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"
@@ -38,6982 +38,6982 @@
     },
     {
       "epoch": 0.0022259321090706734,
-      "grad_norm": 2.553895950317383,
+      "grad_norm": 2.459383487701416,
       "learning_rate": 0.00019999950652018584,
       "loss": 4.1176,
       "step": 5
     },
     {
       "epoch": 0.002671118530884808,
-      "grad_norm": 3.750920534133911,
+      "grad_norm": 3.519275188446045,
       "learning_rate": 0.0001999980260856137,
-      "loss": 4.3848,
+      "loss": 4.3858,
       "step": 6
     },
     {
       "epoch": 0.0031163049526989426,
-      "grad_norm": 4.422790050506592,
+      "grad_norm": 4.532252311706543,
       "learning_rate": 0.000199995558710895,
-      "loss": 5.3942,
+      "loss": 5.3821,
       "step": 7
     },
     {
       "epoch": 0.003561491374513077,
       "grad_norm": NaN,
       "learning_rate": 0.000199995558710895,
-      "loss": 3.3069,
+      "loss": 3.3187,
       "step": 8
     },
     {
       "epoch": 0.004006677796327212,
-      "grad_norm": 6.551366329193115,
+      "grad_norm": 4.433828830718994,
       "learning_rate": 0.00019999210442038162,
-      "loss": 3.3369,
+      "loss": 3.3476,
       "step": 9
     },
     {
       "epoch": 0.004451864218141347,
-      "grad_norm": 5.16669225692749,
+      "grad_norm": 4.996153831481934,
       "learning_rate": 0.00019998766324816607,
-      "loss": 4.0822,
+      "loss": 4.0539,
       "step": 10
     },
     {
       "epoch": 0.004897050639955481,
-      "grad_norm": 4.758444309234619,
+      "grad_norm": 3.9972689151763916,
       "learning_rate": 0.0001999822352380809,
-      "loss": 3.1787,
+      "loss": 3.1474,
       "step": 11
     },
     {
       "epoch": 0.005342237061769616,
-      "grad_norm": 7.418261528015137,
+      "grad_norm": 7.407133102416992,
       "learning_rate": 0.00019997582044369843,
-      "loss": 3.594,
+      "loss": 3.568,
       "step": 12
     },
     {
       "epoch": 0.005787423483583751,
-      "grad_norm": 6.103316307067871,
+      "grad_norm": 5.174762725830078,
       "learning_rate": 0.00019996841892833,
-      "loss": 2.7363,
+      "loss": 2.7407,
       "step": 13
     },
     {
       "epoch": 0.006232609905397885,
-      "grad_norm": 5.146238803863525,
+      "grad_norm": 5.499022960662842,
       "learning_rate": 0.00019996003076502565,
-      "loss": 2.4596,
+      "loss": 2.4789,
       "step": 14
     },
     {
       "epoch": 0.00667779632721202,
-      "grad_norm": 6.732138633728027,
+      "grad_norm": 4.329870223999023,
       "learning_rate": 0.00019995065603657316,
-      "loss": 3.1536,
+      "loss": 3.1731,
       "step": 15
     },
     {
       "epoch": 0.007122982749026154,
-      "grad_norm": 6.138510227203369,
+      "grad_norm": 5.633994102478027,
       "learning_rate": 0.0001999402948354973,
-      "loss": 2.5301,
+      "loss": 2.5078,
       "step": 16
     },
     {
       "epoch": 0.007568169170840289,
-      "grad_norm": 7.237410068511963,
+      "grad_norm": 6.770676136016846,
       "learning_rate": 0.00019992894726405893,
-      "loss": 2.5562,
+      "loss": 2.5148,
       "step": 17
     },
     {
       "epoch": 0.008013355592654424,
-      "grad_norm": 5.183891773223877,
+      "grad_norm": 4.9481964111328125,
       "learning_rate": 0.000199916613434254,
-      "loss": 3.6864,
+      "loss": 3.6853,
       "step": 18
     },
     {
       "epoch": 0.00845854201446856,
-      "grad_norm": 4.217153072357178,
+      "grad_norm": 4.193163871765137,
       "learning_rate": 0.0001999032934678125,
-      "loss": 2.2188,
+      "loss": 2.2103,
       "step": 19
     },
     {
       "epoch": 0.008903728436282694,
-      "grad_norm": 4.953031063079834,
+      "grad_norm": 4.7864813804626465,
       "learning_rate": 0.00019988898749619702,
-      "loss": 2.159,
+      "loss": 2.1588,
       "step": 20
     },
     {
       "epoch": 0.009348914858096828,
-      "grad_norm": 4.839296340942383,
+      "grad_norm": 4.700019836425781,
       "learning_rate": 0.00019987369566060176,
-      "loss": 2.3152,
+      "loss": 2.3093,
       "step": 21
     },
     {
       "epoch": 0.009794101279910962,
-      "grad_norm": 4.408138275146484,
+      "grad_norm": 4.116794109344482,
       "learning_rate": 0.00019985741811195097,
-      "loss": 2.2196,
+      "loss": 2.1852,
       "step": 22
     },
     {
       "epoch": 0.010239287701725098,
-      "grad_norm": 3.8916175365448,
+      "grad_norm": 3.7330455780029297,
       "learning_rate": 0.00019984015501089752,
-      "loss": 2.4175,
+      "loss": 2.4284,
       "step": 23
     },
     {
       "epoch": 0.010684474123539232,
-      "grad_norm": 4.368136882781982,
+      "grad_norm": 4.410067081451416,
       "learning_rate": 0.0001998219065278212,
-      "loss": 2.5028,
+      "loss": 2.4726,
       "step": 24
     },
     {
       "epoch": 0.011129660545353366,
-      "grad_norm": 6.353954792022705,
+      "grad_norm": 6.348716735839844,
       "learning_rate": 0.00019980267284282717,
-      "loss": 2.6495,
+      "loss": 2.6418,
       "step": 25
     },
     {
       "epoch": 0.011574846967167502,
-      "grad_norm": 4.783512115478516,
+      "grad_norm": 4.5325775146484375,
       "learning_rate": 0.00019978245414574417,
-      "loss": 1.9392,
+      "loss": 1.9052,
       "step": 26
     },
     {
       "epoch": 0.012020033388981636,
-      "grad_norm": 3.656338930130005,
+      "grad_norm": 3.46451997756958,
       "learning_rate": 0.00019976125063612252,
-      "loss": 1.8599,
+      "loss": 1.8182,
       "step": 27
     },
     {
       "epoch": 0.01246521981079577,
-      "grad_norm": 5.02943229675293,
+      "grad_norm": 4.992573261260986,
       "learning_rate": 0.00019973906252323238,
-      "loss": 2.272,
+      "loss": 2.2369,
       "step": 28
     },
     {
       "epoch": 0.012910406232609905,
-      "grad_norm": 3.1446378231048584,
+      "grad_norm": 3.1490421295166016,
       "learning_rate": 0.0001997158900260614,
-      "loss": 2.2221,
+      "loss": 2.2188,
       "step": 29
     },
     {
       "epoch": 0.01335559265442404,
-      "grad_norm": 2.948500871658325,
+      "grad_norm": 2.976283550262451,
       "learning_rate": 0.0001996917333733128,
-      "loss": 1.7518,
+      "loss": 1.7364,
       "step": 30
     },
     {
       "epoch": 0.013800779076238175,
-      "grad_norm": 3.700584650039673,
+      "grad_norm": 3.7836174964904785,
       "learning_rate": 0.00019966659280340297,
-      "loss": 2.2607,
+      "loss": 2.2586,
       "step": 31
     },
     {
       "epoch": 0.014245965498052309,
-      "grad_norm": 4.679660797119141,
+      "grad_norm": 4.688706874847412,
       "learning_rate": 0.00019964046856445924,
-      "loss": 1.4464,
+      "loss": 1.4408,
       "step": 32
     },
     {
       "epoch": 0.014691151919866445,
-      "grad_norm": 13.211495399475098,
+      "grad_norm": 4.265240669250488,
       "learning_rate": 0.00019961336091431727,
-      "loss": 2.8803,
+      "loss": 2.8621,
       "step": 33
     },
     {
       "epoch": 0.015136338341680579,
-      "grad_norm": 2.2771530151367188,
+      "grad_norm": 2.2330338954925537,
       "learning_rate": 0.00019958527012051857,
-      "loss": 1.5439,
+      "loss": 1.541,
       "step": 34
     },
     {
       "epoch": 0.015581524763494713,
-      "grad_norm": 4.475869178771973,
+      "grad_norm": 4.225557327270508,
       "learning_rate": 0.00019955619646030802,
-      "loss": 2.488,
+      "loss": 2.5235,
       "step": 35
     },
     {
       "epoch": 0.016026711185308847,
-      "grad_norm": 3.8205974102020264,
+      "grad_norm": 3.765263319015503,
       "learning_rate": 0.00019952614022063084,
-      "loss": 2.0277,
+      "loss": 1.9924,
       "step": 36
     },
     {
       "epoch": 0.016471897607122983,
-      "grad_norm": 3.721785545349121,
+      "grad_norm": 3.4039769172668457,
       "learning_rate": 0.00019949510169813003,
-      "loss": 2.293,
+      "loss": 2.2609,
       "step": 37
     },
     {
       "epoch": 0.01691708402893712,
-      "grad_norm": 5.029823303222656,
+      "grad_norm": 4.773370742797852,
       "learning_rate": 0.00019946308119914323,
-      "loss": 2.5691,
+      "loss": 2.5648,
       "step": 38
     },
     {
       "epoch": 0.01736227045075125,
-      "grad_norm": 3.547927141189575,
+      "grad_norm": 2.5985257625579834,
       "learning_rate": 0.0001994300790396999,
-      "loss": 1.4268,
+      "loss": 1.4,
       "step": 39
     },
     {
       "epoch": 0.017807456872565387,
-      "grad_norm": 4.380621910095215,
+      "grad_norm": 3.532564878463745,
       "learning_rate": 0.000199396095545518,
-      "loss": 2.2319,
+      "loss": 2.2226,
       "step": 40
     },
     {
       "epoch": 0.01825264329437952,
-      "grad_norm": 3.113034248352051,
+      "grad_norm": 3.193943500518799,
       "learning_rate": 0.00019936113105200085,
-      "loss": 2.0469,
+      "loss": 2.0588,
       "step": 41
     },
     {
       "epoch": 0.018697829716193656,
-      "grad_norm": 3.8902220726013184,
+      "grad_norm": 4.202755928039551,
       "learning_rate": 0.00019932518590423394,
-      "loss": 2.7515,
+      "loss": 2.775,
       "step": 42
     },
     {
       "epoch": 0.01914301613800779,
-      "grad_norm": 6.047741413116455,
+      "grad_norm": 3.9249866008758545,
       "learning_rate": 0.00019928826045698136,
-      "loss": 2.1295,
+      "loss": 2.0668,
       "step": 43
     },
     {
       "epoch": 0.019588202559821924,
-      "grad_norm": 1.9819254875183105,
+      "grad_norm": 1.8659396171569824,
       "learning_rate": 0.0001992503550746824,
-      "loss": 1.1775,
+      "loss": 1.1624,
       "step": 44
     },
     {
       "epoch": 0.02003338898163606,
-      "grad_norm": 3.1958131790161133,
+      "grad_norm": 3.1702041625976562,
       "learning_rate": 0.0001992114701314478,
-      "loss": 1.9084,
+      "loss": 1.9217,
       "step": 45
     },
     {
       "epoch": 0.020478575403450196,
-      "grad_norm": 5.241977691650391,
+      "grad_norm": 5.405176639556885,
       "learning_rate": 0.0001991716060110563,
-      "loss": 3.0362,
+      "loss": 3.0466,
       "step": 46
     },
     {
       "epoch": 0.020923761825264328,
-      "grad_norm": 2.977036237716675,
+      "grad_norm": 2.9543397426605225,
       "learning_rate": 0.00019913076310695068,
-      "loss": 1.704,
+      "loss": 1.7083,
       "step": 47
     },
     {
       "epoch": 0.021368948247078464,
-      "grad_norm": 4.862756729125977,
+      "grad_norm": 4.4758687019348145,
       "learning_rate": 0.00019908894182223388,
-      "loss": 1.9846,
+      "loss": 1.9944,
       "step": 48
     },
     {
       "epoch": 0.0218141346688926,
-      "grad_norm": 2.590459108352661,
+      "grad_norm": 2.584315299987793,
       "learning_rate": 0.00019904614256966512,
-      "loss": 2.2383,
+      "loss": 2.2473,
       "step": 49
     },
     {
       "epoch": 0.022259321090706732,
-      "grad_norm": 4.649803161621094,
+      "grad_norm": 4.648918628692627,
       "learning_rate": 0.00019900236577165576,
-      "loss": 2.0824,
+      "loss": 2.0701,
       "step": 50
     },
     {
       "epoch": 0.02270450751252087,
-      "grad_norm": 4.012078285217285,
+      "grad_norm": 3.875636100769043,
       "learning_rate": 0.0001989576118602651,
-      "loss": 1.8357,
+      "loss": 1.8491,
       "step": 51
     },
     {
       "epoch": 0.023149693934335004,
-      "grad_norm": 2.382760763168335,
+      "grad_norm": 2.448580026626587,
       "learning_rate": 0.00019891188127719618,
-      "loss": 1.4587,
+      "loss": 1.4837,
       "step": 52
     },
     {
       "epoch": 0.023594880356149137,
-      "grad_norm": 3.601736545562744,
+      "grad_norm": 3.6881260871887207,
       "learning_rate": 0.0001988651744737914,
-      "loss": 2.118,
+      "loss": 2.1082,
       "step": 53
     },
     {
       "epoch": 0.024040066777963272,
-      "grad_norm": 3.790908098220825,
+      "grad_norm": 3.7397072315216064,
       "learning_rate": 0.00019881749191102808,
-      "loss": 2.4478,
+      "loss": 2.4085,
       "step": 54
     },
     {
       "epoch": 0.02448525319977741,
-      "grad_norm": 5.634518623352051,
+      "grad_norm": 4.061782360076904,
       "learning_rate": 0.00019876883405951377,
-      "loss": 2.0788,
+      "loss": 2.1085,
       "step": 55
     },
     {
       "epoch": 0.02493043962159154,
-      "grad_norm": 2.9153330326080322,
+      "grad_norm": 2.8737947940826416,
       "learning_rate": 0.00019871920139948192,
-      "loss": 1.753,
+      "loss": 1.7602,
       "step": 56
     },
     {
       "epoch": 0.025375626043405677,
-      "grad_norm": 3.083784341812134,
+      "grad_norm": 3.1491646766662598,
       "learning_rate": 0.0001986685944207868,
-      "loss": 2.0111,
+      "loss": 2.0325,
       "step": 57
     },
     {
       "epoch": 0.02582081246521981,
-      "grad_norm": 3.6852848529815674,
+      "grad_norm": 4.124095916748047,
       "learning_rate": 0.0001986170136228989,
-      "loss": 2.4499,
+      "loss": 2.4,
       "step": 58
     },
     {
       "epoch": 0.026265998887033945,
-      "grad_norm": 3.1907293796539307,
+      "grad_norm": 3.113374710083008,
       "learning_rate": 0.00019856445951489982,
-      "loss": 1.6657,
+      "loss": 1.6621,
       "step": 59
     },
     {
       "epoch": 0.02671118530884808,
-      "grad_norm": 4.264678955078125,
+      "grad_norm": 4.29051399230957,
       "learning_rate": 0.0001985109326154774,
-      "loss": 2.1187,
+      "loss": 2.1999,
       "step": 60
     },
     {
       "epoch": 0.027156371730662213,
-      "grad_norm": 2.061347484588623,
+      "grad_norm": 2.0840396881103516,
       "learning_rate": 0.00019845643345292054,
-      "loss": 1.2979,
+      "loss": 1.303,
       "step": 61
     },
     {
       "epoch": 0.02760155815247635,
-      "grad_norm": 2.9782309532165527,
+      "grad_norm": 2.9523775577545166,
       "learning_rate": 0.00019840096256511398,
-      "loss": 1.96,
+      "loss": 1.9373,
       "step": 62
     },
     {
       "epoch": 0.028046744574290485,
-      "grad_norm": 3.8613250255584717,
+      "grad_norm": 3.697802782058716,
       "learning_rate": 0.00019834452049953297,
-      "loss": 3.4576,
+      "loss": 3.4386,
       "step": 63
     },
     {
       "epoch": 0.028491930996104618,
-      "grad_norm": 1.9967275857925415,
+      "grad_norm": 1.9980502128601074,
       "learning_rate": 0.00019828710781323792,
-      "loss": 0.9779,
+      "loss": 0.9815,
       "step": 64
     },
     {
       "epoch": 0.028937117417918753,
-      "grad_norm": 2.790372371673584,
+      "grad_norm": 2.667588233947754,
       "learning_rate": 0.0001982287250728689,
-      "loss": 1.5419,
+      "loss": 1.5538,
       "step": 65
     },
     {
       "epoch": 0.02938230383973289,
-      "grad_norm": 2.1522772312164307,
+      "grad_norm": 2.119844913482666,
       "learning_rate": 0.0001981693728546399,
-      "loss": 1.3176,
+      "loss": 1.3043,
       "step": 66
     },
     {
       "epoch": 0.029827490261547022,
-      "grad_norm": 2.084890842437744,
+      "grad_norm": 2.006866931915283,
       "learning_rate": 0.0001981090517443334,
-      "loss": 0.8775,
+      "loss": 0.8754,
       "step": 67
     },
     {
       "epoch": 0.030272676683361158,
-      "grad_norm": 2.6281416416168213,
+      "grad_norm": 2.604780912399292,
       "learning_rate": 0.00019804776233729444,
-      "loss": 1.7329,
+      "loss": 1.7314,
       "step": 68
     },
     {
       "epoch": 0.030717863105175294,
-      "grad_norm": 2.443356990814209,
+      "grad_norm": 2.333446979522705,
       "learning_rate": 0.0001979855052384247,
-      "loss": 1.6412,
+      "loss": 1.6191,
       "step": 69
     },
     {
       "epoch": 0.031163049526989426,
-      "grad_norm": 3.233210325241089,
+      "grad_norm": 3.268148899078369,
       "learning_rate": 0.00019792228106217658,
-      "loss": 1.9627,
+      "loss": 1.9537,
       "step": 70
     },
     {
       "epoch": 0.03160823594880356,
-      "grad_norm": 4.562886714935303,
+      "grad_norm": 4.541009426116943,
       "learning_rate": 0.00019785809043254722,
-      "loss": 1.3901,
+      "loss": 1.4556,
       "step": 71
     },
     {
       "epoch": 0.032053422370617694,
-      "grad_norm": 4.928150653839111,
+      "grad_norm": 4.054357528686523,
       "learning_rate": 0.0001977929339830722,
-      "loss": 1.8413,
+      "loss": 1.8164,
       "step": 72
     },
     {
       "epoch": 0.032498608792431834,
-      "grad_norm": 2.872011184692383,
+      "grad_norm": 2.9288697242736816,
       "learning_rate": 0.00019772681235681936,
-      "loss": 1.8261,
+      "loss": 1.81,
       "step": 73
     },
     {
       "epoch": 0.032943795214245966,
-      "grad_norm": 4.901350975036621,
+      "grad_norm": 3.799945831298828,
       "learning_rate": 0.00019765972620638248,
-      "loss": 2.5204,
+      "loss": 2.5435,
       "step": 74
     },
     {
       "epoch": 0.0333889816360601,
-      "grad_norm": 4.601914882659912,
+      "grad_norm": 4.2441301345825195,
       "learning_rate": 0.00019759167619387476,
-      "loss": 2.3304,
+      "loss": 2.3859,
       "step": 75
     },
     {
       "epoch": 0.03383416805787424,
-      "grad_norm": 4.285044193267822,
+      "grad_norm": 4.169832229614258,
       "learning_rate": 0.00019752266299092236,
-      "loss": 2.148,
+      "loss": 2.1274,
       "step": 76
     },
     {
       "epoch": 0.03427935447968837,
-      "grad_norm": 3.728804111480713,
+      "grad_norm": 3.6839637756347656,
       "learning_rate": 0.00019745268727865774,
-      "loss": 2.2936,
+      "loss": 2.2896,
       "step": 77
     },
     {
       "epoch": 0.0347245409015025,
-      "grad_norm": 3.0641114711761475,
+      "grad_norm": 2.736715793609619,
       "learning_rate": 0.0001973817497477129,
-      "loss": 1.7878,
+      "loss": 1.7153,
       "step": 78
     },
     {
       "epoch": 0.03516972732331664,
-      "grad_norm": 4.134016990661621,
+      "grad_norm": 4.134005069732666,
       "learning_rate": 0.00019730985109821266,
-      "loss": 2.9227,
+      "loss": 2.9136,
       "step": 79
     },
     {
       "epoch": 0.035614913745130775,
-      "grad_norm": 12.307600975036621,
+      "grad_norm": 4.007108688354492,
       "learning_rate": 0.00019723699203976766,
-      "loss": 2.7279,
+      "loss": 2.6435,
       "step": 80
     },
     {
       "epoch": 0.03606010016694491,
-      "grad_norm": 3.992847204208374,
+      "grad_norm": 4.0692901611328125,
       "learning_rate": 0.0001971631732914674,
-      "loss": 2.5772,
+      "loss": 2.5131,
       "step": 81
     },
     {
       "epoch": 0.03650528658875904,
-      "grad_norm": 3.676469564437866,
+      "grad_norm": 2.820955753326416,
       "learning_rate": 0.0001970883955818731,
-      "loss": 1.5831,
+      "loss": 1.53,
       "step": 82
     },
     {
       "epoch": 0.03695047301057318,
-      "grad_norm": 3.0918328762054443,
+      "grad_norm": 2.4926910400390625,
       "learning_rate": 0.0001970126596490106,
-      "loss": 1.4057,
+      "loss": 1.3355,
       "step": 83
     },
     {
       "epoch": 0.03739565943238731,
-      "grad_norm": 3.945146322250366,
+      "grad_norm": 4.664624214172363,
       "learning_rate": 0.00019693596624036292,
-      "loss": 1.5807,
+      "loss": 1.6793,
       "step": 84
     },
     {
       "epoch": 0.037840845854201444,
-      "grad_norm": 2.956975221633911,
+      "grad_norm": 2.7750489711761475,
       "learning_rate": 0.0001968583161128631,
-      "loss": 1.9335,
+      "loss": 1.9206,
       "step": 85
     },
     {
       "epoch": 0.03828603227601558,
-      "grad_norm": 2.6506834030151367,
+      "grad_norm": 2.659250259399414,
       "learning_rate": 0.00019677971003288655,
-      "loss": 1.8569,
+      "loss": 1.8099,
       "step": 86
     },
     {
       "epoch": 0.038731218697829715,
-      "grad_norm": 3.6588492393493652,
+      "grad_norm": 3.730412483215332,
       "learning_rate": 0.00019670014877624353,
-      "loss": 1.6848,
+      "loss": 1.702,
       "step": 87
     },
     {
       "epoch": 0.03917640511964385,
-      "grad_norm": 3.677953004837036,
+      "grad_norm": 3.6121342182159424,
       "learning_rate": 0.00019661963312817148,
-      "loss": 2.5729,
+      "loss": 2.5358,
       "step": 88
     },
     {
       "epoch": 0.03962159154145799,
-      "grad_norm": 3.1620278358459473,
+      "grad_norm": 3.176912307739258,
       "learning_rate": 0.0001965381638833274,
-      "loss": 1.8112,
+      "loss": 1.8566,
       "step": 89
     },
     {
       "epoch": 0.04006677796327212,
-      "grad_norm": 4.549400806427002,
+      "grad_norm": 3.7813401222229004,
       "learning_rate": 0.00019645574184577982,
-      "loss": 3.0601,
+      "loss": 3.0385,
       "step": 90
     },
     {
       "epoch": 0.04051196438508625,
-      "grad_norm": 4.790261745452881,
+      "grad_norm": 6.105398178100586,
       "learning_rate": 0.000196372367829001,
-      "loss": 2.0025,
+      "loss": 1.9945,
       "step": 91
     },
     {
       "epoch": 0.04095715080690039,
-      "grad_norm": 2.0430498123168945,
+      "grad_norm": 2.48732852935791,
       "learning_rate": 0.00019628804265585877,
-      "loss": 1.757,
+      "loss": 1.7157,
       "step": 92
     },
     {
       "epoch": 0.041402337228714524,
-      "grad_norm": 3.846684217453003,
+      "grad_norm": 3.6320507526397705,
       "learning_rate": 0.0001962027671586086,
-      "loss": 2.8248,
+      "loss": 2.7751,
       "step": 93
     },
     {
       "epoch": 0.041847523650528656,
-      "grad_norm": 2.1094014644622803,
+      "grad_norm": 2.2230594158172607,
       "learning_rate": 0.0001961165421788852,
-      "loss": 1.388,
+      "loss": 1.4204,
       "step": 94
     },
     {
       "epoch": 0.042292710072342796,
-      "grad_norm": 2.475429058074951,
+      "grad_norm": 2.5647616386413574,
       "learning_rate": 0.0001960293685676943,
-      "loss": 2.109,
+      "loss": 2.1416,
       "step": 95
     },
     {
       "epoch": 0.04273789649415693,
-      "grad_norm": 2.539384603500366,
+      "grad_norm": 2.535376787185669,
       "learning_rate": 0.0001959412471854043,
-      "loss": 1.7985,
+      "loss": 1.8542,
       "step": 96
     },
     {
       "epoch": 0.04318308291597106,
-      "grad_norm": 3.7397396564483643,
+      "grad_norm": 4.080014705657959,
       "learning_rate": 0.0001958521789017376,
-      "loss": 2.3767,
+      "loss": 2.2818,
       "step": 97
     },
     {
       "epoch": 0.0436282693377852,
-      "grad_norm": 4.037478446960449,
+      "grad_norm": 4.706260681152344,
       "learning_rate": 0.00019576216459576222,
-      "loss": 2.5889,
+      "loss": 2.6084,
       "step": 98
     },
     {
       "epoch": 0.04407345575959933,
-      "grad_norm": 5.122471809387207,
+      "grad_norm": 4.561995506286621,
       "learning_rate": 0.00019567120515588308,
-      "loss": 2.2834,
+      "loss": 2.2661,
       "step": 99
     },
     {
       "epoch": 0.044518642181413465,
-      "grad_norm": 3.0687732696533203,
+      "grad_norm": 3.076706647872925,
       "learning_rate": 0.00019557930147983302,
-      "loss": 2.6478,
+      "loss": 2.6317,
       "step": 100
     },
     {
       "epoch": 0.044963828603227604,
-      "grad_norm": 2.641993284225464,
+      "grad_norm": 3.0237557888031006,
       "learning_rate": 0.00019548645447466431,
-      "loss": 1.2885,
+      "loss": 1.333,
       "step": 101
     },
     {
       "epoch": 0.04540901502504174,
-      "grad_norm": 2.480499505996704,
+      "grad_norm": 2.5858519077301025,
       "learning_rate": 0.00019539266505673938,
-      "loss": 1.4892,
+      "loss": 1.5241,
       "step": 102
     },
     {
       "epoch": 0.04585420144685587,
-      "grad_norm": 2.8571128845214844,
+      "grad_norm": 2.306894063949585,
       "learning_rate": 0.00019529793415172192,
-      "loss": 1.5001,
+      "loss": 1.4665,
       "step": 103
     },
     {
       "epoch": 0.04629938786867001,
-      "grad_norm": 2.330220937728882,
+      "grad_norm": 3.80251407623291,
       "learning_rate": 0.00019520226269456768,
-      "loss": 1.3509,
+      "loss": 1.4046,
       "step": 104
     },
     {
       "epoch": 0.04674457429048414,
-      "grad_norm": 6.893773555755615,
+      "grad_norm": 4.36609411239624,
       "learning_rate": 0.00019510565162951537,
-      "loss": 1.6334,
+      "loss": 1.6537,
       "step": 105
     },
     {
       "epoch": 0.04718976071229827,
-      "grad_norm": NaN,
-      "learning_rate": 0.00019510565162951537,
-      "loss": 1.9998,
+      "grad_norm": 2.7875075340270996,
+      "learning_rate": 0.00019500810191007718,
+      "loss": 1.9127,
       "step": 106
     },
     {
       "epoch": 0.04763494713411241,
-      "grad_norm": 3.612894058227539,
-      "learning_rate": 0.00019500810191007718,
-      "loss": 2.5597,
+      "grad_norm": 3.589874267578125,
+      "learning_rate": 0.00019490961449902946,
+      "loss": 2.6136,
       "step": 107
     },
     {
       "epoch": 0.048080133555926545,
-      "grad_norm": 2.3237907886505127,
-      "learning_rate": 0.00019490961449902946,
-      "loss": 1.6031,
+      "grad_norm": 2.4417803287506104,
+      "learning_rate": 0.0001948101903684032,
+      "loss": 1.6387,
       "step": 108
     },
     {
       "epoch": 0.04852531997774068,
-      "grad_norm": 2.755427122116089,
-      "learning_rate": 0.0001948101903684032,
-      "loss": 1.8851,
+      "grad_norm": 2.994558811187744,
+      "learning_rate": 0.00019470983049947444,
+      "loss": 1.9393,
       "step": 109
     },
     {
       "epoch": 0.04897050639955482,
-      "grad_norm": 2.8905320167541504,
-      "learning_rate": 0.00019470983049947444,
-      "loss": 1.5392,
+      "grad_norm": 2.747941255569458,
+      "learning_rate": 0.00019460853588275454,
+      "loss": 1.5652,
       "step": 110
     },
     {
       "epoch": 0.04941569282136895,
-      "grad_norm": 10.140717506408691,
-      "learning_rate": 0.00019460853588275454,
-      "loss": 2.005,
+      "grad_norm": 2.574636697769165,
+      "learning_rate": 0.00019450630751798048,
+      "loss": 2.0804,
       "step": 111
     },
     {
       "epoch": 0.04986087924318308,
-      "grad_norm": 3.27565336227417,
-      "learning_rate": 0.00019450630751798048,
-      "loss": 2.2091,
+      "grad_norm": 3.1800804138183594,
+      "learning_rate": 0.000194403146414105,
+      "loss": 2.3238,
       "step": 112
     },
     {
       "epoch": 0.05030606566499722,
-      "grad_norm": 4.00056266784668,
-      "learning_rate": 0.000194403146414105,
-      "loss": 3.019,
+      "grad_norm": 4.0952935218811035,
+      "learning_rate": 0.00019429905358928646,
+      "loss": 3.0053,
       "step": 113
     },
     {
       "epoch": 0.05075125208681135,
-      "grad_norm": 3.5745298862457275,
-      "learning_rate": 0.00019429905358928646,
-      "loss": 2.154,
+      "grad_norm": 3.5660924911499023,
+      "learning_rate": 0.00019419403007087907,
+      "loss": 2.1842,
       "step": 114
     },
     {
       "epoch": 0.051196438508625486,
-      "grad_norm": 3.2333099842071533,
-      "learning_rate": 0.00019419403007087907,
-      "loss": 2.1104,
+      "grad_norm": 2.9287869930267334,
+      "learning_rate": 0.00019408807689542257,
+      "loss": 2.1226,
       "step": 115
     },
     {
       "epoch": 0.05164162493043962,
-      "grad_norm": 2.650325298309326,
-      "learning_rate": 0.00019408807689542257,
-      "loss": 1.1883,
+      "grad_norm": 2.3753621578216553,
+      "learning_rate": 0.00019398119510863197,
+      "loss": 1.1418,
       "step": 116
     },
     {
       "epoch": 0.05208681135225376,
-      "grad_norm": 2.167717218399048,
-      "learning_rate": 0.00019398119510863197,
-      "loss": 2.0315,
+      "grad_norm": 2.0147056579589844,
+      "learning_rate": 0.00019387338576538744,
+      "loss": 2.0313,
       "step": 117
     },
     {
       "epoch": 0.05253199777406789,
-      "grad_norm": 3.0747454166412354,
-      "learning_rate": 0.00019387338576538744,
-      "loss": 1.7789,
+      "grad_norm": 2.659571886062622,
+      "learning_rate": 0.00019376464992972356,
+      "loss": 1.6707,
       "step": 118
     },
     {
       "epoch": 0.05297718419588202,
-      "grad_norm": 1.7313055992126465,
-      "learning_rate": 0.00019376464992972356,
-      "loss": 1.4003,
+      "grad_norm": 2.0537705421447754,
+      "learning_rate": 0.00019365498867481923,
+      "loss": 1.4477,
       "step": 119
     },
     {
       "epoch": 0.05342237061769616,
-      "grad_norm": 2.92069149017334,
-      "learning_rate": 0.00019365498867481923,
-      "loss": 1.3967,
+      "grad_norm": 3.3644461631774902,
+      "learning_rate": 0.00019354440308298675,
+      "loss": 1.3865,
       "step": 120
     },
     {
       "epoch": 0.053867557039510294,
-      "grad_norm": 7.5436553955078125,
-      "learning_rate": 0.00019354440308298675,
-      "loss": 1.9777,
+      "grad_norm": 3.3797600269317627,
+      "learning_rate": 0.00019343289424566122,
+      "loss": 2.0211,
       "step": 121
     },
     {
       "epoch": 0.05431274346132443,
-      "grad_norm": 8.587759971618652,
-      "learning_rate": 0.00019343289424566122,
-      "loss": 1.8453,
+      "grad_norm": 5.750209331512451,
+      "learning_rate": 0.00019332046326338986,
+      "loss": 1.7715,
       "step": 122
     },
     {
       "epoch": 0.054757929883138566,
-      "grad_norm": 2.8797378540039062,
-      "learning_rate": 0.00019332046326338986,
-      "loss": 1.8348,
+      "grad_norm": 2.731649398803711,
+      "learning_rate": 0.0001932071112458211,
+      "loss": 1.8292,
       "step": 123
     },
     {
       "epoch": 0.0552031163049527,
-      "grad_norm": 3.496337652206421,
-      "learning_rate": 0.0001932071112458211,
-      "loss": 2.2295,
+      "grad_norm": 3.3839035034179688,
+      "learning_rate": 0.00019309283931169356,
+      "loss": 2.1492,
       "step": 124
     },
     {
       "epoch": 0.05564830272676683,
-      "grad_norm": 4.732133865356445,
-      "learning_rate": 0.00019309283931169356,
-      "loss": 3.2504,
+      "grad_norm": 4.659741401672363,
+      "learning_rate": 0.00019297764858882514,
+      "loss": 3.1081,
       "step": 125
     },
     {
       "epoch": 0.05609348914858097,
-      "grad_norm": 3.367631196975708,
-      "learning_rate": 0.00019297764858882514,
-      "loss": 1.3877,
+      "grad_norm": 2.6617093086242676,
+      "learning_rate": 0.00019286154021410173,
+      "loss": 1.3619,
       "step": 126
     },
     {
       "epoch": 0.0565386755703951,
-      "grad_norm": 3.1331379413604736,
-      "learning_rate": 0.00019286154021410173,
-      "loss": 1.775,
+      "grad_norm": 2.8840413093566895,
+      "learning_rate": 0.00019274451533346615,
+      "loss": 1.7941,
       "step": 127
     },
     {
       "epoch": 0.056983861992209235,
-      "grad_norm": 2.303420305252075,
-      "learning_rate": 0.00019274451533346615,
-      "loss": 2.3335,
+      "grad_norm": 2.5516390800476074,
+      "learning_rate": 0.00019262657510190666,
+      "loss": 2.355,
       "step": 128
     },
     {
       "epoch": 0.057429048414023375,
-      "grad_norm": 3.3306965827941895,
-      "learning_rate": 0.00019262657510190666,
-      "loss": 2.1583,
+      "grad_norm": 2.9875566959381104,
+      "learning_rate": 0.0001925077206834458,
+      "loss": 2.1655,
       "step": 129
     },
     {
       "epoch": 0.05787423483583751,
-      "grad_norm": 2.900506019592285,
-      "learning_rate": 0.0001925077206834458,
-      "loss": 1.7225,
+      "grad_norm": 1.9098056554794312,
+      "learning_rate": 0.0001923879532511287,
+      "loss": 1.6303,
       "step": 130
     },
     {
       "epoch": 0.05831942125765164,
-      "grad_norm": 2.4355456829071045,
-      "learning_rate": 0.0001923879532511287,
-      "loss": 1.4379,
+      "grad_norm": 2.171577215194702,
+      "learning_rate": 0.0001922672739870115,
+      "loss": 1.3926,
       "step": 131
     },
     {
       "epoch": 0.05876460767946578,
-      "grad_norm": 6.534120559692383,
-      "learning_rate": 0.0001922672739870115,
-      "loss": 3.3114,
+      "grad_norm": 4.862812519073486,
+      "learning_rate": 0.00019214568408214985,
+      "loss": 3.0963,
       "step": 132
     },
     {
       "epoch": 0.05920979410127991,
-      "grad_norm": 4.634522914886475,
-      "learning_rate": 0.00019214568408214985,
-      "loss": 3.683,
+      "grad_norm": 4.10323429107666,
+      "learning_rate": 0.00019202318473658705,
+      "loss": 3.8107,
       "step": 133
     },
     {
       "epoch": 0.059654980523094044,
-      "grad_norm": 3.742173671722412,
-      "learning_rate": 0.00019202318473658705,
-      "loss": 2.1639,
+      "grad_norm": 25.305931091308594,
+      "learning_rate": 0.00019189977715934213,
+      "loss": 1.7825,
       "step": 134
     },
     {
       "epoch": 0.06010016694490818,
-      "grad_norm": 3.062386989593506,
-      "learning_rate": 0.00019189977715934213,
-      "loss": 2.2926,
+      "grad_norm": 2.8035287857055664,
+      "learning_rate": 0.00019177546256839812,
+      "loss": 2.1349,
       "step": 135
     },
     {
       "epoch": 0.060545353366722315,
