145605476938,
+      "learning_rate": 9.061991703302844e-07,
+      "loss": 0.0579,
+      "step": 1250
+    },
+    {
+      "epoch": 0.23921981068935844,
+      "grad_norm": 3.7939708262092573,
+      "learning_rate": 9.063007941962711e-07,
+      "loss": 0.2634,
+      "step": 1251
+    },
+    {
+      "epoch": 0.23941103355961374,
+      "grad_norm": 4.157111745426726,
+      "learning_rate": 9.064023368605984e-07,
+      "loss": 0.615,
+      "step": 1252
+    },
+    {
+      "epoch": 0.23960225642986901,
+      "grad_norm": 8.373536394819759,
+      "learning_rate": 9.065037984529295e-07,
+      "loss": 0.4538,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2397934793001243,
+      "grad_norm": 1.3651349600421687,
+      "learning_rate": 9.066051791026177e-07,
+      "loss": 0.368,
+      "step": 1254
+    },
+    {
+      "epoch": 0.23998470217037957,
+      "grad_norm": 3.536258757422975,
+      "learning_rate": 9.067064789387063e-07,
+      "loss": 0.2567,
+      "step": 1255
+    },
+    {
+      "epoch": 0.24017592504063487,
+      "grad_norm": 6.474391670076702,
+      "learning_rate": 9.068076980899311e-07,
+      "loss": 0.219,
+      "step": 1256
+    },
+    {
+      "epoch": 0.24036714791089014,
+      "grad_norm": 2.940199565212601,
+      "learning_rate": 9.069088366847196e-07,
+      "loss": 0.1743,
+      "step": 1257
+    },
+    {
+      "epoch": 0.24055837078114542,
+      "grad_norm": 2.9010953413120233,
+      "learning_rate": 9.070098948511937e-07,
+      "loss": 0.1936,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2407495936514007,
+      "grad_norm": 5.5281485790485165,
+      "learning_rate": 9.071108727171697e-07,
+      "loss": 0.2619,
+      "step": 1259
+    },
+    {
+      "epoch": 0.240940816521656,
+      "grad_norm": 5.844353139830521,
+      "learning_rate": 9.07211770410159e-07,
+      "loss": 0.1944,
+      "step": 1260
+    },
+    {
+      "epoch": 0.24113203939191127,
+      "grad_norm": 4.754958726641634,
+      "learning_rate": 9.073125880573699e-07,
+      "loss": 0.1428,
+      "step": 1261
+    },
+    {
+      "epoch": 0.24132326226216655,
+      "grad_norm": 4.640459359549112,
+      "learning_rate": 9.07413325785708e-07,
+      "loss": 0.1263,
+      "step": 1262
+    },
+    {
+      "epoch": 0.24151448513242182,
+      "grad_norm": 6.683332053522889,
+      "learning_rate": 9.075139837217782e-07,
+      "loss": 0.2818,
+      "step": 1263
+    },
+    {
+      "epoch": 0.24170570800267713,
+      "grad_norm": 3.048825622596667,
+      "learning_rate": 9.076145619918832e-07,
+      "loss": 0.2549,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2418969308729324,
+      "grad_norm": 3.880407313764282,
+      "learning_rate": 9.077150607220273e-07,
+      "loss": 0.3776,
+      "step": 1265
+    },
+    {
+      "epoch": 0.24208815374318768,
+      "grad_norm": 4.829202179974282,
+      "learning_rate": 9.078154800379158e-07,
+      "loss": 0.1717,
+      "step": 1266
+    },
+    {
+      "epoch": 0.24227937661344298,
+      "grad_norm": 3.888673223842998,
+      "learning_rate": 9.079158200649558e-07,
+      "loss": 0.1636,
+      "step": 1267
+    },
+    {
+      "epoch": 0.24247059948369826,
+      "grad_norm": 2.8602205287500992,
+      "learning_rate": 9.080160809282585e-07,
+      "loss": 0.1063,
+      "step": 1268
+    },
+    {
+      "epoch": 0.24266182235395353,
+      "grad_norm": 4.430841958180637,
+      "learning_rate": 9.081162627526381e-07,
+      "loss": 0.2701,
+      "step": 1269
+    },
+    {
+      "epoch": 0.2428530452242088,
+      "grad_norm": 1.9174480158636449,
+      "learning_rate": 9.082163656626146e-07,
+      "loss": 0.1217,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2430442680944641,
+      "grad_norm": 2.9769914773518558,
+      "learning_rate": 9.083163897824139e-07,
+      "loss": 0.1628,
+      "step": 1271
+    },
+    {
+      "epoch": 0.2432354909647194,
+      "grad_norm": 3.8440236560200813,
+      "learning_rate": 9.084163352359684e-07,
+      "loss": 0.1966,
+      "step": 1272
+    },
+    {
+      "epoch": 0.24342671383497466,
+      "grad_norm": 3.570354436381856,
+      "learning_rate": 9.085162021469187e-07,
+      "loss": 0.2201,
+      "step": 1273
+    },
+    {
+      "epoch": 0.24361793670522994,
+      "grad_norm": 4.897718848362219,
+      "learning_rate": 9.08615990638614e-07,
+      "loss": 0.2151,
+      "step": 1274
+    },
+    {
+      "epoch": 0.24380915957548524,
+      "grad_norm": 9.939868188978052,
+      "learning_rate": 9.087157008341132e-07,
+      "loss": 0.1761,
+      "step": 1275
+    },
+    {
+      "epoch": 0.24400038244574052,
+      "grad_norm": 6.75626929860001,
+      "learning_rate": 9.088153328561857e-07,
+      "loss": 0.6122,
+      "step": 1276
+    },
+    {
+      "epoch": 0.2441916053159958,
+      "grad_norm": 5.863880104015393,
+      "learning_rate": 9.089148868273125e-07,
+      "loss": 0.5453,
+      "step": 1277
+    },
+    {
+      "epoch": 0.24438282818625107,
+      "grad_norm": 5.171260370802445,
+      "learning_rate": 9.090143628696868e-07,
+      "loss": 0.3543,
+      "step": 1278
+    },
+    {
+      "epoch": 0.24457405105650637,
+      "grad_norm": 2.9026654194852677,
+      "learning_rate": 9.091137611052152e-07,
+      "loss": 0.3164,
+      "step": 1279
+    },
+    {
+      "epoch": 0.24476527392676165,
+      "grad_norm": 3.161913332596509,
+      "learning_rate": 9.092130816555184e-07,
+      "loss": 0.3061,
+      "step": 1280
+    },
+    {
+      "epoch": 0.24495649679701692,
+      "grad_norm": 3.0868803877473576,
+      "learning_rate": 9.093123246419321e-07,
+      "loss": 0.2175,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2451477196672722,
+      "grad_norm": 4.135439061537415,
+      "learning_rate": 9.094114901855082e-07,
+      "loss": 0.2442,
+      "step": 1282
+    },
+    {
+      "epoch": 0.2453389425375275,
+      "grad_norm": 2.993577917962973,
+      "learning_rate": 9.095105784070152e-07,
+      "loss": 0.1401,
+      "step": 1283
+    },
+    {
+      "epoch": 0.24553016540778277,
+      "grad_norm": 2.3711529744778685,
+      "learning_rate": 9.096095894269393e-07,
+      "loss": 0.075,
+      "step": 1284
+    },
+    {
+      "epoch": 0.24572138827803805,
+      "grad_norm": 3.6416412375905884,
+      "learning_rate": 9.097085233654856e-07,
+      "loss": 0.0979,
+      "step": 1285
+    },
+    {
+      "epoch": 0.24591261114829333,
+      "grad_norm": 3.2147873865426666,
+      "learning_rate": 9.098073803425782e-07,
+      "loss": 0.0893,
+      "step": 1286
+    },
+    {
+      "epoch": 0.24610383401854863,
+      "grad_norm": 3.117241956956965,
+      "learning_rate": 9.099061604778621e-07,
+      "loss": 0.061,
+      "step": 1287
+    },
+    {
+      "epoch": 0.2462950568888039,
+      "grad_norm": 12.545177150066973,
+      "learning_rate": 9.10004863890703e-07,
+      "loss": 0.6514,
+      "step": 1288
+    },
+    {
+      "epoch": 0.24648627975905918,
+      "grad_norm": 3.5129080212999146,
+      "learning_rate": 9.101034907001891e-07,
+      "loss": 0.2645,
+      "step": 1289
+    },
+    {
+      "epoch": 0.24667750262931445,
+      "grad_norm": 4.932777947445901,
+      "learning_rate": 9.102020410251313e-07,
+      "loss": 0.4276,
+      "step": 1290
+    },
+    {
+      "epoch": 0.24686872549956976,
+      "grad_norm": 2.9490210056326487,
+      "learning_rate": 9.103005149840643e-07,
+      "loss": 0.4109,
+      "step": 1291
+    },
+    {
+      "epoch": 0.24705994836982503,
+      "grad_norm": 5.030104038686939,
+      "learning_rate": 9.103989126952477e-07,
+      "loss": 0.1776,
+      "step": 1292
+    },
+    {
+      "epoch": 0.2472511712400803,
+      "grad_norm": 4.4185260721361015,
+      "learning_rate": 9.104972342766663e-07,
+      "loss": 0.2959,
+      "step": 1293
+    },
+    {
+      "epoch": 0.24744239411033558,
+      "grad_norm": 3.675881525307657,
+      "learning_rate": 9.105954798460311e-07,
+      "loss": 0.1516,
+      "step": 1294
+    },
+    {
+      "epoch": 0.2476336169805909,
+      "grad_norm": 3.2141720706286674,
+      "learning_rate": 9.106936495207812e-07,
+      "loss": 0.1732,
+      "step": 1295
+    },
+    {
+      "epoch": 0.24782483985084616,
+      "grad_norm": 3.8743723699441084,
+      "learning_rate": 9.107917434180827e-07,
+      "loss": 0.1655,
+      "step": 1296
+    },
+    {
+      "epoch": 0.24801606272110144,
+      "grad_norm": 3.4023890180236975,
+      "learning_rate": 9.108897616548312e-07,
+      "loss": 0.189,
+      "step": 1297
+    },
+    {
+      "epoch": 0.2482072855913567,
+      "grad_norm": 4.44714660887354,
+      "learning_rate": 9.109877043476517e-07,
+      "loss": 0.3939,
+      "step": 1298
+    },
+    {
+      "epoch": 0.24839850846161202,
+      "grad_norm": 3.808777664830262,
+      "learning_rate": 9.110855716129001e-07,
+      "loss": 0.0965,
+      "step": 1299
+    },
+    {
+      "epoch": 0.2485897313318673,
+      "grad_norm": 5.328398666736994,
+      "learning_rate": 9.111833635666634e-07,
+      "loss": 0.1255,
+      "step": 1300
+    },
+    {
+      "epoch": 0.24878095420212257,
+      "grad_norm": 4.613376962054863,
+      "learning_rate": 9.112810803247607e-07,
+      "loss": 0.1765,
+      "step": 1301
+    },
+    {
+      "epoch": 0.24897217707237784,
+      "grad_norm": 4.2435642207385325,
+      "learning_rate": 9.113787220027446e-07,
+      "loss": 0.7329,
+      "step": 1302
+    },
+    {
+      "epoch": 0.24916339994263315,
+      "grad_norm": 6.521652251508018,
+      "learning_rate": 9.114762887159012e-07,
+      "loss": 0.3743,
+      "step": 1303
+    },
+    {
+      "epoch": 0.24935462281288842,
+      "grad_norm": 2.9762254258783107,
+      "learning_rate": 9.115737805792514e-07,
+      "loss": 0.2299,
+      "step": 1304
+    },
+    {
+      "epoch": 0.2495458456831437,
+      "grad_norm": 2.952762965056223,
+      "learning_rate": 9.116711977075514e-07,
+      "loss": 0.343,
+      "step": 1305
+    },
+    {
+      "epoch": 0.249737068553399,
+      "grad_norm": 2.22730636217415,
+      "learning_rate": 9.117685402152938e-07,
+      "loss": 0.3029,
+      "step": 1306
+    },
+    {
+      "epoch": 0.24992829142365428,
+      "grad_norm": 2.9458708297163225,
+      "learning_rate": 9.118658082167086e-07,
+      "loss": 0.2458,
+      "step": 1307
+    },
+    {
+      "epoch": 0.25011951429390955,
+      "grad_norm": 3.0703674777401253,
+      "learning_rate": 9.119630018257632e-07,
+      "loss": 0.3242,
+      "step": 1308
+    },
+    {
+      "epoch": 0.25031073716416485,
+      "grad_norm": 3.099917189199579,
+      "learning_rate": 9.120601211561639e-07,
+      "loss": 0.1547,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2505019600344201,
+      "grad_norm": 24.428307094434523,
+      "learning_rate": 9.121571663213565e-07,
+      "loss": 0.2551,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2506931829046754,
+      "grad_norm": 6.26237605233036,
+      "learning_rate": 9.122541374345275e-07,
+      "loss": 0.1633,
+      "step": 1311
+    },
+    {
+      "epoch": 0.2508844057749307,
+      "grad_norm": 6.62749016944159,
+      "learning_rate": 9.123510346086037e-07,
+      "loss": 0.118,
+      "step": 1312
+    },
+    {
+      "epoch": 0.25107562864518596,
+      "grad_norm": 6.644617079012829,
+      "learning_rate": 9.124478579562546e-07,
+      "loss": 0.3344,
+      "step": 1313
+    },
+    {
+      "epoch": 0.25126685151544126,
+      "grad_norm": 3.325023656775277,
+      "learning_rate": 9.125446075898917e-07,
+      "loss": 0.5004,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2514580743856965,
+      "grad_norm": 4.826763467257087,
+      "learning_rate": 9.126412836216704e-07,
+      "loss": 0.6764,
+      "step": 1315
+    },
+    {
+      "epoch": 0.2516492972559518,
+      "grad_norm": 2.183569838063673,
+      "learning_rate": 9.1273788616349e-07,
+      "loss": 0.2421,
+      "step": 1316
+    },
+    {
+      "epoch": 0.2518405201262071,
+      "grad_norm": 4.00164880149335,
+      "learning_rate": 9.12834415326995e-07,
+      "loss": 0.3633,
+      "step": 1317
+    },
+    {
+      "epoch": 0.25203174299646236,
+      "grad_norm": 4.343770637737619,
+      "learning_rate": 9.129308712235758e-07,
+      "loss": 0.3869,
+      "step": 1318
+    },
+    {
+      "epoch": 0.25222296586671766,
+      "grad_norm": 2.8138276251546133,
+      "learning_rate": 9.13027253964369e-07,
+      "loss": 0.2071,
+      "step": 1319
+    },
+    {
+      "epoch": 0.25241418873697297,
+      "grad_norm": 3.7066327313944036,
+      "learning_rate": 9.131235636602589e-07,
+      "loss": 0.2348,
+      "step": 1320
+    },
+    {
+      "epoch": 0.2526054116072282,
+      "grad_norm": 3.66645094208025,
+      "learning_rate": 9.132198004218775e-07,
+      "loss": 0.292,
+      "step": 1321
+    },
+    {
+      "epoch": 0.2527966344774835,
+      "grad_norm": 6.845936883045346,
+      "learning_rate": 9.13315964359606e-07,
+      "loss": 0.1095,
+      "step": 1322
+    },
+    {
+      "epoch": 0.25298785734773876,
+      "grad_norm": 5.802809646410992,
+      "learning_rate": 9.134120555835751e-07,
+      "loss": 0.1243,
+      "step": 1323
+    },
+    {
+      "epoch": 0.25317908021799407,
+      "grad_norm": 4.479745848122361,
+      "learning_rate": 9.135080742036657e-07,
+      "loss": 0.1374,
+      "step": 1324
+    },
+    {
+      "epoch": 0.25337030308824937,
+      "grad_norm": 4.826169306138025,
+      "learning_rate": 9.136040203295101e-07,
+      "loss": 0.1224,
+      "step": 1325
+    },
+    {
+      "epoch": 0.2535615259585046,
+      "grad_norm": 4.284912324985483,
+      "learning_rate": 9.136998940704921e-07,
+      "loss": 0.2691,
+      "step": 1326
+    },
+    {
+      "epoch": 0.2537527488287599,
+      "grad_norm": 4.278228584634965,
+      "learning_rate": 9.137956955357485e-07,
+      "loss": 0.4657,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2539439716990152,
+      "grad_norm": 4.380120904377286,
+      "learning_rate": 9.138914248341692e-07,
+      "loss": 0.5421,
+      "step": 1328
+    },
+    {
+      "epoch": 0.2541351945692705,
+      "grad_norm": 3.614512828538709,
+      "learning_rate": 9.139870820743986e-07,
+      "loss": 0.6621,
+      "step": 1329
+    },
+    {
+      "epoch": 0.2543264174395258,
+      "grad_norm": 3.214163911116229,
+      "learning_rate": 9.140826673648352e-07,
+      "loss": 0.5049,
+      "step": 1330
+    },
+    {
+      "epoch": 0.254517640309781,
+      "grad_norm": 2.820852127651185,
+      "learning_rate": 9.141781808136337e-07,
+      "loss": 0.3808,
+      "step": 1331
+    },
+    {
+      "epoch": 0.2547088631800363,
+      "grad_norm": 7.234303550284991,
+      "learning_rate": 9.14273622528705e-07,
+      "loss": 0.1168,
+      "step": 1332
+    },
+    {
+      "epoch": 0.25490008605029163,
+      "grad_norm": 2.1994647198497015,
+      "learning_rate": 9.143689926177171e-07,
+      "loss": 0.1108,
+      "step": 1333
+    },
+    {
+      "epoch": 0.2550913089205469,
+      "grad_norm": 2.843239371598273,
+      "learning_rate": 9.144642911880956e-07,
+      "loss": 0.1109,
+      "step": 1334
+    },
+    {
+      "epoch": 0.2552825317908022,
+      "grad_norm": 3.5147127451396605,
+      "learning_rate": 9.145595183470245e-07,
+      "loss": 0.0833,
+      "step": 1335
+    },
+    {
+      "epoch": 0.2554737546610575,
+      "grad_norm": 2.7085781769240174,
+      "learning_rate": 9.146546742014474e-07,
+      "loss": 0.1253,
+      "step": 1336
+    },
+    {
+      "epoch": 0.25566497753131273,
+      "grad_norm": 3.5122594024686578,
+      "learning_rate": 9.147497588580676e-07,
+      "loss": 0.1343,
+      "step": 1337
+    },
+    {
+      "epoch": 0.25585620040156803,
+      "grad_norm": 4.806856686085151,
+      "learning_rate": 9.148447724233491e-07,
+      "loss": 0.2369,
+      "step": 1338
+    },
+    {
+      "epoch": 0.2560474232718233,
+      "grad_norm": 5.708207883175667,
+      "learning_rate": 9.149397150035173e-07,
+      "loss": 0.524,
+      "step": 1339
+    },
+    {
+      "epoch": 0.2562386461420786,
+      "grad_norm": 4.200737143723608,
+      "learning_rate": 9.150345867045598e-07,
+      "loss": 0.5379,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2564298690123339,
+      "grad_norm": 1.947054408641099,
+      "learning_rate": 9.151293876322268e-07,
+      "loss": 0.2462,
+      "step": 1341
+    },
+    {
+      "epoch": 0.25662109188258914,
+      "grad_norm": 2.6265671003257136,
+      "learning_rate": 9.152241178920322e-07,
+      "loss": 0.1933,
+      "step": 1342
+    },
+    {
+      "epoch": 0.25681231475284444,
+      "grad_norm": 3.0973988602869715,
+      "learning_rate": 9.153187775892537e-07,
+      "loss": 0.1092,
+      "step": 1343
+    },
+    {
+      "epoch": 0.25700353762309974,
+      "grad_norm": 3.111105279311512,
+      "learning_rate": 9.154133668289347e-07,
+      "loss": 0.2484,
+      "step": 1344
+    },
+    {
+      "epoch": 0.257194760493355,
+      "grad_norm": 4.081697616449424,
+      "learning_rate": 9.155078857158833e-07,
+      "loss": 0.1837,
+      "step": 1345
+    },
+    {
+      "epoch": 0.2573859833636103,
+      "grad_norm": 3.7093595020770653,
+      "learning_rate": 9.156023343546747e-07,
+      "loss": 0.0801,
+      "step": 1346
+    },
+    {
+      "epoch": 0.2575772062338656,
+      "grad_norm": 2.783572138365971,
+      "learning_rate": 9.156967128496507e-07,
+      "loss": 0.0988,
+      "step": 1347
+    },
+    {
+      "epoch": 0.25776842910412084,
+      "grad_norm": 4.6978996372407655,
+      "learning_rate": 9.157910213049209e-07,
+      "loss": 0.1814,
+      "step": 1348
+    },
+    {
+      "epoch": 0.25795965197437615,
+      "grad_norm": 4.486347043998529,
+      "learning_rate": 9.15885259824363e-07,
+      "loss": 0.2382,
+      "step": 1349
+    },
+    {
+      "epoch": 0.2581508748446314,
+      "grad_norm": 5.942508552618937,
+      "learning_rate": 9.159794285116245e-07,
+      "loss": 0.2974,
+      "step": 1350
+    },
+    {
+      "epoch": 0.2583420977148867,
+      "grad_norm": 6.630276576008367,
+      "learning_rate": 9.160735274701218e-07,
+      "loss": 0.2436,
+      "step": 1351
+    },
+    {
+      "epoch": 0.258533320585142,
+      "grad_norm": 3.195496644774126,
+      "learning_rate": 9.161675568030422e-07,
+      "loss": 0.5836,
+      "step": 1352
+    },
+    {
+      "epoch": 0.25872454345539725,
+      "grad_norm": 3.759062434240948,
+      "learning_rate": 9.162615166133443e-07,
+      "loss": 0.39,
+      "step": 1353
+    },
+    {
+      "epoch": 0.25891576632565255,
+      "grad_norm": 3.0010394838628187,
+      "learning_rate": 9.163554070037579e-07,
+      "loss": 0.0986,
+      "step": 1354
+    },
+    {
+      "epoch": 0.25910698919590786,
+      "grad_norm": 2.7670359487621305,
+      "learning_rate": 9.16449228076786e-07,
+      "loss": 0.2618,
+      "step": 1355
+    },
+    {
+      "epoch": 0.2592982120661631,
+      "grad_norm": 4.983810919051351,
+      "learning_rate": 9.165429799347039e-07,
+      "loss": 0.2668,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2594894349364184,
+      "grad_norm": 4.807640098875175,
+      "learning_rate": 9.166366626795616e-07,
+      "loss": 0.3077,
+      "step": 1357
+    },
+    {
+      "epoch": 0.25968065780667365,
+      "grad_norm": 2.792139060822928,
+      "learning_rate": 9.167302764131829e-07,
+      "loss": 0.3285,
+      "step": 1358
+    },
+    {
+      "epoch": 0.25987188067692896,
+      "grad_norm": 2.5428045779439183,
+      "learning_rate": 9.16823821237167e-07,
+      "loss": 0.0837,
+      "step": 1359
+    },
+    {
+      "epoch": 0.26006310354718426,
+      "grad_norm": 5.939437389509738,
+      "learning_rate": 9.169172972528889e-07,
+      "loss": 0.2151,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2602543264174395,
+      "grad_norm": 2.9265826646687514,
+      "learning_rate": 9.170107045615001e-07,
+      "loss": 0.065,
+      "step": 1361
+    },
+    {
+      "epoch": 0.2604455492876948,
+      "grad_norm": 3.3458359531061377,
+      "learning_rate": 9.171040432639292e-07,
+      "loss": 0.2695,
+      "step": 1362
+    },
+    {
+      "epoch": 0.2606367721579501,
+      "grad_norm": 9.00328682115506,
+      "learning_rate": 9.171973134608825e-07,
+      "loss": 0.3906,
+      "step": 1363
+    },
+    {
+      "epoch": 0.26082799502820536,
+      "grad_norm": 8.429048501615004,
+      "learning_rate": 9.172905152528446e-07,
+      "loss": 0.4509,
+      "step": 1364
+    },
+    {
+      "epoch": 0.26101921789846066,
+      "grad_norm": 3.3597392040039207,
+      "learning_rate": 9.173836487400796e-07,
+      "loss": 0.3582,
+      "step": 1365
+    },
+    {
+      "epoch": 0.2612104407687159,
+      "grad_norm": 4.821946224297265,
+      "learning_rate": 9.174767140226308e-07,
+      "loss": 0.7812,
+      "step": 1366
+    },
+    {
+      "epoch": 0.2614016636389712,
+      "grad_norm": 4.196198950517506,
+      "learning_rate": 9.175697112003224e-07,
+      "loss": 0.2161,
+      "step": 1367
+    },
+    {
+      "epoch": 0.2615928865092265,
+      "grad_norm": 5.3023471529314445,
+      "learning_rate": 9.17662640372759e-07,
+      "loss": 0.3288,
+      "step": 1368
+    },
+    {
+      "epoch": 0.26178410937948177,
+      "grad_norm": 4.213472357118202,
+      "learning_rate": 9.177555016393273e-07,
+      "loss": 0.3721,
+      "step": 1369
+    },
+    {
+      "epoch": 0.26197533224973707,
+      "grad_norm": 2.220979027564499,
+      "learning_rate": 9.178482950991964e-07,
+      "loss": 0.3332,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2621665551199924,
+      "grad_norm": 6.523600648221844,
+      "learning_rate": 9.179410208513178e-07,
+      "loss": 0.4091,
+      "step": 1371
+    },
+    {
+      "epoch": 0.2623577779902476,
+      "grad_norm": 3.6357496631923545,
+      "learning_rate": 9.18033678994427e-07,
+      "loss": 0.1123,
+      "step": 1372
+    },
+    {
+      "epoch": 0.2625490008605029,
+      "grad_norm": 2.8409579477443963,
+      "learning_rate": 9.181262696270433e-07,
+      "loss": 0.0774,
+      "step": 1373
+    },
+    {
+      "epoch": 0.26274022373075817,
+      "grad_norm": 6.932885912372611,
+      "learning_rate": 9.182187928474713e-07,
+      "loss": 0.1524,
+      "step": 1374
+    },
+    {
+      "epoch": 0.2629314466010135,
+      "grad_norm": 13.379977717131576,
+      "learning_rate": 9.183112487538006e-07,
+      "loss": 0.1702,
+      "step": 1375
+    },
+    {
+      "epoch": 0.2631226694712688,
+      "grad_norm": 3.56830565564039,
+      "learning_rate": 9.18403637443907e-07,
+      "loss": 0.4685,
+      "step": 1376
+    },
+    {
+      "epoch": 0.263313892341524,
+      "grad_norm": 4.021146429386448,
+      "learning_rate": 9.184959590154532e-07,
+      "loss": 0.4381,
+      "step": 1377
+    },
+    {
+      "epoch": 0.26350511521177933,
+      "grad_norm": 3.529597748269256,
+      "learning_rate": 9.18588213565889e-07,
+      "loss": 0.3432,
+      "step": 1378
+    },
+    {
+      "epoch": 0.26369633808203463,
+      "grad_norm": 2.9573342333116646,
+      "learning_rate": 9.186804011924521e-07,
+      "loss": 0.2637,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2638875609522899,
+      "grad_norm": 4.694336242231372,
+      "learning_rate": 9.187725219921686e-07,
+      "loss": 0.5191,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2640787838225452,
+      "grad_norm": 3.871625230567178,
+      "learning_rate": 9.188645760618543e-07,
+      "loss": 0.3963,
+      "step": 1381
+    },
+    {
+      "epoch": 0.2642700066928005,
+      "grad_norm": 3.029911492946672,
+      "learning_rate": 9.189565634981141e-07,
+      "loss": 0.1242,
+      "step": 1382
+    },
+    {
+      "epoch": 0.26446122956305573,
+      "grad_norm": 3.65100268870547,
+      "learning_rate": 9.190484843973437e-07,
+      "loss": 0.214,
+      "step": 1383
+    },
+    {
+      "epoch": 0.26465245243331104,
+      "grad_norm": 5.200547871704822,
+      "learning_rate": 9.191403388557297e-07,
+      "loss": 0.2868,
+      "step": 1384
+    },
+    {
+      "epoch": 0.2648436753035663,
+      "grad_norm": 2.5707501905486287,
+      "learning_rate": 9.1923212696925e-07,
+      "loss": 0.104,
+      "step": 1385
+    },
+    {
+      "epoch": 0.2650348981738216,
+      "grad_norm": 6.871020743376358,
+      "learning_rate": 9.193238488336751e-07,
+      "loss": 0.2117,
+      "step": 1386
+    },
+    {
+      "epoch": 0.2652261210440769,
+      "grad_norm": 7.659528505241521,
+      "learning_rate": 9.194155045445679e-07,
+      "loss": 0.1783,
+      "step": 1387
+    },
+    {
+      "epoch": 0.26541734391433214,
+      "grad_norm": 5.450244919059544,
+      "learning_rate": 9.195070941972852e-07,
+      "loss": 0.3068,
+      "step": 1388
+    },
+    {
+      "epoch": 0.26560856678458744,
+      "grad_norm": 4.730784045799203,
+      "learning_rate": 9.195986178869771e-07,
+      "loss": 0.4272,
+      "step": 1389
+    },
+    {
+      "epoch": 0.26579978965484274,
+      "grad_norm": 3.9554988684384393,
+      "learning_rate": 9.196900757085885e-07,
+      "loss": 0.385,
+      "step": 1390
+    },
+    {
+      "epoch": 0.265991012525098,
+      "grad_norm": 4.523722110945869,
+      "learning_rate": 9.197814677568599e-07,
+      "loss": 0.3289,
+      "step": 1391
+    },
+    {
+      "epoch": 0.2661822353953533,
+      "grad_norm": 4.984473918064791,
+      "learning_rate": 9.198727941263271e-07,
+      "loss": 0.2927,
+      "step": 1392
+    },
+    {
+      "epoch": 0.26637345826560854,
+      "grad_norm": 3.253514810123037,
+      "learning_rate": 9.199640549113221e-07,
+      "loss": 0.2622,
+      "step": 1393
+    },
+    {
+      "epoch": 0.26656468113586385,
+      "grad_norm": 3.4388113121632045,
+      "learning_rate": 9.200552502059745e-07,
+      "loss": 0.2665,
+      "step": 1394
+    },
+    {
+      "epoch": 0.26675590400611915,
+      "grad_norm": 3.2175180985090663,
+      "learning_rate": 9.201463801042102e-07,
+      "loss": 0.1298,
+      "step": 1395
+    },
+    {
+      "epoch": 0.2669471268763744,
+      "grad_norm": 3.2021177615563285,
+      "learning_rate": 9.202374446997546e-07,
+      "loss": 0.1054,
+      "step": 1396
+    },
+    {
+      "epoch": 0.2671383497466297,
+      "grad_norm": 4.207717372018716,
+      "learning_rate": 9.203284440861308e-07,
+      "loss": 0.2541,
+      "step": 1397
+    },
+    {
+      "epoch": 0.267329572616885,
+      "grad_norm": 3.0676784719018735,
+      "learning_rate": 9.204193783566615e-07,
+      "loss": 0.1275,
+      "step": 1398
+    },
+    {
+      "epoch": 0.26752079548714025,
+      "grad_norm": 2.345503392796288,
+      "learning_rate": 9.205102476044692e-07,
+      "loss": 0.0692,
+      "step": 1399
+    },
+    {
+      "epoch": 0.26771201835739555,
+      "grad_norm": 5.8964135341678645,
+      "learning_rate": 9.206010519224763e-07,
+      "loss": 0.176,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2679032412276508,
+      "grad_norm": 33.86183186460181,
+      "learning_rate": 9.206917914034071e-07,
+      "loss": 0.2145,
+      "step": 1401
+    },
+    {
+      "epoch": 0.2680944640979061,
+      "grad_norm": 6.585945029837036,
+      "learning_rate": 9.207824661397863e-07,
+      "loss": 0.3028,
+      "step": 1402
+    },
+    {
+      "epoch": 0.2682856869681614,
+      "grad_norm": 3.297547330596177,
+      "learning_rate": 9.20873076223942e-07,
+      "loss": 0.2652,
+      "step": 1403
+    },
+    {
+      "epoch": 0.26847690983841666,
+      "grad_norm": 3.028402820042587,
+      "learning_rate": 9.209636217480034e-07,
+      "loss": 0.3933,
+      "step": 1404
+    },
+    {
+      "epoch": 0.26866813270867196,
+      "grad_norm": 4.421340964052666,
+      "learning_rate": 9.210541028039043e-07,
+      "loss": 0.3852,
+      "step": 1405
+    },
+    {
+      "epoch": 0.26885935557892726,
+      "grad_norm": 3.148713736330976,
+      "learning_rate": 9.211445194833814e-07,
+      "loss": 0.3528,
+      "step": 1406
+    },
+    {
+      "epoch": 0.2690505784491825,
+      "grad_norm": 5.234225780580613,
+      "learning_rate": 9.212348718779762e-07,
+      "loss": 0.4719,
+      "step": 1407
+    },
+    {
+      "epoch": 0.2692418013194378,
+      "grad_norm": 2.6752529987232485,
+      "learning_rate": 9.213251600790346e-07,
+      "loss": 0.1235,
+      "step": 1408
+    },
+    {
+      "epoch": 0.26943302418969306,
+      "grad_norm": 2.8745581245856022,
+      "learning_rate": 9.214153841777085e-07,
+      "loss": 0.1013,
+      "step": 1409
+    },
+    {
+      "epoch": 0.26962424705994836,
+      "grad_norm": 5.458065560254308,
+      "learning_rate": 9.215055442649556e-07,
+      "loss": 0.1721,
+      "step": 1410
+    },
+    {
+      "epoch": 0.26981546993020367,
+      "grad_norm": 3.5339199538866914,
+      "learning_rate": 9.2159564043154e-07,
+      "loss": 0.3509,
+      "step": 1411
+    },
+    {
+      "epoch": 0.2700066928004589,
+      "grad_norm": 3.494269311623785,
+      "learning_rate": 9.216856727680326e-07,
+      "loss": 0.106,
+      "step": 1412
+    },
+    {
+      "epoch": 0.2701979156707142,
+      "grad_norm": 7.431509731107464,
+      "learning_rate": 9.217756413648129e-07,
+      "loss": 0.4209,
+      "step": 1413
+    },
+    {
+      "epoch": 0.2703891385409695,
+      "grad_norm": 6.008010285573148,
+      "learning_rate": 9.218655463120676e-07,
+      "loss": 0.787,
+      "step": 1414
+    },
+    {
+      "epoch": 0.27058036141122477,
+      "grad_norm": 4.6516852821698365,
+      "learning_rate": 9.219553876997927e-07,
+      "loss": 0.1811,
+      "step": 1415
+    },
+    {
+      "epoch": 0.27077158428148007,
+      "grad_norm": 6.769042951139048,
+      "learning_rate": 9.220451656177931e-07,
+      "loss": 0.6349,
+      "step": 1416
+    },
+    {
+      "epoch": 0.2709628071517354,
+      "grad_norm": 2.1892835702843394,
+      "learning_rate": 9.221348801556837e-07,
+      "loss": 0.3043,
+      "step": 1417
+    },
+    {
+      "epoch": 0.2711540300219906,
+      "grad_norm": 3.5463497092180294,
+      "learning_rate": 9.222245314028899e-07,
+      "loss": 0.4003,
+      "step": 1418
+    },
+    {
+      "epoch": 0.2713452528922459,
+      "grad_norm": 3.934564374103548,
+      "learning_rate": 9.223141194486475e-07,
+      "loss": 0.1325,
+      "step": 1419
+    },
+    {
+      "epoch": 0.2715364757625012,
+      "grad_norm": 2.4319002434198063,
+      "learning_rate": 9.224036443820044e-07,
+      "loss": 0.1743,
+      "step": 1420
+    },
+    {
+      "epoch": 0.2717276986327565,
+      "grad_norm": 3.8089784711538313,
+      "learning_rate": 9.224931062918194e-07,
+      "loss": 0.4653,
+      "step": 1421
+    },
+    {
+      "epoch": 0.2719189215030118,
+      "grad_norm": 5.049406662634069,
+      "learning_rate": 9.225825052667649e-07,
+      "loss": 0.2979,
+      "step": 1422
+    },
+    {
+      "epoch": 0.272110144373267,
+      "grad_norm": 4.089381092503676,
+      "learning_rate": 9.226718413953257e-07,
+      "loss": 0.2729,
+      "step": 1423
+    },
+    {
+      "epoch": 0.27230136724352233,
+      "grad_norm": 3.163288389100366,
+      "learning_rate": 9.227611147658002e-07,
+      "loss": 0.0661,
+      "step": 1424
+    },
+    {
+      "epoch": 0.27249259011377763,
+      "grad_norm": 4.88614534692925,
+      "learning_rate": 9.228503254663008e-07,
+      "loss": 0.189,
+      "step": 1425
+    },
+    {
+      "epoch": 0.2726838129840329,
+      "grad_norm": 4.342048586242785,
+      "learning_rate": 9.229394735847544e-07,
+      "loss": 0.4163,
+      "step": 1426
+    },
+    {
+      "epoch": 0.2728750358542882,
+      "grad_norm": 4.728302236640085,
+      "learning_rate": 9.230285592089034e-07,
+      "loss": 0.5886,
+      "step": 1427
+    },
+    {
+      "epoch": 0.27306625872454343,
+      "grad_norm": 5.752154734369962,
+      "learning_rate": 9.231175824263052e-07,
+      "loss": 0.6143,
+      "step": 1428
+    },
+    {
+      "epoch": 0.27325748159479873,
+      "grad_norm": 3.1294978007637826,
+      "learning_rate": 9.232065433243338e-07,
+      "loss": 0.3548,
+      "step": 1429
+    },
+    {
+      "epoch": 0.27344870446505404,
+      "grad_norm": 2.55270093853628,
+      "learning_rate": 9.232954419901795e-07,
+      "loss": 0.2944,
+      "step": 1430
+    },
+    {
+      "epoch": 0.2736399273353093,
+      "grad_norm": 3.3991135900315266,
+      "learning_rate": 9.233842785108501e-07,
+      "loss": 0.3526,
+      "step": 1431
+    },
+    {
+      "epoch": 0.2738311502055646,
+      "grad_norm": 3.013048721927345,
+      "learning_rate": 9.234730529731709e-07,
+      "loss": 0.1239,
+      "step": 1432
+    },
+    {
+      "epoch": 0.2740223730758199,
+      "grad_norm": 1.551971055989889,
+      "learning_rate": 9.235617654637852e-07,
+      "loss": 0.2222,
+      "step": 1433
+    },
+    {
+      "epoch": 0.27421359594607514,
+      "grad_norm": 4.910614987645108,
+      "learning_rate": 9.236504160691549e-07,
+      "loss": 0.3565,
+      "step": 1434
+    },
+    {
+      "epoch": 0.27440481881633044,
+      "grad_norm": 4.984616456117723,
+      "learning_rate": 9.237390048755617e-07,
+      "loss": 0.2691,
+      "step": 1435
+    },
+    {
+      "epoch": 0.2745960416865857,
+      "grad_norm": 5.070187418707067,
+      "learning_rate": 9.238275319691065e-07,
+      "loss": 0.1022,
+      "step": 1436
+    },
+    {
+      "epoch": 0.274787264556841,
+      "grad_norm": 5.920082324203085,
+      "learning_rate": 9.239159974357101e-07,
+      "loss": 0.1611,
+      "step": 1437
+    },
+    {
+      "epoch": 0.2749784874270963,
+      "grad_norm": 4.897186265509185,
+      "learning_rate": 9.240044013611146e-07,
+      "loss": 0.2267,
+      "step": 1438
+    },
+    {
+      "epoch": 0.27516971029735154,
+      "grad_norm": 4.683617777122482,
+      "learning_rate": 9.240927438308831e-07,
+      "loss": 0.6445,
+      "step": 1439
+    },
+    {
+      "epoch": 0.27536093316760685,
+      "grad_norm": 4.560704126036257,
+      "learning_rate": 9.241810249304003e-07,
+      "loss": 0.2637,
+      "step": 1440
+    },
+    {
+      "epoch": 0.27555215603786215,
+      "grad_norm": 3.174861610580513,
+      "learning_rate": 9.242692447448727e-07,
+      "loss": 0.3734,
+      "step": 1441
+    },
+    {
+      "epoch": 0.2757433789081174,
+      "grad_norm": 4.319348412223395,
+      "learning_rate": 9.243574033593304e-07,
+      "loss": 0.2233,
+      "step": 1442
+    },
+    {
+      "epoch": 0.2759346017783727,
+      "grad_norm": 4.247320340169963,
+      "learning_rate": 9.244455008586258e-07,
+      "loss": 0.3585,
+      "step": 1443
+    },
+    {
+      "epoch": 0.27612582464862795,
+      "grad_norm": 3.4225183169625035,
+      "learning_rate": 9.245335373274352e-07,
+      "loss": 0.1765,
+      "step": 1444
+    },
+    {
+      "epoch": 0.27631704751888325,
+      "grad_norm": 3.044928452225898,
+      "learning_rate": 9.246215128502595e-07,
+      "loss": 0.1294,
+      "step": 1445
+    },
+    {
+      "epoch": 0.27650827038913856,
+      "grad_norm": 5.409658250714768,
+      "learning_rate": 9.247094275114235e-07,
+      "loss": 0.2514,
+      "step": 1446
+    },
+    {
+      "epoch": 0.2766994932593938,
+      "grad_norm": 2.9612928997521606,
+      "learning_rate": 9.247972813950773e-07,
+      "loss": 0.1215,
+      "step": 1447
+    },
+    {
+      "epoch": 0.2768907161296491,
+      "grad_norm": 3.0287882453496904,
+      "learning_rate": 9.248850745851972e-07,
+      "loss": 0.0889,
+      "step": 1448
+    },
+    {
+      "epoch": 0.2770819389999044,
+      "grad_norm": 2.7163140795575336,
+      "learning_rate": 9.249728071655849e-07,
+      "loss": 0.0404,
+      "step": 1449
+    },
+    {
+      "epoch": 0.27727316187015966,
+      "grad_norm": 6.939371767219396,
+      "learning_rate": 9.250604792198688e-07,
+      "loss": 0.1566,
+      "step": 1450
+    },
+    {
+      "epoch": 0.27746438474041496,
+      "grad_norm": 45.117970096026454,
+      "learning_rate": 9.251480908315046e-07,
+      "loss": 0.3924,
+      "step": 1451
+    },
+    {
+      "epoch": 0.27765560761067026,
+      "grad_norm": 4.061700360754919,
+      "learning_rate": 9.25235642083775e-07,
+      "loss": 0.2831,
+      "step": 1452
+    },
+    {
+      "epoch": 0.2778468304809255,
+      "grad_norm": 3.7536473656341545,
+      "learning_rate": 9.253231330597913e-07,
+      "loss": 0.3318,
+      "step": 1453
+    },
+    {
+      "epoch": 0.2780380533511808,
+      "grad_norm": 4.364203510130117,
+      "learning_rate": 9.25410563842493e-07,
+      "loss": 0.4813,
+      "step": 1454
+    },
+    {
+      "epoch": 0.27822927622143606,
+      "grad_norm": 2.9077843338280203,
+      "learning_rate": 9.254979345146485e-07,
+      "loss": 0.2561,
+      "step": 1455
+    },
+    {
+      "epoch": 0.27842049909169136,
+      "grad_norm": 4.432839323546906,
+      "learning_rate": 9.255852451588552e-07,
+      "loss": 0.2173,
+      "step": 1456
+    },
+    {
+      "epoch": 0.27861172196194667,
+      "grad_norm": 5.772001880352008,
+      "learning_rate": 9.256724958575414e-07,
+      "loss": 0.3879,
+      "step": 1457
+    },
+    {
+      "epoch": 0.2788029448322019,
+      "grad_norm": 3.3465007269958407,
+      "learning_rate": 9.257596866929645e-07,
+      "loss": 0.1935,
+      "step": 1458
+    },
+    {
+      "epoch": 0.2789941677024572,
+      "grad_norm": 6.754851681586678,
+      "learning_rate": 9.258468177472137e-07,
+      "loss": 0.1983,
+      "step": 1459
+    },
+    {
+      "epoch": 0.2791853905727125,
+      "grad_norm": 3.7357788001176147,
+      "learning_rate": 9.259338891022091e-07,
+      "loss": 0.1223,
+      "step": 1460
+    },
+    {
+      "epoch": 0.27937661344296777,
+      "grad_norm": 7.3240177055740725,
+      "learning_rate": 9.260209008397025e-07,
+      "loss": 0.1578,
+      "step": 1461
+    },
+    {
+      "epoch": 0.2795678363132231,
+      "grad_norm": 3.240337386284509,
+      "learning_rate": 9.261078530412776e-07,
+      "loss": 0.114,
+      "step": 1462
+    },
+    {
+      "epoch": 0.2797590591834783,
+      "grad_norm": 11.709452692104477,
+      "learning_rate": 9.261947457883515e-07,
+      "loss": 0.4321,
+      "step": 1463
+    },
+    {
+      "epoch": 0.2799502820537336,
+      "grad_norm": 4.871658549415663,
+      "learning_rate": 9.262815791621734e-07,
+      "loss": 0.3099,
+      "step": 1464
+    },
+    {
+      "epoch": 0.2801415049239889,
+      "grad_norm": 5.269866050451118,
+      "learning_rate": 9.263683532438271e-07,
+      "loss": 0.6329,
+      "step": 1465
+    },
+    {
+      "epoch": 0.2803327277942442,
+      "grad_norm": 3.156480686331084,
+      "learning_rate": 9.264550681142298e-07,
+      "loss": 0.4675,
+      "step": 1466
+    },
+    {
+      "epoch": 0.2805239506644995,
+      "grad_norm": 2.5590316747126476,
+      "learning_rate": 9.265417238541332e-07,
+      "loss": 0.3459,
+      "step": 1467
+    },
+    {
+      "epoch": 0.2807151735347548,
+      "grad_norm": 5.583061534282159,
+      "learning_rate": 9.266283205441241e-07,
+      "loss": 0.5001,
+      "step": 1468
+    },
+    {
+      "epoch": 0.28090639640501003,
+      "grad_norm": 2.706924502427214,
+      "learning_rate": 9.267148582646246e-07,
+      "loss": 0.3894,
+      "step": 1469
+    },
+    {
+      "epoch": 0.28109761927526533,
+      "grad_norm": 3.6736233059432926,
+      "learning_rate": 9.268013370958925e-07,
+      "loss": 0.3753,
+      "step": 1470
+    },
+    {
+      "epoch": 0.2812888421455206,
+      "grad_norm": 2.1413800098435627,
+      "learning_rate": 9.268877571180222e-07,
+      "loss": 0.3059,
+      "step": 1471
+    },
+    {
+      "epoch": 0.2814800650157759,
+      "grad_norm": 3.8876599883003333,
+      "learning_rate": 9.269741184109443e-07,
+      "loss": 0.1065,
+      "step": 1472
+    },
+    {
+      "epoch": 0.2816712878860312,
+      "grad_norm": 4.089158839722357,
+      "learning_rate": 9.270604210544271e-07,
+      "loss": 0.2686,
+      "step": 1473
+    },
+    {
+      "epoch": 0.28186251075628643,
+      "grad_norm": 3.0988493229689054,
+      "learning_rate": 9.27146665128076e-07,
+      "loss": 0.0728,
+      "step": 1474
+    },
+    {
+      "epoch": 0.28205373362654174,
+      "grad_norm": 4.6736410113898055,
+      "learning_rate": 9.272328507113348e-07,
+      "loss": 0.1168,
+      "step": 1475
+    },
+    {
+      "epoch": 0.28224495649679704,
+      "grad_norm": 7.663073181320624,
+      "learning_rate": 9.273189778834857e-07,
+      "loss": 0.6337,
+      "step": 1476
+    },
+    {
+      "epoch": 0.2824361793670523,
+      "grad_norm": 4.5571504349395715,
+      "learning_rate": 9.274050467236495e-07,
+      "loss": 0.3434,
+      "step": 1477
+    },
+    {
+      "epoch": 0.2826274022373076,
+      "grad_norm": 3.77715577976431,
+      "learning_rate": 9.274910573107867e-07,
+      "loss": 0.4867,
+      "step": 1478
+    },
+    {
+      "epoch": 0.28281862510756284,
+      "grad_norm": 3.5949544463899197,
+      "learning_rate": 9.275770097236977e-07,
+      "loss": 0.1223,
+      "step": 1479
+    },
+    {
+      "epoch": 0.28300984797781814,
+      "grad_norm": 3.611471516809129,
+      "learning_rate": 9.276629040410226e-07,
+      "loss": 0.321,
+      "step": 1480
+    },
+    {
+      "epoch": 0.28320107084807344,
+      "grad_norm": 4.069026458301712,
+      "learning_rate": 9.277487403412426e-07,
+      "loss": 0.5484,
+      "step": 1481
+    },
+    {
+      "epoch": 0.2833922937183287,
+      "grad_norm": 4.402067323009565,
+      "learning_rate": 9.278345187026797e-07,
+      "loss": 0.3825,
+      "step": 1482
+    },
+    {
+      "epoch": 0.283583516588584,
+      "grad_norm": 2.0955752487488835,
+      "learning_rate": 9.279202392034976e-07,
+      "loss": 0.1972,
+      "step": 1483
+    },
+    {
+      "epoch": 0.2837747394588393,
+      "grad_norm": 3.011443089636047,
+      "learning_rate": 9.28005901921702e-07,
+      "loss": 0.1704,
+      "step": 1484
+    },
+    {
+      "epoch": 0.28396596232909455,
+      "grad_norm": 4.103579070859474,
+      "learning_rate": 9.280915069351407e-07,
+      "loss": 0.1338,
+      "step": 1485
+    },
+    {
+      "epoch": 0.28415718519934985,
+      "grad_norm": 3.8313525237512254,
+      "learning_rate": 9.281770543215044e-07,
+      "loss": 0.3773,
+      "step": 1486
+    },
+    {
+      "epoch": 0.28434840806960515,
+      "grad_norm": 3.4436548791061514,
+      "learning_rate": 9.282625441583269e-07,
+      "loss": 0.0985,
+      "step": 1487
+    },
+    {
+      "epoch": 0.2845396309398604,
+      "grad_norm": 14.75973241382561,
+      "learning_rate": 9.283479765229859e-07,
+      "loss": 0.3179,
+      "step": 1488
+    },
+    {
+      "epoch": 0.2847308538101157,
+      "grad_norm": 5.819722927878559,
+      "learning_rate": 9.284333514927031e-07,
+      "loss": 0.6252,
+      "step": 1489
+    },
+    {
+      "epoch": 0.28492207668037095,
+      "grad_norm": 3.381417284086821,
+      "learning_rate": 9.285186691445443e-07,
+      "loss": 0.3271,
+      "step": 1490
+    },
+    {
+      "epoch": 0.28511329955062625,
+      "grad_norm": 6.165339077846751,
+      "learning_rate": 9.286039295554204e-07,
+      "loss": 0.32,
+      "step": 1491
+    },
+    {
+      "epoch": 0.28530452242088156,
+      "grad_norm": 4.413593422600728,
+      "learning_rate": 9.286891328020881e-07,
+      "loss": 0.46,
+      "step": 1492
+    },
+    {
+      "epoch": 0.2854957452911368,
+      "grad_norm": 2.7712702490342536,
+      "learning_rate": 9.287742789611489e-07,
+      "loss": 0.4192,
+      "step": 1493
+    },
+    {
+      "epoch": 0.2856869681613921,
+      "grad_norm": 3.8009527467652227,
+      "learning_rate": 9.288593681090512e-07,
+      "loss": 0.279,
+      "step": 1494
+    },
+    {
+      "epoch": 0.2858781910316474,
+      "grad_norm": 2.5815023783780506,
+      "learning_rate": 9.289444003220893e-07,
+      "loss": 0.0828,
+      "step": 1495
+    },
+    {
+      "epoch": 0.28606941390190266,
+      "grad_norm": 3.6145326169255076,
+      "learning_rate": 9.29029375676405e-07,
+      "loss": 0.197,
+      "step": 1496
+    },
+    {
+      "epoch": 0.28626063677215796,
+      "grad_norm": 2.8315311384932422,
+      "learning_rate": 9.291142942479875e-07,
+      "loss": 0.1014,
+      "step": 1497
+    },
+    {
+      "epoch": 0.2864518596424132,
+      "grad_norm": 4.111536422128385,
+      "learning_rate": 9.29199156112673e-07,
+      "loss": 0.1346,
+      "step": 1498
+    },
+    {
+      "epoch": 0.2866430825126685,
+      "grad_norm": 6.5845985001704745,
+      "learning_rate": 9.292839613461466e-07,
+      "loss": 0.2266,
+      "step": 1499
+    },
+    {
+      "epoch": 0.2868343053829238,
+      "grad_norm": 4.861739684140612,
+      "learning_rate": 9.293687100239419e-07,
+      "loss": 0.2155,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2868343053829238,
+      "eval_runtime": 727.3009,
+      "eval_samples_per_second": 2.109,
+      "eval_steps_per_second": 0.528,
+      "step": 1500
+    },
+    {
+      "epoch": 0.28702552825317906,
+      "grad_norm": 3.6943577506873204,
+      "learning_rate": 9.294534022214413e-07,
+      "loss": 0.3501,
+      "step": 1501
+    },
+    {
+      "epoch": 0.28721675112343437,
+      "grad_norm": 4.074966786255077,
+      "learning_rate": 9.295380380138764e-07,
+      "loss": 0.4441,
+      "step": 1502
+    },
+    {
+      "epoch": 0.28740797399368967,
+      "grad_norm": 4.3349854669910295,
+      "learning_rate": 9.296226174763292e-07,
+      "loss": 0.3406,
+      "step": 1503
+    },
+    {
+      "epoch": 0.2875991968639449,
+      "grad_norm": 2.9108500837631683,
+      "learning_rate": 9.297071406837314e-07,
+      "loss": 0.2916,
+      "step": 1504
+    },
+    {
+      "epoch": 0.2877904197342002,
+      "grad_norm": 4.823595805237994,
+      "learning_rate": 9.297916077108651e-07,
+      "loss": 0.7132,
+      "step": 1505
+    },
+    {
+      "epoch": 0.28798164260445547,
+      "grad_norm": 2.128853277544799,
+      "learning_rate": 9.29876018632364e-07,
+      "loss": 0.1237,
+      "step": 1506
+    },
+    {
+      "epoch": 0.28817286547471077,
+      "grad_norm": 2.8847573602412075,
+      "learning_rate": 9.299603735227127e-07,
+      "loss": 0.1397,
+      "step": 1507
+    },
+    {
+      "epoch": 0.2883640883449661,
+      "grad_norm": 2.8152665836274986,
+      "learning_rate": 9.300446724562479e-07,
+      "loss": 0.2468,
+      "step": 1508
+    },
+    {
+      "epoch": 0.2885553112152213,
+      "grad_norm": 3.8416922418787056,
+      "learning_rate": 9.30128915507158e-07,
+      "loss": 0.2996,
+      "step": 1509
+    },
+    {
+      "epoch": 0.2887465340854766,
+      "grad_norm": 4.332475919750142,
+      "learning_rate": 9.302131027494845e-07,
+      "loss": 0.2896,
+      "step": 1510
+    },
+    {
+      "epoch": 0.28893775695573193,
+      "grad_norm": 3.4465812717043796,
+      "learning_rate": 9.302972342571214e-07,
+      "loss": 0.1035,
+      "step": 1511
+    },
+    {
+      "epoch": 0.2891289798259872,
+      "grad_norm": 2.9362796927128536,
+      "learning_rate": 9.303813101038164e-07,
+      "loss": 0.0793,
+      "step": 1512
+    },
+    {
+      "epoch": 0.2893202026962425,
+      "grad_norm": 5.713552645936732,
+      "learning_rate": 9.304653303631707e-07,
+      "loss": 0.1177,
+      "step": 1513
+    },
+    {
+      "epoch": 0.2895114255664977,
+      "grad_norm": 6.826731524358055,
+      "learning_rate": 9.305492951086396e-07,
+      "loss": 0.6811,
+      "step": 1514
+    },
+    {
+      "epoch": 0.28970264843675303,
+      "grad_norm": 3.7934260775240696,
+      "learning_rate": 9.306332044135332e-07,
+      "loss": 0.3587,
+      "step": 1515
+    },
+    {
+      "epoch": 0.28989387130700833,
+      "grad_norm": 2.0283541651303554,
+      "learning_rate": 9.30717058351016e-07,
+      "loss": 0.3073,
+      "step": 1516
+    },
+    {
+      "epoch": 0.2900850941772636,
+      "grad_norm": 2.9896111215680086,
+      "learning_rate": 9.308008569941079e-07,
+      "loss": 0.2364,
+      "step": 1517
+    },
+    {
+      "epoch": 0.2902763170475189,
+      "grad_norm": 2.347978566619423,
+      "learning_rate": 9.308846004156848e-07,
+      "loss": 0.229,
+      "step": 1518
+    },
+    {
+      "epoch": 0.2904675399177742,
+      "grad_norm": 2.563723574408716,
+      "learning_rate": 9.309682886884785e-07,
+      "loss": 0.2361,
+      "step": 1519
+    },
+    {
+      "epoch": 0.29065876278802943,
+      "grad_norm": 4.0615641763179875,
+      "learning_rate": 9.310519218850765e-07,
+      "loss": 0.2611,
+      "step": 1520
+    },
+    {
+      "epoch": 0.29084998565828474,
+      "grad_norm": 2.592751919931899,
+      "learning_rate": 9.311355000779241e-07,
+      "loss": 0.23,
+      "step": 1521
+    },
+    {
+      "epoch": 0.29104120852854004,
+      "grad_norm": 5.129385025908911,
+      "learning_rate": 9.312190233393234e-07,
+      "loss": 0.4076,
+      "step": 1522
+    },
+    {
+      "epoch": 0.2912324313987953,
+      "grad_norm": 2.729732910896512,
+      "learning_rate": 9.313024917414335e-07,
+      "loss": 0.0946,
+      "step": 1523
+    },
+    {
+      "epoch": 0.2914236542690506,
+      "grad_norm": 3.879419452258296,
+      "learning_rate": 9.313859053562722e-07,
+      "loss": 0.0895,
+      "step": 1524
+    },
+    {
+      "epoch": 0.29161487713930584,
+      "grad_norm": 3.089370247576572,
+      "learning_rate": 9.314692642557152e-07,
+      "loss": 0.0745,
+      "step": 1525
+    },
+    {
+      "epoch": 0.29180610000956114,
+      "grad_norm": 5.378926462039912,
+      "learning_rate": 9.315525685114967e-07,
+      "loss": 0.2924,
+      "step": 1526
+    },
+    {
+      "epoch": 0.29199732287981645,
+      "grad_norm": 5.093495555677653,
+      "learning_rate": 9.316358181952104e-07,
+      "loss": 0.7183,
+      "step": 1527
+    },
+    {
+      "epoch": 0.2921885457500717,
+      "grad_norm": 4.683004639699211,
+      "learning_rate": 9.317190133783089e-07,
+      "loss": 0.5668,
+      "step": 1528
+    },
+    {
+      "epoch": 0.292379768620327,
+      "grad_norm": 4.263913883849757,
+      "learning_rate": 9.318021541321048e-07,
+      "loss": 0.6363,
+      "step": 1529
+    },
+    {
+      "epoch": 0.2925709914905823,
+      "grad_norm": 2.6991740552897867,
+      "learning_rate": 9.318852405277707e-07,
+      "loss": 0.2912,
+      "step": 1530
+    },
+    {
+      "epoch": 0.29276221436083755,
+      "grad_norm": 3.730017254255135,
+      "learning_rate": 9.3196827263634e-07,
+      "loss": 0.2249,
+      "step": 1531
+    },
+    {
+      "epoch": 0.29295343723109285,
+      "grad_norm": 3.059244722392175,
+      "learning_rate": 9.320512505287063e-07,
+      "loss": 0.21,
+      "step": 1532
+    },
+    {
+      "epoch": 0.2931446601013481,
+      "grad_norm": 5.678819676651291,
+      "learning_rate": 9.321341742756255e-07,
+      "loss": 0.2028,
+      "step": 1533
+    },
+    {
+      "epoch": 0.2933358829716034,
+      "grad_norm": 4.658094002587161,
+      "learning_rate": 9.322170439477138e-07,
+      "loss": 0.2177,
+      "step": 1534
+    },
+    {
+      "epoch": 0.2935271058418587,
+      "grad_norm": 3.2372344272635782,
+      "learning_rate": 9.322998596154502e-07,
+      "loss": 0.0624,
+      "step": 1535
+    },
+    {
+      "epoch": 0.29371832871211395,
+      "grad_norm": 2.1240262437122266,
+      "learning_rate": 9.32382621349176e-07,
+      "loss": 0.0619,
+      "step": 1536
+    },
+    {
+      "epoch": 0.29390955158236926,
+      "grad_norm": 3.613504004958021,
+      "learning_rate": 9.324653292190946e-07,
+      "loss": 0.0899,
+      "step": 1537
+    },
+    {
+      "epoch": 0.29410077445262456,
+      "grad_norm": 7.512791216342403,
+      "learning_rate": 9.325479832952728e-07,
+      "loss": 0.2453,
+      "step": 1538
+    },
+    {
+      "epoch": 0.2942919973228798,
+      "grad_norm": 8.800620872096184,
+      "learning_rate": 9.326305836476409e-07,
+      "loss": 0.7435,
+      "step": 1539
+    },
+    {
+      "epoch": 0.2944832201931351,
+      "grad_norm": 3.125018157906231,
+      "learning_rate": 9.327131303459925e-07,
+      "loss": 0.2439,
+      "step": 1540
+    },
+    {
+      "epoch": 0.29467444306339036,
+      "grad_norm": 3.0336899523299468,
+      "learning_rate": 9.327956234599859e-07,
+      "loss": 0.2071,
+      "step": 1541
+    },
+    {
+      "epoch": 0.29486566593364566,
+      "grad_norm": 1.6304509365276683,
+      "learning_rate": 9.328780630591431e-07,
+      "loss": 0.2662,
+      "step": 1542
+    },
+    {
+      "epoch": 0.29505688880390096,
+      "grad_norm": 3.125391210864212,
+      "learning_rate": 9.329604492128514e-07,
+      "loss": 0.2784,
+      "step": 1543
+    },
+    {
+      "epoch": 0.2952481116741562,
+      "grad_norm": 4.4031452035000465,
+      "learning_rate": 9.330427819903632e-07,
+      "loss": 0.2116,
+      "step": 1544
+    },
+    {
+      "epoch": 0.2954393345444115,
+      "grad_norm": 3.2512675527812327,
+      "learning_rate": 9.33125061460796e-07,
+      "loss": 0.3194,
+      "step": 1545
+    },
+    {
+      "epoch": 0.2956305574146668,
+      "grad_norm": 6.98565075037752,
+      "learning_rate": 9.332072876931336e-07,
+      "loss": 0.1339,
+      "step": 1546
+    },
+    {
+      "epoch": 0.29582178028492206,
+      "grad_norm": 3.23807980563122,
+      "learning_rate": 9.332894607562258e-07,
+      "loss": 0.1066,
+      "step": 1547
+    },
+    {
+      "epoch": 0.29601300315517737,
+      "grad_norm": 4.055596692531285,
+      "learning_rate": 9.333715807187889e-07,
+      "loss": 0.1065,
+      "step": 1548
+    },
+    {
+      "epoch": 0.2962042260254326,
+      "grad_norm": 5.889923620586914,
+      "learning_rate": 9.33453647649406e-07,
+      "loss": 0.1228,
+      "step": 1549
+    },
+    {
+      "epoch": 0.2963954488956879,
+      "grad_norm": 6.768311141308916,
+      "learning_rate": 9.335356616165278e-07,
+      "loss": 0.1612,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2965866717659432,
+      "grad_norm": 6.5310695203457945,
+      "learning_rate": 9.336176226884718e-07,
+      "loss": 0.3052,
+      "step": 1551
+    },
+    {
+      "epoch": 0.29677789463619847,
+      "grad_norm": 5.455217107688438,
+      "learning_rate": 9.336995309334242e-07,
+      "loss": 0.568,
+      "step": 1552
+    },
+    {
+      "epoch": 0.2969691175064538,
+      "grad_norm": 5.61523692255377,
+      "learning_rate": 9.33781386419439e-07,
+      "loss": 0.5922,
+      "step": 1553
+    },
+    {
+      "epoch": 0.2971603403767091,
+      "grad_norm": 3.0773870677664124,
+      "learning_rate": 9.338631892144387e-07,
+      "loss": 0.2647,
+      "step": 1554
+    },
+    {
+      "epoch": 0.2973515632469643,
+      "grad_norm": 4.916850889386115,
+      "learning_rate": 9.339449393862153e-07,
+      "loss": 0.5543,
+      "step": 1555
+    },
+    {
+      "epoch": 0.2975427861172196,
+      "grad_norm": 4.32164094171151,
+      "learning_rate": 9.340266370024293e-07,
+      "loss": 0.3782,
+      "step": 1556
+    },
+    {
+      "epoch": 0.29773400898747493,
+      "grad_norm": 4.715738105807866,
+      "learning_rate": 9.341082821306116e-07,
+      "loss": 0.4974,
+      "step": 1557
+    },
+    {
+      "epoch": 0.2979252318577302,
+      "grad_norm": 2.7303323561583945,
+      "learning_rate": 9.34189874838162e-07,
+      "loss": 0.0813,
+      "step": 1558
+    },
+    {
+      "epoch": 0.2981164547279855,
+      "grad_norm": 3.2826752973061386,
+      "learning_rate": 9.342714151923513e-07,
+      "loss": 0.1884,
+      "step": 1559
+    },
+    {
+      "epoch": 0.29830767759824073,
+      "grad_norm": 5.118175149202122,
+      "learning_rate": 9.343529032603209e-07,
+      "loss": 0.2798,
+      "step": 1560
+    },
+    {
+      "epoch": 0.29849890046849603,
+      "grad_norm": 3.502864074517928,
+      "learning_rate": 9.344343391090829e-07,
+      "loss": 0.091,
+      "step": 1561
+    },
+    {
+      "epoch": 0.29869012333875133,
+      "grad_norm": 6.480004292239427,
+      "learning_rate": 9.345157228055205e-07,
+      "loss": 0.1013,
+      "step": 1562
+    },
+    {
+      "epoch": 0.2988813462090066,
+      "grad_norm": 14.520813834762436,
+      "learning_rate": 9.345970544163889e-07,
+      "loss": 0.5279,
+      "step": 1563
+    },
+    {
+      "epoch": 0.2990725690792619,
+      "grad_norm": 4.829739692576286,
+      "learning_rate": 9.34678334008315e-07,
+      "loss": 0.3181,
+      "step": 1564
+    },
+    {
+      "epoch": 0.2992637919495172,
+      "grad_norm": 4.695012900240302,
+      "learning_rate": 9.347595616477977e-07,
+      "loss": 0.7812,
+      "step": 1565
+    },
+    {
+      "epoch": 0.29945501481977244,
+      "grad_norm": 4.321652416750571,
+      "learning_rate": 9.348407374012089e-07,
+      "loss": 0.4565,
+      "step": 1566
+    },
+    {
+      "epoch": 0.29964623769002774,
+      "grad_norm": 3.0816808000910645,
+      "learning_rate": 9.349218613347932e-07,
+      "loss": 0.1573,
+      "step": 1567
+    },
+    {
+      "epoch": 0.299837460560283,
+      "grad_norm": 6.286473166493823,
+      "learning_rate": 9.350029335146683e-07,
+      "loss": 0.1515,
+      "step": 1568
+    },
+    {
+      "epoch": 0.3000286834305383,
+      "grad_norm": 3.541046866243872,
+      "learning_rate": 9.350839540068256e-07,
+      "loss": 0.367,
+      "step": 1569
+    },
+    {
+      "epoch": 0.3002199063007936,
+      "grad_norm": 3.1725341412666377,
+      "learning_rate": 9.351649228771303e-07,
+      "loss": 0.2905,
+      "step": 1570
+    },
+    {
+      "epoch": 0.30041112917104884,
+      "grad_norm": 6.290923152553059,
+      "learning_rate": 9.352458401913219e-07,
+      "loss": 0.1424,
+      "step": 1571
+    },
+    {
+      "epoch": 0.30060235204130414,
+      "grad_norm": 3.151034739252888,
+      "learning_rate": 9.353267060150141e-07,
+      "loss": 0.0764,
+      "step": 1572
+    },
+    {
+      "epoch": 0.30079357491155945,
+      "grad_norm": 11.149353802349307,
+      "learning_rate": 9.354075204136958e-07,
+      "loss": 0.2671,
+      "step": 1573
+    },
+    {
+      "epoch": 0.3009847977818147,
+      "grad_norm": 4.622946618447126,
+      "learning_rate": 9.354882834527307e-07,
+      "loss": 0.1091,
+      "step": 1574
+    },
+    {
+      "epoch": 0.30117602065207,
+      "grad_norm": 5.154798823666699,
+      "learning_rate": 9.355689951973582e-07,
+      "loss": 0.2243,
+      "step": 1575
+    },
+    {
+      "epoch": 0.30136724352232525,
+      "grad_norm": 7.307752062286831,
+      "learning_rate": 9.356496557126934e-07,
+      "loss": 0.4047,
+      "step": 1576
+    },
+    {
+      "epoch": 0.30155846639258055,
+      "grad_norm": 6.95229706710504,
+      "learning_rate": 9.357302650637275e-07,
+      "loss": 0.6909,
+      "step": 1577
+    },
+    {
+      "epoch": 0.30174968926283585,
+      "grad_norm": 3.93417508660463,
+      "learning_rate": 9.358108233153279e-07,
+      "loss": 0.1565,
+      "step": 1578
+    },
+    {
+      "epoch": 0.3019409121330911,
+      "grad_norm": 3.28068639365316,
+      "learning_rate": 9.358913305322391e-07,
+      "loss": 0.356,
+      "step": 1579
+    },
+    {
+      "epoch": 0.3021321350033464,
+      "grad_norm": 2.446168499867176,
+      "learning_rate": 9.359717867790825e-07,
+      "loss": 0.2312,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3023233578736017,
+      "grad_norm": 2.6574700638959095,
+      "learning_rate": 9.360521921203565e-07,
+      "loss": 0.1313,
+      "step": 1581
+    },
+    {
+      "epoch": 0.30251458074385695,
+      "grad_norm": 3.920926895133341,
+      "learning_rate": 9.361325466204376e-07,
+      "loss": 0.1869,
+      "step": 1582
+    },
+    {
+      "epoch": 0.30270580361411226,
+      "grad_norm": 2.7863846433408566,
+      "learning_rate": 9.362128503435801e-07,
+      "loss": 0.2515,
+      "step": 1583
+    },
+    {
+      "epoch": 0.3028970264843675,
+      "grad_norm": 3.03528475888258,
+      "learning_rate": 9.362931033539164e-07,
+      "loss": 0.2197,
+      "step": 1584
+    },
+    {
+      "epoch": 0.3030882493546228,
+      "grad_norm": 5.1920001262200435,
+      "learning_rate": 9.363733057154577e-07,
+      "loss": 0.1061,
+      "step": 1585
+    },
+    {
+      "epoch": 0.3032794722248781,
+      "grad_norm": 3.7756620369736282,
+      "learning_rate": 9.36453457492094e-07,
+      "loss": 0.172,
+      "step": 1586
+    },
+    {
+      "epoch": 0.30347069509513336,
+      "grad_norm": 3.582241572171883,
+      "learning_rate": 9.365335587475948e-07,
+      "loss": 0.0701,
+      "step": 1587
+    },
+    {
+      "epoch": 0.30366191796538866,
+      "grad_norm": 3.237285097272928,
+      "learning_rate": 9.366136095456081e-07,
+      "loss": 0.2833,
+      "step": 1588
+    },
+    {
+      "epoch": 0.30385314083564396,
+      "grad_norm": 4.470994719015649,
+      "learning_rate": 9.366936099496629e-07,
+      "loss": 0.3763,
+      "step": 1589
+    },
+    {
+      "epoch": 0.3040443637058992,
+      "grad_norm": 5.304709433059407,
+      "learning_rate": 9.367735600231676e-07,
+      "loss": 0.6461,
+      "step": 1590
+    },
+    {
+      "epoch": 0.3042355865761545,
+      "grad_norm": 5.087242978513659,
+      "learning_rate": 9.368534598294112e-07,
+      "loss": 0.5905,
+      "step": 1591
+    },
+    {
+      "epoch": 0.30442680944640976,
+      "grad_norm": 3.435064389265278,
+      "learning_rate": 9.369333094315635e-07,
+      "loss": 0.4381,
+      "step": 1592
+    },
+    {
+      "epoch": 0.30461803231666507,
+      "grad_norm": 3.184279666330865,
+      "learning_rate": 9.370131088926749e-07,
+      "loss": 0.2088,
+      "step": 1593
+    },
+    {
+      "epoch": 0.30480925518692037,
+      "grad_norm": 8.212257750957654,
+      "learning_rate": 9.370928582756776e-07,
+      "loss": 0.5302,
+      "step": 1594
+    },
+    {
+      "epoch": 0.3050004780571756,
+      "grad_norm": 1.2745827515721226,
+      "learning_rate": 9.37172557643385e-07,
+      "loss": 0.0749,
+      "step": 1595
+    },
+    {
+      "epoch": 0.3051917009274309,
+      "grad_norm": 2.0094782824988364,
+      "learning_rate": 9.372522070584928e-07,
+      "loss": 0.2069,
+      "step": 1596
+    },
+    {
+      "epoch": 0.3053829237976862,
+      "grad_norm": 3.2393186078765024,
+      "learning_rate": 9.373318065835783e-07,
+      "loss": 0.1703,
+      "step": 1597
+    },
+    {
+      "epoch": 0.30557414666794147,
+      "grad_norm": 1.886245131034457,
+      "learning_rate": 9.374113562811018e-07,
+      "loss": 0.0691,
+      "step": 1598
+    },
+    {
+      "epoch": 0.3057653695381968,
+      "grad_norm": 3.7906534846431437,
+      "learning_rate": 9.374908562134062e-07,
+      "loss": 0.0911,
+      "step": 1599
+    },
+    {
+      "epoch": 0.3059565924084521,
+      "grad_norm": 5.7218709465441675,
+      "learning_rate": 9.375703064427176e-07,
+      "loss": 0.1322,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3061478152787073,
+      "grad_norm": 8.630491415908953,
+      "learning_rate": 9.376497070311452e-07,
+      "loss": 0.511,
+      "step": 1601
+    },
+    {
+      "epoch": 0.30633903814896263,
+      "grad_norm": 6.352257647134298,
+      "learning_rate": 9.37729058040682e-07,
+      "loss": 0.1536,
+      "step": 1602
+    },
+    {
+      "epoch": 0.3065302610192179,
+      "grad_norm": 2.8593002893245143,
+      "learning_rate": 9.378083595332049e-07,
+      "loss": 0.3399,
+      "step": 1603
+    },
+    {
+      "epoch": 0.3067214838894732,
+      "grad_norm": 2.767820101371458,
+      "learning_rate": 9.378876115704753e-07,
+      "loss": 0.3,
+      "step": 1604
+    },
+    {
+      "epoch": 0.3069127067597285,
+      "grad_norm": 3.529567081175477,
+      "learning_rate": 9.379668142141386e-07,
+      "loss": 0.2105,
+      "step": 1605
+    },
+    {
+      "epoch": 0.30710392962998373,
+      "grad_norm": 2.607552720119366,
+      "learning_rate": 9.380459675257254e-07,
+      "loss": 0.1606,
+      "step": 1606
+    },
+    {
+      "epoch": 0.30729515250023903,
+      "grad_norm": 4.70276242502952,
+      "learning_rate": 9.381250715666513e-07,
+      "loss": 0.1917,
+      "step": 1607
+    },
+    {
+      "epoch": 0.30748637537049434,
+      "grad_norm": 5.516093550278124,
+      "learning_rate": 9.382041263982174e-07,
+      "loss": 0.2022,
+      "step": 1608
+    },
+    {
+      "epoch": 0.3076775982407496,
+      "grad_norm": 2.5973237936339495,
+      "learning_rate": 9.382831320816102e-07,
+      "loss": 0.0749,
+      "step": 1609
+    },
+    {
+      "epoch": 0.3078688211110049,
+      "grad_norm": 3.0278460079830816,
+      "learning_rate": 9.383620886779025e-07,
+      "loss": 0.0887,
+      "step": 1610
+    },
+    {
+      "epoch": 0.30806004398126013,
+      "grad_norm": 3.8124118231910535,
+      "learning_rate": 9.384409962480527e-07,
+      "loss": 0.1054,
+      "step": 1611
+    },
+    {
+      "epoch": 0.30825126685151544,
+      "grad_norm": 9.333481469568225,
+      "learning_rate": 9.385198548529066e-07,
+      "loss": 0.3175,
+      "step": 1612
+    },
+    {
+      "epoch": 0.30844248972177074,
+      "grad_norm": 6.118339186564886,
+      "learning_rate": 9.385986645531964e-07,
+      "loss": 0.1931,
+      "step": 1613
+    },
+    {
+      "epoch": 0.308633712592026,
+      "grad_norm": 4.6718090652355,
+      "learning_rate": 9.38677425409541e-07,
+      "loss": 0.5771,
+      "step": 1614
+    },
+    {
+      "epoch": 0.3088249354622813,
+      "grad_norm": 3.3098284906859368,
+      "learning_rate": 9.387561374824471e-07,
+      "loss": 0.3008,
+      "step": 1615
+    },
+    {
+      "epoch": 0.3090161583325366,
+      "grad_norm": 3.0593253048843887,
+      "learning_rate": 9.388348008323089e-07,
+      "loss": 0.6208,
+      "step": 1616
+    },
+    {
+      "epoch": 0.30920738120279184,
+      "grad_norm": 3.8595936825221693,
+      "learning_rate": 9.389134155194088e-07,
+      "loss": 0.3747,
+      "step": 1617
+    },
+    {
+      "epoch": 0.30939860407304715,
+      "grad_norm": 2.8391931935168317,
+      "learning_rate": 9.389919816039169e-07,
+      "loss": 0.1721,
+      "step": 1618
+    },
+    {
+      "epoch": 0.3095898269433024,
+      "grad_norm": 4.70762241814136,
+      "learning_rate": 9.390704991458922e-07,
+      "loss": 0.4148,
+      "step": 1619
+    },
+    {
+      "epoch": 0.3097810498135577,
+      "grad_norm": 2.9580060392760448,
+      "learning_rate": 9.39148968205282e-07,
+      "loss": 0.2318,
+      "step": 1620
+    },
+    {
+      "epoch": 0.309972272683813,
+      "grad_norm": 3.7046420695891165,
+      "learning_rate": 9.392273888419232e-07,
+      "loss": 0.1166,
+      "step": 1621
+    },
+    {
+      "epoch": 0.31016349555406825,
+      "grad_norm": 3.3861448160242484,
+      "learning_rate": 9.393057611155414e-07,
+      "loss": 0.0808,
+      "step": 1622
+    },
+    {
+      "epoch": 0.31035471842432355,
+      "grad_norm": 3.252607766549791,
+      "learning_rate": 9.393840850857522e-07,
+      "loss": 0.2439,
+      "step": 1623
+    },
+    {
+      "epoch": 0.31054594129457885,
+      "grad_norm": 2.5071406430027787,
+      "learning_rate": 9.394623608120608e-07,
+      "loss": 0.0501,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3107371641648341,
+      "grad_norm": 4.262561634322883,
+      "learning_rate": 9.395405883538626e-07,
+      "loss": 0.0879,
+      "step": 1625
+    },
+    {
+      "epoch": 0.3109283870350894,
+      "grad_norm": 6.239007361760957,
+      "learning_rate": 9.396187677704435e-07,
+      "loss": 0.5855,
+      "step": 1626
+    },
+    {
+      "epoch": 0.31111960990534465,
+      "grad_norm": 4.0249914038903984,
+      "learning_rate": 9.396968991209798e-07,
+      "loss": 0.4337,
+      "step": 1627
+    },
+    {
+      "epoch": 0.31131083277559995,
+      "grad_norm": 4.331905325322634,
+      "learning_rate": 9.397749824645388e-07,
+      "loss": 0.422,
+      "step": 1628
+    },
+    {
+      "epoch": 0.31150205564585526,
+      "grad_norm": 3.5878067629777997,
+      "learning_rate": 9.398530178600791e-07,
+      "loss": 0.343,
+      "step": 1629
+    },
+    {
+      "epoch": 0.3116932785161105,
+      "grad_norm": 4.427067703518672,
+      "learning_rate": 9.399310053664507e-07,
+      "loss": 0.2456,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3118845013863658,
+      "grad_norm": 4.264486419416378,
+      "learning_rate": 9.400089450423952e-07,
+      "loss": 0.2376,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3120757242566211,
+      "grad_norm": 2.895986502642291,
+      "learning_rate": 9.400868369465464e-07,
+      "loss": 0.1362,
+      "step": 1632
+    },
+    {
+      "epoch": 0.31226694712687636,
+      "grad_norm": 3.3674404111261347,
+      "learning_rate": 9.401646811374303e-07,
+      "loss": 0.1103,
+      "step": 1633
+    },
+    {
+      "epoch": 0.31245816999713166,
+      "grad_norm": 6.357513380197838,
+      "learning_rate": 9.402424776734652e-07,
+      "loss": 0.27,
+      "step": 1634
+    },
+    {
+      "epoch": 0.31264939286738697,
+      "grad_norm": 9.466422320512272,
+      "learning_rate": 9.403202266129624e-07,
+      "loss": 0.5528,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3128406157376422,
+      "grad_norm": 4.346197961644379,
+      "learning_rate": 9.403979280141261e-07,
+      "loss": 0.1048,
+      "step": 1636
+    },
+    {
+      "epoch": 0.3130318386078975,
+      "grad_norm": 7.688473275454175,
+      "learning_rate": 9.404755819350539e-07,
+      "loss": 0.2568,
+      "step": 1637
+    },
+    {
+      "epoch": 0.31322306147815276,
+      "grad_norm": 5.652663131580192,
+      "learning_rate": 9.405531884337373e-07,
+      "loss": 0.2384,
+      "step": 1638
+    },
+    {
+      "epoch": 0.31341428434840807,
+      "grad_norm": 4.674535497752796,
+      "learning_rate": 9.406307475680605e-07,
+      "loss": 0.3073,
+      "step": 1639
+    },
+    {
+      "epoch": 0.31360550721866337,
+      "grad_norm": 4.263221818322006,
+      "learning_rate": 9.40708259395803e-07,
+      "loss": 0.2974,
+      "step": 1640
+    },
+    {
+      "epoch": 0.3137967300889186,
+      "grad_norm": 2.465029653560405,
+      "learning_rate": 9.407857239746384e-07,
+      "loss": 0.1436,
+      "step": 1641
+    },
+    {
+      "epoch": 0.3139879529591739,
+      "grad_norm": 3.574518776896991,
+      "learning_rate": 9.408631413621341e-07,
+      "loss": 0.298,
+      "step": 1642
+    },
+    {
+      "epoch": 0.3141791758294292,
+      "grad_norm": 2.2829796491627365,
+      "learning_rate": 9.409405116157534e-07,
+      "loss": 0.0837,
+      "step": 1643
+    },
+    {
+      "epoch": 0.3143703986996845,
+      "grad_norm": 4.721853082228561,
+      "learning_rate": 9.41017834792854e-07,
+      "loss": 0.2704,
+      "step": 1644
+    },
+    {
+      "epoch": 0.3145616215699398,
+      "grad_norm": 2.451953000537674,
+      "learning_rate": 9.410951109506893e-07,
+      "loss": 0.1395,
+      "step": 1645
+    },
+    {
+      "epoch": 0.314752844440195,
+      "grad_norm": 3.5944185381477727,
+      "learning_rate": 9.411723401464081e-07,
+      "loss": 0.2868,
+      "step": 1646
+    },
+    {
+      "epoch": 0.3149440673104503,
+      "grad_norm": 4.436735275353315,
+      "learning_rate": 9.412495224370553e-07,
+      "loss": 0.1882,
+      "step": 1647
+    },
+    {
+      "epoch": 0.31513529018070563,
+      "grad_norm": 2.987992735255141,
+      "learning_rate": 9.413266578795717e-07,
+      "loss": 0.0761,
+      "step": 1648
+    },
+    {
+      "epoch": 0.3153265130509609,
+      "grad_norm": 4.031517788767941,
+      "learning_rate": 9.414037465307947e-07,
+      "loss": 0.0837,
+      "step": 1649
+    },
+    {
+      "epoch": 0.3155177359212162,
+      "grad_norm": 7.650074239133749,
+      "learning_rate": 9.414807884474582e-07,
+      "loss": 0.3223,
+      "step": 1650
+    },
+    {
+      "epoch": 0.3157089587914715,
+      "grad_norm": 6.101220505980558,
+      "learning_rate": 9.41557783686193e-07,
+      "loss": 0.3708,
+      "step": 1651
+    },
+    {
+      "epoch": 0.31590018166172673,
+      "grad_norm": 6.478215024021735,
+      "learning_rate": 9.416347323035267e-07,
+      "loss": 0.7107,
+      "step": 1652
+    },
+    {
+      "epoch": 0.31609140453198203,
+      "grad_norm": 2.991199619173412,
+      "learning_rate": 9.417116343558852e-07,
+      "loss": 0.4574,
+      "step": 1653
+    },
+    {
+      "epoch": 0.3162826274022373,
+      "grad_norm": 2.3026927526852217,
+      "learning_rate": 9.41788489899591e-07,
+      "loss": 0.2171,
+      "step": 1654
+    },
+    {
+      "epoch": 0.3164738502724926,
+      "grad_norm": 2.8136531585066615,
+      "learning_rate": 9.418652989908651e-07,
+      "loss": 0.1115,
+      "step": 1655
+    },
+    {
+      "epoch": 0.3166650731427479,
+      "grad_norm": 4.951832695977338,
+      "learning_rate": 9.419420616858262e-07,
+      "loss": 0.4853,
+      "step": 1656
+    },
+    {
+      "epoch": 0.31685629601300314,
+      "grad_norm": 2.4325148638656344,
+      "learning_rate": 9.420187780404917e-07,
+      "loss": 0.299,
+      "step": 1657
+    },
+    {
+      "epoch": 0.31704751888325844,
+      "grad_norm": 3.1082953490799126,
+      "learning_rate": 9.420954481107778e-07,
+      "loss": 0.1251,
+      "step": 1658
+    },
+    {
+      "epoch": 0.31723874175351374,
+      "grad_norm": 4.708303850919978,
+      "learning_rate": 9.421720719524987e-07,
+      "loss": 0.2314,
+      "step": 1659
+    },
+    {
+      "epoch": 0.317429964623769,
+      "grad_norm": 3.9395538302076183,
+      "learning_rate": 9.422486496213686e-07,
+      "loss": 0.1497,
+      "step": 1660
+    },
+    {
+      "epoch": 0.3176211874940243,
+      "grad_norm": 2.951320197608771,
+      "learning_rate": 9.423251811730006e-07,
+      "loss": 0.0808,
+      "step": 1661
+    },
+    {
+      "epoch": 0.31781241036427954,
+      "grad_norm": 6.454302738036737,
+      "learning_rate": 9.424016666629077e-07,
+      "loss": 0.0922,
+      "step": 1662
+    },
+    {
+      "epoch": 0.31800363323453484,
+      "grad_norm": 5.188015647905762,
+      "learning_rate": 9.42478106146502e-07,
+      "loss": 0.199,
+      "step": 1663
+    },
+    {
+      "epoch": 0.31819485610479015,
+      "grad_norm": 5.569022480027948,
+      "learning_rate": 9.425544996790966e-07,
+      "loss": 0.7048,
+      "step": 1664
+    },
+    {
+      "epoch": 0.3183860789750454,
+      "grad_norm": 6.089748222771451,
+      "learning_rate": 9.426308473159044e-07,
+      "loss": 0.3742,
+      "step": 1665
+    },
+    {
+      "epoch": 0.3185773018453007,
+      "grad_norm": 4.794388614377119,
+      "learning_rate": 9.427071491120388e-07,
+      "loss": 0.5576,
+      "step": 1666
+    },
+    {
+      "epoch": 0.318768524715556,
+      "grad_norm": 4.961084751136488,
+      "learning_rate": 9.427834051225143e-07,
+      "loss": 0.328,
+      "step": 1667
+    },
+    {
+      "epoch": 0.31895974758581125,
+      "grad_norm": 3.266770741368325,
+      "learning_rate": 9.428596154022461e-07,
+      "loss": 0.3151,
+      "step": 1668
+    },
+    {
+      "epoch": 0.31915097045606655,
+      "grad_norm": 9.510213080264355,
+      "learning_rate": 9.429357800060509e-07,
+      "loss": 0.4425,
+      "step": 1669
+    },
+    {
+      "epoch": 0.31934219332632185,
+      "grad_norm": 4.321109525754854,
+      "learning_rate": 9.430118989886467e-07,
+      "loss": 0.4536,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3195334161965771,
+      "grad_norm": 6.717987443099647,
+      "learning_rate": 9.430879724046533e-07,
+      "loss": 0.2713,
+      "step": 1671
+    },
+    {
+      "epoch": 0.3197246390668324,
+      "grad_norm": 2.8975767929525404,
+      "learning_rate": 9.431640003085927e-07,
+      "loss": 0.0862,
+      "step": 1672
+    },
+    {
+      "epoch": 0.31991586193708765,
+      "grad_norm": 3.149018113433988,
+      "learning_rate": 9.432399827548887e-07,
+      "loss": 0.0548,
+      "step": 1673
+    },
+    {
+      "epoch": 0.32010708480734296,
+      "grad_norm": 5.693499408655227,
+      "learning_rate": 9.433159197978678e-07,
+      "loss": 0.436,
+      "step": 1674
+    },
+    {
+      "epoch": 0.32029830767759826,
+      "grad_norm": 5.8084456547403684,
+      "learning_rate": 9.433918114917591e-07,
+      "loss": 0.0661,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3204895305478535,
+      "grad_norm": 15.803686769324734,
+      "learning_rate": 9.434676578906947e-07,
+      "loss": 0.1854,
+      "step": 1676
+    },
+    {
+      "epoch": 0.3206807534181088,
+      "grad_norm": 3.5202614284015885,
+      "learning_rate": 9.435434590487097e-07,
+      "loss": 0.5066,
+      "step": 1677
+    },
+    {
+      "epoch": 0.3208719762883641,
+      "grad_norm": 5.279263094428088,
+      "learning_rate": 9.436192150197424e-07,
+      "loss": 0.4149,
+      "step": 1678
+    },
+    {
+      "epoch": 0.32106319915861936,
+      "grad_norm": 7.412991445374091,
+      "learning_rate": 9.436949258576352e-07,
+      "loss": 0.196,
+      "step": 1679
+    },
+    {
+      "epoch": 0.32125442202887466,
+      "grad_norm": 2.1867917685465974,
+      "learning_rate": 9.437705916161339e-07,
+      "loss": 0.2595,
+      "step": 1680
+    },
+    {
+      "epoch": 0.3214456448991299,
+      "grad_norm": 3.509008666677092,
+      "learning_rate": 9.438462123488885e-07,
+      "loss": 0.4255,
+      "step": 1681
+    },
+    {
+      "epoch": 0.3216368677693852,
+      "grad_norm": 2.96296642510777,
+      "learning_rate": 9.439217881094533e-07,
+      "loss": 0.1776,
+      "step": 1682
+    },
+    {
+      "epoch": 0.3218280906396405,
+      "grad_norm": 4.483661661333351,
+      "learning_rate": 9.43997318951287e-07,
+      "loss": 0.1404,
+      "step": 1683
+    },
+    {
+      "epoch": 0.32201931350989577,
+      "grad_norm": 6.472653303220512,
+      "learning_rate": 9.440728049277531e-07,
+      "loss": 0.1537,
+      "step": 1684
+    },
+    {
+      "epoch": 0.32221053638015107,
+      "grad_norm": 4.306775924391126,
+      "learning_rate": 9.441482460921202e-07,
+      "loss": 0.3078,
+      "step": 1685
+    },
+    {
+      "epoch": 0.3224017592504064,
+      "grad_norm": 3.8290615882333854,
+      "learning_rate": 9.442236424975618e-07,
+      "loss": 0.099,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3225929821206616,
+      "grad_norm": 3.1898314980994043,
+      "learning_rate": 9.44298994197157e-07,
+      "loss": 0.0913,
+      "step": 1687
+    },
+    {
+      "epoch": 0.3227842049909169,
+      "grad_norm": 8.408641148941944,
+      "learning_rate": 9.443743012438908e-07,
+      "loss": 0.3135,
+      "step": 1688
+    },
+    {
+      "epoch": 0.32297542786117217,
+      "grad_norm": 4.194933769421632,
+      "learning_rate": 9.444495636906533e-07,
+      "loss": 0.2727,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3231666507314275,
+      "grad_norm": 4.940387509240047,
+      "learning_rate": 9.445247815902416e-07,
+      "loss": 0.4452,
+      "step": 1690
+    },
+    {
+      "epoch": 0.3233578736016828,
+      "grad_norm": 3.4130620472608095,
+      "learning_rate": 9.445999549953582e-07,
+      "loss": 0.3903,
+      "step": 1691
+    },
+    {
+      "epoch": 0.323549096471938,
+      "grad_norm": 4.345139535083506,
+      "learning_rate": 9.446750839586131e-07,
+      "loss": 0.6568,
+      "step": 1692
+    },
+    {
+      "epoch": 0.32374031934219333,
+      "grad_norm": 3.696745967467313,
+      "learning_rate": 9.447501685325222e-07,
+      "loss": 0.3152,
+      "step": 1693
+    },
+    {
+      "epoch": 0.32393154221244863,
+      "grad_norm": 3.4602058274452596,
+      "learning_rate": 9.448252087695088e-07,
+      "loss": 0.4549,
+      "step": 1694
+    },
+    {
+      "epoch": 0.3241227650827039,
+      "grad_norm": 2.818468161827598,
+      "learning_rate": 9.449002047219033e-07,
+      "loss": 0.234,
+      "step": 1695
+    },
+    {
+      "epoch": 0.3243139879529592,
+      "grad_norm": 1.5391997455263622,
+      "learning_rate": 9.449751564419434e-07,
+      "loss": 0.12,
+      "step": 1696
+    },
+    {
+      "epoch": 0.32450521082321443,
+      "grad_norm": 3.852062881495208,
+      "learning_rate": 9.450500639817747e-07,
+      "loss": 0.1442,
+      "step": 1697
+    },
+    {
+      "epoch": 0.32469643369346973,
+      "grad_norm": 4.273525425426597,
+      "learning_rate": 9.451249273934502e-07,
+      "loss": 0.1842,
+      "step": 1698
+    },
+    {
+      "epoch": 0.32488765656372504,
+      "grad_norm": 2.7954020111098585,
+      "learning_rate": 9.451997467289317e-07,
+      "loss": 0.055,
+      "step": 1699
+    },
+    {
+      "epoch": 0.3250788794339803,
+      "grad_norm": 3.681842796380408,
+      "learning_rate": 9.452745220400881e-07,
+      "loss": 0.138,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3252701023042356,
+      "grad_norm": 4.138200849198906,
+      "learning_rate": 9.453492533786983e-07,
+      "loss": 0.1958,
+      "step": 1701
+    },
+    {
+      "epoch": 0.3254613251744909,
+      "grad_norm": 4.870027989927847,
+      "learning_rate": 9.454239407964486e-07,
+      "loss": 0.561,
+      "step": 1702
+    },
+    {
+      "epoch": 0.32565254804474614,
+      "grad_norm": 7.3052368610266925,
+      "learning_rate": 9.454985843449348e-07,
+      "loss": 0.4954,
+      "step": 1703
+    },
+    {
+      "epoch": 0.32584377091500144,
+      "grad_norm": 2.552043981962372,
+      "learning_rate": 9.455731840756618e-07,
+      "loss": 0.3179,
+      "step": 1704
+    },
+    {
+      "epoch": 0.32603499378525674,
+      "grad_norm": 3.962824445372122,
+      "learning_rate": 9.45647740040044e-07,
+      "loss": 0.2598,
+      "step": 1705
+    },
+    {
+      "epoch": 0.326226216655512,
+      "grad_norm": 4.7100920416713326,
+      "learning_rate": 9.457222522894048e-07,
+      "loss": 0.5223,
+      "step": 1706
+    },
+    {
+      "epoch": 0.3264174395257673,
+      "grad_norm": 4.478778603473347,
+      "learning_rate": 9.457967208749781e-07,
+      "loss": 0.2378,
+      "step": 1707
+    },
+    {
+      "epoch": 0.32660866239602254,
+      "grad_norm": 4.651642228411658,
+      "learning_rate": 9.458711458479071e-07,
+      "loss": 0.2237,
+      "step": 1708
+    },
+    {
+      "epoch": 0.32679988526627785,
+      "grad_norm": 3.388409712706619,
+      "learning_rate": 9.459455272592458e-07,
+      "loss": 0.3723,
+      "step": 1709
+    },
+    {
+      "epoch": 0.32699110813653315,
+      "grad_norm": 5.408555085152675,
+      "learning_rate": 9.460198651599582e-07,
+      "loss": 0.1365,
+      "step": 1710
+    },
+    {
+      "epoch": 0.3271823310067884,
+      "grad_norm": 4.720792618811919,
+      "learning_rate": 9.460941596009193e-07,
+      "loss": 0.1483,
+      "step": 1711
+    },
+    {
+      "epoch": 0.3273735538770437,
+      "grad_norm": 2.88984083931955,
+      "learning_rate": 9.461684106329143e-07,
+      "loss": 0.0511,
+      "step": 1712
+    },
+    {
+      "epoch": 0.327564776747299,
+      "grad_norm": 5.797856743855062,
+      "learning_rate": 9.4624261830664e-07,
+      "loss": 0.4078,
+      "step": 1713
+    },
+    {
+      "epoch": 0.32775599961755425,
+      "grad_norm": 4.1589165644860495,
+      "learning_rate": 9.463167826727043e-07,
+      "loss": 0.4594,
+      "step": 1714
+    },
+    {
+      "epoch": 0.32794722248780955,
+      "grad_norm": 4.609757540071611,
+      "learning_rate": 9.463909037816262e-07,
+      "loss": 0.5734,
+      "step": 1715
+    },
+    {
+      "epoch": 0.3281384453580648,
+      "grad_norm": 2.92558827801537,
+      "learning_rate": 9.464649816838371e-07,
+      "loss": 0.2488,
+      "step": 1716
+    },
+    {
+      "epoch": 0.3283296682283201,
+      "grad_norm": 2.1267426021416322,
+      "learning_rate": 9.465390164296796e-07,
+      "loss": 0.4441,
+      "step": 1717
+    },
+    {
+      "epoch": 0.3285208910985754,
+      "grad_norm": 3.6047017282801255,
+      "learning_rate": 9.466130080694082e-07,
+      "loss": 0.4332,
+      "step": 1718
+    },
+    {
+      "epoch": 0.32871211396883065,
+      "grad_norm": 3.8334920822664817,
+      "learning_rate": 9.466869566531907e-07,
+      "loss": 0.2376,
+      "step": 1719
+    },
+    {
+      "epoch": 0.32890333683908596,
+      "grad_norm": 3.070095685530102,
+      "learning_rate": 9.467608622311064e-07,
+      "loss": 0.1403,
+      "step": 1720
+    },
+    {
+      "epoch": 0.32909455970934126,
+      "grad_norm": 3.437472117917817,
+      "learning_rate": 9.468347248531475e-07,
+      "loss": 0.1414,
+      "step": 1721
+    },
+    {
+      "epoch": 0.3292857825795965,
+      "grad_norm": 6.045066385590912,
+      "learning_rate": 9.469085445692193e-07,
+      "loss": 0.3908,
+      "step": 1722
+    },
+    {
+      "epoch": 0.3294770054498518,
+      "grad_norm": 3.565955109489673,
+      "learning_rate": 9.469823214291401e-07,
+      "loss": 0.1631,
+      "step": 1723
+    },
+    {
+      "epoch": 0.32966822832010706,
+      "grad_norm": 3.011993910196944,
+      "learning_rate": 9.470560554826412e-07,
+      "loss": 0.0996,
+      "step": 1724
+    },
+    {
+      "epoch": 0.32985945119036236,
+      "grad_norm": 4.918537473631425,
+      "learning_rate": 9.47129746779368e-07,
+      "loss": 0.1012,
+      "step": 1725
+    },
+    {
+      "epoch": 0.33005067406061767,
+      "grad_norm": 9.109263607541484,
+      "learning_rate": 9.472033953688789e-07,
+      "loss": 0.4021,
+      "step": 1726
+    },
+    {
+      "epoch": 0.3302418969308729,
+      "grad_norm": 6.708893399128439,
+      "learning_rate": 9.472770013006465e-07,
+      "loss": 0.5217,
+      "step": 1727
+    },
+    {
+      "epoch": 0.3304331198011282,
+      "grad_norm": 4.7167265134503555,
+      "learning_rate": 9.473505646240577e-07,
+      "loss": 0.45,
+      "step": 1728
+    },
+    {
+      "epoch": 0.3306243426713835,
+      "grad_norm": 3.1167816433293534,
+      "learning_rate": 9.474240853884135e-07,
+      "loss": 0.1615,
+      "step": 1729
+    },
+    {
+      "epoch": 0.33081556554163877,
+      "grad_norm": 3.850677450448502,
+      "learning_rate": 9.474975636429289e-07,
+      "loss": 0.1532,
+      "step": 1730
+    },
+    {
+      "epoch": 0.33100678841189407,
+      "grad_norm": 3.813597974419701,
+      "learning_rate": 9.475709994367345e-07,
+      "loss": 0.4618,
+      "step": 1731
+    },
+    {
+      "epoch": 0.3311980112821493,
+      "grad_norm": 4.4739993548335235,
+      "learning_rate": 9.476443928188751e-07,
+      "loss": 0.1348,
+      "step": 1732
+    },
+    {
+      "epoch": 0.3313892341524046,
+      "grad_norm": 2.4111218365295435,
+      "learning_rate": 9.477177438383109e-07,
+      "loss": 0.2782,
+      "step": 1733
+    },
+    {
+      "epoch": 0.3315804570226599,
+      "grad_norm": 2.0599425672654395,
+      "learning_rate": 9.47791052543917e-07,
+      "loss": 0.1798,
+      "step": 1734
+    },
+    {
+      "epoch": 0.3317716798929152,
+      "grad_norm": 3.1964765820623264,
+      "learning_rate": 9.478643189844845e-07,
+      "loss": 0.0442,
+      "step": 1735
+    },
+    {
+      "epoch": 0.3319629027631705,
+      "grad_norm": 2.748517590559987,
+      "learning_rate": 9.479375432087198e-07,
+      "loss": 0.1787,
+      "step": 1736
+    },
+    {
+      "epoch": 0.3321541256334258,
+      "grad_norm": 2.4755674463066804,
+      "learning_rate": 9.48010725265245e-07,
+      "loss": 0.0599,
+      "step": 1737
+    },
+    {
+      "epoch": 0.332345348503681,
+      "grad_norm": 6.107111440783317,
+      "learning_rate": 9.480838652025987e-07,
+      "loss": 0.6485,
+      "step": 1738
+    },
+    {
+      "epoch": 0.33253657137393633,
+      "grad_norm": 9.333300408804813,
+      "learning_rate": 9.481569630692356e-07,
+      "loss": 0.4915,
+      "step": 1739
+    },
+    {
+      "epoch": 0.33272779424419163,
+      "grad_norm": 5.009238291568764,
+      "learning_rate": 9.482300189135264e-07,
+      "loss": 0.4498,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3329190171144469,
+      "grad_norm": 4.705956098034793,
+      "learning_rate": 9.483030327837591e-07,
+      "loss": 0.5773,
+      "step": 1741
+    },
+    {
+      "epoch": 0.3331102399847022,
+      "grad_norm": 3.1240397694653588,
+      "learning_rate": 9.483760047281381e-07,
+      "loss": 0.3165,
+      "step": 1742
+    },
+    {
+      "epoch": 0.33330146285495743,
+      "grad_norm": 1.9130746588927041,
+      "learning_rate": 9.484489347947848e-07,
+      "loss": 0.1775,
+      "step": 1743
+    },
+    {
+      "epoch": 0.33349268572521273,
+      "grad_norm": 2.8906829003386134,
+      "learning_rate": 9.485218230317381e-07,
+      "loss": 0.2168,
+      "step": 1744
+    },
+    {
+      "epoch": 0.33368390859546804,
+      "grad_norm": 4.671765992797058,
+      "learning_rate": 9.485946694869539e-07,
+      "loss": 0.1084,
+      "step": 1745
+    },
+    {
+      "epoch": 0.3338751314657233,
+      "grad_norm": 2.438610801928022,
+      "learning_rate": 9.486674742083059e-07,
+      "loss": 0.2294,
+      "step": 1746
+    },
+    {
+      "epoch": 0.3340663543359786,
+      "grad_norm": 3.3406711557338227,
+      "learning_rate": 9.487402372435857e-07,
+      "loss": 0.3175,
+      "step": 1747
+    },
+    {
+      "epoch": 0.3342575772062339,
+      "grad_norm": 6.0490554702795984,
+      "learning_rate": 9.488129586405025e-07,
+      "loss": 0.4494,
+      "step": 1748
+    },
+    {
+      "epoch": 0.33444880007648914,
+      "grad_norm": 4.057051305870054,
+      "learning_rate": 9.488856384466839e-07,
+      "loss": 0.1584,
+      "step": 1749
+    },
+    {
+      "epoch": 0.33464002294674444,
+      "grad_norm": 5.570791001921088,
+      "learning_rate": 9.489582767096756e-07,
+      "loss": 0.2046,
+      "step": 1750
+    },
+    {
+      "epoch": 0.3348312458169997,
+      "grad_norm": 5.902993894123769,
+      "learning_rate": 9.490308734769423e-07,
+      "loss": 0.2671,
+      "step": 1751
+    },
+    {
+      "epoch": 0.335022468687255,
+      "grad_norm": 4.21663290520422,
+      "learning_rate": 9.491034287958669e-07,
+      "loss": 0.4533,
+      "step": 1752
+    },
+    {
+      "epoch": 0.3352136915575103,
+      "grad_norm": 2.8552070633445754,
+      "learning_rate": 9.491759427137511e-07,
+      "loss": 0.1132,
+      "step": 1753
+    },
+    {
+      "epoch": 0.33540491442776554,
+      "grad_norm": 2.1174900641090013,
+      "learning_rate": 9.492484152778162e-07,
+      "loss": 0.2825,
+      "step": 1754
+    },
+    {
+      "epoch": 0.33559613729802085,
+      "grad_norm": 1.796854831748228,
+      "learning_rate": 9.493208465352027e-07,
+      "loss": 0.2334,
+      "step": 1755
+    },
+    {
+      "epoch": 0.33578736016827615,
+      "grad_norm": 3.198789612024309,
+      "learning_rate": 9.493932365329701e-07,
+      "loss": 0.0819,
+      "step": 1756
+    },
+    {
+      "epoch": 0.3359785830385314,
+      "grad_norm": 2.1768328387207765,
+      "learning_rate": 9.494655853180976e-07,
+      "loss": 0.3874,
+      "step": 1757
+    },
+    {
+      "epoch": 0.3361698059087867,
+      "grad_norm": 2.3012526956901946,
+      "learning_rate": 9.495378929374846e-07,
+      "loss": 0.2278,
+      "step": 1758
+    },
+    {
+      "epoch": 0.33636102877904195,
+      "grad_norm": 2.190359916375147,
+      "learning_rate": 9.496101594379503e-07,
+      "loss": 0.189,
+      "step": 1759
+    },
+    {
+      "epoch": 0.33655225164929725,
+      "grad_norm": 4.262666563655919,
+      "learning_rate": 9.496823848662338e-07,
+      "loss": 0.1192,
+      "step": 1760
+    },
+    {
+      "epoch": 0.33674347451955255,
+      "grad_norm": 3.3115775785348567,
+      "learning_rate": 9.49754569268995e-07,
+      "loss": 0.0991,
+      "step": 1761
+    },
+    {
+      "epoch": 0.3369346973898078,
+      "grad_norm": 3.3418643661452547,
+      "learning_rate": 9.49826712692814e-07,
+      "loss": 0.0966,
+      "step": 1762
+    },
+    {
+      "epoch": 0.3371259202600631,
+      "grad_norm": 8.400862658572816,
+      "learning_rate": 9.498988151841918e-07,
+      "loss": 0.2155,
+      "step": 1763
+    },
+    {
+      "epoch": 0.3373171431303184,
+      "grad_norm": 5.226127298479916,
+      "learning_rate": 9.499708767895501e-07,
+      "loss": 0.4877,
+      "step": 1764
+    },
+    {
+      "epoch": 0.33750836600057366,
+      "grad_norm": 3.5581558242821587,
+      "learning_rate": 9.500428975552319e-07,
+      "loss": 0.3755,
+      "step": 1765
+    },
+    {
+      "epoch": 0.33769958887082896,
+      "grad_norm": 3.614861879783529,
+      "learning_rate": 9.501148775275014e-07,
+      "loss": 0.2456,
+      "step": 1766
+    },
+    {
+      "epoch": 0.3378908117410842,
+      "grad_norm": 6.104657705647795,
+      "learning_rate": 9.50186816752544e-07,
+      "loss": 0.5583,
+      "step": 1767
+    },
+    {
+      "epoch": 0.3380820346113395,
+      "grad_norm": 4.31305370369725,
+      "learning_rate": 9.502587152764672e-07,
+      "loss": 0.3055,
+      "step": 1768
+    },
+    {
+      "epoch": 0.3382732574815948,
+      "grad_norm": 3.7341852499026906,
+      "learning_rate": 9.503305731452996e-07,
+      "loss": 0.256,
+      "step": 1769
+    },
+    {
+      "epoch": 0.33846448035185006,
+      "grad_norm": 2.901724131842463,
+      "learning_rate": 9.504023904049924e-07,
+      "loss": 0.1851,
+      "step": 1770
+    },
+    {
+      "epoch": 0.33865570322210536,
+      "grad_norm": 3.4792689386438798,
+      "learning_rate": 9.504741671014186e-07,
+      "loss": 0.2877,
+      "step": 1771
+    },
+    {
+      "epoch": 0.33884692609236067,
+      "grad_norm": 5.5634655060942,
+      "learning_rate": 9.505459032803735e-07,
+      "loss": 0.1232,
+      "step": 1772
+    },
+    {
+      "epoch": 0.3390381489626159,
+      "grad_norm": 3.8304727634048676,
+      "learning_rate": 9.506175989875752e-07,
+      "loss": 0.2634,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3392293718328712,
+      "grad_norm": 5.344801414519021,
+      "learning_rate": 9.506892542686641e-07,
+      "loss": 0.108,
+      "step": 1774
+    },
+    {
+      "epoch": 0.3394205947031265,
+      "grad_norm": 8.613006656620682,
+      "learning_rate": 9.507608691692036e-07,
+      "loss": 0.1408,
+      "step": 1775
+    },
+    {
+      "epoch": 0.33961181757338177,
+      "grad_norm": 7.49268238394239,
+      "learning_rate": 9.508324437346801e-07,
+      "loss": 0.3814,
+      "step": 1776
+    },
+    {
+      "epoch": 0.33980304044363707,
+      "grad_norm": 4.334761286480099,
+      "learning_rate": 9.509039780105031e-07,
+      "loss": 0.6733,
+      "step": 1777
+    },
+    {
+      "epoch": 0.3399942633138923,
+      "grad_norm": 4.116094512922006,
+      "learning_rate": 9.509754720420059e-07,
+      "loss": 0.2479,
+      "step": 1778
+    },
+    {
+      "epoch": 0.3401854861841476,
+      "grad_norm": 2.8326509626382235,
+      "learning_rate": 9.510469258744446e-07,
+      "loss": 0.1504,
+      "step": 1779
+    },
+    {
+      "epoch": 0.3403767090544029,
+      "grad_norm": 2.6046782842975373,
+      "learning_rate": 9.511183395529996e-07,
+      "loss": 0.2865,
+      "step": 1780
+    },
+    {
+      "epoch": 0.3405679319246582,
+      "grad_norm": 2.9156574864841738,
+      "learning_rate": 9.511897131227746e-07,
+      "loss": 0.0877,
+      "step": 1781
+    },
+    {
+      "epoch": 0.3407591547949135,
+      "grad_norm": 2.4939120553119203,
+      "learning_rate": 9.512610466287982e-07,
+      "loss": 0.2317,
+      "step": 1782
+    },
+    {
+      "epoch": 0.3409503776651688,
+      "grad_norm": 2.398527559956646,
+      "learning_rate": 9.513323401160224e-07,
+      "loss": 0.1137,
+      "step": 1783
+    },
+    {
+      "epoch": 0.341141600535424,
+      "grad_norm": 5.483513136740057,
+      "learning_rate": 9.514035936293242e-07,
+      "loss": 0.4056,
+      "step": 1784
+    },
+    {
+      "epoch": 0.34133282340567933,
+      "grad_norm": 3.440187686634975,
+      "learning_rate": 9.514748072135045e-07,
+      "loss": 0.1199,
+      "step": 1785
+    },
+    {
+      "epoch": 0.3415240462759346,
+      "grad_norm": 3.526620311204053,
+      "learning_rate": 9.515459809132894e-07,
+      "loss": 0.1543,
+      "step": 1786
+    },
+    {
+      "epoch": 0.3417152691461899,
+      "grad_norm": 3.956837354693956,
+      "learning_rate": 9.516171147733298e-07,
+      "loss": 0.1478,
+      "step": 1787
+    },
+    {
+      "epoch": 0.3419064920164452,
+      "grad_norm": 4.6049740228738285,
+      "learning_rate": 9.516882088382018e-07,
+      "loss": 0.1989,
+      "step": 1788
+    },
+    {
+      "epoch": 0.34209771488670043,
+      "grad_norm": 6.089533985767577,
+      "learning_rate": 9.517592631524064e-07,
+      "loss": 0.2217,
+      "step": 1789
+    },
+    {
+      "epoch": 0.34228893775695574,
+      "grad_norm": 6.010163282566079,
+      "learning_rate": 9.518302777603702e-07,
+      "loss": 0.6252,
+      "step": 1790
+    },
+    {
+      "epoch": 0.34248016062721104,
+      "grad_norm": 6.6754723885360665,
+      "learning_rate": 9.519012527064453e-07,
+      "loss": 0.5431,
+      "step": 1791
+    },
+    {
+      "epoch": 0.3426713834974663,
+      "grad_norm": 3.0793894383049607,
+      "learning_rate": 9.519721880349097e-07,
+      "loss": 0.3158,
+      "step": 1792
+    },
+    {
+      "epoch": 0.3428626063677216,
+      "grad_norm": 4.039607647643446,
+      "learning_rate": 9.520430837899669e-07,
+      "loss": 0.2174,
+      "step": 1793
+    },
+    {
+      "epoch": 0.34305382923797684,
+      "grad_norm": 3.278827899526652,
+      "learning_rate": 9.521139400157469e-07,
+      "loss": 0.2805,
+      "step": 1794
+    },
+    {
+      "epoch": 0.34324505210823214,
+      "grad_norm": 4.1222018374761324,
+      "learning_rate": 9.521847567563057e-07,
+      "loss": 0.195,
+      "step": 1795
+    },
+    {
+      "epoch": 0.34343627497848744,
+      "grad_norm": 2.8492043974138093,
+      "learning_rate": 9.522555340556258e-07,
+      "loss": 0.142,
+      "step": 1796
+    },
+    {
+      "epoch": 0.3436274978487427,
+      "grad_norm": 2.8473447778475207,
+      "learning_rate": 9.52326271957616e-07,
+      "loss": 0.2575,
+      "step": 1797
+    },
+    {
+      "epoch": 0.343818720718998,
+      "grad_norm": 2.9849252243516426,
+      "learning_rate": 9.523969705061121e-07,
+      "loss": 0.0838,
+      "step": 1798
+    },
+    {
+      "epoch": 0.3440099435892533,
+      "grad_norm": 5.192227702597894,
+      "learning_rate": 9.524676297448768e-07,
+      "loss": 0.1424,
+      "step": 1799
+    },
+    {
+      "epoch": 0.34420116645950855,
+      "grad_norm": 5.240680415407017,
+      "learning_rate": 9.525382497175996e-07,
+      "loss": 0.0549,
+      "step": 1800
+    },
+    {
+      "epoch": 0.34439238932976385,
+      "grad_norm": 10.863368336130618,
+      "learning_rate": 9.526088304678972e-07,
+      "loss": 0.4659,
+      "step": 1801
+    },
+    {
+      "epoch": 0.3445836122000191,
+      "grad_norm": 4.244465084813536,
+      "learning_rate": 9.526793720393139e-07,
+      "loss": 0.8077,
+      "step": 1802
+    },
+    {
+      "epoch": 0.3447748350702744,
+      "grad_norm": 5.33484697717727,
+      "learning_rate": 9.527498744753215e-07,
+      "loss": 0.6598,
+      "step": 1803
+    },
+    {
+      "epoch": 0.3449660579405297,
+      "grad_norm": 3.9628597012023494,
+      "learning_rate": 9.528203378193193e-07,
+      "loss": 0.3334,
+      "step": 1804
+    },
+    {
+      "epoch": 0.34515728081078495,
+      "grad_norm": 3.0063530093367357,
+      "learning_rate": 9.528907621146345e-07,
+      "loss": 0.3736,
+      "step": 1805
+    },
+    {
+      "epoch": 0.34534850368104025,
+      "grad_norm": 3.1182181965626636,
+      "learning_rate": 9.529611474045226e-07,
+      "loss": 0.4145,
+      "step": 1806
+    },
+    {
+      "epoch": 0.34553972655129556,
+      "grad_norm": 2.0833994537033473,
+      "learning_rate": 9.530314937321668e-07,
+      "loss": 0.2553,
+      "step": 1807
+    },
+    {
+      "epoch": 0.3457309494215508,
+      "grad_norm": 3.1346856127833953,
+      "learning_rate": 9.531018011406789e-07,
+      "loss": 0.2389,
+      "step": 1808
+    },
+    {
+      "epoch": 0.3459221722918061,
+      "grad_norm": 4.83242833225965,
+      "learning_rate": 9.531720696730991e-07,
+      "loss": 0.3026,
+      "step": 1809
+    },
+    {
+      "epoch": 0.3461133951620614,
+      "grad_norm": 4.219325499664507,
+      "learning_rate": 9.532422993723965e-07,
+      "loss": 0.264,
+      "step": 1810
+    },
+    {
+      "epoch": 0.34630461803231666,
+      "grad_norm": 9.053740799975694,
+      "learning_rate": 9.533124902814686e-07,
+      "loss": 0.3629,
+      "step": 1811
+    },
+    {
+      "epoch": 0.34649584090257196,
+      "grad_norm": 3.6043522222892945,
+      "learning_rate": 9.533826424431419e-07,
+      "loss": 0.1445,
+      "step": 1812
+    },
+    {
+      "epoch": 0.3466870637728272,
+      "grad_norm": 8.694769363371668,
+      "learning_rate": 9.534527559001725e-07,
+      "loss": 0.2309,
+      "step": 1813
+    },
+    {
+      "epoch": 0.3468782866430825,
+      "grad_norm": 4.462604271447886,
+      "learning_rate": 9.53522830695245e-07,
+      "loss": 0.6058,
+      "step": 1814
+    },
+    {
+      "epoch": 0.3470695095133378,
+      "grad_norm": 5.43324832654736,
+      "learning_rate": 9.535928668709744e-07,
+      "loss": 0.1066,
+      "step": 1815
+    },
+    {
+      "epoch": 0.34726073238359306,
+      "grad_norm": 3.1762199521135344,
+      "learning_rate": 9.536628644699042e-07,
+      "loss": 0.2911,
+      "step": 1816
+    },
+    {
+      "epoch": 0.34745195525384837,
+      "grad_norm": 1.8857918114950758,
+      "learning_rate": 9.537328235345085e-07,
+      "loss": 0.1156,
+      "step": 1817
+    },
+    {
+      "epoch": 0.34764317812410367,
+      "grad_norm": 4.627641078420364,
+      "learning_rate": 9.538027441071908e-07,
+      "loss": 0.2947,
+      "step": 1818
+    },
+    {
+      "epoch": 0.3478344009943589,
+      "grad_norm": 4.288615552617686,
+      "learning_rate": 9.538726262302849e-07,
+      "loss": 0.4037,
+      "step": 1819
+    },
+    {
+      "epoch": 0.3480256238646142,
+      "grad_norm": 2.8001071432595057,
+      "learning_rate": 9.539424699460546e-07,
+      "loss": 0.1877,
+      "step": 1820
+    },
+    {
+      "epoch": 0.34821684673486947,
+      "grad_norm": 1.2761870113300688,
+      "learning_rate": 9.540122752966943e-07,
+      "loss": 0.0351,
+      "step": 1821
+    },
+    {
+      "epoch": 0.34840806960512477,
+      "grad_norm": 3.25588266127508,
+      "learning_rate": 9.540820423243287e-07,
+      "loss": 0.1188,
+      "step": 1822
+    },
+    {
+      "epoch": 0.3485992924753801,
+      "grad_norm": 3.323129692942441,
+      "learning_rate": 9.54151771071013e-07,
+      "loss": 0.0829,
+      "step": 1823
+    },
+    {
+      "epoch": 0.3487905153456353,
+      "grad_norm": 5.996910253344628,
+      "learning_rate": 9.54221461578734e-07,
+      "loss": 0.0868,
+      "step": 1824
+    },
+    {
+      "epoch": 0.3489817382158906,
+      "grad_norm": 6.790369327259449,
+      "learning_rate": 9.542911138894085e-07,
+      "loss": 0.1377,
+      "step": 1825
+    },
+    {
+      "epoch": 0.3491729610861459,
+      "grad_norm": 10.977442413575185,
+      "learning_rate": 9.543607280448847e-07,
+      "loss": 0.2559,
+      "step": 1826
+    },
+    {
+      "epoch": 0.3493641839564012,
+      "grad_norm": 4.607680807886779,
+      "learning_rate": 9.544303040869426e-07,
+      "loss": 0.5215,
+      "step": 1827
+    },
+    {
+      "epoch": 0.3495554068266565,
+      "grad_norm": 4.172170107115716,
+      "learning_rate": 9.544998420572929e-07,
+      "loss": 0.4131,
+      "step": 1828
+    },
+    {
+      "epoch": 0.3497466296969117,
+      "grad_norm": 3.2176778544699722,
+      "learning_rate": 9.545693419975782e-07,
+      "loss": 0.2148,
+      "step": 1829
+    },
+    {
+      "epoch": 0.34993785256716703,
+      "grad_norm": 4.400147019444306,
+      "learning_rate": 9.546388039493727e-07,
+      "loss": 0.3942,
+      "step": 1830
+    },
+    {
+      "epoch": 0.35012907543742233,
+      "grad_norm": 4.822223105384133,
+      "learning_rate": 9.547082279541826e-07,
+      "loss": 0.3639,
+      "step": 1831
+    },
+    {
+      "epoch": 0.3503202983076776,
+      "grad_norm": 3.7105115144070586,
+      "learning_rate": 9.547776140534464e-07,
+      "loss": 0.3325,
+      "step": 1832
+    },
+    {
+      "epoch": 0.3505115211779329,
+      "grad_norm": 3.990283609124795,
+      "learning_rate": 9.548469622885338e-07,
+      "loss": 0.4527,
+      "step": 1833
+    },
+    {
+      "epoch": 0.3507027440481882,
+      "grad_norm": 4.154582671542668,
+      "learning_rate": 9.549162727007477e-07,
+      "loss": 0.1002,
+      "step": 1834
+    },
+    {
+      "epoch": 0.35089396691844343,
+      "grad_norm": 4.296151839004215,
+      "learning_rate": 9.549855453313234e-07,
+      "loss": 0.1102,
+      "step": 1835
+    },
+    {
+      "epoch": 0.35108518978869874,
+      "grad_norm": 2.561682407857779,
+      "learning_rate": 9.550547802214284e-07,
+      "loss": 0.032,
+      "step": 1836
+    },
+    {
+      "epoch": 0.351276412658954,
+      "grad_norm": 3.0255966165649766,
+      "learning_rate": 9.551239774121629e-07,
+      "loss": 0.092,
+      "step": 1837
+    },
+    {
+      "epoch": 0.3514676355292093,
+      "grad_norm": 7.666847835003487,
+      "learning_rate": 9.551931369445606e-07,
+      "loss": 0.2445,
+      "step": 1838
+    },
+    {
+      "epoch": 0.3516588583994646,
+      "grad_norm": 6.344214812289434,
+      "learning_rate": 9.552622588595875e-07,
+      "loss": 0.2142,
+      "step": 1839
+    },
+    {
+      "epoch": 0.35185008126971984,
+      "grad_norm": 4.636857786721234,
+      "learning_rate": 9.553313431981437e-07,
+      "loss": 0.4141,
+      "step": 1840
+    },
+    {
+      "epoch": 0.35204130413997514,
+      "grad_norm": 4.671223776858301,
+      "learning_rate": 9.554003900010617e-07,
+      "loss": 0.4288,
+      "step": 1841
+    },
+    {
+      "epoch": 0.35223252701023045,
+      "grad_norm": 3.875327188768366,
+      "learning_rate": 9.554693993091078e-07,
+      "loss": 0.3494,
+      "step": 1842
+    },
+    {
+      "epoch": 0.3524237498804857,
+      "grad_norm": 2.6580395224268587,
+      "learning_rate": 9.555383711629822e-07,
+      "loss": 0.2027,
+      "step": 1843
+    },
+    {
+      "epoch": 0.352614972750741,
+      "grad_norm": 3.975650825242696,
+      "learning_rate": 9.556073056033188e-07,
+      "loss": 0.4348,
+      "step": 1844
+    },
+    {
+      "epoch": 0.35280619562099624,
+      "grad_norm": 4.764596096344759,
+      "learning_rate": 9.55676202670685e-07,
+      "loss": 0.4367,
+      "step": 1845
+    },
+    {
+      "epoch": 0.35299741849125155,
+      "grad_norm": 3.7225317438845993,
+      "learning_rate": 9.557450624055825e-07,
+      "loss": 0.179,
+      "step": 1846
+    },
+    {
+      "epoch": 0.35318864136150685,
+      "grad_norm": 5.5438131128823604,
+      "learning_rate": 9.558138848484475e-07,
+      "loss": 0.2604,
+      "step": 1847
+    },
+    {
+      "epoch": 0.3533798642317621,
+      "grad_norm": 4.068124955740369,
+      "learning_rate": 9.5588267003965e-07,
+      "loss": 0.1906,
+      "step": 1848
+    },
+    {
+      "epoch": 0.3535710871020174,
+      "grad_norm": 5.866446579070493,
+      "learning_rate": 9.55951418019495e-07,
+      "loss": 0.1816,
+      "step": 1849
+    },
+    {
+      "epoch": 0.3537623099722727,
+      "grad_norm": 7.013745028423151,
+      "learning_rate": 9.560201288282218e-07,
+      "loss": 0.0862,
+      "step": 1850
+    },
+    {
+      "epoch": 0.35395353284252795,
+      "grad_norm": 5.414578411848124,
+      "learning_rate": 9.560888025060044e-07,
+      "loss": 0.2168,
+      "step": 1851
+    },
+    {
+      "epoch": 0.35414475571278325,
+      "grad_norm": 5.535939960010832,
+      "learning_rate": 9.56157439092952e-07,
+      "loss": 0.4573,
+      "step": 1852
+    },
+    {
+      "epoch": 0.35433597858303856,
+      "grad_norm": 2.47781879317925,
+      "learning_rate": 9.562260386291086e-07,
+      "loss": 0.3963,
+      "step": 1853
+    },
+    {
+      "epoch": 0.3545272014532938,
+      "grad_norm": 4.086773234394141,
+      "learning_rate": 9.562946011544536e-07,
+      "loss": 0.3497,
+      "step": 1854
+    },
+    {
+      "epoch": 0.3547184243235491,
+      "grad_norm": 3.0837767170528423,
+      "learning_rate": 9.563631267089013e-07,
+      "loss": 0.3723,
+      "step": 1855
+    },
+    {
+      "epoch": 0.35490964719380436,
+      "grad_norm": 4.308354472262814,
+      "learning_rate": 9.564316153323021e-07,
+      "loss": 0.2748,
+      "step": 1856
+    },
+    {
+      "epoch": 0.35510087006405966,
+      "grad_norm": 4.1639024084639535,
+      "learning_rate": 9.565000670644416e-07,
+      "loss": 0.4804,
+      "step": 1857
+    },
+    {
+      "epoch": 0.35529209293431496,
+      "grad_norm": 2.7283283625398007,
+      "learning_rate": 9.56568481945041e-07,
+      "loss": 0.2356,
+      "step": 1858
+    },
+    {
+      "epoch": 0.3554833158045702,
+      "grad_norm": 2.6876218901205897,
+      "learning_rate": 9.566368600137577e-07,
+      "loss": 0.1227,
+      "step": 1859
+    },
+    {
+      "epoch": 0.3556745386748255,
+      "grad_norm": 3.3694156081169577,
+      "learning_rate": 9.567052013101854e-07,
+      "loss": 0.1479,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3558657615450808,
+      "grad_norm": 4.022114656624691,
+      "learning_rate": 9.567735058738528e-07,
+      "loss": 0.1845,
+      "step": 1861
+    },
+    {
+      "epoch": 0.35605698441533606,
+      "grad_norm": 4.349289031409082,
+      "learning_rate": 9.568417737442266e-07,
+      "loss": 0.1084,
+      "step": 1862
+    },
+    {
+      "epoch": 0.35624820728559137,
+      "grad_norm": 4.256451142385479,
+      "learning_rate": 9.569100049607081e-07,
+      "loss": 0.1315,
+      "step": 1863
+    },
+    {
+      "epoch": 0.3564394301558466,
+      "grad_norm": 3.2463151142462876,
+      "learning_rate": 9.569781995626367e-07,
+      "loss": 0.2829,
+      "step": 1864
+    },
+    {
+      "epoch": 0.3566306530261019,
+      "grad_norm": 3.336144612871379,
+      "learning_rate": 9.570463575892874e-07,
+      "loss": 0.4244,
+      "step": 1865
+    },
+    {
+      "epoch": 0.3568218758963572,
+      "grad_norm": 3.5465774747441783,
+      "learning_rate": 9.57114479079873e-07,
+      "loss": 0.2037,
+      "step": 1866
+    },
+    {
+      "epoch": 0.35701309876661247,
+      "grad_norm": 3.478528647636699,
+      "learning_rate": 9.571825640735423e-07,
+      "loss": 0.2668,
+      "step": 1867
+    },
+    {
+      "epoch": 0.35720432163686777,
+      "grad_norm": 5.673964606792226,
+      "learning_rate": 9.57250612609382e-07,
+      "loss": 0.2109,
+      "step": 1868
+    },
+    {
+      "epoch": 0.3573955445071231,
+      "grad_norm": 2.191607297762859,
+      "learning_rate": 9.573186247264158e-07,
+      "loss": 0.0984,
+      "step": 1869
+    },
+    {
+      "epoch": 0.3575867673773783,
+      "grad_norm": 3.086392989525316,
+      "learning_rate": 9.573866004636045e-07,
+      "loss": 0.2029,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3577779902476336,
+      "grad_norm": 3.3701914688188004,
+      "learning_rate": 9.574545398598465e-07,
+      "loss": 0.2479,
+      "step": 1871
+    },
+    {
+      "epoch": 0.3579692131178889,
+      "grad_norm": 3.1452694991801806,
+      "learning_rate": 9.575224429539785e-07,
+      "loss": 0.0902,
+      "step": 1872
+    },
+    {
+      "epoch": 0.3581604359881442,
+      "grad_norm": 3.8676318597292294,
+      "learning_rate": 9.57590309784774e-07,
+      "loss": 0.301,
+      "step": 1873
+    },
+    {
+      "epoch": 0.3583516588583995,
+      "grad_norm": 2.5927095279611665,
+      "learning_rate": 9.576581403909448e-07,
+      "loss": 0.0618,
+      "step": 1874
+    },
+    {
+      "epoch": 0.3585428817286547,
+      "grad_norm": 7.321452602652743,
+      "learning_rate": 9.577259348111412e-07,
+      "loss": 0.2955,
+      "step": 1875
+    },
+    {
+      "epoch": 0.35873410459891003,
+      "grad_norm": 3.8801482671743805,
+      "learning_rate": 9.577936930839512e-07,
+      "loss": 0.164,
+      "step": 1876
+    },
+    {
+      "epoch": 0.35892532746916533,
+      "grad_norm": 4.251215592614962,
+      "learning_rate": 9.578614152479008e-07,
+      "loss": 0.4894,
+      "step": 1877
+    },
+    {
+      "epoch": 0.3591165503394206,
+      "grad_norm": 3.885605603789855,
+      "learning_rate": 9.579291013414553e-07,
+      "loss": 0.5988,
+      "step": 1878
+    },
+    {
+      "epoch": 0.3593077732096759,
+      "grad_norm": 4.788745465138319,
+      "learning_rate": 9.579967514030178e-07,
+      "loss": 0.27,
+      "step": 1879
+    },
+    {
+      "epoch": 0.35949899607993113,
+      "grad_norm": 3.5892192641927907,
+      "learning_rate": 9.580643654709306e-07,
+      "loss": 0.1095,
+      "step": 1880
+    },
+    {
+      "epoch": 0.35969021895018644,
+      "grad_norm": 3.314724589050699,
+      "learning_rate": 9.581319435834745e-07,
+      "loss": 0.2489,
+      "step": 1881
+    },
+    {
+      "epoch": 0.35988144182044174,
+      "grad_norm": 3.780676806172664,
+      "learning_rate": 9.581994857788696e-07,
+      "loss": 0.4269,
+      "step": 1882
+    },
+    {
+      "epoch": 0.360072664690697,
+      "grad_norm": 4.445951982301503,
+      "learning_rate": 9.582669920952746e-07,
+      "loss": 0.1611,
+      "step": 1883
+    },
+    {
+      "epoch": 0.3602638875609523,
+      "grad_norm": 2.378515352351561,
+      "learning_rate": 9.58334462570788e-07,
+      "loss": 0.3323,
+      "step": 1884
+    },
+    {
+      "epoch": 0.3604551104312076,
+      "grad_norm": 4.868226848109414,
+      "learning_rate": 9.58401897243447e-07,
+      "loss": 0.2436,
+      "step": 1885
+    },
+    {
+      "epoch": 0.36064633330146284,
+      "grad_norm": 4.146635709526403,
+      "learning_rate": 9.584692961512293e-07,
+      "loss": 0.1866,
+      "step": 1886
+    },
+    {
+      "epoch": 0.36083755617171814,
+      "grad_norm": 3.114817547789608,
+      "learning_rate": 9.585366593320506e-07,
+      "loss": 0.0843,
+      "step": 1887
+    },
+    {
+      "epoch": 0.36102877904197345,
+      "grad_norm": 8.810326707305077,
+      "learning_rate": 9.58603986823768e-07,
+      "loss": 0.2132,
+      "step": 1888
+    },
+    {
+      "epoch": 0.3612200019122287,
+      "grad_norm": 5.987478223534088,
+      "learning_rate": 9.586712786641778e-07,
+      "loss": 0.6297,
+      "step": 1889
+    },
+    {
+      "epoch": 0.361411224782484,
+      "grad_norm": 5.66447398828566,
+      "learning_rate": 9.587385348910158e-07,
+      "loss": 0.5156,
+      "step": 1890
+    },
+    {
+      "epoch": 0.36160244765273925,
+      "grad_norm": 3.5962521011192594,
+      "learning_rate": 9.588057555419586e-07,
+      "loss": 0.2022,
+      "step": 1891
+    },
+    {
+      "epoch": 0.36179367052299455,
+      "grad_norm": 4.889297841528854,
+      "learning_rate": 9.588729406546226e-07,
+      "loss": 0.5287,
+      "step": 1892
+    },
+    {
+      "epoch": 0.36198489339324985,
+      "grad_norm": 3.704521076996608,
+      "learning_rate": 9.58940090266565e-07,
+      "loss": 0.2588,
+      "step": 1893
+    },
+    {
+      "epoch": 0.3621761162635051,
+      "grad_norm": 3.5016693493849296,
+      "learning_rate": 9.590072044152832e-07,
+      "loss": 0.1738,
+      "step": 1894
+    },
+    {
+      "epoch": 0.3623673391337604,
+      "grad_norm": 4.703616094454741,
+      "learning_rate": 9.590742831382153e-07,
+      "loss": 0.0753,
+      "step": 1895
+    },
+    {
+      "epoch": 0.3625585620040157,
+      "grad_norm": 2.4147451478739757,
+      "learning_rate": 9.591413264727401e-07,
+      "loss": 0.2082,
+      "step": 1896
+    },
+    {
+      "epoch": 0.36274978487427095,
+      "grad_norm": 5.139727389457485,
+      "learning_rate": 9.592083344561772e-07,
+      "loss": 0.1835,
+      "step": 1897
+    },
+    {
+      "epoch": 0.36294100774452626,
+      "grad_norm": 8.106579837250408,
+      "learning_rate": 9.592753071257875e-07,
+      "loss": 0.151,
+      "step": 1898
+    },
+    {
+      "epoch": 0.3631322306147815,
+      "grad_norm": 2.57921702058988,
+      "learning_rate": 9.593422445187727e-07,
+      "loss": 0.0661,
+      "step": 1899
+    },
+    {
+      "epoch": 0.3633234534850368,
+      "grad_norm": 7.633383722708133,
+      "learning_rate": 9.594091466722759e-07,
+      "loss": 0.2719,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3635146763552921,
+      "grad_norm": 5.907709173566634,
+      "learning_rate": 9.594760136233813e-07,
+      "loss": 0.5175,
+      "step": 1901
+    },
+    {
+      "epoch": 0.36370589922554736,
+      "grad_norm": 4.639107093069924,
+      "learning_rate": 9.595428454091154e-07,
+      "loss": 0.5915,
+      "step": 1902
+    },
+    {
+      "epoch": 0.36389712209580266,
+      "grad_norm": 4.4015248170755585,
+      "learning_rate": 9.596096420664452e-07,
+      "loss": 0.4248,
+      "step": 1903
+    },
+    {
+      "epoch": 0.36408834496605796,
+      "grad_norm": 4.291812980651495,
+      "learning_rate": 9.596764036322802e-07,
+      "loss": 0.2761,
+      "step": 1904
+    },
+    {
+      "epoch": 0.3642795678363132,
+      "grad_norm": 3.390718239060388,
+      "learning_rate": 9.597431301434716e-07,
+      "loss": 0.3778,
+      "step": 1905
+    },
+    {
+      "epoch": 0.3644707907065685,
+      "grad_norm": 3.2411695973655434,
+      "learning_rate": 9.598098216368121e-07,
+      "loss": 0.376,
+      "step": 1906
+    },
+    {
+      "epoch": 0.36466201357682376,
+      "grad_norm": 3.6940551941553093,
+      "learning_rate": 9.598764781490377e-07,
+      "loss": 0.235,
+      "step": 1907
+    },
+    {
+      "epoch": 0.36485323644707907,
+      "grad_norm": 4.111133271807721,
+      "learning_rate": 9.599430997168252e-07,
+      "loss": 0.1886,
+      "step": 1908
+    },
+    {
+      "epoch": 0.36504445931733437,
+      "grad_norm": 5.063770841105034,
+      "learning_rate": 9.600096863767947e-07,
+      "loss": 0.1883,
+      "step": 1909
+    },
+    {
+      "epoch": 0.3652356821875896,
+      "grad_norm": 3.0540758383976927,
+      "learning_rate": 9.600762381655082e-07,
+      "loss": 0.097,
+      "step": 1910
+    },
+    {
+      "epoch": 0.3654269050578449,
+      "grad_norm": 3.785420159586903,
+      "learning_rate": 9.60142755119471e-07,
+      "loss": 0.0973,
+      "step": 1911
+    },
+    {
+      "epoch": 0.3656181279281002,
+      "grad_norm": 2.904653459321783,
+      "learning_rate": 9.6020923727513e-07,
+      "loss": 0.0668,
+      "step": 1912
+    },
+    {
+      "epoch": 0.36580935079835547,
+      "grad_norm": 10.710088049262211,
+      "learning_rate": 9.602756846688761e-07,
+      "loss": 0.3137,
+      "step": 1913
+    },
+    {
+      "epoch": 0.3660005736686108,
+      "grad_norm": 3.438264102975378,
+      "learning_rate": 9.603420973370425e-07,
+      "loss": 0.2105,
+      "step": 1914
+    },
+    {
+      "epoch": 0.366191796538866,
+      "grad_norm": 5.031204507752405,
+      "learning_rate": 9.604084753159057e-07,
+      "loss": 0.7346,
+      "step": 1915
+    },
+    {
+      "epoch": 0.3663830194091213,
+      "grad_norm": 2.393259130600954,
+      "learning_rate": 9.604748186416851e-07,
+      "loss": 0.2973,
+      "step": 1916
+    },
+    {
+      "epoch": 0.3665742422793766,
+      "grad_norm": 5.184898183589947,
+      "learning_rate": 9.605411273505437e-07,
+      "loss": 0.3938,
+      "step": 1917
+    },
+    {
+      "epoch": 0.3667654651496319,
+      "grad_norm": 2.721438689797711,
+      "learning_rate": 9.606074014785878e-07,
+      "loss": 0.3781,
+      "step": 1918
+    },
+    {
+      "epoch": 0.3669566880198872,
+      "grad_norm": 4.922401811774169,
+      "learning_rate": 9.60673641061867e-07,
+      "loss": 0.2792,
+      "step": 1919
+    },
+    {
+      "epoch": 0.3671479108901425,
+      "grad_norm": 3.617883471112202,
+      "learning_rate": 9.607398461363754e-07,
+      "loss": 0.2506,
+      "step": 1920
+    },
+    {
+      "epoch": 0.36733913376039773,
+      "grad_norm": 4.603040991985773,
+      "learning_rate": 9.608060167380494e-07,
+      "loss": 0.3263,
+      "step": 1921
+    },
+    {
+      "epoch": 0.36753035663065303,
+      "grad_norm": 4.681980596555111,
+      "learning_rate": 9.60872152902771e-07,
+      "loss": 0.095,
+      "step": 1922
+    },
+    {
+      "epoch": 0.36772157950090834,
+      "grad_norm": 4.857543992860808,
+      "learning_rate": 9.60938254666365e-07,
+      "loss": 0.2976,
+      "step": 1923
+    },
+    {
+      "epoch": 0.3679128023711636,
+      "grad_norm": 3.903271203093618,
+      "learning_rate": 9.610043220646009e-07,
+      "loss": 0.1144,
+      "step": 1924
+    },
+    {
+      "epoch": 0.3681040252414189,
+      "grad_norm": 6.908363812804618,
+      "learning_rate": 9.610703551331918e-07,
+      "loss": 0.1067,
+      "step": 1925
+    },
+    {
+      "epoch": 0.36829524811167413,
+      "grad_norm": 4.71147656242875,
+      "learning_rate": 9.611363539077962e-07,
+      "loss": 0.3665,
+      "step": 1926
+    },
+    {
+      "epoch": 0.36848647098192944,
+      "grad_norm": 3.413554208309698,
+      "learning_rate": 9.61202318424016e-07,
+      "loss": 0.2834,
+      "step": 1927
+    },
+    {
+      "epoch": 0.36867769385218474,
+      "grad_norm": 2.890458055778841,
+      "learning_rate": 9.612682487173984e-07,
+      "loss": 0.3765,
+      "step": 1928
+    },
+    {
+      "epoch": 0.36886891672244,
+      "grad_norm": 5.969358722743258,
+      "learning_rate": 9.61334144823435e-07,
+      "loss": 0.3161,
+      "step": 1929
+    },
+    {
+      "epoch": 0.3690601395926953,
+      "grad_norm": 4.1764581201299285,
+      "learning_rate": 9.614000067775625e-07,
+      "loss": 0.3592,
+      "step": 1930
+    },
+    {
+      "epoch": 0.3692513624629506,
+      "grad_norm": 4.021841025042249,
+      "learning_rate": 9.614658346151619e-07,
+      "loss": 0.3974,
+      "step": 1931
+    },
+    {
+      "epoch": 0.36944258533320584,
+      "grad_norm": 4.173728274060073,
+      "learning_rate": 9.615316283715599e-07,
+      "loss": 0.3173,
+      "step": 1932
+    },
+    {
+      "epoch": 0.36963380820346114,
+      "grad_norm": 4.120738052999782,
+      "learning_rate": 9.615973880820282e-07,
+      "loss": 0.5058,
+      "step": 1933
+    },
+    {
+      "epoch": 0.3698250310737164,
+      "grad_norm": 3.6324624169775515,
+      "learning_rate": 9.616631137817836e-07,
+      "loss": 0.1227,
+      "step": 1934
+    },
+    {
+      "epoch": 0.3700162539439717,
+      "grad_norm": 4.856041454810682,
+      "learning_rate": 9.617288055059883e-07,
+      "loss": 0.1041,
+      "step": 1935
+    },
+    {
+      "epoch": 0.370207476814227,
+      "grad_norm": 4.251679257013233,
+      "learning_rate": 9.6179446328975e-07,
+      "loss": 0.1151,
+      "step": 1936
+    },
+    {
+      "epoch": 0.37039869968448225,
+      "grad_norm": 2.9701544401043,
+      "learning_rate": 9.618600871681225e-07,
+      "loss": 0.0506,
+      "step": 1937
+    },
+    {
+      "epoch": 0.37058992255473755,
+      "grad_norm": 11.92993051897202,
+      "learning_rate": 9.619256771761046e-07,
+      "loss": 0.2304,
+      "step": 1938
+    },
+    {
+      "epoch": 0.37078114542499285,
+      "grad_norm": 5.371056285380702,
+      "learning_rate": 9.619912333486412e-07,
+      "loss": 0.5295,
+      "step": 1939
+    },
+    {
+      "epoch": 0.3709723682952481,
+      "grad_norm": 3.5728240747196724,
+      "learning_rate": 9.620567557206235e-07,
+      "loss": 0.5179,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3711635911655034,
+      "grad_norm": 3.1917718423457697,
+      "learning_rate": 9.62122244326888e-07,
+      "loss": 0.2759,
+      "step": 1941
+    },
+    {
+      "epoch": 0.37135481403575865,
+      "grad_norm": 5.5467916133146815,
+      "learning_rate": 9.621876992022183e-07,
+      "loss": 0.1627,
+      "step": 1942
+    },
+    {
+      "epoch": 0.37154603690601395,
+      "grad_norm": 6.073061837172828,
+      "learning_rate": 9.622531203813436e-07,
+      "loss": 0.1475,
+      "step": 1943
+    },
+    {
+      "epoch": 0.37173725977626926,
+      "grad_norm": 2.8777267130768776,
+      "learning_rate": 9.623185078989397e-07,
+      "loss": 0.1248,
+      "step": 1944
+    },
+    {
+      "epoch": 0.3719284826465245,
+      "grad_norm": 2.054884056291801,
+      "learning_rate": 9.623838617896288e-07,
+      "loss": 0.1233,
+      "step": 1945
+    },
+    {
+      "epoch": 0.3721197055167798,
+      "grad_norm": 2.6686852583525456,
+      "learning_rate": 9.624491820879798e-07,
+      "loss": 0.389,
+      "step": 1946
+    },
+    {
+      "epoch": 0.3723109283870351,
+      "grad_norm": 1.7060941659937516,
+      "learning_rate": 9.625144688285086e-07,
+      "loss": 0.0292,
+      "step": 1947
+    },
+    {
+      "epoch": 0.37250215125729036,
+      "grad_norm": 3.8035421977976327,
+      "learning_rate": 9.625797220456776e-07,
+      "loss": 0.2532,
+      "step": 1948
+    },
+    {
+      "epoch": 0.37269337412754566,
+      "grad_norm": 3.3751178297031177,
+      "learning_rate": 9.626449417738959e-07,
+      "loss": 0.1132,
+      "step": 1949
+    },
+    {
+      "epoch": 0.3728845969978009,
+      "grad_norm": 7.384040851434552,
+      "learning_rate": 9.627101280475203e-07,
+      "loss": 0.099,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3730758198680562,
+      "grad_norm": 25.03401738369588,
+      "learning_rate": 9.62775280900854e-07,
+      "loss": 0.5992,
+      "step": 1951
+    },
+    {
+      "epoch": 0.3732670427383115,
+      "grad_norm": 4.289540304172251,
+      "learning_rate": 9.628404003681484e-07,
+      "loss": 0.2882,
+      "step": 1952
+    },
+    {
+      "epoch": 0.37345826560856676,
+      "grad_norm": 3.3581608907065625,
+      "learning_rate": 9.629054864836015e-07,
+      "loss": 0.2336,
+      "step": 1953
+    },
+    {
+      "epoch": 0.37364948847882207,
+      "grad_norm": 2.239251211217374,
+      "learning_rate": 9.62970539281359e-07,
+      "loss": 0.1873,
+      "step": 1954
+    },
+    {
+      "epoch": 0.37384071134907737,
+      "grad_norm": 5.445938227644644,
+      "learning_rate": 9.630355587955142e-07,
+      "loss": 0.3242,
+      "step": 1955
+    },
+    {
+      "epoch": 0.3740319342193326,
+      "grad_norm": 4.220835530518051,
+      "learning_rate": 9.631005450601081e-07,
+      "loss": 0.2984,
+      "step": 1956
+    },
+    {
+      "epoch": 0.3742231570895879,
+      "grad_norm": 3.2207482810755748,
+      "learning_rate": 9.631654981091298e-07,
+      "loss": 0.4254,
+      "step": 1957
+    },
+    {
+      "epoch": 0.3744143799598432,
+      "grad_norm": 2.957692967985537,
+      "learning_rate": 9.632304179765158e-07,
+      "loss": 0.1127,
+      "step": 1958
+    },
+    {
+      "epoch": 0.37460560283009847,
+      "grad_norm": 5.75927980688931,
+      "learning_rate": 9.632953046961508e-07,
+      "loss": 0.1228,
+      "step": 1959
+    },
+    {
+      "epoch": 0.3747968257003538,
+      "grad_norm": 3.6664060731241968,
+      "learning_rate": 9.633601583018676e-07,
+      "loss": 0.1592,
+      "step": 1960
+    },
+    {
+      "epoch": 0.374988048570609,
+      "grad_norm": 2.9866241448077515,
+      "learning_rate": 9.634249788274476e-07,
+      "loss": 0.1094,
+      "step": 1961
+    },
+    {
+      "epoch": 0.3751792714408643,
+      "grad_norm": 4.7360099764280665,
+      "learning_rate": 9.6348976630662e-07,
+      "loss": 0.1422,
+      "step": 1962
+    },
+    {
+      "epoch": 0.37537049431111963,
+      "grad_norm": 8.927805945143556,
+      "learning_rate": 9.635545207730628e-07,
+      "loss": 0.5081,
+      "step": 1963
+    },
+    {
+      "epoch": 0.3755617171813749,
+      "grad_norm": 3.6708289421710547,
+      "learning_rate": 9.636192422604022e-07,
+      "loss": 0.2974,
+      "step": 1964
+    },
+    {
+      "epoch": 0.3757529400516302,
+      "grad_norm": 4.646099111484316,
+      "learning_rate": 9.636839308022134e-07,
+      "loss": 0.738,
+      "step": 1965
+    },
+    {
+      "epoch": 0.3759441629218855,
+      "grad_norm": 2.7193167961623423,
+      "learning_rate": 9.637485864320203e-07,
+      "loss": 0.3337,
+      "step": 1966
+    },
+    {
+      "epoch": 0.37613538579214073,
+      "grad_norm": 4.377587343539519,
+      "learning_rate": 9.638132091832955e-07,
+      "loss": 0.348,
+      "step": 1967
+    },
+    {
+      "epoch": 0.37632660866239603,
+      "grad_norm": 4.08031489596925,
+      "learning_rate": 9.638777990894607e-07,
+      "loss": 0.3344,
+      "step": 1968
+    },
+    {
+      "epoch": 0.3765178315326513,
+      "grad_norm": 3.572006791854843,
+      "learning_rate": 9.639423561838865e-07,
+      "loss": 0.2262,
+      "step": 1969
+    },
+    {
+      "epoch": 0.3767090544029066,
+      "grad_norm": 3.4550815752614907,
+      "learning_rate": 9.640068804998928e-07,
+      "loss": 0.1021,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3769002772731619,
+      "grad_norm": 3.789634677119016,
+      "learning_rate": 9.640713720707486e-07,
+      "loss": 0.2475,
+      "step": 1971
+    },
+    {
+      "epoch": 0.37709150014341714,
+      "grad_norm": 3.101674113919765,
+      "learning_rate": 9.641358309296726e-07,
+      "loss": 0.2124,
+      "step": 1972
+    },
+    {
+      "epoch": 0.37728272301367244,
+      "grad_norm": 6.88050566345341,
+      "learning_rate": 9.642002571098329e-07,
+      "loss": 0.18,
+      "step": 1973
+    },
+    {
+      "epoch": 0.37747394588392774,
+      "grad_norm": 2.5938200424112985,
+      "learning_rate": 9.64264650644347e-07,
+      "loss": 0.072,
+      "step": 1974
+    },
+    {
+      "epoch": 0.377665168754183,
+      "grad_norm": 4.696698940934203,
+      "learning_rate": 9.643290115662817e-07,
+      "loss": 0.2523,
+      "step": 1975
+    },
+    {
+      "epoch": 0.3778563916244383,
+      "grad_norm": 6.347034524840832,
+      "learning_rate": 9.643933399086547e-07,
+      "loss": 0.1592,
+      "step": 1976
+    },
+    {
+      "epoch": 0.37804761449469354,
+      "grad_norm": 4.752801771541227,
+      "learning_rate": 9.644576357044328e-07,
+      "loss": 0.4108,
+      "step": 1977
+    },
+    {
+      "epoch": 0.37823883736494884,
+      "grad_norm": 5.442381498267157,
+      "learning_rate": 9.645218989865324e-07,
+      "loss": 0.4336,
+      "step": 1978
+    },
+    {
+      "epoch": 0.37843006023520415,
+      "grad_norm": 5.94738300376778,
+      "learning_rate": 9.645861297878211e-07,
+      "loss": 0.4034,
+      "step": 1979
+    },
+    {
+      "epoch": 0.3786212831054594,
+      "grad_norm": 3.0923714697239553,
+      "learning_rate": 9.646503281411158e-07,
+      "loss": 0.2202,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3788125059757147,
+      "grad_norm": 3.9352947296320546,
+      "learning_rate": 9.64714494079184e-07,
+      "loss": 0.6811,
+      "step": 1981
+    },
+    {
+      "epoch": 0.37900372884597,
+      "grad_norm": 3.454591051512598,
+      "learning_rate": 9.647786276347436e-07,
+      "loss": 0.2141,
+      "step": 1982
+    },
+    {
+      "epoch": 0.37919495171622525,
+      "grad_norm": 2.382019211090283,
+      "learning_rate": 9.64842728840463e-07,
+      "loss": 0.1838,
+      "step": 1983
+    },
+    {
+      "epoch": 0.37938617458648055,
+      "grad_norm": 2.4003659922963116,
+      "learning_rate": 9.64906797728961e-07,
+      "loss": 0.1289,
+      "step": 1984
+    },
+    {
+      "epoch": 0.3795773974567358,
+      "grad_norm": 1.8812449407667904,
+      "learning_rate": 9.649708343328074e-07,
+      "loss": 0.024,
+      "step": 1985
+    },
+    {
+      "epoch": 0.3797686203269911,
+      "grad_norm": 3.3717730601318294,
+      "learning_rate": 9.650348386845227e-07,
+      "loss": 0.0858,
+      "step": 1986
+    },
+    {
+      "epoch": 0.3799598431972464,
+      "grad_norm": 5.831346046171936,
+      "learning_rate": 9.650988108165779e-07,
+      "loss": 0.1787,
+      "step": 1987
+    },
+    {
+      "epoch": 0.38015106606750165,
+      "grad_norm": 6.26623166921519,
+      "learning_rate": 9.651627507613957e-07,
+      "loss": 0.2946,
+      "step": 1988
+    },
+    {
+      "epoch": 0.38034228893775696,
+      "grad_norm": 4.135249034389104,
+      "learning_rate": 9.65226658551349e-07,
+      "loss": 0.3043,
+      "step": 1989
+    },
+    {
+      "epoch": 0.38053351180801226,
+      "grad_norm": 3.641492617453891,
+      "learning_rate": 9.652905342187628e-07,
+      "loss": 0.2115,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3807247346782675,
+      "grad_norm": 3.387371981408255,
+      "learning_rate": 9.653543777959127e-07,
+      "loss": 0.3295,
+      "step": 1991
+    },
+    {
+      "epoch": 0.3809159575485228,
+      "grad_norm": 2.658386089634555,
+      "learning_rate": 9.65418189315026e-07,
+      "loss": 0.1479,
+      "step": 1992
+    },
+    {
+      "epoch": 0.3811071804187781,
+      "grad_norm": 3.213492831575771,
+      "learning_rate": 9.654819688082818e-07,
+      "loss": 0.4939,
+      "step": 1993
+    },
+    {
+      "epoch": 0.38129840328903336,
+      "grad_norm": 2.666902342951845,
+      "learning_rate": 9.655457163078096e-07,
+      "loss": 0.146,
+      "step": 1994
+    },
+    {
+      "epoch": 0.38148962615928866,
+      "grad_norm": 3.3064926664143317,
+      "learning_rate": 9.65609431845692e-07,
+      "loss": 0.2955,
+      "step": 1995
+    },
+    {
+      "epoch": 0.3816808490295439,
+      "grad_norm": 2.9287999957803885,
+      "learning_rate": 9.656731154539626e-07,
+      "loss": 0.1631,
+      "step": 1996
+    },
+    {
+      "epoch": 0.3818720718997992,
+      "grad_norm": 2.5461587659383587,
+      "learning_rate": 9.657367671646067e-07,
+      "loss": 0.0974,
+      "step": 1997
+    },
+    {
+      "epoch": 0.3820632947700545,
+      "grad_norm": 2.5683745061044307,
+      "learning_rate": 9.65800387009562e-07,
+      "loss": 0.1465,
+      "step": 1998
+    },
+    {
+      "epoch": 0.38225451764030977,
+      "grad_norm": 2.5921752929400075,
+      "learning_rate": 9.658639750207182e-07,
+      "loss": 0.0552,
+      "step": 1999
+    },
+    {
+      "epoch": 0.38244574051056507,
+      "grad_norm": 7.42631651641331,
+      "learning_rate": 9.659275312299169e-07,
+      "loss": 0.2036,
+      "step": 2000
+    },
+    {
+      "epoch": 0.38244574051056507,
+      "eval_runtime": 724.3537,
+      "eval_samples_per_second": 2.118,
+      "eval_steps_per_second": 0.53,
+      "step": 2000
+    },
+    {
+      "epoch": 0.38263696338082037,
+      "grad_norm": 4.733819406472505,
+      "learning_rate": 9.659910556689525e-07,
+      "loss": 0.426,
+      "step": 2001
+    },
+    {
+      "epoch": 0.3828281862510756,
+      "grad_norm": 8.508288661778492,
+      "learning_rate": 9.660545483695709e-07,
+      "loss": 0.2302,
+      "step": 2002
+    },
+    {
+      "epoch": 0.3830194091213309,
+      "grad_norm": 2.8376772249642936,
+      "learning_rate": 9.661180093634709e-07,
+      "loss": 0.4224,
+      "step": 2003
+    },
+    {
+      "epoch": 0.38321063199158617,
+      "grad_norm": 3.364078936645212,
+      "learning_rate": 9.661814386823043e-07,
+      "loss": 0.2105,
+      "step": 2004
+    },
+    {
+      "epoch": 0.3834018548618415,
+      "grad_norm": 8.06262112312119,
+      "learning_rate": 9.662448363576747e-07,
+      "loss": 0.4889,
+      "step": 2005
+    },
+    {
+      "epoch": 0.3835930777320968,
+      "grad_norm": 4.714128264568096,
+      "learning_rate": 9.663082024211387e-07,
+      "loss": 0.4211,
+      "step": 2006
+    },
+    {
+      "epoch": 0.383784300602352,
+      "grad_norm": 3.338386155115448,
+      "learning_rate": 9.663715369042058e-07,
+      "loss": 0.216,
+      "step": 2007
+    },
+    {
+      "epoch": 0.3839755234726073,
+      "grad_norm": 4.2380458784009924,
+      "learning_rate": 9.66434839838339e-07,
+      "loss": 0.1849,
+      "step": 2008
+    },
+    {
+      "epoch": 0.38416674634286263,
+      "grad_norm": 3.708317588744154,
+      "learning_rate": 9.664981112549531e-07,
+      "loss": 0.1608,
+      "step": 2009
+    },
+    {
+      "epoch": 0.3843579692131179,
+      "grad_norm": 6.886400515105359,
+      "learning_rate": 9.665613511854166e-07,
+      "loss": 0.2693,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3845491920833732,
+      "grad_norm": 2.81619088332247,
+      "learning_rate": 9.666245596610516e-07,
+      "loss": 0.0407,
+      "step": 2011
+    },
+    {
+      "epoch": 0.38474041495362843,
+      "grad_norm": 2.446039938662237,
+      "learning_rate": 9.66687736713133e-07,
+      "loss": 0.0619,
+      "step": 2012
+    },
+    {
+      "epoch": 0.38493163782388373,
+      "grad_norm": 7.344923079927306,
+      "learning_rate": 9.66750882372889e-07,
+      "loss": 0.1749,
+      "step": 2013
+    },
+    {
+      "epoch": 0.38512286069413904,
+      "grad_norm": 5.05703553462278,
+      "learning_rate": 9.668139966715014e-07,
+      "loss": 0.2611,
+      "step": 2014
+    },
+    {
+      "epoch": 0.3853140835643943,
+      "grad_norm": 5.416108420533477,
+      "learning_rate": 9.66877079640106e-07,
+      "loss": 0.4099,
+      "step": 2015
+    },
+    {
+      "epoch": 0.3855053064346496,
+      "grad_norm": 4.912951515161943,
+      "learning_rate": 9.669401313097915e-07,
+      "loss": 0.2948,
+      "step": 2016
+    },
+    {
+      "epoch": 0.3856965293049049,
+      "grad_norm": 5.327505842015014,
+      "learning_rate": 9.670031517116008e-07,
+      "loss": 0.4563,
+      "step": 2017
+    },
+    {
+      "epoch": 0.38588775217516014,
+      "grad_norm": 3.8561878712501114,
+      "learning_rate": 9.670661408765307e-07,
+      "loss": 0.12,
+      "step": 2018
+    },
+    {
+      "epoch": 0.38607897504541544,
+      "grad_norm": 2.497220210535552,
+      "learning_rate": 9.671290988355316e-07,
+      "loss": 0.2282,
+      "step": 2019
+    },
+    {
+      "epoch": 0.3862701979156707,
+      "grad_norm": 4.338682003735814,
+      "learning_rate": 9.671920256195083e-07,
+      "loss": 0.4325,
+      "step": 2020
+    },
+    {
+      "epoch": 0.386461420785926,
+      "grad_norm": 2.269766133902853,
+      "learning_rate": 9.672549212593194e-07,
+      "loss": 0.0659,
+      "step": 2021
+    },
+    {
+      "epoch": 0.3866526436561813,
+      "grad_norm": 3.5653061104816777,
+      "learning_rate": 9.673177857857777e-07,
+      "loss": 0.3118,
+      "step": 2022
+    },
+    {
+      "epoch": 0.38684386652643654,
+      "grad_norm": 3.29186863018091,
+      "learning_rate": 9.673806192296507e-07,
+      "loss": 0.2394,
+      "step": 2023
+    },
+    {
+      "epoch": 0.38703508939669184,
+      "grad_norm": 3.1765278481743318,
+      "learning_rate": 9.674434216216598e-07,
+      "loss": 0.087,
+      "step": 2024
+    },
+    {
+      "epoch": 0.38722631226694715,
+      "grad_norm": 5.189776553009389,
+      "learning_rate": 9.675061929924813e-07,
+      "loss": 0.1452,
+      "step": 2025
+    },
+    {
+      "epoch": 0.3874175351372024,
+      "grad_norm": 5.538613628067994,
+      "learning_rate": 9.675689333727457e-07,
+      "loss": 0.3658,
+      "step": 2026
+    },
+    {
+      "epoch": 0.3876087580074577,
+      "grad_norm": 3.642896475699799,
+      "learning_rate": 9.676316427930383e-07,
+      "loss": 0.3299,
+      "step": 2027
+    },
+    {
+      "epoch": 0.387799980877713,
+      "grad_norm": 4.566728905209588,
+      "learning_rate": 9.67694321283899e-07,
+      "loss": 0.514,
+      "step": 2028
+    },
+    {
+      "epoch": 0.38799120374796825,
+      "grad_norm": 3.2393743973028,
+      "learning_rate": 9.677569688758232e-07,
+      "loss": 0.1606,
+      "step": 2029
+    },
+    {
+      "epoch": 0.38818242661822355,
+      "grad_norm": 3.2014644133034555,
+      "learning_rate": 9.678195855992602e-07,
+      "loss": 0.2163,
+      "step": 2030
+    },
+    {
+      "epoch": 0.3883736494884788,
+      "grad_norm": 3.5479080107503607,
+      "learning_rate": 9.678821714846148e-07,
+      "loss": 0.1766,
+      "step": 2031
+    },
+    {
+      "epoch": 0.3885648723587341,
+      "grad_norm": 2.109778132593167,
+      "learning_rate": 9.679447265622473e-07,
+      "loss": 0.1401,
+      "step": 2032
+    },
+    {
+      "epoch": 0.3887560952289894,
+      "grad_norm": 3.5963609581656923,
+      "learning_rate": 9.680072508624726e-07,
+      "loss": 0.2626,
+      "step": 2033
+    },
+    {
+      "epoch": 0.38894731809924465,
+      "grad_norm": 2.6745746203025726,
+      "learning_rate": 9.680697444155608e-07,
+      "loss": 0.074,
+      "step": 2034
+    },
+    {
+      "epoch": 0.38913854096949996,
+      "grad_norm": 3.6950218941294115,
+      "learning_rate": 9.681322072517381e-07,
+      "loss": 0.1943,
+      "step": 2035
+    },
+    {
+      "epoch": 0.38932976383975526,
+      "grad_norm": 2.9032493605734953,
+      "learning_rate": 9.681946394011855e-07,
+      "loss": 0.1442,
+      "step": 2036
+    },
+    {
+      "epoch": 0.3895209867100105,
+      "grad_norm": 5.324146743253906,
+      "learning_rate": 9.682570408940396e-07,
+      "loss": 0.1173,
+      "step": 2037
+    },
+    {
+      "epoch": 0.3897122095802658,
+      "grad_norm": 6.131291486345054,
+      "learning_rate": 9.683194117603932e-07,
+      "loss": 0.3048,
+      "step": 2038
+    },
+    {
+      "epoch": 0.38990343245052106,
+      "grad_norm": 5.2815439718016375,
+      "learning_rate": 9.68381752030294e-07,
+      "loss": 0.7102,
+      "step": 2039
+    },
+    {
+      "epoch": 0.39009465532077636,
+      "grad_norm": 4.918345902746917,
+      "learning_rate": 9.684440617337459e-07,
+      "loss": 0.2249,
+      "step": 2040
+    },
+    {
+      "epoch": 0.39028587819103167,
+      "grad_norm": 2.2947992819649468,
+      "learning_rate": 9.685063409007086e-07,
+      "loss": 0.2141,
+      "step": 2041
+    },
+    {
+      "epoch": 0.3904771010612869,
+      "grad_norm": 4.639738980876936,
+      "learning_rate": 9.685685895610981e-07,
+      "loss": 0.4358,
+      "step": 2042
+    },
+    {
+      "epoch": 0.3906683239315422,
+      "grad_norm": 3.610613325647224,
+      "learning_rate": 9.68630807744786e-07,
+      "loss": 0.4204,
+      "step": 2043
+    },
+    {
+      "epoch": 0.3908595468017975,
+      "grad_norm": 4.439667078544352,
+      "learning_rate": 9.686929954816005e-07,
+      "loss": 0.2393,
+      "step": 2044
+    },
+    {
+      "epoch": 0.39105076967205277,
+      "grad_norm": 3.1366488210215717,
+      "learning_rate": 9.687551528013255e-07,
+      "loss": 0.3632,
+      "step": 2045
+    },
+    {
+      "epoch": 0.39124199254230807,
+      "grad_norm": 3.1937231014186493,
+      "learning_rate": 9.688172797337015e-07,
+      "loss": 0.2112,
+      "step": 2046
+    },
+    {
+      "epoch": 0.3914332154125633,
+      "grad_norm": 2.7562042820743797,
+      "learning_rate": 9.688793763084255e-07,
+      "loss": 0.1421,
+      "step": 2047
+    },
+    {
+      "epoch": 0.3916244382828186,
+      "grad_norm": 2.544564351565472,
+      "learning_rate": 9.68941442555151e-07,
+      "loss": 0.0743,
+      "step": 2048
+    },
+    {
+      "epoch": 0.3918156611530739,
+      "grad_norm": 4.623779960755948,
+      "learning_rate": 9.690034785034877e-07,
+      "loss": 0.1417,
+      "step": 2049
+    },
+    {
+      "epoch": 0.39200688402332917,
+      "grad_norm": 3.115786742026731,
+      "learning_rate": 9.690654841830024e-07,
+      "loss": 0.1041,
+      "step": 2050
+    },
+    {
+      "epoch": 0.3921981068935845,
+      "grad_norm": 4.471256006933736,
+      "learning_rate": 9.691274596232184e-07,
+      "loss": 0.4931,
+      "step": 2051
+    },
+    {
+      "epoch": 0.3923893297638398,
+      "grad_norm": 4.190941378650175,
+      "learning_rate": 9.69189404853616e-07,
+      "loss": 0.4572,
+      "step": 2052
+    },
+    {
+      "epoch": 0.392580552634095,
+      "grad_norm": 3.5944587340374685,
+      "learning_rate": 9.69251319903632e-07,
+      "loss": 0.2655,
+      "step": 2053
+    },
+    {
+      "epoch": 0.39277177550435033,
+      "grad_norm": 4.283718761711996,
+      "learning_rate": 9.693132048026607e-07,
+      "loss": 0.2868,
+      "step": 2054
+    },
+    {
+      "epoch": 0.3929629983746056,
+      "grad_norm": 5.353677017245294,
+      "learning_rate": 9.693750595800533e-07,
+      "loss": 0.725,
+      "step": 2055
+    },
+    {
+      "epoch": 0.3931542212448609,
+      "grad_norm": 6.201545879469571,
+      "learning_rate": 9.69436884265118e-07,
+      "loss": 0.3215,
+      "step": 2056
+    },
+    {
+      "epoch": 0.3933454441151162,
+      "grad_norm": 10.211335738456382,
+      "learning_rate": 9.694986788871208e-07,
+      "loss": 0.3406,
+      "step": 2057
+    },
+    {
+      "epoch": 0.39353666698537143,
+      "grad_norm": 2.821586680336672,
+      "learning_rate": 9.69560443475284e-07,
+      "loss": 0.1782,
+      "step": 2058
+    },
+    {
+      "epoch": 0.39372788985562673,
+      "grad_norm": 2.2599763104370303,
+      "learning_rate": 9.69622178058788e-07,
+      "loss": 0.1355,
+      "step": 2059
+    },
+    {
+      "epoch": 0.39391911272588204,
+      "grad_norm": 3.7155196320116866,
+      "learning_rate": 9.69683882666771e-07,
+      "loss": 0.1195,
+      "step": 2060
+    },
+    {
+      "epoch": 0.3941103355961373,
+      "grad_norm": 4.5456574897543875,
+      "learning_rate": 9.697455573283282e-07,
+      "loss": 0.1194,
+      "step": 2061
+    },
+    {
+      "epoch": 0.3943015584663926,
+      "grad_norm": 3.4865827878803,
+      "learning_rate": 9.698072020725127e-07,
+      "loss": 0.1215,
+      "step": 2062
+    },
+    {
+      "epoch": 0.3944927813366479,
+      "grad_norm": 7.571919021761864,
+      "learning_rate": 9.69868816928335e-07,
+      "loss": 0.2636,
+      "step": 2063
+    },
+    {
+      "epoch": 0.39468400420690314,
+      "grad_norm": 7.039268160407789,
+      "learning_rate": 9.69930401924764e-07,
+      "loss": 0.3308,
+      "step": 2064
+    },
+    {
+      "epoch": 0.39487522707715844,
+      "grad_norm": 7.241309910083063,
+      "learning_rate": 9.69991957090726e-07,
+      "loss": 0.7763,
+      "step": 2065
+    },
+    {
+      "epoch": 0.3950664499474137,
+      "grad_norm": 3.656222742745287,
+      "learning_rate": 9.700534824551058e-07,
+      "loss": 0.1901,
+      "step": 2066
+    },
+    {
+      "epoch": 0.395257672817669,
+      "grad_norm": 3.641547221327925,
+      "learning_rate": 9.701149780467458e-07,
+      "loss": 0.4165,
+      "step": 2067
+    },
+    {
+      "epoch": 0.3954488956879243,
+      "grad_norm": 4.057050835738276,
+      "learning_rate": 9.70176443894447e-07,
+      "loss": 0.1949,
+      "step": 2068
+    },
+    {
+      "epoch": 0.39564011855817954,
+      "grad_norm": 3.880565092681425,
+      "learning_rate": 9.702378800269676e-07,
+      "loss": 0.3698,
+      "step": 2069
+    },
+    {
+      "epoch": 0.39583134142843485,
+      "grad_norm": 2.7365189348956607,
+      "learning_rate": 9.702992864730256e-07,
+      "loss": 0.1466,
+      "step": 2070
+    },
+    {
+      "epoch": 0.39602256429869015,
+      "grad_norm": 4.890688764747551,
+      "learning_rate": 9.703606632612963e-07,
+      "loss": 0.3284,
+      "step": 2071
+    },
+    {
+      "epoch": 0.3962137871689454,
+      "grad_norm": 3.8257493724901006,
+      "learning_rate": 9.704220104204139e-07,
+      "loss": 0.2968,
+      "step": 2072
+    },
+    {
+      "epoch": 0.3964050100392007,
+      "grad_norm": 2.8164364711288163,
+      "learning_rate": 9.704833279789709e-07,
+      "loss": 0.097,
+      "step": 2073
+    },
+    {
+      "epoch": 0.39659623290945595,
+      "grad_norm": 3.8635283449954145,
+      "learning_rate": 9.705446159655191e-07,
+      "loss": 0.1046,
+      "step": 2074
+    },
+    {
+      "epoch": 0.39678745577971125,
+      "grad_norm": 6.317086847802219,
+      "learning_rate": 9.706058744085678e-07,
+      "loss": 0.1989,
+      "step": 2075
+    },
+    {
+      "epoch": 0.39697867864996655,
+      "grad_norm": 4.176783727744304,
+      "learning_rate": 9.706671033365866e-07,
+      "loss": 0.2049,
+      "step": 2076
+    },
+    {
+      "epoch": 0.3971699015202218,
+      "grad_norm": 4.214909819048901,
+      "learning_rate": 9.707283027780023e-07,
+      "loss": 0.5465,
+      "step": 2077
+    },
+    {
+      "epoch": 0.3973611243904771,
+      "grad_norm": 4.826325213822632,
+      "learning_rate": 9.707894727612025e-07,
+      "loss": 0.37,
+      "step": 2078
+    },
+    {
+      "epoch": 0.3975523472607324,
+      "grad_norm": 3.662433709595992,
+      "learning_rate": 9.70850613314532e-07,
+      "loss": 0.3764,
+      "step": 2079
+    },
+    {
+      "epoch": 0.39774357013098766,
+      "grad_norm": 3.4995239479018396,
+      "learning_rate": 9.70911724466296e-07,
+      "loss": 0.1486,
+      "step": 2080
+    },
+    {
+      "epoch": 0.39793479300124296,
+      "grad_norm": 3.3837363587165434,
+      "learning_rate": 9.709728062447581e-07,
+      "loss": 0.4188,
+      "step": 2081
+    },
+    {
+      "epoch": 0.3981260158714982,
+      "grad_norm": 2.473303253714102,
+      "learning_rate": 9.71033858678142e-07,
+      "loss": 0.1395,
+      "step": 2082
+    },
+    {
+      "epoch": 0.3983172387417535,
+      "grad_norm": 4.587291926932533,
+      "learning_rate": 9.710948817946298e-07,
+      "loss": 0.1711,
+      "step": 2083
+    },
+    {
+      "epoch": 0.3985084616120088,
+      "grad_norm": 1.7700691668687039,
+      "learning_rate": 9.71155875622364e-07,
+      "loss": 0.0581,
+      "step": 2084
+    },
+    {
+      "epoch": 0.39869968448226406,
+      "grad_norm": 5.541694048285658,
+      "learning_rate": 9.712168401894455e-07,
+      "loss": 0.2102,
+      "step": 2085
+    },
+    {
+      "epoch": 0.39889090735251936,
+      "grad_norm": 2.0044244464309937,
+      "learning_rate": 9.712777755239355e-07,
+      "loss": 0.0408,
+      "step": 2086
+    },
+    {
+      "epoch": 0.39908213022277467,
+      "grad_norm": 5.069532243267355,
+      "learning_rate": 9.71338681653855e-07,
+      "loss": 0.1429,
+      "step": 2087
+    },
+    {
+      "epoch": 0.3992733530930299,
+      "grad_norm": 8.918716005021276,
+      "learning_rate": 9.71399558607184e-07,
+      "loss": 0.4086,
+      "step": 2088
+    },
+    {
+      "epoch": 0.3994645759632852,
+      "grad_norm": 5.545253462919218,
+      "learning_rate": 9.71460406411863e-07,
+      "loss": 0.6002,
+      "step": 2089
+    },
+    {
+      "epoch": 0.39965579883354047,
+      "grad_norm": 6.286921886766427,
+      "learning_rate": 9.71521225095792e-07,
+      "loss": 0.5907,
+      "step": 2090
+    },
+    {
+      "epoch": 0.39984702170379577,
+      "grad_norm": 5.623466960960182,
+      "learning_rate": 9.715820146868312e-07,
+      "loss": 0.6547,
+      "step": 2091
+    },
+    {
+      "epoch": 0.40003824457405107,
+      "grad_norm": 1.4325171645702295,
+      "learning_rate": 9.716427752128005e-07,
+      "loss": 0.2525,
+      "step": 2092
+    },
+    {
+      "epoch": 0.4002294674443063,
+      "grad_norm": 4.877806613656336,
+      "learning_rate": 9.717035067014807e-07,
+      "loss": 0.4892,
+      "step": 2093
+    },
+    {
+      "epoch": 0.4004206903145616,
+      "grad_norm": 3.3225017196846007,
+      "learning_rate": 9.717642091806116e-07,
+      "loss": 0.2577,
+      "step": 2094
+    },
+    {
+      "epoch": 0.4006119131848169,
+      "grad_norm": 2.2021809171662423,
+      "learning_rate": 9.71824882677894e-07,
+      "loss": 0.2441,
+      "step": 2095
+    },
+    {
+      "epoch": 0.4008031360550722,
+      "grad_norm": 3.88679738947612,
+      "learning_rate": 9.71885527220989e-07,
+      "loss": 0.2253,
+      "step": 2096
+    },
+    {
+      "epoch": 0.4009943589253275,
+      "grad_norm": 3.488483692202697,
+      "learning_rate": 9.719461428375185e-07,
+      "loss": 0.0915,
+      "step": 2097
+    },
+    {
+      "epoch": 0.4011855817955827,
+      "grad_norm": 2.624159178720619,
+      "learning_rate": 9.720067295550638e-07,
+      "loss": 0.0846,
+      "step": 2098
+    },
+    {
+      "epoch": 0.401376804665838,
+      "grad_norm": 2.243111981396378,
+      "learning_rate": 9.720672874011675e-07,
+      "loss": 0.062,
+      "step": 2099
+    },
+    {
+      "epoch": 0.40156802753609333,
+      "grad_norm": 2.66987749354458,
+      "learning_rate": 9.721278164033332e-07,
+      "loss": 0.0791,
+      "step": 2100
+    },
+    {
+      "epoch": 0.4017592504063486,
+      "grad_norm": 6.19224659123647,
+      "learning_rate": 9.721883165890245e-07,
+      "loss": 0.2049,
+      "step": 2101
+    },
+    {
+      "epoch": 0.4019504732766039,
+      "grad_norm": 3.0564673036423704,
+      "learning_rate": 9.72248787985666e-07,
+      "loss": 0.2176,
+      "step": 2102
+    },
+    {
+      "epoch": 0.4021416961468592,
+      "grad_norm": 3.9149594069284994,
+      "learning_rate": 9.723092306206433e-07,
+      "loss": 0.4629,
+      "step": 2103
+    },
+    {
+      "epoch": 0.40233291901711443,
+      "grad_norm": 2.4193375615559254,
+      "learning_rate": 9.72369644521303e-07,
+      "loss": 0.2531,
+      "step": 2104
+    },
+    {
+      "epoch": 0.40252414188736974,
+      "grad_norm": 4.551747231838432,
+      "learning_rate": 9.724300297149523e-07,
+      "loss": 0.6005,
+      "step": 2105
+    },
+    {
+      "epoch": 0.40271536475762504,
+      "grad_norm": 4.74192726389022,
+      "learning_rate": 9.724903862288603e-07,
+      "loss": 0.4852,
+      "step": 2106
+    },
+    {
+      "epoch": 0.4029065876278803,
+      "grad_norm": 2.9842136199617864,
+      "learning_rate": 9.725507140902564e-07,
+      "loss": 0.3736,
+      "step": 2107
+    },
+    {
+      "epoch": 0.4030978104981356,
+      "grad_norm": 2.913852878425219,
+      "learning_rate": 9.726110133263316e-07,
+      "loss": 0.1555,
+      "step": 2108
+    },
+    {
+      "epoch": 0.40328903336839084,
+      "grad_norm": 2.2366514546899894,
+      "learning_rate": 9.72671283964238e-07,
+      "loss": 0.0538,
+      "step": 2109
+    },
+    {
+      "epoch": 0.40348025623864614,
+      "grad_norm": 2.883319427662505,
+      "learning_rate": 9.727315260310902e-07,
+      "loss": 0.0557,
+      "step": 2110
+    },
+    {
+      "epoch": 0.40367147910890144,
+      "grad_norm": 4.08922064260015,
+      "learning_rate": 9.727917395539621e-07,
+      "loss": 0.0936,
+      "step": 2111
+    },
+    {
+      "epoch": 0.4038627019791567,
+      "grad_norm": 2.9032753107847697,
+      "learning_rate": 9.728519245598914e-07,
+      "loss": 0.0762,
+      "step": 2112
+    },
+    {
+      "epoch": 0.404053924849412,
+      "grad_norm": 7.060599822588469,
+      "learning_rate": 9.729120810758759e-07,
+      "loss": 0.4572,
+      "step": 2113
+    },
+    {
+      "epoch": 0.4042451477196673,
+      "grad_norm": 5.918963277613596,
+      "learning_rate": 9.729722091288758e-07,
+      "loss": 0.3352,
+      "step": 2114
+    },
+    {
+      "epoch": 0.40443637058992254,
+      "grad_norm": 3.4962571430917544,
+      "learning_rate": 9.730323087458124e-07,
+      "loss": 0.2498,
+      "step": 2115
+    },
+    {
+      "epoch": 0.40462759346017785,
+      "grad_norm": 4.467585271823107,
+      "learning_rate": 9.730923799535698e-07,
+      "loss": 0.3258,
+      "step": 2116
+    },
+    {
+      "epoch": 0.4048188163304331,
+      "grad_norm": 4.432381700377552,
+      "learning_rate": 9.73152422778993e-07,
+      "loss": 0.3221,
+      "step": 2117
+    },
+    {
+      "epoch": 0.4050100392006884,
+      "grad_norm": 2.4579616416501966,
+      "learning_rate": 9.732124372488896e-07,
+      "loss": 0.2523,
+      "step": 2118
+    },
+    {
+      "epoch": 0.4052012620709437,
+      "grad_norm": 6.633735799161448,
+      "learning_rate": 9.732724233900291e-07,
+      "loss": 0.3533,
+      "step": 2119
+    },
+    {
+      "epoch": 0.40539248494119895,
+      "grad_norm": 2.8831241098101135,
+      "learning_rate": 9.733323812291426e-07,
+      "loss": 0.1838,
+      "step": 2120
+    },
+    {
+      "epoch": 0.40558370781145425,
+      "grad_norm": 4.036283440851569,
+      "learning_rate": 9.733923107929244e-07,
+      "loss": 0.1838,
+      "step": 2121
+    },
+    {
+      "epoch": 0.40577493068170956,
+      "grad_norm": 2.9502127683818236,
+      "learning_rate": 9.734522121080303e-07,
+      "loss": 0.1278,
+      "step": 2122
+    },
+    {
+      "epoch": 0.4059661535519648,
+      "grad_norm": 2.5086085877147517,
+      "learning_rate": 9.735120852010786e-07,
+      "loss": 0.1889,
+      "step": 2123
+    },
+    {
+      "epoch": 0.4061573764222201,
+      "grad_norm": 2.8914904149526732,
+      "learning_rate": 9.7357193009865e-07,
+      "loss": 0.1052,
+      "step": 2124
+    },
+    {
+      "epoch": 0.40634859929247535,
+      "grad_norm": 6.451395093871291,
+      "learning_rate": 9.736317468272874e-07,
+      "loss": 0.146,
+      "step": 2125
+    },
+    {
+      "epoch": 0.40653982216273066,
+      "grad_norm": 7.7773153016179855,
+      "learning_rate": 9.736915354134971e-07,
+      "loss": 0.2848,
+      "step": 2126
+    },
+    {
+      "epoch": 0.40673104503298596,
+      "grad_norm": 3.3039205035328183,
+      "learning_rate": 9.737512958837469e-07,
+      "loss": 0.1308,
+      "step": 2127
+    },
+    {
+      "epoch": 0.4069222679032412,
+      "grad_norm": 3.1892965900905152,
+      "learning_rate": 9.73811028264468e-07,
+      "loss": 0.2797,
+      "step": 2128
+    },
+    {
+      "epoch": 0.4071134907734965,
+      "grad_norm": 3.1575331393063593,
+      "learning_rate": 9.738707325820538e-07,
+      "loss": 0.1635,
+      "step": 2129
+    },
+    {
+      "epoch": 0.4073047136437518,
+      "grad_norm": 7.7413897521963175,
+      "learning_rate": 9.739304088628611e-07,
+      "loss": 0.2288,
+      "step": 2130
+    },
+    {
+      "epoch": 0.40749593651400706,
+      "grad_norm": 5.084861067202712,
+      "learning_rate": 9.739900571332096e-07,
+      "loss": 0.4711,
+      "step": 2131
+    },
+    {
+      "epoch": 0.40768715938426237,
+      "grad_norm": 1.827896299484216,
+      "learning_rate": 9.740496774193815e-07,
+      "loss": 0.1299,
+      "step": 2132
+    },
+    {
+      "epoch": 0.4078783822545176,
+      "grad_norm": 3.1162582197819226,
+      "learning_rate": 9.74109269747622e-07,
+      "loss": 0.2423,
+      "step": 2133
+    },
+    {
+      "epoch": 0.4080696051247729,
+      "grad_norm": 2.76065720763709,
+      "learning_rate": 9.741688341441398e-07,
+      "loss": 0.2549,
+      "step": 2134
+    },
+    {
+      "epoch": 0.4082608279950282,
+      "grad_norm": 2.380825327253472,
+      "learning_rate": 9.742283706351065e-07,
+      "loss": 0.1679,
+      "step": 2135
+    },
+    {
+      "epoch": 0.40845205086528347,
+      "grad_norm": 5.243127593611529,
+      "learning_rate": 9.742878792466572e-07,
+      "loss": 0.3907,
+      "step": 2136
+    },
+    {
+      "epoch": 0.40864327373553877,
+      "grad_norm": 4.693436589044413,
+      "learning_rate": 9.7434736000489e-07,
+      "loss": 0.1925,
+      "step": 2137
+    },
+    {
+      "epoch": 0.4088344966057941,
+      "grad_norm": 10.062844217229758,
+      "learning_rate": 9.744068129358662e-07,
+      "loss": 0.0838,
+      "step": 2138
+    },
+    {
+      "epoch": 0.4090257194760493,
+      "grad_norm": 7.24738659288697,
+      "learning_rate": 9.744662380656113e-07,
+      "loss": 0.283,
+      "step": 2139
+    },
+    {
+      "epoch": 0.4092169423463046,
+      "grad_norm": 3.8935434203969157,
+      "learning_rate": 9.745256354201136e-07,
+      "loss": 0.3179,
+      "step": 2140
+    },
+    {
+      "epoch": 0.4094081652165599,
+      "grad_norm": 4.633710751610403,
+      "learning_rate": 9.745850050253253e-07,
+      "loss": 0.5113,
+      "step": 2141
+    },
+    {
+      "epoch": 0.4095993880868152,
+      "grad_norm": 3.771021098876779,
+      "learning_rate": 9.74644346907162e-07,
+      "loss": 0.1853,
+      "step": 2142
+    },
+    {
+      "epoch": 0.4097906109570705,
+      "grad_norm": 2.673695529634361,
+      "learning_rate": 9.747036610915033e-07,
+      "loss": 0.0758,
+      "step": 2143
+    },
+    {
+      "epoch": 0.4099818338273257,
+      "grad_norm": 2.685795531812203,
+      "learning_rate": 9.747629476041923e-07,
+      "loss": 0.2452,
+      "step": 2144
+    },
+    {
+      "epoch": 0.41017305669758103,
+      "grad_norm": 3.556042222320521,
+      "learning_rate": 9.748222064710363e-07,
+      "loss": 0.2825,
+      "step": 2145
+    },
+    {
+      "epoch": 0.41036427956783633,
+      "grad_norm": 3.7216872468189433,
+      "learning_rate": 9.748814377178063e-07,
+      "loss": 0.352,
+      "step": 2146
