diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,22687 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 3235,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0015455950540958269,
+      "grad_norm": 5.26565455238215,
+      "learning_rate": 2.469135802469136e-07,
+      "loss": 0.8853,
+      "step": 1
+    },
+    {
+      "epoch": 0.0030911901081916537,
+      "grad_norm": 5.282847760707683,
+      "learning_rate": 4.938271604938272e-07,
+      "loss": 0.8908,
+      "step": 2
+    },
+    {
+      "epoch": 0.00463678516228748,
+      "grad_norm": 5.179577811088595,
+      "learning_rate": 7.407407407407407e-07,
+      "loss": 0.8789,
+      "step": 3
+    },
+    {
+      "epoch": 0.0061823802163833074,
+      "grad_norm": 5.14965313767903,
+      "learning_rate": 9.876543209876544e-07,
+      "loss": 0.8745,
+      "step": 4
+    },
+    {
+      "epoch": 0.0077279752704791345,
+      "grad_norm": 5.0926621078802095,
+      "learning_rate": 1.234567901234568e-06,
+      "loss": 0.8697,
+      "step": 5
+    },
+    {
+      "epoch": 0.00927357032457496,
+      "grad_norm": 4.936823049639591,
+      "learning_rate": 1.4814814814814815e-06,
+      "loss": 0.8677,
+      "step": 6
+    },
+    {
+      "epoch": 0.010819165378670788,
+      "grad_norm": 4.800318401151413,
+      "learning_rate": 1.7283950617283952e-06,
+      "loss": 0.8667,
+      "step": 7
+    },
+    {
+      "epoch": 0.012364760432766615,
+      "grad_norm": 3.9449223561641844,
+      "learning_rate": 1.9753086419753087e-06,
+      "loss": 0.8235,
+      "step": 8
+    },
+    {
+      "epoch": 0.013910355486862442,
+      "grad_norm": 3.7105633047486504,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 0.8259,
+      "step": 9
+    },
+    {
+      "epoch": 0.015455950540958269,
+      "grad_norm": 2.198523539340047,
+      "learning_rate": 2.469135802469136e-06,
+      "loss": 0.7827,
+      "step": 10
+    },
+    {
+      "epoch": 0.017001545595054096,
+      "grad_norm": 2.0963348737509295,
+      "learning_rate": 2.7160493827160496e-06,
+      "loss": 0.776,
+      "step": 11
+    },
+    {
+      "epoch": 0.01854714064914992,
+      "grad_norm": 1.8699088033519315,
+      "learning_rate": 2.962962962962963e-06,
+      "loss": 0.7768,
+      "step": 12
+    },
+    {
+      "epoch": 0.02009273570324575,
+      "grad_norm": 2.7286736237238034,
+      "learning_rate": 3.2098765432098767e-06,
+      "loss": 0.7374,
+      "step": 13
+    },
+    {
+      "epoch": 0.021638330757341576,
+      "grad_norm": 3.3504502146745563,
+      "learning_rate": 3.4567901234567904e-06,
+      "loss": 0.7616,
+      "step": 14
+    },
+    {
+      "epoch": 0.023183925811437404,
+      "grad_norm": 3.3641769676772864,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 0.7633,
+      "step": 15
+    },
+    {
+      "epoch": 0.02472952086553323,
+      "grad_norm": 3.147552036611318,
+      "learning_rate": 3.9506172839506175e-06,
+      "loss": 0.7376,
+      "step": 16
+    },
+    {
+      "epoch": 0.02627511591962906,
+      "grad_norm": 2.8976148119744427,
+      "learning_rate": 4.197530864197531e-06,
+      "loss": 0.736,
+      "step": 17
+    },
+    {
+      "epoch": 0.027820710973724884,
+      "grad_norm": 2.081476960627112,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.6908,
+      "step": 18
+    },
+    {
+      "epoch": 0.02936630602782071,
+      "grad_norm": 1.806761897373877,
+      "learning_rate": 4.691358024691358e-06,
+      "loss": 0.689,
+      "step": 19
+    },
+    {
+      "epoch": 0.030911901081916538,
+      "grad_norm": 1.390550633941569,
+      "learning_rate": 4.938271604938272e-06,
+      "loss": 0.6851,
+      "step": 20
+    },
+    {
+      "epoch": 0.03245749613601236,
+      "grad_norm": 1.1260914755524882,
+      "learning_rate": 5.185185185185185e-06,
+      "loss": 0.6689,
+      "step": 21
+    },
+    {
+      "epoch": 0.03400309119010819,
+      "grad_norm": 1.1828994926159446,
+      "learning_rate": 5.432098765432099e-06,
+      "loss": 0.6484,
+      "step": 22
+    },
+    {
+      "epoch": 0.03554868624420402,
+      "grad_norm": 1.2488339030387958,
+      "learning_rate": 5.6790123456790125e-06,
+      "loss": 0.6608,
+      "step": 23
+    },
+    {
+      "epoch": 0.03709428129829984,
+      "grad_norm": 1.1994124081567479,
+      "learning_rate": 5.925925925925926e-06,
+      "loss": 0.6476,
+      "step": 24
+    },
+    {
+      "epoch": 0.03863987635239567,
+      "grad_norm": 1.0594113595099088,
+      "learning_rate": 6.17283950617284e-06,
+      "loss": 0.6384,
+      "step": 25
+    },
+    {
+      "epoch": 0.0401854714064915,
+      "grad_norm": 0.8136334420177265,
+      "learning_rate": 6.419753086419753e-06,
+      "loss": 0.6363,
+      "step": 26
+    },
+    {
+      "epoch": 0.04173106646058733,
+      "grad_norm": 0.8515726939745646,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.6326,
+      "step": 27
+    },
+    {
+      "epoch": 0.04327666151468315,
+      "grad_norm": 0.8875870370322879,
+      "learning_rate": 6.913580246913581e-06,
+      "loss": 0.6229,
+      "step": 28
+    },
+    {
+      "epoch": 0.04482225656877898,
+      "grad_norm": 0.768439654173595,
+      "learning_rate": 7.160493827160494e-06,
+      "loss": 0.6167,
+      "step": 29
+    },
+    {
+      "epoch": 0.04636785162287481,
+      "grad_norm": 0.5189099049678193,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.5938,
+      "step": 30
+    },
+    {
+      "epoch": 0.04791344667697063,
+      "grad_norm": 0.6395256514127601,
+      "learning_rate": 7.654320987654322e-06,
+      "loss": 0.6152,
+      "step": 31
+    },
+    {
+      "epoch": 0.04945904173106646,
+      "grad_norm": 0.746617550878664,
+      "learning_rate": 7.901234567901235e-06,
+      "loss": 0.5969,
+      "step": 32
+    },
+    {
+      "epoch": 0.05100463678516229,
+      "grad_norm": 0.5858126224692574,
+      "learning_rate": 8.148148148148148e-06,
+      "loss": 0.6024,
+      "step": 33
+    },
+    {
+      "epoch": 0.05255023183925812,
+      "grad_norm": 0.39467639483180406,
+      "learning_rate": 8.395061728395062e-06,
+      "loss": 0.5956,
+      "step": 34
+    },
+    {
+      "epoch": 0.05409582689335394,
+      "grad_norm": 0.5640256567279153,
+      "learning_rate": 8.641975308641975e-06,
+      "loss": 0.5821,
+      "step": 35
+    },
+    {
+      "epoch": 0.05564142194744977,
+      "grad_norm": 0.5987573689680895,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.5833,
+      "step": 36
+    },
+    {
+      "epoch": 0.0571870170015456,
+      "grad_norm": 0.44066358802919997,
+      "learning_rate": 9.135802469135803e-06,
+      "loss": 0.591,
+      "step": 37
+    },
+    {
+      "epoch": 0.05873261205564142,
+      "grad_norm": 0.3852388525791168,
+      "learning_rate": 9.382716049382717e-06,
+      "loss": 0.5778,
+      "step": 38
+    },
+    {
+      "epoch": 0.06027820710973725,
+      "grad_norm": 0.5098271176851854,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 0.5713,
+      "step": 39
+    },
+    {
+      "epoch": 0.061823802163833076,
+      "grad_norm": 0.39503936645354354,
+      "learning_rate": 9.876543209876543e-06,
+      "loss": 0.5714,
+      "step": 40
+    },
+    {
+      "epoch": 0.0633693972179289,
+      "grad_norm": 0.32905075142191664,
+      "learning_rate": 1.0123456790123458e-05,
+      "loss": 0.5728,
+      "step": 41
+    },
+    {
+      "epoch": 0.06491499227202473,
+      "grad_norm": 0.3840798410090591,
+      "learning_rate": 1.037037037037037e-05,
+      "loss": 0.5661,
+      "step": 42
+    },
+    {
+      "epoch": 0.06646058732612056,
+      "grad_norm": 0.38897396040812976,
+      "learning_rate": 1.0617283950617285e-05,
+      "loss": 0.5822,
+      "step": 43
+    },
+    {
+      "epoch": 0.06800618238021638,
+      "grad_norm": 0.29495303808913936,
+      "learning_rate": 1.0864197530864198e-05,
+      "loss": 0.5543,
+      "step": 44
+    },
+    {
+      "epoch": 0.0695517774343122,
+      "grad_norm": 0.3253252000829262,
+      "learning_rate": 1.1111111111111113e-05,
+      "loss": 0.5593,
+      "step": 45
+    },
+    {
+      "epoch": 0.07109737248840804,
+      "grad_norm": 0.34116573815145546,
+      "learning_rate": 1.1358024691358025e-05,
+      "loss": 0.5491,
+      "step": 46
+    },
+    {
+      "epoch": 0.07264296754250386,
+      "grad_norm": 0.24621781702597118,
+      "learning_rate": 1.160493827160494e-05,
+      "loss": 0.5619,
+      "step": 47
+    },
+    {
+      "epoch": 0.07418856259659969,
+      "grad_norm": 0.369173395538977,
+      "learning_rate": 1.1851851851851852e-05,
+      "loss": 0.5563,
+      "step": 48
+    },
+    {
+      "epoch": 0.07573415765069552,
+      "grad_norm": 0.2782501293355875,
+      "learning_rate": 1.2098765432098767e-05,
+      "loss": 0.5456,
+      "step": 49
+    },
+    {
+      "epoch": 0.07727975270479134,
+      "grad_norm": 0.29617283022535873,
+      "learning_rate": 1.234567901234568e-05,
+      "loss": 0.5553,
+      "step": 50
+    },
+    {
+      "epoch": 0.07882534775888717,
+      "grad_norm": 0.3110918507531697,
+      "learning_rate": 1.2592592592592593e-05,
+      "loss": 0.5483,
+      "step": 51
+    },
+    {
+      "epoch": 0.080370942812983,
+      "grad_norm": 0.27084532621368845,
+      "learning_rate": 1.2839506172839507e-05,
+      "loss": 0.5547,
+      "step": 52
+    },
+    {
+      "epoch": 0.08191653786707882,
+      "grad_norm": 0.2457944469090477,
+      "learning_rate": 1.3086419753086422e-05,
+      "loss": 0.5347,
+      "step": 53
+    },
+    {
+      "epoch": 0.08346213292117466,
+      "grad_norm": 0.2541436303825332,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.5428,
+      "step": 54
+    },
+    {
+      "epoch": 0.08500772797527048,
+      "grad_norm": 0.2376504860719097,
+      "learning_rate": 1.3580246913580248e-05,
+      "loss": 0.5388,
+      "step": 55
+    },
+    {
+      "epoch": 0.0865533230293663,
+      "grad_norm": 0.24597957657408892,
+      "learning_rate": 1.3827160493827162e-05,
+      "loss": 0.5513,
+      "step": 56
+    },
+    {
+      "epoch": 0.08809891808346214,
+      "grad_norm": 0.24278390126556001,
+      "learning_rate": 1.4074074074074075e-05,
+      "loss": 0.5549,
+      "step": 57
+    },
+    {
+      "epoch": 0.08964451313755796,
+      "grad_norm": 0.20020826855660456,
+      "learning_rate": 1.4320987654320988e-05,
+      "loss": 0.5328,
+      "step": 58
+    },
+    {
+      "epoch": 0.09119010819165378,
+      "grad_norm": 0.25383977984023004,
+      "learning_rate": 1.4567901234567903e-05,
+      "loss": 0.5505,
+      "step": 59
+    },
+    {
+      "epoch": 0.09273570324574962,
+      "grad_norm": 0.2340720421695047,
+      "learning_rate": 1.4814814814814815e-05,
+      "loss": 0.5386,
+      "step": 60
+    },
+    {
+      "epoch": 0.09428129829984544,
+      "grad_norm": 0.25602875868447883,
+      "learning_rate": 1.506172839506173e-05,
+      "loss": 0.5367,
+      "step": 61
+    },
+    {
+      "epoch": 0.09582689335394126,
+      "grad_norm": 0.27002467070309155,
+      "learning_rate": 1.5308641975308643e-05,
+      "loss": 0.5397,
+      "step": 62
+    },
+    {
+      "epoch": 0.0973724884080371,
+      "grad_norm": 0.3234378757167277,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 0.5358,
+      "step": 63
+    },
+    {
+      "epoch": 0.09891808346213292,
+      "grad_norm": 0.28118242669639826,
+      "learning_rate": 1.580246913580247e-05,
+      "loss": 0.5366,
+      "step": 64
+    },
+    {
+      "epoch": 0.10046367851622875,
+      "grad_norm": 0.24122559938222943,
+      "learning_rate": 1.6049382716049385e-05,
+      "loss": 0.5361,
+      "step": 65
+    },
+    {
+      "epoch": 0.10200927357032458,
+      "grad_norm": 0.28007867616979454,
+      "learning_rate": 1.6296296296296297e-05,
+      "loss": 0.5466,
+      "step": 66
+    },
+    {
+      "epoch": 0.1035548686244204,
+      "grad_norm": 0.251129640949984,
+      "learning_rate": 1.654320987654321e-05,
+      "loss": 0.5303,
+      "step": 67
+    },
+    {
+      "epoch": 0.10510046367851623,
+      "grad_norm": 0.22565320702718436,
+      "learning_rate": 1.6790123456790123e-05,
+      "loss": 0.5244,
+      "step": 68
+    },
+    {
+      "epoch": 0.10664605873261206,
+      "grad_norm": 0.24920566874778105,
+      "learning_rate": 1.7037037037037038e-05,
+      "loss": 0.5332,
+      "step": 69
+    },
+    {
+      "epoch": 0.10819165378670788,
+      "grad_norm": 0.29209867356120256,
+      "learning_rate": 1.728395061728395e-05,
+      "loss": 0.5302,
+      "step": 70
+    },
+    {
+      "epoch": 0.10973724884080371,
+      "grad_norm": 0.27057400482624655,
+      "learning_rate": 1.7530864197530865e-05,
+      "loss": 0.5195,
+      "step": 71
+    },
+    {
+      "epoch": 0.11128284389489954,
+      "grad_norm": 0.35167113888602347,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 0.5163,
+      "step": 72
+    },
+    {
+      "epoch": 0.11282843894899536,
+      "grad_norm": 0.39760550684028373,
+      "learning_rate": 1.802469135802469e-05,
+      "loss": 0.5178,
+      "step": 73
+    },
+    {
+      "epoch": 0.1143740340030912,
+      "grad_norm": 0.6784824026653671,
+      "learning_rate": 1.8271604938271607e-05,
+      "loss": 0.5341,
+      "step": 74
+    },
+    {
+      "epoch": 0.11591962905718702,
+      "grad_norm": 0.9618391802990648,
+      "learning_rate": 1.851851851851852e-05,
+      "loss": 0.5337,
+      "step": 75
+    },
+    {
+      "epoch": 0.11746522411128284,
+      "grad_norm": 1.207818633909923,
+      "learning_rate": 1.8765432098765433e-05,
+      "loss": 0.5251,
+      "step": 76
+    },
+    {
+      "epoch": 0.11901081916537867,
+      "grad_norm": 0.4633663432837794,
+      "learning_rate": 1.901234567901235e-05,
+      "loss": 0.5189,
+      "step": 77
+    },
+    {
+      "epoch": 0.1205564142194745,
+      "grad_norm": 0.8571543214971838,
+      "learning_rate": 1.925925925925926e-05,
+      "loss": 0.5213,
+      "step": 78
+    },
+    {
+      "epoch": 0.12210200927357033,
+      "grad_norm": 1.0464179386041605,
+      "learning_rate": 1.9506172839506175e-05,
+      "loss": 0.5327,
+      "step": 79
+    },
+    {
+      "epoch": 0.12364760432766615,
+      "grad_norm": 0.6030795060182889,
+      "learning_rate": 1.9753086419753087e-05,
+      "loss": 0.5108,
+      "step": 80
+    },
+    {
+      "epoch": 0.125193199381762,
+      "grad_norm": 0.8282902108432075,
+      "learning_rate": 2e-05,
+      "loss": 0.5248,
+      "step": 81
+    },
+    {
+      "epoch": 0.1267387944358578,
+      "grad_norm": 0.6834029854964587,
+      "learning_rate": 2.0246913580246917e-05,
+      "loss": 0.5272,
+      "step": 82
+    },
+    {
+      "epoch": 0.12828438948995363,
+      "grad_norm": 0.9070124398991254,
+      "learning_rate": 2.049382716049383e-05,
+      "loss": 0.5181,
+      "step": 83
+    },
+    {
+      "epoch": 0.12982998454404945,
+      "grad_norm": 0.7509309688845062,
+      "learning_rate": 2.074074074074074e-05,
+      "loss": 0.5108,
+      "step": 84
+    },
+    {
+      "epoch": 0.13137557959814528,
+      "grad_norm": 0.5603999059650706,
+      "learning_rate": 2.0987654320987655e-05,
+      "loss": 0.5101,
+      "step": 85
+    },
+    {
+      "epoch": 0.13292117465224113,
+      "grad_norm": 0.8636069290455202,
+      "learning_rate": 2.123456790123457e-05,
+      "loss": 0.5146,
+      "step": 86
+    },
+    {
+      "epoch": 0.13446676970633695,
+      "grad_norm": 0.8045336877644039,
+      "learning_rate": 2.1481481481481485e-05,
+      "loss": 0.5154,
+      "step": 87
+    },
+    {
+      "epoch": 0.13601236476043277,
+      "grad_norm": 0.6023948125967806,
+      "learning_rate": 2.1728395061728397e-05,
+      "loss": 0.5083,
+      "step": 88
+    },
+    {
+      "epoch": 0.1375579598145286,
+      "grad_norm": 0.7704598071209162,
+      "learning_rate": 2.197530864197531e-05,
+      "loss": 0.5169,
+      "step": 89
+    },
+    {
+      "epoch": 0.1391035548686244,
+      "grad_norm": 0.741700249163537,
+      "learning_rate": 2.2222222222222227e-05,
+      "loss": 0.5198,
+      "step": 90
+    },
+    {
+      "epoch": 0.14064914992272023,
+      "grad_norm": 0.4950278030178324,
+      "learning_rate": 2.246913580246914e-05,
+      "loss": 0.5078,
+      "step": 91
+    },
+    {
+      "epoch": 0.14219474497681608,
+      "grad_norm": 0.5837929907445218,
+      "learning_rate": 2.271604938271605e-05,
+      "loss": 0.5036,
+      "step": 92
+    },
+    {
+      "epoch": 0.1437403400309119,
+      "grad_norm": 0.6238323145413748,
+      "learning_rate": 2.2962962962962965e-05,
+      "loss": 0.4985,
+      "step": 93
+    },
+    {
+      "epoch": 0.14528593508500773,
+      "grad_norm": 0.5532970702371948,
+      "learning_rate": 2.320987654320988e-05,
+      "loss": 0.5134,
+      "step": 94
+    },
+    {
+      "epoch": 0.14683153013910355,
+      "grad_norm": 0.8342787251297591,
+      "learning_rate": 2.3456790123456795e-05,
+      "loss": 0.5168,
+      "step": 95
+    },
+    {
+      "epoch": 0.14837712519319937,
+      "grad_norm": 0.8824818854861024,
+      "learning_rate": 2.3703703703703703e-05,
+      "loss": 0.5091,
+      "step": 96
+    },
+    {
+      "epoch": 0.14992272024729522,
+      "grad_norm": 1.288218263701589,
+      "learning_rate": 2.3950617283950618e-05,
+      "loss": 0.5228,
+      "step": 97
+    },
+    {
+      "epoch": 0.15146831530139104,
+      "grad_norm": 0.7289192134121079,
+      "learning_rate": 2.4197530864197533e-05,
+      "loss": 0.5027,
+      "step": 98
+    },
+    {
+      "epoch": 0.15301391035548687,
+      "grad_norm": 0.6904973969603343,
+      "learning_rate": 2.444444444444445e-05,
+      "loss": 0.5073,
+      "step": 99
+    },
+    {
+      "epoch": 0.1545595054095827,
+      "grad_norm": 1.0630399683570595,
+      "learning_rate": 2.469135802469136e-05,
+      "loss": 0.5096,
+      "step": 100
+    },
+    {
+      "epoch": 0.1561051004636785,
+      "grad_norm": 1.143477539924475,
+      "learning_rate": 2.4938271604938275e-05,
+      "loss": 0.5082,
+      "step": 101
+    },
+    {
+      "epoch": 0.15765069551777433,
+      "grad_norm": 0.8679214652629397,
+      "learning_rate": 2.5185185185185187e-05,
+      "loss": 0.5074,
+      "step": 102
+    },
+    {
+      "epoch": 0.15919629057187018,
+      "grad_norm": 1.1787592552800283,
+      "learning_rate": 2.54320987654321e-05,
+      "loss": 0.4968,
+      "step": 103
+    },
+    {
+      "epoch": 0.160741885625966,
+      "grad_norm": 1.0427291703445587,
+      "learning_rate": 2.5679012345679013e-05,
+      "loss": 0.5009,
+      "step": 104
+    },
+    {
+      "epoch": 0.16228748068006182,
+      "grad_norm": 0.847469777194458,
+      "learning_rate": 2.5925925925925928e-05,
+      "loss": 0.5145,
+      "step": 105
+    },
+    {
+      "epoch": 0.16383307573415765,
+      "grad_norm": 1.032596111298762,
+      "learning_rate": 2.6172839506172843e-05,
+      "loss": 0.4992,
+      "step": 106
+    },
+    {
+      "epoch": 0.16537867078825347,
+      "grad_norm": 0.7112034821814768,
+      "learning_rate": 2.641975308641976e-05,
+      "loss": 0.5096,
+      "step": 107
+    },
+    {
+      "epoch": 0.16692426584234932,
+      "grad_norm": 0.7754993141540937,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.4907,
+      "step": 108
+    },
+    {
+      "epoch": 0.16846986089644514,
+      "grad_norm": 0.7610858355351189,
+      "learning_rate": 2.691358024691358e-05,
+      "loss": 0.5065,
+      "step": 109
+    },
+    {
+      "epoch": 0.17001545595054096,
+      "grad_norm": 0.7407566934069416,
+      "learning_rate": 2.7160493827160497e-05,
+      "loss": 0.5025,
+      "step": 110
+    },
+    {
+      "epoch": 0.17156105100463678,
+      "grad_norm": 0.8022655281725392,
+      "learning_rate": 2.740740740740741e-05,
+      "loss": 0.4963,
+      "step": 111
+    },
+    {
+      "epoch": 0.1731066460587326,
+      "grad_norm": 0.7631247318207706,
+      "learning_rate": 2.7654320987654323e-05,
+      "loss": 0.5018,
+      "step": 112
+    },
+    {
+      "epoch": 0.17465224111282843,
+      "grad_norm": 1.078272532566959,
+      "learning_rate": 2.7901234567901235e-05,
+      "loss": 0.4933,
+      "step": 113
+    },
+    {
+      "epoch": 0.17619783616692428,
+      "grad_norm": 0.9770429338531161,
+      "learning_rate": 2.814814814814815e-05,
+      "loss": 0.5108,
+      "step": 114
+    },
+    {
+      "epoch": 0.1777434312210201,
+      "grad_norm": 1.209883015711933,
+      "learning_rate": 2.8395061728395065e-05,
+      "loss": 0.5048,
+      "step": 115
+    },
+    {
+      "epoch": 0.17928902627511592,
+      "grad_norm": 0.6568854476685991,
+      "learning_rate": 2.8641975308641977e-05,
+      "loss": 0.5063,
+      "step": 116
+    },
+    {
+      "epoch": 0.18083462132921174,
+      "grad_norm": 0.6369454626125819,
+      "learning_rate": 2.888888888888889e-05,
+      "loss": 0.49,
+      "step": 117
+    },
+    {
+      "epoch": 0.18238021638330756,
+      "grad_norm": 0.6945493811214182,
+      "learning_rate": 2.9135802469135807e-05,
+      "loss": 0.5008,
+      "step": 118
+    },
+    {
+      "epoch": 0.1839258114374034,
+      "grad_norm": 0.7084610539302401,
+      "learning_rate": 2.938271604938272e-05,
+      "loss": 0.4928,
+      "step": 119
+    },
+    {
+      "epoch": 0.18547140649149924,
+      "grad_norm": 1.0347139753934802,
+      "learning_rate": 2.962962962962963e-05,
+      "loss": 0.503,
+      "step": 120
+    },
+    {
+      "epoch": 0.18701700154559506,
+      "grad_norm": 1.0981919160573457,
+      "learning_rate": 2.9876543209876545e-05,
+      "loss": 0.5032,
+      "step": 121
+    },
+    {
+      "epoch": 0.18856259659969088,
+      "grad_norm": 0.8082068804313435,
+      "learning_rate": 3.012345679012346e-05,
+      "loss": 0.5002,
+      "step": 122
+    },
+    {
+      "epoch": 0.1901081916537867,
+      "grad_norm": 0.9238973243615479,
+      "learning_rate": 3.0370370370370375e-05,
+      "loss": 0.4982,
+      "step": 123
+    },
+    {
+      "epoch": 0.19165378670788252,
+      "grad_norm": 0.9631424104539678,
+      "learning_rate": 3.061728395061729e-05,
+      "loss": 0.5021,
+      "step": 124
+    },
+    {
+      "epoch": 0.19319938176197837,
+      "grad_norm": 1.1086567942544703,
+      "learning_rate": 3.08641975308642e-05,
+      "loss": 0.4983,
+      "step": 125
+    },
+    {
+      "epoch": 0.1947449768160742,
+      "grad_norm": 1.1628943248183918,
+      "learning_rate": 3.111111111111112e-05,
+      "loss": 0.5024,
+      "step": 126
+    },
+    {
+      "epoch": 0.19629057187017002,
+      "grad_norm": 0.8257350753088295,
+      "learning_rate": 3.135802469135803e-05,
+      "loss": 0.4998,
+      "step": 127
+    },
+    {
+      "epoch": 0.19783616692426584,
+      "grad_norm": 0.8618783547766707,
+      "learning_rate": 3.160493827160494e-05,
+      "loss": 0.4922,
+      "step": 128
+    },
+    {
+      "epoch": 0.19938176197836166,
+      "grad_norm": 1.241169327864676,
+      "learning_rate": 3.185185185185185e-05,
+      "loss": 0.4999,
+      "step": 129
+    },
+    {
+      "epoch": 0.2009273570324575,
+      "grad_norm": 0.828132554003403,
+      "learning_rate": 3.209876543209877e-05,
+      "loss": 0.5026,
+      "step": 130
+    },
+    {
+      "epoch": 0.20247295208655333,
+      "grad_norm": 0.7793563201875642,
+      "learning_rate": 3.234567901234568e-05,
+      "loss": 0.4886,
+      "step": 131
+    },
+    {
+      "epoch": 0.20401854714064915,
+      "grad_norm": 0.9675464742848618,
+      "learning_rate": 3.259259259259259e-05,
+      "loss": 0.4915,
+      "step": 132
+    },
+    {
+      "epoch": 0.20556414219474498,
+      "grad_norm": 0.8431076894503557,
+      "learning_rate": 3.2839506172839505e-05,
+      "loss": 0.5025,
+      "step": 133
+    },
+    {
+      "epoch": 0.2071097372488408,
+      "grad_norm": 0.8812947158826483,
+      "learning_rate": 3.308641975308642e-05,
+      "loss": 0.4903,
+      "step": 134
+    },
+    {
+      "epoch": 0.20865533230293662,
+      "grad_norm": 1.3213441436820172,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.4997,
+      "step": 135
+    },
+    {
+      "epoch": 0.21020092735703247,
+      "grad_norm": 0.5642206183809721,
+      "learning_rate": 3.3580246913580247e-05,
+      "loss": 0.4982,
+      "step": 136
+    },
+    {
+      "epoch": 0.2117465224111283,
+      "grad_norm": 1.3173420862661687,
+      "learning_rate": 3.3827160493827165e-05,
+      "loss": 0.502,
+      "step": 137
+    },
+    {
+      "epoch": 0.2132921174652241,
+      "grad_norm": 0.8513440164344224,
+      "learning_rate": 3.4074074074074077e-05,
+      "loss": 0.4971,
+      "step": 138
+    },
+    {
+      "epoch": 0.21483771251931993,
+      "grad_norm": 0.8672711786334716,
+      "learning_rate": 3.4320987654320995e-05,
+      "loss": 0.489,
+      "step": 139
+    },
+    {
+      "epoch": 0.21638330757341576,
+      "grad_norm": 1.0621643679180979,
+      "learning_rate": 3.45679012345679e-05,
+      "loss": 0.4991,
+      "step": 140
+    },
+    {
+      "epoch": 0.21792890262751158,
+      "grad_norm": 0.7921164646769278,
+      "learning_rate": 3.481481481481482e-05,
+      "loss": 0.4833,
+      "step": 141
+    },
+    {
+      "epoch": 0.21947449768160743,
+      "grad_norm": 1.1313070889701453,
+      "learning_rate": 3.506172839506173e-05,
+      "loss": 0.4983,
+      "step": 142
+    },
+    {
+      "epoch": 0.22102009273570325,
+      "grad_norm": 0.979446436116847,
+      "learning_rate": 3.530864197530865e-05,
+      "loss": 0.4888,
+      "step": 143
+    },
+    {
+      "epoch": 0.22256568778979907,
+      "grad_norm": 0.9631416113357064,
+      "learning_rate": 3.555555555555555e-05,
+      "loss": 0.4973,
+      "step": 144
+    },
+    {
+      "epoch": 0.2241112828438949,
+      "grad_norm": 0.8774746523842448,
+      "learning_rate": 3.580246913580247e-05,
+      "loss": 0.4909,
+      "step": 145
+    },
+    {
+      "epoch": 0.22565687789799072,
+      "grad_norm": 0.9473774265327042,
+      "learning_rate": 3.604938271604938e-05,
+      "loss": 0.4867,
+      "step": 146
+    },
+    {
+      "epoch": 0.22720247295208656,
+      "grad_norm": 1.2382081588447356,
+      "learning_rate": 3.62962962962963e-05,
+      "loss": 0.4895,
+      "step": 147
+    },
+    {
+      "epoch": 0.2287480680061824,
+      "grad_norm": 0.9373600980716077,
+      "learning_rate": 3.654320987654321e-05,
+      "loss": 0.4906,
+      "step": 148
+    },
+    {
+      "epoch": 0.2302936630602782,
+      "grad_norm": 1.1641931805143952,
+      "learning_rate": 3.6790123456790125e-05,
+      "loss": 0.493,
+      "step": 149
+    },
+    {
+      "epoch": 0.23183925811437403,
+      "grad_norm": 0.6192917083811479,
+      "learning_rate": 3.703703703703704e-05,
+      "loss": 0.5018,
+      "step": 150
+    },
+    {
+      "epoch": 0.23338485316846985,
+      "grad_norm": 0.9568940117279771,
+      "learning_rate": 3.7283950617283955e-05,
+      "loss": 0.4859,
+      "step": 151
+    },
+    {
+      "epoch": 0.23493044822256567,
+      "grad_norm": 0.7558753315433521,
+      "learning_rate": 3.7530864197530867e-05,
+      "loss": 0.4963,
+      "step": 152
+    },
+    {
+      "epoch": 0.23647604327666152,
+      "grad_norm": 0.9212501384989878,
+      "learning_rate": 3.777777777777778e-05,
+      "loss": 0.4819,
+      "step": 153
+    },
+    {
+      "epoch": 0.23802163833075735,
+      "grad_norm": 1.1697128542170077,
+      "learning_rate": 3.80246913580247e-05,
+      "loss": 0.4823,
+      "step": 154
+    },
+    {
+      "epoch": 0.23956723338485317,
+      "grad_norm": 1.1974076561986966,
+      "learning_rate": 3.827160493827161e-05,
+      "loss": 0.4898,
+      "step": 155
+    },
+    {
+      "epoch": 0.241112828438949,
+      "grad_norm": 0.5721357290416084,
+      "learning_rate": 3.851851851851852e-05,
+      "loss": 0.4883,
+      "step": 156
+    },
+    {
+      "epoch": 0.2426584234930448,
+      "grad_norm": 0.8279428890745845,
+      "learning_rate": 3.876543209876543e-05,
+      "loss": 0.4913,
+      "step": 157
+    },
+    {
+      "epoch": 0.24420401854714066,
+      "grad_norm": 0.7905766725922289,
+      "learning_rate": 3.901234567901235e-05,
+      "loss": 0.4999,
+      "step": 158
+    },
+    {
+      "epoch": 0.24574961360123648,
+      "grad_norm": 1.1663187217487774,
+      "learning_rate": 3.925925925925926e-05,
+      "loss": 0.4985,
+      "step": 159
+    },
+    {
+      "epoch": 0.2472952086553323,
+      "grad_norm": 1.044068287922803,
+      "learning_rate": 3.950617283950617e-05,
+      "loss": 0.4805,
+      "step": 160
+    },
+    {
+      "epoch": 0.24884080370942813,
+      "grad_norm": 0.9571058865876857,
+      "learning_rate": 3.975308641975309e-05,
+      "loss": 0.4991,
+      "step": 161
+    },
+    {
+      "epoch": 0.250386398763524,
+      "grad_norm": 1.1566279067789822,
+      "learning_rate": 4e-05,
+      "loss": 0.4964,
+      "step": 162
+    },
+    {
+      "epoch": 0.25193199381761977,
+      "grad_norm": 1.1448282250848643,
+      "learning_rate": 4.0246913580246915e-05,
+      "loss": 0.498,
+      "step": 163
+    },
+    {
+      "epoch": 0.2534775888717156,
+      "grad_norm": 1.1203356595739242,
+      "learning_rate": 4.049382716049383e-05,
+      "loss": 0.4873,
+      "step": 164
+    },
+    {
+      "epoch": 0.2550231839258114,
+      "grad_norm": 1.0233529441473044,
+      "learning_rate": 4.0740740740740745e-05,
+      "loss": 0.4926,
+      "step": 165
+    },
+    {
+      "epoch": 0.25656877897990726,
+      "grad_norm": 0.9118723834779173,
+      "learning_rate": 4.098765432098766e-05,
+      "loss": 0.4803,
+      "step": 166
+    },
+    {
+      "epoch": 0.2581143740340031,
+      "grad_norm": 1.1742293487040738,
+      "learning_rate": 4.123456790123457e-05,
+      "loss": 0.4891,
+      "step": 167
+    },
+    {
+      "epoch": 0.2596599690880989,
+      "grad_norm": 0.9090936632777925,
+      "learning_rate": 4.148148148148148e-05,
+      "loss": 0.4814,
+      "step": 168
+    },
+    {
+      "epoch": 0.26120556414219476,
+      "grad_norm": 1.0102483704824479,
+      "learning_rate": 4.17283950617284e-05,
+      "loss": 0.4787,
+      "step": 169
+    },
+    {
+      "epoch": 0.26275115919629055,
+      "grad_norm": 1.0211857026705857,
+      "learning_rate": 4.197530864197531e-05,
+      "loss": 0.488,
+      "step": 170
+    },
+    {
+      "epoch": 0.2642967542503864,
+      "grad_norm": 1.1146681594826109,
+      "learning_rate": 4.222222222222223e-05,
+      "loss": 0.4888,
+      "step": 171
+    },
+    {
+      "epoch": 0.26584234930448225,
+      "grad_norm": 0.9537477496730035,
+      "learning_rate": 4.246913580246914e-05,
+      "loss": 0.4734,
+      "step": 172
+    },
+    {
+      "epoch": 0.26738794435857804,
+      "grad_norm": 1.043638066624262,
+      "learning_rate": 4.271604938271606e-05,
+      "loss": 0.4742,
+      "step": 173
+    },
+    {
+      "epoch": 0.2689335394126739,
+      "grad_norm": 1.4422937022582138,
+      "learning_rate": 4.296296296296297e-05,
+      "loss": 0.4959,
+      "step": 174
+    },
+    {
+      "epoch": 0.2704791344667697,
+      "grad_norm": 1.0299464445089963,
+      "learning_rate": 4.3209876543209875e-05,
+      "loss": 0.481,
+      "step": 175
+    },
+    {
+      "epoch": 0.27202472952086554,
+      "grad_norm": 1.1818391836210722,
+      "learning_rate": 4.345679012345679e-05,
+      "loss": 0.4847,
+      "step": 176
+    },
+    {
+      "epoch": 0.2735703245749614,
+      "grad_norm": 0.7618135468190563,
+      "learning_rate": 4.3703703703703705e-05,
+      "loss": 0.4759,
+      "step": 177
+    },
+    {
+      "epoch": 0.2751159196290572,
+      "grad_norm": 1.6171899862729473,
+      "learning_rate": 4.395061728395062e-05,
+      "loss": 0.4822,
+      "step": 178
+    },
+    {
+      "epoch": 0.27666151468315303,
+      "grad_norm": 0.5985280763844202,
+      "learning_rate": 4.4197530864197535e-05,
+      "loss": 0.4903,
+      "step": 179
+    },
+    {
+      "epoch": 0.2782071097372488,
+      "grad_norm": 1.014644632490285,
+      "learning_rate": 4.444444444444445e-05,
+      "loss": 0.4777,
+      "step": 180
+    },
+    {
+      "epoch": 0.2797527047913447,
+      "grad_norm": 1.4790130521335796,
+      "learning_rate": 4.4691358024691365e-05,
+      "loss": 0.4924,
+      "step": 181
+    },
+    {
+      "epoch": 0.28129829984544047,
+      "grad_norm": 0.9371239153663216,
+      "learning_rate": 4.493827160493828e-05,
+      "loss": 0.4903,
+      "step": 182
+    },
+    {
+      "epoch": 0.2828438948995363,
+      "grad_norm": 0.8819988307673926,
+      "learning_rate": 4.518518518518519e-05,
+      "loss": 0.4857,
+      "step": 183
+    },
+    {
+      "epoch": 0.28438948995363217,
+      "grad_norm": 1.2980448181187785,
+      "learning_rate": 4.54320987654321e-05,
+      "loss": 0.4895,
+      "step": 184
+    },
+    {
+      "epoch": 0.28593508500772796,
+      "grad_norm": 1.0937651336408682,
+      "learning_rate": 4.567901234567901e-05,
+      "loss": 0.4776,
+      "step": 185
+    },
+    {
+      "epoch": 0.2874806800618238,
+      "grad_norm": 1.1682285456048371,
+      "learning_rate": 4.592592592592593e-05,
+      "loss": 0.4917,
+      "step": 186
+    },
+    {
+      "epoch": 0.2890262751159196,
+      "grad_norm": 1.1571200068187466,
+      "learning_rate": 4.617283950617284e-05,
+      "loss": 0.4876,
+      "step": 187
+    },
+    {
+      "epoch": 0.29057187017001546,
+      "grad_norm": 0.728643662456322,
+      "learning_rate": 4.641975308641976e-05,
+      "loss": 0.4752,
+      "step": 188
+    },
+    {
+      "epoch": 0.2921174652241113,
+      "grad_norm": 1.5497374490680917,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 0.4821,
+      "step": 189
+    },
+    {
+      "epoch": 0.2936630602782071,
+      "grad_norm": 0.6173816215753983,
+      "learning_rate": 4.691358024691359e-05,
+      "loss": 0.4824,
+      "step": 190
+    },
+    {
+      "epoch": 0.29520865533230295,
+      "grad_norm": 1.3179137124249143,
+      "learning_rate": 4.7160493827160495e-05,
+      "loss": 0.4944,
+      "step": 191
+    },
+    {
+      "epoch": 0.29675425038639874,
+      "grad_norm": 0.7994479560157165,
+      "learning_rate": 4.7407407407407407e-05,
+      "loss": 0.4809,
+      "step": 192
+    },
+    {
+      "epoch": 0.2982998454404946,
+      "grad_norm": 1.2701249362355918,
+      "learning_rate": 4.7654320987654325e-05,
+      "loss": 0.482,
+      "step": 193
+    },
+    {
+      "epoch": 0.29984544049459044,
+      "grad_norm": 0.7512432473101126,
+      "learning_rate": 4.7901234567901237e-05,
+      "loss": 0.4839,
+      "step": 194
+    },
+    {
+      "epoch": 0.30139103554868624,
+      "grad_norm": 0.929048702574481,
+      "learning_rate": 4.8148148148148155e-05,
+      "loss": 0.4824,
+      "step": 195
+    },
+    {
+      "epoch": 0.3029366306027821,
+      "grad_norm": 1.1727570739816218,
+      "learning_rate": 4.8395061728395067e-05,
+      "loss": 0.4844,
+      "step": 196
+    },
+    {
+      "epoch": 0.3044822256568779,
+      "grad_norm": 1.2592385923726135,
+      "learning_rate": 4.8641975308641985e-05,
+      "loss": 0.4883,
+      "step": 197
+    },
+    {
+      "epoch": 0.30602782071097373,
+      "grad_norm": 1.0274500278429952,
+      "learning_rate": 4.88888888888889e-05,
+      "loss": 0.4737,
+      "step": 198
+    },
+    {
+      "epoch": 0.3075734157650695,
+      "grad_norm": 1.0222774019121448,
+      "learning_rate": 4.91358024691358e-05,
+      "loss": 0.4799,
+      "step": 199
+    },
+    {
+      "epoch": 0.3091190108191654,
+      "grad_norm": 1.0732342293445107,
+      "learning_rate": 4.938271604938272e-05,
+      "loss": 0.4808,
+      "step": 200
+    },
+    {
+      "epoch": 0.3106646058732612,
+      "grad_norm": 1.1148226181132161,
+      "learning_rate": 4.962962962962963e-05,
+      "loss": 0.4764,
+      "step": 201
+    },
+    {
+      "epoch": 0.312210200927357,
+      "grad_norm": 0.8997498855655075,
+      "learning_rate": 4.987654320987655e-05,
+      "loss": 0.4872,
+      "step": 202
+    },
+    {
+      "epoch": 0.31375579598145287,
+      "grad_norm": 1.1821249594599401,
+      "learning_rate": 5.012345679012346e-05,
+      "loss": 0.4697,
+      "step": 203
+    },
+    {
+      "epoch": 0.31530139103554866,
+      "grad_norm": 1.0971633801119358,
+      "learning_rate": 5.037037037037037e-05,
+      "loss": 0.485,
+      "step": 204
+    },
+    {
+      "epoch": 0.3168469860896445,
+      "grad_norm": 0.8255153623316014,
+      "learning_rate": 5.061728395061729e-05,
+      "loss": 0.4744,
+      "step": 205
+    },
+    {
+      "epoch": 0.31839258114374036,
+      "grad_norm": 1.2668927846380267,
+      "learning_rate": 5.08641975308642e-05,
+      "loss": 0.4811,
+      "step": 206
+    },
+    {
+      "epoch": 0.31993817619783615,
+      "grad_norm": 1.2156929653202533,
+      "learning_rate": 5.111111111111111e-05,
+      "loss": 0.4823,
+      "step": 207
+    },
+    {
+      "epoch": 0.321483771251932,
+      "grad_norm": 0.8745911104546992,
+      "learning_rate": 5.1358024691358027e-05,
+      "loss": 0.4777,
+      "step": 208
+    },
+    {
+      "epoch": 0.3230293663060278,
+      "grad_norm": 1.7497488486354917,
+      "learning_rate": 5.160493827160494e-05,
+      "loss": 0.4847,
+      "step": 209
+    },
+    {
+      "epoch": 0.32457496136012365,
+      "grad_norm": 0.8228927010361928,
+      "learning_rate": 5.1851851851851857e-05,
+      "loss": 0.4717,
+      "step": 210
+    },
+    {
+      "epoch": 0.3261205564142195,
+      "grad_norm": 1.9889828694359233,
+      "learning_rate": 5.209876543209877e-05,
+      "loss": 0.4835,
+      "step": 211
+    },
+    {
+      "epoch": 0.3276661514683153,
+      "grad_norm": 1.287587405121619,
+      "learning_rate": 5.234567901234569e-05,
+      "loss": 0.4724,
+      "step": 212
+    },
+    {
+      "epoch": 0.32921174652241114,
+      "grad_norm": 1.6948902446976986,
+      "learning_rate": 5.25925925925926e-05,
+      "loss": 0.4756,
+      "step": 213
+    },
+    {
+      "epoch": 0.33075734157650694,
+      "grad_norm": 1.4119177281832886,
+      "learning_rate": 5.283950617283952e-05,
+      "loss": 0.4826,
+      "step": 214
+    },
+    {
+      "epoch": 0.3323029366306028,
+      "grad_norm": 1.5167781358660195,
+      "learning_rate": 5.308641975308642e-05,
+      "loss": 0.4811,
+      "step": 215
+    },
+    {
+      "epoch": 0.33384853168469864,
+      "grad_norm": 1.1308110283174948,
+      "learning_rate": 5.333333333333333e-05,
+      "loss": 0.4816,
+      "step": 216
+    },
+    {
+      "epoch": 0.33539412673879443,
+      "grad_norm": 1.2278896437711684,
+      "learning_rate": 5.358024691358025e-05,
+      "loss": 0.474,
+      "step": 217
+    },
+    {
+      "epoch": 0.3369397217928903,
+      "grad_norm": 0.8622626370979445,
+      "learning_rate": 5.382716049382716e-05,
+      "loss": 0.4744,
+      "step": 218
+    },
+    {
+      "epoch": 0.3384853168469861,
+      "grad_norm": 1.1986006504365891,
+      "learning_rate": 5.407407407407408e-05,
+      "loss": 0.4798,
+      "step": 219
+    },
+    {
+      "epoch": 0.3400309119010819,
+      "grad_norm": 0.6179519214399118,
+      "learning_rate": 5.432098765432099e-05,
+      "loss": 0.4683,
+      "step": 220
+    },
+    {
+      "epoch": 0.3415765069551777,
+      "grad_norm": 0.9582362210598186,
+      "learning_rate": 5.4567901234567905e-05,
+      "loss": 0.4721,
+      "step": 221
+    },
+    {
+      "epoch": 0.34312210200927357,
+      "grad_norm": 0.8775268539297884,
+      "learning_rate": 5.481481481481482e-05,
+      "loss": 0.4838,
+      "step": 222
+    },
+    {
+      "epoch": 0.3446676970633694,
+      "grad_norm": 1.0835329353629608,
+      "learning_rate": 5.506172839506173e-05,
+      "loss": 0.4836,
+      "step": 223
+    },
+    {
+      "epoch": 0.3462132921174652,
+      "grad_norm": 1.1439631376935295,
+      "learning_rate": 5.5308641975308647e-05,
+      "loss": 0.4767,
+      "step": 224
+    },
+    {
+      "epoch": 0.34775888717156106,
+      "grad_norm": 1.2716091535791791,
+      "learning_rate": 5.555555555555556e-05,
+      "loss": 0.4846,
+      "step": 225
+    },
+    {
+      "epoch": 0.34930448222565685,
+      "grad_norm": 0.8542190983120236,
+      "learning_rate": 5.580246913580247e-05,
+      "loss": 0.4736,
+      "step": 226
+    },
+    {
+      "epoch": 0.3508500772797527,
+      "grad_norm": 0.7750218925853332,
+      "learning_rate": 5.604938271604939e-05,
+      "loss": 0.4782,
+      "step": 227
+    },
+    {
+      "epoch": 0.35239567233384855,
+      "grad_norm": 0.8410137777186895,
+      "learning_rate": 5.62962962962963e-05,
+      "loss": 0.4659,
+      "step": 228
+    },
+    {
+      "epoch": 0.35394126738794435,
+      "grad_norm": 1.1347741397027802,
+      "learning_rate": 5.654320987654322e-05,
+      "loss": 0.4842,
+      "step": 229
+    },
+    {
+      "epoch": 0.3554868624420402,
+      "grad_norm": 1.2640994747112773,
+      "learning_rate": 5.679012345679013e-05,
+      "loss": 0.4811,
+      "step": 230
+    },
+    {
+      "epoch": 0.357032457496136,
+      "grad_norm": 1.0358540531545488,
+      "learning_rate": 5.7037037037037035e-05,
+      "loss": 0.4815,
+      "step": 231
+    },
+    {
+      "epoch": 0.35857805255023184,
+      "grad_norm": 1.1731444912874869,
+      "learning_rate": 5.728395061728395e-05,
+      "loss": 0.4824,
+      "step": 232
+    },
+    {
+      "epoch": 0.3601236476043277,
+      "grad_norm": 1.0963715252451798,
+      "learning_rate": 5.7530864197530865e-05,
+      "loss": 0.4819,
+      "step": 233
+    },
+    {
+      "epoch": 0.3616692426584235,
+      "grad_norm": 1.1234656582964098,
+      "learning_rate": 5.777777777777778e-05,
+      "loss": 0.4788,
+      "step": 234
+    },
+    {
+      "epoch": 0.36321483771251933,
+      "grad_norm": 1.0039863248210223,
+      "learning_rate": 5.8024691358024695e-05,
+      "loss": 0.4848,
+      "step": 235
+    },
+    {
+      "epoch": 0.36476043276661513,
+      "grad_norm": 0.8562522614103883,
+      "learning_rate": 5.827160493827161e-05,
+      "loss": 0.4713,
+      "step": 236
+    },
+    {
+      "epoch": 0.366306027820711,
+      "grad_norm": 1.0344076244948395,
+      "learning_rate": 5.8518518518518525e-05,
+      "loss": 0.4677,
+      "step": 237
+    },
+    {
+      "epoch": 0.3678516228748068,
+      "grad_norm": 1.034057279913179,
+      "learning_rate": 5.876543209876544e-05,
+      "loss": 0.4768,
+      "step": 238
+    },
+    {
+      "epoch": 0.3693972179289026,
+      "grad_norm": 1.7367687508499554,
+      "learning_rate": 5.901234567901235e-05,
+      "loss": 0.4852,
+      "step": 239
+    },
+    {
+      "epoch": 0.37094281298299847,
+      "grad_norm": 0.8118841495079787,
+      "learning_rate": 5.925925925925926e-05,
+      "loss": 0.4634,
+      "step": 240
+    },
+    {
+      "epoch": 0.37248840803709427,
+      "grad_norm": 1.471099919135472,
+      "learning_rate": 5.950617283950618e-05,
+      "loss": 0.4887,
+      "step": 241
+    },
+    {
+      "epoch": 0.3740340030911901,
+      "grad_norm": 1.1587198015371982,
+      "learning_rate": 5.975308641975309e-05,
+      "loss": 0.4717,
+      "step": 242
+    },
+    {
+      "epoch": 0.3755795981452859,
+      "grad_norm": 1.758659389813336,
+      "learning_rate": 6.000000000000001e-05,
+      "loss": 0.4895,
+      "step": 243
+    },
+    {
+      "epoch": 0.37712519319938176,
+      "grad_norm": 1.3315013513741276,
+      "learning_rate": 6.024691358024692e-05,
+      "loss": 0.4873,
+      "step": 244
+    },
+    {
+      "epoch": 0.3786707882534776,
+      "grad_norm": 1.7550563966474997,
+      "learning_rate": 6.049382716049383e-05,
+      "loss": 0.4754,
+      "step": 245
+    },
+    {
+      "epoch": 0.3802163833075734,
+      "grad_norm": 1.4584548318573913,
+      "learning_rate": 6.074074074074075e-05,
+      "loss": 0.4889,
+      "step": 246
+    },
+    {
+      "epoch": 0.38176197836166925,
+      "grad_norm": 1.7243876781798975,
+      "learning_rate": 6.098765432098766e-05,
+      "loss": 0.4765,
+      "step": 247
+    },
+    {
+      "epoch": 0.38330757341576505,
+      "grad_norm": 1.167432430205111,
+      "learning_rate": 6.123456790123457e-05,
+      "loss": 0.4726,
+      "step": 248
+    },
+    {
+      "epoch": 0.3848531684698609,
+      "grad_norm": 1.4858200526822005,
+      "learning_rate": 6.148148148148148e-05,
+      "loss": 0.4828,
+      "step": 249
+    },
+    {
+      "epoch": 0.38639876352395675,
+      "grad_norm": 0.9500856749304806,
+      "learning_rate": 6.17283950617284e-05,
+      "loss": 0.48,
+      "step": 250
+    },
+    {
+      "epoch": 0.38794435857805254,
+      "grad_norm": 0.8681770559849791,
+      "learning_rate": 6.197530864197531e-05,
+      "loss": 0.4703,
+      "step": 251
+    },
+    {
+      "epoch": 0.3894899536321484,
+      "grad_norm": 1.0989972142230926,
+      "learning_rate": 6.222222222222223e-05,
+      "loss": 0.4695,
+      "step": 252
+    },
+    {
+      "epoch": 0.3910355486862442,
+      "grad_norm": 0.9402230268151549,
+      "learning_rate": 6.246913580246914e-05,
+      "loss": 0.4758,
+      "step": 253
+    },
+    {
+      "epoch": 0.39258114374034003,
+      "grad_norm": 0.9848484342795987,
+      "learning_rate": 6.271604938271606e-05,
+      "loss": 0.4668,
+      "step": 254
+    },
+    {
+      "epoch": 0.3941267387944359,
+      "grad_norm": 1.1574464067866512,
+      "learning_rate": 6.296296296296298e-05,
+      "loss": 0.4657,
+      "step": 255
+    },
+    {
+      "epoch": 0.3956723338485317,
+      "grad_norm": 1.1322864127449526,
+      "learning_rate": 6.320987654320988e-05,
+      "loss": 0.4766,
+      "step": 256
+    },
+    {
+      "epoch": 0.3972179289026275,
+      "grad_norm": 1.196737268139593,
+      "learning_rate": 6.34567901234568e-05,
+      "loss": 0.468,
+      "step": 257
+    },
+    {
+      "epoch": 0.3987635239567233,
+      "grad_norm": 0.8615114262443972,
+      "learning_rate": 6.37037037037037e-05,
+      "loss": 0.4786,
+      "step": 258
+    },
+    {
+      "epoch": 0.40030911901081917,
+      "grad_norm": 1.0155583184427355,
+      "learning_rate": 6.395061728395062e-05,
+      "loss": 0.476,
+      "step": 259
+    },
+    {
+      "epoch": 0.401854714064915,
+      "grad_norm": 0.8705021038685753,
+      "learning_rate": 6.419753086419754e-05,
+      "loss": 0.4814,
+      "step": 260
+    },
+    {
+      "epoch": 0.4034003091190108,
+      "grad_norm": 1.0551108697985605,
+      "learning_rate": 6.444444444444446e-05,
+      "loss": 0.4698,
+      "step": 261
+    },
+    {
+      "epoch": 0.40494590417310666,
+      "grad_norm": 1.2069387128869065,
+      "learning_rate": 6.469135802469136e-05,
+      "loss": 0.4743,
+      "step": 262
+    },
+    {
+      "epoch": 0.40649149922720246,
+      "grad_norm": 0.7605544444447253,
+      "learning_rate": 6.493827160493828e-05,
+      "loss": 0.4543,
+      "step": 263
+    },
+    {
+      "epoch": 0.4080370942812983,
+      "grad_norm": 0.8205540585121743,
+      "learning_rate": 6.518518518518519e-05,
+      "loss": 0.4684,
+      "step": 264
+    },
+    {
+      "epoch": 0.4095826893353941,
+      "grad_norm": 1.0619805628658,
+      "learning_rate": 6.54320987654321e-05,
+      "loss": 0.4668,
+      "step": 265
+    },
+    {
+      "epoch": 0.41112828438948995,
+      "grad_norm": 1.232094986638869,
+      "learning_rate": 6.567901234567901e-05,
+      "loss": 0.4657,
+      "step": 266
+    },
+    {
+      "epoch": 0.4126738794435858,
+      "grad_norm": 1.2059342270172937,
+      "learning_rate": 6.592592592592593e-05,
+      "loss": 0.4627,
+      "step": 267
+    },
+    {
+      "epoch": 0.4142194744976816,
+      "grad_norm": 0.9923650404733554,
+      "learning_rate": 6.617283950617285e-05,
+      "loss": 0.4823,
+      "step": 268
+    },
+    {
+      "epoch": 0.41576506955177744,
+      "grad_norm": 0.9006494596530477,
+      "learning_rate": 6.641975308641977e-05,
+      "loss": 0.468,
+      "step": 269
+    },
+    {
+      "epoch": 0.41731066460587324,
+      "grad_norm": 0.9325765412437474,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.4784,
+      "step": 270
+    },
+    {
+      "epoch": 0.4188562596599691,
+      "grad_norm": 1.047774171198826,
+      "learning_rate": 6.691358024691359e-05,
+      "loss": 0.4671,
+      "step": 271
+    },
+    {
+      "epoch": 0.42040185471406494,
+      "grad_norm": 1.2038403751218132,
+      "learning_rate": 6.716049382716049e-05,
+      "loss": 0.4831,
+      "step": 272
+    },
+    {
+      "epoch": 0.42194744976816073,
+      "grad_norm": 0.848323979703427,
+      "learning_rate": 6.740740740740741e-05,
+      "loss": 0.4768,
+      "step": 273
+    },
+    {
+      "epoch": 0.4234930448222566,
+      "grad_norm": 1.468772962629579,
+      "learning_rate": 6.765432098765433e-05,
+      "loss": 0.4699,
+      "step": 274
+    },
+    {
+      "epoch": 0.4250386398763524,
+      "grad_norm": 0.7234151499833007,
+      "learning_rate": 6.790123456790123e-05,
+      "loss": 0.4709,
+      "step": 275
+    },
+    {
+      "epoch": 0.4265842349304482,
+      "grad_norm": 0.7864446856107487,
+      "learning_rate": 6.814814814814815e-05,
+      "loss": 0.4707,
+      "step": 276
+    },
+    {
+      "epoch": 0.4281298299845441,
+      "grad_norm": 1.5551025678321875,
+      "learning_rate": 6.839506172839507e-05,
+      "loss": 0.4828,
+      "step": 277
+    },
+    {
+      "epoch": 0.42967542503863987,
+      "grad_norm": 0.759156333480939,
+      "learning_rate": 6.864197530864199e-05,
+      "loss": 0.4602,
+      "step": 278
+    },
+    {
+      "epoch": 0.4312210200927357,
+      "grad_norm": 0.9162967605158892,
+      "learning_rate": 6.88888888888889e-05,
+      "loss": 0.4676,
+      "step": 279
+    },
+    {
+      "epoch": 0.4327666151468315,
+      "grad_norm": 1.3177533448409955,
+      "learning_rate": 6.91358024691358e-05,
+      "loss": 0.4771,
+      "step": 280
+    },
+    {
+      "epoch": 0.43431221020092736,
+      "grad_norm": 1.0065569840708148,
+      "learning_rate": 6.938271604938272e-05,
+      "loss": 0.4654,
+      "step": 281
+    },
+    {
+      "epoch": 0.43585780525502316,
+      "grad_norm": 1.0099019290554392,
+      "learning_rate": 6.962962962962964e-05,
+      "loss": 0.4736,
+      "step": 282
+    },
+    {
+      "epoch": 0.437403400309119,
+      "grad_norm": 0.9542599858654569,
+      "learning_rate": 6.987654320987655e-05,
+      "loss": 0.4697,
+      "step": 283
+    },
+    {
+      "epoch": 0.43894899536321486,
+      "grad_norm": 1.0993378431582324,
+      "learning_rate": 7.012345679012346e-05,
+      "loss": 0.4646,
+      "step": 284
+    },
+    {
+      "epoch": 0.44049459041731065,
+      "grad_norm": 0.9091639988441267,
+      "learning_rate": 7.037037037037038e-05,
+      "loss": 0.4731,
+      "step": 285
+    },
+    {
+      "epoch": 0.4420401854714065,
+      "grad_norm": 1.1115775076593608,
+      "learning_rate": 7.06172839506173e-05,
+      "loss": 0.4735,
+      "step": 286
+    },
+    {
+      "epoch": 0.4435857805255023,
+      "grad_norm": 1.1534031048862625,
+      "learning_rate": 7.08641975308642e-05,
+      "loss": 0.4748,
+      "step": 287
+    },
+    {
+      "epoch": 0.44513137557959814,
+      "grad_norm": 1.1500223572788075,
+      "learning_rate": 7.11111111111111e-05,
+      "loss": 0.4778,
+      "step": 288
+    },
+    {
+      "epoch": 0.446676970633694,
+      "grad_norm": 1.0303778729219273,
+      "learning_rate": 7.135802469135802e-05,
+      "loss": 0.4688,
+      "step": 289
+    },
+    {
+      "epoch": 0.4482225656877898,
+      "grad_norm": 0.9689431632471549,
+      "learning_rate": 7.160493827160494e-05,
+      "loss": 0.4762,
+      "step": 290
+    },
+    {
+      "epoch": 0.44976816074188564,
+      "grad_norm": 0.9233987172654912,
+      "learning_rate": 7.185185185185186e-05,
+      "loss": 0.4696,
+      "step": 291
+    },
+    {
+      "epoch": 0.45131375579598143,
+      "grad_norm": 1.4934762550254848,
+      "learning_rate": 7.209876543209877e-05,
+      "loss": 0.4752,
+      "step": 292
+    },
+    {
+      "epoch": 0.4528593508500773,
+      "grad_norm": 0.7502253552129426,
+      "learning_rate": 7.234567901234568e-05,
+      "loss": 0.4657,
+      "step": 293
+    },
+    {
+      "epoch": 0.45440494590417313,
+      "grad_norm": 1.117072200683503,
+      "learning_rate": 7.25925925925926e-05,
+      "loss": 0.4663,
+      "step": 294
+    },
+    {
+      "epoch": 0.4559505409582689,
+      "grad_norm": 0.9331429681836702,
+      "learning_rate": 7.283950617283952e-05,
+      "loss": 0.4766,
+      "step": 295
+    },
+    {
+      "epoch": 0.4574961360123648,
+      "grad_norm": 1.068844821636654,
+      "learning_rate": 7.308641975308643e-05,
+      "loss": 0.4699,
+      "step": 296
+    },
+    {
+      "epoch": 0.45904173106646057,
+      "grad_norm": 1.1226087523721424,
+      "learning_rate": 7.333333333333333e-05,
+      "loss": 0.4606,
+      "step": 297
+    },
+    {
+      "epoch": 0.4605873261205564,
+      "grad_norm": 1.0261605921557848,
+      "learning_rate": 7.358024691358025e-05,
+      "loss": 0.4664,
+      "step": 298
+    },
+    {
+      "epoch": 0.46213292117465227,
+      "grad_norm": 1.040973288606845,
+      "learning_rate": 7.382716049382717e-05,
+      "loss": 0.4696,
+      "step": 299
+    },
+    {
+      "epoch": 0.46367851622874806,
+      "grad_norm": 0.9783172608381693,
+      "learning_rate": 7.407407407407409e-05,
+      "loss": 0.4599,
+      "step": 300
+    },
+    {
+      "epoch": 0.4652241112828439,
+      "grad_norm": 1.2289274446920366,
+      "learning_rate": 7.432098765432099e-05,
+      "loss": 0.4676,
+      "step": 301
+    },
+    {
+      "epoch": 0.4667697063369397,
+      "grad_norm": 0.9551069659281435,
+      "learning_rate": 7.456790123456791e-05,
+      "loss": 0.4621,
+      "step": 302
+    },
+    {
+      "epoch": 0.46831530139103555,
+      "grad_norm": 0.9271166884767883,
+      "learning_rate": 7.481481481481483e-05,
+      "loss": 0.4612,
+      "step": 303
+    },
+    {
+      "epoch": 0.46986089644513135,
+      "grad_norm": 1.004864766717887,
+      "learning_rate": 7.506172839506173e-05,
+      "loss": 0.4683,
+      "step": 304
+    },
+    {
+      "epoch": 0.4714064914992272,
+      "grad_norm": 1.128627756522529,
+      "learning_rate": 7.530864197530865e-05,
+      "loss": 0.4709,
+      "step": 305
+    },
+    {
+      "epoch": 0.47295208655332305,
+      "grad_norm": 0.9321358642047468,
+      "learning_rate": 7.555555555555556e-05,
+      "loss": 0.4626,
+      "step": 306
+    },
+    {
+      "epoch": 0.47449768160741884,
+      "grad_norm": 1.3278999945984602,
+      "learning_rate": 7.580246913580247e-05,
+      "loss": 0.4748,
+      "step": 307
+    },
+    {
+      "epoch": 0.4760432766615147,
+      "grad_norm": 0.944025022165049,
+      "learning_rate": 7.60493827160494e-05,
+      "loss": 0.4606,
+      "step": 308
+    },
+    {
+      "epoch": 0.4775888717156105,
+      "grad_norm": 1.0799574982338942,
+      "learning_rate": 7.62962962962963e-05,
+      "loss": 0.4624,
+      "step": 309
+    },
+    {
+      "epoch": 0.47913446676970634,
+      "grad_norm": 0.8620477915536247,
+      "learning_rate": 7.654320987654322e-05,
+      "loss": 0.4704,
+      "step": 310
+    },
+    {
+      "epoch": 0.4806800618238022,
+      "grad_norm": 0.6727562335388572,
+      "learning_rate": 7.679012345679013e-05,
+      "loss": 0.4539,
+      "step": 311
+    },
+    {
+      "epoch": 0.482225656877898,
+      "grad_norm": 0.7125580919951122,
+      "learning_rate": 7.703703703703704e-05,
+      "loss": 0.465,
+      "step": 312
+    },
+    {
+      "epoch": 0.48377125193199383,
+      "grad_norm": 0.8850579621624475,
+      "learning_rate": 7.728395061728396e-05,
+      "loss": 0.4651,
+      "step": 313
+    },
+    {
+      "epoch": 0.4853168469860896,
+      "grad_norm": 1.089814285070669,
+      "learning_rate": 7.753086419753086e-05,
+      "loss": 0.4719,
+      "step": 314
+    },
+    {
+      "epoch": 0.4868624420401855,
+      "grad_norm": 1.0347330093917404,
+      "learning_rate": 7.777777777777778e-05,
+      "loss": 0.469,
+      "step": 315
+    },
+    {
+      "epoch": 0.4884080370942813,
+      "grad_norm": 1.033824468869202,
+      "learning_rate": 7.80246913580247e-05,
+      "loss": 0.4739,
+      "step": 316
+    },
+    {
+      "epoch": 0.4899536321483771,
+      "grad_norm": 1.0214619727971168,
+      "learning_rate": 7.827160493827162e-05,
+      "loss": 0.4636,
+      "step": 317
+    },
+    {
+      "epoch": 0.49149922720247297,
+      "grad_norm": 0.9265658300824663,
+      "learning_rate": 7.851851851851852e-05,
+      "loss": 0.4576,
+      "step": 318
+    },
+    {
+      "epoch": 0.49304482225656876,
+      "grad_norm": 0.9757407254253527,
+      "learning_rate": 7.876543209876544e-05,
+      "loss": 0.4696,
+      "step": 319
+    },
+    {
+      "epoch": 0.4945904173106646,
+      "grad_norm": 1.102852921361581,
+      "learning_rate": 7.901234567901235e-05,
+      "loss": 0.4649,
+      "step": 320
+    },
+    {
+      "epoch": 0.49613601236476046,
+      "grad_norm": 0.9164031059399039,
+      "learning_rate": 7.925925925925926e-05,
+      "loss": 0.468,
+      "step": 321
+    },
+    {
+      "epoch": 0.49768160741885625,
+      "grad_norm": 1.001279432380049,
+      "learning_rate": 7.950617283950618e-05,
+      "loss": 0.4604,
+      "step": 322
+    },
+    {
+      "epoch": 0.4992272024729521,
+      "grad_norm": 1.3988013409604274,
+      "learning_rate": 7.975308641975309e-05,
+      "loss": 0.4725,
+      "step": 323
+    },
+    {
+      "epoch": 0.500772797527048,
+      "grad_norm": 0.9225532294408699,
+      "learning_rate": 8e-05,
+      "loss": 0.4672,
+      "step": 324
+    },
+    {
+      "epoch": 0.5023183925811437,
+      "grad_norm": 1.2366631094121028,
+      "learning_rate": 7.999997670593474e-05,
+      "loss": 0.4743,
+      "step": 325
+    },
+    {
+      "epoch": 0.5038639876352395,
+      "grad_norm": 0.9985754515306058,
+      "learning_rate": 7.999990682376607e-05,
+      "loss": 0.4742,
+      "step": 326
+    },
+    {
+      "epoch": 0.5054095826893354,
+      "grad_norm": 1.77167590709832,
+      "learning_rate": 7.999979035357536e-05,
+      "loss": 0.4716,
+      "step": 327
+    },
+    {
+      "epoch": 0.5069551777434312,
+      "grad_norm": 0.9268029187821334,
+      "learning_rate": 7.99996272954983e-05,
+      "loss": 0.4537,
+      "step": 328
+    },
+    {
+      "epoch": 0.508500772797527,
+      "grad_norm": 2.2199583912571446,
+      "learning_rate": 7.99994176497248e-05,
+      "loss": 0.4732,
+      "step": 329
+    },
+    {
+      "epoch": 0.5100463678516228,
+      "grad_norm": 1.6042414500936655,
+      "learning_rate": 7.999916141649901e-05,
+      "loss": 0.4829,
+      "step": 330
+    },
+    {
+      "epoch": 0.5115919629057187,
+      "grad_norm": 1.4455600037251086,
+      "learning_rate": 7.99988585961194e-05,
+      "loss": 0.4691,
+      "step": 331
+    },
+    {
+      "epoch": 0.5131375579598145,
+      "grad_norm": 1.022348436352116,
+      "learning_rate": 7.999850918893863e-05,
+      "loss": 0.4641,
+      "step": 332
+    },
+    {
+      "epoch": 0.5146831530139103,
+      "grad_norm": 1.1934129252938614,
+      "learning_rate": 7.999811319536369e-05,
+      "loss": 0.4701,
+      "step": 333
+    },
+    {
+      "epoch": 0.5162287480680062,
+      "grad_norm": 0.7743917973611034,
+      "learning_rate": 7.999767061585576e-05,
+      "loss": 0.461,
+      "step": 334
+    },
+    {
+      "epoch": 0.517774343122102,
+      "grad_norm": 0.9949229362004298,
+      "learning_rate": 7.999718145093036e-05,
+      "loss": 0.4625,
+      "step": 335
+    },
+    {
+      "epoch": 0.5193199381761978,
+      "grad_norm": 0.5977321063031327,
+      "learning_rate": 7.999664570115717e-05,
+      "loss": 0.4738,
+      "step": 336
+    },
+    {
+      "epoch": 0.5208655332302936,
+      "grad_norm": 0.8157527643511809,
+      "learning_rate": 7.99960633671602e-05,
+      "loss": 0.4657,
+      "step": 337
+    },
+    {
+      "epoch": 0.5224111282843895,
+      "grad_norm": 0.7880571467871172,
+      "learning_rate": 7.999543444961772e-05,
+      "loss": 0.4755,
+      "step": 338
+    },
+    {
+      "epoch": 0.5239567233384853,
+      "grad_norm": 0.639800468540418,
+      "learning_rate": 7.999475894926221e-05,
+      "loss": 0.4589,
+      "step": 339
+    },
+    {
+      "epoch": 0.5255023183925811,
+      "grad_norm": 0.7688262011165063,
+      "learning_rate": 7.999403686688042e-05,
+      "loss": 0.4667,
+      "step": 340
+    },
+    {
+      "epoch": 0.527047913446677,
+      "grad_norm": 0.9245531439939358,
+      "learning_rate": 7.999326820331338e-05,
+      "loss": 0.4642,
+      "step": 341
+    },
+    {
+      "epoch": 0.5285935085007728,
+      "grad_norm": 1.322503342624274,
+      "learning_rate": 7.999245295945634e-05,
+      "loss": 0.4551,
+      "step": 342
+    },
+    {
+      "epoch": 0.5301391035548686,
+      "grad_norm": 0.6402251237059537,
+      "learning_rate": 7.999159113625881e-05,
+      "loss": 0.4567,
+      "step": 343
+    },
+    {
+      "epoch": 0.5316846986089645,
+      "grad_norm": 0.7517897833225893,
+      "learning_rate": 7.999068273472459e-05,
+      "loss": 0.4661,
+      "step": 344
+    },
+    {
+      "epoch": 0.5332302936630603,
+      "grad_norm": 1.0070553668165176,
+      "learning_rate": 7.998972775591167e-05,
+      "loss": 0.4711,
+      "step": 345
+    },
+    {
+      "epoch": 0.5347758887171561,
+      "grad_norm": 1.2274192513747975,
+      "learning_rate": 7.998872620093235e-05,
+      "loss": 0.466,
+      "step": 346
+    },
+    {
+      "epoch": 0.5363214837712519,
+      "grad_norm": 0.7117374365803301,
+      "learning_rate": 7.99876780709531e-05,
+      "loss": 0.4681,
+      "step": 347
+    },
+    {
+      "epoch": 0.5378670788253478,
+      "grad_norm": 0.8188752537996355,
+      "learning_rate": 7.99865833671947e-05,
+      "loss": 0.465,
+      "step": 348
+    },
+    {
+      "epoch": 0.5394126738794436,
+      "grad_norm": 0.875214173758601,
+      "learning_rate": 7.998544209093217e-05,
+      "loss": 0.4734,
+      "step": 349
+    },
+    {
+      "epoch": 0.5409582689335394,
+      "grad_norm": 0.659624722972033,
+      "learning_rate": 7.998425424349473e-05,
+      "loss": 0.4637,
+      "step": 350
+    },
+    {
+      "epoch": 0.5425038639876353,
+      "grad_norm": 0.7057348490226953,
+      "learning_rate": 7.99830198262659e-05,
+      "loss": 0.4623,
+      "step": 351
+    },
+    {
+      "epoch": 0.5440494590417311,
+      "grad_norm": 0.7050244930924285,
+      "learning_rate": 7.998173884068338e-05,
+      "loss": 0.454,
+      "step": 352
+    },
+    {
+      "epoch": 0.5455950540958269,
+      "grad_norm": 0.6020745367214642,
+      "learning_rate": 7.998041128823916e-05,
+      "loss": 0.4559,
+      "step": 353
+    },
+    {
+      "epoch": 0.5471406491499228,
+      "grad_norm": 0.6218265771052139,
+      "learning_rate": 7.997903717047944e-05,
+      "loss": 0.468,
+      "step": 354
+    },
+    {
+      "epoch": 0.5486862442040186,
+      "grad_norm": 0.7498605169579337,
+      "learning_rate": 7.997761648900466e-05,
+      "loss": 0.4568,
+      "step": 355
+    },
+    {
+      "epoch": 0.5502318392581144,
+      "grad_norm": 0.7386153906816254,
+      "learning_rate": 7.997614924546948e-05,
+      "loss": 0.4622,
+      "step": 356
+    },
+    {
+      "epoch": 0.5517774343122102,
+      "grad_norm": 0.8615583934080182,
+      "learning_rate": 7.997463544158282e-05,
+      "loss": 0.4632,
+      "step": 357
+    },
+    {
+      "epoch": 0.5533230293663061,
+      "grad_norm": 1.2489575879713575,
+      "learning_rate": 7.99730750791078e-05,
+      "loss": 0.4627,
+      "step": 358
+    },
+    {
+      "epoch": 0.5548686244204019,
+      "grad_norm": 0.8146213378195578,
+      "learning_rate": 7.997146815986179e-05,
+      "loss": 0.4568,
+      "step": 359
+    },
+    {
+      "epoch": 0.5564142194744977,
+      "grad_norm": 0.7922732907357454,
+      "learning_rate": 7.996981468571636e-05,
+      "loss": 0.4565,
+      "step": 360
+    },
+    {
+      "epoch": 0.5579598145285936,
+      "grad_norm": 0.6941937896489192,
+      "learning_rate": 7.996811465859733e-05,
+      "loss": 0.4517,
+      "step": 361
+    },
+    {
+      "epoch": 0.5595054095826894,
+      "grad_norm": 0.7332247912397324,
+      "learning_rate": 7.996636808048472e-05,
+      "loss": 0.463,
+      "step": 362
+    },
+    {
+      "epoch": 0.5610510046367851,
+      "grad_norm": 0.7508020875142575,
+      "learning_rate": 7.996457495341277e-05,
+      "loss": 0.4602,
+      "step": 363
+    },
+    {
+      "epoch": 0.5625965996908809,
+      "grad_norm": 0.756110893237292,
+      "learning_rate": 7.996273527946995e-05,
+      "loss": 0.4528,
+      "step": 364
+    },
+    {
+      "epoch": 0.5641421947449768,
+      "grad_norm": 0.7127020025169759,
+      "learning_rate": 7.996084906079895e-05,
+      "loss": 0.4545,
+      "step": 365
+    },
+    {
+      "epoch": 0.5656877897990726,
+      "grad_norm": 0.6905141182528117,
+      "learning_rate": 7.995891629959662e-05,
+      "loss": 0.4604,
+      "step": 366
+    },
+    {
+      "epoch": 0.5672333848531684,
+      "grad_norm": 0.8143923681191669,
+      "learning_rate": 7.995693699811409e-05,
+      "loss": 0.452,
+      "step": 367
+    },
+    {
+      "epoch": 0.5687789799072643,
+      "grad_norm": 1.0747779771959909,
+      "learning_rate": 7.995491115865662e-05,
+      "loss": 0.4471,
+      "step": 368
+    },
+    {
+      "epoch": 0.5703245749613601,
+      "grad_norm": 1.1193263847381563,
+      "learning_rate": 7.995283878358374e-05,
+      "loss": 0.462,
+      "step": 369
+    },
+    {
+      "epoch": 0.5718701700154559,
+      "grad_norm": 0.8604609875694763,
+      "learning_rate": 7.995071987530914e-05,
+      "loss": 0.4547,
+      "step": 370
+    },
+    {
+      "epoch": 0.5734157650695518,
+      "grad_norm": 0.8277466123842292,
+      "learning_rate": 7.994855443630072e-05,
+      "loss": 0.4564,
+      "step": 371
+    },
+    {
+      "epoch": 0.5749613601236476,
+      "grad_norm": 0.9834357821168109,
+      "learning_rate": 7.994634246908058e-05,
+      "loss": 0.4588,
+      "step": 372
+    },
+    {
+      "epoch": 0.5765069551777434,
+      "grad_norm": 1.1250087341862804,
+      "learning_rate": 7.994408397622501e-05,
+      "loss": 0.4592,
+      "step": 373
+    },
+    {
+      "epoch": 0.5780525502318392,
+      "grad_norm": 0.8218693286115759,
+      "learning_rate": 7.994177896036446e-05,
+      "loss": 0.4585,
+      "step": 374
+    },
+    {
+      "epoch": 0.5795981452859351,
+      "grad_norm": 0.8271377679531405,
+      "learning_rate": 7.993942742418362e-05,
+      "loss": 0.4606,
+      "step": 375
+    },
+    {
+      "epoch": 0.5811437403400309,
+      "grad_norm": 1.0470920301230913,
+      "learning_rate": 7.993702937042131e-05,
+      "loss": 0.4532,
+      "step": 376
+    },
+    {
+      "epoch": 0.5826893353941267,
+      "grad_norm": 1.1461008670711745,
+      "learning_rate": 7.993458480187057e-05,
+      "loss": 0.4711,
+      "step": 377
+    },
+    {
+      "epoch": 0.5842349304482226,
+      "grad_norm": 0.8038032056131119,
+      "learning_rate": 7.993209372137858e-05,
+      "loss": 0.4605,
+      "step": 378
+    },
+    {
+      "epoch": 0.5857805255023184,
+      "grad_norm": 0.9733659124496695,
+      "learning_rate": 7.99295561318467e-05,
+      "loss": 0.4585,
+      "step": 379
+    },
+    {
+      "epoch": 0.5873261205564142,
+      "grad_norm": 1.1224303019272235,
+      "learning_rate": 7.99269720362305e-05,
+      "loss": 0.4624,
+      "step": 380
+    },
+    {
+      "epoch": 0.58887171561051,
+      "grad_norm": 0.6177695872757162,
+      "learning_rate": 7.992434143753968e-05,
+      "loss": 0.4459,
+      "step": 381
+    },
+    {
+      "epoch": 0.5904173106646059,
+      "grad_norm": 0.5624887565167997,
+      "learning_rate": 7.992166433883807e-05,
+      "loss": 0.4531,
+      "step": 382
+    },
+    {
+      "epoch": 0.5919629057187017,
+      "grad_norm": 0.6382536533975867,
+      "learning_rate": 7.991894074324374e-05,
+      "loss": 0.4565,
+      "step": 383
+    },
+    {
+      "epoch": 0.5935085007727975,
+      "grad_norm": 0.7222552352500594,
+      "learning_rate": 7.991617065392884e-05,
+      "loss": 0.4451,
+      "step": 384
+    },
+    {
+      "epoch": 0.5950540958268934,
+      "grad_norm": 0.8115172194271355,
+      "learning_rate": 7.99133540741197e-05,
+      "loss": 0.4521,
+      "step": 385
+    },
+    {
+      "epoch": 0.5965996908809892,
+      "grad_norm": 0.9417727734786303,
+      "learning_rate": 7.991049100709684e-05,
+      "loss": 0.4543,
+      "step": 386
+    },
+    {
+      "epoch": 0.598145285935085,
+      "grad_norm": 0.9514296085746278,
+      "learning_rate": 7.990758145619484e-05,
+      "loss": 0.4561,
+      "step": 387
+    },
+    {
+      "epoch": 0.5996908809891809,
+      "grad_norm": 0.9714733033456525,
+      "learning_rate": 7.99046254248025e-05,
+      "loss": 0.451,
+      "step": 388
+    },
+    {
+      "epoch": 0.6012364760432767,
+      "grad_norm": 0.9844237398266047,
+      "learning_rate": 7.990162291636269e-05,
+      "loss": 0.4504,
+      "step": 389
+    },
+    {
+      "epoch": 0.6027820710973725,
+      "grad_norm": 0.9369492734406104,
+      "learning_rate": 7.989857393437245e-05,
+      "loss": 0.4585,
+      "step": 390
+    },
+    {
+      "epoch": 0.6043276661514683,
+      "grad_norm": 0.7551208261075399,
+      "learning_rate": 7.989547848238294e-05,
+      "loss": 0.4452,
+      "step": 391
+    },
+    {
+      "epoch": 0.6058732612055642,
+      "grad_norm": 0.6232383668619544,
+      "learning_rate": 7.989233656399944e-05,
+      "loss": 0.4587,
+      "step": 392
+    },
+    {
+      "epoch": 0.60741885625966,
+      "grad_norm": 0.4901931334082307,
+      "learning_rate": 7.988914818288136e-05,
+      "loss": 0.4572,
+      "step": 393
+    },
+    {
+      "epoch": 0.6089644513137558,
+      "grad_norm": 0.47374939525246373,
+      "learning_rate": 7.988591334274221e-05,
+      "loss": 0.4446,
+      "step": 394
+    },
+    {
+      "epoch": 0.6105100463678517,
+      "grad_norm": 0.4596712513868155,
+      "learning_rate": 7.988263204734962e-05,
+      "loss": 0.4524,
+      "step": 395
+    },
+    {
+      "epoch": 0.6120556414219475,
+      "grad_norm": 0.4282353169289228,
+      "learning_rate": 7.987930430052533e-05,
+      "loss": 0.4577,
+      "step": 396
+    },
+    {
+      "epoch": 0.6136012364760433,
+      "grad_norm": 0.5099867320381566,
+      "learning_rate": 7.98759301061452e-05,
+      "loss": 0.4485,
+      "step": 397
+    },
+    {
+      "epoch": 0.615146831530139,
+      "grad_norm": 0.539515319591019,
+      "learning_rate": 7.987250946813911e-05,
+      "loss": 0.4493,
+      "step": 398
+    },
+    {
+      "epoch": 0.616692426584235,
+      "grad_norm": 0.6309528596296303,
+      "learning_rate": 7.986904239049113e-05,
+      "loss": 0.4644,
+      "step": 399
+    },
+    {
+      "epoch": 0.6182380216383307,
+      "grad_norm": 0.7946789370337108,
+      "learning_rate": 7.986552887723937e-05,
+      "loss": 0.4513,
+      "step": 400
+    },
+    {
+      "epoch": 0.6197836166924265,
+      "grad_norm": 0.9655674223669141,
+      "learning_rate": 7.986196893247603e-05,
+      "loss": 0.4518,
+      "step": 401
+    },
+    {
+      "epoch": 0.6213292117465224,
+      "grad_norm": 1.1337471741612535,
+      "learning_rate": 7.985836256034737e-05,
+      "loss": 0.4604,
+      "step": 402
+    },
+    {
+      "epoch": 0.6228748068006182,
+      "grad_norm": 0.8273182843501121,
+      "learning_rate": 7.985470976505378e-05,
+      "loss": 0.4567,
+      "step": 403
+    },
+    {
+      "epoch": 0.624420401854714,
+      "grad_norm": 0.6777953490489269,
+      "learning_rate": 7.985101055084963e-05,
+      "loss": 0.4463,
+      "step": 404
+    },
+    {
+      "epoch": 0.6259659969088099,
+      "grad_norm": 0.7153814353017776,
+      "learning_rate": 7.984726492204346e-05,
+      "loss": 0.4412,
+      "step": 405
+    },
+    {
+      "epoch": 0.6275115919629057,
+      "grad_norm": 0.7425537671761728,
+      "learning_rate": 7.98434728829978e-05,
+      "loss": 0.4569,
+      "step": 406
+    },
+    {
+      "epoch": 0.6290571870170015,