-      "grad_norm": 3.8586795330047607,
-      "learning_rate": 0.00019177546256839812,
-      "loss": 1.788,
+      "grad_norm": 3.733940601348877,
+      "learning_rate": 0.0001916502421906898,
+      "loss": 1.7692,
       "step": 136
     },
     {
       "epoch": 0.06099053978853645,
-      "grad_norm": 3.739783525466919,
-      "learning_rate": 0.0001916502421906898,
-      "loss": 2.7422,
+      "grad_norm": 3.670928955078125,
+      "learning_rate": 0.00019152411726209176,
+      "loss": 2.6832,
       "step": 137
     },
     {
       "epoch": 0.06143572621035059,
-      "grad_norm": 3.0615429878234863,
-      "learning_rate": 0.00019152411726209176,
-      "loss": 2.0941,
+      "grad_norm": 2.6655819416046143,
+      "learning_rate": 0.00019139708902740613,
+      "loss": 2.0142,
       "step": 138
     },
     {
       "epoch": 0.06188091263216472,
-      "grad_norm": 2.7861218452453613,
-      "learning_rate": 0.00019139708902740613,
-      "loss": 1.3494,
+      "grad_norm": 1.9738538265228271,
+      "learning_rate": 0.0001912691587403503,
+      "loss": 1.318,
       "step": 139
     },
     {
       "epoch": 0.06232609905397885,
-      "grad_norm": 4.011098384857178,
-      "learning_rate": 0.0001912691587403503,
-      "loss": 2.6336,
+      "grad_norm": 4.214688777923584,
+      "learning_rate": 0.00019114032766354453,
+      "loss": 2.6407,
       "step": 140
     },
     {
       "epoch": 0.06277128547579298,
-      "grad_norm": 4.350919723510742,
-      "learning_rate": 0.00019114032766354453,
-      "loss": 1.4581,
+      "grad_norm": 3.1077980995178223,
+      "learning_rate": 0.00019101059706849957,
+      "loss": 1.3253,
       "step": 141
     },
     {
       "epoch": 0.06321647189760712,
-      "grad_norm": 3.9159739017486572,
-      "learning_rate": 0.00019101059706849957,
-      "loss": 2.5859,
+      "grad_norm": 3.848151445388794,
+      "learning_rate": 0.00019087996823560402,
+      "loss": 2.6727,
       "step": 142
     },
     {
       "epoch": 0.06366165831942126,
-      "grad_norm": 4.638038158416748,
-      "learning_rate": 0.00019087996823560402,
-      "loss": 2.6958,
+      "grad_norm": 5.2651047706604,
+      "learning_rate": 0.0001907484424541117,
+      "loss": 2.881,
       "step": 143
     },
     {
       "epoch": 0.06410684474123539,
-      "grad_norm": 4.493806838989258,
-      "learning_rate": 0.0001907484424541117,
-      "loss": 1.7282,
+      "grad_norm": 4.984432220458984,
+      "learning_rate": 0.00019061602102212898,
+      "loss": 1.6423,
       "step": 144
     },
     {
       "epoch": 0.06455203116304953,
-      "grad_norm": 3.276454448699951,
-      "learning_rate": 0.00019061602102212898,
-      "loss": 1.4519,
+      "grad_norm": 2.679222583770752,
+      "learning_rate": 0.00019048270524660196,
+      "loss": 1.4025,
       "step": 145
     },
     {
       "epoch": 0.06499721758486367,
-      "grad_norm": 3.375685691833496,
-      "learning_rate": 0.00019048270524660196,
-      "loss": 2.7348,
+      "grad_norm": 3.5055928230285645,
+      "learning_rate": 0.0001903484964433035,
+      "loss": 2.7786,
       "step": 146
     },
     {
       "epoch": 0.06544240400667779,
-      "grad_norm": 3.7552807331085205,
-      "learning_rate": 0.0001903484964433035,
-      "loss": 2.4439,
+      "grad_norm": 4.194420337677002,
+      "learning_rate": 0.00019021339593682028,
+      "loss": 2.4812,
       "step": 147
     },
     {
       "epoch": 0.06588759042849193,
-      "grad_norm": 3.9915504455566406,
-      "learning_rate": 0.00019021339593682028,
-      "loss": 2.7966,
+      "grad_norm": 8.712601661682129,
+      "learning_rate": 0.00019007740506053983,
+      "loss": 2.9453,
       "step": 148
     },
     {
       "epoch": 0.06633277685030607,
-      "grad_norm": 2.8780786991119385,
-      "learning_rate": 0.00019007740506053983,
-      "loss": 1.7375,
+      "grad_norm": 6.733173370361328,
+      "learning_rate": 0.0001899405251566371,
+      "loss": 1.7089,
       "step": 149
     },
     {
       "epoch": 0.0667779632721202,
-      "grad_norm": 2.792412042617798,
-      "learning_rate": 0.0001899405251566371,
-      "loss": 2.4148,
+      "grad_norm": 4.221855163574219,
+      "learning_rate": 0.00018980275757606157,
+      "loss": 2.4643,
       "step": 150
     },
     {
       "epoch": 0.06722314969393434,
-      "grad_norm": 3.3408055305480957,
-      "learning_rate": 0.00018980275757606157,
-      "loss": 2.6953,
+      "grad_norm": 3.792280673980713,
+      "learning_rate": 0.00018966410367852362,
+      "loss": 2.7927,
       "step": 151
     },
     {
       "epoch": 0.06766833611574848,
-      "grad_norm": 2.7032229900360107,
-      "learning_rate": 0.00018966410367852362,
-      "loss": 2.0699,
+      "grad_norm": 2.9503629207611084,
+      "learning_rate": 0.00018952456483248119,
+      "loss": 2.1709,
       "step": 152
     },
     {
       "epoch": 0.0681135225375626,
-      "grad_norm": 3.501795530319214,
-      "learning_rate": 0.00018952456483248119,
-      "loss": 2.6974,
+      "grad_norm": 7.369802474975586,
+      "learning_rate": 0.0001893841424151264,
+      "loss": 2.7544,
       "step": 153
     },
     {
       "epoch": 0.06855870895937674,
-      "grad_norm": 2.4695076942443848,
-      "learning_rate": 0.0001893841424151264,
-      "loss": 1.7339,
+      "grad_norm": 3.224266767501831,
+      "learning_rate": 0.0001892428378123718,
+      "loss": 1.713,
       "step": 154
     },
     {
       "epoch": 0.06900389538119088,
-      "grad_norm": 4.014305591583252,
-      "learning_rate": 0.0001892428378123718,
-      "loss": 2.4761,
+      "grad_norm": 4.188726902008057,
+      "learning_rate": 0.0001891006524188368,
+      "loss": 2.5254,
       "step": 155
     },
     {
       "epoch": 0.069449081803005,
-      "grad_norm": 2.3375473022460938,
-      "learning_rate": 0.0001891006524188368,
-      "loss": 2.0071,
+      "grad_norm": 2.6371772289276123,
+      "learning_rate": 0.00018895758763783383,
+      "loss": 2.0355,
       "step": 156
     },
     {
       "epoch": 0.06989426822481914,
-      "grad_norm": 3.323019504547119,
-      "learning_rate": 0.00018895758763783383,
-      "loss": 2.226,
+      "grad_norm": 4.633547306060791,
+      "learning_rate": 0.00018881364488135448,
+      "loss": 2.2087,
       "step": 157
     },
     {
       "epoch": 0.07033945464663328,
-      "grad_norm": 7.093875885009766,
-      "learning_rate": 0.00018881364488135448,
-      "loss": 1.7551,
+      "grad_norm": 2.663710117340088,
+      "learning_rate": 0.00018866882557005567,
+      "loss": 1.7024,
       "step": 158
     },
     {
       "epoch": 0.07078464106844741,
-      "grad_norm": 2.1790101528167725,
-      "learning_rate": 0.00018866882557005567,
-      "loss": 1.7356,
+      "grad_norm": 2.3200747966766357,
+      "learning_rate": 0.00018852313113324552,
+      "loss": 1.7767,
       "step": 159
     },
     {
       "epoch": 0.07122982749026155,
-      "grad_norm": 4.966728687286377,
-      "learning_rate": 0.00018852313113324552,
-      "loss": 2.3969,
+      "grad_norm": 4.9632248878479,
+      "learning_rate": 0.00018837656300886937,
+      "loss": 2.361,
       "step": 160
     },
     {
       "epoch": 0.07167501391207569,
-      "grad_norm": 2.207064390182495,
-      "learning_rate": 0.00018837656300886937,
-      "loss": 1.9052,
+      "grad_norm": 2.1868953704833984,
+      "learning_rate": 0.00018822912264349534,
+      "loss": 1.8371,
       "step": 161
     },
     {
       "epoch": 0.07212020033388981,
-      "grad_norm": 2.988813877105713,
-      "learning_rate": 0.00018822912264349534,
-      "loss": 2.2362,
+      "grad_norm": 3.481379747390747,
+      "learning_rate": 0.00018808081149230036,
+      "loss": 2.2402,
       "step": 162
     },
     {
       "epoch": 0.07256538675570395,
-      "grad_norm": 15.190692901611328,
-      "learning_rate": 0.00018808081149230036,
-      "loss": 1.7636,
+      "grad_norm": 2.7832183837890625,
+      "learning_rate": 0.00018793163101905563,
+      "loss": 1.7177,
       "step": 163
     },
     {
       "epoch": 0.07301057317751808,
-      "grad_norm": 2.8087406158447266,
-      "learning_rate": 0.00018793163101905563,
-      "loss": 2.0533,
+      "grad_norm": 3.0585930347442627,
+      "learning_rate": 0.00018778158269611218,
+      "loss": 2.0282,
       "step": 164
     },
     {
       "epoch": 0.07345575959933222,
-      "grad_norm": 2.135378122329712,
-      "learning_rate": 0.00018778158269611218,
-      "loss": 2.2528,
+      "grad_norm": 4.76095724105835,
+      "learning_rate": 0.00018763066800438636,
+      "loss": 2.2408,
       "step": 165
     },
     {
       "epoch": 0.07390094602114636,
-      "grad_norm": 3.5980896949768066,
-      "learning_rate": 0.00018763066800438636,
-      "loss": 2.5453,
+      "grad_norm": 3.7105226516723633,
+      "learning_rate": 0.0001874788884333453,
+      "loss": 2.5698,
       "step": 166
     },
     {
       "epoch": 0.07434613244296048,
-      "grad_norm": 2.6481971740722656,
-      "learning_rate": 0.0001874788884333453,
-      "loss": 1.8122,
+      "grad_norm": 3.511364221572876,
+      "learning_rate": 0.00018732624548099204,
+      "loss": 1.7663,
       "step": 167
     },
     {
       "epoch": 0.07479131886477462,
-      "grad_norm": 2.283841371536255,
-      "learning_rate": 0.00018732624548099204,
-      "loss": 2.0774,
+      "grad_norm": 2.9411141872406006,
+      "learning_rate": 0.0001871727406538509,
+      "loss": 2.062,
       "step": 168
     },
     {
       "epoch": 0.07523650528658876,
-      "grad_norm": 2.59877610206604,
-      "learning_rate": 0.0001871727406538509,
-      "loss": 2.0153,
+      "grad_norm": 2.748162269592285,
+      "learning_rate": 0.0001870183754669526,
+      "loss": 1.9645,
       "step": 169
     },
     {
       "epoch": 0.07568169170840289,
-      "grad_norm": 4.261568546295166,
-      "learning_rate": 0.0001870183754669526,
-      "loss": 1.56,
+      "grad_norm": 9.688993453979492,
+      "learning_rate": 0.00018686315144381913,
+      "loss": 1.6173,
       "step": 170
     },
     {
       "epoch": 0.07612687813021703,
-      "grad_norm": 2.588036060333252,
-      "learning_rate": 0.00018686315144381913,
-      "loss": 2.4187,
+      "grad_norm": 4.454281806945801,
+      "learning_rate": 0.000186707070116449,
+      "loss": 2.343,
       "step": 171
     },
     {
       "epoch": 0.07657206455203117,
-      "grad_norm": 3.837876796722412,
-      "learning_rate": 0.000186707070116449,
-      "loss": 2.1299,
+      "grad_norm": 3.597024440765381,
+      "learning_rate": 0.0001865501330253019,
+      "loss": 2.1327,
       "step": 172
     },
     {
       "epoch": 0.07701725097384529,
-      "grad_norm": 2.0875399112701416,
-      "learning_rate": 0.0001865501330253019,
-      "loss": 2.1619,
+      "grad_norm": 6.372188091278076,
+      "learning_rate": 0.00018639234171928353,
+      "loss": 2.3311,
       "step": 173
     },
     {
       "epoch": 0.07746243739565943,
-      "grad_norm": 2.641989231109619,
-      "learning_rate": 0.00018639234171928353,
-      "loss": 1.7574,
+      "grad_norm": 4.005566596984863,
+      "learning_rate": 0.0001862336977557304,
+      "loss": 1.7461,
       "step": 174
     },
     {
       "epoch": 0.07790762381747357,
-      "grad_norm": 3.0749399662017822,
-      "learning_rate": 0.0001862336977557304,
-      "loss": 1.6736,
+      "grad_norm": 5.076218128204346,
+      "learning_rate": 0.0001860742027003944,
+      "loss": 1.5752,
       "step": 175
     },
     {
       "epoch": 0.0783528102392877,
-      "grad_norm": 2.718775749206543,
-      "learning_rate": 0.0001860742027003944,
-      "loss": 2.1462,
+      "grad_norm": 6.041815757751465,
+      "learning_rate": 0.00018591385812742725,
+      "loss": 2.1361,
       "step": 176
     },
     {
       "epoch": 0.07879799666110184,
-      "grad_norm": 2.1774768829345703,
-      "learning_rate": 0.00018591385812742725,
-      "loss": 1.4246,
+      "grad_norm": 3.361299753189087,
+      "learning_rate": 0.00018575266561936523,
+      "loss": 1.3601,
       "step": 177
     },
     {
       "epoch": 0.07924318308291597,
-      "grad_norm": 3.1828486919403076,
-      "learning_rate": 0.00018575266561936523,
-      "loss": 2.3951,
+      "grad_norm": 6.296206951141357,
+      "learning_rate": 0.00018559062676711332,
+      "loss": 2.4282,
       "step": 178
     },
     {
       "epoch": 0.0796883695047301,
-      "grad_norm": 2.520548105239868,
-      "learning_rate": 0.00018559062676711332,
-      "loss": 1.6368,
+      "grad_norm": 3.221632719039917,
+      "learning_rate": 0.0001854277431699295,
+      "loss": 1.5614,
       "step": 179
     },
     {
       "epoch": 0.08013355592654424,
-      "grad_norm": 1.9750266075134277,
-      "learning_rate": 0.0001854277431699295,
-      "loss": 1.4846,
+      "grad_norm": 3.1220271587371826,
+      "learning_rate": 0.00018526401643540922,
+      "loss": 1.3944,
       "step": 180
     },
     {
       "epoch": 0.08057874234835838,
-      "grad_norm": 5.070473670959473,
-      "learning_rate": 0.00018526401643540922,
-      "loss": 1.8334,
+      "grad_norm": 4.852542400360107,
+      "learning_rate": 0.00018509944817946922,
+      "loss": 1.801,
       "step": 181
     },
     {
       "epoch": 0.0810239287701725,
-      "grad_norm": 2.920872688293457,
-      "learning_rate": 0.00018509944817946922,
-      "loss": 1.6212,
+      "grad_norm": 5.415312767028809,
+      "learning_rate": 0.00018493404002633166,
+      "loss": 1.615,
       "step": 182
     },
     {
       "epoch": 0.08146911519198664,
-      "grad_norm": 2.4279935359954834,
-      "learning_rate": 0.00018493404002633166,
-      "loss": 1.7948,
+      "grad_norm": 2.706400156021118,
+      "learning_rate": 0.00018476779360850832,
+      "loss": 1.8279,
       "step": 183
     },
     {
       "epoch": 0.08191430161380078,
-      "grad_norm": 3.453925609588623,
-      "learning_rate": 0.00018476779360850832,
-      "loss": 2.1735,
+      "grad_norm": 4.849134922027588,
+      "learning_rate": 0.00018460071056678422,
+      "loss": 2.0777,
       "step": 184
     },
     {
       "epoch": 0.08235948803561491,
-      "grad_norm": 1.8669699430465698,
-      "learning_rate": 0.00018460071056678422,
-      "loss": 1.2327,
+      "grad_norm": 2.9566752910614014,
+      "learning_rate": 0.00018443279255020152,
+      "loss": 1.2185,
       "step": 185
     },
     {
       "epoch": 0.08280467445742905,
-      "grad_norm": 2.559237003326416,
-      "learning_rate": 0.00018443279255020152,
-      "loss": 1.6827,
+      "grad_norm": 2.846752882003784,
+      "learning_rate": 0.00018426404121604323,
+      "loss": 1.6251,
       "step": 186
     },
     {
       "epoch": 0.08324986087924319,
-      "grad_norm": 3.643465518951416,
-      "learning_rate": 0.00018426404121604323,
-      "loss": 2.3146,
+      "grad_norm": 4.407576084136963,
+      "learning_rate": 0.00018409445822981693,
+      "loss": 2.3741,
       "step": 187
     },
     {
       "epoch": 0.08369504730105731,
-      "grad_norm": 3.2583975791931152,
-      "learning_rate": 0.00018409445822981693,
-      "loss": 1.9232,
+      "grad_norm": 2.320455551147461,
+      "learning_rate": 0.00018392404526523817,
+      "loss": 1.8316,
       "step": 188
     },
     {
       "epoch": 0.08414023372287145,
-      "grad_norm": 3.837696075439453,
-      "learning_rate": 0.00018392404526523817,
-      "loss": 2.215,
+      "grad_norm": 5.207255840301514,
+      "learning_rate": 0.0001837528040042142,
+      "loss": 2.1853,
       "step": 189
     },
     {
       "epoch": 0.08458542014468559,
-      "grad_norm": 2.9309654235839844,
-      "learning_rate": 0.0001837528040042142,
-      "loss": 1.9777,
+      "grad_norm": 2.795053720474243,
+      "learning_rate": 0.00018358073613682706,
+      "loss": 1.9838,
       "step": 190
     },
     {
       "epoch": 0.08503060656649972,
-      "grad_norm": 3.0458357334136963,
-      "learning_rate": 0.00018358073613682706,
-      "loss": 2.0955,
+      "grad_norm": 2.9807851314544678,
+      "learning_rate": 0.00018340784336131713,
+      "loss": 2.0224,
       "step": 191
     },
     {
       "epoch": 0.08547579298831386,
-      "grad_norm": 3.166865825653076,
-      "learning_rate": 0.00018340784336131713,
-      "loss": 1.3475,
+      "grad_norm": 3.149538278579712,
+      "learning_rate": 0.00018323412738406635,
+      "loss": 1.3808,
       "step": 192
     },
     {
       "epoch": 0.085920979410128,
-      "grad_norm": 3.3954527378082275,
-      "learning_rate": 0.00018323412738406635,
-      "loss": 2.514,
+      "grad_norm": 2.727731704711914,
+      "learning_rate": 0.00018305958991958127,
+      "loss": 2.4561,
       "step": 193
     },
     {
       "epoch": 0.08636616583194212,
-      "grad_norm": 3.6221113204956055,
-      "learning_rate": 0.00018305958991958127,
-      "loss": 2.1971,
+      "grad_norm": 3.7259228229522705,
+      "learning_rate": 0.0001828842326904762,
+      "loss": 2.2377,
       "step": 194
     },
     {
       "epoch": 0.08681135225375626,
-      "grad_norm": 2.5096681118011475,
-      "learning_rate": 0.0001828842326904762,
-      "loss": 1.988,
+      "grad_norm": 2.473024368286133,
+      "learning_rate": 0.00018270805742745617,
+      "loss": 2.0065,
       "step": 195
     },
     {
       "epoch": 0.0872565386755704,
-      "grad_norm": 2.7260031700134277,
-      "learning_rate": 0.00018270805742745617,
-      "loss": 1.7442,
+      "grad_norm": 2.7980408668518066,
+      "learning_rate": 0.00018253106586929997,
+      "loss": 1.7422,
       "step": 196
     },
     {
       "epoch": 0.08770172509738453,
-      "grad_norm": 3.155184507369995,
-      "learning_rate": 0.00018253106586929997,
-      "loss": 2.5275,
+      "grad_norm": 2.401157855987549,
+      "learning_rate": 0.00018235325976284275,
+      "loss": 2.485,
       "step": 197
     },
     {
       "epoch": 0.08814691151919866,
-      "grad_norm": 2.810918092727661,
-      "learning_rate": 0.00018235325976284275,
-      "loss": 2.2599,
+      "grad_norm": 2.2364661693573,
+      "learning_rate": 0.00018217464086295904,
+      "loss": 2.1849,
       "step": 198
     },
     {
       "epoch": 0.0885920979410128,
-      "grad_norm": 3.6741535663604736,
-      "learning_rate": 0.00018217464086295904,
-      "loss": 2.4509,
+      "grad_norm": 3.2548766136169434,
+      "learning_rate": 0.00018199521093254523,
+      "loss": 2.3613,
       "step": 199
     },
     {
       "epoch": 0.08903728436282693,
-      "grad_norm": 3.171786069869995,
-      "learning_rate": 0.00018199521093254523,
-      "loss": 1.5725,
+      "grad_norm": 3.292804718017578,
+      "learning_rate": 0.00018181497174250236,
+      "loss": 1.6004,
       "step": 200
     },
     {
       "epoch": 0.08948247078464107,
-      "grad_norm": 5.2465500831604,
-      "learning_rate": 0.00018181497174250236,
-      "loss": 2.1309,
+      "grad_norm": 7.338946342468262,
+      "learning_rate": 0.00018163392507171842,
+      "loss": 2.1327,
       "step": 201
     },
     {
       "epoch": 0.08992765720645521,
-      "grad_norm": 3.3447489738464355,
-      "learning_rate": 0.00018163392507171842,
-      "loss": 2.7054,
+      "grad_norm": 3.515322685241699,
+      "learning_rate": 0.00018145207270705096,
+      "loss": 2.8209,
       "step": 202
     },
     {
       "epoch": 0.09037284362826933,
-      "grad_norm": 2.2638697624206543,
-      "learning_rate": 0.00018145207270705096,
-      "loss": 1.6347,
+      "grad_norm": 2.8458731174468994,
+      "learning_rate": 0.0001812694164433094,
+      "loss": 1.7502,
       "step": 203
     },
     {
       "epoch": 0.09081803005008347,
-      "grad_norm": 2.3262786865234375,
-      "learning_rate": 0.0001812694164433094,
-      "loss": 1.9896,
+      "grad_norm": 2.307251214981079,
+      "learning_rate": 0.00018108595808323736,
+      "loss": 1.9572,
       "step": 204
     },
     {
       "epoch": 0.09126321647189761,
-      "grad_norm": 2.7014877796173096,
-      "learning_rate": 0.00018108595808323736,
-      "loss": 2.3665,
+      "grad_norm": 2.826714515686035,
+      "learning_rate": 0.00018090169943749476,
+      "loss": 2.3473,
       "step": 205
     },
     {
       "epoch": 0.09170840289371174,
-      "grad_norm": 2.7195963859558105,
-      "learning_rate": 0.00018090169943749476,
-      "loss": 2.0004,
+      "grad_norm": 2.7337214946746826,
+      "learning_rate": 0.00018071664232464002,
+      "loss": 1.9562,
       "step": 206
     },
     {
       "epoch": 0.09215358931552588,
-      "grad_norm": 3.5895092487335205,
-      "learning_rate": 0.00018071664232464002,
-      "loss": 2.7216,
+      "grad_norm": 4.373119831085205,
+      "learning_rate": 0.0001805307885711122,
+      "loss": 2.868,
       "step": 207
     },
     {
       "epoch": 0.09259877573734002,
-      "grad_norm": 6.0607008934021,
-      "learning_rate": 0.0001805307885711122,
-      "loss": 1.8817,
+      "grad_norm": 2.356687307357788,
+      "learning_rate": 0.00018034414001121278,
+      "loss": 1.8623,
       "step": 208
     },
     {
       "epoch": 0.09304396215915414,
-      "grad_norm": 3.0178444385528564,
-      "learning_rate": 0.00018034414001121278,
-      "loss": 1.9194,
+      "grad_norm": 2.2096173763275146,
+      "learning_rate": 0.00018015669848708767,
+      "loss": 1.9387,
       "step": 209
     },
     {
       "epoch": 0.09348914858096828,
-      "grad_norm": 3.298114776611328,
-      "learning_rate": 0.00018015669848708767,
-      "loss": 3.2093,
+      "grad_norm": 3.289705276489258,
+      "learning_rate": 0.00017996846584870908,
+      "loss": 3.1561,
       "step": 210
     },
     {
       "epoch": 0.09393433500278242,
-      "grad_norm": 2.1203761100769043,
-      "learning_rate": 0.00017996846584870908,
-      "loss": 1.5833,
+      "grad_norm": 2.1223509311676025,
+      "learning_rate": 0.0001797794439538571,
+      "loss": 1.5744,
       "step": 211
     },
     {
       "epoch": 0.09437952142459655,
-      "grad_norm": 2.907470464706421,
-      "learning_rate": 0.0001797794439538571,
-      "loss": 1.7891,
+      "grad_norm": 2.5752720832824707,
+      "learning_rate": 0.0001795896346681016,
+      "loss": 1.782,
       "step": 212
     },
     {
       "epoch": 0.09482470784641069,
-      "grad_norm": 3.151555061340332,
-      "learning_rate": 0.0001795896346681016,
-      "loss": 2.0487,
+      "grad_norm": 5.218358993530273,
+      "learning_rate": 0.00017939903986478355,
+      "loss": 2.1056,
       "step": 213
     },
     {
       "epoch": 0.09526989426822483,
-      "grad_norm": 3.080402374267578,
-      "learning_rate": 0.00017939903986478355,
-      "loss": 1.7114,
+      "grad_norm": 2.5816993713378906,
+      "learning_rate": 0.00017920766142499672,
+      "loss": 1.6919,
       "step": 214
     },
     {
       "epoch": 0.09571508069003895,
-      "grad_norm": 4.0168561935424805,
-      "learning_rate": 0.00017920766142499672,
-      "loss": 2.5786,
+      "grad_norm": 6.3435564041137695,
+      "learning_rate": 0.00017901550123756906,
+      "loss": 2.6954,
       "step": 215
     },
     {
       "epoch": 0.09616026711185309,
-      "grad_norm": 1.990530252456665,
-      "learning_rate": 0.00017901550123756906,
-      "loss": 2.1145,
+      "grad_norm": 2.04316782951355,
+      "learning_rate": 0.00017882256119904403,
+      "loss": 2.1028,
       "step": 216
     },
     {
       "epoch": 0.09660545353366723,
-      "grad_norm": 2.382887840270996,
-      "learning_rate": 0.00017882256119904403,
-      "loss": 1.8551,
+      "grad_norm": 5.312159538269043,
+      "learning_rate": 0.00017862884321366188,
+      "loss": 1.8883,
       "step": 217
     },
     {
       "epoch": 0.09705063995548135,
-      "grad_norm": 4.134660243988037,
-      "learning_rate": 0.00017862884321366188,
-      "loss": 2.4086,
+      "grad_norm": 4.3344926834106445,
+      "learning_rate": 0.000178434349193341,
+      "loss": 2.4914,
       "step": 218
     },
     {
       "epoch": 0.0974958263772955,
-      "grad_norm": 2.927734375,
-      "learning_rate": 0.000178434349193341,
-      "loss": 1.9386,
+      "grad_norm": 3.5837326049804688,
+      "learning_rate": 0.0001782390810576588,
+      "loss": 1.9651,
       "step": 219
     },
     {
       "epoch": 0.09794101279910963,
-      "grad_norm": 2.8078081607818604,
-      "learning_rate": 0.0001782390810576588,
-      "loss": 2.0586,
+      "grad_norm": 2.7920126914978027,
+      "learning_rate": 0.000178043040733833,
+      "loss": 1.9839,
       "step": 220
     },
     {
       "epoch": 0.09838619922092376,
-      "grad_norm": 4.8399248123168945,
-      "learning_rate": 0.000178043040733833,
-      "loss": 1.3215,
+      "grad_norm": 1.7593663930892944,
+      "learning_rate": 0.00017784623015670238,
+      "loss": 1.2661,
       "step": 221
     },
     {
       "epoch": 0.0988313856427379,
-      "grad_norm": 2.1833369731903076,
-      "learning_rate": 0.00017784623015670238,
-      "loss": 1.8898,
+      "grad_norm": 2.281625747680664,
+      "learning_rate": 0.00017764865126870786,
+      "loss": 1.9069,
       "step": 222
     },
     {
       "epoch": 0.09927657206455204,
-      "grad_norm": 3.099783420562744,
-      "learning_rate": 0.00017764865126870786,
-      "loss": 2.9137,
+      "grad_norm": 3.2211875915527344,
+      "learning_rate": 0.00017745030601987337,
+      "loss": 2.9048,
       "step": 223
     },
     {
       "epoch": 0.09972175848636616,
-      "grad_norm": 2.1298272609710693,
-      "learning_rate": 0.00017745030601987337,
-      "loss": 1.4856,
+      "grad_norm": 2.8080756664276123,
+      "learning_rate": 0.00017725119636778644,
+      "loss": 1.507,
       "step": 224
     },
     {
       "epoch": 0.1001669449081803,
-      "grad_norm": 2.0361313819885254,
-      "learning_rate": 0.00017725119636778644,
-      "loss": 1.8,
+      "grad_norm": 2.2004809379577637,
+      "learning_rate": 0.00017705132427757895,
+      "loss": 1.9243,
       "step": 225
     },
     {
       "epoch": 0.10061213132999444,
-      "grad_norm": 2.4032180309295654,
-      "learning_rate": 0.00017705132427757895,
-      "loss": 1.8565,
+      "grad_norm": 2.086294174194336,
+      "learning_rate": 0.00017685069172190766,
+      "loss": 1.7783,
       "step": 226
     },
     {
       "epoch": 0.10105731775180857,
-      "grad_norm": 3.0993106365203857,
-      "learning_rate": 0.00017685069172190766,
-      "loss": 2.5554,
+      "grad_norm": 3.3997292518615723,
+      "learning_rate": 0.00017664930068093498,
+      "loss": 2.6626,
       "step": 227
     },
     {
       "epoch": 0.1015025041736227,
-      "grad_norm": 4.992886066436768,
-      "learning_rate": 0.00017664930068093498,
-      "loss": 3.3477,
+      "grad_norm": 4.921217918395996,
+      "learning_rate": 0.00017644715314230918,
+      "loss": 3.3393,
       "step": 228
     },
     {
       "epoch": 0.10194769059543683,
-      "grad_norm": 3.065136194229126,
-      "learning_rate": 0.00017644715314230918,
-      "loss": 1.785,
+      "grad_norm": 3.048692464828491,
+      "learning_rate": 0.0001762442511011448,
+      "loss": 1.7612,
       "step": 229
     },
     {
       "epoch": 0.10239287701725097,
-      "grad_norm": 2.915815830230713,
-      "learning_rate": 0.0001762442511011448,
-      "loss": 1.7885,
+      "grad_norm": 2.9201133251190186,
+      "learning_rate": 0.0001760405965600031,
+      "loss": 1.8635,
       "step": 230
     },
     {
       "epoch": 0.10283806343906511,
-      "grad_norm": 1.5576674938201904,
-      "learning_rate": 0.0001760405965600031,
-      "loss": 1.2228,
+      "grad_norm": 1.5012266635894775,
+      "learning_rate": 0.0001758361915288722,
+      "loss": 1.2037,
       "step": 231
     },
     {
       "epoch": 0.10328324986087924,
-      "grad_norm": 4.22326135635376,
-      "learning_rate": 0.0001758361915288722,
-      "loss": 1.8784,
+      "grad_norm": 3.065877676010132,
+      "learning_rate": 0.0001756310380251472,
+      "loss": 1.925,
       "step": 232
     },
     {
       "epoch": 0.10372843628269338,
-      "grad_norm": 2.074902296066284,
-      "learning_rate": 0.0001756310380251472,
-      "loss": 2.1881,
+      "grad_norm": 2.101668357849121,
+      "learning_rate": 0.00017542513807361037,
+      "loss": 2.2399,
       "step": 233
     },
     {
       "epoch": 0.10417362270450752,
-      "grad_norm": 3.5856900215148926,
-      "learning_rate": 0.00017542513807361037,
-      "loss": 2.452,
+      "grad_norm": 2.6265454292297363,
+      "learning_rate": 0.00017521849370641114,
+      "loss": 2.4111,
       "step": 234
     },
     {
       "epoch": 0.10461880912632164,
-      "grad_norm": 3.115443229675293,
-      "learning_rate": 0.00017521849370641114,
-      "loss": 1.9735,
+      "grad_norm": 2.498668670654297,
+      "learning_rate": 0.00017501110696304596,
+      "loss": 2.0284,
       "step": 235
     },
     {
       "epoch": 0.10506399554813578,
-      "grad_norm": 4.1047868728637695,
-      "learning_rate": 0.00017501110696304596,
-      "loss": 4.1891,
+      "grad_norm": 3.8684022426605225,
+      "learning_rate": 0.00017480297989033825,
+      "loss": 4.1502,
       "step": 236
     },
     {
       "epoch": 0.10550918196994992,
-      "grad_norm": 2.4047348499298096,
-      "learning_rate": 0.00017480297989033825,
-      "loss": 1.5507,
+      "grad_norm": 2.2905540466308594,
+      "learning_rate": 0.00017459411454241822,
+      "loss": 1.5043,
       "step": 237
     },
     {
       "epoch": 0.10595436839176404,
-      "grad_norm": 2.6233081817626953,
-      "learning_rate": 0.00017459411454241822,
-      "loss": 1.6275,
+      "grad_norm": 2.6878786087036133,
+      "learning_rate": 0.00017438451298070252,
+      "loss": 1.5641,
       "step": 238
     },
     {
       "epoch": 0.10639955481357818,
-      "grad_norm": 2.559443235397339,
-      "learning_rate": 0.00017438451298070252,
-      "loss": 2.1019,
+      "grad_norm": 2.6507771015167236,
+      "learning_rate": 0.00017417417727387394,
+      "loss": 2.0586,
       "step": 239
     },
     {
       "epoch": 0.10684474123539232,
-      "grad_norm": 2.711853265762329,
-      "learning_rate": 0.00017417417727387394,
-      "loss": 1.8006,
+      "grad_norm": 2.4303462505340576,
+      "learning_rate": 0.000173963109497861,
+      "loss": 1.7561,
       "step": 240
     },
     {
       "epoch": 0.10728992765720645,
-      "grad_norm": 2.1308255195617676,
-      "learning_rate": 0.000173963109497861,
-      "loss": 1.0008,
+      "grad_norm": 2.0561587810516357,
+      "learning_rate": 0.0001737513117358174,
+      "loss": 0.9511,
       "step": 241
     },
     {
       "epoch": 0.10773511407902059,
-      "grad_norm": 2.0422074794769287,
-      "learning_rate": 0.0001737513117358174,
-      "loss": 2.1842,
+      "grad_norm": 1.9653527736663818,
+      "learning_rate": 0.0001735387860781016,
+      "loss": 2.142,
       "step": 242
     },
     {
       "epoch": 0.10818030050083473,
-      "grad_norm": 1.971974492073059,
-      "learning_rate": 0.0001735387860781016,
-      "loss": 1.1267,
+      "grad_norm": 1.6377851963043213,
+      "learning_rate": 0.00017332553462225602,
+      "loss": 1.075,
       "step": 243
     },
     {
       "epoch": 0.10862548692264885,
-      "grad_norm": 1.90377676486969,
-      "learning_rate": 0.00017332553462225602,
-      "loss": 1.4106,
+      "grad_norm": 1.703354001045227,
+      "learning_rate": 0.00017311155947298643,
+      "loss": 1.3669,
       "step": 244
     },
     {
       "epoch": 0.10907067334446299,
-      "grad_norm": 6.567409992218018,
-      "learning_rate": 0.00017311155947298643,
-      "loss": 3.1737,
+      "grad_norm": 6.472535133361816,
+      "learning_rate": 0.00017289686274214118,
+      "loss": 3.3305,
       "step": 245
     },
     {
       "epoch": 0.10951585976627713,
-      "grad_norm": 2.6171557903289795,
-      "learning_rate": 0.00017289686274214118,
-      "loss": 2.0543,
+      "grad_norm": 3.062089204788208,
+      "learning_rate": 0.0001726814465486903,
+      "loss": 2.014,
       "step": 246
     },
     {
       "epoch": 0.10996104618809126,
-      "grad_norm": 3.360865354537964,
-      "learning_rate": 0.0001726814465486903,
-      "loss": 3.0316,
+      "grad_norm": 3.3952343463897705,
+      "learning_rate": 0.0001724653130187047,
+      "loss": 3.0298,
       "step": 247
     },
     {
       "epoch": 0.1104062326099054,
-      "grad_norm": 5.862345218658447,
-      "learning_rate": 0.0001724653130187047,
-      "loss": 2.1349,
+      "grad_norm": 2.866058349609375,
+      "learning_rate": 0.00017224846428533499,
+      "loss": 2.182,
       "step": 248
     },
     {
       "epoch": 0.11085141903171954,
-      "grad_norm": 5.230323314666748,
-      "learning_rate": 0.00017224846428533499,
-      "loss": 1.959,
+      "grad_norm": 3.6851840019226074,
+      "learning_rate": 0.0001720309024887907,
+      "loss": 2.0767,
       "step": 249
     },
     {
       "epoch": 0.11129660545353366,
-      "grad_norm": 1.8790351152420044,
-      "learning_rate": 0.0001720309024887907,
-      "loss": 1.559,
+      "grad_norm": 2.4382998943328857,