+    },
+    {
+      "epoch": 0.4105555024380916,
+      "grad_norm": 5.515764704764002,
+      "learning_rate": 9.749406413702372e-07,
+      "loss": 0.1461,
+      "step": 2147
+    },
+    {
+      "epoch": 0.4107467253083469,
+      "grad_norm": 3.735233862524737,
+      "learning_rate": 9.749998174540277e-07,
+      "loss": 0.0347,
+      "step": 2148
+    },
+    {
+      "epoch": 0.4109379481786022,
+      "grad_norm": 4.979308611816402,
+      "learning_rate": 9.750589659948416e-07,
+      "loss": 0.0977,
+      "step": 2149
+    },
+    {
+      "epoch": 0.41112917104885743,
+      "grad_norm": 5.434838882486662,
+      "learning_rate": 9.751180870183057e-07,
+      "loss": 0.0846,
+      "step": 2150
+    },
+    {
+      "epoch": 0.41132039391911274,
+      "grad_norm": 5.504327805283102,
+      "learning_rate": 9.751771805500118e-07,
+      "loss": 0.2917,
+      "step": 2151
+    },
+    {
+      "epoch": 0.411511616789368,
+      "grad_norm": 5.576381891841888,
+      "learning_rate": 9.75236246615516e-07,
+      "loss": 0.3092,
+      "step": 2152
+    },
+    {
+      "epoch": 0.4117028396596233,
+      "grad_norm": 3.668850046382626,
+      "learning_rate": 9.752952852403381e-07,
+      "loss": 0.3553,
+      "step": 2153
+    },
+    {
+      "epoch": 0.4118940625298786,
+      "grad_norm": 2.6883014991926637,
+      "learning_rate": 9.753542964499632e-07,
+      "loss": 0.3198,
+      "step": 2154
+    },
+    {
+      "epoch": 0.41208528540013384,
+      "grad_norm": 3.3022678616505856,
+      "learning_rate": 9.754132802698407e-07,
+      "loss": 0.3467,
+      "step": 2155
+    },
+    {
+      "epoch": 0.41227650827038914,
+      "grad_norm": 1.8327601432639287,
+      "learning_rate": 9.754722367253839e-07,
+      "loss": 0.3043,
+      "step": 2156
+    },
+    {
+      "epoch": 0.41246773114064444,
+      "grad_norm": 4.96186303343545,
+      "learning_rate": 9.755311658419714e-07,
+      "loss": 0.5342,
+      "step": 2157
+    },
+    {
+      "epoch": 0.4126589540108997,
+      "grad_norm": 2.832816600428379,
+      "learning_rate": 9.755900676449469e-07,
+      "loss": 0.5614,
+      "step": 2158
+    },
+    {
+      "epoch": 0.412850176881155,
+      "grad_norm": 4.1826362262072525,
+      "learning_rate": 9.756489421596178e-07,
+      "loss": 0.2067,
+      "step": 2159
+    },
+    {
+      "epoch": 0.41304139975141024,
+      "grad_norm": 4.010239845770508,
+      "learning_rate": 9.75707789411257e-07,
+      "loss": 0.1295,
+      "step": 2160
+    },
+    {
+      "epoch": 0.41323262262166555,
+      "grad_norm": 2.626727580204908,
+      "learning_rate": 9.757666094251023e-07,
+      "loss": 0.0568,
+      "step": 2161
+    },
+    {
+      "epoch": 0.41342384549192085,
+      "grad_norm": 2.749571073200031,
+      "learning_rate": 9.758254022263565e-07,
+      "loss": 0.1323,
+      "step": 2162
+    },
+    {
+      "epoch": 0.4136150683621761,
+      "grad_norm": 5.983985828464475,
+      "learning_rate": 9.758841678401873e-07,
+      "loss": 0.1107,
+      "step": 2163
+    },
+    {
+      "epoch": 0.4138062912324314,
+      "grad_norm": 5.260137488885394,
+      "learning_rate": 9.75942906291727e-07,
+      "loss": 0.5862,
+      "step": 2164
+    },
+    {
+      "epoch": 0.4139975141026867,
+      "grad_norm": 4.425255659228756,
+      "learning_rate": 9.760016176060744e-07,
+      "loss": 0.2134,
+      "step": 2165
+    },
+    {
+      "epoch": 0.41418873697294195,
+      "grad_norm": 4.605346574529783,
+      "learning_rate": 9.760603018082921e-07,
+      "loss": 0.3951,
+      "step": 2166
+    },
+    {
+      "epoch": 0.41437995984319725,
+      "grad_norm": 5.110616643916769,
+      "learning_rate": 9.761189589234089e-07,
+      "loss": 0.4961,
+      "step": 2167
+    },
+    {
+      "epoch": 0.4145711827134525,
+      "grad_norm": 4.138448351827311,
+      "learning_rate": 9.761775889764186e-07,
+      "loss": 0.2885,
+      "step": 2168
+    },
+    {
+      "epoch": 0.4147624055837078,
+      "grad_norm": 3.262013459739772,
+      "learning_rate": 9.762361919922804e-07,
+      "loss": 0.3333,
+      "step": 2169
+    },
+    {
+      "epoch": 0.4149536284539631,
+      "grad_norm": 10.157836790104207,
+      "learning_rate": 9.76294767995919e-07,
+      "loss": 0.1366,
+      "step": 2170
+    },
+    {
+      "epoch": 0.41514485132421836,
+      "grad_norm": 5.287461820423993,
+      "learning_rate": 9.763533170122248e-07,
+      "loss": 0.2115,
+      "step": 2171
+    },
+    {
+      "epoch": 0.41533607419447366,
+      "grad_norm": 3.701058370414909,
+      "learning_rate": 9.764118390660541e-07,
+      "loss": 0.1737,
+      "step": 2172
+    },
+    {
+      "epoch": 0.41552729706472896,
+      "grad_norm": 1.7226982544083502,
+      "learning_rate": 9.764703341822282e-07,
+      "loss": 0.0983,
+      "step": 2173
+    },
+    {
+      "epoch": 0.4157185199349842,
+      "grad_norm": 5.980890360097877,
+      "learning_rate": 9.765288023855342e-07,
+      "loss": 0.2478,
+      "step": 2174
+    },
+    {
+      "epoch": 0.4159097428052395,
+      "grad_norm": 4.607673977709717,
+      "learning_rate": 9.765872437007257e-07,
+      "loss": 0.0822,
+      "step": 2175
+    },
+    {
+      "epoch": 0.4161009656754948,
+      "grad_norm": 4.596228521284929,
+      "learning_rate": 9.766456581525215e-07,
+      "loss": 0.4169,
+      "step": 2176
+    },
+    {
+      "epoch": 0.41629218854575006,
+      "grad_norm": 5.586889908699536,
+      "learning_rate": 9.767040457656067e-07,
+      "loss": 0.2865,
+      "step": 2177
+    },
+    {
+      "epoch": 0.41648341141600537,
+      "grad_norm": 3.7462848061697054,
+      "learning_rate": 9.767624065646319e-07,
+      "loss": 0.3586,
+      "step": 2178
+    },
+    {
+      "epoch": 0.4166746342862606,
+      "grad_norm": 2.7718887517990063,
+      "learning_rate": 9.768207405742146e-07,
+      "loss": 0.3307,
+      "step": 2179
+    },
+    {
+      "epoch": 0.4168658571565159,
+      "grad_norm": 3.8499816993179343,
+      "learning_rate": 9.768790478189373e-07,
+      "loss": 0.7148,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4170570800267712,
+      "grad_norm": 3.9411498819078927,
+      "learning_rate": 9.7693732832335e-07,
+      "loss": 0.2921,
+      "step": 2181
+    },
+    {
+      "epoch": 0.41724830289702647,
+      "grad_norm": 3.489959347332389,
+      "learning_rate": 9.769955821119676e-07,
+      "loss": 0.2221,
+      "step": 2182
+    },
+    {
+      "epoch": 0.41743952576728177,
+      "grad_norm": 3.5631927352111785,
+      "learning_rate": 9.770538092092722e-07,
+      "loss": 0.3861,
+      "step": 2183
+    },
+    {
+      "epoch": 0.4176307486375371,
+      "grad_norm": 1.9404014888686156,
+      "learning_rate": 9.771120096397123e-07,
+      "loss": 0.1292,
+      "step": 2184
+    },
+    {
+      "epoch": 0.4178219715077923,
+      "grad_norm": 2.9589866310841235,
+      "learning_rate": 9.771701834277018e-07,
+      "loss": 0.0676,
+      "step": 2185
+    },
+    {
+      "epoch": 0.4180131943780476,
+      "grad_norm": 4.6236130980483585,
+      "learning_rate": 9.772283305976224e-07,
+      "loss": 0.3231,
+      "step": 2186
+    },
+    {
+      "epoch": 0.4182044172483029,
+      "grad_norm": 4.527856858595863,
+      "learning_rate": 9.772864511738216e-07,
+      "loss": 0.2179,
+      "step": 2187
+    },
+    {
+      "epoch": 0.4183956401185582,
+      "grad_norm": 5.94004975834309,
+      "learning_rate": 9.773445451806133e-07,
+      "loss": 0.1616,
+      "step": 2188
+    },
+    {
+      "epoch": 0.4185868629888135,
+      "grad_norm": 5.107649298832417,
+      "learning_rate": 9.77402612642279e-07,
+      "loss": 0.3761,
+      "step": 2189
+    },
+    {
+      "epoch": 0.4187780858590687,
+      "grad_norm": 4.315938711952712,
+      "learning_rate": 9.77460653583066e-07,
+      "loss": 0.6074,
+      "step": 2190
+    },
+    {
+      "epoch": 0.41896930872932403,
+      "grad_norm": 7.8929978834647425,
+      "learning_rate": 9.77518668027189e-07,
+      "loss": 0.351,
+      "step": 2191
+    },
+    {
+      "epoch": 0.41916053159957933,
+      "grad_norm": 2.9193687093606124,
+      "learning_rate": 9.77576655998829e-07,
+      "loss": 0.2492,
+      "step": 2192
+    },
+    {
+      "epoch": 0.4193517544698346,
+      "grad_norm": 3.8029065360120367,
+      "learning_rate": 9.776346175221345e-07,
+      "loss": 0.343,
+      "step": 2193
+    },
+    {
+      "epoch": 0.4195429773400899,
+      "grad_norm": 5.844146225496114,
+      "learning_rate": 9.776925526212206e-07,
+      "loss": 0.2844,
+      "step": 2194
+    },
+    {
+      "epoch": 0.41973420021034513,
+      "grad_norm": 7.128205766099044,
+      "learning_rate": 9.777504613201693e-07,
+      "loss": 0.6386,
+      "step": 2195
+    },
+    {
+      "epoch": 0.41992542308060044,
+      "grad_norm": 3.3504004964986613,
+      "learning_rate": 9.778083436430304e-07,
+      "loss": 0.2445,
+      "step": 2196
+    },
+    {
+      "epoch": 0.42011664595085574,
+      "grad_norm": 1.7785569965151258,
+      "learning_rate": 9.778661996138199e-07,
+      "loss": 0.1015,
+      "step": 2197
+    },
+    {
+      "epoch": 0.420307868821111,
+      "grad_norm": 4.098162420700335,
+      "learning_rate": 9.779240292565216e-07,
+      "loss": 0.3958,
+      "step": 2198
+    },
+    {
+      "epoch": 0.4204990916913663,
+      "grad_norm": 2.408864372163066,
+      "learning_rate": 9.779818325950868e-07,
+      "loss": 0.0292,
+      "step": 2199
+    },
+    {
+      "epoch": 0.4206903145616216,
+      "grad_norm": 13.552154955233654,
+      "learning_rate": 9.780396096534332e-07,
+      "loss": 0.2798,
+      "step": 2200
+    },
+    {
+      "epoch": 0.42088153743187684,
+      "grad_norm": 5.380834384867105,
+      "learning_rate": 9.780973604554468e-07,
+      "loss": 0.4751,
+      "step": 2201
+    },
+    {
+      "epoch": 0.42107276030213214,
+      "grad_norm": 4.990455482197973,
+      "learning_rate": 9.78155085024981e-07,
+      "loss": 0.3426,
+      "step": 2202
+    },
+    {
+      "epoch": 0.4212639831723874,
+      "grad_norm": 5.137030646424517,
+      "learning_rate": 9.78212783385856e-07,
+      "loss": 0.3707,
+      "step": 2203
+    },
+    {
+      "epoch": 0.4214552060426427,
+      "grad_norm": 3.6982849940584397,
+      "learning_rate": 9.782704555618601e-07,
+      "loss": 0.3603,
+      "step": 2204
+    },
+    {
+      "epoch": 0.421646428912898,
+      "grad_norm": 4.748340115893255,
+      "learning_rate": 9.783281015767496e-07,
+      "loss": 0.3051,
+      "step": 2205
+    },
+    {
+      "epoch": 0.42183765178315324,
+      "grad_norm": 3.150853746763898,
+      "learning_rate": 9.78385721454247e-07,
+      "loss": 0.2707,
+      "step": 2206
+    },
+    {
+      "epoch": 0.42202887465340855,
+      "grad_norm": 4.934300991902148,
+      "learning_rate": 9.784433152180448e-07,
+      "loss": 0.142,
+      "step": 2207
+    },
+    {
+      "epoch": 0.42222009752366385,
+      "grad_norm": 10.932690069612471,
+      "learning_rate": 9.785008828918012e-07,
+      "loss": 0.2989,
+      "step": 2208
+    },
+    {
+      "epoch": 0.4224113203939191,
+      "grad_norm": 4.41373106125685,
+      "learning_rate": 9.785584244991436e-07,
+      "loss": 0.2698,
+      "step": 2209
+    },
+    {
+      "epoch": 0.4226025432641744,
+      "grad_norm": 4.999789805762503,
+      "learning_rate": 9.786159400636664e-07,
+      "loss": 0.3007,
+      "step": 2210
+    },
+    {
+      "epoch": 0.4227937661344297,
+      "grad_norm": 5.944804306466908,
+      "learning_rate": 9.78673429608933e-07,
+      "loss": 0.1416,
+      "step": 2211
+    },
+    {
+      "epoch": 0.42298498900468495,
+      "grad_norm": 4.085978812306086,
+      "learning_rate": 9.787308931584737e-07,
+      "loss": 0.1606,
+      "step": 2212
+    },
+    {
+      "epoch": 0.42317621187494026,
+      "grad_norm": 8.584754258350644,
+      "learning_rate": 9.78788330735788e-07,
+      "loss": 0.2281,
+      "step": 2213
+    },
+    {
+      "epoch": 0.4233674347451955,
+      "grad_norm": 5.590887434777203,
+      "learning_rate": 9.788457423643425e-07,
+      "loss": 0.5461,
+      "step": 2214
+    },
+    {
+      "epoch": 0.4235586576154508,
+      "grad_norm": 3.521868100366883,
+      "learning_rate": 9.789031280675729e-07,
+      "loss": 0.367,
+      "step": 2215
+    },
+    {
+      "epoch": 0.4237498804857061,
+      "grad_norm": 3.5821944504295793,
+      "learning_rate": 9.789604878688826e-07,
+      "loss": 0.156,
+      "step": 2216
+    },
+    {
+      "epoch": 0.42394110335596136,
+      "grad_norm": 1.0852622258486955,
+      "learning_rate": 9.790178217916436e-07,
+      "loss": 0.2999,
+      "step": 2217
+    },
+    {
+      "epoch": 0.42413232622621666,
+      "grad_norm": 4.460015356952589,
+      "learning_rate": 9.790751298591963e-07,
+      "loss": 0.4468,
+      "step": 2218
+    },
+    {
+      "epoch": 0.42432354909647196,
+      "grad_norm": 1.8736909747354102,
+      "learning_rate": 9.79132412094849e-07,
+      "loss": 0.1363,
+      "step": 2219
+    },
+    {
+      "epoch": 0.4245147719667272,
+      "grad_norm": 5.943012449310219,
+      "learning_rate": 9.791896685218795e-07,
+      "loss": 0.2645,
+      "step": 2220
+    },
+    {
+      "epoch": 0.4247059948369825,
+      "grad_norm": 3.4292098909777153,
+      "learning_rate": 9.792468991635331e-07,
+      "loss": 0.082,
+      "step": 2221
+    },
+    {
+      "epoch": 0.42489721770723776,
+      "grad_norm": 3.1493522886383576,
+      "learning_rate": 9.793041040430246e-07,
+      "loss": 0.197,
+      "step": 2222
+    },
+    {
+      "epoch": 0.42508844057749307,
+      "grad_norm": 5.267082759661095,
+      "learning_rate": 9.793612831835365e-07,
+      "loss": 0.2161,
+      "step": 2223
+    },
+    {
+      "epoch": 0.42527966344774837,
+      "grad_norm": 3.2973843583312186,
+      "learning_rate": 9.794184366082212e-07,
+      "loss": 0.13,
+      "step": 2224
+    },
+    {
+      "epoch": 0.4254708863180036,
+      "grad_norm": 3.913639129462084,
+      "learning_rate": 9.79475564340199e-07,
+      "loss": 0.0747,
+      "step": 2225
+    },
+    {
+      "epoch": 0.4256621091882589,
+      "grad_norm": 6.568364892299942,
+      "learning_rate": 9.79532666402559e-07,
+      "loss": 0.6796,
+      "step": 2226
+    },
+    {
+      "epoch": 0.4258533320585142,
+      "grad_norm": 4.774646861813897,
+      "learning_rate": 9.795897428183598e-07,
+      "loss": 0.83,
+      "step": 2227
+    },
+    {
+      "epoch": 0.42604455492876947,
+      "grad_norm": 3.7098881894354703,
+      "learning_rate": 9.796467936106283e-07,
+      "loss": 0.2557,
+      "step": 2228
+    },
+    {
+      "epoch": 0.4262357777990248,
+      "grad_norm": 5.286881553608861,
+      "learning_rate": 9.797038188023613e-07,
+      "loss": 0.2727,
+      "step": 2229
+    },
+    {
+      "epoch": 0.42642700066928,
+      "grad_norm": 5.020097966245105,
+      "learning_rate": 9.797608184165235e-07,
+      "loss": 0.2163,
+      "step": 2230
+    },
+    {
+      "epoch": 0.4266182235395353,
+      "grad_norm": 3.392348313001068,
+      "learning_rate": 9.798177924760495e-07,
+      "loss": 0.2915,
+      "step": 2231
+    },
+    {
+      "epoch": 0.4268094464097906,
+      "grad_norm": 2.8213007244372164,
+      "learning_rate": 9.798747410038429e-07,
+      "loss": 0.1399,
+      "step": 2232
+    },
+    {
+      "epoch": 0.4270006692800459,
+      "grad_norm": 2.368000457460772,
+      "learning_rate": 9.799316640227763e-07,
+      "loss": 0.1319,
+      "step": 2233
+    },
+    {
+      "epoch": 0.4271918921503012,
+      "grad_norm": 2.466207812992992,
+      "learning_rate": 9.79988561555692e-07,
+      "loss": 0.1368,
+      "step": 2234
+    },
+    {
+      "epoch": 0.4273831150205565,
+      "grad_norm": 3.668278397722594,
+      "learning_rate": 9.80045433625401e-07,
+      "loss": 0.1053,
+      "step": 2235
+    },
+    {
+      "epoch": 0.42757433789081173,
+      "grad_norm": 2.3920999601905137,
+      "learning_rate": 9.801022802546848e-07,
+      "loss": 0.1394,
+      "step": 2236
+    },
+    {
+      "epoch": 0.42776556076106703,
+      "grad_norm": 3.921124633497188,
+      "learning_rate": 9.801591014662928e-07,
+      "loss": 0.0912,
+      "step": 2237
+    },
+    {
+      "epoch": 0.4279567836313223,
+      "grad_norm": 2.688207754933629,
+      "learning_rate": 9.80215897282945e-07,
+      "loss": 0.1099,
+      "step": 2238
+    },
+    {
+      "epoch": 0.4281480065015776,
+      "grad_norm": 5.38806445357776,
+      "learning_rate": 9.802726677273307e-07,
+      "loss": 0.2604,
+      "step": 2239
+    },
+    {
+      "epoch": 0.4283392293718329,
+      "grad_norm": 4.472069634361164,
+      "learning_rate": 9.80329412822109e-07,
+      "loss": 0.7171,
+      "step": 2240
+    },
+    {
+      "epoch": 0.42853045224208813,
+      "grad_norm": 4.553390396751042,
+      "learning_rate": 9.803861325899081e-07,
+      "loss": 0.378,
+      "step": 2241
+    },
+    {
+      "epoch": 0.42872167511234344,
+      "grad_norm": 2.6194440081386894,
+      "learning_rate": 9.804428270533262e-07,
+      "loss": 0.1007,
+      "step": 2242
+    },
+    {
+      "epoch": 0.42891289798259874,
+      "grad_norm": 4.254872390142445,
+      "learning_rate": 9.804994962349316e-07,
+      "loss": 0.2771,
+      "step": 2243
+    },
+    {
+      "epoch": 0.429104120852854,
+      "grad_norm": 5.429586154348701,
+      "learning_rate": 9.80556140157262e-07,
+      "loss": 0.4871,
+      "step": 2244
+    },
+    {
+      "epoch": 0.4292953437231093,
+      "grad_norm": 3.2373257505442234,
+      "learning_rate": 9.806127588428252e-07,
+      "loss": 0.3666,
+      "step": 2245
+    },
+    {
+      "epoch": 0.4294865665933646,
+      "grad_norm": 2.563866041840987,
+      "learning_rate": 9.806693523140984e-07,
+      "loss": 0.0625,
+      "step": 2246
+    },
+    {
+      "epoch": 0.42967778946361984,
+      "grad_norm": 3.1167671092230753,
+      "learning_rate": 9.8072592059353e-07,
+      "loss": 0.1501,
+      "step": 2247
+    },
+    {
+      "epoch": 0.42986901233387514,
+      "grad_norm": 2.310852598084097,
+      "learning_rate": 9.807824637035367e-07,
+      "loss": 0.145,
+      "step": 2248
+    },
+    {
+      "epoch": 0.4300602352041304,
+      "grad_norm": 3.4443818995905566,
+      "learning_rate": 9.808389816665072e-07,
+      "loss": 0.0982,
+      "step": 2249
+    },
+    {
+      "epoch": 0.4302514580743857,
+      "grad_norm": 5.257780213844965,
+      "learning_rate": 9.80895474504799e-07,
+      "loss": 0.0823,
+      "step": 2250
+    },
+    {
+      "epoch": 0.430442680944641,
+      "grad_norm": 5.634695725032213,
+      "learning_rate": 9.809519422407398e-07,
+      "loss": 0.4395,
+      "step": 2251
+    },
+    {
+      "epoch": 0.43063390381489625,
+      "grad_norm": 6.7699871140868755,
+      "learning_rate": 9.810083848966286e-07,
+      "loss": 0.5543,
+      "step": 2252
+    },
+    {
+      "epoch": 0.43082512668515155,
+      "grad_norm": 4.061152073180592,
+      "learning_rate": 9.810648024947333e-07,
+      "loss": 0.2436,
+      "step": 2253
+    },
+    {
+      "epoch": 0.43101634955540685,
+      "grad_norm": 1.927611435020395,
+      "learning_rate": 9.811211950572937e-07,
+      "loss": 0.1295,
+      "step": 2254
+    },
+    {
+      "epoch": 0.4312075724256621,
+      "grad_norm": 4.335945638013197,
+      "learning_rate": 9.811775626065185e-07,
+      "loss": 0.3094,
+      "step": 2255
+    },
+    {
+      "epoch": 0.4313987952959174,
+      "grad_norm": 4.358356753420708,
+      "learning_rate": 9.812339051645881e-07,
+      "loss": 0.4715,
+      "step": 2256
+    },
+    {
+      "epoch": 0.43159001816617265,
+      "grad_norm": 3.4555601621503746,
+      "learning_rate": 9.812902227536526e-07,
+      "loss": 0.1721,
+      "step": 2257
+    },
+    {
+      "epoch": 0.43178124103642795,
+      "grad_norm": 2.6592380546695304,
+      "learning_rate": 9.81346515395833e-07,
+      "loss": 0.2057,
+      "step": 2258
+    },
+    {
+      "epoch": 0.43197246390668326,
+      "grad_norm": 3.382361683748854,
+      "learning_rate": 9.814027831132209e-07,
+      "loss": 0.3843,
+      "step": 2259
+    },
+    {
+      "epoch": 0.4321636867769385,
+      "grad_norm": 2.0336498462327577,
+      "learning_rate": 9.814590259278782e-07,
+      "loss": 0.0595,
+      "step": 2260
+    },
+    {
+      "epoch": 0.4323549096471938,
+      "grad_norm": 3.480385313735804,
+      "learning_rate": 9.815152438618382e-07,
+      "loss": 0.12,
+      "step": 2261
+    },
+    {
+      "epoch": 0.4325461325174491,
+      "grad_norm": 2.7471590105770236,
+      "learning_rate": 9.815714369371047e-07,
+      "loss": 0.0538,
+      "step": 2262
+    },
+    {
+      "epoch": 0.43273735538770436,
+      "grad_norm": 7.036156057495434,
+      "learning_rate": 9.81627605175652e-07,
+      "loss": 0.3034,
+      "step": 2263
+    },
+    {
+      "epoch": 0.43292857825795966,
+      "grad_norm": 3.963071831197789,
+      "learning_rate": 9.816837485994253e-07,
+      "loss": 0.3245,
+      "step": 2264
+    },
+    {
+      "epoch": 0.4331198011282149,
+      "grad_norm": 6.834764966404923,
+      "learning_rate": 9.817398672303412e-07,
+      "loss": 0.7238,
+      "step": 2265
+    },
+    {
+      "epoch": 0.4333110239984702,
+      "grad_norm": 3.6792563410566497,
+      "learning_rate": 9.817959610902873e-07,
+      "loss": 0.555,
+      "step": 2266
+    },
+    {
+      "epoch": 0.4335022468687255,
+      "grad_norm": 2.051653925187031,
+      "learning_rate": 9.818520302011213e-07,
+      "loss": 0.211,
+      "step": 2267
+    },
+    {
+      "epoch": 0.43369346973898076,
+      "grad_norm": 3.1509035358681152,
+      "learning_rate": 9.819080745846733e-07,
+      "loss": 0.2647,
+      "step": 2268
+    },
+    {
+      "epoch": 0.43388469260923607,
+      "grad_norm": 5.187049754262845,
+      "learning_rate": 9.819640942627433e-07,
+      "loss": 0.7708,
+      "step": 2269
+    },
+    {
+      "epoch": 0.43407591547949137,
+      "grad_norm": 3.8197168303184936,
+      "learning_rate": 9.820200892571035e-07,
+      "loss": 0.3034,
+      "step": 2270
+    },
+    {
+      "epoch": 0.4342671383497466,
+      "grad_norm": 2.9745066193574146,
+      "learning_rate": 9.820760595894966e-07,
+      "loss": 0.3584,
+      "step": 2271
+    },
+    {
+      "epoch": 0.4344583612200019,
+      "grad_norm": 3.0986251181751303,
+      "learning_rate": 9.821320052816368e-07,
+      "loss": 0.1425,
+      "step": 2272
+    },
+    {
+      "epoch": 0.43464958409025717,
+      "grad_norm": 3.0866296692094033,
+      "learning_rate": 9.8218792635521e-07,
+      "loss": 0.0877,
+      "step": 2273
+    },
+    {
+      "epoch": 0.43484080696051247,
+      "grad_norm": 4.218117221424359,
+      "learning_rate": 9.82243822831873e-07,
+      "loss": 0.1122,
+      "step": 2274
+    },
+    {
+      "epoch": 0.4350320298307678,
+      "grad_norm": 6.377620083804742,
+      "learning_rate": 9.822996947332538e-07,
+      "loss": 0.3048,
+      "step": 2275
+    },
+    {
+      "epoch": 0.435223252701023,
+      "grad_norm": 6.11768324439968,
+      "learning_rate": 9.82355542080953e-07,
+      "loss": 0.3476,
+      "step": 2276
+    },
+    {
+      "epoch": 0.4354144755712783,
+      "grad_norm": 4.03288744413788,
+      "learning_rate": 9.824113648965416e-07,
+      "loss": 0.2881,
+      "step": 2277
+    },
+    {
+      "epoch": 0.43560569844153363,
+      "grad_norm": 3.6624568845535066,
+      "learning_rate": 9.82467163201563e-07,
+      "loss": 0.4882,
+      "step": 2278
+    },
+    {
+      "epoch": 0.4357969213117889,
+      "grad_norm": 2.598271484943915,
+      "learning_rate": 9.825229370175315e-07,
+      "loss": 0.276,
+      "step": 2279
+    },
+    {
+      "epoch": 0.4359881441820442,
+      "grad_norm": 1.393013649311547,
+      "learning_rate": 9.825786863659334e-07,
+      "loss": 0.378,
+      "step": 2280
+    },
+    {
+      "epoch": 0.4361793670522995,
+      "grad_norm": 2.497782009420208,
+      "learning_rate": 9.826344112682269e-07,
+      "loss": 0.4941,
+      "step": 2281
+    },
+    {
+      "epoch": 0.43637058992255473,
+      "grad_norm": 4.264938131511991,
+      "learning_rate": 9.826901117458419e-07,
+      "loss": 0.5331,
+      "step": 2282
+    },
+    {
+      "epoch": 0.43656181279281003,
+      "grad_norm": 2.8924988317858182,
+      "learning_rate": 9.827457878201802e-07,
+      "loss": 0.2659,
+      "step": 2283
+    },
+    {
+      "epoch": 0.4367530356630653,
+      "grad_norm": 3.6178072898914877,
+      "learning_rate": 9.82801439512615e-07,
+      "loss": 0.0906,
+      "step": 2284
+    },
+    {
+      "epoch": 0.4369442585333206,
+      "grad_norm": 3.413571250367487,
+      "learning_rate": 9.828570668444921e-07,
+      "loss": 0.0562,
+      "step": 2285
+    },
+    {
+      "epoch": 0.4371354814035759,
+      "grad_norm": 7.448195260623936,
+      "learning_rate": 9.82912669837129e-07,
+      "loss": 0.1994,
+      "step": 2286
+    },
+    {
+      "epoch": 0.43732670427383113,
+      "grad_norm": 3.887130945479543,
+      "learning_rate": 9.829682485118152e-07,
+      "loss": 0.111,
+      "step": 2287
+    },
+    {
+      "epoch": 0.43751792714408644,
+      "grad_norm": 7.041410302305224,
+      "learning_rate": 9.83023802889812e-07,
+      "loss": 0.2429,
+      "step": 2288
+    },
+    {
+      "epoch": 0.43770915001434174,
+      "grad_norm": 5.026662500280244,
+      "learning_rate": 9.830793329923537e-07,
+      "loss": 0.5903,
+      "step": 2289
+    },
+    {
+      "epoch": 0.437900372884597,
+      "grad_norm": 2.370191524991772,
+      "learning_rate": 9.831348388406456e-07,
+      "loss": 0.2714,
+      "step": 2290
+    },
+    {
+      "epoch": 0.4380915957548523,
+      "grad_norm": 4.218338331694087,
+      "learning_rate": 9.831903204558661e-07,
+      "loss": 0.4121,
+      "step": 2291
+    },
+    {
+      "epoch": 0.43828281862510754,
+      "grad_norm": 3.37661986860537,
+      "learning_rate": 9.832457778591657e-07,
+      "loss": 0.4274,
+      "step": 2292
+    },
+    {
+      "epoch": 0.43847404149536284,
+      "grad_norm": 2.7526103548661864,
+      "learning_rate": 9.83301211071667e-07,
+      "loss": 0.2287,
+      "step": 2293
+    },
+    {
+      "epoch": 0.43866526436561815,
+      "grad_norm": 2.9561977648001454,
+      "learning_rate": 9.833566201144652e-07,
+      "loss": 0.1329,
+      "step": 2294
+    },
+    {
+      "epoch": 0.4388564872358734,
+      "grad_norm": 2.737137886379039,
+      "learning_rate": 9.834120050086277e-07,
+      "loss": 0.0963,
+      "step": 2295
+    },
+    {
+      "epoch": 0.4390477101061287,
+      "grad_norm": 3.4437886369202815,
+      "learning_rate": 9.834673657751943e-07,
+      "loss": 0.1179,
+      "step": 2296
+    },
+    {
+      "epoch": 0.439238932976384,
+      "grad_norm": 3.8484698834375184,
+      "learning_rate": 9.83522702435178e-07,
+      "loss": 0.2757,
+      "step": 2297
+    },
+    {
+      "epoch": 0.43943015584663925,
+      "grad_norm": 3.7796406749712297,
+      "learning_rate": 9.835780150095632e-07,
+      "loss": 0.0495,
+      "step": 2298
+    },
+    {
+      "epoch": 0.43962137871689455,
+      "grad_norm": 3.035172588839549,
+      "learning_rate": 9.836333035193083e-07,
+      "loss": 0.1085,
+      "step": 2299
+    },
+    {
+      "epoch": 0.4398126015871498,
+      "grad_norm": 3.0530998611586604,
+      "learning_rate": 9.83688567985343e-07,
+      "loss": 0.0533,
+      "step": 2300
+    },
+    {
+      "epoch": 0.4400038244574051,
+      "grad_norm": 6.851132665393958,
+      "learning_rate": 9.837438084285706e-07,
+      "loss": 0.4296,
+      "step": 2301
+    },
+    {
+      "epoch": 0.4401950473276604,
+      "grad_norm": 4.410722559341428,
+      "learning_rate": 9.83799024869867e-07,
+      "loss": 0.1892,
+      "step": 2302
+    },
+    {
+      "epoch": 0.44038627019791565,
+      "grad_norm": 3.2658067862549562,
+      "learning_rate": 9.838542173300805e-07,
+      "loss": 0.4602,
+      "step": 2303
+    },
+    {
+      "epoch": 0.44057749306817096,
+      "grad_norm": 5.011441301142309,
+      "learning_rate": 9.839093858300329e-07,
+      "loss": 0.5328,
+      "step": 2304
+    },
+    {
+      "epoch": 0.44076871593842626,
+      "grad_norm": 4.9009843499307975,
+      "learning_rate": 9.83964530390518e-07,
+      "loss": 0.4167,
+      "step": 2305
+    },
+    {
+      "epoch": 0.4409599388086815,
+      "grad_norm": 4.217878689681478,
+      "learning_rate": 9.840196510323036e-07,
+      "loss": 0.3591,
+      "step": 2306
+    },
+    {
+      "epoch": 0.4411511616789368,
+      "grad_norm": 3.368257073909648,
+      "learning_rate": 9.840747477761297e-07,
+      "loss": 0.3303,
+      "step": 2307
+    },
+    {
+      "epoch": 0.44134238454919206,
+      "grad_norm": 2.611960038596513,
+      "learning_rate": 9.841298206427097e-07,
+      "loss": 0.3159,
+      "step": 2308
+    },
+    {
+      "epoch": 0.44153360741944736,
+      "grad_norm": 3.354585435840038,
+      "learning_rate": 9.841848696527298e-07,
+      "loss": 0.1088,
+      "step": 2309
+    },
+    {
+      "epoch": 0.44172483028970266,
+      "grad_norm": 4.061294376768386,
+      "learning_rate": 9.842398948268493e-07,
+      "loss": 0.0684,
+      "step": 2310
+    },
+    {
+      "epoch": 0.4419160531599579,
+      "grad_norm": 4.672229153787614,
+      "learning_rate": 9.842948961857016e-07,
+      "loss": 0.0662,
+      "step": 2311
+    },
+    {
+      "epoch": 0.4421072760302132,
+      "grad_norm": 7.292156907812946,
+      "learning_rate": 9.84349873749892e-07,
+      "loss": 0.173,
+      "step": 2312
+    },
+    {
+      "epoch": 0.4422984989004685,
+      "grad_norm": 5.325245188051552,
+      "learning_rate": 9.8440482754e-07,
+      "loss": 0.3724,
+      "step": 2313
+    },
+    {
+      "epoch": 0.44248972177072377,
+      "grad_norm": 4.924334586643745,
+      "learning_rate": 9.844597575765777e-07,
+      "loss": 0.5379,
+      "step": 2314
+    },
+    {
+      "epoch": 0.44268094464097907,
+      "grad_norm": 4.489899638186396,
+      "learning_rate": 9.845146638801513e-07,
+      "loss": 0.3387,
+      "step": 2315
+    },
+    {
+      "epoch": 0.4428721675112343,
+      "grad_norm": 7.094489911852366,
+      "learning_rate": 9.8456954647122e-07,
+      "loss": 0.6249,
+      "step": 2316
+    },
+    {
+      "epoch": 0.4430633903814896,
+      "grad_norm": 3.0129504424994313,
+      "learning_rate": 9.846244053702563e-07,
+      "loss": 0.4851,
+      "step": 2317
+    },
+    {
+      "epoch": 0.4432546132517449,
+      "grad_norm": 2.8114337171447663,
+      "learning_rate": 9.846792405977066e-07,
+      "loss": 0.2633,
+      "step": 2318
+    },
+    {
+      "epoch": 0.44344583612200017,
+      "grad_norm": 1.7751061743560064,
+      "learning_rate": 9.847340521739905e-07,
+      "loss": 0.1768,
+      "step": 2319
+    },
+    {
+      "epoch": 0.4436370589922555,
+      "grad_norm": 2.595300509742631,
+      "learning_rate": 9.847888401195016e-07,
+      "loss": 0.3727,
+      "step": 2320
+    },
+    {
+      "epoch": 0.4438282818625108,
+      "grad_norm": 2.3967538696079562,
+      "learning_rate": 9.848436044546063e-07,
+      "loss": 0.1246,
+      "step": 2321
+    },
+    {
+      "epoch": 0.444019504732766,
+      "grad_norm": 6.577219993694827,
+      "learning_rate": 9.848983451996458e-07,
+      "loss": 0.1213,
+      "step": 2322
+    },
+    {
+      "epoch": 0.4442107276030213,
+      "grad_norm": 2.207945939639235,
+      "learning_rate": 9.849530623749343e-07,
+      "loss": 0.0759,
+      "step": 2323
+    },
+    {
+      "epoch": 0.44440195047327663,
+      "grad_norm": 3.035799682441854,
+      "learning_rate": 9.850077560007599e-07,
+      "loss": 0.0856,
+      "step": 2324
+    },
+    {
+      "epoch": 0.4445931733435319,
+      "grad_norm": 4.372738935115809,
+      "learning_rate": 9.850624260973846e-07,
+      "loss": 0.0681,
+      "step": 2325
+    },
+    {
+      "epoch": 0.4447843962137872,
+      "grad_norm": 5.634710449794856,
+      "learning_rate": 9.851170726850442e-07,
+      "loss": 0.1583,
+      "step": 2326
+    },
+    {
+      "epoch": 0.44497561908404243,
+      "grad_norm": 4.72213320679384,
+      "learning_rate": 9.851716957839485e-07,
+      "loss": 0.3364,
+      "step": 2327
+    },
+    {
+      "epoch": 0.44516684195429773,
+      "grad_norm": 4.866212211362216,
+      "learning_rate": 9.85226295414281e-07,
+      "loss": 0.5181,
+      "step": 2328
+    },
+    {
+      "epoch": 0.44535806482455303,
+      "grad_norm": 2.3904683180183466,
+      "learning_rate": 9.852808715961997e-07,
+      "loss": 0.0786,
+      "step": 2329
+    },
+    {
+      "epoch": 0.4455492876948083,
+      "grad_norm": 5.232961344203838,
+      "learning_rate": 9.85335424349836e-07,
+      "loss": 0.4838,
+      "step": 2330
+    },
+    {
+      "epoch": 0.4457405105650636,
+      "grad_norm": 4.312322419421796,
+      "learning_rate": 9.853899536952957e-07,
+      "loss": 0.5248,
+      "step": 2331
+    },
+    {
+      "epoch": 0.4459317334353189,
+      "grad_norm": 3.069038262589564,
+      "learning_rate": 9.85444459652659e-07,
+      "loss": 0.0926,
+      "step": 2332
+    },
+    {
+      "epoch": 0.44612295630557414,
+      "grad_norm": 3.305303421520177,
+      "learning_rate": 9.854989422419796e-07,
+      "loss": 0.4288,
+      "step": 2333
+    },
+    {
+      "epoch": 0.44631417917582944,
+      "grad_norm": 6.0363106324165745,
+      "learning_rate": 9.855534014832863e-07,
+      "loss": 0.3118,
+      "step": 2334
+    },
+    {
+      "epoch": 0.4465054020460847,
+      "grad_norm": 3.20782260214975,
+      "learning_rate": 9.856078373965813e-07,
+      "loss": 0.1299,
+      "step": 2335
+    },
+    {
+      "epoch": 0.44669662491634,
+      "grad_norm": 4.048271969083716,
+      "learning_rate": 9.856622500018419e-07,
+      "loss": 0.0909,
+      "step": 2336
+    },
+    {
+      "epoch": 0.4468878477865953,
+      "grad_norm": 2.527467798316548,
+      "learning_rate": 9.85716639319019e-07,
+      "loss": 0.0614,
+      "step": 2337
+    },
+    {
+      "epoch": 0.44707907065685054,
+      "grad_norm": 6.749352600699155,
+      "learning_rate": 9.857710053680378e-07,
+      "loss": 0.3266,
+      "step": 2338
+    },
+    {
+      "epoch": 0.44727029352710584,
+      "grad_norm": 6.040514220017724,
+      "learning_rate": 9.858253481687995e-07,
+      "loss": 0.5266,
+      "step": 2339
+    },
+    {
+      "epoch": 0.44746151639736115,
+      "grad_norm": 6.9328903142290415,
+      "learning_rate": 9.858796677411778e-07,
+      "loss": 0.6714,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4476527392676164,
+      "grad_norm": 3.8068568254887794,
+      "learning_rate": 9.85933964105022e-07,
+      "loss": 0.2394,
+      "step": 2341
+    },
+    {
+      "epoch": 0.4478439621378717,
+      "grad_norm": 3.8389737607631074,
+      "learning_rate": 9.859882372801557e-07,
+      "loss": 0.4125,
+      "step": 2342
+    },
+    {
+      "epoch": 0.44803518500812695,
+      "grad_norm": 4.759621562820373,
+      "learning_rate": 9.860424872863772e-07,
+      "loss": 0.2428,
+      "step": 2343
+    },
+    {
+      "epoch": 0.44822640787838225,
+      "grad_norm": 3.960916914124575,
+      "learning_rate": 9.860967141434598e-07,
+      "loss": 0.3285,
+      "step": 2344
+    },
+    {
+      "epoch": 0.44841763074863755,
+      "grad_norm": 2.715395469315483,
+      "learning_rate": 9.861509178711504e-07,
+      "loss": 0.0767,
+      "step": 2345
+    },
+    {
+      "epoch": 0.4486088536188928,
+      "grad_norm": 4.369129274683513,
+      "learning_rate": 9.862050984891717e-07,
+      "loss": 0.364,
+      "step": 2346
+    },
+    {
+      "epoch": 0.4488000764891481,
+      "grad_norm": 4.261636623077257,
+      "learning_rate": 9.86259256017221e-07,
+      "loss": 0.1454,
+      "step": 2347
+    },
+    {
+      "epoch": 0.4489912993594034,
+      "grad_norm": 5.112744634693284,
+      "learning_rate": 9.863133904749702e-07,
+      "loss": 0.1814,
+      "step": 2348
+    },
+    {
+      "epoch": 0.44918252222965865,
+      "grad_norm": 4.826987320767337,
+      "learning_rate": 9.863675018820658e-07,
+      "loss": 0.1716,
+      "step": 2349
+    },
+    {
+      "epoch": 0.44937374509991396,
+      "grad_norm": 6.303224301631641,
+      "learning_rate": 9.8642159025813e-07,
+      "loss": 0.085,
+      "step": 2350
+    },
+    {
+      "epoch": 0.4495649679701692,
+      "grad_norm": 3.509562644120052,
+      "learning_rate": 9.86475655622759e-07,
+      "loss": 0.3086,
+      "step": 2351
+    },
+    {
+      "epoch": 0.4497561908404245,
+      "grad_norm": 5.805243772239625,
+      "learning_rate": 9.865296979955253e-07,
+      "loss": 0.711,
+      "step": 2352
+    },
+    {
+      "epoch": 0.4499474137106798,
+      "grad_norm": 6.94172003986162,
+      "learning_rate": 9.865837173959747e-07,
+      "loss": 0.6843,
+      "step": 2353
+    },
+    {
+      "epoch": 0.45013863658093506,
+      "grad_norm": 3.664051047756623,
+      "learning_rate": 9.8663771384363e-07,
+      "loss": 0.1354,
+      "step": 2354
+    },
+    {
+      "epoch": 0.45032985945119036,
+      "grad_norm": 2.041747798734926,
+      "learning_rate": 9.866916873579873e-07,
+      "loss": 0.2689,
+      "step": 2355
+    },
+    {
+      "epoch": 0.45052108232144567,
+      "grad_norm": 4.28071022807696,
+      "learning_rate": 9.86745637958519e-07,
+      "loss": 0.396,
+      "step": 2356
+    },
+    {
+      "epoch": 0.4507123051917009,
+      "grad_norm": 7.195926669041361,
+      "learning_rate": 9.86799565664673e-07,
+      "loss": 0.1996,
+      "step": 2357
+    },
+    {
+      "epoch": 0.4509035280619562,
+      "grad_norm": 2.385028346630999,
+      "learning_rate": 9.86853470495871e-07,
+      "loss": 0.1216,
+      "step": 2358
+    },
+    {
+      "epoch": 0.4510947509322115,
+      "grad_norm": 2.510034355293209,
+      "learning_rate": 9.869073524715115e-07,
+      "loss": 0.2663,
+      "step": 2359
+    },
+    {
+      "epoch": 0.45128597380246677,
+      "grad_norm": 2.385817836915896,
+      "learning_rate": 9.869612116109674e-07,
+      "loss": 0.0772,
+      "step": 2360
+    },
+    {
+      "epoch": 0.45147719667272207,
+      "grad_norm": 3.04145150316321,
+      "learning_rate": 9.870150479335875e-07,
+      "loss": 0.099,
+      "step": 2361
+    },
+    {
+      "epoch": 0.4516684195429773,
+      "grad_norm": 6.995464217820197,
+      "learning_rate": 9.87068861458696e-07,
+      "loss": 0.1275,
+      "step": 2362
+    },
+    {
+      "epoch": 0.4518596424132326,
+      "grad_norm": 6.378921798434817,
+      "learning_rate": 9.871226522055917e-07,
+      "loss": 0.4388,
+      "step": 2363
+    },
+    {
+      "epoch": 0.4520508652834879,
+      "grad_norm": 4.858220297456484,
+      "learning_rate": 9.871764201935503e-07,
+      "loss": 0.5741,
+      "step": 2364
+    },
+    {
+      "epoch": 0.45224208815374317,
+      "grad_norm": 5.831742787365393,
+      "learning_rate": 9.87230165441822e-07,
+      "loss": 0.6286,
+      "step": 2365
+    },
+    {
+      "epoch": 0.4524333110239985,
+      "grad_norm": 2.933567905439169,
+      "learning_rate": 9.872838879696329e-07,
+      "loss": 0.3694,
+      "step": 2366
+    },
+    {
+      "epoch": 0.4526245338942538,
+      "grad_norm": 2.1812924181467177,
+      "learning_rate": 9.873375877961849e-07,
+      "loss": 0.329,
+      "step": 2367
+    },
+    {
+      "epoch": 0.452815756764509,
+      "grad_norm": 2.6728098165355223,
+      "learning_rate": 9.873912649406552e-07,
+      "loss": 0.2404,
+      "step": 2368
+    },
+    {
+      "epoch": 0.45300697963476433,
+      "grad_norm": 2.909015611194253,
+      "learning_rate": 9.87444919422197e-07,
+      "loss": 0.2729,
+      "step": 2369
+    },
+    {
+      "epoch": 0.4531982025050196,
+      "grad_norm": 6.175238588488372,
+      "learning_rate": 9.874985512599394e-07,
+      "loss": 0.2577,
+      "step": 2370
+    },
+    {
+      "epoch": 0.4533894253752749,
+      "grad_norm": 2.3174264343502875,
+      "learning_rate": 9.875521604729868e-07,
+      "loss": 0.1773,
+      "step": 2371
+    },
+    {
+      "epoch": 0.4535806482455302,
+      "grad_norm": 2.580468794348887,
+      "learning_rate": 9.876057470804195e-07,
+      "loss": 0.1134,
+      "step": 2372
+    },
+    {
+      "epoch": 0.45377187111578543,
+      "grad_norm": 3.564569507945881,
+      "learning_rate": 9.876593111012945e-07,
+      "loss": 0.2147,
+      "step": 2373
+    },
+    {
+      "epoch": 0.45396309398604073,
+      "grad_norm": 2.772143252701955,
+      "learning_rate": 9.877128525546436e-07,
+      "loss": 0.1059,
+      "step": 2374
+    },
+    {
+      "epoch": 0.45415431685629604,
+      "grad_norm": 2.624913713762741,
+      "learning_rate": 9.877663714594751e-07,
+      "loss": 0.1042,
+      "step": 2375
+    },
+    {
+      "epoch": 0.4543455397265513,
+      "grad_norm": 4.375708386428036,
+      "learning_rate": 9.878198678347733e-07,
+      "loss": 0.1463,
+      "step": 2376
+    },
+    {
+      "epoch": 0.4545367625968066,
+      "grad_norm": 3.941199245253111,
+      "learning_rate": 9.878733416994985e-07,
+      "loss": 0.4756,
+      "step": 2377
+    },
+    {
+      "epoch": 0.45472798546706183,
+      "grad_norm": 5.336750108271854,
+      "learning_rate": 9.87926793072587e-07,
+      "loss": 0.4948,
+      "step": 2378
+    },
+    {
+      "epoch": 0.45491920833731714,
+      "grad_norm": 4.806668798805078,
+      "learning_rate": 9.87980221972951e-07,
+      "loss": 0.2236,
+      "step": 2379
+    },
+    {
+      "epoch": 0.45511043120757244,
+      "grad_norm": 3.5696902077093626,
+      "learning_rate": 9.880336284194794e-07,
+      "loss": 0.2345,
+      "step": 2380
+    },
+    {
+      "epoch": 0.4553016540778277,
+      "grad_norm": 3.822706724356376,
+      "learning_rate": 9.88087012431037e-07,
+      "loss": 0.1886,
+      "step": 2381
+    },
+    {
+      "epoch": 0.455492876948083,
+      "grad_norm": 3.398311919599651,
+      "learning_rate": 9.88140374026465e-07,
+      "loss": 0.1323,
+      "step": 2382
+    },
+    {
+      "epoch": 0.4556840998183383,
+      "grad_norm": 3.431328522108564,
+      "learning_rate": 9.881937132245803e-07,
+      "loss": 0.2139,
+      "step": 2383
+    },
+    {
+      "epoch": 0.45587532268859354,
+      "grad_norm": 4.5609736567570565,
+      "learning_rate": 9.882470300441768e-07,
+      "loss": 0.286,
+      "step": 2384
+    },
+    {
+      "epoch": 0.45606654555884885,
+      "grad_norm": 2.825066530237572,
+      "learning_rate": 9.883003245040247e-07,
+      "loss": 0.2235,
+      "step": 2385
+    },
+    {
+      "epoch": 0.4562577684291041,
+      "grad_norm": 1.9646578448645187,
+      "learning_rate": 9.883535966228696e-07,
+      "loss": 0.0654,
+      "step": 2386
+    },
+    {
+      "epoch": 0.4564489912993594,
+      "grad_norm": 2.7890003528669975,
+      "learning_rate": 9.884068464194352e-07,
+      "loss": 0.1171,
+      "step": 2387
+    },
+    {
+      "epoch": 0.4566402141696147,
+      "grad_norm": 5.908558550776087,
+      "learning_rate": 9.884600739124203e-07,
+      "loss": 0.19,
+      "step": 2388
+    },
+    {
+      "epoch": 0.45683143703986995,
+      "grad_norm": 3.4563086837560135,
+      "learning_rate": 9.88513279120501e-07,
+      "loss": 0.291,
+      "step": 2389
+    },
+    {
+      "epoch": 0.45702265991012525,
+      "grad_norm": 6.491941885772098,
+      "learning_rate": 9.885664620623294e-07,
+      "loss": 0.4589,
+      "step": 2390
+    },
+    {
+      "epoch": 0.45721388278038055,
+      "grad_norm": 2.675455828181397,
+      "learning_rate": 9.886196227565344e-07,
+      "loss": 0.224,
+      "step": 2391
+    },
+    {
+      "epoch": 0.4574051056506358,
+      "grad_norm": 3.847260771434711,
+      "learning_rate": 9.88672761221722e-07,
+      "loss": 0.2166,
+      "step": 2392
+    },
+    {
+      "epoch": 0.4575963285208911,
+      "grad_norm": 3.2030907605830072,
+      "learning_rate": 9.88725877476474e-07,
+      "loss": 0.1784,
+      "step": 2393
+    },
+    {
+      "epoch": 0.4577875513911464,
+      "grad_norm": 3.0242515079317625,
+      "learning_rate": 9.887789715393495e-07,
+      "loss": 0.0929,
+      "step": 2394
+    },
+    {
+      "epoch": 0.45797877426140166,
+      "grad_norm": 2.30222253392969,
+      "learning_rate": 9.888320434288844e-07,
+      "loss": 0.1978,
+      "step": 2395
+    },
+    {
+      "epoch": 0.45816999713165696,
+      "grad_norm": 2.6329341147206318,
+      "learning_rate": 9.88885093163591e-07,
+      "loss": 0.0585,
+      "step": 2396
+    },
+    {
+      "epoch": 0.4583612200019122,
+      "grad_norm": 4.170945196083899,
+      "learning_rate": 9.889381207619586e-07,
+      "loss": 0.2377,
+      "step": 2397
+    },
+    {
+      "epoch": 0.4585524428721675,
+      "grad_norm": 3.3832412694673075,
+      "learning_rate": 9.889911262424536e-07,
+      "loss": 0.1492,
+      "step": 2398
+    },
+    {
+      "epoch": 0.4587436657424228,
+      "grad_norm": 2.7592931992360987,
+      "learning_rate": 9.89044109623519e-07,
+      "loss": 0.128,
+      "step": 2399
+    },
+    {
+      "epoch": 0.45893488861267806,
+      "grad_norm": 3.400884760011563,
+      "learning_rate": 9.890970709235746e-07,
+      "loss": 0.0624,
+      "step": 2400
+    },
+    {
+      "epoch": 0.45912611148293336,
+      "grad_norm": 4.774679818328525,
+      "learning_rate": 9.891500101610175e-07,
+      "loss": 0.4011,
+      "step": 2401
+    },
+    {
+      "epoch": 0.45931733435318867,
+      "grad_norm": 4.24968561243834,
+      "learning_rate": 9.89202927354222e-07,
+      "loss": 0.3871,
+      "step": 2402
+    },
+    {
+      "epoch": 0.4595085572234439,
+      "grad_norm": 4.788592914503182,
+      "learning_rate": 9.89255822521539e-07,
+      "loss": 0.3755,
+      "step": 2403
+    },
+    {
+      "epoch": 0.4596997800936992,
+      "grad_norm": 4.336350759796456,
+      "learning_rate": 9.893086956812966e-07,
+      "loss": 0.2489,
+      "step": 2404
+    },
+    {
+      "epoch": 0.45989100296395446,
+      "grad_norm": 6.905708119118911,
+      "learning_rate": 9.893615468518001e-07,
+      "loss": 0.7651,
+      "step": 2405
+    },
+    {
+      "epoch": 0.46008222583420977,
+      "grad_norm": 5.588317790765913,
+      "learning_rate": 9.894143760513322e-07,
+      "loss": 0.5572,
+      "step": 2406
+    },
+    {
+      "epoch": 0.46027344870446507,
+      "grad_norm": 1.7341386788691522,
+      "learning_rate": 9.894671832981523e-07,
+      "loss": 0.2293,
+      "step": 2407
+    },
+    {
+      "epoch": 0.4604646715747203,
+      "grad_norm": 3.8308995993653343,
+      "learning_rate": 9.895199686104978e-07,
+      "loss": 0.1527,
+      "step": 2408
+    },
+    {
+      "epoch": 0.4606558944449756,
+      "grad_norm": 4.606361153864728,
+      "learning_rate": 9.895727320065823e-07,
+      "loss": 0.1425,
+      "step": 2409
+    },
+    {
+      "epoch": 0.4608471173152309,
+      "grad_norm": 2.736032561214393,
+      "learning_rate": 9.896254735045978e-07,
+      "loss": 0.0675,
+      "step": 2410
+    },
+    {
+      "epoch": 0.4610383401854862,
+      "grad_norm": 3.6374240932868727,
+      "learning_rate": 9.896781931227129e-07,
+      "loss": 0.0801,
+      "step": 2411
+    },
+    {
+      "epoch": 0.4612295630557415,
+      "grad_norm": 4.0549662058747336,
+      "learning_rate": 9.897308908790741e-07,
+      "loss": 0.2627,
+      "step": 2412
+    },
+    {
+      "epoch": 0.4614207859259967,
+      "grad_norm": 7.601691469390269,
+      "learning_rate": 9.897835667918053e-07,
+      "loss": 0.4785,
+      "step": 2413
+    },
+    {
+      "epoch": 0.461612008796252,
+      "grad_norm": 5.520577171668703,
+      "learning_rate": 9.898362208790075e-07,
+      "loss": 0.4675,
+      "step": 2414
+    },
+    {
+      "epoch": 0.46180323166650733,
+      "grad_norm": 4.477110611146264,
+      "learning_rate": 9.898888531587593e-07,
+      "loss": 0.3352,
+      "step": 2415
+    },
+    {
+      "epoch": 0.4619944545367626,
+      "grad_norm": 2.726303831580825,
+      "learning_rate": 9.89941463649117e-07,
+      "loss": 0.3039,
+      "step": 2416
+    },
+    {
+      "epoch": 0.4621856774070179,
+      "grad_norm": 2.2025823262965156,
+      "learning_rate": 9.899940523681147e-07,
+      "loss": 0.0776,
+      "step": 2417
+    },
+    {
+      "epoch": 0.4623769002772732,
+      "grad_norm": 4.348294302980988,
+      "learning_rate": 9.900466193337635e-07,
+      "loss": 0.3336,
+      "step": 2418
+    },
+    {
+      "epoch": 0.46256812314752843,
+      "grad_norm": 2.9657646829652755,
+      "learning_rate": 9.90099164564053e-07,
+      "loss": 0.2164,
+      "step": 2419
+    },
+    {
+      "epoch": 0.46275934601778373,
+      "grad_norm": 5.221315404378666,
+      "learning_rate": 9.901516880769493e-07,
+      "loss": 0.6028,
+      "step": 2420
+    },
+    {
+      "epoch": 0.462950568888039,
+      "grad_norm": 4.385713810341562,
+      "learning_rate": 9.902041898903978e-07,
+      "loss": 0.3629,
+      "step": 2421
+    },
+    {
+      "epoch": 0.4631417917582943,
+      "grad_norm": 3.2619542567953483,
+      "learning_rate": 9.902566700223202e-07,
+      "loss": 0.1103,
+      "step": 2422
+    },
+    {
+      "epoch": 0.4633330146285496,
+      "grad_norm": 4.1981412134940745,
+      "learning_rate": 9.903091284906169e-07,
+      "loss": 0.1713,
+      "step": 2423
+    },
+    {
+      "epoch": 0.46352423749880484,
+      "grad_norm": 3.2965037059974294,
+      "learning_rate": 9.903615653131658e-07,
+      "loss": 0.0758,
+      "step": 2424
+    },
+    {
+      "epoch": 0.46371546036906014,
+      "grad_norm": 6.389922982529308,
+      "learning_rate": 9.904139805078227e-07,
+      "loss": 0.1735,
+      "step": 2425
+    },
+    {
+      "epoch": 0.46390668323931544,
+      "grad_norm": 8.776818294101558,
+      "learning_rate": 9.904663740924215e-07,
+      "loss": 0.2891,
+      "step": 2426
+    },
+    {
+      "epoch": 0.4640979061095707,
+      "grad_norm": 4.88546139087282,
+      "learning_rate": 9.905187460847737e-07,
+      "loss": 0.3143,
+      "step": 2427
+    },
+    {
+      "epoch": 0.464289128979826,
+      "grad_norm": 3.614815579006355,
+      "learning_rate": 9.905710965026694e-07,
+      "loss": 0.3499,
+      "step": 2428
+    },
+    {
+      "epoch": 0.4644803518500813,
+      "grad_norm": 1.7256852074454143,
+      "learning_rate": 9.906234253638759e-07,
+      "loss": 0.1839,
+      "step": 2429
+    },
+    {
+      "epoch": 0.46467157472033654,
+      "grad_norm": 3.8154746785963938,
+      "learning_rate": 9.90675732686139e-07,
+      "loss": 0.3288,
+      "step": 2430
+    },
+    {
+      "epoch": 0.46486279759059185,
+      "grad_norm": 3.947155089720586,
+      "learning_rate": 9.907280184871827e-07,
+      "loss": 0.4448,
+      "step": 2431
+    },
+    {
+      "epoch": 0.4650540204608471,
+      "grad_norm": 6.582624984812201,
+      "learning_rate": 9.90780282784709e-07,
+      "loss": 0.7721,
+      "step": 2432
+    },
+    {
+      "epoch": 0.4652452433311024,
+      "grad_norm": 3.9515833807740868,
+      "learning_rate": 9.908325255963983e-07,
+      "loss": 0.1637,
+      "step": 2433
+    },
+    {
+      "epoch": 0.4654364662013577,
+      "grad_norm": 3.785378968228342,
+      "learning_rate": 9.908847469399087e-07,
+      "loss": 0.1841,
+      "step": 2434
+    },
+    {
+      "epoch": 0.46562768907161295,
+      "grad_norm": 5.234799860322906,
+      "learning_rate": 9.909369468328768e-07,
+      "loss": 0.3155,
+      "step": 2435
+    },
+    {
+      "epoch": 0.46581891194186825,
+      "grad_norm": 4.0868530414724695,
+      "learning_rate": 9.909891252929177e-07,
+      "loss": 0.0596,
+      "step": 2436
+    },
+    {
+      "epoch": 0.46601013481212356,
+      "grad_norm": 3.710317845057555,
+      "learning_rate": 9.910412823376243e-07,
+      "loss": 0.1072,
+      "step": 2437
+    },
+    {
+      "epoch": 0.4662013576823788,
+      "grad_norm": 3.621925892114408,
+      "learning_rate": 9.910934179845687e-07,
+      "loss": 0.1508,
+      "step": 2438
+    },
+    {
+      "epoch": 0.4663925805526341,
+      "grad_norm": 10.009404528547805,
+      "learning_rate": 9.911455322513004e-07,
+      "loss": 0.2493,
+      "step": 2439
+    },
+    {
+      "epoch": 0.46658380342288935,
+      "grad_norm": 4.441343979747071,
+      "learning_rate": 9.911976251553478e-07,
+      "loss": 0.5881,
+      "step": 2440
+    },
+    {
+      "epoch": 0.46677502629314466,
+      "grad_norm": 3.3176383705287984,
+      "learning_rate": 9.912496967142178e-07,
+      "loss": 0.402,
+      "step": 2441
+    },
+    {
+      "epoch": 0.46696624916339996,
+      "grad_norm": 5.110267677836039,
+      "learning_rate": 9.913017469453956e-07,
+      "loss": 0.2722,
+      "step": 2442
+    },
+    {
+      "epoch": 0.4671574720336552,
+      "grad_norm": 3.6152581163118347,
+      "learning_rate": 9.913537758663452e-07,
+      "loss": 0.2633,
+      "step": 2443
+    },
+    {
+      "epoch": 0.4673486949039105,
+      "grad_norm": 3.2736892057165417,
+      "learning_rate": 9.914057834945087e-07,
+      "loss": 0.1567,
+      "step": 2444
+    },
+    {
+      "epoch": 0.4675399177741658,
+      "grad_norm": 3.1894006485368385,
+      "learning_rate": 9.914577698473076e-07,
+      "loss": 0.2404,
+      "step": 2445
+    },
+    {
+      "epoch": 0.46773114064442106,
+      "grad_norm": 3.286433603200751,
+      "learning_rate": 9.915097349421409e-07,
+      "loss": 0.1279,
+      "step": 2446
+    },
+    {
+      "epoch": 0.46792236351467636,
+      "grad_norm": 3.474145353317003,
+      "learning_rate": 9.915616787963872e-07,
+      "loss": 0.0945,
+      "step": 2447
+    },
+    {
+      "epoch": 0.4681135863849316,
+      "grad_norm": 3.1567418687745468,
+      "learning_rate": 9.916136014274034e-07,
+      "loss": 0.133,
+      "step": 2448
+    },
+    {
+      "epoch": 0.4683048092551869,
+      "grad_norm": 3.8240525362483964,
+      "learning_rate": 9.91665502852525e-07,
+      "loss": 0.1114,
+      "step": 2449
+    },
+    {
+      "epoch": 0.4684960321254422,
+      "grad_norm": 7.322083803424079,
+      "learning_rate": 9.91717383089067e-07,
+      "loss": 0.1369,
+      "step": 2450
+    },
+    {
+      "epoch": 0.46868725499569747,
+      "grad_norm": 4.621104068696214,
+      "learning_rate": 9.91769242154322e-07,
+      "loss": 0.4703,
+      "step": 2451
+    },
+    {
+      "epoch": 0.46887847786595277,
+      "grad_norm": 7.611638603896544,
+      "learning_rate": 9.918210800655627e-07,
+      "loss": 0.5892,
+      "step": 2452
+    },
+    {
+      "epoch": 0.4690697007362081,
+      "grad_norm": 7.903659803366635,
+      "learning_rate": 9.918728968400398e-07,
+      "loss": 0.1467,
+      "step": 2453
+    },
+    {
+      "epoch": 0.4692609236064633,
+      "grad_norm": 4.246053546339682,
+      "learning_rate": 9.91924692494983e-07,
+      "loss": 0.3653,
+      "step": 2454
+    },
+    {
+      "epoch": 0.4694521464767186,
+      "grad_norm": 3.4552884466440306,
+      "learning_rate": 9.919764670476014e-07,
+      "loss": 0.3267,
+      "step": 2455
+    },
+    {
+      "epoch": 0.46964336934697387,
+      "grad_norm": 3.5488973221359847,
+      "learning_rate": 9.92028220515083e-07,
+      "loss": 0.2941,
+      "step": 2456
+    },
+    {
+      "epoch": 0.4698345922172292,
+      "grad_norm": 2.3951321667735743,
+      "learning_rate": 9.920799529145941e-07,
+      "loss": 0.1786,
+      "step": 2457
+    },
+    {
+      "epoch": 0.4700258150874845,
+      "grad_norm": 3.1186978689358607,
+      "learning_rate": 9.921316642632808e-07,
+      "loss": 0.1902,
+      "step": 2458
+    },
+    {
+      "epoch": 0.4702170379577397,
+      "grad_norm": 7.662496953219979,
+      "learning_rate": 9.92183354578268e-07,
+      "loss": 0.4151,
+      "step": 2459
+    },
+    {
+      "epoch": 0.47040826082799503,
+      "grad_norm": 2.924581977121818,
+      "learning_rate": 9.9223502387666e-07,
+      "loss": 0.1414,
+      "step": 2460
+    },
+    {
+      "epoch": 0.47059948369825033,
+      "grad_norm": 3.0515249911189226,
+      "learning_rate": 9.922866721755397e-07,
+      "loss": 0.0798,
+      "step": 2461
+    },
+    {
+      "epoch": 0.4707907065685056,
+      "grad_norm": 4.113063302927237,
+      "learning_rate": 9.923382994919696e-07,
+      "loss": 0.1519,
+      "step": 2462
+    },
+    {
+      "epoch": 0.4709819294387609,
+      "grad_norm": 8.152267945562755,
+      "learning_rate": 9.92389905842991e-07,
+      "loss": 0.271,
+      "step": 2463
+    },
+    {
+      "epoch": 0.4711731523090162,
+      "grad_norm": 6.442515669223735,
+      "learning_rate": 9.924414912456252e-07,
+      "loss": 0.6266,
+      "step": 2464
+    },
+    {
+      "epoch": 0.47136437517927143,
+      "grad_norm": 3.6131015052420845,
+      "learning_rate": 9.924930557168719e-07,
+      "loss": 0.345,
+      "step": 2465
+    },
+    {
+      "epoch": 0.47155559804952674,
+      "grad_norm": 2.4830667181926165,
+      "learning_rate": 9.92544599273711e-07,
+      "loss": 0.1022,
+      "step": 2466
+    },
+    {
+      "epoch": 0.471746820919782,
+      "grad_norm": 2.9393366489589257,
+      "learning_rate": 9.925961219331006e-07,
+      "loss": 0.2123,
+      "step": 2467
+    },
+    {
+      "epoch": 0.4719380437900373,
+      "grad_norm": 3.2329348983935597,
+      "learning_rate": 9.926476237119796e-07,
+      "loss": 0.4972,
+      "step": 2468
+    },
+    {
+      "epoch": 0.4721292666602926,
+      "grad_norm": 2.3274409165960286,
+      "learning_rate": 9.92699104627265e-07,
+      "loss": 0.1827,
+      "step": 2469
+    },
+    {
+      "epoch": 0.47232048953054784,
+      "grad_norm": 3.561820383665578,
+      "learning_rate": 9.92750564695854e-07,
+      "loss": 0.3903,
+      "step": 2470
+    },
+    {
+      "epoch": 0.47251171240080314,
+      "grad_norm": 1.6235952908271187,
+      "learning_rate": 9.928020039346232e-07,
+      "loss": 0.122,
+      "step": 2471
+    },
+    {
+      "epoch": 0.47270293527105844,
+      "grad_norm": 2.334840753095524,
+      "learning_rate": 9.928534223604285e-07,
+      "loss": 0.071,
+      "step": 2472
+    },
+    {
+      "epoch": 0.4728941581413137,
+      "grad_norm": 2.69328855735397,
+      "learning_rate": 9.929048199901056e-07,
+      "loss": 0.0861,
+      "step": 2473
+    },
+    {
+      "epoch": 0.473085381011569,
+      "grad_norm": 2.6987372271123764,
+      "learning_rate": 9.929561968404695e-07,
+      "loss": 0.0948,
+      "step": 2474
+    },
+    {
+      "epoch": 0.47327660388182424,
+      "grad_norm": 3.5670581234409204,
+      "learning_rate": 9.93007552928315e-07,
+      "loss": 0.0743,
+      "step": 2475
+    },
+    {
+      "epoch": 0.47346782675207955,
+      "grad_norm": 8.64027136411846,
+      "learning_rate": 9.930588882704166e-07,
+      "loss": 0.4149,
+      "step": 2476
+    },
+    {
+      "epoch": 0.47365904962233485,
+      "grad_norm": 7.270836714458682,
+      "learning_rate": 9.93110202883528e-07,
+      "loss": 0.7718,
+      "step": 2477
+    },
+    {
+      "epoch": 0.4738502724925901,
+      "grad_norm": 4.065145716759281,
+      "learning_rate": 9.931614967843836e-07,
+      "loss": 0.234,
+      "step": 2478
+    },
+    {
+      "epoch": 0.4740414953628454,
+      "grad_norm": 2.1711525573123653,
+      "learning_rate": 9.932127699896965e-07,
+      "loss": 0.4205,
+      "step": 2479
+    },
+    {
+      "epoch": 0.4742327182331007,
+      "grad_norm": 2.288076038477573,
+      "learning_rate": 9.932640225161603e-07,
+      "loss": 0.2139,
+      "step": 2480
+    },
+    {
+      "epoch": 0.47442394110335595,
+      "grad_norm": 2.974533230363648,
+      "learning_rate": 9.93315254380448e-07,
+      "loss": 0.3001,
+      "step": 2481
+    },
+    {
+      "epoch": 0.47461516397361125,
+      "grad_norm": 2.368965614188453,
+      "learning_rate": 9.933664655992124e-07,
+      "loss": 0.2043,
+      "step": 2482
+    },
+    {
+      "epoch": 0.4748063868438665,
+      "grad_norm": 4.905559904318111,
+      "learning_rate": 9.934176561890865e-07,
+      "loss": 0.1659,
+      "step": 2483
+    },
+    {
+      "epoch": 0.4749976097141218,
+      "grad_norm": 3.7685906219943925,
+      "learning_rate": 9.93468826166683e-07,
+      "loss": 0.1759,
+      "step": 2484
+    },
+    {
+      "epoch": 0.4751888325843771,
+      "grad_norm": 17.014618814947596,
+      "learning_rate": 9.93519975548595e-07,
+      "loss": 0.1375,
+      "step": 2485
+    },
+    {
+      "epoch": 0.47538005545463236,
+      "grad_norm": 2.818859878056282,
+      "learning_rate": 9.935711043513945e-07,
+      "loss": 0.0993,
+      "step": 2486
+    },
+    {
+      "epoch": 0.47557127832488766,
+      "grad_norm": 2.7844136279302703,
+      "learning_rate": 9.936222125916345e-07,
+      "loss": 0.0429,
+      "step": 2487
+    },
+    {
+      "epoch": 0.47576250119514296,
+      "grad_norm": 9.79272684620952,
+      "learning_rate": 9.93673300285848e-07,
+      "loss": 0.3285,
+      "step": 2488
+    },
+    {
+      "epoch": 0.4759537240653982,
+      "grad_norm": 4.647834289905975,
+      "learning_rate": 9.937243674505473e-07,
+      "loss": 0.414,
+      "step": 2489
+    },
+    {
+      "epoch": 0.4761449469356535,
+      "grad_norm": 5.179876033522062,
+      "learning_rate": 9.937754141022255e-07,
+      "loss": 0.3679,
+      "step": 2490
+    },
+    {
+      "epoch": 0.47633616980590876,
+      "grad_norm": 3.270255424486264,
+      "learning_rate": 9.938264402573556e-07,
+      "loss": 0.355,
+      "step": 2491
+    },
+    {
+      "epoch": 0.47652739267616406,
+      "grad_norm": 3.1951320755811823,
+      "learning_rate": 9.938774459323907e-07,
+      "loss": 0.2719,
+      "step": 2492
+    },
+    {
+      "epoch": 0.47671861554641937,
+      "grad_norm": 3.786009134393425,
+      "learning_rate": 9.939284311437644e-07,
+      "loss": 0.3281,
+      "step": 2493
+    },
+    {
+      "epoch": 0.4769098384166746,
+      "grad_norm": 3.4932597517558444,
+      "learning_rate": 9.939793959078902e-07,
+      "loss": 0.431,
+      "step": 2494
+    },
+    {
+      "epoch": 0.4771010612869299,
+      "grad_norm": 2.0309905840206084,
+      "learning_rate": 9.940303402411618e-07,
+      "loss": 0.1557,
+      "step": 2495
+    },
+    {
+      "epoch": 0.4772922841571852,
+      "grad_norm": 5.6381054267674635,
+      "learning_rate": 9.940812641599534e-07,
+      "loss": 0.1426,
+      "step": 2496
+    },
+    {
+      "epoch": 0.47748350702744047,
+      "grad_norm": 4.263482865929133,
+      "learning_rate": 9.941321676806198e-07,
+      "loss": 0.3392,
+      "step": 2497
+    },
+    {
+      "epoch": 0.47767472989769577,
+      "grad_norm": 2.2874982781742474,
+      "learning_rate": 9.941830508194956e-07,
+      "loss": 0.0731,
+      "step": 2498
+    },
+    {
+      "epoch": 0.4778659527679511,
+      "grad_norm": 3.0810454016302224,
+      "learning_rate": 9.942339135928956e-07,
+      "loss": 0.0913,
+      "step": 2499
+    },
+    {
+      "epoch": 0.4780571756382063,
+      "grad_norm": 2.405717468783859,
+      "learning_rate": 9.942847560171163e-07,
+      "loss": 0.0346,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4780571756382063,
+      "eval_runtime": 753.0242,
+      "eval_samples_per_second": 2.037,
+      "eval_steps_per_second": 0.51,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4782483985084616,
+      "grad_norm": 5.650900993412778,
+      "learning_rate": 9.943355781084332e-07,
+      "loss": 0.3306,
+      "step": 2501
+    },
+    {
+      "epoch": 0.4784396213787169,
+      "grad_norm": 3.0250933546053003,
+      "learning_rate": 9.94386379883103e-07,
+      "loss": 0.2562,
+      "step": 2502
+    },
+    {
+      "epoch": 0.4786308442489722,
+      "grad_norm": 3.2390839579230746,
+      "learning_rate": 9.94437161357363e-07,
+      "loss": 0.2529,
+      "step": 2503
+    },
+    {
+      "epoch": 0.4788220671192275,
+      "grad_norm": 3.1841970047698913,
+      "learning_rate": 9.944879225474304e-07,
+      "loss": 0.3275,
+      "step": 2504
+    },
+    {
+      "epoch": 0.4790132899894827,
+      "grad_norm": 4.756745618144029,
+      "learning_rate": 9.945386634695035e-07,
+      "loss": 0.2194,
+      "step": 2505
+    },
+    {
+      "epoch": 0.47920451285973803,
+      "grad_norm": 2.99554080162729,
+      "learning_rate": 9.945893841397615e-07,
+      "loss": 0.2475,
+      "step": 2506
+    },
+    {
+      "epoch": 0.47939573572999333,
+      "grad_norm": 4.9472667816076425,
+      "learning_rate": 9.946400845743634e-07,
+      "loss": 0.6179,
+      "step": 2507
+    },
+    {
+      "epoch": 0.4795869586002486,
+      "grad_norm": 2.529281604381991,
+      "learning_rate": 9.946907647894497e-07,
+      "loss": 0.2552,
+      "step": 2508
+    },
+    {
+      "epoch": 0.4797781814705039,
+      "grad_norm": 2.1976601208362196,
+      "learning_rate": 9.947414248011406e-07,
+      "loss": 0.1344,
+      "step": 2509
+    },
+    {
+      "epoch": 0.47996940434075913,
+      "grad_norm": 9.962579521522338,
+      "learning_rate": 9.947920646255383e-07,
+      "loss": 0.0658,
+      "step": 2510
+    },
+    {
+      "epoch": 0.48016062721101443,
+      "grad_norm": 3.053540573027961,
+      "learning_rate": 9.948426842787247e-07,
+      "loss": 0.1138,
+      "step": 2511
+    },
+    {
+      "epoch": 0.48035185008126974,
+      "grad_norm": 3.37835470119118,
+      "learning_rate": 9.94893283776763e-07,
+      "loss": 0.1109,
+      "step": 2512
+    },
+    {
+      "epoch": 0.480543072951525,
+      "grad_norm": 6.402961594082302,
+      "learning_rate": 9.94943863135697e-07,
+      "loss": 0.1532,
+      "step": 2513
+    },
+    {
+      "epoch": 0.4807342958217803,
+      "grad_norm": 3.8417543021996132,
+      "learning_rate": 9.949944223715515e-07,
+      "loss": 0.4004,
+      "step": 2514
+    },
+    {
+      "epoch": 0.4809255186920356,
+      "grad_norm": 1.9152862996160835,
+      "learning_rate": 9.950449615003323e-07,
+      "loss": 0.0697,
+      "step": 2515
+    },
+    {
+      "epoch": 0.48111674156229084,
+      "grad_norm": 3.313875308723546,
+      "learning_rate": 9.950954805380255e-07,
+      "loss": 0.4192,
+      "step": 2516
+    },
+    {
+      "epoch": 0.48130796443254614,
+      "grad_norm": 4.48765247479092,
+      "learning_rate": 9.951459795005993e-07,
+      "loss": 0.6735,
+      "step": 2517
+    },
+    {
+      "epoch": 0.4814991873028014,
+      "grad_norm": 2.8938307023793226,
+      "learning_rate": 9.951964584040016e-07,
+      "loss": 0.2039,
+      "step": 2518
+    },
+    {
+      "epoch": 0.4816904101730567,
+      "grad_norm": 3.702727606330848,
+      "learning_rate": 9.95246917264162e-07,
+      "loss": 0.2608,
+      "step": 2519
+    },
+    {
+      "epoch": 0.481881633043312,
+      "grad_norm": 4.315217198710385,
+      "learning_rate": 9.952973560969907e-07,
+      "loss": 0.1234,
+      "step": 2520
+    },
+    {
+      "epoch": 0.48207285591356724,
+      "grad_norm": 4.378445821439303,
+      "learning_rate": 9.9534777491838e-07,
+      "loss": 0.1906,
+      "step": 2521
+    },
+    {
+      "epoch": 0.48226407878382255,
+      "grad_norm": 4.015781503440398,
+      "learning_rate": 9.953981737442017e-07,
+      "loss": 0.3923,
+      "step": 2522
+    },
+    {
+      "epoch": 0.48245530165407785,
+      "grad_norm": 8.376507979181598,
+      "learning_rate": 9.954485525903102e-07,
+      "loss": 0.2564,
+      "step": 2523
+    },
+    {
+      "epoch": 0.4826465245243331,
+      "grad_norm": 3.9472675578298086,
+      "learning_rate": 9.9549891147254e-07,
+      "loss": 0.1338,
+      "step": 2524
+    },
+    {
+      "epoch": 0.4828377473945884,
+      "grad_norm": 6.5389510752528786,
+      "learning_rate": 9.955492504067075e-07,
+      "loss": 0.1009,
+      "step": 2525
+    },
+    {
+      "epoch": 0.48302897026484365,
+      "grad_norm": 4.640623766966376,
+      "learning_rate": 9.9559956940861e-07,
+      "loss": 0.1412,
+      "step": 2526
+    },
+    {
+      "epoch": 0.48322019313509895,
+      "grad_norm": 6.028619538504571,
+      "learning_rate": 9.95649868494026e-07,
+      "loss": 0.4166,
+      "step": 2527
+    },
+    {
+      "epoch": 0.48341141600535426,
+      "grad_norm": 4.612624855729369,
+      "learning_rate": 9.95700147678715e-07,
+      "loss": 0.3214,
+      "step": 2528
+    },
+    {
+      "epoch": 0.4836026388756095,
+      "grad_norm": 2.727291064489736,
+      "learning_rate": 9.957504069784187e-07,
+      "loss": 0.2603,
+      "step": 2529
+    },
+    {
+      "epoch": 0.4837938617458648,
+      "grad_norm": 2.881070569589215,
+      "learning_rate": 9.958006464088593e-07,
+      "loss": 0.1446,
+      "step": 2530
+    },
+    {
+      "epoch": 0.4839850846161201,
+      "grad_norm": 2.105104545842268,
+      "learning_rate": 9.958508659857405e-07,
+      "loss": 0.1558,
+      "step": 2531
+    },
+    {
+      "epoch": 0.48417630748637536,
+      "grad_norm": 1.9224641718706919,
+      "learning_rate": 9.959010657247477e-07,
+      "loss": 0.1772,
+      "step": 2532
+    },
+    {
+      "epoch": 0.48436753035663066,
+      "grad_norm": 2.2030702408282288,
+      "learning_rate": 9.959512456415473e-07,
+      "loss": 0.1712,
+      "step": 2533
+    },
+    {
+      "epoch": 0.48455875322688596,
+      "grad_norm": 4.299355640481472,
+      "learning_rate": 9.960014057517879e-07,
+      "loss": 0.2387,
+      "step": 2534
+    },
+    {
+      "epoch": 0.4847499760971412,
+      "grad_norm": 6.0181283473295615,
+      "learning_rate": 9.960515460710983e-07,
+      "loss": 0.1413,
+      "step": 2535
+    },
+    {
+      "epoch": 0.4849411989673965,
+      "grad_norm": 4.950364363390059,
+      "learning_rate": 9.961016666150903e-07,
+      "loss": 0.1867,
+      "step": 2536
+    },
+    {
+      "epoch": 0.48513242183765176,
+      "grad_norm": 7.967849022289026,
+      "learning_rate": 9.96151767399356e-07,
+      "loss": 0.1297,
+      "step": 2537
+    },
+    {
+      "epoch": 0.48532364470790706,
+      "grad_norm": 6.430675555822114,
+      "learning_rate": 9.9620184843947e-07,
+      "loss": 0.2273,
+      "step": 2538
+    },
+    {
+      "epoch": 0.48551486757816237,
+      "grad_norm": 6.741515301803677,
+      "learning_rate": 9.962519097509877e-07,
+      "loss": 0.3761,
+      "step": 2539
+    },
+    {
+      "epoch": 0.4857060904484176,
+      "grad_norm": 4.630663316969498,
+      "learning_rate": 9.963019513494467e-07,
+      "loss": 0.3374,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4858973133186729,
+      "grad_norm": 3.557626971876128,
+      "learning_rate": 9.963519732503656e-07,
+      "loss": 0.3066,
+      "step": 2541
+    },
+    {
+      "epoch": 0.4860885361889282,
+      "grad_norm": 2.8508197810045957,
+      "learning_rate": 9.964019754692458e-07,
+      "loss": 0.242,
+      "step": 2542
+    },
+    {
+      "epoch": 0.48627975905918347,
+      "grad_norm": 4.796603018042593,
+      "learning_rate": 9.964519580215692e-07,
+      "loss": 0.2054,
+      "step": 2543
+    },
+    {
+      "epoch": 0.4864709819294388,
+      "grad_norm": 3.830321137433406,
+      "learning_rate": 9.965019209228003e-07,
+      "loss": 0.2869,
+      "step": 2544
+    },
+    {
+      "epoch": 0.486662204799694,
+      "grad_norm": 3.07998647414371,
+      "learning_rate": 9.965518641883848e-07,
+      "loss": 0.167,
+      "step": 2545
+    },
+    {
+      "epoch": 0.4868534276699493,
+      "grad_norm": 2.4268366042918226,
+      "learning_rate": 9.966017878337506e-07,
+      "loss": 0.0831,
+      "step": 2546
+    },
+    {
+      "epoch": 0.4870446505402046,
+      "grad_norm": 3.6148006729319526,
+      "learning_rate": 9.96651691874307e-07,
+      "loss": 0.1464,
+      "step": 2547
+    },
+    {
+      "epoch": 0.4872358734104599,
+      "grad_norm": 4.179106541838472,
+      "learning_rate": 9.967015763254459e-07,
+      "loss": 0.0872,
+      "step": 2548
+    },
+    {
+      "epoch": 0.4874270962807152,
+      "grad_norm": 4.5252154978657115,
+      "learning_rate": 9.967514412025403e-07,
+      "loss": 0.125,
+      "step": 2549
+    },
+    {
+      "epoch": 0.4876183191509705,
+      "grad_norm": 6.08975229445439,
+      "learning_rate": 9.96801286520945e-07,
+      "loss": 0.1283,
+      "step": 2550
+    },
+    {
+      "epoch": 0.48780954202122573,
+      "grad_norm": 3.50899738783697,
+      "learning_rate": 9.968511122959978e-07,
+      "loss": 0.3033,
+      "step": 2551
+    },
+    {
+      "epoch": 0.48800076489148103,
+      "grad_norm": 9.103219571524287,
+      "learning_rate": 9.969009185430177e-07,
+      "loss": 0.7004,
+      "step": 2552
+    },
+    {
+      "epoch": 0.4881919877617363,
+      "grad_norm": 3.735579292486195,
+      "learning_rate": 9.969507052773054e-07,
+      "loss": 0.5176,
+      "step": 2553
+    },
+    {
+      "epoch": 0.4883832106319916,
+      "grad_norm": 4.9890308220225315,
+      "learning_rate": 9.970004725141445e-07,
+      "loss": 0.4361,
+      "step": 2554
+    },
+    {
+      "epoch": 0.4885744335022469,
+      "grad_norm": 3.3763532398063996,
+      "learning_rate": 9.970502202687998e-07,
+      "loss": 0.2752,
+      "step": 2555
+    },
+    {
+      "epoch": 0.48876565637250213,
+      "grad_norm": 6.200203849148518,
+      "learning_rate": 9.970999485565186e-07,
+      "loss": 0.5555,
+      "step": 2556
+    },
+    {
+      "epoch": 0.48895687924275744,
+      "grad_norm": 3.5181459331855383,
+      "learning_rate": 9.971496573925307e-07,
+      "loss": 0.2859,
+      "step": 2557
+    },
+    {
+      "epoch": 0.48914810211301274,
+      "grad_norm": 4.292509101231501,
+      "learning_rate": 9.971993467920473e-07,
+      "loss": 0.2636,
+      "step": 2558
+    },
+    {
+      "epoch": 0.489339324983268,
+      "grad_norm": 3.290310656829355,
+      "learning_rate": 9.972490167702616e-07,
+      "loss": 0.155,
+      "step": 2559
+    },
+    {
+      "epoch": 0.4895305478535233,
+      "grad_norm": 5.428449267253361,
+      "learning_rate": 9.972986673423503e-07,
+      "loss": 0.1154,
+      "step": 2560
+    },
+    {
+      "epoch": 0.48972177072377854,
+      "grad_norm": 3.65271033421,
+      "learning_rate": 9.97348298523471e-07,
+      "loss": 0.0702,
+      "step": 2561
+    },
+    {
+      "epoch": 0.48991299359403384,
+      "grad_norm": 4.291835645787152,
+      "learning_rate": 9.97397910328764e-07,
+      "loss": 0.1518,
+      "step": 2562
+    },
+    {
+      "epoch": 0.49010421646428914,
+      "grad_norm": 7.86440909152663,
+      "learning_rate": 9.97447502773352e-07,
+      "loss": 0.1984,
+      "step": 2563
+    },
+    {
+      "epoch": 0.4902954393345444,
+      "grad_norm": 4.887290329594413,
+      "learning_rate": 9.9749707587234e-07,
+      "loss": 0.4265,
+      "step": 2564
+    },
+    {
+      "epoch": 0.4904866622047997,
+      "grad_norm": 4.542561480830932,
+      "learning_rate": 9.975466296408152e-07,
+      "loss": 0.3845,
+      "step": 2565
+    },
+    {
+      "epoch": 0.490677885075055,
+      "grad_norm": 3.011605385552864,
+      "learning_rate": 9.97596164093847e-07,
+      "loss": 0.281,
+      "step": 2566
+    },
+    {
+      "epoch": 0.49086910794531025,
+      "grad_norm": 3.3910135494806313,
+      "learning_rate": 9.976456792464876e-07,
+      "loss": 0.1861,
+      "step": 2567
+    },
+    {
+      "epoch": 0.49106033081556555,
+      "grad_norm": 2.6983914246018847,
+      "learning_rate": 9.976951751137713e-07,
+      "loss": 0.302,
+      "step": 2568
+    },
+    {
+      "epoch": 0.4912515536858208,
+      "grad_norm": 2.5350874081802894,
+      "learning_rate": 9.977446517107147e-07,
+      "loss": 0.1575,
+      "step": 2569
+    },
+    {
+      "epoch": 0.4914427765560761,
+      "grad_norm": 3.5551408751054723,
+      "learning_rate": 9.977941090523173e-07,
+      "loss": 0.0853,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4916339994263314,
+      "grad_norm": 2.4867077318163755,
+      "learning_rate": 9.97843547153561e-07,
+      "loss": 0.1223,
+      "step": 2571
+    },
+    {
+      "epoch": 0.49182522229658665,
+      "grad_norm": 2.059251133807186,
+      "learning_rate": 9.978929660294102e-07,
+      "loss": 0.1032,
+      "step": 2572
+    },
+    {
+      "epoch": 0.49201644516684195,
+      "grad_norm": 2.6898554748226435,
+      "learning_rate": 9.979423656948112e-07,
+      "loss": 0.114,
+      "step": 2573
+    },
+    {
+      "epoch": 0.49220766803709726,
+      "grad_norm": 4.19354290287426,
+      "learning_rate": 9.97991746164694e-07,
+      "loss": 0.0715,
+      "step": 2574
+    },
+    {
+      "epoch": 0.4923988909073525,
+      "grad_norm": 5.1017858856504805,
+      "learning_rate": 9.980411074539704e-07,
+      "loss": 0.0904,
+      "step": 2575
+    },
+    {
+      "epoch": 0.4925901137776078,
+      "grad_norm": 8.261400148998655,
+      "learning_rate": 9.98090449577535e-07,
+      "loss": 0.4373,
+      "step": 2576
+    },
+    {
+      "epoch": 0.4927813366478631,
+      "grad_norm": 5.949468495968483,
+      "learning_rate": 9.98139772550265e-07,
+      "loss": 0.5916,
+      "step": 2577
+    },
+    {
+      "epoch": 0.49297255951811836,
+      "grad_norm": 6.482681972599098,
+      "learning_rate": 9.98189076387021e-07,
+      "loss": 0.5595,
+      "step": 2578
+    },
+    {
+      "epoch": 0.49316378238837366,
+      "grad_norm": 1.6617695228987404,
+      "learning_rate": 9.982383611026453e-07,
+      "loss": 0.1056,
+      "step": 2579
+    },
+    {
+      "epoch": 0.4933550052586289,
+      "grad_norm": 4.725241938457027,
+      "learning_rate": 9.982876267119632e-07,
+      "loss": 0.5067,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4935462281288842,
+      "grad_norm": 4.437968269678243,
+      "learning_rate": 9.983368732297834e-07,
+      "loss": 0.2558,
+      "step": 2581
+    },
+    {
+      "epoch": 0.4937374509991395,
+      "grad_norm": 1.8870798717064698,
+      "learning_rate": 9.983861006708963e-07,
+      "loss": 0.1861,
+      "step": 2582
+    },
+    {
+      "epoch": 0.49392867386939476,
+      "grad_norm": 3.4802781726109826,
+      "learning_rate": 9.984353090500763e-07,
+      "loss": 0.1821,
+      "step": 2583
+    },
+    {
+      "epoch": 0.49411989673965007,
+      "grad_norm": 2.85993476766607,
+      "learning_rate": 9.984844983820796e-07,
+      "loss": 0.1933,
+      "step": 2584
+    },
+    {
+      "epoch": 0.49431111960990537,
+      "grad_norm": 4.700819090224442,
+      "learning_rate": 9.985336686816462e-07,
+      "loss": 0.2313,
+      "step": 2585
+    },
+    {
+      "epoch": 0.4945023424801606,
+      "grad_norm": 5.4452501509612,
+      "learning_rate": 9.985828199634982e-07,
+      "loss": 0.1628,
+      "step": 2586
+    },
+    {
+      "epoch": 0.4946935653504159,
+      "grad_norm": 4.415667270854929,
+      "learning_rate": 9.986319522423411e-07,
+      "loss": 0.0766,
+      "step": 2587
+    },
+    {
+      "epoch": 0.49488478822067117,
+      "grad_norm": 8.452879953270042,
+      "learning_rate": 9.986810655328632e-07,
+      "loss": 0.2904,
+      "step": 2588
+    },
+    {
+      "epoch": 0.49507601109092647,
+      "grad_norm": 3.4089507149277827,
+      "learning_rate": 9.987301598497357e-07,
+      "loss": 0.1595,
+      "step": 2589
+    },
+    {
+      "epoch": 0.4952672339611818,
+      "grad_norm": 6.075469939968025,
+      "learning_rate": 9.98779235207613e-07,
+      "loss": 0.3801,
+      "step": 2590
+    },
+    {
+      "epoch": 0.495458456831437,
+      "grad_norm": 3.0631250794540485,
+      "learning_rate": 9.988282916211325e-07,
+      "loss": 0.2714,
+      "step": 2591
+    },
+    {
+      "epoch": 0.4956496797016923,
+      "grad_norm": 2.2642589791856844,
+      "learning_rate": 9.988773291049147e-07,
+      "loss": 0.2856,
+      "step": 2592
+    },
+    {
+      "epoch": 0.49584090257194763,
+      "grad_norm": 3.207875669129611,
+      "learning_rate": 9.989263476735626e-07,
+      "loss": 0.5664,
+      "step": 2593
+    },
+    {
+      "epoch": 0.4960321254422029,
+      "grad_norm": 4.55992701890333,
+      "learning_rate": 9.989753473416632e-07,
+      "loss": 0.3051,
+      "step": 2594
+    },
+    {
+      "epoch": 0.4962233483124582,
+      "grad_norm": 4.231866184823192,
+      "learning_rate": 9.990243281237859e-07,
+      "loss": 0.2275,
+      "step": 2595
+    },
+    {
+      "epoch": 0.4964145711827134,
+      "grad_norm": 2.5521803753107535,
+      "learning_rate": 9.990732900344835e-07,
+      "loss": 0.1332,
+      "step": 2596
+    },
+    {
+      "epoch": 0.49660579405296873,
+      "grad_norm": 3.8597251331717515,
+      "learning_rate": 9.991222330882926e-07,
+      "loss": 0.1348,
+      "step": 2597
+    },
+    {
+      "epoch": 0.49679701692322403,
+      "grad_norm": 3.071405968313597,
+      "learning_rate": 9.99171157299732e-07,
+      "loss": 0.0984,
+      "step": 2598
+    },
+    {
+      "epoch": 0.4969882397934793,
+      "grad_norm": 2.3723196665781257,
+      "learning_rate": 9.992200626833043e-07,
+      "loss": 0.0766,
+      "step": 2599
+    },
+    {
+      "epoch": 0.4971794626637346,
+      "grad_norm": 6.710988942705191,
+      "learning_rate": 9.992689492534952e-07,
+      "loss": 0.1874,
+      "step": 2600
+    },
+    {
+      "epoch": 0.4973706855339899,
+      "grad_norm": 11.519810075253927,
+      "learning_rate": 9.993178170247739e-07,
+      "loss": 0.2693,
+      "step": 2601
+    },
+    {
+      "epoch": 0.49756190840424513,
+      "grad_norm": 6.438803827782457,
+      "learning_rate": 9.993666660115926e-07,
+      "loss": 0.3396,
+      "step": 2602
+    },
+    {
+      "epoch": 0.49775313127450044,
+      "grad_norm": 3.6591528430093514,
+      "learning_rate": 9.994154962283872e-07,
+      "loss": 0.3824,
+      "step": 2603
+    },
+    {
+      "epoch": 0.4979443541447557,
+      "grad_norm": 2.372580048282935,
+      "learning_rate": 9.994643076895767e-07,
+      "loss": 0.2865,
+      "step": 2604
+    },
+    {
+      "epoch": 0.498135577015011,
+      "grad_norm": 4.745469593089345,
+      "learning_rate": 9.995131004095633e-07,
+      "loss": 0.4305,
+      "step": 2605
+    },
+    {
+      "epoch": 0.4983267998852663,
+      "grad_norm": 3.5624506093667105,
+      "learning_rate": 9.995618744027332e-07,
+      "loss": 0.1393,
+      "step": 2606
+    },
+    {
+      "epoch": 0.49851802275552154,
+      "grad_norm": 1.6911874678076146,
+      "learning_rate": 9.996106296834557e-07,
+      "loss": 0.1633,
+      "step": 2607
+    },
+    {
+      "epoch": 0.49870924562577684,
+      "grad_norm": 4.51005976838222,
+      "learning_rate": 9.996593662660832e-07,
+      "loss": 0.3819,
+      "step": 2608
+    },
+    {
+      "epoch": 0.49890046849603215,
+      "grad_norm": 3.331371890588152,
+      "learning_rate": 9.997080841649525e-07,
+      "loss": 0.1415,
+      "step": 2609
+    },
+    {
+      "epoch": 0.4990916913662874,
+      "grad_norm": 6.744166078362046,
+      "learning_rate": 9.997567833943832e-07,
+      "loss": 0.4676,
+      "step": 2610
+    },
+    {
+      "epoch": 0.4992829142365427,
+      "grad_norm": 6.948522934968496,
+      "learning_rate": 9.998054639686787e-07,
+      "loss": 0.3421,
+      "step": 2611
+    },
+    {
+      "epoch": 0.499474137106798,
+      "grad_norm": 3.2047274419191636,
+      "learning_rate": 9.998541259021257e-07,
+      "loss": 0.1104,
+      "step": 2612
+    },
+    {
+      "epoch": 0.49966535997705325,
+      "grad_norm": 6.575089649125932,
+      "learning_rate": 9.99902769208995e-07,
+      "loss": 0.1345,
+      "step": 2613
+    },
+    {
+      "epoch": 0.49985658284730855,
+      "grad_norm": 4.240535294912936,
+      "learning_rate": 9.999513939035403e-07,
+      "loss": 0.2357,
+      "step": 2614
+    },
+    {
+      "epoch": 0.5000478057175638,
+      "grad_norm": 4.293721987703726,
+      "learning_rate": 1e-06,
+      "loss": 0.3057,
+      "step": 2615
+    },
+    {
+      "epoch": 0.5002390285878191,
+      "grad_norm": 1.5389530510882974,
+      "learning_rate": 1e-06,
+      "loss": 0.2421,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5004302514580744,
+      "grad_norm": 3.768886056345149,
+      "learning_rate": 1e-06,
+      "loss": 0.191,
+      "step": 2617
+    },
+    {
+      "epoch": 0.5006214743283297,
+      "grad_norm": 2.481996081051741,
+      "learning_rate": 1e-06,
+      "loss": 0.1512,
+      "step": 2618
+    },
+    {
+      "epoch": 0.5008126971985849,
+      "grad_norm": 1.6935317711813411,
+      "learning_rate": 1e-06,
+      "loss": 0.22,
+      "step": 2619
+    },
+    {
+      "epoch": 0.5010039200688402,
+      "grad_norm": 1.8653071683889593,
+      "learning_rate": 1e-06,
+      "loss": 0.1881,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5011951429390955,
+      "grad_norm": 3.699321096871531,
+      "learning_rate": 1e-06,
+      "loss": 0.0715,
+      "step": 2621
+    },
+    {
+      "epoch": 0.5013863658093508,
+      "grad_norm": 4.635334531363553,
+      "learning_rate": 1e-06,
+      "loss": 0.1899,
+      "step": 2622
+    },
+    {
+      "epoch": 0.5015775886796061,
+      "grad_norm": 3.894027386577366,
+      "learning_rate": 1e-06,
+      "loss": 0.1044,
+      "step": 2623
+    },
+    {
+      "epoch": 0.5017688115498614,
+      "grad_norm": 4.761269902643141,
+      "learning_rate": 1e-06,
+      "loss": 0.1513,
+      "step": 2624
+    },
+    {
+      "epoch": 0.5019600344201166,
+      "grad_norm": 6.700222569654694,
+      "learning_rate": 1e-06,
+      "loss": 0.0855,
+      "step": 2625
+    },
+    {
+      "epoch": 0.5021512572903719,
+      "grad_norm": 6.423677882858434,
+      "learning_rate": 1e-06,
+      "loss": 0.2757,
+      "step": 2626
+    },
+    {
+      "epoch": 0.5023424801606272,
+      "grad_norm": 4.494239086546355,
+      "learning_rate": 1e-06,
+      "loss": 0.5412,
+      "step": 2627
+    },
+    {
+      "epoch": 0.5025337030308825,
+      "grad_norm": 4.742480298827358,
+      "learning_rate": 1e-06,
+      "loss": 0.1525,
+      "step": 2628
+    },
+    {
+      "epoch": 0.5027249259011378,
+      "grad_norm": 2.8158518844795704,
+      "learning_rate": 1e-06,
+      "loss": 0.2352,
+      "step": 2629
+    },
+    {
+      "epoch": 0.502916148771393,
+      "grad_norm": 2.6190224662828006,
+      "learning_rate": 1e-06,
+      "loss": 0.2122,
+      "step": 2630
+    },
+    {
+      "epoch": 0.5031073716416483,
+      "grad_norm": 2.6360739590655595,
+      "learning_rate": 1e-06,
+      "loss": 0.3462,
+      "step": 2631
+    },
+    {
+      "epoch": 0.5032985945119036,
+      "grad_norm": 2.755044039254843,
+      "learning_rate": 1e-06,
+      "loss": 0.0907,
+      "step": 2632
+    },
+    {
+      "epoch": 0.5034898173821589,
+      "grad_norm": 3.238942924239248,
+      "learning_rate": 1e-06,
+      "loss": 0.2096,
+      "step": 2633
+    },
+    {
+      "epoch": 0.5036810402524142,
+      "grad_norm": 3.175174894547057,
+      "learning_rate": 1e-06,
+      "loss": 0.1243,
+      "step": 2634
+    },
+    {
+      "epoch": 0.5038722631226694,
+      "grad_norm": 2.7886523204266953,
+      "learning_rate": 1e-06,
+      "loss": 0.088,
+      "step": 2635
+    },
+    {
+      "epoch": 0.5040634859929247,
+      "grad_norm": 1.6730343373087377,
+      "learning_rate": 1e-06,
+      "loss": 0.0233,
+      "step": 2636
+    },
+    {
+      "epoch": 0.50425470886318,
+      "grad_norm": 5.01960306672846,
+      "learning_rate": 1e-06,
+      "loss": 0.095,
+      "step": 2637
+    },
+    {
+      "epoch": 0.5044459317334353,
+      "grad_norm": 10.099524579331229,
+      "learning_rate": 1e-06,
+      "loss": 0.1049,
+      "step": 2638
+    },
+    {
+      "epoch": 0.5046371546036906,
+      "grad_norm": 3.2978852518544994,
+      "learning_rate": 1e-06,
+      "loss": 0.4835,
+      "step": 2639
+    },
+    {
+      "epoch": 0.5048283774739459,
+      "grad_norm": 4.187603849575144,
+      "learning_rate": 1e-06,
+      "loss": 0.5152,
+      "step": 2640
+    },
+    {
+      "epoch": 0.5050196003442011,
+      "grad_norm": 4.908067888855179,
+      "learning_rate": 1e-06,
+      "loss": 0.4818,
+      "step": 2641
+    },
+    {
+      "epoch": 0.5052108232144564,
+      "grad_norm": 4.068819500729493,
+      "learning_rate": 1e-06,
+      "loss": 0.5641,
+      "step": 2642
+    },
+    {
+      "epoch": 0.5054020460847117,
+      "grad_norm": 3.9162832572978683,
+      "learning_rate": 1e-06,
+      "loss": 0.2855,
+      "step": 2643
+    },
+    {
+      "epoch": 0.505593268954967,
+      "grad_norm": 2.8555677738579752,
+      "learning_rate": 1e-06,
+      "loss": 0.2098,
+      "step": 2644
+    },
+    {
+      "epoch": 0.5057844918252223,
+      "grad_norm": 5.355499200713485,
+      "learning_rate": 1e-06,
+      "loss": 0.335,
+      "step": 2645
+    },
+    {
+      "epoch": 0.5059757146954775,
+      "grad_norm": 3.813456431004801,
+      "learning_rate": 1e-06,
+      "loss": 0.2268,
+      "step": 2646
+    },
+    {
+      "epoch": 0.5061669375657328,
+      "grad_norm": 2.2518961651919374,
+      "learning_rate": 1e-06,
+      "loss": 0.1571,
+      "step": 2647
+    },
+    {
+      "epoch": 0.5063581604359881,
+      "grad_norm": 2.3661574335224658,
+      "learning_rate": 1e-06,
+      "loss": 0.1401,
+      "step": 2648
+    },
+    {
+      "epoch": 0.5065493833062434,
+      "grad_norm": 3.1349461773841054,
+      "learning_rate": 1e-06,
+      "loss": 0.0832,
+      "step": 2649
+    },
+    {
+      "epoch": 0.5067406061764987,
+      "grad_norm": 2.09977072872009,
+      "learning_rate": 1e-06,
+      "loss": 0.051,
+      "step": 2650
+    },
+    {
+      "epoch": 0.506931829046754,
+      "grad_norm": 4.212640025361337,
+      "learning_rate": 1e-06,
+      "loss": 0.2322,
+      "step": 2651
+    },
+    {
+      "epoch": 0.5071230519170092,
+      "grad_norm": 4.101054423516415,
+      "learning_rate": 1e-06,
+      "loss": 0.3652,
+      "step": 2652
+    },
+    {
+      "epoch": 0.5073142747872645,
+      "grad_norm": 3.2145847665845766,
+      "learning_rate": 1e-06,
+      "loss": 0.3306,
+      "step": 2653
+    },
+    {
+      "epoch": 0.5075054976575198,
+      "grad_norm": 3.1127923177082653,
+      "learning_rate": 1e-06,
+      "loss": 0.2188,
+      "step": 2654
+    },
+    {
+      "epoch": 0.5076967205277751,
+      "grad_norm": 4.218570281545267,
+      "learning_rate": 1e-06,
+      "loss": 0.3952,
+      "step": 2655
+    },
+    {
+      "epoch": 0.5078879433980305,
+      "grad_norm": 2.074474943901998,
+      "learning_rate": 1e-06,
+      "loss": 0.142,
+      "step": 2656
+    },
+    {
+      "epoch": 0.5080791662682856,
+      "grad_norm": 5.9852326179495146,
+      "learning_rate": 1e-06,
+      "loss": 0.1741,
+      "step": 2657
+    },
+    {
+      "epoch": 0.508270389138541,
+      "grad_norm": 3.35099563606545,
+      "learning_rate": 1e-06,
+      "loss": 0.2082,
+      "step": 2658
+    },
+    {
+      "epoch": 0.5084616120087962,
+      "grad_norm": 2.980887247404417,
+      "learning_rate": 1e-06,
+      "loss": 0.0836,
+      "step": 2659
+    },
+    {
+      "epoch": 0.5086528348790516,
+      "grad_norm": 5.461671444275091,
+      "learning_rate": 1e-06,
+      "loss": 0.2421,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5088440577493069,
+      "grad_norm": 2.797859657901683,
+      "learning_rate": 1e-06,
+      "loss": 0.0936,
+      "step": 2661
+    },
+    {
+      "epoch": 0.509035280619562,
+      "grad_norm": 5.1704962699189805,
+      "learning_rate": 1e-06,
+      "loss": 0.131,
+      "step": 2662
+    },
+    {
+      "epoch": 0.5092265034898174,
+      "grad_norm": 6.174406435465863,
+      "learning_rate": 1e-06,
+      "loss": 0.1388,
+      "step": 2663
+    },
+    {
+      "epoch": 0.5094177263600727,
+      "grad_norm": 8.74260115475777,
+      "learning_rate": 1e-06,
+      "loss": 0.5239,
+      "step": 2664
+    },
+    {
+      "epoch": 0.509608949230328,
+      "grad_norm": 2.5645216433689657,
+      "learning_rate": 1e-06,
+      "loss": 0.3126,
+      "step": 2665
+    },
+    {
+      "epoch": 0.5098001721005833,
+      "grad_norm": 3.650908391082708,
+      "learning_rate": 1e-06,
+      "loss": 0.4167,
+      "step": 2666
+    },
+    {
+      "epoch": 0.5099913949708386,
+      "grad_norm": 2.857226622579721,
+      "learning_rate": 1e-06,
+      "loss": 0.096,
+      "step": 2667
+    },
+    {
+      "epoch": 0.5101826178410938,
+      "grad_norm": 3.7256230543694984,
+      "learning_rate": 1e-06,
+      "loss": 0.3193,
+      "step": 2668
+    },
+    {
+      "epoch": 0.5103738407113491,
+      "grad_norm": 2.812183192741859,
+      "learning_rate": 1e-06,
+      "loss": 0.0828,
+      "step": 2669
+    },
+    {
+      "epoch": 0.5105650635816044,
+      "grad_norm": 2.6937601677864222,
+      "learning_rate": 1e-06,
+      "loss": 0.1874,
+      "step": 2670
+    },
+    {
+      "epoch": 0.5107562864518597,
+      "grad_norm": 5.184521291001292,
+      "learning_rate": 1e-06,
+      "loss": 0.2352,
+      "step": 2671
+    },
+    {
+      "epoch": 0.510947509322115,
+      "grad_norm": 2.7151362638895375,
+      "learning_rate": 1e-06,
+      "loss": 0.0666,
+      "step": 2672
+    },
+    {
+      "epoch": 0.5111387321923702,
+      "grad_norm": 4.690681293513177,
+      "learning_rate": 1e-06,
+      "loss": 0.1315,
+      "step": 2673
+    },
+    {
+      "epoch": 0.5113299550626255,
+      "grad_norm": 4.049840128062614,
+      "learning_rate": 1e-06,
+      "loss": 0.1154,
+      "step": 2674
+    },
+    {
+      "epoch": 0.5115211779328808,
+      "grad_norm": 14.349396036975792,
+      "learning_rate": 1e-06,
+      "loss": 0.2024,
+      "step": 2675
+    },
+    {
+      "epoch": 0.5117124008031361,
+      "grad_norm": 5.492044590790523,
+      "learning_rate": 1e-06,
+      "loss": 0.264,
+      "step": 2676
+    },
+    {
+      "epoch": 0.5119036236733914,
+      "grad_norm": 5.766142038872366,
+      "learning_rate": 1e-06,
+      "loss": 0.4151,
+      "step": 2677
+    },
+    {
+      "epoch": 0.5120948465436466,
+      "grad_norm": 2.570192467573885,
+      "learning_rate": 1e-06,
+      "loss": 0.4422,
+      "step": 2678
+    },
+    {
+      "epoch": 0.5122860694139019,
+      "grad_norm": 2.324192591928694,
+      "learning_rate": 1e-06,
+      "loss": 0.2957,
+      "step": 2679
+    },
+    {
+      "epoch": 0.5124772922841572,
+      "grad_norm": 2.8746034099871016,
+      "learning_rate": 1e-06,
+      "loss": 0.1598,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5126685151544125,
+      "grad_norm": 1.9210590398196137,
+      "learning_rate": 1e-06,
+      "loss": 0.256,
+      "step": 2681
+    },
+    {
+      "epoch": 0.5128597380246678,
+      "grad_norm": 2.701307181653827,
+      "learning_rate": 1e-06,
+      "loss": 0.1516,
+      "step": 2682
+    },
+    {
+      "epoch": 0.5130509608949231,
+      "grad_norm": 2.516629794358039,
+      "learning_rate": 1e-06,
+      "loss": 0.1371,
+      "step": 2683
+    },
+    {
+      "epoch": 0.5132421837651783,
+      "grad_norm": 3.4392315751508575,
+      "learning_rate": 1e-06,
+      "loss": 0.1948,
+      "step": 2684
+    },
+    {
+      "epoch": 0.5134334066354336,
+      "grad_norm": 3.229355902920812,
+      "learning_rate": 1e-06,
+      "loss": 0.248,
+      "step": 2685
+    },
+    {
+      "epoch": 0.5136246295056889,
+      "grad_norm": 3.049194392137051,
+      "learning_rate": 1e-06,
+      "loss": 0.1012,
+      "step": 2686
+    },
+    {
+      "epoch": 0.5138158523759442,
+      "grad_norm": 2.159217009443949,
+      "learning_rate": 1e-06,
+      "loss": 0.0361,
+      "step": 2687
+    },
+    {
+      "epoch": 0.5140070752461995,
+      "grad_norm": 13.053989686242119,
+      "learning_rate": 1e-06,
+      "loss": 0.1415,
+      "step": 2688
+    },
+    {
+      "epoch": 0.5141982981164547,
+      "grad_norm": 4.581535657274953,
+      "learning_rate": 1e-06,
+      "loss": 0.3931,
+      "step": 2689
+    },
+    {
+      "epoch": 0.51438952098671,
+      "grad_norm": 4.286019223581727,
+      "learning_rate": 1e-06,
+      "loss": 0.7403,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5145807438569653,
+      "grad_norm": 3.2767429460382793,
+      "learning_rate": 1e-06,
+      "loss": 0.2039,
+      "step": 2691
+    },
+    {
+      "epoch": 0.5147719667272206,
+      "grad_norm": 2.239693879940356,
+      "learning_rate": 1e-06,
+      "loss": 0.076,
+      "step": 2692
+    },
+    {
+      "epoch": 0.5149631895974759,
+      "grad_norm": 2.3705946571425702,
+      "learning_rate": 1e-06,
+      "loss": 0.311,
+      "step": 2693
+    },
+    {
+      "epoch": 0.5151544124677312,
+      "grad_norm": 3.9382973726438606,
+      "learning_rate": 1e-06,
+      "loss": 0.1432,
+      "step": 2694
+    },
+    {
+      "epoch": 0.5153456353379864,
+      "grad_norm": 2.4701846818954447,
+      "learning_rate": 1e-06,
+      "loss": 0.0742,
+      "step": 2695
+    },
+    {
+      "epoch": 0.5155368582082417,
+      "grad_norm": 2.9533305424604586,
+      "learning_rate": 1e-06,
+      "loss": 0.2524,
+      "step": 2696
+    },
+    {
+      "epoch": 0.515728081078497,
+      "grad_norm": 2.5343113036329856,
+      "learning_rate": 1e-06,
+      "loss": 0.112,
+      "step": 2697
+    },
+    {
+      "epoch": 0.5159193039487523,
+      "grad_norm": 2.6358921592082605,
+      "learning_rate": 1e-06,
+      "loss": 0.0676,
+      "step": 2698
+    },
+    {
+      "epoch": 0.5161105268190076,
+      "grad_norm": 2.654001967641393,
+      "learning_rate": 1e-06,
+      "loss": 0.107,
+      "step": 2699
+    },
+    {
+      "epoch": 0.5163017496892628,
+      "grad_norm": 1.8354685732386042,
+      "learning_rate": 1e-06,
+      "loss": 0.0404,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5164929725595181,
+      "grad_norm": 5.966706410423057,
+      "learning_rate": 1e-06,
+      "loss": 0.481,
+      "step": 2701
+    },
+    {
+      "epoch": 0.5166841954297734,
+      "grad_norm": 6.6603595780913105,
+      "learning_rate": 1e-06,
+      "loss": 0.6519,
+      "step": 2702
+    },
+    {
+      "epoch": 0.5168754183000287,
+      "grad_norm": 3.766500284239245,
+      "learning_rate": 1e-06,
+      "loss": 0.3273,
+      "step": 2703
+    },
+    {
+      "epoch": 0.517066641170284,
+      "grad_norm": 14.071858864914134,
+      "learning_rate": 1e-06,
+      "loss": 0.3651,
+      "step": 2704
+    },
+    {
+      "epoch": 0.5172578640405392,
+      "grad_norm": 2.50420531392755,
+      "learning_rate": 1e-06,
+      "loss": 0.3783,
+      "step": 2705
+    },
+    {
+      "epoch": 0.5174490869107945,
+      "grad_norm": 3.6047265972105578,
+      "learning_rate": 1e-06,
+      "loss": 0.2922,
+      "step": 2706
+    },
+    {
+      "epoch": 0.5176403097810498,
+      "grad_norm": 4.420990656654961,
+      "learning_rate": 1e-06,
+      "loss": 0.2491,
+      "step": 2707
+    },
+    {
+      "epoch": 0.5178315326513051,
+      "grad_norm": 2.968093317585303,
+      "learning_rate": 1e-06,
+      "loss": 0.183,
+      "step": 2708
+    },
+    {
+      "epoch": 0.5180227555215604,
+      "grad_norm": 3.0748718421095127,
+      "learning_rate": 1e-06,
+      "loss": 0.0871,
+      "step": 2709
+    },
+    {
+      "epoch": 0.5182139783918157,
+      "grad_norm": 7.349706952747148,
+      "learning_rate": 1e-06,
+      "loss": 0.125,
+      "step": 2710
+    },
+    {
+      "epoch": 0.5184052012620709,
+      "grad_norm": 3.885782069528619,
+      "learning_rate": 1e-06,
+      "loss": 0.0502,
+      "step": 2711
+    },
+    {
+      "epoch": 0.5185964241323262,
+      "grad_norm": 1.8392920183134573,
+      "learning_rate": 1e-06,
+      "loss": 0.0332,
+      "step": 2712
+    },
+    {
+      "epoch": 0.5187876470025815,
+      "grad_norm": 6.021229380041115,
+      "learning_rate": 1e-06,
+      "loss": 0.1028,
+      "step": 2713
+    },
+    {
+      "epoch": 0.5189788698728368,
+      "grad_norm": 6.275864451163282,
+      "learning_rate": 1e-06,
+      "loss": 0.5497,
+      "step": 2714
+    },
+    {
+      "epoch": 0.5191700927430921,
+      "grad_norm": 4.196352355635947,
+      "learning_rate": 1e-06,
+      "loss": 0.3509,
+      "step": 2715
+    },
+    {
+      "epoch": 0.5193613156133473,
+      "grad_norm": 3.3327741789727283,
+      "learning_rate": 1e-06,
+      "loss": 0.2835,
+      "step": 2716
+    },
+    {
+      "epoch": 0.5195525384836026,
+      "grad_norm": 3.0549310065187796,
+      "learning_rate": 1e-06,
+      "loss": 0.3485,
+      "step": 2717
+    },
+    {
+      "epoch": 0.5197437613538579,
+      "grad_norm": 3.8180029505163255,
+      "learning_rate": 1e-06,
+      "loss": 0.1553,
+      "step": 2718
+    },
+    {
+      "epoch": 0.5199349842241132,
+      "grad_norm": 2.2805935686014913,
+      "learning_rate": 1e-06,
+      "loss": 0.1818,
+      "step": 2719
+    },
+    {
+      "epoch": 0.5201262070943685,
+      "grad_norm": 4.370693594144582,
+      "learning_rate": 1e-06,
+      "loss": 0.1695,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5203174299646238,
+      "grad_norm": 3.5383631539462,
+      "learning_rate": 1e-06,
+      "loss": 0.1358,
+      "step": 2721
+    },
+    {
+      "epoch": 0.520508652834879,
+      "grad_norm": 3.9684907122639452,
+      "learning_rate": 1e-06,
+      "loss": 0.2105,
+      "step": 2722
+    },
+    {
+      "epoch": 0.5206998757051343,
+      "grad_norm": 9.292443714807806,
+      "learning_rate": 1e-06,
+      "loss": 0.1365,
+      "step": 2723
+    },
+    {
+      "epoch": 0.5208910985753896,
+      "grad_norm": 5.278348227117444,
+      "learning_rate": 1e-06,
+      "loss": 0.1266,
+      "step": 2724
+    },
+    {
+      "epoch": 0.5210823214456449,
+      "grad_norm": 6.874659026967123,
+      "learning_rate": 1e-06,
+      "loss": 0.1264,
+      "step": 2725
+    },
+    {
+      "epoch": 0.5212735443159002,
+      "grad_norm": 6.803043346414409,
+      "learning_rate": 1e-06,
+      "loss": 0.274,
+      "step": 2726
+    },
+    {
+      "epoch": 0.5214647671861554,
+      "grad_norm": 4.961245069205789,
+      "learning_rate": 1e-06,
+      "loss": 0.3109,
+      "step": 2727
+    },
+    {
+      "epoch": 0.5216559900564107,
+      "grad_norm": 5.233516246895443,
+      "learning_rate": 1e-06,
+      "loss": 0.4121,
+      "step": 2728
+    },
+    {
+      "epoch": 0.521847212926666,
+      "grad_norm": 4.585977427424142,
+      "learning_rate": 1e-06,
+      "loss": 0.2159,
+      "step": 2729
+    },
+    {
+      "epoch": 0.5220384357969213,
+      "grad_norm": 1.5231279107505464,
+      "learning_rate": 1e-06,
+      "loss": 0.2758,
+      "step": 2730
+    },
+    {
+      "epoch": 0.5222296586671766,
+      "grad_norm": 3.390079858790145,
+      "learning_rate": 1e-06,
+      "loss": 0.2439,
+      "step": 2731
+    },
+    {
+      "epoch": 0.5224208815374318,
+      "grad_norm": 1.7193433951021455,
+      "learning_rate": 1e-06,
+      "loss": 0.1457,
+      "step": 2732
+    },
+    {
+      "epoch": 0.5226121044076871,
+      "grad_norm": 2.984037530003393,
+      "learning_rate": 1e-06,
+      "loss": 0.0828,
+      "step": 2733
+    },
+    {
+      "epoch": 0.5228033272779424,
+      "grad_norm": 4.966523637354933,
+      "learning_rate": 1e-06,
+      "loss": 0.1218,
+      "step": 2734
+    },
+    {
+      "epoch": 0.5229945501481977,
+      "grad_norm": 3.614794341129535,
+      "learning_rate": 1e-06,
+      "loss": 0.0887,
+      "step": 2735
+    },
+    {
+      "epoch": 0.523185773018453,
+      "grad_norm": 3.8831535204950254,
+      "learning_rate": 1e-06,
+      "loss": 0.0667,
+      "step": 2736
+    },
+    {
+      "epoch": 0.5233769958887083,
+      "grad_norm": 2.167034545869526,
+      "learning_rate": 1e-06,
+      "loss": 0.0469,
+      "step": 2737
+    },
+    {
+      "epoch": 0.5235682187589635,
+      "grad_norm": 7.091455298507664,
+      "learning_rate": 1e-06,
+      "loss": 0.1952,
+      "step": 2738
+    },
+    {
+      "epoch": 0.5237594416292188,
+      "grad_norm": 7.436543635580745,
+      "learning_rate": 1e-06,
+      "loss": 0.6977,
+      "step": 2739
+    },
+    {
+      "epoch": 0.5239506644994741,
+      "grad_norm": 3.8218627779773784,
+      "learning_rate": 1e-06,
+      "loss": 0.3931,
+      "step": 2740
+    },
+    {
+      "epoch": 0.5241418873697294,
+      "grad_norm": 4.050702380274262,
+      "learning_rate": 1e-06,
+      "loss": 0.2931,
+      "step": 2741
+    },
+    {
+      "epoch": 0.5243331102399847,
+      "grad_norm": 1.9191880710792393,
+      "learning_rate": 1e-06,
+      "loss": 0.1334,
+      "step": 2742
+    },
+    {
+      "epoch": 0.5245243331102399,
+      "grad_norm": 4.468840698175416,
+      "learning_rate": 1e-06,
+      "loss": 0.2162,
+      "step": 2743
+    },
+    {
+      "epoch": 0.5247155559804952,
+      "grad_norm": 6.403451596779266,
+      "learning_rate": 1e-06,
+      "loss": 0.2983,
+      "step": 2744
+    },
+    {
+      "epoch": 0.5249067788507505,
+      "grad_norm": 4.178270559692744,
+      "learning_rate": 1e-06,
+      "loss": 0.1382,
+      "step": 2745
+    },
+    {
+      "epoch": 0.5250980017210058,
+      "grad_norm": 6.0091476643410475,
+      "learning_rate": 1e-06,
+      "loss": 0.1144,
+      "step": 2746
+    },
+    {
+      "epoch": 0.5252892245912612,
+      "grad_norm": 5.437713618850736,
+      "learning_rate": 1e-06,
+      "loss": 0.1221,
+      "step": 2747
+    },
+    {
+      "epoch": 0.5254804474615163,
+      "grad_norm": 5.02763891057589,
+      "learning_rate": 1e-06,
+      "loss": 0.0731,
+      "step": 2748
+    },
+    {
+      "epoch": 0.5256716703317716,
+      "grad_norm": 3.665359495259938,
+      "learning_rate": 1e-06,
+      "loss": 0.0996,
+      "step": 2749
+    },
+    {
+      "epoch": 0.525862893202027,
+      "grad_norm": 6.222600770213286,
+      "learning_rate": 1e-06,
+      "loss": 0.189,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5260541160722823,
+      "grad_norm": 9.360344583412239,
+      "learning_rate": 1e-06,
+      "loss": 0.4123,
+      "step": 2751
+    },
+    {
+      "epoch": 0.5262453389425376,
+      "grad_norm": 3.611902450190534,
+      "learning_rate": 1e-06,
+      "loss": 0.2339,
+      "step": 2752
+    },
+    {
+      "epoch": 0.5264365618127929,
+      "grad_norm": 5.433190051726848,
+      "learning_rate": 1e-06,
+      "loss": 0.6203,
+      "step": 2753
+    },
+    {
+      "epoch": 0.526627784683048,
+      "grad_norm": 2.566939253596164,
+      "learning_rate": 1e-06,
+      "loss": 0.2808,
+      "step": 2754
+    },
+    {
+      "epoch": 0.5268190075533034,
+      "grad_norm": 2.330541802801334,
+      "learning_rate": 1e-06,
+      "loss": 0.1094,
+      "step": 2755
+    },
+    {
+      "epoch": 0.5270102304235587,
+      "grad_norm": 2.0196978207040743,
+      "learning_rate": 1e-06,
+      "loss": 0.3447,
+      "step": 2756
+    },
+    {
+      "epoch": 0.527201453293814,
+      "grad_norm": 3.6050067576690275,
+      "learning_rate": 1e-06,
+      "loss": 0.1413,
+      "step": 2757
+    },
+    {
+      "epoch": 0.5273926761640693,
+      "grad_norm": 3.516696614021921,
+      "learning_rate": 1e-06,
+      "loss": 0.1203,
+      "step": 2758
+    },
+    {
+      "epoch": 0.5275838990343245,
+      "grad_norm": 3.341011137810473,
+      "learning_rate": 1e-06,
+      "loss": 0.2047,
+      "step": 2759
+    },
+    {
+      "epoch": 0.5277751219045798,
+      "grad_norm": 2.6457210326704597,
+      "learning_rate": 1e-06,
+      "loss": 0.1943,
+      "step": 2760
+    },
+    {
+      "epoch": 0.5279663447748351,
+      "grad_norm": 2.868639336739129,
+      "learning_rate": 1e-06,
+      "loss": 0.0666,
+      "step": 2761
+    },
+    {
+      "epoch": 0.5281575676450904,
+      "grad_norm": 2.6176556979766104,
+      "learning_rate": 1e-06,
+      "loss": 0.0551,
+      "step": 2762
+    },
+    {
+      "epoch": 0.5283487905153457,
+      "grad_norm": 7.085024672891303,
+      "learning_rate": 1e-06,
+      "loss": 0.2865,
+      "step": 2763
+    },
+    {
+      "epoch": 0.528540013385601,
+      "grad_norm": 6.636402035004724,
+      "learning_rate": 1e-06,
+      "loss": 0.6295,
+      "step": 2764
+    },
+    {
+      "epoch": 0.5287312362558562,
+      "grad_norm": 3.8663459864090144,
+      "learning_rate": 1e-06,
+      "loss": 0.319,
+      "step": 2765
+    },
+    {
+      "epoch": 0.5289224591261115,
+      "grad_norm": 3.8495824959028,
+      "learning_rate": 1e-06,
+      "loss": 0.4577,
+      "step": 2766
+    },
+    {
+      "epoch": 0.5291136819963668,
+      "grad_norm": 2.5046039150900437,
+      "learning_rate": 1e-06,
+      "loss": 0.1641,
+      "step": 2767
+    },
+    {
+      "epoch": 0.5293049048666221,
+      "grad_norm": 3.5377861904665133,
+      "learning_rate": 1e-06,
+      "loss": 0.3036,
+      "step": 2768
+    },
+    {
+      "epoch": 0.5294961277368774,
+      "grad_norm": 3.87888399945626,
+      "learning_rate": 1e-06,
+      "loss": 0.3282,
+      "step": 2769
+    },
+    {
+      "epoch": 0.5296873506071326,
+      "grad_norm": 3.79725023464117,
+      "learning_rate": 1e-06,
+      "loss": 0.3643,
+      "step": 2770
+    },
+    {
+      "epoch": 0.5298785734773879,
+      "grad_norm": 1.7783437074801385,
+      "learning_rate": 1e-06,
+      "loss": 0.1723,
+      "step": 2771
+    },
+    {
+      "epoch": 0.5300697963476432,
+      "grad_norm": 3.3260931698978773,
+      "learning_rate": 1e-06,
+      "loss": 0.2307,
+      "step": 2772
+    },
+    {
+      "epoch": 0.5302610192178985,
+      "grad_norm": 4.339140717485168,
+      "learning_rate": 1e-06,
+      "loss": 0.1381,
+      "step": 2773
+    },
+    {
+      "epoch": 0.5304522420881538,
+      "grad_norm": 3.7111834314067327,
+      "learning_rate": 1e-06,
+      "loss": 0.0835,
+      "step": 2774
+    },
+    {
+      "epoch": 0.530643464958409,
+      "grad_norm": 9.005357207521454,
+      "learning_rate": 1e-06,
+      "loss": 0.2967,
+      "step": 2775
+    },
+    {
+      "epoch": 0.5308346878286643,
+      "grad_norm": 5.8664199184777175,
+      "learning_rate": 1e-06,
+      "loss": 0.3534,
+      "step": 2776
+    },
+    {
+      "epoch": 0.5310259106989196,
+      "grad_norm": 4.242187050385565,
+      "learning_rate": 1e-06,
+      "loss": 0.3749,
+      "step": 2777
+    },
+    {
+      "epoch": 0.5312171335691749,
+      "grad_norm": 3.4726889300873705,
+      "learning_rate": 1e-06,
+      "loss": 0.4445,
+      "step": 2778
+    },
+    {
+      "epoch": 0.5314083564394302,
+      "grad_norm": 2.327294323086784,
+      "learning_rate": 1e-06,
+      "loss": 0.123,
+      "step": 2779
+    },
+    {
+      "epoch": 0.5315995793096855,
+      "grad_norm": 2.3260709162534545,
+      "learning_rate": 1e-06,
+      "loss": 0.2579,
+      "step": 2780
+    },
+    {
+      "epoch": 0.5317908021799407,
+      "grad_norm": 4.760909552346589,
+      "learning_rate": 1e-06,
+      "loss": 0.3569,
+      "step": 2781
+    },
+    {
+      "epoch": 0.531982025050196,
+      "grad_norm": 1.3296028441971381,
+      "learning_rate": 1e-06,
+      "loss": 0.0747,
+      "step": 2782
+    },
+    {
+      "epoch": 0.5321732479204513,
+      "grad_norm": 1.7566061670114523,
+      "learning_rate": 1e-06,
+      "loss": 0.0814,
+      "step": 2783
+    },
+    {
+      "epoch": 0.5323644707907066,
+      "grad_norm": 4.027696803071898,
+      "learning_rate": 1e-06,
+      "loss": 0.107,
+      "step": 2784
+    },
+    {
+      "epoch": 0.5325556936609619,
+      "grad_norm": 3.3698975881981705,
+      "learning_rate": 1e-06,
+      "loss": 0.1296,
+      "step": 2785
+    },
+    {
+      "epoch": 0.5327469165312171,
+      "grad_norm": 3.5294733219762398,
+      "learning_rate": 1e-06,
+      "loss": 0.0774,
+      "step": 2786
+    },
+    {
+      "epoch": 0.5329381394014724,
+      "grad_norm": 3.925321850063624,
+      "learning_rate": 1e-06,
+      "loss": 0.1048,
+      "step": 2787
+    },
+    {
+      "epoch": 0.5331293622717277,
+      "grad_norm": 7.147806442005341,
+      "learning_rate": 1e-06,
+      "loss": 0.4586,
+      "step": 2788
+    },
+    {
+      "epoch": 0.533320585141983,
+      "grad_norm": 6.029946299957496,
+      "learning_rate": 1e-06,
+      "loss": 0.2669,
+      "step": 2789
+    },
+    {
+      "epoch": 0.5335118080122383,
+      "grad_norm": 4.36976943012289,
+      "learning_rate": 1e-06,
+      "loss": 0.2269,
+      "step": 2790
+    },
+    {
+      "epoch": 0.5337030308824936,
+      "grad_norm": 3.302049237545572,
+      "learning_rate": 1e-06,
+      "loss": 0.1773,
+      "step": 2791
+    },
+    {
+      "epoch": 0.5338942537527488,
+      "grad_norm": 2.7635972515046707,
+      "learning_rate": 1e-06,
+      "loss": 0.1575,
+      "step": 2792
+    },
+    {
+      "epoch": 0.5340854766230041,
+      "grad_norm": 2.850311422900931,
+      "learning_rate": 1e-06,
+      "loss": 0.227,
+      "step": 2793
+    },
+    {
+      "epoch": 0.5342766994932594,
+      "grad_norm": 3.5899995556249316,
+      "learning_rate": 1e-06,
+      "loss": 0.3615,
+      "step": 2794
+    },
+    {
+      "epoch": 0.5344679223635147,
+      "grad_norm": 3.1222163199945534,
+      "learning_rate": 1e-06,
+      "loss": 0.2007,
+      "step": 2795
+    },
+    {
+      "epoch": 0.53465914523377,
+      "grad_norm": 4.8655973870995135,
+      "learning_rate": 1e-06,
+      "loss": 0.3424,
+      "step": 2796
+    },
+    {
+      "epoch": 0.5348503681040252,
+      "grad_norm": 5.64512695548168,
+      "learning_rate": 1e-06,
+      "loss": 0.1198,
+      "step": 2797
+    },
+    {
+      "epoch": 0.5350415909742805,
+      "grad_norm": 2.083988493581919,
+      "learning_rate": 1e-06,
+      "loss": 0.0656,
+      "step": 2798
+    },
+    {
+      "epoch": 0.5352328138445358,
+      "grad_norm": 4.18474226145408,
+      "learning_rate": 1e-06,
+      "loss": 0.0972,
+      "step": 2799
+    },
+    {
+      "epoch": 0.5354240367147911,
+      "grad_norm": 3.3573143605076656,
+      "learning_rate": 1e-06,
+      "loss": 0.0877,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5356152595850464,
+      "grad_norm": 5.99572029069062,
+      "learning_rate": 1e-06,
+      "loss": 0.1708,
+      "step": 2801
+    },
+    {
+      "epoch": 0.5358064824553016,
+      "grad_norm": 6.778171385434442,
+      "learning_rate": 1e-06,
+      "loss": 0.7229,
+      "step": 2802
+    },
+    {
+      "epoch": 0.5359977053255569,
+      "grad_norm": 2.097983336565229,
+      "learning_rate": 1e-06,
+      "loss": 0.1482,
+      "step": 2803
+    },
+    {
+      "epoch": 0.5361889281958122,
+      "grad_norm": 2.8023237293082315,
+      "learning_rate": 1e-06,
+      "loss": 0.2393,
+      "step": 2804
+    },
+    {
+      "epoch": 0.5363801510660675,
+      "grad_norm": 4.315868886234157,
+      "learning_rate": 1e-06,
+      "loss": 0.5796,
+      "step": 2805
+    },
+    {
+      "epoch": 0.5365713739363228,
+      "grad_norm": 3.6526318767951924,
+      "learning_rate": 1e-06,
+      "loss": 0.1448,
+      "step": 2806
+    },
+    {
+      "epoch": 0.5367625968065781,
+      "grad_norm": 2.8492146062337054,
+      "learning_rate": 1e-06,
+      "loss": 0.1753,
+      "step": 2807
+    },
+    {
+      "epoch": 0.5369538196768333,
+      "grad_norm": 2.2212891354867326,
+      "learning_rate": 1e-06,
+      "loss": 0.1686,
+      "step": 2808
+    },
+    {
+      "epoch": 0.5371450425470886,
+      "grad_norm": 2.6320848676338775,
+      "learning_rate": 1e-06,
+      "loss": 0.1031,
+      "step": 2809
+    },
+    {
+      "epoch": 0.5373362654173439,
+      "grad_norm": 4.247386353102286,
+      "learning_rate": 1e-06,
+      "loss": 0.0859,
+      "step": 2810
+    },
+    {
+      "epoch": 0.5375274882875992,
+      "grad_norm": 4.468345357007691,
+      "learning_rate": 1e-06,
+      "loss": 0.0601,
+      "step": 2811
+    },
+    {
+      "epoch": 0.5377187111578545,
+      "grad_norm": 4.092003136582352,
+      "learning_rate": 1e-06,
+      "loss": 0.2063,
+      "step": 2812
+    },
+    {
+      "epoch": 0.5379099340281097,
+      "grad_norm": 5.5388527899079865,
+      "learning_rate": 1e-06,
+      "loss": 0.2434,
+      "step": 2813
+    },
+    {
+      "epoch": 0.538101156898365,
+      "grad_norm": 4.381127779965982,
+      "learning_rate": 1e-06,
+      "loss": 0.3624,
+      "step": 2814
+    },
+    {
+      "epoch": 0.5382923797686203,
+      "grad_norm": 2.9670054881601517,
+      "learning_rate": 1e-06,
+      "loss": 0.4617,
+      "step": 2815
+    },
+    {
+      "epoch": 0.5384836026388756,
+      "grad_norm": 2.2829906146090093,
+      "learning_rate": 1e-06,
+      "loss": 0.252,
+      "step": 2816
+    },
+    {
+      "epoch": 0.5386748255091309,
+      "grad_norm": 2.00814139793105,
+      "learning_rate": 1e-06,
+      "loss": 0.265,
+      "step": 2817
+    },
+    {
+      "epoch": 0.5388660483793861,
+      "grad_norm": 4.965509281526595,
+      "learning_rate": 1e-06,
+      "loss": 0.4496,
+      "step": 2818
+    },
+    {
+      "epoch": 0.5390572712496414,
+      "grad_norm": 2.763393040262406,
+      "learning_rate": 1e-06,
+      "loss": 0.1007,
+      "step": 2819
+    },
+    {
+      "epoch": 0.5392484941198967,
+      "grad_norm": 2.4149961178513437,
+      "learning_rate": 1e-06,
+      "loss": 0.1353,
+      "step": 2820
+    },
+    {
+      "epoch": 0.539439716990152,
+      "grad_norm": 3.5515748197893178,
+      "learning_rate": 1e-06,
+      "loss": 0.0848,
+      "step": 2821
+    },
+    {
+      "epoch": 0.5396309398604073,
+      "grad_norm": 2.6340692163930064,
+      "learning_rate": 1e-06,
+      "loss": 0.3195,
+      "step": 2822
+    },
+    {
+      "epoch": 0.5398221627306626,
+      "grad_norm": 3.663689893309528,
+      "learning_rate": 1e-06,
+      "loss": 0.152,
+      "step": 2823
+    },
+    {
+      "epoch": 0.5400133856009178,
+      "grad_norm": 6.9008027425228535,
+      "learning_rate": 1e-06,
+      "loss": 0.1096,
+      "step": 2824
+    },
+    {
+      "epoch": 0.5402046084711731,
+      "grad_norm": 7.800687847619343,
+      "learning_rate": 1e-06,
+      "loss": 0.1096,
+      "step": 2825
+    },
+    {
+      "epoch": 0.5403958313414284,
+      "grad_norm": 9.052257142558425,
+      "learning_rate": 1e-06,
+      "loss": 0.6124,
+      "step": 2826
+    },
+    {
+      "epoch": 0.5405870542116837,
+      "grad_norm": 6.189939075965139,
+      "learning_rate": 1e-06,
+      "loss": 0.7208,
+      "step": 2827
+    },
+    {
+      "epoch": 0.540778277081939,
+      "grad_norm": 2.5604523642037726,
+      "learning_rate": 1e-06,
+      "loss": 0.2876,
+      "step": 2828
+    },
+    {
+      "epoch": 0.5409694999521942,
+      "grad_norm": 2.8816792397990563,
+      "learning_rate": 1e-06,
+      "loss": 0.0858,
+      "step": 2829
+    },
+    {
+      "epoch": 0.5411607228224495,
+      "grad_norm": 4.073105569476752,
+      "learning_rate": 1e-06,
+      "loss": 0.3622,
+      "step": 2830
+    },
+    {
+      "epoch": 0.5413519456927048,
+      "grad_norm": 3.181015899378626,
+      "learning_rate": 1e-06,
+      "loss": 0.2127,
+      "step": 2831
+    },
+    {
+      "epoch": 0.5415431685629601,
+      "grad_norm": 2.5549229504637507,
+      "learning_rate": 1e-06,
+      "loss": 0.2349,
+      "step": 2832
+    },
+    {
+      "epoch": 0.5417343914332154,
+      "grad_norm": 2.9581738460335516,
+      "learning_rate": 1e-06,
+      "loss": 0.1124,
+      "step": 2833
+    },
+    {
+      "epoch": 0.5419256143034707,
+      "grad_norm": 4.403066364274235,
+      "learning_rate": 1e-06,
+      "loss": 0.1974,
+      "step": 2834
+    },
+    {
+      "epoch": 0.5421168371737259,
+      "grad_norm": 3.789632663891566,
+      "learning_rate": 1e-06,
+      "loss": 0.1086,
+      "step": 2835
+    },
+    {
+      "epoch": 0.5423080600439812,
+      "grad_norm": 2.3748179918116206,
+      "learning_rate": 1e-06,
+      "loss": 0.0523,
+      "step": 2836
+    },
+    {
+      "epoch": 0.5424992829142365,
+      "grad_norm": 2.9205165703970706,
+      "learning_rate": 1e-06,
+      "loss": 0.0776,
+      "step": 2837
+    },
+    {
+      "epoch": 0.5426905057844919,
+      "grad_norm": 8.11241137157866,
+      "learning_rate": 1e-06,
+      "loss": 0.2382,
+      "step": 2838
+    },
+    {
+      "epoch": 0.5428817286547472,
+      "grad_norm": 5.895783692635085,
+      "learning_rate": 1e-06,
+      "loss": 0.9043,
+      "step": 2839
+    },
+    {
+      "epoch": 0.5430729515250023,
+      "grad_norm": 5.133416127284467,
+      "learning_rate": 1e-06,
+      "loss": 0.442,
+      "step": 2840
+    },
+    {
+      "epoch": 0.5432641743952576,
+      "grad_norm": 3.2263615063002375,
+      "learning_rate": 1e-06,
+      "loss": 0.3263,
+      "step": 2841
+    },
+    {
+      "epoch": 0.543455397265513,
+      "grad_norm": 3.3445144207358792,
+      "learning_rate": 1e-06,
+      "loss": 0.3406,
+      "step": 2842
+    },
+    {
+      "epoch": 0.5436466201357683,
+      "grad_norm": 5.785472750457742,
+      "learning_rate": 1e-06,
+      "loss": 0.2932,
+      "step": 2843
+    },
+    {
+      "epoch": 0.5438378430060236,
+      "grad_norm": 4.4748834520219285,
+      "learning_rate": 1e-06,
+      "loss": 0.4489,
+      "step": 2844
+    },
+    {
+      "epoch": 0.5440290658762788,
+      "grad_norm": 3.525189761201236,
+      "learning_rate": 1e-06,
+      "loss": 0.166,
+      "step": 2845
+    },
+    {
+      "epoch": 0.544220288746534,
+      "grad_norm": 2.8395000181525583,
+      "learning_rate": 1e-06,
+      "loss": 0.0681,
+      "step": 2846
+    },
+    {
+      "epoch": 0.5444115116167894,
+      "grad_norm": 3.5221532270036446,
+      "learning_rate": 1e-06,
+      "loss": 0.2015,
+      "step": 2847
+    },
+    {
+      "epoch": 0.5446027344870447,
+      "grad_norm": 3.8178148588729206,
+      "learning_rate": 1e-06,
+      "loss": 0.0813,
+      "step": 2848
+    },
+    {
+      "epoch": 0.5447939573573,
+      "grad_norm": 11.611073016804369,
+      "learning_rate": 1e-06,
+      "loss": 0.108,
+      "step": 2849
+    },
+    {
+      "epoch": 0.5449851802275553,
+      "grad_norm": 4.362399456975117,
+      "learning_rate": 1e-06,
+      "loss": 0.1207,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5451764030978105,
+      "grad_norm": 4.788802023146763,
+      "learning_rate": 1e-06,
+      "loss": 0.2846,
+      "step": 2851
+    },
+    {
+      "epoch": 0.5453676259680658,
+      "grad_norm": 4.201386840371438,
+      "learning_rate": 1e-06,
+      "loss": 0.354,
+      "step": 2852
+    },
+    {
+      "epoch": 0.5455588488383211,
+      "grad_norm": 4.955141060338892,
+      "learning_rate": 1e-06,
+      "loss": 0.4887,
+      "step": 2853
+    },
+    {
+      "epoch": 0.5457500717085764,
+      "grad_norm": 9.26601322898192,
+      "learning_rate": 1e-06,
+      "loss": 0.2676,
+      "step": 2854
+    },
+    {
+      "epoch": 0.5459412945788317,
+      "grad_norm": 3.9018559270292092,
+      "learning_rate": 1e-06,
+      "loss": 0.2831,
+      "step": 2855
+    },
+    {
+      "epoch": 0.5461325174490869,
+      "grad_norm": 3.8090569630171935,
+      "learning_rate": 1e-06,
+      "loss": 0.321,
+      "step": 2856
+    },
+    {
+      "epoch": 0.5463237403193422,
+      "grad_norm": 1.917018056530752,
+      "learning_rate": 1e-06,
+      "loss": 0.2387,
+      "step": 2857
+    },
+    {
+      "epoch": 0.5465149631895975,
+      "grad_norm": 3.6882463120234115,
+      "learning_rate": 1e-06,
+      "loss": 0.1509,
+      "step": 2858
+    },
+    {
+      "epoch": 0.5467061860598528,
+      "grad_norm": 2.911521806292308,
+      "learning_rate": 1e-06,
+      "loss": 0.1848,
+      "step": 2859
+    },
+    {
+      "epoch": 0.5468974089301081,
+      "grad_norm": 3.240364462992912,
+      "learning_rate": 1e-06,
+      "loss": 0.0998,
+      "step": 2860
+    },
+    {
+      "epoch": 0.5470886318003634,
+      "grad_norm": 4.39183250466402,
+      "learning_rate": 1e-06,
+      "loss": 0.1168,
+      "step": 2861
+    },
+    {
+      "epoch": 0.5472798546706186,
+      "grad_norm": 4.006529961621905,
+      "learning_rate": 1e-06,
+      "loss": 0.191,
+      "step": 2862
+    },
+    {
+      "epoch": 0.5474710775408739,
+      "grad_norm": 9.368578937131796,
+      "learning_rate": 1e-06,
+      "loss": 0.4036,
+      "step": 2863
+    },
+    {
+      "epoch": 0.5476623004111292,
+      "grad_norm": 4.3158425908131495,
+      "learning_rate": 1e-06,
+      "loss": 0.3092,
+      "step": 2864
+    },
+    {
+      "epoch": 0.5478535232813845,
+      "grad_norm": 3.996563627448845,
+      "learning_rate": 1e-06,
+      "loss": 0.7786,
+      "step": 2865
+    },
+    {
+      "epoch": 0.5480447461516398,
+      "grad_norm": 6.598286157670479,
+      "learning_rate": 1e-06,
+      "loss": 0.3486,
+      "step": 2866
+    },
+    {
+      "epoch": 0.548235969021895,
+      "grad_norm": 2.994316439309764,
+      "learning_rate": 1e-06,
+      "loss": 0.3763,
+      "step": 2867
+    },
+    {
+      "epoch": 0.5484271918921503,
+      "grad_norm": 2.226221854938543,
+      "learning_rate": 1e-06,
+      "loss": 0.1746,
+      "step": 2868
+    },
+    {
+      "epoch": 0.5486184147624056,
+      "grad_norm": 2.642224882415225,
+      "learning_rate": 1e-06,
+      "loss": 0.0964,
+      "step": 2869
+    },
+    {
+      "epoch": 0.5488096376326609,
+      "grad_norm": 5.86892808475395,
+      "learning_rate": 1e-06,
+      "loss": 0.0951,
+      "step": 2870
+    },
+    {
+      "epoch": 0.5490008605029162,
+      "grad_norm": 2.430803539401728,
+      "learning_rate": 1e-06,
+      "loss": 0.1181,
+      "step": 2871
+    },
+    {
+      "epoch": 0.5491920833731714,
+      "grad_norm": 4.104169558348177,
+      "learning_rate": 1e-06,
+      "loss": 0.1133,
+      "step": 2872
+    },
+    {
+      "epoch": 0.5493833062434267,
+      "grad_norm": 2.000717511217509,
+      "learning_rate": 1e-06,
+      "loss": 0.0461,
+      "step": 2873
+    },
+    {
+      "epoch": 0.549574529113682,
+      "grad_norm": 2.704712770763795,
+      "learning_rate": 1e-06,
+      "loss": 0.0519,
+      "step": 2874
+    },
+    {
+      "epoch": 0.5497657519839373,
+      "grad_norm": 2.8481019643254233,
+      "learning_rate": 1e-06,
+      "loss": 0.061,
+      "step": 2875
+    },
+    {
+      "epoch": 0.5499569748541926,
+      "grad_norm": 5.60767834009466,
+      "learning_rate": 1e-06,
+      "loss": 0.3017,
+      "step": 2876
+    },
+    {
+      "epoch": 0.5501481977244479,
+      "grad_norm": 3.9102589291744447,
+      "learning_rate": 1e-06,
+      "loss": 0.3349,
+      "step": 2877
+    },
+    {
+      "epoch": 0.5503394205947031,
+      "grad_norm": 4.0820740035889065,
+      "learning_rate": 1e-06,
+      "loss": 0.5101,
+      "step": 2878
+    },
+    {
+      "epoch": 0.5505306434649584,
+      "grad_norm": 2.4047616095769793,
+      "learning_rate": 1e-06,
+      "loss": 0.3637,
+      "step": 2879
+    },
+    {
+      "epoch": 0.5507218663352137,
+      "grad_norm": 5.060463578169155,
+      "learning_rate": 1e-06,
+      "loss": 0.3266,
+      "step": 2880
+    },
+    {
+      "epoch": 0.550913089205469,
+      "grad_norm": 2.6356227829335626,
+      "learning_rate": 1e-06,
+      "loss": 0.1636,
+      "step": 2881
+    },
+    {
+      "epoch": 0.5511043120757243,
+      "grad_norm": 2.7040699095576555,
+      "learning_rate": 1e-06,
+      "loss": 0.1842,
+      "step": 2882
+    },
+    {
+      "epoch": 0.5512955349459795,
+      "grad_norm": 2.920495998114019,
+      "learning_rate": 1e-06,
+      "loss": 0.2769,
+      "step": 2883
+    },
+    {
+      "epoch": 0.5514867578162348,
+      "grad_norm": 5.572471168275763,
+      "learning_rate": 1e-06,
+      "loss": 0.1559,
+      "step": 2884
+    },
+    {
+      "epoch": 0.5516779806864901,
+      "grad_norm": 3.767293346789788,
+      "learning_rate": 1e-06,
+      "loss": 0.1298,
+      "step": 2885
+    },
+    {
+      "epoch": 0.5518692035567454,
+      "grad_norm": 4.312505528543218,
+      "learning_rate": 1e-06,
+      "loss": 0.1193,
+      "step": 2886
+    },
+    {
+      "epoch": 0.5520604264270007,
+      "grad_norm": 4.1214369210642445,
+      "learning_rate": 1e-06,
+      "loss": 0.1996,
+      "step": 2887
+    },
+    {
+      "epoch": 0.5522516492972559,
+      "grad_norm": 6.3219919884983025,
+      "learning_rate": 1e-06,
+      "loss": 0.1543,
+      "step": 2888
+    },
+    {
+      "epoch": 0.5524428721675112,
+      "grad_norm": 4.8085409947095945,
+      "learning_rate": 1e-06,
+      "loss": 0.6852,
+      "step": 2889
+    },
+    {
+      "epoch": 0.5526340950377665,
+      "grad_norm": 4.203002927914819,
+      "learning_rate": 1e-06,
+      "loss": 0.5196,
+      "step": 2890
+    },
+    {
+      "epoch": 0.5528253179080218,
+      "grad_norm": 2.7051348848963888,
+      "learning_rate": 1e-06,
+      "loss": 0.4623,
+      "step": 2891
+    },
+    {
+      "epoch": 0.5530165407782771,
+      "grad_norm": 3.2479916749669013,
+      "learning_rate": 1e-06,
+      "loss": 0.2892,
+      "step": 2892
+    },
+    {
+      "epoch": 0.5532077636485324,
+      "grad_norm": 2.0135711141512695,
+      "learning_rate": 1e-06,
+      "loss": 0.2424,
+      "step": 2893
+    },
+    {
+      "epoch": 0.5533989865187876,
+      "grad_norm": 5.631329345144133,
+      "learning_rate": 1e-06,
+      "loss": 0.5644,
+      "step": 2894
+    },
+    {
+      "epoch": 0.5535902093890429,
+      "grad_norm": 4.132319343435104,
+      "learning_rate": 1e-06,
+      "loss": 0.0886,
+      "step": 2895
+    },
+    {
+      "epoch": 0.5537814322592982,
+      "grad_norm": 4.955405303053353,
+      "learning_rate": 1e-06,
+      "loss": 0.1758,
+      "step": 2896
+    },
+    {
+      "epoch": 0.5539726551295535,
+      "grad_norm": 2.717921251286015,
+      "learning_rate": 1e-06,
+      "loss": 0.1771,
+      "step": 2897
+    },
+    {
+      "epoch": 0.5541638779998088,
+      "grad_norm": 2.854876204978414,
+      "learning_rate": 1e-06,
+      "loss": 0.0796,
+      "step": 2898
+    },
+    {
+      "epoch": 0.554355100870064,
+      "grad_norm": 3.7790398562071617,
+      "learning_rate": 1e-06,
+      "loss": 0.1019,
+      "step": 2899
+    },
+    {
+      "epoch": 0.5545463237403193,
+      "grad_norm": 3.0490583373206825,
+      "learning_rate": 1e-06,
+      "loss": 0.0504,
+      "step": 2900
+    },
+    {
+      "epoch": 0.5547375466105746,
+      "grad_norm": 3.909042092960076,
+      "learning_rate": 1e-06,
+      "loss": 0.1944,
+      "step": 2901
+    },
+    {
+      "epoch": 0.5549287694808299,
+      "grad_norm": 7.5374441673020245,
+      "learning_rate": 1e-06,
+      "loss": 0.4923,
+      "step": 2902
+    },
+    {
+      "epoch": 0.5551199923510852,
+      "grad_norm": 3.193946602400675,
+      "learning_rate": 1e-06,
+      "loss": 0.3753,
+      "step": 2903
+    },
+    {
+      "epoch": 0.5553112152213405,
+      "grad_norm": 3.5654839851384565,
+      "learning_rate": 1e-06,
+      "loss": 0.2467,
+      "step": 2904
+    },
+    {
+      "epoch": 0.5555024380915957,
+      "grad_norm": 3.4218885168914683,
+      "learning_rate": 1e-06,
+      "loss": 0.1537,
+      "step": 2905
+    },
+    {
+      "epoch": 0.555693660961851,
+      "grad_norm": 3.283676267681182,
+      "learning_rate": 1e-06,
+      "loss": 0.2694,
+      "step": 2906
+    },
+    {
+      "epoch": 0.5558848838321063,
+      "grad_norm": 4.149582412116463,
+      "learning_rate": 1e-06,
+      "loss": 0.1757,
+      "step": 2907
+    },
+    {
+      "epoch": 0.5560761067023616,
+      "grad_norm": 3.482280426661481,
+      "learning_rate": 1e-06,
+      "loss": 0.2407,
+      "step": 2908
+    },
+    {
+      "epoch": 0.5562673295726169,
+      "grad_norm": 3.3921063962120432,
+      "learning_rate": 1e-06,
+      "loss": 0.1087,
+      "step": 2909
+    },
+    {
+      "epoch": 0.5564585524428721,
+      "grad_norm": 3.2354258428878664,
+      "learning_rate": 1e-06,
+      "loss": 0.0882,
+      "step": 2910
+    },
+    {
+      "epoch": 0.5566497753131274,
+      "grad_norm": 3.7088968227472527,
+      "learning_rate": 1e-06,
+      "loss": 0.1365,
+      "step": 2911
+    },
+    {
+      "epoch": 0.5568409981833827,
+      "grad_norm": 2.5295393072762606,
+      "learning_rate": 1e-06,
+      "loss": 0.0682,
+      "step": 2912
+    },
+    {
+      "epoch": 0.557032221053638,
+      "grad_norm": 15.824259731340417,
+      "learning_rate": 1e-06,
+      "loss": 0.1768,
+      "step": 2913
+    },
+    {
+      "epoch": 0.5572234439238933,
+      "grad_norm": 4.817809309717416,
+      "learning_rate": 1e-06,
+      "loss": 0.6315,
+      "step": 2914
+    },
+    {
+      "epoch": 0.5574146667941485,
+      "grad_norm": 5.702354734477877,
+      "learning_rate": 1e-06,
+      "loss": 0.8182,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5576058896644038,
+      "grad_norm": 3.354687778595485,
+      "learning_rate": 1e-06,
+      "loss": 0.2005,
+      "step": 2916
+    },
+    {
+      "epoch": 0.5577971125346591,
+      "grad_norm": 2.5411375047333915,
+      "learning_rate": 1e-06,
+      "loss": 0.0813,
+      "step": 2917
+    },
+    {
+      "epoch": 0.5579883354049144,
+      "grad_norm": 4.187849115650431,
+      "learning_rate": 1e-06,
+      "loss": 0.5953,
+      "step": 2918
+    },
+    {
+      "epoch": 0.5581795582751697,
+      "grad_norm": 2.5350803546095717,
+      "learning_rate": 1e-06,
+      "loss": 0.1183,
+      "step": 2919
+    },
+    {
+      "epoch": 0.558370781145425,
+      "grad_norm": 5.502094390214785,
+      "learning_rate": 1e-06,
+      "loss": 0.1783,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5585620040156802,
+      "grad_norm": 1.7388124490104913,
+      "learning_rate": 1e-06,
+      "loss": 0.0473,
+      "step": 2921
+    },
+    {
+      "epoch": 0.5587532268859355,
+      "grad_norm": 4.63110731885791,
+      "learning_rate": 1e-06,
+      "loss": 0.0759,
+      "step": 2922
+    },
+    {
+      "epoch": 0.5589444497561908,
+      "grad_norm": 2.0963273767374657,
+      "learning_rate": 1e-06,
+      "loss": 0.0529,
+      "step": 2923
+    },
+    {
+      "epoch": 0.5591356726264461,
+      "grad_norm": 3.2554645081490845,
+      "learning_rate": 1e-06,
+      "loss": 0.0824,
+      "step": 2924
+    },
+    {
+      "epoch": 0.5593268954967014,
+      "grad_norm": 5.721016523295003,
+      "learning_rate": 1e-06,
+      "loss": 0.1021,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5595181183669566,
+      "grad_norm": 4.943491137800957,
+      "learning_rate": 1e-06,
+      "loss": 0.79,
+      "step": 2926
+    },
+    {
+      "epoch": 0.5597093412372119,
+      "grad_norm": 5.050996119390284,
+      "learning_rate": 1e-06,
+      "loss": 0.3179,
+      "step": 2927
+    },
+    {
+      "epoch": 0.5599005641074672,
+      "grad_norm": 3.0274188813761778,
+      "learning_rate": 1e-06,
+      "loss": 0.1714,
+      "step": 2928
+    },
+    {
+      "epoch": 0.5600917869777225,
+      "grad_norm": 4.093974798888076,
+      "learning_rate": 1e-06,
+      "loss": 0.3123,
+      "step": 2929
+    },
+    {
+      "epoch": 0.5602830098479779,
+      "grad_norm": 4.409826247756298,
+      "learning_rate": 1e-06,
+      "loss": 0.3728,
+      "step": 2930
+    },
+    {
+      "epoch": 0.560474232718233,
+      "grad_norm": 3.880149496088973,
+      "learning_rate": 1e-06,
+      "loss": 0.1428,
+      "step": 2931
+    },
+    {
+      "epoch": 0.5606654555884883,
+      "grad_norm": 3.0083302238074916,
+      "learning_rate": 1e-06,
+      "loss": 0.2435,
+      "step": 2932
+    },
+    {
+      "epoch": 0.5608566784587437,
+      "grad_norm": 2.9800607544511113,
+      "learning_rate": 1e-06,
+      "loss": 0.3745,
+      "step": 2933
+    },
+    {
+      "epoch": 0.561047901328999,
+      "grad_norm": 2.427248402058976,
+      "learning_rate": 1e-06,
+      "loss": 0.149,
+      "step": 2934
+    },
+    {
+      "epoch": 0.5612391241992543,
+      "grad_norm": 3.3340910686098812,
+      "learning_rate": 1e-06,
+      "loss": 0.1487,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5614303470695096,
+      "grad_norm": 6.980195503885327,
+      "learning_rate": 1e-06,
+      "loss": 0.1815,
+      "step": 2936
+    },
+    {
+      "epoch": 0.5616215699397648,
+      "grad_norm": 3.429237979288009,
+      "learning_rate": 1e-06,
+      "loss": 0.0811,
+      "step": 2937
+    },
+    {
+      "epoch": 0.5618127928100201,
+      "grad_norm": 6.038213475247149,
+      "learning_rate": 1e-06,
+      "loss": 0.1176,
+      "step": 2938
+    },
+    {
+      "epoch": 0.5620040156802754,
+      "grad_norm": 4.081689438814124,
+      "learning_rate": 1e-06,
+      "loss": 0.6611,
+      "step": 2939
+    },
+    {
+      "epoch": 0.5621952385505307,
+      "grad_norm": 5.207108722405193,
+      "learning_rate": 1e-06,
+      "loss": 0.2512,
+      "step": 2940
+    },
+    {
+      "epoch": 0.562386461420786,
+      "grad_norm": 4.345836549918038,
+      "learning_rate": 1e-06,
+      "loss": 0.2384,
+      "step": 2941
+    },
+    {
+      "epoch": 0.5625776842910412,
+      "grad_norm": 3.933567081122356,
+      "learning_rate": 1e-06,
+      "loss": 0.5559,
+      "step": 2942
+    },
+    {
+      "epoch": 0.5627689071612965,
+      "grad_norm": 3.7742044684286564,
+      "learning_rate": 1e-06,
+      "loss": 0.198,
+      "step": 2943
+    },
+    {
+      "epoch": 0.5629601300315518,
+      "grad_norm": 3.1347628869731032,
+      "learning_rate": 1e-06,
+      "loss": 0.3118,
+      "step": 2944
+    },
+    {
+      "epoch": 0.5631513529018071,
+      "grad_norm": 3.3077355434888047,
+      "learning_rate": 1e-06,
+      "loss": 0.1179,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5633425757720624,
+      "grad_norm": 4.704226748159117,
+      "learning_rate": 1e-06,
+      "loss": 0.2676,
+      "step": 2946
+    },
+    {
+      "epoch": 0.5635337986423177,
+      "grad_norm": 1.2451961716676088,
+      "learning_rate": 1e-06,
+      "loss": 0.0779,
+      "step": 2947
+    },
+    {
+      "epoch": 0.5637250215125729,
+      "grad_norm": 5.528858767083621,
+      "learning_rate": 1e-06,
+      "loss": 0.1241,
+      "step": 2948
+    },
+    {
+      "epoch": 0.5639162443828282,
+      "grad_norm": 3.310441295099241,
+      "learning_rate": 1e-06,
+      "loss": 0.0735,
+      "step": 2949
+    },
+    {
+      "epoch": 0.5641074672530835,
+      "grad_norm": 6.054781060265031,
+      "learning_rate": 1e-06,
+      "loss": 0.1602,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5642986901233388,
+      "grad_norm": 6.048392645064239,
+      "learning_rate": 1e-06,
+      "loss": 0.3786,
+      "step": 2951
+    },
+    {
+      "epoch": 0.5644899129935941,
+      "grad_norm": 4.13499580503018,
+      "learning_rate": 1e-06,
+      "loss": 0.5543,
+      "step": 2952
+    },
+    {
+      "epoch": 0.5646811358638493,
+      "grad_norm": 4.135669204085961,
+      "learning_rate": 1e-06,
+      "loss": 0.3081,
+      "step": 2953
+    },
+    {
+      "epoch": 0.5648723587341046,
+      "grad_norm": 1.6669086598508163,
+      "learning_rate": 1e-06,
+      "loss": 0.5363,
+      "step": 2954
+    },
+    {
+      "epoch": 0.5650635816043599,
+      "grad_norm": 6.212249885912473,
+      "learning_rate": 1e-06,
+      "loss": 0.7076,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5652548044746152,
+      "grad_norm": 4.05017709227137,
+      "learning_rate": 1e-06,
+      "loss": 0.3927,
+      "step": 2956
+    },
+    {
+      "epoch": 0.5654460273448705,
+      "grad_norm": 3.572935647422858,
+      "learning_rate": 1e-06,
+      "loss": 0.1061,
+      "step": 2957
+    },
+    {
+      "epoch": 0.5656372502151257,
+      "grad_norm": 3.7386094350864836,
+      "learning_rate": 1e-06,
+      "loss": 0.0559,
+      "step": 2958
+    },
+    {
+      "epoch": 0.565828473085381,
+      "grad_norm": 6.534137872218055,
+      "learning_rate": 1e-06,
+      "loss": 0.2276,
+      "step": 2959
+    },
+    {
+      "epoch": 0.5660196959556363,
+      "grad_norm": 4.3789393536884305,
+      "learning_rate": 1e-06,
+      "loss": 0.101,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5662109188258916,
+      "grad_norm": 4.4876067847507555,
+      "learning_rate": 1e-06,
+      "loss": 0.1684,
+      "step": 2961
+    },
+    {
+      "epoch": 0.5664021416961469,
+      "grad_norm": 3.0606669954391013,
+      "learning_rate": 1e-06,
+      "loss": 0.0486,
+      "step": 2962
+    },
+    {
+      "epoch": 0.5665933645664022,
+      "grad_norm": 6.669357646797661,
+      "learning_rate": 1e-06,
+      "loss": 0.156,
+      "step": 2963
+    },
+    {
+      "epoch": 0.5667845874366574,
+      "grad_norm": 7.520750459195041,
+      "learning_rate": 1e-06,
+      "loss": 0.8849,
+      "step": 2964
+    },
+    {
+      "epoch": 0.5669758103069127,
+      "grad_norm": 4.3069143195344095,
+      "learning_rate": 1e-06,
+      "loss": 0.4593,
+      "step": 2965
+    },
+    {
+      "epoch": 0.567167033177168,
+      "grad_norm": 5.409923738516781,
+      "learning_rate": 1e-06,
+      "loss": 0.3073,
+      "step": 2966
+    },
+    {
+      "epoch": 0.5673582560474233,
+      "grad_norm": 2.556718677935022,
+      "learning_rate": 1e-06,
+      "loss": 0.136,
+      "step": 2967
+    },
+    {
+      "epoch": 0.5675494789176786,
+      "grad_norm": 5.217761283002173,
+      "learning_rate": 1e-06,
+      "loss": 0.2872,
+      "step": 2968
+    },
+    {
+      "epoch": 0.5677407017879338,
+      "grad_norm": 3.40231754192152,
+      "learning_rate": 1e-06,
+      "loss": 0.1554,
+      "step": 2969
+    },
+    {
+      "epoch": 0.5679319246581891,
+      "grad_norm": 2.279500421034098,
+      "learning_rate": 1e-06,
+      "loss": 0.1675,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5681231475284444,
+      "grad_norm": 4.5503614764148175,
+      "learning_rate": 1e-06,
+      "loss": 0.1292,
+      "step": 2971
+    },
+    {
+      "epoch": 0.5683143703986997,
+      "grad_norm": 2.728001955104362,
+      "learning_rate": 1e-06,
+      "loss": 0.3477,
+      "step": 2972
+    },
+    {
+      "epoch": 0.568505593268955,
+      "grad_norm": 5.947883280734436,
+      "learning_rate": 1e-06,
+      "loss": 0.1731,
+      "step": 2973
+    },
+    {
+      "epoch": 0.5686968161392103,
+      "grad_norm": 4.381336528016139,
+      "learning_rate": 1e-06,
+      "loss": 0.1373,
+      "step": 2974
+    },
+    {
+      "epoch": 0.5688880390094655,
+      "grad_norm": 6.022743353323301,
+      "learning_rate": 1e-06,
+      "loss": 0.1592,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5690792618797208,
+      "grad_norm": 4.119323060801137,
+      "learning_rate": 1e-06,
+      "loss": 0.2735,
+      "step": 2976
+    },
+    {
+      "epoch": 0.5692704847499761,
+      "grad_norm": 5.515243906142165,
+      "learning_rate": 1e-06,
+      "loss": 0.6345,
+      "step": 2977
+    },
+    {
+      "epoch": 0.5694617076202314,
+      "grad_norm": 5.384557875651013,
+      "learning_rate": 1e-06,
+      "loss": 0.4277,
+      "step": 2978
+    },
+    {
+      "epoch": 0.5696529304904867,
+      "grad_norm": 2.1610618748595414,
+      "learning_rate": 1e-06,
+      "loss": 0.2609,
+      "step": 2979
+    },
+    {
+      "epoch": 0.5698441533607419,
+      "grad_norm": 3.46390284134632,
+      "learning_rate": 1e-06,
+      "loss": 0.3334,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5700353762309972,
+      "grad_norm": 3.9439478563107726,
+      "learning_rate": 1e-06,
+      "loss": 0.3128,
+      "step": 2981
+    },
+    {
+      "epoch": 0.5702265991012525,
+      "grad_norm": 3.404508653051299,
+      "learning_rate": 1e-06,
+      "loss": 0.2736,
+      "step": 2982
+    },
+    {
+      "epoch": 0.5704178219715078,
+      "grad_norm": 4.432522413236921,
+      "learning_rate": 1e-06,
+      "loss": 0.1371,
+      "step": 2983
+    },
+    {
+      "epoch": 0.5706090448417631,
+      "grad_norm": 2.9683056950625586,
+      "learning_rate": 1e-06,
+      "loss": 0.31,
+      "step": 2984
+    },
+    {
+      "epoch": 0.5708002677120183,
+      "grad_norm": 3.7405233804384825,
+      "learning_rate": 1e-06,
+      "loss": 0.1191,
+      "step": 2985
+    },
+    {
+      "epoch": 0.5709914905822736,
+      "grad_norm": 4.325715313450766,
+      "learning_rate": 1e-06,
+      "loss": 0.1205,
+      "step": 2986
+    },
+    {
+      "epoch": 0.5711827134525289,
+      "grad_norm": 4.152428957189615,
+      "learning_rate": 1e-06,
+      "loss": 0.0623,
+      "step": 2987
+    },
+    {
+      "epoch": 0.5713739363227842,
+      "grad_norm": 6.471716159287791,
+      "learning_rate": 1e-06,
+      "loss": 0.2802,
+      "step": 2988
+    },
+    {
+      "epoch": 0.5715651591930395,
+      "grad_norm": 3.93600665641625,
+      "learning_rate": 1e-06,
+      "loss": 0.4433,
+      "step": 2989
+    },
+    {
+      "epoch": 0.5717563820632948,
+      "grad_norm": 3.16375233578153,
+      "learning_rate": 1e-06,
+      "loss": 0.2633,
+      "step": 2990
+    },
+    {
+      "epoch": 0.57194760493355,
+      "grad_norm": 4.4321450172964445,
+      "learning_rate": 1e-06,
+      "loss": 0.2572,
+      "step": 2991
+    },
+    {
+      "epoch": 0.5721388278038053,
+      "grad_norm": 2.630842065789616,
+      "learning_rate": 1e-06,
+      "loss": 0.3319,
+      "step": 2992
+    },
+    {
+      "epoch": 0.5723300506740606,
+      "grad_norm": 3.423219738802485,
+      "learning_rate": 1e-06,
+      "loss": 0.48,
+      "step": 2993
+    },
+    {
+      "epoch": 0.5725212735443159,
+      "grad_norm": 5.487323370591303,
+      "learning_rate": 1e-06,
+      "loss": 0.3357,
+      "step": 2994
+    },
+    {
+      "epoch": 0.5727124964145712,
+      "grad_norm": 2.515191461392826,
+      "learning_rate": 1e-06,
+      "loss": 0.1313,
+      "step": 2995
+    },
+    {
+      "epoch": 0.5729037192848264,
+      "grad_norm": 3.2119335105541493,
+      "learning_rate": 1e-06,
+      "loss": 0.15,
+      "step": 2996
+    },
+    {
+      "epoch": 0.5730949421550817,
+      "grad_norm": 2.4972029297559284,
+      "learning_rate": 1e-06,
+      "loss": 0.104,
+      "step": 2997
+    },
+    {
+      "epoch": 0.573286165025337,
+      "grad_norm": 4.397415503144979,
+      "learning_rate": 1e-06,
+      "loss": 0.5011,
+      "step": 2998
+    },
+    {
+      "epoch": 0.5734773878955923,
+      "grad_norm": 3.98671985516317,
+      "learning_rate": 1e-06,
+      "loss": 0.1692,
+      "step": 2999
+    },
+    {
+      "epoch": 0.5736686107658476,
+      "grad_norm": 6.173577877718548,
+      "learning_rate": 1e-06,
+      "loss": 0.1266,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5736686107658476,
+      "eval_runtime": 741.726,
+      "eval_samples_per_second": 2.068,
+      "eval_steps_per_second": 0.518,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5738598336361028,
+      "grad_norm": 3.614675090112537,
+      "learning_rate": 1e-06,
+      "loss": 0.5129,
+      "step": 3001
+    },
+    {
+      "epoch": 0.5740510565063581,
+      "grad_norm": 3.7945295711244227,
+      "learning_rate": 1e-06,
+      "loss": 0.3129,
+      "step": 3002
+    },
+    {
+      "epoch": 0.5742422793766134,
+      "grad_norm": 5.095707990966439,
+      "learning_rate": 1e-06,
+      "loss": 0.3718,
+      "step": 3003
+    },
+    {
+      "epoch": 0.5744335022468687,
+      "grad_norm": 1.2887964147429085,
+      "learning_rate": 1e-06,
+      "loss": 0.1743,
+      "step": 3004
+    },
+    {
+      "epoch": 0.574624725117124,
+      "grad_norm": 2.9163078269016043,
+      "learning_rate": 1e-06,
+      "loss": 0.2492,
+      "step": 3005
+    },
+    {
+      "epoch": 0.5748159479873793,
+      "grad_norm": 2.2536124793319576,
+      "learning_rate": 1e-06,
+      "loss": 0.1466,
+      "step": 3006
+    },
+    {
+      "epoch": 0.5750071708576345,
+      "grad_norm": 3.579929346634105,
+      "learning_rate": 1e-06,
+      "loss": 0.0753,
+      "step": 3007
+    },
+    {
+      "epoch": 0.5751983937278898,
+      "grad_norm": 2.332874956974842,
+      "learning_rate": 1e-06,
+      "loss": 0.1195,
+      "step": 3008
+    },
+    {
+      "epoch": 0.5753896165981451,
+      "grad_norm": 2.789425866314746,
+      "learning_rate": 1e-06,
+      "loss": 0.1667,
+      "step": 3009
+    },
+    {
+      "epoch": 0.5755808394684004,
+      "grad_norm": 4.438654789635339,
+      "learning_rate": 1e-06,
+      "loss": 0.1929,
+      "step": 3010
+    },
+    {
+      "epoch": 0.5757720623386557,
+      "grad_norm": 3.049505887890417,
+      "learning_rate": 1e-06,
+      "loss": 0.1204,
+      "step": 3011
+    },
+    {
+      "epoch": 0.5759632852089109,
+      "grad_norm": 1.7969891055821832,
+      "learning_rate": 1e-06,
+      "loss": 0.0393,
+      "step": 3012
+    },
+    {
+      "epoch": 0.5761545080791662,
+      "grad_norm": 10.314488814497524,
+      "learning_rate": 1e-06,
+      "loss": 0.2632,
+      "step": 3013
+    },
+    {
+      "epoch": 0.5763457309494215,
+      "grad_norm": 4.492139096206613,
+      "learning_rate": 1e-06,
+      "loss": 0.6049,
+      "step": 3014
+    },
+    {
+      "epoch": 0.5765369538196768,
+      "grad_norm": 3.708456073115607,
+      "learning_rate": 1e-06,
+      "loss": 0.1751,
+      "step": 3015
+    },
+    {
+      "epoch": 0.5767281766899321,
+      "grad_norm": 2.55850475826272,
+      "learning_rate": 1e-06,
+      "loss": 0.3028,
+      "step": 3016
+    },
+    {
+      "epoch": 0.5769193995601875,
+      "grad_norm": 3.5919187441090905,
+      "learning_rate": 1e-06,
+      "loss": 0.1174,
+      "step": 3017
+    },
+    {
+      "epoch": 0.5771106224304426,
+      "grad_norm": 3.2339769201357487,
+      "learning_rate": 1e-06,
+      "loss": 0.2926,
+      "step": 3018
+    },
+    {
+      "epoch": 0.577301845300698,
+      "grad_norm": 2.870865045897381,
+      "learning_rate": 1e-06,
+      "loss": 0.2888,
+      "step": 3019
+    },
+    {
+      "epoch": 0.5774930681709532,
+      "grad_norm": 2.310369180019903,
+      "learning_rate": 1e-06,
+      "loss": 0.3,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5776842910412086,
+      "grad_norm": 2.5513448983335603,
+      "learning_rate": 1e-06,
+      "loss": 0.139,
+      "step": 3021
+    },
+    {
+      "epoch": 0.5778755139114639,
+      "grad_norm": 5.210916270337189,
+      "learning_rate": 1e-06,
+      "loss": 0.2538,
+      "step": 3022
+    },
+    {
+      "epoch": 0.578066736781719,
+      "grad_norm": 4.5370587179022035,
+      "learning_rate": 1e-06,
+      "loss": 0.1217,
+      "step": 3023
+    },
+    {
+      "epoch": 0.5782579596519744,
+      "grad_norm": 3.979380869800426,
+      "learning_rate": 1e-06,
+      "loss": 0.2862,
+      "step": 3024
+    },
+    {
+      "epoch": 0.5784491825222297,
+      "grad_norm": 4.499148818153343,
+      "learning_rate": 1e-06,
+      "loss": 0.2231,
+      "step": 3025
+    },
+    {
+      "epoch": 0.578640405392485,
+      "grad_norm": 7.073590508237147,
+      "learning_rate": 1e-06,
+      "loss": 0.2226,
+      "step": 3026
+    },
+    {
+      "epoch": 0.5788316282627403,
+      "grad_norm": 4.322404627368349,
+      "learning_rate": 1e-06,
+      "loss": 0.4774,
+      "step": 3027
+    },
+    {
+      "epoch": 0.5790228511329955,
+      "grad_norm": 3.3237122128488097,
+      "learning_rate": 1e-06,
+      "loss": 0.2193,
+      "step": 3028
+    },
+    {
+      "epoch": 0.5792140740032508,
+      "grad_norm": 3.7472923676303904,
+      "learning_rate": 1e-06,
+      "loss": 0.1139,
+      "step": 3029
+    },
+    {
+      "epoch": 0.5794052968735061,
+      "grad_norm": 4.225763890658673,
+      "learning_rate": 1e-06,
+      "loss": 0.1545,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5795965197437614,
+      "grad_norm": 3.0804310791746214,
+      "learning_rate": 1e-06,
+      "loss": 0.3106,
+      "step": 3031
+    },
+    {
+      "epoch": 0.5797877426140167,
+      "grad_norm": 2.9261078391237807,
+      "learning_rate": 1e-06,
+      "loss": 0.1321,
+      "step": 3032
+    },
+    {
+      "epoch": 0.579978965484272,
+      "grad_norm": 3.5159164646974985,
+      "learning_rate": 1e-06,
+      "loss": 0.083,
+      "step": 3033
+    },
+    {
+      "epoch": 0.5801701883545272,
+      "grad_norm": 4.324813732921094,
+      "learning_rate": 1e-06,
+      "loss": 0.1278,
+      "step": 3034
+    },
+    {
+      "epoch": 0.5803614112247825,
+      "grad_norm": 2.234414160325265,
+      "learning_rate": 1e-06,
+      "loss": 0.0557,
+      "step": 3035
+    },
+    {
+      "epoch": 0.5805526340950378,
+      "grad_norm": 5.069183835279539,
+      "learning_rate": 1e-06,
+      "loss": 0.2409,
+      "step": 3036
+    },
+    {
+      "epoch": 0.5807438569652931,
+      "grad_norm": 11.070796291034268,
+      "learning_rate": 1e-06,
+      "loss": 0.1107,
+      "step": 3037
+    },
+    {
+      "epoch": 0.5809350798355484,
+      "grad_norm": 4.543082107087465,
+      "learning_rate": 1e-06,
+      "loss": 0.175,
+      "step": 3038
+    },
+    {
+      "epoch": 0.5811263027058036,
+      "grad_norm": 5.489771956135436,
+      "learning_rate": 1e-06,
+      "loss": 0.5867,
+      "step": 3039
+    },
+    {
+      "epoch": 0.5813175255760589,
+      "grad_norm": 4.9081730080776165,
+      "learning_rate": 1e-06,
+      "loss": 0.4681,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5815087484463142,
+      "grad_norm": 3.7374975963572044,
+      "learning_rate": 1e-06,
+      "loss": 0.2855,
+      "step": 3041
+    },
+    {
+      "epoch": 0.5816999713165695,
+      "grad_norm": 6.071096401584642,
+      "learning_rate": 1e-06,
+      "loss": 0.1612,
+      "step": 3042
+    },
+    {
+      "epoch": 0.5818911941868248,
+      "grad_norm": 4.103796359685453,
+      "learning_rate": 1e-06,
+      "loss": 0.6905,
+      "step": 3043
+    },
+    {
+      "epoch": 0.5820824170570801,
+      "grad_norm": 3.3873035669558633,
+      "learning_rate": 1e-06,
+      "loss": 0.2094,
+      "step": 3044
+    },
+    {
+      "epoch": 0.5822736399273353,
+      "grad_norm": 5.7318146106158,
+      "learning_rate": 1e-06,
+      "loss": 0.3026,
+      "step": 3045
+    },
+    {
+      "epoch": 0.5824648627975906,
+      "grad_norm": 4.01726240327874,
+      "learning_rate": 1e-06,
+      "loss": 0.1354,
+      "step": 3046
+    },
+    {
+      "epoch": 0.5826560856678459,
+      "grad_norm": 2.3279801906123043,
+      "learning_rate": 1e-06,
+      "loss": 0.0871,
+      "step": 3047
+    },
+    {
+      "epoch": 0.5828473085381012,
+      "grad_norm": 3.3790410185822295,
+      "learning_rate": 1e-06,
+      "loss": 0.1659,
+      "step": 3048
+    },
+    {
+      "epoch": 0.5830385314083565,
+      "grad_norm": 3.903902859292229,
+      "learning_rate": 1e-06,
+      "loss": 0.0916,
+      "step": 3049
+    },
+    {
+      "epoch": 0.5832297542786117,
+      "grad_norm": 4.053629654501567,
+      "learning_rate": 1e-06,
+      "loss": 0.1018,
+      "step": 3050
+    },
+    {
+      "epoch": 0.583420977148867,
+      "grad_norm": 5.801293182902769,
+      "learning_rate": 1e-06,
+      "loss": 0.4126,
+      "step": 3051
+    },
+    {
+      "epoch": 0.5836122000191223,
+      "grad_norm": 4.343305606202168,
+      "learning_rate": 1e-06,
+      "loss": 0.3177,
+      "step": 3052
+    },
+    {
+      "epoch": 0.5838034228893776,
+      "grad_norm": 3.9775864157888963,
+      "learning_rate": 1e-06,
+      "loss": 0.3335,
+      "step": 3053
+    },
+    {
+      "epoch": 0.5839946457596329,
+      "grad_norm": 2.385576989617978,
+      "learning_rate": 1e-06,
+      "loss": 0.1448,
+      "step": 3054
+    },
+    {
+      "epoch": 0.5841858686298881,
+      "grad_norm": 2.15480313616663,
+      "learning_rate": 1e-06,
+      "loss": 0.0961,
+      "step": 3055
+    },
+    {
+      "epoch": 0.5843770915001434,
+      "grad_norm": 5.692030558908211,
+      "learning_rate": 1e-06,
+      "loss": 0.1777,
+      "step": 3056
+    },
+    {
+      "epoch": 0.5845683143703987,
+      "grad_norm": 5.406968967501787,
+      "learning_rate": 1e-06,
+      "loss": 0.3559,
+      "step": 3057
+    },
+    {
+      "epoch": 0.584759537240654,
+      "grad_norm": 2.8920843616068463,
+      "learning_rate": 1e-06,
+      "loss": 0.2263,
+      "step": 3058
+    },
+    {
+      "epoch": 0.5849507601109093,
+      "grad_norm": 3.249687179769373,
+      "learning_rate": 1e-06,
+      "loss": 0.207,
+      "step": 3059
+    },
+    {
+      "epoch": 0.5851419829811646,
+      "grad_norm": 2.2146309240711632,
+      "learning_rate": 1e-06,
+      "loss": 0.0514,
+      "step": 3060
+    },
+    {
+      "epoch": 0.5853332058514198,
+      "grad_norm": 3.8614454391479627,
+      "learning_rate": 1e-06,
+      "loss": 0.1365,
+      "step": 3061
+    },
+    {
+      "epoch": 0.5855244287216751,
+      "grad_norm": 3.4436001836793744,
+      "learning_rate": 1e-06,
+      "loss": 0.0592,
+      "step": 3062
+    },
+    {
+      "epoch": 0.5857156515919304,
+      "grad_norm": 8.191719191089502,
+      "learning_rate": 1e-06,
+      "loss": 0.1626,
+      "step": 3063
+    },
+    {
+      "epoch": 0.5859068744621857,
+      "grad_norm": 4.969484670774869,
+      "learning_rate": 1e-06,
+      "loss": 0.4978,
+      "step": 3064
+    },
+    {
+      "epoch": 0.586098097332441,
+      "grad_norm": 4.30358717575765,
+      "learning_rate": 1e-06,
+      "loss": 0.316,
+      "step": 3065
+    },
+    {
+      "epoch": 0.5862893202026962,
+      "grad_norm": 2.2910311308951505,
+      "learning_rate": 1e-06,
+      "loss": 0.3367,
+      "step": 3066
+    },
+    {
+      "epoch": 0.5864805430729515,
+      "grad_norm": 3.4866491175319925,
+      "learning_rate": 1e-06,
+      "loss": 0.2761,
+      "step": 3067
+    },
+    {
+      "epoch": 0.5866717659432068,
+      "grad_norm": 2.6742437915927857,
+      "learning_rate": 1e-06,
+      "loss": 0.2928,
+      "step": 3068
+    },
+    {
+      "epoch": 0.5868629888134621,
+      "grad_norm": 4.766690094541228,
+      "learning_rate": 1e-06,
+      "loss": 0.3969,
+      "step": 3069
+    },
+    {
+      "epoch": 0.5870542116837174,
+      "grad_norm": 2.852704381282313,
+      "learning_rate": 1e-06,
+      "loss": 0.275,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5872454345539726,
+      "grad_norm": 2.6877542419825904,
+      "learning_rate": 1e-06,
+      "loss": 0.1589,
+      "step": 3071
+    },
+    {
+      "epoch": 0.5874366574242279,
+      "grad_norm": 2.8780652167358314,
+      "learning_rate": 1e-06,
+      "loss": 0.1041,
+      "step": 3072
+    },
+    {
+      "epoch": 0.5876278802944832,
+      "grad_norm": 2.438755103516041,
+      "learning_rate": 1e-06,
+      "loss": 0.0594,
+      "step": 3073
+    },
+    {
+      "epoch": 0.5878191031647385,
+      "grad_norm": 3.63902882961291,
+      "learning_rate": 1e-06,
+      "loss": 0.1233,
+      "step": 3074
+    },
+    {
+      "epoch": 0.5880103260349938,
+      "grad_norm": 4.067982422227121,
+      "learning_rate": 1e-06,
+      "loss": 0.0858,
+      "step": 3075
+    },
+    {
+      "epoch": 0.5882015489052491,
+      "grad_norm": 5.181087345013608,
+      "learning_rate": 1e-06,
+      "loss": 0.1698,
+      "step": 3076
+    },
+    {
+      "epoch": 0.5883927717755043,
+      "grad_norm": 3.739580428920174,
+      "learning_rate": 1e-06,
+      "loss": 0.6241,
+      "step": 3077
+    },
+    {
+      "epoch": 0.5885839946457596,
+      "grad_norm": 4.491082362362381,
+      "learning_rate": 1e-06,
+      "loss": 0.4616,
+      "step": 3078
+    },
+    {
+      "epoch": 0.5887752175160149,
+      "grad_norm": 1.7417224211948,
+      "learning_rate": 1e-06,
+      "loss": 0.2242,
+      "step": 3079
+    },
+    {
+      "epoch": 0.5889664403862702,
+      "grad_norm": 1.7797128168169176,
+      "learning_rate": 1e-06,
+      "loss": 0.1902,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5891576632565255,
+      "grad_norm": 2.569121203552865,
+      "learning_rate": 1e-06,
+      "loss": 0.1421,
+      "step": 3081
+    },
+    {
+      "epoch": 0.5893488861267807,
+      "grad_norm": 3.178157503905196,
+      "learning_rate": 1e-06,
+      "loss": 0.0724,
+      "step": 3082
+    },
+    {
+      "epoch": 0.589540108997036,
+      "grad_norm": 4.041281114783535,
+      "learning_rate": 1e-06,
+      "loss": 0.0746,
+      "step": 3083
+    },
+    {
+      "epoch": 0.5897313318672913,
+      "grad_norm": 3.1777727890271894,
+      "learning_rate": 1e-06,
+      "loss": 0.1351,
+      "step": 3084
+    },
+    {
+      "epoch": 0.5899225547375466,
+      "grad_norm": 2.6342018151967848,
+      "learning_rate": 1e-06,
+      "loss": 0.1593,
+      "step": 3085
+    },
+    {
+      "epoch": 0.5901137776078019,
+      "grad_norm": 3.581270259779899,
+      "learning_rate": 1e-06,