+      "grad_norm": 0.839336233607707,
+      "learning_rate": 7.983963443812924e-05,
+      "loss": 0.446,
+      "step": 407
+    },
+    {
+      "epoch": 0.6306027820710973,
+      "grad_norm": 0.9586866438462934,
+      "learning_rate": 7.983574959190843e-05,
+      "loss": 0.4535,
+      "step": 408
+    },
+    {
+      "epoch": 0.6321483771251932,
+      "grad_norm": 0.9765270261026754,
+      "learning_rate": 7.983181834886006e-05,
+      "loss": 0.4475,
+      "step": 409
+    },
+    {
+      "epoch": 0.633693972179289,
+      "grad_norm": 0.759582420091956,
+      "learning_rate": 7.982784071356289e-05,
+      "loss": 0.452,
+      "step": 410
+    },
+    {
+      "epoch": 0.6352395672333848,
+      "grad_norm": 0.6353310663111404,
+      "learning_rate": 7.982381669064965e-05,
+      "loss": 0.4421,
+      "step": 411
+    },
+    {
+      "epoch": 0.6367851622874807,
+      "grad_norm": 0.7361906306338643,
+      "learning_rate": 7.981974628480715e-05,
+      "loss": 0.4438,
+      "step": 412
+    },
+    {
+      "epoch": 0.6383307573415765,
+      "grad_norm": 0.6339865173716746,
+      "learning_rate": 7.981562950077619e-05,
+      "loss": 0.4485,
+      "step": 413
+    },
+    {
+      "epoch": 0.6398763523956723,
+      "grad_norm": 0.5435774755527445,
+      "learning_rate": 7.981146634335163e-05,
+      "loss": 0.4536,
+      "step": 414
+    },
+    {
+      "epoch": 0.6414219474497682,
+      "grad_norm": 0.5710868607662681,
+      "learning_rate": 7.980725681738229e-05,
+      "loss": 0.4548,
+      "step": 415
+    },
+    {
+      "epoch": 0.642967542503864,
+      "grad_norm": 0.5911308867835612,
+      "learning_rate": 7.980300092777102e-05,
+      "loss": 0.451,
+      "step": 416
+    },
+    {
+      "epoch": 0.6445131375579598,
+      "grad_norm": 0.7225191067183553,
+      "learning_rate": 7.979869867947468e-05,
+      "loss": 0.4417,
+      "step": 417
+    },
+    {
+      "epoch": 0.6460587326120556,
+      "grad_norm": 0.7521628176249268,
+      "learning_rate": 7.97943500775041e-05,
+      "loss": 0.4402,
+      "step": 418
+    },
+    {
+      "epoch": 0.6476043276661515,
+      "grad_norm": 0.678780273863019,
+      "learning_rate": 7.978995512692412e-05,
+      "loss": 0.456,
+      "step": 419
+    },
+    {
+      "epoch": 0.6491499227202473,
+      "grad_norm": 0.7182629873142787,
+      "learning_rate": 7.978551383285354e-05,
+      "loss": 0.4551,
+      "step": 420
+    },
+    {
+      "epoch": 0.6506955177743431,
+      "grad_norm": 0.8393864782623631,
+      "learning_rate": 7.978102620046517e-05,
+      "loss": 0.4504,
+      "step": 421
+    },
+    {
+      "epoch": 0.652241112828439,
+      "grad_norm": 1.133016706790667,
+      "learning_rate": 7.977649223498575e-05,
+      "loss": 0.4514,
+      "step": 422
+    },
+    {
+      "epoch": 0.6537867078825348,
+      "grad_norm": 1.03878671480764,
+      "learning_rate": 7.977191194169603e-05,
+      "loss": 0.4639,
+      "step": 423
+    },
+    {
+      "epoch": 0.6553323029366306,
+      "grad_norm": 0.8444433492995927,
+      "learning_rate": 7.976728532593067e-05,
+      "loss": 0.4491,
+      "step": 424
+    },
+    {
+      "epoch": 0.6568778979907264,
+      "grad_norm": 0.6960863030640867,
+      "learning_rate": 7.97626123930783e-05,
+      "loss": 0.4493,
+      "step": 425
+    },
+    {
+      "epoch": 0.6584234930448223,
+      "grad_norm": 0.7005072676048999,
+      "learning_rate": 7.975789314858153e-05,
+      "loss": 0.4476,
+      "step": 426
+    },
+    {
+      "epoch": 0.6599690880989181,
+      "grad_norm": 0.836746348938439,
+      "learning_rate": 7.975312759793685e-05,
+      "loss": 0.4512,
+      "step": 427
+    },
+    {
+      "epoch": 0.6615146831530139,
+      "grad_norm": 1.0689325787549664,
+      "learning_rate": 7.974831574669472e-05,
+      "loss": 0.4627,
+      "step": 428
+    },
+    {
+      "epoch": 0.6630602782071098,
+      "grad_norm": 1.1186581408986205,
+      "learning_rate": 7.974345760045953e-05,
+      "loss": 0.4488,
+      "step": 429
+    },
+    {
+      "epoch": 0.6646058732612056,
+      "grad_norm": 0.6777940465736243,
+      "learning_rate": 7.973855316488957e-05,
+      "loss": 0.4492,
+      "step": 430
+    },
+    {
+      "epoch": 0.6661514683153014,
+      "grad_norm": 0.5338973438235535,
+      "learning_rate": 7.973360244569705e-05,
+      "loss": 0.4431,
+      "step": 431
+    },
+    {
+      "epoch": 0.6676970633693973,
+      "grad_norm": 0.6210050484906554,
+      "learning_rate": 7.972860544864809e-05,
+      "loss": 0.4541,
+      "step": 432
+    },
+    {
+      "epoch": 0.6692426584234931,
+      "grad_norm": 0.7162208131594553,
+      "learning_rate": 7.972356217956272e-05,
+      "loss": 0.4578,
+      "step": 433
+    },
+    {
+      "epoch": 0.6707882534775889,
+      "grad_norm": 0.708321211019045,
+      "learning_rate": 7.971847264431482e-05,
+      "loss": 0.4498,
+      "step": 434
+    },
+    {
+      "epoch": 0.6723338485316847,
+      "grad_norm": 0.5917552072415284,
+      "learning_rate": 7.971333684883223e-05,
+      "loss": 0.4521,
+      "step": 435
+    },
+    {
+      "epoch": 0.6738794435857806,
+      "grad_norm": 0.5189396718295533,
+      "learning_rate": 7.970815479909661e-05,
+      "loss": 0.4433,
+      "step": 436
+    },
+    {
+      "epoch": 0.6754250386398764,
+      "grad_norm": 0.544043936139772,
+      "learning_rate": 7.970292650114351e-05,
+      "loss": 0.4478,
+      "step": 437
+    },
+    {
+      "epoch": 0.6769706336939721,
+      "grad_norm": 0.4283212741759581,
+      "learning_rate": 7.969765196106233e-05,
+      "loss": 0.4454,
+      "step": 438
+    },
+    {
+      "epoch": 0.678516228748068,
+      "grad_norm": 0.35731818057965875,
+      "learning_rate": 7.969233118499636e-05,
+      "loss": 0.4521,
+      "step": 439
+    },
+    {
+      "epoch": 0.6800618238021638,
+      "grad_norm": 0.49696678982887327,
+      "learning_rate": 7.968696417914274e-05,
+      "loss": 0.4401,
+      "step": 440
+    },
+    {
+      "epoch": 0.6816074188562596,
+      "grad_norm": 0.496159232355586,
+      "learning_rate": 7.968155094975243e-05,
+      "loss": 0.4458,
+      "step": 441
+    },
+    {
+      "epoch": 0.6831530139103554,
+      "grad_norm": 0.3930491418328142,
+      "learning_rate": 7.96760915031302e-05,
+      "loss": 0.4505,
+      "step": 442
+    },
+    {
+      "epoch": 0.6846986089644513,
+      "grad_norm": 0.41749812940055175,
+      "learning_rate": 7.967058584563471e-05,
+      "loss": 0.4625,
+      "step": 443
+    },
+    {
+      "epoch": 0.6862442040185471,
+      "grad_norm": 0.5552461791440614,
+      "learning_rate": 7.966503398367845e-05,
+      "loss": 0.4489,
+      "step": 444
+    },
+    {
+      "epoch": 0.6877897990726429,
+      "grad_norm": 0.539358431362722,
+      "learning_rate": 7.965943592372766e-05,
+      "loss": 0.4422,
+      "step": 445
+    },
+    {
+      "epoch": 0.6893353941267388,
+      "grad_norm": 0.5055158333988253,
+      "learning_rate": 7.965379167230241e-05,
+      "loss": 0.4557,
+      "step": 446
+    },
+    {
+      "epoch": 0.6908809891808346,
+      "grad_norm": 0.6860788834653874,
+      "learning_rate": 7.964810123597659e-05,
+      "loss": 0.4424,
+      "step": 447
+    },
+    {
+      "epoch": 0.6924265842349304,
+      "grad_norm": 1.011035092699708,
+      "learning_rate": 7.964236462137788e-05,
+      "loss": 0.4522,
+      "step": 448
+    },
+    {
+      "epoch": 0.6939721792890263,
+      "grad_norm": 1.3481812963210864,
+      "learning_rate": 7.963658183518772e-05,
+      "loss": 0.4637,
+      "step": 449
+    },
+    {
+      "epoch": 0.6955177743431221,
+      "grad_norm": 0.5038082193164464,
+      "learning_rate": 7.963075288414134e-05,
+      "loss": 0.4402,
+      "step": 450
+    },
+    {
+      "epoch": 0.6970633693972179,
+      "grad_norm": 0.8801962675300073,
+      "learning_rate": 7.962487777502773e-05,
+      "loss": 0.4537,
+      "step": 451
+    },
+    {
+      "epoch": 0.6986089644513137,
+      "grad_norm": 1.3239962632072204,
+      "learning_rate": 7.961895651468967e-05,
+      "loss": 0.4641,
+      "step": 452
+    },
+    {
+      "epoch": 0.7001545595054096,
+      "grad_norm": 0.36296382947469763,
+      "learning_rate": 7.961298911002367e-05,
+      "loss": 0.4539,
+      "step": 453
+    },
+    {
+      "epoch": 0.7017001545595054,
+      "grad_norm": 1.229972375753171,
+      "learning_rate": 7.960697556797997e-05,
+      "loss": 0.4538,
+      "step": 454
+    },
+    {
+      "epoch": 0.7032457496136012,
+      "grad_norm": 0.7635879409253314,
+      "learning_rate": 7.960091589556257e-05,
+      "loss": 0.4514,
+      "step": 455
+    },
+    {
+      "epoch": 0.7047913446676971,
+      "grad_norm": 0.6558156495845362,
+      "learning_rate": 7.959481009982919e-05,
+      "loss": 0.4442,
+      "step": 456
+    },
+    {
+      "epoch": 0.7063369397217929,
+      "grad_norm": 1.0384695968547808,
+      "learning_rate": 7.958865818789127e-05,
+      "loss": 0.4493,
+      "step": 457
+    },
+    {
+      "epoch": 0.7078825347758887,
+      "grad_norm": 1.0235891024753911,
+      "learning_rate": 7.958246016691396e-05,
+      "loss": 0.4405,
+      "step": 458
+    },
+    {
+      "epoch": 0.7094281298299846,
+      "grad_norm": 0.9251493068561275,
+      "learning_rate": 7.957621604411612e-05,
+      "loss": 0.4545,
+      "step": 459
+    },
+    {
+      "epoch": 0.7109737248840804,
+      "grad_norm": 0.8685368960150491,
+      "learning_rate": 7.95699258267703e-05,
+      "loss": 0.4431,
+      "step": 460
+    },
+    {
+      "epoch": 0.7125193199381762,
+      "grad_norm": 0.9488476837307872,
+      "learning_rate": 7.956358952220273e-05,
+      "loss": 0.4419,
+      "step": 461
+    },
+    {
+      "epoch": 0.714064914992272,
+      "grad_norm": 1.2420512500174485,
+      "learning_rate": 7.955720713779332e-05,
+      "loss": 0.4504,
+      "step": 462
+    },
+    {
+      "epoch": 0.7156105100463679,
+      "grad_norm": 0.6758251019013465,
+      "learning_rate": 7.955077868097568e-05,
+      "loss": 0.444,
+      "step": 463
+    },
+    {
+      "epoch": 0.7171561051004637,
+      "grad_norm": 0.6734311630977253,
+      "learning_rate": 7.954430415923703e-05,
+      "loss": 0.4499,
+      "step": 464
+    },
+    {
+      "epoch": 0.7187017001545595,
+      "grad_norm": 0.90449672539994,
+      "learning_rate": 7.953778358011827e-05,
+      "loss": 0.449,
+      "step": 465
+    },
+    {
+      "epoch": 0.7202472952086554,
+      "grad_norm": 1.3061506463069403,
+      "learning_rate": 7.953121695121395e-05,
+      "loss": 0.4499,
+      "step": 466
+    },
+    {
+      "epoch": 0.7217928902627512,
+      "grad_norm": 0.5869778083359343,
+      "learning_rate": 7.952460428017223e-05,
+      "loss": 0.4469,
+      "step": 467
+    },
+    {
+      "epoch": 0.723338485316847,
+      "grad_norm": 0.41164098744813804,
+      "learning_rate": 7.951794557469492e-05,
+      "loss": 0.4514,
+      "step": 468
+    },
+    {
+      "epoch": 0.7248840803709428,
+      "grad_norm": 0.7840446972608712,
+      "learning_rate": 7.951124084253745e-05,
+      "loss": 0.4445,
+      "step": 469
+    },
+    {
+      "epoch": 0.7264296754250387,
+      "grad_norm": 1.1636968066294135,
+      "learning_rate": 7.95044900915088e-05,
+      "loss": 0.4482,
+      "step": 470
+    },
+    {
+      "epoch": 0.7279752704791345,
+      "grad_norm": 0.7797641335885679,
+      "learning_rate": 7.949769332947162e-05,
+      "loss": 0.4338,
+      "step": 471
+    },
+    {
+      "epoch": 0.7295208655332303,
+      "grad_norm": 0.5014085938497415,
+      "learning_rate": 7.949085056434213e-05,
+      "loss": 0.4622,
+      "step": 472
+    },
+    {
+      "epoch": 0.7310664605873262,
+      "grad_norm": 0.7769090079710187,
+      "learning_rate": 7.948396180409012e-05,
+      "loss": 0.4544,
+      "step": 473
+    },
+    {
+      "epoch": 0.732612055641422,
+      "grad_norm": 1.1171659903459716,
+      "learning_rate": 7.947702705673893e-05,
+      "loss": 0.448,
+      "step": 474
+    },
+    {
+      "epoch": 0.7341576506955177,
+      "grad_norm": 0.7619870379947891,
+      "learning_rate": 7.947004633036548e-05,
+      "loss": 0.4439,
+      "step": 475
+    },
+    {
+      "epoch": 0.7357032457496137,
+      "grad_norm": 0.6685466568508059,
+      "learning_rate": 7.946301963310028e-05,
+      "loss": 0.4541,
+      "step": 476
+    },
+    {
+      "epoch": 0.7372488408037094,
+      "grad_norm": 0.643490097890329,
+      "learning_rate": 7.945594697312732e-05,
+      "loss": 0.4511,
+      "step": 477
+    },
+    {
+      "epoch": 0.7387944358578052,
+      "grad_norm": 0.5896476046955875,
+      "learning_rate": 7.944882835868417e-05,
+      "loss": 0.4476,
+      "step": 478
+    },
+    {
+      "epoch": 0.740340030911901,
+      "grad_norm": 0.7100375800765096,
+      "learning_rate": 7.944166379806186e-05,
+      "loss": 0.453,
+      "step": 479
+    },
+    {
+      "epoch": 0.7418856259659969,
+      "grad_norm": 0.8256463904622523,
+      "learning_rate": 7.943445329960503e-05,
+      "loss": 0.4529,
+      "step": 480
+    },
+    {
+      "epoch": 0.7434312210200927,
+      "grad_norm": 0.6962258117516658,
+      "learning_rate": 7.942719687171175e-05,
+      "loss": 0.4492,
+      "step": 481
+    },
+    {
+      "epoch": 0.7449768160741885,
+      "grad_norm": 0.6381012396696243,
+      "learning_rate": 7.941989452283359e-05,
+      "loss": 0.4483,
+      "step": 482
+    },
+    {
+      "epoch": 0.7465224111282844,
+      "grad_norm": 0.8129367906824256,
+      "learning_rate": 7.941254626147565e-05,
+      "loss": 0.4431,
+      "step": 483
+    },
+    {
+      "epoch": 0.7480680061823802,
+      "grad_norm": 0.7781368768779827,
+      "learning_rate": 7.940515209619644e-05,
+      "loss": 0.438,
+      "step": 484
+    },
+    {
+      "epoch": 0.749613601236476,
+      "grad_norm": 0.6249515855342817,
+      "learning_rate": 7.939771203560799e-05,
+      "loss": 0.4431,
+      "step": 485
+    },
+    {
+      "epoch": 0.7511591962905718,
+      "grad_norm": 0.5094482079908089,
+      "learning_rate": 7.939022608837576e-05,
+      "loss": 0.4405,
+      "step": 486
+    },
+    {
+      "epoch": 0.7527047913446677,
+      "grad_norm": 0.575246536145531,
+      "learning_rate": 7.938269426321867e-05,
+      "loss": 0.4394,
+      "step": 487
+    },
+    {
+      "epoch": 0.7542503863987635,
+      "grad_norm": 0.725728800009646,
+      "learning_rate": 7.937511656890903e-05,
+      "loss": 0.4352,
+      "step": 488
+    },
+    {
+      "epoch": 0.7557959814528593,
+      "grad_norm": 0.8217792413157236,
+      "learning_rate": 7.936749301427264e-05,
+      "loss": 0.4437,
+      "step": 489
+    },
+    {
+      "epoch": 0.7573415765069552,
+      "grad_norm": 0.9119629982005905,
+      "learning_rate": 7.935982360818863e-05,
+      "loss": 0.439,
+      "step": 490
+    },
+    {
+      "epoch": 0.758887171561051,
+      "grad_norm": 0.9928588690645079,
+      "learning_rate": 7.935210835958963e-05,
+      "loss": 0.4581,
+      "step": 491
+    },
+    {
+      "epoch": 0.7604327666151468,
+      "grad_norm": 0.9954906017773287,
+      "learning_rate": 7.934434727746161e-05,
+      "loss": 0.4481,
+      "step": 492
+    },
+    {
+      "epoch": 0.7619783616692427,
+      "grad_norm": 0.7651782309999712,
+      "learning_rate": 7.93365403708439e-05,
+      "loss": 0.4524,
+      "step": 493
+    },
+    {
+      "epoch": 0.7635239567233385,
+      "grad_norm": 0.4537098778667196,
+      "learning_rate": 7.932868764882926e-05,
+      "loss": 0.4452,
+      "step": 494
+    },
+    {
+      "epoch": 0.7650695517774343,
+      "grad_norm": 0.41264029448347583,
+      "learning_rate": 7.932078912056377e-05,
+      "loss": 0.446,
+      "step": 495
+    },
+    {
+      "epoch": 0.7666151468315301,
+      "grad_norm": 0.5624666992251612,
+      "learning_rate": 7.931284479524685e-05,
+      "loss": 0.4421,
+      "step": 496
+    },
+    {
+      "epoch": 0.768160741885626,
+      "grad_norm": 0.6132371242178343,
+      "learning_rate": 7.930485468213131e-05,
+      "loss": 0.4372,
+      "step": 497
+    },
+    {
+      "epoch": 0.7697063369397218,
+      "grad_norm": 0.5700958776156412,
+      "learning_rate": 7.929681879052325e-05,
+      "loss": 0.4441,
+      "step": 498
+    },
+    {
+      "epoch": 0.7712519319938176,
+      "grad_norm": 0.6025219052546477,
+      "learning_rate": 7.92887371297821e-05,
+      "loss": 0.4442,
+      "step": 499
+    },
+    {
+      "epoch": 0.7727975270479135,
+      "grad_norm": 0.7623541245497347,
+      "learning_rate": 7.92806097093206e-05,
+      "loss": 0.4405,
+      "step": 500
+    },
+    {
+      "epoch": 0.7743431221020093,
+      "grad_norm": 0.8822025006890616,
+      "learning_rate": 7.927243653860478e-05,
+      "loss": 0.4574,
+      "step": 501
+    },
+    {
+      "epoch": 0.7758887171561051,
+      "grad_norm": 0.82008990123641,
+      "learning_rate": 7.926421762715395e-05,
+      "loss": 0.4386,
+      "step": 502
+    },
+    {
+      "epoch": 0.7774343122102009,
+      "grad_norm": 0.6836316562948456,
+      "learning_rate": 7.925595298454073e-05,
+      "loss": 0.4471,
+      "step": 503
+    },
+    {
+      "epoch": 0.7789799072642968,
+      "grad_norm": 0.6645314827937467,
+      "learning_rate": 7.924764262039095e-05,
+      "loss": 0.4456,
+      "step": 504
+    },
+    {
+      "epoch": 0.7805255023183926,
+      "grad_norm": 0.6024413666963008,
+      "learning_rate": 7.923928654438371e-05,
+      "loss": 0.4477,
+      "step": 505
+    },
+    {
+      "epoch": 0.7820710973724884,
+      "grad_norm": 0.6067806497405188,
+      "learning_rate": 7.92308847662514e-05,
+      "loss": 0.4482,
+      "step": 506
+    },
+    {
+      "epoch": 0.7836166924265843,
+      "grad_norm": 0.8022751574120819,
+      "learning_rate": 7.922243729577957e-05,
+      "loss": 0.4496,
+      "step": 507
+    },
+    {
+      "epoch": 0.7851622874806801,
+      "grad_norm": 1.0187980123821614,
+      "learning_rate": 7.9213944142807e-05,
+      "loss": 0.4443,
+      "step": 508
+    },
+    {
+      "epoch": 0.7867078825347759,
+      "grad_norm": 0.9559480149445991,
+      "learning_rate": 7.920540531722572e-05,
+      "loss": 0.4455,
+      "step": 509
+    },
+    {
+      "epoch": 0.7882534775888718,
+      "grad_norm": 0.653004430829866,
+      "learning_rate": 7.919682082898092e-05,
+      "loss": 0.4399,
+      "step": 510
+    },
+    {
+      "epoch": 0.7897990726429676,
+      "grad_norm": 0.4991146707421027,
+      "learning_rate": 7.9188190688071e-05,
+      "loss": 0.4457,
+      "step": 511
+    },
+    {
+      "epoch": 0.7913446676970634,
+      "grad_norm": 0.7431054178473085,
+      "learning_rate": 7.917951490454748e-05,
+      "loss": 0.4496,
+      "step": 512
+    },
+    {
+      "epoch": 0.7928902627511591,
+      "grad_norm": 0.8725524077390574,
+      "learning_rate": 7.917079348851509e-05,
+      "loss": 0.4477,
+      "step": 513
+    },
+    {
+      "epoch": 0.794435857805255,
+      "grad_norm": 0.6178648850982248,
+      "learning_rate": 7.916202645013169e-05,
+      "loss": 0.4518,
+      "step": 514
+    },
+    {
+      "epoch": 0.7959814528593508,
+      "grad_norm": 0.4298646959576983,
+      "learning_rate": 7.915321379960829e-05,
+      "loss": 0.4499,
+      "step": 515
+    },
+    {
+      "epoch": 0.7975270479134466,
+      "grad_norm": 0.617294586390934,
+      "learning_rate": 7.9144355547209e-05,
+      "loss": 0.4519,
+      "step": 516
+    },
+    {
+      "epoch": 0.7990726429675425,
+      "grad_norm": 0.7476939585049011,
+      "learning_rate": 7.913545170325105e-05,
+      "loss": 0.4412,
+      "step": 517
+    },
+    {
+      "epoch": 0.8006182380216383,
+      "grad_norm": 0.7296163770474515,
+      "learning_rate": 7.912650227810478e-05,
+      "loss": 0.4353,
+      "step": 518
+    },
+    {
+      "epoch": 0.8021638330757341,
+      "grad_norm": 0.7674361690186301,
+      "learning_rate": 7.911750728219362e-05,
+      "loss": 0.4375,
+      "step": 519
+    },
+    {
+      "epoch": 0.80370942812983,
+      "grad_norm": 0.8319004424044151,
+      "learning_rate": 7.910846672599408e-05,
+      "loss": 0.4402,
+      "step": 520
+    },
+    {
+      "epoch": 0.8052550231839258,
+      "grad_norm": 0.7935095223866093,
+      "learning_rate": 7.909938062003572e-05,
+      "loss": 0.4432,
+      "step": 521
+    },
+    {
+      "epoch": 0.8068006182380216,
+      "grad_norm": 0.7275342789621305,
+      "learning_rate": 7.909024897490114e-05,
+      "loss": 0.4422,
+      "step": 522
+    },
+    {
+      "epoch": 0.8083462132921174,
+      "grad_norm": 0.6875408362152193,
+      "learning_rate": 7.9081071801226e-05,
+      "loss": 0.4446,
+      "step": 523
+    },
+    {
+      "epoch": 0.8098918083462133,
+      "grad_norm": 0.5489098475992172,
+      "learning_rate": 7.9071849109699e-05,
+      "loss": 0.4453,
+      "step": 524
+    },
+    {
+      "epoch": 0.8114374034003091,
+      "grad_norm": 0.4948536954876545,
+      "learning_rate": 7.906258091106184e-05,
+      "loss": 0.4439,
+      "step": 525
+    },
+    {
+      "epoch": 0.8129829984544049,
+      "grad_norm": 0.4895935666593852,
+      "learning_rate": 7.90532672161092e-05,
+      "loss": 0.4532,
+      "step": 526
+    },
+    {
+      "epoch": 0.8145285935085008,
+      "grad_norm": 0.4699864148512322,
+      "learning_rate": 7.904390803568879e-05,
+      "loss": 0.4379,
+      "step": 527
+    },
+    {
+      "epoch": 0.8160741885625966,
+      "grad_norm": 0.5369887995106534,
+      "learning_rate": 7.903450338070126e-05,
+      "loss": 0.4465,
+      "step": 528
+    },
+    {
+      "epoch": 0.8176197836166924,
+      "grad_norm": 0.6633108394754598,
+      "learning_rate": 7.902505326210028e-05,
+      "loss": 0.4384,
+      "step": 529
+    },
+    {
+      "epoch": 0.8191653786707882,
+      "grad_norm": 0.5780383800688729,
+      "learning_rate": 7.90155576908924e-05,
+      "loss": 0.4446,
+      "step": 530
+    },
+    {
+      "epoch": 0.8207109737248841,
+      "grad_norm": 0.5181731618383103,
+      "learning_rate": 7.900601667813714e-05,
+      "loss": 0.4508,
+      "step": 531
+    },
+    {
+      "epoch": 0.8222565687789799,
+      "grad_norm": 0.6003510684977706,
+      "learning_rate": 7.899643023494695e-05,
+      "loss": 0.4379,
+      "step": 532
+    },
+    {
+      "epoch": 0.8238021638330757,
+      "grad_norm": 0.6489577208509044,
+      "learning_rate": 7.898679837248723e-05,
+      "loss": 0.4401,
+      "step": 533
+    },
+    {
+      "epoch": 0.8253477588871716,
+      "grad_norm": 0.7106579431009961,
+      "learning_rate": 7.89771211019762e-05,
+      "loss": 0.4481,
+      "step": 534
+    },
+    {
+      "epoch": 0.8268933539412674,
+      "grad_norm": 0.8120044732059997,
+      "learning_rate": 7.896739843468502e-05,
+      "loss": 0.4339,
+      "step": 535
+    },
+    {
+      "epoch": 0.8284389489953632,
+      "grad_norm": 0.8242079009990684,
+      "learning_rate": 7.895763038193771e-05,
+      "loss": 0.4476,
+      "step": 536
+    },
+    {
+      "epoch": 0.8299845440494591,
+      "grad_norm": 0.9028256658800938,
+      "learning_rate": 7.894781695511117e-05,
+      "loss": 0.4407,
+      "step": 537
+    },
+    {
+      "epoch": 0.8315301391035549,
+      "grad_norm": 0.8808561340581151,
+      "learning_rate": 7.893795816563511e-05,
+      "loss": 0.4309,
+      "step": 538
+    },
+    {
+      "epoch": 0.8330757341576507,
+      "grad_norm": 0.8058197940502757,
+      "learning_rate": 7.89280540249921e-05,
+      "loss": 0.4406,
+      "step": 539
+    },
+    {
+      "epoch": 0.8346213292117465,
+      "grad_norm": 0.7700753495449788,
+      "learning_rate": 7.891810454471753e-05,
+      "loss": 0.4502,
+      "step": 540
+    },
+    {
+      "epoch": 0.8361669242658424,
+      "grad_norm": 0.7626009050964863,
+      "learning_rate": 7.890810973639958e-05,
+      "loss": 0.4425,
+      "step": 541
+    },
+    {
+      "epoch": 0.8377125193199382,
+      "grad_norm": 0.5970485832623469,
+      "learning_rate": 7.889806961167924e-05,
+      "loss": 0.4389,
+      "step": 542
+    },
+    {
+      "epoch": 0.839258114374034,
+      "grad_norm": 0.4132167503544843,
+      "learning_rate": 7.88879841822503e-05,
+      "loss": 0.4484,
+      "step": 543
+    },
+    {
+      "epoch": 0.8408037094281299,
+      "grad_norm": 0.561299960687465,
+      "learning_rate": 7.887785345985926e-05,
+      "loss": 0.4493,
+      "step": 544
+    },
+    {
+      "epoch": 0.8423493044822257,
+      "grad_norm": 0.8945919322396804,
+      "learning_rate": 7.886767745630542e-05,
+      "loss": 0.4432,
+      "step": 545
+    },
+    {
+      "epoch": 0.8438948995363215,
+      "grad_norm": 1.128259316099117,
+      "learning_rate": 7.88574561834408e-05,
+      "loss": 0.4399,
+      "step": 546
+    },
+    {
+      "epoch": 0.8454404945904173,
+      "grad_norm": 0.6004881396794288,
+      "learning_rate": 7.884718965317016e-05,
+      "loss": 0.4307,
+      "step": 547
+    },
+    {
+      "epoch": 0.8469860896445132,
+      "grad_norm": 0.501549489983025,
+      "learning_rate": 7.883687787745095e-05,
+      "loss": 0.436,
+      "step": 548
+    },
+    {
+      "epoch": 0.848531684698609,
+      "grad_norm": 0.8233296170013034,
+      "learning_rate": 7.882652086829333e-05,
+      "loss": 0.4281,
+      "step": 549
+    },
+    {
+      "epoch": 0.8500772797527048,
+      "grad_norm": 0.9394743239402724,
+      "learning_rate": 7.881611863776014e-05,
+      "loss": 0.4493,
+      "step": 550
+    },
+    {
+      "epoch": 0.8516228748068007,
+      "grad_norm": 0.9023855999242737,
+      "learning_rate": 7.880567119796692e-05,
+      "loss": 0.4328,
+      "step": 551
+    },
+    {
+      "epoch": 0.8531684698608965,
+      "grad_norm": 0.8111759904343439,
+      "learning_rate": 7.879517856108178e-05,
+      "loss": 0.441,
+      "step": 552
+    },
+    {
+      "epoch": 0.8547140649149922,
+      "grad_norm": 0.6977912406931622,
+      "learning_rate": 7.87846407393256e-05,
+      "loss": 0.4514,
+      "step": 553
+    },
+    {
+      "epoch": 0.8562596599690881,
+      "grad_norm": 0.5293324244486036,
+      "learning_rate": 7.877405774497176e-05,
+      "loss": 0.4334,
+      "step": 554
+    },
+    {
+      "epoch": 0.8578052550231839,
+      "grad_norm": 0.4023348521021668,
+      "learning_rate": 7.87634295903463e-05,
+      "loss": 0.4426,
+      "step": 555
+    },
+    {
+      "epoch": 0.8593508500772797,
+      "grad_norm": 0.48562563002908576,
+      "learning_rate": 7.875275628782794e-05,
+      "loss": 0.4395,
+      "step": 556
+    },
+    {
+      "epoch": 0.8608964451313755,
+      "grad_norm": 0.5600880982432517,
+      "learning_rate": 7.874203784984783e-05,
+      "loss": 0.451,
+      "step": 557
+    },
+    {
+      "epoch": 0.8624420401854714,
+      "grad_norm": 0.4867775949500963,
+      "learning_rate": 7.873127428888982e-05,
+      "loss": 0.4443,
+      "step": 558
+    },
+    {
+      "epoch": 0.8639876352395672,
+      "grad_norm": 0.45670488658685837,
+      "learning_rate": 7.872046561749023e-05,
+      "loss": 0.44,
+      "step": 559
+    },
+    {
+      "epoch": 0.865533230293663,
+      "grad_norm": 0.4761398613593943,
+      "learning_rate": 7.870961184823799e-05,
+      "loss": 0.4342,
+      "step": 560
+    },
+    {
+      "epoch": 0.8670788253477589,
+      "grad_norm": 0.46409040634999327,
+      "learning_rate": 7.869871299377449e-05,
+      "loss": 0.4304,
+      "step": 561
+    },
+    {
+      "epoch": 0.8686244204018547,
+      "grad_norm": 0.5109193076120637,
+      "learning_rate": 7.868776906679369e-05,
+      "loss": 0.4364,
+      "step": 562
+    },
+    {
+      "epoch": 0.8701700154559505,
+      "grad_norm": 0.6469465230363364,
+      "learning_rate": 7.867678008004199e-05,
+      "loss": 0.4416,
+      "step": 563
+    },
+    {
+      "epoch": 0.8717156105100463,
+      "grad_norm": 0.789980590549452,
+      "learning_rate": 7.866574604631832e-05,
+      "loss": 0.4425,
+      "step": 564
+    },
+    {
+      "epoch": 0.8732612055641422,
+      "grad_norm": 0.8933023925413848,
+      "learning_rate": 7.865466697847402e-05,
+      "loss": 0.4346,
+      "step": 565
+    },
+    {
+      "epoch": 0.874806800618238,
+      "grad_norm": 0.8968704012364771,
+      "learning_rate": 7.864354288941296e-05,
+      "loss": 0.4496,
+      "step": 566
+    },
+    {
+      "epoch": 0.8763523956723338,
+      "grad_norm": 0.8582673246978159,
+      "learning_rate": 7.863237379209137e-05,
+      "loss": 0.4329,
+      "step": 567
+    },
+    {
+      "epoch": 0.8778979907264297,
+      "grad_norm": 0.8087136075140331,
+      "learning_rate": 7.862115969951796e-05,
+      "loss": 0.4371,
+      "step": 568
+    },
+    {
+      "epoch": 0.8794435857805255,
+      "grad_norm": 0.7542353694260568,
+      "learning_rate": 7.860990062475382e-05,
+      "loss": 0.4365,
+      "step": 569
+    },
+    {
+      "epoch": 0.8809891808346213,
+      "grad_norm": 0.9014794292836983,
+      "learning_rate": 7.85985965809124e-05,
+      "loss": 0.4342,
+      "step": 570
+    },
+    {
+      "epoch": 0.8825347758887172,
+      "grad_norm": 1.070357819758905,
+      "learning_rate": 7.858724758115959e-05,
+      "loss": 0.4502,
+      "step": 571
+    },
+    {
+      "epoch": 0.884080370942813,
+      "grad_norm": 0.8931616566277254,
+      "learning_rate": 7.857585363871359e-05,
+      "loss": 0.4325,
+      "step": 572
+    },
+    {
+      "epoch": 0.8856259659969088,
+      "grad_norm": 0.7106036244064576,
+      "learning_rate": 7.856441476684495e-05,
+      "loss": 0.434,
+      "step": 573
+    },
+    {
+      "epoch": 0.8871715610510046,
+      "grad_norm": 0.6349342889735747,
+      "learning_rate": 7.85529309788766e-05,
+      "loss": 0.4458,
+      "step": 574
+    },
+    {
+      "epoch": 0.8887171561051005,
+      "grad_norm": 0.7282774160754655,
+      "learning_rate": 7.854140228818371e-05,
+      "loss": 0.4418,
+      "step": 575
+    },
+    {
+      "epoch": 0.8902627511591963,
+      "grad_norm": 0.715674599165362,
+      "learning_rate": 7.85298287081938e-05,
+      "loss": 0.442,
+      "step": 576
+    },
+    {
+      "epoch": 0.8918083462132921,
+      "grad_norm": 0.6093688140005056,
+      "learning_rate": 7.851821025238665e-05,
+      "loss": 0.4389,
+      "step": 577
+    },
+    {
+      "epoch": 0.893353941267388,
+      "grad_norm": 0.4640001330868731,
+      "learning_rate": 7.850654693429432e-05,
+      "loss": 0.4467,
+      "step": 578
+    },
+    {
+      "epoch": 0.8948995363214838,
+      "grad_norm": 0.4329185407466596,
+      "learning_rate": 7.84948387675011e-05,
+      "loss": 0.4422,
+      "step": 579
+    },
+    {
+      "epoch": 0.8964451313755796,
+      "grad_norm": 0.5522433341281833,
+      "learning_rate": 7.848308576564356e-05,
+      "loss": 0.4402,
+      "step": 580
+    },
+    {
+      "epoch": 0.8979907264296755,
+      "grad_norm": 0.5371418584269162,
+      "learning_rate": 7.847128794241041e-05,
+      "loss": 0.4408,
+      "step": 581
+    },
+    {
+      "epoch": 0.8995363214837713,
+      "grad_norm": 0.48480106420137625,
+      "learning_rate": 7.845944531154267e-05,
+      "loss": 0.4348,
+      "step": 582
+    },
+    {
+      "epoch": 0.9010819165378671,
+      "grad_norm": 0.5334433638893552,
+      "learning_rate": 7.844755788683345e-05,
+      "loss": 0.4403,
+      "step": 583
+    },
+    {
+      "epoch": 0.9026275115919629,
+      "grad_norm": 0.5014659034435851,
+      "learning_rate": 7.843562568212809e-05,
+      "loss": 0.4289,
+      "step": 584
+    },
+    {
+      "epoch": 0.9041731066460588,
+      "grad_norm": 0.5686974467379184,
+      "learning_rate": 7.842364871132407e-05,
+      "loss": 0.4374,
+      "step": 585
+    },
+    {
+      "epoch": 0.9057187017001546,
+      "grad_norm": 0.6733417899674484,
+      "learning_rate": 7.841162698837099e-05,
+      "loss": 0.4396,
+      "step": 586
+    },
+    {
+      "epoch": 0.9072642967542504,
+      "grad_norm": 0.46013610351286977,
+      "learning_rate": 7.839956052727061e-05,
+      "loss": 0.4464,
+      "step": 587
+    },
+    {
+      "epoch": 0.9088098918083463,
+      "grad_norm": 0.4602822817401326,
+      "learning_rate": 7.838744934207676e-05,
+      "loss": 0.4349,
+      "step": 588
+    },
+    {
+      "epoch": 0.910355486862442,
+      "grad_norm": 0.6095025922435364,
+      "learning_rate": 7.83752934468954e-05,
+      "loss": 0.4312,
+      "step": 589
+    },
+    {
+      "epoch": 0.9119010819165378,
+      "grad_norm": 0.5896066011664475,
+      "learning_rate": 7.836309285588452e-05,
+      "loss": 0.4355,
+      "step": 590
+    },
+    {
+      "epoch": 0.9134466769706336,
+      "grad_norm": 0.5421806312279646,
+      "learning_rate": 7.835084758325417e-05,
+      "loss": 0.43,
+      "step": 591
+    },
+    {
+      "epoch": 0.9149922720247295,
+      "grad_norm": 0.6621099500827932,
+      "learning_rate": 7.83385576432665e-05,
+      "loss": 0.4289,
+      "step": 592
+    },
+    {
+      "epoch": 0.9165378670788253,
+      "grad_norm": 0.7808387078845036,
+      "learning_rate": 7.832622305023564e-05,
+      "loss": 0.4405,
+      "step": 593
+    },
+    {
+      "epoch": 0.9180834621329211,
+      "grad_norm": 0.8519171092319919,
+      "learning_rate": 7.83138438185277e-05,
+      "loss": 0.4557,
+      "step": 594
+    },
+    {
+      "epoch": 0.919629057187017,
+      "grad_norm": 0.8521530486540375,
+      "learning_rate": 7.830141996256083e-05,
+      "loss": 0.4383,
+      "step": 595
+    },
+    {
+      "epoch": 0.9211746522411128,
+      "grad_norm": 0.736575807908396,
+      "learning_rate": 7.828895149680513e-05,
+      "loss": 0.4417,
+      "step": 596
+    },
+    {
+      "epoch": 0.9227202472952086,
+      "grad_norm": 0.5494295213432342,
+      "learning_rate": 7.827643843578268e-05,
+      "loss": 0.4302,
+      "step": 597
+    },
+    {
+      "epoch": 0.9242658423493045,
+      "grad_norm": 0.5331247439334492,
+      "learning_rate": 7.826388079406746e-05,
+      "loss": 0.4393,
+      "step": 598
+    },
+    {
+      "epoch": 0.9258114374034003,
+      "grad_norm": 0.6935679018267995,
+      "learning_rate": 7.825127858628541e-05,
+      "loss": 0.4483,
+      "step": 599
+    },
+    {
+      "epoch": 0.9273570324574961,
+      "grad_norm": 0.7595133695662849,
+      "learning_rate": 7.823863182711436e-05,
+      "loss": 0.4381,
+      "step": 600
+    },
+    {
+      "epoch": 0.9289026275115919,
+      "grad_norm": 0.6857450673605338,
+      "learning_rate": 7.822594053128403e-05,
+      "loss": 0.4389,
+      "step": 601
+    },
+    {
+      "epoch": 0.9304482225656878,
+      "grad_norm": 0.6824829999836448,
+      "learning_rate": 7.8213204713576e-05,
+      "loss": 0.4417,
+      "step": 602
+    },
+    {
+      "epoch": 0.9319938176197836,
+      "grad_norm": 0.6650583983572376,
+      "learning_rate": 7.820042438882374e-05,
+      "loss": 0.4371,
+      "step": 603
+    },
+    {
+      "epoch": 0.9335394126738794,
+      "grad_norm": 0.4745479272047121,
+      "learning_rate": 7.818759957191253e-05,
+      "loss": 0.435,
+      "step": 604
+    },
+    {
+      "epoch": 0.9350850077279753,
+      "grad_norm": 0.3498563813008576,
+      "learning_rate": 7.817473027777948e-05,
+      "loss": 0.4311,
+      "step": 605
+    },
+    {
+      "epoch": 0.9366306027820711,
+      "grad_norm": 0.445017616125712,
+      "learning_rate": 7.816181652141348e-05,
+      "loss": 0.4308,
+      "step": 606
+    },
+    {
+      "epoch": 0.9381761978361669,
+      "grad_norm": 0.4274818418189949,
+      "learning_rate": 7.814885831785525e-05,
+      "loss": 0.4303,
+      "step": 607
+    },
+    {
+      "epoch": 0.9397217928902627,
+      "grad_norm": 0.3867818825049829,
+      "learning_rate": 7.813585568219723e-05,
+      "loss": 0.4365,
+      "step": 608
+    },
+    {
+      "epoch": 0.9412673879443586,
+      "grad_norm": 0.39201184835391145,
+      "learning_rate": 7.812280862958364e-05,
+      "loss": 0.4356,
+      "step": 609
+    },
+    {
+      "epoch": 0.9428129829984544,
+      "grad_norm": 0.4730224744291155,
+      "learning_rate": 7.810971717521043e-05,
+      "loss": 0.43,
+      "step": 610
+    },
+    {
+      "epoch": 0.9443585780525502,
+      "grad_norm": 0.42281022746677577,
+      "learning_rate": 7.809658133432526e-05,
+      "loss": 0.4397,
+      "step": 611
+    },
+    {
+      "epoch": 0.9459041731066461,
+      "grad_norm": 0.4616372163135228,
+      "learning_rate": 7.808340112222748e-05,
+      "loss": 0.4404,
+      "step": 612
+    },
+    {
+      "epoch": 0.9474497681607419,
+      "grad_norm": 0.4906932188952762,
+      "learning_rate": 7.807017655426812e-05,
+      "loss": 0.4348,
+      "step": 613
+    },
+    {
+      "epoch": 0.9489953632148377,
+      "grad_norm": 0.5433982101934519,
+      "learning_rate": 7.80569076458499e-05,
+      "loss": 0.4351,
+      "step": 614
+    },
+    {
+      "epoch": 0.9505409582689336,
+      "grad_norm": 0.6181933088082244,
+      "learning_rate": 7.804359441242714e-05,
+      "loss": 0.4361,
+      "step": 615
+    },
+    {
+      "epoch": 0.9520865533230294,
+      "grad_norm": 0.7533568986921175,
+      "learning_rate": 7.803023686950582e-05,
+      "loss": 0.4398,
+      "step": 616
+    },
+    {
+      "epoch": 0.9536321483771252,
+      "grad_norm": 0.9369337286385903,
+      "learning_rate": 7.80168350326435e-05,
+      "loss": 0.4495,
+      "step": 617
+    },
+    {
+      "epoch": 0.955177743431221,
+      "grad_norm": 1.1515472785597607,
+      "learning_rate": 7.800338891744934e-05,
+      "loss": 0.4381,
+      "step": 618
+    },
+    {
+      "epoch": 0.9567233384853169,
+      "grad_norm": 0.8492454483579691,
+      "learning_rate": 7.798989853958412e-05,
+      "loss": 0.4355,
+      "step": 619
+    },
+    {
+      "epoch": 0.9582689335394127,
+      "grad_norm": 0.5008569867912441,
+      "learning_rate": 7.797636391476005e-05,
+      "loss": 0.4327,
+      "step": 620
+    },
+    {
+      "epoch": 0.9598145285935085,
+      "grad_norm": 0.42790561003786926,
+      "learning_rate": 7.796278505874102e-05,
+      "loss": 0.4377,
+      "step": 621
+    },
+    {
+      "epoch": 0.9613601236476044,
+      "grad_norm": 0.6704670742916456,
+      "learning_rate": 7.794916198734232e-05,
+      "loss": 0.4408,
+      "step": 622
+    },
+    {
+      "epoch": 0.9629057187017002,
+      "grad_norm": 0.7315174010174182,
+      "learning_rate": 7.793549471643083e-05,
+      "loss": 0.4344,
+      "step": 623
+    },
+    {
+      "epoch": 0.964451313755796,
+      "grad_norm": 0.5180116517867882,
+      "learning_rate": 7.792178326192483e-05,
+      "loss": 0.44,
+      "step": 624
+    },
+    {
+      "epoch": 0.9659969088098919,
+      "grad_norm": 0.3682930959457848,
+      "learning_rate": 7.790802763979411e-05,
+      "loss": 0.4422,
+      "step": 625
+    },
+    {
+      "epoch": 0.9675425038639877,
+      "grad_norm": 0.4827151927848899,
+      "learning_rate": 7.789422786605988e-05,
+      "loss": 0.4414,
+      "step": 626
+    },
+    {
+      "epoch": 0.9690880989180835,
+      "grad_norm": 0.525652375349752,
+      "learning_rate": 7.788038395679479e-05,
+      "loss": 0.4396,
+      "step": 627
+    },
+    {
+      "epoch": 0.9706336939721792,
+      "grad_norm": 0.47711550987418944,
+      "learning_rate": 7.78664959281229e-05,
+      "loss": 0.4417,
+      "step": 628
+    },
+    {
+      "epoch": 0.9721792890262752,
+      "grad_norm": 0.4521350816222783,
+      "learning_rate": 7.78525637962196e-05,
+      "loss": 0.4329,
+      "step": 629
+    },
+    {
+      "epoch": 0.973724884080371,
+      "grad_norm": 0.5311829664198351,
+      "learning_rate": 7.783858757731174e-05,
+      "loss": 0.4328,
+      "step": 630
+    },
+    {
+      "epoch": 0.9752704791344667,
+      "grad_norm": 0.5404316116308204,
+      "learning_rate": 7.782456728767744e-05,
+      "loss": 0.4275,
+      "step": 631
+    },
+    {
+      "epoch": 0.9768160741885626,
+      "grad_norm": 0.6410717302232665,
+      "learning_rate": 7.781050294364615e-05,
+      "loss": 0.435,
+      "step": 632
+    },
+    {
+      "epoch": 0.9783616692426584,
+      "grad_norm": 0.7640852772159162,
+      "learning_rate": 7.779639456159872e-05,
+      "loss": 0.4311,
+      "step": 633
+    },
+    {
+      "epoch": 0.9799072642967542,
+      "grad_norm": 0.8147701250526043,
+      "learning_rate": 7.778224215796719e-05,
+      "loss": 0.4348,
+      "step": 634
+    },
+    {
+      "epoch": 0.98145285935085,
+      "grad_norm": 0.8184848854239374,
+      "learning_rate": 7.77680457492349e-05,
+      "loss": 0.4402,
+      "step": 635
+    },
+    {
+      "epoch": 0.9829984544049459,
+      "grad_norm": 0.8728094725370318,
+      "learning_rate": 7.775380535193649e-05,
+      "loss": 0.4289,
+      "step": 636
+    },
+    {
+      "epoch": 0.9845440494590417,
+      "grad_norm": 0.8091290006835459,
+      "learning_rate": 7.773952098265778e-05,
+      "loss": 0.4368,
+      "step": 637
+    },
+    {
+      "epoch": 0.9860896445131375,
+      "grad_norm": 0.6942530585191017,
+      "learning_rate": 7.772519265803581e-05,
+      "loss": 0.4372,
+      "step": 638
+    },
+    {
+      "epoch": 0.9876352395672334,
+      "grad_norm": 0.5165228454411432,
+      "learning_rate": 7.771082039475884e-05,
+      "loss": 0.4381,
+      "step": 639
+    },
+    {
+      "epoch": 0.9891808346213292,
+      "grad_norm": 0.37777307745002114,
+      "learning_rate": 7.769640420956627e-05,
+      "loss": 0.4326,
+      "step": 640
+    },
+    {
+      "epoch": 0.990726429675425,
+      "grad_norm": 0.36547077198781047,
+      "learning_rate": 7.768194411924873e-05,
+      "loss": 0.4294,
+      "step": 641
+    },
+    {
+      "epoch": 0.9922720247295209,
+      "grad_norm": 0.4503781104554527,
+      "learning_rate": 7.766744014064786e-05,
+      "loss": 0.4378,
+      "step": 642
+    },
+    {
+      "epoch": 0.9938176197836167,
+      "grad_norm": 0.5481159449044525,
+      "learning_rate": 7.765289229065657e-05,
+      "loss": 0.4375,
+      "step": 643
+    },
+    {
+      "epoch": 0.9953632148377125,
+      "grad_norm": 0.6296919301305356,
+      "learning_rate": 7.763830058621873e-05,
+      "loss": 0.4462,
+      "step": 644
+    },
+    {
+      "epoch": 0.9969088098918083,
+      "grad_norm": 0.6488083143648484,
+      "learning_rate": 7.762366504432937e-05,
+      "loss": 0.4436,
+      "step": 645
+    },
+    {
+      "epoch": 0.9984544049459042,
+      "grad_norm": 0.6477864119314514,
+      "learning_rate": 7.760898568203453e-05,
+      "loss": 0.4355,
+      "step": 646
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7056159823492316,
+      "learning_rate": 7.759426251643135e-05,
+      "loss": 0.4445,
+      "step": 647
+    },
+    {
+      "epoch": 1.001545595054096,
+      "grad_norm": 0.729906652382838,
+      "learning_rate": 7.757949556466794e-05,
+      "loss": 0.4175,
+      "step": 648
+    },
+    {
+      "epoch": 1.0030911901081916,
+      "grad_norm": 0.744332000011287,
+      "learning_rate": 7.756468484394339e-05,
+      "loss": 0.4175,
+      "step": 649
+    },
+    {
+      "epoch": 1.0046367851622875,
+      "grad_norm": 0.9054700744643306,
+      "learning_rate": 7.75498303715078e-05,
+      "loss": 0.4243,
+      "step": 650
+    },
+    {
+      "epoch": 1.0061823802163834,
+      "grad_norm": 0.8567705197263996,
+      "learning_rate": 7.753493216466225e-05,
+      "loss": 0.4251,
+      "step": 651
+    },
+    {
+      "epoch": 1.007727975270479,
+      "grad_norm": 0.8080573567421983,
+      "learning_rate": 7.751999024075871e-05,
+      "loss": 0.4259,
+      "step": 652
+    },
+    {
+      "epoch": 1.009273570324575,
+      "grad_norm": 0.8497810423019736,
+      "learning_rate": 7.750500461720008e-05,
+      "loss": 0.4295,
+      "step": 653
+    },
+    {
+      "epoch": 1.010819165378671,
+      "grad_norm": 0.963231768182949,
+      "learning_rate": 7.748997531144019e-05,
+      "loss": 0.4202,
+      "step": 654
+    },
+    {
+      "epoch": 1.0123647604327666,
+      "grad_norm": 0.9131327521435064,
+      "learning_rate": 7.747490234098371e-05,
+      "loss": 0.4234,
+      "step": 655
+    },
+    {
+      "epoch": 1.0139103554868625,
+      "grad_norm": 3.233993309506214,
+      "learning_rate": 7.745978572338615e-05,
+      "loss": 0.4208,
+      "step": 656
+    },
+    {
+      "epoch": 1.0154559505409582,
+      "grad_norm": 1.1478534927734931,
+      "learning_rate": 7.744462547625393e-05,
+      "loss": 0.4313,
+      "step": 657
+    },
+    {
+      "epoch": 1.017001545595054,
+      "grad_norm": 1.2500819953566402,
+      "learning_rate": 7.742942161724422e-05,
+      "loss": 0.4317,
+      "step": 658
+    },
+    {
+      "epoch": 1.01854714064915,
+      "grad_norm": 0.7276767086353634,
+      "learning_rate": 7.7414174164065e-05,
+      "loss": 0.4177,
+      "step": 659
+    },
+    {
+      "epoch": 1.0200927357032457,
+      "grad_norm": 1.169391244438201,
+      "learning_rate": 7.739888313447502e-05,
+      "loss": 0.4258,
+      "step": 660
+    },
+    {
+      "epoch": 1.0216383307573416,
+      "grad_norm": 0.5246199930754715,
+      "learning_rate": 7.738354854628381e-05,
+      "loss": 0.4214,
+      "step": 661
+    },
+    {
+      "epoch": 1.0231839258114375,
+      "grad_norm": 0.8263151377221576,
+      "learning_rate": 7.73681704173516e-05,
+      "loss": 0.4205,
+      "step": 662
+    },
+    {
+      "epoch": 1.0247295208655331,
+      "grad_norm": 0.6551303044770097,
+      "learning_rate": 7.735274876558937e-05,
+      "loss": 0.4266,
+      "step": 663
+    },
+    {
+      "epoch": 1.026275115919629,
+      "grad_norm": 0.5149694864670407,
+      "learning_rate": 7.733728360895875e-05,
+      "loss": 0.4174,
+      "step": 664
+    },
+    {
+      "epoch": 1.027820710973725,
+      "grad_norm": 0.5586430541966018,
+      "learning_rate": 7.732177496547204e-05,
+      "loss": 0.4286,
+      "step": 665
+    },
+    {
+      "epoch": 1.0293663060278206,
+      "grad_norm": 0.5739556085683527,
+      "learning_rate": 7.730622285319225e-05,
+      "loss": 0.4207,
+      "step": 666
+    },
+    {
+      "epoch": 1.0309119010819165,
+      "grad_norm": 0.5090479419066287,
+      "learning_rate": 7.729062729023294e-05,
+      "loss": 0.4195,
+      "step": 667
+    },
+    {
+      "epoch": 1.0324574961360125,
+      "grad_norm": 0.44214443641614104,
+      "learning_rate": 7.727498829475835e-05,
+      "loss": 0.4226,
+      "step": 668
+    },
+    {
+      "epoch": 1.0340030911901081,
+      "grad_norm": 0.4301128111577211,
+      "learning_rate": 7.725930588498324e-05,
+      "loss": 0.424,
+      "step": 669
+    },
+    {
+      "epoch": 1.035548686244204,
+      "grad_norm": 0.4399222127023858,
+      "learning_rate": 7.724358007917296e-05,
+      "loss": 0.4185,
+      "step": 670
+    },
+    {
+      "epoch": 1.0370942812983,
+      "grad_norm": 0.48272414168544736,
+      "learning_rate": 7.722781089564342e-05,
+      "loss": 0.4266,
+      "step": 671
+    },
+    {
+      "epoch": 1.0386398763523956,
+      "grad_norm": 0.44917139741794737,
+      "learning_rate": 7.721199835276105e-05,
+      "loss": 0.4212,
+      "step": 672
+    },
+    {
+      "epoch": 1.0401854714064915,
+      "grad_norm": 0.5647705885835368,
+      "learning_rate": 7.719614246894274e-05,
+      "loss": 0.4228,
+      "step": 673
+    },
+    {
+      "epoch": 1.0417310664605872,
+      "grad_norm": 0.5300385509440503,
+      "learning_rate": 7.718024326265591e-05,
+      "loss": 0.424,
+      "step": 674
+    },
+    {
+      "epoch": 1.0432766615146831,
+      "grad_norm": 0.5033099549666273,
+      "learning_rate": 7.716430075241842e-05,
+      "loss": 0.4135,
+      "step": 675
+    },
+    {
+      "epoch": 1.044822256568779,
+      "grad_norm": 0.5281034352711879,
+      "learning_rate": 7.714831495679857e-05,
+      "loss": 0.4239,
+      "step": 676
+    },
+    {
+      "epoch": 1.0463678516228747,
+      "grad_norm": 0.4854866818040774,
+      "learning_rate": 7.713228589441505e-05,
+      "loss": 0.4269,
+      "step": 677
+    },
+    {
+      "epoch": 1.0479134466769706,
+      "grad_norm": 0.4753894723383696,
+      "learning_rate": 7.711621358393697e-05,
+      "loss": 0.4185,
+      "step": 678
+    },
+    {
+      "epoch": 1.0494590417310665,
+      "grad_norm": 0.4597262766834238,
+      "learning_rate": 7.71000980440838e-05,
+      "loss": 0.4218,
+      "step": 679
+    },
+    {
+      "epoch": 1.0510046367851622,
+      "grad_norm": 0.4835682943503437,
+      "learning_rate": 7.708393929362535e-05,
+      "loss": 0.423,
+      "step": 680
+    },
+    {
+      "epoch": 1.052550231839258,
+      "grad_norm": 0.42980910216068385,
+      "learning_rate": 7.70677373513818e-05,
+      "loss": 0.4077,
+      "step": 681
+    },
+    {
+      "epoch": 1.054095826893354,
+      "grad_norm": 0.4474633900583748,
+      "learning_rate": 7.705149223622357e-05,
+      "loss": 0.412,
+      "step": 682
+    },
+    {
+      "epoch": 1.0556414219474497,
+      "grad_norm": 2.4069816314810333,
+      "learning_rate": 7.703520396707142e-05,
+      "loss": 0.427,
+      "step": 683
+    },
+    {
+      "epoch": 1.0571870170015456,
+      "grad_norm": 93.04219750838124,
+      "learning_rate": 7.701887256289635e-05,
+      "loss": 0.7967,
+      "step": 684
+    },
+    {
+      "epoch": 1.0587326120556415,
+      "grad_norm": 2.1906687985628297,
+      "learning_rate": 7.70024980427196e-05,
+      "loss": 0.4442,
+      "step": 685
+    },
+    {
+      "epoch": 1.0602782071097372,
+      "grad_norm": 272.835975471966,
+      "learning_rate": 7.698608042561262e-05,
+      "loss": 5.4936,
+      "step": 686
+    },
+    {
+      "epoch": 1.061823802163833,
+      "grad_norm": 4.3404427976272535,
+      "learning_rate": 7.696961973069707e-05,
+      "loss": 0.579,
+      "step": 687
+    },
+    {
+      "epoch": 1.063369397217929,
+      "grad_norm": 10.760573360273582,
+      "learning_rate": 7.695311597714476e-05,
+      "loss": 0.7585,
+      "step": 688
+    },
+    {
+      "epoch": 1.0649149922720247,
+      "grad_norm": 188.1831077117581,
+      "learning_rate": 7.693656918417768e-05,
+      "loss": 1.8238,
+      "step": 689
+    },
+    {
+      "epoch": 1.0664605873261206,
+      "grad_norm": 197.49638552527358,
+      "learning_rate": 7.691997937106794e-05,
+      "loss": 2.2759,
+      "step": 690
+    },
+    {
+      "epoch": 1.0680061823802163,
+      "grad_norm": 206.56798798122128,
+      "learning_rate": 7.690334655713772e-05,
+      "loss": 2.0006,
+      "step": 691
+    },
+    {
+      "epoch": 1.0695517774343122,
+      "grad_norm": 100.6833875702295,
+      "learning_rate": 7.688667076175934e-05,
+      "loss": 8.2281,
+      "step": 692
+    },
+    {
+      "epoch": 1.071097372488408,
+      "grad_norm": 105.295675147931,
+      "learning_rate": 7.686995200435513e-05,
+      "loss": 4.6001,
+      "step": 693
+    },
+    {
+      "epoch": 1.0726429675425038,
+      "grad_norm": 14.645438032599293,
+      "learning_rate": 7.685319030439753e-05,
+      "loss": 1.1132,
+      "step": 694
+    },
+    {
+      "epoch": 1.0741885625965997,
+      "grad_norm": 23.371900123313832,
+      "learning_rate": 7.683638568140888e-05,
+      "loss": 0.884,
+      "step": 695
+    },
+    {
+      "epoch": 1.0757341576506956,
+      "grad_norm": 5.890050408513227,
+      "learning_rate": 7.681953815496163e-05,
+      "loss": 0.6911,
+      "step": 696
+    },
+    {
+      "epoch": 1.0772797527047913,
+      "grad_norm": 3.959864486181502,
+      "learning_rate": 7.680264774467813e-05,
+      "loss": 0.6442,
+      "step": 697
+    },
+    {
+      "epoch": 1.0788253477588872,
+      "grad_norm": 13.66560312210619,
+      "learning_rate": 7.678571447023069e-05,
+      "loss": 0.7786,
+      "step": 698
+    },
+    {
+      "epoch": 1.080370942812983,
+      "grad_norm": 3.3446009588671073,
+      "learning_rate": 7.676873835134157e-05,
+      "loss": 0.6308,
+      "step": 699
+    },
+    {
+      "epoch": 1.0819165378670788,
+      "grad_norm": 3.1766326038920907,
+      "learning_rate": 7.675171940778288e-05,
+      "loss": 0.6062,
+      "step": 700
+    },
+    {
+      "epoch": 1.0834621329211747,
+      "grad_norm": 2.8537048610899776,
+      "learning_rate": 7.673465765937667e-05,
+      "loss": 0.5381,
+      "step": 701
+    },
+    {
+      "epoch": 1.0850077279752706,
+      "grad_norm": 15.15550131506517,
+      "learning_rate": 7.67175531259948e-05,
+      "loss": 0.7292,
+      "step": 702
+    },
+    {
+      "epoch": 1.0865533230293662,
+      "grad_norm": 97.51530829433813,
+      "learning_rate": 7.670040582755897e-05,
+      "loss": 9.4639,
+      "step": 703
+    },
+    {
+      "epoch": 1.0880989180834622,
+      "grad_norm": 54.95438868978781,
+      "learning_rate": 7.668321578404071e-05,
+      "loss": 9.5178,
+      "step": 704
+    },
+    {
+      "epoch": 1.089644513137558,
+      "grad_norm": 420.2876754740128,
+      "learning_rate": 7.666598301546132e-05,
+      "loss": 28.2843,
+      "step": 705
+    },
+    {
+      "epoch": 1.0911901081916537,
+      "grad_norm": 52.35848535158038,
+      "learning_rate": 7.664870754189183e-05,
+      "loss": 9.2741,
+      "step": 706
+    },
+    {
+      "epoch": 1.0927357032457496,
+      "grad_norm": 97.93519552001106,
+      "learning_rate": 7.663138938345309e-05,
+      "loss": 13.0573,
+      "step": 707
+    },
+    {
+      "epoch": 1.0942812982998453,
+      "grad_norm": 39.499276377622536,
+      "learning_rate": 7.661402856031557e-05,
+      "loss": 16.4409,
+      "step": 708
+    },
+    {
+      "epoch": 1.0958268933539412,
+      "grad_norm": 70.20801371263504,
+      "learning_rate": 7.659662509269948e-05,
+      "loss": 9.4351,
+      "step": 709
+    },
+    {
+      "epoch": 1.0973724884080371,
+      "grad_norm": 259.3118119719558,
+      "learning_rate": 7.657917900087475e-05,
+      "loss": 22.6132,
+      "step": 710
+    },
+    {
+      "epoch": 1.0989180834621328,
+      "grad_norm": 63.40624302863192,
+      "learning_rate": 7.656169030516085e-05,
+      "loss": 10.2828,
+      "step": 711
+    },
+    {
+      "epoch": 1.1004636785162287,
+      "grad_norm": 24.90925284691798,
+      "learning_rate": 7.654415902592693e-05,
+      "loss": 8.0415,
+      "step": 712
+    },
+    {
+      "epoch": 1.1020092735703246,
+      "grad_norm": 151.42831740527757,
+      "learning_rate": 7.652658518359172e-05,
+      "loss": 8.9223,
+      "step": 713
+    },
+    {
+      "epoch": 1.1035548686244203,
+      "grad_norm": 347.70782039381953,
+      "learning_rate": 7.650896879862355e-05,
+      "loss": 8.2149,
+      "step": 714
+    },
+    {
+      "epoch": 1.1051004636785162,
+      "grad_norm": 48.74238085739299,
+      "learning_rate": 7.649130989154028e-05,
+      "loss": 8.6492,
+      "step": 715
+    },
+    {
+      "epoch": 1.1066460587326121,
+      "grad_norm": 28.495354706861466,
+      "learning_rate": 7.64736084829093e-05,
+      "loss": 7.6341,
+      "step": 716
+    },
+    {
+      "epoch": 1.1081916537867078,
+      "grad_norm": 9.757410329201203,
+      "learning_rate": 7.645586459334746e-05,
+      "loss": 6.7279,
+      "step": 717
+    },
+    {
+      "epoch": 1.1097372488408037,
+      "grad_norm": 13.581780015633562,
+      "learning_rate": 7.643807824352119e-05,
+      "loss": 6.7868,
+      "step": 718
+    },
+    {
+      "epoch": 1.1112828438948996,
+      "grad_norm": 33.35818290868364,
+      "learning_rate": 7.642024945414625e-05,
+      "loss": 7.0932,
+      "step": 719
+    },
+    {
+      "epoch": 1.1128284389489953,
+      "grad_norm": 35.833130776074306,
+      "learning_rate": 7.640237824598792e-05,
+      "loss": 6.8967,
+      "step": 720
+    },
+    {
+      "epoch": 1.1143740340030912,
+      "grad_norm": 17.593556359959898,
+      "learning_rate": 7.638446463986085e-05,
+      "loss": 6.7292,
+      "step": 721
+    },
+    {
+      "epoch": 1.1159196290571871,
+      "grad_norm": 11.628269896069664,
+      "learning_rate": 7.636650865662907e-05,
+      "loss": 6.5477,
+      "step": 722
+    },
+    {
+      "epoch": 1.1174652241112828,
+      "grad_norm": 25.27507341382732,
+      "learning_rate": 7.634851031720598e-05,
+      "loss": 6.5934,
+      "step": 723
+    },
+    {
+      "epoch": 1.1190108191653787,
+      "grad_norm": 22.411429742342815,
+      "learning_rate": 7.633046964255431e-05,
+      "loss": 6.614,
+      "step": 724
+    },
+    {
+      "epoch": 1.1205564142194744,
+      "grad_norm": 9.461470565160505,
+      "learning_rate": 7.631238665368606e-05,
+      "loss": 6.3066,
+      "step": 725
+    },
+    {
+      "epoch": 1.1221020092735703,
+      "grad_norm": 9.156472187617643,
+      "learning_rate": 7.629426137166259e-05,
+      "loss": 6.2147,
+      "step": 726
+    },
+    {
+      "epoch": 1.1236476043276662,
+      "grad_norm": 8.476840375076899,
+      "learning_rate": 7.627609381759445e-05,
+      "loss": 6.1244,
+      "step": 727
+    },
+    {
+      "epoch": 1.125193199381762,
+      "grad_norm": 4.93811170279106,
+      "learning_rate": 7.625788401264146e-05,
+      "loss": 5.9622,
+      "step": 728
+    },
+    {
+      "epoch": 1.1267387944358578,
+      "grad_norm": 8.735095953309674,
+      "learning_rate": 7.623963197801263e-05,
+      "loss": 6.022,
+      "step": 729
+    },
+    {
+      "epoch": 1.1282843894899537,
+      "grad_norm": 6.554505113131336,
+      "learning_rate": 7.622133773496616e-05,
+      "loss": 5.9151,
+      "step": 730
+    },
+    {
+      "epoch": 1.1298299845440494,
+      "grad_norm": 6.807111122961733,
+      "learning_rate": 7.620300130480944e-05,
+      "loss": 5.8511,
+      "step": 731
+    },
+    {
+      "epoch": 1.1313755795981453,
+      "grad_norm": 4.556049123778991,
+      "learning_rate": 7.618462270889893e-05,
+      "loss": 5.7735,
+      "step": 732
+    },
+    {
+      "epoch": 1.1329211746522412,
+      "grad_norm": 8.287766607976652,
+      "learning_rate": 7.616620196864027e-05,
+      "loss": 5.7623,
+      "step": 733
+    },
+    {
+      "epoch": 1.1344667697063369,
+      "grad_norm": 7.100423364114964,
+      "learning_rate": 7.614773910548816e-05,
+      "loss": 5.6987,
+      "step": 734
+    },
+    {
+      "epoch": 1.1360123647604328,
+      "grad_norm": 4.627903031062972,
+      "learning_rate": 7.612923414094634e-05,
+      "loss": 5.6177,
+      "step": 735
+    },
+    {
+      "epoch": 1.1375579598145287,
+      "grad_norm": 4.881674498195489,
+      "learning_rate": 7.61106870965676e-05,
+      "loss": 5.5916,
+      "step": 736
+    },
+    {
+      "epoch": 1.1391035548686244,
+      "grad_norm": 3.6298034646753483,
+      "learning_rate": 7.609209799395377e-05,
+      "loss": 5.5201,
+      "step": 737
+    },
+    {
+      "epoch": 1.1406491499227203,
+      "grad_norm": 5.3843072583896845,
+      "learning_rate": 7.607346685475561e-05,
+      "loss": 5.4628,
+      "step": 738
+    },
+    {
+      "epoch": 1.1421947449768162,
+      "grad_norm": 3.6685012575901506,
+      "learning_rate": 7.605479370067288e-05,
+      "loss": 5.4009,
+      "step": 739
+    },
+    {
+      "epoch": 1.1437403400309119,
+      "grad_norm": 4.115992637127712,
+      "learning_rate": 7.603607855345427e-05,
+      "loss": 5.3897,
+      "step": 740
+    },
+    {
+      "epoch": 1.1452859350850078,
+      "grad_norm": 4.042453089971003,
+      "learning_rate": 7.601732143489735e-05,
+      "loss": 5.325,
+      "step": 741
+    },
+    {
+      "epoch": 1.1468315301391034,
+      "grad_norm": 3.16771761274949,
+      "learning_rate": 7.599852236684863e-05,
+      "loss": 5.2761,
+      "step": 742
+    },
+    {
+      "epoch": 1.1483771251931993,
+      "grad_norm": 5.038776248782134,
+      "learning_rate": 7.597968137120342e-05,
+      "loss": 5.2472,
+      "step": 743
+    },
+    {
+      "epoch": 1.1499227202472952,
+      "grad_norm": 3.0285414515015887,
+      "learning_rate": 7.59607984699059e-05,
+      "loss": 5.1769,
+      "step": 744
+    },
+    {
+      "epoch": 1.1514683153013912,
+      "grad_norm": 8.593385916790467,
+      "learning_rate": 7.594187368494904e-05,
+      "loss": 5.2736,
+      "step": 745
+    },
+    {
+      "epoch": 1.1530139103554868,
+      "grad_norm": 4.918520985824609,
+      "learning_rate": 7.592290703837462e-05,
+      "loss": 5.1426,
+      "step": 746
+    },
+    {
+      "epoch": 1.1545595054095827,
+      "grad_norm": 6.804430302655261,
+      "learning_rate": 7.590389855227312e-05,
+      "loss": 5.1824,
+      "step": 747
+    },
+    {
+      "epoch": 1.1561051004636784,
+      "grad_norm": 4.876833368982366,
+      "learning_rate": 7.58848482487838e-05,
+      "loss": 5.1545,
+      "step": 748
+    },
+    {
+      "epoch": 1.1576506955177743,
+      "grad_norm": 6.15456205410321,
+      "learning_rate": 7.586575615009464e-05,
+      "loss": 5.0909,
+      "step": 749
+    },
+    {
+      "epoch": 1.1591962905718702,
+      "grad_norm": 6.17166000852694,
+      "learning_rate": 7.584662227844223e-05,
+      "loss": 5.0788,
+      "step": 750
+    },
+    {
+      "epoch": 1.160741885625966,
+      "grad_norm": 3.8399589257173505,
+      "learning_rate": 7.582744665611187e-05,
+      "loss": 5.0191,
+      "step": 751
+    },
+    {
+      "epoch": 1.1622874806800618,
+      "grad_norm": 6.150141184578046,
+      "learning_rate": 7.580822930543747e-05,
+      "loss": 5.1103,
+      "step": 752
+    },
+    {
+      "epoch": 1.1638330757341577,
+      "grad_norm": 3.6360701998799088,
+      "learning_rate": 7.578897024880154e-05,
+      "loss": 4.9713,
+      "step": 753
+    },
+    {
+      "epoch": 1.1653786707882534,
+      "grad_norm": 4.362230382599794,
+      "learning_rate": 7.576966950863515e-05,
+      "loss": 4.9301,
+      "step": 754
+    },
+    {
+      "epoch": 1.1669242658423493,
+      "grad_norm": 4.160900154324657,
+      "learning_rate": 7.575032710741796e-05,
+      "loss": 4.935,
+      "step": 755
+    },
+    {
+      "epoch": 1.1684698608964452,
+      "grad_norm": 4.542412791649708,
+      "learning_rate": 7.573094306767812e-05,
+      "loss": 4.9442,
+      "step": 756
+    },
+    {
+      "epoch": 1.170015455950541,
+      "grad_norm": 4.245248645752507,
+      "learning_rate": 7.571151741199227e-05,
+      "loss": 4.9345,
+      "step": 757
+    },
+    {
+      "epoch": 1.1715610510046368,
+      "grad_norm": 2.9414058287299834,
+      "learning_rate": 7.569205016298556e-05,
+      "loss": 4.8409,
+      "step": 758
+    },
+    {
+      "epoch": 1.1731066460587325,
+      "grad_norm": 4.156394586232867,
+      "learning_rate": 7.567254134333155e-05,
+      "loss": 4.8069,
+      "step": 759
+    },
+    {
+      "epoch": 1.1746522411128284,
+      "grad_norm": 5.035582526248614,
+      "learning_rate": 7.565299097575221e-05,
+      "loss": 4.8857,
+      "step": 760
+    },
+    {
+      "epoch": 1.1761978361669243,
+      "grad_norm": 5.045628338913676,
+      "learning_rate": 7.563339908301792e-05,
+      "loss": 4.8286,
+      "step": 761
+    },
+    {
+      "epoch": 1.1777434312210202,
+      "grad_norm": 3.988007515418781,
+      "learning_rate": 7.561376568794745e-05,
+      "loss": 4.7804,
+      "step": 762
+    },
+    {
+      "epoch": 1.179289026275116,
+      "grad_norm": 4.186207353985047,
+      "learning_rate": 7.559409081340784e-05,
+      "loss": 4.7789,
+      "step": 763
+    },
+    {
+      "epoch": 1.1808346213292118,
+      "grad_norm": 3.9396849054608793,
+      "learning_rate": 7.557437448231451e-05,
+      "loss": 4.6864,
+      "step": 764
+    },
+    {
+      "epoch": 1.1823802163833075,
+      "grad_norm": 4.537203137435761,
+      "learning_rate": 7.555461671763112e-05,
+      "loss": 4.7164,
+      "step": 765
+    },
+    {
+      "epoch": 1.1839258114374034,
+      "grad_norm": 4.65880138729529,
+      "learning_rate": 7.55348175423696e-05,
+      "loss": 4.6988,
+      "step": 766
+    },
+    {
+      "epoch": 1.1854714064914993,
+      "grad_norm": 3.4087996084696996,
+      "learning_rate": 7.551497697959013e-05,
+      "loss": 4.6225,
+      "step": 767
+    },
+    {
+      "epoch": 1.187017001545595,
+      "grad_norm": 3.5488113498187057,
+      "learning_rate": 7.549509505240107e-05,
+      "loss": 4.6265,
+      "step": 768
+    },
+    {
+      "epoch": 1.1885625965996909,
+      "grad_norm": 5.495822131358979,
+      "learning_rate": 7.547517178395896e-05,
+      "loss": 4.6234,
+      "step": 769
+    },
+    {
+      "epoch": 1.1901081916537868,
+      "grad_norm": 4.86362818242385,
+      "learning_rate": 7.545520719746851e-05,
+      "loss": 4.6488,
+      "step": 770
+    },
+    {
+      "epoch": 1.1916537867078825,
+      "grad_norm": 2.600222845522659,
+      "learning_rate": 7.543520131618251e-05,
+      "loss": 4.5529,
+      "step": 771
+    },
+    {
+      "epoch": 1.1931993817619784,
+      "grad_norm": 5.555047812446914,
+      "learning_rate": 7.541515416340192e-05,
+      "loss": 4.615,
+      "step": 772
+    },
+    {
+      "epoch": 1.1947449768160743,
+      "grad_norm": 4.40056853472803,
+      "learning_rate": 7.539506576247568e-05,
+      "loss": 4.5056,
+      "step": 773
+    },
+    {
+      "epoch": 1.19629057187017,
+      "grad_norm": 4.982724730651727,
+      "learning_rate": 7.537493613680085e-05,
+      "loss": 4.4974,
+      "step": 774
+    },
+    {
+      "epoch": 1.1978361669242659,
+      "grad_norm": 3.321912005941974,
+      "learning_rate": 7.535476530982244e-05,
+      "loss": 4.5238,
+      "step": 775
+    },
+    {
+      "epoch": 1.1993817619783615,
+      "grad_norm": 4.643808857450901,
+      "learning_rate": 7.533455330503351e-05,
+      "loss": 4.5084,
+      "step": 776
+    },
+    {
+      "epoch": 1.2009273570324575,
+      "grad_norm": 3.0699451748744533,
+      "learning_rate": 7.531430014597503e-05,
+      "loss": 4.4445,
+      "step": 777
+    },
+    {
+      "epoch": 1.2024729520865534,
+      "grad_norm": 4.55358876465052,
+      "learning_rate": 7.529400585623592e-05,
+      "loss": 4.4033,
+      "step": 778
+    },
+    {
+      "epoch": 1.2040185471406493,
+      "grad_norm": 4.112369297404642,
+      "learning_rate": 7.527367045945301e-05,
+      "loss": 4.4221,
+      "step": 779
+    },
+    {
+      "epoch": 1.205564142194745,
+      "grad_norm": 4.311022526229789,
+      "learning_rate": 7.5253293979311e-05,
+      "loss": 4.3886,
+      "step": 780
+    },
+    {
+      "epoch": 1.2071097372488409,
+      "grad_norm": 3.456145921710313,
+      "learning_rate": 7.523287643954243e-05,
+      "loss": 4.3614,
+      "step": 781
+    },
+    {
+      "epoch": 1.2086553323029365,
+      "grad_norm": 3.576154707012333,
+      "learning_rate": 7.52124178639277e-05,
+      "loss": 4.377,
+      "step": 782
+    },
+    {
+      "epoch": 1.2102009273570324,
+      "grad_norm": 4.97025430932979,
+      "learning_rate": 7.519191827629497e-05,
+      "loss": 4.3643,
+      "step": 783
+    },
+    {
+      "epoch": 1.2117465224111283,
+      "grad_norm": 3.9979129091531562,
+      "learning_rate": 7.517137770052017e-05,
+      "loss": 4.3257,
+      "step": 784
+    },
+    {
+      "epoch": 1.213292117465224,
+      "grad_norm": 3.178227194464303,
+      "learning_rate": 7.515079616052699e-05,
+      "loss": 4.2892,
+      "step": 785
+    },
+    {
+      "epoch": 1.21483771251932,
+      "grad_norm": 4.604727245265333,
+      "learning_rate": 7.513017368028681e-05,
+      "loss": 4.2603,
+      "step": 786
+    },
+    {
+      "epoch": 1.2163833075734158,
+      "grad_norm": 3.6145987998432996,
+      "learning_rate": 7.510951028381869e-05,
+      "loss": 4.247,
+      "step": 787
+    },
+    {
+      "epoch": 1.2179289026275115,
+      "grad_norm": 3.5981722391474147,
+      "learning_rate": 7.508880599518936e-05,
+      "loss": 4.2436,
+      "step": 788
+    },
+    {
+      "epoch": 1.2194744976816074,
+      "grad_norm": 4.986630877971159,
+      "learning_rate": 7.506806083851319e-05,
+      "loss": 4.2043,
+      "step": 789
+    },
+    {
+      "epoch": 1.2210200927357033,
+      "grad_norm": 3.7121353858408104,
+      "learning_rate": 7.504727483795211e-05,
+      "loss": 4.1745,
+      "step": 790
+    },
+    {
+      "epoch": 1.222565687789799,
+      "grad_norm": 3.649859347149282,
+      "learning_rate": 7.502644801771564e-05,
+      "loss": 4.1762,
+      "step": 791
+    },
+    {
+      "epoch": 1.224111282843895,
+      "grad_norm": 4.924771754371227,
+      "learning_rate": 7.500558040206086e-05,
+      "loss": 4.1296,
+      "step": 792
+    },
+    {
+      "epoch": 1.2256568778979906,
+      "grad_norm": 3.846939395214013,
+      "learning_rate": 7.498467201529236e-05,
+      "loss": 4.0909,
+      "step": 793
+    },
+    {
+      "epoch": 1.2272024729520865,
+      "grad_norm": 4.1106153126958045,
+      "learning_rate": 7.496372288176219e-05,
+      "loss": 4.1137,
+      "step": 794
+    },
+    {
+      "epoch": 1.2287480680061824,
+      "grad_norm": 4.808135397086257,
+      "learning_rate": 7.494273302586986e-05,
+      "loss": 4.1335,
+      "step": 795
+    },
+    {
+      "epoch": 1.2302936630602783,
+      "grad_norm": 3.6829894131694014,
+      "learning_rate": 7.492170247206234e-05,
+      "loss": 4.0574,
+      "step": 796
+    },
+    {
+      "epoch": 1.231839258114374,
+      "grad_norm": 3.519370566178741,
+      "learning_rate": 7.490063124483401e-05,
+      "loss": 4.0691,
+      "step": 797
+    },
+    {
+      "epoch": 1.23338485316847,
+      "grad_norm": 3.646513485706842,
+      "learning_rate": 7.487951936872655e-05,
+      "loss": 4.0395,
+      "step": 798
+    },
+    {
+      "epoch": 1.2349304482225656,
+      "grad_norm": 4.283259741858885,
+      "learning_rate": 7.485836686832904e-05,
+      "loss": 4.0253,
+      "step": 799
+    },
+    {
+      "epoch": 1.2364760432766615,
+      "grad_norm": 2.691183180900685,
+      "learning_rate": 7.483717376827791e-05,
+      "loss": 3.9774,
+      "step": 800
+    },
+    {
+      "epoch": 1.2380216383307574,
+      "grad_norm": 3.506101475003609,
+      "learning_rate": 7.481594009325679e-05,
+      "loss": 4.0073,
+      "step": 801
+    },
+    {
+      "epoch": 1.239567233384853,
+      "grad_norm": 3.8175643338130296,
+      "learning_rate": 7.479466586799662e-05,
+      "loss": 3.9094,
+      "step": 802
+    },
+    {
+      "epoch": 1.241112828438949,
+      "grad_norm": 4.354755409166524,
+      "learning_rate": 7.477335111727556e-05,
+      "loss": 3.9642,
+      "step": 803
+    },
+    {
+      "epoch": 1.242658423493045,
+      "grad_norm": 3.2877563917762225,
+      "learning_rate": 7.475199586591897e-05,
+      "loss": 3.9092,
+      "step": 804
+    },
+    {
+      "epoch": 1.2442040185471406,
+      "grad_norm": 4.439813730748986,
+      "learning_rate": 7.473060013879938e-05,
+      "loss": 3.9291,
+      "step": 805
+    },
+    {
+      "epoch": 1.2457496136012365,
+      "grad_norm": 4.343013410523542,
+      "learning_rate": 7.470916396083649e-05,
+      "loss": 3.8698,
+      "step": 806
+    },
+    {
+      "epoch": 1.2472952086553324,
+      "grad_norm": 3.7371468862956023,
+      "learning_rate": 7.468768735699703e-05,
+      "loss": 3.8414,
+      "step": 807
+    },
+    {
+      "epoch": 1.248840803709428,
+      "grad_norm": 3.853589896317306,
+      "learning_rate": 7.466617035229493e-05,
+      "loss": 3.8265,
+      "step": 808
+    },
+    {
+      "epoch": 1.250386398763524,
+      "grad_norm": 4.032592828377498,
+      "learning_rate": 7.464461297179107e-05,
+      "loss": 3.7549,
+      "step": 809
+    },
+    {
+      "epoch": 1.2519319938176197,
+      "grad_norm": 3.9821131119978936,
+      "learning_rate": 7.462301524059342e-05,
+      "loss": 3.7342,
+      "step": 810
+    },
+    {
+      "epoch": 1.2534775888717156,
+      "grad_norm": 4.732422166817073,
+      "learning_rate": 7.460137718385694e-05,
+      "loss": 3.7275,
+      "step": 811
+    },
+    {
+      "epoch": 1.2550231839258115,
+      "grad_norm": 4.5088682750508475,
+      "learning_rate": 7.457969882678353e-05,
+      "loss": 3.6924,
+      "step": 812
+    },
+    {
+      "epoch": 1.2565687789799074,
+      "grad_norm": 5.449522575354933,
+      "learning_rate": 7.455798019462204e-05,
+      "loss": 3.7383,
+      "step": 813
+    },
+    {
+      "epoch": 1.258114374034003,
+      "grad_norm": 4.76650173402543,
+      "learning_rate": 7.453622131266824e-05,
+      "loss": 3.7432,
+      "step": 814
+    },
+    {
+      "epoch": 1.259659969088099,
