diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
--- "a/last-checkpoint/trainer_state.json"
+++ "b/last-checkpoint/trainer_state.json"
@@ -38,6974 +38,6982 @@
     },
     {
       "epoch": 0.0022259321090706734,
-      "grad_norm": 2.553895950317383,
+      "grad_norm": 3.0090935230255127,
       "learning_rate": 0.00019999950652018584,
       "loss": 4.1176,
       "step": 5
     },
     {
       "epoch": 0.002671118530884808,
-      "grad_norm": 3.750920534133911,
+      "grad_norm": 5.011908531188965,
       "learning_rate": 0.0001999980260856137,
-      "loss": 4.3848,
+      "loss": 4.3713,
       "step": 6
     },
     {
       "epoch": 0.0031163049526989426,
-      "grad_norm": 4.422790050506592,
-      "learning_rate": 0.000199995558710895,
-      "loss": 5.3942,
+      "grad_norm": NaN,
+      "learning_rate": 0.0001999980260856137,
+      "loss": 5.3384,
       "step": 7
     },
     {
       "epoch": 0.003561491374513077,
       "grad_norm": NaN,
-      "learning_rate": 0.000199995558710895,
-      "loss": 3.3069,
+      "learning_rate": 0.0001999980260856137,
+      "loss": 3.6488,
       "step": 8
     },
     {
       "epoch": 0.004006677796327212,
-      "grad_norm": 6.551366329193115,
-      "learning_rate": 0.00019999210442038162,
-      "loss": 3.3369,
+      "grad_norm": 4.769114971160889,
+      "learning_rate": 0.000199995558710895,
+      "loss": 3.5996,
       "step": 9
     },
     {
       "epoch": 0.004451864218141347,
-      "grad_norm": 5.16669225692749,
-      "learning_rate": 0.00019998766324816607,
-      "loss": 4.0822,
+      "grad_norm": 6.694155216217041,
+      "learning_rate": 0.00019999210442038162,
+      "loss": 4.4521,
       "step": 10
     },
     {
       "epoch": 0.004897050639955481,
-      "grad_norm": 4.758444309234619,
-      "learning_rate": 0.0001999822352380809,
-      "loss": 3.1787,
+      "grad_norm": 3.8457248210906982,
+      "learning_rate": 0.00019998766324816607,
+      "loss": 3.4679,
       "step": 11
     },
     {
       "epoch": 0.005342237061769616,
-      "grad_norm": 7.418261528015137,
-      "learning_rate": 0.00019997582044369843,
-      "loss": 3.594,
+      "grad_norm": 7.944019794464111,
+      "learning_rate": 0.0001999822352380809,
+      "loss": 4.2213,
       "step": 12
     },
     {
       "epoch": 0.005787423483583751,
-      "grad_norm": 6.103316307067871,
-      "learning_rate": 0.00019996841892833,
-      "loss": 2.7363,
+      "grad_norm": 5.603764057159424,
+      "learning_rate": 0.00019997582044369843,
+      "loss": 3.1073,
       "step": 13
     },
     {
       "epoch": 0.006232609905397885,
-      "grad_norm": 5.146238803863525,
-      "learning_rate": 0.00019996003076502565,
-      "loss": 2.4596,
+      "grad_norm": 5.0515265464782715,
+      "learning_rate": 0.00019996841892833,
+      "loss": 2.6351,
       "step": 14
     },
     {
       "epoch": 0.00667779632721202,
-      "grad_norm": 6.732138633728027,
-      "learning_rate": 0.00019995065603657316,
-      "loss": 3.1536,
+      "grad_norm": 4.5042572021484375,
+      "learning_rate": 0.00019996003076502565,
+      "loss": 3.238,
       "step": 15
     },
     {
       "epoch": 0.007122982749026154,
-      "grad_norm": 6.138510227203369,
-      "learning_rate": 0.0001999402948354973,
-      "loss": 2.5301,
+      "grad_norm": 6.955883502960205,
+      "learning_rate": 0.00019995065603657316,
+      "loss": 2.5309,
       "step": 16
     },
     {
       "epoch": 0.007568169170840289,
-      "grad_norm": 7.237410068511963,
-      "learning_rate": 0.00019992894726405893,
-      "loss": 2.5562,
+      "grad_norm": 5.884995937347412,
+      "learning_rate": 0.0001999402948354973,
+      "loss": 2.6348,
       "step": 17
     },
     {
       "epoch": 0.008013355592654424,
-      "grad_norm": 5.183891773223877,
-      "learning_rate": 0.000199916613434254,
-      "loss": 3.6864,
+      "grad_norm": 5.377790927886963,
+      "learning_rate": 0.00019992894726405893,
+      "loss": 3.7344,
       "step": 18
     },
     {
       "epoch": 0.00845854201446856,
-      "grad_norm": 4.217153072357178,
-      "learning_rate": 0.0001999032934678125,
-      "loss": 2.2188,
+      "grad_norm": 4.829436302185059,
+      "learning_rate": 0.000199916613434254,
+      "loss": 2.2568,
       "step": 19
     },
     {
       "epoch": 0.008903728436282694,
-      "grad_norm": 4.953031063079834,
-      "learning_rate": 0.00019988898749619702,
-      "loss": 2.159,
+      "grad_norm": 4.930129051208496,
+      "learning_rate": 0.0001999032934678125,
+      "loss": 2.2034,
       "step": 20
     },
     {
       "epoch": 0.009348914858096828,
-      "grad_norm": 4.839296340942383,
-      "learning_rate": 0.00019987369566060176,
-      "loss": 2.3152,
+      "grad_norm": 4.200410842895508,
+      "learning_rate": 0.00019988898749619702,
+      "loss": 2.2395,
       "step": 21
     },
     {
       "epoch": 0.009794101279910962,
-      "grad_norm": 4.408138275146484,
-      "learning_rate": 0.00019985741811195097,
-      "loss": 2.2196,
+      "grad_norm": 3.6214070320129395,
+      "learning_rate": 0.00019987369566060176,
+      "loss": 2.1656,
       "step": 22
     },
     {
       "epoch": 0.010239287701725098,
-      "grad_norm": 3.8916175365448,
-      "learning_rate": 0.00019984015501089752,
-      "loss": 2.4175,
+      "grad_norm": 3.9661879539489746,
+      "learning_rate": 0.00019985741811195097,
+      "loss": 2.4032,
       "step": 23
     },
     {
       "epoch": 0.010684474123539232,
-      "grad_norm": 4.368136882781982,
-      "learning_rate": 0.0001998219065278212,
-      "loss": 2.5028,
+      "grad_norm": 4.426929473876953,
+      "learning_rate": 0.00019984015501089752,
+      "loss": 2.4649,
       "step": 24
     },
     {
       "epoch": 0.011129660545353366,
-      "grad_norm": 6.353954792022705,
-      "learning_rate": 0.00019980267284282717,
-      "loss": 2.6495,
+      "grad_norm": 6.472476005554199,
+      "learning_rate": 0.0001998219065278212,
+      "loss": 2.5907,
       "step": 25
     },
     {
       "epoch": 0.011574846967167502,
-      "grad_norm": 4.783512115478516,
-      "learning_rate": 0.00019978245414574417,
-      "loss": 1.9392,
+      "grad_norm": 6.279055118560791,
+      "learning_rate": 0.00019980267284282717,
+      "loss": 1.987,
       "step": 26
     },
     {
       "epoch": 0.012020033388981636,
-      "grad_norm": 3.656338930130005,
-      "learning_rate": 0.00019976125063612252,
-      "loss": 1.8599,
+      "grad_norm": 3.898387908935547,
+      "learning_rate": 0.00019978245414574417,
+      "loss": 1.93,
       "step": 27
     },
     {
       "epoch": 0.01246521981079577,
-      "grad_norm": 5.02943229675293,
-      "learning_rate": 0.00019973906252323238,
-      "loss": 2.272,
+      "grad_norm": 8.439789772033691,
+      "learning_rate": 0.00019976125063612252,
+      "loss": 2.1414,
       "step": 28
     },
     {
       "epoch": 0.012910406232609905,
-      "grad_norm": 3.1446378231048584,
-      "learning_rate": 0.0001997158900260614,
-      "loss": 2.2221,
+      "grad_norm": 3.3945260047912598,
+      "learning_rate": 0.00019973906252323238,
+      "loss": 2.2744,
       "step": 29
     },
     {
       "epoch": 0.01335559265442404,
-      "grad_norm": 2.948500871658325,
-      "learning_rate": 0.0001996917333733128,
-      "loss": 1.7518,
+      "grad_norm": 3.3326220512390137,
+      "learning_rate": 0.0001997158900260614,
+      "loss": 1.7339,
       "step": 30
     },
     {
       "epoch": 0.013800779076238175,
-      "grad_norm": 3.700584650039673,
-      "learning_rate": 0.00019966659280340297,
-      "loss": 2.2607,
+      "grad_norm": 3.9388749599456787,
+      "learning_rate": 0.0001996917333733128,
+      "loss": 2.3125,
       "step": 31
     },
     {
       "epoch": 0.014245965498052309,
-      "grad_norm": 4.679660797119141,
-      "learning_rate": 0.00019964046856445924,
-      "loss": 1.4464,
+      "grad_norm": 4.969825744628906,
+      "learning_rate": 0.00019966659280340297,
+      "loss": 1.5139,
       "step": 32
     },
     {
       "epoch": 0.014691151919866445,
-      "grad_norm": 13.211495399475098,
-      "learning_rate": 0.00019961336091431727,
-      "loss": 2.8803,
+      "grad_norm": 4.162886619567871,
+      "learning_rate": 0.00019964046856445924,
+      "loss": 2.8482,
       "step": 33
     },
     {
       "epoch": 0.015136338341680579,
-      "grad_norm": 2.2771530151367188,
-      "learning_rate": 0.00019958527012051857,
-      "loss": 1.5439,
+      "grad_norm": 2.171025037765503,
+      "learning_rate": 0.00019961336091431727,
+      "loss": 1.5546,
       "step": 34
     },
     {
       "epoch": 0.015581524763494713,
-      "grad_norm": 4.475869178771973,
-      "learning_rate": 0.00019955619646030802,
-      "loss": 2.488,
+      "grad_norm": 4.523052215576172,
+      "learning_rate": 0.00019958527012051857,
+      "loss": 2.5814,
       "step": 35
     },
     {
       "epoch": 0.016026711185308847,
-      "grad_norm": 3.8205974102020264,
-      "learning_rate": 0.00019952614022063084,
-      "loss": 2.0277,
+      "grad_norm": 3.6901893615722656,
+      "learning_rate": 0.00019955619646030802,
+      "loss": 1.8633,
       "step": 36
     },
     {
       "epoch": 0.016471897607122983,
-      "grad_norm": 3.721785545349121,
-      "learning_rate": 0.00019949510169813003,
-      "loss": 2.293,
+      "grad_norm": 3.451967477798462,
+      "learning_rate": 0.00019952614022063084,
+      "loss": 2.2273,
       "step": 37
     },
     {
       "epoch": 0.01691708402893712,
-      "grad_norm": 5.029823303222656,
-      "learning_rate": 0.00019946308119914323,
-      "loss": 2.5691,
+      "grad_norm": 5.3913164138793945,
+      "learning_rate": 0.00019949510169813003,
+      "loss": 2.5587,
       "step": 38
     },
     {
       "epoch": 0.01736227045075125,
-      "grad_norm": 3.547927141189575,
-      "learning_rate": 0.0001994300790396999,
-      "loss": 1.4268,
+      "grad_norm": 3.112423896789551,
+      "learning_rate": 0.00019946308119914323,
+      "loss": 1.417,
       "step": 39
     },
     {
       "epoch": 0.017807456872565387,
-      "grad_norm": 4.380621910095215,
-      "learning_rate": 0.000199396095545518,
-      "loss": 2.2319,
+      "grad_norm": 4.377839088439941,
+      "learning_rate": 0.0001994300790396999,
+      "loss": 2.1685,
       "step": 40
     },
     {
       "epoch": 0.01825264329437952,
-      "grad_norm": 3.113034248352051,
-      "learning_rate": 0.00019936113105200085,
-      "loss": 2.0469,
+      "grad_norm": 3.380156993865967,
+      "learning_rate": 0.000199396095545518,
+      "loss": 2.0514,
       "step": 41
     },
     {
       "epoch": 0.018697829716193656,
-      "grad_norm": 3.8902220726013184,
-      "learning_rate": 0.00019932518590423394,
-      "loss": 2.7515,
+      "grad_norm": 4.578589916229248,
+      "learning_rate": 0.00019936113105200085,
+      "loss": 2.7112,
       "step": 42
     },
     {
       "epoch": 0.01914301613800779,
-      "grad_norm": 6.047741413116455,
-      "learning_rate": 0.00019928826045698136,
-      "loss": 2.1295,
+      "grad_norm": 7.042427062988281,
+      "learning_rate": 0.00019932518590423394,
+      "loss": 2.165,
       "step": 43
     },
     {
       "epoch": 0.019588202559821924,
-      "grad_norm": 1.9819254875183105,
-      "learning_rate": 0.0001992503550746824,
-      "loss": 1.1775,
+      "grad_norm": 1.9593662023544312,
+      "learning_rate": 0.00019928826045698136,
+      "loss": 1.1432,
       "step": 44
     },
     {
       "epoch": 0.02003338898163606,
-      "grad_norm": 3.1958131790161133,
-      "learning_rate": 0.0001992114701314478,
-      "loss": 1.9084,
+      "grad_norm": 3.4116411209106445,
+      "learning_rate": 0.0001992503550746824,
+      "loss": 1.9726,
       "step": 45
     },
     {
       "epoch": 0.020478575403450196,
-      "grad_norm": 5.241977691650391,
-      "learning_rate": 0.0001991716060110563,
-      "loss": 3.0362,
+      "grad_norm": 6.172439098358154,
+      "learning_rate": 0.0001992114701314478,
+      "loss": 3.1525,
       "step": 46
     },
     {
       "epoch": 0.020923761825264328,
-      "grad_norm": 2.977036237716675,
-      "learning_rate": 0.00019913076310695068,
-      "loss": 1.704,
+      "grad_norm": 3.173847198486328,
+      "learning_rate": 0.0001991716060110563,
+      "loss": 1.7376,
       "step": 47
     },
     {
       "epoch": 0.021368948247078464,
-      "grad_norm": 4.862756729125977,
-      "learning_rate": 0.00019908894182223388,
-      "loss": 1.9846,
+      "grad_norm": 5.255223751068115,
+      "learning_rate": 0.00019913076310695068,
+      "loss": 1.9631,
       "step": 48
     },
     {
       "epoch": 0.0218141346688926,
-      "grad_norm": 2.590459108352661,
-      "learning_rate": 0.00019904614256966512,
-      "loss": 2.2383,
+      "grad_norm": 3.278393268585205,
+      "learning_rate": 0.00019908894182223388,
+      "loss": 2.2257,
       "step": 49
     },
     {
       "epoch": 0.022259321090706732,
-      "grad_norm": 4.649803161621094,
-      "learning_rate": 0.00019900236577165576,
-      "loss": 2.0824,
+      "grad_norm": 7.197827339172363,
+      "learning_rate": 0.00019904614256966512,
+      "loss": 2.1086,
       "step": 50
     },
     {
       "epoch": 0.02270450751252087,
-      "grad_norm": 4.012078285217285,
-      "learning_rate": 0.0001989576118602651,
-      "loss": 1.8357,
+      "grad_norm": 3.401571273803711,
+      "learning_rate": 0.00019900236577165576,
+      "loss": 1.8664,
       "step": 51
     },
     {
       "epoch": 0.023149693934335004,
-      "grad_norm": 2.382760763168335,
-      "learning_rate": 0.00019891188127719618,
-      "loss": 1.4587,
+      "grad_norm": 3.7931394577026367,
+      "learning_rate": 0.0001989576118602651,
+      "loss": 1.4463,
       "step": 52
     },
     {
       "epoch": 0.023594880356149137,
-      "grad_norm": 3.601736545562744,
-      "learning_rate": 0.0001988651744737914,
-      "loss": 2.118,
+      "grad_norm": 3.639777421951294,
+      "learning_rate": 0.00019891188127719618,
+      "loss": 2.1317,
       "step": 53
     },
     {
       "epoch": 0.024040066777963272,
-      "grad_norm": 3.790908098220825,
-      "learning_rate": 0.00019881749191102808,
-      "loss": 2.4478,
+      "grad_norm": 3.713106155395508,
+      "learning_rate": 0.0001988651744737914,
+      "loss": 2.371,
       "step": 54
     },
     {
       "epoch": 0.02448525319977741,
-      "grad_norm": 5.634518623352051,
-      "learning_rate": 0.00019876883405951377,
-      "loss": 2.0788,
+      "grad_norm": 6.107675552368164,
+      "learning_rate": 0.00019881749191102808,
+      "loss": 2.0907,
       "step": 55
     },
     {
       "epoch": 0.02493043962159154,
-      "grad_norm": 2.9153330326080322,
-      "learning_rate": 0.00019871920139948192,
-      "loss": 1.753,
+      "grad_norm": 3.0008957386016846,
+      "learning_rate": 0.00019876883405951377,
+      "loss": 1.7736,
       "step": 56
     },
     {
       "epoch": 0.025375626043405677,
-      "grad_norm": 3.083784341812134,
-      "learning_rate": 0.0001986685944207868,
-      "loss": 2.0111,
+      "grad_norm": 3.111758232116699,
+      "learning_rate": 0.00019871920139948192,
+      "loss": 1.9832,
       "step": 57
     },
     {
       "epoch": 0.02582081246521981,
-      "grad_norm": 3.6852848529815674,
-      "learning_rate": 0.0001986170136228989,
-      "loss": 2.4499,
+      "grad_norm": 3.809453248977661,
+      "learning_rate": 0.0001986685944207868,
+      "loss": 2.4095,
       "step": 58
     },
     {
       "epoch": 0.026265998887033945,
-      "grad_norm": 3.1907293796539307,
-      "learning_rate": 0.00019856445951489982,
-      "loss": 1.6657,
+      "grad_norm": 3.9784367084503174,
+      "learning_rate": 0.0001986170136228989,
+      "loss": 1.7365,
       "step": 59
     },
     {
       "epoch": 0.02671118530884808,
-      "grad_norm": 4.264678955078125,
-      "learning_rate": 0.0001985109326154774,
-      "loss": 2.1187,
+      "grad_norm": 4.848536014556885,
+      "learning_rate": 0.00019856445951489982,
+      "loss": 2.1358,
       "step": 60
     },
     {
       "epoch": 0.027156371730662213,
-      "grad_norm": 2.061347484588623,
-      "learning_rate": 0.00019845643345292054,
-      "loss": 1.2979,
+      "grad_norm": 2.1415553092956543,
+      "learning_rate": 0.0001985109326154774,
+      "loss": 1.3609,
       "step": 61
     },
     {
       "epoch": 0.02760155815247635,
-      "grad_norm": 2.9782309532165527,
-      "learning_rate": 0.00019840096256511398,
-      "loss": 1.96,
+      "grad_norm": 2.9329774379730225,
+      "learning_rate": 0.00019845643345292054,
+      "loss": 1.9755,
       "step": 62
     },
     {
       "epoch": 0.028046744574290485,
-      "grad_norm": 3.8613250255584717,
-      "learning_rate": 0.00019834452049953297,
-      "loss": 3.4576,
+      "grad_norm": 4.407453536987305,
+      "learning_rate": 0.00019840096256511398,
+      "loss": 3.6484,
       "step": 63
     },
     {
       "epoch": 0.028491930996104618,
-      "grad_norm": 1.9967275857925415,
-      "learning_rate": 0.00019828710781323792,
-      "loss": 0.9779,
+      "grad_norm": 1.9221078157424927,
+      "learning_rate": 0.00019834452049953297,
+      "loss": 0.9992,
       "step": 64
     },
     {
       "epoch": 0.028937117417918753,
-      "grad_norm": 2.790372371673584,
-      "learning_rate": 0.0001982287250728689,
-      "loss": 1.5419,
+      "grad_norm": 3.787074089050293,
+      "learning_rate": 0.00019828710781323792,
+      "loss": 1.6102,
       "step": 65
     },
     {
       "epoch": 0.02938230383973289,
-      "grad_norm": 2.1522772312164307,
-      "learning_rate": 0.0001981693728546399,
-      "loss": 1.3176,
+      "grad_norm": 2.174247980117798,
+      "learning_rate": 0.0001982287250728689,
+      "loss": 1.3549,
       "step": 66
     },
     {
       "epoch": 0.029827490261547022,
-      "grad_norm": 2.084890842437744,
-      "learning_rate": 0.0001981090517443334,
-      "loss": 0.8775,
+      "grad_norm": 1.8840477466583252,
+      "learning_rate": 0.0001981693728546399,
+      "loss": 0.9458,
       "step": 67
     },
     {
       "epoch": 0.030272676683361158,
-      "grad_norm": 2.6281416416168213,
-      "learning_rate": 0.00019804776233729444,
-      "loss": 1.7329,
+      "grad_norm": 2.486109733581543,
+      "learning_rate": 0.0001981090517443334,
+      "loss": 1.7881,
       "step": 68
     },
     {
       "epoch": 0.030717863105175294,
-      "grad_norm": 2.443356990814209,
-      "learning_rate": 0.0001979855052384247,
-      "loss": 1.6412,
+      "grad_norm": 2.583346366882324,
+      "learning_rate": 0.00019804776233729444,
+      "loss": 1.6065,
       "step": 69
     },
     {
       "epoch": 0.031163049526989426,
-      "grad_norm": 3.233210325241089,
-      "learning_rate": 0.00019792228106217658,
-      "loss": 1.9627,
+      "grad_norm": 3.2725138664245605,
+      "learning_rate": 0.0001979855052384247,
+      "loss": 1.8852,
       "step": 70
     },
     {
       "epoch": 0.03160823594880356,
-      "grad_norm": 4.562886714935303,
-      "learning_rate": 0.00019785809043254722,
-      "loss": 1.3901,
+      "grad_norm": 2.918576240539551,
+      "learning_rate": 0.00019792228106217658,
+      "loss": 1.2993,
       "step": 71
     },
     {
       "epoch": 0.032053422370617694,
-      "grad_norm": 4.928150653839111,
-      "learning_rate": 0.0001977929339830722,
-      "loss": 1.8413,
+      "grad_norm": 4.293739318847656,
+      "learning_rate": 0.00019785809043254722,
+      "loss": 1.8252,
       "step": 72
     },
     {
       "epoch": 0.032498608792431834,
-      "grad_norm": 2.872011184692383,
-      "learning_rate": 0.00019772681235681936,
-      "loss": 1.8261,
+      "grad_norm": 3.2163848876953125,
+      "learning_rate": 0.0001977929339830722,
+      "loss": 1.8469,
       "step": 73
     },
     {
       "epoch": 0.032943795214245966,
-      "grad_norm": 4.901350975036621,
-      "learning_rate": 0.00019765972620638248,
-      "loss": 2.5204,
+      "grad_norm": 4.199680328369141,
+      "learning_rate": 0.00019772681235681936,
+      "loss": 2.658,
       "step": 74
     },
     {
       "epoch": 0.0333889816360601,
-      "grad_norm": 4.601914882659912,
-      "learning_rate": 0.00019759167619387476,
-      "loss": 2.3304,
+      "grad_norm": 7.452486991882324,
+      "learning_rate": 0.00019765972620638248,
+      "loss": 2.4333,
       "step": 75
     },
     {
       "epoch": 0.03383416805787424,
-      "grad_norm": 4.285044193267822,
-      "learning_rate": 0.00019752266299092236,
-      "loss": 2.148,
+      "grad_norm": 3.0536255836486816,
+      "learning_rate": 0.00019759167619387476,
+      "loss": 2.1119,
       "step": 76
     },
     {
       "epoch": 0.03427935447968837,
-      "grad_norm": 3.728804111480713,
-      "learning_rate": 0.00019745268727865774,
-      "loss": 2.2936,
+      "grad_norm": 4.044293403625488,
+      "learning_rate": 0.00019752266299092236,
+      "loss": 2.2781,
       "step": 77
     },
     {
       "epoch": 0.0347245409015025,
-      "grad_norm": 3.0641114711761475,
-      "learning_rate": 0.0001973817497477129,
-      "loss": 1.7878,
+      "grad_norm": 2.646242141723633,
+      "learning_rate": 0.00019745268727865774,
+      "loss": 1.7457,
       "step": 78
     },
     {
       "epoch": 0.03516972732331664,
-      "grad_norm": 4.134016990661621,
-      "learning_rate": 0.00019730985109821266,
-      "loss": 2.9227,
+      "grad_norm": 5.419813632965088,
+      "learning_rate": 0.0001973817497477129,
+      "loss": 3.1021,
       "step": 79
     },
     {
       "epoch": 0.035614913745130775,
-      "grad_norm": 12.307600975036621,
-      "learning_rate": 0.00019723699203976766,
-      "loss": 2.7279,
+      "grad_norm": 3.2200138568878174,
+      "learning_rate": 0.00019730985109821266,
+      "loss": 2.5992,
       "step": 80
     },
     {
       "epoch": 0.03606010016694491,
-      "grad_norm": 3.992847204208374,
-      "learning_rate": 0.0001971631732914674,
-      "loss": 2.5772,
+      "grad_norm": 4.368217468261719,
+      "learning_rate": 0.00019723699203976766,
+      "loss": 2.5135,
       "step": 81
     },
     {
       "epoch": 0.03650528658875904,
-      "grad_norm": 3.676469564437866,
-      "learning_rate": 0.0001970883955818731,
-      "loss": 1.5831,
+      "grad_norm": 3.3963205814361572,
+      "learning_rate": 0.0001971631732914674,
+      "loss": 1.7087,
       "step": 82
     },
     {
       "epoch": 0.03695047301057318,
-      "grad_norm": 3.0918328762054443,
-      "learning_rate": 0.0001970126596490106,
-      "loss": 1.4057,
+      "grad_norm": 2.7550411224365234,
+      "learning_rate": 0.0001970883955818731,
+      "loss": 1.307,
       "step": 83
     },
     {
       "epoch": 0.03739565943238731,
-      "grad_norm": 3.945146322250366,
-      "learning_rate": 0.00019693596624036292,
-      "loss": 1.5807,
+      "grad_norm": 3.3400115966796875,
+      "learning_rate": 0.0001970126596490106,
+      "loss": 1.8352,
       "step": 84
     },
     {
       "epoch": 0.037840845854201444,
-      "grad_norm": 2.956975221633911,
-      "learning_rate": 0.0001968583161128631,
-      "loss": 1.9335,
+      "grad_norm": 2.74469256401062,
+      "learning_rate": 0.00019693596624036292,
+      "loss": 1.9022,
       "step": 85
     },
     {
       "epoch": 0.03828603227601558,
-      "grad_norm": 2.6506834030151367,
-      "learning_rate": 0.00019677971003288655,
-      "loss": 1.8569,
+      "grad_norm": 2.932953357696533,
+      "learning_rate": 0.0001968583161128631,
+      "loss": 1.9438,
       "step": 86
     },
     {
       "epoch": 0.038731218697829715,
-      "grad_norm": 3.6588492393493652,
-      "learning_rate": 0.00019670014877624353,
-      "loss": 1.6848,
+      "grad_norm": 4.120993137359619,
+      "learning_rate": 0.00019677971003288655,
+      "loss": 1.8208,
       "step": 87
     },
     {
       "epoch": 0.03917640511964385,
-      "grad_norm": 3.677953004837036,
-      "learning_rate": 0.00019661963312817148,
-      "loss": 2.5729,
+      "grad_norm": 4.330765247344971,
+      "learning_rate": 0.00019670014877624353,
+      "loss": 2.5833,
       "step": 88
     },
     {
       "epoch": 0.03962159154145799,
-      "grad_norm": 3.1620278358459473,
-      "learning_rate": 0.0001965381638833274,
-      "loss": 1.8112,
+      "grad_norm": 3.0520567893981934,
+      "learning_rate": 0.00019661963312817148,
+      "loss": 1.9228,
       "step": 89
     },
     {
       "epoch": 0.04006677796327212,
-      "grad_norm": 4.549400806427002,
-      "learning_rate": 0.00019645574184577982,
-      "loss": 3.0601,
+      "grad_norm": 3.913396120071411,
+      "learning_rate": 0.0001965381638833274,
+      "loss": 2.9555,
       "step": 90
     },
     {
       "epoch": 0.04051196438508625,
-      "grad_norm": 4.790261745452881,
-      "learning_rate": 0.000196372367829001,
-      "loss": 2.0025,
+      "grad_norm": 3.679295539855957,
+      "learning_rate": 0.00019645574184577982,
+      "loss": 1.9205,
       "step": 91
     },
     {
       "epoch": 0.04095715080690039,
-      "grad_norm": 2.0430498123168945,
-      "learning_rate": 0.00019628804265585877,
-      "loss": 1.757,
+      "grad_norm": 2.6744461059570312,
+      "learning_rate": 0.000196372367829001,
+      "loss": 1.8619,
       "step": 92
     },
     {
       "epoch": 0.041402337228714524,
-      "grad_norm": 3.846684217453003,
-      "learning_rate": 0.0001962027671586086,
-      "loss": 2.8248,
+      "grad_norm": 3.623929977416992,
+      "learning_rate": 0.00019628804265585877,
+      "loss": 2.8998,
       "step": 93
     },
     {
       "epoch": 0.041847523650528656,
-      "grad_norm": 2.1094014644622803,
-      "learning_rate": 0.0001961165421788852,
-      "loss": 1.388,
+      "grad_norm": 2.4871249198913574,
+      "learning_rate": 0.0001962027671586086,
+      "loss": 1.41,
       "step": 94
     },
     {
       "epoch": 0.042292710072342796,
-      "grad_norm": 2.475429058074951,
-      "learning_rate": 0.0001960293685676943,
-      "loss": 2.109,
+      "grad_norm": 2.6301159858703613,
+      "learning_rate": 0.0001961165421788852,
+      "loss": 2.1916,
       "step": 95
     },
     {
       "epoch": 0.04273789649415693,
-      "grad_norm": 2.539384603500366,
-      "learning_rate": 0.0001959412471854043,
-      "loss": 1.7985,
+      "grad_norm": 3.1040453910827637,
+      "learning_rate": 0.0001960293685676943,
+      "loss": 1.8362,
       "step": 96
     },
     {
       "epoch": 0.04318308291597106,
-      "grad_norm": 3.7397396564483643,
-      "learning_rate": 0.0001958521789017376,
-      "loss": 2.3767,
+      "grad_norm": 3.9120583534240723,
+      "learning_rate": 0.0001959412471854043,
+      "loss": 2.3609,
       "step": 97
     },
     {
       "epoch": 0.0436282693377852,
-      "grad_norm": 4.037478446960449,
-      "learning_rate": 0.00019576216459576222,
-      "loss": 2.5889,
+      "grad_norm": 4.234198093414307,
+      "learning_rate": 0.0001958521789017376,
+      "loss": 2.615,
       "step": 98
     },
     {
       "epoch": 0.04407345575959933,
-      "grad_norm": 5.122471809387207,
-      "learning_rate": 0.00019567120515588308,
-      "loss": 2.2834,
+      "grad_norm": 3.9530892372131348,
+      "learning_rate": 0.00019576216459576222,
+      "loss": 2.349,
       "step": 99
     },
     {
       "epoch": 0.044518642181413465,
-      "grad_norm": 3.0687732696533203,
-      "learning_rate": 0.00019557930147983302,
-      "loss": 2.6478,
+      "grad_norm": 3.4594247341156006,
+      "learning_rate": 0.00019567120515588308,
+      "loss": 2.6695,
       "step": 100
     },
     {
       "epoch": 0.044963828603227604,
-      "grad_norm": 2.641993284225464,
-      "learning_rate": 0.00019548645447466431,
-      "loss": 1.2885,
+      "grad_norm": 2.3701462745666504,
+      "learning_rate": 0.00019557930147983302,
+      "loss": 1.2688,
       "step": 101
     },
     {
       "epoch": 0.04540901502504174,
-      "grad_norm": 2.480499505996704,
-      "learning_rate": 0.00019539266505673938,
-      "loss": 1.4892,
+      "grad_norm": 2.673154354095459,
+      "learning_rate": 0.00019548645447466431,
+      "loss": 1.5608,
       "step": 102
     },
     {
       "epoch": 0.04585420144685587,
-      "grad_norm": 2.8571128845214844,
-      "learning_rate": 0.00019529793415172192,
-      "loss": 1.5001,
+      "grad_norm": 2.1529808044433594,
+      "learning_rate": 0.00019539266505673938,
+      "loss": 1.4572,
       "step": 103
     },
     {
       "epoch": 0.04629938786867001,
-      "grad_norm": 2.330220937728882,
-      "learning_rate": 0.00019520226269456768,
-      "loss": 1.3509,
+      "grad_norm": 2.52852463722229,
+      "learning_rate": 0.00019529793415172192,
+      "loss": 1.436,
       "step": 104
     },
     {
       "epoch": 0.04674457429048414,
-      "grad_norm": 6.893773555755615,
-      "learning_rate": 0.00019510565162951537,
-      "loss": 1.6334,
+      "grad_norm": 6.402322769165039,
+      "learning_rate": 0.00019520226269456768,
+      "loss": 1.6831,
       "step": 105
     },
     {
       "epoch": 0.04718976071229827,
-      "grad_norm": NaN,
+      "grad_norm": 2.9657320976257324,
       "learning_rate": 0.00019510565162951537,
-      "loss": 1.9998,
+      "loss": 1.9188,
       "step": 106
     },
     {
       "epoch": 0.04763494713411241,
-      "grad_norm": 3.612894058227539,
+      "grad_norm": 3.413729667663574,
       "learning_rate": 0.00019500810191007718,
-      "loss": 2.5597,
+      "loss": 2.6164,
       "step": 107
     },
     {
       "epoch": 0.048080133555926545,
-      "grad_norm": 2.3237907886505127,
+      "grad_norm": 2.1516737937927246,
       "learning_rate": 0.00019490961449902946,
-      "loss": 1.6031,
+      "loss": 1.6051,
       "step": 108
     },
     {
       "epoch": 0.04852531997774068,
-      "grad_norm": 2.755427122116089,
+      "grad_norm": 2.9398021697998047,
       "learning_rate": 0.0001948101903684032,
-      "loss": 1.8851,
+      "loss": 1.9419,
       "step": 109
     },
     {
       "epoch": 0.04897050639955482,
-      "grad_norm": 2.8905320167541504,
+      "grad_norm": 3.36769962310791,
       "learning_rate": 0.00019470983049947444,
-      "loss": 1.5392,
+      "loss": 1.5534,
       "step": 110
     },
     {
       "epoch": 0.04941569282136895,
-      "grad_norm": 10.140717506408691,
+      "grad_norm": 2.4700193405151367,
       "learning_rate": 0.00019460853588275454,
-      "loss": 2.005,
+      "loss": 2.0116,
       "step": 111
     },
     {
       "epoch": 0.04986087924318308,
-      "grad_norm": 3.27565336227417,
+      "grad_norm": 3.086195230484009,
       "learning_rate": 0.00019450630751798048,
-      "loss": 2.2091,
+      "loss": 2.2248,
       "step": 112
     },
     {
       "epoch": 0.05030606566499722,
-      "grad_norm": 4.00056266784668,
+      "grad_norm": 4.200692653656006,
       "learning_rate": 0.000194403146414105,
-      "loss": 3.019,
+      "loss": 3.0367,
       "step": 113
     },
     {
       "epoch": 0.05075125208681135,
-      "grad_norm": 3.5745298862457275,
+      "grad_norm": 3.112865924835205,
       "learning_rate": 0.00019429905358928646,
-      "loss": 2.154,
+      "loss": 2.1513,
       "step": 114
     },
     {
       "epoch": 0.051196438508625486,
-      "grad_norm": 3.2333099842071533,
+      "grad_norm": 2.8886055946350098,
       "learning_rate": 0.00019419403007087907,
-      "loss": 2.1104,
+      "loss": 2.0628,
       "step": 115
     },
     {
       "epoch": 0.05164162493043962,
-      "grad_norm": 2.650325298309326,
+      "grad_norm": 3.146010398864746,
       "learning_rate": 0.00019408807689542257,
-      "loss": 1.1883,
+      "loss": 1.2535,
       "step": 116
     },
     {
       "epoch": 0.05208681135225376,
-      "grad_norm": 2.167717218399048,
+      "grad_norm": 2.0565860271453857,
       "learning_rate": 0.00019398119510863197,
-      "loss": 2.0315,
+      "loss": 1.9781,
       "step": 117
     },
     {
       "epoch": 0.05253199777406789,
-      "grad_norm": 3.0747454166412354,
+      "grad_norm": 3.192265748977661,
       "learning_rate": 0.00019387338576538744,
-      "loss": 1.7789,
+      "loss": 1.7527,
       "step": 118
     },
     {
       "epoch": 0.05297718419588202,
-      "grad_norm": 1.7313055992126465,
+      "grad_norm": 1.8156938552856445,
       "learning_rate": 0.00019376464992972356,
-      "loss": 1.4003,
+      "loss": 1.3864,
       "step": 119
     },
     {
       "epoch": 0.05342237061769616,
-      "grad_norm": 2.92069149017334,
+      "grad_norm": 2.6510698795318604,
       "learning_rate": 0.00019365498867481923,
-      "loss": 1.3967,
+      "loss": 1.3648,
       "step": 120
     },
     {
       "epoch": 0.053867557039510294,
-      "grad_norm": 7.5436553955078125,
+      "grad_norm": 3.3322391510009766,
       "learning_rate": 0.00019354440308298675,
-      "loss": 1.9777,
+      "loss": 2.0549,
       "step": 121
     },
     {
       "epoch": 0.05431274346132443,
-      "grad_norm": 8.587759971618652,
+      "grad_norm": 2.9640331268310547,
       "learning_rate": 0.00019343289424566122,
-      "loss": 1.8453,
+      "loss": 1.7579,
       "step": 122
     },
     {
       "epoch": 0.054757929883138566,
-      "grad_norm": 2.8797378540039062,
+      "grad_norm": 3.904458999633789,
       "learning_rate": 0.00019332046326338986,
-      "loss": 1.8348,
+      "loss": 1.8362,
       "step": 123
     },
     {
       "epoch": 0.0552031163049527,
-      "grad_norm": 3.496337652206421,
+      "grad_norm": 4.279350280761719,
       "learning_rate": 0.0001932071112458211,
-      "loss": 2.2295,
+      "loss": 2.2549,
       "step": 124
     },
     {
       "epoch": 0.05564830272676683,
-      "grad_norm": 4.732133865356445,
+      "grad_norm": 4.328281879425049,
       "learning_rate": 0.00019309283931169356,
-      "loss": 3.2504,
+      "loss": 3.2525,
       "step": 125
     },
     {
       "epoch": 0.05609348914858097,
-      "grad_norm": 3.367631196975708,
+      "grad_norm": 2.8716776371002197,
       "learning_rate": 0.00019297764858882514,
-      "loss": 1.3877,
+      "loss": 1.3097,
       "step": 126
     },
     {
       "epoch": 0.0565386755703951,
-      "grad_norm": 3.1331379413604736,
+      "grad_norm": 3.0296878814697266,
       "learning_rate": 0.00019286154021410173,
-      "loss": 1.775,
+      "loss": 1.7506,
       "step": 127
     },
     {
       "epoch": 0.056983861992209235,
-      "grad_norm": 2.303420305252075,
+      "grad_norm": 2.415759563446045,
       "learning_rate": 0.00019274451533346615,
-      "loss": 2.3335,
+      "loss": 2.2494,
       "step": 128
     },
     {
       "epoch": 0.057429048414023375,
-      "grad_norm": 3.3306965827941895,
+      "grad_norm": 3.0764763355255127,
       "learning_rate": 0.00019262657510190666,
-      "loss": 2.1583,
+      "loss": 2.1633,
       "step": 129
     },
     {
       "epoch": 0.05787423483583751,
-      "grad_norm": 2.900506019592285,
+      "grad_norm": 2.2252795696258545,
       "learning_rate": 0.0001925077206834458,
-      "loss": 1.7225,
+      "loss": 1.7298,
       "step": 130
     },
     {
       "epoch": 0.05831942125765164,
-      "grad_norm": 2.4355456829071045,