+      "learning_rate": 0.00017181262977631888,
+      "loss": 1.5969,
       "step": 250
     },
     {
       "epoch": 0.1117417918753478,
-      "grad_norm": 2.6701748371124268,
-      "learning_rate": 0.00017181262977631888,
-      "loss": 2.3727,
+      "grad_norm": 3.0784506797790527,
+      "learning_rate": 0.00017159364830218312,
+      "loss": 2.485,
       "step": 251
     },
     {
       "epoch": 0.11218697829716194,
-      "grad_norm": 2.2404518127441406,
-      "learning_rate": 0.00017159364830218312,
-      "loss": 1.933,
+      "grad_norm": 2.263517141342163,
+      "learning_rate": 0.00017137396022764214,
+      "loss": 1.9584,
       "step": 252
     },
     {
       "epoch": 0.11263216471897607,
-      "grad_norm": 3.8700356483459473,
-      "learning_rate": 0.00017137396022764214,
-      "loss": 3.0309,
+      "grad_norm": 3.8933019638061523,
+      "learning_rate": 0.00017115356772092857,
+      "loss": 3.008,
       "step": 253
     },
     {
       "epoch": 0.1130773511407902,
-      "grad_norm": 2.1600594520568848,
-      "learning_rate": 0.00017115356772092857,
-      "loss": 1.9069,
+      "grad_norm": 1.9795575141906738,
+      "learning_rate": 0.0001709324729572274,
+      "loss": 1.9471,
       "step": 254
     },
     {
       "epoch": 0.11352253756260434,
-      "grad_norm": 3.375443935394287,
-      "learning_rate": 0.0001709324729572274,
-      "loss": 2.4507,
+      "grad_norm": 6.0236639976501465,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 2.4301,
       "step": 255
     },
     {
       "epoch": 0.11396772398441847,
-      "grad_norm": 2.818181037902832,
-      "learning_rate": 0.00017071067811865476,
-      "loss": 1.9793,
+      "grad_norm": 2.843769073486328,
+      "learning_rate": 0.00017048818539423615,
+      "loss": 1.948,
       "step": 256
     },
     {
       "epoch": 0.11441291040623261,
-      "grad_norm": 3.2607429027557373,
-      "learning_rate": 0.00017048818539423615,
-      "loss": 1.9051,
+      "grad_norm": 3.080153703689575,
+      "learning_rate": 0.00017026499697988493,
+      "loss": 1.8616,
       "step": 257
     },
     {
       "epoch": 0.11485809682804675,
-      "grad_norm": 2.977018117904663,
-      "learning_rate": 0.00017026499697988493,
-      "loss": 1.0261,
+      "grad_norm": 2.4120187759399414,
+      "learning_rate": 0.00017004111507838064,
+      "loss": 0.9854,
       "step": 258
     },
     {
       "epoch": 0.11530328324986087,
-      "grad_norm": 2.8753457069396973,
-      "learning_rate": 0.00017004111507838064,
-      "loss": 1.4664,
+      "grad_norm": 2.5840303897857666,
+      "learning_rate": 0.00016981654189934727,
+      "loss": 1.4495,
       "step": 259
     },
     {
       "epoch": 0.11574846967167501,
-      "grad_norm": 3.7540669441223145,
-      "learning_rate": 0.00016981654189934727,
-      "loss": 2.4403,
+      "grad_norm": 3.637345790863037,
+      "learning_rate": 0.00016959127965923142,
+      "loss": 2.3222,
       "step": 260
     },
     {
       "epoch": 0.11619365609348915,
-      "grad_norm": 2.8768951892852783,
-      "learning_rate": 0.00016959127965923142,
-      "loss": 1.925,
+      "grad_norm": 4.117860794067383,
+      "learning_rate": 0.0001693653305812805,
+      "loss": 2.0434,
       "step": 261
     },
     {
       "epoch": 0.11663884251530328,
-      "grad_norm": 3.1603147983551025,
-      "learning_rate": 0.0001693653305812805,
-      "loss": 2.3041,
+      "grad_norm": 4.139839172363281,
+      "learning_rate": 0.00016913869689552064,
+      "loss": 2.3411,
       "step": 262
     },
     {
       "epoch": 0.11708402893711742,
-      "grad_norm": 3.66261625289917,
-      "learning_rate": 0.00016913869689552064,
-      "loss": 2.9761,
+      "grad_norm": 3.67438006401062,
+      "learning_rate": 0.00016891138083873487,
+      "loss": 2.9323,
       "step": 263
     },
     {
       "epoch": 0.11752921535893156,
-      "grad_norm": 5.0103020668029785,
-      "learning_rate": 0.00016891138083873487,
-      "loss": 2.4806,
+      "grad_norm": 3.891420364379883,
+      "learning_rate": 0.00016868338465444085,
+      "loss": 2.2782,
       "step": 264
     },
     {
       "epoch": 0.11797440178074568,
-      "grad_norm": 3.2384135723114014,
-      "learning_rate": 0.00016868338465444085,
-      "loss": 2.8883,
+      "grad_norm": 5.930166721343994,
+      "learning_rate": 0.00016845471059286887,
+      "loss": 2.8448,
       "step": 265
     },
     {
       "epoch": 0.11841958820255982,
-      "grad_norm": 3.064087390899658,
-      "learning_rate": 0.00016845471059286887,
-      "loss": 2.3382,
+      "grad_norm": 2.611166000366211,
+      "learning_rate": 0.00016822536091093965,
+      "loss": 2.2242,
       "step": 266
     },
     {
       "epoch": 0.11886477462437396,
-      "grad_norm": 3.015291452407837,
-      "learning_rate": 0.00016822536091093965,
-      "loss": 2.1407,
+      "grad_norm": 2.925534248352051,
+      "learning_rate": 0.00016799533787224192,
+      "loss": 2.1772,
       "step": 267
     },
     {
       "epoch": 0.11930996104618809,
-      "grad_norm": 3.0050747394561768,
-      "learning_rate": 0.00016799533787224192,
-      "loss": 3.0019,
+      "grad_norm": 2.9522793292999268,
+      "learning_rate": 0.00016776464374701025,
+      "loss": 3.0509,
       "step": 268
     },
     {
       "epoch": 0.11975514746800223,
-      "grad_norm": 3.889848470687866,
-      "learning_rate": 0.00016776464374701025,
-      "loss": 2.9921,
+      "grad_norm": 4.202291965484619,
+      "learning_rate": 0.00016753328081210245,
+      "loss": 2.9649,
       "step": 269
     },
     {
       "epoch": 0.12020033388981637,
-      "grad_norm": 3.5669052600860596,
-      "learning_rate": 0.00016753328081210245,
-      "loss": 1.9408,
+      "grad_norm": 3.8469340801239014,
+      "learning_rate": 0.00016730125135097735,
+      "loss": 1.9204,
       "step": 270
     },
     {
       "epoch": 0.12064552031163049,
-      "grad_norm": 2.8057496547698975,
-      "learning_rate": 0.00016730125135097735,
-      "loss": 2.0,
+      "grad_norm": 3.0598490238189697,
+      "learning_rate": 0.000167068557653672,
+      "loss": 1.9098,
       "step": 271
     },
     {
       "epoch": 0.12109070673344463,
-      "grad_norm": 3.1653785705566406,
-      "learning_rate": 0.000167068557653672,
-      "loss": 2.188,
+      "grad_norm": 3.7417612075805664,
+      "learning_rate": 0.0001668352020167793,
+      "loss": 2.2238,
       "step": 272
     },
     {
       "epoch": 0.12153589315525877,
-      "grad_norm": 2.8963727951049805,
-      "learning_rate": 0.0001668352020167793,
-      "loss": 1.8975,
+      "grad_norm": 3.315929651260376,
+      "learning_rate": 0.00016660118674342517,
+      "loss": 1.9491,
       "step": 273
     },
     {
       "epoch": 0.1219810795770729,
-      "grad_norm": 2.387599468231201,
-      "learning_rate": 0.00016660118674342517,
-      "loss": 1.5727,
+      "grad_norm": 2.337838888168335,
+      "learning_rate": 0.00016636651414324587,
+      "loss": 1.618,
       "step": 274
     },
     {
       "epoch": 0.12242626599888703,
-      "grad_norm": 2.26202654838562,
-      "learning_rate": 0.00016636651414324587,
-      "loss": 1.0566,
+      "grad_norm": 2.5510923862457275,
+      "learning_rate": 0.00016613118653236518,
+      "loss": 1.0794,
       "step": 275
     },
     {
       "epoch": 0.12287145242070117,
-      "grad_norm": 2.7368710041046143,
-      "learning_rate": 0.00016613118653236518,
-      "loss": 2.203,
+      "grad_norm": 2.8431742191314697,
+      "learning_rate": 0.0001658952062333717,
+      "loss": 2.2171,
       "step": 276
     },
     {
       "epoch": 0.1233166388425153,
-      "grad_norm": 3.0989296436309814,
-      "learning_rate": 0.0001658952062333717,
-      "loss": 2.1975,
+      "grad_norm": 2.607703447341919,
+      "learning_rate": 0.00016565857557529566,
+      "loss": 2.1593,
       "step": 277
     },
     {
       "epoch": 0.12376182526432944,
-      "grad_norm": 2.62036395072937,
-      "learning_rate": 0.00016565857557529566,
-      "loss": 1.5915,
+      "grad_norm": 2.661377191543579,
+      "learning_rate": 0.00016542129689358612,
+      "loss": 1.6088,
       "step": 278
     },
     {
       "epoch": 0.12420701168614358,
-      "grad_norm": 2.7251317501068115,
-      "learning_rate": 0.00016542129689358612,
-      "loss": 1.9822,
+      "grad_norm": 4.326172828674316,
+      "learning_rate": 0.0001651833725300879,
+      "loss": 1.9431,
       "step": 279
     },
     {
       "epoch": 0.1246521981079577,
-      "grad_norm": 2.146713972091675,
-      "learning_rate": 0.0001651833725300879,
-      "loss": 1.5296,
+      "grad_norm": 2.104480266571045,
+      "learning_rate": 0.00016494480483301836,
+      "loss": 1.508,
       "step": 280
     },
     {
       "epoch": 0.12509738452977184,
-      "grad_norm": 2.940171718597412,
-      "learning_rate": 0.00016494480483301836,
-      "loss": 1.4362,
+      "grad_norm": 2.470026731491089,
+      "learning_rate": 0.00016470559615694446,
+      "loss": 1.3881,
       "step": 281
     },
     {
       "epoch": 0.12554257095158597,
-      "grad_norm": 5.681768417358398,
-      "learning_rate": 0.00016470559615694446,
-      "loss": 2.3687,
+      "grad_norm": 4.4281463623046875,
+      "learning_rate": 0.00016446574886275913,
+      "loss": 2.4574,
       "step": 282
     },
     {
       "epoch": 0.12598775737340012,
-      "grad_norm": 3.23738169670105,
-      "learning_rate": 0.00016446574886275913,
-      "loss": 2.8589,
+      "grad_norm": 3.3189570903778076,
+      "learning_rate": 0.00016422526531765846,
+      "loss": 2.8489,
       "step": 283
     },
     {
       "epoch": 0.12643294379521425,
-      "grad_norm": 4.700006008148193,
-      "learning_rate": 0.00016422526531765846,
-      "loss": 2.2194,
+      "grad_norm": 4.1713995933532715,
+      "learning_rate": 0.00016398414789511786,
+      "loss": 2.2512,
       "step": 284
     },
     {
       "epoch": 0.12687813021702837,
-      "grad_norm": 4.28733491897583,
-      "learning_rate": 0.00016398414789511786,
-      "loss": 3.1353,
+      "grad_norm": 4.102217674255371,
+      "learning_rate": 0.000163742398974869,
+      "loss": 3.077,
       "step": 285
     },
     {
       "epoch": 0.12732331663884253,
-      "grad_norm": 2.7851359844207764,
-      "learning_rate": 0.000163742398974869,
-      "loss": 1.5339,
+      "grad_norm": 2.3833987712860107,
+      "learning_rate": 0.00016350002094287609,
+      "loss": 1.5402,
       "step": 286
     },
     {
       "epoch": 0.12776850306065665,
-      "grad_norm": 4.9445624351501465,
-      "learning_rate": 0.00016350002094287609,
-      "loss": 1.8456,
+      "grad_norm": 5.322443008422852,
+      "learning_rate": 0.00016325701619131246,
+      "loss": 1.7873,
       "step": 287
     },
     {
       "epoch": 0.12821368948247078,
-      "grad_norm": 3.781156539916992,
-      "learning_rate": 0.00016325701619131246,
-      "loss": 1.9967,
+      "grad_norm": 3.7488527297973633,
+      "learning_rate": 0.00016301338711853693,
+      "loss": 2.1434,
       "step": 288
     },
     {
       "epoch": 0.12865887590428493,
-      "grad_norm": 4.377411842346191,
-      "learning_rate": 0.00016301338711853693,
-      "loss": 3.0335,
+      "grad_norm": 4.739319801330566,
+      "learning_rate": 0.00016276913612907007,
+      "loss": 3.1044,
       "step": 289
     },
     {
       "epoch": 0.12910406232609906,
-      "grad_norm": 3.9102425575256348,
-      "learning_rate": 0.00016276913612907007,
-      "loss": 2.9909,
+      "grad_norm": 3.856187582015991,
+      "learning_rate": 0.00016252426563357055,
+      "loss": 2.9997,
       "step": 290
     },
     {
       "epoch": 0.12954924874791318,
-      "grad_norm": 4.036067485809326,
-      "learning_rate": 0.00016252426563357055,
-      "loss": 2.1329,
+      "grad_norm": 3.8840365409851074,
+      "learning_rate": 0.00016227877804881127,
+      "loss": 2.1057,
       "step": 291
     },
     {
       "epoch": 0.12999443516972733,
-      "grad_norm": 2.960005283355713,
-      "learning_rate": 0.00016227877804881127,
-      "loss": 1.7441,
+      "grad_norm": 3.2177507877349854,
+      "learning_rate": 0.00016203267579765563,
+      "loss": 1.6866,
       "step": 292
     },
     {
       "epoch": 0.13043962159154146,
-      "grad_norm": 1.907597303390503,
-      "learning_rate": 0.00016203267579765563,
-      "loss": 1.3116,
+      "grad_norm": 2.5322701930999756,
+      "learning_rate": 0.00016178596130903344,
+      "loss": 1.3248,
       "step": 293
     },
     {
       "epoch": 0.13088480801335559,
-      "grad_norm": 3.022691249847412,
-      "learning_rate": 0.00016178596130903344,
-      "loss": 2.5592,
+      "grad_norm": 2.921722173690796,
+      "learning_rate": 0.00016153863701791717,
+      "loss": 2.4703,
       "step": 294
     },
     {
       "epoch": 0.13132999443516974,
-      "grad_norm": 4.502274990081787,
-      "learning_rate": 0.00016153863701791717,
-      "loss": 2.3546,
+      "grad_norm": 5.355931282043457,
+      "learning_rate": 0.00016129070536529766,
+      "loss": 2.4131,
       "step": 295
     },
     {
       "epoch": 0.13177518085698386,
-      "grad_norm": 3.090162754058838,
-      "learning_rate": 0.00016129070536529766,
-      "loss": 2.0902,
+      "grad_norm": 3.7854509353637695,
+      "learning_rate": 0.00016104216879816026,
+      "loss": 2.1055,
       "step": 296
     },
     {
       "epoch": 0.132220367278798,
-      "grad_norm": 2.6545863151550293,
-      "learning_rate": 0.00016104216879816026,
-      "loss": 1.932,
+      "grad_norm": 2.6546759605407715,
+      "learning_rate": 0.00016079302976946055,
+      "loss": 1.9531,
       "step": 297
     },
     {
       "epoch": 0.13266555370061214,
-      "grad_norm": 2.626149892807007,
-      "learning_rate": 0.00016079302976946055,
-      "loss": 1.9929,
+      "grad_norm": 3.101888418197632,
+      "learning_rate": 0.00016054329073810015,
+      "loss": 2.0063,
       "step": 298
     },
     {
       "epoch": 0.13311074012242627,
-      "grad_norm": 2.5994675159454346,
-      "learning_rate": 0.00016054329073810015,
-      "loss": 2.0321,
+      "grad_norm": 21.36931610107422,
+      "learning_rate": 0.00016029295416890248,
+      "loss": 2.0639,
       "step": 299
     },
     {
       "epoch": 0.1335559265442404,
-      "grad_norm": 2.5016086101531982,
-      "learning_rate": 0.00016029295416890248,
-      "loss": 1.6849,
+      "grad_norm": 2.3509132862091064,
+      "learning_rate": 0.00016004202253258842,
+      "loss": 1.6333,
       "step": 300
     },
     {
       "epoch": 0.13400111296605455,
-      "grad_norm": 2.2394087314605713,
-      "learning_rate": 0.00016004202253258842,
-      "loss": 1.6157,
+      "grad_norm": 2.153367280960083,
+      "learning_rate": 0.0001597904983057519,
+      "loss": 1.6092,
       "step": 301
     },
     {
       "epoch": 0.13444629938786867,
-      "grad_norm": 2.9624788761138916,
-      "learning_rate": 0.0001597904983057519,
-      "loss": 1.0671,
+      "grad_norm": 2.392821788787842,
+      "learning_rate": 0.00015953838397083552,
+      "loss": 1.0187,
       "step": 302
     },
     {
       "epoch": 0.1348914858096828,
-      "grad_norm": 3.160810708999634,
-      "learning_rate": 0.00015953838397083552,
-      "loss": 2.1351,
+      "grad_norm": 3.2482519149780273,
+      "learning_rate": 0.00015928568201610595,
+      "loss": 2.0845,
       "step": 303
     },
     {
       "epoch": 0.13533667223149695,
-      "grad_norm": 2.8223936557769775,
-      "learning_rate": 0.00015928568201610595,
-      "loss": 1.9108,
+      "grad_norm": 3.762847661972046,
+      "learning_rate": 0.00015903239493562948,
+      "loss": 1.9434,
       "step": 304
     },
     {
       "epoch": 0.13578185865331108,
-      "grad_norm": 2.559141159057617,
-      "learning_rate": 0.00015903239493562948,
-      "loss": 2.1271,
+      "grad_norm": 3.044520378112793,
+      "learning_rate": 0.00015877852522924732,
+      "loss": 2.1382,
       "step": 305
     },
     {
       "epoch": 0.1362270450751252,
-      "grad_norm": 3.396461009979248,
-      "learning_rate": 0.00015877852522924732,
-      "loss": 1.8675,
+      "grad_norm": 3.8854618072509766,
+      "learning_rate": 0.00015852407540255104,
+      "loss": 1.9244,
       "step": 306
     },
     {
       "epoch": 0.13667223149693936,
-      "grad_norm": 3.4334423542022705,
-      "learning_rate": 0.00015852407540255104,
-      "loss": 1.7112,
+      "grad_norm": 4.318282604217529,
+      "learning_rate": 0.00015826904796685762,
+      "loss": 1.7102,
       "step": 307
     },
     {
       "epoch": 0.13711741791875348,
-      "grad_norm": 5.243045330047607,
-      "learning_rate": 0.00015826904796685762,
-      "loss": 1.5088,
+      "grad_norm": 9.81147289276123,
+      "learning_rate": 0.00015801344543918495,
+      "loss": 1.4987,
       "step": 308
     },
     {
       "epoch": 0.1375626043405676,
-      "grad_norm": 2.0283901691436768,
-      "learning_rate": 0.00015801344543918495,
-      "loss": 1.5021,
+      "grad_norm": 2.0692050457000732,
+      "learning_rate": 0.00015775727034222675,
+      "loss": 1.4957,
       "step": 309
     },
     {
       "epoch": 0.13800779076238176,
-      "grad_norm": 2.131378173828125,
-      "learning_rate": 0.00015775727034222675,
-      "loss": 2.0285,
+      "grad_norm": 2.195147752761841,
+      "learning_rate": 0.00015750052520432787,
+      "loss": 1.9884,
       "step": 310
     },
     {
       "epoch": 0.13845297718419589,
-      "grad_norm": 2.8632185459136963,
-      "learning_rate": 0.00015750052520432787,
-      "loss": 2.1884,
+      "grad_norm": 3.409510850906372,
+      "learning_rate": 0.0001572432125594591,
+      "loss": 2.1815,
       "step": 311
     },
     {
       "epoch": 0.13889816360601,
-      "grad_norm": 3.2297849655151367,
-      "learning_rate": 0.0001572432125594591,
-      "loss": 2.0599,
+      "grad_norm": 3.4458372592926025,
+      "learning_rate": 0.00015698533494719238,
+      "loss": 2.1309,
       "step": 312
     },
     {
       "epoch": 0.13934335002782416,
-      "grad_norm": 4.282790660858154,
-      "learning_rate": 0.00015698533494719238,
-      "loss": 2.6433,
+      "grad_norm": 3.9557583332061768,
+      "learning_rate": 0.00015672689491267567,
+      "loss": 2.6572,
       "step": 313
     },
     {
       "epoch": 0.1397885364496383,
-      "grad_norm": 3.296445369720459,
-      "learning_rate": 0.00015672689491267567,
-      "loss": 2.5076,
+      "grad_norm": 3.585172176361084,
+      "learning_rate": 0.00015646789500660773,
+      "loss": 2.5301,
       "step": 314
     },
     {
       "epoch": 0.14023372287145242,
-      "grad_norm": 2.099283456802368,
-      "learning_rate": 0.00015646789500660773,
-      "loss": 1.3551,
+      "grad_norm": 1.9923994541168213,
+      "learning_rate": 0.00015620833778521307,
+      "loss": 1.3647,
       "step": 315
     },
     {
       "epoch": 0.14067890929326657,
-      "grad_norm": 2.6336355209350586,
-      "learning_rate": 0.00015620833778521307,
-      "loss": 1.5616,
+      "grad_norm": 3.0583300590515137,
+      "learning_rate": 0.0001559482258102167,
+      "loss": 1.5729,
       "step": 316
     },
     {
       "epoch": 0.1411240957150807,
-      "grad_norm": 1.6293888092041016,
-      "learning_rate": 0.0001559482258102167,
+      "grad_norm": 1.7110612392425537,
+      "learning_rate": 0.00015568756164881882,
       "loss": 0.9969,
       "step": 317
     },
     {
       "epoch": 0.14156928213689482,
-      "grad_norm": 2.045325517654419,
-      "learning_rate": 0.00015568756164881882,
-      "loss": 1.6796,
+      "grad_norm": 2.0573203563690186,
+      "learning_rate": 0.00015542634787366942,
+      "loss": 1.6916,
       "step": 318
     },
     {
       "epoch": 0.14201446855870897,
-      "grad_norm": 3.8815982341766357,
-      "learning_rate": 0.00015542634787366942,
-      "loss": 1.6373,
+      "grad_norm": 3.113630533218384,
+      "learning_rate": 0.00015516458706284303,
+      "loss": 1.5547,
       "step": 319
     },
     {
       "epoch": 0.1424596549805231,
-      "grad_norm": 2.5085458755493164,
-      "learning_rate": 0.00015516458706284303,
-      "loss": 2.3436,
+      "grad_norm": 2.5844383239746094,
+      "learning_rate": 0.0001549022817998132,
+      "loss": 2.3636,
       "step": 320
     },
     {
       "epoch": 0.14290484140233722,
-      "grad_norm": 2.1215457916259766,
-      "learning_rate": 0.0001549022817998132,
-      "loss": 1.2887,
+      "grad_norm": 2.303222417831421,
+      "learning_rate": 0.00015463943467342693,
+      "loss": 1.3141,
       "step": 321
     },
     {
       "epoch": 0.14335002782415138,
-      "grad_norm": 1.9667388200759888,
-      "learning_rate": 0.00015463943467342693,
-      "loss": 1.7435,
+      "grad_norm": 1.8629870414733887,
+      "learning_rate": 0.00015437604827787927,
+      "loss": 1.6869,
       "step": 322
     },
     {
       "epoch": 0.1437952142459655,
-      "grad_norm": 4.962423324584961,
-      "learning_rate": 0.00015437604827787927,
-      "loss": 3.1438,
+      "grad_norm": 6.568711757659912,
+      "learning_rate": 0.00015411212521268758,
+      "loss": 2.9446,
       "step": 323
     },
     {
       "epoch": 0.14424040066777963,
-      "grad_norm": 3.1294186115264893,
-      "learning_rate": 0.00015411212521268758,
-      "loss": 1.922,
+      "grad_norm": 2.645054817199707,
+      "learning_rate": 0.00015384766808266602,
+      "loss": 1.9217,
       "step": 324
     },
     {
       "epoch": 0.14468558708959378,
-      "grad_norm": 2.8280906677246094,
-      "learning_rate": 0.00015384766808266602,
-      "loss": 1.8233,
+      "grad_norm": 3.1033923625946045,
+      "learning_rate": 0.00015358267949789966,
+      "loss": 1.832,
       "step": 325
     },
     {
       "epoch": 0.1451307735114079,
-      "grad_norm": 3.0242607593536377,
-      "learning_rate": 0.00015358267949789966,
-      "loss": 2.0155,
+      "grad_norm": 3.0402700901031494,
+      "learning_rate": 0.00015331716207371888,
+      "loss": 2.1234,
       "step": 326
     },
     {
       "epoch": 0.14557595993322203,
-      "grad_norm": 1.9289848804473877,
-      "learning_rate": 0.00015331716207371888,
-      "loss": 1.6602,
+      "grad_norm": 2.067523717880249,
+      "learning_rate": 0.0001530511184306734,
+      "loss": 1.6432,
       "step": 327
     },
     {
       "epoch": 0.14602114635503616,
-      "grad_norm": 3.2559800148010254,
-      "learning_rate": 0.0001530511184306734,
-      "loss": 2.0712,
+      "grad_norm": 3.112488269805908,
+      "learning_rate": 0.00015278455119450664,
+      "loss": 2.0268,
       "step": 328
     },
     {
       "epoch": 0.1464663327768503,
-      "grad_norm": 3.1982977390289307,
-      "learning_rate": 0.00015278455119450664,
-      "loss": 1.9866,
+      "grad_norm": 2.8925583362579346,
+      "learning_rate": 0.0001525174629961296,
+      "loss": 1.9262,
       "step": 329
     },
     {
       "epoch": 0.14691151919866444,
-      "grad_norm": 4.816375255584717,
-      "learning_rate": 0.0001525174629961296,
-      "loss": 3.3914,
+      "grad_norm": 4.412671089172363,
+      "learning_rate": 0.0001522498564715949,
+      "loss": 3.3829,
       "step": 330
     },
     {
       "epoch": 0.14735670562047856,
-      "grad_norm": 2.63909912109375,
-      "learning_rate": 0.0001522498564715949,
-      "loss": 2.4218,
+      "grad_norm": 2.726099729537964,
+      "learning_rate": 0.00015198173426207094,
+      "loss": 2.3954,
       "step": 331
     },
     {
       "epoch": 0.14780189204229272,
-      "grad_norm": 2.412900686264038,
-      "learning_rate": 0.00015198173426207094,
-      "loss": 1.9517,
+      "grad_norm": 2.8025686740875244,
+      "learning_rate": 0.00015171309901381572,
+      "loss": 1.9772,
       "step": 332
     },
     {
       "epoch": 0.14824707846410684,
-      "grad_norm": 2.017374038696289,
-      "learning_rate": 0.00015171309901381572,
-      "loss": 1.4106,
+      "grad_norm": 1.8641215562820435,
+      "learning_rate": 0.00015144395337815064,
+      "loss": 1.369,
       "step": 333
     },
     {
       "epoch": 0.14869226488592097,
-      "grad_norm": 3.504751205444336,
-      "learning_rate": 0.00015144395337815064,
-      "loss": 2.372,
+      "grad_norm": 3.443453550338745,
+      "learning_rate": 0.00015117430001143452,
+      "loss": 2.3748,
       "step": 334
     },
     {
       "epoch": 0.14913745130773512,
-      "grad_norm": 3.916456460952759,
-      "learning_rate": 0.00015117430001143452,
-      "loss": 1.9819,
+      "grad_norm": 3.986330986022949,
+      "learning_rate": 0.00015090414157503714,
+      "loss": 1.936,
       "step": 335
     },
     {
       "epoch": 0.14958263772954924,
-      "grad_norm": 1.9360079765319824,
-      "learning_rate": 0.00015090414157503714,
-      "loss": 1.5803,
+      "grad_norm": 1.7998448610305786,
+      "learning_rate": 0.00015063348073531324,
+      "loss": 1.6217,
       "step": 336
     },
     {
       "epoch": 0.15002782415136337,
-      "grad_norm": 4.142049312591553,
-      "learning_rate": 0.00015063348073531324,
-      "loss": 3.1576,
+      "grad_norm": 3.430058717727661,
+      "learning_rate": 0.0001503623201635761,
+      "loss": 2.9871,
       "step": 337
     },
     {
       "epoch": 0.15047301057317752,
-      "grad_norm": 5.079156875610352,
-      "learning_rate": 0.0001503623201635761,
-      "loss": 2.724,
+      "grad_norm": 7.617468357086182,
+      "learning_rate": 0.000150090662536071,
+      "loss": 2.6147,
       "step": 338
     },
     {
       "epoch": 0.15091819699499165,
-      "grad_norm": 2.2369110584259033,
-      "learning_rate": 0.000150090662536071,
-      "loss": 1.2975,
+      "grad_norm": 2.18721079826355,
+      "learning_rate": 0.0001498185105339491,
+      "loss": 1.3139,
       "step": 339
     },
     {
       "epoch": 0.15136338341680577,
-      "grad_norm": 3.156975746154785,
-      "learning_rate": 0.0001498185105339491,
-      "loss": 1.9116,
+      "grad_norm": 3.0415077209472656,
+      "learning_rate": 0.00014954586684324078,
+      "loss": 1.8715,
       "step": 340
     },
     {
       "epoch": 0.15180856983861993,
-      "grad_norm": 2.597970724105835,
-      "learning_rate": 0.00014954586684324078,
-      "loss": 1.4781,
+      "grad_norm": 3.532071113586426,
+      "learning_rate": 0.00014927273415482915,
+      "loss": 1.5293,
       "step": 341
     },
     {
       "epoch": 0.15225375626043405,
-      "grad_norm": 2.6372573375701904,
-      "learning_rate": 0.00014927273415482915,
-      "loss": 2.4297,
+      "grad_norm": 3.2872228622436523,
+      "learning_rate": 0.00014899911516442365,
+      "loss": 2.4616,
       "step": 342
     },
     {
       "epoch": 0.15269894268224818,
-      "grad_norm": 5.823768615722656,
-      "learning_rate": 0.00014899911516442365,
-      "loss": 3.3218,
+      "grad_norm": 7.112374782562256,
+      "learning_rate": 0.00014872501257253323,
+      "loss": 3.3683,
       "step": 343
     },
     {
       "epoch": 0.15314412910406233,
-      "grad_norm": 1.7671197652816772,
-      "learning_rate": 0.00014872501257253323,
-      "loss": 1.5805,
+      "grad_norm": 1.7861616611480713,
+      "learning_rate": 0.0001484504290844398,
+      "loss": 1.5288,
       "step": 344
     },
     {
       "epoch": 0.15358931552587646,
-      "grad_norm": 4.895401954650879,
-      "learning_rate": 0.0001484504290844398,
-      "loss": 2.6022,
+      "grad_norm": 4.398621559143066,
+      "learning_rate": 0.00014817536741017152,
+      "loss": 2.5975,
       "step": 345
     },
     {
       "epoch": 0.15403450194769058,
-      "grad_norm": 2.013131618499756,
-      "learning_rate": 0.00014817536741017152,
-      "loss": 1.9509,
+      "grad_norm": 2.140319347381592,
+      "learning_rate": 0.00014789983026447612,
+      "loss": 1.931,
       "step": 346
     },
     {
       "epoch": 0.15447968836950474,
-      "grad_norm": 4.518467426300049,
-      "learning_rate": 0.00014789983026447612,
-      "loss": 3.4165,
+      "grad_norm": 4.669970512390137,
+      "learning_rate": 0.0001476238203667939,
+      "loss": 3.2845,
       "step": 347
     },
     {
       "epoch": 0.15492487479131886,
-      "grad_norm": 3.0389232635498047,
-      "learning_rate": 0.0001476238203667939,
-      "loss": 2.2441,
+      "grad_norm": 4.6830573081970215,
+      "learning_rate": 0.0001473473404412312,
+      "loss": 2.3253,
       "step": 348
     },
     {
       "epoch": 0.155370061213133,
-      "grad_norm": 2.6705610752105713,
-      "learning_rate": 0.0001473473404412312,
-      "loss": 1.6084,
+      "grad_norm": 2.430469274520874,
+      "learning_rate": 0.0001470703932165333,
+      "loss": 1.5183,
       "step": 349
     },
     {
       "epoch": 0.15581524763494714,
-      "grad_norm": 2.1380245685577393,
-      "learning_rate": 0.0001470703932165333,
-      "loss": 2.0694,
+      "grad_norm": 2.4476468563079834,
+      "learning_rate": 0.00014679298142605734,
+      "loss": 2.0601,
       "step": 350
     },
     {
       "epoch": 0.15626043405676127,
-      "grad_norm": 2.587960720062256,
-      "learning_rate": 0.00014679298142605734,
-      "loss": 1.7345,
+      "grad_norm": 2.8119711875915527,
+      "learning_rate": 0.00014651510780774583,
+      "loss": 1.6978,
       "step": 351
     },
     {
       "epoch": 0.1567056204785754,
-      "grad_norm": 2.9578754901885986,
-      "learning_rate": 0.00014651510780774583,
-      "loss": 2.6031,
+      "grad_norm": 4.684689998626709,
+      "learning_rate": 0.00014623677510409918,
+      "loss": 2.6597,
       "step": 352
     },
     {
       "epoch": 0.15715080690038954,
-      "grad_norm": 2.348881483078003,
-      "learning_rate": 0.00014623677510409918,
-      "loss": 1.448,
+      "grad_norm": 2.430534839630127,
+      "learning_rate": 0.00014595798606214882,
+      "loss": 1.4758,
       "step": 353
     },
     {
       "epoch": 0.15759599332220367,
-      "grad_norm": 1.7157777547836304,
-      "learning_rate": 0.00014595798606214882,
-      "loss": 1.8623,
+      "grad_norm": 1.7102493047714233,
+      "learning_rate": 0.00014567874343342997,
+      "loss": 1.8347,
       "step": 354
     },
     {
       "epoch": 0.1580411797440178,
-      "grad_norm": 2.54980206489563,
-      "learning_rate": 0.00014567874343342997,
-      "loss": 2.263,
+      "grad_norm": 2.9137520790100098,
+      "learning_rate": 0.00014539904997395468,
+      "loss": 2.2442,
       "step": 355
     },
     {
       "epoch": 0.15848636616583195,
-      "grad_norm": 3.329082727432251,
-      "learning_rate": 0.00014539904997395468,
-      "loss": 2.8837,
+      "grad_norm": 4.17578649520874,
+      "learning_rate": 0.00014511890844418453,
+      "loss": 2.8419,
       "step": 356
     },
     {
       "epoch": 0.15893155258764607,
-      "grad_norm": 2.646977186203003,
-      "learning_rate": 0.00014511890844418453,
-      "loss": 2.1387,
+      "grad_norm": 3.7746939659118652,
+      "learning_rate": 0.00014483832160900326,
+      "loss": 2.0522,
       "step": 357
     },
     {
       "epoch": 0.1593767390094602,
-      "grad_norm": 2.9344465732574463,
-      "learning_rate": 0.00014483832160900326,
-      "loss": 1.9743,
+      "grad_norm": 2.525339126586914,
+      "learning_rate": 0.00014455729223768966,
+      "loss": 1.9249,
       "step": 358
     },
     {
       "epoch": 0.15982192543127435,
-      "grad_norm": 2.056246757507324,
-      "learning_rate": 0.00014455729223768966,
-      "loss": 1.8003,
+      "grad_norm": 2.1002516746520996,
+      "learning_rate": 0.0001442758231038902,
+      "loss": 1.8466,
       "step": 359
     },
     {
       "epoch": 0.16026711185308848,
-      "grad_norm": 2.6241796016693115,
-      "learning_rate": 0.0001442758231038902,
-      "loss": 1.888,
+      "grad_norm": 4.392673492431641,
+      "learning_rate": 0.00014399391698559152,
+      "loss": 1.8609,
       "step": 360
     },
     {
       "epoch": 0.1607122982749026,
-      "grad_norm": 2.479048490524292,
-      "learning_rate": 0.00014399391698559152,
-      "loss": 2.3365,
+      "grad_norm": 2.834761381149292,
+      "learning_rate": 0.0001437115766650933,
+      "loss": 2.2807,
       "step": 361
     },
     {
       "epoch": 0.16115748469671676,
-      "grad_norm": 3.3290910720825195,
-      "learning_rate": 0.0001437115766650933,
-      "loss": 2.5672,
+      "grad_norm": 3.437537670135498,
+      "learning_rate": 0.00014342880492898048,
+      "loss": 2.4952,
       "step": 362
     },
     {
       "epoch": 0.16160267111853088,
-      "grad_norm": 2.197995901107788,
-      "learning_rate": 0.00014342880492898048,
-      "loss": 1.4891,
+      "grad_norm": 2.5581488609313965,
+      "learning_rate": 0.0001431456045680959,
+      "loss": 1.5068,
       "step": 363
     },
     {
       "epoch": 0.162047857540345,
-      "grad_norm": 2.3591930866241455,
-      "learning_rate": 0.0001431456045680959,
-      "loss": 1.7955,
+      "grad_norm": 2.5887176990509033,
+      "learning_rate": 0.00014286197837751286,
+      "loss": 1.8657,
       "step": 364
     },
     {
