diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7785 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 1106,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0018083182640144665,
+      "grad_norm": 124.59085845947266,
+      "learning_rate": 1.7857142857142858e-07,
+      "loss": 1.8767,
+      "step": 1
+    },
+    {
+      "epoch": 0.003616636528028933,
+      "grad_norm": 123.01050567626953,
+      "learning_rate": 3.5714285714285716e-07,
+      "loss": 1.8663,
+      "step": 2
+    },
+    {
+      "epoch": 0.0054249547920434,
+      "grad_norm": 157.79428100585938,
+      "learning_rate": 5.357142857142857e-07,
+      "loss": 1.9201,
+      "step": 3
+    },
+    {
+      "epoch": 0.007233273056057866,
+      "grad_norm": 135.23194885253906,
+      "learning_rate": 7.142857142857143e-07,
+      "loss": 1.9675,
+      "step": 4
+    },
+    {
+      "epoch": 0.009041591320072333,
+      "grad_norm": 104.50148010253906,
+      "learning_rate": 8.928571428571429e-07,
+      "loss": 2.1019,
+      "step": 5
+    },
+    {
+      "epoch": 0.0108499095840868,
+      "grad_norm": 121.24005126953125,
+      "learning_rate": 1.0714285714285714e-06,
+      "loss": 2.043,
+      "step": 6
+    },
+    {
+      "epoch": 0.012658227848101266,
+      "grad_norm": 78.78042602539062,
+      "learning_rate": 1.25e-06,
+      "loss": 1.9918,
+      "step": 7
+    },
+    {
+      "epoch": 0.014466546112115732,
+      "grad_norm": 63.876441955566406,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 2.0925,
+      "step": 8
+    },
+    {
+      "epoch": 0.0162748643761302,
+      "grad_norm": 128.54800415039062,
+      "learning_rate": 1.6071428571428574e-06,
+      "loss": 1.9507,
+      "step": 9
+    },
+    {
+      "epoch": 0.018083182640144666,
+      "grad_norm": 110.11956024169922,
+      "learning_rate": 1.7857142857142859e-06,
+      "loss": 1.9933,
+      "step": 10
+    },
+    {
+      "epoch": 0.019891500904159132,
+      "grad_norm": 186.6797332763672,
+      "learning_rate": 1.9642857142857144e-06,
+      "loss": 1.7431,
+      "step": 11
+    },
+    {
+      "epoch": 0.0216998191681736,
+      "grad_norm": 289.7953796386719,
+      "learning_rate": 2.1428571428571427e-06,
+      "loss": 1.7807,
+      "step": 12
+    },
+    {
+      "epoch": 0.023508137432188065,
+      "grad_norm": 275.05596923828125,
+      "learning_rate": 2.321428571428572e-06,
+      "loss": 1.7687,
+      "step": 13
+    },
+    {
+      "epoch": 0.02531645569620253,
+      "grad_norm": 318.3794250488281,
+      "learning_rate": 2.5e-06,
+      "loss": 1.8591,
+      "step": 14
+    },
+    {
+      "epoch": 0.027124773960216998,
+      "grad_norm": 154.50942993164062,
+      "learning_rate": 2.6785714285714285e-06,
+      "loss": 1.7368,
+      "step": 15
+    },
+    {
+      "epoch": 0.028933092224231464,
+      "grad_norm": 65.9786148071289,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 1.654,
+      "step": 16
+    },
+    {
+      "epoch": 0.03074141048824593,
+      "grad_norm": 97.9104995727539,
+      "learning_rate": 3.0357142857142856e-06,
+      "loss": 1.7364,
+      "step": 17
+    },
+    {
+      "epoch": 0.0325497287522604,
+      "grad_norm": 86.56370544433594,
+      "learning_rate": 3.2142857142857147e-06,
+      "loss": 1.7569,
+      "step": 18
+    },
+    {
+      "epoch": 0.034358047016274866,
+      "grad_norm": 58.56660079956055,
+      "learning_rate": 3.3928571428571435e-06,
+      "loss": 1.6885,
+      "step": 19
+    },
+    {
+      "epoch": 0.03616636528028933,
+      "grad_norm": 41.65616226196289,
+      "learning_rate": 3.5714285714285718e-06,
+      "loss": 1.5784,
+      "step": 20
+    },
+    {
+      "epoch": 0.0379746835443038,
+      "grad_norm": 33.33378601074219,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 1.62,
+      "step": 21
+    },
+    {
+      "epoch": 0.039783001808318265,
+      "grad_norm": 29.60235595703125,
+      "learning_rate": 3.928571428571429e-06,
+      "loss": 1.6855,
+      "step": 22
+    },
+    {
+      "epoch": 0.04159132007233273,
+      "grad_norm": 26.571788787841797,
+      "learning_rate": 4.107142857142857e-06,
+      "loss": 1.7122,
+      "step": 23
+    },
+    {
+      "epoch": 0.0433996383363472,
+      "grad_norm": 19.6992130279541,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 1.6152,
+      "step": 24
+    },
+    {
+      "epoch": 0.045207956600361664,
+      "grad_norm": 40.05808639526367,
+      "learning_rate": 4.464285714285715e-06,
+      "loss": 1.6236,
+      "step": 25
+    },
+    {
+      "epoch": 0.04701627486437613,
+      "grad_norm": 11.833436012268066,
+      "learning_rate": 4.642857142857144e-06,
+      "loss": 1.2085,
+      "step": 26
+    },
+    {
+      "epoch": 0.048824593128390596,
+      "grad_norm": 19.057842254638672,
+      "learning_rate": 4.821428571428572e-06,
+      "loss": 1.1683,
+      "step": 27
+    },
+    {
+      "epoch": 0.05063291139240506,
+      "grad_norm": 21.635419845581055,
+      "learning_rate": 5e-06,
+      "loss": 1.2539,
+      "step": 28
+    },
+    {
+      "epoch": 0.05244122965641953,
+      "grad_norm": 19.195417404174805,
+      "learning_rate": 5.1785714285714296e-06,
+      "loss": 1.2789,
+      "step": 29
+    },
+    {
+      "epoch": 0.054249547920433995,
+      "grad_norm": 13.066848754882812,
+      "learning_rate": 5.357142857142857e-06,
+      "loss": 1.2662,
+      "step": 30
+    },
+    {
+      "epoch": 0.05605786618444846,
+      "grad_norm": 5.356083393096924,
+      "learning_rate": 5.535714285714286e-06,
+      "loss": 1.1504,
+      "step": 31
+    },
+    {
+      "epoch": 0.05786618444846293,
+      "grad_norm": 5.366274833679199,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 1.1215,
+      "step": 32
+    },
+    {
+      "epoch": 0.059674502712477394,
+      "grad_norm": 10.569491386413574,
+      "learning_rate": 5.892857142857144e-06,
+      "loss": 1.1052,
+      "step": 33
+    },
+    {
+      "epoch": 0.06148282097649186,
+      "grad_norm": 3.97001576423645,
+      "learning_rate": 6.071428571428571e-06,
+      "loss": 1.0776,
+      "step": 34
+    },
+    {
+      "epoch": 0.06329113924050633,
+      "grad_norm": 6.67834997177124,
+      "learning_rate": 6.25e-06,
+      "loss": 1.1684,
+      "step": 35
+    },
+    {
+      "epoch": 0.0650994575045208,
+      "grad_norm": 6.470335483551025,
+      "learning_rate": 6.4285714285714295e-06,
+      "loss": 1.0797,
+      "step": 36
+    },
+    {
+      "epoch": 0.06690777576853527,
+      "grad_norm": 4.232357501983643,
+      "learning_rate": 6.607142857142858e-06,
+      "loss": 1.0273,
+      "step": 37
+    },
+    {
+      "epoch": 0.06871609403254973,
+      "grad_norm": 4.908628940582275,
+      "learning_rate": 6.785714285714287e-06,
+      "loss": 1.0514,
+      "step": 38
+    },
+    {
+      "epoch": 0.0705244122965642,
+      "grad_norm": 3.294133424758911,
+      "learning_rate": 6.964285714285714e-06,
+      "loss": 1.1107,
+      "step": 39
+    },
+    {
+      "epoch": 0.07233273056057866,
+      "grad_norm": 3.946671962738037,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 1.0426,
+      "step": 40
+    },
+    {
+      "epoch": 0.07414104882459313,
+      "grad_norm": 5.3715949058532715,
+      "learning_rate": 7.321428571428572e-06,
+      "loss": 1.0513,
+      "step": 41
+    },
+    {
+      "epoch": 0.0759493670886076,
+      "grad_norm": 7.43781852722168,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.06,
+      "step": 42
+    },
+    {
+      "epoch": 0.07775768535262206,
+      "grad_norm": 4.168594837188721,
+      "learning_rate": 7.67857142857143e-06,
+      "loss": 1.1397,
+      "step": 43
+    },
+    {
+      "epoch": 0.07956600361663653,
+      "grad_norm": 38.190696716308594,
+      "learning_rate": 7.857142857142858e-06,
+      "loss": 1.0466,
+      "step": 44
+    },
+    {
+      "epoch": 0.081374321880651,
+      "grad_norm": 5.13964319229126,
+      "learning_rate": 8.035714285714286e-06,
+      "loss": 1.1128,
+      "step": 45
+    },
+    {
+      "epoch": 0.08318264014466546,
+      "grad_norm": 5.730332851409912,
+      "learning_rate": 8.214285714285714e-06,
+      "loss": 1.1362,
+      "step": 46
+    },
+    {
+      "epoch": 0.08499095840867993,
+      "grad_norm": 3.4547622203826904,
+      "learning_rate": 8.392857142857144e-06,
+      "loss": 1.1364,
+      "step": 47
+    },
+    {
+      "epoch": 0.0867992766726944,
+      "grad_norm": 3.945357322692871,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 1.0957,
+      "step": 48
+    },
+    {
+      "epoch": 0.08860759493670886,
+      "grad_norm": 14.470746994018555,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 1.0731,
+      "step": 49
+    },
+    {
+      "epoch": 0.09041591320072333,
+      "grad_norm": 4.687364101409912,
+      "learning_rate": 8.92857142857143e-06,
+      "loss": 1.1294,
+      "step": 50
+    },
+    {
+      "epoch": 0.0922242314647378,
+      "grad_norm": 3.4039835929870605,
+      "learning_rate": 9.107142857142858e-06,
+      "loss": 0.9838,
+      "step": 51
+    },
+    {
+      "epoch": 0.09403254972875226,
+      "grad_norm": 17.923242568969727,
+      "learning_rate": 9.285714285714288e-06,
+      "loss": 0.9088,
+      "step": 52
+    },
+    {
+      "epoch": 0.09584086799276673,
+      "grad_norm": 3.5276126861572266,
+      "learning_rate": 9.464285714285714e-06,
+      "loss": 0.9651,
+      "step": 53
+    },
+    {
+      "epoch": 0.09764918625678119,
+      "grad_norm": 4.005324840545654,
+      "learning_rate": 9.642857142857144e-06,
+      "loss": 1.0155,
+      "step": 54
+    },
+    {
+      "epoch": 0.09945750452079566,
+      "grad_norm": 3.9271018505096436,
+      "learning_rate": 9.821428571428573e-06,
+      "loss": 1.0829,
+      "step": 55
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 4.813779354095459,
+      "learning_rate": 1e-05,
+      "loss": 0.9691,
+      "step": 56
+    },
+    {
+      "epoch": 0.10307414104882459,
+      "grad_norm": 3.8365554809570312,
+      "learning_rate": 9.999977619961366e-06,
+      "loss": 0.9716,
+      "step": 57
+    },
+    {
+      "epoch": 0.10488245931283906,
+      "grad_norm": 4.112904071807861,
+      "learning_rate": 9.999910480045805e-06,
+      "loss": 1.0489,
+      "step": 58
+    },
+    {
+      "epoch": 0.10669077757685352,
+      "grad_norm": 4.034896373748779,
+      "learning_rate": 9.999798580854356e-06,
+      "loss": 1.0521,
+      "step": 59
+    },
+    {
+      "epoch": 0.10849909584086799,
+      "grad_norm": 4.018319129943848,
+      "learning_rate": 9.999641923388745e-06,
+      "loss": 0.9569,
+      "step": 60
+    },
+    {
+      "epoch": 0.11030741410488246,
+      "grad_norm": 4.899073123931885,
+      "learning_rate": 9.999440509051367e-06,
+      "loss": 0.9694,
+      "step": 61
+    },
+    {
+      "epoch": 0.11211573236889692,
+      "grad_norm": 9.774127960205078,
+      "learning_rate": 9.999194339645292e-06,
+      "loss": 1.0075,
+      "step": 62
+    },
+    {
+      "epoch": 0.11392405063291139,
+      "grad_norm": 5.587623596191406,
+      "learning_rate": 9.998903417374228e-06,
+      "loss": 1.0096,
+      "step": 63
+    },
+    {
+      "epoch": 0.11573236889692586,
+      "grad_norm": 4.204263687133789,
+      "learning_rate": 9.998567744842518e-06,
+      "loss": 0.9308,
+      "step": 64
+    },
+    {
+      "epoch": 0.11754068716094032,
+      "grad_norm": 5.104825973510742,
+      "learning_rate": 9.998187325055107e-06,
+      "loss": 0.8768,
+      "step": 65
+    },
+    {
+      "epoch": 0.11934900542495479,
+      "grad_norm": 5.699406623840332,
+      "learning_rate": 9.997762161417517e-06,
+      "loss": 0.9115,
+      "step": 66
+    },
+    {
+      "epoch": 0.12115732368896925,
+      "grad_norm": 4.967496871948242,
+      "learning_rate": 9.997292257735822e-06,
+      "loss": 0.8973,
+      "step": 67
+    },
+    {
+      "epoch": 0.12296564195298372,
+      "grad_norm": 5.5881500244140625,
+      "learning_rate": 9.996777618216608e-06,
+      "loss": 0.9434,
+      "step": 68
+    },
+    {
+      "epoch": 0.12477396021699819,
+      "grad_norm": 5.467623233795166,
+      "learning_rate": 9.996218247466932e-06,
+      "loss": 0.9518,
+      "step": 69
+    },
+    {
+      "epoch": 0.12658227848101267,
+      "grad_norm": 5.854726314544678,
+      "learning_rate": 9.995614150494293e-06,
+      "loss": 0.9395,
+      "step": 70
+    },
+    {
+      "epoch": 0.12839059674502712,
+      "grad_norm": 8.993833541870117,
+      "learning_rate": 9.994965332706574e-06,
+      "loss": 0.9938,
+      "step": 71
+    },
+    {
+      "epoch": 0.1301989150090416,
+      "grad_norm": 6.169260501861572,
+      "learning_rate": 9.994271799912004e-06,
+      "loss": 0.9776,
+      "step": 72
+    },
+    {
+      "epoch": 0.13200723327305605,
+      "grad_norm": 6.879613399505615,
+      "learning_rate": 9.993533558319098e-06,
+      "loss": 0.9327,
+      "step": 73
+    },
+    {
+      "epoch": 0.13381555153707053,
+      "grad_norm": 6.696218490600586,
+      "learning_rate": 9.992750614536606e-06,
+      "loss": 0.972,
+      "step": 74
+    },
+    {
+      "epoch": 0.13562386980108498,
+      "grad_norm": 6.739992618560791,
+      "learning_rate": 9.991922975573453e-06,
+      "loss": 1.0328,
+      "step": 75
+    },
+    {
+      "epoch": 0.13743218806509946,
+      "grad_norm": 5.942216873168945,
+      "learning_rate": 9.991050648838676e-06,
+      "loss": 0.834,
+      "step": 76
+    },
+    {
+      "epoch": 0.13924050632911392,
+      "grad_norm": 5.648139953613281,
+      "learning_rate": 9.990133642141359e-06,
+      "loss": 0.7784,
+      "step": 77
+    },
+    {
+      "epoch": 0.1410488245931284,
+      "grad_norm": 6.231765270233154,
+      "learning_rate": 9.989171963690556e-06,
+      "loss": 0.7989,
+      "step": 78
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 6.385239124298096,
+      "learning_rate": 9.988165622095233e-06,
+      "loss": 0.8615,
+      "step": 79
+    },
+    {
+      "epoch": 0.14466546112115733,
+      "grad_norm": 6.530540943145752,
+      "learning_rate": 9.987114626364172e-06,
+      "loss": 0.8554,
+      "step": 80
+    },
+    {
+      "epoch": 0.14647377938517178,
+      "grad_norm": 6.088751792907715,
+      "learning_rate": 9.986018985905901e-06,
+      "loss": 0.8795,
+      "step": 81
+    },
+    {
+      "epoch": 0.14828209764918626,
+      "grad_norm": 5.742943286895752,
+      "learning_rate": 9.984878710528615e-06,
+      "loss": 0.7886,
+      "step": 82
+    },
+    {
+      "epoch": 0.15009041591320071,
+      "grad_norm": 6.685095310211182,
+      "learning_rate": 9.983693810440073e-06,
+      "loss": 0.7599,
+      "step": 83
+    },
+    {
+      "epoch": 0.1518987341772152,
+      "grad_norm": 6.117989540100098,
+      "learning_rate": 9.982464296247523e-06,
+      "loss": 0.8052,
+      "step": 84
+    },
+    {
+      "epoch": 0.15370705244122965,
+      "grad_norm": 6.1924004554748535,
+      "learning_rate": 9.98119017895759e-06,
+      "loss": 0.7964,
+      "step": 85
+    },
+    {
+      "epoch": 0.15551537070524413,
+      "grad_norm": 5.929360866546631,
+      "learning_rate": 9.979871469976197e-06,
+      "loss": 0.8647,
+      "step": 86
+    },
+    {
+      "epoch": 0.15732368896925858,
+      "grad_norm": 5.7412872314453125,
+      "learning_rate": 9.978508181108442e-06,
+      "loss": 0.7466,
+      "step": 87
+    },
+    {
+      "epoch": 0.15913200723327306,
+      "grad_norm": 5.8075995445251465,
+      "learning_rate": 9.97710032455851e-06,
+      "loss": 0.6711,
+      "step": 88
+    },
+    {
+      "epoch": 0.1609403254972875,
+      "grad_norm": 6.012415409088135,
+      "learning_rate": 9.975647912929558e-06,
+      "loss": 0.6933,
+      "step": 89
+    },
+    {
+      "epoch": 0.162748643761302,
+      "grad_norm": 5.956928253173828,
+      "learning_rate": 9.974150959223591e-06,
+      "loss": 0.7058,
+      "step": 90
+    },
+    {
+      "epoch": 0.16455696202531644,
+      "grad_norm": 5.904573917388916,
+      "learning_rate": 9.972609476841368e-06,
+      "loss": 0.7432,
+      "step": 91
+    },
+    {
+      "epoch": 0.16636528028933092,
+      "grad_norm": 6.147512912750244,
+      "learning_rate": 9.971023479582258e-06,
+      "loss": 0.6556,
+      "step": 92
+    },
+    {
+      "epoch": 0.16817359855334538,
+      "grad_norm": 5.897213935852051,
+      "learning_rate": 9.969392981644138e-06,
+      "loss": 0.7354,
+      "step": 93
+    },
+    {
+      "epoch": 0.16998191681735986,
+      "grad_norm": 6.633100509643555,
+      "learning_rate": 9.967717997623245e-06,
+      "loss": 0.694,
+      "step": 94
+    },
+    {
+      "epoch": 0.1717902350813743,
+      "grad_norm": 6.375577449798584,
+      "learning_rate": 9.965998542514066e-06,
+      "loss": 0.6957,
+      "step": 95
+    },
+    {
+      "epoch": 0.1735985533453888,
+      "grad_norm": 6.096574783325195,
+      "learning_rate": 9.964234631709188e-06,
+      "loss": 0.6565,
+      "step": 96
+    },
+    {
+      "epoch": 0.17540687160940324,
+      "grad_norm": 6.757542610168457,
+      "learning_rate": 9.962426280999168e-06,
+      "loss": 0.6353,
+      "step": 97
+    },
+    {
+      "epoch": 0.17721518987341772,
+      "grad_norm": 6.78255558013916,
+      "learning_rate": 9.960573506572391e-06,
+      "loss": 0.6166,
+      "step": 98
+    },
+    {
+      "epoch": 0.17902350813743217,
+      "grad_norm": 5.72428035736084,
+      "learning_rate": 9.95867632501492e-06,
+      "loss": 0.7272,
+      "step": 99
+    },
+    {
+      "epoch": 0.18083182640144665,
+      "grad_norm": 6.663970470428467,
+      "learning_rate": 9.956734753310355e-06,
+      "loss": 0.6538,
+      "step": 100
+    },
+    {
+      "epoch": 0.18264014466546113,
+      "grad_norm": 5.2808098793029785,
+      "learning_rate": 9.954748808839675e-06,
+      "loss": 0.5545,
+      "step": 101
+    },
+    {
+      "epoch": 0.1844484629294756,
+      "grad_norm": 5.146191596984863,
+      "learning_rate": 9.952718509381086e-06,
+      "loss": 0.6049,
+      "step": 102
+    },
+    {
+      "epoch": 0.18625678119349007,
+      "grad_norm": 5.410797119140625,
+      "learning_rate": 9.950643873109861e-06,
+      "loss": 0.6559,
+      "step": 103
+    },
+    {
+      "epoch": 0.18806509945750452,
+      "grad_norm": 5.563495635986328,
+      "learning_rate": 9.948524918598175e-06,
+      "loss": 0.6172,
+      "step": 104
+    },
+    {
+      "epoch": 0.189873417721519,
+      "grad_norm": 4.960428714752197,
+      "learning_rate": 9.946361664814942e-06,
+      "loss": 0.6353,
+      "step": 105
+    },
+    {
+      "epoch": 0.19168173598553345,
+      "grad_norm": 5.002950191497803,
+      "learning_rate": 9.944154131125643e-06,
+      "loss": 0.6519,
+      "step": 106
+    },
+    {
+      "epoch": 0.19349005424954793,
+      "grad_norm": 5.334290027618408,
+      "learning_rate": 9.941902337292156e-06,
+      "loss": 0.5404,
+      "step": 107
+    },
+    {
+      "epoch": 0.19529837251356238,
+      "grad_norm": 5.085877418518066,
+      "learning_rate": 9.93960630347257e-06,
+      "loss": 0.5654,
+      "step": 108
+    },
+    {
+      "epoch": 0.19710669077757687,
+      "grad_norm": 4.713104724884033,
+      "learning_rate": 9.937266050221015e-06,
+      "loss": 0.4798,
+      "step": 109
+    },
+    {
+      "epoch": 0.19891500904159132,
+      "grad_norm": 4.222567558288574,
+      "learning_rate": 9.934881598487478e-06,
+      "loss": 0.6132,
+      "step": 110
+    },
+    {
+      "epoch": 0.2007233273056058,
+      "grad_norm": 4.574428081512451,
+      "learning_rate": 9.932452969617607e-06,
+      "loss": 0.5248,
+      "step": 111
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 4.324956893920898,
+      "learning_rate": 9.929980185352525e-06,
+      "loss": 0.5395,
+      "step": 112
+    },
+    {
+      "epoch": 0.20433996383363473,
+      "grad_norm": 4.227971076965332,
+      "learning_rate": 9.927463267828635e-06,
+      "loss": 0.4582,
+      "step": 113
+    },
+    {
+      "epoch": 0.20614828209764918,
+      "grad_norm": 4.175861358642578,
+      "learning_rate": 9.924902239577419e-06,
+      "loss": 0.482,
+      "step": 114
+    },
+    {
+      "epoch": 0.20795660036166366,
+      "grad_norm": 3.743654251098633,
+      "learning_rate": 9.922297123525244e-06,
+      "loss": 0.5059,
+      "step": 115
+    },
+    {
+      "epoch": 0.20976491862567812,
+      "grad_norm": 3.767977714538574,
+      "learning_rate": 9.91964794299315e-06,
+      "loss": 0.4781,
+      "step": 116
+    },
+    {
+      "epoch": 0.2115732368896926,
+      "grad_norm": 3.4714298248291016,
+      "learning_rate": 9.91695472169664e-06,
+      "loss": 0.5297,
+      "step": 117
+    },
+    {
+      "epoch": 0.21338155515370705,
+      "grad_norm": 3.7578494548797607,
+      "learning_rate": 9.914217483745472e-06,
+      "loss": 0.4538,
+      "step": 118
+    },
+    {
+      "epoch": 0.21518987341772153,
+      "grad_norm": 3.590576410293579,
+      "learning_rate": 9.911436253643445e-06,
+      "loss": 0.4805,
+      "step": 119
+    },
+    {
+      "epoch": 0.21699819168173598,
+      "grad_norm": 3.2946393489837646,
+      "learning_rate": 9.90861105628817e-06,
+      "loss": 0.4111,
+      "step": 120
+    },
+    {
+      "epoch": 0.21880650994575046,
+      "grad_norm": 3.1186115741729736,
+      "learning_rate": 9.905741916970863e-06,
+      "loss": 0.5042,
+      "step": 121
+    },
+    {
+      "epoch": 0.2206148282097649,
+      "grad_norm": 3.2415828704833984,
+      "learning_rate": 9.902828861376101e-06,
+      "loss": 0.5124,
+      "step": 122
+    },
+    {
+      "epoch": 0.2224231464737794,
+      "grad_norm": 3.1252877712249756,
+      "learning_rate": 9.8998719155816e-06,
+      "loss": 0.4682,
+      "step": 123
+    },
+    {
+      "epoch": 0.22423146473779385,
+      "grad_norm": 2.8531622886657715,
+      "learning_rate": 9.896871106057989e-06,
+      "loss": 0.4516,
+      "step": 124
+    },
+    {
+      "epoch": 0.22603978300180833,
+      "grad_norm": 2.932840585708618,
+      "learning_rate": 9.89382645966856e-06,
+      "loss": 0.4935,
+      "step": 125
+    },
+    {
+      "epoch": 0.22784810126582278,
+      "grad_norm": 2.7010486125946045,
+      "learning_rate": 9.890738003669029e-06,
+      "loss": 0.5096,
+      "step": 126
+    },
+    {
+      "epoch": 0.22965641952983726,
+      "grad_norm": 2.4493417739868164,
+      "learning_rate": 9.887605765707309e-06,
+      "loss": 0.4621,
+      "step": 127
+    },
+    {
+      "epoch": 0.2314647377938517,
+      "grad_norm": 2.3981196880340576,
+      "learning_rate": 9.884429773823238e-06,
+      "loss": 0.3943,
+      "step": 128
+    },
+    {
+      "epoch": 0.2332730560578662,
+      "grad_norm": 2.549567222595215,
+      "learning_rate": 9.88121005644835e-06,
+      "loss": 0.3628,
+      "step": 129
+    },
+    {
+      "epoch": 0.23508137432188064,
+      "grad_norm": 2.2552852630615234,
+      "learning_rate": 9.877946642405598e-06,
+      "loss": 0.4017,
+      "step": 130
+    },
+    {
+      "epoch": 0.23688969258589512,
+      "grad_norm": 2.451272964477539,
+      "learning_rate": 9.874639560909118e-06,
+      "loss": 0.5817,
+      "step": 131
+    },
+    {
+      "epoch": 0.23869801084990958,
+      "grad_norm": 2.375648021697998,
+      "learning_rate": 9.871288841563956e-06,
+      "loss": 0.5272,
+      "step": 132
+    },
+    {
+      "epoch": 0.24050632911392406,
+      "grad_norm": 2.253958225250244,
+      "learning_rate": 9.867894514365802e-06,
+      "loss": 0.4738,
+      "step": 133
+    },
+    {
+      "epoch": 0.2423146473779385,
+      "grad_norm": 2.3769643306732178,
+      "learning_rate": 9.864456609700726e-06,
+      "loss": 0.5198,
+      "step": 134
+    },
+    {
+      "epoch": 0.244122965641953,
+      "grad_norm": 2.0296530723571777,
+      "learning_rate": 9.860975158344902e-06,
+      "loss": 0.421,
+      "step": 135
+    },
+    {
+      "epoch": 0.24593128390596744,
+      "grad_norm": 2.2663729190826416,
+      "learning_rate": 9.857450191464337e-06,
+      "loss": 0.3671,
+      "step": 136
+    },
+    {
+      "epoch": 0.24773960216998192,
+      "grad_norm": 1.969122052192688,
+      "learning_rate": 9.853881740614591e-06,
+      "loss": 0.3928,
+      "step": 137
+    },
+    {
+      "epoch": 0.24954792043399637,
+      "grad_norm": 2.0730412006378174,
+      "learning_rate": 9.85026983774049e-06,
+      "loss": 0.3816,
+      "step": 138
+    },
+    {
+      "epoch": 0.2513562386980108,
+      "grad_norm": 2.104492425918579,
+      "learning_rate": 9.846614515175843e-06,
+      "loss": 0.4365,