+      "grad_norm": 2.305452823638916,
       "learning_rate": 0.0001923879532511287,
-      "loss": 1.4379,
+      "loss": 1.5171,
       "step": 131
     },
     {
       "epoch": 0.05876460767946578,
-      "grad_norm": 6.534120559692383,
+      "grad_norm": 5.750187873840332,
       "learning_rate": 0.0001922672739870115,
-      "loss": 3.3114,
+      "loss": 3.4236,
       "step": 132
     },
     {
       "epoch": 0.05920979410127991,
-      "grad_norm": 4.634522914886475,
+      "grad_norm": 4.094027519226074,
       "learning_rate": 0.00019214568408214985,
-      "loss": 3.683,
+      "loss": 3.6172,
       "step": 133
     },
     {
       "epoch": 0.059654980523094044,
-      "grad_norm": 3.742173671722412,
+      "grad_norm": 4.096864700317383,
       "learning_rate": 0.00019202318473658705,
-      "loss": 2.1639,
+      "loss": 1.8016,
       "step": 134
     },
     {
       "epoch": 0.06010016694490818,
-      "grad_norm": 3.062386989593506,
+      "grad_norm": 2.9913458824157715,
       "learning_rate": 0.00019189977715934213,
-      "loss": 2.2926,
+      "loss": 2.2077,
       "step": 135
     },
     {
       "epoch": 0.060545353366722315,
-      "grad_norm": 3.8586795330047607,
+      "grad_norm": 3.893141508102417,
       "learning_rate": 0.00019177546256839812,
-      "loss": 1.788,
+      "loss": 1.8356,
       "step": 136
     },
     {
       "epoch": 0.06099053978853645,
-      "grad_norm": 3.739783525466919,
+      "grad_norm": 3.7156870365142822,
       "learning_rate": 0.0001916502421906898,
-      "loss": 2.7422,
+      "loss": 2.6727,
       "step": 137
     },
     {
       "epoch": 0.06143572621035059,
-      "grad_norm": 3.0615429878234863,
+      "grad_norm": 2.921959638595581,
       "learning_rate": 0.00019152411726209176,
-      "loss": 2.0941,
+      "loss": 2.133,
       "step": 138
     },
     {
       "epoch": 0.06188091263216472,
-      "grad_norm": 2.7861218452453613,
+      "grad_norm": 1.890904188156128,
       "learning_rate": 0.00019139708902740613,
-      "loss": 1.3494,
+      "loss": 1.2856,
       "step": 139
     },
     {
       "epoch": 0.06232609905397885,
-      "grad_norm": 4.011098384857178,
+      "grad_norm": 4.972781658172607,
       "learning_rate": 0.0001912691587403503,
-      "loss": 2.6336,
+      "loss": 2.561,
       "step": 140
     },
     {
       "epoch": 0.06277128547579298,
-      "grad_norm": 4.350919723510742,
+      "grad_norm": 4.031817436218262,
       "learning_rate": 0.00019114032766354453,
-      "loss": 1.4581,
+      "loss": 1.4916,
       "step": 141
     },
     {
       "epoch": 0.06321647189760712,
-      "grad_norm": 3.9159739017486572,
+      "grad_norm": 3.651207208633423,
       "learning_rate": 0.00019101059706849957,
-      "loss": 2.5859,
+      "loss": 2.6341,
       "step": 142
     },
     {
       "epoch": 0.06366165831942126,
-      "grad_norm": 4.638038158416748,
+      "grad_norm": 4.809238433837891,
       "learning_rate": 0.00019087996823560402,
-      "loss": 2.6958,
+      "loss": 2.7431,
       "step": 143
     },
     {
       "epoch": 0.06410684474123539,
-      "grad_norm": 4.493806838989258,
+      "grad_norm": 3.8684816360473633,
       "learning_rate": 0.0001907484424541117,
-      "loss": 1.7282,
+      "loss": 1.716,
       "step": 144
     },
     {
       "epoch": 0.06455203116304953,
-      "grad_norm": 3.276454448699951,
+      "grad_norm": 4.384403705596924,
       "learning_rate": 0.00019061602102212898,
-      "loss": 1.4519,
+      "loss": 1.511,
       "step": 145
     },
     {
       "epoch": 0.06499721758486367,
-      "grad_norm": 3.375685691833496,
+      "grad_norm": 3.2156827449798584,
       "learning_rate": 0.00019048270524660196,
-      "loss": 2.7348,
+      "loss": 2.7144,
       "step": 146
     },
     {
       "epoch": 0.06544240400667779,
-      "grad_norm": 3.7552807331085205,
+      "grad_norm": 3.423051118850708,
       "learning_rate": 0.0001903484964433035,
-      "loss": 2.4439,
+      "loss": 2.4857,
       "step": 147
     },
     {
       "epoch": 0.06588759042849193,
-      "grad_norm": 3.9915504455566406,
+      "grad_norm": 3.9671273231506348,
       "learning_rate": 0.00019021339593682028,
-      "loss": 2.7966,
+      "loss": 2.7275,
       "step": 148
     },
     {
       "epoch": 0.06633277685030607,
-      "grad_norm": 2.8780786991119385,
+      "grad_norm": 2.884432077407837,
       "learning_rate": 0.00019007740506053983,
-      "loss": 1.7375,
+      "loss": 1.6876,
       "step": 149
     },
     {
       "epoch": 0.0667779632721202,
-      "grad_norm": 2.792412042617798,
+      "grad_norm": 2.6939189434051514,
       "learning_rate": 0.0001899405251566371,
-      "loss": 2.4148,
+      "loss": 2.4106,
       "step": 150
     },
     {
       "epoch": 0.06722314969393434,
-      "grad_norm": 3.3408055305480957,
+      "grad_norm": 3.268193244934082,
       "learning_rate": 0.00018980275757606157,
-      "loss": 2.6953,
+      "loss": 2.7097,
       "step": 151
     },
     {
       "epoch": 0.06766833611574848,
-      "grad_norm": 2.7032229900360107,
+      "grad_norm": 2.9582340717315674,
       "learning_rate": 0.00018966410367852362,
-      "loss": 2.0699,
+      "loss": 2.1443,
       "step": 152
     },
     {
       "epoch": 0.0681135225375626,
-      "grad_norm": 3.501795530319214,
+      "grad_norm": 3.300041913986206,
       "learning_rate": 0.00018952456483248119,
-      "loss": 2.6974,
+      "loss": 2.6703,
       "step": 153
     },
     {
       "epoch": 0.06855870895937674,
-      "grad_norm": 2.4695076942443848,
+      "grad_norm": 2.3364670276641846,
       "learning_rate": 0.0001893841424151264,
-      "loss": 1.7339,
+      "loss": 1.7084,
       "step": 154
     },
     {
       "epoch": 0.06900389538119088,
-      "grad_norm": 4.014305591583252,
+      "grad_norm": 3.9670588970184326,
       "learning_rate": 0.0001892428378123718,
-      "loss": 2.4761,
+      "loss": 2.4215,
       "step": 155
     },
     {
       "epoch": 0.069449081803005,
-      "grad_norm": 2.3375473022460938,
+      "grad_norm": 2.3109562397003174,
       "learning_rate": 0.0001891006524188368,
-      "loss": 2.0071,
+      "loss": 2.001,
       "step": 156
     },
     {
       "epoch": 0.06989426822481914,
-      "grad_norm": 3.323019504547119,
+      "grad_norm": 3.1594605445861816,
       "learning_rate": 0.00018895758763783383,
-      "loss": 2.226,
+      "loss": 2.1514,
       "step": 157
     },
     {
       "epoch": 0.07033945464663328,
-      "grad_norm": 7.093875885009766,
+      "grad_norm": 3.3163321018218994,
       "learning_rate": 0.00018881364488135448,
-      "loss": 1.7551,
+      "loss": 1.7621,
       "step": 158
     },
     {
       "epoch": 0.07078464106844741,
-      "grad_norm": 2.1790101528167725,
+      "grad_norm": 2.604219913482666,
       "learning_rate": 0.00018866882557005567,
-      "loss": 1.7356,
+      "loss": 1.6635,
       "step": 159
     },
     {
       "epoch": 0.07122982749026155,
-      "grad_norm": 4.966728687286377,
+      "grad_norm": 3.6924543380737305,
       "learning_rate": 0.00018852313113324552,
-      "loss": 2.3969,
+      "loss": 2.2035,
       "step": 160
     },
     {
       "epoch": 0.07167501391207569,
-      "grad_norm": 2.207064390182495,
+      "grad_norm": 2.356276035308838,
       "learning_rate": 0.00018837656300886937,
-      "loss": 1.9052,
+      "loss": 1.9358,
       "step": 161
     },
     {
       "epoch": 0.07212020033388981,
-      "grad_norm": 2.988813877105713,
+      "grad_norm": 4.977180004119873,
       "learning_rate": 0.00018822912264349534,
-      "loss": 2.2362,
+      "loss": 2.3076,
       "step": 162
     },
     {
       "epoch": 0.07256538675570395,
-      "grad_norm": 15.190692901611328,
+      "grad_norm": 3.712543487548828,
       "learning_rate": 0.00018808081149230036,
-      "loss": 1.7636,
+      "loss": 1.7823,
       "step": 163
     },
     {
       "epoch": 0.07301057317751808,
-      "grad_norm": 2.8087406158447266,
+      "grad_norm": 2.742771625518799,
       "learning_rate": 0.00018793163101905563,
-      "loss": 2.0533,
+      "loss": 2.1218,
       "step": 164
     },
     {
       "epoch": 0.07345575959933222,
-      "grad_norm": 2.135378122329712,
+      "grad_norm": 2.7484240531921387,
       "learning_rate": 0.00018778158269611218,
-      "loss": 2.2528,
+      "loss": 2.3076,
       "step": 165
     },
     {
       "epoch": 0.07390094602114636,
-      "grad_norm": 3.5980896949768066,
+      "grad_norm": 3.3283402919769287,
       "learning_rate": 0.00018763066800438636,
-      "loss": 2.5453,
+      "loss": 2.5205,
       "step": 166
     },
     {
       "epoch": 0.07434613244296048,
-      "grad_norm": 2.6481971740722656,
+      "grad_norm": 2.589799404144287,
       "learning_rate": 0.0001874788884333453,
-      "loss": 1.8122,
+      "loss": 1.7938,
       "step": 167
     },
     {
       "epoch": 0.07479131886477462,
-      "grad_norm": 2.283841371536255,
+      "grad_norm": 2.3176369667053223,
       "learning_rate": 0.00018732624548099204,
-      "loss": 2.0774,
+      "loss": 2.125,
       "step": 168
     },
     {
       "epoch": 0.07523650528658876,
-      "grad_norm": 2.59877610206604,
+      "grad_norm": 2.814786672592163,
       "learning_rate": 0.0001871727406538509,
-      "loss": 2.0153,
+      "loss": 2.1019,
       "step": 169
     },
     {
       "epoch": 0.07568169170840289,
-      "grad_norm": 4.261568546295166,
+      "grad_norm": 5.515918731689453,
       "learning_rate": 0.0001870183754669526,
-      "loss": 1.56,
+      "loss": 1.652,
       "step": 170
     },
     {
       "epoch": 0.07612687813021703,
-      "grad_norm": 2.588036060333252,
+      "grad_norm": 2.5525944232940674,
       "learning_rate": 0.00018686315144381913,
-      "loss": 2.4187,
+      "loss": 2.3976,
       "step": 171
     },
     {
       "epoch": 0.07657206455203117,
-      "grad_norm": 3.837876796722412,
+      "grad_norm": 3.176581382751465,
       "learning_rate": 0.000186707070116449,
-      "loss": 2.1299,
+      "loss": 2.2002,
       "step": 172
     },
     {
       "epoch": 0.07701725097384529,
-      "grad_norm": 2.0875399112701416,
+      "grad_norm": 4.414255619049072,
       "learning_rate": 0.0001865501330253019,
-      "loss": 2.1619,
+      "loss": 2.2484,
       "step": 173
     },
     {
       "epoch": 0.07746243739565943,
-      "grad_norm": 2.641989231109619,
+      "grad_norm": 2.5007452964782715,
       "learning_rate": 0.00018639234171928353,
-      "loss": 1.7574,
+      "loss": 1.7283,
       "step": 174
     },
     {
       "epoch": 0.07790762381747357,
-      "grad_norm": 3.0749399662017822,
+      "grad_norm": 2.847289800643921,
       "learning_rate": 0.0001862336977557304,
-      "loss": 1.6736,
+      "loss": 1.6054,
       "step": 175
     },
     {
       "epoch": 0.0783528102392877,
-      "grad_norm": 2.718775749206543,
+      "grad_norm": 3.44081449508667,
       "learning_rate": 0.0001860742027003944,
-      "loss": 2.1462,
+      "loss": 2.174,
       "step": 176
     },
     {
       "epoch": 0.07879799666110184,
-      "grad_norm": 2.1774768829345703,
+      "grad_norm": 2.025012254714966,
       "learning_rate": 0.00018591385812742725,
-      "loss": 1.4246,
+      "loss": 1.4577,
       "step": 177
     },
     {
       "epoch": 0.07924318308291597,
-      "grad_norm": 3.1828486919403076,
+      "grad_norm": 2.9062042236328125,
       "learning_rate": 0.00018575266561936523,
-      "loss": 2.3951,
+      "loss": 2.4368,
       "step": 178
     },
     {
       "epoch": 0.0796883695047301,
-      "grad_norm": 2.520548105239868,
+      "grad_norm": 2.5875933170318604,
       "learning_rate": 0.00018559062676711332,
-      "loss": 1.6368,
+      "loss": 1.5715,
       "step": 179
     },
     {
       "epoch": 0.08013355592654424,
-      "grad_norm": 1.9750266075134277,
+      "grad_norm": 1.958118200302124,
       "learning_rate": 0.0001854277431699295,
-      "loss": 1.4846,
+      "loss": 1.3609,
       "step": 180
     },
     {
       "epoch": 0.08057874234835838,
-      "grad_norm": 5.070473670959473,
+      "grad_norm": 3.3691253662109375,
       "learning_rate": 0.00018526401643540922,
-      "loss": 1.8334,
+      "loss": 1.7653,
       "step": 181
     },
     {
       "epoch": 0.0810239287701725,
-      "grad_norm": 2.920872688293457,
+      "grad_norm": 3.036341667175293,
       "learning_rate": 0.00018509944817946922,
-      "loss": 1.6212,
+      "loss": 1.6798,
       "step": 182
     },
     {
       "epoch": 0.08146911519198664,
-      "grad_norm": 2.4279935359954834,
+      "grad_norm": 2.7306416034698486,
       "learning_rate": 0.00018493404002633166,
-      "loss": 1.7948,
+      "loss": 1.8532,
       "step": 183
     },
     {
       "epoch": 0.08191430161380078,
-      "grad_norm": 3.453925609588623,
+      "grad_norm": 3.7763404846191406,
       "learning_rate": 0.00018476779360850832,
-      "loss": 2.1735,
+      "loss": 2.2009,
       "step": 184
     },
     {
       "epoch": 0.08235948803561491,
-      "grad_norm": 1.8669699430465698,
+      "grad_norm": 2.890293836593628,
       "learning_rate": 0.00018460071056678422,
-      "loss": 1.2327,
+      "loss": 1.27,
       "step": 185
     },
     {
       "epoch": 0.08280467445742905,
-      "grad_norm": 2.559237003326416,
+      "grad_norm": 2.616748094558716,
       "learning_rate": 0.00018443279255020152,
-      "loss": 1.6827,
+      "loss": 1.5601,
       "step": 186
     },
     {
       "epoch": 0.08324986087924319,
-      "grad_norm": 3.643465518951416,
+      "grad_norm": 3.860060930252075,
       "learning_rate": 0.00018426404121604323,
-      "loss": 2.3146,
+      "loss": 2.3942,
       "step": 187
     },
     {
       "epoch": 0.08369504730105731,
-      "grad_norm": 3.2583975791931152,
+      "grad_norm": 2.253901243209839,
       "learning_rate": 0.00018409445822981693,
-      "loss": 1.9232,
+      "loss": 1.8034,
       "step": 188
     },
     {
       "epoch": 0.08414023372287145,
-      "grad_norm": 3.837696075439453,
+      "grad_norm": 3.9860846996307373,
       "learning_rate": 0.00018392404526523817,
-      "loss": 2.215,
+      "loss": 2.2563,
       "step": 189
     },
     {
       "epoch": 0.08458542014468559,
-      "grad_norm": 2.9309654235839844,
+      "grad_norm": 3.047419548034668,
       "learning_rate": 0.0001837528040042142,
-      "loss": 1.9777,
+      "loss": 2.0222,
       "step": 190
     },
     {
       "epoch": 0.08503060656649972,
-      "grad_norm": 3.0458357334136963,
+      "grad_norm": 3.0273172855377197,
       "learning_rate": 0.00018358073613682706,
-      "loss": 2.0955,
+      "loss": 1.9349,
       "step": 191
     },
     {
       "epoch": 0.08547579298831386,
-      "grad_norm": 3.166865825653076,
+      "grad_norm": 2.851856231689453,
       "learning_rate": 0.00018340784336131713,
-      "loss": 1.3475,
+      "loss": 1.4659,
       "step": 192
     },
     {
       "epoch": 0.085920979410128,
-      "grad_norm": 3.3954527378082275,
+      "grad_norm": 3.09624981880188,
       "learning_rate": 0.00018323412738406635,
-      "loss": 2.514,
+      "loss": 2.5376,
       "step": 193
     },
     {
       "epoch": 0.08636616583194212,
-      "grad_norm": 3.6221113204956055,
+      "grad_norm": 3.4956867694854736,
       "learning_rate": 0.00018305958991958127,
-      "loss": 2.1971,
+      "loss": 2.1087,
       "step": 194
     },
     {
       "epoch": 0.08681135225375626,
-      "grad_norm": 2.5096681118011475,
+      "grad_norm": 2.766479253768921,
       "learning_rate": 0.0001828842326904762,
-      "loss": 1.988,
+      "loss": 2.0002,
       "step": 195
     },
     {
       "epoch": 0.0872565386755704,
-      "grad_norm": 2.7260031700134277,
+      "grad_norm": 2.8506453037261963,
       "learning_rate": 0.00018270805742745617,
-      "loss": 1.7442,
+      "loss": 1.6929,
       "step": 196
     },
     {
       "epoch": 0.08770172509738453,
-      "grad_norm": 3.155184507369995,
+      "grad_norm": 16.174423217773438,
       "learning_rate": 0.00018253106586929997,
-      "loss": 2.5275,
+      "loss": 2.5689,
       "step": 197
     },
     {
       "epoch": 0.08814691151919866,
-      "grad_norm": 2.810918092727661,
+      "grad_norm": 2.555532932281494,
       "learning_rate": 0.00018235325976284275,
-      "loss": 2.2599,
+      "loss": 2.2229,
       "step": 198
     },
     {
       "epoch": 0.0885920979410128,
-      "grad_norm": 3.6741535663604736,
+      "grad_norm": 3.642751932144165,
       "learning_rate": 0.00018217464086295904,
-      "loss": 2.4509,
+      "loss": 2.5422,
       "step": 199
     },
     {
       "epoch": 0.08903728436282693,
-      "grad_norm": 3.171786069869995,
+      "grad_norm": 3.288435220718384,
       "learning_rate": 0.00018199521093254523,
-      "loss": 1.5725,
+      "loss": 1.628,
       "step": 200
     },
     {
       "epoch": 0.08948247078464107,
-      "grad_norm": 5.2465500831604,
+      "grad_norm": 2.8634743690490723,
       "learning_rate": 0.00018181497174250236,
-      "loss": 2.1309,
+      "loss": 2.0577,
       "step": 201
     },
     {
       "epoch": 0.08992765720645521,
-      "grad_norm": 3.3447489738464355,
+      "grad_norm": 3.5794100761413574,
       "learning_rate": 0.00018163392507171842,
-      "loss": 2.7054,
+      "loss": 2.8942,
       "step": 202
     },
     {
       "epoch": 0.09037284362826933,
-      "grad_norm": 2.2638697624206543,
+      "grad_norm": 2.6064746379852295,
       "learning_rate": 0.00018145207270705096,
-      "loss": 1.6347,
+      "loss": 1.769,
       "step": 203
     },
     {
       "epoch": 0.09081803005008347,
-      "grad_norm": 2.3262786865234375,
+      "grad_norm": 2.4976682662963867,
       "learning_rate": 0.0001812694164433094,
-      "loss": 1.9896,
+      "loss": 2.0236,
       "step": 204
     },
     {
       "epoch": 0.09126321647189761,
-      "grad_norm": 2.7014877796173096,
+      "grad_norm": 3.3556509017944336,
       "learning_rate": 0.00018108595808323736,
-      "loss": 2.3665,
+      "loss": 2.4135,
       "step": 205
     },
     {
       "epoch": 0.09170840289371174,
-      "grad_norm": 2.7195963859558105,
+      "grad_norm": 2.855940580368042,
       "learning_rate": 0.00018090169943749476,
-      "loss": 2.0004,
+      "loss": 2.0044,
       "step": 206
     },
     {
       "epoch": 0.09215358931552588,
-      "grad_norm": 3.5895092487335205,
+      "grad_norm": 5.23508358001709,
       "learning_rate": 0.00018071664232464002,
-      "loss": 2.7216,
+      "loss": 3.1566,
       "step": 207
     },
     {
       "epoch": 0.09259877573734002,
-      "grad_norm": 6.0607008934021,
+      "grad_norm": 2.203725576400757,
       "learning_rate": 0.0001805307885711122,
-      "loss": 1.8817,
+      "loss": 1.8375,
       "step": 208
     },
     {
       "epoch": 0.09304396215915414,
-      "grad_norm": 3.0178444385528564,
+      "grad_norm": 1.9931917190551758,
       "learning_rate": 0.00018034414001121278,
-      "loss": 1.9194,
+      "loss": 1.8984,
       "step": 209
     },
     {
       "epoch": 0.09348914858096828,
-      "grad_norm": 3.298114776611328,
+      "grad_norm": 3.4173572063446045,
       "learning_rate": 0.00018015669848708767,
-      "loss": 3.2093,
+      "loss": 3.1145,
       "step": 210
     },
     {
       "epoch": 0.09393433500278242,
-      "grad_norm": 2.1203761100769043,
+      "grad_norm": 2.123898983001709,
       "learning_rate": 0.00017996846584870908,
-      "loss": 1.5833,
+      "loss": 1.5743,
       "step": 211
     },
     {
       "epoch": 0.09437952142459655,
-      "grad_norm": 2.907470464706421,
+      "grad_norm": 2.5022401809692383,
       "learning_rate": 0.0001797794439538571,
-      "loss": 1.7891,
+      "loss": 1.8138,
       "step": 212
     },
     {
       "epoch": 0.09482470784641069,
-      "grad_norm": 3.151555061340332,
+      "grad_norm": 4.588496685028076,
       "learning_rate": 0.0001795896346681016,
-      "loss": 2.0487,
+      "loss": 2.0505,
       "step": 213
     },
     {
       "epoch": 0.09526989426822483,
-      "grad_norm": 3.080402374267578,
+      "grad_norm": 2.4845874309539795,
       "learning_rate": 0.00017939903986478355,
-      "loss": 1.7114,
+      "loss": 1.7597,
       "step": 214
     },
     {
       "epoch": 0.09571508069003895,
-      "grad_norm": 4.0168561935424805,
+      "grad_norm": 3.9968199729919434,
       "learning_rate": 0.00017920766142499672,
-      "loss": 2.5786,
+      "loss": 2.6499,
       "step": 215
     },
     {
       "epoch": 0.09616026711185309,
-      "grad_norm": 1.990530252456665,
+      "grad_norm": 2.016294479370117,
       "learning_rate": 0.00017901550123756906,
-      "loss": 2.1145,
+      "loss": 2.0903,
       "step": 216
     },
     {
       "epoch": 0.09660545353366723,
-      "grad_norm": 2.382887840270996,
+      "grad_norm": 2.111436605453491,
       "learning_rate": 0.00017882256119904403,
-      "loss": 1.8551,
+      "loss": 1.8273,
       "step": 217
     },
     {
       "epoch": 0.09705063995548135,
-      "grad_norm": 4.134660243988037,
+      "grad_norm": 3.4320459365844727,
       "learning_rate": 0.00017862884321366188,
-      "loss": 2.4086,
+      "loss": 2.3938,
       "step": 218
     },
     {
       "epoch": 0.0974958263772955,
-      "grad_norm": 2.927734375,
+      "grad_norm": 3.473053216934204,
       "learning_rate": 0.000178434349193341,
-      "loss": 1.9386,
+      "loss": 2.0303,
       "step": 219
     },
     {
       "epoch": 0.09794101279910963,
-      "grad_norm": 2.8078081607818604,
+      "grad_norm": 3.058842658996582,
       "learning_rate": 0.0001782390810576588,
-      "loss": 2.0586,
+      "loss": 1.9517,
       "step": 220
     },
     {
       "epoch": 0.09838619922092376,
-      "grad_norm": 4.8399248123168945,
+      "grad_norm": 3.2113101482391357,
       "learning_rate": 0.000178043040733833,
-      "loss": 1.3215,
+      "loss": 1.3175,
       "step": 221
     },
     {
       "epoch": 0.0988313856427379,
-      "grad_norm": 2.1833369731903076,
+      "grad_norm": 2.1675655841827393,
       "learning_rate": 0.00017784623015670238,
-      "loss": 1.8898,
+      "loss": 1.9149,
       "step": 222
     },
     {
       "epoch": 0.09927657206455204,
-      "grad_norm": 3.099783420562744,
+      "grad_norm": 2.9513309001922607,
       "learning_rate": 0.00017764865126870786,
-      "loss": 2.9137,
+      "loss": 2.916,
       "step": 223
     },
     {
       "epoch": 0.09972175848636616,
-      "grad_norm": 2.1298272609710693,
+      "grad_norm": 2.7082040309906006,
       "learning_rate": 0.00017745030601987337,
-      "loss": 1.4856,
+      "loss": 1.5118,
       "step": 224
     },
     {
       "epoch": 0.1001669449081803,
-      "grad_norm": 2.0361313819885254,
+      "grad_norm": 2.039851188659668,
       "learning_rate": 0.00017725119636778644,
-      "loss": 1.8,
+      "loss": 1.8884,
       "step": 225
     },
     {
       "epoch": 0.10061213132999444,
-      "grad_norm": 2.4032180309295654,
+      "grad_norm": 2.2612671852111816,
       "learning_rate": 0.00017705132427757895,
-      "loss": 1.8565,
+      "loss": 1.7989,
       "step": 226
     },
     {
       "epoch": 0.10105731775180857,
-      "grad_norm": 3.0993106365203857,
+      "grad_norm": 3.061992883682251,
       "learning_rate": 0.00017685069172190766,
-      "loss": 2.5554,
+      "loss": 2.6819,
       "step": 227
     },
     {
       "epoch": 0.1015025041736227,
-      "grad_norm": 4.992886066436768,
-      "learning_rate": 0.00017664930068093498,
-      "loss": 3.3477,
+      "grad_norm": NaN,
+      "learning_rate": 0.00017685069172190766,
+      "loss": 3.6896,
       "step": 228
     },
     {
       "epoch": 0.10194769059543683,
-      "grad_norm": 3.065136194229126,
-      "learning_rate": 0.00017644715314230918,
-      "loss": 1.785,
+      "grad_norm": 2.9173364639282227,
+      "learning_rate": 0.00017664930068093498,
+      "loss": 1.7822,
       "step": 229
     },
     {
       "epoch": 0.10239287701725097,
-      "grad_norm": 2.915815830230713,
-      "learning_rate": 0.0001762442511011448,
-      "loss": 1.7885,
+      "grad_norm": 3.1690425872802734,
+      "learning_rate": 0.00017644715314230918,
+      "loss": 1.7367,
       "step": 230
     },
     {
       "epoch": 0.10283806343906511,
-      "grad_norm": 1.5576674938201904,
-      "learning_rate": 0.0001760405965600031,
-      "loss": 1.2228,
+      "grad_norm": 1.5905847549438477,
+      "learning_rate": 0.0001762442511011448,
+      "loss": 1.2002,
       "step": 231
     },
     {
       "epoch": 0.10328324986087924,
-      "grad_norm": 4.22326135635376,
-      "learning_rate": 0.0001758361915288722,
-      "loss": 1.8784,
+      "grad_norm": 2.388805866241455,
+      "learning_rate": 0.0001760405965600031,
+      "loss": 1.911,
       "step": 232
     },
     {
       "epoch": 0.10372843628269338,
-      "grad_norm": 2.074902296066284,
-      "learning_rate": 0.0001756310380251472,
-      "loss": 2.1881,
+      "grad_norm": 2.208155393600464,
+      "learning_rate": 0.0001758361915288722,
+      "loss": 2.1999,
       "step": 233
     },
     {
       "epoch": 0.10417362270450752,
-      "grad_norm": 3.5856900215148926,
-      "learning_rate": 0.00017542513807361037,
-      "loss": 2.452,
+      "grad_norm": 2.7173423767089844,
+      "learning_rate": 0.0001756310380251472,
+      "loss": 2.4822,
       "step": 234
     },
     {
       "epoch": 0.10461880912632164,
-      "grad_norm": 3.115443229675293,
-      "learning_rate": 0.00017521849370641114,
-      "loss": 1.9735,
+      "grad_norm": 2.548349618911743,
+      "learning_rate": 0.00017542513807361037,
+      "loss": 2.043,
       "step": 235
     },
     {
       "epoch": 0.10506399554813578,
-      "grad_norm": 4.1047868728637695,
-      "learning_rate": 0.00017501110696304596,
-      "loss": 4.1891,
+      "grad_norm": 3.7509632110595703,
+      "learning_rate": 0.00017521849370641114,
+      "loss": 4.1206,
       "step": 236
     },
     {
       "epoch": 0.10550918196994992,
-      "grad_norm": 2.4047348499298096,
-      "learning_rate": 0.00017480297989033825,
-      "loss": 1.5507,
+      "grad_norm": 3.5168652534484863,
+      "learning_rate": 0.00017501110696304596,
+      "loss": 1.638,
       "step": 237
     },
     {
       "epoch": 0.10595436839176404,
-      "grad_norm": 2.6233081817626953,
-      "learning_rate": 0.00017459411454241822,
-      "loss": 1.6275,
+      "grad_norm": 2.3908824920654297,
+      "learning_rate": 0.00017480297989033825,
+      "loss": 1.5586,
       "step": 238
     },
     {
       "epoch": 0.10639955481357818,
-      "grad_norm": 2.559443235397339,
-      "learning_rate": 0.00017438451298070252,
-      "loss": 2.1019,
+      "grad_norm": 2.5945377349853516,
+      "learning_rate": 0.00017459411454241822,
+      "loss": 2.0376,
       "step": 239
     },
     {
       "epoch": 0.10684474123539232,
-      "grad_norm": 2.711853265762329,
-      "learning_rate": 0.00017417417727387394,
-      "loss": 1.8006,
+      "grad_norm": 2.566469430923462,
+      "learning_rate": 0.00017438451298070252,
+      "loss": 1.7843,
       "step": 240
     },
     {
       "epoch": 0.10728992765720645,
-      "grad_norm": 2.1308255195617676,
-      "learning_rate": 0.000173963109497861,
-      "loss": 1.0008,
+      "grad_norm": 2.0323879718780518,
+      "learning_rate": 0.00017417417727387394,
+      "loss": 0.9602,
       "step": 241
     },
     {
       "epoch": 0.10773511407902059,
-      "grad_norm": 2.0422074794769287,
-      "learning_rate": 0.0001737513117358174,
-      "loss": 2.1842,
+      "grad_norm": 2.077362537384033,
+      "learning_rate": 0.000173963109497861,
+      "loss": 2.1987,
       "step": 242
     },
     {
       "epoch": 0.10818030050083473,
-      "grad_norm": 1.971974492073059,
-      "learning_rate": 0.0001735387860781016,
-      "loss": 1.1267,
+      "grad_norm": 1.803091049194336,
+      "learning_rate": 0.0001737513117358174,
+      "loss": 1.0866,
       "step": 243
     },
     {
       "epoch": 0.10862548692264885,
-      "grad_norm": 1.90377676486969,
-      "learning_rate": 0.00017332553462225602,
-      "loss": 1.4106,
+      "grad_norm": 1.6945303678512573,
+      "learning_rate": 0.0001735387860781016,
+      "loss": 1.3323,
       "step": 244
     },
     {
       "epoch": 0.10907067334446299,
-      "grad_norm": 6.567409992218018,
-      "learning_rate": 0.00017311155947298643,
-      "loss": 3.1737,
+      "grad_norm": 5.757752418518066,
+      "learning_rate": 0.00017332553462225602,
+      "loss": 3.8289,
       "step": 245
     },
     {
       "epoch": 0.10951585976627713,
-      "grad_norm": 2.6171557903289795,
-      "learning_rate": 0.00017289686274214118,
-      "loss": 2.0543,
+      "grad_norm": 2.5788002014160156,
+      "learning_rate": 0.00017311155947298643,
+      "loss": 1.9719,
       "step": 246
     },
     {
       "epoch": 0.10996104618809126,
-      "grad_norm": 3.360865354537964,
-      "learning_rate": 0.0001726814465486903,
-      "loss": 3.0316,
+      "grad_norm": 3.206275463104248,
+      "learning_rate": 0.00017289686274214118,
+      "loss": 3.0306,
       "step": 247
     },
     {
       "epoch": 0.1104062326099054,
-      "grad_norm": 5.862345218658447,
-      "learning_rate": 0.0001724653130187047,
-      "loss": 2.1349,
+      "grad_norm": 3.6661014556884766,
+      "learning_rate": 0.0001726814465486903,
+      "loss": 2.2205,
       "step": 248
     },
     {
       "epoch": 0.11085141903171954,
-      "grad_norm": 5.230323314666748,
-      "learning_rate": 0.00017224846428533499,
-      "loss": 1.959,
+      "grad_norm": 2.7304251194000244,
+      "learning_rate": 0.0001724653130187047,
+      "loss": 1.9815,
       "step": 249
     },
     {
       "epoch": 0.11129660545353366,
-      "grad_norm": 1.8790351152420044,
-      "learning_rate": 0.0001720309024887907,
-      "loss": 1.559,
+      "grad_norm": 1.8175365924835205,
+      "learning_rate": 0.00017224846428533499,
+      "loss": 1.5728,
       "step": 250
     },
     {
       "epoch": 0.1117417918753478,
-      "grad_norm": 2.6701748371124268,
-      "learning_rate": 0.00017181262977631888,
-      "loss": 2.3727,
+      "grad_norm": 2.614417791366577,
+      "learning_rate": 0.0001720309024887907,
+      "loss": 2.3802,
       "step": 251
     },
     {
       "epoch": 0.11218697829716194,
-      "grad_norm": 2.2404518127441406,
-      "learning_rate": 0.00017159364830218312,
-      "loss": 1.933,
+      "grad_norm": 1.9326399564743042,
+      "learning_rate": 0.00017181262977631888,
+      "loss": 1.9611,
       "step": 252
     },
     {
       "epoch": 0.11263216471897607,
-      "grad_norm": 3.8700356483459473,
-      "learning_rate": 0.00017137396022764214,
-      "loss": 3.0309,
+      "grad_norm": 3.356496572494507,
+      "learning_rate": 0.00017159364830218312,
+      "loss": 2.9872,
       "step": 253
     },
     {
       "epoch": 0.1130773511407902,
-      "grad_norm": 2.1600594520568848,
-      "learning_rate": 0.00017115356772092857,
-      "loss": 1.9069,
+      "grad_norm": 2.0015017986297607,
+      "learning_rate": 0.00017137396022764214,
+      "loss": 1.9562,
       "step": 254
     },
     {
       "epoch": 0.11352253756260434,
-      "grad_norm": 3.375443935394287,
-      "learning_rate": 0.0001709324729572274,
-      "loss": 2.4507,
+      "grad_norm": 3.3400707244873047,
+      "learning_rate": 0.00017115356772092857,
+      "loss": 2.4441,
       "step": 255
     },
     {
       "epoch": 0.11396772398441847,
-      "grad_norm": 2.818181037902832,
-      "learning_rate": 0.00017071067811865476,
-      "loss": 1.9793,
+      "grad_norm": 2.803450345993042,
+      "learning_rate": 0.0001709324729572274,
+      "loss": 1.9406,
       "step": 256
     },
     {
       "epoch": 0.11441291040623261,
-      "grad_norm": 3.2607429027557373,
-      "learning_rate": 0.00017048818539423615,
-      "loss": 1.9051,
+      "grad_norm": 2.935906410217285,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 1.8791,
       "step": 257
     },
     {
       "epoch": 0.11485809682804675,
-      "grad_norm": 2.977018117904663,
-      "learning_rate": 0.00017026499697988493,
-      "loss": 1.0261,
+      "grad_norm": 2.608060598373413,
+      "learning_rate": 0.00017048818539423615,
+      "loss": 0.9475,
       "step": 258
     },
     {
       "epoch": 0.11530328324986087,
-      "grad_norm": 2.8753457069396973,
-      "learning_rate": 0.00017004111507838064,
-      "loss": 1.4664,
+      "grad_norm": 2.941483736038208,
+      "learning_rate": 0.00017026499697988493,
+      "loss": 1.5754,
       "step": 259
     },
     {
       "epoch": 0.11574846967167501,
-      "grad_norm": 3.7540669441223145,
-      "learning_rate": 0.00016981654189934727,
-      "loss": 2.4403,
+      "grad_norm": 3.5497794151306152,
+      "learning_rate": 0.00017004111507838064,
+      "loss": 2.3999,
       "step": 260
     },
     {
       "epoch": 0.11619365609348915,
-      "grad_norm": 2.8768951892852783,
-      "learning_rate": 0.00016959127965923142,
-      "loss": 1.925,
+      "grad_norm": 2.6955342292785645,
+      "learning_rate": 0.00016981654189934727,
+      "loss": 2.0135,
       "step": 261
     },
     {
       "epoch": 0.11663884251530328,
-      "grad_norm": 3.1603147983551025,
-      "learning_rate": 0.0001693653305812805,
-      "loss": 2.3041,
+      "grad_norm": 3.7114124298095703,
+      "learning_rate": 0.00016959127965923142,
+      "loss": 2.4143,
       "step": 262
     },
     {
       "epoch": 0.11708402893711742,
-      "grad_norm": 3.66261625289917,
-      "learning_rate": 0.00016913869689552064,
-      "loss": 2.9761,
+      "grad_norm": 3.3473169803619385,
+      "learning_rate": 0.0001693653305812805,
+      "loss": 2.8757,
       "step": 263
     },
     {
       "epoch": 0.11752921535893156,
-      "grad_norm": 5.0103020668029785,
-      "learning_rate": 0.00016891138083873487,
-      "loss": 2.4806,
+      "grad_norm": 4.7687554359436035,
+      "learning_rate": 0.00016913869689552064,
+      "loss": 2.3467,
       "step": 264
     },
     {
       "epoch": 0.11797440178074568,
-      "grad_norm": 3.2384135723114014,
-      "learning_rate": 0.00016868338465444085,
-      "loss": 2.8883,
+      "grad_norm": 3.388779401779175,
+      "learning_rate": 0.00016891138083873487,
+      "loss": 2.851,
       "step": 265
     },
     {
       "epoch": 0.11841958820255982,
-      "grad_norm": 3.064087390899658,
-      "learning_rate": 0.00016845471059286887,
-      "loss": 2.3382,
+      "grad_norm": 2.956942319869995,
+      "learning_rate": 0.00016868338465444085,
+      "loss": 2.3852,
       "step": 266
     },
     {
       "epoch": 0.11886477462437396,
-      "grad_norm": 3.015291452407837,
-      "learning_rate": 0.00016822536091093965,
-      "loss": 2.1407,
+      "grad_norm": 2.6591806411743164,
+      "learning_rate": 0.00016845471059286887,
+      "loss": 2.0056,
       "step": 267
     },
     {
       "epoch": 0.11930996104618809,
-      "grad_norm": 3.0050747394561768,
-      "learning_rate": 0.00016799533787224192,
-      "loss": 3.0019,
+      "grad_norm": 3.0301601886749268,
+      "learning_rate": 0.00016822536091093965,
+      "loss": 3.0666,
       "step": 268
     },
     {