       "epoch": 0.16249304396215916,
-      "grad_norm": 1.573541283607483,
-      "learning_rate": 0.00014286197837751286,
-      "loss": 1.1049,
+      "grad_norm": 1.8554385900497437,
+      "learning_rate": 0.00014257792915650728,
+      "loss": 1.1389,
       "step": 365
     },
     {
       "epoch": 0.1629382303839733,
-      "grad_norm": 2.382366895675659,
-      "learning_rate": 0.00014257792915650728,
-      "loss": 1.717,
+      "grad_norm": 2.9788575172424316,
+      "learning_rate": 0.00014229345970853032,
+      "loss": 1.7338,
       "step": 366
     },
     {
       "epoch": 0.1633834168057874,
-      "grad_norm": 3.2191505432128906,
-      "learning_rate": 0.00014229345970853032,
-      "loss": 1.6779,
+      "grad_norm": 3.6624138355255127,
+      "learning_rate": 0.00014200857284118066,
+      "loss": 1.6937,
       "step": 367
     },
     {
       "epoch": 0.16382860322760157,
-      "grad_norm": 2.418705940246582,
-      "learning_rate": 0.00014200857284118066,
-      "loss": 2.6337,
+      "grad_norm": 2.463923931121826,
+      "learning_rate": 0.00014172327136617656,
+      "loss": 2.6497,
       "step": 368
     },
     {
       "epoch": 0.1642737896494157,
-      "grad_norm": 5.708262920379639,
-      "learning_rate": 0.00014172327136617656,
-      "loss": 1.7388,
+      "grad_norm": 2.7360804080963135,
+      "learning_rate": 0.00014143755809932845,
+      "loss": 1.7198,
       "step": 369
     },
     {
       "epoch": 0.16471897607122982,
-      "grad_norm": 2.3304173946380615,
-      "learning_rate": 0.00014143755809932845,
-      "loss": 1.4906,
+      "grad_norm": 5.014685153961182,
+      "learning_rate": 0.00014115143586051088,
+      "loss": 1.5489,
       "step": 370
     },
     {
       "epoch": 0.16516416249304397,
-      "grad_norm": 2.6991124153137207,
-      "learning_rate": 0.00014115143586051088,
-      "loss": 2.3806,
+      "grad_norm": 3.2301042079925537,
+      "learning_rate": 0.00014086490747363493,
+      "loss": 2.4075,
       "step": 371
     },
     {
       "epoch": 0.1656093489148581,
-      "grad_norm": 2.831113815307617,
-      "learning_rate": 0.00014086490747363493,
-      "loss": 2.0115,
+      "grad_norm": 3.1608779430389404,
+      "learning_rate": 0.00014057797576662,
+      "loss": 2.0533,
       "step": 372
     },
     {
       "epoch": 0.16605453533667222,
-      "grad_norm": 1.9368008375167847,
-      "learning_rate": 0.00014057797576662,
-      "loss": 1.6859,
+      "grad_norm": 2.133025646209717,
+      "learning_rate": 0.00014029064357136628,
+      "loss": 1.7483,
       "step": 373
     },
     {
       "epoch": 0.16649972175848637,
-      "grad_norm": 2.908867120742798,
-      "learning_rate": 0.00014029064357136628,
-      "loss": 2.2395,
+      "grad_norm": 3.2147281169891357,
+      "learning_rate": 0.00014000291372372647,
+      "loss": 2.279,
       "step": 374
     },
     {
       "epoch": 0.1669449081803005,
-      "grad_norm": 2.90267014503479,
-      "learning_rate": 0.00014000291372372647,
-      "loss": 1.6788,
+      "grad_norm": 3.177180290222168,
+      "learning_rate": 0.00013971478906347806,
+      "loss": 1.6962,
       "step": 375
     },
     {
       "epoch": 0.16739009460211463,
-      "grad_norm": 4.940428256988525,
-      "learning_rate": 0.00013971478906347806,
-      "loss": 3.0354,
+      "grad_norm": 4.325331211090088,
+      "learning_rate": 0.00013942627243429512,
+      "loss": 2.9958,
       "step": 376
     },
     {
       "epoch": 0.16783528102392878,
-      "grad_norm": 3.7296106815338135,
-      "learning_rate": 0.00013942627243429512,
-      "loss": 3.1005,
+      "grad_norm": 3.7681610584259033,
+      "learning_rate": 0.00013913736668372026,
+      "loss": 3.1293,
       "step": 377
     },
     {
       "epoch": 0.1682804674457429,
-      "grad_norm": 1.515480399131775,
-      "learning_rate": 0.00013913736668372026,
-      "loss": 0.9675,
+      "grad_norm": 1.5435031652450562,
+      "learning_rate": 0.00013884807466313663,
+      "loss": 0.9989,
       "step": 378
     },
     {
       "epoch": 0.16872565386755703,
-      "grad_norm": 2.466693878173828,
-      "learning_rate": 0.00013884807466313663,
-      "loss": 2.4831,
+      "grad_norm": 2.3767507076263428,
+      "learning_rate": 0.00013855839922773968,
+      "loss": 2.5113,
       "step": 379
     },
     {
       "epoch": 0.16917084028937118,
-      "grad_norm": 2.1269826889038086,
-      "learning_rate": 0.00013855839922773968,
-      "loss": 1.7787,
+      "grad_norm": 2.171083927154541,
+      "learning_rate": 0.000138268343236509,
+      "loss": 1.7974,
       "step": 380
     },
     {
       "epoch": 0.1696160267111853,
-      "grad_norm": 2.2861855030059814,
-      "learning_rate": 0.000138268343236509,
-      "loss": 1.8964,
+      "grad_norm": 2.643446207046509,
+      "learning_rate": 0.00013797790955218014,
+      "loss": 1.9618,
       "step": 381
     },
     {
       "epoch": 0.17006121313299943,
-      "grad_norm": 7.59018611907959,
-      "learning_rate": 0.00013797790955218014,
-      "loss": 3.1163,
+      "grad_norm": 6.600242614746094,
+      "learning_rate": 0.00013768710104121627,
+      "loss": 3.0334,
       "step": 382
     },
     {
       "epoch": 0.1705063995548136,
-      "grad_norm": 3.9868404865264893,
-      "learning_rate": 0.00013768710104121627,
-      "loss": 2.4488,
+      "grad_norm": 4.860446929931641,
+      "learning_rate": 0.00013739592057378003,
+      "loss": 2.4737,
       "step": 383
     },
     {
       "epoch": 0.1709515859766277,
-      "grad_norm": 1.6617087125778198,
-      "learning_rate": 0.00013739592057378003,
-      "loss": 1.21,
+      "grad_norm": 1.7493665218353271,
+      "learning_rate": 0.0001371043710237051,
+      "loss": 1.2447,
       "step": 384
     },
     {
       "epoch": 0.17139677239844184,
-      "grad_norm": 1.8703160285949707,
-      "learning_rate": 0.0001371043710237051,
-      "loss": 1.5203,
+      "grad_norm": 1.8847475051879883,
+      "learning_rate": 0.00013681245526846783,
+      "loss": 1.5054,
       "step": 385
     },
     {
       "epoch": 0.171841958820256,
-      "grad_norm": 3.014777898788452,
-      "learning_rate": 0.00013681245526846783,
-      "loss": 2.0583,
+      "grad_norm": 3.478487253189087,
+      "learning_rate": 0.0001365201761891588,
+      "loss": 2.1903,
       "step": 386
     },
     {
       "epoch": 0.17228714524207012,
-      "grad_norm": 1.9843742847442627,
-      "learning_rate": 0.0001365201761891588,
-      "loss": 1.2145,
+      "grad_norm": 1.92650306224823,
+      "learning_rate": 0.00013622753667045457,
+      "loss": 1.2139,
       "step": 387
     },
     {
       "epoch": 0.17273233166388424,
-      "grad_norm": 2.274531364440918,
-      "learning_rate": 0.00013622753667045457,
-      "loss": 1.6243,
+      "grad_norm": 2.4054200649261475,
+      "learning_rate": 0.00013593453960058908,
+      "loss": 1.6496,
       "step": 388
     },
     {
       "epoch": 0.1731775180856984,
-      "grad_norm": 2.8857228755950928,
-      "learning_rate": 0.00013593453960058908,
-      "loss": 2.1109,
+      "grad_norm": 3.745361328125,
+      "learning_rate": 0.00013564118787132506,
+      "loss": 2.1288,
       "step": 389
     },
     {
       "epoch": 0.17362270450751252,
-      "grad_norm": 2.410280227661133,
-      "learning_rate": 0.00013564118787132506,
-      "loss": 1.6705,
+      "grad_norm": 2.4529619216918945,
+      "learning_rate": 0.00013534748437792573,
+      "loss": 1.6666,
       "step": 390
     },
     {
       "epoch": 0.17406789092932665,
-      "grad_norm": 2.83205509185791,
-      "learning_rate": 0.00013534748437792573,
-      "loss": 1.6343,
+      "grad_norm": 3.01240873336792,
+      "learning_rate": 0.0001350534320191259,
+      "loss": 1.6192,
       "step": 391
     },
     {
       "epoch": 0.1745130773511408,
-      "grad_norm": 3.435878276824951,
-      "learning_rate": 0.0001350534320191259,
-      "loss": 2.7,
+      "grad_norm": 3.704413652420044,
+      "learning_rate": 0.0001347590336971037,
+      "loss": 2.7263,
       "step": 392
     },
     {
       "epoch": 0.17495826377295493,
-      "grad_norm": 2.2407634258270264,
-      "learning_rate": 0.0001347590336971037,
-      "loss": 1.5352,
+      "grad_norm": 2.4902546405792236,
+      "learning_rate": 0.0001344642923174517,
+      "loss": 1.58,
       "step": 393
     },
     {
       "epoch": 0.17540345019476905,
-      "grad_norm": 2.265737295150757,
-      "learning_rate": 0.0001344642923174517,
-      "loss": 1.8782,
+      "grad_norm": 2.159018039703369,
+      "learning_rate": 0.00013416921078914835,
+      "loss": 1.9158,
       "step": 394
     },
     {
       "epoch": 0.1758486366165832,
-      "grad_norm": 3.784402847290039,
-      "learning_rate": 0.00013416921078914835,
-      "loss": 2.0235,
+      "grad_norm": 3.4280478954315186,
+      "learning_rate": 0.00013387379202452917,
+      "loss": 2.059,
       "step": 395
     },
     {
       "epoch": 0.17629382303839733,
-      "grad_norm": 3.2451465129852295,
-      "learning_rate": 0.00013387379202452917,
-      "loss": 1.9769,
+      "grad_norm": 3.7631638050079346,
+      "learning_rate": 0.00013357803893925807,
+      "loss": 2.0537,
       "step": 396
     },
     {
       "epoch": 0.17673900946021145,
-      "grad_norm": 2.4640157222747803,
-      "learning_rate": 0.00013357803893925807,
-      "loss": 1.5395,
+      "grad_norm": 5.2759833335876465,
+      "learning_rate": 0.00013328195445229868,
+      "loss": 1.5751,
       "step": 397
     },
     {
       "epoch": 0.1771841958820256,
-      "grad_norm": 2.5422592163085938,
-      "learning_rate": 0.00013328195445229868,
-      "loss": 1.4376,
+      "grad_norm": 2.408705711364746,
+      "learning_rate": 0.00013298554148588528,
+      "loss": 1.444,
       "step": 398
     },
     {
       "epoch": 0.17762938230383973,
-      "grad_norm": 2.2282865047454834,
-      "learning_rate": 0.00013298554148588528,
-      "loss": 1.3988,
+      "grad_norm": 2.2729711532592773,
+      "learning_rate": 0.00013268880296549425,
+      "loss": 1.4234,
       "step": 399
     },
     {
       "epoch": 0.17807456872565386,
-      "grad_norm": 3.0729358196258545,
-      "learning_rate": 0.00013268880296549425,
-      "loss": 1.2906,
+      "grad_norm": 16.259376525878906,
+      "learning_rate": 0.00013239174181981495,
+      "loss": 1.3433,
       "step": 400
     },
     {
       "epoch": 0.178519755147468,
-      "grad_norm": 2.901768684387207,
-      "learning_rate": 0.00013239174181981495,
-      "loss": 1.7331,
+      "grad_norm": 2.839766025543213,
+      "learning_rate": 0.00013209436098072095,
+      "loss": 1.7553,
       "step": 401
     },
     {
       "epoch": 0.17896494156928214,
-      "grad_norm": 3.0331332683563232,
-      "learning_rate": 0.00013209436098072095,
-      "loss": 1.2559,
+      "grad_norm": 3.6237425804138184,
+      "learning_rate": 0.00013179666338324108,
+      "loss": 1.2521,
       "step": 402
     },
     {
       "epoch": 0.17941012799109626,
-      "grad_norm": 2.4391226768493652,
-      "learning_rate": 0.00013179666338324108,
-      "loss": 2.0366,
+      "grad_norm": 1.935936689376831,
+      "learning_rate": 0.0001314986519655305,
+      "loss": 1.9945,
       "step": 403
     },
     {
       "epoch": 0.17985531441291042,
-      "grad_norm": 3.8267745971679688,
-      "learning_rate": 0.0001314986519655305,
-      "loss": 2.2878,
+      "grad_norm": 3.6026456356048584,
+      "learning_rate": 0.0001312003296688415,
+      "loss": 2.2332,
       "step": 404
     },
     {
       "epoch": 0.18030050083472454,
-      "grad_norm": 2.9066944122314453,
-      "learning_rate": 0.0001312003296688415,
-      "loss": 2.3687,
+      "grad_norm": 2.802741289138794,
+      "learning_rate": 0.00013090169943749476,
+      "loss": 2.3164,
       "step": 405
     },
     {
       "epoch": 0.18074568725653867,
-      "grad_norm": 2.571004867553711,
-      "learning_rate": 0.00013090169943749476,
-      "loss": 1.4725,
+      "grad_norm": 2.429253101348877,
+      "learning_rate": 0.0001306027642188501,
+      "loss": 1.5159,
       "step": 406
     },
     {
       "epoch": 0.18119087367835282,
-      "grad_norm": 2.621811866760254,
-      "learning_rate": 0.0001306027642188501,
-      "loss": 1.4061,
+      "grad_norm": 2.450249671936035,
+      "learning_rate": 0.00013030352696327742,
+      "loss": 1.3854,
       "step": 407
     },
     {
       "epoch": 0.18163606010016695,
-      "grad_norm": 3.159045457839966,
-      "learning_rate": 0.00013030352696327742,
-      "loss": 2.3518,
+      "grad_norm": 3.1639933586120605,
+      "learning_rate": 0.00013000399062412763,
+      "loss": 2.4286,
       "step": 408
     },
     {
       "epoch": 0.18208124652198107,
-      "grad_norm": 2.415534496307373,
-      "learning_rate": 0.00013000399062412763,
-      "loss": 1.9811,
+      "grad_norm": 2.301260471343994,
+      "learning_rate": 0.0001297041581577035,
+      "loss": 1.9364,
       "step": 409
     },
     {
       "epoch": 0.18252643294379522,
-      "grad_norm": 3.9818074703216553,
-      "learning_rate": 0.0001297041581577035,
-      "loss": 2.3213,
+      "grad_norm": 3.752669095993042,
+      "learning_rate": 0.0001294040325232304,
+      "loss": 2.2948,
       "step": 410
     },
     {
       "epoch": 0.18297161936560935,
-      "grad_norm": 2.050266981124878,
-      "learning_rate": 0.0001294040325232304,
-      "loss": 1.3752,
+      "grad_norm": 3.825401782989502,
+      "learning_rate": 0.00012910361668282719,
+      "loss": 1.3947,
       "step": 411
     },
     {
       "epoch": 0.18341680578742348,
-      "grad_norm": 2.8136508464813232,
-      "learning_rate": 0.00012910361668282719,
-      "loss": 1.8344,
+      "grad_norm": 3.190941572189331,
+      "learning_rate": 0.00012880291360147693,
+      "loss": 1.7847,
       "step": 412
     },
     {
       "epoch": 0.18386199220923763,
-      "grad_norm": 1.4741458892822266,
-      "learning_rate": 0.00012880291360147693,
-      "loss": 1.1769,
+      "grad_norm": 1.5995765924453735,
+      "learning_rate": 0.0001285019262469976,
+      "loss": 1.1607,
       "step": 413
     },
     {
       "epoch": 0.18430717863105175,
-      "grad_norm": 2.022780418395996,
-      "learning_rate": 0.0001285019262469976,
-      "loss": 1.2992,
+      "grad_norm": 2.03053879737854,
+      "learning_rate": 0.00012820065759001293,
+      "loss": 1.3358,
       "step": 414
     },
     {
       "epoch": 0.18475236505286588,
-      "grad_norm": 3.656526565551758,
-      "learning_rate": 0.00012820065759001293,
-      "loss": 1.7559,
+      "grad_norm": 3.097127914428711,
+      "learning_rate": 0.00012789911060392294,
+      "loss": 1.7824,
       "step": 415
     },
     {
       "epoch": 0.18519755147468003,
-      "grad_norm": 1.9175677299499512,
-      "learning_rate": 0.00012789911060392294,
-      "loss": 1.2447,
+      "grad_norm": 1.8788080215454102,
+      "learning_rate": 0.0001275972882648746,
+      "loss": 1.1845,
       "step": 416
     },
     {
       "epoch": 0.18564273789649416,
-      "grad_norm": 3.0030620098114014,
-      "learning_rate": 0.0001275972882648746,
-      "loss": 1.7294,
+      "grad_norm": 3.1509644985198975,
+      "learning_rate": 0.00012729519355173254,
+      "loss": 1.7853,
       "step": 417
     },
     {
       "epoch": 0.18608792431830828,
-      "grad_norm": 2.6003453731536865,
-      "learning_rate": 0.00012729519355173254,
-      "loss": 1.9416,
+      "grad_norm": 2.6359877586364746,
+      "learning_rate": 0.00012699282944604967,
+      "loss": 1.9357,
       "step": 418
     },
     {
       "epoch": 0.18653311074012244,
-      "grad_norm": 3.4832310676574707,
-      "learning_rate": 0.00012699282944604967,
-      "loss": 2.2738,
+      "grad_norm": 3.170910358428955,
+      "learning_rate": 0.00012669019893203759,
+      "loss": 2.2175,
       "step": 419
     },
     {
       "epoch": 0.18697829716193656,
-      "grad_norm": 3.8813576698303223,
-      "learning_rate": 0.00012669019893203759,
-      "loss": 2.3793,
+      "grad_norm": 3.9023044109344482,
+      "learning_rate": 0.0001263873049965373,
+      "loss": 2.362,
       "step": 420
     },
     {
       "epoch": 0.1874234835837507,
-      "grad_norm": 2.7621383666992188,
-      "learning_rate": 0.0001263873049965373,
-      "loss": 1.966,
+      "grad_norm": 2.529301881790161,
+      "learning_rate": 0.00012608415062898972,
+      "loss": 1.9661,
       "step": 421
     },
     {
       "epoch": 0.18786867000556484,
-      "grad_norm": 2.440943717956543,
-      "learning_rate": 0.00012608415062898972,
-      "loss": 1.8048,
+      "grad_norm": 2.5790696144104004,
+      "learning_rate": 0.000125780738821406,
+      "loss": 1.8104,
       "step": 422
     },
     {
       "epoch": 0.18831385642737897,
-      "grad_norm": 1.448773980140686,
-      "learning_rate": 0.000125780738821406,
-      "loss": 1.2906,
+      "grad_norm": 2.633057117462158,
+      "learning_rate": 0.00012547707256833823,
+      "loss": 1.305,
       "step": 423
     },
     {
       "epoch": 0.1887590428491931,
-      "grad_norm": 3.421792507171631,
-      "learning_rate": 0.00012547707256833823,
-      "loss": 2.0552,
+      "grad_norm": 2.9603359699249268,
+      "learning_rate": 0.00012517315486684972,
+      "loss": 2.049,
       "step": 424
     },
     {
       "epoch": 0.18920422927100725,
-      "grad_norm": 3.0485334396362305,
-      "learning_rate": 0.00012517315486684972,
-      "loss": 1.9516,
+      "grad_norm": 3.049309492111206,
+      "learning_rate": 0.0001248689887164855,
+      "loss": 1.8918,
       "step": 425
     },
     {
       "epoch": 0.18964941569282137,
-      "grad_norm": 2.456167221069336,
-      "learning_rate": 0.0001248689887164855,
+      "grad_norm": 2.4798386096954346,
+      "learning_rate": 0.00012456457711924266,
       "loss": 1.821,
       "step": 426
     },
     {
       "epoch": 0.1900946021146355,
-      "grad_norm": 2.6346614360809326,
-      "learning_rate": 0.00012456457711924266,
-      "loss": 2.2551,
+      "grad_norm": 2.461341142654419,
+      "learning_rate": 0.00012425992307954075,
+      "loss": 2.1629,
       "step": 427
     },
     {
       "epoch": 0.19053978853644965,
-      "grad_norm": 3.651096820831299,
-      "learning_rate": 0.00012425992307954075,
-      "loss": 2.8791,
+      "grad_norm": 3.7393171787261963,
+      "learning_rate": 0.0001239550296041922,
+      "loss": 2.9632,
       "step": 428
     },
     {
       "epoch": 0.19098497495826378,
-      "grad_norm": 1.784805417060852,
-      "learning_rate": 0.0001239550296041922,
-      "loss": 1.3371,
+      "grad_norm": 1.8002454042434692,
+      "learning_rate": 0.00012364989970237248,
+      "loss": 1.3328,
       "step": 429
     },
     {
       "epoch": 0.1914301613800779,
-      "grad_norm": 3.8563966751098633,
-      "learning_rate": 0.00012364989970237248,
-      "loss": 1.7833,
+      "grad_norm": 3.2655725479125977,
+      "learning_rate": 0.00012334453638559057,
+      "loss": 1.691,
       "step": 430
     },
     {
       "epoch": 0.19187534780189205,
-      "grad_norm": 4.481318950653076,
-      "learning_rate": 0.00012334453638559057,
-      "loss": 2.8327,
+      "grad_norm": 3.687551259994507,
+      "learning_rate": 0.00012303894266765908,
+      "loss": 2.7816,
       "step": 431
     },
     {
       "epoch": 0.19232053422370618,
-      "grad_norm": 2.5223546028137207,
-      "learning_rate": 0.00012303894266765908,
-      "loss": 2.3534,
+      "grad_norm": 2.6039812564849854,
+      "learning_rate": 0.00012273312156466464,
+      "loss": 2.3608,
       "step": 432
     },
     {
       "epoch": 0.1927657206455203,
-      "grad_norm": 4.295266151428223,
-      "learning_rate": 0.00012273312156466464,
-      "loss": 2.6395,
+      "grad_norm": 3.720024824142456,
+      "learning_rate": 0.00012242707609493814,
+      "loss": 2.5207,
       "step": 433
     },
     {
       "epoch": 0.19321090706733446,
-      "grad_norm": 1.6645418405532837,
-      "learning_rate": 0.00012242707609493814,
-      "loss": 1.0704,
+      "grad_norm": 1.6510679721832275,
+      "learning_rate": 0.00012212080927902474,
+      "loss": 1.0557,
       "step": 434
     },
     {
       "epoch": 0.19365609348914858,
-      "grad_norm": 3.143169641494751,
-      "learning_rate": 0.00012212080927902474,
-      "loss": 1.7698,
+      "grad_norm": 3.315629243850708,
+      "learning_rate": 0.00012181432413965428,
+      "loss": 1.7847,
       "step": 435
     },
     {
       "epoch": 0.1941012799109627,
-      "grad_norm": 2.1398355960845947,
-      "learning_rate": 0.00012181432413965428,
-      "loss": 1.534,
+      "grad_norm": 1.9172368049621582,
+      "learning_rate": 0.00012150762370171136,
+      "loss": 1.4622,
       "step": 436
     },
     {
       "epoch": 0.19454646633277686,
-      "grad_norm": 3.0436127185821533,
-      "learning_rate": 0.00012150762370171136,
-      "loss": 2.5599,
+      "grad_norm": 3.259791612625122,
+      "learning_rate": 0.00012120071099220549,
+      "loss": 2.6262,
       "step": 437
     },
     {
       "epoch": 0.194991652754591,
-      "grad_norm": 2.8342740535736084,
-      "learning_rate": 0.00012120071099220549,
-      "loss": 2.1801,
+      "grad_norm": 2.9717273712158203,
+      "learning_rate": 0.00012089358904024117,
+      "loss": 2.1567,
       "step": 438
     },
     {
       "epoch": 0.1954368391764051,
-      "grad_norm": 2.6189768314361572,
-      "learning_rate": 0.00012089358904024117,
-      "loss": 1.9147,
+      "grad_norm": 2.3790361881256104,
+      "learning_rate": 0.00012058626087698814,
+      "loss": 1.8398,
       "step": 439
     },
     {
       "epoch": 0.19588202559821927,
-      "grad_norm": 3.389474868774414,
-      "learning_rate": 0.00012058626087698814,
-      "loss": 1.3261,
+      "grad_norm": 3.106640100479126,
+      "learning_rate": 0.00012027872953565125,
+      "loss": 1.3114,
       "step": 440
     },
     {
       "epoch": 0.1963272120200334,
-      "grad_norm": 3.9404706954956055,
-      "learning_rate": 0.00012027872953565125,
-      "loss": 2.1684,
+      "grad_norm": 4.137223720550537,
+      "learning_rate": 0.00011997099805144069,
+      "loss": 2.2908,
       "step": 441
     },
     {
       "epoch": 0.19677239844184752,
-      "grad_norm": 2.1697638034820557,
-      "learning_rate": 0.00011997099805144069,
-      "loss": 1.4573,
+      "grad_norm": 2.8714749813079834,
+      "learning_rate": 0.000119663069461542,
+      "loss": 1.5153,
       "step": 442
     },
     {
       "epoch": 0.19721758486366167,
-      "grad_norm": 3.369276762008667,
-      "learning_rate": 0.000119663069461542,
-      "loss": 2.4629,
+      "grad_norm": 3.3624212741851807,
+      "learning_rate": 0.00011935494680508606,
+      "loss": 2.4638,
       "step": 443
     },
     {
       "epoch": 0.1976627712854758,
-      "grad_norm": 3.570361375808716,
-      "learning_rate": 0.00011935494680508606,
-      "loss": 2.0022,
+      "grad_norm": 3.3930442333221436,
+      "learning_rate": 0.00011904663312311901,
+      "loss": 2.0868,
       "step": 444
     },
     {
       "epoch": 0.19810795770728992,
-      "grad_norm": 2.2363481521606445,
-      "learning_rate": 0.00011904663312311901,
-      "loss": 1.5469,
+      "grad_norm": 2.2953524589538574,
+      "learning_rate": 0.00011873813145857249,
+      "loss": 1.6131,
       "step": 445
     },
     {
       "epoch": 0.19855314412910408,
-      "grad_norm": 3.8988430500030518,
-      "learning_rate": 0.00011873813145857249,
-      "loss": 1.5989,
+      "grad_norm": 3.6715896129608154,
+      "learning_rate": 0.00011842944485623335,
+      "loss": 1.633,
       "step": 446
     },
     {
       "epoch": 0.1989983305509182,
-      "grad_norm": 4.510793685913086,
-      "learning_rate": 0.00011842944485623335,
-      "loss": 1.8663,
+      "grad_norm": 4.457159042358398,
+      "learning_rate": 0.00011812057636271374,
+      "loss": 1.9242,
       "step": 447
     },
     {
       "epoch": 0.19944351697273233,
-      "grad_norm": 2.888258457183838,
-      "learning_rate": 0.00011812057636271374,
-      "loss": 1.8436,
+      "grad_norm": 2.747443437576294,
+      "learning_rate": 0.000117811529026421,
+      "loss": 1.8423,
       "step": 448
     },
     {
       "epoch": 0.19988870339454648,
-      "grad_norm": 3.260255813598633,
-      "learning_rate": 0.000117811529026421,
-      "loss": 1.8474,
+      "grad_norm": 3.4309027194976807,
+      "learning_rate": 0.00011750230589752762,
+      "loss": 1.8837,
       "step": 449
     },
     {
       "epoch": 0.2003338898163606,
-      "grad_norm": 3.7456436157226562,
-      "learning_rate": 0.00011750230589752762,
-      "loss": 2.7048,
+      "grad_norm": 3.642226457595825,
+      "learning_rate": 0.00011719291002794096,
+      "loss": 2.5399,
       "step": 450
     },
     {
       "epoch": 0.20077907623817473,
-      "grad_norm": 3.437774181365967,
-      "learning_rate": 0.00011719291002794096,
-      "loss": 1.8771,
+      "grad_norm": 2.9268958568573,
+      "learning_rate": 0.00011688334447127338,
+      "loss": 1.682,
       "step": 451
     },
     {
       "epoch": 0.20122426265998888,
-      "grad_norm": 2.5170540809631348,
-      "learning_rate": 0.00011688334447127338,
-      "loss": 1.2413,
+      "grad_norm": 3.2225277423858643,
+      "learning_rate": 0.00011657361228281199,
+      "loss": 1.2943,
       "step": 452
     },
     {
       "epoch": 0.201669449081803,
-      "grad_norm": 2.3565282821655273,
-      "learning_rate": 0.00011657361228281199,
-      "loss": 1.5339,
+      "grad_norm": 2.1313159465789795,
+      "learning_rate": 0.00011626371651948838,
+      "loss": 1.5448,
       "step": 453
     },
     {
       "epoch": 0.20211463550361713,
-      "grad_norm": 1.9782150983810425,
-      "learning_rate": 0.00011626371651948838,
-      "loss": 1.6498,
+      "grad_norm": 1.9492061138153076,
+      "learning_rate": 0.00011595366023984864,
+      "loss": 1.6555,
       "step": 454
     },
     {
       "epoch": 0.2025598219254313,
-      "grad_norm": 3.0511531829833984,
-      "learning_rate": 0.00011595366023984864,
-      "loss": 2.4361,
+      "grad_norm": 3.278721570968628,
+      "learning_rate": 0.0001156434465040231,
+      "loss": 2.502,
       "step": 455
     },
     {
       "epoch": 0.2030050083472454,
-      "grad_norm": 2.8406717777252197,
-      "learning_rate": 0.0001156434465040231,
-      "loss": 2.3422,
+      "grad_norm": 2.8373258113861084,
+      "learning_rate": 0.00011533307837369607,
+      "loss": 2.3922,
       "step": 456
     },
     {
       "epoch": 0.20345019476905954,
-      "grad_norm": 1.845874309539795,
-      "learning_rate": 0.00011533307837369607,
-      "loss": 1.1649,
+      "grad_norm": 2.0082993507385254,
+      "learning_rate": 0.00011502255891207572,
+      "loss": 1.1809,
       "step": 457
     },
     {
       "epoch": 0.20389538119087366,
-      "grad_norm": 2.0563290119171143,
-      "learning_rate": 0.00011502255891207572,
-      "loss": 2.0534,
+      "grad_norm": 2.080368757247925,
+      "learning_rate": 0.00011471189118386375,
+      "loss": 2.099,
       "step": 458
     },
     {
       "epoch": 0.20434056761268782,
-      "grad_norm": 2.988293170928955,
-      "learning_rate": 0.00011471189118386375,
-      "loss": 2.1835,
+      "grad_norm": 3.027259111404419,
+      "learning_rate": 0.00011440107825522521,
+      "loss": 2.1991,
       "step": 459
     },
     {
       "epoch": 0.20478575403450194,
-      "grad_norm": 2.55254864692688,
-      "learning_rate": 0.00011440107825522521,
-      "loss": 1.6571,
+      "grad_norm": 2.5214195251464844,
+      "learning_rate": 0.00011409012319375827,
+      "loss": 1.6357,
       "step": 460
     },
     {
       "epoch": 0.20523094045631607,
-      "grad_norm": 3.426617383956909,
-      "learning_rate": 0.00011409012319375827,
-      "loss": 2.8862,
+      "grad_norm": 3.444044351577759,
+      "learning_rate": 0.0001137790290684638,
+      "loss": 2.891,
       "step": 461
     },
     {
       "epoch": 0.20567612687813022,
-      "grad_norm": 2.7885515689849854,
-      "learning_rate": 0.0001137790290684638,
-      "loss": 1.8314,
+      "grad_norm": 2.7990877628326416,
+      "learning_rate": 0.00011346779894971527,
+      "loss": 1.8256,
       "step": 462
     },
     {
       "epoch": 0.20612131329994435,
-      "grad_norm": 3.6172313690185547,
-      "learning_rate": 0.00011346779894971527,
-      "loss": 2.8842,
+      "grad_norm": 3.639455795288086,
+      "learning_rate": 0.00011315643590922827,
+      "loss": 2.8392,
       "step": 463
     },
     {
       "epoch": 0.20656649972175847,
-      "grad_norm": 2.387355089187622,
-      "learning_rate": 0.00011315643590922827,
-      "loss": 1.9039,
+      "grad_norm": 2.3463006019592285,
+      "learning_rate": 0.0001128449430200303,
+      "loss": 1.9092,
       "step": 464
     },
     {
       "epoch": 0.20701168614357263,
-      "grad_norm": 3.7346434593200684,
-      "learning_rate": 0.0001128449430200303,
-      "loss": 2.7045,
+      "grad_norm": 4.1158599853515625,
+      "learning_rate": 0.00011253332335643043,
+      "loss": 2.7153,
       "step": 465
     },
     {
       "epoch": 0.20745687256538675,
-      "grad_norm": 2.339855909347534,
-      "learning_rate": 0.00011253332335643043,
-      "loss": 2.2928,
+      "grad_norm": 2.4753026962280273,
+      "learning_rate": 0.00011222157999398895,
+      "loss": 2.2568,
       "step": 466
     },
     {
       "epoch": 0.20790205898720088,
-      "grad_norm": 2.6608927249908447,
-      "learning_rate": 0.00011222157999398895,
-      "loss": 1.6835,
+      "grad_norm": 6.2262725830078125,
+      "learning_rate": 0.00011190971600948699,
+      "loss": 1.5797,
       "step": 467
     },
     {
       "epoch": 0.20834724540901503,
-      "grad_norm": 1.9378608465194702,
-      "learning_rate": 0.00011190971600948699,
-      "loss": 1.2669,
+      "grad_norm": 1.8223447799682617,
+      "learning_rate": 0.00011159773448089614,
+      "loss": 1.2652,
       "step": 468
     },
     {
       "epoch": 0.20879243183082916,
-      "grad_norm": 1.9632965326309204,
-      "learning_rate": 0.00011159773448089614,
-      "loss": 1.8431,
+      "grad_norm": 1.922956109046936,
+      "learning_rate": 0.00011128563848734816,
+      "loss": 1.8199,
       "step": 469
     },
     {
       "epoch": 0.20923761825264328,
-      "grad_norm": 2.015932083129883,
-      "learning_rate": 0.00011128563848734816,
-      "loss": 2.0745,
+      "grad_norm": 1.9727588891983032,
+      "learning_rate": 0.00011097343110910452,
+      "loss": 2.0616,
       "step": 470
     },
     {
       "epoch": 0.20968280467445743,
-      "grad_norm": 3.485055446624756,
-      "learning_rate": 0.00011097343110910452,
-      "loss": 2.3325,
+      "grad_norm": 3.9084508419036865,
+      "learning_rate": 0.000110661115427526,
+      "loss": 2.2877,
       "step": 471
     },
     {
       "epoch": 0.21012799109627156,
-      "grad_norm": 1.4672657251358032,
-      "learning_rate": 0.000110661115427526,
-      "loss": 1.0659,
+      "grad_norm": 1.531131386756897,
+      "learning_rate": 0.00011034869452504226,
+      "loss": 1.0564,
       "step": 472
     },
     {
       "epoch": 0.21057317751808569,
-      "grad_norm": 1.980690360069275,
-      "learning_rate": 0.00011034869452504226,
-      "loss": 1.4239,
+      "grad_norm": 2.049264669418335,
+      "learning_rate": 0.00011003617148512149,
+      "loss": 1.4703,
       "step": 473
     },
     {
       "epoch": 0.21101836393989984,
-      "grad_norm": 3.352896213531494,
-      "learning_rate": 0.00011003617148512149,
-      "loss": 2.4245,
+      "grad_norm": 3.721092700958252,
+      "learning_rate": 0.00010972354939223996,
+      "loss": 2.4984,
       "step": 474
     },
     {
       "epoch": 0.21146355036171396,
-      "grad_norm": 2.4032557010650635,
-      "learning_rate": 0.00010972354939223996,
-      "loss": 2.1085,
+      "grad_norm": 2.4527993202209473,
+      "learning_rate": 0.00010941083133185146,
+      "loss": 2.0951,
       "step": 475
     },
     {
       "epoch": 0.2119087367835281,
-      "grad_norm": 2.4151711463928223,
-      "learning_rate": 0.00010941083133185146,
-      "loss": 1.4129,
+      "grad_norm": 2.467804431915283,
+      "learning_rate": 0.00010909802039035701,
+      "loss": 1.4163,
       "step": 476
     },
     {
       "epoch": 0.21235392320534224,
-      "grad_norm": 2.9115583896636963,
-      "learning_rate": 0.00010909802039035701,
-      "loss": 1.8839,
+      "grad_norm": 3.00038743019104,
+      "learning_rate": 0.00010878511965507434,
+      "loss": 1.9473,
       "step": 477
     },
     {
       "epoch": 0.21279910962715637,
-      "grad_norm": 2.2955610752105713,
-      "learning_rate": 0.00010878511965507434,
-      "loss": 1.5365,
+      "grad_norm": 2.367648124694824,
+      "learning_rate": 0.00010847213221420736,
+      "loss": 1.5799,
       "step": 478
     },
     {
       "epoch": 0.2132442960489705,
-      "grad_norm": 2.006725311279297,
-      "learning_rate": 0.00010847213221420736,
-      "loss": 1.1548,