+      "step": 139
+    },
+    {
+      "epoch": 0.25316455696202533,
+      "grad_norm": 2.0811867713928223,
+      "learning_rate": 9.842915805643156e-06,
+      "loss": 0.3995,
+      "step": 140
+    },
+    {
+      "epoch": 0.2549728752260398,
+      "grad_norm": 2.356283187866211,
+      "learning_rate": 9.839173742253334e-06,
+      "loss": 0.4335,
+      "step": 141
+    },
+    {
+      "epoch": 0.25678119349005424,
+      "grad_norm": 1.9231618642807007,
+      "learning_rate": 9.835388358505383e-06,
+      "loss": 0.4127,
+      "step": 142
+    },
+    {
+      "epoch": 0.2585895117540687,
+      "grad_norm": 2.0608198642730713,
+      "learning_rate": 9.83155968828612e-06,
+      "loss": 0.4257,
+      "step": 143
+    },
+    {
+      "epoch": 0.2603978300180832,
+      "grad_norm": 2.212334394454956,
+      "learning_rate": 9.827687765869859e-06,
+      "loss": 0.3849,
+      "step": 144
+    },
+    {
+      "epoch": 0.26220614828209765,
+      "grad_norm": 2.1471047401428223,
+      "learning_rate": 9.823772625918111e-06,
+      "loss": 0.4147,
+      "step": 145
+    },
+    {
+      "epoch": 0.2640144665461121,
+      "grad_norm": 2.3254005908966064,
+      "learning_rate": 9.819814303479268e-06,
+      "loss": 0.4096,
+      "step": 146
+    },
+    {
+      "epoch": 0.26582278481012656,
+      "grad_norm": 2.1744768619537354,
+      "learning_rate": 9.815812833988292e-06,
+      "loss": 0.4814,
+      "step": 147
+    },
+    {
+      "epoch": 0.26763110307414106,
+      "grad_norm": 2.347172260284424,
+      "learning_rate": 9.811768253266401e-06,
+      "loss": 0.4409,
+      "step": 148
+    },
+    {
+      "epoch": 0.2694394213381555,
+      "grad_norm": 2.213552713394165,
+      "learning_rate": 9.807680597520746e-06,
+      "loss": 0.559,
+      "step": 149
+    },
+    {
+      "epoch": 0.27124773960216997,
+      "grad_norm": 2.426069498062134,
+      "learning_rate": 9.803549903344081e-06,
+      "loss": 0.4766,
+      "step": 150
+    },
+    {
+      "epoch": 0.2730560578661845,
+      "grad_norm": 2.633708953857422,
+      "learning_rate": 9.799376207714446e-06,
+      "loss": 0.4687,
+      "step": 151
+    },
+    {
+      "epoch": 0.27486437613019893,
+      "grad_norm": 2.165501117706299,
+      "learning_rate": 9.79515954799483e-06,
+      "loss": 0.3533,
+      "step": 152
+    },
+    {
+      "epoch": 0.2766726943942134,
+      "grad_norm": 2.4209182262420654,
+      "learning_rate": 9.790899961932833e-06,
+      "loss": 0.378,
+      "step": 153
+    },
+    {
+      "epoch": 0.27848101265822783,
+      "grad_norm": 2.345797538757324,
+      "learning_rate": 9.786597487660336e-06,
+      "loss": 0.5108,
+      "step": 154
+    },
+    {
+      "epoch": 0.28028933092224234,
+      "grad_norm": 4.470305442810059,
+      "learning_rate": 9.782252163693159e-06,
+      "loss": 0.4871,
+      "step": 155
+    },
+    {
+      "epoch": 0.2820976491862568,
+      "grad_norm": 2.4609861373901367,
+      "learning_rate": 9.777864028930705e-06,
+      "loss": 0.4943,
+      "step": 156
+    },
+    {
+      "epoch": 0.28390596745027125,
+      "grad_norm": 1.9806599617004395,
+      "learning_rate": 9.773433122655625e-06,
+      "loss": 0.3348,
+      "step": 157
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 2.1933321952819824,
+      "learning_rate": 9.768959484533461e-06,
+      "loss": 0.4652,
+      "step": 158
+    },
+    {
+      "epoch": 0.2875226039783002,
+      "grad_norm": 2.601301670074463,
+      "learning_rate": 9.76444315461229e-06,
+      "loss": 0.4319,
+      "step": 159
+    },
+    {
+      "epoch": 0.28933092224231466,
+      "grad_norm": 1.9371708631515503,
+      "learning_rate": 9.75988417332237e-06,
+      "loss": 0.3555,
+      "step": 160
+    },
+    {
+      "epoch": 0.2911392405063291,
+      "grad_norm": 2.256004810333252,
+      "learning_rate": 9.755282581475769e-06,
+      "loss": 0.4961,
+      "step": 161
+    },
+    {
+      "epoch": 0.29294755877034356,
+      "grad_norm": 3.1606812477111816,
+      "learning_rate": 9.750638420266008e-06,
+      "loss": 0.3908,
+      "step": 162
+    },
+    {
+      "epoch": 0.29475587703435807,
+      "grad_norm": 2.105372667312622,
+      "learning_rate": 9.745951731267693e-06,
+      "loss": 0.4479,
+      "step": 163
+    },
+    {
+      "epoch": 0.2965641952983725,
+      "grad_norm": 2.355839490890503,
+      "learning_rate": 9.741222556436132e-06,
+      "loss": 0.4198,
+      "step": 164
+    },
+    {
+      "epoch": 0.298372513562387,
+      "grad_norm": 2.2613070011138916,
+      "learning_rate": 9.736450938106976e-06,
+      "loss": 0.4441,
+      "step": 165
+    },
+    {
+      "epoch": 0.30018083182640143,
+      "grad_norm": 2.1388442516326904,
+      "learning_rate": 9.731636918995821e-06,
+      "loss": 0.3449,
+      "step": 166
+    },
+    {
+      "epoch": 0.30198915009041594,
+      "grad_norm": 2.255647659301758,
+      "learning_rate": 9.726780542197845e-06,
+      "loss": 0.4772,
+      "step": 167
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 2.3774516582489014,
+      "learning_rate": 9.721881851187406e-06,
+      "loss": 0.4102,
+      "step": 168
+    },
+    {
+      "epoch": 0.30560578661844484,
+      "grad_norm": 2.328238010406494,
+      "learning_rate": 9.716940889817662e-06,
+      "loss": 0.3546,
+      "step": 169
+    },
+    {
+      "epoch": 0.3074141048824593,
+      "grad_norm": 2.1654934883117676,
+      "learning_rate": 9.711957702320176e-06,
+      "loss": 0.3346,
+      "step": 170
+    },
+    {
+      "epoch": 0.3092224231464738,
+      "grad_norm": 2.319549798965454,
+      "learning_rate": 9.706932333304518e-06,
+      "loss": 0.3936,
+      "step": 171
+    },
+    {
+      "epoch": 0.31103074141048825,
+      "grad_norm": 2.4396815299987793,
+      "learning_rate": 9.701864827757868e-06,
+      "loss": 0.4511,
+      "step": 172
+    },
+    {
+      "epoch": 0.3128390596745027,
+      "grad_norm": 2.3042893409729004,
+      "learning_rate": 9.696755231044618e-06,
+      "loss": 0.356,
+      "step": 173
+    },
+    {
+      "epoch": 0.31464737793851716,
+      "grad_norm": 2.3526265621185303,
+      "learning_rate": 9.691603588905956e-06,
+      "loss": 0.3223,
+      "step": 174
+    },
+    {
+      "epoch": 0.31645569620253167,
+      "grad_norm": 2.7514851093292236,
+      "learning_rate": 9.68640994745946e-06,
+      "loss": 0.551,
+      "step": 175
+    },
+    {
+      "epoch": 0.3182640144665461,
+      "grad_norm": 2.0451977252960205,
+      "learning_rate": 9.681174353198687e-06,
+      "loss": 0.3433,
+      "step": 176
+    },
+    {
+      "epoch": 0.32007233273056057,
+      "grad_norm": 2.3086750507354736,
+      "learning_rate": 9.675896852992762e-06,
+      "loss": 0.4105,
+      "step": 177
+    },
+    {
+      "epoch": 0.321880650994575,
+      "grad_norm": 1.9786276817321777,
+      "learning_rate": 9.670577494085945e-06,
+      "loss": 0.3799,
+      "step": 178
+    },
+    {
+      "epoch": 0.32368896925858953,
+      "grad_norm": 2.3861396312713623,
+      "learning_rate": 9.665216324097222e-06,
+      "loss": 0.5156,
+      "step": 179
+    },
+    {
+      "epoch": 0.325497287522604,
+      "grad_norm": 2.164726734161377,
+      "learning_rate": 9.659813391019867e-06,
+      "loss": 0.4335,
+      "step": 180
+    },
+    {
+      "epoch": 0.32730560578661844,
+      "grad_norm": 2.0450663566589355,
+      "learning_rate": 9.654368743221022e-06,
+      "loss": 0.3984,
+      "step": 181
+    },
+    {
+      "epoch": 0.3291139240506329,
+      "grad_norm": 2.014610767364502,
+      "learning_rate": 9.648882429441258e-06,
+      "loss": 0.3474,
+      "step": 182
+    },
+    {
+      "epoch": 0.3309222423146474,
+      "grad_norm": 1.9683150053024292,
+      "learning_rate": 9.643354498794139e-06,
+      "loss": 0.3614,
+      "step": 183
+    },
+    {
+      "epoch": 0.33273056057866185,
+      "grad_norm": 1.8113188743591309,
+      "learning_rate": 9.637785000765789e-06,
+      "loss": 0.3701,
+      "step": 184
+    },
+    {
+      "epoch": 0.3345388788426763,
+      "grad_norm": 1.8580265045166016,
+      "learning_rate": 9.632173985214438e-06,
+      "loss": 0.2733,
+      "step": 185
+    },
+    {
+      "epoch": 0.33634719710669075,
+      "grad_norm": 1.8365068435668945,
+      "learning_rate": 9.626521502369984e-06,
+      "loss": 0.3113,
+      "step": 186
+    },
+    {
+      "epoch": 0.33815551537070526,
+      "grad_norm": 1.8660293817520142,
+      "learning_rate": 9.620827602833542e-06,
+      "loss": 0.3299,
+      "step": 187
+    },
+    {
+      "epoch": 0.3399638336347197,
+      "grad_norm": 1.9789987802505493,
+      "learning_rate": 9.615092337576987e-06,
+      "loss": 0.3565,
+      "step": 188
+    },
+    {
+      "epoch": 0.34177215189873417,
+      "grad_norm": 1.978827714920044,
+      "learning_rate": 9.609315757942504e-06,
+      "loss": 0.3514,
+      "step": 189
+    },
+    {
+      "epoch": 0.3435804701627486,
+      "grad_norm": 2.199197769165039,
+      "learning_rate": 9.603497915642122e-06,
+      "loss": 0.4354,
+      "step": 190
+    },
+    {
+      "epoch": 0.3453887884267631,
+      "grad_norm": 1.6824909448623657,
+      "learning_rate": 9.597638862757255e-06,
+      "loss": 0.3105,
+      "step": 191
+    },
+    {
+      "epoch": 0.3471971066907776,
+      "grad_norm": 1.8425745964050293,
+      "learning_rate": 9.591738651738235e-06,
+      "loss": 0.3257,
+      "step": 192
+    },
+    {
+      "epoch": 0.34900542495479203,
+      "grad_norm": 1.7892580032348633,
+      "learning_rate": 9.585797335403843e-06,
+      "loss": 0.3323,
+      "step": 193
+    },
+    {
+      "epoch": 0.3508137432188065,
+      "grad_norm": 1.8971889019012451,
+      "learning_rate": 9.579814966940833e-06,
+      "loss": 0.2936,
+      "step": 194
+    },
+    {
+      "epoch": 0.352622061482821,
+      "grad_norm": 2.005496025085449,
+      "learning_rate": 9.573791599903463e-06,
+      "loss": 0.3361,
+      "step": 195
+    },
+    {
+      "epoch": 0.35443037974683544,
+      "grad_norm": 1.9333208799362183,
+      "learning_rate": 9.567727288213005e-06,
+      "loss": 0.3354,
+      "step": 196
+    },
+    {
+      "epoch": 0.3562386980108499,
+      "grad_norm": 1.880614995956421,
+      "learning_rate": 9.561622086157273e-06,
+      "loss": 0.3193,
+      "step": 197
+    },
+    {
+      "epoch": 0.35804701627486435,
+      "grad_norm": 1.9405372142791748,
+      "learning_rate": 9.55547604839013e-06,
+      "loss": 0.3773,
+      "step": 198
+    },
+    {
+      "epoch": 0.35985533453887886,
+      "grad_norm": 1.8647263050079346,
+      "learning_rate": 9.549289229930997e-06,
+      "loss": 0.4083,
+      "step": 199
+    },
+    {
+      "epoch": 0.3616636528028933,
+      "grad_norm": 2.569082498550415,
+      "learning_rate": 9.543061686164374e-06,
+      "loss": 0.5087,
+      "step": 200
+    },
+    {
+      "epoch": 0.36347197106690776,
+      "grad_norm": 1.94440495967865,
+      "learning_rate": 9.536793472839325e-06,
+      "loss": 0.4153,
+      "step": 201
+    },
+    {
+      "epoch": 0.36528028933092227,
+      "grad_norm": 2.031101942062378,
+      "learning_rate": 9.530484646068996e-06,
+      "loss": 0.3542,
+      "step": 202
+    },
+    {
+      "epoch": 0.3670886075949367,
+      "grad_norm": 2.0991175174713135,
+      "learning_rate": 9.524135262330098e-06,
+      "loss": 0.414,
+      "step": 203
+    },
+    {
+      "epoch": 0.3688969258589512,
+      "grad_norm": 2.028451919555664,
+      "learning_rate": 9.517745378462417e-06,
+      "loss": 0.496,
+      "step": 204
+    },
+    {
+      "epoch": 0.3707052441229656,
+      "grad_norm": 1.760362982749939,
+      "learning_rate": 9.511315051668287e-06,
+      "loss": 0.3633,
+      "step": 205
+    },
+    {
+      "epoch": 0.37251356238698013,
+      "grad_norm": 1.908535361289978,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 0.3192,
+      "step": 206
+    },
+    {
+      "epoch": 0.3743218806509946,
+      "grad_norm": 1.8707324266433716,
+      "learning_rate": 9.498333299919759e-06,
+      "loss": 0.3455,
+      "step": 207
+    },
+    {
+      "epoch": 0.37613019891500904,
+      "grad_norm": 1.9523345232009888,
+      "learning_rate": 9.491781991178203e-06,
+      "loss": 0.3584,
+      "step": 208
+    },
+    {
+      "epoch": 0.3779385171790235,
+      "grad_norm": 1.8394471406936646,
+      "learning_rate": 9.485190471934845e-06,
+      "loss": 0.3222,
+      "step": 209
+    },
+    {
+      "epoch": 0.379746835443038,
+      "grad_norm": 2.543618679046631,
+      "learning_rate": 9.478558801197065e-06,
+      "loss": 0.362,
+      "step": 210
+    },
+    {
+      "epoch": 0.38155515370705245,
+      "grad_norm": 1.6976804733276367,
+      "learning_rate": 9.471887038331686e-06,
+      "loss": 0.3006,
+      "step": 211
+    },
+    {
+      "epoch": 0.3833634719710669,
+      "grad_norm": 1.6421040296554565,
+      "learning_rate": 9.465175243064428e-06,
+      "loss": 0.3134,
+      "step": 212
+    },
+    {
+      "epoch": 0.38517179023508136,
+      "grad_norm": 1.725056529045105,
+      "learning_rate": 9.458423475479387e-06,
+      "loss": 0.375,
+      "step": 213
+    },
+    {
+      "epoch": 0.38698010849909587,
+      "grad_norm": 1.7434353828430176,
+      "learning_rate": 9.451631796018495e-06,
+      "loss": 0.355,
+      "step": 214
+    },
+    {
+      "epoch": 0.3887884267631103,
+      "grad_norm": 1.7160333395004272,
+      "learning_rate": 9.444800265480968e-06,
+      "loss": 0.2843,
+      "step": 215
+    },
+    {
+      "epoch": 0.39059674502712477,
+      "grad_norm": 1.8513472080230713,
+      "learning_rate": 9.437928945022772e-06,
+      "loss": 0.3075,
+      "step": 216
+    },
+    {
+      "epoch": 0.3924050632911392,
+      "grad_norm": 1.9343299865722656,
+      "learning_rate": 9.431017896156074e-06,
+      "loss": 0.3384,
+      "step": 217
+    },
+    {
+      "epoch": 0.39421338155515373,
+      "grad_norm": 1.7150647640228271,
+      "learning_rate": 9.424067180748692e-06,
+      "loss": 0.3089,
+      "step": 218
+    },
+    {
+      "epoch": 0.3960216998191682,
+      "grad_norm": 1.6371943950653076,
+      "learning_rate": 9.417076861023539e-06,
+      "loss": 0.2921,
+      "step": 219
+    },
+    {
+      "epoch": 0.39783001808318263,
+      "grad_norm": 1.5865685939788818,
+      "learning_rate": 9.410046999558062e-06,
+      "loss": 0.2884,
+      "step": 220
+    },
+    {
+      "epoch": 0.3996383363471971,
+      "grad_norm": 1.7501705884933472,
+      "learning_rate": 9.40297765928369e-06,
+      "loss": 0.3182,
+      "step": 221
+    },
+    {
+      "epoch": 0.4014466546112116,
+      "grad_norm": 1.8444981575012207,
+      "learning_rate": 9.395868903485269e-06,
+      "loss": 0.3304,
+      "step": 222
+    },
+    {
+      "epoch": 0.40325497287522605,
+      "grad_norm": 1.7558904886245728,
+      "learning_rate": 9.388720795800488e-06,
+      "loss": 0.3376,
+      "step": 223
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 2.039940118789673,
+      "learning_rate": 9.381533400219319e-06,
+      "loss": 0.4301,
+      "step": 224
+    },
+    {
+      "epoch": 0.40687160940325495,
+      "grad_norm": 2.1482834815979004,
+      "learning_rate": 9.374306781083437e-06,
+      "loss": 0.4223,
+      "step": 225
+    },
+    {
+      "epoch": 0.40867992766726946,
+      "grad_norm": 1.8534361124038696,
+      "learning_rate": 9.36704100308565e-06,
+      "loss": 0.3514,
+      "step": 226
+    },
+    {
+      "epoch": 0.4104882459312839,
+      "grad_norm": 1.7480493783950806,
+      "learning_rate": 9.359736131269312e-06,
+      "loss": 0.3,
+      "step": 227
+    },
+    {
+      "epoch": 0.41229656419529837,
+      "grad_norm": 1.631418228149414,
+      "learning_rate": 9.352392231027752e-06,
+      "loss": 0.2817,
+      "step": 228
+    },
+    {
+      "epoch": 0.4141048824593128,
+      "grad_norm": 1.6843211650848389,
+      "learning_rate": 9.345009368103677e-06,
+      "loss": 0.3025,
+      "step": 229
+    },
+    {
+      "epoch": 0.4159132007233273,
+      "grad_norm": 2.5229790210723877,
+      "learning_rate": 9.337587608588588e-06,
+      "loss": 0.3805,
+      "step": 230
+    },
+    {
+      "epoch": 0.4177215189873418,
+      "grad_norm": 1.8178796768188477,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.3752,
+      "step": 231
+    },
+    {
+      "epoch": 0.41952983725135623,
+      "grad_norm": 1.8291960954666138,
+      "learning_rate": 9.322627665891807e-06,
+      "loss": 0.3489,
+      "step": 232
+    },
+    {
+      "epoch": 0.4213381555153707,
+      "grad_norm": 1.6301926374435425,
+      "learning_rate": 9.315089616631752e-06,
+      "loss": 0.4083,
+      "step": 233
+    },
+    {
+      "epoch": 0.4231464737793852,
+      "grad_norm": 1.5303218364715576,
+      "learning_rate": 9.307512938622762e-06,
+      "loss": 0.3023,
+      "step": 234
+    },
+    {
+      "epoch": 0.42495479204339964,
+      "grad_norm": 1.685544490814209,
+      "learning_rate": 9.299897699691377e-06,
+      "loss": 0.3413,
+      "step": 235
+    },
+    {
+      "epoch": 0.4267631103074141,
+      "grad_norm": 1.7197544574737549,
+      "learning_rate": 9.292243968009332e-06,
+      "loss": 0.3755,
+      "step": 236
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 1.3292421102523804,
+      "learning_rate": 9.28455181209295e-06,
+      "loss": 0.2676,
+      "step": 237
+    },
+    {
+      "epoch": 0.43037974683544306,
+      "grad_norm": 1.6216092109680176,
+      "learning_rate": 9.276821300802535e-06,
+      "loss": 0.3359,
+      "step": 238
+    },
+    {
+      "epoch": 0.4321880650994575,
+      "grad_norm": 1.9262677431106567,
+      "learning_rate": 9.269052503341737e-06,
+      "loss": 0.3757,
+      "step": 239
+    },
+    {
+      "epoch": 0.43399638336347196,
+      "grad_norm": 1.5813138484954834,
+      "learning_rate": 9.261245489256956e-06,
+      "loss": 0.3469,
+      "step": 240
+    },
+    {
+      "epoch": 0.4358047016274864,
+      "grad_norm": 1.3660542964935303,
+      "learning_rate": 9.253400328436699e-06,
+      "loss": 0.2521,
+      "step": 241
+    },
+    {
+      "epoch": 0.4376130198915009,
+      "grad_norm": 1.645929217338562,
+      "learning_rate": 9.24551709111097e-06,
+      "loss": 0.278,
+      "step": 242
+    },
+    {
+      "epoch": 0.4394213381555154,
+      "grad_norm": 1.5204540491104126,
+      "learning_rate": 9.237595847850628e-06,
+      "loss": 0.2852,
+      "step": 243
+    },
+    {
+      "epoch": 0.4412296564195298,
+      "grad_norm": 1.4475878477096558,
+      "learning_rate": 9.229636669566769e-06,
+      "loss": 0.2785,
+      "step": 244
+    },
+    {
+      "epoch": 0.4430379746835443,
+      "grad_norm": 1.6262820959091187,
+      "learning_rate": 9.221639627510076e-06,
+      "loss": 0.2624,
+      "step": 245
+    },
+    {
+      "epoch": 0.4448462929475588,
+      "grad_norm": 1.6817715167999268,
+      "learning_rate": 9.213604793270196e-06,
+      "loss": 0.303,
+      "step": 246
+    },
+    {
+      "epoch": 0.44665461121157324,
+      "grad_norm": 1.6763249635696411,
+      "learning_rate": 9.205532238775089e-06,
+      "loss": 0.2872,
+      "step": 247
+    },
+    {
+      "epoch": 0.4484629294755877,
+      "grad_norm": 1.5903000831604004,
+      "learning_rate": 9.197422036290386e-06,
+      "loss": 0.2982,
+      "step": 248
+    },
+    {
+      "epoch": 0.45027124773960214,
+      "grad_norm": 1.6391197443008423,
+      "learning_rate": 9.189274258418748e-06,
+      "loss": 0.3317,
+      "step": 249
+    },
+    {
+      "epoch": 0.45207956600361665,
+      "grad_norm": 1.815727949142456,
+      "learning_rate": 9.181088978099203e-06,
+      "loss": 0.3693,
+      "step": 250
+    },
+    {
+      "epoch": 0.4538878842676311,
+      "grad_norm": 1.7187553644180298,
+      "learning_rate": 9.172866268606514e-06,
+      "loss": 0.3699,
+      "step": 251
+    },
+    {
+      "epoch": 0.45569620253164556,
+      "grad_norm": 1.6256039142608643,
+      "learning_rate": 9.164606203550498e-06,
+      "loss": 0.2668,
+      "step": 252
+    },
+    {
+      "epoch": 0.45750452079566006,
+      "grad_norm": 1.582828402519226,
+      "learning_rate": 9.156308856875386e-06,
+      "loss": 0.2835,
+      "step": 253
+    },
+    {
+      "epoch": 0.4593128390596745,
+      "grad_norm": 1.8006536960601807,
+      "learning_rate": 9.147974302859158e-06,
+      "loss": 0.3347,
+      "step": 254
+    },
+    {
+      "epoch": 0.46112115732368897,
+      "grad_norm": 1.8780992031097412,
+      "learning_rate": 9.139602616112864e-06,
+      "loss": 0.367,
+      "step": 255
+    },
+    {
+      "epoch": 0.4629294755877034,
+      "grad_norm": 1.8903968334197998,
+      "learning_rate": 9.131193871579975e-06,
+      "loss": 0.3908,
+      "step": 256
+    },
+    {
+      "epoch": 0.46473779385171793,
+      "grad_norm": 1.6569807529449463,
+      "learning_rate": 9.122748144535704e-06,
+      "loss": 0.3989,
+      "step": 257
+    },
+    {
+      "epoch": 0.4665461121157324,
+      "grad_norm": 1.561207890510559,
+      "learning_rate": 9.114265510586329e-06,
+      "loss": 0.2794,
+      "step": 258
+    },
+    {
+      "epoch": 0.46835443037974683,
+      "grad_norm": 1.6755136251449585,
+      "learning_rate": 9.10574604566852e-06,
+      "loss": 0.2977,
+      "step": 259
+    },
+    {
+      "epoch": 0.4701627486437613,
+      "grad_norm": 1.5138905048370361,
+      "learning_rate": 9.09718982604866e-06,
+      "loss": 0.3073,
+      "step": 260
+    },
+    {
+      "epoch": 0.4719710669077758,
+      "grad_norm": 1.563460111618042,
+      "learning_rate": 9.088596928322158e-06,
+      "loss": 0.3175,
+      "step": 261
+    },
+    {
+      "epoch": 0.47377938517179025,
+      "grad_norm": 1.6276342868804932,
+      "learning_rate": 9.079967429412766e-06,
+      "loss": 0.3257,
+      "step": 262
+    },
+    {
+      "epoch": 0.4755877034358047,
+      "grad_norm": 1.558644413948059,
+      "learning_rate": 9.071301406571893e-06,
+      "loss": 0.3221,
+      "step": 263
+    },
+    {
+      "epoch": 0.47739602169981915,
+      "grad_norm": 1.5441564321517944,
+      "learning_rate": 9.062598937377911e-06,
+      "loss": 0.3554,
+      "step": 264
+    },
+    {
+      "epoch": 0.47920433996383366,
+      "grad_norm": 1.562873125076294,
+      "learning_rate": 9.053860099735455e-06,
+      "loss": 0.2965,
+      "step": 265
+    },
+    {
+      "epoch": 0.4810126582278481,
+      "grad_norm": 1.6871848106384277,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.322,
+      "step": 266
+    },
+    {
+      "epoch": 0.48282097649186256,
+      "grad_norm": 1.4427196979522705,
+      "learning_rate": 9.036273632350839e-06,
+      "loss": 0.2904,
+      "step": 267
+    },
+    {
+      "epoch": 0.484629294755877,
+      "grad_norm": 1.6650079488754272,
+      "learning_rate": 9.027426160043005e-06,
+      "loss": 0.3755,
+      "step": 268
+    },
+    {
+      "epoch": 0.4864376130198915,
+      "grad_norm": 1.5298763513565063,
+      "learning_rate": 9.018542634153944e-06,
+      "loss": 0.2784,
+      "step": 269
+    },
+    {
+      "epoch": 0.488245931283906,
+      "grad_norm": 1.5398807525634766,
+      "learning_rate": 9.00962313420912e-06,
+      "loss": 0.2627,
+      "step": 270
+    },
+    {
+      "epoch": 0.49005424954792043,
+      "grad_norm": 1.4287172555923462,
+      "learning_rate": 9.000667740056033e-06,
+      "loss": 0.2319,
+      "step": 271
+    },
+    {
+      "epoch": 0.4918625678119349,
+      "grad_norm": 1.5109469890594482,
+      "learning_rate": 8.991676531863507e-06,
+      "loss": 0.2677,
+      "step": 272
+    },
+    {
+      "epoch": 0.4936708860759494,
+      "grad_norm": 1.773703932762146,
+      "learning_rate": 8.982649590120982e-06,
+      "loss": 0.38,
+      "step": 273
+    },
+    {
+      "epoch": 0.49547920433996384,
+      "grad_norm": 1.7830049991607666,
+      "learning_rate": 8.973586995637778e-06,
+      "loss": 0.3287,
+      "step": 274
+    },
+    {
+      "epoch": 0.4972875226039783,
+      "grad_norm": 1.6185346841812134,
+      "learning_rate": 8.964488829542377e-06,
+      "loss": 0.33,
+      "step": 275
+    },
+    {
+      "epoch": 0.49909584086799275,
+      "grad_norm": 1.6162431240081787,
+      "learning_rate": 8.955355173281709e-06,
+      "loss": 0.328,
+      "step": 276
+    },
+    {
+      "epoch": 0.5009041591320073,
+      "grad_norm": 1.9096221923828125,
+      "learning_rate": 8.946186108620397e-06,
+      "loss": 0.3238,
+      "step": 277
+    },
+    {
+      "epoch": 0.5027124773960217,
+      "grad_norm": 1.4491026401519775,