+      "grad_norm": 4.067447342968902,
+      "learning_rate": 7.451442220626477e-05,
+      "loss": 3.6055,
+      "step": 815
+    },
+    {
+      "epoch": 1.2612055641421946,
+      "grad_norm": 6.253282776401617,
+      "learning_rate": 7.449258290080112e-05,
+      "loss": 3.6417,
+      "step": 816
+    },
+    {
+      "epoch": 1.2627511591962906,
+      "grad_norm": 3.2475778126347223,
+      "learning_rate": 7.44707034217136e-05,
+      "loss": 3.5223,
+      "step": 817
+    },
+    {
+      "epoch": 1.2642967542503865,
+      "grad_norm": 6.215818869272131,
+      "learning_rate": 7.44487837944853e-05,
+      "loss": 3.5363,
+      "step": 818
+    },
+    {
+      "epoch": 1.2658423493044824,
+      "grad_norm": 4.126921705951098,
+      "learning_rate": 7.44268240446461e-05,
+      "loss": 3.412,
+      "step": 819
+    },
+    {
+      "epoch": 1.267387944358578,
+      "grad_norm": 3.612290620794877,
+      "learning_rate": 7.440482419777259e-05,
+      "loss": 3.4251,
+      "step": 820
+    },
+    {
+      "epoch": 1.268933539412674,
+      "grad_norm": 6.2150029543395044,
+      "learning_rate": 7.438278427948805e-05,
+      "loss": 3.338,
+      "step": 821
+    },
+    {
+      "epoch": 1.2704791344667696,
+      "grad_norm": 3.483848021370735,
+      "learning_rate": 7.436070431546245e-05,
+      "loss": 3.2848,
+      "step": 822
+    },
+    {
+      "epoch": 1.2720247295208655,
+      "grad_norm": 5.015251343254233,
+      "learning_rate": 7.433858433141242e-05,
+      "loss": 3.2626,
+      "step": 823
+    },
+    {
+      "epoch": 1.2735703245749614,
+      "grad_norm": 4.373403459885209,
+      "learning_rate": 7.431642435310113e-05,
+      "loss": 3.1986,
+      "step": 824
+    },
+    {
+      "epoch": 1.2751159196290571,
+      "grad_norm": 6.518864003736122,
+      "learning_rate": 7.429422440633842e-05,
+      "loss": 3.1361,
+      "step": 825
+    },
+    {
+      "epoch": 1.276661514683153,
+      "grad_norm": 6.3909451423005414,
+      "learning_rate": 7.427198451698062e-05,
+      "loss": 3.0242,
+      "step": 826
+    },
+    {
+      "epoch": 1.2782071097372487,
+      "grad_norm": 6.9704608126309,
+      "learning_rate": 7.42497047109306e-05,
+      "loss": 2.9986,
+      "step": 827
+    },
+    {
+      "epoch": 1.2797527047913446,
+      "grad_norm": 4.655208475479434,
+      "learning_rate": 7.422738501413774e-05,
+      "loss": 2.8159,
+      "step": 828
+    },
+    {
+      "epoch": 1.2812982998454405,
+      "grad_norm": 15.894144967213778,
+      "learning_rate": 7.420502545259785e-05,
+      "loss": 3.3776,
+      "step": 829
+    },
+    {
+      "epoch": 1.2828438948995364,
+      "grad_norm": 7.253150436432715,
+      "learning_rate": 7.418262605235319e-05,
+      "loss": 2.9451,
+      "step": 830
+    },
+    {
+      "epoch": 1.2843894899536321,
+      "grad_norm": 6.263577298039799,
+      "learning_rate": 7.41601868394924e-05,
+      "loss": 2.6416,
+      "step": 831
+    },
+    {
+      "epoch": 1.285935085007728,
+      "grad_norm": 8.962995030592927,
+      "learning_rate": 7.413770784015053e-05,
+      "loss": 2.4536,
+      "step": 832
+    },
+    {
+      "epoch": 1.2874806800618237,
+      "grad_norm": 16.309713333237088,
+      "learning_rate": 7.411518908050893e-05,
+      "loss": 2.0685,
+      "step": 833
+    },
+    {
+      "epoch": 1.2890262751159196,
+      "grad_norm": 25.303288554745563,
+      "learning_rate": 7.409263058679527e-05,
+      "loss": 2.7383,
+      "step": 834
+    },
+    {
+      "epoch": 1.2905718701700155,
+      "grad_norm": 120.89512942503012,
+      "learning_rate": 7.407003238528352e-05,
+      "loss": 2.4238,
+      "step": 835
+    },
+    {
+      "epoch": 1.2921174652241114,
+      "grad_norm": 180.8543756230507,
+      "learning_rate": 7.404739450229385e-05,
+      "loss": 3.3396,
+      "step": 836
+    },
+    {
+      "epoch": 1.293663060278207,
+      "grad_norm": 108.89896521164664,
+      "learning_rate": 7.402471696419268e-05,
+      "loss": 3.2652,
+      "step": 837
+    },
+    {
+      "epoch": 1.295208655332303,
+      "grad_norm": 15.288203222772317,
+      "learning_rate": 7.400199979739263e-05,
+      "loss": 2.9395,
+      "step": 838
+    },
+    {
+      "epoch": 1.2967542503863987,
+      "grad_norm": 8.53625059242779,
+      "learning_rate": 7.397924302835247e-05,
+      "loss": 2.5028,
+      "step": 839
+    },
+    {
+      "epoch": 1.2982998454404946,
+      "grad_norm": 5.39838056054518,
+      "learning_rate": 7.395644668357704e-05,
+      "loss": 2.3153,
+      "step": 840
+    },
+    {
+      "epoch": 1.2998454404945905,
+      "grad_norm": 6.489588147483505,
+      "learning_rate": 7.393361078961735e-05,
+      "loss": 1.9318,
+      "step": 841
+    },
+    {
+      "epoch": 1.3013910355486862,
+      "grad_norm": 4.212253493269106,
+      "learning_rate": 7.391073537307044e-05,
+      "loss": 1.5532,
+      "step": 842
+    },
+    {
+      "epoch": 1.302936630602782,
+      "grad_norm": 3.0261374214192625,
+      "learning_rate": 7.388782046057936e-05,
+      "loss": 1.2898,
+      "step": 843
+    },
+    {
+      "epoch": 1.3044822256568778,
+      "grad_norm": 2.9650921679720463,
+      "learning_rate": 7.386486607883321e-05,
+      "loss": 1.0822,
+      "step": 844
+    },
+    {
+      "epoch": 1.3060278207109737,
+      "grad_norm": 27.710845490194725,
+      "learning_rate": 7.384187225456702e-05,
+      "loss": 0.9661,
+      "step": 845
+    },
+    {
+      "epoch": 1.3075734157650696,
+      "grad_norm": 88.50090182305416,
+      "learning_rate": 7.381883901456177e-05,
+      "loss": 4.9543,
+      "step": 846
+    },
+    {
+      "epoch": 1.3091190108191655,
+      "grad_norm": 191.46826332868898,
+      "learning_rate": 7.379576638564434e-05,
+      "loss": 2.5144,
+      "step": 847
+    },
+    {
+      "epoch": 1.3106646058732612,
+      "grad_norm": 77.57634195299751,
+      "learning_rate": 7.377265439468752e-05,
+      "loss": 2.7009,
+      "step": 848
+    },
+    {
+      "epoch": 1.312210200927357,
+      "grad_norm": 13.33061951011333,
+      "learning_rate": 7.374950306860991e-05,
+      "loss": 1.4283,
+      "step": 849
+    },
+    {
+      "epoch": 1.3137557959814528,
+      "grad_norm": 7.027804561086941,
+      "learning_rate": 7.372631243437593e-05,
+      "loss": 1.117,
+      "step": 850
+    },
+    {
+      "epoch": 1.3153013910355487,
+      "grad_norm": 8.667066051805296,
+      "learning_rate": 7.370308251899581e-05,
+      "loss": 0.9451,
+      "step": 851
+    },
+    {
+      "epoch": 1.3168469860896446,
+      "grad_norm": 6.737402708053703,
+      "learning_rate": 7.367981334952546e-05,
+      "loss": 0.9061,
+      "step": 852
+    },
+    {
+      "epoch": 1.3183925811437405,
+      "grad_norm": 3.17675567551532,
+      "learning_rate": 7.365650495306662e-05,
+      "loss": 0.7909,
+      "step": 853
+    },
+    {
+      "epoch": 1.3199381761978362,
+      "grad_norm": 3.4109002512475457,
+      "learning_rate": 7.36331573567666e-05,
+      "loss": 0.7458,
+      "step": 854
+    },
+    {
+      "epoch": 1.321483771251932,
+      "grad_norm": 0.9923781639303959,
+      "learning_rate": 7.360977058781847e-05,
+      "loss": 0.6871,
+      "step": 855
+    },
+    {
+      "epoch": 1.3230293663060277,
+      "grad_norm": 1.7370513259576752,
+      "learning_rate": 7.358634467346083e-05,
+      "loss": 0.725,
+      "step": 856
+    },
+    {
+      "epoch": 1.3245749613601236,
+      "grad_norm": 2.246456041219754,
+      "learning_rate": 7.356287964097795e-05,
+      "loss": 0.6404,
+      "step": 857
+    },
+    {
+      "epoch": 1.3261205564142196,
+      "grad_norm": 1.1800484045818855,
+      "learning_rate": 7.353937551769962e-05,
+      "loss": 0.6224,
+      "step": 858
+    },
+    {
+      "epoch": 1.3276661514683152,
+      "grad_norm": 1.4459039419605137,
+      "learning_rate": 7.351583233100118e-05,
+      "loss": 0.6058,
+      "step": 859
+    },
+    {
+      "epoch": 1.3292117465224111,
+      "grad_norm": 0.6928273626415625,
+      "learning_rate": 7.349225010830342e-05,
+      "loss": 0.5829,
+      "step": 860
+    },
+    {
+      "epoch": 1.3307573415765068,
+      "grad_norm": 1.0641321797411551,
+      "learning_rate": 7.346862887707269e-05,
+      "loss": 0.5559,
+      "step": 861
+    },
+    {
+      "epoch": 1.3323029366306027,
+      "grad_norm": 1.0485031936258609,
+      "learning_rate": 7.344496866482065e-05,
+      "loss": 0.5363,
+      "step": 862
+    },
+    {
+      "epoch": 1.3338485316846986,
+      "grad_norm": 1.0744847853013832,
+      "learning_rate": 7.342126949910447e-05,
+      "loss": 0.5442,
+      "step": 863
+    },
+    {
+      "epoch": 1.3353941267387945,
+      "grad_norm": 0.7111534457833215,
+      "learning_rate": 7.339753140752664e-05,
+      "loss": 0.5227,
+      "step": 864
+    },
+    {
+      "epoch": 1.3369397217928902,
+      "grad_norm": 0.7942962396605778,
+      "learning_rate": 7.337375441773499e-05,
+      "loss": 0.5243,
+      "step": 865
+    },
+    {
+      "epoch": 1.3384853168469861,
+      "grad_norm": 0.60059709479708,
+      "learning_rate": 7.334993855742265e-05,
+      "loss": 0.5097,
+      "step": 866
+    },
+    {
+      "epoch": 1.3400309119010818,
+      "grad_norm": 0.7581651481181462,
+      "learning_rate": 7.332608385432803e-05,
+      "loss": 0.5181,
+      "step": 867
+    },
+    {
+      "epoch": 1.3415765069551777,
+      "grad_norm": 0.5308804552574883,
+      "learning_rate": 7.33021903362348e-05,
+      "loss": 0.5008,
+      "step": 868
+    },
+    {
+      "epoch": 1.3431221020092736,
+      "grad_norm": 0.9556922253495935,
+      "learning_rate": 7.327825803097179e-05,
+      "loss": 0.4885,
+      "step": 869
+    },
+    {
+      "epoch": 1.3446676970633695,
+      "grad_norm": 1.6647993772515672,
+      "learning_rate": 7.325428696641306e-05,
+      "loss": 0.487,
+      "step": 870
+    },
+    {
+      "epoch": 1.3462132921174652,
+      "grad_norm": 0.557824650859362,
+      "learning_rate": 7.323027717047777e-05,
+      "loss": 0.4965,
+      "step": 871
+    },
+    {
+      "epoch": 1.3477588871715611,
+      "grad_norm": 0.991469443849489,
+      "learning_rate": 7.320622867113023e-05,
+      "loss": 0.4938,
+      "step": 872
+    },
+    {
+      "epoch": 1.3493044822256568,
+      "grad_norm": 0.6046190116577677,
+      "learning_rate": 7.318214149637976e-05,
+      "loss": 0.4816,
+      "step": 873
+    },
+    {
+      "epoch": 1.3508500772797527,
+      "grad_norm": 0.7644227271098579,
+      "learning_rate": 7.315801567428081e-05,
+      "loss": 0.4792,
+      "step": 874
+    },
+    {
+      "epoch": 1.3523956723338486,
+      "grad_norm": 0.6458372967237748,
+      "learning_rate": 7.313385123293281e-05,
+      "loss": 0.482,
+      "step": 875
+    },
+    {
+      "epoch": 1.3539412673879443,
+      "grad_norm": 0.6907800581217399,
+      "learning_rate": 7.310964820048013e-05,
+      "loss": 0.4808,
+      "step": 876
+    },
+    {
+      "epoch": 1.3554868624420402,
+      "grad_norm": 0.6498021432208272,
+      "learning_rate": 7.308540660511214e-05,
+      "loss": 0.4713,
+      "step": 877
+    },
+    {
+      "epoch": 1.3570324574961359,
+      "grad_norm": 0.5333204285641276,
+      "learning_rate": 7.30611264750631e-05,
+      "loss": 0.4673,
+      "step": 878
+    },
+    {
+      "epoch": 1.3585780525502318,
+      "grad_norm": 0.5259273615462304,
+      "learning_rate": 7.303680783861216e-05,
+      "loss": 0.4622,
+      "step": 879
+    },
+    {
+      "epoch": 1.3601236476043277,
+      "grad_norm": 0.4221242373141024,
+      "learning_rate": 7.301245072408333e-05,
+      "loss": 0.467,
+      "step": 880
+    },
+    {
+      "epoch": 1.3616692426584236,
+      "grad_norm": 0.5096799188944314,
+      "learning_rate": 7.29880551598454e-05,
+      "loss": 0.4589,
+      "step": 881
+    },
+    {
+      "epoch": 1.3632148377125193,
+      "grad_norm": 0.4932511435948369,
+      "learning_rate": 7.296362117431197e-05,
+      "loss": 0.4532,
+      "step": 882
+    },
+    {
+      "epoch": 1.3647604327666152,
+      "grad_norm": 0.49911424004834976,
+      "learning_rate": 7.293914879594138e-05,
+      "loss": 0.4632,
+      "step": 883
+    },
+    {
+      "epoch": 1.3663060278207109,
+      "grad_norm": 0.4229006710362692,
+      "learning_rate": 7.291463805323668e-05,
+      "loss": 0.4608,
+      "step": 884
+    },
+    {
+      "epoch": 1.3678516228748068,
+      "grad_norm": 0.5394785447176259,
+      "learning_rate": 7.289008897474564e-05,
+      "loss": 0.4577,
+      "step": 885
+    },
+    {
+      "epoch": 1.3693972179289027,
+      "grad_norm": 0.38824651713858316,
+      "learning_rate": 7.286550158906062e-05,
+      "loss": 0.4386,
+      "step": 886
+    },
+    {
+      "epoch": 1.3709428129829986,
+      "grad_norm": 0.5566933255367444,
+      "learning_rate": 7.284087592481867e-05,
+      "loss": 0.4456,
+      "step": 887
+    },
+    {
+      "epoch": 1.3724884080370943,
+      "grad_norm": 0.3563089736478836,
+      "learning_rate": 7.281621201070133e-05,
+      "loss": 0.4452,
+      "step": 888
+    },
+    {
+      "epoch": 1.3740340030911902,
+      "grad_norm": 0.3561209994399552,
+      "learning_rate": 7.279150987543476e-05,
+      "loss": 0.4557,
+      "step": 889
+    },
+    {
+      "epoch": 1.3755795981452859,
+      "grad_norm": 0.36769309647797255,
+      "learning_rate": 7.276676954778964e-05,
+      "loss": 0.4525,
+      "step": 890
+    },
+    {
+      "epoch": 1.3771251931993818,
+      "grad_norm": 0.3431461137335865,
+      "learning_rate": 7.27419910565811e-05,
+      "loss": 0.4485,
+      "step": 891
+    },
+    {
+      "epoch": 1.3786707882534777,
+      "grad_norm": 0.3652454301040794,
+      "learning_rate": 7.271717443066871e-05,
+      "loss": 0.4506,
+      "step": 892
+    },
+    {
+      "epoch": 1.3802163833075733,
+      "grad_norm": 0.31578645288650126,
+      "learning_rate": 7.26923196989565e-05,
+      "loss": 0.4332,
+      "step": 893
+    },
+    {
+      "epoch": 1.3817619783616693,
+      "grad_norm": 0.34860086522725214,
+      "learning_rate": 7.266742689039284e-05,
+      "loss": 0.4448,
+      "step": 894
+    },
+    {
+      "epoch": 1.383307573415765,
+      "grad_norm": 0.2845785029690209,
+      "learning_rate": 7.264249603397046e-05,
+      "loss": 0.4472,
+      "step": 895
+    },
+    {
+      "epoch": 1.3848531684698608,
+      "grad_norm": 0.28915228241658936,
+      "learning_rate": 7.261752715872643e-05,
+      "loss": 0.4408,
+      "step": 896
+    },
+    {
+      "epoch": 1.3863987635239567,
+      "grad_norm": 0.271133894870028,
+      "learning_rate": 7.259252029374209e-05,
+      "loss": 0.4449,
+      "step": 897
+    },
+    {
+      "epoch": 1.3879443585780527,
+      "grad_norm": 0.2311051594218128,
+      "learning_rate": 7.256747546814298e-05,
+      "loss": 0.4373,
+      "step": 898
+    },
+    {
+      "epoch": 1.3894899536321483,
+      "grad_norm": 0.2853715894695218,
+      "learning_rate": 7.254239271109891e-05,
+      "loss": 0.4466,
+      "step": 899
+    },
+    {
+      "epoch": 1.3910355486862442,
+      "grad_norm": 0.25346013690449665,
+      "learning_rate": 7.251727205182384e-05,
+      "loss": 0.4372,
+      "step": 900
+    },
+    {
+      "epoch": 1.39258114374034,
+      "grad_norm": 0.2785666994423498,
+      "learning_rate": 7.249211351957591e-05,
+      "loss": 0.4345,
+      "step": 901
+    },
+    {
+      "epoch": 1.3941267387944358,
+      "grad_norm": 0.2785254538459984,
+      "learning_rate": 7.246691714365732e-05,
+      "loss": 0.4314,
+      "step": 902
+    },
+    {
+      "epoch": 1.3956723338485317,
+      "grad_norm": 0.36191851798747987,
+      "learning_rate": 7.244168295341437e-05,
+      "loss": 0.4327,
+      "step": 903
+    },
+    {
+      "epoch": 1.3972179289026276,
+      "grad_norm": 0.23887449436457348,
+      "learning_rate": 7.241641097823742e-05,
+      "loss": 0.4418,
+      "step": 904
+    },
+    {
+      "epoch": 1.3987635239567233,
+      "grad_norm": 0.26704044921405956,
+      "learning_rate": 7.239110124756081e-05,
+      "loss": 0.4413,
+      "step": 905
+    },
+    {
+      "epoch": 1.4003091190108192,
+      "grad_norm": 0.2464024352275652,
+      "learning_rate": 7.236575379086286e-05,
+      "loss": 0.4318,
+      "step": 906
+    },
+    {
+      "epoch": 1.401854714064915,
+      "grad_norm": 0.26600554185547204,
+      "learning_rate": 7.234036863766586e-05,
+      "loss": 0.4359,
+      "step": 907
+    },
+    {
+      "epoch": 1.4034003091190108,
+      "grad_norm": 0.27451211025062616,
+      "learning_rate": 7.231494581753596e-05,
+      "loss": 0.4356,
+      "step": 908
+    },
+    {
+      "epoch": 1.4049459041731067,
+      "grad_norm": 0.2998713303413568,
+      "learning_rate": 7.22894853600832e-05,
+      "loss": 0.4493,
+      "step": 909
+    },
+    {
+      "epoch": 1.4064914992272024,
+      "grad_norm": 0.3624005258089661,
+      "learning_rate": 7.226398729496148e-05,
+      "loss": 0.4358,
+      "step": 910
+    },
+    {
+      "epoch": 1.4080370942812983,
+      "grad_norm": 0.37741966314468167,
+      "learning_rate": 7.223845165186845e-05,
+      "loss": 0.4252,
+      "step": 911
+    },
+    {
+      "epoch": 1.409582689335394,
+      "grad_norm": 0.3757885636786183,
+      "learning_rate": 7.221287846054559e-05,
+      "loss": 0.428,
+      "step": 912
+    },
+    {
+      "epoch": 1.41112828438949,
+      "grad_norm": 0.33431463700971414,
+      "learning_rate": 7.218726775077803e-05,
+      "loss": 0.431,
+      "step": 913
+    },
+    {
+      "epoch": 1.4126738794435858,
+      "grad_norm": 0.33944920095897424,
+      "learning_rate": 7.216161955239471e-05,
+      "loss": 0.4315,
+      "step": 914
+    },
+    {
+      "epoch": 1.4142194744976817,
+      "grad_norm": 0.40767393932426144,
+      "learning_rate": 7.213593389526812e-05,
+      "loss": 0.4309,
+      "step": 915
+    },
+    {
+      "epoch": 1.4157650695517774,
+      "grad_norm": 0.4332624192257046,
+      "learning_rate": 7.211021080931446e-05,
+      "loss": 0.4239,
+      "step": 916
+    },
+    {
+      "epoch": 1.4173106646058733,
+      "grad_norm": 0.5367545117142651,
+      "learning_rate": 7.208445032449347e-05,
+      "loss": 0.4356,
+      "step": 917
+    },
+    {
+      "epoch": 1.418856259659969,
+      "grad_norm": 0.6797508269994943,
+      "learning_rate": 7.205865247080848e-05,
+      "loss": 0.4236,
+      "step": 918
+    },
+    {
+      "epoch": 1.4204018547140649,
+      "grad_norm": 0.8065158861855208,
+      "learning_rate": 7.203281727830634e-05,
+      "loss": 0.4273,
+      "step": 919
+    },
+    {
+      "epoch": 1.4219474497681608,
+      "grad_norm": 0.8535877358815611,
+      "learning_rate": 7.200694477707737e-05,
+      "loss": 0.4233,
+      "step": 920
+    },
+    {
+      "epoch": 1.4234930448222567,
+      "grad_norm": 0.7057296559667867,
+      "learning_rate": 7.198103499725538e-05,
+      "loss": 0.4367,
+      "step": 921
+    },
+    {
+      "epoch": 1.4250386398763524,
+      "grad_norm": 0.5329342219908334,
+      "learning_rate": 7.195508796901754e-05,
+      "loss": 0.4394,
+      "step": 922
+    },
+    {
+      "epoch": 1.4265842349304483,
+      "grad_norm": 0.4076659192290988,
+      "learning_rate": 7.192910372258447e-05,
+      "loss": 0.4256,
+      "step": 923
+    },
+    {
+      "epoch": 1.428129829984544,
+      "grad_norm": 0.26064222228962725,
+      "learning_rate": 7.19030822882201e-05,
+      "loss": 0.4323,
+      "step": 924
+    },
+    {
+      "epoch": 1.4296754250386399,
+      "grad_norm": 0.2585035718372981,
+      "learning_rate": 7.187702369623167e-05,
+      "loss": 0.4369,
+      "step": 925
+    },
+    {
+      "epoch": 1.4312210200927358,
+      "grad_norm": 0.403830163370915,
+      "learning_rate": 7.185092797696969e-05,
+      "loss": 0.4256,
+      "step": 926
+    },
+    {
+      "epoch": 1.4327666151468315,
+      "grad_norm": 0.515104682498424,
+      "learning_rate": 7.182479516082797e-05,
+      "loss": 0.4315,
+      "step": 927
+    },
+    {
+      "epoch": 1.4343122102009274,
+      "grad_norm": 0.5590711787713675,
+      "learning_rate": 7.179862527824347e-05,
+      "loss": 0.4314,
+      "step": 928
+    },
+    {
+      "epoch": 1.435857805255023,
+      "grad_norm": 0.5742581145612555,
+      "learning_rate": 7.177241835969632e-05,
+      "loss": 0.4365,
+      "step": 929
+    },
+    {
+      "epoch": 1.437403400309119,
+      "grad_norm": 0.5507747284900131,
+      "learning_rate": 7.174617443570981e-05,
+      "loss": 0.4252,
+      "step": 930
+    },
+    {
+      "epoch": 1.4389489953632149,
+      "grad_norm": 0.43550998310891703,
+      "learning_rate": 7.171989353685035e-05,
+      "loss": 0.4335,
+      "step": 931
+    },
+    {
+      "epoch": 1.4404945904173108,
+      "grad_norm": 0.34969190686092066,
+      "learning_rate": 7.169357569372738e-05,
+      "loss": 0.4292,
+      "step": 932
+    },
+    {
+      "epoch": 1.4420401854714064,
+      "grad_norm": 0.3121259773521068,
+      "learning_rate": 7.166722093699335e-05,
+      "loss": 0.4311,
+      "step": 933
+    },
+    {
+      "epoch": 1.4435857805255023,
+      "grad_norm": 0.3040459077132056,
+      "learning_rate": 7.164082929734375e-05,
+      "loss": 0.4149,
+      "step": 934
+    },
+    {
+      "epoch": 1.445131375579598,
+      "grad_norm": 0.44094007127066864,
+      "learning_rate": 7.161440080551701e-05,
+      "loss": 0.4321,
+      "step": 935
+    },
+    {
+      "epoch": 1.446676970633694,
+      "grad_norm": 0.5279397373158785,
+      "learning_rate": 7.158793549229449e-05,
+      "loss": 0.4201,
+      "step": 936
+    },
+    {
+      "epoch": 1.4482225656877898,
+      "grad_norm": 0.4726234844469759,
+      "learning_rate": 7.156143338850041e-05,
+      "loss": 0.4244,
+      "step": 937
+    },
+    {
+      "epoch": 1.4497681607418857,
+      "grad_norm": 0.4398190703180919,
+      "learning_rate": 7.153489452500186e-05,
+      "loss": 0.4271,
+      "step": 938
+    },
+    {
+      "epoch": 1.4513137557959814,
+      "grad_norm": 0.3944713485012089,
+      "learning_rate": 7.150831893270874e-05,
+      "loss": 0.4336,
+      "step": 939
+    },
+    {
+      "epoch": 1.4528593508500773,
+      "grad_norm": 0.2704413055996092,
+      "learning_rate": 7.148170664257374e-05,
+      "loss": 0.4249,
+      "step": 940
+    },
+    {
+      "epoch": 1.454404945904173,
+      "grad_norm": 0.267487761883538,
+      "learning_rate": 7.145505768559225e-05,
+      "loss": 0.4238,
+      "step": 941
+    },
+    {
+      "epoch": 1.455950540958269,
+      "grad_norm": 0.35855880795815,
+      "learning_rate": 7.142837209280246e-05,
+      "loss": 0.4251,
+      "step": 942
+    },
+    {
+      "epoch": 1.4574961360123648,
+      "grad_norm": 0.43446999920977725,
+      "learning_rate": 7.14016498952851e-05,
+      "loss": 0.4173,
+      "step": 943
+    },
+    {
+      "epoch": 1.4590417310664605,
+      "grad_norm": 0.46506642846282903,
+      "learning_rate": 7.137489112416363e-05,
+      "loss": 0.4343,
+      "step": 944
+    },
+    {
+      "epoch": 1.4605873261205564,
+      "grad_norm": 0.4088930431765722,
+      "learning_rate": 7.134809581060407e-05,
+      "loss": 0.4211,
+      "step": 945
+    },
+    {
+      "epoch": 1.4621329211746523,
+      "grad_norm": 0.27778175203358063,
+      "learning_rate": 7.132126398581504e-05,
+      "loss": 0.4273,
+      "step": 946
+    },
+    {
+      "epoch": 1.463678516228748,
+      "grad_norm": 0.2384600394764093,
+      "learning_rate": 7.129439568104762e-05,
+      "loss": 0.4189,
+      "step": 947
+    },
+    {
+      "epoch": 1.465224111282844,
+      "grad_norm": 0.24158481481830246,
+      "learning_rate": 7.126749092759542e-05,
+      "loss": 0.4222,
+      "step": 948
+    },
+    {
+      "epoch": 1.4667697063369398,
+      "grad_norm": 0.2451523418282528,
+      "learning_rate": 7.124054975679449e-05,
+      "loss": 0.4311,
+      "step": 949
+    },
+    {
+      "epoch": 1.4683153013910355,
+      "grad_norm": 0.32254793087115824,
+      "learning_rate": 7.121357220002331e-05,
+      "loss": 0.4172,
+      "step": 950
+    },
+    {
+      "epoch": 1.4698608964451314,
+      "grad_norm": 0.44144955558690707,
+      "learning_rate": 7.118655828870274e-05,
+      "loss": 0.4185,
+      "step": 951
+    },
+    {
+      "epoch": 1.471406491499227,
+      "grad_norm": 0.537749720995868,
+      "learning_rate": 7.115950805429591e-05,
+      "loss": 0.4321,
+      "step": 952
+    },
+    {
+      "epoch": 1.472952086553323,
+      "grad_norm": 0.5715097587899174,
+      "learning_rate": 7.11324215283084e-05,
+      "loss": 0.4237,
+      "step": 953
+    },
+    {
+      "epoch": 1.474497681607419,
+      "grad_norm": 0.6273212266327671,
+      "learning_rate": 7.110529874228792e-05,
+      "loss": 0.4319,
+      "step": 954
+    },
+    {
+      "epoch": 1.4760432766615148,
+      "grad_norm": 0.7889797687201212,
+      "learning_rate": 7.107813972782448e-05,
+      "loss": 0.4234,
+      "step": 955
+    },
+    {
+      "epoch": 1.4775888717156105,
+      "grad_norm": 0.7264618248619983,
+      "learning_rate": 7.105094451655029e-05,
+      "loss": 0.4232,
+      "step": 956
+    },
+    {
+      "epoch": 1.4791344667697064,
+      "grad_norm": 0.7089214741759883,
+      "learning_rate": 7.102371314013969e-05,
+      "loss": 0.4343,
+      "step": 957
+    },
+    {
+      "epoch": 1.480680061823802,
+      "grad_norm": 0.6617679962496519,
+      "learning_rate": 7.099644563030914e-05,
+      "loss": 0.4282,
+      "step": 958
+    },
+    {
+      "epoch": 1.482225656877898,
+      "grad_norm": 0.5594902654488881,
+      "learning_rate": 7.096914201881722e-05,
+      "loss": 0.4238,
+      "step": 959
+    },
+    {
+      "epoch": 1.4837712519319939,
+      "grad_norm": 0.4634050282986392,
+      "learning_rate": 7.09418023374645e-05,
+      "loss": 0.4209,
+      "step": 960
+    },
+    {
+      "epoch": 1.4853168469860896,
+      "grad_norm": 0.38011653898573167,
+      "learning_rate": 7.091442661809364e-05,
+      "loss": 0.4226,
+      "step": 961
+    },
+    {
+      "epoch": 1.4868624420401855,
+      "grad_norm": 0.28704041077599796,
+      "learning_rate": 7.088701489258921e-05,
+      "loss": 0.4237,
+      "step": 962
+    },
+    {
+      "epoch": 1.4884080370942814,
+      "grad_norm": 0.2830439462631011,
+      "learning_rate": 7.085956719287773e-05,
+      "loss": 0.4239,
+      "step": 963
+    },
+    {
+      "epoch": 1.489953632148377,
+      "grad_norm": 0.3733602880135298,
+      "learning_rate": 7.083208355092763e-05,
+      "loss": 0.4195,
+      "step": 964
+    },
+    {
+      "epoch": 1.491499227202473,
+      "grad_norm": 0.4886140984615729,
+      "learning_rate": 7.080456399874919e-05,
+      "loss": 0.4272,
+      "step": 965
+    },
+    {
+      "epoch": 1.4930448222565689,
+      "grad_norm": 0.5820865498538151,
+      "learning_rate": 7.077700856839453e-05,
+      "loss": 0.4206,
+      "step": 966
+    },
+    {
+      "epoch": 1.4945904173106646,
+      "grad_norm": 0.6735568679899008,
+      "learning_rate": 7.074941729195756e-05,
+      "loss": 0.4216,
+      "step": 967
+    },
+    {
+      "epoch": 1.4961360123647605,
+      "grad_norm": 0.7238533873018045,
+      "learning_rate": 7.072179020157394e-05,
+      "loss": 0.4237,
+      "step": 968
+    },
+    {
+      "epoch": 1.4976816074188561,
+      "grad_norm": 0.639253964759351,
+      "learning_rate": 7.069412732942097e-05,
+      "loss": 0.4267,
+      "step": 969
+    },
+    {
+      "epoch": 1.499227202472952,
+      "grad_norm": 0.517856341056609,
+      "learning_rate": 7.066642870771775e-05,
+      "loss": 0.4221,
+      "step": 970
+    },
+    {
+      "epoch": 1.500772797527048,
+      "grad_norm": 0.3988654450549474,
+      "learning_rate": 7.063869436872495e-05,
+      "loss": 0.4327,
+      "step": 971
+    },
+    {
+      "epoch": 1.5023183925811439,
+      "grad_norm": 0.3364989820133205,
+      "learning_rate": 7.06109243447448e-05,
+      "loss": 0.4195,
+      "step": 972
+    },
+    {
+      "epoch": 1.5038639876352395,
+      "grad_norm": 0.32790952085979724,
+      "learning_rate": 7.05831186681212e-05,
+      "loss": 0.4268,
+      "step": 973
+    },
+    {
+      "epoch": 1.5054095826893354,
+      "grad_norm": 0.3985317931788811,
+      "learning_rate": 7.055527737123947e-05,
+      "loss": 0.4199,
+      "step": 974
+    },
+    {
+      "epoch": 1.5069551777434311,
+      "grad_norm": 0.5260103397842368,
+      "learning_rate": 7.052740048652645e-05,
+      "loss": 0.4263,
+      "step": 975
+    },
+    {
+      "epoch": 1.508500772797527,
+      "grad_norm": 0.6525153290567401,
+      "learning_rate": 7.049948804645047e-05,
+      "loss": 0.4272,
+      "step": 976
+    },
+    {
+      "epoch": 1.510046367851623,
+      "grad_norm": 0.6846903618954039,
+      "learning_rate": 7.047154008352122e-05,
+      "loss": 0.4206,
+      "step": 977
+    },
+    {
+      "epoch": 1.5115919629057188,
+      "grad_norm": 0.6199992700429547,
+      "learning_rate": 7.044355663028981e-05,
+      "loss": 0.4228,
+      "step": 978
+    },
+    {
+      "epoch": 1.5131375579598145,
+      "grad_norm": 0.5578406454030075,
+      "learning_rate": 7.041553771934862e-05,
+      "loss": 0.4163,
+      "step": 979
+    },
+    {
+      "epoch": 1.5146831530139102,
+      "grad_norm": 0.48284322718920997,
+      "learning_rate": 7.03874833833314e-05,
+      "loss": 0.4325,
+      "step": 980
+    },
+    {
+      "epoch": 1.5162287480680061,
+      "grad_norm": 0.4282926517985966,
+      "learning_rate": 7.035939365491312e-05,
+      "loss": 0.4275,
+      "step": 981
+    },
+    {
+      "epoch": 1.517774343122102,
+      "grad_norm": 0.3969906684645652,
+      "learning_rate": 7.033126856680996e-05,
+      "loss": 0.4251,
+      "step": 982
+    },
+    {
+      "epoch": 1.519319938176198,
+      "grad_norm": 0.33465065506818753,
+      "learning_rate": 7.030310815177933e-05,
+      "loss": 0.4177,
+      "step": 983
+    },
+    {
+      "epoch": 1.5208655332302936,
+      "grad_norm": 0.3032412982070374,
+      "learning_rate": 7.027491244261972e-05,
+      "loss": 0.4232,
+      "step": 984
+    },
+    {
+      "epoch": 1.5224111282843895,
+      "grad_norm": 0.43752458748824424,
+      "learning_rate": 7.02466814721708e-05,
+      "loss": 0.4251,
+      "step": 985
+    },
+    {
+      "epoch": 1.5239567233384852,
+      "grad_norm": 0.4924572904287951,
+      "learning_rate": 7.021841527331327e-05,
+      "loss": 0.423,
+      "step": 986
+    },
+    {
+      "epoch": 1.525502318392581,
+      "grad_norm": 0.4526093033505471,
+      "learning_rate": 7.019011387896884e-05,
+      "loss": 0.4196,
+      "step": 987
+    },
+    {
+      "epoch": 1.527047913446677,
+      "grad_norm": 0.49978471023791293,
+      "learning_rate": 7.016177732210025e-05,
+      "loss": 0.4212,
+      "step": 988
+    },
+    {
+      "epoch": 1.528593508500773,
+      "grad_norm": 0.539499544187315,
+      "learning_rate": 7.013340563571119e-05,
+      "loss": 0.4113,
+      "step": 989
+    },
+    {
+      "epoch": 1.5301391035548686,
+      "grad_norm": 0.4868929178302569,
+      "learning_rate": 7.010499885284624e-05,
+      "loss": 0.4136,
+      "step": 990
+    },
+    {
+      "epoch": 1.5316846986089645,
+      "grad_norm": 0.4980297906843877,
+      "learning_rate": 7.007655700659087e-05,
+      "loss": 0.4299,
+      "step": 991
+    },
+    {
+      "epoch": 1.5332302936630602,
+      "grad_norm": 0.582892376266293,
+      "learning_rate": 7.004808013007142e-05,
+      "loss": 0.4242,
+      "step": 992
+    },
+    {
+      "epoch": 1.534775888717156,
+      "grad_norm": 0.5798827136342914,
+      "learning_rate": 7.001956825645496e-05,
+      "loss": 0.4194,
+      "step": 993
+    },
+    {
+      "epoch": 1.536321483771252,
+      "grad_norm": 0.5309078533844069,
+      "learning_rate": 6.999102141894939e-05,
+      "loss": 0.4234,
+      "step": 994
+    },
+    {
+      "epoch": 1.537867078825348,
+      "grad_norm": 0.5430064404970218,
+      "learning_rate": 6.99624396508033e-05,
+      "loss": 0.4254,
+      "step": 995
+    },
+    {
+      "epoch": 1.5394126738794436,
+      "grad_norm": 0.46571511147801864,
+      "learning_rate": 6.993382298530598e-05,
+      "loss": 0.4128,
+      "step": 996
+    },
+    {
+      "epoch": 1.5409582689335393,
+      "grad_norm": 0.29907402931903726,
+      "learning_rate": 6.990517145578733e-05,
+      "loss": 0.4205,
+      "step": 997
+    },
+    {
+      "epoch": 1.5425038639876352,
+      "grad_norm": 0.33136439887119,
+      "learning_rate": 6.987648509561788e-05,
+      "loss": 0.4279,
+      "step": 998
+    },
+    {
+      "epoch": 1.544049459041731,
+      "grad_norm": 0.4477695438478884,
+      "learning_rate": 6.984776393820876e-05,
+      "loss": 0.4212,
+      "step": 999
+    },
+    {
+      "epoch": 1.545595054095827,
+      "grad_norm": 0.4487614235398153,
+      "learning_rate": 6.981900801701156e-05,
+      "loss": 0.4167,
+      "step": 1000
+    },
+    {
+      "epoch": 1.5471406491499229,
+      "grad_norm": 0.36555309719909523,
+      "learning_rate": 6.979021736551841e-05,
+      "loss": 0.4261,
+      "step": 1001
+    },
+    {
+      "epoch": 1.5486862442040186,
+      "grad_norm": 0.3501723419708198,
+      "learning_rate": 6.976139201726189e-05,
+      "loss": 0.418,
+      "step": 1002
+    },
+    {
+      "epoch": 1.5502318392581143,
+      "grad_norm": 0.3651564751235074,
+      "learning_rate": 6.973253200581493e-05,
+      "loss": 0.4221,
+      "step": 1003
+    },
+    {
+      "epoch": 1.5517774343122102,
+      "grad_norm": 0.34570119420075257,
+      "learning_rate": 6.970363736479094e-05,
+      "loss": 0.414,
+      "step": 1004
+    },
+    {
+      "epoch": 1.553323029366306,
+      "grad_norm": 0.4805761331573332,
+      "learning_rate": 6.967470812784356e-05,
+      "loss": 0.4182,
+      "step": 1005
+    },
+    {
+      "epoch": 1.554868624420402,
+      "grad_norm": 0.5649790110710811,
+      "learning_rate": 6.964574432866679e-05,
+      "loss": 0.4298,
+      "step": 1006
+    },
+    {
+      "epoch": 1.5564142194744977,
+      "grad_norm": 0.6221087425963319,
+      "learning_rate": 6.961674600099484e-05,
+      "loss": 0.4216,
+      "step": 1007
+    },
+    {
+      "epoch": 1.5579598145285936,
+      "grad_norm": 0.7181645539552841,
+      "learning_rate": 6.958771317860218e-05,
+      "loss": 0.4316,
+      "step": 1008
+    },
+    {
+      "epoch": 1.5595054095826892,
+      "grad_norm": 0.8221465058092516,
+      "learning_rate": 6.955864589530342e-05,
+      "loss": 0.4168,
+      "step": 1009
+    },
+    {
+      "epoch": 1.5610510046367851,
+      "grad_norm": 0.8898827490422138,
+      "learning_rate": 6.952954418495332e-05,
+      "loss": 0.4158,
+      "step": 1010
+    },
+    {
+      "epoch": 1.562596599690881,
+      "grad_norm": 0.9006817281728228,
+      "learning_rate": 6.950040808144674e-05,
+      "loss": 0.4169,
+      "step": 1011
+    },
+    {
+      "epoch": 1.564142194744977,
+      "grad_norm": 0.7845557662937626,
+      "learning_rate": 6.947123761871858e-05,
+      "loss": 0.4235,
+      "step": 1012
+    },
+    {
+      "epoch": 1.5656877897990726,
+      "grad_norm": 0.6167610107609887,
+      "learning_rate": 6.944203283074381e-05,
+      "loss": 0.4178,
+      "step": 1013
+    },
+    {
+      "epoch": 1.5672333848531683,
+      "grad_norm": 0.45690594363128245,
+      "learning_rate": 6.94127937515373e-05,
+      "loss": 0.4302,
+      "step": 1014
+    },
+    {
+      "epoch": 1.5687789799072642,
+      "grad_norm": 0.3499403023769347,
+      "learning_rate": 6.938352041515393e-05,
+      "loss": 0.4199,
+      "step": 1015
+    },
+    {
+      "epoch": 1.5703245749613601,
+      "grad_norm": 0.3572705438871156,
+      "learning_rate": 6.935421285568842e-05,
+      "loss": 0.4173,
+      "step": 1016
+    },
+    {
+      "epoch": 1.571870170015456,
+      "grad_norm": 0.44950483927508295,
+      "learning_rate": 6.932487110727543e-05,
+      "loss": 0.4209,
+      "step": 1017
+    },
+    {
+      "epoch": 1.573415765069552,
+      "grad_norm": 3.1292477500728544,
+      "learning_rate": 6.929549520408934e-05,
+      "loss": 0.4231,
+      "step": 1018
+    },
+    {
+      "epoch": 1.5749613601236476,
+      "grad_norm": 1.8980381367708854,
+      "learning_rate": 6.926608518034438e-05,
+      "loss": 0.4377,
+      "step": 1019
+    },
+    {
+      "epoch": 1.5765069551777433,
+      "grad_norm": 2.830478244840452,
+      "learning_rate": 6.923664107029452e-05,
+      "loss": 0.4763,
+      "step": 1020
+    },
+    {
+      "epoch": 1.5780525502318392,
+      "grad_norm": 0.9592222026976776,
+      "learning_rate": 6.920716290823337e-05,
+      "loss": 0.4398,
+      "step": 1021
+    },
+    {
+      "epoch": 1.5795981452859351,
+      "grad_norm": 1.6478299306991953,
+      "learning_rate": 6.917765072849427e-05,
+      "loss": 0.4653,
+      "step": 1022
+    },
+    {
+      "epoch": 1.581143740340031,
+      "grad_norm": 0.6840248212604744,
+      "learning_rate": 6.914810456545015e-05,
+      "loss": 0.4435,
+      "step": 1023
+    },
+    {
+      "epoch": 1.5826893353941267,
+      "grad_norm": 1.1177393718394315,
+      "learning_rate": 6.911852445351352e-05,
+      "loss": 0.4419,
+      "step": 1024
+    },
+    {
+      "epoch": 1.5842349304482226,
+      "grad_norm": 0.6531693590989968,
+      "learning_rate": 6.908891042713643e-05,
+      "loss": 0.4516,
+      "step": 1025
+    },
+    {
+      "epoch": 1.5857805255023183,
+      "grad_norm": 0.7123299350546227,
+      "learning_rate": 6.905926252081043e-05,
+      "loss": 0.4379,
+      "step": 1026
+    },
+    {
+      "epoch": 1.5873261205564142,
+      "grad_norm": 0.6393235977873607,
+      "learning_rate": 6.902958076906655e-05,
+      "loss": 0.4366,
+      "step": 1027
+    },
+    {
+      "epoch": 1.58887171561051,
+      "grad_norm": 0.6699748000095743,
+      "learning_rate": 6.899986520647521e-05,
+      "loss": 0.4337,
+      "step": 1028
+    },
+    {
+      "epoch": 1.590417310664606,
+      "grad_norm": 0.5965947765071413,
+      "learning_rate": 6.897011586764622e-05,
+      "loss": 0.4321,
+      "step": 1029
+    },
+    {
+      "epoch": 1.5919629057187017,
+      "grad_norm": 0.5056237819580799,
+      "learning_rate": 6.894033278722874e-05,
+      "loss": 0.4277,
+      "step": 1030
+    },
+    {
+      "epoch": 1.5935085007727974,
+      "grad_norm": 0.5083639625187817,
+      "learning_rate": 6.891051599991119e-05,
+      "loss": 0.4234,
+      "step": 1031
+    },
+    {
+      "epoch": 1.5950540958268933,
+      "grad_norm": 0.5046816637015624,
+      "learning_rate": 6.888066554042134e-05,
+      "loss": 0.4333,
+      "step": 1032
+    },
+    {
+      "epoch": 1.5965996908809892,
+      "grad_norm": 0.40281242552896174,
+      "learning_rate": 6.885078144352606e-05,
+      "loss": 0.4271,
+      "step": 1033
+    },
+    {
+      "epoch": 1.598145285935085,
+      "grad_norm": 0.4000957468169698,
+      "learning_rate": 6.882086374403148e-05,
+      "loss": 0.434,
+      "step": 1034
+    },
+    {
+      "epoch": 1.599690880989181,
+      "grad_norm": 0.4656514897022174,
+      "learning_rate": 6.879091247678285e-05,
+      "loss": 0.4336,
+      "step": 1035
+    },
+    {
+      "epoch": 1.6012364760432767,
+      "grad_norm": 0.3851788185826526,
+      "learning_rate": 6.87609276766645e-05,
+      "loss": 0.4265,
+      "step": 1036
+    },
+    {
+      "epoch": 1.6027820710973724,
+      "grad_norm": 0.3663704499902741,
+      "learning_rate": 6.873090937859982e-05,
+      "loss": 0.4309,
+      "step": 1037
+    },
+    {
+      "epoch": 1.6043276661514683,
+      "grad_norm": 0.7346406351853385,
+      "learning_rate": 6.870085761755122e-05,
+      "loss": 0.4228,
+      "step": 1038
+    },
+    {
+      "epoch": 1.6058732612055642,
+      "grad_norm": 0.46844196197914234,
+      "learning_rate": 6.867077242852012e-05,
+      "loss": 0.425,
+      "step": 1039
+    },
+    {
+      "epoch": 1.60741885625966,
+      "grad_norm": 0.49329122070089737,
+      "learning_rate": 6.864065384654676e-05,
+      "loss": 0.4243,
+      "step": 1040
+    },
+    {
+      "epoch": 1.6089644513137558,
+      "grad_norm": 0.3696669871281922,
+      "learning_rate": 6.861050190671043e-05,
+      "loss": 0.4185,
+      "step": 1041
+    },
+    {
+      "epoch": 1.6105100463678517,
+      "grad_norm": 0.33531974942756926,
+      "learning_rate": 6.858031664412914e-05,
+      "loss": 0.4301,
+      "step": 1042
+    },
+    {
+      "epoch": 1.6120556414219473,
+      "grad_norm": 0.32366833386851807,
+      "learning_rate": 6.85500980939598e-05,
+      "loss": 0.4258,
+      "step": 1043
+    },
+    {
+      "epoch": 1.6136012364760433,
+      "grad_norm": 0.3198516867842109,
+      "learning_rate": 6.851984629139802e-05,
+      "loss": 0.4248,
+      "step": 1044
+    },
+    {
+      "epoch": 1.6151468315301392,
+      "grad_norm": 0.3683588176040148,
+      "learning_rate": 6.84895612716782e-05,
+      "loss": 0.4169,
+      "step": 1045
+    },
+    {
+      "epoch": 1.616692426584235,
+      "grad_norm": 0.3701188524554859,
+      "learning_rate": 6.845924307007339e-05,
+      "loss": 0.4221,
+      "step": 1046
+    },
+    {
+      "epoch": 1.6182380216383307,
+      "grad_norm": 0.2974961436601128,
+      "learning_rate": 6.84288917218953e-05,
+      "loss": 0.4268,
+      "step": 1047
+    },
+    {
+      "epoch": 1.6197836166924264,
+      "grad_norm": 0.28565881219366634,
+      "learning_rate": 6.839850726249423e-05,
+      "loss": 0.4229,
+      "step": 1048
+    },
+    {
+      "epoch": 1.6213292117465223,
+      "grad_norm": 0.3339494717656481,
+      "learning_rate": 6.83680897272591e-05,
+      "loss": 0.4194,
+      "step": 1049
+    },
+    {
+      "epoch": 1.6228748068006182,
+      "grad_norm": 0.37085420309921574,
+      "learning_rate": 6.833763915161727e-05,
+      "loss": 0.4234,
+      "step": 1050
+    },
+    {
+      "epoch": 1.6244204018547141,
+      "grad_norm": 0.3687746373768453,
+      "learning_rate": 6.830715557103465e-05,
+      "loss": 0.4255,
+      "step": 1051
+    },
+    {
+      "epoch": 1.62596599690881,
+      "grad_norm": 0.4853540323313802,
+      "learning_rate": 6.827663902101554e-05,
+      "loss": 0.4263,
+      "step": 1052
+    },
+    {
+      "epoch": 1.6275115919629057,
+      "grad_norm": 0.506980707599996,
+      "learning_rate": 6.82460895371027e-05,
+      "loss": 0.4244,
+      "step": 1053
+    },
+    {
+      "epoch": 1.6290571870170014,
+      "grad_norm": 0.4384308700255955,
+      "learning_rate": 6.82155071548772e-05,
+      "loss": 0.4341,
+      "step": 1054
+    },
+    {
+      "epoch": 1.6306027820710973,
+      "grad_norm": 0.29139242680715716,
+      "learning_rate": 6.818489190995842e-05,
+      "loss": 0.4151,
+      "step": 1055
+    },
+    {
+      "epoch": 1.6321483771251932,
+      "grad_norm": 0.24337771986454898,
+      "learning_rate": 6.815424383800405e-05,
+      "loss": 0.4257,
+      "step": 1056
+    },
+    {
+      "epoch": 1.6336939721792891,
+      "grad_norm": 0.30221847310026523,
+      "learning_rate": 6.812356297471e-05,
+      "loss": 0.4191,
+      "step": 1057
+    },
+    {
+      "epoch": 1.6352395672333848,
+      "grad_norm": 0.34157919884657423,
+      "learning_rate": 6.809284935581039e-05,
+      "loss": 0.4249,
+      "step": 1058
+    },
+    {
+      "epoch": 1.6367851622874807,
+      "grad_norm": 0.3627131538237721,
+      "learning_rate": 6.806210301707745e-05,
+      "loss": 0.4176,
+      "step": 1059
+    },
+    {
+      "epoch": 1.6383307573415764,
+      "grad_norm": 0.38110607158808035,
+      "learning_rate": 6.803132399432154e-05,
+      "loss": 0.4169,
+      "step": 1060
+    },
+    {
+      "epoch": 1.6398763523956723,
+      "grad_norm": 0.33843619857889295,
+      "learning_rate": 6.800051232339111e-05,
+      "loss": 0.4201,
+      "step": 1061
+    },
+    {
+      "epoch": 1.6414219474497682,
+      "grad_norm": 0.2546149999249603,
+      "learning_rate": 6.796966804017258e-05,
+      "loss": 0.413,
+      "step": 1062
+    },
+    {
+      "epoch": 1.6429675425038641,
+      "grad_norm": 0.18453053277347461,
+      "learning_rate": 6.79387911805904e-05,
+      "loss": 0.4175,
+      "step": 1063
+    },
+    {
+      "epoch": 1.6445131375579598,
+      "grad_norm": 0.24557042084259198,
+      "learning_rate": 6.790788178060697e-05,
+      "loss": 0.4165,
+      "step": 1064
+    },
+    {
+      "epoch": 1.6460587326120555,
+      "grad_norm": 0.2269674504241286,
+      "learning_rate": 6.787693987622254e-05,
+      "loss": 0.4185,
+      "step": 1065
+    },
+    {
+      "epoch": 1.6476043276661514,
+      "grad_norm": 0.19447194748578533,
+      "learning_rate": 6.784596550347526e-05,
+      "loss": 0.4366,
+      "step": 1066
+    },
+    {
+      "epoch": 1.6491499227202473,
+      "grad_norm": 0.23111756984294846,
+      "learning_rate": 6.78149586984411e-05,
+      "loss": 0.4316,
+      "step": 1067
+    },
+    {
+      "epoch": 1.6506955177743432,
+      "grad_norm": 0.2256982718893023,
+      "learning_rate": 6.778391949723375e-05,
+      "loss": 0.4127,
+      "step": 1068
+    },
+    {
+      "epoch": 1.652241112828439,
+      "grad_norm": 0.2332090184401883,
+      "learning_rate": 6.77528479360047e-05,
+      "loss": 0.4049,
+      "step": 1069
+    },
+    {
+      "epoch": 1.6537867078825348,
+      "grad_norm": 0.23913692934004846,
+      "learning_rate": 6.772174405094309e-05,
+      "loss": 0.4296,
+      "step": 1070
+    },
+    {
+      "epoch": 1.6553323029366305,
+      "grad_norm": 0.3520928886160808,
+      "learning_rate": 6.769060787827571e-05,
+      "loss": 0.4093,
+      "step": 1071
+    },
+    {
+      "epoch": 1.6568778979907264,
+      "grad_norm": 0.5369071929986187,
+      "learning_rate": 6.765943945426697e-05,
+      "loss": 0.4265,
+      "step": 1072
+    },
+    {
+      "epoch": 1.6584234930448223,
+      "grad_norm": 0.7007120576774098,
+      "learning_rate": 6.762823881521883e-05,
+      "loss": 0.4203,
+      "step": 1073
+    },
+    {
+      "epoch": 1.6599690880989182,
+      "grad_norm": 0.77923246824803,
+      "learning_rate": 6.759700599747077e-05,
+      "loss": 0.4313,
+      "step": 1074
+    },
+    {
+      "epoch": 1.6615146831530139,
+      "grad_norm": 0.740452462478401,
+      "learning_rate": 6.756574103739979e-05,
+      "loss": 0.4212,
+      "step": 1075
+    },
+    {
+      "epoch": 1.6630602782071098,
+      "grad_norm": 0.6678975079052005,
+      "learning_rate": 6.753444397142024e-05,
+      "loss": 0.4264,
+      "step": 1076
+    },
+    {
+      "epoch": 1.6646058732612055,
+      "grad_norm": 0.6013591124678156,
+      "learning_rate": 6.750311483598394e-05,
+      "loss": 0.4174,
+      "step": 1077
+    },
+    {
+      "epoch": 1.6661514683153014,
+      "grad_norm": 0.4321985899818665,
+      "learning_rate": 6.747175366758005e-05,
+      "loss": 0.4174,
+      "step": 1078
+    },
+    {
+      "epoch": 1.6676970633693973,
+      "grad_norm": 0.24934380170419806,
+      "learning_rate": 6.7440360502735e-05,
+      "loss": 0.4294,
+      "step": 1079
+    },
+    {
+      "epoch": 1.6692426584234932,
+      "grad_norm": 0.3200073164107801,
+      "learning_rate": 6.74089353780125e-05,
+      "loss": 0.4118,
+      "step": 1080
+    },
+    {
+      "epoch": 1.6707882534775889,
+      "grad_norm": 0.44053524474232914,
+      "learning_rate": 6.737747833001354e-05,
+      "loss": 0.4102,
+      "step": 1081
+    },
+    {
+      "epoch": 1.6723338485316845,
+      "grad_norm": 0.47983582823985493,
+      "learning_rate": 6.734598939537621e-05,
+      "loss": 0.4181,
+      "step": 1082
+    },
+    {
+      "epoch": 1.6738794435857804,
+      "grad_norm": 0.5191238281114948,
+      "learning_rate": 6.731446861077581e-05,
+      "loss": 0.4165,
+      "step": 1083
+    },
+    {
+      "epoch": 1.6754250386398764,
+      "grad_norm": 0.5029319945077619,
+      "learning_rate": 6.728291601292466e-05,
+      "loss": 0.4303,
+      "step": 1084
+    },
+    {
+      "epoch": 1.6769706336939723,
+      "grad_norm": 0.4309530318214376,
+      "learning_rate": 6.725133163857219e-05,
+      "loss": 0.4228,
+      "step": 1085
+    },
+    {
+      "epoch": 1.6785162287480682,
+      "grad_norm": 0.4503679207752796,
+      "learning_rate": 6.721971552450483e-05,
+      "loss": 0.4153,
+      "step": 1086
+    },
+    {
+      "epoch": 1.6800618238021638,
+      "grad_norm": 0.49171363614832303,
+      "learning_rate": 6.7188067707546e-05,
+      "loss": 0.4173,
+      "step": 1087
+    },
+    {
+      "epoch": 1.6816074188562595,
+      "grad_norm": 0.4889423713612924,
+      "learning_rate": 6.715638822455594e-05,
+      "loss": 0.4163,
+      "step": 1088
+    },
+    {
+      "epoch": 1.6831530139103554,
+      "grad_norm": 0.44497904897047286,
+      "learning_rate": 6.712467711243191e-05,
+      "loss": 0.4188,
+      "step": 1089
+    },
+    {
+      "epoch": 1.6846986089644513,
+      "grad_norm": 0.39806898970401583,
+      "learning_rate": 6.709293440810792e-05,
+      "loss": 0.4131,
+      "step": 1090
+    },
+    {
+      "epoch": 1.6862442040185472,
+      "grad_norm": 0.33195307395967655,
+      "learning_rate": 6.70611601485548e-05,
+      "loss": 0.4189,
+      "step": 1091
+    },
+    {
+      "epoch": 1.687789799072643,
+      "grad_norm": 0.8871998849464271,
+      "learning_rate": 6.702935437078017e-05,
+      "loss": 0.4168,
+      "step": 1092
+    },
+    {
+      "epoch": 1.6893353941267388,
+      "grad_norm": 0.44193443814333394,
+      "learning_rate": 6.699751711182827e-05,
+      "loss": 0.4101,
+      "step": 1093
+    },
+    {
+      "epoch": 1.6908809891808345,
+      "grad_norm": 0.4141123061157703,
+      "learning_rate": 6.696564840878009e-05,
+      "loss": 0.4262,
+      "step": 1094
+    },
+    {
+      "epoch": 1.6924265842349304,
+      "grad_norm": 0.4973649245570431,
+      "learning_rate": 6.693374829875322e-05,
+      "loss": 0.4225,
+      "step": 1095
+    },
+    {
+      "epoch": 1.6939721792890263,
+      "grad_norm": 0.6688557554900625,
+      "learning_rate": 6.690181681890177e-05,
+      "loss": 0.4114,
+      "step": 1096
+    },
+    {
+      "epoch": 1.6955177743431222,
+      "grad_norm": 0.6857166826042449,
+      "learning_rate": 6.686985400641651e-05,
+      "loss": 0.4217,
+      "step": 1097
+    },
+    {
+      "epoch": 1.697063369397218,
+      "grad_norm": 0.5830106585335436,
+      "learning_rate": 6.68378598985246e-05,
+      "loss": 0.4219,
+      "step": 1098
+    },
+    {
+      "epoch": 1.6986089644513136,
+      "grad_norm": 0.501857289501519,
+      "learning_rate": 6.680583453248967e-05,
+      "loss": 0.419,
+      "step": 1099
+    },
+    {
+      "epoch": 1.7001545595054095,
+      "grad_norm": 0.4796633764215541,
+      "learning_rate": 6.677377794561178e-05,
+      "loss": 0.423,
+      "step": 1100
+    },
+    {
+      "epoch": 1.7017001545595054,
+      "grad_norm": 0.43334114327907036,
+      "learning_rate": 6.674169017522734e-05,
+      "loss": 0.4201,
+      "step": 1101
+    },
+    {
+      "epoch": 1.7032457496136013,
+      "grad_norm": 0.3364871560607749,
+      "learning_rate": 6.670957125870911e-05,
+      "loss": 0.4278,
+      "step": 1102
+    },
+    {
+      "epoch": 1.7047913446676972,
+      "grad_norm": 0.4011619253995148,
+      "learning_rate": 6.667742123346603e-05,
+      "loss": 0.4216,
+      "step": 1103
+    },
+    {
+      "epoch": 1.706336939721793,
+      "grad_norm": 0.40592447432225237,
+      "learning_rate": 6.664524013694341e-05,
+      "loss": 0.4153,
+      "step": 1104
+    },
+    {
+      "epoch": 1.7078825347758886,
+      "grad_norm": 0.40454456250038306,
+      "learning_rate": 6.661302800662261e-05,
+      "loss": 0.424,
+      "step": 1105
+    },
+    {
+      "epoch": 1.7094281298299845,
+      "grad_norm": 0.391031986132384,
+      "learning_rate": 6.658078488002128e-05,
+      "loss": 0.416,
+      "step": 1106
+    },
+    {
+      "epoch": 1.7109737248840804,
+      "grad_norm": 0.318555248550208,
+      "learning_rate": 6.654851079469303e-05,
+      "loss": 0.4194,
+      "step": 1107
+    },
+    {
+      "epoch": 1.7125193199381763,
+      "grad_norm": 0.2833193496747006,
+      "learning_rate": 6.651620578822762e-05,
+      "loss": 0.4226,
+      "step": 1108
+    },
+    {
+      "epoch": 1.714064914992272,
+      "grad_norm": 0.4102788785301504,
+      "learning_rate": 6.64838698982508e-05,
+      "loss": 0.4209,
+      "step": 1109
+    },
+    {
+      "epoch": 1.7156105100463679,
+      "grad_norm": 0.43559268625858316,
+      "learning_rate": 6.645150316242426e-05,
+      "loss": 0.414,
+      "step": 1110
+    },
+    {
+      "epoch": 1.7171561051004636,
+      "grad_norm": 0.3400148358964731,
+      "learning_rate": 6.641910561844569e-05,
+      "loss": 0.4091,
+      "step": 1111
+    },
+    {
+      "epoch": 1.7187017001545595,
+      "grad_norm": 0.3106594308436494,
+      "learning_rate": 6.638667730404856e-05,
+      "loss": 0.4236,
+      "step": 1112
+    },
+    {
+      "epoch": 1.7202472952086554,
+      "grad_norm": 0.2581600125356283,
+      "learning_rate": 6.635421825700228e-05,
+      "loss": 0.4238,
+      "step": 1113
+    },
+    {
+      "epoch": 1.7217928902627513,
+      "grad_norm": 0.3077881881891871,
+      "learning_rate": 6.6321728515112e-05,
+      "loss": 0.4231,
+      "step": 1114
+    },
+    {
+      "epoch": 1.723338485316847,
+      "grad_norm": 0.4573839706064513,
+      "learning_rate": 6.62892081162186e-05,
+      "loss": 0.4123,
+      "step": 1115
+    },
+    {
+      "epoch": 1.7248840803709427,
+      "grad_norm": 0.42187828947492784,
+      "learning_rate": 6.625665709819871e-05,
+      "loss": 0.4278,
+      "step": 1116
+    },
+    {
+      "epoch": 1.7264296754250386,
+      "grad_norm": 0.3485554331780731,
+      "learning_rate": 6.622407549896462e-05,
+      "loss": 0.4191,
+      "step": 1117
+    },
+    {
+      "epoch": 1.7279752704791345,
+      "grad_norm": 0.237263508252921,
+      "learning_rate": 6.619146335646422e-05,
+      "loss": 0.4095,
+      "step": 1118
+    },
+    {
+      "epoch": 1.7295208655332304,
+      "grad_norm": 0.3190353533959669,
+      "learning_rate": 6.615882070868096e-05,
+      "loss": 0.4143,
+      "step": 1119
+    },
+    {
+      "epoch": 1.7310664605873263,
+      "grad_norm": 0.4023411179851799,
+      "learning_rate": 6.612614759363386e-05,
+      "loss": 0.4245,
+      "step": 1120
+    },
+    {
+      "epoch": 1.732612055641422,
+      "grad_norm": 0.296489833215759,
+      "learning_rate": 6.609344404937739e-05,
+      "loss": 0.4166,
+      "step": 1121
+    },
+    {
+      "epoch": 1.7341576506955176,
+      "grad_norm": 0.34685746967755765,
+      "learning_rate": 6.606071011400148e-05,
+      "loss": 0.4272,
+      "step": 1122
+    },
+    {
+      "epoch": 1.7357032457496135,
+      "grad_norm": 0.3189024251074944,
+      "learning_rate": 6.602794582563146e-05,
+      "loss": 0.411,
+      "step": 1123
+    },
+    {
+      "epoch": 1.7372488408037094,
+      "grad_norm": 0.34829829224625086,
+      "learning_rate": 6.599515122242799e-05,
+      "loss": 0.4158,
+      "step": 1124
+    },
+    {
+      "epoch": 1.7387944358578054,
+      "grad_norm": 0.3899462123118788,
+      "learning_rate": 6.596232634258706e-05,
+      "loss": 0.4176,
+      "step": 1125
+    },
+    {
+      "epoch": 1.740340030911901,
+      "grad_norm": 0.3589594431918154,
+      "learning_rate": 6.592947122433992e-05,
+      "loss": 0.4153,
+      "step": 1126
+    },
+    {
+      "epoch": 1.741885625965997,
+      "grad_norm": 0.32624874047244823,
+      "learning_rate": 6.5896585905953e-05,
+      "loss": 0.4257,
+      "step": 1127
+    },
+    {
+      "epoch": 1.7434312210200926,
+      "grad_norm": 0.3573946165560419,
+      "learning_rate": 6.586367042572797e-05,
+      "loss": 0.4186,
+      "step": 1128
+    },
+    {
+      "epoch": 1.7449768160741885,
+      "grad_norm": 0.33050528839284254,
+      "learning_rate": 6.583072482200158e-05,
+      "loss": 0.4205,
+      "step": 1129
+    },
+    {
+      "epoch": 1.7465224111282844,
+      "grad_norm": 0.28947624973873576,
+      "learning_rate": 6.579774913314571e-05,
+      "loss": 0.4116,
+      "step": 1130
+    },
+    {
+      "epoch": 1.7480680061823803,
+      "grad_norm": 0.36295085848997277,
+      "learning_rate": 6.576474339756723e-05,
+      "loss": 0.4151,
+      "step": 1131
+    },
+    {
+      "epoch": 1.749613601236476,
+      "grad_norm": 0.3375864893741672,
+      "learning_rate": 6.573170765370805e-05,
+      "loss": 0.4207,
+      "step": 1132
+    },
+    {
+      "epoch": 1.7511591962905717,
+      "grad_norm": 0.2574122901993273,
+      "learning_rate": 6.569864194004498e-05,
+      "loss": 0.4147,
+      "step": 1133
+    },
+    {
+      "epoch": 1.7527047913446676,
+      "grad_norm": 0.23678817072974698,
+      "learning_rate": 6.566554629508976e-05,
+      "loss": 0.425,
+      "step": 1134
+    },
+    {
+      "epoch": 1.7542503863987635,
+      "grad_norm": 0.29487274917096445,
+      "learning_rate": 6.563242075738902e-05,
+      "loss": 0.4144,
+      "step": 1135
+    },
+    {
+      "epoch": 1.7557959814528594,
+      "grad_norm": 0.37107506017243763,
+      "learning_rate": 6.559926536552417e-05,
+      "loss": 0.4229,
+      "step": 1136
+    },
+    {
+      "epoch": 1.7573415765069553,
+      "grad_norm": 0.3231852598757117,
+      "learning_rate": 6.556608015811142e-05,
+      "loss": 0.4154,
+      "step": 1137
+    },
+    {
+      "epoch": 1.758887171561051,
+      "grad_norm": 0.34463743975189143,
+      "learning_rate": 6.553286517380166e-05,
+      "loss": 0.4136,
+      "step": 1138
+    },
+    {
+      "epoch": 1.7604327666151467,
+      "grad_norm": 0.241224707262446,
+      "learning_rate": 6.54996204512805e-05,
+      "loss": 0.4152,
+      "step": 1139
+    },
+    {
+      "epoch": 1.7619783616692426,
+      "grad_norm": 0.27379722315518734,
+      "learning_rate": 6.546634602926821e-05,
+      "loss": 0.4247,
+      "step": 1140
+    },
+    {
+      "epoch": 1.7635239567233385,
+      "grad_norm": 0.3486256104244915,
+      "learning_rate": 6.543304194651956e-05,
+      "loss": 0.4129,
+      "step": 1141
+    },
+    {
+      "epoch": 1.7650695517774344,
+      "grad_norm": 0.2647142219521478,
+      "learning_rate": 6.539970824182398e-05,
+      "loss": 0.4174,
+      "step": 1142
+    },
+    {
+      "epoch": 1.76661514683153,
+      "grad_norm": 0.26381420485930107,
+      "learning_rate": 6.536634495400531e-05,
+      "loss": 0.4086,
+      "step": 1143
+    },
+    {
+      "epoch": 1.768160741885626,
+      "grad_norm": 0.3078250326202281,
+      "learning_rate": 6.533295212192189e-05,
+      "loss": 0.4258,
+      "step": 1144
+    },
+    {
+      "epoch": 1.7697063369397217,
+      "grad_norm": 0.37360398510483267,
+      "learning_rate": 6.529952978446645e-05,
+      "loss": 0.4109,
+      "step": 1145
+    },
+    {
+      "epoch": 1.7712519319938176,
+      "grad_norm": 0.44167793126435473,
+      "learning_rate": 6.526607798056612e-05,
+      "loss": 0.4211,
+      "step": 1146
+    },
+    {
+      "epoch": 1.7727975270479135,
+      "grad_norm": 0.454035260878904,
+      "learning_rate": 6.52325967491823e-05,
+      "loss": 0.4215,
+      "step": 1147
+    },
+    {
+      "epoch": 1.7743431221020094,
+      "grad_norm": 0.4246007986639795,
+      "learning_rate": 6.519908612931071e-05,
+      "loss": 0.4097,
+      "step": 1148
+    },
+    {
+      "epoch": 1.775888717156105,
+      "grad_norm": 0.5040633836956859,
+      "learning_rate": 6.516554615998126e-05,
+      "loss": 0.4154,
+      "step": 1149
+    },
+    {
+      "epoch": 1.7774343122102008,
+      "grad_norm": 0.5323469886608114,
+      "learning_rate": 6.513197688025808e-05,
+      "loss": 0.4101,
+      "step": 1150
+    },
+    {
+      "epoch": 1.7789799072642967,
+      "grad_norm": 0.540753440436067,
+      "learning_rate": 6.509837832923939e-05,
+      "loss": 0.4176,
+      "step": 1151
+    },
+    {
+      "epoch": 1.7805255023183926,
+      "grad_norm": 0.5436390782696222,
+      "learning_rate": 6.506475054605757e-05,
+      "loss": 0.4108,
+      "step": 1152
+    },
+    {
+      "epoch": 1.7820710973724885,
+      "grad_norm": 0.5167138471794511,
+      "learning_rate": 6.503109356987899e-05,
+      "loss": 0.4093,
+      "step": 1153
+    },
+    {
+      "epoch": 1.7836166924265844,
+      "grad_norm": 0.5670069583521865,
+      "learning_rate": 6.499740743990404e-05,
+      "loss": 0.4169,
+      "step": 1154
+    },
+    {
+      "epoch": 1.78516228748068,
+      "grad_norm": 0.6272309850462795,
+      "learning_rate": 6.496369219536706e-05,
+      "loss": 0.4182,
+      "step": 1155
+    },
+    {
+      "epoch": 1.7867078825347757,
+      "grad_norm": 0.5321810006856598,
+      "learning_rate": 6.492994787553632e-05,
+      "loss": 0.4153,
+      "step": 1156
+    },
+    {
+      "epoch": 1.7882534775888717,
+      "grad_norm": 0.3947246482796357,
+      "learning_rate": 6.489617451971392e-05,
+      "loss": 0.4185,
+      "step": 1157
+    },
+    {
+      "epoch": 1.7897990726429676,
+      "grad_norm": 0.2919535338412543,
+      "learning_rate": 6.486237216723581e-05,
+      "loss": 0.4234,
+      "step": 1158
+    },
+    {
+      "epoch": 1.7913446676970635,
+      "grad_norm": 0.31757558124492435,
+      "learning_rate": 6.482854085747171e-05,
+      "loss": 0.4107,
+      "step": 1159
+    },
+    {
+      "epoch": 1.7928902627511591,
+      "grad_norm": 0.41985931089669054,
+      "learning_rate": 6.479468062982505e-05,
+      "loss": 0.4116,
+      "step": 1160
+    },
+    {
+      "epoch": 1.794435857805255,
+      "grad_norm": 0.43819798291205037,
+      "learning_rate": 6.476079152373292e-05,
+      "loss": 0.4077,
+      "step": 1161
+    },
+    {
+      "epoch": 1.7959814528593507,
+      "grad_norm": 0.4394654469509475,
+      "learning_rate": 6.472687357866611e-05,
+      "loss": 0.4076,
+      "step": 1162
+    },
+    {
+      "epoch": 1.7975270479134466,
+      "grad_norm": 0.3421834960647247,
+      "learning_rate": 6.469292683412895e-05,
+      "loss": 0.4157,
+      "step": 1163
+    },
+    {
+      "epoch": 1.7990726429675425,
+      "grad_norm": 0.261891084008789,
+      "learning_rate": 6.465895132965931e-05,
+      "loss": 0.4156,
+      "step": 1164
+    },
+    {
+      "epoch": 1.8006182380216385,
+      "grad_norm": 0.2566986603550055,
+      "learning_rate": 6.46249471048286e-05,
+      "loss": 0.4144,
+      "step": 1165
+    },
+    {
+      "epoch": 1.8021638330757341,
+      "grad_norm": 0.3715925083121105,
+      "learning_rate": 6.459091419924162e-05,
+      "loss": 0.4127,
+      "step": 1166
+    },
+    {
+      "epoch": 1.80370942812983,
+      "grad_norm": 0.44997059607774637,
+      "learning_rate": 6.455685265253663e-05,
+      "loss": 0.4108,
+      "step": 1167
+    },
+    {
+      "epoch": 1.8052550231839257,
+      "grad_norm": 0.39558533003170193,
+      "learning_rate": 6.452276250438521e-05,
+      "loss": 0.4082,
+      "step": 1168
+    },
+    {
+      "epoch": 1.8068006182380216,
+      "grad_norm": 0.3205412972742549,
+      "learning_rate": 6.448864379449227e-05,
+      "loss": 0.4145,
+      "step": 1169
+    },
+    {
+      "epoch": 1.8083462132921175,
+      "grad_norm": 0.33560562911213015,
+      "learning_rate": 6.445449656259598e-05,
+      "loss": 0.4077,
+      "step": 1170
+    },
+    {
+      "epoch": 1.8098918083462134,
+      "grad_norm": 0.3408721236384103,
+      "learning_rate": 6.442032084846775e-05,
+      "loss": 0.4204,
+      "step": 1171
+    },
+    {
+      "epoch": 1.8114374034003091,
+      "grad_norm": 0.32642351132837616,
+      "learning_rate": 6.438611669191212e-05,
+      "loss": 0.4099,
+      "step": 1172
+    },
+    {
+      "epoch": 1.8129829984544048,
+      "grad_norm": 0.3008497794435818,
+      "learning_rate": 6.435188413276679e-05,
+      "loss": 0.4175,
+      "step": 1173
+    },
+    {
+      "epoch": 1.8145285935085007,
+      "grad_norm": 0.3161866790828201,
+      "learning_rate": 6.431762321090255e-05,
+      "loss": 0.426,
+      "step": 1174
+    },
+    {
+      "epoch": 1.8160741885625966,
+      "grad_norm": 0.34293499287346135,
+      "learning_rate": 6.42833339662232e-05,
+      "loss": 0.4227,
+      "step": 1175
+    },
+    {
+      "epoch": 1.8176197836166925,
+      "grad_norm": 0.32808353325790235,
+      "learning_rate": 6.424901643866553e-05,
+      "loss": 0.4163,
+      "step": 1176
+    },
+    {
+      "epoch": 1.8191653786707882,
+      "grad_norm": 0.3659363750905557,
+      "learning_rate": 6.421467066819927e-05,
+      "loss": 0.4181,
+      "step": 1177
+    },
+    {
+      "epoch": 1.820710973724884,
+      "grad_norm": 0.4687589192970821,
+      "learning_rate": 6.418029669482707e-05,
+      "loss": 0.4099,
+      "step": 1178
+    },
+    {
+      "epoch": 1.8222565687789798,
+      "grad_norm": 0.46175089574388317,
+      "learning_rate": 6.414589455858438e-05,
+      "loss": 0.4068,
+      "step": 1179
+    },
+    {
+      "epoch": 1.8238021638330757,
+      "grad_norm": 0.47340277937673997,
+      "learning_rate": 6.411146429953951e-05,
+      "loss": 0.4033,
+      "step": 1180
+    },
+    {
+      "epoch": 1.8253477588871716,
+      "grad_norm": 0.5999195250605643,
+      "learning_rate": 6.407700595779348e-05,
+      "loss": 0.4199,
+      "step": 1181
+    },
+    {
+      "epoch": 1.8268933539412675,
+      "grad_norm": 0.6650778602296572,
+      "learning_rate": 6.404251957348005e-05,
+      "loss": 0.4128,
+      "step": 1182
+    },
+    {
+      "epoch": 1.8284389489953632,
+      "grad_norm": 0.6605952338300355,
+      "learning_rate": 6.400800518676558e-05,
+      "loss": 0.4236,
+      "step": 1183
+    },
+    {
+      "epoch": 1.829984544049459,
+      "grad_norm": 0.6526342849663198,
+      "learning_rate": 6.397346283784915e-05,
+      "loss": 0.412,
+      "step": 1184
+    },
+    {
+      "epoch": 1.8315301391035548,
+      "grad_norm": 0.595465682627906,
+      "learning_rate": 6.393889256696231e-05,
+      "loss": 0.4133,
+      "step": 1185
+    },
+    {
+      "epoch": 1.8330757341576507,
+      "grad_norm": 0.4792886863134555,
+      "learning_rate": 6.390429441436915e-05,
+      "loss": 0.4013,
+      "step": 1186
+    },
+    {
+      "epoch": 1.8346213292117466,
+      "grad_norm": 0.35199181474614083,
+      "learning_rate": 6.38696684203663e-05,
+      "loss": 0.4032,
+      "step": 1187
+    },
+    {
+      "epoch": 1.8361669242658425,
+      "grad_norm": 0.2625679966109965,
+      "learning_rate": 6.383501462528273e-05,
+      "loss": 0.4114,
+      "step": 1188
+    },
+    {
+      "epoch": 1.8377125193199382,
+      "grad_norm": 0.20920678505260878,
+      "learning_rate": 6.380033306947985e-05,
+      "loss": 0.4133,
+      "step": 1189
+    },
+    {
+      "epoch": 1.8392581143740339,
+      "grad_norm": 0.25469585355967583,
+      "learning_rate": 6.376562379335136e-05,
+      "loss": 0.4138,
+      "step": 1190
+    },
+    {
+      "epoch": 1.8408037094281298,
+      "grad_norm": 0.3612273794583833,
+      "learning_rate": 6.373088683732329e-05,
+      "loss": 0.4182,
+      "step": 1191
+    },
+    {
+      "epoch": 1.8423493044822257,
+      "grad_norm": 0.39257267944161245,
+      "learning_rate": 6.369612224185387e-05,
+      "loss": 0.4093,
+      "step": 1192
+    },
+    {
+      "epoch": 1.8438948995363216,
+      "grad_norm": 0.42380522893852823,
+      "learning_rate": 6.366133004743354e-05,
+      "loss": 0.4129,
+      "step": 1193
+    },
+    {
+      "epoch": 1.8454404945904173,
+      "grad_norm": 0.373374444311683,
+      "learning_rate": 6.362651029458488e-05,
+      "loss": 0.4111,
+      "step": 1194
+    },
+    {
+      "epoch": 1.8469860896445132,
+      "grad_norm": 0.2804690905657256,
+      "learning_rate": 6.35916630238626e-05,
+      "loss": 0.4162,
+      "step": 1195
+    },
+    {
+      "epoch": 1.8485316846986088,
+      "grad_norm": 0.22179321091728624,
+      "learning_rate": 6.355678827585339e-05,
+      "loss": 0.4196,
+      "step": 1196
+    },
+    {
+      "epoch": 1.8500772797527048,
+      "grad_norm": 0.25619803053508716,
+      "learning_rate": 6.352188609117601e-05,
+      "loss": 0.4165,
+      "step": 1197
+    },
+    {
+      "epoch": 1.8516228748068007,
+      "grad_norm": 0.29265947672612824,
+      "learning_rate": 6.348695651048113e-05,
+      "loss": 0.412,
+      "step": 1198
+    },
+    {
+      "epoch": 1.8531684698608966,
+      "grad_norm": 0.3248585262131145,
+      "learning_rate": 6.345199957445134e-05,
+      "loss": 0.4044,
+      "step": 1199
+    },
+    {
+      "epoch": 1.8547140649149922,
+      "grad_norm": 0.33661124084441724,
+      "learning_rate": 6.341701532380111e-05,
+      "loss": 0.4213,
+      "step": 1200
+    },
+    {
+      "epoch": 1.8562596599690881,
+      "grad_norm": 0.37602417264309734,
+      "learning_rate": 6.338200379927673e-05,
+      "loss": 0.4123,
+      "step": 1201
+    },
+    {
+      "epoch": 1.8578052550231838,
+      "grad_norm": 0.4273071512644445,
+      "learning_rate": 6.33469650416562e-05,
+      "loss": 0.417,
+      "step": 1202
+    },
+    {
+      "epoch": 1.8593508500772797,
+      "grad_norm": 0.42079618075158465,
+      "learning_rate": 6.33118990917493e-05,
+      "loss": 0.4178,
+      "step": 1203
+    },
+    {
+      "epoch": 1.8608964451313756,
+      "grad_norm": 0.40759642036475224,
+      "learning_rate": 6.327680599039745e-05,
+      "loss": 0.4094,
+      "step": 1204
+    },
+    {
+      "epoch": 1.8624420401854715,
+      "grad_norm": 0.39745878772659676,
+      "learning_rate": 6.324168577847371e-05,
+      "loss": 0.4079,
+      "step": 1205
+    },
+    {
+      "epoch": 1.8639876352395672,
+      "grad_norm": 0.3415056835125822,
+      "learning_rate": 6.320653849688268e-05,
+      "loss": 0.4157,
+      "step": 1206
+    },
+    {
+      "epoch": 1.865533230293663,
+      "grad_norm": 0.24947747232526643,
+      "learning_rate": 6.317136418656054e-05,
+      "loss": 0.4028,
+      "step": 1207
+    },
+    {
+      "epoch": 1.8670788253477588,
+      "grad_norm": 0.3005682819226224,
+      "learning_rate": 6.31361628884749e-05,
+      "loss": 0.4221,
+      "step": 1208
+    },
+    {
+      "epoch": 1.8686244204018547,
+      "grad_norm": 0.40116056221954854,
+      "learning_rate": 6.310093464362485e-05,
+      "loss": 0.4091,
+      "step": 1209
+    },
+    {
+      "epoch": 1.8701700154559506,
+      "grad_norm": 0.4979343197776133,
+      "learning_rate": 6.306567949304085e-05,
+      "loss": 0.4143,
+      "step": 1210
+    },
+    {
+      "epoch": 1.8717156105100463,
+      "grad_norm": 0.5484171918207302,
+      "learning_rate": 6.303039747778464e-05,
+      "loss": 0.42,
+      "step": 1211
+    },
+    {
+      "epoch": 1.8732612055641422,
+      "grad_norm": 0.4410930297414067,
+      "learning_rate": 6.299508863894933e-05,
+      "loss": 0.4075,
+      "step": 1212
+    },
+    {
+      "epoch": 1.874806800618238,
+      "grad_norm": 0.4015790888559192,
+      "learning_rate": 6.295975301765925e-05,
+      "loss": 0.4144,
+      "step": 1213
+    },
+    {
+      "epoch": 1.8763523956723338,
+      "grad_norm": 0.4815391033114978,
+      "learning_rate": 6.29243906550699e-05,
+      "loss": 0.4105,
+      "step": 1214
+    },
+    {
+      "epoch": 1.8778979907264297,
+      "grad_norm": 0.5005456093567422,
+      "learning_rate": 6.288900159236793e-05,
+      "loss": 0.4103,
+      "step": 1215
+    },
+    {
+      "epoch": 1.8794435857805256,
+      "grad_norm": 0.5730406470111321,
+      "learning_rate": 6.285358587077112e-05,
+      "loss": 0.4181,
+      "step": 1216
+    },
+    {
+      "epoch": 1.8809891808346213,
+      "grad_norm": 0.5962056972704061,
+      "learning_rate": 6.281814353152825e-05,
+      "loss": 0.4178,
+      "step": 1217
+    },
+    {
+      "epoch": 1.8825347758887172,
+      "grad_norm": 0.42427410875375193,
+      "learning_rate": 6.278267461591915e-05,
+      "loss": 0.4128,
+      "step": 1218
+    },
+    {
+      "epoch": 1.8840803709428129,
+      "grad_norm": 0.3196237136374311,
+      "learning_rate": 6.274717916525457e-05,
+      "loss": 0.4141,
+      "step": 1219
+    },
+    {
+      "epoch": 1.8856259659969088,
+      "grad_norm": 0.3488472721495409,
+      "learning_rate": 6.271165722087621e-05,
+      "loss": 0.4155,
+      "step": 1220