+      "grad_norm": 1.9528924226760864,
+      "learning_rate": 0.00010815906115681578,
+      "loss": 1.1888,
       "step": 479
     },
     {
       "epoch": 0.21368948247078465,
-      "grad_norm": 2.6657001972198486,
-      "learning_rate": 0.00010815906115681578,
-      "loss": 2.425,
+      "grad_norm": 2.714282274246216,
+      "learning_rate": 0.0001078459095727845,
+      "loss": 2.428,
       "step": 480
     },
     {
       "epoch": 0.21413466889259877,
-      "grad_norm": 2.453512668609619,
-      "learning_rate": 0.0001078459095727845,
-      "loss": 1.5641,
+      "grad_norm": 2.3494949340820312,
+      "learning_rate": 0.00010753268055279329,
+      "loss": 1.5414,
       "step": 481
     },
     {
       "epoch": 0.2145798553144129,
-      "grad_norm": 2.2438836097717285,
-      "learning_rate": 0.00010753268055279329,
-      "loss": 1.736,
+      "grad_norm": 2.4214608669281006,
+      "learning_rate": 0.0001072193771882861,
+      "loss": 1.7672,
       "step": 482
     },
     {
       "epoch": 0.21502504173622705,
-      "grad_norm": 2.286283254623413,
-      "learning_rate": 0.0001072193771882861,
-      "loss": 1.2366,
+      "grad_norm": 2.17545223236084,
+      "learning_rate": 0.00010690600257144061,
+      "loss": 1.3416,
       "step": 483
     },
     {
       "epoch": 0.21547022815804118,
-      "grad_norm": 2.357325553894043,
-      "learning_rate": 0.00010690600257144061,
-      "loss": 1.9888,
+      "grad_norm": 2.161003351211548,
+      "learning_rate": 0.0001065925597951378,
+      "loss": 1.9831,
       "step": 484
     },
     {
       "epoch": 0.2159154145798553,
-      "grad_norm": 4.799047470092773,
-      "learning_rate": 0.0001065925597951378,
-      "loss": 2.1147,
+      "grad_norm": 4.312062740325928,
+      "learning_rate": 0.00010627905195293135,
+      "loss": 2.153,
       "step": 485
     },
     {
       "epoch": 0.21636060100166946,
-      "grad_norm": 3.4145114421844482,
-      "learning_rate": 0.00010627905195293135,
-      "loss": 2.2844,
+      "grad_norm": 2.86545467376709,
+      "learning_rate": 0.00010596548213901708,
+      "loss": 2.1744,
       "step": 486
     },
     {
       "epoch": 0.21680578742348358,
-      "grad_norm": 2.5542376041412354,
-      "learning_rate": 0.00010596548213901708,
-      "loss": 1.5962,
+      "grad_norm": 13.845744132995605,
+      "learning_rate": 0.00010565185344820247,
+      "loss": 1.6032,
       "step": 487
     },
     {
       "epoch": 0.2172509738452977,
-      "grad_norm": 3.4638633728027344,
-      "learning_rate": 0.00010565185344820247,
-      "loss": 2.0081,
+      "grad_norm": 3.2781753540039062,
+      "learning_rate": 0.00010533816897587606,
+      "loss": 1.9962,
       "step": 488
     },
     {
       "epoch": 0.21769616026711186,
-      "grad_norm": 3.015550374984741,
-      "learning_rate": 0.00010533816897587606,
-      "loss": 1.978,
+      "grad_norm": 3.188302516937256,
+      "learning_rate": 0.00010502443181797697,
+      "loss": 1.9948,
       "step": 489
     },
     {
       "epoch": 0.21814134668892599,
-      "grad_norm": 3.2160592079162598,
-      "learning_rate": 0.00010502443181797697,
-      "loss": 1.6013,
+      "grad_norm": 3.2082247734069824,
+      "learning_rate": 0.00010471064507096426,
+      "loss": 1.6113,
       "step": 490
     },
     {
       "epoch": 0.2185865331107401,
-      "grad_norm": 3.16475510597229,
-      "learning_rate": 0.00010471064507096426,
-      "loss": 2.1781,
+      "grad_norm": 3.1600544452667236,
+      "learning_rate": 0.0001043968118317865,
+      "loss": 2.151,
       "step": 491
     },
     {
       "epoch": 0.21903171953255426,
-      "grad_norm": 3.783090591430664,
-      "learning_rate": 0.0001043968118317865,
-      "loss": 2.6753,
+      "grad_norm": 3.8899669647216797,
+      "learning_rate": 0.00010408293519785101,
+      "loss": 2.5905,
       "step": 492
     },
     {
       "epoch": 0.2194769059543684,
-      "grad_norm": 3.5325706005096436,
-      "learning_rate": 0.00010408293519785101,
-      "loss": 2.4122,
+      "grad_norm": 3.558845043182373,
+      "learning_rate": 0.00010376901826699348,
+      "loss": 2.3854,
       "step": 493
     },
     {
       "epoch": 0.21992209237618252,
-      "grad_norm": 2.7166507244110107,
-      "learning_rate": 0.00010376901826699348,
-      "loss": 1.2042,
+      "grad_norm": 2.09258770942688,
+      "learning_rate": 0.00010345506413744726,
+      "loss": 1.1808,
       "step": 494
     },
     {
       "epoch": 0.22036727879799667,
-      "grad_norm": 3.5661749839782715,
-      "learning_rate": 0.00010345506413744726,
-      "loss": 2.304,
+      "grad_norm": 3.715296506881714,
+      "learning_rate": 0.00010314107590781284,
+      "loss": 2.2847,
       "step": 495
     },
     {
       "epoch": 0.2208124652198108,
-      "grad_norm": 2.4347705841064453,
-      "learning_rate": 0.00010314107590781284,
-      "loss": 1.6218,
+      "grad_norm": 2.6141772270202637,
+      "learning_rate": 0.00010282705667702734,
+      "loss": 1.6611,
       "step": 496
     },
     {
       "epoch": 0.22125765164162492,
-      "grad_norm": 2.070305109024048,
-      "learning_rate": 0.00010282705667702734,
-      "loss": 2.0498,
+      "grad_norm": 2.0929834842681885,
+      "learning_rate": 0.00010251300954433376,
+      "loss": 2.094,
       "step": 497
     },
     {
       "epoch": 0.22170283806343907,
-      "grad_norm": 2.101881980895996,
-      "learning_rate": 0.00010251300954433376,
-      "loss": 1.6202,
+      "grad_norm": 2.0069286823272705,
+      "learning_rate": 0.00010219893760925052,
+      "loss": 1.6007,
       "step": 498
     },
     {
       "epoch": 0.2221480244852532,
-      "grad_norm": 1.2417051792144775,
-      "learning_rate": 0.00010219893760925052,
-      "loss": 0.9997,
+      "grad_norm": 1.2862166166305542,
+      "learning_rate": 0.00010188484397154084,
+      "loss": 0.963,
       "step": 499
     },
     {
       "epoch": 0.22259321090706732,
-      "grad_norm": 2.754783868789673,
-      "learning_rate": 0.00010188484397154084,
-      "loss": 2.2696,
+      "grad_norm": 3.608480930328369,
+      "learning_rate": 0.00010157073173118208,
+      "loss": 2.4212,
       "step": 500
     },
     {
       "epoch": 0.22303839732888148,
-      "grad_norm": 2.7350590229034424,
-      "learning_rate": 0.00010157073173118208,
-      "loss": 1.888,
+      "grad_norm": 3.028559446334839,
+      "learning_rate": 0.00010125660398833528,
+      "loss": 1.9149,
       "step": 501
     },
     {
       "epoch": 0.2234835837506956,
-      "grad_norm": 2.365584373474121,
-      "learning_rate": 0.00010125660398833528,
-      "loss": 1.5181,
+      "grad_norm": 2.9070098400115967,
+      "learning_rate": 0.00010094246384331442,
+      "loss": 1.5629,
       "step": 502
     },
     {
       "epoch": 0.22392877017250973,
-      "grad_norm": 1.7909997701644897,
-      "learning_rate": 0.00010094246384331442,
-      "loss": 1.8491,
+      "grad_norm": 1.9340059757232666,
+      "learning_rate": 0.00010062831439655591,
+      "loss": 1.8523,
       "step": 503
     },
     {
       "epoch": 0.22437395659432388,
-      "grad_norm": 2.5978925228118896,
-      "learning_rate": 0.00010062831439655591,
-      "loss": 2.0812,
+      "grad_norm": 2.7267184257507324,
+      "learning_rate": 0.00010031415874858797,
+      "loss": 2.148,
       "step": 504
     },
     {
       "epoch": 0.224819143016138,
-      "grad_norm": 1.8115602731704712,
-      "learning_rate": 0.00010031415874858797,
-      "loss": 1.5947,
+      "grad_norm": 2.565450668334961,
+      "learning_rate": 0.0001,
+      "loss": 1.5885,
       "step": 505
     },
     {
       "epoch": 0.22526432943795213,
-      "grad_norm": 3.372662305831909,
-      "learning_rate": 0.0001,
-      "loss": 1.846,
+      "grad_norm": 2.329294204711914,
+      "learning_rate": 9.968584125141204e-05,
+      "loss": 1.9223,
       "step": 506
     },
     {
       "epoch": 0.22570951585976629,
-      "grad_norm": 2.081552505493164,
-      "learning_rate": 9.968584125141204e-05,
-      "loss": 1.7191,
+      "grad_norm": 1.89286208152771,
+      "learning_rate": 9.937168560344412e-05,
+      "loss": 1.7222,
       "step": 507
     },
     {
       "epoch": 0.2261547022815804,
-      "grad_norm": 2.355083465576172,
-      "learning_rate": 9.937168560344412e-05,
-      "loss": 1.8958,
+      "grad_norm": 2.4377524852752686,
+      "learning_rate": 9.90575361566856e-05,
+      "loss": 1.922,
       "step": 508
     },
     {
       "epoch": 0.22659988870339454,
-      "grad_norm": 2.8670079708099365,
-      "learning_rate": 9.90575361566856e-05,
-      "loss": 2.0792,
+      "grad_norm": 3.0375113487243652,
+      "learning_rate": 9.874339601166473e-05,
+      "loss": 2.1823,
       "step": 509
     },
     {
       "epoch": 0.2270450751252087,
-      "grad_norm": 3.7465760707855225,
-      "learning_rate": 9.874339601166473e-05,
-      "loss": 2.3301,
+      "grad_norm": 3.9398081302642822,
+      "learning_rate": 9.842926826881796e-05,
+      "loss": 2.2468,
       "step": 510
     },
     {
       "epoch": 0.22749026154702282,
-      "grad_norm": 3.3605611324310303,
-      "learning_rate": 9.842926826881796e-05,
-      "loss": 3.4267,
+      "grad_norm": 3.3195459842681885,
+      "learning_rate": 9.81151560284592e-05,
+      "loss": 3.3873,
       "step": 511
     },
     {
       "epoch": 0.22793544796883694,
-      "grad_norm": 2.507578134536743,
-      "learning_rate": 9.81151560284592e-05,
-      "loss": 1.9814,
+      "grad_norm": 2.422905921936035,
+      "learning_rate": 9.78010623907495e-05,
+      "loss": 1.9593,
       "step": 512
     },
     {
       "epoch": 0.2283806343906511,
-      "grad_norm": 4.343908786773682,
-      "learning_rate": 9.78010623907495e-05,
-      "loss": 2.2251,
+      "grad_norm": 4.870426654815674,
+      "learning_rate": 9.748699045566626e-05,
+      "loss": 2.3262,
       "step": 513
     },
     {
       "epoch": 0.22882582081246522,
-      "grad_norm": 2.300027370452881,
-      "learning_rate": 9.748699045566626e-05,
-      "loss": 1.7964,
+      "grad_norm": 2.3376801013946533,
+      "learning_rate": 9.717294332297268e-05,
+      "loss": 1.8018,
       "step": 514
     },
     {
       "epoch": 0.22927100723427934,
-      "grad_norm": 2.0999529361724854,
-      "learning_rate": 9.717294332297268e-05,
-      "loss": 2.1174,
+      "grad_norm": 2.2101001739501953,
+      "learning_rate": 9.685892409218717e-05,
+      "loss": 2.1126,
       "step": 515
     },
     {
       "epoch": 0.2297161936560935,
-      "grad_norm": 2.6414802074432373,
-      "learning_rate": 9.685892409218717e-05,
-      "loss": 2.0422,
+      "grad_norm": 2.68068790435791,
+      "learning_rate": 9.654493586255278e-05,
+      "loss": 2.0475,
       "step": 516
     },
     {
       "epoch": 0.23016138007790762,
-      "grad_norm": 2.8301937580108643,
-      "learning_rate": 9.654493586255278e-05,
-      "loss": 2.0245,
+      "grad_norm": 2.847536325454712,
+      "learning_rate": 9.623098173300654e-05,
+      "loss": 2.039,
       "step": 517
     },
     {
       "epoch": 0.23060656649972175,
-      "grad_norm": 1.4676358699798584,
-      "learning_rate": 9.623098173300654e-05,
-      "loss": 1.3448,
+      "grad_norm": 1.75150728225708,
+      "learning_rate": 9.591706480214901e-05,
+      "loss": 1.3167,
       "step": 518
     },
     {
       "epoch": 0.2310517529215359,
-      "grad_norm": 3.920909881591797,
-      "learning_rate": 9.591706480214901e-05,
-      "loss": 2.7722,
+      "grad_norm": 4.02113151550293,
+      "learning_rate": 9.560318816821353e-05,
+      "loss": 2.7966,
       "step": 519
     },
     {
       "epoch": 0.23149693934335003,
-      "grad_norm": 4.668717861175537,
-      "learning_rate": 9.560318816821353e-05,
-      "loss": 1.8226,
+      "grad_norm": 4.969461441040039,
+      "learning_rate": 9.528935492903575e-05,
+      "loss": 1.8454,
       "step": 520
     },
     {
       "epoch": 0.23194212576516415,
-      "grad_norm": 1.8420554399490356,
-      "learning_rate": 9.528935492903575e-05,
-      "loss": 1.0351,
+      "grad_norm": 1.8146275281906128,
+      "learning_rate": 9.497556818202306e-05,
+      "loss": 1.0192,
       "step": 521
     },
     {
       "epoch": 0.2323873121869783,
-      "grad_norm": 3.2772724628448486,
-      "learning_rate": 9.497556818202306e-05,
-      "loss": 2.7337,
+      "grad_norm": 3.0495381355285645,
+      "learning_rate": 9.466183102412395e-05,
+      "loss": 2.642,
       "step": 522
     },
     {
       "epoch": 0.23283249860879243,
-      "grad_norm": 2.2478621006011963,
-      "learning_rate": 9.466183102412395e-05,
-      "loss": 1.9922,
+      "grad_norm": 2.3191003799438477,
+      "learning_rate": 9.434814655179755e-05,
+      "loss": 1.9564,
       "step": 523
     },
     {
       "epoch": 0.23327768503060656,
-      "grad_norm": 1.188016653060913,
-      "learning_rate": 9.434814655179755e-05,
-      "loss": 1.0313,
+      "grad_norm": 1.3154622316360474,
+      "learning_rate": 9.403451786098294e-05,
+      "loss": 1.0393,
       "step": 524
     },
     {
       "epoch": 0.2337228714524207,
-      "grad_norm": 3.9148828983306885,
-      "learning_rate": 9.403451786098294e-05,
-      "loss": 2.2554,
+      "grad_norm": 3.91440749168396,
+      "learning_rate": 9.372094804706867e-05,
+      "loss": 2.2118,
       "step": 525
     },
     {
       "epoch": 0.23416805787423484,
-      "grad_norm": 2.4724814891815186,
-      "learning_rate": 9.372094804706867e-05,
-      "loss": 1.8411,
+      "grad_norm": 2.475968360900879,
+      "learning_rate": 9.340744020486222e-05,
+      "loss": 1.8973,
       "step": 526
     },
     {
       "epoch": 0.23461324429604896,
-      "grad_norm": 4.861220359802246,
-      "learning_rate": 9.340744020486222e-05,
-      "loss": 1.7102,
+      "grad_norm": 5.9066596031188965,
+      "learning_rate": 9.309399742855942e-05,
+      "loss": 1.704,
       "step": 527
     },
     {
       "epoch": 0.23505843071786311,
-      "grad_norm": 1.9564683437347412,
-      "learning_rate": 9.309399742855942e-05,
-      "loss": 1.119,
+      "grad_norm": 1.908116340637207,
+      "learning_rate": 9.278062281171393e-05,
+      "loss": 1.1249,
       "step": 528
     },
     {
       "epoch": 0.23550361713967724,
-      "grad_norm": 2.6515750885009766,
-      "learning_rate": 9.278062281171393e-05,
-      "loss": 2.0485,
+      "grad_norm": 2.5091328620910645,
+      "learning_rate": 9.246731944720675e-05,
+      "loss": 1.9685,
       "step": 529
     },
     {
       "epoch": 0.23594880356149137,
-      "grad_norm": 3.663888454437256,
-      "learning_rate": 9.246731944720675e-05,
-      "loss": 2.0759,
+      "grad_norm": 3.161684036254883,
+      "learning_rate": 9.215409042721552e-05,
+      "loss": 2.0443,
       "step": 530
     },
     {
       "epoch": 0.23639398998330552,
-      "grad_norm": 1.989341378211975,
-      "learning_rate": 9.215409042721552e-05,
-      "loss": 1.3344,
+      "grad_norm": 1.9064785242080688,
+      "learning_rate": 9.184093884318425e-05,
+      "loss": 1.3001,
       "step": 531
     },
     {
       "epoch": 0.23683917640511964,
-      "grad_norm": 2.359320878982544,
-      "learning_rate": 9.184093884318425e-05,
-      "loss": 1.3044,
+      "grad_norm": 2.3057310581207275,
+      "learning_rate": 9.152786778579267e-05,
+      "loss": 1.3416,
       "step": 532
     },
     {
       "epoch": 0.23728436282693377,
-      "grad_norm": 4.149430274963379,
-      "learning_rate": 9.152786778579267e-05,
-      "loss": 2.1484,
+      "grad_norm": 4.578581809997559,
+      "learning_rate": 9.121488034492569e-05,
+      "loss": 2.1121,
       "step": 533
     },
     {
       "epoch": 0.23772954924874792,
-      "grad_norm": 3.5896191596984863,
-      "learning_rate": 9.121488034492569e-05,
-      "loss": 2.8477,
+      "grad_norm": 3.51808500289917,
+      "learning_rate": 9.090197960964301e-05,
+      "loss": 2.7852,
       "step": 534
     },
     {
       "epoch": 0.23817473567056205,
-      "grad_norm": 2.467137336730957,
-      "learning_rate": 9.090197960964301e-05,
-      "loss": 1.5795,
+      "grad_norm": 3.1831729412078857,
+      "learning_rate": 9.058916866814858e-05,
+      "loss": 1.5904,
       "step": 535
     },
     {
       "epoch": 0.23861992209237617,
-      "grad_norm": 3.346538543701172,
-      "learning_rate": 9.058916866814858e-05,
-      "loss": 2.2605,
+      "grad_norm": 3.1725871562957764,
+      "learning_rate": 9.027645060776006e-05,
+      "loss": 2.2413,
       "step": 536
     },
     {
       "epoch": 0.23906510851419033,
-      "grad_norm": 3.115483522415161,
-      "learning_rate": 9.027645060776006e-05,
-      "loss": 1.9898,
+      "grad_norm": 2.9285264015197754,
+      "learning_rate": 8.99638285148785e-05,
+      "loss": 1.9292,
       "step": 537
     },
     {
       "epoch": 0.23951029493600445,
-      "grad_norm": 2.0012478828430176,
-      "learning_rate": 8.99638285148785e-05,
-      "loss": 1.3743,
+      "grad_norm": 1.8293259143829346,
+      "learning_rate": 8.965130547495776e-05,
+      "loss": 1.3782,
       "step": 538
     },
     {
       "epoch": 0.23995548135781858,
-      "grad_norm": 3.6035563945770264,
-      "learning_rate": 8.965130547495776e-05,
-      "loss": 1.6191,
+      "grad_norm": 3.355395793914795,
+      "learning_rate": 8.933888457247402e-05,
+      "loss": 1.5259,
       "step": 539
     },
     {
       "epoch": 0.24040066777963273,
-      "grad_norm": 1.9701684713363647,
-      "learning_rate": 8.933888457247402e-05,
-      "loss": 1.7474,
+      "grad_norm": 2.001899480819702,
+      "learning_rate": 8.902656889089548e-05,
+      "loss": 1.7848,
       "step": 540
     },
     {
       "epoch": 0.24084585420144686,
-      "grad_norm": 3.0901401042938232,
-      "learning_rate": 8.902656889089548e-05,
-      "loss": 2.3074,
+      "grad_norm": 3.229248046875,
+      "learning_rate": 8.871436151265184e-05,
+      "loss": 2.3367,
       "step": 541
     },
     {
       "epoch": 0.24129104062326098,
-      "grad_norm": 3.2117514610290527,
-      "learning_rate": 8.871436151265184e-05,
-      "loss": 2.0538,
+      "grad_norm": 3.0638716220855713,
+      "learning_rate": 8.840226551910387e-05,
+      "loss": 2.0436,
       "step": 542
     },
     {
       "epoch": 0.24173622704507514,
-      "grad_norm": 2.3342740535736084,
-      "learning_rate": 8.840226551910387e-05,
-      "loss": 1.1496,
+      "grad_norm": 2.2944014072418213,
+      "learning_rate": 8.809028399051302e-05,
+      "loss": 1.1552,
       "step": 543
     },
     {
       "epoch": 0.24218141346688926,
-      "grad_norm": 2.8328073024749756,
-      "learning_rate": 8.809028399051302e-05,
-      "loss": 1.2791,
+      "grad_norm": 3.305636405944824,
+      "learning_rate": 8.777842000601105e-05,
+      "loss": 1.3073,
       "step": 544
     },
     {
       "epoch": 0.2426265998887034,
-      "grad_norm": 2.9056038856506348,
-      "learning_rate": 8.777842000601105e-05,
-      "loss": 2.2656,
+      "grad_norm": 3.025535821914673,
+      "learning_rate": 8.746667664356956e-05,
+      "loss": 2.267,
       "step": 545
     },
     {
       "epoch": 0.24307178631051754,
-      "grad_norm": 2.6859729290008545,
-      "learning_rate": 8.746667664356956e-05,
-      "loss": 2.004,
+      "grad_norm": 2.5709593296051025,
+      "learning_rate": 8.715505697996971e-05,
+      "loss": 1.9966,
       "step": 546
     },
     {
       "epoch": 0.24351697273233167,
-      "grad_norm": 2.280202627182007,
-      "learning_rate": 8.715505697996971e-05,
-      "loss": 2.0117,
+      "grad_norm": 2.39162278175354,
+      "learning_rate": 8.684356409077176e-05,
+      "loss": 2.0166,
       "step": 547
     },
     {
       "epoch": 0.2439621591541458,
-      "grad_norm": 3.593543291091919,
-      "learning_rate": 8.684356409077176e-05,
-      "loss": 2.4651,
+      "grad_norm": 3.4689042568206787,
+      "learning_rate": 8.653220105028474e-05,
+      "loss": 2.5257,
       "step": 548
     },
     {
       "epoch": 0.24440734557595994,
-      "grad_norm": 2.155273199081421,
-      "learning_rate": 8.653220105028474e-05,
-      "loss": 0.9082,
+      "grad_norm": 1.7875288724899292,
+      "learning_rate": 8.62209709315362e-05,
+      "loss": 0.9387,
       "step": 549
     },
     {
       "epoch": 0.24485253199777407,
-      "grad_norm": 3.196293354034424,
-      "learning_rate": 8.62209709315362e-05,
-      "loss": 2.4122,
+      "grad_norm": 3.2240853309631348,
+      "learning_rate": 8.590987680624174e-05,
+      "loss": 2.4149,
       "step": 550
     },
     {
       "epoch": 0.2452977184195882,
-      "grad_norm": 2.0001351833343506,
-      "learning_rate": 8.590987680624174e-05,
-      "loss": 1.6249,
+      "grad_norm": 1.9656553268432617,
+      "learning_rate": 8.559892174477479e-05,
+      "loss": 1.6305,
       "step": 551
     },
     {
       "epoch": 0.24574290484140235,
-      "grad_norm": 2.7972264289855957,
-      "learning_rate": 8.559892174477479e-05,
-      "loss": 1.9658,
+      "grad_norm": 2.66410756111145,
+      "learning_rate": 8.528810881613626e-05,
+      "loss": 1.9304,
       "step": 552
     },
     {
       "epoch": 0.24618809126321647,
-      "grad_norm": 2.3301680088043213,
-      "learning_rate": 8.528810881613626e-05,
-      "loss": 2.2419,
+      "grad_norm": 2.319838047027588,
+      "learning_rate": 8.497744108792429e-05,
+      "loss": 2.2429,
       "step": 553
     },
     {
       "epoch": 0.2466332776850306,
-      "grad_norm": 3.339153289794922,
-      "learning_rate": 8.497744108792429e-05,
-      "loss": 2.1285,
+      "grad_norm": 3.2388088703155518,
+      "learning_rate": 8.466692162630392e-05,
+      "loss": 2.1559,
       "step": 554
     },
     {
       "epoch": 0.24707846410684475,
-      "grad_norm": 4.093669891357422,
-      "learning_rate": 8.466692162630392e-05,
-      "loss": 1.7672,
+      "grad_norm": 4.480194568634033,
+      "learning_rate": 8.435655349597689e-05,
+      "loss": 1.7788,
       "step": 555
     },
     {
       "epoch": 0.24752365052865888,
-      "grad_norm": 3.632969379425049,
-      "learning_rate": 8.435655349597689e-05,
-      "loss": 1.7379,
+      "grad_norm": 3.7542636394500732,
+      "learning_rate": 8.404633976015134e-05,
+      "loss": 1.6982,
       "step": 556
     },
     {
       "epoch": 0.247968836950473,
-      "grad_norm": 2.793414831161499,
-      "learning_rate": 8.404633976015134e-05,
-      "loss": 1.859,
+      "grad_norm": 2.8315999507904053,
+      "learning_rate": 8.373628348051165e-05,
+      "loss": 1.8328,
       "step": 557
     },
     {
       "epoch": 0.24841402337228716,
-      "grad_norm": 2.6596624851226807,
-      "learning_rate": 8.373628348051165e-05,
-      "loss": 1.9903,
+      "grad_norm": 2.8203070163726807,
+      "learning_rate": 8.342638771718802e-05,
+      "loss": 2.0157,
       "step": 558
     },
     {
       "epoch": 0.24885920979410128,
-      "grad_norm": 2.221126079559326,
-      "learning_rate": 8.342638771718802e-05,
-      "loss": 2.6768,
+      "grad_norm": 2.3228957653045654,
+      "learning_rate": 8.311665552872662e-05,
+      "loss": 2.7434,
       "step": 559
     },
     {
       "epoch": 0.2493043962159154,
-      "grad_norm": 2.701017379760742,
-      "learning_rate": 8.311665552872662e-05,
-      "loss": 1.6827,
+      "grad_norm": 2.91206693649292,
+      "learning_rate": 8.280708997205904e-05,
+      "loss": 1.6866,
       "step": 560
     },
     {
       "epoch": 0.24974958263772956,
-      "grad_norm": 1.7237666845321655,
-      "learning_rate": 8.280708997205904e-05,
-      "loss": 1.5455,
+      "grad_norm": 1.778489589691162,
+      "learning_rate": 8.249769410247239e-05,
+      "loss": 1.5339,
       "step": 561
     },
     {
       "epoch": 0.2501947690595437,
-      "grad_norm": 2.3445723056793213,
-      "learning_rate": 8.249769410247239e-05,
-      "loss": 1.7404,
+      "grad_norm": 2.1387264728546143,
+      "learning_rate": 8.218847097357898e-05,
+      "loss": 1.6759,
       "step": 562
     },
     {
       "epoch": 0.2506399554813578,
-      "grad_norm": 1.6703085899353027,
-      "learning_rate": 8.218847097357898e-05,
-      "loss": 1.5689,
+      "grad_norm": 1.6116218566894531,
+      "learning_rate": 8.187942363728625e-05,
+      "loss": 1.5866,
       "step": 563
     },
     {
       "epoch": 0.25108514190317194,
-      "grad_norm": 2.455984115600586,
-      "learning_rate": 8.187942363728625e-05,
-      "loss": 1.366,
+      "grad_norm": 2.747645854949951,
+      "learning_rate": 8.157055514376666e-05,
+      "loss": 1.4102,
       "step": 564
     },
     {
       "epoch": 0.25153032832498606,
-      "grad_norm": 2.8952829837799072,
-      "learning_rate": 8.157055514376666e-05,
-      "loss": 2.6693,
+      "grad_norm": 2.8808135986328125,
+      "learning_rate": 8.126186854142752e-05,
+      "loss": 2.6266,
       "step": 565
     },
     {
       "epoch": 0.25197551474680024,
-      "grad_norm": 2.0182900428771973,
-      "learning_rate": 8.126186854142752e-05,
-      "loss": 1.7768,
+      "grad_norm": 2.3907628059387207,
+      "learning_rate": 8.095336687688102e-05,
+      "loss": 1.7627,
       "step": 566
     },
     {
       "epoch": 0.25242070116861437,
-      "grad_norm": 3.7107861042022705,
-      "learning_rate": 8.095336687688102e-05,
-      "loss": 2.4294,
+      "grad_norm": 3.7652440071105957,
+      "learning_rate": 8.064505319491398e-05,
+      "loss": 2.3349,
       "step": 567
     },
     {
       "epoch": 0.2528658875904285,
-      "grad_norm": 2.1526546478271484,
-      "learning_rate": 8.064505319491398e-05,
-      "loss": 2.0468,
+      "grad_norm": 10.634536743164062,
+      "learning_rate": 8.033693053845801e-05,
+      "loss": 2.0711,
       "step": 568
     },
     {
       "epoch": 0.2533110740122426,
-      "grad_norm": 2.4397218227386475,
-      "learning_rate": 8.033693053845801e-05,
-      "loss": 2.1954,
+      "grad_norm": 2.42941951751709,
+      "learning_rate": 8.002900194855932e-05,
+      "loss": 2.193,
       "step": 569
     },
     {
       "epoch": 0.25375626043405675,
-      "grad_norm": 2.494946241378784,
-      "learning_rate": 8.002900194855932e-05,
-      "loss": 1.9958,
+      "grad_norm": 2.9347145557403564,
+      "learning_rate": 7.972127046434878e-05,
+      "loss": 2.0439,
       "step": 570
     },
     {
       "epoch": 0.25420144685587087,
-      "grad_norm": 4.778120994567871,
-      "learning_rate": 7.972127046434878e-05,
-      "loss": 2.9802,
+      "grad_norm": 4.5750627517700195,
+      "learning_rate": 7.941373912301189e-05,
+      "loss": 2.9297,
       "step": 571
     },
     {
       "epoch": 0.25464663327768505,
-      "grad_norm": 7.366669178009033,
-      "learning_rate": 7.941373912301189e-05,
-      "loss": 1.6644,
+      "grad_norm": 2.949826717376709,
+      "learning_rate": 7.910641095975886e-05,
+      "loss": 1.7383,
       "step": 572
     },
     {
       "epoch": 0.2550918196994992,
-      "grad_norm": 2.677617311477661,
-      "learning_rate": 7.910641095975886e-05,
-      "loss": 2.0434,
+      "grad_norm": 2.601602554321289,
+      "learning_rate": 7.879928900779456e-05,
+      "loss": 1.934,
       "step": 573
     },
     {
       "epoch": 0.2555370061213133,
-      "grad_norm": 1.901222825050354,
-      "learning_rate": 7.879928900779456e-05,
-      "loss": 1.0331,
+      "grad_norm": 2.0207157135009766,
+      "learning_rate": 7.849237629828869e-05,
+      "loss": 1.08,
       "step": 574
     },
     {
       "epoch": 0.25598219254312743,
-      "grad_norm": 4.581434726715088,
-      "learning_rate": 7.849237629828869e-05,
-      "loss": 3.2133,
+      "grad_norm": 5.089204788208008,
+      "learning_rate": 7.818567586034577e-05,
+      "loss": 3.2924,
       "step": 575
     },
     {
       "epoch": 0.25642737896494155,
-      "grad_norm": 3.0468733310699463,
-      "learning_rate": 7.818567586034577e-05,
-      "loss": 1.6714,
+      "grad_norm": 3.2320477962493896,
+      "learning_rate": 7.787919072097531e-05,
+      "loss": 1.6713,
       "step": 576
     },
     {
       "epoch": 0.2568725653867557,
-      "grad_norm": 2.401651620864868,
-      "learning_rate": 7.787919072097531e-05,
-      "loss": 1.4202,
+      "grad_norm": 2.1552391052246094,
+      "learning_rate": 7.75729239050619e-05,
+      "loss": 1.3664,
       "step": 577
     },
     {
       "epoch": 0.25731775180856986,
-      "grad_norm": 3.3343799114227295,
-      "learning_rate": 7.75729239050619e-05,
-      "loss": 1.9961,
+      "grad_norm": 3.1360971927642822,
+      "learning_rate": 7.726687843533538e-05,
+      "loss": 1.9872,
       "step": 578
     },
     {
       "epoch": 0.257762938230384,
-      "grad_norm": 2.1830644607543945,
-      "learning_rate": 7.726687843533538e-05,
-      "loss": 1.8748,
+      "grad_norm": 2.541607141494751,
+      "learning_rate": 7.696105733234098e-05,
+      "loss": 1.8889,
       "step": 579
     },
     {
       "epoch": 0.2582081246521981,
-      "grad_norm": 3.9290175437927246,
-      "learning_rate": 7.696105733234098e-05,
-      "loss": 2.0981,
+      "grad_norm": 3.5571320056915283,
+      "learning_rate": 7.66554636144095e-05,
+      "loss": 2.1595,
       "step": 580
     },
     {
       "epoch": 0.25865331107401224,
-      "grad_norm": 2.6848762035369873,
-      "learning_rate": 7.66554636144095e-05,
-      "loss": 2.07,
+      "grad_norm": 2.6313812732696533,
+      "learning_rate": 7.635010029762756e-05,
+      "loss": 2.1058,
       "step": 581
     },
     {
       "epoch": 0.25909849749582636,
-      "grad_norm": 3.8313651084899902,
-      "learning_rate": 7.635010029762756e-05,
-      "loss": 1.8205,
+      "grad_norm": 4.335572719573975,
+      "learning_rate": 7.604497039580785e-05,
+      "loss": 1.8649,
       "step": 582
     },
     {
       "epoch": 0.2595436839176405,
-      "grad_norm": 2.1311826705932617,
-      "learning_rate": 7.604497039580785e-05,
-      "loss": 1.5328,
+      "grad_norm": 2.399296522140503,
+      "learning_rate": 7.574007692045928e-05,
+      "loss": 1.5448,
       "step": 583
     },
     {
       "epoch": 0.25998887033945467,
-      "grad_norm": 2.774587631225586,
-      "learning_rate": 7.574007692045928e-05,
-      "loss": 1.3775,
+      "grad_norm": 2.475374221801758,
+      "learning_rate": 7.543542288075739e-05,
+      "loss": 1.3653,
       "step": 584
     },
     {
       "epoch": 0.2604340567612688,
-      "grad_norm": 2.86027193069458,
-      "learning_rate": 7.543542288075739e-05,
-      "loss": 2.5958,
+      "grad_norm": 2.7806360721588135,
+      "learning_rate": 7.513101128351454e-05,
+      "loss": 2.6151,
       "step": 585
     },
     {
       "epoch": 0.2608792431830829,
-      "grad_norm": 3.274968147277832,
-      "learning_rate": 7.513101128351454e-05,
-      "loss": 2.5216,
+      "grad_norm": 7.9013190269470215,
+      "learning_rate": 7.48268451331503e-05,
+      "loss": 2.4676,
       "step": 586
     },
     {
       "epoch": 0.26132442960489705,
-      "grad_norm": 2.597700595855713,
-      "learning_rate": 7.48268451331503e-05,
-      "loss": 2.5979,
+      "grad_norm": 2.787186861038208,
+      "learning_rate": 7.45229274316618e-05,
+      "loss": 2.5793,
       "step": 587
     },
     {
       "epoch": 0.26176961602671117,
-      "grad_norm": 1.9474818706512451,
-      "learning_rate": 7.45229274316618e-05,
-      "loss": 1.4696,
+      "grad_norm": 2.341843605041504,
+      "learning_rate": 7.421926117859403e-05,
+      "loss": 1.4365,
       "step": 588
     },
     {
       "epoch": 0.2622148024485253,
-      "grad_norm": 3.209158182144165,
-      "learning_rate": 7.421926117859403e-05,
-      "loss": 2.4483,
+      "grad_norm": 3.3418471813201904,
+      "learning_rate": 7.391584937101033e-05,
+      "loss": 2.4653,
       "step": 589
     },
     {
       "epoch": 0.2626599888703395,
-      "grad_norm": 2.9960036277770996,
-      "learning_rate": 7.391584937101033e-05,
-      "loss": 2.4985,
+      "grad_norm": 3.7794909477233887,
+      "learning_rate": 7.361269500346274e-05,
+      "loss": 2.6623,
       "step": 590
     },
     {
       "epoch": 0.2631051752921536,
-      "grad_norm": 1.7019833326339722,
-      "learning_rate": 7.361269500346274e-05,
-      "loss": 1.3692,
+      "grad_norm": 2.570446729660034,
+      "learning_rate": 7.330980106796246e-05,
+      "loss": 1.4194,
       "step": 591
     },
     {
       "epoch": 0.26355036171396773,
-      "grad_norm": 3.178032398223877,
-      "learning_rate": 7.330980106796246e-05,
-      "loss": 1.9495,
+      "grad_norm": 3.370253324508667,
+      "learning_rate": 7.300717055395039e-05,
+      "loss": 1.9577,
       "step": 592
     },
     {
       "epoch": 0.26399554813578185,
-      "grad_norm": 3.838534355163574,
-      "learning_rate": 7.300717055395039e-05,
-      "loss": 2.3653,
+      "grad_norm": 4.153658390045166,
+      "learning_rate": 7.270480644826749e-05,
+      "loss": 2.4074,
       "step": 593
     },
     {
       "epoch": 0.264440734557596,
-      "grad_norm": 2.432239294052124,
-      "learning_rate": 7.270480644826749e-05,
-      "loss": 1.0926,