+      "learning_rate": 8.936981717640061e-06,
+      "loss": 0.3222,
+      "step": 278
+    },
+    {
+      "epoch": 0.5045207956600362,
+      "grad_norm": 1.4854817390441895,
+      "learning_rate": 8.927742082738542e-06,
+      "loss": 0.2992,
+      "step": 279
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 1.7579244375228882,
+      "learning_rate": 8.9184672866292e-06,
+      "loss": 0.3724,
+      "step": 280
+    },
+    {
+      "epoch": 0.5081374321880651,
+      "grad_norm": 1.8710993528366089,
+      "learning_rate": 8.90915741234015e-06,
+      "loss": 0.3663,
+      "step": 281
+    },
+    {
+      "epoch": 0.5099457504520796,
+      "grad_norm": 1.777894377708435,
+      "learning_rate": 8.899812543213532e-06,
+      "loss": 0.4384,
+      "step": 282
+    },
+    {
+      "epoch": 0.5117540687160941,
+      "grad_norm": 1.5931551456451416,
+      "learning_rate": 8.890432762904757e-06,
+      "loss": 0.3171,
+      "step": 283
+    },
+    {
+      "epoch": 0.5135623869801085,
+      "grad_norm": 1.4698400497436523,
+      "learning_rate": 8.881018155381766e-06,
+      "loss": 0.3111,
+      "step": 284
+    },
+    {
+      "epoch": 0.515370705244123,
+      "grad_norm": 1.8043538331985474,
+      "learning_rate": 8.871568804924269e-06,
+      "loss": 0.4004,
+      "step": 285
+    },
+    {
+      "epoch": 0.5171790235081374,
+      "grad_norm": 1.707382321357727,
+      "learning_rate": 8.862084796122998e-06,
+      "loss": 0.3684,
+      "step": 286
+    },
+    {
+      "epoch": 0.5189873417721519,
+      "grad_norm": 1.702929973602295,
+      "learning_rate": 8.852566213878947e-06,
+      "loss": 0.3919,
+      "step": 287
+    },
+    {
+      "epoch": 0.5207956600361664,
+      "grad_norm": 1.3536096811294556,
+      "learning_rate": 8.84301314340261e-06,
+      "loss": 0.2297,
+      "step": 288
+    },
+    {
+      "epoch": 0.5226039783001808,
+      "grad_norm": 1.5708818435668945,
+      "learning_rate": 8.833425670213223e-06,
+      "loss": 0.3132,
+      "step": 289
+    },
+    {
+      "epoch": 0.5244122965641953,
+      "grad_norm": 1.352744460105896,
+      "learning_rate": 8.823803880137993e-06,
+      "loss": 0.2602,
+      "step": 290
+    },
+    {
+      "epoch": 0.5262206148282098,
+      "grad_norm": 1.6007314920425415,
+      "learning_rate": 8.814147859311333e-06,
+      "loss": 0.3156,
+      "step": 291
+    },
+    {
+      "epoch": 0.5280289330922242,
+      "grad_norm": 1.6960111856460571,
+      "learning_rate": 8.804457694174093e-06,
+      "loss": 0.4026,
+      "step": 292
+    },
+    {
+      "epoch": 0.5298372513562387,
+      "grad_norm": 1.5450199842453003,
+      "learning_rate": 8.794733471472778e-06,
+      "loss": 0.3557,
+      "step": 293
+    },
+    {
+      "epoch": 0.5316455696202531,
+      "grad_norm": 1.5667040348052979,
+      "learning_rate": 8.784975278258783e-06,
+      "loss": 0.3074,
+      "step": 294
+    },
+    {
+      "epoch": 0.5334538878842676,
+      "grad_norm": 1.5064713954925537,
+      "learning_rate": 8.775183201887603e-06,
+      "loss": 0.3242,
+      "step": 295
+    },
+    {
+      "epoch": 0.5352622061482821,
+      "grad_norm": 1.6140296459197998,
+      "learning_rate": 8.765357330018056e-06,
+      "loss": 0.2771,
+      "step": 296
+    },
+    {
+      "epoch": 0.5370705244122965,
+      "grad_norm": 1.8816146850585938,
+      "learning_rate": 8.755497750611498e-06,
+      "loss": 0.3906,
+      "step": 297
+    },
+    {
+      "epoch": 0.538878842676311,
+      "grad_norm": 1.6953233480453491,
+      "learning_rate": 8.745604551931042e-06,
+      "loss": 0.3338,
+      "step": 298
+    },
+    {
+      "epoch": 0.5406871609403255,
+      "grad_norm": 1.7361221313476562,
+      "learning_rate": 8.73567782254075e-06,
+      "loss": 0.3105,
+      "step": 299
+    },
+    {
+      "epoch": 0.5424954792043399,
+      "grad_norm": 1.6798964738845825,
+      "learning_rate": 8.725717651304856e-06,
+      "loss": 0.3645,
+      "step": 300
+    },
+    {
+      "epoch": 0.5443037974683544,
+      "grad_norm": 1.4772828817367554,
+      "learning_rate": 8.715724127386971e-06,
+      "loss": 0.2823,
+      "step": 301
+    },
+    {
+      "epoch": 0.546112115732369,
+      "grad_norm": 1.754899263381958,
+      "learning_rate": 8.705697340249275e-06,
+      "loss": 0.3611,
+      "step": 302
+    },
+    {
+      "epoch": 0.5479204339963833,
+      "grad_norm": 1.6467196941375732,
+      "learning_rate": 8.695637379651719e-06,
+      "loss": 0.2694,
+      "step": 303
+    },
+    {
+      "epoch": 0.5497287522603979,
+      "grad_norm": 1.663810133934021,
+      "learning_rate": 8.685544335651226e-06,
+      "loss": 0.2803,
+      "step": 304
+    },
+    {
+      "epoch": 0.5515370705244123,
+      "grad_norm": 1.6183619499206543,
+      "learning_rate": 8.675418298600884e-06,
+      "loss": 0.3227,
+      "step": 305
+    },
+    {
+      "epoch": 0.5533453887884268,
+      "grad_norm": 1.7311010360717773,
+      "learning_rate": 8.665259359149132e-06,
+      "loss": 0.4449,
+      "step": 306
+    },
+    {
+      "epoch": 0.5551537070524413,
+      "grad_norm": 2.268200159072876,
+      "learning_rate": 8.655067608238953e-06,
+      "loss": 0.298,
+      "step": 307
+    },
+    {
+      "epoch": 0.5569620253164557,
+      "grad_norm": 1.66535222530365,
+      "learning_rate": 8.644843137107058e-06,
+      "loss": 0.3834,
+      "step": 308
+    },
+    {
+      "epoch": 0.5587703435804702,
+      "grad_norm": 1.6682361364364624,
+      "learning_rate": 8.634586037283072e-06,
+      "loss": 0.3886,
+      "step": 309
+    },
+    {
+      "epoch": 0.5605786618444847,
+      "grad_norm": 1.3878757953643799,
+      "learning_rate": 8.62429640058871e-06,
+      "loss": 0.2728,
+      "step": 310
+    },
+    {
+      "epoch": 0.5623869801084991,
+      "grad_norm": 1.6548659801483154,
+      "learning_rate": 8.613974319136959e-06,
+      "loss": 0.3578,
+      "step": 311
+    },
+    {
+      "epoch": 0.5641952983725136,
+      "grad_norm": 1.6845568418502808,
+      "learning_rate": 8.603619885331251e-06,
+      "loss": 0.3421,
+      "step": 312
+    },
+    {
+      "epoch": 0.566003616636528,
+      "grad_norm": 1.4604007005691528,
+      "learning_rate": 8.593233191864638e-06,
+      "loss": 0.3022,
+      "step": 313
+    },
+    {
+      "epoch": 0.5678119349005425,
+      "grad_norm": 1.6256940364837646,
+      "learning_rate": 8.582814331718961e-06,
+      "loss": 0.3088,
+      "step": 314
+    },
+    {
+      "epoch": 0.569620253164557,
+      "grad_norm": 1.6187587976455688,
+      "learning_rate": 8.572363398164017e-06,
+      "loss": 0.28,
+      "step": 315
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 10.702666282653809,
+      "learning_rate": 8.561880484756726e-06,
+      "loss": 0.3774,
+      "step": 316
+    },
+    {
+      "epoch": 0.5732368896925859,
+      "grad_norm": 1.3536791801452637,
+      "learning_rate": 8.551365685340285e-06,
+      "loss": 0.2602,
+      "step": 317
+    },
+    {
+      "epoch": 0.5750452079566004,
+      "grad_norm": 1.6421338319778442,
+      "learning_rate": 8.540819094043349e-06,
+      "loss": 0.3563,
+      "step": 318
+    },
+    {
+      "epoch": 0.5768535262206148,
+      "grad_norm": 1.4128696918487549,
+      "learning_rate": 8.530240805279159e-06,
+      "loss": 0.2518,
+      "step": 319
+    },
+    {
+      "epoch": 0.5786618444846293,
+      "grad_norm": 1.573652982711792,
+      "learning_rate": 8.519630913744726e-06,
+      "loss": 0.3026,
+      "step": 320
+    },
+    {
+      "epoch": 0.5804701627486437,
+      "grad_norm": 1.4500043392181396,
+      "learning_rate": 8.508989514419959e-06,
+      "loss": 0.2553,
+      "step": 321
+    },
+    {
+      "epoch": 0.5822784810126582,
+      "grad_norm": 1.375697374343872,
+      "learning_rate": 8.498316702566828e-06,
+      "loss": 0.2347,
+      "step": 322
+    },
+    {
+      "epoch": 0.5840867992766727,
+      "grad_norm": 1.601431965827942,
+      "learning_rate": 8.487612573728513e-06,
+      "loss": 0.2794,
+      "step": 323
+    },
+    {
+      "epoch": 0.5858951175406871,
+      "grad_norm": 1.7059786319732666,
+      "learning_rate": 8.476877223728539e-06,
+      "loss": 0.358,
+      "step": 324
+    },
+    {
+      "epoch": 0.5877034358047016,
+      "grad_norm": 1.8292579650878906,
+      "learning_rate": 8.466110748669926e-06,
+      "loss": 0.4206,
+      "step": 325
+    },
+    {
+      "epoch": 0.5895117540687161,
+      "grad_norm": 1.5839455127716064,
+      "learning_rate": 8.455313244934324e-06,
+      "loss": 0.3653,
+      "step": 326
+    },
+    {
+      "epoch": 0.5913200723327305,
+      "grad_norm": 1.629156231880188,
+      "learning_rate": 8.444484809181155e-06,
+      "loss": 0.3159,
+      "step": 327
+    },
+    {
+      "epoch": 0.593128390596745,
+      "grad_norm": 1.526873230934143,
+      "learning_rate": 8.433625538346742e-06,
+      "loss": 0.3332,
+      "step": 328
+    },
+    {
+      "epoch": 0.5949367088607594,
+      "grad_norm": 1.5268447399139404,
+      "learning_rate": 8.422735529643445e-06,
+      "loss": 0.2919,
+      "step": 329
+    },
+    {
+      "epoch": 0.596745027124774,
+      "grad_norm": 1.5196579694747925,
+      "learning_rate": 8.41181488055879e-06,
+      "loss": 0.3301,
+      "step": 330
+    },
+    {
+      "epoch": 0.5985533453887885,
+      "grad_norm": 1.5749566555023193,
+      "learning_rate": 8.400863688854598e-06,
+      "loss": 0.3386,
+      "step": 331
+    },
+    {
+      "epoch": 0.6003616636528029,
+      "grad_norm": 1.5672210454940796,
+      "learning_rate": 8.389882052566106e-06,
+      "loss": 0.3103,
+      "step": 332
+    },
+    {
+      "epoch": 0.6021699819168174,
+      "grad_norm": 1.8071767091751099,
+      "learning_rate": 8.37887007000109e-06,
+      "loss": 0.3571,
+      "step": 333
+    },
+    {
+      "epoch": 0.6039783001808319,
+      "grad_norm": 1.4983811378479004,
+      "learning_rate": 8.36782783973899e-06,
+      "loss": 0.2772,
+      "step": 334
+    },
+    {
+      "epoch": 0.6057866184448463,
+      "grad_norm": 1.5349783897399902,
+      "learning_rate": 8.35675546063002e-06,
+      "loss": 0.3161,
+      "step": 335
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 1.4982757568359375,
+      "learning_rate": 8.345653031794292e-06,
+      "loss": 0.3483,
+      "step": 336
+    },
+    {
+      "epoch": 0.6094032549728752,
+      "grad_norm": 1.4851738214492798,
+      "learning_rate": 8.334520652620918e-06,
+      "loss": 0.2911,
+      "step": 337
+    },
+    {
+      "epoch": 0.6112115732368897,
+      "grad_norm": 1.4346152544021606,
+      "learning_rate": 8.32335842276713e-06,
+      "loss": 0.2647,
+      "step": 338
+    },
+    {
+      "epoch": 0.6130198915009042,
+      "grad_norm": 1.5562641620635986,
+      "learning_rate": 8.31216644215738e-06,
+      "loss": 0.3653,
+      "step": 339
+    },
+    {
+      "epoch": 0.6148282097649186,
+      "grad_norm": 1.502779245376587,
+      "learning_rate": 8.300944810982452e-06,
+      "loss": 0.3021,
+      "step": 340
+    },
+    {
+      "epoch": 0.6166365280289331,
+      "grad_norm": 1.447108507156372,
+      "learning_rate": 8.289693629698564e-06,
+      "loss": 0.2713,
+      "step": 341
+    },
+    {
+      "epoch": 0.6184448462929476,
+      "grad_norm": 1.3462806940078735,
+      "learning_rate": 8.278412999026462e-06,
+      "loss": 0.2314,
+      "step": 342
+    },
+    {
+      "epoch": 0.620253164556962,
+      "grad_norm": 1.360318660736084,
+      "learning_rate": 8.267103019950529e-06,
+      "loss": 0.2538,
+      "step": 343
+    },
+    {
+      "epoch": 0.6220614828209765,
+      "grad_norm": 1.5165733098983765,
+      "learning_rate": 8.255763793717868e-06,
+      "loss": 0.3008,
+      "step": 344
+    },
+    {
+      "epoch": 0.6238698010849909,
+      "grad_norm": 1.3933101892471313,
+      "learning_rate": 8.244395421837412e-06,
+      "loss": 0.2313,
+      "step": 345
+    },
+    {
+      "epoch": 0.6256781193490054,
+      "grad_norm": 1.4917360544204712,
+      "learning_rate": 8.232998006078998e-06,
+      "loss": 0.3104,
+      "step": 346
+    },
+    {
+      "epoch": 0.6274864376130199,
+      "grad_norm": 1.4749935865402222,
+      "learning_rate": 8.221571648472473e-06,
+      "loss": 0.2528,
+      "step": 347
+    },
+    {
+      "epoch": 0.6292947558770343,
+      "grad_norm": 1.531565546989441,
+      "learning_rate": 8.210116451306762e-06,
+      "loss": 0.2578,
+      "step": 348
+    },
+    {
+      "epoch": 0.6311030741410488,
+      "grad_norm": 1.7813491821289062,
+      "learning_rate": 8.198632517128968e-06,
+      "loss": 0.358,
+      "step": 349
+    },
+    {
+      "epoch": 0.6329113924050633,
+      "grad_norm": 1.770026445388794,
+      "learning_rate": 8.18711994874345e-06,
+      "loss": 0.3252,
+      "step": 350
+    },
+    {
+      "epoch": 0.6347197106690777,
+      "grad_norm": 1.589426040649414,
+      "learning_rate": 8.175578849210894e-06,
+      "loss": 0.3258,
+      "step": 351
+    },
+    {
+      "epoch": 0.6365280289330922,
+      "grad_norm": 1.7726151943206787,
+      "learning_rate": 8.164009321847405e-06,
+      "loss": 0.3183,
+      "step": 352
+    },
+    {
+      "epoch": 0.6383363471971067,
+      "grad_norm": 1.4787203073501587,
+      "learning_rate": 8.15241147022357e-06,
+      "loss": 0.2637,
+      "step": 353
+    },
+    {
+      "epoch": 0.6401446654611211,
+      "grad_norm": 1.7074470520019531,
+      "learning_rate": 8.140785398163535e-06,
+      "loss": 0.3754,
+      "step": 354
+    },
+    {
+      "epoch": 0.6419529837251357,
+      "grad_norm": 1.7455577850341797,
+      "learning_rate": 8.129131209744075e-06,
+      "loss": 0.3453,
+      "step": 355
+    },
+    {
+      "epoch": 0.64376130198915,
+      "grad_norm": 1.5032719373703003,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.301,
+      "step": 356
+    },
+    {
+      "epoch": 0.6455696202531646,
+      "grad_norm": 1.5343469381332397,
+      "learning_rate": 8.105738901391553e-06,
+      "loss": 0.3515,
+      "step": 357
+    },
+    {
+      "epoch": 0.6473779385171791,
+      "grad_norm": 1.6261870861053467,
+      "learning_rate": 8.094000990866795e-06,
+      "loss": 0.2817,
+      "step": 358
+    },
+    {
+      "epoch": 0.6491862567811935,
+      "grad_norm": 1.3367079496383667,
+      "learning_rate": 8.08223538279735e-06,
+      "loss": 0.2207,
+      "step": 359
+    },
+    {
+      "epoch": 0.650994575045208,
+      "grad_norm": 1.5360643863677979,
+      "learning_rate": 8.070442182509127e-06,
+      "loss": 0.31,
+      "step": 360
+    },
+    {
+      "epoch": 0.6528028933092225,
+      "grad_norm": 1.673229455947876,
+      "learning_rate": 8.058621495575032e-06,
+      "loss": 0.3752,
+      "step": 361
+    },
+    {
+      "epoch": 0.6546112115732369,
+      "grad_norm": 1.747746467590332,
+      "learning_rate": 8.046773427814043e-06,
+      "loss": 0.373,
+      "step": 362
+    },
+    {
+      "epoch": 0.6564195298372514,
+      "grad_norm": 1.3916759490966797,
+      "learning_rate": 8.034898085290239e-06,
+      "loss": 0.2682,
+      "step": 363
+    },
+    {
+      "epoch": 0.6582278481012658,
+      "grad_norm": 1.4220057725906372,
+      "learning_rate": 8.022995574311876e-06,
+      "loss": 0.2696,
+      "step": 364
+    },
+    {
+      "epoch": 0.6600361663652803,
+      "grad_norm": 1.483280062675476,
+      "learning_rate": 8.011066001430412e-06,
+      "loss": 0.3125,
+      "step": 365
+    },
+    {
+      "epoch": 0.6618444846292948,
+      "grad_norm": 1.436482310295105,
+      "learning_rate": 7.99910947343957e-06,
+      "loss": 0.2781,
+      "step": 366
+    },
+    {
+      "epoch": 0.6636528028933092,
+      "grad_norm": 1.4785102605819702,
+      "learning_rate": 7.987126097374372e-06,
+      "loss": 0.2998,
+      "step": 367
+    },
+    {
+      "epoch": 0.6654611211573237,
+      "grad_norm": 1.388868808746338,
+      "learning_rate": 7.975115980510187e-06,
+      "loss": 0.2345,
+      "step": 368
+    },
+    {
+      "epoch": 0.6672694394213382,
+      "grad_norm": 1.443105936050415,
+      "learning_rate": 7.963079230361765e-06,
+      "loss": 0.2669,
+      "step": 369
+    },
+    {
+      "epoch": 0.6690777576853526,
+      "grad_norm": 1.4718849658966064,
+      "learning_rate": 7.951015954682281e-06,
+      "loss": 0.293,
+      "step": 370
+    },
+    {
+      "epoch": 0.6708860759493671,
+      "grad_norm": 1.5544575452804565,
+      "learning_rate": 7.938926261462366e-06,
+      "loss": 0.2738,
+      "step": 371
+    },
+    {
+      "epoch": 0.6726943942133815,
+      "grad_norm": 1.6500585079193115,
+      "learning_rate": 7.926810258929138e-06,
+      "loss": 0.3018,
+      "step": 372
+    },
+    {
+      "epoch": 0.674502712477396,
+      "grad_norm": 1.8269141912460327,
+      "learning_rate": 7.914668055545242e-06,
+      "loss": 0.3061,
+      "step": 373
+    },
+    {
+      "epoch": 0.6763110307414105,
+      "grad_norm": 1.6963152885437012,
+      "learning_rate": 7.902499760007867e-06,
+      "loss": 0.3793,
+      "step": 374
+    },
+    {
+      "epoch": 0.6781193490054249,
+      "grad_norm": 1.9636969566345215,
+      "learning_rate": 7.890305481247786e-06,
+      "loss": 0.3418,
+      "step": 375
+    },
+    {
+      "epoch": 0.6799276672694394,
+      "grad_norm": 1.7126883268356323,
+      "learning_rate": 7.87808532842837e-06,
+      "loss": 0.3795,
+      "step": 376
+    },
+    {
+      "epoch": 0.6817359855334539,
+      "grad_norm": 1.4325624704360962,
+      "learning_rate": 7.865839410944613e-06,
+      "loss": 0.2684,
+      "step": 377
+    },
+    {
+      "epoch": 0.6835443037974683,
+      "grad_norm": 1.4512115716934204,
+      "learning_rate": 7.85356783842216e-06,
+      "loss": 0.2926,
+      "step": 378
+    },
+    {
+      "epoch": 0.6853526220614828,
+      "grad_norm": 1.6614792346954346,
+      "learning_rate": 7.841270720716318e-06,
+      "loss": 0.289,
+      "step": 379
+    },
+    {
+      "epoch": 0.6871609403254972,
+      "grad_norm": 1.7151885032653809,
+      "learning_rate": 7.828948167911073e-06,
+      "loss": 0.3802,
+      "step": 380
+    },
+    {
+      "epoch": 0.6889692585895117,
+      "grad_norm": 1.5852060317993164,
+      "learning_rate": 7.81660029031811e-06,
+      "loss": 0.3071,
+      "step": 381
+    },
+    {
+      "epoch": 0.6907775768535263,
+      "grad_norm": 1.5491100549697876,
+      "learning_rate": 7.804227198475823e-06,
+      "loss": 0.3158,
+      "step": 382
+    },
+    {
+      "epoch": 0.6925858951175407,
+      "grad_norm": 1.4715349674224854,
+      "learning_rate": 7.791829003148313e-06,
+      "loss": 0.2502,
+      "step": 383
+    },
+    {
+      "epoch": 0.6943942133815552,
+      "grad_norm": 1.4923640489578247,
+      "learning_rate": 7.779405815324424e-06,
+      "loss": 0.2891,
+      "step": 384
+    },
+    {
+      "epoch": 0.6962025316455697,
+      "grad_norm": 1.5427296161651611,
+      "learning_rate": 7.76695774621672e-06,
+      "loss": 0.2912,
+      "step": 385
+    },
+    {
+      "epoch": 0.6980108499095841,
+      "grad_norm": 1.5590311288833618,
+      "learning_rate": 7.754484907260513e-06,
+      "loss": 0.2663,
+      "step": 386
+    },
+    {
+      "epoch": 0.6998191681735986,
+      "grad_norm": 1.5675221681594849,
+      "learning_rate": 7.741987410112847e-06,
+      "loss": 0.267,
+      "step": 387
+    },
+    {
+      "epoch": 0.701627486437613,
+      "grad_norm": 1.506213903427124,
+      "learning_rate": 7.72946536665151e-06,
+      "loss": 0.2658,
+      "step": 388
+    },
+    {
+      "epoch": 0.7034358047016275,
+      "grad_norm": 1.3946253061294556,
+      "learning_rate": 7.716918888974029e-06,
+      "loss": 0.2686,
+      "step": 389
+    },
+    {
+      "epoch": 0.705244122965642,
+      "grad_norm": 1.6965469121932983,
+      "learning_rate": 7.704348089396667e-06,
+      "loss": 0.2232,
+      "step": 390
+    },
+    {
+      "epoch": 0.7070524412296564,
+      "grad_norm": 1.5790448188781738,
+      "learning_rate": 7.691753080453413e-06,
+      "loss": 0.2911,
+      "step": 391
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 1.4819265604019165,
+      "learning_rate": 7.679133974894984e-06,
+      "loss": 0.2719,
+      "step": 392
+    },
+    {
+      "epoch": 0.7106690777576854,
+      "grad_norm": 1.4775956869125366,
+      "learning_rate": 7.666490885687807e-06,
+      "loss": 0.2602,
+      "step": 393
+    },
+    {
+      "epoch": 0.7124773960216998,
+      "grad_norm": 1.3449031114578247,
+      "learning_rate": 7.653823926013016e-06,
+      "loss": 0.2391,
+      "step": 394
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 1.365142583847046,
+      "learning_rate": 7.641133209265423e-06,
+      "loss": 0.2492,
+      "step": 395
+    },
+    {
+      "epoch": 0.7160940325497287,
+      "grad_norm": 1.6121798753738403,
+      "learning_rate": 7.628418849052523e-06,
+      "loss": 0.3359,
+      "step": 396
+    },
+    {
+      "epoch": 0.7179023508137432,
+      "grad_norm": 1.517443299293518,
+      "learning_rate": 7.615680959193469e-06,
+      "loss": 0.2862,
+      "step": 397
+    },
+    {
+      "epoch": 0.7197106690777577,
+      "grad_norm": 1.458894968032837,
+      "learning_rate": 7.602919653718044e-06,
+      "loss": 0.283,
+      "step": 398
+    },
+    {
+      "epoch": 0.7215189873417721,
+      "grad_norm": 1.5772570371627808,
+      "learning_rate": 7.590135046865652e-06,
+      "loss": 0.3005,
+      "step": 399
+    },
+    {
+      "epoch": 0.7233273056057866,
+      "grad_norm": 1.8053752183914185,
+      "learning_rate": 7.577327253084292e-06,
+      "loss": 0.3303,
+      "step": 400
+    },
+    {
+      "epoch": 0.7251356238698011,
+      "grad_norm": 1.523858904838562,
+      "learning_rate": 7.564496387029532e-06,
+      "loss": 0.245,
+      "step": 401
+    },
+    {
+      "epoch": 0.7269439421338155,
+      "grad_norm": 1.5387307405471802,
+      "learning_rate": 7.551642563563481e-06,
+      "loss": 0.2924,
+      "step": 402
+    },
+    {
+      "epoch": 0.72875226039783,
+      "grad_norm": 1.4865695238113403,
+      "learning_rate": 7.5387658977537695e-06,
+      "loss": 0.2731,
+      "step": 403
+    },
+    {
+      "epoch": 0.7305605786618445,
+      "grad_norm": 1.6363080739974976,
+      "learning_rate": 7.5258665048725065e-06,
+      "loss": 0.3144,
+      "step": 404
+    },
+    {
+      "epoch": 0.7323688969258589,
+      "grad_norm": 1.5519919395446777,
+      "learning_rate": 7.512944500395255e-06,
+      "loss": 0.3401,
+      "step": 405
+    },
+    {
+      "epoch": 0.7341772151898734,
+      "grad_norm": 1.5395323038101196,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.2855,
+      "step": 406
+    },
+    {
+      "epoch": 0.7359855334538878,
+      "grad_norm": 1.7250220775604248,
+      "learning_rate": 7.48703311956611e-06,
+      "loss": 0.3112,
+      "step": 407
+    },
+    {
+      "epoch": 0.7377938517179023,
+      "grad_norm": 1.4305680990219116,
+      "learning_rate": 7.4740439751732994e-06,
+      "loss": 0.3191,
+      "step": 408
+    },
+    {
+      "epoch": 0.7396021699819169,
+      "grad_norm": 1.5562596321105957,
+      "learning_rate": 7.461032683100587e-06,
+      "loss": 0.3067,
+      "step": 409
+    },
+    {
+      "epoch": 0.7414104882459313,
+      "grad_norm": 1.390204906463623,
+      "learning_rate": 7.447999359825263e-06,
+      "loss": 0.2646,
+      "step": 410
+    },
+    {
+      "epoch": 0.7432188065099458,
+      "grad_norm": 1.5041637420654297,
+      "learning_rate": 7.434944122021837e-06,
+      "loss": 0.2877,
+      "step": 411
+    },
+    {
+      "epoch": 0.7450271247739603,
+      "grad_norm": 1.5278867483139038,
+      "learning_rate": 7.421867086561001e-06,
+      "loss": 0.2353,
+      "step": 412
+    },
+    {
+      "epoch": 0.7468354430379747,
+      "grad_norm": 1.4603511095046997,
+      "learning_rate": 7.408768370508577e-06,
+      "loss": 0.3148,
+      "step": 413
+    },
+    {
+      "epoch": 0.7486437613019892,
+      "grad_norm": 1.4478647708892822,
+      "learning_rate": 7.395648091124476e-06,
+      "loss": 0.2963,
+      "step": 414
+    },
+    {
+      "epoch": 0.7504520795660036,
+      "grad_norm": 1.669955849647522,
+      "learning_rate": 7.382506365861639e-06,
+      "loss": 0.2516,
+      "step": 415
+    },
+    {
+      "epoch": 0.7522603978300181,
+      "grad_norm": 1.3429478406906128,
+      "learning_rate": 7.369343312364994e-06,
+      "loss": 0.2837,
+      "step": 416
+    },
+    {
+      "epoch": 0.7540687160940326,
+      "grad_norm": 1.557304859161377,