       "epoch": 0.11975514746800223,
-      "grad_norm": 3.889848470687866,
-      "learning_rate": 0.00016776464374701025,
-      "loss": 2.9921,
+      "grad_norm": 4.273592948913574,
+      "learning_rate": 0.00016799533787224192,
+      "loss": 3.084,
       "step": 269
     },
     {
       "epoch": 0.12020033388981637,
-      "grad_norm": 3.5669052600860596,
-      "learning_rate": 0.00016753328081210245,
-      "loss": 1.9408,
+      "grad_norm": 3.4681904315948486,
+      "learning_rate": 0.00016776464374701025,
+      "loss": 1.8521,
       "step": 270
     },
     {
       "epoch": 0.12064552031163049,
-      "grad_norm": 2.8057496547698975,
-      "learning_rate": 0.00016730125135097735,
-      "loss": 2.0,
+      "grad_norm": 2.7582857608795166,
+      "learning_rate": 0.00016753328081210245,
+      "loss": 1.9692,
       "step": 271
     },
     {
       "epoch": 0.12109070673344463,
-      "grad_norm": 3.1653785705566406,
-      "learning_rate": 0.000167068557653672,
-      "loss": 2.188,
+      "grad_norm": 5.249884605407715,
+      "learning_rate": 0.00016730125135097735,
+      "loss": 2.2755,
       "step": 272
     },
     {
       "epoch": 0.12153589315525877,
-      "grad_norm": 2.8963727951049805,
-      "learning_rate": 0.0001668352020167793,
-      "loss": 1.8975,
+      "grad_norm": 2.75067400932312,
+      "learning_rate": 0.000167068557653672,
+      "loss": 1.8455,
       "step": 273
     },
     {
       "epoch": 0.1219810795770729,
-      "grad_norm": 2.387599468231201,
-      "learning_rate": 0.00016660118674342517,
-      "loss": 1.5727,
+      "grad_norm": 3.4320733547210693,
+      "learning_rate": 0.0001668352020167793,
+      "loss": 1.5247,
       "step": 274
     },
     {
       "epoch": 0.12242626599888703,
-      "grad_norm": 2.26202654838562,
-      "learning_rate": 0.00016636651414324587,
-      "loss": 1.0566,
+      "grad_norm": 4.216282844543457,
+      "learning_rate": 0.00016660118674342517,
+      "loss": 1.0498,
       "step": 275
     },
     {
       "epoch": 0.12287145242070117,
-      "grad_norm": 2.7368710041046143,
-      "learning_rate": 0.00016613118653236518,
-      "loss": 2.203,
+      "grad_norm": 2.9553184509277344,
+      "learning_rate": 0.00016636651414324587,
+      "loss": 2.1679,
       "step": 276
     },
     {
       "epoch": 0.1233166388425153,
-      "grad_norm": 3.0989296436309814,
-      "learning_rate": 0.0001658952062333717,
-      "loss": 2.1975,
+      "grad_norm": 3.903003215789795,
+      "learning_rate": 0.00016613118653236518,
+      "loss": 2.1111,
       "step": 277
     },
     {
       "epoch": 0.12376182526432944,
-      "grad_norm": 2.62036395072937,
-      "learning_rate": 0.00016565857557529566,
-      "loss": 1.5915,
+      "grad_norm": 2.639263868331909,
+      "learning_rate": 0.0001658952062333717,
+      "loss": 1.6075,
       "step": 278
     },
     {
       "epoch": 0.12420701168614358,
-      "grad_norm": 2.7251317501068115,
-      "learning_rate": 0.00016542129689358612,
-      "loss": 1.9822,
+      "grad_norm": 4.093780517578125,
+      "learning_rate": 0.00016565857557529566,
+      "loss": 1.9396,
       "step": 279
     },
     {
       "epoch": 0.1246521981079577,
-      "grad_norm": 2.146713972091675,
-      "learning_rate": 0.0001651833725300879,
-      "loss": 1.5296,
+      "grad_norm": 2.228785514831543,
+      "learning_rate": 0.00016542129689358612,
+      "loss": 1.4554,
       "step": 280
     },
     {
       "epoch": 0.12509738452977184,
-      "grad_norm": 2.940171718597412,
-      "learning_rate": 0.00016494480483301836,
-      "loss": 1.4362,
+      "grad_norm": 3.1095426082611084,
+      "learning_rate": 0.0001651833725300879,
+      "loss": 1.5483,
       "step": 281
     },
     {
       "epoch": 0.12554257095158597,
-      "grad_norm": 5.681768417358398,
-      "learning_rate": 0.00016470559615694446,
-      "loss": 2.3687,
+      "grad_norm": 5.153223514556885,
+      "learning_rate": 0.00016494480483301836,
+      "loss": 2.3421,
       "step": 282
     },
     {
       "epoch": 0.12598775737340012,
-      "grad_norm": 3.23738169670105,
-      "learning_rate": 0.00016446574886275913,
-      "loss": 2.8589,
+      "grad_norm": 4.7425360679626465,
+      "learning_rate": 0.00016470559615694446,
+      "loss": 2.9086,
       "step": 283
     },
     {
       "epoch": 0.12643294379521425,
-      "grad_norm": 4.700006008148193,
-      "learning_rate": 0.00016422526531765846,
-      "loss": 2.2194,
+      "grad_norm": 5.0155720710754395,
+      "learning_rate": 0.00016446574886275913,
+      "loss": 2.037,
       "step": 284
     },
     {
       "epoch": 0.12687813021702837,
-      "grad_norm": 4.28733491897583,
-      "learning_rate": 0.00016398414789511786,
-      "loss": 3.1353,
+      "grad_norm": 4.832873344421387,
+      "learning_rate": 0.00016422526531765846,
+      "loss": 3.072,
       "step": 285
     },
     {
       "epoch": 0.12732331663884253,
-      "grad_norm": 2.7851359844207764,
-      "learning_rate": 0.000163742398974869,
-      "loss": 1.5339,
+      "grad_norm": 2.0803449153900146,
+      "learning_rate": 0.00016398414789511786,
+      "loss": 1.438,
       "step": 286
     },
     {
       "epoch": 0.12776850306065665,
-      "grad_norm": 4.9445624351501465,
-      "learning_rate": 0.00016350002094287609,
-      "loss": 1.8456,
+      "grad_norm": 5.869335174560547,
+      "learning_rate": 0.000163742398974869,
+      "loss": 1.9414,
       "step": 287
     },
     {
       "epoch": 0.12821368948247078,
-      "grad_norm": 3.781156539916992,
-      "learning_rate": 0.00016325701619131246,
-      "loss": 1.9967,
+      "grad_norm": 4.237220287322998,
+      "learning_rate": 0.00016350002094287609,
+      "loss": 2.1311,
       "step": 288
     },
     {
       "epoch": 0.12865887590428493,
-      "grad_norm": 4.377411842346191,
-      "learning_rate": 0.00016301338711853693,
-      "loss": 3.0335,
+      "grad_norm": 4.472570896148682,
+      "learning_rate": 0.00016325701619131246,
+      "loss": 2.9077,
       "step": 289
     },
     {
       "epoch": 0.12910406232609906,
-      "grad_norm": 3.9102425575256348,
-      "learning_rate": 0.00016276913612907007,
-      "loss": 2.9909,
+      "grad_norm": 5.2327728271484375,
+      "learning_rate": 0.00016301338711853693,
+      "loss": 2.9067,
       "step": 290
     },
     {
       "epoch": 0.12954924874791318,
-      "grad_norm": 4.036067485809326,
-      "learning_rate": 0.00016252426563357055,
-      "loss": 2.1329,
+      "grad_norm": 4.579474925994873,
+      "learning_rate": 0.00016276913612907007,
+      "loss": 2.2166,
       "step": 291
     },
     {
       "epoch": 0.12999443516972733,
-      "grad_norm": 2.960005283355713,
-      "learning_rate": 0.00016227877804881127,
-      "loss": 1.7441,
+      "grad_norm": 3.3287553787231445,
+      "learning_rate": 0.00016252426563357055,
+      "loss": 1.8055,
       "step": 292
     },
     {
       "epoch": 0.13043962159154146,
-      "grad_norm": 1.907597303390503,
-      "learning_rate": 0.00016203267579765563,
-      "loss": 1.3116,
+      "grad_norm": 1.91664457321167,
+      "learning_rate": 0.00016227877804881127,
+      "loss": 1.2989,
       "step": 293
     },
     {
       "epoch": 0.13088480801335559,
-      "grad_norm": 3.022691249847412,
-      "learning_rate": 0.00016178596130903344,
-      "loss": 2.5592,
+      "grad_norm": 2.971626043319702,
+      "learning_rate": 0.00016203267579765563,
+      "loss": 2.4503,
       "step": 294
     },
     {
       "epoch": 0.13132999443516974,
-      "grad_norm": 4.502274990081787,
-      "learning_rate": 0.00016153863701791717,
-      "loss": 2.3546,
+      "grad_norm": 5.920880317687988,
+      "learning_rate": 0.00016178596130903344,
+      "loss": 2.3126,
       "step": 295
     },
     {
       "epoch": 0.13177518085698386,
-      "grad_norm": 3.090162754058838,
-      "learning_rate": 0.00016129070536529766,
-      "loss": 2.0902,
+      "grad_norm": 4.491692066192627,
+      "learning_rate": 0.00016153863701791717,
+      "loss": 2.0189,
       "step": 296
     },
     {
       "epoch": 0.132220367278798,
-      "grad_norm": 2.6545863151550293,
-      "learning_rate": 0.00016104216879816026,
-      "loss": 1.932,
+      "grad_norm": 3.163790702819824,
+      "learning_rate": 0.00016129070536529766,
+      "loss": 2.0132,
       "step": 297
     },
     {
       "epoch": 0.13266555370061214,
-      "grad_norm": 2.626149892807007,
-      "learning_rate": 0.00016079302976946055,
-      "loss": 1.9929,
+      "grad_norm": 3.0796022415161133,
+      "learning_rate": 0.00016104216879816026,
+      "loss": 1.9845,
       "step": 298
     },
     {
       "epoch": 0.13311074012242627,
-      "grad_norm": 2.5994675159454346,
-      "learning_rate": 0.00016054329073810015,
-      "loss": 2.0321,
+      "grad_norm": 5.9430646896362305,
+      "learning_rate": 0.00016079302976946055,
+      "loss": 1.9985,
       "step": 299
     },
     {
       "epoch": 0.1335559265442404,
-      "grad_norm": 2.5016086101531982,
-      "learning_rate": 0.00016029295416890248,
-      "loss": 1.6849,
+      "grad_norm": 3.705900192260742,
+      "learning_rate": 0.00016054329073810015,
+      "loss": 1.7144,
       "step": 300
     },
     {
       "epoch": 0.13400111296605455,
-      "grad_norm": 2.2394087314605713,
-      "learning_rate": 0.00016004202253258842,
-      "loss": 1.6157,
+      "grad_norm": 2.9216806888580322,
+      "learning_rate": 0.00016029295416890248,
+      "loss": 1.5975,
       "step": 301
     },
     {
       "epoch": 0.13444629938786867,
-      "grad_norm": 2.9624788761138916,
-      "learning_rate": 0.0001597904983057519,
-      "loss": 1.0671,
+      "grad_norm": 2.5859384536743164,
+      "learning_rate": 0.00016004202253258842,
+      "loss": 0.9565,
       "step": 302
     },
     {
       "epoch": 0.1348914858096828,
-      "grad_norm": 3.160810708999634,
-      "learning_rate": 0.00015953838397083552,
-      "loss": 2.1351,
+      "grad_norm": 3.337036609649658,
+      "learning_rate": 0.0001597904983057519,
+      "loss": 2.1839,
       "step": 303
     },
     {
       "epoch": 0.13533667223149695,
-      "grad_norm": 2.8223936557769775,
-      "learning_rate": 0.00015928568201610595,
-      "loss": 1.9108,
+      "grad_norm": 3.213444709777832,
+      "learning_rate": 0.00015953838397083552,
+      "loss": 2.0066,
       "step": 304
     },
     {
       "epoch": 0.13578185865331108,
-      "grad_norm": 2.559141159057617,
-      "learning_rate": 0.00015903239493562948,
-      "loss": 2.1271,
+      "grad_norm": 3.1022112369537354,
+      "learning_rate": 0.00015928568201610595,
+      "loss": 2.1541,
       "step": 305
     },
     {
       "epoch": 0.1362270450751252,
-      "grad_norm": 3.396461009979248,
-      "learning_rate": 0.00015877852522924732,
-      "loss": 1.8675,
+      "grad_norm": 3.345914840698242,
+      "learning_rate": 0.00015903239493562948,
+      "loss": 1.9975,
       "step": 306
     },
     {
       "epoch": 0.13667223149693936,
-      "grad_norm": 3.4334423542022705,
-      "learning_rate": 0.00015852407540255104,
-      "loss": 1.7112,
+      "grad_norm": 3.352506399154663,
+      "learning_rate": 0.00015877852522924732,
+      "loss": 1.6579,
       "step": 307
     },
     {
       "epoch": 0.13711741791875348,
-      "grad_norm": 5.243045330047607,
-      "learning_rate": 0.00015826904796685762,
-      "loss": 1.5088,
+      "grad_norm": 3.894771099090576,
+      "learning_rate": 0.00015852407540255104,
+      "loss": 1.441,
       "step": 308
     },
     {
       "epoch": 0.1375626043405676,
-      "grad_norm": 2.0283901691436768,
-      "learning_rate": 0.00015801344543918495,
-      "loss": 1.5021,
+      "grad_norm": 2.2265045642852783,
+      "learning_rate": 0.00015826904796685762,
+      "loss": 1.5189,
       "step": 309
     },
     {
       "epoch": 0.13800779076238176,
-      "grad_norm": 2.131378173828125,
-      "learning_rate": 0.00015775727034222675,
-      "loss": 2.0285,
+      "grad_norm": 3.3169307708740234,
+      "learning_rate": 0.00015801344543918495,
+      "loss": 1.9754,
       "step": 310
     },
     {
       "epoch": 0.13845297718419589,
-      "grad_norm": 2.8632185459136963,
-      "learning_rate": 0.00015750052520432787,
-      "loss": 2.1884,
+      "grad_norm": 3.269655466079712,
+      "learning_rate": 0.00015775727034222675,
+      "loss": 2.1918,
       "step": 311
     },
     {
       "epoch": 0.13889816360601,
-      "grad_norm": 3.2297849655151367,
-      "learning_rate": 0.0001572432125594591,
-      "loss": 2.0599,
+      "grad_norm": 3.389246702194214,
+      "learning_rate": 0.00015750052520432787,
+      "loss": 2.1055,
       "step": 312
     },
     {
       "epoch": 0.13934335002782416,
-      "grad_norm": 4.282790660858154,
-      "learning_rate": 0.00015698533494719238,
-      "loss": 2.6433,
+      "grad_norm": 3.9781131744384766,
+      "learning_rate": 0.0001572432125594591,
+      "loss": 2.5854,
       "step": 313
     },
     {
       "epoch": 0.1397885364496383,
-      "grad_norm": 3.296445369720459,
-      "learning_rate": 0.00015672689491267567,
-      "loss": 2.5076,
+      "grad_norm": 3.8214235305786133,
+      "learning_rate": 0.00015698533494719238,
+      "loss": 2.4634,
       "step": 314
     },
     {
       "epoch": 0.14023372287145242,
-      "grad_norm": 2.099283456802368,
-      "learning_rate": 0.00015646789500660773,
-      "loss": 1.3551,
+      "grad_norm": 3.8951199054718018,
+      "learning_rate": 0.00015672689491267567,
+      "loss": 1.3678,
       "step": 315
     },
     {
       "epoch": 0.14067890929326657,
-      "grad_norm": 2.6336355209350586,
-      "learning_rate": 0.00015620833778521307,
-      "loss": 1.5616,
+      "grad_norm": 3.5023081302642822,
+      "learning_rate": 0.00015646789500660773,
+      "loss": 1.4961,
       "step": 316
     },
     {
       "epoch": 0.1411240957150807,
-      "grad_norm": 1.6293888092041016,
-      "learning_rate": 0.0001559482258102167,
-      "loss": 0.9969,
+      "grad_norm": 1.9665567874908447,
+      "learning_rate": 0.00015620833778521307,
+      "loss": 1.0397,
       "step": 317
     },
     {
       "epoch": 0.14156928213689482,
-      "grad_norm": 2.045325517654419,
-      "learning_rate": 0.00015568756164881882,
-      "loss": 1.6796,
+      "grad_norm": 2.015803575515747,
+      "learning_rate": 0.0001559482258102167,
+      "loss": 1.6607,
       "step": 318
     },
     {
       "epoch": 0.14201446855870897,
-      "grad_norm": 3.8815982341766357,
-      "learning_rate": 0.00015542634787366942,
-      "loss": 1.6373,
+      "grad_norm": 3.766350746154785,
+      "learning_rate": 0.00015568756164881882,
+      "loss": 1.6466,
       "step": 319
     },
     {
       "epoch": 0.1424596549805231,
-      "grad_norm": 2.5085458755493164,
-      "learning_rate": 0.00015516458706284303,
-      "loss": 2.3436,
+      "grad_norm": 2.1854238510131836,
+      "learning_rate": 0.00015542634787366942,
+      "loss": 2.359,
       "step": 320
     },
     {
       "epoch": 0.14290484140233722,
-      "grad_norm": 2.1215457916259766,
-      "learning_rate": 0.0001549022817998132,
-      "loss": 1.2887,
+      "grad_norm": 2.5207645893096924,
+      "learning_rate": 0.00015516458706284303,
+      "loss": 1.3911,
       "step": 321
     },
     {
       "epoch": 0.14335002782415138,
-      "grad_norm": 1.9667388200759888,
-      "learning_rate": 0.00015463943467342693,
-      "loss": 1.7435,
+      "grad_norm": 2.2049384117126465,
+      "learning_rate": 0.0001549022817998132,
+      "loss": 1.7231,
       "step": 322
     },
     {
       "epoch": 0.1437952142459655,
-      "grad_norm": 4.962423324584961,
-      "learning_rate": 0.00015437604827787927,
-      "loss": 3.1438,
+      "grad_norm": 6.676117420196533,
+      "learning_rate": 0.00015463943467342693,
+      "loss": 2.9389,
       "step": 323
     },
     {
       "epoch": 0.14424040066777963,
-      "grad_norm": 3.1294186115264893,
-      "learning_rate": 0.00015411212521268758,
-      "loss": 1.922,
+      "grad_norm": 2.7219367027282715,
+      "learning_rate": 0.00015437604827787927,
+      "loss": 1.9098,
       "step": 324
     },
     {
       "epoch": 0.14468558708959378,
-      "grad_norm": 2.8280906677246094,
-      "learning_rate": 0.00015384766808266602,
-      "loss": 1.8233,
+      "grad_norm": 2.9971964359283447,
+      "learning_rate": 0.00015411212521268758,
+      "loss": 1.8154,
       "step": 325
     },
     {
       "epoch": 0.1451307735114079,
-      "grad_norm": 3.0242607593536377,
-      "learning_rate": 0.00015358267949789966,
-      "loss": 2.0155,
+      "grad_norm": 3.812910795211792,
+      "learning_rate": 0.00015384766808266602,
+      "loss": 2.1349,
       "step": 326
     },
     {
       "epoch": 0.14557595993322203,
-      "grad_norm": 1.9289848804473877,
-      "learning_rate": 0.00015331716207371888,
-      "loss": 1.6602,
+      "grad_norm": 1.8337472677230835,
+      "learning_rate": 0.00015358267949789966,
+      "loss": 1.7524,
       "step": 327
     },
     {
       "epoch": 0.14602114635503616,
-      "grad_norm": 3.2559800148010254,
-      "learning_rate": 0.0001530511184306734,
-      "loss": 2.0712,
+      "grad_norm": 2.9395253658294678,
+      "learning_rate": 0.00015331716207371888,
+      "loss": 1.9965,
       "step": 328
     },
     {
       "epoch": 0.1464663327768503,
-      "grad_norm": 3.1982977390289307,
-      "learning_rate": 0.00015278455119450664,
-      "loss": 1.9866,
+      "grad_norm": 2.7258734703063965,
+      "learning_rate": 0.0001530511184306734,
+      "loss": 1.824,
       "step": 329
     },
     {
       "epoch": 0.14691151919866444,
-      "grad_norm": 4.816375255584717,
-      "learning_rate": 0.0001525174629961296,
-      "loss": 3.3914,
+      "grad_norm": 5.46239709854126,
+      "learning_rate": 0.00015278455119450664,
+      "loss": 3.3704,
       "step": 330
     },
     {
       "epoch": 0.14735670562047856,
-      "grad_norm": 2.63909912109375,
-      "learning_rate": 0.0001522498564715949,
-      "loss": 2.4218,
+      "grad_norm": 2.961097002029419,
+      "learning_rate": 0.0001525174629961296,
+      "loss": 2.4186,
       "step": 331
     },
     {
       "epoch": 0.14780189204229272,
-      "grad_norm": 2.412900686264038,
-      "learning_rate": 0.00015198173426207094,
-      "loss": 1.9517,
+      "grad_norm": 3.7786366939544678,
+      "learning_rate": 0.0001522498564715949,
+      "loss": 1.9667,
       "step": 332
     },
     {
       "epoch": 0.14824707846410684,
-      "grad_norm": 2.017374038696289,
-      "learning_rate": 0.00015171309901381572,
-      "loss": 1.4106,
+      "grad_norm": 1.9760524034500122,
+      "learning_rate": 0.00015198173426207094,
+      "loss": 1.4929,
       "step": 333
     },
     {
       "epoch": 0.14869226488592097,
-      "grad_norm": 3.504751205444336,
-      "learning_rate": 0.00015144395337815064,
-      "loss": 2.372,
+      "grad_norm": 3.354715347290039,
+      "learning_rate": 0.00015171309901381572,
+      "loss": 2.3631,
       "step": 334
     },
     {
       "epoch": 0.14913745130773512,
-      "grad_norm": 3.916456460952759,
-      "learning_rate": 0.00015117430001143452,
-      "loss": 1.9819,
+      "grad_norm": 3.134253740310669,
+      "learning_rate": 0.00015144395337815064,
+      "loss": 1.9308,
       "step": 335
     },
     {
       "epoch": 0.14958263772954924,
-      "grad_norm": 1.9360079765319824,
-      "learning_rate": 0.00015090414157503714,
-      "loss": 1.5803,
+      "grad_norm": 1.8403511047363281,
+      "learning_rate": 0.00015117430001143452,
+      "loss": 1.5778,
       "step": 336
     },
     {
       "epoch": 0.15002782415136337,
-      "grad_norm": 4.142049312591553,
-      "learning_rate": 0.00015063348073531324,
-      "loss": 3.1576,
+      "grad_norm": 3.9706308841705322,
+      "learning_rate": 0.00015090414157503714,
+      "loss": 3.1083,
       "step": 337
     },
     {
       "epoch": 0.15047301057317752,
-      "grad_norm": 5.079156875610352,
-      "learning_rate": 0.0001503623201635761,
-      "loss": 2.724,
+      "grad_norm": 5.338191509246826,
+      "learning_rate": 0.00015063348073531324,
+      "loss": 2.761,
       "step": 338
     },
     {
       "epoch": 0.15091819699499165,
-      "grad_norm": 2.2369110584259033,
-      "learning_rate": 0.000150090662536071,
-      "loss": 1.2975,
+      "grad_norm": 2.094376802444458,
+      "learning_rate": 0.0001503623201635761,
+      "loss": 1.3146,
       "step": 339
     },
     {
       "epoch": 0.15136338341680577,
-      "grad_norm": 3.156975746154785,
-      "learning_rate": 0.0001498185105339491,
-      "loss": 1.9116,
+      "grad_norm": 3.034886360168457,
+      "learning_rate": 0.000150090662536071,
+      "loss": 1.9302,
       "step": 340
     },
     {
       "epoch": 0.15180856983861993,
-      "grad_norm": 2.597970724105835,
-      "learning_rate": 0.00014954586684324078,
-      "loss": 1.4781,
+      "grad_norm": 2.9534401893615723,
+      "learning_rate": 0.0001498185105339491,
+      "loss": 1.6089,
       "step": 341
     },
     {
       "epoch": 0.15225375626043405,
-      "grad_norm": 2.6372573375701904,
-      "learning_rate": 0.00014927273415482915,
-      "loss": 2.4297,
+      "grad_norm": 2.7792069911956787,
+      "learning_rate": 0.00014954586684324078,
+      "loss": 2.5257,
       "step": 342
     },
     {
       "epoch": 0.15269894268224818,
-      "grad_norm": 5.823768615722656,
-      "learning_rate": 0.00014899911516442365,
-      "loss": 3.3218,
+      "grad_norm": 6.320833206176758,
+      "learning_rate": 0.00014927273415482915,
+      "loss": 3.6341,
       "step": 343
     },
     {
       "epoch": 0.15314412910406233,
-      "grad_norm": 1.7671197652816772,
-      "learning_rate": 0.00014872501257253323,
-      "loss": 1.5805,
+      "grad_norm": 1.7262303829193115,
+      "learning_rate": 0.00014899911516442365,
+      "loss": 1.5187,
       "step": 344
     },
     {
       "epoch": 0.15358931552587646,
-      "grad_norm": 4.895401954650879,
-      "learning_rate": 0.0001484504290844398,
-      "loss": 2.6022,
+      "grad_norm": 4.237273216247559,
+      "learning_rate": 0.00014872501257253323,
+      "loss": 2.6972,
       "step": 345
     },
     {
       "epoch": 0.15403450194769058,
-      "grad_norm": 2.013131618499756,
-      "learning_rate": 0.00014817536741017152,
-      "loss": 1.9509,
+      "grad_norm": 2.0071420669555664,
+      "learning_rate": 0.0001484504290844398,
+      "loss": 1.9343,
       "step": 346
     },
     {
       "epoch": 0.15447968836950474,
-      "grad_norm": 4.518467426300049,
-      "learning_rate": 0.00014789983026447612,
-      "loss": 3.4165,
+      "grad_norm": 4.065686225891113,
+      "learning_rate": 0.00014817536741017152,
+      "loss": 3.2849,
       "step": 347
     },
     {
       "epoch": 0.15492487479131886,
-      "grad_norm": 3.0389232635498047,
-      "learning_rate": 0.0001476238203667939,
-      "loss": 2.2441,
+      "grad_norm": 3.3194217681884766,
+      "learning_rate": 0.00014789983026447612,
+      "loss": 2.3392,
       "step": 348
     },
     {
       "epoch": 0.155370061213133,
-      "grad_norm": 2.6705610752105713,
-      "learning_rate": 0.0001473473404412312,
-      "loss": 1.6084,
+      "grad_norm": 2.6346261501312256,
+      "learning_rate": 0.0001476238203667939,
+      "loss": 1.5951,
       "step": 349
     },
     {
       "epoch": 0.15581524763494714,
-      "grad_norm": 2.1380245685577393,
-      "learning_rate": 0.0001470703932165333,
-      "loss": 2.0694,
+      "grad_norm": 2.0923845767974854,
+      "learning_rate": 0.0001473473404412312,
+      "loss": 2.1046,
       "step": 350
     },
     {
       "epoch": 0.15626043405676127,
-      "grad_norm": 2.587960720062256,
-      "learning_rate": 0.00014679298142605734,
-      "loss": 1.7345,
+      "grad_norm": 2.5452301502227783,
+      "learning_rate": 0.0001470703932165333,
+      "loss": 1.6933,
       "step": 351
     },
     {
       "epoch": 0.1567056204785754,
-      "grad_norm": 2.9578754901885986,
-      "learning_rate": 0.00014651510780774583,
-      "loss": 2.6031,
+      "grad_norm": 3.4971985816955566,
+      "learning_rate": 0.00014679298142605734,
+      "loss": 2.6632,
       "step": 352
     },
     {
       "epoch": 0.15715080690038954,
-      "grad_norm": 2.348881483078003,
-      "learning_rate": 0.00014623677510409918,
-      "loss": 1.448,
+      "grad_norm": 2.550041913986206,
+      "learning_rate": 0.00014651510780774583,
+      "loss": 1.4933,
       "step": 353
     },
     {
       "epoch": 0.15759599332220367,
-      "grad_norm": 1.7157777547836304,
-      "learning_rate": 0.00014595798606214882,
-      "loss": 1.8623,
+      "grad_norm": 1.7662994861602783,
+      "learning_rate": 0.00014623677510409918,
+      "loss": 1.876,
       "step": 354
     },
     {
       "epoch": 0.1580411797440178,
-      "grad_norm": 2.54980206489563,
-      "learning_rate": 0.00014567874343342997,
-      "loss": 2.263,
+      "grad_norm": 2.723386764526367,
+      "learning_rate": 0.00014595798606214882,
+      "loss": 2.3278,
       "step": 355
     },
     {
       "epoch": 0.15848636616583195,
-      "grad_norm": 3.329082727432251,
-      "learning_rate": 0.00014539904997395468,
-      "loss": 2.8837,
+      "grad_norm": 4.134690761566162,
+      "learning_rate": 0.00014567874343342997,
+      "loss": 3.0174,
       "step": 356
     },
     {
       "epoch": 0.15893155258764607,
-      "grad_norm": 2.646977186203003,
-      "learning_rate": 0.00014511890844418453,
-      "loss": 2.1387,
+      "grad_norm": 2.940185785293579,
+      "learning_rate": 0.00014539904997395468,
+      "loss": 2.019,
       "step": 357
     },
     {
       "epoch": 0.1593767390094602,
-      "grad_norm": 2.9344465732574463,
-      "learning_rate": 0.00014483832160900326,
-      "loss": 1.9743,
+      "grad_norm": 2.911198854446411,
+      "learning_rate": 0.00014511890844418453,
+      "loss": 2.0522,
       "step": 358
     },
     {
       "epoch": 0.15982192543127435,
-      "grad_norm": 2.056246757507324,
-      "learning_rate": 0.00014455729223768966,
-      "loss": 1.8003,
+      "grad_norm": 1.9790756702423096,
+      "learning_rate": 0.00014483832160900326,
+      "loss": 1.8049,
       "step": 359
     },
     {
       "epoch": 0.16026711185308848,
-      "grad_norm": 2.6241796016693115,
-      "learning_rate": 0.0001442758231038902,
-      "loss": 1.888,
+      "grad_norm": 3.8680615425109863,
+      "learning_rate": 0.00014455729223768966,
+      "loss": 1.9089,
       "step": 360
     },
     {
       "epoch": 0.1607122982749026,
-      "grad_norm": 2.479048490524292,
-      "learning_rate": 0.00014399391698559152,
-      "loss": 2.3365,
+      "grad_norm": 2.533252477645874,
+      "learning_rate": 0.0001442758231038902,
+      "loss": 2.2968,
       "step": 361
     },
     {
       "epoch": 0.16115748469671676,
-      "grad_norm": 3.3290910720825195,
-      "learning_rate": 0.0001437115766650933,
-      "loss": 2.5672,
+      "grad_norm": 3.094660758972168,
+      "learning_rate": 0.00014399391698559152,
+      "loss": 2.6112,
       "step": 362
     },
     {
       "epoch": 0.16160267111853088,
-      "grad_norm": 2.197995901107788,
-      "learning_rate": 0.00014342880492898048,
-      "loss": 1.4891,
+      "grad_norm": 2.3190393447875977,
+      "learning_rate": 0.0001437115766650933,
+      "loss": 1.5047,
       "step": 363
     },
     {
       "epoch": 0.162047857540345,
-      "grad_norm": 2.3591930866241455,
-      "learning_rate": 0.0001431456045680959,
-      "loss": 1.7955,
+      "grad_norm": 2.6006054878234863,
+      "learning_rate": 0.00014342880492898048,
+      "loss": 1.8432,
       "step": 364
     },
     {
       "epoch": 0.16249304396215916,
-      "grad_norm": 1.573541283607483,
-      "learning_rate": 0.00014286197837751286,
-      "loss": 1.1049,
+      "grad_norm": 1.5267677307128906,
+      "learning_rate": 0.0001431456045680959,
+      "loss": 1.1295,
       "step": 365
     },
     {
       "epoch": 0.1629382303839733,
-      "grad_norm": 2.382366895675659,
-      "learning_rate": 0.00014257792915650728,
-      "loss": 1.717,
+      "grad_norm": 2.2319841384887695,
+      "learning_rate": 0.00014286197837751286,
+      "loss": 1.7568,
       "step": 366
     },
     {
       "epoch": 0.1633834168057874,
-      "grad_norm": 3.2191505432128906,
-      "learning_rate": 0.00014229345970853032,
-      "loss": 1.6779,
+      "grad_norm": 3.55242919921875,
+      "learning_rate": 0.00014257792915650728,
+      "loss": 1.6084,
       "step": 367
     },
     {
       "epoch": 0.16382860322760157,
-      "grad_norm": 2.418705940246582,
-      "learning_rate": 0.00014200857284118066,
-      "loss": 2.6337,
+      "grad_norm": 2.2210030555725098,
+      "learning_rate": 0.00014229345970853032,
+      "loss": 2.615,
       "step": 368
     },
     {
       "epoch": 0.1642737896494157,
-      "grad_norm": 5.708262920379639,
-      "learning_rate": 0.00014172327136617656,
-      "loss": 1.7388,
+      "grad_norm": 2.4402711391448975,
+      "learning_rate": 0.00014200857284118066,
+      "loss": 1.7663,
       "step": 369
     },
     {
       "epoch": 0.16471897607122982,
-      "grad_norm": 2.3304173946380615,
-      "learning_rate": 0.00014143755809932845,
-      "loss": 1.4906,
+      "grad_norm": 2.149660110473633,
+      "learning_rate": 0.00014172327136617656,
+      "loss": 1.4058,
       "step": 370
     },
     {
       "epoch": 0.16516416249304397,
-      "grad_norm": 2.6991124153137207,
-      "learning_rate": 0.00014115143586051088,
-      "loss": 2.3806,
+      "grad_norm": 2.59019136428833,
+      "learning_rate": 0.00014143755809932845,
+      "loss": 2.3203,
       "step": 371
     },
     {
       "epoch": 0.1656093489148581,
-      "grad_norm": 2.831113815307617,
-      "learning_rate": 0.00014086490747363493,
-      "loss": 2.0115,
+      "grad_norm": 2.770490884780884,
+      "learning_rate": 0.00014115143586051088,
+      "loss": 2.0733,
       "step": 372
     },
     {
       "epoch": 0.16605453533667222,
-      "grad_norm": 1.9368008375167847,
-      "learning_rate": 0.00014057797576662,
-      "loss": 1.6859,
+      "grad_norm": 1.6338651180267334,
+      "learning_rate": 0.00014086490747363493,
+      "loss": 1.7117,
       "step": 373
     },
     {
       "epoch": 0.16649972175848637,
-      "grad_norm": 2.908867120742798,
-      "learning_rate": 0.00014029064357136628,
-      "loss": 2.2395,
+      "grad_norm": 2.856350898742676,
+      "learning_rate": 0.00014057797576662,
+      "loss": 2.2055,
       "step": 374
     },
     {
       "epoch": 0.1669449081803005,
-      "grad_norm": 2.90267014503479,
-      "learning_rate": 0.00014000291372372647,
-      "loss": 1.6788,
+      "grad_norm": 2.6866633892059326,
+      "learning_rate": 0.00014029064357136628,
+      "loss": 1.5605,
       "step": 375
     },
     {
       "epoch": 0.16739009460211463,
-      "grad_norm": 4.940428256988525,
-      "learning_rate": 0.00013971478906347806,
-      "loss": 3.0354,
+      "grad_norm": 3.7765796184539795,
+      "learning_rate": 0.00014000291372372647,
+      "loss": 2.924,
       "step": 376
     },
     {
       "epoch": 0.16783528102392878,
-      "grad_norm": 3.7296106815338135,
-      "learning_rate": 0.00013942627243429512,
-      "loss": 3.1005,
+      "grad_norm": 3.816124439239502,
+      "learning_rate": 0.00013971478906347806,
+      "loss": 3.0261,
       "step": 377
     },
     {
       "epoch": 0.1682804674457429,
-      "grad_norm": 1.515480399131775,
-      "learning_rate": 0.00013913736668372026,
-      "loss": 0.9675,
+      "grad_norm": 1.5007274150848389,
+      "learning_rate": 0.00013942627243429512,
+      "loss": 1.0276,
       "step": 378
     },
     {
       "epoch": 0.16872565386755703,
-      "grad_norm": 2.466693878173828,
-      "learning_rate": 0.00013884807466313663,
-      "loss": 2.4831,
+      "grad_norm": 2.2790749073028564,
+      "learning_rate": 0.00013913736668372026,
+      "loss": 2.5021,
       "step": 379
     },
     {
       "epoch": 0.16917084028937118,
-      "grad_norm": 2.1269826889038086,
-      "learning_rate": 0.00013855839922773968,
-      "loss": 1.7787,
+      "grad_norm": 2.2356507778167725,
+      "learning_rate": 0.00013884807466313663,
+      "loss": 1.8988,
       "step": 380
     },
     {
       "epoch": 0.1696160267111853,
-      "grad_norm": 2.2861855030059814,
-      "learning_rate": 0.000138268343236509,
-      "loss": 1.8964,
+      "grad_norm": 2.7027039527893066,
+      "learning_rate": 0.00013855839922773968,
+      "loss": 1.9612,
       "step": 381
     },
     {
       "epoch": 0.17006121313299943,
-      "grad_norm": 7.59018611907959,
-      "learning_rate": 0.00013797790955218014,
-      "loss": 3.1163,
+      "grad_norm": 7.827208042144775,
+      "learning_rate": 0.000138268343236509,
+      "loss": 3.1185,
       "step": 382
     },
     {
       "epoch": 0.1705063995548136,
-      "grad_norm": 3.9868404865264893,
-      "learning_rate": 0.00013768710104121627,
-      "loss": 2.4488,
+      "grad_norm": 4.195032119750977,
+      "learning_rate": 0.00013797790955218014,
+      "loss": 2.3052,
       "step": 383
     },
     {
       "epoch": 0.1709515859766277,
-      "grad_norm": 1.6617087125778198,
-      "learning_rate": 0.00013739592057378003,
-      "loss": 1.21,
+      "grad_norm": 1.5661183595657349,
+      "learning_rate": 0.00013768710104121627,
+      "loss": 1.2437,
       "step": 384
     },
     {
       "epoch": 0.17139677239844184,
-      "grad_norm": 1.8703160285949707,
-      "learning_rate": 0.0001371043710237051,
-      "loss": 1.5203,
+      "grad_norm": 1.8515113592147827,
+      "learning_rate": 0.00013739592057378003,
+      "loss": 1.4918,
       "step": 385
     },
     {
       "epoch": 0.171841958820256,
-      "grad_norm": 3.014777898788452,
-      "learning_rate": 0.00013681245526846783,
-      "loss": 2.0583,
+      "grad_norm": 2.7802538871765137,
+      "learning_rate": 0.0001371043710237051,
+      "loss": 1.9919,
       "step": 386
     },
     {
       "epoch": 0.17228714524207012,
-      "grad_norm": 1.9843742847442627,
-      "learning_rate": 0.0001365201761891588,
-      "loss": 1.2145,
+      "grad_norm": 2.7814526557922363,
+      "learning_rate": 0.00013681245526846783,
+      "loss": 1.1622,
       "step": 387
     },
     {
       "epoch": 0.17273233166388424,
-      "grad_norm": 2.274531364440918,
-      "learning_rate": 0.00013622753667045457,
-      "loss": 1.6243,
+      "grad_norm": 2.9777095317840576,
+      "learning_rate": 0.0001365201761891588,
+      "loss": 1.6679,
       "step": 388
     },
     {
       "epoch": 0.1731775180856984,
-      "grad_norm": 2.8857228755950928,
-      "learning_rate": 0.00013593453960058908,
-      "loss": 2.1109,
+      "grad_norm": 3.84218692779541,
+      "learning_rate": 0.00013622753667045457,
+      "loss": 2.1753,
       "step": 389
     },
     {
       "epoch": 0.17362270450751252,
-      "grad_norm": 2.410280227661133,
-      "learning_rate": 0.00013564118787132506,
-      "loss": 1.6705,
+      "grad_norm": 2.0334346294403076,
+      "learning_rate": 0.00013593453960058908,