+    },
+    {
+      "epoch": 1.8871715610510047,
+      "grad_norm": 0.46154481832376043,
+      "learning_rate": 6.267610882415654e-05,
+      "loss": 0.4129,
+      "step": 1221
+    },
+    {
+      "epoch": 1.8887171561051006,
+      "grad_norm": 0.5639974047085498,
+      "learning_rate": 6.264053401649892e-05,
+      "loss": 0.4075,
+      "step": 1222
+    },
+    {
+      "epoch": 1.8902627511591963,
+      "grad_norm": 0.5253893444511016,
+      "learning_rate": 6.260493283933744e-05,
+      "loss": 0.4147,
+      "step": 1223
+    },
+    {
+      "epoch": 1.891808346213292,
+      "grad_norm": 0.4951660291809643,
+      "learning_rate": 6.256930533413692e-05,
+      "loss": 0.4233,
+      "step": 1224
+    },
+    {
+      "epoch": 1.8933539412673879,
+      "grad_norm": 0.591961666560772,
+      "learning_rate": 6.253365154239282e-05,
+      "loss": 0.4232,
+      "step": 1225
+    },
+    {
+      "epoch": 1.8948995363214838,
+      "grad_norm": 0.5675281152594379,
+      "learning_rate": 6.249797150563124e-05,
+      "loss": 0.4129,
+      "step": 1226
+    },
+    {
+      "epoch": 1.8964451313755797,
+      "grad_norm": 0.5417143579552858,
+      "learning_rate": 6.246226526540881e-05,
+      "loss": 0.4273,
+      "step": 1227
+    },
+    {
+      "epoch": 1.8979907264296756,
+      "grad_norm": 0.5979033416426762,
+      "learning_rate": 6.242653286331273e-05,
+      "loss": 0.4311,
+      "step": 1228
+    },
+    {
+      "epoch": 1.8995363214837713,
+      "grad_norm": 0.5412056674221714,
+      "learning_rate": 6.239077434096063e-05,
+      "loss": 0.4107,
+      "step": 1229
+    },
+    {
+      "epoch": 1.901081916537867,
+      "grad_norm": 0.34544360856265593,
+      "learning_rate": 6.235498974000057e-05,
+      "loss": 0.4103,
+      "step": 1230
+    },
+    {
+      "epoch": 1.9026275115919629,
+      "grad_norm": 0.3093068702664324,
+      "learning_rate": 6.231917910211104e-05,
+      "loss": 0.4187,
+      "step": 1231
+    },
+    {
+      "epoch": 1.9041731066460588,
+      "grad_norm": 0.4606991022107261,
+      "learning_rate": 6.228334246900075e-05,
+      "loss": 0.4165,
+      "step": 1232
+    },
+    {
+      "epoch": 1.9057187017001547,
+      "grad_norm": 0.5144169510223493,
+      "learning_rate": 6.224747988240877e-05,
+      "loss": 0.4062,
+      "step": 1233
+    },
+    {
+      "epoch": 1.9072642967542504,
+      "grad_norm": 0.5291361309953686,
+      "learning_rate": 6.221159138410436e-05,
+      "loss": 0.4127,
+      "step": 1234
+    },
+    {
+      "epoch": 1.9088098918083463,
+      "grad_norm": 0.5486330265411696,
+      "learning_rate": 6.217567701588699e-05,
+      "loss": 0.41,
+      "step": 1235
+    },
+    {
+      "epoch": 1.910355486862442,
+      "grad_norm": 0.5042173545118099,
+      "learning_rate": 6.213973681958622e-05,
+      "loss": 0.4079,
+      "step": 1236
+    },
+    {
+      "epoch": 1.9119010819165378,
+      "grad_norm": 0.4301677403077317,
+      "learning_rate": 6.210377083706173e-05,
+      "loss": 0.4187,
+      "step": 1237
+    },
+    {
+      "epoch": 1.9134466769706338,
+      "grad_norm": 0.3221844288424395,
+      "learning_rate": 6.206777911020319e-05,
+      "loss": 0.417,
+      "step": 1238
+    },
+    {
+      "epoch": 1.9149922720247297,
+      "grad_norm": 0.35860210442670204,
+      "learning_rate": 6.203176168093033e-05,
+      "loss": 0.4071,
+      "step": 1239
+    },
+    {
+      "epoch": 1.9165378670788253,
+      "grad_norm": 0.455646522032718,
+      "learning_rate": 6.199571859119273e-05,
+      "loss": 0.413,
+      "step": 1240
+    },
+    {
+      "epoch": 1.918083462132921,
+      "grad_norm": 0.5114258237030155,
+      "learning_rate": 6.19596498829699e-05,
+      "loss": 0.4103,
+      "step": 1241
+    },
+    {
+      "epoch": 1.919629057187017,
+      "grad_norm": 0.5018901743826315,
+      "learning_rate": 6.192355559827119e-05,
+      "loss": 0.4138,
+      "step": 1242
+    },
+    {
+      "epoch": 1.9211746522411128,
+      "grad_norm": 0.3770510003359604,
+      "learning_rate": 6.188743577913573e-05,
+      "loss": 0.4088,
+      "step": 1243
+    },
+    {
+      "epoch": 1.9227202472952087,
+      "grad_norm": 0.3075133758449342,
+      "learning_rate": 6.185129046763238e-05,
+      "loss": 0.412,
+      "step": 1244
+    },
+    {
+      "epoch": 1.9242658423493046,
+      "grad_norm": 0.31888203899725603,
+      "learning_rate": 6.181511970585972e-05,
+      "loss": 0.4169,
+      "step": 1245
+    },
+    {
+      "epoch": 1.9258114374034003,
+      "grad_norm": 0.3680308830373179,
+      "learning_rate": 6.177892353594594e-05,
+      "loss": 0.4126,
+      "step": 1246
+    },
+    {
+      "epoch": 1.927357032457496,
+      "grad_norm": 0.3419862027786341,
+      "learning_rate": 6.174270200004885e-05,
+      "loss": 0.4106,
+      "step": 1247
+    },
+    {
+      "epoch": 1.928902627511592,
+      "grad_norm": 0.4074337933081065,
+      "learning_rate": 6.170645514035578e-05,
+      "loss": 0.4114,
+      "step": 1248
+    },
+    {
+      "epoch": 1.9304482225656878,
+      "grad_norm": 0.402449633126962,
+      "learning_rate": 6.167018299908358e-05,
+      "loss": 0.4231,
+      "step": 1249
+    },
+    {
+      "epoch": 1.9319938176197837,
+      "grad_norm": 0.45296020240879586,
+      "learning_rate": 6.163388561847849e-05,
+      "loss": 0.4144,
+      "step": 1250
+    },
+    {
+      "epoch": 1.9335394126738794,
+      "grad_norm": 0.3919980871130073,
+      "learning_rate": 6.159756304081625e-05,
+      "loss": 0.4154,
+      "step": 1251
+    },
+    {
+      "epoch": 1.9350850077279753,
+      "grad_norm": 0.3488830984928472,
+      "learning_rate": 6.156121530840183e-05,
+      "loss": 0.4104,
+      "step": 1252
+    },
+    {
+      "epoch": 1.936630602782071,
+      "grad_norm": 0.33868722953018976,
+      "learning_rate": 6.15248424635696e-05,
+      "loss": 0.4142,
+      "step": 1253
+    },
+    {
+      "epoch": 1.938176197836167,
+      "grad_norm": 0.2855567516301163,
+      "learning_rate": 6.14884445486831e-05,
+      "loss": 0.4172,
+      "step": 1254
+    },
+    {
+      "epoch": 1.9397217928902628,
+      "grad_norm": 0.2922986903670274,
+      "learning_rate": 6.145202160613509e-05,
+      "loss": 0.4105,
+      "step": 1255
+    },
+    {
+      "epoch": 1.9412673879443587,
+      "grad_norm": 0.29727295624694355,
+      "learning_rate": 6.141557367834752e-05,
+      "loss": 0.4135,
+      "step": 1256
+    },
+    {
+      "epoch": 1.9428129829984544,
+      "grad_norm": 0.33154582980383096,
+      "learning_rate": 6.13791008077714e-05,
+      "loss": 0.4168,
+      "step": 1257
+    },
+    {
+      "epoch": 1.94435857805255,
+      "grad_norm": 0.29921155830945473,
+      "learning_rate": 6.13426030368868e-05,
+      "loss": 0.401,
+      "step": 1258
+    },
+    {
+      "epoch": 1.945904173106646,
+      "grad_norm": 0.3000493708088317,
+      "learning_rate": 6.130608040820277e-05,
+      "loss": 0.421,
+      "step": 1259
+    },
+    {
+      "epoch": 1.947449768160742,
+      "grad_norm": 0.31902396993102805,
+      "learning_rate": 6.126953296425738e-05,
+      "loss": 0.408,
+      "step": 1260
+    },
+    {
+      "epoch": 1.9489953632148378,
+      "grad_norm": 0.33261146579060874,
+      "learning_rate": 6.123296074761753e-05,
+      "loss": 0.4095,
+      "step": 1261
+    },
+    {
+      "epoch": 1.9505409582689337,
+      "grad_norm": 0.36866099521520945,
+      "learning_rate": 6.119636380087899e-05,
+      "loss": 0.4178,
+      "step": 1262
+    },
+    {
+      "epoch": 1.9520865533230294,
+      "grad_norm": 0.341692553925293,
+      "learning_rate": 6.115974216666638e-05,
+      "loss": 0.4156,
+      "step": 1263
+    },
+    {
+      "epoch": 1.953632148377125,
+      "grad_norm": 0.32121154073975516,
+      "learning_rate": 6.1123095887633e-05,
+      "loss": 0.4094,
+      "step": 1264
+    },
+    {
+      "epoch": 1.955177743431221,
+      "grad_norm": 0.41034270168967524,
+      "learning_rate": 6.108642500646091e-05,
+      "loss": 0.4196,
+      "step": 1265
+    },
+    {
+      "epoch": 1.9567233384853169,
+      "grad_norm": 0.41868186295386195,
+      "learning_rate": 6.10497295658608e-05,
+      "loss": 0.407,
+      "step": 1266
+    },
+    {
+      "epoch": 1.9582689335394128,
+      "grad_norm": 0.3773719691008005,
+      "learning_rate": 6.1013009608571954e-05,
+      "loss": 0.4121,
+      "step": 1267
+    },
+    {
+      "epoch": 1.9598145285935085,
+      "grad_norm": 0.4011699259879193,
+      "learning_rate": 6.097626517736226e-05,
+      "loss": 0.412,
+      "step": 1268
+    },
+    {
+      "epoch": 1.9613601236476044,
+      "grad_norm": 0.43043830359425667,
+      "learning_rate": 6.0939496315028044e-05,
+      "loss": 0.4108,
+      "step": 1269
+    },
+    {
+      "epoch": 1.9629057187017,
+      "grad_norm": 0.4408474459144848,
+      "learning_rate": 6.090270306439415e-05,
+      "loss": 0.4158,
+      "step": 1270
+    },
+    {
+      "epoch": 1.964451313755796,
+      "grad_norm": 0.5261256456743819,
+      "learning_rate": 6.086588546831378e-05,
+      "loss": 0.4226,
+      "step": 1271
+    },
+    {
+      "epoch": 1.9659969088098919,
+      "grad_norm": 0.5996786938777356,
+      "learning_rate": 6.082904356966849e-05,
+      "loss": 0.4084,
+      "step": 1272
+    },
+    {
+      "epoch": 1.9675425038639878,
+      "grad_norm": 0.6343211777203427,
+      "learning_rate": 6.07921774113682e-05,
+      "loss": 0.4164,
+      "step": 1273
+    },
+    {
+      "epoch": 1.9690880989180835,
+      "grad_norm": 0.6506414593550145,
+      "learning_rate": 6.075528703635102e-05,
+      "loss": 0.4154,
+      "step": 1274
+    },
+    {
+      "epoch": 1.9706336939721791,
+      "grad_norm": 0.5323031068469932,
+      "learning_rate": 6.071837248758329e-05,
+      "loss": 0.4165,
+      "step": 1275
+    },
+    {
+      "epoch": 1.972179289026275,
+      "grad_norm": 0.4033522028561435,
+      "learning_rate": 6.06814338080595e-05,
+      "loss": 0.4153,
+      "step": 1276
+    },
+    {
+      "epoch": 1.973724884080371,
+      "grad_norm": 0.35960287534280405,
+      "learning_rate": 6.064447104080229e-05,
+      "loss": 0.4101,
+      "step": 1277
+    },
+    {
+      "epoch": 1.9752704791344669,
+      "grad_norm": 0.3645860122710083,
+      "learning_rate": 6.060748422886225e-05,
+      "loss": 0.411,
+      "step": 1278
+    },
+    {
+      "epoch": 1.9768160741885628,
+      "grad_norm": 0.3184631350197533,
+      "learning_rate": 6.05704734153181e-05,
+      "loss": 0.4145,
+      "step": 1279
+    },
+    {
+      "epoch": 1.9783616692426584,
+      "grad_norm": 0.26820918607939453,
+      "learning_rate": 6.053343864327643e-05,
+      "loss": 0.4105,
+      "step": 1280
+    },
+    {
+      "epoch": 1.9799072642967541,
+      "grad_norm": 0.2548959130980497,
+      "learning_rate": 6.0496379955871756e-05,
+      "loss": 0.4101,
+      "step": 1281
+    },
+    {
+      "epoch": 1.98145285935085,
+      "grad_norm": 0.3122132325464997,
+      "learning_rate": 6.045929739626646e-05,
+      "loss": 0.4194,
+      "step": 1282
+    },
+    {
+      "epoch": 1.982998454404946,
+      "grad_norm": 0.3760475329731319,
+      "learning_rate": 6.042219100765071e-05,
+      "loss": 0.408,
+      "step": 1283
+    },
+    {
+      "epoch": 1.9845440494590418,
+      "grad_norm": 0.39407874614851257,
+      "learning_rate": 6.0385060833242454e-05,
+      "loss": 0.4126,
+      "step": 1284
+    },
+    {
+      "epoch": 1.9860896445131375,
+      "grad_norm": 0.4158714884606941,
+      "learning_rate": 6.0347906916287325e-05,
+      "loss": 0.4102,
+      "step": 1285
+    },
+    {
+      "epoch": 1.9876352395672334,
+      "grad_norm": 0.39739963706387743,
+      "learning_rate": 6.0310729300058594e-05,
+      "loss": 0.4074,
+      "step": 1286
+    },
+    {
+      "epoch": 1.989180834621329,
+      "grad_norm": 0.40956072410450617,
+      "learning_rate": 6.0273528027857174e-05,
+      "loss": 0.4035,
+      "step": 1287
+    },
+    {
+      "epoch": 1.990726429675425,
+      "grad_norm": 0.4235509784543519,
+      "learning_rate": 6.02363031430115e-05,
+      "loss": 0.4063,
+      "step": 1288
+    },
+    {
+      "epoch": 1.992272024729521,
+      "grad_norm": 0.35569231358204384,
+      "learning_rate": 6.01990546888775e-05,
+      "loss": 0.4088,
+      "step": 1289
+    },
+    {
+      "epoch": 1.9938176197836168,
+      "grad_norm": 0.3277558087857754,
+      "learning_rate": 6.016178270883859e-05,
+      "loss": 0.4124,
+      "step": 1290
+    },
+    {
+      "epoch": 1.9953632148377125,
+      "grad_norm": 0.39938016792163783,
+      "learning_rate": 6.012448724630557e-05,
+      "loss": 0.4065,
+      "step": 1291
+    },
+    {
+      "epoch": 1.9969088098918082,
+      "grad_norm": 0.5074637792535728,
+      "learning_rate": 6.008716834471657e-05,
+      "loss": 0.4113,
+      "step": 1292
+    },
+    {
+      "epoch": 1.998454404945904,
+      "grad_norm": 0.5378561174080769,
+      "learning_rate": 6.004982604753705e-05,
+      "loss": 0.406,
+      "step": 1293
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.5512949661008176,
+      "learning_rate": 6.001246039825969e-05,
+      "loss": 0.3932,
+      "step": 1294
+    },
+    {
+      "epoch": 2.001545595054096,
+      "grad_norm": 0.48164343529296594,
+      "learning_rate": 5.99750714404044e-05,
+      "loss": 0.3942,
+      "step": 1295
+    },
+    {
+      "epoch": 2.003091190108192,
+      "grad_norm": 0.4173541891289462,
+      "learning_rate": 5.993765921751822e-05,
+      "loss": 0.3933,
+      "step": 1296
+    },
+    {
+      "epoch": 2.0046367851622873,
+      "grad_norm": 0.44874736906998797,
+      "learning_rate": 5.990022377317527e-05,
+      "loss": 0.399,
+      "step": 1297
+    },
+    {
+      "epoch": 2.006182380216383,
+      "grad_norm": 0.48852803899547276,
+      "learning_rate": 5.986276515097676e-05,
+      "loss": 0.3895,
+      "step": 1298
+    },
+    {
+      "epoch": 2.007727975270479,
+      "grad_norm": 0.47210391121002704,
+      "learning_rate": 5.982528339455086e-05,
+      "loss": 0.3965,
+      "step": 1299
+    },
+    {
+      "epoch": 2.009273570324575,
+      "grad_norm": 0.4071856200785768,
+      "learning_rate": 5.978777854755268e-05,
+      "loss": 0.3877,
+      "step": 1300
+    },
+    {
+      "epoch": 2.010819165378671,
+      "grad_norm": 0.39801788539688204,
+      "learning_rate": 5.975025065366425e-05,
+      "loss": 0.3901,
+      "step": 1301
+    },
+    {
+      "epoch": 2.012364760432767,
+      "grad_norm": 0.44597876167965794,
+      "learning_rate": 5.971269975659444e-05,
+      "loss": 0.389,
+      "step": 1302
+    },
+    {
+      "epoch": 2.0139103554868623,
+      "grad_norm": 0.39663282709561176,
+      "learning_rate": 5.9675125900078876e-05,
+      "loss": 0.3886,
+      "step": 1303
+    },
+    {
+      "epoch": 2.015455950540958,
+      "grad_norm": 0.3271883215905907,
+      "learning_rate": 5.963752912787998e-05,
+      "loss": 0.3952,
+      "step": 1304
+    },
+    {
+      "epoch": 2.017001545595054,
+      "grad_norm": 0.3443570775972986,
+      "learning_rate": 5.959990948378682e-05,
+      "loss": 0.3916,
+      "step": 1305
+    },
+    {
+      "epoch": 2.01854714064915,
+      "grad_norm": 0.43047965620621526,
+      "learning_rate": 5.9562267011615106e-05,
+      "loss": 0.3958,
+      "step": 1306
+    },
+    {
+      "epoch": 2.020092735703246,
+      "grad_norm": 0.48135344420690257,
+      "learning_rate": 5.952460175520719e-05,
+      "loss": 0.403,
+      "step": 1307
+    },
+    {
+      "epoch": 2.021638330757342,
+      "grad_norm": 0.4731700325734638,
+      "learning_rate": 5.948691375843187e-05,
+      "loss": 0.39,
+      "step": 1308
+    },
+    {
+      "epoch": 2.0231839258114372,
+      "grad_norm": 0.3648430550938961,
+      "learning_rate": 5.94492030651845e-05,
+      "loss": 0.3952,
+      "step": 1309
+    },
+    {
+      "epoch": 2.024729520865533,
+      "grad_norm": 0.36491524034826517,
+      "learning_rate": 5.9411469719386865e-05,
+      "loss": 0.387,
+      "step": 1310
+    },
+    {
+      "epoch": 2.026275115919629,
+      "grad_norm": 0.4363293882642952,
+      "learning_rate": 5.937371376498707e-05,
+      "loss": 0.3953,
+      "step": 1311
+    },
+    {
+      "epoch": 2.027820710973725,
+      "grad_norm": 0.49219155345852056,
+      "learning_rate": 5.9335935245959657e-05,
+      "loss": 0.3904,
+      "step": 1312
+    },
+    {
+      "epoch": 2.029366306027821,
+      "grad_norm": 0.46819033388509657,
+      "learning_rate": 5.9298134206305344e-05,
+      "loss": 0.405,
+      "step": 1313
+    },
+    {
+      "epoch": 2.0309119010819163,
+      "grad_norm": 0.3731531854940534,
+      "learning_rate": 5.926031069005115e-05,
+      "loss": 0.398,
+      "step": 1314
+    },
+    {
+      "epoch": 2.0324574961360122,
+      "grad_norm": 0.34717209437762364,
+      "learning_rate": 5.9222464741250234e-05,
+      "loss": 0.3906,
+      "step": 1315
+    },
+    {
+      "epoch": 2.034003091190108,
+      "grad_norm": 0.32135536723818864,
+      "learning_rate": 5.9184596403981907e-05,
+      "loss": 0.3919,
+      "step": 1316
+    },
+    {
+      "epoch": 2.035548686244204,
+      "grad_norm": 0.3187246820508031,
+      "learning_rate": 5.9146705722351535e-05,
+      "loss": 0.387,
+      "step": 1317
+    },
+    {
+      "epoch": 2.0370942812983,
+      "grad_norm": 0.3844105515744366,
+      "learning_rate": 5.910879274049052e-05,
+      "loss": 0.3858,
+      "step": 1318
+    },
+    {
+      "epoch": 2.038639876352396,
+      "grad_norm": 0.3212345741912804,
+      "learning_rate": 5.9070857502556255e-05,
+      "loss": 0.3982,
+      "step": 1319
+    },
+    {
+      "epoch": 2.0401854714064913,
+      "grad_norm": 0.2837636853315771,
+      "learning_rate": 5.903290005273199e-05,
+      "loss": 0.3967,
+      "step": 1320
+    },
+    {
+      "epoch": 2.041731066460587,
+      "grad_norm": 0.2753993787315415,
+      "learning_rate": 5.899492043522695e-05,
+      "loss": 0.3999,
+      "step": 1321
+    },
+    {
+      "epoch": 2.043276661514683,
+      "grad_norm": 0.23306084091444113,
+      "learning_rate": 5.895691869427608e-05,
+      "loss": 0.3884,
+      "step": 1322
+    },
+    {
+      "epoch": 2.044822256568779,
+      "grad_norm": 0.2752161170160501,
+      "learning_rate": 5.8918894874140136e-05,
+      "loss": 0.3999,
+      "step": 1323
+    },
+    {
+      "epoch": 2.046367851622875,
+      "grad_norm": 0.2893023265733322,
+      "learning_rate": 5.88808490191056e-05,
+      "loss": 0.3912,
+      "step": 1324
+    },
+    {
+      "epoch": 2.047913446676971,
+      "grad_norm": 0.39089214093172914,
+      "learning_rate": 5.884278117348458e-05,
+      "loss": 0.3886,
+      "step": 1325
+    },
+    {
+      "epoch": 2.0494590417310663,
+      "grad_norm": 0.47685298138721893,
+      "learning_rate": 5.880469138161484e-05,
+      "loss": 0.4072,
+      "step": 1326
+    },
+    {
+      "epoch": 2.051004636785162,
+      "grad_norm": 0.36289657252757496,
+      "learning_rate": 5.876657968785969e-05,
+      "loss": 0.3961,
+      "step": 1327
+    },
+    {
+      "epoch": 2.052550231839258,
+      "grad_norm": 0.2614837486118298,
+      "learning_rate": 5.872844613660792e-05,
+      "loss": 0.3826,
+      "step": 1328
+    },
+    {
+      "epoch": 2.054095826893354,
+      "grad_norm": 0.23327559900974454,
+      "learning_rate": 5.8690290772273815e-05,
+      "loss": 0.395,
+      "step": 1329
+    },
+    {
+      "epoch": 2.05564142194745,
+      "grad_norm": 0.25665314076887247,
+      "learning_rate": 5.865211363929707e-05,
+      "loss": 0.3943,
+      "step": 1330
+    },
+    {
+      "epoch": 2.0571870170015454,
+      "grad_norm": 0.28734827929141427,
+      "learning_rate": 5.861391478214268e-05,
+      "loss": 0.387,
+      "step": 1331
+    },
+    {
+      "epoch": 2.0587326120556413,
+      "grad_norm": 0.2927672387638488,
+      "learning_rate": 5.8575694245301005e-05,
+      "loss": 0.3867,
+      "step": 1332
+    },
+    {
+      "epoch": 2.060278207109737,
+      "grad_norm": 0.28953425176851866,
+      "learning_rate": 5.853745207328762e-05,
+      "loss": 0.4011,
+      "step": 1333
+    },
+    {
+      "epoch": 2.061823802163833,
+      "grad_norm": 0.28360602802351886,
+      "learning_rate": 5.8499188310643293e-05,
+      "loss": 0.3866,
+      "step": 1334
+    },
+    {
+      "epoch": 2.063369397217929,
+      "grad_norm": 0.28169766603083646,
+      "learning_rate": 5.8460903001933996e-05,
+      "loss": 0.3889,
+      "step": 1335
+    },
+    {
+      "epoch": 2.064914992272025,
+      "grad_norm": 0.29187158559313897,
+      "learning_rate": 5.842259619175071e-05,
+      "loss": 0.3887,
+      "step": 1336
+    },
+    {
+      "epoch": 2.0664605873261204,
+      "grad_norm": 0.30502182980252696,
+      "learning_rate": 5.838426792470952e-05,
+      "loss": 0.3931,
+      "step": 1337
+    },
+    {
+      "epoch": 2.0680061823802163,
+      "grad_norm": 0.2797816265394829,
+      "learning_rate": 5.834591824545149e-05,
+      "loss": 0.3849,
+      "step": 1338
+    },
+    {
+      "epoch": 2.069551777434312,
+      "grad_norm": 0.23022841047299905,
+      "learning_rate": 5.830754719864259e-05,
+      "loss": 0.3943,
+      "step": 1339
+    },
+    {
+      "epoch": 2.071097372488408,
+      "grad_norm": 0.25594596278299064,
+      "learning_rate": 5.826915482897374e-05,
+      "loss": 0.3889,
+      "step": 1340
+    },
+    {
+      "epoch": 2.072642967542504,
+      "grad_norm": 0.29263312885541526,
+      "learning_rate": 5.8230741181160625e-05,
+      "loss": 0.394,
+      "step": 1341
+    },
+    {
+      "epoch": 2.0741885625966,
+      "grad_norm": 0.3132155920309119,
+      "learning_rate": 5.8192306299943764e-05,
+      "loss": 0.3876,
+      "step": 1342
+    },
+    {
+      "epoch": 2.0757341576506954,
+      "grad_norm": 0.37803460940583933,
+      "learning_rate": 5.815385023008839e-05,
+      "loss": 0.3967,
+      "step": 1343
+    },
+    {
+      "epoch": 2.0772797527047913,
+      "grad_norm": 0.39599324509331707,
+      "learning_rate": 5.8115373016384404e-05,
+      "loss": 0.3831,
+      "step": 1344
+    },
+    {
+      "epoch": 2.078825347758887,
+      "grad_norm": 0.36883394820473075,
+      "learning_rate": 5.807687470364635e-05,
+      "loss": 0.4026,
+      "step": 1345
+    },
+    {
+      "epoch": 2.080370942812983,
+      "grad_norm": 0.3163780061294479,
+      "learning_rate": 5.803835533671334e-05,
+      "loss": 0.3954,
+      "step": 1346
+    },
+    {
+      "epoch": 2.081916537867079,
+      "grad_norm": 0.2548836687691033,
+      "learning_rate": 5.7999814960449e-05,
+      "loss": 0.3931,
+      "step": 1347
+    },
+    {
+      "epoch": 2.0834621329211744,
+      "grad_norm": 0.3777335328739351,
+      "learning_rate": 5.7961253619741415e-05,
+      "loss": 0.3921,
+      "step": 1348
+    },
+    {
+      "epoch": 2.0850077279752703,
+      "grad_norm": 0.4696687530915752,
+      "learning_rate": 5.7922671359503145e-05,
+      "loss": 0.3995,
+      "step": 1349
+    },
+    {
+      "epoch": 2.0865533230293662,
+      "grad_norm": 0.4653528251375817,
+      "learning_rate": 5.788406822467105e-05,
+      "loss": 0.3991,
+      "step": 1350
+    },
+    {
+      "epoch": 2.088098918083462,
+      "grad_norm": 0.4146975919005292,
+      "learning_rate": 5.784544426020632e-05,
+      "loss": 0.3966,
+      "step": 1351
+    },
+    {
+      "epoch": 2.089644513137558,
+      "grad_norm": 0.3602980967706405,
+      "learning_rate": 5.780679951109442e-05,
+      "loss": 0.4004,
+      "step": 1352
+    },
+    {
+      "epoch": 2.091190108191654,
+      "grad_norm": 0.3812258112220021,
+      "learning_rate": 5.776813402234503e-05,
+      "loss": 0.3895,
+      "step": 1353
+    },
+    {
+      "epoch": 2.0927357032457494,
+      "grad_norm": 0.5348490661093533,
+      "learning_rate": 5.772944783899195e-05,
+      "loss": 0.3868,
+      "step": 1354
+    },
+    {
+      "epoch": 2.0942812982998453,
+      "grad_norm": 0.5341805710201996,
+      "learning_rate": 5.769074100609311e-05,
+      "loss": 0.4031,
+      "step": 1355
+    },
+    {
+      "epoch": 2.0958268933539412,
+      "grad_norm": 0.42718412388594196,
+      "learning_rate": 5.765201356873049e-05,
+      "loss": 0.3924,
+      "step": 1356
+    },
+    {
+      "epoch": 2.097372488408037,
+      "grad_norm": 0.33149597465277597,
+      "learning_rate": 5.761326557201006e-05,
+      "loss": 0.3913,
+      "step": 1357
+    },
+    {
+      "epoch": 2.098918083462133,
+      "grad_norm": 0.26099816268109227,
+      "learning_rate": 5.7574497061061745e-05,
+      "loss": 0.391,
+      "step": 1358
+    },
+    {
+      "epoch": 2.100463678516229,
+      "grad_norm": 0.32062129249672705,
+      "learning_rate": 5.753570808103934e-05,
+      "loss": 0.3899,
+      "step": 1359
+    },
+    {
+      "epoch": 2.1020092735703244,
+      "grad_norm": 0.48520151913932597,
+      "learning_rate": 5.7496898677120506e-05,
+      "loss": 0.3891,
+      "step": 1360
+    },
+    {
+      "epoch": 2.1035548686244203,
+      "grad_norm": 0.4735338135968625,
+      "learning_rate": 5.745806889450669e-05,
+      "loss": 0.3927,
+      "step": 1361
+    },
+    {
+      "epoch": 2.105100463678516,
+      "grad_norm": 0.47126817539463034,
+      "learning_rate": 5.7419218778423045e-05,
+      "loss": 0.3799,
+      "step": 1362
+    },
+    {
+      "epoch": 2.106646058732612,
+      "grad_norm": 0.4650326573097983,
+      "learning_rate": 5.738034837411845e-05,
+      "loss": 0.4049,
+      "step": 1363
+    },
+    {
+      "epoch": 2.108191653786708,
+      "grad_norm": 0.5052600261623205,
+      "learning_rate": 5.734145772686539e-05,
+      "loss": 0.3907,
+      "step": 1364
+    },
+    {
+      "epoch": 2.109737248840804,
+      "grad_norm": 0.4492378071242691,
+      "learning_rate": 5.7302546881959905e-05,
+      "loss": 0.3929,
+      "step": 1365
+    },
+    {
+      "epoch": 2.1112828438948994,
+      "grad_norm": 0.4022561190795431,
+      "learning_rate": 5.726361588472161e-05,
+      "loss": 0.3957,
+      "step": 1366
+    },
+    {
+      "epoch": 2.1128284389489953,
+      "grad_norm": 0.3232004812396111,
+      "learning_rate": 5.722466478049356e-05,
+      "loss": 0.3951,
+      "step": 1367
+    },
+    {
+      "epoch": 2.114374034003091,
+      "grad_norm": 0.29120789738606623,
+      "learning_rate": 5.718569361464221e-05,
+      "loss": 0.3973,
+      "step": 1368
+    },
+    {
+      "epoch": 2.115919629057187,
+      "grad_norm": 0.3297836661004876,
+      "learning_rate": 5.714670243255745e-05,
+      "loss": 0.4035,
+      "step": 1369
+    },
+    {
+      "epoch": 2.117465224111283,
+      "grad_norm": 0.33666164737551685,
+      "learning_rate": 5.710769127965238e-05,
+      "loss": 0.4048,
+      "step": 1370
+    },
+    {
+      "epoch": 2.1190108191653785,
+      "grad_norm": 0.32318941343611973,
+      "learning_rate": 5.7068660201363464e-05,
+      "loss": 0.3886,
+      "step": 1371
+    },
+    {
+      "epoch": 2.1205564142194744,
+      "grad_norm": 0.31605456636520374,
+      "learning_rate": 5.7029609243150306e-05,
+      "loss": 0.3986,
+      "step": 1372
+    },
+    {
+      "epoch": 2.1221020092735703,
+      "grad_norm": 0.3444017106165271,
+      "learning_rate": 5.699053845049568e-05,
+      "loss": 0.3974,
+      "step": 1373
+    },
+    {
+      "epoch": 2.123647604327666,
+      "grad_norm": 0.32076054536360127,
+      "learning_rate": 5.695144786890547e-05,
+      "loss": 0.3941,
+      "step": 1374
+    },
+    {
+      "epoch": 2.125193199381762,
+      "grad_norm": 0.3375559972286297,
+      "learning_rate": 5.6912337543908616e-05,
+      "loss": 0.4049,
+      "step": 1375
+    },
+    {
+      "epoch": 2.126738794435858,
+      "grad_norm": 0.3398886977192635,
+      "learning_rate": 5.687320752105702e-05,
+      "loss": 0.4028,
+      "step": 1376
+    },
+    {
+      "epoch": 2.1282843894899535,
+      "grad_norm": 0.28589222478831905,
+      "learning_rate": 5.683405784592557e-05,
+      "loss": 0.4076,
+      "step": 1377
+    },
+    {
+      "epoch": 2.1298299845440494,
+      "grad_norm": 0.2991766255827722,
+      "learning_rate": 5.679488856411201e-05,
+      "loss": 0.3986,
+      "step": 1378
+    },
+    {
+      "epoch": 2.1313755795981453,
+      "grad_norm": 0.2730775297870511,
+      "learning_rate": 5.6755699721236927e-05,
+      "loss": 0.3962,
+      "step": 1379
+    },
+    {
+      "epoch": 2.132921174652241,
+      "grad_norm": 0.2279078127037609,
+      "learning_rate": 5.67164913629437e-05,
+      "loss": 0.3951,
+      "step": 1380
+    },
+    {
+      "epoch": 2.134466769706337,
+      "grad_norm": 0.26190056792962724,
+      "learning_rate": 5.6677263534898425e-05,
+      "loss": 0.3918,
+      "step": 1381
+    },
+    {
+      "epoch": 2.1360123647604325,
+      "grad_norm": 0.30977615037207307,
+      "learning_rate": 5.663801628278988e-05,
+      "loss": 0.388,
+      "step": 1382
+    },
+    {
+      "epoch": 2.1375579598145285,
+      "grad_norm": 0.2902462039984759,
+      "learning_rate": 5.659874965232948e-05,
+      "loss": 0.3905,
+      "step": 1383
+    },
+    {
+      "epoch": 2.1391035548686244,
+      "grad_norm": 0.27677467210321927,
+      "learning_rate": 5.655946368925118e-05,
+      "loss": 0.3923,
+      "step": 1384
+    },
+    {
+      "epoch": 2.1406491499227203,
+      "grad_norm": 0.24262357446186741,
+      "learning_rate": 5.652015843931149e-05,
+      "loss": 0.3868,
+      "step": 1385
+    },
+    {
+      "epoch": 2.142194744976816,
+      "grad_norm": 0.21439927480680235,
+      "learning_rate": 5.648083394828935e-05,
+      "loss": 0.396,
+      "step": 1386
+    },
+    {
+      "epoch": 2.143740340030912,
+      "grad_norm": 0.3024814027973054,
+      "learning_rate": 5.644149026198612e-05,
+      "loss": 0.3937,
+      "step": 1387
+    },
+    {
+      "epoch": 2.1452859350850075,
+      "grad_norm": 0.33105463220970677,
+      "learning_rate": 5.640212742622553e-05,
+      "loss": 0.4014,
+      "step": 1388
+    },
+    {
+      "epoch": 2.1468315301391034,
+      "grad_norm": 0.2393384112444071,
+      "learning_rate": 5.636274548685361e-05,
+      "loss": 0.3896,
+      "step": 1389
+    },
+    {
+      "epoch": 2.1483771251931993,
+      "grad_norm": 0.2586734232627636,
+      "learning_rate": 5.63233444897386e-05,
+      "loss": 0.3975,
+      "step": 1390
+    },
+    {
+      "epoch": 2.1499227202472952,
+      "grad_norm": 0.2303543368209657,
+      "learning_rate": 5.628392448077101e-05,
+      "loss": 0.3936,
+      "step": 1391
+    },
+    {
+      "epoch": 2.151468315301391,
+      "grad_norm": 0.2293507554950428,
+      "learning_rate": 5.624448550586342e-05,
+      "loss": 0.396,
+      "step": 1392
+    },
+    {
+      "epoch": 2.153013910355487,
+      "grad_norm": 0.2548338874339664,
+      "learning_rate": 5.620502761095056e-05,
+      "loss": 0.397,
+      "step": 1393
+    },
+    {
+      "epoch": 2.1545595054095825,
+      "grad_norm": 0.22722623652559307,
+      "learning_rate": 5.616555084198915e-05,
+      "loss": 0.3955,
+      "step": 1394
+    },
+    {
+      "epoch": 2.1561051004636784,
+      "grad_norm": 0.1918331492110652,
+      "learning_rate": 5.6126055244957925e-05,
+      "loss": 0.3882,
+      "step": 1395
+    },
+    {
+      "epoch": 2.1576506955177743,
+      "grad_norm": 0.2166050594320556,
+      "learning_rate": 5.608654086585752e-05,
+      "loss": 0.4004,
+      "step": 1396
+    },
+    {
+      "epoch": 2.1591962905718702,
+      "grad_norm": 0.2809903221060055,
+      "learning_rate": 5.604700775071048e-05,
+      "loss": 0.4012,
+      "step": 1397
+    },
+    {
+      "epoch": 2.160741885625966,
+      "grad_norm": 0.36215355227293844,
+      "learning_rate": 5.600745594556113e-05,
+      "loss": 0.3953,
+      "step": 1398
+    },
+    {
+      "epoch": 2.162287480680062,
+      "grad_norm": 0.401216351650045,
+      "learning_rate": 5.596788549647562e-05,
+      "loss": 0.3986,
+      "step": 1399
+    },
+    {
+      "epoch": 2.1638330757341575,
+      "grad_norm": 0.36923309321816083,
+      "learning_rate": 5.592829644954175e-05,
+      "loss": 0.3896,
+      "step": 1400
+    },
+    {
+      "epoch": 2.1653786707882534,
+      "grad_norm": 0.28053308499077567,
+      "learning_rate": 5.588868885086902e-05,
+      "loss": 0.3945,
+      "step": 1401
+    },
+    {
+      "epoch": 2.1669242658423493,
+      "grad_norm": 0.2347720773966801,
+      "learning_rate": 5.584906274658854e-05,
+      "loss": 0.3896,
+      "step": 1402
+    },
+    {
+      "epoch": 2.1684698608964452,
+      "grad_norm": 0.2996575204398839,
+      "learning_rate": 5.580941818285295e-05,
+      "loss": 0.3956,
+      "step": 1403
+    },
+    {
+      "epoch": 2.170015455950541,
+      "grad_norm": 0.24214511700057684,
+      "learning_rate": 5.576975520583642e-05,
+      "loss": 0.3948,
+      "step": 1404
+    },
+    {
+      "epoch": 2.1715610510046366,
+      "grad_norm": 0.20950657680483423,
+      "learning_rate": 5.573007386173452e-05,
+      "loss": 0.3925,
+      "step": 1405
+    },
+    {
+      "epoch": 2.1731066460587325,
+      "grad_norm": 0.26792625106666884,
+      "learning_rate": 5.569037419676427e-05,
+      "loss": 0.3992,
+      "step": 1406
+    },
+    {
+      "epoch": 2.1746522411128284,
+      "grad_norm": 0.2712494902998535,
+      "learning_rate": 5.5650656257163985e-05,
+      "loss": 0.3859,
+      "step": 1407
+    },
+    {
+      "epoch": 2.1761978361669243,
+      "grad_norm": 0.2930171260012353,
+      "learning_rate": 5.561092008919328e-05,
+      "loss": 0.3958,
+      "step": 1408
+    },
+    {
+      "epoch": 2.17774343122102,
+      "grad_norm": 0.304917573510114,
+      "learning_rate": 5.557116573913301e-05,
+      "loss": 0.3848,
+      "step": 1409
+    },
+    {
+      "epoch": 2.179289026275116,
+      "grad_norm": 0.2414499333351336,
+      "learning_rate": 5.553139325328518e-05,
+      "loss": 0.4,
+      "step": 1410
+    },
+    {
+      "epoch": 2.1808346213292116,
+      "grad_norm": 0.34294961997680773,
+      "learning_rate": 5.5491602677972945e-05,
+      "loss": 0.3981,
+      "step": 1411
+    },
+    {
+      "epoch": 2.1823802163833075,
+      "grad_norm": 0.37602037518813536,
+      "learning_rate": 5.545179405954051e-05,
+      "loss": 0.3921,
+      "step": 1412
+    },
+    {
+      "epoch": 2.1839258114374034,
+      "grad_norm": 0.25752242189578534,
+      "learning_rate": 5.5411967444353105e-05,
+      "loss": 0.3996,
+      "step": 1413
+    },
+    {
+      "epoch": 2.1854714064914993,
+      "grad_norm": 0.31666571784081177,
+      "learning_rate": 5.537212287879693e-05,
+      "loss": 0.3922,
+      "step": 1414
+    },
+    {
+      "epoch": 2.187017001545595,
+      "grad_norm": 0.3736689968910478,
+      "learning_rate": 5.533226040927906e-05,
+      "loss": 0.3975,
+      "step": 1415
+    },
+    {
+      "epoch": 2.1885625965996907,
+      "grad_norm": 0.2916315544781542,
+      "learning_rate": 5.5292380082227476e-05,
+      "loss": 0.3898,
+      "step": 1416
+    },
+    {
+      "epoch": 2.1901081916537866,
+      "grad_norm": 0.23508392644454182,
+      "learning_rate": 5.5252481944090876e-05,
+      "loss": 0.3905,
+      "step": 1417
+    },
+    {
+      "epoch": 2.1916537867078825,
+      "grad_norm": 0.2089635516135591,
+      "learning_rate": 5.521256604133879e-05,
+      "loss": 0.396,
+      "step": 1418
+    },
+    {
+      "epoch": 2.1931993817619784,
+      "grad_norm": 0.283463538403399,
+      "learning_rate": 5.51726324204614e-05,
+      "loss": 0.3932,
+      "step": 1419
+    },
+    {
+      "epoch": 2.1947449768160743,
+      "grad_norm": 0.20312496600238453,
+      "learning_rate": 5.513268112796951e-05,
+      "loss": 0.3871,
+      "step": 1420
+    },
+    {
+      "epoch": 2.19629057187017,
+      "grad_norm": 0.23945133879207264,
+      "learning_rate": 5.509271221039452e-05,
+      "loss": 0.4051,
+      "step": 1421
+    },
+    {
+      "epoch": 2.1978361669242656,
+      "grad_norm": 0.26650563107564185,
+      "learning_rate": 5.505272571428837e-05,
+      "loss": 0.3929,
+      "step": 1422
+    },
+    {
+      "epoch": 2.1993817619783615,
+      "grad_norm": 0.29965664042142814,
+      "learning_rate": 5.501272168622345e-05,
+      "loss": 0.4062,
+      "step": 1423
+    },
+    {
+      "epoch": 2.2009273570324575,
+      "grad_norm": 0.2829338590370037,
+      "learning_rate": 5.497270017279259e-05,
+      "loss": 0.3996,
+      "step": 1424
+    },
+    {
+      "epoch": 2.2024729520865534,
+      "grad_norm": 0.29676225707278125,
+      "learning_rate": 5.493266122060897e-05,
+      "loss": 0.3936,
+      "step": 1425
+    },
+    {
+      "epoch": 2.2040185471406493,
+      "grad_norm": 0.35621600063358044,
+      "learning_rate": 5.48926048763061e-05,
+      "loss": 0.3995,
+      "step": 1426
+    },
+    {
+      "epoch": 2.205564142194745,
+      "grad_norm": 0.382789454234158,
+      "learning_rate": 5.485253118653774e-05,
+      "loss": 0.3828,
+      "step": 1427
+    },
+    {
+      "epoch": 2.2071097372488406,
+      "grad_norm": 0.32438131697044115,
+      "learning_rate": 5.481244019797782e-05,
+      "loss": 0.3964,
+      "step": 1428
+    },
+    {
+      "epoch": 2.2086553323029365,
+      "grad_norm": 0.2522710585082952,
+      "learning_rate": 5.477233195732046e-05,
+      "loss": 0.3998,
+      "step": 1429
+    },
+    {
+      "epoch": 2.2102009273570324,
+      "grad_norm": 0.22076782465890474,
+      "learning_rate": 5.473220651127988e-05,
+      "loss": 0.3961,
+      "step": 1430
+    },
+    {
+      "epoch": 2.2117465224111283,
+      "grad_norm": 0.24445337462545094,
+      "learning_rate": 5.469206390659028e-05,
+      "loss": 0.3788,
+      "step": 1431
+    },
+    {
+      "epoch": 2.2132921174652243,
+      "grad_norm": 0.23976966421294135,
+      "learning_rate": 5.46519041900059e-05,
+      "loss": 0.3933,
+      "step": 1432
+    },
+    {
+      "epoch": 2.21483771251932,
+      "grad_norm": 0.2508674942889716,
+      "learning_rate": 5.461172740830091e-05,
+      "loss": 0.3875,
+      "step": 1433
+    },
+    {
+      "epoch": 2.2163833075734156,
+      "grad_norm": 0.2786405104063059,
+      "learning_rate": 5.45715336082693e-05,
+      "loss": 0.3798,
+      "step": 1434
+    },
+    {
+      "epoch": 2.2179289026275115,
+      "grad_norm": 0.3579458488758703,
+      "learning_rate": 5.453132283672496e-05,
+      "loss": 0.4041,
+      "step": 1435
+    },
+    {
+      "epoch": 2.2194744976816074,
+      "grad_norm": 0.3935960503455614,
+      "learning_rate": 5.4491095140501486e-05,
+      "loss": 0.3933,
+      "step": 1436
+    },
+    {
+      "epoch": 2.2210200927357033,
+      "grad_norm": 0.3697590188908626,
+      "learning_rate": 5.4450850566452213e-05,
+      "loss": 0.3921,
+      "step": 1437
+    },
+    {
+      "epoch": 2.2225656877897992,
+      "grad_norm": 0.3605308818179636,
+      "learning_rate": 5.441058916145011e-05,
+      "loss": 0.3999,
+      "step": 1438
+    },
+    {
+      "epoch": 2.2241112828438947,
+      "grad_norm": 0.3356567615719234,
+      "learning_rate": 5.43703109723878e-05,
+      "loss": 0.3909,
+      "step": 1439
+    },
+    {
+      "epoch": 2.2256568778979906,
+      "grad_norm": 0.30531472053735503,
+      "learning_rate": 5.43300160461774e-05,
+      "loss": 0.3928,
+      "step": 1440
+    },
+    {
+      "epoch": 2.2272024729520865,
+      "grad_norm": 0.2565363158264292,
+      "learning_rate": 5.428970442975053e-05,
+      "loss": 0.3945,
+      "step": 1441
+    },
+    {
+      "epoch": 2.2287480680061824,
+      "grad_norm": 0.2380223524712181,
+      "learning_rate": 5.424937617005829e-05,
+      "loss": 0.3793,
+      "step": 1442
+    },
+    {
+      "epoch": 2.2302936630602783,
+      "grad_norm": 0.2560952406565104,
+      "learning_rate": 5.4209031314071116e-05,
+      "loss": 0.3911,
+      "step": 1443
+    },
+    {
+      "epoch": 2.2318392581143742,
+      "grad_norm": 0.24313728437682017,
+      "learning_rate": 5.4168669908778814e-05,
+      "loss": 0.4018,
+      "step": 1444
+    },
+    {
+      "epoch": 2.2333848531684697,
+      "grad_norm": 0.21664727345403825,
+      "learning_rate": 5.412829200119041e-05,
+      "loss": 0.3902,
+      "step": 1445
+    },
+    {
+      "epoch": 2.2349304482225656,
+      "grad_norm": 0.2675734440025409,
+      "learning_rate": 5.408789763833421e-05,
+      "loss": 0.3898,
+      "step": 1446
+    },
+    {
+      "epoch": 2.2364760432766615,
+      "grad_norm": 0.2923302849334493,
+      "learning_rate": 5.4047486867257664e-05,
+      "loss": 0.3893,
+      "step": 1447
+    },
+    {
+      "epoch": 2.2380216383307574,
+      "grad_norm": 0.26972423411484947,
+      "learning_rate": 5.4007059735027315e-05,
+      "loss": 0.397,
+      "step": 1448
+    },
+    {
+      "epoch": 2.2395672333848533,
+      "grad_norm": 0.304336195544412,
+      "learning_rate": 5.3966616288728774e-05,
+      "loss": 0.4051,
+      "step": 1449
+    },
+    {
+      "epoch": 2.2411128284389488,
+      "grad_norm": 0.28943902807642036,
+      "learning_rate": 5.3926156575466675e-05,
+      "loss": 0.3941,
+      "step": 1450
+    },
+    {
+      "epoch": 2.2426584234930447,
+      "grad_norm": 0.28694545302948715,
+      "learning_rate": 5.388568064236457e-05,
+      "loss": 0.3915,
+      "step": 1451
+    },
+    {
+      "epoch": 2.2442040185471406,
+      "grad_norm": 0.33675761551956246,
+      "learning_rate": 5.3845188536564884e-05,
+      "loss": 0.3978,
+      "step": 1452
+    },
+    {
+      "epoch": 2.2457496136012365,
+      "grad_norm": 0.38081008970903646,
+      "learning_rate": 5.3804680305228934e-05,
+      "loss": 0.3881,
+      "step": 1453
+    },
+    {
+      "epoch": 2.2472952086553324,
+      "grad_norm": 0.34809364873567517,
+      "learning_rate": 5.3764155995536796e-05,
+      "loss": 0.396,
+      "step": 1454
+    },
+    {
+      "epoch": 2.2488408037094283,
+      "grad_norm": 0.3633482183310246,
+      "learning_rate": 5.3723615654687224e-05,
+      "loss": 0.3879,
+      "step": 1455
+    },
+    {
+      "epoch": 2.250386398763524,
+      "grad_norm": 0.36431906421062726,
+      "learning_rate": 5.368305932989773e-05,
+      "loss": 0.3941,
+      "step": 1456
+    },
+    {
+      "epoch": 2.2519319938176197,
+      "grad_norm": 0.4049586685164246,
+      "learning_rate": 5.364248706840437e-05,
+      "loss": 0.3918,
+      "step": 1457
+    },
+    {
+      "epoch": 2.2534775888717156,
+      "grad_norm": 0.39800391551422143,
+      "learning_rate": 5.3601898917461826e-05,
+      "loss": 0.3888,
+      "step": 1458
+    },
+    {
+      "epoch": 2.2550231839258115,
+      "grad_norm": 0.31734587330195074,
+      "learning_rate": 5.35612949243432e-05,
+      "loss": 0.3917,
+      "step": 1459
+    },
+    {
+      "epoch": 2.2565687789799074,
+      "grad_norm": 0.2570755668387734,
+      "learning_rate": 5.352067513634011e-05,
+      "loss": 0.391,
+      "step": 1460
+    },
+    {
+      "epoch": 2.2581143740340033,
+      "grad_norm": 0.24640982169832737,
+      "learning_rate": 5.3480039600762586e-05,
+      "loss": 0.3946,
+      "step": 1461
+    },
+    {
+      "epoch": 2.2596599690880987,
+      "grad_norm": 0.23778783202732132,
+      "learning_rate": 5.343938836493894e-05,
+      "loss": 0.3869,
+      "step": 1462
+    },
+    {
+      "epoch": 2.2612055641421946,
+      "grad_norm": 0.2653726286543613,
+      "learning_rate": 5.339872147621579e-05,
+      "loss": 0.3963,
+      "step": 1463
+    },
+    {
+      "epoch": 2.2627511591962906,
+      "grad_norm": 0.23187671980103855,
+      "learning_rate": 5.335803898195803e-05,
+      "loss": 0.3837,
+      "step": 1464
+    },
+    {
+      "epoch": 2.2642967542503865,
+      "grad_norm": 0.2480098286152096,
+      "learning_rate": 5.3317340929548676e-05,
+      "loss": 0.3948,
+      "step": 1465
+    },
+    {
+      "epoch": 2.2658423493044824,
+      "grad_norm": 0.2768215559146363,
+      "learning_rate": 5.327662736638886e-05,
+      "loss": 0.3904,
+      "step": 1466
+    },
+    {
+      "epoch": 2.2673879443585783,
+      "grad_norm": 0.3541058579001065,
+      "learning_rate": 5.3235898339897846e-05,
+      "loss": 0.3868,
+      "step": 1467
+    },
+    {
+      "epoch": 2.2689335394126737,
+      "grad_norm": 0.3455770642766049,
+      "learning_rate": 5.319515389751283e-05,
+      "loss": 0.3956,
+      "step": 1468
+    },
+    {
+      "epoch": 2.2704791344667696,
+      "grad_norm": 0.24872912183460227,
+      "learning_rate": 5.3154394086688995e-05,
+      "loss": 0.3934,
+      "step": 1469
+    },
+    {
+      "epoch": 2.2720247295208655,
+      "grad_norm": 0.22593842204717707,
+      "learning_rate": 5.311361895489946e-05,
+      "loss": 0.4029,
+      "step": 1470
+    },
+    {
+      "epoch": 2.2735703245749614,
+      "grad_norm": 0.18875541559897221,
+      "learning_rate": 5.3072828549635107e-05,
+      "loss": 0.3868,
+      "step": 1471
+    },
+    {
+      "epoch": 2.2751159196290573,
+      "grad_norm": 0.20398086887059988,
+      "learning_rate": 5.303202291840471e-05,
+      "loss": 0.3923,
+      "step": 1472
+    },
+    {
+      "epoch": 2.276661514683153,
+      "grad_norm": 0.23263084748729548,
+      "learning_rate": 5.299120210873467e-05,
+      "loss": 0.3837,
+      "step": 1473
+    },
+    {
+      "epoch": 2.2782071097372487,
+      "grad_norm": 0.21563000272531446,
+      "learning_rate": 5.295036616816914e-05,
+      "loss": 0.3931,
+      "step": 1474
+    },
+    {
+      "epoch": 2.2797527047913446,
+      "grad_norm": 0.22171142965832003,
+      "learning_rate": 5.290951514426988e-05,
+      "loss": 0.3956,
+      "step": 1475
+    },
+    {
+      "epoch": 2.2812982998454405,
+      "grad_norm": 0.22587717378598995,
+      "learning_rate": 5.286864908461619e-05,
+      "loss": 0.3872,
+      "step": 1476
+    },
+    {
+      "epoch": 2.2828438948995364,
+      "grad_norm": 0.1971272547786867,
+      "learning_rate": 5.282776803680492e-05,
+      "loss": 0.4036,
+      "step": 1477
+    },
+    {
+      "epoch": 2.2843894899536323,
+      "grad_norm": 0.2224845402260878,
+      "learning_rate": 5.278687204845035e-05,
+      "loss": 0.3897,
+      "step": 1478
+    },
+    {
+      "epoch": 2.285935085007728,
+      "grad_norm": 0.19441003764474904,
+      "learning_rate": 5.2745961167184185e-05,
+      "loss": 0.3889,
+      "step": 1479
+    },
+    {
+      "epoch": 2.2874806800618237,
+      "grad_norm": 0.18171125267942248,
+      "learning_rate": 5.270503544065544e-05,
+      "loss": 0.402,
+      "step": 1480
+    },
+    {
+      "epoch": 2.2890262751159196,
+      "grad_norm": 0.17983694611736392,
+      "learning_rate": 5.266409491653046e-05,
+      "loss": 0.3854,
+      "step": 1481
+    },
+    {
+      "epoch": 2.2905718701700155,
+      "grad_norm": 0.20254679567986317,
+      "learning_rate": 5.262313964249281e-05,
+      "loss": 0.3988,
+      "step": 1482
+    },
+    {
+      "epoch": 2.2921174652241114,
+      "grad_norm": 0.19184410298081583,
+      "learning_rate": 5.258216966624321e-05,
+      "loss": 0.3851,
+      "step": 1483
+    },
+    {
+      "epoch": 2.293663060278207,
+      "grad_norm": 0.254767017913148,
+      "learning_rate": 5.254118503549954e-05,
+      "loss": 0.3928,
+      "step": 1484
+    },
+    {
+      "epoch": 2.295208655332303,
+      "grad_norm": 0.31469261555728983,
+      "learning_rate": 5.2500185797996734e-05,
+      "loss": 0.39,
+      "step": 1485
+    },
+    {
+      "epoch": 2.2967542503863987,
+      "grad_norm": 0.28076586148567945,
+      "learning_rate": 5.245917200148675e-05,
+      "loss": 0.394,
+      "step": 1486
+    },
+    {
+      "epoch": 2.2982998454404946,
+      "grad_norm": 0.25246618561021705,
+      "learning_rate": 5.241814369373845e-05,
+      "loss": 0.3939,
+      "step": 1487
+    },
+    {
+      "epoch": 2.2998454404945905,
+      "grad_norm": 0.22736004566702298,
+      "learning_rate": 5.237710092253768e-05,
+      "loss": 0.3891,
+      "step": 1488
+    },
+    {
+      "epoch": 2.3013910355486864,
+      "grad_norm": 0.25674536214096144,
+      "learning_rate": 5.233604373568707e-05,
+      "loss": 0.3954,
+      "step": 1489
+    },
+    {
+      "epoch": 2.3029366306027823,
+      "grad_norm": 0.24910563016368026,
+      "learning_rate": 5.229497218100606e-05,
+      "loss": 0.3961,
+      "step": 1490
+    },
+    {
+      "epoch": 2.3044822256568778,
+      "grad_norm": 0.25669220171173646,
+      "learning_rate": 5.225388630633083e-05,
+      "loss": 0.4008,
+      "step": 1491
+    },
+    {
+      "epoch": 2.3060278207109737,
+      "grad_norm": 0.21963865673874589,
+      "learning_rate": 5.221278615951422e-05,
+      "loss": 0.3926,
+      "step": 1492
+    },
+    {
+      "epoch": 2.3075734157650696,
+      "grad_norm": 0.2736157715412973,
+      "learning_rate": 5.217167178842573e-05,
+      "loss": 0.396,
+      "step": 1493
+    },
+    {
+      "epoch": 2.3091190108191655,
+      "grad_norm": 0.38381019611426875,
+      "learning_rate": 5.2130543240951366e-05,
+      "loss": 0.3904,
+      "step": 1494
+    },
+    {
+      "epoch": 2.3106646058732614,
+      "grad_norm": 0.3546302751151582,
+      "learning_rate": 5.208940056499371e-05,
+      "loss": 0.3997,
+      "step": 1495
+    },
+    {
+      "epoch": 2.312210200927357,
+      "grad_norm": 0.29651312604337876,
+      "learning_rate": 5.204824380847177e-05,
+      "loss": 0.393,
+      "step": 1496
+    },
+    {
+      "epoch": 2.3137557959814528,
+      "grad_norm": 0.25451612140716673,
+      "learning_rate": 5.2007073019320935e-05,
+      "loss": 0.3971,
+      "step": 1497
+    },
+    {
+      "epoch": 2.3153013910355487,
+      "grad_norm": 0.29775277849373727,
+      "learning_rate": 5.196588824549298e-05,
+      "loss": 0.395,
+      "step": 1498
+    },
+    {
+      "epoch": 2.3168469860896446,
+      "grad_norm": 0.30367259317781947,
+      "learning_rate": 5.192468953495592e-05,
+      "loss": 0.3982,
+      "step": 1499
+    },
+    {
+      "epoch": 2.3183925811437405,
+      "grad_norm": 0.34598867891208274,
+      "learning_rate": 5.188347693569407e-05,
+      "loss": 0.395,
+      "step": 1500
+    },
+    {
+      "epoch": 2.3199381761978364,
+      "grad_norm": 0.37348045757641857,
+      "learning_rate": 5.184225049570783e-05,
+      "loss": 0.3961,
+      "step": 1501
+    },
+    {
+      "epoch": 2.321483771251932,
+      "grad_norm": 0.3362956096767625,
+      "learning_rate": 5.180101026301379e-05,
+      "loss": 0.3966,
+      "step": 1502
+    },
+    {
+      "epoch": 2.3230293663060277,
+      "grad_norm": 0.31799218245762445,
+      "learning_rate": 5.17597562856446e-05,
+      "loss": 0.3978,
+      "step": 1503
+    },
+    {
+      "epoch": 2.3245749613601236,
+      "grad_norm": 0.3464467388056877,
+      "learning_rate": 5.171848861164887e-05,
+      "loss": 0.3948,
+      "step": 1504
+    },
+    {
+      "epoch": 2.3261205564142196,
+      "grad_norm": 0.41609624207931256,
+      "learning_rate": 5.167720728909121e-05,
+      "loss": 0.3844,
+      "step": 1505
+    },
+    {
+      "epoch": 2.3276661514683155,
+      "grad_norm": 0.4745365547147811,
+      "learning_rate": 5.163591236605212e-05,
+      "loss": 0.3882,
+      "step": 1506
+    },
+    {
+      "epoch": 2.329211746522411,
+      "grad_norm": 0.4825019829743042,
+      "learning_rate": 5.1594603890627934e-05,
+      "loss": 0.3853,
+      "step": 1507
+    },
+    {
+      "epoch": 2.330757341576507,
+      "grad_norm": 0.38029637854706505,
+      "learning_rate": 5.155328191093074e-05,
+      "loss": 0.3867,
+      "step": 1508
+    },
+    {
+      "epoch": 2.3323029366306027,
+      "grad_norm": 0.3240935750344444,
+      "learning_rate": 5.151194647508841e-05,
+      "loss": 0.3859,
+      "step": 1509
+    },
+    {
+      "epoch": 2.3338485316846986,
+      "grad_norm": 0.3248722572663255,
+      "learning_rate": 5.147059763124445e-05,
+      "loss": 0.3928,
+      "step": 1510
+    },
+    {
+      "epoch": 2.3353941267387945,
+      "grad_norm": 0.4688604609285743,
+      "learning_rate": 5.1429235427557994e-05,
+      "loss": 0.3982,
+      "step": 1511
+    },
+    {
+      "epoch": 2.3369397217928904,
+      "grad_norm": 0.42273469107871264,
+      "learning_rate": 5.1387859912203734e-05,
+      "loss": 0.3942,
+      "step": 1512
+    },
+    {
+      "epoch": 2.338485316846986,
+      "grad_norm": 0.3433548338511978,
+      "learning_rate": 5.134647113337187e-05,
+      "loss": 0.3836,
+      "step": 1513
+    },
+    {
+      "epoch": 2.340030911901082,
+      "grad_norm": 0.2851349484514058,
+      "learning_rate": 5.130506913926806e-05,
+      "loss": 0.3954,
+      "step": 1514
+    },
+    {
+      "epoch": 2.3415765069551777,
+      "grad_norm": 0.4135287612524575,
+      "learning_rate": 5.1263653978113315e-05,
+      "loss": 0.3851,
+      "step": 1515
+    },
+    {
+      "epoch": 2.3431221020092736,
+      "grad_norm": 0.5691535220332999,
+      "learning_rate": 5.1222225698144024e-05,
+      "loss": 0.3911,
+      "step": 1516
+    },
+    {
+      "epoch": 2.3446676970633695,
+      "grad_norm": 0.5566667027836483,
+      "learning_rate": 5.118078434761185e-05,
+      "loss": 0.398,
+      "step": 1517
+    },
+    {
+      "epoch": 2.346213292117465,
+      "grad_norm": 0.405028370843713,
+      "learning_rate": 5.113932997478364e-05,
+      "loss": 0.3893,
+      "step": 1518
+    },
+    {
+      "epoch": 2.347758887171561,
+      "grad_norm": 0.231212331290742,
+      "learning_rate": 5.1097862627941465e-05,
+      "loss": 0.3956,
+      "step": 1519
+    },
+    {
+      "epoch": 2.349304482225657,
+      "grad_norm": 0.2972009251282469,
+      "learning_rate": 5.105638235538248e-05,
+      "loss": 0.388,
+      "step": 1520
+    },
+    {
+      "epoch": 2.3508500772797527,
+      "grad_norm": 0.39737209705772275,
+      "learning_rate": 5.1014889205418875e-05,
+      "loss": 0.3994,
+      "step": 1521
+    },
+    {
+      "epoch": 2.3523956723338486,
+      "grad_norm": 0.3992741406688027,
+      "learning_rate": 5.097338322637786e-05,
+      "loss": 0.39,
+      "step": 1522
+    },
+    {
+      "epoch": 2.3539412673879445,
+      "grad_norm": 0.3236551759909285,
+      "learning_rate": 5.093186446660159e-05,
+      "loss": 0.3868,
+      "step": 1523
+    },
+    {
+      "epoch": 2.3554868624420404,
+      "grad_norm": 0.25157713529976583,
+      "learning_rate": 5.0890332974447117e-05,
+      "loss": 0.3962,
+      "step": 1524
+    },
+    {
+      "epoch": 2.357032457496136,
+      "grad_norm": 0.27996376356540936,
+      "learning_rate": 5.084878879828627e-05,
+      "loss": 0.3941,
+      "step": 1525
+    },
+    {
+      "epoch": 2.358578052550232,
+      "grad_norm": 0.2992424334167224,
+      "learning_rate": 5.080723198650571e-05,
+      "loss": 0.3969,
+      "step": 1526
+    },
+    {
+      "epoch": 2.3601236476043277,
+      "grad_norm": 0.30496741070379774,
+      "learning_rate": 5.07656625875068e-05,
+      "loss": 0.3882,
+      "step": 1527
+    },
+    {
+      "epoch": 2.3616692426584236,
+      "grad_norm": 0.3024436138778408,
+      "learning_rate": 5.0724080649705526e-05,
+      "loss": 0.3963,
+      "step": 1528
+    },
+    {
+      "epoch": 2.3632148377125195,
+      "grad_norm": 0.32770408845151705,
+      "learning_rate": 5.068248622153252e-05,
+      "loss": 0.3842,
+      "step": 1529
+    },
+    {
+      "epoch": 2.364760432766615,
+      "grad_norm": 0.34329092443466325,
+      "learning_rate": 5.0640879351432934e-05,
+      "loss": 0.3972,
+      "step": 1530
+    },
+    {
+      "epoch": 2.366306027820711,
+      "grad_norm": 0.28538432097026656,
+      "learning_rate": 5.059926008786648e-05,
+      "loss": 0.3848,
+      "step": 1531
+    },
+    {
+      "epoch": 2.3678516228748068,
+      "grad_norm": 0.2351912104958466,
+      "learning_rate": 5.055762847930719e-05,
+      "loss": 0.3931,
+      "step": 1532
+    },
+    {
+      "epoch": 2.3693972179289027,
+      "grad_norm": 0.2675290659579395,
+      "learning_rate": 5.051598457424355e-05,
+      "loss": 0.4034,
+      "step": 1533
+    },
+    {
+      "epoch": 2.3709428129829986,
+      "grad_norm": 0.2450247585979996,
+      "learning_rate": 5.047432842117836e-05,
+      "loss": 0.3845,
+      "step": 1534
+    },
+    {
+      "epoch": 2.3724884080370945,
+      "grad_norm": 0.24341720101474015,
+      "learning_rate": 5.043266006862869e-05,
+      "loss": 0.3928,
+      "step": 1535
+    },
+    {
+      "epoch": 2.37403400309119,
+      "grad_norm": 0.2913711114752356,
+      "learning_rate": 5.039097956512577e-05,
+      "loss": 0.3888,
+      "step": 1536
+    },
+    {
+      "epoch": 2.375579598145286,
+      "grad_norm": 0.3530909123361894,
+      "learning_rate": 5.034928695921506e-05,
+      "loss": 0.3904,
+      "step": 1537
+    },
+    {
+      "epoch": 2.3771251931993818,
+      "grad_norm": 0.3945139153800936,
+      "learning_rate": 5.030758229945605e-05,
+      "loss": 0.3854,
+      "step": 1538
+    },
+    {
+      "epoch": 2.3786707882534777,
+      "grad_norm": 0.39017734529459125,
+      "learning_rate": 5.0265865634422284e-05,
+      "loss": 0.393,
+      "step": 1539
+    },
+    {
+      "epoch": 2.3802163833075736,
+      "grad_norm": 0.305288369267672,
+      "learning_rate": 5.0224137012701314e-05,
+      "loss": 0.3962,
+      "step": 1540
+    },
+    {
+      "epoch": 2.381761978361669,
+      "grad_norm": 0.3139832140402103,
+      "learning_rate": 5.018239648289461e-05,
+      "loss": 0.4092,
+      "step": 1541
+    },
+    {
+      "epoch": 2.383307573415765,
+      "grad_norm": 0.3163870471362883,
+      "learning_rate": 5.014064409361749e-05,
+      "loss": 0.3893,
+      "step": 1542
+    },
+    {
+      "epoch": 2.384853168469861,
+      "grad_norm": 0.30771143522996813,
+      "learning_rate": 5.009887989349909e-05,
+      "loss": 0.3893,
+      "step": 1543
+    },
+    {
+      "epoch": 2.3863987635239567,
+      "grad_norm": 0.304189291631867,
+      "learning_rate": 5.0057103931182324e-05,
+      "loss": 0.3891,
+      "step": 1544
+    },
+    {
+      "epoch": 2.3879443585780527,
+      "grad_norm": 0.229975198322786,
+      "learning_rate": 5.0015316255323803e-05,
+      "loss": 0.3998,
+      "step": 1545
+    },
+    {
+      "epoch": 2.3894899536321486,
+      "grad_norm": 0.28369565359041216,
+      "learning_rate": 4.997351691459374e-05,
+      "loss": 0.4022,
+      "step": 1546
+    },
+    {
+      "epoch": 2.391035548686244,
+      "grad_norm": 0.3174888594737512,
+      "learning_rate": 4.993170595767598e-05,
+      "loss": 0.396,
+      "step": 1547
+    },
+    {
+      "epoch": 2.39258114374034,
+      "grad_norm": 0.2909324953054256,
+      "learning_rate": 4.988988343326788e-05,
+      "loss": 0.3893,
+      "step": 1548
+    },
+    {
+      "epoch": 2.394126738794436,
+      "grad_norm": 0.24681751451704773,
+      "learning_rate": 4.984804939008027e-05,
+      "loss": 0.3949,
+      "step": 1549
+    },
+    {
+      "epoch": 2.3956723338485317,
+      "grad_norm": 0.3624724434626657,
+      "learning_rate": 4.9806203876837386e-05,
+      "loss": 0.3933,
+      "step": 1550
+    },
+    {
+      "epoch": 2.3972179289026276,
+      "grad_norm": 0.248485801335781,
+      "learning_rate": 4.976434694227687e-05,
+      "loss": 0.3993,
+      "step": 1551
+    },
+    {
+      "epoch": 2.398763523956723,
+      "grad_norm": 0.2535610648412525,
+      "learning_rate": 4.972247863514959e-05,
+      "loss": 0.394,
+      "step": 1552
+    },
+    {
+      "epoch": 2.400309119010819,
+      "grad_norm": 0.24641241219834106,
+      "learning_rate": 4.968059900421971e-05,
+      "loss": 0.3857,
+      "step": 1553
+    },
+    {
+      "epoch": 2.401854714064915,
+      "grad_norm": 0.16300375810319317,
+      "learning_rate": 4.963870809826458e-05,
+      "loss": 0.386,
+      "step": 1554
+    },
+    {
+      "epoch": 2.403400309119011,
+      "grad_norm": 0.19116905649294372,
+      "learning_rate": 4.959680596607467e-05,
+      "loss": 0.3924,
+      "step": 1555
+    },
+    {
+      "epoch": 2.4049459041731067,
+      "grad_norm": 0.23101898261927714,
+      "learning_rate": 4.955489265645355e-05,
+      "loss": 0.3986,
+      "step": 1556
+    },
+    {
+      "epoch": 2.4064914992272026,
+      "grad_norm": 0.2654842304461216,
+      "learning_rate": 4.951296821821775e-05,
+      "loss": 0.3832,
+      "step": 1557
+    },
+    {
+      "epoch": 2.4080370942812985,
+      "grad_norm": 0.24896210454474568,
+      "learning_rate": 4.947103270019682e-05,
+      "loss": 0.3985,
+      "step": 1558
+    },
+    {
+      "epoch": 2.409582689335394,
+      "grad_norm": 0.22508988746256473,
+      "learning_rate": 4.942908615123322e-05,
+      "loss": 0.3877,
+      "step": 1559
+    },
+    {
+      "epoch": 2.41112828438949,
+      "grad_norm": 0.20552555820521468,
+      "learning_rate": 4.938712862018218e-05,
+      "loss": 0.3911,
+      "step": 1560
+    },
+    {
+      "epoch": 2.412673879443586,
+      "grad_norm": 0.20543558148018728,
+      "learning_rate": 4.93451601559118e-05,
+      "loss": 0.3902,
+      "step": 1561
+    },
+    {
+      "epoch": 2.4142194744976817,
+      "grad_norm": 0.21293359127356493,
+      "learning_rate": 4.930318080730292e-05,
+      "loss": 0.3934,
+      "step": 1562
+    },
+    {
+      "epoch": 2.4157650695517776,
+      "grad_norm": 0.22528425535355565,
+      "learning_rate": 4.926119062324898e-05,
+      "loss": 0.3886,
+      "step": 1563
+    },
+    {
+      "epoch": 2.417310664605873,
+      "grad_norm": 0.2740527567402077,
+      "learning_rate": 4.921918965265608e-05,
+      "loss": 0.3922,
+      "step": 1564
+    },
+    {
+      "epoch": 2.418856259659969,
+      "grad_norm": 0.222753478064478,
+      "learning_rate": 4.917717794444292e-05,
+      "loss": 0.3864,
+      "step": 1565
+    },
+    {
+      "epoch": 2.420401854714065,
+      "grad_norm": 0.2317187265658363,
+      "learning_rate": 4.913515554754066e-05,
+      "loss": 0.3954,
+      "step": 1566
+    },
+    {
+      "epoch": 2.421947449768161,
+      "grad_norm": 0.23958271070659026,
+      "learning_rate": 4.9093122510892916e-05,
+      "loss": 0.3967,
+      "step": 1567
+    },
+    {
+      "epoch": 2.4234930448222567,
+      "grad_norm": 0.24081649181744408,
+      "learning_rate": 4.905107888345571e-05,
+      "loss": 0.3885,
+      "step": 1568
+    },
+    {
+      "epoch": 2.4250386398763526,
+      "grad_norm": 0.28559014304553987,
+      "learning_rate": 4.900902471419739e-05,
+      "loss": 0.3812,
+      "step": 1569
+    },
+    {
+      "epoch": 2.426584234930448,
+      "grad_norm": 0.2706678308636684,
+      "learning_rate": 4.896696005209858e-05,
+      "loss": 0.3955,
+      "step": 1570
+    },
+    {
+      "epoch": 2.428129829984544,
+      "grad_norm": 0.269485650400652,
+      "learning_rate": 4.8924884946152146e-05,
+      "loss": 0.3817,
+      "step": 1571
+    },
+    {
+      "epoch": 2.42967542503864,
+      "grad_norm": 0.2657977637708371,
+      "learning_rate": 4.888279944536307e-05,
+      "loss": 0.3978,
+      "step": 1572
+    },
+    {
+      "epoch": 2.4312210200927358,
+      "grad_norm": 0.27799517604598306,
+      "learning_rate": 4.884070359874852e-05,
+      "loss": 0.3939,
+      "step": 1573
+    },
+    {
+      "epoch": 2.4327666151468317,
+      "grad_norm": 0.2675807845283423,
+      "learning_rate": 4.879859745533761e-05,
+      "loss": 0.3952,
+      "step": 1574
+    },
+    {
+      "epoch": 2.434312210200927,
+      "grad_norm": 0.2086648242083336,
+      "learning_rate": 4.875648106417153e-05,
+      "loss": 0.3973,
+      "step": 1575
+    },
+    {
+      "epoch": 2.435857805255023,
+      "grad_norm": 0.21399432663226547,
+      "learning_rate": 4.871435447430339e-05,
+      "loss": 0.3939,
+      "step": 1576
+    },
+    {
+      "epoch": 2.437403400309119,
+      "grad_norm": 0.25115123792120214,
+      "learning_rate": 4.867221773479815e-05,
+      "loss": 0.3817,
+      "step": 1577
+    },
+    {
+      "epoch": 2.438948995363215,
+      "grad_norm": 0.2668161649482835,
+      "learning_rate": 4.863007089473262e-05,
+      "loss": 0.389,
+      "step": 1578
+    },
+    {
+      "epoch": 2.4404945904173108,
+      "grad_norm": 0.2502869198766568,
+      "learning_rate": 4.858791400319534e-05,
+      "loss": 0.3836,
+      "step": 1579
+    },
+    {
+      "epoch": 2.4420401854714067,
+      "grad_norm": 0.19761123703119207,
+      "learning_rate": 4.854574710928662e-05,
+      "loss": 0.3918,
+      "step": 1580
+    },
+    {
+      "epoch": 2.443585780525502,
+      "grad_norm": 0.20489041079125325,
+      "learning_rate": 4.8503570262118336e-05,
+      "loss": 0.388,
+      "step": 1581
+    },
+    {
+      "epoch": 2.445131375579598,
+      "grad_norm": 0.23761986430786358,
+      "learning_rate": 4.8461383510814e-05,
+      "loss": 0.3989,
+      "step": 1582
+    },
+    {
+      "epoch": 2.446676970633694,
+      "grad_norm": 0.26799255466712524,
+      "learning_rate": 4.841918690450871e-05,
+      "loss": 0.3953,
+      "step": 1583
+    },
+    {
+      "epoch": 2.44822256568779,
+      "grad_norm": 0.22044145729388834,
+      "learning_rate": 4.837698049234893e-05,
+      "loss": 0.3918,
+      "step": 1584
+    },
+    {
+      "epoch": 2.4497681607418857,
+      "grad_norm": 0.19638779708616366,
+      "learning_rate": 4.833476432349264e-05,
+      "loss": 0.3941,
+      "step": 1585
+    },
+    {
+      "epoch": 2.451313755795981,
+      "grad_norm": 0.20539014072105105,
+      "learning_rate": 4.829253844710914e-05,
+      "loss": 0.3895,
+      "step": 1586
+    },
+    {
+      "epoch": 2.452859350850077,
+      "grad_norm": 0.2615520342957774,
+      "learning_rate": 4.825030291237906e-05,
+      "loss": 0.3857,
+      "step": 1587
+    },
+    {
+      "epoch": 2.454404945904173,
+      "grad_norm": 0.2353556811492111,
+      "learning_rate": 4.820805776849424e-05,
+      "loss": 0.3931,
+      "step": 1588
+    },
+    {
+      "epoch": 2.455950540958269,
+      "grad_norm": 0.26594630149950554,
+      "learning_rate": 4.816580306465775e-05,
+      "loss": 0.4015,
+      "step": 1589
+    },
+    {
+      "epoch": 2.457496136012365,
+      "grad_norm": 0.2847862158071832,
+      "learning_rate": 4.8123538850083796e-05,
+      "loss": 0.3939,
+      "step": 1590
+    },
+    {
+      "epoch": 2.4590417310664607,
+      "grad_norm": 0.2325110176480486,
+      "learning_rate": 4.8081265173997624e-05,
+      "loss": 0.387,
+      "step": 1591
+    },
+    {
+      "epoch": 2.4605873261205566,
+      "grad_norm": 0.2605313564717108,
+      "learning_rate": 4.803898208563552e-05,
+      "loss": 0.3938,
+      "step": 1592
+    },
+    {
+      "epoch": 2.462132921174652,
+      "grad_norm": 0.251069170835732,
+      "learning_rate": 4.799668963424476e-05,
+      "loss": 0.4004,
+      "step": 1593
+    },
+    {
+      "epoch": 2.463678516228748,
+      "grad_norm": 0.1985495939966303,
+      "learning_rate": 4.7954387869083486e-05,
+      "loss": 0.3906,
+      "step": 1594
+    },
+    {
+      "epoch": 2.465224111282844,
+      "grad_norm": 0.24250392489168981,
+      "learning_rate": 4.7912076839420695e-05,
+      "loss": 0.3962,
+      "step": 1595
+    },
+    {
+      "epoch": 2.46676970633694,
+      "grad_norm": 0.24210751742714964,
+      "learning_rate": 4.786975659453619e-05,
+      "loss": 0.3926,
+      "step": 1596
+    },
+    {
+      "epoch": 2.4683153013910357,
+      "grad_norm": 0.2516813510196297,
+      "learning_rate": 4.78274271837205e-05,
+      "loss": 0.3971,
+      "step": 1597
+    },
+    {
+      "epoch": 2.469860896445131,
+      "grad_norm": 0.231444499963526,
+      "learning_rate": 4.7785088656274824e-05,
+      "loss": 0.3922,
+      "step": 1598
+    },
+    {
+      "epoch": 2.471406491499227,
+      "grad_norm": 0.1858894539903597,
+      "learning_rate": 4.7742741061510975e-05,
+      "loss": 0.394,
+      "step": 1599
+    },
+    {
+      "epoch": 2.472952086553323,
+      "grad_norm": 0.2753734441818444,
+      "learning_rate": 4.770038444875136e-05,
+      "loss": 0.3913,
+      "step": 1600
+    },
+    {
+      "epoch": 2.474497681607419,
+      "grad_norm": 0.2530319282868681,
+      "learning_rate": 4.765801886732884e-05,
+      "loss": 0.4057,
+      "step": 1601
+    },
+    {
+      "epoch": 2.476043276661515,
+      "grad_norm": 0.24971066262343666,
+      "learning_rate": 4.761564436658674e-05,
+      "loss": 0.3892,
+      "step": 1602
+    },
+    {
+      "epoch": 2.4775888717156107,
+      "grad_norm": 0.21172394410295473,
+      "learning_rate": 4.75732609958788e-05,
+      "loss": 0.4007,
+      "step": 1603
+    },
+    {
+      "epoch": 2.479134466769706,
+      "grad_norm": 0.21123359781475406,
+      "learning_rate": 4.753086880456907e-05,
+      "loss": 0.3986,
+      "step": 1604
+    },
+    {
+      "epoch": 2.480680061823802,
+      "grad_norm": 0.21942737233497195,
+      "learning_rate": 4.748846784203187e-05,
+      "loss": 0.4015,
+      "step": 1605
+    },
+    {
+      "epoch": 2.482225656877898,
+      "grad_norm": 0.30372087235313455,
+      "learning_rate": 4.7446058157651715e-05,
+      "loss": 0.3889,
+      "step": 1606
+    },
+    {
+      "epoch": 2.483771251931994,
+      "grad_norm": 0.20412899046160612,
+      "learning_rate": 4.7403639800823344e-05,
+      "loss": 0.3865,
+      "step": 1607
+    },
+    {
+      "epoch": 2.48531684698609,
+      "grad_norm": 0.2193097389895047,
+      "learning_rate": 4.7361212820951525e-05,
+      "loss": 0.3887,
+      "step": 1608
+    },
+    {
+      "epoch": 2.4868624420401853,
+      "grad_norm": 0.2721648103907066,
+      "learning_rate": 4.731877726745112e-05,
+      "loss": 0.4021,
+      "step": 1609
+    },
+    {
+      "epoch": 2.488408037094281,
+      "grad_norm": 0.22312355301605005,
+      "learning_rate": 4.7276333189746944e-05,
+      "loss": 0.3939,
+      "step": 1610
+    },
+    {
+      "epoch": 2.489953632148377,
+      "grad_norm": 0.22679494324181823,
+      "learning_rate": 4.723388063727375e-05,
+      "loss": 0.3909,
+      "step": 1611
+    },
+    {
+      "epoch": 2.491499227202473,
+      "grad_norm": 0.22064137384238353,
+      "learning_rate": 4.719141965947618e-05,
+      "loss": 0.3897,
+      "step": 1612
+    },
+    {
+      "epoch": 2.493044822256569,
+      "grad_norm": 0.24253869301521516,
+      "learning_rate": 4.714895030580865e-05,
+      "loss": 0.3903,
+      "step": 1613
+    },
+    {
+      "epoch": 2.4945904173106648,
+      "grad_norm": 0.23241390259659883,
+      "learning_rate": 4.7106472625735375e-05,
+      "loss": 0.3929,
+      "step": 1614
+    },
+    {
+      "epoch": 2.4961360123647607,
+      "grad_norm": 0.2771522090480515,
+      "learning_rate": 4.7063986668730246e-05,
+      "loss": 0.3861,
+      "step": 1615
+    },
+    {
+      "epoch": 2.497681607418856,
+      "grad_norm": 0.3106756875403178,
+      "learning_rate": 4.702149248427677e-05,
+      "loss": 0.3895,
+      "step": 1616
+    },
+    {
+      "epoch": 2.499227202472952,
+      "grad_norm": 0.34350424156535503,
+      "learning_rate": 4.6978990121868096e-05,
+      "loss": 0.3862,
+      "step": 1617
+    },
+    {
+      "epoch": 2.500772797527048,
+      "grad_norm": 0.3223694433694756,
+      "learning_rate": 4.6936479631006873e-05,
+      "loss": 0.3907,
+      "step": 1618
+    },
+    {
+      "epoch": 2.502318392581144,
+      "grad_norm": 0.2726783290551887,
+      "learning_rate": 4.689396106120516e-05,
+      "loss": 0.3985,
+      "step": 1619
+    },
+    {
+      "epoch": 2.5038639876352393,
+      "grad_norm": 0.23225676032680767,
+      "learning_rate": 4.6851434461984505e-05,
+      "loss": 0.3953,
+      "step": 1620
+    },
+    {
+      "epoch": 2.5054095826893352,
+      "grad_norm": 0.32044583909189767,
+      "learning_rate": 4.680889988287578e-05,
+      "loss": 0.39,
+      "step": 1621
+    },
+    {
+      "epoch": 2.506955177743431,
+      "grad_norm": 0.28909449180223523,
+      "learning_rate": 4.676635737341913e-05,
+      "loss": 0.3915,
+      "step": 1622
+    },
+    {
+      "epoch": 2.508500772797527,
+      "grad_norm": 0.3373093849569923,
+      "learning_rate": 4.672380698316398e-05,
+      "loss": 0.3957,
+      "step": 1623
+    },
+    {
+      "epoch": 2.510046367851623,
+      "grad_norm": 0.2934474440801747,
+      "learning_rate": 4.6681248761668886e-05,
+      "loss": 0.3944,
+      "step": 1624
+    },
+    {
+      "epoch": 2.511591962905719,