+      "learning_rate": 7.356159048470402e-06,
+      "loss": 0.3204,
+      "step": 417
+    },
+    {
+      "epoch": 0.755877034358047,
+      "grad_norm": 1.5032353401184082,
+      "learning_rate": 7.342953692203594e-06,
+      "loss": 0.2217,
+      "step": 418
+    },
+    {
+      "epoch": 0.7576853526220615,
+      "grad_norm": 1.5901117324829102,
+      "learning_rate": 7.3297273617791246e-06,
+      "loss": 0.2959,
+      "step": 419
+    },
+    {
+      "epoch": 0.759493670886076,
+      "grad_norm": 1.5197011232376099,
+      "learning_rate": 7.31648017559931e-06,
+      "loss": 0.2872,
+      "step": 420
+    },
+    {
+      "epoch": 0.7613019891500904,
+      "grad_norm": 1.4186420440673828,
+      "learning_rate": 7.303212252253163e-06,
+      "loss": 0.2769,
+      "step": 421
+    },
+    {
+      "epoch": 0.7631103074141049,
+      "grad_norm": 1.5729889869689941,
+      "learning_rate": 7.289923710515338e-06,
+      "loss": 0.319,
+      "step": 422
+    },
+    {
+      "epoch": 0.7649186256781193,
+      "grad_norm": 1.560356616973877,
+      "learning_rate": 7.276614669345069e-06,
+      "loss": 0.3268,
+      "step": 423
+    },
+    {
+      "epoch": 0.7667269439421338,
+      "grad_norm": 1.403883695602417,
+      "learning_rate": 7.263285247885097e-06,
+      "loss": 0.2321,
+      "step": 424
+    },
+    {
+      "epoch": 0.7685352622061483,
+      "grad_norm": 1.9532519578933716,
+      "learning_rate": 7.249935565460606e-06,
+      "loss": 0.4592,
+      "step": 425
+    },
+    {
+      "epoch": 0.7703435804701627,
+      "grad_norm": 1.6798689365386963,
+      "learning_rate": 7.236565741578163e-06,
+      "loss": 0.3231,
+      "step": 426
+    },
+    {
+      "epoch": 0.7721518987341772,
+      "grad_norm": 1.5938643217086792,
+      "learning_rate": 7.223175895924638e-06,
+      "loss": 0.2551,
+      "step": 427
+    },
+    {
+      "epoch": 0.7739602169981917,
+      "grad_norm": 1.386902093887329,
+      "learning_rate": 7.2097661483661355e-06,
+      "loss": 0.3029,
+      "step": 428
+    },
+    {
+      "epoch": 0.7757685352622061,
+      "grad_norm": 1.6395827531814575,
+      "learning_rate": 7.19633661894692e-06,
+      "loss": 0.3045,
+      "step": 429
+    },
+    {
+      "epoch": 0.7775768535262206,
+      "grad_norm": 1.7160226106643677,
+      "learning_rate": 7.182887427888351e-06,
+      "loss": 0.4528,
+      "step": 430
+    },
+    {
+      "epoch": 0.779385171790235,
+      "grad_norm": 1.6600942611694336,
+      "learning_rate": 7.169418695587791e-06,
+      "loss": 0.3538,
+      "step": 431
+    },
+    {
+      "epoch": 0.7811934900542495,
+      "grad_norm": 1.5458256006240845,
+      "learning_rate": 7.155930542617543e-06,
+      "loss": 0.3465,
+      "step": 432
+    },
+    {
+      "epoch": 0.783001808318264,
+      "grad_norm": 1.226886510848999,
+      "learning_rate": 7.142423089723758e-06,
+      "loss": 0.2268,
+      "step": 433
+    },
+    {
+      "epoch": 0.7848101265822784,
+      "grad_norm": 1.5652273893356323,
+      "learning_rate": 7.128896457825364e-06,
+      "loss": 0.3008,
+      "step": 434
+    },
+    {
+      "epoch": 0.786618444846293,
+      "grad_norm": 1.3933650255203247,
+      "learning_rate": 7.11535076801298e-06,
+      "loss": 0.2755,
+      "step": 435
+    },
+    {
+      "epoch": 0.7884267631103075,
+      "grad_norm": 1.3730487823486328,
+      "learning_rate": 7.101786141547829e-06,
+      "loss": 0.2982,
+      "step": 436
+    },
+    {
+      "epoch": 0.7902350813743219,
+      "grad_norm": 1.4305229187011719,
+      "learning_rate": 7.088202699860656e-06,
+      "loss": 0.3213,
+      "step": 437
+    },
+    {
+      "epoch": 0.7920433996383364,
+      "grad_norm": 1.4953081607818604,
+      "learning_rate": 7.074600564550643e-06,
+      "loss": 0.2981,
+      "step": 438
+    },
+    {
+      "epoch": 0.7938517179023508,
+      "grad_norm": 1.3501685857772827,
+      "learning_rate": 7.060979857384316e-06,
+      "loss": 0.2656,
+      "step": 439
+    },
+    {
+      "epoch": 0.7956600361663653,
+      "grad_norm": 1.3591465950012207,
+      "learning_rate": 7.047340700294454e-06,
+      "loss": 0.2656,
+      "step": 440
+    },
+    {
+      "epoch": 0.7974683544303798,
+      "grad_norm": 1.309065818786621,
+      "learning_rate": 7.033683215379002e-06,
+      "loss": 0.2299,
+      "step": 441
+    },
+    {
+      "epoch": 0.7992766726943942,
+      "grad_norm": 1.33230459690094,
+      "learning_rate": 7.020007524899976e-06,
+      "loss": 0.2238,
+      "step": 442
+    },
+    {
+      "epoch": 0.8010849909584087,
+      "grad_norm": 1.5571876764297485,
+      "learning_rate": 7.006313751282372e-06,
+      "loss": 0.2758,
+      "step": 443
+    },
+    {
+      "epoch": 0.8028933092224232,
+      "grad_norm": 1.3781603574752808,
+      "learning_rate": 6.992602017113058e-06,
+      "loss": 0.2044,
+      "step": 444
+    },
+    {
+      "epoch": 0.8047016274864376,
+      "grad_norm": 1.4426606893539429,
+      "learning_rate": 6.978872445139695e-06,
+      "loss": 0.2365,
+      "step": 445
+    },
+    {
+      "epoch": 0.8065099457504521,
+      "grad_norm": 1.4027334451675415,
+      "learning_rate": 6.965125158269619e-06,
+      "loss": 0.2407,
+      "step": 446
+    },
+    {
+      "epoch": 0.8083182640144665,
+      "grad_norm": 1.5120912790298462,
+      "learning_rate": 6.951360279568758e-06,
+      "loss": 0.2799,
+      "step": 447
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 1.738258719444275,
+      "learning_rate": 6.9375779322605154e-06,
+      "loss": 0.3353,
+      "step": 448
+    },
+    {
+      "epoch": 0.8119349005424955,
+      "grad_norm": 1.58566415309906,
+      "learning_rate": 6.923778239724681e-06,
+      "loss": 0.2767,
+      "step": 449
+    },
+    {
+      "epoch": 0.8137432188065099,
+      "grad_norm": 1.571824073791504,
+      "learning_rate": 6.909961325496312e-06,
+      "loss": 0.3004,
+      "step": 450
+    },
+    {
+      "epoch": 0.8155515370705244,
+      "grad_norm": 1.6151618957519531,
+      "learning_rate": 6.896127313264643e-06,
+      "loss": 0.3023,
+      "step": 451
+    },
+    {
+      "epoch": 0.8173598553345389,
+      "grad_norm": 1.4722002744674683,
+      "learning_rate": 6.88227632687196e-06,
+      "loss": 0.2301,
+      "step": 452
+    },
+    {
+      "epoch": 0.8191681735985533,
+      "grad_norm": 1.4254121780395508,
+      "learning_rate": 6.868408490312511e-06,
+      "loss": 0.2401,
+      "step": 453
+    },
+    {
+      "epoch": 0.8209764918625678,
+      "grad_norm": 1.4507607221603394,
+      "learning_rate": 6.854523927731383e-06,
+      "loss": 0.2683,
+      "step": 454
+    },
+    {
+      "epoch": 0.8227848101265823,
+      "grad_norm": 1.6184362173080444,
+      "learning_rate": 6.840622763423391e-06,
+      "loss": 0.301,
+      "step": 455
+    },
+    {
+      "epoch": 0.8245931283905967,
+      "grad_norm": 1.561452031135559,
+      "learning_rate": 6.8267051218319766e-06,
+      "loss": 0.2869,
+      "step": 456
+    },
+    {
+      "epoch": 0.8264014466546112,
+      "grad_norm": 1.7055240869522095,
+      "learning_rate": 6.8127711275480805e-06,
+      "loss": 0.3714,
+      "step": 457
+    },
+    {
+      "epoch": 0.8282097649186256,
+      "grad_norm": 1.7333972454071045,
+      "learning_rate": 6.798820905309036e-06,
+      "loss": 0.4099,
+      "step": 458
+    },
+    {
+      "epoch": 0.8300180831826401,
+      "grad_norm": 1.6271618604660034,
+      "learning_rate": 6.784854579997446e-06,
+      "loss": 0.3044,
+      "step": 459
+    },
+    {
+      "epoch": 0.8318264014466547,
+      "grad_norm": 1.4839351177215576,
+      "learning_rate": 6.7708722766400745e-06,
+      "loss": 0.2909,
+      "step": 460
+    },
+    {
+      "epoch": 0.833634719710669,
+      "grad_norm": 1.429251790046692,
+      "learning_rate": 6.7568741204067145e-06,
+      "loss": 0.2359,
+      "step": 461
+    },
+    {
+      "epoch": 0.8354430379746836,
+      "grad_norm": 1.4327332973480225,
+      "learning_rate": 6.7428602366090764e-06,
+      "loss": 0.2806,
+      "step": 462
+    },
+    {
+      "epoch": 0.8372513562386981,
+      "grad_norm": 1.3575445413589478,
+      "learning_rate": 6.728830750699667e-06,
+      "loss": 0.2825,
+      "step": 463
+    },
+    {
+      "epoch": 0.8390596745027125,
+      "grad_norm": 1.378030776977539,
+      "learning_rate": 6.714785788270658e-06,
+      "loss": 0.2553,
+      "step": 464
+    },
+    {
+      "epoch": 0.840867992766727,
+      "grad_norm": 1.5035066604614258,
+      "learning_rate": 6.700725475052773e-06,
+      "loss": 0.3266,
+      "step": 465
+    },
+    {
+      "epoch": 0.8426763110307414,
+      "grad_norm": 1.504051923751831,
+      "learning_rate": 6.686649936914151e-06,
+      "loss": 0.2739,
+      "step": 466
+    },
+    {
+      "epoch": 0.8444846292947559,
+      "grad_norm": 1.616783857345581,
+      "learning_rate": 6.672559299859228e-06,
+      "loss": 0.2851,
+      "step": 467
+    },
+    {
+      "epoch": 0.8462929475587704,
+      "grad_norm": 1.441886305809021,
+      "learning_rate": 6.658453690027604e-06,
+      "loss": 0.2652,
+      "step": 468
+    },
+    {
+      "epoch": 0.8481012658227848,
+      "grad_norm": 1.53322172164917,
+      "learning_rate": 6.644333233692917e-06,
+      "loss": 0.2995,
+      "step": 469
+    },
+    {
+      "epoch": 0.8499095840867993,
+      "grad_norm": 1.403977632522583,
+      "learning_rate": 6.63019805726171e-06,
+      "loss": 0.2575,
+      "step": 470
+    },
+    {
+      "epoch": 0.8517179023508138,
+      "grad_norm": 1.598618745803833,
+      "learning_rate": 6.616048287272301e-06,
+      "loss": 0.3201,
+      "step": 471
+    },
+    {
+      "epoch": 0.8535262206148282,
+      "grad_norm": 1.41034996509552,
+      "learning_rate": 6.601884050393649e-06,
+      "loss": 0.2529,
+      "step": 472
+    },
+    {
+      "epoch": 0.8553345388788427,
+      "grad_norm": 1.719070315361023,
+      "learning_rate": 6.587705473424223e-06,
+      "loss": 0.3794,
+      "step": 473
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 1.633659839630127,
+      "learning_rate": 6.57351268329086e-06,
+      "loss": 0.3025,
+      "step": 474
+    },
+    {
+      "epoch": 0.8589511754068716,
+      "grad_norm": 1.910181999206543,
+      "learning_rate": 6.55930580704764e-06,
+      "loss": 0.3444,
+      "step": 475
+    },
+    {
+      "epoch": 0.8607594936708861,
+      "grad_norm": 1.4851632118225098,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.2492,
+      "step": 476
+    },
+    {
+      "epoch": 0.8625678119349005,
+      "grad_norm": 1.38944673538208,
+      "learning_rate": 6.5308503050772884e-06,
+      "loss": 0.2595,
+      "step": 477
+    },
+    {
+      "epoch": 0.864376130198915,
+      "grad_norm": 1.2743439674377441,
+      "learning_rate": 6.51660193408425e-06,
+      "loss": 0.186,
+      "step": 478
+    },
+    {
+      "epoch": 0.8661844484629295,
+      "grad_norm": 1.5795000791549683,
+      "learning_rate": 6.50233998644726e-06,
+      "loss": 0.3064,
+      "step": 479
+    },
+    {
+      "epoch": 0.8679927667269439,
+      "grad_norm": 1.6655584573745728,
+      "learning_rate": 6.4880645898394935e-06,
+      "loss": 0.2734,
+      "step": 480
+    },
+    {
+      "epoch": 0.8698010849909584,
+      "grad_norm": 1.4860823154449463,
+      "learning_rate": 6.473775872054522e-06,
+      "loss": 0.2903,
+      "step": 481
+    },
+    {
+      "epoch": 0.8716094032549728,
+      "grad_norm": 1.5311415195465088,
+      "learning_rate": 6.459473961005168e-06,
+      "loss": 0.3421,
+      "step": 482
+    },
+    {
+      "epoch": 0.8734177215189873,
+      "grad_norm": 1.5267893075942993,
+      "learning_rate": 6.445158984722358e-06,
+      "loss": 0.3595,
+      "step": 483
+    },
+    {
+      "epoch": 0.8752260397830018,
+      "grad_norm": 1.5918548107147217,
+      "learning_rate": 6.4308310713539845e-06,
+      "loss": 0.2756,
+      "step": 484
+    },
+    {
+      "epoch": 0.8770343580470162,
+      "grad_norm": 1.3084057569503784,
+      "learning_rate": 6.4164903491637475e-06,
+      "loss": 0.2272,
+      "step": 485
+    },
+    {
+      "epoch": 0.8788426763110307,
+      "grad_norm": 1.5899354219436646,
+      "learning_rate": 6.402136946530014e-06,
+      "loss": 0.2936,
+      "step": 486
+    },
+    {
+      "epoch": 0.8806509945750453,
+      "grad_norm": 1.3249406814575195,
+      "learning_rate": 6.387770991944667e-06,
+      "loss": 0.2621,
+      "step": 487
+    },
+    {
+      "epoch": 0.8824593128390597,
+      "grad_norm": 1.8586225509643555,
+      "learning_rate": 6.373392614011952e-06,
+      "loss": 0.3753,
+      "step": 488
+    },
+    {
+      "epoch": 0.8842676311030742,
+      "grad_norm": 1.362688660621643,
+      "learning_rate": 6.359001941447331e-06,
+      "loss": 0.2727,
+      "step": 489
+    },
+    {
+      "epoch": 0.8860759493670886,
+      "grad_norm": 1.4210124015808105,
+      "learning_rate": 6.344599103076329e-06,
+      "loss": 0.262,
+      "step": 490
+    },
+    {
+      "epoch": 0.8878842676311031,
+      "grad_norm": 1.268723964691162,
+      "learning_rate": 6.330184227833376e-06,
+      "loss": 0.2268,
+      "step": 491
+    },
+    {
+      "epoch": 0.8896925858951176,
+      "grad_norm": 1.4708172082901,
+      "learning_rate": 6.315757444760659e-06,
+      "loss": 0.2857,
+      "step": 492
+    },
+    {
+      "epoch": 0.891500904159132,
+      "grad_norm": 1.5065549612045288,
+      "learning_rate": 6.301318883006962e-06,
+      "loss": 0.2527,
+      "step": 493
+    },
+    {
+      "epoch": 0.8933092224231465,
+      "grad_norm": 1.4784266948699951,
+      "learning_rate": 6.286868671826513e-06,
+      "loss": 0.2291,
+      "step": 494
+    },
+    {
+      "epoch": 0.895117540687161,
+      "grad_norm": 1.5903269052505493,
+      "learning_rate": 6.272406940577827e-06,
+      "loss": 0.3356,
+      "step": 495
+    },
+    {
+      "epoch": 0.8969258589511754,
+      "grad_norm": 1.5249438285827637,
+      "learning_rate": 6.257933818722544e-06,
+      "loss": 0.2397,
+      "step": 496
+    },
+    {
+      "epoch": 0.8987341772151899,
+      "grad_norm": 1.426729679107666,
+      "learning_rate": 6.243449435824276e-06,
+      "loss": 0.236,
+      "step": 497
+    },
+    {
+      "epoch": 0.9005424954792043,
+      "grad_norm": 1.7714941501617432,
+      "learning_rate": 6.228953921547441e-06,
+      "loss": 0.3306,
+      "step": 498
+    },
+    {
+      "epoch": 0.9023508137432188,
+      "grad_norm": 1.5219871997833252,
+      "learning_rate": 6.2144474056561076e-06,
+      "loss": 0.3083,
+      "step": 499
+    },
+    {
+      "epoch": 0.9041591320072333,
+      "grad_norm": 1.5918591022491455,
+      "learning_rate": 6.19993001801283e-06,
+      "loss": 0.351,
+      "step": 500
+    },
+    {
+      "epoch": 0.9059674502712477,
+      "grad_norm": 1.496276617050171,
+      "learning_rate": 6.185401888577488e-06,
+      "loss": 0.3046,
+      "step": 501
+    },
+    {
+      "epoch": 0.9077757685352622,
+      "grad_norm": 1.41849946975708,
+      "learning_rate": 6.17086314740612e-06,
+      "loss": 0.254,
+      "step": 502
+    },
+    {
+      "epoch": 0.9095840867992767,
+      "grad_norm": 1.437718391418457,
+      "learning_rate": 6.1563139246497615e-06,
+      "loss": 0.2972,
+      "step": 503
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 1.3922340869903564,
+      "learning_rate": 6.141754350553279e-06,
+      "loss": 0.2571,
+      "step": 504
+    },
+    {
+      "epoch": 0.9132007233273056,
+      "grad_norm": 1.597642183303833,
+      "learning_rate": 6.1271845554542065e-06,
+      "loss": 0.314,
+      "step": 505
+    },
+    {
+      "epoch": 0.9150090415913201,
+      "grad_norm": 1.5225824117660522,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.3122,
+      "step": 506
+    },
+    {
+      "epoch": 0.9168173598553345,
+      "grad_norm": 1.4691764116287231,
+      "learning_rate": 6.098014824054741e-06,
+      "loss": 0.2968,
+      "step": 507
+    },
+    {
+      "epoch": 0.918625678119349,
+      "grad_norm": 1.5843572616577148,
+      "learning_rate": 6.083415148882236e-06,
+      "loss": 0.3542,
+      "step": 508
+    },
+    {
+      "epoch": 0.9204339963833634,
+      "grad_norm": 1.489445686340332,
+      "learning_rate": 6.068805774960574e-06,
+      "loss": 0.3083,
+      "step": 509
+    },
+    {
+      "epoch": 0.9222423146473779,
+      "grad_norm": 1.3550089597702026,
+      "learning_rate": 6.054186833073096e-06,
+      "loss": 0.2574,
+      "step": 510
+    },
+    {
+      "epoch": 0.9240506329113924,
+      "grad_norm": 1.2581572532653809,
+      "learning_rate": 6.039558454088796e-06,
+      "loss": 0.2264,
+      "step": 511
+    },
+    {
+      "epoch": 0.9258589511754068,
+      "grad_norm": 1.4372060298919678,
+      "learning_rate": 6.024920768961153e-06,
+      "loss": 0.1999,
+      "step": 512
+    },
+    {
+      "epoch": 0.9276672694394213,
+      "grad_norm": 1.2871983051300049,
+      "learning_rate": 6.010273908726944e-06,
+      "loss": 0.2595,
+      "step": 513
+    },
+    {
+      "epoch": 0.9294755877034359,
+      "grad_norm": 1.4556411504745483,
+      "learning_rate": 5.995618004505091e-06,
+      "loss": 0.3164,
+      "step": 514
+    },
+    {
+      "epoch": 0.9312839059674503,
+      "grad_norm": 1.4673136472702026,
+      "learning_rate": 5.980953187495476e-06,
+      "loss": 0.2846,
+      "step": 515
+    },
+    {
+      "epoch": 0.9330922242314648,
+      "grad_norm": 1.3011901378631592,
+      "learning_rate": 5.9662795889777666e-06,
+      "loss": 0.2311,
+      "step": 516
+    },
+    {
+      "epoch": 0.9349005424954792,
+      "grad_norm": 1.5057899951934814,
+      "learning_rate": 5.951597340310244e-06,
+      "loss": 0.2718,
+      "step": 517
+    },
+    {
+      "epoch": 0.9367088607594937,
+      "grad_norm": 1.4500908851623535,
+      "learning_rate": 5.936906572928625e-06,
+      "loss": 0.2973,
+      "step": 518
+    },
+    {
+      "epoch": 0.9385171790235082,
+      "grad_norm": 1.53549063205719,
+      "learning_rate": 5.922207418344884e-06,
+      "loss": 0.3095,
+      "step": 519
+    },
+    {
+      "epoch": 0.9403254972875226,
+      "grad_norm": 1.6090471744537354,
+      "learning_rate": 5.907500008146082e-06,
+      "loss": 0.2837,
+      "step": 520
+    },
+    {
+      "epoch": 0.9421338155515371,
+      "grad_norm": 1.4988715648651123,
+      "learning_rate": 5.892784473993184e-06,
+      "loss": 0.2474,
+      "step": 521
+    },
+    {
+      "epoch": 0.9439421338155516,
+      "grad_norm": 1.4724313020706177,
+      "learning_rate": 5.878060947619877e-06,
+      "loss": 0.2491,
+      "step": 522
+    },
+    {
+      "epoch": 0.945750452079566,
+      "grad_norm": 1.5639798641204834,
+      "learning_rate": 5.863329560831397e-06,
+      "loss": 0.2355,
+      "step": 523
+    },
+    {
+      "epoch": 0.9475587703435805,
+      "grad_norm": 1.585341215133667,
+      "learning_rate": 5.848590445503345e-06,
+      "loss": 0.2999,
+      "step": 524
+    },
+    {
+      "epoch": 0.9493670886075949,
+      "grad_norm": 1.9147390127182007,
+      "learning_rate": 5.8338437335805124e-06,
+      "loss": 0.3447,
+      "step": 525
+    },
+    {
+      "epoch": 0.9511754068716094,
+      "grad_norm": 1.3038002252578735,
+      "learning_rate": 5.819089557075689e-06,
+      "loss": 0.2896,
+      "step": 526
+    },
+    {
+      "epoch": 0.9529837251356239,
+      "grad_norm": 1.4564861059188843,
+      "learning_rate": 5.8043280480684925e-06,
+      "loss": 0.2513,
+      "step": 527
+    },
+    {
+      "epoch": 0.9547920433996383,
+      "grad_norm": 1.4620134830474854,
+      "learning_rate": 5.78955933870418e-06,
+      "loss": 0.2145,
+      "step": 528
+    },
+    {
+      "epoch": 0.9566003616636528,
+      "grad_norm": 1.608670949935913,
+      "learning_rate": 5.7747835611924655e-06,
+      "loss": 0.3446,
+      "step": 529
+    },
+    {
+      "epoch": 0.9584086799276673,
+      "grad_norm": 1.5054386854171753,
+      "learning_rate": 5.760000847806337e-06,
+      "loss": 0.3055,
+      "step": 530
+    },
+    {
+      "epoch": 0.9602169981916817,
+      "grad_norm": 1.5572552680969238,
+      "learning_rate": 5.745211330880872e-06,
+      "loss": 0.3188,
+      "step": 531
+    },
+    {
+      "epoch": 0.9620253164556962,
+      "grad_norm": 1.5822899341583252,
+      "learning_rate": 5.730415142812059e-06,
+      "loss": 0.3309,
+      "step": 532
+    },
+    {
+      "epoch": 0.9638336347197106,
+      "grad_norm": 1.582491159439087,
+      "learning_rate": 5.7156124160555985e-06,
+      "loss": 0.3307,
+      "step": 533
+    },
+    {
+      "epoch": 0.9656419529837251,
+      "grad_norm": 1.3483372926712036,
+      "learning_rate": 5.70080328312573e-06,
+      "loss": 0.2641,
+      "step": 534
+    },
+    {
+      "epoch": 0.9674502712477396,
+      "grad_norm": 1.3354854583740234,
+      "learning_rate": 5.68598787659404e-06,
+      "loss": 0.2254,
+      "step": 535
+    },
+    {
+      "epoch": 0.969258589511754,
+      "grad_norm": 1.2679874897003174,
+      "learning_rate": 5.671166329088278e-06,
+      "loss": 0.2232,
+      "step": 536
+    },
+    {
+      "epoch": 0.9710669077757685,
+      "grad_norm": 1.419390320777893,
+      "learning_rate": 5.656338773291165e-06,
+      "loss": 0.2335,
+      "step": 537
+    },
+    {
+      "epoch": 0.972875226039783,
+      "grad_norm": 1.31120765209198,
+      "learning_rate": 5.641505341939212e-06,
+      "loss": 0.2434,
+      "step": 538
+    },
+    {
+      "epoch": 0.9746835443037974,
+      "grad_norm": 1.42831289768219,
+      "learning_rate": 5.626666167821522e-06,
+      "loss": 0.2789,
+      "step": 539
+    },
+    {
+      "epoch": 0.976491862567812,
+      "grad_norm": 1.4863039255142212,
+      "learning_rate": 5.611821383778614e-06,
+      "loss": 0.2905,
+      "step": 540
+    },
+    {
+      "epoch": 0.9783001808318263,
+      "grad_norm": 1.2998374700546265,
+      "learning_rate": 5.596971122701221e-06,
+      "loss": 0.241,
+      "step": 541
+    },
+    {
+      "epoch": 0.9801084990958409,
+      "grad_norm": 1.4642691612243652,
+      "learning_rate": 5.582115517529114e-06,
+      "loss": 0.3114,
+      "step": 542
+    },
+    {
+      "epoch": 0.9819168173598554,
+      "grad_norm": 1.541292667388916,
+      "learning_rate": 5.5672547012499e-06,
+      "loss": 0.2984,
+      "step": 543
+    },
+    {
+      "epoch": 0.9837251356238698,
+      "grad_norm": 1.4272589683532715,
+      "learning_rate": 5.55238880689783e-06,
+      "loss": 0.307,
+      "step": 544
+    },
+    {
+      "epoch": 0.9855334538878843,
+      "grad_norm": 1.4981824159622192,
+      "learning_rate": 5.537517967552626e-06,
+      "loss": 0.2646,
+      "step": 545
+    },
+    {
+      "epoch": 0.9873417721518988,
+      "grad_norm": 1.4142934083938599,
+      "learning_rate": 5.522642316338268e-06,
+      "loss": 0.2581,
+      "step": 546
+    },
+    {
+      "epoch": 0.9891500904159132,
+      "grad_norm": 1.5048015117645264,
+      "learning_rate": 5.507761986421818e-06,
+      "loss": 0.2628,
+      "step": 547
+    },
+    {
+      "epoch": 0.9909584086799277,
+      "grad_norm": 1.488046646118164,
+      "learning_rate": 5.4928771110122185e-06,
+      "loss": 0.3166,
+      "step": 548
+    },
+    {
+      "epoch": 0.9927667269439421,
+      "grad_norm": 1.4790452718734741,
+      "learning_rate": 5.477987823359104e-06,
+      "loss": 0.2557,
+      "step": 549
+    },
+    {
+      "epoch": 0.9945750452079566,
+      "grad_norm": 1.6512081623077393,
+      "learning_rate": 5.463094256751608e-06,
+      "loss": 0.3687,
+      "step": 550
+    },
+    {
+      "epoch": 0.9963833634719711,
+      "grad_norm": 1.5028005838394165,
+      "learning_rate": 5.448196544517168e-06,
+      "loss": 0.2967,
+      "step": 551
+    },
+    {
+      "epoch": 0.9981916817359855,
+      "grad_norm": 1.5956478118896484,
+      "learning_rate": 5.433294820020335e-06,
+      "loss": 0.3197,
+      "step": 552
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.464298129081726,
+      "learning_rate": 5.41838921666158e-06,
+      "loss": 0.263,
+      "step": 553
+    },
+    {
+      "epoch": 1.0018083182640145,
+      "grad_norm": 1.3117297887802124,
+      "learning_rate": 5.403479867876087e-06,
+      "loss": 0.2027,
+      "step": 554
+    },
+    {
+      "epoch": 1.003616636528029,
+      "grad_norm": 1.3781763315200806,
+      "learning_rate": 5.388566907132583e-06,
+      "loss": 0.1909,
+      "step": 555
+    },
+    {
+      "epoch": 1.0054249547920433,
+      "grad_norm": 1.4087004661560059,