+      "loss": 1.5844,
       "step": 390
     },
     {
       "epoch": 0.17406789092932665,
-      "grad_norm": 2.83205509185791,
-      "learning_rate": 0.00013534748437792573,
-      "loss": 1.6343,
+      "grad_norm": 2.971250057220459,
+      "learning_rate": 0.00013564118787132506,
+      "loss": 1.6105,
       "step": 391
     },
     {
       "epoch": 0.1745130773511408,
-      "grad_norm": 3.435878276824951,
-      "learning_rate": 0.0001350534320191259,
-      "loss": 2.7,
+      "grad_norm": 3.5902481079101562,
+      "learning_rate": 0.00013534748437792573,
+      "loss": 2.641,
       "step": 392
     },
     {
       "epoch": 0.17495826377295493,
-      "grad_norm": 2.2407634258270264,
-      "learning_rate": 0.0001347590336971037,
-      "loss": 1.5352,
+      "grad_norm": 2.412871837615967,
+      "learning_rate": 0.0001350534320191259,
+      "loss": 1.54,
       "step": 393
     },
     {
       "epoch": 0.17540345019476905,
-      "grad_norm": 2.265737295150757,
-      "learning_rate": 0.0001344642923174517,
-      "loss": 1.8782,
+      "grad_norm": 2.215141534805298,
+      "learning_rate": 0.0001347590336971037,
+      "loss": 1.8944,
       "step": 394
     },
     {
       "epoch": 0.1758486366165832,
-      "grad_norm": 3.784402847290039,
-      "learning_rate": 0.00013416921078914835,
-      "loss": 2.0235,
+      "grad_norm": 3.511486768722534,
+      "learning_rate": 0.0001344642923174517,
+      "loss": 1.9907,
       "step": 395
     },
     {
       "epoch": 0.17629382303839733,
-      "grad_norm": 3.2451465129852295,
-      "learning_rate": 0.00013387379202452917,
-      "loss": 1.9769,
+      "grad_norm": 3.9856975078582764,
+      "learning_rate": 0.00013416921078914835,
+      "loss": 2.0401,
       "step": 396
     },
     {
       "epoch": 0.17673900946021145,
-      "grad_norm": 2.4640157222747803,
-      "learning_rate": 0.00013357803893925807,
-      "loss": 1.5395,
+      "grad_norm": 4.489439010620117,
+      "learning_rate": 0.00013387379202452917,
+      "loss": 1.5596,
       "step": 397
     },
     {
       "epoch": 0.1771841958820256,
-      "grad_norm": 2.5422592163085938,
-      "learning_rate": 0.00013328195445229868,
-      "loss": 1.4376,
+      "grad_norm": 2.6045870780944824,
+      "learning_rate": 0.00013357803893925807,
+      "loss": 1.3855,
       "step": 398
     },
     {
       "epoch": 0.17762938230383973,
-      "grad_norm": 2.2282865047454834,
-      "learning_rate": 0.00013298554148588528,
-      "loss": 1.3988,
+      "grad_norm": 2.2161664962768555,
+      "learning_rate": 0.00013328195445229868,
+      "loss": 1.391,
       "step": 399
     },
     {
       "epoch": 0.17807456872565386,
-      "grad_norm": 3.0729358196258545,
-      "learning_rate": 0.00013268880296549425,
-      "loss": 1.2906,
+      "grad_norm": 2.083101987838745,
+      "learning_rate": 0.00013298554148588528,
+      "loss": 1.3238,
       "step": 400
     },
     {
       "epoch": 0.178519755147468,
-      "grad_norm": 2.901768684387207,
-      "learning_rate": 0.00013239174181981495,
-      "loss": 1.7331,
+      "grad_norm": 2.943004846572876,
+      "learning_rate": 0.00013268880296549425,
+      "loss": 1.7093,
       "step": 401
     },
     {
       "epoch": 0.17896494156928214,
-      "grad_norm": 3.0331332683563232,
-      "learning_rate": 0.00013209436098072095,
-      "loss": 1.2559,
+      "grad_norm": 2.9320528507232666,
+      "learning_rate": 0.00013239174181981495,
+      "loss": 1.2862,
       "step": 402
     },
     {
       "epoch": 0.17941012799109626,
-      "grad_norm": 2.4391226768493652,
-      "learning_rate": 0.00013179666338324108,
-      "loss": 2.0366,
+      "grad_norm": 2.143991470336914,
+      "learning_rate": 0.00013209436098072095,
+      "loss": 1.9759,
       "step": 403
     },
     {
       "epoch": 0.17985531441291042,
-      "grad_norm": 3.8267745971679688,
-      "learning_rate": 0.0001314986519655305,
-      "loss": 2.2878,
+      "grad_norm": 3.475919485092163,
+      "learning_rate": 0.00013179666338324108,
+      "loss": 2.2233,
       "step": 404
     },
     {
       "epoch": 0.18030050083472454,
-      "grad_norm": 2.9066944122314453,
-      "learning_rate": 0.0001312003296688415,
-      "loss": 2.3687,
+      "grad_norm": 2.583845853805542,
+      "learning_rate": 0.0001314986519655305,
+      "loss": 2.3193,
       "step": 405
     },
     {
       "epoch": 0.18074568725653867,
-      "grad_norm": 2.571004867553711,
-      "learning_rate": 0.00013090169943749476,
-      "loss": 1.4725,
+      "grad_norm": 2.5285942554473877,
+      "learning_rate": 0.0001312003296688415,
+      "loss": 1.5733,
       "step": 406
     },
     {
       "epoch": 0.18119087367835282,
-      "grad_norm": 2.621811866760254,
-      "learning_rate": 0.0001306027642188501,
-      "loss": 1.4061,
+      "grad_norm": 2.6708908081054688,
+      "learning_rate": 0.00013090169943749476,
+      "loss": 1.3843,
       "step": 407
     },
     {
       "epoch": 0.18163606010016695,
-      "grad_norm": 3.159045457839966,
-      "learning_rate": 0.00013030352696327742,
-      "loss": 2.3518,
+      "grad_norm": 3.256469249725342,
+      "learning_rate": 0.0001306027642188501,
+      "loss": 2.4698,
       "step": 408
     },
     {
       "epoch": 0.18208124652198107,
-      "grad_norm": 2.415534496307373,
-      "learning_rate": 0.00013000399062412763,
-      "loss": 1.9811,
+      "grad_norm": 2.4278006553649902,
+      "learning_rate": 0.00013030352696327742,
+      "loss": 1.987,
       "step": 409
     },
     {
       "epoch": 0.18252643294379522,
-      "grad_norm": 3.9818074703216553,
-      "learning_rate": 0.0001297041581577035,
-      "loss": 2.3213,
+      "grad_norm": 3.7837703227996826,
+      "learning_rate": 0.00013000399062412763,
+      "loss": 2.4242,
       "step": 410
     },
     {
       "epoch": 0.18297161936560935,
-      "grad_norm": 2.050266981124878,
-      "learning_rate": 0.0001294040325232304,
-      "loss": 1.3752,
+      "grad_norm": 2.245476245880127,
+      "learning_rate": 0.0001297041581577035,
+      "loss": 1.385,
       "step": 411
     },
     {
       "epoch": 0.18341680578742348,
-      "grad_norm": 2.8136508464813232,
-      "learning_rate": 0.00012910361668282719,
-      "loss": 1.8344,
+      "grad_norm": 2.3928139209747314,
+      "learning_rate": 0.0001294040325232304,
+      "loss": 1.8256,
       "step": 412
     },
     {
       "epoch": 0.18386199220923763,
-      "grad_norm": 1.4741458892822266,
-      "learning_rate": 0.00012880291360147693,
-      "loss": 1.1769,
+      "grad_norm": 1.4292818307876587,
+      "learning_rate": 0.00012910361668282719,
+      "loss": 1.1733,
       "step": 413
     },
     {
       "epoch": 0.18430717863105175,
-      "grad_norm": 2.022780418395996,
-      "learning_rate": 0.0001285019262469976,
-      "loss": 1.2992,
+      "grad_norm": 2.100421905517578,
+      "learning_rate": 0.00012880291360147693,
+      "loss": 1.2818,
       "step": 414
     },
     {
       "epoch": 0.18475236505286588,
-      "grad_norm": 3.656526565551758,
-      "learning_rate": 0.00012820065759001293,
-      "loss": 1.7559,
+      "grad_norm": 2.6342415809631348,
+      "learning_rate": 0.0001285019262469976,
+      "loss": 1.7543,
       "step": 415
     },
     {
       "epoch": 0.18519755147468003,
-      "grad_norm": 1.9175677299499512,
-      "learning_rate": 0.00012789911060392294,
-      "loss": 1.2447,
+      "grad_norm": 2.015045642852783,
+      "learning_rate": 0.00012820065759001293,
+      "loss": 1.2298,
       "step": 416
     },
     {
       "epoch": 0.18564273789649416,
-      "grad_norm": 3.0030620098114014,
-      "learning_rate": 0.0001275972882648746,
-      "loss": 1.7294,
+      "grad_norm": 3.4002110958099365,
+      "learning_rate": 0.00012789911060392294,
+      "loss": 1.7277,
       "step": 417
     },
     {
       "epoch": 0.18608792431830828,
-      "grad_norm": 2.6003453731536865,
-      "learning_rate": 0.00012729519355173254,
-      "loss": 1.9416,
+      "grad_norm": 2.60213565826416,
+      "learning_rate": 0.0001275972882648746,
+      "loss": 1.916,
       "step": 418
     },
     {
       "epoch": 0.18653311074012244,
-      "grad_norm": 3.4832310676574707,
-      "learning_rate": 0.00012699282944604967,
-      "loss": 2.2738,
+      "grad_norm": 3.615614891052246,
+      "learning_rate": 0.00012729519355173254,
+      "loss": 2.3712,
       "step": 419
     },
     {
       "epoch": 0.18697829716193656,
-      "grad_norm": 3.8813576698303223,
-      "learning_rate": 0.00012669019893203759,
-      "loss": 2.3793,
+      "grad_norm": 3.994493246078491,
+      "learning_rate": 0.00012699282944604967,
+      "loss": 2.2538,
       "step": 420
     },
     {
       "epoch": 0.1874234835837507,
-      "grad_norm": 2.7621383666992188,
-      "learning_rate": 0.0001263873049965373,
-      "loss": 1.966,
+      "grad_norm": 2.8048856258392334,
+      "learning_rate": 0.00012669019893203759,
+      "loss": 2.0137,
       "step": 421
     },
     {
       "epoch": 0.18786867000556484,
-      "grad_norm": 2.440943717956543,
-      "learning_rate": 0.00012608415062898972,
-      "loss": 1.8048,
+      "grad_norm": 2.4496052265167236,
+      "learning_rate": 0.0001263873049965373,
+      "loss": 1.7419,
       "step": 422
     },
     {
       "epoch": 0.18831385642737897,
-      "grad_norm": 1.448773980140686,
-      "learning_rate": 0.000125780738821406,
-      "loss": 1.2906,
+      "grad_norm": 1.7106574773788452,
+      "learning_rate": 0.00012608415062898972,
+      "loss": 1.32,
       "step": 423
     },
     {
       "epoch": 0.1887590428491931,
-      "grad_norm": 3.421792507171631,
-      "learning_rate": 0.00012547707256833823,
-      "loss": 2.0552,
+      "grad_norm": 2.7997119426727295,
+      "learning_rate": 0.000125780738821406,
+      "loss": 2.0131,
       "step": 424
     },
     {
       "epoch": 0.18920422927100725,
-      "grad_norm": 3.0485334396362305,
-      "learning_rate": 0.00012517315486684972,
-      "loss": 1.9516,
+      "grad_norm": 3.381044864654541,
+      "learning_rate": 0.00012547707256833823,
+      "loss": 1.9459,
       "step": 425
     },
     {
       "epoch": 0.18964941569282137,
-      "grad_norm": 2.456167221069336,
-      "learning_rate": 0.0001248689887164855,
-      "loss": 1.821,
+      "grad_norm": 2.4843053817749023,
+      "learning_rate": 0.00012517315486684972,
+      "loss": 1.8847,
       "step": 426
     },
     {
       "epoch": 0.1900946021146355,
-      "grad_norm": 2.6346614360809326,
-      "learning_rate": 0.00012456457711924266,
-      "loss": 2.2551,
+      "grad_norm": 2.7945895195007324,
+      "learning_rate": 0.0001248689887164855,
+      "loss": 2.2822,
       "step": 427
     },
     {
       "epoch": 0.19053978853644965,
-      "grad_norm": 3.651096820831299,
-      "learning_rate": 0.00012425992307954075,
-      "loss": 2.8791,
+      "grad_norm": 3.6827800273895264,
+      "learning_rate": 0.00012456457711924266,
+      "loss": 3.0184,
       "step": 428
     },
     {
       "epoch": 0.19098497495826378,
-      "grad_norm": 1.784805417060852,
-      "learning_rate": 0.0001239550296041922,
-      "loss": 1.3371,
+      "grad_norm": 1.9260412454605103,
+      "learning_rate": 0.00012425992307954075,
+      "loss": 1.2819,
       "step": 429
     },
     {
       "epoch": 0.1914301613800779,
-      "grad_norm": 3.8563966751098633,
-      "learning_rate": 0.00012364989970237248,
-      "loss": 1.7833,
+      "grad_norm": 3.411238431930542,
+      "learning_rate": 0.0001239550296041922,
+      "loss": 1.7672,
       "step": 430
     },
     {
       "epoch": 0.19187534780189205,
-      "grad_norm": 4.481318950653076,
-      "learning_rate": 0.00012334453638559057,
-      "loss": 2.8327,
+      "grad_norm": 5.014751434326172,
+      "learning_rate": 0.00012364989970237248,
+      "loss": 2.9558,
       "step": 431
     },
     {
       "epoch": 0.19232053422370618,
-      "grad_norm": 2.5223546028137207,
-      "learning_rate": 0.00012303894266765908,
-      "loss": 2.3534,
+      "grad_norm": 2.5855774879455566,
+      "learning_rate": 0.00012334453638559057,
+      "loss": 2.3954,
       "step": 432
     },
     {
       "epoch": 0.1927657206455203,
-      "grad_norm": 4.295266151428223,
-      "learning_rate": 0.00012273312156466464,
-      "loss": 2.6395,
+      "grad_norm": 3.727348566055298,
+      "learning_rate": 0.00012303894266765908,
+      "loss": 2.7071,
       "step": 433
     },
     {
       "epoch": 0.19321090706733446,
-      "grad_norm": 1.6645418405532837,
-      "learning_rate": 0.00012242707609493814,
-      "loss": 1.0704,
+      "grad_norm": 1.514435887336731,
+      "learning_rate": 0.00012273312156466464,
+      "loss": 1.0601,
       "step": 434
     },
     {
       "epoch": 0.19365609348914858,
-      "grad_norm": 3.143169641494751,
-      "learning_rate": 0.00012212080927902474,
-      "loss": 1.7698,
+      "grad_norm": 3.1652112007141113,
+      "learning_rate": 0.00012242707609493814,
+      "loss": 1.8037,
       "step": 435
     },
     {
       "epoch": 0.1941012799109627,
-      "grad_norm": 2.1398355960845947,
-      "learning_rate": 0.00012181432413965428,
-      "loss": 1.534,
+      "grad_norm": 2.0706193447113037,
+      "learning_rate": 0.00012212080927902474,
+      "loss": 1.4704,
       "step": 436
     },
     {
       "epoch": 0.19454646633277686,
-      "grad_norm": 3.0436127185821533,
-      "learning_rate": 0.00012150762370171136,
-      "loss": 2.5599,
+      "grad_norm": 2.991508722305298,
+      "learning_rate": 0.00012181432413965428,
+      "loss": 2.6486,
       "step": 437
     },
     {
       "epoch": 0.194991652754591,
-      "grad_norm": 2.8342740535736084,
-      "learning_rate": 0.00012120071099220549,
-      "loss": 2.1801,
+      "grad_norm": 2.859058141708374,
+      "learning_rate": 0.00012150762370171136,
+      "loss": 2.1157,
       "step": 438
     },
     {
       "epoch": 0.1954368391764051,
-      "grad_norm": 2.6189768314361572,
-      "learning_rate": 0.00012089358904024117,
-      "loss": 1.9147,
+      "grad_norm": 2.6109142303466797,
+      "learning_rate": 0.00012120071099220549,
+      "loss": 1.9313,
       "step": 439
     },
     {
       "epoch": 0.19588202559821927,
-      "grad_norm": 3.389474868774414,
-      "learning_rate": 0.00012058626087698814,
-      "loss": 1.3261,
+      "grad_norm": 3.324293375015259,
+      "learning_rate": 0.00012089358904024117,
+      "loss": 1.4513,
       "step": 440
     },
     {
       "epoch": 0.1963272120200334,
-      "grad_norm": 3.9404706954956055,
-      "learning_rate": 0.00012027872953565125,
-      "loss": 2.1684,
+      "grad_norm": 4.9398722648620605,
+      "learning_rate": 0.00012058626087698814,
+      "loss": 2.2132,
       "step": 441
     },
     {
       "epoch": 0.19677239844184752,
-      "grad_norm": 2.1697638034820557,
-      "learning_rate": 0.00011997099805144069,
-      "loss": 1.4573,
+      "grad_norm": 2.100003480911255,
+      "learning_rate": 0.00012027872953565125,
+      "loss": 1.4436,
       "step": 442
     },
     {
       "epoch": 0.19721758486366167,
-      "grad_norm": 3.369276762008667,
-      "learning_rate": 0.000119663069461542,
-      "loss": 2.4629,
+      "grad_norm": 3.5456857681274414,
+      "learning_rate": 0.00011997099805144069,
+      "loss": 2.3848,
       "step": 443
     },
     {
       "epoch": 0.1976627712854758,
-      "grad_norm": 3.570361375808716,
-      "learning_rate": 0.00011935494680508606,
-      "loss": 2.0022,
+      "grad_norm": 3.4976227283477783,
+      "learning_rate": 0.000119663069461542,
+      "loss": 2.0566,
       "step": 444
     },
     {
       "epoch": 0.19810795770728992,
-      "grad_norm": 2.2363481521606445,
-      "learning_rate": 0.00011904663312311901,
-      "loss": 1.5469,
+      "grad_norm": 2.21435284614563,
+      "learning_rate": 0.00011935494680508606,
+      "loss": 1.6255,
       "step": 445
     },
     {
       "epoch": 0.19855314412910408,
-      "grad_norm": 3.8988430500030518,
-      "learning_rate": 0.00011873813145857249,
-      "loss": 1.5989,
+      "grad_norm": 3.514770984649658,
+      "learning_rate": 0.00011904663312311901,
+      "loss": 1.6344,
       "step": 446
     },
     {
       "epoch": 0.1989983305509182,
-      "grad_norm": 4.510793685913086,
-      "learning_rate": 0.00011842944485623335,
-      "loss": 1.8663,
+      "grad_norm": 4.509243488311768,
+      "learning_rate": 0.00011873813145857249,
+      "loss": 1.8616,
       "step": 447
     },
     {
       "epoch": 0.19944351697273233,
-      "grad_norm": 2.888258457183838,
-      "learning_rate": 0.00011812057636271374,
-      "loss": 1.8436,
+      "grad_norm": 2.5948128700256348,
+      "learning_rate": 0.00011842944485623335,
+      "loss": 1.8176,
       "step": 448
     },
     {
       "epoch": 0.19988870339454648,
-      "grad_norm": 3.260255813598633,
-      "learning_rate": 0.000117811529026421,
-      "loss": 1.8474,
+      "grad_norm": 3.1495344638824463,
+      "learning_rate": 0.00011812057636271374,
+      "loss": 1.964,
       "step": 449
     },
     {
       "epoch": 0.2003338898163606,
-      "grad_norm": 3.7456436157226562,
-      "learning_rate": 0.00011750230589752762,
-      "loss": 2.7048,
+      "grad_norm": 3.056257486343384,
+      "learning_rate": 0.000117811529026421,
+      "loss": 2.5386,
       "step": 450
     },
     {
       "epoch": 0.20077907623817473,
-      "grad_norm": 3.437774181365967,
-      "learning_rate": 0.00011719291002794096,
-      "loss": 1.8771,
+      "grad_norm": 3.0034544467926025,
+      "learning_rate": 0.00011750230589752762,
+      "loss": 1.7149,
       "step": 451
     },
     {
       "epoch": 0.20122426265998888,
-      "grad_norm": 2.5170540809631348,
-      "learning_rate": 0.00011688334447127338,
-      "loss": 1.2413,
+      "grad_norm": 3.060884714126587,
+      "learning_rate": 0.00011719291002794096,
+      "loss": 1.2396,
       "step": 452
     },
     {
       "epoch": 0.201669449081803,
-      "grad_norm": 2.3565282821655273,
-      "learning_rate": 0.00011657361228281199,
-      "loss": 1.5339,
+      "grad_norm": 2.2568047046661377,
+      "learning_rate": 0.00011688334447127338,
+      "loss": 1.534,
       "step": 453
     },
     {
       "epoch": 0.20211463550361713,
-      "grad_norm": 1.9782150983810425,
-      "learning_rate": 0.00011626371651948838,
-      "loss": 1.6498,
+      "grad_norm": 2.048640727996826,
+      "learning_rate": 0.00011657361228281199,
+      "loss": 1.709,
       "step": 454
     },
     {
       "epoch": 0.2025598219254313,
-      "grad_norm": 3.0511531829833984,
-      "learning_rate": 0.00011595366023984864,
-      "loss": 2.4361,
+      "grad_norm": 2.9789860248565674,
+      "learning_rate": 0.00011626371651948838,
+      "loss": 2.4558,
       "step": 455
     },
     {
       "epoch": 0.2030050083472454,
-      "grad_norm": 2.8406717777252197,
-      "learning_rate": 0.0001156434465040231,
-      "loss": 2.3422,
+      "grad_norm": 2.8929457664489746,
+      "learning_rate": 0.00011595366023984864,
+      "loss": 2.3674,
       "step": 456
     },
     {
       "epoch": 0.20345019476905954,
-      "grad_norm": 1.845874309539795,
-      "learning_rate": 0.00011533307837369607,
-      "loss": 1.1649,
+      "grad_norm": 2.2031073570251465,
+      "learning_rate": 0.0001156434465040231,
+      "loss": 1.1918,
       "step": 457
     },
     {
       "epoch": 0.20389538119087366,
-      "grad_norm": 2.0563290119171143,
-      "learning_rate": 0.00011502255891207572,
-      "loss": 2.0534,
+      "grad_norm": 2.2434680461883545,
+      "learning_rate": 0.00011533307837369607,
+      "loss": 2.1253,
       "step": 458
     },
     {
       "epoch": 0.20434056761268782,
-      "grad_norm": 2.988293170928955,
-      "learning_rate": 0.00011471189118386375,
-      "loss": 2.1835,
+      "grad_norm": 2.9706263542175293,
+      "learning_rate": 0.00011502255891207572,
+      "loss": 2.2269,
       "step": 459
     },
     {
       "epoch": 0.20478575403450194,
-      "grad_norm": 2.55254864692688,
-      "learning_rate": 0.00011440107825522521,
-      "loss": 1.6571,
+      "grad_norm": 2.349477767944336,
+      "learning_rate": 0.00011471189118386375,
+      "loss": 1.6526,
       "step": 460
     },
     {
       "epoch": 0.20523094045631607,
-      "grad_norm": 3.426617383956909,
-      "learning_rate": 0.00011409012319375827,
-      "loss": 2.8862,
+      "grad_norm": 4.1461181640625,
+      "learning_rate": 0.00011440107825522521,
+      "loss": 2.8745,
       "step": 461
     },
     {
       "epoch": 0.20567612687813022,
-      "grad_norm": 2.7885515689849854,
-      "learning_rate": 0.0001137790290684638,
-      "loss": 1.8314,
+      "grad_norm": 2.6058058738708496,
+      "learning_rate": 0.00011409012319375827,
+      "loss": 1.8548,
       "step": 462
     },
     {
       "epoch": 0.20612131329994435,
-      "grad_norm": 3.6172313690185547,
-      "learning_rate": 0.00011346779894971527,
-      "loss": 2.8842,
+      "grad_norm": 3.3619179725646973,
+      "learning_rate": 0.0001137790290684638,
+      "loss": 2.7486,
       "step": 463
     },
     {
       "epoch": 0.20656649972175847,
-      "grad_norm": 2.387355089187622,
-      "learning_rate": 0.00011315643590922827,
-      "loss": 1.9039,
+      "grad_norm": 2.5370032787323,
+      "learning_rate": 0.00011346779894971527,
+      "loss": 1.9156,
       "step": 464
     },
     {
       "epoch": 0.20701168614357263,
-      "grad_norm": 3.7346434593200684,
-      "learning_rate": 0.0001128449430200303,
-      "loss": 2.7045,
+      "grad_norm": 3.8200554847717285,
+      "learning_rate": 0.00011315643590922827,
+      "loss": 2.6811,
       "step": 465
     },
     {
       "epoch": 0.20745687256538675,
-      "grad_norm": 2.339855909347534,
-      "learning_rate": 0.00011253332335643043,
-      "loss": 2.2928,
+      "grad_norm": 2.52695369720459,
+      "learning_rate": 0.0001128449430200303,
+      "loss": 2.3269,
       "step": 466
     },
     {
       "epoch": 0.20790205898720088,
-      "grad_norm": 2.6608927249908447,
-      "learning_rate": 0.00011222157999398895,
-      "loss": 1.6835,
+      "grad_norm": 3.2745585441589355,
+      "learning_rate": 0.00011253332335643043,
+      "loss": 1.6872,
       "step": 467
     },
     {
       "epoch": 0.20834724540901503,
-      "grad_norm": 1.9378608465194702,
-      "learning_rate": 0.00011190971600948699,
-      "loss": 1.2669,
+      "grad_norm": 2.0885512828826904,
+      "learning_rate": 0.00011222157999398895,
+      "loss": 1.2987,
       "step": 468
     },
     {
       "epoch": 0.20879243183082916,
-      "grad_norm": 1.9632965326309204,
-      "learning_rate": 0.00011159773448089614,
-      "loss": 1.8431,
+      "grad_norm": 1.8260122537612915,
+      "learning_rate": 0.00011190971600948699,
+      "loss": 1.8033,
       "step": 469
     },
     {
       "epoch": 0.20923761825264328,
-      "grad_norm": 2.015932083129883,
-      "learning_rate": 0.00011128563848734816,
-      "loss": 2.0745,
+      "grad_norm": 2.1838929653167725,
+      "learning_rate": 0.00011159773448089614,
+      "loss": 2.0944,
       "step": 470
     },
     {
       "epoch": 0.20968280467445743,
-      "grad_norm": 3.485055446624756,
-      "learning_rate": 0.00011097343110910452,
-      "loss": 2.3325,
+      "grad_norm": 3.3871777057647705,
+      "learning_rate": 0.00011128563848734816,
+      "loss": 2.2967,
       "step": 471
     },
     {
       "epoch": 0.21012799109627156,
-      "grad_norm": 1.4672657251358032,
-      "learning_rate": 0.000110661115427526,
-      "loss": 1.0659,
+      "grad_norm": 1.639346718788147,
+      "learning_rate": 0.00011097343110910452,
+      "loss": 1.0565,
       "step": 472
     },
     {
       "epoch": 0.21057317751808569,
-      "grad_norm": 1.980690360069275,
-      "learning_rate": 0.00011034869452504226,
-      "loss": 1.4239,
+      "grad_norm": 2.0020196437835693,
+      "learning_rate": 0.000110661115427526,
+      "loss": 1.4049,
       "step": 473
     },
     {
       "epoch": 0.21101836393989984,
-      "grad_norm": 3.352896213531494,
-      "learning_rate": 0.00011003617148512149,
-      "loss": 2.4245,
+      "grad_norm": 3.8000659942626953,
+      "learning_rate": 0.00011034869452504226,
+      "loss": 2.369,
       "step": 474
     },
     {
       "epoch": 0.21146355036171396,
-      "grad_norm": 2.4032557010650635,
-      "learning_rate": 0.00010972354939223996,
-      "loss": 2.1085,
+      "grad_norm": 2.5505828857421875,
+      "learning_rate": 0.00011003617148512149,
+      "loss": 2.0883,
       "step": 475
     },
     {
       "epoch": 0.2119087367835281,
-      "grad_norm": 2.4151711463928223,
-      "learning_rate": 0.00010941083133185146,
-      "loss": 1.4129,
+      "grad_norm": 2.576690196990967,
+      "learning_rate": 0.00010972354939223996,
+      "loss": 1.3852,
       "step": 476
     },
     {
       "epoch": 0.21235392320534224,
-      "grad_norm": 2.9115583896636963,
-      "learning_rate": 0.00010909802039035701,
-      "loss": 1.8839,
+      "grad_norm": 3.0029373168945312,
+      "learning_rate": 0.00010941083133185146,
+      "loss": 1.9155,
       "step": 477
     },
     {
       "epoch": 0.21279910962715637,
-      "grad_norm": 2.2955610752105713,
-      "learning_rate": 0.00010878511965507434,
-      "loss": 1.5365,
+      "grad_norm": 2.3589234352111816,
+      "learning_rate": 0.00010909802039035701,
+      "loss": 1.57,
       "step": 478
     },
     {
       "epoch": 0.2132442960489705,
-      "grad_norm": 2.006725311279297,
-      "learning_rate": 0.00010847213221420736,
-      "loss": 1.1548,
+      "grad_norm": 1.7912325859069824,
+      "learning_rate": 0.00010878511965507434,
+      "loss": 1.1136,
       "step": 479
     },
     {
       "epoch": 0.21368948247078465,
-      "grad_norm": 2.6657001972198486,
-      "learning_rate": 0.00010815906115681578,
-      "loss": 2.425,
+      "grad_norm": 3.358750820159912,
+      "learning_rate": 0.00010847213221420736,
+      "loss": 2.4167,
       "step": 480
     },
     {
       "epoch": 0.21413466889259877,
-      "grad_norm": 2.453512668609619,
-      "learning_rate": 0.0001078459095727845,
-      "loss": 1.5641,
+      "grad_norm": 2.7186977863311768,
+      "learning_rate": 0.00010815906115681578,
+      "loss": 1.6694,
       "step": 481
     },
     {
       "epoch": 0.2145798553144129,
-      "grad_norm": 2.2438836097717285,
-      "learning_rate": 0.00010753268055279329,
-      "loss": 1.736,
+      "grad_norm": 2.201023817062378,
+      "learning_rate": 0.0001078459095727845,
+      "loss": 1.7678,
       "step": 482
     },
     {
       "epoch": 0.21502504173622705,
-      "grad_norm": 2.286283254623413,
-      "learning_rate": 0.0001072193771882861,
-      "loss": 1.2366,
+      "grad_norm": 2.136697769165039,
+      "learning_rate": 0.00010753268055279329,
+      "loss": 1.2536,
       "step": 483
     },
     {
       "epoch": 0.21547022815804118,
-      "grad_norm": 2.357325553894043,
-      "learning_rate": 0.00010690600257144061,
-      "loss": 1.9888,
+      "grad_norm": 2.5579357147216797,
+      "learning_rate": 0.0001072193771882861,
+      "loss": 1.9963,
       "step": 484
     },
     {
       "epoch": 0.2159154145798553,
-      "grad_norm": 4.799047470092773,
-      "learning_rate": 0.0001065925597951378,
-      "loss": 2.1147,
+      "grad_norm": 4.265329837799072,
+      "learning_rate": 0.00010690600257144061,
+      "loss": 2.2109,
       "step": 485
     },
     {
       "epoch": 0.21636060100166946,
-      "grad_norm": 3.4145114421844482,
-      "learning_rate": 0.00010627905195293135,
-      "loss": 2.2844,
+      "grad_norm": 3.6019909381866455,
+      "learning_rate": 0.0001065925597951378,
+      "loss": 2.2145,
       "step": 486
     },
     {
       "epoch": 0.21680578742348358,
-      "grad_norm": 2.5542376041412354,
-      "learning_rate": 0.00010596548213901708,
-      "loss": 1.5962,
+      "grad_norm": 17.6555118560791,
+      "learning_rate": 0.00010627905195293135,
+      "loss": 1.8007,
       "step": 487
     },
     {
       "epoch": 0.2172509738452977,
-      "grad_norm": 3.4638633728027344,
-      "learning_rate": 0.00010565185344820247,
-      "loss": 2.0081,
+      "grad_norm": 3.238166570663452,
+      "learning_rate": 0.00010596548213901708,
+      "loss": 1.9748,
       "step": 488
     },
     {
       "epoch": 0.21769616026711186,
-      "grad_norm": 3.015550374984741,
-      "learning_rate": 0.00010533816897587606,
-      "loss": 1.978,
+      "grad_norm": 2.759385824203491,
+      "learning_rate": 0.00010565185344820247,
+      "loss": 1.9488,
       "step": 489
     },
     {
       "epoch": 0.21814134668892599,
-      "grad_norm": 3.2160592079162598,
-      "learning_rate": 0.00010502443181797697,
-      "loss": 1.6013,
+      "grad_norm": 3.2196052074432373,
+      "learning_rate": 0.00010533816897587606,
+      "loss": 1.697,
       "step": 490
     },
     {
       "epoch": 0.2185865331107401,
-      "grad_norm": 3.16475510597229,
-      "learning_rate": 0.00010471064507096426,
-      "loss": 2.1781,
+      "grad_norm": 3.861132860183716,
+      "learning_rate": 0.00010502443181797697,
+      "loss": 2.3017,
       "step": 491
     },
     {
       "epoch": 0.21903171953255426,
-      "grad_norm": 3.783090591430664,
-      "learning_rate": 0.0001043968118317865,
-      "loss": 2.6753,
+      "grad_norm": 3.6513118743896484,
+      "learning_rate": 0.00010471064507096426,
+      "loss": 2.5697,
       "step": 492
     },
     {
       "epoch": 0.2194769059543684,
-      "grad_norm": 3.5325706005096436,
-      "learning_rate": 0.00010408293519785101,
-      "loss": 2.4122,
+      "grad_norm": 3.5032055377960205,
+      "learning_rate": 0.0001043968118317865,
+      "loss": 2.4045,
       "step": 493
     },
     {
       "epoch": 0.21992209237618252,
-      "grad_norm": 2.7166507244110107,
-      "learning_rate": 0.00010376901826699348,
-      "loss": 1.2042,
+      "grad_norm": 1.864540696144104,
+      "learning_rate": 0.00010408293519785101,
+      "loss": 1.1518,
       "step": 494
     },
     {
       "epoch": 0.22036727879799667,
-      "grad_norm": 3.5661749839782715,
-      "learning_rate": 0.00010345506413744726,
-      "loss": 2.304,
+      "grad_norm": 3.349238872528076,
+      "learning_rate": 0.00010376901826699348,
+      "loss": 2.2104,
       "step": 495
     },
     {
       "epoch": 0.2208124652198108,
-      "grad_norm": 2.4347705841064453,
-      "learning_rate": 0.00010314107590781284,
-      "loss": 1.6218,
+      "grad_norm": 2.841289520263672,
+      "learning_rate": 0.00010345506413744726,
+      "loss": 1.696,
       "step": 496
     },
     {
       "epoch": 0.22125765164162492,
-      "grad_norm": 2.070305109024048,
-      "learning_rate": 0.00010282705667702734,
-      "loss": 2.0498,
+      "grad_norm": 2.52150297164917,
+      "learning_rate": 0.00010314107590781284,
+      "loss": 2.0612,
       "step": 497
     },
     {
       "epoch": 0.22170283806343907,
-      "grad_norm": 2.101881980895996,
-      "learning_rate": 0.00010251300954433376,
-      "loss": 1.6202,
+      "grad_norm": 2.2152016162872314,
+      "learning_rate": 0.00010282705667702734,
+      "loss": 1.6126,
       "step": 498
     },
     {
       "epoch": 0.2221480244852532,
-      "grad_norm": 1.2417051792144775,
-      "learning_rate": 0.00010219893760925052,
-      "loss": 0.9997,
+      "grad_norm": 1.3674941062927246,
+      "learning_rate": 0.00010251300954433376,
+      "loss": 0.9631,
       "step": 499
     },
     {
       "epoch": 0.22259321090706732,
-      "grad_norm": 2.754783868789673,
-      "learning_rate": 0.00010188484397154084,
-      "loss": 2.2696,
+      "grad_norm": 3.501506805419922,
+      "learning_rate": 0.00010219893760925052,
+      "loss": 2.3345,
       "step": 500
     },
     {
       "epoch": 0.22303839732888148,
-      "grad_norm": 2.7350590229034424,
-      "learning_rate": 0.00010157073173118208,
-      "loss": 1.888,
+      "grad_norm": 5.884027004241943,
+      "learning_rate": 0.00010188484397154084,
+      "loss": 1.8717,
       "step": 501
     },
     {
       "epoch": 0.2234835837506956,
-      "grad_norm": 2.365584373474121,
-      "learning_rate": 0.00010125660398833528,
-      "loss": 1.5181,
+      "grad_norm": 2.9893875122070312,
+      "learning_rate": 0.00010157073173118208,
+      "loss": 1.5762,
       "step": 502
     },
     {
       "epoch": 0.22392877017250973,
-      "grad_norm": 1.7909997701644897,
-      "learning_rate": 0.00010094246384331442,
-      "loss": 1.8491,
+      "grad_norm": 2.135575532913208,
+      "learning_rate": 0.00010125660398833528,
+      "loss": 1.922,
       "step": 503
     },
     {
       "epoch": 0.22437395659432388,
-      "grad_norm": 2.5978925228118896,
-      "learning_rate": 0.00010062831439655591,
-      "loss": 2.0812,
+      "grad_norm": 2.745893955230713,
+      "learning_rate": 0.00010094246384331442,
+      "loss": 2.1196,
       "step": 504
     },
     {
       "epoch": 0.224819143016138,
-      "grad_norm": 1.8115602731704712,
-      "learning_rate": 0.00010031415874858797,
-      "loss": 1.5947,
+      "grad_norm": 2.79962158203125,
+      "learning_rate": 0.00010062831439655591,
+      "loss": 1.5614,
       "step": 505
     },
     {
       "epoch": 0.22526432943795213,
-      "grad_norm": 3.372662305831909,
-      "learning_rate": 0.0001,
-      "loss": 1.846,
+      "grad_norm": 2.848219871520996,
+      "learning_rate": 0.00010031415874858797,
+      "loss": 2.0875,
       "step": 506
     },
     {
       "epoch": 0.22570951585976629,
-      "grad_norm": 2.081552505493164,
-      "learning_rate": 9.968584125141204e-05,
-      "loss": 1.7191,
+      "grad_norm": 1.9362163543701172,
+      "learning_rate": 0.0001,
+      "loss": 1.7142,
       "step": 507
     },
     {
       "epoch": 0.2261547022815804,
-      "grad_norm": 2.355083465576172,
-      "learning_rate": 9.937168560344412e-05,
-      "loss": 1.8958,
+      "grad_norm": 2.283254623413086,
+      "learning_rate": 9.968584125141204e-05,
+      "loss": 1.8632,
       "step": 508
     },
     {
       "epoch": 0.22659988870339454,
-      "grad_norm": 2.8670079708099365,
-      "learning_rate": 9.90575361566856e-05,
-      "loss": 2.0792,
+      "grad_norm": 3.0278165340423584,
+      "learning_rate": 9.937168560344412e-05,
+      "loss": 2.1201,
       "step": 509
     },
     {
       "epoch": 0.2270450751252087,
-      "grad_norm": 3.7465760707855225,
-      "learning_rate": 9.874339601166473e-05,
-      "loss": 2.3301,
+      "grad_norm": 3.8611364364624023,
+      "learning_rate": 9.90575361566856e-05,
+      "loss": 2.3123,
       "step": 510
     },
     {
       "epoch": 0.22749026154702282,
-      "grad_norm": 3.3605611324310303,
-      "learning_rate": 9.842926826881796e-05,
-      "loss": 3.4267,
+      "grad_norm": 3.6970300674438477,
+      "learning_rate": 9.874339601166473e-05,
+      "loss": 3.4317,
       "step": 511
     },
     {
       "epoch": 0.22793544796883694,
-      "grad_norm": 2.507578134536743,
-      "learning_rate": 9.81151560284592e-05,
-      "loss": 1.9814,
+      "grad_norm": 2.2782552242279053,
+      "learning_rate": 9.842926826881796e-05,