+      "loss": 0.0635,
+      "step": 3086
+    },
+    {
+      "epoch": 0.5903050004780572,
+      "grad_norm": 5.4837486148482055,
+      "learning_rate": 1e-06,
+      "loss": 0.1119,
+      "step": 3087
+    },
+    {
+      "epoch": 0.5904962233483124,
+      "grad_norm": 4.64505379242233,
+      "learning_rate": 1e-06,
+      "loss": 0.1639,
+      "step": 3088
+    },
+    {
+      "epoch": 0.5906874462185677,
+      "grad_norm": 5.674810989526057,
+      "learning_rate": 1e-06,
+      "loss": 0.2037,
+      "step": 3089
+    },
+    {
+      "epoch": 0.590878669088823,
+      "grad_norm": 3.398574440488109,
+      "learning_rate": 1e-06,
+      "loss": 0.2101,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5910698919590783,
+      "grad_norm": 5.840191252202678,
+      "learning_rate": 1e-06,
+      "loss": 0.5838,
+      "step": 3091
+    },
+    {
+      "epoch": 0.5912611148293336,
+      "grad_norm": 3.320713728055505,
+      "learning_rate": 1e-06,
+      "loss": 0.2676,
+      "step": 3092
+    },
+    {
+      "epoch": 0.5914523376995888,
+      "grad_norm": 4.474389206104024,
+      "learning_rate": 1e-06,
+      "loss": 0.1185,
+      "step": 3093
+    },
+    {
+      "epoch": 0.5916435605698441,
+      "grad_norm": 3.5342296081373474,
+      "learning_rate": 1e-06,
+      "loss": 0.1546,
+      "step": 3094
+    },
+    {
+      "epoch": 0.5918347834400994,
+      "grad_norm": 3.1118603458890783,
+      "learning_rate": 1e-06,
+      "loss": 0.1651,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5920260063103547,
+      "grad_norm": 4.121691677595044,
+      "learning_rate": 1e-06,
+      "loss": 0.4556,
+      "step": 3096
+    },
+    {
+      "epoch": 0.59221722918061,
+      "grad_norm": 5.032770152300201,
+      "learning_rate": 1e-06,
+      "loss": 0.0892,
+      "step": 3097
+    },
+    {
+      "epoch": 0.5924084520508652,
+      "grad_norm": 11.94120499401063,
+      "learning_rate": 1e-06,
+      "loss": 0.1702,
+      "step": 3098
+    },
+    {
+      "epoch": 0.5925996749211205,
+      "grad_norm": 3.386941482277206,
+      "learning_rate": 1e-06,
+      "loss": 0.113,
+      "step": 3099
+    },
+    {
+      "epoch": 0.5927908977913758,
+      "grad_norm": 7.891507241056333,
+      "learning_rate": 1e-06,
+      "loss": 0.1494,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5929821206616311,
+      "grad_norm": 4.282793198506876,
+      "learning_rate": 1e-06,
+      "loss": 0.2858,
+      "step": 3101
+    },
+    {
+      "epoch": 0.5931733435318864,
+      "grad_norm": 5.18215023084189,
+      "learning_rate": 1e-06,
+      "loss": 0.5662,
+      "step": 3102
+    },
+    {
+      "epoch": 0.5933645664021417,
+      "grad_norm": 5.242685763333668,
+      "learning_rate": 1e-06,
+      "loss": 0.6137,
+      "step": 3103
+    },
+    {
+      "epoch": 0.5935557892723969,
+      "grad_norm": 4.5153100983003895,
+      "learning_rate": 1e-06,
+      "loss": 0.1645,
+      "step": 3104
+    },
+    {
+      "epoch": 0.5937470121426522,
+      "grad_norm": 3.306192114166442,
+      "learning_rate": 1e-06,
+      "loss": 0.3559,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5939382350129075,
+      "grad_norm": 4.250482419627277,
+      "learning_rate": 1e-06,
+      "loss": 0.2569,
+      "step": 3106
+    },
+    {
+      "epoch": 0.5941294578831628,
+      "grad_norm": 2.2217715468646904,
+      "learning_rate": 1e-06,
+      "loss": 0.1705,
+      "step": 3107
+    },
+    {
+      "epoch": 0.5943206807534182,
+      "grad_norm": 8.312104409348697,
+      "learning_rate": 1e-06,
+      "loss": 0.1882,
+      "step": 3108
+    },
+    {
+      "epoch": 0.5945119036236733,
+      "grad_norm": 4.34771218482381,
+      "learning_rate": 1e-06,
+      "loss": 0.1111,
+      "step": 3109
+    },
+    {
+      "epoch": 0.5947031264939286,
+      "grad_norm": 2.927918574040012,
+      "learning_rate": 1e-06,
+      "loss": 0.3147,
+      "step": 3110
+    },
+    {
+      "epoch": 0.594894349364184,
+      "grad_norm": 3.4728902216596507,
+      "learning_rate": 1e-06,
+      "loss": 0.121,
+      "step": 3111
+    },
+    {
+      "epoch": 0.5950855722344393,
+      "grad_norm": 1.5519839602752536,
+      "learning_rate": 1e-06,
+      "loss": 0.0251,
+      "step": 3112
+    },
+    {
+      "epoch": 0.5952767951046946,
+      "grad_norm": 6.727353004145302,
+      "learning_rate": 1e-06,
+      "loss": 0.4214,
+      "step": 3113
+    },
+    {
+      "epoch": 0.5954680179749499,
+      "grad_norm": 8.394355385111055,
+      "learning_rate": 1e-06,
+      "loss": 0.2435,
+      "step": 3114
+    },
+    {
+      "epoch": 0.595659240845205,
+      "grad_norm": 5.719761826685086,
+      "learning_rate": 1e-06,
+      "loss": 0.596,
+      "step": 3115
+    },
+    {
+      "epoch": 0.5958504637154604,
+      "grad_norm": 3.221146071344709,
+      "learning_rate": 1e-06,
+      "loss": 0.282,
+      "step": 3116
+    },
+    {
+      "epoch": 0.5960416865857157,
+      "grad_norm": 2.739693488358917,
+      "learning_rate": 1e-06,
+      "loss": 0.3748,
+      "step": 3117
+    },
+    {
+      "epoch": 0.596232909455971,
+      "grad_norm": 3.5295237144779197,
+      "learning_rate": 1e-06,
+      "loss": 0.3971,
+      "step": 3118
+    },
+    {
+      "epoch": 0.5964241323262263,
+      "grad_norm": 5.282698032117221,
+      "learning_rate": 1e-06,
+      "loss": 0.4169,
+      "step": 3119
+    },
+    {
+      "epoch": 0.5966153551964815,
+      "grad_norm": 3.6883047486542258,
+      "learning_rate": 1e-06,
+      "loss": 0.3997,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5968065780667368,
+      "grad_norm": 2.7906390524037286,
+      "learning_rate": 1e-06,
+      "loss": 0.0798,
+      "step": 3121
+    },
+    {
+      "epoch": 0.5969978009369921,
+      "grad_norm": 3.238392127408572,
+      "learning_rate": 1e-06,
+      "loss": 0.1008,
+      "step": 3122
+    },
+    {
+      "epoch": 0.5971890238072474,
+      "grad_norm": 4.212264211151278,
+      "learning_rate": 1e-06,
+      "loss": 0.3561,
+      "step": 3123
+    },
+    {
+      "epoch": 0.5973802466775027,
+      "grad_norm": 3.288659585299299,
+      "learning_rate": 1e-06,
+      "loss": 0.0707,
+      "step": 3124
+    },
+    {
+      "epoch": 0.5975714695477579,
+      "grad_norm": 4.5944867224531025,
+      "learning_rate": 1e-06,
+      "loss": 0.0565,
+      "step": 3125
+    },
+    {
+      "epoch": 0.5977626924180132,
+      "grad_norm": 3.991650091679723,
+      "learning_rate": 1e-06,
+      "loss": 0.1663,
+      "step": 3126
+    },
+    {
+      "epoch": 0.5979539152882685,
+      "grad_norm": 6.989356532351754,
+      "learning_rate": 1e-06,
+      "loss": 0.6475,
+      "step": 3127
+    },
+    {
+      "epoch": 0.5981451381585238,
+      "grad_norm": 3.0738534480089363,
+      "learning_rate": 1e-06,
+      "loss": 0.1423,
+      "step": 3128
+    },
+    {
+      "epoch": 0.5983363610287791,
+      "grad_norm": 2.3470818805560287,
+      "learning_rate": 1e-06,
+      "loss": 0.2423,
+      "step": 3129
+    },
+    {
+      "epoch": 0.5985275838990344,
+      "grad_norm": 2.3953361216436067,
+      "learning_rate": 1e-06,
+      "loss": 0.2402,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5987188067692896,
+      "grad_norm": 2.722014560909972,
+      "learning_rate": 1e-06,
+      "loss": 0.2903,
+      "step": 3131
+    },
+    {
+      "epoch": 0.5989100296395449,
+      "grad_norm": 4.53559446230203,
+      "learning_rate": 1e-06,
+      "loss": 0.3419,
+      "step": 3132
+    },
+    {
+      "epoch": 0.5991012525098002,
+      "grad_norm": 1.7495130133830334,
+      "learning_rate": 1e-06,
+      "loss": 0.0694,
+      "step": 3133
+    },
+    {
+      "epoch": 0.5992924753800555,
+      "grad_norm": 2.5712840554333134,
+      "learning_rate": 1e-06,
+      "loss": 0.1685,
+      "step": 3134
+    },
+    {
+      "epoch": 0.5994836982503108,
+      "grad_norm": 3.0763496662411334,
+      "learning_rate": 1e-06,
+      "loss": 0.2056,
+      "step": 3135
+    },
+    {
+      "epoch": 0.599674921120566,
+      "grad_norm": 7.660401009204817,
+      "learning_rate": 1e-06,
+      "loss": 0.1576,
+      "step": 3136
+    },
+    {
+      "epoch": 0.5998661439908213,
+      "grad_norm": 3.324511791899674,
+      "learning_rate": 1e-06,
+      "loss": 0.0742,
+      "step": 3137
+    },
+    {
+      "epoch": 0.6000573668610766,
+      "grad_norm": 5.567420411999198,
+      "learning_rate": 1e-06,
+      "loss": 0.1819,
+      "step": 3138
+    },
+    {
+      "epoch": 0.6002485897313319,
+      "grad_norm": 4.959420328107381,
+      "learning_rate": 1e-06,
+      "loss": 0.4759,
+      "step": 3139
+    },
+    {
+      "epoch": 0.6004398126015872,
+      "grad_norm": 5.505916707476862,
+      "learning_rate": 1e-06,
+      "loss": 0.5914,
+      "step": 3140
+    },
+    {
+      "epoch": 0.6006310354718424,
+      "grad_norm": 5.271859030255617,
+      "learning_rate": 1e-06,
+      "loss": 0.2614,
+      "step": 3141
+    },
+    {
+      "epoch": 0.6008222583420977,
+      "grad_norm": 5.415921947440849,
+      "learning_rate": 1e-06,
+      "loss": 0.4672,
+      "step": 3142
+    },
+    {
+      "epoch": 0.601013481212353,
+      "grad_norm": 4.54255224337476,
+      "learning_rate": 1e-06,
+      "loss": 0.3971,
+      "step": 3143
+    },
+    {
+      "epoch": 0.6012047040826083,
+      "grad_norm": 3.962657186477158,
+      "learning_rate": 1e-06,
+      "loss": 0.2165,
+      "step": 3144
+    },
+    {
+      "epoch": 0.6013959269528636,
+      "grad_norm": 2.992263036028455,
+      "learning_rate": 1e-06,
+      "loss": 0.1656,
+      "step": 3145
+    },
+    {
+      "epoch": 0.6015871498231189,
+      "grad_norm": 4.659665282511604,
+      "learning_rate": 1e-06,
+      "loss": 0.1335,
+      "step": 3146
+    },
+    {
+      "epoch": 0.6017783726933741,
+      "grad_norm": 3.961201976847355,
+      "learning_rate": 1e-06,
+      "loss": 0.3033,
+      "step": 3147
+    },
+    {
+      "epoch": 0.6019695955636294,
+      "grad_norm": 3.2114221805728067,
+      "learning_rate": 1e-06,
+      "loss": 0.0863,
+      "step": 3148
+    },
+    {
+      "epoch": 0.6021608184338847,
+      "grad_norm": 4.839211977656824,
+      "learning_rate": 1e-06,
+      "loss": 0.1437,
+      "step": 3149
+    },
+    {
+      "epoch": 0.60235204130414,
+      "grad_norm": 2.854692972112212,
+      "learning_rate": 1e-06,
+      "loss": 0.1078,
+      "step": 3150
+    },
+    {
+      "epoch": 0.6025432641743953,
+      "grad_norm": 7.097517046825634,
+      "learning_rate": 1e-06,
+      "loss": 0.5228,
+      "step": 3151
+    },
+    {
+      "epoch": 0.6027344870446505,
+      "grad_norm": 5.768260322526845,
+      "learning_rate": 1e-06,
+      "loss": 0.5304,
+      "step": 3152
+    },
+    {
+      "epoch": 0.6029257099149058,
+      "grad_norm": 3.080975447178332,
+      "learning_rate": 1e-06,
+      "loss": 0.2995,
+      "step": 3153
+    },
+    {
+      "epoch": 0.6031169327851611,
+      "grad_norm": 2.1137008955455543,
+      "learning_rate": 1e-06,
+      "loss": 0.1735,
+      "step": 3154
+    },
+    {
+      "epoch": 0.6033081556554164,
+      "grad_norm": 4.495733781949191,
+      "learning_rate": 1e-06,
+      "loss": 0.643,
+      "step": 3155
+    },
+    {
+      "epoch": 0.6034993785256717,
+      "grad_norm": 2.5683891729794714,
+      "learning_rate": 1e-06,
+      "loss": 0.3311,
+      "step": 3156
+    },
+    {
+      "epoch": 0.603690601395927,
+      "grad_norm": 6.954151960676981,
+      "learning_rate": 1e-06,
+      "loss": 0.2338,
+      "step": 3157
+    },
+    {
+      "epoch": 0.6038818242661822,
+      "grad_norm": 3.0709510508058107,
+      "learning_rate": 1e-06,
+      "loss": 0.1662,
+      "step": 3158
+    },
+    {
+      "epoch": 0.6040730471364375,
+      "grad_norm": 2.018066465321043,
+      "learning_rate": 1e-06,
+      "loss": 0.1763,
+      "step": 3159
+    },
+    {
+      "epoch": 0.6042642700066928,
+      "grad_norm": 3.245757341454301,
+      "learning_rate": 1e-06,
+      "loss": 0.0838,
+      "step": 3160
+    },
+    {
+      "epoch": 0.6044554928769481,
+      "grad_norm": 2.4935922520203535,
+      "learning_rate": 1e-06,
+      "loss": 0.0698,
+      "step": 3161
+    },
+    {
+      "epoch": 0.6046467157472034,
+      "grad_norm": 4.6664695471049225,
+      "learning_rate": 1e-06,
+      "loss": 0.0977,
+      "step": 3162
+    },
+    {
+      "epoch": 0.6048379386174586,
+      "grad_norm": 6.964725444171889,
+      "learning_rate": 1e-06,
+      "loss": 0.3655,
+      "step": 3163
+    },
+    {
+      "epoch": 0.6050291614877139,
+      "grad_norm": 5.423373056620695,
+      "learning_rate": 1e-06,
+      "loss": 0.6169,
+      "step": 3164
+    },
+    {
+      "epoch": 0.6052203843579692,
+      "grad_norm": 4.622080319520912,
+      "learning_rate": 1e-06,
+      "loss": 0.4968,
+      "step": 3165
+    },
+    {
+      "epoch": 0.6054116072282245,
+      "grad_norm": 2.690265651901595,
+      "learning_rate": 1e-06,
+      "loss": 0.1727,
+      "step": 3166
+    },
+    {
+      "epoch": 0.6056028300984798,
+      "grad_norm": 3.777316482903136,
+      "learning_rate": 1e-06,
+      "loss": 0.4536,
+      "step": 3167
+    },
+    {
+      "epoch": 0.605794052968735,
+      "grad_norm": 2.3339034700935777,
+      "learning_rate": 1e-06,
+      "loss": 0.1879,
+      "step": 3168
+    },
+    {
+      "epoch": 0.6059852758389903,
+      "grad_norm": 3.514985429020033,
+      "learning_rate": 1e-06,
+      "loss": 0.1485,
+      "step": 3169
+    },
+    {
+      "epoch": 0.6061764987092456,
+      "grad_norm": 3.463367306711243,
+      "learning_rate": 1e-06,
+      "loss": 0.1494,
+      "step": 3170
+    },
+    {
+      "epoch": 0.6063677215795009,
+      "grad_norm": 2.809317419970535,
+      "learning_rate": 1e-06,
+      "loss": 0.1549,
+      "step": 3171
+    },
+    {
+      "epoch": 0.6065589444497562,
+      "grad_norm": 5.444937169016382,
+      "learning_rate": 1e-06,
+      "loss": 0.2323,
+      "step": 3172
+    },
+    {
+      "epoch": 0.6067501673200115,
+      "grad_norm": 4.654274540638292,
+      "learning_rate": 1e-06,
+      "loss": 0.1504,
+      "step": 3173
+    },
+    {
+      "epoch": 0.6069413901902667,
+      "grad_norm": 2.5530086682266244,
+      "learning_rate": 1e-06,
+      "loss": 0.0765,
+      "step": 3174
+    },
+    {
+      "epoch": 0.607132613060522,
+      "grad_norm": 4.295593514517845,
+      "learning_rate": 1e-06,
+      "loss": 0.1104,
+      "step": 3175
+    },
+    {
+      "epoch": 0.6073238359307773,
+      "grad_norm": 4.45891187009876,
+      "learning_rate": 1e-06,
+      "loss": 0.2518,
+      "step": 3176
+    },
+    {
+      "epoch": 0.6075150588010326,
+      "grad_norm": 6.833991073339101,
+      "learning_rate": 1e-06,
+      "loss": 0.4992,
+      "step": 3177
+    },
+    {
+      "epoch": 0.6077062816712879,
+      "grad_norm": 3.0424514350369023,
+      "learning_rate": 1e-06,
+      "loss": 0.2258,
+      "step": 3178
+    },
+    {
+      "epoch": 0.6078975045415431,
+      "grad_norm": 3.1271798732493408,
+      "learning_rate": 1e-06,
+      "loss": 0.3215,
+      "step": 3179
+    },
+    {
+      "epoch": 0.6080887274117984,
+      "grad_norm": 2.719773373979999,
+      "learning_rate": 1e-06,
+      "loss": 0.0869,
+      "step": 3180
+    },
+    {
+      "epoch": 0.6082799502820537,
+      "grad_norm": 3.354106515538546,
+      "learning_rate": 1e-06,
+      "loss": 0.2738,
+      "step": 3181
+    },
+    {
+      "epoch": 0.608471173152309,
+      "grad_norm": 5.912660702221111,
+      "learning_rate": 1e-06,
+      "loss": 0.7063,
+      "step": 3182
+    },
+    {
+      "epoch": 0.6086623960225643,
+      "grad_norm": 2.6602623648071586,
+      "learning_rate": 1e-06,
+      "loss": 0.0955,
+      "step": 3183
+    },
+    {
+      "epoch": 0.6088536188928195,
+      "grad_norm": 2.313671949194347,
+      "learning_rate": 1e-06,
+      "loss": 0.2069,
+      "step": 3184
+    },
+    {
+      "epoch": 0.6090448417630748,
+      "grad_norm": 7.3154319688831375,
+      "learning_rate": 1e-06,
+      "loss": 0.1579,
+      "step": 3185
+    },
+    {
+      "epoch": 0.6092360646333301,
+      "grad_norm": 5.740393865809135,
+      "learning_rate": 1e-06,
+      "loss": 0.0723,
+      "step": 3186
+    },
+    {
+      "epoch": 0.6094272875035854,
+      "grad_norm": 3.5595013228482437,
+      "learning_rate": 1e-06,
+      "loss": 0.1064,
+      "step": 3187
+    },
+    {
+      "epoch": 0.6096185103738407,
+      "grad_norm": 10.811311689578593,
+      "learning_rate": 1e-06,
+      "loss": 0.2602,
+      "step": 3188
+    },
+    {
+      "epoch": 0.609809733244096,
+      "grad_norm": 5.335426495994958,
+      "learning_rate": 1e-06,
+      "loss": 0.3361,
+      "step": 3189
+    },
+    {
+      "epoch": 0.6100009561143512,
+      "grad_norm": 5.240458400855657,
+      "learning_rate": 1e-06,
+      "loss": 0.4819,
+      "step": 3190
+    },
+    {
+      "epoch": 0.6101921789846065,
+      "grad_norm": 2.4433694255326364,
+      "learning_rate": 1e-06,
+      "loss": 0.0972,
+      "step": 3191
+    },
+    {
+      "epoch": 0.6103834018548618,
+      "grad_norm": 3.2896646777470497,
+      "learning_rate": 1e-06,
+      "loss": 0.2867,
+      "step": 3192
+    },
+    {
+      "epoch": 0.6105746247251171,
+      "grad_norm": 2.012583841034605,
+      "learning_rate": 1e-06,
+      "loss": 0.2029,
+      "step": 3193
+    },
+    {
+      "epoch": 0.6107658475953724,
+      "grad_norm": 8.174829827733562,
+      "learning_rate": 1e-06,
+      "loss": 0.3254,
+      "step": 3194
+    },
+    {
+      "epoch": 0.6109570704656276,
+      "grad_norm": 2.9156448935988153,
+      "learning_rate": 1e-06,
+      "loss": 0.2857,
+      "step": 3195
+    },
+    {
+      "epoch": 0.6111482933358829,
+      "grad_norm": 1.727754483503757,
+      "learning_rate": 1e-06,
+      "loss": 0.0589,
+      "step": 3196
+    },
+    {
+      "epoch": 0.6113395162061382,
+      "grad_norm": 3.900451751367538,
+      "learning_rate": 1e-06,
+      "loss": 0.3669,
+      "step": 3197
+    },
+    {
+      "epoch": 0.6115307390763935,
+      "grad_norm": 2.799112884453017,
+      "learning_rate": 1e-06,
+      "loss": 0.0646,
+      "step": 3198
+    },
+    {
+      "epoch": 0.6117219619466489,
+      "grad_norm": 3.921855546515498,
+      "learning_rate": 1e-06,
+      "loss": 0.1019,
+      "step": 3199
+    },
+    {
+      "epoch": 0.6119131848169042,
+      "grad_norm": 2.924651922398202,
+      "learning_rate": 1e-06,
+      "loss": 0.0713,
+      "step": 3200
+    },
+    {
+      "epoch": 0.6121044076871593,
+      "grad_norm": 6.307314716510146,
+      "learning_rate": 1e-06,
+      "loss": 0.2465,
+      "step": 3201
+    },
+    {
+      "epoch": 0.6122956305574146,
+      "grad_norm": 4.271727504011729,
+      "learning_rate": 1e-06,
+      "loss": 0.3547,
+      "step": 3202
+    },
+    {
+      "epoch": 0.61248685342767,
+      "grad_norm": 4.075991719335768,
+      "learning_rate": 1e-06,
+      "loss": 0.379,
+      "step": 3203
+    },
+    {
+      "epoch": 0.6126780762979253,
+      "grad_norm": 3.358024858138483,
+      "learning_rate": 1e-06,
+      "loss": 0.2146,
+      "step": 3204
+    },
+    {
+      "epoch": 0.6128692991681806,
+      "grad_norm": 3.1499590734820377,
+      "learning_rate": 1e-06,
+      "loss": 0.2448,
+      "step": 3205
+    },
+    {
+      "epoch": 0.6130605220384358,
+      "grad_norm": 5.253244578165237,
+      "learning_rate": 1e-06,
+      "loss": 0.3826,
+      "step": 3206
+    },
+    {
+      "epoch": 0.613251744908691,
+      "grad_norm": 3.4343644146326575,
+      "learning_rate": 1e-06,
+      "loss": 0.288,
+      "step": 3207
+    },
+    {
+      "epoch": 0.6134429677789464,
+      "grad_norm": 2.517133177745921,
+      "learning_rate": 1e-06,
+      "loss": 0.168,
+      "step": 3208
+    },
+    {
+      "epoch": 0.6136341906492017,
+      "grad_norm": 3.4128794420724278,
+      "learning_rate": 1e-06,
+      "loss": 0.2944,
+      "step": 3209
+    },
+    {
+      "epoch": 0.613825413519457,
+      "grad_norm": 2.5635942123300137,
+      "learning_rate": 1e-06,
+      "loss": 0.1219,
+      "step": 3210
+    },
+    {
+      "epoch": 0.6140166363897122,
+      "grad_norm": 2.898961433704147,
+      "learning_rate": 1e-06,
+      "loss": 0.1046,
+      "step": 3211
+    },
+    {
+      "epoch": 0.6142078592599675,
+      "grad_norm": 2.6629133234991404,
+      "learning_rate": 1e-06,
+      "loss": 0.0461,
+      "step": 3212
+    },
+    {
+      "epoch": 0.6143990821302228,
+      "grad_norm": 5.823785470239193,
+      "learning_rate": 1e-06,
+      "loss": 0.2644,
+      "step": 3213
+    },
+    {
+      "epoch": 0.6145903050004781,
+      "grad_norm": 6.12743687323648,
+      "learning_rate": 1e-06,
+      "loss": 0.5311,
+      "step": 3214
+    },
+    {
+      "epoch": 0.6147815278707334,
+      "grad_norm": 3.6545327873476285,
+      "learning_rate": 1e-06,
+      "loss": 0.4936,
+      "step": 3215
+    },
+    {
+      "epoch": 0.6149727507409887,
+      "grad_norm": 2.507983335576523,
+      "learning_rate": 1e-06,
+      "loss": 0.3217,
+      "step": 3216
+    },
+    {
+      "epoch": 0.6151639736112439,
+      "grad_norm": 5.613612197750952,
+      "learning_rate": 1e-06,
+      "loss": 0.5688,
+      "step": 3217
+    },
+    {
+      "epoch": 0.6153551964814992,
+      "grad_norm": 3.4607514962185744,
+      "learning_rate": 1e-06,
+      "loss": 0.2929,
+      "step": 3218
+    },
+    {
+      "epoch": 0.6155464193517545,
+      "grad_norm": 4.3389365332138485,
+      "learning_rate": 1e-06,
+      "loss": 0.2887,
+      "step": 3219
+    },
+    {
+      "epoch": 0.6157376422220098,
+      "grad_norm": 3.3820819726684084,
+      "learning_rate": 1e-06,
+      "loss": 0.1727,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6159288650922651,
+      "grad_norm": 2.174921985849435,
+      "learning_rate": 1e-06,
+      "loss": 0.1057,
+      "step": 3221
+    },
+    {
+      "epoch": 0.6161200879625203,
+      "grad_norm": 3.460613984773703,
+      "learning_rate": 1e-06,
+      "loss": 0.0914,
+      "step": 3222
+    },
+    {
+      "epoch": 0.6163113108327756,
+      "grad_norm": 1.4809949437994616,
+      "learning_rate": 1e-06,
+      "loss": 0.0308,
+      "step": 3223
+    },
+    {
+      "epoch": 0.6165025337030309,
+      "grad_norm": 5.05508746727075,
+      "learning_rate": 1e-06,
+      "loss": 0.0973,
+      "step": 3224
+    },
+    {
+      "epoch": 0.6166937565732862,
+      "grad_norm": 5.422485668960142,
+      "learning_rate": 1e-06,
+      "loss": 0.1418,
+      "step": 3225
+    },
+    {
+      "epoch": 0.6168849794435415,
+      "grad_norm": 4.232380640157347,
+      "learning_rate": 1e-06,
+      "loss": 0.3009,
+      "step": 3226
+    },
+    {
+      "epoch": 0.6170762023137968,
+      "grad_norm": 5.926802786661005,
+      "learning_rate": 1e-06,
+      "loss": 0.4941,
+      "step": 3227
+    },
+    {
+      "epoch": 0.617267425184052,
+      "grad_norm": 3.4898415692939375,
+      "learning_rate": 1e-06,
+      "loss": 0.2643,
+      "step": 3228
+    },
+    {
+      "epoch": 0.6174586480543073,
+      "grad_norm": 1.8696502660416388,
+      "learning_rate": 1e-06,
+      "loss": 0.2707,
+      "step": 3229
+    },
+    {
+      "epoch": 0.6176498709245626,
+      "grad_norm": 3.8624062931088536,
+      "learning_rate": 1e-06,
+      "loss": 0.2768,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6178410937948179,
+      "grad_norm": 3.2079366133175817,
+      "learning_rate": 1e-06,
+      "loss": 0.2204,
+      "step": 3231
+    },
+    {
+      "epoch": 0.6180323166650732,
+      "grad_norm": 3.678650144880284,
+      "learning_rate": 1e-06,
+      "loss": 0.094,
+      "step": 3232
+    },
+    {
+      "epoch": 0.6182235395353284,
+      "grad_norm": 2.816281214040316,
+      "learning_rate": 1e-06,
+      "loss": 0.1943,
+      "step": 3233
+    },
+    {
+      "epoch": 0.6184147624055837,
+      "grad_norm": 3.651923595067929,
+      "learning_rate": 1e-06,
+      "loss": 0.0984,
+      "step": 3234
+    },
+    {
+      "epoch": 0.618605985275839,
+      "grad_norm": 3.6940429313160075,
+      "learning_rate": 1e-06,
+      "loss": 0.1046,
+      "step": 3235
+    },
+    {
+      "epoch": 0.6187972081460943,
+      "grad_norm": 3.534958773835254,
+      "learning_rate": 1e-06,
+      "loss": 0.0896,
+      "step": 3236
+    },
+    {
+      "epoch": 0.6189884310163496,
+      "grad_norm": 4.21802452348695,
+      "learning_rate": 1e-06,
+      "loss": 0.1076,
+      "step": 3237
+    },
+    {
+      "epoch": 0.6191796538866048,
+      "grad_norm": 6.22713101231468,
+      "learning_rate": 1e-06,
+      "loss": 0.4088,
+      "step": 3238
+    },
+    {
+      "epoch": 0.6193708767568601,
+      "grad_norm": 3.8503248845537397,
+      "learning_rate": 1e-06,
+      "loss": 0.4244,
+      "step": 3239
+    },
+    {
+      "epoch": 0.6195620996271154,
+      "grad_norm": 3.4690400423326935,
+      "learning_rate": 1e-06,
+      "loss": 0.3349,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6197533224973707,
+      "grad_norm": 3.648860669968473,
+      "learning_rate": 1e-06,
+      "loss": 0.3192,
+      "step": 3241
+    },
+    {
+      "epoch": 0.619944545367626,
+      "grad_norm": 3.8357119989248787,
+      "learning_rate": 1e-06,
+      "loss": 0.1266,
+      "step": 3242
+    },
+    {
+      "epoch": 0.6201357682378813,
+      "grad_norm": 5.15974428090303,
+      "learning_rate": 1e-06,
+      "loss": 0.4352,
+      "step": 3243
+    },
+    {
+      "epoch": 0.6203269911081365,
+      "grad_norm": 3.23496878513093,
+      "learning_rate": 1e-06,
+      "loss": 0.1273,
+      "step": 3244
+    },
+    {
+      "epoch": 0.6205182139783918,
+      "grad_norm": 1.6278442287154002,
+      "learning_rate": 1e-06,
+      "loss": 0.0855,
+      "step": 3245
+    },
+    {
+      "epoch": 0.6207094368486471,
+      "grad_norm": 1.938771599674481,
+      "learning_rate": 1e-06,
+      "loss": 0.0532,
+      "step": 3246
+    },
+    {
+      "epoch": 0.6209006597189024,
+      "grad_norm": 4.088458652827956,
+      "learning_rate": 1e-06,
+      "loss": 0.3237,
+      "step": 3247
+    },
+    {
+      "epoch": 0.6210918825891577,
+      "grad_norm": 1.9617478113669025,
+      "learning_rate": 1e-06,
+      "loss": 0.0389,
+      "step": 3248
+    },
+    {
+      "epoch": 0.6212831054594129,
+      "grad_norm": 3.797743709697364,
+      "learning_rate": 1e-06,
+      "loss": 0.2956,
+      "step": 3249
+    },
+    {
+      "epoch": 0.6214743283296682,
+      "grad_norm": 6.409305792722711,
+      "learning_rate": 1e-06,
+      "loss": 0.1196,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6216655511999235,
+      "grad_norm": 8.454108725684993,
+      "learning_rate": 1e-06,
+      "loss": 0.3046,
+      "step": 3251
+    },
+    {
+      "epoch": 0.6218567740701788,
+      "grad_norm": 4.260780126213455,
+      "learning_rate": 1e-06,
+      "loss": 0.3659,
+      "step": 3252
+    },
+    {
+      "epoch": 0.6220479969404341,
+      "grad_norm": 4.5437852776796355,
+      "learning_rate": 1e-06,
+      "loss": 0.337,
+      "step": 3253
+    },
+    {
+      "epoch": 0.6222392198106893,
+      "grad_norm": 2.406164291638905,
+      "learning_rate": 1e-06,
+      "loss": 0.1685,
+      "step": 3254
+    },
+    {
+      "epoch": 0.6224304426809446,
+      "grad_norm": 4.2202026126908985,
+      "learning_rate": 1e-06,
+      "loss": 0.1361,
+      "step": 3255
+    },
+    {
+      "epoch": 0.6226216655511999,
+      "grad_norm": 2.529223914548555,
+      "learning_rate": 1e-06,
+      "loss": 0.1574,
+      "step": 3256
+    },
+    {
+      "epoch": 0.6228128884214552,
+      "grad_norm": 4.7924277102053345,
+      "learning_rate": 1e-06,
+      "loss": 0.3791,
+      "step": 3257
+    },
+    {
+      "epoch": 0.6230041112917105,
+      "grad_norm": 1.9070709211887227,
+      "learning_rate": 1e-06,
+      "loss": 0.222,
+      "step": 3258
+    },
+    {
+      "epoch": 0.6231953341619658,
+      "grad_norm": 3.3838269692591765,
+      "learning_rate": 1e-06,
+      "loss": 0.1843,
+      "step": 3259
+    },
+    {
+      "epoch": 0.623386557032221,
+      "grad_norm": 3.368647495614083,
+      "learning_rate": 1e-06,
+      "loss": 0.1311,
+      "step": 3260
+    },
+    {
+      "epoch": 0.6235777799024763,
+      "grad_norm": 3.738056746132232,
+      "learning_rate": 1e-06,
+      "loss": 0.0838,
+      "step": 3261
+    },
+    {
+      "epoch": 0.6237690027727316,
+      "grad_norm": 2.579543030065116,
+      "learning_rate": 1e-06,
+      "loss": 0.037,
+      "step": 3262
+    },
+    {
+      "epoch": 0.6239602256429869,
+      "grad_norm": 4.926827981258244,
+      "learning_rate": 1e-06,
+      "loss": 0.2276,
+      "step": 3263
+    },
+    {
+      "epoch": 0.6241514485132422,
+      "grad_norm": 11.2270062597143,
+      "learning_rate": 1e-06,
+      "loss": 0.2369,
+      "step": 3264
+    },
+    {
+      "epoch": 0.6243426713834974,
+      "grad_norm": 4.851637196234558,
+      "learning_rate": 1e-06,
+      "loss": 0.4108,
+      "step": 3265
+    },
+    {
+      "epoch": 0.6245338942537527,
+      "grad_norm": 3.3967009920052362,
+      "learning_rate": 1e-06,
+      "loss": 0.2438,
+      "step": 3266
+    },
+    {
+      "epoch": 0.624725117124008,
+      "grad_norm": 2.9996927421898376,
+      "learning_rate": 1e-06,
+      "loss": 0.254,
+      "step": 3267
+    },
+    {
+      "epoch": 0.6249163399942633,
+      "grad_norm": 7.1801130731121345,
+      "learning_rate": 1e-06,
+      "loss": 0.4218,
+      "step": 3268
+    },
+    {
+      "epoch": 0.6251075628645186,
+      "grad_norm": 3.2994228840970092,
+      "learning_rate": 1e-06,
+      "loss": 0.0951,
+      "step": 3269
+    },
+    {
+      "epoch": 0.6252987857347739,
+      "grad_norm": 2.6370956038560593,
+      "learning_rate": 1e-06,
+      "loss": 0.1642,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6254900086050291,
+      "grad_norm": 3.0531818552492322,
+      "learning_rate": 1e-06,
+      "loss": 0.1531,
+      "step": 3271
+    },
+    {
+      "epoch": 0.6256812314752844,
+      "grad_norm": 5.098235980864082,
+      "learning_rate": 1e-06,
+      "loss": 0.0916,
+      "step": 3272
+    },
+    {
+      "epoch": 0.6258724543455397,
+      "grad_norm": 4.371628906653143,
+      "learning_rate": 1e-06,
+      "loss": 0.1572,
+      "step": 3273
+    },
+    {
+      "epoch": 0.626063677215795,
+      "grad_norm": 1.9635445943012892,
+      "learning_rate": 1e-06,
+      "loss": 0.0729,
+      "step": 3274
+    },
+    {
+      "epoch": 0.6262549000860503,
+      "grad_norm": 4.700270080922815,
+      "learning_rate": 1e-06,
+      "loss": 0.1552,
+      "step": 3275
+    },
+    {
+      "epoch": 0.6264461229563055,
+      "grad_norm": 6.706573097430732,
+      "learning_rate": 1e-06,
+      "loss": 0.3018,
+      "step": 3276
+    },
+    {
+      "epoch": 0.6266373458265608,
+      "grad_norm": 5.611770500912098,
+      "learning_rate": 1e-06,
+      "loss": 0.4701,
+      "step": 3277
+    },
+    {
+      "epoch": 0.6268285686968161,
+      "grad_norm": 6.577197084197417,
+      "learning_rate": 1e-06,
+      "loss": 0.5662,
+      "step": 3278
+    },
+    {
+      "epoch": 0.6270197915670714,
+      "grad_norm": 2.718050461144195,
+      "learning_rate": 1e-06,
+      "loss": 0.2517,
+      "step": 3279
+    },
+    {
+      "epoch": 0.6272110144373267,
+      "grad_norm": 2.9826305144052667,
+      "learning_rate": 1e-06,
+      "loss": 0.3392,
+      "step": 3280
+    },
+    {
+      "epoch": 0.6274022373075819,
+      "grad_norm": 2.6270768260547497,
+      "learning_rate": 1e-06,
+      "loss": 0.0965,
+      "step": 3281
+    },
+    {
+      "epoch": 0.6275934601778372,
+      "grad_norm": 3.4775029999797504,
+      "learning_rate": 1e-06,
+      "loss": 0.1214,
+      "step": 3282
+    },
+    {
+      "epoch": 0.6277846830480925,
+      "grad_norm": 1.9287451786913044,
+      "learning_rate": 1e-06,
+      "loss": 0.2218,
+      "step": 3283
+    },
+    {
+      "epoch": 0.6279759059183478,
+      "grad_norm": 4.464363986827773,
+      "learning_rate": 1e-06,
+      "loss": 0.119,
+      "step": 3284
+    },
+    {
+      "epoch": 0.6281671287886031,
+      "grad_norm": 3.047365902210558,
+      "learning_rate": 1e-06,
+      "loss": 0.0825,
+      "step": 3285
+    },
+    {
+      "epoch": 0.6283583516588584,
+      "grad_norm": 3.0247106114200273,
+      "learning_rate": 1e-06,
+      "loss": 0.0957,
+      "step": 3286
+    },
+    {
+      "epoch": 0.6285495745291136,
+      "grad_norm": 2.765534135435731,
+      "learning_rate": 1e-06,
+      "loss": 0.0683,
+      "step": 3287
+    },
+    {
+      "epoch": 0.628740797399369,
+      "grad_norm": 3.8472735374385585,
+      "learning_rate": 1e-06,
+      "loss": 0.4176,
+      "step": 3288
+    },
+    {
+      "epoch": 0.6289320202696242,
+      "grad_norm": 4.05600794935482,
+      "learning_rate": 1e-06,
+      "loss": 0.2874,
+      "step": 3289
+    },
+    {
+      "epoch": 0.6291232431398796,
+      "grad_norm": 4.583609046734957,
+      "learning_rate": 1e-06,
+      "loss": 0.1854,
+      "step": 3290
+    },
+    {
+      "epoch": 0.6293144660101349,
+      "grad_norm": 2.950593862726829,
+      "learning_rate": 1e-06,
+      "loss": 0.3411,
+      "step": 3291
+    },
+    {
+      "epoch": 0.62950568888039,
+      "grad_norm": 3.183302565287813,
+      "learning_rate": 1e-06,
+      "loss": 0.2853,
+      "step": 3292
+    },
+    {
+      "epoch": 0.6296969117506453,
+      "grad_norm": 2.8235730373527836,
+      "learning_rate": 1e-06,
+      "loss": 0.1802,
+      "step": 3293
+    },
+    {
+      "epoch": 0.6298881346209007,
+      "grad_norm": 4.0932395558586,
+      "learning_rate": 1e-06,
+      "loss": 0.1848,
+      "step": 3294
+    },
+    {
+      "epoch": 0.630079357491156,
+      "grad_norm": 3.3651966838973135,
+      "learning_rate": 1e-06,
+      "loss": 0.0866,
+      "step": 3295
+    },
+    {
+      "epoch": 0.6302705803614113,
+      "grad_norm": 3.8803426766234064,
+      "learning_rate": 1e-06,
+      "loss": 0.0953,
+      "step": 3296
+    },
+    {
+      "epoch": 0.6304618032316666,
+      "grad_norm": 4.473044299487294,
+      "learning_rate": 1e-06,
+      "loss": 0.173,
+      "step": 3297
+    },
+    {
+      "epoch": 0.6306530261019218,
+      "grad_norm": 2.6108567833194853,
+      "learning_rate": 1e-06,
+      "loss": 0.0736,
+      "step": 3298
+    },
+    {
+      "epoch": 0.6308442489721771,
+      "grad_norm": 5.909809634374933,
+      "learning_rate": 1e-06,
+      "loss": 0.0972,
+      "step": 3299
+    },
+    {
+      "epoch": 0.6310354718424324,
+      "grad_norm": 3.843426450432324,
+      "learning_rate": 1e-06,
+      "loss": 0.0774,
+      "step": 3300
+    },
+    {
+      "epoch": 0.6312266947126877,
+      "grad_norm": 5.319782247460923,
+      "learning_rate": 1e-06,
+      "loss": 0.29,
+      "step": 3301
+    },
+    {
+      "epoch": 0.631417917582943,
+      "grad_norm": 5.439879674983253,
+      "learning_rate": 1e-06,
+      "loss": 0.4256,
+      "step": 3302
+    },
+    {
+      "epoch": 0.6316091404531982,
+      "grad_norm": 2.7133380913595087,
+      "learning_rate": 1e-06,
+      "loss": 0.4205,
+      "step": 3303
+    },
+    {
+      "epoch": 0.6318003633234535,
+      "grad_norm": 1.4095386198692872,
+      "learning_rate": 1e-06,
+      "loss": 0.3012,
+      "step": 3304
+    },
+    {
+      "epoch": 0.6319915861937088,
+      "grad_norm": 2.6177119854012174,
+      "learning_rate": 1e-06,
+      "loss": 0.2322,
+      "step": 3305
+    },
+    {
+      "epoch": 0.6321828090639641,
+      "grad_norm": 2.9927869549860326,
+      "learning_rate": 1e-06,
+      "loss": 0.1338,
+      "step": 3306
+    },
+    {
+      "epoch": 0.6323740319342194,
+      "grad_norm": 2.7307026644205314,
+      "learning_rate": 1e-06,
+      "loss": 0.1806,
+      "step": 3307
+    },
+    {
+      "epoch": 0.6325652548044746,
+      "grad_norm": 2.7359159650668574,
+      "learning_rate": 1e-06,
+      "loss": 0.1339,
+      "step": 3308
+    },
+    {
+      "epoch": 0.6327564776747299,
+      "grad_norm": 3.6002148405182965,
+      "learning_rate": 1e-06,
+      "loss": 0.2708,
+      "step": 3309
+    },
+    {
+      "epoch": 0.6329477005449852,
+      "grad_norm": 5.186302506444692,
+      "learning_rate": 1e-06,
+      "loss": 0.1602,
+      "step": 3310
+    },
+    {
+      "epoch": 0.6331389234152405,
+      "grad_norm": 3.1495694002389256,
+      "learning_rate": 1e-06,
+      "loss": 0.1054,
+      "step": 3311
+    },
+    {
+      "epoch": 0.6333301462854958,
+      "grad_norm": 4.177279853795854,
+      "learning_rate": 1e-06,
+      "loss": 0.0995,
+      "step": 3312
+    },
+    {
+      "epoch": 0.6335213691557511,
+      "grad_norm": 5.138553839864437,
+      "learning_rate": 1e-06,
+      "loss": 0.1812,
+      "step": 3313
+    },
+    {
+      "epoch": 0.6337125920260063,
+      "grad_norm": 3.1278467563350545,
+      "learning_rate": 1e-06,
+      "loss": 0.2391,
+      "step": 3314
+    },
+    {
+      "epoch": 0.6339038148962616,
+      "grad_norm": 3.395822926760839,
+      "learning_rate": 1e-06,
+      "loss": 0.2659,
+      "step": 3315
+    },
+    {
+      "epoch": 0.6340950377665169,
+      "grad_norm": 2.6841396348506574,
+      "learning_rate": 1e-06,
+      "loss": 0.2783,
+      "step": 3316
+    },
+    {
+      "epoch": 0.6342862606367722,
+      "grad_norm": 3.296216564507969,
+      "learning_rate": 1e-06,
+      "loss": 0.2323,
+      "step": 3317
+    },
+    {
+      "epoch": 0.6344774835070275,
+      "grad_norm": 3.843295683024313,
+      "learning_rate": 1e-06,
+      "loss": 0.4406,
+      "step": 3318
+    },
+    {
+      "epoch": 0.6346687063772827,
+      "grad_norm": 3.274608757585979,
+      "learning_rate": 1e-06,
+      "loss": 0.2437,
+      "step": 3319
+    },
+    {
+      "epoch": 0.634859929247538,
+      "grad_norm": 2.292015904489562,
+      "learning_rate": 1e-06,
+      "loss": 0.1293,
+      "step": 3320
+    },
+    {
+      "epoch": 0.6350511521177933,
+      "grad_norm": 2.7351777560476274,
+      "learning_rate": 1e-06,
+      "loss": 0.1268,
+      "step": 3321
+    },
+    {
+      "epoch": 0.6352423749880486,
+      "grad_norm": 3.9599808058610697,
+      "learning_rate": 1e-06,
+      "loss": 0.096,
+      "step": 3322
+    },
+    {
+      "epoch": 0.6354335978583039,
+      "grad_norm": 3.0205283350822367,
+      "learning_rate": 1e-06,
+      "loss": 0.164,
+      "step": 3323
+    },
+    {
+      "epoch": 0.6356248207285591,
+      "grad_norm": 2.2573006125570516,
+      "learning_rate": 1e-06,
+      "loss": 0.0431,
+      "step": 3324
+    },
+    {
+      "epoch": 0.6358160435988144,
+      "grad_norm": 4.3024894516509145,
+      "learning_rate": 1e-06,
+      "loss": 0.1585,
+      "step": 3325
+    },
+    {
+      "epoch": 0.6360072664690697,
+      "grad_norm": 4.962301039851112,
+      "learning_rate": 1e-06,
+      "loss": 0.4718,
+      "step": 3326
+    },
+    {
+      "epoch": 0.636198489339325,
+      "grad_norm": 3.7664526825159172,
+      "learning_rate": 1e-06,
+      "loss": 0.2735,
+      "step": 3327
+    },
+    {
+      "epoch": 0.6363897122095803,
+      "grad_norm": 4.898818687196873,
+      "learning_rate": 1e-06,
+      "loss": 0.4993,
+      "step": 3328
+    },
+    {
+      "epoch": 0.6365809350798356,
+      "grad_norm": 2.7671695857273217,
+      "learning_rate": 1e-06,
+      "loss": 0.1766,
+      "step": 3329
+    },
+    {
+      "epoch": 0.6367721579500908,
+      "grad_norm": 4.611901990380115,
+      "learning_rate": 1e-06,
+      "loss": 0.1813,
+      "step": 3330
+    },
+    {
+      "epoch": 0.6369633808203461,
+      "grad_norm": 2.904148121085442,
+      "learning_rate": 1e-06,
+      "loss": 0.1909,
+      "step": 3331
+    },
+    {
+      "epoch": 0.6371546036906014,
+      "grad_norm": 2.5757282560553136,
+      "learning_rate": 1e-06,
+      "loss": 0.0971,
+      "step": 3332
+    },
+    {
+      "epoch": 0.6373458265608567,
+      "grad_norm": 3.3124929464013197,
+      "learning_rate": 1e-06,
+      "loss": 0.1528,
+      "step": 3333
+    },
+    {
+      "epoch": 0.637537049431112,
+      "grad_norm": 6.70081067091594,
+      "learning_rate": 1e-06,
+      "loss": 0.1548,
+      "step": 3334
+    },
+    {
+      "epoch": 0.6377282723013672,
+      "grad_norm": 7.87691196059167,
+      "learning_rate": 1e-06,
+      "loss": 0.4014,
+      "step": 3335
+    },
+    {
+      "epoch": 0.6379194951716225,
+      "grad_norm": 3.883381915027342,
+      "learning_rate": 1e-06,
+      "loss": 0.1098,
+      "step": 3336
+    },
+    {
+      "epoch": 0.6381107180418778,
+      "grad_norm": 2.909725776939274,
+      "learning_rate": 1e-06,
+      "loss": 0.0581,
+      "step": 3337
+    },
+    {
+      "epoch": 0.6383019409121331,
+      "grad_norm": 7.045208854530508,
+      "learning_rate": 1e-06,
+      "loss": 0.2969,
+      "step": 3338
+    },
+    {
+      "epoch": 0.6384931637823884,
+      "grad_norm": 4.337127247385044,
+      "learning_rate": 1e-06,
+      "loss": 0.4583,
+      "step": 3339
+    },
+    {
+      "epoch": 0.6386843866526437,
+      "grad_norm": 4.124419431578584,
+      "learning_rate": 1e-06,
+      "loss": 0.3568,
+      "step": 3340
+    },
+    {
+      "epoch": 0.6388756095228989,
+      "grad_norm": 4.3027512198943985,
+      "learning_rate": 1e-06,
+      "loss": 0.4397,
+      "step": 3341
+    },
+    {
+      "epoch": 0.6390668323931542,
+      "grad_norm": 6.555356160955151,
+      "learning_rate": 1e-06,
+      "loss": 0.5993,
+      "step": 3342
+    },
+    {
+      "epoch": 0.6392580552634095,
+      "grad_norm": 3.003396654495273,
+      "learning_rate": 1e-06,
+      "loss": 0.1676,
+      "step": 3343
+    },
+    {
+      "epoch": 0.6394492781336648,
+      "grad_norm": 2.895245956751421,
+      "learning_rate": 1e-06,
+      "loss": 0.0937,
+      "step": 3344
+    },
+    {
+      "epoch": 0.6396405010039201,
+      "grad_norm": 3.41126155528414,
+      "learning_rate": 1e-06,
+      "loss": 0.0998,
+      "step": 3345
+    },
+    {
+      "epoch": 0.6398317238741753,
+      "grad_norm": 3.2659968844303706,
+      "learning_rate": 1e-06,
+      "loss": 0.0656,
+      "step": 3346
+    },
+    {
+      "epoch": 0.6400229467444306,
+      "grad_norm": 3.058081261193207,
+      "learning_rate": 1e-06,
+      "loss": 0.0905,
+      "step": 3347
+    },
+    {
+      "epoch": 0.6402141696146859,
+      "grad_norm": 5.322421693525967,
+      "learning_rate": 1e-06,
+      "loss": 0.169,
+      "step": 3348
+    },
+    {
+      "epoch": 0.6404053924849412,
+      "grad_norm": 2.3268740576681464,
+      "learning_rate": 1e-06,
+      "loss": 0.0355,
+      "step": 3349
+    },
+    {
+      "epoch": 0.6405966153551965,
+      "grad_norm": 3.9054282582933553,
+      "learning_rate": 1e-06,
+      "loss": 0.1048,
+      "step": 3350
+    },
+    {
+      "epoch": 0.6407878382254517,
+      "grad_norm": 6.450541794370297,
+      "learning_rate": 1e-06,
+      "loss": 0.4569,
+      "step": 3351
+    },
+    {
+      "epoch": 0.640979061095707,
+      "grad_norm": 9.511104719015254,
+      "learning_rate": 1e-06,
+      "loss": 0.4195,
+      "step": 3352
+    },
+    {
+      "epoch": 0.6411702839659623,
+      "grad_norm": 1.9751778075903166,
+      "learning_rate": 1e-06,
+      "loss": 0.3689,
+      "step": 3353
+    },
+    {
+      "epoch": 0.6413615068362176,
+      "grad_norm": 5.630131224869713,
+      "learning_rate": 1e-06,
+      "loss": 0.3228,
+      "step": 3354
+    },
+    {
+      "epoch": 0.6415527297064729,
+      "grad_norm": 2.4508236337534877,
+      "learning_rate": 1e-06,
+      "loss": 0.2002,
+      "step": 3355
+    },
+    {
+      "epoch": 0.6417439525767282,
+      "grad_norm": 2.5981391629954174,
+      "learning_rate": 1e-06,
+      "loss": 0.1526,
+      "step": 3356
+    },
+    {
+      "epoch": 0.6419351754469834,
+      "grad_norm": 4.998918034790598,
+      "learning_rate": 1e-06,
+      "loss": 0.1685,
+      "step": 3357
+    },
+    {
+      "epoch": 0.6421263983172387,
+      "grad_norm": 2.686197719429121,
+      "learning_rate": 1e-06,
+      "loss": 0.101,
+      "step": 3358
+    },
+    {
+      "epoch": 0.642317621187494,
+      "grad_norm": 2.7882031738431743,
+      "learning_rate": 1e-06,
+      "loss": 0.0772,
+      "step": 3359
+    },
+    {
+      "epoch": 0.6425088440577493,
+      "grad_norm": 3.7735960869803407,
+      "learning_rate": 1e-06,
+      "loss": 0.0582,
+      "step": 3360
+    },
+    {
+      "epoch": 0.6427000669280046,
+      "grad_norm": 3.995402555081082,
+      "learning_rate": 1e-06,
+      "loss": 0.1276,
+      "step": 3361
+    },
+    {
+      "epoch": 0.6428912897982598,
+      "grad_norm": 3.76433024960462,
+      "learning_rate": 1e-06,
+      "loss": 0.09,
+      "step": 3362
+    },
+    {
+      "epoch": 0.6430825126685151,
+      "grad_norm": 2.568848537959006,
+      "learning_rate": 1e-06,
+      "loss": 0.1368,
+      "step": 3363
+    },
+    {
+      "epoch": 0.6432737355387704,
+      "grad_norm": 5.098323524035847,
+      "learning_rate": 1e-06,
+      "loss": 0.3835,
+      "step": 3364
+    },
+    {
+      "epoch": 0.6434649584090257,
+      "grad_norm": 4.113690217382666,
+      "learning_rate": 1e-06,
+      "loss": 0.2715,
+      "step": 3365
+    },
+    {
+      "epoch": 0.643656181279281,
+      "grad_norm": 2.5829185542529314,
+      "learning_rate": 1e-06,
+      "loss": 0.2287,
+      "step": 3366
+    },
+    {
+      "epoch": 0.6438474041495363,
+      "grad_norm": 2.6201122737444607,
+      "learning_rate": 1e-06,
+      "loss": 0.1984,
+      "step": 3367
+    },
+    {
+      "epoch": 0.6440386270197915,
+      "grad_norm": 1.7491542270988034,
+      "learning_rate": 1e-06,
+      "loss": 0.1261,
+      "step": 3368
+    },
+    {
+      "epoch": 0.6442298498900468,
+      "grad_norm": 2.6711429239964417,
+      "learning_rate": 1e-06,
+      "loss": 0.1071,
+      "step": 3369
+    },
+    {
+      "epoch": 0.6444210727603021,
+      "grad_norm": 2.6235317483663656,
+      "learning_rate": 1e-06,
+      "loss": 0.1741,
+      "step": 3370
+    },
+    {
+      "epoch": 0.6446122956305574,
+      "grad_norm": 2.2086031287108527,
+      "learning_rate": 1e-06,
+      "loss": 0.1256,
+      "step": 3371
+    },
+    {
+      "epoch": 0.6448035185008127,
+      "grad_norm": 3.500459232175202,
+      "learning_rate": 1e-06,
+      "loss": 0.2336,
+      "step": 3372
+    },
+    {
+      "epoch": 0.6449947413710679,
+      "grad_norm": 4.085420710775477,
+      "learning_rate": 1e-06,
+      "loss": 0.3297,
+      "step": 3373
+    },
+    {
+      "epoch": 0.6451859642413232,
+      "grad_norm": 4.036863218940852,
+      "learning_rate": 1e-06,
+      "loss": 0.0997,
+      "step": 3374
+    },
+    {
+      "epoch": 0.6453771871115785,
+      "grad_norm": 4.979326423838051,
+      "learning_rate": 1e-06,
+      "loss": 0.0731,
+      "step": 3375
+    },
+    {
+      "epoch": 0.6455684099818338,
+      "grad_norm": 5.492243412433212,
+      "learning_rate": 1e-06,
+      "loss": 0.6198,
+      "step": 3376
+    },
+    {
+      "epoch": 0.6457596328520891,
+      "grad_norm": 3.890624387197178,
+      "learning_rate": 1e-06,
+      "loss": 0.3617,
+      "step": 3377
+    },
+    {
+      "epoch": 0.6459508557223443,
+      "grad_norm": 4.000632474487359,
+      "learning_rate": 1e-06,
+      "loss": 0.3049,
+      "step": 3378
+    },
+    {
+      "epoch": 0.6461420785925996,
+      "grad_norm": 4.052633658387074,
+      "learning_rate": 1e-06,
+      "loss": 0.1839,
+      "step": 3379
+    },
+    {
+      "epoch": 0.646333301462855,
+      "grad_norm": 4.756548734048144,
+      "learning_rate": 1e-06,
+      "loss": 0.6115,
+      "step": 3380
+    },
+    {
+      "epoch": 0.6465245243331103,
+      "grad_norm": 6.736199293400556,
+      "learning_rate": 1e-06,
+      "loss": 0.5059,
+      "step": 3381
+    },
+    {
+      "epoch": 0.6467157472033656,
+      "grad_norm": 3.5492418097406704,
+      "learning_rate": 1e-06,
+      "loss": 0.319,
+      "step": 3382
+    },
+    {
+      "epoch": 0.6469069700736209,
+      "grad_norm": 2.5108548068261363,
+      "learning_rate": 1e-06,
+      "loss": 0.2767,
+      "step": 3383
+    },
+    {
+      "epoch": 0.647098192943876,
+      "grad_norm": 2.9955024384406053,
+      "learning_rate": 1e-06,
+      "loss": 0.2151,
+      "step": 3384
+    },
+    {
+      "epoch": 0.6472894158141314,
+      "grad_norm": 3.128302698109007,
+      "learning_rate": 1e-06,
+      "loss": 0.3204,
+      "step": 3385
+    },
+    {
+      "epoch": 0.6474806386843867,
+      "grad_norm": 2.4835447447562813,
+      "learning_rate": 1e-06,
+      "loss": 0.0912,
+      "step": 3386
+    },
+    {
+      "epoch": 0.647671861554642,
+      "grad_norm": 4.271096992107673,
+      "learning_rate": 1e-06,
+      "loss": 0.0866,
+      "step": 3387
+    },
+    {
+      "epoch": 0.6478630844248973,
+      "grad_norm": 6.856442461120508,
+      "learning_rate": 1e-06,
+      "loss": 0.1412,
+      "step": 3388
+    },
+    {
+      "epoch": 0.6480543072951525,
+      "grad_norm": 9.2000248452556,
+      "learning_rate": 1e-06,
+      "loss": 0.4413,
+      "step": 3389
+    },
+    {
+      "epoch": 0.6482455301654078,
+      "grad_norm": 3.9161377542910176,
+      "learning_rate": 1e-06,
+      "loss": 0.3246,
+      "step": 3390
+    },
+    {
+      "epoch": 0.6484367530356631,
+      "grad_norm": 2.427155675069176,
+      "learning_rate": 1e-06,
+      "loss": 0.3419,
+      "step": 3391
+    },
+    {
+      "epoch": 0.6486279759059184,
+      "grad_norm": 3.1328041285833446,
+      "learning_rate": 1e-06,
+      "loss": 0.4643,
+      "step": 3392
+    },
+    {
+      "epoch": 0.6488191987761737,
+      "grad_norm": 3.1158719836321565,
+      "learning_rate": 1e-06,
+      "loss": 0.1587,
+      "step": 3393
+    },
+    {
+      "epoch": 0.6490104216464289,
+      "grad_norm": 4.814783916754631,
+      "learning_rate": 1e-06,
+      "loss": 0.344,
+      "step": 3394
+    },
+    {
+      "epoch": 0.6492016445166842,
+      "grad_norm": 3.008802374909693,
+      "learning_rate": 1e-06,
+      "loss": 0.1251,
+      "step": 3395
+    },
+    {
+      "epoch": 0.6493928673869395,
+      "grad_norm": 5.684048789773998,
+      "learning_rate": 1e-06,
+      "loss": 0.2001,
+      "step": 3396
+    },
+    {
+      "epoch": 0.6495840902571948,
+      "grad_norm": 2.5199129510407206,
+      "learning_rate": 1e-06,
+      "loss": 0.1753,
+      "step": 3397
+    },
+    {
+      "epoch": 0.6497753131274501,
+      "grad_norm": 2.332512529509195,
+      "learning_rate": 1e-06,
+      "loss": 0.047,
+      "step": 3398
+    },
+    {
+      "epoch": 0.6499665359977054,
+      "grad_norm": 2.400067399985904,
+      "learning_rate": 1e-06,
+      "loss": 0.0352,
+      "step": 3399
+    },
+    {
+      "epoch": 0.6501577588679606,
+      "grad_norm": 4.438876945809499,
+      "learning_rate": 1e-06,
+      "loss": 0.1331,
+      "step": 3400
+    },
+    {
+      "epoch": 0.6503489817382159,
+      "grad_norm": 4.257327550866413,
+      "learning_rate": 1e-06,
+      "loss": 0.7091,
+      "step": 3401
+    },
+    {
+      "epoch": 0.6505402046084712,
+      "grad_norm": 2.4280086494367152,
+      "learning_rate": 1e-06,
+      "loss": 0.3524,
+      "step": 3402
+    },
+    {
+      "epoch": 0.6507314274787265,
+      "grad_norm": 3.7111658287075375,
+      "learning_rate": 1e-06,
+      "loss": 0.3749,
+      "step": 3403
+    },
+    {
+      "epoch": 0.6509226503489818,
+      "grad_norm": 2.843784373987392,
+      "learning_rate": 1e-06,
+      "loss": 0.2291,
+      "step": 3404
+    },
+    {
+      "epoch": 0.651113873219237,
+      "grad_norm": 3.916178787903857,
+      "learning_rate": 1e-06,
+      "loss": 0.1643,
+      "step": 3405
+    },
+    {
+      "epoch": 0.6513050960894923,
+      "grad_norm": 2.3676822762175043,
+      "learning_rate": 1e-06,
+      "loss": 0.1816,
+      "step": 3406
+    },
+    {
+      "epoch": 0.6514963189597476,
+      "grad_norm": 4.233662788206926,
+      "learning_rate": 1e-06,
+      "loss": 0.2477,
+      "step": 3407
+    },
+    {
+      "epoch": 0.6516875418300029,
+      "grad_norm": 2.4519638909672836,
+      "learning_rate": 1e-06,
+      "loss": 0.0519,
+      "step": 3408
+    },
+    {
+      "epoch": 0.6518787647002582,
+      "grad_norm": 3.4393792304224933,
+      "learning_rate": 1e-06,
+      "loss": 0.1264,
+      "step": 3409
+    },
+    {
+      "epoch": 0.6520699875705135,
+      "grad_norm": 4.84311598043061,
+      "learning_rate": 1e-06,
+      "loss": 0.3691,
+      "step": 3410
+    },
+    {
+      "epoch": 0.6522612104407687,
+      "grad_norm": 5.499132261315544,
+      "learning_rate": 1e-06,
+      "loss": 0.1198,
+      "step": 3411
+    },
+    {
+      "epoch": 0.652452433311024,
+      "grad_norm": 2.910444288912314,
+      "learning_rate": 1e-06,
+      "loss": 0.07,
+      "step": 3412
+    },
+    {
+      "epoch": 0.6526436561812793,
+      "grad_norm": 2.9941081364431117,
+      "learning_rate": 1e-06,
+      "loss": 0.2883,
+      "step": 3413
+    },
+    {
+      "epoch": 0.6528348790515346,
+      "grad_norm": 5.9180470678604875,
+      "learning_rate": 1e-06,
+      "loss": 0.3649,
+      "step": 3414
+    },
+    {
+      "epoch": 0.6530261019217899,
+      "grad_norm": 4.012793110156799,
+      "learning_rate": 1e-06,
+      "loss": 0.2772,
+      "step": 3415
+    },
+    {
+      "epoch": 0.6532173247920451,
+      "grad_norm": 3.4342543104380328,
+      "learning_rate": 1e-06,
+      "loss": 0.3314,
+      "step": 3416
+    },
+    {
+      "epoch": 0.6534085476623004,
+      "grad_norm": 2.7405745498380836,
+      "learning_rate": 1e-06,
+      "loss": 0.3376,
+      "step": 3417
+    },
+    {
+      "epoch": 0.6535997705325557,
+      "grad_norm": 4.622389804316784,
+      "learning_rate": 1e-06,
+      "loss": 0.4087,
+      "step": 3418
+    },
+    {
+      "epoch": 0.653790993402811,
+      "grad_norm": 2.535711900567942,
+      "learning_rate": 1e-06,
+      "loss": 0.1171,
+      "step": 3419
+    },
+    {
+      "epoch": 0.6539822162730663,
+      "grad_norm": 1.871021946303788,
+      "learning_rate": 1e-06,
+      "loss": 0.1137,
+      "step": 3420
+    },
+    {
+      "epoch": 0.6541734391433215,
+      "grad_norm": 1.7909622434026076,
+      "learning_rate": 1e-06,
+      "loss": 0.0665,
+      "step": 3421
+    },
+    {
+      "epoch": 0.6543646620135768,
+      "grad_norm": 3.1515871863091007,
+      "learning_rate": 1e-06,
+      "loss": 0.1027,
+      "step": 3422
+    },
+    {
+      "epoch": 0.6545558848838321,
+      "grad_norm": 4.830890147636383,
+      "learning_rate": 1e-06,
+      "loss": 0.0998,
+      "step": 3423
+    },
+    {
+      "epoch": 0.6547471077540874,
+      "grad_norm": 3.766122618913048,
+      "learning_rate": 1e-06,
+      "loss": 0.1006,
+      "step": 3424
+    },
+    {
+      "epoch": 0.6549383306243427,
+      "grad_norm": 5.508319458051895,
+      "learning_rate": 1e-06,
+      "loss": 0.101,
+      "step": 3425
+    },
+    {
+      "epoch": 0.655129553494598,
+      "grad_norm": 4.078438114700814,
+      "learning_rate": 1e-06,
+      "loss": 0.4507,
+      "step": 3426
+    },
+    {
+      "epoch": 0.6553207763648532,
+      "grad_norm": 4.057983470042682,
+      "learning_rate": 1e-06,
+      "loss": 0.3246,
+      "step": 3427
+    },
+    {
+      "epoch": 0.6555119992351085,
+      "grad_norm": 2.7541988569062674,
+      "learning_rate": 1e-06,
+      "loss": 0.2978,
+      "step": 3428
+    },
+    {
+      "epoch": 0.6557032221053638,
+      "grad_norm": 3.0485332964730087,
+      "learning_rate": 1e-06,
+      "loss": 0.2142,
+      "step": 3429
+    },
+    {
+      "epoch": 0.6558944449756191,
+      "grad_norm": 4.893097371398441,
+      "learning_rate": 1e-06,
+      "loss": 0.3477,
+      "step": 3430
+    },
+    {
+      "epoch": 0.6560856678458744,
+      "grad_norm": 3.612704900190689,
+      "learning_rate": 1e-06,
+      "loss": 0.4726,
+      "step": 3431
+    },
+    {
+      "epoch": 0.6562768907161296,
+      "grad_norm": 3.3111684750132957,
+      "learning_rate": 1e-06,
+      "loss": 0.2055,
+      "step": 3432
+    },
+    {
+      "epoch": 0.6564681135863849,
+      "grad_norm": 3.156630993500187,
+      "learning_rate": 1e-06,
+      "loss": 0.327,
+      "step": 3433
+    },
+    {
+      "epoch": 0.6566593364566402,
+      "grad_norm": 2.921330365506639,
+      "learning_rate": 1e-06,
+      "loss": 0.1545,
+      "step": 3434
+    },
+    {
+      "epoch": 0.6568505593268955,
+      "grad_norm": 6.404907085807883,
+      "learning_rate": 1e-06,
+      "loss": 0.1701,
+      "step": 3435
+    },
+    {
+      "epoch": 0.6570417821971508,
+      "grad_norm": 4.788194387889429,
+      "learning_rate": 1e-06,
+      "loss": 0.1522,
+      "step": 3436
+    },
+    {
+      "epoch": 0.657233005067406,
+      "grad_norm": 3.511350075588095,
+      "learning_rate": 1e-06,
+      "loss": 0.0527,
+      "step": 3437
+    },
+    {
+      "epoch": 0.6574242279376613,
+      "grad_norm": 2.9525195919100677,
+      "learning_rate": 1e-06,
+      "loss": 0.2773,
+      "step": 3438
+    },
+    {
+      "epoch": 0.6576154508079166,
+      "grad_norm": 3.6885190945638215,
+      "learning_rate": 1e-06,
+      "loss": 0.3408,
+      "step": 3439
+    },
+    {
+      "epoch": 0.6578066736781719,
+      "grad_norm": 4.049841070000246,
+      "learning_rate": 1e-06,
+      "loss": 0.34,
+      "step": 3440
+    },
+    {
+      "epoch": 0.6579978965484272,
+      "grad_norm": 3.575929192851161,
+      "learning_rate": 1e-06,
+      "loss": 0.4275,
+      "step": 3441
+    },
+    {
+      "epoch": 0.6581891194186825,
+      "grad_norm": 2.3403120129574955,
+      "learning_rate": 1e-06,
+      "loss": 0.0688,
+      "step": 3442
+    },
+    {
+      "epoch": 0.6583803422889377,
+      "grad_norm": 4.485229410791787,
+      "learning_rate": 1e-06,
+      "loss": 0.3005,
+      "step": 3443
+    },
+    {
+      "epoch": 0.658571565159193,
+      "grad_norm": 3.871087098332454,
+      "learning_rate": 1e-06,
+      "loss": 0.1917,
+      "step": 3444
+    },
+    {
+      "epoch": 0.6587627880294483,
+      "grad_norm": 1.5247154157912917,
+      "learning_rate": 1e-06,
+      "loss": 0.1218,
+      "step": 3445
+    },
+    {
+      "epoch": 0.6589540108997036,
+      "grad_norm": 2.9449519333859446,
+      "learning_rate": 1e-06,
+      "loss": 0.1864,
+      "step": 3446
+    },
+    {
+      "epoch": 0.6591452337699589,
+      "grad_norm": 2.5351628331361944,
+      "learning_rate": 1e-06,
+      "loss": 0.0684,
+      "step": 3447
+    },
+    {
+      "epoch": 0.6593364566402141,
+      "grad_norm": 2.952342742168722,
+      "learning_rate": 1e-06,
+      "loss": 0.1682,
+      "step": 3448
+    },
+    {
+      "epoch": 0.6595276795104694,
+      "grad_norm": 4.25243375319562,
+      "learning_rate": 1e-06,
+      "loss": 0.1286,
+      "step": 3449
+    },
+    {
+      "epoch": 0.6597189023807247,
+      "grad_norm": 5.11056140809244,
+      "learning_rate": 1e-06,
+      "loss": 0.1175,
+      "step": 3450
+    },
+    {
+      "epoch": 0.65991012525098,
+      "grad_norm": 6.699889782810359,
+      "learning_rate": 1e-06,
+      "loss": 0.2859,
+      "step": 3451
+    },
+    {
+      "epoch": 0.6601013481212353,
+      "grad_norm": 7.506682915317076,
+      "learning_rate": 1e-06,
+      "loss": 0.3428,
+      "step": 3452
+    },
+    {
+      "epoch": 0.6602925709914906,
+      "grad_norm": 2.916070895428142,
+      "learning_rate": 1e-06,
+      "loss": 0.2816,
+      "step": 3453
+    },
+    {
+      "epoch": 0.6604837938617458,
+      "grad_norm": 3.677262010997037,
+      "learning_rate": 1e-06,
+      "loss": 0.326,
+      "step": 3454
+    },
+    {
+      "epoch": 0.6606750167320011,
+      "grad_norm": 1.7871845500539496,
+      "learning_rate": 1e-06,
+      "loss": 0.1391,
+      "step": 3455
+    },
+    {
+      "epoch": 0.6608662396022564,
+      "grad_norm": 4.13813286404193,
+      "learning_rate": 1e-06,
+      "loss": 0.5514,
+      "step": 3456
+    },
+    {
+      "epoch": 0.6610574624725117,
+      "grad_norm": 2.647913428023724,
+      "learning_rate": 1e-06,
+      "loss": 0.1045,
+      "step": 3457
+    },
+    {
+      "epoch": 0.661248685342767,
+      "grad_norm": 4.031522046749557,
+      "learning_rate": 1e-06,
+      "loss": 0.1587,
+      "step": 3458
+    },
+    {
+      "epoch": 0.6614399082130222,
+      "grad_norm": 2.627557144777686,
+      "learning_rate": 1e-06,
+      "loss": 0.1491,
+      "step": 3459
+    },
+    {
+      "epoch": 0.6616311310832775,
+      "grad_norm": 3.7240529437373455,
+      "learning_rate": 1e-06,
+      "loss": 0.1289,
+      "step": 3460
+    },
+    {
+      "epoch": 0.6618223539535328,
+      "grad_norm": 2.7635846558951425,
+      "learning_rate": 1e-06,
+      "loss": 0.0618,
+      "step": 3461
+    },
+    {
+      "epoch": 0.6620135768237881,
+      "grad_norm": 4.696383996655615,
+      "learning_rate": 1e-06,
+      "loss": 0.0833,
+      "step": 3462
+    },
+    {
+      "epoch": 0.6622047996940434,
+      "grad_norm": 5.671117979030244,
+      "learning_rate": 1e-06,
+      "loss": 0.3005,
+      "step": 3463
+    },
+    {
+      "epoch": 0.6623960225642986,
+      "grad_norm": 3.9082060532222913,
+      "learning_rate": 1e-06,
+      "loss": 0.4302,
+      "step": 3464
+    },
+    {
+      "epoch": 0.6625872454345539,
+      "grad_norm": 4.738463597197185,
+      "learning_rate": 1e-06,
+      "loss": 0.6757,
+      "step": 3465
+    },
+    {
+      "epoch": 0.6627784683048092,
+      "grad_norm": 1.9653163819252724,
+      "learning_rate": 1e-06,
+      "loss": 0.2696,
+      "step": 3466
+    },
+    {
+      "epoch": 0.6629696911750645,
+      "grad_norm": 1.8985674150386156,
+      "learning_rate": 1e-06,
+      "loss": 0.0922,
+      "step": 3467
+    },
+    {
+      "epoch": 0.6631609140453198,
+      "grad_norm": 3.4384954225022386,
+      "learning_rate": 1e-06,
+      "loss": 0.1198,
+      "step": 3468
+    },
+    {
+      "epoch": 0.6633521369155752,
+      "grad_norm": 4.479997259207296,
+      "learning_rate": 1e-06,
+      "loss": 0.2711,
+      "step": 3469
+    },
+    {
+      "epoch": 0.6635433597858303,
+      "grad_norm": 1.8742809188346126,
+      "learning_rate": 1e-06,
+      "loss": 0.0597,
+      "step": 3470
+    },
+    {
+      "epoch": 0.6637345826560856,
+      "grad_norm": 1.4595846348236812,
+      "learning_rate": 1e-06,
+      "loss": 0.2144,
+      "step": 3471
+    },
+    {
+      "epoch": 0.663925805526341,
+      "grad_norm": 3.0919531025457494,
+      "learning_rate": 1e-06,
+      "loss": 0.1235,
+      "step": 3472
+    },
+    {
+      "epoch": 0.6641170283965963,
+      "grad_norm": 6.475756991995976,
+      "learning_rate": 1e-06,
+      "loss": 0.1389,
+      "step": 3473
+    },
+    {
+      "epoch": 0.6643082512668516,
+      "grad_norm": 2.415493735316149,
+      "learning_rate": 1e-06,
+      "loss": 0.0578,
+      "step": 3474
+    },
+    {
+      "epoch": 0.6644994741371067,
+      "grad_norm": 11.318233828304994,
+      "learning_rate": 1e-06,
+      "loss": 0.3916,
+      "step": 3475
+    },
+    {
+      "epoch": 0.664690697007362,
+      "grad_norm": 5.114375009845583,
+      "learning_rate": 1e-06,
+      "loss": 0.4083,
+      "step": 3476
+    },
+    {
+      "epoch": 0.6648819198776174,
+      "grad_norm": 7.763162326045817,
+      "learning_rate": 1e-06,
+      "loss": 0.3499,
+      "step": 3477
+    },
+    {
+      "epoch": 0.6650731427478727,
+      "grad_norm": 3.903802455875803,
+      "learning_rate": 1e-06,
+      "loss": 0.2312,
+      "step": 3478
+    },
+    {
+      "epoch": 0.665264365618128,
+      "grad_norm": 5.131911293006447,
+      "learning_rate": 1e-06,
+      "loss": 0.1771,
+      "step": 3479
+    },
+    {
+      "epoch": 0.6654555884883833,
+      "grad_norm": 2.7505179697686515,
+      "learning_rate": 1e-06,
+      "loss": 0.2893,
+      "step": 3480
+    },
+    {
+      "epoch": 0.6656468113586385,
+      "grad_norm": 3.248214377868598,
+      "learning_rate": 1e-06,
+      "loss": 0.3961,
+      "step": 3481
+    },
+    {
+      "epoch": 0.6658380342288938,
+      "grad_norm": 3.303903040204459,
+      "learning_rate": 1e-06,
+      "loss": 0.1901,
+      "step": 3482
+    },
+    {
+      "epoch": 0.6660292570991491,
+      "grad_norm": 3.657821839070935,
+      "learning_rate": 1e-06,
+      "loss": 0.2165,
+      "step": 3483
+    },
+    {
+      "epoch": 0.6662204799694044,
+      "grad_norm": 3.2599376452514752,
+      "learning_rate": 1e-06,
+      "loss": 0.2144,
+      "step": 3484
+    },
+    {
+      "epoch": 0.6664117028396597,
+      "grad_norm": 2.5767264532018386,
+      "learning_rate": 1e-06,
+      "loss": 0.0604,
+      "step": 3485
+    },
+    {
+      "epoch": 0.6666029257099149,
+      "grad_norm": 5.223191615464195,
+      "learning_rate": 1e-06,
+      "loss": 0.2082,
+      "step": 3486
+    },
+    {
+      "epoch": 0.6667941485801702,
+      "grad_norm": 3.5259692110912533,
+      "learning_rate": 1e-06,
+      "loss": 0.088,
+      "step": 3487
+    },
+    {
+      "epoch": 0.6669853714504255,
+      "grad_norm": 6.555169361757797,
+      "learning_rate": 1e-06,
+      "loss": 0.1267,
+      "step": 3488
+    },
+    {
+      "epoch": 0.6671765943206808,
+      "grad_norm": 4.666146839799023,
+      "learning_rate": 1e-06,
+      "loss": 0.2038,
+      "step": 3489
+    },
+    {
+      "epoch": 0.6673678171909361,
+      "grad_norm": 5.147334859161948,
+      "learning_rate": 1e-06,
+      "loss": 0.3294,
+      "step": 3490
+    },
+    {
+      "epoch": 0.6675590400611913,
+      "grad_norm": 3.9523792392685415,
+      "learning_rate": 1e-06,
+      "loss": 0.4719,
+      "step": 3491
+    },
+    {
+      "epoch": 0.6677502629314466,
+      "grad_norm": 2.8351316541885723,
+      "learning_rate": 1e-06,
+      "loss": 0.4441,
+      "step": 3492
+    },
+    {
+      "epoch": 0.6679414858017019,
+      "grad_norm": 2.840614350721753,
+      "learning_rate": 1e-06,
+      "loss": 0.2686,
+      "step": 3493
+    },
+    {
+      "epoch": 0.6681327086719572,
+      "grad_norm": 3.362237323016354,
+      "learning_rate": 1e-06,
+      "loss": 0.2948,
+      "step": 3494
+    },
+    {
+      "epoch": 0.6683239315422125,
+      "grad_norm": 2.339703538161272,
+      "learning_rate": 1e-06,
+      "loss": 0.0684,
+      "step": 3495
+    },
+    {
+      "epoch": 0.6685151544124678,
+      "grad_norm": 3.573042278843592,
+      "learning_rate": 1e-06,
+      "loss": 0.2189,
+      "step": 3496
+    },
+    {
+      "epoch": 0.668706377282723,
+      "grad_norm": 3.160455149695193,
+      "learning_rate": 1e-06,
+      "loss": 0.243,
+      "step": 3497
+    },
+    {
+      "epoch": 0.6688976001529783,
+      "grad_norm": 3.4775349489234304,
+      "learning_rate": 1e-06,
+      "loss": 0.0776,
+      "step": 3498
+    },
+    {
+      "epoch": 0.6690888230232336,
+      "grad_norm": 2.767866185608608,
+      "learning_rate": 1e-06,
+      "loss": 0.0455,
+      "step": 3499
+    },
+    {
+      "epoch": 0.6692800458934889,
+      "grad_norm": 4.943241885148147,
+      "learning_rate": 1e-06,
+      "loss": 0.0968,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6692800458934889,
+      "eval_runtime": 742.1299,
+      "eval_samples_per_second": 2.067,
+      "eval_steps_per_second": 0.517,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6694712687637442,
+      "grad_norm": 5.608488645098798,
+      "learning_rate": 1e-06,
+      "loss": 0.1927,
+      "step": 3501
+    },
+    {
+      "epoch": 0.6696624916339994,
+      "grad_norm": 3.134360827397778,
+      "learning_rate": 1e-06,
+      "loss": 0.2417,
+      "step": 3502
+    },
+    {
+      "epoch": 0.6698537145042547,
+      "grad_norm": 3.0998248942972855,
+      "learning_rate": 1e-06,
+      "loss": 0.3531,
+      "step": 3503
+    },
+    {
+      "epoch": 0.67004493737451,
+      "grad_norm": 4.5449795786920255,
+      "learning_rate": 1e-06,
+      "loss": 0.3574,
+      "step": 3504
+    },
+    {
+      "epoch": 0.6702361602447653,
+      "grad_norm": 4.2202374132094205,
+      "learning_rate": 1e-06,
+      "loss": 0.3426,
+      "step": 3505
+    },
+    {
+      "epoch": 0.6704273831150206,
+      "grad_norm": 3.874518826050901,
+      "learning_rate": 1e-06,
+      "loss": 0.2763,
+      "step": 3506
+    },
+    {
+      "epoch": 0.6706186059852758,
+      "grad_norm": 4.716362355438548,
+      "learning_rate": 1e-06,
+      "loss": 0.2044,
+      "step": 3507
+    },
+    {
+      "epoch": 0.6708098288555311,
+      "grad_norm": 3.5888205513551528,
+      "learning_rate": 1e-06,
+      "loss": 0.107,
+      "step": 3508
+    },
+    {
+      "epoch": 0.6710010517257864,
+      "grad_norm": 3.5346574125779138,
+      "learning_rate": 1e-06,
+      "loss": 0.11,
+      "step": 3509
+    },
+    {
+      "epoch": 0.6711922745960417,
+      "grad_norm": 3.8934584261990683,
+      "learning_rate": 1e-06,
+      "loss": 0.0938,
+      "step": 3510
+    },
+    {
+      "epoch": 0.671383497466297,
+      "grad_norm": 18.042366184596258,
+      "learning_rate": 1e-06,
+      "loss": 0.2986,
+      "step": 3511
+    },
+    {
+      "epoch": 0.6715747203365523,
+      "grad_norm": 3.2894620312936076,
+      "learning_rate": 1e-06,
+      "loss": 0.0889,
+      "step": 3512
+    },
+    {
+      "epoch": 0.6717659432068075,
+      "grad_norm": 4.36519264696219,
+      "learning_rate": 1e-06,
+      "loss": 0.0813,
+      "step": 3513
+    },
+    {
+      "epoch": 0.6719571660770628,
+      "grad_norm": 6.910363538704787,
+      "learning_rate": 1e-06,
+      "loss": 0.4009,
+      "step": 3514
+    },
+    {
+      "epoch": 0.6721483889473181,
+      "grad_norm": 5.2311311149988935,
+      "learning_rate": 1e-06,
+      "loss": 0.6676,
+      "step": 3515
+    },
+    {
+      "epoch": 0.6723396118175734,
+      "grad_norm": 4.185825753366811,
+      "learning_rate": 1e-06,
+      "loss": 0.1744,
+      "step": 3516
+    },
+    {
+      "epoch": 0.6725308346878287,
+      "grad_norm": 2.6671745492165995,
+      "learning_rate": 1e-06,
+      "loss": 0.2366,
+      "step": 3517
+    },
+    {
+      "epoch": 0.6727220575580839,
+      "grad_norm": 2.4234560820053437,
+      "learning_rate": 1e-06,
+      "loss": 0.1312,
+      "step": 3518
+    },
+    {
+      "epoch": 0.6729132804283392,
+      "grad_norm": 4.711888054494285,
+      "learning_rate": 1e-06,
+      "loss": 0.6447,
+      "step": 3519
+    },
+    {
+      "epoch": 0.6731045032985945,
+      "grad_norm": 1.8580231440471071,
+      "learning_rate": 1e-06,
+      "loss": 0.0857,
+      "step": 3520
+    },
+    {
+      "epoch": 0.6732957261688498,
+      "grad_norm": 2.6057022653508013,
+      "learning_rate": 1e-06,
+      "loss": 0.1556,
+      "step": 3521
+    },
+    {
+      "epoch": 0.6734869490391051,
+      "grad_norm": 3.719193632447061,
+      "learning_rate": 1e-06,
+      "loss": 0.1863,
+      "step": 3522
+    },
+    {
+      "epoch": 0.6736781719093604,
+      "grad_norm": 1.9660233886856506,
+      "learning_rate": 1e-06,
+      "loss": 0.053,
+      "step": 3523
+    },
+    {
+      "epoch": 0.6738693947796156,
+      "grad_norm": 2.9128268078993957,
+      "learning_rate": 1e-06,
+      "loss": 0.1586,
+      "step": 3524
+    },
+    {
+      "epoch": 0.6740606176498709,
+      "grad_norm": 4.188649034139907,
+      "learning_rate": 1e-06,
+      "loss": 0.1393,
+      "step": 3525
+    },
+    {
+      "epoch": 0.6742518405201262,
+      "grad_norm": 3.1728371349510387,
+      "learning_rate": 1e-06,
+      "loss": 0.3855,
+      "step": 3526
+    },
+    {
+      "epoch": 0.6744430633903815,
+      "grad_norm": 3.3810025277950535,
+      "learning_rate": 1e-06,
+      "loss": 0.1827,
+      "step": 3527
+    },
+    {
+      "epoch": 0.6746342862606368,
+      "grad_norm": 4.491757154998202,
+      "learning_rate": 1e-06,
+      "loss": 0.4295,
+      "step": 3528
+    },
+    {
+      "epoch": 0.674825509130892,
+      "grad_norm": 4.696572843922628,
+      "learning_rate": 1e-06,
+      "loss": 0.4616,
+      "step": 3529
+    },
+    {
+      "epoch": 0.6750167320011473,
+      "grad_norm": 3.1074818737824086,
+      "learning_rate": 1e-06,
+      "loss": 0.1184,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6752079548714026,
+      "grad_norm": 3.625380660492313,
+      "learning_rate": 1e-06,
+      "loss": 0.1275,
+      "step": 3531
+    },
+    {
+      "epoch": 0.6753991777416579,
+      "grad_norm": 2.6907518806624617,
+      "learning_rate": 1e-06,
+      "loss": 0.1091,
+      "step": 3532
+    },
+    {
+      "epoch": 0.6755904006119132,
+      "grad_norm": 4.304262116005629,
+      "learning_rate": 1e-06,
+      "loss": 0.1116,
+      "step": 3533
+    },
+    {
+      "epoch": 0.6757816234821684,
+      "grad_norm": 2.262592048247857,
+      "learning_rate": 1e-06,
+      "loss": 0.085,
+      "step": 3534
+    },
+    {
+      "epoch": 0.6759728463524237,
+      "grad_norm": 4.759710324873667,
+      "learning_rate": 1e-06,
+      "loss": 0.1284,
+      "step": 3535
+    },
+    {
+      "epoch": 0.676164069222679,
+      "grad_norm": 5.164574763619331,
+      "learning_rate": 1e-06,
+      "loss": 0.1518,
+      "step": 3536
+    },
+    {
+      "epoch": 0.6763552920929343,
+      "grad_norm": 8.794919923176,
+      "learning_rate": 1e-06,
+      "loss": 0.1832,
+      "step": 3537
+    },
+    {
+      "epoch": 0.6765465149631896,
+      "grad_norm": 5.634032732342044,
+      "learning_rate": 1e-06,
+      "loss": 0.2417,
+      "step": 3538
+    },
+    {
+      "epoch": 0.6767377378334449,
+      "grad_norm": 5.474372806942776,
+      "learning_rate": 1e-06,
+      "loss": 0.4297,
+      "step": 3539
+    },
+    {
+      "epoch": 0.6769289607037001,
+      "grad_norm": 5.485546541203792,
+      "learning_rate": 1e-06,
+      "loss": 0.7151,
+      "step": 3540
+    },
+    {
+      "epoch": 0.6771201835739554,
+      "grad_norm": 6.7183011636962995,
+      "learning_rate": 1e-06,
+      "loss": 0.325,
+      "step": 3541
+    },
+    {
+      "epoch": 0.6773114064442107,
+      "grad_norm": 2.7417668294340536,
+      "learning_rate": 1e-06,
+      "loss": 0.1494,
+      "step": 3542
+    },
+    {
+      "epoch": 0.677502629314466,
+      "grad_norm": 2.5830785153957705,
+      "learning_rate": 1e-06,
+      "loss": 0.2081,
+      "step": 3543
+    },
+    {
+      "epoch": 0.6776938521847213,
+      "grad_norm": 3.665718273852593,
+      "learning_rate": 1e-06,
+      "loss": 0.34,
+      "step": 3544
+    },
+    {
+      "epoch": 0.6778850750549765,
+      "grad_norm": 2.1502148853526566,
+      "learning_rate": 1e-06,
+      "loss": 0.181,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6780762979252318,
+      "grad_norm": 2.4418066077983522,
+      "learning_rate": 1e-06,
+      "loss": 0.0817,
+      "step": 3546
+    },
+    {
+      "epoch": 0.6782675207954871,
+      "grad_norm": 2.6711087382928964,
+      "learning_rate": 1e-06,
+      "loss": 0.1311,
+      "step": 3547
+    },
+    {
+      "epoch": 0.6784587436657424,
+      "grad_norm": 2.875547439865748,
+      "learning_rate": 1e-06,
+      "loss": 0.057,
+      "step": 3548
+    },
+    {
+      "epoch": 0.6786499665359977,
+      "grad_norm": 5.580268683032126,
+      "learning_rate": 1e-06,
+      "loss": 0.195,
+      "step": 3549
+    },
+    {
+      "epoch": 0.678841189406253,
+      "grad_norm": 4.640550400876095,
+      "learning_rate": 1e-06,
+      "loss": 0.0975,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6790324122765082,
+      "grad_norm": 5.720592097868481,
+      "learning_rate": 1e-06,
+      "loss": 0.307,
+      "step": 3551
+    },
+    {
+      "epoch": 0.6792236351467635,
+      "grad_norm": 4.109617741018612,
+      "learning_rate": 1e-06,
+      "loss": 0.5548,
+      "step": 3552
+    },
+    {
+      "epoch": 0.6794148580170188,
+      "grad_norm": 4.7316733731478395,
+      "learning_rate": 1e-06,
+      "loss": 0.5056,
+      "step": 3553
+    },
+    {
+      "epoch": 0.6796060808872741,
+      "grad_norm": 2.190567808180132,
+      "learning_rate": 1e-06,
+      "loss": 0.1443,
+      "step": 3554
+    },
+    {
+      "epoch": 0.6797973037575294,
+      "grad_norm": 6.122207627891566,
+      "learning_rate": 1e-06,
+      "loss": 0.4866,
+      "step": 3555
+    },
+    {
+      "epoch": 0.6799885266277846,
+      "grad_norm": 2.5452296084868675,
+      "learning_rate": 1e-06,
+      "loss": 0.1926,
+      "step": 3556
+    },
+    {
+      "epoch": 0.6801797494980399,
+      "grad_norm": 3.5230620664207066,
+      "learning_rate": 1e-06,
+      "loss": 0.347,
+      "step": 3557
+    },
+    {
+      "epoch": 0.6803709723682952,
+      "grad_norm": 2.4218009691306244,
+      "learning_rate": 1e-06,
+      "loss": 0.0922,
+      "step": 3558
+    },
+    {
+      "epoch": 0.6805621952385505,
+      "grad_norm": 3.100028400906318,
+      "learning_rate": 1e-06,
+      "loss": 0.0896,
+      "step": 3559
+    },
+    {
+      "epoch": 0.6807534181088059,
+      "grad_norm": 4.468867800567087,
+      "learning_rate": 1e-06,
+      "loss": 0.2192,
+      "step": 3560
+    },
+    {
+      "epoch": 0.680944640979061,
+      "grad_norm": 3.1384394238557345,
+      "learning_rate": 1e-06,
+      "loss": 0.1589,
+      "step": 3561
+    },
+    {
+      "epoch": 0.6811358638493163,
+      "grad_norm": 1.9907636752848743,
+      "learning_rate": 1e-06,
+      "loss": 0.0381,
+      "step": 3562
+    },
+    {
+      "epoch": 0.6813270867195717,
+      "grad_norm": 5.3251191103513085,
+      "learning_rate": 1e-06,
+      "loss": 0.2147,
+      "step": 3563
+    },
+    {
+      "epoch": 0.681518309589827,
+      "grad_norm": 3.023608614394322,
+      "learning_rate": 1e-06,
+      "loss": 0.3378,
+      "step": 3564
+    },
+    {
+      "epoch": 0.6817095324600823,
+      "grad_norm": 3.9610139436888088,
+      "learning_rate": 1e-06,
+      "loss": 0.5397,
+      "step": 3565
+    },
+    {
+      "epoch": 0.6819007553303376,
+      "grad_norm": 3.9168446412468994,
+      "learning_rate": 1e-06,
+      "loss": 0.6644,
+      "step": 3566
+    },
+    {
+      "epoch": 0.6820919782005928,
+      "grad_norm": 2.9998029008332625,
+      "learning_rate": 1e-06,
+      "loss": 0.2162,
+      "step": 3567
+    },
+    {
+      "epoch": 0.682283201070848,
+      "grad_norm": 5.219395945884638,
+      "learning_rate": 1e-06,
+      "loss": 0.649,
+      "step": 3568
+    },
+    {
+      "epoch": 0.6824744239411034,
+      "grad_norm": 1.9426511768049923,
+      "learning_rate": 1e-06,
+      "loss": 0.195,
+      "step": 3569
+    },
+    {
+      "epoch": 0.6826656468113587,
+      "grad_norm": 2.753645561270679,
+      "learning_rate": 1e-06,
+      "loss": 0.2148,
+      "step": 3570
+    },
+    {
+      "epoch": 0.682856869681614,
+      "grad_norm": 5.014943109191543,
+      "learning_rate": 1e-06,
+      "loss": 0.3169,
+      "step": 3571
+    },
+    {
+      "epoch": 0.6830480925518692,
+      "grad_norm": 4.735976750819387,
+      "learning_rate": 1e-06,
+      "loss": 0.0856,
+      "step": 3572
+    },
+    {
+      "epoch": 0.6832393154221245,
+      "grad_norm": 3.3561534960066,
+      "learning_rate": 1e-06,
+      "loss": 0.1041,
+      "step": 3573
+    },
+    {
+      "epoch": 0.6834305382923798,
+      "grad_norm": 3.963952716995767,
+      "learning_rate": 1e-06,
+      "loss": 0.0619,
+      "step": 3574
+    },
+    {
+      "epoch": 0.6836217611626351,
+      "grad_norm": 3.5952012821984645,
+      "learning_rate": 1e-06,
+      "loss": 0.1261,
+      "step": 3575
+    },
+    {
+      "epoch": 0.6838129840328904,
+      "grad_norm": 6.021607592296717,
+      "learning_rate": 1e-06,
+      "loss": 0.3223,
+      "step": 3576
+    },
+    {
+      "epoch": 0.6840042069031456,
+      "grad_norm": 4.5394837385891185,
+      "learning_rate": 1e-06,
+      "loss": 0.3696,
+      "step": 3577
+    },
+    {
+      "epoch": 0.6841954297734009,
+      "grad_norm": 4.45846355402545,
+      "learning_rate": 1e-06,
+      "loss": 0.5343,
+      "step": 3578
+    },
+    {
+      "epoch": 0.6843866526436562,
+      "grad_norm": 2.9137283419962383,
+      "learning_rate": 1e-06,
+      "loss": 0.194,
+      "step": 3579
+    },
+    {
+      "epoch": 0.6845778755139115,
+      "grad_norm": 2.4644886371631536,
+      "learning_rate": 1e-06,
+      "loss": 0.2729,
+      "step": 3580
+    },
+    {
+      "epoch": 0.6847690983841668,
+      "grad_norm": 4.389625082494878,
+      "learning_rate": 1e-06,
+      "loss": 0.4207,
+      "step": 3581
+    },
+    {
+      "epoch": 0.6849603212544221,
+      "grad_norm": 3.4329187556453076,
+      "learning_rate": 1e-06,
+      "loss": 0.1338,
+      "step": 3582
+    },
+    {
+      "epoch": 0.6851515441246773,
+      "grad_norm": 2.5901911575883974,
+      "learning_rate": 1e-06,
+      "loss": 0.0648,
+      "step": 3583
+    },
+    {
+      "epoch": 0.6853427669949326,
+      "grad_norm": 3.863939435567146,
+      "learning_rate": 1e-06,
+      "loss": 0.2222,
+      "step": 3584
+    },
+    {
+      "epoch": 0.6855339898651879,
+      "grad_norm": 2.790169290581663,
+      "learning_rate": 1e-06,
+      "loss": 0.1247,
+      "step": 3585
+    },
+    {
+      "epoch": 0.6857252127354432,
+      "grad_norm": 4.073607767436615,
+      "learning_rate": 1e-06,
+      "loss": 0.1314,
+      "step": 3586
+    },
+    {
+      "epoch": 0.6859164356056985,
+      "grad_norm": 2.6605701096589516,
+      "learning_rate": 1e-06,
+      "loss": 0.0563,
+      "step": 3587
+    },
+    {
+      "epoch": 0.6861076584759537,
+      "grad_norm": 5.021372891423388,
+      "learning_rate": 1e-06,
+      "loss": 0.2255,
+      "step": 3588
+    },
+    {
+      "epoch": 0.686298881346209,
+      "grad_norm": 5.625387390360417,
+      "learning_rate": 1e-06,
+      "loss": 0.6646,
+      "step": 3589
+    },
+    {
+      "epoch": 0.6864901042164643,
+      "grad_norm": 5.371622310640329,
+      "learning_rate": 1e-06,
+      "loss": 0.3696,
+      "step": 3590
+    },
+    {
+      "epoch": 0.6866813270867196,
+      "grad_norm": 3.5425949824810488,
+      "learning_rate": 1e-06,
+      "loss": 0.2778,
+      "step": 3591
+    },
+    {
+      "epoch": 0.6868725499569749,
+      "grad_norm": 3.5204675849383364,
+      "learning_rate": 1e-06,
+      "loss": 0.292,
+      "step": 3592
+    },
+    {
+      "epoch": 0.6870637728272302,
+      "grad_norm": 4.071257790829469,
+      "learning_rate": 1e-06,
+      "loss": 0.1751,
+      "step": 3593
+    },
+    {
+      "epoch": 0.6872549956974854,
+      "grad_norm": 3.027903646561245,
+      "learning_rate": 1e-06,
+      "loss": 0.3315,
+      "step": 3594
+    },
+    {
+      "epoch": 0.6874462185677407,
+      "grad_norm": 3.892565875074804,
+      "learning_rate": 1e-06,
+      "loss": 0.2079,
+      "step": 3595
+    },
+    {
+      "epoch": 0.687637441437996,
+      "grad_norm": 3.1860526295617597,
+      "learning_rate": 1e-06,
+      "loss": 0.2472,
+      "step": 3596
+    },
+    {
+      "epoch": 0.6878286643082513,
+      "grad_norm": 3.5192929270812345,
+      "learning_rate": 1e-06,
+      "loss": 0.1242,
+      "step": 3597
+    },
+    {
+      "epoch": 0.6880198871785066,
+      "grad_norm": 4.079482280196652,
+      "learning_rate": 1e-06,
+      "loss": 0.1014,
+      "step": 3598
+    },
+    {
+      "epoch": 0.6882111100487618,
+      "grad_norm": 2.5316390692198874,
+      "learning_rate": 1e-06,
+      "loss": 0.0794,
+      "step": 3599
+    },
+    {
+      "epoch": 0.6884023329190171,
+      "grad_norm": 2.907397792131633,
+      "learning_rate": 1e-06,
+      "loss": 0.1013,
+      "step": 3600
+    },
+    {
+      "epoch": 0.6885935557892724,
+      "grad_norm": 4.767240056761256,
+      "learning_rate": 1e-06,
+      "loss": 0.3063,
+      "step": 3601
+    },
+    {
+      "epoch": 0.6887847786595277,
+      "grad_norm": 4.2270591006924745,
+      "learning_rate": 1e-06,
+      "loss": 0.5272,
+      "step": 3602
+    },
+    {
+      "epoch": 0.688976001529783,
+      "grad_norm": 9.589366395960505,
+      "learning_rate": 1e-06,
+      "loss": 0.3713,
+      "step": 3603
+    },
+    {
+      "epoch": 0.6891672244000382,
+      "grad_norm": 4.721138255054584,
+      "learning_rate": 1e-06,
+      "loss": 0.1815,
+      "step": 3604
+    },
+    {
+      "epoch": 0.6893584472702935,
+      "grad_norm": 3.082505415689703,
+      "learning_rate": 1e-06,
+      "loss": 0.3611,
+      "step": 3605
+    },
+    {
+      "epoch": 0.6895496701405488,
+      "grad_norm": 4.465156444102771,
+      "learning_rate": 1e-06,
+      "loss": 0.4753,
+      "step": 3606
+    },
+    {
+      "epoch": 0.6897408930108041,
+      "grad_norm": 3.2408859398017524,
+      "learning_rate": 1e-06,
+      "loss": 0.0995,
+      "step": 3607
+    },
+    {
+      "epoch": 0.6899321158810594,
+      "grad_norm": 2.9593411372179896,
+      "learning_rate": 1e-06,
+      "loss": 0.1035,
+      "step": 3608
+    },
+    {
+      "epoch": 0.6901233387513147,
+      "grad_norm": 3.034660858612447,
+      "learning_rate": 1e-06,
+      "loss": 0.1468,
+      "step": 3609
+    },
+    {
+      "epoch": 0.6903145616215699,
+      "grad_norm": 3.9790164592946082,
+      "learning_rate": 1e-06,
+      "loss": 0.111,
+      "step": 3610
+    },
+    {
+      "epoch": 0.6905057844918252,
+      "grad_norm": 4.681126444959714,
+      "learning_rate": 1e-06,
+      "loss": 0.2284,
+      "step": 3611
+    },
+    {
+      "epoch": 0.6906970073620805,
+      "grad_norm": 7.544042336046797,
+      "learning_rate": 1e-06,
+      "loss": 0.1574,
+      "step": 3612
+    },
+    {
+      "epoch": 0.6908882302323358,
+      "grad_norm": 6.198510747908998,
+      "learning_rate": 1e-06,
+      "loss": 0.1913,
+      "step": 3613
+    },
+    {
+      "epoch": 0.6910794531025911,
+      "grad_norm": 4.335107782381343,
+      "learning_rate": 1e-06,
+      "loss": 0.2072,
+      "step": 3614
+    },
+    {
+      "epoch": 0.6912706759728463,
+      "grad_norm": 2.66498780052728,
+      "learning_rate": 1e-06,
+      "loss": 0.1666,
+      "step": 3615
+    },
+    {
+      "epoch": 0.6914618988431016,
+      "grad_norm": 2.2194100056462847,
+      "learning_rate": 1e-06,
+      "loss": 0.2665,
+      "step": 3616
+    },
+    {
+      "epoch": 0.6916531217133569,
+      "grad_norm": 2.710349423167888,
+      "learning_rate": 1e-06,
+      "loss": 0.2897,
+      "step": 3617
+    },
+    {
+      "epoch": 0.6918443445836122,
+      "grad_norm": 5.274349060392596,
+      "learning_rate": 1e-06,
+      "loss": 0.6655,
+      "step": 3618
+    },
+    {
+      "epoch": 0.6920355674538675,
+      "grad_norm": 1.7986695989118748,
+      "learning_rate": 1e-06,
+      "loss": 0.1618,
+      "step": 3619
+    },
+    {
+      "epoch": 0.6922267903241228,
+      "grad_norm": 3.7030768411458417,
+      "learning_rate": 1e-06,
+      "loss": 0.1698,
+      "step": 3620
+    },
+    {
+      "epoch": 0.692418013194378,
+      "grad_norm": 2.6514770709849604,
+      "learning_rate": 1e-06,
+      "loss": 0.077,
+      "step": 3621
+    },
+    {
+      "epoch": 0.6926092360646333,
+      "grad_norm": 3.579407574528881,
+      "learning_rate": 1e-06,
+      "loss": 0.0546,
+      "step": 3622
+    },
+    {
+      "epoch": 0.6928004589348886,
+      "grad_norm": 1.8800552249467268,
+      "learning_rate": 1e-06,
+      "loss": 0.0359,
+      "step": 3623
+    },
+    {
+      "epoch": 0.6929916818051439,
+      "grad_norm": 2.468760623184719,
+      "learning_rate": 1e-06,
+      "loss": 0.053,
+      "step": 3624
+    },
+    {
+      "epoch": 0.6931829046753992,
+      "grad_norm": 4.598089957705838,
+      "learning_rate": 1e-06,
+      "loss": 0.1439,
+      "step": 3625
+    },
+    {
+      "epoch": 0.6933741275456544,
+      "grad_norm": 7.590175713677645,
+      "learning_rate": 1e-06,
+      "loss": 0.5166,
+      "step": 3626
+    },
+    {
+      "epoch": 0.6935653504159097,
+      "grad_norm": 5.695483630295059,
+      "learning_rate": 1e-06,
+      "loss": 0.8735,
+      "step": 3627
+    },
+    {
+      "epoch": 0.693756573286165,
+      "grad_norm": 5.717073788988579,
+      "learning_rate": 1e-06,
+      "loss": 0.4077,
+      "step": 3628
+    },
+    {
+      "epoch": 0.6939477961564203,
+      "grad_norm": 3.5632398405974715,
+      "learning_rate": 1e-06,
+      "loss": 0.1981,
+      "step": 3629
+    },
+    {
+      "epoch": 0.6941390190266756,
+      "grad_norm": 2.770431565911955,
+      "learning_rate": 1e-06,
+      "loss": 0.0995,
+      "step": 3630
+    },
+    {
+      "epoch": 0.6943302418969308,
+      "grad_norm": 3.974200133453854,
+      "learning_rate": 1e-06,
+      "loss": 0.1242,
+      "step": 3631
+    },
+    {
+      "epoch": 0.6945214647671861,
+      "grad_norm": 2.3182563276777377,
+      "learning_rate": 1e-06,
+      "loss": 0.1634,
+      "step": 3632
+    },
+    {
+      "epoch": 0.6947126876374414,
+      "grad_norm": 2.343440734804075,
+      "learning_rate": 1e-06,
+      "loss": 0.086,
+      "step": 3633
+    },
+    {
+      "epoch": 0.6949039105076967,
+      "grad_norm": 3.5183097929884566,
+      "learning_rate": 1e-06,
+      "loss": 0.0948,
+      "step": 3634
+    },
+    {
+      "epoch": 0.695095133377952,
+      "grad_norm": 3.0826149352736816,
+      "learning_rate": 1e-06,
+      "loss": 0.1404,
+      "step": 3635
+    },
+    {
+      "epoch": 0.6952863562482073,
+      "grad_norm": 6.2795180025217014,
+      "learning_rate": 1e-06,
+      "loss": 0.1313,
+      "step": 3636
+    },
+    {
+      "epoch": 0.6954775791184625,
+      "grad_norm": 3.4563251011227054,
+      "learning_rate": 1e-06,
+      "loss": 0.1008,
+      "step": 3637
+    },
+    {
+      "epoch": 0.6956688019887178,
+      "grad_norm": 10.773720054987608,
+      "learning_rate": 1e-06,
+      "loss": 0.2611,
+      "step": 3638
+    },
+    {
+      "epoch": 0.6958600248589731,
+      "grad_norm": 6.024234466670751,
+      "learning_rate": 1e-06,
+      "loss": 0.4018,
+      "step": 3639
+    },
+    {
+      "epoch": 0.6960512477292284,
+      "grad_norm": 5.490005254968793,
+      "learning_rate": 1e-06,
+      "loss": 0.5113,
+      "step": 3640
+    },
+    {
+      "epoch": 0.6962424705994837,
+      "grad_norm": 2.780531458242668,
+      "learning_rate": 1e-06,
+      "loss": 0.3551,
+      "step": 3641
+    },
+    {
+      "epoch": 0.6964336934697389,
+      "grad_norm": 3.1449876536989687,
+      "learning_rate": 1e-06,
+      "loss": 0.3709,
+      "step": 3642
+    },
+    {
+      "epoch": 0.6966249163399942,
+      "grad_norm": 3.2520135364133442,
+      "learning_rate": 1e-06,
+      "loss": 0.2607,
+      "step": 3643
+    },
+    {
+      "epoch": 0.6968161392102495,
+      "grad_norm": 1.8122712352685744,
+      "learning_rate": 1e-06,
+      "loss": 0.2936,
+      "step": 3644
+    },
+    {
+      "epoch": 0.6970073620805048,
+      "grad_norm": 2.9185947221707083,
+      "learning_rate": 1e-06,
+      "loss": 0.0886,
+      "step": 3645
+    },
+    {
+      "epoch": 0.6971985849507601,
+      "grad_norm": 3.852538691183442,
+      "learning_rate": 1e-06,
+      "loss": 0.2023,
+      "step": 3646
+    },
+    {
+      "epoch": 0.6973898078210153,
+      "grad_norm": 5.82300725421719,
+      "learning_rate": 1e-06,
+      "loss": 0.3201,
+      "step": 3647
+    },
+    {
+      "epoch": 0.6975810306912706,
+      "grad_norm": 6.355793762337261,
+      "learning_rate": 1e-06,
+      "loss": 0.2787,
+      "step": 3648
+    },
+    {
+      "epoch": 0.697772253561526,
+      "grad_norm": 2.5941794166370076,
+      "learning_rate": 1e-06,
+      "loss": 0.0626,
+      "step": 3649
+    },
+    {
+      "epoch": 0.6979634764317812,
+      "grad_norm": 5.641679237310397,
+      "learning_rate": 1e-06,
+      "loss": 0.1641,
+      "step": 3650
+    },
+    {
+      "epoch": 0.6981546993020366,
+      "grad_norm": 4.732408574707889,
+      "learning_rate": 1e-06,
+      "loss": 0.1582,
+      "step": 3651
+    },
+    {
+      "epoch": 0.6983459221722919,
+      "grad_norm": 4.1502497678303625,
+      "learning_rate": 1e-06,
+      "loss": 0.2968,
+      "step": 3652
+    },
+    {
+      "epoch": 0.698537145042547,
+      "grad_norm": 7.0943845221447415,
+      "learning_rate": 1e-06,
+      "loss": 0.6905,
+      "step": 3653
+    },
+    {
+      "epoch": 0.6987283679128024,
+      "grad_norm": 3.313784080511429,
+      "learning_rate": 1e-06,
+      "loss": 0.1772,
+      "step": 3654
+    },
+    {
+      "epoch": 0.6989195907830577,
+      "grad_norm": 3.278506049018691,
+      "learning_rate": 1e-06,
+      "loss": 0.2057,
+      "step": 3655
+    },
+    {
+      "epoch": 0.699110813653313,
+      "grad_norm": 5.743851733214768,
+      "learning_rate": 1e-06,
+      "loss": 0.4775,
+      "step": 3656
+    },
+    {
+      "epoch": 0.6993020365235683,
+      "grad_norm": 2.6156867891622273,
+      "learning_rate": 1e-06,
+      "loss": 0.1562,
+      "step": 3657
+    },
+    {
+      "epoch": 0.6994932593938235,
+      "grad_norm": 3.4364739013639927,
+      "learning_rate": 1e-06,
+      "loss": 0.1587,
+      "step": 3658
+    },
+    {
+      "epoch": 0.6996844822640788,
+      "grad_norm": 3.047863765595564,
+      "learning_rate": 1e-06,
+      "loss": 0.1183,
+      "step": 3659
+    },
+    {
+      "epoch": 0.6998757051343341,
+      "grad_norm": 6.041073088152177,
+      "learning_rate": 1e-06,
+      "loss": 0.2073,
+      "step": 3660
+    },
+    {
+      "epoch": 0.7000669280045894,
+      "grad_norm": 4.02594473040677,
+      "learning_rate": 1e-06,
+      "loss": 0.311,
+      "step": 3661
+    },
+    {
+      "epoch": 0.7002581508748447,
+      "grad_norm": 2.9574411327343215,
+      "learning_rate": 1e-06,
+      "loss": 0.0642,
+      "step": 3662
+    },
+    {
+      "epoch": 0.7004493737451,
+      "grad_norm": 5.640833800617292,
+      "learning_rate": 1e-06,
+      "loss": 0.1445,
+      "step": 3663
+    },
+    {
+      "epoch": 0.7006405966153552,
+      "grad_norm": 4.248310426450989,
+      "learning_rate": 1e-06,
+      "loss": 0.5405,
+      "step": 3664
+    },
+    {
+      "epoch": 0.7008318194856105,
+      "grad_norm": 2.439139963761281,
+      "learning_rate": 1e-06,
+      "loss": 0.1992,
+      "step": 3665
+    },
+    {
+      "epoch": 0.7010230423558658,
+      "grad_norm": 8.052303046370394,
+      "learning_rate": 1e-06,
+      "loss": 0.6034,
+      "step": 3666
+    },
+    {
+      "epoch": 0.7012142652261211,
+      "grad_norm": 4.762578071494993,
+      "learning_rate": 1e-06,
+      "loss": 0.4782,
+      "step": 3667
+    },
+    {
+      "epoch": 0.7014054880963764,
+      "grad_norm": 4.082230996533344,
+      "learning_rate": 1e-06,
+      "loss": 0.3235,
+      "step": 3668
+    },
+    {
+      "epoch": 0.7015967109666316,
+      "grad_norm": 2.734927486734618,
+      "learning_rate": 1e-06,
+      "loss": 0.1699,
+      "step": 3669
+    },
+    {
+      "epoch": 0.7017879338368869,
+      "grad_norm": 2.846876165764546,
+      "learning_rate": 1e-06,
+      "loss": 0.1545,
+      "step": 3670
+    },
+    {
+      "epoch": 0.7019791567071422,
+      "grad_norm": 5.089760183144568,
+      "learning_rate": 1e-06,
+      "loss": 0.1231,
+      "step": 3671
+    },
+    {
+      "epoch": 0.7021703795773975,
+      "grad_norm": 3.4082513931485345,
+      "learning_rate": 1e-06,
+      "loss": 0.0636,
+      "step": 3672
+    },
+    {
+      "epoch": 0.7023616024476528,
+      "grad_norm": 2.5153234075934643,
+      "learning_rate": 1e-06,
+      "loss": 0.0434,
+      "step": 3673
+    },
+    {
+      "epoch": 0.702552825317908,
+      "grad_norm": 2.8080458232585914,
+      "learning_rate": 1e-06,
+      "loss": 0.0744,
+      "step": 3674
+    },
+    {
+      "epoch": 0.7027440481881633,
+      "grad_norm": 4.105501735565633,
+      "learning_rate": 1e-06,
+      "loss": 0.1013,
+      "step": 3675
+    },
+    {
+      "epoch": 0.7029352710584186,
+      "grad_norm": 6.4273468243846725,
+      "learning_rate": 1e-06,
+      "loss": 0.3538,
+      "step": 3676
+    },
+    {
+      "epoch": 0.7031264939286739,
+      "grad_norm": 4.3652983863455495,
+      "learning_rate": 1e-06,
+      "loss": 0.427,
+      "step": 3677
+    },
+    {
+      "epoch": 0.7033177167989292,
+      "grad_norm": 3.825194813872191,
+      "learning_rate": 1e-06,
+      "loss": 0.3109,
+      "step": 3678
+    },
+    {
+      "epoch": 0.7035089396691845,
+      "grad_norm": 2.081193193456578,
+      "learning_rate": 1e-06,
+      "loss": 0.1879,
+      "step": 3679
+    },
+    {
+      "epoch": 0.7037001625394397,
+      "grad_norm": 2.9018138540391685,
+      "learning_rate": 1e-06,
+      "loss": 0.2357,
+      "step": 3680
+    },
+    {
+      "epoch": 0.703891385409695,
+      "grad_norm": 4.492757903663697,
+      "learning_rate": 1e-06,
+      "loss": 0.1447,
+      "step": 3681
+    },
+    {
+      "epoch": 0.7040826082799503,
+      "grad_norm": 2.8220648987875436,
+      "learning_rate": 1e-06,
+      "loss": 0.083,
+      "step": 3682
+    },
+    {
+      "epoch": 0.7042738311502056,
+      "grad_norm": 3.8818157920074743,
+      "learning_rate": 1e-06,
+      "loss": 0.1193,
+      "step": 3683
+    },
+    {
+      "epoch": 0.7044650540204609,
+      "grad_norm": 1.4877275853802643,
+      "learning_rate": 1e-06,
+      "loss": 0.0996,
+      "step": 3684
+    },
+    {
+      "epoch": 0.7046562768907161,
+      "grad_norm": 2.506925527043939,
+      "learning_rate": 1e-06,
+      "loss": 0.1232,
+      "step": 3685
+    },
+    {
+      "epoch": 0.7048474997609714,
+      "grad_norm": 3.159960086425576,
+      "learning_rate": 1e-06,
+      "loss": 0.0863,
+      "step": 3686
+    },
+    {
+      "epoch": 0.7050387226312267,
+      "grad_norm": 4.538156441948819,
+      "learning_rate": 1e-06,
+      "loss": 0.0738,
+      "step": 3687
+    },
+    {
+      "epoch": 0.705229945501482,
+      "grad_norm": 5.572079585540104,
+      "learning_rate": 1e-06,
+      "loss": 0.206,
+      "step": 3688
+    },
+    {
+      "epoch": 0.7054211683717373,
+      "grad_norm": 6.114801440879087,
+      "learning_rate": 1e-06,
+      "loss": 0.4607,
+      "step": 3689
+    },
+    {
+      "epoch": 0.7056123912419925,
+      "grad_norm": 4.615184988693129,
+      "learning_rate": 1e-06,
+      "loss": 0.3339,
+      "step": 3690
+    },
+    {
+      "epoch": 0.7058036141122478,
+      "grad_norm": 2.9806247335391047,
+      "learning_rate": 1e-06,
+      "loss": 0.3153,
+      "step": 3691
+    },
+    {
+      "epoch": 0.7059948369825031,
+      "grad_norm": 3.0906777588571472,
+      "learning_rate": 1e-06,
+      "loss": 0.3438,
+      "step": 3692
+    },
+    {
+      "epoch": 0.7061860598527584,
+      "grad_norm": 4.985322007288739,
+      "learning_rate": 1e-06,
+      "loss": 0.7767,
+      "step": 3693
+    },
+    {
+      "epoch": 0.7063772827230137,
+      "grad_norm": 2.9459509524362684,
+      "learning_rate": 1e-06,
+      "loss": 0.1286,
+      "step": 3694
+    },
+    {
+      "epoch": 0.706568505593269,
+      "grad_norm": 2.3066340098338673,
+      "learning_rate": 1e-06,
+      "loss": 0.1426,
+      "step": 3695
+    },
+    {
+      "epoch": 0.7067597284635242,
+      "grad_norm": 4.083642727295611,
+      "learning_rate": 1e-06,
+      "loss": 0.1346,
+      "step": 3696
+    },
+    {
+      "epoch": 0.7069509513337795,
+      "grad_norm": 3.117562254748323,
+      "learning_rate": 1e-06,
+      "loss": 0.1594,
+      "step": 3697
+    },
+    {
+      "epoch": 0.7071421742040348,
+      "grad_norm": 2.475915385037616,
+      "learning_rate": 1e-06,
+      "loss": 0.1083,
+      "step": 3698
+    },
+    {
+      "epoch": 0.7073333970742901,
+      "grad_norm": 1.976034342648841,
+      "learning_rate": 1e-06,
+      "loss": 0.0499,
+      "step": 3699
+    },
+    {
+      "epoch": 0.7075246199445454,
+      "grad_norm": 4.217336912501173,
+      "learning_rate": 1e-06,
+      "loss": 0.1611,
+      "step": 3700
+    },
+    {
+      "epoch": 0.7077158428148006,
+      "grad_norm": 4.648453912786151,
+      "learning_rate": 1e-06,
+      "loss": 0.2239,
+      "step": 3701
+    },
+    {
+      "epoch": 0.7079070656850559,
+      "grad_norm": 10.400998551774894,
+      "learning_rate": 1e-06,
+      "loss": 0.571,
+      "step": 3702
+    },
+    {
+      "epoch": 0.7080982885553112,
+      "grad_norm": 2.836742444518572,
+      "learning_rate": 1e-06,
+      "loss": 0.3033,
+      "step": 3703
+    },
+    {
+      "epoch": 0.7082895114255665,
+      "grad_norm": 2.121617486128517,
+      "learning_rate": 1e-06,
+      "loss": 0.2761,
+      "step": 3704
+    },
+    {
+      "epoch": 0.7084807342958218,
+      "grad_norm": 2.7339645513651667,
+      "learning_rate": 1e-06,
+      "loss": 0.2234,
+      "step": 3705
+    },
+    {
+      "epoch": 0.7086719571660771,
+      "grad_norm": 2.811214068273473,
+      "learning_rate": 1e-06,
+      "loss": 0.2057,
+      "step": 3706
+    },
+    {
+      "epoch": 0.7088631800363323,
+      "grad_norm": 2.2833012803698236,
+      "learning_rate": 1e-06,
+      "loss": 0.2222,
+      "step": 3707
+    },
+    {
+      "epoch": 0.7090544029065876,
+      "grad_norm": 4.464604729289475,
+      "learning_rate": 1e-06,
+      "loss": 0.2384,
+      "step": 3708
+    },
+    {
+      "epoch": 0.7092456257768429,
+      "grad_norm": 3.9325180022257484,
+      "learning_rate": 1e-06,
+      "loss": 0.0894,
+      "step": 3709
+    },
+    {
+      "epoch": 0.7094368486470982,
+      "grad_norm": 7.284326517343885,
+      "learning_rate": 1e-06,
+      "loss": 0.3116,
+      "step": 3710
+    },
+    {
+      "epoch": 0.7096280715173535,
+      "grad_norm": 3.8709102090538883,
+      "learning_rate": 1e-06,
+      "loss": 0.0608,
+      "step": 3711
+    },
+    {
+      "epoch": 0.7098192943876087,
+      "grad_norm": 5.006891270021683,
+      "learning_rate": 1e-06,
+      "loss": 0.1034,
+      "step": 3712
+    },
+    {
+      "epoch": 0.710010517257864,
+      "grad_norm": 6.5810082430148125,
+      "learning_rate": 1e-06,
+      "loss": 0.1165,
+      "step": 3713
+    },
+    {
+      "epoch": 0.7102017401281193,
+      "grad_norm": 4.409962706462232,
+      "learning_rate": 1e-06,
+      "loss": 0.533,
+      "step": 3714
+    },
+    {
+      "epoch": 0.7103929629983746,
+      "grad_norm": 4.755367859932223,
+      "learning_rate": 1e-06,
+      "loss": 0.8542,
+      "step": 3715
+    },
+    {
+      "epoch": 0.7105841858686299,
+      "grad_norm": 5.915145069449012,
+      "learning_rate": 1e-06,
+      "loss": 0.6715,
+      "step": 3716
+    },
+    {
+      "epoch": 0.7107754087388851,
+      "grad_norm": 1.712429354247697,
+      "learning_rate": 1e-06,
+      "loss": 0.2831,
+      "step": 3717
+    },
+    {
+      "epoch": 0.7109666316091404,
+      "grad_norm": 4.538659923251111,
+      "learning_rate": 1e-06,
+      "loss": 0.5849,
+      "step": 3718
+    },
+    {
+      "epoch": 0.7111578544793957,
+      "grad_norm": 3.243480158112853,
+      "learning_rate": 1e-06,
+      "loss": 0.3638,
+      "step": 3719
+    },
+    {
+      "epoch": 0.711349077349651,
+      "grad_norm": 2.2435451743481387,
+      "learning_rate": 1e-06,
+      "loss": 0.2031,
+      "step": 3720
+    },
+    {
+      "epoch": 0.7115403002199063,
+      "grad_norm": 2.826652607238041,
+      "learning_rate": 1e-06,
+      "loss": 0.2336,
+      "step": 3721
+    },
+    {
+      "epoch": 0.7117315230901616,
+      "grad_norm": 2.2830182890725483,
+      "learning_rate": 1e-06,
+      "loss": 0.1022,
+      "step": 3722
+    },
+    {
+      "epoch": 0.7119227459604168,
+      "grad_norm": 4.010363742759896,
+      "learning_rate": 1e-06,
+      "loss": 0.1452,
+      "step": 3723
+    },
+    {
+      "epoch": 0.7121139688306721,
+      "grad_norm": 3.8079688239304006,
+      "learning_rate": 1e-06,
+      "loss": 0.1079,
+      "step": 3724
+    },
+    {
+      "epoch": 0.7123051917009274,
+      "grad_norm": 3.914848446695696,
+      "learning_rate": 1e-06,
+      "loss": 0.1238,
+      "step": 3725
+    },
+    {
+      "epoch": 0.7124964145711827,
+      "grad_norm": 5.752033703276852,
+      "learning_rate": 1e-06,
+      "loss": 0.4358,
+      "step": 3726
+    },
+    {
+      "epoch": 0.712687637441438,
+      "grad_norm": 4.9634805243916915,
+      "learning_rate": 1e-06,
+      "loss": 0.5357,
+      "step": 3727
+    },
+    {
+      "epoch": 0.7128788603116932,
+      "grad_norm": 3.933814609846727,
+      "learning_rate": 1e-06,
+      "loss": 0.4144,
+      "step": 3728
+    },
+    {
+      "epoch": 0.7130700831819485,
+      "grad_norm": 3.1631175949712493,
+      "learning_rate": 1e-06,
+      "loss": 0.2481,
+      "step": 3729
+    },
+    {
+      "epoch": 0.7132613060522038,
+      "grad_norm": 2.4951130786720084,
+      "learning_rate": 1e-06,
+      "loss": 0.2447,
+      "step": 3730
+    },
+    {
+      "epoch": 0.7134525289224591,
+      "grad_norm": 1.6535826022657196,
+      "learning_rate": 1e-06,
+      "loss": 0.3353,
+      "step": 3731
+    },
+    {
+      "epoch": 0.7136437517927144,
+      "grad_norm": 4.443633233318864,
+      "learning_rate": 1e-06,
+      "loss": 0.4297,
+      "step": 3732
+    },
+    {
+      "epoch": 0.7138349746629697,
+      "grad_norm": 3.996464478121898,
+      "learning_rate": 1e-06,
+      "loss": 0.2599,
+      "step": 3733
+    },
+    {
+      "epoch": 0.7140261975332249,
+      "grad_norm": 5.072625477584546,
+      "learning_rate": 1e-06,
+      "loss": 0.1956,
+      "step": 3734
+    },
+    {
+      "epoch": 0.7142174204034802,
+      "grad_norm": 4.732101246941332,
+      "learning_rate": 1e-06,
+      "loss": 0.0832,
+      "step": 3735
+    },
+    {
+      "epoch": 0.7144086432737355,
+      "grad_norm": 6.679378248049558,
+      "learning_rate": 1e-06,
+      "loss": 0.2988,
+      "step": 3736
+    },
+    {
+      "epoch": 0.7145998661439908,
+      "grad_norm": 2.802584569031066,
+      "learning_rate": 1e-06,
+      "loss": 0.1092,
+      "step": 3737
+    },
+    {
+      "epoch": 0.7147910890142462,
+      "grad_norm": 5.5701668012699335,
+      "learning_rate": 1e-06,
+      "loss": 0.2171,
+      "step": 3738
+    },
+    {
+      "epoch": 0.7149823118845013,
+      "grad_norm": 5.140761909864725,
+      "learning_rate": 1e-06,
+      "loss": 0.3739,
+      "step": 3739
+    },
+    {
+      "epoch": 0.7151735347547566,
+      "grad_norm": 4.638430092382964,
+      "learning_rate": 1e-06,
+      "loss": 0.5262,
+      "step": 3740
+    },
+    {
+      "epoch": 0.715364757625012,
+      "grad_norm": 5.968254028804719,
+      "learning_rate": 1e-06,
+      "loss": 0.343,
+      "step": 3741
+    },
+    {
+      "epoch": 0.7155559804952673,
+      "grad_norm": 2.2247775234479317,
+      "learning_rate": 1e-06,
+      "loss": 0.3247,
+      "step": 3742
+    },
+    {
+      "epoch": 0.7157472033655226,
+      "grad_norm": 4.457559084583592,
+      "learning_rate": 1e-06,
+      "loss": 0.3415,
+      "step": 3743
+    },
+    {
+      "epoch": 0.7159384262357777,
+      "grad_norm": 3.80035103130232,
+      "learning_rate": 1e-06,
+      "loss": 0.3027,
+      "step": 3744
+    },
+    {
+      "epoch": 0.716129649106033,
+      "grad_norm": 2.004635921537245,
+      "learning_rate": 1e-06,
+      "loss": 0.1311,
+      "step": 3745
+    },
+    {
+      "epoch": 0.7163208719762884,
+      "grad_norm": 3.3778092737012186,
+      "learning_rate": 1e-06,
+      "loss": 0.0971,
+      "step": 3746
+    },
+    {
+      "epoch": 0.7165120948465437,
+      "grad_norm": 5.251484978786655,
+      "learning_rate": 1e-06,
+      "loss": 0.1317,
+      "step": 3747
+    },
+    {
+      "epoch": 0.716703317716799,
+      "grad_norm": 3.8986209680963935,
+      "learning_rate": 1e-06,
+      "loss": 0.0804,
+      "step": 3748
+    },
+    {
+      "epoch": 0.7168945405870543,
+      "grad_norm": 3.469398369206178,
+      "learning_rate": 1e-06,
+      "loss": 0.0935,
+      "step": 3749
+    },
+    {
+      "epoch": 0.7170857634573095,
+      "grad_norm": 4.152984483579202,
+      "learning_rate": 1e-06,
+      "loss": 0.1812,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7172769863275648,
+      "grad_norm": 5.220424908703984,
+      "learning_rate": 1e-06,
+      "loss": 0.4423,
+      "step": 3751
+    },
+    {
+      "epoch": 0.7174682091978201,
+      "grad_norm": 4.394881713455842,
+      "learning_rate": 1e-06,
+      "loss": 0.2685,
+      "step": 3752
+    },
+    {
+      "epoch": 0.7176594320680754,
+      "grad_norm": 2.7006771933825124,
+      "learning_rate": 1e-06,
+      "loss": 0.2306,
+      "step": 3753
+    },
+    {
+      "epoch": 0.7178506549383307,
+      "grad_norm": 3.4519677618030897,
+      "learning_rate": 1e-06,
+      "loss": 0.3092,
+      "step": 3754
+    },
+    {
+      "epoch": 0.7180418778085859,
+      "grad_norm": 6.042794203502935,
+      "learning_rate": 1e-06,
+      "loss": 0.4386,
+      "step": 3755
+    },
+    {
+      "epoch": 0.7182331006788412,
+      "grad_norm": 3.8431470444574805,
+      "learning_rate": 1e-06,
+      "loss": 0.154,
+      "step": 3756
+    },
+    {
+      "epoch": 0.7184243235490965,
+      "grad_norm": 2.6766954628364137,
+      "learning_rate": 1e-06,
+      "loss": 0.0916,
+      "step": 3757
+    },
+    {
+      "epoch": 0.7186155464193518,
+      "grad_norm": 3.0142665827563073,
+      "learning_rate": 1e-06,
+      "loss": 0.1034,
+      "step": 3758
+    },
+    {
+      "epoch": 0.7188067692896071,
+      "grad_norm": 1.9529384066141684,
+      "learning_rate": 1e-06,
+      "loss": 0.0706,
+      "step": 3759
+    },
+    {
+      "epoch": 0.7189979921598623,
+      "grad_norm": 3.8880714707889426,
+      "learning_rate": 1e-06,
+      "loss": 0.1811,
+      "step": 3760
+    },
+    {
+      "epoch": 0.7191892150301176,
+      "grad_norm": 3.6724097329219627,
+      "learning_rate": 1e-06,
+      "loss": 0.0929,
+      "step": 3761
+    },
+    {
+      "epoch": 0.7193804379003729,
+      "grad_norm": 3.3870548138751495,
+      "learning_rate": 1e-06,
+      "loss": 0.0843,
+      "step": 3762
+    },
+    {
+      "epoch": 0.7195716607706282,
+      "grad_norm": 7.2555649544909935,
+      "learning_rate": 1e-06,
+      "loss": 0.2736,
+      "step": 3763
+    },
+    {
+      "epoch": 0.7197628836408835,
+      "grad_norm": 5.154285403430774,
+      "learning_rate": 1e-06,
+      "loss": 0.6091,
+      "step": 3764
+    },
+    {
+      "epoch": 0.7199541065111388,
+      "grad_norm": 6.2000112718049,
+      "learning_rate": 1e-06,
+      "loss": 0.8397,
+      "step": 3765
+    },
+    {
+      "epoch": 0.720145329381394,
+      "grad_norm": 7.174758896448575,
+      "learning_rate": 1e-06,
+      "loss": 0.5008,
+      "step": 3766
+    },
+    {
+      "epoch": 0.7203365522516493,
+      "grad_norm": 3.899569238201865,
+      "learning_rate": 1e-06,
+      "loss": 0.1519,
+      "step": 3767
+    },
+    {
+      "epoch": 0.7205277751219046,
+      "grad_norm": 3.5148159876445813,
+      "learning_rate": 1e-06,
+      "loss": 0.3045,
+      "step": 3768
+    },
+    {
+      "epoch": 0.7207189979921599,
+      "grad_norm": 4.191853324318796,
+      "learning_rate": 1e-06,
+      "loss": 0.1793,
+      "step": 3769
+    },
+    {
+      "epoch": 0.7209102208624152,
+      "grad_norm": 2.392414794329407,
+      "learning_rate": 1e-06,
+      "loss": 0.1817,
+      "step": 3770
+    },
+    {
+      "epoch": 0.7211014437326704,
+      "grad_norm": 3.786135457303371,
+      "learning_rate": 1e-06,
+      "loss": 0.1715,
+      "step": 3771
+    },
+    {
+      "epoch": 0.7212926666029257,
+      "grad_norm": 2.5597321984169175,
+      "learning_rate": 1e-06,
+      "loss": 0.1199,
+      "step": 3772
+    },
+    {
+      "epoch": 0.721483889473181,
+      "grad_norm": 3.2475678072832954,
+      "learning_rate": 1e-06,
+      "loss": 0.0902,
+      "step": 3773
+    },
+    {
+      "epoch": 0.7216751123434363,
+      "grad_norm": 3.516498372418599,
+      "learning_rate": 1e-06,
+      "loss": 0.0805,
+      "step": 3774
+    },
+    {
+      "epoch": 0.7218663352136916,
+      "grad_norm": 3.7277198130053257,
+      "learning_rate": 1e-06,
+      "loss": 0.1258,
+      "step": 3775
+    },
+    {
+      "epoch": 0.7220575580839469,
+      "grad_norm": 5.169126581314049,
+      "learning_rate": 1e-06,
+      "loss": 0.6791,
+      "step": 3776
+    },
+    {
+      "epoch": 0.7222487809542021,
+      "grad_norm": 3.670543283065722,
+      "learning_rate": 1e-06,
+      "loss": 0.257,
+      "step": 3777
+    },
+    {
+      "epoch": 0.7224400038244574,
+      "grad_norm": 3.772640799693979,
+      "learning_rate": 1e-06,
+      "loss": 0.2448,
+      "step": 3778
+    },
+    {
+      "epoch": 0.7226312266947127,
+      "grad_norm": 3.749079018985941,
+      "learning_rate": 1e-06,
+      "loss": 0.2704,
+      "step": 3779
+    },
+    {
+      "epoch": 0.722822449564968,
+      "grad_norm": 3.5940072672435583,
+      "learning_rate": 1e-06,
+      "loss": 0.4338,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7230136724352233,
+      "grad_norm": 3.121778124278944,
+      "learning_rate": 1e-06,
+      "loss": 0.2162,
+      "step": 3781
+    },
+    {
+      "epoch": 0.7232048953054785,
+      "grad_norm": 2.2249347291438,
+      "learning_rate": 1e-06,
+      "loss": 0.1755,
+      "step": 3782
+    },
+    {
+      "epoch": 0.7233961181757338,
+      "grad_norm": 3.828927216845075,
+      "learning_rate": 1e-06,
+      "loss": 0.1208,
+      "step": 3783
+    },
+    {
+      "epoch": 0.7235873410459891,
+      "grad_norm": 2.8178219775247486,
+      "learning_rate": 1e-06,
+      "loss": 0.2488,
+      "step": 3784
+    },
+    {
+      "epoch": 0.7237785639162444,
+      "grad_norm": 1.6307359111901867,
+      "learning_rate": 1e-06,
+      "loss": 0.0507,
+      "step": 3785
+    },
+    {
+      "epoch": 0.7239697867864997,
+      "grad_norm": 3.3402531958451167,
+      "learning_rate": 1e-06,
+      "loss": 0.1066,
+      "step": 3786
+    },
+    {
+      "epoch": 0.7241610096567549,
+      "grad_norm": 5.6048953730051005,
+      "learning_rate": 1e-06,
+      "loss": 0.1152,
+      "step": 3787
+    },
+    {
+      "epoch": 0.7243522325270102,
+      "grad_norm": 5.580539212989623,
+      "learning_rate": 1e-06,
+      "loss": 0.2232,
+      "step": 3788
+    },
+    {
+      "epoch": 0.7245434553972655,
+      "grad_norm": 4.51065074615721,
+      "learning_rate": 1e-06,
+      "loss": 0.3734,
+      "step": 3789
+    },
+    {
+      "epoch": 0.7247346782675208,
+      "grad_norm": 2.637618028793064,
+      "learning_rate": 1e-06,
+      "loss": 0.172,
+      "step": 3790
+    },
+    {
+      "epoch": 0.7249259011377761,
+      "grad_norm": 2.372097148719902,
+      "learning_rate": 1e-06,
+      "loss": 0.0727,
+      "step": 3791
+    },
+    {
+      "epoch": 0.7251171240080314,
+      "grad_norm": 2.1600733597978223,
+      "learning_rate": 1e-06,
+      "loss": 0.3617,
+      "step": 3792
+    },
+    {
+      "epoch": 0.7253083468782866,
+      "grad_norm": 4.852087708085368,
+      "learning_rate": 1e-06,
+      "loss": 0.6421,
+      "step": 3793
+    },
+    {
+      "epoch": 0.7254995697485419,
+      "grad_norm": 3.654739328630935,
+      "learning_rate": 1e-06,
+      "loss": 0.3633,
+      "step": 3794
+    },
+    {
+      "epoch": 0.7256907926187972,
+      "grad_norm": 4.850739979088403,
+      "learning_rate": 1e-06,
+      "loss": 0.2938,
+      "step": 3795
+    },
+    {
+      "epoch": 0.7258820154890525,
+      "grad_norm": 3.1477740703392914,
+      "learning_rate": 1e-06,
+      "loss": 0.1405,
+      "step": 3796
+    },
+    {
+      "epoch": 0.7260732383593078,
+      "grad_norm": 3.8479749828326404,
+      "learning_rate": 1e-06,
+      "loss": 0.1698,
+      "step": 3797
+    },
+    {
+      "epoch": 0.726264461229563,
+      "grad_norm": 2.049777121174514,
+      "learning_rate": 1e-06,
+      "loss": 0.1221,
+      "step": 3798
+    },
+    {
+      "epoch": 0.7264556840998183,
+      "grad_norm": 4.620066098817882,
+      "learning_rate": 1e-06,
+      "loss": 0.1274,
+      "step": 3799
+    },
+    {
+      "epoch": 0.7266469069700736,
+      "grad_norm": 9.736053273139088,
+      "learning_rate": 1e-06,
+      "loss": 0.1565,
+      "step": 3800
+    },
+    {
+      "epoch": 0.7268381298403289,
+      "grad_norm": 7.115588480671822,
+      "learning_rate": 1e-06,
+      "loss": 0.3699,
+      "step": 3801
+    },
+    {
+      "epoch": 0.7270293527105842,
+      "grad_norm": 4.303926210189239,
+      "learning_rate": 1e-06,
+      "loss": 0.4497,
+      "step": 3802
+    },
+    {
+      "epoch": 0.7272205755808395,
+      "grad_norm": 4.464945420703738,
+      "learning_rate": 1e-06,
+      "loss": 0.288,
+      "step": 3803
+    },
+    {
+      "epoch": 0.7274117984510947,
+      "grad_norm": 3.4725512047789895,
+      "learning_rate": 1e-06,
+      "loss": 0.5472,
+      "step": 3804
+    },
+    {
+      "epoch": 0.72760302132135,
+      "grad_norm": 4.625979345609517,
+      "learning_rate": 1e-06,
+      "loss": 0.1918,
+      "step": 3805
+    },
+    {
+      "epoch": 0.7277942441916053,
+      "grad_norm": 2.652634617358374,
+      "learning_rate": 1e-06,
+      "loss": 0.1899,
+      "step": 3806
+    },
+    {
+      "epoch": 0.7279854670618606,
+      "grad_norm": 2.6665747646549094,
+      "learning_rate": 1e-06,
+      "loss": 0.1038,
+      "step": 3807
+    },
+    {
+      "epoch": 0.7281766899321159,
+      "grad_norm": 4.197859064027227,
+      "learning_rate": 1e-06,
+      "loss": 0.1659,
+      "step": 3808
+    },
+    {
+      "epoch": 0.7283679128023711,
+      "grad_norm": 3.4480147026084875,
+      "learning_rate": 1e-06,
+      "loss": 0.1581,
+      "step": 3809
+    },
+    {
+      "epoch": 0.7285591356726264,
+      "grad_norm": 3.777779239454797,
+      "learning_rate": 1e-06,
+      "loss": 0.1363,
+      "step": 3810
+    },
+    {
+      "epoch": 0.7287503585428817,
+      "grad_norm": 4.41274199862298,
+      "learning_rate": 1e-06,
+      "loss": 0.0691,
+      "step": 3811
+    },
+    {
+      "epoch": 0.728941581413137,
+      "grad_norm": 16.35439349298645,
+      "learning_rate": 1e-06,
+      "loss": 0.067,
+      "step": 3812
+    },
+    {
+      "epoch": 0.7291328042833923,
+      "grad_norm": 6.784629717894378,
+      "learning_rate": 1e-06,
+      "loss": 0.2447,
+      "step": 3813
+    },
+    {
+      "epoch": 0.7293240271536475,
+      "grad_norm": 5.00601940219242,
+      "learning_rate": 1e-06,
+      "loss": 0.421,
+      "step": 3814
+    },
+    {
+      "epoch": 0.7295152500239028,
+      "grad_norm": 5.3626581484364815,
+      "learning_rate": 1e-06,
+      "loss": 0.6263,
+      "step": 3815
+    },
+    {
+      "epoch": 0.7297064728941581,
+      "grad_norm": 5.625414684692664,
+      "learning_rate": 1e-06,
+      "loss": 0.4979,
+      "step": 3816
+    },
+    {
+      "epoch": 0.7298976957644134,
+      "grad_norm": 2.819910755233305,
+      "learning_rate": 1e-06,
+      "loss": 0.2035,
+      "step": 3817
+    },
+    {
+      "epoch": 0.7300889186346687,
+      "grad_norm": 3.5642020191945307,
+      "learning_rate": 1e-06,
+      "loss": 0.1773,
+      "step": 3818
+    },
+    {
+      "epoch": 0.730280141504924,
+      "grad_norm": 4.099867813956804,
+      "learning_rate": 1e-06,
+      "loss": 0.3236,
+      "step": 3819
+    },
+    {
+      "epoch": 0.7304713643751792,
+      "grad_norm": 2.644698392693933,
+      "learning_rate": 1e-06,
+      "loss": 0.1284,
+      "step": 3820
+    },
+    {
+      "epoch": 0.7306625872454345,
+      "grad_norm": 3.0326775404368362,
+      "learning_rate": 1e-06,
+      "loss": 0.2231,
+      "step": 3821
+    },
+    {
+      "epoch": 0.7308538101156898,
+      "grad_norm": 4.540326098523389,
+      "learning_rate": 1e-06,
+      "loss": 0.1499,
+      "step": 3822
+    },
+    {
+      "epoch": 0.7310450329859451,
+      "grad_norm": 6.6825974221768565,
+      "learning_rate": 1e-06,
+      "loss": 0.0814,
+      "step": 3823
+    },
+    {
+      "epoch": 0.7312362558562004,
+      "grad_norm": 4.059194540545349,
+      "learning_rate": 1e-06,
+      "loss": 0.2513,
+      "step": 3824
+    },
+    {
+      "epoch": 0.7314274787264556,
+      "grad_norm": 4.113973039522971,
+      "learning_rate": 1e-06,
+      "loss": 0.072,
+      "step": 3825
+    },
+    {
+      "epoch": 0.7316187015967109,
+      "grad_norm": 7.159395950666367,
+      "learning_rate": 1e-06,
+      "loss": 0.4882,
+      "step": 3826
+    },
+    {
+      "epoch": 0.7318099244669662,
+      "grad_norm": 3.4662602742727593,
+      "learning_rate": 1e-06,
+      "loss": 0.265,
+      "step": 3827
+    },
+    {
+      "epoch": 0.7320011473372215,
+      "grad_norm": 5.517642595032642,
+      "learning_rate": 1e-06,
+      "loss": 0.4044,
+      "step": 3828
+    },
+    {
+      "epoch": 0.7321923702074769,
+      "grad_norm": 4.387350035275422,
+      "learning_rate": 1e-06,
+      "loss": 0.2526,
+      "step": 3829
+    },
+    {
+      "epoch": 0.732383593077732,
+      "grad_norm": 3.9288466295234623,
+      "learning_rate": 1e-06,
+      "loss": 0.1933,
+      "step": 3830
+    },
+    {
+      "epoch": 0.7325748159479873,
+      "grad_norm": 4.905666049295145,
+      "learning_rate": 1e-06,
+      "loss": 0.4543,
+      "step": 3831
+    },
+    {
+      "epoch": 0.7327660388182426,
+      "grad_norm": 4.391468550805624,
+      "learning_rate": 1e-06,
+      "loss": 0.3817,
+      "step": 3832
+    },
+    {
+      "epoch": 0.732957261688498,
+      "grad_norm": 4.256547932553329,
+      "learning_rate": 1e-06,
+      "loss": 0.2142,
+      "step": 3833
+    },
+    {
+      "epoch": 0.7331484845587533,
+      "grad_norm": 2.8585544096602886,
+      "learning_rate": 1e-06,
+      "loss": 0.0623,
+      "step": 3834
+    },
+    {
+      "epoch": 0.7333397074290086,
+      "grad_norm": 3.5985510559318783,
+      "learning_rate": 1e-06,
+      "loss": 0.156,
+      "step": 3835
+    },
+    {
+      "epoch": 0.7335309302992638,
+      "grad_norm": 3.689497325949495,
+      "learning_rate": 1e-06,
+      "loss": 0.1277,
+      "step": 3836
+    },
+    {
+      "epoch": 0.733722153169519,
+      "grad_norm": 6.3612075901997756,
+      "learning_rate": 1e-06,
+      "loss": 0.1052,
+      "step": 3837
+    },
+    {
+      "epoch": 0.7339133760397744,
+      "grad_norm": 6.411895497403939,
+      "learning_rate": 1e-06,
+      "loss": 0.2323,
+      "step": 3838
+    },
+    {
+      "epoch": 0.7341045989100297,
+      "grad_norm": 4.33950312639975,
+      "learning_rate": 1e-06,
+      "loss": 0.3727,
+      "step": 3839
+    },
+    {
+      "epoch": 0.734295821780285,
+      "grad_norm": 3.4880895957113776,
+      "learning_rate": 1e-06,
+      "loss": 0.2749,
+      "step": 3840
+    },
+    {
+      "epoch": 0.7344870446505402,
+      "grad_norm": 3.7208684449000518,
+      "learning_rate": 1e-06,
+      "loss": 0.3232,
+      "step": 3841
+    },
+    {
+      "epoch": 0.7346782675207955,
+      "grad_norm": 2.1410460788876464,
+      "learning_rate": 1e-06,
+      "loss": 0.134,
+      "step": 3842
+    },
+    {
+      "epoch": 0.7348694903910508,
+      "grad_norm": 2.672536756871155,
+      "learning_rate": 1e-06,
+      "loss": 0.0956,
+      "step": 3843
+    },
+    {
+      "epoch": 0.7350607132613061,
+      "grad_norm": 2.2673656222766074,
+      "learning_rate": 1e-06,
+      "loss": 0.1685,
+      "step": 3844
+    },
+    {
+      "epoch": 0.7352519361315614,
+      "grad_norm": 4.1272357314945305,
+      "learning_rate": 1e-06,
+      "loss": 0.214,
+      "step": 3845
+    },
+    {
+      "epoch": 0.7354431590018167,
+      "grad_norm": 2.7460000685534154,
+      "learning_rate": 1e-06,
+      "loss": 0.0743,
+      "step": 3846
+    },
+    {
+      "epoch": 0.7356343818720719,
+      "grad_norm": 3.3331039508732188,
+      "learning_rate": 1e-06,
+      "loss": 0.1344,
+      "step": 3847
+    },
+    {
+      "epoch": 0.7358256047423272,
+      "grad_norm": 2.338900724739981,
+      "learning_rate": 1e-06,
+      "loss": 0.1635,
+      "step": 3848
+    },
+    {
+      "epoch": 0.7360168276125825,
+      "grad_norm": 2.066671833934272,
+      "learning_rate": 1e-06,
+      "loss": 0.0482,
+      "step": 3849
+    },
+    {
+      "epoch": 0.7362080504828378,
+      "grad_norm": 3.8878044720608154,
+      "learning_rate": 1e-06,
+      "loss": 0.1154,
+      "step": 3850
+    },
+    {
+      "epoch": 0.7363992733530931,
+      "grad_norm": 5.07786301157113,
+      "learning_rate": 1e-06,
+      "loss": 0.187,
+      "step": 3851
+    },
+    {
+      "epoch": 0.7365904962233483,
+      "grad_norm": 3.343521395560495,
+      "learning_rate": 1e-06,
+      "loss": 0.4081,
+      "step": 3852
+    },
+    {
+      "epoch": 0.7367817190936036,
+      "grad_norm": 4.076677906044797,
+      "learning_rate": 1e-06,
+      "loss": 0.3749,
+      "step": 3853
+    },
+    {
+      "epoch": 0.7369729419638589,
+      "grad_norm": 5.120292921031761,
+      "learning_rate": 1e-06,
+      "loss": 0.2391,
+      "step": 3854
+    },
+    {
+      "epoch": 0.7371641648341142,
+      "grad_norm": 2.836219290192588,
+      "learning_rate": 1e-06,
+      "loss": 0.2278,
+      "step": 3855
+    },
+    {
+      "epoch": 0.7373553877043695,
+      "grad_norm": 4.313969679583099,
+      "learning_rate": 1e-06,
+      "loss": 0.3763,
+      "step": 3856
+    },
+    {
+      "epoch": 0.7375466105746247,
+      "grad_norm": 2.67069463702007,
+      "learning_rate": 1e-06,
+      "loss": 0.1748,
+      "step": 3857
+    },
+    {
+      "epoch": 0.73773783344488,
+      "grad_norm": 4.027647552723545,
+      "learning_rate": 1e-06,
+      "loss": 0.2934,
+      "step": 3858
+    },
+    {
+      "epoch": 0.7379290563151353,
+      "grad_norm": 2.1053034163691433,
+      "learning_rate": 1e-06,
+      "loss": 0.1476,
+      "step": 3859
+    },
+    {
+      "epoch": 0.7381202791853906,
+      "grad_norm": 2.9721283011486515,
+      "learning_rate": 1e-06,
+      "loss": 0.258,
+      "step": 3860
+    },
+    {
+      "epoch": 0.7383115020556459,
+      "grad_norm": 2.646314371990148,
+      "learning_rate": 1e-06,
+      "loss": 0.0554,
+      "step": 3861
+    },
+    {
+      "epoch": 0.7385027249259012,
+      "grad_norm": 3.1806133899862865,
+      "learning_rate": 1e-06,
+      "loss": 0.0852,
+      "step": 3862
+    },
+    {
+      "epoch": 0.7386939477961564,
+      "grad_norm": 4.344085680388863,
+      "learning_rate": 1e-06,
+      "loss": 0.0809,
+      "step": 3863
+    },
+    {
+      "epoch": 0.7388851706664117,
+      "grad_norm": 6.098873547215246,
+      "learning_rate": 1e-06,
+      "loss": 0.3027,
+      "step": 3864
+    },
+    {
+      "epoch": 0.739076393536667,
+      "grad_norm": 4.691511751714952,
+      "learning_rate": 1e-06,
+      "loss": 0.4246,
+      "step": 3865
+    },
+    {
+      "epoch": 0.7392676164069223,
+      "grad_norm": 6.7236594072867275,
+      "learning_rate": 1e-06,
+      "loss": 0.2501,
+      "step": 3866
+    },
+    {
+      "epoch": 0.7394588392771776,
+      "grad_norm": 3.088954715652459,
+      "learning_rate": 1e-06,
+      "loss": 0.4093,
+      "step": 3867
+    },
+    {
+      "epoch": 0.7396500621474328,
+      "grad_norm": 2.04534983415094,
+      "learning_rate": 1e-06,
+      "loss": 0.1304,
+      "step": 3868
+    },
+    {
+      "epoch": 0.7398412850176881,
+      "grad_norm": 2.842331584950312,
+      "learning_rate": 1e-06,
+      "loss": 0.1453,
+      "step": 3869
+    },
+    {
+      "epoch": 0.7400325078879434,
+      "grad_norm": 1.8715044498660778,
+      "learning_rate": 1e-06,
+      "loss": 0.2635,
+      "step": 3870
+    },
+    {
+      "epoch": 0.7402237307581987,
+      "grad_norm": 1.7866125531447274,
+      "learning_rate": 1e-06,
+      "loss": 0.1301,
+      "step": 3871
+    },
+    {
+      "epoch": 0.740414953628454,
+      "grad_norm": 3.1665049227181292,
+      "learning_rate": 1e-06,
+      "loss": 0.1836,
+      "step": 3872
+    },
+    {
+      "epoch": 0.7406061764987093,
+      "grad_norm": 1.6438928665328447,
+      "learning_rate": 1e-06,
+      "loss": 0.0481,
+      "step": 3873
+    },
+    {
+      "epoch": 0.7407973993689645,
+      "grad_norm": 2.1884478014341697,
+      "learning_rate": 1e-06,
+      "loss": 0.0312,
+      "step": 3874
+    },
+    {
+      "epoch": 0.7409886222392198,
+      "grad_norm": 3.75725552878448,
+      "learning_rate": 1e-06,
+      "loss": 0.1558,
+      "step": 3875
+    },
+    {
+      "epoch": 0.7411798451094751,
+      "grad_norm": 3.771862388406448,
+      "learning_rate": 1e-06,
+      "loss": 0.2538,
+      "step": 3876
+    },
+    {
+      "epoch": 0.7413710679797304,
+      "grad_norm": 3.3819271629426346,
+      "learning_rate": 1e-06,
+      "loss": 0.242,
+      "step": 3877
+    },
+    {
+      "epoch": 0.7415622908499857,
+      "grad_norm": 5.011801905547954,
+      "learning_rate": 1e-06,
+      "loss": 0.4475,
+      "step": 3878
+    },
+    {
+      "epoch": 0.7417535137202409,
+      "grad_norm": 1.3841008795427365,
+      "learning_rate": 1e-06,
+      "loss": 0.1493,
+      "step": 3879
+    },
+    {
+      "epoch": 0.7419447365904962,
+      "grad_norm": 2.738856539514797,
+      "learning_rate": 1e-06,
+      "loss": 0.2779,
+      "step": 3880
+    },
+    {
+      "epoch": 0.7421359594607515,
+      "grad_norm": 5.021199488709836,
+      "learning_rate": 1e-06,
+      "loss": 0.5298,
+      "step": 3881
+    },
+    {
+      "epoch": 0.7423271823310068,
+      "grad_norm": 2.2351532761075004,
+      "learning_rate": 1e-06,
+      "loss": 0.1359,
+      "step": 3882
+    },
+    {
+      "epoch": 0.7425184052012621,
+      "grad_norm": 2.641359486655336,
+      "learning_rate": 1e-06,
+      "loss": 0.0981,
+      "step": 3883
+    },
+    {
+      "epoch": 0.7427096280715173,
+      "grad_norm": 2.4243200941720437,
+      "learning_rate": 1e-06,
+      "loss": 0.0496,
+      "step": 3884
+    },
+    {
+      "epoch": 0.7429008509417726,
+      "grad_norm": 4.324669956557876,
+      "learning_rate": 1e-06,
+      "loss": 0.3625,
+      "step": 3885
+    },
+    {
+      "epoch": 0.7430920738120279,
+      "grad_norm": 2.0271514624986464,
+      "learning_rate": 1e-06,
+      "loss": 0.0565,
+      "step": 3886
+    },
+    {
+      "epoch": 0.7432832966822832,
+      "grad_norm": 3.472484468517352,
+      "learning_rate": 1e-06,
+      "loss": 0.0935,
+      "step": 3887
+    },
+    {
+      "epoch": 0.7434745195525385,
+      "grad_norm": 2.3244387738884216,
+      "learning_rate": 1e-06,
+      "loss": 0.1888,
+      "step": 3888
+    },
+    {
+      "epoch": 0.7436657424227938,
+      "grad_norm": 6.891527320101904,
+      "learning_rate": 1e-06,
+      "loss": 0.6096,
+      "step": 3889
+    },
+    {
+      "epoch": 0.743856965293049,
+      "grad_norm": 3.936019013436457,
+      "learning_rate": 1e-06,
+      "loss": 0.4575,
+      "step": 3890
+    },
+    {
+      "epoch": 0.7440481881633043,
+      "grad_norm": 3.1316204895004103,
+      "learning_rate": 1e-06,
+      "loss": 0.1593,
+      "step": 3891
+    },
+    {
+      "epoch": 0.7442394110335596,
+      "grad_norm": 5.707170279791822,
+      "learning_rate": 1e-06,
+      "loss": 0.1155,
+      "step": 3892
+    },
+    {
+      "epoch": 0.7444306339038149,
+      "grad_norm": 2.5703779867346244,
+      "learning_rate": 1e-06,
+      "loss": 0.299,
+      "step": 3893
+    },
+    {
+      "epoch": 0.7446218567740702,
+      "grad_norm": 3.2846699648448854,
+      "learning_rate": 1e-06,
+      "loss": 0.3836,
+      "step": 3894
+    },
+    {
+      "epoch": 0.7448130796443254,
+      "grad_norm": 4.55061527326079,
+      "learning_rate": 1e-06,
+      "loss": 0.1949,
+      "step": 3895
+    },
+    {
+      "epoch": 0.7450043025145807,
+      "grad_norm": 2.2488903912353693,
+      "learning_rate": 1e-06,
+      "loss": 0.2664,
+      "step": 3896
+    },
+    {
+      "epoch": 0.745195525384836,
+      "grad_norm": 5.203040136613209,
+      "learning_rate": 1e-06,
+      "loss": 0.2468,
+      "step": 3897
+    },
+    {
+      "epoch": 0.7453867482550913,
+      "grad_norm": 1.7261923829116688,
+      "learning_rate": 1e-06,
+      "loss": 0.0772,
+      "step": 3898
+    },
+    {
+      "epoch": 0.7455779711253466,
+      "grad_norm": 3.667116368777817,
+      "learning_rate": 1e-06,
+      "loss": 0.076,
+      "step": 3899
+    },
+    {
+      "epoch": 0.7457691939956018,
+      "grad_norm": 7.484818160728417,
+      "learning_rate": 1e-06,
+      "loss": 0.2887,
+      "step": 3900
+    },
+    {
+      "epoch": 0.7459604168658571,
+      "grad_norm": 4.08697554930599,
+      "learning_rate": 1e-06,
+      "loss": 0.2248,
+      "step": 3901
+    },
+    {
+      "epoch": 0.7461516397361124,
+      "grad_norm": 4.038814810088527,
+      "learning_rate": 1e-06,
+      "loss": 0.3166,
+      "step": 3902
+    },
+    {
+      "epoch": 0.7463428626063677,
+      "grad_norm": 2.947349107291823,
+      "learning_rate": 1e-06,
+      "loss": 0.204,
+      "step": 3903
+    },
+    {
+      "epoch": 0.746534085476623,
+      "grad_norm": 2.85792412975238,
+      "learning_rate": 1e-06,
+      "loss": 0.1626,
+      "step": 3904
+    },
+    {
+      "epoch": 0.7467253083468783,
+      "grad_norm": 3.218251458234004,
+      "learning_rate": 1e-06,
+      "loss": 0.4827,
+      "step": 3905
+    },
+    {
+      "epoch": 0.7469165312171335,
+      "grad_norm": 3.6690002875921888,
+      "learning_rate": 1e-06,
+      "loss": 0.2663,
+      "step": 3906
+    },
+    {
+      "epoch": 0.7471077540873888,
+      "grad_norm": 4.2825987981388485,
+      "learning_rate": 1e-06,
+      "loss": 0.4475,
+      "step": 3907
+    },
+    {
+      "epoch": 0.7472989769576441,
+      "grad_norm": 5.7201096216114555,
+      "learning_rate": 1e-06,
+      "loss": 0.3254,
+      "step": 3908
+    },
+    {
+      "epoch": 0.7474901998278994,
+      "grad_norm": 3.938736010120744,
+      "learning_rate": 1e-06,
+      "loss": 0.1733,
+      "step": 3909
+    },
+    {
+      "epoch": 0.7476814226981547,
+      "grad_norm": 2.973880711136515,
+      "learning_rate": 1e-06,
+      "loss": 0.3524,
+      "step": 3910
+    },
+    {
+      "epoch": 0.7478726455684099,
+      "grad_norm": 3.4908429657478477,
+      "learning_rate": 1e-06,
+      "loss": 0.1949,
+      "step": 3911
+    },
+    {
+      "epoch": 0.7480638684386652,
+      "grad_norm": 3.6829226257665,
+      "learning_rate": 1e-06,
+      "loss": 0.092,
+      "step": 3912
+    },
+    {
+      "epoch": 0.7482550913089205,
+      "grad_norm": 6.85171616489598,
+      "learning_rate": 1e-06,
+      "loss": 0.1539,
+      "step": 3913
+    },
+    {
+      "epoch": 0.7484463141791758,
+      "grad_norm": 5.358090616585538,
+      "learning_rate": 1e-06,
+      "loss": 0.6355,
+      "step": 3914
+    },
+    {
+      "epoch": 0.7486375370494311,
+      "grad_norm": 4.858708670214402,
+      "learning_rate": 1e-06,
+      "loss": 0.5292,
+      "step": 3915
+    },
+    {
+      "epoch": 0.7488287599196864,
+      "grad_norm": 5.374295033200339,
+      "learning_rate": 1e-06,
+      "loss": 0.1079,
+      "step": 3916
+    },
+    {
+      "epoch": 0.7490199827899416,
+      "grad_norm": 4.993190897364571,
+      "learning_rate": 1e-06,
+      "loss": 0.3542,
+      "step": 3917
+    },
+    {
+      "epoch": 0.7492112056601969,
+      "grad_norm": 2.7980368135115063,
+      "learning_rate": 1e-06,
+      "loss": 0.1128,
+      "step": 3918
+    },
+    {
+      "epoch": 0.7494024285304522,
+      "grad_norm": 3.0264160309869403,
+      "learning_rate": 1e-06,
+      "loss": 0.2286,
+      "step": 3919
+    },
+    {
+      "epoch": 0.7495936514007076,
+      "grad_norm": 9.831642867651915,
+      "learning_rate": 1e-06,
+      "loss": 0.1857,
+      "step": 3920
+    },
+    {
+      "epoch": 0.7497848742709629,
+      "grad_norm": 2.3625449807057968,
+      "learning_rate": 1e-06,
+      "loss": 0.1252,
+      "step": 3921
+    },
+    {
+      "epoch": 0.749976097141218,
+      "grad_norm": 3.993891343076666,
+      "learning_rate": 1e-06,
+      "loss": 0.0835,
+      "step": 3922
+    },
+    {
+      "epoch": 0.7501673200114733,
+      "grad_norm": 3.6245352348459985,
+      "learning_rate": 1e-06,
+      "loss": 0.1108,
+      "step": 3923
+    },
+    {
+      "epoch": 0.7503585428817287,
+      "grad_norm": 4.027876040842199,
+      "learning_rate": 1e-06,
+      "loss": 0.1197,
+      "step": 3924
+    },
+    {
+      "epoch": 0.750549765751984,
+      "grad_norm": 6.423034414803489,
+      "learning_rate": 1e-06,
+      "loss": 0.1596,
+      "step": 3925
+    },
+    {
+      "epoch": 0.7507409886222393,
+      "grad_norm": 4.438933664753071,
+      "learning_rate": 1e-06,
+      "loss": 0.4673,
+      "step": 3926
+    },
+    {
+      "epoch": 0.7509322114924945,
+      "grad_norm": 4.140668760824178,
+      "learning_rate": 1e-06,
+      "loss": 0.4117,
+      "step": 3927
+    },
+    {
+      "epoch": 0.7511234343627498,
+      "grad_norm": 3.5010283185679247,
+      "learning_rate": 1e-06,
+      "loss": 0.2118,
+      "step": 3928
+    },
+    {
+      "epoch": 0.7513146572330051,
+      "grad_norm": 2.5697948549083596,
+      "learning_rate": 1e-06,
+      "loss": 0.0658,
+      "step": 3929
+    },
+    {
+      "epoch": 0.7515058801032604,
+      "grad_norm": 4.4784849609177,
+      "learning_rate": 1e-06,
+      "loss": 0.4215,
+      "step": 3930
+    },
+    {
+      "epoch": 0.7516971029735157,
+      "grad_norm": 3.9157484615069373,
+      "learning_rate": 1e-06,
+      "loss": 0.4887,
+      "step": 3931
+    },
+    {
+      "epoch": 0.751888325843771,
+      "grad_norm": 3.3244350555727062,
+      "learning_rate": 1e-06,
+      "loss": 0.1955,
+      "step": 3932
+    },
+    {
+      "epoch": 0.7520795487140262,
+      "grad_norm": 4.216708219019393,
+      "learning_rate": 1e-06,
+      "loss": 0.1661,
+      "step": 3933
+    },
+    {
+      "epoch": 0.7522707715842815,
+      "grad_norm": 1.8768198718226292,
+      "learning_rate": 1e-06,
+      "loss": 0.0385,
+      "step": 3934
+    },
+    {
+      "epoch": 0.7524619944545368,
+      "grad_norm": 5.541165016050995,
+      "learning_rate": 1e-06,
+      "loss": 0.3766,
+      "step": 3935
+    },
+    {
+      "epoch": 0.7526532173247921,
+      "grad_norm": 4.22891164427307,
+      "learning_rate": 1e-06,
+      "loss": 0.1596,
+      "step": 3936
+    },
+    {
+      "epoch": 0.7528444401950474,
+      "grad_norm": 2.440666098743169,
+      "learning_rate": 1e-06,
+      "loss": 0.0813,
+      "step": 3937
+    },
+    {
+      "epoch": 0.7530356630653026,
+      "grad_norm": 7.371704140979796,
+      "learning_rate": 1e-06,
+      "loss": 0.2154,
+      "step": 3938
+    },
+    {
+      "epoch": 0.7532268859355579,
+      "grad_norm": 6.1741883396971895,
+      "learning_rate": 1e-06,
+      "loss": 0.571,
+      "step": 3939
+    },
+    {
+      "epoch": 0.7534181088058132,
+      "grad_norm": 4.239020582836203,
+      "learning_rate": 1e-06,
+      "loss": 0.3484,
+      "step": 3940
+    },
+    {
+      "epoch": 0.7536093316760685,
+      "grad_norm": 1.5877379171554096,
+      "learning_rate": 1e-06,
+      "loss": 0.162,
+      "step": 3941
+    },
+    {
+      "epoch": 0.7538005545463238,
+      "grad_norm": 2.360255197207977,
+      "learning_rate": 1e-06,
+      "loss": 0.1181,
+      "step": 3942
+    },
+    {
+      "epoch": 0.753991777416579,
+      "grad_norm": 2.335605967798594,
+      "learning_rate": 1e-06,
+      "loss": 0.096,
+      "step": 3943
+    },
+    {
+      "epoch": 0.7541830002868343,
+      "grad_norm": 5.106725163103992,
+      "learning_rate": 1e-06,
+      "loss": 0.3826,
+      "step": 3944
+    },
+    {
+      "epoch": 0.7543742231570896,
+      "grad_norm": 4.280013525397775,
+      "learning_rate": 1e-06,
+      "loss": 0.1836,
+      "step": 3945
+    },
+    {
+      "epoch": 0.7545654460273449,
+      "grad_norm": 4.416394399252188,
+      "learning_rate": 1e-06,
+      "loss": 0.2044,
+      "step": 3946
+    },
+    {
+      "epoch": 0.7547566688976002,
+      "grad_norm": 4.403078060292515,
+      "learning_rate": 1e-06,
+      "loss": 0.3302,
+      "step": 3947
+    },
+    {
+      "epoch": 0.7549478917678555,
+      "grad_norm": 3.5918394898318233,
+      "learning_rate": 1e-06,
+      "loss": 0.1086,
+      "step": 3948
+    },
+    {
+      "epoch": 0.7551391146381107,
+      "grad_norm": 2.1384899323293127,
+      "learning_rate": 1e-06,
+      "loss": 0.0537,
+      "step": 3949
+    },
+    {
+      "epoch": 0.755330337508366,
+      "grad_norm": 6.101701292835646,
+      "learning_rate": 1e-06,
+      "loss": 0.1033,
+      "step": 3950
+    },
+    {
+      "epoch": 0.7555215603786213,
+      "grad_norm": 4.0469209528490175,
+      "learning_rate": 1e-06,
+      "loss": 0.4927,
+      "step": 3951
+    },
+    {
+      "epoch": 0.7557127832488766,
+      "grad_norm": 3.9862536022018085,
+      "learning_rate": 1e-06,
+      "loss": 0.4655,
+      "step": 3952
+    },
+    {
+      "epoch": 0.7559040061191319,
+      "grad_norm": 3.616036744935352,
+      "learning_rate": 1e-06,
+      "loss": 0.1194,
+      "step": 3953
+    },
+    {
+      "epoch": 0.7560952289893871,
+      "grad_norm": 6.401845916142719,
+      "learning_rate": 1e-06,
+      "loss": 0.1846,
+      "step": 3954
+    },
+    {
+      "epoch": 0.7562864518596424,
+      "grad_norm": 3.088881081023133,
+      "learning_rate": 1e-06,
+      "loss": 0.1605,
+      "step": 3955
+    },
+    {
+      "epoch": 0.7564776747298977,
+      "grad_norm": 2.8523095393473605,
+      "learning_rate": 1e-06,
+      "loss": 0.1083,
+      "step": 3956
+    },
+    {
+      "epoch": 0.756668897600153,
+      "grad_norm": 3.365716245471644,
+      "learning_rate": 1e-06,
+      "loss": 0.1275,
+      "step": 3957
+    },
+    {
+      "epoch": 0.7568601204704083,
+      "grad_norm": 2.8640658564321577,
+      "learning_rate": 1e-06,
+      "loss": 0.0977,
+      "step": 3958
+    },
+    {
+      "epoch": 0.7570513433406636,
+      "grad_norm": 4.298001006582932,
+      "learning_rate": 1e-06,
+      "loss": 0.2105,
+      "step": 3959
+    },
+    {
+      "epoch": 0.7572425662109188,
+      "grad_norm": 3.7400017011098003,
+      "learning_rate": 1e-06,
+      "loss": 0.073,
+      "step": 3960
+    },
+    {
+      "epoch": 0.7574337890811741,
+      "grad_norm": 3.1368154318459345,
+      "learning_rate": 1e-06,
+      "loss": 0.0443,
+      "step": 3961
+    },
+    {
+      "epoch": 0.7576250119514294,
+      "grad_norm": 7.927345809567831,
+      "learning_rate": 1e-06,
+      "loss": 0.1182,
+      "step": 3962
+    },
+    {
+      "epoch": 0.7578162348216847,
+      "grad_norm": 6.752092107876784,
+      "learning_rate": 1e-06,
+      "loss": 0.3041,
+      "step": 3963
+    },
+    {
+      "epoch": 0.75800745769194,
+      "grad_norm": 4.583797754975726,
+      "learning_rate": 1e-06,
+      "loss": 0.3701,
+      "step": 3964
+    },
+    {
+      "epoch": 0.7581986805621952,
+      "grad_norm": 5.556148374505623,
+      "learning_rate": 1e-06,
+      "loss": 0.5907,
+      "step": 3965
+    },
+    {
+      "epoch": 0.7583899034324505,
+      "grad_norm": 1.987641596998384,
+      "learning_rate": 1e-06,
+      "loss": 0.1896,
+      "step": 3966
+    },
+    {
+      "epoch": 0.7585811263027058,
+      "grad_norm": 4.469464865106327,
+      "learning_rate": 1e-06,
+      "loss": 0.2959,
+      "step": 3967
+    },
+    {
+      "epoch": 0.7587723491729611,
+      "grad_norm": 3.3436069636188765,
+      "learning_rate": 1e-06,
+      "loss": 0.2297,
+      "step": 3968
+    },
+    {
+      "epoch": 0.7589635720432164,
+      "grad_norm": 2.6805273580523004,
+      "learning_rate": 1e-06,
+      "loss": 0.1419,
+      "step": 3969
+    },
+    {
+      "epoch": 0.7591547949134716,
+      "grad_norm": 3.7800163589607956,
+      "learning_rate": 1e-06,
+      "loss": 0.2329,
+      "step": 3970
+    },
+    {
+      "epoch": 0.7593460177837269,
+      "grad_norm": 2.835455903494897,
+      "learning_rate": 1e-06,
+      "loss": 0.2086,
+      "step": 3971
+    },
+    {
+      "epoch": 0.7595372406539822,
+      "grad_norm": 3.919414097732828,
+      "learning_rate": 1e-06,
+      "loss": 0.1233,
+      "step": 3972
+    },
+    {
+      "epoch": 0.7597284635242375,
+      "grad_norm": 3.2384966696951327,
+      "learning_rate": 1e-06,
+      "loss": 0.2097,
+      "step": 3973
+    },
+    {
+      "epoch": 0.7599196863944928,
+      "grad_norm": 3.3342136333284556,
+      "learning_rate": 1e-06,
+      "loss": 0.0643,
+      "step": 3974
+    },
+    {
+      "epoch": 0.7601109092647481,
+      "grad_norm": 3.1902939762037685,
+      "learning_rate": 1e-06,
+      "loss": 0.0854,
+      "step": 3975
+    },
+    {
+      "epoch": 0.7603021321350033,
+      "grad_norm": 5.165981103677191,
+      "learning_rate": 1e-06,
+      "loss": 0.1403,
+      "step": 3976
+    },
+    {
+      "epoch": 0.7604933550052586,
+      "grad_norm": 4.983642332943844,
+      "learning_rate": 1e-06,
+      "loss": 0.4221,
+      "step": 3977
+    },
+    {
+      "epoch": 0.7606845778755139,
+      "grad_norm": 5.31932365649973,
+      "learning_rate": 1e-06,
+      "loss": 0.5743,
+      "step": 3978
+    },
+    {
+      "epoch": 0.7608758007457692,
+      "grad_norm": 4.079912868018154,
+      "learning_rate": 1e-06,
+      "loss": 0.4267,
+      "step": 3979
+    },
+    {
+      "epoch": 0.7610670236160245,
+      "grad_norm": 2.3068570545028724,
+      "learning_rate": 1e-06,
+      "loss": 0.3533,
+      "step": 3980
+    },
+    {
+      "epoch": 0.7612582464862797,
+      "grad_norm": 3.4294427945162425,
+      "learning_rate": 1e-06,
+      "loss": 0.166,
+      "step": 3981
+    },
+    {
+      "epoch": 0.761449469356535,
+      "grad_norm": 4.473589644008324,
+      "learning_rate": 1e-06,
+      "loss": 0.1379,
+      "step": 3982
+    },
+    {
+      "epoch": 0.7616406922267903,
+      "grad_norm": 2.853643958869806,
+      "learning_rate": 1e-06,
+      "loss": 0.0839,
+      "step": 3983
+    },
+    {
+      "epoch": 0.7618319150970456,
+      "grad_norm": 2.5309564867683996,
+      "learning_rate": 1e-06,
+      "loss": 0.1385,
+      "step": 3984
+    },
+    {
+      "epoch": 0.7620231379673009,
+      "grad_norm": 3.6616536174733225,
+      "learning_rate": 1e-06,
+      "loss": 0.2128,
+      "step": 3985
+    },
+    {
+      "epoch": 0.7622143608375562,
+      "grad_norm": 4.913135726212392,
+      "learning_rate": 1e-06,
+      "loss": 0.1027,
+      "step": 3986
+    },
+    {
+      "epoch": 0.7624055837078114,
+      "grad_norm": 3.6338139948083126,
+      "learning_rate": 1e-06,
+      "loss": 0.0961,
+      "step": 3987
+    },
+    {
+      "epoch": 0.7625968065780667,
+      "grad_norm": 4.596368575321195,
+      "learning_rate": 1e-06,
+      "loss": 0.1924,
+      "step": 3988
+    },
+    {
+      "epoch": 0.762788029448322,
+      "grad_norm": 4.398383294376247,
+      "learning_rate": 1e-06,
+      "loss": 0.4758,
+      "step": 3989
+    },
+    {
+      "epoch": 0.7629792523185773,
+      "grad_norm": 6.071374434542319,
+      "learning_rate": 1e-06,
+      "loss": 0.6171,
+      "step": 3990
+    },
+    {
+      "epoch": 0.7631704751888326,
+      "grad_norm": 3.4912775889616734,
+      "learning_rate": 1e-06,
+      "loss": 0.192,
+      "step": 3991
+    },
+    {
+      "epoch": 0.7633616980590878,
+      "grad_norm": 3.4863271675856526,
+      "learning_rate": 1e-06,
+      "loss": 0.3949,
+      "step": 3992
+    },
+    {
+      "epoch": 0.7635529209293431,
+      "grad_norm": 2.159092674072128,
+      "learning_rate": 1e-06,
+      "loss": 0.1555,
+      "step": 3993
+    },
+    {
+      "epoch": 0.7637441437995984,
+      "grad_norm": 2.381616210812365,
+      "learning_rate": 1e-06,
+      "loss": 0.1444,
+      "step": 3994
+    },
+    {
+      "epoch": 0.7639353666698537,
+      "grad_norm": 3.6773033760663405,
+      "learning_rate": 1e-06,
+      "loss": 0.2426,
+      "step": 3995
+    },
+    {
+      "epoch": 0.764126589540109,
+      "grad_norm": 3.381283456351275,
+      "learning_rate": 1e-06,
+      "loss": 0.1036,
+      "step": 3996
+    },
+    {
+      "epoch": 0.7643178124103642,
+      "grad_norm": 3.8651184095476583,
+      "learning_rate": 1e-06,
+      "loss": 0.1651,
+      "step": 3997
+    },
+    {
+      "epoch": 0.7645090352806195,
+      "grad_norm": 3.0570271695482925,
+      "learning_rate": 1e-06,
+      "loss": 0.1559,
+      "step": 3998
+    },
+    {
+      "epoch": 0.7647002581508748,
+      "grad_norm": 5.030097592508632,
+      "learning_rate": 1e-06,
+      "loss": 0.2049,
+      "step": 3999
+    },
+    {
+      "epoch": 0.7648914810211301,
+      "grad_norm": 13.036943614348838,
+      "learning_rate": 1e-06,
+      "loss": 0.356,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7648914810211301,
+      "eval_runtime": 719.4886,
+      "eval_samples_per_second": 2.132,
+      "eval_steps_per_second": 0.534,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7650827038913854,
+      "grad_norm": 3.070391549604594,
+      "learning_rate": 1e-06,
+      "loss": 0.4446,
+      "step": 4001
+    },
+    {
+      "epoch": 0.7652739267616407,
+      "grad_norm": 4.158847312786241,
+      "learning_rate": 1e-06,
+      "loss": 0.464,
+      "step": 4002
+    },
+    {
+      "epoch": 0.7654651496318959,
+      "grad_norm": 3.1187370101260665,
+      "learning_rate": 1e-06,
+      "loss": 0.3173,
+      "step": 4003
+    },
+    {
+      "epoch": 0.7656563725021512,
+      "grad_norm": 2.565788577925182,
+      "learning_rate": 1e-06,
+      "loss": 0.1629,
+      "step": 4004
+    },
+    {
+      "epoch": 0.7658475953724065,
+      "grad_norm": 3.355968078479653,
+      "learning_rate": 1e-06,
+      "loss": 0.3918,
+      "step": 4005
+    },
+    {
+      "epoch": 0.7660388182426618,
+      "grad_norm": 5.486461275195402,
+      "learning_rate": 1e-06,
+      "loss": 0.4689,
+      "step": 4006
+    },
+    {
+      "epoch": 0.7662300411129171,
+      "grad_norm": 3.68044018994471,
+      "learning_rate": 1e-06,
+      "loss": 0.1995,
+      "step": 4007
+    },
+    {
+      "epoch": 0.7664212639831723,
+      "grad_norm": 2.973650130799323,
+      "learning_rate": 1e-06,
+      "loss": 0.2526,
+      "step": 4008
+    },
+    {
+      "epoch": 0.7666124868534276,
+      "grad_norm": 2.0252513635299043,
+      "learning_rate": 1e-06,
+      "loss": 0.1607,
+      "step": 4009
+    },
+    {
+      "epoch": 0.766803709723683,
+      "grad_norm": 4.645218447834352,
+      "learning_rate": 1e-06,
+      "loss": 0.0927,
+      "step": 4010
+    },
+    {
+      "epoch": 0.7669949325939383,
+      "grad_norm": 2.810538731312038,
+      "learning_rate": 1e-06,
+      "loss": 0.1818,
+      "step": 4011
+    },
+    {
+      "epoch": 0.7671861554641936,
+      "grad_norm": 2.4854928626643096,
+      "learning_rate": 1e-06,
+      "loss": 0.1088,
+      "step": 4012
+    },
+    {
+      "epoch": 0.7673773783344487,
+      "grad_norm": 6.088019077486683,
+      "learning_rate": 1e-06,
+      "loss": 0.2576,
+      "step": 4013
+    },
+    {
+      "epoch": 0.767568601204704,
+      "grad_norm": 3.4923662314878183,
+      "learning_rate": 1e-06,
+      "loss": 0.4229,
+      "step": 4014
+    },
+    {
+      "epoch": 0.7677598240749594,
+      "grad_norm": 3.827896298174766,
+      "learning_rate": 1e-06,
+      "loss": 0.259,
+      "step": 4015
+    },
+    {
+      "epoch": 0.7679510469452147,
+      "grad_norm": 6.5472682934895525,
+      "learning_rate": 1e-06,
+      "loss": 0.8165,
+      "step": 4016
+    },
+    {
+      "epoch": 0.76814226981547,
+      "grad_norm": 2.755866382571043,
+      "learning_rate": 1e-06,
+      "loss": 0.1697,
+      "step": 4017
+    },
+    {
+      "epoch": 0.7683334926857253,
+      "grad_norm": 3.323991671380765,
+      "learning_rate": 1e-06,
+      "loss": 0.2198,
+      "step": 4018
+    },
+    {
+      "epoch": 0.7685247155559805,
+      "grad_norm": 1.5006465313251316,
+      "learning_rate": 1e-06,
+      "loss": 0.0544,
+      "step": 4019
+    },
+    {
+      "epoch": 0.7687159384262358,
+      "grad_norm": 4.570772588019623,
+      "learning_rate": 1e-06,
+      "loss": 0.3775,
+      "step": 4020
+    },
+    {
+      "epoch": 0.7689071612964911,
+      "grad_norm": 2.5302264180851726,
+      "learning_rate": 1e-06,
+      "loss": 0.1115,
+      "step": 4021
+    },
+    {
+      "epoch": 0.7690983841667464,
+      "grad_norm": 3.622311614662082,
+      "learning_rate": 1e-06,
+      "loss": 0.0777,
+      "step": 4022
+    },
+    {
+      "epoch": 0.7692896070370017,
+      "grad_norm": 3.163747512774204,
+      "learning_rate": 1e-06,
+      "loss": 0.1562,
+      "step": 4023
+    },
+    {
+      "epoch": 0.7694808299072569,
+      "grad_norm": 2.770716404180518,
+      "learning_rate": 1e-06,
+      "loss": 0.0423,
+      "step": 4024
+    },
+    {
+      "epoch": 0.7696720527775122,
+      "grad_norm": 5.863314756212557,
+      "learning_rate": 1e-06,
+      "loss": 0.1281,
+      "step": 4025
+    },
+    {
+      "epoch": 0.7698632756477675,
+      "grad_norm": 4.86408420046908,
+      "learning_rate": 1e-06,
+      "loss": 0.5083,
+      "step": 4026
+    },
+    {
+      "epoch": 0.7700544985180228,
+      "grad_norm": 5.983411746277226,
+      "learning_rate": 1e-06,
+      "loss": 0.7271,
+      "step": 4027
+    },
+    {
+      "epoch": 0.7702457213882781,
+      "grad_norm": 3.6611345056594513,
+      "learning_rate": 1e-06,
+      "loss": 0.2173,
+      "step": 4028
+    },
+    {
+      "epoch": 0.7704369442585334,
+      "grad_norm": 3.1920213235032917,
+      "learning_rate": 1e-06,
+      "loss": 0.2751,
+      "step": 4029
+    },
+    {
+      "epoch": 0.7706281671287886,
+      "grad_norm": 3.2753321006063945,
+      "learning_rate": 1e-06,
+      "loss": 0.2839,
+      "step": 4030
+    },
+    {
+      "epoch": 0.7708193899990439,
+      "grad_norm": 3.1176630484396415,
+      "learning_rate": 1e-06,
+      "loss": 0.2292,
+      "step": 4031
+    },
+    {
+      "epoch": 0.7710106128692992,
+      "grad_norm": 3.3870591781217274,
+      "learning_rate": 1e-06,
+      "loss": 0.247,
+      "step": 4032
+    },
+    {
+      "epoch": 0.7712018357395545,
+      "grad_norm": 2.8069872764839916,
+      "learning_rate": 1e-06,
+      "loss": 0.2026,
+      "step": 4033
+    },
+    {
+      "epoch": 0.7713930586098098,
+      "grad_norm": 2.3021358524275293,
+      "learning_rate": 1e-06,
+      "loss": 0.1535,
+      "step": 4034
+    },
+    {
+      "epoch": 0.771584281480065,
+      "grad_norm": 5.062564425588323,
+      "learning_rate": 1e-06,
+      "loss": 0.2458,
+      "step": 4035
+    },
+    {
+      "epoch": 0.7717755043503203,
+      "grad_norm": 7.214398017667463,
+      "learning_rate": 1e-06,
+      "loss": 0.1844,
+      "step": 4036
+    },
+    {
+      "epoch": 0.7719667272205756,
+      "grad_norm": 3.075107547771116,
+      "learning_rate": 1e-06,
+      "loss": 0.0489,
+      "step": 4037
+    },
+    {
+      "epoch": 0.7721579500908309,
+      "grad_norm": 5.631961160309122,
+      "learning_rate": 1e-06,
+      "loss": 0.3761,
+      "step": 4038
+    },
+    {
+      "epoch": 0.7723491729610862,
+      "grad_norm": 5.781497764433234,
+      "learning_rate": 1e-06,
+      "loss": 0.6925,
+      "step": 4039
+    },
+    {
+      "epoch": 0.7725403958313414,
+      "grad_norm": 4.5522446010954996,
+      "learning_rate": 1e-06,
+      "loss": 0.5236,
+      "step": 4040
+    },
+    {
+      "epoch": 0.7727316187015967,
+      "grad_norm": 3.687362151074764,
+      "learning_rate": 1e-06,
+      "loss": 0.2348,
+      "step": 4041
+    },
+    {
+      "epoch": 0.772922841571852,
+      "grad_norm": 2.7584085225655137,
+      "learning_rate": 1e-06,
+      "loss": 0.2487,
+      "step": 4042
+    },
+    {
+      "epoch": 0.7731140644421073,
+      "grad_norm": 3.7957569878134545,
+      "learning_rate": 1e-06,
+      "loss": 0.2761,
+      "step": 4043
+    },
+    {
+      "epoch": 0.7733052873123626,
+      "grad_norm": 4.958617813967921,
+      "learning_rate": 1e-06,
+      "loss": 0.2424,
+      "step": 4044
+    },
+    {
+      "epoch": 0.7734965101826179,
+      "grad_norm": 5.397150220187277,
+      "learning_rate": 1e-06,
+      "loss": 0.1221,
+      "step": 4045
+    },
+    {
+      "epoch": 0.7736877330528731,
+      "grad_norm": 2.21067555092306,
+      "learning_rate": 1e-06,
+      "loss": 0.1235,
+      "step": 4046
+    },
+    {
+      "epoch": 0.7738789559231284,
+      "grad_norm": 5.620574906018949,
+      "learning_rate": 1e-06,
+      "loss": 0.2865,
+      "step": 4047
+    },
+    {
+      "epoch": 0.7740701787933837,
+      "grad_norm": 2.9929183983606378,
+      "learning_rate": 1e-06,
+      "loss": 0.1014,
+      "step": 4048
+    },
+    {
+      "epoch": 0.774261401663639,
+      "grad_norm": 3.118101516607677,
+      "learning_rate": 1e-06,
+      "loss": 0.0636,
+      "step": 4049
+    },
+    {
+      "epoch": 0.7744526245338943,
+      "grad_norm": 8.709021825510979,
+      "learning_rate": 1e-06,
+      "loss": 0.1353,
+      "step": 4050
+    },
+    {
+      "epoch": 0.7746438474041495,
+      "grad_norm": 3.9482194822962033,
+      "learning_rate": 1e-06,
+      "loss": 0.1608,
+      "step": 4051
+    },
+    {
+      "epoch": 0.7748350702744048,
+      "grad_norm": 5.720092282384449,
+      "learning_rate": 1e-06,
+      "loss": 0.5042,
+      "step": 4052
+    },
+    {
+      "epoch": 0.7750262931446601,
+      "grad_norm": 5.8493569958076845,
+      "learning_rate": 1e-06,
+      "loss": 0.7138,
+      "step": 4053
+    },
+    {
+      "epoch": 0.7752175160149154,
+      "grad_norm": 3.689608730671752,
+      "learning_rate": 1e-06,
+      "loss": 0.3434,
+      "step": 4054
+    },
+    {
+      "epoch": 0.7754087388851707,
+      "grad_norm": 4.417024250070177,
+      "learning_rate": 1e-06,
+      "loss": 0.3933,
+      "step": 4055
+    },
+    {
+      "epoch": 0.775599961755426,
+      "grad_norm": 1.7614275571126077,
+      "learning_rate": 1e-06,
+      "loss": 0.0626,
+      "step": 4056
+    },
+    {
+      "epoch": 0.7757911846256812,
+      "grad_norm": 2.177443685931191,
+      "learning_rate": 1e-06,
+      "loss": 0.092,
+      "step": 4057
+    },
+    {
+      "epoch": 0.7759824074959365,
+      "grad_norm": 4.449407156554658,
+      "learning_rate": 1e-06,
+      "loss": 0.0826,
+      "step": 4058
+    },
+    {
+      "epoch": 0.7761736303661918,
+      "grad_norm": 4.524506068801983,
+      "learning_rate": 1e-06,
+      "loss": 0.0664,
+      "step": 4059
+    },
+    {
+      "epoch": 0.7763648532364471,
+      "grad_norm": 2.2000005765393977,
+      "learning_rate": 1e-06,
+      "loss": 0.0639,
+      "step": 4060
+    },
+    {
+      "epoch": 0.7765560761067024,
+      "grad_norm": 2.6059003526807785,
+      "learning_rate": 1e-06,
+      "loss": 0.0606,
+      "step": 4061
+    },
+    {
+      "epoch": 0.7767472989769576,
+      "grad_norm": 4.779463091268296,
+      "learning_rate": 1e-06,
+      "loss": 0.0605,
+      "step": 4062
+    },
+    {
+      "epoch": 0.7769385218472129,
+      "grad_norm": 7.148337174842439,
+      "learning_rate": 1e-06,
+      "loss": 0.1794,
+      "step": 4063
+    },
+    {
+      "epoch": 0.7771297447174682,
+      "grad_norm": 3.6786922720690525,
+      "learning_rate": 1e-06,
+      "loss": 0.1845,
+      "step": 4064
+    },
+    {
+      "epoch": 0.7773209675877235,
+      "grad_norm": 3.960327221808368,
+      "learning_rate": 1e-06,
+      "loss": 0.3532,
+      "step": 4065
+    },
+    {
+      "epoch": 0.7775121904579788,
+      "grad_norm": 3.255857617538518,
+      "learning_rate": 1e-06,
+      "loss": 0.2789,
+      "step": 4066
+    },
+    {
+      "epoch": 0.777703413328234,
+      "grad_norm": 3.225296621949762,
+      "learning_rate": 1e-06,
+      "loss": 0.2371,
+      "step": 4067
+    },
+    {
+      "epoch": 0.7778946361984893,
+      "grad_norm": 2.7363588820150877,
+      "learning_rate": 1e-06,
+      "loss": 0.2663,
+      "step": 4068
+    },
+    {
+      "epoch": 0.7780858590687446,
+      "grad_norm": 4.385994529557197,
+      "learning_rate": 1e-06,
+      "loss": 0.353,
+      "step": 4069
+    },
+    {
+      "epoch": 0.7782770819389999,
+      "grad_norm": 3.0259882297929885,
+      "learning_rate": 1e-06,
+      "loss": 0.1429,
+      "step": 4070
+    },
+    {
+      "epoch": 0.7784683048092552,
+      "grad_norm": 3.7364950506687244,
+      "learning_rate": 1e-06,
+      "loss": 0.0976,
+      "step": 4071
+    },
+    {