+      "grad_norm": 0.20853548303800792,
+      "learning_rate": 4.6638682758501553e-05,
+      "loss": 0.3867,
+      "step": 1625
+    },
+    {
+      "epoch": 2.5131375579598147,
+      "grad_norm": 0.20643856276794834,
+      "learning_rate": 4.659610902323875e-05,
+      "loss": 0.3841,
+      "step": 1626
+    },
+    {
+      "epoch": 2.51468315301391,
+      "grad_norm": 0.26311445438533454,
+      "learning_rate": 4.655352760546624e-05,
+      "loss": 0.3925,
+      "step": 1627
+    },
+    {
+      "epoch": 2.516228748068006,
+      "grad_norm": 0.2668275480704203,
+      "learning_rate": 4.651093855477874e-05,
+      "loss": 0.3919,
+      "step": 1628
+    },
+    {
+      "epoch": 2.517774343122102,
+      "grad_norm": 0.2694379532597783,
+      "learning_rate": 4.646834192077986e-05,
+      "loss": 0.3899,
+      "step": 1629
+    },
+    {
+      "epoch": 2.519319938176198,
+      "grad_norm": 0.26228893960462624,
+      "learning_rate": 4.6425737753082024e-05,
+      "loss": 0.3928,
+      "step": 1630
+    },
+    {
+      "epoch": 2.5208655332302934,
+      "grad_norm": 0.1936290956124116,
+      "learning_rate": 4.638312610130646e-05,
+      "loss": 0.3941,
+      "step": 1631
+    },
+    {
+      "epoch": 2.5224111282843893,
+      "grad_norm": 0.21327089164691612,
+      "learning_rate": 4.6340507015083116e-05,
+      "loss": 0.3924,
+      "step": 1632
+    },
+    {
+      "epoch": 2.523956723338485,
+      "grad_norm": 0.24903276404302455,
+      "learning_rate": 4.629788054405054e-05,
+      "loss": 0.3955,
+      "step": 1633
+    },
+    {
+      "epoch": 2.525502318392581,
+      "grad_norm": 0.25245491821379856,
+      "learning_rate": 4.625524673785593e-05,
+      "loss": 0.3829,
+      "step": 1634
+    },
+    {
+      "epoch": 2.527047913446677,
+      "grad_norm": 0.2844301896804582,
+      "learning_rate": 4.621260564615505e-05,
+      "loss": 0.3859,
+      "step": 1635
+    },
+    {
+      "epoch": 2.528593508500773,
+      "grad_norm": 0.21833037149589407,
+      "learning_rate": 4.6169957318612096e-05,
+      "loss": 0.3898,
+      "step": 1636
+    },
+    {
+      "epoch": 2.530139103554869,
+      "grad_norm": 0.19449687548089142,
+      "learning_rate": 4.6127301804899717e-05,
+      "loss": 0.3936,
+      "step": 1637
+    },
+    {
+      "epoch": 2.5316846986089647,
+      "grad_norm": 0.18782982149635674,
+      "learning_rate": 4.608463915469893e-05,
+      "loss": 0.3954,
+      "step": 1638
+    },
+    {
+      "epoch": 2.53323029366306,
+      "grad_norm": 0.2159678201262451,
+      "learning_rate": 4.604196941769906e-05,
+      "loss": 0.3918,
+      "step": 1639
+    },
+    {
+      "epoch": 2.534775888717156,
+      "grad_norm": 0.2563281079039888,
+      "learning_rate": 4.5999292643597693e-05,
+      "loss": 0.3836,
+      "step": 1640
+    },
+    {
+      "epoch": 2.536321483771252,
+      "grad_norm": 0.3491491507041857,
+      "learning_rate": 4.59566088821006e-05,
+      "loss": 0.3939,
+      "step": 1641
+    },
+    {
+      "epoch": 2.537867078825348,
+      "grad_norm": 0.25422624748850614,
+      "learning_rate": 4.591391818292171e-05,
+      "loss": 0.3817,
+      "step": 1642
+    },
+    {
+      "epoch": 2.5394126738794434,
+      "grad_norm": 0.1737983690973426,
+      "learning_rate": 4.5871220595783016e-05,
+      "loss": 0.388,
+      "step": 1643
+    },
+    {
+      "epoch": 2.5409582689335393,
+      "grad_norm": 0.2024156139334753,
+      "learning_rate": 4.582851617041452e-05,
+      "loss": 0.3919,
+      "step": 1644
+    },
+    {
+      "epoch": 2.542503863987635,
+      "grad_norm": 0.202932388910506,
+      "learning_rate": 4.578580495655423e-05,
+      "loss": 0.3953,
+      "step": 1645
+    },
+    {
+      "epoch": 2.544049459041731,
+      "grad_norm": 0.243242339398608,
+      "learning_rate": 4.5743087003948024e-05,
+      "loss": 0.3933,
+      "step": 1646
+    },
+    {
+      "epoch": 2.545595054095827,
+      "grad_norm": 0.24488570408288615,
+      "learning_rate": 4.570036236234963e-05,
+      "loss": 0.3927,
+      "step": 1647
+    },
+    {
+      "epoch": 2.547140649149923,
+      "grad_norm": 0.2685531776215014,
+      "learning_rate": 4.56576310815206e-05,
+      "loss": 0.3909,
+      "step": 1648
+    },
+    {
+      "epoch": 2.548686244204019,
+      "grad_norm": 0.2728965982324487,
+      "learning_rate": 4.561489321123018e-05,
+      "loss": 0.3876,
+      "step": 1649
+    },
+    {
+      "epoch": 2.5502318392581143,
+      "grad_norm": 0.2253682041586724,
+      "learning_rate": 4.557214880125533e-05,
+      "loss": 0.3976,
+      "step": 1650
+    },
+    {
+      "epoch": 2.55177743431221,
+      "grad_norm": 0.20483244888828098,
+      "learning_rate": 4.552939790138058e-05,
+      "loss": 0.3928,
+      "step": 1651
+    },
+    {
+      "epoch": 2.553323029366306,
+      "grad_norm": 0.2235110600526468,
+      "learning_rate": 4.548664056139803e-05,
+      "loss": 0.3915,
+      "step": 1652
+    },
+    {
+      "epoch": 2.554868624420402,
+      "grad_norm": 0.25680000871236164,
+      "learning_rate": 4.544387683110733e-05,
+      "loss": 0.4003,
+      "step": 1653
+    },
+    {
+      "epoch": 2.5564142194744974,
+      "grad_norm": 0.3330721283056668,
+      "learning_rate": 4.540110676031552e-05,
+      "loss": 0.3862,
+      "step": 1654
+    },
+    {
+      "epoch": 2.5579598145285933,
+      "grad_norm": 0.3367498008963851,
+      "learning_rate": 4.5358330398837035e-05,
+      "loss": 0.3884,
+      "step": 1655
+    },
+    {
+      "epoch": 2.5595054095826892,
+      "grad_norm": 0.285531764031909,
+      "learning_rate": 4.531554779649365e-05,
+      "loss": 0.3902,
+      "step": 1656
+    },
+    {
+      "epoch": 2.561051004636785,
+      "grad_norm": 0.25585937876834175,
+      "learning_rate": 4.52727590031144e-05,
+      "loss": 0.397,
+      "step": 1657
+    },
+    {
+      "epoch": 2.562596599690881,
+      "grad_norm": 0.2765152500473313,
+      "learning_rate": 4.522996406853552e-05,
+      "loss": 0.3977,
+      "step": 1658
+    },
+    {
+      "epoch": 2.564142194744977,
+      "grad_norm": 0.2121356374509123,
+      "learning_rate": 4.5187163042600416e-05,
+      "loss": 0.3953,
+      "step": 1659
+    },
+    {
+      "epoch": 2.565687789799073,
+      "grad_norm": 0.23605796283812225,
+      "learning_rate": 4.514435597515962e-05,
+      "loss": 0.3942,
+      "step": 1660
+    },
+    {
+      "epoch": 2.5672333848531683,
+      "grad_norm": 0.280273284639065,
+      "learning_rate": 4.510154291607061e-05,
+      "loss": 0.3974,
+      "step": 1661
+    },
+    {
+      "epoch": 2.5687789799072642,
+      "grad_norm": 0.2862636237262634,
+      "learning_rate": 4.5058723915197916e-05,
+      "loss": 0.3992,
+      "step": 1662
+    },
+    {
+      "epoch": 2.57032457496136,
+      "grad_norm": 0.20355355298012318,
+      "learning_rate": 4.501589902241298e-05,
+      "loss": 0.3969,
+      "step": 1663
+    },
+    {
+      "epoch": 2.571870170015456,
+      "grad_norm": 0.22242162610416033,
+      "learning_rate": 4.497306828759407e-05,
+      "loss": 0.3834,
+      "step": 1664
+    },
+    {
+      "epoch": 2.573415765069552,
+      "grad_norm": 0.28927596076059225,
+      "learning_rate": 4.493023176062631e-05,
+      "loss": 0.3973,
+      "step": 1665
+    },
+    {
+      "epoch": 2.5749613601236474,
+      "grad_norm": 0.2620957900323579,
+      "learning_rate": 4.488738949140151e-05,
+      "loss": 0.3883,
+      "step": 1666
+    },
+    {
+      "epoch": 2.5765069551777433,
+      "grad_norm": 0.2039195819917358,
+      "learning_rate": 4.4844541529818234e-05,
+      "loss": 0.3883,
+      "step": 1667
+    },
+    {
+      "epoch": 2.578052550231839,
+      "grad_norm": 0.23742806986642045,
+      "learning_rate": 4.480168792578163e-05,
+      "loss": 0.3871,
+      "step": 1668
+    },
+    {
+      "epoch": 2.579598145285935,
+      "grad_norm": 0.23110172448224695,
+      "learning_rate": 4.475882872920341e-05,
+      "loss": 0.3844,
+      "step": 1669
+    },
+    {
+      "epoch": 2.581143740340031,
+      "grad_norm": 0.27447312756803394,
+      "learning_rate": 4.471596399000186e-05,
+      "loss": 0.3908,
+      "step": 1670
+    },
+    {
+      "epoch": 2.582689335394127,
+      "grad_norm": 0.23463566543332323,
+      "learning_rate": 4.4673093758101655e-05,
+      "loss": 0.3887,
+      "step": 1671
+    },
+    {
+      "epoch": 2.584234930448223,
+      "grad_norm": 0.20136632940367377,
+      "learning_rate": 4.463021808343389e-05,
+      "loss": 0.3942,
+      "step": 1672
+    },
+    {
+      "epoch": 2.5857805255023183,
+      "grad_norm": 0.18778083320382974,
+      "learning_rate": 4.458733701593603e-05,
+      "loss": 0.3932,
+      "step": 1673
+    },
+    {
+      "epoch": 2.587326120556414,
+      "grad_norm": 0.19782531967338815,
+      "learning_rate": 4.454445060555176e-05,
+      "loss": 0.3945,
+      "step": 1674
+    },
+    {
+      "epoch": 2.58887171561051,
+      "grad_norm": 0.22508762999604104,
+      "learning_rate": 4.450155890223104e-05,
+      "loss": 0.3873,
+      "step": 1675
+    },
+    {
+      "epoch": 2.590417310664606,
+      "grad_norm": 0.21168664081062777,
+      "learning_rate": 4.445866195592998e-05,
+      "loss": 0.3926,
+      "step": 1676
+    },
+    {
+      "epoch": 2.5919629057187015,
+      "grad_norm": 0.16234233057601102,
+      "learning_rate": 4.44157598166108e-05,
+      "loss": 0.3957,
+      "step": 1677
+    },
+    {
+      "epoch": 2.5935085007727974,
+      "grad_norm": 0.2006506183967543,
+      "learning_rate": 4.437285253424174e-05,
+      "loss": 0.3987,
+      "step": 1678
+    },
+    {
+      "epoch": 2.5950540958268933,
+      "grad_norm": 0.19266544117512432,
+      "learning_rate": 4.432994015879707e-05,
+      "loss": 0.3885,
+      "step": 1679
+    },
+    {
+      "epoch": 2.596599690880989,
+      "grad_norm": 0.23946286724356225,
+      "learning_rate": 4.428702274025694e-05,
+      "loss": 0.402,
+      "step": 1680
+    },
+    {
+      "epoch": 2.598145285935085,
+      "grad_norm": 0.26911538712111643,
+      "learning_rate": 4.424410032860747e-05,
+      "loss": 0.3854,
+      "step": 1681
+    },
+    {
+      "epoch": 2.599690880989181,
+      "grad_norm": 0.24576664996872866,
+      "learning_rate": 4.420117297384048e-05,
+      "loss": 0.3955,
+      "step": 1682
+    },
+    {
+      "epoch": 2.601236476043277,
+      "grad_norm": 0.20026359969497476,
+      "learning_rate": 4.415824072595361e-05,
+      "loss": 0.3966,
+      "step": 1683
+    },
+    {
+      "epoch": 2.6027820710973724,
+      "grad_norm": 0.24261706436109684,
+      "learning_rate": 4.411530363495021e-05,
+      "loss": 0.3913,
+      "step": 1684
+    },
+    {
+      "epoch": 2.6043276661514683,
+      "grad_norm": 0.5551795310711124,
+      "learning_rate": 4.407236175083923e-05,
+      "loss": 0.3955,
+      "step": 1685
+    },
+    {
+      "epoch": 2.605873261205564,
+      "grad_norm": 0.29021409934929493,
+      "learning_rate": 4.402941512363522e-05,
+      "loss": 0.3961,
+      "step": 1686
+    },
+    {
+      "epoch": 2.60741885625966,
+      "grad_norm": 0.25950368123392215,
+      "learning_rate": 4.3986463803358284e-05,
+      "loss": 0.3964,
+      "step": 1687
+    },
+    {
+      "epoch": 2.6089644513137555,
+      "grad_norm": 0.24088722690149555,
+      "learning_rate": 4.394350784003394e-05,
+      "loss": 0.3969,
+      "step": 1688
+    },
+    {
+      "epoch": 2.6105100463678514,
+      "grad_norm": 0.2324338024413383,
+      "learning_rate": 4.3900547283693126e-05,
+      "loss": 0.379,
+      "step": 1689
+    },
+    {
+      "epoch": 2.6120556414219473,
+      "grad_norm": 0.21735680121166714,
+      "learning_rate": 4.385758218437218e-05,
+      "loss": 0.4003,
+      "step": 1690
+    },
+    {
+      "epoch": 2.6136012364760433,
+      "grad_norm": 0.18670609149054954,
+      "learning_rate": 4.3814612592112675e-05,
+      "loss": 0.3933,
+      "step": 1691
+    },
+    {
+      "epoch": 2.615146831530139,
+      "grad_norm": 0.23190433118432097,
+      "learning_rate": 4.377163855696143e-05,
+      "loss": 0.3833,
+      "step": 1692
+    },
+    {
+      "epoch": 2.616692426584235,
+      "grad_norm": 0.23098885325708501,
+      "learning_rate": 4.3728660128970454e-05,
+      "loss": 0.3941,
+      "step": 1693
+    },
+    {
+      "epoch": 2.618238021638331,
+      "grad_norm": 0.18041231125937515,
+      "learning_rate": 4.3685677358196855e-05,
+      "loss": 0.3935,
+      "step": 1694
+    },
+    {
+      "epoch": 2.6197836166924264,
+      "grad_norm": 0.18907903531096312,
+      "learning_rate": 4.364269029470281e-05,
+      "loss": 0.3927,
+      "step": 1695
+    },
+    {
+      "epoch": 2.6213292117465223,
+      "grad_norm": 0.21197360112624988,
+      "learning_rate": 4.359969898855549e-05,
+      "loss": 0.3862,
+      "step": 1696
+    },
+    {
+      "epoch": 2.6228748068006182,
+      "grad_norm": 0.2415173234059811,
+      "learning_rate": 4.355670348982701e-05,
+      "loss": 0.3972,
+      "step": 1697
+    },
+    {
+      "epoch": 2.624420401854714,
+      "grad_norm": 0.2538082855558849,
+      "learning_rate": 4.3513703848594366e-05,
+      "loss": 0.3971,
+      "step": 1698
+    },
+    {
+      "epoch": 2.62596599690881,
+      "grad_norm": 0.21427677280173618,
+      "learning_rate": 4.347070011493939e-05,
+      "loss": 0.3892,
+      "step": 1699
+    },
+    {
+      "epoch": 2.6275115919629055,
+      "grad_norm": 0.27485089107298366,
+      "learning_rate": 4.342769233894865e-05,
+      "loss": 0.3873,
+      "step": 1700
+    },
+    {
+      "epoch": 2.6290571870170014,
+      "grad_norm": 0.2374833907537212,
+      "learning_rate": 4.338468057071347e-05,
+      "loss": 0.4006,
+      "step": 1701
+    },
+    {
+      "epoch": 2.6306027820710973,
+      "grad_norm": 0.22520297589709645,
+      "learning_rate": 4.334166486032977e-05,
+      "loss": 0.3925,
+      "step": 1702
+    },
+    {
+      "epoch": 2.6321483771251932,
+      "grad_norm": 0.3218468827633287,
+      "learning_rate": 4.329864525789811e-05,
+      "loss": 0.3889,
+      "step": 1703
+    },
+    {
+      "epoch": 2.633693972179289,
+      "grad_norm": 0.3019997561853618,
+      "learning_rate": 4.325562181352356e-05,
+      "loss": 0.3912,
+      "step": 1704
+    },
+    {
+      "epoch": 2.635239567233385,
+      "grad_norm": 0.2093153275570432,
+      "learning_rate": 4.321259457731565e-05,
+      "loss": 0.3863,
+      "step": 1705
+    },
+    {
+      "epoch": 2.636785162287481,
+      "grad_norm": 0.20006225048976253,
+      "learning_rate": 4.316956359938835e-05,
+      "loss": 0.3932,
+      "step": 1706
+    },
+    {
+      "epoch": 2.6383307573415764,
+      "grad_norm": 0.22304616722385706,
+      "learning_rate": 4.3126528929859994e-05,
+      "loss": 0.3925,
+      "step": 1707
+    },
+    {
+      "epoch": 2.6398763523956723,
+      "grad_norm": 0.21878151455286357,
+      "learning_rate": 4.3083490618853175e-05,
+      "loss": 0.3902,
+      "step": 1708
+    },
+    {
+      "epoch": 2.641421947449768,
+      "grad_norm": 0.22444441987583222,
+      "learning_rate": 4.3040448716494775e-05,
+      "loss": 0.3961,
+      "step": 1709
+    },
+    {
+      "epoch": 2.642967542503864,
+      "grad_norm": 0.2483241102745166,
+      "learning_rate": 4.2997403272915836e-05,
+      "loss": 0.3934,
+      "step": 1710
+    },
+    {
+      "epoch": 2.6445131375579596,
+      "grad_norm": 0.22959455078957128,
+      "learning_rate": 4.295435433825152e-05,
+      "loss": 0.3824,
+      "step": 1711
+    },
+    {
+      "epoch": 2.6460587326120555,
+      "grad_norm": 0.21699514755570257,
+      "learning_rate": 4.2911301962641065e-05,
+      "loss": 0.3939,
+      "step": 1712
+    },
+    {
+      "epoch": 2.6476043276661514,
+      "grad_norm": 0.270050550781768,
+      "learning_rate": 4.2868246196227715e-05,
+      "loss": 0.3947,
+      "step": 1713
+    },
+    {
+      "epoch": 2.6491499227202473,
+      "grad_norm": 0.2272440301136524,
+      "learning_rate": 4.2825187089158656e-05,
+      "loss": 0.3932,
+      "step": 1714
+    },
+    {
+      "epoch": 2.650695517774343,
+      "grad_norm": 0.20918266050275014,
+      "learning_rate": 4.2782124691584984e-05,
+      "loss": 0.3977,
+      "step": 1715
+    },
+    {
+      "epoch": 2.652241112828439,
+      "grad_norm": 0.19535768556289942,
+      "learning_rate": 4.273905905366159e-05,
+      "loss": 0.3919,
+      "step": 1716
+    },
+    {
+      "epoch": 2.653786707882535,
+      "grad_norm": 0.27365474283090385,
+      "learning_rate": 4.269599022554718e-05,
+      "loss": 0.3978,
+      "step": 1717
+    },
+    {
+      "epoch": 2.6553323029366305,
+      "grad_norm": 0.2862309974660244,
+      "learning_rate": 4.265291825740416e-05,
+      "loss": 0.3865,
+      "step": 1718
+    },
+    {
+      "epoch": 2.6568778979907264,
+      "grad_norm": 0.24978807145564993,
+      "learning_rate": 4.2609843199398594e-05,
+      "loss": 0.3866,
+      "step": 1719
+    },
+    {
+      "epoch": 2.6584234930448223,
+      "grad_norm": 0.18804780266739948,
+      "learning_rate": 4.256676510170013e-05,
+      "loss": 0.3896,
+      "step": 1720
+    },
+    {
+      "epoch": 2.659969088098918,
+      "grad_norm": 0.20114039643341255,
+      "learning_rate": 4.252368401448198e-05,
+      "loss": 0.3877,
+      "step": 1721
+    },
+    {
+      "epoch": 2.6615146831530136,
+      "grad_norm": 0.22568422793760368,
+      "learning_rate": 4.248059998792081e-05,
+      "loss": 0.3953,
+      "step": 1722
+    },
+    {
+      "epoch": 2.6630602782071096,
+      "grad_norm": 0.23913012062540562,
+      "learning_rate": 4.243751307219675e-05,
+      "loss": 0.3958,
+      "step": 1723
+    },
+    {
+      "epoch": 2.6646058732612055,
+      "grad_norm": 0.23926122468307615,
+      "learning_rate": 4.2394423317493256e-05,
+      "loss": 0.3941,
+      "step": 1724
+    },
+    {
+      "epoch": 2.6661514683153014,
+      "grad_norm": 0.26739200174714023,
+      "learning_rate": 4.2351330773997116e-05,
+      "loss": 0.3938,
+      "step": 1725
+    },
+    {
+      "epoch": 2.6676970633693973,
+      "grad_norm": 0.2521231378359259,
+      "learning_rate": 4.230823549189834e-05,
+      "loss": 0.3964,
+      "step": 1726
+    },
+    {
+      "epoch": 2.669242658423493,
+      "grad_norm": 0.21798757466882637,
+      "learning_rate": 4.226513752139016e-05,
+      "loss": 0.3797,
+      "step": 1727
+    },
+    {
+      "epoch": 2.670788253477589,
+      "grad_norm": 0.24296505447418829,
+      "learning_rate": 4.2222036912668915e-05,
+      "loss": 0.3993,
+      "step": 1728
+    },
+    {
+      "epoch": 2.6723338485316845,
+      "grad_norm": 0.275836142582003,
+      "learning_rate": 4.2178933715934027e-05,
+      "loss": 0.3995,
+      "step": 1729
+    },
+    {
+      "epoch": 2.6738794435857804,
+      "grad_norm": 0.23257862603468263,
+      "learning_rate": 4.2135827981387936e-05,
+      "loss": 0.3852,
+      "step": 1730
+    },
+    {
+      "epoch": 2.6754250386398764,
+      "grad_norm": 0.2509912628829087,
+      "learning_rate": 4.2092719759236013e-05,
+      "loss": 0.3926,
+      "step": 1731
+    },
+    {
+      "epoch": 2.6769706336939723,
+      "grad_norm": 0.29179187494496706,
+      "learning_rate": 4.204960909968656e-05,
+      "loss": 0.3911,
+      "step": 1732
+    },
+    {
+      "epoch": 2.678516228748068,
+      "grad_norm": 0.25691820918430236,
+      "learning_rate": 4.200649605295071e-05,
+      "loss": 0.3886,
+      "step": 1733
+    },
+    {
+      "epoch": 2.6800618238021636,
+      "grad_norm": 0.19850809916040388,
+      "learning_rate": 4.1963380669242335e-05,
+      "loss": 0.3859,
+      "step": 1734
+    },
+    {
+      "epoch": 2.6816074188562595,
+      "grad_norm": 0.2584257513915739,
+      "learning_rate": 4.192026299877811e-05,
+      "loss": 0.3816,
+      "step": 1735
+    },
+    {
+      "epoch": 2.6831530139103554,
+      "grad_norm": 0.25829112210072946,
+      "learning_rate": 4.187714309177729e-05,
+      "loss": 0.3903,
+      "step": 1736
+    },
+    {
+      "epoch": 2.6846986089644513,
+      "grad_norm": 0.22790216466337682,
+      "learning_rate": 4.1834020998461785e-05,
+      "loss": 0.3861,
+      "step": 1737
+    },
+    {
+      "epoch": 2.6862442040185472,
+      "grad_norm": 0.2914850077764589,
+      "learning_rate": 4.179089676905605e-05,
+      "loss": 0.394,
+      "step": 1738
+    },
+    {
+      "epoch": 2.687789799072643,
+      "grad_norm": 0.2671371421689647,
+      "learning_rate": 4.174777045378699e-05,
+      "loss": 0.3896,
+      "step": 1739
+    },
+    {
+      "epoch": 2.689335394126739,
+      "grad_norm": 0.19241631176557675,
+      "learning_rate": 4.170464210288399e-05,
+      "loss": 0.3955,
+      "step": 1740
+    },
+    {
+      "epoch": 2.6908809891808345,
+      "grad_norm": 0.2086286450779931,
+      "learning_rate": 4.1661511766578774e-05,
+      "loss": 0.3967,
+      "step": 1741
+    },
+    {
+      "epoch": 2.6924265842349304,
+      "grad_norm": 0.23025192323619,
+      "learning_rate": 4.1618379495105364e-05,
+      "loss": 0.3981,
+      "step": 1742
+    },
+    {
+      "epoch": 2.6939721792890263,
+      "grad_norm": 0.2413035288042119,
+      "learning_rate": 4.157524533870009e-05,
+      "loss": 0.3985,
+      "step": 1743
+    },
+    {
+      "epoch": 2.6955177743431222,
+      "grad_norm": 0.22474391007406155,
+      "learning_rate": 4.153210934760142e-05,
+      "loss": 0.3896,
+      "step": 1744
+    },
+    {
+      "epoch": 2.6970633693972177,
+      "grad_norm": 0.20021352917828603,
+      "learning_rate": 4.148897157204999e-05,
+      "loss": 0.386,
+      "step": 1745
+    },
+    {
+      "epoch": 2.6986089644513136,
+      "grad_norm": 0.26474447032524473,
+      "learning_rate": 4.144583206228852e-05,
+      "loss": 0.3998,
+      "step": 1746
+    },
+    {
+      "epoch": 2.7001545595054095,
+      "grad_norm": 0.2158672684708272,
+      "learning_rate": 4.1402690868561735e-05,
+      "loss": 0.3887,
+      "step": 1747
+    },
+    {
+      "epoch": 2.7017001545595054,
+      "grad_norm": 0.24266533972938326,
+      "learning_rate": 4.13595480411163e-05,
+      "loss": 0.3943,
+      "step": 1748
+    },
+    {
+      "epoch": 2.7032457496136013,
+      "grad_norm": 0.24274619528355063,
+      "learning_rate": 4.1316403630200825e-05,
+      "loss": 0.3789,
+      "step": 1749
+    },
+    {
+      "epoch": 2.704791344667697,
+      "grad_norm": 0.2916946178505382,
+      "learning_rate": 4.127325768606576e-05,
+      "loss": 0.3962,
+      "step": 1750
+    },
+    {
+      "epoch": 2.706336939721793,
+      "grad_norm": 0.2413892655948878,
+      "learning_rate": 4.12301102589633e-05,
+      "loss": 0.3913,
+      "step": 1751
+    },
+    {
+      "epoch": 2.7078825347758886,
+      "grad_norm": 0.1904043535480117,
+      "learning_rate": 4.118696139914742e-05,
+      "loss": 0.3911,
+      "step": 1752
+    },
+    {
+      "epoch": 2.7094281298299845,
+      "grad_norm": 0.22388949005518227,
+      "learning_rate": 4.114381115687371e-05,
+      "loss": 0.3956,
+      "step": 1753
+    },
+    {
+      "epoch": 2.7109737248840804,
+      "grad_norm": 0.17867764017934373,
+      "learning_rate": 4.110065958239942e-05,
+      "loss": 0.3878,
+      "step": 1754
+    },
+    {
+      "epoch": 2.7125193199381763,
+      "grad_norm": 0.20689302795190623,
+      "learning_rate": 4.105750672598333e-05,
+      "loss": 0.4045,
+      "step": 1755
+    },
+    {
+      "epoch": 2.7140649149922718,
+      "grad_norm": 0.23874824045849408,
+      "learning_rate": 4.101435263788569e-05,
+      "loss": 0.3894,
+      "step": 1756
+    },
+    {
+      "epoch": 2.7156105100463677,
+      "grad_norm": 0.26648071941463974,
+      "learning_rate": 4.097119736836823e-05,
+      "loss": 0.3873,
+      "step": 1757
+    },
+    {
+      "epoch": 2.7171561051004636,
+      "grad_norm": 0.21952269929540558,
+      "learning_rate": 4.092804096769404e-05,
+      "loss": 0.3863,
+      "step": 1758
+    },
+    {
+      "epoch": 2.7187017001545595,
+      "grad_norm": 0.24690525236175953,
+      "learning_rate": 4.088488348612748e-05,
+      "loss": 0.3897,
+      "step": 1759
+    },
+    {
+      "epoch": 2.7202472952086554,
+      "grad_norm": 0.17406867030472212,
+      "learning_rate": 4.084172497393425e-05,
+      "loss": 0.3921,
+      "step": 1760
+    },
+    {
+      "epoch": 2.7217928902627513,
+      "grad_norm": 0.18210134548010445,
+      "learning_rate": 4.0798565481381194e-05,
+      "loss": 0.3967,
+      "step": 1761
+    },
+    {
+      "epoch": 2.723338485316847,
+      "grad_norm": 0.21945415295395937,
+      "learning_rate": 4.0755405058736314e-05,
+      "loss": 0.3954,
+      "step": 1762
+    },
+    {
+      "epoch": 2.7248840803709427,
+      "grad_norm": 0.232237741090731,
+      "learning_rate": 4.07122437562687e-05,
+      "loss": 0.3967,
+      "step": 1763
+    },
+    {
+      "epoch": 2.7264296754250386,
+      "grad_norm": 0.2974379407362513,
+      "learning_rate": 4.0669081624248464e-05,
+      "loss": 0.401,
+      "step": 1764
+    },
+    {
+      "epoch": 2.7279752704791345,
+      "grad_norm": 0.21490090328087502,
+      "learning_rate": 4.0625918712946665e-05,
+      "loss": 0.3857,
+      "step": 1765
+    },
+    {
+      "epoch": 2.7295208655332304,
+      "grad_norm": 0.19018711214547304,
+      "learning_rate": 4.05827550726353e-05,
+      "loss": 0.4022,
+      "step": 1766
+    },
+    {
+      "epoch": 2.7310664605873263,
+      "grad_norm": 0.2161740494195666,
+      "learning_rate": 4.0539590753587195e-05,
+      "loss": 0.3945,
+      "step": 1767
+    },
+    {
+      "epoch": 2.7326120556414217,
+      "grad_norm": 0.22327255615725017,
+      "learning_rate": 4.049642580607599e-05,
+      "loss": 0.3919,
+      "step": 1768
+    },
+    {
+      "epoch": 2.7341576506955176,
+      "grad_norm": 0.2002937348555192,
+      "learning_rate": 4.045326028037603e-05,
+      "loss": 0.391,
+      "step": 1769
+    },
+    {
+      "epoch": 2.7357032457496135,
+      "grad_norm": 0.21879066736924463,
+      "learning_rate": 4.041009422676234e-05,
+      "loss": 0.3932,
+      "step": 1770
+    },
+    {
+      "epoch": 2.7372488408037094,
+      "grad_norm": 0.21984116168203174,
+      "learning_rate": 4.036692769551056e-05,
+      "loss": 0.3815,
+      "step": 1771
+    },
+    {
+      "epoch": 2.7387944358578054,
+      "grad_norm": 0.22773966477195914,
+      "learning_rate": 4.0323760736896904e-05,
+      "loss": 0.3871,
+      "step": 1772
+    },
+    {
+      "epoch": 2.7403400309119013,
+      "grad_norm": 0.17651812601554587,
+      "learning_rate": 4.0280593401198046e-05,
+      "loss": 0.3874,
+      "step": 1773
+    },
+    {
+      "epoch": 2.741885625965997,
+      "grad_norm": 0.2452970756456186,
+      "learning_rate": 4.0237425738691156e-05,
+      "loss": 0.3911,
+      "step": 1774
+    },
+    {
+      "epoch": 2.7434312210200926,
+      "grad_norm": 0.28920746138391956,
+      "learning_rate": 4.0194257799653724e-05,
+      "loss": 0.3971,
+      "step": 1775
+    },
+    {
+      "epoch": 2.7449768160741885,
+      "grad_norm": 0.2139771033872057,
+      "learning_rate": 4.0151089634363593e-05,
+      "loss": 0.3817,
+      "step": 1776
+    },
+    {
+      "epoch": 2.7465224111282844,
+      "grad_norm": 0.20423584050338486,
+      "learning_rate": 4.010792129309888e-05,
+      "loss": 0.3928,
+      "step": 1777
+    },
+    {
+      "epoch": 2.7480680061823803,
+      "grad_norm": 0.3212385675865011,
+      "learning_rate": 4.006475282613789e-05,
+      "loss": 0.3874,
+      "step": 1778
+    },
+    {
+      "epoch": 2.749613601236476,
+      "grad_norm": 0.2704711707718232,
+      "learning_rate": 4.0021584283759066e-05,
+      "loss": 0.3989,
+      "step": 1779
+    },
+    {
+      "epoch": 2.7511591962905717,
+      "grad_norm": 0.2682116360791284,
+      "learning_rate": 3.997841571624094e-05,
+      "loss": 0.3916,
+      "step": 1780
+    },
+    {
+      "epoch": 2.7527047913446676,
+      "grad_norm": 0.32279677803920165,
+      "learning_rate": 3.9935247173862124e-05,
+      "loss": 0.3866,
+      "step": 1781
+    },
+    {
+      "epoch": 2.7542503863987635,
+      "grad_norm": 0.2830715557450085,
+      "learning_rate": 3.9892078706901135e-05,
+      "loss": 0.3924,
+      "step": 1782
+    },
+    {
+      "epoch": 2.7557959814528594,
+      "grad_norm": 0.24500749618625672,
+      "learning_rate": 3.9848910365636406e-05,
+      "loss": 0.3923,
+      "step": 1783
+    },
+    {
+      "epoch": 2.7573415765069553,
+      "grad_norm": 0.264674266420782,
+      "learning_rate": 3.980574220034629e-05,
+      "loss": 0.3849,
+      "step": 1784
+    },
+    {
+      "epoch": 2.7588871715610512,
+      "grad_norm": 0.21841249256542083,
+      "learning_rate": 3.976257426130886e-05,
+      "loss": 0.3891,
+      "step": 1785
+    },
+    {
+      "epoch": 2.7604327666151467,
+      "grad_norm": 0.18420110960683264,
+      "learning_rate": 3.971940659880196e-05,
+      "loss": 0.3944,
+      "step": 1786
+    },
+    {
+      "epoch": 2.7619783616692426,
+      "grad_norm": 0.2877075785983188,
+      "learning_rate": 3.967623926310311e-05,
+      "loss": 0.3906,
+      "step": 1787
+    },
+    {
+      "epoch": 2.7635239567233385,
+      "grad_norm": 0.24290364105842152,
+      "learning_rate": 3.963307230448945e-05,
+      "loss": 0.4077,
+      "step": 1788
+    },
+    {
+      "epoch": 2.7650695517774344,
+      "grad_norm": 0.21695406242088836,
+      "learning_rate": 3.958990577323767e-05,
+      "loss": 0.3937,
+      "step": 1789
+    },
+    {
+      "epoch": 2.76661514683153,
+      "grad_norm": 0.25856486111006705,
+      "learning_rate": 3.954673971962398e-05,
+      "loss": 0.3959,
+      "step": 1790
+    },
+    {
+      "epoch": 2.7681607418856258,
+      "grad_norm": 0.2491166348704581,
+      "learning_rate": 3.950357419392402e-05,
+      "loss": 0.3879,
+      "step": 1791
+    },
+    {
+      "epoch": 2.7697063369397217,
+      "grad_norm": 0.20871545391562396,
+      "learning_rate": 3.9460409246412805e-05,
+      "loss": 0.4034,
+      "step": 1792
+    },
+    {
+      "epoch": 2.7712519319938176,
+      "grad_norm": 0.20508228510925783,
+      "learning_rate": 3.941724492736471e-05,
+      "loss": 0.4042,
+      "step": 1793
+    },
+    {
+      "epoch": 2.7727975270479135,
+      "grad_norm": 0.18720477368958455,
+      "learning_rate": 3.9374081287053355e-05,
+      "loss": 0.3893,
+      "step": 1794
+    },
+    {
+      "epoch": 2.7743431221020094,
+      "grad_norm": 0.19308946884703032,
+      "learning_rate": 3.9330918375751556e-05,
+      "loss": 0.3915,
+      "step": 1795
+    },
+    {
+      "epoch": 2.7758887171561053,
+      "grad_norm": 0.18226166599343516,
+      "learning_rate": 3.928775624373131e-05,
+      "loss": 0.3918,
+      "step": 1796
+    },
+    {
+      "epoch": 2.7774343122102008,
+      "grad_norm": 0.17465834204444608,
+      "learning_rate": 3.92445949412637e-05,
+      "loss": 0.3876,
+      "step": 1797
+    },
+    {
+      "epoch": 2.7789799072642967,
+      "grad_norm": 0.1813088489376211,
+      "learning_rate": 3.9201434518618826e-05,
+      "loss": 0.3988,
+      "step": 1798
+    },
+    {
+      "epoch": 2.7805255023183926,
+      "grad_norm": 0.18100302492637613,
+      "learning_rate": 3.915827502606576e-05,
+      "loss": 0.3802,
+      "step": 1799
+    },
+    {
+      "epoch": 2.7820710973724885,
+      "grad_norm": 0.19417838342895236,
+      "learning_rate": 3.911511651387253e-05,
+      "loss": 0.377,
+      "step": 1800
+    },
+    {
+      "epoch": 2.7836166924265844,
+      "grad_norm": 0.2573213340117975,
+      "learning_rate": 3.907195903230599e-05,
+      "loss": 0.3912,
+      "step": 1801
+    },
+    {
+      "epoch": 2.78516228748068,
+      "grad_norm": 0.284000471614278,
+      "learning_rate": 3.902880263163178e-05,
+      "loss": 0.3848,
+      "step": 1802
+    },
+    {
+      "epoch": 2.7867078825347757,
+      "grad_norm": 0.22616621719359473,
+      "learning_rate": 3.898564736211432e-05,
+      "loss": 0.4022,
+      "step": 1803
+    },
+    {
+      "epoch": 2.7882534775888717,
+      "grad_norm": 0.20526110287847993,
+      "learning_rate": 3.894249327401669e-05,
+      "loss": 0.397,
+      "step": 1804
+    },
+    {
+      "epoch": 2.7897990726429676,
+      "grad_norm": 0.17058234330188535,
+      "learning_rate": 3.889934041760058e-05,
+      "loss": 0.3926,
+      "step": 1805
+    },
+    {
+      "epoch": 2.7913446676970635,
+      "grad_norm": 0.22396294493426347,
+      "learning_rate": 3.88561888431263e-05,
+      "loss": 0.3968,
+      "step": 1806
+    },
+    {
+      "epoch": 2.7928902627511594,
+      "grad_norm": 0.18761021761388325,
+      "learning_rate": 3.88130386008526e-05,
+      "loss": 0.4027,
+      "step": 1807
+    },
+    {
+      "epoch": 2.7944358578052553,
+      "grad_norm": 0.16755062760736492,
+      "learning_rate": 3.87698897410367e-05,
+      "loss": 0.3972,
+      "step": 1808
+    },
+    {
+      "epoch": 2.7959814528593507,
+      "grad_norm": 0.24803898568626884,
+      "learning_rate": 3.872674231393425e-05,
+      "loss": 0.3876,
+      "step": 1809
+    },
+    {
+      "epoch": 2.7975270479134466,
+      "grad_norm": 0.2607238677140293,
+      "learning_rate": 3.868359636979919e-05,
+      "loss": 0.396,
+      "step": 1810
+    },
+    {
+      "epoch": 2.7990726429675425,
+      "grad_norm": 0.24051231824413313,
+      "learning_rate": 3.8640451958883705e-05,
+      "loss": 0.394,
+      "step": 1811
+    },
+    {
+      "epoch": 2.8006182380216385,
+      "grad_norm": 0.2198760157566278,
+      "learning_rate": 3.8597309131438285e-05,
+      "loss": 0.3929,
+      "step": 1812
+    },
+    {
+      "epoch": 2.802163833075734,
+      "grad_norm": 0.21247205512843312,
+      "learning_rate": 3.85541679377115e-05,
+      "loss": 0.39,
+      "step": 1813
+    },
+    {
+      "epoch": 2.80370942812983,
+      "grad_norm": 0.22974063721913382,
+      "learning_rate": 3.851102842795001e-05,
+      "loss": 0.3873,
+      "step": 1814
+    },
+    {
+      "epoch": 2.8052550231839257,
+      "grad_norm": 0.24422117496900178,
+      "learning_rate": 3.846789065239859e-05,
+      "loss": 0.3866,
+      "step": 1815
+    },
+    {
+      "epoch": 2.8068006182380216,
+      "grad_norm": 0.1879780673874634,
+      "learning_rate": 3.8424754661299926e-05,
+      "loss": 0.3915,
+      "step": 1816
+    },
+    {
+      "epoch": 2.8083462132921175,
+      "grad_norm": 0.28131521964208156,
+      "learning_rate": 3.8381620504894635e-05,
+      "loss": 0.3965,
+      "step": 1817
+    },
+    {
+      "epoch": 2.8098918083462134,
+      "grad_norm": 0.4062594396282389,
+      "learning_rate": 3.833848823342124e-05,
+      "loss": 0.3882,
+      "step": 1818
+    },
+    {
+      "epoch": 2.8114374034003093,
+      "grad_norm": 0.3737581895445919,
+      "learning_rate": 3.8295357897116015e-05,
+      "loss": 0.3889,
+      "step": 1819
+    },
+    {
+      "epoch": 2.812982998454405,
+      "grad_norm": 0.3180846973132513,
+      "learning_rate": 3.825222954621301e-05,
+      "loss": 0.4007,
+      "step": 1820
+    },
+    {
+      "epoch": 2.8145285935085007,
+      "grad_norm": 0.23924262459817525,
+      "learning_rate": 3.8209103230943955e-05,
+      "loss": 0.396,
+      "step": 1821
+    },
+    {
+      "epoch": 2.8160741885625966,
+      "grad_norm": 0.17186725441862907,
+      "learning_rate": 3.816597900153822e-05,
+      "loss": 0.3825,
+      "step": 1822
+    },
+    {
+      "epoch": 2.8176197836166925,
+      "grad_norm": 0.2019171990332553,
+      "learning_rate": 3.812285690822272e-05,
+      "loss": 0.3911,
+      "step": 1823
+    },
+    {
+      "epoch": 2.819165378670788,
+      "grad_norm": 0.17955446107446282,
+      "learning_rate": 3.8079737001221905e-05,
+      "loss": 0.3931,
+      "step": 1824
+    },
+    {
+      "epoch": 2.820710973724884,
+      "grad_norm": 0.20078132013690356,
+      "learning_rate": 3.803661933075768e-05,
+      "loss": 0.3899,
+      "step": 1825
+    },
+    {
+      "epoch": 2.82225656877898,
+      "grad_norm": 0.2281153173227666,
+      "learning_rate": 3.799350394704932e-05,
+      "loss": 0.3774,
+      "step": 1826
+    },
+    {
+      "epoch": 2.8238021638330757,
+      "grad_norm": 0.21250268042646084,
+      "learning_rate": 3.7950390900313455e-05,
+      "loss": 0.3952,
+      "step": 1827
+    },
+    {
+      "epoch": 2.8253477588871716,
+      "grad_norm": 0.17452701520242536,
+      "learning_rate": 3.790728024076401e-05,
+      "loss": 0.393,
+      "step": 1828
+    },
+    {
+      "epoch": 2.8268933539412675,
+      "grad_norm": 0.18226256585586875,
+      "learning_rate": 3.786417201861209e-05,
+      "loss": 0.3896,
+      "step": 1829
+    },
+    {
+      "epoch": 2.8284389489953634,
+      "grad_norm": 0.1699200232766844,
+      "learning_rate": 3.782106628406599e-05,
+      "loss": 0.3881,
+      "step": 1830
+    },
+    {
+      "epoch": 2.8299845440494593,
+      "grad_norm": 0.17347852322248702,
+      "learning_rate": 3.77779630873311e-05,
+      "loss": 0.3945,
+      "step": 1831
+    },
+    {
+      "epoch": 2.8315301391035548,
+      "grad_norm": 0.19679976534755758,
+      "learning_rate": 3.773486247860986e-05,
+      "loss": 0.3935,
+      "step": 1832
+    },
+    {
+      "epoch": 2.8330757341576507,
+      "grad_norm": 0.18755022647775246,
+      "learning_rate": 3.769176450810167e-05,
+      "loss": 0.3909,
+      "step": 1833
+    },
+    {
+      "epoch": 2.8346213292117466,
+      "grad_norm": 0.16961535389711938,
+      "learning_rate": 3.7648669226002904e-05,
+      "loss": 0.396,
+      "step": 1834
+    },
+    {
+      "epoch": 2.8361669242658425,
+      "grad_norm": 0.18048002267687868,
+      "learning_rate": 3.760557668250676e-05,
+      "loss": 0.3845,
+      "step": 1835
+    },
+    {
+      "epoch": 2.837712519319938,
+      "grad_norm": 0.1736727737721342,
+      "learning_rate": 3.756248692780325e-05,
+      "loss": 0.3993,
+      "step": 1836
+    },
+    {
+      "epoch": 2.839258114374034,
+      "grad_norm": 0.17581172468210995,
+      "learning_rate": 3.7519400012079203e-05,
+      "loss": 0.3921,
+      "step": 1837
+    },
+    {
+      "epoch": 2.8408037094281298,
+      "grad_norm": 0.1894157962541827,
+      "learning_rate": 3.747631598551804e-05,
+      "loss": 0.3949,
+      "step": 1838
+    },
+    {
+      "epoch": 2.8423493044822257,
+      "grad_norm": 0.1921294070326651,
+      "learning_rate": 3.743323489829988e-05,
+      "loss": 0.3856,
+      "step": 1839
+    },
+    {
+      "epoch": 2.8438948995363216,
+      "grad_norm": 0.1772283655136702,
+      "learning_rate": 3.739015680060141e-05,
+      "loss": 0.3907,
+      "step": 1840
+    },
+    {
+      "epoch": 2.8454404945904175,
+      "grad_norm": 0.22891639350098605,
+      "learning_rate": 3.734708174259585e-05,
+      "loss": 0.3873,
+      "step": 1841
+    },
+    {
+      "epoch": 2.8469860896445134,
+      "grad_norm": 0.17898775501928832,
+      "learning_rate": 3.730400977445282e-05,
+      "loss": 0.3833,
+      "step": 1842
+    },
+    {
+      "epoch": 2.848531684698609,
+      "grad_norm": 0.1838794786145881,
+      "learning_rate": 3.7260940946338416e-05,
+      "loss": 0.3836,
+      "step": 1843
+    },
+    {
+      "epoch": 2.8500772797527048,
+      "grad_norm": 0.23873368949862525,
+      "learning_rate": 3.7217875308415036e-05,
+      "loss": 0.3953,
+      "step": 1844
+    },
+    {
+      "epoch": 2.8516228748068007,
+      "grad_norm": 0.2037754497050098,
+      "learning_rate": 3.717481291084135e-05,
+      "loss": 0.3879,
+      "step": 1845
+    },
+    {
+      "epoch": 2.8531684698608966,
+      "grad_norm": 0.19247884951391822,
+      "learning_rate": 3.713175380377229e-05,
+      "loss": 0.3782,
+      "step": 1846
+    },
+    {
+      "epoch": 2.854714064914992,
+      "grad_norm": 0.3062130201916458,
+      "learning_rate": 3.708869803735895e-05,
+      "loss": 0.3905,
+      "step": 1847
+    },
+    {
+      "epoch": 2.856259659969088,
+      "grad_norm": 0.19763069985499787,
+      "learning_rate": 3.704564566174849e-05,
+      "loss": 0.3832,
+      "step": 1848
+    },
+    {
+      "epoch": 2.857805255023184,
+      "grad_norm": 0.18346313517919666,
+      "learning_rate": 3.700259672708417e-05,
+      "loss": 0.3867,
+      "step": 1849
+    },
+    {
+      "epoch": 2.8593508500772797,
+      "grad_norm": 0.1961417697891756,
+      "learning_rate": 3.695955128350524e-05,
+      "loss": 0.3923,
+      "step": 1850
+    },
+    {
+      "epoch": 2.8608964451313756,
+      "grad_norm": 0.20811812847397376,
+      "learning_rate": 3.691650938114684e-05,
+      "loss": 0.4045,
+      "step": 1851
+    },
+    {
+      "epoch": 2.8624420401854715,
+      "grad_norm": 0.19118895495750038,
+      "learning_rate": 3.6873471070140026e-05,
+      "loss": 0.3877,
+      "step": 1852
+    },
+    {
+      "epoch": 2.8639876352395675,
+      "grad_norm": 0.2446781595702205,
+      "learning_rate": 3.6830436400611657e-05,
+      "loss": 0.3968,
+      "step": 1853
+    },
+    {
+      "epoch": 2.865533230293663,
+      "grad_norm": 0.27637008537499236,
+      "learning_rate": 3.678740542268437e-05,
+      "loss": 0.4017,
+      "step": 1854
+    },
+    {
+      "epoch": 2.867078825347759,
+      "grad_norm": 0.1716813443429829,
+      "learning_rate": 3.674437818647645e-05,
+      "loss": 0.3945,
+      "step": 1855
+    },
+    {
+      "epoch": 2.8686244204018547,
+      "grad_norm": 0.2843102701296005,
+      "learning_rate": 3.67013547421019e-05,
+      "loss": 0.3903,
+      "step": 1856
+    },
+    {
+      "epoch": 2.8701700154559506,
+      "grad_norm": 0.3605604803916667,
+      "learning_rate": 3.665833513967024e-05,
+      "loss": 0.3914,
+      "step": 1857
+    },
+    {
+      "epoch": 2.871715610510046,
+      "grad_norm": 0.281028305368157,
+      "learning_rate": 3.6615319429286544e-05,
+      "loss": 0.3907,
+      "step": 1858
+    },
+    {
+      "epoch": 2.873261205564142,
+      "grad_norm": 0.3706718088832441,
+      "learning_rate": 3.657230766105136e-05,
+      "loss": 0.387,
+      "step": 1859
+    },
+    {
+      "epoch": 2.874806800618238,
+      "grad_norm": 0.3322333261169484,
+      "learning_rate": 3.6529299885060634e-05,
+      "loss": 0.3863,
+      "step": 1860
+    },
+    {
+      "epoch": 2.876352395672334,
+      "grad_norm": 0.32728465454278505,
+      "learning_rate": 3.648629615140565e-05,
+      "loss": 0.3792,
+      "step": 1861
+    },
+    {
+      "epoch": 2.8778979907264297,
+      "grad_norm": 0.4173074392701693,
+      "learning_rate": 3.6443296510173006e-05,
+      "loss": 0.3881,
+      "step": 1862
+    },
+    {
+      "epoch": 2.8794435857805256,
+      "grad_norm": 0.3376004538998364,
+      "learning_rate": 3.640030101144453e-05,
+      "loss": 0.3905,
+      "step": 1863
+    },
+    {
+      "epoch": 2.8809891808346215,
+      "grad_norm": 0.21414190118493256,
+      "learning_rate": 3.635730970529719e-05,
+      "loss": 0.3845,
+      "step": 1864
+    },
+    {
+      "epoch": 2.8825347758887174,
+      "grad_norm": 0.3558180588155539,
+      "learning_rate": 3.631432264180316e-05,
+      "loss": 0.3941,
+      "step": 1865
+    },
+    {
+      "epoch": 2.884080370942813,
+      "grad_norm": 0.39320314157434616,
+      "learning_rate": 3.627133987102956e-05,
+      "loss": 0.3904,
+      "step": 1866
+    },
+    {
+      "epoch": 2.885625965996909,
+      "grad_norm": 0.32800947779896955,
+      "learning_rate": 3.622836144303857e-05,
+      "loss": 0.3979,
+      "step": 1867
+    },
+    {
+      "epoch": 2.8871715610510047,
+      "grad_norm": 0.40344014390857996,
+      "learning_rate": 3.618538740788733e-05,
+      "loss": 0.3879,
+      "step": 1868
+    },
+    {
+      "epoch": 2.8887171561051006,
+      "grad_norm": 0.3092449027860355,
+      "learning_rate": 3.614241781562783e-05,
+      "loss": 0.4004,
+      "step": 1869
+    },
+    {
+      "epoch": 2.890262751159196,
+      "grad_norm": 0.3191402898079322,
+      "learning_rate": 3.609945271630687e-05,
+      "loss": 0.387,
+      "step": 1870
+    },
+    {
+      "epoch": 2.891808346213292,
+      "grad_norm": 0.3404716830049869,
+      "learning_rate": 3.6056492159966076e-05,
+      "loss": 0.3861,
+      "step": 1871
+    },
+    {
+      "epoch": 2.893353941267388,
+      "grad_norm": 0.23888696627268902,
+      "learning_rate": 3.601353619664173e-05,
+      "loss": 0.3842,
+      "step": 1872
+    },
+    {
+      "epoch": 2.894899536321484,
+      "grad_norm": 0.3112003266748598,
+      "learning_rate": 3.597058487636478e-05,
+      "loss": 0.3816,
+      "step": 1873
+    },
+    {
+      "epoch": 2.8964451313755797,
+      "grad_norm": 0.27628158618508,
+      "learning_rate": 3.592763824916078e-05,
+      "loss": 0.3889,
+      "step": 1874
+    },
+    {
+      "epoch": 2.8979907264296756,
+      "grad_norm": 0.21858501891722934,
+      "learning_rate": 3.58846963650498e-05,
+      "loss": 0.3929,
+      "step": 1875
+    },
+    {
+      "epoch": 2.8995363214837715,
+      "grad_norm": 0.20230449524397381,
+      "learning_rate": 3.584175927404641e-05,
+      "loss": 0.3843,
+      "step": 1876
+    },
+    {
+      "epoch": 2.901081916537867,
+      "grad_norm": 0.1996976045353686,
+      "learning_rate": 3.5798827026159536e-05,
+      "loss": 0.3846,
+      "step": 1877
+    },
+    {
+      "epoch": 2.902627511591963,
+      "grad_norm": 0.2539315574524483,
+      "learning_rate": 3.5755899671392546e-05,
+      "loss": 0.3863,
+      "step": 1878
+    },
+    {
+      "epoch": 2.9041731066460588,
+      "grad_norm": 0.27347730889148514,
+      "learning_rate": 3.5712977259743065e-05,
+      "loss": 0.3809,
+      "step": 1879
+    },
+    {
+      "epoch": 2.9057187017001547,
+      "grad_norm": 0.2670957397189431,
+      "learning_rate": 3.5670059841202945e-05,
+      "loss": 0.394,
+      "step": 1880
+    },
+    {
+      "epoch": 2.90726429675425,
+      "grad_norm": 0.32819821406555794,
+      "learning_rate": 3.5627147465758276e-05,
+      "loss": 0.3888,
+      "step": 1881
+    },
+    {
+      "epoch": 2.908809891808346,
+      "grad_norm": 0.34145982364945043,
+      "learning_rate": 3.558424018338922e-05,
+      "loss": 0.3899,
+      "step": 1882
+    },
+    {
+      "epoch": 2.910355486862442,
+      "grad_norm": 0.3096000913364954,
+      "learning_rate": 3.554133804407002e-05,
+      "loss": 0.3969,
+      "step": 1883
+    },
+    {
+      "epoch": 2.911901081916538,
+      "grad_norm": 0.34394438543226763,
+      "learning_rate": 3.549844109776897e-05,
+      "loss": 0.3908,
+      "step": 1884
+    },
+    {
+      "epoch": 2.9134466769706338,
+      "grad_norm": 0.28540515538147054,
+      "learning_rate": 3.5455549394448255e-05,
+      "loss": 0.3825,
+      "step": 1885
+    },
+    {
+      "epoch": 2.9149922720247297,
+      "grad_norm": 0.3004810072116905,
+      "learning_rate": 3.541266298406399e-05,
+      "loss": 0.3892,
+      "step": 1886
+    },
+    {
+      "epoch": 2.9165378670788256,
+      "grad_norm": 0.21089141561599284,
+      "learning_rate": 3.5369781916566115e-05,
+      "loss": 0.3884,
+      "step": 1887
+    },
+    {
+      "epoch": 2.918083462132921,
+      "grad_norm": 0.26942714783902477,
+      "learning_rate": 3.5326906241898365e-05,
+      "loss": 0.3984,
+      "step": 1888
+    },
+    {
+      "epoch": 2.919629057187017,
+      "grad_norm": 0.3442399952953209,
+      "learning_rate": 3.528403600999815e-05,
+      "loss": 0.3901,
+      "step": 1889
+    },
+    {
+      "epoch": 2.921174652241113,
+      "grad_norm": 0.3001915641726886,
+      "learning_rate": 3.5241171270796594e-05,
+      "loss": 0.404,
+      "step": 1890
+    },
+    {
+      "epoch": 2.9227202472952087,
+      "grad_norm": 0.3761387251745968,
+      "learning_rate": 3.519831207421839e-05,
+      "loss": 0.3971,
+      "step": 1891
+    },
+    {
+      "epoch": 2.9242658423493046,
+      "grad_norm": 0.29892816120859234,
+      "learning_rate": 3.5155458470181766e-05,
+      "loss": 0.4035,
+      "step": 1892
+    },
+    {
+      "epoch": 2.9258114374034,
+      "grad_norm": 0.3150415273440005,
+      "learning_rate": 3.51126105085985e-05,
+      "loss": 0.3884,
+      "step": 1893
+    },
+    {
+      "epoch": 2.927357032457496,
+      "grad_norm": 0.34568619548067764,
+      "learning_rate": 3.506976823937371e-05,
+      "loss": 0.388,
+      "step": 1894
+    },
+    {
+      "epoch": 2.928902627511592,
+      "grad_norm": 0.31895937702164295,
+      "learning_rate": 3.502693171240594e-05,
+      "loss": 0.3945,
+      "step": 1895
+    },
+    {
+      "epoch": 2.930448222565688,
+      "grad_norm": 0.2747416006092413,
+      "learning_rate": 3.498410097758703e-05,
+      "loss": 0.3884,
+      "step": 1896
+    },
+    {
+      "epoch": 2.9319938176197837,
+      "grad_norm": 0.36800883335054185,
+      "learning_rate": 3.4941276084802104e-05,
+      "loss": 0.3885,
+      "step": 1897
+    },
+    {
+      "epoch": 2.9335394126738796,
+      "grad_norm": 0.3740803469722144,
+      "learning_rate": 3.48984570839294e-05,
+      "loss": 0.3886,
+      "step": 1898
+    },
+    {
+      "epoch": 2.9350850077279755,
+      "grad_norm": 0.35763463357124636,
+      "learning_rate": 3.485564402484039e-05,
+      "loss": 0.3915,
+      "step": 1899
+    },
+    {
+      "epoch": 2.936630602782071,
+      "grad_norm": 0.25083149731119164,
+      "learning_rate": 3.481283695739959e-05,
+      "loss": 0.3965,
+      "step": 1900
+    },
+    {
+      "epoch": 2.938176197836167,
+      "grad_norm": 0.23201087393763298,
+      "learning_rate": 3.477003593146448e-05,
+      "loss": 0.3846,
+      "step": 1901
+    },
+    {
+      "epoch": 2.939721792890263,
+      "grad_norm": 0.26829347016684496,
+      "learning_rate": 3.4727240996885616e-05,
+      "loss": 0.3727,
+      "step": 1902
+    },
+    {
+      "epoch": 2.9412673879443587,
+      "grad_norm": 0.25541284401326064,
+      "learning_rate": 3.4684452203506366e-05,
+      "loss": 0.3936,
+      "step": 1903
+    },
+    {
+      "epoch": 2.942812982998454,
+      "grad_norm": 0.2881274195973795,
+      "learning_rate": 3.464166960116299e-05,
+      "loss": 0.3876,
+      "step": 1904
+    },
+    {
+      "epoch": 2.94435857805255,
+      "grad_norm": 0.22051538054770017,
+      "learning_rate": 3.459889323968449e-05,
+      "loss": 0.3869,
+      "step": 1905
+    },
+    {
+      "epoch": 2.945904173106646,
+      "grad_norm": 0.2531089166492334,
+      "learning_rate": 3.455612316889268e-05,
+      "loss": 0.3905,
+      "step": 1906
+    },
+    {
+      "epoch": 2.947449768160742,
+      "grad_norm": 0.26532487494222146,
+      "learning_rate": 3.4513359438601976e-05,
+      "loss": 0.3933,
+      "step": 1907
+    },
+    {
+      "epoch": 2.948995363214838,
+      "grad_norm": 0.24694988073087612,
+      "learning_rate": 3.4470602098619437e-05,
+      "loss": 0.3883,
+      "step": 1908
+    },
+    {
+      "epoch": 2.9505409582689337,
+      "grad_norm": 0.21275726411920523,
+      "learning_rate": 3.442785119874469e-05,
+      "loss": 0.3997,
+      "step": 1909
+    },
+    {
+      "epoch": 2.9520865533230296,
+      "grad_norm": 0.21271410672737606,
+      "learning_rate": 3.438510678876983e-05,
+      "loss": 0.3777,
+      "step": 1910
+    },
+    {
+      "epoch": 2.953632148377125,
+      "grad_norm": 0.2801489130994963,
+      "learning_rate": 3.434236891847941e-05,
+      "loss": 0.3876,
+      "step": 1911
+    },
+    {
+      "epoch": 2.955177743431221,
+      "grad_norm": 0.25117899176243486,
+      "learning_rate": 3.4299637637650374e-05,
+      "loss": 0.3878,
+      "step": 1912
+    },
+    {
+      "epoch": 2.956723338485317,
+      "grad_norm": 0.29618242491492724,
+      "learning_rate": 3.4256912996051996e-05,
+      "loss": 0.394,
+      "step": 1913
+    },
+    {
+      "epoch": 2.958268933539413,
+      "grad_norm": 0.18860005187869416,
+      "learning_rate": 3.421419504344578e-05,
+      "loss": 0.3948,
+      "step": 1914
+    },
+    {
+      "epoch": 2.9598145285935082,
+      "grad_norm": 0.21394031751835493,
+      "learning_rate": 3.417148382958549e-05,
+      "loss": 0.3974,
+      "step": 1915
+    },
+    {
+      "epoch": 2.961360123647604,
+      "grad_norm": 0.3149830463145957,
+      "learning_rate": 3.4128779404217004e-05,
+      "loss": 0.3893,
+      "step": 1916
+    },
+    {
+      "epoch": 2.9629057187017,
+      "grad_norm": 0.2104129047081311,
+      "learning_rate": 3.40860818170783e-05,
+      "loss": 0.3975,
+      "step": 1917
+    },
+    {
+      "epoch": 2.964451313755796,
+      "grad_norm": 0.22771617164783442,
+      "learning_rate": 3.404339111789941e-05,
+      "loss": 0.3776,
+      "step": 1918
+    },
+    {
+      "epoch": 2.965996908809892,
+      "grad_norm": 0.2884790016562261,
+      "learning_rate": 3.400070735640233e-05,
+      "loss": 0.3941,
+      "step": 1919
+    },
+    {
+      "epoch": 2.9675425038639878,
+      "grad_norm": 0.18913219534335085,
+      "learning_rate": 3.395803058230095e-05,
+      "loss": 0.3933,
+      "step": 1920
+    },
+    {
+      "epoch": 2.9690880989180837,
+      "grad_norm": 0.2384230589963729,
+      "learning_rate": 3.3915360845301085e-05,
+      "loss": 0.3852,
+      "step": 1921
+    },
+    {
+      "epoch": 2.970633693972179,
+      "grad_norm": 0.2633334096132983,
+      "learning_rate": 3.38726981951003e-05,
+      "loss": 0.3947,
+      "step": 1922
+    },
+    {
+      "epoch": 2.972179289026275,
+      "grad_norm": 0.20814728125813683,
+      "learning_rate": 3.3830042681387904e-05,
+      "loss": 0.39,
+      "step": 1923
+    },
+    {
+      "epoch": 2.973724884080371,
+      "grad_norm": 0.24367474670415484,
+      "learning_rate": 3.378739435384496e-05,
+      "loss": 0.3791,
+      "step": 1924
+    },
+    {
+      "epoch": 2.975270479134467,
+      "grad_norm": 0.2821750783223297,
+      "learning_rate": 3.374475326214408e-05,
+      "loss": 0.3889,
+      "step": 1925
+    },
+    {
+      "epoch": 2.9768160741885628,
+      "grad_norm": 0.27530067994928065,
+      "learning_rate": 3.370211945594947e-05,
+      "loss": 0.3902,
+      "step": 1926
+    },
+    {
+      "epoch": 2.978361669242658,
+      "grad_norm": 0.2720303840999367,
+      "learning_rate": 3.36594929849169e-05,
+      "loss": 0.3842,
+      "step": 1927
+    },
+    {
+      "epoch": 2.979907264296754,
+      "grad_norm": 0.19930232210254215,
+      "learning_rate": 3.3616873898693554e-05,
+      "loss": 0.3792,
+      "step": 1928
+    },
+    {
+      "epoch": 2.98145285935085,
+      "grad_norm": 0.36206977990301475,
+      "learning_rate": 3.3574262246917976e-05,
+      "loss": 0.4014,
+      "step": 1929
+    },
+    {
+      "epoch": 2.982998454404946,
+      "grad_norm": 0.21907688820073476,
+      "learning_rate": 3.3531658079220154e-05,
+      "loss": 0.3806,
+      "step": 1930
+    },
+    {
+      "epoch": 2.984544049459042,
+      "grad_norm": 0.23541598495245927,
+      "learning_rate": 3.348906144522127e-05,
+      "loss": 0.3932,
+      "step": 1931
+    },
+    {
+      "epoch": 2.9860896445131377,
+      "grad_norm": 0.2548801503252262,
+      "learning_rate": 3.344647239453379e-05,
+      "loss": 0.3869,
+      "step": 1932
+    },
+    {
+      "epoch": 2.9876352395672336,
+      "grad_norm": 0.22769350023934656,
+      "learning_rate": 3.340389097676126e-05,
+      "loss": 0.3839,
+      "step": 1933
+    },
+    {
+      "epoch": 2.989180834621329,
+      "grad_norm": 0.26547226925373146,
+      "learning_rate": 3.336131724149846e-05,
+      "loss": 0.3886,
+      "step": 1934
+    },
+    {
+      "epoch": 2.990726429675425,
+      "grad_norm": 0.3073755040234777,
+      "learning_rate": 3.3318751238331134e-05,
+      "loss": 0.3892,
+      "step": 1935
+    },
+    {
+      "epoch": 2.992272024729521,
+      "grad_norm": 0.23743365918197448,
+      "learning_rate": 3.327619301683603e-05,
+      "loss": 0.3868,
+      "step": 1936
+    },
+    {
+      "epoch": 2.993817619783617,
+      "grad_norm": 0.2129623169710384,
+      "learning_rate": 3.3233642626580874e-05,
+      "loss": 0.3827,
+      "step": 1937
+    },
+    {
+      "epoch": 2.9953632148377123,
+      "grad_norm": 0.19930478449122238,
+      "learning_rate": 3.3191100117124234e-05,
+      "loss": 0.3922,
+      "step": 1938
+    },
+    {
+      "epoch": 2.996908809891808,
+      "grad_norm": 0.2818998655532697,
+      "learning_rate": 3.31485655380155e-05,
+      "loss": 0.3969,
+      "step": 1939
+    },
+    {
+      "epoch": 2.998454404945904,
+      "grad_norm": 0.32138858700796147,
+      "learning_rate": 3.3106038938794854e-05,
+      "loss": 0.3918,
+      "step": 1940
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.23813170146158089,
+      "learning_rate": 3.306352036899315e-05,
+      "loss": 0.3779,
+      "step": 1941
+    },
+    {
+      "epoch": 3.001545595054096,
+      "grad_norm": 0.31171465584784497,
+      "learning_rate": 3.30210098781319e-05,
+      "loss": 0.3679,
+      "step": 1942
+    },
+    {
+      "epoch": 3.003091190108192,
+      "grad_norm": 0.2869157618738416,
+      "learning_rate": 3.2978507515723235e-05,
+      "loss": 0.3756,
+      "step": 1943
+    },
+    {
+      "epoch": 3.0046367851622873,
+      "grad_norm": 0.20074930726465806,
+      "learning_rate": 3.2936013331269775e-05,
+      "loss": 0.3651,
+      "step": 1944
+    },
+    {
+      "epoch": 3.006182380216383,
+      "grad_norm": 0.2148586383545826,
+      "learning_rate": 3.289352737426464e-05,
+      "loss": 0.3655,
+      "step": 1945
+    },
+    {
+      "epoch": 3.007727975270479,
+      "grad_norm": 0.2288239820290731,
+      "learning_rate": 3.285104969419136e-05,
+      "loss": 0.3619,
+      "step": 1946
+    },
+    {
+      "epoch": 3.009273570324575,
+      "grad_norm": 0.24053953157264535,
+      "learning_rate": 3.280858034052385e-05,
+      "loss": 0.3636,
+      "step": 1947
+    },
+    {
+      "epoch": 3.010819165378671,
+      "grad_norm": 0.2697801971339192,
+      "learning_rate": 3.2766119362726264e-05,
+      "loss": 0.3689,
+      "step": 1948
+    },
+    {
+      "epoch": 3.012364760432767,
+      "grad_norm": 0.21726444358702826,
+      "learning_rate": 3.2723666810253076e-05,
+      "loss": 0.3634,
+      "step": 1949
+    },
+    {
+      "epoch": 3.0139103554868623,
+      "grad_norm": 0.294339356595773,
+      "learning_rate": 3.26812227325489e-05,
+      "loss": 0.3645,
+      "step": 1950
+    },
+    {
+      "epoch": 3.015455950540958,
+      "grad_norm": 0.2410788387087104,
+      "learning_rate": 3.263878717904847e-05,
+      "loss": 0.3757,
+      "step": 1951
+    },
+    {
+      "epoch": 3.017001545595054,
+      "grad_norm": 0.2751603563610299,
+      "learning_rate": 3.259636019917667e-05,
+      "loss": 0.3646,
+      "step": 1952
+    },
+    {
+      "epoch": 3.01854714064915,
+      "grad_norm": 0.17361571571114168,
+      "learning_rate": 3.255394184234829e-05,
+      "loss": 0.3613,
+      "step": 1953
+    },
+    {
+      "epoch": 3.020092735703246,
+      "grad_norm": 0.2241225829426769,
+      "learning_rate": 3.251153215796814e-05,
+      "loss": 0.3681,
+      "step": 1954
+    },
+    {
+      "epoch": 3.021638330757342,
+      "grad_norm": 0.2365359130191026,
+      "learning_rate": 3.2469131195430933e-05,
+      "loss": 0.3717,
+      "step": 1955
+    },
+    {
+      "epoch": 3.0231839258114372,
+      "grad_norm": 0.172259454607108,
+      "learning_rate": 3.2426739004121214e-05,
+      "loss": 0.3717,
+      "step": 1956
+    },
+    {
+      "epoch": 3.024729520865533,
+      "grad_norm": 0.27340648916561816,
+      "learning_rate": 3.238435563341326e-05,
+      "loss": 0.371,
+      "step": 1957
+    },
+    {
+      "epoch": 3.026275115919629,
+      "grad_norm": 0.2413768979772997,
+      "learning_rate": 3.234198113267118e-05,
+      "loss": 0.3661,
+      "step": 1958
+    },
+    {
+      "epoch": 3.027820710973725,
+      "grad_norm": 0.26528013813158247,
+      "learning_rate": 3.2299615551248655e-05,
+      "loss": 0.3707,
+      "step": 1959
+    },
+    {
+      "epoch": 3.029366306027821,
+      "grad_norm": 0.20935732880513103,
+      "learning_rate": 3.2257258938489045e-05,
+      "loss": 0.3633,
+      "step": 1960
+    },
+    {
+      "epoch": 3.0309119010819163,
+      "grad_norm": 0.20152247145768049,
+      "learning_rate": 3.221491134372518e-05,
+      "loss": 0.357,
+      "step": 1961
+    },
+    {
+      "epoch": 3.0324574961360122,
+      "grad_norm": 0.24715380895695596,
+      "learning_rate": 3.217257281627951e-05,
+      "loss": 0.3744,
+      "step": 1962
+    },
+    {
+      "epoch": 3.034003091190108,
+      "grad_norm": 0.17849954414560545,
+      "learning_rate": 3.2130243405463834e-05,
+      "loss": 0.3602,
+      "step": 1963
+    },
+    {
+      "epoch": 3.035548686244204,
+      "grad_norm": 0.24389877635398147,
+      "learning_rate": 3.208792316057931e-05,
+      "loss": 0.3606,
+      "step": 1964
+    },
+    {
+      "epoch": 3.0370942812983,
+      "grad_norm": 0.24691720825848798,
+      "learning_rate": 3.204561213091653e-05,
+      "loss": 0.357,
+      "step": 1965
+    },
+    {
+      "epoch": 3.038639876352396,
+      "grad_norm": 0.18331040793836595,
+      "learning_rate": 3.200331036575525e-05,
+      "loss": 0.3664,
+      "step": 1966
+    },
+    {
+      "epoch": 3.0401854714064913,
+      "grad_norm": 0.2769177026846328,
+      "learning_rate": 3.196101791436448e-05,
+      "loss": 0.3713,
+      "step": 1967
+    },
+    {
+      "epoch": 3.041731066460587,
+      "grad_norm": 0.1820364877136236,
+      "learning_rate": 3.191873482600239e-05,
+      "loss": 0.3654,
+      "step": 1968
+    },
+    {
+      "epoch": 3.043276661514683,
+      "grad_norm": 0.18851113345357065,
+      "learning_rate": 3.187646114991622e-05,
+      "loss": 0.3619,
+      "step": 1969
+    },
+    {
+      "epoch": 3.044822256568779,
+      "grad_norm": 0.17214646935687045,
+      "learning_rate": 3.1834196935342254e-05,
+      "loss": 0.368,
+      "step": 1970
+    },
+    {
+      "epoch": 3.046367851622875,
+      "grad_norm": 0.16470405943858682,
+      "learning_rate": 3.179194223150577e-05,
+      "loss": 0.373,
+      "step": 1971
+    },
+    {
+      "epoch": 3.047913446676971,
+      "grad_norm": 0.1776058167891048,
+      "learning_rate": 3.1749697087620954e-05,
+      "loss": 0.3704,
+      "step": 1972
+    },
+    {
+      "epoch": 3.0494590417310663,
+      "grad_norm": 0.18969738255695962,
+      "learning_rate": 3.1707461552890864e-05,
+      "loss": 0.3698,
+      "step": 1973
+    },
+    {
+      "epoch": 3.051004636785162,
+      "grad_norm": 0.1544017040360298,
+      "learning_rate": 3.166523567650737e-05,
+      "loss": 0.3644,
+      "step": 1974
+    },
+    {
+      "epoch": 3.052550231839258,
+      "grad_norm": 0.16631194905369284,
+      "learning_rate": 3.162301950765109e-05,
+      "loss": 0.3682,
+      "step": 1975
+    },
+    {
+      "epoch": 3.054095826893354,
+      "grad_norm": 0.1645757603835096,
+      "learning_rate": 3.1580813095491306e-05,
+      "loss": 0.3614,
+      "step": 1976
+    },
+    {
+      "epoch": 3.05564142194745,
+      "grad_norm": 0.16666207079870846,
+      "learning_rate": 3.1538616489186e-05,
+      "loss": 0.3654,
+      "step": 1977
+    },
+    {
+      "epoch": 3.0571870170015454,
+      "grad_norm": 0.23611859921736,
+      "learning_rate": 3.1496429737881685e-05,
+      "loss": 0.3596,
+      "step": 1978
+    },
+    {
+      "epoch": 3.0587326120556413,
+      "grad_norm": 0.16628564713517838,
+      "learning_rate": 3.145425289071339e-05,
+      "loss": 0.3599,
+      "step": 1979
+    },
+    {
+      "epoch": 3.060278207109737,
+      "grad_norm": 0.14586628384056954,
+      "learning_rate": 3.1412085996804665e-05,
+      "loss": 0.3668,
+      "step": 1980
+    },
+    {
+      "epoch": 3.061823802163833,
+      "grad_norm": 0.17648238084761203,
+      "learning_rate": 3.1369929105267397e-05,
+      "loss": 0.3706,
+      "step": 1981
+    },
+    {
+      "epoch": 3.063369397217929,
+      "grad_norm": 0.17339756474227566,
+      "learning_rate": 3.132778226520185e-05,
+      "loss": 0.355,
+      "step": 1982
+    },
+    {
+      "epoch": 3.064914992272025,
+      "grad_norm": 0.17418137231230973,
+      "learning_rate": 3.128564552569662e-05,
+      "loss": 0.3642,
+      "step": 1983
+    },
+    {
+      "epoch": 3.0664605873261204,
+      "grad_norm": 0.18383584794423377,
+      "learning_rate": 3.124351893582848e-05,
+      "loss": 0.3696,
+      "step": 1984
+    },
+    {
+      "epoch": 3.0680061823802163,
+      "grad_norm": 0.15943423486156244,
+      "learning_rate": 3.120140254466239e-05,
+      "loss": 0.3652,
+      "step": 1985
+    },
+    {
+      "epoch": 3.069551777434312,
+      "grad_norm": 0.13558986598558756,
+      "learning_rate": 3.11592964012515e-05,
+      "loss": 0.3744,
+      "step": 1986
+    },
+    {
+      "epoch": 3.071097372488408,
+      "grad_norm": 0.17240632669153214,
+      "learning_rate": 3.1117200554636944e-05,
+      "loss": 0.3723,
+      "step": 1987
+    },
+    {
+      "epoch": 3.072642967542504,
+      "grad_norm": 0.1298321730384647,
+      "learning_rate": 3.107511505384788e-05,
+      "loss": 0.3668,
+      "step": 1988
+    },
+    {
+      "epoch": 3.0741885625966,
+      "grad_norm": 0.1407898135550854,
+      "learning_rate": 3.1033039947901425e-05,
+      "loss": 0.3674,
+      "step": 1989
+    },
+    {
+      "epoch": 3.0757341576506954,
+      "grad_norm": 0.1451842599816864,
+      "learning_rate": 3.099097528580262e-05,
+      "loss": 0.357,
+      "step": 1990
+    },
+    {
+      "epoch": 3.0772797527047913,
+      "grad_norm": 0.15038363856026724,
+      "learning_rate": 3.0948921116544315e-05,
+      "loss": 0.3619,
+      "step": 1991
+    },
+    {
+      "epoch": 3.078825347758887,
+      "grad_norm": 0.15593874991279674,
+      "learning_rate": 3.090687748910709e-05,
+      "loss": 0.3739,
+      "step": 1992
+    },
+    {
+      "epoch": 3.080370942812983,
+      "grad_norm": 0.17760818245046628,
+      "learning_rate": 3.086484445245935e-05,
+      "loss": 0.3598,
+      "step": 1993
+    },
+    {
+      "epoch": 3.081916537867079,
+      "grad_norm": 0.16189925967772206,
+      "learning_rate": 3.082282205555709e-05,
+      "loss": 0.3683,
+      "step": 1994
+    },
+    {
+      "epoch": 3.0834621329211744,
+      "grad_norm": 0.16397882352854634,
+      "learning_rate": 3.078081034734393e-05,
+      "loss": 0.3653,
+      "step": 1995
+    },
+    {
+      "epoch": 3.0850077279752703,
+      "grad_norm": 0.16873160126894193,
+      "learning_rate": 3.073880937675104e-05,
+      "loss": 0.3707,
+      "step": 1996
+    },
+    {
+      "epoch": 3.0865533230293662,
+      "grad_norm": 0.4887966427758655,
+      "learning_rate": 3.0696819192697094e-05,
+      "loss": 0.3704,
+      "step": 1997
+    },
+    {
+      "epoch": 3.088098918083462,
+      "grad_norm": 0.16174679696377703,
+      "learning_rate": 3.065483984408819e-05,
+      "loss": 0.3672,
+      "step": 1998
+    },
+    {
+      "epoch": 3.089644513137558,
+      "grad_norm": 0.1614323899965815,
+      "learning_rate": 3.061287137981783e-05,
+      "loss": 0.3619,
+      "step": 1999
+    },
+    {
+      "epoch": 3.091190108191654,
+      "grad_norm": 0.16531952527462507,
+      "learning_rate": 3.05709138487668e-05,
+      "loss": 0.3608,
+      "step": 2000
+    },
+    {
+      "epoch": 3.0927357032457494,
+      "grad_norm": 0.1665889090936738,
+      "learning_rate": 3.052896729980318e-05,
+      "loss": 0.3628,
+      "step": 2001
+    },
+    {
+      "epoch": 3.0942812982998453,
+      "grad_norm": 0.17925722528029708,
+      "learning_rate": 3.0487031781782265e-05,
+      "loss": 0.3638,
+      "step": 2002
+    },
+    {
+      "epoch": 3.0958268933539412,
+      "grad_norm": 0.15251213414087034,
+      "learning_rate": 3.0445107343546474e-05,
+      "loss": 0.3719,
+      "step": 2003
+    },
+    {
+      "epoch": 3.097372488408037,
+      "grad_norm": 0.15268078066226246,
+      "learning_rate": 3.040319403392533e-05,
+      "loss": 0.3639,
+      "step": 2004
+    },
+    {
+      "epoch": 3.098918083462133,
+      "grad_norm": 0.1919213576284206,
+      "learning_rate": 3.0361291901735428e-05,
+      "loss": 0.3778,
+      "step": 2005
+    },
+    {
+      "epoch": 3.100463678516229,
+      "grad_norm": 0.1661459402079912,
+      "learning_rate": 3.0319400995780305e-05,
+      "loss": 0.3642,
+      "step": 2006
+    },
+    {
+      "epoch": 3.1020092735703244,
+      "grad_norm": 0.17391255717889237,
+      "learning_rate": 3.0277521364850426e-05,
+      "loss": 0.3732,
+      "step": 2007
+    },
+    {
+      "epoch": 3.1035548686244203,
+      "grad_norm": 0.15570388212781644,
+      "learning_rate": 3.0235653057723148e-05,
+      "loss": 0.3627,
+      "step": 2008
+    },
+    {
+      "epoch": 3.105100463678516,
+      "grad_norm": 0.1657068281966001,
+      "learning_rate": 3.0193796123162617e-05,
+      "loss": 0.365,
+      "step": 2009
+    },
+    {
+      "epoch": 3.106646058732612,
+      "grad_norm": 0.1644518030320822,
+      "learning_rate": 3.0151950609919733e-05,
+      "loss": 0.3595,
+      "step": 2010
+    },
+    {
+      "epoch": 3.108191653786708,
+      "grad_norm": 0.15149853097947524,
+      "learning_rate": 3.0110116566732132e-05,
+      "loss": 0.3613,
+      "step": 2011
+    },
+    {
+      "epoch": 3.109737248840804,
+      "grad_norm": 0.20400354964088088,
+      "learning_rate": 3.006829404232404e-05,
+      "loss": 0.3682,
+      "step": 2012
+    },
+    {
+      "epoch": 3.1112828438948994,
+      "grad_norm": 0.1591073786268792,
+      "learning_rate": 3.0026483085406283e-05,
+      "loss": 0.3732,
+      "step": 2013
+    },
+    {
+      "epoch": 3.1128284389489953,
+      "grad_norm": 0.1570119612980249,
+      "learning_rate": 2.9984683744676206e-05,
+      "loss": 0.3659,
+      "step": 2014
+    },
+    {
+      "epoch": 3.114374034003091,
+      "grad_norm": 0.17084269419017303,
+      "learning_rate": 2.994289606881769e-05,
+      "loss": 0.3738,
+      "step": 2015
+    },
+    {
+      "epoch": 3.115919629057187,
+      "grad_norm": 0.14029104939344528,
+      "learning_rate": 2.9901120106500924e-05,
+      "loss": 0.3612,
+      "step": 2016
+    },
+    {
+      "epoch": 3.117465224111283,
+      "grad_norm": 0.1589995205079017,
+      "learning_rate": 2.9859355906382517e-05,
+      "loss": 0.3613,
+      "step": 2017
+    },
+    {
+      "epoch": 3.1190108191653785,
+      "grad_norm": 0.16915103629839312,
+      "learning_rate": 2.98176035171054e-05,
+      "loss": 0.373,
+      "step": 2018
+    },
+    {
+      "epoch": 3.1205564142194744,
+      "grad_norm": 0.13703361557235325,
+      "learning_rate": 2.97758629872987e-05,
+      "loss": 0.359,
+      "step": 2019
+    },
+    {
+      "epoch": 3.1221020092735703,
+      "grad_norm": 0.22441100620131804,
+      "learning_rate": 2.9734134365577723e-05,
+      "loss": 0.3694,
+      "step": 2020
+    },
+    {
+      "epoch": 3.123647604327666,
+      "grad_norm": 0.1478916041558205,
+      "learning_rate": 2.9692417700543966e-05,
+      "loss": 0.3638,
+      "step": 2021
+    },
+    {
+      "epoch": 3.125193199381762,
+      "grad_norm": 0.2096432660700668,
+      "learning_rate": 2.9650713040784954e-05,
+      "loss": 0.3644,
+      "step": 2022
+    },
+    {
+      "epoch": 3.126738794435858,
+      "grad_norm": 0.16526543441751307,
+      "learning_rate": 2.960902043487423e-05,
+      "loss": 0.3618,
+      "step": 2023
+    },
+    {
+      "epoch": 3.1282843894899535,
+      "grad_norm": 0.17324589785159916,
+      "learning_rate": 2.9567339931371327e-05,
+      "loss": 0.3743,
+      "step": 2024
+    },
+    {
+      "epoch": 3.1298299845440494,
+      "grad_norm": 0.1389638279469882,
+      "learning_rate": 2.9525671578821647e-05,
+      "loss": 0.3629,
+      "step": 2025
+    },
+    {
+      "epoch": 3.1313755795981453,
+      "grad_norm": 0.18503881736460734,
+      "learning_rate": 2.9484015425756454e-05,
+      "loss": 0.374,
+      "step": 2026
+    },
+    {
+      "epoch": 3.132921174652241,
+      "grad_norm": 0.12587601505049495,
+      "learning_rate": 2.9442371520692825e-05,
+      "loss": 0.3579,
+      "step": 2027
+    },
+    {
+      "epoch": 3.134466769706337,
+      "grad_norm": 0.1654549667985164,
+      "learning_rate": 2.9400739912133543e-05,
+      "loss": 0.3711,