+      "loss": 1.9245,
       "step": 512
     },
     {
       "epoch": 0.2283806343906511,
-      "grad_norm": 4.343908786773682,
-      "learning_rate": 9.78010623907495e-05,
-      "loss": 2.2251,
+      "grad_norm": 4.838232040405273,
+      "learning_rate": 9.81151560284592e-05,
+      "loss": 2.3966,
       "step": 513
     },
     {
       "epoch": 0.22882582081246522,
-      "grad_norm": 2.300027370452881,
-      "learning_rate": 9.748699045566626e-05,
-      "loss": 1.7964,
+      "grad_norm": 2.7265279293060303,
+      "learning_rate": 9.78010623907495e-05,
+      "loss": 1.7596,
       "step": 514
     },
     {
       "epoch": 0.22927100723427934,
-      "grad_norm": 2.0999529361724854,
-      "learning_rate": 9.717294332297268e-05,
-      "loss": 2.1174,
+      "grad_norm": 2.08231782913208,
+      "learning_rate": 9.748699045566626e-05,
+      "loss": 2.1319,
       "step": 515
     },
     {
       "epoch": 0.2297161936560935,
-      "grad_norm": 2.6414802074432373,
-      "learning_rate": 9.685892409218717e-05,
-      "loss": 2.0422,
+      "grad_norm": 2.751110315322876,
+      "learning_rate": 9.717294332297268e-05,
+      "loss": 2.1303,
       "step": 516
     },
     {
       "epoch": 0.23016138007790762,
-      "grad_norm": 2.8301937580108643,
-      "learning_rate": 9.654493586255278e-05,
-      "loss": 2.0245,
+      "grad_norm": 2.737786293029785,
+      "learning_rate": 9.685892409218717e-05,
+      "loss": 2.0345,
       "step": 517
     },
     {
       "epoch": 0.23060656649972175,
-      "grad_norm": 1.4676358699798584,
-      "learning_rate": 9.623098173300654e-05,
-      "loss": 1.3448,
+      "grad_norm": 1.4824129343032837,
+      "learning_rate": 9.654493586255278e-05,
+      "loss": 1.308,
       "step": 518
     },
     {
       "epoch": 0.2310517529215359,
-      "grad_norm": 3.920909881591797,
-      "learning_rate": 9.591706480214901e-05,
-      "loss": 2.7722,
+      "grad_norm": 3.695892333984375,
+      "learning_rate": 9.623098173300654e-05,
+      "loss": 2.7335,
       "step": 519
     },
     {
       "epoch": 0.23149693934335003,
-      "grad_norm": 4.668717861175537,
-      "learning_rate": 9.560318816821353e-05,
-      "loss": 1.8226,
+      "grad_norm": 7.445088863372803,
+      "learning_rate": 9.591706480214901e-05,
+      "loss": 1.8408,
       "step": 520
     },
     {
       "epoch": 0.23194212576516415,
-      "grad_norm": 1.8420554399490356,
-      "learning_rate": 9.528935492903575e-05,
-      "loss": 1.0351,
+      "grad_norm": 1.7280323505401611,
+      "learning_rate": 9.560318816821353e-05,
+      "loss": 0.9995,
       "step": 521
     },
     {
       "epoch": 0.2323873121869783,
-      "grad_norm": 3.2772724628448486,
-      "learning_rate": 9.497556818202306e-05,
-      "loss": 2.7337,
+      "grad_norm": 5.599155426025391,
+      "learning_rate": 9.528935492903575e-05,
+      "loss": 2.7664,
       "step": 522
     },
     {
       "epoch": 0.23283249860879243,
-      "grad_norm": 2.2478621006011963,
-      "learning_rate": 9.466183102412395e-05,
-      "loss": 1.9922,
+      "grad_norm": 2.3892910480499268,
+      "learning_rate": 9.497556818202306e-05,
+      "loss": 1.9636,
       "step": 523
     },
     {
       "epoch": 0.23327768503060656,
-      "grad_norm": 1.188016653060913,
-      "learning_rate": 9.434814655179755e-05,
-      "loss": 1.0313,
+      "grad_norm": 1.4825445413589478,
+      "learning_rate": 9.466183102412395e-05,
+      "loss": 0.9973,
       "step": 524
     },
     {
       "epoch": 0.2337228714524207,
-      "grad_norm": 3.9148828983306885,
-      "learning_rate": 9.403451786098294e-05,
-      "loss": 2.2554,
+      "grad_norm": 3.159691572189331,
+      "learning_rate": 9.434814655179755e-05,
+      "loss": 2.2085,
       "step": 525
     },
     {
       "epoch": 0.23416805787423484,
-      "grad_norm": 2.4724814891815186,
-      "learning_rate": 9.372094804706867e-05,
-      "loss": 1.8411,
+      "grad_norm": 2.600054979324341,
+      "learning_rate": 9.403451786098294e-05,
+      "loss": 1.877,
       "step": 526
     },
     {
       "epoch": 0.23461324429604896,
-      "grad_norm": 4.861220359802246,
-      "learning_rate": 9.340744020486222e-05,
-      "loss": 1.7102,
+      "grad_norm": 5.829646587371826,
+      "learning_rate": 9.372094804706867e-05,
+      "loss": 1.6461,
       "step": 527
     },
     {
       "epoch": 0.23505843071786311,
-      "grad_norm": 1.9564683437347412,
-      "learning_rate": 9.309399742855942e-05,
-      "loss": 1.119,
+      "grad_norm": 1.8458348512649536,
+      "learning_rate": 9.340744020486222e-05,
+      "loss": 1.0875,
       "step": 528
     },
     {
       "epoch": 0.23550361713967724,
-      "grad_norm": 2.6515750885009766,
-      "learning_rate": 9.278062281171393e-05,
-      "loss": 2.0485,
+      "grad_norm": 2.3733482360839844,
+      "learning_rate": 9.309399742855942e-05,
+      "loss": 1.9736,
       "step": 529
     },
     {
       "epoch": 0.23594880356149137,
-      "grad_norm": 3.663888454437256,
-      "learning_rate": 9.246731944720675e-05,
-      "loss": 2.0759,
+      "grad_norm": 4.013689041137695,
+      "learning_rate": 9.278062281171393e-05,
+      "loss": 2.1294,
       "step": 530
     },
     {
       "epoch": 0.23639398998330552,
-      "grad_norm": 1.989341378211975,
-      "learning_rate": 9.215409042721552e-05,
-      "loss": 1.3344,
+      "grad_norm": 2.0923335552215576,
+      "learning_rate": 9.246731944720675e-05,
+      "loss": 1.3463,
       "step": 531
     },
     {
       "epoch": 0.23683917640511964,
-      "grad_norm": 2.359320878982544,
-      "learning_rate": 9.184093884318425e-05,
-      "loss": 1.3044,
+      "grad_norm": 2.23264479637146,
+      "learning_rate": 9.215409042721552e-05,
+      "loss": 1.3212,
       "step": 532
     },
     {
       "epoch": 0.23728436282693377,
-      "grad_norm": 4.149430274963379,
-      "learning_rate": 9.152786778579267e-05,
-      "loss": 2.1484,
+      "grad_norm": 4.511864185333252,
+      "learning_rate": 9.184093884318425e-05,
+      "loss": 2.1561,
       "step": 533
     },
     {
       "epoch": 0.23772954924874792,
-      "grad_norm": 3.5896191596984863,
-      "learning_rate": 9.121488034492569e-05,
-      "loss": 2.8477,
+      "grad_norm": 3.2981691360473633,
+      "learning_rate": 9.152786778579267e-05,
+      "loss": 2.9609,
       "step": 534
     },
     {
       "epoch": 0.23817473567056205,
-      "grad_norm": 2.467137336730957,
-      "learning_rate": 9.090197960964301e-05,
-      "loss": 1.5795,
+      "grad_norm": 3.0374643802642822,
+      "learning_rate": 9.121488034492569e-05,
+      "loss": 1.5647,
       "step": 535
     },
     {
       "epoch": 0.23861992209237617,
-      "grad_norm": 3.346538543701172,
-      "learning_rate": 9.058916866814858e-05,
-      "loss": 2.2605,
+      "grad_norm": 3.0681955814361572,
+      "learning_rate": 9.090197960964301e-05,
+      "loss": 2.2617,
       "step": 536
     },
     {
       "epoch": 0.23906510851419033,
-      "grad_norm": 3.115483522415161,
-      "learning_rate": 9.027645060776006e-05,
-      "loss": 1.9898,
+      "grad_norm": 2.6293389797210693,
+      "learning_rate": 9.058916866814858e-05,
+      "loss": 2.0123,
       "step": 537
     },
     {
       "epoch": 0.23951029493600445,
-      "grad_norm": 2.0012478828430176,
-      "learning_rate": 8.99638285148785e-05,
-      "loss": 1.3743,
+      "grad_norm": 1.9579615592956543,
+      "learning_rate": 9.027645060776006e-05,
+      "loss": 1.4156,
       "step": 538
     },
     {
       "epoch": 0.23995548135781858,
-      "grad_norm": 3.6035563945770264,
-      "learning_rate": 8.965130547495776e-05,
-      "loss": 1.6191,
+      "grad_norm": 3.5243144035339355,
+      "learning_rate": 8.99638285148785e-05,
+      "loss": 1.5838,
       "step": 539
     },
     {
       "epoch": 0.24040066777963273,
-      "grad_norm": 1.9701684713363647,
-      "learning_rate": 8.933888457247402e-05,
-      "loss": 1.7474,
+      "grad_norm": 2.244328022003174,
+      "learning_rate": 8.965130547495776e-05,
+      "loss": 1.7732,
       "step": 540
     },
     {
       "epoch": 0.24084585420144686,
-      "grad_norm": 3.0901401042938232,
-      "learning_rate": 8.902656889089548e-05,
-      "loss": 2.3074,
+      "grad_norm": 3.2589073181152344,
+      "learning_rate": 8.933888457247402e-05,
+      "loss": 2.2833,
       "step": 541
     },
     {
       "epoch": 0.24129104062326098,
-      "grad_norm": 3.2117514610290527,
-      "learning_rate": 8.871436151265184e-05,
-      "loss": 2.0538,
+      "grad_norm": 5.108296871185303,
+      "learning_rate": 8.902656889089548e-05,
+      "loss": 2.0562,
       "step": 542
     },
     {
       "epoch": 0.24173622704507514,
-      "grad_norm": 2.3342740535736084,
-      "learning_rate": 8.840226551910387e-05,
-      "loss": 1.1496,
+      "grad_norm": 2.2016584873199463,
+      "learning_rate": 8.871436151265184e-05,
+      "loss": 1.1455,
       "step": 543
     },
     {
       "epoch": 0.24218141346688926,
-      "grad_norm": 2.8328073024749756,
-      "learning_rate": 8.809028399051302e-05,
-      "loss": 1.2791,
+      "grad_norm": 3.027379035949707,
+      "learning_rate": 8.840226551910387e-05,
+      "loss": 1.3459,
       "step": 544
     },
     {
       "epoch": 0.2426265998887034,
-      "grad_norm": 2.9056038856506348,
-      "learning_rate": 8.777842000601105e-05,
-      "loss": 2.2656,
+      "grad_norm": 3.1470062732696533,
+      "learning_rate": 8.809028399051302e-05,
+      "loss": 2.2754,
       "step": 545
     },
     {
       "epoch": 0.24307178631051754,
-      "grad_norm": 2.6859729290008545,
-      "learning_rate": 8.746667664356956e-05,
-      "loss": 2.004,
+      "grad_norm": 2.895399570465088,
+      "learning_rate": 8.777842000601105e-05,
+      "loss": 2.0696,
       "step": 546
     },
     {
       "epoch": 0.24351697273233167,
-      "grad_norm": 2.280202627182007,
-      "learning_rate": 8.715505697996971e-05,
-      "loss": 2.0117,
+      "grad_norm": 2.655974864959717,
+      "learning_rate": 8.746667664356956e-05,
+      "loss": 2.0362,
       "step": 547
     },
     {
       "epoch": 0.2439621591541458,
-      "grad_norm": 3.593543291091919,
-      "learning_rate": 8.684356409077176e-05,
-      "loss": 2.4651,
+      "grad_norm": 3.3608624935150146,
+      "learning_rate": 8.715505697996971e-05,
+      "loss": 2.4649,
       "step": 548
     },
     {
       "epoch": 0.24440734557595994,
-      "grad_norm": 2.155273199081421,
-      "learning_rate": 8.653220105028474e-05,
-      "loss": 0.9082,
+      "grad_norm": 1.6335474252700806,
+      "learning_rate": 8.684356409077176e-05,
+      "loss": 0.9378,
       "step": 549
     },
     {
       "epoch": 0.24485253199777407,
-      "grad_norm": 3.196293354034424,
-      "learning_rate": 8.62209709315362e-05,
-      "loss": 2.4122,
+      "grad_norm": 3.400189161300659,
+      "learning_rate": 8.653220105028474e-05,
+      "loss": 2.4242,
       "step": 550
     },
     {
       "epoch": 0.2452977184195882,
-      "grad_norm": 2.0001351833343506,
-      "learning_rate": 8.590987680624174e-05,
-      "loss": 1.6249,
+      "grad_norm": 1.9661957025527954,
+      "learning_rate": 8.62209709315362e-05,
+      "loss": 1.6024,
       "step": 551
     },
     {
       "epoch": 0.24574290484140235,
-      "grad_norm": 2.7972264289855957,
-      "learning_rate": 8.559892174477479e-05,
-      "loss": 1.9658,
+      "grad_norm": 2.605260133743286,
+      "learning_rate": 8.590987680624174e-05,
+      "loss": 1.9463,
       "step": 552
     },
     {
       "epoch": 0.24618809126321647,
-      "grad_norm": 2.3301680088043213,
-      "learning_rate": 8.528810881613626e-05,
-      "loss": 2.2419,
+      "grad_norm": 2.2247166633605957,
+      "learning_rate": 8.559892174477479e-05,
+      "loss": 2.2634,
       "step": 553
     },
     {
       "epoch": 0.2466332776850306,
-      "grad_norm": 3.339153289794922,
-      "learning_rate": 8.497744108792429e-05,
-      "loss": 2.1285,
+      "grad_norm": 3.3085622787475586,
+      "learning_rate": 8.528810881613626e-05,
+      "loss": 2.0767,
       "step": 554
     },
     {
       "epoch": 0.24707846410684475,
-      "grad_norm": 4.093669891357422,
-      "learning_rate": 8.466692162630392e-05,
-      "loss": 1.7672,
+      "grad_norm": 3.5857832431793213,
+      "learning_rate": 8.497744108792429e-05,
+      "loss": 1.9394,
       "step": 555
     },
     {
       "epoch": 0.24752365052865888,
-      "grad_norm": 3.632969379425049,
-      "learning_rate": 8.435655349597689e-05,
-      "loss": 1.7379,
+      "grad_norm": 3.7488949298858643,
+      "learning_rate": 8.466692162630392e-05,
+      "loss": 1.7949,
       "step": 556
     },
     {
       "epoch": 0.247968836950473,
-      "grad_norm": 2.793414831161499,
-      "learning_rate": 8.404633976015134e-05,
-      "loss": 1.859,
+      "grad_norm": 2.594888925552368,
+      "learning_rate": 8.435655349597689e-05,
+      "loss": 1.8403,
       "step": 557
     },
     {
       "epoch": 0.24841402337228716,
-      "grad_norm": 2.6596624851226807,
-      "learning_rate": 8.373628348051165e-05,
-      "loss": 1.9903,
+      "grad_norm": 2.8361175060272217,
+      "learning_rate": 8.404633976015134e-05,
+      "loss": 2.0542,
       "step": 558
     },
     {
       "epoch": 0.24885920979410128,
-      "grad_norm": 2.221126079559326,
-      "learning_rate": 8.342638771718802e-05,
-      "loss": 2.6768,
+      "grad_norm": 2.4254066944122314,
+      "learning_rate": 8.373628348051165e-05,
+      "loss": 2.7412,
       "step": 559
     },
     {
       "epoch": 0.2493043962159154,
-      "grad_norm": 2.701017379760742,
-      "learning_rate": 8.311665552872662e-05,
-      "loss": 1.6827,
+      "grad_norm": 2.855569362640381,
+      "learning_rate": 8.342638771718802e-05,
+      "loss": 1.6349,
       "step": 560
     },
     {
       "epoch": 0.24974958263772956,
-      "grad_norm": 1.7237666845321655,
-      "learning_rate": 8.280708997205904e-05,
-      "loss": 1.5455,
+      "grad_norm": 1.6968199014663696,
+      "learning_rate": 8.311665552872662e-05,
+      "loss": 1.5019,
       "step": 561
     },
     {
       "epoch": 0.2501947690595437,
-      "grad_norm": 2.3445723056793213,
-      "learning_rate": 8.249769410247239e-05,
-      "loss": 1.7404,
+      "grad_norm": 2.3553712368011475,
+      "learning_rate": 8.280708997205904e-05,
+      "loss": 1.7379,
       "step": 562
     },
     {
       "epoch": 0.2506399554813578,
-      "grad_norm": 1.6703085899353027,
-      "learning_rate": 8.218847097357898e-05,
-      "loss": 1.5689,
+      "grad_norm": 1.8103502988815308,
+      "learning_rate": 8.249769410247239e-05,
+      "loss": 1.5911,
       "step": 563
     },
     {
       "epoch": 0.25108514190317194,
-      "grad_norm": 2.455984115600586,
-      "learning_rate": 8.187942363728625e-05,
-      "loss": 1.366,
+      "grad_norm": 2.619528293609619,
+      "learning_rate": 8.218847097357898e-05,
+      "loss": 1.4286,
       "step": 564
     },
     {
       "epoch": 0.25153032832498606,
-      "grad_norm": 2.8952829837799072,
-      "learning_rate": 8.157055514376666e-05,
-      "loss": 2.6693,
+      "grad_norm": 2.888099431991577,
+      "learning_rate": 8.187942363728625e-05,
+      "loss": 2.643,
       "step": 565
     },
     {
       "epoch": 0.25197551474680024,
-      "grad_norm": 2.0182900428771973,
-      "learning_rate": 8.126186854142752e-05,
-      "loss": 1.7768,
+      "grad_norm": 2.008424758911133,
+      "learning_rate": 8.157055514376666e-05,
+      "loss": 1.7771,
       "step": 566
     },
     {
       "epoch": 0.25242070116861437,
-      "grad_norm": 3.7107861042022705,
-      "learning_rate": 8.095336687688102e-05,
-      "loss": 2.4294,
+      "grad_norm": 3.2794687747955322,
+      "learning_rate": 8.126186854142752e-05,
+      "loss": 2.4204,
       "step": 567
     },
     {
       "epoch": 0.2528658875904285,
-      "grad_norm": 2.1526546478271484,
-      "learning_rate": 8.064505319491398e-05,
-      "loss": 2.0468,
+      "grad_norm": 4.072325229644775,
+      "learning_rate": 8.095336687688102e-05,
+      "loss": 2.0796,
       "step": 568
     },
     {
       "epoch": 0.2533110740122426,
-      "grad_norm": 2.4397218227386475,
-      "learning_rate": 8.033693053845801e-05,
-      "loss": 2.1954,
+      "grad_norm": 2.2015626430511475,
+      "learning_rate": 8.064505319491398e-05,
+      "loss": 2.1884,
       "step": 569
     },
     {
       "epoch": 0.25375626043405675,
-      "grad_norm": 2.494946241378784,
-      "learning_rate": 8.002900194855932e-05,
-      "loss": 1.9958,
+      "grad_norm": 2.523545742034912,
+      "learning_rate": 8.033693053845801e-05,
+      "loss": 2.0094,
       "step": 570
     },
     {
       "epoch": 0.25420144685587087,
-      "grad_norm": 4.778120994567871,
-      "learning_rate": 7.972127046434878e-05,
-      "loss": 2.9802,
+      "grad_norm": 4.743039608001709,
+      "learning_rate": 8.002900194855932e-05,
+      "loss": 3.0156,
       "step": 571
     },
     {
       "epoch": 0.25464663327768505,
-      "grad_norm": 7.366669178009033,
-      "learning_rate": 7.941373912301189e-05,
-      "loss": 1.6644,
+      "grad_norm": 2.6447269916534424,
+      "learning_rate": 7.972127046434878e-05,
+      "loss": 1.6751,
       "step": 572
     },
     {
       "epoch": 0.2550918196994992,
-      "grad_norm": 2.677617311477661,
-      "learning_rate": 7.910641095975886e-05,
-      "loss": 2.0434,
+      "grad_norm": 2.480916738510132,
+      "learning_rate": 7.941373912301189e-05,
+      "loss": 1.9236,
       "step": 573
     },
     {
       "epoch": 0.2555370061213133,
-      "grad_norm": 1.901222825050354,
-      "learning_rate": 7.879928900779456e-05,
-      "loss": 1.0331,
+      "grad_norm": 2.005185842514038,
+      "learning_rate": 7.910641095975886e-05,
+      "loss": 1.0807,
       "step": 574
     },
     {
       "epoch": 0.25598219254312743,
-      "grad_norm": 4.581434726715088,
-      "learning_rate": 7.849237629828869e-05,
-      "loss": 3.2133,
+      "grad_norm": 4.589725971221924,
+      "learning_rate": 7.879928900779456e-05,
+      "loss": 3.2364,
       "step": 575
     },
     {
       "epoch": 0.25642737896494155,
-      "grad_norm": 3.0468733310699463,
-      "learning_rate": 7.818567586034577e-05,
-      "loss": 1.6714,
+      "grad_norm": 3.0145387649536133,
+      "learning_rate": 7.849237629828869e-05,
+      "loss": 1.6721,
       "step": 576
     },
     {
       "epoch": 0.2568725653867557,
-      "grad_norm": 2.401651620864868,
-      "learning_rate": 7.787919072097531e-05,
-      "loss": 1.4202,
+      "grad_norm": 2.3507320880889893,
+      "learning_rate": 7.818567586034577e-05,
+      "loss": 1.4727,
       "step": 577
     },
     {
       "epoch": 0.25731775180856986,
-      "grad_norm": 3.3343799114227295,
-      "learning_rate": 7.75729239050619e-05,
-      "loss": 1.9961,
+      "grad_norm": 3.227512836456299,
+      "learning_rate": 7.787919072097531e-05,
+      "loss": 2.0266,
       "step": 578
     },
     {
       "epoch": 0.257762938230384,
-      "grad_norm": 2.1830644607543945,
-      "learning_rate": 7.726687843533538e-05,
-      "loss": 1.8748,
+      "grad_norm": 2.394864082336426,
+      "learning_rate": 7.75729239050619e-05,
+      "loss": 1.8211,
       "step": 579
     },
     {
       "epoch": 0.2582081246521981,
-      "grad_norm": 3.9290175437927246,
-      "learning_rate": 7.696105733234098e-05,
-      "loss": 2.0981,
+      "grad_norm": 2.832284688949585,
+      "learning_rate": 7.726687843533538e-05,
+      "loss": 2.1158,
       "step": 580
     },
     {
       "epoch": 0.25865331107401224,
-      "grad_norm": 2.6848762035369873,
-      "learning_rate": 7.66554636144095e-05,
-      "loss": 2.07,
+      "grad_norm": 2.7681491374969482,
+      "learning_rate": 7.696105733234098e-05,
+      "loss": 2.0653,
       "step": 581
     },
     {
       "epoch": 0.25909849749582636,
-      "grad_norm": 3.8313651084899902,
-      "learning_rate": 7.635010029762756e-05,
-      "loss": 1.8205,
+      "grad_norm": 4.447993278503418,
+      "learning_rate": 7.66554636144095e-05,
+      "loss": 1.7896,
       "step": 582
     },
     {
       "epoch": 0.2595436839176405,
-      "grad_norm": 2.1311826705932617,
-      "learning_rate": 7.604497039580785e-05,
-      "loss": 1.5328,
+      "grad_norm": 2.437788248062134,
+      "learning_rate": 7.635010029762756e-05,
+      "loss": 1.5824,
       "step": 583
     },
     {
       "epoch": 0.25998887033945467,
-      "grad_norm": 2.774587631225586,
-      "learning_rate": 7.574007692045928e-05,
-      "loss": 1.3775,
+      "grad_norm": 2.262251853942871,
+      "learning_rate": 7.604497039580785e-05,
+      "loss": 1.3313,
       "step": 584
     },
     {
       "epoch": 0.2604340567612688,
-      "grad_norm": 2.86027193069458,
-      "learning_rate": 7.543542288075739e-05,
-      "loss": 2.5958,
+      "grad_norm": 2.7486987113952637,
+      "learning_rate": 7.574007692045928e-05,
+      "loss": 2.56,
       "step": 585
     },
     {
       "epoch": 0.2608792431830829,
-      "grad_norm": 3.274968147277832,
-      "learning_rate": 7.513101128351454e-05,
-      "loss": 2.5216,
+      "grad_norm": 4.241098403930664,
+      "learning_rate": 7.543542288075739e-05,
+      "loss": 2.5701,
       "step": 586
     },
     {
       "epoch": 0.26132442960489705,
-      "grad_norm": 2.597700595855713,
-      "learning_rate": 7.48268451331503e-05,
-      "loss": 2.5979,
+      "grad_norm": 2.46081805229187,
+      "learning_rate": 7.513101128351454e-05,
+      "loss": 2.6092,
       "step": 587
     },
     {
       "epoch": 0.26176961602671117,
-      "grad_norm": 1.9474818706512451,
-      "learning_rate": 7.45229274316618e-05,
-      "loss": 1.4696,
+      "grad_norm": 1.9531766176223755,
+      "learning_rate": 7.48268451331503e-05,
+      "loss": 1.479,
       "step": 588
     },
     {
       "epoch": 0.2622148024485253,
-      "grad_norm": 3.209158182144165,
-      "learning_rate": 7.421926117859403e-05,
-      "loss": 2.4483,
+      "grad_norm": 3.3573310375213623,
+      "learning_rate": 7.45229274316618e-05,
+      "loss": 2.4908,
       "step": 589
     },
     {
       "epoch": 0.2626599888703395,
-      "grad_norm": 2.9960036277770996,
-      "learning_rate": 7.391584937101033e-05,
-      "loss": 2.4985,
+      "grad_norm": 3.4041333198547363,
+      "learning_rate": 7.421926117859403e-05,
+      "loss": 2.7018,
       "step": 590
     },
     {
       "epoch": 0.2631051752921536,
-      "grad_norm": 1.7019833326339722,
-      "learning_rate": 7.361269500346274e-05,
-      "loss": 1.3692,
+      "grad_norm": 1.837587833404541,
+      "learning_rate": 7.391584937101033e-05,
+      "loss": 1.3851,
       "step": 591
     },
     {
       "epoch": 0.26355036171396773,
-      "grad_norm": 3.178032398223877,
-      "learning_rate": 7.330980106796246e-05,
-      "loss": 1.9495,
+      "grad_norm": 3.824810028076172,
+      "learning_rate": 7.361269500346274e-05,
+      "loss": 2.0024,
       "step": 592
     },
     {
       "epoch": 0.26399554813578185,
-      "grad_norm": 3.838534355163574,
-      "learning_rate": 7.300717055395039e-05,
-      "loss": 2.3653,
+      "grad_norm": 3.523655652999878,
+      "learning_rate": 7.330980106796246e-05,
+      "loss": 2.3259,
       "step": 593
     },
     {
       "epoch": 0.264440734557596,
-      "grad_norm": 2.432239294052124,
-      "learning_rate": 7.270480644826749e-05,
-      "loss": 1.0926,
+      "grad_norm": 2.3373308181762695,
+      "learning_rate": 7.300717055395039e-05,
+      "loss": 1.0866,
       "step": 594
     },
     {
       "epoch": 0.2648859209794101,
-      "grad_norm": 3.833432197570801,
-      "learning_rate": 7.240271173512546e-05,
-      "loss": 2.6508,
+      "grad_norm": 3.652315139770508,
+      "learning_rate": 7.270480644826749e-05,
+      "loss": 2.5877,
       "step": 595
     },
     {
       "epoch": 0.2653311074012243,
-      "grad_norm": 3.8783698081970215,
-      "learning_rate": 7.210088939607708e-05,
-      "loss": 2.1285,
+      "grad_norm": 3.174226760864258,
+      "learning_rate": 7.240271173512546e-05,
+      "loss": 2.0778,
       "step": 596
     },
     {
       "epoch": 0.2657762938230384,
-      "grad_norm": 3.5125503540039062,
-      "learning_rate": 7.179934240998706e-05,
-      "loss": 2.6731,
+      "grad_norm": 3.0619466304779053,
+      "learning_rate": 7.210088939607708e-05,
+      "loss": 2.7003,
       "step": 597
     },
     {
       "epoch": 0.26622148024485254,
-      "grad_norm": 2.598796844482422,
-      "learning_rate": 7.149807375300239e-05,
-      "loss": 2.4519,
+      "grad_norm": 2.6117167472839355,
+      "learning_rate": 7.179934240998706e-05,
+      "loss": 2.4971,
       "step": 598
     },
     {
       "epoch": 0.26666666666666666,
-      "grad_norm": 2.9682698249816895,
-      "learning_rate": 7.119708639852312e-05,
-      "loss": 1.7751,
+      "grad_norm": 3.0407607555389404,
+      "learning_rate": 7.149807375300239e-05,
+      "loss": 1.8027,
       "step": 599
     },
     {
       "epoch": 0.2671118530884808,
-      "grad_norm": 2.220308780670166,
-      "learning_rate": 7.089638331717284e-05,
-      "loss": 1.7819,
+      "grad_norm": 1.9023070335388184,
+      "learning_rate": 7.119708639852312e-05,
+      "loss": 1.7886,
       "step": 600
     },
     {
       "epoch": 0.2675570395102949,
-      "grad_norm": 2.421492338180542,
-      "learning_rate": 7.059596747676962e-05,
-      "loss": 2.3129,
+      "grad_norm": 2.4321749210357666,
+      "learning_rate": 7.089638331717284e-05,
+      "loss": 2.2824,
       "step": 601
     },
     {
       "epoch": 0.2680022259321091,
-      "grad_norm": 2.011636257171631,
-      "learning_rate": 7.029584184229653e-05,
-      "loss": 1.264,
+      "grad_norm": 2.2802674770355225,
+      "learning_rate": 7.059596747676962e-05,
+      "loss": 1.2629,
       "step": 602
     },
     {
       "epoch": 0.2684474123539232,
-      "grad_norm": 1.722353458404541,
-      "learning_rate": 6.999600937587239e-05,
-      "loss": 1.4916,
+      "grad_norm": 1.8967430591583252,
+      "learning_rate": 7.029584184229653e-05,
+      "loss": 1.5382,
       "step": 603
     },
     {
       "epoch": 0.26889259877573735,
-      "grad_norm": 2.3344128131866455,
-      "learning_rate": 6.969647303672262e-05,
-      "loss": 2.0492,
+      "grad_norm": 2.444330930709839,
+      "learning_rate": 6.999600937587239e-05,
+      "loss": 2.0665,
       "step": 604
     },
     {
       "epoch": 0.26933778519755147,
-      "grad_norm": 10.07349681854248,
-      "learning_rate": 6.939723578114993e-05,
-      "loss": 2.213,
+      "grad_norm": 7.994617462158203,
+      "learning_rate": 6.969647303672262e-05,
+      "loss": 2.1398,
       "step": 605
     },
     {
       "epoch": 0.2697829716193656,
-      "grad_norm": 1.6907522678375244,
-      "learning_rate": 6.909830056250527e-05,
-      "loss": 1.3918,
+      "grad_norm": 1.7072683572769165,
+      "learning_rate": 6.939723578114993e-05,
+      "loss": 1.3992,
       "step": 606
     },
     {
       "epoch": 0.2702281580411797,
-      "grad_norm": 3.627065896987915,
-      "learning_rate": 6.879967033115853e-05,
-      "loss": 2.4359,
+      "grad_norm": 3.6613316535949707,
+      "learning_rate": 6.909830056250527e-05,
+      "loss": 2.4227,
       "step": 607
     },
     {
       "epoch": 0.2706733444629939,
-      "grad_norm": 1.5359236001968384,
-      "learning_rate": 6.850134803446954e-05,
-      "loss": 1.1468,
+      "grad_norm": 1.5923817157745361,
+      "learning_rate": 6.879967033115853e-05,
+      "loss": 1.1673,
       "step": 608
     },
     {
       "epoch": 0.27111853088480803,
-      "grad_norm": 2.957019567489624,
-      "learning_rate": 6.820333661675893e-05,
-      "loss": 1.8714,
+      "grad_norm": 2.9673140048980713,
+      "learning_rate": 6.850134803446954e-05,
+      "loss": 1.8865,
       "step": 609
     },
     {
       "epoch": 0.27156371730662215,
-      "grad_norm": 2.876551389694214,
-      "learning_rate": 6.790563901927907e-05,
-      "loss": 1.946,
+      "grad_norm": 2.5711653232574463,
+      "learning_rate": 6.820333661675893e-05,
+      "loss": 1.924,
       "step": 610
     },
     {
       "epoch": 0.2720089037284363,
-      "grad_norm": 2.3107662200927734,
-      "learning_rate": 6.760825818018508e-05,
-      "loss": 2.3033,
+      "grad_norm": 2.23542857170105,
+      "learning_rate": 6.790563901927907e-05,
+      "loss": 2.3644,
       "step": 611
     },
     {
       "epoch": 0.2724540901502504,
-      "grad_norm": 1.7891398668289185,
-      "learning_rate": 6.731119703450577e-05,
-      "loss": 1.8722,
+      "grad_norm": 1.7758067846298218,
+      "learning_rate": 6.760825818018508e-05,
+      "loss": 1.8761,
       "step": 612
     },
     {
       "epoch": 0.27289927657206453,
-      "grad_norm": 1.788131594657898,
-      "learning_rate": 6.701445851411472e-05,
-      "loss": 1.3048,
+      "grad_norm": 1.9661537408828735,
+      "learning_rate": 6.731119703450577e-05,
+      "loss": 1.3787,
       "step": 613
     },
     {
       "epoch": 0.2733444629938787,
-      "grad_norm": 2.7209267616271973,
-      "learning_rate": 6.671804554770135e-05,
-      "loss": 2.1986,
+      "grad_norm": 2.7200584411621094,
+      "learning_rate": 6.701445851411472e-05,
+      "loss": 2.1951,
       "step": 614
     },
     {
       "epoch": 0.27378964941569284,
-      "grad_norm": 3.6044182777404785,
-      "learning_rate": 6.642196106074194e-05,
-      "loss": 2.3438,
+      "grad_norm": 3.6363039016723633,
+      "learning_rate": 6.671804554770135e-05,
+      "loss": 2.3486,
       "step": 615
     },
     {
       "epoch": 0.27423483583750696,
-      "grad_norm": 3.2944424152374268,
-      "learning_rate": 6.612620797547087e-05,
-      "loss": 2.0854,
+      "grad_norm": 8.95522403717041,
+      "learning_rate": 6.642196106074194e-05,
+      "loss": 2.2566,
       "step": 616
     },
     {
       "epoch": 0.2746800222593211,
-      "grad_norm": 3.1867449283599854,
-      "learning_rate": 6.583078921085167e-05,
-      "loss": 1.9628,
+      "grad_norm": 3.6754941940307617,
+      "learning_rate": 6.612620797547087e-05,
+      "loss": 2.0391,
       "step": 617
     },
     {
       "epoch": 0.2751252086811352,
-      "grad_norm": 1.6597883701324463,
-      "learning_rate": 6.55357076825483e-05,
-      "loss": 1.0473,
+      "grad_norm": 1.4028700590133667,
+      "learning_rate": 6.583078921085167e-05,
+      "loss": 1.0158,
       "step": 618
     },
     {
       "epoch": 0.27557039510294934,
-      "grad_norm": 4.027482986450195,
-      "learning_rate": 6.52409663028963e-05,
-      "loss": 2.1282,
+      "grad_norm": 2.4848408699035645,
+      "learning_rate": 6.55357076825483e-05,
+      "loss": 2.1528,
       "step": 619
     },
     {
       "epoch": 0.2760155815247635,
-      "grad_norm": 2.803708791732788,
-      "learning_rate": 6.494656798087412e-05,
-      "loss": 1.8455,
+      "grad_norm": 2.78027081489563,
+      "learning_rate": 6.52409663028963e-05,
+      "loss": 1.8844,
       "step": 620
     },
     {
       "epoch": 0.27646076794657765,
-      "grad_norm": 3.2381043434143066,
-      "learning_rate": 6.465251562207431e-05,
-      "loss": 2.3748,
+      "grad_norm": 2.730515480041504,
+      "learning_rate": 6.494656798087412e-05,
+      "loss": 2.3197,
       "step": 621
     },
     {
       "epoch": 0.27690595436839177,
-      "grad_norm": 3.580366373062134,
-      "learning_rate": 6.435881212867493e-05,
-      "loss": 2.5845,
+      "grad_norm": 9.125295639038086,
+      "learning_rate": 6.465251562207431e-05,
+      "loss": 2.5809,
       "step": 622
     },
     {
       "epoch": 0.2773511407902059,
-      "grad_norm": 4.986792087554932,
-      "learning_rate": 6.406546039941094e-05,
-      "loss": 2.0802,
+      "grad_norm": 4.174372673034668,
+      "learning_rate": 6.435881212867493e-05,
+      "loss": 2.1089,
       "step": 623
     },
     {
       "epoch": 0.27779632721202,
-      "grad_norm": 1.9550381898880005,
-      "learning_rate": 6.377246332954544e-05,
-      "loss": 1.6063,
+      "grad_norm": 1.8563311100006104,
+      "learning_rate": 6.406546039941094e-05,
+      "loss": 1.6145,
       "step": 624
     },
     {
       "epoch": 0.27824151363383415,
-      "grad_norm": 2.505039930343628,
-      "learning_rate": 6.347982381084123e-05,
-      "loss": 1.6332,
+      "grad_norm": 2.578172445297241,
+      "learning_rate": 6.377246332954544e-05,
+      "loss": 1.5995,
       "step": 625
     },
     {
       "epoch": 0.27868670005564833,
-      "grad_norm": 1.7756378650665283,
-      "learning_rate": 6.318754473153221e-05,
-      "loss": 1.3104,
+      "grad_norm": 1.9171854257583618,
+      "learning_rate": 6.347982381084123e-05,
+      "loss": 1.2731,
       "step": 626
     },
     {
       "epoch": 0.27913188647746245,
-      "grad_norm": 2.1896908283233643,
-      "learning_rate": 6.289562897629492e-05,
-      "loss": 1.7569,
+      "grad_norm": 2.240983724594116,
+      "learning_rate": 6.318754473153221e-05,
+      "loss": 1.7384,
       "step": 627
     },
     {
       "epoch": 0.2795770728992766,
-      "grad_norm": 1.8758255243301392,
-      "learning_rate": 6.260407942621998e-05,
-      "loss": 1.756,
+      "grad_norm": 1.9845826625823975,
+      "learning_rate": 6.289562897629492e-05,
+      "loss": 1.7666,
       "step": 628
     },
     {
       "epoch": 0.2800222593210907,
-      "grad_norm": 3.3655426502227783,
-      "learning_rate": 6.231289895878375e-05,
-      "loss": 2.0823,
+      "grad_norm": 3.1939802169799805,
+      "learning_rate": 6.260407942621998e-05,
+      "loss": 2.0571,
       "step": 629
     },
     {
       "epoch": 0.28046744574290483,
-      "grad_norm": 3.293968915939331,
-      "learning_rate": 6.20220904478199e-05,
-      "loss": 1.9035,
+      "grad_norm": 3.2095515727996826,
+      "learning_rate": 6.231289895878375e-05,
+      "loss": 1.8829,
       "step": 630
     },
     {
       "epoch": 0.28091263216471896,
-      "grad_norm": 3.167731285095215,
-      "learning_rate": 6.173165676349103e-05,
-      "loss": 1.5836,
+      "grad_norm": 2.775292158126831,
+      "learning_rate": 6.20220904478199e-05,
+      "loss": 1.5986,
       "step": 631
     },
     {
       "epoch": 0.28135781858653314,
-      "grad_norm": 3.473268985748291,
-      "learning_rate": 6.144160077226036e-05,
-      "loss": 1.7259,
+      "grad_norm": 3.0598955154418945,
+      "learning_rate": 6.173165676349103e-05,
+      "loss": 1.6686,
       "step": 632
     },
     {
       "epoch": 0.28180300500834726,
-      "grad_norm": 2.428987979888916,
-      "learning_rate": 6.11519253368634e-05,
-      "loss": 1.5845,
+      "grad_norm": 1.9700332880020142,
+      "learning_rate": 6.144160077226036e-05,
+      "loss": 1.5556,
       "step": 633
     },
     {
       "epoch": 0.2822481914301614,
-      "grad_norm": 3.5516490936279297,
-      "learning_rate": 6.086263331627976e-05,
-      "loss": 1.5608,
+      "grad_norm": 2.914013624191284,
+      "learning_rate": 6.11519253368634e-05,
+      "loss": 1.5223,
       "step": 634
     },
     {
       "epoch": 0.2826933778519755,
-      "grad_norm": 4.088588714599609,