+      "epoch": 0.7786595276795105,
+      "grad_norm": 4.164006617530746,
+      "learning_rate": 1e-06,
+      "loss": 0.2052,
+      "step": 4072
+    },
+    {
+      "epoch": 0.7788507505497657,
+      "grad_norm": 3.2675730174515523,
+      "learning_rate": 1e-06,
+      "loss": 0.0752,
+      "step": 4073
+    },
+    {
+      "epoch": 0.779041973420021,
+      "grad_norm": 4.311004158104732,
+      "learning_rate": 1e-06,
+      "loss": 0.1102,
+      "step": 4074
+    },
+    {
+      "epoch": 0.7792331962902763,
+      "grad_norm": 4.320110295618457,
+      "learning_rate": 1e-06,
+      "loss": 0.052,
+      "step": 4075
+    },
+    {
+      "epoch": 0.7794244191605316,
+      "grad_norm": 7.897631867295925,
+      "learning_rate": 1e-06,
+      "loss": 0.2689,
+      "step": 4076
+    },
+    {
+      "epoch": 0.7796156420307869,
+      "grad_norm": 6.2423919147732905,
+      "learning_rate": 1e-06,
+      "loss": 0.639,
+      "step": 4077
+    },
+    {
+      "epoch": 0.7798068649010421,
+      "grad_norm": 2.8255864336452787,
+      "learning_rate": 1e-06,
+      "loss": 0.2039,
+      "step": 4078
+    },
+    {
+      "epoch": 0.7799980877712974,
+      "grad_norm": 3.700933008865251,
+      "learning_rate": 1e-06,
+      "loss": 0.2236,
+      "step": 4079
+    },
+    {
+      "epoch": 0.7801893106415527,
+      "grad_norm": 3.1519599690388422,
+      "learning_rate": 1e-06,
+      "loss": 0.1624,
+      "step": 4080
+    },
+    {
+      "epoch": 0.780380533511808,
+      "grad_norm": 2.7222362396848148,
+      "learning_rate": 1e-06,
+      "loss": 0.102,
+      "step": 4081
+    },
+    {
+      "epoch": 0.7805717563820633,
+      "grad_norm": 3.3242642212390194,
+      "learning_rate": 1e-06,
+      "loss": 0.2081,
+      "step": 4082
+    },
+    {
+      "epoch": 0.7807629792523185,
+      "grad_norm": 1.7714525449137004,
+      "learning_rate": 1e-06,
+      "loss": 0.0809,
+      "step": 4083
+    },
+    {
+      "epoch": 0.7809542021225738,
+      "grad_norm": 5.419180692564339,
+      "learning_rate": 1e-06,
+      "loss": 0.5617,
+      "step": 4084
+    },
+    {
+      "epoch": 0.7811454249928291,
+      "grad_norm": 2.0724692572693715,
+      "learning_rate": 1e-06,
+      "loss": 0.1644,
+      "step": 4085
+    },
+    {
+      "epoch": 0.7813366478630844,
+      "grad_norm": 2.6694569493359337,
+      "learning_rate": 1e-06,
+      "loss": 0.0795,
+      "step": 4086
+    },
+    {
+      "epoch": 0.7815278707333397,
+      "grad_norm": 2.5988904272673548,
+      "learning_rate": 1e-06,
+      "loss": 0.0341,
+      "step": 4087
+    },
+    {
+      "epoch": 0.781719093603595,
+      "grad_norm": 4.847580312031127,
+      "learning_rate": 1e-06,
+      "loss": 0.1375,
+      "step": 4088
+    },
+    {
+      "epoch": 0.7819103164738502,
+      "grad_norm": 4.87679453367743,
+      "learning_rate": 1e-06,
+      "loss": 0.9582,
+      "step": 4089
+    },
+    {
+      "epoch": 0.7821015393441055,
+      "grad_norm": 8.15599556595484,
+      "learning_rate": 1e-06,
+      "loss": 0.3661,
+      "step": 4090
+    },
+    {
+      "epoch": 0.7822927622143608,
+      "grad_norm": 3.5682941633340692,
+      "learning_rate": 1e-06,
+      "loss": 0.3445,
+      "step": 4091
+    },
+    {
+      "epoch": 0.7824839850846161,
+      "grad_norm": 4.8818145464561855,
+      "learning_rate": 1e-06,
+      "loss": 0.4113,
+      "step": 4092
+    },
+    {
+      "epoch": 0.7826752079548714,
+      "grad_norm": 3.3651235676221773,
+      "learning_rate": 1e-06,
+      "loss": 0.2707,
+      "step": 4093
+    },
+    {
+      "epoch": 0.7828664308251266,
+      "grad_norm": 4.134743252109405,
+      "learning_rate": 1e-06,
+      "loss": 0.1398,
+      "step": 4094
+    },
+    {
+      "epoch": 0.7830576536953819,
+      "grad_norm": 3.364850784553028,
+      "learning_rate": 1e-06,
+      "loss": 0.2491,
+      "step": 4095
+    },
+    {
+      "epoch": 0.7832488765656372,
+      "grad_norm": 3.44712868221269,
+      "learning_rate": 1e-06,
+      "loss": 0.1312,
+      "step": 4096
+    },
+    {
+      "epoch": 0.7834400994358925,
+      "grad_norm": 3.2339968252927243,
+      "learning_rate": 1e-06,
+      "loss": 0.1136,
+      "step": 4097
+    },
+    {
+      "epoch": 0.7836313223061478,
+      "grad_norm": 3.7200492063980097,
+      "learning_rate": 1e-06,
+      "loss": 0.1182,
+      "step": 4098
+    },
+    {
+      "epoch": 0.7838225451764032,
+      "grad_norm": 2.7200264400711767,
+      "learning_rate": 1e-06,
+      "loss": 0.0369,
+      "step": 4099
+    },
+    {
+      "epoch": 0.7840137680466583,
+      "grad_norm": 4.496073811451229,
+      "learning_rate": 1e-06,
+      "loss": 0.075,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7842049909169136,
+      "grad_norm": 5.795695634746963,
+      "learning_rate": 1e-06,
+      "loss": 0.6525,
+      "step": 4101
+    },
+    {
+      "epoch": 0.784396213787169,
+      "grad_norm": 2.8485732204728422,
+      "learning_rate": 1e-06,
+      "loss": 0.2259,
+      "step": 4102
+    },
+    {
+      "epoch": 0.7845874366574243,
+      "grad_norm": 2.8489136817310894,
+      "learning_rate": 1e-06,
+      "loss": 0.239,
+      "step": 4103
+    },
+    {
+      "epoch": 0.7847786595276796,
+      "grad_norm": 3.950599798705413,
+      "learning_rate": 1e-06,
+      "loss": 0.2081,
+      "step": 4104
+    },
+    {
+      "epoch": 0.7849698823979347,
+      "grad_norm": 3.5174071817413357,
+      "learning_rate": 1e-06,
+      "loss": 0.147,
+      "step": 4105
+    },
+    {
+      "epoch": 0.78516110526819,
+      "grad_norm": 4.89820737161283,
+      "learning_rate": 1e-06,
+      "loss": 0.1402,
+      "step": 4106
+    },
+    {
+      "epoch": 0.7853523281384454,
+      "grad_norm": 2.543508633496386,
+      "learning_rate": 1e-06,
+      "loss": 0.0946,
+      "step": 4107
+    },
+    {
+      "epoch": 0.7855435510087007,
+      "grad_norm": 2.4606498429040133,
+      "learning_rate": 1e-06,
+      "loss": 0.1383,
+      "step": 4108
+    },
+    {
+      "epoch": 0.785734773878956,
+      "grad_norm": 1.9117480551309571,
+      "learning_rate": 1e-06,
+      "loss": 0.0459,
+      "step": 4109
+    },
+    {
+      "epoch": 0.7859259967492112,
+      "grad_norm": 4.037330713650027,
+      "learning_rate": 1e-06,
+      "loss": 0.1211,
+      "step": 4110
+    },
+    {
+      "epoch": 0.7861172196194665,
+      "grad_norm": 3.2312739889704303,
+      "learning_rate": 1e-06,
+      "loss": 0.108,
+      "step": 4111
+    },
+    {
+      "epoch": 0.7863084424897218,
+      "grad_norm": 4.792493776447711,
+      "learning_rate": 1e-06,
+      "loss": 0.1044,
+      "step": 4112
+    },
+    {
+      "epoch": 0.7864996653599771,
+      "grad_norm": 7.057964117428634,
+      "learning_rate": 1e-06,
+      "loss": 0.1973,
+      "step": 4113
+    },
+    {
+      "epoch": 0.7866908882302324,
+      "grad_norm": 8.15440798015625,
+      "learning_rate": 1e-06,
+      "loss": 0.7052,
+      "step": 4114
+    },
+    {
+      "epoch": 0.7868821111004877,
+      "grad_norm": 3.865865955537212,
+      "learning_rate": 1e-06,
+      "loss": 0.3045,
+      "step": 4115
+    },
+    {
+      "epoch": 0.7870733339707429,
+      "grad_norm": 3.1177043438516403,
+      "learning_rate": 1e-06,
+      "loss": 0.2474,
+      "step": 4116
+    },
+    {
+      "epoch": 0.7872645568409982,
+      "grad_norm": 4.324450092967588,
+      "learning_rate": 1e-06,
+      "loss": 0.5683,
+      "step": 4117
+    },
+    {
+      "epoch": 0.7874557797112535,
+      "grad_norm": 4.459158681369153,
+      "learning_rate": 1e-06,
+      "loss": 0.2612,
+      "step": 4118
+    },
+    {
+      "epoch": 0.7876470025815088,
+      "grad_norm": 2.4705162972841292,
+      "learning_rate": 1e-06,
+      "loss": 0.083,
+      "step": 4119
+    },
+    {
+      "epoch": 0.7878382254517641,
+      "grad_norm": 2.7399763881751915,
+      "learning_rate": 1e-06,
+      "loss": 0.1201,
+      "step": 4120
+    },
+    {
+      "epoch": 0.7880294483220193,
+      "grad_norm": 8.035226987565393,
+      "learning_rate": 1e-06,
+      "loss": 0.1123,
+      "step": 4121
+    },
+    {
+      "epoch": 0.7882206711922746,
+      "grad_norm": 1.9856291885019612,
+      "learning_rate": 1e-06,
+      "loss": 0.0439,
+      "step": 4122
+    },
+    {
+      "epoch": 0.7884118940625299,
+      "grad_norm": 3.0254768374074468,
+      "learning_rate": 1e-06,
+      "loss": 0.0615,
+      "step": 4123
+    },
+    {
+      "epoch": 0.7886031169327852,
+      "grad_norm": 4.088787536355497,
+      "learning_rate": 1e-06,
+      "loss": 0.112,
+      "step": 4124
+    },
+    {
+      "epoch": 0.7887943398030405,
+      "grad_norm": 3.394558080077729,
+      "learning_rate": 1e-06,
+      "loss": 0.0833,
+      "step": 4125
+    },
+    {
+      "epoch": 0.7889855626732958,
+      "grad_norm": 5.226520532459497,
+      "learning_rate": 1e-06,
+      "loss": 0.4389,
+      "step": 4126
+    },
+    {
+      "epoch": 0.789176785543551,
+      "grad_norm": 4.312499336574337,
+      "learning_rate": 1e-06,
+      "loss": 0.3849,
+      "step": 4127
+    },
+    {
+      "epoch": 0.7893680084138063,
+      "grad_norm": 4.29151565631747,
+      "learning_rate": 1e-06,
+      "loss": 0.4521,
+      "step": 4128
+    },
+    {
+      "epoch": 0.7895592312840616,
+      "grad_norm": 3.1330523696355455,
+      "learning_rate": 1e-06,
+      "loss": 0.1806,
+      "step": 4129
+    },
+    {
+      "epoch": 0.7897504541543169,
+      "grad_norm": 4.767954772798721,
+      "learning_rate": 1e-06,
+      "loss": 0.3644,
+      "step": 4130
+    },
+    {
+      "epoch": 0.7899416770245722,
+      "grad_norm": 2.6364365264179623,
+      "learning_rate": 1e-06,
+      "loss": 0.1209,
+      "step": 4131
+    },
+    {
+      "epoch": 0.7901328998948274,
+      "grad_norm": 4.241275583960745,
+      "learning_rate": 1e-06,
+      "loss": 0.4508,
+      "step": 4132
+    },
+    {
+      "epoch": 0.7903241227650827,
+      "grad_norm": 3.099412764026683,
+      "learning_rate": 1e-06,
+      "loss": 0.1919,
+      "step": 4133
+    },
+    {
+      "epoch": 0.790515345635338,
+      "grad_norm": 12.965236252155945,
+      "learning_rate": 1e-06,
+      "loss": 0.3709,
+      "step": 4134
+    },
+    {
+      "epoch": 0.7907065685055933,
+      "grad_norm": 4.551869169852438,
+      "learning_rate": 1e-06,
+      "loss": 0.1674,
+      "step": 4135
+    },
+    {
+      "epoch": 0.7908977913758486,
+      "grad_norm": 3.600658112876755,
+      "learning_rate": 1e-06,
+      "loss": 0.0954,
+      "step": 4136
+    },
+    {
+      "epoch": 0.7910890142461038,
+      "grad_norm": 3.9082266727254744,
+      "learning_rate": 1e-06,
+      "loss": 0.1791,
+      "step": 4137
+    },
+    {
+      "epoch": 0.7912802371163591,
+      "grad_norm": 4.691775393990664,
+      "learning_rate": 1e-06,
+      "loss": 0.179,
+      "step": 4138
+    },
+    {
+      "epoch": 0.7914714599866144,
+      "grad_norm": 5.5325171102266975,
+      "learning_rate": 1e-06,
+      "loss": 0.2083,
+      "step": 4139
+    },
+    {
+      "epoch": 0.7916626828568697,
+      "grad_norm": 8.273426895206848,
+      "learning_rate": 1e-06,
+      "loss": 0.3575,
+      "step": 4140
+    },
+    {
+      "epoch": 0.791853905727125,
+      "grad_norm": 3.8506489689790997,
+      "learning_rate": 1e-06,
+      "loss": 0.1254,
+      "step": 4141
+    },
+    {
+      "epoch": 0.7920451285973803,
+      "grad_norm": 4.339579384457529,
+      "learning_rate": 1e-06,
+      "loss": 0.2043,
+      "step": 4142
+    },
+    {
+      "epoch": 0.7922363514676355,
+      "grad_norm": 2.322094371182351,
+      "learning_rate": 1e-06,
+      "loss": 0.2544,
+      "step": 4143
+    },
+    {
+      "epoch": 0.7924275743378908,
+      "grad_norm": 2.504793149893796,
+      "learning_rate": 1e-06,
+      "loss": 0.1341,
+      "step": 4144
+    },
+    {
+      "epoch": 0.7926187972081461,
+      "grad_norm": 2.784081378701057,
+      "learning_rate": 1e-06,
+      "loss": 0.1573,
+      "step": 4145
+    },
+    {
+      "epoch": 0.7928100200784014,
+      "grad_norm": 2.9279856710760823,
+      "learning_rate": 1e-06,
+      "loss": 0.1314,
+      "step": 4146
+    },
+    {
+      "epoch": 0.7930012429486567,
+      "grad_norm": 2.910580106245249,
+      "learning_rate": 1e-06,
+      "loss": 0.0653,
+      "step": 4147
+    },
+    {
+      "epoch": 0.7931924658189119,
+      "grad_norm": 4.451474817603466,
+      "learning_rate": 1e-06,
+      "loss": 0.1097,
+      "step": 4148
+    },
+    {
+      "epoch": 0.7933836886891672,
+      "grad_norm": 3.2353521520336015,
+      "learning_rate": 1e-06,
+      "loss": 0.0621,
+      "step": 4149
+    },
+    {
+      "epoch": 0.7935749115594225,
+      "grad_norm": 3.8445922425022396,
+      "learning_rate": 1e-06,
+      "loss": 0.0675,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7937661344296778,
+      "grad_norm": 6.365716794774905,
+      "learning_rate": 1e-06,
+      "loss": 0.4904,
+      "step": 4151
+    },
+    {
+      "epoch": 0.7939573572999331,
+      "grad_norm": 2.0750244644745095,
+      "learning_rate": 1e-06,
+      "loss": 0.2245,
+      "step": 4152
+    },
+    {
+      "epoch": 0.7941485801701883,
+      "grad_norm": 2.9655509979694528,
+      "learning_rate": 1e-06,
+      "loss": 0.1524,
+      "step": 4153
+    },
+    {
+      "epoch": 0.7943398030404436,
+      "grad_norm": 4.635756073716562,
+      "learning_rate": 1e-06,
+      "loss": 0.2564,
+      "step": 4154
+    },
+    {
+      "epoch": 0.7945310259106989,
+      "grad_norm": 4.412560671640471,
+      "learning_rate": 1e-06,
+      "loss": 0.3785,
+      "step": 4155
+    },
+    {
+      "epoch": 0.7947222487809542,
+      "grad_norm": 4.100213691911184,
+      "learning_rate": 1e-06,
+      "loss": 0.3473,
+      "step": 4156
+    },
+    {
+      "epoch": 0.7949134716512095,
+      "grad_norm": 1.8753411300601976,
+      "learning_rate": 1e-06,
+      "loss": 0.1616,
+      "step": 4157
+    },
+    {
+      "epoch": 0.7951046945214648,
+      "grad_norm": 2.5221815260947675,
+      "learning_rate": 1e-06,
+      "loss": 0.1181,
+      "step": 4158
+    },
+    {
+      "epoch": 0.79529591739172,
+      "grad_norm": 2.98294528447815,
+      "learning_rate": 1e-06,
+      "loss": 0.2526,
+      "step": 4159
+    },
+    {
+      "epoch": 0.7954871402619753,
+      "grad_norm": 2.214829433065224,
+      "learning_rate": 1e-06,
+      "loss": 0.1926,
+      "step": 4160
+    },
+    {
+      "epoch": 0.7956783631322306,
+      "grad_norm": 2.7739612353475995,
+      "learning_rate": 1e-06,
+      "loss": 0.1462,
+      "step": 4161
+    },
+    {
+      "epoch": 0.7958695860024859,
+      "grad_norm": 2.5048365538491777,
+      "learning_rate": 1e-06,
+      "loss": 0.0574,
+      "step": 4162
+    },
+    {
+      "epoch": 0.7960608088727412,
+      "grad_norm": 4.567437204849045,
+      "learning_rate": 1e-06,
+      "loss": 0.0704,
+      "step": 4163
+    },
+    {
+      "epoch": 0.7962520317429964,
+      "grad_norm": 8.687702862683174,
+      "learning_rate": 1e-06,
+      "loss": 0.3776,
+      "step": 4164
+    },
+    {
+      "epoch": 0.7964432546132517,
+      "grad_norm": 4.56131195904421,
+      "learning_rate": 1e-06,
+      "loss": 0.4243,
+      "step": 4165
+    },
+    {
+      "epoch": 0.796634477483507,
+      "grad_norm": 2.8373488600217627,
+      "learning_rate": 1e-06,
+      "loss": 0.1884,
+      "step": 4166
+    },
+    {
+      "epoch": 0.7968257003537623,
+      "grad_norm": 2.7733024456828925,
+      "learning_rate": 1e-06,
+      "loss": 0.194,
+      "step": 4167
+    },
+    {
+      "epoch": 0.7970169232240176,
+      "grad_norm": 6.515358898848977,
+      "learning_rate": 1e-06,
+      "loss": 0.4005,
+      "step": 4168
+    },
+    {
+      "epoch": 0.7972081460942729,
+      "grad_norm": 4.746789951830791,
+      "learning_rate": 1e-06,
+      "loss": 0.5418,
+      "step": 4169
+    },
+    {
+      "epoch": 0.7973993689645281,
+      "grad_norm": 3.892930661328446,
+      "learning_rate": 1e-06,
+      "loss": 0.3463,
+      "step": 4170
+    },
+    {
+      "epoch": 0.7975905918347834,
+      "grad_norm": 3.165420203254441,
+      "learning_rate": 1e-06,
+      "loss": 0.1386,
+      "step": 4171
+    },
+    {
+      "epoch": 0.7977818147050387,
+      "grad_norm": 3.6580991063194657,
+      "learning_rate": 1e-06,
+      "loss": 0.1415,
+      "step": 4172
+    },
+    {
+      "epoch": 0.797973037575294,
+      "grad_norm": 2.987349859957989,
+      "learning_rate": 1e-06,
+      "loss": 0.0567,
+      "step": 4173
+    },
+    {
+      "epoch": 0.7981642604455493,
+      "grad_norm": 3.0611762475085533,
+      "learning_rate": 1e-06,
+      "loss": 0.0623,
+      "step": 4174
+    },
+    {
+      "epoch": 0.7983554833158045,
+      "grad_norm": 5.5877117620360215,
+      "learning_rate": 1e-06,
+      "loss": 0.1142,
+      "step": 4175
+    },
+    {
+      "epoch": 0.7985467061860598,
+      "grad_norm": 3.0988888687341083,
+      "learning_rate": 1e-06,
+      "loss": 0.4484,
+      "step": 4176
+    },
+    {
+      "epoch": 0.7987379290563151,
+      "grad_norm": 5.216175472221658,
+      "learning_rate": 1e-06,
+      "loss": 0.3216,
+      "step": 4177
+    },
+    {
+      "epoch": 0.7989291519265704,
+      "grad_norm": 4.5258464295566485,
+      "learning_rate": 1e-06,
+      "loss": 0.5322,
+      "step": 4178
+    },
+    {
+      "epoch": 0.7991203747968257,
+      "grad_norm": 3.833808510118984,
+      "learning_rate": 1e-06,
+      "loss": 0.2577,
+      "step": 4179
+    },
+    {
+      "epoch": 0.7993115976670809,
+      "grad_norm": 2.9464305101528643,
+      "learning_rate": 1e-06,
+      "loss": 0.2721,
+      "step": 4180
+    },
+    {
+      "epoch": 0.7995028205373362,
+      "grad_norm": 2.6759187314988977,
+      "learning_rate": 1e-06,
+      "loss": 0.104,
+      "step": 4181
+    },
+    {
+      "epoch": 0.7996940434075915,
+      "grad_norm": 4.0673531520902335,
+      "learning_rate": 1e-06,
+      "loss": 0.4249,
+      "step": 4182
+    },
+    {
+      "epoch": 0.7998852662778468,
+      "grad_norm": 1.9241907078213214,
+      "learning_rate": 1e-06,
+      "loss": 0.0743,
+      "step": 4183
+    },
+    {
+      "epoch": 0.8000764891481021,
+      "grad_norm": 2.922454776462964,
+      "learning_rate": 1e-06,
+      "loss": 0.088,
+      "step": 4184
+    },
+    {
+      "epoch": 0.8002677120183574,
+      "grad_norm": 4.3632991725515975,
+      "learning_rate": 1e-06,
+      "loss": 0.2822,
+      "step": 4185
+    },
+    {
+      "epoch": 0.8004589348886126,
+      "grad_norm": 5.292989830735234,
+      "learning_rate": 1e-06,
+      "loss": 0.2912,
+      "step": 4186
+    },
+    {
+      "epoch": 0.8006501577588679,
+      "grad_norm": 4.5989412830882035,
+      "learning_rate": 1e-06,
+      "loss": 0.1713,
+      "step": 4187
+    },
+    {
+      "epoch": 0.8008413806291232,
+      "grad_norm": 6.515305618764803,
+      "learning_rate": 1e-06,
+      "loss": 0.1007,
+      "step": 4188
+    },
+    {
+      "epoch": 0.8010326034993785,
+      "grad_norm": 7.652052620163939,
+      "learning_rate": 1e-06,
+      "loss": 0.5043,
+      "step": 4189
+    },
+    {
+      "epoch": 0.8012238263696339,
+      "grad_norm": 3.4826724421739033,
+      "learning_rate": 1e-06,
+      "loss": 0.2995,
+      "step": 4190
+    },
+    {
+      "epoch": 0.801415049239889,
+      "grad_norm": 1.9179213259674412,
+      "learning_rate": 1e-06,
+      "loss": 0.1627,
+      "step": 4191
+    },
+    {
+      "epoch": 0.8016062721101443,
+      "grad_norm": 4.4418509201185055,
+      "learning_rate": 1e-06,
+      "loss": 0.2881,
+      "step": 4192
+    },
+    {
+      "epoch": 0.8017974949803996,
+      "grad_norm": 2.3707763861692595,
+      "learning_rate": 1e-06,
+      "loss": 0.2167,
+      "step": 4193
+    },
+    {
+      "epoch": 0.801988717850655,
+      "grad_norm": 4.4537402347069,
+      "learning_rate": 1e-06,
+      "loss": 0.4418,
+      "step": 4194
+    },
+    {
+      "epoch": 0.8021799407209103,
+      "grad_norm": 2.0730130989481577,
+      "learning_rate": 1e-06,
+      "loss": 0.084,
+      "step": 4195
+    },
+    {
+      "epoch": 0.8023711635911654,
+      "grad_norm": 4.145872540823688,
+      "learning_rate": 1e-06,
+      "loss": 0.1366,
+      "step": 4196
+    },
+    {
+      "epoch": 0.8025623864614208,
+      "grad_norm": 5.567032585376219,
+      "learning_rate": 1e-06,
+      "loss": 0.1538,
+      "step": 4197
+    },
+    {
+      "epoch": 0.802753609331676,
+      "grad_norm": 2.404978032001877,
+      "learning_rate": 1e-06,
+      "loss": 0.0509,
+      "step": 4198
+    },
+    {
+      "epoch": 0.8029448322019314,
+      "grad_norm": 3.462691230864042,
+      "learning_rate": 1e-06,
+      "loss": 0.1086,
+      "step": 4199
+    },
+    {
+      "epoch": 0.8031360550721867,
+      "grad_norm": 2.271188783970113,
+      "learning_rate": 1e-06,
+      "loss": 0.0844,
+      "step": 4200
+    },
+    {
+      "epoch": 0.803327277942442,
+      "grad_norm": 7.757529753942345,
+      "learning_rate": 1e-06,
+      "loss": 0.4912,
+      "step": 4201
+    },
+    {
+      "epoch": 0.8035185008126972,
+      "grad_norm": 3.6365688797076166,
+      "learning_rate": 1e-06,
+      "loss": 0.3617,
+      "step": 4202
+    },
+    {
+      "epoch": 0.8037097236829525,
+      "grad_norm": 2.033066157730482,
+      "learning_rate": 1e-06,
+      "loss": 0.2105,
+      "step": 4203
+    },
+    {
+      "epoch": 0.8039009465532078,
+      "grad_norm": 3.8411908818421234,
+      "learning_rate": 1e-06,
+      "loss": 0.3284,
+      "step": 4204
+    },
+    {
+      "epoch": 0.8040921694234631,
+      "grad_norm": 5.071596238801333,
+      "learning_rate": 1e-06,
+      "loss": 0.5064,
+      "step": 4205
+    },
+    {
+      "epoch": 0.8042833922937184,
+      "grad_norm": 5.303693945177392,
+      "learning_rate": 1e-06,
+      "loss": 0.2759,
+      "step": 4206
+    },
+    {
+      "epoch": 0.8044746151639736,
+      "grad_norm": 1.4503307754492876,
+      "learning_rate": 1e-06,
+      "loss": 0.0269,
+      "step": 4207
+    },
+    {
+      "epoch": 0.8046658380342289,
+      "grad_norm": 5.487725867139419,
+      "learning_rate": 1e-06,
+      "loss": 0.1273,
+      "step": 4208
+    },
+    {
+      "epoch": 0.8048570609044842,
+      "grad_norm": 3.225641077478673,
+      "learning_rate": 1e-06,
+      "loss": 0.1279,
+      "step": 4209
+    },
+    {
+      "epoch": 0.8050482837747395,
+      "grad_norm": 4.360806794020505,
+      "learning_rate": 1e-06,
+      "loss": 0.2912,
+      "step": 4210
+    },
+    {
+      "epoch": 0.8052395066449948,
+      "grad_norm": 3.56760268600231,
+      "learning_rate": 1e-06,
+      "loss": 0.104,
+      "step": 4211
+    },
+    {
+      "epoch": 0.8054307295152501,
+      "grad_norm": 2.8616295769638658,
+      "learning_rate": 1e-06,
+      "loss": 0.0733,
+      "step": 4212
+    },
+    {
+      "epoch": 0.8056219523855053,
+      "grad_norm": 4.82061202789504,
+      "learning_rate": 1e-06,
+      "loss": 0.1585,
+      "step": 4213
+    },
+    {
+      "epoch": 0.8058131752557606,
+      "grad_norm": 4.058249024775923,
+      "learning_rate": 1e-06,
+      "loss": 0.6207,
+      "step": 4214
+    },
+    {
+      "epoch": 0.8060043981260159,
+      "grad_norm": 3.207815318378115,
+      "learning_rate": 1e-06,
+      "loss": 0.1162,
+      "step": 4215
+    },
+    {
+      "epoch": 0.8061956209962712,
+      "grad_norm": 3.3953078317763725,
+      "learning_rate": 1e-06,
+      "loss": 0.2017,
+      "step": 4216
+    },
+    {
+      "epoch": 0.8063868438665265,
+      "grad_norm": 2.8317436545782426,
+      "learning_rate": 1e-06,
+      "loss": 0.1752,
+      "step": 4217
+    },
+    {
+      "epoch": 0.8065780667367817,
+      "grad_norm": 3.5725029501025203,
+      "learning_rate": 1e-06,
+      "loss": 0.2997,
+      "step": 4218
+    },
+    {
+      "epoch": 0.806769289607037,
+      "grad_norm": 10.807497487999939,
+      "learning_rate": 1e-06,
+      "loss": 0.4871,
+      "step": 4219
+    },
+    {
+      "epoch": 0.8069605124772923,
+      "grad_norm": 3.9583105521634057,
+      "learning_rate": 1e-06,
+      "loss": 0.1913,
+      "step": 4220
+    },
+    {
+      "epoch": 0.8071517353475476,
+      "grad_norm": 2.737617793227026,
+      "learning_rate": 1e-06,
+      "loss": 0.1407,
+      "step": 4221
+    },
+    {
+      "epoch": 0.8073429582178029,
+      "grad_norm": 2.4122001105783326,
+      "learning_rate": 1e-06,
+      "loss": 0.1692,
+      "step": 4222
+    },
+    {
+      "epoch": 0.8075341810880581,
+      "grad_norm": 3.423980622450641,
+      "learning_rate": 1e-06,
+      "loss": 0.156,
+      "step": 4223
+    },
+    {
+      "epoch": 0.8077254039583134,
+      "grad_norm": 2.5436465155619348,
+      "learning_rate": 1e-06,
+      "loss": 0.0981,
+      "step": 4224
+    },
+    {
+      "epoch": 0.8079166268285687,
+      "grad_norm": 4.1272738576268,
+      "learning_rate": 1e-06,
+      "loss": 0.1418,
+      "step": 4225
+    },
+    {
+      "epoch": 0.808107849698824,
+      "grad_norm": 3.4132073420378686,
+      "learning_rate": 1e-06,
+      "loss": 0.1086,
+      "step": 4226
+    },
+    {
+      "epoch": 0.8082990725690793,
+      "grad_norm": 7.071547662742262,
+      "learning_rate": 1e-06,
+      "loss": 0.4395,
+      "step": 4227
+    },
+    {
+      "epoch": 0.8084902954393346,
+      "grad_norm": 3.500111850586174,
+      "learning_rate": 1e-06,
+      "loss": 0.457,
+      "step": 4228
+    },
+    {
+      "epoch": 0.8086815183095898,
+      "grad_norm": 1.5510714088937592,
+      "learning_rate": 1e-06,
+      "loss": 0.1071,
+      "step": 4229
+    },
+    {
+      "epoch": 0.8088727411798451,
+      "grad_norm": 3.0079364701522007,
+      "learning_rate": 1e-06,
+      "loss": 0.155,
+      "step": 4230
+    },
+    {
+      "epoch": 0.8090639640501004,
+      "grad_norm": 2.883869091232085,
+      "learning_rate": 1e-06,
+      "loss": 0.2206,
+      "step": 4231
+    },
+    {
+      "epoch": 0.8092551869203557,
+      "grad_norm": 4.555184129496115,
+      "learning_rate": 1e-06,
+      "loss": 0.1359,
+      "step": 4232
+    },
+    {
+      "epoch": 0.809446409790611,
+      "grad_norm": 3.8345716107364862,
+      "learning_rate": 1e-06,
+      "loss": 0.125,
+      "step": 4233
+    },
+    {
+      "epoch": 0.8096376326608662,
+      "grad_norm": 1.8870460747580828,
+      "learning_rate": 1e-06,
+      "loss": 0.1517,
+      "step": 4234
+    },
+    {
+      "epoch": 0.8098288555311215,
+      "grad_norm": 4.013757412409702,
+      "learning_rate": 1e-06,
+      "loss": 0.0966,
+      "step": 4235
+    },
+    {
+      "epoch": 0.8100200784013768,
+      "grad_norm": 5.243970405756985,
+      "learning_rate": 1e-06,
+      "loss": 0.2266,
+      "step": 4236
+    },
+    {
+      "epoch": 0.8102113012716321,
+      "grad_norm": 2.967615171670496,
+      "learning_rate": 1e-06,
+      "loss": 0.086,
+      "step": 4237
+    },
+    {
+      "epoch": 0.8104025241418874,
+      "grad_norm": 6.737533925949488,
+      "learning_rate": 1e-06,
+      "loss": 0.1248,
+      "step": 4238
+    },
+    {
+      "epoch": 0.8105937470121427,
+      "grad_norm": 4.910236658425747,
+      "learning_rate": 1e-06,
+      "loss": 0.4106,
+      "step": 4239
+    },
+    {
+      "epoch": 0.8107849698823979,
+      "grad_norm": 5.742340293460362,
+      "learning_rate": 1e-06,
+      "loss": 0.4135,
+      "step": 4240
+    },
+    {
+      "epoch": 0.8109761927526532,
+      "grad_norm": 3.033100467868012,
+      "learning_rate": 1e-06,
+      "loss": 0.212,
+      "step": 4241
+    },
+    {
+      "epoch": 0.8111674156229085,
+      "grad_norm": 2.6040817755532006,
+      "learning_rate": 1e-06,
+      "loss": 0.2244,
+      "step": 4242
+    },
+    {
+      "epoch": 0.8113586384931638,
+      "grad_norm": 2.7999078939819992,
+      "learning_rate": 1e-06,
+      "loss": 0.1094,
+      "step": 4243
+    },
+    {
+      "epoch": 0.8115498613634191,
+      "grad_norm": 4.894080358197057,
+      "learning_rate": 1e-06,
+      "loss": 0.2101,
+      "step": 4244
+    },
+    {
+      "epoch": 0.8117410842336743,
+      "grad_norm": 4.864621975709185,
+      "learning_rate": 1e-06,
+      "loss": 0.3675,
+      "step": 4245
+    },
+    {
+      "epoch": 0.8119323071039296,
+      "grad_norm": 4.126037004919708,
+      "learning_rate": 1e-06,
+      "loss": 0.2568,
+      "step": 4246
+    },
+    {
+      "epoch": 0.8121235299741849,
+      "grad_norm": 3.132462099033264,
+      "learning_rate": 1e-06,
+      "loss": 0.0447,
+      "step": 4247
+    },
+    {
+      "epoch": 0.8123147528444402,
+      "grad_norm": 3.711101199103219,
+      "learning_rate": 1e-06,
+      "loss": 0.0777,
+      "step": 4248
+    },
+    {
+      "epoch": 0.8125059757146955,
+      "grad_norm": 3.812625320125889,
+      "learning_rate": 1e-06,
+      "loss": 0.0763,
+      "step": 4249
+    },
+    {
+      "epoch": 0.8126971985849507,
+      "grad_norm": 4.5001195255936155,
+      "learning_rate": 1e-06,
+      "loss": 0.0738,
+      "step": 4250
+    },
+    {
+      "epoch": 0.812888421455206,
+      "grad_norm": 6.0304254946669,
+      "learning_rate": 1e-06,
+      "loss": 0.4134,
+      "step": 4251
+    },
+    {
+      "epoch": 0.8130796443254613,
+      "grad_norm": 3.7354461538268287,
+      "learning_rate": 1e-06,
+      "loss": 0.3487,
+      "step": 4252
+    },
+    {
+      "epoch": 0.8132708671957166,
+      "grad_norm": 4.679314141512721,
+      "learning_rate": 1e-06,
+      "loss": 0.2992,
+      "step": 4253
+    },
+    {
+      "epoch": 0.8134620900659719,
+      "grad_norm": 2.636800581050781,
+      "learning_rate": 1e-06,
+      "loss": 0.342,
+      "step": 4254
+    },
+    {
+      "epoch": 0.8136533129362272,
+      "grad_norm": 3.9602452261933117,
+      "learning_rate": 1e-06,
+      "loss": 0.2229,
+      "step": 4255
+    },
+    {
+      "epoch": 0.8138445358064824,
+      "grad_norm": 4.009416701589809,
+      "learning_rate": 1e-06,
+      "loss": 0.1791,
+      "step": 4256
+    },
+    {
+      "epoch": 0.8140357586767377,
+      "grad_norm": 3.7494282604516433,
+      "learning_rate": 1e-06,
+      "loss": 0.1113,
+      "step": 4257
+    },
+    {
+      "epoch": 0.814226981546993,
+      "grad_norm": 2.993953970250138,
+      "learning_rate": 1e-06,
+      "loss": 0.3216,
+      "step": 4258
+    },
+    {
+      "epoch": 0.8144182044172483,
+      "grad_norm": 1.64125325117639,
+      "learning_rate": 1e-06,
+      "loss": 0.0361,
+      "step": 4259
+    },
+    {
+      "epoch": 0.8146094272875036,
+      "grad_norm": 4.204109615446701,
+      "learning_rate": 1e-06,
+      "loss": 0.137,
+      "step": 4260
+    },
+    {
+      "epoch": 0.8148006501577588,
+      "grad_norm": 4.165016050823008,
+      "learning_rate": 1e-06,
+      "loss": 0.2609,
+      "step": 4261
+    },
+    {
+      "epoch": 0.8149918730280141,
+      "grad_norm": 3.3678048756090044,
+      "learning_rate": 1e-06,
+      "loss": 0.1179,
+      "step": 4262
+    },
+    {
+      "epoch": 0.8151830958982694,
+      "grad_norm": 9.181845657211385,
+      "learning_rate": 1e-06,
+      "loss": 0.4304,
+      "step": 4263
+    },
+    {
+      "epoch": 0.8153743187685247,
+      "grad_norm": 3.466985030105352,
+      "learning_rate": 1e-06,
+      "loss": 0.5777,
+      "step": 4264
+    },
+    {
+      "epoch": 0.81556554163878,
+      "grad_norm": 4.1770294008396105,
+      "learning_rate": 1e-06,
+      "loss": 0.226,
+      "step": 4265
+    },
+    {
+      "epoch": 0.8157567645090352,
+      "grad_norm": 2.4452345960596187,
+      "learning_rate": 1e-06,
+      "loss": 0.1399,
+      "step": 4266
+    },
+    {
+      "epoch": 0.8159479873792905,
+      "grad_norm": 2.459943104217813,
+      "learning_rate": 1e-06,
+      "loss": 0.1764,
+      "step": 4267
+    },
+    {
+      "epoch": 0.8161392102495458,
+      "grad_norm": 3.742129522523544,
+      "learning_rate": 1e-06,
+      "loss": 0.2436,
+      "step": 4268
+    },
+    {
+      "epoch": 0.8163304331198011,
+      "grad_norm": 2.4316810206232993,
+      "learning_rate": 1e-06,
+      "loss": 0.0779,
+      "step": 4269
+    },
+    {
+      "epoch": 0.8165216559900564,
+      "grad_norm": 4.603821800515137,
+      "learning_rate": 1e-06,
+      "loss": 0.251,
+      "step": 4270
+    },
+    {
+      "epoch": 0.8167128788603117,
+      "grad_norm": 2.855764642458344,
+      "learning_rate": 1e-06,
+      "loss": 0.1419,
+      "step": 4271
+    },
+    {
+      "epoch": 0.8169041017305669,
+      "grad_norm": 2.669794622373241,
+      "learning_rate": 1e-06,
+      "loss": 0.0649,
+      "step": 4272
+    },
+    {
+      "epoch": 0.8170953246008222,
+      "grad_norm": 2.3220166457086386,
+      "learning_rate": 1e-06,
+      "loss": 0.058,
+      "step": 4273
+    },
+    {
+      "epoch": 0.8172865474710775,
+      "grad_norm": 3.985677827745493,
+      "learning_rate": 1e-06,
+      "loss": 0.1054,
+      "step": 4274
+    },
+    {
+      "epoch": 0.8174777703413328,
+      "grad_norm": 8.180922333965468,
+      "learning_rate": 1e-06,
+      "loss": 0.2625,
+      "step": 4275
+    },
+    {
+      "epoch": 0.8176689932115881,
+      "grad_norm": 25.87241457493869,
+      "learning_rate": 1e-06,
+      "loss": 0.3211,
+      "step": 4276
+    },
+    {
+      "epoch": 0.8178602160818433,
+      "grad_norm": 3.9475725706748737,
+      "learning_rate": 1e-06,
+      "loss": 0.5625,
+      "step": 4277
+    },
+    {
+      "epoch": 0.8180514389520986,
+      "grad_norm": 5.445394113112396,
+      "learning_rate": 1e-06,
+      "loss": 0.6764,
+      "step": 4278
+    },
+    {
+      "epoch": 0.818242661822354,
+      "grad_norm": 3.2438188608023073,
+      "learning_rate": 1e-06,
+      "loss": 0.3959,
+      "step": 4279
+    },
+    {
+      "epoch": 0.8184338846926092,
+      "grad_norm": 4.963535859890797,
+      "learning_rate": 1e-06,
+      "loss": 0.2405,
+      "step": 4280
+    },
+    {
+      "epoch": 0.8186251075628646,
+      "grad_norm": 4.016187813945554,
+      "learning_rate": 1e-06,
+      "loss": 0.2322,
+      "step": 4281
+    },
+    {
+      "epoch": 0.8188163304331199,
+      "grad_norm": 2.8035681570934754,
+      "learning_rate": 1e-06,
+      "loss": 0.1662,
+      "step": 4282
+    },
+    {
+      "epoch": 0.819007553303375,
+      "grad_norm": 1.596534689756061,
+      "learning_rate": 1e-06,
+      "loss": 0.1497,
+      "step": 4283
+    },
+    {
+      "epoch": 0.8191987761736303,
+      "grad_norm": 2.2831716935055817,
+      "learning_rate": 1e-06,
+      "loss": 0.1404,
+      "step": 4284
+    },
+    {
+      "epoch": 0.8193899990438857,
+      "grad_norm": 4.135699873418424,
+      "learning_rate": 1e-06,
+      "loss": 0.0856,
+      "step": 4285
+    },
+    {
+      "epoch": 0.819581221914141,
+      "grad_norm": 4.084972960663935,
+      "learning_rate": 1e-06,
+      "loss": 0.1243,
+      "step": 4286
+    },
+    {
+      "epoch": 0.8197724447843963,
+      "grad_norm": 3.3793539996417983,
+      "learning_rate": 1e-06,
+      "loss": 0.1084,
+      "step": 4287
+    },
+    {
+      "epoch": 0.8199636676546515,
+      "grad_norm": 11.551959537889624,
+      "learning_rate": 1e-06,
+      "loss": 0.2477,
+      "step": 4288
+    },
+    {
+      "epoch": 0.8201548905249068,
+      "grad_norm": 5.008982600539244,
+      "learning_rate": 1e-06,
+      "loss": 0.6292,
+      "step": 4289
+    },
+    {
+      "epoch": 0.8203461133951621,
+      "grad_norm": 2.92067428641975,
+      "learning_rate": 1e-06,
+      "loss": 0.3245,
+      "step": 4290
+    },
+    {
+      "epoch": 0.8205373362654174,
+      "grad_norm": 6.4402228957750856,
+      "learning_rate": 1e-06,
+      "loss": 0.4456,
+      "step": 4291
+    },
+    {
+      "epoch": 0.8207285591356727,
+      "grad_norm": 3.296303215884765,
+      "learning_rate": 1e-06,
+      "loss": 0.168,
+      "step": 4292
+    },
+    {
+      "epoch": 0.8209197820059279,
+      "grad_norm": 2.4951218696445796,
+      "learning_rate": 1e-06,
+      "loss": 0.1433,
+      "step": 4293
+    },
+    {
+      "epoch": 0.8211110048761832,
+      "grad_norm": 3.490980515778891,
+      "learning_rate": 1e-06,
+      "loss": 0.2184,
+      "step": 4294
+    },
+    {
+      "epoch": 0.8213022277464385,
+      "grad_norm": 3.278477978316933,
+      "learning_rate": 1e-06,
+      "loss": 0.1353,
+      "step": 4295
+    },
+    {
+      "epoch": 0.8214934506166938,
+      "grad_norm": 2.9951224412551736,
+      "learning_rate": 1e-06,
+      "loss": 0.1188,
+      "step": 4296
+    },
+    {
+      "epoch": 0.8216846734869491,
+      "grad_norm": 2.147652888512227,
+      "learning_rate": 1e-06,
+      "loss": 0.1198,
+      "step": 4297
+    },
+    {
+      "epoch": 0.8218758963572044,
+      "grad_norm": 3.0999687193246026,
+      "learning_rate": 1e-06,
+      "loss": 0.0884,
+      "step": 4298
+    },
+    {
+      "epoch": 0.8220671192274596,
+      "grad_norm": 4.706721657349798,
+      "learning_rate": 1e-06,
+      "loss": 0.4826,
+      "step": 4299
+    },
+    {
+      "epoch": 0.8222583420977149,
+      "grad_norm": 8.6793038105732,
+      "learning_rate": 1e-06,
+      "loss": 0.229,
+      "step": 4300
+    },
+    {
+      "epoch": 0.8224495649679702,
+      "grad_norm": 5.403196144849075,
+      "learning_rate": 1e-06,
+      "loss": 0.4801,
+      "step": 4301
+    },
+    {
+      "epoch": 0.8226407878382255,
+      "grad_norm": 4.636195473842716,
+      "learning_rate": 1e-06,
+      "loss": 0.4351,
+      "step": 4302
+    },
+    {
+      "epoch": 0.8228320107084808,
+      "grad_norm": 4.554054491081464,
+      "learning_rate": 1e-06,
+      "loss": 0.3452,
+      "step": 4303
+    },
+    {
+      "epoch": 0.823023233578736,
+      "grad_norm": 3.881379813021355,
+      "learning_rate": 1e-06,
+      "loss": 0.4029,
+      "step": 4304
+    },
+    {
+      "epoch": 0.8232144564489913,
+      "grad_norm": 3.451419184598907,
+      "learning_rate": 1e-06,
+      "loss": 0.112,
+      "step": 4305
+    },
+    {
+      "epoch": 0.8234056793192466,
+      "grad_norm": 2.689485481859869,
+      "learning_rate": 1e-06,
+      "loss": 0.2408,
+      "step": 4306
+    },
+    {
+      "epoch": 0.8235969021895019,
+      "grad_norm": 4.087779809549715,
+      "learning_rate": 1e-06,
+      "loss": 0.354,
+      "step": 4307
+    },
+    {
+      "epoch": 0.8237881250597572,
+      "grad_norm": 3.371908255228372,
+      "learning_rate": 1e-06,
+      "loss": 0.1796,
+      "step": 4308
+    },
+    {
+      "epoch": 0.8239793479300125,
+      "grad_norm": 3.3234698912311393,
+      "learning_rate": 1e-06,
+      "loss": 0.1358,
+      "step": 4309
+    },
+    {
+      "epoch": 0.8241705708002677,
+      "grad_norm": 3.7395387326328535,
+      "learning_rate": 1e-06,
+      "loss": 0.3338,
+      "step": 4310
+    },
+    {
+      "epoch": 0.824361793670523,
+      "grad_norm": 2.436229570360875,
+      "learning_rate": 1e-06,
+      "loss": 0.0923,
+      "step": 4311
+    },
+    {
+      "epoch": 0.8245530165407783,
+      "grad_norm": 6.019981491212495,
+      "learning_rate": 1e-06,
+      "loss": 0.0876,
+      "step": 4312
+    },
+    {
+      "epoch": 0.8247442394110336,
+      "grad_norm": 4.133747420539432,
+      "learning_rate": 1e-06,
+      "loss": 0.2874,
+      "step": 4313
+    },
+    {
+      "epoch": 0.8249354622812889,
+      "grad_norm": 5.0778926854913164,
+      "learning_rate": 1e-06,
+      "loss": 0.6965,
+      "step": 4314
+    },
+    {
+      "epoch": 0.8251266851515441,
+      "grad_norm": 3.8111900205262126,
+      "learning_rate": 1e-06,
+      "loss": 0.3291,
+      "step": 4315
+    },
+    {
+      "epoch": 0.8253179080217994,
+      "grad_norm": 2.6534775570469695,
+      "learning_rate": 1e-06,
+      "loss": 0.1081,
+      "step": 4316
+    },
+    {
+      "epoch": 0.8255091308920547,
+      "grad_norm": 4.042655247129485,
+      "learning_rate": 1e-06,
+      "loss": 0.3858,
+      "step": 4317
+    },
+    {
+      "epoch": 0.82570035376231,
+      "grad_norm": 2.579803567697838,
+      "learning_rate": 1e-06,
+      "loss": 0.2313,
+      "step": 4318
+    },
+    {
+      "epoch": 0.8258915766325653,
+      "grad_norm": 5.686529055708413,
+      "learning_rate": 1e-06,
+      "loss": 0.2059,
+      "step": 4319
+    },
+    {
+      "epoch": 0.8260827995028205,
+      "grad_norm": 4.166306111312415,
+      "learning_rate": 1e-06,
+      "loss": 0.301,
+      "step": 4320
+    },
+    {
+      "epoch": 0.8262740223730758,
+      "grad_norm": 3.422856020414358,
+      "learning_rate": 1e-06,
+      "loss": 0.1679,
+      "step": 4321
+    },
+    {
+      "epoch": 0.8264652452433311,
+      "grad_norm": 2.878755355876358,
+      "learning_rate": 1e-06,
+      "loss": 0.0974,
+      "step": 4322
+    },
+    {
+      "epoch": 0.8266564681135864,
+      "grad_norm": 3.208285013470769,
+      "learning_rate": 1e-06,
+      "loss": 0.1882,
+      "step": 4323
+    },
+    {
+      "epoch": 0.8268476909838417,
+      "grad_norm": 3.5925975776221204,
+      "learning_rate": 1e-06,
+      "loss": 0.1364,
+      "step": 4324
+    },
+    {
+      "epoch": 0.827038913854097,
+      "grad_norm": 6.3401981426287435,
+      "learning_rate": 1e-06,
+      "loss": 0.238,
+      "step": 4325
+    },
+    {
+      "epoch": 0.8272301367243522,
+      "grad_norm": 8.701704748343497,
+      "learning_rate": 1e-06,
+      "loss": 0.1477,
+      "step": 4326
+    },
+    {
+      "epoch": 0.8274213595946075,
+      "grad_norm": 5.051116200586378,
+      "learning_rate": 1e-06,
+      "loss": 0.4821,
+      "step": 4327
+    },
+    {
+      "epoch": 0.8276125824648628,
+      "grad_norm": 4.422505795403323,
+      "learning_rate": 1e-06,
+      "loss": 0.1805,
+      "step": 4328
+    },
+    {
+      "epoch": 0.8278038053351181,
+      "grad_norm": 2.7604138260352915,
+      "learning_rate": 1e-06,
+      "loss": 0.295,
+      "step": 4329
+    },
+    {
+      "epoch": 0.8279950282053734,
+      "grad_norm": 4.1983724619601945,
+      "learning_rate": 1e-06,
+      "loss": 0.3177,
+      "step": 4330
+    },
+    {
+      "epoch": 0.8281862510756286,
+      "grad_norm": 1.665456984528773,
+      "learning_rate": 1e-06,
+      "loss": 0.1192,
+      "step": 4331
+    },
+    {
+      "epoch": 0.8283774739458839,
+      "grad_norm": 3.775964494990868,
+      "learning_rate": 1e-06,
+      "loss": 0.1113,
+      "step": 4332
+    },
+    {
+      "epoch": 0.8285686968161392,
+      "grad_norm": 3.4135976514201625,
+      "learning_rate": 1e-06,
+      "loss": 0.1548,
+      "step": 4333
+    },
+    {
+      "epoch": 0.8287599196863945,
+      "grad_norm": 2.631226557176411,
+      "learning_rate": 1e-06,
+      "loss": 0.0949,
+      "step": 4334
+    },
+    {
+      "epoch": 0.8289511425566498,
+      "grad_norm": 4.711480205900352,
+      "learning_rate": 1e-06,
+      "loss": 0.0925,
+      "step": 4335
+    },
+    {
+      "epoch": 0.829142365426905,
+      "grad_norm": 2.2343823626203543,
+      "learning_rate": 1e-06,
+      "loss": 0.0675,
+      "step": 4336
+    },
+    {
+      "epoch": 0.8293335882971603,
+      "grad_norm": 3.3738672686813995,
+      "learning_rate": 1e-06,
+      "loss": 0.2337,
+      "step": 4337
+    },
+    {
+      "epoch": 0.8295248111674156,
+      "grad_norm": 9.066208494988926,
+      "learning_rate": 1e-06,
+      "loss": 0.1362,
+      "step": 4338
+    },
+    {
+      "epoch": 0.8297160340376709,
+      "grad_norm": 7.554531975519107,
+      "learning_rate": 1e-06,
+      "loss": 0.683,
+      "step": 4339
+    },
+    {
+      "epoch": 0.8299072569079262,
+      "grad_norm": 3.068521302319423,
+      "learning_rate": 1e-06,
+      "loss": 0.238,
+      "step": 4340
+    },
+    {
+      "epoch": 0.8300984797781815,
+      "grad_norm": 2.1943169233502524,
+      "learning_rate": 1e-06,
+      "loss": 0.0989,
+      "step": 4341
+    },
+    {
+      "epoch": 0.8302897026484367,
+      "grad_norm": 2.9357778492590674,
+      "learning_rate": 1e-06,
+      "loss": 0.1768,
+      "step": 4342
+    },
+    {
+      "epoch": 0.830480925518692,
+      "grad_norm": 5.106260697950085,
+      "learning_rate": 1e-06,
+      "loss": 0.392,
+      "step": 4343
+    },
+    {
+      "epoch": 0.8306721483889473,
+      "grad_norm": 2.017809608082843,
+      "learning_rate": 1e-06,
+      "loss": 0.2399,
+      "step": 4344
+    },
+    {
+      "epoch": 0.8308633712592026,
+      "grad_norm": 3.0775559572265845,
+      "learning_rate": 1e-06,
+      "loss": 0.3477,
+      "step": 4345
+    },
+    {
+      "epoch": 0.8310545941294579,
+      "grad_norm": 3.585043669026318,
+      "learning_rate": 1e-06,
+      "loss": 0.1852,
+      "step": 4346
+    },
+    {
+      "epoch": 0.8312458169997131,
+      "grad_norm": 4.319235144829247,
+      "learning_rate": 1e-06,
+      "loss": 0.0809,
+      "step": 4347
+    },
+    {
+      "epoch": 0.8314370398699684,
+      "grad_norm": 3.743894438592223,
+      "learning_rate": 1e-06,
+      "loss": 0.2364,
+      "step": 4348
+    },
+    {
+      "epoch": 0.8316282627402237,
+      "grad_norm": 3.056006105516736,
+      "learning_rate": 1e-06,
+      "loss": 0.1178,
+      "step": 4349
+    },
+    {
+      "epoch": 0.831819485610479,
+      "grad_norm": 7.567825586014902,
+      "learning_rate": 1e-06,
+      "loss": 0.0731,
+      "step": 4350
+    },
+    {
+      "epoch": 0.8320107084807343,
+      "grad_norm": 5.592245741458553,
+      "learning_rate": 1e-06,
+      "loss": 0.332,
+      "step": 4351
+    },
+    {
+      "epoch": 0.8322019313509896,
+      "grad_norm": 5.034054560914654,
+      "learning_rate": 1e-06,
+      "loss": 0.3668,
+      "step": 4352
+    },
+    {
+      "epoch": 0.8323931542212448,
+      "grad_norm": 3.866281854181943,
+      "learning_rate": 1e-06,
+      "loss": 0.3118,
+      "step": 4353
+    },
+    {
+      "epoch": 0.8325843770915001,
+      "grad_norm": 2.350974278687777,
+      "learning_rate": 1e-06,
+      "loss": 0.1812,
+      "step": 4354
+    },
+    {
+      "epoch": 0.8327755999617554,
+      "grad_norm": 3.283167251672326,
+      "learning_rate": 1e-06,
+      "loss": 0.1775,
+      "step": 4355
+    },
+    {
+      "epoch": 0.8329668228320107,
+      "grad_norm": 3.272008178682062,
+      "learning_rate": 1e-06,
+      "loss": 0.2156,
+      "step": 4356
+    },
+    {
+      "epoch": 0.833158045702266,
+      "grad_norm": 3.154193170218608,
+      "learning_rate": 1e-06,
+      "loss": 0.2686,
+      "step": 4357
+    },
+    {
+      "epoch": 0.8333492685725212,
+      "grad_norm": 2.4706311362412534,
+      "learning_rate": 1e-06,
+      "loss": 0.0971,
+      "step": 4358
+    },
+    {
+      "epoch": 0.8335404914427765,
+      "grad_norm": 2.2958799529376055,
+      "learning_rate": 1e-06,
+      "loss": 0.1156,
+      "step": 4359
+    },
+    {
+      "epoch": 0.8337317143130318,
+      "grad_norm": 5.319030876229266,
+      "learning_rate": 1e-06,
+      "loss": 0.1387,
+      "step": 4360
+    },
+    {
+      "epoch": 0.8339229371832871,
+      "grad_norm": 3.0263096772226024,
+      "learning_rate": 1e-06,
+      "loss": 0.1035,
+      "step": 4361
+    },
+    {
+      "epoch": 0.8341141600535424,
+      "grad_norm": 2.709606800481227,
+      "learning_rate": 1e-06,
+      "loss": 0.0392,
+      "step": 4362
+    },
+    {
+      "epoch": 0.8343053829237976,
+      "grad_norm": 5.8976413214723635,
+      "learning_rate": 1e-06,
+      "loss": 0.341,
+      "step": 4363
+    },
+    {
+      "epoch": 0.8344966057940529,
+      "grad_norm": 3.848942916212921,
+      "learning_rate": 1e-06,
+      "loss": 0.2419,
+      "step": 4364
+    },
+    {
+      "epoch": 0.8346878286643082,
+      "grad_norm": 4.482645691579944,
+      "learning_rate": 1e-06,
+      "loss": 0.6515,
+      "step": 4365
+    },
+    {
+      "epoch": 0.8348790515345635,
+      "grad_norm": 3.5451539018848415,
+      "learning_rate": 1e-06,
+      "loss": 0.4109,
+      "step": 4366
+    },
+    {
+      "epoch": 0.8350702744048188,
+      "grad_norm": 4.1675854750702594,
+      "learning_rate": 1e-06,
+      "loss": 0.2119,
+      "step": 4367
+    },
+    {
+      "epoch": 0.8352614972750741,
+      "grad_norm": 2.0011675526149517,
+      "learning_rate": 1e-06,
+      "loss": 0.2161,
+      "step": 4368
+    },
+    {
+      "epoch": 0.8354527201453293,
+      "grad_norm": 4.836140025986856,
+      "learning_rate": 1e-06,
+      "loss": 0.3567,
+      "step": 4369
+    },
+    {
+      "epoch": 0.8356439430155846,
+      "grad_norm": 2.8375212815616786,
+      "learning_rate": 1e-06,
+      "loss": 0.2795,
+      "step": 4370
+    },
+    {
+      "epoch": 0.83583516588584,
+      "grad_norm": 5.295457858569433,
+      "learning_rate": 1e-06,
+      "loss": 0.1729,
+      "step": 4371
+    },
+    {
+      "epoch": 0.8360263887560953,
+      "grad_norm": 2.301256943443055,
+      "learning_rate": 1e-06,
+      "loss": 0.0622,
+      "step": 4372
+    },
+    {
+      "epoch": 0.8362176116263506,
+      "grad_norm": 3.353562689635885,
+      "learning_rate": 1e-06,
+      "loss": 0.2312,
+      "step": 4373
+    },
+    {
+      "epoch": 0.8364088344966057,
+      "grad_norm": 3.91914692257231,
+      "learning_rate": 1e-06,
+      "loss": 0.0982,
+      "step": 4374
+    },
+    {
+      "epoch": 0.836600057366861,
+      "grad_norm": 3.828490072475882,
+      "learning_rate": 1e-06,
+      "loss": 0.1125,
+      "step": 4375
+    },
+    {
+      "epoch": 0.8367912802371164,
+      "grad_norm": 6.635616505087367,
+      "learning_rate": 1e-06,
+      "loss": 0.545,
+      "step": 4376
+    },
+    {
+      "epoch": 0.8369825031073717,
+      "grad_norm": 5.908106597868503,
+      "learning_rate": 1e-06,
+      "loss": 0.3749,
+      "step": 4377
+    },
+    {
+      "epoch": 0.837173725977627,
+      "grad_norm": 2.364658427341426,
+      "learning_rate": 1e-06,
+      "loss": 0.1766,
+      "step": 4378
+    },
+    {
+      "epoch": 0.8373649488478822,
+      "grad_norm": 2.799063822371288,
+      "learning_rate": 1e-06,
+      "loss": 0.3355,
+      "step": 4379
+    },
+    {
+      "epoch": 0.8375561717181375,
+      "grad_norm": 3.319751934527536,
+      "learning_rate": 1e-06,
+      "loss": 0.4323,
+      "step": 4380
+    },
+    {
+      "epoch": 0.8377473945883928,
+      "grad_norm": 3.926916764805432,
+      "learning_rate": 1e-06,
+      "loss": 0.1354,
+      "step": 4381
+    },
+    {
+      "epoch": 0.8379386174586481,
+      "grad_norm": 5.6180741800073655,
+      "learning_rate": 1e-06,
+      "loss": 0.3689,
+      "step": 4382
+    },
+    {
+      "epoch": 0.8381298403289034,
+      "grad_norm": 3.0774361861156856,
+      "learning_rate": 1e-06,
+      "loss": 0.1955,
+      "step": 4383
+    },
+    {
+      "epoch": 0.8383210631991587,
+      "grad_norm": 3.767127279514654,
+      "learning_rate": 1e-06,
+      "loss": 0.1091,
+      "step": 4384
+    },
+    {
+      "epoch": 0.8385122860694139,
+      "grad_norm": 4.653630691216214,
+      "learning_rate": 1e-06,
+      "loss": 0.2503,
+      "step": 4385
+    },
+    {
+      "epoch": 0.8387035089396692,
+      "grad_norm": 3.9789185504884235,
+      "learning_rate": 1e-06,
+      "loss": 0.1257,
+      "step": 4386
+    },
+    {
+      "epoch": 0.8388947318099245,
+      "grad_norm": 5.364957965497737,
+      "learning_rate": 1e-06,
+      "loss": 0.0961,
+      "step": 4387
+    },
+    {
+      "epoch": 0.8390859546801798,
+      "grad_norm": 6.9575954510483875,
+      "learning_rate": 1e-06,
+      "loss": 0.1094,
+      "step": 4388
+    },
+    {
+      "epoch": 0.8392771775504351,
+      "grad_norm": 3.27636311242295,
+      "learning_rate": 1e-06,
+      "loss": 0.1489,
+      "step": 4389
+    },
+    {
+      "epoch": 0.8394684004206903,
+      "grad_norm": 5.328356069437361,
+      "learning_rate": 1e-06,
+      "loss": 0.2922,
+      "step": 4390
+    },
+    {
+      "epoch": 0.8396596232909456,
+      "grad_norm": 2.4048841488984043,
+      "learning_rate": 1e-06,
+      "loss": 0.0944,
+      "step": 4391
+    },
+    {
+      "epoch": 0.8398508461612009,
+      "grad_norm": 4.374277545678927,
+      "learning_rate": 1e-06,
+      "loss": 0.2179,
+      "step": 4392
+    },
+    {
+      "epoch": 0.8400420690314562,
+      "grad_norm": 3.725197980087062,
+      "learning_rate": 1e-06,
+      "loss": 0.4594,
+      "step": 4393
+    },
+    {
+      "epoch": 0.8402332919017115,
+      "grad_norm": 2.7022514034129603,
+      "learning_rate": 1e-06,
+      "loss": 0.2438,
+      "step": 4394
+    },
+    {
+      "epoch": 0.8404245147719668,
+      "grad_norm": 3.697762255959558,
+      "learning_rate": 1e-06,
+      "loss": 0.1488,
+      "step": 4395
+    },
+    {
+      "epoch": 0.840615737642222,
+      "grad_norm": 4.524007336395548,
+      "learning_rate": 1e-06,
+      "loss": 0.1629,
+      "step": 4396
+    },
+    {
+      "epoch": 0.8408069605124773,
+      "grad_norm": 2.351684060234056,
+      "learning_rate": 1e-06,
+      "loss": 0.099,
+      "step": 4397
+    },
+    {
+      "epoch": 0.8409981833827326,
+      "grad_norm": 3.32100147952788,
+      "learning_rate": 1e-06,
+      "loss": 0.0637,
+      "step": 4398
+    },
+    {
+      "epoch": 0.8411894062529879,
+      "grad_norm": 3.573767795151513,
+      "learning_rate": 1e-06,
+      "loss": 0.0664,
+      "step": 4399
+    },
+    {
+      "epoch": 0.8413806291232432,
+      "grad_norm": 3.08300812183242,
+      "learning_rate": 1e-06,
+      "loss": 0.0378,
+      "step": 4400
+    },
+    {
+      "epoch": 0.8415718519934984,
+      "grad_norm": 5.387960023999328,
+      "learning_rate": 1e-06,
+      "loss": 0.3599,
+      "step": 4401
+    },
+    {
+      "epoch": 0.8417630748637537,
+      "grad_norm": 4.915512768530654,
+      "learning_rate": 1e-06,
+      "loss": 0.4601,
+      "step": 4402
+    },
+    {
+      "epoch": 0.841954297734009,
+      "grad_norm": 1.6628808495967005,
+      "learning_rate": 1e-06,
+      "loss": 0.0678,
+      "step": 4403
+    },
+    {
+      "epoch": 0.8421455206042643,
+      "grad_norm": 3.9539634810494464,
+      "learning_rate": 1e-06,
+      "loss": 0.3294,
+      "step": 4404
+    },
+    {
+      "epoch": 0.8423367434745196,
+      "grad_norm": 3.9702449596174385,
+      "learning_rate": 1e-06,
+      "loss": 0.3595,
+      "step": 4405
+    },
+    {
+      "epoch": 0.8425279663447748,
+      "grad_norm": 4.999335817092161,
+      "learning_rate": 1e-06,
+      "loss": 0.1967,
+      "step": 4406
+    },
+    {
+      "epoch": 0.8427191892150301,
+      "grad_norm": 3.763772342530686,
+      "learning_rate": 1e-06,
+      "loss": 0.211,
+      "step": 4407
+    },
+    {
+      "epoch": 0.8429104120852854,
+      "grad_norm": 2.0290649399007594,
+      "learning_rate": 1e-06,
+      "loss": 0.062,
+      "step": 4408
+    },
+    {
+      "epoch": 0.8431016349555407,
+      "grad_norm": 3.4712673366224975,
+      "learning_rate": 1e-06,
+      "loss": 0.2406,
+      "step": 4409
+    },
+    {
+      "epoch": 0.843292857825796,
+      "grad_norm": 4.384364759306799,
+      "learning_rate": 1e-06,
+      "loss": 0.1529,
+      "step": 4410
+    },
+    {
+      "epoch": 0.8434840806960513,
+      "grad_norm": 6.48364253675614,
+      "learning_rate": 1e-06,
+      "loss": 0.209,
+      "step": 4411
+    },
+    {
+      "epoch": 0.8436753035663065,
+      "grad_norm": 4.884944651669856,
+      "learning_rate": 1e-06,
+      "loss": 0.1246,
+      "step": 4412
+    },
+    {
+      "epoch": 0.8438665264365618,
+      "grad_norm": 7.747849935264929,
+      "learning_rate": 1e-06,
+      "loss": 0.3416,
+      "step": 4413
+    },
+    {
+      "epoch": 0.8440577493068171,
+      "grad_norm": 3.60592605484537,
+      "learning_rate": 1e-06,
+      "loss": 0.2592,
+      "step": 4414
+    },
+    {
+      "epoch": 0.8442489721770724,
+      "grad_norm": 3.9469441580735194,
+      "learning_rate": 1e-06,
+      "loss": 0.3651,
+      "step": 4415
+    },
+    {
+      "epoch": 0.8444401950473277,
+      "grad_norm": 3.605741314902204,
+      "learning_rate": 1e-06,
+      "loss": 0.1682,
+      "step": 4416
+    },
+    {
+      "epoch": 0.8446314179175829,
+      "grad_norm": 2.440357098791651,
+      "learning_rate": 1e-06,
+      "loss": 0.1546,
+      "step": 4417
+    },
+    {
+      "epoch": 0.8448226407878382,
+      "grad_norm": 4.475886695007255,
+      "learning_rate": 1e-06,
+      "loss": 0.3014,
+      "step": 4418
+    },
+    {
+      "epoch": 0.8450138636580935,
+      "grad_norm": 3.7387872271787206,
+      "learning_rate": 1e-06,
+      "loss": 0.442,
+      "step": 4419
+    },
+    {
+      "epoch": 0.8452050865283488,
+      "grad_norm": 4.1693871327210665,
+      "learning_rate": 1e-06,
+      "loss": 0.2406,
+      "step": 4420
+    },
+    {
+      "epoch": 0.8453963093986041,
+      "grad_norm": 3.9079077293021833,
+      "learning_rate": 1e-06,
+      "loss": 0.1167,
+      "step": 4421
+    },
+    {
+      "epoch": 0.8455875322688594,
+      "grad_norm": 4.48518008139571,
+      "learning_rate": 1e-06,
+      "loss": 0.1592,
+      "step": 4422
+    },
+    {
+      "epoch": 0.8457787551391146,
+      "grad_norm": 1.9009046810038128,
+      "learning_rate": 1e-06,
+      "loss": 0.0396,
+      "step": 4423
+    },
+    {
+      "epoch": 0.8459699780093699,
+      "grad_norm": 2.3833738572559127,
+      "learning_rate": 1e-06,
+      "loss": 0.0932,
+      "step": 4424
+    },
+    {
+      "epoch": 0.8461612008796252,
+      "grad_norm": 4.357435444426429,
+      "learning_rate": 1e-06,
+      "loss": 0.1612,
+      "step": 4425
+    },
+    {
+      "epoch": 0.8463524237498805,
+      "grad_norm": 5.224510440067941,
+      "learning_rate": 1e-06,
+      "loss": 0.2995,
+      "step": 4426
+    },
+    {
+      "epoch": 0.8465436466201358,
+      "grad_norm": 5.806641610431299,
+      "learning_rate": 1e-06,
+      "loss": 0.6601,
+      "step": 4427
+    },
+    {
+      "epoch": 0.846734869490391,
+      "grad_norm": 4.549856774727335,
+      "learning_rate": 1e-06,
+      "loss": 0.3025,
+      "step": 4428
+    },
+    {
+      "epoch": 0.8469260923606463,
+      "grad_norm": 2.0273811469184504,
+      "learning_rate": 1e-06,
+      "loss": 0.2278,
+      "step": 4429
+    },
+    {
+      "epoch": 0.8471173152309016,
+      "grad_norm": 3.8277319881565295,
+      "learning_rate": 1e-06,
+      "loss": 0.4615,
+      "step": 4430
+    },
+    {
+      "epoch": 0.8473085381011569,
+      "grad_norm": 2.4667244822797443,
+      "learning_rate": 1e-06,
+      "loss": 0.1333,
+      "step": 4431
+    },
+    {
+      "epoch": 0.8474997609714122,
+      "grad_norm": 4.4420503738068415,
+      "learning_rate": 1e-06,
+      "loss": 0.442,
+      "step": 4432
+    },
+    {
+      "epoch": 0.8476909838416674,
+      "grad_norm": 3.3984929222761715,
+      "learning_rate": 1e-06,
+      "loss": 0.1482,
+      "step": 4433
+    },
+    {
+      "epoch": 0.8478822067119227,
+      "grad_norm": 1.868075299385165,
+      "learning_rate": 1e-06,
+      "loss": 0.0742,
+      "step": 4434
+    },
+    {
+      "epoch": 0.848073429582178,
+      "grad_norm": 2.4514716341235925,
+      "learning_rate": 1e-06,
+      "loss": 0.1368,
+      "step": 4435
+    },
+    {
+      "epoch": 0.8482646524524333,
+      "grad_norm": 5.17556548046709,
+      "learning_rate": 1e-06,
+      "loss": 0.2065,
+      "step": 4436
+    },
+    {
+      "epoch": 0.8484558753226886,
+      "grad_norm": 4.586188288367304,
+      "learning_rate": 1e-06,
+      "loss": 0.0838,
+      "step": 4437
+    },
+    {
+      "epoch": 0.8486470981929439,
+      "grad_norm": 3.793366620521669,
+      "learning_rate": 1e-06,
+      "loss": 0.0776,
+      "step": 4438
+    },
+    {
+      "epoch": 0.8488383210631991,
+      "grad_norm": 5.724551067491355,
+      "learning_rate": 1e-06,
+      "loss": 0.5873,
+      "step": 4439
+    },
+    {
+      "epoch": 0.8490295439334544,
+      "grad_norm": 3.0339519610490084,
+      "learning_rate": 1e-06,
+      "loss": 0.1479,
+      "step": 4440
+    },
+    {
+      "epoch": 0.8492207668037097,
+      "grad_norm": 5.217178930576595,
+      "learning_rate": 1e-06,
+      "loss": 0.1149,
+      "step": 4441
+    },
+    {
+      "epoch": 0.849411989673965,
+      "grad_norm": 2.7259755199561844,
+      "learning_rate": 1e-06,
+      "loss": 0.1614,
+      "step": 4442
+    },
+    {
+      "epoch": 0.8496032125442203,
+      "grad_norm": 3.414192389280444,
+      "learning_rate": 1e-06,
+      "loss": 0.4206,
+      "step": 4443
+    },
+    {
+      "epoch": 0.8497944354144755,
+      "grad_norm": 3.0049385112748928,
+      "learning_rate": 1e-06,
+      "loss": 0.237,
+      "step": 4444
+    },
+    {
+      "epoch": 0.8499856582847308,
+      "grad_norm": 4.214016667938219,
+      "learning_rate": 1e-06,
+      "loss": 0.2292,
+      "step": 4445
+    },
+    {
+      "epoch": 0.8501768811549861,
+      "grad_norm": 1.4049965676506717,
+      "learning_rate": 1e-06,
+      "loss": 0.0951,
+      "step": 4446
+    },
+    {
+      "epoch": 0.8503681040252414,
+      "grad_norm": 2.7787536390437753,
+      "learning_rate": 1e-06,
+      "loss": 0.1204,
+      "step": 4447
+    },
+    {
+      "epoch": 0.8505593268954967,
+      "grad_norm": 4.492710355101791,
+      "learning_rate": 1e-06,
+      "loss": 0.0765,
+      "step": 4448
+    },
+    {
+      "epoch": 0.8507505497657519,
+      "grad_norm": 2.366213758768438,
+      "learning_rate": 1e-06,
+      "loss": 0.1124,
+      "step": 4449
+    },
+    {
+      "epoch": 0.8509417726360072,
+      "grad_norm": 3.575008984534438,
+      "learning_rate": 1e-06,
+      "loss": 0.1275,
+      "step": 4450
+    },
+    {
+      "epoch": 0.8511329955062625,
+      "grad_norm": 5.1595633293913545,
+      "learning_rate": 1e-06,
+      "loss": 0.4614,
+      "step": 4451
+    },
+    {
+      "epoch": 0.8513242183765178,
+      "grad_norm": 4.310662361813854,
+      "learning_rate": 1e-06,
+      "loss": 0.7021,
+      "step": 4452
+    },
+    {
+      "epoch": 0.8515154412467731,
+      "grad_norm": 3.917075210258264,
+      "learning_rate": 1e-06,
+      "loss": 0.3964,
+      "step": 4453
+    },
+    {
+      "epoch": 0.8517066641170284,
+      "grad_norm": 7.8725882576204524,
+      "learning_rate": 1e-06,
+      "loss": 0.5535,
+      "step": 4454
+    },
+    {
+      "epoch": 0.8518978869872836,
+      "grad_norm": 3.824324983362311,
+      "learning_rate": 1e-06,
+      "loss": 0.2659,
+      "step": 4455
+    },
+    {
+      "epoch": 0.8520891098575389,
+      "grad_norm": 3.231858310955145,
+      "learning_rate": 1e-06,
+      "loss": 0.1146,
+      "step": 4456
+    },
+    {
+      "epoch": 0.8522803327277942,
+      "grad_norm": 4.946922100563437,
+      "learning_rate": 1e-06,
+      "loss": 0.336,
+      "step": 4457
+    },
+    {
+      "epoch": 0.8524715555980495,
+      "grad_norm": 2.6904936684202716,
+      "learning_rate": 1e-06,
+      "loss": 0.1808,
+      "step": 4458
+    },
+    {
+      "epoch": 0.8526627784683048,
+      "grad_norm": 1.8430207960308127,
+      "learning_rate": 1e-06,
+      "loss": 0.1683,
+      "step": 4459
+    },
+    {
+      "epoch": 0.85285400133856,
+      "grad_norm": 3.2413540791588153,
+      "learning_rate": 1e-06,
+      "loss": 0.1131,
+      "step": 4460
+    },
+    {
+      "epoch": 0.8530452242088153,
+      "grad_norm": 2.875099677970167,
+      "learning_rate": 1e-06,
+      "loss": 0.0841,
+      "step": 4461
+    },
+    {
+      "epoch": 0.8532364470790706,
+      "grad_norm": 2.519321829202347,
+      "learning_rate": 1e-06,
+      "loss": 0.0589,
+      "step": 4462
+    },
+    {
+      "epoch": 0.853427669949326,
+      "grad_norm": 8.283749991158329,
+      "learning_rate": 1e-06,
+      "loss": 0.2486,
+      "step": 4463
+    },
+    {
+      "epoch": 0.8536188928195813,
+      "grad_norm": 3.1755845500469784,
+      "learning_rate": 1e-06,
+      "loss": 0.287,
+      "step": 4464
+    },
+    {
+      "epoch": 0.8538101156898366,
+      "grad_norm": 3.655889248779402,
+      "learning_rate": 1e-06,
+      "loss": 0.1803,
+      "step": 4465
+    },
+    {
+      "epoch": 0.8540013385600917,
+      "grad_norm": 2.339068811442512,
+      "learning_rate": 1e-06,
+      "loss": 0.244,
+      "step": 4466
+    },
+    {
+      "epoch": 0.854192561430347,
+      "grad_norm": 2.997209363628338,
+      "learning_rate": 1e-06,
+      "loss": 0.2386,
+      "step": 4467
+    },
+    {
+      "epoch": 0.8543837843006024,
+      "grad_norm": 3.3160870401810385,
+      "learning_rate": 1e-06,
+      "loss": 0.3643,
+      "step": 4468
+    },
+    {
+      "epoch": 0.8545750071708577,
+      "grad_norm": 3.2632645102216085,
+      "learning_rate": 1e-06,
+      "loss": 0.2975,
+      "step": 4469
+    },
+    {
+      "epoch": 0.854766230041113,
+      "grad_norm": 4.069835437193971,
+      "learning_rate": 1e-06,
+      "loss": 0.2891,
+      "step": 4470
+    },
+    {
+      "epoch": 0.8549574529113682,
+      "grad_norm": 4.848665561162063,
+      "learning_rate": 1e-06,
+      "loss": 0.1196,
+      "step": 4471
+    },
+    {
+      "epoch": 0.8551486757816235,
+      "grad_norm": 3.7957915341825186,
+      "learning_rate": 1e-06,
+      "loss": 0.1465,
+      "step": 4472
+    },
+    {
+      "epoch": 0.8553398986518788,
+      "grad_norm": 3.1162793359091,
+      "learning_rate": 1e-06,
+      "loss": 0.1702,
+      "step": 4473
+    },
+    {
+      "epoch": 0.8555311215221341,
+      "grad_norm": 3.53336480204646,
+      "learning_rate": 1e-06,
+      "loss": 0.0958,
+      "step": 4474
+    },
+    {
+      "epoch": 0.8557223443923894,
+      "grad_norm": 6.009079103926498,
+      "learning_rate": 1e-06,
+      "loss": 0.1261,
+      "step": 4475
+    },
+    {
+      "epoch": 0.8559135672626446,
+      "grad_norm": 7.367403417600789,
+      "learning_rate": 1e-06,
+      "loss": 0.3776,
+      "step": 4476
+    },
+    {
+      "epoch": 0.8561047901328999,
+      "grad_norm": 5.2399699357313265,
+      "learning_rate": 1e-06,
+      "loss": 0.5846,
+      "step": 4477
+    },
+    {
+      "epoch": 0.8562960130031552,
+      "grad_norm": 3.413084680123159,
+      "learning_rate": 1e-06,
+      "loss": 0.2262,
+      "step": 4478
+    },
+    {
+      "epoch": 0.8564872358734105,
+      "grad_norm": 3.7722667828791456,
+      "learning_rate": 1e-06,
+      "loss": 0.2954,
+      "step": 4479
+    },
+    {
+      "epoch": 0.8566784587436658,
+      "grad_norm": 3.5753772296485793,
+      "learning_rate": 1e-06,
+      "loss": 0.1737,
+      "step": 4480
+    },
+    {
+      "epoch": 0.8568696816139211,
+      "grad_norm": 3.3811514566837535,
+      "learning_rate": 1e-06,
+      "loss": 0.1099,
+      "step": 4481
+    },
+    {
+      "epoch": 0.8570609044841763,
+      "grad_norm": 2.0522210580014377,
+      "learning_rate": 1e-06,
+      "loss": 0.0941,
+      "step": 4482
+    },
+    {
+      "epoch": 0.8572521273544316,
+      "grad_norm": 2.1567759218012363,
+      "learning_rate": 1e-06,
+      "loss": 0.0709,
+      "step": 4483
+    },
+    {
+      "epoch": 0.8574433502246869,
+      "grad_norm": 2.4906671843552033,
+      "learning_rate": 1e-06,
+      "loss": 0.0592,
+      "step": 4484
+    },
+    {
+      "epoch": 0.8576345730949422,
+      "grad_norm": 2.81938684547691,
+      "learning_rate": 1e-06,
+      "loss": 0.0857,
+      "step": 4485
+    },
+    {
+      "epoch": 0.8578257959651975,
+      "grad_norm": 3.1164771778893514,
+      "learning_rate": 1e-06,
+      "loss": 0.0788,
+      "step": 4486
+    },
+    {
+      "epoch": 0.8580170188354527,
+      "grad_norm": 3.623312195473796,
+      "learning_rate": 1e-06,
+      "loss": 0.1331,
+      "step": 4487
+    },
+    {
+      "epoch": 0.858208241705708,
+      "grad_norm": 3.6359593697723374,
+      "learning_rate": 1e-06,
+      "loss": 0.2386,
+      "step": 4488
+    },
+    {
+      "epoch": 0.8583994645759633,
+      "grad_norm": 3.656323684862812,
+      "learning_rate": 1e-06,
+      "loss": 0.2467,
+      "step": 4489
+    },
+    {
+      "epoch": 0.8585906874462186,
+      "grad_norm": 3.1499826884929347,
+      "learning_rate": 1e-06,
+      "loss": 0.1184,
+      "step": 4490
+    },
+    {
+      "epoch": 0.8587819103164739,
+      "grad_norm": 2.2533508356908154,
+      "learning_rate": 1e-06,
+      "loss": 0.1138,
+      "step": 4491
+    },
+    {
+      "epoch": 0.8589731331867292,
+      "grad_norm": 2.847957307366977,
+      "learning_rate": 1e-06,
+      "loss": 0.1518,
+      "step": 4492
+    },
+    {
+      "epoch": 0.8591643560569844,
+      "grad_norm": 2.173173145970975,
+      "learning_rate": 1e-06,
+      "loss": 0.2146,
+      "step": 4493
+    },
+    {
+      "epoch": 0.8593555789272397,
+      "grad_norm": 3.796515930528199,
+      "learning_rate": 1e-06,
+      "loss": 0.4675,
+      "step": 4494
+    },
+    {
+      "epoch": 0.859546801797495,
+      "grad_norm": 2.3136810173744116,
+      "learning_rate": 1e-06,
+      "loss": 0.1071,
+      "step": 4495
+    },
+    {
+      "epoch": 0.8597380246677503,
+      "grad_norm": 3.71068307104767,
+      "learning_rate": 1e-06,
+      "loss": 0.2381,
+      "step": 4496
+    },
+    {
+      "epoch": 0.8599292475380056,
+      "grad_norm": 5.11243871824662,
+      "learning_rate": 1e-06,
+      "loss": 0.208,
+      "step": 4497
+    },
+    {
+      "epoch": 0.8601204704082608,
+      "grad_norm": 4.015370876717105,
+      "learning_rate": 1e-06,
+      "loss": 0.3657,
+      "step": 4498
+    },
+    {
+      "epoch": 0.8603116932785161,
+      "grad_norm": 5.664466917135474,
+      "learning_rate": 1e-06,
+      "loss": 0.3118,
+      "step": 4499
+    },
+    {
+      "epoch": 0.8605029161487714,
+      "grad_norm": 4.52091590817654,
+      "learning_rate": 1e-06,
+      "loss": 0.1433,
+      "step": 4500
+    },
+    {
+      "epoch": 0.8605029161487714,
+      "eval_runtime": 740.2315,
+      "eval_samples_per_second": 2.072,
+      "eval_steps_per_second": 0.519,
+      "step": 4500
+    },
+    {
+      "epoch": 0.8606941390190267,
+      "grad_norm": 3.8129000844682754,
+      "learning_rate": 1e-06,
+      "loss": 0.6294,
+      "step": 4501
+    },
+    {
+      "epoch": 0.860885361889282,
+      "grad_norm": 6.646049025651796,
+      "learning_rate": 1e-06,
+      "loss": 0.4869,
+      "step": 4502
+    },
+    {
+      "epoch": 0.8610765847595372,
+      "grad_norm": 5.53208908354664,
+      "learning_rate": 1e-06,
+      "loss": 0.3487,
+      "step": 4503
+    },
+    {
+      "epoch": 0.8612678076297925,
+      "grad_norm": 7.276484030672562,
+      "learning_rate": 1e-06,
+      "loss": 0.3109,
+      "step": 4504
+    },
+    {
+      "epoch": 0.8614590305000478,
+      "grad_norm": 5.506341919147993,
+      "learning_rate": 1e-06,
+      "loss": 0.4174,
+      "step": 4505
+    },
+    {
+      "epoch": 0.8616502533703031,
+      "grad_norm": 2.7905915500733567,
+      "learning_rate": 1e-06,
+      "loss": 0.1306,
+      "step": 4506
+    },
+    {
+      "epoch": 0.8618414762405584,
+      "grad_norm": 5.019473111895802,
+      "learning_rate": 1e-06,
+      "loss": 0.2481,
+      "step": 4507
+    },
+    {
+      "epoch": 0.8620326991108137,
+      "grad_norm": 2.589224210487866,
+      "learning_rate": 1e-06,
+      "loss": 0.0646,
+      "step": 4508
+    },
+    {
+      "epoch": 0.8622239219810689,
+      "grad_norm": 3.777711079201903,
+      "learning_rate": 1e-06,
+      "loss": 0.0888,
+      "step": 4509
+    },
+    {
+      "epoch": 0.8624151448513242,
+      "grad_norm": 3.0851703704333104,
+      "learning_rate": 1e-06,
+      "loss": 0.0884,
+      "step": 4510
+    },
+    {
+      "epoch": 0.8626063677215795,
+      "grad_norm": 2.8262440870612076,
+      "learning_rate": 1e-06,
+      "loss": 0.0658,
+      "step": 4511
+    },
+    {
+      "epoch": 0.8627975905918348,
+      "grad_norm": 4.218116769244244,
+      "learning_rate": 1e-06,
+      "loss": 0.0863,
+      "step": 4512
+    },
+    {
+      "epoch": 0.8629888134620901,
+      "grad_norm": 4.068166448933178,
+      "learning_rate": 1e-06,
+      "loss": 0.2247,
+      "step": 4513
+    },
+    {
+      "epoch": 0.8631800363323453,
+      "grad_norm": 5.28652059083055,
+      "learning_rate": 1e-06,
+      "loss": 0.5513,
+      "step": 4514
+    },
+    {
+      "epoch": 0.8633712592026006,
+      "grad_norm": 3.9913808227852123,
+      "learning_rate": 1e-06,
+      "loss": 0.189,
+      "step": 4515
+    },
+    {
+      "epoch": 0.8635624820728559,
+      "grad_norm": 3.0112375550351396,
+      "learning_rate": 1e-06,
+      "loss": 0.2089,
+      "step": 4516
+    },
+    {
+      "epoch": 0.8637537049431112,
+      "grad_norm": 3.4739626646610615,
+      "learning_rate": 1e-06,
+      "loss": 0.3417,
+      "step": 4517
+    },
+    {
+      "epoch": 0.8639449278133665,
+      "grad_norm": 3.882968461955117,
+      "learning_rate": 1e-06,
+      "loss": 0.2188,
+      "step": 4518
+    },
+    {
+      "epoch": 0.8641361506836217,
+      "grad_norm": 2.259213128629366,
+      "learning_rate": 1e-06,
+      "loss": 0.1414,
+      "step": 4519
+    },
+    {
+      "epoch": 0.864327373553877,
+      "grad_norm": 1.8753714511417972,
+      "learning_rate": 1e-06,
+      "loss": 0.1367,
+      "step": 4520
+    },
+    {
+      "epoch": 0.8645185964241323,
+      "grad_norm": 3.00822798596442,
+      "learning_rate": 1e-06,
+      "loss": 0.0702,
+      "step": 4521
+    },
+    {
+      "epoch": 0.8647098192943876,
+      "grad_norm": 3.248334311075261,
+      "learning_rate": 1e-06,
+      "loss": 0.1534,
+      "step": 4522
+    },
+    {
+      "epoch": 0.8649010421646429,
+      "grad_norm": 4.314058326351657,
+      "learning_rate": 1e-06,
+      "loss": 0.1108,
+      "step": 4523
+    },
+    {
+      "epoch": 0.8650922650348982,
+      "grad_norm": 2.4847173874719575,
+      "learning_rate": 1e-06,
+      "loss": 0.1442,
+      "step": 4524
+    },
+    {
+      "epoch": 0.8652834879051534,
+      "grad_norm": 10.07416346558115,
+      "learning_rate": 1e-06,
+      "loss": 0.2403,
+      "step": 4525
+    },
+    {
+      "epoch": 0.8654747107754087,
+      "grad_norm": 7.310902999858488,
+      "learning_rate": 1e-06,
+      "loss": 0.2615,
+      "step": 4526
+    },
+    {
+      "epoch": 0.865665933645664,
+      "grad_norm": 3.9846737020386938,
+      "learning_rate": 1e-06,
+      "loss": 0.2641,
+      "step": 4527
+    },
+    {
+      "epoch": 0.8658571565159193,
+      "grad_norm": 6.2839595587173935,
+      "learning_rate": 1e-06,
+      "loss": 0.6039,
+      "step": 4528
+    },
+    {
+      "epoch": 0.8660483793861746,
+      "grad_norm": 3.030457383183922,
+      "learning_rate": 1e-06,
+      "loss": 0.5185,
+      "step": 4529
+    },
+    {
+      "epoch": 0.8662396022564298,
+      "grad_norm": 2.8017806181539595,
+      "learning_rate": 1e-06,
+      "loss": 0.1514,
+      "step": 4530
+    },
+    {
+      "epoch": 0.8664308251266851,
+      "grad_norm": 3.5343545415608295,
+      "learning_rate": 1e-06,
+      "loss": 0.3157,
+      "step": 4531
+    },
+    {
+      "epoch": 0.8666220479969404,
+      "grad_norm": 5.897082121283883,
+      "learning_rate": 1e-06,
+      "loss": 0.3133,
+      "step": 4532
+    },
+    {
+      "epoch": 0.8668132708671957,
+      "grad_norm": 5.476487446269324,
+      "learning_rate": 1e-06,
+      "loss": 0.3666,
+      "step": 4533
+    },
+    {
+      "epoch": 0.867004493737451,
+      "grad_norm": 2.3264039097644216,
+      "learning_rate": 1e-06,
+      "loss": 0.162,
+      "step": 4534
+    },
+    {
+      "epoch": 0.8671957166077063,
+      "grad_norm": 7.251160890439075,
+      "learning_rate": 1e-06,
+      "loss": 0.1306,
+      "step": 4535
+    },
+    {
+      "epoch": 0.8673869394779615,
+      "grad_norm": 2.208919399395302,
+      "learning_rate": 1e-06,
+      "loss": 0.0544,
+      "step": 4536
+    },
+    {
+      "epoch": 0.8675781623482168,
+      "grad_norm": 2.6438865608314988,
+      "learning_rate": 1e-06,
+      "loss": 0.0987,
+      "step": 4537
+    },
+    {
+      "epoch": 0.8677693852184721,
+      "grad_norm": 4.93022546991556,
+      "learning_rate": 1e-06,
+      "loss": 0.1424,
+      "step": 4538
+    },
+    {
+      "epoch": 0.8679606080887274,
+      "grad_norm": 4.407647094003039,
+      "learning_rate": 1e-06,
+      "loss": 0.3132,
+      "step": 4539
+    },
+    {
+      "epoch": 0.8681518309589827,
+      "grad_norm": 4.012095046479745,
+      "learning_rate": 1e-06,
+      "loss": 0.2829,
+      "step": 4540
+    },
+    {
+      "epoch": 0.8683430538292379,
+      "grad_norm": 2.6567393132722907,
+      "learning_rate": 1e-06,
+      "loss": 0.2763,
+      "step": 4541
+    },
+    {
+      "epoch": 0.8685342766994932,
+      "grad_norm": 2.7696695871601356,
+      "learning_rate": 1e-06,
+      "loss": 0.1385,
+      "step": 4542
+    },
+    {
+      "epoch": 0.8687254995697485,
+      "grad_norm": 7.491218766800269,
+      "learning_rate": 1e-06,
+      "loss": 0.2274,
+      "step": 4543
+    },
+    {
+      "epoch": 0.8689167224400038,
+      "grad_norm": 6.542791800312488,
+      "learning_rate": 1e-06,
+      "loss": 0.2541,
+      "step": 4544
+    },
+    {
+      "epoch": 0.8691079453102591,
+      "grad_norm": 5.436498264688218,
+      "learning_rate": 1e-06,
+      "loss": 0.1401,
+      "step": 4545
+    },
+    {
+      "epoch": 0.8692991681805143,
+      "grad_norm": 3.689611444665937,
+      "learning_rate": 1e-06,
+      "loss": 0.2903,
+      "step": 4546
+    },
+    {
+      "epoch": 0.8694903910507696,
+      "grad_norm": 2.8030394065827826,
+      "learning_rate": 1e-06,
+      "loss": 0.1241,
+      "step": 4547
+    },
+    {
+      "epoch": 0.8696816139210249,
+      "grad_norm": 2.1926305278344995,
+      "learning_rate": 1e-06,
+      "loss": 0.0673,
+      "step": 4548
+    },
+    {
+      "epoch": 0.8698728367912802,
+      "grad_norm": 3.358362657467878,
+      "learning_rate": 1e-06,
+      "loss": 0.0737,
+      "step": 4549
+    },
+    {
+      "epoch": 0.8700640596615355,
+      "grad_norm": 11.403778895423356,
+      "learning_rate": 1e-06,
+      "loss": 0.1566,
+      "step": 4550
+    },
+    {
+      "epoch": 0.8702552825317909,
+      "grad_norm": 4.155201256166547,
+      "learning_rate": 1e-06,
+      "loss": 0.352,
+      "step": 4551
+    },
+    {
+      "epoch": 0.870446505402046,
+      "grad_norm": 3.440508548938279,
+      "learning_rate": 1e-06,
+      "loss": 0.4484,
+      "step": 4552
+    },
+    {
+      "epoch": 0.8706377282723013,
+      "grad_norm": 5.151381349950911,
+      "learning_rate": 1e-06,
+      "loss": 0.6452,
+      "step": 4553
+    },
+    {
+      "epoch": 0.8708289511425567,
+      "grad_norm": 2.4922962703108005,
+      "learning_rate": 1e-06,
+      "loss": 0.171,
+      "step": 4554
+    },
+    {
+      "epoch": 0.871020174012812,
+      "grad_norm": 2.733538603246838,
+      "learning_rate": 1e-06,
+      "loss": 0.1929,
+      "step": 4555
+    },
+    {
+      "epoch": 0.8712113968830673,
+      "grad_norm": 2.332611063023129,
+      "learning_rate": 1e-06,
+      "loss": 0.1169,
+      "step": 4556
+    },
+    {
+      "epoch": 0.8714026197533224,
+      "grad_norm": 3.0562552797236795,
+      "learning_rate": 1e-06,
+      "loss": 0.2705,
+      "step": 4557
+    },
+    {
+      "epoch": 0.8715938426235778,
+      "grad_norm": 1.9834750680474398,
+      "learning_rate": 1e-06,
+      "loss": 0.0727,
+      "step": 4558
+    },
+    {
+      "epoch": 0.8717850654938331,
+      "grad_norm": 3.0969941055690113,
+      "learning_rate": 1e-06,
+      "loss": 0.1979,
+      "step": 4559
+    },
+    {
+      "epoch": 0.8719762883640884,
+      "grad_norm": 4.549096053321335,
+      "learning_rate": 1e-06,
+      "loss": 0.107,
+      "step": 4560
+    },
+    {
+      "epoch": 0.8721675112343437,
+      "grad_norm": 3.757257305538439,
+      "learning_rate": 1e-06,
+      "loss": 0.2258,
+      "step": 4561
+    },
+    {
+      "epoch": 0.872358734104599,
+      "grad_norm": 2.6199995278219714,
+      "learning_rate": 1e-06,
+      "loss": 0.0729,
+      "step": 4562
+    },
+    {
+      "epoch": 0.8725499569748542,
+      "grad_norm": 3.5015452924558366,
+      "learning_rate": 1e-06,
+      "loss": 0.1813,
+      "step": 4563
+    },
+    {
+      "epoch": 0.8727411798451095,
+      "grad_norm": 6.221628107741185,
+      "learning_rate": 1e-06,
+      "loss": 0.4207,
+      "step": 4564
+    },
+    {
+      "epoch": 0.8729324027153648,
+      "grad_norm": 2.7612814550304834,
+      "learning_rate": 1e-06,
+      "loss": 0.3177,
+      "step": 4565
+    },
+    {
+      "epoch": 0.8731236255856201,
+      "grad_norm": 3.4694217340767723,
+      "learning_rate": 1e-06,
+      "loss": 0.2023,
+      "step": 4566
+    },
+    {
+      "epoch": 0.8733148484558754,
+      "grad_norm": 2.441524704157622,
+      "learning_rate": 1e-06,
+      "loss": 0.0865,
+      "step": 4567
+    },
+    {
+      "epoch": 0.8735060713261306,
+      "grad_norm": 3.2065732677707572,
+      "learning_rate": 1e-06,
+      "loss": 0.1661,
+      "step": 4568
+    },
+    {
+      "epoch": 0.8736972941963859,
+      "grad_norm": 3.6760650752204946,
+      "learning_rate": 1e-06,
+      "loss": 0.2424,
+      "step": 4569
+    },
+    {
+      "epoch": 0.8738885170666412,
+      "grad_norm": 4.431473198826104,
+      "learning_rate": 1e-06,
+      "loss": 0.3379,
+      "step": 4570
+    },
+    {
+      "epoch": 0.8740797399368965,
+      "grad_norm": 1.7785486182644874,
+      "learning_rate": 1e-06,
+      "loss": 0.0831,
+      "step": 4571
+    },
+    {
+      "epoch": 0.8742709628071518,
+      "grad_norm": 5.87718549619539,
+      "learning_rate": 1e-06,
+      "loss": 0.3408,
+      "step": 4572
+    },
+    {
+      "epoch": 0.874462185677407,
+      "grad_norm": 2.133099882445731,
+      "learning_rate": 1e-06,
+      "loss": 0.097,
+      "step": 4573
+    },
+    {
+      "epoch": 0.8746534085476623,
+      "grad_norm": 5.688603556120006,
+      "learning_rate": 1e-06,
+      "loss": 0.0756,
+      "step": 4574
+    },
+    {
+      "epoch": 0.8748446314179176,
+      "grad_norm": 8.329808621446347,
+      "learning_rate": 1e-06,
+      "loss": 0.1498,
+      "step": 4575
+    },
+    {
+      "epoch": 0.8750358542881729,
+      "grad_norm": 6.138698596028385,
+      "learning_rate": 1e-06,
+      "loss": 0.3289,
+      "step": 4576
+    },
+    {
+      "epoch": 0.8752270771584282,
+      "grad_norm": 3.20924767350275,
+      "learning_rate": 1e-06,
+      "loss": 0.5044,
+      "step": 4577
+    },
+    {
+      "epoch": 0.8754183000286835,
+      "grad_norm": 4.292991853327278,
+      "learning_rate": 1e-06,
+      "loss": 0.2372,
+      "step": 4578
+    },
+    {
+      "epoch": 0.8756095228989387,
+      "grad_norm": 3.444842040450445,
+      "learning_rate": 1e-06,
+      "loss": 0.1849,
+      "step": 4579
+    },
+    {
+      "epoch": 0.875800745769194,
+      "grad_norm": 1.5501948480267747,
+      "learning_rate": 1e-06,
+      "loss": 0.0492,
+      "step": 4580
+    },
+    {
+      "epoch": 0.8759919686394493,
+      "grad_norm": 2.8155892254486274,
+      "learning_rate": 1e-06,
+      "loss": 0.2358,
+      "step": 4581
+    },
+    {
+      "epoch": 0.8761831915097046,
+      "grad_norm": 2.800348522430095,
+      "learning_rate": 1e-06,
+      "loss": 0.1309,
+      "step": 4582
+    },
+    {
+      "epoch": 0.8763744143799599,
+      "grad_norm": 1.5886432886189779,
+      "learning_rate": 1e-06,
+      "loss": 0.1153,
+      "step": 4583
+    },
+    {
+      "epoch": 0.8765656372502151,
+      "grad_norm": 2.7318986414905533,
+      "learning_rate": 1e-06,
+      "loss": 0.1656,
+      "step": 4584
+    },
+    {
+      "epoch": 0.8767568601204704,
+      "grad_norm": 2.4507035860567523,
+      "learning_rate": 1e-06,
+      "loss": 0.125,
+      "step": 4585
+    },
+    {
+      "epoch": 0.8769480829907257,
+      "grad_norm": 3.5661189119171426,
+      "learning_rate": 1e-06,
+      "loss": 0.0927,
+      "step": 4586
+    },
+    {
+      "epoch": 0.877139305860981,
+      "grad_norm": 6.620817483635335,
+      "learning_rate": 1e-06,
+      "loss": 0.115,
+      "step": 4587
+    },
+    {
+      "epoch": 0.8773305287312363,
+      "grad_norm": 5.019341443699799,
+      "learning_rate": 1e-06,
+      "loss": 0.3508,
+      "step": 4588
+    },
+    {
+      "epoch": 0.8775217516014915,
+      "grad_norm": 5.163426439411609,
+      "learning_rate": 1e-06,
+      "loss": 0.3256,
+      "step": 4589
+    },
+    {
+      "epoch": 0.8777129744717468,
+      "grad_norm": 3.6179506885042314,
+      "learning_rate": 1e-06,
+      "loss": 0.2041,
+      "step": 4590
+    },
+    {
+      "epoch": 0.8779041973420021,
+      "grad_norm": 4.413335635673484,
+      "learning_rate": 1e-06,
+      "loss": 0.4161,
+      "step": 4591
+    },
+    {
+      "epoch": 0.8780954202122574,
+      "grad_norm": 4.923497686378398,
+      "learning_rate": 1e-06,
+      "loss": 0.1461,
+      "step": 4592
+    },
+    {
+      "epoch": 0.8782866430825127,
+      "grad_norm": 5.631302926241379,
+      "learning_rate": 1e-06,
+      "loss": 0.3009,
+      "step": 4593
+    },
+    {
+      "epoch": 0.878477865952768,
+      "grad_norm": 2.6060699730040087,
+      "learning_rate": 1e-06,
+      "loss": 0.1152,
+      "step": 4594
+    },
+    {
+      "epoch": 0.8786690888230232,
+      "grad_norm": 1.868055197878413,
+      "learning_rate": 1e-06,
+      "loss": 0.1278,
+      "step": 4595
+    },
+    {
+      "epoch": 0.8788603116932785,
+      "grad_norm": 16.907704853067358,
+      "learning_rate": 1e-06,
+      "loss": 0.1793,
+      "step": 4596
+    },
+    {
+      "epoch": 0.8790515345635338,
+      "grad_norm": 2.60646552788737,
+      "learning_rate": 1e-06,
+      "loss": 0.0838,
+      "step": 4597
+    },
+    {
+      "epoch": 0.8792427574337891,
+      "grad_norm": 2.7743240873576966,
+      "learning_rate": 1e-06,
+      "loss": 0.0878,
+      "step": 4598
+    },
+    {
+      "epoch": 0.8794339803040444,
+      "grad_norm": 3.130161904970822,
+      "learning_rate": 1e-06,
+      "loss": 0.0624,
+      "step": 4599
+    },
+    {
+      "epoch": 0.8796252031742996,
+      "grad_norm": 7.9030033728191595,
+      "learning_rate": 1e-06,
+      "loss": 0.1118,
+      "step": 4600
+    },
+    {
+      "epoch": 0.8798164260445549,
+      "grad_norm": 6.723898826550523,
+      "learning_rate": 1e-06,
+      "loss": 0.5424,
+      "step": 4601
+    },
+    {
+      "epoch": 0.8800076489148102,
+      "grad_norm": 5.314520418020762,
+      "learning_rate": 1e-06,
+      "loss": 0.6778,
+      "step": 4602
+    },
+    {
+      "epoch": 0.8801988717850655,
+      "grad_norm": 6.286215873621079,
+      "learning_rate": 1e-06,
+      "loss": 0.6467,
+      "step": 4603
+    },
+    {
+      "epoch": 0.8803900946553208,
+      "grad_norm": 3.052020926157554,
+      "learning_rate": 1e-06,
+      "loss": 0.1809,
+      "step": 4604
+    },
+    {
+      "epoch": 0.8805813175255761,
+      "grad_norm": 3.132121250061282,
+      "learning_rate": 1e-06,
+      "loss": 0.3014,
+      "step": 4605
+    },
+    {
+      "epoch": 0.8807725403958313,
+      "grad_norm": 2.294989090604762,
+      "learning_rate": 1e-06,
+      "loss": 0.1469,
+      "step": 4606
+    },
+    {
+      "epoch": 0.8809637632660866,
+      "grad_norm": 5.7168866232919315,
+      "learning_rate": 1e-06,
+      "loss": 0.2211,
+      "step": 4607
+    },
+    {
+      "epoch": 0.8811549861363419,
+      "grad_norm": 2.635073271280996,
+      "learning_rate": 1e-06,
+      "loss": 0.1214,
+      "step": 4608
+    },
+    {
+      "epoch": 0.8813462090065972,
+      "grad_norm": 4.695735360144326,
+      "learning_rate": 1e-06,
+      "loss": 0.1282,
+      "step": 4609
+    },
+    {
+      "epoch": 0.8815374318768525,
+      "grad_norm": 3.5090192660359705,
+      "learning_rate": 1e-06,
+      "loss": 0.0522,
+      "step": 4610
+    },
+    {
+      "epoch": 0.8817286547471077,
+      "grad_norm": 4.03540959998467,
+      "learning_rate": 1e-06,
+      "loss": 0.1128,
+      "step": 4611
+    },
+    {
+      "epoch": 0.881919877617363,
+      "grad_norm": 2.6822135741090807,
+      "learning_rate": 1e-06,
+      "loss": 0.0805,
+      "step": 4612
+    },
+    {
+      "epoch": 0.8821111004876183,
+      "grad_norm": 4.318607594064743,
+      "learning_rate": 1e-06,
+      "loss": 0.1743,
+      "step": 4613
+    },
+    {
+      "epoch": 0.8823023233578736,
+      "grad_norm": 4.753745961053393,
+      "learning_rate": 1e-06,
+      "loss": 0.3387,
+      "step": 4614
+    },
+    {
+      "epoch": 0.8824935462281289,
+      "grad_norm": 3.448412134503306,
+      "learning_rate": 1e-06,
+      "loss": 0.4789,
+      "step": 4615
+    },
+    {
+      "epoch": 0.8826847690983841,
+      "grad_norm": 2.7574807634832603,
+      "learning_rate": 1e-06,
+      "loss": 0.1738,
+      "step": 4616
+    },
+    {
+      "epoch": 0.8828759919686394,
+      "grad_norm": 3.0582788142057815,
+      "learning_rate": 1e-06,
+      "loss": 0.182,
+      "step": 4617
+    },
+    {
+      "epoch": 0.8830672148388947,
+      "grad_norm": 1.8637316657225675,
+      "learning_rate": 1e-06,
+      "loss": 0.0727,
+      "step": 4618
+    },
+    {
+      "epoch": 0.88325843770915,
+      "grad_norm": 4.630965432344674,
+      "learning_rate": 1e-06,
+      "loss": 0.4249,
+      "step": 4619
+    },
+    {
+      "epoch": 0.8834496605794053,
+      "grad_norm": 3.818333399847711,
+      "learning_rate": 1e-06,
+      "loss": 0.2533,
+      "step": 4620
+    },
+    {
+      "epoch": 0.8836408834496606,
+      "grad_norm": 4.94252974814308,
+      "learning_rate": 1e-06,
+      "loss": 0.1185,
+      "step": 4621
+    },
+    {
+      "epoch": 0.8838321063199158,
+      "grad_norm": 2.5527664100837555,
+      "learning_rate": 1e-06,
+      "loss": 0.0643,
+      "step": 4622
+    },
+    {
+      "epoch": 0.8840233291901711,
+      "grad_norm": 1.7125582051653314,
+      "learning_rate": 1e-06,
+      "loss": 0.0963,
+      "step": 4623
+    },
+    {
+      "epoch": 0.8842145520604264,
+      "grad_norm": 3.5110733067077486,
+      "learning_rate": 1e-06,
+      "loss": 0.0651,
+      "step": 4624
+    },
+    {
+      "epoch": 0.8844057749306817,
+      "grad_norm": 3.8571657351668107,
+      "learning_rate": 1e-06,
+      "loss": 0.129,
+      "step": 4625
+    },
+    {
+      "epoch": 0.884596997800937,
+      "grad_norm": 5.24794274849018,
+      "learning_rate": 1e-06,
+      "loss": 0.2719,
+      "step": 4626
+    },
+    {
+      "epoch": 0.8847882206711922,
+      "grad_norm": 4.345531729415002,
+      "learning_rate": 1e-06,
+      "loss": 0.4991,
+      "step": 4627
+    },
+    {
+      "epoch": 0.8849794435414475,
+      "grad_norm": 5.213594002576922,
+      "learning_rate": 1e-06,
+      "loss": 0.2939,
+      "step": 4628
+    },
+    {
+      "epoch": 0.8851706664117028,
+      "grad_norm": 2.315403739046925,
+      "learning_rate": 1e-06,
+      "loss": 0.1472,
+      "step": 4629
+    },
+    {
+      "epoch": 0.8853618892819581,
+      "grad_norm": 3.0418222640141304,
+      "learning_rate": 1e-06,
+      "loss": 0.2796,
+      "step": 4630
+    },
+    {
+      "epoch": 0.8855531121522134,
+      "grad_norm": 1.5348807905547548,
+      "learning_rate": 1e-06,
+      "loss": 0.1665,
+      "step": 4631
+    },
+    {
+      "epoch": 0.8857443350224686,
+      "grad_norm": 3.6510971144938518,
+      "learning_rate": 1e-06,
+      "loss": 0.143,
+      "step": 4632
+    },
+    {
+      "epoch": 0.8859355578927239,
+      "grad_norm": 3.639141124172442,
+      "learning_rate": 1e-06,
+      "loss": 0.208,
+      "step": 4633
+    },
+    {
+      "epoch": 0.8861267807629792,
+      "grad_norm": 3.070834593511674,
+      "learning_rate": 1e-06,
+      "loss": 0.0835,
+      "step": 4634
+    },
+    {
+      "epoch": 0.8863180036332345,
+      "grad_norm": 4.544768904212864,
+      "learning_rate": 1e-06,
+      "loss": 0.0616,
+      "step": 4635
+    },
+    {
+      "epoch": 0.8865092265034898,
+      "grad_norm": 2.3576738907081185,
+      "learning_rate": 1e-06,
+      "loss": 0.0536,
+      "step": 4636
+    },
+    {
+      "epoch": 0.8867004493737451,
+      "grad_norm": 5.091302763592675,
+      "learning_rate": 1e-06,
+      "loss": 0.1807,
+      "step": 4637
+    },
+    {
+      "epoch": 0.8868916722440003,
+      "grad_norm": 5.526725685961022,
+      "learning_rate": 1e-06,
+      "loss": 0.2765,
+      "step": 4638
+    },
+    {
+      "epoch": 0.8870828951142556,
+      "grad_norm": 4.518964436826078,
+      "learning_rate": 1e-06,
+      "loss": 0.3939,
+      "step": 4639
+    },
+    {
+      "epoch": 0.887274117984511,
+      "grad_norm": 3.1321115066244367,
+      "learning_rate": 1e-06,
+      "loss": 0.258,
+      "step": 4640
+    },
+    {
+      "epoch": 0.8874653408547662,
+      "grad_norm": 3.9038303052574013,
+      "learning_rate": 1e-06,
+      "loss": 0.1803,
+      "step": 4641
+    },
+    {
+      "epoch": 0.8876565637250216,
+      "grad_norm": 1.523552679940765,
+      "learning_rate": 1e-06,
+      "loss": 0.0411,
+      "step": 4642
+    },
+    {
+      "epoch": 0.8878477865952767,
+      "grad_norm": 2.4216202140425676,
+      "learning_rate": 1e-06,
+      "loss": 0.0902,
+      "step": 4643
+    },
+    {
+      "epoch": 0.888039009465532,
+      "grad_norm": 3.8189471404639415,
+      "learning_rate": 1e-06,
+      "loss": 0.134,
+      "step": 4644
+    },
+    {
+      "epoch": 0.8882302323357874,
+      "grad_norm": 4.486250322437035,
+      "learning_rate": 1e-06,
+      "loss": 0.296,
+      "step": 4645
+    },
+    {
+      "epoch": 0.8884214552060427,
+      "grad_norm": 3.6579537619812093,
+      "learning_rate": 1e-06,
+      "loss": 0.2355,
+      "step": 4646
+    },
+    {
+      "epoch": 0.888612678076298,
+      "grad_norm": 3.2013126363290447,
+      "learning_rate": 1e-06,
+      "loss": 0.0657,
+      "step": 4647
+    },
+    {
+      "epoch": 0.8888039009465533,
+      "grad_norm": 2.651945058889311,
+      "learning_rate": 1e-06,
+      "loss": 0.0566,
+      "step": 4648
+    },
+    {
+      "epoch": 0.8889951238168085,
+      "grad_norm": 3.3069620443772347,
+      "learning_rate": 1e-06,
+      "loss": 0.0913,
+      "step": 4649
+    },
+    {
+      "epoch": 0.8891863466870638,
+      "grad_norm": 5.975530956730435,
+      "learning_rate": 1e-06,
+      "loss": 0.0826,
+      "step": 4650
+    },
+    {
+      "epoch": 0.8893775695573191,
+      "grad_norm": 4.608432686386031,
+      "learning_rate": 1e-06,
+      "loss": 0.2832,
+      "step": 4651
+    },
+    {
+      "epoch": 0.8895687924275744,
+      "grad_norm": 4.253365025975252,
+      "learning_rate": 1e-06,
+      "loss": 0.1559,
+      "step": 4652
+    },
+    {
+      "epoch": 0.8897600152978297,
+      "grad_norm": 5.756218243075401,
+      "learning_rate": 1e-06,
+      "loss": 0.4796,
+      "step": 4653
+    },
+    {
+      "epoch": 0.8899512381680849,
+      "grad_norm": 5.881783971758032,
+      "learning_rate": 1e-06,
+      "loss": 0.34,
+      "step": 4654
+    },
+    {
+      "epoch": 0.8901424610383402,
+      "grad_norm": 3.152958428299672,
+      "learning_rate": 1e-06,
+      "loss": 0.218,
+      "step": 4655
+    },
+    {
+      "epoch": 0.8903336839085955,
+      "grad_norm": 4.250393344403941,
+      "learning_rate": 1e-06,
+      "loss": 0.1706,
+      "step": 4656
+    },
+    {
+      "epoch": 0.8905249067788508,
+      "grad_norm": 5.976222289132249,
+      "learning_rate": 1e-06,
+      "loss": 0.4155,
+      "step": 4657
+    },
+    {
+      "epoch": 0.8907161296491061,
+      "grad_norm": 4.498377189483691,
+      "learning_rate": 1e-06,
+      "loss": 0.1164,
+      "step": 4658
+    },
+    {
+      "epoch": 0.8909073525193613,
+      "grad_norm": 3.2288781970393243,
+      "learning_rate": 1e-06,
+      "loss": 0.1163,
+      "step": 4659
+    },
+    {
+      "epoch": 0.8910985753896166,
+      "grad_norm": 2.615095800100098,
+      "learning_rate": 1e-06,
+      "loss": 0.1059,
+      "step": 4660
+    },
+    {
+      "epoch": 0.8912897982598719,
+      "grad_norm": 3.090307921835268,
+      "learning_rate": 1e-06,
+      "loss": 0.2767,
+      "step": 4661
+    },
+    {
+      "epoch": 0.8914810211301272,
+      "grad_norm": 3.1613990405250356,
+      "learning_rate": 1e-06,
+      "loss": 0.0745,
+      "step": 4662
+    },
+    {
+      "epoch": 0.8916722440003825,
+      "grad_norm": 7.609498573255465,
+      "learning_rate": 1e-06,
+      "loss": 0.1154,
+      "step": 4663
+    },
+    {
+      "epoch": 0.8918634668706378,
+      "grad_norm": 8.118436539187845,
+      "learning_rate": 1e-06,
+      "loss": 0.4736,
+      "step": 4664
+    },
+    {
+      "epoch": 0.892054689740893,
+      "grad_norm": 4.151876801928684,
+      "learning_rate": 1e-06,
+      "loss": 0.2142,
+      "step": 4665
+    },
+    {
+      "epoch": 0.8922459126111483,
+      "grad_norm": 4.470721043542073,
+      "learning_rate": 1e-06,
+      "loss": 0.1649,
+      "step": 4666
+    },
+    {
+      "epoch": 0.8924371354814036,
+      "grad_norm": 4.723836822236784,
+      "learning_rate": 1e-06,
+      "loss": 0.4443,
+      "step": 4667
+    },
+    {
+      "epoch": 0.8926283583516589,
+      "grad_norm": 3.023234673530216,
+      "learning_rate": 1e-06,
+      "loss": 0.1046,
+      "step": 4668
+    },
+    {
+      "epoch": 0.8928195812219142,
+      "grad_norm": 2.7824375317631236,
+      "learning_rate": 1e-06,
+      "loss": 0.096,
+      "step": 4669
+    },
+    {
+      "epoch": 0.8930108040921694,
+      "grad_norm": 2.123261806694612,
+      "learning_rate": 1e-06,
+      "loss": 0.1491,
+      "step": 4670
+    },
+    {
+      "epoch": 0.8932020269624247,
+      "grad_norm": 3.1670865901459693,
+      "learning_rate": 1e-06,
+      "loss": 0.099,
+      "step": 4671
+    },
+    {
+      "epoch": 0.89339324983268,
+      "grad_norm": 1.9745631302154658,
+      "learning_rate": 1e-06,
+      "loss": 0.0297,
+      "step": 4672
+    },
+    {
+      "epoch": 0.8935844727029353,
+      "grad_norm": 2.198311599998061,
+      "learning_rate": 1e-06,
+      "loss": 0.0479,
+      "step": 4673
+    },
+    {
+      "epoch": 0.8937756955731906,
+      "grad_norm": 4.19883514781311,
+      "learning_rate": 1e-06,
+      "loss": 0.2287,
+      "step": 4674
+    },
+    {
+      "epoch": 0.8939669184434459,
+      "grad_norm": 6.793539557404341,
+      "learning_rate": 1e-06,
+      "loss": 0.1516,
+      "step": 4675
+    },
+    {
+      "epoch": 0.8941581413137011,
+      "grad_norm": 5.73422731630331,
+      "learning_rate": 1e-06,
+      "loss": 0.2545,
+      "step": 4676
+    },
+    {
+      "epoch": 0.8943493641839564,
+      "grad_norm": 4.340334860193209,
+      "learning_rate": 1e-06,
+      "loss": 0.6095,
+      "step": 4677
+    },
+    {
+      "epoch": 0.8945405870542117,
+      "grad_norm": 4.786421030236112,
+      "learning_rate": 1e-06,
+      "loss": 0.5843,
+      "step": 4678
+    },
+    {
+      "epoch": 0.894731809924467,
+      "grad_norm": 4.374095278245986,
+      "learning_rate": 1e-06,
+      "loss": 0.1741,
+      "step": 4679
+    },
+    {
+      "epoch": 0.8949230327947223,
+      "grad_norm": 4.331333995186292,
+      "learning_rate": 1e-06,
+      "loss": 0.1968,
+      "step": 4680
+    },
+    {
+      "epoch": 0.8951142556649775,
+      "grad_norm": 4.433370683746943,
+      "learning_rate": 1e-06,
+      "loss": 0.3565,
+      "step": 4681
+    },
+    {
+      "epoch": 0.8953054785352328,
+      "grad_norm": 3.4707834967568725,
+      "learning_rate": 1e-06,
+      "loss": 0.2445,
+      "step": 4682
+    },
+    {
+      "epoch": 0.8954967014054881,
+      "grad_norm": 2.9706536865133426,
+      "learning_rate": 1e-06,
+      "loss": 0.1499,
+      "step": 4683
+    },
+    {
+      "epoch": 0.8956879242757434,
+      "grad_norm": 1.9700995769604257,
+      "learning_rate": 1e-06,
+      "loss": 0.0487,
+      "step": 4684
+    },
+    {
+      "epoch": 0.8958791471459987,
+      "grad_norm": 2.2847296944302466,
+      "learning_rate": 1e-06,
+      "loss": 0.1098,
+      "step": 4685
+    },
+    {
+      "epoch": 0.8960703700162539,
+      "grad_norm": 2.4577590034420065,
+      "learning_rate": 1e-06,
+      "loss": 0.0822,
+      "step": 4686
+    },
+    {
+      "epoch": 0.8962615928865092,
+      "grad_norm": 3.464978616491614,
+      "learning_rate": 1e-06,
+      "loss": 0.0869,
+      "step": 4687
+    },
+    {
+      "epoch": 0.8964528157567645,
+      "grad_norm": 6.077856376615494,
+      "learning_rate": 1e-06,
+      "loss": 0.1781,
+      "step": 4688
+    },
+    {
+      "epoch": 0.8966440386270198,
+      "grad_norm": 5.210115885553176,
+      "learning_rate": 1e-06,
+      "loss": 0.6373,
+      "step": 4689
+    },
+    {
+      "epoch": 0.8968352614972751,
+      "grad_norm": 3.5333063670394926,
+      "learning_rate": 1e-06,
+      "loss": 0.1664,
+      "step": 4690
+    },
+    {
+      "epoch": 0.8970264843675304,
+      "grad_norm": 5.803397978821195,
+      "learning_rate": 1e-06,
+      "loss": 0.423,
+      "step": 4691
+    },
+    {
+      "epoch": 0.8972177072377856,
+      "grad_norm": 2.8661177715905954,
+      "learning_rate": 1e-06,
+      "loss": 0.2531,
+      "step": 4692
+    },
+    {
+      "epoch": 0.8974089301080409,
+      "grad_norm": 3.766621058252631,
+      "learning_rate": 1e-06,
+      "loss": 0.2741,
+      "step": 4693
+    },
+    {
+      "epoch": 0.8976001529782962,
+      "grad_norm": 4.174033531664775,
+      "learning_rate": 1e-06,
+      "loss": 0.2262,
+      "step": 4694
+    },
+    {
+      "epoch": 0.8977913758485515,
+      "grad_norm": 2.6992521380555323,
+      "learning_rate": 1e-06,
+      "loss": 0.1984,
+      "step": 4695
+    },
+    {
+      "epoch": 0.8979825987188068,
+      "grad_norm": 3.0473630856558382,
+      "learning_rate": 1e-06,
+      "loss": 0.1136,
+      "step": 4696
+    },
+    {
+      "epoch": 0.898173821589062,
+      "grad_norm": 2.7943548004437337,
+      "learning_rate": 1e-06,
+      "loss": 0.1164,
+      "step": 4697
+    },
+    {
+      "epoch": 0.8983650444593173,
+      "grad_norm": 4.924182750455927,
+      "learning_rate": 1e-06,
+      "loss": 0.0768,
+      "step": 4698
+    },
+    {
+      "epoch": 0.8985562673295726,
+      "grad_norm": 2.4807716965592013,
+      "learning_rate": 1e-06,
+      "loss": 0.0738,
+      "step": 4699
+    },
+    {
+      "epoch": 0.8987474901998279,
+      "grad_norm": 8.496112158744424,
+      "learning_rate": 1e-06,
+      "loss": 0.1659,
+      "step": 4700
+    },
+    {
+      "epoch": 0.8989387130700832,
+      "grad_norm": 5.250114984615534,
+      "learning_rate": 1e-06,
+      "loss": 0.4338,
+      "step": 4701
+    },
+    {
+      "epoch": 0.8991299359403384,
+      "grad_norm": 3.8972453586081635,
+      "learning_rate": 1e-06,
+      "loss": 0.4434,
+      "step": 4702
+    },
+    {
+      "epoch": 0.8993211588105937,
+      "grad_norm": 2.6640684961156404,
+      "learning_rate": 1e-06,
+      "loss": 0.2679,
+      "step": 4703
+    },
+    {
+      "epoch": 0.899512381680849,
+      "grad_norm": 4.236076210309005,
+      "learning_rate": 1e-06,
+      "loss": 0.4743,
+      "step": 4704
+    },
+    {
+      "epoch": 0.8997036045511043,
+      "grad_norm": 4.371315412721896,
+      "learning_rate": 1e-06,
+      "loss": 0.3705,
+      "step": 4705
+    },
+    {
+      "epoch": 0.8998948274213596,
+      "grad_norm": 5.880620662084733,
+      "learning_rate": 1e-06,
+      "loss": 0.5302,
+      "step": 4706
+    },
+    {
+      "epoch": 0.9000860502916149,
+      "grad_norm": 2.152674267102234,
+      "learning_rate": 1e-06,
+      "loss": 0.1248,
+      "step": 4707
+    },
+    {
+      "epoch": 0.9002772731618701,
+      "grad_norm": 2.5152137372787142,
+      "learning_rate": 1e-06,
+      "loss": 0.0908,
+      "step": 4708
+    },
+    {
+      "epoch": 0.9004684960321254,
+      "grad_norm": 7.817323462161493,
+      "learning_rate": 1e-06,
+      "loss": 0.1406,
+      "step": 4709
+    },
+    {
+      "epoch": 0.9006597189023807,
+      "grad_norm": 3.3695156607387755,
+      "learning_rate": 1e-06,
+      "loss": 0.0831,
+      "step": 4710
+    },
+    {
+      "epoch": 0.900850941772636,
+      "grad_norm": 4.84205568275121,
+      "learning_rate": 1e-06,
+      "loss": 0.085,
+      "step": 4711
+    },
+    {
+      "epoch": 0.9010421646428913,
+      "grad_norm": 1.5515287120939478,
+      "learning_rate": 1e-06,
+      "loss": 0.0283,
+      "step": 4712
+    },
+    {
+      "epoch": 0.9012333875131465,
+      "grad_norm": 5.2858280000274345,
+      "learning_rate": 1e-06,
+      "loss": 0.2013,
+      "step": 4713
+    },
+    {
+      "epoch": 0.9014246103834018,
+      "grad_norm": 4.564853427033856,
+      "learning_rate": 1e-06,
+      "loss": 0.489,
+      "step": 4714
+    },
+    {
+      "epoch": 0.9016158332536571,
+      "grad_norm": 3.858583438640935,
+      "learning_rate": 1e-06,
+      "loss": 0.2872,
+      "step": 4715
+    },
+    {
+      "epoch": 0.9018070561239124,
+      "grad_norm": 2.682060947803829,
+      "learning_rate": 1e-06,
+      "loss": 0.1227,
+      "step": 4716
+    },
+    {
+      "epoch": 0.9019982789941677,
+      "grad_norm": 2.162877939527038,
+      "learning_rate": 1e-06,
+      "loss": 0.2511,
+      "step": 4717
+    },
+    {
+      "epoch": 0.902189501864423,
+      "grad_norm": 4.9290976194101495,
+      "learning_rate": 1e-06,
+      "loss": 0.3859,
+      "step": 4718
+    },
+    {
+      "epoch": 0.9023807247346782,
+      "grad_norm": 3.3678448028870047,
+      "learning_rate": 1e-06,
+      "loss": 0.2067,
+      "step": 4719
+    },
+    {
+      "epoch": 0.9025719476049335,
+      "grad_norm": 4.983696870503895,
+      "learning_rate": 1e-06,
+      "loss": 0.465,
+      "step": 4720
+    },
+    {
+      "epoch": 0.9027631704751888,
+      "grad_norm": 3.5347183883702824,
+      "learning_rate": 1e-06,
+      "loss": 0.1151,
+      "step": 4721
+    },
+    {
+      "epoch": 0.9029543933454441,
+      "grad_norm": 4.425288416197109,
+      "learning_rate": 1e-06,
+      "loss": 0.1654,
+      "step": 4722
+    },
+    {
+      "epoch": 0.9031456162156994,
+      "grad_norm": 3.3910427978748725,
+      "learning_rate": 1e-06,
+      "loss": 0.1243,
+      "step": 4723
+    },
+    {
+      "epoch": 0.9033368390859546,
+      "grad_norm": 2.791346280136799,
+      "learning_rate": 1e-06,
+      "loss": 0.0886,
+      "step": 4724
+    },
+    {
+      "epoch": 0.9035280619562099,
+      "grad_norm": 5.791165483410917,
+      "learning_rate": 1e-06,
+      "loss": 0.13,
+      "step": 4725
+    },
+    {
+      "epoch": 0.9037192848264652,
+      "grad_norm": 3.596618171225893,
+      "learning_rate": 1e-06,
+      "loss": 0.1714,
+      "step": 4726
+    },
+    {
+      "epoch": 0.9039105076967205,
+      "grad_norm": 5.15629216957188,
+      "learning_rate": 1e-06,
+      "loss": 0.5149,
+      "step": 4727
+    },
+    {
+      "epoch": 0.9041017305669758,
+      "grad_norm": 3.6873895660933074,
+      "learning_rate": 1e-06,
+      "loss": 0.3126,
+      "step": 4728
+    },
+    {
+      "epoch": 0.904292953437231,
+      "grad_norm": 4.455198641714395,
+      "learning_rate": 1e-06,
+      "loss": 0.2494,
+      "step": 4729
+    },
+    {
+      "epoch": 0.9044841763074863,
+      "grad_norm": 4.2952063077286695,
+      "learning_rate": 1e-06,
+      "loss": 0.3949,
+      "step": 4730
+    },
+    {
+      "epoch": 0.9046753991777416,
+      "grad_norm": 3.9853839868274847,
+      "learning_rate": 1e-06,
+      "loss": 0.3868,
+      "step": 4731
+    },
+    {
+      "epoch": 0.904866622047997,
+      "grad_norm": 3.2273834883552657,
+      "learning_rate": 1e-06,
+      "loss": 0.1403,
+      "step": 4732
+    },
+    {
+      "epoch": 0.9050578449182523,
+      "grad_norm": 4.018185522798802,
+      "learning_rate": 1e-06,
+      "loss": 0.3649,
+      "step": 4733
+    },
+    {
+      "epoch": 0.9052490677885076,
+      "grad_norm": 2.5971728800156844,
+      "learning_rate": 1e-06,
+      "loss": 0.1242,
+      "step": 4734
+    },
+    {
+      "epoch": 0.9054402906587627,
+      "grad_norm": 4.570894853055628,
+      "learning_rate": 1e-06,
+      "loss": 0.2653,
+      "step": 4735
+    },
+    {
+      "epoch": 0.905631513529018,
+      "grad_norm": 5.703156771636673,
+      "learning_rate": 1e-06,
+      "loss": 0.2234,
+      "step": 4736
+    },
+    {
+      "epoch": 0.9058227363992734,
+      "grad_norm": 2.808512425206893,
+      "learning_rate": 1e-06,
+      "loss": 0.0555,
+      "step": 4737
+    },
+    {
+      "epoch": 0.9060139592695287,
+      "grad_norm": 12.839253698984026,
+      "learning_rate": 1e-06,
+      "loss": 0.1211,
+      "step": 4738
+    },
+    {
+      "epoch": 0.906205182139784,
+      "grad_norm": 3.892913145509913,
+      "learning_rate": 1e-06,
+      "loss": 0.6343,
+      "step": 4739
+    },
+    {
+      "epoch": 0.9063964050100392,
+      "grad_norm": 3.940768505027577,
+      "learning_rate": 1e-06,
+      "loss": 0.3834,
+      "step": 4740
+    },
+    {
+      "epoch": 0.9065876278802945,
+      "grad_norm": 2.756519479185598,
+      "learning_rate": 1e-06,
+      "loss": 0.2732,
+      "step": 4741
+    },
+    {
+      "epoch": 0.9067788507505498,
+      "grad_norm": 2.2279696117431294,
+      "learning_rate": 1e-06,
+      "loss": 0.117,
+      "step": 4742
+    },
+    {
+      "epoch": 0.9069700736208051,
+      "grad_norm": 2.9699370570850276,
+      "learning_rate": 1e-06,
+      "loss": 0.2891,
+      "step": 4743
+    },
+    {
+      "epoch": 0.9071612964910604,
+      "grad_norm": 3.3165356506336288,
+      "learning_rate": 1e-06,
+      "loss": 0.1284,
+      "step": 4744
+    },
+    {
+      "epoch": 0.9073525193613157,
+      "grad_norm": 4.44802103508291,
+      "learning_rate": 1e-06,
+      "loss": 0.3284,
+      "step": 4745
+    },
+    {
+      "epoch": 0.9075437422315709,
+      "grad_norm": 2.9966828762997926,
+      "learning_rate": 1e-06,
+      "loss": 0.1378,
+      "step": 4746
+    },
+    {
+      "epoch": 0.9077349651018262,
+      "grad_norm": 3.387065654090158,
+      "learning_rate": 1e-06,
+      "loss": 0.0657,
+      "step": 4747
+    },
+    {
+      "epoch": 0.9079261879720815,
+      "grad_norm": 3.7589581620754173,
+      "learning_rate": 1e-06,
+      "loss": 0.0745,
+      "step": 4748
+    },
+    {
+      "epoch": 0.9081174108423368,
+      "grad_norm": 3.5817084213764914,
+      "learning_rate": 1e-06,
+      "loss": 0.1338,
+      "step": 4749
+    },
+    {
+      "epoch": 0.9083086337125921,
+      "grad_norm": 12.030597460997301,
+      "learning_rate": 1e-06,
+      "loss": 0.0595,
+      "step": 4750
+    },
+    {
+      "epoch": 0.9084998565828473,
+      "grad_norm": 5.30337099079131,
+      "learning_rate": 1e-06,
+      "loss": 0.1873,
+      "step": 4751
+    },
+    {
+      "epoch": 0.9086910794531026,
+      "grad_norm": 6.587773946027963,
+      "learning_rate": 1e-06,
+      "loss": 0.4296,
+      "step": 4752
+    },
+    {
+      "epoch": 0.9088823023233579,
+      "grad_norm": 4.012153044819416,
+      "learning_rate": 1e-06,
+      "loss": 0.4728,
+      "step": 4753
+    },
+    {
+      "epoch": 0.9090735251936132,
+      "grad_norm": 3.566685142350399,
+      "learning_rate": 1e-06,
+      "loss": 0.3959,
+      "step": 4754
+    },
+    {
+      "epoch": 0.9092647480638685,
+      "grad_norm": 2.5638942182087594,
+      "learning_rate": 1e-06,
+      "loss": 0.2472,
+      "step": 4755
+    },
+    {
+      "epoch": 0.9094559709341237,
+      "grad_norm": 2.8916858273035344,
+      "learning_rate": 1e-06,
+      "loss": 0.1838,
+      "step": 4756
+    },
+    {
+      "epoch": 0.909647193804379,
+      "grad_norm": 2.781840572448396,
+      "learning_rate": 1e-06,
+      "loss": 0.0774,
+      "step": 4757
+    },
+    {
+      "epoch": 0.9098384166746343,
+      "grad_norm": 3.1525037835965906,
+      "learning_rate": 1e-06,
+      "loss": 0.274,
+      "step": 4758
+    },
+    {
+      "epoch": 0.9100296395448896,
+      "grad_norm": 3.1784364080760064,
+      "learning_rate": 1e-06,
+      "loss": 0.0879,
+      "step": 4759
+    },
+    {
+      "epoch": 0.9102208624151449,
+      "grad_norm": 4.37439854438551,
+      "learning_rate": 1e-06,
+      "loss": 0.2127,
+      "step": 4760
+    },
+    {
+      "epoch": 0.9104120852854002,
+      "grad_norm": 2.997984367986417,
+      "learning_rate": 1e-06,
+      "loss": 0.1568,
+      "step": 4761
+    },
+    {
+      "epoch": 0.9106033081556554,
+      "grad_norm": 4.227245903116477,
+      "learning_rate": 1e-06,
+      "loss": 0.0839,
+      "step": 4762
+    },
+    {
+      "epoch": 0.9107945310259107,
+      "grad_norm": 5.270432086292423,
+      "learning_rate": 1e-06,
+      "loss": 0.2816,
+      "step": 4763
+    },
+    {
+      "epoch": 0.910985753896166,
+      "grad_norm": 7.879042057347776,
+      "learning_rate": 1e-06,
+      "loss": 0.6824,
+      "step": 4764
+    },
+    {
+      "epoch": 0.9111769767664213,
+      "grad_norm": 3.562006464451659,
+      "learning_rate": 1e-06,
+      "loss": 0.2676,
+      "step": 4765
+    },
+    {
+      "epoch": 0.9113681996366766,
+      "grad_norm": 3.201098289365497,
+      "learning_rate": 1e-06,
+      "loss": 0.1453,
+      "step": 4766
+    },
+    {
+      "epoch": 0.9115594225069318,
+      "grad_norm": 5.042900480935561,
+      "learning_rate": 1e-06,
+      "loss": 0.4997,
+      "step": 4767
+    },
+    {
+      "epoch": 0.9117506453771871,
+      "grad_norm": 3.7360223305012386,
+      "learning_rate": 1e-06,
+      "loss": 0.295,
+      "step": 4768
+    },
+    {
+      "epoch": 0.9119418682474424,
+      "grad_norm": 4.946155349716613,
+      "learning_rate": 1e-06,
+      "loss": 0.1702,
+      "step": 4769
+    },
+    {
+      "epoch": 0.9121330911176977,
+      "grad_norm": 3.936003143138716,
+      "learning_rate": 1e-06,
+      "loss": 0.0965,
+      "step": 4770
+    },
+    {
+      "epoch": 0.912324313987953,
+      "grad_norm": 3.0837762531698467,
+      "learning_rate": 1e-06,
+      "loss": 0.1853,
+      "step": 4771
+    },
+    {
+      "epoch": 0.9125155368582082,
+      "grad_norm": 2.8978580193763577,
+      "learning_rate": 1e-06,
+      "loss": 0.0781,
+      "step": 4772
+    },
+    {
+      "epoch": 0.9127067597284635,
+      "grad_norm": 2.222526345370394,
+      "learning_rate": 1e-06,
+      "loss": 0.0556,
+      "step": 4773
+    },
+    {
+      "epoch": 0.9128979825987188,
+      "grad_norm": 3.57691741955533,
+      "learning_rate": 1e-06,
+      "loss": 0.053,
+      "step": 4774
+    },
+    {
+      "epoch": 0.9130892054689741,
+      "grad_norm": 4.449588696620374,
+      "learning_rate": 1e-06,
+      "loss": 0.0915,
+      "step": 4775
+    },
+    {
+      "epoch": 0.9132804283392294,
+      "grad_norm": 5.362887729952693,
+      "learning_rate": 1e-06,
+      "loss": 0.1407,
+      "step": 4776
+    },
+    {
+      "epoch": 0.9134716512094847,
+      "grad_norm": 5.28142804630073,
+      "learning_rate": 1e-06,
+      "loss": 0.185,
+      "step": 4777
+    },
+    {
+      "epoch": 0.9136628740797399,
+      "grad_norm": 4.282994270016268,
+      "learning_rate": 1e-06,
+      "loss": 0.2732,
+      "step": 4778
+    },
+    {
+      "epoch": 0.9138540969499952,
+      "grad_norm": 2.6468097557124444,
+      "learning_rate": 1e-06,
+      "loss": 0.2036,
+      "step": 4779
+    },
+    {
+      "epoch": 0.9140453198202505,
+      "grad_norm": 3.5137740624471387,
+      "learning_rate": 1e-06,
+      "loss": 0.4022,
+      "step": 4780
+    },
+    {
+      "epoch": 0.9142365426905058,
+      "grad_norm": 3.1292079822912933,
+      "learning_rate": 1e-06,
+      "loss": 0.1082,
+      "step": 4781
+    },
+    {
+      "epoch": 0.9144277655607611,
+      "grad_norm": 4.808610012792362,
+      "learning_rate": 1e-06,
+      "loss": 0.3288,
+      "step": 4782
+    },
+    {
+      "epoch": 0.9146189884310163,
+      "grad_norm": 3.485810535808384,
+      "learning_rate": 1e-06,
+      "loss": 0.2356,
+      "step": 4783
+    },
+    {
+      "epoch": 0.9148102113012716,
+      "grad_norm": 2.506260376716118,
+      "learning_rate": 1e-06,
+      "loss": 0.1366,
+      "step": 4784
+    },
+    {
+      "epoch": 0.9150014341715269,
+      "grad_norm": 2.6660737530865446,
+      "learning_rate": 1e-06,
+      "loss": 0.0384,
+      "step": 4785
+    },
+    {
+      "epoch": 0.9151926570417822,
+      "grad_norm": 3.099236911252167,
+      "learning_rate": 1e-06,
+      "loss": 0.0732,
+      "step": 4786
+    },
+    {
+      "epoch": 0.9153838799120375,
+      "grad_norm": 4.331832674805176,
+      "learning_rate": 1e-06,
+      "loss": 0.0942,
+      "step": 4787
+    },
+    {
+      "epoch": 0.9155751027822928,
+      "grad_norm": 6.056161614905879,
+      "learning_rate": 1e-06,
+      "loss": 0.3136,
+      "step": 4788
+    },
+    {
+      "epoch": 0.915766325652548,
+      "grad_norm": 5.1357643778340245,
+      "learning_rate": 1e-06,
+      "loss": 0.6126,
+      "step": 4789
+    },
+    {
+      "epoch": 0.9159575485228033,
+      "grad_norm": 3.1614850130584156,
+      "learning_rate": 1e-06,
+      "loss": 0.4478,
+      "step": 4790
+    },
+    {
+      "epoch": 0.9161487713930586,
+      "grad_norm": 2.9254239810239473,
+      "learning_rate": 1e-06,
+      "loss": 0.2176,
+      "step": 4791
+    },
+    {
+      "epoch": 0.9163399942633139,
+      "grad_norm": 3.543321619262663,
+      "learning_rate": 1e-06,
+      "loss": 0.1065,
+      "step": 4792
+    },
+    {
+      "epoch": 0.9165312171335692,
+      "grad_norm": 3.7205535618315615,
+      "learning_rate": 1e-06,
+      "loss": 0.2656,
+      "step": 4793
+    },
+    {
+      "epoch": 0.9167224400038244,
+      "grad_norm": 3.26644259352539,
+      "learning_rate": 1e-06,
+      "loss": 0.1374,
+      "step": 4794
+    },
+    {
+      "epoch": 0.9169136628740797,
+      "grad_norm": 3.8979504114698944,
+      "learning_rate": 1e-06,
+      "loss": 0.106,
+      "step": 4795
+    },
+    {
+      "epoch": 0.917104885744335,
+      "grad_norm": 3.3397726486400976,
+      "learning_rate": 1e-06,
+      "loss": 0.0862,
+      "step": 4796
+    },
+    {
+      "epoch": 0.9172961086145903,
+      "grad_norm": 6.995828202840057,
+      "learning_rate": 1e-06,
+      "loss": 0.1711,
+      "step": 4797
+    },
+    {
+      "epoch": 0.9174873314848456,
+      "grad_norm": 5.011483547432976,
+      "learning_rate": 1e-06,
+      "loss": 0.283,
+      "step": 4798
+    },
+    {
+      "epoch": 0.9176785543551008,
+      "grad_norm": 4.591236906253663,
+      "learning_rate": 1e-06,
+      "loss": 0.1254,
+      "step": 4799
+    },
+    {
+      "epoch": 0.9178697772253561,
+      "grad_norm": 4.785542474846969,
+      "learning_rate": 1e-06,
+      "loss": 0.0771,
+      "step": 4800
+    },
+    {
+      "epoch": 0.9180610000956114,
+      "grad_norm": 3.102060888109716,
+      "learning_rate": 1e-06,
+      "loss": 0.208,
+      "step": 4801
+    },
+    {
+      "epoch": 0.9182522229658667,
+      "grad_norm": 5.067932416686566,
+      "learning_rate": 1e-06,
+      "loss": 0.2903,
+      "step": 4802
+    },
+    {
+      "epoch": 0.918443445836122,
+      "grad_norm": 2.315596697965117,
+      "learning_rate": 1e-06,
+      "loss": 0.2332,
+      "step": 4803
+    },
+    {
+      "epoch": 0.9186346687063773,
+      "grad_norm": 3.1395545767463022,
+      "learning_rate": 1e-06,
+      "loss": 0.2447,
+      "step": 4804
+    },
+    {
+      "epoch": 0.9188258915766325,
+      "grad_norm": 2.747982325334638,
+      "learning_rate": 1e-06,
+      "loss": 0.1843,
+      "step": 4805
+    },
+    {
+      "epoch": 0.9190171144468878,
+      "grad_norm": 4.84074318373975,
+      "learning_rate": 1e-06,
+      "loss": 0.1553,
+      "step": 4806
+    },
+    {
+      "epoch": 0.9192083373171431,
+      "grad_norm": 2.7257595683509557,
+      "learning_rate": 1e-06,
+      "loss": 0.1994,
+      "step": 4807
+    },
+    {
+      "epoch": 0.9193995601873984,
+      "grad_norm": 4.216793256425034,
+      "learning_rate": 1e-06,
+      "loss": 0.2236,
+      "step": 4808
+    },
+    {
+      "epoch": 0.9195907830576537,
+      "grad_norm": 4.737380682013887,
+      "learning_rate": 1e-06,
+      "loss": 0.2382,
+      "step": 4809
+    },
+    {
+      "epoch": 0.9197820059279089,
+      "grad_norm": 3.7711880062218435,
+      "learning_rate": 1e-06,
+      "loss": 0.0848,
+      "step": 4810
+    },
+    {
+      "epoch": 0.9199732287981642,
+      "grad_norm": 2.664371595601988,
+      "learning_rate": 1e-06,
+      "loss": 0.0817,
+      "step": 4811
+    },
+    {
+      "epoch": 0.9201644516684195,
+      "grad_norm": 2.517087617830611,
+      "learning_rate": 1e-06,
+      "loss": 0.0548,
+      "step": 4812
+    },
+    {
+      "epoch": 0.9203556745386748,
+      "grad_norm": 9.654283638153478,
+      "learning_rate": 1e-06,
+      "loss": 0.1026,
+      "step": 4813
+    },
+    {
+      "epoch": 0.9205468974089301,
+      "grad_norm": 4.732493413808746,
+      "learning_rate": 1e-06,
+      "loss": 0.5561,
+      "step": 4814
+    },
+    {
+      "epoch": 0.9207381202791854,
+      "grad_norm": 3.755662457664512,
+      "learning_rate": 1e-06,
+      "loss": 0.3585,
+      "step": 4815
+    },
+    {
+      "epoch": 0.9209293431494406,
+      "grad_norm": 2.353677400592696,
+      "learning_rate": 1e-06,
+      "loss": 0.1989,
+      "step": 4816
+    },
+    {
+      "epoch": 0.9211205660196959,
+      "grad_norm": 4.211460625988096,
+      "learning_rate": 1e-06,
+      "loss": 0.3175,
+      "step": 4817
+    },
+    {
+      "epoch": 0.9213117888899512,
+      "grad_norm": 4.1307153122042335,
+      "learning_rate": 1e-06,
+      "loss": 0.1876,
+      "step": 4818
+    },
+    {
+      "epoch": 0.9215030117602065,
+      "grad_norm": 4.353093917089428,
+      "learning_rate": 1e-06,
+      "loss": 0.2672,
+      "step": 4819
+    },
+    {
+      "epoch": 0.9216942346304619,
+      "grad_norm": 3.11989374688984,
+      "learning_rate": 1e-06,
+      "loss": 0.2457,
+      "step": 4820
+    },
+    {
+      "epoch": 0.921885457500717,
+      "grad_norm": 2.2804256674293093,
+      "learning_rate": 1e-06,
+      "loss": 0.0501,
+      "step": 4821
+    },
+    {
+      "epoch": 0.9220766803709723,
+      "grad_norm": 2.9376467404339195,
+      "learning_rate": 1e-06,
+      "loss": 0.1518,
+      "step": 4822
+    },
+    {
+      "epoch": 0.9222679032412276,
+      "grad_norm": 2.2800654641172917,
+      "learning_rate": 1e-06,
+      "loss": 0.0523,
+      "step": 4823
+    },
+    {
+      "epoch": 0.922459126111483,
+      "grad_norm": 3.0355167837569628,
+      "learning_rate": 1e-06,
+      "loss": 0.0556,
+      "step": 4824
+    },
+    {
+      "epoch": 0.9226503489817383,
+      "grad_norm": 3.166233099987591,
+      "learning_rate": 1e-06,
+      "loss": 0.054,
+      "step": 4825
+    },
+    {
+      "epoch": 0.9228415718519934,
+      "grad_norm": 7.381509268242399,
+      "learning_rate": 1e-06,
+      "loss": 0.3393,
+      "step": 4826
+    },
+    {
+      "epoch": 0.9230327947222488,
+      "grad_norm": 3.404468315390532,
+      "learning_rate": 1e-06,
+      "loss": 0.5754,
+      "step": 4827
+    },
+    {
+      "epoch": 0.923224017592504,
+      "grad_norm": 3.5760516027187887,
+      "learning_rate": 1e-06,
+      "loss": 0.3427,
+      "step": 4828
+    },
+    {
+      "epoch": 0.9234152404627594,
+      "grad_norm": 3.841389248996343,
+      "learning_rate": 1e-06,
+      "loss": 0.2175,
+      "step": 4829
+    },
+    {
+      "epoch": 0.9236064633330147,
+      "grad_norm": 2.4192726181784714,
+      "learning_rate": 1e-06,
+      "loss": 0.0801,
+      "step": 4830
+    },
+    {
+      "epoch": 0.92379768620327,
+      "grad_norm": 2.840558619295388,
+      "learning_rate": 1e-06,
+      "loss": 0.1034,
+      "step": 4831
+    },
+    {
+      "epoch": 0.9239889090735252,
+      "grad_norm": 1.7002507024919775,
+      "learning_rate": 1e-06,
+      "loss": 0.1265,
+      "step": 4832
+    },
+    {
+      "epoch": 0.9241801319437805,
+      "grad_norm": 5.00757349547883,
+      "learning_rate": 1e-06,
+      "loss": 0.0683,
+      "step": 4833
+    },
+    {
+      "epoch": 0.9243713548140358,
+      "grad_norm": 2.8046756839768943,
+      "learning_rate": 1e-06,
+      "loss": 0.0606,
+      "step": 4834
+    },
+    {
+      "epoch": 0.9245625776842911,
+      "grad_norm": 3.812904586586658,
+      "learning_rate": 1e-06,
+      "loss": 0.2422,
+      "step": 4835
+    },
+    {
+      "epoch": 0.9247538005545464,
+      "grad_norm": 3.0650924297789746,
+      "learning_rate": 1e-06,
+      "loss": 0.0651,
+      "step": 4836
+    },
+    {
+      "epoch": 0.9249450234248016,
+      "grad_norm": 8.585315463136023,
+      "learning_rate": 1e-06,
+      "loss": 0.2405,
+      "step": 4837
+    },
+    {
+      "epoch": 0.9251362462950569,
+      "grad_norm": 5.842908176435084,
+      "learning_rate": 1e-06,
+      "loss": 0.3144,
+      "step": 4838
+    },
+    {
+      "epoch": 0.9253274691653122,
+      "grad_norm": 6.618509820285148,
+      "learning_rate": 1e-06,
+      "loss": 0.2449,
+      "step": 4839
+    },
+    {
+      "epoch": 0.9255186920355675,
+      "grad_norm": 3.5927171632756467,
+      "learning_rate": 1e-06,
+      "loss": 0.4225,
+      "step": 4840
+    },
+    {
+      "epoch": 0.9257099149058228,
+      "grad_norm": 4.523008019807307,
+      "learning_rate": 1e-06,
+      "loss": 0.2148,
+      "step": 4841
+    },
+    {
+      "epoch": 0.925901137776078,
+      "grad_norm": 5.331207904895449,
+      "learning_rate": 1e-06,
+      "loss": 0.1758,
+      "step": 4842
+    },
+    {
+      "epoch": 0.9260923606463333,
+      "grad_norm": 3.2637290013608253,
+      "learning_rate": 1e-06,
+      "loss": 0.4056,
+      "step": 4843
+    },
+    {
+      "epoch": 0.9262835835165886,
+      "grad_norm": 3.565754257386825,
+      "learning_rate": 1e-06,
+      "loss": 0.2526,
+      "step": 4844
+    },
+    {
+      "epoch": 0.9264748063868439,
+      "grad_norm": 2.9780130531429747,
+      "learning_rate": 1e-06,
+      "loss": 0.2344,
+      "step": 4845
+    },
+    {
+      "epoch": 0.9266660292570992,
+      "grad_norm": 3.4535508389506586,
+      "learning_rate": 1e-06,
+      "loss": 0.1323,
+      "step": 4846
+    },
+    {
+      "epoch": 0.9268572521273545,
+      "grad_norm": 2.7924860610545728,
+      "learning_rate": 1e-06,
+      "loss": 0.096,
+      "step": 4847
+    },
+    {
+      "epoch": 0.9270484749976097,
+      "grad_norm": 2.662856648516419,
+      "learning_rate": 1e-06,
+      "loss": 0.0714,
+      "step": 4848
+    },
+    {
+      "epoch": 0.927239697867865,
+      "grad_norm": 2.129656234366173,
+      "learning_rate": 1e-06,
+      "loss": 0.0306,
+      "step": 4849
+    },
+    {
+      "epoch": 0.9274309207381203,
+      "grad_norm": 6.469529551835304,
+      "learning_rate": 1e-06,
+      "loss": 0.304,
+      "step": 4850
+    },
+    {
+      "epoch": 0.9276221436083756,
+      "grad_norm": 6.084681566192506,
+      "learning_rate": 1e-06,
+      "loss": 0.2814,
+      "step": 4851
+    },
+    {
+      "epoch": 0.9278133664786309,
+      "grad_norm": 4.330247269193009,
+      "learning_rate": 1e-06,
+      "loss": 0.1425,
+      "step": 4852
+    },
+    {
+      "epoch": 0.9280045893488861,
+      "grad_norm": 3.791311953095339,
+      "learning_rate": 1e-06,
+      "loss": 0.3191,
+      "step": 4853
+    },
+    {
+      "epoch": 0.9281958122191414,
+      "grad_norm": 2.757977877837292,
+      "learning_rate": 1e-06,
+      "loss": 0.138,
+      "step": 4854
+    },
+    {
+      "epoch": 0.9283870350893967,
+      "grad_norm": 3.937551831101074,
+      "learning_rate": 1e-06,
+      "loss": 0.1388,
+      "step": 4855
+    },
+    {
+      "epoch": 0.928578257959652,
+      "grad_norm": 3.4945407252414578,
+      "learning_rate": 1e-06,
+      "loss": 0.3251,
+      "step": 4856
+    },
+    {
+      "epoch": 0.9287694808299073,
+      "grad_norm": 3.2209306755670917,
+      "learning_rate": 1e-06,
+      "loss": 0.1427,
+      "step": 4857
+    },
+    {
+      "epoch": 0.9289607037001626,
+      "grad_norm": 3.6808616239016,
+      "learning_rate": 1e-06,
+      "loss": 0.2077,
+      "step": 4858
+    },
+    {
+      "epoch": 0.9291519265704178,
+      "grad_norm": 3.7871035831086446,
+      "learning_rate": 1e-06,
+      "loss": 0.3044,
+      "step": 4859
+    },
+    {
+      "epoch": 0.9293431494406731,
+      "grad_norm": 2.194819493442494,
+      "learning_rate": 1e-06,
+      "loss": 0.0792,
+      "step": 4860
+    },
+    {
+      "epoch": 0.9295343723109284,
+      "grad_norm": 3.310420553222321,
+      "learning_rate": 1e-06,
+      "loss": 0.129,
+      "step": 4861
+    },
+    {
+      "epoch": 0.9297255951811837,
+      "grad_norm": 3.629209901980633,
+      "learning_rate": 1e-06,
+      "loss": 0.0877,
+      "step": 4862
+    },
+    {
+      "epoch": 0.929916818051439,
+      "grad_norm": 3.7044528560267054,
+      "learning_rate": 1e-06,
+      "loss": 0.0869,
+      "step": 4863
+    },
+    {
+      "epoch": 0.9301080409216942,
+      "grad_norm": 4.843905341826086,
+      "learning_rate": 1e-06,
+      "loss": 0.4323,
+      "step": 4864
+    },
+    {
+      "epoch": 0.9302992637919495,
+      "grad_norm": 3.10090488024947,
+      "learning_rate": 1e-06,
+      "loss": 0.3198,
+      "step": 4865
+    },
+    {
+      "epoch": 0.9304904866622048,
+      "grad_norm": 3.686553057621072,
+      "learning_rate": 1e-06,
+      "loss": 0.2792,
+      "step": 4866
+    },
+    {
+      "epoch": 0.9306817095324601,
+      "grad_norm": 2.8582536343361484,
+      "learning_rate": 1e-06,
+      "loss": 0.1512,
+      "step": 4867
+    },
+    {
+      "epoch": 0.9308729324027154,
+      "grad_norm": 2.2622025520845197,
+      "learning_rate": 1e-06,
+      "loss": 0.0756,
+      "step": 4868
+    },
+    {
+      "epoch": 0.9310641552729706,
+      "grad_norm": 2.8888947128171987,
+      "learning_rate": 1e-06,
+      "loss": 0.1844,
+      "step": 4869
+    },
+    {
+      "epoch": 0.9312553781432259,
+      "grad_norm": 2.569504909548548,
+      "learning_rate": 1e-06,
+      "loss": 0.156,
+      "step": 4870
+    },
+    {
+      "epoch": 0.9314466010134812,
+      "grad_norm": 3.2533153349998694,
+      "learning_rate": 1e-06,
+      "loss": 0.1682,
+      "step": 4871
+    },
+    {
+      "epoch": 0.9316378238837365,
+      "grad_norm": 2.2264543473596006,
+      "learning_rate": 1e-06,
+      "loss": 0.1132,
+      "step": 4872
+    },
+    {
+      "epoch": 0.9318290467539918,
+      "grad_norm": 3.1958199916463297,
+      "learning_rate": 1e-06,
+      "loss": 0.0755,
+      "step": 4873
+    },
+    {
+      "epoch": 0.9320202696242471,
+      "grad_norm": 3.20554598985549,
+      "learning_rate": 1e-06,
+      "loss": 0.0564,
+      "step": 4874
+    },
+    {
+      "epoch": 0.9322114924945023,
+      "grad_norm": 4.036215865533407,
+      "learning_rate": 1e-06,
+      "loss": 0.1062,
+      "step": 4875
+    },
+    {
+      "epoch": 0.9324027153647576,
+      "grad_norm": 4.229247645212,
+      "learning_rate": 1e-06,
+      "loss": 0.2687,
+      "step": 4876
+    },
+    {
+      "epoch": 0.9325939382350129,
+      "grad_norm": 4.537418139851615,
+      "learning_rate": 1e-06,
+      "loss": 0.5196,
+      "step": 4877
+    },
+    {
+      "epoch": 0.9327851611052682,
+      "grad_norm": 2.400648756557706,
+      "learning_rate": 1e-06,
+      "loss": 0.2584,
+      "step": 4878
+    },
+    {
+      "epoch": 0.9329763839755235,
+      "grad_norm": 4.40610411923294,
+      "learning_rate": 1e-06,
+      "loss": 0.2301,
+      "step": 4879
+    },
+    {
+      "epoch": 0.9331676068457787,
+      "grad_norm": 4.249528297725976,
+      "learning_rate": 1e-06,
+      "loss": 0.2113,
+      "step": 4880
+    },
+    {
+      "epoch": 0.933358829716034,
+      "grad_norm": 2.224234880201983,
+      "learning_rate": 1e-06,
+      "loss": 0.2533,
+      "step": 4881
+    },
+    {
+      "epoch": 0.9335500525862893,
+      "grad_norm": 2.75926390755831,
+      "learning_rate": 1e-06,
+      "loss": 0.1343,
+      "step": 4882
+    },
+    {
+      "epoch": 0.9337412754565446,
+      "grad_norm": 3.126033612022374,
+      "learning_rate": 1e-06,
+      "loss": 0.1323,
+      "step": 4883
+    },
+    {
+      "epoch": 0.9339324983267999,
+      "grad_norm": 2.511933740221842,
+      "learning_rate": 1e-06,
+      "loss": 0.0364,
+      "step": 4884
+    },
+    {
+      "epoch": 0.9341237211970551,
+      "grad_norm": 1.1698691431048305,
+      "learning_rate": 1e-06,
+      "loss": 0.0297,
+      "step": 4885
+    },
+    {
+      "epoch": 0.9343149440673104,
+      "grad_norm": 2.4855786172991308,
+      "learning_rate": 1e-06,
+      "loss": 0.0811,
+      "step": 4886
+    },
+    {
+      "epoch": 0.9345061669375657,
+      "grad_norm": 3.231143092262895,
+      "learning_rate": 1e-06,
+      "loss": 0.0869,
+      "step": 4887
+    },
+    {
+      "epoch": 0.934697389807821,
+      "grad_norm": 3.4490709629561587,
+      "learning_rate": 1e-06,
+      "loss": 0.136,
+      "step": 4888
+    },
+    {
+      "epoch": 0.9348886126780763,
+      "grad_norm": 3.3255219644989187,
+      "learning_rate": 1e-06,
+      "loss": 0.2258,
+      "step": 4889
+    },
+    {
+      "epoch": 0.9350798355483316,
+      "grad_norm": 4.409266095671864,
+      "learning_rate": 1e-06,
+      "loss": 0.1908,
+      "step": 4890
+    },
+    {
+      "epoch": 0.9352710584185868,
+      "grad_norm": 3.7287775330353434,
+      "learning_rate": 1e-06,
+      "loss": 0.1335,
+      "step": 4891
+    },
+    {
+      "epoch": 0.9354622812888421,
+      "grad_norm": 3.0768780191496967,
+      "learning_rate": 1e-06,
+      "loss": 0.2278,
+      "step": 4892
+    },
+    {
+      "epoch": 0.9356535041590974,
+      "grad_norm": 5.969582704308084,
+      "learning_rate": 1e-06,
+      "loss": 0.6035,
+      "step": 4893
+    },
+    {
+      "epoch": 0.9358447270293527,
+      "grad_norm": 4.350386784337669,
+      "learning_rate": 1e-06,
+      "loss": 0.1205,
+      "step": 4894
+    },
+    {
+      "epoch": 0.936035949899608,
+      "grad_norm": 3.650701633061137,
+      "learning_rate": 1e-06,
+      "loss": 0.1099,
+      "step": 4895
+    },
+    {
+      "epoch": 0.9362271727698632,
+      "grad_norm": 2.625118797883808,
+      "learning_rate": 1e-06,
+      "loss": 0.1113,
+      "step": 4896
+    },
+    {
+      "epoch": 0.9364183956401185,
+      "grad_norm": 2.760776817771173,
+      "learning_rate": 1e-06,
+      "loss": 0.1318,
+      "step": 4897
+    },
+    {
+      "epoch": 0.9366096185103738,
+      "grad_norm": 3.7652012756773447,
+      "learning_rate": 1e-06,
+      "loss": 0.1437,
+      "step": 4898
+    },
+    {
+      "epoch": 0.9368008413806291,
+      "grad_norm": 2.8131327447146868,
+      "learning_rate": 1e-06,
+      "loss": 0.0384,
+      "step": 4899
+    },
+    {
+      "epoch": 0.9369920642508844,
+      "grad_norm": 5.498170721851141,
+      "learning_rate": 1e-06,
+      "loss": 0.1918,
+      "step": 4900
+    },
+    {
+      "epoch": 0.9371832871211397,
+      "grad_norm": 6.106500520125657,
+      "learning_rate": 1e-06,
+      "loss": 0.3012,
+      "step": 4901
+    },
+    {
+      "epoch": 0.9373745099913949,
+      "grad_norm": 4.251313847929103,
+      "learning_rate": 1e-06,
+      "loss": 0.4521,
+      "step": 4902
+    },
+    {
+      "epoch": 0.9375657328616502,
+      "grad_norm": 3.5216219468413428,
+      "learning_rate": 1e-06,
+      "loss": 0.2028,
+      "step": 4903
+    },
+    {
+      "epoch": 0.9377569557319055,
+      "grad_norm": 4.603035190843064,
+      "learning_rate": 1e-06,
+      "loss": 0.2594,
+      "step": 4904
+    },
+    {
+      "epoch": 0.9379481786021608,
+      "grad_norm": 3.917322928998031,
+      "learning_rate": 1e-06,
+      "loss": 0.1667,
+      "step": 4905
+    },
+    {
+      "epoch": 0.9381394014724161,
+      "grad_norm": 3.855325899115549,
+      "learning_rate": 1e-06,
+      "loss": 0.2138,
+      "step": 4906
+    },
+    {
+      "epoch": 0.9383306243426713,
+      "grad_norm": 2.383120057315738,
+      "learning_rate": 1e-06,
+      "loss": 0.1847,
+      "step": 4907
+    },
+    {
+      "epoch": 0.9385218472129266,
+      "grad_norm": 5.397064520198056,
+      "learning_rate": 1e-06,
+      "loss": 0.1474,
+      "step": 4908
+    },
+    {
+      "epoch": 0.9387130700831819,
+      "grad_norm": 2.419963352934734,
+      "learning_rate": 1e-06,
+      "loss": 0.0683,
+      "step": 4909
+    },
+    {
+      "epoch": 0.9389042929534372,
+      "grad_norm": 2.73497055107595,
+      "learning_rate": 1e-06,
+      "loss": 0.048,
+      "step": 4910
+    },
+    {
+      "epoch": 0.9390955158236926,
+      "grad_norm": 4.833461343232048,
+      "learning_rate": 1e-06,
+      "loss": 0.1789,
+      "step": 4911
+    },
+    {
+      "epoch": 0.9392867386939477,
+      "grad_norm": 4.704043479277337,
+      "learning_rate": 1e-06,
+      "loss": 0.0923,
+      "step": 4912
+    },
+    {
+      "epoch": 0.939477961564203,
+      "grad_norm": 17.418899514686178,
+      "learning_rate": 1e-06,
+      "loss": 0.2228,
+      "step": 4913
+    },
+    {
+      "epoch": 0.9396691844344583,
+      "grad_norm": 5.968845226866345,
+      "learning_rate": 1e-06,
+      "loss": 0.5672,
+      "step": 4914
+    },
+    {
+      "epoch": 0.9398604073047137,
+      "grad_norm": 3.9148823075611947,
+      "learning_rate": 1e-06,
+      "loss": 0.3331,
+      "step": 4915
+    },
+    {
+      "epoch": 0.940051630174969,
+      "grad_norm": 2.9831172829365733,
+      "learning_rate": 1e-06,
+      "loss": 0.25,
+      "step": 4916
+    },
+    {
+      "epoch": 0.9402428530452243,
+      "grad_norm": 3.6584337022318016,
+      "learning_rate": 1e-06,
+      "loss": 0.4552,
+      "step": 4917
+    },
+    {
+      "epoch": 0.9404340759154795,
+      "grad_norm": 4.958389516987633,
+      "learning_rate": 1e-06,
+      "loss": 0.4985,
+      "step": 4918
+    },
+    {
+      "epoch": 0.9406252987857348,
+      "grad_norm": 4.0303506957021344,
+      "learning_rate": 1e-06,
+      "loss": 0.2553,
+      "step": 4919
+    },
+    {
+      "epoch": 0.9408165216559901,
+      "grad_norm": 2.797029629487321,
+      "learning_rate": 1e-06,
+      "loss": 0.1887,
+      "step": 4920
+    },
+    {
+      "epoch": 0.9410077445262454,
+      "grad_norm": 1.6903865991629006,
+      "learning_rate": 1e-06,
+      "loss": 0.0875,
+      "step": 4921
+    },
+    {
+      "epoch": 0.9411989673965007,
+      "grad_norm": 5.865426887658126,
+      "learning_rate": 1e-06,
+      "loss": 0.0663,
+      "step": 4922
+    },
+    {
+      "epoch": 0.9413901902667559,
+      "grad_norm": 2.4508427980359766,
+      "learning_rate": 1e-06,
+      "loss": 0.0595,
+      "step": 4923
+    },
+    {
+      "epoch": 0.9415814131370112,
+      "grad_norm": 3.755666901431984,
+      "learning_rate": 1e-06,
+      "loss": 0.0886,
+      "step": 4924
+    },
+    {
+      "epoch": 0.9417726360072665,
+      "grad_norm": 4.925212588296772,
+      "learning_rate": 1e-06,
+      "loss": 0.3128,
+      "step": 4925
+    },
+    {
+      "epoch": 0.9419638588775218,
+      "grad_norm": 5.417740622450325,
+      "learning_rate": 1e-06,
+      "loss": 0.4622,
+      "step": 4926
+    },
+    {
+      "epoch": 0.9421550817477771,
+      "grad_norm": 3.8407042306622308,
+      "learning_rate": 1e-06,
+      "loss": 0.3678,
+      "step": 4927
+    },
+    {
+      "epoch": 0.9423463046180324,
+      "grad_norm": 4.1357328484620055,
+      "learning_rate": 1e-06,
+      "loss": 0.1648,
+      "step": 4928
+    },
+    {
+      "epoch": 0.9425375274882876,
+      "grad_norm": 2.7930689653677176,
+      "learning_rate": 1e-06,
+      "loss": 0.1356,
+      "step": 4929
+    },
+    {
+      "epoch": 0.9427287503585429,
+      "grad_norm": 2.878183178308488,
+      "learning_rate": 1e-06,
+      "loss": 0.1545,
+      "step": 4930
+    },
+    {
+      "epoch": 0.9429199732287982,
+      "grad_norm": 4.937624579680817,
+      "learning_rate": 1e-06,
+      "loss": 0.4348,
+      "step": 4931
+    },
+    {
+      "epoch": 0.9431111960990535,
+      "grad_norm": 4.516962084428866,
+      "learning_rate": 1e-06,
+      "loss": 0.4966,
+      "step": 4932
+    },
+    {
+      "epoch": 0.9433024189693088,
+      "grad_norm": 2.426249828938041,
+      "learning_rate": 1e-06,
+      "loss": 0.0825,
+      "step": 4933
+    },
+    {
+      "epoch": 0.943493641839564,
+      "grad_norm": 3.4177554341916827,
+      "learning_rate": 1e-06,
+      "loss": 0.2981,
+      "step": 4934
+    },
+    {
+      "epoch": 0.9436848647098193,
+      "grad_norm": 3.3911295572526075,
+      "learning_rate": 1e-06,
+      "loss": 0.1113,
+      "step": 4935
+    },
+    {
+      "epoch": 0.9438760875800746,
+      "grad_norm": 2.0374711063887254,
+      "learning_rate": 1e-06,
+      "loss": 0.0604,
+      "step": 4936
+    },
+    {
+      "epoch": 0.9440673104503299,
+      "grad_norm": 5.834413955141611,
+      "learning_rate": 1e-06,
+      "loss": 0.1729,
+      "step": 4937
+    },
+    {
+      "epoch": 0.9442585333205852,
+      "grad_norm": 4.386838535973058,
+      "learning_rate": 1e-06,
+      "loss": 0.0978,
+      "step": 4938
+    },
+    {
+      "epoch": 0.9444497561908404,
+      "grad_norm": 5.048720834208164,
+      "learning_rate": 1e-06,
+      "loss": 0.2446,
+      "step": 4939
+    },
+    {
+      "epoch": 0.9446409790610957,
+      "grad_norm": 5.401109489620819,
+      "learning_rate": 1e-06,
+      "loss": 0.2787,
+      "step": 4940
+    },
+    {
+      "epoch": 0.944832201931351,
+      "grad_norm": 3.343920124039968,
+      "learning_rate": 1e-06,
+      "loss": 0.2109,
+      "step": 4941
+    },
+    {
+      "epoch": 0.9450234248016063,
+      "grad_norm": 3.6836521309911663,
+      "learning_rate": 1e-06,
+      "loss": 0.2285,
+      "step": 4942
+    },
+    {
+      "epoch": 0.9452146476718616,
+      "grad_norm": 2.8305035278643342,
+      "learning_rate": 1e-06,
+      "loss": 0.2394,
+      "step": 4943
+    },
+    {
+      "epoch": 0.9454058705421169,
+      "grad_norm": 10.597964008501268,
+      "learning_rate": 1e-06,
+      "loss": 0.1955,
+      "step": 4944
+    },
+    {
+      "epoch": 0.9455970934123721,
+      "grad_norm": 2.3923247035749986,
+      "learning_rate": 1e-06,
+      "loss": 0.0988,
+      "step": 4945
+    },
+    {
+      "epoch": 0.9457883162826274,
+      "grad_norm": 3.8715523179687295,
+      "learning_rate": 1e-06,
+      "loss": 0.1415,
+      "step": 4946
+    },
+    {
+      "epoch": 0.9459795391528827,
+      "grad_norm": 5.286734176980779,
+      "learning_rate": 1e-06,
+      "loss": 0.0681,
+      "step": 4947
+    },
+    {
+      "epoch": 0.946170762023138,
+      "grad_norm": 4.255100779611844,
+      "learning_rate": 1e-06,
+      "loss": 0.1147,
+      "step": 4948
+    },
+    {
+      "epoch": 0.9463619848933933,
+      "grad_norm": 6.48924539419899,
+      "learning_rate": 1e-06,
+      "loss": 0.1378,
+      "step": 4949
+    },
+    {
+      "epoch": 0.9465532077636485,
+      "grad_norm": 4.549063768585168,
+      "learning_rate": 1e-06,
+      "loss": 0.0565,
+      "step": 4950
+    },
+    {
+      "epoch": 0.9467444306339038,
+      "grad_norm": 5.21760354612406,
+      "learning_rate": 1e-06,
+      "loss": 0.6296,
+      "step": 4951
+    },
+    {
+      "epoch": 0.9469356535041591,
+      "grad_norm": 4.367251919457238,
+      "learning_rate": 1e-06,
+      "loss": 0.4485,
+      "step": 4952
+    },
+    {
+      "epoch": 0.9471268763744144,
+      "grad_norm": 5.959659384396674,
+      "learning_rate": 1e-06,
+      "loss": 0.2304,
+      "step": 4953
+    },
+    {
+      "epoch": 0.9473180992446697,
+      "grad_norm": 2.0458077713352885,
+      "learning_rate": 1e-06,
+      "loss": 0.2029,
+      "step": 4954
+    },
+    {
+      "epoch": 0.9475093221149249,
+      "grad_norm": 3.1435433109358444,
+      "learning_rate": 1e-06,
+      "loss": 0.3104,
+      "step": 4955
+    },
+    {
+      "epoch": 0.9477005449851802,
+      "grad_norm": 3.8126528349712943,
+      "learning_rate": 1e-06,
+      "loss": 0.2196,
+      "step": 4956
+    },
+    {
+      "epoch": 0.9478917678554355,
+      "grad_norm": 1.8366968410780264,
+      "learning_rate": 1e-06,
+      "loss": 0.1363,
+      "step": 4957
+    },
+    {
+      "epoch": 0.9480829907256908,
+      "grad_norm": 3.979247140653177,
+      "learning_rate": 1e-06,
+      "loss": 0.1754,
+      "step": 4958
+    },
+    {
+      "epoch": 0.9482742135959461,
+      "grad_norm": 5.137953971612075,
+      "learning_rate": 1e-06,
+      "loss": 0.3041,
+      "step": 4959
+    },
+    {
+      "epoch": 0.9484654364662014,
+      "grad_norm": 2.291561286121704,
+      "learning_rate": 1e-06,
+      "loss": 0.0874,
+      "step": 4960
+    },
+    {
+      "epoch": 0.9486566593364566,
+      "grad_norm": 1.9363916364746465,
+      "learning_rate": 1e-06,
+      "loss": 0.0218,
+      "step": 4961
+    },
+    {
+      "epoch": 0.9488478822067119,
+      "grad_norm": 5.26832072363389,
+      "learning_rate": 1e-06,
+      "loss": 0.1387,
+      "step": 4962
+    },
+    {
+      "epoch": 0.9490391050769672,
+      "grad_norm": 3.3284725670368083,
+      "learning_rate": 1e-06,
+      "loss": 0.2401,
+      "step": 4963
+    },
+    {
+      "epoch": 0.9492303279472225,
+      "grad_norm": 4.580676100356469,
+      "learning_rate": 1e-06,
+      "loss": 0.5466,
+      "step": 4964
+    },
+    {
+      "epoch": 0.9494215508174778,
+      "grad_norm": 3.446850491779866,
+      "learning_rate": 1e-06,
+      "loss": 0.2742,
+      "step": 4965
+    },
+    {
+      "epoch": 0.949612773687733,
+      "grad_norm": 3.223898464421441,
+      "learning_rate": 1e-06,
+      "loss": 0.2117,
+      "step": 4966
+    },
+    {
+      "epoch": 0.9498039965579883,
+      "grad_norm": 3.001426198824443,
+      "learning_rate": 1e-06,
+      "loss": 0.2244,
+      "step": 4967
+    },
+    {
+      "epoch": 0.9499952194282436,
+      "grad_norm": 3.888360279729494,
+      "learning_rate": 1e-06,
+      "loss": 0.4231,
+      "step": 4968
+    },
+    {
+      "epoch": 0.9501864422984989,
+      "grad_norm": 2.436263528847942,
+      "learning_rate": 1e-06,
+      "loss": 0.185,
+      "step": 4969
+    },
+    {
+      "epoch": 0.9503776651687542,
+      "grad_norm": 4.143294254976974,
+      "learning_rate": 1e-06,
+      "loss": 0.1174,
+      "step": 4970
+    },
+    {
+      "epoch": 0.9505688880390095,
+      "grad_norm": 3.181525521628797,
+      "learning_rate": 1e-06,
+      "loss": 0.0673,
+      "step": 4971
+    },
+    {
+      "epoch": 0.9507601109092647,
+      "grad_norm": 2.3824397561621455,
+      "learning_rate": 1e-06,
+      "loss": 0.0792,
+      "step": 4972
+    },
+    {
+      "epoch": 0.95095133377952,
+      "grad_norm": 2.6605744110237244,
+      "learning_rate": 1e-06,
+      "loss": 0.0553,
+      "step": 4973
+    },
+    {
+      "epoch": 0.9511425566497753,
+      "grad_norm": 4.459329773019275,
+      "learning_rate": 1e-06,
+      "loss": 0.1697,
+      "step": 4974
+    },
+    {
+      "epoch": 0.9513337795200306,
+      "grad_norm": 5.729298020215335,
+      "learning_rate": 1e-06,
+      "loss": 0.1203,
+      "step": 4975
+    },
+    {
+      "epoch": 0.9515250023902859,
+      "grad_norm": 5.854454639813914,
+      "learning_rate": 1e-06,
+      "loss": 0.1041,
+      "step": 4976
+    },
+    {
+      "epoch": 0.9517162252605411,
+      "grad_norm": 3.6958929981001565,
+      "learning_rate": 1e-06,
+      "loss": 0.4263,
+      "step": 4977
+    },
+    {
+      "epoch": 0.9519074481307964,
+      "grad_norm": 4.074614786300453,
+      "learning_rate": 1e-06,
+      "loss": 0.3331,
+      "step": 4978
+    },
+    {
+      "epoch": 0.9520986710010517,
+      "grad_norm": 3.0279055363321166,
+      "learning_rate": 1e-06,
+      "loss": 0.1516,
+      "step": 4979
+    },
+    {
+      "epoch": 0.952289893871307,
+      "grad_norm": 4.244595625705331,
+      "learning_rate": 1e-06,
+      "loss": 0.4831,
+      "step": 4980
+    },
+    {
+      "epoch": 0.9524811167415623,
+      "grad_norm": 3.6553916249594556,
+      "learning_rate": 1e-06,
+      "loss": 0.2921,
+      "step": 4981
+    },
+    {
+      "epoch": 0.9526723396118175,
+      "grad_norm": 4.526213273576532,
+      "learning_rate": 1e-06,
+      "loss": 0.1697,
+      "step": 4982
+    },
+    {
+      "epoch": 0.9528635624820728,
+      "grad_norm": 3.4735409754253634,
+      "learning_rate": 1e-06,
+      "loss": 0.1639,
+      "step": 4983
+    },
+    {
+      "epoch": 0.9530547853523281,
+      "grad_norm": 5.7972944986983945,
+      "learning_rate": 1e-06,
+      "loss": 0.1427,
+      "step": 4984
+    },
+    {
+      "epoch": 0.9532460082225834,
+      "grad_norm": 3.0051649455217064,
+      "learning_rate": 1e-06,
+      "loss": 0.0695,
+      "step": 4985
+    },
+    {
+      "epoch": 0.9534372310928387,
+      "grad_norm": 2.4892269714767563,
+      "learning_rate": 1e-06,
+      "loss": 0.0782,
+      "step": 4986
+    },
+    {
+      "epoch": 0.953628453963094,
+      "grad_norm": 4.628449674212603,
+      "learning_rate": 1e-06,
+      "loss": 0.1096,
+      "step": 4987
+    },
+    {
+      "epoch": 0.9538196768333492,
+      "grad_norm": 9.159671079235606,
+      "learning_rate": 1e-06,
+      "loss": 0.2102,
+      "step": 4988
+    },
+    {
+      "epoch": 0.9540108997036045,
+      "grad_norm": 3.5513409298323277,
+      "learning_rate": 1e-06,
+      "loss": 0.1934,
+      "step": 4989
+    },
+    {
+      "epoch": 0.9542021225738598,
+      "grad_norm": 4.049926549928339,
+      "learning_rate": 1e-06,
+      "loss": 0.4053,
+      "step": 4990
+    },
+    {
+      "epoch": 0.9543933454441151,
+      "grad_norm": 3.8499914838027895,
+      "learning_rate": 1e-06,
+      "loss": 0.2051,
+      "step": 4991
+    },
+    {
+      "epoch": 0.9545845683143704,
+      "grad_norm": 2.201287118489395,
+      "learning_rate": 1e-06,
+      "loss": 0.2502,
+      "step": 4992
+    },
+    {
+      "epoch": 0.9547757911846256,
+      "grad_norm": 4.487225945480238,
+      "learning_rate": 1e-06,
+      "loss": 0.3212,
+      "step": 4993
+    },
+    {
+      "epoch": 0.9549670140548809,
+      "grad_norm": 4.603176073670218,
+      "learning_rate": 1e-06,
+      "loss": 0.2469,
+      "step": 4994
+    },
+    {
+      "epoch": 0.9551582369251362,
+      "grad_norm": 2.0006240824703556,
+      "learning_rate": 1e-06,
+      "loss": 0.0505,
+      "step": 4995
+    },
+    {
+      "epoch": 0.9553494597953915,
+      "grad_norm": 3.426153843451368,
+      "learning_rate": 1e-06,
+      "loss": 0.2252,
+      "step": 4996
+    },
+    {
+      "epoch": 0.9555406826656468,
+      "grad_norm": 2.925937867397329,
+      "learning_rate": 1e-06,
+      "loss": 0.0962,
+      "step": 4997
+    },
+    {
+      "epoch": 0.9557319055359021,
+      "grad_norm": 4.202267924566313,
+      "learning_rate": 1e-06,
+      "loss": 0.0836,
+      "step": 4998
+    },
+    {
+      "epoch": 0.9559231284061573,
+      "grad_norm": 4.452182388929276,
+      "learning_rate": 1e-06,
+      "loss": 0.1588,
+      "step": 4999
+    },
+    {
+      "epoch": 0.9561143512764126,
+      "grad_norm": 4.250284746392298,
+      "learning_rate": 1e-06,
+      "loss": 0.0982,
+      "step": 5000
+    },
+    {
+      "epoch": 0.9561143512764126,
+      "eval_runtime": 720.9848,
+      "eval_samples_per_second": 2.128,
+      "eval_steps_per_second": 0.533,
+      "step": 5000
+    },
+    {
+      "epoch": 0.956305574146668,
+      "grad_norm": 3.893773287340701,
+      "learning_rate": 1e-06,
+      "loss": 0.2052,
+      "step": 5001
+    },
+    {
+      "epoch": 0.9564967970169233,
+      "grad_norm": 4.620878703127109,
+      "learning_rate": 1e-06,
+      "loss": 0.4605,
+      "step": 5002
+    },
+    {
+      "epoch": 0.9566880198871786,
+      "grad_norm": 4.601117097965921,
+      "learning_rate": 1e-06,
+      "loss": 0.4317,
+      "step": 5003
+    },
+    {
+      "epoch": 0.9568792427574337,
+      "grad_norm": 3.4851682298903746,
+      "learning_rate": 1e-06,
+      "loss": 0.2666,
+      "step": 5004
+    },
+    {
+      "epoch": 0.957070465627689,
+      "grad_norm": 3.5254528431144045,
+      "learning_rate": 1e-06,
+      "loss": 0.3047,
+      "step": 5005
+    },
+    {
+      "epoch": 0.9572616884979444,
+      "grad_norm": 6.0962597495981345,
+      "learning_rate": 1e-06,
+      "loss": 0.222,
+      "step": 5006
+    },
+    {
+      "epoch": 0.9574529113681997,
+      "grad_norm": 2.686584937869535,
+      "learning_rate": 1e-06,
+      "loss": 0.2338,
+      "step": 5007
+    },
+    {
+      "epoch": 0.957644134238455,
+      "grad_norm": 4.075259316273389,
+      "learning_rate": 1e-06,
+      "loss": 0.2208,
+      "step": 5008
+    },
+    {
+      "epoch": 0.9578353571087102,
+      "grad_norm": 2.7019613771311204,
+      "learning_rate": 1e-06,
+      "loss": 0.1701,
+      "step": 5009
+    },
+    {
+      "epoch": 0.9580265799789655,
+      "grad_norm": 4.243949622514092,
+      "learning_rate": 1e-06,
+      "loss": 0.1093,
+      "step": 5010
+    },
+    {
+      "epoch": 0.9582178028492208,
+      "grad_norm": 4.023416167267475,
+      "learning_rate": 1e-06,
+      "loss": 0.1414,
+      "step": 5011
+    },
+    {
+      "epoch": 0.9584090257194761,
+      "grad_norm": 3.5790766487451444,
+      "learning_rate": 1e-06,
+      "loss": 0.0987,
+      "step": 5012
+    },
+    {
+      "epoch": 0.9586002485897314,
+      "grad_norm": 5.941716674480105,
+      "learning_rate": 1e-06,
+      "loss": 0.2096,
+      "step": 5013
+    },
+    {
+      "epoch": 0.9587914714599867,
+      "grad_norm": 3.349366506774692,
+      "learning_rate": 1e-06,
+      "loss": 0.1542,
+      "step": 5014
+    },
+    {
+      "epoch": 0.9589826943302419,
+      "grad_norm": 8.162984472037335,
+      "learning_rate": 1e-06,
+      "loss": 0.676,
+      "step": 5015
+    },
+    {
+      "epoch": 0.9591739172004972,
+      "grad_norm": 4.183931110930918,
+      "learning_rate": 1e-06,
+      "loss": 0.2452,
+      "step": 5016
+    },
+    {
+      "epoch": 0.9593651400707525,
+      "grad_norm": 1.7675790691900102,
+      "learning_rate": 1e-06,
+      "loss": 0.1387,
+      "step": 5017
+    },
+    {
+      "epoch": 0.9595563629410078,
+      "grad_norm": 6.08472607838468,
+      "learning_rate": 1e-06,
+      "loss": 0.4339,
+      "step": 5018
+    },
+    {
+      "epoch": 0.9597475858112631,
+      "grad_norm": 4.426229213524666,
+      "learning_rate": 1e-06,
+      "loss": 0.4419,
+      "step": 5019
+    },
+    {
+      "epoch": 0.9599388086815183,
+      "grad_norm": 2.444691246992417,
+      "learning_rate": 1e-06,
+      "loss": 0.188,
+      "step": 5020
+    },
+    {
+      "epoch": 0.9601300315517736,
+      "grad_norm": 3.685622302942711,
+      "learning_rate": 1e-06,
+      "loss": 0.2726,
+      "step": 5021
+    },
+    {
+      "epoch": 0.9603212544220289,
+      "grad_norm": 5.565792181041173,
+      "learning_rate": 1e-06,
+      "loss": 0.1817,
+      "step": 5022
+    },
+    {
+      "epoch": 0.9605124772922842,
+      "grad_norm": 2.640154012175769,
+      "learning_rate": 1e-06,
+      "loss": 0.1153,
+      "step": 5023
+    },
+    {
+      "epoch": 0.9607037001625395,
+      "grad_norm": 2.450777424841861,
+      "learning_rate": 1e-06,
+      "loss": 0.095,
+      "step": 5024
+    },
+    {
+      "epoch": 0.9608949230327947,
+      "grad_norm": 2.489203122100515,
+      "learning_rate": 1e-06,
+      "loss": 0.0543,
+      "step": 5025
+    },
+    {
+      "epoch": 0.96108614590305,
+      "grad_norm": 5.493634615262535,
+      "learning_rate": 1e-06,
+      "loss": 0.1254,
+      "step": 5026
+    },
+    {
+      "epoch": 0.9612773687733053,
+      "grad_norm": 3.0803454760217965,
+      "learning_rate": 1e-06,
+      "loss": 0.3706,
+      "step": 5027
+    },
+    {
+      "epoch": 0.9614685916435606,
+      "grad_norm": 4.548134960201922,
+      "learning_rate": 1e-06,
+      "loss": 0.2501,
+      "step": 5028
+    },
+    {
+      "epoch": 0.9616598145138159,
+      "grad_norm": 2.391242253021278,
+      "learning_rate": 1e-06,
+      "loss": 0.2839,
+      "step": 5029
+    },
+    {
+      "epoch": 0.9618510373840712,
+      "grad_norm": 1.73371861039551,
+      "learning_rate": 1e-06,
+      "loss": 0.1923,
+      "step": 5030
+    },
+    {
+      "epoch": 0.9620422602543264,
+      "grad_norm": 3.885172136536241,
+      "learning_rate": 1e-06,
+      "loss": 0.2695,
+      "step": 5031
+    },
+    {
+      "epoch": 0.9622334831245817,
+      "grad_norm": 2.866891455893133,
+      "learning_rate": 1e-06,
+      "loss": 0.1349,
+      "step": 5032
+    },
+    {
+      "epoch": 0.962424705994837,
+      "grad_norm": 2.424664767188621,
+      "learning_rate": 1e-06,
+      "loss": 0.0887,
+      "step": 5033
+    },
+    {
+      "epoch": 0.9626159288650923,
+      "grad_norm": 2.7933194027039696,
+      "learning_rate": 1e-06,
+      "loss": 0.1558,
+      "step": 5034
+    },
+    {
+      "epoch": 0.9628071517353476,
+      "grad_norm": 2.6706117019304383,
+      "learning_rate": 1e-06,
+      "loss": 0.0607,
+      "step": 5035
+    },
+    {
+      "epoch": 0.9629983746056028,
+      "grad_norm": 2.6536156550005714,
+      "learning_rate": 1e-06,
+      "loss": 0.0367,
+      "step": 5036
+    },
+    {
+      "epoch": 0.9631895974758581,
+      "grad_norm": 4.764641732259834,
+      "learning_rate": 1e-06,
+      "loss": 0.1413,
+      "step": 5037
+    },
+    {
+      "epoch": 0.9633808203461134,
+      "grad_norm": 4.548623499315522,
+      "learning_rate": 1e-06,
+      "loss": 0.1568,
+      "step": 5038
+    },
+    {
+      "epoch": 0.9635720432163687,
+      "grad_norm": 4.915084756797578,
+      "learning_rate": 1e-06,
+      "loss": 0.4637,
+      "step": 5039
+    },
+    {
+      "epoch": 0.963763266086624,
+      "grad_norm": 3.9472380820723876,
+      "learning_rate": 1e-06,
+      "loss": 0.2488,
+      "step": 5040
+    },
+    {
+      "epoch": 0.9639544889568793,
+      "grad_norm": 2.395274807607925,
+      "learning_rate": 1e-06,
+      "loss": 0.0689,
+      "step": 5041
+    },
+    {
+      "epoch": 0.9641457118271345,
+      "grad_norm": 3.233846132835026,
+      "learning_rate": 1e-06,
+      "loss": 0.1814,
+      "step": 5042
+    },
+    {
+      "epoch": 0.9643369346973898,
+      "grad_norm": 4.260127398954434,
+      "learning_rate": 1e-06,
+      "loss": 0.3066,
+      "step": 5043
+    },
+    {
+      "epoch": 0.9645281575676451,
+      "grad_norm": 4.335384299239772,
+      "learning_rate": 1e-06,
+      "loss": 0.2364,
+      "step": 5044
+    },
+    {
+      "epoch": 0.9647193804379004,
+      "grad_norm": 3.5070800016451655,
+      "learning_rate": 1e-06,
+      "loss": 0.258,
+      "step": 5045
+    },
+    {
+      "epoch": 0.9649106033081557,
+      "grad_norm": 3.0315606656101908,
+      "learning_rate": 1e-06,
+      "loss": 0.3129,
+      "step": 5046
+    },
+    {
+      "epoch": 0.9651018261784109,
+      "grad_norm": 2.7773597169570037,
+      "learning_rate": 1e-06,
+      "loss": 0.2526,
+      "step": 5047
+    },
+    {
+      "epoch": 0.9652930490486662,
+      "grad_norm": 2.4623684068557945,
+      "learning_rate": 1e-06,
+      "loss": 0.0559,
+      "step": 5048
+    },
+    {
+      "epoch": 0.9654842719189215,
+      "grad_norm": 4.017196407182123,
+      "learning_rate": 1e-06,
+      "loss": 0.2765,
+      "step": 5049
+    },
+    {
+      "epoch": 0.9656754947891768,
+      "grad_norm": 7.060666276652911,
+      "learning_rate": 1e-06,
+      "loss": 0.0829,
+      "step": 5050
+    },
+    {
+      "epoch": 0.9658667176594321,
+      "grad_norm": 3.6131731666689357,
+      "learning_rate": 1e-06,
+      "loss": 0.2708,
+      "step": 5051
+    },
+    {
+      "epoch": 0.9660579405296873,
+      "grad_norm": 3.597304203730856,
+      "learning_rate": 1e-06,
+      "loss": 0.2903,
+      "step": 5052
+    },
+    {
+      "epoch": 0.9662491633999426,
+      "grad_norm": 1.9000371151360842,
+      "learning_rate": 1e-06,
+      "loss": 0.1253,
+      "step": 5053
+    },
+    {
+      "epoch": 0.9664403862701979,
+      "grad_norm": 2.7640003669617075,
+      "learning_rate": 1e-06,
+      "loss": 0.2722,
+      "step": 5054
+    },
+    {
+      "epoch": 0.9666316091404532,
+      "grad_norm": 2.219928911806439,
+      "learning_rate": 1e-06,
+      "loss": 0.0907,
+      "step": 5055
+    },
+    {
+      "epoch": 0.9668228320107085,
+      "grad_norm": 4.493801085880737,
+      "learning_rate": 1e-06,
+      "loss": 0.2735,
+      "step": 5056
+    },
+    {
+      "epoch": 0.9670140548809638,
+      "grad_norm": 3.9214304637879556,
+      "learning_rate": 1e-06,