+      "grad_norm": 2.401904344558716,
+      "learning_rate": 7.240271173512546e-05,
+      "loss": 1.0986,
       "step": 594
     },
     {
       "epoch": 0.2648859209794101,
-      "grad_norm": 3.833432197570801,
-      "learning_rate": 7.240271173512546e-05,
-      "loss": 2.6508,
+      "grad_norm": 3.75860595703125,
+      "learning_rate": 7.210088939607708e-05,
+      "loss": 2.6992,
       "step": 595
     },
     {
       "epoch": 0.2653311074012243,
-      "grad_norm": 3.8783698081970215,
-      "learning_rate": 7.210088939607708e-05,
-      "loss": 2.1285,
+      "grad_norm": 4.001908302307129,
+      "learning_rate": 7.179934240998706e-05,
+      "loss": 2.1339,
       "step": 596
     },
     {
       "epoch": 0.2657762938230384,
-      "grad_norm": 3.5125503540039062,
-      "learning_rate": 7.179934240998706e-05,
-      "loss": 2.6731,
+      "grad_norm": 3.6537344455718994,
+      "learning_rate": 7.149807375300239e-05,
+      "loss": 2.705,
       "step": 597
     },
     {
       "epoch": 0.26622148024485254,
-      "grad_norm": 2.598796844482422,
-      "learning_rate": 7.149807375300239e-05,
-      "loss": 2.4519,
+      "grad_norm": 2.6432173252105713,
+      "learning_rate": 7.119708639852312e-05,
+      "loss": 2.4516,
       "step": 598
     },
     {
       "epoch": 0.26666666666666666,
-      "grad_norm": 2.9682698249816895,
-      "learning_rate": 7.119708639852312e-05,
-      "loss": 1.7751,
+      "grad_norm": 4.261600971221924,
+      "learning_rate": 7.089638331717284e-05,
+      "loss": 1.8103,
       "step": 599
     },
     {
       "epoch": 0.2671118530884808,
-      "grad_norm": 2.220308780670166,
-      "learning_rate": 7.089638331717284e-05,
-      "loss": 1.7819,
+      "grad_norm": 1.7668631076812744,
+      "learning_rate": 7.059596747676962e-05,
+      "loss": 1.7337,
       "step": 600
     },
     {
       "epoch": 0.2675570395102949,
-      "grad_norm": 2.421492338180542,
-      "learning_rate": 7.059596747676962e-05,
-      "loss": 2.3129,
+      "grad_norm": 2.6390795707702637,
+      "learning_rate": 7.029584184229653e-05,
+      "loss": 2.217,
       "step": 601
     },
     {
       "epoch": 0.2680022259321091,
-      "grad_norm": 2.011636257171631,
-      "learning_rate": 7.029584184229653e-05,
-      "loss": 1.264,
+      "grad_norm": 1.76444411277771,
+      "learning_rate": 6.999600937587239e-05,
+      "loss": 1.2096,
       "step": 602
     },
     {
       "epoch": 0.2684474123539232,
-      "grad_norm": 1.722353458404541,
-      "learning_rate": 6.999600937587239e-05,
-      "loss": 1.4916,
+      "grad_norm": 2.0564613342285156,
+      "learning_rate": 6.969647303672262e-05,
+      "loss": 1.5429,
       "step": 603
     },
     {
       "epoch": 0.26889259877573735,
-      "grad_norm": 2.3344128131866455,
-      "learning_rate": 6.969647303672262e-05,
-      "loss": 2.0492,
+      "grad_norm": 2.3482933044433594,
+      "learning_rate": 6.939723578114993e-05,
+      "loss": 2.0948,
       "step": 604
     },
     {
       "epoch": 0.26933778519755147,
-      "grad_norm": 10.07349681854248,
-      "learning_rate": 6.939723578114993e-05,
-      "loss": 2.213,
+      "grad_norm": 6.831959247589111,
+      "learning_rate": 6.909830056250527e-05,
+      "loss": 2.2643,
       "step": 605
     },
     {
       "epoch": 0.2697829716193656,
-      "grad_norm": 1.6907522678375244,
-      "learning_rate": 6.909830056250527e-05,
-      "loss": 1.3918,
+      "grad_norm": 1.8372883796691895,
+      "learning_rate": 6.879967033115853e-05,
+      "loss": 1.411,
       "step": 606
     },
     {
       "epoch": 0.2702281580411797,
-      "grad_norm": 3.627065896987915,
-      "learning_rate": 6.879967033115853e-05,
-      "loss": 2.4359,
+      "grad_norm": 3.655836582183838,
+      "learning_rate": 6.850134803446954e-05,
+      "loss": 2.4844,
       "step": 607
     },
     {
       "epoch": 0.2706733444629939,
-      "grad_norm": 1.5359236001968384,
-      "learning_rate": 6.850134803446954e-05,
-      "loss": 1.1468,
+      "grad_norm": 1.5599287748336792,
+      "learning_rate": 6.820333661675893e-05,
+      "loss": 1.1673,
       "step": 608
     },
     {
       "epoch": 0.27111853088480803,
-      "grad_norm": 2.957019567489624,
-      "learning_rate": 6.820333661675893e-05,
-      "loss": 1.8714,
+      "grad_norm": 2.9432828426361084,
+      "learning_rate": 6.790563901927907e-05,
+      "loss": 1.8728,
       "step": 609
     },
     {
       "epoch": 0.27156371730662215,
-      "grad_norm": 2.876551389694214,
-      "learning_rate": 6.790563901927907e-05,
-      "loss": 1.946,
+      "grad_norm": 2.8540334701538086,
+      "learning_rate": 6.760825818018508e-05,
+      "loss": 1.904,
       "step": 610
     },
     {
       "epoch": 0.2720089037284363,
-      "grad_norm": 2.3107662200927734,
-      "learning_rate": 6.760825818018508e-05,
-      "loss": 2.3033,
+      "grad_norm": 2.4347307682037354,
+      "learning_rate": 6.731119703450577e-05,
+      "loss": 2.3533,
       "step": 611
     },
     {
       "epoch": 0.2724540901502504,
-      "grad_norm": 1.7891398668289185,
-      "learning_rate": 6.731119703450577e-05,
-      "loss": 1.8722,
+      "grad_norm": 1.7807879447937012,
+      "learning_rate": 6.701445851411472e-05,
+      "loss": 1.869,
       "step": 612
     },
     {
       "epoch": 0.27289927657206453,
-      "grad_norm": 1.788131594657898,
-      "learning_rate": 6.701445851411472e-05,
-      "loss": 1.3048,
+      "grad_norm": 1.9673492908477783,
+      "learning_rate": 6.671804554770135e-05,
+      "loss": 1.321,
       "step": 613
     },
     {
       "epoch": 0.2733444629938787,
-      "grad_norm": 2.7209267616271973,
-      "learning_rate": 6.671804554770135e-05,
-      "loss": 2.1986,
+      "grad_norm": 2.595839262008667,
+      "learning_rate": 6.642196106074194e-05,
+      "loss": 2.2138,
       "step": 614
     },
     {
       "epoch": 0.27378964941569284,
-      "grad_norm": 3.6044182777404785,
-      "learning_rate": 6.642196106074194e-05,
-      "loss": 2.3438,
+      "grad_norm": 3.112882137298584,
+      "learning_rate": 6.612620797547087e-05,
+      "loss": 2.3044,
       "step": 615
     },
     {
       "epoch": 0.27423483583750696,
-      "grad_norm": 3.2944424152374268,
-      "learning_rate": 6.612620797547087e-05,
-      "loss": 2.0854,
+      "grad_norm": 3.4389781951904297,
+      "learning_rate": 6.583078921085167e-05,
+      "loss": 2.1193,
       "step": 616
     },
     {
       "epoch": 0.2746800222593211,
-      "grad_norm": 3.1867449283599854,
-      "learning_rate": 6.583078921085167e-05,
-      "loss": 1.9628,
+      "grad_norm": 5.003129482269287,
+      "learning_rate": 6.55357076825483e-05,
+      "loss": 1.9993,
       "step": 617
     },
     {
       "epoch": 0.2751252086811352,
-      "grad_norm": 1.6597883701324463,
-      "learning_rate": 6.55357076825483e-05,
-      "loss": 1.0473,
+      "grad_norm": 1.3485205173492432,
+      "learning_rate": 6.52409663028963e-05,
+      "loss": 1.0218,
       "step": 618
     },
     {
       "epoch": 0.27557039510294934,
-      "grad_norm": 4.027482986450195,
-      "learning_rate": 6.52409663028963e-05,
-      "loss": 2.1282,
+      "grad_norm": 2.5111501216888428,
+      "learning_rate": 6.494656798087412e-05,
+      "loss": 2.0669,
       "step": 619
     },
     {
       "epoch": 0.2760155815247635,
-      "grad_norm": 2.803708791732788,
-      "learning_rate": 6.494656798087412e-05,
-      "loss": 1.8455,
+      "grad_norm": 2.7159624099731445,
+      "learning_rate": 6.465251562207431e-05,
+      "loss": 1.886,
       "step": 620
     },
     {
       "epoch": 0.27646076794657765,
-      "grad_norm": 3.2381043434143066,
-      "learning_rate": 6.465251562207431e-05,
-      "loss": 2.3748,
+      "grad_norm": 3.150252342224121,
+      "learning_rate": 6.435881212867493e-05,
+      "loss": 2.3275,
       "step": 621
     },
     {
       "epoch": 0.27690595436839177,
-      "grad_norm": 3.580366373062134,
-      "learning_rate": 6.435881212867493e-05,
-      "loss": 2.5845,
+      "grad_norm": 4.344287395477295,
+      "learning_rate": 6.406546039941094e-05,
+      "loss": 2.5912,
       "step": 622
     },
     {
       "epoch": 0.2773511407902059,
-      "grad_norm": 4.986792087554932,
-      "learning_rate": 6.406546039941094e-05,
-      "loss": 2.0802,
+      "grad_norm": 3.325969934463501,
+      "learning_rate": 6.377246332954544e-05,
+      "loss": 2.0392,
       "step": 623
     },
     {
       "epoch": 0.27779632721202,
-      "grad_norm": 1.9550381898880005,
-      "learning_rate": 6.377246332954544e-05,
-      "loss": 1.6063,
+      "grad_norm": 1.8683552742004395,
+      "learning_rate": 6.347982381084123e-05,
+      "loss": 1.6495,
       "step": 624
     },
     {
       "epoch": 0.27824151363383415,
-      "grad_norm": 2.505039930343628,
-      "learning_rate": 6.347982381084123e-05,
-      "loss": 1.6332,
+      "grad_norm": 2.8429386615753174,
+      "learning_rate": 6.318754473153221e-05,
+      "loss": 1.7563,
       "step": 625
     },
     {
       "epoch": 0.27868670005564833,
-      "grad_norm": 1.7756378650665283,
-      "learning_rate": 6.318754473153221e-05,
-      "loss": 1.3104,
+      "grad_norm": 1.8203266859054565,
+      "learning_rate": 6.289562897629492e-05,
+      "loss": 1.2633,
       "step": 626
     },
     {
       "epoch": 0.27913188647746245,
-      "grad_norm": 2.1896908283233643,
-      "learning_rate": 6.289562897629492e-05,
-      "loss": 1.7569,
+      "grad_norm": 2.1837453842163086,
+      "learning_rate": 6.260407942621998e-05,
+      "loss": 1.6992,
       "step": 627
     },
     {
       "epoch": 0.2795770728992766,
-      "grad_norm": 1.8758255243301392,
-      "learning_rate": 6.260407942621998e-05,
-      "loss": 1.756,
+      "grad_norm": 1.8465784788131714,
+      "learning_rate": 6.231289895878375e-05,
+      "loss": 1.7235,
       "step": 628
     },
     {
       "epoch": 0.2800222593210907,
-      "grad_norm": 3.3655426502227783,
-      "learning_rate": 6.231289895878375e-05,
-      "loss": 2.0823,
+      "grad_norm": 3.2776570320129395,
+      "learning_rate": 6.20220904478199e-05,
+      "loss": 2.0223,
       "step": 629
     },
     {
       "epoch": 0.28046744574290483,
-      "grad_norm": 3.293968915939331,
-      "learning_rate": 6.20220904478199e-05,
-      "loss": 1.9035,
+      "grad_norm": 5.350255966186523,
+      "learning_rate": 6.173165676349103e-05,
+      "loss": 1.855,
       "step": 630
     },
     {
       "epoch": 0.28091263216471896,
-      "grad_norm": 3.167731285095215,
-      "learning_rate": 6.173165676349103e-05,
-      "loss": 1.5836,
+      "grad_norm": 2.7620725631713867,
+      "learning_rate": 6.144160077226036e-05,
+      "loss": 1.5578,
       "step": 631
     },
     {
       "epoch": 0.28135781858653314,
-      "grad_norm": 3.473268985748291,
-      "learning_rate": 6.144160077226036e-05,
-      "loss": 1.7259,
+      "grad_norm": 4.09578275680542,
+      "learning_rate": 6.11519253368634e-05,
+      "loss": 1.7264,
       "step": 632
     },
     {
       "epoch": 0.28180300500834726,
-      "grad_norm": 2.428987979888916,
-      "learning_rate": 6.11519253368634e-05,
-      "loss": 1.5845,
+      "grad_norm": 2.946509599685669,
+      "learning_rate": 6.086263331627976e-05,
+      "loss": 1.5729,
       "step": 633
     },
     {
       "epoch": 0.2822481914301614,
-      "grad_norm": 3.5516490936279297,
-      "learning_rate": 6.086263331627976e-05,
-      "loss": 1.5608,
+      "grad_norm": 2.881894588470459,
+      "learning_rate": 6.05737275657049e-05,
+      "loss": 1.6147,
       "step": 634
     },
     {
       "epoch": 0.2826933778519755,
-      "grad_norm": 4.088588714599609,
-      "learning_rate": 6.05737275657049e-05,
-      "loss": 2.4316,
+      "grad_norm": 3.856121778488159,
+      "learning_rate": 6.0285210936521955e-05,
+      "loss": 2.3838,
       "step": 635
     },
     {
-      "epoch": 0.28313856427378964,
-      "grad_norm": 3.0667595863342285,
-      "learning_rate": 6.0285210936521955e-05,
-      "loss": 1.3786,
+      "epoch": 0.28313856427378964,
+      "grad_norm": 3.899914026260376,
+      "learning_rate": 5.999708627627354e-05,
+      "loss": 1.4203,
       "step": 636
     },
     {
       "epoch": 0.28358375069560376,
-      "grad_norm": 2.605963945388794,
-      "learning_rate": 5.999708627627354e-05,
-      "loss": 1.7549,
+      "grad_norm": 2.6259968280792236,
+      "learning_rate": 5.9709356428633746e-05,
+      "loss": 1.7286,
       "step": 637
     },
     {
       "epoch": 0.28402893711741795,
-      "grad_norm": 3.6883420944213867,
-      "learning_rate": 5.9709356428633746e-05,
-      "loss": 2.2541,
+      "grad_norm": 3.2001712322235107,
+      "learning_rate": 5.9422024233380013e-05,
+      "loss": 2.1947,
       "step": 638
     },
     {
       "epoch": 0.28447412353923207,
-      "grad_norm": 2.800830364227295,
-      "learning_rate": 5.9422024233380013e-05,
-      "loss": 1.8305,
+      "grad_norm": 2.7082347869873047,
+      "learning_rate": 5.913509252636511e-05,
+      "loss": 1.7782,
       "step": 639
     },
     {
       "epoch": 0.2849193099610462,
-      "grad_norm": 2.8093373775482178,
-      "learning_rate": 5.913509252636511e-05,
-      "loss": 2.8368,
+      "grad_norm": 2.737236738204956,
+      "learning_rate": 5.884856413948913e-05,
+      "loss": 2.7653,
       "step": 640
     },
     {
       "epoch": 0.2853644963828603,
-      "grad_norm": 2.3125829696655273,
-      "learning_rate": 5.884856413948913e-05,
-      "loss": 1.0694,
+      "grad_norm": 2.4625205993652344,
+      "learning_rate": 5.856244190067159e-05,
+      "loss": 1.1071,
       "step": 641
     },
     {
       "epoch": 0.28580968280467445,
-      "grad_norm": 2.6709086894989014,
-      "learning_rate": 5.856244190067159e-05,
-      "loss": 1.6645,
+      "grad_norm": 2.5604653358459473,
+      "learning_rate": 5.82767286338235e-05,
+      "loss": 1.651,
       "step": 642
     },
     {
       "epoch": 0.2862548692264886,
-      "grad_norm": 3.269890546798706,
-      "learning_rate": 5.82767286338235e-05,
-      "loss": 1.9574,
+      "grad_norm": 2.799736499786377,
+      "learning_rate": 5.799142715881938e-05,
+      "loss": 1.9749,
       "step": 643
     },
     {
       "epoch": 0.28670005564830275,
-      "grad_norm": 6.454814910888672,
-      "learning_rate": 5.799142715881938e-05,
-      "loss": 1.9828,
+      "grad_norm": 3.1290884017944336,
+      "learning_rate": 5.770654029146969e-05,
+      "loss": 1.8804,
       "step": 644
     },
     {
       "epoch": 0.2871452420701169,
-      "grad_norm": 2.483273983001709,
-      "learning_rate": 5.770654029146969e-05,
-      "loss": 1.1946,
+      "grad_norm": 2.4735965728759766,
+      "learning_rate": 5.7422070843492734e-05,
+      "loss": 1.2096,
       "step": 645
     },
     {
       "epoch": 0.287590428491931,
-      "grad_norm": 1.7462166547775269,
-      "learning_rate": 5.7422070843492734e-05,
-      "loss": 1.4047,
+      "grad_norm": 1.8366683721542358,
+      "learning_rate": 5.713802162248718e-05,
+      "loss": 1.3955,
       "step": 646
     },
     {
       "epoch": 0.28803561491374513,
-      "grad_norm": 4.918879508972168,
-      "learning_rate": 5.713802162248718e-05,
-      "loss": 2.362,
+      "grad_norm": 3.0377116203308105,
+      "learning_rate": 5.6854395431904094e-05,
+      "loss": 2.328,
       "step": 647
     },
     {
       "epoch": 0.28848080133555926,
-      "grad_norm": 2.5706145763397217,
-      "learning_rate": 5.6854395431904094e-05,
-      "loss": 2.0059,
+      "grad_norm": 3.051595687866211,
+      "learning_rate": 5.657119507101954e-05,
+      "loss": 1.9927,
       "step": 648
     },
     {
       "epoch": 0.2889259877573734,
-      "grad_norm": 3.0318686962127686,
-      "learning_rate": 5.657119507101954e-05,
-      "loss": 2.7449,
+      "grad_norm": 3.1508376598358154,
+      "learning_rate": 5.6288423334906735e-05,
+      "loss": 2.6943,
       "step": 649
     },
     {
       "epoch": 0.28937117417918756,
-      "grad_norm": 1.3213516473770142,
-      "learning_rate": 5.6288423334906735e-05,
-      "loss": 1.1325,
+      "grad_norm": 1.3692312240600586,
+      "learning_rate": 5.6006083014408484e-05,
+      "loss": 1.151,
       "step": 650
     },
     {
       "epoch": 0.2898163606010017,
-      "grad_norm": 3.3162622451782227,
-      "learning_rate": 5.6006083014408484e-05,
-      "loss": 2.1244,
+      "grad_norm": 3.5010199546813965,
+      "learning_rate": 5.572417689610987e-05,
+      "loss": 2.1075,
       "step": 651
     },
     {
       "epoch": 0.2902615470228158,
-      "grad_norm": 2.8271548748016357,
-      "learning_rate": 5.572417689610987e-05,
-      "loss": 2.6029,
+      "grad_norm": 2.759070873260498,
+      "learning_rate": 5.544270776231038e-05,
+      "loss": 2.575,
       "step": 652
     },
     {
       "epoch": 0.29070673344462994,
-      "grad_norm": 3.4697909355163574,
-      "learning_rate": 5.544270776231038e-05,
-      "loss": 1.8302,
+      "grad_norm": 3.8564367294311523,
+      "learning_rate": 5.5161678390996796e-05,
+      "loss": 1.8527,
       "step": 653
     },
     {
       "epoch": 0.29115191986644406,
-      "grad_norm": 2.9080650806427,
-      "learning_rate": 5.5161678390996796e-05,
-      "loss": 2.4099,
+      "grad_norm": 3.3082222938537598,
+      "learning_rate": 5.488109155581549e-05,
+      "loss": 2.5134,
       "step": 654
     },
     {
       "epoch": 0.2915971062882582,
-      "grad_norm": 2.8306384086608887,
-      "learning_rate": 5.488109155581549e-05,
-      "loss": 2.1103,
+      "grad_norm": 3.0657942295074463,
+      "learning_rate": 5.4600950026045326e-05,
+      "loss": 2.1126,
       "step": 655
     },
     {
       "epoch": 0.2920422927100723,
-      "grad_norm": 3.779367208480835,
-      "learning_rate": 5.4600950026045326e-05,
-      "loss": 2.3853,
+      "grad_norm": 3.670198678970337,
+      "learning_rate": 5.4321256566570036e-05,
+      "loss": 2.4895,
       "step": 656
     },
     {
       "epoch": 0.2924874791318865,
-      "grad_norm": 3.2461814880371094,
-      "learning_rate": 5.4321256566570036e-05,
-      "loss": 1.6209,
+      "grad_norm": 2.923703908920288,
+      "learning_rate": 5.404201393785122e-05,
+      "loss": 1.5419,
       "step": 657
     },
     {
       "epoch": 0.2929326655537006,
-      "grad_norm": 3.401667356491089,
-      "learning_rate": 5.404201393785122e-05,
-      "loss": 1.8184,
+      "grad_norm": 4.151619911193848,
+      "learning_rate": 5.3763224895900846e-05,
+      "loss": 1.8035,
       "step": 658
     },
     {
       "epoch": 0.29337785197551475,
-      "grad_norm": 3.3337910175323486,
-      "learning_rate": 5.3763224895900846e-05,
-      "loss": 2.288,
+      "grad_norm": 3.982928514480591,
+      "learning_rate": 5.348489219225416e-05,
+      "loss": 2.2633,
       "step": 659
     },
     {
       "epoch": 0.2938230383973289,
-      "grad_norm": 4.496949672698975,
-      "learning_rate": 5.348489219225416e-05,
-      "loss": 2.0749,
+      "grad_norm": 4.458322048187256,
+      "learning_rate": 5.320701857394268e-05,
+      "loss": 2.0028,
       "step": 660
     },
     {
       "epoch": 0.294268224819143,
-      "grad_norm": 4.434914588928223,
-      "learning_rate": 5.320701857394268e-05,
-      "loss": 2.3961,
+      "grad_norm": 3.433276653289795,
+      "learning_rate": 5.292960678346675e-05,
+      "loss": 2.3767,
       "step": 661
     },
     {
       "epoch": 0.2947134112409571,
-      "grad_norm": 1.705828070640564,
-      "learning_rate": 5.292960678346675e-05,
-      "loss": 2.0444,
+      "grad_norm": 1.869184970855713,
+      "learning_rate": 5.265265955876879e-05,
+      "loss": 2.0265,
       "step": 662
     },
     {
       "epoch": 0.2951585976627713,
-      "grad_norm": 3.431540012359619,
-      "learning_rate": 5.265265955876879e-05,
-      "loss": 1.4793,
+      "grad_norm": 3.0243453979492188,
+      "learning_rate": 5.237617963320608e-05,
+      "loss": 1.4633,
       "step": 663
     },
     {
       "epoch": 0.29560378408458543,
-      "grad_norm": 2.5585508346557617,
-      "learning_rate": 5.237617963320608e-05,
-      "loss": 1.8598,
+      "grad_norm": 3.066246271133423,
+      "learning_rate": 5.210016973552391e-05,
+      "loss": 1.862,
       "step": 664
     },
     {
       "epoch": 0.29604897050639956,
-      "grad_norm": 2.0425503253936768,
-      "learning_rate": 5.210016973552391e-05,
-      "loss": 1.6394,
+      "grad_norm": 1.9982380867004395,
+      "learning_rate": 5.182463258982846e-05,
+      "loss": 1.6254,
       "step": 665
     },
     {
       "epoch": 0.2964941569282137,
-      "grad_norm": 1.9641846418380737,
-      "learning_rate": 5.182463258982846e-05,
-      "loss": 1.1607,
+      "grad_norm": 1.776050090789795,
+      "learning_rate": 5.1549570915560206e-05,
+      "loss": 1.1617,
       "step": 666
     },
     {
       "epoch": 0.2969393433500278,
-      "grad_norm": 2.71909499168396,
-      "learning_rate": 5.1549570915560206e-05,
-      "loss": 1.9043,
+      "grad_norm": 2.6614208221435547,
+      "learning_rate": 5.127498742746675e-05,
+      "loss": 1.8786,
       "step": 667
     },
     {
       "epoch": 0.29738452977184193,
-      "grad_norm": 1.7965915203094482,
-      "learning_rate": 5.127498742746675e-05,
-      "loss": 1.4473,
+      "grad_norm": 1.8079196214675903,
+      "learning_rate": 5.100088483557634e-05,
+      "loss": 1.4484,
       "step": 668
     },
     {
       "epoch": 0.2978297161936561,
-      "grad_norm": 2.792185068130493,
-      "learning_rate": 5.100088483557634e-05,
-      "loss": 2.1568,
+      "grad_norm": 2.559936285018921,
+      "learning_rate": 5.072726584517086e-05,
+      "loss": 2.1928,
       "step": 669
     },
     {
       "epoch": 0.29827490261547024,
-      "grad_norm": 2.522604465484619,
-      "learning_rate": 5.072726584517086e-05,
-      "loss": 1.4188,
+      "grad_norm": 2.427232503890991,
+      "learning_rate": 5.045413315675924e-05,
+      "loss": 1.4675,
       "step": 670
     },
     {
       "epoch": 0.29872008903728436,
-      "grad_norm": 2.1286098957061768,
-      "learning_rate": 5.045413315675924e-05,
-      "loss": 1.2097,
+      "grad_norm": 2.2989280223846436,
+      "learning_rate": 5.018148946605092e-05,
+      "loss": 1.2586,
       "step": 671
     },
     {
       "epoch": 0.2991652754590985,
-      "grad_norm": 2.812614917755127,
-      "learning_rate": 5.018148946605092e-05,
-      "loss": 2.0791,
+      "grad_norm": 2.920217275619507,
+      "learning_rate": 4.990933746392899e-05,
+      "loss": 2.0559,
       "step": 672
     },
     {
       "epoch": 0.2996104618809126,
-      "grad_norm": 1.8847200870513916,
-      "learning_rate": 4.990933746392899e-05,
-      "loss": 1.6048,
+      "grad_norm": 1.9449331760406494,
+      "learning_rate": 4.9637679836423924e-05,
+      "loss": 1.6264,
       "step": 673
     },
     {
       "epoch": 0.30005564830272674,
-      "grad_norm": 2.700967788696289,
-      "learning_rate": 4.9637679836423924e-05,
-      "loss": 1.9822,
+      "grad_norm": 2.7529637813568115,
+      "learning_rate": 4.9366519264686725e-05,
+      "loss": 1.9742,
       "step": 674
     },
     {
       "epoch": 0.3005008347245409,
-      "grad_norm": 5.683379650115967,
-      "learning_rate": 4.9366519264686725e-05,
-      "loss": 2.3686,
+      "grad_norm": 4.343723297119141,
+      "learning_rate": 4.909585842496287e-05,
+      "loss": 2.4238,
       "step": 675
     },
     {
       "epoch": 0.30094602114635505,
-      "grad_norm": 2.7878029346466064,
-      "learning_rate": 4.909585842496287e-05,
-      "loss": 2.0972,
+      "grad_norm": 2.7013916969299316,
+      "learning_rate": 4.8825699988565485e-05,
+      "loss": 2.1105,
       "step": 676
     },
     {
       "epoch": 0.3013912075681692,
-      "grad_norm": 1.970714807510376,
-      "learning_rate": 4.8825699988565485e-05,
-      "loss": 1.8908,
+      "grad_norm": 2.1688084602355957,
+      "learning_rate": 4.8556046621849346e-05,
+      "loss": 1.8924,
       "step": 677
     },
     {
       "epoch": 0.3018363939899833,
-      "grad_norm": 3.5123045444488525,
-      "learning_rate": 4.8556046621849346e-05,
-      "loss": 2.4987,
+      "grad_norm": 3.5238544940948486,
+      "learning_rate": 4.828690098618429e-05,
+      "loss": 2.4906,
       "step": 678
     },
     {
       "epoch": 0.3022815804117974,
-      "grad_norm": 3.1310598850250244,
-      "learning_rate": 4.828690098618429e-05,
-      "loss": 2.2711,
+      "grad_norm": 2.9248886108398438,
+      "learning_rate": 4.8018265737929044e-05,
+      "loss": 2.2373,
       "step": 679
     },
     {
       "epoch": 0.30272676683361155,
-      "grad_norm": 2.584655284881592,
-      "learning_rate": 4.8018265737929044e-05,
-      "loss": 1.9538,
+      "grad_norm": 2.773230791091919,
+      "learning_rate": 4.7750143528405126e-05,
+      "loss": 1.9962,
       "step": 680
     },
     {
       "epoch": 0.30317195325542573,
-      "grad_norm": 2.380634307861328,
-      "learning_rate": 4.7750143528405126e-05,
-      "loss": 1.322,
+      "grad_norm": 2.1851563453674316,
+      "learning_rate": 4.748253700387042e-05,
+      "loss": 1.3326,
       "step": 681
     },
     {
       "epoch": 0.30361713967723986,
-      "grad_norm": 3.004396438598633,
-      "learning_rate": 4.748253700387042e-05,
-      "loss": 2.1901,
+      "grad_norm": 3.1086177825927734,
+      "learning_rate": 4.721544880549337e-05,
+      "loss": 2.2943,
       "step": 682
     },
     {
       "epoch": 0.304062326099054,
-      "grad_norm": 2.5817222595214844,
-      "learning_rate": 4.721544880549337e-05,
-      "loss": 1.9884,
+      "grad_norm": 2.3710122108459473,
+      "learning_rate": 4.694888156932658e-05,
+      "loss": 1.9881,
       "step": 683
     },
     {
       "epoch": 0.3045075125208681,
-      "grad_norm": 1.8147649765014648,
-      "learning_rate": 4.694888156932658e-05,
-      "loss": 1.6065,
+      "grad_norm": 2.8495447635650635,
+      "learning_rate": 4.668283792628114e-05,
+      "loss": 1.6848,
       "step": 684
     },
     {
       "epoch": 0.30495269894268223,
-      "grad_norm": 2.1298556327819824,
-      "learning_rate": 4.668283792628114e-05,
-      "loss": 2.1248,
+      "grad_norm": 2.106743812561035,
+      "learning_rate": 4.6417320502100316e-05,
+      "loss": 2.1312,
       "step": 685
     },
     {
       "epoch": 0.30539788536449636,
-      "grad_norm": 2.9568581581115723,
-      "learning_rate": 4.6417320502100316e-05,
-      "loss": 2.8798,
+      "grad_norm": 2.764496326446533,
+      "learning_rate": 4.615233191733398e-05,
+      "loss": 2.9265,
       "step": 686
     },
     {
       "epoch": 0.30584307178631054,
-      "grad_norm": 3.2827939987182617,
-      "learning_rate": 4.615233191733398e-05,
-      "loss": 2.6892,
+      "grad_norm": 3.204009771347046,
+      "learning_rate": 4.588787478731242e-05,
+      "loss": 2.6597,
       "step": 687
     },
     {
       "epoch": 0.30628825820812466,
-      "grad_norm": 2.5867035388946533,
-      "learning_rate": 4.588787478731242e-05,
-      "loss": 2.105,
+      "grad_norm": 2.7305068969726562,
+      "learning_rate": 4.5623951722120736e-05,
+      "loss": 2.1332,
       "step": 688
     },
     {
       "epoch": 0.3067334446299388,
-      "grad_norm": 3.4361913204193115,
-      "learning_rate": 4.5623951722120736e-05,
-      "loss": 2.7065,
+      "grad_norm": 3.599640130996704,
+      "learning_rate": 4.5360565326573104e-05,
+      "loss": 2.6595,
       "step": 689
     },
     {
       "epoch": 0.3071786310517529,
-      "grad_norm": 3.75927996635437,
-      "learning_rate": 4.5360565326573104e-05,
-      "loss": 1.6896,
+      "grad_norm": 3.23399019241333,
+      "learning_rate": 4.5097718200186814e-05,
+      "loss": 1.6178,
       "step": 690
     },
     {
       "epoch": 0.30762381747356704,
-      "grad_norm": 3.1970183849334717,
-      "learning_rate": 4.5097718200186814e-05,
-      "loss": 2.5938,
+      "grad_norm": 3.274613618850708,
+      "learning_rate": 4.483541293715698e-05,
+      "loss": 2.5681,
       "step": 691
     },
     {
       "epoch": 0.30806900389538117,
-      "grad_norm": 2.316333770751953,
-      "learning_rate": 4.483541293715698e-05,
-      "loss": 2.1561,
+      "grad_norm": 2.2831473350524902,
+      "learning_rate": 4.457365212633058e-05,
+      "loss": 2.1433,
       "step": 692
     },
     {
       "epoch": 0.30851419031719535,
-      "grad_norm": 5.406709671020508,
-      "learning_rate": 4.457365212633058e-05,
-      "loss": 1.39,
+      "grad_norm": 2.0501179695129395,
+      "learning_rate": 4.431243835118124e-05,
+      "loss": 1.3321,
       "step": 693
     },
     {
       "epoch": 0.3089593767390095,
-      "grad_norm": 1.8153971433639526,
-      "learning_rate": 4.431243835118124e-05,
-      "loss": 1.6361,
+      "grad_norm": 1.8365594148635864,
+      "learning_rate": 4.4051774189783315e-05,
+      "loss": 1.6827,
       "step": 694
     },
     {
       "epoch": 0.3094045631608236,
-      "grad_norm": 3.764073133468628,
-      "learning_rate": 4.4051774189783315e-05,
-      "loss": 2.0475,
+      "grad_norm": 2.847127676010132,
+      "learning_rate": 4.379166221478697e-05,
+      "loss": 2.0208,
       "step": 695
     },
     {
       "epoch": 0.3098497495826377,
-      "grad_norm": 2.5098719596862793,
-      "learning_rate": 4.379166221478697e-05,
-      "loss": 1.5775,
+      "grad_norm": 2.437344551086426,
+      "learning_rate": 4.3532104993392306e-05,
+      "loss": 1.551,
       "step": 696
     },
     {
       "epoch": 0.31029493600445185,
-      "grad_norm": 2.3882479667663574,
-      "learning_rate": 4.3532104993392306e-05,
-      "loss": 2.1238,
+      "grad_norm": 2.4390649795532227,
+      "learning_rate": 4.327310508732437e-05,
+      "loss": 2.1749,
       "step": 697
     },
     {
       "epoch": 0.310740122426266,
-      "grad_norm": 2.4165163040161133,
-      "learning_rate": 4.327310508732437e-05,
-      "loss": 2.2308,
+      "grad_norm": 2.5979692935943604,
+      "learning_rate": 4.301466505280762e-05,
+      "loss": 2.2946,
       "step": 698
     },
     {
       "epoch": 0.31118530884808016,
-      "grad_norm": 2.2906312942504883,
-      "learning_rate": 4.301466505280762e-05,
-      "loss": 2.2186,
+      "grad_norm": 2.2395007610321045,
+      "learning_rate": 4.2756787440540936e-05,
+      "loss": 2.2389,
       "step": 699
     },
     {
       "epoch": 0.3116304952698943,
-      "grad_norm": 3.0668275356292725,
-      "learning_rate": 4.2756787440540936e-05,
-      "loss": 1.7851,
+      "grad_norm": 2.821206569671631,
+      "learning_rate": 4.249947479567218e-05,
+      "loss": 1.8095,
       "step": 700
     },
     {
       "epoch": 0.3120756816917084,
-      "grad_norm": 2.195206642150879,
-      "learning_rate": 4.249947479567218e-05,
-      "loss": 1.6858,
+      "grad_norm": 2.19750714302063,
+      "learning_rate": 4.224272965777326e-05,
+      "loss": 1.6818,
       "step": 701
     },
     {
       "epoch": 0.31252086811352253,
-      "grad_norm": 2.6582798957824707,
-      "learning_rate": 4.224272965777326e-05,
-      "loss": 1.5366,
+      "grad_norm": 2.464684247970581,
+      "learning_rate": 4.1986554560815096e-05,
+      "loss": 1.5108,
       "step": 702
     },
     {
       "epoch": 0.31296605453533666,
-      "grad_norm": 1.982883095741272,
-      "learning_rate": 4.1986554560815096e-05,
-      "loss": 1.6462,
+      "grad_norm": 2.2051353454589844,
+      "learning_rate": 4.173095203314241e-05,
+      "loss": 1.7233,
       "step": 703
     },
     {
       "epoch": 0.3134112409571508,
-      "grad_norm": 2.2738630771636963,
-      "learning_rate": 4.173095203314241e-05,
-      "loss": 1.7683,
+      "grad_norm": 2.5886502265930176,
+      "learning_rate": 4.1475924597449024e-05,
+      "loss": 1.8085,
       "step": 704
     },
     {
       "epoch": 0.31385642737896496,
-      "grad_norm": 2.549669027328491,
-      "learning_rate": 4.1475924597449024e-05,
-      "loss": 2.3111,
+      "grad_norm": 2.4677674770355225,
+      "learning_rate": 4.12214747707527e-05,
+      "loss": 2.2383,
       "step": 705
     },
     {
       "epoch": 0.3143016138007791,
-      "grad_norm": 2.8641653060913086,
-      "learning_rate": 4.12214747707527e-05,
-      "loss": 1.7643,
+      "grad_norm": 2.8653159141540527,
+      "learning_rate": 4.096760506437057e-05,
+      "loss": 1.7662,
       "step": 706
     },
     {
       "epoch": 0.3147468002225932,
-      "grad_norm": 1.6562206745147705,
-      "learning_rate": 4.096760506437057e-05,
-      "loss": 1.6724,
+      "grad_norm": 1.6294842958450317,
+      "learning_rate": 4.071431798389408e-05,
+      "loss": 1.6328,
       "step": 707
     },
     {
       "epoch": 0.31519198664440734,
-      "grad_norm": 2.6927621364593506,
-      "learning_rate": 4.071431798389408e-05,
-      "loss": 1.4428,
+      "grad_norm": 2.834298610687256,
+      "learning_rate": 4.0461616029164526e-05,
+      "loss": 1.4878,
       "step": 708
     },
     {
       "epoch": 0.31563717306622147,
-      "grad_norm": 2.4343008995056152,