+      "learning_rate": 5.373650467932122e-06,
+      "loss": 0.2242,
+      "step": 556
+    },
+    {
+      "epoch": 1.0072332730560578,
+      "grad_norm": 1.3254539966583252,
+      "learning_rate": 5.358730683806897e-06,
+      "loss": 0.221,
+      "step": 557
+    },
+    {
+      "epoch": 1.0090415913200723,
+      "grad_norm": 1.4249013662338257,
+      "learning_rate": 5.343807688319047e-06,
+      "loss": 0.2352,
+      "step": 558
+    },
+    {
+      "epoch": 1.0108499095840868,
+      "grad_norm": 1.4847030639648438,
+      "learning_rate": 5.328881615059458e-06,
+      "loss": 0.2912,
+      "step": 559
+    },
+    {
+      "epoch": 1.0126582278481013,
+      "grad_norm": 1.2479192018508911,
+      "learning_rate": 5.3139525976465675e-06,
+      "loss": 0.1962,
+      "step": 560
+    },
+    {
+      "epoch": 1.0144665461121158,
+      "grad_norm": 1.4667129516601562,
+      "learning_rate": 5.299020769725172e-06,
+      "loss": 0.2087,
+      "step": 561
+    },
+    {
+      "epoch": 1.0162748643761301,
+      "grad_norm": 1.1846363544464111,
+      "learning_rate": 5.284086264965224e-06,
+      "loss": 0.1374,
+      "step": 562
+    },
+    {
+      "epoch": 1.0180831826401446,
+      "grad_norm": 1.2141779661178589,
+      "learning_rate": 5.269149217060642e-06,
+      "loss": 0.1698,
+      "step": 563
+    },
+    {
+      "epoch": 1.0198915009041591,
+      "grad_norm": 1.181235671043396,
+      "learning_rate": 5.2542097597281095e-06,
+      "loss": 0.1543,
+      "step": 564
+    },
+    {
+      "epoch": 1.0216998191681737,
+      "grad_norm": 1.3612666130065918,
+      "learning_rate": 5.239268026705878e-06,
+      "loss": 0.2075,
+      "step": 565
+    },
+    {
+      "epoch": 1.0235081374321882,
+      "grad_norm": 1.3640803098678589,
+      "learning_rate": 5.224324151752575e-06,
+      "loss": 0.1809,
+      "step": 566
+    },
+    {
+      "epoch": 1.0253164556962024,
+      "grad_norm": 1.2999763488769531,
+      "learning_rate": 5.209378268645998e-06,
+      "loss": 0.1773,
+      "step": 567
+    },
+    {
+      "epoch": 1.027124773960217,
+      "grad_norm": 1.4859157800674438,
+      "learning_rate": 5.194430511181925e-06,
+      "loss": 0.1986,
+      "step": 568
+    },
+    {
+      "epoch": 1.0289330922242315,
+      "grad_norm": 1.568980097770691,
+      "learning_rate": 5.179481013172912e-06,
+      "loss": 0.2389,
+      "step": 569
+    },
+    {
+      "epoch": 1.030741410488246,
+      "grad_norm": 1.532390832901001,
+      "learning_rate": 5.1645299084470936e-06,
+      "loss": 0.1698,
+      "step": 570
+    },
+    {
+      "epoch": 1.0325497287522605,
+      "grad_norm": 1.5008491277694702,
+      "learning_rate": 5.1495773308469935e-06,
+      "loss": 0.1706,
+      "step": 571
+    },
+    {
+      "epoch": 1.0343580470162748,
+      "grad_norm": 1.5538599491119385,
+      "learning_rate": 5.134623414228315e-06,
+      "loss": 0.2351,
+      "step": 572
+    },
+    {
+      "epoch": 1.0361663652802893,
+      "grad_norm": 1.3042097091674805,
+      "learning_rate": 5.119668292458751e-06,
+      "loss": 0.1678,
+      "step": 573
+    },
+    {
+      "epoch": 1.0379746835443038,
+      "grad_norm": 1.3612871170043945,
+      "learning_rate": 5.1047120994167855e-06,
+      "loss": 0.1975,
+      "step": 574
+    },
+    {
+      "epoch": 1.0397830018083183,
+      "grad_norm": 1.5851926803588867,
+      "learning_rate": 5.0897549689904865e-06,
+      "loss": 0.2288,
+      "step": 575
+    },
+    {
+      "epoch": 1.0415913200723328,
+      "grad_norm": 1.3465512990951538,
+      "learning_rate": 5.074797035076319e-06,
+      "loss": 0.1772,
+      "step": 576
+    },
+    {
+      "epoch": 1.0433996383363473,
+      "grad_norm": 1.5474019050598145,
+      "learning_rate": 5.059838431577937e-06,
+      "loss": 0.2423,
+      "step": 577
+    },
+    {
+      "epoch": 1.0452079566003616,
+      "grad_norm": 1.7244025468826294,
+      "learning_rate": 5.04487929240499e-06,
+      "loss": 0.291,
+      "step": 578
+    },
+    {
+      "epoch": 1.047016274864376,
+      "grad_norm": 1.509497046470642,
+      "learning_rate": 5.029919751471925e-06,
+      "loss": 0.2104,
+      "step": 579
+    },
+    {
+      "epoch": 1.0488245931283906,
+      "grad_norm": 1.299607515335083,
+      "learning_rate": 5.014959942696782e-06,
+      "loss": 0.1796,
+      "step": 580
+    },
+    {
+      "epoch": 1.0506329113924051,
+      "grad_norm": 1.3673267364501953,
+      "learning_rate": 5e-06,
+      "loss": 0.1826,
+      "step": 581
+    },
+    {
+      "epoch": 1.0524412296564196,
+      "grad_norm": 1.3643478155136108,
+      "learning_rate": 4.98504005730322e-06,
+      "loss": 0.1796,
+      "step": 582
+    },
+    {
+      "epoch": 1.054249547920434,
+      "grad_norm": 1.5472909212112427,
+      "learning_rate": 4.970080248528077e-06,
+      "loss": 0.1736,
+      "step": 583
+    },
+    {
+      "epoch": 1.0560578661844484,
+      "grad_norm": 1.5176197290420532,
+      "learning_rate": 4.955120707595011e-06,
+      "loss": 0.2345,
+      "step": 584
+    },
+    {
+      "epoch": 1.057866184448463,
+      "grad_norm": 1.6651631593704224,
+      "learning_rate": 4.940161568422065e-06,
+      "loss": 0.2461,
+      "step": 585
+    },
+    {
+      "epoch": 1.0596745027124774,
+      "grad_norm": 1.336642861366272,
+      "learning_rate": 4.9252029649236835e-06,
+      "loss": 0.1985,
+      "step": 586
+    },
+    {
+      "epoch": 1.061482820976492,
+      "grad_norm": 1.5216953754425049,
+      "learning_rate": 4.910245031009515e-06,
+      "loss": 0.2382,
+      "step": 587
+    },
+    {
+      "epoch": 1.0632911392405062,
+      "grad_norm": 1.5934653282165527,
+      "learning_rate": 4.895287900583216e-06,
+      "loss": 0.234,
+      "step": 588
+    },
+    {
+      "epoch": 1.0650994575045207,
+      "grad_norm": 1.3796063661575317,
+      "learning_rate": 4.88033170754125e-06,
+      "loss": 0.15,
+      "step": 589
+    },
+    {
+      "epoch": 1.0669077757685352,
+      "grad_norm": 1.261986255645752,
+      "learning_rate": 4.865376585771687e-06,
+      "loss": 0.1732,
+      "step": 590
+    },
+    {
+      "epoch": 1.0687160940325497,
+      "grad_norm": 1.2686052322387695,
+      "learning_rate": 4.850422669153009e-06,
+      "loss": 0.1862,
+      "step": 591
+    },
+    {
+      "epoch": 1.0705244122965643,
+      "grad_norm": 1.3567222356796265,
+      "learning_rate": 4.835470091552906e-06,
+      "loss": 0.2148,
+      "step": 592
+    },
+    {
+      "epoch": 1.0723327305605788,
+      "grad_norm": 1.3897775411605835,
+      "learning_rate": 4.8205189868270894e-06,
+      "loss": 0.1955,
+      "step": 593
+    },
+    {
+      "epoch": 1.074141048824593,
+      "grad_norm": 1.4903334379196167,
+      "learning_rate": 4.805569488818077e-06,
+      "loss": 0.2222,
+      "step": 594
+    },
+    {
+      "epoch": 1.0759493670886076,
+      "grad_norm": 1.172352910041809,
+      "learning_rate": 4.7906217313540035e-06,
+      "loss": 0.1574,
+      "step": 595
+    },
+    {
+      "epoch": 1.077757685352622,
+      "grad_norm": 1.4119564294815063,
+      "learning_rate": 4.775675848247427e-06,
+      "loss": 0.1851,
+      "step": 596
+    },
+    {
+      "epoch": 1.0795660036166366,
+      "grad_norm": 1.2973700761795044,
+      "learning_rate": 4.760731973294122e-06,
+      "loss": 0.1554,
+      "step": 597
+    },
+    {
+      "epoch": 1.081374321880651,
+      "grad_norm": 1.3571135997772217,
+      "learning_rate": 4.745790240271892e-06,
+      "loss": 0.1954,
+      "step": 598
+    },
+    {
+      "epoch": 1.0831826401446654,
+      "grad_norm": 1.3942945003509521,
+      "learning_rate": 4.73085078293936e-06,
+      "loss": 0.2016,
+      "step": 599
+    },
+    {
+      "epoch": 1.0849909584086799,
+      "grad_norm": 1.5880961418151855,
+      "learning_rate": 4.715913735034779e-06,
+      "loss": 0.1762,
+      "step": 600
+    },
+    {
+      "epoch": 1.0867992766726944,
+      "grad_norm": 1.5360944271087646,
+      "learning_rate": 4.700979230274829e-06,
+      "loss": 0.21,
+      "step": 601
+    },
+    {
+      "epoch": 1.0886075949367089,
+      "grad_norm": 1.6026498079299927,
+      "learning_rate": 4.686047402353433e-06,
+      "loss": 0.2701,
+      "step": 602
+    },
+    {
+      "epoch": 1.0904159132007234,
+      "grad_norm": 1.4952352046966553,
+      "learning_rate": 4.6711183849405435e-06,
+      "loss": 0.2294,
+      "step": 603
+    },
+    {
+      "epoch": 1.092224231464738,
+      "grad_norm": 1.3100388050079346,
+      "learning_rate": 4.6561923116809545e-06,
+      "loss": 0.161,
+      "step": 604
+    },
+    {
+      "epoch": 1.0940325497287522,
+      "grad_norm": 1.4728491306304932,
+      "learning_rate": 4.641269316193104e-06,
+      "loss": 0.2186,
+      "step": 605
+    },
+    {
+      "epoch": 1.0958408679927667,
+      "grad_norm": 1.9371023178100586,
+      "learning_rate": 4.626349532067879e-06,
+      "loss": 0.1844,
+      "step": 606
+    },
+    {
+      "epoch": 1.0976491862567812,
+      "grad_norm": 1.62328040599823,
+      "learning_rate": 4.611433092867417e-06,
+      "loss": 0.2775,
+      "step": 607
+    },
+    {
+      "epoch": 1.0994575045207957,
+      "grad_norm": 1.4356878995895386,
+      "learning_rate": 4.596520132123915e-06,
+      "loss": 0.2043,
+      "step": 608
+    },
+    {
+      "epoch": 1.1012658227848102,
+      "grad_norm": 1.4942317008972168,
+      "learning_rate": 4.581610783338424e-06,
+      "loss": 0.2003,
+      "step": 609
+    },
+    {
+      "epoch": 1.1030741410488245,
+      "grad_norm": 1.425881028175354,
+      "learning_rate": 4.566705179979665e-06,
+      "loss": 0.2133,
+      "step": 610
+    },
+    {
+      "epoch": 1.104882459312839,
+      "grad_norm": 1.2177973985671997,
+      "learning_rate": 4.551803455482833e-06,
+      "loss": 0.1606,
+      "step": 611
+    },
+    {
+      "epoch": 1.1066907775768535,
+      "grad_norm": 1.517218828201294,
+      "learning_rate": 4.536905743248394e-06,
+      "loss": 0.2364,
+      "step": 612
+    },
+    {
+      "epoch": 1.108499095840868,
+      "grad_norm": 1.442273736000061,
+      "learning_rate": 4.522012176640898e-06,
+      "loss": 0.2205,
+      "step": 613
+    },
+    {
+      "epoch": 1.1103074141048825,
+      "grad_norm": 1.3430852890014648,
+      "learning_rate": 4.507122888987782e-06,
+      "loss": 0.1889,
+      "step": 614
+    },
+    {
+      "epoch": 1.1121157323688968,
+      "grad_norm": 1.5666782855987549,
+      "learning_rate": 4.4922380135781835e-06,
+      "loss": 0.2544,
+      "step": 615
+    },
+    {
+      "epoch": 1.1139240506329113,
+      "grad_norm": 1.3658379316329956,
+      "learning_rate": 4.477357683661734e-06,
+      "loss": 0.1932,
+      "step": 616
+    },
+    {
+      "epoch": 1.1157323688969258,
+      "grad_norm": 1.3235565423965454,
+      "learning_rate": 4.462482032447377e-06,
+      "loss": 0.163,
+      "step": 617
+    },
+    {
+      "epoch": 1.1175406871609403,
+      "grad_norm": 1.2962212562561035,
+      "learning_rate": 4.447611193102171e-06,
+      "loss": 0.1839,
+      "step": 618
+    },
+    {
+      "epoch": 1.1193490054249549,
+      "grad_norm": 1.3023827075958252,
+      "learning_rate": 4.432745298750102e-06,
+      "loss": 0.1856,
+      "step": 619
+    },
+    {
+      "epoch": 1.1211573236889691,
+      "grad_norm": 1.5122865438461304,
+      "learning_rate": 4.417884482470887e-06,
+      "loss": 0.2504,
+      "step": 620
+    },
+    {
+      "epoch": 1.1229656419529837,
+      "grad_norm": 1.4467133283615112,
+      "learning_rate": 4.4030288772987795e-06,
+      "loss": 0.2083,
+      "step": 621
+    },
+    {
+      "epoch": 1.1247739602169982,
+      "grad_norm": 1.2714449167251587,
+      "learning_rate": 4.388178616221389e-06,
+      "loss": 0.143,
+      "step": 622
+    },
+    {
+      "epoch": 1.1265822784810127,
+      "grad_norm": 1.4074504375457764,
+      "learning_rate": 4.373333832178478e-06,
+      "loss": 0.1898,
+      "step": 623
+    },
+    {
+      "epoch": 1.1283905967450272,
+      "grad_norm": 1.278922200202942,
+      "learning_rate": 4.35849465806079e-06,
+      "loss": 0.1375,
+      "step": 624
+    },
+    {
+      "epoch": 1.1301989150090417,
+      "grad_norm": 1.2364168167114258,
+      "learning_rate": 4.3436612267088355e-06,
+      "loss": 0.1439,
+      "step": 625
+    },
+    {
+      "epoch": 1.132007233273056,
+      "grad_norm": 1.4670965671539307,
+      "learning_rate": 4.3288336709117246e-06,
+      "loss": 0.1772,
+      "step": 626
+    },
+    {
+      "epoch": 1.1338155515370705,
+      "grad_norm": 1.5321455001831055,
+      "learning_rate": 4.314012123405961e-06,
+      "loss": 0.2492,
+      "step": 627
+    },
+    {
+      "epoch": 1.135623869801085,
+      "grad_norm": 1.592399001121521,
+      "learning_rate": 4.299196716874271e-06,
+      "loss": 0.2118,
+      "step": 628
+    },
+    {
+      "epoch": 1.1374321880650995,
+      "grad_norm": 1.4796591997146606,
+      "learning_rate": 4.284387583944403e-06,
+      "loss": 0.188,
+      "step": 629
+    },
+    {
+      "epoch": 1.139240506329114,
+      "grad_norm": 1.2441701889038086,
+      "learning_rate": 4.269584857187942e-06,
+      "loss": 0.16,
+      "step": 630
+    },
+    {
+      "epoch": 1.1410488245931285,
+      "grad_norm": 1.564576506614685,
+      "learning_rate": 4.254788669119127e-06,
+      "loss": 0.2022,
+      "step": 631
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 1.6278743743896484,
+      "learning_rate": 4.239999152193664e-06,
+      "loss": 0.2145,
+      "step": 632
+    },
+    {
+      "epoch": 1.1446654611211573,
+      "grad_norm": 1.6288756132125854,
+      "learning_rate": 4.225216438807536e-06,
+      "loss": 0.2328,
+      "step": 633
+    },
+    {
+      "epoch": 1.1464737793851718,
+      "grad_norm": 1.6530628204345703,
+      "learning_rate": 4.2104406612958216e-06,
+      "loss": 0.2808,
+      "step": 634
+    },
+    {
+      "epoch": 1.1482820976491863,
+      "grad_norm": 1.503882646560669,
+      "learning_rate": 4.195671951931509e-06,
+      "loss": 0.2263,
+      "step": 635
+    },
+    {
+      "epoch": 1.1500904159132008,
+      "grad_norm": 1.1974093914031982,
+      "learning_rate": 4.180910442924312e-06,
+      "loss": 0.1463,
+      "step": 636
+    },
+    {
+      "epoch": 1.1518987341772151,
+      "grad_norm": 1.47310209274292,
+      "learning_rate": 4.166156266419489e-06,
+      "loss": 0.193,
+      "step": 637
+    },
+    {
+      "epoch": 1.1537070524412296,
+      "grad_norm": 1.2455514669418335,
+      "learning_rate": 4.1514095544966556e-06,
+      "loss": 0.1488,
+      "step": 638
+    },
+    {
+      "epoch": 1.1555153707052441,
+      "grad_norm": 1.3934590816497803,
+      "learning_rate": 4.136670439168605e-06,
+      "loss": 0.2162,
+      "step": 639
+    },
+    {
+      "epoch": 1.1573236889692586,
+      "grad_norm": 1.3384346961975098,
+      "learning_rate": 4.121939052380125e-06,
+      "loss": 0.2076,
+      "step": 640
+    },
+    {
+      "epoch": 1.1591320072332731,
+      "grad_norm": 1.489557147026062,
+      "learning_rate": 4.107215526006818e-06,
+      "loss": 0.2155,
+      "step": 641
+    },
+    {
+      "epoch": 1.1609403254972874,
+      "grad_norm": 1.3429855108261108,
+      "learning_rate": 4.092499991853919e-06,
+      "loss": 0.1869,
+      "step": 642
+    },
+    {
+      "epoch": 1.162748643761302,
+      "grad_norm": 1.3671791553497314,
+      "learning_rate": 4.0777925816551175e-06,
+      "loss": 0.1861,
+      "step": 643
+    },
+    {
+      "epoch": 1.1645569620253164,
+      "grad_norm": 1.3407803773880005,
+      "learning_rate": 4.063093427071376e-06,
+      "loss": 0.164,
+      "step": 644
+    },
+    {
+      "epoch": 1.166365280289331,
+      "grad_norm": 1.5938869714736938,
+      "learning_rate": 4.048402659689757e-06,
+      "loss": 0.2856,
+      "step": 645
+    },
+    {
+      "epoch": 1.1681735985533455,
+      "grad_norm": 1.3737971782684326,
+      "learning_rate": 4.033720411022235e-06,
+      "loss": 0.1738,
+      "step": 646
+    },
+    {
+      "epoch": 1.1699819168173597,
+      "grad_norm": 1.4273154735565186,
+      "learning_rate": 4.019046812504526e-06,
+      "loss": 0.1974,
+      "step": 647
+    },
+    {
+      "epoch": 1.1717902350813743,
+      "grad_norm": 1.3090167045593262,
+      "learning_rate": 4.0043819954949105e-06,
+      "loss": 0.1572,
+      "step": 648
+    },
+    {
+      "epoch": 1.1735985533453888,
+      "grad_norm": 1.5619733333587646,
+      "learning_rate": 3.989726091273056e-06,
+      "loss": 0.2376,
+      "step": 649
+    },
+    {
+      "epoch": 1.1754068716094033,
+      "grad_norm": 1.6391355991363525,
+      "learning_rate": 3.975079231038848e-06,
+      "loss": 0.2051,
+      "step": 650
+    },
+    {
+      "epoch": 1.1772151898734178,
+      "grad_norm": 1.3856724500656128,
+      "learning_rate": 3.960441545911205e-06,
+      "loss": 0.1677,
+      "step": 651
+    },
+    {
+      "epoch": 1.179023508137432,
+      "grad_norm": 1.6323115825653076,
+      "learning_rate": 3.9458131669269066e-06,
+      "loss": 0.2374,
+      "step": 652
+    },
+    {
+      "epoch": 1.1808318264014466,
+      "grad_norm": 1.5338612794876099,
+      "learning_rate": 3.931194225039427e-06,
+      "loss": 0.2115,
+      "step": 653
+    },
+    {
+      "epoch": 1.182640144665461,
+      "grad_norm": 1.3400039672851562,
+      "learning_rate": 3.916584851117766e-06,
+      "loss": 0.1853,
+      "step": 654
+    },
+    {
+      "epoch": 1.1844484629294756,
+      "grad_norm": 1.3426306247711182,
+      "learning_rate": 3.90198517594526e-06,
+      "loss": 0.1883,
+      "step": 655
+    },
+    {
+      "epoch": 1.18625678119349,
+      "grad_norm": 1.176185965538025,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 0.1475,
+      "step": 656
+    },
+    {
+      "epoch": 1.1880650994575046,
+      "grad_norm": 1.4213191270828247,
+      "learning_rate": 3.872815444545794e-06,
+      "loss": 0.1868,
+      "step": 657
+    },
+    {
+      "epoch": 1.189873417721519,
+      "grad_norm": 1.6255871057510376,
+      "learning_rate": 3.8582456494467214e-06,
+      "loss": 0.2424,
+      "step": 658
+    },
+    {
+      "epoch": 1.1916817359855334,
+      "grad_norm": 1.3936301469802856,
+      "learning_rate": 3.843686075350239e-06,
+      "loss": 0.2065,
+      "step": 659
+    },
+    {
+      "epoch": 1.193490054249548,
+      "grad_norm": 1.6474289894104004,
+      "learning_rate": 3.829136852593881e-06,
+      "loss": 0.2248,
+      "step": 660
+    },
+    {
+      "epoch": 1.1952983725135624,
+      "grad_norm": 1.485768437385559,
+      "learning_rate": 3.8145981114225135e-06,
+      "loss": 0.2007,
+      "step": 661
+    },
+    {
+      "epoch": 1.197106690777577,
+      "grad_norm": 1.2493586540222168,
+      "learning_rate": 3.8000699819871704e-06,
+      "loss": 0.1713,
+      "step": 662
+    },
+    {
+      "epoch": 1.1989150090415914,
+      "grad_norm": 1.2929006814956665,
+      "learning_rate": 3.785552594343894e-06,
+      "loss": 0.1464,
+      "step": 663
+    },
+    {
+      "epoch": 1.2007233273056057,
+      "grad_norm": 1.2843424081802368,
+      "learning_rate": 3.7710460784525617e-06,
+      "loss": 0.1572,
+      "step": 664
+    },
+    {
+      "epoch": 1.2025316455696202,
+      "grad_norm": 1.338903546333313,
+      "learning_rate": 3.756550564175727e-06,
+      "loss": 0.1904,
+      "step": 665
+    },
+    {
+      "epoch": 1.2043399638336347,
+      "grad_norm": 1.3234328031539917,
+      "learning_rate": 3.7420661812774577e-06,
+      "loss": 0.1794,
+      "step": 666
+    },
+    {
+      "epoch": 1.2061482820976492,
+      "grad_norm": 1.2833119630813599,
+      "learning_rate": 3.7275930594221752e-06,
+      "loss": 0.1809,
+      "step": 667
+    },
+    {
+      "epoch": 1.2079566003616637,
+      "grad_norm": 1.5440844297409058,
+      "learning_rate": 3.7131313281734895e-06,
+      "loss": 0.202,
+      "step": 668
+    },
+    {
+      "epoch": 1.209764918625678,
+      "grad_norm": 1.699369192123413,
+      "learning_rate": 3.69868111699304e-06,
+      "loss": 0.237,
+      "step": 669
+    },
+    {
+      "epoch": 1.2115732368896925,
+      "grad_norm": 1.37570321559906,
+      "learning_rate": 3.6842425552393424e-06,
+      "loss": 0.1718,
+      "step": 670
+    },
+    {
+      "epoch": 1.213381555153707,
+      "grad_norm": 1.528974175453186,
+      "learning_rate": 3.669815772166625e-06,
+      "loss": 0.1925,
+      "step": 671
+    },
+    {
+      "epoch": 1.2151898734177216,
+      "grad_norm": 1.3920633792877197,
+      "learning_rate": 3.655400896923672e-06,
+      "loss": 0.1758,
+      "step": 672
+    },
+    {
+      "epoch": 1.216998191681736,
+      "grad_norm": 1.432664394378662,
+      "learning_rate": 3.6409980585526707e-06,
+      "loss": 0.1672,
+      "step": 673
+    },
+    {
+      "epoch": 1.2188065099457503,
+      "grad_norm": 1.4724292755126953,
+      "learning_rate": 3.62660738598805e-06,
+      "loss": 0.1828,
+      "step": 674
+    },
+    {
+      "epoch": 1.2206148282097649,
+      "grad_norm": 1.3286408185958862,
+      "learning_rate": 3.6122290080553334e-06,
+      "loss": 0.1661,
+      "step": 675
+    },
+    {
+      "epoch": 1.2224231464737794,
+      "grad_norm": 1.4410812854766846,
+      "learning_rate": 3.5978630534699873e-06,
+      "loss": 0.1767,
+      "step": 676
+    },
+    {
+      "epoch": 1.2242314647377939,
+      "grad_norm": 1.597132921218872,
+      "learning_rate": 3.583509650836254e-06,
+      "loss": 0.2085,
+      "step": 677
+    },
+    {
+      "epoch": 1.2260397830018084,
+      "grad_norm": 1.8338674306869507,
+      "learning_rate": 3.5691689286460172e-06,
+      "loss": 0.2821,
+      "step": 678
+    },
+    {
+      "epoch": 1.2278481012658227,
+      "grad_norm": 1.3840421438217163,
+      "learning_rate": 3.5548410152776414e-06,
+      "loss": 0.1592,
+      "step": 679
+    },
+    {
+      "epoch": 1.2296564195298372,
+      "grad_norm": 1.298305630683899,
+      "learning_rate": 3.540526038994834e-06,
+      "loss": 0.1629,
+      "step": 680
+    },
+    {
+      "epoch": 1.2314647377938517,
+      "grad_norm": 1.3622138500213623,
+      "learning_rate": 3.526224127945479e-06,
+      "loss": 0.163,
+      "step": 681
+    },
+    {
+      "epoch": 1.2332730560578662,
+      "grad_norm": 1.664790153503418,
+      "learning_rate": 3.5119354101605086e-06,
+      "loss": 0.2225,
+      "step": 682
+    },
+    {
+      "epoch": 1.2350813743218807,
+      "grad_norm": 1.718390941619873,
+      "learning_rate": 3.4976600135527403e-06,
+      "loss": 0.2209,
+      "step": 683
+    },
+    {
+      "epoch": 1.2368896925858952,
+      "grad_norm": 1.634655237197876,
+      "learning_rate": 3.4833980659157507e-06,
+      "loss": 0.2358,
+      "step": 684
+    },
+    {
+      "epoch": 1.2386980108499095,
+      "grad_norm": 1.5706018209457397,
+      "learning_rate": 3.469149694922712e-06,
+      "loss": 0.2219,
+      "step": 685
+    },
+    {
+      "epoch": 1.240506329113924,
+      "grad_norm": 1.480089545249939,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.2048,
+      "step": 686
+    },
+    {
+      "epoch": 1.2423146473779385,
+      "grad_norm": 1.5125480890274048,
+      "learning_rate": 3.4406941929523607e-06,
+      "loss": 0.2325,
+      "step": 687
+    },
+    {
+      "epoch": 1.244122965641953,
+      "grad_norm": 1.3126697540283203,
+      "learning_rate": 3.4264873167091405e-06,
+      "loss": 0.1528,
+      "step": 688
+    },
+    {
+      "epoch": 1.2459312839059675,
+      "grad_norm": 1.4432624578475952,
+      "learning_rate": 3.412294526575779e-06,
+      "loss": 0.2041,
+      "step": 689
+    },
+    {
+      "epoch": 1.247739602169982,
+      "grad_norm": 1.4925624132156372,
+      "learning_rate": 3.398115949606352e-06,
+      "loss": 0.22,
+      "step": 690
+    },
+    {
+      "epoch": 1.2495479204339963,
+      "grad_norm": 1.4070854187011719,
+      "learning_rate": 3.383951712727701e-06,
+      "loss": 0.2023,
+      "step": 691
+    },
+    {
+      "epoch": 1.2513562386980108,
+      "grad_norm": 1.3051189184188843,
+      "learning_rate": 3.3698019427382912e-06,
+      "loss": 0.1776,
+      "step": 692
+    },
+    {
+      "epoch": 1.2531645569620253,
+      "grad_norm": 1.3715490102767944,
+      "learning_rate": 3.355666766307084e-06,
+      "loss": 0.1957,
+      "step": 693
+    },
+    {
+      "epoch": 1.2549728752260398,
+      "grad_norm": 1.4365156888961792,
+      "learning_rate": 3.341546309972398e-06,
+      "loss": 0.1963,
+      "step": 694
+    },
+    {