+      "step": 2028
+    },
+    {
+      "epoch": 3.1360123647604325,
+      "grad_norm": 0.13126514642819836,
+      "learning_rate": 2.9359120648567062e-05,
+      "loss": 0.3665,
+      "step": 2029
+    },
+    {
+      "epoch": 3.1375579598145285,
+      "grad_norm": 0.1433166542494949,
+      "learning_rate": 2.9317513778467498e-05,
+      "loss": 0.3651,
+      "step": 2030
+    },
+    {
+      "epoch": 3.1391035548686244,
+      "grad_norm": 0.14785582958072702,
+      "learning_rate": 2.9275919350294498e-05,
+      "loss": 0.3599,
+      "step": 2031
+    },
+    {
+      "epoch": 3.1406491499227203,
+      "grad_norm": 0.1627068357890892,
+      "learning_rate": 2.9234337412493218e-05,
+      "loss": 0.372,
+      "step": 2032
+    },
+    {
+      "epoch": 3.142194744976816,
+      "grad_norm": 0.15647817397173375,
+      "learning_rate": 2.9192768013494297e-05,
+      "loss": 0.3666,
+      "step": 2033
+    },
+    {
+      "epoch": 3.143740340030912,
+      "grad_norm": 0.14213938932745243,
+      "learning_rate": 2.9151211201713744e-05,
+      "loss": 0.3661,
+      "step": 2034
+    },
+    {
+      "epoch": 3.1452859350850075,
+      "grad_norm": 0.13130075114364653,
+      "learning_rate": 2.91096670255529e-05,
+      "loss": 0.3577,
+      "step": 2035
+    },
+    {
+      "epoch": 3.1468315301391034,
+      "grad_norm": 0.14255253995836262,
+      "learning_rate": 2.9068135533398415e-05,
+      "loss": 0.3575,
+      "step": 2036
+    },
+    {
+      "epoch": 3.1483771251931993,
+      "grad_norm": 0.14334987938477375,
+      "learning_rate": 2.902661677362215e-05,
+      "loss": 0.3687,
+      "step": 2037
+    },
+    {
+      "epoch": 3.1499227202472952,
+      "grad_norm": 0.1493354947333077,
+      "learning_rate": 2.898511079458113e-05,
+      "loss": 0.3758,
+      "step": 2038
+    },
+    {
+      "epoch": 3.151468315301391,
+      "grad_norm": 0.12377396402553244,
+      "learning_rate": 2.8943617644617536e-05,
+      "loss": 0.368,
+      "step": 2039
+    },
+    {
+      "epoch": 3.153013910355487,
+      "grad_norm": 0.14593946865969992,
+      "learning_rate": 2.890213737205854e-05,
+      "loss": 0.3654,
+      "step": 2040
+    },
+    {
+      "epoch": 3.1545595054095825,
+      "grad_norm": 0.13530777880989286,
+      "learning_rate": 2.886067002521637e-05,
+      "loss": 0.3715,
+      "step": 2041
+    },
+    {
+      "epoch": 3.1561051004636784,
+      "grad_norm": 0.14577397239207665,
+      "learning_rate": 2.8819215652388162e-05,
+      "loss": 0.3669,
+      "step": 2042
+    },
+    {
+      "epoch": 3.1576506955177743,
+      "grad_norm": 0.15068942998498627,
+      "learning_rate": 2.877777430185599e-05,
+      "loss": 0.3603,
+      "step": 2043
+    },
+    {
+      "epoch": 3.1591962905718702,
+      "grad_norm": 0.1403873402782044,
+      "learning_rate": 2.873634602188671e-05,
+      "loss": 0.3618,
+      "step": 2044
+    },
+    {
+      "epoch": 3.160741885625966,
+      "grad_norm": 0.14312343061212757,
+      "learning_rate": 2.8694930860731954e-05,
+      "loss": 0.3609,
+      "step": 2045
+    },
+    {
+      "epoch": 3.162287480680062,
+      "grad_norm": 0.12439940271897412,
+      "learning_rate": 2.8653528866628132e-05,
+      "loss": 0.366,
+      "step": 2046
+    },
+    {
+      "epoch": 3.1638330757341575,
+      "grad_norm": 0.15363023129281814,
+      "learning_rate": 2.8612140087796282e-05,
+      "loss": 0.3599,
+      "step": 2047
+    },
+    {
+      "epoch": 3.1653786707882534,
+      "grad_norm": 0.15970831874172828,
+      "learning_rate": 2.8570764572442012e-05,
+      "loss": 0.3708,
+      "step": 2048
+    },
+    {
+      "epoch": 3.1669242658423493,
+      "grad_norm": 0.14084145298957124,
+      "learning_rate": 2.852940236875556e-05,
+      "loss": 0.3683,
+      "step": 2049
+    },
+    {
+      "epoch": 3.1684698608964452,
+      "grad_norm": 0.17499188472712682,
+      "learning_rate": 2.848805352491161e-05,
+      "loss": 0.361,
+      "step": 2050
+    },
+    {
+      "epoch": 3.170015455950541,
+      "grad_norm": 0.14425219613022813,
+      "learning_rate": 2.8446718089069264e-05,
+      "loss": 0.37,
+      "step": 2051
+    },
+    {
+      "epoch": 3.1715610510046366,
+      "grad_norm": 0.1251859772791234,
+      "learning_rate": 2.840539610937208e-05,
+      "loss": 0.3623,
+      "step": 2052
+    },
+    {
+      "epoch": 3.1731066460587325,
+      "grad_norm": 0.12579663061900387,
+      "learning_rate": 2.8364087633947887e-05,
+      "loss": 0.3583,
+      "step": 2053
+    },
+    {
+      "epoch": 3.1746522411128284,
+      "grad_norm": 0.12251358173410634,
+      "learning_rate": 2.8322792710908787e-05,
+      "loss": 0.3623,
+      "step": 2054
+    },
+    {
+      "epoch": 3.1761978361669243,
+      "grad_norm": 0.13829595686979887,
+      "learning_rate": 2.828151138835114e-05,
+      "loss": 0.3613,
+      "step": 2055
+    },
+    {
+      "epoch": 3.17774343122102,
+      "grad_norm": 0.1381224917859969,
+      "learning_rate": 2.8240243714355416e-05,
+      "loss": 0.3658,
+      "step": 2056
+    },
+    {
+      "epoch": 3.179289026275116,
+      "grad_norm": 0.14834068382748958,
+      "learning_rate": 2.8198989736986212e-05,
+      "loss": 0.3774,
+      "step": 2057
+    },
+    {
+      "epoch": 3.1808346213292116,
+      "grad_norm": 0.12234098986699041,
+      "learning_rate": 2.8157749504292185e-05,
+      "loss": 0.3632,
+      "step": 2058
+    },
+    {
+      "epoch": 3.1823802163833075,
+      "grad_norm": 0.14951922074387244,
+      "learning_rate": 2.811652306430595e-05,
+      "loss": 0.3695,
+      "step": 2059
+    },
+    {
+      "epoch": 3.1839258114374034,
+      "grad_norm": 0.12717372719527187,
+      "learning_rate": 2.8075310465044076e-05,
+      "loss": 0.368,
+      "step": 2060
+    },
+    {
+      "epoch": 3.1854714064914993,
+      "grad_norm": 0.13522718209919893,
+      "learning_rate": 2.8034111754507032e-05,
+      "loss": 0.368,
+      "step": 2061
+    },
+    {
+      "epoch": 3.187017001545595,
+      "grad_norm": 0.12239250181348509,
+      "learning_rate": 2.799292698067907e-05,
+      "loss": 0.3636,
+      "step": 2062
+    },
+    {
+      "epoch": 3.1885625965996907,
+      "grad_norm": 0.13502088259022943,
+      "learning_rate": 2.795175619152824e-05,
+      "loss": 0.3646,
+      "step": 2063
+    },
+    {
+      "epoch": 3.1901081916537866,
+      "grad_norm": 0.12908259748447184,
+      "learning_rate": 2.7910599435006297e-05,
+      "loss": 0.3724,
+      "step": 2064
+    },
+    {
+      "epoch": 3.1916537867078825,
+      "grad_norm": 0.12814919445837764,
+      "learning_rate": 2.7869456759048644e-05,
+      "loss": 0.3683,
+      "step": 2065
+    },
+    {
+      "epoch": 3.1931993817619784,
+      "grad_norm": 0.14163890351082115,
+      "learning_rate": 2.7828328211574277e-05,
+      "loss": 0.3675,
+      "step": 2066
+    },
+    {
+      "epoch": 3.1947449768160743,
+      "grad_norm": 0.1372701812449613,
+      "learning_rate": 2.778721384048579e-05,
+      "loss": 0.3705,
+      "step": 2067
+    },
+    {
+      "epoch": 3.19629057187017,
+      "grad_norm": 0.15204661234673036,
+      "learning_rate": 2.7746113693669186e-05,
+      "loss": 0.3674,
+      "step": 2068
+    },
+    {
+      "epoch": 3.1978361669242656,
+      "grad_norm": 0.15115300068077478,
+      "learning_rate": 2.7705027818993958e-05,
+      "loss": 0.3729,
+      "step": 2069
+    },
+    {
+      "epoch": 3.1993817619783615,
+      "grad_norm": 0.14370368896428348,
+      "learning_rate": 2.7663956264312943e-05,
+      "loss": 0.367,
+      "step": 2070
+    },
+    {
+      "epoch": 3.2009273570324575,
+      "grad_norm": 0.14968474740202153,
+      "learning_rate": 2.7622899077462335e-05,
+      "loss": 0.3713,
+      "step": 2071
+    },
+    {
+      "epoch": 3.2024729520865534,
+      "grad_norm": 0.1686462967770937,
+      "learning_rate": 2.758185630626156e-05,
+      "loss": 0.3684,
+      "step": 2072
+    },
+    {
+      "epoch": 3.2040185471406493,
+      "grad_norm": 0.1536991268040143,
+      "learning_rate": 2.7540827998513263e-05,
+      "loss": 0.3653,
+      "step": 2073
+    },
+    {
+      "epoch": 3.205564142194745,
+      "grad_norm": 0.14785756813895054,
+      "learning_rate": 2.7499814202003272e-05,
+      "loss": 0.3672,
+      "step": 2074
+    },
+    {
+      "epoch": 3.2071097372488406,
+      "grad_norm": 0.16013222543831185,
+      "learning_rate": 2.7458814964500473e-05,
+      "loss": 0.3689,
+      "step": 2075
+    },
+    {
+      "epoch": 3.2086553323029365,
+      "grad_norm": 0.1534977441030092,
+      "learning_rate": 2.741783033375679e-05,
+      "loss": 0.3704,
+      "step": 2076
+    },
+    {
+      "epoch": 3.2102009273570324,
+      "grad_norm": 0.17365777916889796,
+      "learning_rate": 2.73768603575072e-05,
+      "loss": 0.3659,
+      "step": 2077
+    },
+    {
+      "epoch": 3.2117465224111283,
+      "grad_norm": 0.14489892162203102,
+      "learning_rate": 2.7335905083469555e-05,
+      "loss": 0.3658,
+      "step": 2078
+    },
+    {
+      "epoch": 3.2132921174652243,
+      "grad_norm": 0.1777579032834708,
+      "learning_rate": 2.7294964559344563e-05,
+      "loss": 0.3678,
+      "step": 2079
+    },
+    {
+      "epoch": 3.21483771251932,
+      "grad_norm": 0.17040592808564636,
+      "learning_rate": 2.725403883281583e-05,
+      "loss": 0.3642,
+      "step": 2080
+    },
+    {
+      "epoch": 3.2163833075734156,
+      "grad_norm": 0.1881164082243543,
+      "learning_rate": 2.721312795154966e-05,
+      "loss": 0.3731,
+      "step": 2081
+    },
+    {
+      "epoch": 3.2179289026275115,
+      "grad_norm": 0.18418639695158592,
+      "learning_rate": 2.7172231963195085e-05,
+      "loss": 0.3681,
+      "step": 2082
+    },
+    {
+      "epoch": 3.2194744976816074,
+      "grad_norm": 0.2082446090608705,
+      "learning_rate": 2.713135091538382e-05,
+      "loss": 0.3653,
+      "step": 2083
+    },
+    {
+      "epoch": 3.2210200927357033,
+      "grad_norm": 0.16220777622045765,
+      "learning_rate": 2.7090484855730136e-05,
+      "loss": 0.3738,
+      "step": 2084
+    },
+    {
+      "epoch": 3.2225656877897992,
+      "grad_norm": 0.15783064457328178,
+      "learning_rate": 2.7049633831830867e-05,
+      "loss": 0.3643,
+      "step": 2085
+    },
+    {
+      "epoch": 3.2241112828438947,
+      "grad_norm": 0.15693558187552553,
+      "learning_rate": 2.7008797891265336e-05,
+      "loss": 0.3704,
+      "step": 2086
+    },
+    {
+      "epoch": 3.2256568778979906,
+      "grad_norm": 0.14758748067377753,
+      "learning_rate": 2.6967977081595304e-05,
+      "loss": 0.3652,
+      "step": 2087
+    },
+    {
+      "epoch": 3.2272024729520865,
+      "grad_norm": 0.16748703002722784,
+      "learning_rate": 2.692717145036489e-05,
+      "loss": 0.3653,
+      "step": 2088
+    },
+    {
+      "epoch": 3.2287480680061824,
+      "grad_norm": 0.15125269311815615,
+      "learning_rate": 2.6886381045100552e-05,
+      "loss": 0.3656,
+      "step": 2089
+    },
+    {
+      "epoch": 3.2302936630602783,
+      "grad_norm": 0.15544533844342912,
+      "learning_rate": 2.6845605913311012e-05,
+      "loss": 0.3682,
+      "step": 2090
+    },
+    {
+      "epoch": 3.2318392581143742,
+      "grad_norm": 0.1562434423567851,
+      "learning_rate": 2.6804846102487182e-05,
+      "loss": 0.3717,
+      "step": 2091
+    },
+    {
+      "epoch": 3.2333848531684697,
+      "grad_norm": 0.14291583957621115,
+      "learning_rate": 2.676410166010217e-05,
+      "loss": 0.3767,
+      "step": 2092
+    },
+    {
+      "epoch": 3.2349304482225656,
+      "grad_norm": 0.1357194755115203,
+      "learning_rate": 2.6723372633611147e-05,
+      "loss": 0.3651,
+      "step": 2093
+    },
+    {
+      "epoch": 3.2364760432766615,
+      "grad_norm": 0.17378316610340208,
+      "learning_rate": 2.6682659070451338e-05,
+      "loss": 0.3632,
+      "step": 2094
+    },
+    {
+      "epoch": 3.2380216383307574,
+      "grad_norm": 0.12682019037929812,
+      "learning_rate": 2.6641961018041978e-05,
+      "loss": 0.3672,
+      "step": 2095
+    },
+    {
+      "epoch": 3.2395672333848533,
+      "grad_norm": 0.17406933720096976,
+      "learning_rate": 2.6601278523784214e-05,
+      "loss": 0.368,
+      "step": 2096
+    },
+    {
+      "epoch": 3.2411128284389488,
+      "grad_norm": 0.14091612975788875,
+      "learning_rate": 2.6560611635061083e-05,
+      "loss": 0.3734,
+      "step": 2097
+    },
+    {
+      "epoch": 3.2426584234930447,
+      "grad_norm": 0.16253732307206356,
+      "learning_rate": 2.6519960399237434e-05,
+      "loss": 0.3733,
+      "step": 2098
+    },
+    {
+      "epoch": 3.2442040185471406,
+      "grad_norm": 0.15080109666228644,
+      "learning_rate": 2.6479324863659897e-05,
+      "loss": 0.3675,
+      "step": 2099
+    },
+    {
+      "epoch": 3.2457496136012365,
+      "grad_norm": 0.15434815098317825,
+      "learning_rate": 2.6438705075656822e-05,
+      "loss": 0.3657,
+      "step": 2100
+    },
+    {
+      "epoch": 3.2472952086553324,
+      "grad_norm": 0.15767668819755418,
+      "learning_rate": 2.6398101082538187e-05,
+      "loss": 0.3838,
+      "step": 2101
+    },
+    {
+      "epoch": 3.2488408037094283,
+      "grad_norm": 0.13494582357863177,
+      "learning_rate": 2.6357512931595634e-05,
+      "loss": 0.3638,
+      "step": 2102
+    },
+    {
+      "epoch": 3.250386398763524,
+      "grad_norm": 0.15718928860212597,
+      "learning_rate": 2.631694067010228e-05,
+      "loss": 0.3752,
+      "step": 2103
+    },
+    {
+      "epoch": 3.2519319938176197,
+      "grad_norm": 0.1535896044874011,
+      "learning_rate": 2.6276384345312772e-05,
+      "loss": 0.3688,
+      "step": 2104
+    },
+    {
+      "epoch": 3.2534775888717156,
+      "grad_norm": 0.14398376836227514,
+      "learning_rate": 2.6235844004463217e-05,
+      "loss": 0.3624,
+      "step": 2105
+    },
+    {
+      "epoch": 3.2550231839258115,
+      "grad_norm": 0.12886124949716404,
+      "learning_rate": 2.619531969477108e-05,
+      "loss": 0.3604,
+      "step": 2106
+    },
+    {
+      "epoch": 3.2565687789799074,
+      "grad_norm": 0.20057890482614704,
+      "learning_rate": 2.6154811463435116e-05,
+      "loss": 0.3754,
+      "step": 2107
+    },
+    {
+      "epoch": 3.2581143740340033,
+      "grad_norm": 0.13665360927560344,
+      "learning_rate": 2.6114319357635447e-05,
+      "loss": 0.3701,
+      "step": 2108
+    },
+    {
+      "epoch": 3.2596599690880987,
+      "grad_norm": 0.16106687173917297,
+      "learning_rate": 2.6073843424533332e-05,
+      "loss": 0.3626,
+      "step": 2109
+    },
+    {
+      "epoch": 3.2612055641421946,
+      "grad_norm": 0.13092642496388196,
+      "learning_rate": 2.6033383711271222e-05,
+      "loss": 0.3602,
+      "step": 2110
+    },
+    {
+      "epoch": 3.2627511591962906,
+      "grad_norm": 0.15358668926841232,
+      "learning_rate": 2.599294026497269e-05,
+      "loss": 0.3643,
+      "step": 2111
+    },
+    {
+      "epoch": 3.2642967542503865,
+      "grad_norm": 0.14606690153826582,
+      "learning_rate": 2.5952513132742346e-05,
+      "loss": 0.3658,
+      "step": 2112
+    },
+    {
+      "epoch": 3.2658423493044824,
+      "grad_norm": 0.1378753572686257,
+      "learning_rate": 2.5912102361665793e-05,
+      "loss": 0.37,
+      "step": 2113
+    },
+    {
+      "epoch": 3.2673879443585783,
+      "grad_norm": 0.14687750965345156,
+      "learning_rate": 2.5871707998809594e-05,
+      "loss": 0.365,
+      "step": 2114
+    },
+    {
+      "epoch": 3.2689335394126737,
+      "grad_norm": 0.14819762397571287,
+      "learning_rate": 2.5831330091221203e-05,
+      "loss": 0.3797,
+      "step": 2115
+    },
+    {
+      "epoch": 3.2704791344667696,
+      "grad_norm": 0.12710955138828492,
+      "learning_rate": 2.5790968685928887e-05,
+      "loss": 0.3732,
+      "step": 2116
+    },
+    {
+      "epoch": 3.2720247295208655,
+      "grad_norm": 0.15343239107124537,
+      "learning_rate": 2.575062382994172e-05,
+      "loss": 0.3739,
+      "step": 2117
+    },
+    {
+      "epoch": 3.2735703245749614,
+      "grad_norm": 0.1416408721693487,
+      "learning_rate": 2.5710295570249485e-05,
+      "loss": 0.3694,
+      "step": 2118
+    },
+    {
+      "epoch": 3.2751159196290573,
+      "grad_norm": 0.13009832790542777,
+      "learning_rate": 2.5669983953822616e-05,
+      "loss": 0.3742,
+      "step": 2119
+    },
+    {
+      "epoch": 3.276661514683153,
+      "grad_norm": 0.14417898261561116,
+      "learning_rate": 2.5629689027612213e-05,
+      "loss": 0.3711,
+      "step": 2120
+    },
+    {
+      "epoch": 3.2782071097372487,
+      "grad_norm": 0.13507266005557989,
+      "learning_rate": 2.5589410838549902e-05,
+      "loss": 0.3756,
+      "step": 2121
+    },
+    {
+      "epoch": 3.2797527047913446,
+      "grad_norm": 0.1650121483203225,
+      "learning_rate": 2.5549149433547803e-05,
+      "loss": 0.3703,
+      "step": 2122
+    },
+    {
+      "epoch": 3.2812982998454405,
+      "grad_norm": 0.14953876290809184,
+      "learning_rate": 2.5508904859498528e-05,
+      "loss": 0.3604,
+      "step": 2123
+    },
+    {
+      "epoch": 3.2828438948995364,
+      "grad_norm": 0.14221545791884788,
+      "learning_rate": 2.5468677163275056e-05,
+      "loss": 0.3672,
+      "step": 2124
+    },
+    {
+      "epoch": 3.2843894899536323,
+      "grad_norm": 0.14399205076304955,
+      "learning_rate": 2.542846639173071e-05,
+      "loss": 0.3681,
+      "step": 2125
+    },
+    {
+      "epoch": 3.285935085007728,
+      "grad_norm": 0.14904816147074224,
+      "learning_rate": 2.538827259169911e-05,
+      "loss": 0.3717,
+      "step": 2126
+    },
+    {
+      "epoch": 3.2874806800618237,
+      "grad_norm": 0.14830345453722532,
+      "learning_rate": 2.5348095809994113e-05,
+      "loss": 0.3694,
+      "step": 2127
+    },
+    {
+      "epoch": 3.2890262751159196,
+      "grad_norm": 0.13985328837797856,
+      "learning_rate": 2.530793609340974e-05,
+      "loss": 0.3779,
+      "step": 2128
+    },
+    {
+      "epoch": 3.2905718701700155,
+      "grad_norm": 0.17710821167389063,
+      "learning_rate": 2.526779348872013e-05,
+      "loss": 0.366,
+      "step": 2129
+    },
+    {
+      "epoch": 3.2921174652241114,
+      "grad_norm": 0.14343141353829914,
+      "learning_rate": 2.522766804267955e-05,
+      "loss": 0.3699,
+      "step": 2130
+    },
+    {
+      "epoch": 3.293663060278207,
+      "grad_norm": 0.19980441120996223,
+      "learning_rate": 2.5187559802022196e-05,
+      "loss": 0.3693,
+      "step": 2131
+    },
+    {
+      "epoch": 3.295208655332303,
+      "grad_norm": 0.15717682802009292,
+      "learning_rate": 2.514746881346227e-05,
+      "loss": 0.3682,
+      "step": 2132
+    },
+    {
+      "epoch": 3.2967542503863987,
+      "grad_norm": 0.1886639544522163,
+      "learning_rate": 2.5107395123693908e-05,
+      "loss": 0.3641,
+      "step": 2133
+    },
+    {
+      "epoch": 3.2982998454404946,
+      "grad_norm": 0.17588396501052606,
+      "learning_rate": 2.5067338779391044e-05,
+      "loss": 0.3603,
+      "step": 2134
+    },
+    {
+      "epoch": 3.2998454404945905,
+      "grad_norm": 0.17319305501363041,
+      "learning_rate": 2.502729982720742e-05,
+      "loss": 0.3762,
+      "step": 2135
+    },
+    {
+      "epoch": 3.3013910355486864,
+      "grad_norm": 0.17074748630118594,
+      "learning_rate": 2.498727831377656e-05,
+      "loss": 0.3734,
+      "step": 2136
+    },
+    {
+      "epoch": 3.3029366306027823,
+      "grad_norm": 0.13590596256129128,
+      "learning_rate": 2.4947274285711657e-05,
+      "loss": 0.3637,
+      "step": 2137
+    },
+    {
+      "epoch": 3.3044822256568778,
+      "grad_norm": 0.18090013616044767,
+      "learning_rate": 2.4907287789605485e-05,
+      "loss": 0.3706,
+      "step": 2138
+    },
+    {
+      "epoch": 3.3060278207109737,
+      "grad_norm": 0.144784935494973,
+      "learning_rate": 2.4867318872030503e-05,
+      "loss": 0.373,
+      "step": 2139
+    },
+    {
+      "epoch": 3.3075734157650696,
+      "grad_norm": 0.1891658448891262,
+      "learning_rate": 2.482736757953861e-05,
+      "loss": 0.3631,
+      "step": 2140
+    },
+    {
+      "epoch": 3.3091190108191655,
+      "grad_norm": 0.14667963718248692,
+      "learning_rate": 2.478743395866121e-05,
+      "loss": 0.3708,
+      "step": 2141
+    },
+    {
+      "epoch": 3.3106646058732614,
+      "grad_norm": 0.21035594167766744,
+      "learning_rate": 2.4747518055909127e-05,
+      "loss": 0.377,
+      "step": 2142
+    },
+    {
+      "epoch": 3.312210200927357,
+      "grad_norm": 0.16983171943181535,
+      "learning_rate": 2.4707619917772547e-05,
+      "loss": 0.3666,
+      "step": 2143
+    },
+    {
+      "epoch": 3.3137557959814528,
+      "grad_norm": 0.20078565648335242,
+      "learning_rate": 2.466773959072094e-05,
+      "loss": 0.3706,
+      "step": 2144
+    },
+    {
+      "epoch": 3.3153013910355487,
+      "grad_norm": 0.1877400727983853,
+      "learning_rate": 2.4627877121203082e-05,
+      "loss": 0.3703,
+      "step": 2145
+    },
+    {
+      "epoch": 3.3168469860896446,
+      "grad_norm": 0.16510685650538093,
+      "learning_rate": 2.4588032555646905e-05,
+      "loss": 0.3726,
+      "step": 2146
+    },
+    {
+      "epoch": 3.3183925811437405,
+      "grad_norm": 0.1528367488160912,
+      "learning_rate": 2.4548205940459502e-05,
+      "loss": 0.365,
+      "step": 2147
+    },
+    {
+      "epoch": 3.3199381761978364,
+      "grad_norm": 0.15434010432237918,
+      "learning_rate": 2.450839732202707e-05,
+      "loss": 0.3642,
+      "step": 2148
+    },
+    {
+      "epoch": 3.321483771251932,
+      "grad_norm": 0.1552317292573876,
+      "learning_rate": 2.4468606746714835e-05,
+      "loss": 0.3656,
+      "step": 2149
+    },
+    {
+      "epoch": 3.3230293663060277,
+      "grad_norm": 0.1575056447538431,
+      "learning_rate": 2.4428834260867e-05,
+      "loss": 0.3735,
+      "step": 2150
+    },
+    {
+      "epoch": 3.3245749613601236,
+      "grad_norm": 0.1623900587467734,
+      "learning_rate": 2.4389079910806728e-05,
+      "loss": 0.3677,
+      "step": 2151
+    },
+    {
+      "epoch": 3.3261205564142196,
+      "grad_norm": 0.1321736194562474,
+      "learning_rate": 2.4349343742836028e-05,
+      "loss": 0.3663,
+      "step": 2152
+    },
+    {
+      "epoch": 3.3276661514683155,
+      "grad_norm": 0.126420491100854,
+      "learning_rate": 2.4309625803235744e-05,
+      "loss": 0.3665,
+      "step": 2153
+    },
+    {
+      "epoch": 3.329211746522411,
+      "grad_norm": 0.12932422188472273,
+      "learning_rate": 2.426992613826549e-05,
+      "loss": 0.362,
+      "step": 2154
+    },
+    {
+      "epoch": 3.330757341576507,
+      "grad_norm": 0.1373101315845367,
+      "learning_rate": 2.42302447941636e-05,
+      "loss": 0.3698,
+      "step": 2155
+    },
+    {
+      "epoch": 3.3323029366306027,
+      "grad_norm": 0.16669707516861867,
+      "learning_rate": 2.4190581817147063e-05,
+      "loss": 0.358,
+      "step": 2156
+    },
+    {
+      "epoch": 3.3338485316846986,
+      "grad_norm": 0.13589432515355804,
+      "learning_rate": 2.415093725341147e-05,
+      "loss": 0.3626,
+      "step": 2157
+    },
+    {
+      "epoch": 3.3353941267387945,
+      "grad_norm": 0.16452255693638304,
+      "learning_rate": 2.4111311149130993e-05,
+      "loss": 0.3535,
+      "step": 2158
+    },
+    {
+      "epoch": 3.3369397217928904,
+      "grad_norm": 0.14103684315072043,
+      "learning_rate": 2.4071703550458273e-05,
+      "loss": 0.374,
+      "step": 2159
+    },
+    {
+      "epoch": 3.338485316846986,
+      "grad_norm": 0.1466722795459836,
+      "learning_rate": 2.403211450352439e-05,
+      "loss": 0.3605,
+      "step": 2160
+    },
+    {
+      "epoch": 3.340030911901082,
+      "grad_norm": 0.14758435816430387,
+      "learning_rate": 2.3992544054438877e-05,
+      "loss": 0.3643,
+      "step": 2161
+    },
+    {
+      "epoch": 3.3415765069551777,
+      "grad_norm": 0.14523145992398567,
+      "learning_rate": 2.3952992249289542e-05,
+      "loss": 0.3682,
+      "step": 2162
+    },
+    {
+      "epoch": 3.3431221020092736,
+      "grad_norm": 0.14600152871059888,
+      "learning_rate": 2.3913459134142482e-05,
+      "loss": 0.3656,
+      "step": 2163
+    },
+    {
+      "epoch": 3.3446676970633695,
+      "grad_norm": 0.1205791485784816,
+      "learning_rate": 2.3873944755042088e-05,
+      "loss": 0.364,
+      "step": 2164
+    },
+    {
+      "epoch": 3.346213292117465,
+      "grad_norm": 0.1429536650084363,
+      "learning_rate": 2.3834449158010866e-05,
+      "loss": 0.3645,
+      "step": 2165
+    },
+    {
+      "epoch": 3.347758887171561,
+      "grad_norm": 0.13917271002419995,
+      "learning_rate": 2.3794972389049448e-05,
+      "loss": 0.3688,
+      "step": 2166
+    },
+    {
+      "epoch": 3.349304482225657,
+      "grad_norm": 0.1374963736951724,
+      "learning_rate": 2.3755514494136583e-05,
+      "loss": 0.3737,
+      "step": 2167
+    },
+    {
+      "epoch": 3.3508500772797527,
+      "grad_norm": 0.13850266898961983,
+      "learning_rate": 2.3716075519229004e-05,
+      "loss": 0.3593,
+      "step": 2168
+    },
+    {
+      "epoch": 3.3523956723338486,
+      "grad_norm": 0.1299321402219489,
+      "learning_rate": 2.3676655510261402e-05,
+      "loss": 0.3601,
+      "step": 2169
+    },
+    {
+      "epoch": 3.3539412673879445,
+      "grad_norm": 0.1534785499695775,
+      "learning_rate": 2.3637254513146406e-05,
+      "loss": 0.3618,
+      "step": 2170
+    },
+    {
+      "epoch": 3.3554868624420404,
+      "grad_norm": 0.13334938950825415,
+      "learning_rate": 2.3597872573774477e-05,
+      "loss": 0.3752,
+      "step": 2171
+    },
+    {
+      "epoch": 3.357032457496136,
+      "grad_norm": 0.1400025344417481,
+      "learning_rate": 2.3558509738013884e-05,
+      "loss": 0.361,
+      "step": 2172
+    },
+    {
+      "epoch": 3.358578052550232,
+      "grad_norm": 0.11592633920458482,
+      "learning_rate": 2.3519166051710664e-05,
+      "loss": 0.3666,
+      "step": 2173
+    },
+    {
+      "epoch": 3.3601236476043277,
+      "grad_norm": 0.1495106821718199,
+      "learning_rate": 2.3479841560688527e-05,
+      "loss": 0.3732,
+      "step": 2174
+    },
+    {
+      "epoch": 3.3616692426584236,
+      "grad_norm": 0.12148103208951651,
+      "learning_rate": 2.3440536310748825e-05,
+      "loss": 0.3678,
+      "step": 2175
+    },
+    {
+      "epoch": 3.3632148377125195,
+      "grad_norm": 0.1560380485682236,
+      "learning_rate": 2.3401250347670534e-05,
+      "loss": 0.3653,
+      "step": 2176
+    },
+    {
+      "epoch": 3.364760432766615,
+      "grad_norm": 0.12671564592640389,
+      "learning_rate": 2.3361983717210136e-05,
+      "loss": 0.3607,
+      "step": 2177
+    },
+    {
+      "epoch": 3.366306027820711,
+      "grad_norm": 0.15976617220407052,
+      "learning_rate": 2.3322736465101595e-05,
+      "loss": 0.3768,
+      "step": 2178
+    },
+    {
+      "epoch": 3.3678516228748068,
+      "grad_norm": 0.13554607763701795,
+      "learning_rate": 2.3283508637056305e-05,
+      "loss": 0.3661,
+      "step": 2179
+    },
+    {
+      "epoch": 3.3693972179289027,
+      "grad_norm": 0.16597557158850385,
+      "learning_rate": 2.324430027876308e-05,
+      "loss": 0.367,
+      "step": 2180
+    },
+    {
+      "epoch": 3.3709428129829986,
+      "grad_norm": 0.14013672295171226,
+      "learning_rate": 2.3205111435888012e-05,
+      "loss": 0.3676,
+      "step": 2181
+    },
+    {
+      "epoch": 3.3724884080370945,
+      "grad_norm": 0.16247932045000632,
+      "learning_rate": 2.3165942154074435e-05,
+      "loss": 0.3652,
+      "step": 2182
+    },
+    {
+      "epoch": 3.37403400309119,
+      "grad_norm": 0.127265358952448,
+      "learning_rate": 2.3126792478942984e-05,
+      "loss": 0.3609,
+      "step": 2183
+    },
+    {
+      "epoch": 3.375579598145286,
+      "grad_norm": 0.146921149877785,
+      "learning_rate": 2.3087662456091398e-05,
+      "loss": 0.3657,
+      "step": 2184
+    },
+    {
+      "epoch": 3.3771251931993818,
+      "grad_norm": 0.13121490181217182,
+      "learning_rate": 2.3048552131094527e-05,
+      "loss": 0.3603,
+      "step": 2185
+    },
+    {
+      "epoch": 3.3786707882534777,
+      "grad_norm": 0.17031366399895273,
+      "learning_rate": 2.3009461549504327e-05,
+      "loss": 0.3618,
+      "step": 2186
+    },
+    {
+      "epoch": 3.3802163833075736,
+      "grad_norm": 0.14391343264973547,
+      "learning_rate": 2.2970390756849707e-05,
+      "loss": 0.3677,
+      "step": 2187
+    },
+    {
+      "epoch": 3.381761978361669,
+      "grad_norm": 0.16743317678418054,
+      "learning_rate": 2.293133979863655e-05,
+      "loss": 0.3746,
+      "step": 2188
+    },
+    {
+      "epoch": 3.383307573415765,
+      "grad_norm": 0.4433956956052001,
+      "learning_rate": 2.2892308720347633e-05,
+      "loss": 0.3724,
+      "step": 2189
+    },
+    {
+      "epoch": 3.384853168469861,
+      "grad_norm": 0.17711364532400312,
+      "learning_rate": 2.2853297567442576e-05,
+      "loss": 0.3737,
+      "step": 2190
+    },
+    {
+      "epoch": 3.3863987635239567,
+      "grad_norm": 0.12860192982862037,
+      "learning_rate": 2.2814306385357784e-05,
+      "loss": 0.3679,
+      "step": 2191
+    },
+    {
+      "epoch": 3.3879443585780527,
+      "grad_norm": 0.18114097397886067,
+      "learning_rate": 2.277533521950645e-05,
+      "loss": 0.3689,
+      "step": 2192
+    },
+    {
+      "epoch": 3.3894899536321486,
+      "grad_norm": 0.143617498461099,
+      "learning_rate": 2.2736384115278397e-05,
+      "loss": 0.3653,
+      "step": 2193
+    },
+    {
+      "epoch": 3.391035548686244,
+      "grad_norm": 0.14934037051494214,
+      "learning_rate": 2.2697453118040108e-05,
+      "loss": 0.3705,
+      "step": 2194
+    },
+    {
+      "epoch": 3.39258114374034,
+      "grad_norm": 0.12063351961554758,
+      "learning_rate": 2.2658542273134635e-05,
+      "loss": 0.3649,
+      "step": 2195
+    },
+    {
+      "epoch": 3.394126738794436,
+      "grad_norm": 0.15418817321104553,
+      "learning_rate": 2.2619651625881565e-05,
+      "loss": 0.3695,
+      "step": 2196
+    },
+    {
+      "epoch": 3.3956723338485317,
+      "grad_norm": 0.1294006613911471,
+      "learning_rate": 2.258078122157696e-05,
+      "loss": 0.3669,
+      "step": 2197
+    },
+    {
+      "epoch": 3.3972179289026276,
+      "grad_norm": 0.14619418705114054,
+      "learning_rate": 2.2541931105493325e-05,
+      "loss": 0.372,
+      "step": 2198
+    },
+    {
+      "epoch": 3.398763523956723,
+      "grad_norm": 0.14290869803965706,
+      "learning_rate": 2.2503101322879504e-05,
+      "loss": 0.3659,
+      "step": 2199
+    },
+    {
+      "epoch": 3.400309119010819,
+      "grad_norm": 0.15248398259469637,
+      "learning_rate": 2.2464291918960654e-05,
+      "loss": 0.3582,
+      "step": 2200
+    },
+    {
+      "epoch": 3.401854714064915,
+      "grad_norm": 0.13962906567160044,
+      "learning_rate": 2.242550293893827e-05,
+      "loss": 0.3693,
+      "step": 2201
+    },
+    {
+      "epoch": 3.403400309119011,
+      "grad_norm": 0.15266168076238193,
+      "learning_rate": 2.2386734427989953e-05,
+      "loss": 0.362,
+      "step": 2202
+    },
+    {
+      "epoch": 3.4049459041731067,
+      "grad_norm": 0.13968077225056355,
+      "learning_rate": 2.234798643126951e-05,
+      "loss": 0.371,
+      "step": 2203
+    },
+    {
+      "epoch": 3.4064914992272026,
+      "grad_norm": 0.18309654478650297,
+      "learning_rate": 2.2309258993906893e-05,
+      "loss": 0.3575,
+      "step": 2204
+    },
+    {
+      "epoch": 3.4080370942812985,
+      "grad_norm": 0.15824859847422365,
+      "learning_rate": 2.227055216100806e-05,
+      "loss": 0.3744,
+      "step": 2205
+    },
+    {
+      "epoch": 3.409582689335394,
+      "grad_norm": 0.1799621641151411,
+      "learning_rate": 2.2231865977654986e-05,
+      "loss": 0.3659,
+      "step": 2206
+    },
+    {
+      "epoch": 3.41112828438949,
+      "grad_norm": 0.16819288754850809,
+      "learning_rate": 2.2193200488905577e-05,
+      "loss": 0.3668,
+      "step": 2207
+    },
+    {
+      "epoch": 3.412673879443586,
+      "grad_norm": 0.14129559366228042,
+      "learning_rate": 2.2154555739793704e-05,
+      "loss": 0.3697,
+      "step": 2208
+    },
+    {
+      "epoch": 3.4142194744976817,
+      "grad_norm": 0.16403415378386554,
+      "learning_rate": 2.2115931775328977e-05,
+      "loss": 0.363,
+      "step": 2209
+    },
+    {
+      "epoch": 3.4157650695517776,
+      "grad_norm": 0.13802924828710786,
+      "learning_rate": 2.207732864049686e-05,
+      "loss": 0.3746,
+      "step": 2210
+    },
+    {
+      "epoch": 3.417310664605873,
+      "grad_norm": 0.19682770170330607,
+      "learning_rate": 2.2038746380258588e-05,
+      "loss": 0.3564,
+      "step": 2211
+    },
+    {
+      "epoch": 3.418856259659969,
+      "grad_norm": 0.17313372028050691,
+      "learning_rate": 2.2000185039551023e-05,
+      "loss": 0.365,
+      "step": 2212
+    },
+    {
+      "epoch": 3.420401854714065,
+      "grad_norm": 0.15480950449114608,
+      "learning_rate": 2.1961644663286666e-05,
+      "loss": 0.3591,
+      "step": 2213
+    },
+    {
+      "epoch": 3.421947449768161,
+      "grad_norm": 0.14347961963031264,
+      "learning_rate": 2.1923125296353652e-05,
+      "loss": 0.3721,
+      "step": 2214
+    },
+    {
+      "epoch": 3.4234930448222567,
+      "grad_norm": 0.21703862095555077,
+      "learning_rate": 2.1884626983615602e-05,
+      "loss": 0.3807,
+      "step": 2215
+    },
+    {
+      "epoch": 3.4250386398763526,
+      "grad_norm": 0.24410686962854097,
+      "learning_rate": 2.184614976991162e-05,
+      "loss": 0.3628,
+      "step": 2216
+    },
+    {
+      "epoch": 3.426584234930448,
+      "grad_norm": 0.1363111022100698,
+      "learning_rate": 2.1807693700056246e-05,
+      "loss": 0.3581,
+      "step": 2217
+    },
+    {
+      "epoch": 3.428129829984544,
+      "grad_norm": 0.23085033749351025,
+      "learning_rate": 2.176925881883939e-05,
+      "loss": 0.3686,
+      "step": 2218
+    },
+    {
+      "epoch": 3.42967542503864,
+      "grad_norm": 0.12923862126584543,
+      "learning_rate": 2.173084517102627e-05,
+      "loss": 0.3668,
+      "step": 2219
+    },
+    {
+      "epoch": 3.4312210200927358,
+      "grad_norm": 0.1651531123248392,
+      "learning_rate": 2.1692452801357415e-05,
+      "loss": 0.3668,
+      "step": 2220
+    },
+    {
+      "epoch": 3.4327666151468317,
+      "grad_norm": 0.1424252177597652,
+      "learning_rate": 2.1654081754548528e-05,
+      "loss": 0.3668,
+      "step": 2221
+    },
+    {
+      "epoch": 3.434312210200927,
+      "grad_norm": 0.13295987830609388,
+      "learning_rate": 2.161573207529049e-05,
+      "loss": 0.3693,
+      "step": 2222
+    },
+    {
+      "epoch": 3.435857805255023,
+      "grad_norm": 0.13862922963229274,
+      "learning_rate": 2.1577403808249304e-05,
+      "loss": 0.3681,
+      "step": 2223
+    },
+    {
+      "epoch": 3.437403400309119,
+      "grad_norm": 0.13194165522460818,
+      "learning_rate": 2.1539096998066028e-05,
+      "loss": 0.3591,
+      "step": 2224
+    },
+    {
+      "epoch": 3.438948995363215,
+      "grad_norm": 0.13499690881078524,
+      "learning_rate": 2.1500811689356703e-05,
+      "loss": 0.3642,
+      "step": 2225
+    },
+    {
+      "epoch": 3.4404945904173108,
+      "grad_norm": 0.1339915648072994,
+      "learning_rate": 2.1462547926712393e-05,
+      "loss": 0.3554,
+      "step": 2226
+    },
+    {
+      "epoch": 3.4420401854714067,
+      "grad_norm": 0.13631810446849388,
+      "learning_rate": 2.1424305754699012e-05,
+      "loss": 0.3682,
+      "step": 2227
+    },
+    {
+      "epoch": 3.443585780525502,
+      "grad_norm": 0.12801192112518459,
+      "learning_rate": 2.138608521785732e-05,
+      "loss": 0.3745,
+      "step": 2228
+    },
+    {
+      "epoch": 3.445131375579598,
+      "grad_norm": 0.12447615819235312,
+      "learning_rate": 2.1347886360702958e-05,
+      "loss": 0.3668,
+      "step": 2229
+    },
+    {
+      "epoch": 3.446676970633694,
+      "grad_norm": 0.1281798142444195,
+      "learning_rate": 2.13097092277262e-05,
+      "loss": 0.3592,
+      "step": 2230
+    },
+    {
+      "epoch": 3.44822256568779,
+      "grad_norm": 0.18194444810246324,
+      "learning_rate": 2.1271553863392084e-05,
+      "loss": 0.378,
+      "step": 2231
+    },
+    {
+      "epoch": 3.4497681607418857,
+      "grad_norm": 0.16293128387063224,
+      "learning_rate": 2.1233420312140325e-05,
+      "loss": 0.3609,
+      "step": 2232
+    },
+    {
+      "epoch": 3.451313755795981,
+      "grad_norm": 0.19059816836888333,
+      "learning_rate": 2.1195308618385164e-05,
+      "loss": 0.3676,
+      "step": 2233
+    },
+    {
+      "epoch": 3.452859350850077,
+      "grad_norm": 0.23789578340396286,
+      "learning_rate": 2.1157218826515428e-05,
+      "loss": 0.3533,
+      "step": 2234
+    },
+    {
+      "epoch": 3.454404945904173,
+      "grad_norm": 0.1305828277942505,
+      "learning_rate": 2.1119150980894405e-05,
+      "loss": 0.3593,
+      "step": 2235
+    },
+    {
+      "epoch": 3.455950540958269,
+      "grad_norm": 0.15005871190242023,
+      "learning_rate": 2.1081105125859884e-05,
+      "loss": 0.3739,
+      "step": 2236
+    },
+    {
+      "epoch": 3.457496136012365,
+      "grad_norm": 0.12897613860424506,
+      "learning_rate": 2.1043081305723945e-05,
+      "loss": 0.37,
+      "step": 2237
+    },
+    {
+      "epoch": 3.4590417310664607,
+      "grad_norm": 0.13550501095396109,
+      "learning_rate": 2.1005079564773057e-05,
+      "loss": 0.3722,
+      "step": 2238
+    },
+    {
+      "epoch": 3.4605873261205566,
+      "grad_norm": 0.11760751538202946,
+      "learning_rate": 2.0967099947268013e-05,
+      "loss": 0.3663,
+      "step": 2239
+    },
+    {
+      "epoch": 3.462132921174652,
+      "grad_norm": 0.14338053026336461,
+      "learning_rate": 2.092914249744377e-05,
+      "loss": 0.3662,
+      "step": 2240
+    },
+    {
+      "epoch": 3.463678516228748,
+      "grad_norm": 0.1213786296882088,
+      "learning_rate": 2.089120725950948e-05,
+      "loss": 0.3667,
+      "step": 2241
+    },
+    {
+      "epoch": 3.465224111282844,
+      "grad_norm": 0.1389070497931083,
+      "learning_rate": 2.0853294277648475e-05,
+      "loss": 0.3622,
+      "step": 2242
+    },
+    {
+      "epoch": 3.46676970633694,
+      "grad_norm": 0.3067028646142398,
+      "learning_rate": 2.0815403596018103e-05,
+      "loss": 0.3719,
+      "step": 2243
+    },
+    {
+      "epoch": 3.4683153013910357,
+      "grad_norm": 0.13083023438257277,
+      "learning_rate": 2.077753525874978e-05,
+      "loss": 0.3748,
+      "step": 2244
+    },
+    {
+      "epoch": 3.469860896445131,
+      "grad_norm": 0.2117395750689371,
+      "learning_rate": 2.0739689309948865e-05,
+      "loss": 0.3644,
+      "step": 2245
+    },
+    {
+      "epoch": 3.471406491499227,
+      "grad_norm": 0.1514838520767274,
+      "learning_rate": 2.070186579369467e-05,
+      "loss": 0.363,
+      "step": 2246
+    },
+    {
+      "epoch": 3.472952086553323,
+      "grad_norm": 0.13751256580191215,
+      "learning_rate": 2.066406475404035e-05,
+      "loss": 0.3631,
+      "step": 2247
+    },
+    {
+      "epoch": 3.474497681607419,
+      "grad_norm": 0.14978411561899216,
+      "learning_rate": 2.0626286235012928e-05,
+      "loss": 0.3615,
+      "step": 2248
+    },
+    {
+      "epoch": 3.476043276661515,
+      "grad_norm": 0.15003822080242218,
+      "learning_rate": 2.0588530280613152e-05,
+      "loss": 0.3667,
+      "step": 2249
+    },
+    {
+      "epoch": 3.4775888717156107,
+      "grad_norm": 0.1624011793089801,
+      "learning_rate": 2.0550796934815507e-05,
+      "loss": 0.363,
+      "step": 2250
+    },
+    {
+      "epoch": 3.479134466769706,
+      "grad_norm": 0.11050380571513596,
+      "learning_rate": 2.051308624156815e-05,
+      "loss": 0.3668,
+      "step": 2251
+    },
+    {
+      "epoch": 3.480680061823802,
+      "grad_norm": 0.1561929256372977,
+      "learning_rate": 2.0475398244792835e-05,
+      "loss": 0.3669,
+      "step": 2252
+    },
+    {
+      "epoch": 3.482225656877898,
+      "grad_norm": 0.1187676222355437,
+      "learning_rate": 2.043773298838489e-05,
+      "loss": 0.3565,
+      "step": 2253
+    },
+    {
+      "epoch": 3.483771251931994,
+      "grad_norm": 0.17276171169900728,
+      "learning_rate": 2.0400090516213192e-05,
+      "loss": 0.3709,
+      "step": 2254
+    },
+    {
+      "epoch": 3.48531684698609,
+      "grad_norm": 0.14069491642052284,
+      "learning_rate": 2.0362470872120032e-05,
+      "loss": 0.3766,
+      "step": 2255
+    },
+    {
+      "epoch": 3.4868624420401853,
+      "grad_norm": 0.14822815796650152,
+      "learning_rate": 2.032487409992113e-05,
+      "loss": 0.37,
+      "step": 2256
+    },
+    {
+      "epoch": 3.488408037094281,
+      "grad_norm": 0.1413867069961767,
+      "learning_rate": 2.0287300243405578e-05,
+      "loss": 0.3644,
+      "step": 2257
+    },
+    {
+      "epoch": 3.489953632148377,
+      "grad_norm": 0.15082419920184817,
+      "learning_rate": 2.0249749346335764e-05,
+      "loss": 0.37,
+      "step": 2258
+    },
+    {
+      "epoch": 3.491499227202473,
+      "grad_norm": 0.14103776048876207,
+      "learning_rate": 2.0212221452447328e-05,
+      "loss": 0.3717,
+      "step": 2259
+    },
+    {
+      "epoch": 3.493044822256569,
+      "grad_norm": 0.13815955422925869,
+      "learning_rate": 2.0174716605449155e-05,
+      "loss": 0.3677,
+      "step": 2260
+    },
+    {
+      "epoch": 3.4945904173106648,
+      "grad_norm": 0.15131638558783417,
+      "learning_rate": 2.0137234849023248e-05,
+      "loss": 0.3662,
+      "step": 2261
+    },
+    {
+      "epoch": 3.4961360123647607,
+      "grad_norm": 0.12480028276492205,
+      "learning_rate": 2.009977622682474e-05,
+      "loss": 0.362,
+      "step": 2262
+    },
+    {
+      "epoch": 3.497681607418856,
+      "grad_norm": 0.15266972037670695,
+      "learning_rate": 2.0062340782481785e-05,
+      "loss": 0.3693,
+      "step": 2263
+    },
+    {
+      "epoch": 3.499227202472952,
+      "grad_norm": 0.12720931793188006,
+      "learning_rate": 2.002492855959562e-05,
+      "loss": 0.3665,
+      "step": 2264
+    },
+    {
+      "epoch": 3.500772797527048,
+      "grad_norm": 0.1272550131960794,
+      "learning_rate": 1.9987539601740332e-05,
+      "loss": 0.3645,
+      "step": 2265
+    },
+    {
+      "epoch": 3.502318392581144,
+      "grad_norm": 0.12189387184243682,
+      "learning_rate": 1.9950173952462966e-05,
+      "loss": 0.3663,
+      "step": 2266
+    },
+    {
+      "epoch": 3.5038639876352393,
+      "grad_norm": 0.13492177300182276,
+      "learning_rate": 1.9912831655283446e-05,
+      "loss": 0.3555,
+      "step": 2267
+    },
+    {
+      "epoch": 3.5054095826893352,
+      "grad_norm": 0.12115252260644373,
+      "learning_rate": 1.987551275369445e-05,
+      "loss": 0.3732,
+      "step": 2268
+    },
+    {
+      "epoch": 3.506955177743431,
+      "grad_norm": 0.16506182656469232,
+      "learning_rate": 1.983821729116141e-05,
+      "loss": 0.3622,
+      "step": 2269
+    },
+    {
+      "epoch": 3.508500772797527,
+      "grad_norm": 0.12253683731579625,
+      "learning_rate": 1.9800945311122505e-05,
+      "loss": 0.3672,
+      "step": 2270
+    },
+    {
+      "epoch": 3.510046367851623,
+      "grad_norm": 0.1407742820609905,
+      "learning_rate": 1.9763696856988516e-05,
+      "loss": 0.37,
+      "step": 2271
+    },
+    {
+      "epoch": 3.511591962905719,
+      "grad_norm": 0.13681226830576138,
+      "learning_rate": 1.972647197214284e-05,
+      "loss": 0.3738,
+      "step": 2272
+    },
+    {
+      "epoch": 3.5131375579598147,
+      "grad_norm": 0.12079535836295369,
+      "learning_rate": 1.9689270699941416e-05,
+      "loss": 0.3649,
+      "step": 2273
+    },
+    {
+      "epoch": 3.51468315301391,
+      "grad_norm": 0.13307052757071333,
+      "learning_rate": 1.965209308371269e-05,
+      "loss": 0.3666,
+      "step": 2274
+    },
+    {
+      "epoch": 3.516228748068006,
+      "grad_norm": 0.12190116044810183,
+      "learning_rate": 1.9614939166757546e-05,
+      "loss": 0.3661,
+      "step": 2275
+    },
+    {
+      "epoch": 3.517774343122102,
+      "grad_norm": 0.12424625343826458,
+      "learning_rate": 1.957780899234929e-05,
+      "loss": 0.3676,
+      "step": 2276
+    },
+    {
+      "epoch": 3.519319938176198,
+      "grad_norm": 0.13310891983612444,
+      "learning_rate": 1.954070260373355e-05,
+      "loss": 0.3682,
+      "step": 2277
+    },
+    {
+      "epoch": 3.5208655332302934,
+      "grad_norm": 0.12058985909010728,
+      "learning_rate": 1.9503620044128254e-05,
+      "loss": 0.3736,
+      "step": 2278
+    },
+    {
+      "epoch": 3.5224111282843893,
+      "grad_norm": 0.1328700160544732,
+      "learning_rate": 1.9466561356723586e-05,
+      "loss": 0.3666,
+      "step": 2279
+    },
+    {
+      "epoch": 3.523956723338485,
+      "grad_norm": 0.12698907820026423,
+      "learning_rate": 1.9429526584681916e-05,
+      "loss": 0.3713,
+      "step": 2280
+    },
+    {
+      "epoch": 3.525502318392581,
+      "grad_norm": 0.11932373478013461,
+      "learning_rate": 1.9392515771137747e-05,
+      "loss": 0.3698,
+      "step": 2281
+    },
+    {
+      "epoch": 3.527047913446677,
+      "grad_norm": 0.13157331967424146,
+      "learning_rate": 1.9355528959197726e-05,
+      "loss": 0.369,
+      "step": 2282
+    },
+    {
+      "epoch": 3.528593508500773,
+      "grad_norm": 0.13566800341641627,
+      "learning_rate": 1.9318566191940504e-05,
+      "loss": 0.3691,
+      "step": 2283
+    },
+    {
+      "epoch": 3.530139103554869,
+      "grad_norm": 0.13162448475838684,
+      "learning_rate": 1.9281627512416724e-05,
+      "loss": 0.3729,
+      "step": 2284
+    },
+    {
+      "epoch": 3.5316846986089647,
+      "grad_norm": 0.11949373719861449,
+      "learning_rate": 1.9244712963649e-05,
+      "loss": 0.3534,
+      "step": 2285
+    },
+    {
+      "epoch": 3.53323029366306,
+      "grad_norm": 0.11736053310407618,
+      "learning_rate": 1.9207822588631817e-05,
+      "loss": 0.3707,
+      "step": 2286
+    },
+    {
+      "epoch": 3.534775888717156,
+      "grad_norm": 0.12394050065357538,
+      "learning_rate": 1.9170956430331507e-05,
+      "loss": 0.3698,
+      "step": 2287
+    },
+    {
+      "epoch": 3.536321483771252,
+      "grad_norm": 0.10797427831830231,
+      "learning_rate": 1.9134114531686235e-05,
+      "loss": 0.3545,
+      "step": 2288
+    },
+    {
+      "epoch": 3.537867078825348,
+      "grad_norm": 0.12618049750436605,
+      "learning_rate": 1.909729693560586e-05,
+      "loss": 0.3642,
+      "step": 2289
+    },
+    {
+      "epoch": 3.5394126738794434,
+      "grad_norm": 0.12076226937002642,
+      "learning_rate": 1.906050368497196e-05,
+      "loss": 0.3717,
+      "step": 2290
+    },
+    {
+      "epoch": 3.5409582689335393,
+      "grad_norm": 0.12473790053926571,
+      "learning_rate": 1.9023734822637743e-05,
+      "loss": 0.3679,
+      "step": 2291
+    },
+    {
+      "epoch": 3.542503863987635,
+      "grad_norm": 0.11560859256389514,
+      "learning_rate": 1.8986990391428056e-05,
+      "loss": 0.3687,
+      "step": 2292
+    },
+    {
+      "epoch": 3.544049459041731,
+      "grad_norm": 0.12606767866219007,
+      "learning_rate": 1.8950270434139227e-05,
+      "loss": 0.3667,
+      "step": 2293
+    },
+    {
+      "epoch": 3.545595054095827,
+      "grad_norm": 0.12389947684873036,
+      "learning_rate": 1.89135749935391e-05,
+      "loss": 0.3673,
+      "step": 2294
+    },
+    {
+      "epoch": 3.547140649149923,
+      "grad_norm": 0.1425149378605872,
+      "learning_rate": 1.8876904112367014e-05,
+      "loss": 0.3776,
+      "step": 2295
+    },
+    {
+      "epoch": 3.548686244204019,
+      "grad_norm": 0.15493403968430905,
+      "learning_rate": 1.884025783333364e-05,
+      "loss": 0.367,
+      "step": 2296
+    },
+    {
+      "epoch": 3.5502318392581143,
+      "grad_norm": 0.11972366321998125,
+      "learning_rate": 1.8803636199121006e-05,
+      "loss": 0.3698,
+      "step": 2297
+    },
+    {
+      "epoch": 3.55177743431221,
+      "grad_norm": 0.16216599108368399,
+      "learning_rate": 1.8767039252382474e-05,
+      "loss": 0.3649,
+      "step": 2298
+    },
+    {
+      "epoch": 3.553323029366306,
+      "grad_norm": 0.13076290262596807,
+      "learning_rate": 1.873046703574264e-05,
+      "loss": 0.372,
+      "step": 2299
+    },
+    {
+      "epoch": 3.554868624420402,
+      "grad_norm": 0.1801491112133388,
+      "learning_rate": 1.869391959179723e-05,
+      "loss": 0.363,
+      "step": 2300
+    },
+    {
+      "epoch": 3.5564142194744974,
+      "grad_norm": 0.1122279839845766,
+      "learning_rate": 1.865739696311322e-05,
+      "loss": 0.3632,
+      "step": 2301
+    },
+    {
+      "epoch": 3.5579598145285933,
+      "grad_norm": 0.16441164572650369,
+      "learning_rate": 1.8620899192228622e-05,
+      "loss": 0.376,
+      "step": 2302
+    },
+    {
+      "epoch": 3.5595054095826892,
+      "grad_norm": 0.12016647909976266,
+      "learning_rate": 1.858442632165248e-05,
+      "loss": 0.363,
+      "step": 2303
+    },
+    {
+      "epoch": 3.561051004636785,
+      "grad_norm": 0.13389180792505737,
+      "learning_rate": 1.8547978393864912e-05,
+      "loss": 0.3679,
+      "step": 2304
+    },
+    {
+      "epoch": 3.562596599690881,
+      "grad_norm": 0.13337931765783254,
+      "learning_rate": 1.851155545131692e-05,
+      "loss": 0.3662,
+      "step": 2305
+    },
+    {
+      "epoch": 3.564142194744977,
+      "grad_norm": 0.12029691360682021,
+      "learning_rate": 1.8475157536430415e-05,
+      "loss": 0.3775,
+      "step": 2306
+    },
+    {
+      "epoch": 3.565687789799073,
+      "grad_norm": 0.11962427742341881,
+      "learning_rate": 1.8438784691598176e-05,
+      "loss": 0.3626,
+      "step": 2307
+    },
+    {
+      "epoch": 3.5672333848531683,
+      "grad_norm": 0.12076196646675953,
+      "learning_rate": 1.840243695918377e-05,
+      "loss": 0.3503,
+      "step": 2308
+    },
+    {
+      "epoch": 3.5687789799072642,
+      "grad_norm": 0.1139843872027316,
+      "learning_rate": 1.8366114381521504e-05,
+      "loss": 0.3536,
+      "step": 2309
+    },
+    {
+      "epoch": 3.57032457496136,
+      "grad_norm": 0.1373164349875881,
+      "learning_rate": 1.8329817000916437e-05,
+      "loss": 0.3663,
+      "step": 2310
+    },
+    {
+      "epoch": 3.571870170015456,
+      "grad_norm": 0.14416586034482695,
+      "learning_rate": 1.8293544859644225e-05,
+      "loss": 0.3702,
+      "step": 2311
+    },
+    {
+      "epoch": 3.573415765069552,
+      "grad_norm": 0.12982911810173123,
+      "learning_rate": 1.825729799995116e-05,
+      "loss": 0.3662,
+      "step": 2312
+    },
+    {
+      "epoch": 3.5749613601236474,
+      "grad_norm": 0.1391599086123799,
+      "learning_rate": 1.8221076464054066e-05,
+      "loss": 0.3743,
+      "step": 2313
+    },
+    {
+      "epoch": 3.5765069551777433,
+      "grad_norm": 0.13603704128187755,
+      "learning_rate": 1.8184880294140293e-05,
+      "loss": 0.3624,
+      "step": 2314
+    },
+    {
+      "epoch": 3.578052550231839,
+      "grad_norm": 0.123219953191658,
+      "learning_rate": 1.814870953236762e-05,
+      "loss": 0.3734,
+      "step": 2315
+    },
+    {
+      "epoch": 3.579598145285935,
+      "grad_norm": 0.1545600721091344,
+      "learning_rate": 1.811256422086428e-05,
+      "loss": 0.3752,
+      "step": 2316
+    },
+    {
+      "epoch": 3.581143740340031,
+      "grad_norm": 0.12844492582796185,
+      "learning_rate": 1.807644440172882e-05,
+      "loss": 0.3571,
+      "step": 2317
+    },
+    {
+      "epoch": 3.582689335394127,
+      "grad_norm": 0.13017833846477858,
+      "learning_rate": 1.804035011703011e-05,
+      "loss": 0.3691,
+      "step": 2318
+    },
+    {
+      "epoch": 3.584234930448223,
+      "grad_norm": 0.1444866065809688,
+      "learning_rate": 1.8004281408807287e-05,
+      "loss": 0.3695,
+      "step": 2319
+    },
+    {
+      "epoch": 3.5857805255023183,
+      "grad_norm": 0.12692361732698557,
+      "learning_rate": 1.796823831906969e-05,
+      "loss": 0.3669,
+      "step": 2320
+    },
+    {
+      "epoch": 3.587326120556414,
+      "grad_norm": 0.12298294915988707,
+      "learning_rate": 1.793222088979682e-05,
+      "loss": 0.3554,
+      "step": 2321
+    },
+    {
+      "epoch": 3.58887171561051,
+      "grad_norm": 0.14366650700441572,
+      "learning_rate": 1.7896229162938286e-05,
+      "loss": 0.3647,
+      "step": 2322
+    },
+    {
+      "epoch": 3.590417310664606,
+      "grad_norm": 0.11991611653840754,
+      "learning_rate": 1.7860263180413797e-05,
+      "loss": 0.3682,
+      "step": 2323
+    },
+    {
+      "epoch": 3.5919629057187015,
+      "grad_norm": 0.12616629641714278,
+      "learning_rate": 1.782432298411303e-05,
+      "loss": 0.353,
+      "step": 2324
+    },
+    {
+      "epoch": 3.5935085007727974,
+      "grad_norm": 0.11510384978951838,
+      "learning_rate": 1.778840861589564e-05,
+      "loss": 0.3682,
+      "step": 2325
+    },
+    {
+      "epoch": 3.5950540958268933,
+      "grad_norm": 0.1427137699243558,
+      "learning_rate": 1.7752520117591237e-05,
+      "loss": 0.3586,
+      "step": 2326
+    },
+    {
+      "epoch": 3.596599690880989,
+      "grad_norm": 0.12137664650483156,
+      "learning_rate": 1.771665753099927e-05,
+      "loss": 0.3699,
+      "step": 2327
+    },
+    {
+      "epoch": 3.598145285935085,
+      "grad_norm": 0.1593699503856505,
+      "learning_rate": 1.768082089788897e-05,
+      "loss": 0.3726,
+      "step": 2328
+    },
+    {
+      "epoch": 3.599690880989181,
+      "grad_norm": 0.1132503843017808,
+      "learning_rate": 1.7645010259999425e-05,
+      "loss": 0.3631,
+      "step": 2329
+    },
+    {
+      "epoch": 3.601236476043277,
+      "grad_norm": 0.12430097431032934,
+      "learning_rate": 1.7609225659039384e-05,
+      "loss": 0.3601,
+      "step": 2330
+    },
+    {
+      "epoch": 3.6027820710973724,
+      "grad_norm": 0.12808034790865092,
+      "learning_rate": 1.7573467136687276e-05,
+      "loss": 0.3719,
+      "step": 2331
+    },
+    {
+      "epoch": 3.6043276661514683,
+      "grad_norm": 0.11166963969343933,
+      "learning_rate": 1.7537734734591193e-05,
+      "loss": 0.369,
+      "step": 2332
+    },
+    {
+      "epoch": 3.605873261205564,
+      "grad_norm": 0.12119834161313824,
+      "learning_rate": 1.7502028494368777e-05,
+      "loss": 0.3612,
+      "step": 2333
+    },
+    {
+      "epoch": 3.60741885625966,
+      "grad_norm": 0.1338599900749187,
+      "learning_rate": 1.746634845760719e-05,
+      "loss": 0.3659,
+      "step": 2334
+    },
+    {
+      "epoch": 3.6089644513137555,
+      "grad_norm": 0.1202207743667235,
+      "learning_rate": 1.7430694665863096e-05,
+      "loss": 0.3683,
+      "step": 2335
+    },
+    {
+      "epoch": 3.6105100463678514,
+      "grad_norm": 0.1237073823219968,
+      "learning_rate": 1.7395067160662575e-05,
+      "loss": 0.3714,
+      "step": 2336
+    },
+    {
+      "epoch": 3.6120556414219473,
+      "grad_norm": 0.12007218125289,
+      "learning_rate": 1.7359465983501088e-05,
+      "loss": 0.371,
+      "step": 2337
+    },
+    {
+      "epoch": 3.6136012364760433,
+      "grad_norm": 0.12842443914781965,
+      "learning_rate": 1.732389117584347e-05,
+      "loss": 0.364,
+      "step": 2338
+    },
+    {
+      "epoch": 3.615146831530139,
+      "grad_norm": 0.10596904265581793,
+      "learning_rate": 1.728834277912381e-05,
+      "loss": 0.3602,
+      "step": 2339
+    },
+    {
+      "epoch": 3.616692426584235,
+      "grad_norm": 0.10869023415659597,
+      "learning_rate": 1.725282083474543e-05,
+      "loss": 0.3621,
+      "step": 2340
+    },
+    {
+      "epoch": 3.618238021638331,
+      "grad_norm": 0.1165075477749802,
+      "learning_rate": 1.7217325384080856e-05,
+      "loss": 0.3655,
+      "step": 2341
+    },
+    {
+      "epoch": 3.6197836166924264,
+      "grad_norm": 0.10882163379175609,
+      "learning_rate": 1.718185646847176e-05,
+      "loss": 0.3596,
+      "step": 2342
+    },
+    {
+      "epoch": 3.6213292117465223,
+      "grad_norm": 0.1178456252177068,
+      "learning_rate": 1.7146414129228907e-05,
+      "loss": 0.3657,
+      "step": 2343
+    },
+    {
+      "epoch": 3.6228748068006182,
+      "grad_norm": 0.10240211125646784,
+      "learning_rate": 1.7110998407632075e-05,
+      "loss": 0.361,
+      "step": 2344
+    },
+    {
+      "epoch": 3.624420401854714,
+      "grad_norm": 0.13934509863883984,
+      "learning_rate": 1.707560934493012e-05,
+      "loss": 0.362,
+      "step": 2345
+    },
+    {
+      "epoch": 3.62596599690881,
+      "grad_norm": 0.12235121349091291,
+      "learning_rate": 1.7040246982340768e-05,
+      "loss": 0.3724,
+      "step": 2346
+    },
+    {
+      "epoch": 3.6275115919629055,
+      "grad_norm": 0.1289959917030072,
+      "learning_rate": 1.7004911361050688e-05,
+      "loss": 0.3619,
+      "step": 2347
+    },
+    {
+      "epoch": 3.6290571870170014,
+      "grad_norm": 0.12656070349013324,
+      "learning_rate": 1.696960252221539e-05,
+      "loss": 0.3765,
+      "step": 2348
+    },
+    {
+      "epoch": 3.6306027820710973,
+      "grad_norm": 0.13164208414661552,
+      "learning_rate": 1.6934320506959186e-05,
+      "loss": 0.3569,
+      "step": 2349
+    },
+    {
+      "epoch": 3.6321483771251932,
+      "grad_norm": 0.14018696204791378,
+      "learning_rate": 1.6899065356375157e-05,
+      "loss": 0.3696,
+      "step": 2350
+    },
+    {
+      "epoch": 3.633693972179289,
+      "grad_norm": 0.13335338503516184,
+      "learning_rate": 1.6863837111525102e-05,
+      "loss": 0.3583,
+      "step": 2351
+    },
+    {
+      "epoch": 3.635239567233385,
+      "grad_norm": 0.14189327242128436,
+      "learning_rate": 1.6828635813439474e-05,
+      "loss": 0.376,
+      "step": 2352
+    },
+    {
+      "epoch": 3.636785162287481,
+      "grad_norm": 0.1424767988489207,
+      "learning_rate": 1.6793461503117317e-05,
+      "loss": 0.3672,
+      "step": 2353
+    },
+    {
+      "epoch": 3.6383307573415764,
+      "grad_norm": 0.12716271067818344,
+      "learning_rate": 1.6758314221526295e-05,
+      "loss": 0.3703,
+      "step": 2354
+    },
+    {
+      "epoch": 3.6398763523956723,
+      "grad_norm": 0.14231456486144842,
+      "learning_rate": 1.6723194009602565e-05,
+      "loss": 0.362,
+      "step": 2355
+    },
+    {
+      "epoch": 3.641421947449768,
+      "grad_norm": 0.1155046849682613,
+      "learning_rate": 1.6688100908250706e-05,
+      "loss": 0.3692,
+      "step": 2356
+    },
+    {
+      "epoch": 3.642967542503864,
+      "grad_norm": 0.11984742917495439,
+      "learning_rate": 1.665303495834381e-05,
+      "loss": 0.364,
+      "step": 2357
+    },
+    {
+      "epoch": 3.6445131375579596,
+      "grad_norm": 0.11095599152364047,
+      "learning_rate": 1.661799620072329e-05,
+      "loss": 0.3625,
+      "step": 2358
+    },
+    {
+      "epoch": 3.6460587326120555,
+      "grad_norm": 0.11742773718301099,
+      "learning_rate": 1.658298467619889e-05,
+      "loss": 0.3677,
+      "step": 2359
+    },
+    {
+      "epoch": 3.6476043276661514,
+      "grad_norm": 0.1370819526528459,
+      "learning_rate": 1.6548000425548673e-05,
+      "loss": 0.3657,
+      "step": 2360
+    },
+    {
+      "epoch": 3.6491499227202473,
+      "grad_norm": 0.1310253884603917,
+      "learning_rate": 1.651304348951889e-05,
+      "loss": 0.372,
+      "step": 2361
+    },
+    {
+      "epoch": 3.650695517774343,
+      "grad_norm": 0.1352718464038083,
+      "learning_rate": 1.6478113908824007e-05,
+      "loss": 0.3683,
+      "step": 2362
+    },
+    {
+      "epoch": 3.652241112828439,
+      "grad_norm": 0.12912938776115337,
+      "learning_rate": 1.644321172414662e-05,
+      "loss": 0.3635,
+      "step": 2363
+    },
+    {
+      "epoch": 3.653786707882535,
+      "grad_norm": 0.14573174937900993,
+      "learning_rate": 1.6408336976137413e-05,
+      "loss": 0.3612,
+      "step": 2364
+    },
+    {
+      "epoch": 3.6553323029366305,
+      "grad_norm": 0.13518683413753524,
+      "learning_rate": 1.6373489705415105e-05,
+      "loss": 0.3549,
+      "step": 2365
+    },
+    {
+      "epoch": 3.6568778979907264,
+      "grad_norm": 0.14780922237985136,
+      "learning_rate": 1.6338669952566458e-05,
+      "loss": 0.3699,
+      "step": 2366
+    },
+    {
+      "epoch": 3.6584234930448223,
+      "grad_norm": 0.13451760149116115,
+      "learning_rate": 1.6303877758146135e-05,
+      "loss": 0.357,
+      "step": 2367
+    },
+    {
+      "epoch": 3.659969088098918,
+      "grad_norm": 0.1538837938353939,
+      "learning_rate": 1.6269113162676718e-05,
+      "loss": 0.3697,
+      "step": 2368
+    },
+    {
+      "epoch": 3.6615146831530136,
+      "grad_norm": 0.14661770256363155,
+      "learning_rate": 1.623437620664865e-05,
+      "loss": 0.3579,
+      "step": 2369
+    },
+    {
+      "epoch": 3.6630602782071096,
+      "grad_norm": 0.12274379462840326,
+      "learning_rate": 1.6199666930520167e-05,
+      "loss": 0.3643,
+      "step": 2370
+    },
+    {
+      "epoch": 3.6646058732612055,
+      "grad_norm": 0.1410997977440352,
+      "learning_rate": 1.6164985374717287e-05,
+      "loss": 0.3568,
+      "step": 2371
+    },
+    {
+      "epoch": 3.6661514683153014,
+      "grad_norm": 0.10481209766859342,
+      "learning_rate": 1.6130331579633708e-05,
+      "loss": 0.3653,
+      "step": 2372
+    },
+    {
+      "epoch": 3.6676970633693973,
+      "grad_norm": 0.13728656717393564,
+      "learning_rate": 1.609570558563085e-05,
+      "loss": 0.3703,
+      "step": 2373
+    },
+    {
+      "epoch": 3.669242658423493,
+      "grad_norm": 0.34913408623266173,
+      "learning_rate": 1.6061107433037707e-05,
+      "loss": 0.3802,
+      "step": 2374
+    },
+    {
+      "epoch": 3.670788253477589,
+      "grad_norm": 0.17026625701091253,
+      "learning_rate": 1.6026537162150866e-05,
+      "loss": 0.3699,
+      "step": 2375
+    },
+    {
+      "epoch": 3.6723338485316845,
+      "grad_norm": 0.10603428596954531,
+      "learning_rate": 1.5991994813234427e-05,
+      "loss": 0.3586,
+      "step": 2376
+    },
+    {
+      "epoch": 3.6738794435857804,
+      "grad_norm": 0.14456360618655534,
+      "learning_rate": 1.5957480426519974e-05,
+      "loss": 0.3641,
+      "step": 2377
+    },
+    {
+      "epoch": 3.6754250386398764,
+      "grad_norm": 0.12460666932722352,
+      "learning_rate": 1.5922994042206524e-05,
+      "loss": 0.3663,
+      "step": 2378
+    },
+    {
+      "epoch": 3.6769706336939723,
+      "grad_norm": 0.13224902476417935,
+      "learning_rate": 1.58885357004605e-05,
+      "loss": 0.3641,
+      "step": 2379
+    },
+    {
+      "epoch": 3.678516228748068,
+      "grad_norm": 0.11835773755558501,
+      "learning_rate": 1.5854105441415626e-05,
+      "loss": 0.3725,
+      "step": 2380
+    },
+    {
+      "epoch": 3.6800618238021636,
+      "grad_norm": 0.1399784920859329,
+      "learning_rate": 1.581970330517294e-05,
+      "loss": 0.3726,
+      "step": 2381
+    },
+    {
+      "epoch": 3.6816074188562595,
+      "grad_norm": 0.10863356986620387,
+      "learning_rate": 1.5785329331800748e-05,
+      "loss": 0.3699,
+      "step": 2382
+    },
+    {
+      "epoch": 3.6831530139103554,
+      "grad_norm": 0.14914152304207584,
+      "learning_rate": 1.5750983561334493e-05,
+      "loss": 0.3649,
+      "step": 2383
+    },
+    {
+      "epoch": 3.6846986089644513,
+      "grad_norm": 0.10983839097615365,
+      "learning_rate": 1.571666603377681e-05,
+      "loss": 0.3616,
+      "step": 2384
+    },
+    {
+      "epoch": 3.6862442040185472,
+      "grad_norm": 0.11864062319572434,
+      "learning_rate": 1.5682376789097454e-05,
+      "loss": 0.3686,
+      "step": 2385
+    },
+    {
+      "epoch": 3.687789799072643,
+      "grad_norm": 0.12921683619582586,
+      "learning_rate": 1.5648115867233217e-05,
+      "loss": 0.369,
+      "step": 2386
+    },
+    {
+      "epoch": 3.689335394126739,
+      "grad_norm": 0.13351544550888958,
+      "learning_rate": 1.5613883308087885e-05,
+      "loss": 0.3577,
+      "step": 2387
+    },
+    {
+      "epoch": 3.6908809891808345,
+      "grad_norm": 0.11574422097680881,
+      "learning_rate": 1.557967915153226e-05,
+      "loss": 0.3693,
+      "step": 2388
+    },
+    {
+      "epoch": 3.6924265842349304,
+      "grad_norm": 0.11967192135535802,
+      "learning_rate": 1.554550343740403e-05,
+      "loss": 0.3713,
+      "step": 2389
+    },
+    {
+      "epoch": 3.6939721792890263,
+      "grad_norm": 0.11774543564588924,
+      "learning_rate": 1.5511356205507745e-05,
+      "loss": 0.3664,
+      "step": 2390
+    },
+    {
+      "epoch": 3.6955177743431222,
+      "grad_norm": 0.11217961944837937,
+      "learning_rate": 1.547723749561481e-05,
+      "loss": 0.3731,
+      "step": 2391
+    },
+    {
+      "epoch": 3.6970633693972177,
+      "grad_norm": 0.1128032148337808,
+      "learning_rate": 1.544314734746339e-05,
+      "loss": 0.3599,
+      "step": 2392
+    },
+    {
+      "epoch": 3.6986089644513136,
+      "grad_norm": 0.11791019137532446,
+      "learning_rate": 1.5409085800758382e-05,
+      "loss": 0.3727,
+      "step": 2393
+    },
+    {
+      "epoch": 3.7001545595054095,
+      "grad_norm": 0.11014852879861646,
+      "learning_rate": 1.537505289517141e-05,
+      "loss": 0.3631,
+      "step": 2394
+    },
+    {
+      "epoch": 3.7017001545595054,
+      "grad_norm": 0.1189598231949291,
+      "learning_rate": 1.534104867034069e-05,
+      "loss": 0.3662,
+      "step": 2395
+    },
+    {
+      "epoch": 3.7032457496136013,
+      "grad_norm": 0.1064272086248934,
+      "learning_rate": 1.5307073165871064e-05,
+      "loss": 0.3714,
+      "step": 2396
+    },
+    {
+      "epoch": 3.704791344667697,
+      "grad_norm": 0.1175254940810905,
+      "learning_rate": 1.5273126421333902e-05,
+      "loss": 0.3622,
+      "step": 2397
+    },
+    {
+      "epoch": 3.706336939721793,
+      "grad_norm": 0.11869856783550627,
+      "learning_rate": 1.5239208476267094e-05,
+      "loss": 0.3855,
+      "step": 2398
+    },
+    {
+      "epoch": 3.7078825347758886,
+      "grad_norm": 0.10585418877519753,
+      "learning_rate": 1.520531937017498e-05,
+      "loss": 0.3679,
+      "step": 2399
+    },
+    {
+      "epoch": 3.7094281298299845,
+      "grad_norm": 0.10414374016494432,
+      "learning_rate": 1.5171459142528298e-05,
+      "loss": 0.3646,
+      "step": 2400
+    },
+    {
+      "epoch": 3.7109737248840804,
+      "grad_norm": 0.1154751459568018,
+      "learning_rate": 1.5137627832764192e-05,
+      "loss": 0.3663,
+      "step": 2401
+    },
+    {
+      "epoch": 3.7125193199381763,
+      "grad_norm": 0.12972038175619321,
+      "learning_rate": 1.5103825480286092e-05,
+      "loss": 0.3634,
+      "step": 2402
+    },
+    {
+      "epoch": 3.7140649149922718,
+      "grad_norm": 0.10544766399278752,
+      "learning_rate": 1.50700521244637e-05,
+      "loss": 0.363,
+      "step": 2403
+    },
+    {
+      "epoch": 3.7156105100463677,
+      "grad_norm": 0.14105149020865543,
+      "learning_rate": 1.5036307804632958e-05,
+      "loss": 0.3599,
+      "step": 2404
+    },
+    {
+      "epoch": 3.7171561051004636,
+      "grad_norm": 0.10578826342035111,
+      "learning_rate": 1.5002592560095984e-05,
+      "loss": 0.3768,
+      "step": 2405
+    },
+    {
+      "epoch": 3.7187017001545595,
+      "grad_norm": 0.1246590706351202,
+      "learning_rate": 1.4968906430121015e-05,
+      "loss": 0.3727,
+      "step": 2406
+    },
+    {
+      "epoch": 3.7202472952086554,
+      "grad_norm": 0.10149076338372942,
+      "learning_rate": 1.4935249453942437e-05,
+      "loss": 0.3615,
+      "step": 2407
+    },
+    {
+      "epoch": 3.7217928902627513,
+      "grad_norm": 0.11756217507445449,
+      "learning_rate": 1.4901621670760613e-05,
+      "loss": 0.3607,
+      "step": 2408
+    },
+    {
+      "epoch": 3.723338485316847,
+      "grad_norm": 0.11375314071535139,
+      "learning_rate": 1.4868023119741927e-05,
+      "loss": 0.3734,
+      "step": 2409
+    },
+    {
+      "epoch": 3.7248840803709427,
+      "grad_norm": 0.1192254435232237,
+      "learning_rate": 1.4834453840018754e-05,
+      "loss": 0.3741,
+      "step": 2410
+    },
+    {
+      "epoch": 3.7264296754250386,
+      "grad_norm": 0.1350920020566422,
+      "learning_rate": 1.480091387068931e-05,
+      "loss": 0.3703,
+      "step": 2411
+    },
+    {
+      "epoch": 3.7279752704791345,
+      "grad_norm": 0.12556924033907177,
+      "learning_rate": 1.4767403250817704e-05,
+      "loss": 0.362,
+      "step": 2412
+    },
+    {
+      "epoch": 3.7295208655332304,
+      "grad_norm": 0.11766856952811641,
+      "learning_rate": 1.4733922019433893e-05,
+      "loss": 0.3663,
+      "step": 2413
+    },
+    {
+      "epoch": 3.7310664605873263,
+      "grad_norm": 0.1286354996512215,
+      "learning_rate": 1.4700470215533558e-05,
+      "loss": 0.3648,
+      "step": 2414
+    },
+    {
+      "epoch": 3.7326120556414217,
+      "grad_norm": 0.11158946005891823,
+      "learning_rate": 1.4667047878078115e-05,
+      "loss": 0.3733,
+      "step": 2415
+    },
+    {
+      "epoch": 3.7341576506955176,
+      "grad_norm": 0.11544437169649772,
+      "learning_rate": 1.46336550459947e-05,
+      "loss": 0.3605,
+      "step": 2416
+    },
+    {
+      "epoch": 3.7357032457496135,
+      "grad_norm": 0.11471496762611157,
+      "learning_rate": 1.4600291758176028e-05,
+      "loss": 0.3704,
+      "step": 2417
+    },
+    {
+      "epoch": 3.7372488408037094,
+      "grad_norm": 0.1238451434723756,
+      "learning_rate": 1.4566958053480442e-05,
+      "loss": 0.3697,
+      "step": 2418
+    },
+    {
+      "epoch": 3.7387944358578054,
+      "grad_norm": 0.13135744878532335,
+      "learning_rate": 1.4533653970731808e-05,
+      "loss": 0.3663,
+      "step": 2419
+    },
+    {
+      "epoch": 3.7403400309119013,
+      "grad_norm": 0.14567429305716154,
+      "learning_rate": 1.4500379548719505e-05,
+      "loss": 0.378,
+      "step": 2420
+    },
+    {
+      "epoch": 3.741885625965997,
+      "grad_norm": 0.11847751544960321,
+      "learning_rate": 1.4467134826198343e-05,
+      "loss": 0.3637,
+      "step": 2421
+    },
+    {
+      "epoch": 3.7434312210200926,
+      "grad_norm": 0.12265362965387298,
+      "learning_rate": 1.443391984188859e-05,
+      "loss": 0.3652,
+      "step": 2422
+    },
+    {
+      "epoch": 3.7449768160741885,
+      "grad_norm": 0.10576200048700815,
+      "learning_rate": 1.4400734634475835e-05,
+      "loss": 0.3615,
+      "step": 2423
+    },
+    {
+      "epoch": 3.7465224111282844,
+      "grad_norm": 0.09971084048285343,
+      "learning_rate": 1.4367579242610989e-05,
+      "loss": 0.3642,
+      "step": 2424
+    },
+    {
+      "epoch": 3.7480680061823803,
+      "grad_norm": 0.11389348686792579,
+      "learning_rate": 1.4334453704910254e-05,
+      "loss": 0.3694,
+      "step": 2425
+    },
+    {
+      "epoch": 3.749613601236476,
+      "grad_norm": 0.11100882415194324,
+      "learning_rate": 1.4301358059955045e-05,
+      "loss": 0.3722,
+      "step": 2426
+    },
+    {
+      "epoch": 3.7511591962905717,
+      "grad_norm": 0.10612153251344403,
+      "learning_rate": 1.4268292346291972e-05,
+      "loss": 0.3722,
+      "step": 2427
+    },
+    {
+      "epoch": 3.7527047913446676,
+      "grad_norm": 0.10280001895994094,
+      "learning_rate": 1.4235256602432763e-05,
+      "loss": 0.362,
+      "step": 2428
+    },
+    {
+      "epoch": 3.7542503863987635,
+      "grad_norm": 0.11268140539427714,
+      "learning_rate": 1.4202250866854285e-05,
+      "loss": 0.3675,
+      "step": 2429
+    },
+    {
+      "epoch": 3.7557959814528594,
+      "grad_norm": 0.10771453126577249,
+      "learning_rate": 1.4169275177998416e-05,
+      "loss": 0.3659,
+      "step": 2430
+    },
+    {