-      "learning_rate": 4.0461616029164526e-05,
-      "loss": 1.2359,
+      "grad_norm": 2.328455924987793,
+      "learning_rate": 4.020950169424815e-05,
+      "loss": 1.2361,
       "step": 709
     },
     {
       "epoch": 0.3160823594880356,
-      "grad_norm": 2.6044414043426514,
-      "learning_rate": 4.020950169424815e-05,
-      "loss": 1.8076,
+      "grad_norm": 2.746863842010498,
+      "learning_rate": 3.9957977467411615e-05,
+      "loss": 1.8509,
       "step": 710
     },
     {
       "epoch": 0.3165275459098498,
-      "grad_norm": 4.747753143310547,
-      "learning_rate": 3.9957977467411615e-05,
-      "loss": 1.983,
+      "grad_norm": 3.194289445877075,
+      "learning_rate": 3.9707045831097555e-05,
+      "loss": 1.9686,
       "step": 711
     },
     {
       "epoch": 0.3169727323316639,
-      "grad_norm": 4.754560947418213,
-      "learning_rate": 3.9707045831097555e-05,
-      "loss": 2.7843,
+      "grad_norm": 5.094242572784424,
+      "learning_rate": 3.945670926189987e-05,
+      "loss": 2.8198,
       "step": 712
     },
     {
       "epoch": 0.317417918753478,
-      "grad_norm": 2.5929174423217773,
-      "learning_rate": 3.945670926189987e-05,
-      "loss": 2.3036,
+      "grad_norm": 2.7125210762023926,
+      "learning_rate": 3.920697023053949e-05,
+      "loss": 2.3279,
       "step": 713
     },
     {
       "epoch": 0.31786310517529215,
-      "grad_norm": 3.1126303672790527,
-      "learning_rate": 3.920697023053949e-05,
-      "loss": 2.2455,
+      "grad_norm": 3.3546807765960693,
+      "learning_rate": 3.895783120183976e-05,
+      "loss": 2.2521,
       "step": 714
     },
     {
       "epoch": 0.3183082915971063,
-      "grad_norm": 3.110023260116577,
-      "learning_rate": 3.895783120183976e-05,
-      "loss": 1.8107,
+      "grad_norm": 3.4734482765197754,
+      "learning_rate": 3.8709294634702376e-05,
+      "loss": 1.8874,
       "step": 715
     },
     {
       "epoch": 0.3187534780189204,
-      "grad_norm": 2.862748622894287,
-      "learning_rate": 3.8709294634702376e-05,
-      "loss": 1.9125,
+      "grad_norm": 3.120919942855835,
+      "learning_rate": 3.846136298208285e-05,
+      "loss": 1.9086,
       "step": 716
     },
     {
       "epoch": 0.3191986644407346,
-      "grad_norm": 2.372352361679077,
-      "learning_rate": 3.846136298208285e-05,
-      "loss": 1.9124,
+      "grad_norm": 2.432473659515381,
+      "learning_rate": 3.821403869096658e-05,
+      "loss": 1.9659,
       "step": 717
     },
     {
       "epoch": 0.3196438508625487,
-      "grad_norm": 3.2221293449401855,
-      "learning_rate": 3.821403869096658e-05,
-      "loss": 1.8325,
+      "grad_norm": 3.055663585662842,
+      "learning_rate": 3.796732420234443e-05,
+      "loss": 1.8126,
       "step": 718
     },
     {
       "epoch": 0.32008903728436283,
-      "grad_norm": 2.2187700271606445,
-      "learning_rate": 3.796732420234443e-05,
-      "loss": 1.388,
+      "grad_norm": 2.0997750759124756,
+      "learning_rate": 3.7721221951188765e-05,
+      "loss": 1.3377,
       "step": 719
     },
     {
       "epoch": 0.32053422370617696,
-      "grad_norm": 2.560734510421753,
-      "learning_rate": 3.7721221951188765e-05,
-      "loss": 1.374,
+      "grad_norm": 2.715914487838745,
+      "learning_rate": 3.747573436642951e-05,
+      "loss": 1.4096,
       "step": 720
     },
     {
       "epoch": 0.3209794101279911,
-      "grad_norm": 2.4914464950561523,
-      "learning_rate": 3.747573436642951e-05,
-      "loss": 2.0269,
+      "grad_norm": 2.5205202102661133,
+      "learning_rate": 3.7230863870929964e-05,
+      "loss": 1.9885,
       "step": 721
     },
     {
       "epoch": 0.3214245965498052,
-      "grad_norm": 2.3180055618286133,
-      "learning_rate": 3.7230863870929964e-05,
-      "loss": 2.0662,
+      "grad_norm": 2.2237253189086914,
+      "learning_rate": 3.698661288146311e-05,
+      "loss": 2.0693,
       "step": 722
     },
     {
       "epoch": 0.3218697829716194,
-      "grad_norm": 2.14855694770813,
-      "learning_rate": 3.698661288146311e-05,
-      "loss": 1.3372,
+      "grad_norm": 2.0891478061676025,
+      "learning_rate": 3.674298380868756e-05,
+      "loss": 1.304,
       "step": 723
     },
     {
       "epoch": 0.3223149693934335,
-      "grad_norm": 3.403369426727295,
-      "learning_rate": 3.674298380868756e-05,
-      "loss": 2.3522,
+      "grad_norm": 3.3524973392486572,
+      "learning_rate": 3.649997905712396e-05,
+      "loss": 2.2968,
       "step": 724
     },
     {
       "epoch": 0.32276015581524764,
-      "grad_norm": 6.225140571594238,
-      "learning_rate": 3.649997905712396e-05,
-      "loss": 1.1184,
+      "grad_norm": 2.234802722930908,
+      "learning_rate": 3.6257601025131026e-05,
+      "loss": 1.2229,
       "step": 725
     },
     {
       "epoch": 0.32320534223706177,
-      "grad_norm": 1.9929250478744507,
-      "learning_rate": 3.6257601025131026e-05,
-      "loss": 1.4805,
+      "grad_norm": 2.095039129257202,
+      "learning_rate": 3.601585210488218e-05,
+      "loss": 1.4692,
       "step": 726
     },
     {
       "epoch": 0.3236505286588759,
-      "grad_norm": 1.9394887685775757,
-      "learning_rate": 3.601585210488218e-05,
-      "loss": 1.9297,
+      "grad_norm": 1.8265604972839355,
+      "learning_rate": 3.577473468234156e-05,
+      "loss": 1.9447,
       "step": 727
     },
     {
       "epoch": 0.32409571508069,
-      "grad_norm": 2.5257530212402344,
-      "learning_rate": 3.577473468234156e-05,
-      "loss": 1.6671,
+      "grad_norm": 1.9364336729049683,
+      "learning_rate": 3.553425113724088e-05,
+      "loss": 1.5901,
       "step": 728
     },
     {
       "epoch": 0.3245409015025042,
-      "grad_norm": 3.4196643829345703,
-      "learning_rate": 3.553425113724088e-05,
-      "loss": 2.7964,
+      "grad_norm": 3.5687179565429688,
+      "learning_rate": 3.52944038430556e-05,
+      "loss": 2.8283,
       "step": 729
     },
     {
       "epoch": 0.3249860879243183,
-      "grad_norm": 2.4060604572296143,
-      "learning_rate": 3.52944038430556e-05,
-      "loss": 1.7696,
+      "grad_norm": 2.2928483486175537,
+      "learning_rate": 3.5055195166981645e-05,
+      "loss": 1.8126,
       "step": 730
     },
     {
       "epoch": 0.32543127434613245,
-      "grad_norm": 2.9572129249572754,
-      "learning_rate": 3.5055195166981645e-05,
-      "loss": 1.6015,
+      "grad_norm": 2.6834230422973633,
+      "learning_rate": 3.481662746991214e-05,
+      "loss": 1.5685,
       "step": 731
     },
     {
       "epoch": 0.3258764607679466,
-      "grad_norm": 2.2669548988342285,
-      "learning_rate": 3.481662746991214e-05,
-      "loss": 2.3187,
+      "grad_norm": 2.1744472980499268,
+      "learning_rate": 3.4578703106413904e-05,
+      "loss": 2.3411,
       "step": 732
     },
     {
       "epoch": 0.3263216471897607,
-      "grad_norm": 4.523863792419434,
-      "learning_rate": 3.4578703106413904e-05,
-      "loss": 3.4186,
+      "grad_norm": 4.877960205078125,
+      "learning_rate": 3.4341424424704375e-05,
+      "loss": 3.4117,
       "step": 733
     },
     {
       "epoch": 0.3267668336115748,
-      "grad_norm": 2.4980828762054443,
-      "learning_rate": 3.4341424424704375e-05,
-      "loss": 1.3333,
+      "grad_norm": 2.355116844177246,
+      "learning_rate": 3.4104793766628304e-05,
+      "loss": 1.321,
       "step": 734
     },
     {
       "epoch": 0.327212020033389,
-      "grad_norm": 1.4893592596054077,
-      "learning_rate": 3.4104793766628304e-05,
-      "loss": 1.2549,
+      "grad_norm": 1.589286208152771,
+      "learning_rate": 3.386881346763483e-05,
+      "loss": 1.2843,
       "step": 735
     },
     {
       "epoch": 0.32765720645520313,
-      "grad_norm": 2.372570753097534,
-      "learning_rate": 3.386881346763483e-05,
-      "loss": 1.3731,
+      "grad_norm": 4.362774848937988,
+      "learning_rate": 3.363348585675414e-05,
+      "loss": 1.4054,
       "step": 736
     },
     {
       "epoch": 0.32810239287701726,
-      "grad_norm": 1.388790249824524,
-      "learning_rate": 3.363348585675414e-05,
-      "loss": 1.0765,
+      "grad_norm": 1.3995791673660278,
+      "learning_rate": 3.339881325657484e-05,
+      "loss": 1.0818,
       "step": 737
     },
     {
       "epoch": 0.3285475792988314,
-      "grad_norm": 3.7632062435150146,
-      "learning_rate": 3.339881325657484e-05,
-      "loss": 2.1809,
+      "grad_norm": 2.9872403144836426,
+      "learning_rate": 3.316479798322072e-05,
+      "loss": 2.1892,
       "step": 738
     },
     {
       "epoch": 0.3289927657206455,
-      "grad_norm": 3.2846572399139404,
-      "learning_rate": 3.316479798322072e-05,
-      "loss": 1.7677,
+      "grad_norm": 3.312777042388916,
+      "learning_rate": 3.2931442346328004e-05,
+      "loss": 1.8328,
       "step": 739
     },
     {
       "epoch": 0.32943795214245963,
-      "grad_norm": 9.13051986694336,
-      "learning_rate": 3.2931442346328004e-05,
-      "loss": 1.6814,
+      "grad_norm": 2.1134469509124756,
+      "learning_rate": 3.269874864902269e-05,
+      "loss": 1.6851,
       "step": 740
     },
     {
       "epoch": 0.3298831385642738,
-      "grad_norm": 1.820805311203003,
-      "learning_rate": 3.269874864902269e-05,
-      "loss": 1.8115,
+      "grad_norm": 1.7910422086715698,
+      "learning_rate": 3.246671918789755e-05,
+      "loss": 1.8255,
       "step": 741
     },
     {
       "epoch": 0.33032832498608794,
-      "grad_norm": 4.556207180023193,
-      "learning_rate": 3.246671918789755e-05,
-      "loss": 2.7417,
+      "grad_norm": 4.106746196746826,
+      "learning_rate": 3.223535625298979e-05,
+      "loss": 2.6915,
       "step": 742
     },
     {
       "epoch": 0.33077351140790207,
-      "grad_norm": 2.135354995727539,
-      "learning_rate": 3.223535625298979e-05,
-      "loss": 1.5414,
+      "grad_norm": 2.1480355262756348,
+      "learning_rate": 3.200466212775808e-05,
+      "loss": 1.5085,
       "step": 743
     },
     {
       "epoch": 0.3312186978297162,
-      "grad_norm": 12.734091758728027,
-      "learning_rate": 3.200466212775808e-05,
-      "loss": 1.7974,
+      "grad_norm": 3.8113248348236084,
+      "learning_rate": 3.1774639089060363e-05,
+      "loss": 1.797,
       "step": 744
     },
     {
       "epoch": 0.3316638842515303,
-      "grad_norm": 2.428943634033203,
-      "learning_rate": 3.1774639089060363e-05,
-      "loss": 1.732,
+      "grad_norm": 2.449369430541992,
+      "learning_rate": 3.154528940713113e-05,
+      "loss": 1.7518,
       "step": 745
     },
     {
       "epoch": 0.33210907067334444,
-      "grad_norm": 2.2564094066619873,
-      "learning_rate": 3.154528940713113e-05,
-      "loss": 1.8338,
+      "grad_norm": 2.2306559085845947,
+      "learning_rate": 3.1316615345559185e-05,
+      "loss": 1.7888,
       "step": 746
     },
     {
       "epoch": 0.3325542570951586,
-      "grad_norm": 2.3983113765716553,
-      "learning_rate": 3.1316615345559185e-05,
-      "loss": 1.7076,
+      "grad_norm": 2.4093217849731445,
+      "learning_rate": 3.108861916126518e-05,
+      "loss": 1.7471,
       "step": 747
     },
     {
       "epoch": 0.33299944351697275,
-      "grad_norm": 1.7274196147918701,
-      "learning_rate": 3.108861916126518e-05,
-      "loss": 1.2812,
+      "grad_norm": 1.6302881240844727,
+      "learning_rate": 3.086130310447937e-05,
+      "loss": 1.2229,
       "step": 748
     },
     {
       "epoch": 0.3334446299387869,
-      "grad_norm": 2.3063457012176514,
-      "learning_rate": 3.086130310447937e-05,
-      "loss": 2.0274,
+      "grad_norm": 2.4067959785461426,
+      "learning_rate": 3.063466941871952e-05,
+      "loss": 2.0926,
       "step": 749
     },
     {
       "epoch": 0.333889816360601,
-      "grad_norm": 2.640960931777954,
-      "learning_rate": 3.063466941871952e-05,
-      "loss": 1.4717,
+      "grad_norm": 2.2470877170562744,
+      "learning_rate": 3.0408720340768572e-05,
+      "loss": 1.4138,
       "step": 750
     },
     {
       "epoch": 0.3343350027824151,
-      "grad_norm": 2.7002780437469482,
-      "learning_rate": 3.0408720340768572e-05,
-      "loss": 2.0358,
+      "grad_norm": 2.531765937805176,
+      "learning_rate": 3.018345810065275e-05,
+      "loss": 2.0289,
       "step": 751
     },
     {
       "epoch": 0.33478018920422925,
-      "grad_norm": 2.129476547241211,
-      "learning_rate": 3.018345810065275e-05,
-      "loss": 1.7658,
+      "grad_norm": 2.2208447456359863,
+      "learning_rate": 2.9958884921619367e-05,
+      "loss": 1.7873,
       "step": 752
     },
     {
       "epoch": 0.33522537562604343,
-      "grad_norm": 3.172895669937134,
-      "learning_rate": 2.9958884921619367e-05,
-      "loss": 2.3525,
+      "grad_norm": 3.0990521907806396,
+      "learning_rate": 2.9735003020115092e-05,
+      "loss": 2.318,
       "step": 753
     },
     {
       "epoch": 0.33567056204785756,
-      "grad_norm": 3.458550453186035,
-      "learning_rate": 2.9735003020115092e-05,
-      "loss": 2.2845,
+      "grad_norm": 3.5115699768066406,
+      "learning_rate": 2.9511814605763855e-05,
+      "loss": 2.3048,
       "step": 754
     },
     {
       "epoch": 0.3361157484696717,
-      "grad_norm": 2.624788284301758,
-      "learning_rate": 2.9511814605763855e-05,
-      "loss": 2.0448,
+      "grad_norm": 2.530822515487671,
+      "learning_rate": 2.9289321881345254e-05,
+      "loss": 2.09,
       "step": 755
     },
     {
       "epoch": 0.3365609348914858,
-      "grad_norm": 4.643282413482666,
-      "learning_rate": 2.9289321881345254e-05,
-      "loss": 2.8465,
+      "grad_norm": 4.581959247589111,
+      "learning_rate": 2.9067527042772636e-05,
+      "loss": 2.8179,
       "step": 756
     },
     {
       "epoch": 0.33700612131329993,
-      "grad_norm": 2.141824960708618,
-      "learning_rate": 2.9067527042772636e-05,
-      "loss": 1.795,
+      "grad_norm": 2.0303454399108887,
+      "learning_rate": 2.8846432279071467e-05,
+      "loss": 1.7881,
       "step": 757
     },
     {
       "epoch": 0.33745130773511406,
-      "grad_norm": 1.945853352546692,
-      "learning_rate": 2.8846432279071467e-05,
-      "loss": 1.4118,
+      "grad_norm": 2.0643062591552734,
+      "learning_rate": 2.8626039772357882e-05,
+      "loss": 1.4213,
       "step": 758
     },
     {
       "epoch": 0.33789649415692824,
-      "grad_norm": 2.440788507461548,
-      "learning_rate": 2.8626039772357882e-05,
-      "loss": 1.8631,
+      "grad_norm": 2.3261709213256836,
+      "learning_rate": 2.840635169781688e-05,
+      "loss": 1.8827,
       "step": 759
     },
     {
       "epoch": 0.33834168057874237,
-      "grad_norm": 2.524973154067993,
-      "learning_rate": 2.840635169781688e-05,
-      "loss": 1.9311,
+      "grad_norm": 2.7478396892547607,
+      "learning_rate": 2.8187370223681132e-05,
+      "loss": 1.9566,
       "step": 760
     },
     {
       "epoch": 0.3387868670005565,
-      "grad_norm": 2.88173508644104,
-      "learning_rate": 2.8187370223681132e-05,
-      "loss": 1.8423,
+      "grad_norm": 2.7778799533843994,
+      "learning_rate": 2.7969097511209308e-05,
+      "loss": 1.8509,
       "step": 761
     },
     {
       "epoch": 0.3392320534223706,
-      "grad_norm": 9.78935718536377,
-      "learning_rate": 2.7969097511209308e-05,
-      "loss": 1.7553,
+      "grad_norm": 3.287660598754883,
+      "learning_rate": 2.775153571466502e-05,
+      "loss": 1.7451,
       "step": 762
     },
     {
       "epoch": 0.33967723984418474,
-      "grad_norm": 4.4023356437683105,
-      "learning_rate": 2.775153571466502e-05,
-      "loss": 2.0394,
+      "grad_norm": 4.2368268966674805,
+      "learning_rate": 2.753468698129533e-05,
+      "loss": 2.0127,
       "step": 763
     },
     {
       "epoch": 0.34012242626599887,
-      "grad_norm": 2.8718042373657227,
-      "learning_rate": 2.753468698129533e-05,
-      "loss": 1.5894,
+      "grad_norm": 2.852130889892578,
+      "learning_rate": 2.7318553451309726e-05,
+      "loss": 1.6444,
       "step": 764
     },
     {
       "epoch": 0.34056761268781305,
-      "grad_norm": 1.903527855873108,
-      "learning_rate": 2.7318553451309726e-05,
-      "loss": 1.3534,
+      "grad_norm": 1.9484717845916748,
+      "learning_rate": 2.7103137257858868e-05,
+      "loss": 1.3003,
       "step": 765
     },
     {
       "epoch": 0.3410127991096272,
-      "grad_norm": 2.23759388923645,
-      "learning_rate": 2.7103137257858868e-05,
-      "loss": 1.871,
+      "grad_norm": 2.2469749450683594,
+      "learning_rate": 2.688844052701359e-05,
+      "loss": 1.8344,
       "step": 766
     },
     {
       "epoch": 0.3414579855314413,
-      "grad_norm": 3.0346007347106934,
-      "learning_rate": 2.688844052701359e-05,
-      "loss": 2.4884,
+      "grad_norm": 3.1364622116088867,
+      "learning_rate": 2.6674465377744017e-05,
+      "loss": 2.5384,
       "step": 767
     },
     {
       "epoch": 0.3419031719532554,
-      "grad_norm": 2.180615186691284,
-      "learning_rate": 2.6674465377744017e-05,
-      "loss": 1.3251,
+      "grad_norm": 1.9614194631576538,
+      "learning_rate": 2.646121392189841e-05,
+      "loss": 1.3528,
       "step": 768
     },
     {
       "epoch": 0.34234835837506955,
-      "grad_norm": 5.008700847625732,
-      "learning_rate": 2.646121392189841e-05,
-      "loss": 2.3742,
+      "grad_norm": 6.426572322845459,
+      "learning_rate": 2.624868826418262e-05,
+      "loss": 2.3194,
       "step": 769
     },
     {
       "epoch": 0.3427935447968837,
-      "grad_norm": 2.506002426147461,
-      "learning_rate": 2.624868826418262e-05,
-      "loss": 1.756,
+      "grad_norm": 2.49417781829834,
+      "learning_rate": 2.603689050213902e-05,
+      "loss": 1.7783,
       "step": 770
     },
     {
       "epoch": 0.34323873121869786,
-      "grad_norm": 3.1411874294281006,
-      "learning_rate": 2.603689050213902e-05,
-      "loss": 2.8913,
+      "grad_norm": 2.8509533405303955,
+      "learning_rate": 2.582582272612609e-05,
+      "loss": 2.9057,
       "step": 771
     },
     {
       "epoch": 0.343683917640512,
-      "grad_norm": 2.647324800491333,
-      "learning_rate": 2.582582272612609e-05,
-      "loss": 2.6637,
+      "grad_norm": 3.2252373695373535,
+      "learning_rate": 2.561548701929749e-05,
+      "loss": 2.6851,
       "step": 772
     },
     {
       "epoch": 0.3441291040623261,
-      "grad_norm": 2.1316630840301514,
-      "learning_rate": 2.561548701929749e-05,
-      "loss": 1.3719,
+      "grad_norm": 2.132861375808716,
+      "learning_rate": 2.540588545758179e-05,
+      "loss": 1.3824,
       "step": 773
     },
     {
       "epoch": 0.34457429048414023,
-      "grad_norm": 3.7572953701019287,
-      "learning_rate": 2.540588545758179e-05,
-      "loss": 1.7199,
+      "grad_norm": 4.197037220001221,
+      "learning_rate": 2.5197020109661772e-05,
+      "loss": 1.6613,
       "step": 774
     },
     {
       "epoch": 0.34501947690595436,
-      "grad_norm": 7.4356369972229,
-      "learning_rate": 2.5197020109661772e-05,
-      "loss": 1.4686,
+      "grad_norm": 2.0976722240448,
+      "learning_rate": 2.4988893036954043e-05,
+      "loss": 1.4904,
       "step": 775
     },
     {
       "epoch": 0.3454646633277685,
-      "grad_norm": 2.2626800537109375,
-      "learning_rate": 2.4988893036954043e-05,
-      "loss": 1.8888,
+      "grad_norm": 2.280292510986328,
+      "learning_rate": 2.4781506293588873e-05,
+      "loss": 1.8493,
       "step": 776
     },
     {
       "epoch": 0.34590984974958267,
-      "grad_norm": 3.9464282989501953,
-      "learning_rate": 2.4781506293588873e-05,
-      "loss": 2.2344,
+      "grad_norm": 4.074158191680908,
+      "learning_rate": 2.4574861926389615e-05,
+      "loss": 2.3215,
       "step": 777
     },
     {
       "epoch": 0.3463550361713968,
-      "grad_norm": 3.29831862449646,
-      "learning_rate": 2.4574861926389615e-05,
-      "loss": 2.4997,
+      "grad_norm": 3.060152769088745,
+      "learning_rate": 2.436896197485282e-05,
+      "loss": 2.5575,
       "step": 778
     },
     {
       "epoch": 0.3468002225932109,
-      "grad_norm": 2.7223682403564453,
-      "learning_rate": 2.436896197485282e-05,
-      "loss": 2.3799,
+      "grad_norm": 2.5705432891845703,
+      "learning_rate": 2.4163808471127812e-05,
+      "loss": 2.3104,
       "step": 779
     },
     {
       "epoch": 0.34724540901502504,
-      "grad_norm": 2.570220947265625,
-      "learning_rate": 2.4163808471127812e-05,
-      "loss": 1.6605,
+      "grad_norm": 2.781761884689331,
+      "learning_rate": 2.3959403439996907e-05,
+      "loss": 1.7279,
       "step": 780
     },
     {
       "epoch": 0.34769059543683917,
-      "grad_norm": 2.4514687061309814,
-      "learning_rate": 2.3959403439996907e-05,
-      "loss": 1.3153,
+      "grad_norm": 2.3290603160858154,
+      "learning_rate": 2.37557488988552e-05,
+      "loss": 1.3618,
       "step": 781
     },
     {
       "epoch": 0.3481357818586533,
-      "grad_norm": 4.2457733154296875,
-      "learning_rate": 2.37557488988552e-05,
-      "loss": 2.2081,
+      "grad_norm": 2.3041293621063232,
+      "learning_rate": 2.3552846857690846e-05,
+      "loss": 2.2214,
       "step": 782
     },
     {
       "epoch": 0.3485809682804674,
-      "grad_norm": 2.6355888843536377,
-      "learning_rate": 2.3552846857690846e-05,
-      "loss": 2.1335,
+      "grad_norm": 2.8323187828063965,
+      "learning_rate": 2.3350699319065026e-05,
+      "loss": 2.1625,
       "step": 783
     },
     {
       "epoch": 0.3490261547022816,
-      "grad_norm": 3.205240488052368,
-      "learning_rate": 2.3350699319065026e-05,
-      "loss": 2.0054,
+      "grad_norm": 2.8128693103790283,
+      "learning_rate": 2.3149308278092342e-05,
+      "loss": 2.0094,
       "step": 784
     },
     {
       "epoch": 0.3494713411240957,
-      "grad_norm": 2.160902976989746,
-      "learning_rate": 2.3149308278092342e-05,
-      "loss": 1.7905,
+      "grad_norm": 2.1774494647979736,
+      "learning_rate": 2.2948675722421086e-05,
+      "loss": 1.8026,
       "step": 785
     },
     {
       "epoch": 0.34991652754590985,
-      "grad_norm": 2.5220730304718018,
-      "learning_rate": 2.2948675722421086e-05,
-      "loss": 1.7417,
+      "grad_norm": 2.4783859252929688,
+      "learning_rate": 2.2748803632213557e-05,
+      "loss": 1.7025,
       "step": 786
     },
     {
       "epoch": 0.350361713967724,
-      "grad_norm": 4.050189018249512,
-      "learning_rate": 2.2748803632213557e-05,
-      "loss": 2.9296,
+      "grad_norm": 3.7645058631896973,
+      "learning_rate": 2.254969398012663e-05,
+      "loss": 2.9119,
       "step": 787
     },
     {
       "epoch": 0.3508069003895381,
-      "grad_norm": 1.7380865812301636,
-      "learning_rate": 2.254969398012663e-05,
-      "loss": 1.3506,
+      "grad_norm": 1.7522432804107666,
+      "learning_rate": 2.235134873129213e-05,
+      "loss": 1.3106,
       "step": 788
     },
     {
       "epoch": 0.3512520868113522,
-      "grad_norm": 3.115523099899292,
-      "learning_rate": 2.235134873129213e-05,
-      "loss": 2.3948,
+      "grad_norm": 2.829122543334961,
+      "learning_rate": 2.2153769843297667e-05,
+      "loss": 2.4193,
       "step": 789
     },
     {
       "epoch": 0.3516972732331664,
-      "grad_norm": 2.615837812423706,
-      "learning_rate": 2.2153769843297667e-05,
-      "loss": 1.865,
+      "grad_norm": 2.5540711879730225,
+      "learning_rate": 2.195695926616702e-05,
+      "loss": 1.8474,
       "step": 790
     },
     {
       "epoch": 0.35214245965498053,
-      "grad_norm": 2.9471611976623535,
-      "learning_rate": 2.195695926616702e-05,
-      "loss": 1.7431,
+      "grad_norm": 3.003200054168701,
+      "learning_rate": 2.1760918942341192e-05,
+      "loss": 1.8296,
       "step": 791
     },
     {
       "epoch": 0.35258764607679466,
-      "grad_norm": 2.246422529220581,
-      "learning_rate": 2.1760918942341192e-05,
-      "loss": 1.0834,
+      "grad_norm": 2.2994792461395264,
+      "learning_rate": 2.1565650806658975e-05,
+      "loss": 1.1001,
       "step": 792
     },
     {
       "epoch": 0.3530328324986088,
-      "grad_norm": 3.3071248531341553,
-      "learning_rate": 2.1565650806658975e-05,
-      "loss": 2.3679,
+      "grad_norm": 2.9518396854400635,
+      "learning_rate": 2.137115678633811e-05,
+      "loss": 2.4107,
       "step": 793
     },
     {
       "epoch": 0.3534780189204229,
-      "grad_norm": 1.6390622854232788,
-      "learning_rate": 2.137115678633811e-05,
-      "loss": 1.3217,
+      "grad_norm": 1.6138437986373901,
+      "learning_rate": 2.1177438800956007e-05,
+      "loss": 1.3177,
       "step": 794
     },
     {
       "epoch": 0.35392320534223703,
-      "grad_norm": 4.03280782699585,
-      "learning_rate": 2.1177438800956007e-05,
-      "loss": 1.9452,
+      "grad_norm": 3.4545297622680664,
+      "learning_rate": 2.098449876243096e-05,
+      "loss": 1.8813,
       "step": 795
     },
     {
       "epoch": 0.3543683917640512,
-      "grad_norm": 15.388529777526855,
-      "learning_rate": 2.098449876243096e-05,
-      "loss": 2.3951,
+      "grad_norm": 3.7924232482910156,
+      "learning_rate": 2.07923385750033e-05,
+      "loss": 2.3388,
       "step": 796
     },
     {
       "epoch": 0.35481357818586534,
-      "grad_norm": 3.017653703689575,
-      "learning_rate": 2.07923385750033e-05,
-      "loss": 1.4409,
+      "grad_norm": 2.952394485473633,
+      "learning_rate": 2.0600960135216462e-05,
+      "loss": 1.493,
       "step": 797
     },
     {
       "epoch": 0.35525876460767947,
-      "grad_norm": 2.619394540786743,
-      "learning_rate": 2.0600960135216462e-05,
-      "loss": 1.7445,
+      "grad_norm": 2.618921995162964,
+      "learning_rate": 2.0410365331898416e-05,
+      "loss": 1.7359,
       "step": 798
     },
     {
       "epoch": 0.3557039510294936,
-      "grad_norm": 3.4079883098602295,
-      "learning_rate": 2.0410365331898416e-05,
-      "loss": 2.4086,
+      "grad_norm": 3.029841184616089,
+      "learning_rate": 2.0220556046142893e-05,
+      "loss": 2.2872,
       "step": 799
     },
     {
       "epoch": 0.3561491374513077,
-      "grad_norm": 1.3419363498687744,
-      "learning_rate": 2.0220556046142893e-05,
-      "loss": 0.9088,
+      "grad_norm": 1.313881516456604,
+      "learning_rate": 2.0031534151290943e-05,
+      "loss": 0.8875,
       "step": 800
     },
     {
       "epoch": 0.35659432387312184,
-      "grad_norm": 3.1316444873809814,
-      "learning_rate": 2.0031534151290943e-05,
-      "loss": 1.4489,
+      "grad_norm": 2.9203662872314453,
+      "learning_rate": 1.9843301512912327e-05,
+      "loss": 1.4551,
       "step": 801
     },
     {
       "epoch": 0.357039510294936,
-      "grad_norm": 2.7865233421325684,
-      "learning_rate": 1.9843301512912327e-05,
-      "loss": 1.5738,
+      "grad_norm": 6.416241645812988,
+      "learning_rate": 1.965585998878724e-05,
+      "loss": 1.5389,
       "step": 802
     },
     {
       "epoch": 0.35748469671675015,
-      "grad_norm": 2.195303201675415,
-      "learning_rate": 1.965585998878724e-05,
-      "loss": 1.1416,
+      "grad_norm": 2.0492238998413086,
+      "learning_rate": 1.946921142888781e-05,
+      "loss": 1.1651,
       "step": 803
     },
     {
       "epoch": 0.3579298831385643,
-      "grad_norm": 1.7528187036514282,
-      "learning_rate": 1.946921142888781e-05,
-      "loss": 1.3798,
+      "grad_norm": 1.750148892402649,
+      "learning_rate": 1.928335767535997e-05,
+      "loss": 1.3607,
       "step": 804
     },
     {
       "epoch": 0.3583750695603784,
-      "grad_norm": 1.894110083580017,
-      "learning_rate": 1.928335767535997e-05,
-      "loss": 1.6355,
+      "grad_norm": 1.8103892803192139,
+      "learning_rate": 1.9098300562505266e-05,
+      "loss": 1.6197,
       "step": 805
     },
     {
       "epoch": 0.3588202559821925,
-      "grad_norm": 3.0293827056884766,
-      "learning_rate": 1.9098300562505266e-05,
-      "loss": 2.377,
+      "grad_norm": 3.195742607116699,
+      "learning_rate": 1.891404191676265e-05,
+      "loss": 2.3358,
       "step": 806
     },
     {
       "epoch": 0.35926544240400665,
-      "grad_norm": 3.3845605850219727,
-      "learning_rate": 1.891404191676265e-05,
-      "loss": 1.7129,
+      "grad_norm": 3.590696096420288,
+      "learning_rate": 1.8730583556690605e-05,
+      "loss": 1.6947,
       "step": 807
     },
     {
       "epoch": 0.35971062882582083,
-      "grad_norm": 2.3603949546813965,
-      "learning_rate": 1.8730583556690605e-05,
-      "loss": 1.7646,
+      "grad_norm": 2.2915265560150146,
+      "learning_rate": 1.854792729294905e-05,
+      "loss": 1.7559,
       "step": 808
     },
     {
       "epoch": 0.36015581524763496,
-      "grad_norm": 2.9297170639038086,
-      "learning_rate": 1.854792729294905e-05,
-      "loss": 2.088,
+      "grad_norm": 2.9653847217559814,
+      "learning_rate": 1.8366074928281607e-05,
+      "loss": 2.1048,
       "step": 809
     },
     {
       "epoch": 0.3606010016694491,
-      "grad_norm": 2.4962754249572754,
-      "learning_rate": 1.8366074928281607e-05,
-      "loss": 1.7159,
+      "grad_norm": 2.5163042545318604,
+      "learning_rate": 1.818502825749764e-05,
+      "loss": 1.7355,
       "step": 810
     },
     {
       "epoch": 0.3610461880912632,
-      "grad_norm": 2.170757532119751,
-      "learning_rate": 1.818502825749764e-05,
-      "loss": 1.5805,
+      "grad_norm": 1.8697280883789062,
+      "learning_rate": 1.8004789067454764e-05,
+      "loss": 1.5879,
       "step": 811
     },
     {
       "epoch": 0.36149137451307733,
-      "grad_norm": 3.127051830291748,
-      "learning_rate": 1.8004789067454764e-05,
-      "loss": 2.4222,
+      "grad_norm": 3.06601619720459,
+      "learning_rate": 1.7825359137040988e-05,
+      "loss": 2.4572,
       "step": 812
     },
     {
       "epoch": 0.36193656093489146,
-      "grad_norm": 2.367252826690674,
-      "learning_rate": 1.7825359137040988e-05,
-      "loss": 2.109,
+      "grad_norm": 2.5294203758239746,
+      "learning_rate": 1.7646740237157256e-05,
+      "loss": 2.0837,
       "step": 813
     },
     {
       "epoch": 0.36238174735670564,
-      "grad_norm": 1.9050239324569702,
-      "learning_rate": 1.7646740237157256e-05,
-      "loss": 1.2665,
+      "grad_norm": 1.993363857269287,
+      "learning_rate": 1.7468934130700044e-05,
+      "loss": 1.2753,
       "step": 814
     },
     {
       "epoch": 0.36282693377851977,
-      "grad_norm": 2.2031283378601074,
-      "learning_rate": 1.7468934130700044e-05,
-      "loss": 1.2599,
+      "grad_norm": 2.295287609100342,
+      "learning_rate": 1.7291942572543807e-05,
+      "loss": 1.3016,
       "step": 815
     },
     {
       "epoch": 0.3632721202003339,
-      "grad_norm": 3.291349172592163,
-      "learning_rate": 1.7291942572543807e-05,
-      "loss": 2.1776,
+      "grad_norm": 3.3601508140563965,
+      "learning_rate": 1.7115767309523812e-05,
+      "loss": 2.2146,
       "step": 816
     },
     {
       "epoch": 0.363717306622148,
-      "grad_norm": 4.794470310211182,
-      "learning_rate": 1.7115767309523812e-05,
-      "loss": 2.1294,
+      "grad_norm": 3.0300424098968506,
+      "learning_rate": 1.6940410080418723e-05,
+      "loss": 2.2092,
       "step": 817
     },
     {
       "epoch": 0.36416249304396214,
-      "grad_norm": 1.8986549377441406,
-      "learning_rate": 1.6940410080418723e-05,
-      "loss": 1.5655,
+      "grad_norm": 1.729293704032898,
+      "learning_rate": 1.6765872615933677e-05,
+      "loss": 1.5451,
       "step": 818
     },
     {
       "epoch": 0.36460767946577627,
-      "grad_norm": 3.300102472305298,
-      "learning_rate": 1.6765872615933677e-05,
-      "loss": 2.4501,
+      "grad_norm": 3.0013277530670166,
+      "learning_rate": 1.6592156638682886e-05,
+      "loss": 2.537,
       "step": 819
     },
     {
       "epoch": 0.36505286588759045,
-      "grad_norm": 2.3774209022521973,
-      "learning_rate": 1.6592156638682886e-05,
-      "loss": 2.0107,
+      "grad_norm": 2.4126980304718018,
+      "learning_rate": 1.6419263863172997e-05,
+      "loss": 1.9869,
       "step": 820
     },
     {
       "epoch": 0.3654980523094046,
-      "grad_norm": 1.869045615196228,
-      "learning_rate": 1.6419263863172997e-05,
-      "loss": 1.6384,
+      "grad_norm": 1.8156296014785767,
+      "learning_rate": 1.6247195995785837e-05,
+      "loss": 1.6322,
       "step": 821
     },
     {
       "epoch": 0.3659432387312187,
-      "grad_norm": 2.376424551010132,
-      "learning_rate": 1.6247195995785837e-05,
-      "loss": 1.6731,
+      "grad_norm": 2.5480713844299316,
+      "learning_rate": 1.6075954734761845e-05,
+      "loss": 1.6801,
       "step": 822
     },
     {
       "epoch": 0.3663884251530328,
-      "grad_norm": 2.211214303970337,
-      "learning_rate": 1.6075954734761845e-05,
-      "loss": 1.6778,
+      "grad_norm": 2.2491114139556885,
+      "learning_rate": 1.5905541770183096e-05,