+      "epoch": 1.2567811934900543,
+      "grad_norm": 1.3215476274490356,
+      "learning_rate": 3.327440700140774e-06,
+      "loss": 0.1727,
+      "step": 695
+    },
+    {
+      "epoch": 1.2585895117540686,
+      "grad_norm": 1.251080870628357,
+      "learning_rate": 3.3133500630858507e-06,
+      "loss": 0.1531,
+      "step": 696
+    },
+    {
+      "epoch": 1.2603978300180831,
+      "grad_norm": 1.3705614805221558,
+      "learning_rate": 3.299274524947229e-06,
+      "loss": 0.2043,
+      "step": 697
+    },
+    {
+      "epoch": 1.2622061482820977,
+      "grad_norm": 1.36355721950531,
+      "learning_rate": 3.2852142117293435e-06,
+      "loss": 0.1954,
+      "step": 698
+    },
+    {
+      "epoch": 1.2640144665461122,
+      "grad_norm": 1.3615798950195312,
+      "learning_rate": 3.2711692493003357e-06,
+      "loss": 0.1588,
+      "step": 699
+    },
+    {
+      "epoch": 1.2658227848101267,
+      "grad_norm": 1.5433591604232788,
+      "learning_rate": 3.2571397633909252e-06,
+      "loss": 0.2013,
+      "step": 700
+    },
+    {
+      "epoch": 1.267631103074141,
+      "grad_norm": 1.4738589525222778,
+      "learning_rate": 3.2431258795932863e-06,
+      "loss": 0.2119,
+      "step": 701
+    },
+    {
+      "epoch": 1.2694394213381555,
+      "grad_norm": 1.5291759967803955,
+      "learning_rate": 3.229127723359927e-06,
+      "loss": 0.2092,
+      "step": 702
+    },
+    {
+      "epoch": 1.27124773960217,
+      "grad_norm": 1.6856266260147095,
+      "learning_rate": 3.215145420002555e-06,
+      "loss": 0.2396,
+      "step": 703
+    },
+    {
+      "epoch": 1.2730560578661845,
+      "grad_norm": 1.3590456247329712,
+      "learning_rate": 3.2011790946909673e-06,
+      "loss": 0.2145,
+      "step": 704
+    },
+    {
+      "epoch": 1.274864376130199,
+      "grad_norm": 1.4624722003936768,
+      "learning_rate": 3.1872288724519207e-06,
+      "loss": 0.1623,
+      "step": 705
+    },
+    {
+      "epoch": 1.2766726943942133,
+      "grad_norm": 1.468032717704773,
+      "learning_rate": 3.173294878168025e-06,
+      "loss": 0.2106,
+      "step": 706
+    },
+    {
+      "epoch": 1.2784810126582278,
+      "grad_norm": 1.3411674499511719,
+      "learning_rate": 3.1593772365766107e-06,
+      "loss": 0.1563,
+      "step": 707
+    },
+    {
+      "epoch": 1.2802893309222423,
+      "grad_norm": 1.5718902349472046,
+      "learning_rate": 3.1454760722686206e-06,
+      "loss": 0.2691,
+      "step": 708
+    },
+    {
+      "epoch": 1.2820976491862568,
+      "grad_norm": 1.5618810653686523,
+      "learning_rate": 3.1315915096874894e-06,
+      "loss": 0.1944,
+      "step": 709
+    },
+    {
+      "epoch": 1.2839059674502713,
+      "grad_norm": 1.3696106672286987,
+      "learning_rate": 3.11772367312804e-06,
+      "loss": 0.2013,
+      "step": 710
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 1.5091700553894043,
+      "learning_rate": 3.1038726867353587e-06,
+      "loss": 0.2409,
+      "step": 711
+    },
+    {
+      "epoch": 1.2875226039783003,
+      "grad_norm": 1.5826542377471924,
+      "learning_rate": 3.090038674503688e-06,
+      "loss": 0.2375,
+      "step": 712
+    },
+    {
+      "epoch": 1.2893309222423146,
+      "grad_norm": 1.4372962713241577,
+      "learning_rate": 3.076221760275321e-06,
+      "loss": 0.1882,
+      "step": 713
+    },
+    {
+      "epoch": 1.2911392405063291,
+      "grad_norm": 1.3111985921859741,
+      "learning_rate": 3.0624220677394854e-06,
+      "loss": 0.1689,
+      "step": 714
+    },
+    {
+      "epoch": 1.2929475587703436,
+      "grad_norm": 1.4454139471054077,
+      "learning_rate": 3.048639720431244e-06,
+      "loss": 0.1667,
+      "step": 715
+    },
+    {
+      "epoch": 1.2947558770343581,
+      "grad_norm": 1.4132417440414429,
+      "learning_rate": 3.0348748417303826e-06,
+      "loss": 0.1781,
+      "step": 716
+    },
+    {
+      "epoch": 1.2965641952983726,
+      "grad_norm": 1.3847126960754395,
+      "learning_rate": 3.0211275548603076e-06,
+      "loss": 0.1729,
+      "step": 717
+    },
+    {
+      "epoch": 1.298372513562387,
+      "grad_norm": 1.21110200881958,
+      "learning_rate": 3.007397982886942e-06,
+      "loss": 0.1512,
+      "step": 718
+    },
+    {
+      "epoch": 1.3001808318264014,
+      "grad_norm": 1.4430028200149536,
+      "learning_rate": 2.9936862487176295e-06,
+      "loss": 0.1658,
+      "step": 719
+    },
+    {
+      "epoch": 1.301989150090416,
+      "grad_norm": 1.3040170669555664,
+      "learning_rate": 2.979992475100024e-06,
+      "loss": 0.1705,
+      "step": 720
+    },
+    {
+      "epoch": 1.3037974683544304,
+      "grad_norm": 1.420000672340393,
+      "learning_rate": 2.966316784621e-06,
+      "loss": 0.19,
+      "step": 721
+    },
+    {
+      "epoch": 1.305605786618445,
+      "grad_norm": 1.5596671104431152,
+      "learning_rate": 2.9526592997055488e-06,
+      "loss": 0.1929,
+      "step": 722
+    },
+    {
+      "epoch": 1.3074141048824592,
+      "grad_norm": 1.2487597465515137,
+      "learning_rate": 2.9390201426156855e-06,
+      "loss": 0.135,
+      "step": 723
+    },
+    {
+      "epoch": 1.3092224231464737,
+      "grad_norm": 1.2281571626663208,
+      "learning_rate": 2.9253994354493575e-06,
+      "loss": 0.1226,
+      "step": 724
+    },
+    {
+      "epoch": 1.3110307414104883,
+      "grad_norm": 1.6670477390289307,
+      "learning_rate": 2.911797300139345e-06,
+      "loss": 0.2273,
+      "step": 725
+    },
+    {
+      "epoch": 1.3128390596745028,
+      "grad_norm": 1.5783288478851318,
+      "learning_rate": 2.8982138584521734e-06,
+      "loss": 0.2369,
+      "step": 726
+    },
+    {
+      "epoch": 1.3146473779385173,
+      "grad_norm": 1.5759886503219604,
+      "learning_rate": 2.884649231987021e-06,
+      "loss": 0.1865,
+      "step": 727
+    },
+    {
+      "epoch": 1.3164556962025316,
+      "grad_norm": 1.6388565301895142,
+      "learning_rate": 2.871103542174637e-06,
+      "loss": 0.248,
+      "step": 728
+    },
+    {
+      "epoch": 1.318264014466546,
+      "grad_norm": 1.4341042041778564,
+      "learning_rate": 2.857576910276243e-06,
+      "loss": 0.209,
+      "step": 729
+    },
+    {
+      "epoch": 1.3200723327305606,
+      "grad_norm": 1.2234702110290527,
+      "learning_rate": 2.844069457382459e-06,
+      "loss": 0.1543,
+      "step": 730
+    },
+    {
+      "epoch": 1.321880650994575,
+      "grad_norm": 1.287217378616333,
+      "learning_rate": 2.83058130441221e-06,
+      "loss": 0.1763,
+      "step": 731
+    },
+    {
+      "epoch": 1.3236889692585896,
+      "grad_norm": 1.4264429807662964,
+      "learning_rate": 2.817112572111651e-06,
+      "loss": 0.1983,
+      "step": 732
+    },
+    {
+      "epoch": 1.3254972875226039,
+      "grad_norm": 1.4583526849746704,
+      "learning_rate": 2.803663381053081e-06,
+      "loss": 0.2137,
+      "step": 733
+    },
+    {
+      "epoch": 1.3273056057866184,
+      "grad_norm": 1.4866368770599365,
+      "learning_rate": 2.790233851633868e-06,
+      "loss": 0.2016,
+      "step": 734
+    },
+    {
+      "epoch": 1.3291139240506329,
+      "grad_norm": 1.5707135200500488,
+      "learning_rate": 2.776824104075364e-06,
+      "loss": 0.1965,
+      "step": 735
+    },
+    {
+      "epoch": 1.3309222423146474,
+      "grad_norm": 1.2514961957931519,
+      "learning_rate": 2.7634342584218364e-06,
+      "loss": 0.16,
+      "step": 736
+    },
+    {
+      "epoch": 1.332730560578662,
+      "grad_norm": 1.4211190938949585,
+      "learning_rate": 2.7500644345393945e-06,
+      "loss": 0.1857,
+      "step": 737
+    },
+    {
+      "epoch": 1.3345388788426762,
+      "grad_norm": 1.4186261892318726,
+      "learning_rate": 2.7367147521149052e-06,
+      "loss": 0.1926,
+      "step": 738
+    },
+    {
+      "epoch": 1.3363471971066907,
+      "grad_norm": 1.3444159030914307,
+      "learning_rate": 2.723385330654933e-06,
+      "loss": 0.1747,
+      "step": 739
+    },
+    {
+      "epoch": 1.3381555153707052,
+      "grad_norm": 1.4014556407928467,
+      "learning_rate": 2.7100762894846633e-06,
+      "loss": 0.1942,
+      "step": 740
+    },
+    {
+      "epoch": 1.3399638336347197,
+      "grad_norm": 1.349727749824524,
+      "learning_rate": 2.6967877477468394e-06,
+      "loss": 0.1751,
+      "step": 741
+    },
+    {
+      "epoch": 1.3417721518987342,
+      "grad_norm": 1.6538288593292236,
+      "learning_rate": 2.683519824400693e-06,
+      "loss": 0.229,
+      "step": 742
+    },
+    {
+      "epoch": 1.3435804701627485,
+      "grad_norm": 1.4319345951080322,
+      "learning_rate": 2.6702726382208775e-06,
+      "loss": 0.1716,
+      "step": 743
+    },
+    {
+      "epoch": 1.3453887884267632,
+      "grad_norm": 1.2382214069366455,
+      "learning_rate": 2.657046307796407e-06,
+      "loss": 0.1349,
+      "step": 744
+    },
+    {
+      "epoch": 1.3471971066907775,
+      "grad_norm": 1.239733338356018,
+      "learning_rate": 2.6438409515295997e-06,
+      "loss": 0.1595,
+      "step": 745
+    },
+    {
+      "epoch": 1.349005424954792,
+      "grad_norm": 1.3710426092147827,
+      "learning_rate": 2.6306566876350072e-06,
+      "loss": 0.1621,
+      "step": 746
+    },
+    {
+      "epoch": 1.3508137432188065,
+      "grad_norm": 1.6545993089675903,
+      "learning_rate": 2.617493634138363e-06,
+      "loss": 0.2161,
+      "step": 747
+    },
+    {
+      "epoch": 1.352622061482821,
+      "grad_norm": 1.4005903005599976,
+      "learning_rate": 2.6043519088755263e-06,
+      "loss": 0.1447,
+      "step": 748
+    },
+    {
+      "epoch": 1.3544303797468356,
+      "grad_norm": 1.3730400800704956,
+      "learning_rate": 2.5912316294914232e-06,
+      "loss": 0.192,
+      "step": 749
+    },
+    {
+      "epoch": 1.3562386980108498,
+      "grad_norm": 1.4984370470046997,
+      "learning_rate": 2.578132913439e-06,
+      "loss": 0.186,
+      "step": 750
+    },
+    {
+      "epoch": 1.3580470162748643,
+      "grad_norm": 1.4603348970413208,
+      "learning_rate": 2.5650558779781635e-06,
+      "loss": 0.1537,
+      "step": 751
+    },
+    {
+      "epoch": 1.3598553345388789,
+      "grad_norm": 1.3268773555755615,
+      "learning_rate": 2.55200064017474e-06,
+      "loss": 0.1829,
+      "step": 752
+    },
+    {
+      "epoch": 1.3616636528028934,
+      "grad_norm": 1.7684950828552246,
+      "learning_rate": 2.538967316899414e-06,
+      "loss": 0.2551,
+      "step": 753
+    },
+    {
+      "epoch": 1.3634719710669079,
+      "grad_norm": 1.3952032327651978,
+      "learning_rate": 2.5259560248267022e-06,
+      "loss": 0.1929,
+      "step": 754
+    },
+    {
+      "epoch": 1.3652802893309222,
+      "grad_norm": 1.3161765336990356,
+      "learning_rate": 2.512966880433891e-06,
+      "loss": 0.1539,
+      "step": 755
+    },
+    {
+      "epoch": 1.3670886075949367,
+      "grad_norm": 1.2540570497512817,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.1687,
+      "step": 756
+    },
+    {
+      "epoch": 1.3688969258589512,
+      "grad_norm": 1.4456546306610107,
+      "learning_rate": 2.4870554996047454e-06,
+      "loss": 0.1795,
+      "step": 757
+    },
+    {
+      "epoch": 1.3707052441229657,
+      "grad_norm": 1.5679326057434082,
+      "learning_rate": 2.4741334951274948e-06,
+      "loss": 0.2109,
+      "step": 758
+    },
+    {
+      "epoch": 1.3725135623869802,
+      "grad_norm": 1.5739426612854004,
+      "learning_rate": 2.461234102246231e-06,
+      "loss": 0.2011,
+      "step": 759
+    },
+    {
+      "epoch": 1.3743218806509945,
+      "grad_norm": 1.5570074319839478,
+      "learning_rate": 2.448357436436519e-06,
+      "loss": 0.2254,
+      "step": 760
+    },
+    {
+      "epoch": 1.376130198915009,
+      "grad_norm": 1.3705545663833618,
+      "learning_rate": 2.43550361297047e-06,
+      "loss": 0.1858,
+      "step": 761
+    },
+    {
+      "epoch": 1.3779385171790235,
+      "grad_norm": 1.437455177307129,
+      "learning_rate": 2.4226727469157097e-06,
+      "loss": 0.1658,
+      "step": 762
+    },
+    {
+      "epoch": 1.379746835443038,
+      "grad_norm": 1.4281333684921265,
+      "learning_rate": 2.40986495313435e-06,
+      "loss": 0.1851,
+      "step": 763
+    },
+    {
+      "epoch": 1.3815551537070525,
+      "grad_norm": 1.3015307188034058,
+      "learning_rate": 2.3970803462819586e-06,
+      "loss": 0.1678,
+      "step": 764
+    },
+    {
+      "epoch": 1.3833634719710668,
+      "grad_norm": 1.3309510946273804,
+      "learning_rate": 2.384319040806533e-06,
+      "loss": 0.1429,
+      "step": 765
+    },
+    {
+      "epoch": 1.3851717902350813,
+      "grad_norm": 1.27116060256958,
+      "learning_rate": 2.371581150947476e-06,
+      "loss": 0.1446,
+      "step": 766
+    },
+    {
+      "epoch": 1.3869801084990958,
+      "grad_norm": 1.4609099626541138,
+      "learning_rate": 2.3588667907345787e-06,
+      "loss": 0.2185,
+      "step": 767
+    },
+    {
+      "epoch": 1.3887884267631103,
+      "grad_norm": 1.1688401699066162,
+      "learning_rate": 2.3461760739869865e-06,
+      "loss": 0.1317,
+      "step": 768
+    },
+    {
+      "epoch": 1.3905967450271248,
+      "grad_norm": 1.430402159690857,
+      "learning_rate": 2.3335091143121935e-06,
+      "loss": 0.1736,
+      "step": 769
+    },
+    {
+      "epoch": 1.3924050632911391,
+      "grad_norm": 1.3158042430877686,
+      "learning_rate": 2.320866025105016e-06,
+      "loss": 0.1469,
+      "step": 770
+    },
+    {
+      "epoch": 1.3942133815551538,
+      "grad_norm": 1.4294010400772095,
+      "learning_rate": 2.3082469195465893e-06,
+      "loss": 0.1801,
+      "step": 771
+    },
+    {
+      "epoch": 1.3960216998191681,
+      "grad_norm": 1.2591050863265991,
+      "learning_rate": 2.2956519106033366e-06,
+      "loss": 0.1469,
+      "step": 772
+    },
+    {
+      "epoch": 1.3978300180831826,
+      "grad_norm": 1.422146201133728,
+      "learning_rate": 2.283081111025973e-06,
+      "loss": 0.1506,
+      "step": 773
+    },
+    {
+      "epoch": 1.3996383363471971,
+      "grad_norm": 1.3352669477462769,
+      "learning_rate": 2.2705346333484925e-06,
+      "loss": 0.1529,
+      "step": 774
+    },
+    {
+      "epoch": 1.4014466546112117,
+      "grad_norm": 1.4001466035842896,
+      "learning_rate": 2.258012589887154e-06,
+      "loss": 0.1719,
+      "step": 775
+    },
+    {
+      "epoch": 1.4032549728752262,
+      "grad_norm": 1.6978325843811035,
+      "learning_rate": 2.245515092739488e-06,
+      "loss": 0.2443,
+      "step": 776
+    },
+    {
+      "epoch": 1.4050632911392404,
+      "grad_norm": 1.5607898235321045,
+      "learning_rate": 2.23304225378328e-06,
+      "loss": 0.1895,
+      "step": 777
+    },
+    {
+      "epoch": 1.406871609403255,
+      "grad_norm": 1.7388428449630737,
+      "learning_rate": 2.2205941846755787e-06,
+      "loss": 0.2613,
+      "step": 778
+    },
+    {
+      "epoch": 1.4086799276672695,
+      "grad_norm": 1.5756665468215942,
+      "learning_rate": 2.2081709968516867e-06,
+      "loss": 0.2497,
+      "step": 779
+    },
+    {
+      "epoch": 1.410488245931284,
+      "grad_norm": 1.3747678995132446,
+      "learning_rate": 2.1957728015241793e-06,
+      "loss": 0.1611,
+      "step": 780
+    },
+    {
+      "epoch": 1.4122965641952985,
+      "grad_norm": 1.6689574718475342,
+      "learning_rate": 2.1833997096818897e-06,
+      "loss": 0.1418,
+      "step": 781
+    },
+    {
+      "epoch": 1.4141048824593128,
+      "grad_norm": 1.572873592376709,
+      "learning_rate": 2.171051832088928e-06,
+      "loss": 0.2104,
+      "step": 782
+    },
+    {
+      "epoch": 1.4159132007233273,
+      "grad_norm": 1.6118040084838867,
+      "learning_rate": 2.158729279283684e-06,
+      "loss": 0.2243,
+      "step": 783
+    },
+    {
+      "epoch": 1.4177215189873418,
+      "grad_norm": 1.4820952415466309,
+      "learning_rate": 2.146432161577842e-06,
+      "loss": 0.1776,
+      "step": 784
+    },
+    {
+      "epoch": 1.4195298372513563,
+      "grad_norm": 1.289957880973816,
+      "learning_rate": 2.1341605890553895e-06,
+      "loss": 0.1816,
+      "step": 785
+    },
+    {
+      "epoch": 1.4213381555153708,
+      "grad_norm": 1.524835228919983,
+      "learning_rate": 2.1219146715716332e-06,
+      "loss": 0.2344,
+      "step": 786
+    },
+    {
+      "epoch": 1.423146473779385,
+      "grad_norm": 1.2140674591064453,
+      "learning_rate": 2.109694518752216e-06,
+      "loss": 0.1522,
+      "step": 787
+    },
+    {
+      "epoch": 1.4249547920433996,
+      "grad_norm": 1.5748006105422974,
+      "learning_rate": 2.097500239992132e-06,
+      "loss": 0.2018,
+      "step": 788
+    },
+    {
+      "epoch": 1.426763110307414,
+      "grad_norm": 1.516115427017212,
+      "learning_rate": 2.085331944454759e-06,
+      "loss": 0.1788,
+      "step": 789
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 1.3118647336959839,
+      "learning_rate": 2.0731897410708618e-06,
+      "loss": 0.1735,
+      "step": 790
+    },
+    {
+      "epoch": 1.4303797468354431,
+      "grad_norm": 1.3635505437850952,
+      "learning_rate": 2.061073738537635e-06,
+      "loss": 0.1881,
+      "step": 791
+    },
+    {
+      "epoch": 1.4321880650994574,
+      "grad_norm": 1.2797064781188965,
+      "learning_rate": 2.0489840453177198e-06,
+      "loss": 0.1558,
+      "step": 792
+    },
+    {
+      "epoch": 1.433996383363472,
+      "grad_norm": 1.5476677417755127,
+      "learning_rate": 2.0369207696382366e-06,
+      "loss": 0.2168,
+      "step": 793
+    },
+    {
+      "epoch": 1.4358047016274864,
+      "grad_norm": 1.2250149250030518,
+      "learning_rate": 2.0248840194898155e-06,
+      "loss": 0.1283,
+      "step": 794
+    },
+    {
+      "epoch": 1.437613019891501,
+      "grad_norm": 1.4220918416976929,
+      "learning_rate": 2.0128739026256306e-06,
+      "loss": 0.1769,
+      "step": 795
+    },
+    {
+      "epoch": 1.4394213381555154,
+      "grad_norm": 1.3836525678634644,
+      "learning_rate": 2.0008905265604316e-06,
+      "loss": 0.1575,
+      "step": 796
+    },
+    {
+      "epoch": 1.4412296564195297,
+      "grad_norm": 1.3742254972457886,
+      "learning_rate": 1.9889339985695894e-06,
+      "loss": 0.1642,
+      "step": 797
+    },
+    {
+      "epoch": 1.4430379746835442,
+      "grad_norm": 1.3974021673202515,
+      "learning_rate": 1.977004425688126e-06,
+      "loss": 0.1714,
+      "step": 798
+    },
+    {
+      "epoch": 1.4448462929475587,
+      "grad_norm": 1.432606816291809,
+      "learning_rate": 1.9651019147097624e-06,
+      "loss": 0.1734,
+      "step": 799
+    },
+    {
+      "epoch": 1.4466546112115732,
+      "grad_norm": 1.3600192070007324,
+      "learning_rate": 1.95322657218596e-06,
+      "loss": 0.1499,
+      "step": 800
+    },
+    {
+      "epoch": 1.4484629294755877,
+      "grad_norm": 1.4031591415405273,
+      "learning_rate": 1.941378504424968e-06,
+      "loss": 0.1517,
+      "step": 801
+    },
+    {
+      "epoch": 1.450271247739602,
+      "grad_norm": 1.366796851158142,
+      "learning_rate": 1.929557817490874e-06,
+      "loss": 0.1869,
+      "step": 802
+    },
+    {
+      "epoch": 1.4520795660036168,
+      "grad_norm": 1.7163602113723755,
+      "learning_rate": 1.9177646172026513e-06,
+      "loss": 0.2516,
+      "step": 803
+    },
+    {
+      "epoch": 1.453887884267631,
+      "grad_norm": 1.390589952468872,
+      "learning_rate": 1.9059990091332082e-06,
+      "loss": 0.2248,
+      "step": 804
+    },
+    {
+      "epoch": 1.4556962025316456,
+      "grad_norm": 1.2839597463607788,
+      "learning_rate": 1.8942610986084487e-06,
+      "loss": 0.1712,
+      "step": 805
+    },
+    {
+      "epoch": 1.45750452079566,
+      "grad_norm": 1.2960193157196045,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 0.1505,
+      "step": 806
+    },
+    {
+      "epoch": 1.4593128390596746,
+      "grad_norm": 1.5872551202774048,
+      "learning_rate": 1.8708687902559264e-06,
+      "loss": 0.2169,
+      "step": 807
+    },
+    {
+      "epoch": 1.461121157323689,
+      "grad_norm": 1.5269652605056763,
+      "learning_rate": 1.8592146018364682e-06,
+      "loss": 0.1994,
+      "step": 808
+    },
+    {
+      "epoch": 1.4629294755877034,
+      "grad_norm": 1.3784176111221313,
+      "learning_rate": 1.8475885297764307e-06,
+      "loss": 0.1962,
+      "step": 809
+    },
+    {
+      "epoch": 1.4647377938517179,
+      "grad_norm": 1.4608467817306519,
+      "learning_rate": 1.8359906781525955e-06,
+      "loss": 0.2519,
+      "step": 810
+    },
+    {
+      "epoch": 1.4665461121157324,
+      "grad_norm": 1.257627010345459,
+      "learning_rate": 1.8244211507891064e-06,
+      "loss": 0.1389,
+      "step": 811
+    },
+    {
+      "epoch": 1.4683544303797469,
+      "grad_norm": 1.4731613397598267,
+      "learning_rate": 1.8128800512565514e-06,
+      "loss": 0.2222,
+      "step": 812
+    },
+    {
+      "epoch": 1.4701627486437614,
+      "grad_norm": 1.4045157432556152,
+      "learning_rate": 1.8013674828710326e-06,
+      "loss": 0.1505,
+      "step": 813
+    },
+    {
+      "epoch": 1.4719710669077757,
+      "grad_norm": 1.346962571144104,
+      "learning_rate": 1.7898835486932398e-06,
+      "loss": 0.173,
+      "step": 814
+    },
+    {
+      "epoch": 1.4737793851717902,
+      "grad_norm": 1.5003893375396729,
+      "learning_rate": 1.7784283515275292e-06,
+      "loss": 0.1661,
+      "step": 815
+    },
+    {
+      "epoch": 1.4755877034358047,
+      "grad_norm": 1.3266466856002808,
+      "learning_rate": 1.7670019939210025e-06,
+      "loss": 0.1681,
+      "step": 816
+    },
+    {
+      "epoch": 1.4773960216998192,
+      "grad_norm": 1.4711284637451172,
+      "learning_rate": 1.7556045781625902e-06,
+      "loss": 0.2244,
+      "step": 817
+    },
+    {
+      "epoch": 1.4792043399638337,
+      "grad_norm": 1.3809276819229126,
+      "learning_rate": 1.7442362062821323e-06,
+      "loss": 0.1702,
+      "step": 818
+    },
+    {
+      "epoch": 1.481012658227848,
+      "grad_norm": 1.4505352973937988,
+      "learning_rate": 1.7328969800494727e-06,
+      "loss": 0.187,
+      "step": 819
+    },
+    {
+      "epoch": 1.4828209764918625,
+      "grad_norm": 1.3470630645751953,
+      "learning_rate": 1.7215870009735386e-06,
+      "loss": 0.1699,
+      "step": 820
+    },
+    {
+      "epoch": 1.484629294755877,
+      "grad_norm": 1.3713916540145874,
+      "learning_rate": 1.7103063703014372e-06,
+      "loss": 0.177,
+      "step": 821
+    },
+    {
+      "epoch": 1.4864376130198915,
+      "grad_norm": 1.39594566822052,
+      "learning_rate": 1.6990551890175488e-06,
+      "loss": 0.1623,
+      "step": 822
+    },
+    {
+      "epoch": 1.488245931283906,
+      "grad_norm": 1.512831211090088,
+      "learning_rate": 1.6878335578426225e-06,
+      "loss": 0.1603,
+      "step": 823
+    },
+    {
+      "epoch": 1.4900542495479203,
+      "grad_norm": 1.5878227949142456,
+      "learning_rate": 1.6766415772328732e-06,
+      "loss": 0.1859,
+      "step": 824
+    },
+    {
+      "epoch": 1.4918625678119348,
+      "grad_norm": 1.2097469568252563,
+      "learning_rate": 1.6654793473790842e-06,
+      "loss": 0.1448,
+      "step": 825
+    },
+    {
+      "epoch": 1.4936708860759493,
+      "grad_norm": 1.4577940702438354,
+      "learning_rate": 1.6543469682057105e-06,
+      "loss": 0.1748,
+      "step": 826
+    },
+    {
+      "epoch": 1.4954792043399638,
+      "grad_norm": 1.5824118852615356,
+      "learning_rate": 1.6432445393699803e-06,
+      "loss": 0.2448,
+      "step": 827
+    },
+    {
+      "epoch": 1.4972875226039783,
+      "grad_norm": 1.4926061630249023,
+      "learning_rate": 1.632172160261012e-06,
+      "loss": 0.2114,
+      "step": 828
+    },
+    {
+      "epoch": 1.4990958408679926,
+      "grad_norm": 1.3919429779052734,
+      "learning_rate": 1.621129929998912e-06,
+      "loss": 0.1974,
+      "step": 829
+    },
+    {
+      "epoch": 1.5009041591320074,
+      "grad_norm": 1.515310525894165,
+      "learning_rate": 1.610117947433897e-06,
+      "loss": 0.2138,
+      "step": 830
+    },
+    {
+      "epoch": 1.5027124773960217,
+      "grad_norm": 1.385372281074524,
+      "learning_rate": 1.5991363111454023e-06,
+      "loss": 0.1616,
+      "step": 831
+    },
+    {
+      "epoch": 1.5045207956600362,
+      "grad_norm": 1.4240636825561523,
+      "learning_rate": 1.5881851194412106e-06,
+      "loss": 0.1761,
+      "step": 832
+    },
+    {
+      "epoch": 1.5063291139240507,
+      "grad_norm": 1.4712889194488525,