-      "learning_rate": 6.05737275657049e-05,
-      "loss": 2.4316,
+      "grad_norm": 3.7892260551452637,
+      "learning_rate": 6.086263331627976e-05,
+      "loss": 2.5045,
       "step": 635
     },
     {
       "epoch": 0.28313856427378964,
-      "grad_norm": 3.0667595863342285,
-      "learning_rate": 6.0285210936521955e-05,
-      "loss": 1.3786,
+      "grad_norm": 3.4939780235290527,
+      "learning_rate": 6.05737275657049e-05,
+      "loss": 1.3862,
       "step": 636
     },
     {
       "epoch": 0.28358375069560376,
-      "grad_norm": 2.605963945388794,
-      "learning_rate": 5.999708627627354e-05,
-      "loss": 1.7549,
+      "grad_norm": 2.4467647075653076,
+      "learning_rate": 6.0285210936521955e-05,
+      "loss": 1.7271,
       "step": 637
     },
     {
       "epoch": 0.28402893711741795,
-      "grad_norm": 3.6883420944213867,
-      "learning_rate": 5.9709356428633746e-05,
-      "loss": 2.2541,
+      "grad_norm": 3.2485249042510986,
+      "learning_rate": 5.999708627627354e-05,
+      "loss": 2.1898,
       "step": 638
     },
     {
       "epoch": 0.28447412353923207,
-      "grad_norm": 2.800830364227295,
-      "learning_rate": 5.9422024233380013e-05,
-      "loss": 1.8305,
+      "grad_norm": 2.478696584701538,
+      "learning_rate": 5.9709356428633746e-05,
+      "loss": 1.7902,
       "step": 639
     },
     {
       "epoch": 0.2849193099610462,
-      "grad_norm": 2.8093373775482178,
-      "learning_rate": 5.913509252636511e-05,
-      "loss": 2.8368,
+      "grad_norm": 2.8872196674346924,
+      "learning_rate": 5.9422024233380013e-05,
+      "loss": 2.8058,
       "step": 640
     },
     {
       "epoch": 0.2853644963828603,
-      "grad_norm": 2.3125829696655273,
-      "learning_rate": 5.884856413948913e-05,
-      "loss": 1.0694,
+      "grad_norm": 2.1622273921966553,
+      "learning_rate": 5.913509252636511e-05,
+      "loss": 1.0766,
       "step": 641
     },
     {
       "epoch": 0.28580968280467445,
-      "grad_norm": 2.6709086894989014,
-      "learning_rate": 5.856244190067159e-05,
-      "loss": 1.6645,
+      "grad_norm": 2.5097506046295166,
+      "learning_rate": 5.884856413948913e-05,
+      "loss": 1.652,
       "step": 642
     },
     {
       "epoch": 0.2862548692264886,
-      "grad_norm": 3.269890546798706,
-      "learning_rate": 5.82767286338235e-05,
-      "loss": 1.9574,
+      "grad_norm": 2.859157085418701,
+      "learning_rate": 5.856244190067159e-05,
+      "loss": 1.946,
       "step": 643
     },
     {
       "epoch": 0.28670005564830275,
-      "grad_norm": 6.454814910888672,
-      "learning_rate": 5.799142715881938e-05,
-      "loss": 1.9828,
+      "grad_norm": 2.902848958969116,
+      "learning_rate": 5.82767286338235e-05,
+      "loss": 1.8221,
       "step": 644
     },
     {
       "epoch": 0.2871452420701169,
-      "grad_norm": 2.483273983001709,
-      "learning_rate": 5.770654029146969e-05,
-      "loss": 1.1946,
+      "grad_norm": 2.4782211780548096,
+      "learning_rate": 5.799142715881938e-05,
+      "loss": 1.178,
       "step": 645
     },
     {
       "epoch": 0.287590428491931,
-      "grad_norm": 1.7462166547775269,
-      "learning_rate": 5.7422070843492734e-05,
-      "loss": 1.4047,
+      "grad_norm": 1.7475439310073853,
+      "learning_rate": 5.770654029146969e-05,
+      "loss": 1.3781,
       "step": 646
     },
     {
       "epoch": 0.28803561491374513,
-      "grad_norm": 4.918879508972168,
-      "learning_rate": 5.713802162248718e-05,
-      "loss": 2.362,
+      "grad_norm": 3.172783136367798,
+      "learning_rate": 5.7422070843492734e-05,
+      "loss": 2.3546,
       "step": 647
     },
     {
       "epoch": 0.28848080133555926,
-      "grad_norm": 2.5706145763397217,
-      "learning_rate": 5.6854395431904094e-05,
-      "loss": 2.0059,
+      "grad_norm": 2.8005614280700684,
+      "learning_rate": 5.713802162248718e-05,
+      "loss": 1.9669,
       "step": 648
     },
     {
       "epoch": 0.2889259877573734,
-      "grad_norm": 3.0318686962127686,
-      "learning_rate": 5.657119507101954e-05,
-      "loss": 2.7449,
+      "grad_norm": 3.17270565032959,
+      "learning_rate": 5.6854395431904094e-05,
+      "loss": 2.7537,
       "step": 649
     },
     {
       "epoch": 0.28937117417918756,
-      "grad_norm": 1.3213516473770142,
-      "learning_rate": 5.6288423334906735e-05,
-      "loss": 1.1325,
+      "grad_norm": 1.226949691772461,
+      "learning_rate": 5.657119507101954e-05,
+      "loss": 1.1319,
       "step": 650
     },
     {
       "epoch": 0.2898163606010017,
-      "grad_norm": 3.3162622451782227,
-      "learning_rate": 5.6006083014408484e-05,
-      "loss": 2.1244,
+      "grad_norm": 3.1507577896118164,
+      "learning_rate": 5.6288423334906735e-05,
+      "loss": 2.1694,
       "step": 651
     },
     {
       "epoch": 0.2902615470228158,
-      "grad_norm": 2.8271548748016357,
-      "learning_rate": 5.572417689610987e-05,
-      "loss": 2.6029,
+      "grad_norm": 3.0338058471679688,
+      "learning_rate": 5.6006083014408484e-05,
+      "loss": 2.559,
       "step": 652
     },
     {
       "epoch": 0.29070673344462994,
-      "grad_norm": 3.4697909355163574,
-      "learning_rate": 5.544270776231038e-05,
-      "loss": 1.8302,
+      "grad_norm": 2.516258955001831,
+      "learning_rate": 5.572417689610987e-05,
+      "loss": 1.8034,
       "step": 653
     },
     {
       "epoch": 0.29115191986644406,
-      "grad_norm": 2.9080650806427,
-      "learning_rate": 5.5161678390996796e-05,
-      "loss": 2.4099,
+      "grad_norm": 3.0269787311553955,
+      "learning_rate": 5.544270776231038e-05,
+      "loss": 2.4885,
       "step": 654
     },
     {
       "epoch": 0.2915971062882582,
-      "grad_norm": 2.8306384086608887,
-      "learning_rate": 5.488109155581549e-05,
-      "loss": 2.1103,
+      "grad_norm": 3.1893317699432373,
+      "learning_rate": 5.5161678390996796e-05,
+      "loss": 2.0873,
       "step": 655
     },
     {
       "epoch": 0.2920422927100723,
-      "grad_norm": 3.779367208480835,
-      "learning_rate": 5.4600950026045326e-05,
-      "loss": 2.3853,
+      "grad_norm": 3.51753306388855,
+      "learning_rate": 5.488109155581549e-05,
+      "loss": 2.4275,
       "step": 656
     },
     {
       "epoch": 0.2924874791318865,
-      "grad_norm": 3.2461814880371094,
-      "learning_rate": 5.4321256566570036e-05,
-      "loss": 1.6209,
+      "grad_norm": 4.858884811401367,
+      "learning_rate": 5.4600950026045326e-05,
+      "loss": 1.5764,
       "step": 657
     },
     {
       "epoch": 0.2929326655537006,
-      "grad_norm": 3.401667356491089,
-      "learning_rate": 5.404201393785122e-05,
-      "loss": 1.8184,
+      "grad_norm": 3.3828039169311523,
+      "learning_rate": 5.4321256566570036e-05,
+      "loss": 1.8998,
       "step": 658
     },
     {
       "epoch": 0.29337785197551475,
-      "grad_norm": 3.3337910175323486,
-      "learning_rate": 5.3763224895900846e-05,
-      "loss": 2.288,
+      "grad_norm": 3.035909652709961,
+      "learning_rate": 5.404201393785122e-05,
+      "loss": 2.2468,
       "step": 659
     },
     {
       "epoch": 0.2938230383973289,
-      "grad_norm": 4.496949672698975,
-      "learning_rate": 5.348489219225416e-05,
-      "loss": 2.0749,
+      "grad_norm": 4.460206985473633,
+      "learning_rate": 5.3763224895900846e-05,
+      "loss": 1.9738,
       "step": 660
     },
     {
       "epoch": 0.294268224819143,
-      "grad_norm": 4.434914588928223,
-      "learning_rate": 5.320701857394268e-05,
-      "loss": 2.3961,
+      "grad_norm": 3.6244258880615234,
+      "learning_rate": 5.348489219225416e-05,
+      "loss": 2.4416,
       "step": 661
     },
     {
       "epoch": 0.2947134112409571,
-      "grad_norm": 1.705828070640564,
-      "learning_rate": 5.292960678346675e-05,
-      "loss": 2.0444,
+      "grad_norm": 1.7858965396881104,
+      "learning_rate": 5.320701857394268e-05,
+      "loss": 2.0355,
       "step": 662
     },
     {
       "epoch": 0.2951585976627713,
-      "grad_norm": 3.431540012359619,
-      "learning_rate": 5.265265955876879e-05,
-      "loss": 1.4793,
+      "grad_norm": 3.171755075454712,
+      "learning_rate": 5.292960678346675e-05,
+      "loss": 1.4905,
       "step": 663
     },
     {
       "epoch": 0.29560378408458543,
-      "grad_norm": 2.5585508346557617,
-      "learning_rate": 5.237617963320608e-05,
-      "loss": 1.8598,
+      "grad_norm": 2.858747959136963,
+      "learning_rate": 5.265265955876879e-05,
+      "loss": 1.8204,
       "step": 664
     },
     {
       "epoch": 0.29604897050639956,
-      "grad_norm": 2.0425503253936768,
-      "learning_rate": 5.210016973552391e-05,
-      "loss": 1.6394,
+      "grad_norm": 2.019382953643799,
+      "learning_rate": 5.237617963320608e-05,
+      "loss": 1.6461,
       "step": 665
     },
     {
       "epoch": 0.2964941569282137,
-      "grad_norm": 1.9641846418380737,
-      "learning_rate": 5.182463258982846e-05,
-      "loss": 1.1607,
+      "grad_norm": 1.86653733253479,
+      "learning_rate": 5.210016973552391e-05,
+      "loss": 1.1889,
       "step": 666
     },
     {
       "epoch": 0.2969393433500278,
-      "grad_norm": 2.71909499168396,
-      "learning_rate": 5.1549570915560206e-05,
-      "loss": 1.9043,
+      "grad_norm": 2.653991937637329,
+      "learning_rate": 5.182463258982846e-05,
+      "loss": 1.9415,
       "step": 667
     },
     {
       "epoch": 0.29738452977184193,
-      "grad_norm": 1.7965915203094482,
-      "learning_rate": 5.127498742746675e-05,
-      "loss": 1.4473,
+      "grad_norm": 1.676102638244629,
+      "learning_rate": 5.1549570915560206e-05,
+      "loss": 1.4348,
       "step": 668
     },
     {
       "epoch": 0.2978297161936561,
-      "grad_norm": 2.792185068130493,
-      "learning_rate": 5.100088483557634e-05,
-      "loss": 2.1568,
+      "grad_norm": 2.8851091861724854,
+      "learning_rate": 5.127498742746675e-05,
+      "loss": 2.2143,
       "step": 669
     },
     {
       "epoch": 0.29827490261547024,
-      "grad_norm": 2.522604465484619,
-      "learning_rate": 5.072726584517086e-05,
-      "loss": 1.4188,
+      "grad_norm": 2.4972410202026367,
+      "learning_rate": 5.100088483557634e-05,
+      "loss": 1.5418,
       "step": 670
     },
     {
       "epoch": 0.29872008903728436,
-      "grad_norm": 2.1286098957061768,
-      "learning_rate": 5.045413315675924e-05,
-      "loss": 1.2097,
+      "grad_norm": 2.2214536666870117,
+      "learning_rate": 5.072726584517086e-05,
+      "loss": 1.2384,
       "step": 671
     },
     {
       "epoch": 0.2991652754590985,
-      "grad_norm": 2.812614917755127,
-      "learning_rate": 5.018148946605092e-05,
-      "loss": 2.0791,
+      "grad_norm": 2.9166228771209717,
+      "learning_rate": 5.045413315675924e-05,
+      "loss": 2.0722,
       "step": 672
     },
     {
       "epoch": 0.2996104618809126,
-      "grad_norm": 1.8847200870513916,
-      "learning_rate": 4.990933746392899e-05,
-      "loss": 1.6048,
+      "grad_norm": 1.8348429203033447,
+      "learning_rate": 5.018148946605092e-05,
+      "loss": 1.6178,
       "step": 673
     },
     {
       "epoch": 0.30005564830272674,
-      "grad_norm": 2.700967788696289,
-      "learning_rate": 4.9637679836423924e-05,
-      "loss": 1.9822,
+      "grad_norm": 2.423125982284546,
+      "learning_rate": 4.990933746392899e-05,
+      "loss": 1.9328,
       "step": 674
     },
     {
       "epoch": 0.3005008347245409,
-      "grad_norm": 5.683379650115967,
-      "learning_rate": 4.9366519264686725e-05,
-      "loss": 2.3686,
+      "grad_norm": 7.734646797180176,
+      "learning_rate": 4.9637679836423924e-05,
+      "loss": 2.3782,
       "step": 675
     },
     {
       "epoch": 0.30094602114635505,
-      "grad_norm": 2.7878029346466064,
-      "learning_rate": 4.909585842496287e-05,
-      "loss": 2.0972,
+      "grad_norm": 2.6672346591949463,
+      "learning_rate": 4.9366519264686725e-05,
+      "loss": 2.1491,
       "step": 676
     },
     {
       "epoch": 0.3013912075681692,
-      "grad_norm": 1.970714807510376,
-      "learning_rate": 4.8825699988565485e-05,
-      "loss": 1.8908,
+      "grad_norm": 2.255370616912842,
+      "learning_rate": 4.909585842496287e-05,
+      "loss": 1.8773,
       "step": 677
     },
     {
       "epoch": 0.3018363939899833,
-      "grad_norm": 3.5123045444488525,
-      "learning_rate": 4.8556046621849346e-05,
-      "loss": 2.4987,
+      "grad_norm": 4.085122108459473,
+      "learning_rate": 4.8825699988565485e-05,
+      "loss": 2.5306,
       "step": 678
     },
     {
       "epoch": 0.3022815804117974,
-      "grad_norm": 3.1310598850250244,
-      "learning_rate": 4.828690098618429e-05,
-      "loss": 2.2711,
+      "grad_norm": 3.448923110961914,
+      "learning_rate": 4.8556046621849346e-05,
+      "loss": 2.3004,
       "step": 679
     },
     {
       "epoch": 0.30272676683361155,
-      "grad_norm": 2.584655284881592,
-      "learning_rate": 4.8018265737929044e-05,
-      "loss": 1.9538,
+      "grad_norm": 2.694648265838623,
+      "learning_rate": 4.828690098618429e-05,
+      "loss": 1.9922,
       "step": 680
     },
     {
       "epoch": 0.30317195325542573,
-      "grad_norm": 2.380634307861328,
-      "learning_rate": 4.7750143528405126e-05,
-      "loss": 1.322,
+      "grad_norm": 2.3106231689453125,
+      "learning_rate": 4.8018265737929044e-05,
+      "loss": 1.3284,
       "step": 681
     },
     {
       "epoch": 0.30361713967723986,
-      "grad_norm": 3.004396438598633,
-      "learning_rate": 4.748253700387042e-05,
-      "loss": 2.1901,
+      "grad_norm": 3.892047882080078,
+      "learning_rate": 4.7750143528405126e-05,
+      "loss": 2.185,
       "step": 682
     },
     {
       "epoch": 0.304062326099054,
-      "grad_norm": 2.5817222595214844,
-      "learning_rate": 4.721544880549337e-05,
-      "loss": 1.9884,
+      "grad_norm": 2.5504989624023438,
+      "learning_rate": 4.748253700387042e-05,
+      "loss": 2.0292,
       "step": 683
     },
     {
       "epoch": 0.3045075125208681,
-      "grad_norm": 1.8147649765014648,
-      "learning_rate": 4.694888156932658e-05,
-      "loss": 1.6065,
+      "grad_norm": 1.9344456195831299,
+      "learning_rate": 4.721544880549337e-05,
+      "loss": 1.6198,
       "step": 684
     },
     {
       "epoch": 0.30495269894268223,
-      "grad_norm": 2.1298556327819824,
-      "learning_rate": 4.668283792628114e-05,
-      "loss": 2.1248,
+      "grad_norm": 2.216498613357544,
+      "learning_rate": 4.694888156932658e-05,
+      "loss": 2.1632,
       "step": 685
     },
     {
       "epoch": 0.30539788536449636,
-      "grad_norm": 2.9568581581115723,
-      "learning_rate": 4.6417320502100316e-05,
-      "loss": 2.8798,
+      "grad_norm": 2.8042759895324707,
+      "learning_rate": 4.668283792628114e-05,
+      "loss": 2.94,
       "step": 686
     },
     {
       "epoch": 0.30584307178631054,
-      "grad_norm": 3.2827939987182617,
-      "learning_rate": 4.615233191733398e-05,
-      "loss": 2.6892,
+      "grad_norm": 3.1668999195098877,
+      "learning_rate": 4.6417320502100316e-05,
+      "loss": 2.6903,
       "step": 687
     },
     {
       "epoch": 0.30628825820812466,
-      "grad_norm": 2.5867035388946533,
-      "learning_rate": 4.588787478731242e-05,
-      "loss": 2.105,
+      "grad_norm": 2.5685513019561768,
+      "learning_rate": 4.615233191733398e-05,
+      "loss": 2.1633,
       "step": 688
     },
     {
       "epoch": 0.3067334446299388,
-      "grad_norm": 3.4361913204193115,
-      "learning_rate": 4.5623951722120736e-05,
-      "loss": 2.7065,
+      "grad_norm": 3.140465497970581,
+      "learning_rate": 4.588787478731242e-05,
+      "loss": 2.6672,
       "step": 689
     },
     {
       "epoch": 0.3071786310517529,
-      "grad_norm": 3.75927996635437,
-      "learning_rate": 4.5360565326573104e-05,
-      "loss": 1.6896,
+      "grad_norm": 5.06754207611084,
+      "learning_rate": 4.5623951722120736e-05,
+      "loss": 1.5968,
       "step": 690
     },
     {
       "epoch": 0.30762381747356704,
-      "grad_norm": 3.1970183849334717,
-      "learning_rate": 4.5097718200186814e-05,
-      "loss": 2.5938,
+      "grad_norm": 3.0402727127075195,
+      "learning_rate": 4.5360565326573104e-05,
+      "loss": 2.5534,
       "step": 691
     },
     {
       "epoch": 0.30806900389538117,
-      "grad_norm": 2.316333770751953,
-      "learning_rate": 4.483541293715698e-05,
-      "loss": 2.1561,
+      "grad_norm": 2.324404239654541,
+      "learning_rate": 4.5097718200186814e-05,
+      "loss": 2.1212,
       "step": 692
     },
     {
       "epoch": 0.30851419031719535,
-      "grad_norm": 5.406709671020508,
-      "learning_rate": 4.457365212633058e-05,
-      "loss": 1.39,
+      "grad_norm": 2.1986916065216064,
+      "learning_rate": 4.483541293715698e-05,
+      "loss": 1.3312,
       "step": 693
     },
     {
       "epoch": 0.3089593767390095,
-      "grad_norm": 1.8153971433639526,
-      "learning_rate": 4.431243835118124e-05,
-      "loss": 1.6361,
+      "grad_norm": 1.8250253200531006,
+      "learning_rate": 4.457365212633058e-05,
+      "loss": 1.6241,
       "step": 694
     },
     {
       "epoch": 0.3094045631608236,
-      "grad_norm": 3.764073133468628,
-      "learning_rate": 4.4051774189783315e-05,
-      "loss": 2.0475,
+      "grad_norm": 2.8651695251464844,
+      "learning_rate": 4.431243835118124e-05,
+      "loss": 1.9434,
       "step": 695
     },
     {
       "epoch": 0.3098497495826377,
-      "grad_norm": 2.5098719596862793,
-      "learning_rate": 4.379166221478697e-05,
-      "loss": 1.5775,
+      "grad_norm": 2.2517638206481934,
+      "learning_rate": 4.4051774189783315e-05,
+      "loss": 1.5921,
       "step": 696
     },
     {
       "epoch": 0.31029493600445185,
-      "grad_norm": 2.3882479667663574,
-      "learning_rate": 4.3532104993392306e-05,
-      "loss": 2.1238,
+      "grad_norm": 2.332831382751465,
+      "learning_rate": 4.379166221478697e-05,
+      "loss": 2.1504,
       "step": 697
     },
     {
       "epoch": 0.310740122426266,
-      "grad_norm": 2.4165163040161133,
-      "learning_rate": 4.327310508732437e-05,
-      "loss": 2.2308,
+      "grad_norm": 2.4555563926696777,
+      "learning_rate": 4.3532104993392306e-05,
+      "loss": 2.285,
       "step": 698
     },
     {
       "epoch": 0.31118530884808016,
-      "grad_norm": 2.2906312942504883,
-      "learning_rate": 4.301466505280762e-05,
-      "loss": 2.2186,
+      "grad_norm": 2.427152156829834,
+      "learning_rate": 4.327310508732437e-05,
+      "loss": 2.2453,
       "step": 699
     },
     {
       "epoch": 0.3116304952698943,
-      "grad_norm": 3.0668275356292725,
-      "learning_rate": 4.2756787440540936e-05,
-      "loss": 1.7851,
+      "grad_norm": 3.151796340942383,
+      "learning_rate": 4.301466505280762e-05,
+      "loss": 1.7703,
       "step": 700
     },
     {
       "epoch": 0.3120756816917084,
-      "grad_norm": 2.195206642150879,
-      "learning_rate": 4.249947479567218e-05,
-      "loss": 1.6858,
+      "grad_norm": 2.0332770347595215,
+      "learning_rate": 4.2756787440540936e-05,
+      "loss": 1.6522,
       "step": 701
     },
     {
       "epoch": 0.31252086811352253,
-      "grad_norm": 2.6582798957824707,
-      "learning_rate": 4.224272965777326e-05,
-      "loss": 1.5366,
+      "grad_norm": 2.4257619380950928,
+      "learning_rate": 4.249947479567218e-05,
+      "loss": 1.5509,
       "step": 702
     },
     {
       "epoch": 0.31296605453533666,
-      "grad_norm": 1.982883095741272,
-      "learning_rate": 4.1986554560815096e-05,
-      "loss": 1.6462,
+      "grad_norm": 2.1774954795837402,
+      "learning_rate": 4.224272965777326e-05,
+      "loss": 1.637,
       "step": 703
     },
     {
       "epoch": 0.3134112409571508,
-      "grad_norm": 2.2738630771636963,
-      "learning_rate": 4.173095203314241e-05,
-      "loss": 1.7683,
+      "grad_norm": 2.6510019302368164,
+      "learning_rate": 4.1986554560815096e-05,
+      "loss": 1.8324,
       "step": 704
     },
     {
       "epoch": 0.31385642737896496,
-      "grad_norm": 2.549669027328491,
-      "learning_rate": 4.1475924597449024e-05,
-      "loss": 2.3111,
+      "grad_norm": 2.63783860206604,
+      "learning_rate": 4.173095203314241e-05,
+      "loss": 2.2903,
       "step": 705
     },
     {
       "epoch": 0.3143016138007791,
-      "grad_norm": 2.8641653060913086,
-      "learning_rate": 4.12214747707527e-05,
-      "loss": 1.7643,
+      "grad_norm": 2.778245210647583,
+      "learning_rate": 4.1475924597449024e-05,
+      "loss": 1.8128,
       "step": 706
     },
     {
       "epoch": 0.3147468002225932,
-      "grad_norm": 1.6562206745147705,
-      "learning_rate": 4.096760506437057e-05,
-      "loss": 1.6724,
+      "grad_norm": 1.6933324337005615,
+      "learning_rate": 4.12214747707527e-05,
+      "loss": 1.6582,
       "step": 707
     },
     {
       "epoch": 0.31519198664440734,
-      "grad_norm": 2.6927621364593506,
-      "learning_rate": 4.071431798389408e-05,
-      "loss": 1.4428,
+      "grad_norm": 2.9144859313964844,
+      "learning_rate": 4.096760506437057e-05,
+      "loss": 1.4754,
       "step": 708
     },
     {
       "epoch": 0.31563717306622147,
-      "grad_norm": 2.4343008995056152,
-      "learning_rate": 4.0461616029164526e-05,
-      "loss": 1.2359,
+      "grad_norm": 2.8968605995178223,
+      "learning_rate": 4.071431798389408e-05,
+      "loss": 1.2524,
       "step": 709
     },
     {
       "epoch": 0.3160823594880356,
-      "grad_norm": 2.6044414043426514,
-      "learning_rate": 4.020950169424815e-05,
-      "loss": 1.8076,
+      "grad_norm": 3.108628273010254,
+      "learning_rate": 4.0461616029164526e-05,
+      "loss": 1.9153,
       "step": 710
     },
     {
       "epoch": 0.3165275459098498,
-      "grad_norm": 4.747753143310547,
-      "learning_rate": 3.9957977467411615e-05,
-      "loss": 1.983,
+      "grad_norm": 4.1388092041015625,
+      "learning_rate": 4.020950169424815e-05,
+      "loss": 1.956,
       "step": 711
     },
     {
       "epoch": 0.3169727323316639,
-      "grad_norm": 4.754560947418213,
-      "learning_rate": 3.9707045831097555e-05,
-      "loss": 2.7843,
+      "grad_norm": 4.648570537567139,
+      "learning_rate": 3.9957977467411615e-05,
+      "loss": 2.7696,
       "step": 712
     },
     {
       "epoch": 0.317417918753478,
-      "grad_norm": 2.5929174423217773,
-      "learning_rate": 3.945670926189987e-05,
-      "loss": 2.3036,
+      "grad_norm": 2.6900453567504883,
+      "learning_rate": 3.9707045831097555e-05,
+      "loss": 2.2918,
       "step": 713
     },
     {
       "epoch": 0.31786310517529215,
-      "grad_norm": 3.1126303672790527,
-      "learning_rate": 3.920697023053949e-05,
-      "loss": 2.2455,
+      "grad_norm": 3.3735849857330322,
+      "learning_rate": 3.945670926189987e-05,
+      "loss": 2.1827,
       "step": 714
     },
     {
       "epoch": 0.3183082915971063,
-      "grad_norm": 3.110023260116577,
-      "learning_rate": 3.895783120183976e-05,
-      "loss": 1.8107,
+      "grad_norm": 3.3142940998077393,
+      "learning_rate": 3.920697023053949e-05,
+      "loss": 1.8947,
       "step": 715
     },
     {
       "epoch": 0.3187534780189204,
-      "grad_norm": 2.862748622894287,
-      "learning_rate": 3.8709294634702376e-05,
-      "loss": 1.9125,
+      "grad_norm": 2.8147215843200684,
+      "learning_rate": 3.895783120183976e-05,
+      "loss": 1.9302,
       "step": 716
     },
     {
       "epoch": 0.3191986644407346,
-      "grad_norm": 2.372352361679077,
-      "learning_rate": 3.846136298208285e-05,
-      "loss": 1.9124,
+      "grad_norm": 2.287973642349243,
+      "learning_rate": 3.8709294634702376e-05,
+      "loss": 1.8818,
       "step": 717
     },
     {
       "epoch": 0.3196438508625487,
-      "grad_norm": 3.2221293449401855,
-      "learning_rate": 3.821403869096658e-05,
-      "loss": 1.8325,
+      "grad_norm": 3.5175046920776367,
+      "learning_rate": 3.846136298208285e-05,
+      "loss": 1.8394,
       "step": 718
     },
     {
       "epoch": 0.32008903728436283,
-      "grad_norm": 2.2187700271606445,
-      "learning_rate": 3.796732420234443e-05,
-      "loss": 1.388,
+      "grad_norm": 1.9966739416122437,
+      "learning_rate": 3.821403869096658e-05,
+      "loss": 1.3228,
       "step": 719
     },
     {
       "epoch": 0.32053422370617696,
-      "grad_norm": 2.560734510421753,
-      "learning_rate": 3.7721221951188765e-05,
-      "loss": 1.374,
+      "grad_norm": 2.7500932216644287,
+      "learning_rate": 3.796732420234443e-05,
+      "loss": 1.41,
       "step": 720
     },
     {
       "epoch": 0.3209794101279911,
-      "grad_norm": 2.4914464950561523,
-      "learning_rate": 3.747573436642951e-05,
-      "loss": 2.0269,
+      "grad_norm": 2.388460159301758,
+      "learning_rate": 3.7721221951188765e-05,
+      "loss": 2.0217,
       "step": 721
     },
     {
       "epoch": 0.3214245965498052,
-      "grad_norm": 2.3180055618286133,
-      "learning_rate": 3.7230863870929964e-05,
-      "loss": 2.0662,
+      "grad_norm": 2.3415017127990723,
+      "learning_rate": 3.747573436642951e-05,
+      "loss": 2.0357,
       "step": 722
     },
     {
       "epoch": 0.3218697829716194,
-      "grad_norm": 2.14855694770813,
-      "learning_rate": 3.698661288146311e-05,
-      "loss": 1.3372,
+      "grad_norm": 1.972862958908081,
+      "learning_rate": 3.7230863870929964e-05,
+      "loss": 1.322,
       "step": 723
     },
     {
       "epoch": 0.3223149693934335,
-      "grad_norm": 3.403369426727295,
-      "learning_rate": 3.674298380868756e-05,
-      "loss": 2.3522,
+      "grad_norm": 3.0771045684814453,
+      "learning_rate": 3.698661288146311e-05,
+      "loss": 2.3352,
       "step": 724
     },
     {
       "epoch": 0.32276015581524764,
-      "grad_norm": 6.225140571594238,
-      "learning_rate": 3.649997905712396e-05,
-      "loss": 1.1184,
+      "grad_norm": 4.927875518798828,
+      "learning_rate": 3.674298380868756e-05,
+      "loss": 1.1276,
       "step": 725
     },
     {
       "epoch": 0.32320534223706177,
-      "grad_norm": 1.9929250478744507,
-      "learning_rate": 3.6257601025131026e-05,
-      "loss": 1.4805,
+      "grad_norm": 2.021491527557373,
+      "learning_rate": 3.649997905712396e-05,
+      "loss": 1.5137,
       "step": 726
     },
     {
       "epoch": 0.3236505286588759,
-      "grad_norm": 1.9394887685775757,
-      "learning_rate": 3.601585210488218e-05,
-      "loss": 1.9297,
+      "grad_norm": 1.8273261785507202,
+      "learning_rate": 3.6257601025131026e-05,
+      "loss": 1.8926,
       "step": 727
     },
     {
       "epoch": 0.32409571508069,
-      "grad_norm": 2.5257530212402344,
-      "learning_rate": 3.577473468234156e-05,
-      "loss": 1.6671,
+      "grad_norm": 1.9073346853256226,
+      "learning_rate": 3.601585210488218e-05,
+      "loss": 1.6412,
       "step": 728
     },
     {
       "epoch": 0.3245409015025042,
-      "grad_norm": 3.4196643829345703,
-      "learning_rate": 3.553425113724088e-05,
-      "loss": 2.7964,
+      "grad_norm": 3.6371872425079346,
+      "learning_rate": 3.577473468234156e-05,
+      "loss": 2.7159,
       "step": 729
     },
     {
       "epoch": 0.3249860879243183,
-      "grad_norm": 2.4060604572296143,
-      "learning_rate": 3.52944038430556e-05,
-      "loss": 1.7696,
+      "grad_norm": 2.4259378910064697,
+      "learning_rate": 3.553425113724088e-05,
+      "loss": 1.8411,
       "step": 730
     },
     {
       "epoch": 0.32543127434613245,
-      "grad_norm": 2.9572129249572754,
-      "learning_rate": 3.5055195166981645e-05,
-      "loss": 1.6015,
+      "grad_norm": 3.0658254623413086,
+      "learning_rate": 3.52944038430556e-05,
+      "loss": 1.5821,
       "step": 731
     },
     {
       "epoch": 0.3258764607679466,
-      "grad_norm": 2.2669548988342285,
-      "learning_rate": 3.481662746991214e-05,
-      "loss": 2.3187,
+      "grad_norm": 2.293248414993286,
+      "learning_rate": 3.5055195166981645e-05,
+      "loss": 2.3265,
       "step": 732
     },
     {
       "epoch": 0.3263216471897607,
-      "grad_norm": 4.523863792419434,
-      "learning_rate": 3.4578703106413904e-05,
-      "loss": 3.4186,
+      "grad_norm": 5.806252479553223,
+      "learning_rate": 3.481662746991214e-05,
+      "loss": 3.428,
       "step": 733
     },
     {
       "epoch": 0.3267668336115748,
-      "grad_norm": 2.4980828762054443,
-      "learning_rate": 3.4341424424704375e-05,
-      "loss": 1.3333,
+      "grad_norm": 2.4732987880706787,
+      "learning_rate": 3.4578703106413904e-05,
+      "loss": 1.3308,
       "step": 734
     },
     {
       "epoch": 0.327212020033389,
-      "grad_norm": 1.4893592596054077,
-      "learning_rate": 3.4104793766628304e-05,
-      "loss": 1.2549,
+      "grad_norm": 1.5391736030578613,
+      "learning_rate": 3.4341424424704375e-05,
+      "loss": 1.2824,
       "step": 735
     },
     {
       "epoch": 0.32765720645520313,
-      "grad_norm": 2.372570753097534,
-      "learning_rate": 3.386881346763483e-05,
-      "loss": 1.3731,
+      "grad_norm": 2.5730936527252197,
+      "learning_rate": 3.4104793766628304e-05,
+      "loss": 1.4633,
       "step": 736
     },
     {
       "epoch": 0.32810239287701726,
-      "grad_norm": 1.388790249824524,
-      "learning_rate": 3.363348585675414e-05,
-      "loss": 1.0765,
+      "grad_norm": 1.4847195148468018,
+      "learning_rate": 3.386881346763483e-05,
+      "loss": 1.0787,
       "step": 737
     },
     {
       "epoch": 0.3285475792988314,
-      "grad_norm": 3.7632062435150146,
-      "learning_rate": 3.339881325657484e-05,
-      "loss": 2.1809,
+      "grad_norm": 5.096831321716309,
+      "learning_rate": 3.363348585675414e-05,
+      "loss": 2.0878,
       "step": 738
     },
     {
       "epoch": 0.3289927657206455,
-      "grad_norm": 3.2846572399139404,
-      "learning_rate": 3.316479798322072e-05,
-      "loss": 1.7677,
+      "grad_norm": 2.9128997325897217,
+      "learning_rate": 3.339881325657484e-05,
+      "loss": 1.7954,
       "step": 739
     },
     {
       "epoch": 0.32943795214245963,
-      "grad_norm": 9.13051986694336,
-      "learning_rate": 3.2931442346328004e-05,
-      "loss": 1.6814,
+      "grad_norm": 3.4480271339416504,
+      "learning_rate": 3.316479798322072e-05,
+      "loss": 1.6608,
       "step": 740
     },
     {
       "epoch": 0.3298831385642738,
-      "grad_norm": 1.820805311203003,
-      "learning_rate": 3.269874864902269e-05,
-      "loss": 1.8115,
+      "grad_norm": 1.7849931716918945,
+      "learning_rate": 3.2931442346328004e-05,
+      "loss": 1.8315,
       "step": 741
     },
     {
       "epoch": 0.33032832498608794,
-      "grad_norm": 4.556207180023193,
-      "learning_rate": 3.246671918789755e-05,
-      "loss": 2.7417,
+      "grad_norm": 5.605867862701416,
+      "learning_rate": 3.269874864902269e-05,
+      "loss": 2.7352,
       "step": 742
     },
     {
       "epoch": 0.33077351140790207,
-      "grad_norm": 2.135354995727539,
-      "learning_rate": 3.223535625298979e-05,
-      "loss": 1.5414,
+      "grad_norm": 2.2710115909576416,
+      "learning_rate": 3.246671918789755e-05,
+      "loss": 1.4972,
       "step": 743
     },
     {
       "epoch": 0.3312186978297162,
-      "grad_norm": 12.734091758728027,
-      "learning_rate": 3.200466212775808e-05,
-      "loss": 1.7974,
+      "grad_norm": 40.97933578491211,
+      "learning_rate": 3.223535625298979e-05,
+      "loss": 2.0871,
       "step": 744
     },
     {
       "epoch": 0.3316638842515303,
-      "grad_norm": 2.428943634033203,
-      "learning_rate": 3.1774639089060363e-05,
-      "loss": 1.732,
+      "grad_norm": 2.290867567062378,
+      "learning_rate": 3.200466212775808e-05,
+      "loss": 1.7768,
       "step": 745
     },
     {
       "epoch": 0.33210907067334444,
-      "grad_norm": 2.2564094066619873,
-      "learning_rate": 3.154528940713113e-05,
-      "loss": 1.8338,
+      "grad_norm": 2.331289291381836,
+      "learning_rate": 3.1774639089060363e-05,
+      "loss": 1.7775,
       "step": 746
     },
     {
       "epoch": 0.3325542570951586,
-      "grad_norm": 2.3983113765716553,
-      "learning_rate": 3.1316615345559185e-05,
-      "loss": 1.7076,
+      "grad_norm": 2.2418811321258545,
+      "learning_rate": 3.154528940713113e-05,
+      "loss": 1.7201,
       "step": 747
     },
     {
       "epoch": 0.33299944351697275,
-      "grad_norm": 1.7274196147918701,
-      "learning_rate": 3.108861916126518e-05,
-      "loss": 1.2812,
+      "grad_norm": 2.4780850410461426,
+      "learning_rate": 3.1316615345559185e-05,
+      "loss": 1.2194,
       "step": 748
     },
     {
       "epoch": 0.3334446299387869,
-      "grad_norm": 2.3063457012176514,
-      "learning_rate": 3.086130310447937e-05,
-      "loss": 2.0274,
+      "grad_norm": 2.2363805770874023,
+      "learning_rate": 3.108861916126518e-05,
+      "loss": 2.0451,
       "step": 749
     },
     {
       "epoch": 0.333889816360601,
-      "grad_norm": 2.640960931777954,
-      "learning_rate": 3.063466941871952e-05,
-      "loss": 1.4717,
+      "grad_norm": 3.651272773742676,
+      "learning_rate": 3.086130310447937e-05,
+      "loss": 1.4697,
       "step": 750
     },
     {
       "epoch": 0.3343350027824151,
-      "grad_norm": 2.7002780437469482,
-      "learning_rate": 3.0408720340768572e-05,
-      "loss": 2.0358,
+      "grad_norm": 2.626664161682129,
+      "learning_rate": 3.063466941871952e-05,
+      "loss": 2.0324,
       "step": 751
     },
     {
       "epoch": 0.33478018920422925,
-      "grad_norm": 2.129476547241211,
-      "learning_rate": 3.018345810065275e-05,
-      "loss": 1.7658,
+      "grad_norm": 2.636204957962036,
+      "learning_rate": 3.0408720340768572e-05,
+      "loss": 1.7141,
       "step": 752
     },
     {
       "epoch": 0.33522537562604343,
-      "grad_norm": 3.172895669937134,
-      "learning_rate": 2.9958884921619367e-05,
-      "loss": 2.3525,
+      "grad_norm": 3.1656999588012695,
+      "learning_rate": 3.018345810065275e-05,
+      "loss": 2.3117,
       "step": 753
     },
     {
       "epoch": 0.33567056204785756,
-      "grad_norm": 3.458550453186035,
-      "learning_rate": 2.9735003020115092e-05,
-      "loss": 2.2845,
+      "grad_norm": 4.307328224182129,
+      "learning_rate": 2.9958884921619367e-05,
+      "loss": 2.2475,
       "step": 754
     },
     {
       "epoch": 0.3361157484696717,
-      "grad_norm": 2.624788284301758,
-      "learning_rate": 2.9511814605763855e-05,
-      "loss": 2.0448,
+      "grad_norm": 2.5681815147399902,
+      "learning_rate": 2.9735003020115092e-05,
+      "loss": 2.0617,
       "step": 755
     },
     {
       "epoch": 0.3365609348914858,
-      "grad_norm": 4.643282413482666,
-      "learning_rate": 2.9289321881345254e-05,
-      "loss": 2.8465,
+      "grad_norm": 4.749504566192627,
+      "learning_rate": 2.9511814605763855e-05,
+      "loss": 2.7351,
       "step": 756
     },
     {
       "epoch": 0.33700612131329993,
-      "grad_norm": 2.141824960708618,
-      "learning_rate": 2.9067527042772636e-05,