+      "epoch": 3.7573415765069553,
+      "grad_norm": 0.10835455375616225,
+      "learning_rate": 1.4136329574272041e-05,
+      "loss": 0.3684,
+      "step": 2431
+    },
+    {
+      "epoch": 3.7588871715610512,
+      "grad_norm": 0.10972159383995192,
+      "learning_rate": 1.4103414094047021e-05,
+      "loss": 0.368,
+      "step": 2432
+    },
+    {
+      "epoch": 3.7604327666151467,
+      "grad_norm": 0.10179697696083068,
+      "learning_rate": 1.4070528775660113e-05,
+      "loss": 0.3571,
+      "step": 2433
+    },
+    {
+      "epoch": 3.7619783616692426,
+      "grad_norm": 0.11923762129516918,
+      "learning_rate": 1.4037673657412949e-05,
+      "loss": 0.3663,
+      "step": 2434
+    },
+    {
+      "epoch": 3.7635239567233385,
+      "grad_norm": 0.10963970754030641,
+      "learning_rate": 1.4004848777572022e-05,
+      "loss": 0.367,
+      "step": 2435
+    },
+    {
+      "epoch": 3.7650695517774344,
+      "grad_norm": 0.1083508491663517,
+      "learning_rate": 1.3972054174368555e-05,
+      "loss": 0.3719,
+      "step": 2436
+    },
+    {
+      "epoch": 3.76661514683153,
+      "grad_norm": 0.11886956305124681,
+      "learning_rate": 1.3939289885998516e-05,
+      "loss": 0.3582,
+      "step": 2437
+    },
+    {
+      "epoch": 3.7681607418856258,
+      "grad_norm": 0.09179304143810182,
+      "learning_rate": 1.3906555950622624e-05,
+      "loss": 0.361,
+      "step": 2438
+    },
+    {
+      "epoch": 3.7697063369397217,
+      "grad_norm": 0.11000935440391167,
+      "learning_rate": 1.3873852406366162e-05,
+      "loss": 0.3558,
+      "step": 2439
+    },
+    {
+      "epoch": 3.7712519319938176,
+      "grad_norm": 0.09878863374993056,
+      "learning_rate": 1.3841179291319047e-05,
+      "loss": 0.3699,
+      "step": 2440
+    },
+    {
+      "epoch": 3.7727975270479135,
+      "grad_norm": 0.11157323343796777,
+      "learning_rate": 1.3808536643535794e-05,
+      "loss": 0.3653,
+      "step": 2441
+    },
+    {
+      "epoch": 3.7743431221020094,
+      "grad_norm": 0.10868499354404498,
+      "learning_rate": 1.377592450103539e-05,
+      "loss": 0.3681,
+      "step": 2442
+    },
+    {
+      "epoch": 3.7758887171561053,
+      "grad_norm": 0.10663509845694374,
+      "learning_rate": 1.3743342901801287e-05,
+      "loss": 0.3622,
+      "step": 2443
+    },
+    {
+      "epoch": 3.7774343122102008,
+      "grad_norm": 0.12573443101677276,
+      "learning_rate": 1.3710791883781407e-05,
+      "loss": 0.3705,
+      "step": 2444
+    },
+    {
+      "epoch": 3.7789799072642967,
+      "grad_norm": 0.11492129231588556,
+      "learning_rate": 1.3678271484888024e-05,
+      "loss": 0.365,
+      "step": 2445
+    },
+    {
+      "epoch": 3.7805255023183926,
+      "grad_norm": 0.14807717541559526,
+      "learning_rate": 1.3645781742997723e-05,
+      "loss": 0.3656,
+      "step": 2446
+    },
+    {
+      "epoch": 3.7820710973724885,
+      "grad_norm": 0.12550994463979237,
+      "learning_rate": 1.361332269595144e-05,
+      "loss": 0.3654,
+      "step": 2447
+    },
+    {
+      "epoch": 3.7836166924265844,
+      "grad_norm": 0.11062588072470007,
+      "learning_rate": 1.358089438155433e-05,
+      "loss": 0.3602,
+      "step": 2448
+    },
+    {
+      "epoch": 3.78516228748068,
+      "grad_norm": 0.12207981186966346,
+      "learning_rate": 1.3548496837575734e-05,
+      "loss": 0.37,
+      "step": 2449
+    },
+    {
+      "epoch": 3.7867078825347757,
+      "grad_norm": 0.10031037388559602,
+      "learning_rate": 1.3516130101749214e-05,
+      "loss": 0.3717,
+      "step": 2450
+    },
+    {
+      "epoch": 3.7882534775888717,
+      "grad_norm": 0.13408802490768343,
+      "learning_rate": 1.3483794211772389e-05,
+      "loss": 0.3596,
+      "step": 2451
+    },
+    {
+      "epoch": 3.7897990726429676,
+      "grad_norm": 0.1369605229020428,
+      "learning_rate": 1.3451489205306985e-05,
+      "loss": 0.3711,
+      "step": 2452
+    },
+    {
+      "epoch": 3.7913446676970635,
+      "grad_norm": 0.1416302196359165,
+      "learning_rate": 1.3419215119978742e-05,
+      "loss": 0.3625,
+      "step": 2453
+    },
+    {
+      "epoch": 3.7928902627511594,
+      "grad_norm": 0.10716684360570666,
+      "learning_rate": 1.338697199337739e-05,
+      "loss": 0.3757,
+      "step": 2454
+    },
+    {
+      "epoch": 3.7944358578052553,
+      "grad_norm": 0.15103251616508864,
+      "learning_rate": 1.3354759863056614e-05,
+      "loss": 0.3762,
+      "step": 2455
+    },
+    {
+      "epoch": 3.7959814528593507,
+      "grad_norm": 0.13185075362002102,
+      "learning_rate": 1.3322578766533968e-05,
+      "loss": 0.3619,
+      "step": 2456
+    },
+    {
+      "epoch": 3.7975270479134466,
+      "grad_norm": 0.12239289098180552,
+      "learning_rate": 1.3290428741290908e-05,
+      "loss": 0.3644,
+      "step": 2457
+    },
+    {
+      "epoch": 3.7990726429675425,
+      "grad_norm": 0.13086101877955567,
+      "learning_rate": 1.3258309824772656e-05,
+      "loss": 0.3685,
+      "step": 2458
+    },
+    {
+      "epoch": 3.8006182380216385,
+      "grad_norm": 0.12886510139841528,
+      "learning_rate": 1.3226222054388225e-05,
+      "loss": 0.3613,
+      "step": 2459
+    },
+    {
+      "epoch": 3.802163833075734,
+      "grad_norm": 0.12681154241348283,
+      "learning_rate": 1.3194165467510337e-05,
+      "loss": 0.3616,
+      "step": 2460
+    },
+    {
+      "epoch": 3.80370942812983,
+      "grad_norm": 0.1374953428671208,
+      "learning_rate": 1.3162140101475416e-05,
+      "loss": 0.363,
+      "step": 2461
+    },
+    {
+      "epoch": 3.8052550231839257,
+      "grad_norm": 0.13472113941668032,
+      "learning_rate": 1.3130145993583487e-05,
+      "loss": 0.3692,
+      "step": 2462
+    },
+    {
+      "epoch": 3.8068006182380216,
+      "grad_norm": 0.12865828223260975,
+      "learning_rate": 1.3098183181098225e-05,
+      "loss": 0.3712,
+      "step": 2463
+    },
+    {
+      "epoch": 3.8083462132921175,
+      "grad_norm": 0.1320572537937874,
+      "learning_rate": 1.3066251701246801e-05,
+      "loss": 0.3695,
+      "step": 2464
+    },
+    {
+      "epoch": 3.8098918083462134,
+      "grad_norm": 0.12834562006677214,
+      "learning_rate": 1.303435159121992e-05,
+      "loss": 0.3709,
+      "step": 2465
+    },
+    {
+      "epoch": 3.8114374034003093,
+      "grad_norm": 0.12331051533044547,
+      "learning_rate": 1.3002482888171745e-05,
+      "loss": 0.3731,
+      "step": 2466
+    },
+    {
+      "epoch": 3.812982998454405,
+      "grad_norm": 0.12987931317976803,
+      "learning_rate": 1.2970645629219854e-05,
+      "loss": 0.3782,
+      "step": 2467
+    },
+    {
+      "epoch": 3.8145285935085007,
+      "grad_norm": 0.11810538966464726,
+      "learning_rate": 1.2938839851445195e-05,
+      "loss": 0.369,
+      "step": 2468
+    },
+    {
+      "epoch": 3.8160741885625966,
+      "grad_norm": 0.12288577780247337,
+      "learning_rate": 1.2907065591892085e-05,
+      "loss": 0.3606,
+      "step": 2469
+    },
+    {
+      "epoch": 3.8176197836166925,
+      "grad_norm": 0.11085041347305545,
+      "learning_rate": 1.2875322887568102e-05,
+      "loss": 0.365,
+      "step": 2470
+    },
+    {
+      "epoch": 3.819165378670788,
+      "grad_norm": 0.11836388424213261,
+      "learning_rate": 1.2843611775444056e-05,
+      "loss": 0.3711,
+      "step": 2471
+    },
+    {
+      "epoch": 3.820710973724884,
+      "grad_norm": 0.11440744882214117,
+      "learning_rate": 1.2811932292454015e-05,
+      "loss": 0.3673,
+      "step": 2472
+    },
+    {
+      "epoch": 3.82225656877898,
+      "grad_norm": 0.10902753212932297,
+      "learning_rate": 1.2780284475495175e-05,
+      "loss": 0.3581,
+      "step": 2473
+    },
+    {
+      "epoch": 3.8238021638330757,
+      "grad_norm": 0.10797678553836423,
+      "learning_rate": 1.2748668361427808e-05,
+      "loss": 0.3622,
+      "step": 2474
+    },
+    {
+      "epoch": 3.8253477588871716,
+      "grad_norm": 0.11778675829973131,
+      "learning_rate": 1.271708398707535e-05,
+      "loss": 0.3675,
+      "step": 2475
+    },
+    {
+      "epoch": 3.8268933539412675,
+      "grad_norm": 0.10984096594392173,
+      "learning_rate": 1.2685531389224206e-05,
+      "loss": 0.3613,
+      "step": 2476
+    },
+    {
+      "epoch": 3.8284389489953634,
+      "grad_norm": 0.11855504166646279,
+      "learning_rate": 1.2654010604623782e-05,
+      "loss": 0.3656,
+      "step": 2477
+    },
+    {
+      "epoch": 3.8299845440494593,
+      "grad_norm": 0.10105851607114989,
+      "learning_rate": 1.2622521669986463e-05,
+      "loss": 0.3587,
+      "step": 2478
+    },
+    {
+      "epoch": 3.8315301391035548,
+      "grad_norm": 0.11929478675749929,
+      "learning_rate": 1.2591064621987501e-05,
+      "loss": 0.3639,
+      "step": 2479
+    },
+    {
+      "epoch": 3.8330757341576507,
+      "grad_norm": 0.10465017419082835,
+      "learning_rate": 1.2559639497265024e-05,
+      "loss": 0.3708,
+      "step": 2480
+    },
+    {
+      "epoch": 3.8346213292117466,
+      "grad_norm": 0.0967672404214158,
+      "learning_rate": 1.2528246332419971e-05,
+      "loss": 0.3564,
+      "step": 2481
+    },
+    {
+      "epoch": 3.8361669242658425,
+      "grad_norm": 0.10544790052485256,
+      "learning_rate": 1.2496885164016073e-05,
+      "loss": 0.3764,
+      "step": 2482
+    },
+    {
+      "epoch": 3.837712519319938,
+      "grad_norm": 0.09617536404393245,
+      "learning_rate": 1.2465556028579777e-05,
+      "loss": 0.3646,
+      "step": 2483
+    },
+    {
+      "epoch": 3.839258114374034,
+      "grad_norm": 0.10297967559422405,
+      "learning_rate": 1.243425896260022e-05,
+      "loss": 0.3726,
+      "step": 2484
+    },
+    {
+      "epoch": 3.8408037094281298,
+      "grad_norm": 0.10244677724678106,
+      "learning_rate": 1.2402994002529223e-05,
+      "loss": 0.3705,
+      "step": 2485
+    },
+    {
+      "epoch": 3.8423493044822257,
+      "grad_norm": 0.10131219568374498,
+      "learning_rate": 1.2371761184781175e-05,
+      "loss": 0.3639,
+      "step": 2486
+    },
+    {
+      "epoch": 3.8438948995363216,
+      "grad_norm": 0.09707008193138299,
+      "learning_rate": 1.234056054573304e-05,
+      "loss": 0.3718,
+      "step": 2487
+    },
+    {
+      "epoch": 3.8454404945904175,
+      "grad_norm": 0.09992089678593088,
+      "learning_rate": 1.2309392121724302e-05,
+      "loss": 0.3625,
+      "step": 2488
+    },
+    {
+      "epoch": 3.8469860896445134,
+      "grad_norm": 0.09690112449159632,
+      "learning_rate": 1.2278255949056925e-05,
+      "loss": 0.3583,
+      "step": 2489
+    },
+    {
+      "epoch": 3.848531684698609,
+      "grad_norm": 0.10740608520702238,
+      "learning_rate": 1.2247152063995302e-05,
+      "loss": 0.3732,
+      "step": 2490
+    },
+    {
+      "epoch": 3.8500772797527048,
+      "grad_norm": 0.10377366745909304,
+      "learning_rate": 1.2216080502766254e-05,
+      "loss": 0.3644,
+      "step": 2491
+    },
+    {
+      "epoch": 3.8516228748068007,
+      "grad_norm": 0.10178958110475811,
+      "learning_rate": 1.2185041301558918e-05,
+      "loss": 0.3592,
+      "step": 2492
+    },
+    {
+      "epoch": 3.8531684698608966,
+      "grad_norm": 0.10540236073406219,
+      "learning_rate": 1.2154034496524748e-05,
+      "loss": 0.3627,
+      "step": 2493
+    },
+    {
+      "epoch": 3.854714064914992,
+      "grad_norm": 0.11341760050262721,
+      "learning_rate": 1.2123060123777477e-05,
+      "loss": 0.3642,
+      "step": 2494
+    },
+    {
+      "epoch": 3.856259659969088,
+      "grad_norm": 0.11106701813295258,
+      "learning_rate": 1.2092118219393058e-05,
+      "loss": 0.3639,
+      "step": 2495
+    },
+    {
+      "epoch": 3.857805255023184,
+      "grad_norm": 0.10845039770754565,
+      "learning_rate": 1.2061208819409611e-05,
+      "loss": 0.3657,
+      "step": 2496
+    },
+    {
+      "epoch": 3.8593508500772797,
+      "grad_norm": 0.11101560958602312,
+      "learning_rate": 1.203033195982744e-05,
+      "loss": 0.3691,
+      "step": 2497
+    },
+    {
+      "epoch": 3.8608964451313756,
+      "grad_norm": 0.10730901242536212,
+      "learning_rate": 1.1999487676608914e-05,
+      "loss": 0.3602,
+      "step": 2498
+    },
+    {
+      "epoch": 3.8624420401854715,
+      "grad_norm": 0.09987418290859781,
+      "learning_rate": 1.1968676005678459e-05,
+      "loss": 0.365,
+      "step": 2499
+    },
+    {
+      "epoch": 3.8639876352395675,
+      "grad_norm": 0.09839768397643736,
+      "learning_rate": 1.1937896982922555e-05,
+      "loss": 0.3511,
+      "step": 2500
+    },
+    {
+      "epoch": 3.865533230293663,
+      "grad_norm": 0.1141607541008281,
+      "learning_rate": 1.1907150644189627e-05,
+      "loss": 0.3595,
+      "step": 2501
+    },
+    {
+      "epoch": 3.867078825347759,
+      "grad_norm": 0.10095235543244717,
+      "learning_rate": 1.1876437025289995e-05,
+      "loss": 0.3674,
+      "step": 2502
+    },
+    {
+      "epoch": 3.8686244204018547,
+      "grad_norm": 0.11541566839456155,
+      "learning_rate": 1.1845756161995956e-05,
+      "loss": 0.3628,
+      "step": 2503
+    },
+    {
+      "epoch": 3.8701700154559506,
+      "grad_norm": 0.1838395361908257,
+      "learning_rate": 1.1815108090041596e-05,
+      "loss": 0.3647,
+      "step": 2504
+    },
+    {
+      "epoch": 3.871715610510046,
+      "grad_norm": 0.12030001829742068,
+      "learning_rate": 1.178449284512281e-05,
+      "loss": 0.3691,
+      "step": 2505
+    },
+    {
+      "epoch": 3.873261205564142,
+      "grad_norm": 0.11368358105501487,
+      "learning_rate": 1.1753910462897302e-05,
+      "loss": 0.3609,
+      "step": 2506
+    },
+    {
+      "epoch": 3.874806800618238,
+      "grad_norm": 0.10934822052637712,
+      "learning_rate": 1.1723360978984456e-05,
+      "loss": 0.3686,
+      "step": 2507
+    },
+    {
+      "epoch": 3.876352395672334,
+      "grad_norm": 0.11761633146968824,
+      "learning_rate": 1.169284442896537e-05,
+      "loss": 0.37,
+      "step": 2508
+    },
+    {
+      "epoch": 3.8778979907264297,
+      "grad_norm": 0.105574566148553,
+      "learning_rate": 1.166236084838274e-05,
+      "loss": 0.3648,
+      "step": 2509
+    },
+    {
+      "epoch": 3.8794435857805256,
+      "grad_norm": 0.1078432391498194,
+      "learning_rate": 1.1631910272740915e-05,
+      "loss": 0.3592,
+      "step": 2510
+    },
+    {
+      "epoch": 3.8809891808346215,
+      "grad_norm": 0.11406201758434457,
+      "learning_rate": 1.1601492737505775e-05,
+      "loss": 0.3789,
+      "step": 2511
+    },
+    {
+      "epoch": 3.8825347758887174,
+      "grad_norm": 0.11427280856537803,
+      "learning_rate": 1.1571108278104708e-05,
+      "loss": 0.3713,
+      "step": 2512
+    },
+    {
+      "epoch": 3.884080370942813,
+      "grad_norm": 0.10674616626343066,
+      "learning_rate": 1.1540756929926618e-05,
+      "loss": 0.3637,
+      "step": 2513
+    },
+    {
+      "epoch": 3.885625965996909,
+      "grad_norm": 0.10230105720022321,
+      "learning_rate": 1.1510438728321813e-05,
+      "loss": 0.364,
+      "step": 2514
+    },
+    {
+      "epoch": 3.8871715610510047,
+      "grad_norm": 0.10518339393882155,
+      "learning_rate": 1.148015370860199e-05,
+      "loss": 0.3669,
+      "step": 2515
+    },
+    {
+      "epoch": 3.8887171561051006,
+      "grad_norm": 0.10001859936936507,
+      "learning_rate": 1.1449901906040223e-05,
+      "loss": 0.3634,
+      "step": 2516
+    },
+    {
+      "epoch": 3.890262751159196,
+      "grad_norm": 0.11298110745277655,
+      "learning_rate": 1.1419683355870874e-05,
+      "loss": 0.362,
+      "step": 2517
+    },
+    {
+      "epoch": 3.891808346213292,
+      "grad_norm": 0.1037426756168707,
+      "learning_rate": 1.1389498093289579e-05,
+      "loss": 0.3644,
+      "step": 2518
+    },
+    {
+      "epoch": 3.893353941267388,
+      "grad_norm": 0.10133774840519594,
+      "learning_rate": 1.135934615345324e-05,
+      "loss": 0.3732,
+      "step": 2519
+    },
+    {
+      "epoch": 3.894899536321484,
+      "grad_norm": 0.09748273087606743,
+      "learning_rate": 1.1329227571479904e-05,
+      "loss": 0.3675,
+      "step": 2520
+    },
+    {
+      "epoch": 3.8964451313755797,
+      "grad_norm": 0.10083462430348453,
+      "learning_rate": 1.129914238244878e-05,
+      "loss": 0.3517,
+      "step": 2521
+    },
+    {
+      "epoch": 3.8979907264296756,
+      "grad_norm": 0.09849897729988676,
+      "learning_rate": 1.126909062140019e-05,
+      "loss": 0.3559,
+      "step": 2522
+    },
+    {
+      "epoch": 3.8995363214837715,
+      "grad_norm": 0.10555870666575547,
+      "learning_rate": 1.123907232333551e-05,
+      "loss": 0.3594,
+      "step": 2523
+    },
+    {
+      "epoch": 3.901081916537867,
+      "grad_norm": 0.11146760888284764,
+      "learning_rate": 1.1209087523217153e-05,
+      "loss": 0.3653,
+      "step": 2524
+    },
+    {
+      "epoch": 3.902627511591963,
+      "grad_norm": 0.11997983295919977,
+      "learning_rate": 1.1179136255968523e-05,
+      "loss": 0.3664,
+      "step": 2525
+    },
+    {
+      "epoch": 3.9041731066460588,
+      "grad_norm": 0.09861075251718329,
+      "learning_rate": 1.114921855647395e-05,
+      "loss": 0.3632,
+      "step": 2526
+    },
+    {
+      "epoch": 3.9057187017001547,
+      "grad_norm": 0.11435240577883474,
+      "learning_rate": 1.1119334459578668e-05,
+      "loss": 0.3712,
+      "step": 2527
+    },
+    {
+      "epoch": 3.90726429675425,
+      "grad_norm": 0.12655769296532293,
+      "learning_rate": 1.1089484000088805e-05,
+      "loss": 0.3663,
+      "step": 2528
+    },
+    {
+      "epoch": 3.908809891808346,
+      "grad_norm": 0.09803040820276109,
+      "learning_rate": 1.1059667212771288e-05,
+      "loss": 0.3654,
+      "step": 2529
+    },
+    {
+      "epoch": 3.910355486862442,
+      "grad_norm": 0.10975258191024564,
+      "learning_rate": 1.1029884132353793e-05,
+      "loss": 0.3656,
+      "step": 2530
+    },
+    {
+      "epoch": 3.911901081916538,
+      "grad_norm": 0.11451097258362002,
+      "learning_rate": 1.1000134793524802e-05,
+      "loss": 0.3545,
+      "step": 2531
+    },
+    {
+      "epoch": 3.9134466769706338,
+      "grad_norm": 0.10739152900892397,
+      "learning_rate": 1.0970419230933458e-05,
+      "loss": 0.3596,
+      "step": 2532
+    },
+    {
+      "epoch": 3.9149922720247297,
+      "grad_norm": 0.1046841267992002,
+      "learning_rate": 1.0940737479189555e-05,
+      "loss": 0.364,
+      "step": 2533
+    },
+    {
+      "epoch": 3.9165378670788256,
+      "grad_norm": 0.09670283606500873,
+      "learning_rate": 1.0911089572863562e-05,
+      "loss": 0.3617,
+      "step": 2534
+    },
+    {
+      "epoch": 3.918083462132921,
+      "grad_norm": 0.11150108076988523,
+      "learning_rate": 1.0881475546486478e-05,
+      "loss": 0.366,
+      "step": 2535
+    },
+    {
+      "epoch": 3.919629057187017,
+      "grad_norm": 0.10239205892448629,
+      "learning_rate": 1.0851895434549857e-05,
+      "loss": 0.3798,
+      "step": 2536
+    },
+    {
+      "epoch": 3.921174652241113,
+      "grad_norm": 0.10268938062417501,
+      "learning_rate": 1.082234927150573e-05,
+      "loss": 0.3629,
+      "step": 2537
+    },
+    {
+      "epoch": 3.9227202472952087,
+      "grad_norm": 0.10235283480388616,
+      "learning_rate": 1.0792837091766635e-05,
+      "loss": 0.3709,
+      "step": 2538
+    },
+    {
+      "epoch": 3.9242658423493046,
+      "grad_norm": 0.09634602615965598,
+      "learning_rate": 1.0763358929705499e-05,
+      "loss": 0.3712,
+      "step": 2539
+    },
+    {
+      "epoch": 3.9258114374034,
+      "grad_norm": 0.09301068968552884,
+      "learning_rate": 1.0733914819655618e-05,
+      "loss": 0.3667,
+      "step": 2540
+    },
+    {
+      "epoch": 3.927357032457496,
+      "grad_norm": 0.09545970724699197,
+      "learning_rate": 1.0704504795910667e-05,
+      "loss": 0.3632,
+      "step": 2541
+    },
+    {
+      "epoch": 3.928902627511592,
+      "grad_norm": 0.101607272338685,
+      "learning_rate": 1.0675128892724587e-05,
+      "loss": 0.371,
+      "step": 2542
+    },
+    {
+      "epoch": 3.930448222565688,
+      "grad_norm": 0.09517284569366352,
+      "learning_rate": 1.0645787144311583e-05,
+      "loss": 0.3602,
+      "step": 2543
+    },
+    {
+      "epoch": 3.9319938176197837,
+      "grad_norm": 0.09360351163832138,
+      "learning_rate": 1.0616479584846094e-05,
+      "loss": 0.3579,
+      "step": 2544
+    },
+    {
+      "epoch": 3.9335394126738796,
+      "grad_norm": 0.10538953893885825,
+      "learning_rate": 1.0587206248462718e-05,
+      "loss": 0.3641,
+      "step": 2545
+    },
+    {
+      "epoch": 3.9350850077279755,
+      "grad_norm": 0.10187825493669692,
+      "learning_rate": 1.0557967169256206e-05,
+      "loss": 0.3603,
+      "step": 2546
+    },
+    {
+      "epoch": 3.936630602782071,
+      "grad_norm": 0.10448244218839056,
+      "learning_rate": 1.0528762381281425e-05,
+      "loss": 0.3655,
+      "step": 2547
+    },
+    {
+      "epoch": 3.938176197836167,
+      "grad_norm": 0.09479663821005888,
+      "learning_rate": 1.0499591918553281e-05,
+      "loss": 0.3601,
+      "step": 2548
+    },
+    {
+      "epoch": 3.939721792890263,
+      "grad_norm": 0.09783003802874986,
+      "learning_rate": 1.04704558150467e-05,
+      "loss": 0.3625,
+      "step": 2549
+    },
+    {
+      "epoch": 3.9412673879443587,
+      "grad_norm": 0.10151351630021069,
+      "learning_rate": 1.04413541046966e-05,
+      "loss": 0.3661,
+      "step": 2550
+    },
+    {
+      "epoch": 3.942812982998454,
+      "grad_norm": 0.09724859515650129,
+      "learning_rate": 1.0412286821397838e-05,
+      "loss": 0.3702,
+      "step": 2551
+    },
+    {
+      "epoch": 3.94435857805255,
+      "grad_norm": 0.10697273241386306,
+      "learning_rate": 1.0383253999005163e-05,
+      "loss": 0.3755,
+      "step": 2552
+    },
+    {
+      "epoch": 3.945904173106646,
+      "grad_norm": 0.0999074538581681,
+      "learning_rate": 1.0354255671333222e-05,
+      "loss": 0.3682,
+      "step": 2553
+    },
+    {
+      "epoch": 3.947449768160742,
+      "grad_norm": 0.10390728660398883,
+      "learning_rate": 1.0325291872156447e-05,
+      "loss": 0.358,
+      "step": 2554
+    },
+    {
+      "epoch": 3.948995363214838,
+      "grad_norm": 0.1015880940924953,
+      "learning_rate": 1.029636263520906e-05,
+      "loss": 0.3661,
+      "step": 2555
+    },
+    {
+      "epoch": 3.9505409582689337,
+      "grad_norm": 0.09913974603917472,
+      "learning_rate": 1.0267467994185077e-05,
+      "loss": 0.3671,
+      "step": 2556
+    },
+    {
+      "epoch": 3.9520865533230296,
+      "grad_norm": 0.09285619168664734,
+      "learning_rate": 1.0238607982738133e-05,
+      "loss": 0.36,
+      "step": 2557
+    },
+    {
+      "epoch": 3.953632148377125,
+      "grad_norm": 0.09932654996319276,
+      "learning_rate": 1.0209782634481589e-05,
+      "loss": 0.3652,
+      "step": 2558
+    },
+    {
+      "epoch": 3.955177743431221,
+      "grad_norm": 0.10159746896733102,
+      "learning_rate": 1.0180991982988444e-05,
+      "loss": 0.3669,
+      "step": 2559
+    },
+    {
+      "epoch": 3.956723338485317,
+      "grad_norm": 0.10027230146989236,
+      "learning_rate": 1.015223606179125e-05,
+      "loss": 0.3758,
+      "step": 2560
+    },
+    {
+      "epoch": 3.958268933539413,
+      "grad_norm": 0.09630876921012405,
+      "learning_rate": 1.0123514904382108e-05,
+      "loss": 0.3578,
+      "step": 2561
+    },
+    {
+      "epoch": 3.9598145285935082,
+      "grad_norm": 0.10232753308969458,
+      "learning_rate": 1.0094828544212668e-05,
+      "loss": 0.3707,
+      "step": 2562
+    },
+    {
+      "epoch": 3.961360123647604,
+      "grad_norm": 0.10292052209086937,
+      "learning_rate": 1.0066177014694025e-05,
+      "loss": 0.3636,
+      "step": 2563
+    },
+    {
+      "epoch": 3.9629057187017,
+      "grad_norm": 0.10487621364529856,
+      "learning_rate": 1.003756034919671e-05,
+      "loss": 0.3728,
+      "step": 2564
+    },
+    {
+      "epoch": 3.964451313755796,
+      "grad_norm": 0.09723305903976502,
+      "learning_rate": 1.000897858105061e-05,
+      "loss": 0.3655,
+      "step": 2565
+    },
+    {
+      "epoch": 3.965996908809892,
+      "grad_norm": 0.11362968585616648,
+      "learning_rate": 9.980431743545047e-06,
+      "loss": 0.3637,
+      "step": 2566
+    },
+    {
+      "epoch": 3.9675425038639878,
+      "grad_norm": 0.09909344509120414,
+      "learning_rate": 9.951919869928601e-06,
+      "loss": 0.3672,
+      "step": 2567
+    },
+    {
+      "epoch": 3.9690880989180837,
+      "grad_norm": 0.11315416634975581,
+      "learning_rate": 9.923442993409131e-06,
+      "loss": 0.3691,
+      "step": 2568
+    },
+    {
+      "epoch": 3.970633693972179,
+      "grad_norm": 0.10533606955922994,
+      "learning_rate": 9.895001147153774e-06,
+      "loss": 0.3572,
+      "step": 2569
+    },
+    {
+      "epoch": 3.972179289026275,
+      "grad_norm": 0.12170280346016124,
+      "learning_rate": 9.866594364288825e-06,
+      "loss": 0.3546,
+      "step": 2570
+    },
+    {
+      "epoch": 3.973724884080371,
+      "grad_norm": 0.10999283026501391,
+      "learning_rate": 9.83822267789976e-06,
+      "loss": 0.3682,
+      "step": 2571
+    },
+    {
+      "epoch": 3.975270479134467,
+      "grad_norm": 0.10466572095659792,
+      "learning_rate": 9.809886121031177e-06,
+      "loss": 0.3671,
+      "step": 2572
+    },
+    {
+      "epoch": 3.9768160741885628,
+      "grad_norm": 0.11362183054589346,
+      "learning_rate": 9.781584726686751e-06,
+      "loss": 0.363,
+      "step": 2573
+    },
+    {
+      "epoch": 3.978361669242658,
+      "grad_norm": 0.09498875334782204,
+      "learning_rate": 9.7533185278292e-06,
+      "loss": 0.3614,
+      "step": 2574
+    },
+    {
+      "epoch": 3.979907264296754,
+      "grad_norm": 0.11332292262691414,
+      "learning_rate": 9.725087557380282e-06,
+      "loss": 0.3699,
+      "step": 2575
+    },
+    {
+      "epoch": 3.98145285935085,
+      "grad_norm": 0.10488757950290599,
+      "learning_rate": 9.696891848220686e-06,
+      "loss": 0.3602,
+      "step": 2576
+    },
+    {
+      "epoch": 3.982998454404946,
+      "grad_norm": 0.09354094065496711,
+      "learning_rate": 9.668731433190048e-06,
+      "loss": 0.3584,
+      "step": 2577
+    },
+    {
+      "epoch": 3.984544049459042,
+      "grad_norm": 0.10307736561633109,
+      "learning_rate": 9.640606345086896e-06,
+      "loss": 0.358,
+      "step": 2578
+    },
+    {
+      "epoch": 3.9860896445131377,
+      "grad_norm": 0.0991188352213659,
+      "learning_rate": 9.612516616668612e-06,
+      "loss": 0.3636,
+      "step": 2579
+    },
+    {
+      "epoch": 3.9876352395672336,
+      "grad_norm": 0.09447328499261065,
+      "learning_rate": 9.58446228065138e-06,
+      "loss": 0.3643,
+      "step": 2580
+    },
+    {
+      "epoch": 3.989180834621329,
+      "grad_norm": 0.10263294541166457,
+      "learning_rate": 9.5564433697102e-06,
+      "loss": 0.372,
+      "step": 2581
+    },
+    {
+      "epoch": 3.990726429675425,
+      "grad_norm": 0.09626087265886038,
+      "learning_rate": 9.528459916478785e-06,
+      "loss": 0.3605,
+      "step": 2582
+    },
+    {
+      "epoch": 3.992272024729521,
+      "grad_norm": 0.10037091578716863,
+      "learning_rate": 9.500511953549534e-06,
+      "loss": 0.3724,
+      "step": 2583
+    },
+    {
+      "epoch": 3.993817619783617,
+      "grad_norm": 0.10011278941164424,
+      "learning_rate": 9.472599513473563e-06,
+      "loss": 0.3709,
+      "step": 2584
+    },
+    {
+      "epoch": 3.9953632148377123,
+      "grad_norm": 0.09600870645580256,
+      "learning_rate": 9.44472262876056e-06,
+      "loss": 0.3695,
+      "step": 2585
+    },
+    {
+      "epoch": 3.996908809891808,
+      "grad_norm": 0.09299273510515463,
+      "learning_rate": 9.416881331878812e-06,
+      "loss": 0.3692,
+      "step": 2586
+    },
+    {
+      "epoch": 3.998454404945904,
+      "grad_norm": 0.09811585389206162,
+      "learning_rate": 9.389075655255197e-06,
+      "loss": 0.3739,
+      "step": 2587
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.13954531960001063,
+      "learning_rate": 9.361305631275068e-06,
+      "loss": 0.3544,
+      "step": 2588
+    },
+    {
+      "epoch": 4.001545595054096,
+      "grad_norm": 0.10940343660912923,
+      "learning_rate": 9.333571292282243e-06,
+      "loss": 0.3502,
+      "step": 2589
+    },
+    {
+      "epoch": 4.003091190108192,
+      "grad_norm": 0.11320766478293215,
+      "learning_rate": 9.305872670579025e-06,
+      "loss": 0.3412,
+      "step": 2590
+    },
+    {
+      "epoch": 4.004636785162288,
+      "grad_norm": 0.11547323433966702,
+      "learning_rate": 9.278209798426082e-06,
+      "loss": 0.3428,
+      "step": 2591
+    },
+    {
+      "epoch": 4.006182380216384,
+      "grad_norm": 0.116703241057822,
+      "learning_rate": 9.250582708042452e-06,
+      "loss": 0.3461,
+      "step": 2592
+    },
+    {
+      "epoch": 4.0077279752704795,
+      "grad_norm": 0.11477518483073378,
+      "learning_rate": 9.222991431605472e-06,
+      "loss": 0.3538,
+      "step": 2593
+    },
+    {
+      "epoch": 4.0092735703245745,
+      "grad_norm": 0.11852802709349432,
+      "learning_rate": 9.195436001250822e-06,
+      "loss": 0.3511,
+      "step": 2594
+    },
+    {
+      "epoch": 4.0108191653786704,
+      "grad_norm": 0.1091164637954775,
+      "learning_rate": 9.167916449072392e-06,
+      "loss": 0.3484,
+      "step": 2595
+    },
+    {
+      "epoch": 4.012364760432766,
+      "grad_norm": 0.11766480809475463,
+      "learning_rate": 9.140432807122282e-06,
+      "loss": 0.3516,
+      "step": 2596
+    },
+    {
+      "epoch": 4.013910355486862,
+      "grad_norm": 0.11255519633655399,
+      "learning_rate": 9.112985107410801e-06,
+      "loss": 0.3436,
+      "step": 2597
+    },
+    {
+      "epoch": 4.015455950540958,
+      "grad_norm": 0.11371740300321591,
+      "learning_rate": 9.085573381906366e-06,
+      "loss": 0.3399,
+      "step": 2598
+    },
+    {
+      "epoch": 4.017001545595054,
+      "grad_norm": 0.12427309020683548,
+      "learning_rate": 9.058197662535501e-06,
+      "loss": 0.3488,
+      "step": 2599
+    },
+    {
+      "epoch": 4.01854714064915,
+      "grad_norm": 0.10365954811937023,
+      "learning_rate": 9.0308579811828e-06,
+      "loss": 0.3464,
+      "step": 2600
+    },
+    {
+      "epoch": 4.020092735703246,
+      "grad_norm": 0.14011142718575734,
+      "learning_rate": 9.00355436969087e-06,
+      "loss": 0.3524,
+      "step": 2601
+    },
+    {
+      "epoch": 4.021638330757342,
+      "grad_norm": 0.11021612748827103,
+      "learning_rate": 8.976286859860313e-06,
+      "loss": 0.3448,
+      "step": 2602
+    },
+    {
+      "epoch": 4.023183925811438,
+      "grad_norm": 0.10846563108738036,
+      "learning_rate": 8.949055483449708e-06,
+      "loss": 0.347,
+      "step": 2603
+    },
+    {
+      "epoch": 4.024729520865534,
+      "grad_norm": 0.10992941854538306,
+      "learning_rate": 8.921860272175515e-06,
+      "loss": 0.3508,
+      "step": 2604
+    },
+    {
+      "epoch": 4.0262751159196295,
+      "grad_norm": 0.1078991784580418,
+      "learning_rate": 8.894701257712084e-06,
+      "loss": 0.3507,
+      "step": 2605
+    },
+    {
+      "epoch": 4.0278207109737245,
+      "grad_norm": 0.09653065621329056,
+      "learning_rate": 8.867578471691609e-06,
+      "loss": 0.3538,
+      "step": 2606
+    },
+    {
+      "epoch": 4.02936630602782,
+      "grad_norm": 0.10408734736322096,
+      "learning_rate": 8.840491945704093e-06,
+      "loss": 0.3561,
+      "step": 2607
+    },
+    {
+      "epoch": 4.030911901081916,
+      "grad_norm": 0.10521353689745504,
+      "learning_rate": 8.813441711297281e-06,
+      "loss": 0.3406,
+      "step": 2608
+    },
+    {
+      "epoch": 4.032457496136012,
+      "grad_norm": 0.10747064265940268,
+      "learning_rate": 8.786427799976694e-06,
+      "loss": 0.3501,
+      "step": 2609
+    },
+    {
+      "epoch": 4.034003091190108,
+      "grad_norm": 0.116818989842645,
+      "learning_rate": 8.759450243205515e-06,
+      "loss": 0.3496,
+      "step": 2610
+    },
+    {
+      "epoch": 4.035548686244204,
+      "grad_norm": 0.09669902138624194,
+      "learning_rate": 8.73250907240458e-06,
+      "loss": 0.3513,
+      "step": 2611
+    },
+    {
+      "epoch": 4.0370942812983,
+      "grad_norm": 0.10804532239497597,
+      "learning_rate": 8.705604318952394e-06,
+      "loss": 0.3482,
+      "step": 2612
+    },
+    {
+      "epoch": 4.038639876352396,
+      "grad_norm": 0.3617095473258044,
+      "learning_rate": 8.678736014184972e-06,
+      "loss": 0.3485,
+      "step": 2613
+    },
+    {
+      "epoch": 4.040185471406492,
+      "grad_norm": 0.10391160408838113,
+      "learning_rate": 8.651904189395922e-06,
+      "loss": 0.3468,
+      "step": 2614
+    },
+    {
+      "epoch": 4.041731066460588,
+      "grad_norm": 0.10674311351942735,
+      "learning_rate": 8.625108875836377e-06,
+      "loss": 0.3543,
+      "step": 2615
+    },
+    {
+      "epoch": 4.043276661514684,
+      "grad_norm": 0.10698409102885607,
+      "learning_rate": 8.598350104714913e-06,
+      "loss": 0.3481,
+      "step": 2616
+    },
+    {
+      "epoch": 4.044822256568779,
+      "grad_norm": 0.09395026714733962,
+      "learning_rate": 8.571627907197548e-06,
+      "loss": 0.3511,
+      "step": 2617
+    },
+    {
+      "epoch": 4.0463678516228745,
+      "grad_norm": 0.11108836569277418,
+      "learning_rate": 8.54494231440774e-06,
+      "loss": 0.3469,
+      "step": 2618
+    },
+    {
+      "epoch": 4.04791344667697,
+      "grad_norm": 0.10152230122551807,
+      "learning_rate": 8.518293357426284e-06,
+      "loss": 0.3448,
+      "step": 2619
+    },
+    {
+      "epoch": 4.049459041731066,
+      "grad_norm": 0.10285114022610746,
+      "learning_rate": 8.491681067291279e-06,
+      "loss": 0.3462,
+      "step": 2620
+    },
+    {
+      "epoch": 4.051004636785162,
+      "grad_norm": 0.10585982621795423,
+      "learning_rate": 8.465105474998152e-06,
+      "loss": 0.3525,
+      "step": 2621
+    },
+    {
+      "epoch": 4.052550231839258,
+      "grad_norm": 0.09619262083884662,
+      "learning_rate": 8.438566611499603e-06,
+      "loss": 0.3469,
+      "step": 2622
+    },
+    {
+      "epoch": 4.054095826893354,
+      "grad_norm": 0.10071571746712828,
+      "learning_rate": 8.412064507705521e-06,
+      "loss": 0.3526,
+      "step": 2623
+    },
+    {
+      "epoch": 4.05564142194745,
+      "grad_norm": 0.10717668758266125,
+      "learning_rate": 8.385599194482985e-06,
+      "loss": 0.3541,
+      "step": 2624
+    },
+    {
+      "epoch": 4.057187017001546,
+      "grad_norm": 0.12277739259442917,
+      "learning_rate": 8.359170702656252e-06,
+      "loss": 0.3452,
+      "step": 2625
+    },
+    {
+      "epoch": 4.058732612055642,
+      "grad_norm": 0.10635331885322058,
+      "learning_rate": 8.33277906300666e-06,
+      "loss": 0.3546,
+      "step": 2626
+    },
+    {
+      "epoch": 4.060278207109738,
+      "grad_norm": 0.10375750544283012,
+      "learning_rate": 8.306424306272634e-06,
+      "loss": 0.3512,
+      "step": 2627
+    },
+    {
+      "epoch": 4.061823802163833,
+      "grad_norm": 0.10776883013627268,
+      "learning_rate": 8.280106463149655e-06,
+      "loss": 0.358,
+      "step": 2628
+    },
+    {
+      "epoch": 4.063369397217929,
+      "grad_norm": 0.08684956610614249,
+      "learning_rate": 8.253825564290192e-06,
+      "loss": 0.3526,
+      "step": 2629
+    },
+    {
+      "epoch": 4.0649149922720245,
+      "grad_norm": 0.1215722816083166,
+      "learning_rate": 8.227581640303688e-06,
+      "loss": 0.3542,
+      "step": 2630
+    },
+    {
+      "epoch": 4.06646058732612,
+      "grad_norm": 0.09308772726213017,
+      "learning_rate": 8.201374721756545e-06,
+      "loss": 0.3467,
+      "step": 2631
+    },
+    {
+      "epoch": 4.068006182380216,
+      "grad_norm": 0.09641533812883978,
+      "learning_rate": 8.175204839172041e-06,
+      "loss": 0.3521,
+      "step": 2632
+    },
+    {
+      "epoch": 4.069551777434312,
+      "grad_norm": 0.11116876668932495,
+      "learning_rate": 8.14907202303032e-06,
+      "loss": 0.3569,
+      "step": 2633
+    },
+    {
+      "epoch": 4.071097372488408,
+      "grad_norm": 0.08431105343847627,
+      "learning_rate": 8.122976303768357e-06,
+      "loss": 0.3429,
+      "step": 2634
+    },
+    {
+      "epoch": 4.072642967542504,
+      "grad_norm": 0.11091869787119317,
+      "learning_rate": 8.09691771177992e-06,
+      "loss": 0.343,
+      "step": 2635
+    },
+    {
+      "epoch": 4.0741885625966,
+      "grad_norm": 0.09404511826178034,
+      "learning_rate": 8.070896277415535e-06,
+      "loss": 0.3557,
+      "step": 2636
+    },
+    {
+      "epoch": 4.075734157650696,
+      "grad_norm": 0.09675306207406818,
+      "learning_rate": 8.044912030982464e-06,
+      "loss": 0.3477,
+      "step": 2637
+    },
+    {
+      "epoch": 4.077279752704792,
+      "grad_norm": 0.09294999121926695,
+      "learning_rate": 8.018965002744634e-06,
+      "loss": 0.345,
+      "step": 2638
+    },
+    {
+      "epoch": 4.078825347758887,
+      "grad_norm": 0.09093174893958986,
+      "learning_rate": 7.993055222922632e-06,
+      "loss": 0.3526,
+      "step": 2639
+    },
+    {
+      "epoch": 4.080370942812983,
+      "grad_norm": 0.09348658095917288,
+      "learning_rate": 7.96718272169367e-06,
+      "loss": 0.3451,
+      "step": 2640
+    },
+    {
+      "epoch": 4.0819165378670785,
+      "grad_norm": 0.09341498315647132,
+      "learning_rate": 7.94134752919153e-06,
+      "loss": 0.3467,
+      "step": 2641
+    },
+    {
+      "epoch": 4.083462132921174,
+      "grad_norm": 0.09435245349039612,
+      "learning_rate": 7.915549675506536e-06,
+      "loss": 0.3407,
+      "step": 2642
+    },
+    {
+      "epoch": 4.08500772797527,
+      "grad_norm": 0.09125878214059678,
+      "learning_rate": 7.889789190685548e-06,
+      "loss": 0.3475,
+      "step": 2643
+    },
+    {
+      "epoch": 4.086553323029366,
+      "grad_norm": 0.096903975224843,
+      "learning_rate": 7.864066104731885e-06,
+      "loss": 0.3398,
+      "step": 2644
+    },
+    {
+      "epoch": 4.088098918083462,
+      "grad_norm": 0.0901626823503655,
+      "learning_rate": 7.838380447605303e-06,
+      "loss": 0.3465,
+      "step": 2645
+    },
+    {
+      "epoch": 4.089644513137558,
+      "grad_norm": 0.0902459868780099,
+      "learning_rate": 7.812732249221962e-06,
+      "loss": 0.3485,
+      "step": 2646
+    },
+    {
+      "epoch": 4.091190108191654,
+      "grad_norm": 0.11066707787908245,
+      "learning_rate": 7.787121539454431e-06,
+      "loss": 0.3498,
+      "step": 2647
+    },
+    {
+      "epoch": 4.09273570324575,
+      "grad_norm": 0.10240484679057471,
+      "learning_rate": 7.761548348131564e-06,
+      "loss": 0.3459,
+      "step": 2648
+    },
+    {
+      "epoch": 4.094281298299846,
+      "grad_norm": 0.09512902311653947,
+      "learning_rate": 7.736012705038529e-06,
+      "loss": 0.3568,
+      "step": 2649
+    },
+    {
+      "epoch": 4.095826893353942,
+      "grad_norm": 0.1217732544379211,
+      "learning_rate": 7.710514639916807e-06,
+      "loss": 0.3466,
+      "step": 2650
+    },
+    {
+      "epoch": 4.097372488408037,
+      "grad_norm": 0.09070024923190873,
+      "learning_rate": 7.685054182464054e-06,
+      "loss": 0.3446,
+      "step": 2651
+    },
+    {
+      "epoch": 4.098918083462133,
+      "grad_norm": 0.0840054326889453,
+      "learning_rate": 7.659631362334146e-06,
+      "loss": 0.3453,
+      "step": 2652
+    },
+    {
+      "epoch": 4.1004636785162285,
+      "grad_norm": 0.1604455292133467,
+      "learning_rate": 7.634246209137139e-06,
+      "loss": 0.3526,
+      "step": 2653
+    },
+    {
+      "epoch": 4.102009273570324,
+      "grad_norm": 0.09067855239132425,
+      "learning_rate": 7.6088987524392025e-06,
+      "loss": 0.3548,
+      "step": 2654
+    },
+    {
+      "epoch": 4.10355486862442,
+      "grad_norm": 0.08831267056428002,
+      "learning_rate": 7.583589021762594e-06,
+      "loss": 0.3539,
+      "step": 2655
+    },
+    {
+      "epoch": 4.105100463678516,
+      "grad_norm": 0.10656832461232695,
+      "learning_rate": 7.5583170465856415e-06,
+      "loss": 0.3547,
+      "step": 2656
+    },
+    {
+      "epoch": 4.106646058732612,
+      "grad_norm": 0.11272606443660689,
+      "learning_rate": 7.5330828563427015e-06,
+      "loss": 0.3437,
+      "step": 2657
+    },
+    {
+      "epoch": 4.108191653786708,
+      "grad_norm": 0.10006764500407289,
+      "learning_rate": 7.5078864804241e-06,
+      "loss": 0.3581,
+      "step": 2658
+    },
+    {
+      "epoch": 4.109737248840804,
+      "grad_norm": 0.08699335455579792,
+      "learning_rate": 7.482727948176163e-06,
+      "loss": 0.343,
+      "step": 2659
+    },
+    {
+      "epoch": 4.1112828438949,
+      "grad_norm": 0.0891116336540287,
+      "learning_rate": 7.457607288901103e-06,
+      "loss": 0.3457,
+      "step": 2660
+    },
+    {
+      "epoch": 4.112828438948996,
+      "grad_norm": 0.10163097347052266,
+      "learning_rate": 7.432524531857037e-06,
+      "loss": 0.3531,
+      "step": 2661
+    },
+    {
+      "epoch": 4.114374034003091,
+      "grad_norm": 0.09494169578201946,
+      "learning_rate": 7.40747970625793e-06,
+      "loss": 0.3603,
+      "step": 2662
+    },
+    {
+      "epoch": 4.115919629057187,
+      "grad_norm": 0.09402303727235024,
+      "learning_rate": 7.382472841273576e-06,
+      "loss": 0.3387,
+      "step": 2663
+    },
+    {
+      "epoch": 4.117465224111283,
+      "grad_norm": 0.11541775732047659,
+      "learning_rate": 7.35750396602954e-06,
+      "loss": 0.3529,
+      "step": 2664
+    },
+    {
+      "epoch": 4.1190108191653785,
+      "grad_norm": 0.0968165489644101,
+      "learning_rate": 7.332573109607173e-06,
+      "loss": 0.3537,
+      "step": 2665
+    },
+    {
+      "epoch": 4.120556414219474,
+      "grad_norm": 0.09467829102180705,
+      "learning_rate": 7.307680301043514e-06,
+      "loss": 0.3503,
+      "step": 2666
+    },
+    {
+      "epoch": 4.12210200927357,
+      "grad_norm": 0.09864251003053498,
+      "learning_rate": 7.282825569331296e-06,
+      "loss": 0.3514,
+      "step": 2667
+    },
+    {
+      "epoch": 4.123647604327666,
+      "grad_norm": 0.10469599803133331,
+      "learning_rate": 7.258008943418913e-06,
+      "loss": 0.3499,
+      "step": 2668
+    },
+    {
+      "epoch": 4.125193199381762,
+      "grad_norm": 0.08812596124563764,
+      "learning_rate": 7.233230452210364e-06,
+      "loss": 0.3521,
+      "step": 2669
+    },
+    {
+      "epoch": 4.126738794435858,
+      "grad_norm": 0.10214262863411051,
+      "learning_rate": 7.208490124565238e-06,
+      "loss": 0.346,
+      "step": 2670
+    },
+    {
+      "epoch": 4.128284389489954,
+      "grad_norm": 0.09835108239310617,
+      "learning_rate": 7.183787989298681e-06,
+      "loss": 0.3507,
+      "step": 2671
+    },
+    {
+      "epoch": 4.12982998454405,
+      "grad_norm": 0.09020994042023699,
+      "learning_rate": 7.159124075181347e-06,
+      "loss": 0.3485,
+      "step": 2672
+    },
+    {
+      "epoch": 4.131375579598146,
+      "grad_norm": 0.10060564050468568,
+      "learning_rate": 7.134498410939383e-06,
+      "loss": 0.3437,
+      "step": 2673
+    },
+    {
+      "epoch": 4.132921174652241,
+      "grad_norm": 0.1186331079330418,
+      "learning_rate": 7.109911025254366e-06,
+      "loss": 0.3568,
+      "step": 2674
+    },
+    {
+      "epoch": 4.134466769706337,
+      "grad_norm": 0.09767985012419406,
+      "learning_rate": 7.08536194676333e-06,
+      "loss": 0.3458,
+      "step": 2675
+    },
+    {
+      "epoch": 4.1360123647604325,
+      "grad_norm": 0.1028522770072495,
+      "learning_rate": 7.060851204058643e-06,
+      "loss": 0.3484,
+      "step": 2676
+    },
+    {
+      "epoch": 4.1375579598145285,
+      "grad_norm": 0.09264338114853868,
+      "learning_rate": 7.036378825688044e-06,
+      "loss": 0.3522,
+      "step": 2677
+    },
+    {
+      "epoch": 4.139103554868624,
+      "grad_norm": 0.09341069540225162,
+      "learning_rate": 7.011944840154612e-06,
+      "loss": 0.346,
+      "step": 2678
+    },
+    {
+      "epoch": 4.14064914992272,
+      "grad_norm": 0.09763292400983055,
+      "learning_rate": 6.987549275916681e-06,
+      "loss": 0.3411,
+      "step": 2679
+    },
+    {
+      "epoch": 4.142194744976816,
+      "grad_norm": 0.09191041021765126,
+      "learning_rate": 6.963192161387837e-06,
+      "loss": 0.3496,
+      "step": 2680
+    },
+    {
+      "epoch": 4.143740340030912,
+      "grad_norm": 0.09712023598507023,
+      "learning_rate": 6.938873524936905e-06,
+      "loss": 0.3461,
+      "step": 2681
+    },
+    {
+      "epoch": 4.145285935085008,
+      "grad_norm": 0.09889849220678522,
+      "learning_rate": 6.914593394887883e-06,
+      "loss": 0.3498,
+      "step": 2682
+    },
+    {
+      "epoch": 4.146831530139104,
+      "grad_norm": 0.09735515745641965,
+      "learning_rate": 6.890351799519885e-06,
+      "loss": 0.3454,
+      "step": 2683
+    },
+    {
+      "epoch": 4.1483771251932,
+      "grad_norm": 0.1000739488906435,
+      "learning_rate": 6.866148767067211e-06,
+      "loss": 0.3579,
+      "step": 2684
+    },
+    {
+      "epoch": 4.149922720247295,
+      "grad_norm": 0.09416506863271783,
+      "learning_rate": 6.8419843257192e-06,
+      "loss": 0.3531,
+      "step": 2685
+    },
+    {
+      "epoch": 4.151468315301391,
+      "grad_norm": 0.08661446077850428,
+      "learning_rate": 6.817858503620245e-06,
+      "loss": 0.3456,
+      "step": 2686
+    },
+    {
+      "epoch": 4.153013910355487,
+      "grad_norm": 0.08864727511068461,
+      "learning_rate": 6.79377132886979e-06,
+      "loss": 0.3539,
+      "step": 2687
+    },
+    {
+      "epoch": 4.1545595054095825,
+      "grad_norm": 0.09814646368049065,
+      "learning_rate": 6.769722829522236e-06,
+      "loss": 0.3495,
+      "step": 2688
+    },
+    {
+      "epoch": 4.156105100463678,
+      "grad_norm": 0.10086265879889979,
+      "learning_rate": 6.74571303358695e-06,
+      "loss": 0.3524,
+      "step": 2689
+    },
+    {
+      "epoch": 4.157650695517774,
+      "grad_norm": 0.08681264189018105,
+      "learning_rate": 6.721741969028213e-06,
+      "loss": 0.348,
+      "step": 2690
+    },
+    {
+      "epoch": 4.15919629057187,
+      "grad_norm": 0.09231883798002785,
+      "learning_rate": 6.6978096637652134e-06,
+      "loss": 0.3439,
+      "step": 2691
+    },
+    {
+      "epoch": 4.160741885625966,
+      "grad_norm": 0.09883183887807469,
+      "learning_rate": 6.673916145671966e-06,
+      "loss": 0.352,
+      "step": 2692
+    },
+    {
+      "epoch": 4.162287480680062,
+      "grad_norm": 0.0929331735681632,
+      "learning_rate": 6.650061442577356e-06,
+      "loss": 0.344,
+      "step": 2693
+    },
+    {
+      "epoch": 4.163833075734158,
+      "grad_norm": 0.08868671769328039,
+      "learning_rate": 6.626245582265021e-06,
+      "loss": 0.3426,
+      "step": 2694
+    },
+    {
+      "epoch": 4.165378670788254,
+      "grad_norm": 0.09008028624081477,
+      "learning_rate": 6.602468592473368e-06,
+      "loss": 0.3479,
+      "step": 2695
+    },
+    {
+      "epoch": 4.166924265842349,
+      "grad_norm": 0.09592653299689341,
+      "learning_rate": 6.578730500895539e-06,
+      "loss": 0.3454,
+      "step": 2696
+    },
+    {
+      "epoch": 4.168469860896445,
+      "grad_norm": 0.09195938812380217,
+      "learning_rate": 6.555031335179362e-06,
+      "loss": 0.3449,
+      "step": 2697
+    },
+    {
+      "epoch": 4.170015455950541,
+      "grad_norm": 0.0979280950028967,
+      "learning_rate": 6.53137112292733e-06,
+      "loss": 0.3526,
+      "step": 2698
+    },
+    {
+      "epoch": 4.171561051004637,
+      "grad_norm": 0.09009923736562045,
+      "learning_rate": 6.507749891696584e-06,
+      "loss": 0.3509,
+      "step": 2699
+    },
+    {
+      "epoch": 4.1731066460587325,
+      "grad_norm": 0.09900863694145302,
+      "learning_rate": 6.484167668998838e-06,
+      "loss": 0.3573,
+      "step": 2700
+    },
+    {
+      "epoch": 4.174652241112828,
+      "grad_norm": 0.09786973451180507,
+      "learning_rate": 6.46062448230039e-06,
+      "loss": 0.3495,
+      "step": 2701
+    },
+    {
+      "epoch": 4.176197836166924,
+      "grad_norm": 0.08362971408681666,
+      "learning_rate": 6.43712035902206e-06,
+      "loss": 0.3458,
+      "step": 2702
+    },
+    {
+      "epoch": 4.17774343122102,
+      "grad_norm": 0.09525291321654723,
+      "learning_rate": 6.413655326539179e-06,
+      "loss": 0.3511,
+      "step": 2703
+    },
+    {
+      "epoch": 4.179289026275116,
+      "grad_norm": 0.10093476900173766,
+      "learning_rate": 6.390229412181548e-06,
+      "loss": 0.3585,
+      "step": 2704
+    },
+    {
+      "epoch": 4.180834621329212,
+      "grad_norm": 0.08978894281932101,
+      "learning_rate": 6.3668426432333995e-06,
+      "loss": 0.3548,
+      "step": 2705
+    },
+    {
+      "epoch": 4.182380216383308,
+      "grad_norm": 0.09246920589687538,
+      "learning_rate": 6.34349504693339e-06,
+      "loss": 0.3423,
+      "step": 2706
+    },
+    {
+      "epoch": 4.183925811437404,
+      "grad_norm": 0.09130903497926467,
+      "learning_rate": 6.32018665047454e-06,
+      "loss": 0.3501,
+      "step": 2707
+    },
+    {
+      "epoch": 4.185471406491499,
+      "grad_norm": 0.08660881253621892,
+      "learning_rate": 6.296917481004201e-06,
+      "loss": 0.3468,
+      "step": 2708
+    },
+    {
+      "epoch": 4.187017001545595,
+      "grad_norm": 0.08377527635446212,
+      "learning_rate": 6.273687565624067e-06,
+      "loss": 0.3542,
+      "step": 2709
+    },
+    {
+      "epoch": 4.188562596599691,
+      "grad_norm": 0.08866327627783151,
+      "learning_rate": 6.250496931390104e-06,
+      "loss": 0.3493,
+      "step": 2710
+    },
+    {
+      "epoch": 4.190108191653787,
+      "grad_norm": 0.08945807864644795,
+      "learning_rate": 6.227345605312489e-06,
+      "loss": 0.3403,
+      "step": 2711
+    },
+    {
+      "epoch": 4.1916537867078825,
+      "grad_norm": 0.0887623289261429,
+      "learning_rate": 6.20423361435567e-06,
+      "loss": 0.3528,
+      "step": 2712
+    },
+    {
+      "epoch": 4.193199381761978,
+      "grad_norm": 0.08680136607627459,
+      "learning_rate": 6.181160985438253e-06,
+      "loss": 0.3475,
+      "step": 2713
+    },
+    {
+      "epoch": 4.194744976816074,
+      "grad_norm": 0.09505680088025242,
+      "learning_rate": 6.158127745432994e-06,
+      "loss": 0.3474,
+      "step": 2714
+    },
+    {
+      "epoch": 4.19629057187017,
+      "grad_norm": 0.0910181437160273,
+      "learning_rate": 6.135133921166798e-06,
+      "loss": 0.3462,
+      "step": 2715
+    },
+    {
+      "epoch": 4.197836166924266,
+      "grad_norm": 0.08606960293504413,
+      "learning_rate": 6.112179539420644e-06,
+      "loss": 0.3506,
+      "step": 2716
+    },
+    {
+      "epoch": 4.199381761978362,
+      "grad_norm": 0.09129921444068308,
+      "learning_rate": 6.089264626929572e-06,
+      "loss": 0.3532,
+      "step": 2717
+    },
+    {
+      "epoch": 4.200927357032458,
+      "grad_norm": 0.09510092578536367,
+      "learning_rate": 6.0663892103826595e-06,
+      "loss": 0.353,
+      "step": 2718
+    },
+    {
+      "epoch": 4.202472952086553,
+      "grad_norm": 0.09601044093874106,
+      "learning_rate": 6.043553316422972e-06,
+      "loss": 0.3507,
+      "step": 2719
+    },
+    {
+      "epoch": 4.204018547140649,
+      "grad_norm": 0.10019579809178113,
+      "learning_rate": 6.0207569716475456e-06,
+      "loss": 0.3391,
+      "step": 2720
+    },
+    {
+      "epoch": 4.205564142194745,
+      "grad_norm": 0.09149772335796025,
+      "learning_rate": 5.998000202607372e-06,
+      "loss": 0.3525,
+      "step": 2721
+    },
+    {
+      "epoch": 4.207109737248841,
+      "grad_norm": 0.09678861359711653,
+      "learning_rate": 5.975283035807326e-06,
+      "loss": 0.3431,
+      "step": 2722
+    },
+    {
+      "epoch": 4.2086553323029365,
+      "grad_norm": 0.10770337561171722,
+      "learning_rate": 5.952605497706172e-06,
+      "loss": 0.3512,
+      "step": 2723
+    },
+    {
+      "epoch": 4.210200927357032,
+      "grad_norm": 0.09254643928336213,
+      "learning_rate": 5.929967614716506e-06,
+      "loss": 0.3488,
+      "step": 2724
+    },
+    {
+      "epoch": 4.211746522411128,
+      "grad_norm": 0.09719212904806131,
+      "learning_rate": 5.907369413204742e-06,
+      "loss": 0.3486,
+      "step": 2725
+    },
+    {
+      "epoch": 4.213292117465224,
+      "grad_norm": 0.10351728017384307,
+      "learning_rate": 5.884810919491078e-06,
+      "loss": 0.3498,
+      "step": 2726
+    },
+    {
+      "epoch": 4.21483771251932,
+      "grad_norm": 0.09774772910336196,
+      "learning_rate": 5.8622921598494764e-06,
+      "loss": 0.3556,
+      "step": 2727
+    },
+    {
+      "epoch": 4.216383307573416,
+      "grad_norm": 0.08249445366236399,
+      "learning_rate": 5.839813160507608e-06,
+      "loss": 0.3495,
+      "step": 2728
+    },
+    {
+      "epoch": 4.217928902627512,
+      "grad_norm": 0.08774296533632595,
+      "learning_rate": 5.817373947646827e-06,
+      "loss": 0.3456,
+      "step": 2729
+    },
+    {
+      "epoch": 4.219474497681608,
+      "grad_norm": 0.09044142173089888,
+      "learning_rate": 5.794974547402166e-06,
+      "loss": 0.3484,
+      "step": 2730
+    },
+    {
+      "epoch": 4.221020092735703,
+      "grad_norm": 0.0925796711635186,
+      "learning_rate": 5.77261498586228e-06,
+      "loss": 0.3553,
+      "step": 2731
+    },
+    {
+      "epoch": 4.222565687789799,
+      "grad_norm": 0.08649124670615871,
+      "learning_rate": 5.750295289069416e-06,
+      "loss": 0.3545,
+      "step": 2732
+    },
+    {
+      "epoch": 4.224111282843895,
+      "grad_norm": 0.0867973445003958,
+      "learning_rate": 5.728015483019396e-06,
+      "loss": 0.3471,
+      "step": 2733
+    },
+    {
+      "epoch": 4.225656877897991,
+      "grad_norm": 0.08862548922716557,
+      "learning_rate": 5.7057755936615975e-06,
+      "loss": 0.3452,
+      "step": 2734
+    },
+    {
+      "epoch": 4.2272024729520865,
+      "grad_norm": 0.0915530648024646,
+      "learning_rate": 5.683575646898884e-06,
+      "loss": 0.3545,
+      "step": 2735
+    },
+    {
+      "epoch": 4.228748068006182,
+      "grad_norm": 0.08613890956706916,
+      "learning_rate": 5.6614156685875954e-06,
+      "loss": 0.3463,
+      "step": 2736
+    },
+    {
+      "epoch": 4.230293663060278,
+      "grad_norm": 0.08691297495344168,
+      "learning_rate": 5.639295684537547e-06,
+      "loss": 0.3445,
+      "step": 2737
+    },
+    {
+      "epoch": 4.231839258114374,
+      "grad_norm": 0.08310811067759995,
+      "learning_rate": 5.617215720511962e-06,
+      "loss": 0.3526,
+      "step": 2738
+    },
+    {
+      "epoch": 4.23338485316847,
+      "grad_norm": 0.08343953697177237,
+      "learning_rate": 5.595175802227419e-06,
+      "loss": 0.3484,
+      "step": 2739
+    },
+    {
+      "epoch": 4.234930448222566,
+      "grad_norm": 0.0861910567941286,
+      "learning_rate": 5.573175955353903e-06,
+      "loss": 0.3522,
+      "step": 2740
+    },
+    {
+      "epoch": 4.236476043276662,
+      "grad_norm": 0.0777186794836491,
+      "learning_rate": 5.551216205514704e-06,
+      "loss": 0.3505,
+      "step": 2741
+    },
+    {
+      "epoch": 4.238021638330757,
+      "grad_norm": 0.08144096840432163,
+      "learning_rate": 5.529296578286407e-06,
+      "loss": 0.3446,
+      "step": 2742
+    },
+    {
+      "epoch": 4.239567233384853,
+      "grad_norm": 0.08835441368792356,
+      "learning_rate": 5.507417099198886e-06,
+      "loss": 0.3484,
+      "step": 2743
+    },
+    {
+      "epoch": 4.241112828438949,
+      "grad_norm": 0.08126870596595956,
+      "learning_rate": 5.485577793735237e-06,
+      "loss": 0.3505,
+      "step": 2744
+    },
+    {
+      "epoch": 4.242658423493045,
+      "grad_norm": 0.08121575444343013,
+      "learning_rate": 5.463778687331771e-06,
+      "loss": 0.352,
+      "step": 2745
+    },
+    {
+      "epoch": 4.244204018547141,
+      "grad_norm": 0.08322868248497733,
+      "learning_rate": 5.4420198053779735e-06,
+      "loss": 0.3401,
+      "step": 2746
+    },
+    {
+      "epoch": 4.2457496136012365,
+      "grad_norm": 0.09184739288112194,
+      "learning_rate": 5.420301173216484e-06,
+      "loss": 0.3507,
+      "step": 2747
+    },
+    {
+      "epoch": 4.247295208655332,
+      "grad_norm": 0.084403341690368,
+      "learning_rate": 5.398622816143064e-06,
+      "loss": 0.3518,
+      "step": 2748
+    },
+    {
+      "epoch": 4.248840803709428,
+      "grad_norm": 0.08224459987007703,
+      "learning_rate": 5.376984759406578e-06,
+      "loss": 0.3516,
+      "step": 2749
+    },
+    {
+      "epoch": 4.250386398763524,
+      "grad_norm": 0.08629277157699956,
+      "learning_rate": 5.355387028208938e-06,
+      "loss": 0.3549,
+      "step": 2750
+    },
+    {
+      "epoch": 4.25193199381762,
+      "grad_norm": 0.08185089574331944,
+      "learning_rate": 5.333829647705085e-06,
+      "loss": 0.3457,
+      "step": 2751
+    },
+    {
+      "epoch": 4.253477588871716,
+      "grad_norm": 0.08190279325500587,
+      "learning_rate": 5.312312643002972e-06,
+      "loss": 0.352,
+      "step": 2752
+    },
+    {
+      "epoch": 4.255023183925811,
+      "grad_norm": 0.0892876994651363,
+      "learning_rate": 5.290836039163529e-06,
+      "loss": 0.349,
+      "step": 2753
+    },
+    {
+      "epoch": 4.256568778979907,
+      "grad_norm": 0.08447915478996387,
+      "learning_rate": 5.269399861200612e-06,
+      "loss": 0.3459,
+      "step": 2754
+    },
+    {
+      "epoch": 4.258114374034003,
+      "grad_norm": 0.07748852606445458,
+      "learning_rate": 5.248004134081033e-06,
+      "loss": 0.3524,
+      "step": 2755
+    },
+    {
+      "epoch": 4.259659969088099,
+      "grad_norm": 0.08730300384433576,
+      "learning_rate": 5.226648882724448e-06,
+      "loss": 0.3443,
+      "step": 2756
+    },
+    {
+      "epoch": 4.261205564142195,
+      "grad_norm": 0.08505194826000924,
+      "learning_rate": 5.205334132003392e-06,
+      "loss": 0.3457,
+      "step": 2757
+    },
+    {
+      "epoch": 4.2627511591962906,
+      "grad_norm": 0.08362832295747517,
+      "learning_rate": 5.1840599067432215e-06,
+      "loss": 0.3556,
+      "step": 2758
+    },
+    {
+      "epoch": 4.2642967542503865,
+      "grad_norm": 0.08149843969736577,
+      "learning_rate": 5.162826231722102e-06,
+      "loss": 0.3517,
+      "step": 2759
+    },
+    {
+      "epoch": 4.265842349304482,
+      "grad_norm": 0.08443772007896898,
+      "learning_rate": 5.141633131670962e-06,
+      "loss": 0.35,
+      "step": 2760
+    },
+    {
+      "epoch": 4.267387944358578,
+      "grad_norm": 0.07929956288700393,
+      "learning_rate": 5.120480631273461e-06,
+      "loss": 0.3424,
+      "step": 2761
+    },
+    {
+      "epoch": 4.268933539412674,
+      "grad_norm": 0.08258796502296727,
+      "learning_rate": 5.0993687551660075e-06,
+      "loss": 0.3442,
+      "step": 2762
+    },
+    {
+      "epoch": 4.27047913446677,
+      "grad_norm": 0.08508110512188652,
+      "learning_rate": 5.078297527937661e-06,
+      "loss": 0.3528,
+      "step": 2763
+    },
+    {
+      "epoch": 4.272024729520865,
+      "grad_norm": 0.08895252203795366,
+      "learning_rate": 5.057266974130142e-06,
+      "loss": 0.3432,
+      "step": 2764
+    },
+    {
+      "epoch": 4.273570324574961,
+      "grad_norm": 0.08386139591937235,
+      "learning_rate": 5.036277118237833e-06,
+      "loss": 0.3468,
+      "step": 2765
+    },
+    {
+      "epoch": 4.275115919629057,
+      "grad_norm": 0.08920822772839218,
+      "learning_rate": 5.015327984707656e-06,
+      "loss": 0.3499,
+      "step": 2766
+    },
+    {
+      "epoch": 4.276661514683153,
+      "grad_norm": 0.08772240930867956,
+      "learning_rate": 4.9944195979391374e-06,
+      "loss": 0.3429,
+      "step": 2767
+    },
+    {
+      "epoch": 4.278207109737249,
+      "grad_norm": 0.09196911526263683,
+      "learning_rate": 4.973551982284366e-06,
+      "loss": 0.3497,
+      "step": 2768
+    },
+    {
+      "epoch": 4.279752704791345,
+      "grad_norm": 0.08052787163782076,
+      "learning_rate": 4.952725162047904e-06,
+      "loss": 0.3494,
+      "step": 2769
+    },
+    {
+      "epoch": 4.2812982998454405,
+      "grad_norm": 0.09452001276413037,
+      "learning_rate": 4.931939161486816e-06,
+      "loss": 0.3607,
+      "step": 2770
+    },
+    {
+      "epoch": 4.282843894899536,
+      "grad_norm": 0.08611093316743922,
+      "learning_rate": 4.911194004810642e-06,
+      "loss": 0.3569,
+      "step": 2771
+    },
+    {
+      "epoch": 4.284389489953632,
+      "grad_norm": 0.08298827219953163,
+      "learning_rate": 4.890489716181317e-06,
+      "loss": 0.3457,
+      "step": 2772
+    },
+    {
+      "epoch": 4.285935085007728,
+      "grad_norm": 0.08747580977147676,
+      "learning_rate": 4.8698263197132e-06,
+      "loss": 0.3484,
+      "step": 2773
+    },
+    {
+      "epoch": 4.287480680061824,
+      "grad_norm": 0.08509843362345738,
+      "learning_rate": 4.849203839473018e-06,
+      "loss": 0.3511,
+      "step": 2774
+    },
+    {
+      "epoch": 4.289026275115919,
+      "grad_norm": 0.08384040845348897,
+      "learning_rate": 4.828622299479833e-06,
+      "loss": 0.3518,
+      "step": 2775
+    },
+    {
+      "epoch": 4.290571870170015,
+      "grad_norm": 0.08225738324803471,
+      "learning_rate": 4.808081723705029e-06,
+      "loss": 0.3494,
+      "step": 2776
+    },
+    {
+      "epoch": 4.292117465224111,
+      "grad_norm": 0.08929428170089522,
+      "learning_rate": 4.787582136072302e-06,
+      "loss": 0.3468,
+      "step": 2777
+    },
+    {
+      "epoch": 4.293663060278207,
+      "grad_norm": 0.07893963360893906,
+      "learning_rate": 4.7671235604575735e-06,
+      "loss": 0.3458,
+      "step": 2778
+    },
+    {
+      "epoch": 4.295208655332303,
+      "grad_norm": 0.08680221156527582,
+      "learning_rate": 4.746706020689016e-06,
+      "loss": 0.3426,
+      "step": 2779
+    },
+    {
+      "epoch": 4.296754250386399,
+      "grad_norm": 0.0799084807848644,
+      "learning_rate": 4.726329540547005e-06,
+      "loss": 0.3573,
+      "step": 2780
+    },
+    {
+      "epoch": 4.298299845440495,
+      "grad_norm": 0.08603943550132623,
+      "learning_rate": 4.705994143764092e-06,
+      "loss": 0.3516,
+      "step": 2781
+    },
+    {
+      "epoch": 4.2998454404945905,
+      "grad_norm": 0.0848692785094291,
+      "learning_rate": 4.685699854024983e-06,
+      "loss": 0.3532,
+      "step": 2782
+    },
+    {
+      "epoch": 4.301391035548686,
+      "grad_norm": 0.08413350754023934,
+      "learning_rate": 4.665446694966491e-06,
+      "loss": 0.3563,
+      "step": 2783
+    },
+    {
+      "epoch": 4.302936630602782,
+      "grad_norm": 0.08726381849055063,
+      "learning_rate": 4.645234690177556e-06,
+      "loss": 0.3511,
+      "step": 2784
+    },
+    {
+      "epoch": 4.304482225656878,
+      "grad_norm": 0.07768866576856157,
+      "learning_rate": 4.625063863199159e-06,
+      "loss": 0.3455,
+      "step": 2785
+    },
+    {
+      "epoch": 4.306027820710974,
+      "grad_norm": 0.08596411111222517,
+      "learning_rate": 4.604934237524324e-06,
+      "loss": 0.3454,
+      "step": 2786
+    },
+    {
+      "epoch": 4.307573415765069,
+      "grad_norm": 0.07657915076559339,
+      "learning_rate": 4.584845836598093e-06,
+      "loss": 0.3411,
+      "step": 2787
+    },
+    {
+      "epoch": 4.309119010819165,
+      "grad_norm": 0.08908154360320228,
+      "learning_rate": 4.5647986838174904e-06,
+      "loss": 0.3553,
+      "step": 2788
+    },
+    {
+      "epoch": 4.310664605873261,
+      "grad_norm": 0.08381492332227229,
+      "learning_rate": 4.544792802531501e-06,
+      "loss": 0.3442,
+      "step": 2789
+    },
+    {
+      "epoch": 4.312210200927357,
+      "grad_norm": 0.07893020049897356,
+      "learning_rate": 4.5248282160410465e-06,
+      "loss": 0.3505,
+      "step": 2790
+    },
+    {
+      "epoch": 4.313755795981453,
+      "grad_norm": 0.09488830046072966,
+      "learning_rate": 4.504904947598938e-06,
+      "loss": 0.3581,
+      "step": 2791
+    },
+    {
+      "epoch": 4.315301391035549,
+      "grad_norm": 0.08400346949719387,
+      "learning_rate": 4.485023020409874e-06,
+      "loss": 0.3469,
+      "step": 2792
+    },
+    {
+      "epoch": 4.316846986089645,
+      "grad_norm": 0.09019417776449458,
+      "learning_rate": 4.465182457630408e-06,
+      "loss": 0.344,
+      "step": 2793
+    },
+    {
+      "epoch": 4.3183925811437405,
+      "grad_norm": 0.08258223886785825,
+      "learning_rate": 4.445383282368898e-06,
+      "loss": 0.3455,
+      "step": 2794
+    },
+    {
+      "epoch": 4.319938176197836,
+      "grad_norm": 0.08248713459639526,
+      "learning_rate": 4.425625517685501e-06,
+      "loss": 0.3499,
+      "step": 2795
+    },
+    {
+      "epoch": 4.321483771251932,
+      "grad_norm": 0.08953310369330902,
+      "learning_rate": 4.40590918659217e-06,
+      "loss": 0.3556,
+      "step": 2796
+    },
+    {
+      "epoch": 4.323029366306028,
+      "grad_norm": 0.0844215981666415,
+      "learning_rate": 4.3862343120525664e-06,
+      "loss": 0.3414,
+      "step": 2797
+    },
+    {
+      "epoch": 4.324574961360124,
+      "grad_norm": 0.0810130805737031,
+      "learning_rate": 4.36660091698208e-06,
+      "loss": 0.3603,
+      "step": 2798
+    },
+    {
+      "epoch": 4.326120556414219,
+      "grad_norm": 0.08542137093061943,
+      "learning_rate": 4.347009024247801e-06,
+      "loss": 0.346,
+      "step": 2799
+    },
+    {
+      "epoch": 4.327666151468315,
+      "grad_norm": 0.08815077895427041,
+      "learning_rate": 4.327458656668464e-06,
+      "loss": 0.3579,
+      "step": 2800
+    },
+    {
+      "epoch": 4.329211746522411,
+      "grad_norm": 0.07818746188271519,
+      "learning_rate": 4.307949837014445e-06,
+      "loss": 0.3483,
+      "step": 2801
+    },
+    {
+      "epoch": 4.330757341576507,
+      "grad_norm": 0.0808695254180279,
+      "learning_rate": 4.288482588007732e-06,
+      "loss": 0.3522,
+      "step": 2802
+    },
+    {
+      "epoch": 4.332302936630603,
+      "grad_norm": 0.08608424581458866,
+      "learning_rate": 4.269056932321891e-06,
+      "loss": 0.3478,
+      "step": 2803
+    },
+    {
+      "epoch": 4.333848531684699,
+      "grad_norm": 0.07895846412020385,
+      "learning_rate": 4.249672892582042e-06,
+      "loss": 0.348,
+      "step": 2804
+    },
+    {
+      "epoch": 4.3353941267387945,
+      "grad_norm": 0.07942966336374079,
+      "learning_rate": 4.230330491364853e-06,
+      "loss": 0.3481,
+      "step": 2805
+    },
+    {
+      "epoch": 4.3369397217928904,
+      "grad_norm": 0.0800726247678195,
+      "learning_rate": 4.211029751198474e-06,
+      "loss": 0.3468,
+      "step": 2806
+    },
+    {
+      "epoch": 4.338485316846986,
+      "grad_norm": 0.07880525487945812,
+      "learning_rate": 4.19177069456254e-06,
+      "loss": 0.3501,
+      "step": 2807
+    },
+    {
+      "epoch": 4.340030911901082,
+      "grad_norm": 0.07822112072857317,
+      "learning_rate": 4.172553343888139e-06,
+      "loss": 0.3456,
+      "step": 2808
+    },
+    {
+      "epoch": 4.341576506955178,
+      "grad_norm": 0.07982983239806374,
+      "learning_rate": 4.153377721557781e-06,
+      "loss": 0.3499,
+      "step": 2809
+    },
+    {
+      "epoch": 4.343122102009273,
+      "grad_norm": 0.08397108661721098,
+      "learning_rate": 4.134243849905373e-06,
+      "loss": 0.3489,
+      "step": 2810
+    },
+    {
+      "epoch": 4.344667697063369,
+      "grad_norm": 0.08417824581971808,
+      "learning_rate": 4.115151751216195e-06,
+      "loss": 0.3409,
+      "step": 2811
+    },
+    {
+      "epoch": 4.346213292117465,
+      "grad_norm": 0.08356380649469387,
+      "learning_rate": 4.0961014477268876e-06,
+      "loss": 0.3527,
+      "step": 2812
+    },
+    {
+      "epoch": 4.347758887171561,
+      "grad_norm": 0.08337450375715438,
+      "learning_rate": 4.0770929616253954e-06,
+      "loss": 0.3513,
+      "step": 2813
+    },
+    {
+      "epoch": 4.349304482225657,
+      "grad_norm": 0.0794790520176344,
+      "learning_rate": 4.0581263150509634e-06,
+      "loss": 0.3441,
+      "step": 2814
+    },
+    {
+      "epoch": 4.350850077279753,
+      "grad_norm": 0.07989596049121159,
+      "learning_rate": 4.0392015300941075e-06,
+      "loss": 0.3535,
+      "step": 2815
+    },
+    {
+      "epoch": 4.352395672333849,
+      "grad_norm": 0.08258184784831214,
+      "learning_rate": 4.020318628796589e-06,
+      "loss": 0.3538,
+      "step": 2816
+    },
+    {
+      "epoch": 4.3539412673879445,
+      "grad_norm": 0.08321302309092998,
+      "learning_rate": 4.0014776331513786e-06,
+      "loss": 0.3413,
+      "step": 2817
+    },
+    {
+      "epoch": 4.35548686244204,
+      "grad_norm": 0.07972304848581438,
+      "learning_rate": 3.982678565102651e-06,
+      "loss": 0.3432,
+      "step": 2818
+    },
+    {
+      "epoch": 4.357032457496136,
+      "grad_norm": 0.07740206915020002,
+      "learning_rate": 3.963921446545742e-06,
+      "loss": 0.3469,
+      "step": 2819
+    },
+    {
+      "epoch": 4.358578052550232,
+      "grad_norm": 0.08194551709179719,
+      "learning_rate": 3.945206299327118e-06,
+      "loss": 0.3483,
+      "step": 2820
+    },
+    {
+      "epoch": 4.360123647604327,
+      "grad_norm": 0.07644868697025502,
+      "learning_rate": 3.9265331452444e-06,
+      "loss": 0.3476,
+      "step": 2821
+    },
+    {
+      "epoch": 4.361669242658423,
+      "grad_norm": 0.08526880526334188,
+      "learning_rate": 3.9079020060462406e-06,
+      "loss": 0.3516,
+      "step": 2822
+    },
+    {
+      "epoch": 4.363214837712519,
+      "grad_norm": 0.08008458630230256,
+      "learning_rate": 3.889312903432396e-06,
+      "loss": 0.352,
+      "step": 2823
+    },
+    {
+      "epoch": 4.364760432766615,
+      "grad_norm": 0.07887376778827801,
+      "learning_rate": 3.870765859053669e-06,
+      "loss": 0.3506,
+      "step": 2824
+    },
+    {
+      "epoch": 4.366306027820711,
+      "grad_norm": 0.08405949873945731,
+      "learning_rate": 3.85226089451185e-06,
+      "loss": 0.3445,
+      "step": 2825
+    },
+    {
+      "epoch": 4.367851622874807,
+      "grad_norm": 0.07764345620712967,
+      "learning_rate": 3.833798031359725e-06,
+      "loss": 0.3501,
+      "step": 2826
+    },
+    {
+      "epoch": 4.369397217928903,
+      "grad_norm": 0.08110470084261369,
+      "learning_rate": 3.815377291101072e-06,
+      "loss": 0.353,
+      "step": 2827
+    },
+    {
+      "epoch": 4.370942812982999,
+      "grad_norm": 0.0837793846570553,
+      "learning_rate": 3.7969986951905725e-06,
+      "loss": 0.3516,
+      "step": 2828
+    },
+    {
+      "epoch": 4.3724884080370945,
+      "grad_norm": 0.07721899926190336,
+      "learning_rate": 3.7786622650338412e-06,
+      "loss": 0.3487,
+      "step": 2829
+    },
+    {
+      "epoch": 4.37403400309119,
+      "grad_norm": 0.08428388240342731,
+      "learning_rate": 3.7603680219873816e-06,
+      "loss": 0.3467,
+      "step": 2830
+    },
+    {
+      "epoch": 4.375579598145286,
+      "grad_norm": 0.08230952833359399,
+      "learning_rate": 3.74211598735855e-06,
+      "loss": 0.3516,
+      "step": 2831
+    },
+    {
+      "epoch": 4.377125193199381,
+      "grad_norm": 0.08388381109891997,
+      "learning_rate": 3.7239061824055456e-06,
+      "loss": 0.3575,
+      "step": 2832
+    },
+    {
+      "epoch": 4.378670788253477,
+      "grad_norm": 0.08382982365844328,
+      "learning_rate": 3.7057386283374074e-06,
+      "loss": 0.3484,
+      "step": 2833
+    },
+    {
+      "epoch": 4.380216383307573,
+      "grad_norm": 0.07995433671336721,
+      "learning_rate": 3.687613346313934e-06,
+      "loss": 0.3388,