+      "loss": 0.1786,
+      "step": 5057
+    },
+    {
+      "epoch": 0.967205277751219,
+      "grad_norm": 3.9135369045027892,
+      "learning_rate": 1e-06,
+      "loss": 0.2828,
+      "step": 5058
+    },
+    {
+      "epoch": 0.9673965006214743,
+      "grad_norm": 3.996598704961019,
+      "learning_rate": 1e-06,
+      "loss": 0.2022,
+      "step": 5059
+    },
+    {
+      "epoch": 0.9675877234917296,
+      "grad_norm": 3.7152997849110716,
+      "learning_rate": 1e-06,
+      "loss": 0.3866,
+      "step": 5060
+    },
+    {
+      "epoch": 0.9677789463619849,
+      "grad_norm": 3.2979731606524316,
+      "learning_rate": 1e-06,
+      "loss": 0.1204,
+      "step": 5061
+    },
+    {
+      "epoch": 0.9679701692322402,
+      "grad_norm": 2.9984858188443293,
+      "learning_rate": 1e-06,
+      "loss": 0.1003,
+      "step": 5062
+    },
+    {
+      "epoch": 0.9681613921024954,
+      "grad_norm": 4.376097405264902,
+      "learning_rate": 1e-06,
+      "loss": 0.148,
+      "step": 5063
+    },
+    {
+      "epoch": 0.9683526149727507,
+      "grad_norm": 5.974337693541004,
+      "learning_rate": 1e-06,
+      "loss": 0.4376,
+      "step": 5064
+    },
+    {
+      "epoch": 0.968543837843006,
+      "grad_norm": 5.936136510795947,
+      "learning_rate": 1e-06,
+      "loss": 0.5015,
+      "step": 5065
+    },
+    {
+      "epoch": 0.9687350607132613,
+      "grad_norm": 3.0905895080688497,
+      "learning_rate": 1e-06,
+      "loss": 0.1621,
+      "step": 5066
+    },
+    {
+      "epoch": 0.9689262835835166,
+      "grad_norm": 9.992203916009776,
+      "learning_rate": 1e-06,
+      "loss": 0.537,
+      "step": 5067
+    },
+    {
+      "epoch": 0.9691175064537719,
+      "grad_norm": 2.5472829293112396,
+      "learning_rate": 1e-06,
+      "loss": 0.1445,
+      "step": 5068
+    },
+    {
+      "epoch": 0.9693087293240271,
+      "grad_norm": 2.029286183539268,
+      "learning_rate": 1e-06,
+      "loss": 0.0735,
+      "step": 5069
+    },
+    {
+      "epoch": 0.9694999521942824,
+      "grad_norm": 2.139667380321784,
+      "learning_rate": 1e-06,
+      "loss": 0.0626,
+      "step": 5070
+    },
+    {
+      "epoch": 0.9696911750645377,
+      "grad_norm": 2.9016963601240953,
+      "learning_rate": 1e-06,
+      "loss": 0.0766,
+      "step": 5071
+    },
+    {
+      "epoch": 0.969882397934793,
+      "grad_norm": 3.492737728867051,
+      "learning_rate": 1e-06,
+      "loss": 0.0941,
+      "step": 5072
+    },
+    {
+      "epoch": 0.9700736208050483,
+      "grad_norm": 3.41057403207605,
+      "learning_rate": 1e-06,
+      "loss": 0.0691,
+      "step": 5073
+    },
+    {
+      "epoch": 0.9702648436753035,
+      "grad_norm": 6.0592784384026945,
+      "learning_rate": 1e-06,
+      "loss": 0.1847,
+      "step": 5074
+    },
+    {
+      "epoch": 0.9704560665455588,
+      "grad_norm": 3.130395732396505,
+      "learning_rate": 1e-06,
+      "loss": 0.0776,
+      "step": 5075
+    },
+    {
+      "epoch": 0.9706472894158141,
+      "grad_norm": 5.325568964833179,
+      "learning_rate": 1e-06,
+      "loss": 0.165,
+      "step": 5076
+    },
+    {
+      "epoch": 0.9708385122860694,
+      "grad_norm": 3.6956052764980427,
+      "learning_rate": 1e-06,
+      "loss": 0.3951,
+      "step": 5077
+    },
+    {
+      "epoch": 0.9710297351563247,
+      "grad_norm": 2.7466987388398785,
+      "learning_rate": 1e-06,
+      "loss": 0.1767,
+      "step": 5078
+    },
+    {
+      "epoch": 0.9712209580265799,
+      "grad_norm": 2.353048064454705,
+      "learning_rate": 1e-06,
+      "loss": 0.2891,
+      "step": 5079
+    },
+    {
+      "epoch": 0.9714121808968352,
+      "grad_norm": 2.739408731638017,
+      "learning_rate": 1e-06,
+      "loss": 0.2805,
+      "step": 5080
+    },
+    {
+      "epoch": 0.9716034037670905,
+      "grad_norm": 4.348508136160482,
+      "learning_rate": 1e-06,
+      "loss": 0.2776,
+      "step": 5081
+    },
+    {
+      "epoch": 0.9717946266373458,
+      "grad_norm": 2.6776308224700527,
+      "learning_rate": 1e-06,
+      "loss": 0.0707,
+      "step": 5082
+    },
+    {
+      "epoch": 0.9719858495076011,
+      "grad_norm": 3.832189154215131,
+      "learning_rate": 1e-06,
+      "loss": 0.1243,
+      "step": 5083
+    },
+    {
+      "epoch": 0.9721770723778564,
+      "grad_norm": 1.8367345500154504,
+      "learning_rate": 1e-06,
+      "loss": 0.1154,
+      "step": 5084
+    },
+    {
+      "epoch": 0.9723682952481116,
+      "grad_norm": 3.0544720742052314,
+      "learning_rate": 1e-06,
+      "loss": 0.1205,
+      "step": 5085
+    },
+    {
+      "epoch": 0.9725595181183669,
+      "grad_norm": 15.496705720365014,
+      "learning_rate": 1e-06,
+      "loss": 0.1402,
+      "step": 5086
+    },
+    {
+      "epoch": 0.9727507409886222,
+      "grad_norm": 3.809986098657113,
+      "learning_rate": 1e-06,
+      "loss": 0.1104,
+      "step": 5087
+    },
+    {
+      "epoch": 0.9729419638588775,
+      "grad_norm": 6.585170859609046,
+      "learning_rate": 1e-06,
+      "loss": 0.1614,
+      "step": 5088
+    },
+    {
+      "epoch": 0.9731331867291328,
+      "grad_norm": 4.946677262191187,
+      "learning_rate": 1e-06,
+      "loss": 0.1873,
+      "step": 5089
+    },
+    {
+      "epoch": 0.973324409599388,
+      "grad_norm": 3.818690292965491,
+      "learning_rate": 1e-06,
+      "loss": 0.4412,
+      "step": 5090
+    },
+    {
+      "epoch": 0.9735156324696433,
+      "grad_norm": 3.591368043408775,
+      "learning_rate": 1e-06,
+      "loss": 0.3774,
+      "step": 5091
+    },
+    {
+      "epoch": 0.9737068553398986,
+      "grad_norm": 2.6306018094386268,
+      "learning_rate": 1e-06,
+      "loss": 0.1513,
+      "step": 5092
+    },
+    {
+      "epoch": 0.973898078210154,
+      "grad_norm": 2.363629477831007,
+      "learning_rate": 1e-06,
+      "loss": 0.1172,
+      "step": 5093
+    },
+    {
+      "epoch": 0.9740893010804093,
+      "grad_norm": 3.0006992796666894,
+      "learning_rate": 1e-06,
+      "loss": 0.3242,
+      "step": 5094
+    },
+    {
+      "epoch": 0.9742805239506644,
+      "grad_norm": 3.5037501543843197,
+      "learning_rate": 1e-06,
+      "loss": 0.1462,
+      "step": 5095
+    },
+    {
+      "epoch": 0.9744717468209197,
+      "grad_norm": 2.9268534469381136,
+      "learning_rate": 1e-06,
+      "loss": 0.0555,
+      "step": 5096
+    },
+    {
+      "epoch": 0.974662969691175,
+      "grad_norm": 2.9005372700722405,
+      "learning_rate": 1e-06,
+      "loss": 0.1779,
+      "step": 5097
+    },
+    {
+      "epoch": 0.9748541925614304,
+      "grad_norm": 3.8206266496362407,
+      "learning_rate": 1e-06,
+      "loss": 0.2463,
+      "step": 5098
+    },
+    {
+      "epoch": 0.9750454154316857,
+      "grad_norm": 4.379743783502736,
+      "learning_rate": 1e-06,
+      "loss": 0.1091,
+      "step": 5099
+    },
+    {
+      "epoch": 0.975236638301941,
+      "grad_norm": 5.130905098887434,
+      "learning_rate": 1e-06,
+      "loss": 0.082,
+      "step": 5100
+    },
+    {
+      "epoch": 0.9754278611721962,
+      "grad_norm": 5.768461692606814,
+      "learning_rate": 1e-06,
+      "loss": 0.1297,
+      "step": 5101
+    },
+    {
+      "epoch": 0.9756190840424515,
+      "grad_norm": 2.990485999884121,
+      "learning_rate": 1e-06,
+      "loss": 0.2032,
+      "step": 5102
+    },
+    {
+      "epoch": 0.9758103069127068,
+      "grad_norm": 4.881211065510736,
+      "learning_rate": 1e-06,
+      "loss": 0.362,
+      "step": 5103
+    },
+    {
+      "epoch": 0.9760015297829621,
+      "grad_norm": 8.023001505772083,
+      "learning_rate": 1e-06,
+      "loss": 0.1579,
+      "step": 5104
+    },
+    {
+      "epoch": 0.9761927526532174,
+      "grad_norm": 4.24193143686499,
+      "learning_rate": 1e-06,
+      "loss": 0.3104,
+      "step": 5105
+    },
+    {
+      "epoch": 0.9763839755234726,
+      "grad_norm": 4.657769077000083,
+      "learning_rate": 1e-06,
+      "loss": 0.1758,
+      "step": 5106
+    },
+    {
+      "epoch": 0.9765751983937279,
+      "grad_norm": 2.6819966768191557,
+      "learning_rate": 1e-06,
+      "loss": 0.0807,
+      "step": 5107
+    },
+    {
+      "epoch": 0.9767664212639832,
+      "grad_norm": 3.6623630120499047,
+      "learning_rate": 1e-06,
+      "loss": 0.1429,
+      "step": 5108
+    },
+    {
+      "epoch": 0.9769576441342385,
+      "grad_norm": 1.908192942034786,
+      "learning_rate": 1e-06,
+      "loss": 0.0885,
+      "step": 5109
+    },
+    {
+      "epoch": 0.9771488670044938,
+      "grad_norm": 5.699331532147439,
+      "learning_rate": 1e-06,
+      "loss": 0.1404,
+      "step": 5110
+    },
+    {
+      "epoch": 0.9773400898747491,
+      "grad_norm": 3.6303040734924155,
+      "learning_rate": 1e-06,
+      "loss": 0.0847,
+      "step": 5111
+    },
+    {
+      "epoch": 0.9775313127450043,
+      "grad_norm": 7.361818592293608,
+      "learning_rate": 1e-06,
+      "loss": 0.1147,
+      "step": 5112
+    },
+    {
+      "epoch": 0.9777225356152596,
+      "grad_norm": 7.389709803106012,
+      "learning_rate": 1e-06,
+      "loss": 0.2934,
+      "step": 5113
+    },
+    {
+      "epoch": 0.9779137584855149,
+      "grad_norm": 4.286427729123862,
+      "learning_rate": 1e-06,
+      "loss": 0.2604,
+      "step": 5114
+    },
+    {
+      "epoch": 0.9781049813557702,
+      "grad_norm": 2.45727887141535,
+      "learning_rate": 1e-06,
+      "loss": 0.3246,
+      "step": 5115
+    },
+    {
+      "epoch": 0.9782962042260255,
+      "grad_norm": 5.186124872154082,
+      "learning_rate": 1e-06,
+      "loss": 0.1833,
+      "step": 5116
+    },
+    {
+      "epoch": 0.9784874270962807,
+      "grad_norm": 4.815446026228365,
+      "learning_rate": 1e-06,
+      "loss": 0.2777,
+      "step": 5117
+    },
+    {
+      "epoch": 0.978678649966536,
+      "grad_norm": 2.2196706017129304,
+      "learning_rate": 1e-06,
+      "loss": 0.213,
+      "step": 5118
+    },
+    {
+      "epoch": 0.9788698728367913,
+      "grad_norm": 3.554636927653648,
+      "learning_rate": 1e-06,
+      "loss": 0.0948,
+      "step": 5119
+    },
+    {
+      "epoch": 0.9790610957070466,
+      "grad_norm": 2.2484099809925575,
+      "learning_rate": 1e-06,
+      "loss": 0.1499,
+      "step": 5120
+    },
+    {
+      "epoch": 0.9792523185773019,
+      "grad_norm": 3.534305162355375,
+      "learning_rate": 1e-06,
+      "loss": 0.1469,
+      "step": 5121
+    },
+    {
+      "epoch": 0.9794435414475571,
+      "grad_norm": 4.808394823832622,
+      "learning_rate": 1e-06,
+      "loss": 0.1388,
+      "step": 5122
+    },
+    {
+      "epoch": 0.9796347643178124,
+      "grad_norm": 4.128200791938839,
+      "learning_rate": 1e-06,
+      "loss": 0.103,
+      "step": 5123
+    },
+    {
+      "epoch": 0.9798259871880677,
+      "grad_norm": 3.9857661193586784,
+      "learning_rate": 1e-06,
+      "loss": 0.255,
+      "step": 5124
+    },
+    {
+      "epoch": 0.980017210058323,
+      "grad_norm": 6.511319208273486,
+      "learning_rate": 1e-06,
+      "loss": 0.1367,
+      "step": 5125
+    },
+    {
+      "epoch": 0.9802084329285783,
+      "grad_norm": 6.265458482922116,
+      "learning_rate": 1e-06,
+      "loss": 0.3733,
+      "step": 5126
+    },
+    {
+      "epoch": 0.9803996557988336,
+      "grad_norm": 4.426333494823617,
+      "learning_rate": 1e-06,
+      "loss": 0.412,
+      "step": 5127
+    },
+    {
+      "epoch": 0.9805908786690888,
+      "grad_norm": 3.589384130499172,
+      "learning_rate": 1e-06,
+      "loss": 0.1736,
+      "step": 5128
+    },
+    {
+      "epoch": 0.9807821015393441,
+      "grad_norm": 4.614530105226427,
+      "learning_rate": 1e-06,
+      "loss": 0.2654,
+      "step": 5129
+    },
+    {
+      "epoch": 0.9809733244095994,
+      "grad_norm": 6.959529507398863,
+      "learning_rate": 1e-06,
+      "loss": 0.135,
+      "step": 5130
+    },
+    {
+      "epoch": 0.9811645472798547,
+      "grad_norm": 3.3742960442965972,
+      "learning_rate": 1e-06,
+      "loss": 0.2433,
+      "step": 5131
+    },
+    {
+      "epoch": 0.98135577015011,
+      "grad_norm": 3.0996695250051425,
+      "learning_rate": 1e-06,
+      "loss": 0.2215,
+      "step": 5132
+    },
+    {
+      "epoch": 0.9815469930203652,
+      "grad_norm": 2.290928103180912,
+      "learning_rate": 1e-06,
+      "loss": 0.1276,
+      "step": 5133
+    },
+    {
+      "epoch": 0.9817382158906205,
+      "grad_norm": 3.4766501426365055,
+      "learning_rate": 1e-06,
+      "loss": 0.0694,
+      "step": 5134
+    },
+    {
+      "epoch": 0.9819294387608758,
+      "grad_norm": 3.1546843016231776,
+      "learning_rate": 1e-06,
+      "loss": 0.206,
+      "step": 5135
+    },
+    {
+      "epoch": 0.9821206616311311,
+      "grad_norm": 2.933442417919685,
+      "learning_rate": 1e-06,
+      "loss": 0.0809,
+      "step": 5136
+    },
+    {
+      "epoch": 0.9823118845013864,
+      "grad_norm": 5.983418121728883,
+      "learning_rate": 1e-06,
+      "loss": 0.159,
+      "step": 5137
+    },
+    {
+      "epoch": 0.9825031073716416,
+      "grad_norm": 6.276009418409974,
+      "learning_rate": 1e-06,
+      "loss": 0.2065,
+      "step": 5138
+    },
+    {
+      "epoch": 0.9826943302418969,
+      "grad_norm": 4.81603042113876,
+      "learning_rate": 1e-06,
+      "loss": 0.386,
+      "step": 5139
+    },
+    {
+      "epoch": 0.9828855531121522,
+      "grad_norm": 4.559839387374018,
+      "learning_rate": 1e-06,
+      "loss": 0.5429,
+      "step": 5140
+    },
+    {
+      "epoch": 0.9830767759824075,
+      "grad_norm": 1.7719663996729238,
+      "learning_rate": 1e-06,
+      "loss": 0.1321,
+      "step": 5141
+    },
+    {
+      "epoch": 0.9832679988526628,
+      "grad_norm": 4.1419809100688765,
+      "learning_rate": 1e-06,
+      "loss": 0.2576,
+      "step": 5142
+    },
+    {
+      "epoch": 0.9834592217229181,
+      "grad_norm": 2.5032439166716776,
+      "learning_rate": 1e-06,
+      "loss": 0.28,
+      "step": 5143
+    },
+    {
+      "epoch": 0.9836504445931733,
+      "grad_norm": 2.0270131453476945,
+      "learning_rate": 1e-06,
+      "loss": 0.1437,
+      "step": 5144
+    },
+    {
+      "epoch": 0.9838416674634286,
+      "grad_norm": 2.429440893678561,
+      "learning_rate": 1e-06,
+      "loss": 0.121,
+      "step": 5145
+    },
+    {
+      "epoch": 0.9840328903336839,
+      "grad_norm": 3.6790739872204776,
+      "learning_rate": 1e-06,
+      "loss": 0.083,
+      "step": 5146
+    },
+    {
+      "epoch": 0.9842241132039392,
+      "grad_norm": 4.515328051033199,
+      "learning_rate": 1e-06,
+      "loss": 0.2184,
+      "step": 5147
+    },
+    {
+      "epoch": 0.9844153360741945,
+      "grad_norm": 4.804251004372346,
+      "learning_rate": 1e-06,
+      "loss": 0.1434,
+      "step": 5148
+    },
+    {
+      "epoch": 0.9846065589444497,
+      "grad_norm": 2.9305763625560863,
+      "learning_rate": 1e-06,
+      "loss": 0.2984,
+      "step": 5149
+    },
+    {
+      "epoch": 0.984797781814705,
+      "grad_norm": 4.987273231647668,
+      "learning_rate": 1e-06,
+      "loss": 0.0704,
+      "step": 5150
+    },
+    {
+      "epoch": 0.9849890046849603,
+      "grad_norm": 3.018368910592628,
+      "learning_rate": 1e-06,
+      "loss": 0.1091,
+      "step": 5151
+    },
+    {
+      "epoch": 0.9851802275552156,
+      "grad_norm": 4.107607849525487,
+      "learning_rate": 1e-06,
+      "loss": 0.3421,
+      "step": 5152
+    },
+    {
+      "epoch": 0.9853714504254709,
+      "grad_norm": 3.6190951694983418,
+      "learning_rate": 1e-06,
+      "loss": 0.18,
+      "step": 5153
+    },
+    {
+      "epoch": 0.9855626732957262,
+      "grad_norm": 3.756521149812259,
+      "learning_rate": 1e-06,
+      "loss": 0.0846,
+      "step": 5154
+    },
+    {
+      "epoch": 0.9857538961659814,
+      "grad_norm": 3.61183683650814,
+      "learning_rate": 1e-06,
+      "loss": 0.3654,
+      "step": 5155
+    },
+    {
+      "epoch": 0.9859451190362367,
+      "grad_norm": 4.8254232902844825,
+      "learning_rate": 1e-06,
+      "loss": 0.119,
+      "step": 5156
+    },
+    {
+      "epoch": 0.986136341906492,
+      "grad_norm": 2.0927590829389877,
+      "learning_rate": 1e-06,
+      "loss": 0.1782,
+      "step": 5157
+    },
+    {
+      "epoch": 0.9863275647767473,
+      "grad_norm": 3.3644135769909465,
+      "learning_rate": 1e-06,
+      "loss": 0.1168,
+      "step": 5158
+    },
+    {
+      "epoch": 0.9865187876470026,
+      "grad_norm": 3.426196152585075,
+      "learning_rate": 1e-06,
+      "loss": 0.1292,
+      "step": 5159
+    },
+    {
+      "epoch": 0.9867100105172578,
+      "grad_norm": 7.543169268842388,
+      "learning_rate": 1e-06,
+      "loss": 0.0521,
+      "step": 5160
+    },
+    {
+      "epoch": 0.9869012333875131,
+      "grad_norm": 3.872161194551352,
+      "learning_rate": 1e-06,
+      "loss": 0.0723,
+      "step": 5161
+    },
+    {
+      "epoch": 0.9870924562577684,
+      "grad_norm": 2.480440587161726,
+      "learning_rate": 1e-06,
+      "loss": 0.0482,
+      "step": 5162
+    },
+    {
+      "epoch": 0.9872836791280237,
+      "grad_norm": 5.44606921330409,
+      "learning_rate": 1e-06,
+      "loss": 0.2525,
+      "step": 5163
+    },
+    {
+      "epoch": 0.987474901998279,
+      "grad_norm": 3.6063138853658754,
+      "learning_rate": 1e-06,
+      "loss": 0.3108,
+      "step": 5164
+    },
+    {
+      "epoch": 0.9876661248685342,
+      "grad_norm": 4.247507374146938,
+      "learning_rate": 1e-06,
+      "loss": 0.4639,
+      "step": 5165
+    },
+    {
+      "epoch": 0.9878573477387895,
+      "grad_norm": 6.067918381445709,
+      "learning_rate": 1e-06,
+      "loss": 0.8219,
+      "step": 5166
+    },
+    {
+      "epoch": 0.9880485706090448,
+      "grad_norm": 3.9910924197139,
+      "learning_rate": 1e-06,
+      "loss": 0.355,
+      "step": 5167
+    },
+    {
+      "epoch": 0.9882397934793001,
+      "grad_norm": 2.0758218263817376,
+      "learning_rate": 1e-06,
+      "loss": 0.1579,
+      "step": 5168
+    },
+    {
+      "epoch": 0.9884310163495554,
+      "grad_norm": 4.193195398444501,
+      "learning_rate": 1e-06,
+      "loss": 0.2998,
+      "step": 5169
+    },
+    {
+      "epoch": 0.9886222392198107,
+      "grad_norm": 1.9132969162074902,
+      "learning_rate": 1e-06,
+      "loss": 0.1008,
+      "step": 5170
+    },
+    {
+      "epoch": 0.9888134620900659,
+      "grad_norm": 3.4040974106080166,
+      "learning_rate": 1e-06,
+      "loss": 0.0887,
+      "step": 5171
+    },
+    {
+      "epoch": 0.9890046849603212,
+      "grad_norm": 3.8665380067357074,
+      "learning_rate": 1e-06,
+      "loss": 0.2293,
+      "step": 5172
+    },
+    {
+      "epoch": 0.9891959078305765,
+      "grad_norm": 3.885815692828574,
+      "learning_rate": 1e-06,
+      "loss": 0.094,
+      "step": 5173
+    },
+    {
+      "epoch": 0.9893871307008318,
+      "grad_norm": 3.495937714744228,
+      "learning_rate": 1e-06,
+      "loss": 0.1037,
+      "step": 5174
+    },
+    {
+      "epoch": 0.9895783535710871,
+      "grad_norm": 14.141306419182456,
+      "learning_rate": 1e-06,
+      "loss": 0.2238,
+      "step": 5175
+    },
+    {
+      "epoch": 0.9897695764413423,
+      "grad_norm": 5.563195345847086,
+      "learning_rate": 1e-06,
+      "loss": 0.2886,
+      "step": 5176
+    },
+    {
+      "epoch": 0.9899607993115976,
+      "grad_norm": 3.0995329166589984,
+      "learning_rate": 1e-06,
+      "loss": 0.1619,
+      "step": 5177
+    },
+    {
+      "epoch": 0.9901520221818529,
+      "grad_norm": 1.9737298140360289,
+      "learning_rate": 1e-06,
+      "loss": 0.3278,
+      "step": 5178
+    },
+    {
+      "epoch": 0.9903432450521082,
+      "grad_norm": 3.093781750448661,
+      "learning_rate": 1e-06,
+      "loss": 0.1921,
+      "step": 5179
+    },
+    {
+      "epoch": 0.9905344679223635,
+      "grad_norm": 4.294881684313431,
+      "learning_rate": 1e-06,
+      "loss": 0.3125,
+      "step": 5180
+    },
+    {
+      "epoch": 0.9907256907926189,
+      "grad_norm": 3.825328069841716,
+      "learning_rate": 1e-06,
+      "loss": 0.1569,
+      "step": 5181
+    },
+    {
+      "epoch": 0.990916913662874,
+      "grad_norm": 4.425444654979555,
+      "learning_rate": 1e-06,
+      "loss": 0.1953,
+      "step": 5182
+    },
+    {
+      "epoch": 0.9911081365331293,
+      "grad_norm": 3.33138777857667,
+      "learning_rate": 1e-06,
+      "loss": 0.2802,
+      "step": 5183
+    },
+    {
+      "epoch": 0.9912993594033847,
+      "grad_norm": 2.996603155286979,
+      "learning_rate": 1e-06,
+      "loss": 0.0495,
+      "step": 5184
+    },
+    {
+      "epoch": 0.99149058227364,
+      "grad_norm": 1.4598972472520728,
+      "learning_rate": 1e-06,
+      "loss": 0.0585,
+      "step": 5185
+    },
+    {
+      "epoch": 0.9916818051438953,
+      "grad_norm": 3.64791742248934,
+      "learning_rate": 1e-06,
+      "loss": 0.083,
+      "step": 5186
+    },
+    {
+      "epoch": 0.9918730280141504,
+      "grad_norm": 7.851384190294937,
+      "learning_rate": 1e-06,
+      "loss": 0.1299,
+      "step": 5187
+    },
+    {
+      "epoch": 0.9920642508844058,
+      "grad_norm": 8.330737765133756,
+      "learning_rate": 1e-06,
+      "loss": 0.2726,
+      "step": 5188
+    },
+    {
+      "epoch": 0.992255473754661,
+      "grad_norm": 5.771110592274395,
+      "learning_rate": 1e-06,
+      "loss": 0.3954,
+      "step": 5189
+    },
+    {
+      "epoch": 0.9924466966249164,
+      "grad_norm": 3.543064170822608,
+      "learning_rate": 1e-06,
+      "loss": 0.4873,
+      "step": 5190
+    },
+    {
+      "epoch": 0.9926379194951717,
+      "grad_norm": 3.7454039064790403,
+      "learning_rate": 1e-06,
+      "loss": 0.3328,
+      "step": 5191
+    },
+    {
+      "epoch": 0.9928291423654269,
+      "grad_norm": 2.392613500216798,
+      "learning_rate": 1e-06,
+      "loss": 0.2761,
+      "step": 5192
+    },
+    {
+      "epoch": 0.9930203652356822,
+      "grad_norm": 3.1331342498624495,
+      "learning_rate": 1e-06,
+      "loss": 0.1942,
+      "step": 5193
+    },
+    {
+      "epoch": 0.9932115881059375,
+      "grad_norm": 4.459928328276069,
+      "learning_rate": 1e-06,
+      "loss": 0.4141,
+      "step": 5194
+    },
+    {
+      "epoch": 0.9934028109761928,
+      "grad_norm": 4.375876420655929,
+      "learning_rate": 1e-06,
+      "loss": 0.3205,
+      "step": 5195
+    },
+    {
+      "epoch": 0.9935940338464481,
+      "grad_norm": 2.163551132433673,
+      "learning_rate": 1e-06,
+      "loss": 0.1796,
+      "step": 5196
+    },
+    {
+      "epoch": 0.9937852567167034,
+      "grad_norm": 2.7146508896199983,
+      "learning_rate": 1e-06,
+      "loss": 0.0706,
+      "step": 5197
+    },
+    {
+      "epoch": 0.9939764795869586,
+      "grad_norm": 3.3542572072447747,
+      "learning_rate": 1e-06,
+      "loss": 0.2306,
+      "step": 5198
+    },
+    {
+      "epoch": 0.9941677024572139,
+      "grad_norm": 1.7833429722510603,
+      "learning_rate": 1e-06,
+      "loss": 0.0477,
+      "step": 5199
+    },
+    {
+      "epoch": 0.9943589253274692,
+      "grad_norm": 4.262919144144843,
+      "learning_rate": 1e-06,
+      "loss": 0.0796,
+      "step": 5200
+    },
+    {
+      "epoch": 0.9945501481977245,
+      "grad_norm": 5.631802662816779,
+      "learning_rate": 1e-06,
+      "loss": 0.2585,
+      "step": 5201
+    },
+    {
+      "epoch": 0.9947413710679798,
+      "grad_norm": 5.788006767348081,
+      "learning_rate": 1e-06,
+      "loss": 0.3468,
+      "step": 5202
+    },
+    {
+      "epoch": 0.994932593938235,
+      "grad_norm": 4.795141845132572,
+      "learning_rate": 1e-06,
+      "loss": 0.4457,
+      "step": 5203
+    },
+    {
+      "epoch": 0.9951238168084903,
+      "grad_norm": 4.188515838844468,
+      "learning_rate": 1e-06,
+      "loss": 0.2001,
+      "step": 5204
+    },
+    {
+      "epoch": 0.9953150396787456,
+      "grad_norm": 4.570940753760614,
+      "learning_rate": 1e-06,
+      "loss": 0.3501,
+      "step": 5205
+    },
+    {
+      "epoch": 0.9955062625490009,
+      "grad_norm": 3.113929365176512,
+      "learning_rate": 1e-06,
+      "loss": 0.2083,
+      "step": 5206
+    },
+    {
+      "epoch": 0.9956974854192562,
+      "grad_norm": 7.0152038583750755,
+      "learning_rate": 1e-06,
+      "loss": 0.1716,
+      "step": 5207
+    },
+    {
+      "epoch": 0.9958887082895114,
+      "grad_norm": 2.6254645345279775,
+      "learning_rate": 1e-06,
+      "loss": 0.1276,
+      "step": 5208
+    },
+    {
+      "epoch": 0.9960799311597667,
+      "grad_norm": 2.554860427089981,
+      "learning_rate": 1e-06,
+      "loss": 0.1696,
+      "step": 5209
+    },
+    {
+      "epoch": 0.996271154030022,
+      "grad_norm": 2.0588217795388615,
+      "learning_rate": 1e-06,
+      "loss": 0.0718,
+      "step": 5210
+    },
+    {
+      "epoch": 0.9964623769002773,
+      "grad_norm": 2.7031572814072256,
+      "learning_rate": 1e-06,
+      "loss": 0.0702,
+      "step": 5211
+    },
+    {
+      "epoch": 0.9966535997705326,
+      "grad_norm": 3.7714489738110664,
+      "learning_rate": 1e-06,
+      "loss": 0.0612,
+      "step": 5212
+    },
+    {
+      "epoch": 0.9968448226407879,
+      "grad_norm": 8.310348346997104,
+      "learning_rate": 1e-06,
+      "loss": 0.1973,
+      "step": 5213
+    },
+    {
+      "epoch": 0.9970360455110431,
+      "grad_norm": 5.129116173172091,
+      "learning_rate": 1e-06,
+      "loss": 0.6715,
+      "step": 5214
+    },
+    {
+      "epoch": 0.9972272683812984,
+      "grad_norm": 2.8724212691471163,
+      "learning_rate": 1e-06,
+      "loss": 0.1084,
+      "step": 5215
+    },
+    {
+      "epoch": 0.9974184912515537,
+      "grad_norm": 3.1104899952023484,
+      "learning_rate": 1e-06,
+      "loss": 0.3318,
+      "step": 5216
+    },
+    {
+      "epoch": 0.997609714121809,
+      "grad_norm": 1.9741296685443255,
+      "learning_rate": 1e-06,
+      "loss": 0.0646,
+      "step": 5217
+    },
+    {
+      "epoch": 0.9978009369920643,
+      "grad_norm": 1.5977445918424256,
+      "learning_rate": 1e-06,
+      "loss": 0.2218,
+      "step": 5218
+    },
+    {
+      "epoch": 0.9979921598623195,
+      "grad_norm": 3.7327112298308283,
+      "learning_rate": 1e-06,
+      "loss": 0.1794,
+      "step": 5219
+    },
+    {
+      "epoch": 0.9981833827325748,
+      "grad_norm": 3.0922923651942065,
+      "learning_rate": 1e-06,
+      "loss": 0.0765,
+      "step": 5220
+    },
+    {
+      "epoch": 0.9983746056028301,
+      "grad_norm": 2.5091763884542804,
+      "learning_rate": 1e-06,
+      "loss": 0.0431,
+      "step": 5221
+    },
+    {
+      "epoch": 0.9985658284730854,
+      "grad_norm": 4.243124842489917,
+      "learning_rate": 1e-06,
+      "loss": 0.1821,
+      "step": 5222
+    },
+    {
+      "epoch": 0.9987570513433407,
+      "grad_norm": 5.389217469378182,
+      "learning_rate": 1e-06,
+      "loss": 0.1097,
+      "step": 5223
+    },
+    {
+      "epoch": 0.998948274213596,
+      "grad_norm": 4.775115821592094,
+      "learning_rate": 1e-06,
+      "loss": 0.1479,
+      "step": 5224
+    },
+    {
+      "epoch": 0.9991394970838512,
+      "grad_norm": 2.9344739545848277,
+      "learning_rate": 1e-06,
+      "loss": 0.0435,
+      "step": 5225
+    },
+    {
+      "epoch": 0.9993307199541065,
+      "grad_norm": 5.0850334242900965,
+      "learning_rate": 1e-06,
+      "loss": 0.3603,
+      "step": 5226
+    },
+    {
+      "epoch": 0.9995219428243618,
+      "grad_norm": 4.7213113666775,
+      "learning_rate": 1e-06,
+      "loss": 0.5337,
+      "step": 5227
+    },
+    {
+      "epoch": 0.9997131656946171,
+      "grad_norm": 4.428689511214707,
+      "learning_rate": 1e-06,
+      "loss": 0.0706,
+      "step": 5228
+    },
+    {
+      "epoch": 0.9999043885648724,
+      "grad_norm": 7.101807710717727,
+      "learning_rate": 1e-06,
+      "loss": 0.1516,
+      "step": 5229
+    },
+    {
+      "epoch": 1.0000956114351276,
+      "grad_norm": 5.962110091602198,
+      "learning_rate": 1e-06,
+      "loss": 0.3023,
+      "step": 5230
+    },
+    {
+      "epoch": 1.000286834305383,
+      "grad_norm": 3.6716718901604133,
+      "learning_rate": 1e-06,
+      "loss": 0.531,
+      "step": 5231
+    },
+    {
+      "epoch": 1.0004780571756382,
+      "grad_norm": 3.22439524576745,
+      "learning_rate": 1e-06,
+      "loss": 0.3513,
+      "step": 5232
+    },
+    {
+      "epoch": 1.0006692800458934,
+      "grad_norm": 3.685237254938019,
+      "learning_rate": 1e-06,
+      "loss": 0.366,
+      "step": 5233
+    },
+    {
+      "epoch": 1.0008605029161488,
+      "grad_norm": 2.439485768083738,
+      "learning_rate": 1e-06,
+      "loss": 0.1135,
+      "step": 5234
+    },
+    {
+      "epoch": 1.001051725786404,
+      "grad_norm": 2.670432521205372,
+      "learning_rate": 1e-06,
+      "loss": 0.1382,
+      "step": 5235
+    },
+    {
+      "epoch": 1.0012429486566594,
+      "grad_norm": 1.5135212878723487,
+      "learning_rate": 1e-06,
+      "loss": 0.0594,
+      "step": 5236
+    },
+    {
+      "epoch": 1.0014341715269146,
+      "grad_norm": 2.7223228567902784,
+      "learning_rate": 1e-06,
+      "loss": 0.1171,
+      "step": 5237
+    },
+    {
+      "epoch": 1.0016253943971698,
+      "grad_norm": 2.776788884144885,
+      "learning_rate": 1e-06,
+      "loss": 0.1113,
+      "step": 5238
+    },
+    {
+      "epoch": 1.0018166172674252,
+      "grad_norm": 2.2962545997270376,
+      "learning_rate": 1e-06,
+      "loss": 0.1895,
+      "step": 5239
+    },
+    {
+      "epoch": 1.0020078401376804,
+      "grad_norm": 0.8093315481062006,
+      "learning_rate": 1e-06,
+      "loss": 0.0088,
+      "step": 5240
+    },
+    {
+      "epoch": 1.0021990630079358,
+      "grad_norm": 1.825581860213799,
+      "learning_rate": 1e-06,
+      "loss": 0.0552,
+      "step": 5241
+    },
+    {
+      "epoch": 1.002390285878191,
+      "grad_norm": 4.801405963028096,
+      "learning_rate": 1e-06,
+      "loss": 0.1697,
+      "step": 5242
+    },
+    {
+      "epoch": 1.0025815087484462,
+      "grad_norm": 4.2224911526033395,
+      "learning_rate": 1e-06,
+      "loss": 0.1799,
+      "step": 5243
+    },
+    {
+      "epoch": 1.0027727316187016,
+      "grad_norm": 4.712720846571539,
+      "learning_rate": 1e-06,
+      "loss": 0.3335,
+      "step": 5244
+    },
+    {
+      "epoch": 1.0029639544889568,
+      "grad_norm": 6.291937546446158,
+      "learning_rate": 1e-06,
+      "loss": 0.1789,
+      "step": 5245
+    },
+    {
+      "epoch": 1.0031551773592122,
+      "grad_norm": 3.328079724787406,
+      "learning_rate": 1e-06,
+      "loss": 0.1344,
+      "step": 5246
+    },
+    {
+      "epoch": 1.0033464002294674,
+      "grad_norm": 3.5597058766195206,
+      "learning_rate": 1e-06,
+      "loss": 0.2795,
+      "step": 5247
+    },
+    {
+      "epoch": 1.0035376230997228,
+      "grad_norm": 3.0544572436096646,
+      "learning_rate": 1e-06,
+      "loss": 0.1922,
+      "step": 5248
+    },
+    {
+      "epoch": 1.003728845969978,
+      "grad_norm": 1.380705051890246,
+      "learning_rate": 1e-06,
+      "loss": 0.0482,
+      "step": 5249
+    },
+    {
+      "epoch": 1.0039200688402332,
+      "grad_norm": 3.3328332684849418,
+      "learning_rate": 1e-06,
+      "loss": 0.1886,
+      "step": 5250
+    },
+    {
+      "epoch": 1.0041112917104886,
+      "grad_norm": 2.2392025527756925,
+      "learning_rate": 1e-06,
+      "loss": 0.0998,
+      "step": 5251
+    },
+    {
+      "epoch": 1.0043025145807438,
+      "grad_norm": 3.7605240965878766,
+      "learning_rate": 1e-06,
+      "loss": 0.1122,
+      "step": 5252
+    },
+    {
+      "epoch": 1.0044937374509992,
+      "grad_norm": 2.376595011732643,
+      "learning_rate": 1e-06,
+      "loss": 0.0457,
+      "step": 5253
+    },
+    {
+      "epoch": 1.0046849603212544,
+      "grad_norm": 1.9646478938325622,
+      "learning_rate": 1e-06,
+      "loss": 0.0485,
+      "step": 5254
+    },
+    {
+      "epoch": 1.0048761831915096,
+      "grad_norm": 3.5444170127320014,
+      "learning_rate": 1e-06,
+      "loss": 0.1307,
+      "step": 5255
+    },
+    {
+      "epoch": 1.005067406061765,
+      "grad_norm": 2.927099116996665,
+      "learning_rate": 1e-06,
+      "loss": 0.228,
+      "step": 5256
+    },
+    {
+      "epoch": 1.0052586289320202,
+      "grad_norm": 4.116872004485731,
+      "learning_rate": 1e-06,
+      "loss": 0.3688,
+      "step": 5257
+    },
+    {
+      "epoch": 1.0054498518022756,
+      "grad_norm": 3.269065249468974,
+      "learning_rate": 1e-06,
+      "loss": 0.2055,
+      "step": 5258
+    },
+    {
+      "epoch": 1.0056410746725308,
+      "grad_norm": 2.674474332979038,
+      "learning_rate": 1e-06,
+      "loss": 0.2206,
+      "step": 5259
+    },
+    {
+      "epoch": 1.005832297542786,
+      "grad_norm": 2.0007152470520078,
+      "learning_rate": 1e-06,
+      "loss": 0.1089,
+      "step": 5260
+    },
+    {
+      "epoch": 1.0060235204130414,
+      "grad_norm": 2.8261115562682413,
+      "learning_rate": 1e-06,
+      "loss": 0.1653,
+      "step": 5261
+    },
+    {
+      "epoch": 1.0062147432832966,
+      "grad_norm": 2.707398316789008,
+      "learning_rate": 1e-06,
+      "loss": 0.0366,
+      "step": 5262
+    },
+    {
+      "epoch": 1.006405966153552,
+      "grad_norm": 4.2775192155077555,
+      "learning_rate": 1e-06,
+      "loss": 0.1451,
+      "step": 5263
+    },
+    {
+      "epoch": 1.0065971890238072,
+      "grad_norm": 4.2879014519569285,
+      "learning_rate": 1e-06,
+      "loss": 0.0323,
+      "step": 5264
+    },
+    {
+      "epoch": 1.0067884118940624,
+      "grad_norm": 4.089079544095998,
+      "learning_rate": 1e-06,
+      "loss": 0.0803,
+      "step": 5265
+    },
+    {
+      "epoch": 1.0069796347643178,
+      "grad_norm": 2.716971155276765,
+      "learning_rate": 1e-06,
+      "loss": 0.066,
+      "step": 5266
+    },
+    {
+      "epoch": 1.007170857634573,
+      "grad_norm": 2.097382540045356,
+      "learning_rate": 1e-06,
+      "loss": 0.0202,
+      "step": 5267
+    },
+    {
+      "epoch": 1.0073620805048284,
+      "grad_norm": 5.031396330606896,
+      "learning_rate": 1e-06,
+      "loss": 0.3684,
+      "step": 5268
+    },
+    {
+      "epoch": 1.0075533033750836,
+      "grad_norm": 3.661871867298167,
+      "learning_rate": 1e-06,
+      "loss": 0.3494,
+      "step": 5269
+    },
+    {
+      "epoch": 1.0077445262453388,
+      "grad_norm": 5.127020088848402,
+      "learning_rate": 1e-06,
+      "loss": 0.2748,
+      "step": 5270
+    },
+    {
+      "epoch": 1.0079357491155942,
+      "grad_norm": 3.8014014621152388,
+      "learning_rate": 1e-06,
+      "loss": 0.2902,
+      "step": 5271
+    },
+    {
+      "epoch": 1.0081269719858494,
+      "grad_norm": 1.067402356973111,
+      "learning_rate": 1e-06,
+      "loss": 0.1151,
+      "step": 5272
+    },
+    {
+      "epoch": 1.0083181948561049,
+      "grad_norm": 1.2144955445761874,
+      "learning_rate": 1e-06,
+      "loss": 0.033,
+      "step": 5273
+    },
+    {
+      "epoch": 1.00850941772636,
+      "grad_norm": 5.370575969360808,
+      "learning_rate": 1e-06,
+      "loss": 0.1264,
+      "step": 5274
+    },
+    {
+      "epoch": 1.0087006405966155,
+      "grad_norm": 2.7929129214809216,
+      "learning_rate": 1e-06,
+      "loss": 0.2124,
+      "step": 5275
+    },
+    {
+      "epoch": 1.0088918634668707,
+      "grad_norm": 1.5482156511854812,
+      "learning_rate": 1e-06,
+      "loss": 0.0332,
+      "step": 5276
+    },
+    {
+      "epoch": 1.0090830863371258,
+      "grad_norm": 3.7664277420111243,
+      "learning_rate": 1e-06,
+      "loss": 0.1186,
+      "step": 5277
+    },
+    {
+      "epoch": 1.0092743092073813,
+      "grad_norm": 2.4151056005163642,
+      "learning_rate": 1e-06,
+      "loss": 0.176,
+      "step": 5278
+    },
+    {
+      "epoch": 1.0094655320776365,
+      "grad_norm": 3.282879234191258,
+      "learning_rate": 1e-06,
+      "loss": 0.081,
+      "step": 5279
+    },
+    {
+      "epoch": 1.0096567549478919,
+      "grad_norm": 7.170426498773724,
+      "learning_rate": 1e-06,
+      "loss": 0.1861,
+      "step": 5280
+    },
+    {
+      "epoch": 1.009847977818147,
+      "grad_norm": 4.383907296975674,
+      "learning_rate": 1e-06,
+      "loss": 0.2701,
+      "step": 5281
+    },
+    {
+      "epoch": 1.0100392006884022,
+      "grad_norm": 4.250303594180091,
+      "learning_rate": 1e-06,
+      "loss": 0.3736,
+      "step": 5282
+    },
+    {
+      "epoch": 1.0102304235586577,
+      "grad_norm": 4.927565225124023,
+      "learning_rate": 1e-06,
+      "loss": 0.5554,
+      "step": 5283
+    },
+    {
+      "epoch": 1.0104216464289129,
+      "grad_norm": 1.7330852214816825,
+      "learning_rate": 1e-06,
+      "loss": 0.1904,
+      "step": 5284
+    },
+    {
+      "epoch": 1.0106128692991683,
+      "grad_norm": 3.5888371597383255,
+      "learning_rate": 1e-06,
+      "loss": 0.1787,
+      "step": 5285
+    },
+    {
+      "epoch": 1.0108040921694235,
+      "grad_norm": 3.826925408945453,
+      "learning_rate": 1e-06,
+      "loss": 0.2863,
+      "step": 5286
+    },
+    {
+      "epoch": 1.0109953150396787,
+      "grad_norm": 2.7388243306744653,
+      "learning_rate": 1e-06,
+      "loss": 0.1904,
+      "step": 5287
+    },
+    {
+      "epoch": 1.011186537909934,
+      "grad_norm": 1.7727969857271164,
+      "learning_rate": 1e-06,
+      "loss": 0.0998,
+      "step": 5288
+    },
+    {
+      "epoch": 1.0113777607801893,
+      "grad_norm": 2.9792991733204555,
+      "learning_rate": 1e-06,
+      "loss": 0.0664,
+      "step": 5289
+    },
+    {
+      "epoch": 1.0115689836504447,
+      "grad_norm": 2.0106779202180887,
+      "learning_rate": 1e-06,
+      "loss": 0.0389,
+      "step": 5290
+    },
+    {
+      "epoch": 1.0117602065206999,
+      "grad_norm": 2.4289141341038816,
+      "learning_rate": 1e-06,
+      "loss": 0.0333,
+      "step": 5291
+    },
+    {
+      "epoch": 1.011951429390955,
+      "grad_norm": 2.5240400321312793,
+      "learning_rate": 1e-06,
+      "loss": 0.0573,
+      "step": 5292
+    },
+    {
+      "epoch": 1.0121426522612105,
+      "grad_norm": 6.763327827482707,
+      "learning_rate": 1e-06,
+      "loss": 0.2015,
+      "step": 5293
+    },
+    {
+      "epoch": 1.0123338751314657,
+      "grad_norm": 4.382060865389539,
+      "learning_rate": 1e-06,
+      "loss": 0.391,
+      "step": 5294
+    },
+    {
+      "epoch": 1.012525098001721,
+      "grad_norm": 3.9931155803794414,
+      "learning_rate": 1e-06,
+      "loss": 0.2973,
+      "step": 5295
+    },
+    {
+      "epoch": 1.0127163208719763,
+      "grad_norm": 3.0484186419241475,
+      "learning_rate": 1e-06,
+      "loss": 0.3766,
+      "step": 5296
+    },
+    {
+      "epoch": 1.0129075437422315,
+      "grad_norm": 4.827544439774031,
+      "learning_rate": 1e-06,
+      "loss": 0.3648,
+      "step": 5297
+    },
+    {
+      "epoch": 1.0130987666124869,
+      "grad_norm": 2.984765152119286,
+      "learning_rate": 1e-06,
+      "loss": 0.1899,
+      "step": 5298
+    },
+    {
+      "epoch": 1.013289989482742,
+      "grad_norm": 3.1467015329412082,
+      "learning_rate": 1e-06,
+      "loss": 0.1543,
+      "step": 5299
+    },
+    {
+      "epoch": 1.0134812123529975,
+      "grad_norm": 3.663882253154163,
+      "learning_rate": 1e-06,
+      "loss": 0.2461,
+      "step": 5300
+    },
+    {
+      "epoch": 1.0136724352232527,
+      "grad_norm": 4.301419826570531,
+      "learning_rate": 1e-06,
+      "loss": 0.0755,
+      "step": 5301
+    },
+    {
+      "epoch": 1.013863658093508,
+      "grad_norm": 2.82152584108305,
+      "learning_rate": 1e-06,
+      "loss": 0.1287,
+      "step": 5302
+    },
+    {
+      "epoch": 1.0140548809637633,
+      "grad_norm": 3.6783258141834154,
+      "learning_rate": 1e-06,
+      "loss": 0.0816,
+      "step": 5303
+    },
+    {
+      "epoch": 1.0142461038340185,
+      "grad_norm": 2.417753687647797,
+      "learning_rate": 1e-06,
+      "loss": 0.0326,
+      "step": 5304
+    },
+    {
+      "epoch": 1.014437326704274,
+      "grad_norm": 4.793724591164225,
+      "learning_rate": 1e-06,
+      "loss": 0.3066,
+      "step": 5305
+    },
+    {
+      "epoch": 1.014628549574529,
+      "grad_norm": 5.64724250277353,
+      "learning_rate": 1e-06,
+      "loss": 0.1661,
+      "step": 5306
+    },
+    {
+      "epoch": 1.0148197724447845,
+      "grad_norm": 2.874836295069635,
+      "learning_rate": 1e-06,
+      "loss": 0.2391,
+      "step": 5307
+    },
+    {
+      "epoch": 1.0150109953150397,
+      "grad_norm": 2.9306606456682016,
+      "learning_rate": 1e-06,
+      "loss": 0.1184,
+      "step": 5308
+    },
+    {
+      "epoch": 1.0152022181852949,
+      "grad_norm": 2.171544221347411,
+      "learning_rate": 1e-06,
+      "loss": 0.1547,
+      "step": 5309
+    },
+    {
+      "epoch": 1.0153934410555503,
+      "grad_norm": 4.434386329555757,
+      "learning_rate": 1e-06,
+      "loss": 0.1129,
+      "step": 5310
+    },
+    {
+      "epoch": 1.0155846639258055,
+      "grad_norm": 3.061328274531067,
+      "learning_rate": 1e-06,
+      "loss": 0.1215,
+      "step": 5311
+    },
+    {
+      "epoch": 1.015775886796061,
+      "grad_norm": 2.770274591159657,
+      "learning_rate": 1e-06,
+      "loss": 0.0764,
+      "step": 5312
+    },
+    {
+      "epoch": 1.015967109666316,
+      "grad_norm": 2.1376067553179325,
+      "learning_rate": 1e-06,
+      "loss": 0.0746,
+      "step": 5313
+    },
+    {
+      "epoch": 1.0161583325365713,
+      "grad_norm": 2.0283344178313105,
+      "learning_rate": 1e-06,
+      "loss": 0.0613,
+      "step": 5314
+    },
+    {
+      "epoch": 1.0163495554068267,
+      "grad_norm": 3.8723487397925878,
+      "learning_rate": 1e-06,
+      "loss": 0.0976,
+      "step": 5315
+    },
+    {
+      "epoch": 1.016540778277082,
+      "grad_norm": 2.1075218820397494,
+      "learning_rate": 1e-06,
+      "loss": 0.0504,
+      "step": 5316
+    },
+    {
+      "epoch": 1.0167320011473373,
+      "grad_norm": 7.712785771792982,
+      "learning_rate": 1e-06,
+      "loss": 0.0709,
+      "step": 5317
+    },
+    {
+      "epoch": 1.0169232240175925,
+      "grad_norm": 8.118086000560723,
+      "learning_rate": 1e-06,
+      "loss": 0.4497,
+      "step": 5318
+    },
+    {
+      "epoch": 1.0171144468878477,
+      "grad_norm": 4.027251397007302,
+      "learning_rate": 1e-06,
+      "loss": 0.4166,
+      "step": 5319
+    },
+    {
+      "epoch": 1.017305669758103,
+      "grad_norm": 2.3217510041388434,
+      "learning_rate": 1e-06,
+      "loss": 0.2393,
+      "step": 5320
+    },
+    {
+      "epoch": 1.0174968926283583,
+      "grad_norm": 2.304199580899659,
+      "learning_rate": 1e-06,
+      "loss": 0.1026,
+      "step": 5321
+    },
+    {
+      "epoch": 1.0176881154986137,
+      "grad_norm": 5.193689721006716,
+      "learning_rate": 1e-06,
+      "loss": 0.3945,
+      "step": 5322
+    },
+    {
+      "epoch": 1.017879338368869,
+      "grad_norm": 3.584532086256836,
+      "learning_rate": 1e-06,
+      "loss": 0.223,
+      "step": 5323
+    },
+    {
+      "epoch": 1.018070561239124,
+      "grad_norm": 4.6015457127107835,
+      "learning_rate": 1e-06,
+      "loss": 0.0867,
+      "step": 5324
+    },
+    {
+      "epoch": 1.0182617841093795,
+      "grad_norm": 6.126785329528128,
+      "learning_rate": 1e-06,
+      "loss": 0.2948,
+      "step": 5325
+    },
+    {
+      "epoch": 1.0184530069796347,
+      "grad_norm": 4.137128397224586,
+      "learning_rate": 1e-06,
+      "loss": 0.083,
+      "step": 5326
+    },
+    {
+      "epoch": 1.0186442298498901,
+      "grad_norm": 3.702463855554105,
+      "learning_rate": 1e-06,
+      "loss": 0.0679,
+      "step": 5327
+    },
+    {
+      "epoch": 1.0188354527201453,
+      "grad_norm": 3.697764319205347,
+      "learning_rate": 1e-06,
+      "loss": 0.1607,
+      "step": 5328
+    },
+    {
+      "epoch": 1.0190266755904007,
+      "grad_norm": 1.679092798745959,
+      "learning_rate": 1e-06,
+      "loss": 0.0245,
+      "step": 5329
+    },
+    {
+      "epoch": 1.019217898460656,
+      "grad_norm": 5.575757975006592,
+      "learning_rate": 1e-06,
+      "loss": 0.1111,
+      "step": 5330
+    },
+    {
+      "epoch": 1.019409121330911,
+      "grad_norm": 4.308788969540421,
+      "learning_rate": 1e-06,
+      "loss": 0.3898,
+      "step": 5331
+    },
+    {
+      "epoch": 1.0196003442011665,
+      "grad_norm": 3.2403765297139913,
+      "learning_rate": 1e-06,
+      "loss": 0.1422,
+      "step": 5332
+    },
+    {
+      "epoch": 1.0197915670714217,
+      "grad_norm": 4.204640849013082,
+      "learning_rate": 1e-06,
+      "loss": 0.2461,
+      "step": 5333
+    },
+    {
+      "epoch": 1.0199827899416771,
+      "grad_norm": 3.859567367145655,
+      "learning_rate": 1e-06,
+      "loss": 0.0959,
+      "step": 5334
+    },
+    {
+      "epoch": 1.0201740128119323,
+      "grad_norm": 3.868694712994117,
+      "learning_rate": 1e-06,
+      "loss": 0.2374,
+      "step": 5335
+    },
+    {
+      "epoch": 1.0203652356821875,
+      "grad_norm": 4.53555156814954,
+      "learning_rate": 1e-06,
+      "loss": 0.2376,
+      "step": 5336
+    },
+    {
+      "epoch": 1.020556458552443,
+      "grad_norm": 1.5382666684491733,
+      "learning_rate": 1e-06,
+      "loss": 0.071,
+      "step": 5337
+    },
+    {
+      "epoch": 1.0207476814226981,
+      "grad_norm": 2.333916852285567,
+      "learning_rate": 1e-06,
+      "loss": 0.0976,
+      "step": 5338
+    },
+    {
+      "epoch": 1.0209389042929535,
+      "grad_norm": 2.7103897112933706,
+      "learning_rate": 1e-06,
+      "loss": 0.0609,
+      "step": 5339
+    },
+    {
+      "epoch": 1.0211301271632087,
+      "grad_norm": 2.5519407479035534,
+      "learning_rate": 1e-06,
+      "loss": 0.046,
+      "step": 5340
+    },
+    {
+      "epoch": 1.021321350033464,
+      "grad_norm": 2.6328213745329845,
+      "learning_rate": 1e-06,
+      "loss": 0.0639,
+      "step": 5341
+    },
+    {
+      "epoch": 1.0215125729037193,
+      "grad_norm": 6.97407335258221,
+      "learning_rate": 1e-06,
+      "loss": 0.0957,
+      "step": 5342
+    },
+    {
+      "epoch": 1.0217037957739745,
+      "grad_norm": 10.841442379304034,
+      "learning_rate": 1e-06,
+      "loss": 0.3333,
+      "step": 5343
+    },
+    {
+      "epoch": 1.02189501864423,
+      "grad_norm": 4.045446195194406,
+      "learning_rate": 1e-06,
+      "loss": 0.3196,
+      "step": 5344
+    },
+    {
+      "epoch": 1.0220862415144851,
+      "grad_norm": 2.6425718094552213,
+      "learning_rate": 1e-06,
+      "loss": 0.1747,
+      "step": 5345
+    },
+    {
+      "epoch": 1.0222774643847403,
+      "grad_norm": 3.9896484183770253,
+      "learning_rate": 1e-06,
+      "loss": 0.1293,
+      "step": 5346
+    },
+    {
+      "epoch": 1.0224686872549957,
+      "grad_norm": 2.5087250092178297,
+      "learning_rate": 1e-06,
+      "loss": 0.0512,
+      "step": 5347
+    },
+    {
+      "epoch": 1.022659910125251,
+      "grad_norm": 4.657384337434173,
+      "learning_rate": 1e-06,
+      "loss": 0.2727,
+      "step": 5348
+    },
+    {
+      "epoch": 1.0228511329955063,
+      "grad_norm": 3.267364768606029,
+      "learning_rate": 1e-06,
+      "loss": 0.0975,
+      "step": 5349
+    },
+    {
+      "epoch": 1.0230423558657615,
+      "grad_norm": 2.3507841222408925,
+      "learning_rate": 1e-06,
+      "loss": 0.086,
+      "step": 5350
+    },
+    {
+      "epoch": 1.0232335787360167,
+      "grad_norm": 1.4609508003812597,
+      "learning_rate": 1e-06,
+      "loss": 0.058,
+      "step": 5351
+    },
+    {
+      "epoch": 1.0234248016062721,
+      "grad_norm": 3.4176477248795716,
+      "learning_rate": 1e-06,
+      "loss": 0.2549,
+      "step": 5352
+    },
+    {
+      "epoch": 1.0236160244765273,
+      "grad_norm": 2.441403222654373,
+      "learning_rate": 1e-06,
+      "loss": 0.0827,
+      "step": 5353
+    },
+    {
+      "epoch": 1.0238072473467827,
+      "grad_norm": 1.8112111606474766,
+      "learning_rate": 1e-06,
+      "loss": 0.0244,
+      "step": 5354
+    },
+    {
+      "epoch": 1.023998470217038,
+      "grad_norm": 5.607549088595289,
+      "learning_rate": 1e-06,
+      "loss": 0.1293,
+      "step": 5355
+    },
+    {
+      "epoch": 1.0241896930872931,
+      "grad_norm": 4.1105616491585675,
+      "learning_rate": 1e-06,
+      "loss": 0.271,
+      "step": 5356
+    },
+    {
+      "epoch": 1.0243809159575485,
+      "grad_norm": 8.886386927184125,
+      "learning_rate": 1e-06,
+      "loss": 0.3615,
+      "step": 5357
+    },
+    {
+      "epoch": 1.0245721388278037,
+      "grad_norm": 4.831038796371933,
+      "learning_rate": 1e-06,
+      "loss": 0.3904,
+      "step": 5358
+    },
+    {
+      "epoch": 1.0247633616980591,
+      "grad_norm": 4.846463138831728,
+      "learning_rate": 1e-06,
+      "loss": 0.4093,
+      "step": 5359
+    },
+    {
+      "epoch": 1.0249545845683143,
+      "grad_norm": 1.9857038719162934,
+      "learning_rate": 1e-06,
+      "loss": 0.0633,
+      "step": 5360
+    },
+    {
+      "epoch": 1.0251458074385698,
+      "grad_norm": 2.3871822125710525,
+      "learning_rate": 1e-06,
+      "loss": 0.0867,
+      "step": 5361
+    },
+    {
+      "epoch": 1.025337030308825,
+      "grad_norm": 3.381465934585214,
+      "learning_rate": 1e-06,
+      "loss": 0.0934,
+      "step": 5362
+    },
+    {
+      "epoch": 1.0255282531790801,
+      "grad_norm": 2.2780303181882187,
+      "learning_rate": 1e-06,
+      "loss": 0.0433,
+      "step": 5363
+    },
+    {
+      "epoch": 1.0257194760493356,
+      "grad_norm": 4.03448333003824,
+      "learning_rate": 1e-06,
+      "loss": 0.0426,
+      "step": 5364
+    },
+    {
+      "epoch": 1.0259106989195907,
+      "grad_norm": 2.8039238615378395,
+      "learning_rate": 1e-06,
+      "loss": 0.0694,
+      "step": 5365
+    },
+    {
+      "epoch": 1.0261019217898462,
+      "grad_norm": 3.5102318659435516,
+      "learning_rate": 1e-06,
+      "loss": 0.0369,
+      "step": 5366
+    },
+    {
+      "epoch": 1.0262931446601014,
+      "grad_norm": 6.203577289379309,
+      "learning_rate": 1e-06,
+      "loss": 0.0985,
+      "step": 5367
+    },
+    {
+      "epoch": 1.0264843675303565,
+      "grad_norm": 9.43179492425297,
+      "learning_rate": 1e-06,
+      "loss": 0.3127,
+      "step": 5368
+    },
+    {
+      "epoch": 1.026675590400612,
+      "grad_norm": 2.705088260783132,
+      "learning_rate": 1e-06,
+      "loss": 0.3217,
+      "step": 5369
+    },
+    {
+      "epoch": 1.0268668132708672,
+      "grad_norm": 3.8712642258754397,
+      "learning_rate": 1e-06,
+      "loss": 0.2439,
+      "step": 5370
+    },
+    {
+      "epoch": 1.0270580361411226,
+      "grad_norm": 1.838354848676419,
+      "learning_rate": 1e-06,
+      "loss": 0.1397,
+      "step": 5371
+    },
+    {
+      "epoch": 1.0272492590113778,
+      "grad_norm": 2.981151817512422,
+      "learning_rate": 1e-06,
+      "loss": 0.1864,
+      "step": 5372
+    },
+    {
+      "epoch": 1.027440481881633,
+      "grad_norm": 3.037159929744032,
+      "learning_rate": 1e-06,
+      "loss": 0.0832,
+      "step": 5373
+    },
+    {
+      "epoch": 1.0276317047518884,
+      "grad_norm": 3.655539940900863,
+      "learning_rate": 1e-06,
+      "loss": 0.1339,
+      "step": 5374
+    },
+    {
+      "epoch": 1.0278229276221436,
+      "grad_norm": 2.444267171394623,
+      "learning_rate": 1e-06,
+      "loss": 0.0724,
+      "step": 5375
+    },
+    {
+      "epoch": 1.028014150492399,
+      "grad_norm": 1.9784532633195484,
+      "learning_rate": 1e-06,
+      "loss": 0.0365,
+      "step": 5376
+    },
+    {
+      "epoch": 1.0282053733626542,
+      "grad_norm": 2.207359411098062,
+      "learning_rate": 1e-06,
+      "loss": 0.1362,
+      "step": 5377
+    },
+    {
+      "epoch": 1.0283965962329094,
+      "grad_norm": 3.043025475341786,
+      "learning_rate": 1e-06,
+      "loss": 0.125,
+      "step": 5378
+    },
+    {
+      "epoch": 1.0285878191031648,
+      "grad_norm": 2.974489947280402,
+      "learning_rate": 1e-06,
+      "loss": 0.0388,
+      "step": 5379
+    },
+    {
+      "epoch": 1.02877904197342,
+      "grad_norm": 6.246302922642802,
+      "learning_rate": 1e-06,
+      "loss": 0.2256,
+      "step": 5380
+    },
+    {
+      "epoch": 1.0289702648436754,
+      "grad_norm": 7.451454215243827,
+      "learning_rate": 1e-06,
+      "loss": 0.3423,
+      "step": 5381
+    },
+    {
+      "epoch": 1.0291614877139306,
+      "grad_norm": 4.227513458392954,
+      "learning_rate": 1e-06,
+      "loss": 0.2645,
+      "step": 5382
+    },
+    {
+      "epoch": 1.0293527105841858,
+      "grad_norm": 3.092736588774405,
+      "learning_rate": 1e-06,
+      "loss": 0.1687,
+      "step": 5383
+    },
+    {
+      "epoch": 1.0295439334544412,
+      "grad_norm": 2.469171222532792,
+      "learning_rate": 1e-06,
+      "loss": 0.0854,
+      "step": 5384
+    },
+    {
+      "epoch": 1.0297351563246964,
+      "grad_norm": 2.4333674293463394,
+      "learning_rate": 1e-06,
+      "loss": 0.0886,
+      "step": 5385
+    },
+    {
+      "epoch": 1.0299263791949518,
+      "grad_norm": 2.596769574500783,
+      "learning_rate": 1e-06,
+      "loss": 0.1047,
+      "step": 5386
+    },
+    {
+      "epoch": 1.030117602065207,
+      "grad_norm": 2.143845412018965,
+      "learning_rate": 1e-06,
+      "loss": 0.0496,
+      "step": 5387
+    },
+    {
+      "epoch": 1.0303088249354624,
+      "grad_norm": 1.806758468496181,
+      "learning_rate": 1e-06,
+      "loss": 0.1042,
+      "step": 5388
+    },
+    {
+      "epoch": 1.0305000478057176,
+      "grad_norm": 2.346638743042981,
+      "learning_rate": 1e-06,
+      "loss": 0.0389,
+      "step": 5389
+    },
+    {
+      "epoch": 1.0306912706759728,
+      "grad_norm": 4.154115515993127,
+      "learning_rate": 1e-06,
+      "loss": 0.2609,
+      "step": 5390
+    },
+    {
+      "epoch": 1.0308824935462282,
+      "grad_norm": 2.2245624326286277,
+      "learning_rate": 1e-06,
+      "loss": 0.0202,
+      "step": 5391
+    },
+    {
+      "epoch": 1.0310737164164834,
+      "grad_norm": 3.0368107394375246,
+      "learning_rate": 1e-06,
+      "loss": 0.0479,
+      "step": 5392
+    },
+    {
+      "epoch": 1.0312649392867388,
+      "grad_norm": 5.820361000217389,
+      "learning_rate": 1e-06,
+      "loss": 0.2622,
+      "step": 5393
+    },
+    {
+      "epoch": 1.031456162156994,
+      "grad_norm": 3.3346486516489033,
+      "learning_rate": 1e-06,
+      "loss": 0.1348,
+      "step": 5394
+    },
+    {
+      "epoch": 1.0316473850272492,
+      "grad_norm": 4.107827014687056,
+      "learning_rate": 1e-06,
+      "loss": 0.1811,
+      "step": 5395
+    },
+    {
+      "epoch": 1.0318386078975046,
+      "grad_norm": 6.075015020155686,
+      "learning_rate": 1e-06,
+      "loss": 0.2541,
+      "step": 5396
+    },
+    {
+      "epoch": 1.0320298307677598,
+      "grad_norm": 4.261015361078699,
+      "learning_rate": 1e-06,
+      "loss": 0.1688,
+      "step": 5397
+    },
+    {
+      "epoch": 1.0322210536380152,
+      "grad_norm": 3.802101713510617,
+      "learning_rate": 1e-06,
+      "loss": 0.3383,
+      "step": 5398
+    },
+    {
+      "epoch": 1.0324122765082704,
+      "grad_norm": 2.6704243966324683,
+      "learning_rate": 1e-06,
+      "loss": 0.1005,
+      "step": 5399
+    },
+    {
+      "epoch": 1.0326034993785256,
+      "grad_norm": 2.9860755110005943,
+      "learning_rate": 1e-06,
+      "loss": 0.108,
+      "step": 5400
+    },
+    {
+      "epoch": 1.032794722248781,
+      "grad_norm": 2.5838385877477292,
+      "learning_rate": 1e-06,
+      "loss": 0.1961,
+      "step": 5401
+    },
+    {
+      "epoch": 1.0329859451190362,
+      "grad_norm": 2.712284336218385,
+      "learning_rate": 1e-06,
+      "loss": 0.0605,
+      "step": 5402
+    },
+    {
+      "epoch": 1.0331771679892916,
+      "grad_norm": 3.9348915785087573,
+      "learning_rate": 1e-06,
+      "loss": 0.109,
+      "step": 5403
+    },
+    {
+      "epoch": 1.0333683908595468,
+      "grad_norm": 3.0727816449690484,
+      "learning_rate": 1e-06,
+      "loss": 0.0654,
+      "step": 5404
+    },
+    {
+      "epoch": 1.033559613729802,
+      "grad_norm": 12.528519711136816,
+      "learning_rate": 1e-06,
+      "loss": 0.3577,
+      "step": 5405
+    },
+    {
+      "epoch": 1.0337508366000574,
+      "grad_norm": 3.500469993642511,
+      "learning_rate": 1e-06,
+      "loss": 0.458,
+      "step": 5406
+    },
+    {
+      "epoch": 1.0339420594703126,
+      "grad_norm": 5.748986859834591,
+      "learning_rate": 1e-06,
+      "loss": 0.5398,
+      "step": 5407
+    },
+    {
+      "epoch": 1.034133282340568,
+      "grad_norm": 4.026313538972261,
+      "learning_rate": 1e-06,
+      "loss": 0.3088,
+      "step": 5408
+    },
+    {
+      "epoch": 1.0343245052108232,
+      "grad_norm": 2.400072068880752,
+      "learning_rate": 1e-06,
+      "loss": 0.079,
+      "step": 5409
+    },
+    {
+      "epoch": 1.0345157280810784,
+      "grad_norm": 4.613296960970284,
+      "learning_rate": 1e-06,
+      "loss": 0.6028,
+      "step": 5410
+    },
+    {
+      "epoch": 1.0347069509513338,
+      "grad_norm": 3.0580666040286455,
+      "learning_rate": 1e-06,
+      "loss": 0.101,
+      "step": 5411
+    },
+    {
+      "epoch": 1.034898173821589,
+      "grad_norm": 1.737522053750129,
+      "learning_rate": 1e-06,
+      "loss": 0.0292,
+      "step": 5412
+    },
+    {
+      "epoch": 1.0350893966918444,
+      "grad_norm": 3.2584977195489433,
+      "learning_rate": 1e-06,
+      "loss": 0.1469,
+      "step": 5413
+    },
+    {
+      "epoch": 1.0352806195620996,
+      "grad_norm": 2.5874816341024776,
+      "learning_rate": 1e-06,
+      "loss": 0.0515,
+      "step": 5414
+    },
+    {
+      "epoch": 1.035471842432355,
+      "grad_norm": 2.7987133782577516,
+      "learning_rate": 1e-06,
+      "loss": 0.0772,
+      "step": 5415
+    },
+    {
+      "epoch": 1.0356630653026102,
+      "grad_norm": 3.0355291935247815,
+      "learning_rate": 1e-06,
+      "loss": 0.0308,
+      "step": 5416
+    },
+    {
+      "epoch": 1.0358542881728654,
+      "grad_norm": 2.0371244728062883,
+      "learning_rate": 1e-06,
+      "loss": 0.0353,
+      "step": 5417
+    },
+    {
+      "epoch": 1.0360455110431208,
+      "grad_norm": 7.599857589993618,
+      "learning_rate": 1e-06,
+      "loss": 0.2347,
+      "step": 5418
+    },
+    {
+      "epoch": 1.036236733913376,
+      "grad_norm": 3.8802084835317014,
+      "learning_rate": 1e-06,
+      "loss": 0.274,
+      "step": 5419
+    },
+    {
+      "epoch": 1.0364279567836314,
+      "grad_norm": 3.0643999084169895,
+      "learning_rate": 1e-06,
+      "loss": 0.1625,
+      "step": 5420
+    },
+    {
+      "epoch": 1.0366191796538866,
+      "grad_norm": 2.9156733501714323,
+      "learning_rate": 1e-06,
+      "loss": 0.2735,
+      "step": 5421
+    },
+    {
+      "epoch": 1.0368104025241418,
+      "grad_norm": 1.767332753385096,
+      "learning_rate": 1e-06,
+      "loss": 0.0618,
+      "step": 5422
+    },
+    {
+      "epoch": 1.0370016253943972,
+      "grad_norm": 2.44368428876097,
+      "learning_rate": 1e-06,
+      "loss": 0.2296,
+      "step": 5423
+    },
+    {
+      "epoch": 1.0371928482646524,
+      "grad_norm": 2.550708997558967,
+      "learning_rate": 1e-06,
+      "loss": 0.0776,
+      "step": 5424
+    },
+    {
+      "epoch": 1.0373840711349078,
+      "grad_norm": 3.91124741366122,
+      "learning_rate": 1e-06,
+      "loss": 0.1642,
+      "step": 5425
+    },
+    {
+      "epoch": 1.037575294005163,
+      "grad_norm": 2.708417969383807,
+      "learning_rate": 1e-06,
+      "loss": 0.074,
+      "step": 5426
+    },
+    {
+      "epoch": 1.0377665168754182,
+      "grad_norm": 3.2321312534572595,
+      "learning_rate": 1e-06,
+      "loss": 0.0511,
+      "step": 5427
+    },
+    {
+      "epoch": 1.0379577397456736,
+      "grad_norm": 3.6890108278590783,
+      "learning_rate": 1e-06,
+      "loss": 0.1613,
+      "step": 5428
+    },
+    {
+      "epoch": 1.0381489626159288,
+      "grad_norm": 3.8883280273959415,
+      "learning_rate": 1e-06,
+      "loss": 0.0303,
+      "step": 5429
+    },
+    {
+      "epoch": 1.0383401854861842,
+      "grad_norm": 6.0163282107245575,
+      "learning_rate": 1e-06,
+      "loss": 0.3175,
+      "step": 5430
+    },
+    {
+      "epoch": 1.0385314083564394,
+      "grad_norm": 4.730509473529294,
+      "learning_rate": 1e-06,
+      "loss": 0.3146,
+      "step": 5431
+    },
+    {
+      "epoch": 1.0387226312266946,
+      "grad_norm": 4.367947807391442,
+      "learning_rate": 1e-06,
+      "loss": 0.5422,
+      "step": 5432
+    },
+    {
+      "epoch": 1.03891385409695,
+      "grad_norm": 5.173836235365481,
+      "learning_rate": 1e-06,
+      "loss": 0.5044,
+      "step": 5433
+    },
+    {
+      "epoch": 1.0391050769672052,
+      "grad_norm": 1.6647550905662256,
+      "learning_rate": 1e-06,
+      "loss": 0.1233,
+      "step": 5434
+    },
+    {
+      "epoch": 1.0392962998374606,
+      "grad_norm": 2.795555698653869,
+      "learning_rate": 1e-06,
+      "loss": 0.1308,
+      "step": 5435
+    },
+    {
+      "epoch": 1.0394875227077158,
+      "grad_norm": 3.408829858206077,
+      "learning_rate": 1e-06,
+      "loss": 0.1415,
+      "step": 5436
+    },
+    {
+      "epoch": 1.039678745577971,
+      "grad_norm": 3.3150371695573604,
+      "learning_rate": 1e-06,
+      "loss": 0.2453,
+      "step": 5437
+    },
+    {
+      "epoch": 1.0398699684482264,
+      "grad_norm": 3.0215293520913074,
+      "learning_rate": 1e-06,
+      "loss": 0.2041,
+      "step": 5438
+    },
+    {
+      "epoch": 1.0400611913184816,
+      "grad_norm": 2.235162982843866,
+      "learning_rate": 1e-06,
+      "loss": 0.1327,
+      "step": 5439
+    },
+    {
+      "epoch": 1.040252414188737,
+      "grad_norm": 6.1294859982134655,
+      "learning_rate": 1e-06,
+      "loss": 0.0279,
+      "step": 5440
+    },
+    {
+      "epoch": 1.0404436370589922,
+      "grad_norm": 5.11227213531303,
+      "learning_rate": 1e-06,
+      "loss": 0.3391,
+      "step": 5441
+    },
+    {
+      "epoch": 1.0406348599292476,
+      "grad_norm": 4.407111766764698,
+      "learning_rate": 1e-06,
+      "loss": 0.0897,
+      "step": 5442
+    },
+    {
+      "epoch": 1.0408260827995028,
+      "grad_norm": 6.294420999406811,
+      "learning_rate": 1e-06,
+      "loss": 0.4102,
+      "step": 5443
+    },
+    {
+      "epoch": 1.041017305669758,
+      "grad_norm": 4.302139885309802,
+      "learning_rate": 1e-06,
+      "loss": 0.4665,
+      "step": 5444
+    },
+    {
+      "epoch": 1.0412085285400134,
+      "grad_norm": 3.2967349922137688,
+      "learning_rate": 1e-06,
+      "loss": 0.2434,
+      "step": 5445
+    },
+    {
+      "epoch": 1.0413997514102686,
+      "grad_norm": 1.9965399973030062,
+      "learning_rate": 1e-06,
+      "loss": 0.0934,
+      "step": 5446
+    },
+    {
+      "epoch": 1.041590974280524,
+      "grad_norm": 3.1330342582926307,
+      "learning_rate": 1e-06,
+      "loss": 0.2961,
+      "step": 5447
+    },
+    {
+      "epoch": 1.0417821971507792,
+      "grad_norm": 4.620374892906525,
+      "learning_rate": 1e-06,
+      "loss": 0.3154,
+      "step": 5448
+    },
+    {
+      "epoch": 1.0419734200210344,
+      "grad_norm": 3.4754437993378566,
+      "learning_rate": 1e-06,
+      "loss": 0.1742,
+      "step": 5449
+    },
+    {
+      "epoch": 1.0421646428912898,
+      "grad_norm": 3.2303590767451027,
+      "learning_rate": 1e-06,
+      "loss": 0.1539,
+      "step": 5450
+    },
+    {
+      "epoch": 1.042355865761545,
+      "grad_norm": 3.887813670760188,
+      "learning_rate": 1e-06,
+      "loss": 0.1999,
+      "step": 5451
+    },
+    {
+      "epoch": 1.0425470886318005,
+      "grad_norm": 4.790014872358643,
+      "learning_rate": 1e-06,
+      "loss": 0.1326,
+      "step": 5452
+    },
+    {
+      "epoch": 1.0427383115020556,
+      "grad_norm": 2.332794217998105,
+      "learning_rate": 1e-06,
+      "loss": 0.1261,
+      "step": 5453
+    },
+    {
+      "epoch": 1.0429295343723108,
+      "grad_norm": 3.8869807936162233,
+      "learning_rate": 1e-06,
+      "loss": 0.0889,
+      "step": 5454
+    },
+    {
+      "epoch": 1.0431207572425663,
+      "grad_norm": 5.994285723439325,
+      "learning_rate": 1e-06,
+      "loss": 0.1574,
+      "step": 5455
+    },
+    {
+      "epoch": 1.0433119801128214,
+      "grad_norm": 5.390616154317923,
+      "learning_rate": 1e-06,
+      "loss": 0.2774,
+      "step": 5456
+    },
+    {
+      "epoch": 1.0435032029830769,
+      "grad_norm": 3.40680934968019,
+      "learning_rate": 1e-06,
+      "loss": 0.2888,
+      "step": 5457
+    },
+    {
+      "epoch": 1.043694425853332,
+      "grad_norm": 4.15135282964025,
+      "learning_rate": 1e-06,
+      "loss": 0.0725,
+      "step": 5458
+    },
+    {
+      "epoch": 1.0438856487235872,
+      "grad_norm": 2.5150540100729417,
+      "learning_rate": 1e-06,
+      "loss": 0.1099,
+      "step": 5459
+    },
+    {
+      "epoch": 1.0440768715938427,
+      "grad_norm": 2.1395589582570875,
+      "learning_rate": 1e-06,
+      "loss": 0.0475,
+      "step": 5460
+    },
+    {
+      "epoch": 1.0442680944640979,
+      "grad_norm": 4.742057383730275,
+      "learning_rate": 1e-06,
+      "loss": 0.2765,
+      "step": 5461
+    },
+    {
+      "epoch": 1.0444593173343533,
+      "grad_norm": 2.884384595001562,
+      "learning_rate": 1e-06,
+      "loss": 0.0463,
+      "step": 5462
+    },
+    {
+      "epoch": 1.0446505402046085,
+      "grad_norm": 2.566284548401018,
+      "learning_rate": 1e-06,
+      "loss": 0.1473,
+      "step": 5463
+    },
+    {
+      "epoch": 1.0448417630748636,
+      "grad_norm": 3.3048635305141216,
+      "learning_rate": 1e-06,
+      "loss": 0.0944,
+      "step": 5464
+    },
+    {
+      "epoch": 1.045032985945119,
+      "grad_norm": 7.086583756792903,
+      "learning_rate": 1e-06,
+      "loss": 0.1153,
+      "step": 5465
+    },
+    {
+      "epoch": 1.0452242088153743,
+      "grad_norm": 2.0524993965250258,
+      "learning_rate": 1e-06,
+      "loss": 0.0446,
+      "step": 5466
+    },
+    {
+      "epoch": 1.0454154316856297,
+      "grad_norm": 4.32149551501886,
+      "learning_rate": 1e-06,
+      "loss": 0.0926,
+      "step": 5467
+    },
+    {
+      "epoch": 1.0456066545558849,
+      "grad_norm": 8.220694076304714,
+      "learning_rate": 1e-06,
+      "loss": 0.4189,
+      "step": 5468
+    },
+    {
+      "epoch": 1.04579787742614,
+      "grad_norm": 5.23002402684836,
+      "learning_rate": 1e-06,
+      "loss": 0.177,
+      "step": 5469
+    },
+    {
+      "epoch": 1.0459891002963955,
+      "grad_norm": 3.555346082106467,
+      "learning_rate": 1e-06,
+      "loss": 0.2192,
+      "step": 5470
+    },
+    {
+      "epoch": 1.0461803231666507,
+      "grad_norm": 3.875425992130852,
+      "learning_rate": 1e-06,
+      "loss": 0.2219,
+      "step": 5471
+    },
+    {
+      "epoch": 1.046371546036906,
+      "grad_norm": 2.4236807705384855,
+      "learning_rate": 1e-06,
+      "loss": 0.1366,
+      "step": 5472
+    },
+    {
+      "epoch": 1.0465627689071613,
+      "grad_norm": 3.1023349520717765,
+      "learning_rate": 1e-06,
+      "loss": 0.122,
+      "step": 5473
+    },
+    {
+      "epoch": 1.0467539917774167,
+      "grad_norm": 3.6533974499424633,
+      "learning_rate": 1e-06,
+      "loss": 0.2821,
+      "step": 5474
+    },
+    {
+      "epoch": 1.0469452146476719,
+      "grad_norm": 2.1107142965027696,
+      "learning_rate": 1e-06,
+      "loss": 0.1063,
+      "step": 5475
+    },
+    {
+      "epoch": 1.047136437517927,
+      "grad_norm": 2.2946523694810805,
+      "learning_rate": 1e-06,
+      "loss": 0.1082,
+      "step": 5476
+    },
+    {
+      "epoch": 1.0473276603881825,
+      "grad_norm": 3.6706192791317904,
+      "learning_rate": 1e-06,
+      "loss": 0.1436,
+      "step": 5477
+    },
+    {
+      "epoch": 1.0475188832584377,
+      "grad_norm": 2.417545410960546,
+      "learning_rate": 1e-06,
+      "loss": 0.0767,
+      "step": 5478
+    },
+    {
+      "epoch": 1.047710106128693,
+      "grad_norm": 3.1327886033514254,
+      "learning_rate": 1e-06,
+      "loss": 0.0619,
+      "step": 5479
+    },
+    {
+      "epoch": 1.0479013289989483,
+      "grad_norm": 5.172755048390501,
+      "learning_rate": 1e-06,
+      "loss": 0.1217,
+      "step": 5480
+    },
+    {
+      "epoch": 1.0480925518692035,
+      "grad_norm": 3.3966385212469334,
+      "learning_rate": 1e-06,
+      "loss": 0.2466,
+      "step": 5481
+    },
+    {
+      "epoch": 1.0482837747394589,
+      "grad_norm": 3.8962622557696065,
+      "learning_rate": 1e-06,
+      "loss": 0.1498,
+      "step": 5482
+    },
+    {
+      "epoch": 1.048474997609714,
+      "grad_norm": 3.1700393551570665,
+      "learning_rate": 1e-06,
+      "loss": 0.3427,
+      "step": 5483
+    },
+    {
+      "epoch": 1.0486662204799695,
+      "grad_norm": 3.975218180300756,
+      "learning_rate": 1e-06,
+      "loss": 0.2104,
+      "step": 5484
+    },
+    {
+      "epoch": 1.0488574433502247,
+      "grad_norm": 3.8808984471328296,
+      "learning_rate": 1e-06,
+      "loss": 0.1821,
+      "step": 5485
+    },
+    {
+      "epoch": 1.0490486662204799,
+      "grad_norm": 1.9379626306290954,
+      "learning_rate": 1e-06,
+      "loss": 0.1375,
+      "step": 5486
+    },
+    {
+      "epoch": 1.0492398890907353,
+      "grad_norm": 4.9733951376244265,
+      "learning_rate": 1e-06,
+      "loss": 0.5489,
+      "step": 5487
+    },
+    {
+      "epoch": 1.0494311119609905,
+      "grad_norm": 2.732222227338133,
+      "learning_rate": 1e-06,
+      "loss": 0.0977,
+      "step": 5488
+    },
+    {
+      "epoch": 1.049622334831246,
+      "grad_norm": 3.1198460611856484,
+      "learning_rate": 1e-06,
+      "loss": 0.0487,
+      "step": 5489
+    },
+    {
+      "epoch": 1.049813557701501,
+      "grad_norm": 2.653571989150704,
+      "learning_rate": 1e-06,
+      "loss": 0.0377,
+      "step": 5490
+    },
+    {
+      "epoch": 1.0500047805717563,
+      "grad_norm": 1.8952829030014668,
+      "learning_rate": 1e-06,
+      "loss": 0.0451,
+      "step": 5491
+    },
+    {
+      "epoch": 1.0501960034420117,
+      "grad_norm": 3.6622545544139453,
+      "learning_rate": 1e-06,
+      "loss": 0.1108,
+      "step": 5492
+    },
+    {
+      "epoch": 1.0503872263122669,
+      "grad_norm": 4.919284490024828,
+      "learning_rate": 1e-06,
+      "loss": 0.2978,
+      "step": 5493
+    },
+    {
+      "epoch": 1.0505784491825223,
+      "grad_norm": 3.620821945705146,
+      "learning_rate": 1e-06,
+      "loss": 0.2949,
+      "step": 5494
+    },
+    {
+      "epoch": 1.0507696720527775,
+      "grad_norm": 4.689316461201037,
+      "learning_rate": 1e-06,
+      "loss": 0.22,
+      "step": 5495
+    },
+    {
+      "epoch": 1.050960894923033,
+      "grad_norm": 4.017052304023043,
+      "learning_rate": 1e-06,
+      "loss": 0.1856,
+      "step": 5496
+    },
+    {
+      "epoch": 1.051152117793288,
+      "grad_norm": 2.4707268634605652,
+      "learning_rate": 1e-06,
+      "loss": 0.1098,
+      "step": 5497
+    },
+    {
+      "epoch": 1.0513433406635433,
+      "grad_norm": 1.9933684794199173,
+      "learning_rate": 1e-06,
+      "loss": 0.1024,
+      "step": 5498
+    },
+    {
+      "epoch": 1.0515345635337987,
+      "grad_norm": 2.6262487438549322,
+      "learning_rate": 1e-06,
+      "loss": 0.0958,
+      "step": 5499
+    },
+    {
+      "epoch": 1.051725786404054,
+      "grad_norm": 1.9245088718652827,
+      "learning_rate": 1e-06,
+      "loss": 0.0934,
+      "step": 5500
+    },
+    {
+      "epoch": 1.051725786404054,
+      "eval_runtime": 754.3184,
+      "eval_samples_per_second": 2.034,
+      "eval_steps_per_second": 0.509,
+      "step": 5500
+    },
+    {
+      "epoch": 1.0519170092743093,
+      "grad_norm": 2.119611977749291,
+      "learning_rate": 1e-06,
+      "loss": 0.0811,
+      "step": 5501
+    },
+    {
+      "epoch": 1.0521082321445645,
+      "grad_norm": 4.728705810856696,
+      "learning_rate": 1e-06,
+      "loss": 0.0508,
+      "step": 5502
+    },
+    {
+      "epoch": 1.0522994550148197,
+      "grad_norm": 3.1869359732864337,
+      "learning_rate": 1e-06,
+      "loss": 0.106,
+      "step": 5503
+    },
+    {
+      "epoch": 1.052490677885075,
+      "grad_norm": 2.7526014334605833,
+      "learning_rate": 1e-06,
+      "loss": 0.0648,
+      "step": 5504
+    },
+    {
+      "epoch": 1.0526819007553303,
+      "grad_norm": 6.997030173144638,
+      "learning_rate": 1e-06,
+      "loss": 0.1222,
+      "step": 5505
+    },
+    {
+      "epoch": 1.0528731236255857,
+      "grad_norm": 8.396863951425557,
+      "learning_rate": 1e-06,
+      "loss": 0.3153,
+      "step": 5506
+    },
+    {
+      "epoch": 1.053064346495841,
+      "grad_norm": 5.832090772259885,
+      "learning_rate": 1e-06,
+      "loss": 0.2164,
+      "step": 5507
+    },
+    {
+      "epoch": 1.053255569366096,
+      "grad_norm": 2.7374482241398086,
+      "learning_rate": 1e-06,
+      "loss": 0.1925,
+      "step": 5508
+    },
+    {
+      "epoch": 1.0534467922363515,
+      "grad_norm": 1.7438175119983104,
+      "learning_rate": 1e-06,
+      "loss": 0.0625,
+      "step": 5509
+    },
+    {
+      "epoch": 1.0536380151066067,
+      "grad_norm": 2.7663986072364684,
+      "learning_rate": 1e-06,
+      "loss": 0.1159,
+      "step": 5510
+    },
+    {
+      "epoch": 1.0538292379768621,
+      "grad_norm": 1.9966429908422885,
+      "learning_rate": 1e-06,
+      "loss": 0.078,
+      "step": 5511
+    },
+    {
+      "epoch": 1.0540204608471173,
+      "grad_norm": 2.063013648605131,
+      "learning_rate": 1e-06,
+      "loss": 0.088,
+      "step": 5512
+    },
+    {
+      "epoch": 1.0542116837173725,
+      "grad_norm": 1.527428510026236,
+      "learning_rate": 1e-06,
+      "loss": 0.0181,
+      "step": 5513
+    },
+    {
+      "epoch": 1.054402906587628,
+      "grad_norm": 6.364204690994543,
+      "learning_rate": 1e-06,
+      "loss": 0.271,
+      "step": 5514
+    },
+    {
+      "epoch": 1.0545941294578831,
+      "grad_norm": 4.903789121776597,
+      "learning_rate": 1e-06,
+      "loss": 0.0836,
+      "step": 5515
+    },
+    {
+      "epoch": 1.0547853523281385,
+      "grad_norm": 2.829364483920998,
+      "learning_rate": 1e-06,
+      "loss": 0.0251,
+      "step": 5516
+    },
+    {
+      "epoch": 1.0549765751983937,
+      "grad_norm": 2.6812558936563375,
+      "learning_rate": 1e-06,
+      "loss": 0.0267,
+      "step": 5517
+    },
+    {
+      "epoch": 1.055167798068649,
+      "grad_norm": 7.727818648285622,
+      "learning_rate": 1e-06,
+      "loss": 0.3197,
+      "step": 5518
+    },
+    {
+      "epoch": 1.0553590209389043,
+      "grad_norm": 3.739630030320095,
+      "learning_rate": 1e-06,
+      "loss": 0.2431,
+      "step": 5519
+    },
+    {
+      "epoch": 1.0555502438091595,
+      "grad_norm": 4.37666158458437,
+      "learning_rate": 1e-06,
+      "loss": 0.2578,
+      "step": 5520
+    },
+    {
+      "epoch": 1.055741466679415,
+      "grad_norm": 2.247963619596206,
+      "learning_rate": 1e-06,
+      "loss": 0.2477,
+      "step": 5521
+    },
+    {
+      "epoch": 1.0559326895496701,
+      "grad_norm": 2.645543590710677,
+      "learning_rate": 1e-06,
+      "loss": 0.1484,
+      "step": 5522
+    },
+    {
+      "epoch": 1.0561239124199253,
+      "grad_norm": 3.36321759136002,
+      "learning_rate": 1e-06,
+      "loss": 0.2254,
+      "step": 5523
+    },
+    {
+      "epoch": 1.0563151352901807,
+      "grad_norm": 2.6096797011410438,
+      "learning_rate": 1e-06,
+      "loss": 0.0889,
+      "step": 5524
+    },
+    {
+      "epoch": 1.056506358160436,
+      "grad_norm": 2.750678325570429,
+      "learning_rate": 1e-06,
+      "loss": 0.0723,
+      "step": 5525
+    },
+    {
+      "epoch": 1.0566975810306913,
+      "grad_norm": 2.37431526351937,
+      "learning_rate": 1e-06,
+      "loss": 0.112,
+      "step": 5526
+    },
+    {
+      "epoch": 1.0568888039009465,
+      "grad_norm": 2.981133583083044,
+      "learning_rate": 1e-06,
+      "loss": 0.0771,
+      "step": 5527
+    },
+    {
+      "epoch": 1.057080026771202,
+      "grad_norm": 2.223918536244813,
+      "learning_rate": 1e-06,
+      "loss": 0.0285,
+      "step": 5528
+    },
+    {
+      "epoch": 1.0572712496414571,
+      "grad_norm": 3.0578456465572206,
+      "learning_rate": 1e-06,
+      "loss": 0.05,
+      "step": 5529
+    },
+    {
+      "epoch": 1.0574624725117123,
+      "grad_norm": 5.069046309120561,
+      "learning_rate": 1e-06,
+      "loss": 0.2004,
+      "step": 5530
+    },
+    {
+      "epoch": 1.0576536953819677,
+      "grad_norm": 4.309481684060614,
+      "learning_rate": 1e-06,
+      "loss": 0.2828,
+      "step": 5531
+    },
+    {
+      "epoch": 1.057844918252223,
+      "grad_norm": 6.656091535947512,
+      "learning_rate": 1e-06,
+      "loss": 0.2338,
+      "step": 5532
+    },
+    {
+      "epoch": 1.0580361411224783,
+      "grad_norm": 2.981805304226695,
+      "learning_rate": 1e-06,
+      "loss": 0.194,
+      "step": 5533
+    },
+    {
+      "epoch": 1.0582273639927335,
+      "grad_norm": 2.794882464180994,
+      "learning_rate": 1e-06,
+      "loss": 0.276,
+      "step": 5534
+    },
+    {
+      "epoch": 1.0584185868629887,
+      "grad_norm": 2.6910708457894796,
+      "learning_rate": 1e-06,
+      "loss": 0.1365,
+      "step": 5535
+    },
+    {
+      "epoch": 1.0586098097332441,
+      "grad_norm": 3.0609812181825626,
+      "learning_rate": 1e-06,
+      "loss": 0.1667,
+      "step": 5536
+    },
+    {
+      "epoch": 1.0588010326034993,
+      "grad_norm": 2.2743523189276487,
+      "learning_rate": 1e-06,
+      "loss": 0.0915,
+      "step": 5537
+    },
+    {
+      "epoch": 1.0589922554737548,
+      "grad_norm": 2.988453579589832,
+      "learning_rate": 1e-06,
+      "loss": 0.1088,
+      "step": 5538
+    },
+    {
+      "epoch": 1.05918347834401,
+      "grad_norm": 2.4174439287438716,
+      "learning_rate": 1e-06,
+      "loss": 0.0212,
+      "step": 5539
+    },
+    {
+      "epoch": 1.0593747012142651,
+      "grad_norm": 1.5315305491759157,
+      "learning_rate": 1e-06,
+      "loss": 0.0207,
+      "step": 5540
+    },
+    {
+      "epoch": 1.0595659240845205,
+      "grad_norm": 3.5957207624923893,
+      "learning_rate": 1e-06,
+      "loss": 0.0612,
+      "step": 5541
+    },
+    {
+      "epoch": 1.0597571469547757,
+      "grad_norm": 2.991015969354003,
+      "learning_rate": 1e-06,
+      "loss": 0.0709,
+      "step": 5542
+    },
+    {
+      "epoch": 1.0599483698250312,
+      "grad_norm": 4.835963136669728,
+      "learning_rate": 1e-06,
+      "loss": 0.198,
+      "step": 5543
+    },
+    {
+      "epoch": 1.0601395926952863,
+      "grad_norm": 3.9299038472922465,
+      "learning_rate": 1e-06,
+      "loss": 0.4752,
+      "step": 5544
+    },
+    {
+      "epoch": 1.0603308155655415,
+      "grad_norm": 4.047385165191848,
+      "learning_rate": 1e-06,
+      "loss": 0.5869,
+      "step": 5545
+    },
+    {
+      "epoch": 1.060522038435797,
+      "grad_norm": 2.8924853138068616,
+      "learning_rate": 1e-06,
+      "loss": 0.1681,
+      "step": 5546
+    },
+    {
+      "epoch": 1.0607132613060521,
+      "grad_norm": 2.916761196966072,
+      "learning_rate": 1e-06,
+      "loss": 0.2147,
+      "step": 5547
+    },
+    {
+      "epoch": 1.0609044841763076,
+      "grad_norm": 3.964602368152579,
+      "learning_rate": 1e-06,
+      "loss": 0.21,
+      "step": 5548
+    },
+    {
+      "epoch": 1.0610957070465628,
+      "grad_norm": 2.602072817493831,
+      "learning_rate": 1e-06,
+      "loss": 0.0494,
+      "step": 5549
+    },
+    {
+      "epoch": 1.061286929916818,
+      "grad_norm": 2.5850376741774004,
+      "learning_rate": 1e-06,
+      "loss": 0.1309,
+      "step": 5550
+    },
+    {
+      "epoch": 1.0614781527870734,
+      "grad_norm": 3.7286272706650854,
+      "learning_rate": 1e-06,
+      "loss": 0.0985,
+      "step": 5551
+    },
+    {
+      "epoch": 1.0616693756573286,
+      "grad_norm": 4.851308917361368,
+      "learning_rate": 1e-06,
+      "loss": 0.2462,
+      "step": 5552
+    },
+    {
+      "epoch": 1.061860598527584,
+      "grad_norm": 1.800819049033031,
+      "learning_rate": 1e-06,
+      "loss": 0.035,
+      "step": 5553
+    },
+    {
+      "epoch": 1.0620518213978392,
+      "grad_norm": 3.212474742745586,
+      "learning_rate": 1e-06,
+      "loss": 0.0511,
+      "step": 5554
+    },
+    {
+      "epoch": 1.0622430442680946,
+      "grad_norm": 5.167645802574818,
+      "learning_rate": 1e-06,
+      "loss": 0.1197,
+      "step": 5555
+    },
+    {
+      "epoch": 1.0624342671383498,
+      "grad_norm": 4.736301546317177,
+      "learning_rate": 1e-06,
+      "loss": 0.5203,
+      "step": 5556
+    },
+    {
+      "epoch": 1.062625490008605,
+      "grad_norm": 3.504085336316662,
+      "learning_rate": 1e-06,
+      "loss": 0.2016,
+      "step": 5557
+    },
+    {
+      "epoch": 1.0628167128788604,
+      "grad_norm": 2.5352952445886627,
+      "learning_rate": 1e-06,
+      "loss": 0.1601,
+      "step": 5558
+    },
+    {
+      "epoch": 1.0630079357491156,
+      "grad_norm": 2.0654997540937585,
+      "learning_rate": 1e-06,
+      "loss": 0.079,
+      "step": 5559
+    },
+    {
+      "epoch": 1.063199158619371,
+      "grad_norm": 3.1760716229297303,
+      "learning_rate": 1e-06,
+      "loss": 0.2965,
+      "step": 5560
+    },
+    {
+      "epoch": 1.0633903814896262,
+      "grad_norm": 2.061577677408497,
+      "learning_rate": 1e-06,
+      "loss": 0.1104,
+      "step": 5561
+    },
+    {
+      "epoch": 1.0635816043598814,
+      "grad_norm": 3.320510104414023,
+      "learning_rate": 1e-06,
+      "loss": 0.0563,
+      "step": 5562
+    },
+    {
+      "epoch": 1.0637728272301368,
+      "grad_norm": 4.102519884543666,
+      "learning_rate": 1e-06,
+      "loss": 0.0886,
+      "step": 5563
+    },
+    {
+      "epoch": 1.063964050100392,
+      "grad_norm": 2.845635931076547,
+      "learning_rate": 1e-06,
+      "loss": 0.0793,
+      "step": 5564
+    },
+    {
+      "epoch": 1.0641552729706474,
+      "grad_norm": 1.8693341799723675,
+      "learning_rate": 1e-06,
+      "loss": 0.0189,
+      "step": 5565
+    },
+    {
+      "epoch": 1.0643464958409026,
+      "grad_norm": 1.9675013503151202,
+      "learning_rate": 1e-06,
+      "loss": 0.0343,
+      "step": 5566
+    },
+    {
+      "epoch": 1.0645377187111578,
+      "grad_norm": 3.4585772661945535,
+      "learning_rate": 1e-06,
+      "loss": 0.0765,
+      "step": 5567
+    },
+    {
+      "epoch": 1.0647289415814132,
+      "grad_norm": 2.9143426264487053,
+      "learning_rate": 1e-06,
+      "loss": 0.2552,
+      "step": 5568
+    },
+    {
+      "epoch": 1.0649201644516684,
+      "grad_norm": 4.534464991259499,
+      "learning_rate": 1e-06,
+      "loss": 0.336,
+      "step": 5569
+    },
+    {
+      "epoch": 1.0651113873219238,
+      "grad_norm": 4.44409613569023,
+      "learning_rate": 1e-06,
+      "loss": 0.2242,
+      "step": 5570
+    },
+    {
+      "epoch": 1.065302610192179,
+      "grad_norm": 2.4735674632973725,
+      "learning_rate": 1e-06,
+      "loss": 0.2133,
+      "step": 5571
+    },
+    {
+      "epoch": 1.0654938330624342,
+      "grad_norm": 5.10941512999101,
+      "learning_rate": 1e-06,
+      "loss": 0.3382,
+      "step": 5572
+    },
+    {
+      "epoch": 1.0656850559326896,
+      "grad_norm": 3.621314839318215,
+      "learning_rate": 1e-06,
+      "loss": 0.23,
+      "step": 5573
+    },
+    {
+      "epoch": 1.0658762788029448,
+      "grad_norm": 3.0526806417215995,
+      "learning_rate": 1e-06,
+      "loss": 0.205,
+      "step": 5574
+    },
+    {
+      "epoch": 1.0660675016732002,
+      "grad_norm": 3.06511498738045,
+      "learning_rate": 1e-06,
+      "loss": 0.1588,
+      "step": 5575
+    },
+    {
+      "epoch": 1.0662587245434554,
+      "grad_norm": 1.7624138182002693,
+      "learning_rate": 1e-06,
+      "loss": 0.0474,
+      "step": 5576
+    },
+    {
+      "epoch": 1.0664499474137106,
+      "grad_norm": 1.797701504571653,
+      "learning_rate": 1e-06,
+      "loss": 0.0224,
+      "step": 5577
+    },
+    {
+      "epoch": 1.066641170283966,
+      "grad_norm": 2.6038342886846095,
+      "learning_rate": 1e-06,
+      "loss": 0.0641,
+      "step": 5578
+    },
+    {
+      "epoch": 1.0668323931542212,
+      "grad_norm": 3.3034469404294087,
+      "learning_rate": 1e-06,
+      "loss": 0.0797,
+      "step": 5579
+    },
+    {
+      "epoch": 1.0670236160244766,
+      "grad_norm": 30.536723057298158,
+      "learning_rate": 1e-06,
+      "loss": 0.2204,
+      "step": 5580
+    },
+    {
+      "epoch": 1.0672148388947318,
+      "grad_norm": 4.2410574684773765,
+      "learning_rate": 1e-06,
+      "loss": 0.4152,
+      "step": 5581
+    },
+    {
+      "epoch": 1.067406061764987,
+      "grad_norm": 4.194734842606893,
+      "learning_rate": 1e-06,
+      "loss": 0.3324,
+      "step": 5582
+    },
+    {
+      "epoch": 1.0675972846352424,
+      "grad_norm": 2.97346940617324,
+      "learning_rate": 1e-06,
+      "loss": 0.2028,
+      "step": 5583
+    },
+    {
+      "epoch": 1.0677885075054976,
+      "grad_norm": 2.568736233639549,
+      "learning_rate": 1e-06,
+      "loss": 0.0899,
+      "step": 5584
+    },
+    {
+      "epoch": 1.067979730375753,
+      "grad_norm": 4.233422879651181,
+      "learning_rate": 1e-06,
+      "loss": 0.3406,
+      "step": 5585
+    },
+    {
+      "epoch": 1.0681709532460082,
+      "grad_norm": 3.5553558727327004,
+      "learning_rate": 1e-06,
+      "loss": 0.1326,
+      "step": 5586
+    },
+    {
+      "epoch": 1.0683621761162636,
+      "grad_norm": 1.6803923081835581,
+      "learning_rate": 1e-06,
+      "loss": 0.0417,
+      "step": 5587
+    },
+    {
+      "epoch": 1.0685533989865188,
+      "grad_norm": 3.619501745382458,
+      "learning_rate": 1e-06,
+      "loss": 0.1466,
+      "step": 5588
+    },
+    {
+      "epoch": 1.068744621856774,
+      "grad_norm": 2.8200436619834743,
+      "learning_rate": 1e-06,
+      "loss": 0.0948,
+      "step": 5589
+    },
+    {
+      "epoch": 1.0689358447270294,
+      "grad_norm": 2.4979032306665063,
+      "learning_rate": 1e-06,
+      "loss": 0.0483,
+      "step": 5590
+    },
+    {
+      "epoch": 1.0691270675972846,
+      "grad_norm": 2.037918997902149,
+      "learning_rate": 1e-06,
+      "loss": 0.0248,
+      "step": 5591
+    },
+    {
+      "epoch": 1.06931829046754,
+      "grad_norm": 8.769685236677006,
+      "learning_rate": 1e-06,
+      "loss": 0.1429,
+      "step": 5592
+    },
+    {
+      "epoch": 1.0695095133377952,
+      "grad_norm": 5.537812903112377,
+      "learning_rate": 1e-06,
+      "loss": 0.2018,
+      "step": 5593
+    },
+    {
+      "epoch": 1.0697007362080504,
+      "grad_norm": 4.299081021070948,
+      "learning_rate": 1e-06,
+      "loss": 0.3155,
+      "step": 5594
+    },
+    {
+      "epoch": 1.0698919590783058,
+      "grad_norm": 4.101507393603025,
+      "learning_rate": 1e-06,
+      "loss": 0.4299,
+      "step": 5595
+    },
+    {
+      "epoch": 1.070083181948561,
+      "grad_norm": 4.3855483276223435,
+      "learning_rate": 1e-06,
+      "loss": 0.1762,
+      "step": 5596
+    },
+    {
+      "epoch": 1.0702744048188164,
+      "grad_norm": 3.5717195038144056,
+      "learning_rate": 1e-06,
+      "loss": 0.2695,
+      "step": 5597
+    },
+    {
+      "epoch": 1.0704656276890716,
+      "grad_norm": 1.7303829817689438,
+      "learning_rate": 1e-06,
+      "loss": 0.1267,
+      "step": 5598
+    },
+    {
+      "epoch": 1.0706568505593268,
+      "grad_norm": 4.635368478360246,
+      "learning_rate": 1e-06,
+      "loss": 0.1419,
+      "step": 5599
+    },
+    {
+      "epoch": 1.0708480734295822,
+      "grad_norm": 3.7766437059351508,
+      "learning_rate": 1e-06,
+      "loss": 0.2238,
+      "step": 5600
+    },
+    {
+      "epoch": 1.0710392962998374,
+      "grad_norm": 2.3599196051788955,
+      "learning_rate": 1e-06,
+      "loss": 0.1413,
+      "step": 5601
+    },
+    {
+      "epoch": 1.0712305191700928,
+      "grad_norm": 3.9836232392438484,
+      "learning_rate": 1e-06,
+      "loss": 0.0748,
+      "step": 5602
+    },
+    {
+      "epoch": 1.071421742040348,
+      "grad_norm": 2.554606398482117,
+      "learning_rate": 1e-06,
+      "loss": 0.0853,
+      "step": 5603
+    },
+    {
+      "epoch": 1.0716129649106032,
+      "grad_norm": 1.6696037480237755,
+      "learning_rate": 1e-06,
+      "loss": 0.0312,
+      "step": 5604
+    },
+    {
+      "epoch": 1.0718041877808586,
+      "grad_norm": 2.324042408127832,
+      "learning_rate": 1e-06,
+      "loss": 0.0702,
+      "step": 5605
+    },
+    {
+      "epoch": 1.0719954106511138,
+      "grad_norm": 6.016321236089402,
+      "learning_rate": 1e-06,
+      "loss": 0.3191,
+      "step": 5606
+    },
+    {
+      "epoch": 1.0721866335213692,
+      "grad_norm": 6.061439342723606,
+      "learning_rate": 1e-06,
+      "loss": 0.6396,
+      "step": 5607
+    },
+    {
+      "epoch": 1.0723778563916244,
+      "grad_norm": 3.7182851548937217,
+      "learning_rate": 1e-06,
+      "loss": 0.2923,
+      "step": 5608
+    },
+    {
+      "epoch": 1.0725690792618798,
+      "grad_norm": 4.144406991834277,
+      "learning_rate": 1e-06,
+      "loss": 0.3817,
+      "step": 5609
+    },
+    {
+      "epoch": 1.072760302132135,
+      "grad_norm": 3.7481471252551697,
+      "learning_rate": 1e-06,
+      "loss": 0.3513,
+      "step": 5610
+    },
+    {
+      "epoch": 1.0729515250023902,
+      "grad_norm": 3.6501975541672045,
+      "learning_rate": 1e-06,
+      "loss": 0.217,
+      "step": 5611
+    },
+    {
+      "epoch": 1.0731427478726456,
+      "grad_norm": 1.5810629813228263,
+      "learning_rate": 1e-06,
+      "loss": 0.0738,
+      "step": 5612
+    },
+    {
+      "epoch": 1.0733339707429008,
+      "grad_norm": 1.5959314115313665,
+      "learning_rate": 1e-06,
+      "loss": 0.0784,
+      "step": 5613
+    },
+    {
+      "epoch": 1.0735251936131562,
+      "grad_norm": 4.055115546709855,
+      "learning_rate": 1e-06,
+      "loss": 0.0387,
+      "step": 5614
+    },
+    {
+      "epoch": 1.0737164164834114,
+      "grad_norm": 2.564758979821058,
+      "learning_rate": 1e-06,
+      "loss": 0.0478,
+      "step": 5615
+    },
+    {
+      "epoch": 1.0739076393536666,
+      "grad_norm": 1.330992576174306,
+      "learning_rate": 1e-06,
+      "loss": 0.0139,
+      "step": 5616
+    },
+    {
+      "epoch": 1.074098862223922,
+      "grad_norm": 4.393505739718172,
+      "learning_rate": 1e-06,
+      "loss": 0.0442,
+      "step": 5617
+    },
+    {
+      "epoch": 1.0742900850941772,
+      "grad_norm": 7.415519997182482,
+      "learning_rate": 1e-06,
+      "loss": 0.2741,
+      "step": 5618
+    },
+    {
+      "epoch": 1.0744813079644326,
+      "grad_norm": 9.419244959026415,
+      "learning_rate": 1e-06,
+      "loss": 0.5119,
+      "step": 5619
+    },
+    {
+      "epoch": 1.0746725308346878,
+      "grad_norm": 2.54612177844801,
+      "learning_rate": 1e-06,
+      "loss": 0.2908,
+      "step": 5620
+    },
+    {
+      "epoch": 1.074863753704943,
+      "grad_norm": 3.6636077663752062,
+      "learning_rate": 1e-06,
+      "loss": 0.1563,
+      "step": 5621
+    },
+    {
+      "epoch": 1.0750549765751984,
+      "grad_norm": 3.4601562169260838,
+      "learning_rate": 1e-06,
+      "loss": 0.224,
+      "step": 5622
+    },
+    {
+      "epoch": 1.0752461994454536,
+      "grad_norm": 3.394235824178862,
+      "learning_rate": 1e-06,
+      "loss": 0.089,
+      "step": 5623
+    },
+    {
+      "epoch": 1.075437422315709,
+      "grad_norm": 1.9566424625119267,
+      "learning_rate": 1e-06,
+      "loss": 0.0347,
+      "step": 5624
+    },
+    {
+      "epoch": 1.0756286451859642,
+      "grad_norm": 2.2537454584110854,
+      "learning_rate": 1e-06,
+      "loss": 0.056,
+      "step": 5625
+    },
+    {
+      "epoch": 1.0758198680562194,
+      "grad_norm": 3.988821503067968,
+      "learning_rate": 1e-06,
+      "loss": 0.1853,
+      "step": 5626
+    },
+    {
+      "epoch": 1.0760110909264748,
+      "grad_norm": 3.228337794747041,
+      "learning_rate": 1e-06,
+      "loss": 0.1639,
+      "step": 5627
+    },
+    {
+      "epoch": 1.07620231379673,
+      "grad_norm": 3.3808131134905417,
+      "learning_rate": 1e-06,
+      "loss": 0.0536,
+      "step": 5628
+    },
+    {
+      "epoch": 1.0763935366669855,
+      "grad_norm": 4.111055559922836,
+      "learning_rate": 1e-06,
+      "loss": 0.0857,
+      "step": 5629
+    },
+    {
+      "epoch": 1.0765847595372406,
+      "grad_norm": 7.98347936933366,
+      "learning_rate": 1e-06,
+      "loss": 0.1527,
+      "step": 5630
+    },
+    {
+      "epoch": 1.0767759824074958,
+      "grad_norm": 6.650830278694239,
+      "learning_rate": 1e-06,
+      "loss": 0.4131,
+      "step": 5631
+    },
+    {
+      "epoch": 1.0769672052777512,
+      "grad_norm": 4.479896355860583,
+      "learning_rate": 1e-06,
+      "loss": 0.2896,
+      "step": 5632
+    },
+    {
+      "epoch": 1.0771584281480064,
+      "grad_norm": 6.589129379255085,
+      "learning_rate": 1e-06,
+      "loss": 0.2028,
+      "step": 5633
+    },
+    {
+      "epoch": 1.0773496510182619,
+      "grad_norm": 2.344878268467919,
+      "learning_rate": 1e-06,
+      "loss": 0.1227,
+      "step": 5634
+    },
+    {
+      "epoch": 1.077540873888517,
+      "grad_norm": 3.7323505867897726,
+      "learning_rate": 1e-06,
+      "loss": 0.1396,
+      "step": 5635
+    },
+    {
+      "epoch": 1.0777320967587722,
+      "grad_norm": 3.646582971434177,
+      "learning_rate": 1e-06,
+      "loss": 0.187,
+      "step": 5636
+    },
+    {
+      "epoch": 1.0779233196290277,
+      "grad_norm": 1.5616564190072704,
+      "learning_rate": 1e-06,
+      "loss": 0.0289,
+      "step": 5637
+    },
+    {
+      "epoch": 1.0781145424992828,
+      "grad_norm": 2.7854063716252124,
+      "learning_rate": 1e-06,
+      "loss": 0.0585,
+      "step": 5638
+    },
+    {
+      "epoch": 1.0783057653695383,
+      "grad_norm": 4.08292337552461,
+      "learning_rate": 1e-06,
+      "loss": 0.0396,
+      "step": 5639
+    },
+    {
+      "epoch": 1.0784969882397935,
+      "grad_norm": 1.5361145755682,
+      "learning_rate": 1e-06,
+      "loss": 0.0179,
+      "step": 5640
+    },
+    {
+      "epoch": 1.0786882111100486,
+      "grad_norm": 7.185941046116487,
+      "learning_rate": 1e-06,
+      "loss": 0.1526,
+      "step": 5641
+    },
+    {
+      "epoch": 1.078879433980304,
+      "grad_norm": 4.834675810713905,
+      "learning_rate": 1e-06,
+      "loss": 0.1189,
+      "step": 5642
+    },
+    {
+      "epoch": 1.0790706568505593,
+      "grad_norm": 3.7977838880493384,
+      "learning_rate": 1e-06,
+      "loss": 0.3031,
+      "step": 5643
+    },
+    {
+      "epoch": 1.0792618797208147,
+      "grad_norm": 3.225149071375271,
+      "learning_rate": 1e-06,
+      "loss": 0.276,
+      "step": 5644
+    },
+    {
+      "epoch": 1.0794531025910699,
+      "grad_norm": 3.874311447496149,
+      "learning_rate": 1e-06,
+      "loss": 0.3118,
+      "step": 5645
+    },
+    {
+      "epoch": 1.0796443254613253,
+      "grad_norm": 3.0437573254632233,
+      "learning_rate": 1e-06,
+      "loss": 0.1114,
+      "step": 5646
+    },
+    {
+      "epoch": 1.0798355483315805,
+      "grad_norm": 6.142051472852211,
+      "learning_rate": 1e-06,
+      "loss": 0.367,
+      "step": 5647
+    },
+    {
+      "epoch": 1.0800267712018357,
+      "grad_norm": 3.689107899949952,
+      "learning_rate": 1e-06,
+      "loss": 0.1741,
+      "step": 5648
+    },
+    {
+      "epoch": 1.080217994072091,
+      "grad_norm": 2.325284928121335,
+      "learning_rate": 1e-06,
+      "loss": 0.0931,
+      "step": 5649
+    },
+    {
+      "epoch": 1.0804092169423463,
+      "grad_norm": 4.099161080852088,
+      "learning_rate": 1e-06,
+      "loss": 0.1862,
+      "step": 5650
+    },
+    {
+      "epoch": 1.0806004398126017,
+      "grad_norm": 3.946951527582266,
+      "learning_rate": 1e-06,
+      "loss": 0.0846,
+      "step": 5651
+    },
+    {
+      "epoch": 1.0807916626828569,
+      "grad_norm": 1.3827246083059548,
+      "learning_rate": 1e-06,
+      "loss": 0.0223,
+      "step": 5652
+    },
+    {
+      "epoch": 1.080982885553112,
+      "grad_norm": 2.4753466970197886,
+      "learning_rate": 1e-06,
+      "loss": 0.067,
+      "step": 5653
+    },
+    {
+      "epoch": 1.0811741084233675,
+      "grad_norm": 3.540147088868997,
+      "learning_rate": 1e-06,
+      "loss": 0.2016,
+      "step": 5654
+    },
+    {
+      "epoch": 1.0813653312936227,
+      "grad_norm": 7.060848616909866,
+      "learning_rate": 1e-06,
+      "loss": 0.1307,
+      "step": 5655
+    },
+    {
+      "epoch": 1.081556554163878,
+      "grad_norm": 4.47651605765067,
+      "learning_rate": 1e-06,
+      "loss": 0.2634,
+      "step": 5656
+    },
+    {
+      "epoch": 1.0817477770341333,
+      "grad_norm": 4.387916027896511,
+      "learning_rate": 1e-06,
+      "loss": 0.3676,
+      "step": 5657
+    },
+    {
+      "epoch": 1.0819389999043885,
+      "grad_norm": 3.0848562929881633,
+      "learning_rate": 1e-06,
+      "loss": 0.2566,
+      "step": 5658
+    },
+    {
+      "epoch": 1.0821302227746439,
+      "grad_norm": 2.9690742717124516,
+      "learning_rate": 1e-06,
+      "loss": 0.1426,
+      "step": 5659
+    },
+    {
+      "epoch": 1.082321445644899,
+      "grad_norm": 2.6327294582775074,
+      "learning_rate": 1e-06,
+      "loss": 0.0761,
+      "step": 5660
+    },
+    {
+      "epoch": 1.0825126685151545,
+      "grad_norm": 3.5055595928584684,
+      "learning_rate": 1e-06,
+      "loss": 0.0682,
+      "step": 5661
+    },
+    {
+      "epoch": 1.0827038913854097,
+      "grad_norm": 2.6301171516753516,
+      "learning_rate": 1e-06,
+      "loss": 0.0751,
+      "step": 5662
+    },
+    {
+      "epoch": 1.082895114255665,
+      "grad_norm": 2.422012718961097,
+      "learning_rate": 1e-06,
+      "loss": 0.1154,
+      "step": 5663
+    },
+    {
+      "epoch": 1.0830863371259203,
+      "grad_norm": 4.790374625892685,
+      "learning_rate": 1e-06,
+      "loss": 0.2421,
+      "step": 5664
+    },
+    {
+      "epoch": 1.0832775599961755,
+      "grad_norm": 7.423706599653193,
+      "learning_rate": 1e-06,
+      "loss": 0.0604,
+      "step": 5665
+    },
+    {
+      "epoch": 1.083468782866431,
+      "grad_norm": 3.6164739906096184,
+      "learning_rate": 1e-06,
+      "loss": 0.0295,
+      "step": 5666
+    },
+    {
+      "epoch": 1.083660005736686,
+      "grad_norm": 5.802355044821207,
+      "learning_rate": 1e-06,
+      "loss": 0.099,
+      "step": 5667
+    },
+    {
+      "epoch": 1.0838512286069415,
+      "grad_norm": 10.105034629011206,
+      "learning_rate": 1e-06,
+      "loss": 0.3827,
+      "step": 5668
+    },
+    {
+      "epoch": 1.0840424514771967,
+      "grad_norm": 5.106504608119349,
+      "learning_rate": 1e-06,
+      "loss": 0.3219,
+      "step": 5669
+    },
+    {
+      "epoch": 1.0842336743474519,
+      "grad_norm": 2.992283592971937,
+      "learning_rate": 1e-06,
+      "loss": 0.1671,
+      "step": 5670
+    },
+    {
+      "epoch": 1.0844248972177073,
+      "grad_norm": 2.3690922190476873,
+      "learning_rate": 1e-06,
+      "loss": 0.1584,
+      "step": 5671
+    },
+    {
+      "epoch": 1.0846161200879625,
+      "grad_norm": 3.7042331244761457,
+      "learning_rate": 1e-06,
+      "loss": 0.3414,
+      "step": 5672
+    },
+    {
+      "epoch": 1.084807342958218,
+      "grad_norm": 2.449901625546119,
+      "learning_rate": 1e-06,
+      "loss": 0.0676,
+      "step": 5673
+    },
+    {
+      "epoch": 1.084998565828473,
+      "grad_norm": 2.8820258827397334,
+      "learning_rate": 1e-06,
+      "loss": 0.0634,
+      "step": 5674
+    },
+    {
+      "epoch": 1.0851897886987283,
+      "grad_norm": 2.0632993131841375,
+      "learning_rate": 1e-06,
+      "loss": 0.0342,
+      "step": 5675
+    },
+    {
+      "epoch": 1.0853810115689837,
+      "grad_norm": 1.9451973509508278,
+      "learning_rate": 1e-06,
+      "loss": 0.0605,
+      "step": 5676
+    },
+    {
+      "epoch": 1.085572234439239,
+      "grad_norm": 1.3306893220175653,
+      "learning_rate": 1e-06,
+      "loss": 0.0699,
+      "step": 5677
+    },
+    {
+      "epoch": 1.0857634573094943,
+      "grad_norm": 2.537680193887515,
+      "learning_rate": 1e-06,
+      "loss": 0.0385,
+      "step": 5678
+    },
+    {
+      "epoch": 1.0859546801797495,
+      "grad_norm": 1.672570983908536,
+      "learning_rate": 1e-06,
+      "loss": 0.0261,
+      "step": 5679
+    },
+    {
+      "epoch": 1.0861459030500047,
+      "grad_norm": 7.4176023371688675,
+      "learning_rate": 1e-06,
+      "loss": 0.1562,
+      "step": 5680
+    },
+    {
+      "epoch": 1.08633712592026,
+      "grad_norm": 7.453234785698888,
+      "learning_rate": 1e-06,
+      "loss": 0.2076,
+      "step": 5681
+    },
+    {
+      "epoch": 1.0865283487905153,
+      "grad_norm": 3.3281015028146466,
+      "learning_rate": 1e-06,
+      "loss": 0.2519,
+      "step": 5682
+    },
+    {
+      "epoch": 1.0867195716607707,
+      "grad_norm": 2.505141692884501,
+      "learning_rate": 1e-06,
+      "loss": 0.0949,
+      "step": 5683
+    },
+    {
+      "epoch": 1.086910794531026,
+      "grad_norm": 2.3309139469714526,
+      "learning_rate": 1e-06,
+      "loss": 0.0816,
+      "step": 5684
+    },
+    {
+      "epoch": 1.087102017401281,
+      "grad_norm": 4.330179656356448,
+      "learning_rate": 1e-06,
+      "loss": 0.2717,
+      "step": 5685
+    },
+    {
+      "epoch": 1.0872932402715365,
+      "grad_norm": 3.1835729305780136,
+      "learning_rate": 1e-06,
+      "loss": 0.0721,
+      "step": 5686
+    },
+    {
+      "epoch": 1.0874844631417917,
+      "grad_norm": 3.0673100591300404,
+      "learning_rate": 1e-06,
+      "loss": 0.0622,
+      "step": 5687
+    },
+    {
+      "epoch": 1.0876756860120471,
+      "grad_norm": 3.924074931379068,
+      "learning_rate": 1e-06,
+      "loss": 0.1948,
+      "step": 5688
+    },
+    {
+      "epoch": 1.0878669088823023,
+      "grad_norm": 2.5849017232663436,
+      "learning_rate": 1e-06,
+      "loss": 0.1471,
+      "step": 5689
+    },
+    {
+      "epoch": 1.0880581317525575,
+      "grad_norm": 2.1989021683216396,
+      "learning_rate": 1e-06,
+      "loss": 0.0489,
+      "step": 5690
+    },
+    {
+      "epoch": 1.088249354622813,
+      "grad_norm": 2.820155427315365,
+      "learning_rate": 1e-06,
+      "loss": 0.0766,
+      "step": 5691
+    },
+    {
+      "epoch": 1.088440577493068,
+      "grad_norm": 4.544560318464172,
+      "learning_rate": 1e-06,
+      "loss": 0.0669,
+      "step": 5692
+    },
+    {
+      "epoch": 1.0886318003633235,
+      "grad_norm": 4.414148272465876,
+      "learning_rate": 1e-06,
+      "loss": 0.287,
+      "step": 5693
+    },
+    {
+      "epoch": 1.0888230232335787,
+      "grad_norm": 3.0217236138721906,
+      "learning_rate": 1e-06,
+      "loss": 0.1247,
+      "step": 5694
+    },
+    {
+      "epoch": 1.089014246103834,
+      "grad_norm": 4.1459709924179045,
+      "learning_rate": 1e-06,
+      "loss": 0.3588,
+      "step": 5695
+    },
+    {
+      "epoch": 1.0892054689740893,
+      "grad_norm": 3.686927363012045,
+      "learning_rate": 1e-06,
+      "loss": 0.119,
+      "step": 5696
+    },
+    {
+      "epoch": 1.0893966918443445,
+      "grad_norm": 2.2742257865395907,
+      "learning_rate": 1e-06,
+      "loss": 0.2092,
+      "step": 5697
+    },
+    {
+      "epoch": 1.0895879147146,
+      "grad_norm": 1.8792869515276933,
+      "learning_rate": 1e-06,
+      "loss": 0.1255,
+      "step": 5698
+    },
+    {
+      "epoch": 1.0897791375848551,
+      "grad_norm": 3.9290819554722174,
+      "learning_rate": 1e-06,
+      "loss": 0.104,
+      "step": 5699
+    },
+    {
+      "epoch": 1.0899703604551105,
+      "grad_norm": 2.854411502184594,
+      "learning_rate": 1e-06,
+      "loss": 0.1327,
+      "step": 5700
+    },
+    {
+      "epoch": 1.0901615833253657,
+      "grad_norm": 4.751051334682677,
+      "learning_rate": 1e-06,
+      "loss": 0.133,
+      "step": 5701
+    },
+    {
+      "epoch": 1.090352806195621,
+      "grad_norm": 4.063173678367216,
+      "learning_rate": 1e-06,
+      "loss": 0.2002,
+      "step": 5702
+    },
+    {
+      "epoch": 1.0905440290658763,
+      "grad_norm": 2.638131493635213,
+      "learning_rate": 1e-06,
+      "loss": 0.1023,
+      "step": 5703
+    },
+    {
+      "epoch": 1.0907352519361315,
+      "grad_norm": 3.400082946214356,
+      "learning_rate": 1e-06,
+      "loss": 0.1987,
+      "step": 5704
+    },
+    {
+      "epoch": 1.090926474806387,
+      "grad_norm": 11.887885561365284,
+      "learning_rate": 1e-06,
+      "loss": 0.1969,
+      "step": 5705
+    },
+    {
+      "epoch": 1.0911176976766421,
+      "grad_norm": 6.006375739997399,
+      "learning_rate": 1e-06,
+      "loss": 0.6176,
+      "step": 5706
+    },
+    {
+      "epoch": 1.0913089205468973,
+      "grad_norm": 4.31131894068402,
+      "learning_rate": 1e-06,
+      "loss": 0.2824,
+      "step": 5707
+    },
+    {
+      "epoch": 1.0915001434171527,
+      "grad_norm": 2.6277054967834443,
+      "learning_rate": 1e-06,
+      "loss": 0.0801,
+      "step": 5708
+    },
+    {
+      "epoch": 1.091691366287408,
+      "grad_norm": 1.899794042868651,
+      "learning_rate": 1e-06,
+      "loss": 0.101,
+      "step": 5709
+    },
+    {
+      "epoch": 1.0918825891576633,
+      "grad_norm": 4.278535971118198,
+      "learning_rate": 1e-06,
+      "loss": 0.3349,
+      "step": 5710
+    },
+    {
+      "epoch": 1.0920738120279185,
+      "grad_norm": 4.350228945966433,
+      "learning_rate": 1e-06,
+      "loss": 0.2881,
+      "step": 5711
+    },
+    {
+      "epoch": 1.0922650348981737,
+      "grad_norm": 3.7243604891936544,
+      "learning_rate": 1e-06,
+      "loss": 0.1035,
+      "step": 5712
+    },
+    {
+      "epoch": 1.0924562577684291,
+      "grad_norm": 2.5410616941622686,
+      "learning_rate": 1e-06,
+      "loss": 0.0628,
+      "step": 5713
+    },
+    {
+      "epoch": 1.0926474806386843,
+      "grad_norm": 2.3995467115538625,
+      "learning_rate": 1e-06,
+      "loss": 0.081,
+      "step": 5714
+    },
+    {
+      "epoch": 1.0928387035089397,
+      "grad_norm": 3.3647568286928777,
+      "learning_rate": 1e-06,
+      "loss": 0.0603,
+      "step": 5715
+    },
+    {
+      "epoch": 1.093029926379195,
+      "grad_norm": 2.0868749904946746,
+      "learning_rate": 1e-06,
+      "loss": 0.028,
+      "step": 5716
+    },
+    {
+      "epoch": 1.0932211492494501,
+      "grad_norm": 4.643796511349131,
+      "learning_rate": 1e-06,
+      "loss": 0.0413,
+      "step": 5717
+    },
+    {
+      "epoch": 1.0934123721197055,
+      "grad_norm": 4.565291295529466,
+      "learning_rate": 1e-06,
+      "loss": 0.5042,
+      "step": 5718
+    },
+    {
+      "epoch": 1.0936035949899607,
+      "grad_norm": 4.24837126096245,
+      "learning_rate": 1e-06,
+      "loss": 0.4593,
+      "step": 5719
+    },
+    {
+      "epoch": 1.0937948178602162,
+      "grad_norm": 3.624587857724889,
+      "learning_rate": 1e-06,
+      "loss": 0.2837,
+      "step": 5720
+    },
+    {
+      "epoch": 1.0939860407304713,
+      "grad_norm": 3.1281906624075098,
+      "learning_rate": 1e-06,
+      "loss": 0.1746,
+      "step": 5721
+    },
+    {
+      "epoch": 1.0941772636007268,
+      "grad_norm": 3.4267015949898463,
+      "learning_rate": 1e-06,
+      "loss": 0.2406,
+      "step": 5722
+    },
+    {
+      "epoch": 1.094368486470982,
+      "grad_norm": 4.2021941675192975,
+      "learning_rate": 1e-06,
+      "loss": 0.2204,
+      "step": 5723
+    },
+    {
+      "epoch": 1.0945597093412371,
+      "grad_norm": 2.47038089626145,
+      "learning_rate": 1e-06,
+      "loss": 0.1228,
+      "step": 5724
+    },
+    {
+      "epoch": 1.0947509322114926,
+      "grad_norm": 2.3832694194022133,
+      "learning_rate": 1e-06,
+      "loss": 0.0856,
+      "step": 5725
+    },
+    {
+      "epoch": 1.0949421550817477,
+      "grad_norm": 5.25049806684577,
+      "learning_rate": 1e-06,
+      "loss": 0.1433,
+      "step": 5726
+    },
+    {
+      "epoch": 1.0951333779520032,
+      "grad_norm": 2.059447717612858,
+      "learning_rate": 1e-06,
+      "loss": 0.0519,
+      "step": 5727
+    },
+    {
+      "epoch": 1.0953246008222584,
+      "grad_norm": 2.357064435735991,
+      "learning_rate": 1e-06,
+      "loss": 0.0667,
+      "step": 5728
+    },
+    {
+      "epoch": 1.0955158236925135,
+      "grad_norm": 1.9098282670912805,
+      "learning_rate": 1e-06,
+      "loss": 0.0507,
+      "step": 5729
+    },
+    {
+      "epoch": 1.095707046562769,
+      "grad_norm": 6.779319101537318,
+      "learning_rate": 1e-06,
+      "loss": 0.0889,
+      "step": 5730
+    },
+    {
+      "epoch": 1.0958982694330242,
+      "grad_norm": 7.1659051653213295,
+      "learning_rate": 1e-06,
+      "loss": 0.2706,
+      "step": 5731
+    },
+    {
+      "epoch": 1.0960894923032796,
+      "grad_norm": 4.849170432964741,
+      "learning_rate": 1e-06,
+      "loss": 0.3099,
+      "step": 5732
+    },
+    {
+      "epoch": 1.0962807151735348,
+      "grad_norm": 2.9231107821800837,
+      "learning_rate": 1e-06,
+      "loss": 0.2792,
+      "step": 5733
+    },
+    {
+      "epoch": 1.09647193804379,
+      "grad_norm": 2.369572511257758,
+      "learning_rate": 1e-06,
+      "loss": 0.1609,
+      "step": 5734
+    },
+    {
+      "epoch": 1.0966631609140454,
+      "grad_norm": 3.4023095533226924,
+      "learning_rate": 1e-06,
+      "loss": 0.0759,
+      "step": 5735
+    },
+    {
+      "epoch": 1.0968543837843006,
+      "grad_norm": 3.101366781598311,
+      "learning_rate": 1e-06,
+      "loss": 0.0713,
+      "step": 5736
+    },
+    {
+      "epoch": 1.097045606654556,
+      "grad_norm": 1.969938373666781,
+      "learning_rate": 1e-06,
+      "loss": 0.0485,
+      "step": 5737
+    },
+    {
+      "epoch": 1.0972368295248112,
+      "grad_norm": 2.9526253736784893,
+      "learning_rate": 1e-06,
+      "loss": 0.1186,
+      "step": 5738
+    },
+    {
+      "epoch": 1.0974280523950664,
+      "grad_norm": 3.9242603601295967,
+      "learning_rate": 1e-06,
+      "loss": 0.0924,
+      "step": 5739
+    },
+    {
+      "epoch": 1.0976192752653218,
+      "grad_norm": 2.808532374617334,
+      "learning_rate": 1e-06,
+      "loss": 0.0977,
+      "step": 5740
+    },
+    {
+      "epoch": 1.097810498135577,
+      "grad_norm": 3.362714518117601,
+      "learning_rate": 1e-06,
+      "loss": 0.106,
+      "step": 5741
+    },
+    {
+      "epoch": 1.0980017210058324,
+      "grad_norm": 3.369154778049983,
+      "learning_rate": 1e-06,
+      "loss": 0.0754,
+      "step": 5742
+    },
+    {
+      "epoch": 1.0981929438760876,
+      "grad_norm": 3.4922940026754685,
+      "learning_rate": 1e-06,
+      "loss": 0.2642,
+      "step": 5743
+    },
+    {
+      "epoch": 1.0983841667463428,
+      "grad_norm": 5.858411704670133,
+      "learning_rate": 1e-06,
+      "loss": 0.5864,
+      "step": 5744
+    },
+    {
+      "epoch": 1.0985753896165982,
+      "grad_norm": 2.9665398652770256,
+      "learning_rate": 1e-06,
+      "loss": 0.1252,
+      "step": 5745
+    },
+    {
+      "epoch": 1.0987666124868534,
+      "grad_norm": 2.2368223218958283,
+      "learning_rate": 1e-06,
+      "loss": 0.132,
+      "step": 5746
+    },
+    {
+      "epoch": 1.0989578353571088,
+      "grad_norm": 3.139940025825922,
+      "learning_rate": 1e-06,
+      "loss": 0.1261,
+      "step": 5747
+    },
+    {
+      "epoch": 1.099149058227364,
+      "grad_norm": 2.516494316447855,
+      "learning_rate": 1e-06,
+      "loss": 0.0685,
+      "step": 5748
+    },
+    {
+      "epoch": 1.0993402810976192,
+      "grad_norm": 2.2156247976972865,
+      "learning_rate": 1e-06,
+      "loss": 0.1416,
+      "step": 5749
+    },
+    {
+      "epoch": 1.0995315039678746,
+      "grad_norm": 4.4093169232172595,
+      "learning_rate": 1e-06,
+      "loss": 0.0685,
+      "step": 5750
+    },
+    {
+      "epoch": 1.0997227268381298,
+      "grad_norm": 2.9074356777853203,
+      "learning_rate": 1e-06,
+      "loss": 0.2433,
+      "step": 5751
+    },
+    {
+      "epoch": 1.0999139497083852,
+      "grad_norm": 8.125975095484936,
+      "learning_rate": 1e-06,
+      "loss": 0.2018,
+      "step": 5752
+    },
+    {
+      "epoch": 1.1001051725786404,
+      "grad_norm": 1.5381566979010863,
+      "learning_rate": 1e-06,
+      "loss": 0.0286,
+      "step": 5753
+    },
+    {
+      "epoch": 1.1002963954488956,
+      "grad_norm": 8.07044815880052,
+      "learning_rate": 1e-06,
+      "loss": 0.082,
+      "step": 5754
+    },
+    {
+      "epoch": 1.100487618319151,
+      "grad_norm": 2.419113553741282,
+      "learning_rate": 1e-06,
+      "loss": 0.132,
+      "step": 5755
+    },
+    {
+      "epoch": 1.1006788411894062,
+      "grad_norm": 5.83558386123144,
+      "learning_rate": 1e-06,
+      "loss": 0.3841,
+      "step": 5756
+    },
+    {
+      "epoch": 1.1008700640596616,
+      "grad_norm": 4.933624871549209,
+      "learning_rate": 1e-06,
+      "loss": 0.4327,
+      "step": 5757
+    },
+    {
+      "epoch": 1.1010612869299168,
+      "grad_norm": 3.4096545082994547,
+      "learning_rate": 1e-06,
+      "loss": 0.2285,
+      "step": 5758
+    },
+    {
+      "epoch": 1.1012525098001722,
+      "grad_norm": 5.79894653487223,
+      "learning_rate": 1e-06,
+      "loss": 0.1102,
+      "step": 5759
+    },
+    {
+      "epoch": 1.1014437326704274,
+      "grad_norm": 4.059744971773454,
+      "learning_rate": 1e-06,
+      "loss": 0.2231,
+      "step": 5760
+    },
+    {
+      "epoch": 1.1016349555406826,
+      "grad_norm": 2.8997401680137114,
+      "learning_rate": 1e-06,
+      "loss": 0.1029,
+      "step": 5761
+    },
+    {
+      "epoch": 1.101826178410938,
+      "grad_norm": 3.5628297887375457,
+      "learning_rate": 1e-06,
+      "loss": 0.0858,
+      "step": 5762
+    },
+    {
+      "epoch": 1.1020174012811932,
+      "grad_norm": 3.4734505087282406,
+      "learning_rate": 1e-06,
+      "loss": 0.0874,
+      "step": 5763
+    },
+    {
+      "epoch": 1.1022086241514486,
+      "grad_norm": 3.489681291865896,
+      "learning_rate": 1e-06,
+      "loss": 0.2648,
+      "step": 5764
+    },
+    {
+      "epoch": 1.1023998470217038,
+      "grad_norm": 4.296008213355029,
+      "learning_rate": 1e-06,
+      "loss": 0.0966,
+      "step": 5765
+    },
+    {
+      "epoch": 1.102591069891959,
+      "grad_norm": 2.1785022425330287,
+      "learning_rate": 1e-06,
+      "loss": 0.0659,
+      "step": 5766
+    },
+    {
+      "epoch": 1.1027822927622144,
+      "grad_norm": 5.246969165604529,
+      "learning_rate": 1e-06,
+      "loss": 0.1545,
+      "step": 5767
+    },
+    {
+      "epoch": 1.1029735156324696,
+      "grad_norm": 7.38240507101638,
+      "learning_rate": 1e-06,
+      "loss": 0.2674,
+      "step": 5768
+    },
+    {
+      "epoch": 1.103164738502725,
+      "grad_norm": 4.203097318536867,
+      "learning_rate": 1e-06,
+      "loss": 0.4866,
+      "step": 5769
+    },
+    {
+      "epoch": 1.1033559613729802,
+      "grad_norm": 2.2941646024521716,
+      "learning_rate": 1e-06,
+      "loss": 0.112,
+      "step": 5770
+    },
+    {
+      "epoch": 1.1035471842432354,
+      "grad_norm": 2.424706360647361,
+      "learning_rate": 1e-06,
+      "loss": 0.3012,
+      "step": 5771
+    },
+    {
+      "epoch": 1.1037384071134908,
+      "grad_norm": 2.6857182118639322,
+      "learning_rate": 1e-06,
+      "loss": 0.0913,
+      "step": 5772
+    },
+    {
+      "epoch": 1.103929629983746,
+      "grad_norm": 1.9839365317052653,
+      "learning_rate": 1e-06,
+      "loss": 0.0704,
+      "step": 5773
+    },
+    {
+      "epoch": 1.1041208528540014,
+      "grad_norm": 2.7433203289083745,
+      "learning_rate": 1e-06,
+      "loss": 0.1582,
+      "step": 5774
+    },
+    {
+      "epoch": 1.1043120757242566,
+      "grad_norm": 5.387270560988193,
+      "learning_rate": 1e-06,
+      "loss": 0.1358,
+      "step": 5775
+    },
+    {
+      "epoch": 1.104503298594512,
+      "grad_norm": 3.7420354824208677,
+      "learning_rate": 1e-06,
+      "loss": 0.2347,
+      "step": 5776
+    },
+    {
+      "epoch": 1.1046945214647672,
+      "grad_norm": 1.6355030579638954,
+      "learning_rate": 1e-06,
+      "loss": 0.0207,
+      "step": 5777
+    },
+    {
+      "epoch": 1.1048857443350224,
+      "grad_norm": 4.1726811805184605,
+      "learning_rate": 1e-06,
+      "loss": 0.0781,
+      "step": 5778
+    },
+    {
+      "epoch": 1.1050769672052778,
+      "grad_norm": 5.442438097607456,
+      "learning_rate": 1e-06,
+      "loss": 0.1394,
+      "step": 5779
+    },
+    {
+      "epoch": 1.105268190075533,
+      "grad_norm": 4.480800360303557,
+      "learning_rate": 1e-06,
+      "loss": 0.2144,
+      "step": 5780
+    },
+    {
+      "epoch": 1.1054594129457884,
+      "grad_norm": 5.299195448070451,
+      "learning_rate": 1e-06,
+      "loss": 0.3611,
+      "step": 5781
+    },
+    {
+      "epoch": 1.1056506358160436,
+      "grad_norm": 4.158923214423031,
+      "learning_rate": 1e-06,
+      "loss": 0.2073,
+      "step": 5782
+    },
+    {
+      "epoch": 1.1058418586862988,
+      "grad_norm": 3.8998814051276045,
+      "learning_rate": 1e-06,
+      "loss": 0.3035,
+      "step": 5783
+    },
+    {
+      "epoch": 1.1060330815565542,
+      "grad_norm": 3.4855030098523017,
+      "learning_rate": 1e-06,
+      "loss": 0.1274,
+      "step": 5784
+    },
+    {
+      "epoch": 1.1062243044268094,
+      "grad_norm": 3.9857360908587665,
+      "learning_rate": 1e-06,
+      "loss": 0.2577,
+      "step": 5785
+    },
+    {
+      "epoch": 1.1064155272970648,
+      "grad_norm": 4.00927564894808,
+      "learning_rate": 1e-06,
+      "loss": 0.2156,
+      "step": 5786
+    },
+    {
+      "epoch": 1.10660675016732,
+      "grad_norm": 3.8922968569981977,
+      "learning_rate": 1e-06,
+      "loss": 0.1755,
+      "step": 5787
+    },
+    {
+      "epoch": 1.1067979730375752,
+      "grad_norm": 3.0201429608593564,
+      "learning_rate": 1e-06,
+      "loss": 0.1193,
+      "step": 5788
+    },
+    {
+      "epoch": 1.1069891959078306,
+      "grad_norm": 4.566418780724894,
+      "learning_rate": 1e-06,
+      "loss": 0.1119,
+      "step": 5789
+    },
+    {
+      "epoch": 1.1071804187780858,
+      "grad_norm": 2.408671065236304,
+      "learning_rate": 1e-06,
+      "loss": 0.034,
+      "step": 5790
+    },
+    {
+      "epoch": 1.1073716416483412,
+      "grad_norm": 2.265402947429263,
+      "learning_rate": 1e-06,
+      "loss": 0.0285,
+      "step": 5791
+    },
+    {
+      "epoch": 1.1075628645185964,
+      "grad_norm": 1.9265363655885608,
+      "learning_rate": 1e-06,
+      "loss": 0.0236,
+      "step": 5792
+    },
+    {
+      "epoch": 1.1077540873888516,
+      "grad_norm": 11.70543253702768,
+      "learning_rate": 1e-06,
+      "loss": 0.2803,
+      "step": 5793
+    },
+    {
+      "epoch": 1.107945310259107,
+      "grad_norm": 2.9379716859703526,
+      "learning_rate": 1e-06,
+      "loss": 0.1085,
+      "step": 5794
+    },
+    {
+      "epoch": 1.1081365331293622,
+      "grad_norm": 4.1790421887943,
+      "learning_rate": 1e-06,
+      "loss": 0.4274,
+      "step": 5795
+    },
+    {
+      "epoch": 1.1083277559996176,
+      "grad_norm": 4.517828064043581,
+      "learning_rate": 1e-06,
+      "loss": 0.3119,
+      "step": 5796
+    },
+    {
+      "epoch": 1.1085189788698728,
+      "grad_norm": 2.9451706748429842,
+      "learning_rate": 1e-06,
+      "loss": 0.1681,
+      "step": 5797
+    },
+    {
+      "epoch": 1.108710201740128,
+      "grad_norm": 2.557689151574433,
+      "learning_rate": 1e-06,
+      "loss": 0.1734,
+      "step": 5798
+    },
+    {
+      "epoch": 1.1089014246103834,
+      "grad_norm": 3.4576614064173814,
+      "learning_rate": 1e-06,
+      "loss": 0.2532,
+      "step": 5799
+    },
+    {
+      "epoch": 1.1090926474806386,
+      "grad_norm": 3.3898217440750487,
+      "learning_rate": 1e-06,
+      "loss": 0.1212,
+      "step": 5800
+    },
+    {
+      "epoch": 1.109283870350894,
+      "grad_norm": 2.9939722858642077,
+      "learning_rate": 1e-06,
+      "loss": 0.1635,
+      "step": 5801
+    },
+    {
+      "epoch": 1.1094750932211492,
+      "grad_norm": 4.27524942484685,
+      "learning_rate": 1e-06,
+      "loss": 0.1861,
+      "step": 5802
+    },
+    {
+      "epoch": 1.1096663160914044,
+      "grad_norm": 2.6395375196359527,
+      "learning_rate": 1e-06,
+      "loss": 0.0492,
+      "step": 5803
+    },
+    {
+      "epoch": 1.1098575389616598,
+      "grad_norm": 1.492632095651011,
+      "learning_rate": 1e-06,
+      "loss": 0.0154,
+      "step": 5804
+    },
+    {
+      "epoch": 1.110048761831915,
+      "grad_norm": 8.727692861270043,
+      "learning_rate": 1e-06,
+      "loss": 0.103,
+      "step": 5805
+    },
+    {
+      "epoch": 1.1102399847021704,
+      "grad_norm": 3.8327067457399258,
+      "learning_rate": 1e-06,
+      "loss": 0.2842,
+      "step": 5806
+    },
+    {
+      "epoch": 1.1104312075724256,
+      "grad_norm": 3.7258781268530656,
+      "learning_rate": 1e-06,
+      "loss": 0.0798,
+      "step": 5807
+    },
+    {
+      "epoch": 1.1106224304426808,
+      "grad_norm": 2.648565598712324,
+      "learning_rate": 1e-06,
+      "loss": 0.1735,
+      "step": 5808
+    },
+    {
+      "epoch": 1.1108136533129362,
+      "grad_norm": 2.9920061420030843,
+      "learning_rate": 1e-06,
+      "loss": 0.0958,
+      "step": 5809
+    },
+    {
+      "epoch": 1.1110048761831914,
+      "grad_norm": 4.512764523009895,
+      "learning_rate": 1e-06,
+      "loss": 0.5014,
+      "step": 5810
+    },
+    {
+      "epoch": 1.1111960990534469,
+      "grad_norm": 2.632471079436548,
+      "learning_rate": 1e-06,
+      "loss": 0.0878,
+      "step": 5811
+    },
+    {
+      "epoch": 1.111387321923702,
+      "grad_norm": 3.4359260423581133,
+      "learning_rate": 1e-06,
+      "loss": 0.0934,
+      "step": 5812
+    },
+    {
+      "epoch": 1.1115785447939575,
+      "grad_norm": 1.29661948203179,
+      "learning_rate": 1e-06,
+      "loss": 0.0165,
+      "step": 5813
+    },
+    {
+      "epoch": 1.1117697676642126,
+      "grad_norm": 4.9698064868487215,
+      "learning_rate": 1e-06,
+      "loss": 0.4,
+      "step": 5814
+    },
+    {
+      "epoch": 1.1119609905344678,
+      "grad_norm": 3.084130484832499,
+      "learning_rate": 1e-06,
+      "loss": 0.1317,
+      "step": 5815
+    },
+    {
+      "epoch": 1.1121522134047233,
+      "grad_norm": 2.941652163576967,
+      "learning_rate": 1e-06,
+      "loss": 0.0473,
+      "step": 5816
+    },
+    {
+      "epoch": 1.1123434362749784,
+      "grad_norm": 3.25303478867099,
+      "learning_rate": 1e-06,
+      "loss": 0.0469,
+      "step": 5817
+    },
+    {
+      "epoch": 1.1125346591452339,
+      "grad_norm": 9.902705476011539,
+      "learning_rate": 1e-06,
+      "loss": 0.4418,
+      "step": 5818
+    },
+    {
+      "epoch": 1.112725882015489,
+      "grad_norm": 4.240302072596965,
+      "learning_rate": 1e-06,
+      "loss": 0.4366,
+      "step": 5819
+    },
+    {
+      "epoch": 1.1129171048857442,
+      "grad_norm": 2.1809542130216477,
+      "learning_rate": 1e-06,
+      "loss": 0.0898,
+      "step": 5820
+    },
+    {
+      "epoch": 1.1131083277559997,
+      "grad_norm": 2.432838579943353,
+      "learning_rate": 1e-06,
+      "loss": 0.1213,
+      "step": 5821
+    },
+    {
+      "epoch": 1.1132995506262549,
+      "grad_norm": 2.0828513414405676,
+      "learning_rate": 1e-06,
+      "loss": 0.1354,
+      "step": 5822
+    },
+    {
+      "epoch": 1.1134907734965103,
+      "grad_norm": 3.2808552822483104,
+      "learning_rate": 1e-06,
+      "loss": 0.2359,
+      "step": 5823
+    },
+    {
+      "epoch": 1.1136819963667655,
+      "grad_norm": 3.338856762057795,
+      "learning_rate": 1e-06,
+      "loss": 0.113,
+      "step": 5824
+    },
+    {
+      "epoch": 1.1138732192370207,
+      "grad_norm": 2.2105419222918985,
+      "learning_rate": 1e-06,
+      "loss": 0.1167,
+      "step": 5825
+    },
+    {
+      "epoch": 1.114064442107276,
+      "grad_norm": 1.674752726735869,
+      "learning_rate": 1e-06,
+      "loss": 0.0503,
+      "step": 5826
+    },
+    {
+      "epoch": 1.1142556649775313,
+      "grad_norm": 5.481025216203105,
+      "learning_rate": 1e-06,
+      "loss": 0.1428,
+      "step": 5827
+    },
+    {
+      "epoch": 1.1144468878477867,
+      "grad_norm": 3.226370373933365,
+      "learning_rate": 1e-06,
+      "loss": 0.1205,
+      "step": 5828
+    },
+    {
+      "epoch": 1.1146381107180419,
+      "grad_norm": 2.9130833184911125,
+      "learning_rate": 1e-06,
+      "loss": 0.0319,
+      "step": 5829
+    },
+    {
+      "epoch": 1.1148293335882973,
+      "grad_norm": 5.521756921443728,
+      "learning_rate": 1e-06,
+      "loss": 0.0694,
+      "step": 5830
+    },
+    {
+      "epoch": 1.1150205564585525,
+      "grad_norm": 8.073653675195152,
+      "learning_rate": 1e-06,
+      "loss": 0.6017,
+      "step": 5831
+    },
+    {
+      "epoch": 1.1152117793288077,
+      "grad_norm": 2.90192296300914,
+      "learning_rate": 1e-06,
+      "loss": 0.307,
+      "step": 5832
+    },
+    {
+      "epoch": 1.115403002199063,
+      "grad_norm": 3.6738394169732764,
+      "learning_rate": 1e-06,
+      "loss": 0.1847,
+      "step": 5833
+    },
+    {
+      "epoch": 1.1155942250693183,
+      "grad_norm": 4.310163887961671,
+      "learning_rate": 1e-06,
+      "loss": 0.3578,
+      "step": 5834
+    },
+    {
+      "epoch": 1.1157854479395737,
+      "grad_norm": 4.698564009744053,
+      "learning_rate": 1e-06,
+      "loss": 0.3206,
+      "step": 5835
+    },
+    {
+      "epoch": 1.1159766708098289,
+      "grad_norm": 3.2999854578795946,
+      "learning_rate": 1e-06,
+      "loss": 0.0769,
+      "step": 5836
+    },
+    {
+      "epoch": 1.116167893680084,
+      "grad_norm": 3.340994011077491,
+      "learning_rate": 1e-06,
+      "loss": 0.2069,
+      "step": 5837
+    },
+    {
+      "epoch": 1.1163591165503395,
+      "grad_norm": 4.194290576551863,
+      "learning_rate": 1e-06,
+      "loss": 0.1386,
+      "step": 5838
+    },
+    {
+      "epoch": 1.1165503394205947,
+      "grad_norm": 3.002227750786128,
+      "learning_rate": 1e-06,
+      "loss": 0.1123,
+      "step": 5839
+    },
+    {
+      "epoch": 1.11674156229085,
+      "grad_norm": 6.987054844087536,
+      "learning_rate": 1e-06,
+      "loss": 0.3047,
+      "step": 5840
+    },
+    {
+      "epoch": 1.1169327851611053,
+      "grad_norm": 2.536716539424853,
+      "learning_rate": 1e-06,
+      "loss": 0.0378,
+      "step": 5841
+    },
+    {
+      "epoch": 1.1171240080313605,
+      "grad_norm": 7.072186064779636,
+      "learning_rate": 1e-06,
+      "loss": 0.0685,
+      "step": 5842
+    },
+    {
+      "epoch": 1.1173152309016159,
+      "grad_norm": 7.864202453956562,
+      "learning_rate": 1e-06,
+      "loss": 0.2688,
+      "step": 5843
+    },
+    {
+      "epoch": 1.117506453771871,
+      "grad_norm": 4.861701629132838,
+      "learning_rate": 1e-06,
+      "loss": 0.3948,
+      "step": 5844
+    },
+    {
+      "epoch": 1.1176976766421265,
+      "grad_norm": 4.950715354323176,
+      "learning_rate": 1e-06,
+      "loss": 0.2667,
+      "step": 5845
+    },
+    {
+      "epoch": 1.1178888995123817,
+      "grad_norm": 2.737553433110174,
+      "learning_rate": 1e-06,
+      "loss": 0.1687,
+      "step": 5846
+    },
+    {
+      "epoch": 1.1180801223826369,
+      "grad_norm": 4.034289729534329,
+      "learning_rate": 1e-06,
+      "loss": 0.165,
+      "step": 5847
+    },
+    {
+      "epoch": 1.1182713452528923,
+      "grad_norm": 3.5916999608832567,
+      "learning_rate": 1e-06,
+      "loss": 0.2742,
+      "step": 5848
+    },
+    {
+      "epoch": 1.1184625681231475,
+      "grad_norm": 3.6083832406226,
+      "learning_rate": 1e-06,
+      "loss": 0.2094,
+      "step": 5849
+    },
+    {
+      "epoch": 1.118653790993403,
+      "grad_norm": 3.443623585120199,
+      "learning_rate": 1e-06,
+      "loss": 0.0923,
+      "step": 5850
+    },
+    {
+      "epoch": 1.118845013863658,
+      "grad_norm": 2.596701539784415,
+      "learning_rate": 1e-06,
+      "loss": 0.0475,
+      "step": 5851
+    },
+    {
+      "epoch": 1.1190362367339133,
+      "grad_norm": 2.7359067277904314,
+      "learning_rate": 1e-06,
+      "loss": 0.0569,
+      "step": 5852
+    },
+    {
+      "epoch": 1.1192274596041687,
+      "grad_norm": 3.6879670445424075,
+      "learning_rate": 1e-06,
+      "loss": 0.0547,
+      "step": 5853
+    },
+    {
+      "epoch": 1.1194186824744239,
+      "grad_norm": 2.4189513248373506,
+      "learning_rate": 1e-06,
+      "loss": 0.0438,
+      "step": 5854
+    },
+    {
+      "epoch": 1.1196099053446793,
+      "grad_norm": 4.263935131660288,
+      "learning_rate": 1e-06,
+      "loss": 0.1814,
+      "step": 5855
+    },
+    {
+      "epoch": 1.1198011282149345,
+      "grad_norm": 6.690328694980165,
+      "learning_rate": 1e-06,
+      "loss": 0.348,
+      "step": 5856
+    },
+    {
+      "epoch": 1.1199923510851897,
+      "grad_norm": 3.91490082129187,
+      "learning_rate": 1e-06,
+      "loss": 0.2978,
+      "step": 5857
+    },
+    {
+      "epoch": 1.120183573955445,
+      "grad_norm": 2.669853918386945,
+      "learning_rate": 1e-06,
+      "loss": 0.136,
+      "step": 5858
+    },
+    {
+      "epoch": 1.1203747968257003,
+      "grad_norm": 2.448674921868119,
+      "learning_rate": 1e-06,
+      "loss": 0.0795,
+      "step": 5859
+    },
+    {
+      "epoch": 1.1205660196959557,
+      "grad_norm": 2.6149910435422905,
+      "learning_rate": 1e-06,
+      "loss": 0.097,
+      "step": 5860
+    },
+    {
+      "epoch": 1.120757242566211,
+      "grad_norm": 3.109555186509282,
+      "learning_rate": 1e-06,
+      "loss": 0.0767,
+      "step": 5861
+    },
+    {
+      "epoch": 1.120948465436466,
+      "grad_norm": 4.5013235053215785,
+      "learning_rate": 1e-06,
+      "loss": 0.0805,
+      "step": 5862
+    },
+    {
+      "epoch": 1.1211396883067215,
+      "grad_norm": 3.4950845806927604,
+      "learning_rate": 1e-06,
+      "loss": 0.2048,
+      "step": 5863
+    },
+    {
+      "epoch": 1.1213309111769767,
+      "grad_norm": 1.8430341850459326,
+      "learning_rate": 1e-06,
+      "loss": 0.0465,
+      "step": 5864
+    },
+    {
+      "epoch": 1.121522134047232,
+      "grad_norm": 2.4222831136105683,
+      "learning_rate": 1e-06,
+      "loss": 0.0466,
+      "step": 5865
+    },
+    {
+      "epoch": 1.1217133569174873,
+      "grad_norm": 2.9834169777451733,
+      "learning_rate": 1e-06,
+      "loss": 0.0796,
+      "step": 5866
+    },
+    {
+      "epoch": 1.1219045797877427,
+      "grad_norm": 3.7168866786641814,
+      "learning_rate": 1e-06,
+      "loss": 0.0983,
+      "step": 5867
+    },
+    {
+      "epoch": 1.122095802657998,
+      "grad_norm": 4.991428657848044,
+      "learning_rate": 1e-06,
+      "loss": 0.4817,
+      "step": 5868
+    },
+    {
+      "epoch": 1.122287025528253,
+      "grad_norm": 4.360259154430941,
+      "learning_rate": 1e-06,
+      "loss": 0.311,
+      "step": 5869
+    },
+    {
+      "epoch": 1.1224782483985085,
+      "grad_norm": 4.807855916687292,
+      "learning_rate": 1e-06,
+      "loss": 0.2376,
+      "step": 5870
+    },
+    {
+      "epoch": 1.1226694712687637,
+      "grad_norm": 4.17216347829272,
+      "learning_rate": 1e-06,
+      "loss": 0.379,
+      "step": 5871
+    },
+    {
+      "epoch": 1.1228606941390191,
+      "grad_norm": 2.442940923900129,
+      "learning_rate": 1e-06,
+      "loss": 0.1114,
+      "step": 5872
+    },
+    {
+      "epoch": 1.1230519170092743,
+      "grad_norm": 1.706461586601178,
+      "learning_rate": 1e-06,
+      "loss": 0.0738,
+      "step": 5873
+    },
+    {
+      "epoch": 1.1232431398795295,
+      "grad_norm": 4.041573959283608,
+      "learning_rate": 1e-06,
+      "loss": 0.2904,
+      "step": 5874
+    },
+    {
+      "epoch": 1.123434362749785,
+      "grad_norm": 2.4877707827611077,
+      "learning_rate": 1e-06,
+      "loss": 0.0994,
+      "step": 5875
+    },
+    {
+      "epoch": 1.1236255856200401,
+      "grad_norm": 3.886199395067141,
+      "learning_rate": 1e-06,
+      "loss": 0.1011,
+      "step": 5876
+    },
+    {
+      "epoch": 1.1238168084902955,
+      "grad_norm": 3.9148048412648144,
+      "learning_rate": 1e-06,
+      "loss": 0.1117,
+      "step": 5877
+    },
+    {
+      "epoch": 1.1240080313605507,
+      "grad_norm": 3.3483108361265184,
+      "learning_rate": 1e-06,
+      "loss": 0.0448,
+      "step": 5878
+    },
+    {
+      "epoch": 1.124199254230806,
+      "grad_norm": 3.5091361286396014,
+      "learning_rate": 1e-06,
+      "loss": 0.0654,
+      "step": 5879
+    },
+    {
+      "epoch": 1.1243904771010613,
+      "grad_norm": 8.308116703659206,
+      "learning_rate": 1e-06,
+      "loss": 0.1549,
+      "step": 5880
+    },
+    {
+      "epoch": 1.1245816999713165,
+      "grad_norm": 5.226950045189101,
+      "learning_rate": 1e-06,
+      "loss": 0.4874,
+      "step": 5881
+    },
+    {
+      "epoch": 1.124772922841572,
+      "grad_norm": 2.1996134375031855,
+      "learning_rate": 1e-06,
+      "loss": 0.1917,
+      "step": 5882
+    },
+    {
+      "epoch": 1.1249641457118271,
+      "grad_norm": 3.107535887040664,
+      "learning_rate": 1e-06,
+      "loss": 0.2165,
+      "step": 5883
+    },
+    {
+      "epoch": 1.1251553685820825,
+      "grad_norm": 2.217978222360838,
+      "learning_rate": 1e-06,
+      "loss": 0.0597,
+      "step": 5884
+    },
+    {
+      "epoch": 1.1253465914523377,
+      "grad_norm": 3.7669182926852938,
+      "learning_rate": 1e-06,
+      "loss": 0.1761,
+      "step": 5885
+    },
+    {
+      "epoch": 1.125537814322593,
+      "grad_norm": 2.611996732724353,
+      "learning_rate": 1e-06,
+      "loss": 0.1252,
+      "step": 5886
+    },
+    {
+      "epoch": 1.1257290371928483,
+      "grad_norm": 1.9089286619658183,
+      "learning_rate": 1e-06,
+      "loss": 0.0905,
+      "step": 5887
+    },
+    {
+      "epoch": 1.1259202600631035,
+      "grad_norm": 3.0287155095245044,
+      "learning_rate": 1e-06,
+      "loss": 0.0301,
+      "step": 5888
+    },
+    {
+      "epoch": 1.126111482933359,
+      "grad_norm": 3.765190890920188,
+      "learning_rate": 1e-06,
+      "loss": 0.1712,
+      "step": 5889
+    },
+    {
+      "epoch": 1.1263027058036141,
+      "grad_norm": 2.9589481162691835,
+      "learning_rate": 1e-06,
+      "loss": 0.0843,
+      "step": 5890
+    },
+    {
+      "epoch": 1.1264939286738693,
+      "grad_norm": 2.7877773882464965,
+      "learning_rate": 1e-06,
+      "loss": 0.0439,
+      "step": 5891
+    },
+    {
+      "epoch": 1.1266851515441247,
+      "grad_norm": 5.066325496597666,
+      "learning_rate": 1e-06,
+      "loss": 0.228,
+      "step": 5892
+    },
+    {
+      "epoch": 1.12687637441438,
+      "grad_norm": 11.951679217097421,
+      "learning_rate": 1e-06,
+      "loss": 0.1362,
+      "step": 5893
+    },
+    {
+      "epoch": 1.1270675972846353,
+      "grad_norm": 4.29906726745665,
+      "learning_rate": 1e-06,
+      "loss": 0.5217,
+      "step": 5894
+    },
+    {
+      "epoch": 1.1272588201548905,
+      "grad_norm": 4.88825165805331,
+      "learning_rate": 1e-06,
+      "loss": 0.1767,
+      "step": 5895
+    },
+    {
+      "epoch": 1.1274500430251457,
+      "grad_norm": 1.7093017980882228,
+      "learning_rate": 1e-06,
+      "loss": 0.1664,
+      "step": 5896
+    },
+    {
+      "epoch": 1.1276412658954011,
+      "grad_norm": 4.633419929811206,
+      "learning_rate": 1e-06,
+      "loss": 0.3494,
+      "step": 5897
+    },
+    {
+      "epoch": 1.1278324887656563,
+      "grad_norm": 3.3347713388637796,
+      "learning_rate": 1e-06,
+      "loss": 0.2288,
+      "step": 5898
+    },
+    {
+      "epoch": 1.1280237116359118,
+      "grad_norm": 2.349482126191111,
+      "learning_rate": 1e-06,
+      "loss": 0.1477,
+      "step": 5899
+    },
+    {
+      "epoch": 1.128214934506167,
+      "grad_norm": 3.033133167607213,
+      "learning_rate": 1e-06,
+      "loss": 0.0782,
+      "step": 5900
+    },
+    {
+      "epoch": 1.1284061573764221,
+      "grad_norm": 2.9831067331307475,
+      "learning_rate": 1e-06,
+      "loss": 0.1197,
+      "step": 5901
+    },
+    {
+      "epoch": 1.1285973802466776,
+      "grad_norm": 3.418635677120816,
+      "learning_rate": 1e-06,
+      "loss": 0.1216,
+      "step": 5902
+    },
+    {
+      "epoch": 1.1287886031169327,
+      "grad_norm": 1.8191203523554345,
+      "learning_rate": 1e-06,
+      "loss": 0.0535,
+      "step": 5903
+    },
+    {
+      "epoch": 1.1289798259871882,
+      "grad_norm": 3.592240920629927,
+      "learning_rate": 1e-06,
+      "loss": 0.0774,
+      "step": 5904
+    },
+    {
+      "epoch": 1.1291710488574433,
+      "grad_norm": 6.475695138976123,
+      "learning_rate": 1e-06,
+      "loss": 0.1869,
+      "step": 5905
+    },
+    {
+      "epoch": 1.1293622717276985,
+      "grad_norm": 4.6044823493731055,
+      "learning_rate": 1e-06,
+      "loss": 0.3478,
+      "step": 5906
+    },
+    {
+      "epoch": 1.129553494597954,
+      "grad_norm": 3.7367033144412534,
+      "learning_rate": 1e-06,
+      "loss": 0.2026,
+      "step": 5907
+    },
+    {
+      "epoch": 1.1297447174682091,
+      "grad_norm": 3.8652447377008636,
+      "learning_rate": 1e-06,
+      "loss": 0.2068,
+      "step": 5908
+    },
+    {
+      "epoch": 1.1299359403384646,
+      "grad_norm": 2.785016541902736,
+      "learning_rate": 1e-06,
+      "loss": 0.1362,
+      "step": 5909
+    },
+    {
+      "epoch": 1.1301271632087198,
+      "grad_norm": 4.093140534904288,
+      "learning_rate": 1e-06,
+      "loss": 0.3379,
+      "step": 5910
+    },
+    {
+      "epoch": 1.130318386078975,
+      "grad_norm": 3.443964618842273,
+      "learning_rate": 1e-06,
+      "loss": 0.1713,
+      "step": 5911
+    },
+    {
+      "epoch": 1.1305096089492304,
+      "grad_norm": 3.40097799820865,
+      "learning_rate": 1e-06,
+      "loss": 0.1159,
+      "step": 5912
+    },
+    {
+      "epoch": 1.1307008318194856,
+      "grad_norm": 3.200563774514022,
+      "learning_rate": 1e-06,
+      "loss": 0.1202,
+      "step": 5913
+    },
+    {
+      "epoch": 1.130892054689741,
+      "grad_norm": 3.1149078676366573,
+      "learning_rate": 1e-06,
+      "loss": 0.0821,
+      "step": 5914
+    },
+    {
+      "epoch": 1.1310832775599962,
+      "grad_norm": 4.56433026952763,
+      "learning_rate": 1e-06,
+      "loss": 0.1621,
+      "step": 5915
+    },
+    {
+      "epoch": 1.1312745004302514,
+      "grad_norm": 4.396905608568833,
+      "learning_rate": 1e-06,
+      "loss": 0.1163,
+      "step": 5916
+    },
+    {
+      "epoch": 1.1314657233005068,
+      "grad_norm": 7.321192864534904,
+      "learning_rate": 1e-06,
+      "loss": 0.1535,
+      "step": 5917
+    },
+    {
+      "epoch": 1.131656946170762,
+      "grad_norm": 4.993716296868824,
+      "learning_rate": 1e-06,
+      "loss": 0.1324,
+      "step": 5918
+    },
+    {
+      "epoch": 1.1318481690410174,
+      "grad_norm": 3.5259704282129656,
+      "learning_rate": 1e-06,
+      "loss": 0.2881,
+      "step": 5919
+    },
+    {
+      "epoch": 1.1320393919112726,
+      "grad_norm": 5.962104333189246,
+      "learning_rate": 1e-06,
+      "loss": 0.6417,
+      "step": 5920
+    },
+    {
+      "epoch": 1.1322306147815278,
+      "grad_norm": 3.5328689509946933,
+      "learning_rate": 1e-06,
+      "loss": 0.1784,
+      "step": 5921
+    },
+    {
+      "epoch": 1.1324218376517832,
+      "grad_norm": 3.222724756610865,
+      "learning_rate": 1e-06,
+      "loss": 0.3,
+      "step": 5922
+    },
+    {
+      "epoch": 1.1326130605220384,
+      "grad_norm": 3.468165855366434,
+      "learning_rate": 1e-06,
+      "loss": 0.2297,
+      "step": 5923
+    },
+    {
+      "epoch": 1.1328042833922938,
+      "grad_norm": 3.722487294705077,
+      "learning_rate": 1e-06,
+      "loss": 0.1786,
+      "step": 5924
+    },
+    {
+      "epoch": 1.132995506262549,
+      "grad_norm": 2.5444559415847174,
+      "learning_rate": 1e-06,
+      "loss": 0.1137,
+      "step": 5925
+    },
+    {
+      "epoch": 1.1331867291328042,
+      "grad_norm": 1.944997922901133,
+      "learning_rate": 1e-06,
+      "loss": 0.0392,
+      "step": 5926
+    },
+    {
+      "epoch": 1.1333779520030596,
+      "grad_norm": 4.313561129286203,
+      "learning_rate": 1e-06,
+      "loss": 0.0689,
+      "step": 5927
+    },
+    {
+      "epoch": 1.1335691748733148,
+      "grad_norm": 2.8514723672078093,
+      "learning_rate": 1e-06,
+      "loss": 0.0998,
+      "step": 5928
+    },
+    {
+      "epoch": 1.1337603977435702,
+      "grad_norm": 1.7488132267416001,
+      "learning_rate": 1e-06,
+      "loss": 0.0198,
+      "step": 5929
+    },
+    {
+      "epoch": 1.1339516206138254,
+      "grad_norm": 5.62195894900669,
+      "learning_rate": 1e-06,
+      "loss": 0.2165,
+      "step": 5930
+    },
+    {
+      "epoch": 1.1341428434840808,
+      "grad_norm": 4.7722434190825345,
+      "learning_rate": 1e-06,
+      "loss": 0.2747,
+      "step": 5931
+    },
+    {
+      "epoch": 1.134334066354336,
+      "grad_norm": 5.88728185034904,
+      "learning_rate": 1e-06,
+      "loss": 0.3255,
+      "step": 5932
+    },
+    {
+      "epoch": 1.1345252892245912,
+      "grad_norm": 2.4502348183966376,
+      "learning_rate": 1e-06,
+      "loss": 0.1303,
+      "step": 5933
+    },
+    {
+      "epoch": 1.1347165120948466,
+      "grad_norm": 2.7019857310067286,
+      "learning_rate": 1e-06,
+      "loss": 0.1178,
+      "step": 5934
+    },
+    {
+      "epoch": 1.1349077349651018,
+      "grad_norm": 5.626562622036412,
+      "learning_rate": 1e-06,
+      "loss": 0.48,
+      "step": 5935
+    },
+    {
+      "epoch": 1.1350989578353572,
+      "grad_norm": 2.429984022488284,
+      "learning_rate": 1e-06,
+      "loss": 0.0978,
+      "step": 5936
+    },
+    {
+      "epoch": 1.1352901807056124,
+      "grad_norm": 2.8660994707882415,
+      "learning_rate": 1e-06,
+      "loss": 0.2096,
+      "step": 5937
+    },
+    {
+      "epoch": 1.1354814035758678,
+      "grad_norm": 2.116759084451144,
+      "learning_rate": 1e-06,
+      "loss": 0.1153,
+      "step": 5938
+    },
+    {
+      "epoch": 1.135672626446123,
+      "grad_norm": 1.7983744460253726,
+      "learning_rate": 1e-06,
+      "loss": 0.0212,
+      "step": 5939
+    },
+    {
+      "epoch": 1.1358638493163782,
+      "grad_norm": 1.6238299338831037,
+      "learning_rate": 1e-06,
+      "loss": 0.0332,
+      "step": 5940
+    },
+    {
+      "epoch": 1.1360550721866336,
+      "grad_norm": 2.292722204623679,
+      "learning_rate": 1e-06,
+      "loss": 0.0269,
+      "step": 5941
+    },
+    {
+      "epoch": 1.1362462950568888,
+      "grad_norm": 6.602329837060835,
+      "learning_rate": 1e-06,
+      "loss": 0.0923,
+      "step": 5942
+    },
+    {
+      "epoch": 1.1364375179271442,
+      "grad_norm": 3.4964450766211277,
+      "learning_rate": 1e-06,
+      "loss": 0.2447,
+      "step": 5943
+    },
+    {
+      "epoch": 1.1366287407973994,
+      "grad_norm": 3.3864188407974014,
+      "learning_rate": 1e-06,
+      "loss": 0.1431,
+      "step": 5944
+    },
+    {
+      "epoch": 1.1368199636676546,
+      "grad_norm": 3.144688345260681,
+      "learning_rate": 1e-06,
+      "loss": 0.2147,
+      "step": 5945
+    },
+    {
+      "epoch": 1.13701118653791,
+      "grad_norm": 2.4881551521963754,
+      "learning_rate": 1e-06,
+      "loss": 0.1438,
+      "step": 5946
+    },
+    {
+      "epoch": 1.1372024094081652,
+      "grad_norm": 3.4533532528934856,
+      "learning_rate": 1e-06,
+      "loss": 0.3627,
+      "step": 5947
+    },
+    {
+      "epoch": 1.1373936322784206,
+      "grad_norm": 3.08052333596704,
+      "learning_rate": 1e-06,
+      "loss": 0.0961,
+      "step": 5948
+    },
+    {
+      "epoch": 1.1375848551486758,
+      "grad_norm": 1.4868249852225812,
+      "learning_rate": 1e-06,
+      "loss": 0.1013,
+      "step": 5949
+    },
+    {
+      "epoch": 1.137776078018931,
+      "grad_norm": 3.8873432826991734,
+      "learning_rate": 1e-06,
+      "loss": 0.1812,
+      "step": 5950
+    },
+    {
+      "epoch": 1.1379673008891864,
+      "grad_norm": 2.934090929022642,
+      "learning_rate": 1e-06,
+      "loss": 0.1203,
+      "step": 5951
+    },
+    {
+      "epoch": 1.1381585237594416,
+      "grad_norm": 2.003283547064715,
+      "learning_rate": 1e-06,
+      "loss": 0.028,
+      "step": 5952
+    },
+    {
+      "epoch": 1.138349746629697,
+      "grad_norm": 3.581752088145557,
+      "learning_rate": 1e-06,
+      "loss": 0.0407,
+      "step": 5953
+    },
+    {
+      "epoch": 1.1385409694999522,
+      "grad_norm": 2.113881475419178,
+      "learning_rate": 1e-06,
+      "loss": 0.0382,
+      "step": 5954
+    },
+    {
+      "epoch": 1.1387321923702074,
+      "grad_norm": 9.915350647697583,
+      "learning_rate": 1e-06,
+      "loss": 0.232,
+      "step": 5955
+    },
+    {
+      "epoch": 1.1389234152404628,
+      "grad_norm": 3.444426615132227,
+      "learning_rate": 1e-06,
+      "loss": 0.0842,
+      "step": 5956
+    },
+    {
+      "epoch": 1.139114638110718,
+      "grad_norm": 4.84809650939462,
+      "learning_rate": 1e-06,
+      "loss": 0.4183,
+      "step": 5957
+    },
+    {
+      "epoch": 1.1393058609809734,
+      "grad_norm": 2.624445084546548,
+      "learning_rate": 1e-06,
+      "loss": 0.206,
+      "step": 5958
+    },
+    {
+      "epoch": 1.1394970838512286,
+      "grad_norm": 2.4156874832422166,
+      "learning_rate": 1e-06,
+      "loss": 0.1971,
+      "step": 5959
+    },
+    {
+      "epoch": 1.1396883067214838,
+      "grad_norm": 3.133982756118708,
+      "learning_rate": 1e-06,
+      "loss": 0.1396,
+      "step": 5960
+    },
+    {
+      "epoch": 1.1398795295917392,
+      "grad_norm": 2.165908301732761,
+      "learning_rate": 1e-06,
+      "loss": 0.0911,
+      "step": 5961
+    },
+    {
+      "epoch": 1.1400707524619944,
+      "grad_norm": 3.217458225029436,
+      "learning_rate": 1e-06,
+      "loss": 0.1572,
+      "step": 5962
+    },
+    {
+      "epoch": 1.1402619753322498,
+      "grad_norm": 1.5623545769729201,
+      "learning_rate": 1e-06,
+      "loss": 0.0409,
+      "step": 5963
+    },
+    {
+      "epoch": 1.140453198202505,
+      "grad_norm": 4.147630516054327,
+      "learning_rate": 1e-06,
+      "loss": 0.1152,
+      "step": 5964
+    },
+    {
+      "epoch": 1.1406444210727602,
+      "grad_norm": 2.251579895587674,
+      "learning_rate": 1e-06,
+      "loss": 0.11,
+      "step": 5965
+    },
+    {
+      "epoch": 1.1408356439430156,
+      "grad_norm": 2.5816829290630987,
+      "learning_rate": 1e-06,
+      "loss": 0.0338,
+      "step": 5966
+    },
+    {
+      "epoch": 1.1410268668132708,
+      "grad_norm": 6.08094608374696,
+      "learning_rate": 1e-06,
+      "loss": 0.1107,
+      "step": 5967
+    },
+    {
+      "epoch": 1.1412180896835262,
+      "grad_norm": 3.8941559605961826,
+      "learning_rate": 1e-06,
+      "loss": 0.3986,
+      "step": 5968
+    },
+    {
+      "epoch": 1.1414093125537814,
+      "grad_norm": 4.21336416555141,
+      "learning_rate": 1e-06,
+      "loss": 0.535,
+      "step": 5969
+    },
+    {
+      "epoch": 1.1416005354240366,
+      "grad_norm": 4.858591488783844,
+      "learning_rate": 1e-06,
+      "loss": 0.5009,
+      "step": 5970
+    },
+    {
+      "epoch": 1.141791758294292,
+      "grad_norm": 1.985146680356214,
+      "learning_rate": 1e-06,
+      "loss": 0.1714,
+      "step": 5971
+    },
+    {
+      "epoch": 1.1419829811645472,
+      "grad_norm": 3.467288284934249,
+      "learning_rate": 1e-06,
+      "loss": 0.2576,
+      "step": 5972
+    },
+    {
+      "epoch": 1.1421742040348026,
+      "grad_norm": 6.354372348479056,
+      "learning_rate": 1e-06,
+      "loss": 0.477,
+      "step": 5973
+    },
+    {
+      "epoch": 1.1423654269050578,
+      "grad_norm": 3.1470605001922216,
+      "learning_rate": 1e-06,
+      "loss": 0.1948,
+      "step": 5974
+    },
+    {
+      "epoch": 1.142556649775313,
+      "grad_norm": 3.641133285207639,
+      "learning_rate": 1e-06,
+      "loss": 0.0435,
+      "step": 5975
+    },
+    {
+      "epoch": 1.1427478726455684,
+      "grad_norm": 4.286806830156015,
+      "learning_rate": 1e-06,
+      "loss": 0.1993,
+      "step": 5976
+    },
+    {
+      "epoch": 1.1429390955158236,
+      "grad_norm": 4.030704905007312,
+      "learning_rate": 1e-06,
+      "loss": 0.1342,
+      "step": 5977
+    },
+    {
+      "epoch": 1.143130318386079,
+      "grad_norm": 3.3972287889571176,
+      "learning_rate": 1e-06,
+      "loss": 0.0546,
+      "step": 5978
+    },
+    {
+      "epoch": 1.1433215412563342,
+      "grad_norm": 1.54837264185934,
+      "learning_rate": 1e-06,
+      "loss": 0.0179,
+      "step": 5979
+    },
+    {
+      "epoch": 1.1435127641265894,
+      "grad_norm": 8.717169943212513,
+      "learning_rate": 1e-06,
+      "loss": 0.1311,
+      "step": 5980
+    },
+    {
+      "epoch": 1.1437039869968448,
+      "grad_norm": 5.39430831901557,
+      "learning_rate": 1e-06,
+      "loss": 0.5897,
+      "step": 5981
+    },
+    {
+      "epoch": 1.1438952098671,
+      "grad_norm": 6.0016566214794755,
+      "learning_rate": 1e-06,
+      "loss": 0.5352,
+      "step": 5982
+    },
+    {
+      "epoch": 1.1440864327373554,
+      "grad_norm": 3.6361381157388006,
+      "learning_rate": 1e-06,
+      "loss": 0.2728,
+      "step": 5983
+    },
+    {
+      "epoch": 1.1442776556076106,
+      "grad_norm": 3.746913657096378,
+      "learning_rate": 1e-06,
+      "loss": 0.1152,
+      "step": 5984
+    },
+    {
+      "epoch": 1.144468878477866,
+      "grad_norm": 3.816151855486212,
+      "learning_rate": 1e-06,
+      "loss": 0.3375,
+      "step": 5985
+    },
+    {
+      "epoch": 1.1446601013481212,
+      "grad_norm": 1.8869648964370647,
+      "learning_rate": 1e-06,
+      "loss": 0.0482,
+      "step": 5986
+    },
+    {
+      "epoch": 1.1448513242183764,
+      "grad_norm": 2.8592669357141123,
+      "learning_rate": 1e-06,
+      "loss": 0.1321,
+      "step": 5987
+    },
+    {
+      "epoch": 1.1450425470886318,
+      "grad_norm": 2.7363549611683684,
+      "learning_rate": 1e-06,
+      "loss": 0.0986,
+      "step": 5988
+    },
+    {
+      "epoch": 1.145233769958887,
+      "grad_norm": 3.9186064333375623,
+      "learning_rate": 1e-06,
+      "loss": 0.1091,
+      "step": 5989
+    },
+    {
+      "epoch": 1.1454249928291425,
+      "grad_norm": 1.8844399917048105,
+      "learning_rate": 1e-06,
+      "loss": 0.104,
+      "step": 5990
+    },
+    {
+      "epoch": 1.1456162156993976,
+      "grad_norm": 4.936132120834191,
+      "learning_rate": 1e-06,
+      "loss": 0.043,
+      "step": 5991
+    },
+    {
+      "epoch": 1.1458074385696528,
+      "grad_norm": 5.328876520668703,
+      "learning_rate": 1e-06,
+      "loss": 0.0494,
+      "step": 5992
+    },
+    {
+      "epoch": 1.1459986614399083,
+      "grad_norm": 25.1107507790488,
+      "learning_rate": 1e-06,
+      "loss": 0.4191,
+      "step": 5993
+    },
+    {
+      "epoch": 1.1461898843101634,
+      "grad_norm": 4.536292276361516,
+      "learning_rate": 1e-06,
+      "loss": 0.2056,
+      "step": 5994
+    },
+    {
+      "epoch": 1.1463811071804189,
+      "grad_norm": 2.723880968520865,
+      "learning_rate": 1e-06,
+      "loss": 0.1087,
+      "step": 5995
+    },
+    {
+      "epoch": 1.146572330050674,
+      "grad_norm": 3.369395936884264,
+      "learning_rate": 1e-06,
+      "loss": 0.1501,
+      "step": 5996
+    },
+    {
+      "epoch": 1.1467635529209295,
+      "grad_norm": 1.5234083123956772,
+      "learning_rate": 1e-06,
+      "loss": 0.0192,
+      "step": 5997
+    },
+    {
+      "epoch": 1.1469547757911847,
+      "grad_norm": 4.356253432678884,
+      "learning_rate": 1e-06,
+      "loss": 0.3481,
+      "step": 5998
+    },
+    {
+      "epoch": 1.1471459986614398,
+      "grad_norm": 1.2849336564716127,
+      "learning_rate": 1e-06,
+      "loss": 0.0375,
+      "step": 5999
+    },
+    {
+      "epoch": 1.1473372215316953,
+      "grad_norm": 2.8461585272831798,
+      "learning_rate": 1e-06,
+      "loss": 0.0475,
+      "step": 6000
+    },
+    {
+      "epoch": 1.1473372215316953,
+      "eval_runtime": 733.3562,
+      "eval_samples_per_second": 2.092,
+      "eval_steps_per_second": 0.524,
+      "step": 6000
+    },
+    {
+      "epoch": 1.1475284444019505,
+      "grad_norm": 1.8117320966208408,
+      "learning_rate": 1e-06,
+      "loss": 0.0279,
+      "step": 6001
+    },
+    {
+      "epoch": 1.1477196672722059,
+      "grad_norm": 2.2457856870939303,
+      "learning_rate": 1e-06,
+      "loss": 0.0672,
+      "step": 6002
+    },
+    {
+      "epoch": 1.147910890142461,
+      "grad_norm": 2.882193480817042,
+      "learning_rate": 1e-06,
+      "loss": 0.1762,
+      "step": 6003
+    },
+    {
+      "epoch": 1.1481021130127163,
+      "grad_norm": 2.24309072345555,
+      "learning_rate": 1e-06,
+      "loss": 0.0642,
+      "step": 6004
+    },
+    {
+      "epoch": 1.1482933358829717,
+      "grad_norm": 13.935614077667324,
+      "learning_rate": 1e-06,
+      "loss": 0.1527,
+      "step": 6005
+    },
+    {
+      "epoch": 1.1484845587532269,
+      "grad_norm": 3.883022248956175,
+      "learning_rate": 1e-06,
+      "loss": 0.1376,
+      "step": 6006
+    },
+    {
+      "epoch": 1.1486757816234823,
+      "grad_norm": 3.7629749898086993,
+      "learning_rate": 1e-06,
+      "loss": 0.1096,
+      "step": 6007
+    },
+    {
+      "epoch": 1.1488670044937375,
+      "grad_norm": 3.637631605985782,
+      "learning_rate": 1e-06,
+      "loss": 0.1982,
+      "step": 6008
+    },
+    {
+      "epoch": 1.1490582273639927,
+      "grad_norm": 4.750444893330552,
+      "learning_rate": 1e-06,
+      "loss": 0.5968,
+      "step": 6009
+    },
+    {
+      "epoch": 1.149249450234248,
+      "grad_norm": 3.0490406653875595,
+      "learning_rate": 1e-06,
+      "loss": 0.1959,
+      "step": 6010
+    },
+    {
+      "epoch": 1.1494406731045033,
+      "grad_norm": 4.366147267170241,
+      "learning_rate": 1e-06,
+      "loss": 0.2912,
+      "step": 6011
+    },
+    {
+      "epoch": 1.1496318959747587,
+      "grad_norm": 2.1359268199371795,
+      "learning_rate": 1e-06,
+      "loss": 0.1617,
+      "step": 6012
+    },
+    {
+      "epoch": 1.1498231188450139,
+      "grad_norm": 2.4933995853978477,
+      "learning_rate": 1e-06,
+      "loss": 0.0599,
+      "step": 6013
+    },
+    {
+      "epoch": 1.150014341715269,
+      "grad_norm": 5.273821962258454,
+      "learning_rate": 1e-06,
+      "loss": 0.1522,
+      "step": 6014
+    },
+    {
+      "epoch": 1.1502055645855245,
+      "grad_norm": 3.8497858805378735,
+      "learning_rate": 1e-06,
+      "loss": 0.2162,
+      "step": 6015
+    },
+    {
+      "epoch": 1.1503967874557797,
+      "grad_norm": 2.900266990210031,
+      "learning_rate": 1e-06,
+      "loss": 0.0628,
+      "step": 6016
+    },
+    {
+      "epoch": 1.150588010326035,
+      "grad_norm": 4.993945843385761,
+      "learning_rate": 1e-06,
+      "loss": 0.1286,
+      "step": 6017
+    },
+    {
+      "epoch": 1.1507792331962903,
+      "grad_norm": 3.6838758083538203,
+      "learning_rate": 1e-06,
+      "loss": 0.2447,
+      "step": 6018
+    },
+    {
+      "epoch": 1.1509704560665455,
+      "grad_norm": 2.729632291797797,
+      "learning_rate": 1e-06,
+      "loss": 0.1516,
+      "step": 6019
+    },
+    {
+      "epoch": 1.1511616789368009,
+      "grad_norm": 2.5937895254777374,
+      "learning_rate": 1e-06,
+      "loss": 0.0824,
+      "step": 6020
+    },
+    {
+      "epoch": 1.151352901807056,
+      "grad_norm": 1.5971156467134904,
+      "learning_rate": 1e-06,
+      "loss": 0.0472,
+      "step": 6021
+    },
+    {
+      "epoch": 1.1515441246773115,
+      "grad_norm": 5.590267354614566,
+      "learning_rate": 1e-06,
+      "loss": 0.4349,
+      "step": 6022
+    },
+    {
+      "epoch": 1.1517353475475667,
+      "grad_norm": 3.5011451755072978,
+      "learning_rate": 1e-06,
+      "loss": 0.095,
+      "step": 6023
+    },
+    {
+      "epoch": 1.1519265704178219,
+      "grad_norm": 1.8978845866479446,
+      "learning_rate": 1e-06,
+      "loss": 0.0659,
+      "step": 6024
+    },
+    {
+      "epoch": 1.1521177932880773,
+      "grad_norm": 3.8191151995308545,
+      "learning_rate": 1e-06,
+      "loss": 0.0721,
+      "step": 6025
+    },
+    {
+      "epoch": 1.1523090161583325,
+      "grad_norm": 2.090824303585343,
+      "learning_rate": 1e-06,
+      "loss": 0.0674,
+      "step": 6026
+    },
+    {
+      "epoch": 1.152500239028588,
+      "grad_norm": 3.499106156789813,
+      "learning_rate": 1e-06,
+      "loss": 0.0889,
+      "step": 6027
+    },
+    {
+      "epoch": 1.152691461898843,
+      "grad_norm": 10.894721028510206,
+      "learning_rate": 1e-06,
+      "loss": 0.133,
+      "step": 6028
+    },
+    {
+      "epoch": 1.1528826847690983,
+      "grad_norm": 4.023053967611963,
+      "learning_rate": 1e-06,
+      "loss": 0.1927,
+      "step": 6029
+    },
+    {
+      "epoch": 1.1530739076393537,
+      "grad_norm": 3.97572794552625,
+      "learning_rate": 1e-06,
+      "loss": 0.1847,
+      "step": 6030
+    },
+    {
+      "epoch": 1.1532651305096089,
+      "grad_norm": 5.099842570249891,
+      "learning_rate": 1e-06,
+      "loss": 0.3231,
+      "step": 6031
+    },
+    {
+      "epoch": 1.1534563533798643,
+      "grad_norm": 3.9756698955609027,
+      "learning_rate": 1e-06,
+      "loss": 0.1166,
+      "step": 6032
+    },
+    {
+      "epoch": 1.1536475762501195,
+      "grad_norm": 3.014140499958688,
+      "learning_rate": 1e-06,
+      "loss": 0.1818,
+      "step": 6033
+    },
+    {
+      "epoch": 1.1538387991203747,
+      "grad_norm": 3.637175272181874,
+      "learning_rate": 1e-06,
+      "loss": 0.1418,
+      "step": 6034
+    },
+    {
+      "epoch": 1.15403002199063,
+      "grad_norm": 4.75154189130735,
+      "learning_rate": 1e-06,
+      "loss": 0.2032,
+      "step": 6035
+    },
+    {
+      "epoch": 1.1542212448608853,
+      "grad_norm": 3.514799843484448,
+      "learning_rate": 1e-06,
+      "loss": 0.0889,
+      "step": 6036
+    },
+    {
+      "epoch": 1.1544124677311407,
+      "grad_norm": 2.5761589183004348,
+      "learning_rate": 1e-06,
+      "loss": 0.0897,
+      "step": 6037
+    },
+    {
+      "epoch": 1.154603690601396,
+      "grad_norm": 3.6101194992786962,
+      "learning_rate": 1e-06,
+      "loss": 0.0698,
+      "step": 6038
+    },
+    {
+      "epoch": 1.154794913471651,
+      "grad_norm": 1.2019827238101273,
+      "learning_rate": 1e-06,
+      "loss": 0.0466,
+      "step": 6039
+    },
+    {
+      "epoch": 1.1549861363419065,
+      "grad_norm": 2.5098946267308673,
+      "learning_rate": 1e-06,
+      "loss": 0.0291,
+      "step": 6040
+    },
+    {
+      "epoch": 1.1551773592121617,
+      "grad_norm": 2.398817895800074,
+      "learning_rate": 1e-06,
+      "loss": 0.0474,
+      "step": 6041
+    },
+    {
+      "epoch": 1.155368582082417,
+      "grad_norm": 5.310253072049296,
+      "learning_rate": 1e-06,
+      "loss": 0.0821,
+      "step": 6042
+    },
+    {
+      "epoch": 1.1555598049526723,
+      "grad_norm": 2.8066920180848314,
+      "learning_rate": 1e-06,
+      "loss": 0.229,