-      "loss": 1.795,
+      "grad_norm": 3.8484888076782227,
+      "learning_rate": 2.9289321881345254e-05,
+      "loss": 1.8341,
       "step": 757
     },
     {
       "epoch": 0.33745130773511406,
-      "grad_norm": 1.945853352546692,
-      "learning_rate": 2.8846432279071467e-05,
-      "loss": 1.4118,
+      "grad_norm": 2.8804805278778076,
+      "learning_rate": 2.9067527042772636e-05,
+      "loss": 1.3797,
       "step": 758
     },
     {
       "epoch": 0.33789649415692824,
-      "grad_norm": 2.440788507461548,
-      "learning_rate": 2.8626039772357882e-05,
-      "loss": 1.8631,
+      "grad_norm": 2.2779343128204346,
+      "learning_rate": 2.8846432279071467e-05,
+      "loss": 1.9275,
       "step": 759
     },
     {
       "epoch": 0.33834168057874237,
-      "grad_norm": 2.524973154067993,
-      "learning_rate": 2.840635169781688e-05,
-      "loss": 1.9311,
+      "grad_norm": 2.432508707046509,
+      "learning_rate": 2.8626039772357882e-05,
+      "loss": 1.8774,
       "step": 760
     },
     {
       "epoch": 0.3387868670005565,
-      "grad_norm": 2.88173508644104,
-      "learning_rate": 2.8187370223681132e-05,
-      "loss": 1.8423,
+      "grad_norm": 4.494417190551758,
+      "learning_rate": 2.840635169781688e-05,
+      "loss": 2.0044,
       "step": 761
     },
     {
       "epoch": 0.3392320534223706,
-      "grad_norm": 9.78935718536377,
-      "learning_rate": 2.7969097511209308e-05,
-      "loss": 1.7553,
+      "grad_norm": 9.79641056060791,
+      "learning_rate": 2.8187370223681132e-05,
+      "loss": 1.7594,
       "step": 762
     },
     {
       "epoch": 0.33967723984418474,
-      "grad_norm": 4.4023356437683105,
-      "learning_rate": 2.775153571466502e-05,
-      "loss": 2.0394,
+      "grad_norm": 4.837268829345703,
+      "learning_rate": 2.7969097511209308e-05,
+      "loss": 2.1129,
       "step": 763
     },
     {
       "epoch": 0.34012242626599887,
-      "grad_norm": 2.8718042373657227,
-      "learning_rate": 2.753468698129533e-05,
-      "loss": 1.5894,
+      "grad_norm": 3.084362268447876,
+      "learning_rate": 2.775153571466502e-05,
+      "loss": 1.6698,
       "step": 764
     },
     {
       "epoch": 0.34056761268781305,
-      "grad_norm": 1.903527855873108,
-      "learning_rate": 2.7318553451309726e-05,
-      "loss": 1.3534,
+      "grad_norm": 1.84263014793396,
+      "learning_rate": 2.753468698129533e-05,
+      "loss": 1.3117,
       "step": 765
     },
     {
       "epoch": 0.3410127991096272,
-      "grad_norm": 2.23759388923645,
-      "learning_rate": 2.7103137257858868e-05,
-      "loss": 1.871,
+      "grad_norm": 2.350165843963623,
+      "learning_rate": 2.7318553451309726e-05,
+      "loss": 1.8815,
       "step": 766
     },
     {
       "epoch": 0.3414579855314413,
-      "grad_norm": 3.0346007347106934,
-      "learning_rate": 2.688844052701359e-05,
-      "loss": 2.4884,
+      "grad_norm": 3.2090234756469727,
+      "learning_rate": 2.7103137257858868e-05,
+      "loss": 2.5512,
       "step": 767
     },
     {
       "epoch": 0.3419031719532554,
-      "grad_norm": 2.180615186691284,
-      "learning_rate": 2.6674465377744017e-05,
-      "loss": 1.3251,
+      "grad_norm": 1.9890962839126587,
+      "learning_rate": 2.688844052701359e-05,
+      "loss": 1.3836,
       "step": 768
     },
     {
       "epoch": 0.34234835837506955,
-      "grad_norm": 5.008700847625732,
-      "learning_rate": 2.646121392189841e-05,
-      "loss": 2.3742,
+      "grad_norm": 4.618078708648682,
+      "learning_rate": 2.6674465377744017e-05,
+      "loss": 2.4582,
       "step": 769
     },
     {
       "epoch": 0.3427935447968837,
-      "grad_norm": 2.506002426147461,
-      "learning_rate": 2.624868826418262e-05,
-      "loss": 1.756,
+      "grad_norm": 2.055314779281616,
+      "learning_rate": 2.646121392189841e-05,
+      "loss": 1.7066,
       "step": 770
     },
     {
       "epoch": 0.34323873121869786,
-      "grad_norm": 3.1411874294281006,
-      "learning_rate": 2.603689050213902e-05,
-      "loss": 2.8913,
+      "grad_norm": 2.973315477371216,
+      "learning_rate": 2.624868826418262e-05,
+      "loss": 2.8903,
       "step": 771
     },
     {
       "epoch": 0.343683917640512,
-      "grad_norm": 2.647324800491333,
-      "learning_rate": 2.582582272612609e-05,
-      "loss": 2.6637,
+      "grad_norm": 3.261991500854492,
+      "learning_rate": 2.603689050213902e-05,
+      "loss": 2.6735,
       "step": 772
     },
     {
       "epoch": 0.3441291040623261,
-      "grad_norm": 2.1316630840301514,
-      "learning_rate": 2.561548701929749e-05,
-      "loss": 1.3719,
+      "grad_norm": 2.076058864593506,
+      "learning_rate": 2.582582272612609e-05,
+      "loss": 1.381,
       "step": 773
     },
     {
       "epoch": 0.34457429048414023,
-      "grad_norm": 3.7572953701019287,
-      "learning_rate": 2.540588545758179e-05,
-      "loss": 1.7199,
+      "grad_norm": 3.1660914421081543,
+      "learning_rate": 2.561548701929749e-05,
+      "loss": 1.6826,
       "step": 774
     },
     {
       "epoch": 0.34501947690595436,
-      "grad_norm": 7.4356369972229,
-      "learning_rate": 2.5197020109661772e-05,
-      "loss": 1.4686,
+      "grad_norm": 5.427780628204346,
+      "learning_rate": 2.540588545758179e-05,
+      "loss": 1.5369,
       "step": 775
     },
     {
       "epoch": 0.3454646633277685,
-      "grad_norm": 2.2626800537109375,
-      "learning_rate": 2.4988893036954043e-05,
-      "loss": 1.8888,
+      "grad_norm": 2.2618162631988525,
+      "learning_rate": 2.5197020109661772e-05,
+      "loss": 1.8258,
       "step": 776
     },
     {
       "epoch": 0.34590984974958267,
-      "grad_norm": 3.9464282989501953,
-      "learning_rate": 2.4781506293588873e-05,
-      "loss": 2.2344,
+      "grad_norm": 4.402366638183594,
+      "learning_rate": 2.4988893036954043e-05,
+      "loss": 2.3216,
       "step": 777
     },
     {
       "epoch": 0.3463550361713968,
-      "grad_norm": 3.29831862449646,
-      "learning_rate": 2.4574861926389615e-05,
-      "loss": 2.4997,
+      "grad_norm": 2.900803804397583,
+      "learning_rate": 2.4781506293588873e-05,
+      "loss": 2.4731,
       "step": 778
     },
     {
       "epoch": 0.3468002225932109,
-      "grad_norm": 2.7223682403564453,
-      "learning_rate": 2.436896197485282e-05,
-      "loss": 2.3799,
+      "grad_norm": 2.6909384727478027,
+      "learning_rate": 2.4574861926389615e-05,
+      "loss": 2.3296,
       "step": 779
     },
     {
       "epoch": 0.34724540901502504,
-      "grad_norm": 2.570220947265625,
-      "learning_rate": 2.4163808471127812e-05,
-      "loss": 1.6605,
+      "grad_norm": 2.7630159854888916,
+      "learning_rate": 2.436896197485282e-05,
+      "loss": 1.7521,
       "step": 780
     },
     {
       "epoch": 0.34769059543683917,
-      "grad_norm": 2.4514687061309814,
-      "learning_rate": 2.3959403439996907e-05,
-      "loss": 1.3153,
+      "grad_norm": 2.2162787914276123,
+      "learning_rate": 2.4163808471127812e-05,
+      "loss": 1.3404,
       "step": 781
     },
     {
       "epoch": 0.3481357818586533,
-      "grad_norm": 4.2457733154296875,
-      "learning_rate": 2.37557488988552e-05,
-      "loss": 2.2081,
+      "grad_norm": 7.932774543762207,
+      "learning_rate": 2.3959403439996907e-05,
+      "loss": 2.2814,
       "step": 782
     },
     {
       "epoch": 0.3485809682804674,
-      "grad_norm": 2.6355888843536377,
-      "learning_rate": 2.3552846857690846e-05,
-      "loss": 2.1335,
+      "grad_norm": 2.8942651748657227,
+      "learning_rate": 2.37557488988552e-05,
+      "loss": 2.1944,
       "step": 783
     },
     {
       "epoch": 0.3490261547022816,
-      "grad_norm": 3.205240488052368,
-      "learning_rate": 2.3350699319065026e-05,
-      "loss": 2.0054,
+      "grad_norm": 3.0805366039276123,
+      "learning_rate": 2.3552846857690846e-05,
+      "loss": 1.9972,
       "step": 784
     },
     {
       "epoch": 0.3494713411240957,
-      "grad_norm": 2.160902976989746,
-      "learning_rate": 2.3149308278092342e-05,
-      "loss": 1.7905,
+      "grad_norm": 2.1117780208587646,
+      "learning_rate": 2.3350699319065026e-05,
+      "loss": 1.7653,
       "step": 785
     },
     {
       "epoch": 0.34991652754590985,
-      "grad_norm": 2.5220730304718018,
-      "learning_rate": 2.2948675722421086e-05,
-      "loss": 1.7417,
+      "grad_norm": 2.495792865753174,
+      "learning_rate": 2.3149308278092342e-05,
+      "loss": 1.7668,
       "step": 786
     },
     {
       "epoch": 0.350361713967724,
-      "grad_norm": 4.050189018249512,
-      "learning_rate": 2.2748803632213557e-05,
-      "loss": 2.9296,
+      "grad_norm": 3.875079870223999,
+      "learning_rate": 2.2948675722421086e-05,
+      "loss": 2.92,
       "step": 787
     },
     {
       "epoch": 0.3508069003895381,
-      "grad_norm": 1.7380865812301636,
-      "learning_rate": 2.254969398012663e-05,
-      "loss": 1.3506,
+      "grad_norm": 1.6618883609771729,
+      "learning_rate": 2.2748803632213557e-05,
+      "loss": 1.3489,
       "step": 788
     },
     {
       "epoch": 0.3512520868113522,
-      "grad_norm": 3.115523099899292,
-      "learning_rate": 2.235134873129213e-05,
-      "loss": 2.3948,
+      "grad_norm": 2.81689190864563,
+      "learning_rate": 2.254969398012663e-05,
+      "loss": 2.3773,
       "step": 789
     },
     {
       "epoch": 0.3516972732331664,
-      "grad_norm": 2.615837812423706,
-      "learning_rate": 2.2153769843297667e-05,
-      "loss": 1.865,
+      "grad_norm": 2.7152371406555176,
+      "learning_rate": 2.235134873129213e-05,
+      "loss": 1.8859,
       "step": 790
     },
     {
       "epoch": 0.35214245965498053,
-      "grad_norm": 2.9471611976623535,
-      "learning_rate": 2.195695926616702e-05,
-      "loss": 1.7431,
+      "grad_norm": 2.9427664279937744,
+      "learning_rate": 2.2153769843297667e-05,
+      "loss": 1.7977,
       "step": 791
     },
     {
       "epoch": 0.35258764607679466,
-      "grad_norm": 2.246422529220581,
-      "learning_rate": 2.1760918942341192e-05,
-      "loss": 1.0834,
+      "grad_norm": 2.521787643432617,
+      "learning_rate": 2.195695926616702e-05,
+      "loss": 1.0702,
       "step": 792
     },
     {
       "epoch": 0.3530328324986088,
-      "grad_norm": 3.3071248531341553,
-      "learning_rate": 2.1565650806658975e-05,
-      "loss": 2.3679,
+      "grad_norm": 3.2543270587921143,
+      "learning_rate": 2.1760918942341192e-05,
+      "loss": 2.4008,
       "step": 793
     },
     {
       "epoch": 0.3534780189204229,
-      "grad_norm": 1.6390622854232788,
-      "learning_rate": 2.137115678633811e-05,
-      "loss": 1.3217,
+      "grad_norm": 1.7347383499145508,
+      "learning_rate": 2.1565650806658975e-05,
+      "loss": 1.3265,
       "step": 794
     },
     {
       "epoch": 0.35392320534223703,
-      "grad_norm": 4.03280782699585,
-      "learning_rate": 2.1177438800956007e-05,
-      "loss": 1.9452,
+      "grad_norm": 3.650092363357544,
+      "learning_rate": 2.137115678633811e-05,
+      "loss": 1.9098,
       "step": 795
     },
     {
       "epoch": 0.3543683917640512,
-      "grad_norm": 15.388529777526855,
-      "learning_rate": 2.098449876243096e-05,
-      "loss": 2.3951,
+      "grad_norm": 3.2362852096557617,
+      "learning_rate": 2.1177438800956007e-05,
+      "loss": 2.4286,
       "step": 796
     },
     {
       "epoch": 0.35481357818586534,
-      "grad_norm": 3.017653703689575,
-      "learning_rate": 2.07923385750033e-05,
-      "loss": 1.4409,
+      "grad_norm": 2.8814990520477295,
+      "learning_rate": 2.098449876243096e-05,
+      "loss": 1.453,
       "step": 797
     },
     {
       "epoch": 0.35525876460767947,
-      "grad_norm": 2.619394540786743,
-      "learning_rate": 2.0600960135216462e-05,
-      "loss": 1.7445,
+      "grad_norm": 2.3015761375427246,
+      "learning_rate": 2.07923385750033e-05,
+      "loss": 1.7499,
       "step": 798
     },
     {
       "epoch": 0.3557039510294936,
-      "grad_norm": 3.4079883098602295,
-      "learning_rate": 2.0410365331898416e-05,
-      "loss": 2.4086,
+      "grad_norm": 2.598961353302002,
+      "learning_rate": 2.0600960135216462e-05,
+      "loss": 2.2751,
       "step": 799
     },
     {
       "epoch": 0.3561491374513077,
-      "grad_norm": 1.3419363498687744,
-      "learning_rate": 2.0220556046142893e-05,
-      "loss": 0.9088,
+      "grad_norm": 1.2370599508285522,
+      "learning_rate": 2.0410365331898416e-05,
+      "loss": 0.9129,
       "step": 800
     },
     {
       "epoch": 0.35659432387312184,
-      "grad_norm": 3.1316444873809814,
-      "learning_rate": 2.0031534151290943e-05,
-      "loss": 1.4489,
+      "grad_norm": 3.447509527206421,
+      "learning_rate": 2.0220556046142893e-05,
+      "loss": 1.4874,
       "step": 801
     },
     {
       "epoch": 0.357039510294936,
-      "grad_norm": 2.7865233421325684,
-      "learning_rate": 1.9843301512912327e-05,
-      "loss": 1.5738,
+      "grad_norm": 3.167005777359009,
+      "learning_rate": 2.0031534151290943e-05,
+      "loss": 1.5479,
       "step": 802
     },
     {
       "epoch": 0.35748469671675015,
-      "grad_norm": 2.195303201675415,
-      "learning_rate": 1.965585998878724e-05,
-      "loss": 1.1416,
+      "grad_norm": 2.0752930641174316,
+      "learning_rate": 1.9843301512912327e-05,
+      "loss": 1.1736,
       "step": 803
     },
     {
       "epoch": 0.3579298831385643,
-      "grad_norm": 1.7528187036514282,
-      "learning_rate": 1.946921142888781e-05,
-      "loss": 1.3798,
+      "grad_norm": 1.605804204940796,
+      "learning_rate": 1.965585998878724e-05,
+      "loss": 1.336,
       "step": 804
     },
     {
       "epoch": 0.3583750695603784,
-      "grad_norm": 1.894110083580017,
-      "learning_rate": 1.928335767535997e-05,
-      "loss": 1.6355,
+      "grad_norm": 1.9330936670303345,
+      "learning_rate": 1.946921142888781e-05,
+      "loss": 1.6619,
       "step": 805
     },
     {
       "epoch": 0.3588202559821925,
-      "grad_norm": 3.0293827056884766,
-      "learning_rate": 1.9098300562505266e-05,
-      "loss": 2.377,
+      "grad_norm": 2.9422507286071777,
+      "learning_rate": 1.928335767535997e-05,
+      "loss": 2.3493,
       "step": 806
     },
     {
       "epoch": 0.35926544240400665,
-      "grad_norm": 3.3845605850219727,
-      "learning_rate": 1.891404191676265e-05,
-      "loss": 1.7129,
+      "grad_norm": 4.052491664886475,
+      "learning_rate": 1.9098300562505266e-05,
+      "loss": 1.7118,
       "step": 807
     },
     {
       "epoch": 0.35971062882582083,
-      "grad_norm": 2.3603949546813965,
-      "learning_rate": 1.8730583556690605e-05,
-      "loss": 1.7646,
+      "grad_norm": 2.3159127235412598,
+      "learning_rate": 1.891404191676265e-05,
+      "loss": 1.7483,
       "step": 808
     },
     {
       "epoch": 0.36015581524763496,
-      "grad_norm": 2.9297170639038086,
-      "learning_rate": 1.854792729294905e-05,
-      "loss": 2.088,
+      "grad_norm": 2.81177020072937,
+      "learning_rate": 1.8730583556690605e-05,
+      "loss": 2.1032,
       "step": 809
     },
     {
       "epoch": 0.3606010016694491,
-      "grad_norm": 2.4962754249572754,
-      "learning_rate": 1.8366074928281607e-05,
-      "loss": 1.7159,
+      "grad_norm": 2.669597625732422,
+      "learning_rate": 1.854792729294905e-05,
+      "loss": 1.7215,
       "step": 810
     },
     {
       "epoch": 0.3610461880912632,
-      "grad_norm": 2.170757532119751,
-      "learning_rate": 1.818502825749764e-05,
-      "loss": 1.5805,
+      "grad_norm": 1.9452595710754395,
+      "learning_rate": 1.8366074928281607e-05,
+      "loss": 1.5899,
       "step": 811
     },
     {
       "epoch": 0.36149137451307733,
-      "grad_norm": 3.127051830291748,
-      "learning_rate": 1.8004789067454764e-05,
-      "loss": 2.4222,
+      "grad_norm": 3.1938717365264893,
+      "learning_rate": 1.818502825749764e-05,
+      "loss": 2.4713,
       "step": 812
     },
     {
       "epoch": 0.36193656093489146,
-      "grad_norm": 2.367252826690674,
-      "learning_rate": 1.7825359137040988e-05,
-      "loss": 2.109,
+      "grad_norm": 2.9424474239349365,
+      "learning_rate": 1.8004789067454764e-05,
+      "loss": 2.0817,
       "step": 813
     },
     {
       "epoch": 0.36238174735670564,
-      "grad_norm": 1.9050239324569702,
-      "learning_rate": 1.7646740237157256e-05,
-      "loss": 1.2665,
+      "grad_norm": 1.8135279417037964,
+      "learning_rate": 1.7825359137040988e-05,
+      "loss": 1.2497,
       "step": 814
     },
     {
       "epoch": 0.36282693377851977,
-      "grad_norm": 2.2031283378601074,
-      "learning_rate": 1.7468934130700044e-05,
-      "loss": 1.2599,
+      "grad_norm": 2.171995162963867,
+      "learning_rate": 1.7646740237157256e-05,
+      "loss": 1.2897,
       "step": 815
     },
     {
       "epoch": 0.3632721202003339,
-      "grad_norm": 3.291349172592163,
-      "learning_rate": 1.7291942572543807e-05,
-      "loss": 2.1776,
+      "grad_norm": 3.4379913806915283,
+      "learning_rate": 1.7468934130700044e-05,
+      "loss": 2.2097,
       "step": 816
     },
     {
       "epoch": 0.363717306622148,
-      "grad_norm": 4.794470310211182,
-      "learning_rate": 1.7115767309523812e-05,
-      "loss": 2.1294,
+      "grad_norm": 6.185148239135742,
+      "learning_rate": 1.7291942572543807e-05,
+      "loss": 2.1192,
       "step": 817
     },
     {
       "epoch": 0.36416249304396214,
-      "grad_norm": 1.8986549377441406,
-      "learning_rate": 1.6940410080418723e-05,
-      "loss": 1.5655,
+      "grad_norm": 2.1236114501953125,
+      "learning_rate": 1.7115767309523812e-05,
+      "loss": 1.6035,
       "step": 818
     },
     {
       "epoch": 0.36460767946577627,
-      "grad_norm": 3.300102472305298,
-      "learning_rate": 1.6765872615933677e-05,
-      "loss": 2.4501,
+      "grad_norm": 3.2910990715026855,
+      "learning_rate": 1.6940410080418723e-05,
+      "loss": 2.4617,
       "step": 819
     },
     {
       "epoch": 0.36505286588759045,
-      "grad_norm": 2.3774209022521973,
-      "learning_rate": 1.6592156638682886e-05,
-      "loss": 2.0107,
+      "grad_norm": 2.3775298595428467,
+      "learning_rate": 1.6765872615933677e-05,
+      "loss": 1.9845,
       "step": 820
     },
     {
       "epoch": 0.3654980523094046,
-      "grad_norm": 1.869045615196228,
-      "learning_rate": 1.6419263863172997e-05,
-      "loss": 1.6384,
+      "grad_norm": 1.7690722942352295,
+      "learning_rate": 1.6592156638682886e-05,
+      "loss": 1.5974,
       "step": 821
     },
     {
       "epoch": 0.3659432387312187,
-      "grad_norm": 2.376424551010132,
-      "learning_rate": 1.6247195995785837e-05,
-      "loss": 1.6731,
+      "grad_norm": 2.4220054149627686,
+      "learning_rate": 1.6419263863172997e-05,
+      "loss": 1.6484,
       "step": 822
     },
     {
       "epoch": 0.3663884251530328,
-      "grad_norm": 2.211214303970337,
-      "learning_rate": 1.6075954734761845e-05,
-      "loss": 1.6778,
+      "grad_norm": 2.0820889472961426,
+      "learning_rate": 1.6247195995785837e-05,
+      "loss": 1.675,
       "step": 823
     },
     {
       "epoch": 0.36683361157484695,
-      "grad_norm": 2.649627685546875,
-      "learning_rate": 1.5905541770183096e-05,
-      "loss": 2.41,
+      "grad_norm": 2.522348642349243,
+      "learning_rate": 1.6075954734761845e-05,
+      "loss": 2.4212,
       "step": 824
     },
     {
       "epoch": 0.3672787979966611,
-      "grad_norm": 1.5385890007019043,
-      "learning_rate": 1.5735958783956794e-05,
-      "loss": 1.1177,
+      "grad_norm": 1.4906548261642456,
+      "learning_rate": 1.5905541770183096e-05,
+      "loss": 1.0927,
       "step": 825
     },
     {
       "epoch": 0.36772398441847526,
-      "grad_norm": 2.307979106903076,
-      "learning_rate": 1.5567207449798515e-05,
-      "loss": 1.6581,
+      "grad_norm": 2.353506565093994,
+      "learning_rate": 1.5735958783956794e-05,
+      "loss": 1.6243,
       "step": 826
     },
     {
       "epoch": 0.3681691708402894,
-      "grad_norm": 2.140033006668091,
-      "learning_rate": 1.539928943321579e-05,
-      "loss": 1.9136,
+      "grad_norm": 2.1150946617126465,
+      "learning_rate": 1.5567207449798515e-05,
+      "loss": 1.8702,
       "step": 827
     },
     {
       "epoch": 0.3686143572621035,
-      "grad_norm": 3.384453773498535,
-      "learning_rate": 1.5232206391491699e-05,
-      "loss": 2.8381,
+      "grad_norm": 3.695554256439209,
+      "learning_rate": 1.539928943321579e-05,
+      "loss": 2.8629,
       "step": 828
     },
     {
       "epoch": 0.36905954368391763,
-      "grad_norm": 2.655250072479248,
-      "learning_rate": 1.5065959973668353e-05,
-      "loss": 1.5791,
+      "grad_norm": 2.1603634357452393,
+      "learning_rate": 1.5232206391491699e-05,
+      "loss": 1.5569,
       "step": 829
     },
     {
       "epoch": 0.36950473010573176,
-      "grad_norm": 1.3919587135314941,
-      "learning_rate": 1.4900551820530828e-05,
-      "loss": 1.1324,
+      "grad_norm": 1.2208006381988525,
+      "learning_rate": 1.5065959973668353e-05,
+      "loss": 1.1422,
       "step": 830
     },
     {
       "epoch": 0.3699499165275459,
-      "grad_norm": 4.468016147613525,
-      "learning_rate": 1.4735983564590783e-05,
-      "loss": 2.4508,
+      "grad_norm": 4.04674768447876,
+      "learning_rate": 1.4900551820530828e-05,
+      "loss": 2.4647,
       "step": 831
     },
     {
       "epoch": 0.37039510294936007,
-      "grad_norm": 2.089130163192749,
-      "learning_rate": 1.4572256830070497e-05,
-      "loss": 1.5146,
+      "grad_norm": 2.1937062740325928,
+      "learning_rate": 1.4735983564590783e-05,
+      "loss": 1.5351,
       "step": 832
     },
     {
       "epoch": 0.3708402893711742,
-      "grad_norm": 1.788549780845642,
-      "learning_rate": 1.4409373232886702e-05,
-      "loss": 1.2323,
+      "grad_norm": 1.7129528522491455,
+      "learning_rate": 1.4572256830070497e-05,
+      "loss": 1.1443,
       "step": 833
     },
     {
       "epoch": 0.3712854757929883,
-      "grad_norm": 3.411046266555786,
-      "learning_rate": 1.4247334380634792e-05,
-      "loss": 2.3234,
+      "grad_norm": 3.113435745239258,
+      "learning_rate": 1.4409373232886702e-05,
+      "loss": 2.2822,
       "step": 834
     },
     {
       "epoch": 0.37173066221480244,
-      "grad_norm": 3.8933844566345215,
-      "learning_rate": 1.4086141872572789e-05,
-      "loss": 1.9897,
+      "grad_norm": 3.254114866256714,
+      "learning_rate": 1.4247334380634792e-05,
+      "loss": 1.9679,
       "step": 835
     },
     {
       "epoch": 0.37217584863661657,
-      "grad_norm": 2.343925952911377,
-      "learning_rate": 1.3925797299605647e-05,
-      "loss": 1.4688,
+      "grad_norm": 2.4401051998138428,
+      "learning_rate": 1.4086141872572789e-05,
+      "loss": 1.435,
       "step": 836
     },
     {
       "epoch": 0.3726210350584307,
-      "grad_norm": 3.1801838874816895,
-      "learning_rate": 1.3766302244269624e-05,
-      "loss": 2.4004,
+      "grad_norm": 3.798807382583618,
+      "learning_rate": 1.3925797299605647e-05,
+      "loss": 2.3262,
       "step": 837
     },
     {
       "epoch": 0.3730662214802449,
-      "grad_norm": 1.550551414489746,
-      "learning_rate": 1.3607658280716473e-05,
-      "loss": 1.0341,
+      "grad_norm": 1.5593875646591187,
+      "learning_rate": 1.3766302244269624e-05,
+      "loss": 1.0388,
       "step": 838
     },
     {
       "epoch": 0.373511407902059,
-      "grad_norm": 2.117818593978882,
-      "learning_rate": 1.3449866974698122e-05,
-      "loss": 1.7987,
+      "grad_norm": 2.430449962615967,
+      "learning_rate": 1.3607658280716473e-05,
+      "loss": 1.7574,
       "step": 839
     },
     {
       "epoch": 0.3739565943238731,
-      "grad_norm": 4.381692409515381,
-      "learning_rate": 1.3292929883550998e-05,
-      "loss": 2.4703,
+      "grad_norm": 4.148708343505859,
+      "learning_rate": 1.3449866974698122e-05,
+      "loss": 2.53,
       "step": 840
     },
     {
       "epoch": 0.37440178074568725,
-      "grad_norm": 2.0128989219665527,
-      "learning_rate": 1.3136848556180892e-05,
-      "loss": 1.8252,
+      "grad_norm": 2.0616252422332764,
+      "learning_rate": 1.3292929883550998e-05,
+      "loss": 1.8342,
       "step": 841
     },
     {
       "epoch": 0.3748469671675014,
-      "grad_norm": 1.8894273042678833,
-      "learning_rate": 1.2981624533047432e-05,
-      "loss": 1.7092,
+      "grad_norm": 1.8223658800125122,
+      "learning_rate": 1.3136848556180892e-05,
+      "loss": 1.6677,
       "step": 842
     },
     {
       "epoch": 0.3752921535893155,
-      "grad_norm": 3.7159740924835205,
-      "learning_rate": 1.2827259346149122e-05,
-      "loss": 2.0822,
+      "grad_norm": 3.7050740718841553,
+      "learning_rate": 1.2981624533047432e-05,
+      "loss": 2.0134,
       "step": 843
     },
     {
       "epoch": 0.3757373400111297,
-      "grad_norm": 3.0898549556732178,
-      "learning_rate": 1.2673754519008008e-05,
-      "loss": 2.6043,
+      "grad_norm": 3.2520573139190674,
+      "learning_rate": 1.2827259346149122e-05,
+      "loss": 2.7152,
       "step": 844
     },
     {
       "epoch": 0.3761825264329438,
-      "grad_norm": 2.4676640033721924,
-      "learning_rate": 1.2521111566654731e-05,
-      "loss": 2.4497,
+      "grad_norm": 3.439446449279785,
+      "learning_rate": 1.2673754519008008e-05,
+      "loss": 2.5235,
       "step": 845
     },
     {
       "epoch": 0.37662771285475793,
-      "grad_norm": 3.5587337017059326,
-      "learning_rate": 1.2369331995613665e-05,
-      "loss": 1.6656,
+      "grad_norm": 3.737055778503418,
+      "learning_rate": 1.2521111566654731e-05,
+      "loss": 1.692,
       "step": 846
     },
     {
       "epoch": 0.37707289927657206,
-      "grad_norm": 7.141547679901123,
-      "learning_rate": 1.2218417303887842e-05,
-      "loss": 1.7547,
+      "grad_norm": 5.928926944732666,
+      "learning_rate": 1.2369331995613665e-05,
+      "loss": 1.8147,
       "step": 847
     },
     {
       "epoch": 0.3775180856983862,
-      "grad_norm": 1.7326031923294067,
-      "learning_rate": 1.206836898094439e-05,
-      "loss": 1.4759,
+      "grad_norm": 1.5324987173080444,
+      "learning_rate": 1.2218417303887842e-05,
+      "loss": 1.4275,
       "step": 848
     },
     {
       "epoch": 0.3779632721202003,
-      "grad_norm": 3.1554577350616455,
-      "learning_rate": 1.191918850769964e-05,
-      "loss": 2.4164,
+      "grad_norm": 3.3122949600219727,
+      "learning_rate": 1.206836898094439e-05,
+      "loss": 2.386,
       "step": 849
     },
     {
       "epoch": 0.3784084585420145,
-      "grad_norm": 1.9731732606887817,
-      "learning_rate": 1.1770877356504683e-05,
-      "loss": 1.0614,
+      "grad_norm": 1.8216137886047363,
+      "learning_rate": 1.191918850769964e-05,
+      "loss": 1.0651,
       "step": 850
     },
     {
       "epoch": 0.3788536449638286,
-      "grad_norm": 2.997244119644165,
-      "learning_rate": 1.1623436991130654e-05,
-      "loss": 2.0257,
+      "grad_norm": 2.842923641204834,
+      "learning_rate": 1.1770877356504683e-05,
+      "loss": 1.9827,
       "step": 851
     },
     {
       "epoch": 0.37929883138564274,
-      "grad_norm": 5.286291599273682,
-      "learning_rate": 1.1476868866754486e-05,
-      "loss": 2.0731,
+      "grad_norm": 4.741792678833008,
+      "learning_rate": 1.1623436991130654e-05,
+      "loss": 1.9713,
       "step": 852
     },
     {
       "epoch": 0.37974401780745687,
-      "grad_norm": 2.4582889080047607,
-      "learning_rate": 1.1331174429944347e-05,
-      "loss": 1.406,
+      "grad_norm": 2.5982000827789307,
+      "learning_rate": 1.1476868866754486e-05,
+      "loss": 1.3794,
       "step": 853
     },
     {
       "epoch": 0.380189204229271,
-      "grad_norm": 2.530067205429077,
-      "learning_rate": 1.1186355118645554e-05,
-      "loss": 1.6258,
+      "grad_norm": 2.446061611175537,
+      "learning_rate": 1.1331174429944347e-05,
+      "loss": 1.6679,
       "step": 854
     },
     {
       "epoch": 0.3806343906510851,
-      "grad_norm": 3.028555393218994,
-      "learning_rate": 1.1042412362166222e-05,
-      "loss": 2.8568,
+      "grad_norm": 3.426534414291382,
+      "learning_rate": 1.1186355118645554e-05,
+      "loss": 2.7746,
       "step": 855
     },
     {
       "epoch": 0.3810795770728993,
-      "grad_norm": 2.7910170555114746,
-      "learning_rate": 1.0899347581163221e-05,
-      "loss": 1.8254,
+      "grad_norm": 3.0752968788146973,
+      "learning_rate": 1.1042412362166222e-05,
+      "loss": 1.8427,
       "step": 856
     },
     {
       "epoch": 0.3815247634947134,
-      "grad_norm": 39.85447692871094,
-      "learning_rate": 1.0757162187628222e-05,
-      "loss": 2.8449,
+      "grad_norm": 7.70589542388916,
+      "learning_rate": 1.0899347581163221e-05,
+      "loss": 2.7767,
       "step": 857
     },
     {
       "epoch": 0.38196994991652755,
-      "grad_norm": 3.291005849838257,
-      "learning_rate": 1.0615857584873623e-05,
-      "loss": 2.0754,
+      "grad_norm": 2.923967123031616,
+      "learning_rate": 1.0757162187628222e-05,
+      "loss": 2.0666,
       "step": 858
     },
     {
       "epoch": 0.3824151363383417,
-      "grad_norm": 4.250852584838867,
-      "learning_rate": 1.0475435167518843e-05,
-      "loss": 2.1429,
+      "grad_norm": 4.278809547424316,
+      "learning_rate": 1.0615857584873623e-05,
+      "loss": 2.1183,
       "step": 859
     },
     {
       "epoch": 0.3828603227601558,
-      "grad_norm": 2.714912176132202,
-      "learning_rate": 1.0335896321476413e-05,
-      "loss": 1.7616,
+      "grad_norm": 2.635345935821533,
+      "learning_rate": 1.0475435167518843e-05,
+      "loss": 1.6819,
       "step": 860
     },
     {
       "epoch": 0.38330550918196993,
-      "grad_norm": 2.5076630115509033,
-      "learning_rate": 1.0197242423938446e-05,
-      "loss": 1.864,
+      "grad_norm": 2.319185733795166,
+      "learning_rate": 1.0335896321476413e-05,
+      "loss": 1.7721,
       "step": 861
     },
     {
       "epoch": 0.3837506956037841,
-      "grad_norm": 3.874208450317383,
-      "learning_rate": 1.0059474843362892e-05,
-      "loss": 3.0877,
+      "grad_norm": 3.1770641803741455,
+      "learning_rate": 1.0197242423938446e-05,
+      "loss": 3.0742,
       "step": 862
     },
     {
       "epoch": 0.38419588202559823,
-      "grad_norm": 3.2415525913238525,
-      "learning_rate": 9.922594939460194e-06,
-      "loss": 2.3881,
+      "grad_norm": 3.24739933013916,
+      "learning_rate": 1.0059474843362892e-05,
+      "loss": 2.3675,
       "step": 863
     },
     {
       "epoch": 0.38464106844741236,
-      "grad_norm": 2.6872646808624268,
-      "learning_rate": 9.786604063179728e-06,
-      "loss": 1.953,
+      "grad_norm": 2.699204683303833,
+      "learning_rate": 9.922594939460194e-06,
+      "loss": 1.9462,
       "step": 864
     },
     {
       "epoch": 0.3850862548692265,
-      "grad_norm": 2.7855000495910645,
-      "learning_rate": 9.651503556696516e-06,
-      "loss": 2.1969,
+      "grad_norm": 2.9695088863372803,
+      "learning_rate": 9.786604063179728e-06,
+      "loss": 2.204,
       "step": 865
     },
     {
       "epoch": 0.3855314412910406,
-      "grad_norm": 3.2661259174346924,
-      "learning_rate": 9.517294753398064e-06,
-      "loss": 2.2943,
+      "grad_norm": 4.052135467529297,
+      "learning_rate": 9.651503556696516e-06,
+      "loss": 2.354,
       "step": 866
     },
     {
       "epoch": 0.38597662771285474,
-      "grad_norm": 2.2386388778686523,
-      "learning_rate": 9.383978977871021e-06,
-      "loss": 1.3356,
+      "grad_norm": 2.3410732746124268,
+      "learning_rate": 9.517294753398064e-06,
+      "loss": 1.3547,
       "step": 867
     },
     {
       "epoch": 0.3864218141346689,
-      "grad_norm": 2.699187755584717,
-      "learning_rate": 9.251557545888312e-06,
-      "loss": 2.4968,
+      "grad_norm": 2.827927827835083,
+      "learning_rate": 9.383978977871021e-06,
+      "loss": 2.5106,
       "step": 868
     },
     {
       "epoch": 0.38686700055648304,
-      "grad_norm": 2.743082046508789,
-      "learning_rate": 9.120031764395987e-06,
-      "loss": 4.0225,
+      "grad_norm": 2.7326576709747314,
+      "learning_rate": 9.251557545888312e-06,
+      "loss": 4.0232,
       "step": 869
     },
     {
       "epoch": 0.38731218697829717,
-      "grad_norm": 2.0932228565216064,
-      "learning_rate": 8.989402931500434e-06,
-      "loss": 1.4646,
+      "grad_norm": 2.105860710144043,
+      "learning_rate": 9.120031764395987e-06,
+      "loss": 1.4466,
       "step": 870
     },
     {
       "epoch": 0.3877573734001113,
-      "grad_norm": 2.3609516620635986,
-      "learning_rate": 8.85967233645547e-06,
-      "loss": 2.0499,
+      "grad_norm": 2.9740326404571533,
+      "learning_rate": 8.989402931500434e-06,
+      "loss": 2.0867,
       "step": 871
     },
     {
       "epoch": 0.3882025598219254,
-      "grad_norm": 8.27224063873291,
-      "learning_rate": 8.730841259649725e-06,
-      "loss": 2.7773,
+      "grad_norm": 8.592368125915527,
+      "learning_rate": 8.85967233645547e-06,
+      "loss": 2.7982,
       "step": 872
     },
     {
       "epoch": 0.38864774624373954,
-      "grad_norm": 2.892641544342041,
-      "learning_rate": 8.602910972593892e-06,
-      "loss": 2.2208,
+      "grad_norm": 2.596602201461792,
+      "learning_rate": 8.730841259649725e-06,
+      "loss": 2.2047,
       "step": 873
     },
     {
       "epoch": 0.3890929326655537,
-      "grad_norm": 2.17576265335083,
-      "learning_rate": 8.475882737908248e-06,
-      "loss": 1.9523,
+      "grad_norm": 2.2201178073883057,
+      "learning_rate": 8.602910972593892e-06,
+      "loss": 1.9367,
       "step": 874
     },
     {
       "epoch": 0.38953811908736785,
-      "grad_norm": 3.0030341148376465,
-      "learning_rate": 8.34975780931021e-06,
-      "loss": 2.3068,
+      "grad_norm": 3.1746363639831543,
+      "learning_rate": 8.475882737908248e-06,
+      "loss": 2.2672,
       "step": 875
     },
     {
       "epoch": 0.389983305509182,
-      "grad_norm": 3.5457253456115723,
-      "learning_rate": 8.224537431601886e-06,
-      "loss": 1.9086,
+      "grad_norm": 3.8068130016326904,
+      "learning_rate": 8.34975780931021e-06,
+      "loss": 1.9147,
       "step": 876
     },
     {
       "epoch": 0.3904284919309961,
-      "grad_norm": 2.699582815170288,
-      "learning_rate": 8.100222840657878e-06,
-      "loss": 2.1184,
+      "grad_norm": 2.8877782821655273,
+      "learning_rate": 8.224537431601886e-06,
+      "loss": 2.1708,
       "step": 877
     },
     {
       "epoch": 0.3908736783528102,
-      "grad_norm": 2.5596392154693604,
-      "learning_rate": 7.976815263412963e-06,
-      "loss": 1.7771,
+      "grad_norm": 2.659292221069336,
+      "learning_rate": 8.100222840657878e-06,
+      "loss": 1.7599,
       "step": 878
     },
     {
       "epoch": 0.39131886477462435,
-      "grad_norm": 2.6684186458587646,
-      "learning_rate": 7.854315917850163e-06,
-      "loss": 1.7569,