+      "learning_rate": 1.5772644703565564e-06,
+      "loss": 0.2036,
+      "step": 833
+    },
+    {
+      "epoch": 1.508137432188065,
+      "grad_norm": 1.475898265838623,
+      "learning_rate": 1.5663744616532612e-06,
+      "loss": 0.2174,
+      "step": 834
+    },
+    {
+      "epoch": 1.5099457504520797,
+      "grad_norm": 1.3733068704605103,
+      "learning_rate": 1.5555151908188465e-06,
+      "loss": 0.183,
+      "step": 835
+    },
+    {
+      "epoch": 1.511754068716094,
+      "grad_norm": 1.2728660106658936,
+      "learning_rate": 1.544686755065677e-06,
+      "loss": 0.163,
+      "step": 836
+    },
+    {
+      "epoch": 1.5135623869801085,
+      "grad_norm": 1.255317211151123,
+      "learning_rate": 1.5338892513300757e-06,
+      "loss": 0.1694,
+      "step": 837
+    },
+    {
+      "epoch": 1.515370705244123,
+      "grad_norm": 1.50929856300354,
+      "learning_rate": 1.523122776271463e-06,
+      "loss": 0.2136,
+      "step": 838
+    },
+    {
+      "epoch": 1.5171790235081373,
+      "grad_norm": 1.4814870357513428,
+      "learning_rate": 1.5123874262714893e-06,
+      "loss": 0.2196,
+      "step": 839
+    },
+    {
+      "epoch": 1.518987341772152,
+      "grad_norm": 1.267453908920288,
+      "learning_rate": 1.5016832974331725e-06,
+      "loss": 0.1548,
+      "step": 840
+    },
+    {
+      "epoch": 1.5207956600361663,
+      "grad_norm": 1.3543283939361572,
+      "learning_rate": 1.4910104855800429e-06,
+      "loss": 0.1848,
+      "step": 841
+    },
+    {
+      "epoch": 1.5226039783001808,
+      "grad_norm": 1.2082693576812744,
+      "learning_rate": 1.4803690862552755e-06,
+      "loss": 0.1516,
+      "step": 842
+    },
+    {
+      "epoch": 1.5244122965641953,
+      "grad_norm": 1.3485254049301147,
+      "learning_rate": 1.4697591947208412e-06,
+      "loss": 0.1513,
+      "step": 843
+    },
+    {
+      "epoch": 1.5262206148282098,
+      "grad_norm": 1.4419184923171997,
+      "learning_rate": 1.459180905956653e-06,
+      "loss": 0.1783,
+      "step": 844
+    },
+    {
+      "epoch": 1.5280289330922243,
+      "grad_norm": 1.2863306999206543,
+      "learning_rate": 1.4486343146597154e-06,
+      "loss": 0.1569,
+      "step": 845
+    },
+    {
+      "epoch": 1.5298372513562386,
+      "grad_norm": 1.366868257522583,
+      "learning_rate": 1.438119515243277e-06,
+      "loss": 0.1732,
+      "step": 846
+    },
+    {
+      "epoch": 1.5316455696202531,
+      "grad_norm": 1.3020880222320557,
+      "learning_rate": 1.4276366018359845e-06,
+      "loss": 0.1486,
+      "step": 847
+    },
+    {
+      "epoch": 1.5334538878842676,
+      "grad_norm": 1.4607349634170532,
+      "learning_rate": 1.4171856682810386e-06,
+      "loss": 0.1761,
+      "step": 848
+    },
+    {
+      "epoch": 1.5352622061482821,
+      "grad_norm": 1.2814878225326538,
+      "learning_rate": 1.4067668081353625e-06,
+      "loss": 0.1529,
+      "step": 849
+    },
+    {
+      "epoch": 1.5370705244122966,
+      "grad_norm": 1.507812738418579,
+      "learning_rate": 1.39638011466875e-06,
+      "loss": 0.1911,
+      "step": 850
+    },
+    {
+      "epoch": 1.538878842676311,
+      "grad_norm": 1.6292840242385864,
+      "learning_rate": 1.3860256808630429e-06,
+      "loss": 0.2177,
+      "step": 851
+    },
+    {
+      "epoch": 1.5406871609403257,
+      "grad_norm": 1.4293454885482788,
+      "learning_rate": 1.3757035994112915e-06,
+      "loss": 0.1967,
+      "step": 852
+    },
+    {
+      "epoch": 1.54249547920434,
+      "grad_norm": 1.595735788345337,
+      "learning_rate": 1.3654139627169282e-06,
+      "loss": 0.2231,
+      "step": 853
+    },
+    {
+      "epoch": 1.5443037974683544,
+      "grad_norm": 1.3294328451156616,
+      "learning_rate": 1.3551568628929434e-06,
+      "loss": 0.1933,
+      "step": 854
+    },
+    {
+      "epoch": 1.546112115732369,
+      "grad_norm": 1.385629653930664,
+      "learning_rate": 1.344932391761049e-06,
+      "loss": 0.187,
+      "step": 855
+    },
+    {
+      "epoch": 1.5479204339963832,
+      "grad_norm": 1.2658005952835083,
+      "learning_rate": 1.3347406408508695e-06,
+      "loss": 0.1646,
+      "step": 856
+    },
+    {
+      "epoch": 1.549728752260398,
+      "grad_norm": 1.4734207391738892,
+      "learning_rate": 1.3245817013991164e-06,
+      "loss": 0.1971,
+      "step": 857
+    },
+    {
+      "epoch": 1.5515370705244123,
+      "grad_norm": 1.4878017902374268,
+      "learning_rate": 1.3144556643487743e-06,
+      "loss": 0.2085,
+      "step": 858
+    },
+    {
+      "epoch": 1.5533453887884268,
+      "grad_norm": 1.627400517463684,
+      "learning_rate": 1.3043626203482823e-06,
+      "loss": 0.232,
+      "step": 859
+    },
+    {
+      "epoch": 1.5551537070524413,
+      "grad_norm": 1.4053066968917847,
+      "learning_rate": 1.2943026597507268e-06,
+      "loss": 0.1906,
+      "step": 860
+    },
+    {
+      "epoch": 1.5569620253164556,
+      "grad_norm": 1.381201148033142,
+      "learning_rate": 1.2842758726130283e-06,
+      "loss": 0.1822,
+      "step": 861
+    },
+    {
+      "epoch": 1.5587703435804703,
+      "grad_norm": 1.4989362955093384,
+      "learning_rate": 1.2742823486951434e-06,
+      "loss": 0.2128,
+      "step": 862
+    },
+    {
+      "epoch": 1.5605786618444846,
+      "grad_norm": 1.1638606786727905,
+      "learning_rate": 1.2643221774592517e-06,
+      "loss": 0.1255,
+      "step": 863
+    },
+    {
+      "epoch": 1.562386980108499,
+      "grad_norm": 1.4444135427474976,
+      "learning_rate": 1.254395448068959e-06,
+      "loss": 0.1892,
+      "step": 864
+    },
+    {
+      "epoch": 1.5641952983725136,
+      "grad_norm": 1.5755318403244019,
+      "learning_rate": 1.2445022493885017e-06,
+      "loss": 0.243,
+      "step": 865
+    },
+    {
+      "epoch": 1.5660036166365279,
+      "grad_norm": 1.4121593236923218,
+      "learning_rate": 1.234642669981946e-06,
+      "loss": 0.1952,
+      "step": 866
+    },
+    {
+      "epoch": 1.5678119349005426,
+      "grad_norm": 1.2563629150390625,
+      "learning_rate": 1.2248167981123992e-06,
+      "loss": 0.1354,
+      "step": 867
+    },
+    {
+      "epoch": 1.5696202531645569,
+      "grad_norm": 1.3229103088378906,
+      "learning_rate": 1.2150247217412186e-06,
+      "loss": 0.1531,
+      "step": 868
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 1.481399655342102,
+      "learning_rate": 1.205266528527223e-06,
+      "loss": 0.1914,
+      "step": 869
+    },
+    {
+      "epoch": 1.573236889692586,
+      "grad_norm": 1.36403489112854,
+      "learning_rate": 1.195542305825908e-06,
+      "loss": 0.1684,
+      "step": 870
+    },
+    {
+      "epoch": 1.5750452079566004,
+      "grad_norm": 1.4882696866989136,
+      "learning_rate": 1.1858521406886674e-06,
+      "loss": 0.2198,
+      "step": 871
+    },
+    {
+      "epoch": 1.576853526220615,
+      "grad_norm": 1.2685580253601074,
+      "learning_rate": 1.1761961198620081e-06,
+      "loss": 0.1541,
+      "step": 872
+    },
+    {
+      "epoch": 1.5786618444846292,
+      "grad_norm": 1.526816487312317,
+      "learning_rate": 1.1665743297867781e-06,
+      "loss": 0.2088,
+      "step": 873
+    },
+    {
+      "epoch": 1.5804701627486437,
+      "grad_norm": 1.7556816339492798,
+      "learning_rate": 1.1569868565973912e-06,
+      "loss": 0.2518,
+      "step": 874
+    },
+    {
+      "epoch": 1.5822784810126582,
+      "grad_norm": 1.544098138809204,
+      "learning_rate": 1.1474337861210543e-06,
+      "loss": 0.1968,
+      "step": 875
+    },
+    {
+      "epoch": 1.5840867992766727,
+      "grad_norm": 1.4241015911102295,
+      "learning_rate": 1.137915203877003e-06,
+      "loss": 0.142,
+      "step": 876
+    },
+    {
+      "epoch": 1.5858951175406872,
+      "grad_norm": 1.4717199802398682,
+      "learning_rate": 1.1284311950757326e-06,
+      "loss": 0.2006,
+      "step": 877
+    },
+    {
+      "epoch": 1.5877034358047015,
+      "grad_norm": 1.4794663190841675,
+      "learning_rate": 1.118981844618236e-06,
+      "loss": 0.1872,
+      "step": 878
+    },
+    {
+      "epoch": 1.5895117540687163,
+      "grad_norm": 1.3753726482391357,
+      "learning_rate": 1.1095672370952431e-06,
+      "loss": 0.1922,
+      "step": 879
+    },
+    {
+      "epoch": 1.5913200723327305,
+      "grad_norm": 1.4431695938110352,
+      "learning_rate": 1.1001874567864696e-06,
+      "loss": 0.1903,
+      "step": 880
+    },
+    {
+      "epoch": 1.593128390596745,
+      "grad_norm": 1.121346354484558,
+      "learning_rate": 1.0908425876598512e-06,
+      "loss": 0.1054,
+      "step": 881
+    },
+    {
+      "epoch": 1.5949367088607596,
+      "grad_norm": 1.3335767984390259,
+      "learning_rate": 1.0815327133708015e-06,
+      "loss": 0.1544,
+      "step": 882
+    },
+    {
+      "epoch": 1.5967450271247738,
+      "grad_norm": 1.4335230588912964,
+      "learning_rate": 1.0722579172614577e-06,
+      "loss": 0.2006,
+      "step": 883
+    },
+    {
+      "epoch": 1.5985533453887886,
+      "grad_norm": 1.5786991119384766,
+      "learning_rate": 1.06301828235994e-06,
+      "loss": 0.2021,
+      "step": 884
+    },
+    {
+      "epoch": 1.6003616636528029,
+      "grad_norm": 1.3353679180145264,
+      "learning_rate": 1.0538138913796032e-06,
+      "loss": 0.1874,
+      "step": 885
+    },
+    {
+      "epoch": 1.6021699819168174,
+      "grad_norm": 1.3402432203292847,
+      "learning_rate": 1.044644826718295e-06,
+      "loss": 0.1683,
+      "step": 886
+    },
+    {
+      "epoch": 1.6039783001808319,
+      "grad_norm": 1.2367477416992188,
+      "learning_rate": 1.0355111704576237e-06,
+      "loss": 0.1509,
+      "step": 887
+    },
+    {
+      "epoch": 1.6057866184448462,
+      "grad_norm": 1.4592540264129639,
+      "learning_rate": 1.0264130043622245e-06,
+      "loss": 0.1983,
+      "step": 888
+    },
+    {
+      "epoch": 1.6075949367088609,
+      "grad_norm": 1.438301682472229,
+      "learning_rate": 1.0173504098790188e-06,
+      "loss": 0.1741,
+      "step": 889
+    },
+    {
+      "epoch": 1.6094032549728752,
+      "grad_norm": 1.4034836292266846,
+      "learning_rate": 1.0083234681364934e-06,
+      "loss": 0.1669,
+      "step": 890
+    },
+    {
+      "epoch": 1.6112115732368897,
+      "grad_norm": 1.4361186027526855,
+      "learning_rate": 9.993322599439692e-07,
+      "loss": 0.1862,
+      "step": 891
+    },
+    {
+      "epoch": 1.6130198915009042,
+      "grad_norm": 1.259032964706421,
+      "learning_rate": 9.903768657908803e-07,
+      "loss": 0.1595,
+      "step": 892
+    },
+    {
+      "epoch": 1.6148282097649185,
+      "grad_norm": 1.3813611268997192,
+      "learning_rate": 9.814573658460564e-07,
+      "loss": 0.1753,
+      "step": 893
+    },
+    {
+      "epoch": 1.6166365280289332,
+      "grad_norm": 1.337997317314148,
+      "learning_rate": 9.725738399569968e-07,
+      "loss": 0.1527,
+      "step": 894
+    },
+    {
+      "epoch": 1.6184448462929475,
+      "grad_norm": 1.438178539276123,
+      "learning_rate": 9.637263676491627e-07,
+      "loss": 0.2042,
+      "step": 895
+    },
+    {
+      "epoch": 1.620253164556962,
+      "grad_norm": 1.3483699560165405,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.1884,
+      "step": 896
+    },
+    {
+      "epoch": 1.6220614828209765,
+      "grad_norm": 1.3791651725769043,
+      "learning_rate": 9.46139900264546e-07,
+      "loss": 0.1636,
+      "step": 897
+    },
+    {
+      "epoch": 1.6238698010849908,
+      "grad_norm": 1.3592714071273804,
+      "learning_rate": 9.374010626220908e-07,
+      "loss": 0.1724,
+      "step": 898
+    },
+    {
+      "epoch": 1.6256781193490055,
+      "grad_norm": 1.2142133712768555,
+      "learning_rate": 9.286985934281079e-07,
+      "loss": 0.1284,
+      "step": 899
+    },
+    {
+      "epoch": 1.6274864376130198,
+      "grad_norm": 1.467472791671753,
+      "learning_rate": 9.200325705872342e-07,
+      "loss": 0.2012,
+      "step": 900
+    },
+    {
+      "epoch": 1.6292947558770343,
+      "grad_norm": 1.6717779636383057,
+      "learning_rate": 9.114030716778433e-07,
+      "loss": 0.2178,
+      "step": 901
+    },
+    {
+      "epoch": 1.6311030741410488,
+      "grad_norm": 1.3586127758026123,
+      "learning_rate": 9.028101739513406e-07,
+      "loss": 0.1548,
+      "step": 902
+    },
+    {
+      "epoch": 1.6329113924050633,
+      "grad_norm": 1.5966354608535767,
+      "learning_rate": 8.942539543314799e-07,
+      "loss": 0.242,
+      "step": 903
+    },
+    {
+      "epoch": 1.6347197106690778,
+      "grad_norm": 1.280997633934021,
+      "learning_rate": 8.857344894136715e-07,
+      "loss": 0.2019,
+      "step": 904
+    },
+    {
+      "epoch": 1.6365280289330921,
+      "grad_norm": 1.2589819431304932,
+      "learning_rate": 8.772518554642973e-07,
+      "loss": 0.1418,
+      "step": 905
+    },
+    {
+      "epoch": 1.6383363471971069,
+      "grad_norm": 1.241722822189331,
+      "learning_rate": 8.688061284200266e-07,
+      "loss": 0.1535,
+      "step": 906
+    },
+    {
+      "epoch": 1.6401446654611211,
+      "grad_norm": 1.5112919807434082,
+      "learning_rate": 8.603973838871388e-07,
+      "loss": 0.1941,
+      "step": 907
+    },
+    {
+      "epoch": 1.6419529837251357,
+      "grad_norm": 1.5479835271835327,
+      "learning_rate": 8.520256971408453e-07,
+      "loss": 0.1946,
+      "step": 908
+    },
+    {
+      "epoch": 1.6437613019891502,
+      "grad_norm": 1.4731577634811401,
+      "learning_rate": 8.436911431246136e-07,
+      "loss": 0.1527,
+      "step": 909
+    },
+    {
+      "epoch": 1.6455696202531644,
+      "grad_norm": 1.6067286729812622,
+      "learning_rate": 8.353937964495029e-07,
+      "loss": 0.2203,
+      "step": 910
+    },
+    {
+      "epoch": 1.6473779385171792,
+      "grad_norm": 1.4132425785064697,
+      "learning_rate": 8.271337313934869e-07,
+      "loss": 0.2019,
+      "step": 911
+    },
+    {
+      "epoch": 1.6491862567811935,
+      "grad_norm": 1.3919837474822998,
+      "learning_rate": 8.189110219007967e-07,
+      "loss": 0.1666,
+      "step": 912
+    },
+    {
+      "epoch": 1.650994575045208,
+      "grad_norm": 1.260819911956787,
+      "learning_rate": 8.107257415812526e-07,
+      "loss": 0.1517,
+      "step": 913
+    },
+    {
+      "epoch": 1.6528028933092225,
+      "grad_norm": 1.4202353954315186,
+      "learning_rate": 8.025779637096138e-07,
+      "loss": 0.2081,
+      "step": 914
+    },
+    {
+      "epoch": 1.6546112115732368,
+      "grad_norm": 1.3754041194915771,
+      "learning_rate": 7.944677612249113e-07,
+      "loss": 0.1866,
+      "step": 915
+    },
+    {
+      "epoch": 1.6564195298372515,
+      "grad_norm": 1.6516166925430298,
+      "learning_rate": 7.863952067298042e-07,
+      "loss": 0.2307,
+      "step": 916
+    },
+    {
+      "epoch": 1.6582278481012658,
+      "grad_norm": 1.5449033975601196,
+      "learning_rate": 7.783603724899258e-07,
+      "loss": 0.174,
+      "step": 917
+    },
+    {
+      "epoch": 1.6600361663652803,
+      "grad_norm": 1.790122389793396,
+      "learning_rate": 7.70363330433233e-07,
+      "loss": 0.2007,
+      "step": 918
+    },
+    {
+      "epoch": 1.6618444846292948,
+      "grad_norm": 1.3342410326004028,
+      "learning_rate": 7.624041521493735e-07,
+      "loss": 0.1553,
+      "step": 919
+    },
+    {
+      "epoch": 1.663652802893309,
+      "grad_norm": 1.3326550722122192,
+      "learning_rate": 7.544829088890326e-07,
+      "loss": 0.1913,
+      "step": 920
+    },
+    {
+      "epoch": 1.6654611211573238,
+      "grad_norm": 1.4879108667373657,
+      "learning_rate": 7.465996715633028e-07,
+      "loss": 0.2249,
+      "step": 921
+    },
+    {
+      "epoch": 1.667269439421338,
+      "grad_norm": 1.4543988704681396,
+      "learning_rate": 7.387545107430455e-07,
+      "loss": 0.1916,
+      "step": 922
+    },
+    {
+      "epoch": 1.6690777576853526,
+      "grad_norm": 1.49525785446167,
+      "learning_rate": 7.309474966582636e-07,
+      "loss": 0.1715,
+      "step": 923
+    },
+    {
+      "epoch": 1.6708860759493671,
+      "grad_norm": 1.3753548860549927,
+      "learning_rate": 7.23178699197467e-07,
+      "loss": 0.1475,
+      "step": 924
+    },
+    {
+      "epoch": 1.6726943942133814,
+      "grad_norm": 1.4657784700393677,
+      "learning_rate": 7.154481879070502e-07,
+      "loss": 0.1763,
+      "step": 925
+    },
+    {
+      "epoch": 1.6745027124773961,
+      "grad_norm": 1.640970230102539,
+      "learning_rate": 7.077560319906696e-07,
+      "loss": 0.2059,
+      "step": 926
+    },
+    {
+      "epoch": 1.6763110307414104,
+      "grad_norm": 1.5941320657730103,
+      "learning_rate": 7.001023003086243e-07,
+      "loss": 0.2146,
+      "step": 927
+    },
+    {
+      "epoch": 1.678119349005425,
+      "grad_norm": 1.595048189163208,
+      "learning_rate": 6.924870613772388e-07,
+      "loss": 0.2296,
+      "step": 928
+    },
+    {
+      "epoch": 1.6799276672694394,
+      "grad_norm": 1.2889961004257202,
+      "learning_rate": 6.849103833682491e-07,
+      "loss": 0.1861,
+      "step": 929
+    },
+    {
+      "epoch": 1.681735985533454,
+      "grad_norm": 1.2051788568496704,
+      "learning_rate": 6.773723341081945e-07,
+      "loss": 0.1704,
+      "step": 930
+    },
+    {
+      "epoch": 1.6835443037974684,
+      "grad_norm": 1.346713900566101,
+      "learning_rate": 6.698729810778065e-07,
+      "loss": 0.1746,
+      "step": 931
+    },
+    {
+      "epoch": 1.6853526220614827,
+      "grad_norm": 1.4051306247711182,
+      "learning_rate": 6.624123914114122e-07,
+      "loss": 0.161,
+      "step": 932
+    },
+    {
+      "epoch": 1.6871609403254972,
+      "grad_norm": 1.3659924268722534,
+      "learning_rate": 6.549906318963245e-07,
+      "loss": 0.1757,
+      "step": 933
+    },
+    {
+      "epoch": 1.6889692585895117,
+      "grad_norm": 1.442762017250061,
+      "learning_rate": 6.476077689722487e-07,
+      "loss": 0.1918,
+      "step": 934
+    },
+    {
+      "epoch": 1.6907775768535263,
+      "grad_norm": 1.488540768623352,
+      "learning_rate": 6.402638687306872e-07,
+      "loss": 0.2175,
+      "step": 935
+    },
+    {
+      "epoch": 1.6925858951175408,
+      "grad_norm": 1.3435027599334717,
+      "learning_rate": 6.329589969143518e-07,
+      "loss": 0.1643,
+      "step": 936
+    },
+    {
+      "epoch": 1.694394213381555,
+      "grad_norm": 1.1935302019119263,
+      "learning_rate": 6.256932189165644e-07,
+      "loss": 0.1306,
+      "step": 937
+    },
+    {
+      "epoch": 1.6962025316455698,
+      "grad_norm": 1.3391468524932861,
+      "learning_rate": 6.184665997806832e-07,
+      "loss": 0.1497,
+      "step": 938
+    },
+    {
+      "epoch": 1.698010849909584,
+      "grad_norm": 1.5649363994598389,
+      "learning_rate": 6.112792041995125e-07,
+      "loss": 0.2096,
+      "step": 939
+    },
+    {
+      "epoch": 1.6998191681735986,
+      "grad_norm": 1.3080732822418213,
+      "learning_rate": 6.041310965147318e-07,
+      "loss": 0.1562,
+      "step": 940
+    },
+    {
+      "epoch": 1.701627486437613,
+      "grad_norm": 1.4739066362380981,
+      "learning_rate": 5.9702234071631e-07,
+      "loss": 0.196,
+      "step": 941
+    },
+    {
+      "epoch": 1.7034358047016274,
+      "grad_norm": 1.3502253293991089,
+      "learning_rate": 5.899530004419396e-07,
+      "loss": 0.1557,
+      "step": 942
+    },
+    {
+      "epoch": 1.705244122965642,
+      "grad_norm": 1.2946463823318481,
+      "learning_rate": 5.829231389764628e-07,
+      "loss": 0.1789,
+      "step": 943
+    },
+    {
+      "epoch": 1.7070524412296564,
+      "grad_norm": 1.3189704418182373,
+      "learning_rate": 5.759328192513075e-07,
+      "loss": 0.1559,
+      "step": 944
+    },
+    {
+      "epoch": 1.7088607594936709,
+      "grad_norm": 1.6325633525848389,
+      "learning_rate": 5.689821038439264e-07,
+      "loss": 0.2148,
+      "step": 945
+    },
+    {
+      "epoch": 1.7106690777576854,
+      "grad_norm": 1.3703547716140747,
+      "learning_rate": 5.620710549772295e-07,
+      "loss": 0.1663,
+      "step": 946
+    },
+    {
+      "epoch": 1.7124773960216997,
+      "grad_norm": 1.351824164390564,
+      "learning_rate": 5.55199734519034e-07,
+      "loss": 0.1754,
+      "step": 947
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 1.2838019132614136,
+      "learning_rate": 5.483682039815059e-07,
+      "loss": 0.1472,
+      "step": 948
+    },
+    {
+      "epoch": 1.7160940325497287,
+      "grad_norm": 1.3122061491012573,
+      "learning_rate": 5.415765245206128e-07,
+      "loss": 0.1506,
+      "step": 949
+    },
+    {
+      "epoch": 1.7179023508137432,
+      "grad_norm": 1.4882829189300537,
+      "learning_rate": 5.348247569355736e-07,
+      "loss": 0.1914,
+      "step": 950
+    },
+    {
+      "epoch": 1.7197106690777577,
+      "grad_norm": 1.3123871088027954,
+      "learning_rate": 5.281129616683167e-07,
+      "loss": 0.1368,
+      "step": 951
+    },
+    {
+      "epoch": 1.721518987341772,
+      "grad_norm": 1.3857144117355347,
+      "learning_rate": 5.214411988029355e-07,
+      "loss": 0.1688,
+      "step": 952
+    },
+    {
+      "epoch": 1.7233273056057867,
+      "grad_norm": 1.731164574623108,
+      "learning_rate": 5.148095280651566e-07,
+      "loss": 0.2503,
+      "step": 953
+    },
+    {
+      "epoch": 1.725135623869801,
+      "grad_norm": 1.3402376174926758,
+      "learning_rate": 5.082180088217981e-07,
+      "loss": 0.1786,
+      "step": 954
+    },
+    {
+      "epoch": 1.7269439421338155,
+      "grad_norm": 1.2338224649429321,
+      "learning_rate": 5.016667000802417e-07,
+      "loss": 0.144,
+      "step": 955
+    },
+    {
+      "epoch": 1.72875226039783,
+      "grad_norm": 1.353575587272644,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 0.1458,
+      "step": 956
+    },
+    {
+      "epoch": 1.7305605786618445,
+      "grad_norm": 1.6582491397857666,
+      "learning_rate": 4.886849483317141e-07,
+      "loss": 0.1942,
+      "step": 957
+    },
+    {
+      "epoch": 1.732368896925859,
+      "grad_norm": 1.782560110092163,
+      "learning_rate": 4.822546215375851e-07,
+      "loss": 0.2759,
+      "step": 958
+    },
+    {
+      "epoch": 1.7341772151898733,
+      "grad_norm": 1.6829458475112915,
+      "learning_rate": 4.758647376699033e-07,
+      "loss": 0.2606,
+      "step": 959
+    },
+    {
+      "epoch": 1.7359855334538878,
+      "grad_norm": 1.549209713935852,
+      "learning_rate": 4.6951535393100654e-07,
+      "loss": 0.2303,
+      "step": 960
+    },
+    {
+      "epoch": 1.7377938517179023,
+      "grad_norm": 1.3191518783569336,
+      "learning_rate": 4.632065271606756e-07,
+      "loss": 0.1533,
+      "step": 961
+    },
+    {
+      "epoch": 1.7396021699819169,
+      "grad_norm": 1.565401554107666,
+      "learning_rate": 4.569383138356276e-07,
+      "loss": 0.2127,
+      "step": 962
+    },
+    {
+      "epoch": 1.7414104882459314,
+      "grad_norm": 1.5010247230529785,
+      "learning_rate": 4.507107700690044e-07,
+      "loss": 0.2363,
+      "step": 963
+    },
+    {
+      "epoch": 1.7432188065099457,
+      "grad_norm": 1.3151311874389648,
+      "learning_rate": 4.4452395160987314e-07,
+      "loss": 0.1479,
+      "step": 964
+    },
+    {
+      "epoch": 1.7450271247739604,
+      "grad_norm": 1.4224631786346436,
+      "learning_rate": 4.383779138427274e-07,
+      "loss": 0.1784,
+      "step": 965
+    },
+    {
+      "epoch": 1.7468354430379747,
+      "grad_norm": 1.3394750356674194,
+      "learning_rate": 4.322727117869951e-07,
+      "loss": 0.1527,
+      "step": 966
+    },
+    {
+      "epoch": 1.7486437613019892,
+      "grad_norm": 1.3689308166503906,
+      "learning_rate": 4.2620840009653827e-07,
+      "loss": 0.1804,
+      "step": 967
+    },
+    {
+      "epoch": 1.7504520795660037,
+      "grad_norm": 1.4480156898498535,
+      "learning_rate": 4.201850330591678e-07,
+      "loss": 0.2136,
+      "step": 968
+    },
+    {
+      "epoch": 1.752260397830018,
+      "grad_norm": 1.5066252946853638,
+      "learning_rate": 4.1420266459615944e-07,
+      "loss": 0.2062,
+      "step": 969
+    },
+    {
+      "epoch": 1.7540687160940327,
+      "grad_norm": 1.3773577213287354,
+      "learning_rate": 4.082613482617664e-07,
+      "loss": 0.1553,
+      "step": 970
+    },
+    {
+      "epoch": 1.755877034358047,
+      "grad_norm": 1.287730097770691,
+      "learning_rate": 4.0236113724274716e-07,
+      "loss": 0.1399,