+      "step": 6043
+    },
+    {
+      "epoch": 1.1557510278229277,
+      "grad_norm": 4.1501293575854215,
+      "learning_rate": 1e-06,
+      "loss": 0.346,
+      "step": 6044
+    },
+    {
+      "epoch": 1.155942250693183,
+      "grad_norm": 2.360131249899918,
+      "learning_rate": 1e-06,
+      "loss": 0.1327,
+      "step": 6045
+    },
+    {
+      "epoch": 1.156133473563438,
+      "grad_norm": 1.9030577398732555,
+      "learning_rate": 1e-06,
+      "loss": 0.1026,
+      "step": 6046
+    },
+    {
+      "epoch": 1.1563246964336935,
+      "grad_norm": 2.7627167146351987,
+      "learning_rate": 1e-06,
+      "loss": 0.1488,
+      "step": 6047
+    },
+    {
+      "epoch": 1.1565159193039487,
+      "grad_norm": 2.745612459073775,
+      "learning_rate": 1e-06,
+      "loss": 0.0719,
+      "step": 6048
+    },
+    {
+      "epoch": 1.1567071421742041,
+      "grad_norm": 2.504933067843818,
+      "learning_rate": 1e-06,
+      "loss": 0.0757,
+      "step": 6049
+    },
+    {
+      "epoch": 1.1568983650444593,
+      "grad_norm": 1.5273784038388076,
+      "learning_rate": 1e-06,
+      "loss": 0.0746,
+      "step": 6050
+    },
+    {
+      "epoch": 1.1570895879147147,
+      "grad_norm": 6.799576005619849,
+      "learning_rate": 1e-06,
+      "loss": 0.0743,
+      "step": 6051
+    },
+    {
+      "epoch": 1.15728081078497,
+      "grad_norm": 2.989543174624567,
+      "learning_rate": 1e-06,
+      "loss": 0.0359,
+      "step": 6052
+    },
+    {
+      "epoch": 1.157472033655225,
+      "grad_norm": 2.7794396155369188,
+      "learning_rate": 1e-06,
+      "loss": 0.0259,
+      "step": 6053
+    },
+    {
+      "epoch": 1.1576632565254805,
+      "grad_norm": 4.063678277144577,
+      "learning_rate": 1e-06,
+      "loss": 0.2538,
+      "step": 6054
+    },
+    {
+      "epoch": 1.1578544793957357,
+      "grad_norm": 6.6960850451320155,
+      "learning_rate": 1e-06,
+      "loss": 0.0766,
+      "step": 6055
+    },
+    {
+      "epoch": 1.1580457022659911,
+      "grad_norm": 3.253045635739395,
+      "learning_rate": 1e-06,
+      "loss": 0.2029,
+      "step": 6056
+    },
+    {
+      "epoch": 1.1582369251362463,
+      "grad_norm": 4.587524347474559,
+      "learning_rate": 1e-06,
+      "loss": 0.4593,
+      "step": 6057
+    },
+    {
+      "epoch": 1.1584281480065015,
+      "grad_norm": 3.6003826467817377,
+      "learning_rate": 1e-06,
+      "loss": 0.4927,
+      "step": 6058
+    },
+    {
+      "epoch": 1.158619370876757,
+      "grad_norm": 1.458030215051015,
+      "learning_rate": 1e-06,
+      "loss": 0.1267,
+      "step": 6059
+    },
+    {
+      "epoch": 1.1588105937470121,
+      "grad_norm": 3.5051691847604083,
+      "learning_rate": 1e-06,
+      "loss": 0.3105,
+      "step": 6060
+    },
+    {
+      "epoch": 1.1590018166172675,
+      "grad_norm": 2.965863722105759,
+      "learning_rate": 1e-06,
+      "loss": 0.2621,
+      "step": 6061
+    },
+    {
+      "epoch": 1.1591930394875227,
+      "grad_norm": 2.172650205733401,
+      "learning_rate": 1e-06,
+      "loss": 0.0546,
+      "step": 6062
+    },
+    {
+      "epoch": 1.159384262357778,
+      "grad_norm": 2.9460823813507115,
+      "learning_rate": 1e-06,
+      "loss": 0.0267,
+      "step": 6063
+    },
+    {
+      "epoch": 1.1595754852280333,
+      "grad_norm": 2.6870641133106123,
+      "learning_rate": 1e-06,
+      "loss": 0.1157,
+      "step": 6064
+    },
+    {
+      "epoch": 1.1597667080982885,
+      "grad_norm": 4.905169131039374,
+      "learning_rate": 1e-06,
+      "loss": 0.1574,
+      "step": 6065
+    },
+    {
+      "epoch": 1.159957930968544,
+      "grad_norm": 4.6327134840678275,
+      "learning_rate": 1e-06,
+      "loss": 0.0502,
+      "step": 6066
+    },
+    {
+      "epoch": 1.1601491538387991,
+      "grad_norm": 3.7958194222947332,
+      "learning_rate": 1e-06,
+      "loss": 0.0614,
+      "step": 6067
+    },
+    {
+      "epoch": 1.1603403767090543,
+      "grad_norm": 6.218089313011662,
+      "learning_rate": 1e-06,
+      "loss": 0.6075,
+      "step": 6068
+    },
+    {
+      "epoch": 1.1605315995793097,
+      "grad_norm": 5.430925801635133,
+      "learning_rate": 1e-06,
+      "loss": 0.614,
+      "step": 6069
+    },
+    {
+      "epoch": 1.160722822449565,
+      "grad_norm": 4.688883666225938,
+      "learning_rate": 1e-06,
+      "loss": 0.3064,
+      "step": 6070
+    },
+    {
+      "epoch": 1.1609140453198203,
+      "grad_norm": 2.5751003208802676,
+      "learning_rate": 1e-06,
+      "loss": 0.1138,
+      "step": 6071
+    },
+    {
+      "epoch": 1.1611052681900755,
+      "grad_norm": 3.828118522794265,
+      "learning_rate": 1e-06,
+      "loss": 0.3617,
+      "step": 6072
+    },
+    {
+      "epoch": 1.1612964910603307,
+      "grad_norm": 2.7108963405331843,
+      "learning_rate": 1e-06,
+      "loss": 0.0654,
+      "step": 6073
+    },
+    {
+      "epoch": 1.1614877139305861,
+      "grad_norm": 2.9404890395869785,
+      "learning_rate": 1e-06,
+      "loss": 0.0982,
+      "step": 6074
+    },
+    {
+      "epoch": 1.1616789368008413,
+      "grad_norm": 4.267930075232629,
+      "learning_rate": 1e-06,
+      "loss": 0.0837,
+      "step": 6075
+    },
+    {
+      "epoch": 1.1618701596710967,
+      "grad_norm": 4.69927313845292,
+      "learning_rate": 1e-06,
+      "loss": 0.1952,
+      "step": 6076
+    },
+    {
+      "epoch": 1.162061382541352,
+      "grad_norm": 2.3028634822201077,
+      "learning_rate": 1e-06,
+      "loss": 0.0742,
+      "step": 6077
+    },
+    {
+      "epoch": 1.1622526054116071,
+      "grad_norm": 3.0721562021101034,
+      "learning_rate": 1e-06,
+      "loss": 0.0411,
+      "step": 6078
+    },
+    {
+      "epoch": 1.1624438282818625,
+      "grad_norm": 0.8527075392396835,
+      "learning_rate": 1e-06,
+      "loss": 0.0073,
+      "step": 6079
+    },
+    {
+      "epoch": 1.1626350511521177,
+      "grad_norm": 4.201638338164226,
+      "learning_rate": 1e-06,
+      "loss": 0.2946,
+      "step": 6080
+    },
+    {
+      "epoch": 1.1628262740223732,
+      "grad_norm": 3.056380873298283,
+      "learning_rate": 1e-06,
+      "loss": 0.2119,
+      "step": 6081
+    },
+    {
+      "epoch": 1.1630174968926283,
+      "grad_norm": 3.873516537135779,
+      "learning_rate": 1e-06,
+      "loss": 0.4816,
+      "step": 6082
+    },
+    {
+      "epoch": 1.1632087197628835,
+      "grad_norm": 3.6083679115335996,
+      "learning_rate": 1e-06,
+      "loss": 0.1102,
+      "step": 6083
+    },
+    {
+      "epoch": 1.163399942633139,
+      "grad_norm": 3.8318460868873307,
+      "learning_rate": 1e-06,
+      "loss": 0.3264,
+      "step": 6084
+    },
+    {
+      "epoch": 1.1635911655033941,
+      "grad_norm": 6.8895810951926775,
+      "learning_rate": 1e-06,
+      "loss": 0.361,
+      "step": 6085
+    },
+    {
+      "epoch": 1.1637823883736496,
+      "grad_norm": 1.298981575756667,
+      "learning_rate": 1e-06,
+      "loss": 0.0313,
+      "step": 6086
+    },
+    {
+      "epoch": 1.1639736112439047,
+      "grad_norm": 1.634238975218113,
+      "learning_rate": 1e-06,
+      "loss": 0.0326,
+      "step": 6087
+    },
+    {
+      "epoch": 1.16416483411416,
+      "grad_norm": 3.914241337450047,
+      "learning_rate": 1e-06,
+      "loss": 0.207,
+      "step": 6088
+    },
+    {
+      "epoch": 1.1643560569844154,
+      "grad_norm": 3.7580481314589114,
+      "learning_rate": 1e-06,
+      "loss": 0.1803,
+      "step": 6089
+    },
+    {
+      "epoch": 1.1645472798546705,
+      "grad_norm": 3.1600332719524,
+      "learning_rate": 1e-06,
+      "loss": 0.135,
+      "step": 6090
+    },
+    {
+      "epoch": 1.164738502724926,
+      "grad_norm": 5.011081812665559,
+      "learning_rate": 1e-06,
+      "loss": 0.0846,
+      "step": 6091
+    },
+    {
+      "epoch": 1.1649297255951812,
+      "grad_norm": 3.9078907686896476,
+      "learning_rate": 1e-06,
+      "loss": 0.1068,
+      "step": 6092
+    },
+    {
+      "epoch": 1.1651209484654363,
+      "grad_norm": 4.551971829908792,
+      "learning_rate": 1e-06,
+      "loss": 0.2366,
+      "step": 6093
+    },
+    {
+      "epoch": 1.1653121713356918,
+      "grad_norm": 4.759697100835222,
+      "learning_rate": 1e-06,
+      "loss": 0.2825,
+      "step": 6094
+    },
+    {
+      "epoch": 1.165503394205947,
+      "grad_norm": 3.7071175600381046,
+      "learning_rate": 1e-06,
+      "loss": 0.2263,
+      "step": 6095
+    },
+    {
+      "epoch": 1.1656946170762024,
+      "grad_norm": 5.209684069640764,
+      "learning_rate": 1e-06,
+      "loss": 0.3862,
+      "step": 6096
+    },
+    {
+      "epoch": 1.1658858399464576,
+      "grad_norm": 1.977050901843044,
+      "learning_rate": 1e-06,
+      "loss": 0.0823,
+      "step": 6097
+    },
+    {
+      "epoch": 1.166077062816713,
+      "grad_norm": 2.279132330413246,
+      "learning_rate": 1e-06,
+      "loss": 0.1658,
+      "step": 6098
+    },
+    {
+      "epoch": 1.1662682856869682,
+      "grad_norm": 2.1312654914181075,
+      "learning_rate": 1e-06,
+      "loss": 0.1352,
+      "step": 6099
+    },
+    {
+      "epoch": 1.1664595085572234,
+      "grad_norm": 2.42462258302297,
+      "learning_rate": 1e-06,
+      "loss": 0.1147,
+      "step": 6100
+    },
+    {
+      "epoch": 1.1666507314274788,
+      "grad_norm": 3.033280469273067,
+      "learning_rate": 1e-06,
+      "loss": 0.0529,
+      "step": 6101
+    },
+    {
+      "epoch": 1.166841954297734,
+      "grad_norm": 5.19699220966741,
+      "learning_rate": 1e-06,
+      "loss": 0.1998,
+      "step": 6102
+    },
+    {
+      "epoch": 1.1670331771679894,
+      "grad_norm": 3.496527447785193,
+      "learning_rate": 1e-06,
+      "loss": 0.065,
+      "step": 6103
+    },
+    {
+      "epoch": 1.1672244000382446,
+      "grad_norm": 3.987565741400659,
+      "learning_rate": 1e-06,
+      "loss": 0.0611,
+      "step": 6104
+    },
+    {
+      "epoch": 1.1674156229084998,
+      "grad_norm": 5.587080917173907,
+      "learning_rate": 1e-06,
+      "loss": 0.1653,
+      "step": 6105
+    },
+    {
+      "epoch": 1.1676068457787552,
+      "grad_norm": 5.66501372669843,
+      "learning_rate": 1e-06,
+      "loss": 0.1289,
+      "step": 6106
+    },
+    {
+      "epoch": 1.1677980686490104,
+      "grad_norm": 4.1760982325316425,
+      "learning_rate": 1e-06,
+      "loss": 0.4504,
+      "step": 6107
+    },
+    {
+      "epoch": 1.1679892915192658,
+      "grad_norm": 3.3877044616030245,
+      "learning_rate": 1e-06,
+      "loss": 0.2744,
+      "step": 6108
+    },
+    {
+      "epoch": 1.168180514389521,
+      "grad_norm": 3.3350454383968855,
+      "learning_rate": 1e-06,
+      "loss": 0.12,
+      "step": 6109
+    },
+    {
+      "epoch": 1.1683717372597764,
+      "grad_norm": 3.689352313835786,
+      "learning_rate": 1e-06,
+      "loss": 0.2732,
+      "step": 6110
+    },
+    {
+      "epoch": 1.1685629601300316,
+      "grad_norm": 1.6297799015640468,
+      "learning_rate": 1e-06,
+      "loss": 0.0917,
+      "step": 6111
+    },
+    {
+      "epoch": 1.1687541830002868,
+      "grad_norm": 2.1580136591396837,
+      "learning_rate": 1e-06,
+      "loss": 0.0528,
+      "step": 6112
+    },
+    {
+      "epoch": 1.1689454058705422,
+      "grad_norm": 3.5306718741270005,
+      "learning_rate": 1e-06,
+      "loss": 0.1235,
+      "step": 6113
+    },
+    {
+      "epoch": 1.1691366287407974,
+      "grad_norm": 2.7003464723738353,
+      "learning_rate": 1e-06,
+      "loss": 0.0449,
+      "step": 6114
+    },
+    {
+      "epoch": 1.1693278516110528,
+      "grad_norm": 1.2672632705105018,
+      "learning_rate": 1e-06,
+      "loss": 0.0176,
+      "step": 6115
+    },
+    {
+      "epoch": 1.169519074481308,
+      "grad_norm": 5.036413919068266,
+      "learning_rate": 1e-06,
+      "loss": 0.1848,
+      "step": 6116
+    },
+    {
+      "epoch": 1.1697102973515632,
+      "grad_norm": 3.7667128388492217,
+      "learning_rate": 1e-06,
+      "loss": 0.0536,
+      "step": 6117
+    },
+    {
+      "epoch": 1.1699015202218186,
+      "grad_norm": 9.432279443899683,
+      "learning_rate": 1e-06,
+      "loss": 0.1973,
+      "step": 6118
+    },
+    {
+      "epoch": 1.1700927430920738,
+      "grad_norm": 4.452338540749716,
+      "learning_rate": 1e-06,
+      "loss": 0.5127,
+      "step": 6119
+    },
+    {
+      "epoch": 1.1702839659623292,
+      "grad_norm": 3.0346795570600342,
+      "learning_rate": 1e-06,
+      "loss": 0.117,
+      "step": 6120
+    },
+    {
+      "epoch": 1.1704751888325844,
+      "grad_norm": 4.162587088742085,
+      "learning_rate": 1e-06,
+      "loss": 0.2295,
+      "step": 6121
+    },
+    {
+      "epoch": 1.1706664117028396,
+      "grad_norm": 2.769343575600846,
+      "learning_rate": 1e-06,
+      "loss": 0.1444,
+      "step": 6122
+    },
+    {
+      "epoch": 1.170857634573095,
+      "grad_norm": 4.544344797780345,
+      "learning_rate": 1e-06,
+      "loss": 0.3575,
+      "step": 6123
+    },
+    {
+      "epoch": 1.1710488574433502,
+      "grad_norm": 3.5787443599463287,
+      "learning_rate": 1e-06,
+      "loss": 0.0784,
+      "step": 6124
+    },
+    {
+      "epoch": 1.1712400803136056,
+      "grad_norm": 2.5673229142344693,
+      "learning_rate": 1e-06,
+      "loss": 0.0965,
+      "step": 6125
+    },
+    {
+      "epoch": 1.1714313031838608,
+      "grad_norm": 4.171788132581104,
+      "learning_rate": 1e-06,
+      "loss": 0.0899,
+      "step": 6126
+    },
+    {
+      "epoch": 1.171622526054116,
+      "grad_norm": 3.5324630679220004,
+      "learning_rate": 1e-06,
+      "loss": 0.2433,
+      "step": 6127
+    },
+    {
+      "epoch": 1.1718137489243714,
+      "grad_norm": 3.498045511847166,
+      "learning_rate": 1e-06,
+      "loss": 0.0949,
+      "step": 6128
+    },
+    {
+      "epoch": 1.1720049717946266,
+      "grad_norm": 3.117636435550961,
+      "learning_rate": 1e-06,
+      "loss": 0.0501,
+      "step": 6129
+    },
+    {
+      "epoch": 1.172196194664882,
+      "grad_norm": 6.095399916862254,
+      "learning_rate": 1e-06,
+      "loss": 0.1825,
+      "step": 6130
+    },
+    {
+      "epoch": 1.1723874175351372,
+      "grad_norm": 3.8100330313304505,
+      "learning_rate": 1e-06,
+      "loss": 0.2788,
+      "step": 6131
+    },
+    {
+      "epoch": 1.1725786404053924,
+      "grad_norm": 4.859598136268112,
+      "learning_rate": 1e-06,
+      "loss": 0.2356,
+      "step": 6132
+    },
+    {
+      "epoch": 1.1727698632756478,
+      "grad_norm": 2.3176190191197614,
+      "learning_rate": 1e-06,
+      "loss": 0.0521,
+      "step": 6133
+    },
+    {
+      "epoch": 1.172961086145903,
+      "grad_norm": 2.889958155239983,
+      "learning_rate": 1e-06,
+      "loss": 0.2477,
+      "step": 6134
+    },
+    {
+      "epoch": 1.1731523090161584,
+      "grad_norm": 2.413614373448974,
+      "learning_rate": 1e-06,
+      "loss": 0.0421,
+      "step": 6135
+    },
+    {
+      "epoch": 1.1733435318864136,
+      "grad_norm": 3.8251096722207327,
+      "learning_rate": 1e-06,
+      "loss": 0.2534,
+      "step": 6136
+    },
+    {
+      "epoch": 1.1735347547566688,
+      "grad_norm": 2.059840365955924,
+      "learning_rate": 1e-06,
+      "loss": 0.0747,
+      "step": 6137
+    },
+    {
+      "epoch": 1.1737259776269242,
+      "grad_norm": 3.397188926305569,
+      "learning_rate": 1e-06,
+      "loss": 0.1005,
+      "step": 6138
+    },
+    {
+      "epoch": 1.1739172004971794,
+      "grad_norm": 3.7607679264650877,
+      "learning_rate": 1e-06,
+      "loss": 0.1486,
+      "step": 6139
+    },
+    {
+      "epoch": 1.1741084233674348,
+      "grad_norm": 3.7305862168596695,
+      "learning_rate": 1e-06,
+      "loss": 0.0793,
+      "step": 6140
+    },
+    {
+      "epoch": 1.17429964623769,
+      "grad_norm": 3.7922170190719267,
+      "learning_rate": 1e-06,
+      "loss": 0.0417,
+      "step": 6141
+    },
+    {
+      "epoch": 1.1744908691079452,
+      "grad_norm": 7.47073554935774,
+      "learning_rate": 1e-06,
+      "loss": 0.2529,
+      "step": 6142
+    },
+    {
+      "epoch": 1.1746820919782006,
+      "grad_norm": 6.064347447750748,
+      "learning_rate": 1e-06,
+      "loss": 0.3083,
+      "step": 6143
+    },
+    {
+      "epoch": 1.1748733148484558,
+      "grad_norm": 4.2401833199511545,
+      "learning_rate": 1e-06,
+      "loss": 0.3269,
+      "step": 6144
+    },
+    {
+      "epoch": 1.1750645377187112,
+      "grad_norm": 2.3788631541791205,
+      "learning_rate": 1e-06,
+      "loss": 0.1524,
+      "step": 6145
+    },
+    {
+      "epoch": 1.1752557605889664,
+      "grad_norm": 4.680089764508521,
+      "learning_rate": 1e-06,
+      "loss": 0.1911,
+      "step": 6146
+    },
+    {
+      "epoch": 1.1754469834592216,
+      "grad_norm": 2.938612057828664,
+      "learning_rate": 1e-06,
+      "loss": 0.2583,
+      "step": 6147
+    },
+    {
+      "epoch": 1.175638206329477,
+      "grad_norm": 4.046355343008032,
+      "learning_rate": 1e-06,
+      "loss": 0.135,
+      "step": 6148
+    },
+    {
+      "epoch": 1.1758294291997322,
+      "grad_norm": 2.5196772570036963,
+      "learning_rate": 1e-06,
+      "loss": 0.1758,
+      "step": 6149
+    },
+    {
+      "epoch": 1.1760206520699876,
+      "grad_norm": 2.1220260453994233,
+      "learning_rate": 1e-06,
+      "loss": 0.057,
+      "step": 6150
+    },
+    {
+      "epoch": 1.1762118749402428,
+      "grad_norm": 1.9048984927069184,
+      "learning_rate": 1e-06,
+      "loss": 0.0463,
+      "step": 6151
+    },
+    {
+      "epoch": 1.176403097810498,
+      "grad_norm": 3.2595621437483775,
+      "learning_rate": 1e-06,
+      "loss": 0.1027,
+      "step": 6152
+    },
+    {
+      "epoch": 1.1765943206807534,
+      "grad_norm": 1.8732187711341723,
+      "learning_rate": 1e-06,
+      "loss": 0.0302,
+      "step": 6153
+    },
+    {
+      "epoch": 1.1767855435510086,
+      "grad_norm": 11.910092832522208,
+      "learning_rate": 1e-06,
+      "loss": 0.0938,
+      "step": 6154
+    },
+    {
+      "epoch": 1.176976766421264,
+      "grad_norm": 8.739291368471866,
+      "learning_rate": 1e-06,
+      "loss": 0.3536,
+      "step": 6155
+    },
+    {
+      "epoch": 1.1771679892915192,
+      "grad_norm": 6.945276249553879,
+      "learning_rate": 1e-06,
+      "loss": 0.3873,
+      "step": 6156
+    },
+    {
+      "epoch": 1.1773592121617746,
+      "grad_norm": 5.143969158264062,
+      "learning_rate": 1e-06,
+      "loss": 0.3855,
+      "step": 6157
+    },
+    {
+      "epoch": 1.1775504350320298,
+      "grad_norm": 3.302396227977136,
+      "learning_rate": 1e-06,
+      "loss": 0.0863,
+      "step": 6158
+    },
+    {
+      "epoch": 1.177741657902285,
+      "grad_norm": 4.200726927565121,
+      "learning_rate": 1e-06,
+      "loss": 0.3113,
+      "step": 6159
+    },
+    {
+      "epoch": 1.1779328807725404,
+      "grad_norm": 1.395161932225447,
+      "learning_rate": 1e-06,
+      "loss": 0.1202,
+      "step": 6160
+    },
+    {
+      "epoch": 1.1781241036427956,
+      "grad_norm": 3.219917456001305,
+      "learning_rate": 1e-06,
+      "loss": 0.1637,
+      "step": 6161
+    },
+    {
+      "epoch": 1.178315326513051,
+      "grad_norm": 2.1232875206804414,
+      "learning_rate": 1e-06,
+      "loss": 0.0935,
+      "step": 6162
+    },
+    {
+      "epoch": 1.1785065493833062,
+      "grad_norm": 7.327349851560792,
+      "learning_rate": 1e-06,
+      "loss": 0.2337,
+      "step": 6163
+    },
+    {
+      "epoch": 1.1786977722535616,
+      "grad_norm": 3.5453642598018296,
+      "learning_rate": 1e-06,
+      "loss": 0.1045,
+      "step": 6164
+    },
+    {
+      "epoch": 1.1788889951238168,
+      "grad_norm": 2.131361918713018,
+      "learning_rate": 1e-06,
+      "loss": 0.0443,
+      "step": 6165
+    },
+    {
+      "epoch": 1.179080217994072,
+      "grad_norm": 2.8969882519095,
+      "learning_rate": 1e-06,
+      "loss": 0.0337,
+      "step": 6166
+    },
+    {
+      "epoch": 1.1792714408643274,
+      "grad_norm": 7.648674405006961,
+      "learning_rate": 1e-06,
+      "loss": 0.1021,
+      "step": 6167
+    },
+    {
+      "epoch": 1.1794626637345826,
+      "grad_norm": 8.286696690540545,
+      "learning_rate": 1e-06,
+      "loss": 0.2962,
+      "step": 6168
+    },
+    {
+      "epoch": 1.179653886604838,
+      "grad_norm": 5.805347266509181,
+      "learning_rate": 1e-06,
+      "loss": 0.4973,
+      "step": 6169
+    },
+    {
+      "epoch": 1.1798451094750932,
+      "grad_norm": 2.448142464687214,
+      "learning_rate": 1e-06,
+      "loss": 0.2713,
+      "step": 6170
+    },
+    {
+      "epoch": 1.1800363323453484,
+      "grad_norm": 7.603456173107542,
+      "learning_rate": 1e-06,
+      "loss": 0.2119,
+      "step": 6171
+    },
+    {
+      "epoch": 1.1802275552156039,
+      "grad_norm": 4.230795836198288,
+      "learning_rate": 1e-06,
+      "loss": 0.3238,
+      "step": 6172
+    },
+    {
+      "epoch": 1.180418778085859,
+      "grad_norm": 5.642174674320354,
+      "learning_rate": 1e-06,
+      "loss": 0.1055,
+      "step": 6173
+    },
+    {
+      "epoch": 1.1806100009561145,
+      "grad_norm": 2.6768667427842328,
+      "learning_rate": 1e-06,
+      "loss": 0.0549,
+      "step": 6174
+    },
+    {
+      "epoch": 1.1808012238263697,
+      "grad_norm": 2.9013675161982313,
+      "learning_rate": 1e-06,
+      "loss": 0.1684,
+      "step": 6175
+    },
+    {
+      "epoch": 1.1809924466966248,
+      "grad_norm": 3.116000683325324,
+      "learning_rate": 1e-06,
+      "loss": 0.0274,
+      "step": 6176
+    },
+    {
+      "epoch": 1.1811836695668803,
+      "grad_norm": 3.5193160961914334,
+      "learning_rate": 1e-06,
+      "loss": 0.2376,
+      "step": 6177
+    },
+    {
+      "epoch": 1.1813748924371354,
+      "grad_norm": 1.7196586287866982,
+      "learning_rate": 1e-06,
+      "loss": 0.0159,
+      "step": 6178
+    },
+    {
+      "epoch": 1.1815661153073909,
+      "grad_norm": 4.154155920710514,
+      "learning_rate": 1e-06,
+      "loss": 0.059,
+      "step": 6179
+    },
+    {
+      "epoch": 1.181757338177646,
+      "grad_norm": 4.848030610733837,
+      "learning_rate": 1e-06,
+      "loss": 0.2168,
+      "step": 6180
+    },
+    {
+      "epoch": 1.1819485610479012,
+      "grad_norm": 3.142653991276596,
+      "learning_rate": 1e-06,
+      "loss": 0.1814,
+      "step": 6181
+    },
+    {
+      "epoch": 1.1821397839181567,
+      "grad_norm": 4.829249970071698,
+      "learning_rate": 1e-06,
+      "loss": 0.4583,
+      "step": 6182
+    },
+    {
+      "epoch": 1.1823310067884119,
+      "grad_norm": 6.848361756386551,
+      "learning_rate": 1e-06,
+      "loss": 0.4834,
+      "step": 6183
+    },
+    {
+      "epoch": 1.1825222296586673,
+      "grad_norm": 2.982001034486362,
+      "learning_rate": 1e-06,
+      "loss": 0.0566,
+      "step": 6184
+    },
+    {
+      "epoch": 1.1827134525289225,
+      "grad_norm": 3.1549450282221265,
+      "learning_rate": 1e-06,
+      "loss": 0.0951,
+      "step": 6185
+    },
+    {
+      "epoch": 1.1829046753991777,
+      "grad_norm": 2.9202966353882047,
+      "learning_rate": 1e-06,
+      "loss": 0.2051,
+      "step": 6186
+    },
+    {
+      "epoch": 1.183095898269433,
+      "grad_norm": 3.256597278549462,
+      "learning_rate": 1e-06,
+      "loss": 0.2064,
+      "step": 6187
+    },
+    {
+      "epoch": 1.1832871211396883,
+      "grad_norm": 2.6173436360259252,
+      "learning_rate": 1e-06,
+      "loss": 0.0351,
+      "step": 6188
+    },
+    {
+      "epoch": 1.1834783440099437,
+      "grad_norm": 5.555900859697999,
+      "learning_rate": 1e-06,
+      "loss": 0.132,
+      "step": 6189
+    },
+    {
+      "epoch": 1.1836695668801989,
+      "grad_norm": 3.228741738858585,
+      "learning_rate": 1e-06,
+      "loss": 0.0867,
+      "step": 6190
+    },
+    {
+      "epoch": 1.183860789750454,
+      "grad_norm": 3.0151463262119202,
+      "learning_rate": 1e-06,
+      "loss": 0.0585,
+      "step": 6191
+    },
+    {
+      "epoch": 1.1840520126207095,
+      "grad_norm": 7.7987263837904015,
+      "learning_rate": 1e-06,
+      "loss": 0.0958,
+      "step": 6192
+    },
+    {
+      "epoch": 1.1842432354909647,
+      "grad_norm": 4.2681541909534975,
+      "learning_rate": 1e-06,
+      "loss": 0.3384,
+      "step": 6193
+    },
+    {
+      "epoch": 1.18443445836122,
+      "grad_norm": 3.498179507230458,
+      "learning_rate": 1e-06,
+      "loss": 0.2219,
+      "step": 6194
+    },
+    {
+      "epoch": 1.1846256812314753,
+      "grad_norm": 4.4513869792255045,
+      "learning_rate": 1e-06,
+      "loss": 0.1052,
+      "step": 6195
+    },
+    {
+      "epoch": 1.1848169041017305,
+      "grad_norm": 2.9417265658093177,
+      "learning_rate": 1e-06,
+      "loss": 0.162,
+      "step": 6196
+    },
+    {
+      "epoch": 1.1850081269719859,
+      "grad_norm": 3.002460424462555,
+      "learning_rate": 1e-06,
+      "loss": 0.212,
+      "step": 6197
+    },
+    {
+      "epoch": 1.185199349842241,
+      "grad_norm": 4.230264956425602,
+      "learning_rate": 1e-06,
+      "loss": 0.2342,
+      "step": 6198
+    },
+    {
+      "epoch": 1.1853905727124965,
+      "grad_norm": 3.929929934355119,
+      "learning_rate": 1e-06,
+      "loss": 0.2477,
+      "step": 6199
+    },
+    {
+      "epoch": 1.1855817955827517,
+      "grad_norm": 2.557217526316092,
+      "learning_rate": 1e-06,
+      "loss": 0.0967,
+      "step": 6200
+    },
+    {
+      "epoch": 1.1857730184530069,
+      "grad_norm": 1.7840183897139756,
+      "learning_rate": 1e-06,
+      "loss": 0.0802,
+      "step": 6201
+    },
+    {
+      "epoch": 1.1859642413232623,
+      "grad_norm": 2.65378285442037,
+      "learning_rate": 1e-06,
+      "loss": 0.0364,
+      "step": 6202
+    },
+    {
+      "epoch": 1.1861554641935175,
+      "grad_norm": 1.892924400121784,
+      "learning_rate": 1e-06,
+      "loss": 0.095,
+      "step": 6203
+    },
+    {
+      "epoch": 1.1863466870637729,
+      "grad_norm": 1.3996941658162791,
+      "learning_rate": 1e-06,
+      "loss": 0.0248,
+      "step": 6204
+    },
+    {
+      "epoch": 1.186537909934028,
+      "grad_norm": 7.391305001007598,
+      "learning_rate": 1e-06,
+      "loss": 0.1837,
+      "step": 6205
+    },
+    {
+      "epoch": 1.1867291328042833,
+      "grad_norm": 4.840073303739447,
+      "learning_rate": 1e-06,
+      "loss": 0.4051,
+      "step": 6206
+    },
+    {
+      "epoch": 1.1869203556745387,
+      "grad_norm": 4.597884827978777,
+      "learning_rate": 1e-06,
+      "loss": 0.3804,
+      "step": 6207
+    },
+    {
+      "epoch": 1.1871115785447939,
+      "grad_norm": 2.4308461067074214,
+      "learning_rate": 1e-06,
+      "loss": 0.2315,
+      "step": 6208
+    },
+    {
+      "epoch": 1.1873028014150493,
+      "grad_norm": 2.9605364766978295,
+      "learning_rate": 1e-06,
+      "loss": 0.1941,
+      "step": 6209
+    },
+    {
+      "epoch": 1.1874940242853045,
+      "grad_norm": 2.1266012329789676,
+      "learning_rate": 1e-06,
+      "loss": 0.0637,
+      "step": 6210
+    },
+    {
+      "epoch": 1.18768524715556,
+      "grad_norm": 2.765331532915805,
+      "learning_rate": 1e-06,
+      "loss": 0.0898,
+      "step": 6211
+    },
+    {
+      "epoch": 1.187876470025815,
+      "grad_norm": 2.4219044345174527,
+      "learning_rate": 1e-06,
+      "loss": 0.1135,
+      "step": 6212
+    },
+    {
+      "epoch": 1.1880676928960703,
+      "grad_norm": 3.185828574780153,
+      "learning_rate": 1e-06,
+      "loss": 0.0764,
+      "step": 6213
+    },
+    {
+      "epoch": 1.1882589157663257,
+      "grad_norm": 2.8529041217319957,
+      "learning_rate": 1e-06,
+      "loss": 0.0397,
+      "step": 6214
+    },
+    {
+      "epoch": 1.1884501386365809,
+      "grad_norm": 1.645070039829187,
+      "learning_rate": 1e-06,
+      "loss": 0.0476,
+      "step": 6215
+    },
+    {
+      "epoch": 1.1886413615068363,
+      "grad_norm": 2.8635020671801716,
+      "learning_rate": 1e-06,
+      "loss": 0.0772,
+      "step": 6216
+    },
+    {
+      "epoch": 1.1888325843770915,
+      "grad_norm": 9.472967134073789,
+      "learning_rate": 1e-06,
+      "loss": 0.0797,
+      "step": 6217
+    },
+    {
+      "epoch": 1.1890238072473467,
+      "grad_norm": 8.017199147940282,
+      "learning_rate": 1e-06,
+      "loss": 0.3214,
+      "step": 6218
+    },
+    {
+      "epoch": 1.189215030117602,
+      "grad_norm": 4.670198171523743,
+      "learning_rate": 1e-06,
+      "loss": 0.3498,
+      "step": 6219
+    },
+    {
+      "epoch": 1.1894062529878573,
+      "grad_norm": 3.9093874119257404,
+      "learning_rate": 1e-06,
+      "loss": 0.3513,
+      "step": 6220
+    },
+    {
+      "epoch": 1.1895974758581127,
+      "grad_norm": 7.6156381136562565,
+      "learning_rate": 1e-06,
+      "loss": 0.2326,
+      "step": 6221
+    },
+    {
+      "epoch": 1.189788698728368,
+      "grad_norm": 3.495671865911962,
+      "learning_rate": 1e-06,
+      "loss": 0.1708,
+      "step": 6222
+    },
+    {
+      "epoch": 1.1899799215986233,
+      "grad_norm": 2.150948905022293,
+      "learning_rate": 1e-06,
+      "loss": 0.0551,
+      "step": 6223
+    },
+    {
+      "epoch": 1.1901711444688785,
+      "grad_norm": 3.3032650602992413,
+      "learning_rate": 1e-06,
+      "loss": 0.1812,
+      "step": 6224
+    },
+    {
+      "epoch": 1.1903623673391337,
+      "grad_norm": 4.640150463361273,
+      "learning_rate": 1e-06,
+      "loss": 0.0895,
+      "step": 6225
+    },
+    {
+      "epoch": 1.1905535902093891,
+      "grad_norm": 3.3491697506693847,
+      "learning_rate": 1e-06,
+      "loss": 0.0796,
+      "step": 6226
+    },
+    {
+      "epoch": 1.1907448130796443,
+      "grad_norm": 2.206298382366372,
+      "learning_rate": 1e-06,
+      "loss": 0.047,
+      "step": 6227
+    },
+    {
+      "epoch": 1.1909360359498997,
+      "grad_norm": 3.448169587569421,
+      "learning_rate": 1e-06,
+      "loss": 0.0625,
+      "step": 6228
+    },
+    {
+      "epoch": 1.191127258820155,
+      "grad_norm": 3.1295482343597265,
+      "learning_rate": 1e-06,
+      "loss": 0.0806,
+      "step": 6229
+    },
+    {
+      "epoch": 1.19131848169041,
+      "grad_norm": 8.188888315620327,
+      "learning_rate": 1e-06,
+      "loss": 0.3707,
+      "step": 6230
+    },
+    {
+      "epoch": 1.1915097045606655,
+      "grad_norm": 2.7462411680847865,
+      "learning_rate": 1e-06,
+      "loss": 0.1043,
+      "step": 6231
+    },
+    {
+      "epoch": 1.1917009274309207,
+      "grad_norm": 3.8054315047143805,
+      "learning_rate": 1e-06,
+      "loss": 0.2878,
+      "step": 6232
+    },
+    {
+      "epoch": 1.1918921503011761,
+      "grad_norm": 1.9889720623495768,
+      "learning_rate": 1e-06,
+      "loss": 0.0527,
+      "step": 6233
+    },
+    {
+      "epoch": 1.1920833731714313,
+      "grad_norm": 5.149220802099519,
+      "learning_rate": 1e-06,
+      "loss": 0.2338,
+      "step": 6234
+    },
+    {
+      "epoch": 1.1922745960416865,
+      "grad_norm": 4.578249776656583,
+      "learning_rate": 1e-06,
+      "loss": 0.2038,
+      "step": 6235
+    },
+    {
+      "epoch": 1.192465818911942,
+      "grad_norm": 1.5915012576583318,
+      "learning_rate": 1e-06,
+      "loss": 0.0917,
+      "step": 6236
+    },
+    {
+      "epoch": 1.1926570417821971,
+      "grad_norm": 3.620879363428441,
+      "learning_rate": 1e-06,
+      "loss": 0.2892,
+      "step": 6237
+    },
+    {
+      "epoch": 1.1928482646524525,
+      "grad_norm": 6.044511673279175,
+      "learning_rate": 1e-06,
+      "loss": 0.2949,
+      "step": 6238
+    },
+    {
+      "epoch": 1.1930394875227077,
+      "grad_norm": 2.2823992407997062,
+      "learning_rate": 1e-06,
+      "loss": 0.0477,
+      "step": 6239
+    },
+    {
+      "epoch": 1.193230710392963,
+      "grad_norm": 4.7503246397208745,
+      "learning_rate": 1e-06,
+      "loss": 0.1058,
+      "step": 6240
+    },
+    {
+      "epoch": 1.1934219332632183,
+      "grad_norm": 2.617663166604211,
+      "learning_rate": 1e-06,
+      "loss": 0.0793,
+      "step": 6241
+    },
+    {
+      "epoch": 1.1936131561334735,
+      "grad_norm": 7.144611873349089,
+      "learning_rate": 1e-06,
+      "loss": 0.1022,
+      "step": 6242
+    },
+    {
+      "epoch": 1.193804379003729,
+      "grad_norm": 6.3405847028313795,
+      "learning_rate": 1e-06,
+      "loss": 0.2172,
+      "step": 6243
+    },
+    {
+      "epoch": 1.1939956018739841,
+      "grad_norm": 9.604071516667586,
+      "learning_rate": 1e-06,
+      "loss": 0.4007,
+      "step": 6244
+    },
+    {
+      "epoch": 1.1941868247442393,
+      "grad_norm": 1.8985308086556811,
+      "learning_rate": 1e-06,
+      "loss": 0.1387,
+      "step": 6245
+    },
+    {
+      "epoch": 1.1943780476144947,
+      "grad_norm": 4.32038622719076,
+      "learning_rate": 1e-06,
+      "loss": 0.2813,
+      "step": 6246
+    },
+    {
+      "epoch": 1.19456927048475,
+      "grad_norm": 3.754823761117262,
+      "learning_rate": 1e-06,
+      "loss": 0.1874,
+      "step": 6247
+    },
+    {
+      "epoch": 1.1947604933550053,
+      "grad_norm": 1.5576740505637507,
+      "learning_rate": 1e-06,
+      "loss": 0.1377,
+      "step": 6248
+    },
+    {
+      "epoch": 1.1949517162252605,
+      "grad_norm": 4.859496087531242,
+      "learning_rate": 1e-06,
+      "loss": 0.1677,
+      "step": 6249
+    },
+    {
+      "epoch": 1.1951429390955157,
+      "grad_norm": 4.02322203420331,
+      "learning_rate": 1e-06,
+      "loss": 0.2898,
+      "step": 6250
+    },
+    {
+      "epoch": 1.1953341619657711,
+      "grad_norm": 4.865002283956339,
+      "learning_rate": 1e-06,
+      "loss": 0.1332,
+      "step": 6251
+    },
+    {
+      "epoch": 1.1955253848360263,
+      "grad_norm": 6.487726214218974,
+      "learning_rate": 1e-06,
+      "loss": 0.0358,
+      "step": 6252
+    },
+    {
+      "epoch": 1.1957166077062817,
+      "grad_norm": 2.9336087039815264,
+      "learning_rate": 1e-06,
+      "loss": 0.0527,
+      "step": 6253
+    },
+    {
+      "epoch": 1.195907830576537,
+      "grad_norm": 3.1841332101416167,
+      "learning_rate": 1e-06,
+      "loss": 0.0469,
+      "step": 6254
+    },
+    {
+      "epoch": 1.1960990534467921,
+      "grad_norm": 6.4730897066502795,
+      "learning_rate": 1e-06,
+      "loss": 0.1034,
+      "step": 6255
+    },
+    {
+      "epoch": 1.1962902763170475,
+      "grad_norm": 4.782875532690147,
+      "learning_rate": 1e-06,
+      "loss": 0.3601,
+      "step": 6256
+    },
+    {
+      "epoch": 1.1964814991873027,
+      "grad_norm": 4.7244281079582136,
+      "learning_rate": 1e-06,
+      "loss": 0.4177,
+      "step": 6257
+    },
+    {
+      "epoch": 1.1966727220575581,
+      "grad_norm": 3.5262918688733667,
+      "learning_rate": 1e-06,
+      "loss": 0.3078,
+      "step": 6258
+    },
+    {
+      "epoch": 1.1968639449278133,
+      "grad_norm": 2.224414203981085,
+      "learning_rate": 1e-06,
+      "loss": 0.0989,
+      "step": 6259
+    },
+    {
+      "epoch": 1.1970551677980685,
+      "grad_norm": 2.6606772832592562,
+      "learning_rate": 1e-06,
+      "loss": 0.1074,
+      "step": 6260
+    },
+    {
+      "epoch": 1.197246390668324,
+      "grad_norm": 3.833212822941021,
+      "learning_rate": 1e-06,
+      "loss": 0.1402,
+      "step": 6261
+    },
+    {
+      "epoch": 1.1974376135385791,
+      "grad_norm": 2.4952794330199852,
+      "learning_rate": 1e-06,
+      "loss": 0.094,
+      "step": 6262
+    },
+    {
+      "epoch": 1.1976288364088346,
+      "grad_norm": 4.289100521732087,
+      "learning_rate": 1e-06,
+      "loss": 0.383,
+      "step": 6263
+    },
+    {
+      "epoch": 1.1978200592790897,
+      "grad_norm": 5.181407245969379,
+      "learning_rate": 1e-06,
+      "loss": 0.1429,
+      "step": 6264
+    },
+    {
+      "epoch": 1.1980112821493452,
+      "grad_norm": 2.319978529403255,
+      "learning_rate": 1e-06,
+      "loss": 0.0179,
+      "step": 6265
+    },
+    {
+      "epoch": 1.1982025050196004,
+      "grad_norm": 3.419256314403954,
+      "learning_rate": 1e-06,
+      "loss": 0.0856,
+      "step": 6266
+    },
+    {
+      "epoch": 1.1983937278898555,
+      "grad_norm": 3.679187392410074,
+      "learning_rate": 1e-06,
+      "loss": 0.0558,
+      "step": 6267
+    },
+    {
+      "epoch": 1.198584950760111,
+      "grad_norm": 5.806069046666445,
+      "learning_rate": 1e-06,
+      "loss": 0.4942,
+      "step": 6268
+    },
+    {
+      "epoch": 1.1987761736303661,
+      "grad_norm": 5.088428174578255,
+      "learning_rate": 1e-06,
+      "loss": 0.5731,
+      "step": 6269
+    },
+    {
+      "epoch": 1.1989673965006216,
+      "grad_norm": 3.286047344031549,
+      "learning_rate": 1e-06,
+      "loss": 0.1636,
+      "step": 6270
+    },
+    {
+      "epoch": 1.1991586193708768,
+      "grad_norm": 4.024512997183877,
+      "learning_rate": 1e-06,
+      "loss": 0.1157,
+      "step": 6271
+    },
+    {
+      "epoch": 1.199349842241132,
+      "grad_norm": 4.000772878365826,
+      "learning_rate": 1e-06,
+      "loss": 0.2659,
+      "step": 6272
+    },
+    {
+      "epoch": 1.1995410651113874,
+      "grad_norm": 5.458707646161293,
+      "learning_rate": 1e-06,
+      "loss": 0.4278,
+      "step": 6273
+    },
+    {
+      "epoch": 1.1997322879816426,
+      "grad_norm": 3.757612893241738,
+      "learning_rate": 1e-06,
+      "loss": 0.2149,
+      "step": 6274
+    },
+    {
+      "epoch": 1.199923510851898,
+      "grad_norm": 2.852156569379115,
+      "learning_rate": 1e-06,
+      "loss": 0.0464,
+      "step": 6275
+    },
+    {
+      "epoch": 1.2001147337221532,
+      "grad_norm": 2.2633686273975675,
+      "learning_rate": 1e-06,
+      "loss": 0.0572,
+      "step": 6276
+    },
+    {
+      "epoch": 1.2003059565924086,
+      "grad_norm": 1.880396071877076,
+      "learning_rate": 1e-06,
+      "loss": 0.0247,
+      "step": 6277
+    },
+    {
+      "epoch": 1.2004971794626638,
+      "grad_norm": 1.7418054409804575,
+      "learning_rate": 1e-06,
+      "loss": 0.0286,
+      "step": 6278
+    },
+    {
+      "epoch": 1.200688402332919,
+      "grad_norm": 5.331232412114709,
+      "learning_rate": 1e-06,
+      "loss": 0.1148,
+      "step": 6279
+    },
+    {
+      "epoch": 1.2008796252031744,
+      "grad_norm": 3.547332330871949,
+      "learning_rate": 1e-06,
+      "loss": 0.0665,
+      "step": 6280
+    },
+    {
+      "epoch": 1.2010708480734296,
+      "grad_norm": 4.726359314137364,
+      "learning_rate": 1e-06,
+      "loss": 0.7016,
+      "step": 6281
+    },
+    {
+      "epoch": 1.201262070943685,
+      "grad_norm": 5.207916344798459,
+      "learning_rate": 1e-06,
+      "loss": 0.2426,
+      "step": 6282
+    },
+    {
+      "epoch": 1.2014532938139402,
+      "grad_norm": 4.489351813888315,
+      "learning_rate": 1e-06,
+      "loss": 0.3406,
+      "step": 6283
+    },
+    {
+      "epoch": 1.2016445166841954,
+      "grad_norm": 2.1971732836499287,
+      "learning_rate": 1e-06,
+      "loss": 0.1627,
+      "step": 6284
+    },
+    {
+      "epoch": 1.2018357395544508,
+      "grad_norm": 3.86819598987234,
+      "learning_rate": 1e-06,
+      "loss": 0.1652,
+      "step": 6285
+    },
+    {
+      "epoch": 1.202026962424706,
+      "grad_norm": 3.6714264291329783,
+      "learning_rate": 1e-06,
+      "loss": 0.2054,
+      "step": 6286
+    },
+    {
+      "epoch": 1.2022181852949614,
+      "grad_norm": 2.898156000744418,
+      "learning_rate": 1e-06,
+      "loss": 0.0468,
+      "step": 6287
+    },
+    {
+      "epoch": 1.2024094081652166,
+      "grad_norm": 1.9674782051129829,
+      "learning_rate": 1e-06,
+      "loss": 0.0377,
+      "step": 6288
+    },
+    {
+      "epoch": 1.2026006310354718,
+      "grad_norm": 3.538128121082713,
+      "learning_rate": 1e-06,
+      "loss": 0.0867,
+      "step": 6289
+    },
+    {
+      "epoch": 1.2027918539057272,
+      "grad_norm": 1.772053922469978,
+      "learning_rate": 1e-06,
+      "loss": 0.021,
+      "step": 6290
+    },
+    {
+      "epoch": 1.2029830767759824,
+      "grad_norm": 7.798236491066777,
+      "learning_rate": 1e-06,
+      "loss": 0.0573,
+      "step": 6291
+    },
+    {
+      "epoch": 1.2031742996462378,
+      "grad_norm": 4.098203609772023,
+      "learning_rate": 1e-06,
+      "loss": 0.0869,
+      "step": 6292
+    },
+    {
+      "epoch": 1.203365522516493,
+      "grad_norm": 4.466484936295792,
+      "learning_rate": 1e-06,
+      "loss": 0.331,
+      "step": 6293
+    },
+    {
+      "epoch": 1.2035567453867482,
+      "grad_norm": 5.51423571328874,
+      "learning_rate": 1e-06,
+      "loss": 0.2364,
+      "step": 6294
+    },
+    {
+      "epoch": 1.2037479682570036,
+      "grad_norm": 2.31791475784065,
+      "learning_rate": 1e-06,
+      "loss": 0.2221,
+      "step": 6295
+    },
+    {
+      "epoch": 1.2039391911272588,
+      "grad_norm": 4.280765478185372,
+      "learning_rate": 1e-06,
+      "loss": 0.3772,
+      "step": 6296
+    },
+    {
+      "epoch": 1.2041304139975142,
+      "grad_norm": 3.7577085738374008,
+      "learning_rate": 1e-06,
+      "loss": 0.2387,
+      "step": 6297
+    },
+    {
+      "epoch": 1.2043216368677694,
+      "grad_norm": 4.047937437833045,
+      "learning_rate": 1e-06,
+      "loss": 0.2273,
+      "step": 6298
+    },
+    {
+      "epoch": 1.2045128597380246,
+      "grad_norm": 2.219043336201863,
+      "learning_rate": 1e-06,
+      "loss": 0.0691,
+      "step": 6299
+    },
+    {
+      "epoch": 1.20470408260828,
+      "grad_norm": 4.947179264198857,
+      "learning_rate": 1e-06,
+      "loss": 0.297,
+      "step": 6300
+    },
+    {
+      "epoch": 1.2048953054785352,
+      "grad_norm": 5.135550826707995,
+      "learning_rate": 1e-06,
+      "loss": 0.0519,
+      "step": 6301
+    },
+    {
+      "epoch": 1.2050865283487906,
+      "grad_norm": 3.320354290867516,
+      "learning_rate": 1e-06,
+      "loss": 0.1433,
+      "step": 6302
+    },
+    {
+      "epoch": 1.2052777512190458,
+      "grad_norm": 3.1479552398950674,
+      "learning_rate": 1e-06,
+      "loss": 0.0594,
+      "step": 6303
+    },
+    {
+      "epoch": 1.205468974089301,
+      "grad_norm": 3.7488120422622484,
+      "learning_rate": 1e-06,
+      "loss": 0.0203,
+      "step": 6304
+    },
+    {
+      "epoch": 1.2056601969595564,
+      "grad_norm": 6.843935053862243,
+      "learning_rate": 1e-06,
+      "loss": 0.1673,
+      "step": 6305
+    },
+    {
+      "epoch": 1.2058514198298116,
+      "grad_norm": 3.1651806774666595,
+      "learning_rate": 1e-06,
+      "loss": 0.2022,
+      "step": 6306
+    },
+    {
+      "epoch": 1.206042642700067,
+      "grad_norm": 4.283960305921375,
+      "learning_rate": 1e-06,
+      "loss": 0.1656,
+      "step": 6307
+    },
+    {
+      "epoch": 1.2062338655703222,
+      "grad_norm": 2.130659644733996,
+      "learning_rate": 1e-06,
+      "loss": 0.1423,
+      "step": 6308
+    },
+    {
+      "epoch": 1.2064250884405774,
+      "grad_norm": 2.4117529231184545,
+      "learning_rate": 1e-06,
+      "loss": 0.1062,
+      "step": 6309
+    },
+    {
+      "epoch": 1.2066163113108328,
+      "grad_norm": 5.067579946691841,
+      "learning_rate": 1e-06,
+      "loss": 0.2634,
+      "step": 6310
+    },
+    {
+      "epoch": 1.206807534181088,
+      "grad_norm": 4.106383884981998,
+      "learning_rate": 1e-06,
+      "loss": 0.1727,
+      "step": 6311
+    },
+    {
+      "epoch": 1.2069987570513434,
+      "grad_norm": 4.788406103058981,
+      "learning_rate": 1e-06,
+      "loss": 0.0534,
+      "step": 6312
+    },
+    {
+      "epoch": 1.2071899799215986,
+      "grad_norm": 3.596936612633074,
+      "learning_rate": 1e-06,
+      "loss": 0.0824,
+      "step": 6313
+    },
+    {
+      "epoch": 1.2073812027918538,
+      "grad_norm": 3.4994312914248393,
+      "learning_rate": 1e-06,
+      "loss": 0.0584,
+      "step": 6314
+    },
+    {
+      "epoch": 1.2075724256621092,
+      "grad_norm": 3.5874541489066885,
+      "learning_rate": 1e-06,
+      "loss": 0.0422,
+      "step": 6315
+    },
+    {
+      "epoch": 1.2077636485323644,
+      "grad_norm": 1.824640915438196,
+      "learning_rate": 1e-06,
+      "loss": 0.0245,
+      "step": 6316
+    },
+    {
+      "epoch": 1.2079548714026198,
+      "grad_norm": 5.70793989547383,
+      "learning_rate": 1e-06,
+      "loss": 0.0771,
+      "step": 6317
+    },
+    {
+      "epoch": 1.208146094272875,
+      "grad_norm": 4.337539074636564,
+      "learning_rate": 1e-06,
+      "loss": 0.2966,
+      "step": 6318
+    },
+    {
+      "epoch": 1.2083373171431302,
+      "grad_norm": 3.1958177535478463,
+      "learning_rate": 1e-06,
+      "loss": 0.1584,
+      "step": 6319
+    },
+    {
+      "epoch": 1.2085285400133856,
+      "grad_norm": 3.1857972926653093,
+      "learning_rate": 1e-06,
+      "loss": 0.3555,
+      "step": 6320
+    },
+    {
+      "epoch": 1.2087197628836408,
+      "grad_norm": 2.239928699448552,
+      "learning_rate": 1e-06,
+      "loss": 0.1079,
+      "step": 6321
+    },
+    {
+      "epoch": 1.2089109857538962,
+      "grad_norm": 3.5322430319978224,
+      "learning_rate": 1e-06,
+      "loss": 0.4156,
+      "step": 6322
+    },
+    {
+      "epoch": 1.2091022086241514,
+      "grad_norm": 4.826936939814655,
+      "learning_rate": 1e-06,
+      "loss": 0.3362,
+      "step": 6323
+    },
+    {
+      "epoch": 1.2092934314944068,
+      "grad_norm": 3.2113660540311186,
+      "learning_rate": 1e-06,
+      "loss": 0.1322,
+      "step": 6324
+    },
+    {
+      "epoch": 1.209484654364662,
+      "grad_norm": 3.0192406676156796,
+      "learning_rate": 1e-06,
+      "loss": 0.0828,
+      "step": 6325
+    },
+    {
+      "epoch": 1.2096758772349172,
+      "grad_norm": 3.471769513605854,
+      "learning_rate": 1e-06,
+      "loss": 0.0726,
+      "step": 6326
+    },
+    {
+      "epoch": 1.2098671001051726,
+      "grad_norm": 2.443835900399422,
+      "learning_rate": 1e-06,
+      "loss": 0.0352,
+      "step": 6327
+    },
+    {
+      "epoch": 1.2100583229754278,
+      "grad_norm": 3.361102276694998,
+      "learning_rate": 1e-06,
+      "loss": 0.2602,
+      "step": 6328
+    },
+    {
+      "epoch": 1.2102495458456832,
+      "grad_norm": 4.8262434073766,
+      "learning_rate": 1e-06,
+      "loss": 0.0647,
+      "step": 6329
+    },
+    {
+      "epoch": 1.2104407687159384,
+      "grad_norm": 2.9955680216816933,
+      "learning_rate": 1e-06,
+      "loss": 0.1364,
+      "step": 6330
+    },
+    {
+      "epoch": 1.2106319915861936,
+      "grad_norm": 5.79051595915616,
+      "learning_rate": 1e-06,
+      "loss": 0.4762,
+      "step": 6331
+    },
+    {
+      "epoch": 1.210823214456449,
+      "grad_norm": 3.5151001941189137,
+      "learning_rate": 1e-06,
+      "loss": 0.4477,
+      "step": 6332
+    },
+    {
+      "epoch": 1.2110144373267042,
+      "grad_norm": 3.9118235384141395,
+      "learning_rate": 1e-06,
+      "loss": 0.1067,
+      "step": 6333
+    },
+    {
+      "epoch": 1.2112056601969596,
+      "grad_norm": 2.874876600188014,
+      "learning_rate": 1e-06,
+      "loss": 0.1642,
+      "step": 6334
+    },
+    {
+      "epoch": 1.2113968830672148,
+      "grad_norm": 1.9385580434898322,
+      "learning_rate": 1e-06,
+      "loss": 0.0361,
+      "step": 6335
+    },
+    {
+      "epoch": 1.2115881059374702,
+      "grad_norm": 4.2998795425930885,
+      "learning_rate": 1e-06,
+      "loss": 0.0915,
+      "step": 6336
+    },
+    {
+      "epoch": 1.2117793288077254,
+      "grad_norm": 2.8462136464567895,
+      "learning_rate": 1e-06,
+      "loss": 0.048,
+      "step": 6337
+    },
+    {
+      "epoch": 1.2119705516779806,
+      "grad_norm": 3.5818081352610536,
+      "learning_rate": 1e-06,
+      "loss": 0.1185,
+      "step": 6338
+    },
+    {
+      "epoch": 1.212161774548236,
+      "grad_norm": 2.805283846985578,
+      "learning_rate": 1e-06,
+      "loss": 0.0709,
+      "step": 6339
+    },
+    {
+      "epoch": 1.2123529974184912,
+      "grad_norm": 3.3820516598465207,
+      "learning_rate": 1e-06,
+      "loss": 0.0463,
+      "step": 6340
+    },
+    {
+      "epoch": 1.2125442202887466,
+      "grad_norm": 2.2343905788492076,
+      "learning_rate": 1e-06,
+      "loss": 0.032,
+      "step": 6341
+    },
+    {
+      "epoch": 1.2127354431590018,
+      "grad_norm": 6.362434120061984,
+      "learning_rate": 1e-06,
+      "loss": 0.1308,
+      "step": 6342
+    },
+    {
+      "epoch": 1.212926666029257,
+      "grad_norm": 7.895104313379313,
+      "learning_rate": 1e-06,
+      "loss": 0.4712,
+      "step": 6343
+    },
+    {
+      "epoch": 1.2131178888995124,
+      "grad_norm": 3.549781717386136,
+      "learning_rate": 1e-06,
+      "loss": 0.3189,
+      "step": 6344
+    },
+    {
+      "epoch": 1.2133091117697676,
+      "grad_norm": 3.5234887908111996,
+      "learning_rate": 1e-06,
+      "loss": 0.1862,
+      "step": 6345
+    },
+    {
+      "epoch": 1.213500334640023,
+      "grad_norm": 3.631799831554527,
+      "learning_rate": 1e-06,
+      "loss": 0.1933,
+      "step": 6346
+    },
+    {
+      "epoch": 1.2136915575102782,
+      "grad_norm": 5.145403186359191,
+      "learning_rate": 1e-06,
+      "loss": 0.3527,
+      "step": 6347
+    },
+    {
+      "epoch": 1.2138827803805334,
+      "grad_norm": 3.9161106012485063,
+      "learning_rate": 1e-06,
+      "loss": 0.1271,
+      "step": 6348
+    },
+    {
+      "epoch": 1.2140740032507888,
+      "grad_norm": 4.914824167484533,
+      "learning_rate": 1e-06,
+      "loss": 0.0513,
+      "step": 6349
+    },
+    {
+      "epoch": 1.214265226121044,
+      "grad_norm": 1.9588024781384177,
+      "learning_rate": 1e-06,
+      "loss": 0.0704,
+      "step": 6350
+    },
+    {
+      "epoch": 1.2144564489912995,
+      "grad_norm": 2.5971860990771813,
+      "learning_rate": 1e-06,
+      "loss": 0.206,
+      "step": 6351
+    },
+    {
+      "epoch": 1.2146476718615546,
+      "grad_norm": 1.6102516971526482,
+      "learning_rate": 1e-06,
+      "loss": 0.0247,
+      "step": 6352
+    },
+    {
+      "epoch": 1.2148388947318098,
+      "grad_norm": 1.5887021176545768,
+      "learning_rate": 1e-06,
+      "loss": 0.0262,
+      "step": 6353
+    },
+    {
+      "epoch": 1.2150301176020653,
+      "grad_norm": 2.1849076305217925,
+      "learning_rate": 1e-06,
+      "loss": 0.0304,
+      "step": 6354
+    },
+    {
+      "epoch": 1.2152213404723204,
+      "grad_norm": 5.85371015223565,
+      "learning_rate": 1e-06,
+      "loss": 0.2031,
+      "step": 6355
+    },
+    {
+      "epoch": 1.2154125633425759,
+      "grad_norm": 5.983865980208096,
+      "learning_rate": 1e-06,
+      "loss": 0.1658,
+      "step": 6356
+    },
+    {
+      "epoch": 1.215603786212831,
+      "grad_norm": 5.257311135317669,
+      "learning_rate": 1e-06,
+      "loss": 0.4358,
+      "step": 6357
+    },
+    {
+      "epoch": 1.2157950090830862,
+      "grad_norm": 3.8517278980079794,
+      "learning_rate": 1e-06,
+      "loss": 0.1275,
+      "step": 6358
+    },
+    {
+      "epoch": 1.2159862319533417,
+      "grad_norm": 2.545792706587347,
+      "learning_rate": 1e-06,
+      "loss": 0.1083,
+      "step": 6359
+    },
+    {
+      "epoch": 1.2161774548235968,
+      "grad_norm": 3.138056373658742,
+      "learning_rate": 1e-06,
+      "loss": 0.2142,
+      "step": 6360
+    },
+    {
+      "epoch": 1.2163686776938523,
+      "grad_norm": 2.342637370182387,
+      "learning_rate": 1e-06,
+      "loss": 0.0868,
+      "step": 6361
+    },
+    {
+      "epoch": 1.2165599005641075,
+      "grad_norm": 2.8306354318848954,
+      "learning_rate": 1e-06,
+      "loss": 0.1651,
+      "step": 6362
+    },
+    {
+      "epoch": 1.2167511234343626,
+      "grad_norm": 2.041420110749622,
+      "learning_rate": 1e-06,
+      "loss": 0.0909,
+      "step": 6363
+    },
+    {
+      "epoch": 1.216942346304618,
+      "grad_norm": 2.338621912700196,
+      "learning_rate": 1e-06,
+      "loss": 0.0624,
+      "step": 6364
+    },
+    {
+      "epoch": 1.2171335691748733,
+      "grad_norm": 4.339367309018462,
+      "learning_rate": 1e-06,
+      "loss": 0.0788,
+      "step": 6365
+    },
+    {
+      "epoch": 1.2173247920451287,
+      "grad_norm": 3.4175501975870586,
+      "learning_rate": 1e-06,
+      "loss": 0.096,
+      "step": 6366
+    },
+    {
+      "epoch": 1.2175160149153839,
+      "grad_norm": 5.362369334800849,
+      "learning_rate": 1e-06,
+      "loss": 0.087,
+      "step": 6367
+    },
+    {
+      "epoch": 1.217707237785639,
+      "grad_norm": 4.352358185431433,
+      "learning_rate": 1e-06,
+      "loss": 0.1206,
+      "step": 6368
+    },
+    {
+      "epoch": 1.2178984606558945,
+      "grad_norm": 2.9278982166255276,
+      "learning_rate": 1e-06,
+      "loss": 0.2119,
+      "step": 6369
+    },
+    {
+      "epoch": 1.2180896835261497,
+      "grad_norm": 2.9795296365640627,
+      "learning_rate": 1e-06,
+      "loss": 0.0831,
+      "step": 6370
+    },
+    {
+      "epoch": 1.218280906396405,
+      "grad_norm": 4.707099933044027,
+      "learning_rate": 1e-06,
+      "loss": 0.2097,
+      "step": 6371
+    },
+    {
+      "epoch": 1.2184721292666603,
+      "grad_norm": 5.000316419126078,
+      "learning_rate": 1e-06,
+      "loss": 0.2246,
+      "step": 6372
+    },
+    {
+      "epoch": 1.2186633521369155,
+      "grad_norm": 2.6379642059986574,
+      "learning_rate": 1e-06,
+      "loss": 0.1338,
+      "step": 6373
+    },
+    {
+      "epoch": 1.2188545750071709,
+      "grad_norm": 2.1552061788610573,
+      "learning_rate": 1e-06,
+      "loss": 0.1427,
+      "step": 6374
+    },
+    {
+      "epoch": 1.219045797877426,
+      "grad_norm": 2.509168596912031,
+      "learning_rate": 1e-06,
+      "loss": 0.0476,
+      "step": 6375
+    },
+    {
+      "epoch": 1.2192370207476815,
+      "grad_norm": 3.046469793518364,
+      "learning_rate": 1e-06,
+      "loss": 0.0697,
+      "step": 6376
+    },
+    {
+      "epoch": 1.2194282436179367,
+      "grad_norm": 4.108678490388112,
+      "learning_rate": 1e-06,
+      "loss": 0.1518,
+      "step": 6377
+    },
+    {
+      "epoch": 1.219619466488192,
+      "grad_norm": 3.342215783212199,
+      "learning_rate": 1e-06,
+      "loss": 0.0733,
+      "step": 6378
+    },
+    {
+      "epoch": 1.2198106893584473,
+      "grad_norm": 1.6737729289369607,
+      "learning_rate": 1e-06,
+      "loss": 0.022,
+      "step": 6379
+    },
+    {
+      "epoch": 1.2200019122287025,
+      "grad_norm": 6.110395185361505,
+      "learning_rate": 1e-06,
+      "loss": 0.1848,
+      "step": 6380
+    },
+    {
+      "epoch": 1.2201931350989579,
+      "grad_norm": 9.030193966697054,
+      "learning_rate": 1e-06,
+      "loss": 0.5446,
+      "step": 6381
+    },
+    {
+      "epoch": 1.220384357969213,
+      "grad_norm": 4.58883549431091,
+      "learning_rate": 1e-06,
+      "loss": 0.2286,
+      "step": 6382
+    },
+    {
+      "epoch": 1.2205755808394685,
+      "grad_norm": 3.8862606219480575,
+      "learning_rate": 1e-06,
+      "loss": 0.4932,
+      "step": 6383
+    },
+    {
+      "epoch": 1.2207668037097237,
+      "grad_norm": 3.371747180117581,
+      "learning_rate": 1e-06,
+      "loss": 0.26,
+      "step": 6384
+    },
+    {
+      "epoch": 1.2209580265799789,
+      "grad_norm": 2.774085600441901,
+      "learning_rate": 1e-06,
+      "loss": 0.1889,
+      "step": 6385
+    },
+    {
+      "epoch": 1.2211492494502343,
+      "grad_norm": 5.805390963535699,
+      "learning_rate": 1e-06,
+      "loss": 0.1421,
+      "step": 6386
+    },
+    {
+      "epoch": 1.2213404723204895,
+      "grad_norm": 3.8570058863908683,
+      "learning_rate": 1e-06,
+      "loss": 0.0878,
+      "step": 6387
+    },
+    {
+      "epoch": 1.221531695190745,
+      "grad_norm": 1.8812751666948873,
+      "learning_rate": 1e-06,
+      "loss": 0.166,
+      "step": 6388
+    },
+    {
+      "epoch": 1.221722918061,
+      "grad_norm": 2.4686295685184767,
+      "learning_rate": 1e-06,
+      "loss": 0.0724,
+      "step": 6389
+    },
+    {
+      "epoch": 1.2219141409312555,
+      "grad_norm": 1.6654341511416597,
+      "learning_rate": 1e-06,
+      "loss": 0.0401,
+      "step": 6390
+    },
+    {
+      "epoch": 1.2221053638015107,
+      "grad_norm": 2.1654226937967698,
+      "learning_rate": 1e-06,
+      "loss": 0.0287,
+      "step": 6391
+    },
+    {
+      "epoch": 1.2222965866717659,
+      "grad_norm": 3.9316714305021416,
+      "learning_rate": 1e-06,
+      "loss": 0.0931,
+      "step": 6392
+    },
+    {
+      "epoch": 1.2224878095420213,
+      "grad_norm": 4.54207774877081,
+      "learning_rate": 1e-06,
+      "loss": 0.2236,
+      "step": 6393
+    },
+    {
+      "epoch": 1.2226790324122765,
+      "grad_norm": 5.458152748595623,
+      "learning_rate": 1e-06,
+      "loss": 0.3743,
+      "step": 6394
+    },
+    {
+      "epoch": 1.222870255282532,
+      "grad_norm": 4.279219096951208,
+      "learning_rate": 1e-06,
+      "loss": 0.2209,
+      "step": 6395
+    },
+    {
+      "epoch": 1.223061478152787,
+      "grad_norm": 3.5506352729292865,
+      "learning_rate": 1e-06,
+      "loss": 0.107,
+      "step": 6396
+    },
+    {
+      "epoch": 1.2232527010230423,
+      "grad_norm": 2.6418167224066846,
+      "learning_rate": 1e-06,
+      "loss": 0.1735,
+      "step": 6397
+    },
+    {
+      "epoch": 1.2234439238932977,
+      "grad_norm": 2.296414595152271,
+      "learning_rate": 1e-06,
+      "loss": 0.3036,
+      "step": 6398
+    },
+    {
+      "epoch": 1.223635146763553,
+      "grad_norm": 3.213581416761642,
+      "learning_rate": 1e-06,
+      "loss": 0.2004,
+      "step": 6399
+    },
+    {
+      "epoch": 1.2238263696338083,
+      "grad_norm": 2.7339819925436557,
+      "learning_rate": 1e-06,
+      "loss": 0.2066,
+      "step": 6400
+    },
+    {
+      "epoch": 1.2240175925040635,
+      "grad_norm": 1.629987106506278,
+      "learning_rate": 1e-06,
+      "loss": 0.0477,
+      "step": 6401
+    },
+    {
+      "epoch": 1.2242088153743187,
+      "grad_norm": 4.146275993497999,
+      "learning_rate": 1e-06,
+      "loss": 0.0351,
+      "step": 6402
+    },
+    {
+      "epoch": 1.224400038244574,
+      "grad_norm": 3.163648187332221,
+      "learning_rate": 1e-06,
+      "loss": 0.0442,
+      "step": 6403
+    },
+    {
+      "epoch": 1.2245912611148293,
+      "grad_norm": 2.3182723712744693,
+      "learning_rate": 1e-06,
+      "loss": 0.0324,
+      "step": 6404
+    },
+    {
+      "epoch": 1.2247824839850847,
+      "grad_norm": 4.928925807525293,
+      "learning_rate": 1e-06,
+      "loss": 0.314,
+      "step": 6405
+    },
+    {
+      "epoch": 1.22497370685534,
+      "grad_norm": 5.057217326012353,
+      "learning_rate": 1e-06,
+      "loss": 0.4812,
+      "step": 6406
+    },
+    {
+      "epoch": 1.225164929725595,
+      "grad_norm": 5.1906515452643776,
+      "learning_rate": 1e-06,
+      "loss": 0.4173,
+      "step": 6407
+    },
+    {
+      "epoch": 1.2253561525958505,
+      "grad_norm": 3.2938994227180682,
+      "learning_rate": 1e-06,
+      "loss": 0.2464,
+      "step": 6408
+    },
+    {
+      "epoch": 1.2255473754661057,
+      "grad_norm": 2.1787946506286806,
+      "learning_rate": 1e-06,
+      "loss": 0.098,
+      "step": 6409
+    },
+    {
+      "epoch": 1.2257385983363611,
+      "grad_norm": 5.495700542859582,
+      "learning_rate": 1e-06,
+      "loss": 0.5433,
+      "step": 6410
+    },
+    {
+      "epoch": 1.2259298212066163,
+      "grad_norm": 2.5111413650978256,
+      "learning_rate": 1e-06,
+      "loss": 0.1471,
+      "step": 6411
+    },
+    {
+      "epoch": 1.2261210440768715,
+      "grad_norm": 3.233086370756885,
+      "learning_rate": 1e-06,
+      "loss": 0.0634,
+      "step": 6412
+    },
+    {
+      "epoch": 1.226312266947127,
+      "grad_norm": 3.1292076775258804,
+      "learning_rate": 1e-06,
+      "loss": 0.1886,
+      "step": 6413
+    },
+    {
+      "epoch": 1.226503489817382,
+      "grad_norm": 2.412830025824552,
+      "learning_rate": 1e-06,
+      "loss": 0.0551,
+      "step": 6414
+    },
+    {
+      "epoch": 1.2266947126876375,
+      "grad_norm": 2.9175716267793144,
+      "learning_rate": 1e-06,
+      "loss": 0.0661,
+      "step": 6415
+    },
+    {
+      "epoch": 1.2268859355578927,
+      "grad_norm": 2.386411956198829,
+      "learning_rate": 1e-06,
+      "loss": 0.0421,
+      "step": 6416
+    },
+    {
+      "epoch": 1.227077158428148,
+      "grad_norm": 4.261784987455007,
+      "learning_rate": 1e-06,
+      "loss": 0.0417,
+      "step": 6417
+    },
+    {
+      "epoch": 1.2272683812984033,
+      "grad_norm": 7.76482990577232,
+      "learning_rate": 1e-06,
+      "loss": 0.4466,
+      "step": 6418
+    },
+    {
+      "epoch": 1.2274596041686585,
+      "grad_norm": 7.997040678084212,
+      "learning_rate": 1e-06,
+      "loss": 0.5149,
+      "step": 6419
+    },
+    {
+      "epoch": 1.227650827038914,
+      "grad_norm": 4.040213149840312,
+      "learning_rate": 1e-06,
+      "loss": 0.2486,
+      "step": 6420
+    },
+    {
+      "epoch": 1.2278420499091691,
+      "grad_norm": 3.6155849391923636,
+      "learning_rate": 1e-06,
+      "loss": 0.2708,
+      "step": 6421
+    },
+    {
+      "epoch": 1.2280332727794243,
+      "grad_norm": 2.937619632963259,
+      "learning_rate": 1e-06,
+      "loss": 0.1598,
+      "step": 6422
+    },
+    {
+      "epoch": 1.2282244956496797,
+      "grad_norm": 3.346694647936725,
+      "learning_rate": 1e-06,
+      "loss": 0.1229,
+      "step": 6423
+    },
+    {
+      "epoch": 1.228415718519935,
+      "grad_norm": 2.22565149777286,
+      "learning_rate": 1e-06,
+      "loss": 0.1245,
+      "step": 6424
+    },
+    {
+      "epoch": 1.2286069413901903,
+      "grad_norm": 4.91867106470077,
+      "learning_rate": 1e-06,
+      "loss": 0.2197,
+      "step": 6425
+    },
+    {
+      "epoch": 1.2287981642604455,
+      "grad_norm": 2.6463622117791465,
+      "learning_rate": 1e-06,
+      "loss": 0.148,
+      "step": 6426
+    },
+    {
+      "epoch": 1.2289893871307007,
+      "grad_norm": 3.308503096778669,
+      "learning_rate": 1e-06,
+      "loss": 0.1017,
+      "step": 6427
+    },
+    {
+      "epoch": 1.2291806100009561,
+      "grad_norm": 1.4450213164766883,
+      "learning_rate": 1e-06,
+      "loss": 0.0134,
+      "step": 6428
+    },
+    {
+      "epoch": 1.2293718328712113,
+      "grad_norm": 5.252942486825594,
+      "learning_rate": 1e-06,
+      "loss": 0.1687,
+      "step": 6429
+    },
+    {
+      "epoch": 1.2295630557414667,
+      "grad_norm": 16.99395644527967,
+      "learning_rate": 1e-06,
+      "loss": 0.1591,
+      "step": 6430
+    },
+    {
+      "epoch": 1.229754278611722,
+      "grad_norm": 4.280133401093048,
+      "learning_rate": 1e-06,
+      "loss": 0.3949,
+      "step": 6431
+    },
+    {
+      "epoch": 1.2299455014819771,
+      "grad_norm": 4.547740735879877,
+      "learning_rate": 1e-06,
+      "loss": 0.2617,
+      "step": 6432
+    },
+    {
+      "epoch": 1.2301367243522325,
+      "grad_norm": 2.782902515975296,
+      "learning_rate": 1e-06,
+      "loss": 0.2691,
+      "step": 6433
+    },
+    {
+      "epoch": 1.2303279472224877,
+      "grad_norm": 4.266941635887731,
+      "learning_rate": 1e-06,
+      "loss": 0.0973,
+      "step": 6434
+    },
+    {
+      "epoch": 1.2305191700927431,
+      "grad_norm": 4.144361659687072,
+      "learning_rate": 1e-06,
+      "loss": 0.3449,
+      "step": 6435
+    },
+    {
+      "epoch": 1.2307103929629983,
+      "grad_norm": 2.124651936467537,
+      "learning_rate": 1e-06,
+      "loss": 0.0601,
+      "step": 6436
+    },
+    {
+      "epoch": 1.2309016158332537,
+      "grad_norm": 2.0845789682313502,
+      "learning_rate": 1e-06,
+      "loss": 0.0361,
+      "step": 6437
+    },
+    {
+      "epoch": 1.231092838703509,
+      "grad_norm": 2.15855362061329,
+      "learning_rate": 1e-06,
+      "loss": 0.1443,
+      "step": 6438
+    },
+    {
+      "epoch": 1.2312840615737641,
+      "grad_norm": 9.682747542696026,
+      "learning_rate": 1e-06,
+      "loss": 0.0601,
+      "step": 6439
+    },
+    {
+      "epoch": 1.2314752844440195,
+      "grad_norm": 2.780988359467397,
+      "learning_rate": 1e-06,
+      "loss": 0.0473,
+      "step": 6440
+    },
+    {
+      "epoch": 1.2316665073142747,
+      "grad_norm": 3.7154054106816736,
+      "learning_rate": 1e-06,
+      "loss": 0.212,
+      "step": 6441
+    },
+    {
+      "epoch": 1.2318577301845302,
+      "grad_norm": 3.0033934791743606,
+      "learning_rate": 1e-06,
+      "loss": 0.0443,
+      "step": 6442
+    },
+    {
+      "epoch": 1.2320489530547853,
+      "grad_norm": 4.436442934313911,
+      "learning_rate": 1e-06,
+      "loss": 0.2291,
+      "step": 6443
+    },
+    {
+      "epoch": 1.2322401759250408,
+      "grad_norm": 7.139319235780862,
+      "learning_rate": 1e-06,
+      "loss": 0.6294,
+      "step": 6444
+    },
+    {
+      "epoch": 1.232431398795296,
+      "grad_norm": 3.0877270881445815,
+      "learning_rate": 1e-06,
+      "loss": 0.1036,
+      "step": 6445
+    },
+    {
+      "epoch": 1.2326226216655511,
+      "grad_norm": 3.0501114308998223,
+      "learning_rate": 1e-06,
+      "loss": 0.0621,
+      "step": 6446
+    },
+    {
+      "epoch": 1.2328138445358066,
+      "grad_norm": 2.7374564981705594,
+      "learning_rate": 1e-06,
+      "loss": 0.1461,
+      "step": 6447
+    },
+    {
+      "epoch": 1.2330050674060617,
+      "grad_norm": 3.5296727261649923,
+      "learning_rate": 1e-06,
+      "loss": 0.168,
+      "step": 6448
+    },
+    {
+      "epoch": 1.2331962902763172,
+      "grad_norm": 2.248375412095004,
+      "learning_rate": 1e-06,
+      "loss": 0.1135,
+      "step": 6449
+    },
+    {
+      "epoch": 1.2333875131465724,
+      "grad_norm": 3.84167672664183,
+      "learning_rate": 1e-06,
+      "loss": 0.1843,
+      "step": 6450
+    },
+    {
+      "epoch": 1.2335787360168275,
+      "grad_norm": 3.209314237648997,
+      "learning_rate": 1e-06,
+      "loss": 0.054,
+      "step": 6451
+    },
+    {
+      "epoch": 1.233769958887083,
+      "grad_norm": 4.6465668833311256,
+      "learning_rate": 1e-06,
+      "loss": 0.0498,
+      "step": 6452
+    },
+    {
+      "epoch": 1.2339611817573382,
+      "grad_norm": 3.2074456080807336,
+      "learning_rate": 1e-06,
+      "loss": 0.1112,
+      "step": 6453
+    },
+    {
+      "epoch": 1.2341524046275936,
+      "grad_norm": 1.889977903993811,
+      "learning_rate": 1e-06,
+      "loss": 0.0255,
+      "step": 6454
+    },
+    {
+      "epoch": 1.2343436274978488,
+      "grad_norm": 9.936209348921043,
+      "learning_rate": 1e-06,
+      "loss": 0.1436,
+      "step": 6455
+    },
+    {
+      "epoch": 1.234534850368104,
+      "grad_norm": 6.832673606970024,
+      "learning_rate": 1e-06,
+      "loss": 0.5468,
+      "step": 6456
+    },
+    {
+      "epoch": 1.2347260732383594,
+      "grad_norm": 5.013413176272767,
+      "learning_rate": 1e-06,
+      "loss": 0.2787,
+      "step": 6457
+    },
+    {
+      "epoch": 1.2349172961086146,
+      "grad_norm": 2.7936604530269316,
+      "learning_rate": 1e-06,
+      "loss": 0.1063,
+      "step": 6458
+    },
+    {
+      "epoch": 1.23510851897887,
+      "grad_norm": 2.8123775031646083,
+      "learning_rate": 1e-06,
+      "loss": 0.1677,
+      "step": 6459
+    },
+    {
+      "epoch": 1.2352997418491252,
+      "grad_norm": 2.677957671556706,
+      "learning_rate": 1e-06,
+      "loss": 0.1091,
+      "step": 6460
+    },
+    {
+      "epoch": 1.2354909647193804,
+      "grad_norm": 5.011020912116117,
+      "learning_rate": 1e-06,
+      "loss": 0.4431,
+      "step": 6461
+    },
+    {
+      "epoch": 1.2356821875896358,
+      "grad_norm": 3.379800843460318,
+      "learning_rate": 1e-06,
+      "loss": 0.2948,
+      "step": 6462
+    },
+    {
+      "epoch": 1.235873410459891,
+      "grad_norm": 3.010602494326545,
+      "learning_rate": 1e-06,
+      "loss": 0.055,
+      "step": 6463
+    },
+    {
+      "epoch": 1.2360646333301464,
+      "grad_norm": 6.158387098821862,
+      "learning_rate": 1e-06,
+      "loss": 0.0755,
+      "step": 6464
+    },
+    {
+      "epoch": 1.2362558562004016,
+      "grad_norm": 2.4427218135198796,
+      "learning_rate": 1e-06,
+      "loss": 0.0408,
+      "step": 6465
+    },
+    {
+      "epoch": 1.2364470790706568,
+      "grad_norm": 5.7159667199182405,
+      "learning_rate": 1e-06,
+      "loss": 0.0287,
+      "step": 6466
+    },
+    {
+      "epoch": 1.2366383019409122,
+      "grad_norm": 5.425904988689819,
+      "learning_rate": 1e-06,
+      "loss": 0.0707,
+      "step": 6467
+    },
+    {
+      "epoch": 1.2368295248111674,
+      "grad_norm": 6.745509879463193,
+      "learning_rate": 1e-06,
+      "loss": 0.3763,
+      "step": 6468
+    },
+    {
+      "epoch": 1.2370207476814228,
+      "grad_norm": 4.878261942044042,
+      "learning_rate": 1e-06,
+      "loss": 0.4395,
+      "step": 6469
+    },
+    {
+      "epoch": 1.237211970551678,
+      "grad_norm": 5.093956317116722,
+      "learning_rate": 1e-06,
+      "loss": 0.302,
+      "step": 6470
+    },
+    {
+      "epoch": 1.2374031934219332,
+      "grad_norm": 2.995255851707413,
+      "learning_rate": 1e-06,
+      "loss": 0.1478,
+      "step": 6471
+    },
+    {
+      "epoch": 1.2375944162921886,
+      "grad_norm": 3.803398148845016,
+      "learning_rate": 1e-06,
+      "loss": 0.3354,
+      "step": 6472
+    },
+    {
+      "epoch": 1.2377856391624438,
+      "grad_norm": 2.4465317291690813,
+      "learning_rate": 1e-06,
+      "loss": 0.1611,
+      "step": 6473
+    },
+    {
+      "epoch": 1.2379768620326992,
+      "grad_norm": 3.026767523285303,
+      "learning_rate": 1e-06,
+      "loss": 0.1746,
+      "step": 6474
+    },
+    {
+      "epoch": 1.2381680849029544,
+      "grad_norm": 6.290549913878288,
+      "learning_rate": 1e-06,
+      "loss": 0.1004,
+      "step": 6475
+    },
+    {
+      "epoch": 1.2383593077732096,
+      "grad_norm": 4.374610883574724,
+      "learning_rate": 1e-06,
+      "loss": 0.124,
+      "step": 6476
+    },
+    {
+      "epoch": 1.238550530643465,
+      "grad_norm": 3.6532580531733823,
+      "learning_rate": 1e-06,
+      "loss": 0.2187,
+      "step": 6477
+    },
+    {
+      "epoch": 1.2387417535137202,
+      "grad_norm": 4.260245819521336,
+      "learning_rate": 1e-06,
+      "loss": 0.089,
+      "step": 6478
+    },
+    {
+      "epoch": 1.2389329763839756,
+      "grad_norm": 4.29499936868754,
+      "learning_rate": 1e-06,
+      "loss": 0.079,
+      "step": 6479
+    },
+    {
+      "epoch": 1.2391241992542308,
+      "grad_norm": 4.9705406177389575,
+      "learning_rate": 1e-06,
+      "loss": 0.163,
+      "step": 6480
+    },
+    {
+      "epoch": 1.239315422124486,
+      "grad_norm": 4.472596120638394,
+      "learning_rate": 1e-06,
+      "loss": 0.1803,
+      "step": 6481
+    },
+    {
+      "epoch": 1.2395066449947414,
+      "grad_norm": 3.2169797899763393,
+      "learning_rate": 1e-06,
+      "loss": 0.2079,
+      "step": 6482
+    },
+    {
+      "epoch": 1.2396978678649966,
+      "grad_norm": 4.079664619349496,
+      "learning_rate": 1e-06,
+      "loss": 0.1282,
+      "step": 6483
+    },
+    {
+      "epoch": 1.239889090735252,
+      "grad_norm": 1.4510992553120168,
+      "learning_rate": 1e-06,
+      "loss": 0.0458,
+      "step": 6484
+    },
+    {
+      "epoch": 1.2400803136055072,
+      "grad_norm": 3.2447814124829613,
+      "learning_rate": 1e-06,
+      "loss": 0.2116,
+      "step": 6485
+    },
+    {
+      "epoch": 1.2402715364757624,
+      "grad_norm": 4.967540281811111,
+      "learning_rate": 1e-06,
+      "loss": 0.2643,
+      "step": 6486
+    },
+    {
+      "epoch": 1.2404627593460178,
+      "grad_norm": 4.446622900073055,
+      "learning_rate": 1e-06,
+      "loss": 0.099,
+      "step": 6487
+    },
+    {
+      "epoch": 1.240653982216273,
+      "grad_norm": 2.180484202271083,
+      "learning_rate": 1e-06,
+      "loss": 0.0356,
+      "step": 6488
+    },
+    {
+      "epoch": 1.2408452050865284,
+      "grad_norm": 2.128543927919336,
+      "learning_rate": 1e-06,
+      "loss": 0.0414,
+      "step": 6489
+    },
+    {
+      "epoch": 1.2410364279567836,
+      "grad_norm": 3.614223773819719,
+      "learning_rate": 1e-06,
+      "loss": 0.0332,
+      "step": 6490
+    },
+    {
+      "epoch": 1.241227650827039,
+      "grad_norm": 2.5546145180826274,
+      "learning_rate": 1e-06,
+      "loss": 0.0722,
+      "step": 6491
+    },
+    {
+      "epoch": 1.2414188736972942,
+      "grad_norm": 7.01666265685389,
+      "learning_rate": 1e-06,
+      "loss": 0.1586,
+      "step": 6492
+    },
+    {
+      "epoch": 1.2416100965675494,
+      "grad_norm": 6.808819817944401,
+      "learning_rate": 1e-06,
+      "loss": 0.3024,
+      "step": 6493
+    },
+    {
+      "epoch": 1.2418013194378048,
+      "grad_norm": 5.749358099891762,
+      "learning_rate": 1e-06,
+      "loss": 0.3554,
+      "step": 6494
+    },
+    {
+      "epoch": 1.24199254230806,
+      "grad_norm": 4.279817216125285,
+      "learning_rate": 1e-06,
+      "loss": 0.3917,
+      "step": 6495
+    },
+    {
+      "epoch": 1.2421837651783154,
+      "grad_norm": 1.7741899952475355,
+      "learning_rate": 1e-06,
+      "loss": 0.1148,
+      "step": 6496
+    },
+    {
+      "epoch": 1.2423749880485706,
+      "grad_norm": 3.4738634240465216,
+      "learning_rate": 1e-06,
+      "loss": 0.2501,
+      "step": 6497
+    },
+    {
+      "epoch": 1.2425662109188258,
+      "grad_norm": 4.527307097160734,
+      "learning_rate": 1e-06,
+      "loss": 0.3482,
+      "step": 6498
+    },
+    {
+      "epoch": 1.2427574337890812,
+      "grad_norm": 1.9956668165767664,
+      "learning_rate": 1e-06,
+      "loss": 0.0845,
+      "step": 6499
+    },
+    {
+      "epoch": 1.2429486566593364,
+      "grad_norm": 2.456162242923299,
+      "learning_rate": 1e-06,
+      "loss": 0.0986,
+      "step": 6500
+    },
+    {
+      "epoch": 1.2429486566593364,
+      "eval_runtime": 729.7111,
+      "eval_samples_per_second": 2.102,
+      "eval_steps_per_second": 0.526,
+      "step": 6500
+    },
+    {
+      "epoch": 1.2431398795295918,
+      "grad_norm": 3.097997504699372,
+      "learning_rate": 1e-06,
+      "loss": 0.096,
+      "step": 6501
+    },
+    {
+      "epoch": 1.243331102399847,
+      "grad_norm": 9.834691702777056,
+      "learning_rate": 1e-06,
+      "loss": 0.2294,
+      "step": 6502
+    },
+    {
+      "epoch": 1.2435223252701024,
+      "grad_norm": 2.5626391861494415,
+      "learning_rate": 1e-06,
+      "loss": 0.0338,
+      "step": 6503
+    },
+    {
+      "epoch": 1.2437135481403576,
+      "grad_norm": 1.5969712110825571,
+      "learning_rate": 1e-06,
+      "loss": 0.0188,
+      "step": 6504
+    },
+    {
+      "epoch": 1.2439047710106128,
+      "grad_norm": 9.15131211175202,
+      "learning_rate": 1e-06,
+      "loss": 0.1198,
+      "step": 6505
+    },
+    {
+      "epoch": 1.2440959938808682,
+      "grad_norm": 6.000396397529936,
+      "learning_rate": 1e-06,
+      "loss": 0.3501,
+      "step": 6506
+    },
+    {
+      "epoch": 1.2442872167511234,
+      "grad_norm": 4.937641576354834,
+      "learning_rate": 1e-06,
+      "loss": 0.2687,
+      "step": 6507
+    },
+    {
+      "epoch": 1.2444784396213788,
+      "grad_norm": 2.3039357139175447,
+      "learning_rate": 1e-06,
+      "loss": 0.2368,
+      "step": 6508
+    },
+    {
+      "epoch": 1.244669662491634,
+      "grad_norm": 3.8276084084235764,
+      "learning_rate": 1e-06,
+      "loss": 0.3902,
+      "step": 6509
+    },
+    {
+      "epoch": 1.2448608853618892,
+      "grad_norm": 3.5241690339370417,
+      "learning_rate": 1e-06,
+      "loss": 0.2661,
+      "step": 6510
+    },
+    {
+      "epoch": 1.2450521082321446,
+      "grad_norm": 4.450274495434725,
+      "learning_rate": 1e-06,
+      "loss": 0.2062,
+      "step": 6511
+    },
+    {
+      "epoch": 1.2452433311023998,
+      "grad_norm": 2.982841057538288,
+      "learning_rate": 1e-06,
+      "loss": 0.1274,
+      "step": 6512
+    },
+    {
+      "epoch": 1.2454345539726552,
+      "grad_norm": 2.4036661121729437,
+      "learning_rate": 1e-06,
+      "loss": 0.0763,
+      "step": 6513
+    },
+    {
+      "epoch": 1.2456257768429104,
+      "grad_norm": 2.16763781521903,
+      "learning_rate": 1e-06,
+      "loss": 0.1666,
+      "step": 6514
+    },
+    {
+      "epoch": 1.2458169997131656,
+      "grad_norm": 2.6084006311999386,
+      "learning_rate": 1e-06,
+      "loss": 0.1079,
+      "step": 6515
+    },
+    {
+      "epoch": 1.246008222583421,
+      "grad_norm": 3.1791924318508613,
+      "learning_rate": 1e-06,
+      "loss": 0.0557,
+      "step": 6516
+    },
+    {
+      "epoch": 1.2461994454536762,
+      "grad_norm": 4.87811590532314,
+      "learning_rate": 1e-06,
+      "loss": 0.1801,
+      "step": 6517
+    },
+    {
+      "epoch": 1.2463906683239316,
+      "grad_norm": 5.988294946066366,
+      "learning_rate": 1e-06,
+      "loss": 0.2305,
+      "step": 6518
+    },
+    {
+      "epoch": 1.2465818911941868,
+      "grad_norm": 5.340273970244458,
+      "learning_rate": 1e-06,
+      "loss": 0.5855,
+      "step": 6519
+    },
+    {
+      "epoch": 1.246773114064442,
+      "grad_norm": 6.077029092638395,
+      "learning_rate": 1e-06,
+      "loss": 0.4405,
+      "step": 6520
+    },
+    {
+      "epoch": 1.2469643369346974,
+      "grad_norm": 2.8938455322842787,
+      "learning_rate": 1e-06,
+      "loss": 0.1439,
+      "step": 6521
+    },
+    {
+      "epoch": 1.2471555598049526,
+      "grad_norm": 5.842607845540173,
+      "learning_rate": 1e-06,
+      "loss": 0.4677,
+      "step": 6522
+    },
+    {
+      "epoch": 1.247346782675208,
+      "grad_norm": 1.5322020548798874,
+      "learning_rate": 1e-06,
+      "loss": 0.0533,
+      "step": 6523
+    },
+    {
+      "epoch": 1.2475380055454632,
+      "grad_norm": 2.881238388733213,
+      "learning_rate": 1e-06,
+      "loss": 0.0639,
+      "step": 6524
+    },
+    {
+      "epoch": 1.2477292284157184,
+      "grad_norm": 1.8431546818052984,
+      "learning_rate": 1e-06,
+      "loss": 0.0467,
+      "step": 6525
+    },
+    {
+      "epoch": 1.2479204512859738,
+      "grad_norm": 1.9483189770855176,
+      "learning_rate": 1e-06,
+      "loss": 0.0289,
+      "step": 6526
+    },
+    {
+      "epoch": 1.248111674156229,
+      "grad_norm": 4.235013681364873,
+      "learning_rate": 1e-06,
+      "loss": 0.1979,
+      "step": 6527
+    },
+    {
+      "epoch": 1.2483028970264844,
+      "grad_norm": 2.270956146932182,
+      "learning_rate": 1e-06,
+      "loss": 0.0466,
+      "step": 6528
+    },
+    {
+      "epoch": 1.2484941198967396,
+      "grad_norm": 2.3143601024621057,
+      "learning_rate": 1e-06,
+      "loss": 0.0343,
+      "step": 6529
+    },
+    {
+      "epoch": 1.2486853427669948,
+      "grad_norm": 6.6556725766454115,
+      "learning_rate": 1e-06,
+      "loss": 0.2922,
+      "step": 6530
+    },
+    {
+      "epoch": 1.2488765656372502,
+      "grad_norm": 4.424218732755386,
+      "learning_rate": 1e-06,
+      "loss": 0.4405,
+      "step": 6531
+    },
+    {
+      "epoch": 1.2490677885075054,
+      "grad_norm": 5.121520745511445,
+      "learning_rate": 1e-06,
+      "loss": 0.4694,
+      "step": 6532
+    },
+    {
+      "epoch": 1.2492590113777609,
+      "grad_norm": 2.6652271935417775,
+      "learning_rate": 1e-06,
+      "loss": 0.1866,
+      "step": 6533
+    },
+    {
+      "epoch": 1.249450234248016,
+      "grad_norm": 3.6387655724231873,
+      "learning_rate": 1e-06,
+      "loss": 0.0563,
+      "step": 6534
+    },
+    {
+      "epoch": 1.2496414571182712,
+      "grad_norm": 4.26522344578575,
+      "learning_rate": 1e-06,
+      "loss": 0.2478,
+      "step": 6535
+    },
+    {
+      "epoch": 1.2498326799885267,
+      "grad_norm": 4.148144382962027,
+      "learning_rate": 1e-06,
+      "loss": 0.2854,
+      "step": 6536
+    },
+    {
+      "epoch": 1.2500239028587818,
+      "grad_norm": 2.5702781790902933,
+      "learning_rate": 1e-06,
+      "loss": 0.0549,
+      "step": 6537
+    },
+    {
+      "epoch": 1.2502151257290373,
+      "grad_norm": 6.515015790895309,
+      "learning_rate": 1e-06,
+      "loss": 0.0468,
+      "step": 6538
+    },
+    {
+      "epoch": 1.2504063485992924,
+      "grad_norm": 3.285884962328526,
+      "learning_rate": 1e-06,
+      "loss": 0.0409,
+      "step": 6539
+    },
+    {
+      "epoch": 1.2505975714695476,
+      "grad_norm": 1.7964981347320412,
+      "learning_rate": 1e-06,
+      "loss": 0.0404,
+      "step": 6540
+    },
+    {
+      "epoch": 1.250788794339803,
+      "grad_norm": 3.302910942713352,
+      "learning_rate": 1e-06,
+      "loss": 0.0605,
+      "step": 6541
+    },
+    {
+      "epoch": 1.2509800172100582,
+      "grad_norm": 3.566334049960576,
+      "learning_rate": 1e-06,
+      "loss": 0.0746,
+      "step": 6542
+    },
+    {
+      "epoch": 1.2511712400803137,
+      "grad_norm": 6.360872359517385,
+      "learning_rate": 1e-06,
+      "loss": 0.2226,
+      "step": 6543
+    },
+    {
+      "epoch": 1.2513624629505689,
+      "grad_norm": 6.06470818958379,
+      "learning_rate": 1e-06,
+      "loss": 0.4932,
+      "step": 6544
+    },
+    {
+      "epoch": 1.251553685820824,
+      "grad_norm": 4.94855064212155,
+      "learning_rate": 1e-06,
+      "loss": 0.2584,
+      "step": 6545
+    },
+    {
+      "epoch": 1.2517449086910795,
+      "grad_norm": 3.086767003081223,
+      "learning_rate": 1e-06,
+      "loss": 0.1447,
+      "step": 6546
+    },
+    {
+      "epoch": 1.2519361315613347,
+      "grad_norm": 6.342540907616987,
+      "learning_rate": 1e-06,
+      "loss": 0.3622,
+      "step": 6547
+    },
+    {
+      "epoch": 1.25212735443159,
+      "grad_norm": 3.048166011294278,
+      "learning_rate": 1e-06,
+      "loss": 0.2135,
+      "step": 6548
+    },
+    {
+      "epoch": 1.2523185773018453,
+      "grad_norm": 3.9374241897308866,
+      "learning_rate": 1e-06,
+      "loss": 0.1358,
+      "step": 6549
+    },
+    {
+      "epoch": 1.2525098001721005,
+      "grad_norm": 2.2633051077309823,
+      "learning_rate": 1e-06,
+      "loss": 0.1207,
+      "step": 6550
+    },
+    {
+      "epoch": 1.2527010230423559,
+      "grad_norm": 3.2319787035703116,
+      "learning_rate": 1e-06,
+      "loss": 0.1634,
+      "step": 6551
+    },
+    {
+      "epoch": 1.2528922459126113,
+      "grad_norm": 1.7583243070012167,
+      "learning_rate": 1e-06,
+      "loss": 0.0771,
+      "step": 6552
+    },
+    {
+      "epoch": 1.2530834687828665,
+      "grad_norm": 3.099585837723724,
+      "learning_rate": 1e-06,
+      "loss": 0.0438,
+      "step": 6553
+    },
+    {
+      "epoch": 1.2532746916531217,
+      "grad_norm": 4.284141065581268,
+      "learning_rate": 1e-06,
+      "loss": 0.0495,
+      "step": 6554
+    },
+    {
+      "epoch": 1.253465914523377,
+      "grad_norm": 5.171898233874361,
+      "learning_rate": 1e-06,
+      "loss": 0.1754,
+      "step": 6555
+    },
+    {
+      "epoch": 1.2536571373936323,
+      "grad_norm": 8.95116569557118,
+      "learning_rate": 1e-06,
+      "loss": 0.5053,
+      "step": 6556
+    },
+    {
+      "epoch": 1.2538483602638877,
+      "grad_norm": 3.816556679641482,
+      "learning_rate": 1e-06,
+      "loss": 0.3305,
+      "step": 6557
+    },
+    {
+      "epoch": 1.2540395831341429,
+      "grad_norm": 2.214029044556087,
+      "learning_rate": 1e-06,
+      "loss": 0.0886,
+      "step": 6558
+    },
+    {
+      "epoch": 1.254230806004398,
+      "grad_norm": 3.1043776028712253,
+      "learning_rate": 1e-06,
+      "loss": 0.1522,
+      "step": 6559
+    },
+    {
+      "epoch": 1.2544220288746535,
+      "grad_norm": 3.839115605435316,
+      "learning_rate": 1e-06,
+      "loss": 0.3895,
+      "step": 6560
+    },
+    {
+      "epoch": 1.2546132517449087,
+      "grad_norm": 5.120806024704643,
+      "learning_rate": 1e-06,
+      "loss": 0.43,
+      "step": 6561
+    },
+    {
+      "epoch": 1.254804474615164,
+      "grad_norm": 4.008760633844083,
+      "learning_rate": 1e-06,
+      "loss": 0.1149,
+      "step": 6562
+    },
+    {
+      "epoch": 1.2549956974854193,
+      "grad_norm": 3.4790086691305118,
+      "learning_rate": 1e-06,
+      "loss": 0.0674,
+      "step": 6563
+    },
+    {
+      "epoch": 1.2551869203556745,
+      "grad_norm": 2.443151231860595,
+      "learning_rate": 1e-06,
+      "loss": 0.0763,
+      "step": 6564
+    },
+    {
+      "epoch": 1.2553781432259299,
+      "grad_norm": 3.3635859157612606,
+      "learning_rate": 1e-06,
+      "loss": 0.2028,
+      "step": 6565
+    },
+    {
+      "epoch": 1.255569366096185,
+      "grad_norm": 2.726273835649497,
+      "learning_rate": 1e-06,
+      "loss": 0.0534,
+      "step": 6566
+    },
+    {
+      "epoch": 1.2557605889664405,
+      "grad_norm": 10.073433001176868,
+      "learning_rate": 1e-06,
+      "loss": 0.1196,
+      "step": 6567
+    },
+    {
+      "epoch": 1.2559518118366957,
+      "grad_norm": 5.687889001978513,
+      "learning_rate": 1e-06,
+      "loss": 0.1211,
+      "step": 6568
+    },
+    {
+      "epoch": 1.2561430347069509,
+      "grad_norm": 4.6617288805770185,
+      "learning_rate": 1e-06,
+      "loss": 0.4689,
+      "step": 6569
+    },
+    {
+      "epoch": 1.2563342575772063,
+      "grad_norm": 2.446969442871264,
+      "learning_rate": 1e-06,
+      "loss": 0.1921,
+      "step": 6570
+    },
+    {
+      "epoch": 1.2565254804474615,
+      "grad_norm": 3.9933591791716547,
+      "learning_rate": 1e-06,
+      "loss": 0.2344,
+      "step": 6571
+    },
+    {
+      "epoch": 1.256716703317717,
+      "grad_norm": 3.6734405426700767,
+      "learning_rate": 1e-06,
+      "loss": 0.2106,
+      "step": 6572
+    },
+    {
+      "epoch": 1.256907926187972,
+      "grad_norm": 3.381061479597909,
+      "learning_rate": 1e-06,
+      "loss": 0.2114,
+      "step": 6573
+    },
+    {
+      "epoch": 1.2570991490582273,
+      "grad_norm": 2.8284923673152185,
+      "learning_rate": 1e-06,
+      "loss": 0.0928,
+      "step": 6574
+    },
+    {
+      "epoch": 1.2572903719284827,
+      "grad_norm": 6.217626115465512,
+      "learning_rate": 1e-06,
+      "loss": 0.1963,
+      "step": 6575
+    },
+    {
+      "epoch": 1.257481594798738,
+      "grad_norm": 2.932428567964338,
+      "learning_rate": 1e-06,
+      "loss": 0.0655,
+      "step": 6576
+    },
+    {
+      "epoch": 1.2576728176689933,
+      "grad_norm": 3.6323944179368097,
+      "learning_rate": 1e-06,
+      "loss": 0.0684,
+      "step": 6577
+    },
+    {
+      "epoch": 1.2578640405392485,
+      "grad_norm": 3.8119994288280976,
+      "learning_rate": 1e-06,
+      "loss": 0.0985,
+      "step": 6578
+    },
+    {
+      "epoch": 1.2580552634095037,
+      "grad_norm": 5.57826929413336,
+      "learning_rate": 1e-06,
+      "loss": 0.0376,
+      "step": 6579
+    },
+    {
+      "epoch": 1.258246486279759,
+      "grad_norm": 5.254735718096678,
+      "learning_rate": 1e-06,
+      "loss": 0.1495,
+      "step": 6580
+    },
+    {
+      "epoch": 1.2584377091500143,
+      "grad_norm": 9.547686503332521,
+      "learning_rate": 1e-06,
+      "loss": 0.2302,
+      "step": 6581
+    },
+    {
+      "epoch": 1.2586289320202697,
+      "grad_norm": 5.44756289245798,
+      "learning_rate": 1e-06,
+      "loss": 0.2727,
+      "step": 6582
+    },
+    {
+      "epoch": 1.258820154890525,
+      "grad_norm": 2.790303784857841,
+      "learning_rate": 1e-06,
+      "loss": 0.1562,
+      "step": 6583
+    },
+    {
+      "epoch": 1.25901137776078,
+      "grad_norm": 3.281233723917966,
+      "learning_rate": 1e-06,
+      "loss": 0.1508,
+      "step": 6584
+    },
+    {
+      "epoch": 1.2592026006310355,
+      "grad_norm": 2.6600696700874655,
+      "learning_rate": 1e-06,
+      "loss": 0.1839,
+      "step": 6585
+    },
+    {
+      "epoch": 1.2593938235012907,
+      "grad_norm": 3.3366163934305035,
+      "learning_rate": 1e-06,
+      "loss": 0.1905,
+      "step": 6586
+    },
+    {
+      "epoch": 1.2595850463715461,
+      "grad_norm": 4.187982332150447,
+      "learning_rate": 1e-06,
+      "loss": 0.1314,
+      "step": 6587
+    },
+    {
+      "epoch": 1.2597762692418013,
+      "grad_norm": 2.863283414956975,
+      "learning_rate": 1e-06,
+      "loss": 0.0477,
+      "step": 6588
+    },
+    {
+      "epoch": 1.2599674921120565,
+      "grad_norm": 2.065843126838781,
+      "learning_rate": 1e-06,
+      "loss": 0.0296,
+      "step": 6589
+    },
+    {
+      "epoch": 1.260158714982312,
+      "grad_norm": 2.981894215969838,
+      "learning_rate": 1e-06,
+      "loss": 0.0832,
+      "step": 6590
+    },
+    {
+      "epoch": 1.260349937852567,
+      "grad_norm": 2.223646643875791,
+      "learning_rate": 1e-06,
+      "loss": 0.0339,
+      "step": 6591
+    },
+    {
+      "epoch": 1.2605411607228225,
+      "grad_norm": 5.76262439916227,
+      "learning_rate": 1e-06,
+      "loss": 0.0924,
+      "step": 6592
+    },
+    {
+      "epoch": 1.2607323835930777,
+      "grad_norm": 4.409032281416109,
+      "learning_rate": 1e-06,
+      "loss": 0.1954,
+      "step": 6593
+    },
+    {
+      "epoch": 1.260923606463333,
+      "grad_norm": 6.394367791511918,
+      "learning_rate": 1e-06,
+      "loss": 0.734,
+      "step": 6594
+    },
+    {
+      "epoch": 1.2611148293335883,
+      "grad_norm": 4.098164515069902,
+      "learning_rate": 1e-06,
+      "loss": 0.4214,
+      "step": 6595
+    },
+    {
+      "epoch": 1.2613060522038435,
+      "grad_norm": 4.517414306098064,
+      "learning_rate": 1e-06,
+      "loss": 0.1235,
+      "step": 6596
+    },
+    {
+      "epoch": 1.261497275074099,
+      "grad_norm": 4.020087349165672,
+      "learning_rate": 1e-06,
+      "loss": 0.3804,
+      "step": 6597
+    },
+    {
+      "epoch": 1.2616884979443541,
+      "grad_norm": 3.8398396520117375,
+      "learning_rate": 1e-06,
+      "loss": 0.2374,
+      "step": 6598
+    },
+    {
+      "epoch": 1.2618797208146093,
+      "grad_norm": 2.7368059088363577,
+      "learning_rate": 1e-06,
+      "loss": 0.0448,
+      "step": 6599
+    },
+    {
+      "epoch": 1.2620709436848647,
+      "grad_norm": 3.772618554377839,
+      "learning_rate": 1e-06,
+      "loss": 0.1298,
+      "step": 6600
+    },
+    {
+      "epoch": 1.26226216655512,
+      "grad_norm": 2.435842194812415,
+      "learning_rate": 1e-06,
+      "loss": 0.0521,
+      "step": 6601
+    },
+    {
+      "epoch": 1.2624533894253753,
+      "grad_norm": 3.7864117661052132,
+      "learning_rate": 1e-06,
+      "loss": 0.1037,
+      "step": 6602
+    },
+    {
+      "epoch": 1.2626446122956305,
+      "grad_norm": 3.084907301834706,
+      "learning_rate": 1e-06,
+      "loss": 0.028,
+      "step": 6603
+    },
+    {
+      "epoch": 1.2628358351658857,
+      "grad_norm": 4.290076964428234,
+      "learning_rate": 1e-06,
+      "loss": 0.08,
+      "step": 6604
+    },
+    {
+      "epoch": 1.2630270580361411,
+      "grad_norm": 3.312195098087951,
+      "learning_rate": 1e-06,
+      "loss": 0.0729,
+      "step": 6605
+    },
+    {
+      "epoch": 1.2632182809063965,
+      "grad_norm": 4.976184201796099,
+      "learning_rate": 1e-06,
+      "loss": 0.2319,
+      "step": 6606
+    },
+    {
+      "epoch": 1.2634095037766517,
+      "grad_norm": 3.529343892361139,
+      "learning_rate": 1e-06,
+      "loss": 0.3152,
+      "step": 6607
+    },
+    {
+      "epoch": 1.263600726646907,
+      "grad_norm": 2.9292942444397445,
+      "learning_rate": 1e-06,
+      "loss": 0.1298,
+      "step": 6608
+    },
+    {
+      "epoch": 1.2637919495171621,
+      "grad_norm": 1.97810478501439,
+      "learning_rate": 1e-06,
+      "loss": 0.0978,
+      "step": 6609
+    },
+    {
+      "epoch": 1.2639831723874175,
+      "grad_norm": 11.778837488192211,
+      "learning_rate": 1e-06,
+      "loss": 0.2025,
+      "step": 6610
+    },
+    {
+      "epoch": 1.264174395257673,
+      "grad_norm": 3.5355993172365356,
+      "learning_rate": 1e-06,
+      "loss": 0.2535,
+      "step": 6611
+    },
+    {
+      "epoch": 1.2643656181279281,
+      "grad_norm": 3.7858665595057994,
+      "learning_rate": 1e-06,
+      "loss": 0.2491,
+      "step": 6612
+    },
+    {
+      "epoch": 1.2645568409981833,
+      "grad_norm": 2.6401231277536485,
+      "learning_rate": 1e-06,
+      "loss": 0.1206,
+      "step": 6613
+    },
+    {
+      "epoch": 1.2647480638684387,
+      "grad_norm": 3.402142909677466,
+      "learning_rate": 1e-06,
+      "loss": 0.1674,
+      "step": 6614
+    },
+    {
+      "epoch": 1.264939286738694,
+      "grad_norm": 5.01280232799633,
+      "learning_rate": 1e-06,
+      "loss": 0.1039,
+      "step": 6615
+    },
+    {
+      "epoch": 1.2651305096089493,
+      "grad_norm": 2.930393469628408,
+      "learning_rate": 1e-06,
+      "loss": 0.0479,
+      "step": 6616
+    },
+    {
+      "epoch": 1.2653217324792045,
+      "grad_norm": 6.499957157874136,
+      "learning_rate": 1e-06,
+      "loss": 0.1088,
+      "step": 6617
+    },
+    {
+      "epoch": 1.2655129553494597,
+      "grad_norm": 6.7469687543241355,
+      "learning_rate": 1e-06,
+      "loss": 0.339,
+      "step": 6618
+    },
+    {
+      "epoch": 1.2657041782197151,
+      "grad_norm": 4.185100722109232,
+      "learning_rate": 1e-06,
+      "loss": 0.3834,
+      "step": 6619
+    },
+    {
+      "epoch": 1.2658954010899703,
+      "grad_norm": 3.2166803613348143,
+      "learning_rate": 1e-06,
+      "loss": 0.2616,
+      "step": 6620
+    },
+    {
+      "epoch": 1.2660866239602258,
+      "grad_norm": 3.991665501816698,
+      "learning_rate": 1e-06,
+      "loss": 0.2428,
+      "step": 6621
+    },
+    {
+      "epoch": 1.266277846830481,
+      "grad_norm": 2.286710292225775,
+      "learning_rate": 1e-06,
+      "loss": 0.1798,
+      "step": 6622
+    },
+    {
+      "epoch": 1.2664690697007361,
+      "grad_norm": 3.2876462084381752,
+      "learning_rate": 1e-06,
+      "loss": 0.1552,
+      "step": 6623
+    },
+    {
+      "epoch": 1.2666602925709916,
+      "grad_norm": 1.9473960167054176,
+      "learning_rate": 1e-06,
+      "loss": 0.1364,
+      "step": 6624
+    },
+    {
+      "epoch": 1.2668515154412467,
+      "grad_norm": 1.6740816474063076,
+      "learning_rate": 1e-06,
+      "loss": 0.029,
+      "step": 6625
+    },
+    {
+      "epoch": 1.2670427383115022,
+      "grad_norm": 0.9944781796423947,
+      "learning_rate": 1e-06,
+      "loss": 0.0214,
+      "step": 6626
+    },
+    {
+      "epoch": 1.2672339611817574,
+      "grad_norm": 3.0879062215470285,
+      "learning_rate": 1e-06,
+      "loss": 0.0511,
+      "step": 6627
+    },
+    {
+      "epoch": 1.2674251840520125,
+      "grad_norm": 2.830023534023171,
+      "learning_rate": 1e-06,
+      "loss": 0.0472,
+      "step": 6628
+    },
+    {
+      "epoch": 1.267616406922268,
+      "grad_norm": 2.731769213874465,
+      "learning_rate": 1e-06,
+      "loss": 0.0503,
+      "step": 6629
+    },
+    {
+      "epoch": 1.2678076297925231,
+      "grad_norm": 9.385633617833328,
+      "learning_rate": 1e-06,
+      "loss": 0.1401,
+      "step": 6630
+    },
+    {
+      "epoch": 1.2679988526627786,
+      "grad_norm": 5.149764727801551,
+      "learning_rate": 1e-06,
+      "loss": 0.4489,
+      "step": 6631
+    },
+    {
+      "epoch": 1.2681900755330338,
+      "grad_norm": 2.0995805321420056,
+      "learning_rate": 1e-06,
+      "loss": 0.2049,
+      "step": 6632
+    },
+    {
+      "epoch": 1.268381298403289,
+      "grad_norm": 5.126357294246702,
+      "learning_rate": 1e-06,
+      "loss": 0.1097,
+      "step": 6633
+    },
+    {
+      "epoch": 1.2685725212735444,
+      "grad_norm": 3.0558622398982322,
+      "learning_rate": 1e-06,
+      "loss": 0.0946,
+      "step": 6634
+    },
+    {
+      "epoch": 1.2687637441437996,
+      "grad_norm": 3.099110131058777,
+      "learning_rate": 1e-06,
+      "loss": 0.0865,
+      "step": 6635
+    },
+    {
+      "epoch": 1.268954967014055,
+      "grad_norm": 2.238650732517998,
+      "learning_rate": 1e-06,
+      "loss": 0.1025,
+      "step": 6636
+    },
+    {
+      "epoch": 1.2691461898843102,
+      "grad_norm": 2.067700159423456,
+      "learning_rate": 1e-06,
+      "loss": 0.032,
+      "step": 6637
+    },
+    {
+      "epoch": 1.2693374127545654,
+      "grad_norm": 2.2266939894014453,
+      "learning_rate": 1e-06,
+      "loss": 0.0444,
+      "step": 6638
+    },
+    {
+      "epoch": 1.2695286356248208,
+      "grad_norm": 5.436588254761127,
+      "learning_rate": 1e-06,
+      "loss": 0.1622,
+      "step": 6639
+    },
+    {
+      "epoch": 1.269719858495076,
+      "grad_norm": 2.199460167109848,
+      "learning_rate": 1e-06,
+      "loss": 0.0406,
+      "step": 6640
+    },
+    {
+      "epoch": 1.2699110813653314,
+      "grad_norm": 9.589346107887412,
+      "learning_rate": 1e-06,
+      "loss": 0.2941,
+      "step": 6641
+    },
+    {
+      "epoch": 1.2701023042355866,
+      "grad_norm": 4.181584293247965,
+      "learning_rate": 1e-06,
+      "loss": 0.0621,
+      "step": 6642
+    },
+    {
+      "epoch": 1.2702935271058418,
+      "grad_norm": 6.384949343965071,
+      "learning_rate": 1e-06,
+      "loss": 0.3362,
+      "step": 6643
+    },
+    {
+      "epoch": 1.2704847499760972,
+      "grad_norm": 4.765570968571777,
+      "learning_rate": 1e-06,
+      "loss": 0.3151,
+      "step": 6644
+    },
+    {
+      "epoch": 1.2706759728463524,
+      "grad_norm": 4.278543326724321,
+      "learning_rate": 1e-06,
+      "loss": 0.3371,
+      "step": 6645
+    },
+    {
+      "epoch": 1.2708671957166078,
+      "grad_norm": 3.9619701476496134,
+      "learning_rate": 1e-06,
+      "loss": 0.3231,
+      "step": 6646
+    },
+    {
+      "epoch": 1.271058418586863,
+      "grad_norm": 1.61305083691492,
+      "learning_rate": 1e-06,
+      "loss": 0.1107,
+      "step": 6647
+    },
+    {
+      "epoch": 1.2712496414571182,
+      "grad_norm": 5.711087789788893,
+      "learning_rate": 1e-06,
+      "loss": 0.1099,
+      "step": 6648
+    },
+    {
+      "epoch": 1.2714408643273736,
+      "grad_norm": 2.5801568549920186,
+      "learning_rate": 1e-06,
+      "loss": 0.0968,
+      "step": 6649
+    },
+    {
+      "epoch": 1.2716320871976288,
+      "grad_norm": 4.237985570763313,
+      "learning_rate": 1e-06,
+      "loss": 0.1537,
+      "step": 6650
+    },
+    {
+      "epoch": 1.2718233100678842,
+      "grad_norm": 3.1077390421398845,
+      "learning_rate": 1e-06,
+      "loss": 0.0729,
+      "step": 6651
+    },
+    {
+      "epoch": 1.2720145329381394,
+      "grad_norm": 2.6590723638769647,
+      "learning_rate": 1e-06,
+      "loss": 0.0315,
+      "step": 6652
+    },
+    {
+      "epoch": 1.2722057558083946,
+      "grad_norm": 2.3106173506990113,
+      "learning_rate": 1e-06,
+      "loss": 0.0503,
+      "step": 6653
+    },
+    {
+      "epoch": 1.27239697867865,
+      "grad_norm": 1.8422558436120522,
+      "learning_rate": 1e-06,
+      "loss": 0.033,
+      "step": 6654
+    },
+    {
+      "epoch": 1.2725882015489052,
+      "grad_norm": 3.1337752159147976,
+      "learning_rate": 1e-06,
+      "loss": 0.1705,
+      "step": 6655
+    },
+    {
+      "epoch": 1.2727794244191606,
+      "grad_norm": 6.7402561643917585,
+      "learning_rate": 1e-06,
+      "loss": 0.5971,
+      "step": 6656
+    },
+    {
+      "epoch": 1.2729706472894158,
+      "grad_norm": 3.3986847184685662,
+      "learning_rate": 1e-06,
+      "loss": 0.1698,
+      "step": 6657
+    },
+    {
+      "epoch": 1.273161870159671,
+      "grad_norm": 2.987188321378954,
+      "learning_rate": 1e-06,
+      "loss": 0.0864,
+      "step": 6658
+    },
+    {
+      "epoch": 1.2733530930299264,
+      "grad_norm": 3.6186343897666364,
+      "learning_rate": 1e-06,
+      "loss": 0.2991,
+      "step": 6659
+    },
+    {
+      "epoch": 1.2735443159001818,
+      "grad_norm": 5.381380596636378,
+      "learning_rate": 1e-06,
+      "loss": 0.379,
+      "step": 6660
+    },
+    {
+      "epoch": 1.273735538770437,
+      "grad_norm": 2.34726695357317,
+      "learning_rate": 1e-06,
+      "loss": 0.1259,
+      "step": 6661
+    },
+    {
+      "epoch": 1.2739267616406922,
+      "grad_norm": 3.855740337603842,
+      "learning_rate": 1e-06,
+      "loss": 0.0709,
+      "step": 6662
+    },
+    {
+      "epoch": 1.2741179845109474,
+      "grad_norm": 2.1166743821531635,
+      "learning_rate": 1e-06,
+      "loss": 0.0608,
+      "step": 6663
+    },
+    {
+      "epoch": 1.2743092073812028,
+      "grad_norm": 2.8543408380163484,
+      "learning_rate": 1e-06,
+      "loss": 0.0439,
+      "step": 6664
+    },
+    {
+      "epoch": 1.2745004302514582,
+      "grad_norm": 1.79681150904729,
+      "learning_rate": 1e-06,
+      "loss": 0.0447,
+      "step": 6665
+    },
+    {
+      "epoch": 1.2746916531217134,
+      "grad_norm": 2.760122569030103,
+      "learning_rate": 1e-06,
+      "loss": 0.0741,
+      "step": 6666
+    },
+    {
+      "epoch": 1.2748828759919686,
+      "grad_norm": 3.105655458223805,
+      "learning_rate": 1e-06,
+      "loss": 0.0948,
+      "step": 6667
+    },
+    {
+      "epoch": 1.275074098862224,
+      "grad_norm": 7.522911425743366,
+      "learning_rate": 1e-06,
+      "loss": 0.2378,
+      "step": 6668
+    },
+    {
+      "epoch": 1.2752653217324792,
+      "grad_norm": 3.122599633538458,
+      "learning_rate": 1e-06,
+      "loss": 0.1197,
+      "step": 6669
+    },
+    {
+      "epoch": 1.2754565446027346,
+      "grad_norm": 4.000340924039465,
+      "learning_rate": 1e-06,
+      "loss": 0.1227,
+      "step": 6670
+    },
+    {
+      "epoch": 1.2756477674729898,
+      "grad_norm": 4.538583438036965,
+      "learning_rate": 1e-06,
+      "loss": 0.101,
+      "step": 6671
+    },
+    {
+      "epoch": 1.275838990343245,
+      "grad_norm": 4.876589198163159,
+      "learning_rate": 1e-06,
+      "loss": 0.291,
+      "step": 6672
+    },
+    {
+      "epoch": 1.2760302132135004,
+      "grad_norm": 1.8196322762339319,
+      "learning_rate": 1e-06,
+      "loss": 0.0917,
+      "step": 6673
+    },
+    {
+      "epoch": 1.2762214360837556,
+      "grad_norm": 3.4761677378578315,
+      "learning_rate": 1e-06,
+      "loss": 0.1292,
+      "step": 6674
+    },
+    {
+      "epoch": 1.276412658954011,
+      "grad_norm": 2.7807284091096514,
+      "learning_rate": 1e-06,
+      "loss": 0.1065,
+      "step": 6675
+    },
+    {
+      "epoch": 1.2766038818242662,
+      "grad_norm": 2.6852853210700345,
+      "learning_rate": 1e-06,
+      "loss": 0.0956,
+      "step": 6676
+    },
+    {
+      "epoch": 1.2767951046945214,
+      "grad_norm": 6.010962961141487,
+      "learning_rate": 1e-06,
+      "loss": 0.4787,
+      "step": 6677
+    },
+    {
+      "epoch": 1.2769863275647768,
+      "grad_norm": 1.9747313811721976,
+      "learning_rate": 1e-06,
+      "loss": 0.042,
+      "step": 6678
+    },
+    {
+      "epoch": 1.277177550435032,
+      "grad_norm": 7.790782558716247,
+      "learning_rate": 1e-06,
+      "loss": 0.1073,
+      "step": 6679
+    },
+    {
+      "epoch": 1.2773687733052874,
+      "grad_norm": 5.790670112479427,
+      "learning_rate": 1e-06,
+      "loss": 0.0658,
+      "step": 6680
+    },
+    {
+      "epoch": 1.2775599961755426,
+      "grad_norm": 4.446509657627652,
+      "learning_rate": 1e-06,
+      "loss": 0.1278,
+      "step": 6681
+    },
+    {
+      "epoch": 1.2777512190457978,
+      "grad_norm": 5.239826699805456,
+      "learning_rate": 1e-06,
+      "loss": 0.4149,
+      "step": 6682
+    },
+    {
+      "epoch": 1.2779424419160532,
+      "grad_norm": 3.5799346745247664,
+      "learning_rate": 1e-06,
+      "loss": 0.0723,
+      "step": 6683
+    },
+    {
+      "epoch": 1.2781336647863084,
+      "grad_norm": 3.2168775130299707,
+      "learning_rate": 1e-06,
+      "loss": 0.1602,
+      "step": 6684
+    },
+    {
+      "epoch": 1.2783248876565638,
+      "grad_norm": 3.2976673493857125,
+      "learning_rate": 1e-06,
+      "loss": 0.2449,
+      "step": 6685
+    },
+    {
+      "epoch": 1.278516110526819,
+      "grad_norm": 2.0943288429977525,
+      "learning_rate": 1e-06,
+      "loss": 0.0945,
+      "step": 6686
+    },
+    {
+      "epoch": 1.2787073333970742,
+      "grad_norm": 1.004992302063265,
+      "learning_rate": 1e-06,
+      "loss": 0.0078,
+      "step": 6687
+    },
+    {
+      "epoch": 1.2788985562673296,
+      "grad_norm": 4.415037010443839,
+      "learning_rate": 1e-06,
+      "loss": 0.0878,
+      "step": 6688
+    },
+    {
+      "epoch": 1.2790897791375848,
+      "grad_norm": 2.127434401974498,
+      "learning_rate": 1e-06,
+      "loss": 0.0429,
+      "step": 6689
+    },
+    {
+      "epoch": 1.2792810020078402,
+      "grad_norm": 3.4911565773739937,
+      "learning_rate": 1e-06,
+      "loss": 0.0559,
+      "step": 6690
+    },
+    {
+      "epoch": 1.2794722248780954,
+      "grad_norm": 3.239659511901175,
+      "learning_rate": 1e-06,
+      "loss": 0.0303,
+      "step": 6691
+    },
+    {
+      "epoch": 1.2796634477483506,
+      "grad_norm": 5.489951056868497,
+      "learning_rate": 1e-06,
+      "loss": 0.0849,
+      "step": 6692
+    },
+    {
+      "epoch": 1.279854670618606,
+      "grad_norm": 6.159266938026569,
+      "learning_rate": 1e-06,
+      "loss": 0.5947,
+      "step": 6693
+    },
+    {
+      "epoch": 1.2800458934888612,
+      "grad_norm": 4.13004202906765,
+      "learning_rate": 1e-06,
+      "loss": 0.3078,
+      "step": 6694
+    },
+    {
+      "epoch": 1.2802371163591166,
+      "grad_norm": 4.686423623162168,
+      "learning_rate": 1e-06,
+      "loss": 0.2296,
+      "step": 6695
+    },
+    {
+      "epoch": 1.2804283392293718,
+      "grad_norm": 3.13082495448562,
+      "learning_rate": 1e-06,
+      "loss": 0.2855,
+      "step": 6696
+    },
+    {
+      "epoch": 1.280619562099627,
+      "grad_norm": 4.0854109065540305,
+      "learning_rate": 1e-06,
+      "loss": 0.0896,
+      "step": 6697
+    },
+    {
+      "epoch": 1.2808107849698824,
+      "grad_norm": 2.4796530516262165,
+      "learning_rate": 1e-06,
+      "loss": 0.082,
+      "step": 6698
+    },
+    {
+      "epoch": 1.2810020078401376,
+      "grad_norm": 1.8846283067083576,
+      "learning_rate": 1e-06,
+      "loss": 0.0761,
+      "step": 6699
+    },
+    {
+      "epoch": 1.281193230710393,
+      "grad_norm": 2.312700005207531,
+      "learning_rate": 1e-06,
+      "loss": 0.1921,
+      "step": 6700
+    },
+    {
+      "epoch": 1.2813844535806482,
+      "grad_norm": 6.16716778068009,
+      "learning_rate": 1e-06,
+      "loss": 0.0576,
+      "step": 6701
+    },
+    {
+      "epoch": 1.2815756764509034,
+      "grad_norm": 2.6895035660722377,
+      "learning_rate": 1e-06,
+      "loss": 0.1402,
+      "step": 6702
+    },
+    {
+      "epoch": 1.2817668993211588,
+      "grad_norm": 2.012397607129592,
+      "learning_rate": 1e-06,
+      "loss": 0.0393,
+      "step": 6703
+    },
+    {
+      "epoch": 1.281958122191414,
+      "grad_norm": 2.772714182059573,
+      "learning_rate": 1e-06,
+      "loss": 0.0544,
+      "step": 6704
+    },
+    {
+      "epoch": 1.2821493450616694,
+      "grad_norm": 3.042755628769134,
+      "learning_rate": 1e-06,
+      "loss": 0.0967,
+      "step": 6705
+    },
+    {
+      "epoch": 1.2823405679319246,
+      "grad_norm": 3.231768751284625,
+      "learning_rate": 1e-06,
+      "loss": 0.6103,
+      "step": 6706
+    },
+    {
+      "epoch": 1.2825317908021798,
+      "grad_norm": 3.6604741140658397,
+      "learning_rate": 1e-06,
+      "loss": 0.4139,
+      "step": 6707
+    },
+    {
+      "epoch": 1.2827230136724352,
+      "grad_norm": 3.985997726003378,
+      "learning_rate": 1e-06,
+      "loss": 0.2431,
+      "step": 6708
+    },
+    {
+      "epoch": 1.2829142365426904,
+      "grad_norm": 2.1435663661768984,
+      "learning_rate": 1e-06,
+      "loss": 0.0693,
+      "step": 6709
+    },
+    {
+      "epoch": 1.2831054594129458,
+      "grad_norm": 1.8149567758126466,
+      "learning_rate": 1e-06,
+      "loss": 0.0908,
+      "step": 6710
+    },
+    {
+      "epoch": 1.283296682283201,
+      "grad_norm": 4.202862699471096,
+      "learning_rate": 1e-06,
+      "loss": 0.2793,
+      "step": 6711
+    },
+    {
+      "epoch": 1.2834879051534562,
+      "grad_norm": 3.0755559589254644,
+      "learning_rate": 1e-06,
+      "loss": 0.1915,
+      "step": 6712
+    },
+    {
+      "epoch": 1.2836791280237116,
+      "grad_norm": 3.9350628046358067,
+      "learning_rate": 1e-06,
+      "loss": 0.1452,
+      "step": 6713
+    },
+    {
+      "epoch": 1.2838703508939668,
+      "grad_norm": 2.5222775654408984,
+      "learning_rate": 1e-06,
+      "loss": 0.0415,
+      "step": 6714
+    },
+    {
+      "epoch": 1.2840615737642223,
+      "grad_norm": 4.108646690896558,
+      "learning_rate": 1e-06,
+      "loss": 0.2305,
+      "step": 6715
+    },
+    {
+      "epoch": 1.2842527966344774,
+      "grad_norm": 1.8470976874392322,
+      "learning_rate": 1e-06,
+      "loss": 0.0388,
+      "step": 6716
+    },
+    {
+      "epoch": 1.2844440195047326,
+      "grad_norm": 6.165893437477185,
+      "learning_rate": 1e-06,
+      "loss": 0.235,
+      "step": 6717
+    },
+    {
+      "epoch": 1.284635242374988,
+      "grad_norm": 7.37766256599908,
+      "learning_rate": 1e-06,
+      "loss": 0.1935,
+      "step": 6718
+    },
+    {
+      "epoch": 1.2848264652452435,
+      "grad_norm": 4.7212933892197215,
+      "learning_rate": 1e-06,
+      "loss": 0.2834,
+      "step": 6719
+    },
+    {
+      "epoch": 1.2850176881154987,
+      "grad_norm": 2.3274850669898073,
+      "learning_rate": 1e-06,
+      "loss": 0.134,
+      "step": 6720
+    },
+    {
+      "epoch": 1.2852089109857538,
+      "grad_norm": 2.687517565292797,
+      "learning_rate": 1e-06,
+      "loss": 0.1339,
+      "step": 6721
+    },
+    {
+      "epoch": 1.285400133856009,
+      "grad_norm": 5.780708246364786,
+      "learning_rate": 1e-06,
+      "loss": 0.6566,
+      "step": 6722
+    },
+    {
+      "epoch": 1.2855913567262645,
+      "grad_norm": 3.8692569628521802,
+      "learning_rate": 1e-06,
+      "loss": 0.2157,
+      "step": 6723
+    },
+    {
+      "epoch": 1.2857825795965199,
+      "grad_norm": 1.7089675640914246,
+      "learning_rate": 1e-06,
+      "loss": 0.0859,
+      "step": 6724
+    },
+    {
+      "epoch": 1.285973802466775,
+      "grad_norm": 4.051762637643247,
+      "learning_rate": 1e-06,
+      "loss": 0.2589,
+      "step": 6725
+    },
+    {
+      "epoch": 1.2861650253370303,
+      "grad_norm": 2.5815564063775716,
+      "learning_rate": 1e-06,
+      "loss": 0.0804,
+      "step": 6726
+    },
+    {
+      "epoch": 1.2863562482072857,
+      "grad_norm": 2.054966780679446,
+      "learning_rate": 1e-06,
+      "loss": 0.015,
+      "step": 6727
+    },
+    {
+      "epoch": 1.2865474710775409,
+      "grad_norm": 1.7974568627371503,
+      "learning_rate": 1e-06,
+      "loss": 0.0239,
+      "step": 6728
+    },
+    {
+      "epoch": 1.2867386939477963,
+      "grad_norm": 1.6976365714549906,
+      "learning_rate": 1e-06,
+      "loss": 0.0181,
+      "step": 6729
+    },
+    {
+      "epoch": 1.2869299168180515,
+      "grad_norm": 4.374081760502265,
+      "learning_rate": 1e-06,
+      "loss": 0.1109,
+      "step": 6730
+    },
+    {
+      "epoch": 1.2871211396883067,
+      "grad_norm": 5.05104388788098,
+      "learning_rate": 1e-06,
+      "loss": 0.2905,
+      "step": 6731
+    },
+    {
+      "epoch": 1.287312362558562,
+      "grad_norm": 4.1970938482764755,
+      "learning_rate": 1e-06,
+      "loss": 0.3356,
+      "step": 6732
+    },
+    {
+      "epoch": 1.2875035854288173,
+      "grad_norm": 5.6311244260481175,
+      "learning_rate": 1e-06,
+      "loss": 0.155,
+      "step": 6733
+    },
+    {
+      "epoch": 1.2876948082990727,
+      "grad_norm": 2.089356781026843,
+      "learning_rate": 1e-06,
+      "loss": 0.1177,
+      "step": 6734
+    },
+    {
+      "epoch": 1.2878860311693279,
+      "grad_norm": 2.676045069253693,
+      "learning_rate": 1e-06,
+      "loss": 0.1465,
+      "step": 6735
+    },
+    {
+      "epoch": 1.288077254039583,
+      "grad_norm": 3.541201991096815,
+      "learning_rate": 1e-06,
+      "loss": 0.2649,
+      "step": 6736
+    },
+    {
+      "epoch": 1.2882684769098385,
+      "grad_norm": 3.0905319585574964,
+      "learning_rate": 1e-06,
+      "loss": 0.1666,
+      "step": 6737
+    },
+    {
+      "epoch": 1.2884596997800937,
+      "grad_norm": 1.981701529281495,
+      "learning_rate": 1e-06,
+      "loss": 0.0482,
+      "step": 6738
+    },
+    {
+      "epoch": 1.288650922650349,
+      "grad_norm": 1.3135117309853834,
+      "learning_rate": 1e-06,
+      "loss": 0.0156,
+      "step": 6739
+    },
+    {
+      "epoch": 1.2888421455206043,
+      "grad_norm": 3.160325393582559,
+      "learning_rate": 1e-06,
+      "loss": 0.0591,
+      "step": 6740
+    },
+    {
+      "epoch": 1.2890333683908595,
+      "grad_norm": 6.18269793732313,
+      "learning_rate": 1e-06,
+      "loss": 0.1078,
+      "step": 6741
+    },
+    {
+      "epoch": 1.2892245912611149,
+      "grad_norm": 6.427146511498974,
+      "learning_rate": 1e-06,
+      "loss": 0.1089,
+      "step": 6742
+    },
+    {
+      "epoch": 1.28941581413137,
+      "grad_norm": 7.257214606781099,
+      "learning_rate": 1e-06,
+      "loss": 0.2907,
+      "step": 6743
+    },
+    {
+      "epoch": 1.2896070370016255,
+      "grad_norm": 2.6783044209687246,
+      "learning_rate": 1e-06,
+      "loss": 0.2633,
+      "step": 6744
+    },
+    {
+      "epoch": 1.2897982598718807,
+      "grad_norm": 7.8280097511320434,
+      "learning_rate": 1e-06,
+      "loss": 0.2829,
+      "step": 6745
+    },
+    {
+      "epoch": 1.2899894827421359,
+      "grad_norm": 2.8495322328810437,
+      "learning_rate": 1e-06,
+      "loss": 0.1555,
+      "step": 6746
+    },
+    {
+      "epoch": 1.2901807056123913,
+      "grad_norm": 4.59335429245403,
+      "learning_rate": 1e-06,
+      "loss": 0.554,
+      "step": 6747
+    },
+    {
+      "epoch": 1.2903719284826465,
+      "grad_norm": 2.625965304224658,
+      "learning_rate": 1e-06,
+      "loss": 0.1104,
+      "step": 6748
+    },
+    {
+      "epoch": 1.290563151352902,
+      "grad_norm": 2.2541169542435466,
+      "learning_rate": 1e-06,
+      "loss": 0.078,
+      "step": 6749
+    },
+    {
+      "epoch": 1.290754374223157,
+      "grad_norm": 1.612338180702425,
+      "learning_rate": 1e-06,
+      "loss": 0.1579,
+      "step": 6750
+    },
+    {
+      "epoch": 1.2909455970934123,
+      "grad_norm": 3.067601994915009,
+      "learning_rate": 1e-06,
+      "loss": 0.0969,
+      "step": 6751
+    },
+    {
+      "epoch": 1.2911368199636677,
+      "grad_norm": 2.7560488326863615,
+      "learning_rate": 1e-06,
+      "loss": 0.083,
+      "step": 6752
+    },
+    {
+      "epoch": 1.2913280428339229,
+      "grad_norm": 2.6105534981691103,
+      "learning_rate": 1e-06,
+      "loss": 0.1182,
+      "step": 6753
+    },
+    {
+      "epoch": 1.2915192657041783,
+      "grad_norm": 1.9103770372224755,
+      "learning_rate": 1e-06,
+      "loss": 0.0247,
+      "step": 6754
+    },
+    {
+      "epoch": 1.2917104885744335,
+      "grad_norm": 2.2067608422911578,
+      "learning_rate": 1e-06,
+      "loss": 0.0257,
+      "step": 6755
+    },
+    {
+      "epoch": 1.2919017114446887,
+      "grad_norm": 5.205069147213177,
+      "learning_rate": 1e-06,
+      "loss": 0.457,
+      "step": 6756
+    },
+    {
+      "epoch": 1.292092934314944,
+      "grad_norm": 5.514244360666348,
+      "learning_rate": 1e-06,
+      "loss": 0.2864,
+      "step": 6757
+    },
+    {
+      "epoch": 1.2922841571851993,
+      "grad_norm": 3.07643460559579,
+      "learning_rate": 1e-06,
+      "loss": 0.0801,
+      "step": 6758
+    },
+    {
+      "epoch": 1.2924753800554547,
+      "grad_norm": 4.085119570379413,
+      "learning_rate": 1e-06,
+      "loss": 0.197,
+      "step": 6759
+    },
+    {
+      "epoch": 1.29266660292571,
+      "grad_norm": 3.9124116442761463,
+      "learning_rate": 1e-06,
+      "loss": 0.2054,
+      "step": 6760
+    },
+    {
+      "epoch": 1.292857825795965,
+      "grad_norm": 2.833027673577399,
+      "learning_rate": 1e-06,
+      "loss": 0.1679,
+      "step": 6761
+    },
+    {
+      "epoch": 1.2930490486662205,
+      "grad_norm": 5.37634331528853,
+      "learning_rate": 1e-06,
+      "loss": 0.129,
+      "step": 6762
+    },
+    {
+      "epoch": 1.2932402715364757,
+      "grad_norm": 3.3794067717706544,
+      "learning_rate": 1e-06,
+      "loss": 0.1047,
+      "step": 6763
+    },
+    {
+      "epoch": 1.293431494406731,
+      "grad_norm": 3.2816353707849855,
+      "learning_rate": 1e-06,
+      "loss": 0.0498,
+      "step": 6764
+    },
+    {
+      "epoch": 1.2936227172769863,
+      "grad_norm": 4.773255889484559,
+      "learning_rate": 1e-06,
+      "loss": 0.0856,
+      "step": 6765
+    },
+    {
+      "epoch": 1.2938139401472415,
+      "grad_norm": 2.358677438854046,
+      "learning_rate": 1e-06,
+      "loss": 0.0379,
+      "step": 6766
+    },
+    {
+      "epoch": 1.294005163017497,
+      "grad_norm": 3.7062086955401554,
+      "learning_rate": 1e-06,
+      "loss": 0.083,
+      "step": 6767
+    },
+    {
+      "epoch": 1.294196385887752,
+      "grad_norm": 3.4711867013283104,
+      "learning_rate": 1e-06,
+      "loss": 0.1204,
+      "step": 6768
+    },
+    {
+      "epoch": 1.2943876087580075,
+      "grad_norm": 3.713086086468029,
+      "learning_rate": 1e-06,
+      "loss": 0.1903,
+      "step": 6769
+    },
+    {
+      "epoch": 1.2945788316282627,
+      "grad_norm": 5.0222607030679,
+      "learning_rate": 1e-06,
+      "loss": 0.3857,
+      "step": 6770
+    },
+    {
+      "epoch": 1.294770054498518,
+      "grad_norm": 2.002670769809595,
+      "learning_rate": 1e-06,
+      "loss": 0.109,
+      "step": 6771
+    },
+    {
+      "epoch": 1.2949612773687733,
+      "grad_norm": 2.707809886721997,
+      "learning_rate": 1e-06,
+      "loss": 0.1258,
+      "step": 6772
+    },
+    {
+      "epoch": 1.2951525002390287,
+      "grad_norm": 3.7223161540033116,
+      "learning_rate": 1e-06,
+      "loss": 0.2418,
+      "step": 6773
+    },
+    {
+      "epoch": 1.295343723109284,
+      "grad_norm": 3.022167007870663,
+      "learning_rate": 1e-06,
+      "loss": 0.0934,
+      "step": 6774
+    },
+    {
+      "epoch": 1.295534945979539,
+      "grad_norm": 2.4987316728981788,
+      "learning_rate": 1e-06,
+      "loss": 0.1893,
+      "step": 6775
+    },
+    {
+      "epoch": 1.2957261688497943,
+      "grad_norm": 2.247546129442817,
+      "learning_rate": 1e-06,
+      "loss": 0.0761,
+      "step": 6776
+    },
+    {
+      "epoch": 1.2959173917200497,
+      "grad_norm": 2.045740526610248,
+      "learning_rate": 1e-06,
+      "loss": 0.0406,
+      "step": 6777
+    },
+    {
+      "epoch": 1.2961086145903051,
+      "grad_norm": 5.431651335221821,
+      "learning_rate": 1e-06,
+      "loss": 0.0578,
+      "step": 6778
+    },
+    {
+      "epoch": 1.2962998374605603,
+      "grad_norm": 1.441517895352274,
+      "learning_rate": 1e-06,
+      "loss": 0.0186,
+      "step": 6779
+    },
+    {
+      "epoch": 1.2964910603308155,
+      "grad_norm": 6.354482507412521,
+      "learning_rate": 1e-06,
+      "loss": 0.1878,
+      "step": 6780
+    },
+    {
+      "epoch": 1.296682283201071,
+      "grad_norm": 3.8646174975150656,
+      "learning_rate": 1e-06,
+      "loss": 0.2328,
+      "step": 6781
+    },
+    {
+      "epoch": 1.2968735060713261,
+      "grad_norm": 4.160554650084276,
+      "learning_rate": 1e-06,
+      "loss": 0.2139,
+      "step": 6782
+    },
+    {
+      "epoch": 1.2970647289415815,
+      "grad_norm": 3.432652332456997,
+      "learning_rate": 1e-06,
+      "loss": 0.0836,
+      "step": 6783
+    },
+    {
+      "epoch": 1.2972559518118367,
+      "grad_norm": 3.03604988679371,
+      "learning_rate": 1e-06,
+      "loss": 0.1494,
+      "step": 6784
+    },
+    {
+      "epoch": 1.297447174682092,
+      "grad_norm": 2.9183914715872348,
+      "learning_rate": 1e-06,
+      "loss": 0.0656,
+      "step": 6785
+    },
+    {
+      "epoch": 1.2976383975523473,
+      "grad_norm": 5.0289904336252365,
+      "learning_rate": 1e-06,
+      "loss": 0.3491,
+      "step": 6786
+    },
+    {
+      "epoch": 1.2978296204226025,
+      "grad_norm": 2.553364262289797,
+      "learning_rate": 1e-06,
+      "loss": 0.0786,
+      "step": 6787
+    },
+    {
+      "epoch": 1.298020843292858,
+      "grad_norm": 1.6399172028133402,
+      "learning_rate": 1e-06,
+      "loss": 0.0415,
+      "step": 6788
+    },
+    {
+      "epoch": 1.2982120661631131,
+      "grad_norm": 11.668167526344982,
+      "learning_rate": 1e-06,
+      "loss": 0.054,
+      "step": 6789
+    },
+    {
+      "epoch": 1.2984032890333683,
+      "grad_norm": 3.6010236026633424,
+      "learning_rate": 1e-06,
+      "loss": 0.0609,
+      "step": 6790
+    },
+    {
+      "epoch": 1.2985945119036237,
+      "grad_norm": 1.8802896906474291,
+      "learning_rate": 1e-06,
+      "loss": 0.0172,
+      "step": 6791
+    },
+    {
+      "epoch": 1.298785734773879,
+      "grad_norm": 7.194260842407893,
+      "learning_rate": 1e-06,
+      "loss": 0.2257,
+      "step": 6792
+    },
+    {
+      "epoch": 1.2989769576441343,
+      "grad_norm": 8.24055217411293,
+      "learning_rate": 1e-06,
+      "loss": 0.2565,
+      "step": 6793
+    },
+    {
+      "epoch": 1.2991681805143895,
+      "grad_norm": 5.106630854136602,
+      "learning_rate": 1e-06,
+      "loss": 0.1395,
+      "step": 6794
+    },
+    {
+      "epoch": 1.2993594033846447,
+      "grad_norm": 2.7732122625309006,
+      "learning_rate": 1e-06,
+      "loss": 0.2242,
+      "step": 6795
+    },
+    {
+      "epoch": 1.2995506262549001,
+      "grad_norm": 5.893151669758993,
+      "learning_rate": 1e-06,
+      "loss": 0.1652,
+      "step": 6796
+    },
+    {
+      "epoch": 1.2997418491251553,
+      "grad_norm": 2.9922007269106317,
+      "learning_rate": 1e-06,
+      "loss": 0.0867,
+      "step": 6797
+    },
+    {
+      "epoch": 1.2999330719954107,
+      "grad_norm": 4.122986967012068,
+      "learning_rate": 1e-06,
+      "loss": 0.1649,
+      "step": 6798
+    },
+    {
+      "epoch": 1.300124294865666,
+      "grad_norm": 1.9986975720598916,
+      "learning_rate": 1e-06,
+      "loss": 0.0631,
+      "step": 6799
+    },
+    {
+      "epoch": 1.3003155177359211,
+      "grad_norm": 3.3525559880234708,
+      "learning_rate": 1e-06,
+      "loss": 0.1706,
+      "step": 6800
+    },
+    {
+      "epoch": 1.3005067406061765,
+      "grad_norm": 3.227304147086279,
+      "learning_rate": 1e-06,
+      "loss": 0.1275,
+      "step": 6801
+    },
+    {
+      "epoch": 1.3006979634764317,
+      "grad_norm": 1.4139467054977193,
+      "learning_rate": 1e-06,
+      "loss": 0.0225,
+      "step": 6802
+    },
+    {
+      "epoch": 1.3008891863466872,
+      "grad_norm": 2.4217656264680603,
+      "learning_rate": 1e-06,
+      "loss": 0.0289,
+      "step": 6803
+    },
+    {
+      "epoch": 1.3010804092169423,
+      "grad_norm": 2.9782553036854753,
+      "learning_rate": 1e-06,
+      "loss": 0.0553,
+      "step": 6804
+    },
+    {
+      "epoch": 1.3012716320871975,
+      "grad_norm": 5.506822862388051,
+      "learning_rate": 1e-06,
+      "loss": 0.3145,
+      "step": 6805
+    },
+    {
+      "epoch": 1.301462854957453,
+      "grad_norm": 4.178948852299085,
+      "learning_rate": 1e-06,
+      "loss": 0.2456,
+      "step": 6806
+    },
+    {
+      "epoch": 1.3016540778277081,
+      "grad_norm": 3.4279118766363257,
+      "learning_rate": 1e-06,
+      "loss": 0.3083,
+      "step": 6807
+    },
+    {
+      "epoch": 1.3018453006979636,
+      "grad_norm": 3.792342883723661,
+      "learning_rate": 1e-06,
+      "loss": 0.2346,
+      "step": 6808
+    },
+    {
+      "epoch": 1.3020365235682188,
+      "grad_norm": 3.960373697276877,
+      "learning_rate": 1e-06,
+      "loss": 0.2938,
+      "step": 6809
+    },
+    {
+      "epoch": 1.302227746438474,
+      "grad_norm": 2.4168252783278223,
+      "learning_rate": 1e-06,
+      "loss": 0.2093,
+      "step": 6810
+    },
+    {
+      "epoch": 1.3024189693087294,
+      "grad_norm": 4.397279956337133,
+      "learning_rate": 1e-06,
+      "loss": 0.3559,
+      "step": 6811
+    },
+    {
+      "epoch": 1.3026101921789845,
+      "grad_norm": 3.5062460343445574,
+      "learning_rate": 1e-06,
+      "loss": 0.0582,
+      "step": 6812
+    },
+    {
+      "epoch": 1.30280141504924,
+      "grad_norm": 2.461943308843043,
+      "learning_rate": 1e-06,
+      "loss": 0.08,
+      "step": 6813
+    },
+    {
+      "epoch": 1.3029926379194952,
+      "grad_norm": 2.967463967699889,
+      "learning_rate": 1e-06,
+      "loss": 0.1253,
+      "step": 6814
+    },
+    {
+      "epoch": 1.3031838607897503,
+      "grad_norm": 2.1264326090948025,
+      "learning_rate": 1e-06,
+      "loss": 0.0912,
+      "step": 6815
+    },
+    {
+      "epoch": 1.3033750836600058,
+      "grad_norm": 3.7217882070983155,
+      "learning_rate": 1e-06,
+      "loss": 0.2267,
+      "step": 6816
+    },
+    {
+      "epoch": 1.303566306530261,
+      "grad_norm": 2.707998391983479,
+      "learning_rate": 1e-06,
+      "loss": 0.0333,
+      "step": 6817
+    },
+    {
+      "epoch": 1.3037575294005164,
+      "grad_norm": 3.909749287130853,
+      "learning_rate": 1e-06,
+      "loss": 0.2666,
+      "step": 6818
+    },
+    {
+      "epoch": 1.3039487522707716,
+      "grad_norm": 6.447775926014322,
+      "learning_rate": 1e-06,
+      "loss": 0.2929,
+      "step": 6819
+    },
+    {
+      "epoch": 1.3041399751410268,
+      "grad_norm": 3.4729850965202624,
+      "learning_rate": 1e-06,
+      "loss": 0.0829,
+      "step": 6820
+    },
+    {
+      "epoch": 1.3043311980112822,
+      "grad_norm": 3.044297913538843,
+      "learning_rate": 1e-06,
+      "loss": 0.1522,
+      "step": 6821
+    },
+    {
+      "epoch": 1.3045224208815374,
+      "grad_norm": 2.9347758546465865,
+      "learning_rate": 1e-06,
+      "loss": 0.1552,
+      "step": 6822
+    },
+    {
+      "epoch": 1.3047136437517928,
+      "grad_norm": 2.660989908003566,
+      "learning_rate": 1e-06,
+      "loss": 0.0706,
+      "step": 6823
+    },
+    {
+      "epoch": 1.304904866622048,
+      "grad_norm": 2.987394552883768,
+      "learning_rate": 1e-06,
+      "loss": 0.0755,
+      "step": 6824
+    },
+    {
+      "epoch": 1.3050960894923032,
+      "grad_norm": 2.597835219105045,
+      "learning_rate": 1e-06,
+      "loss": 0.1113,
+      "step": 6825
+    },
+    {
+      "epoch": 1.3052873123625586,
+      "grad_norm": 3.111342224740494,
+      "learning_rate": 1e-06,
+      "loss": 0.0988,
+      "step": 6826
+    },
+    {
+      "epoch": 1.3054785352328138,
+      "grad_norm": 1.7231783605483104,
+      "learning_rate": 1e-06,
+      "loss": 0.0704,
+      "step": 6827
+    },
+    {
+      "epoch": 1.3056697581030692,
+      "grad_norm": 2.562620020591052,
+      "learning_rate": 1e-06,
+      "loss": 0.0858,
+      "step": 6828
+    },
+    {
+      "epoch": 1.3058609809733244,
+      "grad_norm": 1.4835729741172996,
+      "learning_rate": 1e-06,
+      "loss": 0.0096,
+      "step": 6829
+    },
+    {
+      "epoch": 1.3060522038435796,
+      "grad_norm": 7.8573429825302386,
+      "learning_rate": 1e-06,
+      "loss": 0.1286,
+      "step": 6830
+    },
+    {
+      "epoch": 1.306243426713835,
+      "grad_norm": 8.922630050964075,
+      "learning_rate": 1e-06,
+      "loss": 0.5149,
+      "step": 6831
+    },
+    {
+      "epoch": 1.3064346495840904,
+      "grad_norm": 4.674543862351162,
+      "learning_rate": 1e-06,
+      "loss": 0.6069,
+      "step": 6832
+    },
+    {
+      "epoch": 1.3066258724543456,
+      "grad_norm": 5.835054343525391,
+      "learning_rate": 1e-06,
+      "loss": 0.2508,
+      "step": 6833
+    },
+    {
+      "epoch": 1.3068170953246008,
+      "grad_norm": 2.101623534269961,
+      "learning_rate": 1e-06,
+      "loss": 0.1008,
+      "step": 6834
+    },
+    {
+      "epoch": 1.307008318194856,
+      "grad_norm": 3.8767250897200136,
+      "learning_rate": 1e-06,
+      "loss": 0.1263,
+      "step": 6835
+    },
+    {
+      "epoch": 1.3071995410651114,
+      "grad_norm": 4.651357038456018,
+      "learning_rate": 1e-06,
+      "loss": 0.0982,
+      "step": 6836
+    },
+    {
+      "epoch": 1.3073907639353668,
+      "grad_norm": 4.118630085937561,
+      "learning_rate": 1e-06,
+      "loss": 0.2304,
+      "step": 6837
+    },
+    {
+      "epoch": 1.307581986805622,
+      "grad_norm": 4.001498656860264,
+      "learning_rate": 1e-06,
+      "loss": 0.1481,
+      "step": 6838
+    },
+    {
+      "epoch": 1.3077732096758772,
+      "grad_norm": 3.3972024010570165,
+      "learning_rate": 1e-06,
+      "loss": 0.0657,
+      "step": 6839
+    },
+    {
+      "epoch": 1.3079644325461326,
+      "grad_norm": 4.138064416840308,
+      "learning_rate": 1e-06,
+      "loss": 0.0813,
+      "step": 6840
+    },
+    {
+      "epoch": 1.3081556554163878,
+      "grad_norm": 3.43722449845913,
+      "learning_rate": 1e-06,
+      "loss": 0.0724,
+      "step": 6841
+    },
+    {
+      "epoch": 1.3083468782866432,
+      "grad_norm": 8.482232766499521,
+      "learning_rate": 1e-06,
+      "loss": 0.0961,
+      "step": 6842
+    },
+    {
+      "epoch": 1.3085381011568984,
+      "grad_norm": 5.444150694017829,
+      "learning_rate": 1e-06,
+      "loss": 0.6303,
+      "step": 6843
+    },
+    {
+      "epoch": 1.3087293240271536,
+      "grad_norm": 5.0488844139753875,
+      "learning_rate": 1e-06,
+      "loss": 0.4141,
+      "step": 6844
+    },
+    {
+      "epoch": 1.308920546897409,
+      "grad_norm": 2.536024418138704,
+      "learning_rate": 1e-06,
+      "loss": 0.0557,
+      "step": 6845
+    },
+    {
+      "epoch": 1.3091117697676642,
+      "grad_norm": 3.362131664423399,
+      "learning_rate": 1e-06,
+      "loss": 0.1446,
+      "step": 6846
+    },
+    {
+      "epoch": 1.3093029926379196,
+      "grad_norm": 3.1730624076315626,
+      "learning_rate": 1e-06,
+      "loss": 0.0918,
+      "step": 6847
+    },
+    {
+      "epoch": 1.3094942155081748,
+      "grad_norm": 2.462121684645672,
+      "learning_rate": 1e-06,
+      "loss": 0.0535,
+      "step": 6848
+    },
+    {
+      "epoch": 1.30968543837843,
+      "grad_norm": 2.7093289134891885,
+      "learning_rate": 1e-06,
+      "loss": 0.049,
+      "step": 6849
+    },
+    {
+      "epoch": 1.3098766612486854,
+      "grad_norm": 2.757686968748001,
+      "learning_rate": 1e-06,
+      "loss": 0.087,
+      "step": 6850
+    },
+    {
+      "epoch": 1.3100678841189406,
+      "grad_norm": 2.64890440479566,
+      "learning_rate": 1e-06,
+      "loss": 0.0346,
+      "step": 6851
+    },
+    {
+      "epoch": 1.310259106989196,
+      "grad_norm": 2.0143148023018194,
+      "learning_rate": 1e-06,
+      "loss": 0.0722,
+      "step": 6852
+    },
+    {
+      "epoch": 1.3104503298594512,
+      "grad_norm": 1.4119417083097996,
+      "learning_rate": 1e-06,
+      "loss": 0.0193,
+      "step": 6853
+    },
+    {
+      "epoch": 1.3106415527297064,
+      "grad_norm": 3.4045118744410994,
+      "learning_rate": 1e-06,
+      "loss": 0.1626,
+      "step": 6854
+    },
+    {
+      "epoch": 1.3108327755999618,
+      "grad_norm": 7.3977999226167475,
+      "learning_rate": 1e-06,
+      "loss": 0.1864,
+      "step": 6855
+    },
+    {
+      "epoch": 1.311023998470217,
+      "grad_norm": 4.178482138549564,
+      "learning_rate": 1e-06,
+      "loss": 0.4195,
+      "step": 6856
+    },
+    {
+      "epoch": 1.3112152213404724,
+      "grad_norm": 3.802000502986162,
+      "learning_rate": 1e-06,
+      "loss": 0.2099,
+      "step": 6857
+    },
+    {
+      "epoch": 1.3114064442107276,
+      "grad_norm": 2.2256242883515607,
+      "learning_rate": 1e-06,
+      "loss": 0.1218,
+      "step": 6858
+    },
+    {
+      "epoch": 1.3115976670809828,
+      "grad_norm": 3.8630048847765037,
+      "learning_rate": 1e-06,
+      "loss": 0.1954,
+      "step": 6859
+    },
+    {
+      "epoch": 1.3117888899512382,
+      "grad_norm": 4.2677665057821494,
+      "learning_rate": 1e-06,
+      "loss": 0.2879,
+      "step": 6860
+    },
+    {
+      "epoch": 1.3119801128214934,
+      "grad_norm": 2.843112916113599,
+      "learning_rate": 1e-06,
+      "loss": 0.2127,
+      "step": 6861
+    },
+    {
+      "epoch": 1.3121713356917488,
+      "grad_norm": 1.643173781084752,
+      "learning_rate": 1e-06,
+      "loss": 0.1179,
+      "step": 6862
+    },
+    {
+      "epoch": 1.312362558562004,
+      "grad_norm": 1.4104948192151057,
+      "learning_rate": 1e-06,
+      "loss": 0.017,
+      "step": 6863
+    },
+    {
+      "epoch": 1.3125537814322592,
+      "grad_norm": 3.755041167215144,
+      "learning_rate": 1e-06,
+      "loss": 0.1917,
+      "step": 6864
+    },
+    {
+      "epoch": 1.3127450043025146,
+      "grad_norm": 1.8752728581573146,
+      "learning_rate": 1e-06,
+      "loss": 0.0797,
+      "step": 6865
+    },
+    {
+      "epoch": 1.3129362271727698,
+      "grad_norm": 2.525970038483224,
+      "learning_rate": 1e-06,
+      "loss": 0.0565,
+      "step": 6866
+    },
+    {
+      "epoch": 1.3131274500430252,
+      "grad_norm": 3.436126295786181,
+      "learning_rate": 1e-06,
+      "loss": 0.0461,
+      "step": 6867
+    },
+    {
+      "epoch": 1.3133186729132804,
+      "grad_norm": 5.215240760578164,
+      "learning_rate": 1e-06,
+      "loss": 0.1533,
+      "step": 6868
+    },
+    {
+      "epoch": 1.3135098957835356,
+      "grad_norm": 3.3431239032155204,
+      "learning_rate": 1e-06,
+      "loss": 0.2859,
+      "step": 6869
+    },
+    {
+      "epoch": 1.313701118653791,
+      "grad_norm": 4.687456461386343,
+      "learning_rate": 1e-06,
+      "loss": 0.4582,
+      "step": 6870
+    },
+    {
+      "epoch": 1.3138923415240462,
+      "grad_norm": 2.313907710326135,
+      "learning_rate": 1e-06,
+      "loss": 0.0942,
+      "step": 6871
+    },
+    {
+      "epoch": 1.3140835643943016,
+      "grad_norm": 4.744884145305148,
+      "learning_rate": 1e-06,
+      "loss": 0.3728,
+      "step": 6872
+    },
+    {
+      "epoch": 1.3142747872645568,
+      "grad_norm": 3.4603648519263492,
+      "learning_rate": 1e-06,
+      "loss": 0.085,
+      "step": 6873
+    },
+    {
+      "epoch": 1.314466010134812,
+      "grad_norm": 2.0103966377591465,
+      "learning_rate": 1e-06,
+      "loss": 0.0911,
+      "step": 6874
+    },
+    {
+      "epoch": 1.3146572330050674,
+      "grad_norm": 2.9286673028384995,
+      "learning_rate": 1e-06,
+      "loss": 0.061,
+      "step": 6875
+    },
+    {
+      "epoch": 1.3148484558753226,
+      "grad_norm": 2.2366298155536892,
+      "learning_rate": 1e-06,
+      "loss": 0.0375,
+      "step": 6876
+    },
+    {
+      "epoch": 1.315039678745578,
+      "grad_norm": 3.748748443287739,
+      "learning_rate": 1e-06,
+      "loss": 0.0877,
+      "step": 6877
+    },
+    {
+      "epoch": 1.3152309016158332,
+      "grad_norm": 4.268341429162878,
+      "learning_rate": 1e-06,
+      "loss": 0.0773,
+      "step": 6878
+    },
+    {
+      "epoch": 1.3154221244860884,
+      "grad_norm": 3.312944490372433,
+      "learning_rate": 1e-06,
+      "loss": 0.0845,
+      "step": 6879
+    },
+    {
+      "epoch": 1.3156133473563438,
+      "grad_norm": 7.886015438860941,
+      "learning_rate": 1e-06,
+      "loss": 0.224,
+      "step": 6880
+    },
+    {
+      "epoch": 1.315804570226599,
+      "grad_norm": 4.448262876426915,
+      "learning_rate": 1e-06,
+      "loss": 0.4397,
+      "step": 6881
+    },
+    {
+      "epoch": 1.3159957930968544,
+      "grad_norm": 4.443145813892533,
+      "learning_rate": 1e-06,
+      "loss": 0.5915,
+      "step": 6882
+    },
+    {
+      "epoch": 1.3161870159671096,
+      "grad_norm": 1.9370251965949365,
+      "learning_rate": 1e-06,
+      "loss": 0.1619,
+      "step": 6883
+    },
+    {
+      "epoch": 1.3163782388373648,
+      "grad_norm": 5.770654815593497,
+      "learning_rate": 1e-06,
+      "loss": 0.3432,
+      "step": 6884
+    },
+    {
+      "epoch": 1.3165694617076202,
+      "grad_norm": 2.7139611874801077,
+      "learning_rate": 1e-06,
+      "loss": 0.1455,
+      "step": 6885
+    },
+    {
+      "epoch": 1.3167606845778757,
+      "grad_norm": 1.5161683200054052,
+      "learning_rate": 1e-06,
+      "loss": 0.0744,
+      "step": 6886
+    },
+    {
+      "epoch": 1.3169519074481308,
+      "grad_norm": 3.06901480159627,
+      "learning_rate": 1e-06,
+      "loss": 0.0852,
+      "step": 6887
+    },
+    {
+      "epoch": 1.317143130318386,
+      "grad_norm": 2.9622517992822748,
+      "learning_rate": 1e-06,
+      "loss": 0.0496,
+      "step": 6888
+    },
+    {
+      "epoch": 1.3173343531886412,
+      "grad_norm": 5.183127709303732,
+      "learning_rate": 1e-06,
+      "loss": 0.0879,
+      "step": 6889
+    },
+    {
+      "epoch": 1.3175255760588966,
+      "grad_norm": 2.608434998924781,
+      "learning_rate": 1e-06,
+      "loss": 0.0492,
+      "step": 6890
+    },
+    {
+      "epoch": 1.317716798929152,
+      "grad_norm": 3.899280403458428,
+      "learning_rate": 1e-06,
+      "loss": 0.1,
+      "step": 6891
+    },
+    {
+      "epoch": 1.3179080217994072,
+      "grad_norm": 11.538688266189396,
+      "learning_rate": 1e-06,
+      "loss": 0.0931,
+      "step": 6892
+    },
+    {
+      "epoch": 1.3180992446696624,
+      "grad_norm": 13.747188211260523,
+      "learning_rate": 1e-06,
+      "loss": 0.3019,
+      "step": 6893
+    },
+    {
+      "epoch": 1.3182904675399179,
+      "grad_norm": 4.195345461128083,
+      "learning_rate": 1e-06,
+      "loss": 0.3147,
+      "step": 6894
+    },
+    {
+      "epoch": 1.318481690410173,
+      "grad_norm": 3.1866136702383,
+      "learning_rate": 1e-06,
+      "loss": 0.08,
+      "step": 6895
+    },
+    {
+      "epoch": 1.3186729132804285,
+      "grad_norm": 1.8092713374088987,
+      "learning_rate": 1e-06,
+      "loss": 0.1718,
+      "step": 6896
+    },
+    {
+      "epoch": 1.3188641361506837,
+      "grad_norm": 4.955307151905577,
+      "learning_rate": 1e-06,
+      "loss": 0.3247,
+      "step": 6897
+    },
+    {
+      "epoch": 1.3190553590209388,
+      "grad_norm": 3.6689959987848337,
+      "learning_rate": 1e-06,
+      "loss": 0.3318,
+      "step": 6898
+    },
+    {
+      "epoch": 1.3192465818911943,
+      "grad_norm": 1.58039669790944,
+      "learning_rate": 1e-06,
+      "loss": 0.0694,
+      "step": 6899
+    },
+    {
+      "epoch": 1.3194378047614495,
+      "grad_norm": 2.844354250921697,
+      "learning_rate": 1e-06,
+      "loss": 0.1188,
+      "step": 6900
+    },
+    {
+      "epoch": 1.3196290276317049,
+      "grad_norm": 2.2829420529470865,
+      "learning_rate": 1e-06,
+      "loss": 0.1469,
+      "step": 6901
+    },
+    {
+      "epoch": 1.31982025050196,
+      "grad_norm": 8.998903843565737,
+      "learning_rate": 1e-06,
+      "loss": 0.076,
+      "step": 6902
+    },
+    {
+      "epoch": 1.3200114733722152,
+      "grad_norm": 2.636654097152952,
+      "learning_rate": 1e-06,
+      "loss": 0.0438,
+      "step": 6903
+    },
+    {
+      "epoch": 1.3202026962424707,
+      "grad_norm": 2.5500465725404395,
+      "learning_rate": 1e-06,
+      "loss": 0.0598,
+      "step": 6904
+    },
+    {
+      "epoch": 1.3203939191127259,
+      "grad_norm": 5.5224194053754365,
+      "learning_rate": 1e-06,
+      "loss": 0.1125,
+      "step": 6905
+    },
+    {
+      "epoch": 1.3205851419829813,
+      "grad_norm": 7.105952182994335,
+      "learning_rate": 1e-06,
+      "loss": 0.5765,
+      "step": 6906
+    },
+    {
+      "epoch": 1.3207763648532365,
+      "grad_norm": 3.7450618018639257,
+      "learning_rate": 1e-06,
+      "loss": 0.1954,
+      "step": 6907
+    },
+    {
+      "epoch": 1.3209675877234917,
+      "grad_norm": 3.14754865419386,
+      "learning_rate": 1e-06,
+      "loss": 0.1318,
+      "step": 6908
+    },
+    {
+      "epoch": 1.321158810593747,
+      "grad_norm": 2.794354629800703,
+      "learning_rate": 1e-06,
+      "loss": 0.1514,
+      "step": 6909
+    },
+    {
+      "epoch": 1.3213500334640023,
+      "grad_norm": 3.650357445759459,
+      "learning_rate": 1e-06,
+      "loss": 0.2373,
+      "step": 6910
+    },
+    {
+      "epoch": 1.3215412563342577,
+      "grad_norm": 2.402318740342379,
+      "learning_rate": 1e-06,
+      "loss": 0.1747,
+      "step": 6911
+    },
+    {
+      "epoch": 1.3217324792045129,
+      "grad_norm": 2.1877005348929943,
+      "learning_rate": 1e-06,
+      "loss": 0.1051,
+      "step": 6912
+    },
+    {
+      "epoch": 1.321923702074768,
+      "grad_norm": 3.3809084566411287,
+      "learning_rate": 1e-06,
+      "loss": 0.0733,
+      "step": 6913
+    },
+    {
+      "epoch": 1.3221149249450235,
+      "grad_norm": 1.6494770839795654,
+      "learning_rate": 1e-06,
+      "loss": 0.0376,
+      "step": 6914
+    },
+    {
+      "epoch": 1.3223061478152787,
+      "grad_norm": 3.3742672160033678,
+      "learning_rate": 1e-06,
+      "loss": 0.1944,
+      "step": 6915
+    },
+    {
+      "epoch": 1.322497370685534,
+      "grad_norm": 2.8766226128674495,
+      "learning_rate": 1e-06,
+      "loss": 0.0502,
+      "step": 6916
+    },
+    {
+      "epoch": 1.3226885935557893,
+      "grad_norm": 8.428324365573985,
+      "learning_rate": 1e-06,
+      "loss": 0.1328,
+      "step": 6917
+    },
+    {
+      "epoch": 1.3228798164260445,
+      "grad_norm": 7.031361218738439,
+      "learning_rate": 1e-06,
+      "loss": 0.4775,
+      "step": 6918
+    },
+    {
+      "epoch": 1.3230710392962999,
+      "grad_norm": 4.333927382807451,
+      "learning_rate": 1e-06,
+      "loss": 0.4158,
+      "step": 6919
+    },
+    {
+      "epoch": 1.323262262166555,
+      "grad_norm": 4.448939662147784,
+      "learning_rate": 1e-06,
+      "loss": 0.4064,
+      "step": 6920
+    },
+    {
+      "epoch": 1.3234534850368105,
+      "grad_norm": 2.2552371004896687,
+      "learning_rate": 1e-06,
+      "loss": 0.1078,
+      "step": 6921
+    },
+    {
+      "epoch": 1.3236447079070657,
+      "grad_norm": 2.0669970181706168,
+      "learning_rate": 1e-06,
+      "loss": 0.1382,
+      "step": 6922
+    },
+    {
+      "epoch": 1.3238359307773209,
+      "grad_norm": 2.280655038666634,
+      "learning_rate": 1e-06,
+      "loss": 0.0635,
+      "step": 6923
+    },
+    {
+      "epoch": 1.3240271536475763,
+      "grad_norm": 2.881238885225054,
+      "learning_rate": 1e-06,
+      "loss": 0.1308,
+      "step": 6924
+    },
+    {
+      "epoch": 1.3242183765178315,
+      "grad_norm": 3.1343147310070525,
+      "learning_rate": 1e-06,
+      "loss": 0.0941,
+      "step": 6925
+    },
+    {
+      "epoch": 1.3244095993880869,
+      "grad_norm": 9.17100528577212,
+      "learning_rate": 1e-06,
+      "loss": 0.0103,
+      "step": 6926
+    },
+    {
+      "epoch": 1.324600822258342,
+      "grad_norm": 3.2011607032008382,
+      "learning_rate": 1e-06,
+      "loss": 0.1716,
+      "step": 6927
+    },
+    {
+      "epoch": 1.3247920451285973,
+      "grad_norm": 3.894027264123896,
+      "learning_rate": 1e-06,
+      "loss": 0.0473,
+      "step": 6928
+    },
+    {
+      "epoch": 1.3249832679988527,
+      "grad_norm": 1.9558914124044078,
+      "learning_rate": 1e-06,
+      "loss": 0.0232,
+      "step": 6929
+    },
+    {
+      "epoch": 1.3251744908691079,
+      "grad_norm": 7.625693617994207,
+      "learning_rate": 1e-06,
+      "loss": 0.1596,
+      "step": 6930
+    },
+    {
+      "epoch": 1.3253657137393633,
+      "grad_norm": 4.1149757423473865,
+      "learning_rate": 1e-06,
+      "loss": 0.1972,
+      "step": 6931
+    },
+    {
+      "epoch": 1.3255569366096185,
+      "grad_norm": 4.418053799054273,
+      "learning_rate": 1e-06,
+      "loss": 0.2959,
+      "step": 6932
+    },
+    {
+      "epoch": 1.3257481594798737,
+      "grad_norm": 3.6257785092692587,
+      "learning_rate": 1e-06,
+      "loss": 0.2712,
+      "step": 6933
+    },
+    {
+      "epoch": 1.325939382350129,
+      "grad_norm": 2.0812649059048263,
+      "learning_rate": 1e-06,
+      "loss": 0.1022,
+      "step": 6934
+    },
+    {
+      "epoch": 1.3261306052203843,
+      "grad_norm": 2.4625843169208106,
+      "learning_rate": 1e-06,
+      "loss": 0.0887,
+      "step": 6935
+    },
+    {
+      "epoch": 1.3263218280906397,
+      "grad_norm": 2.2702933786509805,
+      "learning_rate": 1e-06,
+      "loss": 0.0844,
+      "step": 6936
+    },
+    {
+      "epoch": 1.326513050960895,
+      "grad_norm": 2.946945102217861,
+      "learning_rate": 1e-06,
+      "loss": 0.1477,
+      "step": 6937
+    },
+    {
+      "epoch": 1.32670427383115,
+      "grad_norm": 3.534065541987856,
+      "learning_rate": 1e-06,
+      "loss": 0.0928,
+      "step": 6938
+    },
+    {
+      "epoch": 1.3268954967014055,
+      "grad_norm": 2.968499745311601,
+      "learning_rate": 1e-06,
+      "loss": 0.163,
+      "step": 6939
+    },
+    {
+      "epoch": 1.3270867195716607,
+      "grad_norm": 3.3023461238292127,
+      "learning_rate": 1e-06,
+      "loss": 0.0783,
+      "step": 6940
+    },
+    {
+      "epoch": 1.327277942441916,
+      "grad_norm": 2.685157198291175,
+      "learning_rate": 1e-06,
+      "loss": 0.0372,
+      "step": 6941
+    },
+    {
+      "epoch": 1.3274691653121713,
+      "grad_norm": 3.566065426237514,
+      "learning_rate": 1e-06,
+      "loss": 0.0562,
+      "step": 6942
+    },
+    {
+      "epoch": 1.3276603881824265,
+      "grad_norm": 5.942144244004242,
+      "learning_rate": 1e-06,
+      "loss": 0.2789,
+      "step": 6943
+    },
+    {
+      "epoch": 1.327851611052682,
+      "grad_norm": 5.637966216236128,
+      "learning_rate": 1e-06,
+      "loss": 0.4549,
+      "step": 6944
+    },
+    {
+      "epoch": 1.3280428339229373,
+      "grad_norm": 3.0783615408697775,
+      "learning_rate": 1e-06,
+      "loss": 0.1568,
+      "step": 6945
+    },
+    {
+      "epoch": 1.3282340567931925,
+      "grad_norm": 2.613780614015944,
+      "learning_rate": 1e-06,
+      "loss": 0.1232,
+      "step": 6946
+    },
+    {
+      "epoch": 1.3284252796634477,
+      "grad_norm": 2.748781454436298,
+      "learning_rate": 1e-06,
+      "loss": 0.3121,
+      "step": 6947
+    },
+    {
+      "epoch": 1.328616502533703,
+      "grad_norm": 3.4311389675832373,
+      "learning_rate": 1e-06,
+      "loss": 0.2941,
+      "step": 6948
+    },
+    {
+      "epoch": 1.3288077254039583,
+      "grad_norm": 2.0923146124494476,
+      "learning_rate": 1e-06,
+      "loss": 0.1453,
+      "step": 6949
+    },
+    {
+      "epoch": 1.3289989482742137,
+      "grad_norm": 3.415239423678071,
+      "learning_rate": 1e-06,
+      "loss": 0.0673,
+      "step": 6950
+    },
+    {
+      "epoch": 1.329190171144469,
+      "grad_norm": 2.5763798211124653,
+      "learning_rate": 1e-06,
+      "loss": 0.0771,
+      "step": 6951
+    },
+    {
+      "epoch": 1.329381394014724,
+      "grad_norm": 3.177352010070404,
+      "learning_rate": 1e-06,
+      "loss": 0.1242,
+      "step": 6952
+    },
+    {
+      "epoch": 1.3295726168849795,
+      "grad_norm": 2.077449028959802,
+      "learning_rate": 1e-06,
+      "loss": 0.0668,
+      "step": 6953
+    },
+    {
+      "epoch": 1.3297638397552347,
+      "grad_norm": 2.5601159147007584,
+      "learning_rate": 1e-06,
+      "loss": 0.0401,
+      "step": 6954
+    },
+    {
+      "epoch": 1.3299550626254901,
+      "grad_norm": 3.9541243545203204,
+      "learning_rate": 1e-06,
+      "loss": 0.0941,
+      "step": 6955
+    },
+    {
+      "epoch": 1.3301462854957453,
+      "grad_norm": 4.189294459043813,
+      "learning_rate": 1e-06,
+      "loss": 0.223,
+      "step": 6956
+    },
+    {
+      "epoch": 1.3303375083660005,
+      "grad_norm": 3.962091582516986,
+      "learning_rate": 1e-06,
+      "loss": 0.1908,
+      "step": 6957
+    },
+    {
+      "epoch": 1.330528731236256,
+      "grad_norm": 2.5163905714388064,
+      "learning_rate": 1e-06,
+      "loss": 0.0664,
+      "step": 6958
+    },
+    {
+      "epoch": 1.3307199541065111,
+      "grad_norm": 4.609045812199114,
+      "learning_rate": 1e-06,
+      "loss": 0.1076,
+      "step": 6959
+    },
+    {
+      "epoch": 1.3309111769767665,
+      "grad_norm": 5.615658781515757,
+      "learning_rate": 1e-06,
+      "loss": 0.2321,
+      "step": 6960
+    },
+    {
+      "epoch": 1.3311023998470217,
+      "grad_norm": 2.2563357688262116,
+      "learning_rate": 1e-06,
+      "loss": 0.0925,
+      "step": 6961
+    },
+    {
+      "epoch": 1.331293622717277,
+      "grad_norm": 2.8423039039736664,
+      "learning_rate": 1e-06,
+      "loss": 0.1145,
+      "step": 6962
+    },
+    {
+      "epoch": 1.3314848455875323,
+      "grad_norm": 2.403176363432061,
+      "learning_rate": 1e-06,
+      "loss": 0.093,
+      "step": 6963
+    },
+    {
+      "epoch": 1.3316760684577875,
+      "grad_norm": 1.0628910467055164,
+      "learning_rate": 1e-06,
+      "loss": 0.02,
+      "step": 6964
+    },
+    {
+      "epoch": 1.331867291328043,
+      "grad_norm": 2.4140209206762577,
+      "learning_rate": 1e-06,
+      "loss": 0.0402,
+      "step": 6965
+    },
+    {
+      "epoch": 1.3320585141982981,
+      "grad_norm": 4.617511563619645,
+      "learning_rate": 1e-06,
+      "loss": 0.0774,
+      "step": 6966
+    },
+    {
+      "epoch": 1.3322497370685533,
+      "grad_norm": 2.112853622183035,
+      "learning_rate": 1e-06,
+      "loss": 0.03,
+      "step": 6967
+    },
+    {
+      "epoch": 1.3324409599388087,
+      "grad_norm": 5.182873972728596,
+      "learning_rate": 1e-06,
+      "loss": 0.4505,
+      "step": 6968
+    },
+    {
+      "epoch": 1.332632182809064,
+      "grad_norm": 4.568305724286823,
+      "learning_rate": 1e-06,
+      "loss": 0.3899,
+      "step": 6969
+    },
+    {
+      "epoch": 1.3328234056793193,
+      "grad_norm": 1.6882548586553623,
+      "learning_rate": 1e-06,
+      "loss": 0.1535,
+      "step": 6970
+    },
+    {
+      "epoch": 1.3330146285495745,
+      "grad_norm": 3.3076442900978478,
+      "learning_rate": 1e-06,
+      "loss": 0.2774,
+      "step": 6971
+    },
+    {
+      "epoch": 1.3332058514198297,
+      "grad_norm": 4.534567414400429,
+      "learning_rate": 1e-06,
+      "loss": 0.2971,
+      "step": 6972
+    },
+    {
+      "epoch": 1.3333970742900851,
+      "grad_norm": 3.345496176283474,
+      "learning_rate": 1e-06,
+      "loss": 0.1124,
+      "step": 6973
+    },
+    {
+      "epoch": 1.3335882971603403,
+      "grad_norm": 3.1128958699290736,
+      "learning_rate": 1e-06,
+      "loss": 0.1966,
+      "step": 6974
+    },
+    {
+      "epoch": 1.3337795200305957,
+      "grad_norm": 3.757563148521194,
+      "learning_rate": 1e-06,
+      "loss": 0.1349,
+      "step": 6975
+    },
+    {
+      "epoch": 1.333970742900851,
+      "grad_norm": 2.9214175269064353,
+      "learning_rate": 1e-06,
+      "loss": 0.0711,
+      "step": 6976
+    },
+    {
+      "epoch": 1.3341619657711061,
+      "grad_norm": 4.067094522908998,
+      "learning_rate": 1e-06,
+      "loss": 0.1902,
+      "step": 6977
+    },
+    {
+      "epoch": 1.3343531886413615,
+      "grad_norm": 4.433808846654519,
+      "learning_rate": 1e-06,
+      "loss": 0.1863,
+      "step": 6978
+    },
+    {
+      "epoch": 1.3345444115116167,
+      "grad_norm": 3.5319993565905494,
+      "learning_rate": 1e-06,
+      "loss": 0.0401,
+      "step": 6979
+    },
+    {
+      "epoch": 1.3347356343818721,
+      "grad_norm": 6.213622773486608,
+      "learning_rate": 1e-06,
+      "loss": 0.3544,
+      "step": 6980
+    },
+    {
+      "epoch": 1.3349268572521273,
+      "grad_norm": 6.234798713377834,
+      "learning_rate": 1e-06,
+      "loss": 0.5699,
+      "step": 6981
+    },
+    {
+      "epoch": 1.3351180801223825,
+      "grad_norm": 3.4092732583030467,
+      "learning_rate": 1e-06,
+      "loss": 0.2463,
+      "step": 6982
+    },
+    {
+      "epoch": 1.335309302992638,
+      "grad_norm": 5.430562471654747,
+      "learning_rate": 1e-06,
+      "loss": 0.2286,
+      "step": 6983
+    },
+    {
+      "epoch": 1.3355005258628931,
+      "grad_norm": 2.982591345668043,
+      "learning_rate": 1e-06,
+      "loss": 0.1682,
+      "step": 6984
+    },
+    {
+      "epoch": 1.3356917487331486,
+      "grad_norm": 4.858700033823529,
+      "learning_rate": 1e-06,
+      "loss": 0.2448,
+      "step": 6985
+    },
+    {
+      "epoch": 1.3358829716034037,
+      "grad_norm": 3.231049675699205,
+      "learning_rate": 1e-06,
+      "loss": 0.1203,
+      "step": 6986
+    },
+    {
+      "epoch": 1.336074194473659,
+      "grad_norm": 1.539875648357813,
+      "learning_rate": 1e-06,
+      "loss": 0.05,
+      "step": 6987
+    },
+    {
+      "epoch": 1.3362654173439144,
+      "grad_norm": 1.9852148966888084,
+      "learning_rate": 1e-06,
+      "loss": 0.1237,
+      "step": 6988
+    },
+    {
+      "epoch": 1.3364566402141695,
+      "grad_norm": 4.66086467464792,
+      "learning_rate": 1e-06,
+      "loss": 0.1015,
+      "step": 6989
+    },
+    {
+      "epoch": 1.336647863084425,
+      "grad_norm": 2.7956499368368597,
+      "learning_rate": 1e-06,
+      "loss": 0.0841,
+      "step": 6990
+    },
+    {
+      "epoch": 1.3368390859546802,
+      "grad_norm": 3.0074570482949325,
+      "learning_rate": 1e-06,
+      "loss": 0.068,
+      "step": 6991
+    },
+    {
+      "epoch": 1.3370303088249353,
+      "grad_norm": 4.561797832646947,
+      "learning_rate": 1e-06,
+      "loss": 0.1149,
+      "step": 6992
+    },
+    {
+      "epoch": 1.3372215316951908,
+      "grad_norm": 7.7151522881341466,
+      "learning_rate": 1e-06,
+      "loss": 0.281,
+      "step": 6993
+    },
+    {
+      "epoch": 1.337412754565446,
+      "grad_norm": 4.9175370733650725,
+      "learning_rate": 1e-06,
+      "loss": 0.2856,
+      "step": 6994
+    },
+    {
+      "epoch": 1.3376039774357014,
+      "grad_norm": 4.018362574197351,
+      "learning_rate": 1e-06,
+      "loss": 0.294,
+      "step": 6995
+    },
+    {
+      "epoch": 1.3377952003059566,
+      "grad_norm": 2.193385680288051,
+      "learning_rate": 1e-06,
+      "loss": 0.1166,
+      "step": 6996
+    },
+    {
+      "epoch": 1.3379864231762117,
+      "grad_norm": 2.253404796227074,
+      "learning_rate": 1e-06,
+      "loss": 0.0841,
+      "step": 6997
+    },
+    {
+      "epoch": 1.3381776460464672,
+      "grad_norm": 4.070049138252875,
+      "learning_rate": 1e-06,
+      "loss": 0.3291,
+      "step": 6998
+    },
+    {
+      "epoch": 1.3383688689167226,
+      "grad_norm": 1.7947251437259737,
+      "learning_rate": 1e-06,
+      "loss": 0.0644,
+      "step": 6999
+    },
+    {
+      "epoch": 1.3385600917869778,
+      "grad_norm": 2.7762455868331686,
+      "learning_rate": 1e-06,
+      "loss": 0.1041,
+      "step": 7000
+    },
+    {
+      "epoch": 1.3385600917869778,
+      "eval_runtime": 733.9072,
+      "eval_samples_per_second": 2.09,
+      "eval_steps_per_second": 0.523,
+      "step": 7000
+    },
+    {
+      "epoch": 1.338751314657233,
+      "grad_norm": 2.28496197257755,
+      "learning_rate": 1e-06,
+      "loss": 0.0823,
+      "step": 7001
+    },
+    {
+      "epoch": 1.3389425375274882,
+      "grad_norm": 2.478928070982197,
+      "learning_rate": 1e-06,
+      "loss": 0.0498,
+      "step": 7002
+    },
+    {
+      "epoch": 1.3391337603977436,
+      "grad_norm": 1.9563356179705926,
+      "learning_rate": 1e-06,
+      "loss": 0.0335,
+      "step": 7003
+    },
+    {
+      "epoch": 1.339324983267999,
+      "grad_norm": 2.1664067870564834,
+      "learning_rate": 1e-06,
+      "loss": 0.0285,
+      "step": 7004
+    },
+    {
+      "epoch": 1.3395162061382542,
+      "grad_norm": 4.376112878128161,
+      "learning_rate": 1e-06,
+      "loss": 0.0779,
+      "step": 7005
+    },
+    {
+      "epoch": 1.3397074290085094,
+      "grad_norm": 6.10382874196838,
+      "learning_rate": 1e-06,
+      "loss": 0.2204,
+      "step": 7006
+    },
+    {
+      "epoch": 1.3398986518787648,
+      "grad_norm": 5.559322692727944,
+      "learning_rate": 1e-06,
+      "loss": 0.54,
+      "step": 7007
+    },
+    {
+      "epoch": 1.34008987474902,
+      "grad_norm": 2.6813007981261143,
+      "learning_rate": 1e-06,
+      "loss": 0.1366,
+      "step": 7008
+    },
+    {
+      "epoch": 1.3402810976192754,
+      "grad_norm": 1.6426826289586256,
+      "learning_rate": 1e-06,
+      "loss": 0.0712,
+      "step": 7009
+    },
+    {
+      "epoch": 1.3404723204895306,
+      "grad_norm": 2.269215967336471,
+      "learning_rate": 1e-06,
+      "loss": 0.1366,
+      "step": 7010
+    },
+    {
+      "epoch": 1.3406635433597858,
+      "grad_norm": 4.260133667044261,
+      "learning_rate": 1e-06,
+      "loss": 0.1541,
+      "step": 7011
+    },
+    {
+      "epoch": 1.3408547662300412,
+      "grad_norm": 3.9185471721037684,
+      "learning_rate": 1e-06,
+      "loss": 0.253,
+      "step": 7012
+    },
+    {
+      "epoch": 1.3410459891002964,
+      "grad_norm": 2.9267288120017994,
+      "learning_rate": 1e-06,
+      "loss": 0.1563,
+      "step": 7013
+    },
+    {
+      "epoch": 1.3412372119705518,
+      "grad_norm": 1.8458270641877752,
+      "learning_rate": 1e-06,
+      "loss": 0.0286,
+      "step": 7014
+    },
+    {
+      "epoch": 1.341428434840807,
+      "grad_norm": 3.865177873118484,
+      "learning_rate": 1e-06,
+      "loss": 0.0285,
+      "step": 7015
+    },
+    {
+      "epoch": 1.3416196577110622,
+      "grad_norm": 7.031752097958138,
+      "learning_rate": 1e-06,
+      "loss": 0.0668,
+      "step": 7016
+    },
+    {
+      "epoch": 1.3418108805813176,
+      "grad_norm": 4.231989677834857,
+      "learning_rate": 1e-06,
+      "loss": 0.0514,
+      "step": 7017
+    },
+    {
+      "epoch": 1.3420021034515728,
+      "grad_norm": 5.369638297119698,
+      "learning_rate": 1e-06,
+      "loss": 0.1152,
+      "step": 7018
+    },
+    {
+      "epoch": 1.3421933263218282,
+      "grad_norm": 4.93126816353396,
+      "learning_rate": 1e-06,
+      "loss": 0.3471,
+      "step": 7019
+    },
+    {
+      "epoch": 1.3423845491920834,
+      "grad_norm": 4.002798055479344,
+      "learning_rate": 1e-06,
+      "loss": 0.2878,
+      "step": 7020
+    },
+    {
+      "epoch": 1.3425757720623386,
+      "grad_norm": 4.399337649343329,
+      "learning_rate": 1e-06,
+      "loss": 0.4579,
+      "step": 7021
+    },
+    {
+      "epoch": 1.342766994932594,
+      "grad_norm": 3.264047634151164,
+      "learning_rate": 1e-06,
+      "loss": 0.1401,
+      "step": 7022
+    },
+    {
+      "epoch": 1.3429582178028492,
+      "grad_norm": 3.1540394208497022,
+      "learning_rate": 1e-06,
+      "loss": 0.0823,
+      "step": 7023
+    },
+    {
+      "epoch": 1.3431494406731046,
+      "grad_norm": 3.7127909639438847,
+      "learning_rate": 1e-06,
+      "loss": 0.2534,
+      "step": 7024
+    },
+    {
+      "epoch": 1.3433406635433598,
+      "grad_norm": 2.772157615523763,
+      "learning_rate": 1e-06,
+      "loss": 0.0581,
+      "step": 7025
+    },
+    {
+      "epoch": 1.343531886413615,
+      "grad_norm": 3.1547721197711804,
+      "learning_rate": 1e-06,
+      "loss": 0.1228,
+      "step": 7026
+    },
+    {
+      "epoch": 1.3437231092838704,
+      "grad_norm": 2.8626596731104335,
+      "learning_rate": 1e-06,
+      "loss": 0.0564,
+      "step": 7027
+    },
+    {
+      "epoch": 1.3439143321541256,
+      "grad_norm": 2.2191971878930548,
+      "learning_rate": 1e-06,
+      "loss": 0.049,
+      "step": 7028
+    },
+    {
+      "epoch": 1.344105555024381,
+      "grad_norm": 1.7649450427568634,
+      "learning_rate": 1e-06,
+      "loss": 0.019,
+      "step": 7029
+    },