+      "grad_norm": 2.6380879878997803,
+      "learning_rate": 7.976815263412963e-06,
+      "loss": 1.6982,
       "step": 879
     },
     {
       "epoch": 0.39176405119643853,
-      "grad_norm": 2.5746593475341797,
-      "learning_rate": 7.73272601298851e-06,
-      "loss": 2.0132,
+      "grad_norm": 2.541360378265381,
+      "learning_rate": 7.854315917850163e-06,
+      "loss": 2.0257,
       "step": 880
     },
     {
       "epoch": 0.39220923761825266,
-      "grad_norm": 2.5082550048828125,
-      "learning_rate": 7.612046748871327e-06,
-      "loss": 1.7144,
+      "grad_norm": 2.434187650680542,
+      "learning_rate": 7.73272601298851e-06,
+      "loss": 1.7154,
       "step": 881
     },
     {
       "epoch": 0.3926544240400668,
-      "grad_norm": 1.8464329242706299,
-      "learning_rate": 7.492279316554207e-06,
-      "loss": 1.1433,
+      "grad_norm": 1.727720856666565,
+      "learning_rate": 7.612046748871327e-06,
+      "loss": 1.1091,
       "step": 882
     },
     {
       "epoch": 0.3930996104618809,
-      "grad_norm": 3.3372461795806885,
-      "learning_rate": 7.3734248980933395e-06,
-      "loss": 2.8916,
+      "grad_norm": 3.544562578201294,
+      "learning_rate": 7.492279316554207e-06,
+      "loss": 2.8993,
       "step": 883
     },
     {
       "epoch": 0.39354479688369504,
-      "grad_norm": 3.623055934906006,
-      "learning_rate": 7.255484666533874e-06,
-      "loss": 1.7107,
+      "grad_norm": 2.7014670372009277,
+      "learning_rate": 7.3734248980933395e-06,
+      "loss": 1.7446,
       "step": 884
     },
     {
       "epoch": 0.39398998330550916,
-      "grad_norm": 2.4824955463409424,
-      "learning_rate": 7.138459785898266e-06,
-      "loss": 1.79,
+      "grad_norm": 2.4355177879333496,
+      "learning_rate": 7.255484666533874e-06,
+      "loss": 1.7447,
       "step": 885
     },
     {
       "epoch": 0.39443516972732334,
-      "grad_norm": 2.6194679737091064,
-      "learning_rate": 7.022351411174866e-06,
-      "loss": 1.6792,
+      "grad_norm": 2.3996999263763428,
+      "learning_rate": 7.138459785898266e-06,
+      "loss": 1.7492,
       "step": 886
     },
     {
       "epoch": 0.39488035614913747,
-      "grad_norm": 2.1351640224456787,
-      "learning_rate": 6.907160688306425e-06,
-      "loss": 1.992,
+      "grad_norm": 2.208282232284546,
+      "learning_rate": 7.022351411174866e-06,
+      "loss": 1.9822,
       "step": 887
     },
     {
       "epoch": 0.3953255425709516,
-      "grad_norm": 3.6620707511901855,
-      "learning_rate": 6.7928887541789055e-06,
-      "loss": 1.9295,
+      "grad_norm": 2.583613395690918,
+      "learning_rate": 6.907160688306425e-06,
+      "loss": 1.9487,
       "step": 888
     },
     {
       "epoch": 0.3957707289927657,
-      "grad_norm": 2.735880136489868,
-      "learning_rate": 6.679536736610137e-06,
-      "loss": 2.5191,
+      "grad_norm": 2.559744358062744,
+      "learning_rate": 6.7928887541789055e-06,
+      "loss": 2.5036,
       "step": 889
     },
     {
       "epoch": 0.39621591541457984,
-      "grad_norm": 2.25370192527771,
-      "learning_rate": 6.5671057543387985e-06,
-      "loss": 1.7509,
+      "grad_norm": 2.1296982765197754,
+      "learning_rate": 6.679536736610137e-06,
+      "loss": 1.7707,
       "step": 890
     },
     {
       "epoch": 0.39666110183639397,
-      "grad_norm": 3.4820597171783447,
-      "learning_rate": 6.455596917013273e-06,
-      "loss": 2.0117,
+      "grad_norm": 2.857872724533081,
+      "learning_rate": 6.5671057543387985e-06,
+      "loss": 2.0831,
       "step": 891
     },
     {
       "epoch": 0.39710628825820815,
-      "grad_norm": 2.8001556396484375,
-      "learning_rate": 6.345011325180772e-06,
-      "loss": 1.6458,
+      "grad_norm": 2.2674055099487305,
+      "learning_rate": 6.455596917013273e-06,
+      "loss": 1.7374,
       "step": 892
     },
     {
       "epoch": 0.3975514746800223,
-      "grad_norm": 1.6623258590698242,
-      "learning_rate": 6.235350070276447e-06,
-      "loss": 1.6371,
+      "grad_norm": 1.8583672046661377,
+      "learning_rate": 6.345011325180772e-06,
+      "loss": 1.6326,
       "step": 893
     },
     {
       "epoch": 0.3979966611018364,
-      "grad_norm": 2.0735392570495605,
-      "learning_rate": 6.126614234612593e-06,
-      "loss": 1.8299,
+      "grad_norm": 2.2417795658111572,
+      "learning_rate": 6.235350070276447e-06,
+      "loss": 1.8506,
       "step": 894
     },
     {
       "epoch": 0.3984418475236505,
-      "grad_norm": 2.833833694458008,
-      "learning_rate": 6.018804891368035e-06,
-      "loss": 1.6262,
+      "grad_norm": 2.7897722721099854,
+      "learning_rate": 6.126614234612593e-06,
+      "loss": 1.5891,
       "step": 895
     },
     {
       "epoch": 0.39888703394546465,
-      "grad_norm": 1.4123960733413696,
-      "learning_rate": 5.911923104577455e-06,
-      "loss": 1.2736,
+      "grad_norm": 1.4530127048492432,
+      "learning_rate": 6.018804891368035e-06,
+      "loss": 1.2882,
       "step": 896
     },
     {
       "epoch": 0.3993322203672788,
-      "grad_norm": 4.113855838775635,
-      "learning_rate": 5.805969929120947e-06,
-      "loss": 2.449,
+      "grad_norm": 4.077309608459473,
+      "learning_rate": 5.911923104577455e-06,
+      "loss": 2.4773,
       "step": 897
     },
     {
       "epoch": 0.39977740678909296,
-      "grad_norm": 4.152754306793213,
-      "learning_rate": 5.700946410713548e-06,
-      "loss": 1.9736,
+      "grad_norm": 2.6346118450164795,
+      "learning_rate": 5.805969929120947e-06,
+      "loss": 1.9211,
       "step": 898
     },
     {
       "epoch": 0.4002225932109071,
-      "grad_norm": 2.434448480606079,
-      "learning_rate": 5.5968535858950345e-06,
-      "loss": 2.2586,
+      "grad_norm": 2.4013400077819824,
+      "learning_rate": 5.700946410713548e-06,
+      "loss": 2.2058,
       "step": 899
     },
     {
       "epoch": 0.4006677796327212,
-      "grad_norm": 2.730743169784546,
-      "learning_rate": 5.49369248201953e-06,
-      "loss": 2.1945,
+      "grad_norm": 2.5028560161590576,
+      "learning_rate": 5.5968535858950345e-06,
+      "loss": 2.1112,
       "step": 900
     },
     {
       "epoch": 0.40111296605453534,
-      "grad_norm": 2.3305442333221436,
-      "learning_rate": 5.39146411724547e-06,
-      "loss": 1.844,
+      "grad_norm": 2.262838363647461,
+      "learning_rate": 5.49369248201953e-06,
+      "loss": 1.792,
       "step": 901
     },
     {
       "epoch": 0.40155815247634946,
-      "grad_norm": 2.125002861022949,
-      "learning_rate": 5.290169500525577e-06,
-      "loss": 1.8199,
+      "grad_norm": 3.0734193325042725,
+      "learning_rate": 5.39146411724547e-06,
+      "loss": 1.8429,
       "step": 902
     },
     {
       "epoch": 0.4020033388981636,
-      "grad_norm": 2.352309226989746,
-      "learning_rate": 5.189809631596798e-06,
-      "loss": 2.143,
+      "grad_norm": 2.408612012863159,
+      "learning_rate": 5.290169500525577e-06,
+      "loss": 2.1481,
       "step": 903
     },
     {
       "epoch": 0.40244852531997777,
-      "grad_norm": 2.7642769813537598,
-      "learning_rate": 5.0903855009705514e-06,
-      "loss": 1.75,
+      "grad_norm": 2.9177372455596924,
+      "learning_rate": 5.189809631596798e-06,
+      "loss": 1.653,
       "step": 904
     },
     {
       "epoch": 0.4028937117417919,
-      "grad_norm": 4.074885845184326,
-      "learning_rate": 4.991898089922819e-06,
-      "loss": 2.1392,
+      "grad_norm": 3.3481597900390625,
+      "learning_rate": 5.0903855009705514e-06,
+      "loss": 2.0301,
       "step": 905
     },
     {
       "epoch": 0.403338898163606,
-      "grad_norm": 2.652728319168091,
-      "learning_rate": 4.8943483704846475e-06,
-      "loss": 1.9152,
+      "grad_norm": 3.205288887023926,
+      "learning_rate": 4.991898089922819e-06,
+      "loss": 1.9313,
       "step": 906
     },
     {
       "epoch": 0.40378408458542014,
-      "grad_norm": 2.720862627029419,
-      "learning_rate": 4.797737305432337e-06,
-      "loss": 2.3358,
+      "grad_norm": 4.120809078216553,
+      "learning_rate": 4.8943483704846475e-06,
+      "loss": 2.4249,
       "step": 907
     },
     {
       "epoch": 0.40422927100723427,
-      "grad_norm": 2.838879346847534,
-      "learning_rate": 4.702065848278126e-06,
-      "loss": 2.1309,
+      "grad_norm": 2.6746208667755127,
+      "learning_rate": 4.797737305432337e-06,
+      "loss": 2.1356,
       "step": 908
     },
     {
       "epoch": 0.4046744574290484,
-      "grad_norm": 2.1942715644836426,
-      "learning_rate": 4.607334943260655e-06,
-      "loss": 1.4136,
+      "grad_norm": 2.167707920074463,
+      "learning_rate": 4.702065848278126e-06,
+      "loss": 1.4489,
       "step": 909
     },
     {
       "epoch": 0.4051196438508626,
-      "grad_norm": 3.1751365661621094,
-      "learning_rate": 4.513545525335705e-06,
-      "loss": 2.0721,
+      "grad_norm": 3.323460578918457,
+      "learning_rate": 4.607334943260655e-06,
+      "loss": 2.0664,
       "step": 910
     },
     {
       "epoch": 0.4055648302726767,
-      "grad_norm": 2.386066436767578,
-      "learning_rate": 4.420698520166988e-06,
-      "loss": 1.7174,
+      "grad_norm": 2.3525609970092773,
+      "learning_rate": 4.513545525335705e-06,
+      "loss": 1.7016,
       "step": 911
     },
     {
       "epoch": 0.4060100166944908,
-      "grad_norm": 2.3867030143737793,
-      "learning_rate": 4.328794844116946e-06,
-      "loss": 1.9815,
+      "grad_norm": 2.3468728065490723,
+      "learning_rate": 4.420698520166988e-06,
+      "loss": 1.9747,
       "step": 912
     },
     {
       "epoch": 0.40645520311630495,
-      "grad_norm": 2.3891232013702393,
-      "learning_rate": 4.237835404237778e-06,
-      "loss": 1.9543,
+      "grad_norm": 3.5238194465637207,
+      "learning_rate": 4.328794844116946e-06,
+      "loss": 2.0,
       "step": 913
     },
     {
       "epoch": 0.4069003895381191,
-      "grad_norm": 2.0092246532440186,
-      "learning_rate": 4.147821098262405e-06,
-      "loss": 1.7221,
+      "grad_norm": 2.4445719718933105,
+      "learning_rate": 4.237835404237778e-06,
+      "loss": 1.7141,
       "step": 914
     },
     {
       "epoch": 0.4073455759599332,
-      "grad_norm": 13.39091968536377,
-      "learning_rate": 4.0587528145957235e-06,
-      "loss": 1.7269,
+      "grad_norm": 6.485393047332764,
+      "learning_rate": 4.147821098262405e-06,
+      "loss": 1.7179,
       "step": 915
     },
     {
       "epoch": 0.40779076238174733,
-      "grad_norm": 2.8730368614196777,
-      "learning_rate": 3.970631432305694e-06,
-      "loss": 1.867,
+      "grad_norm": 3.072964906692505,
+      "learning_rate": 4.0587528145957235e-06,
+      "loss": 1.9091,
       "step": 916
     },
     {
       "epoch": 0.4082359488035615,
-      "grad_norm": 4.520809173583984,
-      "learning_rate": 3.883457821114811e-06,
-      "loss": 2.3426,
+      "grad_norm": 3.583833932876587,
+      "learning_rate": 3.970631432305694e-06,
+      "loss": 2.2719,
       "step": 917
     },
     {
       "epoch": 0.40868113522537564,
-      "grad_norm": 2.5547308921813965,
-      "learning_rate": 3.797232841391407e-06,
-      "loss": 1.6473,
+      "grad_norm": 2.661090850830078,
+      "learning_rate": 3.883457821114811e-06,
+      "loss": 1.6855,
       "step": 918
     },
     {
       "epoch": 0.40912632164718976,
-      "grad_norm": 2.6396074295043945,
-      "learning_rate": 3.711957344141237e-06,
-      "loss": 1.5152,
+      "grad_norm": 2.690599203109741,
+      "learning_rate": 3.797232841391407e-06,
+      "loss": 1.5438,
       "step": 919
     },
     {
       "epoch": 0.4095715080690039,
-      "grad_norm": 2.5423967838287354,
-      "learning_rate": 3.627632170999029e-06,
-      "loss": 1.8634,
+      "grad_norm": 2.358945846557617,
+      "learning_rate": 3.711957344141237e-06,
+      "loss": 1.8729,
       "step": 920
     },
     {
       "epoch": 0.410016694490818,
-      "grad_norm": 2.359440565109253,
-      "learning_rate": 3.5442581542201923e-06,
-      "loss": 1.6601,
+      "grad_norm": 2.3557281494140625,
+      "learning_rate": 3.627632170999029e-06,
+      "loss": 1.6729,
       "step": 921
     },
     {
       "epoch": 0.41046188091263214,
-      "grad_norm": 3.3366010189056396,
-      "learning_rate": 3.461836116672612e-06,
-      "loss": 2.3835,
+      "grad_norm": 3.3151612281799316,
+      "learning_rate": 3.5442581542201923e-06,
+      "loss": 2.3729,
       "step": 922
     },
     {
       "epoch": 0.4109070673344463,
-      "grad_norm": 1.9757188558578491,
-      "learning_rate": 3.380366871828522e-06,
-      "loss": 1.3504,
+      "grad_norm": 2.004413604736328,
+      "learning_rate": 3.461836116672612e-06,
+      "loss": 1.3912,
       "step": 923
     },
     {
       "epoch": 0.41135225375626044,
-      "grad_norm": 1.7242034673690796,
-      "learning_rate": 3.2998512237565005e-06,
-      "loss": 0.9853,
+      "grad_norm": 1.6248087882995605,
+      "learning_rate": 3.380366871828522e-06,
+      "loss": 1.0217,
       "step": 924
     },
     {
       "epoch": 0.41179744017807457,
-      "grad_norm": 4.0235161781311035,
-      "learning_rate": 3.2202899671134546e-06,
-      "loss": 2.5185,
+      "grad_norm": 3.879448652267456,
+      "learning_rate": 3.2998512237565005e-06,
+      "loss": 2.5093,
       "step": 925
     },
     {
       "epoch": 0.4122426265998887,
-      "grad_norm": 3.6245474815368652,
-      "learning_rate": 3.1416838871368924e-06,
-      "loss": 2.2566,
+      "grad_norm": 2.8478055000305176,
+      "learning_rate": 3.2202899671134546e-06,
+      "loss": 2.1615,
       "step": 926
     },
     {
       "epoch": 0.4126878130217028,
-      "grad_norm": 2.94282865524292,
-      "learning_rate": 3.064033759637064e-06,
-      "loss": 1.7241,
+      "grad_norm": 3.20760178565979,
+      "learning_rate": 3.1416838871368924e-06,
+      "loss": 1.8088,
       "step": 927
     },
     {
       "epoch": 0.41313299944351695,
-      "grad_norm": 1.7278319597244263,
-      "learning_rate": 2.9873403509894203e-06,
-      "loss": 1.6847,
+      "grad_norm": 1.640320897102356,
+      "learning_rate": 3.064033759637064e-06,
+      "loss": 1.7017,
       "step": 928
     },
     {
       "epoch": 0.4135781858653311,
-      "grad_norm": 2.720810651779175,
-      "learning_rate": 2.9116044181269007e-06,
-      "loss": 1.7949,
+      "grad_norm": 3.4903106689453125,
+      "learning_rate": 2.9873403509894203e-06,
+      "loss": 1.799,
       "step": 929
     },
     {
       "epoch": 0.41402337228714525,
-      "grad_norm": 2.830357789993286,
-      "learning_rate": 2.836826708532603e-06,
-      "loss": 1.8299,
+      "grad_norm": 2.6346209049224854,
+      "learning_rate": 2.9116044181269007e-06,
+      "loss": 1.8041,
       "step": 930
     },
     {
       "epoch": 0.4144685587089594,
-      "grad_norm": 2.6716318130493164,
-      "learning_rate": 2.7630079602323442e-06,
-      "loss": 1.4774,
+      "grad_norm": 2.9832823276519775,
+      "learning_rate": 2.836826708532603e-06,
+      "loss": 1.5375,
       "step": 931
     },
     {
       "epoch": 0.4149137451307735,
-      "grad_norm": 2.116966724395752,
-      "learning_rate": 2.690148901787337e-06,
-      "loss": 0.5776,
+      "grad_norm": 1.6513136625289917,
+      "learning_rate": 2.7630079602323442e-06,
+      "loss": 0.5627,
       "step": 932
     },
     {
       "epoch": 0.41535893155258763,
-      "grad_norm": 1.9328233003616333,
-      "learning_rate": 2.618250252287113e-06,
-      "loss": 1.4808,
+      "grad_norm": 5.0608015060424805,
+      "learning_rate": 2.690148901787337e-06,
+      "loss": 1.5045,
       "step": 933
     },
     {
       "epoch": 0.41580411797440175,
-      "grad_norm": 2.859921932220459,
-      "learning_rate": 2.5473127213422763e-06,
-      "loss": 2.1348,
+      "grad_norm": 2.82723331451416,
+      "learning_rate": 2.618250252287113e-06,
+      "loss": 2.0713,
       "step": 934
     },
     {
       "epoch": 0.41624930439621594,
-      "grad_norm": 2.039973020553589,
-      "learning_rate": 2.4773370090776626e-06,
-      "loss": 1.8692,
+      "grad_norm": 1.8556444644927979,
+      "learning_rate": 2.5473127213422763e-06,
+      "loss": 1.8779,
       "step": 935
     },
     {
       "epoch": 0.41669449081803006,
-      "grad_norm": 3.830960273742676,
-      "learning_rate": 2.4083238061252567e-06,
-      "loss": 1.9331,
+      "grad_norm": 3.8842921257019043,
+      "learning_rate": 2.4773370090776626e-06,
+      "loss": 1.9951,
       "step": 936
     },
     {
       "epoch": 0.4171396772398442,
-      "grad_norm": 2.213256597518921,
-      "learning_rate": 2.3402737936175425e-06,
-      "loss": 1.8484,
+      "grad_norm": 2.5429670810699463,
+      "learning_rate": 2.4083238061252567e-06,
+      "loss": 1.9011,
       "step": 937
     },
     {
       "epoch": 0.4175848636616583,
-      "grad_norm": 2.631376266479492,
-      "learning_rate": 2.273187643180652e-06,
-      "loss": 2.4657,
+      "grad_norm": 2.3649942874908447,
+      "learning_rate": 2.3402737936175425e-06,
+      "loss": 2.4083,
       "step": 938
     },
     {
       "epoch": 0.41803005008347244,
-      "grad_norm": 2.208151340484619,
-      "learning_rate": 2.2070660169278166e-06,
-      "loss": 1.9784,
+      "grad_norm": 2.2096188068389893,
+      "learning_rate": 2.273187643180652e-06,
+      "loss": 1.9797,
       "step": 939
     },
     {
       "epoch": 0.41847523650528656,
-      "grad_norm": 2.4797284603118896,
-      "learning_rate": 2.141909567452793e-06,
-      "loss": 2.1487,
+      "grad_norm": 2.303684949874878,
+      "learning_rate": 2.2070660169278166e-06,
+      "loss": 2.1345,
       "step": 940
     },
     {
       "epoch": 0.41892042292710074,
-      "grad_norm": 2.340677261352539,
-      "learning_rate": 2.0777189378234143e-06,
-      "loss": 1.695,
+      "grad_norm": 2.5213372707366943,
+      "learning_rate": 2.141909567452793e-06,
+      "loss": 1.7136,
       "step": 941
     },
     {
       "epoch": 0.41936560934891487,
-      "grad_norm": 2.2142953872680664,
-      "learning_rate": 2.014494761575314e-06,
-      "loss": 1.9991,
+      "grad_norm": 2.287384033203125,
+      "learning_rate": 2.0777189378234143e-06,
+      "loss": 2.0467,
       "step": 942
     },
     {
       "epoch": 0.419810795770729,
-      "grad_norm": 2.083629846572876,
-      "learning_rate": 1.9522376627055583e-06,
-      "loss": 1.3691,
+      "grad_norm": 2.4261367321014404,
+      "learning_rate": 2.014494761575314e-06,
+      "loss": 1.3849,
       "step": 943
     },
     {
       "epoch": 0.4202559821925431,
-      "grad_norm": 2.2179574966430664,
-      "learning_rate": 1.8909482556666024e-06,
-      "loss": 1.5776,
+      "grad_norm": 2.379718065261841,
+      "learning_rate": 1.9522376627055583e-06,
+      "loss": 1.5418,
       "step": 944
     },
     {
       "epoch": 0.42070116861435725,
-      "grad_norm": 2.716609239578247,
-      "learning_rate": 1.8306271453601199e-06,
-      "loss": 1.6952,
+      "grad_norm": 2.5570926666259766,
+      "learning_rate": 1.8909482556666024e-06,
+      "loss": 1.6895,
       "step": 945
     },
     {
       "epoch": 0.42114635503617137,
-      "grad_norm": 3.5457558631896973,
-      "learning_rate": 1.771274927131139e-06,
-      "loss": 2.2282,
+      "grad_norm": 3.3661398887634277,
+      "learning_rate": 1.8306271453601199e-06,
+      "loss": 2.2852,
       "step": 946
     },
     {
       "epoch": 0.42159154145798555,
-      "grad_norm": 2.859788656234741,
-      "learning_rate": 1.712892186762083e-06,
-      "loss": 2.4899,
+      "grad_norm": 2.8509480953216553,
+      "learning_rate": 1.771274927131139e-06,
+      "loss": 2.4994,
       "step": 947
     },
     {
       "epoch": 0.4220367278797997,
-      "grad_norm": 2.126269817352295,
-      "learning_rate": 1.6554795004670388e-06,
-      "loss": 1.7597,
+      "grad_norm": 2.0693445205688477,
+      "learning_rate": 1.712892186762083e-06,
+      "loss": 1.7586,
       "step": 948
     },
     {
       "epoch": 0.4224819143016138,
-      "grad_norm": 11.28099250793457,
-      "learning_rate": 1.5990374348860305e-06,
-      "loss": 2.0871,
+      "grad_norm": 3.6371309757232666,
+      "learning_rate": 1.6554795004670388e-06,
+      "loss": 2.0685,
       "step": 949
     },
     {
       "epoch": 0.42292710072342793,
-      "grad_norm": 2.5182433128356934,
-      "learning_rate": 1.543566547079467e-06,
-      "loss": 1.6924,
+      "grad_norm": 2.525317907333374,
+      "learning_rate": 1.5990374348860305e-06,
+      "loss": 1.6713,
       "step": 950
     },
     {
       "epoch": 0.42337228714524205,
-      "grad_norm": 2.587977886199951,
-      "learning_rate": 1.4890673845226133e-06,
-      "loss": 2.0768,
+      "grad_norm": 2.3136887550354004,
+      "learning_rate": 1.543566547079467e-06,
+      "loss": 2.0899,
       "step": 951
     },
     {
       "epoch": 0.4238174735670562,
-      "grad_norm": 2.2390692234039307,
-      "learning_rate": 1.4355404851001952e-06,
-      "loss": 1.8168,
+      "grad_norm": 2.1782240867614746,
+      "learning_rate": 1.4890673845226133e-06,
+      "loss": 1.8084,
       "step": 952
     },
     {
       "epoch": 0.42426265998887036,
-      "grad_norm": 2.229036331176758,
-      "learning_rate": 1.3829863771011253e-06,
-      "loss": 1.4062,
+      "grad_norm": 1.9394268989562988,
+      "learning_rate": 1.4355404851001952e-06,
+      "loss": 1.3746,
       "step": 953
     },
     {
       "epoch": 0.4247078464106845,
-      "grad_norm": 2.124692440032959,
-      "learning_rate": 1.3314055792131964e-06,
-      "loss": 1.9792,
+      "grad_norm": 2.159578323364258,
+      "learning_rate": 1.3829863771011253e-06,
+      "loss": 1.9826,
       "step": 954
     },
     {
       "epoch": 0.4251530328324986,
-      "grad_norm": 9.60531997680664,
-      "learning_rate": 1.280798600518085e-06,
-      "loss": 2.1784,
+      "grad_norm": 6.865352153778076,
+      "learning_rate": 1.3314055792131964e-06,
+      "loss": 2.2541,
       "step": 955
     },
     {
       "epoch": 0.42559821925431274,
-      "grad_norm": 3.135800838470459,
-      "learning_rate": 1.231165940486234e-06,
-      "loss": 2.4013,
+      "grad_norm": 3.5491487979888916,
+      "learning_rate": 1.280798600518085e-06,
+      "loss": 2.3958,
       "step": 956
     },
     {
       "epoch": 0.42604340567612686,
-      "grad_norm": 1.8015056848526,
-      "learning_rate": 1.1825080889719563e-06,
-      "loss": 2.0513,
+      "grad_norm": 1.782197117805481,
+      "learning_rate": 1.231165940486234e-06,
+      "loss": 2.0068,
       "step": 957
     },
     {
       "epoch": 0.426488592097941,
-      "grad_norm": 2.0358049869537354,
-      "learning_rate": 1.134825526208605e-06,
-      "loss": 1.6947,
+      "grad_norm": 2.2239012718200684,
+      "learning_rate": 1.1825080889719563e-06,
+      "loss": 1.695,
       "step": 958
     },
     {
       "epoch": 0.42693377851975517,
-      "grad_norm": 2.1654505729675293,
-      "learning_rate": 1.0881187228038215e-06,
-      "loss": 1.5756,
+      "grad_norm": 2.1969902515411377,
+      "learning_rate": 1.134825526208605e-06,
+      "loss": 1.5873,
       "step": 959
     },
     {
       "epoch": 0.4273789649415693,
-      "grad_norm": 1.4749082326889038,
-      "learning_rate": 1.0423881397349068e-06,
-      "loss": 1.0004,
+      "grad_norm": 1.4430875778198242,
+      "learning_rate": 1.0881187228038215e-06,
+      "loss": 0.9703,
       "step": 960
     },
     {
       "epoch": 0.4278241513633834,
-      "grad_norm": 2.168402671813965,
-      "learning_rate": 9.976342283442463e-07,
-      "loss": 1.8201,
+      "grad_norm": 2.2270007133483887,
+      "learning_rate": 1.0423881397349068e-06,
+      "loss": 1.8217,
       "step": 961
     },
     {
       "epoch": 0.42826933778519755,
-      "grad_norm": 2.350822925567627,
-      "learning_rate": 9.538574303348813e-07,
-      "loss": 1.9771,
+      "grad_norm": 2.846492290496826,
+      "learning_rate": 9.976342283442463e-07,
+      "loss": 2.0558,
       "step": 962
     },
     {
       "epoch": 0.42871452420701167,
-      "grad_norm": 3.4289748668670654,
-      "learning_rate": 9.110581777661331e-07,
-      "loss": 2.1341,
+      "grad_norm": 3.4389944076538086,
+      "learning_rate": 9.538574303348813e-07,
+      "loss": 2.1505,
       "step": 963
     },
     {
       "epoch": 0.4291597106288258,
-      "grad_norm": 2.5226845741271973,
-      "learning_rate": 8.692368930493521e-07,
-      "loss": 1.4104,
+      "grad_norm": 2.7002594470977783,
+      "learning_rate": 9.110581777661331e-07,
+      "loss": 1.4517,
       "step": 964
     },
     {
       "epoch": 0.42960489705064,
-      "grad_norm": 3.7586185932159424,
-      "learning_rate": 8.283939889437209e-07,
-      "loss": 2.83,
+      "grad_norm": 3.454148292541504,
+      "learning_rate": 8.692368930493521e-07,
+      "loss": 2.8175,
       "step": 965
     },
     {
       "epoch": 0.4300500834724541,
-      "grad_norm": 2.972186803817749,
-      "learning_rate": 7.885298685522235e-07,
-      "loss": 2.0237,
+      "grad_norm": 3.0087950229644775,
+      "learning_rate": 8.283939889437209e-07,
+      "loss": 2.0103,
       "step": 966
     },
     {
       "epoch": 0.43049526989426823,
-      "grad_norm": 2.904418468475342,
-      "learning_rate": 7.496449253176274e-07,
-      "loss": 1.6953,
+      "grad_norm": 2.8295722007751465,
+      "learning_rate": 7.885298685522235e-07,
+      "loss": 1.6394,
       "step": 967
     },
     {
       "epoch": 0.43094045631608235,
-      "grad_norm": 2.9558122158050537,
-      "learning_rate": 7.117395430186414e-07,
-      "loss": 3.0708,
+      "grad_norm": 3.0573501586914062,
+      "learning_rate": 7.496449253176274e-07,
+      "loss": 3.1388,
       "step": 968
     },
     {
       "epoch": 0.4313856427378965,
-      "grad_norm": 2.732140064239502,
-      "learning_rate": 6.748140957660631e-07,
-      "loss": 1.7392,
+      "grad_norm": 2.893523931503296,
+      "learning_rate": 7.117395430186414e-07,
+      "loss": 1.8017,
       "step": 969
     },
     {
       "epoch": 0.4318308291597106,
-      "grad_norm": 2.7709476947784424,
-      "learning_rate": 6.388689479991605e-07,
-      "loss": 1.7855,
+      "grad_norm": 4.128028869628906,
+      "learning_rate": 6.748140957660631e-07,
+      "loss": 1.7412,
       "step": 970
     },
     {
       "epoch": 0.4322760155815248,
-      "grad_norm": 1.9800370931625366,
-      "learning_rate": 6.039044544820404e-07,
-      "loss": 0.8279,
+      "grad_norm": 1.7751336097717285,
+      "learning_rate": 6.388689479991605e-07,
+      "loss": 0.8345,
       "step": 971
     },
     {
       "epoch": 0.4327212020033389,
-      "grad_norm": 3.503222703933716,
-      "learning_rate": 5.699209603001076e-07,
-      "loss": 1.2507,
+      "grad_norm": 3.487071990966797,
+      "learning_rate": 6.039044544820404e-07,
+      "loss": 1.2831,
       "step": 972
     },
     {
       "epoch": 0.43316638842515304,
-      "grad_norm": 3.6716744899749756,
-      "learning_rate": 5.369188008567672e-07,
-      "loss": 1.5849,
+      "grad_norm": 4.514010906219482,
+      "learning_rate": 5.699209603001076e-07,
+      "loss": 1.6334,
       "step": 973
     },
     {
       "epoch": 0.43361157484696716,
-      "grad_norm": 3.8361189365386963,
-      "learning_rate": 5.048983018699827e-07,
-      "loss": 2.7786,
+      "grad_norm": 4.002710819244385,
+      "learning_rate": 5.369188008567672e-07,
+      "loss": 2.7366,
       "step": 974
     },
     {
       "epoch": 0.4340567612687813,
-      "grad_norm": 2.1607441902160645,
-      "learning_rate": 4.738597793691679e-07,
-      "loss": 1.6388,
+      "grad_norm": 2.2446699142456055,
+      "learning_rate": 5.048983018699827e-07,
+      "loss": 1.6826,
       "step": 975
     },
     {
       "epoch": 0.4345019476905954,
-      "grad_norm": 2.7278549671173096,
-      "learning_rate": 4.438035396920004e-07,
-      "loss": 2.5643,
+      "grad_norm": 2.6733949184417725,
+      "learning_rate": 4.738597793691679e-07,
+      "loss": 2.561,
       "step": 976
     },
     {
       "epoch": 0.4349471341124096,
-      "grad_norm": 2.0202202796936035,
-      "learning_rate": 4.1472987948143473e-07,
-      "loss": 1.6711,
+      "grad_norm": 1.9742475748062134,
+      "learning_rate": 4.438035396920004e-07,
+      "loss": 1.632,
       "step": 977
     },
     {
       "epoch": 0.4353923205342237,
-      "grad_norm": 2.3101208209991455,
-      "learning_rate": 3.866390856827495e-07,
-      "loss": 1.8005,
+      "grad_norm": 2.39536714553833,
+      "learning_rate": 4.1472987948143473e-07,
+      "loss": 1.8493,
       "step": 978
     },
     {
       "epoch": 0.43583750695603785,
-      "grad_norm": 1.7602624893188477,
-      "learning_rate": 3.595314355407609e-07,
-      "loss": 1.4123,
+      "grad_norm": 2.0998411178588867,
+      "learning_rate": 3.866390856827495e-07,
+      "loss": 1.4321,
       "step": 979
     },
     {
       "epoch": 0.43628269337785197,
-      "grad_norm": 2.8393287658691406,
-      "learning_rate": 3.3340719659701313e-07,
-      "loss": 1.9375,
+      "grad_norm": 4.382582187652588,
+      "learning_rate": 3.595314355407609e-07,
+      "loss": 1.9031,
       "step": 980
     },
     {
       "epoch": 0.4367278797996661,
-      "grad_norm": 2.671792507171631,
-      "learning_rate": 3.0826662668720364e-07,
-      "loss": 2.1422,
+      "grad_norm": 2.58941912651062,
+      "learning_rate": 3.3340719659701313e-07,
+      "loss": 2.1195,
       "step": 981
     },
     {
       "epoch": 0.4371730662214802,
-      "grad_norm": 2.798468828201294,
-      "learning_rate": 2.841099739386066e-07,
-      "loss": 1.1786,
+      "grad_norm": 2.468071460723877,
+      "learning_rate": 3.0826662668720364e-07,
+      "loss": 1.1861,
       "step": 982
     },
     {
       "epoch": 0.4376182526432944,
-      "grad_norm": 2.4670653343200684,
-      "learning_rate": 2.609374767676309e-07,
-      "loss": 2.0231,
+      "grad_norm": 2.3609464168548584,
+      "learning_rate": 2.841099739386066e-07,
+      "loss": 2.0167,
       "step": 983
     },
     {
       "epoch": 0.43806343906510853,
-      "grad_norm": 2.650512218475342,
-      "learning_rate": 2.387493638774774e-07,
-      "loss": 2.1189,
+      "grad_norm": 2.617172956466675,
+      "learning_rate": 2.609374767676309e-07,
+      "loss": 2.0974,
       "step": 984
     },
     {
       "epoch": 0.43850862548692265,
-      "grad_norm": 2.984624147415161,
-      "learning_rate": 2.175458542558517e-07,
-      "loss": 2.3214,
+      "grad_norm": 2.8748812675476074,
+      "learning_rate": 2.387493638774774e-07,
+      "loss": 2.28,
       "step": 985
     },
     {
       "epoch": 0.4389538119087368,
-      "grad_norm": 2.806220769882202,
-      "learning_rate": 1.973271571728441e-07,
-      "loss": 2.8719,
+      "grad_norm": 2.907883644104004,
+      "learning_rate": 2.175458542558517e-07,
+      "loss": 2.9018,
       "step": 986
     },
     {
       "epoch": 0.4393989983305509,
-      "grad_norm": 2.488363742828369,
-      "learning_rate": 1.7809347217881966e-07,
-      "loss": 2.2061,
+      "grad_norm": 2.2215635776519775,
+      "learning_rate": 1.973271571728441e-07,
+      "loss": 2.2156,
       "step": 987
     },
     {
       "epoch": 0.43984418475236503,
-      "grad_norm": 4.78094482421875,
-      "learning_rate": 1.598449891024978e-07,
-      "loss": 2.4017,
+      "grad_norm": 4.6469597816467285,
+      "learning_rate": 1.7809347217881966e-07,
+      "loss": 2.4033,
       "step": 988
     },
     {
       "epoch": 0.4402893711741792,
-      "grad_norm": 1.6316055059432983,
-      "learning_rate": 1.425818880490315e-07,
-      "loss": 1.8267,
+      "grad_norm": 1.8456226587295532,
+      "learning_rate": 1.598449891024978e-07,
+      "loss": 1.8562,
       "step": 989
     },
     {
       "epoch": 0.44073455759599334,
-      "grad_norm": 2.0801150798797607,
-      "learning_rate": 1.2630433939825327e-07,
-      "loss": 1.6163,
+      "grad_norm": 2.2966156005859375,
+      "learning_rate": 1.425818880490315e-07,
+      "loss": 1.5724,
       "step": 990
     },
     {
       "epoch": 0.44117974401780746,
-      "grad_norm": 1.5451546907424927,
-      "learning_rate": 1.1101250380300965e-07,
-      "loss": 1.3939,
+      "grad_norm": 1.6308859586715698,
+      "learning_rate": 1.2630433939825327e-07,
+      "loss": 1.4085,
       "step": 991
     },
     {
       "epoch": 0.4416249304396216,
-      "grad_norm": 1.971220850944519,
-      "learning_rate": 9.670653218752934e-08,
-      "loss": 1.7467,
+      "grad_norm": 1.9520773887634277,
+      "learning_rate": 1.1101250380300965e-07,
+      "loss": 1.746,
       "step": 992
     },
     {
       "epoch": 0.4420701168614357,
-      "grad_norm": 3.593325138092041,
-      "learning_rate": 8.33865657459909e-08,
-      "loss": 2.2187,
+      "grad_norm": 3.1653501987457275,
+      "learning_rate": 9.670653218752934e-08,
+      "loss": 2.2131,
       "step": 993
     },
     {
       "epoch": 0.44251530328324984,
-      "grad_norm": 1.9523143768310547,
-      "learning_rate": 7.105273594107953e-08,
-      "loss": 1.2635,
+      "grad_norm": 3.58504319190979,
+      "learning_rate": 8.33865657459909e-08,
+      "loss": 1.3358,
       "step": 994
     },
     {
       "epoch": 0.442960489705064,
-      "grad_norm": 2.172438144683838,
-      "learning_rate": 5.970516450271025e-08,
-      "loss": 1.3526,
+      "grad_norm": 2.2316412925720215,
+      "learning_rate": 7.105273594107953e-08,
+      "loss": 1.3569,
       "step": 995
     },
     {
       "epoch": 0.44340567612687815,
-      "grad_norm": 1.9198912382125854,
-      "learning_rate": 4.934396342684e-08,
-      "loss": 1.1006,
+      "grad_norm": 1.7121682167053223,
+      "learning_rate": 5.970516450271025e-08,
+      "loss": 1.0737,
       "step": 996
     },
     {
       "epoch": 0.44385086254869227,
-      "grad_norm": 2.753998041152954,
-      "learning_rate": 3.996923497434635e-08,
-      "loss": 2.1173,
+      "grad_norm": 3.018528699874878,
+      "learning_rate": 4.934396342684e-08,
+      "loss": 2.1253,
       "step": 997
     },
     {
       "epoch": 0.4442960489705064,
-      "grad_norm": 2.4736664295196533,
-      "learning_rate": 3.1581071670006015e-08,
-      "loss": 2.3068,
+      "grad_norm": 2.544288396835327,
+      "learning_rate": 3.996923497434635e-08,
+      "loss": 2.3852,
       "step": 998
     },
     {
       "epoch": 0.4447412353923205,
-      "grad_norm": 2.1527669429779053,
-      "learning_rate": 2.417955630159563e-08,
-      "loss": 1.4739,
+      "grad_norm": 2.447824716567993,
+      "learning_rate": 3.1581071670006015e-08,
+      "loss": 1.4901,
       "step": 999
     },
     {
       "epoch": 0.44518642181413465,
-      "grad_norm": 2.12872052192688,
-      "learning_rate": 1.7764761919103477e-08,
-      "loss": 2.0056,
+      "grad_norm": 2.373067855834961,
+      "learning_rate": 2.417955630159563e-08,
+      "loss": 1.9986,
+      "step": 1000
+    },
+    {
+      "epoch": 0.44518642181413465,
+      "eval_loss": 1.9380669593811035,
+      "eval_runtime": 136.1191,
+      "eval_samples_per_second": 3.475,
+      "eval_steps_per_second": 3.475,
       "step": 1000
     }
   ],