+      "step": 971
+    },
+    {
+      "epoch": 1.7576853526220615,
+      "grad_norm": 1.420283555984497,
+      "learning_rate": 3.965020843578804e-07,
+      "loss": 0.1636,
+      "step": 972
+    },
+    {
+      "epoch": 1.759493670886076,
+      "grad_norm": 1.3530566692352295,
+      "learning_rate": 3.90684242057498e-07,
+      "loss": 0.1625,
+      "step": 973
+    },
+    {
+      "epoch": 1.7613019891500903,
+      "grad_norm": 1.2307391166687012,
+      "learning_rate": 3.8490766242301356e-07,
+      "loss": 0.135,
+      "step": 974
+    },
+    {
+      "epoch": 1.763110307414105,
+      "grad_norm": 1.3423430919647217,
+      "learning_rate": 3.7917239716645906e-07,
+      "loss": 0.1531,
+      "step": 975
+    },
+    {
+      "epoch": 1.7649186256781193,
+      "grad_norm": 1.2529070377349854,
+      "learning_rate": 3.734784976300165e-07,
+      "loss": 0.1566,
+      "step": 976
+    },
+    {
+      "epoch": 1.7667269439421338,
+      "grad_norm": 1.437445044517517,
+      "learning_rate": 3.678260147855628e-07,
+      "loss": 0.1802,
+      "step": 977
+    },
+    {
+      "epoch": 1.7685352622061483,
+      "grad_norm": 1.633385181427002,
+      "learning_rate": 3.6221499923421164e-07,
+      "loss": 0.2335,
+      "step": 978
+    },
+    {
+      "epoch": 1.7703435804701626,
+      "grad_norm": 1.2325958013534546,
+      "learning_rate": 3.566455012058612e-07,
+      "loss": 0.1435,
+      "step": 979
+    },
+    {
+      "epoch": 1.7721518987341773,
+      "grad_norm": 1.4155337810516357,
+      "learning_rate": 3.511175705587433e-07,
+      "loss": 0.1985,
+      "step": 980
+    },
+    {
+      "epoch": 1.7739602169981916,
+      "grad_norm": 1.4826092720031738,
+      "learning_rate": 3.4563125677897936e-07,
+      "loss": 0.2048,
+      "step": 981
+    },
+    {
+      "epoch": 1.7757685352622061,
+      "grad_norm": 1.4274451732635498,
+      "learning_rate": 3.4018660898013423e-07,
+      "loss": 0.1825,
+      "step": 982
+    },
+    {
+      "epoch": 1.7775768535262206,
+      "grad_norm": 1.6296826601028442,
+      "learning_rate": 3.347836759027789e-07,
+      "loss": 0.238,
+      "step": 983
+    },
+    {
+      "epoch": 1.779385171790235,
+      "grad_norm": 1.4893583059310913,
+      "learning_rate": 3.2942250591405546e-07,
+      "loss": 0.2111,
+      "step": 984
+    },
+    {
+      "epoch": 1.7811934900542497,
+      "grad_norm": 1.403668999671936,
+      "learning_rate": 3.241031470072398e-07,
+      "loss": 0.2029,
+      "step": 985
+    },
+    {
+      "epoch": 1.783001808318264,
+      "grad_norm": 1.5171066522598267,
+      "learning_rate": 3.18825646801314e-07,
+      "loss": 0.2289,
+      "step": 986
+    },
+    {
+      "epoch": 1.7848101265822784,
+      "grad_norm": 1.2951208353042603,
+      "learning_rate": 3.135900525405428e-07,
+      "loss": 0.1638,
+      "step": 987
+    },
+    {
+      "epoch": 1.786618444846293,
+      "grad_norm": 1.4618269205093384,
+      "learning_rate": 3.0839641109404627e-07,
+      "loss": 0.2098,
+      "step": 988
+    },
+    {
+      "epoch": 1.7884267631103075,
+      "grad_norm": 1.3721157312393188,
+      "learning_rate": 3.03244768955383e-07,
+      "loss": 0.1716,
+      "step": 989
+    },
+    {
+      "epoch": 1.790235081374322,
+      "grad_norm": 1.3801895380020142,
+      "learning_rate": 2.9813517224213274e-07,
+      "loss": 0.1815,
+      "step": 990
+    },
+    {
+      "epoch": 1.7920433996383363,
+      "grad_norm": 1.4312127828598022,
+      "learning_rate": 2.930676666954846e-07,
+      "loss": 0.1884,
+      "step": 991
+    },
+    {
+      "epoch": 1.7938517179023508,
+      "grad_norm": 1.4009455442428589,
+      "learning_rate": 2.8804229767982637e-07,
+      "loss": 0.1819,
+      "step": 992
+    },
+    {
+      "epoch": 1.7956600361663653,
+      "grad_norm": 1.3529410362243652,
+      "learning_rate": 2.8305911018233935e-07,
+      "loss": 0.157,
+      "step": 993
+    },
+    {
+      "epoch": 1.7974683544303798,
+      "grad_norm": 1.4529396295547485,
+      "learning_rate": 2.7811814881259503e-07,
+      "loss": 0.1873,
+      "step": 994
+    },
+    {
+      "epoch": 1.7992766726943943,
+      "grad_norm": 1.3773640394210815,
+      "learning_rate": 2.7321945780215576e-07,
+      "loss": 0.1396,
+      "step": 995
+    },
+    {
+      "epoch": 1.8010849909584086,
+      "grad_norm": 1.3521469831466675,
+      "learning_rate": 2.6836308100417874e-07,
+      "loss": 0.1619,
+      "step": 996
+    },
+    {
+      "epoch": 1.8028933092224233,
+      "grad_norm": 1.3587416410446167,
+      "learning_rate": 2.6354906189302534e-07,
+      "loss": 0.14,
+      "step": 997
+    },
+    {
+      "epoch": 1.8047016274864376,
+      "grad_norm": 1.3548296689987183,
+      "learning_rate": 2.587774435638679e-07,
+      "loss": 0.1268,
+      "step": 998
+    },
+    {
+      "epoch": 1.806509945750452,
+      "grad_norm": 1.5057969093322754,
+      "learning_rate": 2.5404826873230926e-07,
+      "loss": 0.1788,
+      "step": 999
+    },
+    {
+      "epoch": 1.8083182640144666,
+      "grad_norm": 1.6407438516616821,
+      "learning_rate": 2.4936157973399266e-07,
+      "loss": 0.2138,
+      "step": 1000
+    },
+    {
+      "epoch": 1.810126582278481,
+      "grad_norm": 1.5798081159591675,
+      "learning_rate": 2.447174185242324e-07,
+      "loss": 0.1921,
+      "step": 1001
+    },
+    {
+      "epoch": 1.8119349005424956,
+      "grad_norm": 1.5625442266464233,
+      "learning_rate": 2.40115826677631e-07,
+      "loss": 0.1757,
+      "step": 1002
+    },
+    {
+      "epoch": 1.81374321880651,
+      "grad_norm": 1.638500452041626,
+      "learning_rate": 2.3555684538770995e-07,
+      "loss": 0.2254,
+      "step": 1003
+    },
+    {
+      "epoch": 1.8155515370705244,
+      "grad_norm": 1.3669939041137695,
+      "learning_rate": 2.3104051546654016e-07,
+      "loss": 0.1808,
+      "step": 1004
+    },
+    {
+      "epoch": 1.817359855334539,
+      "grad_norm": 1.4740699529647827,
+      "learning_rate": 2.2656687734437577e-07,
+      "loss": 0.2454,
+      "step": 1005
+    },
+    {
+      "epoch": 1.8191681735985532,
+      "grad_norm": 1.3462810516357422,
+      "learning_rate": 2.2213597106929608e-07,
+      "loss": 0.1583,
+      "step": 1006
+    },
+    {
+      "epoch": 1.820976491862568,
+      "grad_norm": 1.6884946823120117,
+      "learning_rate": 2.177478363068425e-07,
+      "loss": 0.2415,
+      "step": 1007
+    },
+    {
+      "epoch": 1.8227848101265822,
+      "grad_norm": 1.511620044708252,
+      "learning_rate": 2.134025123396638e-07,
+      "loss": 0.2227,
+      "step": 1008
+    },
+    {
+      "epoch": 1.8245931283905967,
+      "grad_norm": 1.4712649583816528,
+      "learning_rate": 2.0910003806716817e-07,
+      "loss": 0.1888,
+      "step": 1009
+    },
+    {
+      "epoch": 1.8264014466546112,
+      "grad_norm": 1.5498689413070679,
+      "learning_rate": 2.0484045200517222e-07,
+      "loss": 0.2506,
+      "step": 1010
+    },
+    {
+      "epoch": 1.8282097649186255,
+      "grad_norm": 1.3635046482086182,
+      "learning_rate": 2.006237922855553e-07,
+      "loss": 0.1767,
+      "step": 1011
+    },
+    {
+      "epoch": 1.8300180831826403,
+      "grad_norm": 1.5200732946395874,
+      "learning_rate": 1.9645009665592073e-07,
+      "loss": 0.1996,
+      "step": 1012
+    },
+    {
+      "epoch": 1.8318264014466545,
+      "grad_norm": 1.483102798461914,
+      "learning_rate": 1.9231940247925573e-07,
+      "loss": 0.1994,
+      "step": 1013
+    },
+    {
+      "epoch": 1.833634719710669,
+      "grad_norm": 1.6003260612487793,
+      "learning_rate": 1.882317467335998e-07,
+      "loss": 0.2617,
+      "step": 1014
+    },
+    {
+      "epoch": 1.8354430379746836,
+      "grad_norm": 1.3227473497390747,
+      "learning_rate": 1.841871660117095e-07,
+      "loss": 0.1602,
+      "step": 1015
+    },
+    {
+      "epoch": 1.837251356238698,
+      "grad_norm": 1.5393905639648438,
+      "learning_rate": 1.801856965207338e-07,
+      "loss": 0.229,
+      "step": 1016
+    },
+    {
+      "epoch": 1.8390596745027126,
+      "grad_norm": 1.2200417518615723,
+      "learning_rate": 1.7622737408188984e-07,
+      "loss": 0.1448,
+      "step": 1017
+    },
+    {
+      "epoch": 1.8408679927667269,
+      "grad_norm": 1.4272844791412354,
+      "learning_rate": 1.7231223413014086e-07,
+      "loss": 0.2081,
+      "step": 1018
+    },
+    {
+      "epoch": 1.8426763110307414,
+      "grad_norm": 1.3564151525497437,
+      "learning_rate": 1.6844031171388054e-07,
+      "loss": 0.1992,
+      "step": 1019
+    },
+    {
+      "epoch": 1.8444846292947559,
+      "grad_norm": 1.2938915491104126,
+      "learning_rate": 1.6461164149461805e-07,
+      "loss": 0.163,
+      "step": 1020
+    },
+    {
+      "epoch": 1.8462929475587704,
+      "grad_norm": 1.307634711265564,
+      "learning_rate": 1.6082625774666793e-07,
+      "loss": 0.1393,
+      "step": 1021
+    },
+    {
+      "epoch": 1.8481012658227849,
+      "grad_norm": 1.332777738571167,
+      "learning_rate": 1.5708419435684463e-07,
+      "loss": 0.1685,
+      "step": 1022
+    },
+    {
+      "epoch": 1.8499095840867992,
+      "grad_norm": 1.3043736219406128,
+      "learning_rate": 1.5338548482415726e-07,
+      "loss": 0.1499,
+      "step": 1023
+    },
+    {
+      "epoch": 1.851717902350814,
+      "grad_norm": 1.3902978897094727,
+      "learning_rate": 1.4973016225951097e-07,
+      "loss": 0.175,
+      "step": 1024
+    },
+    {
+      "epoch": 1.8535262206148282,
+      "grad_norm": 1.3513139486312866,
+      "learning_rate": 1.4611825938540936e-07,
+      "loss": 0.1508,
+      "step": 1025
+    },
+    {
+      "epoch": 1.8553345388788427,
+      "grad_norm": 1.4719350337982178,
+      "learning_rate": 1.4254980853566248e-07,
+      "loss": 0.1741,
+      "step": 1026
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 1.6112737655639648,
+      "learning_rate": 1.3902484165509877e-07,
+      "loss": 0.2054,
+      "step": 1027
+    },
+    {
+      "epoch": 1.8589511754068715,
+      "grad_norm": 1.7364201545715332,
+      "learning_rate": 1.3554339029927532e-07,
+      "loss": 0.2887,
+      "step": 1028
+    },
+    {
+      "epoch": 1.8607594936708862,
+      "grad_norm": 1.2992795705795288,
+      "learning_rate": 1.3210548563419857e-07,
+      "loss": 0.1532,
+      "step": 1029
+    },
+    {
+      "epoch": 1.8625678119349005,
+      "grad_norm": 1.2579344511032104,
+      "learning_rate": 1.2871115843604508e-07,
+      "loss": 0.1324,
+      "step": 1030
+    },
+    {
+      "epoch": 1.864376130198915,
+      "grad_norm": 1.2674459218978882,
+      "learning_rate": 1.253604390908819e-07,
+      "loss": 0.1439,
+      "step": 1031
+    },
+    {
+      "epoch": 1.8661844484629295,
+      "grad_norm": 1.2449123859405518,
+      "learning_rate": 1.220533575944033e-07,
+      "loss": 0.1407,
+      "step": 1032
+    },
+    {
+      "epoch": 1.8679927667269438,
+      "grad_norm": 1.49626886844635,
+      "learning_rate": 1.1878994355165207e-07,
+      "loss": 0.2121,
+      "step": 1033
+    },
+    {
+      "epoch": 1.8698010849909585,
+      "grad_norm": 1.4758660793304443,
+      "learning_rate": 1.1557022617676217e-07,
+      "loss": 0.1855,
+      "step": 1034
+    },
+    {
+      "epoch": 1.8716094032549728,
+      "grad_norm": 1.4279465675354004,
+      "learning_rate": 1.1239423429269203e-07,
+      "loss": 0.1983,
+      "step": 1035
+    },
+    {
+      "epoch": 1.8734177215189873,
+      "grad_norm": 1.5149667263031006,
+      "learning_rate": 1.0926199633097156e-07,
+      "loss": 0.2079,
+      "step": 1036
+    },
+    {
+      "epoch": 1.8752260397830018,
+      "grad_norm": 1.47335946559906,
+      "learning_rate": 1.0617354033144289e-07,
+      "loss": 0.215,
+      "step": 1037
+    },
+    {
+      "epoch": 1.8770343580470161,
+      "grad_norm": 1.147404432296753,
+      "learning_rate": 1.031288939420122e-07,
+      "loss": 0.1251,
+      "step": 1038
+    },
+    {
+      "epoch": 1.8788426763110309,
+      "grad_norm": 1.3628894090652466,
+      "learning_rate": 1.0012808441839994e-07,
+      "loss": 0.1638,
+      "step": 1039
+    },
+    {
+      "epoch": 1.8806509945750451,
+      "grad_norm": 1.299794316291809,
+      "learning_rate": 9.717113862389993e-08,
+      "loss": 0.1765,
+      "step": 1040
+    },
+    {
+      "epoch": 1.8824593128390597,
+      "grad_norm": 1.4835070371627808,
+      "learning_rate": 9.42580830291373e-08,
+      "loss": 0.2342,
+      "step": 1041
+    },
+    {
+      "epoch": 1.8842676311030742,
+      "grad_norm": 1.4349442720413208,
+      "learning_rate": 9.138894371182983e-08,
+      "loss": 0.2125,
+      "step": 1042
+    },
+    {
+      "epoch": 1.8860759493670884,
+      "grad_norm": 1.4553990364074707,
+      "learning_rate": 8.856374635655696e-08,
+      "loss": 0.1792,
+      "step": 1043
+    },
+    {
+      "epoch": 1.8878842676311032,
+      "grad_norm": 1.5098941326141357,
+      "learning_rate": 8.57825162545295e-08,
+      "loss": 0.1675,
+      "step": 1044
+    },
+    {
+      "epoch": 1.8896925858951175,
+      "grad_norm": 1.490625262260437,
+      "learning_rate": 8.304527830336251e-08,
+      "loss": 0.1983,
+      "step": 1045
+    },
+    {
+      "epoch": 1.891500904159132,
+      "grad_norm": 1.379252552986145,
+      "learning_rate": 8.035205700685167e-08,
+      "loss": 0.1606,
+      "step": 1046
+    },
+    {
+      "epoch": 1.8933092224231465,
+      "grad_norm": 1.3355510234832764,
+      "learning_rate": 7.770287647475672e-08,
+      "loss": 0.1503,
+      "step": 1047
+    },
+    {
+      "epoch": 1.895117540687161,
+      "grad_norm": 1.3615237474441528,
+      "learning_rate": 7.509776042258166e-08,
+      "loss": 0.1521,
+      "step": 1048
+    },
+    {
+      "epoch": 1.8969258589511755,
+      "grad_norm": 1.5093348026275635,
+      "learning_rate": 7.253673217136659e-08,
+      "loss": 0.1781,
+      "step": 1049
+    },
+    {
+      "epoch": 1.8987341772151898,
+      "grad_norm": 1.4346864223480225,
+      "learning_rate": 7.001981464747565e-08,
+      "loss": 0.152,
+      "step": 1050
+    },
+    {
+      "epoch": 1.9005424954792043,
+      "grad_norm": 1.5285435914993286,
+      "learning_rate": 6.75470303823933e-08,
+      "loss": 0.1815,
+      "step": 1051
+    },
+    {
+      "epoch": 1.9023508137432188,
+      "grad_norm": 1.6307448148727417,
+      "learning_rate": 6.511840151252169e-08,
+      "loss": 0.2082,
+      "step": 1052
+    },
+    {
+      "epoch": 1.9041591320072333,
+      "grad_norm": 1.6566197872161865,
+      "learning_rate": 6.273394977898528e-08,
+      "loss": 0.2519,
+      "step": 1053
+    },
+    {
+      "epoch": 1.9059674502712478,
+      "grad_norm": 1.3278251886367798,
+      "learning_rate": 6.039369652743266e-08,
+      "loss": 0.1927,
+      "step": 1054
+    },
+    {
+      "epoch": 1.907775768535262,
+      "grad_norm": 1.1688714027404785,
+      "learning_rate": 5.809766270784667e-08,
+      "loss": 0.1433,
+      "step": 1055
+    },
+    {
+      "epoch": 1.9095840867992768,
+      "grad_norm": 1.3987767696380615,
+      "learning_rate": 5.584586887435739e-08,
+      "loss": 0.1864,
+      "step": 1056
+    },
+    {
+      "epoch": 1.9113924050632911,
+      "grad_norm": 1.4437639713287354,
+      "learning_rate": 5.363833518505834e-08,
+      "loss": 0.196,
+      "step": 1057
+    },
+    {
+      "epoch": 1.9132007233273056,
+      "grad_norm": 1.4916951656341553,
+      "learning_rate": 5.1475081401825553e-08,
+      "loss": 0.1743,
+      "step": 1058
+    },
+    {
+      "epoch": 1.9150090415913201,
+      "grad_norm": 1.4764553308486938,
+      "learning_rate": 4.9356126890139356e-08,
+      "loss": 0.2376,
+      "step": 1059
+    },
+    {
+      "epoch": 1.9168173598553344,
+      "grad_norm": 1.5004583597183228,
+      "learning_rate": 4.7281490618914516e-08,
+      "loss": 0.2351,
+      "step": 1060
+    },
+    {
+      "epoch": 1.9186256781193491,
+      "grad_norm": 1.1736469268798828,
+      "learning_rate": 4.52511911603265e-08,
+      "loss": 0.1474,
+      "step": 1061
+    },
+    {
+      "epoch": 1.9204339963833634,
+      "grad_norm": 1.4233418703079224,
+      "learning_rate": 4.32652466896466e-08,
+      "loss": 0.1801,
+      "step": 1062
+    },
+    {
+      "epoch": 1.922242314647378,
+      "grad_norm": 1.3795275688171387,
+      "learning_rate": 4.13236749850815e-08,
+      "loss": 0.1719,
+      "step": 1063
+    },
+    {
+      "epoch": 1.9240506329113924,
+      "grad_norm": 1.4570521116256714,
+      "learning_rate": 3.9426493427611177e-08,
+      "loss": 0.1823,
+      "step": 1064
+    },
+    {
+      "epoch": 1.9258589511754067,
+      "grad_norm": 1.4998348951339722,
+      "learning_rate": 3.7573719000832954e-08,
+      "loss": 0.2146,
+      "step": 1065
+    },
+    {
+      "epoch": 1.9276672694394215,
+      "grad_norm": 1.483872652053833,
+      "learning_rate": 3.576536829081323e-08,
+      "loss": 0.1949,
+      "step": 1066
+    },
+    {
+      "epoch": 1.9294755877034357,
+      "grad_norm": 1.3987336158752441,
+      "learning_rate": 3.400145748593542e-08,
+      "loss": 0.1459,
+      "step": 1067
+    },
+    {
+      "epoch": 1.9312839059674503,
+      "grad_norm": 1.3769676685333252,
+      "learning_rate": 3.2282002376756163e-08,
+      "loss": 0.151,
+      "step": 1068
+    },
+    {
+      "epoch": 1.9330922242314648,
+      "grad_norm": 1.3427252769470215,
+      "learning_rate": 3.0607018355864326e-08,
+      "loss": 0.1649,
+      "step": 1069
+    },
+    {
+      "epoch": 1.934900542495479,
+      "grad_norm": 1.5869094133377075,
+      "learning_rate": 2.8976520417742794e-08,
+      "loss": 0.1515,
+      "step": 1070
+    },
+    {
+      "epoch": 1.9367088607594938,
+      "grad_norm": 1.3570787906646729,
+      "learning_rate": 2.7390523158633552e-08,
+      "loss": 0.1441,
+      "step": 1071
+    },
+    {
+      "epoch": 1.938517179023508,
+      "grad_norm": 1.270442008972168,
+      "learning_rate": 2.584904077640893e-08,
+      "loss": 0.1294,
+      "step": 1072
+    },
+    {
+      "epoch": 1.9403254972875226,
+      "grad_norm": 1.5576837062835693,
+      "learning_rate": 2.43520870704439e-08,
+      "loss": 0.1987,
+      "step": 1073
+    },
+    {
+      "epoch": 1.942133815551537,
+      "grad_norm": 1.3428254127502441,
+      "learning_rate": 2.2899675441490078e-08,
+      "loss": 0.1622,
+      "step": 1074
+    },
+    {
+      "epoch": 1.9439421338155516,
+      "grad_norm": 1.6970995664596558,
+      "learning_rate": 2.1491818891559156e-08,
+      "loss": 0.242,
+      "step": 1075
+    },
+    {
+      "epoch": 1.945750452079566,
+      "grad_norm": 1.391972541809082,
+      "learning_rate": 2.012853002380466e-08,
+      "loss": 0.1615,
+      "step": 1076
+    },
+    {
+      "epoch": 1.9475587703435804,
+      "grad_norm": 1.4631078243255615,
+      "learning_rate": 1.8809821042410358e-08,
+      "loss": 0.1719,
+      "step": 1077
+    },
+    {
+      "epoch": 1.9493670886075949,
+      "grad_norm": 1.5060595273971558,
+      "learning_rate": 1.753570375247815e-08,
+      "loss": 0.2042,
+      "step": 1078
+    },
+    {
+      "epoch": 1.9511754068716094,
+      "grad_norm": 1.3056021928787231,
+      "learning_rate": 1.630618955992702e-08,
+      "loss": 0.1851,
+      "step": 1079
+    },
+    {
+      "epoch": 1.952983725135624,
+      "grad_norm": 1.3215388059616089,
+      "learning_rate": 1.5121289471385915e-08,
+      "loss": 0.1705,
+      "step": 1080
+    },
+    {
+      "epoch": 1.9547920433996384,
+      "grad_norm": 1.4620890617370605,
+      "learning_rate": 1.3981014094099354e-08,
+      "loss": 0.1952,
+      "step": 1081
+    },
+    {
+      "epoch": 1.9566003616636527,
+      "grad_norm": 1.6308116912841797,
+      "learning_rate": 1.2885373635829756e-08,
+      "loss": 0.2325,
+      "step": 1082
+    },
+    {
+      "epoch": 1.9584086799276674,
+      "grad_norm": 1.5802830457687378,
+      "learning_rate": 1.1834377904768046e-08,
+      "loss": 0.264,
+      "step": 1083
+    },
+    {
+      "epoch": 1.9602169981916817,
+      "grad_norm": 1.3415133953094482,
+      "learning_rate": 1.0828036309443735e-08,
+      "loss": 0.1476,
+      "step": 1084
+    },
+    {
+      "epoch": 1.9620253164556962,
+      "grad_norm": 1.47640061378479,
+      "learning_rate": 9.866357858642206e-09,
+      "loss": 0.1971,
+      "step": 1085
+    },
+    {
+      "epoch": 1.9638336347197107,
+      "grad_norm": 1.5643129348754883,
+      "learning_rate": 8.949351161324227e-09,
+      "loss": 0.2245,
+      "step": 1086
+    },
+    {
+      "epoch": 1.965641952983725,
+      "grad_norm": 1.3774830102920532,
+      "learning_rate": 8.077024426547675e-09,
+      "loss": 0.1857,
+      "step": 1087
+    },
+    {
+      "epoch": 1.9674502712477397,
+      "grad_norm": 1.5181736946105957,
+      "learning_rate": 7.249385463395375e-09,
+      "loss": 0.1833,
+      "step": 1088
+    },
+    {
+      "epoch": 1.969258589511754,
+      "grad_norm": 1.389775276184082,
+      "learning_rate": 6.46644168090349e-09,
+      "loss": 0.1817,
+      "step": 1089
+    },
+    {
+      "epoch": 1.9710669077757685,
+      "grad_norm": 1.3835821151733398,
+      "learning_rate": 5.728200087997126e-09,
+      "loss": 0.1965,
+      "step": 1090
+    },
+    {
+      "epoch": 1.972875226039783,
+      "grad_norm": 1.4697110652923584,
+      "learning_rate": 5.034667293427053e-09,
+      "loss": 0.2245,
+      "step": 1091
+    },
+    {
+      "epoch": 1.9746835443037973,
+      "grad_norm": 1.4795719385147095,
+      "learning_rate": 4.385849505708084e-09,
+      "loss": 0.2116,
+      "step": 1092
+    },
+    {
+      "epoch": 1.976491862567812,
+      "grad_norm": 1.6337482929229736,
+      "learning_rate": 3.781752533068561e-09,
+      "loss": 0.2097,
+      "step": 1093
+    },
+    {
+      "epoch": 1.9783001808318263,
+      "grad_norm": 1.4029808044433594,
+      "learning_rate": 3.2223817833931803e-09,
+      "loss": 0.165,
+      "step": 1094
+    },
+    {
+      "epoch": 1.9801084990958409,
+      "grad_norm": 1.5177561044692993,
+      "learning_rate": 2.707742264178026e-09,
+      "loss": 0.1931,
+      "step": 1095
+    },
+    {
+      "epoch": 1.9819168173598554,
+      "grad_norm": 1.3459399938583374,
+      "learning_rate": 2.237838582483387e-09,
+      "loss": 0.1714,
+      "step": 1096
+    },
+    {
+      "epoch": 1.9837251356238697,
+      "grad_norm": 1.4402574300765991,
+      "learning_rate": 1.8126749448943437e-09,
+      "loss": 0.1958,
+      "step": 1097
+    },
+    {
+      "epoch": 1.9855334538878844,
+      "grad_norm": 1.4353395700454712,
+      "learning_rate": 1.4322551574830202e-09,
+      "loss": 0.153,
+      "step": 1098
+    },
+    {
+      "epoch": 1.9873417721518987,
+      "grad_norm": 1.5452730655670166,
+      "learning_rate": 1.096582625772502e-09,
+      "loss": 0.2276,
+      "step": 1099
+    },
+    {
+      "epoch": 1.9891500904159132,
+      "grad_norm": 1.3681756258010864,
+      "learning_rate": 8.056603547090813e-10,
+      "loss": 0.1353,
+      "step": 1100
+    },
+    {
+      "epoch": 1.9909584086799277,
+      "grad_norm": 1.120079755783081,
+      "learning_rate": 5.594909486328348e-10,
+      "loss": 0.1033,
+      "step": 1101
+    },
+    {
+      "epoch": 1.992766726943942,
+      "grad_norm": 1.3574172258377075,
+      "learning_rate": 3.580766112565304e-10,
+      "loss": 0.1766,
+      "step": 1102
+    },
+    {
+      "epoch": 1.9945750452079567,
+      "grad_norm": 1.516481876373291,
+      "learning_rate": 2.0141914564453247e-10,
+      "loss": 0.2305,
+      "step": 1103
+    },
+    {
+      "epoch": 1.996383363471971,
+      "grad_norm": 1.2223807573318481,
+      "learning_rate": 8.951995419614889e-11,
+      "loss": 0.1527,
+      "step": 1104
+    },
+    {
+      "epoch": 1.9981916817359855,
+      "grad_norm": 1.495068907737732,
+      "learning_rate": 2.2380038635638758e-11,
+      "loss": 0.1962,
+      "step": 1105
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.2624982595443726,
+      "learning_rate": 0.0,
+      "loss": 0.1585,
+      "step": 1106
+    },
+    {
+      "epoch": 2.0,
+      "step": 1106,
+      "total_flos": 329840792174592.0,
+      "train_loss": 0.33624019250138976,
+      "train_runtime": 8393.9432,
+      "train_samples_per_second": 2.106,
+      "train_steps_per_second": 0.132
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1106,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "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": 329840792174592.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}