+    {
+      "epoch": 1.3442967778946362,
+      "grad_norm": 16.092205440141957,
+      "learning_rate": 1e-06,
+      "loss": 0.2976,
+      "step": 7030
+    },
+    {
+      "epoch": 1.3444880007648914,
+      "grad_norm": 6.106314358328157,
+      "learning_rate": 1e-06,
+      "loss": 0.3927,
+      "step": 7031
+    },
+    {
+      "epoch": 1.3446792236351468,
+      "grad_norm": 5.310087666048289,
+      "learning_rate": 1e-06,
+      "loss": 0.2855,
+      "step": 7032
+    },
+    {
+      "epoch": 1.344870446505402,
+      "grad_norm": 3.0198500688231835,
+      "learning_rate": 1e-06,
+      "loss": 0.2372,
+      "step": 7033
+    },
+    {
+      "epoch": 1.3450616693756574,
+      "grad_norm": 4.031396907524856,
+      "learning_rate": 1e-06,
+      "loss": 0.15,
+      "step": 7034
+    },
+    {
+      "epoch": 1.3452528922459126,
+      "grad_norm": 2.2723224435397698,
+      "learning_rate": 1e-06,
+      "loss": 0.096,
+      "step": 7035
+    },
+    {
+      "epoch": 1.3454441151161678,
+      "grad_norm": 3.2959512437899368,
+      "learning_rate": 1e-06,
+      "loss": 0.1001,
+      "step": 7036
+    },
+    {
+      "epoch": 1.3456353379864232,
+      "grad_norm": 2.4082183713654643,
+      "learning_rate": 1e-06,
+      "loss": 0.1056,
+      "step": 7037
+    },
+    {
+      "epoch": 1.3458265608566784,
+      "grad_norm": 2.642371959778188,
+      "learning_rate": 1e-06,
+      "loss": 0.041,
+      "step": 7038
+    },
+    {
+      "epoch": 1.3460177837269338,
+      "grad_norm": 2.10500566984668,
+      "learning_rate": 1e-06,
+      "loss": 0.1315,
+      "step": 7039
+    },
+    {
+      "epoch": 1.346209006597189,
+      "grad_norm": 1.8857447793588156,
+      "learning_rate": 1e-06,
+      "loss": 0.0376,
+      "step": 7040
+    },
+    {
+      "epoch": 1.3464002294674442,
+      "grad_norm": 2.8862788338788357,
+      "learning_rate": 1e-06,
+      "loss": 0.112,
+      "step": 7041
+    },
+    {
+      "epoch": 1.3465914523376996,
+      "grad_norm": 5.865978700585492,
+      "learning_rate": 1e-06,
+      "loss": 0.0475,
+      "step": 7042
+    },
+    {
+      "epoch": 1.3467826752079548,
+      "grad_norm": 4.477511267123377,
+      "learning_rate": 1e-06,
+      "loss": 0.1517,
+      "step": 7043
+    },
+    {
+      "epoch": 1.3469738980782102,
+      "grad_norm": 4.317367018709802,
+      "learning_rate": 1e-06,
+      "loss": 0.2282,
+      "step": 7044
+    },
+    {
+      "epoch": 1.3471651209484654,
+      "grad_norm": 2.8712538486400114,
+      "learning_rate": 1e-06,
+      "loss": 0.1044,
+      "step": 7045
+    },
+    {
+      "epoch": 1.3473563438187206,
+      "grad_norm": 1.958998127638977,
+      "learning_rate": 1e-06,
+      "loss": 0.0674,
+      "step": 7046
+    },
+    {
+      "epoch": 1.347547566688976,
+      "grad_norm": 2.6338918946601333,
+      "learning_rate": 1e-06,
+      "loss": 0.1677,
+      "step": 7047
+    },
+    {
+      "epoch": 1.3477387895592312,
+      "grad_norm": 1.8617828550199444,
+      "learning_rate": 1e-06,
+      "loss": 0.1174,
+      "step": 7048
+    },
+    {
+      "epoch": 1.3479300124294866,
+      "grad_norm": 2.9618966755795317,
+      "learning_rate": 1e-06,
+      "loss": 0.0544,
+      "step": 7049
+    },
+    {
+      "epoch": 1.3481212352997418,
+      "grad_norm": 2.7702723535166522,
+      "learning_rate": 1e-06,
+      "loss": 0.0615,
+      "step": 7050
+    },
+    {
+      "epoch": 1.348312458169997,
+      "grad_norm": 3.8545278156606497,
+      "learning_rate": 1e-06,
+      "loss": 0.347,
+      "step": 7051
+    },
+    {
+      "epoch": 1.3485036810402524,
+      "grad_norm": 4.641923613544753,
+      "learning_rate": 1e-06,
+      "loss": 0.1485,
+      "step": 7052
+    },
+    {
+      "epoch": 1.3486949039105076,
+      "grad_norm": 2.694253199192883,
+      "learning_rate": 1e-06,
+      "loss": 0.0834,
+      "step": 7053
+    },
+    {
+      "epoch": 1.348886126780763,
+      "grad_norm": 2.5138230593254156,
+      "learning_rate": 1e-06,
+      "loss": 0.0251,
+      "step": 7054
+    },
+    {
+      "epoch": 1.3490773496510182,
+      "grad_norm": 4.841895357509926,
+      "learning_rate": 1e-06,
+      "loss": 0.07,
+      "step": 7055
+    },
+    {
+      "epoch": 1.3492685725212734,
+      "grad_norm": 4.048642277075085,
+      "learning_rate": 1e-06,
+      "loss": 0.3826,
+      "step": 7056
+    },
+    {
+      "epoch": 1.3494597953915288,
+      "grad_norm": 4.570554756199617,
+      "learning_rate": 1e-06,
+      "loss": 0.2971,
+      "step": 7057
+    },
+    {
+      "epoch": 1.3496510182617842,
+      "grad_norm": 1.781671975998897,
+      "learning_rate": 1e-06,
+      "loss": 0.0774,
+      "step": 7058
+    },
+    {
+      "epoch": 1.3498422411320394,
+      "grad_norm": 2.7228382109994724,
+      "learning_rate": 1e-06,
+      "loss": 0.0914,
+      "step": 7059
+    },
+    {
+      "epoch": 1.3500334640022946,
+      "grad_norm": 4.60659116312323,
+      "learning_rate": 1e-06,
+      "loss": 0.224,
+      "step": 7060
+    },
+    {
+      "epoch": 1.35022468687255,
+      "grad_norm": 3.0187870505134438,
+      "learning_rate": 1e-06,
+      "loss": 0.0908,
+      "step": 7061
+    },
+    {
+      "epoch": 1.3504159097428052,
+      "grad_norm": 3.459157929678623,
+      "learning_rate": 1e-06,
+      "loss": 0.0836,
+      "step": 7062
+    },
+    {
+      "epoch": 1.3506071326130606,
+      "grad_norm": 3.299838356914274,
+      "learning_rate": 1e-06,
+      "loss": 0.1564,
+      "step": 7063
+    },
+    {
+      "epoch": 1.3507983554833158,
+      "grad_norm": 1.582339861874358,
+      "learning_rate": 1e-06,
+      "loss": 0.0296,
+      "step": 7064
+    },
+    {
+      "epoch": 1.350989578353571,
+      "grad_norm": 3.0054567301593864,
+      "learning_rate": 1e-06,
+      "loss": 0.0791,
+      "step": 7065
+    },
+    {
+      "epoch": 1.3511808012238264,
+      "grad_norm": 2.544447883268621,
+      "learning_rate": 1e-06,
+      "loss": 0.0465,
+      "step": 7066
+    },
+    {
+      "epoch": 1.3513720240940816,
+      "grad_norm": 4.666020121518344,
+      "learning_rate": 1e-06,
+      "loss": 0.0635,
+      "step": 7067
+    },
+    {
+      "epoch": 1.351563246964337,
+      "grad_norm": 3.6011117913823787,
+      "learning_rate": 1e-06,
+      "loss": 0.1785,
+      "step": 7068
+    },
+    {
+      "epoch": 1.3517544698345922,
+      "grad_norm": 4.174090422268244,
+      "learning_rate": 1e-06,
+      "loss": 0.418,
+      "step": 7069
+    },
+    {
+      "epoch": 1.3519456927048474,
+      "grad_norm": 3.1348684513898717,
+      "learning_rate": 1e-06,
+      "loss": 0.0976,
+      "step": 7070
+    },
+    {
+      "epoch": 1.3521369155751028,
+      "grad_norm": 2.647154911017639,
+      "learning_rate": 1e-06,
+      "loss": 0.2533,
+      "step": 7071
+    },
+    {
+      "epoch": 1.352328138445358,
+      "grad_norm": 4.150900933701185,
+      "learning_rate": 1e-06,
+      "loss": 0.2062,
+      "step": 7072
+    },
+    {
+      "epoch": 1.3525193613156135,
+      "grad_norm": 2.4738763628313034,
+      "learning_rate": 1e-06,
+      "loss": 0.0574,
+      "step": 7073
+    },
+    {
+      "epoch": 1.3527105841858686,
+      "grad_norm": 4.631925604135509,
+      "learning_rate": 1e-06,
+      "loss": 0.093,
+      "step": 7074
+    },
+    {
+      "epoch": 1.3529018070561238,
+      "grad_norm": 0.9719441576328552,
+      "learning_rate": 1e-06,
+      "loss": 0.0094,
+      "step": 7075
+    },
+    {
+      "epoch": 1.3530930299263793,
+      "grad_norm": 2.85916387059931,
+      "learning_rate": 1e-06,
+      "loss": 0.0955,
+      "step": 7076
+    },
+    {
+      "epoch": 1.3532842527966344,
+      "grad_norm": 1.1912960314188,
+      "learning_rate": 1e-06,
+      "loss": 0.0192,
+      "step": 7077
+    },
+    {
+      "epoch": 1.3534754756668899,
+      "grad_norm": 6.667148604457232,
+      "learning_rate": 1e-06,
+      "loss": 0.2044,
+      "step": 7078
+    },
+    {
+      "epoch": 1.353666698537145,
+      "grad_norm": 2.729907413351895,
+      "learning_rate": 1e-06,
+      "loss": 0.0983,
+      "step": 7079
+    },
+    {
+      "epoch": 1.3538579214074002,
+      "grad_norm": 6.33910120613029,
+      "learning_rate": 1e-06,
+      "loss": 0.1683,
+      "step": 7080
+    },
+    {
+      "epoch": 1.3540491442776557,
+      "grad_norm": 4.404451043587763,
+      "learning_rate": 1e-06,
+      "loss": 0.1647,
+      "step": 7081
+    },
+    {
+      "epoch": 1.3542403671479109,
+      "grad_norm": 5.124595626136847,
+      "learning_rate": 1e-06,
+      "loss": 0.2368,
+      "step": 7082
+    },
+    {
+      "epoch": 1.3544315900181663,
+      "grad_norm": 3.4669740271639324,
+      "learning_rate": 1e-06,
+      "loss": 0.1094,
+      "step": 7083
+    },
+    {
+      "epoch": 1.3546228128884215,
+      "grad_norm": 4.040443995788763,
+      "learning_rate": 1e-06,
+      "loss": 0.2769,
+      "step": 7084
+    },
+    {
+      "epoch": 1.3548140357586766,
+      "grad_norm": 3.2454868672031303,
+      "learning_rate": 1e-06,
+      "loss": 0.1996,
+      "step": 7085
+    },
+    {
+      "epoch": 1.355005258628932,
+      "grad_norm": 2.5740857473464986,
+      "learning_rate": 1e-06,
+      "loss": 0.133,
+      "step": 7086
+    },
+    {
+      "epoch": 1.3551964814991873,
+      "grad_norm": 2.1693708954653848,
+      "learning_rate": 1e-06,
+      "loss": 0.1155,
+      "step": 7087
+    },
+    {
+      "epoch": 1.3553877043694427,
+      "grad_norm": 1.9782921984069572,
+      "learning_rate": 1e-06,
+      "loss": 0.0174,
+      "step": 7088
+    },
+    {
+      "epoch": 1.3555789272396979,
+      "grad_norm": 9.101572139882235,
+      "learning_rate": 1e-06,
+      "loss": 0.0721,
+      "step": 7089
+    },
+    {
+      "epoch": 1.355770150109953,
+      "grad_norm": 5.403696328045952,
+      "learning_rate": 1e-06,
+      "loss": 0.037,
+      "step": 7090
+    },
+    {
+      "epoch": 1.3559613729802085,
+      "grad_norm": 3.7633933898170353,
+      "learning_rate": 1e-06,
+      "loss": 0.0457,
+      "step": 7091
+    },
+    {
+      "epoch": 1.3561525958504637,
+      "grad_norm": 4.234921353273069,
+      "learning_rate": 1e-06,
+      "loss": 0.0961,
+      "step": 7092
+    },
+    {
+      "epoch": 1.356343818720719,
+      "grad_norm": 4.042008113530173,
+      "learning_rate": 1e-06,
+      "loss": 0.122,
+      "step": 7093
+    },
+    {
+      "epoch": 1.3565350415909743,
+      "grad_norm": 4.181905168827296,
+      "learning_rate": 1e-06,
+      "loss": 0.2318,
+      "step": 7094
+    },
+    {
+      "epoch": 1.3567262644612295,
+      "grad_norm": 6.869084535596992,
+      "learning_rate": 1e-06,
+      "loss": 0.5752,
+      "step": 7095
+    },
+    {
+      "epoch": 1.3569174873314849,
+      "grad_norm": 1.7237241034606072,
+      "learning_rate": 1e-06,
+      "loss": 0.089,
+      "step": 7096
+    },
+    {
+      "epoch": 1.35710871020174,
+      "grad_norm": 3.156716793780794,
+      "learning_rate": 1e-06,
+      "loss": 0.2344,
+      "step": 7097
+    },
+    {
+      "epoch": 1.3572999330719955,
+      "grad_norm": 3.521173735345546,
+      "learning_rate": 1e-06,
+      "loss": 0.2534,
+      "step": 7098
+    },
+    {
+      "epoch": 1.3574911559422507,
+      "grad_norm": 2.5172756300403845,
+      "learning_rate": 1e-06,
+      "loss": 0.0522,
+      "step": 7099
+    },
+    {
+      "epoch": 1.3576823788125059,
+      "grad_norm": 4.227865585861189,
+      "learning_rate": 1e-06,
+      "loss": 0.1607,
+      "step": 7100
+    },
+    {
+      "epoch": 1.3578736016827613,
+      "grad_norm": 3.4770680263602154,
+      "learning_rate": 1e-06,
+      "loss": 0.1233,
+      "step": 7101
+    },
+    {
+      "epoch": 1.3580648245530165,
+      "grad_norm": 4.295289124959132,
+      "learning_rate": 1e-06,
+      "loss": 0.1228,
+      "step": 7102
+    },
+    {
+      "epoch": 1.3582560474232719,
+      "grad_norm": 1.5048639592122883,
+      "learning_rate": 1e-06,
+      "loss": 0.0231,
+      "step": 7103
+    },
+    {
+      "epoch": 1.358447270293527,
+      "grad_norm": 1.7696058918618596,
+      "learning_rate": 1e-06,
+      "loss": 0.041,
+      "step": 7104
+    },
+    {
+      "epoch": 1.3586384931637823,
+      "grad_norm": 5.096200552450893,
+      "learning_rate": 1e-06,
+      "loss": 0.1049,
+      "step": 7105
+    },
+    {
+      "epoch": 1.3588297160340377,
+      "grad_norm": 5.912957474992884,
+      "learning_rate": 1e-06,
+      "loss": 0.416,
+      "step": 7106
+    },
+    {
+      "epoch": 1.3590209389042929,
+      "grad_norm": 6.992620120170946,
+      "learning_rate": 1e-06,
+      "loss": 0.2682,
+      "step": 7107
+    },
+    {
+      "epoch": 1.3592121617745483,
+      "grad_norm": 4.74100425576158,
+      "learning_rate": 1e-06,
+      "loss": 0.3051,
+      "step": 7108
+    },
+    {
+      "epoch": 1.3594033846448035,
+      "grad_norm": 1.7660911417346774,
+      "learning_rate": 1e-06,
+      "loss": 0.0755,
+      "step": 7109
+    },
+    {
+      "epoch": 1.3595946075150587,
+      "grad_norm": 4.774286122908684,
+      "learning_rate": 1e-06,
+      "loss": 0.4242,
+      "step": 7110
+    },
+    {
+      "epoch": 1.359785830385314,
+      "grad_norm": 2.7337779237902233,
+      "learning_rate": 1e-06,
+      "loss": 0.1736,
+      "step": 7111
+    },
+    {
+      "epoch": 1.3599770532555695,
+      "grad_norm": 2.4302804106940563,
+      "learning_rate": 1e-06,
+      "loss": 0.0836,
+      "step": 7112
+    },
+    {
+      "epoch": 1.3601682761258247,
+      "grad_norm": 3.211830775989568,
+      "learning_rate": 1e-06,
+      "loss": 0.2383,
+      "step": 7113
+    },
+    {
+      "epoch": 1.3603594989960799,
+      "grad_norm": 3.361558213635044,
+      "learning_rate": 1e-06,
+      "loss": 0.0781,
+      "step": 7114
+    },
+    {
+      "epoch": 1.360550721866335,
+      "grad_norm": 2.550809383967407,
+      "learning_rate": 1e-06,
+      "loss": 0.0776,
+      "step": 7115
+    },
+    {
+      "epoch": 1.3607419447365905,
+      "grad_norm": 2.86773975677031,
+      "learning_rate": 1e-06,
+      "loss": 0.1954,
+      "step": 7116
+    },
+    {
+      "epoch": 1.360933167606846,
+      "grad_norm": 6.4649853647631375,
+      "learning_rate": 1e-06,
+      "loss": 0.1438,
+      "step": 7117
+    },
+    {
+      "epoch": 1.361124390477101,
+      "grad_norm": 15.739548379121649,
+      "learning_rate": 1e-06,
+      "loss": 0.4582,
+      "step": 7118
+    },
+    {
+      "epoch": 1.3613156133473563,
+      "grad_norm": 3.9612992401770217,
+      "learning_rate": 1e-06,
+      "loss": 0.1758,
+      "step": 7119
+    },
+    {
+      "epoch": 1.3615068362176117,
+      "grad_norm": 3.3110947417239323,
+      "learning_rate": 1e-06,
+      "loss": 0.247,
+      "step": 7120
+    },
+    {
+      "epoch": 1.361698059087867,
+      "grad_norm": 4.182369904441515,
+      "learning_rate": 1e-06,
+      "loss": 0.2929,
+      "step": 7121
+    },
+    {
+      "epoch": 1.3618892819581223,
+      "grad_norm": 2.5686591027877226,
+      "learning_rate": 1e-06,
+      "loss": 0.1734,
+      "step": 7122
+    },
+    {
+      "epoch": 1.3620805048283775,
+      "grad_norm": 2.3780134057067355,
+      "learning_rate": 1e-06,
+      "loss": 0.0659,
+      "step": 7123
+    },
+    {
+      "epoch": 1.3622717276986327,
+      "grad_norm": 2.2968452088376776,
+      "learning_rate": 1e-06,
+      "loss": 0.09,
+      "step": 7124
+    },
+    {
+      "epoch": 1.362462950568888,
+      "grad_norm": 1.7387293548784386,
+      "learning_rate": 1e-06,
+      "loss": 0.0781,
+      "step": 7125
+    },
+    {
+      "epoch": 1.3626541734391433,
+      "grad_norm": 2.0037965502837856,
+      "learning_rate": 1e-06,
+      "loss": 0.0461,
+      "step": 7126
+    },
+    {
+      "epoch": 1.3628453963093987,
+      "grad_norm": 5.481221479735344,
+      "learning_rate": 1e-06,
+      "loss": 0.2444,
+      "step": 7127
+    },
+    {
+      "epoch": 1.363036619179654,
+      "grad_norm": 4.217766767294737,
+      "learning_rate": 1e-06,
+      "loss": 0.2459,
+      "step": 7128
+    },
+    {
+      "epoch": 1.363227842049909,
+      "grad_norm": 3.1188767523990544,
+      "learning_rate": 1e-06,
+      "loss": 0.0621,
+      "step": 7129
+    },
+    {
+      "epoch": 1.3634190649201645,
+      "grad_norm": 5.111814144176654,
+      "learning_rate": 1e-06,
+      "loss": 0.1054,
+      "step": 7130
+    },
+    {
+      "epoch": 1.3636102877904197,
+      "grad_norm": 5.648160766744965,
+      "learning_rate": 1e-06,
+      "loss": 0.6313,
+      "step": 7131
+    },
+    {
+      "epoch": 1.3638015106606751,
+      "grad_norm": 2.741131962480096,
+      "learning_rate": 1e-06,
+      "loss": 0.0732,
+      "step": 7132
+    },
+    {
+      "epoch": 1.3639927335309303,
+      "grad_norm": 5.089537019189677,
+      "learning_rate": 1e-06,
+      "loss": 0.2555,
+      "step": 7133
+    },
+    {
+      "epoch": 1.3641839564011855,
+      "grad_norm": 3.1289755900458904,
+      "learning_rate": 1e-06,
+      "loss": 0.1326,
+      "step": 7134
+    },
+    {
+      "epoch": 1.364375179271441,
+      "grad_norm": 4.033598696173488,
+      "learning_rate": 1e-06,
+      "loss": 0.2854,
+      "step": 7135
+    },
+    {
+      "epoch": 1.364566402141696,
+      "grad_norm": 1.9756663707250008,
+      "learning_rate": 1e-06,
+      "loss": 0.084,
+      "step": 7136
+    },
+    {
+      "epoch": 1.3647576250119515,
+      "grad_norm": 3.8540433434353867,
+      "learning_rate": 1e-06,
+      "loss": 0.1148,
+      "step": 7137
+    },
+    {
+      "epoch": 1.3649488478822067,
+      "grad_norm": 3.6839493289656184,
+      "learning_rate": 1e-06,
+      "loss": 0.0659,
+      "step": 7138
+    },
+    {
+      "epoch": 1.365140070752462,
+      "grad_norm": 5.192530754927614,
+      "learning_rate": 1e-06,
+      "loss": 0.1798,
+      "step": 7139
+    },
+    {
+      "epoch": 1.3653312936227173,
+      "grad_norm": 1.8906441049162235,
+      "learning_rate": 1e-06,
+      "loss": 0.0355,
+      "step": 7140
+    },
+    {
+      "epoch": 1.3655225164929725,
+      "grad_norm": 2.1518482115510427,
+      "learning_rate": 1e-06,
+      "loss": 0.0155,
+      "step": 7141
+    },
+    {
+      "epoch": 1.365713739363228,
+      "grad_norm": 6.215859206108679,
+      "learning_rate": 1e-06,
+      "loss": 0.0518,
+      "step": 7142
+    },
+    {
+      "epoch": 1.3659049622334831,
+      "grad_norm": 3.7952662697964663,
+      "learning_rate": 1e-06,
+      "loss": 0.1097,
+      "step": 7143
+    },
+    {
+      "epoch": 1.3660961851037383,
+      "grad_norm": 3.9862509705526485,
+      "learning_rate": 1e-06,
+      "loss": 0.4479,
+      "step": 7144
+    },
+    {
+      "epoch": 1.3662874079739937,
+      "grad_norm": 3.076361601292068,
+      "learning_rate": 1e-06,
+      "loss": 0.0839,
+      "step": 7145
+    },
+    {
+      "epoch": 1.366478630844249,
+      "grad_norm": 2.460280460869548,
+      "learning_rate": 1e-06,
+      "loss": 0.0895,
+      "step": 7146
+    },
+    {
+      "epoch": 1.3666698537145043,
+      "grad_norm": 2.6149058858591134,
+      "learning_rate": 1e-06,
+      "loss": 0.1664,
+      "step": 7147
+    },
+    {
+      "epoch": 1.3668610765847595,
+      "grad_norm": 3.1742924840589324,
+      "learning_rate": 1e-06,
+      "loss": 0.1758,
+      "step": 7148
+    },
+    {
+      "epoch": 1.3670522994550147,
+      "grad_norm": 3.76436774448061,
+      "learning_rate": 1e-06,
+      "loss": 0.2882,
+      "step": 7149
+    },
+    {
+      "epoch": 1.3672435223252701,
+      "grad_norm": 3.3816577094423033,
+      "learning_rate": 1e-06,
+      "loss": 0.2119,
+      "step": 7150
+    },
+    {
+      "epoch": 1.3674347451955253,
+      "grad_norm": 2.592409545880156,
+      "learning_rate": 1e-06,
+      "loss": 0.0714,
+      "step": 7151
+    },
+    {
+      "epoch": 1.3676259680657807,
+      "grad_norm": 3.4635776756919654,
+      "learning_rate": 1e-06,
+      "loss": 0.05,
+      "step": 7152
+    },
+    {
+      "epoch": 1.367817190936036,
+      "grad_norm": 1.922886280000195,
+      "learning_rate": 1e-06,
+      "loss": 0.0685,
+      "step": 7153
+    },
+    {
+      "epoch": 1.3680084138062911,
+      "grad_norm": 3.657565801188935,
+      "learning_rate": 1e-06,
+      "loss": 0.0708,
+      "step": 7154
+    },
+    {
+      "epoch": 1.3681996366765465,
+      "grad_norm": 7.236559280636619,
+      "learning_rate": 1e-06,
+      "loss": 0.1368,
+      "step": 7155
+    },
+    {
+      "epoch": 1.3683908595468017,
+      "grad_norm": 55.365204851089686,
+      "learning_rate": 1e-06,
+      "loss": 0.517,
+      "step": 7156
+    },
+    {
+      "epoch": 1.3685820824170571,
+      "grad_norm": 4.457499821310419,
+      "learning_rate": 1e-06,
+      "loss": 0.3531,
+      "step": 7157
+    },
+    {
+      "epoch": 1.3687733052873123,
+      "grad_norm": 2.392127568386828,
+      "learning_rate": 1e-06,
+      "loss": 0.1025,
+      "step": 7158
+    },
+    {
+      "epoch": 1.3689645281575675,
+      "grad_norm": 3.7386837924922376,
+      "learning_rate": 1e-06,
+      "loss": 0.1868,
+      "step": 7159
+    },
+    {
+      "epoch": 1.369155751027823,
+      "grad_norm": 2.624501408046607,
+      "learning_rate": 1e-06,
+      "loss": 0.2114,
+      "step": 7160
+    },
+    {
+      "epoch": 1.3693469738980781,
+      "grad_norm": 2.3454833678154126,
+      "learning_rate": 1e-06,
+      "loss": 0.2016,
+      "step": 7161
+    },
+    {
+      "epoch": 1.3695381967683335,
+      "grad_norm": 2.452708020579765,
+      "learning_rate": 1e-06,
+      "loss": 0.1052,
+      "step": 7162
+    },
+    {
+      "epoch": 1.3697294196385887,
+      "grad_norm": 2.3926227674406575,
+      "learning_rate": 1e-06,
+      "loss": 0.0489,
+      "step": 7163
+    },
+    {
+      "epoch": 1.369920642508844,
+      "grad_norm": 5.149354334976247,
+      "learning_rate": 1e-06,
+      "loss": 0.133,
+      "step": 7164
+    },
+    {
+      "epoch": 1.3701118653790993,
+      "grad_norm": 4.542337047318311,
+      "learning_rate": 1e-06,
+      "loss": 0.2788,
+      "step": 7165
+    },
+    {
+      "epoch": 1.3703030882493548,
+      "grad_norm": 3.742244329797477,
+      "learning_rate": 1e-06,
+      "loss": 0.1145,
+      "step": 7166
+    },
+    {
+      "epoch": 1.37049431111961,
+      "grad_norm": 3.987387561707633,
+      "learning_rate": 1e-06,
+      "loss": 0.0751,
+      "step": 7167
+    },
+    {
+      "epoch": 1.3706855339898651,
+      "grad_norm": 6.076398511122093,
+      "learning_rate": 1e-06,
+      "loss": 0.3998,
+      "step": 7168
+    },
+    {
+      "epoch": 1.3708767568601203,
+      "grad_norm": 3.4426145478294194,
+      "learning_rate": 1e-06,
+      "loss": 0.1093,
+      "step": 7169
+    },
+    {
+      "epoch": 1.3710679797303758,
+      "grad_norm": 3.523147198883061,
+      "learning_rate": 1e-06,
+      "loss": 0.1495,
+      "step": 7170
+    },
+    {
+      "epoch": 1.3712592026006312,
+      "grad_norm": 3.713772174944906,
+      "learning_rate": 1e-06,
+      "loss": 0.1258,
+      "step": 7171
+    },
+    {
+      "epoch": 1.3714504254708864,
+      "grad_norm": 3.260395590294589,
+      "learning_rate": 1e-06,
+      "loss": 0.3113,
+      "step": 7172
+    },
+    {
+      "epoch": 1.3716416483411416,
+      "grad_norm": 3.6767869925385908,
+      "learning_rate": 1e-06,
+      "loss": 0.2826,
+      "step": 7173
+    },
+    {
+      "epoch": 1.371832871211397,
+      "grad_norm": 3.725962464590407,
+      "learning_rate": 1e-06,
+      "loss": 0.2861,
+      "step": 7174
+    },
+    {
+      "epoch": 1.3720240940816522,
+      "grad_norm": 2.199563685413306,
+      "learning_rate": 1e-06,
+      "loss": 0.0927,
+      "step": 7175
+    },
+    {
+      "epoch": 1.3722153169519076,
+      "grad_norm": 3.3595257392597877,
+      "learning_rate": 1e-06,
+      "loss": 0.1133,
+      "step": 7176
+    },
+    {
+      "epoch": 1.3724065398221628,
+      "grad_norm": 2.106535066506057,
+      "learning_rate": 1e-06,
+      "loss": 0.0739,
+      "step": 7177
+    },
+    {
+      "epoch": 1.372597762692418,
+      "grad_norm": 2.0029523040074886,
+      "learning_rate": 1e-06,
+      "loss": 0.076,
+      "step": 7178
+    },
+    {
+      "epoch": 1.3727889855626734,
+      "grad_norm": 5.060237426012458,
+      "learning_rate": 1e-06,
+      "loss": 0.2109,
+      "step": 7179
+    },
+    {
+      "epoch": 1.3729802084329286,
+      "grad_norm": 4.372229870783185,
+      "learning_rate": 1e-06,
+      "loss": 0.1363,
+      "step": 7180
+    },
+    {
+      "epoch": 1.373171431303184,
+      "grad_norm": 5.550365456585733,
+      "learning_rate": 1e-06,
+      "loss": 0.1293,
+      "step": 7181
+    },
+    {
+      "epoch": 1.3733626541734392,
+      "grad_norm": 5.004889958074167,
+      "learning_rate": 1e-06,
+      "loss": 0.4529,
+      "step": 7182
+    },
+    {
+      "epoch": 1.3735538770436944,
+      "grad_norm": 4.708478042044171,
+      "learning_rate": 1e-06,
+      "loss": 0.1389,
+      "step": 7183
+    },
+    {
+      "epoch": 1.3737450999139498,
+      "grad_norm": 5.703465114537314,
+      "learning_rate": 1e-06,
+      "loss": 0.21,
+      "step": 7184
+    },
+    {
+      "epoch": 1.373936322784205,
+      "grad_norm": 2.524849039229561,
+      "learning_rate": 1e-06,
+      "loss": 0.0866,
+      "step": 7185
+    },
+    {
+      "epoch": 1.3741275456544604,
+      "grad_norm": 6.453060565469775,
+      "learning_rate": 1e-06,
+      "loss": 0.4313,
+      "step": 7186
+    },
+    {
+      "epoch": 1.3743187685247156,
+      "grad_norm": 2.7962136792077996,
+      "learning_rate": 1e-06,
+      "loss": 0.1392,
+      "step": 7187
+    },
+    {
+      "epoch": 1.3745099913949708,
+      "grad_norm": 2.649329471327423,
+      "learning_rate": 1e-06,
+      "loss": 0.0554,
+      "step": 7188
+    },
+    {
+      "epoch": 1.3747012142652262,
+      "grad_norm": 3.089278872393364,
+      "learning_rate": 1e-06,
+      "loss": 0.116,
+      "step": 7189
+    },
+    {
+      "epoch": 1.3748924371354814,
+      "grad_norm": 2.845095640964956,
+      "learning_rate": 1e-06,
+      "loss": 0.0282,
+      "step": 7190
+    },
+    {
+      "epoch": 1.3750836600057368,
+      "grad_norm": 2.094338519382851,
+      "learning_rate": 1e-06,
+      "loss": 0.0359,
+      "step": 7191
+    },
+    {
+      "epoch": 1.375274882875992,
+      "grad_norm": 4.886303829932007,
+      "learning_rate": 1e-06,
+      "loss": 0.1115,
+      "step": 7192
+    },
+    {
+      "epoch": 1.3754661057462472,
+      "grad_norm": 5.293921443320336,
+      "learning_rate": 1e-06,
+      "loss": 0.2804,
+      "step": 7193
+    },
+    {
+      "epoch": 1.3756573286165026,
+      "grad_norm": 5.902521480406501,
+      "learning_rate": 1e-06,
+      "loss": 0.5185,
+      "step": 7194
+    },
+    {
+      "epoch": 1.3758485514867578,
+      "grad_norm": 4.368339263861535,
+      "learning_rate": 1e-06,
+      "loss": 0.3545,
+      "step": 7195
+    },
+    {
+      "epoch": 1.3760397743570132,
+      "grad_norm": 3.202866408912086,
+      "learning_rate": 1e-06,
+      "loss": 0.1041,
+      "step": 7196
+    },
+    {
+      "epoch": 1.3762309972272684,
+      "grad_norm": 4.133282063279014,
+      "learning_rate": 1e-06,
+      "loss": 0.351,
+      "step": 7197
+    },
+    {
+      "epoch": 1.3764222200975236,
+      "grad_norm": 2.9884569303444755,
+      "learning_rate": 1e-06,
+      "loss": 0.1859,
+      "step": 7198
+    },
+    {
+      "epoch": 1.376613442967779,
+      "grad_norm": 4.498482766276462,
+      "learning_rate": 1e-06,
+      "loss": 0.2145,
+      "step": 7199
+    },
+    {
+      "epoch": 1.3768046658380342,
+      "grad_norm": 2.330097770767976,
+      "learning_rate": 1e-06,
+      "loss": 0.0991,
+      "step": 7200
+    },
+    {
+      "epoch": 1.3769958887082896,
+      "grad_norm": 2.74863382563223,
+      "learning_rate": 1e-06,
+      "loss": 0.0596,
+      "step": 7201
+    },
+    {
+      "epoch": 1.3771871115785448,
+      "grad_norm": 2.401550856830714,
+      "learning_rate": 1e-06,
+      "loss": 0.0915,
+      "step": 7202
+    },
+    {
+      "epoch": 1.3773783344488,
+      "grad_norm": 3.1568016854252208,
+      "learning_rate": 1e-06,
+      "loss": 0.0693,
+      "step": 7203
+    },
+    {
+      "epoch": 1.3775695573190554,
+      "grad_norm": 0.7603644827786837,
+      "learning_rate": 1e-06,
+      "loss": 0.0065,
+      "step": 7204
+    },
+    {
+      "epoch": 1.3777607801893106,
+      "grad_norm": 4.4971420961629684,
+      "learning_rate": 1e-06,
+      "loss": 0.1956,
+      "step": 7205
+    },
+    {
+      "epoch": 1.377952003059566,
+      "grad_norm": 8.360286256552698,
+      "learning_rate": 1e-06,
+      "loss": 0.3248,
+      "step": 7206
+    },
+    {
+      "epoch": 1.3781432259298212,
+      "grad_norm": 2.8603512290884066,
+      "learning_rate": 1e-06,
+      "loss": 0.175,
+      "step": 7207
+    },
+    {
+      "epoch": 1.3783344488000764,
+      "grad_norm": 5.277946025553617,
+      "learning_rate": 1e-06,
+      "loss": 0.1931,
+      "step": 7208
+    },
+    {
+      "epoch": 1.3785256716703318,
+      "grad_norm": 2.779053667355357,
+      "learning_rate": 1e-06,
+      "loss": 0.2088,
+      "step": 7209
+    },
+    {
+      "epoch": 1.378716894540587,
+      "grad_norm": 2.6842241060525613,
+      "learning_rate": 1e-06,
+      "loss": 0.0979,
+      "step": 7210
+    },
+    {
+      "epoch": 1.3789081174108424,
+      "grad_norm": 1.3005711841219079,
+      "learning_rate": 1e-06,
+      "loss": 0.0453,
+      "step": 7211
+    },
+    {
+      "epoch": 1.3790993402810976,
+      "grad_norm": 2.0365024441545874,
+      "learning_rate": 1e-06,
+      "loss": 0.0623,
+      "step": 7212
+    },
+    {
+      "epoch": 1.3792905631513528,
+      "grad_norm": 1.8258912875601432,
+      "learning_rate": 1e-06,
+      "loss": 0.058,
+      "step": 7213
+    },
+    {
+      "epoch": 1.3794817860216082,
+      "grad_norm": 3.9390145673789942,
+      "learning_rate": 1e-06,
+      "loss": 0.0844,
+      "step": 7214
+    },
+    {
+      "epoch": 1.3796730088918634,
+      "grad_norm": 6.135093624330184,
+      "learning_rate": 1e-06,
+      "loss": 0.2218,
+      "step": 7215
+    },
+    {
+      "epoch": 1.3798642317621188,
+      "grad_norm": 3.765908764796097,
+      "learning_rate": 1e-06,
+      "loss": 0.0761,
+      "step": 7216
+    },
+    {
+      "epoch": 1.380055454632374,
+      "grad_norm": 8.22281444696396,
+      "learning_rate": 1e-06,
+      "loss": 0.154,
+      "step": 7217
+    },
+    {
+      "epoch": 1.3802466775026292,
+      "grad_norm": 4.140452716050222,
+      "learning_rate": 1e-06,
+      "loss": 0.1198,
+      "step": 7218
+    },
+    {
+      "epoch": 1.3804379003728846,
+      "grad_norm": 3.751753714890521,
+      "learning_rate": 1e-06,
+      "loss": 0.2139,
+      "step": 7219
+    },
+    {
+      "epoch": 1.3806291232431398,
+      "grad_norm": 2.254145512037969,
+      "learning_rate": 1e-06,
+      "loss": 0.1295,
+      "step": 7220
+    },
+    {
+      "epoch": 1.3808203461133952,
+      "grad_norm": 3.465015222204991,
+      "learning_rate": 1e-06,
+      "loss": 0.2116,
+      "step": 7221
+    },
+    {
+      "epoch": 1.3810115689836504,
+      "grad_norm": 1.6081898593980903,
+      "learning_rate": 1e-06,
+      "loss": 0.1103,
+      "step": 7222
+    },
+    {
+      "epoch": 1.3812027918539056,
+      "grad_norm": 3.480929326024635,
+      "learning_rate": 1e-06,
+      "loss": 0.1688,
+      "step": 7223
+    },
+    {
+      "epoch": 1.381394014724161,
+      "grad_norm": 3.0126842331643924,
+      "learning_rate": 1e-06,
+      "loss": 0.2466,
+      "step": 7224
+    },
+    {
+      "epoch": 1.3815852375944164,
+      "grad_norm": 4.822665686561823,
+      "learning_rate": 1e-06,
+      "loss": 0.2379,
+      "step": 7225
+    },
+    {
+      "epoch": 1.3817764604646716,
+      "grad_norm": 2.2571960452437043,
+      "learning_rate": 1e-06,
+      "loss": 0.0895,
+      "step": 7226
+    },
+    {
+      "epoch": 1.3819676833349268,
+      "grad_norm": 2.599566269423331,
+      "learning_rate": 1e-06,
+      "loss": 0.0528,
+      "step": 7227
+    },
+    {
+      "epoch": 1.382158906205182,
+      "grad_norm": 2.313616818016573,
+      "learning_rate": 1e-06,
+      "loss": 0.1337,
+      "step": 7228
+    },
+    {
+      "epoch": 1.3823501290754374,
+      "grad_norm": 2.7340588196103894,
+      "learning_rate": 1e-06,
+      "loss": 0.0327,
+      "step": 7229
+    },
+    {
+      "epoch": 1.3825413519456928,
+      "grad_norm": 7.820408838614268,
+      "learning_rate": 1e-06,
+      "loss": 0.2625,
+      "step": 7230
+    },
+    {
+      "epoch": 1.382732574815948,
+      "grad_norm": 3.7026947672923143,
+      "learning_rate": 1e-06,
+      "loss": 0.3215,
+      "step": 7231
+    },
+    {
+      "epoch": 1.3829237976862032,
+      "grad_norm": 2.937503084221702,
+      "learning_rate": 1e-06,
+      "loss": 0.1951,
+      "step": 7232
+    },
+    {
+      "epoch": 1.3831150205564586,
+      "grad_norm": 1.9062037227829205,
+      "learning_rate": 1e-06,
+      "loss": 0.0937,
+      "step": 7233
+    },
+    {
+      "epoch": 1.3833062434267138,
+      "grad_norm": 4.68151452981262,
+      "learning_rate": 1e-06,
+      "loss": 0.3651,
+      "step": 7234
+    },
+    {
+      "epoch": 1.3834974662969692,
+      "grad_norm": 3.829688799891724,
+      "learning_rate": 1e-06,
+      "loss": 0.2825,
+      "step": 7235
+    },
+    {
+      "epoch": 1.3836886891672244,
+      "grad_norm": 3.63335521294775,
+      "learning_rate": 1e-06,
+      "loss": 0.2636,
+      "step": 7236
+    },
+    {
+      "epoch": 1.3838799120374796,
+      "grad_norm": 2.416214067455383,
+      "learning_rate": 1e-06,
+      "loss": 0.1743,
+      "step": 7237
+    },
+    {
+      "epoch": 1.384071134907735,
+      "grad_norm": 1.3532549894766983,
+      "learning_rate": 1e-06,
+      "loss": 0.0781,
+      "step": 7238
+    },
+    {
+      "epoch": 1.3842623577779902,
+      "grad_norm": 3.071061758778698,
+      "learning_rate": 1e-06,
+      "loss": 0.107,
+      "step": 7239
+    },
+    {
+      "epoch": 1.3844535806482456,
+      "grad_norm": 3.0331235778163843,
+      "learning_rate": 1e-06,
+      "loss": 0.0941,
+      "step": 7240
+    },
+    {
+      "epoch": 1.3846448035185008,
+      "grad_norm": 4.519769054015453,
+      "learning_rate": 1e-06,
+      "loss": 0.0826,
+      "step": 7241
+    },
+    {
+      "epoch": 1.384836026388756,
+      "grad_norm": 3.6119003378973957,
+      "learning_rate": 1e-06,
+      "loss": 0.0369,
+      "step": 7242
+    },
+    {
+      "epoch": 1.3850272492590114,
+      "grad_norm": 4.026910146168819,
+      "learning_rate": 1e-06,
+      "loss": 0.1903,
+      "step": 7243
+    },
+    {
+      "epoch": 1.3852184721292666,
+      "grad_norm": 4.199977674879049,
+      "learning_rate": 1e-06,
+      "loss": 0.3091,
+      "step": 7244
+    },
+    {
+      "epoch": 1.385409694999522,
+      "grad_norm": 3.3532724708913957,
+      "learning_rate": 1e-06,
+      "loss": 0.0871,
+      "step": 7245
+    },
+    {
+      "epoch": 1.3856009178697772,
+      "grad_norm": 2.5423231579880503,
+      "learning_rate": 1e-06,
+      "loss": 0.2534,
+      "step": 7246
+    },
+    {
+      "epoch": 1.3857921407400324,
+      "grad_norm": 2.4480841288540356,
+      "learning_rate": 1e-06,
+      "loss": 0.0909,
+      "step": 7247
+    },
+    {
+      "epoch": 1.3859833636102878,
+      "grad_norm": 2.2092494383963723,
+      "learning_rate": 1e-06,
+      "loss": 0.0728,
+      "step": 7248
+    },
+    {
+      "epoch": 1.386174586480543,
+      "grad_norm": 7.651487278193657,
+      "learning_rate": 1e-06,
+      "loss": 0.3972,
+      "step": 7249
+    },
+    {
+      "epoch": 1.3863658093507985,
+      "grad_norm": 1.3373685611777009,
+      "learning_rate": 1e-06,
+      "loss": 0.0281,
+      "step": 7250
+    },
+    {
+      "epoch": 1.3865570322210536,
+      "grad_norm": 1.990968817657811,
+      "learning_rate": 1e-06,
+      "loss": 0.0795,
+      "step": 7251
+    },
+    {
+      "epoch": 1.3867482550913088,
+      "grad_norm": 4.122655491253246,
+      "learning_rate": 1e-06,
+      "loss": 0.3008,
+      "step": 7252
+    },
+    {
+      "epoch": 1.3869394779615642,
+      "grad_norm": 3.972655889910566,
+      "learning_rate": 1e-06,
+      "loss": 0.0947,
+      "step": 7253
+    },
+    {
+      "epoch": 1.3871307008318194,
+      "grad_norm": 3.153226102601267,
+      "learning_rate": 1e-06,
+      "loss": 0.021,
+      "step": 7254
+    },
+    {
+      "epoch": 1.3873219237020749,
+      "grad_norm": 4.552139223688223,
+      "learning_rate": 1e-06,
+      "loss": 0.1859,
+      "step": 7255
+    },
+    {
+      "epoch": 1.38751314657233,
+      "grad_norm": 7.302179862856825,
+      "learning_rate": 1e-06,
+      "loss": 0.474,
+      "step": 7256
+    },
+    {
+      "epoch": 1.3877043694425852,
+      "grad_norm": 5.709620792336839,
+      "learning_rate": 1e-06,
+      "loss": 0.4125,
+      "step": 7257
+    },
+    {
+      "epoch": 1.3878955923128407,
+      "grad_norm": 4.236564042243809,
+      "learning_rate": 1e-06,
+      "loss": 0.1338,
+      "step": 7258
+    },
+    {
+      "epoch": 1.3880868151830958,
+      "grad_norm": 1.9474847148055476,
+      "learning_rate": 1e-06,
+      "loss": 0.1958,
+      "step": 7259
+    },
+    {
+      "epoch": 1.3882780380533513,
+      "grad_norm": 2.3527985932821656,
+      "learning_rate": 1e-06,
+      "loss": 0.0989,
+      "step": 7260
+    },
+    {
+      "epoch": 1.3884692609236065,
+      "grad_norm": 2.7160317874615734,
+      "learning_rate": 1e-06,
+      "loss": 0.0714,
+      "step": 7261
+    },
+    {
+      "epoch": 1.3886604837938616,
+      "grad_norm": 5.370667596730439,
+      "learning_rate": 1e-06,
+      "loss": 0.0525,
+      "step": 7262
+    },
+    {
+      "epoch": 1.388851706664117,
+      "grad_norm": 4.5434868107987665,
+      "learning_rate": 1e-06,
+      "loss": 0.1821,
+      "step": 7263
+    },
+    {
+      "epoch": 1.3890429295343723,
+      "grad_norm": 3.5120867068594737,
+      "learning_rate": 1e-06,
+      "loss": 0.2345,
+      "step": 7264
+    },
+    {
+      "epoch": 1.3892341524046277,
+      "grad_norm": 7.945076277830621,
+      "learning_rate": 1e-06,
+      "loss": 0.0741,
+      "step": 7265
+    },
+    {
+      "epoch": 1.3894253752748829,
+      "grad_norm": 0.8527896683188759,
+      "learning_rate": 1e-06,
+      "loss": 0.0155,
+      "step": 7266
+    },
+    {
+      "epoch": 1.389616598145138,
+      "grad_norm": 4.78360784897073,
+      "learning_rate": 1e-06,
+      "loss": 0.0886,
+      "step": 7267
+    },
+    {
+      "epoch": 1.3898078210153935,
+      "grad_norm": 3.914368394787944,
+      "learning_rate": 1e-06,
+      "loss": 0.4296,
+      "step": 7268
+    },
+    {
+      "epoch": 1.3899990438856487,
+      "grad_norm": 4.280019541544288,
+      "learning_rate": 1e-06,
+      "loss": 0.3266,
+      "step": 7269
+    },
+    {
+      "epoch": 1.390190266755904,
+      "grad_norm": 4.580848066280703,
+      "learning_rate": 1e-06,
+      "loss": 0.4763,
+      "step": 7270
+    },
+    {
+      "epoch": 1.3903814896261593,
+      "grad_norm": 6.76162509864377,
+      "learning_rate": 1e-06,
+      "loss": 0.3226,
+      "step": 7271
+    },
+    {
+      "epoch": 1.3905727124964145,
+      "grad_norm": 4.079270008454929,
+      "learning_rate": 1e-06,
+      "loss": 0.3477,
+      "step": 7272
+    },
+    {
+      "epoch": 1.3907639353666699,
+      "grad_norm": 1.8426780817522008,
+      "learning_rate": 1e-06,
+      "loss": 0.1071,
+      "step": 7273
+    },
+    {
+      "epoch": 1.390955158236925,
+      "grad_norm": 2.0591662658583982,
+      "learning_rate": 1e-06,
+      "loss": 0.0894,
+      "step": 7274
+    },
+    {
+      "epoch": 1.3911463811071805,
+      "grad_norm": 3.869851414393377,
+      "learning_rate": 1e-06,
+      "loss": 0.0622,
+      "step": 7275
+    },
+    {
+      "epoch": 1.3913376039774357,
+      "grad_norm": 1.7214851810461225,
+      "learning_rate": 1e-06,
+      "loss": 0.0596,
+      "step": 7276
+    },
+    {
+      "epoch": 1.3915288268476909,
+      "grad_norm": 4.612888871834708,
+      "learning_rate": 1e-06,
+      "loss": 0.0605,
+      "step": 7277
+    },
+    {
+      "epoch": 1.3917200497179463,
+      "grad_norm": 2.4634108935454173,
+      "learning_rate": 1e-06,
+      "loss": 0.0534,
+      "step": 7278
+    },
+    {
+      "epoch": 1.3919112725882017,
+      "grad_norm": 5.720685120746677,
+      "learning_rate": 1e-06,
+      "loss": 0.1808,
+      "step": 7279
+    },
+    {
+      "epoch": 1.3921024954584569,
+      "grad_norm": 3.9810782163439637,
+      "learning_rate": 1e-06,
+      "loss": 0.1794,
+      "step": 7280
+    },
+    {
+      "epoch": 1.392293718328712,
+      "grad_norm": 5.305063046180526,
+      "learning_rate": 1e-06,
+      "loss": 0.5927,
+      "step": 7281
+    },
+    {
+      "epoch": 1.3924849411989673,
+      "grad_norm": 5.564848157759643,
+      "learning_rate": 1e-06,
+      "loss": 0.6657,
+      "step": 7282
+    },
+    {
+      "epoch": 1.3926761640692227,
+      "grad_norm": 4.404435020699307,
+      "learning_rate": 1e-06,
+      "loss": 0.4235,
+      "step": 7283
+    },
+    {
+      "epoch": 1.392867386939478,
+      "grad_norm": 2.9884586855001705,
+      "learning_rate": 1e-06,
+      "loss": 0.3499,
+      "step": 7284
+    },
+    {
+      "epoch": 1.3930586098097333,
+      "grad_norm": 3.741525738843868,
+      "learning_rate": 1e-06,
+      "loss": 0.0717,
+      "step": 7285
+    },
+    {
+      "epoch": 1.3932498326799885,
+      "grad_norm": 3.927869129572401,
+      "learning_rate": 1e-06,
+      "loss": 0.3284,
+      "step": 7286
+    },
+    {
+      "epoch": 1.393441055550244,
+      "grad_norm": 1.5766439758963806,
+      "learning_rate": 1e-06,
+      "loss": 0.1318,
+      "step": 7287
+    },
+    {
+      "epoch": 1.393632278420499,
+      "grad_norm": 2.028488512186718,
+      "learning_rate": 1e-06,
+      "loss": 0.1513,
+      "step": 7288
+    },
+    {
+      "epoch": 1.3938235012907545,
+      "grad_norm": 5.3794528348266075,
+      "learning_rate": 1e-06,
+      "loss": 0.0777,
+      "step": 7289
+    },
+    {
+      "epoch": 1.3940147241610097,
+      "grad_norm": 2.485134080765773,
+      "learning_rate": 1e-06,
+      "loss": 0.0795,
+      "step": 7290
+    },
+    {
+      "epoch": 1.3942059470312649,
+      "grad_norm": 1.633479661939272,
+      "learning_rate": 1e-06,
+      "loss": 0.0199,
+      "step": 7291
+    },
+    {
+      "epoch": 1.3943971699015203,
+      "grad_norm": 5.799490202499416,
+      "learning_rate": 1e-06,
+      "loss": 0.065,
+      "step": 7292
+    },
+    {
+      "epoch": 1.3945883927717755,
+      "grad_norm": 5.866479742077781,
+      "learning_rate": 1e-06,
+      "loss": 0.2069,
+      "step": 7293
+    },
+    {
+      "epoch": 1.394779615642031,
+      "grad_norm": 4.944979447031846,
+      "learning_rate": 1e-06,
+      "loss": 0.5539,
+      "step": 7294
+    },
+    {
+      "epoch": 1.394970838512286,
+      "grad_norm": 2.6934200111296716,
+      "learning_rate": 1e-06,
+      "loss": 0.1179,
+      "step": 7295
+    },
+    {
+      "epoch": 1.3951620613825413,
+      "grad_norm": 1.4933828788582053,
+      "learning_rate": 1e-06,
+      "loss": 0.0846,
+      "step": 7296
+    },
+    {
+      "epoch": 1.3953532842527967,
+      "grad_norm": 4.952804218097333,
+      "learning_rate": 1e-06,
+      "loss": 0.4644,
+      "step": 7297
+    },
+    {
+      "epoch": 1.395544507123052,
+      "grad_norm": 4.750823100200062,
+      "learning_rate": 1e-06,
+      "loss": 0.2627,
+      "step": 7298
+    },
+    {
+      "epoch": 1.3957357299933073,
+      "grad_norm": 3.6007796820754887,
+      "learning_rate": 1e-06,
+      "loss": 0.1966,
+      "step": 7299
+    },
+    {
+      "epoch": 1.3959269528635625,
+      "grad_norm": 3.9467240329602715,
+      "learning_rate": 1e-06,
+      "loss": 0.1072,
+      "step": 7300
+    },
+    {
+      "epoch": 1.3961181757338177,
+      "grad_norm": 1.227174399752595,
+      "learning_rate": 1e-06,
+      "loss": 0.0046,
+      "step": 7301
+    },
+    {
+      "epoch": 1.396309398604073,
+      "grad_norm": 3.154421286004323,
+      "learning_rate": 1e-06,
+      "loss": 0.1873,
+      "step": 7302
+    },
+    {
+      "epoch": 1.3965006214743283,
+      "grad_norm": 3.150851022717259,
+      "learning_rate": 1e-06,
+      "loss": 0.0508,
+      "step": 7303
+    },
+    {
+      "epoch": 1.3966918443445837,
+      "grad_norm": 2.8499177552787427,
+      "learning_rate": 1e-06,
+      "loss": 0.1265,
+      "step": 7304
+    },
+    {
+      "epoch": 1.396883067214839,
+      "grad_norm": 10.42587638116958,
+      "learning_rate": 1e-06,
+      "loss": 0.1308,
+      "step": 7305
+    },
+    {
+      "epoch": 1.397074290085094,
+      "grad_norm": 3.4933393315513697,
+      "learning_rate": 1e-06,
+      "loss": 0.2334,
+      "step": 7306
+    },
+    {
+      "epoch": 1.3972655129553495,
+      "grad_norm": 4.504213056346026,
+      "learning_rate": 1e-06,
+      "loss": 0.3883,
+      "step": 7307
+    },
+    {
+      "epoch": 1.3974567358256047,
+      "grad_norm": 3.3298177936181865,
+      "learning_rate": 1e-06,
+      "loss": 0.2694,
+      "step": 7308
+    },
+    {
+      "epoch": 1.3976479586958601,
+      "grad_norm": 3.872287539225101,
+      "learning_rate": 1e-06,
+      "loss": 0.2173,
+      "step": 7309
+    },
+    {
+      "epoch": 1.3978391815661153,
+      "grad_norm": 2.2747364855728645,
+      "learning_rate": 1e-06,
+      "loss": 0.0479,
+      "step": 7310
+    },
+    {
+      "epoch": 1.3980304044363705,
+      "grad_norm": 5.655301683711406,
+      "learning_rate": 1e-06,
+      "loss": 0.1816,
+      "step": 7311
+    },
+    {
+      "epoch": 1.398221627306626,
+      "grad_norm": 2.5641149688689624,
+      "learning_rate": 1e-06,
+      "loss": 0.0886,
+      "step": 7312
+    },
+    {
+      "epoch": 1.398412850176881,
+      "grad_norm": 5.150422658520499,
+      "learning_rate": 1e-06,
+      "loss": 0.1343,
+      "step": 7313
+    },
+    {
+      "epoch": 1.3986040730471365,
+      "grad_norm": 2.3329050034248895,
+      "learning_rate": 1e-06,
+      "loss": 0.0408,
+      "step": 7314
+    },
+    {
+      "epoch": 1.3987952959173917,
+      "grad_norm": 6.585537247809412,
+      "learning_rate": 1e-06,
+      "loss": 0.144,
+      "step": 7315
+    },
+    {
+      "epoch": 1.398986518787647,
+      "grad_norm": 2.922089757523781,
+      "learning_rate": 1e-06,
+      "loss": 0.1857,
+      "step": 7316
+    },
+    {
+      "epoch": 1.3991777416579023,
+      "grad_norm": 5.2528789437675005,
+      "learning_rate": 1e-06,
+      "loss": 0.0776,
+      "step": 7317
+    },
+    {
+      "epoch": 1.3993689645281575,
+      "grad_norm": 3.7385983387430306,
+      "learning_rate": 1e-06,
+      "loss": 0.1128,
+      "step": 7318
+    },
+    {
+      "epoch": 1.399560187398413,
+      "grad_norm": 3.3276914215629363,
+      "learning_rate": 1e-06,
+      "loss": 0.1632,
+      "step": 7319
+    },
+    {
+      "epoch": 1.3997514102686681,
+      "grad_norm": 3.1353273262822383,
+      "learning_rate": 1e-06,
+      "loss": 0.1321,
+      "step": 7320
+    },
+    {
+      "epoch": 1.3999426331389233,
+      "grad_norm": 4.446065453449536,
+      "learning_rate": 1e-06,
+      "loss": 0.4092,
+      "step": 7321
+    },
+    {
+      "epoch": 1.4001338560091787,
+      "grad_norm": 4.0409599289298574,
+      "learning_rate": 1e-06,
+      "loss": 0.2369,
+      "step": 7322
+    },
+    {
+      "epoch": 1.400325078879434,
+      "grad_norm": 2.2379352508532766,
+      "learning_rate": 1e-06,
+      "loss": 0.1015,
+      "step": 7323
+    },
+    {
+      "epoch": 1.4005163017496893,
+      "grad_norm": 2.1098681968273274,
+      "learning_rate": 1e-06,
+      "loss": 0.0493,
+      "step": 7324
+    },
+    {
+      "epoch": 1.4007075246199445,
+      "grad_norm": 3.057541550535651,
+      "learning_rate": 1e-06,
+      "loss": 0.0439,
+      "step": 7325
+    },
+    {
+      "epoch": 1.4008987474901997,
+      "grad_norm": 2.9629023733390185,
+      "learning_rate": 1e-06,
+      "loss": 0.0457,
+      "step": 7326
+    },
+    {
+      "epoch": 1.4010899703604551,
+      "grad_norm": 2.9214482123234013,
+      "learning_rate": 1e-06,
+      "loss": 0.1119,
+      "step": 7327
+    },
+    {
+      "epoch": 1.4012811932307103,
+      "grad_norm": 2.581424613015907,
+      "learning_rate": 1e-06,
+      "loss": 0.0508,
+      "step": 7328
+    },
+    {
+      "epoch": 1.4014724161009657,
+      "grad_norm": 4.615987295758038,
+      "learning_rate": 1e-06,
+      "loss": 0.0626,
+      "step": 7329
+    },
+    {
+      "epoch": 1.401663638971221,
+      "grad_norm": 4.448302753258096,
+      "learning_rate": 1e-06,
+      "loss": 0.1782,
+      "step": 7330
+    },
+    {
+      "epoch": 1.4018548618414761,
+      "grad_norm": 6.12881319888334,
+      "learning_rate": 1e-06,
+      "loss": 0.4366,
+      "step": 7331
+    },
+    {
+      "epoch": 1.4020460847117315,
+      "grad_norm": 4.793106437763559,
+      "learning_rate": 1e-06,
+      "loss": 0.3021,
+      "step": 7332
+    },
+    {
+      "epoch": 1.4022373075819867,
+      "grad_norm": 4.5029247104383145,
+      "learning_rate": 1e-06,
+      "loss": 0.2094,
+      "step": 7333
+    },
+    {
+      "epoch": 1.4024285304522421,
+      "grad_norm": 2.6802288431706565,
+      "learning_rate": 1e-06,
+      "loss": 0.1487,
+      "step": 7334
+    },
+    {
+      "epoch": 1.4026197533224973,
+      "grad_norm": 2.6700630513436634,
+      "learning_rate": 1e-06,
+      "loss": 0.0674,
+      "step": 7335
+    },
+    {
+      "epoch": 1.4028109761927525,
+      "grad_norm": 3.6554863287088684,
+      "learning_rate": 1e-06,
+      "loss": 0.177,
+      "step": 7336
+    },
+    {
+      "epoch": 1.403002199063008,
+      "grad_norm": 4.741753899297591,
+      "learning_rate": 1e-06,
+      "loss": 0.2628,
+      "step": 7337
+    },
+    {
+      "epoch": 1.4031934219332634,
+      "grad_norm": 2.4360089754650214,
+      "learning_rate": 1e-06,
+      "loss": 0.0527,
+      "step": 7338
+    },
+    {
+      "epoch": 1.4033846448035185,
+      "grad_norm": 5.186965547169598,
+      "learning_rate": 1e-06,
+      "loss": 0.0794,
+      "step": 7339
+    },
+    {
+      "epoch": 1.4035758676737737,
+      "grad_norm": 3.1108230975970743,
+      "learning_rate": 1e-06,
+      "loss": 0.0522,
+      "step": 7340
+    },
+    {
+      "epoch": 1.403767090544029,
+      "grad_norm": 3.839747880765324,
+      "learning_rate": 1e-06,
+      "loss": 0.0985,
+      "step": 7341
+    },
+    {
+      "epoch": 1.4039583134142843,
+      "grad_norm": 6.037435729512016,
+      "learning_rate": 1e-06,
+      "loss": 0.0719,
+      "step": 7342
+    },
+    {
+      "epoch": 1.4041495362845398,
+      "grad_norm": 3.462022873818891,
+      "learning_rate": 1e-06,
+      "loss": 0.1487,
+      "step": 7343
+    },
+    {
+      "epoch": 1.404340759154795,
+      "grad_norm": 10.64461044982004,
+      "learning_rate": 1e-06,
+      "loss": 0.3798,
+      "step": 7344
+    },
+    {
+      "epoch": 1.4045319820250501,
+      "grad_norm": 2.263056278865702,
+      "learning_rate": 1e-06,
+      "loss": 0.129,
+      "step": 7345
+    },
+    {
+      "epoch": 1.4047232048953056,
+      "grad_norm": 2.725160866984928,
+      "learning_rate": 1e-06,
+      "loss": 0.2541,
+      "step": 7346
+    },
+    {
+      "epoch": 1.4049144277655607,
+      "grad_norm": 5.194100282978159,
+      "learning_rate": 1e-06,
+      "loss": 0.4013,
+      "step": 7347
+    },
+    {
+      "epoch": 1.4051056506358162,
+      "grad_norm": 3.477900078450857,
+      "learning_rate": 1e-06,
+      "loss": 0.0979,
+      "step": 7348
+    },
+    {
+      "epoch": 1.4052968735060714,
+      "grad_norm": 2.4583431071286497,
+      "learning_rate": 1e-06,
+      "loss": 0.0405,
+      "step": 7349
+    },
+    {
+      "epoch": 1.4054880963763265,
+      "grad_norm": 1.953561840795988,
+      "learning_rate": 1e-06,
+      "loss": 0.0203,
+      "step": 7350
+    },
+    {
+      "epoch": 1.405679319246582,
+      "grad_norm": 8.635425845249822,
+      "learning_rate": 1e-06,
+      "loss": 0.0868,
+      "step": 7351
+    },
+    {
+      "epoch": 1.4058705421168372,
+      "grad_norm": 3.1266996720557283,
+      "learning_rate": 1e-06,
+      "loss": 0.134,
+      "step": 7352
+    },
+    {
+      "epoch": 1.4060617649870926,
+      "grad_norm": 1.8719915414235981,
+      "learning_rate": 1e-06,
+      "loss": 0.0794,
+      "step": 7353
+    },
+    {
+      "epoch": 1.4062529878573478,
+      "grad_norm": 2.820226356268596,
+      "learning_rate": 1e-06,
+      "loss": 0.0563,
+      "step": 7354
+    },
+    {
+      "epoch": 1.406444210727603,
+      "grad_norm": 4.370526888092672,
+      "learning_rate": 1e-06,
+      "loss": 0.147,
+      "step": 7355
+    },
+    {
+      "epoch": 1.4066354335978584,
+      "grad_norm": 3.972860655500631,
+      "learning_rate": 1e-06,
+      "loss": 0.2418,
+      "step": 7356
+    },
+    {
+      "epoch": 1.4068266564681136,
+      "grad_norm": 5.045630712034578,
+      "learning_rate": 1e-06,
+      "loss": 0.3633,
+      "step": 7357
+    },
+    {
+      "epoch": 1.407017879338369,
+      "grad_norm": 3.566180686878791,
+      "learning_rate": 1e-06,
+      "loss": 0.3524,
+      "step": 7358
+    },
+    {
+      "epoch": 1.4072091022086242,
+      "grad_norm": 3.729526069129613,
+      "learning_rate": 1e-06,
+      "loss": 0.2611,
+      "step": 7359
+    },
+    {
+      "epoch": 1.4074003250788794,
+      "grad_norm": 3.578266657394798,
+      "learning_rate": 1e-06,
+      "loss": 0.2727,
+      "step": 7360
+    },
+    {
+      "epoch": 1.4075915479491348,
+      "grad_norm": 1.8946341889721015,
+      "learning_rate": 1e-06,
+      "loss": 0.1029,
+      "step": 7361
+    },
+    {
+      "epoch": 1.40778277081939,
+      "grad_norm": 1.6387791013134172,
+      "learning_rate": 1e-06,
+      "loss": 0.0778,
+      "step": 7362
+    },
+    {
+      "epoch": 1.4079739936896454,
+      "grad_norm": 3.669245391100506,
+      "learning_rate": 1e-06,
+      "loss": 0.0698,
+      "step": 7363
+    },
+    {
+      "epoch": 1.4081652165599006,
+      "grad_norm": 4.75128919272653,
+      "learning_rate": 1e-06,
+      "loss": 0.1631,
+      "step": 7364
+    },
+    {
+      "epoch": 1.4083564394301558,
+      "grad_norm": 3.4808548050658756,
+      "learning_rate": 1e-06,
+      "loss": 0.3209,
+      "step": 7365
+    },
+    {
+      "epoch": 1.4085476623004112,
+      "grad_norm": 4.123900064569183,
+      "learning_rate": 1e-06,
+      "loss": 0.1672,
+      "step": 7366
+    },
+    {
+      "epoch": 1.4087388851706664,
+      "grad_norm": 3.484252927549142,
+      "learning_rate": 1e-06,
+      "loss": 0.0992,
+      "step": 7367
+    },
+    {
+      "epoch": 1.4089301080409218,
+      "grad_norm": 8.108056815732635,
+      "learning_rate": 1e-06,
+      "loss": 0.2948,
+      "step": 7368
+    },
+    {
+      "epoch": 1.409121330911177,
+      "grad_norm": 3.525533995526533,
+      "learning_rate": 1e-06,
+      "loss": 0.2491,
+      "step": 7369
+    },
+    {
+      "epoch": 1.4093125537814322,
+      "grad_norm": 4.032434570387116,
+      "learning_rate": 1e-06,
+      "loss": 0.3284,
+      "step": 7370
+    },
+    {
+      "epoch": 1.4095037766516876,
+      "grad_norm": 3.9770108253153262,
+      "learning_rate": 1e-06,
+      "loss": 0.1268,
+      "step": 7371
+    },
+    {
+      "epoch": 1.4096949995219428,
+      "grad_norm": 3.873741683926416,
+      "learning_rate": 1e-06,
+      "loss": 0.4015,
+      "step": 7372
+    },
+    {
+      "epoch": 1.4098862223921982,
+      "grad_norm": 3.242212649328288,
+      "learning_rate": 1e-06,
+      "loss": 0.1706,
+      "step": 7373
+    },
+    {
+      "epoch": 1.4100774452624534,
+      "grad_norm": 3.5507983048871905,
+      "learning_rate": 1e-06,
+      "loss": 0.1569,
+      "step": 7374
+    },
+    {
+      "epoch": 1.4102686681327086,
+      "grad_norm": 4.815402852256144,
+      "learning_rate": 1e-06,
+      "loss": 0.0978,
+      "step": 7375
+    },
+    {
+      "epoch": 1.410459891002964,
+      "grad_norm": 3.4413533201271553,
+      "learning_rate": 1e-06,
+      "loss": 0.0426,
+      "step": 7376
+    },
+    {
+      "epoch": 1.4106511138732192,
+      "grad_norm": 4.62243457473495,
+      "learning_rate": 1e-06,
+      "loss": 0.0662,
+      "step": 7377
+    },
+    {
+      "epoch": 1.4108423367434746,
+      "grad_norm": 2.6271040522106004,
+      "learning_rate": 1e-06,
+      "loss": 0.0283,
+      "step": 7378
+    },
+    {
+      "epoch": 1.4110335596137298,
+      "grad_norm": 6.64013813702394,
+      "learning_rate": 1e-06,
+      "loss": 0.1418,
+      "step": 7379
+    },
+    {
+      "epoch": 1.411224782483985,
+      "grad_norm": 2.3753057584201422,
+      "learning_rate": 1e-06,
+      "loss": 0.0428,
+      "step": 7380
+    },
+    {
+      "epoch": 1.4114160053542404,
+      "grad_norm": 6.549938568526032,
+      "learning_rate": 1e-06,
+      "loss": 0.5308,
+      "step": 7381
+    },
+    {
+      "epoch": 1.4116072282244956,
+      "grad_norm": 4.804815324773241,
+      "learning_rate": 1e-06,
+      "loss": 0.17,
+      "step": 7382
+    },
+    {
+      "epoch": 1.411798451094751,
+      "grad_norm": 2.1084062400344856,
+      "learning_rate": 1e-06,
+      "loss": 0.0569,
+      "step": 7383
+    },
+    {
+      "epoch": 1.4119896739650062,
+      "grad_norm": 2.399605631216048,
+      "learning_rate": 1e-06,
+      "loss": 0.1164,
+      "step": 7384
+    },
+    {
+      "epoch": 1.4121808968352614,
+      "grad_norm": 4.879352485035056,
+      "learning_rate": 1e-06,
+      "loss": 0.334,
+      "step": 7385
+    },
+    {
+      "epoch": 1.4123721197055168,
+      "grad_norm": 3.2533984389070323,
+      "learning_rate": 1e-06,
+      "loss": 0.2106,
+      "step": 7386
+    },
+    {
+      "epoch": 1.412563342575772,
+      "grad_norm": 2.4253786400996367,
+      "learning_rate": 1e-06,
+      "loss": 0.0716,
+      "step": 7387
+    },
+    {
+      "epoch": 1.4127545654460274,
+      "grad_norm": 2.095793680330365,
+      "learning_rate": 1e-06,
+      "loss": 0.1088,
+      "step": 7388
+    },
+    {
+      "epoch": 1.4129457883162826,
+      "grad_norm": 3.3357309460625033,
+      "learning_rate": 1e-06,
+      "loss": 0.1096,
+      "step": 7389
+    },
+    {
+      "epoch": 1.4131370111865378,
+      "grad_norm": 6.404346909113793,
+      "learning_rate": 1e-06,
+      "loss": 0.1875,
+      "step": 7390
+    },
+    {
+      "epoch": 1.4133282340567932,
+      "grad_norm": 4.308030839602454,
+      "learning_rate": 1e-06,
+      "loss": 0.1138,
+      "step": 7391
+    },
+    {
+      "epoch": 1.4135194569270486,
+      "grad_norm": 5.425711645746848,
+      "learning_rate": 1e-06,
+      "loss": 0.0342,
+      "step": 7392
+    },
+    {
+      "epoch": 1.4137106797973038,
+      "grad_norm": 4.828204404307609,
+      "learning_rate": 1e-06,
+      "loss": 0.2981,
+      "step": 7393
+    },
+    {
+      "epoch": 1.413901902667559,
+      "grad_norm": 3.8191500340287594,
+      "learning_rate": 1e-06,
+      "loss": 0.496,
+      "step": 7394
+    },
+    {
+      "epoch": 1.4140931255378142,
+      "grad_norm": 4.221251021891417,
+      "learning_rate": 1e-06,
+      "loss": 0.1857,
+      "step": 7395
+    },
+    {
+      "epoch": 1.4142843484080696,
+      "grad_norm": 1.7057103124454955,
+      "learning_rate": 1e-06,
+      "loss": 0.1225,
+      "step": 7396
+    },
+    {
+      "epoch": 1.414475571278325,
+      "grad_norm": 2.3694066891252925,
+      "learning_rate": 1e-06,
+      "loss": 0.1692,
+      "step": 7397
+    },
+    {
+      "epoch": 1.4146667941485802,
+      "grad_norm": 5.445072557370498,
+      "learning_rate": 1e-06,
+      "loss": 0.3675,
+      "step": 7398
+    },
+    {
+      "epoch": 1.4148580170188354,
+      "grad_norm": 4.895304917841938,
+      "learning_rate": 1e-06,
+      "loss": 0.3139,
+      "step": 7399
+    },
+    {
+      "epoch": 1.4150492398890908,
+      "grad_norm": 2.1277110813379227,
+      "learning_rate": 1e-06,
+      "loss": 0.1137,
+      "step": 7400
+    },
+    {
+      "epoch": 1.415240462759346,
+      "grad_norm": 8.133360801191797,
+      "learning_rate": 1e-06,
+      "loss": 0.1275,
+      "step": 7401
+    },
+    {
+      "epoch": 1.4154316856296014,
+      "grad_norm": 4.0300814095502515,
+      "learning_rate": 1e-06,
+      "loss": 0.0612,
+      "step": 7402
+    },
+    {
+      "epoch": 1.4156229084998566,
+      "grad_norm": 2.684047077652956,
+      "learning_rate": 1e-06,
+      "loss": 0.0525,
+      "step": 7403
+    },
+    {
+      "epoch": 1.4158141313701118,
+      "grad_norm": 7.865112152173023,
+      "learning_rate": 1e-06,
+      "loss": 0.1493,
+      "step": 7404
+    },
+    {
+      "epoch": 1.4160053542403672,
+      "grad_norm": 2.1750154911234993,
+      "learning_rate": 1e-06,
+      "loss": 0.1392,
+      "step": 7405
+    },
+    {
+      "epoch": 1.4161965771106224,
+      "grad_norm": 6.418932163881444,
+      "learning_rate": 1e-06,
+      "loss": 0.2192,
+      "step": 7406
+    },
+    {
+      "epoch": 1.4163877999808778,
+      "grad_norm": 5.743710602634538,
+      "learning_rate": 1e-06,
+      "loss": 0.446,
+      "step": 7407
+    },
+    {
+      "epoch": 1.416579022851133,
+      "grad_norm": 5.369655524771763,
+      "learning_rate": 1e-06,
+      "loss": 0.5145,
+      "step": 7408
+    },
+    {
+      "epoch": 1.4167702457213882,
+      "grad_norm": 2.1188766666685077,
+      "learning_rate": 1e-06,
+      "loss": 0.0547,
+      "step": 7409
+    },
+    {
+      "epoch": 1.4169614685916436,
+      "grad_norm": 1.4558016827787015,
+      "learning_rate": 1e-06,
+      "loss": 0.0749,
+      "step": 7410
+    },
+    {
+      "epoch": 1.4171526914618988,
+      "grad_norm": 2.6501970325852073,
+      "learning_rate": 1e-06,
+      "loss": 0.0986,
+      "step": 7411
+    },
+    {
+      "epoch": 1.4173439143321542,
+      "grad_norm": 2.3981830076708595,
+      "learning_rate": 1e-06,
+      "loss": 0.1167,
+      "step": 7412
+    },
+    {
+      "epoch": 1.4175351372024094,
+      "grad_norm": 2.780850756846589,
+      "learning_rate": 1e-06,
+      "loss": 0.0555,
+      "step": 7413
+    },
+    {
+      "epoch": 1.4177263600726646,
+      "grad_norm": 3.1724397504179236,
+      "learning_rate": 1e-06,
+      "loss": 0.2508,
+      "step": 7414
+    },
+    {
+      "epoch": 1.41791758294292,
+      "grad_norm": 1.5513043422233201,
+      "learning_rate": 1e-06,
+      "loss": 0.0251,
+      "step": 7415
+    },
+    {
+      "epoch": 1.4181088058131752,
+      "grad_norm": 3.1774066364489153,
+      "learning_rate": 1e-06,
+      "loss": 0.0474,
+      "step": 7416
+    },
+    {
+      "epoch": 1.4183000286834306,
+      "grad_norm": 6.820829007056683,
+      "learning_rate": 1e-06,
+      "loss": 0.0659,
+      "step": 7417
+    },
+    {
+      "epoch": 1.4184912515536858,
+      "grad_norm": 3.137956083059708,
+      "learning_rate": 1e-06,
+      "loss": 0.1441,
+      "step": 7418
+    },
+    {
+      "epoch": 1.418682474423941,
+      "grad_norm": 5.538473466509829,
+      "learning_rate": 1e-06,
+      "loss": 0.3216,
+      "step": 7419
+    },
+    {
+      "epoch": 1.4188736972941964,
+      "grad_norm": 1.7762466565932205,
+      "learning_rate": 1e-06,
+      "loss": 0.1223,
+      "step": 7420
+    },
+    {
+      "epoch": 1.4190649201644516,
+      "grad_norm": 3.0563793131613113,
+      "learning_rate": 1e-06,
+      "loss": 0.1892,
+      "step": 7421
+    },
+    {
+      "epoch": 1.419256143034707,
+      "grad_norm": 5.652181774631506,
+      "learning_rate": 1e-06,
+      "loss": 0.3369,
+      "step": 7422
+    },
+    {
+      "epoch": 1.4194473659049622,
+      "grad_norm": 5.207590380766573,
+      "learning_rate": 1e-06,
+      "loss": 0.2712,
+      "step": 7423
+    },
+    {
+      "epoch": 1.4196385887752174,
+      "grad_norm": 3.8128373043747024,
+      "learning_rate": 1e-06,
+      "loss": 0.2238,
+      "step": 7424
+    },
+    {
+      "epoch": 1.4198298116454728,
+      "grad_norm": 2.495912643811299,
+      "learning_rate": 1e-06,
+      "loss": 0.0276,
+      "step": 7425
+    },
+    {
+      "epoch": 1.420021034515728,
+      "grad_norm": 6.984009325665761,
+      "learning_rate": 1e-06,
+      "loss": 0.0411,
+      "step": 7426
+    },
+    {
+      "epoch": 1.4202122573859834,
+      "grad_norm": 3.7037492056983203,
+      "learning_rate": 1e-06,
+      "loss": 0.1588,
+      "step": 7427
+    },
+    {
+      "epoch": 1.4204034802562386,
+      "grad_norm": 4.79196681312698,
+      "learning_rate": 1e-06,
+      "loss": 0.0824,
+      "step": 7428
+    },
+    {
+      "epoch": 1.4205947031264938,
+      "grad_norm": 4.778017035612424,
+      "learning_rate": 1e-06,
+      "loss": 0.0869,
+      "step": 7429
+    },
+    {
+      "epoch": 1.4207859259967492,
+      "grad_norm": 3.298545522651225,
+      "learning_rate": 1e-06,
+      "loss": 0.1804,
+      "step": 7430
+    },
+    {
+      "epoch": 1.4209771488670044,
+      "grad_norm": 3.6209591672839525,
+      "learning_rate": 1e-06,
+      "loss": 0.2072,
+      "step": 7431
+    },
+    {
+      "epoch": 1.4211683717372599,
+      "grad_norm": 4.951516260212268,
+      "learning_rate": 1e-06,
+      "loss": 0.2858,
+      "step": 7432
+    },
+    {
+      "epoch": 1.421359594607515,
+      "grad_norm": 2.316581423538211,
+      "learning_rate": 1e-06,
+      "loss": 0.1959,
+      "step": 7433
+    },
+    {
+      "epoch": 1.4215508174777702,
+      "grad_norm": 4.987467508998962,
+      "learning_rate": 1e-06,
+      "loss": 0.2839,
+      "step": 7434
+    },
+    {
+      "epoch": 1.4217420403480256,
+      "grad_norm": 2.337725615195444,
+      "learning_rate": 1e-06,
+      "loss": 0.1018,
+      "step": 7435
+    },
+    {
+      "epoch": 1.4219332632182808,
+      "grad_norm": 2.1258223569260877,
+      "learning_rate": 1e-06,
+      "loss": 0.0584,
+      "step": 7436
+    },
+    {
+      "epoch": 1.4221244860885363,
+      "grad_norm": 1.9838514463353119,
+      "learning_rate": 1e-06,
+      "loss": 0.0721,
+      "step": 7437
+    },
+    {
+      "epoch": 1.4223157089587914,
+      "grad_norm": 2.784708337492877,
+      "learning_rate": 1e-06,
+      "loss": 0.0839,
+      "step": 7438
+    },
+    {
+      "epoch": 1.4225069318290466,
+      "grad_norm": 1.466120699931898,
+      "learning_rate": 1e-06,
+      "loss": 0.0159,
+      "step": 7439
+    },
+    {
+      "epoch": 1.422698154699302,
+      "grad_norm": 2.82235474759261,
+      "learning_rate": 1e-06,
+      "loss": 0.1393,
+      "step": 7440
+    },
+    {
+      "epoch": 1.4228893775695572,
+      "grad_norm": 2.0335922821543755,
+      "learning_rate": 1e-06,
+      "loss": 0.0598,
+      "step": 7441
+    },
+    {
+      "epoch": 1.4230806004398127,
+      "grad_norm": 3.2865757908615385,
+      "learning_rate": 1e-06,
+      "loss": 0.0623,
+      "step": 7442
+    },
+    {
+      "epoch": 1.4232718233100679,
+      "grad_norm": 5.162050071488242,
+      "learning_rate": 1e-06,
+      "loss": 0.4538,
+      "step": 7443
+    },
+    {
+      "epoch": 1.423463046180323,
+      "grad_norm": 5.35356265370624,
+      "learning_rate": 1e-06,
+      "loss": 0.3143,
+      "step": 7444
+    },
+    {
+      "epoch": 1.4236542690505785,
+      "grad_norm": 3.008910139867484,
+      "learning_rate": 1e-06,
+      "loss": 0.2609,
+      "step": 7445
+    },
+    {
+      "epoch": 1.4238454919208337,
+      "grad_norm": 2.4180642942616024,
+      "learning_rate": 1e-06,
+      "loss": 0.1808,
+      "step": 7446
+    },
+    {
+      "epoch": 1.424036714791089,
+      "grad_norm": 3.373809321851186,
+      "learning_rate": 1e-06,
+      "loss": 0.1921,
+      "step": 7447
+    },
+    {
+      "epoch": 1.4242279376613443,
+      "grad_norm": 3.6849932151121205,
+      "learning_rate": 1e-06,
+      "loss": 0.2964,
+      "step": 7448
+    },
+    {
+      "epoch": 1.4244191605315994,
+      "grad_norm": 1.6769142855472332,
+      "learning_rate": 1e-06,
+      "loss": 0.0547,
+      "step": 7449
+    },
+    {
+      "epoch": 1.4246103834018549,
+      "grad_norm": 1.3134898131882733,
+      "learning_rate": 1e-06,
+      "loss": 0.0133,
+      "step": 7450
+    },
+    {
+      "epoch": 1.4248016062721103,
+      "grad_norm": 2.980775269911638,
+      "learning_rate": 1e-06,
+      "loss": 0.1053,
+      "step": 7451
+    },
+    {
+      "epoch": 1.4249928291423655,
+      "grad_norm": 3.972613519159323,
+      "learning_rate": 1e-06,
+      "loss": 0.2872,
+      "step": 7452
+    },
+    {
+      "epoch": 1.4251840520126207,
+      "grad_norm": 0.9813519941156832,
+      "learning_rate": 1e-06,
+      "loss": 0.0099,
+      "step": 7453
+    },
+    {
+      "epoch": 1.4253752748828759,
+      "grad_norm": 3.291338715130931,
+      "learning_rate": 1e-06,
+      "loss": 0.0664,
+      "step": 7454
+    },
+    {
+      "epoch": 1.4255664977531313,
+      "grad_norm": 4.2955191397752825,
+      "learning_rate": 1e-06,
+      "loss": 0.1407,
+      "step": 7455
+    },
+    {
+      "epoch": 1.4257577206233867,
+      "grad_norm": 4.664652162526513,
+      "learning_rate": 1e-06,
+      "loss": 0.5313,
+      "step": 7456
+    },
+    {
+      "epoch": 1.4259489434936419,
+      "grad_norm": 4.416134832258799,
+      "learning_rate": 1e-06,
+      "loss": 0.2273,
+      "step": 7457
+    },
+    {
+      "epoch": 1.426140166363897,
+      "grad_norm": 4.243121471125424,
+      "learning_rate": 1e-06,
+      "loss": 0.2708,
+      "step": 7458
+    },
+    {
+      "epoch": 1.4263313892341525,
+      "grad_norm": 2.441822425465727,
+      "learning_rate": 1e-06,
+      "loss": 0.1824,
+      "step": 7459
+    },
+    {
+      "epoch": 1.4265226121044077,
+      "grad_norm": 4.894073927719661,
+      "learning_rate": 1e-06,
+      "loss": 0.3725,
+      "step": 7460
+    },
+    {
+      "epoch": 1.426713834974663,
+      "grad_norm": 8.024736783948752,
+      "learning_rate": 1e-06,
+      "loss": 0.1023,
+      "step": 7461
+    },
+    {
+      "epoch": 1.4269050578449183,
+      "grad_norm": 2.7674245206193864,
+      "learning_rate": 1e-06,
+      "loss": 0.1098,
+      "step": 7462
+    },
+    {
+      "epoch": 1.4270962807151735,
+      "grad_norm": 2.2852178059951753,
+      "learning_rate": 1e-06,
+      "loss": 0.1127,
+      "step": 7463
+    },
+    {
+      "epoch": 1.4272875035854289,
+      "grad_norm": 2.8391066989176994,
+      "learning_rate": 1e-06,
+      "loss": 0.0574,
+      "step": 7464
+    },
+    {
+      "epoch": 1.427478726455684,
+      "grad_norm": 3.6198346398453367,
+      "learning_rate": 1e-06,
+      "loss": 0.1177,
+      "step": 7465
+    },
+    {
+      "epoch": 1.4276699493259395,
+      "grad_norm": 3.4983013663270914,
+      "learning_rate": 1e-06,
+      "loss": 0.0958,
+      "step": 7466
+    },
+    {
+      "epoch": 1.4278611721961947,
+      "grad_norm": 6.130337316495209,
+      "learning_rate": 1e-06,
+      "loss": 0.2986,
+      "step": 7467
+    },
+    {
+      "epoch": 1.4280523950664499,
+      "grad_norm": 7.189297923905496,
+      "learning_rate": 1e-06,
+      "loss": 0.2366,
+      "step": 7468
+    },
+    {
+      "epoch": 1.4282436179367053,
+      "grad_norm": 4.610228792417207,
+      "learning_rate": 1e-06,
+      "loss": 0.4017,
+      "step": 7469
+    },
+    {
+      "epoch": 1.4284348408069605,
+      "grad_norm": 4.785366903845404,
+      "learning_rate": 1e-06,
+      "loss": 0.2554,
+      "step": 7470
+    },
+    {
+      "epoch": 1.428626063677216,
+      "grad_norm": 3.5036936751814984,
+      "learning_rate": 1e-06,
+      "loss": 0.0852,
+      "step": 7471
+    },
+    {
+      "epoch": 1.428817286547471,
+      "grad_norm": 1.8203769394106502,
+      "learning_rate": 1e-06,
+      "loss": 0.1579,
+      "step": 7472
+    },
+    {
+      "epoch": 1.4290085094177263,
+      "grad_norm": 2.591512243284273,
+      "learning_rate": 1e-06,
+      "loss": 0.1426,
+      "step": 7473
+    },
+    {
+      "epoch": 1.4291997322879817,
+      "grad_norm": 3.644742937062984,
+      "learning_rate": 1e-06,
+      "loss": 0.0564,
+      "step": 7474
+    },
+    {
+      "epoch": 1.4293909551582369,
+      "grad_norm": 2.082909604532635,
+      "learning_rate": 1e-06,
+      "loss": 0.0692,
+      "step": 7475
+    },
+    {
+      "epoch": 1.4295821780284923,
+      "grad_norm": 2.6932620002566345,
+      "learning_rate": 1e-06,
+      "loss": 0.106,
+      "step": 7476
+    },
+    {
+      "epoch": 1.4297734008987475,
+      "grad_norm": 2.4213254243508677,
+      "learning_rate": 1e-06,
+      "loss": 0.0448,
+      "step": 7477
+    },
+    {
+      "epoch": 1.4299646237690027,
+      "grad_norm": 2.7669652073047573,
+      "learning_rate": 1e-06,
+      "loss": 0.0421,
+      "step": 7478
+    },
+    {
+      "epoch": 1.430155846639258,
+      "grad_norm": 2.275781111712309,
+      "learning_rate": 1e-06,
+      "loss": 0.0384,
+      "step": 7479
+    },
+    {
+      "epoch": 1.4303470695095133,
+      "grad_norm": 7.9497166133251405,
+      "learning_rate": 1e-06,
+      "loss": 0.2885,
+      "step": 7480
+    },
+    {
+      "epoch": 1.4305382923797687,
+      "grad_norm": 3.786078908485894,
+      "learning_rate": 1e-06,
+      "loss": 0.4948,
+      "step": 7481
+    },
+    {
+      "epoch": 1.430729515250024,
+      "grad_norm": 5.344771973444716,
+      "learning_rate": 1e-06,
+      "loss": 0.3707,
+      "step": 7482
+    },
+    {
+      "epoch": 1.430920738120279,
+      "grad_norm": 5.4021975814011185,
+      "learning_rate": 1e-06,
+      "loss": 0.4423,
+      "step": 7483
+    },
+    {
+      "epoch": 1.4311119609905345,
+      "grad_norm": 2.4052912796232273,
+      "learning_rate": 1e-06,
+      "loss": 0.0963,
+      "step": 7484
+    },
+    {
+      "epoch": 1.4313031838607897,
+      "grad_norm": 3.617047762385902,
+      "learning_rate": 1e-06,
+      "loss": 0.1077,
+      "step": 7485
+    },
+    {
+      "epoch": 1.431494406731045,
+      "grad_norm": 3.025746176250695,
+      "learning_rate": 1e-06,
+      "loss": 0.0588,
+      "step": 7486
+    },
+    {
+      "epoch": 1.4316856296013003,
+      "grad_norm": 4.486465232790084,
+      "learning_rate": 1e-06,
+      "loss": 0.2561,
+      "step": 7487
+    },
+    {
+      "epoch": 1.4318768524715555,
+      "grad_norm": 1.9079142950270536,
+      "learning_rate": 1e-06,
+      "loss": 0.023,
+      "step": 7488
+    },
+    {
+      "epoch": 1.432068075341811,
+      "grad_norm": 2.5612639260460344,
+      "learning_rate": 1e-06,
+      "loss": 0.0761,
+      "step": 7489
+    },
+    {
+      "epoch": 1.432259298212066,
+      "grad_norm": 4.052920270585959,
+      "learning_rate": 1e-06,
+      "loss": 0.1314,
+      "step": 7490
+    },
+    {
+      "epoch": 1.4324505210823215,
+      "grad_norm": 1.9745194804239357,
+      "learning_rate": 1e-06,
+      "loss": 0.0241,
+      "step": 7491
+    },
+    {
+      "epoch": 1.4326417439525767,
+      "grad_norm": 6.897449530665387,
+      "learning_rate": 1e-06,
+      "loss": 0.1785,
+      "step": 7492
+    },
+    {
+      "epoch": 1.432832966822832,
+      "grad_norm": 11.067274194916685,
+      "learning_rate": 1e-06,
+      "loss": 0.4598,
+      "step": 7493
+    },
+    {
+      "epoch": 1.4330241896930873,
+      "grad_norm": 4.265966380784934,
+      "learning_rate": 1e-06,
+      "loss": 0.4219,
+      "step": 7494
+    },
+    {
+      "epoch": 1.4332154125633425,
+      "grad_norm": 4.4904858359884505,
+      "learning_rate": 1e-06,
+      "loss": 0.1767,
+      "step": 7495
+    },
+    {
+      "epoch": 1.433406635433598,
+      "grad_norm": 4.869094597468474,
+      "learning_rate": 1e-06,
+      "loss": 0.4522,
+      "step": 7496
+    },
+    {
+      "epoch": 1.4335978583038531,
+      "grad_norm": 4.262581993947677,
+      "learning_rate": 1e-06,
+      "loss": 0.1301,
+      "step": 7497
+    },
+    {
+      "epoch": 1.4337890811741083,
+      "grad_norm": 5.8962566461903725,
+      "learning_rate": 1e-06,
+      "loss": 0.2271,
+      "step": 7498
+    },
+    {
+      "epoch": 1.4339803040443637,
+      "grad_norm": 2.186249948003211,
+      "learning_rate": 1e-06,
+      "loss": 0.0619,
+      "step": 7499
+    },
+    {
+      "epoch": 1.434171526914619,
+      "grad_norm": 1.6541611076408906,
+      "learning_rate": 1e-06,
+      "loss": 0.0398,
+      "step": 7500
+    },
+    {
+      "epoch": 1.434171526914619,
+      "eval_runtime": 759.7984,
+      "eval_samples_per_second": 2.019,
+      "eval_steps_per_second": 0.505,
+      "step": 7500
+    },
+    {
+      "epoch": 1.4343627497848743,
+      "grad_norm": 2.1469470593659765,
+      "learning_rate": 1e-06,
+      "loss": 0.0386,
+      "step": 7501
+    },
+    {
+      "epoch": 1.4345539726551295,
+      "grad_norm": 2.32249127471566,
+      "learning_rate": 1e-06,
+      "loss": 0.108,
+      "step": 7502
+    },
+    {
+      "epoch": 1.4347451955253847,
+      "grad_norm": 3.2375042653884183,
+      "learning_rate": 1e-06,
+      "loss": 0.0574,
+      "step": 7503
+    },
+    {
+      "epoch": 1.4349364183956401,
+      "grad_norm": 1.9438115959588538,
+      "learning_rate": 1e-06,
+      "loss": 0.0346,
+      "step": 7504
+    },
+    {
+      "epoch": 1.4351276412658955,
+      "grad_norm": 7.0809374478511575,
+      "learning_rate": 1e-06,
+      "loss": 0.1253,
+      "step": 7505
+    },
+    {
+      "epoch": 1.4353188641361507,
+      "grad_norm": 4.756578407529162,
+      "learning_rate": 1e-06,
+      "loss": 0.3003,
+      "step": 7506
+    },
+    {
+      "epoch": 1.435510087006406,
+      "grad_norm": 5.959654263702811,
+      "learning_rate": 1e-06,
+      "loss": 0.5841,
+      "step": 7507
+    },
+    {
+      "epoch": 1.4357013098766611,
+      "grad_norm": 4.580742514111749,
+      "learning_rate": 1e-06,
+      "loss": 0.4353,
+      "step": 7508
+    },
+    {
+      "epoch": 1.4358925327469165,
+      "grad_norm": 2.4821606250217174,
+      "learning_rate": 1e-06,
+      "loss": 0.1474,
+      "step": 7509
+    },
+    {
+      "epoch": 1.436083755617172,
+      "grad_norm": 2.350960790763953,
+      "learning_rate": 1e-06,
+      "loss": 0.0917,
+      "step": 7510
+    },
+    {
+      "epoch": 1.4362749784874271,
+      "grad_norm": 4.028819216264705,
+      "learning_rate": 1e-06,
+      "loss": 0.1999,
+      "step": 7511
+    },
+    {
+      "epoch": 1.4364662013576823,
+      "grad_norm": 3.2910196819337014,
+      "learning_rate": 1e-06,
+      "loss": 0.1794,
+      "step": 7512
+    },
+    {
+      "epoch": 1.4366574242279377,
+      "grad_norm": 2.513767101318826,
+      "learning_rate": 1e-06,
+      "loss": 0.068,
+      "step": 7513
+    },
+    {
+      "epoch": 1.436848647098193,
+      "grad_norm": 2.0644330300730847,
+      "learning_rate": 1e-06,
+      "loss": 0.0969,
+      "step": 7514
+    },
+    {
+      "epoch": 1.4370398699684483,
+      "grad_norm": 3.4442073234510815,
+      "learning_rate": 1e-06,
+      "loss": 0.0799,
+      "step": 7515
+    },
+    {
+      "epoch": 1.4372310928387035,
+      "grad_norm": 3.019251880824261,
+      "learning_rate": 1e-06,
+      "loss": 0.0848,
+      "step": 7516
+    },
+    {
+      "epoch": 1.4374223157089587,
+      "grad_norm": 2.8095743008911125,
+      "learning_rate": 1e-06,
+      "loss": 0.0209,
+      "step": 7517
+    },
+    {
+      "epoch": 1.4376135385792141,
+      "grad_norm": 3.8622688841576935,
+      "learning_rate": 1e-06,
+      "loss": 0.3468,
+      "step": 7518
+    },
+    {
+      "epoch": 1.4378047614494693,
+      "grad_norm": 2.813510204724451,
+      "learning_rate": 1e-06,
+      "loss": 0.0623,
+      "step": 7519
+    },
+    {
+      "epoch": 1.4379959843197248,
+      "grad_norm": 2.4203409566370007,
+      "learning_rate": 1e-06,
+      "loss": 0.1098,
+      "step": 7520
+    },
+    {
+      "epoch": 1.43818720718998,
+      "grad_norm": 3.647622256305445,
+      "learning_rate": 1e-06,
+      "loss": 0.2174,
+      "step": 7521
+    },
+    {
+      "epoch": 1.4383784300602351,
+      "grad_norm": 2.476030455036856,
+      "learning_rate": 1e-06,
+      "loss": 0.125,
+      "step": 7522
+    },
+    {
+      "epoch": 1.4385696529304905,
+      "grad_norm": 3.3042750901612234,
+      "learning_rate": 1e-06,
+      "loss": 0.2626,
+      "step": 7523
+    },
+    {
+      "epoch": 1.4387608758007457,
+      "grad_norm": 2.5334624993707275,
+      "learning_rate": 1e-06,
+      "loss": 0.1729,
+      "step": 7524
+    },
+    {
+      "epoch": 1.4389520986710012,
+      "grad_norm": 3.4928287881630284,
+      "learning_rate": 1e-06,
+      "loss": 0.2161,
+      "step": 7525
+    },
+    {
+      "epoch": 1.4391433215412563,
+      "grad_norm": 3.1712960151699052,
+      "learning_rate": 1e-06,
+      "loss": 0.2464,
+      "step": 7526
+    },
+    {
+      "epoch": 1.4393345444115115,
+      "grad_norm": 1.025962452162393,
+      "learning_rate": 1e-06,
+      "loss": 0.0236,
+      "step": 7527
+    },
+    {
+      "epoch": 1.439525767281767,
+      "grad_norm": 1.6921373861822864,
+      "learning_rate": 1e-06,
+      "loss": 0.0489,
+      "step": 7528
+    },
+    {
+      "epoch": 1.4397169901520221,
+      "grad_norm": 4.203992562647664,
+      "learning_rate": 1e-06,
+      "loss": 0.058,
+      "step": 7529
+    },
+    {
+      "epoch": 1.4399082130222776,
+      "grad_norm": 8.854688585643792,
+      "learning_rate": 1e-06,
+      "loss": 0.2389,
+      "step": 7530
+    },
+    {
+      "epoch": 1.4400994358925328,
+      "grad_norm": 4.131836976150602,
+      "learning_rate": 1e-06,
+      "loss": 0.4251,
+      "step": 7531
+    },
+    {
+      "epoch": 1.440290658762788,
+      "grad_norm": 5.489393063263272,
+      "learning_rate": 1e-06,
+      "loss": 0.3683,
+      "step": 7532
+    },
+    {
+      "epoch": 1.4404818816330434,
+      "grad_norm": 4.0962689053957995,
+      "learning_rate": 1e-06,
+      "loss": 0.0867,
+      "step": 7533
+    },
+    {
+      "epoch": 1.4406731045032986,
+      "grad_norm": 3.3434239879220664,
+      "learning_rate": 1e-06,
+      "loss": 0.2271,
+      "step": 7534
+    },
+    {
+      "epoch": 1.440864327373554,
+      "grad_norm": 3.8788334742669415,
+      "learning_rate": 1e-06,
+      "loss": 0.1776,
+      "step": 7535
+    },
+    {
+      "epoch": 1.4410555502438092,
+      "grad_norm": 3.948805186485891,
+      "learning_rate": 1e-06,
+      "loss": 0.2009,
+      "step": 7536
+    },
+    {
+      "epoch": 1.4412467731140644,
+      "grad_norm": 3.2944633739228517,
+      "learning_rate": 1e-06,
+      "loss": 0.1118,
+      "step": 7537
+    },
+    {
+      "epoch": 1.4414379959843198,
+      "grad_norm": 4.081732663272281,
+      "learning_rate": 1e-06,
+      "loss": 0.2937,
+      "step": 7538
+    },
+    {
+      "epoch": 1.441629218854575,
+      "grad_norm": 4.356809893712505,
+      "learning_rate": 1e-06,
+      "loss": 0.0889,
+      "step": 7539
+    },
+    {
+      "epoch": 1.4418204417248304,
+      "grad_norm": 1.1897975379150176,
+      "learning_rate": 1e-06,
+      "loss": 0.0196,
+      "step": 7540
+    },
+    {
+      "epoch": 1.4420116645950856,
+      "grad_norm": 3.7908070746184825,
+      "learning_rate": 1e-06,
+      "loss": 0.1165,
+      "step": 7541
+    },
+    {
+      "epoch": 1.4422028874653408,
+      "grad_norm": 3.400751838583191,
+      "learning_rate": 1e-06,
+      "loss": 0.0714,
+      "step": 7542
+    },
+    {
+      "epoch": 1.4423941103355962,
+      "grad_norm": 8.144698926631282,
+      "learning_rate": 1e-06,
+      "loss": 0.1613,
+      "step": 7543
+    },
+    {
+      "epoch": 1.4425853332058514,
+      "grad_norm": 6.595463390020565,
+      "learning_rate": 1e-06,
+      "loss": 0.4869,
+      "step": 7544
+    },
+    {
+      "epoch": 1.4427765560761068,
+      "grad_norm": 3.493303978143875,
+      "learning_rate": 1e-06,
+      "loss": 0.307,
+      "step": 7545
+    },
+    {
+      "epoch": 1.442967778946362,
+      "grad_norm": 3.613494899722404,
+      "learning_rate": 1e-06,
+      "loss": 0.1374,
+      "step": 7546
+    },
+    {
+      "epoch": 1.4431590018166172,
+      "grad_norm": 3.1085404852871763,
+      "learning_rate": 1e-06,
+      "loss": 0.2206,
+      "step": 7547
+    },
+    {
+      "epoch": 1.4433502246868726,
+      "grad_norm": 3.8030832029755652,
+      "learning_rate": 1e-06,
+      "loss": 0.2992,
+      "step": 7548
+    },
+    {
+      "epoch": 1.4435414475571278,
+      "grad_norm": 1.7751482136056511,
+      "learning_rate": 1e-06,
+      "loss": 0.0508,
+      "step": 7549
+    },
+    {
+      "epoch": 1.4437326704273832,
+      "grad_norm": 2.785043165738742,
+      "learning_rate": 1e-06,
+      "loss": 0.0889,
+      "step": 7550
+    },
+    {
+      "epoch": 1.4439238932976384,
+      "grad_norm": 2.736963844921855,
+      "learning_rate": 1e-06,
+      "loss": 0.0702,
+      "step": 7551
+    },
+    {
+      "epoch": 1.4441151161678936,
+      "grad_norm": 3.0183946609656105,
+      "learning_rate": 1e-06,
+      "loss": 0.0444,
+      "step": 7552
+    },
+    {
+      "epoch": 1.444306339038149,
+      "grad_norm": 2.7087691812227805,
+      "learning_rate": 1e-06,
+      "loss": 0.0511,
+      "step": 7553
+    },
+    {
+      "epoch": 1.4444975619084042,
+      "grad_norm": 2.6310233084190986,
+      "learning_rate": 1e-06,
+      "loss": 0.0693,
+      "step": 7554
+    },
+    {
+      "epoch": 1.4446887847786596,
+      "grad_norm": 7.069360961954085,
+      "learning_rate": 1e-06,
+      "loss": 0.0752,
+      "step": 7555
+    },
+    {
+      "epoch": 1.4448800076489148,
+      "grad_norm": 6.51275849808624,
+      "learning_rate": 1e-06,
+      "loss": 0.3377,
+      "step": 7556
+    },
+    {
+      "epoch": 1.44507123051917,
+      "grad_norm": 2.714234383893633,
+      "learning_rate": 1e-06,
+      "loss": 0.1111,
+      "step": 7557
+    },
+    {
+      "epoch": 1.4452624533894254,
+      "grad_norm": 1.5338850120806464,
+      "learning_rate": 1e-06,
+      "loss": 0.1601,
+      "step": 7558
+    },
+    {
+      "epoch": 1.4454536762596806,
+      "grad_norm": 5.1941355354448815,
+      "learning_rate": 1e-06,
+      "loss": 0.2016,
+      "step": 7559
+    },
+    {
+      "epoch": 1.445644899129936,
+      "grad_norm": 3.5971447377019805,
+      "learning_rate": 1e-06,
+      "loss": 0.3229,
+      "step": 7560
+    },
+    {
+      "epoch": 1.4458361220001912,
+      "grad_norm": 5.304137885451861,
+      "learning_rate": 1e-06,
+      "loss": 0.3207,
+      "step": 7561
+    },
+    {
+      "epoch": 1.4460273448704464,
+      "grad_norm": 4.381123861763584,
+      "learning_rate": 1e-06,
+      "loss": 0.1958,
+      "step": 7562
+    },
+    {
+      "epoch": 1.4462185677407018,
+      "grad_norm": 2.4077700483081403,
+      "learning_rate": 1e-06,
+      "loss": 0.0802,
+      "step": 7563
+    },
+    {
+      "epoch": 1.4464097906109572,
+      "grad_norm": 3.1269788961402885,
+      "learning_rate": 1e-06,
+      "loss": 0.07,
+      "step": 7564
+    },
+    {
+      "epoch": 1.4466010134812124,
+      "grad_norm": 4.434584613601417,
+      "learning_rate": 1e-06,
+      "loss": 0.0287,
+      "step": 7565
+    },
+    {
+      "epoch": 1.4467922363514676,
+      "grad_norm": 1.4117789191811467,
+      "learning_rate": 1e-06,
+      "loss": 0.0265,
+      "step": 7566
+    },
+    {
+      "epoch": 1.446983459221723,
+      "grad_norm": 7.298579341659179,
+      "learning_rate": 1e-06,
+      "loss": 0.1531,
+      "step": 7567
+    },
+    {
+      "epoch": 1.4471746820919782,
+      "grad_norm": 2.9983153381429646,
+      "learning_rate": 1e-06,
+      "loss": 0.2311,
+      "step": 7568
+    },
+    {
+      "epoch": 1.4473659049622336,
+      "grad_norm": 4.950783160899899,
+      "learning_rate": 1e-06,
+      "loss": 0.3411,
+      "step": 7569
+    },
+    {
+      "epoch": 1.4475571278324888,
+      "grad_norm": 1.9058344653879176,
+      "learning_rate": 1e-06,
+      "loss": 0.128,
+      "step": 7570
+    },
+    {
+      "epoch": 1.447748350702744,
+      "grad_norm": 3.0919862594003087,
+      "learning_rate": 1e-06,
+      "loss": 0.1186,
+      "step": 7571
+    },
+    {
+      "epoch": 1.4479395735729994,
+      "grad_norm": 4.010302864871076,
+      "learning_rate": 1e-06,
+      "loss": 0.1883,
+      "step": 7572
+    },
+    {
+      "epoch": 1.4481307964432546,
+      "grad_norm": 2.9611202777420864,
+      "learning_rate": 1e-06,
+      "loss": 0.2212,
+      "step": 7573
+    },
+    {
+      "epoch": 1.44832201931351,
+      "grad_norm": 4.111548715504393,
+      "learning_rate": 1e-06,
+      "loss": 0.1164,
+      "step": 7574
+    },
+    {
+      "epoch": 1.4485132421837652,
+      "grad_norm": 3.5667235117390153,
+      "learning_rate": 1e-06,
+      "loss": 0.0677,
+      "step": 7575
+    },
+    {
+      "epoch": 1.4487044650540204,
+      "grad_norm": 3.0836760528073026,
+      "learning_rate": 1e-06,
+      "loss": 0.1728,
+      "step": 7576
+    },
+    {
+      "epoch": 1.4488956879242758,
+      "grad_norm": 2.102661526130451,
+      "learning_rate": 1e-06,
+      "loss": 0.0454,
+      "step": 7577
+    },
+    {
+      "epoch": 1.449086910794531,
+      "grad_norm": 4.026121914332374,
+      "learning_rate": 1e-06,
+      "loss": 0.0737,
+      "step": 7578
+    },
+    {
+      "epoch": 1.4492781336647864,
+      "grad_norm": 3.186425270955723,
+      "learning_rate": 1e-06,
+      "loss": 0.1202,
+      "step": 7579
+    },
+    {
+      "epoch": 1.4494693565350416,
+      "grad_norm": 3.751508028079003,
+      "learning_rate": 1e-06,
+      "loss": 0.1333,
+      "step": 7580
+    },
+    {
+      "epoch": 1.4496605794052968,
+      "grad_norm": 8.951731202181692,
+      "learning_rate": 1e-06,
+      "loss": 0.3778,
+      "step": 7581
+    },
+    {
+      "epoch": 1.4498518022755522,
+      "grad_norm": 5.469538343011131,
+      "learning_rate": 1e-06,
+      "loss": 0.2409,
+      "step": 7582
+    },
+    {
+      "epoch": 1.4500430251458074,
+      "grad_norm": 4.651075316759862,
+      "learning_rate": 1e-06,
+      "loss": 0.0881,
+      "step": 7583
+    },
+    {
+      "epoch": 1.4502342480160628,
+      "grad_norm": 2.839998390573395,
+      "learning_rate": 1e-06,
+      "loss": 0.0857,
+      "step": 7584
+    },
+    {
+      "epoch": 1.450425470886318,
+      "grad_norm": 4.187009868120946,
+      "learning_rate": 1e-06,
+      "loss": 0.2879,
+      "step": 7585
+    },
+    {
+      "epoch": 1.4506166937565732,
+      "grad_norm": 1.435274972921317,
+      "learning_rate": 1e-06,
+      "loss": 0.09,
+      "step": 7586
+    },
+    {
+      "epoch": 1.4508079166268286,
+      "grad_norm": 4.57653718565624,
+      "learning_rate": 1e-06,
+      "loss": 0.0939,
+      "step": 7587
+    },
+    {
+      "epoch": 1.4509991394970838,
+      "grad_norm": 3.4114756964710016,
+      "learning_rate": 1e-06,
+      "loss": 0.1938,
+      "step": 7588
+    },
+    {
+      "epoch": 1.4511903623673392,
+      "grad_norm": 4.807993574826839,
+      "learning_rate": 1e-06,
+      "loss": 0.0713,
+      "step": 7589
+    },
+    {
+      "epoch": 1.4513815852375944,
+      "grad_norm": 3.0516524981828335,
+      "learning_rate": 1e-06,
+      "loss": 0.0532,
+      "step": 7590
+    },
+    {
+      "epoch": 1.4515728081078496,
+      "grad_norm": 2.1850492645575983,
+      "learning_rate": 1e-06,
+      "loss": 0.0319,
+      "step": 7591
+    },
+    {
+      "epoch": 1.451764030978105,
+      "grad_norm": 5.5908594601522426,
+      "learning_rate": 1e-06,
+      "loss": 0.0201,
+      "step": 7592
+    },
+    {
+      "epoch": 1.4519552538483602,
+      "grad_norm": 4.636293592620892,
+      "learning_rate": 1e-06,
+      "loss": 0.1592,
+      "step": 7593
+    },
+    {
+      "epoch": 1.4521464767186156,
+      "grad_norm": 4.441856072963734,
+      "learning_rate": 1e-06,
+      "loss": 0.3128,
+      "step": 7594
+    },
+    {
+      "epoch": 1.4523376995888708,
+      "grad_norm": 3.60826800656522,
+      "learning_rate": 1e-06,
+      "loss": 0.1125,
+      "step": 7595
+    },
+    {
+      "epoch": 1.452528922459126,
+      "grad_norm": 3.6350996139708758,
+      "learning_rate": 1e-06,
+      "loss": 0.2052,
+      "step": 7596
+    },
+    {
+      "epoch": 1.4527201453293814,
+      "grad_norm": 2.8805440439182033,
+      "learning_rate": 1e-06,
+      "loss": 0.2695,
+      "step": 7597
+    },
+    {
+      "epoch": 1.4529113681996366,
+      "grad_norm": 4.8645163074698035,
+      "learning_rate": 1e-06,
+      "loss": 0.2113,
+      "step": 7598
+    },
+    {
+      "epoch": 1.453102591069892,
+      "grad_norm": 3.4879095513839307,
+      "learning_rate": 1e-06,
+      "loss": 0.0814,
+      "step": 7599
+    },
+    {
+      "epoch": 1.4532938139401472,
+      "grad_norm": 5.9624597448113725,
+      "learning_rate": 1e-06,
+      "loss": 0.0822,
+      "step": 7600
+    },
+    {
+      "epoch": 1.4534850368104024,
+      "grad_norm": 2.709727080281218,
+      "learning_rate": 1e-06,
+      "loss": 0.2416,
+      "step": 7601
+    },
+    {
+      "epoch": 1.4536762596806578,
+      "grad_norm": 1.7887724262273894,
+      "learning_rate": 1e-06,
+      "loss": 0.0607,
+      "step": 7602
+    },
+    {
+      "epoch": 1.453867482550913,
+      "grad_norm": 4.539923003631917,
+      "learning_rate": 1e-06,
+      "loss": 0.1057,
+      "step": 7603
+    },
+    {
+      "epoch": 1.4540587054211684,
+      "grad_norm": 12.670258330712606,
+      "learning_rate": 1e-06,
+      "loss": 0.0728,
+      "step": 7604
+    },
+    {
+      "epoch": 1.4542499282914236,
+      "grad_norm": 4.95847048957108,
+      "learning_rate": 1e-06,
+      "loss": 0.2851,
+      "step": 7605
+    },
+    {
+      "epoch": 1.4544411511616788,
+      "grad_norm": 3.527190584255914,
+      "learning_rate": 1e-06,
+      "loss": 0.1697,
+      "step": 7606
+    },
+    {
+      "epoch": 1.4546323740319342,
+      "grad_norm": 5.498563752346884,
+      "learning_rate": 1e-06,
+      "loss": 0.4821,
+      "step": 7607
+    },
+    {
+      "epoch": 1.4548235969021894,
+      "grad_norm": 4.993449115879321,
+      "learning_rate": 1e-06,
+      "loss": 0.1293,
+      "step": 7608
+    },
+    {
+      "epoch": 1.4550148197724448,
+      "grad_norm": 3.224860751672199,
+      "learning_rate": 1e-06,
+      "loss": 0.3029,
+      "step": 7609
+    },
+    {
+      "epoch": 1.4552060426427,
+      "grad_norm": 2.2352831935468007,
+      "learning_rate": 1e-06,
+      "loss": 0.1121,
+      "step": 7610
+    },
+    {
+      "epoch": 1.4553972655129552,
+      "grad_norm": 2.031243544348215,
+      "learning_rate": 1e-06,
+      "loss": 0.118,
+      "step": 7611
+    },
+    {
+      "epoch": 1.4555884883832106,
+      "grad_norm": 4.2005534761046945,
+      "learning_rate": 1e-06,
+      "loss": 0.1112,
+      "step": 7612
+    },
+    {
+      "epoch": 1.4557797112534658,
+      "grad_norm": 1.9348781442839078,
+      "learning_rate": 1e-06,
+      "loss": 0.0981,
+      "step": 7613
+    },
+    {
+      "epoch": 1.4559709341237212,
+      "grad_norm": 3.45564695757332,
+      "learning_rate": 1e-06,
+      "loss": 0.0401,
+      "step": 7614
+    },
+    {
+      "epoch": 1.4561621569939764,
+      "grad_norm": 4.785981075276549,
+      "learning_rate": 1e-06,
+      "loss": 0.1068,
+      "step": 7615
+    },
+    {
+      "epoch": 1.4563533798642316,
+      "grad_norm": 4.252532653145,
+      "learning_rate": 1e-06,
+      "loss": 0.1436,
+      "step": 7616
+    },
+    {
+      "epoch": 1.456544602734487,
+      "grad_norm": 6.154022278466127,
+      "learning_rate": 1e-06,
+      "loss": 0.0811,
+      "step": 7617
+    },
+    {
+      "epoch": 1.4567358256047425,
+      "grad_norm": 3.890209252385243,
+      "learning_rate": 1e-06,
+      "loss": 0.1888,
+      "step": 7618
+    },
+    {
+      "epoch": 1.4569270484749977,
+      "grad_norm": 4.037669666610918,
+      "learning_rate": 1e-06,
+      "loss": 0.3895,
+      "step": 7619
+    },
+    {
+      "epoch": 1.4571182713452528,
+      "grad_norm": 3.6587862526596666,
+      "learning_rate": 1e-06,
+      "loss": 0.2386,
+      "step": 7620
+    },
+    {
+      "epoch": 1.457309494215508,
+      "grad_norm": 1.836796141778403,
+      "learning_rate": 1e-06,
+      "loss": 0.1284,
+      "step": 7621
+    },
+    {
+      "epoch": 1.4575007170857635,
+      "grad_norm": 5.5319773739940254,
+      "learning_rate": 1e-06,
+      "loss": 0.4562,
+      "step": 7622
+    },
+    {
+      "epoch": 1.4576919399560189,
+      "grad_norm": 3.0285790070595735,
+      "learning_rate": 1e-06,
+      "loss": 0.0832,
+      "step": 7623
+    },
+    {
+      "epoch": 1.457883162826274,
+      "grad_norm": 3.726540491450996,
+      "learning_rate": 1e-06,
+      "loss": 0.1857,
+      "step": 7624
+    },
+    {
+      "epoch": 1.4580743856965293,
+      "grad_norm": 4.227094070373305,
+      "learning_rate": 1e-06,
+      "loss": 0.2038,
+      "step": 7625
+    },
+    {
+      "epoch": 1.4582656085667847,
+      "grad_norm": 3.14873312042667,
+      "learning_rate": 1e-06,
+      "loss": 0.1802,
+      "step": 7626
+    },
+    {
+      "epoch": 1.4584568314370399,
+      "grad_norm": 3.477543450309644,
+      "learning_rate": 1e-06,
+      "loss": 0.1886,
+      "step": 7627
+    },
+    {
+      "epoch": 1.4586480543072953,
+      "grad_norm": 2.5272061105596917,
+      "learning_rate": 1e-06,
+      "loss": 0.0689,
+      "step": 7628
+    },
+    {
+      "epoch": 1.4588392771775505,
+      "grad_norm": 2.807385520333435,
+      "learning_rate": 1e-06,
+      "loss": 0.0313,
+      "step": 7629
+    },
+    {
+      "epoch": 1.4590305000478057,
+      "grad_norm": 5.154608077830472,
+      "learning_rate": 1e-06,
+      "loss": 0.1036,
+      "step": 7630
+    },
+    {
+      "epoch": 1.459221722918061,
+      "grad_norm": 4.436152723787757,
+      "learning_rate": 1e-06,
+      "loss": 0.1469,
+      "step": 7631
+    },
+    {
+      "epoch": 1.4594129457883163,
+      "grad_norm": 4.663624907497604,
+      "learning_rate": 1e-06,
+      "loss": 0.2868,
+      "step": 7632
+    },
+    {
+      "epoch": 1.4596041686585717,
+      "grad_norm": 2.2522834211375042,
+      "learning_rate": 1e-06,
+      "loss": 0.067,
+      "step": 7633
+    },
+    {
+      "epoch": 1.4597953915288269,
+      "grad_norm": 3.336152045254409,
+      "learning_rate": 1e-06,
+      "loss": 0.2664,
+      "step": 7634
+    },
+    {
+      "epoch": 1.459986614399082,
+      "grad_norm": 4.01884859477313,
+      "learning_rate": 1e-06,
+      "loss": 0.1303,
+      "step": 7635
+    },
+    {
+      "epoch": 1.4601778372693375,
+      "grad_norm": 2.2034506049907345,
+      "learning_rate": 1e-06,
+      "loss": 0.0681,
+      "step": 7636
+    },
+    {
+      "epoch": 1.4603690601395927,
+      "grad_norm": 2.691485621108315,
+      "learning_rate": 1e-06,
+      "loss": 0.0972,
+      "step": 7637
+    },
+    {
+      "epoch": 1.460560283009848,
+      "grad_norm": 5.220960502973414,
+      "learning_rate": 1e-06,
+      "loss": 0.1346,
+      "step": 7638
+    },
+    {
+      "epoch": 1.4607515058801033,
+      "grad_norm": 1.6486636182991061,
+      "learning_rate": 1e-06,
+      "loss": 0.0391,
+      "step": 7639
+    },
+    {
+      "epoch": 1.4609427287503585,
+      "grad_norm": 2.189052575774368,
+      "learning_rate": 1e-06,
+      "loss": 0.0325,
+      "step": 7640
+    },
+    {
+      "epoch": 1.4611339516206139,
+      "grad_norm": 3.2067128998033385,
+      "learning_rate": 1e-06,
+      "loss": 0.0692,
+      "step": 7641
+    },
+    {
+      "epoch": 1.461325174490869,
+      "grad_norm": 2.4143465239217923,
+      "learning_rate": 1e-06,
+      "loss": 0.0361,
+      "step": 7642
+    },
+    {
+      "epoch": 1.4615163973611245,
+      "grad_norm": 4.7358561298323565,
+      "learning_rate": 1e-06,
+      "loss": 0.1421,
+      "step": 7643
+    },
+    {
+      "epoch": 1.4617076202313797,
+      "grad_norm": 4.991281537635345,
+      "learning_rate": 1e-06,
+      "loss": 0.3836,
+      "step": 7644
+    },
+    {
+      "epoch": 1.4618988431016349,
+      "grad_norm": 3.212155299117621,
+      "learning_rate": 1e-06,
+      "loss": 0.2397,
+      "step": 7645
+    },
+    {
+      "epoch": 1.4620900659718903,
+      "grad_norm": 2.4318214196092027,
+      "learning_rate": 1e-06,
+      "loss": 0.0733,
+      "step": 7646
+    },
+    {
+      "epoch": 1.4622812888421455,
+      "grad_norm": 2.383269719517545,
+      "learning_rate": 1e-06,
+      "loss": 0.0842,
+      "step": 7647
+    },
+    {
+      "epoch": 1.462472511712401,
+      "grad_norm": 2.14647004640776,
+      "learning_rate": 1e-06,
+      "loss": 0.0745,
+      "step": 7648
+    },
+    {
+      "epoch": 1.462663734582656,
+      "grad_norm": 3.0621547893665544,
+      "learning_rate": 1e-06,
+      "loss": 0.1114,
+      "step": 7649
+    },
+    {
+      "epoch": 1.4628549574529113,
+      "grad_norm": 1.9285141315728722,
+      "learning_rate": 1e-06,
+      "loss": 0.0967,
+      "step": 7650
+    },
+    {
+      "epoch": 1.4630461803231667,
+      "grad_norm": 1.9883391666033676,
+      "learning_rate": 1e-06,
+      "loss": 0.0546,
+      "step": 7651
+    },
+    {
+      "epoch": 1.4632374031934219,
+      "grad_norm": 2.037493105449596,
+      "learning_rate": 1e-06,
+      "loss": 0.0444,
+      "step": 7652
+    },
+    {
+      "epoch": 1.4634286260636773,
+      "grad_norm": 2.8827724712456373,
+      "learning_rate": 1e-06,
+      "loss": 0.0731,
+      "step": 7653
+    },
+    {
+      "epoch": 1.4636198489339325,
+      "grad_norm": 2.721160521606652,
+      "learning_rate": 1e-06,
+      "loss": 0.0521,
+      "step": 7654
+    },
+    {
+      "epoch": 1.4638110718041877,
+      "grad_norm": 4.555479945870846,
+      "learning_rate": 1e-06,
+      "loss": 0.0894,
+      "step": 7655
+    },
+    {
+      "epoch": 1.464002294674443,
+      "grad_norm": 4.052288661484531,
+      "learning_rate": 1e-06,
+      "loss": 0.2873,
+      "step": 7656
+    },
+    {
+      "epoch": 1.4641935175446983,
+      "grad_norm": 5.006736985045925,
+      "learning_rate": 1e-06,
+      "loss": 0.2321,
+      "step": 7657
+    },
+    {
+      "epoch": 1.4643847404149537,
+      "grad_norm": 2.061145915959722,
+      "learning_rate": 1e-06,
+      "loss": 0.0524,
+      "step": 7658
+    },
+    {
+      "epoch": 1.464575963285209,
+      "grad_norm": 2.629540467529172,
+      "learning_rate": 1e-06,
+      "loss": 0.1985,
+      "step": 7659
+    },
+    {
+      "epoch": 1.464767186155464,
+      "grad_norm": 4.446347724499087,
+      "learning_rate": 1e-06,
+      "loss": 0.2155,
+      "step": 7660
+    },
+    {
+      "epoch": 1.4649584090257195,
+      "grad_norm": 3.2741943067194863,
+      "learning_rate": 1e-06,
+      "loss": 0.1065,
+      "step": 7661
+    },
+    {
+      "epoch": 1.4651496318959747,
+      "grad_norm": 2.9855683023117314,
+      "learning_rate": 1e-06,
+      "loss": 0.0718,
+      "step": 7662
+    },
+    {
+      "epoch": 1.46534085476623,
+      "grad_norm": 1.8928514722137646,
+      "learning_rate": 1e-06,
+      "loss": 0.1085,
+      "step": 7663
+    },
+    {
+      "epoch": 1.4655320776364853,
+      "grad_norm": 4.199150162957316,
+      "learning_rate": 1e-06,
+      "loss": 0.1551,
+      "step": 7664
+    },
+    {
+      "epoch": 1.4657233005067405,
+      "grad_norm": 7.085148238500267,
+      "learning_rate": 1e-06,
+      "loss": 0.368,
+      "step": 7665
+    },
+    {
+      "epoch": 1.465914523376996,
+      "grad_norm": 2.3333142711224064,
+      "learning_rate": 1e-06,
+      "loss": 0.0482,
+      "step": 7666
+    },
+    {
+      "epoch": 1.466105746247251,
+      "grad_norm": 3.389261701189566,
+      "learning_rate": 1e-06,
+      "loss": 0.0908,
+      "step": 7667
+    },
+    {
+      "epoch": 1.4662969691175065,
+      "grad_norm": 7.975448605347725,
+      "learning_rate": 1e-06,
+      "loss": 0.1283,
+      "step": 7668
+    },
+    {
+      "epoch": 1.4664881919877617,
+      "grad_norm": 6.773627853798717,
+      "learning_rate": 1e-06,
+      "loss": 0.4382,
+      "step": 7669
+    },
+    {
+      "epoch": 1.466679414858017,
+      "grad_norm": 6.154634990017813,
+      "learning_rate": 1e-06,
+      "loss": 0.6565,
+      "step": 7670
+    },
+    {
+      "epoch": 1.4668706377282723,
+      "grad_norm": 3.1288932961947054,
+      "learning_rate": 1e-06,
+      "loss": 0.1243,
+      "step": 7671
+    },
+    {
+      "epoch": 1.4670618605985275,
+      "grad_norm": 2.842019529239311,
+      "learning_rate": 1e-06,
+      "loss": 0.0894,
+      "step": 7672
+    },
+    {
+      "epoch": 1.467253083468783,
+      "grad_norm": 4.561968837807205,
+      "learning_rate": 1e-06,
+      "loss": 0.3037,
+      "step": 7673
+    },
+    {
+      "epoch": 1.467444306339038,
+      "grad_norm": 2.7683287082597503,
+      "learning_rate": 1e-06,
+      "loss": 0.203,
+      "step": 7674
+    },
+    {
+      "epoch": 1.4676355292092933,
+      "grad_norm": 3.5362456776985947,
+      "learning_rate": 1e-06,
+      "loss": 0.2473,
+      "step": 7675
+    },
+    {
+      "epoch": 1.4678267520795487,
+      "grad_norm": 3.7008855945516554,
+      "learning_rate": 1e-06,
+      "loss": 0.2264,
+      "step": 7676
+    },
+    {
+      "epoch": 1.4680179749498041,
+      "grad_norm": 4.240056916700571,
+      "learning_rate": 1e-06,
+      "loss": 0.249,
+      "step": 7677
+    },
+    {
+      "epoch": 1.4682091978200593,
+      "grad_norm": 3.1318625819824653,
+      "learning_rate": 1e-06,
+      "loss": 0.0604,
+      "step": 7678
+    },
+    {
+      "epoch": 1.4684004206903145,
+      "grad_norm": 2.3973716330745036,
+      "learning_rate": 1e-06,
+      "loss": 0.0454,
+      "step": 7679
+    },
+    {
+      "epoch": 1.46859164356057,
+      "grad_norm": 9.125565184765613,
+      "learning_rate": 1e-06,
+      "loss": 0.1709,
+      "step": 7680
+    },
+    {
+      "epoch": 1.4687828664308251,
+      "grad_norm": 3.7069376058377648,
+      "learning_rate": 1e-06,
+      "loss": 0.285,
+      "step": 7681
+    },
+    {
+      "epoch": 1.4689740893010805,
+      "grad_norm": 4.003075847577095,
+      "learning_rate": 1e-06,
+      "loss": 0.2711,
+      "step": 7682
+    },
+    {
+      "epoch": 1.4691653121713357,
+      "grad_norm": 2.3787240395505003,
+      "learning_rate": 1e-06,
+      "loss": 0.1294,
+      "step": 7683
+    },
+    {
+      "epoch": 1.469356535041591,
+      "grad_norm": 4.127800424364505,
+      "learning_rate": 1e-06,
+      "loss": 0.1475,
+      "step": 7684
+    },
+    {
+      "epoch": 1.4695477579118463,
+      "grad_norm": 2.5885847204505303,
+      "learning_rate": 1e-06,
+      "loss": 0.0645,
+      "step": 7685
+    },
+    {
+      "epoch": 1.4697389807821015,
+      "grad_norm": 3.728425589863688,
+      "learning_rate": 1e-06,
+      "loss": 0.2972,
+      "step": 7686
+    },
+    {
+      "epoch": 1.469930203652357,
+      "grad_norm": 1.9324801234004054,
+      "learning_rate": 1e-06,
+      "loss": 0.0359,
+      "step": 7687
+    },
+    {
+      "epoch": 1.4701214265226121,
+      "grad_norm": 4.046732425019246,
+      "learning_rate": 1e-06,
+      "loss": 0.0518,
+      "step": 7688
+    },
+    {
+      "epoch": 1.4703126493928673,
+      "grad_norm": 3.6631722415936205,
+      "learning_rate": 1e-06,
+      "loss": 0.0483,
+      "step": 7689
+    },
+    {
+      "epoch": 1.4705038722631227,
+      "grad_norm": 3.0119484740836233,
+      "learning_rate": 1e-06,
+      "loss": 0.0586,
+      "step": 7690
+    },
+    {
+      "epoch": 1.470695095133378,
+      "grad_norm": 3.08069762565523,
+      "learning_rate": 1e-06,
+      "loss": 0.1095,
+      "step": 7691
+    },
+    {
+      "epoch": 1.4708863180036333,
+      "grad_norm": 2.649784162116689,
+      "learning_rate": 1e-06,
+      "loss": 0.0511,
+      "step": 7692
+    },
+    {
+      "epoch": 1.4710775408738885,
+      "grad_norm": 7.014790576682728,
+      "learning_rate": 1e-06,
+      "loss": 0.3084,
+      "step": 7693
+    },
+    {
+      "epoch": 1.4712687637441437,
+      "grad_norm": 3.6221149407890287,
+      "learning_rate": 1e-06,
+      "loss": 0.122,
+      "step": 7694
+    },
+    {
+      "epoch": 1.4714599866143991,
+      "grad_norm": 3.424488480510544,
+      "learning_rate": 1e-06,
+      "loss": 0.4483,
+      "step": 7695
+    },
+    {
+      "epoch": 1.4716512094846543,
+      "grad_norm": 5.854429553590103,
+      "learning_rate": 1e-06,
+      "loss": 0.3234,
+      "step": 7696
+    },
+    {
+      "epoch": 1.4718424323549097,
+      "grad_norm": 3.6603081507800668,
+      "learning_rate": 1e-06,
+      "loss": 0.268,
+      "step": 7697
+    },
+    {
+      "epoch": 1.472033655225165,
+      "grad_norm": 2.4058425360163262,
+      "learning_rate": 1e-06,
+      "loss": 0.0864,
+      "step": 7698
+    },
+    {
+      "epoch": 1.4722248780954201,
+      "grad_norm": 3.6069968840160187,
+      "learning_rate": 1e-06,
+      "loss": 0.2177,
+      "step": 7699
+    },
+    {
+      "epoch": 1.4724161009656755,
+      "grad_norm": 2.5720256441478284,
+      "learning_rate": 1e-06,
+      "loss": 0.0751,
+      "step": 7700
+    },
+    {
+      "epoch": 1.4726073238359307,
+      "grad_norm": 2.0829943063175587,
+      "learning_rate": 1e-06,
+      "loss": 0.0748,
+      "step": 7701
+    },
+    {
+      "epoch": 1.4727985467061862,
+      "grad_norm": 5.261448321511996,
+      "learning_rate": 1e-06,
+      "loss": 0.2628,
+      "step": 7702
+    },
+    {
+      "epoch": 1.4729897695764413,
+      "grad_norm": 2.7379043907486724,
+      "learning_rate": 1e-06,
+      "loss": 0.0632,
+      "step": 7703
+    },
+    {
+      "epoch": 1.4731809924466965,
+      "grad_norm": 3.965301336933675,
+      "learning_rate": 1e-06,
+      "loss": 0.0696,
+      "step": 7704
+    },
+    {
+      "epoch": 1.473372215316952,
+      "grad_norm": 4.234663049113396,
+      "learning_rate": 1e-06,
+      "loss": 0.1699,
+      "step": 7705
+    },
+    {
+      "epoch": 1.4735634381872071,
+      "grad_norm": 2.634739201938485,
+      "learning_rate": 1e-06,
+      "loss": 0.2001,
+      "step": 7706
+    },
+    {
+      "epoch": 1.4737546610574626,
+      "grad_norm": 5.249415410509975,
+      "learning_rate": 1e-06,
+      "loss": 0.4988,
+      "step": 7707
+    },
+    {
+      "epoch": 1.4739458839277177,
+      "grad_norm": 4.606544996569546,
+      "learning_rate": 1e-06,
+      "loss": 0.3826,
+      "step": 7708
+    },
+    {
+      "epoch": 1.474137106797973,
+      "grad_norm": 4.903582193856771,
+      "learning_rate": 1e-06,
+      "loss": 0.2175,
+      "step": 7709
+    },
+    {
+      "epoch": 1.4743283296682284,
+      "grad_norm": 2.688995477527179,
+      "learning_rate": 1e-06,
+      "loss": 0.073,
+      "step": 7710
+    },
+    {
+      "epoch": 1.4745195525384835,
+      "grad_norm": 3.5646428138739132,
+      "learning_rate": 1e-06,
+      "loss": 0.0928,
+      "step": 7711
+    },
+    {
+      "epoch": 1.474710775408739,
+      "grad_norm": 4.201963356333993,
+      "learning_rate": 1e-06,
+      "loss": 0.2886,
+      "step": 7712
+    },
+    {
+      "epoch": 1.4749019982789942,
+      "grad_norm": 2.074467933184446,
+      "learning_rate": 1e-06,
+      "loss": 0.0253,
+      "step": 7713
+    },
+    {
+      "epoch": 1.4750932211492493,
+      "grad_norm": 4.389967248051498,
+      "learning_rate": 1e-06,
+      "loss": 0.2059,
+      "step": 7714
+    },
+    {
+      "epoch": 1.4752844440195048,
+      "grad_norm": 2.183360406353468,
+      "learning_rate": 1e-06,
+      "loss": 0.0378,
+      "step": 7715
+    },
+    {
+      "epoch": 1.47547566688976,
+      "grad_norm": 4.116440649173028,
+      "learning_rate": 1e-06,
+      "loss": 0.0875,
+      "step": 7716
+    },
+    {
+      "epoch": 1.4756668897600154,
+      "grad_norm": 3.4517157945831385,
+      "learning_rate": 1e-06,
+      "loss": 0.09,
+      "step": 7717
+    },
+    {
+      "epoch": 1.4758581126302706,
+      "grad_norm": 4.709576715453855,
+      "learning_rate": 1e-06,
+      "loss": 0.1826,
+      "step": 7718
+    },
+    {
+      "epoch": 1.4760493355005257,
+      "grad_norm": 3.4730361713488613,
+      "learning_rate": 1e-06,
+      "loss": 0.4056,
+      "step": 7719
+    },
+    {
+      "epoch": 1.4762405583707812,
+      "grad_norm": 5.005140332551641,
+      "learning_rate": 1e-06,
+      "loss": 0.4567,
+      "step": 7720
+    },
+    {
+      "epoch": 1.4764317812410364,
+      "grad_norm": 1.612032649601725,
+      "learning_rate": 1e-06,
+      "loss": 0.0546,
+      "step": 7721
+    },
+    {
+      "epoch": 1.4766230041112918,
+      "grad_norm": 2.730805775985767,
+      "learning_rate": 1e-06,
+      "loss": 0.0842,
+      "step": 7722
+    },
+    {
+      "epoch": 1.476814226981547,
+      "grad_norm": 4.5186683404701204,
+      "learning_rate": 1e-06,
+      "loss": 0.5249,
+      "step": 7723
+    },
+    {
+      "epoch": 1.4770054498518022,
+      "grad_norm": 4.114298492744825,
+      "learning_rate": 1e-06,
+      "loss": 0.2454,
+      "step": 7724
+    },
+    {
+      "epoch": 1.4771966727220576,
+      "grad_norm": 3.6053957451303327,
+      "learning_rate": 1e-06,
+      "loss": 0.0636,
+      "step": 7725
+    },
+    {
+      "epoch": 1.4773878955923128,
+      "grad_norm": 3.387280198739896,
+      "learning_rate": 1e-06,
+      "loss": 0.1124,
+      "step": 7726
+    },
+    {
+      "epoch": 1.4775791184625682,
+      "grad_norm": 2.4168511243557815,
+      "learning_rate": 1e-06,
+      "loss": 0.1444,
+      "step": 7727
+    },
+    {
+      "epoch": 1.4777703413328234,
+      "grad_norm": 2.927714342044955,
+      "learning_rate": 1e-06,
+      "loss": 0.0533,
+      "step": 7728
+    },
+    {
+      "epoch": 1.4779615642030786,
+      "grad_norm": 3.014593709010837,
+      "learning_rate": 1e-06,
+      "loss": 0.0565,
+      "step": 7729
+    },
+    {
+      "epoch": 1.478152787073334,
+      "grad_norm": 6.146677698441299,
+      "learning_rate": 1e-06,
+      "loss": 0.0706,
+      "step": 7730
+    },
+    {
+      "epoch": 1.4783440099435894,
+      "grad_norm": 5.091803055771565,
+      "learning_rate": 1e-06,
+      "loss": 0.1631,
+      "step": 7731
+    },
+    {
+      "epoch": 1.4785352328138446,
+      "grad_norm": 5.167731616288079,
+      "learning_rate": 1e-06,
+      "loss": 0.4061,
+      "step": 7732
+    },
+    {
+      "epoch": 1.4787264556840998,
+      "grad_norm": 1.9442152122827743,
+      "learning_rate": 1e-06,
+      "loss": 0.0736,
+      "step": 7733
+    },
+    {
+      "epoch": 1.478917678554355,
+      "grad_norm": 1.9245146944709512,
+      "learning_rate": 1e-06,
+      "loss": 0.1105,
+      "step": 7734
+    },
+    {
+      "epoch": 1.4791089014246104,
+      "grad_norm": 4.240317366256581,
+      "learning_rate": 1e-06,
+      "loss": 0.1469,
+      "step": 7735
+    },
+    {
+      "epoch": 1.4793001242948658,
+      "grad_norm": 4.6017830088294644,
+      "learning_rate": 1e-06,
+      "loss": 0.4328,
+      "step": 7736
+    },
+    {
+      "epoch": 1.479491347165121,
+      "grad_norm": 3.0306015107512523,
+      "learning_rate": 1e-06,
+      "loss": 0.0765,
+      "step": 7737
+    },
+    {
+      "epoch": 1.4796825700353762,
+      "grad_norm": 2.812955522735506,
+      "learning_rate": 1e-06,
+      "loss": 0.1329,
+      "step": 7738
+    },
+    {
+      "epoch": 1.4798737929056316,
+      "grad_norm": 3.23397618290536,
+      "learning_rate": 1e-06,
+      "loss": 0.0934,
+      "step": 7739
+    },
+    {
+      "epoch": 1.4800650157758868,
+      "grad_norm": 2.5041389535734058,
+      "learning_rate": 1e-06,
+      "loss": 0.0967,
+      "step": 7740
+    },
+    {
+      "epoch": 1.4802562386461422,
+      "grad_norm": 1.1627318181888495,
+      "learning_rate": 1e-06,
+      "loss": 0.0144,
+      "step": 7741
+    },
+    {
+      "epoch": 1.4804474615163974,
+      "grad_norm": 2.999244594835246,
+      "learning_rate": 1e-06,
+      "loss": 0.0722,
+      "step": 7742
+    },
+    {
+      "epoch": 1.4806386843866526,
+      "grad_norm": 5.295847745326031,
+      "learning_rate": 1e-06,
+      "loss": 0.3194,
+      "step": 7743
+    },
+    {
+      "epoch": 1.480829907256908,
+      "grad_norm": 4.422109213513415,
+      "learning_rate": 1e-06,
+      "loss": 0.2449,
+      "step": 7744
+    },
+    {
+      "epoch": 1.4810211301271632,
+      "grad_norm": 3.180602295896273,
+      "learning_rate": 1e-06,
+      "loss": 0.1368,
+      "step": 7745
+    },
+    {
+      "epoch": 1.4812123529974186,
+      "grad_norm": 1.6572898623370553,
+      "learning_rate": 1e-06,
+      "loss": 0.0566,
+      "step": 7746
+    },
+    {
+      "epoch": 1.4814035758676738,
+      "grad_norm": 2.172351592759772,
+      "learning_rate": 1e-06,
+      "loss": 0.0379,
+      "step": 7747
+    },
+    {
+      "epoch": 1.481594798737929,
+      "grad_norm": 3.848776655163691,
+      "learning_rate": 1e-06,
+      "loss": 0.2796,
+      "step": 7748
+    },
+    {
+      "epoch": 1.4817860216081844,
+      "grad_norm": 0.7313989438688494,
+      "learning_rate": 1e-06,
+      "loss": 0.0251,
+      "step": 7749
+    },
+    {
+      "epoch": 1.4819772444784396,
+      "grad_norm": 3.6616636447491775,
+      "learning_rate": 1e-06,
+      "loss": 0.0485,
+      "step": 7750
+    },
+    {
+      "epoch": 1.482168467348695,
+      "grad_norm": 2.7093414973351235,
+      "learning_rate": 1e-06,
+      "loss": 0.0515,
+      "step": 7751
+    },
+    {
+      "epoch": 1.4823596902189502,
+      "grad_norm": 1.7131671051674033,
+      "learning_rate": 1e-06,
+      "loss": 0.0722,
+      "step": 7752
+    },
+    {
+      "epoch": 1.4825509130892054,
+      "grad_norm": 2.914209275218248,
+      "learning_rate": 1e-06,
+      "loss": 0.0571,
+      "step": 7753
+    },
+    {
+      "epoch": 1.4827421359594608,
+      "grad_norm": 2.156413417649738,
+      "learning_rate": 1e-06,
+      "loss": 0.0397,
+      "step": 7754
+    },
+    {
+      "epoch": 1.482933358829716,
+      "grad_norm": 5.015761519145758,
+      "learning_rate": 1e-06,
+      "loss": 0.1802,
+      "step": 7755
+    },
+    {
+      "epoch": 1.4831245816999714,
+      "grad_norm": 9.240237884101532,
+      "learning_rate": 1e-06,
+      "loss": 0.577,
+      "step": 7756
+    },
+    {
+      "epoch": 1.4833158045702266,
+      "grad_norm": 2.8942874285803546,
+      "learning_rate": 1e-06,
+      "loss": 0.1035,
+      "step": 7757
+    },
+    {
+      "epoch": 1.4835070274404818,
+      "grad_norm": 1.7317335617146539,
+      "learning_rate": 1e-06,
+      "loss": 0.1028,
+      "step": 7758
+    },
+    {
+      "epoch": 1.4836982503107372,
+      "grad_norm": 3.933270195815255,
+      "learning_rate": 1e-06,
+      "loss": 0.4461,
+      "step": 7759
+    },
+    {
+      "epoch": 1.4838894731809924,
+      "grad_norm": 2.3246014448263637,
+      "learning_rate": 1e-06,
+      "loss": 0.1101,
+      "step": 7760
+    },
+    {
+      "epoch": 1.4840806960512478,
+      "grad_norm": 2.9328935907663443,
+      "learning_rate": 1e-06,
+      "loss": 0.0813,
+      "step": 7761
+    },
+    {
+      "epoch": 1.484271918921503,
+      "grad_norm": 2.2222484891716614,
+      "learning_rate": 1e-06,
+      "loss": 0.072,
+      "step": 7762
+    },
+    {
+      "epoch": 1.4844631417917582,
+      "grad_norm": 1.7728232105455408,
+      "learning_rate": 1e-06,
+      "loss": 0.101,
+      "step": 7763
+    },
+    {
+      "epoch": 1.4846543646620136,
+      "grad_norm": 3.4278631897982206,
+      "learning_rate": 1e-06,
+      "loss": 0.0518,
+      "step": 7764
+    },
+    {
+      "epoch": 1.4848455875322688,
+      "grad_norm": 2.816041877868679,
+      "learning_rate": 1e-06,
+      "loss": 0.0486,
+      "step": 7765
+    },
+    {
+      "epoch": 1.4850368104025242,
+      "grad_norm": 2.303731429076413,
+      "learning_rate": 1e-06,
+      "loss": 0.0305,
+      "step": 7766
+    },
+    {
+      "epoch": 1.4852280332727794,
+      "grad_norm": 5.002089826629165,
+      "learning_rate": 1e-06,
+      "loss": 0.0555,
+      "step": 7767
+    },
+    {
+      "epoch": 1.4854192561430346,
+      "grad_norm": 7.009101128298049,
+      "learning_rate": 1e-06,
+      "loss": 0.2224,
+      "step": 7768
+    },
+    {
+      "epoch": 1.48561047901329,
+      "grad_norm": 3.3271276564602705,
+      "learning_rate": 1e-06,
+      "loss": 0.0931,
+      "step": 7769
+    },
+    {
+      "epoch": 1.4858017018835452,
+      "grad_norm": 3.7725963089305288,
+      "learning_rate": 1e-06,
+      "loss": 0.2406,
+      "step": 7770
+    },
+    {
+      "epoch": 1.4859929247538006,
+      "grad_norm": 2.277344796914961,
+      "learning_rate": 1e-06,
+      "loss": 0.1229,
+      "step": 7771
+    },
+    {
+      "epoch": 1.4861841476240558,
+      "grad_norm": 3.1694926819254228,
+      "learning_rate": 1e-06,
+      "loss": 0.0929,
+      "step": 7772
+    },
+    {
+      "epoch": 1.486375370494311,
+      "grad_norm": 1.798909900972705,
+      "learning_rate": 1e-06,
+      "loss": 0.0622,
+      "step": 7773
+    },
+    {
+      "epoch": 1.4865665933645664,
+      "grad_norm": 1.8428509993572213,
+      "learning_rate": 1e-06,
+      "loss": 0.0809,
+      "step": 7774
+    },
+    {
+      "epoch": 1.4867578162348216,
+      "grad_norm": 2.9771253799519597,
+      "learning_rate": 1e-06,
+      "loss": 0.086,
+      "step": 7775
+    },
+    {
+      "epoch": 1.486949039105077,
+      "grad_norm": 6.358529449411491,
+      "learning_rate": 1e-06,
+      "loss": 0.0643,
+      "step": 7776
+    },
+    {
+      "epoch": 1.4871402619753322,
+      "grad_norm": 4.014537144414873,
+      "learning_rate": 1e-06,
+      "loss": 0.0933,
+      "step": 7777
+    },
+    {
+      "epoch": 1.4873314848455874,
+      "grad_norm": 3.51393486934394,
+      "learning_rate": 1e-06,
+      "loss": 0.1243,
+      "step": 7778
+    },
+    {
+      "epoch": 1.4875227077158428,
+      "grad_norm": 1.7084970705913183,
+      "learning_rate": 1e-06,
+      "loss": 0.0226,
+      "step": 7779
+    },
+    {
+      "epoch": 1.487713930586098,
+      "grad_norm": 4.5431150640803315,
+      "learning_rate": 1e-06,
+      "loss": 0.1018,
+      "step": 7780
+    },
+    {
+      "epoch": 1.4879051534563534,
+      "grad_norm": 6.848457285221249,
+      "learning_rate": 1e-06,
+      "loss": 0.2199,
+      "step": 7781
+    },
+    {
+      "epoch": 1.4880963763266086,
+      "grad_norm": 3.9083387969954204,
+      "learning_rate": 1e-06,
+      "loss": 0.2909,
+      "step": 7782
+    },
+    {
+      "epoch": 1.4882875991968638,
+      "grad_norm": 5.239230053660154,
+      "learning_rate": 1e-06,
+      "loss": 0.3612,
+      "step": 7783
+    },
+    {
+      "epoch": 1.4884788220671192,
+      "grad_norm": 3.070958348649194,
+      "learning_rate": 1e-06,
+      "loss": 0.3445,
+      "step": 7784
+    },
+    {
+      "epoch": 1.4886700449373746,
+      "grad_norm": 2.1093266234325325,
+      "learning_rate": 1e-06,
+      "loss": 0.1181,
+      "step": 7785
+    },
+    {
+      "epoch": 1.4888612678076298,
+      "grad_norm": 4.3850409678303786,
+      "learning_rate": 1e-06,
+      "loss": 0.2778,
+      "step": 7786
+    },
+    {
+      "epoch": 1.489052490677885,
+      "grad_norm": 2.9172269192255844,
+      "learning_rate": 1e-06,
+      "loss": 0.1937,
+      "step": 7787
+    },
+    {
+      "epoch": 1.4892437135481402,
+      "grad_norm": 5.819296851925504,
+      "learning_rate": 1e-06,
+      "loss": 0.064,
+      "step": 7788
+    },
+    {
+      "epoch": 1.4894349364183956,
+      "grad_norm": 2.9559438663877486,
+      "learning_rate": 1e-06,
+      "loss": 0.0638,
+      "step": 7789
+    },
+    {
+      "epoch": 1.489626159288651,
+      "grad_norm": 1.7308792075241497,
+      "learning_rate": 1e-06,
+      "loss": 0.033,
+      "step": 7790
+    },
+    {
+      "epoch": 1.4898173821589062,
+      "grad_norm": 2.3894438973685643,
+      "learning_rate": 1e-06,
+      "loss": 0.0536,
+      "step": 7791
+    },
+    {
+      "epoch": 1.4900086050291614,
+      "grad_norm": 4.347272802890785,
+      "learning_rate": 1e-06,
+      "loss": 0.0459,
+      "step": 7792
+    },
+    {
+      "epoch": 1.4901998278994169,
+      "grad_norm": 8.537579211290662,
+      "learning_rate": 1e-06,
+      "loss": 0.3277,
+      "step": 7793
+    },
+    {
+      "epoch": 1.490391050769672,
+      "grad_norm": 4.047306700550418,
+      "learning_rate": 1e-06,
+      "loss": 0.3168,
+      "step": 7794
+    },
+    {
+      "epoch": 1.4905822736399275,
+      "grad_norm": 4.130684836738212,
+      "learning_rate": 1e-06,
+      "loss": 0.1711,
+      "step": 7795
+    },
+    {
+      "epoch": 1.4907734965101826,
+      "grad_norm": 4.3175940903051195,
+      "learning_rate": 1e-06,
+      "loss": 0.3653,
+      "step": 7796
+    },
+    {
+      "epoch": 1.4909647193804378,
+      "grad_norm": 3.4109643631915727,
+      "learning_rate": 1e-06,
+      "loss": 0.1401,
+      "step": 7797
+    },
+    {
+      "epoch": 1.4911559422506933,
+      "grad_norm": 3.905554137237952,
+      "learning_rate": 1e-06,
+      "loss": 0.1212,
+      "step": 7798
+    },
+    {
+      "epoch": 1.4913471651209484,
+      "grad_norm": 1.881046907735411,
+      "learning_rate": 1e-06,
+      "loss": 0.0432,
+      "step": 7799
+    },
+    {
+      "epoch": 1.4915383879912039,
+      "grad_norm": 4.96509575973317,
+      "learning_rate": 1e-06,
+      "loss": 0.0668,
+      "step": 7800
+    },
+    {
+      "epoch": 1.491729610861459,
+      "grad_norm": 4.019731018025384,
+      "learning_rate": 1e-06,
+      "loss": 0.2851,
+      "step": 7801
+    },
+    {
+      "epoch": 1.4919208337317142,
+      "grad_norm": 4.324689362235626,
+      "learning_rate": 1e-06,
+      "loss": 0.0644,
+      "step": 7802
+    },
+    {
+      "epoch": 1.4921120566019697,
+      "grad_norm": 2.471443155908746,
+      "learning_rate": 1e-06,
+      "loss": 0.03,
+      "step": 7803
+    },
+    {
+      "epoch": 1.4923032794722249,
+      "grad_norm": 2.663230161884646,
+      "learning_rate": 1e-06,
+      "loss": 0.0542,
+      "step": 7804
+    },
+    {
+      "epoch": 1.4924945023424803,
+      "grad_norm": 5.896651284280613,
+      "learning_rate": 1e-06,
+      "loss": 0.1141,
+      "step": 7805
+    },
+    {
+      "epoch": 1.4926857252127355,
+      "grad_norm": 4.030716735108241,
+      "learning_rate": 1e-06,
+      "loss": 0.2529,
+      "step": 7806
+    },
+    {
+      "epoch": 1.4928769480829907,
+      "grad_norm": 4.141872693060179,
+      "learning_rate": 1e-06,
+      "loss": 0.3221,
+      "step": 7807
+    },
+    {
+      "epoch": 1.493068170953246,
+      "grad_norm": 1.7168611724929927,
+      "learning_rate": 1e-06,
+      "loss": 0.1728,
+      "step": 7808
+    },
+    {
+      "epoch": 1.4932593938235013,
+      "grad_norm": 3.6122579579297027,
+      "learning_rate": 1e-06,
+      "loss": 0.3246,
+      "step": 7809
+    },
+    {
+      "epoch": 1.4934506166937567,
+      "grad_norm": 4.437293464259928,
+      "learning_rate": 1e-06,
+      "loss": 0.2259,
+      "step": 7810
+    },
+    {
+      "epoch": 1.4936418395640119,
+      "grad_norm": 2.7482255933134683,
+      "learning_rate": 1e-06,
+      "loss": 0.2025,
+      "step": 7811
+    },
+    {
+      "epoch": 1.493833062434267,
+      "grad_norm": 3.027701276771258,
+      "learning_rate": 1e-06,
+      "loss": 0.1043,
+      "step": 7812
+    },
+    {
+      "epoch": 1.4940242853045225,
+      "grad_norm": 1.7888659906534305,
+      "learning_rate": 1e-06,
+      "loss": 0.0574,
+      "step": 7813
+    },
+    {
+      "epoch": 1.4942155081747777,
+      "grad_norm": 5.334419716020201,
+      "learning_rate": 1e-06,
+      "loss": 0.0635,
+      "step": 7814
+    },
+    {
+      "epoch": 1.494406731045033,
+      "grad_norm": 3.080886608925766,
+      "learning_rate": 1e-06,
+      "loss": 0.2008,
+      "step": 7815
+    },
+    {
+      "epoch": 1.4945979539152883,
+      "grad_norm": 3.5205493940418835,
+      "learning_rate": 1e-06,
+      "loss": 0.1488,
+      "step": 7816
+    },
+    {
+      "epoch": 1.4947891767855435,
+      "grad_norm": 11.062041019896776,
+      "learning_rate": 1e-06,
+      "loss": 0.1446,
+      "step": 7817
+    },
+    {
+      "epoch": 1.4949803996557989,
+      "grad_norm": 6.698833238725525,
+      "learning_rate": 1e-06,
+      "loss": 0.3904,
+      "step": 7818
+    },
+    {
+      "epoch": 1.495171622526054,
+      "grad_norm": 5.161187599490431,
+      "learning_rate": 1e-06,
+      "loss": 0.4504,
+      "step": 7819
+    },
+    {
+      "epoch": 1.4953628453963095,
+      "grad_norm": 2.529538741753665,
+      "learning_rate": 1e-06,
+      "loss": 0.1289,
+      "step": 7820
+    },
+    {
+      "epoch": 1.4955540682665647,
+      "grad_norm": 2.999876019777018,
+      "learning_rate": 1e-06,
+      "loss": 0.0817,
+      "step": 7821
+    },
+    {
+      "epoch": 1.4957452911368199,
+      "grad_norm": 4.356898982171233,
+      "learning_rate": 1e-06,
+      "loss": 0.2629,
+      "step": 7822
+    },
+    {
+      "epoch": 1.4959365140070753,
+      "grad_norm": 4.129599578879866,
+      "learning_rate": 1e-06,
+      "loss": 0.2038,
+      "step": 7823
+    },
+    {
+      "epoch": 1.4961277368773305,
+      "grad_norm": 4.154011977110977,
+      "learning_rate": 1e-06,
+      "loss": 0.0978,
+      "step": 7824
+    },
+    {
+      "epoch": 1.4963189597475859,
+      "grad_norm": 2.4030890572468024,
+      "learning_rate": 1e-06,
+      "loss": 0.0912,
+      "step": 7825
+    },
+    {
+      "epoch": 1.496510182617841,
+      "grad_norm": 2.8650399878675477,
+      "learning_rate": 1e-06,
+      "loss": 0.2182,
+      "step": 7826
+    },
+    {
+      "epoch": 1.4967014054880963,
+      "grad_norm": 1.436127297494468,
+      "learning_rate": 1e-06,
+      "loss": 0.0351,
+      "step": 7827
+    },
+    {
+      "epoch": 1.4968926283583517,
+      "grad_norm": 2.489728330169888,
+      "learning_rate": 1e-06,
+      "loss": 0.0516,
+      "step": 7828
+    },
+    {
+      "epoch": 1.4970838512286069,
+      "grad_norm": 4.863077697170867,
+      "learning_rate": 1e-06,
+      "loss": 0.0436,
+      "step": 7829
+    },
+    {
+      "epoch": 1.4972750740988623,
+      "grad_norm": 7.556448793281042,
+      "learning_rate": 1e-06,
+      "loss": 0.2364,
+      "step": 7830
+    },
+    {
+      "epoch": 1.4974662969691175,
+      "grad_norm": 2.632161951485675,
+      "learning_rate": 1e-06,
+      "loss": 0.1437,
+      "step": 7831
+    },
+    {
+      "epoch": 1.4976575198393727,
+      "grad_norm": 2.9496499257968596,
+      "learning_rate": 1e-06,
+      "loss": 0.1743,
+      "step": 7832
+    },
+    {
+      "epoch": 1.497848742709628,
+      "grad_norm": 3.5539233381925954,
+      "learning_rate": 1e-06,
+      "loss": 0.0881,
+      "step": 7833
+    },
+    {
+      "epoch": 1.4980399655798833,
+      "grad_norm": 4.818457742776302,
+      "learning_rate": 1e-06,
+      "loss": 0.4057,
+      "step": 7834
+    },
+    {
+      "epoch": 1.4982311884501387,
+      "grad_norm": 3.957914323547379,
+      "learning_rate": 1e-06,
+      "loss": 0.3558,
+      "step": 7835
+    },
+    {
+      "epoch": 1.4984224113203939,
+      "grad_norm": 2.3790735145232698,
+      "learning_rate": 1e-06,
+      "loss": 0.0538,
+      "step": 7836
+    },
+    {
+      "epoch": 1.498613634190649,
+      "grad_norm": 2.5279505383022793,
+      "learning_rate": 1e-06,
+      "loss": 0.1286,
+      "step": 7837
+    },
+    {
+      "epoch": 1.4988048570609045,
+      "grad_norm": 1.9454501306342735,
+      "learning_rate": 1e-06,
+      "loss": 0.0721,
+      "step": 7838
+    },
+    {
+      "epoch": 1.4989960799311597,
+      "grad_norm": 3.252358534461127,
+      "learning_rate": 1e-06,
+      "loss": 0.0622,
+      "step": 7839
+    },
+    {
+      "epoch": 1.499187302801415,
+      "grad_norm": 1.6848162043139083,
+      "learning_rate": 1e-06,
+      "loss": 0.0456,
+      "step": 7840
+    },
+    {
+      "epoch": 1.4993785256716703,
+      "grad_norm": 2.3697999947090675,
+      "learning_rate": 1e-06,
+      "loss": 0.0636,
+      "step": 7841
+    },
+    {
+      "epoch": 1.4995697485419255,
+      "grad_norm": 8.406205528170004,
+      "learning_rate": 1e-06,
+      "loss": 0.0922,
+      "step": 7842
+    },
+    {
+      "epoch": 1.499760971412181,
+      "grad_norm": 5.747131959232878,
+      "learning_rate": 1e-06,
+      "loss": 0.3326,
+      "step": 7843
+    },
+    {
+      "epoch": 1.4999521942824363,
+      "grad_norm": 5.325979746202384,
+      "learning_rate": 1e-06,
+      "loss": 0.3692,
+      "step": 7844
+    },
+    {
+      "epoch": 1.5001434171526915,
+      "grad_norm": 3.4255227629729736,
+      "learning_rate": 1e-06,
+      "loss": 0.1991,
+      "step": 7845
+    },
+    {
+      "epoch": 1.5003346400229467,
+      "grad_norm": 3.045146588835677,
+      "learning_rate": 1e-06,
+      "loss": 0.0705,
+      "step": 7846
+    },
+    {
+      "epoch": 1.500525862893202,
+      "grad_norm": 4.995915842952449,
+      "learning_rate": 1e-06,
+      "loss": 0.2628,
+      "step": 7847
+    },
+    {
+      "epoch": 1.5007170857634573,
+      "grad_norm": 2.2028230500776984,
+      "learning_rate": 1e-06,
+      "loss": 0.144,
+      "step": 7848
+    },
+    {
+      "epoch": 1.5009083086337127,
+      "grad_norm": 1.6587460261558187,
+      "learning_rate": 1e-06,
+      "loss": 0.0325,
+      "step": 7849
+    },
+    {
+      "epoch": 1.501099531503968,
+      "grad_norm": 2.6531321459203308,
+      "learning_rate": 1e-06,
+      "loss": 0.0299,
+      "step": 7850
+    },
+    {
+      "epoch": 1.501290754374223,
+      "grad_norm": 6.7000173198419155,
+      "learning_rate": 1e-06,
+      "loss": 0.1505,
+      "step": 7851
+    },
+    {
+      "epoch": 1.5014819772444783,
+      "grad_norm": 6.799126894379815,
+      "learning_rate": 1e-06,
+      "loss": 0.0588,
+      "step": 7852
+    },
+    {
+      "epoch": 1.5016732001147337,
+      "grad_norm": 2.879476131106458,
+      "learning_rate": 1e-06,
+      "loss": 0.0703,
+      "step": 7853
+    },
+    {
+      "epoch": 1.5018644229849891,
+      "grad_norm": 5.086516504100264,
+      "learning_rate": 1e-06,
+      "loss": 0.101,
+      "step": 7854
+    },
+    {
+      "epoch": 1.5020556458552443,
+      "grad_norm": 9.138407850762352,
+      "learning_rate": 1e-06,
+      "loss": 0.1784,
+      "step": 7855
+    },
+    {
+      "epoch": 1.5022468687254995,
+      "grad_norm": 4.159479936956934,
+      "learning_rate": 1e-06,
+      "loss": 0.3063,
+      "step": 7856
+    },
+    {
+      "epoch": 1.5024380915957547,
+      "grad_norm": 3.387322571109021,
+      "learning_rate": 1e-06,
+      "loss": 0.2315,
+      "step": 7857
+    },
+    {
+      "epoch": 1.5026293144660101,
+      "grad_norm": 6.287556836457201,
+      "learning_rate": 1e-06,
+      "loss": 0.5455,
+      "step": 7858
+    },
+    {
+      "epoch": 1.5028205373362655,
+      "grad_norm": 4.876958795864244,
+      "learning_rate": 1e-06,
+      "loss": 0.2918,
+      "step": 7859
+    },
+    {
+      "epoch": 1.5030117602065207,
+      "grad_norm": 3.593345420288631,
+      "learning_rate": 1e-06,
+      "loss": 0.2033,
+      "step": 7860
+    },
+    {
+      "epoch": 1.503202983076776,
+      "grad_norm": 2.2428028711030157,
+      "learning_rate": 1e-06,
+      "loss": 0.0781,
+      "step": 7861
+    },
+    {
+      "epoch": 1.503394205947031,
+      "grad_norm": 1.8681621481710657,
+      "learning_rate": 1e-06,
+      "loss": 0.1327,
+      "step": 7862
+    },
+    {
+      "epoch": 1.5035854288172865,
+      "grad_norm": 3.709985357908572,
+      "learning_rate": 1e-06,
+      "loss": 0.0156,
+      "step": 7863
+    },
+    {
+      "epoch": 1.503776651687542,
+      "grad_norm": 3.418120811349629,
+      "learning_rate": 1e-06,
+      "loss": 0.2134,
+      "step": 7864
+    },
+    {
+      "epoch": 1.5039678745577971,
+      "grad_norm": 2.928886283670108,
+      "learning_rate": 1e-06,
+      "loss": 0.1047,
+      "step": 7865
+    },
+    {
+      "epoch": 1.5041590974280523,
+      "grad_norm": 3.8283808408609112,
+      "learning_rate": 1e-06,
+      "loss": 0.0874,
+      "step": 7866
+    },
+    {
+      "epoch": 1.5043503202983077,
+      "grad_norm": 5.616425485083211,
+      "learning_rate": 1e-06,
+      "loss": 0.1215,
+      "step": 7867
+    },
+    {
+      "epoch": 1.504541543168563,
+      "grad_norm": 4.025298227196476,
+      "learning_rate": 1e-06,
+      "loss": 0.0893,
+      "step": 7868
+    },
+    {
+      "epoch": 1.5047327660388183,
+      "grad_norm": 5.019803882126736,
+      "learning_rate": 1e-06,
+      "loss": 0.3669,
+      "step": 7869
+    },
+    {
+      "epoch": 1.5049239889090735,
+      "grad_norm": 4.93754616244985,
+      "learning_rate": 1e-06,
+      "loss": 0.3574,
+      "step": 7870
+    },
+    {
+      "epoch": 1.5051152117793287,
+      "grad_norm": 4.595311321218184,
+      "learning_rate": 1e-06,
+      "loss": 0.2693,
+      "step": 7871
+    },
+    {
+      "epoch": 1.5053064346495841,
+      "grad_norm": 4.751569187569796,
+      "learning_rate": 1e-06,
+      "loss": 0.4948,
+      "step": 7872
+    },
+    {
+      "epoch": 1.5054976575198393,
+      "grad_norm": 8.313214493641143,
+      "learning_rate": 1e-06,
+      "loss": 0.3662,
+      "step": 7873
+    },
+    {
+      "epoch": 1.5056888803900947,
+      "grad_norm": 1.839870315003613,
+      "learning_rate": 1e-06,
+      "loss": 0.0455,
+      "step": 7874
+    },
+    {
+      "epoch": 1.50588010326035,
+      "grad_norm": 10.228368496232747,
+      "learning_rate": 1e-06,
+      "loss": 0.0794,
+      "step": 7875
+    },
+    {
+      "epoch": 1.5060713261306051,
+      "grad_norm": 2.937589928082422,
+      "learning_rate": 1e-06,
+      "loss": 0.1613,
+      "step": 7876
+    },
+    {
+      "epoch": 1.5062625490008605,
+      "grad_norm": 2.78228269114035,
+      "learning_rate": 1e-06,
+      "loss": 0.0133,
+      "step": 7877
+    },
+    {
+      "epoch": 1.5064537718711157,
+      "grad_norm": 1.5259278117162731,
+      "learning_rate": 1e-06,
+      "loss": 0.0222,
+      "step": 7878
+    },
+    {
+      "epoch": 1.5066449947413711,
+      "grad_norm": 6.975958273631603,
+      "learning_rate": 1e-06,
+      "loss": 0.0365,
+      "step": 7879
+    },
+    {
+      "epoch": 1.5068362176116263,
+      "grad_norm": 6.192937754195802,
+      "learning_rate": 1e-06,
+      "loss": 0.1759,
+      "step": 7880
+    },
+    {
+      "epoch": 1.5070274404818815,
+      "grad_norm": 11.492095884662016,
+      "learning_rate": 1e-06,
+      "loss": 0.5521,
+      "step": 7881
+    },
+    {
+      "epoch": 1.507218663352137,
+      "grad_norm": 4.572749083394747,
+      "learning_rate": 1e-06,
+      "loss": 0.3605,
+      "step": 7882
+    },
+    {
+      "epoch": 1.5074098862223924,
+      "grad_norm": 3.3232140642176167,
+      "learning_rate": 1e-06,
+      "loss": 0.1722,
+      "step": 7883
+    },
+    {
+      "epoch": 1.5076011090926476,
+      "grad_norm": 2.7578227877762695,
+      "learning_rate": 1e-06,
+      "loss": 0.116,
+      "step": 7884
+    },
+    {
+      "epoch": 1.5077923319629027,
+      "grad_norm": 6.777809782247945,
+      "learning_rate": 1e-06,
+      "loss": 0.4464,
+      "step": 7885
+    },
+    {
+      "epoch": 1.507983554833158,
+      "grad_norm": 2.241146683447131,
+      "learning_rate": 1e-06,
+      "loss": 0.107,
+      "step": 7886
+    },
+    {
+      "epoch": 1.5081747777034133,
+      "grad_norm": 1.5218481135881823,
+      "learning_rate": 1e-06,
+      "loss": 0.0807,
+      "step": 7887
+    },
+    {
+      "epoch": 1.5083660005736688,
+      "grad_norm": 3.349652793313718,
+      "learning_rate": 1e-06,
+      "loss": 0.1733,
+      "step": 7888
+    },
+    {
+      "epoch": 1.508557223443924,
+      "grad_norm": 1.84365631124103,
+      "learning_rate": 1e-06,
+      "loss": 0.0379,
+      "step": 7889
+    },
+    {
+      "epoch": 1.5087484463141791,
+      "grad_norm": 3.4872143043891537,
+      "learning_rate": 1e-06,
+      "loss": 0.106,
+      "step": 7890
+    },
+    {
+      "epoch": 1.5089396691844343,
+      "grad_norm": 4.127899769046712,
+      "learning_rate": 1e-06,
+      "loss": 0.0743,
+      "step": 7891
+    },
+    {
+      "epoch": 1.5091308920546898,
+      "grad_norm": 5.083952298028368,
+      "learning_rate": 1e-06,
+      "loss": 0.0954,
+      "step": 7892
+    },
+    {
+      "epoch": 1.5093221149249452,
+      "grad_norm": 7.746273498450269,
+      "learning_rate": 1e-06,
+      "loss": 0.5623,
+      "step": 7893
+    },
+    {
+      "epoch": 1.5095133377952004,
+      "grad_norm": 5.598501222861598,
+      "learning_rate": 1e-06,
+      "loss": 0.3889,
+      "step": 7894
+    },
+    {
+      "epoch": 1.5097045606654556,
+      "grad_norm": 3.773194994350685,
+      "learning_rate": 1e-06,
+      "loss": 0.2391,
+      "step": 7895
+    },
+    {
+      "epoch": 1.5098957835357107,
+      "grad_norm": 2.6859389806080944,
+      "learning_rate": 1e-06,
+      "loss": 0.1767,
+      "step": 7896
+    },
+    {
+      "epoch": 1.5100870064059662,
+      "grad_norm": 4.092521890944072,
+      "learning_rate": 1e-06,
+      "loss": 0.1317,
+      "step": 7897
+    },
+    {
+      "epoch": 1.5102782292762216,
+      "grad_norm": 4.843033276106676,
+      "learning_rate": 1e-06,
+      "loss": 0.0528,
+      "step": 7898
+    },
+    {
+      "epoch": 1.5104694521464768,
+      "grad_norm": 3.572537920185328,
+      "learning_rate": 1e-06,
+      "loss": 0.0736,
+      "step": 7899
+    },
+    {
+      "epoch": 1.510660675016732,
+      "grad_norm": 2.8462325777199955,
+      "learning_rate": 1e-06,
+      "loss": 0.0724,
+      "step": 7900
+    },
+    {
+      "epoch": 1.5108518978869871,
+      "grad_norm": 3.807614682850548,
+      "learning_rate": 1e-06,
+      "loss": 0.1966,
+      "step": 7901
+    },
+    {
+      "epoch": 1.5110431207572426,
+      "grad_norm": 3.7498063355346916,
+      "learning_rate": 1e-06,
+      "loss": 0.178,
+      "step": 7902
+    },
+    {
+      "epoch": 1.511234343627498,
+      "grad_norm": 1.1149582930321253,
+      "learning_rate": 1e-06,
+      "loss": 0.012,
+      "step": 7903
+    },
+    {
+      "epoch": 1.5114255664977532,
+      "grad_norm": 3.0578225675094552,
+      "learning_rate": 1e-06,
+      "loss": 0.0416,
+      "step": 7904
+    },
+    {
+      "epoch": 1.5116167893680084,
+      "grad_norm": 6.595579065404933,
+      "learning_rate": 1e-06,
+      "loss": 0.1577,
+      "step": 7905
+    },
+    {
+      "epoch": 1.5118080122382636,
+      "grad_norm": 8.879894787276788,
+      "learning_rate": 1e-06,
+      "loss": 0.3397,
+      "step": 7906
+    },
+    {
+      "epoch": 1.511999235108519,
+      "grad_norm": 4.017732653430101,
+      "learning_rate": 1e-06,
+      "loss": 0.2474,
+      "step": 7907
+    },
+    {
+      "epoch": 1.5121904579787744,
+      "grad_norm": 3.9442122634346544,
+      "learning_rate": 1e-06,
+      "loss": 0.1713,
+      "step": 7908
+    },
+    {
+      "epoch": 1.5123816808490296,
+      "grad_norm": 2.6974272020078383,
+      "learning_rate": 1e-06,
+      "loss": 0.1206,
+      "step": 7909
+    },
+    {
+      "epoch": 1.5125729037192848,
+      "grad_norm": 3.888570833066989,
+      "learning_rate": 1e-06,
+      "loss": 0.2857,
+      "step": 7910
+    },
+    {
+      "epoch": 1.51276412658954,
+      "grad_norm": 2.6568772416878206,
+      "learning_rate": 1e-06,
+      "loss": 0.1523,
+      "step": 7911
+    },
+    {
+      "epoch": 1.5129553494597954,
+      "grad_norm": 2.4490151936755638,
+      "learning_rate": 1e-06,
+      "loss": 0.0686,
+      "step": 7912
+    },
+    {
+      "epoch": 1.5131465723300508,
+      "grad_norm": 1.2944040773552663,
+      "learning_rate": 1e-06,
+      "loss": 0.0182,
+      "step": 7913
+    },
+    {
+      "epoch": 1.513337795200306,
+      "grad_norm": 2.0132929596210696,
+      "learning_rate": 1e-06,
+      "loss": 0.0243,
+      "step": 7914
+    },
+    {
+      "epoch": 1.5135290180705612,
+      "grad_norm": 2.9848320895561677,
+      "learning_rate": 1e-06,
+      "loss": 0.0604,
+      "step": 7915
+    },
+    {
+      "epoch": 1.5137202409408164,
+      "grad_norm": 3.041064546812655,
+      "learning_rate": 1e-06,
+      "loss": 0.0417,
+      "step": 7916
+    },
+    {
+      "epoch": 1.5139114638110718,
+      "grad_norm": 7.616187584610467,
+      "learning_rate": 1e-06,
+      "loss": 0.158,
+      "step": 7917
+    },
+    {
+      "epoch": 1.5141026866813272,
+      "grad_norm": 8.22437027488774,
+      "learning_rate": 1e-06,
+      "loss": 0.3264,
+      "step": 7918
+    },
+    {
+      "epoch": 1.5142939095515824,
+      "grad_norm": 6.047950826354167,
+      "learning_rate": 1e-06,
+      "loss": 0.599,
+      "step": 7919
+    },
+    {
+      "epoch": 1.5144851324218376,
+      "grad_norm": 4.445130356206327,
+      "learning_rate": 1e-06,
+      "loss": 0.5454,
+      "step": 7920
+    },
+    {
+      "epoch": 1.5146763552920928,
+      "grad_norm": 4.23879357682566,
+      "learning_rate": 1e-06,
+      "loss": 0.3094,
+      "step": 7921
+    },
+    {
+      "epoch": 1.5148675781623482,
+      "grad_norm": 2.7075774287232743,
+      "learning_rate": 1e-06,
+      "loss": 0.1958,
+      "step": 7922
+    },
+    {
+      "epoch": 1.5150588010326036,
+      "grad_norm": 2.5595688221332935,
+      "learning_rate": 1e-06,
+      "loss": 0.1199,
+      "step": 7923
+    },
+    {
+      "epoch": 1.5152500239028588,
+      "grad_norm": 4.938428187949006,
+      "learning_rate": 1e-06,
+      "loss": 0.4539,
+      "step": 7924
+    },
+    {
+      "epoch": 1.515441246773114,
+      "grad_norm": 2.972113059664573,
+      "learning_rate": 1e-06,
+      "loss": 0.0986,
+      "step": 7925
+    },
+    {
+      "epoch": 1.5156324696433694,
+      "grad_norm": 2.4046926041041408,
+      "learning_rate": 1e-06,
+      "loss": 0.0342,
+      "step": 7926
+    },
+    {
+      "epoch": 1.5158236925136246,
+      "grad_norm": 7.355131379593704,
+      "learning_rate": 1e-06,
+      "loss": 0.1078,
+      "step": 7927
+    },
+    {
+      "epoch": 1.51601491538388,
+      "grad_norm": 4.096671190582666,
+      "learning_rate": 1e-06,
+      "loss": 0.0449,
+      "step": 7928
+    },
+    {
+      "epoch": 1.5162061382541352,
+      "grad_norm": 1.5738921098728138,
+      "learning_rate": 1e-06,
+      "loss": 0.0167,
+      "step": 7929
+    },
+    {
+      "epoch": 1.5163973611243904,
+      "grad_norm": 5.544454728913096,
+      "learning_rate": 1e-06,
+      "loss": 0.2379,
+      "step": 7930
+    },
+    {
+      "epoch": 1.5165885839946458,
+      "grad_norm": 3.397299670333606,
+      "learning_rate": 1e-06,
+      "loss": 0.1841,
+      "step": 7931
+    },
+    {
+      "epoch": 1.516779806864901,
+      "grad_norm": 4.531815907886554,
+      "learning_rate": 1e-06,
+      "loss": 0.4935,
+      "step": 7932
+    },
+    {
+      "epoch": 1.5169710297351564,
+      "grad_norm": 7.047564352282322,
+      "learning_rate": 1e-06,
+      "loss": 0.4103,
+      "step": 7933
+    },
+    {
+      "epoch": 1.5171622526054116,
+      "grad_norm": 3.7350670081566038,
+      "learning_rate": 1e-06,
+      "loss": 0.1336,
+      "step": 7934
+    },
+    {
+      "epoch": 1.5173534754756668,
+      "grad_norm": 2.0604759457208637,
+      "learning_rate": 1e-06,
+      "loss": 0.0619,
+      "step": 7935
+    },
+    {
+      "epoch": 1.5175446983459222,
+      "grad_norm": 3.3139427929615763,
+      "learning_rate": 1e-06,
+      "loss": 0.1913,
+      "step": 7936
+    },
+    {
+      "epoch": 1.5177359212161776,
+      "grad_norm": 3.2141880928837296,
+      "learning_rate": 1e-06,
+      "loss": 0.1807,
+      "step": 7937
+    },
+    {
+      "epoch": 1.5179271440864328,
+      "grad_norm": 4.800253408418579,
+      "learning_rate": 1e-06,
+      "loss": 0.2141,
+      "step": 7938
+    },
+    {
+      "epoch": 1.518118366956688,
+      "grad_norm": 3.315070971941812,
+      "learning_rate": 1e-06,
+      "loss": 0.07,
+      "step": 7939
+    },
+    {
+      "epoch": 1.5183095898269432,
+      "grad_norm": 4.955698975073342,
+      "learning_rate": 1e-06,
+      "loss": 0.2504,
+      "step": 7940
+    },
+    {
+      "epoch": 1.5185008126971986,
+      "grad_norm": 2.2707262159742196,
+      "learning_rate": 1e-06,
+      "loss": 0.0327,
+      "step": 7941
+    },
+    {
+      "epoch": 1.518692035567454,
+      "grad_norm": 3.1130862683511964,
+      "learning_rate": 1e-06,
+      "loss": 0.0339,
+      "step": 7942
+    },
+    {
+      "epoch": 1.5188832584377092,
+      "grad_norm": 12.614506052139067,
+      "learning_rate": 1e-06,
+      "loss": 0.2793,
+      "step": 7943
+    },
+    {
+      "epoch": 1.5190744813079644,
+      "grad_norm": 4.908937532103235,
+      "learning_rate": 1e-06,
+      "loss": 0.3026,
+      "step": 7944
+    },
+    {
+      "epoch": 1.5192657041782196,
+      "grad_norm": 7.142084352378119,
+      "learning_rate": 1e-06,
+      "loss": 0.3581,
+      "step": 7945
+    },
+    {
+      "epoch": 1.519456927048475,
+      "grad_norm": 2.7156582495582087,
+      "learning_rate": 1e-06,
+      "loss": 0.1488,
+      "step": 7946
+    },
+    {
+      "epoch": 1.5196481499187304,
+      "grad_norm": 3.7833037040547324,
+      "learning_rate": 1e-06,
+      "loss": 0.4268,
+      "step": 7947
+    },
+    {
+      "epoch": 1.5198393727889856,
+      "grad_norm": 3.991902977645445,
+      "learning_rate": 1e-06,
+      "loss": 0.2265,
+      "step": 7948
+    },
+    {
+      "epoch": 1.5200305956592408,
+      "grad_norm": 3.950302262149228,
+      "learning_rate": 1e-06,
+      "loss": 0.2246,
+      "step": 7949
+    },
+    {
+      "epoch": 1.520221818529496,
+      "grad_norm": 2.1145523394738923,
+      "learning_rate": 1e-06,
+      "loss": 0.0656,
+      "step": 7950
+    },
+    {
+      "epoch": 1.5204130413997514,
+      "grad_norm": 1.68113603453823,
+      "learning_rate": 1e-06,
+      "loss": 0.0168,
+      "step": 7951
+    },
+    {
+      "epoch": 1.5206042642700068,
+      "grad_norm": 4.9256730215236395,
+      "learning_rate": 1e-06,
+      "loss": 0.0976,
+      "step": 7952
+    },
+    {
+      "epoch": 1.520795487140262,
+      "grad_norm": 4.834885885002468,
+      "learning_rate": 1e-06,
+      "loss": 0.1258,
+      "step": 7953
+    },
+    {
+      "epoch": 1.5209867100105172,
+      "grad_norm": 3.669358840150487,
+      "learning_rate": 1e-06,
+      "loss": 0.0429,
+      "step": 7954
+    },
+    {
+      "epoch": 1.5211779328807724,
+      "grad_norm": 5.434411410384691,
+      "learning_rate": 1e-06,
+      "loss": 0.4609,
+      "step": 7955
+    },
+    {
+      "epoch": 1.5213691557510278,
+      "grad_norm": 2.8537490613470484,
+      "learning_rate": 1e-06,
+      "loss": 0.1422,
+      "step": 7956
+    },
+    {
+      "epoch": 1.5215603786212832,
+      "grad_norm": 4.6017214580858345,
+      "learning_rate": 1e-06,
+      "loss": 0.1943,
+      "step": 7957
+    },
+    {
+      "epoch": 1.5217516014915384,
+      "grad_norm": 2.4241992257524783,
+      "learning_rate": 1e-06,
+      "loss": 0.1253,
+      "step": 7958
+    },
+    {
+      "epoch": 1.5219428243617936,
+      "grad_norm": 2.1327899191083715,
+      "learning_rate": 1e-06,
+      "loss": 0.0374,
+      "step": 7959
+    },
+    {
+      "epoch": 1.5221340472320488,
+      "grad_norm": 3.9693130934736045,
+      "learning_rate": 1e-06,
+      "loss": 0.2849,
+      "step": 7960
+    },
+    {
+      "epoch": 1.5223252701023042,
+      "grad_norm": 5.630133257520334,
+      "learning_rate": 1e-06,
+      "loss": 0.4493,
+      "step": 7961
+    },
+    {
+      "epoch": 1.5225164929725596,
+      "grad_norm": 2.584430543609406,
+      "learning_rate": 1e-06,
+      "loss": 0.0448,
+      "step": 7962
+    },
+    {
+      "epoch": 1.5227077158428148,
+      "grad_norm": 1.879566037129395,
+      "learning_rate": 1e-06,
+      "loss": 0.048,
+      "step": 7963
+    },
+    {
+      "epoch": 1.52289893871307,
+      "grad_norm": 2.7701262143021683,
+      "learning_rate": 1e-06,
+      "loss": 0.053,
+      "step": 7964
+    },
+    {
+      "epoch": 1.5230901615833252,
+      "grad_norm": 3.2324569838538952,
+      "learning_rate": 1e-06,
+      "loss": 0.0371,
+      "step": 7965
+    },
+    {
+      "epoch": 1.5232813844535806,
+      "grad_norm": 2.09074276990141,
+      "learning_rate": 1e-06,
+      "loss": 0.0617,
+      "step": 7966
+    },
+    {
+      "epoch": 1.523472607323836,
+      "grad_norm": 1.6884105486345304,
+      "learning_rate": 1e-06,
+      "loss": 0.0226,
+      "step": 7967
+    },
+    {
+      "epoch": 1.5236638301940912,
+      "grad_norm": 5.715121092360318,
+      "learning_rate": 1e-06,
+      "loss": 0.4134,
+      "step": 7968
+    },
+    {
+      "epoch": 1.5238550530643464,
+      "grad_norm": 4.53170500476853,
+      "learning_rate": 1e-06,
+      "loss": 0.3999,
+      "step": 7969
+    },
+    {
+      "epoch": 1.5240462759346016,
+      "grad_norm": 2.031382512391896,
+      "learning_rate": 1e-06,
+      "loss": 0.0693,
+      "step": 7970
+    },
+    {
+      "epoch": 1.524237498804857,
+      "grad_norm": 2.0610647120341423,
+      "learning_rate": 1e-06,
+      "loss": 0.083,
+      "step": 7971
+    },
+    {
+      "epoch": 1.5244287216751125,
+      "grad_norm": 4.477457805798951,
+      "learning_rate": 1e-06,
+      "loss": 0.1311,
+      "step": 7972
+    },
+    {
+      "epoch": 1.5246199445453676,
+      "grad_norm": 5.355971609779651,
+      "learning_rate": 1e-06,
+      "loss": 0.3101,
+      "step": 7973
+    },
+    {
+      "epoch": 1.5248111674156228,
+      "grad_norm": 1.7715244141483646,
+      "learning_rate": 1e-06,
+      "loss": 0.059,
+      "step": 7974
+    },
+    {
+      "epoch": 1.525002390285878,
+      "grad_norm": 3.2442598269321765,
+      "learning_rate": 1e-06,
+      "loss": 0.0719,
+      "step": 7975
+    },
+    {
+      "epoch": 1.5251936131561334,
+      "grad_norm": 4.267023884204205,
+      "learning_rate": 1e-06,
+      "loss": 0.0778,
+      "step": 7976
+    },
+    {
+      "epoch": 1.5253848360263889,
+      "grad_norm": 4.076745980323761,
+      "learning_rate": 1e-06,
+      "loss": 0.1121,
+      "step": 7977
+    },
+    {
+      "epoch": 1.525576058896644,
+      "grad_norm": 2.6970099812097685,
+      "learning_rate": 1e-06,
+      "loss": 0.0343,
+      "step": 7978
+    },
+    {
+      "epoch": 1.5257672817668992,
+      "grad_norm": 3.635733400263018,
+      "learning_rate": 1e-06,
+      "loss": 0.0494,
+      "step": 7979
+    },
+    {
+      "epoch": 1.5259585046371547,
+      "grad_norm": 3.946252450298797,
+      "learning_rate": 1e-06,
+      "loss": 0.1781,
+      "step": 7980
+    },
+    {
+      "epoch": 1.5261497275074098,
+      "grad_norm": 4.3029524669535135,
+      "learning_rate": 1e-06,
+      "loss": 0.4227,
+      "step": 7981
+    },
+    {
+      "epoch": 1.5263409503776653,
+      "grad_norm": 3.600831063030092,
+      "learning_rate": 1e-06,
+      "loss": 0.236,
+      "step": 7982
+    },
+    {
+      "epoch": 1.5265321732479205,
+      "grad_norm": 1.8768135201480194,
+      "learning_rate": 1e-06,
+      "loss": 0.0624,
+      "step": 7983
+    },
+    {
+      "epoch": 1.5267233961181756,
+      "grad_norm": 2.0349142036816232,
+      "learning_rate": 1e-06,
+      "loss": 0.0433,
+      "step": 7984
+    },
+    {
+      "epoch": 1.526914618988431,
+      "grad_norm": 4.468198302032284,
+      "learning_rate": 1e-06,
+      "loss": 0.2561,
+      "step": 7985
+    },
+    {
+      "epoch": 1.5271058418586863,
+      "grad_norm": 3.6346169861047466,
+      "learning_rate": 1e-06,
+      "loss": 0.1231,
+      "step": 7986
+    },
+    {
+      "epoch": 1.5272970647289417,
+      "grad_norm": 2.2823969426867,
+      "learning_rate": 1e-06,
+      "loss": 0.1325,
+      "step": 7987
+    },
+    {
+      "epoch": 1.5274882875991969,
+      "grad_norm": 2.1118898447310355,
+      "learning_rate": 1e-06,
+      "loss": 0.0726,
+      "step": 7988
+    },
+    {
+      "epoch": 1.527679510469452,
+      "grad_norm": 4.680657033200699,
+      "learning_rate": 1e-06,
+      "loss": 0.1743,
+      "step": 7989
+    },
+    {
+      "epoch": 1.5278707333397075,
+      "grad_norm": 2.2104979169068253,
+      "learning_rate": 1e-06,
+      "loss": 0.0427,
+      "step": 7990
+    },
+    {
+      "epoch": 1.5280619562099627,
+      "grad_norm": 3.483120542776999,
+      "learning_rate": 1e-06,
+      "loss": 0.1175,
+      "step": 7991
+    },
+    {
+      "epoch": 1.528253179080218,
+      "grad_norm": 7.902203274363136,
+      "learning_rate": 1e-06,
+      "loss": 0.0897,
+      "step": 7992
+    },
+    {
+      "epoch": 1.5284444019504733,
+      "grad_norm": 8.224274261829372,
+      "learning_rate": 1e-06,
+      "loss": 0.2944,
+      "step": 7993
+    },
+    {
+      "epoch": 1.5286356248207285,
+      "grad_norm": 8.397377749271783,
+      "learning_rate": 1e-06,
+      "loss": 0.5248,
+      "step": 7994
+    },
+    {
+      "epoch": 1.5288268476909839,
+      "grad_norm": 4.236512267638045,
+      "learning_rate": 1e-06,
+      "loss": 0.1824,
+      "step": 7995
+    },
+    {
+      "epoch": 1.5290180705612393,
+      "grad_norm": 3.9224412972778597,
+      "learning_rate": 1e-06,
+      "loss": 0.2374,
+      "step": 7996
+    },
+    {
+      "epoch": 1.5292092934314945,
+      "grad_norm": 4.758445660033043,
+      "learning_rate": 1e-06,
+      "loss": 0.259,
+      "step": 7997
+    },
+    {
+      "epoch": 1.5294005163017497,
+      "grad_norm": 2.57733983886274,
+      "learning_rate": 1e-06,
+      "loss": 0.0841,
+      "step": 7998
+    },
+    {
+      "epoch": 1.5295917391720049,
+      "grad_norm": 3.761904829736931,
+      "learning_rate": 1e-06,
+      "loss": 0.1579,
+      "step": 7999
+    },
+    {
+      "epoch": 1.5297829620422603,
+      "grad_norm": 3.631708055190054,
+      "learning_rate": 1e-06,
+      "loss": 0.1669,
+      "step": 8000
+    },
+    {
+      "epoch": 1.5297829620422603,
+      "eval_runtime": 749.8887,
+      "eval_samples_per_second": 2.046,
+      "eval_steps_per_second": 0.512,
+      "step": 8000
+    },
+    {
+      "epoch": 1.5299741849125157,
+      "grad_norm": 3.5790557317154335,
+      "learning_rate": 1e-06,
+      "loss": 0.043,
+      "step": 8001
+    },
+    {
+      "epoch": 1.5301654077827709,
+      "grad_norm": 1.9204302063415892,
+      "learning_rate": 1e-06,
+      "loss": 0.0323,
+      "step": 8002
+    },
+    {
+      "epoch": 1.530356630653026,
+      "grad_norm": 2.2480857971693,
+      "learning_rate": 1e-06,
+      "loss": 0.0367,
+      "step": 8003
+    },
+    {
+      "epoch": 1.5305478535232813,
+      "grad_norm": 6.038279177852536,
+      "learning_rate": 1e-06,
+      "loss": 0.0917,
+      "step": 8004
+    },
+    {
+      "epoch": 1.5307390763935367,
+      "grad_norm": 5.097642593073181,
+      "learning_rate": 1e-06,
+      "loss": 0.1803,
+      "step": 8005
+    },
+    {
+      "epoch": 1.530930299263792,
+      "grad_norm": 7.729884468192365,
+      "learning_rate": 1e-06,
+      "loss": 0.4537,
+      "step": 8006
+    },
+    {
+      "epoch": 1.5311215221340473,
+      "grad_norm": 0.8703297228110656,
+      "learning_rate": 1e-06,
+      "loss": 0.0129,
+      "step": 8007
+    },
+    {
+      "epoch": 1.5313127450043025,
+      "grad_norm": 2.4376187417867583,
+      "learning_rate": 1e-06,
+      "loss": 0.0649,
+      "step": 8008
+    },
+    {
+      "epoch": 1.5315039678745577,
+      "grad_norm": 3.216768562271046,
+      "learning_rate": 1e-06,
+      "loss": 0.2719,
+      "step": 8009
+    },
+    {
+      "epoch": 1.531695190744813,
+      "grad_norm": 4.127221174151117,
+      "learning_rate": 1e-06,
+      "loss": 0.2849,
+      "step": 8010
+    },
+    {
+      "epoch": 1.5318864136150685,
+      "grad_norm": 2.8477562152914278,
+      "learning_rate": 1e-06,
+      "loss": 0.1079,
+      "step": 8011
+    },
+    {
+      "epoch": 1.5320776364853237,
+      "grad_norm": 3.1373580718139853,
+      "learning_rate": 1e-06,
+      "loss": 0.0938,
+      "step": 8012
+    },
+    {
+      "epoch": 1.5322688593555789,
+      "grad_norm": 3.123169477297596,
+      "learning_rate": 1e-06,
+      "loss": 0.091,
+      "step": 8013
+    },
+    {
+      "epoch": 1.532460082225834,
+      "grad_norm": 2.5517632316125964,
+      "learning_rate": 1e-06,
+      "loss": 0.0609,
+      "step": 8014
+    },
+    {
+      "epoch": 1.5326513050960895,
+      "grad_norm": 2.993537618155729,
+      "learning_rate": 1e-06,
+      "loss": 0.049,
+      "step": 8015