+      "step": 2834
+    },
+    {
+      "epoch": 4.381761978361669,
+      "grad_norm": 0.07553437652817582,
+      "learning_rate": 3.6695303574457008e-06,
+      "loss": 0.3466,
+      "step": 2835
+    },
+    {
+      "epoch": 4.383307573415765,
+      "grad_norm": 0.0832900184978018,
+      "learning_rate": 3.651489682794025e-06,
+      "loss": 0.3517,
+      "step": 2836
+    },
+    {
+      "epoch": 4.384853168469861,
+      "grad_norm": 0.08166599537889928,
+      "learning_rate": 3.633491343370934e-06,
+      "loss": 0.3519,
+      "step": 2837
+    },
+    {
+      "epoch": 4.386398763523957,
+      "grad_norm": 0.08317565714983777,
+      "learning_rate": 3.615535360139162e-06,
+      "loss": 0.3503,
+      "step": 2838
+    },
+    {
+      "epoch": 4.387944358578053,
+      "grad_norm": 0.08332970138051007,
+      "learning_rate": 3.5976217540120863e-06,
+      "loss": 0.3444,
+      "step": 2839
+    },
+    {
+      "epoch": 4.3894899536321486,
+      "grad_norm": 0.07937650477051997,
+      "learning_rate": 3.5797505458537597e-06,
+      "loss": 0.3473,
+      "step": 2840
+    },
+    {
+      "epoch": 4.3910355486862445,
+      "grad_norm": 0.084930685850289,
+      "learning_rate": 3.5619217564788255e-06,
+      "loss": 0.3528,
+      "step": 2841
+    },
+    {
+      "epoch": 4.39258114374034,
+      "grad_norm": 0.08409427647011232,
+      "learning_rate": 3.54413540665254e-06,
+      "loss": 0.3444,
+      "step": 2842
+    },
+    {
+      "epoch": 4.394126738794436,
+      "grad_norm": 0.08577284327271625,
+      "learning_rate": 3.526391517090719e-06,
+      "loss": 0.3357,
+      "step": 2843
+    },
+    {
+      "epoch": 4.395672333848531,
+      "grad_norm": 0.08333543187416191,
+      "learning_rate": 3.5086901084597293e-06,
+      "loss": 0.3453,
+      "step": 2844
+    },
+    {
+      "epoch": 4.397217928902627,
+      "grad_norm": 0.07992037276703519,
+      "learning_rate": 3.491031201376451e-06,
+      "loss": 0.3454,
+      "step": 2845
+    },
+    {
+      "epoch": 4.398763523956723,
+      "grad_norm": 0.08209319555463147,
+      "learning_rate": 3.473414816408287e-06,
+      "loss": 0.3402,
+      "step": 2846
+    },
+    {
+      "epoch": 4.400309119010819,
+      "grad_norm": 0.08490154321268921,
+      "learning_rate": 3.455840974073086e-06,
+      "loss": 0.3427,
+      "step": 2847
+    },
+    {
+      "epoch": 4.401854714064915,
+      "grad_norm": 0.07698769938158478,
+      "learning_rate": 3.4383096948391593e-06,
+      "loss": 0.3441,
+      "step": 2848
+    },
+    {
+      "epoch": 4.403400309119011,
+      "grad_norm": 0.07880916167937885,
+      "learning_rate": 3.4208209991252626e-06,
+      "loss": 0.3433,
+      "step": 2849
+    },
+    {
+      "epoch": 4.404945904173107,
+      "grad_norm": 0.08174402671114551,
+      "learning_rate": 3.4033749073005205e-06,
+      "loss": 0.3469,
+      "step": 2850
+    },
+    {
+      "epoch": 4.406491499227203,
+      "grad_norm": 0.08622213318277683,
+      "learning_rate": 3.3859714396844434e-06,
+      "loss": 0.3458,
+      "step": 2851
+    },
+    {
+      "epoch": 4.4080370942812985,
+      "grad_norm": 0.0837012341239019,
+      "learning_rate": 3.3686106165469323e-06,
+      "loss": 0.3588,
+      "step": 2852
+    },
+    {
+      "epoch": 4.409582689335394,
+      "grad_norm": 0.07944231475429726,
+      "learning_rate": 3.3512924581081775e-06,
+      "loss": 0.3477,
+      "step": 2853
+    },
+    {
+      "epoch": 4.41112828438949,
+      "grad_norm": 0.08178179824143911,
+      "learning_rate": 3.3340169845386926e-06,
+      "loss": 0.356,
+      "step": 2854
+    },
+    {
+      "epoch": 4.412673879443586,
+      "grad_norm": 0.08582291436062328,
+      "learning_rate": 3.3167842159592946e-06,
+      "loss": 0.3455,
+      "step": 2855
+    },
+    {
+      "epoch": 4.414219474497681,
+      "grad_norm": 0.08762946770722586,
+      "learning_rate": 3.2995941724410384e-06,
+      "loss": 0.3526,
+      "step": 2856
+    },
+    {
+      "epoch": 4.415765069551777,
+      "grad_norm": 0.08316740590899953,
+      "learning_rate": 3.2824468740052072e-06,
+      "loss": 0.3476,
+      "step": 2857
+    },
+    {
+      "epoch": 4.417310664605873,
+      "grad_norm": 0.08262977766404131,
+      "learning_rate": 3.2653423406233364e-06,
+      "loss": 0.3479,
+      "step": 2858
+    },
+    {
+      "epoch": 4.418856259659969,
+      "grad_norm": 0.08268620916274816,
+      "learning_rate": 3.2482805922171257e-06,
+      "loss": 0.3493,
+      "step": 2859
+    },
+    {
+      "epoch": 4.420401854714065,
+      "grad_norm": 0.08850892909930494,
+      "learning_rate": 3.231261648658439e-06,
+      "loss": 0.3508,
+      "step": 2860
+    },
+    {
+      "epoch": 4.421947449768161,
+      "grad_norm": 0.08479502639571267,
+      "learning_rate": 3.2142855297693145e-06,
+      "loss": 0.3495,
+      "step": 2861
+    },
+    {
+      "epoch": 4.423493044822257,
+      "grad_norm": 0.08113628583032982,
+      "learning_rate": 3.1973522553218774e-06,
+      "loss": 0.3584,
+      "step": 2862
+    },
+    {
+      "epoch": 4.425038639876353,
+      "grad_norm": 0.08072752020645717,
+      "learning_rate": 3.180461845038374e-06,
+      "loss": 0.3488,
+      "step": 2863
+    },
+    {
+      "epoch": 4.4265842349304485,
+      "grad_norm": 0.07667822418785249,
+      "learning_rate": 3.1636143185911216e-06,
+      "loss": 0.3478,
+      "step": 2864
+    },
+    {
+      "epoch": 4.428129829984544,
+      "grad_norm": 0.07762761353502987,
+      "learning_rate": 3.1468096956024863e-06,
+      "loss": 0.3513,
+      "step": 2865
+    },
+    {
+      "epoch": 4.42967542503864,
+      "grad_norm": 0.08132451497698748,
+      "learning_rate": 3.1300479956448693e-06,
+      "loss": 0.3463,
+      "step": 2866
+    },
+    {
+      "epoch": 4.431221020092735,
+      "grad_norm": 0.08253670758213165,
+      "learning_rate": 3.1133292382406678e-06,
+      "loss": 0.3528,
+      "step": 2867
+    },
+    {
+      "epoch": 4.432766615146831,
+      "grad_norm": 0.08377799094948238,
+      "learning_rate": 3.096653442862292e-06,
+      "loss": 0.3543,
+      "step": 2868
+    },
+    {
+      "epoch": 4.434312210200927,
+      "grad_norm": 0.08446725971646331,
+      "learning_rate": 3.080020628932081e-06,
+      "loss": 0.352,
+      "step": 2869
+    },
+    {
+      "epoch": 4.435857805255023,
+      "grad_norm": 0.08136347996807272,
+      "learning_rate": 3.0634308158223303e-06,
+      "loss": 0.3443,
+      "step": 2870
+    },
+    {
+      "epoch": 4.437403400309119,
+      "grad_norm": 0.08057277380348657,
+      "learning_rate": 3.04688402285525e-06,
+      "loss": 0.3569,
+      "step": 2871
+    },
+    {
+      "epoch": 4.438948995363215,
+      "grad_norm": 0.08140459786575811,
+      "learning_rate": 3.030380269302944e-06,
+      "loss": 0.3377,
+      "step": 2872
+    },
+    {
+      "epoch": 4.440494590417311,
+      "grad_norm": 0.08261305046186905,
+      "learning_rate": 3.0139195743873826e-06,
+      "loss": 0.349,
+      "step": 2873
+    },
+    {
+      "epoch": 4.442040185471407,
+      "grad_norm": 0.08485741910587392,
+      "learning_rate": 2.9975019572804042e-06,
+      "loss": 0.3546,
+      "step": 2874
+    },
+    {
+      "epoch": 4.443585780525503,
+      "grad_norm": 0.07873815679095189,
+      "learning_rate": 2.98112743710365e-06,
+      "loss": 0.3492,
+      "step": 2875
+    },
+    {
+      "epoch": 4.4451313755795985,
+      "grad_norm": 0.07944709390269125,
+      "learning_rate": 2.9647960329285806e-06,
+      "loss": 0.3489,
+      "step": 2876
+    },
+    {
+      "epoch": 4.446676970633694,
+      "grad_norm": 0.0812646396088516,
+      "learning_rate": 2.948507763776438e-06,
+      "loss": 0.3465,
+      "step": 2877
+    },
+    {
+      "epoch": 4.448222565687789,
+      "grad_norm": 0.07652264016970232,
+      "learning_rate": 2.9322626486182158e-06,
+      "loss": 0.3574,
+      "step": 2878
+    },
+    {
+      "epoch": 4.449768160741885,
+      "grad_norm": 0.08365648675702712,
+      "learning_rate": 2.916060706374655e-06,
+      "loss": 0.3457,
+      "step": 2879
+    },
+    {
+      "epoch": 4.451313755795981,
+      "grad_norm": 0.07451328373134032,
+      "learning_rate": 2.8999019559162156e-06,
+      "loss": 0.3497,
+      "step": 2880
+    },
+    {
+      "epoch": 4.452859350850077,
+      "grad_norm": 0.07992407614754655,
+      "learning_rate": 2.8837864160630435e-06,
+      "loss": 0.3532,
+      "step": 2881
+    },
+    {
+      "epoch": 4.454404945904173,
+      "grad_norm": 0.07905390351920381,
+      "learning_rate": 2.8677141055849554e-06,
+      "loss": 0.3463,
+      "step": 2882
+    },
+    {
+      "epoch": 4.455950540958269,
+      "grad_norm": 0.07547533981426742,
+      "learning_rate": 2.851685043201431e-06,
+      "loss": 0.3494,
+      "step": 2883
+    },
+    {
+      "epoch": 4.457496136012365,
+      "grad_norm": 0.07804805683329721,
+      "learning_rate": 2.83569924758158e-06,
+      "loss": 0.3424,
+      "step": 2884
+    },
+    {
+      "epoch": 4.459041731066461,
+      "grad_norm": 0.08078161181159066,
+      "learning_rate": 2.8197567373440882e-06,
+      "loss": 0.3435,
+      "step": 2885
+    },
+    {
+      "epoch": 4.460587326120557,
+      "grad_norm": 0.08360221957028523,
+      "learning_rate": 2.803857531057266e-06,
+      "loss": 0.3538,
+      "step": 2886
+    },
+    {
+      "epoch": 4.4621329211746525,
+      "grad_norm": 0.09720199779553902,
+      "learning_rate": 2.788001647238967e-06,
+      "loss": 0.3486,
+      "step": 2887
+    },
+    {
+      "epoch": 4.4636785162287484,
+      "grad_norm": 0.07824968323511246,
+      "learning_rate": 2.7721891043565795e-06,
+      "loss": 0.3566,
+      "step": 2888
+    },
+    {
+      "epoch": 4.4652241112828435,
+      "grad_norm": 0.07520904847589219,
+      "learning_rate": 2.756419920827047e-06,
+      "loss": 0.3447,
+      "step": 2889
+    },
+    {
+      "epoch": 4.466769706336939,
+      "grad_norm": 0.07927334007453044,
+      "learning_rate": 2.740694115016771e-06,
+      "loss": 0.3518,
+      "step": 2890
+    },
+    {
+      "epoch": 4.468315301391035,
+      "grad_norm": 0.07346839525159184,
+      "learning_rate": 2.7250117052416557e-06,
+      "loss": 0.3524,
+      "step": 2891
+    },
+    {
+      "epoch": 4.469860896445131,
+      "grad_norm": 0.08347394736389456,
+      "learning_rate": 2.709372709767055e-06,
+      "loss": 0.3476,
+      "step": 2892
+    },
+    {
+      "epoch": 4.471406491499227,
+      "grad_norm": 0.07785722815714517,
+      "learning_rate": 2.693777146807759e-06,
+      "loss": 0.3524,
+      "step": 2893
+    },
+    {
+      "epoch": 4.472952086553323,
+      "grad_norm": 0.07551340491347229,
+      "learning_rate": 2.6782250345279616e-06,
+      "loss": 0.3573,
+      "step": 2894
+    },
+    {
+      "epoch": 4.474497681607419,
+      "grad_norm": 0.07976315821083518,
+      "learning_rate": 2.662716391041262e-06,
+      "loss": 0.3475,
+      "step": 2895
+    },
+    {
+      "epoch": 4.476043276661515,
+      "grad_norm": 0.07716251071719255,
+      "learning_rate": 2.6472512344106394e-06,
+      "loss": 0.3579,
+      "step": 2896
+    },
+    {
+      "epoch": 4.477588871715611,
+      "grad_norm": 0.0829097665496094,
+      "learning_rate": 2.631829582648404e-06,
+      "loss": 0.3471,
+      "step": 2897
+    },
+    {
+      "epoch": 4.479134466769707,
+      "grad_norm": 0.08103423871659719,
+      "learning_rate": 2.616451453716202e-06,
+      "loss": 0.3559,
+      "step": 2898
+    },
+    {
+      "epoch": 4.4806800618238025,
+      "grad_norm": 0.08553687067102614,
+      "learning_rate": 2.6011168655249906e-06,
+      "loss": 0.3552,
+      "step": 2899
+    },
+    {
+      "epoch": 4.4822256568778975,
+      "grad_norm": 0.08428872138223713,
+      "learning_rate": 2.5858258359350207e-06,
+      "loss": 0.3479,
+      "step": 2900
+    },
+    {
+      "epoch": 4.483771251931993,
+      "grad_norm": 0.07736643977622588,
+      "learning_rate": 2.5705783827557886e-06,
+      "loss": 0.3443,
+      "step": 2901
+    },
+    {
+      "epoch": 4.485316846986089,
+      "grad_norm": 0.07671781625871242,
+      "learning_rate": 2.555374523746075e-06,
+      "loss": 0.3378,
+      "step": 2902
+    },
+    {
+      "epoch": 4.486862442040185,
+      "grad_norm": 0.08055811720332556,
+      "learning_rate": 2.5402142766138528e-06,
+      "loss": 0.3573,
+      "step": 2903
+    },
+    {
+      "epoch": 4.488408037094281,
+      "grad_norm": 0.0799311392726142,
+      "learning_rate": 2.525097659016309e-06,
+      "loss": 0.3501,
+      "step": 2904
+    },
+    {
+      "epoch": 4.489953632148377,
+      "grad_norm": 0.07329363549243337,
+      "learning_rate": 2.510024688559822e-06,
+      "loss": 0.3442,
+      "step": 2905
+    },
+    {
+      "epoch": 4.491499227202473,
+      "grad_norm": 0.0786719091481724,
+      "learning_rate": 2.4949953827999276e-06,
+      "loss": 0.3476,
+      "step": 2906
+    },
+    {
+      "epoch": 4.493044822256569,
+      "grad_norm": 0.08454925530472679,
+      "learning_rate": 2.480009759241302e-06,
+      "loss": 0.3542,
+      "step": 2907
+    },
+    {
+      "epoch": 4.494590417310665,
+      "grad_norm": 0.08074136205065753,
+      "learning_rate": 2.4650678353377622e-06,
+      "loss": 0.3495,
+      "step": 2908
+    },
+    {
+      "epoch": 4.496136012364761,
+      "grad_norm": 0.07288391733839879,
+      "learning_rate": 2.4501696284922094e-06,
+      "loss": 0.3407,
+      "step": 2909
+    },
+    {
+      "epoch": 4.497681607418857,
+      "grad_norm": 0.0756241829800099,
+      "learning_rate": 2.4353151560566256e-06,
+      "loss": 0.3507,
+      "step": 2910
+    },
+    {
+      "epoch": 4.4992272024729525,
+      "grad_norm": 0.07722698742509931,
+      "learning_rate": 2.420504435332074e-06,
+      "loss": 0.3567,
+      "step": 2911
+    },
+    {
+      "epoch": 4.500772797527048,
+      "grad_norm": 0.08092853920519424,
+      "learning_rate": 2.4057374835686575e-06,
+      "loss": 0.3529,
+      "step": 2912
+    },
+    {
+      "epoch": 4.502318392581143,
+      "grad_norm": 0.07564872276180502,
+      "learning_rate": 2.391014317965468e-06,
+      "loss": 0.3551,
+      "step": 2913
+    },
+    {
+      "epoch": 4.503863987635239,
+      "grad_norm": 0.0747734888669493,
+      "learning_rate": 2.376334955670645e-06,
+      "loss": 0.3502,
+      "step": 2914
+    },
+    {
+      "epoch": 4.505409582689335,
+      "grad_norm": 0.07262912762423096,
+      "learning_rate": 2.361699413781282e-06,
+      "loss": 0.3466,
+      "step": 2915
+    },
+    {
+      "epoch": 4.506955177743431,
+      "grad_norm": 0.087234776281415,
+      "learning_rate": 2.3471077093434412e-06,
+      "loss": 0.3452,
+      "step": 2916
+    },
+    {
+      "epoch": 4.508500772797527,
+      "grad_norm": 0.07682097557507851,
+      "learning_rate": 2.3325598593521324e-06,
+      "loss": 0.3456,
+      "step": 2917
+    },
+    {
+      "epoch": 4.510046367851623,
+      "grad_norm": 0.07599485976755793,
+      "learning_rate": 2.3180558807512864e-06,
+      "loss": 0.3437,
+      "step": 2918
+    },
+    {
+      "epoch": 4.511591962905719,
+      "grad_norm": 0.07910867327561365,
+      "learning_rate": 2.303595790433728e-06,
+      "loss": 0.3495,
+      "step": 2919
+    },
+    {
+      "epoch": 4.513137557959815,
+      "grad_norm": 0.07673517152884492,
+      "learning_rate": 2.289179605241172e-06,
+      "loss": 0.3449,
+      "step": 2920
+    },
+    {
+      "epoch": 4.514683153013911,
+      "grad_norm": 0.08393290485261232,
+      "learning_rate": 2.2748073419642e-06,
+      "loss": 0.3519,
+      "step": 2921
+    },
+    {
+      "epoch": 4.516228748068007,
+      "grad_norm": 0.08053111068877837,
+      "learning_rate": 2.260479017342232e-06,
+      "loss": 0.3416,
+      "step": 2922
+    },
+    {
+      "epoch": 4.5177743431221025,
+      "grad_norm": 0.07617643912713123,
+      "learning_rate": 2.2461946480635087e-06,
+      "loss": 0.3552,
+      "step": 2923
+    },
+    {
+      "epoch": 4.5193199381761975,
+      "grad_norm": 0.08085550641476205,
+      "learning_rate": 2.231954250765096e-06,
+      "loss": 0.3627,
+      "step": 2924
+    },
+    {
+      "epoch": 4.520865533230293,
+      "grad_norm": 0.07409947109933004,
+      "learning_rate": 2.2177578420328195e-06,
+      "loss": 0.349,
+      "step": 2925
+    },
+    {
+      "epoch": 4.522411128284389,
+      "grad_norm": 0.0753170008154408,
+      "learning_rate": 2.203605438401288e-06,
+      "loss": 0.342,
+      "step": 2926
+    },
+    {
+      "epoch": 4.523956723338485,
+      "grad_norm": 0.07808859057256458,
+      "learning_rate": 2.1894970563538466e-06,
+      "loss": 0.3515,
+      "step": 2927
+    },
+    {
+      "epoch": 4.525502318392581,
+      "grad_norm": 0.08034407241190529,
+      "learning_rate": 2.1754327123225807e-06,
+      "loss": 0.3475,
+      "step": 2928
+    },
+    {
+      "epoch": 4.527047913446677,
+      "grad_norm": 0.07768682765516363,
+      "learning_rate": 2.161412422688267e-06,
+      "loss": 0.3479,
+      "step": 2929
+    },
+    {
+      "epoch": 4.528593508500773,
+      "grad_norm": 0.07794096057951751,
+      "learning_rate": 2.1474362037804e-06,
+      "loss": 0.3468,
+      "step": 2930
+    },
+    {
+      "epoch": 4.530139103554869,
+      "grad_norm": 0.07296325597840433,
+      "learning_rate": 2.133504071877113e-06,
+      "loss": 0.3465,
+      "step": 2931
+    },
+    {
+      "epoch": 4.531684698608965,
+      "grad_norm": 0.07538263425855553,
+      "learning_rate": 2.119616043205208e-06,
+      "loss": 0.3457,
+      "step": 2932
+    },
+    {
+      "epoch": 4.533230293663061,
+      "grad_norm": 0.08607373910052346,
+      "learning_rate": 2.105772133940125e-06,
+      "loss": 0.3536,
+      "step": 2933
+    },
+    {
+      "epoch": 4.5347758887171565,
+      "grad_norm": 0.07759588754302951,
+      "learning_rate": 2.091972360205903e-06,
+      "loss": 0.3473,
+      "step": 2934
+    },
+    {
+      "epoch": 4.5363214837712516,
+      "grad_norm": 0.07607434840567591,
+      "learning_rate": 2.078216738075178e-06,
+      "loss": 0.3453,
+      "step": 2935
+    },
+    {
+      "epoch": 4.5378670788253475,
+      "grad_norm": 0.07154002196069166,
+      "learning_rate": 2.0645052835691804e-06,
+      "loss": 0.3433,
+      "step": 2936
+    },
+    {
+      "epoch": 4.539412673879443,
+      "grad_norm": 0.07558474458105977,
+      "learning_rate": 2.050838012657681e-06,
+      "loss": 0.3453,
+      "step": 2937
+    },
+    {
+      "epoch": 4.540958268933539,
+      "grad_norm": 0.07190207377118617,
+      "learning_rate": 2.037214941258987e-06,
+      "loss": 0.3505,
+      "step": 2938
+    },
+    {
+      "epoch": 4.542503863987635,
+      "grad_norm": 0.07655013874782704,
+      "learning_rate": 2.023636085239957e-06,
+      "loss": 0.3546,
+      "step": 2939
+    },
+    {
+      "epoch": 4.544049459041731,
+      "grad_norm": 0.08127970977854401,
+      "learning_rate": 2.0101014604159053e-06,
+      "loss": 0.3489,
+      "step": 2940
+    },
+    {
+      "epoch": 4.545595054095827,
+      "grad_norm": 0.07679116403672862,
+      "learning_rate": 1.996611082550657e-06,
+      "loss": 0.3424,
+      "step": 2941
+    },
+    {
+      "epoch": 4.547140649149923,
+      "grad_norm": 0.07395008637835969,
+      "learning_rate": 1.9831649673565102e-06,
+      "loss": 0.3481,
+      "step": 2942
+    },
+    {
+      "epoch": 4.548686244204019,
+      "grad_norm": 0.12737252523291856,
+      "learning_rate": 1.9697631304941913e-06,
+      "loss": 0.357,
+      "step": 2943
+    },
+    {
+      "epoch": 4.550231839258115,
+      "grad_norm": 0.07563447201100297,
+      "learning_rate": 1.956405587572863e-06,
+      "loss": 0.3421,
+      "step": 2944
+    },
+    {
+      "epoch": 4.551777434312211,
+      "grad_norm": 0.07965408053300205,
+      "learning_rate": 1.9430923541501024e-06,
+      "loss": 0.3471,
+      "step": 2945
+    },
+    {
+      "epoch": 4.553323029366306,
+      "grad_norm": 0.07390792975317896,
+      "learning_rate": 1.92982344573188e-06,
+      "loss": 0.3479,
+      "step": 2946
+    },
+    {
+      "epoch": 4.5548686244204015,
+      "grad_norm": 0.0766260842638226,
+      "learning_rate": 1.916598877772531e-06,
+      "loss": 0.3483,
+      "step": 2947
+    },
+    {
+      "epoch": 4.556414219474497,
+      "grad_norm": 0.07754899335989722,
+      "learning_rate": 1.9034186656747478e-06,
+      "loss": 0.3486,
+      "step": 2948
+    },
+    {
+      "epoch": 4.557959814528593,
+      "grad_norm": 0.07480599837503225,
+      "learning_rate": 1.8902828247895754e-06,
+      "loss": 0.3468,
+      "step": 2949
+    },
+    {
+      "epoch": 4.559505409582689,
+      "grad_norm": 0.07774194619712663,
+      "learning_rate": 1.877191370416367e-06,
+      "loss": 0.3494,
+      "step": 2950
+    },
+    {
+      "epoch": 4.561051004636785,
+      "grad_norm": 0.0782130516852238,
+      "learning_rate": 1.8641443178027784e-06,
+      "loss": 0.3416,
+      "step": 2951
+    },
+    {
+      "epoch": 4.562596599690881,
+      "grad_norm": 0.07633523238815748,
+      "learning_rate": 1.8511416821447615e-06,
+      "loss": 0.3521,
+      "step": 2952
+    },
+    {
+      "epoch": 4.564142194744977,
+      "grad_norm": 0.07520877479556337,
+      "learning_rate": 1.838183478586526e-06,
+      "loss": 0.3526,
+      "step": 2953
+    },
+    {
+      "epoch": 4.565687789799073,
+      "grad_norm": 0.07414563679406655,
+      "learning_rate": 1.8252697222205328e-06,
+      "loss": 0.349,
+      "step": 2954
+    },
+    {
+      "epoch": 4.567233384853169,
+      "grad_norm": 0.07843850991489067,
+      "learning_rate": 1.8124004280874795e-06,
+      "loss": 0.3524,
+      "step": 2955
+    },
+    {
+      "epoch": 4.568778979907265,
+      "grad_norm": 0.0764585912548359,
+      "learning_rate": 1.7995756111762697e-06,
+      "loss": 0.3497,
+      "step": 2956
+    },
+    {
+      "epoch": 4.57032457496136,
+      "grad_norm": 0.0736996722848919,
+      "learning_rate": 1.786795286424008e-06,
+      "loss": 0.3523,
+      "step": 2957
+    },
+    {
+      "epoch": 4.571870170015456,
+      "grad_norm": 0.0742984756485489,
+      "learning_rate": 1.7740594687159872e-06,
+      "loss": 0.3439,
+      "step": 2958
+    },
+    {
+      "epoch": 4.5734157650695515,
+      "grad_norm": 0.07351571761008804,
+      "learning_rate": 1.7613681728856536e-06,
+      "loss": 0.3435,
+      "step": 2959
+    },
+    {
+      "epoch": 4.574961360123647,
+      "grad_norm": 0.07529382981935444,
+      "learning_rate": 1.7487214137146003e-06,
+      "loss": 0.347,
+      "step": 2960
+    },
+    {
+      "epoch": 4.576506955177743,
+      "grad_norm": 0.07416963134278001,
+      "learning_rate": 1.7361192059325472e-06,
+      "loss": 0.345,
+      "step": 2961
+    },
+    {
+      "epoch": 4.578052550231839,
+      "grad_norm": 0.07422932118402725,
+      "learning_rate": 1.7235615642173309e-06,
+      "loss": 0.3455,
+      "step": 2962
+    },
+    {
+      "epoch": 4.579598145285935,
+      "grad_norm": 0.07331993504490666,
+      "learning_rate": 1.7110485031948698e-06,
+      "loss": 0.3458,
+      "step": 2963
+    },
+    {
+      "epoch": 4.581143740340031,
+      "grad_norm": 0.07421717020868522,
+      "learning_rate": 1.698580037439177e-06,
+      "loss": 0.3474,
+      "step": 2964
+    },
+    {
+      "epoch": 4.582689335394127,
+      "grad_norm": 0.07909801885727724,
+      "learning_rate": 1.686156181472307e-06,
+      "loss": 0.3585,
+      "step": 2965
+    },
+    {
+      "epoch": 4.584234930448223,
+      "grad_norm": 0.07327136305060182,
+      "learning_rate": 1.6737769497643696e-06,
+      "loss": 0.3518,
+      "step": 2966
+    },
+    {
+      "epoch": 4.585780525502319,
+      "grad_norm": 0.07486603082792806,
+      "learning_rate": 1.6614423567335026e-06,
+      "loss": 0.3508,
+      "step": 2967
+    },
+    {
+      "epoch": 4.587326120556414,
+      "grad_norm": 0.07127235342834087,
+      "learning_rate": 1.6491524167458318e-06,
+      "loss": 0.3451,
+      "step": 2968
+    },
+    {
+      "epoch": 4.58887171561051,
+      "grad_norm": 0.07671073502051187,
+      "learning_rate": 1.6369071441154983e-06,
+      "loss": 0.3404,
+      "step": 2969
+    },
+    {
+      "epoch": 4.590417310664606,
+      "grad_norm": 0.07235205912469946,
+      "learning_rate": 1.6247065531046135e-06,
+      "loss": 0.3537,
+      "step": 2970
+    },
+    {
+      "epoch": 4.5919629057187015,
+      "grad_norm": 0.07405134880832523,
+      "learning_rate": 1.6125506579232418e-06,
+      "loss": 0.3473,
+      "step": 2971
+    },
+    {
+      "epoch": 4.593508500772797,
+      "grad_norm": 0.07586263584542462,
+      "learning_rate": 1.600439472729396e-06,
+      "loss": 0.3481,
+      "step": 2972
+    },
+    {
+      "epoch": 4.595054095826893,
+      "grad_norm": 0.07437360593397314,
+      "learning_rate": 1.5883730116290142e-06,
+      "loss": 0.353,
+      "step": 2973
+    },
+    {
+      "epoch": 4.596599690880989,
+      "grad_norm": 0.07336863791455472,
+      "learning_rate": 1.576351288675939e-06,
+      "loss": 0.3537,
+      "step": 2974
+    },
+    {
+      "epoch": 4.598145285935085,
+      "grad_norm": 0.07294086828924651,
+      "learning_rate": 1.5643743178719216e-06,
+      "loss": 0.348,
+      "step": 2975
+    },
+    {
+      "epoch": 4.599690880989181,
+      "grad_norm": 0.08671386820576314,
+      "learning_rate": 1.552442113166559e-06,
+      "loss": 0.3527,
+      "step": 2976
+    },
+    {
+      "epoch": 4.601236476043277,
+      "grad_norm": 0.0735449095800472,
+      "learning_rate": 1.540554688457343e-06,
+      "loss": 0.3452,
+      "step": 2977
+    },
+    {
+      "epoch": 4.602782071097373,
+      "grad_norm": 0.07394435022116418,
+      "learning_rate": 1.528712057589594e-06,
+      "loss": 0.3598,
+      "step": 2978
+    },
+    {
+      "epoch": 4.604327666151468,
+      "grad_norm": 0.07358602608758982,
+      "learning_rate": 1.5169142343564569e-06,
+      "loss": 0.3441,
+      "step": 2979
+    },
+    {
+      "epoch": 4.605873261205565,
+      "grad_norm": 0.07224280424680504,
+      "learning_rate": 1.505161232498904e-06,
+      "loss": 0.3494,
+      "step": 2980
+    },
+    {
+      "epoch": 4.60741885625966,
+      "grad_norm": 0.07744999403223182,
+      "learning_rate": 1.4934530657056922e-06,
+      "loss": 0.3557,
+      "step": 2981
+    },
+    {
+      "epoch": 4.6089644513137555,
+      "grad_norm": 0.07386541801531468,
+      "learning_rate": 1.481789747613358e-06,
+      "loss": 0.3466,
+      "step": 2982
+    },
+    {
+      "epoch": 4.6105100463678514,
+      "grad_norm": 0.07777540929230686,
+      "learning_rate": 1.4701712918062084e-06,
+      "loss": 0.3617,
+      "step": 2983
+    },
+    {
+      "epoch": 4.612055641421947,
+      "grad_norm": 0.07614144875206677,
+      "learning_rate": 1.4585977118162985e-06,
+      "loss": 0.3602,
+      "step": 2984
+    },
+    {
+      "epoch": 4.613601236476043,
+      "grad_norm": 0.07275324191348273,
+      "learning_rate": 1.4470690211234062e-06,
+      "loss": 0.3478,
+      "step": 2985
+    },
+    {
+      "epoch": 4.615146831530139,
+      "grad_norm": 0.07180343424786612,
+      "learning_rate": 1.4355852331550479e-06,
+      "loss": 0.3454,
+      "step": 2986
+    },
+    {
+      "epoch": 4.616692426584235,
+      "grad_norm": 0.07515994489808035,
+      "learning_rate": 1.4241463612864225e-06,
+      "loss": 0.3487,
+      "step": 2987
+    },
+    {
+      "epoch": 4.618238021638331,
+      "grad_norm": 0.07784492999599076,
+      "learning_rate": 1.4127524188404196e-06,
+      "loss": 0.3488,
+      "step": 2988
+    },
+    {
+      "epoch": 4.619783616692427,
+      "grad_norm": 0.07140157858302701,
+      "learning_rate": 1.4014034190876057e-06,
+      "loss": 0.3451,
+      "step": 2989
+    },
+    {
+      "epoch": 4.621329211746523,
+      "grad_norm": 2.0047265286912013,
+      "learning_rate": 1.3900993752461944e-06,
+      "loss": 0.3687,
+      "step": 2990
+    },
+    {
+      "epoch": 4.622874806800619,
+      "grad_norm": 0.07418079401508136,
+      "learning_rate": 1.3788403004820362e-06,
+      "loss": 0.353,
+      "step": 2991
+    },
+    {
+      "epoch": 4.624420401854714,
+      "grad_norm": 0.0711941253914657,
+      "learning_rate": 1.3676262079086278e-06,
+      "loss": 0.3534,
+      "step": 2992
+    },
+    {
+      "epoch": 4.62596599690881,
+      "grad_norm": 0.07313686483599284,
+      "learning_rate": 1.3564571105870505e-06,
+      "loss": 0.3492,
+      "step": 2993
+    },
+    {
+      "epoch": 4.6275115919629055,
+      "grad_norm": 0.0758597731241708,
+      "learning_rate": 1.345333021525983e-06,
+      "loss": 0.359,
+      "step": 2994
+    },
+    {
+      "epoch": 4.629057187017001,
+      "grad_norm": 0.0719067589383518,
+      "learning_rate": 1.3342539536817012e-06,
+      "loss": 0.3435,
+      "step": 2995
+    },
+    {
+      "epoch": 4.630602782071097,
+      "grad_norm": 0.07533804168035232,
+      "learning_rate": 1.323219919958021e-06,
+      "loss": 0.3449,
+      "step": 2996
+    },
+    {
+      "epoch": 4.632148377125193,
+      "grad_norm": 0.11897090431385539,
+      "learning_rate": 1.3122309332063198e-06,
+      "loss": 0.351,
+      "step": 2997
+    },
+    {
+      "epoch": 4.633693972179289,
+      "grad_norm": 0.07333757116667064,
+      "learning_rate": 1.3012870062255067e-06,
+      "loss": 0.3542,
+      "step": 2998
+    },
+    {
+      "epoch": 4.635239567233385,
+      "grad_norm": 0.0745402534552004,
+      "learning_rate": 1.290388151762012e-06,
+      "loss": 0.3538,
+      "step": 2999
+    },
+    {
+      "epoch": 4.636785162287481,
+      "grad_norm": 0.07571100275937757,
+      "learning_rate": 1.2795343825097618e-06,
+      "loss": 0.3585,
+      "step": 3000
+    },
+    {
+      "epoch": 4.638330757341577,
+      "grad_norm": 0.07176624771608246,
+      "learning_rate": 1.2687257111101859e-06,
+      "loss": 0.3499,
+      "step": 3001
+    },
+    {
+      "epoch": 4.639876352395673,
+      "grad_norm": 0.07402891349893713,
+      "learning_rate": 1.2579621501521743e-06,
+      "loss": 0.3459,
+      "step": 3002
+    },
+    {
+      "epoch": 4.641421947449768,
+      "grad_norm": 0.07579896281869608,
+      "learning_rate": 1.2472437121720727e-06,
+      "loss": 0.3528,
+      "step": 3003
+    },
+    {
+      "epoch": 4.642967542503864,
+      "grad_norm": 0.07434597609548542,
+      "learning_rate": 1.2365704096536901e-06,
+      "loss": 0.3514,
+      "step": 3004
+    },
+    {
+      "epoch": 4.64451313755796,
+      "grad_norm": 0.07328808377517211,
+      "learning_rate": 1.2259422550282562e-06,
+      "loss": 0.3402,
+      "step": 3005
+    },
+    {
+      "epoch": 4.6460587326120555,
+      "grad_norm": 0.07193528453543194,
+      "learning_rate": 1.2153592606744157e-06,
+      "loss": 0.3497,
+      "step": 3006
+    },
+    {
+      "epoch": 4.647604327666151,
+      "grad_norm": 0.0818183336209685,
+      "learning_rate": 1.2048214389182156e-06,
+      "loss": 0.3499,
+      "step": 3007
+    },
+    {
+      "epoch": 4.649149922720247,
+      "grad_norm": 0.07299007521326033,
+      "learning_rate": 1.1943288020330957e-06,
+      "loss": 0.3507,
+      "step": 3008
+    },
+    {
+      "epoch": 4.650695517774343,
+      "grad_norm": 0.07741361279679244,
+      "learning_rate": 1.183881362239858e-06,
+      "loss": 0.3461,
+      "step": 3009
+    },
+    {
+      "epoch": 4.652241112828439,
+      "grad_norm": 0.07190872237292077,
+      "learning_rate": 1.1734791317066763e-06,
+      "loss": 0.3416,
+      "step": 3010
+    },
+    {
+      "epoch": 4.653786707882535,
+      "grad_norm": 0.07112171217981218,
+      "learning_rate": 1.163122122549063e-06,
+      "loss": 0.3527,
+      "step": 3011
+    },
+    {
+      "epoch": 4.655332302936631,
+      "grad_norm": 0.06967890467673124,
+      "learning_rate": 1.1528103468298535e-06,
+      "loss": 0.3493,
+      "step": 3012
+    },
+    {
+      "epoch": 4.656877897990727,
+      "grad_norm": 0.07155519086256591,
+      "learning_rate": 1.142543816559205e-06,
+      "loss": 0.3483,
+      "step": 3013
+    },
+    {
+      "epoch": 4.658423493044822,
+      "grad_norm": 0.07391645581094995,
+      "learning_rate": 1.1323225436945885e-06,
+      "loss": 0.3535,
+      "step": 3014
+    },
+    {
+      "epoch": 4.659969088098918,
+      "grad_norm": 0.0720187041809313,
+      "learning_rate": 1.122146540140747e-06,
+      "loss": 0.3513,
+      "step": 3015
+    },
+    {
+      "epoch": 4.661514683153014,
+      "grad_norm": 0.07501960034559438,
+      "learning_rate": 1.112015817749712e-06,
+      "loss": 0.35,
+      "step": 3016
+    },
+    {
+      "epoch": 4.6630602782071096,
+      "grad_norm": 0.07153130718507449,
+      "learning_rate": 1.1019303883207598e-06,
+      "loss": 0.3513,
+      "step": 3017
+    },
+    {
+      "epoch": 4.6646058732612055,
+      "grad_norm": 0.07017952743086443,
+      "learning_rate": 1.0918902636004325e-06,
+      "loss": 0.3485,
+      "step": 3018
+    },
+    {
+      "epoch": 4.666151468315301,
+      "grad_norm": 0.07789273950609461,
+      "learning_rate": 1.0818954552824822e-06,
+      "loss": 0.3472,
+      "step": 3019
+    },
+    {
+      "epoch": 4.667697063369397,
+      "grad_norm": 0.07399218057264262,
+      "learning_rate": 1.0719459750079086e-06,
+      "loss": 0.35,
+      "step": 3020
+    },
+    {
+      "epoch": 4.669242658423493,
+      "grad_norm": 0.07127857715673765,
+      "learning_rate": 1.0620418343648998e-06,
+      "loss": 0.3464,
+      "step": 3021
+    },
+    {
+      "epoch": 4.670788253477589,
+      "grad_norm": 0.07483385161854272,
+      "learning_rate": 1.0521830448888414e-06,
+      "loss": 0.3456,
+      "step": 3022
+    },
+    {
+      "epoch": 4.672333848531685,
+      "grad_norm": 0.07204304647984132,
+      "learning_rate": 1.042369618062291e-06,
+      "loss": 0.3525,
+      "step": 3023
+    },
+    {
+      "epoch": 4.673879443585781,
+      "grad_norm": 0.07014879739671494,
+      "learning_rate": 1.03260156531499e-06,
+      "loss": 0.349,
+      "step": 3024
+    },
+    {
+      "epoch": 4.675425038639876,
+      "grad_norm": 0.07473680544412793,
+      "learning_rate": 1.022878898023807e-06,
+      "loss": 0.3618,
+      "step": 3025
+    },
+    {
+      "epoch": 4.676970633693972,
+      "grad_norm": 0.07421624584199558,
+      "learning_rate": 1.0132016275127766e-06,
+      "loss": 0.3485,
+      "step": 3026
+    },
+    {
+      "epoch": 4.678516228748068,
+      "grad_norm": 0.07477796206290063,
+      "learning_rate": 1.003569765053043e-06,
+      "loss": 0.3584,
+      "step": 3027
+    },
+    {
+      "epoch": 4.680061823802164,
+      "grad_norm": 0.07749692593469164,
+      "learning_rate": 9.939833218628637e-07,
+      "loss": 0.344,
+      "step": 3028
+    },
+    {
+      "epoch": 4.6816074188562595,
+      "grad_norm": 0.0718024898827332,
+      "learning_rate": 9.84442309107614e-07,
+      "loss": 0.3483,
+      "step": 3029
+    },
+    {
+      "epoch": 4.683153013910355,
+      "grad_norm": 0.0706931768229241,
+      "learning_rate": 9.74946737899729e-07,
+      "loss": 0.3463,
+      "step": 3030
+    },
+    {
+      "epoch": 4.684698608964451,
+      "grad_norm": 0.07347425880468403,
+      "learning_rate": 9.654966192987403e-07,
+      "loss": 0.3515,
+      "step": 3031
+    },
+    {
+      "epoch": 4.686244204018547,
+      "grad_norm": 0.07435454364586866,
+      "learning_rate": 9.560919643112165e-07,
+      "loss": 0.3462,
+      "step": 3032
+    },
+    {
+      "epoch": 4.687789799072643,
+      "grad_norm": 0.07490827768341446,
+      "learning_rate": 9.467327838908092e-07,
+      "loss": 0.3434,
+      "step": 3033
+    },
+    {
+      "epoch": 4.689335394126739,
+      "grad_norm": 0.0724556841352236,
+      "learning_rate": 9.374190889381762e-07,
+      "loss": 0.3571,
+      "step": 3034
+    },
+    {
+      "epoch": 4.690880989180835,
+      "grad_norm": 0.07406491541893734,
+      "learning_rate": 9.28150890301005e-07,
+      "loss": 0.3582,
+      "step": 3035
+    },
+    {
+      "epoch": 4.69242658423493,
+      "grad_norm": 0.07374468202194766,
+      "learning_rate": 9.189281987740073e-07,
+      "loss": 0.3525,
+      "step": 3036
+    },
+    {
+      "epoch": 4.693972179289027,
+      "grad_norm": 0.06877988053161671,
+      "learning_rate": 9.097510250988795e-07,
+      "loss": 0.3563,
+      "step": 3037
+    },
+    {
+      "epoch": 4.695517774343122,
+      "grad_norm": 0.07102803965671498,
+      "learning_rate": 9.006193799642982e-07,
+      "loss": 0.3479,
+      "step": 3038
+    },
+    {
+      "epoch": 4.697063369397218,
+      "grad_norm": 0.07113184558693575,
+      "learning_rate": 8.915332740059291e-07,
+      "loss": 0.351,
+      "step": 3039
+    },
+    {
+      "epoch": 4.698608964451314,
+      "grad_norm": 0.0718632619148234,
+      "learning_rate": 8.824927178063824e-07,
+      "loss": 0.3499,
+      "step": 3040
+    },
+    {
+      "epoch": 4.7001545595054095,
+      "grad_norm": 0.07329925111631844,
+      "learning_rate": 8.734977218952223e-07,
+      "loss": 0.359,
+      "step": 3041
+    },
+    {
+      "epoch": 4.701700154559505,
+      "grad_norm": 0.07547314176689787,
+      "learning_rate": 8.645482967489571e-07,
+      "loss": 0.3515,
+      "step": 3042
+    },
+    {
+      "epoch": 4.703245749613601,
+      "grad_norm": 0.07306475815997367,
+      "learning_rate": 8.556444527910135e-07,
+      "loss": 0.3501,
+      "step": 3043
+    },
+    {
+      "epoch": 4.704791344667697,
+      "grad_norm": 0.07526389425229102,
+      "learning_rate": 8.467862003917182e-07,
+      "loss": 0.3526,
+      "step": 3044
+    },
+    {
+      "epoch": 4.706336939721793,
+      "grad_norm": 0.07258969878008852,
+      "learning_rate": 8.379735498683117e-07,
+      "loss": 0.3429,
+      "step": 3045
+    },
+    {
+      "epoch": 4.707882534775889,
+      "grad_norm": 0.07342674835803169,
+      "learning_rate": 8.292065114849168e-07,
+      "loss": 0.3458,
+      "step": 3046
+    },
+    {
+      "epoch": 4.709428129829985,
+      "grad_norm": 0.07146671285867212,
+      "learning_rate": 8.204850954525256e-07,
+      "loss": 0.3502,
+      "step": 3047
+    },
+    {
+      "epoch": 4.710973724884081,
+      "grad_norm": 0.07026909475371605,
+      "learning_rate": 8.118093119290127e-07,
+      "loss": 0.3413,
+      "step": 3048
+    },
+    {
+      "epoch": 4.712519319938176,
+      "grad_norm": 0.07409553775894827,
+      "learning_rate": 8.031791710190817e-07,
+      "loss": 0.3557,
+      "step": 3049
+    },
+    {
+      "epoch": 4.714064914992272,
+      "grad_norm": 0.07232324119465391,
+      "learning_rate": 7.945946827742878e-07,
+      "loss": 0.3474,
+      "step": 3050
+    },
+    {
+      "epoch": 4.715610510046368,
+      "grad_norm": 0.07059867978942962,
+      "learning_rate": 7.86055857193011e-07,
+      "loss": 0.35,
+      "step": 3051
+    },
+    {
+      "epoch": 4.717156105100464,
+      "grad_norm": 0.0749773680999401,
+      "learning_rate": 7.775627042204559e-07,
+      "loss": 0.3506,
+      "step": 3052
+    },
+    {
+      "epoch": 4.7187017001545595,
+      "grad_norm": 0.07290342990514166,
+      "learning_rate": 7.691152337486118e-07,
+      "loss": 0.3443,
+      "step": 3053
+    },
+    {
+      "epoch": 4.720247295208655,
+      "grad_norm": 0.0752869756287177,
+      "learning_rate": 7.607134556162932e-07,
+      "loss": 0.3527,
+      "step": 3054
+    },
+    {
+      "epoch": 4.721792890262751,
+      "grad_norm": 0.07255599822428015,
+      "learning_rate": 7.523573796090678e-07,
+      "loss": 0.3488,
+      "step": 3055
+    },
+    {
+      "epoch": 4.723338485316847,
+      "grad_norm": 0.07165098224485363,
+      "learning_rate": 7.440470154592793e-07,
+      "loss": 0.3407,
+      "step": 3056
+    },
+    {
+      "epoch": 4.724884080370943,
+      "grad_norm": 0.0690313954012584,
+      "learning_rate": 7.357823728460478e-07,
+      "loss": 0.3533,
+      "step": 3057
+    },
+    {
+      "epoch": 4.726429675425039,
+      "grad_norm": 0.07346441068826698,
+      "learning_rate": 7.275634613952287e-07,
+      "loss": 0.3456,
+      "step": 3058
+    },
+    {
+      "epoch": 4.727975270479135,
+      "grad_norm": 0.06992603922784864,
+      "learning_rate": 7.193902906794092e-07,
+      "loss": 0.3544,
+      "step": 3059
+    },
+    {
+      "epoch": 4.72952086553323,
+      "grad_norm": 0.07167968338403391,
+      "learning_rate": 7.112628702179036e-07,
+      "loss": 0.348,
+      "step": 3060
+    },
+    {
+      "epoch": 4.731066460587326,
+      "grad_norm": 0.07244329403694023,
+      "learning_rate": 7.031812094767576e-07,
+      "loss": 0.348,
+      "step": 3061
+    },
+    {
+      "epoch": 4.732612055641422,
+      "grad_norm": 0.0695438720429314,
+      "learning_rate": 6.951453178686995e-07,
+      "loss": 0.3497,
+      "step": 3062
+    },
+    {
+      "epoch": 4.734157650695518,
+      "grad_norm": 0.06905910992436456,
+      "learning_rate": 6.871552047531537e-07,
+      "loss": 0.3374,
+      "step": 3063
+    },
+    {
+      "epoch": 4.7357032457496135,
+      "grad_norm": 0.07038312733282957,
+      "learning_rate": 6.792108794362451e-07,
+      "loss": 0.3539,
+      "step": 3064
+    },
+    {
+      "epoch": 4.7372488408037094,
+      "grad_norm": 0.07111918247888585,
+      "learning_rate": 6.713123511707498e-07,
+      "loss": 0.3443,
+      "step": 3065
+    },
+    {
+      "epoch": 4.738794435857805,
+      "grad_norm": 0.07029419548120323,
+      "learning_rate": 6.634596291561002e-07,
+      "loss": 0.3461,
+      "step": 3066
+    },
+    {
+      "epoch": 4.740340030911901,
+      "grad_norm": 0.07236006854199122,
+      "learning_rate": 6.556527225383979e-07,
+      "loss": 0.342,
+      "step": 3067
+    },
+    {
+      "epoch": 4.741885625965997,
+      "grad_norm": 0.07028641780741396,
+      "learning_rate": 6.478916404103697e-07,
+      "loss": 0.3466,
+      "step": 3068
+    },
+    {
+      "epoch": 4.743431221020093,
+      "grad_norm": 0.06913665604329292,
+      "learning_rate": 6.401763918113712e-07,
+      "loss": 0.3472,
+      "step": 3069
+    },
+    {
+      "epoch": 4.744976816074189,
+      "grad_norm": 0.07503151033597957,
+      "learning_rate": 6.325069857273791e-07,
+      "loss": 0.3533,
+      "step": 3070
+    },
+    {
+      "epoch": 4.746522411128284,
+      "grad_norm": 0.07241753183805473,
+      "learning_rate": 6.248834310909768e-07,
+      "loss": 0.3501,
+      "step": 3071
+    },
+    {
+      "epoch": 4.74806800618238,
+      "grad_norm": 0.07151901626987163,
+      "learning_rate": 6.173057367813418e-07,
+      "loss": 0.3555,
+      "step": 3072
+    },
+    {
+      "epoch": 4.749613601236476,
+      "grad_norm": 0.0724383131896621,
+      "learning_rate": 6.097739116242407e-07,
+      "loss": 0.3548,
+      "step": 3073
+    },
+    {
+      "epoch": 4.751159196290572,
+      "grad_norm": 0.07344169003089844,
+      "learning_rate": 6.022879643920121e-07,
+      "loss": 0.3525,
+      "step": 3074
+    },
+    {
+      "epoch": 4.752704791344668,
+      "grad_norm": 0.07373091575353383,
+      "learning_rate": 5.948479038035615e-07,
+      "loss": 0.3482,
+      "step": 3075
+    },
+    {
+      "epoch": 4.7542503863987635,
+      "grad_norm": 0.07454194242432469,
+      "learning_rate": 5.874537385243617e-07,
+      "loss": 0.3453,
+      "step": 3076
+    },
+    {
+      "epoch": 4.755795981452859,
+      "grad_norm": 0.06939072118384791,
+      "learning_rate": 5.801054771664128e-07,
+      "loss": 0.3447,
+      "step": 3077
+    },
+    {
+      "epoch": 4.757341576506955,
+      "grad_norm": 0.07059689129139902,
+      "learning_rate": 5.728031282882596e-07,
+      "loss": 0.344,
+      "step": 3078
+    },
+    {
+      "epoch": 4.758887171561051,
+      "grad_norm": 0.06795840612265545,
+      "learning_rate": 5.655467003949743e-07,
+      "loss": 0.3411,
+      "step": 3079
+    },
+    {
+      "epoch": 4.760432766615147,
+      "grad_norm": 0.06866144946942429,
+      "learning_rate": 5.583362019381389e-07,
+      "loss": 0.3488,
+      "step": 3080
+    },
+    {
+      "epoch": 4.761978361669243,
+      "grad_norm": 0.07448348792534418,
+      "learning_rate": 5.511716413158441e-07,
+      "loss": 0.3507,
+      "step": 3081
+    },
+    {
+      "epoch": 4.763523956723338,
+      "grad_norm": 0.07097827849903926,
+      "learning_rate": 5.440530268726863e-07,
+      "loss": 0.3545,
+      "step": 3082
+    },
+    {
+      "epoch": 4.765069551777434,
+      "grad_norm": 0.07223978699338694,
+      "learning_rate": 5.369803668997265e-07,
+      "loss": 0.3526,
+      "step": 3083
+    },
+    {
+      "epoch": 4.76661514683153,
+      "grad_norm": 0.07319239486591154,
+      "learning_rate": 5.29953669634522e-07,
+      "loss": 0.3395,
+      "step": 3084
+    },
+    {
+      "epoch": 4.768160741885626,
+      "grad_norm": 0.07189051429090473,
+      "learning_rate": 5.229729432610819e-07,
+      "loss": 0.3452,
+      "step": 3085
+    },
+    {
+      "epoch": 4.769706336939722,
+      "grad_norm": 0.06897687535342906,
+      "learning_rate": 5.160381959098981e-07,
+      "loss": 0.3446,
+      "step": 3086
+    },
+    {
+      "epoch": 4.771251931993818,
+      "grad_norm": 0.07503284062047047,
+      "learning_rate": 5.091494356578741e-07,
+      "loss": 0.3567,
+      "step": 3087
+    },
+    {
+      "epoch": 4.7727975270479135,
+      "grad_norm": 0.07270867686038726,
+      "learning_rate": 5.02306670528383e-07,
+      "loss": 0.3618,
+      "step": 3088
+    },
+    {
+      "epoch": 4.774343122102009,
+      "grad_norm": 0.0697258178900146,
+      "learning_rate": 4.955099084912097e-07,
+      "loss": 0.3509,
+      "step": 3089
+    },
+    {
+      "epoch": 4.775888717156105,
+      "grad_norm": 0.07336882901055865,
+      "learning_rate": 4.887591574625727e-07,
+      "loss": 0.3542,
+      "step": 3090
+    },
+    {
+      "epoch": 4.777434312210201,
+      "grad_norm": 0.06974428211111179,
+      "learning_rate": 4.820544253050851e-07,
+      "loss": 0.3478,
+      "step": 3091
+    },
+    {
+      "epoch": 4.778979907264297,
+      "grad_norm": 0.07163824194929422,
+      "learning_rate": 4.753957198277759e-07,
+      "loss": 0.3421,
+      "step": 3092
+    },
+    {
+      "epoch": 4.780525502318392,
+      "grad_norm": 0.07001588261468611,
+      "learning_rate": 4.687830487860634e-07,
+      "loss": 0.3476,
+      "step": 3093
+    },
+    {
+      "epoch": 4.782071097372488,
+      "grad_norm": 0.07257955809332253,
+      "learning_rate": 4.622164198817336e-07,
+      "loss": 0.3518,
+      "step": 3094
+    },
+    {
+      "epoch": 4.783616692426584,
+      "grad_norm": 0.07419073933373933,
+      "learning_rate": 4.556958407629797e-07,
+      "loss": 0.3516,
+      "step": 3095
+    },
+    {
+      "epoch": 4.78516228748068,
+      "grad_norm": 0.07323695296781084,
+      "learning_rate": 4.4922131902432664e-07,
+      "loss": 0.3482,
+      "step": 3096
+    },
+    {
+      "epoch": 4.786707882534776,
+      "grad_norm": 0.0703451573368961,
+      "learning_rate": 4.4279286220667574e-07,
+      "loss": 0.3475,
+      "step": 3097
+    },
+    {
+      "epoch": 4.788253477588872,
+      "grad_norm": 0.07283872288973346,
+      "learning_rate": 4.3641047779727777e-07,
+      "loss": 0.3593,
+      "step": 3098
+    },
+    {
+      "epoch": 4.789799072642968,
+      "grad_norm": 0.07067693341439762,
+      "learning_rate": 4.300741732297065e-07,
+      "loss": 0.345,
+      "step": 3099
+    },
+    {
+      "epoch": 4.7913446676970635,
+      "grad_norm": 0.06936899504891707,
+      "learning_rate": 4.2378395588388523e-07,
+      "loss": 0.3543,
+      "step": 3100
+    },
+    {
+      "epoch": 4.792890262751159,
+      "grad_norm": 0.06955688998075936,
+      "learning_rate": 4.175398330860425e-07,
+      "loss": 0.3473,
+      "step": 3101
+    },
+    {
+      "epoch": 4.794435857805255,
+      "grad_norm": 0.07243694849945517,
+      "learning_rate": 4.113418121087387e-07,
+      "loss": 0.3412,
+      "step": 3102
+    },
+    {
+      "epoch": 4.795981452859351,
+      "grad_norm": 0.07157169025779929,
+      "learning_rate": 4.0518990017081706e-07,
+      "loss": 0.3448,
+      "step": 3103
+    },
+    {
+      "epoch": 4.797527047913446,
+      "grad_norm": 0.07018241550207491,
+      "learning_rate": 3.9908410443743493e-07,
+      "loss": 0.3405,
+      "step": 3104
+    },
+    {
+      "epoch": 4.799072642967543,
+      "grad_norm": 0.07326576573687611,
+      "learning_rate": 3.9302443202003714e-07,
+      "loss": 0.3535,
+      "step": 3105
+    },
+    {
+      "epoch": 4.800618238021638,
+      "grad_norm": 0.0707235481534224,
+      "learning_rate": 3.8701088997633805e-07,
+      "loss": 0.3516,
+      "step": 3106
+    },
+    {
+      "epoch": 4.802163833075734,
+      "grad_norm": 0.07237622064497551,
+      "learning_rate": 3.810434853103351e-07,
+      "loss": 0.3479,
+      "step": 3107
+    },
+    {
+      "epoch": 4.80370942812983,
+      "grad_norm": 0.06873407666523473,
+      "learning_rate": 3.7512222497227747e-07,
+      "loss": 0.3555,
+      "step": 3108
+    },
+    {
+      "epoch": 4.805255023183926,
+      "grad_norm": 1.2980980480943682,
+      "learning_rate": 3.692471158586752e-07,
+      "loss": 0.3572,
+      "step": 3109
+    },
+    {
+      "epoch": 4.806800618238022,
+      "grad_norm": 0.07110026591967687,
+      "learning_rate": 3.6341816481229477e-07,
+      "loss": 0.3569,
+      "step": 3110
+    },
+    {
+      "epoch": 4.8083462132921175,
+      "grad_norm": 0.07272273956076963,
+      "learning_rate": 3.57635378622132e-07,
+      "loss": 0.3452,
+      "step": 3111
+    },
+    {
+      "epoch": 4.809891808346213,
+      "grad_norm": 0.07070547568026381,
+      "learning_rate": 3.5189876402341727e-07,
+      "loss": 0.34,
+      "step": 3112
+    },
+    {
+      "epoch": 4.811437403400309,
+      "grad_norm": 0.06861689944938604,
+      "learning_rate": 3.462083276976014e-07,
+      "loss": 0.3391,
+      "step": 3113
+    },
+    {
+      "epoch": 4.812982998454405,
+      "grad_norm": 0.07035516615838779,
+      "learning_rate": 3.4056407627235613e-07,
+      "loss": 0.3552,
+      "step": 3114
+    },
+    {
+      "epoch": 4.814528593508501,
+      "grad_norm": 0.07093647662836161,
+      "learning_rate": 3.3496601632155625e-07,
+      "loss": 0.3496,
+      "step": 3115
+    },
+    {
+      "epoch": 4.816074188562597,
+      "grad_norm": 0.07155958462422801,
+      "learning_rate": 3.2941415436528403e-07,
+      "loss": 0.3557,
+      "step": 3116
+    },
+    {
+      "epoch": 4.817619783616692,
+      "grad_norm": 0.07242485193251162,
+      "learning_rate": 3.239084968698114e-07,
+      "loss": 0.3461,
+      "step": 3117
+    },
+    {
+      "epoch": 4.819165378670788,
+      "grad_norm": 0.07226733900010342,
+      "learning_rate": 3.184490502475912e-07,
+      "loss": 0.3526,
+      "step": 3118
+    },
+    {
+      "epoch": 4.820710973724884,
+      "grad_norm": 0.07125796713408636,
+      "learning_rate": 3.1303582085726593e-07,
+      "loss": 0.3457,
+      "step": 3119
+    },
+    {
+      "epoch": 4.82225656877898,
+      "grad_norm": 0.0725996246570851,
+      "learning_rate": 3.076688150036367e-07,
+      "loss": 0.3488,
+      "step": 3120
+    },
+    {
+      "epoch": 4.823802163833076,
+      "grad_norm": 0.07052611899488413,
+      "learning_rate": 3.023480389376765e-07,
+      "loss": 0.346,
+      "step": 3121
+    },
+    {
+      "epoch": 4.825347758887172,
+      "grad_norm": 0.06981281270196775,
+      "learning_rate": 2.9707349885649937e-07,
+      "loss": 0.3412,
+      "step": 3122
+    },
+    {
+      "epoch": 4.8268933539412675,
+      "grad_norm": 0.069439337031512,
+      "learning_rate": 2.9184520090339565e-07,
+      "loss": 0.3567,
+      "step": 3123
+    },
+    {
+      "epoch": 4.828438948995363,
+      "grad_norm": 0.06951443984608201,
+      "learning_rate": 2.8666315116776975e-07,
+      "loss": 0.3479,
+      "step": 3124
+    },
+    {
+      "epoch": 4.829984544049459,
+      "grad_norm": 0.06735297642300035,
+      "learning_rate": 2.8152735568517606e-07,
+      "loss": 0.3474,
+      "step": 3125
+    },
+    {
+      "epoch": 4.831530139103555,
+      "grad_norm": 0.06971082920706823,
+      "learning_rate": 2.7643782043729197e-07,
+      "loss": 0.346,
+      "step": 3126
+    },
+    {
+      "epoch": 4.833075734157651,
+      "grad_norm": 0.08271390144614732,
+      "learning_rate": 2.713945513519134e-07,
+      "loss": 0.3502,
+      "step": 3127
+    },
+    {
+      "epoch": 4.834621329211746,
+      "grad_norm": 0.06943164183825769,
+      "learning_rate": 2.6639755430295957e-07,
+      "loss": 0.3395,
+      "step": 3128
+    },
+    {
+      "epoch": 4.836166924265842,
+      "grad_norm": 0.06967064529885418,
+      "learning_rate": 2.614468351104371e-07,
+      "loss": 0.3435,
+      "step": 3129
+    },
+    {
+      "epoch": 4.837712519319938,
+      "grad_norm": 0.06966768895389379,
+      "learning_rate": 2.565423995404759e-07,
+      "loss": 0.3436,
+      "step": 3130
+    },
+    {
+      "epoch": 4.839258114374034,
+      "grad_norm": 0.06997659712202021,
+      "learning_rate": 2.5168425330528434e-07,
+      "loss": 0.3374,
+      "step": 3131
+    },
+    {
+      "epoch": 4.84080370942813,
+      "grad_norm": 0.07170201260888213,
+      "learning_rate": 2.4687240206315855e-07,
+      "loss": 0.3523,
+      "step": 3132
+    },
+    {
+      "epoch": 4.842349304482226,
+      "grad_norm": 0.07350875287591517,
+      "learning_rate": 2.4210685141848213e-07,
+      "loss": 0.3526,
+      "step": 3133
+    },
+    {
+      "epoch": 4.843894899536322,
+      "grad_norm": 0.07151950805751307,
+      "learning_rate": 2.3738760692170404e-07,
+      "loss": 0.3509,
+      "step": 3134
+    },
+    {
+      "epoch": 4.8454404945904175,
+      "grad_norm": 0.06977292893693249,
+      "learning_rate": 2.327146740693431e-07,
+      "loss": 0.3533,
+      "step": 3135
+    },
+    {
+      "epoch": 4.846986089644513,
+      "grad_norm": 0.07230874587314665,
+      "learning_rate": 2.280880583039835e-07,
+      "loss": 0.3491,
+      "step": 3136
+    },
+    {
+      "epoch": 4.848531684698609,
+      "grad_norm": 0.07114464201845196,
+      "learning_rate": 2.2350776501425253e-07,
+      "loss": 0.345,
+      "step": 3137
+    },
+    {
+      "epoch": 4.850077279752705,
+      "grad_norm": 0.06958736284066182,
+      "learning_rate": 2.1897379953483844e-07,
+      "loss": 0.3485,
+      "step": 3138
+    },
+    {
+      "epoch": 4.8516228748068,
+      "grad_norm": 0.07376727963501929,
+      "learning_rate": 2.144861671464682e-07,
+      "loss": 0.3479,
+      "step": 3139
+    },
+    {
+      "epoch": 4.853168469860896,
+      "grad_norm": 0.07167646732802174,
+      "learning_rate": 2.1004487307589861e-07,
+      "loss": 0.3459,
+      "step": 3140
+    },
+    {
+      "epoch": 4.854714064914992,
+      "grad_norm": 0.0678011743605068,
+      "learning_rate": 2.056499224959163e-07,
+      "loss": 0.3486,
+      "step": 3141
+    },
+    {
+      "epoch": 4.856259659969088,
+      "grad_norm": 0.07083801332503054,
+      "learning_rate": 2.013013205253378e-07,
+      "loss": 0.3544,
+      "step": 3142
+    },
+    {
+      "epoch": 4.857805255023184,
+      "grad_norm": 0.07296463515942934,
+      "learning_rate": 1.9699907222899162e-07,
+      "loss": 0.345,
+      "step": 3143
+    },
+    {
+      "epoch": 4.85935085007728,
+      "grad_norm": 0.07337285719937855,
+      "learning_rate": 1.9274318261772286e-07,
+      "loss": 0.3465,
+      "step": 3144
+    },
+    {
+      "epoch": 4.860896445131376,
+      "grad_norm": 0.06945046410104723,
+      "learning_rate": 1.8853365664837975e-07,
+      "loss": 0.3545,
+      "step": 3145
+    },
+    {
+      "epoch": 4.8624420401854715,
+      "grad_norm": 0.07003201459282013,
+      "learning_rate": 1.8437049922381378e-07,
+      "loss": 0.3463,
+      "step": 3146
+    },
+    {
+      "epoch": 4.8639876352395675,
+      "grad_norm": 0.07152799831202193,
+      "learning_rate": 1.8025371519286182e-07,
+      "loss": 0.3509,
+      "step": 3147
+    },
+    {
+      "epoch": 4.865533230293663,
+      "grad_norm": 0.06784201123029486,
+      "learning_rate": 1.7618330935035955e-07,
+      "loss": 0.3495,
+      "step": 3148
+    },
+    {
+      "epoch": 4.867078825347759,
+      "grad_norm": 0.06876345540883695,
+      "learning_rate": 1.7215928643712353e-07,
+      "loss": 0.3522,
+      "step": 3149
+    },
+    {
+      "epoch": 4.868624420401854,
+      "grad_norm": 0.06917992969160941,
+      "learning_rate": 1.6818165113993812e-07,
+      "loss": 0.344,
+      "step": 3150
+    },
+    {
+      "epoch": 4.87017001545595,
+      "grad_norm": 0.07128084847857968,
+      "learning_rate": 1.6425040809157745e-07,
+      "loss": 0.3445,
+      "step": 3151
+    },
+    {
+      "epoch": 4.871715610510046,
+      "grad_norm": 0.06912986125990882,
+      "learning_rate": 1.6036556187077e-07,
+      "loss": 0.3585,
+      "step": 3152
+    },
+    {
+      "epoch": 4.873261205564142,
+      "grad_norm": 0.0697507160019747,
+      "learning_rate": 1.5652711700220756e-07,
+      "loss": 0.3442,
+      "step": 3153
+    },
+    {
+      "epoch": 4.874806800618238,
+      "grad_norm": 0.07002166993275444,
+      "learning_rate": 1.5273507795654063e-07,
+      "loss": 0.3582,
+      "step": 3154
+    },
+    {
+      "epoch": 4.876352395672334,
+      "grad_norm": 0.07164887990027259,
+      "learning_rate": 1.489894491503696e-07,
+      "loss": 0.3534,
+      "step": 3155
+    },
+    {
+      "epoch": 4.87789799072643,
+      "grad_norm": 0.07023687429723947,
+      "learning_rate": 1.4529023494623596e-07,
+      "loss": 0.3444,
+      "step": 3156
+    },
+    {
+      "epoch": 4.879443585780526,
+      "grad_norm": 0.07036932592856734,
+      "learning_rate": 1.4163743965263544e-07,
+      "loss": 0.3313,
+      "step": 3157
+    },
+    {
+      "epoch": 4.8809891808346215,
+      "grad_norm": 0.07116839284433644,
+      "learning_rate": 1.380310675239871e-07,
+      "loss": 0.3461,
+      "step": 3158
+    },
+    {
+      "epoch": 4.882534775888717,
+      "grad_norm": 0.06867559877545402,
+      "learning_rate": 1.3447112276063768e-07,
+      "loss": 0.3429,
+      "step": 3159
+    },
+    {
+      "epoch": 4.884080370942813,
+      "grad_norm": 0.07080825908539166,
+      "learning_rate": 1.3095760950887493e-07,
+      "loss": 0.3601,
+      "step": 3160
+    },
+    {
+      "epoch": 4.885625965996908,
+      "grad_norm": 0.0672162448443747,
+      "learning_rate": 1.2749053186089655e-07,
+      "loss": 0.3477,
+      "step": 3161
+    },
+    {
+      "epoch": 4.887171561051004,
+      "grad_norm": 0.06969539242372545,
+      "learning_rate": 1.2406989385481904e-07,
+      "loss": 0.3523,
+      "step": 3162
+    },
+    {
+      "epoch": 4.8887171561051,
+      "grad_norm": 0.07364444652155135,
+      "learning_rate": 1.2069569947466885e-07,
+      "loss": 0.3497,
+      "step": 3163
+    },
+    {
+      "epoch": 4.890262751159196,
+      "grad_norm": 0.06959759375214015,
+      "learning_rate": 1.1736795265038237e-07,
+      "loss": 0.3504,
+      "step": 3164
+    },
+    {
+      "epoch": 4.891808346213292,
+      "grad_norm": 0.07028677984651312,
+      "learning_rate": 1.14086657257797e-07,
+      "loss": 0.3473,
+      "step": 3165
+    },
+    {
+      "epoch": 4.893353941267388,
+      "grad_norm": 0.0720746843915264,
+      "learning_rate": 1.1085181711865123e-07,
+      "loss": 0.3506,
+      "step": 3166
+    },
+    {
+      "epoch": 4.894899536321484,
+      "grad_norm": 0.06850477026674287,
+      "learning_rate": 1.0766343600057128e-07,
+      "loss": 0.3497,
+      "step": 3167
+    },
+    {
+      "epoch": 4.89644513137558,
+      "grad_norm": 0.07195527681480882,
+      "learning_rate": 1.0452151761707552e-07,
+      "loss": 0.3477,
+      "step": 3168
+    },
+    {
+      "epoch": 4.897990726429676,
+      "grad_norm": 0.0727530453591956,
+      "learning_rate": 1.0142606562756119e-07,
+      "loss": 0.3476,
+      "step": 3169
+    },
+    {
+      "epoch": 4.8995363214837715,
+      "grad_norm": 0.0693586325825528,
+      "learning_rate": 9.837708363732213e-08,
+      "loss": 0.3578,
+      "step": 3170
+    },
+    {
+      "epoch": 4.901081916537867,
+      "grad_norm": 0.07160392575837254,
+      "learning_rate": 9.537457519750881e-08,
+      "loss": 0.3446,
+      "step": 3171
+    },
+    {
+      "epoch": 4.902627511591962,
+      "grad_norm": 0.07199585832686192,
+      "learning_rate": 9.241854380515946e-08,
+      "loss": 0.3446,
+      "step": 3172
+    },
+    {
+      "epoch": 4.904173106646059,
+      "grad_norm": 0.07078789282421856,
+      "learning_rate": 8.95089929031645e-08,
+      "loss": 0.35,
+      "step": 3173
+    },
+    {
+      "epoch": 4.905718701700154,
+      "grad_norm": 0.06870033302291263,
+      "learning_rate": 8.664592588029764e-08,
+      "loss": 0.3422,
+      "step": 3174
+    },
+    {
+      "epoch": 4.90726429675425,
+      "grad_norm": 0.07181153777857369,
+      "learning_rate": 8.382934607117144e-08,
+      "loss": 0.3469,
+      "step": 3175
+    },
+    {
+      "epoch": 4.908809891808346,
+      "grad_norm": 0.06931529355649897,
+      "learning_rate": 8.105925675627735e-08,
+      "loss": 0.3449,
+      "step": 3176
+    },
+    {
+      "epoch": 4.910355486862442,
+      "grad_norm": 0.06889547051762213,
+      "learning_rate": 7.83356611619368e-08,
+      "loss": 0.355,
+      "step": 3177
+    },
+    {
+      "epoch": 4.911901081916538,
+      "grad_norm": 0.06997885222535463,
+      "learning_rate": 7.565856246033676e-08,
+      "loss": 0.349,
+      "step": 3178
+    },
+    {
+      "epoch": 4.913446676970634,
+      "grad_norm": 0.07031322677659328,
+      "learning_rate": 7.302796376950305e-08,
+      "loss": 0.3499,
+      "step": 3179
+    },
+    {
+      "epoch": 4.91499227202473,
+      "grad_norm": 0.07026799992091529,
+      "learning_rate": 7.044386815330039e-08,
+      "loss": 0.3482,
+      "step": 3180
+    },
+    {
+      "epoch": 4.916537867078826,
+      "grad_norm": 0.07275397852011724,
+      "learning_rate": 6.79062786214324e-08,
+      "loss": 0.3436,
+      "step": 3181
+    },
+    {
+      "epoch": 4.9180834621329215,
+      "grad_norm": 0.07110659178980795,
+      "learning_rate": 6.541519812943709e-08,
+      "loss": 0.3517,
+      "step": 3182
+    },
+    {
+      "epoch": 4.919629057187017,
+      "grad_norm": 0.06934394961030939,
+      "learning_rate": 6.297062957869138e-08,
+      "loss": 0.3549,
+      "step": 3183
+    },
+    {
+      "epoch": 4.921174652241113,
+      "grad_norm": 0.06846822812054754,
+      "learning_rate": 6.057257581638443e-08,
+      "loss": 0.335,
+      "step": 3184
+    },
+    {
+      "epoch": 4.922720247295208,
+      "grad_norm": 0.06636240898387039,
+      "learning_rate": 5.8221039635539823e-08,
+      "loss": 0.3482,
+      "step": 3185
+    },
+    {
+      "epoch": 4.924265842349304,
+      "grad_norm": 0.06912364928780351,
+      "learning_rate": 5.591602377499783e-08,
+      "loss": 0.3484,
+      "step": 3186
+    },
+    {
+      "epoch": 4.9258114374034,
+      "grad_norm": 0.07010649629366013,
+      "learning_rate": 5.365753091941983e-08,
+      "loss": 0.3538,
+      "step": 3187
+    },
+    {
+      "epoch": 4.927357032457496,
+      "grad_norm": 0.07228072658478765,
+      "learning_rate": 5.1445563699283884e-08,
+      "loss": 0.3424,
+      "step": 3188
+    },
+    {
+      "epoch": 4.928902627511592,
+      "grad_norm": 0.07074974699598284,
+      "learning_rate": 4.928012469086696e-08,
+      "loss": 0.3546,
+      "step": 3189
+    },
+    {
+      "epoch": 4.930448222565688,
+      "grad_norm": 0.06738689125844004,
+      "learning_rate": 4.716121641626714e-08,
+      "loss": 0.3419,
+      "step": 3190
+    },
+    {
+      "epoch": 4.931993817619784,
+      "grad_norm": 0.06934501885583151,
+      "learning_rate": 4.5088841343385866e-08,
+      "loss": 0.3577,
+      "step": 3191
+    },
+    {
+      "epoch": 4.93353941267388,
+      "grad_norm": 0.074852165035623,
+      "learning_rate": 4.3063001885923496e-08,
+      "loss": 0.35,
+      "step": 3192
+    },
+    {
+      "epoch": 4.9350850077279755,
+      "grad_norm": 0.06897115937842231,
+      "learning_rate": 4.108370040337928e-08,
+      "loss": 0.3401,
+      "step": 3193
+    },
+    {
+      "epoch": 4.936630602782071,
+      "grad_norm": 0.07089437912697523,
+      "learning_rate": 3.915093920105584e-08,
+      "loss": 0.3567,
+      "step": 3194
+    },
+    {
+      "epoch": 4.938176197836167,
+      "grad_norm": 0.0711008694012866,
+      "learning_rate": 3.7264720530045815e-08,
+      "loss": 0.35,
+      "step": 3195
+    },
+    {
+      "epoch": 4.939721792890262,
+      "grad_norm": 0.06897510467703488,
+      "learning_rate": 3.542504658723189e-08,
+      "loss": 0.3466,
+      "step": 3196
+    },
+    {
+      "epoch": 4.941267387944358,
+      "grad_norm": 0.07161874661079988,
+      "learning_rate": 3.363191951528677e-08,
+      "loss": 0.343,
+      "step": 3197
+    },
+    {
+      "epoch": 4.942812982998454,
+      "grad_norm": 0.06800251521260453,
+      "learning_rate": 3.188534140267763e-08,
+      "loss": 0.3518,
+      "step": 3198
+    },
+    {
+      "epoch": 4.94435857805255,
+      "grad_norm": 0.06857550082527371,
+      "learning_rate": 3.018531428364835e-08,
+      "loss": 0.3541,
+      "step": 3199
+    },
+    {
+      "epoch": 4.945904173106646,
+      "grad_norm": 0.06842598322376241,
+      "learning_rate": 2.8531840138219523e-08,
+      "loss": 0.3483,
+      "step": 3200
+    },
+    {
+      "epoch": 4.947449768160742,
+      "grad_norm": 0.0699432865457825,
+      "learning_rate": 2.692492089220622e-08,
+      "loss": 0.3455,
+      "step": 3201
+    },
+    {
+      "epoch": 4.948995363214838,
+      "grad_norm": 0.07118104377541498,
+      "learning_rate": 2.5364558417186878e-08,
+      "loss": 0.3459,
+      "step": 3202
+    },
+    {
+      "epoch": 4.950540958268934,
+      "grad_norm": 0.07014681702978057,
+      "learning_rate": 2.3850754530521104e-08,
+      "loss": 0.3394,
+      "step": 3203
+    },
+    {
+      "epoch": 4.95208655332303,
+      "grad_norm": 0.07187162690120859,
+      "learning_rate": 2.2383510995349634e-08,
+      "loss": 0.3541,
+      "step": 3204
+    },
+    {
+      "epoch": 4.9536321483771255,
+      "grad_norm": 0.0686991664958423,
+      "learning_rate": 2.0962829520563277e-08,
+      "loss": 0.3551,
+      "step": 3205
+    },
+    {
+      "epoch": 4.955177743431221,
+      "grad_norm": 0.07159243273125415,
+      "learning_rate": 1.958871176083843e-08,
+      "loss": 0.3533,
+      "step": 3206
+    },
+    {
+      "epoch": 4.956723338485316,
+      "grad_norm": 0.07336825106437933,
+      "learning_rate": 1.8261159316619316e-08,
+      "loss": 0.3494,
+      "step": 3207
+    },
+    {
+      "epoch": 4.958268933539412,
+      "grad_norm": 0.06911403449022502,
+      "learning_rate": 1.6980173734109095e-08,
+      "loss": 0.3404,
+      "step": 3208
+    },
+    {
+      "epoch": 4.959814528593508,
+      "grad_norm": 0.07107073653552722,
+      "learning_rate": 1.5745756505269883e-08,
+      "loss": 0.341,
+      "step": 3209
+    },
+    {
+      "epoch": 4.961360123647604,
+      "grad_norm": 0.06968927460508988,
+      "learning_rate": 1.4557909067840493e-08,
+      "loss": 0.3577,
+      "step": 3210
+    },
+    {
+      "epoch": 4.9629057187017,
+      "grad_norm": 0.07308536629162159,
+      "learning_rate": 1.3416632805305363e-08,
+      "loss": 0.3539,
+      "step": 3211
+    },
+    {
+      "epoch": 4.964451313755796,
+      "grad_norm": 0.07152542735395838,
+      "learning_rate": 1.2321929046907877e-08,
+      "loss": 0.3455,
+      "step": 3212
+    },
+    {
+      "epoch": 4.965996908809892,
+      "grad_norm": 0.07195876470005634,
+      "learning_rate": 1.1273799067663683e-08,
+      "loss": 0.3532,
+      "step": 3213
+    },
+    {
+      "epoch": 4.967542503863988,
+      "grad_norm": 0.06941177989015325,
+      "learning_rate": 1.0272244088329608e-08,
+      "loss": 0.3445,
+      "step": 3214
+    },
+    {
+      "epoch": 4.969088098918084,
+      "grad_norm": 0.07016913867065504,
+      "learning_rate": 9.317265275412546e-09,
+      "loss": 0.353,
+      "step": 3215
+    },
+    {
+      "epoch": 4.97063369397218,
+      "grad_norm": 0.07421619000696215,
+      "learning_rate": 8.408863741191653e-09,
+      "loss": 0.3461,
+      "step": 3216
+    },
+    {
+      "epoch": 4.9721792890262755,
+      "grad_norm": 0.06969975222182093,
+      "learning_rate": 7.547040543673945e-09,
+      "loss": 0.3423,
+      "step": 3217
+    },
+    {
+      "epoch": 4.9737248840803705,
+      "grad_norm": 0.06875874300514737,
+      "learning_rate": 6.731796686634262e-09,
+      "loss": 0.3467,
+      "step": 3218
+    },
+    {
+      "epoch": 4.975270479134466,
+      "grad_norm": 0.06726335077458415,
+      "learning_rate": 5.963133119588627e-09,
+      "loss": 0.3497,
+      "step": 3219
+    },
+    {
+      "epoch": 4.976816074188562,
+      "grad_norm": 0.06995506586532436,
+      "learning_rate": 5.241050737803122e-09,
+      "loss": 0.3489,
+      "step": 3220
+    },
+    {
+      "epoch": 4.978361669242658,
+      "grad_norm": 0.06932164730463274,
+      "learning_rate": 4.5655503822850114e-09,
+      "loss": 0.3453,
+      "step": 3221
+    },
+    {
+      "epoch": 4.979907264296754,
+      "grad_norm": 0.06964687506918937,
+      "learning_rate": 3.936632839796062e-09,
+      "loss": 0.349,
+      "step": 3222
+    },
+    {
+      "epoch": 4.98145285935085,
+      "grad_norm": 0.06731212312953543,
+      "learning_rate": 3.35429884283478e-09,
+      "loss": 0.3504,
+      "step": 3223
+    },
+    {
+      "epoch": 4.982998454404946,
+      "grad_norm": 0.066728749601903,
+      "learning_rate": 2.8185490696497342e-09,
+      "loss": 0.3427,
+      "step": 3224
+    },
+    {
+      "epoch": 4.984544049459042,
+      "grad_norm": 0.06773974916352922,
+      "learning_rate": 2.3293841442351138e-09,
+      "loss": 0.3514,
+      "step": 3225
+    },
+    {
+      "epoch": 4.986089644513138,
+      "grad_norm": 0.06744823029918116,
+      "learning_rate": 1.8868046363174073e-09,
+      "loss": 0.3477,
+      "step": 3226
+    },
+    {
+      "epoch": 4.987635239567234,
+      "grad_norm": 0.07051766584886511,
+      "learning_rate": 1.4908110613687243e-09,
+      "loss": 0.3605,
+      "step": 3227
+    },
+    {
+      "epoch": 4.9891808346213296,
+      "grad_norm": 0.06998152135885344,
+      "learning_rate": 1.141403880606795e-09,
+      "loss": 0.3528,
+      "step": 3228
+    },
+    {
+      "epoch": 4.990726429675425,
+      "grad_norm": 0.06874012390063793,
+      "learning_rate": 8.385835009860899e-10,
+      "loss": 0.3448,
+      "step": 3229
+    },
+    {
+      "epoch": 4.992272024729521,
+      "grad_norm": 0.0684893242405293,
+      "learning_rate": 5.823502752067001e-10,
+      "loss": 0.3579,
+      "step": 3230
+    },
+    {
+      "epoch": 4.993817619783616,
+      "grad_norm": 0.0666169378166932,
+      "learning_rate": 3.7270450170101556e-10,
+      "loss": 0.3434,
+      "step": 3231
+    },
+    {
+      "epoch": 4.995363214837712,
+      "grad_norm": 0.0703101634456694,
+      "learning_rate": 2.0964642464260666e-10,
+      "loss": 0.3524,
+      "step": 3232
+    },
+    {
+      "epoch": 4.996908809891808,
+      "grad_norm": 0.06758631576854585,
+      "learning_rate": 9.317623394622388e-11,
+      "loss": 0.3586,
+      "step": 3233
+    },
+    {
+      "epoch": 4.998454404945904,
+      "grad_norm": 0.0689968575427259,
+      "learning_rate": 2.3294065267798203e-11,
+      "loss": 0.3527,
+      "step": 3234
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.08107722730902595,
+      "learning_rate": 0.0,
+      "loss": 0.3426,
+      "step": 3235
+    },
+    {
+      "epoch": 5.0,
+      "step": 3235,
+      "total_flos": 5.427082020519936e+16,
+      "train_loss": 0.6229784268500079,
+      "train_runtime": 70315.6352,
+      "train_samples_per_second": 23.525,
+      "train_steps_per_second": 0.046
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3235,
+  "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": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.427082020519936e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}