+    },
+    {
+      "epoch": 1.532842527966345,
+      "grad_norm": 4.880207899073931,
+      "learning_rate": 1e-06,
+      "loss": 0.0613,
+      "step": 8016
+    },
+    {
+      "epoch": 1.5330337508366,
+      "grad_norm": 5.3107111892449295,
+      "learning_rate": 1e-06,
+      "loss": 0.043,
+      "step": 8017
+    },
+    {
+      "epoch": 1.5332249737068553,
+      "grad_norm": 3.668648328785785,
+      "learning_rate": 1e-06,
+      "loss": 0.1781,
+      "step": 8018
+    },
+    {
+      "epoch": 1.5334161965771105,
+      "grad_norm": 4.276755765451113,
+      "learning_rate": 1e-06,
+      "loss": 0.3121,
+      "step": 8019
+    },
+    {
+      "epoch": 1.533607419447366,
+      "grad_norm": 3.4323958910263213,
+      "learning_rate": 1e-06,
+      "loss": 0.2573,
+      "step": 8020
+    },
+    {
+      "epoch": 1.5337986423176213,
+      "grad_norm": 2.110444991388315,
+      "learning_rate": 1e-06,
+      "loss": 0.2825,
+      "step": 8021
+    },
+    {
+      "epoch": 1.5339898651878765,
+      "grad_norm": 3.76872258975055,
+      "learning_rate": 1e-06,
+      "loss": 0.3426,
+      "step": 8022
+    },
+    {
+      "epoch": 1.5341810880581317,
+      "grad_norm": 2.7872397399653286,
+      "learning_rate": 1e-06,
+      "loss": 0.1324,
+      "step": 8023
+    },
+    {
+      "epoch": 1.5343723109283869,
+      "grad_norm": 3.4166712101851857,
+      "learning_rate": 1e-06,
+      "loss": 0.2751,
+      "step": 8024
+    },
+    {
+      "epoch": 1.5345635337986423,
+      "grad_norm": 3.1148169354427235,
+      "learning_rate": 1e-06,
+      "loss": 0.0489,
+      "step": 8025
+    },
+    {
+      "epoch": 1.5347547566688977,
+      "grad_norm": 2.552252959757959,
+      "learning_rate": 1e-06,
+      "loss": 0.0777,
+      "step": 8026
+    },
+    {
+      "epoch": 1.534945979539153,
+      "grad_norm": 2.5934690759954186,
+      "learning_rate": 1e-06,
+      "loss": 0.0326,
+      "step": 8027
+    },
+    {
+      "epoch": 1.535137202409408,
+      "grad_norm": 4.786198268241198,
+      "learning_rate": 1e-06,
+      "loss": 0.067,
+      "step": 8028
+    },
+    {
+      "epoch": 1.5353284252796633,
+      "grad_norm": 3.9632813029380416,
+      "learning_rate": 1e-06,
+      "loss": 0.0653,
+      "step": 8029
+    },
+    {
+      "epoch": 1.5355196481499187,
+      "grad_norm": 7.586274160188578,
+      "learning_rate": 1e-06,
+      "loss": 0.1586,
+      "step": 8030
+    },
+    {
+      "epoch": 1.5357108710201741,
+      "grad_norm": 4.165120689960673,
+      "learning_rate": 1e-06,
+      "loss": 0.384,
+      "step": 8031
+    },
+    {
+      "epoch": 1.5359020938904293,
+      "grad_norm": 3.8166000332630885,
+      "learning_rate": 1e-06,
+      "loss": 0.1972,
+      "step": 8032
+    },
+    {
+      "epoch": 1.5360933167606845,
+      "grad_norm": 3.3858603475289333,
+      "learning_rate": 1e-06,
+      "loss": 0.0763,
+      "step": 8033
+    },
+    {
+      "epoch": 1.5362845396309397,
+      "grad_norm": 3.9636073397368365,
+      "learning_rate": 1e-06,
+      "loss": 0.0645,
+      "step": 8034
+    },
+    {
+      "epoch": 1.536475762501195,
+      "grad_norm": 2.3004734630053223,
+      "learning_rate": 1e-06,
+      "loss": 0.0696,
+      "step": 8035
+    },
+    {
+      "epoch": 1.5366669853714505,
+      "grad_norm": 2.530985971503113,
+      "learning_rate": 1e-06,
+      "loss": 0.0638,
+      "step": 8036
+    },
+    {
+      "epoch": 1.5368582082417057,
+      "grad_norm": 3.24588808952898,
+      "learning_rate": 1e-06,
+      "loss": 0.0691,
+      "step": 8037
+    },
+    {
+      "epoch": 1.537049431111961,
+      "grad_norm": 2.4930108602679737,
+      "learning_rate": 1e-06,
+      "loss": 0.0655,
+      "step": 8038
+    },
+    {
+      "epoch": 1.5372406539822163,
+      "grad_norm": 5.97025907123682,
+      "learning_rate": 1e-06,
+      "loss": 0.0817,
+      "step": 8039
+    },
+    {
+      "epoch": 1.5374318768524715,
+      "grad_norm": 2.5426964679979287,
+      "learning_rate": 1e-06,
+      "loss": 0.0544,
+      "step": 8040
+    },
+    {
+      "epoch": 1.537623099722727,
+      "grad_norm": 2.5840188829304847,
+      "learning_rate": 1e-06,
+      "loss": 0.0817,
+      "step": 8041
+    },
+    {
+      "epoch": 1.5378143225929821,
+      "grad_norm": 2.645777443923785,
+      "learning_rate": 1e-06,
+      "loss": 0.0255,
+      "step": 8042
+    },
+    {
+      "epoch": 1.5380055454632373,
+      "grad_norm": 6.31777552364269,
+      "learning_rate": 1e-06,
+      "loss": 0.2438,
+      "step": 8043
+    },
+    {
+      "epoch": 1.5381967683334927,
+      "grad_norm": 5.507335799129205,
+      "learning_rate": 1e-06,
+      "loss": 0.3629,
+      "step": 8044
+    },
+    {
+      "epoch": 1.538387991203748,
+      "grad_norm": 2.947787674057561,
+      "learning_rate": 1e-06,
+      "loss": 0.2071,
+      "step": 8045
+    },
+    {
+      "epoch": 1.5385792140740033,
+      "grad_norm": 3.6109305328473598,
+      "learning_rate": 1e-06,
+      "loss": 0.1156,
+      "step": 8046
+    },
+    {
+      "epoch": 1.5387704369442585,
+      "grad_norm": 3.020757862770653,
+      "learning_rate": 1e-06,
+      "loss": 0.2012,
+      "step": 8047
+    },
+    {
+      "epoch": 1.5389616598145137,
+      "grad_norm": 3.495405587605363,
+      "learning_rate": 1e-06,
+      "loss": 0.1107,
+      "step": 8048
+    },
+    {
+      "epoch": 1.5391528826847691,
+      "grad_norm": 4.798560626066034,
+      "learning_rate": 1e-06,
+      "loss": 0.158,
+      "step": 8049
+    },
+    {
+      "epoch": 1.5393441055550245,
+      "grad_norm": 4.131208891094164,
+      "learning_rate": 1e-06,
+      "loss": 0.1284,
+      "step": 8050
+    },
+    {
+      "epoch": 1.5395353284252797,
+      "grad_norm": 3.1826998900922256,
+      "learning_rate": 1e-06,
+      "loss": 0.0626,
+      "step": 8051
+    },
+    {
+      "epoch": 1.539726551295535,
+      "grad_norm": 3.174631959504614,
+      "learning_rate": 1e-06,
+      "loss": 0.0621,
+      "step": 8052
+    },
+    {
+      "epoch": 1.5399177741657901,
+      "grad_norm": 4.1912742800170015,
+      "learning_rate": 1e-06,
+      "loss": 0.1492,
+      "step": 8053
+    },
+    {
+      "epoch": 1.5401089970360455,
+      "grad_norm": 1.688304214992141,
+      "learning_rate": 1e-06,
+      "loss": 0.0213,
+      "step": 8054
+    },
+    {
+      "epoch": 1.540300219906301,
+      "grad_norm": 6.480973787960755,
+      "learning_rate": 1e-06,
+      "loss": 0.1127,
+      "step": 8055
+    },
+    {
+      "epoch": 1.5404914427765561,
+      "grad_norm": 3.645372169029168,
+      "learning_rate": 1e-06,
+      "loss": 0.2461,
+      "step": 8056
+    },
+    {
+      "epoch": 1.5406826656468113,
+      "grad_norm": 4.232282170629243,
+      "learning_rate": 1e-06,
+      "loss": 0.5681,
+      "step": 8057
+    },
+    {
+      "epoch": 1.5408738885170665,
+      "grad_norm": 4.0245084948199485,
+      "learning_rate": 1e-06,
+      "loss": 0.2151,
+      "step": 8058
+    },
+    {
+      "epoch": 1.541065111387322,
+      "grad_norm": 2.7902177400292048,
+      "learning_rate": 1e-06,
+      "loss": 0.0889,
+      "step": 8059
+    },
+    {
+      "epoch": 1.5412563342575774,
+      "grad_norm": 5.873649279568452,
+      "learning_rate": 1e-06,
+      "loss": 0.1827,
+      "step": 8060
+    },
+    {
+      "epoch": 1.5414475571278325,
+      "grad_norm": 2.6473975376261505,
+      "learning_rate": 1e-06,
+      "loss": 0.1162,
+      "step": 8061
+    },
+    {
+      "epoch": 1.5416387799980877,
+      "grad_norm": 3.6333233217807224,
+      "learning_rate": 1e-06,
+      "loss": 0.2437,
+      "step": 8062
+    },
+    {
+      "epoch": 1.541830002868343,
+      "grad_norm": 3.2428142631197323,
+      "learning_rate": 1e-06,
+      "loss": 0.0409,
+      "step": 8063
+    },
+    {
+      "epoch": 1.5420212257385983,
+      "grad_norm": 1.3473631166846634,
+      "learning_rate": 1e-06,
+      "loss": 0.0418,
+      "step": 8064
+    },
+    {
+      "epoch": 1.5422124486088538,
+      "grad_norm": 3.1430023457065994,
+      "learning_rate": 1e-06,
+      "loss": 0.0469,
+      "step": 8065
+    },
+    {
+      "epoch": 1.542403671479109,
+      "grad_norm": 3.760611588066014,
+      "learning_rate": 1e-06,
+      "loss": 0.0598,
+      "step": 8066
+    },
+    {
+      "epoch": 1.5425948943493641,
+      "grad_norm": 3.370548562248233,
+      "learning_rate": 1e-06,
+      "loss": 0.0762,
+      "step": 8067
+    },
+    {
+      "epoch": 1.5427861172196193,
+      "grad_norm": 5.485735515286602,
+      "learning_rate": 1e-06,
+      "loss": 0.2091,
+      "step": 8068
+    },
+    {
+      "epoch": 1.5429773400898747,
+      "grad_norm": 6.880970807002902,
+      "learning_rate": 1e-06,
+      "loss": 0.887,
+      "step": 8069
+    },
+    {
+      "epoch": 1.5431685629601302,
+      "grad_norm": 2.830141813142076,
+      "learning_rate": 1e-06,
+      "loss": 0.1102,
+      "step": 8070
+    },
+    {
+      "epoch": 1.5433597858303854,
+      "grad_norm": 4.423322353124226,
+      "learning_rate": 1e-06,
+      "loss": 0.1992,
+      "step": 8071
+    },
+    {
+      "epoch": 1.5435510087006405,
+      "grad_norm": 2.343987414733118,
+      "learning_rate": 1e-06,
+      "loss": 0.0569,
+      "step": 8072
+    },
+    {
+      "epoch": 1.5437422315708957,
+      "grad_norm": 2.4582255603852143,
+      "learning_rate": 1e-06,
+      "loss": 0.0876,
+      "step": 8073
+    },
+    {
+      "epoch": 1.5439334544411512,
+      "grad_norm": 4.01496757142687,
+      "learning_rate": 1e-06,
+      "loss": 0.1731,
+      "step": 8074
+    },
+    {
+      "epoch": 1.5441246773114066,
+      "grad_norm": 3.6221876086848077,
+      "learning_rate": 1e-06,
+      "loss": 0.0654,
+      "step": 8075
+    },
+    {
+      "epoch": 1.5443159001816618,
+      "grad_norm": 2.7184181449316,
+      "learning_rate": 1e-06,
+      "loss": 0.0976,
+      "step": 8076
+    },
+    {
+      "epoch": 1.544507123051917,
+      "grad_norm": 2.669529024368265,
+      "learning_rate": 1e-06,
+      "loss": 0.0453,
+      "step": 8077
+    },
+    {
+      "epoch": 1.5446983459221721,
+      "grad_norm": 9.139049009436102,
+      "learning_rate": 1e-06,
+      "loss": 0.072,
+      "step": 8078
+    },
+    {
+      "epoch": 1.5448895687924276,
+      "grad_norm": 4.051360601123641,
+      "learning_rate": 1e-06,
+      "loss": 0.0888,
+      "step": 8079
+    },
+    {
+      "epoch": 1.545080791662683,
+      "grad_norm": 4.590639891058217,
+      "learning_rate": 1e-06,
+      "loss": 0.0914,
+      "step": 8080
+    },
+    {
+      "epoch": 1.5452720145329382,
+      "grad_norm": 3.4693350083828203,
+      "learning_rate": 1e-06,
+      "loss": 0.2594,
+      "step": 8081
+    },
+    {
+      "epoch": 1.5454632374031934,
+      "grad_norm": 5.804716744227309,
+      "learning_rate": 1e-06,
+      "loss": 0.4247,
+      "step": 8082
+    },
+    {
+      "epoch": 1.5456544602734485,
+      "grad_norm": 2.8369802864473552,
+      "learning_rate": 1e-06,
+      "loss": 0.1067,
+      "step": 8083
+    },
+    {
+      "epoch": 1.545845683143704,
+      "grad_norm": 4.730363713932457,
+      "learning_rate": 1e-06,
+      "loss": 0.2616,
+      "step": 8084
+    },
+    {
+      "epoch": 1.5460369060139594,
+      "grad_norm": 4.054972085539179,
+      "learning_rate": 1e-06,
+      "loss": 0.1935,
+      "step": 8085
+    },
+    {
+      "epoch": 1.5462281288842146,
+      "grad_norm": 1.9271496598110418,
+      "learning_rate": 1e-06,
+      "loss": 0.0669,
+      "step": 8086
+    },
+    {
+      "epoch": 1.5464193517544698,
+      "grad_norm": 2.999942620046553,
+      "learning_rate": 1e-06,
+      "loss": 0.2375,
+      "step": 8087
+    },
+    {
+      "epoch": 1.546610574624725,
+      "grad_norm": 3.208207577969712,
+      "learning_rate": 1e-06,
+      "loss": 0.2085,
+      "step": 8088
+    },
+    {
+      "epoch": 1.5468017974949804,
+      "grad_norm": 3.963551398059396,
+      "learning_rate": 1e-06,
+      "loss": 0.1622,
+      "step": 8089
+    },
+    {
+      "epoch": 1.5469930203652358,
+      "grad_norm": 3.7953363762548555,
+      "learning_rate": 1e-06,
+      "loss": 0.0508,
+      "step": 8090
+    },
+    {
+      "epoch": 1.547184243235491,
+      "grad_norm": 2.4489853061719646,
+      "learning_rate": 1e-06,
+      "loss": 0.1283,
+      "step": 8091
+    },
+    {
+      "epoch": 1.5473754661057462,
+      "grad_norm": 2.836083946965576,
+      "learning_rate": 1e-06,
+      "loss": 0.0294,
+      "step": 8092
+    },
+    {
+      "epoch": 1.5475666889760016,
+      "grad_norm": 5.615340182427053,
+      "learning_rate": 1e-06,
+      "loss": 0.2178,
+      "step": 8093
+    },
+    {
+      "epoch": 1.5477579118462568,
+      "grad_norm": 3.5077442595846997,
+      "learning_rate": 1e-06,
+      "loss": 0.202,
+      "step": 8094
+    },
+    {
+      "epoch": 1.5479491347165122,
+      "grad_norm": 4.984422067641209,
+      "learning_rate": 1e-06,
+      "loss": 0.5072,
+      "step": 8095
+    },
+    {
+      "epoch": 1.5481403575867674,
+      "grad_norm": 4.034404378085207,
+      "learning_rate": 1e-06,
+      "loss": 0.1082,
+      "step": 8096
+    },
+    {
+      "epoch": 1.5483315804570226,
+      "grad_norm": 4.592126468591652,
+      "learning_rate": 1e-06,
+      "loss": 0.282,
+      "step": 8097
+    },
+    {
+      "epoch": 1.548522803327278,
+      "grad_norm": 2.6975484667958964,
+      "learning_rate": 1e-06,
+      "loss": 0.0845,
+      "step": 8098
+    },
+    {
+      "epoch": 1.5487140261975332,
+      "grad_norm": 3.0052776009142863,
+      "learning_rate": 1e-06,
+      "loss": 0.0501,
+      "step": 8099
+    },
+    {
+      "epoch": 1.5489052490677886,
+      "grad_norm": 2.8237168326420488,
+      "learning_rate": 1e-06,
+      "loss": 0.0705,
+      "step": 8100
+    },
+    {
+      "epoch": 1.5490964719380438,
+      "grad_norm": 3.588854565241409,
+      "learning_rate": 1e-06,
+      "loss": 0.1323,
+      "step": 8101
+    },
+    {
+      "epoch": 1.549287694808299,
+      "grad_norm": 3.413862907808119,
+      "learning_rate": 1e-06,
+      "loss": 0.0803,
+      "step": 8102
+    },
+    {
+      "epoch": 1.5494789176785544,
+      "grad_norm": 1.6677127257493183,
+      "learning_rate": 1e-06,
+      "loss": 0.0279,
+      "step": 8103
+    },
+    {
+      "epoch": 1.5496701405488096,
+      "grad_norm": 3.735526829017243,
+      "learning_rate": 1e-06,
+      "loss": 0.0493,
+      "step": 8104
+    },
+    {
+      "epoch": 1.549861363419065,
+      "grad_norm": 7.111572462586949,
+      "learning_rate": 1e-06,
+      "loss": 0.0619,
+      "step": 8105
+    },
+    {
+      "epoch": 1.5500525862893202,
+      "grad_norm": 3.507589694213344,
+      "learning_rate": 1e-06,
+      "loss": 0.3299,
+      "step": 8106
+    },
+    {
+      "epoch": 1.5502438091595754,
+      "grad_norm": 4.167013586224936,
+      "learning_rate": 1e-06,
+      "loss": 0.2345,
+      "step": 8107
+    },
+    {
+      "epoch": 1.5504350320298308,
+      "grad_norm": 3.2679137463994135,
+      "learning_rate": 1e-06,
+      "loss": 0.2531,
+      "step": 8108
+    },
+    {
+      "epoch": 1.5506262549000862,
+      "grad_norm": 2.768308899748854,
+      "learning_rate": 1e-06,
+      "loss": 0.1341,
+      "step": 8109
+    },
+    {
+      "epoch": 1.5508174777703414,
+      "grad_norm": 3.4551708668304872,
+      "learning_rate": 1e-06,
+      "loss": 0.1094,
+      "step": 8110
+    },
+    {
+      "epoch": 1.5510087006405966,
+      "grad_norm": 3.461688027044325,
+      "learning_rate": 1e-06,
+      "loss": 0.0869,
+      "step": 8111
+    },
+    {
+      "epoch": 1.5511999235108518,
+      "grad_norm": 3.7325865481154437,
+      "learning_rate": 1e-06,
+      "loss": 0.2519,
+      "step": 8112
+    },
+    {
+      "epoch": 1.5513911463811072,
+      "grad_norm": 2.5608328071127775,
+      "learning_rate": 1e-06,
+      "loss": 0.1213,
+      "step": 8113
+    },
+    {
+      "epoch": 1.5515823692513626,
+      "grad_norm": 3.35574144314483,
+      "learning_rate": 1e-06,
+      "loss": 0.1996,
+      "step": 8114
+    },
+    {
+      "epoch": 1.5517735921216178,
+      "grad_norm": 2.4824700879833244,
+      "learning_rate": 1e-06,
+      "loss": 0.075,
+      "step": 8115
+    },
+    {
+      "epoch": 1.551964814991873,
+      "grad_norm": 2.516311267574961,
+      "learning_rate": 1e-06,
+      "loss": 0.0488,
+      "step": 8116
+    },
+    {
+      "epoch": 1.5521560378621282,
+      "grad_norm": 6.62292851145708,
+      "learning_rate": 1e-06,
+      "loss": 0.0753,
+      "step": 8117
+    },
+    {
+      "epoch": 1.5523472607323836,
+      "grad_norm": 4.731842068472057,
+      "learning_rate": 1e-06,
+      "loss": 0.2558,
+      "step": 8118
+    },
+    {
+      "epoch": 1.552538483602639,
+      "grad_norm": 6.537407851435891,
+      "learning_rate": 1e-06,
+      "loss": 0.7056,
+      "step": 8119
+    },
+    {
+      "epoch": 1.5527297064728942,
+      "grad_norm": 5.584546578903551,
+      "learning_rate": 1e-06,
+      "loss": 0.2839,
+      "step": 8120
+    },
+    {
+      "epoch": 1.5529209293431494,
+      "grad_norm": 2.007162500992235,
+      "learning_rate": 1e-06,
+      "loss": 0.1796,
+      "step": 8121
+    },
+    {
+      "epoch": 1.5531121522134046,
+      "grad_norm": 3.4397923542162414,
+      "learning_rate": 1e-06,
+      "loss": 0.2464,
+      "step": 8122
+    },
+    {
+      "epoch": 1.55330337508366,
+      "grad_norm": 4.993755255590163,
+      "learning_rate": 1e-06,
+      "loss": 0.6054,
+      "step": 8123
+    },
+    {
+      "epoch": 1.5534945979539154,
+      "grad_norm": 4.805839388090399,
+      "learning_rate": 1e-06,
+      "loss": 0.1926,
+      "step": 8124
+    },
+    {
+      "epoch": 1.5536858208241706,
+      "grad_norm": 2.423818485262358,
+      "learning_rate": 1e-06,
+      "loss": 0.0355,
+      "step": 8125
+    },
+    {
+      "epoch": 1.5538770436944258,
+      "grad_norm": 5.426772134929139,
+      "learning_rate": 1e-06,
+      "loss": 0.1124,
+      "step": 8126
+    },
+    {
+      "epoch": 1.554068266564681,
+      "grad_norm": 2.6575605861758933,
+      "learning_rate": 1e-06,
+      "loss": 0.0294,
+      "step": 8127
+    },
+    {
+      "epoch": 1.5542594894349364,
+      "grad_norm": 3.0863080152371363,
+      "learning_rate": 1e-06,
+      "loss": 0.0629,
+      "step": 8128
+    },
+    {
+      "epoch": 1.5544507123051918,
+      "grad_norm": 3.106114196878913,
+      "learning_rate": 1e-06,
+      "loss": 0.0773,
+      "step": 8129
+    },
+    {
+      "epoch": 1.554641935175447,
+      "grad_norm": 4.528616936013623,
+      "learning_rate": 1e-06,
+      "loss": 0.0956,
+      "step": 8130
+    },
+    {
+      "epoch": 1.5548331580457022,
+      "grad_norm": 4.156490691823551,
+      "learning_rate": 1e-06,
+      "loss": 0.3703,
+      "step": 8131
+    },
+    {
+      "epoch": 1.5550243809159574,
+      "grad_norm": 5.84229019617548,
+      "learning_rate": 1e-06,
+      "loss": 0.3535,
+      "step": 8132
+    },
+    {
+      "epoch": 1.5552156037862128,
+      "grad_norm": 3.6718500664047067,
+      "learning_rate": 1e-06,
+      "loss": 0.1955,
+      "step": 8133
+    },
+    {
+      "epoch": 1.5554068266564682,
+      "grad_norm": 1.851531724633807,
+      "learning_rate": 1e-06,
+      "loss": 0.0624,
+      "step": 8134
+    },
+    {
+      "epoch": 1.5555980495267234,
+      "grad_norm": 3.056762770872175,
+      "learning_rate": 1e-06,
+      "loss": 0.2172,
+      "step": 8135
+    },
+    {
+      "epoch": 1.5557892723969786,
+      "grad_norm": 3.025158769208396,
+      "learning_rate": 1e-06,
+      "loss": 0.0932,
+      "step": 8136
+    },
+    {
+      "epoch": 1.5559804952672338,
+      "grad_norm": 2.1525481139154468,
+      "learning_rate": 1e-06,
+      "loss": 0.0689,
+      "step": 8137
+    },
+    {
+      "epoch": 1.5561717181374892,
+      "grad_norm": 4.336664582197513,
+      "learning_rate": 1e-06,
+      "loss": 0.0691,
+      "step": 8138
+    },
+    {
+      "epoch": 1.5563629410077446,
+      "grad_norm": 4.040717311669089,
+      "learning_rate": 1e-06,
+      "loss": 0.0826,
+      "step": 8139
+    },
+    {
+      "epoch": 1.5565541638779998,
+      "grad_norm": 4.504348137676055,
+      "learning_rate": 1e-06,
+      "loss": 0.1973,
+      "step": 8140
+    },
+    {
+      "epoch": 1.556745386748255,
+      "grad_norm": 2.848133607055704,
+      "learning_rate": 1e-06,
+      "loss": 0.0427,
+      "step": 8141
+    },
+    {
+      "epoch": 1.5569366096185102,
+      "grad_norm": 5.0034999042174375,
+      "learning_rate": 1e-06,
+      "loss": 0.0336,
+      "step": 8142
+    },
+    {
+      "epoch": 1.5571278324887656,
+      "grad_norm": 4.553234778175888,
+      "learning_rate": 1e-06,
+      "loss": 0.2631,
+      "step": 8143
+    },
+    {
+      "epoch": 1.557319055359021,
+      "grad_norm": 4.322645995463844,
+      "learning_rate": 1e-06,
+      "loss": 0.2206,
+      "step": 8144
+    },
+    {
+      "epoch": 1.5575102782292762,
+      "grad_norm": 3.3665174627447474,
+      "learning_rate": 1e-06,
+      "loss": 0.0826,
+      "step": 8145
+    },
+    {
+      "epoch": 1.5577015010995314,
+      "grad_norm": 4.983753321033957,
+      "learning_rate": 1e-06,
+      "loss": 0.235,
+      "step": 8146
+    },
+    {
+      "epoch": 1.5578927239697868,
+      "grad_norm": 4.804796667331959,
+      "learning_rate": 1e-06,
+      "loss": 0.2117,
+      "step": 8147
+    },
+    {
+      "epoch": 1.558083946840042,
+      "grad_norm": 3.6367521847868938,
+      "learning_rate": 1e-06,
+      "loss": 0.1202,
+      "step": 8148
+    },
+    {
+      "epoch": 1.5582751697102974,
+      "grad_norm": 3.686621092456574,
+      "learning_rate": 1e-06,
+      "loss": 0.1443,
+      "step": 8149
+    },
+    {
+      "epoch": 1.5584663925805526,
+      "grad_norm": 2.599733276624474,
+      "learning_rate": 1e-06,
+      "loss": 0.1132,
+      "step": 8150
+    },
+    {
+      "epoch": 1.5586576154508078,
+      "grad_norm": 1.88440810849788,
+      "learning_rate": 1e-06,
+      "loss": 0.0646,
+      "step": 8151
+    },
+    {
+      "epoch": 1.5588488383210632,
+      "grad_norm": 3.2021515646428007,
+      "learning_rate": 1e-06,
+      "loss": 0.0333,
+      "step": 8152
+    },
+    {
+      "epoch": 1.5590400611913184,
+      "grad_norm": 3.701408149203125,
+      "learning_rate": 1e-06,
+      "loss": 0.041,
+      "step": 8153
+    },
+    {
+      "epoch": 1.5592312840615739,
+      "grad_norm": 3.646126319375238,
+      "learning_rate": 1e-06,
+      "loss": 0.0726,
+      "step": 8154
+    },
+    {
+      "epoch": 1.559422506931829,
+      "grad_norm": 8.223264433416606,
+      "learning_rate": 1e-06,
+      "loss": 0.1186,
+      "step": 8155
+    },
+    {
+      "epoch": 1.5596137298020842,
+      "grad_norm": 4.539008292893833,
+      "learning_rate": 1e-06,
+      "loss": 0.2932,
+      "step": 8156
+    },
+    {
+      "epoch": 1.5598049526723397,
+      "grad_norm": 5.912125183225479,
+      "learning_rate": 1e-06,
+      "loss": 0.4431,
+      "step": 8157
+    },
+    {
+      "epoch": 1.5599961755425948,
+      "grad_norm": 2.8869349665833726,
+      "learning_rate": 1e-06,
+      "loss": 0.1009,
+      "step": 8158
+    },
+    {
+      "epoch": 1.5601873984128503,
+      "grad_norm": 2.4680647079974523,
+      "learning_rate": 1e-06,
+      "loss": 0.2227,
+      "step": 8159
+    },
+    {
+      "epoch": 1.5603786212831054,
+      "grad_norm": 5.560480329773925,
+      "learning_rate": 1e-06,
+      "loss": 0.3118,
+      "step": 8160
+    },
+    {
+      "epoch": 1.5605698441533606,
+      "grad_norm": 4.36941542462537,
+      "learning_rate": 1e-06,
+      "loss": 0.3053,
+      "step": 8161
+    },
+    {
+      "epoch": 1.560761067023616,
+      "grad_norm": 3.4678179889246037,
+      "learning_rate": 1e-06,
+      "loss": 0.2566,
+      "step": 8162
+    },
+    {
+      "epoch": 1.5609522898938715,
+      "grad_norm": 3.116415669199149,
+      "learning_rate": 1e-06,
+      "loss": 0.0875,
+      "step": 8163
+    },
+    {
+      "epoch": 1.5611435127641267,
+      "grad_norm": 2.4664908587032905,
+      "learning_rate": 1e-06,
+      "loss": 0.1143,
+      "step": 8164
+    },
+    {
+      "epoch": 1.5613347356343819,
+      "grad_norm": 2.826320683819882,
+      "learning_rate": 1e-06,
+      "loss": 0.186,
+      "step": 8165
+    },
+    {
+      "epoch": 1.561525958504637,
+      "grad_norm": 2.24781894936943,
+      "learning_rate": 1e-06,
+      "loss": 0.0763,
+      "step": 8166
+    },
+    {
+      "epoch": 1.5617171813748925,
+      "grad_norm": 6.304038240074708,
+      "learning_rate": 1e-06,
+      "loss": 0.0641,
+      "step": 8167
+    },
+    {
+      "epoch": 1.5619084042451479,
+      "grad_norm": 20.61820659308744,
+      "learning_rate": 1e-06,
+      "loss": 0.1941,
+      "step": 8168
+    },
+    {
+      "epoch": 1.562099627115403,
+      "grad_norm": 4.550526414538659,
+      "learning_rate": 1e-06,
+      "loss": 0.3567,
+      "step": 8169
+    },
+    {
+      "epoch": 1.5622908499856583,
+      "grad_norm": 2.7255856750772445,
+      "learning_rate": 1e-06,
+      "loss": 0.2034,
+      "step": 8170
+    },
+    {
+      "epoch": 1.5624820728559135,
+      "grad_norm": 4.5973168882295665,
+      "learning_rate": 1e-06,
+      "loss": 0.1993,
+      "step": 8171
+    },
+    {
+      "epoch": 1.5626732957261689,
+      "grad_norm": 1.7985431391548212,
+      "learning_rate": 1e-06,
+      "loss": 0.1702,
+      "step": 8172
+    },
+    {
+      "epoch": 1.5628645185964243,
+      "grad_norm": 3.361276913241126,
+      "learning_rate": 1e-06,
+      "loss": 0.1001,
+      "step": 8173
+    },
+    {
+      "epoch": 1.5630557414666795,
+      "grad_norm": 3.201637927721848,
+      "learning_rate": 1e-06,
+      "loss": 0.1241,
+      "step": 8174
+    },
+    {
+      "epoch": 1.5632469643369347,
+      "grad_norm": 5.4168405162593505,
+      "learning_rate": 1e-06,
+      "loss": 0.1456,
+      "step": 8175
+    },
+    {
+      "epoch": 1.5634381872071899,
+      "grad_norm": 2.700187454074234,
+      "learning_rate": 1e-06,
+      "loss": 0.0645,
+      "step": 8176
+    },
+    {
+      "epoch": 1.5636294100774453,
+      "grad_norm": 2.9878350616646983,
+      "learning_rate": 1e-06,
+      "loss": 0.0375,
+      "step": 8177
+    },
+    {
+      "epoch": 1.5638206329477007,
+      "grad_norm": 4.95215161503423,
+      "learning_rate": 1e-06,
+      "loss": 0.0666,
+      "step": 8178
+    },
+    {
+      "epoch": 1.5640118558179559,
+      "grad_norm": 3.7706925732946996,
+      "learning_rate": 1e-06,
+      "loss": 0.0707,
+      "step": 8179
+    },
+    {
+      "epoch": 1.564203078688211,
+      "grad_norm": 9.012018021011452,
+      "learning_rate": 1e-06,
+      "loss": 0.199,
+      "step": 8180
+    },
+    {
+      "epoch": 1.5643943015584663,
+      "grad_norm": 4.698549801721571,
+      "learning_rate": 1e-06,
+      "loss": 0.5418,
+      "step": 8181
+    },
+    {
+      "epoch": 1.5645855244287217,
+      "grad_norm": 4.115996968600952,
+      "learning_rate": 1e-06,
+      "loss": 0.3926,
+      "step": 8182
+    },
+    {
+      "epoch": 1.564776747298977,
+      "grad_norm": 4.570699978566454,
+      "learning_rate": 1e-06,
+      "loss": 0.0829,
+      "step": 8183
+    },
+    {
+      "epoch": 1.5649679701692323,
+      "grad_norm": 2.8775006904236657,
+      "learning_rate": 1e-06,
+      "loss": 0.1761,
+      "step": 8184
+    },
+    {
+      "epoch": 1.5651591930394875,
+      "grad_norm": 3.190488723101263,
+      "learning_rate": 1e-06,
+      "loss": 0.1963,
+      "step": 8185
+    },
+    {
+      "epoch": 1.5653504159097427,
+      "grad_norm": 2.996887181548659,
+      "learning_rate": 1e-06,
+      "loss": 0.0927,
+      "step": 8186
+    },
+    {
+      "epoch": 1.565541638779998,
+      "grad_norm": 3.379831317552316,
+      "learning_rate": 1e-06,
+      "loss": 0.1886,
+      "step": 8187
+    },
+    {
+      "epoch": 1.5657328616502535,
+      "grad_norm": 1.785385684705648,
+      "learning_rate": 1e-06,
+      "loss": 0.0662,
+      "step": 8188
+    },
+    {
+      "epoch": 1.5659240845205087,
+      "grad_norm": 3.4895695168900143,
+      "learning_rate": 1e-06,
+      "loss": 0.2392,
+      "step": 8189
+    },
+    {
+      "epoch": 1.5661153073907639,
+      "grad_norm": 1.2561174427146404,
+      "learning_rate": 1e-06,
+      "loss": 0.0105,
+      "step": 8190
+    },
+    {
+      "epoch": 1.566306530261019,
+      "grad_norm": 5.045582514271786,
+      "learning_rate": 1e-06,
+      "loss": 0.0883,
+      "step": 8191
+    },
+    {
+      "epoch": 1.5664977531312745,
+      "grad_norm": 4.281344447521166,
+      "learning_rate": 1e-06,
+      "loss": 0.1725,
+      "step": 8192
+    },
+    {
+      "epoch": 1.56668897600153,
+      "grad_norm": 5.383574809670916,
+      "learning_rate": 1e-06,
+      "loss": 0.1465,
+      "step": 8193
+    },
+    {
+      "epoch": 1.566880198871785,
+      "grad_norm": 2.4587415798025622,
+      "learning_rate": 1e-06,
+      "loss": 0.2547,
+      "step": 8194
+    },
+    {
+      "epoch": 1.5670714217420403,
+      "grad_norm": 5.0704573271159195,
+      "learning_rate": 1e-06,
+      "loss": 0.3276,
+      "step": 8195
+    },
+    {
+      "epoch": 1.5672626446122955,
+      "grad_norm": 2.0003676076651256,
+      "learning_rate": 1e-06,
+      "loss": 0.0779,
+      "step": 8196
+    },
+    {
+      "epoch": 1.5674538674825509,
+      "grad_norm": 3.4304671042326955,
+      "learning_rate": 1e-06,
+      "loss": 0.1231,
+      "step": 8197
+    },
+    {
+      "epoch": 1.5676450903528063,
+      "grad_norm": 2.0684695696271755,
+      "learning_rate": 1e-06,
+      "loss": 0.2373,
+      "step": 8198
+    },
+    {
+      "epoch": 1.5678363132230615,
+      "grad_norm": 3.1625775655650457,
+      "learning_rate": 1e-06,
+      "loss": 0.1837,
+      "step": 8199
+    },
+    {
+      "epoch": 1.5680275360933167,
+      "grad_norm": 4.548348204543108,
+      "learning_rate": 1e-06,
+      "loss": 0.2914,
+      "step": 8200
+    },
+    {
+      "epoch": 1.5682187589635719,
+      "grad_norm": 3.558328696135478,
+      "learning_rate": 1e-06,
+      "loss": 0.0931,
+      "step": 8201
+    },
+    {
+      "epoch": 1.5684099818338273,
+      "grad_norm": 4.568786261919515,
+      "learning_rate": 1e-06,
+      "loss": 0.2049,
+      "step": 8202
+    },
+    {
+      "epoch": 1.5686012047040827,
+      "grad_norm": 4.103621832350581,
+      "learning_rate": 1e-06,
+      "loss": 0.0803,
+      "step": 8203
+    },
+    {
+      "epoch": 1.568792427574338,
+      "grad_norm": 2.9217446038020145,
+      "learning_rate": 1e-06,
+      "loss": 0.0359,
+      "step": 8204
+    },
+    {
+      "epoch": 1.568983650444593,
+      "grad_norm": 5.857832479252071,
+      "learning_rate": 1e-06,
+      "loss": 0.0799,
+      "step": 8205
+    },
+    {
+      "epoch": 1.5691748733148485,
+      "grad_norm": 5.448959902286274,
+      "learning_rate": 1e-06,
+      "loss": 0.4985,
+      "step": 8206
+    },
+    {
+      "epoch": 1.5693660961851037,
+      "grad_norm": 5.200598300846507,
+      "learning_rate": 1e-06,
+      "loss": 0.4642,
+      "step": 8207
+    },
+    {
+      "epoch": 1.5695573190553591,
+      "grad_norm": 2.4427162501051636,
+      "learning_rate": 1e-06,
+      "loss": 0.0785,
+      "step": 8208
+    },
+    {
+      "epoch": 1.5697485419256143,
+      "grad_norm": 2.705153393300761,
+      "learning_rate": 1e-06,
+      "loss": 0.2053,
+      "step": 8209
+    },
+    {
+      "epoch": 1.5699397647958695,
+      "grad_norm": 4.392617421803434,
+      "learning_rate": 1e-06,
+      "loss": 0.3436,
+      "step": 8210
+    },
+    {
+      "epoch": 1.570130987666125,
+      "grad_norm": 2.8905586853668623,
+      "learning_rate": 1e-06,
+      "loss": 0.1145,
+      "step": 8211
+    },
+    {
+      "epoch": 1.57032221053638,
+      "grad_norm": 2.0223383329870517,
+      "learning_rate": 1e-06,
+      "loss": 0.0786,
+      "step": 8212
+    },
+    {
+      "epoch": 1.5705134334066355,
+      "grad_norm": 3.9230583199107425,
+      "learning_rate": 1e-06,
+      "loss": 0.1884,
+      "step": 8213
+    },
+    {
+      "epoch": 1.5707046562768907,
+      "grad_norm": 5.906123084263873,
+      "learning_rate": 1e-06,
+      "loss": 0.1426,
+      "step": 8214
+    },
+    {
+      "epoch": 1.570895879147146,
+      "grad_norm": 2.598825108560646,
+      "learning_rate": 1e-06,
+      "loss": 0.053,
+      "step": 8215
+    },
+    {
+      "epoch": 1.5710871020174013,
+      "grad_norm": 1.4682153580922783,
+      "learning_rate": 1e-06,
+      "loss": 0.025,
+      "step": 8216
+    },
+    {
+      "epoch": 1.5712783248876567,
+      "grad_norm": 5.471484388943921,
+      "learning_rate": 1e-06,
+      "loss": 0.1037,
+      "step": 8217
+    },
+    {
+      "epoch": 1.571469547757912,
+      "grad_norm": 4.977609281772456,
+      "learning_rate": 1e-06,
+      "loss": 0.1771,
+      "step": 8218
+    },
+    {
+      "epoch": 1.5716607706281671,
+      "grad_norm": 3.7276098031259055,
+      "learning_rate": 1e-06,
+      "loss": 0.2942,
+      "step": 8219
+    },
+    {
+      "epoch": 1.5718519934984223,
+      "grad_norm": 7.049876586699607,
+      "learning_rate": 1e-06,
+      "loss": 0.4577,
+      "step": 8220
+    },
+    {
+      "epoch": 1.5720432163686777,
+      "grad_norm": 2.36123947343151,
+      "learning_rate": 1e-06,
+      "loss": 0.0723,
+      "step": 8221
+    },
+    {
+      "epoch": 1.5722344392389331,
+      "grad_norm": 2.8016631840720723,
+      "learning_rate": 1e-06,
+      "loss": 0.2424,
+      "step": 8222
+    },
+    {
+      "epoch": 1.5724256621091883,
+      "grad_norm": 4.211554600611354,
+      "learning_rate": 1e-06,
+      "loss": 0.1917,
+      "step": 8223
+    },
+    {
+      "epoch": 1.5726168849794435,
+      "grad_norm": 3.7769166691167095,
+      "learning_rate": 1e-06,
+      "loss": 0.209,
+      "step": 8224
+    },
+    {
+      "epoch": 1.5728081078496987,
+      "grad_norm": 3.9205723761139337,
+      "learning_rate": 1e-06,
+      "loss": 0.1909,
+      "step": 8225
+    },
+    {
+      "epoch": 1.5729993307199541,
+      "grad_norm": 2.5473190575786764,
+      "learning_rate": 1e-06,
+      "loss": 0.0579,
+      "step": 8226
+    },
+    {
+      "epoch": 1.5731905535902095,
+      "grad_norm": 3.4252121908352033,
+      "learning_rate": 1e-06,
+      "loss": 0.2162,
+      "step": 8227
+    },
+    {
+      "epoch": 1.5733817764604647,
+      "grad_norm": 1.6485891408008928,
+      "learning_rate": 1e-06,
+      "loss": 0.0596,
+      "step": 8228
+    },
+    {
+      "epoch": 1.57357299933072,
+      "grad_norm": 3.2705980609095624,
+      "learning_rate": 1e-06,
+      "loss": 0.082,
+      "step": 8229
+    },
+    {
+      "epoch": 1.5737642222009751,
+      "grad_norm": 6.850817510452562,
+      "learning_rate": 1e-06,
+      "loss": 0.1843,
+      "step": 8230
+    },
+    {
+      "epoch": 1.5739554450712305,
+      "grad_norm": 4.359413939818245,
+      "learning_rate": 1e-06,
+      "loss": 0.3856,
+      "step": 8231
+    },
+    {
+      "epoch": 1.574146667941486,
+      "grad_norm": 3.601059090275425,
+      "learning_rate": 1e-06,
+      "loss": 0.2649,
+      "step": 8232
+    },
+    {
+      "epoch": 1.5743378908117411,
+      "grad_norm": 3.2432280192146745,
+      "learning_rate": 1e-06,
+      "loss": 0.1886,
+      "step": 8233
+    },
+    {
+      "epoch": 1.5745291136819963,
+      "grad_norm": 3.717961844902576,
+      "learning_rate": 1e-06,
+      "loss": 0.1869,
+      "step": 8234
+    },
+    {
+      "epoch": 1.5747203365522515,
+      "grad_norm": 3.7772705323501183,
+      "learning_rate": 1e-06,
+      "loss": 0.2525,
+      "step": 8235
+    },
+    {
+      "epoch": 1.574911559422507,
+      "grad_norm": 4.5149552531681145,
+      "learning_rate": 1e-06,
+      "loss": 0.3211,
+      "step": 8236
+    },
+    {
+      "epoch": 1.5751027822927623,
+      "grad_norm": 2.7333251681632302,
+      "learning_rate": 1e-06,
+      "loss": 0.1797,
+      "step": 8237
+    },
+    {
+      "epoch": 1.5752940051630175,
+      "grad_norm": 2.9090551149268604,
+      "learning_rate": 1e-06,
+      "loss": 0.1543,
+      "step": 8238
+    },
+    {
+      "epoch": 1.5754852280332727,
+      "grad_norm": 2.234072044481301,
+      "learning_rate": 1e-06,
+      "loss": 0.0649,
+      "step": 8239
+    },
+    {
+      "epoch": 1.575676450903528,
+      "grad_norm": 4.933330658748692,
+      "learning_rate": 1e-06,
+      "loss": 0.0861,
+      "step": 8240
+    },
+    {
+      "epoch": 1.5758676737737833,
+      "grad_norm": 1.808151487650266,
+      "learning_rate": 1e-06,
+      "loss": 0.0317,
+      "step": 8241
+    },
+    {
+      "epoch": 1.5760588966440388,
+      "grad_norm": 6.927480524669682,
+      "learning_rate": 1e-06,
+      "loss": 0.0614,
+      "step": 8242
+    },
+    {
+      "epoch": 1.576250119514294,
+      "grad_norm": 4.319352607255025,
+      "learning_rate": 1e-06,
+      "loss": 0.1147,
+      "step": 8243
+    },
+    {
+      "epoch": 1.5764413423845491,
+      "grad_norm": 4.307951809329337,
+      "learning_rate": 1e-06,
+      "loss": 0.3865,
+      "step": 8244
+    },
+    {
+      "epoch": 1.5766325652548043,
+      "grad_norm": 5.732894332106227,
+      "learning_rate": 1e-06,
+      "loss": 0.3478,
+      "step": 8245
+    },
+    {
+      "epoch": 1.5768237881250597,
+      "grad_norm": 3.6587402470820027,
+      "learning_rate": 1e-06,
+      "loss": 0.2658,
+      "step": 8246
+    },
+    {
+      "epoch": 1.5770150109953152,
+      "grad_norm": 5.849140149690607,
+      "learning_rate": 1e-06,
+      "loss": 0.2551,
+      "step": 8247
+    },
+    {
+      "epoch": 1.5772062338655704,
+      "grad_norm": 4.108173615379657,
+      "learning_rate": 1e-06,
+      "loss": 0.2436,
+      "step": 8248
+    },
+    {
+      "epoch": 1.5773974567358255,
+      "grad_norm": 2.922009959648438,
+      "learning_rate": 1e-06,
+      "loss": 0.2036,
+      "step": 8249
+    },
+    {
+      "epoch": 1.5775886796060807,
+      "grad_norm": 2.8740375400499225,
+      "learning_rate": 1e-06,
+      "loss": 0.0354,
+      "step": 8250
+    },
+    {
+      "epoch": 1.5777799024763361,
+      "grad_norm": 2.905048901512369,
+      "learning_rate": 1e-06,
+      "loss": 0.216,
+      "step": 8251
+    },
+    {
+      "epoch": 1.5779711253465916,
+      "grad_norm": 2.4246046864885393,
+      "learning_rate": 1e-06,
+      "loss": 0.0437,
+      "step": 8252
+    },
+    {
+      "epoch": 1.5781623482168468,
+      "grad_norm": 2.808408685988238,
+      "learning_rate": 1e-06,
+      "loss": 0.0722,
+      "step": 8253
+    },
+    {
+      "epoch": 1.578353571087102,
+      "grad_norm": 2.2267873550113575,
+      "learning_rate": 1e-06,
+      "loss": 0.0261,
+      "step": 8254
+    },
+    {
+      "epoch": 1.5785447939573571,
+      "grad_norm": 5.7278103789371855,
+      "learning_rate": 1e-06,
+      "loss": 0.1001,
+      "step": 8255
+    },
+    {
+      "epoch": 1.5787360168276126,
+      "grad_norm": 7.058695080467329,
+      "learning_rate": 1e-06,
+      "loss": 0.2852,
+      "step": 8256
+    },
+    {
+      "epoch": 1.578927239697868,
+      "grad_norm": 3.5644318546026317,
+      "learning_rate": 1e-06,
+      "loss": 0.3349,
+      "step": 8257
+    },
+    {
+      "epoch": 1.5791184625681232,
+      "grad_norm": 5.572349999635039,
+      "learning_rate": 1e-06,
+      "loss": 0.1758,
+      "step": 8258
+    },
+    {
+      "epoch": 1.5793096854383784,
+      "grad_norm": 2.8133274132845574,
+      "learning_rate": 1e-06,
+      "loss": 0.0787,
+      "step": 8259
+    },
+    {
+      "epoch": 1.5795009083086338,
+      "grad_norm": 9.254097520712193,
+      "learning_rate": 1e-06,
+      "loss": 0.5389,
+      "step": 8260
+    },
+    {
+      "epoch": 1.579692131178889,
+      "grad_norm": 2.6300213334991493,
+      "learning_rate": 1e-06,
+      "loss": 0.0757,
+      "step": 8261
+    },
+    {
+      "epoch": 1.5798833540491444,
+      "grad_norm": 2.1994304656709502,
+      "learning_rate": 1e-06,
+      "loss": 0.0914,
+      "step": 8262
+    },
+    {
+      "epoch": 1.5800745769193996,
+      "grad_norm": 2.1060081456272295,
+      "learning_rate": 1e-06,
+      "loss": 0.0985,
+      "step": 8263
+    },
+    {
+      "epoch": 1.5802657997896548,
+      "grad_norm": 1.2784384606580872,
+      "learning_rate": 1e-06,
+      "loss": 0.0178,
+      "step": 8264
+    },
+    {
+      "epoch": 1.5804570226599102,
+      "grad_norm": 6.618789641033763,
+      "learning_rate": 1e-06,
+      "loss": 0.2897,
+      "step": 8265
+    },
+    {
+      "epoch": 1.5806482455301654,
+      "grad_norm": 2.0029957507922105,
+      "learning_rate": 1e-06,
+      "loss": 0.0265,
+      "step": 8266
+    },
+    {
+      "epoch": 1.5808394684004208,
+      "grad_norm": 6.394650858399548,
+      "learning_rate": 1e-06,
+      "loss": 0.1244,
+      "step": 8267
+    },
+    {
+      "epoch": 1.581030691270676,
+      "grad_norm": 5.157176541398441,
+      "learning_rate": 1e-06,
+      "loss": 0.3213,
+      "step": 8268
+    },
+    {
+      "epoch": 1.5812219141409312,
+      "grad_norm": 3.5562394663247443,
+      "learning_rate": 1e-06,
+      "loss": 0.231,
+      "step": 8269
+    },
+    {
+      "epoch": 1.5814131370111866,
+      "grad_norm": 5.3755803349385385,
+      "learning_rate": 1e-06,
+      "loss": 0.702,
+      "step": 8270
+    },
+    {
+      "epoch": 1.5816043598814418,
+      "grad_norm": 2.0953190176888747,
+      "learning_rate": 1e-06,
+      "loss": 0.0837,
+      "step": 8271
+    },
+    {
+      "epoch": 1.5817955827516972,
+      "grad_norm": 2.2130391945589336,
+      "learning_rate": 1e-06,
+      "loss": 0.0963,
+      "step": 8272
+    },
+    {
+      "epoch": 1.5819868056219524,
+      "grad_norm": 2.6113444679450724,
+      "learning_rate": 1e-06,
+      "loss": 0.0847,
+      "step": 8273
+    },
+    {
+      "epoch": 1.5821780284922076,
+      "grad_norm": 3.195938757824093,
+      "learning_rate": 1e-06,
+      "loss": 0.1072,
+      "step": 8274
+    },
+    {
+      "epoch": 1.582369251362463,
+      "grad_norm": 1.3254523818572452,
+      "learning_rate": 1e-06,
+      "loss": 0.1084,
+      "step": 8275
+    },
+    {
+      "epoch": 1.5825604742327184,
+      "grad_norm": 2.602327159964991,
+      "learning_rate": 1e-06,
+      "loss": 0.1372,
+      "step": 8276
+    },
+    {
+      "epoch": 1.5827516971029736,
+      "grad_norm": 5.968376108012901,
+      "learning_rate": 1e-06,
+      "loss": 0.2142,
+      "step": 8277
+    },
+    {
+      "epoch": 1.5829429199732288,
+      "grad_norm": 4.653687251895227,
+      "learning_rate": 1e-06,
+      "loss": 0.3122,
+      "step": 8278
+    },
+    {
+      "epoch": 1.583134142843484,
+      "grad_norm": 4.197028180449275,
+      "learning_rate": 1e-06,
+      "loss": 0.0549,
+      "step": 8279
+    },
+    {
+      "epoch": 1.5833253657137394,
+      "grad_norm": 3.032240558290969,
+      "learning_rate": 1e-06,
+      "loss": 0.0512,
+      "step": 8280
+    },
+    {
+      "epoch": 1.5835165885839948,
+      "grad_norm": 5.272500640160064,
+      "learning_rate": 1e-06,
+      "loss": 0.2532,
+      "step": 8281
+    },
+    {
+      "epoch": 1.58370781145425,
+      "grad_norm": 3.8764896605850825,
+      "learning_rate": 1e-06,
+      "loss": 0.5613,
+      "step": 8282
+    },
+    {
+      "epoch": 1.5838990343245052,
+      "grad_norm": 4.209622151143201,
+      "learning_rate": 1e-06,
+      "loss": 0.2247,
+      "step": 8283
+    },
+    {
+      "epoch": 1.5840902571947604,
+      "grad_norm": 2.6058197471362883,
+      "learning_rate": 1e-06,
+      "loss": 0.1297,
+      "step": 8284
+    },
+    {
+      "epoch": 1.5842814800650158,
+      "grad_norm": 2.822461015214128,
+      "learning_rate": 1e-06,
+      "loss": 0.1698,
+      "step": 8285
+    },
+    {
+      "epoch": 1.5844727029352712,
+      "grad_norm": 4.71686683113237,
+      "learning_rate": 1e-06,
+      "loss": 0.229,
+      "step": 8286
+    },
+    {
+      "epoch": 1.5846639258055264,
+      "grad_norm": 2.6515920750907984,
+      "learning_rate": 1e-06,
+      "loss": 0.1225,
+      "step": 8287
+    },
+    {
+      "epoch": 1.5848551486757816,
+      "grad_norm": 2.596019625114944,
+      "learning_rate": 1e-06,
+      "loss": 0.0265,
+      "step": 8288
+    },
+    {
+      "epoch": 1.5850463715460368,
+      "grad_norm": 2.2121900912551373,
+      "learning_rate": 1e-06,
+      "loss": 0.0419,
+      "step": 8289
+    },
+    {
+      "epoch": 1.5852375944162922,
+      "grad_norm": 3.6808624011714186,
+      "learning_rate": 1e-06,
+      "loss": 0.2315,
+      "step": 8290
+    },
+    {
+      "epoch": 1.5854288172865476,
+      "grad_norm": 2.510302012082214,
+      "learning_rate": 1e-06,
+      "loss": 0.0523,
+      "step": 8291
+    },
+    {
+      "epoch": 1.5856200401568028,
+      "grad_norm": 3.0327346154991663,
+      "learning_rate": 1e-06,
+      "loss": 0.0542,
+      "step": 8292
+    },
+    {
+      "epoch": 1.585811263027058,
+      "grad_norm": 5.55141725057577,
+      "learning_rate": 1e-06,
+      "loss": 0.2483,
+      "step": 8293
+    },
+    {
+      "epoch": 1.5860024858973132,
+      "grad_norm": 5.039542998832827,
+      "learning_rate": 1e-06,
+      "loss": 0.5249,
+      "step": 8294
+    },
+    {
+      "epoch": 1.5861937087675686,
+      "grad_norm": 2.903431992575382,
+      "learning_rate": 1e-06,
+      "loss": 0.2309,
+      "step": 8295
+    },
+    {
+      "epoch": 1.586384931637824,
+      "grad_norm": 2.448997767455585,
+      "learning_rate": 1e-06,
+      "loss": 0.1838,
+      "step": 8296
+    },
+    {
+      "epoch": 1.5865761545080792,
+      "grad_norm": 2.9866195147345866,
+      "learning_rate": 1e-06,
+      "loss": 0.2181,
+      "step": 8297
+    },
+    {
+      "epoch": 1.5867673773783344,
+      "grad_norm": 3.29609302087646,
+      "learning_rate": 1e-06,
+      "loss": 0.2711,
+      "step": 8298
+    },
+    {
+      "epoch": 1.5869586002485896,
+      "grad_norm": 3.444136161407593,
+      "learning_rate": 1e-06,
+      "loss": 0.1347,
+      "step": 8299
+    },
+    {
+      "epoch": 1.587149823118845,
+      "grad_norm": 1.4262549880506399,
+      "learning_rate": 1e-06,
+      "loss": 0.0472,
+      "step": 8300
+    },
+    {
+      "epoch": 1.5873410459891004,
+      "grad_norm": 3.1805579191492472,
+      "learning_rate": 1e-06,
+      "loss": 0.054,
+      "step": 8301
+    },
+    {
+      "epoch": 1.5875322688593556,
+      "grad_norm": 4.24380625264747,
+      "learning_rate": 1e-06,
+      "loss": 0.022,
+      "step": 8302
+    },
+    {
+      "epoch": 1.5877234917296108,
+      "grad_norm": 1.5785694818293596,
+      "learning_rate": 1e-06,
+      "loss": 0.0271,
+      "step": 8303
+    },
+    {
+      "epoch": 1.587914714599866,
+      "grad_norm": 5.457112338338124,
+      "learning_rate": 1e-06,
+      "loss": 0.0911,
+      "step": 8304
+    },
+    {
+      "epoch": 1.5881059374701214,
+      "grad_norm": 5.963823932036725,
+      "learning_rate": 1e-06,
+      "loss": 0.1008,
+      "step": 8305
+    },
+    {
+      "epoch": 1.5882971603403768,
+      "grad_norm": 5.241954451495894,
+      "learning_rate": 1e-06,
+      "loss": 0.4737,
+      "step": 8306
+    },
+    {
+      "epoch": 1.588488383210632,
+      "grad_norm": 5.682489220960995,
+      "learning_rate": 1e-06,
+      "loss": 0.3236,
+      "step": 8307
+    },
+    {
+      "epoch": 1.5886796060808872,
+      "grad_norm": 3.064488602208598,
+      "learning_rate": 1e-06,
+      "loss": 0.1465,
+      "step": 8308
+    },
+    {
+      "epoch": 1.5888708289511424,
+      "grad_norm": 2.8349759071577734,
+      "learning_rate": 1e-06,
+      "loss": 0.1271,
+      "step": 8309
+    },
+    {
+      "epoch": 1.5890620518213978,
+      "grad_norm": 2.665060443729502,
+      "learning_rate": 1e-06,
+      "loss": 0.0876,
+      "step": 8310
+    },
+    {
+      "epoch": 1.5892532746916532,
+      "grad_norm": 2.8800208989020883,
+      "learning_rate": 1e-06,
+      "loss": 0.0503,
+      "step": 8311
+    },
+    {
+      "epoch": 1.5894444975619084,
+      "grad_norm": 2.8730510242258447,
+      "learning_rate": 1e-06,
+      "loss": 0.0548,
+      "step": 8312
+    },
+    {
+      "epoch": 1.5896357204321636,
+      "grad_norm": 5.918934920099976,
+      "learning_rate": 1e-06,
+      "loss": 0.1768,
+      "step": 8313
+    },
+    {
+      "epoch": 1.5898269433024188,
+      "grad_norm": 3.6911353602242167,
+      "learning_rate": 1e-06,
+      "loss": 0.1096,
+      "step": 8314
+    },
+    {
+      "epoch": 1.5900181661726742,
+      "grad_norm": 2.2673489030084526,
+      "learning_rate": 1e-06,
+      "loss": 0.0369,
+      "step": 8315
+    },
+    {
+      "epoch": 1.5902093890429296,
+      "grad_norm": 2.0664910513598507,
+      "learning_rate": 1e-06,
+      "loss": 0.0513,
+      "step": 8316
+    },
+    {
+      "epoch": 1.5904006119131848,
+      "grad_norm": 3.859904264537721,
+      "learning_rate": 1e-06,
+      "loss": 0.053,
+      "step": 8317
+    },
+    {
+      "epoch": 1.59059183478344,
+      "grad_norm": 4.7792194517635656,
+      "learning_rate": 1e-06,
+      "loss": 0.1547,
+      "step": 8318
+    },
+    {
+      "epoch": 1.5907830576536954,
+      "grad_norm": 2.9792042620938766,
+      "learning_rate": 1e-06,
+      "loss": 0.0996,
+      "step": 8319
+    },
+    {
+      "epoch": 1.5909742805239506,
+      "grad_norm": 4.071757640303764,
+      "learning_rate": 1e-06,
+      "loss": 0.2174,
+      "step": 8320
+    },
+    {
+      "epoch": 1.591165503394206,
+      "grad_norm": 4.506714791336281,
+      "learning_rate": 1e-06,
+      "loss": 0.239,
+      "step": 8321
+    },
+    {
+      "epoch": 1.5913567262644612,
+      "grad_norm": 4.5337483851848255,
+      "learning_rate": 1e-06,
+      "loss": 0.139,
+      "step": 8322
+    },
+    {
+      "epoch": 1.5915479491347164,
+      "grad_norm": 3.44891017341382,
+      "learning_rate": 1e-06,
+      "loss": 0.1207,
+      "step": 8323
+    },
+    {
+      "epoch": 1.5917391720049718,
+      "grad_norm": 3.2521104562635585,
+      "learning_rate": 1e-06,
+      "loss": 0.0722,
+      "step": 8324
+    },
+    {
+      "epoch": 1.591930394875227,
+      "grad_norm": 4.007919101896114,
+      "learning_rate": 1e-06,
+      "loss": 0.195,
+      "step": 8325
+    },
+    {
+      "epoch": 1.5921216177454824,
+      "grad_norm": 2.2279179245437697,
+      "learning_rate": 1e-06,
+      "loss": 0.0734,
+      "step": 8326
+    },
+    {
+      "epoch": 1.5923128406157376,
+      "grad_norm": 3.6966865033602696,
+      "learning_rate": 1e-06,
+      "loss": 0.059,
+      "step": 8327
+    },
+    {
+      "epoch": 1.5925040634859928,
+      "grad_norm": 3.9487245214593676,
+      "learning_rate": 1e-06,
+      "loss": 0.2272,
+      "step": 8328
+    },
+    {
+      "epoch": 1.5926952863562482,
+      "grad_norm": 1.1474384971209972,
+      "learning_rate": 1e-06,
+      "loss": 0.0222,
+      "step": 8329
+    },
+    {
+      "epoch": 1.5928865092265037,
+      "grad_norm": 11.97201007636373,
+      "learning_rate": 1e-06,
+      "loss": 0.3028,
+      "step": 8330
+    },
+    {
+      "epoch": 1.5930777320967588,
+      "grad_norm": 4.924397914894647,
+      "learning_rate": 1e-06,
+      "loss": 0.3132,
+      "step": 8331
+    },
+    {
+      "epoch": 1.593268954967014,
+      "grad_norm": 6.235033043572693,
+      "learning_rate": 1e-06,
+      "loss": 0.1568,
+      "step": 8332
+    },
+    {
+      "epoch": 1.5934601778372692,
+      "grad_norm": 3.373003616453212,
+      "learning_rate": 1e-06,
+      "loss": 0.2103,
+      "step": 8333
+    },
+    {
+      "epoch": 1.5936514007075246,
+      "grad_norm": 3.5834256685174246,
+      "learning_rate": 1e-06,
+      "loss": 0.2164,
+      "step": 8334
+    },
+    {
+      "epoch": 1.59384262357778,
+      "grad_norm": 2.0507034722974273,
+      "learning_rate": 1e-06,
+      "loss": 0.1846,
+      "step": 8335
+    },
+    {
+      "epoch": 1.5940338464480353,
+      "grad_norm": 2.5179578496239587,
+      "learning_rate": 1e-06,
+      "loss": 0.0798,
+      "step": 8336
+    },
+    {
+      "epoch": 1.5942250693182904,
+      "grad_norm": 2.3606881030765847,
+      "learning_rate": 1e-06,
+      "loss": 0.0747,
+      "step": 8337
+    },
+    {
+      "epoch": 1.5944162921885456,
+      "grad_norm": 2.469049749405439,
+      "learning_rate": 1e-06,
+      "loss": 0.1012,
+      "step": 8338
+    },
+    {
+      "epoch": 1.594607515058801,
+      "grad_norm": 1.9654684419856887,
+      "learning_rate": 1e-06,
+      "loss": 0.0591,
+      "step": 8339
+    },
+    {
+      "epoch": 1.5947987379290565,
+      "grad_norm": 3.007854193856721,
+      "learning_rate": 1e-06,
+      "loss": 0.2039,
+      "step": 8340
+    },
+    {
+      "epoch": 1.5949899607993117,
+      "grad_norm": 2.848542587007454,
+      "learning_rate": 1e-06,
+      "loss": 0.0395,
+      "step": 8341
+    },
+    {
+      "epoch": 1.5951811836695668,
+      "grad_norm": 7.462371784175837,
+      "learning_rate": 1e-06,
+      "loss": 0.2598,
+      "step": 8342
+    },
+    {
+      "epoch": 1.595372406539822,
+      "grad_norm": 3.4240278326133304,
+      "learning_rate": 1e-06,
+      "loss": 0.1473,
+      "step": 8343
+    },
+    {
+      "epoch": 1.5955636294100775,
+      "grad_norm": 4.6629899163870325,
+      "learning_rate": 1e-06,
+      "loss": 0.3628,
+      "step": 8344
+    },
+    {
+      "epoch": 1.5957548522803329,
+      "grad_norm": 2.3774830737549344,
+      "learning_rate": 1e-06,
+      "loss": 0.1466,
+      "step": 8345
+    },
+    {
+      "epoch": 1.595946075150588,
+      "grad_norm": 2.2815708300589055,
+      "learning_rate": 1e-06,
+      "loss": 0.1723,
+      "step": 8346
+    },
+    {
+      "epoch": 1.5961372980208433,
+      "grad_norm": 4.105470840642308,
+      "learning_rate": 1e-06,
+      "loss": 0.2337,
+      "step": 8347
+    },
+    {
+      "epoch": 1.5963285208910984,
+      "grad_norm": 2.287362883580487,
+      "learning_rate": 1e-06,
+      "loss": 0.1071,
+      "step": 8348
+    },
+    {
+      "epoch": 1.5965197437613539,
+      "grad_norm": 5.525455694263819,
+      "learning_rate": 1e-06,
+      "loss": 0.2275,
+      "step": 8349
+    },
+    {
+      "epoch": 1.5967109666316093,
+      "grad_norm": 2.216016885940997,
+      "learning_rate": 1e-06,
+      "loss": 0.1112,
+      "step": 8350
+    },
+    {
+      "epoch": 1.5969021895018645,
+      "grad_norm": 3.475611318702983,
+      "learning_rate": 1e-06,
+      "loss": 0.0291,
+      "step": 8351
+    },
+    {
+      "epoch": 1.5970934123721197,
+      "grad_norm": 3.2963740976210945,
+      "learning_rate": 1e-06,
+      "loss": 0.0467,
+      "step": 8352
+    },
+    {
+      "epoch": 1.5972846352423749,
+      "grad_norm": 2.167302319621102,
+      "learning_rate": 1e-06,
+      "loss": 0.0243,
+      "step": 8353
+    },
+    {
+      "epoch": 1.5974758581126303,
+      "grad_norm": 5.799075796222363,
+      "learning_rate": 1e-06,
+      "loss": 0.0994,
+      "step": 8354
+    },
+    {
+      "epoch": 1.5976670809828857,
+      "grad_norm": 5.670791732706503,
+      "learning_rate": 1e-06,
+      "loss": 0.3566,
+      "step": 8355
+    },
+    {
+      "epoch": 1.5978583038531409,
+      "grad_norm": 4.19941655602427,
+      "learning_rate": 1e-06,
+      "loss": 0.3171,
+      "step": 8356
+    },
+    {
+      "epoch": 1.598049526723396,
+      "grad_norm": 4.537720788885731,
+      "learning_rate": 1e-06,
+      "loss": 0.3584,
+      "step": 8357
+    },
+    {
+      "epoch": 1.5982407495936513,
+      "grad_norm": 3.5787603488869753,
+      "learning_rate": 1e-06,
+      "loss": 0.1991,
+      "step": 8358
+    },
+    {
+      "epoch": 1.5984319724639067,
+      "grad_norm": 3.534703953930393,
+      "learning_rate": 1e-06,
+      "loss": 0.1107,
+      "step": 8359
+    },
+    {
+      "epoch": 1.598623195334162,
+      "grad_norm": 2.8001918624810913,
+      "learning_rate": 1e-06,
+      "loss": 0.1208,
+      "step": 8360
+    },
+    {
+      "epoch": 1.5988144182044173,
+      "grad_norm": 2.906565167145202,
+      "learning_rate": 1e-06,
+      "loss": 0.1311,
+      "step": 8361
+    },
+    {
+      "epoch": 1.5990056410746725,
+      "grad_norm": 2.3177925580326497,
+      "learning_rate": 1e-06,
+      "loss": 0.0737,
+      "step": 8362
+    },
+    {
+      "epoch": 1.5991968639449277,
+      "grad_norm": 2.6645221034081854,
+      "learning_rate": 1e-06,
+      "loss": 0.0527,
+      "step": 8363
+    },
+    {
+      "epoch": 1.599388086815183,
+      "grad_norm": 2.6269504929993786,
+      "learning_rate": 1e-06,
+      "loss": 0.0498,
+      "step": 8364
+    },
+    {
+      "epoch": 1.5995793096854385,
+      "grad_norm": 6.471572039267108,
+      "learning_rate": 1e-06,
+      "loss": 0.2844,
+      "step": 8365
+    },
+    {
+      "epoch": 1.5997705325556937,
+      "grad_norm": 1.7844918849094662,
+      "learning_rate": 1e-06,
+      "loss": 0.0465,
+      "step": 8366
+    },
+    {
+      "epoch": 1.5999617554259489,
+      "grad_norm": 6.691632286480684,
+      "learning_rate": 1e-06,
+      "loss": 0.1441,
+      "step": 8367
+    },
+    {
+      "epoch": 1.600152978296204,
+      "grad_norm": 10.240680086082579,
+      "learning_rate": 1e-06,
+      "loss": 0.3047,
+      "step": 8368
+    },
+    {
+      "epoch": 1.6003442011664595,
+      "grad_norm": 5.512799887598679,
+      "learning_rate": 1e-06,
+      "loss": 0.6199,
+      "step": 8369
+    },
+    {
+      "epoch": 1.600535424036715,
+      "grad_norm": 3.5025799642906086,
+      "learning_rate": 1e-06,
+      "loss": 0.0946,
+      "step": 8370
+    },
+    {
+      "epoch": 1.60072664690697,
+      "grad_norm": 1.6766008271371957,
+      "learning_rate": 1e-06,
+      "loss": 0.1699,
+      "step": 8371
+    },
+    {
+      "epoch": 1.6009178697772253,
+      "grad_norm": 4.745990064453837,
+      "learning_rate": 1e-06,
+      "loss": 0.5115,
+      "step": 8372
+    },
+    {
+      "epoch": 1.6011090926474807,
+      "grad_norm": 3.4958711521314236,
+      "learning_rate": 1e-06,
+      "loss": 0.0971,
+      "step": 8373
+    },
+    {
+      "epoch": 1.6013003155177359,
+      "grad_norm": 2.943400685831114,
+      "learning_rate": 1e-06,
+      "loss": 0.0487,
+      "step": 8374
+    },
+    {
+      "epoch": 1.6014915383879913,
+      "grad_norm": 4.472781197081173,
+      "learning_rate": 1e-06,
+      "loss": 0.1845,
+      "step": 8375
+    },
+    {
+      "epoch": 1.6016827612582465,
+      "grad_norm": 2.3203932622860717,
+      "learning_rate": 1e-06,
+      "loss": 0.0667,
+      "step": 8376
+    },
+    {
+      "epoch": 1.6018739841285017,
+      "grad_norm": 4.729624363377322,
+      "learning_rate": 1e-06,
+      "loss": 0.2613,
+      "step": 8377
+    },
+    {
+      "epoch": 1.602065206998757,
+      "grad_norm": 2.6003724088255704,
+      "learning_rate": 1e-06,
+      "loss": 0.0345,
+      "step": 8378
+    },
+    {
+      "epoch": 1.6022564298690123,
+      "grad_norm": 2.9036688058169635,
+      "learning_rate": 1e-06,
+      "loss": 0.067,
+      "step": 8379
+    },
+    {
+      "epoch": 1.6024476527392677,
+      "grad_norm": 4.455166104663395,
+      "learning_rate": 1e-06,
+      "loss": 0.1638,
+      "step": 8380
+    },
+    {
+      "epoch": 1.602638875609523,
+      "grad_norm": 3.6793653343077612,
+      "learning_rate": 1e-06,
+      "loss": 0.3703,
+      "step": 8381
+    },
+    {
+      "epoch": 1.602830098479778,
+      "grad_norm": 4.570294971921516,
+      "learning_rate": 1e-06,
+      "loss": 0.3665,
+      "step": 8382
+    },
+    {
+      "epoch": 1.6030213213500335,
+      "grad_norm": 4.470732775881905,
+      "learning_rate": 1e-06,
+      "loss": 0.4337,
+      "step": 8383
+    },
+    {
+      "epoch": 1.6032125442202887,
+      "grad_norm": 4.127485277968733,
+      "learning_rate": 1e-06,
+      "loss": 0.0715,
+      "step": 8384
+    },
+    {
+      "epoch": 1.603403767090544,
+      "grad_norm": 3.1965543017430034,
+      "learning_rate": 1e-06,
+      "loss": 0.1652,
+      "step": 8385
+    },
+    {
+      "epoch": 1.6035949899607993,
+      "grad_norm": 4.346723238766038,
+      "learning_rate": 1e-06,
+      "loss": 0.2095,
+      "step": 8386
+    },
+    {
+      "epoch": 1.6037862128310545,
+      "grad_norm": 2.5732877723666316,
+      "learning_rate": 1e-06,
+      "loss": 0.0749,
+      "step": 8387
+    },
+    {
+      "epoch": 1.60397743570131,
+      "grad_norm": 4.793807350422404,
+      "learning_rate": 1e-06,
+      "loss": 0.1951,
+      "step": 8388
+    },
+    {
+      "epoch": 1.6041686585715653,
+      "grad_norm": 3.991006396535834,
+      "learning_rate": 1e-06,
+      "loss": 0.0779,
+      "step": 8389
+    },
+    {
+      "epoch": 1.6043598814418205,
+      "grad_norm": 5.618444904195421,
+      "learning_rate": 1e-06,
+      "loss": 0.0665,
+      "step": 8390
+    },
+    {
+      "epoch": 1.6045511043120757,
+      "grad_norm": 1.1448191518263873,
+      "learning_rate": 1e-06,
+      "loss": 0.0117,
+      "step": 8391
+    },
+    {
+      "epoch": 1.604742327182331,
+      "grad_norm": 3.2916791569597548,
+      "learning_rate": 1e-06,
+      "loss": 0.0122,
+      "step": 8392
+    },
+    {
+      "epoch": 1.6049335500525863,
+      "grad_norm": 8.223166087964776,
+      "learning_rate": 1e-06,
+      "loss": 0.3834,
+      "step": 8393
+    },
+    {
+      "epoch": 1.6051247729228417,
+      "grad_norm": 6.171542156274226,
+      "learning_rate": 1e-06,
+      "loss": 0.4023,
+      "step": 8394
+    },
+    {
+      "epoch": 1.605315995793097,
+      "grad_norm": 4.983598511103502,
+      "learning_rate": 1e-06,
+      "loss": 0.463,
+      "step": 8395
+    },
+    {
+      "epoch": 1.605507218663352,
+      "grad_norm": 2.3172002914188834,
+      "learning_rate": 1e-06,
+      "loss": 0.1698,
+      "step": 8396
+    },
+    {
+      "epoch": 1.6056984415336073,
+      "grad_norm": 1.626815515310729,
+      "learning_rate": 1e-06,
+      "loss": 0.0575,
+      "step": 8397
+    },
+    {
+      "epoch": 1.6058896644038627,
+      "grad_norm": 4.2291939770548055,
+      "learning_rate": 1e-06,
+      "loss": 0.0831,
+      "step": 8398
+    },
+    {
+      "epoch": 1.6060808872741181,
+      "grad_norm": 3.3216878757542987,
+      "learning_rate": 1e-06,
+      "loss": 0.1626,
+      "step": 8399
+    },
+    {
+      "epoch": 1.6062721101443733,
+      "grad_norm": 3.577454749886823,
+      "learning_rate": 1e-06,
+      "loss": 0.0658,
+      "step": 8400
+    },
+    {
+      "epoch": 1.6064633330146285,
+      "grad_norm": 3.3864979743495254,
+      "learning_rate": 1e-06,
+      "loss": 0.0768,
+      "step": 8401
+    },
+    {
+      "epoch": 1.6066545558848837,
+      "grad_norm": 3.166947469474993,
+      "learning_rate": 1e-06,
+      "loss": 0.0857,
+      "step": 8402
+    },
+    {
+      "epoch": 1.6068457787551391,
+      "grad_norm": 2.6827234803295736,
+      "learning_rate": 1e-06,
+      "loss": 0.0636,
+      "step": 8403
+    },
+    {
+      "epoch": 1.6070370016253945,
+      "grad_norm": 5.473231980333638,
+      "learning_rate": 1e-06,
+      "loss": 0.0698,
+      "step": 8404
+    },
+    {
+      "epoch": 1.6072282244956497,
+      "grad_norm": 8.70310220766764,
+      "learning_rate": 1e-06,
+      "loss": 0.356,
+      "step": 8405
+    },
+    {
+      "epoch": 1.607419447365905,
+      "grad_norm": 9.347918649028117,
+      "learning_rate": 1e-06,
+      "loss": 0.5722,
+      "step": 8406
+    },
+    {
+      "epoch": 1.60761067023616,
+      "grad_norm": 9.355648566252093,
+      "learning_rate": 1e-06,
+      "loss": 0.2948,
+      "step": 8407
+    },
+    {
+      "epoch": 1.6078018931064155,
+      "grad_norm": 4.984244317999348,
+      "learning_rate": 1e-06,
+      "loss": 0.2381,
+      "step": 8408
+    },
+    {
+      "epoch": 1.607993115976671,
+      "grad_norm": 3.7582617984795372,
+      "learning_rate": 1e-06,
+      "loss": 0.1982,
+      "step": 8409
+    },
+    {
+      "epoch": 1.6081843388469261,
+      "grad_norm": 1.904591949294297,
+      "learning_rate": 1e-06,
+      "loss": 0.1469,
+      "step": 8410
+    },
+    {
+      "epoch": 1.6083755617171813,
+      "grad_norm": 3.4348208997919327,
+      "learning_rate": 1e-06,
+      "loss": 0.0789,
+      "step": 8411
+    },
+    {
+      "epoch": 1.6085667845874365,
+      "grad_norm": 2.0524688461857314,
+      "learning_rate": 1e-06,
+      "loss": 0.089,
+      "step": 8412
+    },
+    {
+      "epoch": 1.608758007457692,
+      "grad_norm": 3.165983996506883,
+      "learning_rate": 1e-06,
+      "loss": 0.144,
+      "step": 8413
+    },
+    {
+      "epoch": 1.6089492303279473,
+      "grad_norm": 5.143709596496087,
+      "learning_rate": 1e-06,
+      "loss": 0.0536,
+      "step": 8414
+    },
+    {
+      "epoch": 1.6091404531982025,
+      "grad_norm": 2.5954799168632245,
+      "learning_rate": 1e-06,
+      "loss": 0.0335,
+      "step": 8415
+    },
+    {
+      "epoch": 1.6093316760684577,
+      "grad_norm": 2.9318704432425853,
+      "learning_rate": 1e-06,
+      "loss": 0.0483,
+      "step": 8416
+    },
+    {
+      "epoch": 1.609522898938713,
+      "grad_norm": 4.5287548693326185,
+      "learning_rate": 1e-06,
+      "loss": 0.0658,
+      "step": 8417
+    },
+    {
+      "epoch": 1.6097141218089683,
+      "grad_norm": 5.182942422107244,
+      "learning_rate": 1e-06,
+      "loss": 0.1066,
+      "step": 8418
+    },
+    {
+      "epoch": 1.6099053446792237,
+      "grad_norm": 9.059448991402956,
+      "learning_rate": 1e-06,
+      "loss": 0.7449,
+      "step": 8419
+    },
+    {
+      "epoch": 1.610096567549479,
+      "grad_norm": 4.424194805788442,
+      "learning_rate": 1e-06,
+      "loss": 0.3044,
+      "step": 8420
+    },
+    {
+      "epoch": 1.6102877904197341,
+      "grad_norm": 1.8537235444869808,
+      "learning_rate": 1e-06,
+      "loss": 0.0789,
+      "step": 8421
+    },
+    {
+      "epoch": 1.6104790132899893,
+      "grad_norm": 6.232298235023995,
+      "learning_rate": 1e-06,
+      "loss": 0.1687,
+      "step": 8422
+    },
+    {
+      "epoch": 1.6106702361602447,
+      "grad_norm": 3.2198423735617085,
+      "learning_rate": 1e-06,
+      "loss": 0.251,
+      "step": 8423
+    },
+    {
+      "epoch": 1.6108614590305002,
+      "grad_norm": 3.9734550880705903,
+      "learning_rate": 1e-06,
+      "loss": 0.2437,
+      "step": 8424
+    },
+    {
+      "epoch": 1.6110526819007553,
+      "grad_norm": 2.8242969211929645,
+      "learning_rate": 1e-06,
+      "loss": 0.0792,
+      "step": 8425
+    },
+    {
+      "epoch": 1.6112439047710105,
+      "grad_norm": 2.9013142665991523,
+      "learning_rate": 1e-06,
+      "loss": 0.092,
+      "step": 8426
+    },
+    {
+      "epoch": 1.6114351276412657,
+      "grad_norm": 3.3563038116630746,
+      "learning_rate": 1e-06,
+      "loss": 0.127,
+      "step": 8427
+    },
+    {
+      "epoch": 1.6116263505115211,
+      "grad_norm": 2.6556355382755554,
+      "learning_rate": 1e-06,
+      "loss": 0.0545,
+      "step": 8428
+    },
+    {
+      "epoch": 1.6118175733817766,
+      "grad_norm": 2.1339053752981454,
+      "learning_rate": 1e-06,
+      "loss": 0.0256,
+      "step": 8429
+    },
+    {
+      "epoch": 1.6120087962520318,
+      "grad_norm": 7.595855797674957,
+      "learning_rate": 1e-06,
+      "loss": 0.085,
+      "step": 8430
+    },
+    {
+      "epoch": 1.612200019122287,
+      "grad_norm": 4.599957067869777,
+      "learning_rate": 1e-06,
+      "loss": 0.2029,
+      "step": 8431
+    },
+    {
+      "epoch": 1.6123912419925424,
+      "grad_norm": 3.637389484837209,
+      "learning_rate": 1e-06,
+      "loss": 0.1684,
+      "step": 8432
+    },
+    {
+      "epoch": 1.6125824648627975,
+      "grad_norm": 14.601296542354499,
+      "learning_rate": 1e-06,
+      "loss": 0.1355,
+      "step": 8433
+    },
+    {
+      "epoch": 1.612773687733053,
+      "grad_norm": 4.008237938859383,
+      "learning_rate": 1e-06,
+      "loss": 0.095,
+      "step": 8434
+    },
+    {
+      "epoch": 1.6129649106033082,
+      "grad_norm": 2.136208649442141,
+      "learning_rate": 1e-06,
+      "loss": 0.1166,
+      "step": 8435
+    },
+    {
+      "epoch": 1.6131561334735633,
+      "grad_norm": 3.8743708930476095,
+      "learning_rate": 1e-06,
+      "loss": 0.1588,
+      "step": 8436
+    },
+    {
+      "epoch": 1.6133473563438188,
+      "grad_norm": 4.25553152599362,
+      "learning_rate": 1e-06,
+      "loss": 0.1396,
+      "step": 8437
+    },
+    {
+      "epoch": 1.613538579214074,
+      "grad_norm": 3.6852262566722023,
+      "learning_rate": 1e-06,
+      "loss": 0.1398,
+      "step": 8438
+    },
+    {
+      "epoch": 1.6137298020843294,
+      "grad_norm": 2.795502053961616,
+      "learning_rate": 1e-06,
+      "loss": 0.0262,
+      "step": 8439
+    },
+    {
+      "epoch": 1.6139210249545846,
+      "grad_norm": 3.7316680418429127,
+      "learning_rate": 1e-06,
+      "loss": 0.0474,
+      "step": 8440
+    },
+    {
+      "epoch": 1.6141122478248398,
+      "grad_norm": 4.298168972423365,
+      "learning_rate": 1e-06,
+      "loss": 0.0883,
+      "step": 8441
+    },
+    {
+      "epoch": 1.6143034706950952,
+      "grad_norm": 9.177790694988397,
+      "learning_rate": 1e-06,
+      "loss": 0.0954,
+      "step": 8442
+    },
+    {
+      "epoch": 1.6144946935653506,
+      "grad_norm": 5.798957388989542,
+      "learning_rate": 1e-06,
+      "loss": 0.4108,
+      "step": 8443
+    },
+    {
+      "epoch": 1.6146859164356058,
+      "grad_norm": 3.3187819599688893,
+      "learning_rate": 1e-06,
+      "loss": 0.2805,
+      "step": 8444
+    },
+    {
+      "epoch": 1.614877139305861,
+      "grad_norm": 2.4450464073410783,
+      "learning_rate": 1e-06,
+      "loss": 0.0751,
+      "step": 8445
+    },
+    {
+      "epoch": 1.6150683621761162,
+      "grad_norm": 1.5576138200394634,
+      "learning_rate": 1e-06,
+      "loss": 0.1119,
+      "step": 8446
+    },
+    {
+      "epoch": 1.6152595850463716,
+      "grad_norm": 3.5915494069181317,
+      "learning_rate": 1e-06,
+      "loss": 0.0739,
+      "step": 8447
+    },
+    {
+      "epoch": 1.615450807916627,
+      "grad_norm": 2.1689143137839975,
+      "learning_rate": 1e-06,
+      "loss": 0.122,
+      "step": 8448
+    },
+    {
+      "epoch": 1.6156420307868822,
+      "grad_norm": 3.6877270644801805,
+      "learning_rate": 1e-06,
+      "loss": 0.1805,
+      "step": 8449
+    },
+    {
+      "epoch": 1.6158332536571374,
+      "grad_norm": 2.351010077087795,
+      "learning_rate": 1e-06,
+      "loss": 0.0539,
+      "step": 8450
+    },
+    {
+      "epoch": 1.6160244765273926,
+      "grad_norm": 3.7458327981158517,
+      "learning_rate": 1e-06,
+      "loss": 0.1069,
+      "step": 8451
+    },
+    {
+      "epoch": 1.616215699397648,
+      "grad_norm": 3.512695719740385,
+      "learning_rate": 1e-06,
+      "loss": 0.0824,
+      "step": 8452
+    },
+    {
+      "epoch": 1.6164069222679034,
+      "grad_norm": 1.9755260177478569,
+      "learning_rate": 1e-06,
+      "loss": 0.0303,
+      "step": 8453
+    },
+    {
+      "epoch": 1.6165981451381586,
+      "grad_norm": 1.8976768571057483,
+      "learning_rate": 1e-06,
+      "loss": 0.0347,
+      "step": 8454
+    },
+    {
+      "epoch": 1.6167893680084138,
+      "grad_norm": 5.2413262059071,
+      "learning_rate": 1e-06,
+      "loss": 0.0501,
+      "step": 8455
+    },
+    {
+      "epoch": 1.616980590878669,
+      "grad_norm": 4.4105779076270775,
+      "learning_rate": 1e-06,
+      "loss": 0.4885,
+      "step": 8456
+    },
+    {
+      "epoch": 1.6171718137489244,
+      "grad_norm": 5.413626693205644,
+      "learning_rate": 1e-06,
+      "loss": 0.4027,
+      "step": 8457
+    },
+    {
+      "epoch": 1.6173630366191798,
+      "grad_norm": 2.7885827258258655,
+      "learning_rate": 1e-06,
+      "loss": 0.1926,
+      "step": 8458
+    },
+    {
+      "epoch": 1.617554259489435,
+      "grad_norm": 3.370121432190894,
+      "learning_rate": 1e-06,
+      "loss": 0.1149,
+      "step": 8459
+    },
+    {
+      "epoch": 1.6177454823596902,
+      "grad_norm": 3.9421219114577015,
+      "learning_rate": 1e-06,
+      "loss": 0.3114,
+      "step": 8460
+    },
+    {
+      "epoch": 1.6179367052299454,
+      "grad_norm": 3.3442204224706575,
+      "learning_rate": 1e-06,
+      "loss": 0.1077,
+      "step": 8461
+    },
+    {
+      "epoch": 1.6181279281002008,
+      "grad_norm": 2.2153567303732506,
+      "learning_rate": 1e-06,
+      "loss": 0.0619,
+      "step": 8462
+    },
+    {
+      "epoch": 1.6183191509704562,
+      "grad_norm": 2.027795524461448,
+      "learning_rate": 1e-06,
+      "loss": 0.0434,
+      "step": 8463
+    },
+    {
+      "epoch": 1.6185103738407114,
+      "grad_norm": 1.4877221366364126,
+      "learning_rate": 1e-06,
+      "loss": 0.0251,
+      "step": 8464
+    },
+    {
+      "epoch": 1.6187015967109666,
+      "grad_norm": 3.9819021891484616,
+      "learning_rate": 1e-06,
+      "loss": 0.1524,
+      "step": 8465
+    },
+    {
+      "epoch": 1.6188928195812218,
+      "grad_norm": 2.072300025302066,
+      "learning_rate": 1e-06,
+      "loss": 0.0325,
+      "step": 8466
+    },
+    {
+      "epoch": 1.6190840424514772,
+      "grad_norm": 4.658644502149967,
+      "learning_rate": 1e-06,
+      "loss": 0.1243,
+      "step": 8467
+    },
+    {
+      "epoch": 1.6192752653217326,
+      "grad_norm": 5.400553618415058,
+      "learning_rate": 1e-06,
+      "loss": 0.392,
+      "step": 8468
+    },
+    {
+      "epoch": 1.6194664881919878,
+      "grad_norm": 3.937933761682815,
+      "learning_rate": 1e-06,
+      "loss": 0.1948,
+      "step": 8469
+    },
+    {
+      "epoch": 1.619657711062243,
+      "grad_norm": 5.153220656636617,
+      "learning_rate": 1e-06,
+      "loss": 0.2306,
+      "step": 8470
+    },
+    {
+      "epoch": 1.6198489339324982,
+      "grad_norm": 5.0275074562385464,
+      "learning_rate": 1e-06,
+      "loss": 0.1951,
+      "step": 8471
+    },
+    {
+      "epoch": 1.6200401568027536,
+      "grad_norm": 4.37232627911784,
+      "learning_rate": 1e-06,
+      "loss": 0.1501,
+      "step": 8472
+    },
+    {
+      "epoch": 1.620231379673009,
+      "grad_norm": 2.6121826599711504,
+      "learning_rate": 1e-06,
+      "loss": 0.1422,
+      "step": 8473
+    },
+    {
+      "epoch": 1.6204226025432642,
+      "grad_norm": 4.068629878039124,
+      "learning_rate": 1e-06,
+      "loss": 0.1749,
+      "step": 8474
+    },
+    {
+      "epoch": 1.6206138254135194,
+      "grad_norm": 2.338550547786124,
+      "learning_rate": 1e-06,
+      "loss": 0.0597,
+      "step": 8475
+    },
+    {
+      "epoch": 1.6208050482837746,
+      "grad_norm": 1.9455577894856768,
+      "learning_rate": 1e-06,
+      "loss": 0.0373,
+      "step": 8476
+    },
+    {
+      "epoch": 1.62099627115403,
+      "grad_norm": 2.884839345600853,
+      "learning_rate": 1e-06,
+      "loss": 0.0178,
+      "step": 8477
+    },
+    {
+      "epoch": 1.6211874940242854,
+      "grad_norm": 4.1660401191116545,
+      "learning_rate": 1e-06,
+      "loss": 0.0511,
+      "step": 8478
+    },
+    {
+      "epoch": 1.6213787168945406,
+      "grad_norm": 2.4491721214224174,
+      "learning_rate": 1e-06,
+      "loss": 0.0555,
+      "step": 8479
+    },
+    {
+      "epoch": 1.6215699397647958,
+      "grad_norm": 10.27850694790469,
+      "learning_rate": 1e-06,
+      "loss": 0.1636,
+      "step": 8480
+    },
+    {
+      "epoch": 1.621761162635051,
+      "grad_norm": 5.202392893196894,
+      "learning_rate": 1e-06,
+      "loss": 0.4606,
+      "step": 8481
+    },
+    {
+      "epoch": 1.6219523855053064,
+      "grad_norm": 4.551905834424067,
+      "learning_rate": 1e-06,
+      "loss": 0.303,
+      "step": 8482
+    },
+    {
+      "epoch": 1.6221436083755618,
+      "grad_norm": 6.6713901316788915,
+      "learning_rate": 1e-06,
+      "loss": 0.1206,
+      "step": 8483
+    },
+    {
+      "epoch": 1.622334831245817,
+      "grad_norm": 1.8374383513819978,
+      "learning_rate": 1e-06,
+      "loss": 0.084,
+      "step": 8484
+    },
+    {
+      "epoch": 1.6225260541160722,
+      "grad_norm": 3.0642869368178207,
+      "learning_rate": 1e-06,
+      "loss": 0.2738,
+      "step": 8485
+    },
+    {
+      "epoch": 1.6227172769863276,
+      "grad_norm": 3.443724666439968,
+      "learning_rate": 1e-06,
+      "loss": 0.2471,
+      "step": 8486
+    },
+    {
+      "epoch": 1.6229084998565828,
+      "grad_norm": 2.7338751635790137,
+      "learning_rate": 1e-06,
+      "loss": 0.0492,
+      "step": 8487
+    },
+    {
+      "epoch": 1.6230997227268382,
+      "grad_norm": 1.6045557846495229,
+      "learning_rate": 1e-06,
+      "loss": 0.058,
+      "step": 8488
+    },
+    {
+      "epoch": 1.6232909455970934,
+      "grad_norm": 2.747344555609793,
+      "learning_rate": 1e-06,
+      "loss": 0.0657,
+      "step": 8489
+    },
+    {
+      "epoch": 1.6234821684673486,
+      "grad_norm": 2.7125574378335555,
+      "learning_rate": 1e-06,
+      "loss": 0.0654,
+      "step": 8490
+    },
+    {
+      "epoch": 1.623673391337604,
+      "grad_norm": 1.965782290231678,
+      "learning_rate": 1e-06,
+      "loss": 0.0511,
+      "step": 8491
+    },
+    {
+      "epoch": 1.6238646142078592,
+      "grad_norm": 7.415925864037372,
+      "learning_rate": 1e-06,
+      "loss": 0.265,
+      "step": 8492
+    },
+    {
+      "epoch": 1.6240558370781146,
+      "grad_norm": 6.401262826308913,
+      "learning_rate": 1e-06,
+      "loss": 0.306,
+      "step": 8493
+    },
+    {
+      "epoch": 1.6242470599483698,
+      "grad_norm": 5.095338164273376,
+      "learning_rate": 1e-06,
+      "loss": 0.4789,
+      "step": 8494
+    },
+    {
+      "epoch": 1.624438282818625,
+      "grad_norm": 5.178172907035006,
+      "learning_rate": 1e-06,
+      "loss": 0.4535,
+      "step": 8495
+    },
+    {
+      "epoch": 1.6246295056888804,
+      "grad_norm": 6.802532779007585,
+      "learning_rate": 1e-06,
+      "loss": 0.2597,
+      "step": 8496
+    },
+    {
+      "epoch": 1.6248207285591356,
+      "grad_norm": 3.548045095619462,
+      "learning_rate": 1e-06,
+      "loss": 0.2459,
+      "step": 8497
+    },
+    {
+      "epoch": 1.625011951429391,
+      "grad_norm": 4.868929482360493,
+      "learning_rate": 1e-06,
+      "loss": 0.4234,
+      "step": 8498
+    },
+    {
+      "epoch": 1.6252031742996462,
+      "grad_norm": 3.7873745334100826,
+      "learning_rate": 1e-06,
+      "loss": 0.2048,
+      "step": 8499
+    },
+    {
+      "epoch": 1.6253943971699014,
+      "grad_norm": 5.603321092007118,
+      "learning_rate": 1e-06,
+      "loss": 0.0675,
+      "step": 8500
+    },
+    {
+      "epoch": 1.6253943971699014,
+      "eval_runtime": 748.5032,
+      "eval_samples_per_second": 2.049,
+      "eval_steps_per_second": 0.513,
+      "step": 8500
+    },
+    {
+      "epoch": 1.6255856200401568,
+      "grad_norm": 2.53231563036329,
+      "learning_rate": 1e-06,
+      "loss": 0.0653,
+      "step": 8501
+    },
+    {
+      "epoch": 1.6257768429104122,
+      "grad_norm": 3.350076412638974,
+      "learning_rate": 1e-06,
+      "loss": 0.0434,
+      "step": 8502
+    },
+    {
+      "epoch": 1.6259680657806674,
+      "grad_norm": 15.581965639746553,
+      "learning_rate": 1e-06,
+      "loss": 0.0758,
+      "step": 8503
+    },
+    {
+      "epoch": 1.6261592886509226,
+      "grad_norm": 3.894763507249803,
+      "learning_rate": 1e-06,
+      "loss": 0.0724,
+      "step": 8504
+    },
+    {
+      "epoch": 1.6263505115211778,
+      "grad_norm": 7.983105702297367,
+      "learning_rate": 1e-06,
+      "loss": 0.1596,
+      "step": 8505
+    },
+    {
+      "epoch": 1.6265417343914332,
+      "grad_norm": 3.650380305504742,
+      "learning_rate": 1e-06,
+      "loss": 0.2725,
+      "step": 8506
+    },
+    {
+      "epoch": 1.6267329572616887,
+      "grad_norm": 3.9981161688319036,
+      "learning_rate": 1e-06,
+      "loss": 0.2739,
+      "step": 8507
+    },
+    {
+      "epoch": 1.6269241801319438,
+      "grad_norm": 3.6741941230170845,
+      "learning_rate": 1e-06,
+      "loss": 0.2039,
+      "step": 8508
+    },
+    {
+      "epoch": 1.627115403002199,
+      "grad_norm": 2.1248643495234543,
+      "learning_rate": 1e-06,
+      "loss": 0.0564,
+      "step": 8509
+    },
+    {
+      "epoch": 1.6273066258724542,
+      "grad_norm": 4.886469138621485,
+      "learning_rate": 1e-06,
+      "loss": 0.4645,
+      "step": 8510
+    },
+    {
+      "epoch": 1.6274978487427096,
+      "grad_norm": 4.026735128448075,
+      "learning_rate": 1e-06,
+      "loss": 0.2402,
+      "step": 8511
+    },
+    {
+      "epoch": 1.627689071612965,
+      "grad_norm": 4.915465429084156,
+      "learning_rate": 1e-06,
+      "loss": 0.0871,
+      "step": 8512
+    },
+    {
+      "epoch": 1.6278802944832202,
+      "grad_norm": 1.6796542142187039,
+      "learning_rate": 1e-06,
+      "loss": 0.0851,
+      "step": 8513
+    },
+    {
+      "epoch": 1.6280715173534754,
+      "grad_norm": 1.933564465233572,
+      "learning_rate": 1e-06,
+      "loss": 0.04,
+      "step": 8514
+    },
+    {
+      "epoch": 1.6282627402237306,
+      "grad_norm": 3.2851973269113692,
+      "learning_rate": 1e-06,
+      "loss": 0.1228,
+      "step": 8515
+    },
+    {
+      "epoch": 1.628453963093986,
+      "grad_norm": 2.770399207809207,
+      "learning_rate": 1e-06,
+      "loss": 0.0608,
+      "step": 8516
+    },
+    {
+      "epoch": 1.6286451859642415,
+      "grad_norm": 3.331513734723067,
+      "learning_rate": 1e-06,
+      "loss": 0.079,
+      "step": 8517
+    },
+    {
+      "epoch": 1.6288364088344967,
+      "grad_norm": 5.1827979781390425,
+      "learning_rate": 1e-06,
+      "loss": 0.2521,
+      "step": 8518
+    },
+    {
+      "epoch": 1.6290276317047518,
+      "grad_norm": 4.569480259428701,
+      "learning_rate": 1e-06,
+      "loss": 0.4772,
+      "step": 8519
+    },
+    {
+      "epoch": 1.629218854575007,
+      "grad_norm": 5.9706168725369615,
+      "learning_rate": 1e-06,
+      "loss": 0.4241,
+      "step": 8520
+    },
+    {
+      "epoch": 1.6294100774452625,
+      "grad_norm": 2.094091472963264,
+      "learning_rate": 1e-06,
+      "loss": 0.0755,
+      "step": 8521
+    },
+    {
+      "epoch": 1.6296013003155179,
+      "grad_norm": 1.972231919379714,
+      "learning_rate": 1e-06,
+      "loss": 0.0938,
+      "step": 8522
+    },
+    {
+      "epoch": 1.629792523185773,
+      "grad_norm": 1.635885604720256,
+      "learning_rate": 1e-06,
+      "loss": 0.0564,
+      "step": 8523
+    },
+    {
+      "epoch": 1.6299837460560282,
+      "grad_norm": 3.1867689902611454,
+      "learning_rate": 1e-06,
+      "loss": 0.1098,
+      "step": 8524
+    },
+    {
+      "epoch": 1.6301749689262834,
+      "grad_norm": 2.154134292900367,
+      "learning_rate": 1e-06,
+      "loss": 0.0471,
+      "step": 8525
+    },
+    {
+      "epoch": 1.6303661917965389,
+      "grad_norm": 2.820974932619056,
+      "learning_rate": 1e-06,
+      "loss": 0.0659,
+      "step": 8526
+    },
+    {
+      "epoch": 1.6305574146667943,
+      "grad_norm": 6.137345930912123,
+      "learning_rate": 1e-06,
+      "loss": 0.0918,
+      "step": 8527
+    },
+    {
+      "epoch": 1.6307486375370495,
+      "grad_norm": 5.0370955539142495,
+      "learning_rate": 1e-06,
+      "loss": 0.0971,
+      "step": 8528
+    },
+    {
+      "epoch": 1.6309398604073047,
+      "grad_norm": 2.1912315149392563,
+      "learning_rate": 1e-06,
+      "loss": 0.0712,
+      "step": 8529
+    },
+    {
+      "epoch": 1.6311310832775598,
+      "grad_norm": 5.408673222955408,
+      "learning_rate": 1e-06,
+      "loss": 0.0646,
+      "step": 8530
+    },
+    {
+      "epoch": 1.6313223061478153,
+      "grad_norm": 6.580143344790316,
+      "learning_rate": 1e-06,
+      "loss": 0.1563,
+      "step": 8531
+    },
+    {
+      "epoch": 1.6315135290180707,
+      "grad_norm": 4.828983563870541,
+      "learning_rate": 1e-06,
+      "loss": 0.4645,
+      "step": 8532
+    },
+    {
+      "epoch": 1.6317047518883259,
+      "grad_norm": 4.046193186938838,
+      "learning_rate": 1e-06,
+      "loss": 0.3577,
+      "step": 8533
+    },
+    {
+      "epoch": 1.631895974758581,
+      "grad_norm": 4.283468520904534,
+      "learning_rate": 1e-06,
+      "loss": 0.2294,
+      "step": 8534
+    },
+    {
+      "epoch": 1.6320871976288363,
+      "grad_norm": 2.4784797443411013,
+      "learning_rate": 1e-06,
+      "loss": 0.0761,
+      "step": 8535
+    },
+    {
+      "epoch": 1.6322784204990917,
+      "grad_norm": 3.4961888135207455,
+      "learning_rate": 1e-06,
+      "loss": 0.3625,
+      "step": 8536
+    },
+    {
+      "epoch": 1.632469643369347,
+      "grad_norm": 4.67215424280435,
+      "learning_rate": 1e-06,
+      "loss": 0.3106,
+      "step": 8537
+    },
+    {
+      "epoch": 1.6326608662396023,
+      "grad_norm": 1.940419612219062,
+      "learning_rate": 1e-06,
+      "loss": 0.05,
+      "step": 8538
+    },
+    {
+      "epoch": 1.6328520891098575,
+      "grad_norm": 8.451518754075405,
+      "learning_rate": 1e-06,
+      "loss": 0.1029,
+      "step": 8539
+    },
+    {
+      "epoch": 1.6330433119801127,
+      "grad_norm": 3.820474882516818,
+      "learning_rate": 1e-06,
+      "loss": 0.2143,
+      "step": 8540
+    },
+    {
+      "epoch": 1.633234534850368,
+      "grad_norm": 1.1257272594882037,
+      "learning_rate": 1e-06,
+      "loss": 0.0152,
+      "step": 8541
+    },
+    {
+      "epoch": 1.6334257577206235,
+      "grad_norm": 2.9048868899886497,
+      "learning_rate": 1e-06,
+      "loss": 0.045,
+      "step": 8542
+    },
+    {
+      "epoch": 1.6336169805908787,
+      "grad_norm": 6.98745038555731,
+      "learning_rate": 1e-06,
+      "loss": 0.3351,
+      "step": 8543
+    },
+    {
+      "epoch": 1.6338082034611339,
+      "grad_norm": 4.322561054846534,
+      "learning_rate": 1e-06,
+      "loss": 0.3107,
+      "step": 8544
+    },
+    {
+      "epoch": 1.6339994263313893,
+      "grad_norm": 3.157280026805284,
+      "learning_rate": 1e-06,
+      "loss": 0.2404,
+      "step": 8545
+    },
+    {
+      "epoch": 1.6341906492016445,
+      "grad_norm": 3.942492030201599,
+      "learning_rate": 1e-06,
+      "loss": 0.4523,
+      "step": 8546
+    },
+    {
+      "epoch": 1.6343818720718999,
+      "grad_norm": 4.914684456429977,
+      "learning_rate": 1e-06,
+      "loss": 0.4157,
+      "step": 8547
+    },
+    {
+      "epoch": 1.634573094942155,
+      "grad_norm": 3.0465288014455254,
+      "learning_rate": 1e-06,
+      "loss": 0.0705,
+      "step": 8548
+    },
+    {
+      "epoch": 1.6347643178124103,
+      "grad_norm": 3.6479209517899114,
+      "learning_rate": 1e-06,
+      "loss": 0.1347,
+      "step": 8549
+    },
+    {
+      "epoch": 1.6349555406826657,
+      "grad_norm": 3.848220582425204,
+      "learning_rate": 1e-06,
+      "loss": 0.3,
+      "step": 8550
+    },
+    {
+      "epoch": 1.6351467635529209,
+      "grad_norm": 3.613347893177648,
+      "learning_rate": 1e-06,
+      "loss": 0.0329,
+      "step": 8551
+    },
+    {
+      "epoch": 1.6353379864231763,
+      "grad_norm": 5.010204392640112,
+      "learning_rate": 1e-06,
+      "loss": 0.197,
+      "step": 8552
+    },
+    {
+      "epoch": 1.6355292092934315,
+      "grad_norm": 3.7110645794441757,
+      "learning_rate": 1e-06,
+      "loss": 0.0686,
+      "step": 8553
+    },
+    {
+      "epoch": 1.6357204321636867,
+      "grad_norm": 2.3507088667481764,
+      "learning_rate": 1e-06,
+      "loss": 0.0682,
+      "step": 8554
+    },
+    {
+      "epoch": 1.635911655033942,
+      "grad_norm": 3.3091704711533554,
+      "learning_rate": 1e-06,
+      "loss": 0.0695,
+      "step": 8555
+    },
+    {
+      "epoch": 1.6361028779041975,
+      "grad_norm": 9.311383686107149,
+      "learning_rate": 1e-06,
+      "loss": 0.4167,
+      "step": 8556
+    },
+    {
+      "epoch": 1.6362941007744527,
+      "grad_norm": 8.220976669317395,
+      "learning_rate": 1e-06,
+      "loss": 0.468,
+      "step": 8557
+    },
+    {
+      "epoch": 1.636485323644708,
+      "grad_norm": 2.5281903648649746,
+      "learning_rate": 1e-06,
+      "loss": 0.0559,
+      "step": 8558
+    },
+    {
+      "epoch": 1.636676546514963,
+      "grad_norm": 2.0637578885293304,
+      "learning_rate": 1e-06,
+      "loss": 0.1838,
+      "step": 8559
+    },
+    {
+      "epoch": 1.6368677693852185,
+      "grad_norm": 4.3582670383018565,
+      "learning_rate": 1e-06,
+      "loss": 0.2689,
+      "step": 8560
+    },
+    {
+      "epoch": 1.637058992255474,
+      "grad_norm": 3.2914698256213857,
+      "learning_rate": 1e-06,
+      "loss": 0.1724,
+      "step": 8561
+    },
+    {
+      "epoch": 1.637250215125729,
+      "grad_norm": 3.045775073147,
+      "learning_rate": 1e-06,
+      "loss": 0.1357,
+      "step": 8562
+    },
+    {
+      "epoch": 1.6374414379959843,
+      "grad_norm": 2.9166497002971004,
+      "learning_rate": 1e-06,
+      "loss": 0.0928,
+      "step": 8563
+    },
+    {
+      "epoch": 1.6376326608662395,
+      "grad_norm": 3.4555524346302438,
+      "learning_rate": 1e-06,
+      "loss": 0.07,
+      "step": 8564
+    },
+    {
+      "epoch": 1.637823883736495,
+      "grad_norm": 1.7368334280600377,
+      "learning_rate": 1e-06,
+      "loss": 0.02,
+      "step": 8565
+    },
+    {
+      "epoch": 1.6380151066067503,
+      "grad_norm": 2.394939543373142,
+      "learning_rate": 1e-06,
+      "loss": 0.03,
+      "step": 8566
+    },
+    {
+      "epoch": 1.6382063294770055,
+      "grad_norm": 3.6129352135824337,
+      "learning_rate": 1e-06,
+      "loss": 0.0655,
+      "step": 8567
+    },
+    {
+      "epoch": 1.6383975523472607,
+      "grad_norm": 4.307511249866286,
+      "learning_rate": 1e-06,
+      "loss": 0.5739,
+      "step": 8568
+    },
+    {
+      "epoch": 1.638588775217516,
+      "grad_norm": 2.36818107677522,
+      "learning_rate": 1e-06,
+      "loss": 0.1399,
+      "step": 8569
+    },
+    {
+      "epoch": 1.6387799980877713,
+      "grad_norm": 3.8714225005854828,
+      "learning_rate": 1e-06,
+      "loss": 0.3258,
+      "step": 8570
+    },
+    {
+      "epoch": 1.6389712209580267,
+      "grad_norm": 2.6199512976567916,
+      "learning_rate": 1e-06,
+      "loss": 0.0814,
+      "step": 8571
+    },
+    {
+      "epoch": 1.639162443828282,
+      "grad_norm": 3.4315134805433707,
+      "learning_rate": 1e-06,
+      "loss": 0.2373,
+      "step": 8572
+    },
+    {
+      "epoch": 1.639353666698537,
+      "grad_norm": 3.921689580052516,
+      "learning_rate": 1e-06,
+      "loss": 0.2686,
+      "step": 8573
+    },
+    {
+      "epoch": 1.6395448895687923,
+      "grad_norm": 3.1066895202756624,
+      "learning_rate": 1e-06,
+      "loss": 0.1604,
+      "step": 8574
+    },
+    {
+      "epoch": 1.6397361124390477,
+      "grad_norm": 3.161191942700758,
+      "learning_rate": 1e-06,
+      "loss": 0.1379,
+      "step": 8575
+    },
+    {
+      "epoch": 1.6399273353093031,
+      "grad_norm": 2.987960658609799,
+      "learning_rate": 1e-06,
+      "loss": 0.068,
+      "step": 8576
+    },
+    {
+      "epoch": 1.6401185581795583,
+      "grad_norm": 2.341126970519956,
+      "learning_rate": 1e-06,
+      "loss": 0.0965,
+      "step": 8577
+    },
+    {
+      "epoch": 1.6403097810498135,
+      "grad_norm": 2.1117289654043288,
+      "learning_rate": 1e-06,
+      "loss": 0.0522,
+      "step": 8578
+    },
+    {
+      "epoch": 1.6405010039200687,
+      "grad_norm": 7.10199758877575,
+      "learning_rate": 1e-06,
+      "loss": 0.0878,
+      "step": 8579
+    },
+    {
+      "epoch": 1.6406922267903241,
+      "grad_norm": 7.4411991440995875,
+      "learning_rate": 1e-06,
+      "loss": 0.1005,
+      "step": 8580
+    },
+    {
+      "epoch": 1.6408834496605795,
+      "grad_norm": 6.96230986796119,
+      "learning_rate": 1e-06,
+      "loss": 0.2643,
+      "step": 8581
+    },
+    {
+      "epoch": 1.6410746725308347,
+      "grad_norm": 5.220574887945247,
+      "learning_rate": 1e-06,
+      "loss": 0.4591,
+      "step": 8582
+    },
+    {
+      "epoch": 1.64126589540109,
+      "grad_norm": 1.5634850258154367,
+      "learning_rate": 1e-06,
+      "loss": 0.0736,
+      "step": 8583
+    },
+    {
+      "epoch": 1.641457118271345,
+      "grad_norm": 2.045253664806828,
+      "learning_rate": 1e-06,
+      "loss": 0.1361,
+      "step": 8584
+    },
+    {
+      "epoch": 1.6416483411416005,
+      "grad_norm": 2.010128718757309,
+      "learning_rate": 1e-06,
+      "loss": 0.064,
+      "step": 8585
+    },
+    {
+      "epoch": 1.641839564011856,
+      "grad_norm": 3.3638208113765224,
+      "learning_rate": 1e-06,
+      "loss": 0.1541,
+      "step": 8586
+    },
+    {
+      "epoch": 1.6420307868821111,
+      "grad_norm": 2.5507129233586667,
+      "learning_rate": 1e-06,
+      "loss": 0.0866,
+      "step": 8587
+    },
+    {
+      "epoch": 1.6422220097523663,
+      "grad_norm": 2.7538689360559405,
+      "learning_rate": 1e-06,
+      "loss": 0.1018,
+      "step": 8588
+    },
+    {
+      "epoch": 1.6424132326226215,
+      "grad_norm": 2.969597304045341,
+      "learning_rate": 1e-06,
+      "loss": 0.1457,
+      "step": 8589
+    },
+    {
+      "epoch": 1.642604455492877,
+      "grad_norm": 7.015778877592804,
+      "learning_rate": 1e-06,
+      "loss": 0.2658,
+      "step": 8590
+    },
+    {
+      "epoch": 1.6427956783631323,
+      "grad_norm": 6.2012158586038595,
+      "learning_rate": 1e-06,
+      "loss": 0.1271,
+      "step": 8591
+    },
+    {
+      "epoch": 1.6429869012333875,
+      "grad_norm": 8.055012856395651,
+      "learning_rate": 1e-06,
+      "loss": 0.1653,
+      "step": 8592
+    },
+    {
+      "epoch": 1.6431781241036427,
+      "grad_norm": 6.061553881252089,
+      "learning_rate": 1e-06,
+      "loss": 0.2087,
+      "step": 8593
+    },
+    {
+      "epoch": 1.643369346973898,
+      "grad_norm": 7.3070642857828,
+      "learning_rate": 1e-06,
+      "loss": 0.484,
+      "step": 8594
+    },
+    {
+      "epoch": 1.6435605698441533,
+      "grad_norm": 3.266219673585309,
+      "learning_rate": 1e-06,
+      "loss": 0.1299,
+      "step": 8595
+    },
+    {
+      "epoch": 1.6437517927144087,
+      "grad_norm": 2.5437041594852703,
+      "learning_rate": 1e-06,
+      "loss": 0.0917,
+      "step": 8596
+    },
+    {
+      "epoch": 1.643943015584664,
+      "grad_norm": 1.9748525347160764,
+      "learning_rate": 1e-06,
+      "loss": 0.0682,
+      "step": 8597
+    },
+    {
+      "epoch": 1.6441342384549191,
+      "grad_norm": 4.8907667633082905,
+      "learning_rate": 1e-06,
+      "loss": 0.1924,
+      "step": 8598
+    },
+    {
+      "epoch": 1.6443254613251745,
+      "grad_norm": 3.3020456273827263,
+      "learning_rate": 1e-06,
+      "loss": 0.1055,
+      "step": 8599
+    },
+    {
+      "epoch": 1.6445166841954297,
+      "grad_norm": 3.937111971672921,
+      "learning_rate": 1e-06,
+      "loss": 0.0461,
+      "step": 8600
+    },
+    {
+      "epoch": 1.6447079070656851,
+      "grad_norm": 2.3063350672139427,
+      "learning_rate": 1e-06,
+      "loss": 0.0684,
+      "step": 8601
+    },
+    {
+      "epoch": 1.6448991299359403,
+      "grad_norm": 3.974400500308572,
+      "learning_rate": 1e-06,
+      "loss": 0.1725,
+      "step": 8602
+    },
+    {
+      "epoch": 1.6450903528061955,
+      "grad_norm": 1.6599220739021492,
+      "learning_rate": 1e-06,
+      "loss": 0.0319,
+      "step": 8603
+    },
+    {
+      "epoch": 1.645281575676451,
+      "grad_norm": 3.958121056469655,
+      "learning_rate": 1e-06,
+      "loss": 0.0993,
+      "step": 8604
+    },
+    {
+      "epoch": 1.6454727985467061,
+      "grad_norm": 3.244685743306267,
+      "learning_rate": 1e-06,
+      "loss": 0.1359,
+      "step": 8605
+    },
+    {
+      "epoch": 1.6456640214169616,
+      "grad_norm": 5.400747685415737,
+      "learning_rate": 1e-06,
+      "loss": 0.2885,
+      "step": 8606
+    },
+    {
+      "epoch": 1.6458552442872167,
+      "grad_norm": 4.944404313761602,
+      "learning_rate": 1e-06,
+      "loss": 0.423,
+      "step": 8607
+    },
+    {
+      "epoch": 1.646046467157472,
+      "grad_norm": 3.2687873677730845,
+      "learning_rate": 1e-06,
+      "loss": 0.2067,
+      "step": 8608
+    },
+    {
+      "epoch": 1.6462376900277274,
+      "grad_norm": 3.0497115014381926,
+      "learning_rate": 1e-06,
+      "loss": 0.1777,
+      "step": 8609
+    },
+    {
+      "epoch": 1.6464289128979825,
+      "grad_norm": 3.3295456983641354,
+      "learning_rate": 1e-06,
+      "loss": 0.3234,
+      "step": 8610
+    },
+    {
+      "epoch": 1.646620135768238,
+      "grad_norm": 4.259830325692154,
+      "learning_rate": 1e-06,
+      "loss": 0.3671,
+      "step": 8611
+    },
+    {
+      "epoch": 1.6468113586384932,
+      "grad_norm": 2.36309702762089,
+      "learning_rate": 1e-06,
+      "loss": 0.072,
+      "step": 8612
+    },
+    {
+      "epoch": 1.6470025815087483,
+      "grad_norm": 1.7813420188291367,
+      "learning_rate": 1e-06,
+      "loss": 0.0286,
+      "step": 8613
+    },
+    {
+      "epoch": 1.6471938043790038,
+      "grad_norm": 3.371034482432728,
+      "learning_rate": 1e-06,
+      "loss": 0.1142,
+      "step": 8614
+    },
+    {
+      "epoch": 1.6473850272492592,
+      "grad_norm": 3.5332720883177746,
+      "learning_rate": 1e-06,
+      "loss": 0.1642,
+      "step": 8615
+    },
+    {
+      "epoch": 1.6475762501195144,
+      "grad_norm": 3.9265869528467854,
+      "learning_rate": 1e-06,
+      "loss": 0.0616,
+      "step": 8616
+    },
+    {
+      "epoch": 1.6477674729897696,
+      "grad_norm": 4.816179528511628,
+      "learning_rate": 1e-06,
+      "loss": 0.0449,
+      "step": 8617
+    },
+    {
+      "epoch": 1.6479586958600247,
+      "grad_norm": 8.361265392999314,
+      "learning_rate": 1e-06,
+      "loss": 0.3003,
+      "step": 8618
+    },
+    {
+      "epoch": 1.6481499187302802,
+      "grad_norm": 4.147640862995922,
+      "learning_rate": 1e-06,
+      "loss": 0.3194,
+      "step": 8619
+    },
+    {
+      "epoch": 1.6483411416005356,
+      "grad_norm": 5.161014644313023,
+      "learning_rate": 1e-06,
+      "loss": 0.3774,
+      "step": 8620
+    },
+    {
+      "epoch": 1.6485323644707908,
+      "grad_norm": 1.2999660010660326,
+      "learning_rate": 1e-06,
+      "loss": 0.0892,
+      "step": 8621
+    },
+    {
+      "epoch": 1.648723587341046,
+      "grad_norm": 2.319316920152297,
+      "learning_rate": 1e-06,
+      "loss": 0.1042,
+      "step": 8622
+    },
+    {
+      "epoch": 1.6489148102113012,
+      "grad_norm": 2.841173618832301,
+      "learning_rate": 1e-06,
+      "loss": 0.169,
+      "step": 8623
+    },
+    {
+      "epoch": 1.6491060330815566,
+      "grad_norm": 2.6151936238014812,
+      "learning_rate": 1e-06,
+      "loss": 0.0603,
+      "step": 8624
+    },
+    {
+      "epoch": 1.649297255951812,
+      "grad_norm": 3.21882599676098,
+      "learning_rate": 1e-06,
+      "loss": 0.0771,
+      "step": 8625
+    },
+    {
+      "epoch": 1.6494884788220672,
+      "grad_norm": 2.3904692156528364,
+      "learning_rate": 1e-06,
+      "loss": 0.0697,
+      "step": 8626
+    },
+    {
+      "epoch": 1.6496797016923224,
+      "grad_norm": 2.8705125700571243,
+      "learning_rate": 1e-06,
+      "loss": 0.0702,
+      "step": 8627
+    },
+    {
+      "epoch": 1.6498709245625776,
+      "grad_norm": 1.6775684563303037,
+      "learning_rate": 1e-06,
+      "loss": 0.0635,
+      "step": 8628
+    },
+    {
+      "epoch": 1.650062147432833,
+      "grad_norm": 3.0151985143591546,
+      "learning_rate": 1e-06,
+      "loss": 0.0717,
+      "step": 8629
+    },
+    {
+      "epoch": 1.6502533703030884,
+      "grad_norm": 5.2585066043862705,
+      "learning_rate": 1e-06,
+      "loss": 0.0726,
+      "step": 8630
+    },
+    {
+      "epoch": 1.6504445931733436,
+      "grad_norm": 6.187406597010456,
+      "learning_rate": 1e-06,
+      "loss": 0.3333,
+      "step": 8631
+    },
+    {
+      "epoch": 1.6506358160435988,
+      "grad_norm": 4.281316826291865,
+      "learning_rate": 1e-06,
+      "loss": 0.3392,
+      "step": 8632
+    },
+    {
+      "epoch": 1.650827038913854,
+      "grad_norm": 6.230171794451491,
+      "learning_rate": 1e-06,
+      "loss": 0.2437,
+      "step": 8633
+    },
+    {
+      "epoch": 1.6510182617841094,
+      "grad_norm": 2.7039331540078377,
+      "learning_rate": 1e-06,
+      "loss": 0.1828,
+      "step": 8634
+    },
+    {
+      "epoch": 1.6512094846543648,
+      "grad_norm": 3.258153371564813,
+      "learning_rate": 1e-06,
+      "loss": 0.2284,
+      "step": 8635
+    },
+    {
+      "epoch": 1.65140070752462,
+      "grad_norm": 2.587570274246444,
+      "learning_rate": 1e-06,
+      "loss": 0.0961,
+      "step": 8636
+    },
+    {
+      "epoch": 1.6515919303948752,
+      "grad_norm": 4.520837646107294,
+      "learning_rate": 1e-06,
+      "loss": 0.3458,
+      "step": 8637
+    },
+    {
+      "epoch": 1.6517831532651304,
+      "grad_norm": 4.043615729860637,
+      "learning_rate": 1e-06,
+      "loss": 0.1524,
+      "step": 8638
+    },
+    {
+      "epoch": 1.6519743761353858,
+      "grad_norm": 3.0769290052870346,
+      "learning_rate": 1e-06,
+      "loss": 0.1898,
+      "step": 8639
+    },
+    {
+      "epoch": 1.6521655990056412,
+      "grad_norm": 2.5579954423304585,
+      "learning_rate": 1e-06,
+      "loss": 0.0818,
+      "step": 8640
+    },
+    {
+      "epoch": 1.6523568218758964,
+      "grad_norm": 4.3873932915125335,
+      "learning_rate": 1e-06,
+      "loss": 0.1523,
+      "step": 8641
+    },
+    {
+      "epoch": 1.6525480447461516,
+      "grad_norm": 3.9883771356158535,
+      "learning_rate": 1e-06,
+      "loss": 0.1313,
+      "step": 8642
+    },
+    {
+      "epoch": 1.6527392676164068,
+      "grad_norm": 3.6735987736174898,
+      "learning_rate": 1e-06,
+      "loss": 0.2059,
+      "step": 8643
+    },
+    {
+      "epoch": 1.6529304904866622,
+      "grad_norm": 7.914968877407053,
+      "learning_rate": 1e-06,
+      "loss": 0.1514,
+      "step": 8644
+    },
+    {
+      "epoch": 1.6531217133569176,
+      "grad_norm": 4.858740467703036,
+      "learning_rate": 1e-06,
+      "loss": 0.3481,
+      "step": 8645
+    },
+    {
+      "epoch": 1.6533129362271728,
+      "grad_norm": 3.741628585202504,
+      "learning_rate": 1e-06,
+      "loss": 0.2035,
+      "step": 8646
+    },
+    {
+      "epoch": 1.653504159097428,
+      "grad_norm": 3.3186073861667498,
+      "learning_rate": 1e-06,
+      "loss": 0.1612,
+      "step": 8647
+    },
+    {
+      "epoch": 1.6536953819676832,
+      "grad_norm": 3.176004512207131,
+      "learning_rate": 1e-06,
+      "loss": 0.0533,
+      "step": 8648
+    },
+    {
+      "epoch": 1.6538866048379386,
+      "grad_norm": 2.344073362296054,
+      "learning_rate": 1e-06,
+      "loss": 0.0946,
+      "step": 8649
+    },
+    {
+      "epoch": 1.654077827708194,
+      "grad_norm": 3.759462244253398,
+      "learning_rate": 1e-06,
+      "loss": 0.0855,
+      "step": 8650
+    },
+    {
+      "epoch": 1.6542690505784492,
+      "grad_norm": 2.498648564320268,
+      "learning_rate": 1e-06,
+      "loss": 0.067,
+      "step": 8651
+    },
+    {
+      "epoch": 1.6544602734487044,
+      "grad_norm": 2.4082599518715946,
+      "learning_rate": 1e-06,
+      "loss": 0.0437,
+      "step": 8652
+    },
+    {
+      "epoch": 1.6546514963189598,
+      "grad_norm": 4.8206901711473105,
+      "learning_rate": 1e-06,
+      "loss": 0.1731,
+      "step": 8653
+    },
+    {
+      "epoch": 1.654842719189215,
+      "grad_norm": 3.2240335006866485,
+      "learning_rate": 1e-06,
+      "loss": 0.0396,
+      "step": 8654
+    },
+    {
+      "epoch": 1.6550339420594704,
+      "grad_norm": 2.7536160363483444,
+      "learning_rate": 1e-06,
+      "loss": 0.0317,
+      "step": 8655
+    },
+    {
+      "epoch": 1.6552251649297256,
+      "grad_norm": 3.427017041215472,
+      "learning_rate": 1e-06,
+      "loss": 0.2844,
+      "step": 8656
+    },
+    {
+      "epoch": 1.6554163877999808,
+      "grad_norm": 4.442919792816475,
+      "learning_rate": 1e-06,
+      "loss": 0.2504,
+      "step": 8657
+    },
+    {
+      "epoch": 1.6556076106702362,
+      "grad_norm": 3.9663442694386095,
+      "learning_rate": 1e-06,
+      "loss": 0.4001,
+      "step": 8658
+    },
+    {
+      "epoch": 1.6557988335404914,
+      "grad_norm": 5.033282325604066,
+      "learning_rate": 1e-06,
+      "loss": 0.2079,
+      "step": 8659
+    },
+    {
+      "epoch": 1.6559900564107468,
+      "grad_norm": 3.731270876623182,
+      "learning_rate": 1e-06,
+      "loss": 0.2762,
+      "step": 8660
+    },
+    {
+      "epoch": 1.656181279281002,
+      "grad_norm": 4.421215426045858,
+      "learning_rate": 1e-06,
+      "loss": 0.211,
+      "step": 8661
+    },
+    {
+      "epoch": 1.6563725021512572,
+      "grad_norm": 2.847569845011145,
+      "learning_rate": 1e-06,
+      "loss": 0.045,
+      "step": 8662
+    },
+    {
+      "epoch": 1.6565637250215126,
+      "grad_norm": 2.7361298951495527,
+      "learning_rate": 1e-06,
+      "loss": 0.1065,
+      "step": 8663
+    },
+    {
+      "epoch": 1.6567549478917678,
+      "grad_norm": 2.792480596818169,
+      "learning_rate": 1e-06,
+      "loss": 0.0598,
+      "step": 8664
+    },
+    {
+      "epoch": 1.6569461707620232,
+      "grad_norm": 3.6817786881440444,
+      "learning_rate": 1e-06,
+      "loss": 0.1935,
+      "step": 8665
+    },
+    {
+      "epoch": 1.6571373936322784,
+      "grad_norm": 1.6439807541317035,
+      "learning_rate": 1e-06,
+      "loss": 0.0208,
+      "step": 8666
+    },
+    {
+      "epoch": 1.6573286165025336,
+      "grad_norm": 5.977929531158107,
+      "learning_rate": 1e-06,
+      "loss": 0.1572,
+      "step": 8667
+    },
+    {
+      "epoch": 1.657519839372789,
+      "grad_norm": 6.849393285905677,
+      "learning_rate": 1e-06,
+      "loss": 0.2205,
+      "step": 8668
+    },
+    {
+      "epoch": 1.6577110622430444,
+      "grad_norm": 5.090718683793998,
+      "learning_rate": 1e-06,
+      "loss": 0.3544,
+      "step": 8669
+    },
+    {
+      "epoch": 1.6579022851132996,
+      "grad_norm": 4.24152494142559,
+      "learning_rate": 1e-06,
+      "loss": 0.1052,
+      "step": 8670
+    },
+    {
+      "epoch": 1.6580935079835548,
+      "grad_norm": 5.106052823502845,
+      "learning_rate": 1e-06,
+      "loss": 0.1822,
+      "step": 8671
+    },
+    {
+      "epoch": 1.65828473085381,
+      "grad_norm": 5.622687309405148,
+      "learning_rate": 1e-06,
+      "loss": 0.2665,
+      "step": 8672
+    },
+    {
+      "epoch": 1.6584759537240654,
+      "grad_norm": 2.12217805521322,
+      "learning_rate": 1e-06,
+      "loss": 0.1318,
+      "step": 8673
+    },
+    {
+      "epoch": 1.6586671765943208,
+      "grad_norm": 3.084853510663185,
+      "learning_rate": 1e-06,
+      "loss": 0.085,
+      "step": 8674
+    },
+    {
+      "epoch": 1.658858399464576,
+      "grad_norm": 2.673657898816901,
+      "learning_rate": 1e-06,
+      "loss": 0.1126,
+      "step": 8675
+    },
+    {
+      "epoch": 1.6590496223348312,
+      "grad_norm": 2.1803158090355,
+      "learning_rate": 1e-06,
+      "loss": 0.0688,
+      "step": 8676
+    },
+    {
+      "epoch": 1.6592408452050864,
+      "grad_norm": 4.917679030510139,
+      "learning_rate": 1e-06,
+      "loss": 0.0395,
+      "step": 8677
+    },
+    {
+      "epoch": 1.6594320680753418,
+      "grad_norm": 4.806479316866131,
+      "learning_rate": 1e-06,
+      "loss": 0.0738,
+      "step": 8678
+    },
+    {
+      "epoch": 1.6596232909455972,
+      "grad_norm": 4.1823370690769295,
+      "learning_rate": 1e-06,
+      "loss": 0.0621,
+      "step": 8679
+    },
+    {
+      "epoch": 1.6598145138158524,
+      "grad_norm": 9.049053028785577,
+      "learning_rate": 1e-06,
+      "loss": 0.2442,
+      "step": 8680
+    },
+    {
+      "epoch": 1.6600057366861076,
+      "grad_norm": 6.900749950841083,
+      "learning_rate": 1e-06,
+      "loss": 0.2216,
+      "step": 8681
+    },
+    {
+      "epoch": 1.6601969595563628,
+      "grad_norm": 3.2129232758599295,
+      "learning_rate": 1e-06,
+      "loss": 0.2502,
+      "step": 8682
+    },
+    {
+      "epoch": 1.6603881824266182,
+      "grad_norm": 5.876081001985467,
+      "learning_rate": 1e-06,
+      "loss": 0.0848,
+      "step": 8683
+    },
+    {
+      "epoch": 1.6605794052968736,
+      "grad_norm": 4.33622957912639,
+      "learning_rate": 1e-06,
+      "loss": 0.2919,
+      "step": 8684
+    },
+    {
+      "epoch": 1.6607706281671288,
+      "grad_norm": 1.65738315325765,
+      "learning_rate": 1e-06,
+      "loss": 0.0351,
+      "step": 8685
+    },
+    {
+      "epoch": 1.660961851037384,
+      "grad_norm": 4.1938730963706945,
+      "learning_rate": 1e-06,
+      "loss": 0.161,
+      "step": 8686
+    },
+    {
+      "epoch": 1.6611530739076392,
+      "grad_norm": 2.8082611353920104,
+      "learning_rate": 1e-06,
+      "loss": 0.0905,
+      "step": 8687
+    },
+    {
+      "epoch": 1.6613442967778946,
+      "grad_norm": 2.956758877141684,
+      "learning_rate": 1e-06,
+      "loss": 0.0965,
+      "step": 8688
+    },
+    {
+      "epoch": 1.66153551964815,
+      "grad_norm": 3.7637396560369707,
+      "learning_rate": 1e-06,
+      "loss": 0.2169,
+      "step": 8689
+    },
+    {
+      "epoch": 1.6617267425184052,
+      "grad_norm": 3.2839049724346405,
+      "learning_rate": 1e-06,
+      "loss": 0.1162,
+      "step": 8690
+    },
+    {
+      "epoch": 1.6619179653886604,
+      "grad_norm": 2.3603969154530398,
+      "learning_rate": 1e-06,
+      "loss": 0.0491,
+      "step": 8691
+    },
+    {
+      "epoch": 1.6621091882589156,
+      "grad_norm": 3.031714649572196,
+      "learning_rate": 1e-06,
+      "loss": 0.0634,
+      "step": 8692
+    },
+    {
+      "epoch": 1.662300411129171,
+      "grad_norm": 4.381727223655577,
+      "learning_rate": 1e-06,
+      "loss": 0.1089,
+      "step": 8693
+    },
+    {
+      "epoch": 1.6624916339994265,
+      "grad_norm": 4.1236304119122815,
+      "learning_rate": 1e-06,
+      "loss": 0.1828,
+      "step": 8694
+    },
+    {
+      "epoch": 1.6626828568696816,
+      "grad_norm": 4.176996523493333,
+      "learning_rate": 1e-06,
+      "loss": 0.2738,
+      "step": 8695
+    },
+    {
+      "epoch": 1.6628740797399368,
+      "grad_norm": 2.4857536665141176,
+      "learning_rate": 1e-06,
+      "loss": 0.0937,
+      "step": 8696
+    },
+    {
+      "epoch": 1.663065302610192,
+      "grad_norm": 1.9790881158064546,
+      "learning_rate": 1e-06,
+      "loss": 0.1004,
+      "step": 8697
+    },
+    {
+      "epoch": 1.6632565254804474,
+      "grad_norm": 3.4537990826968636,
+      "learning_rate": 1e-06,
+      "loss": 0.1755,
+      "step": 8698
+    },
+    {
+      "epoch": 1.6634477483507029,
+      "grad_norm": 5.689762880985964,
+      "learning_rate": 1e-06,
+      "loss": 0.2608,
+      "step": 8699
+    },
+    {
+      "epoch": 1.663638971220958,
+      "grad_norm": 5.349429497649014,
+      "learning_rate": 1e-06,
+      "loss": 0.1369,
+      "step": 8700
+    },
+    {
+      "epoch": 1.6638301940912132,
+      "grad_norm": 4.128733708034505,
+      "learning_rate": 1e-06,
+      "loss": 0.0999,
+      "step": 8701
+    },
+    {
+      "epoch": 1.6640214169614684,
+      "grad_norm": 1.3366926412524103,
+      "learning_rate": 1e-06,
+      "loss": 0.0173,
+      "step": 8702
+    },
+    {
+      "epoch": 1.6642126398317239,
+      "grad_norm": 4.655555212424991,
+      "learning_rate": 1e-06,
+      "loss": 0.0809,
+      "step": 8703
+    },
+    {
+      "epoch": 1.6644038627019793,
+      "grad_norm": 2.26526496920488,
+      "learning_rate": 1e-06,
+      "loss": 0.0391,
+      "step": 8704
+    },
+    {
+      "epoch": 1.6645950855722345,
+      "grad_norm": 6.44195224906991,
+      "learning_rate": 1e-06,
+      "loss": 0.1934,
+      "step": 8705
+    },
+    {
+      "epoch": 1.6647863084424896,
+      "grad_norm": 3.115362182674085,
+      "learning_rate": 1e-06,
+      "loss": 0.1201,
+      "step": 8706
+    },
+    {
+      "epoch": 1.6649775313127448,
+      "grad_norm": 3.640401153886514,
+      "learning_rate": 1e-06,
+      "loss": 0.273,
+      "step": 8707
+    },
+    {
+      "epoch": 1.6651687541830003,
+      "grad_norm": 1.7639866882738857,
+      "learning_rate": 1e-06,
+      "loss": 0.1382,
+      "step": 8708
+    },
+    {
+      "epoch": 1.6653599770532557,
+      "grad_norm": 2.664798578949191,
+      "learning_rate": 1e-06,
+      "loss": 0.1621,
+      "step": 8709
+    },
+    {
+      "epoch": 1.6655511999235109,
+      "grad_norm": 2.9484268843273225,
+      "learning_rate": 1e-06,
+      "loss": 0.1311,
+      "step": 8710
+    },
+    {
+      "epoch": 1.665742422793766,
+      "grad_norm": 3.016467831488637,
+      "learning_rate": 1e-06,
+      "loss": 0.1021,
+      "step": 8711
+    },
+    {
+      "epoch": 1.6659336456640215,
+      "grad_norm": 5.03594713532665,
+      "learning_rate": 1e-06,
+      "loss": 0.1065,
+      "step": 8712
+    },
+    {
+      "epoch": 1.6661248685342767,
+      "grad_norm": 1.2816890801568683,
+      "learning_rate": 1e-06,
+      "loss": 0.0444,
+      "step": 8713
+    },
+    {
+      "epoch": 1.666316091404532,
+      "grad_norm": 2.4495722806053624,
+      "learning_rate": 1e-06,
+      "loss": 0.036,
+      "step": 8714
+    },
+    {
+      "epoch": 1.6665073142747873,
+      "grad_norm": 3.6734957100991283,
+      "learning_rate": 1e-06,
+      "loss": 0.053,
+      "step": 8715
+    },
+    {
+      "epoch": 1.6666985371450425,
+      "grad_norm": 2.4069857772895076,
+      "learning_rate": 1e-06,
+      "loss": 0.0459,
+      "step": 8716
+    },
+    {
+      "epoch": 1.6668897600152979,
+      "grad_norm": 4.613912747514283,
+      "learning_rate": 1e-06,
+      "loss": 0.0475,
+      "step": 8717
+    },
+    {
+      "epoch": 1.667080982885553,
+      "grad_norm": 4.472127851565959,
+      "learning_rate": 1e-06,
+      "loss": 0.361,
+      "step": 8718
+    },
+    {
+      "epoch": 1.6672722057558085,
+      "grad_norm": 5.050470636173483,
+      "learning_rate": 1e-06,
+      "loss": 0.1379,
+      "step": 8719
+    },
+    {
+      "epoch": 1.6674634286260637,
+      "grad_norm": 3.9475099996641845,
+      "learning_rate": 1e-06,
+      "loss": 0.1774,
+      "step": 8720
+    },
+    {
+      "epoch": 1.6676546514963189,
+      "grad_norm": 4.5333268240339635,
+      "learning_rate": 1e-06,
+      "loss": 0.4564,
+      "step": 8721
+    },
+    {
+      "epoch": 1.6678458743665743,
+      "grad_norm": 3.145992418186682,
+      "learning_rate": 1e-06,
+      "loss": 0.1156,
+      "step": 8722
+    },
+    {
+      "epoch": 1.6680370972368297,
+      "grad_norm": 2.209135041971629,
+      "learning_rate": 1e-06,
+      "loss": 0.076,
+      "step": 8723
+    },
+    {
+      "epoch": 1.6682283201070849,
+      "grad_norm": 2.2442524934380663,
+      "learning_rate": 1e-06,
+      "loss": 0.0366,
+      "step": 8724
+    },
+    {
+      "epoch": 1.66841954297734,
+      "grad_norm": 2.453253967393905,
+      "learning_rate": 1e-06,
+      "loss": 0.0354,
+      "step": 8725
+    },
+    {
+      "epoch": 1.6686107658475953,
+      "grad_norm": 2.795575740524356,
+      "learning_rate": 1e-06,
+      "loss": 0.0804,
+      "step": 8726
+    },
+    {
+      "epoch": 1.6688019887178507,
+      "grad_norm": 4.066049287136092,
+      "learning_rate": 1e-06,
+      "loss": 0.232,
+      "step": 8727
+    },
+    {
+      "epoch": 1.668993211588106,
+      "grad_norm": 1.5823795641589122,
+      "learning_rate": 1e-06,
+      "loss": 0.035,
+      "step": 8728
+    },
+    {
+      "epoch": 1.6691844344583613,
+      "grad_norm": 4.998494684596401,
+      "learning_rate": 1e-06,
+      "loss": 0.053,
+      "step": 8729
+    },
+    {
+      "epoch": 1.6693756573286165,
+      "grad_norm": 7.866467167323593,
+      "learning_rate": 1e-06,
+      "loss": 0.1752,
+      "step": 8730
+    },
+    {
+      "epoch": 1.6695668801988717,
+      "grad_norm": 4.537799810587911,
+      "learning_rate": 1e-06,
+      "loss": 0.3989,
+      "step": 8731
+    },
+    {
+      "epoch": 1.669758103069127,
+      "grad_norm": 2.574660499191338,
+      "learning_rate": 1e-06,
+      "loss": 0.1008,
+      "step": 8732
+    },
+    {
+      "epoch": 1.6699493259393825,
+      "grad_norm": 5.210273483039677,
+      "learning_rate": 1e-06,
+      "loss": 0.1522,
+      "step": 8733
+    },
+    {
+      "epoch": 1.6701405488096377,
+      "grad_norm": 4.03955523733776,
+      "learning_rate": 1e-06,
+      "loss": 0.145,
+      "step": 8734
+    },
+    {
+      "epoch": 1.6703317716798929,
+      "grad_norm": 1.9672009868326152,
+      "learning_rate": 1e-06,
+      "loss": 0.047,
+      "step": 8735
+    },
+    {
+      "epoch": 1.670522994550148,
+      "grad_norm": 2.852170111313852,
+      "learning_rate": 1e-06,
+      "loss": 0.0926,
+      "step": 8736
+    },
+    {
+      "epoch": 1.6707142174204035,
+      "grad_norm": 6.4980125323145685,
+      "learning_rate": 1e-06,
+      "loss": 0.416,
+      "step": 8737
+    },
+    {
+      "epoch": 1.670905440290659,
+      "grad_norm": 2.4697517583790396,
+      "learning_rate": 1e-06,
+      "loss": 0.0719,
+      "step": 8738
+    },
+    {
+      "epoch": 1.671096663160914,
+      "grad_norm": 2.9480380694908743,
+      "learning_rate": 1e-06,
+      "loss": 0.105,
+      "step": 8739
+    },
+    {
+      "epoch": 1.6712878860311693,
+      "grad_norm": 3.262716503925119,
+      "learning_rate": 1e-06,
+      "loss": 0.031,
+      "step": 8740
+    },
+    {
+      "epoch": 1.6714791089014245,
+      "grad_norm": 3.1025928544282313,
+      "learning_rate": 1e-06,
+      "loss": 0.0754,
+      "step": 8741
+    },
+    {
+      "epoch": 1.67167033177168,
+      "grad_norm": 1.783125642762334,
+      "learning_rate": 1e-06,
+      "loss": 0.021,
+      "step": 8742
+    },
+    {
+      "epoch": 1.6718615546419353,
+      "grad_norm": 6.156412035363611,
+      "learning_rate": 1e-06,
+      "loss": 0.3501,
+      "step": 8743
+    },
+    {
+      "epoch": 1.6720527775121905,
+      "grad_norm": 3.424607392244493,
+      "learning_rate": 1e-06,
+      "loss": 0.3059,
+      "step": 8744
+    },
+    {
+      "epoch": 1.6722440003824457,
+      "grad_norm": 2.7840389027453463,
+      "learning_rate": 1e-06,
+      "loss": 0.1754,
+      "step": 8745
+    },
+    {
+      "epoch": 1.6724352232527009,
+      "grad_norm": 3.2219977428539663,
+      "learning_rate": 1e-06,
+      "loss": 0.1243,
+      "step": 8746
+    },
+    {
+      "epoch": 1.6726264461229563,
+      "grad_norm": 6.245531239328307,
+      "learning_rate": 1e-06,
+      "loss": 0.0829,
+      "step": 8747
+    },
+    {
+      "epoch": 1.6728176689932117,
+      "grad_norm": 3.1859864774178734,
+      "learning_rate": 1e-06,
+      "loss": 0.087,
+      "step": 8748
+    },
+    {
+      "epoch": 1.673008891863467,
+      "grad_norm": 3.902135528416319,
+      "learning_rate": 1e-06,
+      "loss": 0.2114,
+      "step": 8749
+    },
+    {
+      "epoch": 1.673200114733722,
+      "grad_norm": 3.357967631935493,
+      "learning_rate": 1e-06,
+      "loss": 0.051,
+      "step": 8750
+    },
+    {
+      "epoch": 1.6733913376039773,
+      "grad_norm": 2.7668450891650083,
+      "learning_rate": 1e-06,
+      "loss": 0.0547,
+      "step": 8751
+    },
+    {
+      "epoch": 1.6735825604742327,
+      "grad_norm": 5.0773038699099295,
+      "learning_rate": 1e-06,
+      "loss": 0.1611,
+      "step": 8752
+    },
+    {
+      "epoch": 1.6737737833444881,
+      "grad_norm": 2.0877386310803665,
+      "learning_rate": 1e-06,
+      "loss": 0.035,
+      "step": 8753
+    },
+    {
+      "epoch": 1.6739650062147433,
+      "grad_norm": 4.432336731564688,
+      "learning_rate": 1e-06,
+      "loss": 0.1422,
+      "step": 8754
+    },
+    {
+      "epoch": 1.6741562290849985,
+      "grad_norm": 5.178495934598512,
+      "learning_rate": 1e-06,
+      "loss": 0.1076,
+      "step": 8755
+    },
+    {
+      "epoch": 1.6743474519552537,
+      "grad_norm": 3.381686474732853,
+      "learning_rate": 1e-06,
+      "loss": 0.2731,
+      "step": 8756
+    },
+    {
+      "epoch": 1.674538674825509,
+      "grad_norm": 3.8306689470520916,
+      "learning_rate": 1e-06,
+      "loss": 0.1578,
+      "step": 8757
+    },
+    {
+      "epoch": 1.6747298976957645,
+      "grad_norm": 1.7885851493915925,
+      "learning_rate": 1e-06,
+      "loss": 0.0392,
+      "step": 8758
+    },
+    {
+      "epoch": 1.6749211205660197,
+      "grad_norm": 2.3198343420078795,
+      "learning_rate": 1e-06,
+      "loss": 0.0592,
+      "step": 8759
+    },
+    {
+      "epoch": 1.675112343436275,
+      "grad_norm": 4.003659005323531,
+      "learning_rate": 1e-06,
+      "loss": 0.2234,
+      "step": 8760
+    },
+    {
+      "epoch": 1.67530356630653,
+      "grad_norm": 2.678471948502047,
+      "learning_rate": 1e-06,
+      "loss": 0.1035,
+      "step": 8761
+    },
+    {
+      "epoch": 1.6754947891767855,
+      "grad_norm": 3.5391413193571895,
+      "learning_rate": 1e-06,
+      "loss": 0.1057,
+      "step": 8762
+    },
+    {
+      "epoch": 1.675686012047041,
+      "grad_norm": 2.6566258781617247,
+      "learning_rate": 1e-06,
+      "loss": 0.0315,
+      "step": 8763
+    },
+    {
+      "epoch": 1.6758772349172961,
+      "grad_norm": 6.158368206140515,
+      "learning_rate": 1e-06,
+      "loss": 0.2601,
+      "step": 8764
+    },
+    {
+      "epoch": 1.6760684577875513,
+      "grad_norm": 2.6280957087677304,
+      "learning_rate": 1e-06,
+      "loss": 0.1614,
+      "step": 8765
+    },
+    {
+      "epoch": 1.6762596806578067,
+      "grad_norm": 1.963112768252309,
+      "learning_rate": 1e-06,
+      "loss": 0.0411,
+      "step": 8766
+    },
+    {
+      "epoch": 1.676450903528062,
+      "grad_norm": 4.024096625698009,
+      "learning_rate": 1e-06,
+      "loss": 0.0367,
+      "step": 8767
+    },
+    {
+      "epoch": 1.6766421263983173,
+      "grad_norm": 4.9499283833573,
+      "learning_rate": 1e-06,
+      "loss": 0.389,
+      "step": 8768
+    },
+    {
+      "epoch": 1.6768333492685725,
+      "grad_norm": 6.338450656593788,
+      "learning_rate": 1e-06,
+      "loss": 0.5094,
+      "step": 8769
+    },
+    {
+      "epoch": 1.6770245721388277,
+      "grad_norm": 6.341892816201446,
+      "learning_rate": 1e-06,
+      "loss": 0.4929,
+      "step": 8770
+    },
+    {
+      "epoch": 1.6772157950090831,
+      "grad_norm": 2.3926261554490225,
+      "learning_rate": 1e-06,
+      "loss": 0.1246,
+      "step": 8771
+    },
+    {
+      "epoch": 1.6774070178793383,
+      "grad_norm": 2.554982859484964,
+      "learning_rate": 1e-06,
+      "loss": 0.1211,
+      "step": 8772
+    },
+    {
+      "epoch": 1.6775982407495937,
+      "grad_norm": 3.1917976877064196,
+      "learning_rate": 1e-06,
+      "loss": 0.0725,
+      "step": 8773
+    },
+    {
+      "epoch": 1.677789463619849,
+      "grad_norm": 2.81859352371077,
+      "learning_rate": 1e-06,
+      "loss": 0.0913,
+      "step": 8774
+    },
+    {
+      "epoch": 1.6779806864901041,
+      "grad_norm": 1.544764842312591,
+      "learning_rate": 1e-06,
+      "loss": 0.0406,
+      "step": 8775
+    },
+    {
+      "epoch": 1.6781719093603595,
+      "grad_norm": 2.082321862413225,
+      "learning_rate": 1e-06,
+      "loss": 0.0495,
+      "step": 8776
+    },
+    {
+      "epoch": 1.6783631322306147,
+      "grad_norm": 2.3504382516182827,
+      "learning_rate": 1e-06,
+      "loss": 0.0397,
+      "step": 8777
+    },
+    {
+      "epoch": 1.6785543551008701,
+      "grad_norm": 3.4649995340894684,
+      "learning_rate": 1e-06,
+      "loss": 0.0687,
+      "step": 8778
+    },
+    {
+      "epoch": 1.6787455779711253,
+      "grad_norm": 3.05249584825688,
+      "learning_rate": 1e-06,
+      "loss": 0.0516,
+      "step": 8779
+    },
+    {
+      "epoch": 1.6789368008413805,
+      "grad_norm": 4.3069955830902344,
+      "learning_rate": 1e-06,
+      "loss": 0.1193,
+      "step": 8780
+    },
+    {
+      "epoch": 1.679128023711636,
+      "grad_norm": 6.563197507574909,
+      "learning_rate": 1e-06,
+      "loss": 0.7686,
+      "step": 8781
+    },
+    {
+      "epoch": 1.6793192465818914,
+      "grad_norm": 4.7233084582470415,
+      "learning_rate": 1e-06,
+      "loss": 0.2209,
+      "step": 8782
+    },
+    {
+      "epoch": 1.6795104694521465,
+      "grad_norm": 1.2897766909158381,
+      "learning_rate": 1e-06,
+      "loss": 0.1267,
+      "step": 8783
+    },
+    {
+      "epoch": 1.6797016923224017,
+      "grad_norm": 2.520769155226565,
+      "learning_rate": 1e-06,
+      "loss": 0.1654,
+      "step": 8784
+    },
+    {
+      "epoch": 1.679892915192657,
+      "grad_norm": 3.4944602176193698,
+      "learning_rate": 1e-06,
+      "loss": 0.2002,
+      "step": 8785
+    },
+    {
+      "epoch": 1.6800841380629123,
+      "grad_norm": 3.9802593441912153,
+      "learning_rate": 1e-06,
+      "loss": 0.4025,
+      "step": 8786
+    },
+    {
+      "epoch": 1.6802753609331678,
+      "grad_norm": 3.6906528248977555,
+      "learning_rate": 1e-06,
+      "loss": 0.0467,
+      "step": 8787
+    },
+    {
+      "epoch": 1.680466583803423,
+      "grad_norm": 3.326701688206271,
+      "learning_rate": 1e-06,
+      "loss": 0.0708,
+      "step": 8788
+    },
+    {
+      "epoch": 1.6806578066736781,
+      "grad_norm": 3.5883443230937444,
+      "learning_rate": 1e-06,
+      "loss": 0.0421,
+      "step": 8789
+    },
+    {
+      "epoch": 1.6808490295439333,
+      "grad_norm": 1.9935278478762666,
+      "learning_rate": 1e-06,
+      "loss": 0.0288,
+      "step": 8790
+    },
+    {
+      "epoch": 1.6810402524141888,
+      "grad_norm": 6.486267843084077,
+      "learning_rate": 1e-06,
+      "loss": 0.047,
+      "step": 8791
+    },
+    {
+      "epoch": 1.6812314752844442,
+      "grad_norm": 6.2208202500747625,
+      "learning_rate": 1e-06,
+      "loss": 0.0816,
+      "step": 8792
+    },
+    {
+      "epoch": 1.6814226981546994,
+      "grad_norm": 6.423533278901593,
+      "learning_rate": 1e-06,
+      "loss": 0.2091,
+      "step": 8793
+    },
+    {
+      "epoch": 1.6816139210249545,
+      "grad_norm": 3.6159203041966053,
+      "learning_rate": 1e-06,
+      "loss": 0.1359,
+      "step": 8794
+    },
+    {
+      "epoch": 1.6818051438952097,
+      "grad_norm": 4.439701232824475,
+      "learning_rate": 1e-06,
+      "loss": 0.2792,
+      "step": 8795
+    },
+    {
+      "epoch": 1.6819963667654652,
+      "grad_norm": 3.431793331076934,
+      "learning_rate": 1e-06,
+      "loss": 0.1055,
+      "step": 8796
+    },
+    {
+      "epoch": 1.6821875896357206,
+      "grad_norm": 3.152376422925474,
+      "learning_rate": 1e-06,
+      "loss": 0.1566,
+      "step": 8797
+    },
+    {
+      "epoch": 1.6823788125059758,
+      "grad_norm": 2.196130023386676,
+      "learning_rate": 1e-06,
+      "loss": 0.033,
+      "step": 8798
+    },
+    {
+      "epoch": 1.682570035376231,
+      "grad_norm": 1.71920659329378,
+      "learning_rate": 1e-06,
+      "loss": 0.0558,
+      "step": 8799
+    },
+    {
+      "epoch": 1.6827612582464861,
+      "grad_norm": 2.1592059675158946,
+      "learning_rate": 1e-06,
+      "loss": 0.0974,
+      "step": 8800
+    },
+    {
+      "epoch": 1.6829524811167416,
+      "grad_norm": 3.090893746659242,
+      "learning_rate": 1e-06,
+      "loss": 0.2498,
+      "step": 8801
+    },
+    {
+      "epoch": 1.683143703986997,
+      "grad_norm": 5.766220103413142,
+      "learning_rate": 1e-06,
+      "loss": 0.1231,
+      "step": 8802
+    },
+    {
+      "epoch": 1.6833349268572522,
+      "grad_norm": 2.1119467422749585,
+      "learning_rate": 1e-06,
+      "loss": 0.0353,
+      "step": 8803
+    },
+    {
+      "epoch": 1.6835261497275074,
+      "grad_norm": 2.2678606249993862,
+      "learning_rate": 1e-06,
+      "loss": 0.031,
+      "step": 8804
+    },
+    {
+      "epoch": 1.6837173725977626,
+      "grad_norm": 2.7614516327348704,
+      "learning_rate": 1e-06,
+      "loss": 0.1551,
+      "step": 8805
+    },
+    {
+      "epoch": 1.683908595468018,
+      "grad_norm": 5.965210828996323,
+      "learning_rate": 1e-06,
+      "loss": 0.2741,
+      "step": 8806
+    },
+    {
+      "epoch": 1.6840998183382734,
+      "grad_norm": 3.4396394487436495,
+      "learning_rate": 1e-06,
+      "loss": 0.1776,
+      "step": 8807
+    },
+    {
+      "epoch": 1.6842910412085286,
+      "grad_norm": 3.67118341953792,
+      "learning_rate": 1e-06,
+      "loss": 0.3076,
+      "step": 8808
+    },
+    {
+      "epoch": 1.6844822640787838,
+      "grad_norm": 3.6507560991971504,
+      "learning_rate": 1e-06,
+      "loss": 0.1527,
+      "step": 8809
+    },
+    {
+      "epoch": 1.684673486949039,
+      "grad_norm": 2.829475881077038,
+      "learning_rate": 1e-06,
+      "loss": 0.2216,
+      "step": 8810
+    },
+    {
+      "epoch": 1.6848647098192944,
+      "grad_norm": 2.7189596687235476,
+      "learning_rate": 1e-06,
+      "loss": 0.114,
+      "step": 8811
+    },
+    {
+      "epoch": 1.6850559326895498,
+      "grad_norm": 2.6735725587209345,
+      "learning_rate": 1e-06,
+      "loss": 0.0875,
+      "step": 8812
+    },
+    {
+      "epoch": 1.685247155559805,
+      "grad_norm": 2.391540950257857,
+      "learning_rate": 1e-06,
+      "loss": 0.0571,
+      "step": 8813
+    },
+    {
+      "epoch": 1.6854383784300602,
+      "grad_norm": 4.59870093702858,
+      "learning_rate": 1e-06,
+      "loss": 0.2054,
+      "step": 8814
+    },
+    {
+      "epoch": 1.6856296013003154,
+      "grad_norm": 5.007220204954861,
+      "learning_rate": 1e-06,
+      "loss": 0.0799,
+      "step": 8815
+    },
+    {
+      "epoch": 1.6858208241705708,
+      "grad_norm": 1.7721235474341526,
+      "learning_rate": 1e-06,
+      "loss": 0.0414,
+      "step": 8816
+    },
+    {
+      "epoch": 1.6860120470408262,
+      "grad_norm": 2.572793244607559,
+      "learning_rate": 1e-06,
+      "loss": 0.0998,
+      "step": 8817
+    },
+    {
+      "epoch": 1.6862032699110814,
+      "grad_norm": 5.1918006442903515,
+      "learning_rate": 1e-06,
+      "loss": 0.1496,
+      "step": 8818
+    },
+    {
+      "epoch": 1.6863944927813366,
+      "grad_norm": 8.406911604134217,
+      "learning_rate": 1e-06,
+      "loss": 0.5331,
+      "step": 8819
+    },
+    {
+      "epoch": 1.6865857156515918,
+      "grad_norm": 4.989178199752613,
+      "learning_rate": 1e-06,
+      "loss": 0.0815,
+      "step": 8820
+    },
+    {
+      "epoch": 1.6867769385218472,
+      "grad_norm": 1.8574605109431357,
+      "learning_rate": 1e-06,
+      "loss": 0.1788,
+      "step": 8821
+    },
+    {
+      "epoch": 1.6869681613921026,
+      "grad_norm": 6.571103514456691,
+      "learning_rate": 1e-06,
+      "loss": 0.3537,
+      "step": 8822
+    },
+    {
+      "epoch": 1.6871593842623578,
+      "grad_norm": 3.7262006220062083,
+      "learning_rate": 1e-06,
+      "loss": 0.1112,
+      "step": 8823
+    },
+    {
+      "epoch": 1.687350607132613,
+      "grad_norm": 4.188779037771091,
+      "learning_rate": 1e-06,
+      "loss": 0.3326,
+      "step": 8824
+    },
+    {
+      "epoch": 1.6875418300028684,
+      "grad_norm": 5.253644677213066,
+      "learning_rate": 1e-06,
+      "loss": 0.1341,
+      "step": 8825
+    },
+    {
+      "epoch": 1.6877330528731236,
+      "grad_norm": 5.224215997164951,
+      "learning_rate": 1e-06,
+      "loss": 0.1168,
+      "step": 8826
+    },
+    {
+      "epoch": 1.687924275743379,
+      "grad_norm": 5.299983819900863,
+      "learning_rate": 1e-06,
+      "loss": 0.0646,
+      "step": 8827
+    },
+    {
+      "epoch": 1.6881154986136342,
+      "grad_norm": 2.0519229290488936,
+      "learning_rate": 1e-06,
+      "loss": 0.0291,
+      "step": 8828
+    },
+    {
+      "epoch": 1.6883067214838894,
+      "grad_norm": 3.4200618591068186,
+      "learning_rate": 1e-06,
+      "loss": 0.0652,
+      "step": 8829
+    },
+    {
+      "epoch": 1.6884979443541448,
+      "grad_norm": 3.1877350066075243,
+      "learning_rate": 1e-06,
+      "loss": 0.0724,
+      "step": 8830
+    },
+    {
+      "epoch": 1.6886891672244,
+      "grad_norm": 7.764015568257229,
+      "learning_rate": 1e-06,
+      "loss": 0.3297,
+      "step": 8831
+    },
+    {
+      "epoch": 1.6888803900946554,
+      "grad_norm": 4.080598862772157,
+      "learning_rate": 1e-06,
+      "loss": 0.3337,
+      "step": 8832
+    },
+    {
+      "epoch": 1.6890716129649106,
+      "grad_norm": 1.642198661671898,
+      "learning_rate": 1e-06,
+      "loss": 0.1174,
+      "step": 8833
+    },
+    {
+      "epoch": 1.6892628358351658,
+      "grad_norm": 1.6031532389215324,
+      "learning_rate": 1e-06,
+      "loss": 0.1745,
+      "step": 8834
+    },
+    {
+      "epoch": 1.6894540587054212,
+      "grad_norm": 5.085069995499655,
+      "learning_rate": 1e-06,
+      "loss": 0.2619,
+      "step": 8835
+    },
+    {
+      "epoch": 1.6896452815756766,
+      "grad_norm": 4.254574725799526,
+      "learning_rate": 1e-06,
+      "loss": 0.2404,
+      "step": 8836
+    },
+    {
+      "epoch": 1.6898365044459318,
+      "grad_norm": 3.1952149019515605,
+      "learning_rate": 1e-06,
+      "loss": 0.1145,
+      "step": 8837
+    },
+    {
+      "epoch": 1.690027727316187,
+      "grad_norm": 6.481047068171432,
+      "learning_rate": 1e-06,
+      "loss": 0.2507,
+      "step": 8838
+    },
+    {
+      "epoch": 1.6902189501864422,
+      "grad_norm": 2.398474776882295,
+      "learning_rate": 1e-06,
+      "loss": 0.0547,
+      "step": 8839
+    },
+    {
+      "epoch": 1.6904101730566976,
+      "grad_norm": 3.165431953109728,
+      "learning_rate": 1e-06,
+      "loss": 0.0574,
+      "step": 8840
+    },
+    {
+      "epoch": 1.690601395926953,
+      "grad_norm": 2.08993284802039,
+      "learning_rate": 1e-06,
+      "loss": 0.0399,
+      "step": 8841
+    },
+    {
+      "epoch": 1.6907926187972082,
+      "grad_norm": 4.133590769185398,
+      "learning_rate": 1e-06,
+      "loss": 0.0756,
+      "step": 8842
+    },
+    {
+      "epoch": 1.6909838416674634,
+      "grad_norm": 5.310087127258108,
+      "learning_rate": 1e-06,
+      "loss": 0.3585,
+      "step": 8843
+    },
+    {
+      "epoch": 1.6911750645377186,
+      "grad_norm": 3.923883782348546,
+      "learning_rate": 1e-06,
+      "loss": 0.1425,
+      "step": 8844
+    },
+    {
+      "epoch": 1.691366287407974,
+      "grad_norm": 6.658648564686968,
+      "learning_rate": 1e-06,
+      "loss": 0.0925,
+      "step": 8845
+    },
+    {
+      "epoch": 1.6915575102782294,
+      "grad_norm": 1.8675876332343975,
+      "learning_rate": 1e-06,
+      "loss": 0.0946,
+      "step": 8846
+    },
+    {
+      "epoch": 1.6917487331484846,
+      "grad_norm": 3.4658211379965205,
+      "learning_rate": 1e-06,
+      "loss": 0.1055,
+      "step": 8847
+    },
+    {
+      "epoch": 1.6919399560187398,
+      "grad_norm": 4.983832159348237,
+      "learning_rate": 1e-06,
+      "loss": 0.2888,
+      "step": 8848
+    },
+    {
+      "epoch": 1.692131178888995,
+      "grad_norm": 2.0237283976240126,
+      "learning_rate": 1e-06,
+      "loss": 0.0826,
+      "step": 8849
+    },
+    {
+      "epoch": 1.6923224017592504,
+      "grad_norm": 2.172206386834735,
+      "learning_rate": 1e-06,
+      "loss": 0.0502,
+      "step": 8850
+    },
+    {
+      "epoch": 1.6925136246295058,
+      "grad_norm": 2.8809747394803598,
+      "learning_rate": 1e-06,
+      "loss": 0.0444,
+      "step": 8851
+    },
+    {
+      "epoch": 1.692704847499761,
+      "grad_norm": 2.720371595833055,
+      "learning_rate": 1e-06,
+      "loss": 0.1036,
+      "step": 8852
+    },
+    {
+      "epoch": 1.6928960703700162,
+      "grad_norm": 1.6206596804496634,
+      "learning_rate": 1e-06,
+      "loss": 0.02,
+      "step": 8853
+    },
+    {
+      "epoch": 1.6930872932402714,
+      "grad_norm": 2.4621324332632883,
+      "learning_rate": 1e-06,
+      "loss": 0.057,
+      "step": 8854
+    },
+    {
+      "epoch": 1.6932785161105268,
+      "grad_norm": 4.938483997269221,
+      "learning_rate": 1e-06,
+      "loss": 0.2527,
+      "step": 8855
+    },
+    {
+      "epoch": 1.6934697389807822,
+      "grad_norm": 3.868197592397597,
+      "learning_rate": 1e-06,
+      "loss": 0.2452,
+      "step": 8856
+    },
+    {
+      "epoch": 1.6936609618510374,
+      "grad_norm": 3.9114578321569184,
+      "learning_rate": 1e-06,
+      "loss": 0.3181,
+      "step": 8857
+    },
+    {
+      "epoch": 1.6938521847212926,
+      "grad_norm": 3.4476305025558234,
+      "learning_rate": 1e-06,
+      "loss": 0.4435,
+      "step": 8858
+    },
+    {
+      "epoch": 1.6940434075915478,
+      "grad_norm": 3.0221954081162368,
+      "learning_rate": 1e-06,
+      "loss": 0.154,
+      "step": 8859
+    },
+    {
+      "epoch": 1.6942346304618032,
+      "grad_norm": 4.794650180057049,
+      "learning_rate": 1e-06,
+      "loss": 0.2091,
+      "step": 8860
+    },
+    {
+      "epoch": 1.6944258533320586,
+      "grad_norm": 3.1879179437432263,
+      "learning_rate": 1e-06,
+      "loss": 0.2438,
+      "step": 8861
+    },
+    {
+      "epoch": 1.6946170762023138,
+      "grad_norm": 3.2354077150933764,
+      "learning_rate": 1e-06,
+      "loss": 0.2267,
+      "step": 8862
+    },
+    {
+      "epoch": 1.694808299072569,
+      "grad_norm": 1.6627027660423723,
+      "learning_rate": 1e-06,
+      "loss": 0.0923,
+      "step": 8863
+    },
+    {
+      "epoch": 1.6949995219428242,
+      "grad_norm": 4.067101088495107,
+      "learning_rate": 1e-06,
+      "loss": 0.11,
+      "step": 8864
+    },
+    {
+      "epoch": 1.6951907448130796,
+      "grad_norm": 2.696220709332572,
+      "learning_rate": 1e-06,
+      "loss": 0.0791,
+      "step": 8865
+    },
+    {
+      "epoch": 1.695381967683335,
+      "grad_norm": 2.0443913191421283,
+      "learning_rate": 1e-06,
+      "loss": 0.0389,
+      "step": 8866
+    },
+    {
+      "epoch": 1.6955731905535902,
+      "grad_norm": 7.334169802499431,
+      "learning_rate": 1e-06,
+      "loss": 0.1235,
+      "step": 8867
+    },
+    {
+      "epoch": 1.6957644134238454,
+      "grad_norm": 4.876556148115457,
+      "learning_rate": 1e-06,
+      "loss": 0.1201,
+      "step": 8868
+    },
+    {
+      "epoch": 1.6959556362941006,
+      "grad_norm": 5.420415060613098,
+      "learning_rate": 1e-06,
+      "loss": 0.5486,
+      "step": 8869
+    },
+    {
+      "epoch": 1.696146859164356,
+      "grad_norm": 4.903772785679342,
+      "learning_rate": 1e-06,
+      "loss": 0.2215,
+      "step": 8870
+    },
+    {
+      "epoch": 1.6963380820346114,
+      "grad_norm": 3.405261675040626,
+      "learning_rate": 1e-06,
+      "loss": 0.2626,
+      "step": 8871
+    },
+    {
+      "epoch": 1.6965293049048666,
+      "grad_norm": 4.9074557821154094,
+      "learning_rate": 1e-06,
+      "loss": 0.3396,
+      "step": 8872
+    },
+    {
+      "epoch": 1.6967205277751218,
+      "grad_norm": 2.736878649423226,
+      "learning_rate": 1e-06,
+      "loss": 0.0871,
+      "step": 8873
+    },
+    {
+      "epoch": 1.696911750645377,
+      "grad_norm": 4.200960076817751,
+      "learning_rate": 1e-06,
+      "loss": 0.3108,
+      "step": 8874
+    },
+    {
+      "epoch": 1.6971029735156324,
+      "grad_norm": 2.7010715159806744,
+      "learning_rate": 1e-06,
+      "loss": 0.2928,
+      "step": 8875
+    },
+    {
+      "epoch": 1.6972941963858879,
+      "grad_norm": 2.5602811185146748,
+      "learning_rate": 1e-06,
+      "loss": 0.0771,
+      "step": 8876
+    },
+    {
+      "epoch": 1.697485419256143,
+      "grad_norm": 2.9573822820973583,
+      "learning_rate": 1e-06,
+      "loss": 0.1859,
+      "step": 8877
+    },
+    {
+      "epoch": 1.6976766421263982,
+      "grad_norm": 3.353285979915543,
+      "learning_rate": 1e-06,
+      "loss": 0.2348,
+      "step": 8878
+    },
+    {
+      "epoch": 1.6978678649966537,
+      "grad_norm": 2.6272482553778094,
+      "learning_rate": 1e-06,
+      "loss": 0.0624,
+      "step": 8879
+    },
+    {
+      "epoch": 1.6980590878669088,
+      "grad_norm": 3.2427469161585614,
+      "learning_rate": 1e-06,
+      "loss": 0.0855,
+      "step": 8880
+    },
+    {
+      "epoch": 1.6982503107371643,
+      "grad_norm": 5.756855071135506,
+      "learning_rate": 1e-06,
+      "loss": 0.3811,
+      "step": 8881
+    },
+    {
+      "epoch": 1.6984415336074195,
+      "grad_norm": 3.011454173537902,
+      "learning_rate": 1e-06,
+      "loss": 0.3098,
+      "step": 8882
+    },
+    {
+      "epoch": 1.6986327564776746,
+      "grad_norm": 2.839118623572432,
+      "learning_rate": 1e-06,
+      "loss": 0.1444,
+      "step": 8883
+    },
+    {
+      "epoch": 1.69882397934793,
+      "grad_norm": 2.7372790800510205,
+      "learning_rate": 1e-06,
+      "loss": 0.1859,
+      "step": 8884
+    },
+    {
+      "epoch": 1.6990152022181852,
+      "grad_norm": 3.373783316120901,
+      "learning_rate": 1e-06,
+      "loss": 0.0945,
+      "step": 8885
+    },
+    {
+      "epoch": 1.6992064250884407,
+      "grad_norm": 2.4986831057641155,
+      "learning_rate": 1e-06,
+      "loss": 0.0998,
+      "step": 8886
+    },
+    {
+      "epoch": 1.6993976479586959,
+      "grad_norm": 3.209936575337274,
+      "learning_rate": 1e-06,
+      "loss": 0.1246,
+      "step": 8887
+    },
+    {
+      "epoch": 1.699588870828951,
+      "grad_norm": 2.8457104978027346,
+      "learning_rate": 1e-06,
+      "loss": 0.0992,
+      "step": 8888
+    },
+    {
+      "epoch": 1.6997800936992065,
+      "grad_norm": 3.5292964696751983,
+      "learning_rate": 1e-06,
+      "loss": 0.0515,
+      "step": 8889
+    },
+    {
+      "epoch": 1.6999713165694617,
+      "grad_norm": 1.9906179191427487,
+      "learning_rate": 1e-06,
+      "loss": 0.0583,
+      "step": 8890
+    },
+    {
+      "epoch": 1.700162539439717,
+      "grad_norm": 2.475131322285958,
+      "learning_rate": 1e-06,
+      "loss": 0.0771,
+      "step": 8891
+    },
+    {
+      "epoch": 1.7003537623099723,
+      "grad_norm": 4.358532677219327,
+      "learning_rate": 1e-06,
+      "loss": 0.0871,
+      "step": 8892
+    },
+    {
+      "epoch": 1.7005449851802275,
+      "grad_norm": 4.561778599391707,
+      "learning_rate": 1e-06,
+      "loss": 0.2925,
+      "step": 8893
+    },
+    {
+      "epoch": 1.7007362080504829,
+      "grad_norm": 7.834832338877477,
+      "learning_rate": 1e-06,
+      "loss": 0.4757,
+      "step": 8894
+    },
+    {
+      "epoch": 1.7009274309207383,
+      "grad_norm": 2.893474267465812,
+      "learning_rate": 1e-06,
+      "loss": 0.1052,
+      "step": 8895
+    },
+    {
+      "epoch": 1.7011186537909935,
+      "grad_norm": 1.6048144564231337,
+      "learning_rate": 1e-06,
+      "loss": 0.1283,
+      "step": 8896
+    },
+    {
+      "epoch": 1.7013098766612487,
+      "grad_norm": 2.6885299483325555,
+      "learning_rate": 1e-06,
+      "loss": 0.1662,
+      "step": 8897
+    },
+    {
+      "epoch": 1.7015010995315039,
+      "grad_norm": 4.379108135051209,
+      "learning_rate": 1e-06,
+      "loss": 0.1419,
+      "step": 8898
+    },
+    {
+      "epoch": 1.7016923224017593,
+      "grad_norm": 3.8353984632721425,
+      "learning_rate": 1e-06,
+      "loss": 0.2321,
+      "step": 8899
+    },
+    {
+      "epoch": 1.7018835452720147,
+      "grad_norm": 2.5233598348753605,
+      "learning_rate": 1e-06,
+      "loss": 0.0873,
+      "step": 8900
+    },
+    {
+      "epoch": 1.7020747681422699,
+      "grad_norm": 2.614542521326665,
+      "learning_rate": 1e-06,
+      "loss": 0.0459,
+      "step": 8901
+    },
+    {
+      "epoch": 1.702265991012525,
+      "grad_norm": 2.9947822017796537,
+      "learning_rate": 1e-06,
+      "loss": 0.1441,
+      "step": 8902
+    },
+    {
+      "epoch": 1.7024572138827803,
+      "grad_norm": 3.293419930353019,
+      "learning_rate": 1e-06,
+      "loss": 0.0816,
+      "step": 8903
+    },
+    {
+      "epoch": 1.7026484367530357,
+      "grad_norm": 3.3369613612849918,
+      "learning_rate": 1e-06,
+      "loss": 0.0578,
+      "step": 8904
+    },
+    {
+      "epoch": 1.702839659623291,
+      "grad_norm": 4.234608323541694,
+      "learning_rate": 1e-06,
+      "loss": 0.21,
+      "step": 8905
+    },
+    {
+      "epoch": 1.7030308824935463,
+      "grad_norm": 5.480947787610524,
+      "learning_rate": 1e-06,
+      "loss": 0.5018,
+      "step": 8906
+    },
+    {
+      "epoch": 1.7032221053638015,
+      "grad_norm": 4.47061843758529,
+      "learning_rate": 1e-06,
+      "loss": 0.2719,
+      "step": 8907
+    },
+    {
+      "epoch": 1.7034133282340567,
+      "grad_norm": 2.7665586458129994,
+      "learning_rate": 1e-06,
+      "loss": 0.235,
+      "step": 8908
+    },
+    {
+      "epoch": 1.703604551104312,
+      "grad_norm": 4.418925781206837,
+      "learning_rate": 1e-06,
+      "loss": 0.4495,
+      "step": 8909
+    },
+    {
+      "epoch": 1.7037957739745675,
+      "grad_norm": 4.728539020491482,
+      "learning_rate": 1e-06,
+      "loss": 0.3859,
+      "step": 8910
+    },
+    {
+      "epoch": 1.7039869968448227,
+      "grad_norm": 4.1176078914734955,
+      "learning_rate": 1e-06,
+      "loss": 0.1447,
+      "step": 8911
+    },
+    {
+      "epoch": 1.7041782197150779,
+      "grad_norm": 2.2016064977208996,
+      "learning_rate": 1e-06,
+      "loss": 0.0881,
+      "step": 8912
+    },
+    {
+      "epoch": 1.704369442585333,
+      "grad_norm": 4.102762101359469,
+      "learning_rate": 1e-06,
+      "loss": 0.1599,
+      "step": 8913
+    },
+    {
+      "epoch": 1.7045606654555885,
+      "grad_norm": 3.474858996048535,
+      "learning_rate": 1e-06,
+      "loss": 0.0748,
+      "step": 8914
+    },
+    {
+      "epoch": 1.704751888325844,
+      "grad_norm": 3.7016115596768,
+      "learning_rate": 1e-06,
+      "loss": 0.0636,
+      "step": 8915
+    },
+    {
+      "epoch": 1.704943111196099,
+      "grad_norm": 2.504851021667171,
+      "learning_rate": 1e-06,
+      "loss": 0.0585,
+      "step": 8916
+    },
+    {
+      "epoch": 1.7051343340663543,
+      "grad_norm": 5.070349365476533,
+      "learning_rate": 1e-06,
+      "loss": 0.27,
+      "step": 8917
+    },
+    {
+      "epoch": 1.7053255569366095,
+      "grad_norm": 4.479000047053894,
+      "learning_rate": 1e-06,
+      "loss": 0.2026,
+      "step": 8918
+    },
+    {
+      "epoch": 1.705516779806865,
+      "grad_norm": 3.5544343627673296,
+      "learning_rate": 1e-06,
+      "loss": 0.4393,
+      "step": 8919
+    },
+    {
+      "epoch": 1.7057080026771203,
+      "grad_norm": 1.6457956124662945,
+      "learning_rate": 1e-06,
+      "loss": 0.0809,
+      "step": 8920
+    },
+    {
+      "epoch": 1.7058992255473755,
+      "grad_norm": 1.9435842412695614,
+      "learning_rate": 1e-06,
+      "loss": 0.0864,
+      "step": 8921
+    },
+    {
+      "epoch": 1.7060904484176307,
+      "grad_norm": 1.9710465010136937,
+      "learning_rate": 1e-06,
+      "loss": 0.0649,
+      "step": 8922
+    },
+    {
+      "epoch": 1.7062816712878859,
+      "grad_norm": 3.1312510720744173,
+      "learning_rate": 1e-06,
+      "loss": 0.0491,
+      "step": 8923
+    },
+    {
+      "epoch": 1.7064728941581413,
+      "grad_norm": 4.760774138514577,
+      "learning_rate": 1e-06,
+      "loss": 0.3251,
+      "step": 8924
+    },
+    {
+      "epoch": 1.7066641170283967,
+      "grad_norm": 2.004526142812924,
+      "learning_rate": 1e-06,
+      "loss": 0.0282,
+      "step": 8925
+    },
+    {
+      "epoch": 1.706855339898652,
+      "grad_norm": 2.145938821316956,
+      "learning_rate": 1e-06,
+      "loss": 0.0499,
+      "step": 8926
+    },
+    {
+      "epoch": 1.707046562768907,
+      "grad_norm": 5.308014176304507,
+      "learning_rate": 1e-06,
+      "loss": 0.0974,
+      "step": 8927
+    },
+    {
+      "epoch": 1.7072377856391623,
+      "grad_norm": 3.8717339810975835,
+      "learning_rate": 1e-06,
+      "loss": 0.0938,
+      "step": 8928
+    },
+    {
+      "epoch": 1.7074290085094177,
+      "grad_norm": 8.57623259023999,
+      "learning_rate": 1e-06,
+      "loss": 0.1785,
+      "step": 8929
+    },
+    {
+      "epoch": 1.7076202313796731,
+      "grad_norm": 13.621396209360581,
+      "learning_rate": 1e-06,
+      "loss": 0.2805,
+      "step": 8930
+    },
+    {
+      "epoch": 1.7078114542499283,
+      "grad_norm": 4.617878766926463,
+      "learning_rate": 1e-06,
+      "loss": 0.3629,
+      "step": 8931
+    },
+    {
+      "epoch": 1.7080026771201835,
+      "grad_norm": 3.908843621382275,
+      "learning_rate": 1e-06,
+      "loss": 0.2345,
+      "step": 8932
+    },
+    {
+      "epoch": 1.7081938999904387,
+      "grad_norm": 3.0990553554872995,
+      "learning_rate": 1e-06,
+      "loss": 0.205,
+      "step": 8933
+    },
+    {
+      "epoch": 1.708385122860694,
+      "grad_norm": 3.4171530834779484,
+      "learning_rate": 1e-06,
+      "loss": 0.1182,
+      "step": 8934
+    },
+    {
+      "epoch": 1.7085763457309495,
+      "grad_norm": 2.839298830918881,
+      "learning_rate": 1e-06,
+      "loss": 0.2434,
+      "step": 8935
+    },
+    {
+      "epoch": 1.7087675686012047,
+      "grad_norm": 4.003548240472936,
+      "learning_rate": 1e-06,
+      "loss": 0.2926,
+      "step": 8936
+    },
+    {
+      "epoch": 1.70895879147146,
+      "grad_norm": 2.6213035897455748,
+      "learning_rate": 1e-06,
+      "loss": 0.1519,
+      "step": 8937
+    },
+    {
+      "epoch": 1.7091500143417153,
+      "grad_norm": 2.871744385780312,
+      "learning_rate": 1e-06,
+      "loss": 0.0506,
+      "step": 8938
+    },
+    {
+      "epoch": 1.7093412372119705,
+      "grad_norm": 2.20976501216681,
+      "learning_rate": 1e-06,
+      "loss": 0.0469,
+      "step": 8939
+    },
+    {
+      "epoch": 1.709532460082226,
+      "grad_norm": 4.638061226253548,
+      "learning_rate": 1e-06,
+      "loss": 0.1013,
+      "step": 8940
+    },
+    {
+      "epoch": 1.7097236829524811,
+      "grad_norm": 5.4964407761800445,
+      "learning_rate": 1e-06,
+      "loss": 0.0947,
+      "step": 8941
+    },
+    {
+      "epoch": 1.7099149058227363,
+      "grad_norm": 5.133278278369361,
+      "learning_rate": 1e-06,
+      "loss": 0.0659,
+      "step": 8942
+    },
+    {
+      "epoch": 1.7101061286929917,
+      "grad_norm": 4.4565735430654545,
+      "learning_rate": 1e-06,
+      "loss": 0.1509,
+      "step": 8943
+    },
+    {
+      "epoch": 1.710297351563247,
+      "grad_norm": 4.392072228676808,
+      "learning_rate": 1e-06,
+      "loss": 0.2694,
+      "step": 8944
+    },
+    {
+      "epoch": 1.7104885744335023,
+      "grad_norm": 2.4435388146977175,
+      "learning_rate": 1e-06,
+      "loss": 0.1111,
+      "step": 8945
+    },
+    {
+      "epoch": 1.7106797973037575,
+      "grad_norm": 3.266766070455923,
+      "learning_rate": 1e-06,
+      "loss": 0.0979,
+      "step": 8946
+    },
+    {
+      "epoch": 1.7108710201740127,
+      "grad_norm": 2.78645540516542,
+      "learning_rate": 1e-06,
+      "loss": 0.1595,
+      "step": 8947
+    },
+    {
+      "epoch": 1.7110622430442681,
+      "grad_norm": 3.4194235185330113,
+      "learning_rate": 1e-06,
+      "loss": 0.1873,
+      "step": 8948
+    },
+    {
+      "epoch": 1.7112534659145235,
+      "grad_norm": 1.9682451236586438,
+      "learning_rate": 1e-06,
+      "loss": 0.0511,
+      "step": 8949
+    },
+    {
+      "epoch": 1.7114446887847787,
+      "grad_norm": 2.440686221948407,
+      "learning_rate": 1e-06,
+      "loss": 0.0726,
+      "step": 8950
+    },
+    {
+      "epoch": 1.711635911655034,
+      "grad_norm": 1.5006667880365117,
+      "learning_rate": 1e-06,
+      "loss": 0.066,
+      "step": 8951
+    },
+    {
+      "epoch": 1.7118271345252891,
+      "grad_norm": 2.500815353947547,
+      "learning_rate": 1e-06,
+      "loss": 0.0695,
+      "step": 8952
+    },
+    {
+      "epoch": 1.7120183573955445,
+      "grad_norm": 1.2283772449762091,
+      "learning_rate": 1e-06,
+      "loss": 0.0155,
+      "step": 8953
+    },
+    {
+      "epoch": 1.7122095802658,
+      "grad_norm": 1.5997546156828426,
+      "learning_rate": 1e-06,
+      "loss": 0.0274,
+      "step": 8954
+    },
+    {
+      "epoch": 1.7124008031360551,
+      "grad_norm": 4.231794295808355,
+      "learning_rate": 1e-06,
+      "loss": 0.2674,
+      "step": 8955
+    },
+    {
+      "epoch": 1.7125920260063103,
+      "grad_norm": 5.802829696709501,
+      "learning_rate": 1e-06,
+      "loss": 0.5209,
+      "step": 8956
+    },
+    {
+      "epoch": 1.7127832488765655,
+      "grad_norm": 3.0063915512798616,
+      "learning_rate": 1e-06,
+      "loss": 0.1135,
+      "step": 8957
+    },
+    {
+      "epoch": 1.712974471746821,
+      "grad_norm": 3.7642689396108824,
+      "learning_rate": 1e-06,
+      "loss": 0.119,
+      "step": 8958
+    },
+    {
+      "epoch": 1.7131656946170764,
+      "grad_norm": 3.590392293934024,
+      "learning_rate": 1e-06,
+      "loss": 0.1253,
+      "step": 8959
+    },
+    {
+      "epoch": 1.7133569174873315,
+      "grad_norm": 6.361473843126851,
+      "learning_rate": 1e-06,
+      "loss": 0.5065,
+      "step": 8960
+    },
+    {
+      "epoch": 1.7135481403575867,
+      "grad_norm": 3.71833093670184,
+      "learning_rate": 1e-06,
+      "loss": 0.2708,
+      "step": 8961
+    },
+    {
+      "epoch": 1.713739363227842,
+      "grad_norm": 4.756037037877656,
+      "learning_rate": 1e-06,
+      "loss": 0.2661,
+      "step": 8962
+    },
+    {
+      "epoch": 1.7139305860980973,
+      "grad_norm": 2.2228881500426643,
+      "learning_rate": 1e-06,
+      "loss": 0.0442,
+      "step": 8963
+    },
+    {
+      "epoch": 1.7141218089683528,
+      "grad_norm": 3.850408106390074,
+      "learning_rate": 1e-06,
+      "loss": 0.0466,
+      "step": 8964
+    },
+    {
+      "epoch": 1.714313031838608,
+      "grad_norm": 4.702478510124591,
+      "learning_rate": 1e-06,
+      "loss": 0.0441,
+      "step": 8965
+    },
+    {
+      "epoch": 1.7145042547088631,
+      "grad_norm": 1.8738826601427756,
+      "learning_rate": 1e-06,
+      "loss": 0.0294,
+      "step": 8966
+    },
+    {
+      "epoch": 1.7146954775791183,
+      "grad_norm": 4.446119077748101,
+      "learning_rate": 1e-06,
+      "loss": 0.0469,
+      "step": 8967
+    },
+    {
+      "epoch": 1.7148867004493737,
+      "grad_norm": 7.453782154597968,
+      "learning_rate": 1e-06,
+      "loss": 0.3001,
+      "step": 8968
+    },
+    {
+      "epoch": 1.7150779233196292,
+      "grad_norm": 7.5630129805678825,
+      "learning_rate": 1e-06,
+      "loss": 0.7467,
+      "step": 8969
+    },
+    {
+      "epoch": 1.7152691461898844,
+      "grad_norm": 2.7628582407109055,
+      "learning_rate": 1e-06,
+      "loss": 0.0761,
+      "step": 8970
+    },
+    {
+      "epoch": 1.7154603690601395,
+      "grad_norm": 5.851447110636327,
+      "learning_rate": 1e-06,
+      "loss": 0.2417,
+      "step": 8971
+    },
+    {
+      "epoch": 1.7156515919303947,
+      "grad_norm": 3.836845088580545,
+      "learning_rate": 1e-06,
+      "loss": 0.2962,
+      "step": 8972
+    },
+    {
+      "epoch": 1.7158428148006502,
+      "grad_norm": 3.4703138795426542,
+      "learning_rate": 1e-06,
+      "loss": 0.2534,
+      "step": 8973
+    },
+    {
+      "epoch": 1.7160340376709056,
+      "grad_norm": 2.6126347812764976,
+      "learning_rate": 1e-06,
+      "loss": 0.1434,
+      "step": 8974
+    },
+    {
+      "epoch": 1.7162252605411608,
+      "grad_norm": 3.9685207510991165,
+      "learning_rate": 1e-06,
+      "loss": 0.0796,
+      "step": 8975
+    },
+    {
+      "epoch": 1.716416483411416,
+      "grad_norm": 4.0954476833045454,
+      "learning_rate": 1e-06,
+      "loss": 0.2127,
+      "step": 8976
+    },
+    {
+      "epoch": 1.7166077062816711,
+      "grad_norm": 2.0582319082787626,
+      "learning_rate": 1e-06,
+      "loss": 0.043,
+      "step": 8977
+    },
+    {
+      "epoch": 1.7167989291519266,
+      "grad_norm": 2.6555389350129803,
+      "learning_rate": 1e-06,
+      "loss": 0.058,
+      "step": 8978
+    },
+    {
+      "epoch": 1.716990152022182,
+      "grad_norm": 3.78345910478049,
+      "learning_rate": 1e-06,
+      "loss": 0.096,
+      "step": 8979
+    },
+    {
+      "epoch": 1.7171813748924372,
+      "grad_norm": 6.67151306742981,
+      "learning_rate": 1e-06,
+      "loss": 0.2445,
+      "step": 8980
+    },
+    {
+      "epoch": 1.7173725977626924,
+      "grad_norm": 4.306998683032031,
+      "learning_rate": 1e-06,
+      "loss": 0.582,
+      "step": 8981
+    },
+    {
+      "epoch": 1.7175638206329475,
+      "grad_norm": 3.8724175738827693,
+      "learning_rate": 1e-06,
+      "loss": 0.2791,
+      "step": 8982
+    },
+    {
+      "epoch": 1.717755043503203,
+      "grad_norm": 2.9499785600303388,
+      "learning_rate": 1e-06,
+      "loss": 0.1575,
+      "step": 8983
+    },
+    {
+      "epoch": 1.7179462663734584,
+      "grad_norm": 3.3983342297527734,
+      "learning_rate": 1e-06,
+      "loss": 0.1776,
+      "step": 8984
+    },
+    {
+      "epoch": 1.7181374892437136,
+      "grad_norm": 4.045837269292319,
+      "learning_rate": 1e-06,
+      "loss": 0.2118,
+      "step": 8985
+    },
+    {
+      "epoch": 1.7183287121139688,
+      "grad_norm": 2.197193792228603,
+      "learning_rate": 1e-06,
+      "loss": 0.0844,
+      "step": 8986
+    },
+    {
+      "epoch": 1.718519934984224,
+      "grad_norm": 3.88384713524442,
+      "learning_rate": 1e-06,
+      "loss": 0.1006,
+      "step": 8987
+    },
+    {
+      "epoch": 1.7187111578544794,
+      "grad_norm": 2.8807024586502226,
+      "learning_rate": 1e-06,
+      "loss": 0.0652,
+      "step": 8988
+    },
+    {
+      "epoch": 1.7189023807247348,
+      "grad_norm": 2.432159343989296,
+      "learning_rate": 1e-06,
+      "loss": 0.0305,
+      "step": 8989
+    },
+    {
+      "epoch": 1.71909360359499,
+      "grad_norm": 2.646861279664853,
+      "learning_rate": 1e-06,
+      "loss": 0.0511,
+      "step": 8990
+    },
+    {
+      "epoch": 1.7192848264652452,
+      "grad_norm": 6.3584058618304455,
+      "learning_rate": 1e-06,
+      "loss": 0.0904,
+      "step": 8991
+    },
+    {
+      "epoch": 1.7194760493355006,
+      "grad_norm": 5.278427543574286,
+      "learning_rate": 1e-06,
+      "loss": 0.0522,
+      "step": 8992
+    },
+    {
+      "epoch": 1.7196672722057558,
+      "grad_norm": 6.913725108734827,
+      "learning_rate": 1e-06,
+      "loss": 0.0973,
+      "step": 8993
+    },
+    {
+      "epoch": 1.7198584950760112,
+      "grad_norm": 4.988657675022367,
+      "learning_rate": 1e-06,
+      "loss": 0.437,
+      "step": 8994
+    },
+    {
+      "epoch": 1.7200497179462664,
+      "grad_norm": 3.5036298457817203,
+      "learning_rate": 1e-06,
+      "loss": 0.3803,
+      "step": 8995
+    },
+    {
+      "epoch": 1.7202409408165216,
+      "grad_norm": 2.9414472640421607,
+      "learning_rate": 1e-06,
+      "loss": 0.1664,
+      "step": 8996
+    },
+    {
+      "epoch": 1.720432163686777,
+      "grad_norm": 2.1379258339546015,
+      "learning_rate": 1e-06,
+      "loss": 0.1256,
+      "step": 8997
+    },
+    {
+      "epoch": 1.7206233865570322,
+      "grad_norm": 2.269853001174776,
+      "learning_rate": 1e-06,
+      "loss": 0.1096,
+      "step": 8998
+    },
+    {
+      "epoch": 1.7208146094272876,
+      "grad_norm": 2.729236242751807,
+      "learning_rate": 1e-06,
+      "loss": 0.1619,
+      "step": 8999
+    },
+    {
+      "epoch": 1.7210058322975428,
+      "grad_norm": 1.232004859849708,
+      "learning_rate": 1e-06,
+      "loss": 0.02,
+      "step": 9000
+    },
+    {
+      "epoch": 1.7210058322975428,
+      "eval_runtime": 757.5802,
+      "eval_samples_per_second": 2.025,
+      "eval_steps_per_second": 0.507,
+      "step": 9000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 26145,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.959116055984472e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}