+      "loss": 1.6766,
       "step": 823
     },
     {
       "epoch": 0.36683361157484695,
-      "grad_norm": 2.649627685546875,
-      "learning_rate": 1.5905541770183096e-05,
-      "loss": 2.41,
+      "grad_norm": 2.629528284072876,
+      "learning_rate": 1.5735958783956794e-05,
+      "loss": 2.3958,
       "step": 824
     },
     {
       "epoch": 0.3672787979966611,
-      "grad_norm": 1.5385890007019043,
-      "learning_rate": 1.5735958783956794e-05,
-      "loss": 1.1177,
+      "grad_norm": 1.5279780626296997,
+      "learning_rate": 1.5567207449798515e-05,
+      "loss": 1.1355,
       "step": 825
     },
     {
       "epoch": 0.36772398441847526,
-      "grad_norm": 2.307979106903076,
-      "learning_rate": 1.5567207449798515e-05,
-      "loss": 1.6581,
+      "grad_norm": 2.3699798583984375,
+      "learning_rate": 1.539928943321579e-05,
+      "loss": 1.6162,
       "step": 826
     },
     {
       "epoch": 0.3681691708402894,
-      "grad_norm": 2.140033006668091,
-      "learning_rate": 1.539928943321579e-05,
-      "loss": 1.9136,
+      "grad_norm": 2.1733739376068115,
+      "learning_rate": 1.5232206391491699e-05,
+      "loss": 1.8902,
       "step": 827
     },
     {
       "epoch": 0.3686143572621035,
-      "grad_norm": 3.384453773498535,
-      "learning_rate": 1.5232206391491699e-05,
-      "loss": 2.8381,
+      "grad_norm": 3.5806779861450195,
+      "learning_rate": 1.5065959973668353e-05,
+      "loss": 2.8429,
       "step": 828
     },
     {
       "epoch": 0.36905954368391763,
-      "grad_norm": 2.655250072479248,
-      "learning_rate": 1.5065959973668353e-05,
-      "loss": 1.5791,
+      "grad_norm": 2.532106876373291,
+      "learning_rate": 1.4900551820530828e-05,
+      "loss": 1.5751,
       "step": 829
     },
     {
       "epoch": 0.36950473010573176,
-      "grad_norm": 1.3919587135314941,
-      "learning_rate": 1.4900551820530828e-05,
-      "loss": 1.1324,
+      "grad_norm": 1.3643015623092651,
+      "learning_rate": 1.4735983564590783e-05,
+      "loss": 1.1445,
       "step": 830
     },
     {
       "epoch": 0.3699499165275459,
-      "grad_norm": 4.468016147613525,
-      "learning_rate": 1.4735983564590783e-05,
-      "loss": 2.4508,
+      "grad_norm": 4.0691819190979,
+      "learning_rate": 1.4572256830070497e-05,
+      "loss": 2.4074,
       "step": 831
     },
     {
       "epoch": 0.37039510294936007,
-      "grad_norm": 2.089130163192749,
-      "learning_rate": 1.4572256830070497e-05,
-      "loss": 1.5146,
+      "grad_norm": 2.234241485595703,
+      "learning_rate": 1.4409373232886702e-05,
+      "loss": 1.5424,
       "step": 832
     },
     {
       "epoch": 0.3708402893711742,
-      "grad_norm": 1.788549780845642,
-      "learning_rate": 1.4409373232886702e-05,
-      "loss": 1.2323,
+      "grad_norm": 1.9068703651428223,
+      "learning_rate": 1.4247334380634792e-05,
+      "loss": 1.2191,
       "step": 833
     },
     {
       "epoch": 0.3712854757929883,
-      "grad_norm": 3.411046266555786,
-      "learning_rate": 1.4247334380634792e-05,
-      "loss": 2.3234,
+      "grad_norm": 3.4625084400177,
+      "learning_rate": 1.4086141872572789e-05,
+      "loss": 2.333,
       "step": 834
     },
     {
       "epoch": 0.37173066221480244,
-      "grad_norm": 3.8933844566345215,
-      "learning_rate": 1.4086141872572789e-05,
-      "loss": 1.9897,
+      "grad_norm": 2.7276928424835205,
+      "learning_rate": 1.3925797299605647e-05,
+      "loss": 1.8938,
       "step": 835
     },
     {
       "epoch": 0.37217584863661657,
-      "grad_norm": 2.343925952911377,
-      "learning_rate": 1.3925797299605647e-05,
-      "loss": 1.4688,
+      "grad_norm": 2.2975573539733887,
+      "learning_rate": 1.3766302244269624e-05,
+      "loss": 1.4582,
       "step": 836
     },
     {
       "epoch": 0.3726210350584307,
-      "grad_norm": 3.1801838874816895,
-      "learning_rate": 1.3766302244269624e-05,
-      "loss": 2.4004,
+      "grad_norm": 3.531290292739868,
+      "learning_rate": 1.3607658280716473e-05,
+      "loss": 2.3825,
       "step": 837
     },
     {
       "epoch": 0.3730662214802449,
-      "grad_norm": 1.550551414489746,
-      "learning_rate": 1.3607658280716473e-05,
-      "loss": 1.0341,
+      "grad_norm": 1.5778701305389404,
+      "learning_rate": 1.3449866974698122e-05,
+      "loss": 1.0383,
       "step": 838
     },
     {
       "epoch": 0.373511407902059,
-      "grad_norm": 2.117818593978882,
-      "learning_rate": 1.3449866974698122e-05,
-      "loss": 1.7987,
+      "grad_norm": 1.928511619567871,
+      "learning_rate": 1.3292929883550998e-05,
+      "loss": 1.798,
       "step": 839
     },
     {
       "epoch": 0.3739565943238731,
-      "grad_norm": 4.381692409515381,
-      "learning_rate": 1.3292929883550998e-05,
-      "loss": 2.4703,
+      "grad_norm": 4.289917469024658,
+      "learning_rate": 1.3136848556180892e-05,
+      "loss": 2.4614,
       "step": 840
     },
     {
       "epoch": 0.37440178074568725,
-      "grad_norm": 2.0128989219665527,
-      "learning_rate": 1.3136848556180892e-05,
-      "loss": 1.8252,
+      "grad_norm": 2.0936968326568604,
+      "learning_rate": 1.2981624533047432e-05,
+      "loss": 1.8072,
       "step": 841
     },
     {
       "epoch": 0.3748469671675014,
-      "grad_norm": 1.8894273042678833,
-      "learning_rate": 1.2981624533047432e-05,
-      "loss": 1.7092,
+      "grad_norm": 1.897525429725647,
+      "learning_rate": 1.2827259346149122e-05,
+      "loss": 1.6898,
       "step": 842
     },
     {
       "epoch": 0.3752921535893155,
-      "grad_norm": 3.7159740924835205,
-      "learning_rate": 1.2827259346149122e-05,
-      "loss": 2.0822,
+      "grad_norm": 3.6533870697021484,
+      "learning_rate": 1.2673754519008008e-05,
+      "loss": 2.0014,
       "step": 843
     },
     {
       "epoch": 0.3757373400111297,
-      "grad_norm": 3.0898549556732178,
-      "learning_rate": 1.2673754519008008e-05,
-      "loss": 2.6043,
+      "grad_norm": 3.2471959590911865,
+      "learning_rate": 1.2521111566654731e-05,
+      "loss": 2.6087,
       "step": 844
     },
     {
       "epoch": 0.3761825264329438,
-      "grad_norm": 2.4676640033721924,
-      "learning_rate": 1.2521111566654731e-05,
-      "loss": 2.4497,
+      "grad_norm": 2.473552703857422,
+      "learning_rate": 1.2369331995613665e-05,
+      "loss": 2.4239,
       "step": 845
     },
     {
       "epoch": 0.37662771285475793,
-      "grad_norm": 3.5587337017059326,
-      "learning_rate": 1.2369331995613665e-05,
-      "loss": 1.6656,
+      "grad_norm": 3.8729305267333984,
+      "learning_rate": 1.2218417303887842e-05,
+      "loss": 1.6321,
       "step": 846
     },
     {
       "epoch": 0.37707289927657206,
-      "grad_norm": 7.141547679901123,
-      "learning_rate": 1.2218417303887842e-05,
-      "loss": 1.7547,
+      "grad_norm": 2.8680293560028076,
+      "learning_rate": 1.206836898094439e-05,
+      "loss": 1.7427,
       "step": 847
     },
     {
       "epoch": 0.3775180856983862,
-      "grad_norm": 1.7326031923294067,
-      "learning_rate": 1.206836898094439e-05,
-      "loss": 1.4759,
+      "grad_norm": 1.601547122001648,
+      "learning_rate": 1.191918850769964e-05,
+      "loss": 1.4566,
       "step": 848
     },
     {
       "epoch": 0.3779632721202003,
-      "grad_norm": 3.1554577350616455,
-      "learning_rate": 1.191918850769964e-05,
-      "loss": 2.4164,
+      "grad_norm": 3.3136868476867676,
+      "learning_rate": 1.1770877356504683e-05,
+      "loss": 2.4073,
       "step": 849
     },
     {
       "epoch": 0.3784084585420145,
-      "grad_norm": 1.9731732606887817,
-      "learning_rate": 1.1770877356504683e-05,
-      "loss": 1.0614,
+      "grad_norm": 1.8992241621017456,
+      "learning_rate": 1.1623436991130654e-05,
+      "loss": 1.0409,
       "step": 850
     },
     {
       "epoch": 0.3788536449638286,
-      "grad_norm": 2.997244119644165,
-      "learning_rate": 1.1623436991130654e-05,
-      "loss": 2.0257,
+      "grad_norm": 2.848297595977783,
+      "learning_rate": 1.1476868866754486e-05,
+      "loss": 2.008,
       "step": 851
     },
     {
       "epoch": 0.37929883138564274,
-      "grad_norm": 5.286291599273682,
-      "learning_rate": 1.1476868866754486e-05,
-      "loss": 2.0731,
+      "grad_norm": 5.80834436416626,
+      "learning_rate": 1.1331174429944347e-05,
+      "loss": 2.0277,
       "step": 852
     },
     {
       "epoch": 0.37974401780745687,
-      "grad_norm": 2.4582889080047607,
-      "learning_rate": 1.1331174429944347e-05,
-      "loss": 1.406,
+      "grad_norm": 2.033008575439453,
+      "learning_rate": 1.1186355118645554e-05,
+      "loss": 1.3786,
       "step": 853
     },
     {
       "epoch": 0.380189204229271,
-      "grad_norm": 2.530067205429077,
-      "learning_rate": 1.1186355118645554e-05,
-      "loss": 1.6258,
+      "grad_norm": 2.5872247219085693,
+      "learning_rate": 1.1042412362166222e-05,
+      "loss": 1.7005,
       "step": 854
     },
     {
       "epoch": 0.3806343906510851,
-      "grad_norm": 3.028555393218994,
-      "learning_rate": 1.1042412362166222e-05,
-      "loss": 2.8568,
+      "grad_norm": 3.4668571949005127,
+      "learning_rate": 1.0899347581163221e-05,
+      "loss": 2.8316,
       "step": 855
     },
     {
       "epoch": 0.3810795770728993,
-      "grad_norm": 2.7910170555114746,
-      "learning_rate": 1.0899347581163221e-05,
-      "loss": 1.8254,
+      "grad_norm": 2.7666592597961426,
+      "learning_rate": 1.0757162187628222e-05,
+      "loss": 1.8368,
       "step": 856
     },
     {
       "epoch": 0.3815247634947134,
-      "grad_norm": 39.85447692871094,
-      "learning_rate": 1.0757162187628222e-05,
-      "loss": 2.8449,
+      "grad_norm": 4.064805030822754,
+      "learning_rate": 1.0615857584873623e-05,
+      "loss": 2.7671,
       "step": 857
     },
     {
       "epoch": 0.38196994991652755,
-      "grad_norm": 3.291005849838257,
-      "learning_rate": 1.0615857584873623e-05,
-      "loss": 2.0754,
+      "grad_norm": 2.88974666595459,
+      "learning_rate": 1.0475435167518843e-05,
+      "loss": 2.0095,
       "step": 858
     },
     {
       "epoch": 0.3824151363383417,
-      "grad_norm": 4.250852584838867,
-      "learning_rate": 1.0475435167518843e-05,
-      "loss": 2.1429,
+      "grad_norm": 4.876190662384033,
+      "learning_rate": 1.0335896321476413e-05,
+      "loss": 2.1341,
       "step": 859
     },
     {
       "epoch": 0.3828603227601558,
-      "grad_norm": 2.714912176132202,
-      "learning_rate": 1.0335896321476413e-05,
-      "loss": 1.7616,
+      "grad_norm": 3.0236520767211914,
+      "learning_rate": 1.0197242423938446e-05,
+      "loss": 1.7555,
       "step": 860
     },
     {
       "epoch": 0.38330550918196993,
-      "grad_norm": 2.5076630115509033,
-      "learning_rate": 1.0197242423938446e-05,
-      "loss": 1.864,
+      "grad_norm": 2.457872152328491,
+      "learning_rate": 1.0059474843362892e-05,
+      "loss": 1.8167,
       "step": 861
     },
     {
       "epoch": 0.3837506956037841,
-      "grad_norm": 3.874208450317383,
-      "learning_rate": 1.0059474843362892e-05,
-      "loss": 3.0877,
+      "grad_norm": 3.7359426021575928,
+      "learning_rate": 9.922594939460194e-06,
+      "loss": 3.0536,
       "step": 862
     },
     {
       "epoch": 0.38419588202559823,
-      "grad_norm": 3.2415525913238525,
-      "learning_rate": 9.922594939460194e-06,
-      "loss": 2.3881,
+      "grad_norm": 3.145775556564331,
+      "learning_rate": 9.786604063179728e-06,
+      "loss": 2.3369,
       "step": 863
     },
     {
       "epoch": 0.38464106844741236,
-      "grad_norm": 2.6872646808624268,
-      "learning_rate": 9.786604063179728e-06,
-      "loss": 1.953,
+      "grad_norm": 2.927863597869873,
+      "learning_rate": 9.651503556696516e-06,
+      "loss": 1.9876,
       "step": 864
     },
     {
       "epoch": 0.3850862548692265,
-      "grad_norm": 2.7855000495910645,
-      "learning_rate": 9.651503556696516e-06,
-      "loss": 2.1969,
+      "grad_norm": 2.7322986125946045,
+      "learning_rate": 9.517294753398064e-06,
+      "loss": 2.1556,
       "step": 865
     },
     {
       "epoch": 0.3855314412910406,
-      "grad_norm": 3.2661259174346924,
-      "learning_rate": 9.517294753398064e-06,
-      "loss": 2.2943,
+      "grad_norm": 3.337082862854004,
+      "learning_rate": 9.383978977871021e-06,
+      "loss": 2.3229,
       "step": 866
     },
     {
       "epoch": 0.38597662771285474,
-      "grad_norm": 2.2386388778686523,
-      "learning_rate": 9.383978977871021e-06,
-      "loss": 1.3356,
+      "grad_norm": 2.271052598953247,
+      "learning_rate": 9.251557545888312e-06,
+      "loss": 1.3285,
       "step": 867
     },
     {
       "epoch": 0.3864218141346689,
-      "grad_norm": 2.699187755584717,
-      "learning_rate": 9.251557545888312e-06,
-      "loss": 2.4968,
+      "grad_norm": 3.045947790145874,
+      "learning_rate": 9.120031764395987e-06,
+      "loss": 2.5944,
       "step": 868
     },
     {
       "epoch": 0.38686700055648304,
-      "grad_norm": 2.743082046508789,
-      "learning_rate": 9.120031764395987e-06,
-      "loss": 4.0225,
+      "grad_norm": 2.6999034881591797,
+      "learning_rate": 8.989402931500434e-06,
+      "loss": 4.0293,
       "step": 869
     },
     {
       "epoch": 0.38731218697829717,
-      "grad_norm": 2.0932228565216064,
-      "learning_rate": 8.989402931500434e-06,
-      "loss": 1.4646,
+      "grad_norm": 1.9985932111740112,
+      "learning_rate": 8.85967233645547e-06,
+      "loss": 1.4189,
       "step": 870
     },
     {
       "epoch": 0.3877573734001113,
-      "grad_norm": 2.3609516620635986,
-      "learning_rate": 8.85967233645547e-06,
-      "loss": 2.0499,
+      "grad_norm": 2.2697956562042236,
+      "learning_rate": 8.730841259649725e-06,
+      "loss": 2.1461,
       "step": 871
     },
     {
       "epoch": 0.3882025598219254,
-      "grad_norm": 8.27224063873291,
-      "learning_rate": 8.730841259649725e-06,
-      "loss": 2.7773,
+      "grad_norm": 7.573481559753418,
+      "learning_rate": 8.602910972593892e-06,
+      "loss": 2.7597,
       "step": 872
     },
     {
       "epoch": 0.38864774624373954,
-      "grad_norm": 2.892641544342041,
-      "learning_rate": 8.602910972593892e-06,
-      "loss": 2.2208,
+      "grad_norm": 3.326749086380005,
+      "learning_rate": 8.475882737908248e-06,
+      "loss": 2.156,
       "step": 873
     },
     {
       "epoch": 0.3890929326655537,
-      "grad_norm": 2.17576265335083,
-      "learning_rate": 8.475882737908248e-06,
-      "loss": 1.9523,
+      "grad_norm": 2.2427914142608643,
+      "learning_rate": 8.34975780931021e-06,
+      "loss": 1.9286,
       "step": 874
     },
     {
       "epoch": 0.38953811908736785,
-      "grad_norm": 3.0030341148376465,
-      "learning_rate": 8.34975780931021e-06,
-      "loss": 2.3068,
+      "grad_norm": 3.511756181716919,
+      "learning_rate": 8.224537431601886e-06,
+      "loss": 2.37,
       "step": 875
     },
     {
       "epoch": 0.389983305509182,
-      "grad_norm": 3.5457253456115723,
-      "learning_rate": 8.224537431601886e-06,
-      "loss": 1.9086,
+      "grad_norm": 3.885831356048584,
+      "learning_rate": 8.100222840657878e-06,
+      "loss": 1.9569,
       "step": 876
     },
     {
       "epoch": 0.3904284919309961,
-      "grad_norm": 2.699582815170288,
-      "learning_rate": 8.100222840657878e-06,
-      "loss": 2.1184,
+      "grad_norm": 2.932791233062744,
+      "learning_rate": 7.976815263412963e-06,
+      "loss": 2.1272,
       "step": 877
     },
     {
       "epoch": 0.3908736783528102,
-      "grad_norm": 2.5596392154693604,
-      "learning_rate": 7.976815263412963e-06,
-      "loss": 1.7771,
+      "grad_norm": 2.672161102294922,
+      "learning_rate": 7.854315917850163e-06,
+      "loss": 1.8023,
       "step": 878
     },
     {
       "epoch": 0.39131886477462435,
-      "grad_norm": 2.6684186458587646,
-      "learning_rate": 7.854315917850163e-06,
-      "loss": 1.7569,
+      "grad_norm": 2.610121488571167,
+      "learning_rate": 7.73272601298851e-06,
+      "loss": 1.7252,
       "step": 879
     },
     {
       "epoch": 0.39176405119643853,
-      "grad_norm": 2.5746593475341797,
-      "learning_rate": 7.73272601298851e-06,
-      "loss": 2.0132,
+      "grad_norm": 2.4881973266601562,
+      "learning_rate": 7.612046748871327e-06,
+      "loss": 2.007,
       "step": 880
     },
     {
       "epoch": 0.39220923761825266,
-      "grad_norm": 2.5082550048828125,
-      "learning_rate": 7.612046748871327e-06,
-      "loss": 1.7144,
+      "grad_norm": 2.8911471366882324,
+      "learning_rate": 7.492279316554207e-06,
+      "loss": 1.7142,
       "step": 881
     },
     {
       "epoch": 0.3926544240400668,
-      "grad_norm": 1.8464329242706299,
-      "learning_rate": 7.492279316554207e-06,
-      "loss": 1.1433,
+      "grad_norm": 1.823811411857605,
+      "learning_rate": 7.3734248980933395e-06,
+      "loss": 1.1544,
       "step": 882
     },
     {
       "epoch": 0.3930996104618809,
-      "grad_norm": 3.3372461795806885,
-      "learning_rate": 7.3734248980933395e-06,
-      "loss": 2.8916,
+      "grad_norm": 3.5623137950897217,
+      "learning_rate": 7.255484666533874e-06,
+      "loss": 2.8305,
       "step": 883
     },
     {
       "epoch": 0.39354479688369504,
-      "grad_norm": 3.623055934906006,
-      "learning_rate": 7.255484666533874e-06,
-      "loss": 1.7107,
+      "grad_norm": 2.1440281867980957,
+      "learning_rate": 7.138459785898266e-06,
+      "loss": 1.6987,
       "step": 884
     },
     {
       "epoch": 0.39398998330550916,
-      "grad_norm": 2.4824955463409424,
-      "learning_rate": 7.138459785898266e-06,
-      "loss": 1.79,
+      "grad_norm": 2.6287648677825928,
+      "learning_rate": 7.022351411174866e-06,
+      "loss": 1.7781,
       "step": 885
     },
     {
       "epoch": 0.39443516972732334,
-      "grad_norm": 2.6194679737091064,
-      "learning_rate": 7.022351411174866e-06,
-      "loss": 1.6792,
+      "grad_norm": 2.2390339374542236,
+      "learning_rate": 6.907160688306425e-06,
+      "loss": 1.7095,
       "step": 886
     },
     {
       "epoch": 0.39488035614913747,
-      "grad_norm": 2.1351640224456787,
-      "learning_rate": 6.907160688306425e-06,
-      "loss": 1.992,
+      "grad_norm": 2.168720245361328,
+      "learning_rate": 6.7928887541789055e-06,
+      "loss": 2.0094,
       "step": 887
     },
     {
       "epoch": 0.3953255425709516,
-      "grad_norm": 3.6620707511901855,
-      "learning_rate": 6.7928887541789055e-06,
-      "loss": 1.9295,
+      "grad_norm": 2.6436660289764404,
+      "learning_rate": 6.679536736610137e-06,
+      "loss": 1.9816,
       "step": 888
     },
     {
       "epoch": 0.3957707289927657,
-      "grad_norm": 2.735880136489868,
-      "learning_rate": 6.679536736610137e-06,
-      "loss": 2.5191,
+      "grad_norm": 2.9050190448760986,
+      "learning_rate": 6.5671057543387985e-06,
+      "loss": 2.5014,
       "step": 889
     },
     {
       "epoch": 0.39621591541457984,
-      "grad_norm": 2.25370192527771,
-      "learning_rate": 6.5671057543387985e-06,
-      "loss": 1.7509,
+      "grad_norm": 2.3916468620300293,
+      "learning_rate": 6.455596917013273e-06,
+      "loss": 1.7533,
       "step": 890
     },
     {
       "epoch": 0.39666110183639397,
-      "grad_norm": 3.4820597171783447,
-      "learning_rate": 6.455596917013273e-06,
-      "loss": 2.0117,
+      "grad_norm": 2.5925655364990234,
+      "learning_rate": 6.345011325180772e-06,
+      "loss": 2.0431,
       "step": 891
     },
     {
       "epoch": 0.39710628825820815,
-      "grad_norm": 2.8001556396484375,
-      "learning_rate": 6.345011325180772e-06,
-      "loss": 1.6458,
+      "grad_norm": 2.6889488697052,
+      "learning_rate": 6.235350070276447e-06,
+      "loss": 1.6722,
       "step": 892
     },
     {
       "epoch": 0.3975514746800223,
-      "grad_norm": 1.6623258590698242,
-      "learning_rate": 6.235350070276447e-06,
-      "loss": 1.6371,
+      "grad_norm": 1.676353931427002,
+      "learning_rate": 6.126614234612593e-06,
+      "loss": 1.617,
       "step": 893
     },
     {
       "epoch": 0.3979966611018364,
-      "grad_norm": 2.0735392570495605,
-      "learning_rate": 6.126614234612593e-06,
-      "loss": 1.8299,
+      "grad_norm": 2.1050057411193848,
+      "learning_rate": 6.018804891368035e-06,
+      "loss": 1.8159,
       "step": 894
     },
     {
       "epoch": 0.3984418475236505,
-      "grad_norm": 2.833833694458008,
-      "learning_rate": 6.018804891368035e-06,
-      "loss": 1.6262,
+      "grad_norm": 2.851128101348877,
+      "learning_rate": 5.911923104577455e-06,
+      "loss": 1.6424,
       "step": 895
     },
     {
       "epoch": 0.39888703394546465,
-      "grad_norm": 1.4123960733413696,
-      "learning_rate": 5.911923104577455e-06,
-      "loss": 1.2736,
+      "grad_norm": 1.3776746988296509,
+      "learning_rate": 5.805969929120947e-06,
+      "loss": 1.2788,
       "step": 896
     },
     {
       "epoch": 0.3993322203672788,
-      "grad_norm": 4.113855838775635,
-      "learning_rate": 5.805969929120947e-06,
-      "loss": 2.449,
+      "grad_norm": 4.327455520629883,
+      "learning_rate": 5.700946410713548e-06,
+      "loss": 2.3312,
       "step": 897
     },
     {
       "epoch": 0.39977740678909296,
-      "grad_norm": 4.152754306793213,
-      "learning_rate": 5.700946410713548e-06,
-      "loss": 1.9736,
+      "grad_norm": 2.629577159881592,
+      "learning_rate": 5.5968535858950345e-06,
+      "loss": 1.8854,
       "step": 898
     },
     {
       "epoch": 0.4002225932109071,
-      "grad_norm": 2.434448480606079,
-      "learning_rate": 5.5968535858950345e-06,
-      "loss": 2.2586,
+      "grad_norm": 2.564117670059204,
+      "learning_rate": 5.49369248201953e-06,
+      "loss": 2.2783,
       "step": 899
     },
     {
       "epoch": 0.4006677796327212,
-      "grad_norm": 2.730743169784546,
-      "learning_rate": 5.49369248201953e-06,
-      "loss": 2.1945,
+      "grad_norm": 2.6416311264038086,
+      "learning_rate": 5.39146411724547e-06,
+      "loss": 2.1379,
       "step": 900
     },
     {
       "epoch": 0.40111296605453534,
-      "grad_norm": 2.3305442333221436,
-      "learning_rate": 5.39146411724547e-06,
-      "loss": 1.844,
+      "grad_norm": 2.190145254135132,
+      "learning_rate": 5.290169500525577e-06,
+      "loss": 1.7978,
       "step": 901
     },
     {
       "epoch": 0.40155815247634946,
-      "grad_norm": 2.125002861022949,
-      "learning_rate": 5.290169500525577e-06,
-      "loss": 1.8199,
+      "grad_norm": 2.145873785018921,
+      "learning_rate": 5.189809631596798e-06,
+      "loss": 1.7841,
       "step": 902
     },
     {
       "epoch": 0.4020033388981636,
-      "grad_norm": 2.352309226989746,
-      "learning_rate": 5.189809631596798e-06,
-      "loss": 2.143,
+      "grad_norm": 2.5149333477020264,
+      "learning_rate": 5.0903855009705514e-06,
+      "loss": 2.2296,
       "step": 903
     },
     {
       "epoch": 0.40244852531997777,
-      "grad_norm": 2.7642769813537598,
-      "learning_rate": 5.0903855009705514e-06,
-      "loss": 1.75,
+      "grad_norm": 2.739926815032959,
+      "learning_rate": 4.991898089922819e-06,
+      "loss": 1.7751,
       "step": 904
     },
     {
       "epoch": 0.4028937117417919,
-      "grad_norm": 4.074885845184326,
-      "learning_rate": 4.991898089922819e-06,
-      "loss": 2.1392,
+      "grad_norm": 3.803953170776367,
+      "learning_rate": 4.8943483704846475e-06,
+      "loss": 2.1071,
       "step": 905
     },
     {
       "epoch": 0.403338898163606,
-      "grad_norm": 2.652728319168091,
-      "learning_rate": 4.8943483704846475e-06,
-      "loss": 1.9152,
+      "grad_norm": 5.201776027679443,
+      "learning_rate": 4.797737305432337e-06,
+      "loss": 2.0633,
       "step": 906
     },
     {
       "epoch": 0.40378408458542014,
-      "grad_norm": 2.720862627029419,
-      "learning_rate": 4.797737305432337e-06,
-      "loss": 2.3358,
+      "grad_norm": 2.998929738998413,
+      "learning_rate": 4.702065848278126e-06,
+      "loss": 2.3406,
       "step": 907
     },
     {
       "epoch": 0.40422927100723427,
-      "grad_norm": 2.838879346847534,
-      "learning_rate": 4.702065848278126e-06,
-      "loss": 2.1309,
+      "grad_norm": 2.622312068939209,
+      "learning_rate": 4.607334943260655e-06,
+      "loss": 2.0794,
       "step": 908
     },
     {
       "epoch": 0.4046744574290484,
-      "grad_norm": 2.1942715644836426,
-      "learning_rate": 4.607334943260655e-06,
-      "loss": 1.4136,
+      "grad_norm": 2.2471535205841064,
+      "learning_rate": 4.513545525335705e-06,
+      "loss": 1.4662,
       "step": 909
     },
     {
       "epoch": 0.4051196438508626,
-      "grad_norm": 3.1751365661621094,
-      "learning_rate": 4.513545525335705e-06,
-      "loss": 2.0721,
+      "grad_norm": 3.173253059387207,
+      "learning_rate": 4.420698520166988e-06,
+      "loss": 2.133,
       "step": 910
     },
     {
       "epoch": 0.4055648302726767,
-      "grad_norm": 2.386066436767578,
-      "learning_rate": 4.420698520166988e-06,
-      "loss": 1.7174,
+      "grad_norm": 2.3637583255767822,
+      "learning_rate": 4.328794844116946e-06,
+      "loss": 1.7295,
       "step": 911
     },
     {
       "epoch": 0.4060100166944908,
-      "grad_norm": 2.3867030143737793,
-      "learning_rate": 4.328794844116946e-06,
-      "loss": 1.9815,
+      "grad_norm": 2.423197031021118,
+      "learning_rate": 4.237835404237778e-06,
+      "loss": 1.9654,
       "step": 912
     },
     {
       "epoch": 0.40645520311630495,
-      "grad_norm": 2.3891232013702393,
-      "learning_rate": 4.237835404237778e-06,
-      "loss": 1.9543,
+      "grad_norm": 2.4029150009155273,
+      "learning_rate": 4.147821098262405e-06,
+      "loss": 1.9868,
       "step": 913
     },
     {
       "epoch": 0.4069003895381191,
-      "grad_norm": 2.0092246532440186,
-      "learning_rate": 4.147821098262405e-06,
-      "loss": 1.7221,
+      "grad_norm": 2.146833658218384,
+      "learning_rate": 4.0587528145957235e-06,
+      "loss": 1.7029,
       "step": 914
     },
     {
       "epoch": 0.4073455759599332,
-      "grad_norm": 13.39091968536377,
-      "learning_rate": 4.0587528145957235e-06,
-      "loss": 1.7269,
+      "grad_norm": 5.961406707763672,
+      "learning_rate": 3.970631432305694e-06,
+      "loss": 1.5826,
       "step": 915
     },
     {
       "epoch": 0.40779076238174733,
-      "grad_norm": 2.8730368614196777,
-      "learning_rate": 3.970631432305694e-06,
-      "loss": 1.867,
+      "grad_norm": 2.9087412357330322,
+      "learning_rate": 3.883457821114811e-06,
+      "loss": 1.8822,
       "step": 916
     },
     {
       "epoch": 0.4082359488035615,
-      "grad_norm": 4.520809173583984,
-      "learning_rate": 3.883457821114811e-06,
-      "loss": 2.3426,
+      "grad_norm": 4.532018661499023,
+      "learning_rate": 3.797232841391407e-06,
+      "loss": 2.3841,
       "step": 917
     },
     {
       "epoch": 0.40868113522537564,
-      "grad_norm": 2.5547308921813965,
-      "learning_rate": 3.797232841391407e-06,
-      "loss": 1.6473,
+      "grad_norm": 2.428201913833618,
+      "learning_rate": 3.711957344141237e-06,
+      "loss": 1.6608,
       "step": 918
     },
     {
       "epoch": 0.40912632164718976,
-      "grad_norm": 2.6396074295043945,
-      "learning_rate": 3.711957344141237e-06,
-      "loss": 1.5152,
+      "grad_norm": 2.5778157711029053,
+      "learning_rate": 3.627632170999029e-06,
+      "loss": 1.5169,
       "step": 919
     },
     {
       "epoch": 0.4095715080690039,
-      "grad_norm": 2.5423967838287354,
-      "learning_rate": 3.627632170999029e-06,
-      "loss": 1.8634,
+      "grad_norm": 3.019365072250366,
+      "learning_rate": 3.5442581542201923e-06,
+      "loss": 1.9217,
       "step": 920
     },
     {
       "epoch": 0.410016694490818,
-      "grad_norm": 2.359440565109253,
-      "learning_rate": 3.5442581542201923e-06,
-      "loss": 1.6601,
+      "grad_norm": 2.3321633338928223,
+      "learning_rate": 3.461836116672612e-06,
+      "loss": 1.6613,
       "step": 921
     },
     {
       "epoch": 0.41046188091263214,
-      "grad_norm": 3.3366010189056396,
-      "learning_rate": 3.461836116672612e-06,
-      "loss": 2.3835,
+      "grad_norm": 3.6057512760162354,
+      "learning_rate": 3.380366871828522e-06,
+      "loss": 2.411,
       "step": 922
     },
     {
       "epoch": 0.4109070673344463,
-      "grad_norm": 1.9757188558578491,
-      "learning_rate": 3.380366871828522e-06,
-      "loss": 1.3504,
+      "grad_norm": 2.040179967880249,
+      "learning_rate": 3.2998512237565005e-06,
+      "loss": 1.3784,
       "step": 923
     },
     {
       "epoch": 0.41135225375626044,
-      "grad_norm": 1.7242034673690796,
-      "learning_rate": 3.2998512237565005e-06,
-      "loss": 0.9853,
+      "grad_norm": 1.748306155204773,
+      "learning_rate": 3.2202899671134546e-06,
+      "loss": 1.025,
       "step": 924
     },
     {
       "epoch": 0.41179744017807457,
-      "grad_norm": 4.0235161781311035,
-      "learning_rate": 3.2202899671134546e-06,
-      "loss": 2.5185,
+      "grad_norm": 4.816791534423828,
+      "learning_rate": 3.1416838871368924e-06,
+      "loss": 2.4774,
       "step": 925
     },
     {
       "epoch": 0.4122426265998887,
-      "grad_norm": 3.6245474815368652,
-      "learning_rate": 3.1416838871368924e-06,
-      "loss": 2.2566,
+      "grad_norm": 3.0103750228881836,
+      "learning_rate": 3.064033759637064e-06,
+      "loss": 2.1963,
       "step": 926
     },
     {
       "epoch": 0.4126878130217028,
-      "grad_norm": 2.94282865524292,
-      "learning_rate": 3.064033759637064e-06,
-      "loss": 1.7241,
+      "grad_norm": 2.694369316101074,
+      "learning_rate": 2.9873403509894203e-06,
+      "loss": 1.7994,
       "step": 927
     },
     {
       "epoch": 0.41313299944351695,
-      "grad_norm": 1.7278319597244263,
-      "learning_rate": 2.9873403509894203e-06,
-      "loss": 1.6847,
+      "grad_norm": 1.5631622076034546,
+      "learning_rate": 2.9116044181269007e-06,
+      "loss": 1.6822,
       "step": 928
     },
     {
       "epoch": 0.4135781858653311,
-      "grad_norm": 2.720810651779175,
-      "learning_rate": 2.9116044181269007e-06,
-      "loss": 1.7949,
+      "grad_norm": 2.746835947036743,
+      "learning_rate": 2.836826708532603e-06,
+      "loss": 1.8388,
       "step": 929
     },
     {
       "epoch": 0.41402337228714525,
-      "grad_norm": 2.830357789993286,
-      "learning_rate": 2.836826708532603e-06,
-      "loss": 1.8299,
+      "grad_norm": 2.7418153285980225,
+      "learning_rate": 2.7630079602323442e-06,
+      "loss": 1.8356,
       "step": 930
     },
     {
       "epoch": 0.4144685587089594,
-      "grad_norm": 2.6716318130493164,
-      "learning_rate": 2.7630079602323442e-06,
-      "loss": 1.4774,
+      "grad_norm": 2.965754270553589,
+      "learning_rate": 2.690148901787337e-06,
+      "loss": 1.475,
       "step": 931
     },
     {
       "epoch": 0.4149137451307735,
-      "grad_norm": 2.116966724395752,
-      "learning_rate": 2.690148901787337e-06,
-      "loss": 0.5776,
+      "grad_norm": 1.6221050024032593,
+      "learning_rate": 2.618250252287113e-06,
+      "loss": 0.5562,
       "step": 932
     },
     {
       "epoch": 0.41535893155258763,
-      "grad_norm": 1.9328233003616333,
-      "learning_rate": 2.618250252287113e-06,
-      "loss": 1.4808,
+      "grad_norm": 1.9528764486312866,
+      "learning_rate": 2.5473127213422763e-06,
+      "loss": 1.4934,
       "step": 933
     },
     {
       "epoch": 0.41580411797440175,
-      "grad_norm": 2.859921932220459,
-      "learning_rate": 2.5473127213422763e-06,
-      "loss": 2.1348,
+      "grad_norm": 2.7346739768981934,
+      "learning_rate": 2.4773370090776626e-06,
+      "loss": 2.1929,
       "step": 934
     },
     {
       "epoch": 0.41624930439621594,
-      "grad_norm": 2.039973020553589,
-      "learning_rate": 2.4773370090776626e-06,
-      "loss": 1.8692,
+      "grad_norm": 2.0011751651763916,
+      "learning_rate": 2.4083238061252567e-06,
+      "loss": 1.8707,
       "step": 935
     },
     {
       "epoch": 0.41669449081803006,
-      "grad_norm": 3.830960273742676,
-      "learning_rate": 2.4083238061252567e-06,
-      "loss": 1.9331,
+      "grad_norm": 3.6254782676696777,
+      "learning_rate": 2.3402737936175425e-06,
+      "loss": 1.9401,
       "step": 936
     },
     {
       "epoch": 0.4171396772398442,
-      "grad_norm": 2.213256597518921,
-      "learning_rate": 2.3402737936175425e-06,
-      "loss": 1.8484,
+      "grad_norm": 2.0803048610687256,
+      "learning_rate": 2.273187643180652e-06,
+      "loss": 1.837,
       "step": 937
     },
     {
       "epoch": 0.4175848636616583,
-      "grad_norm": 2.631376266479492,
-      "learning_rate": 2.273187643180652e-06,
-      "loss": 2.4657,
+      "grad_norm": 2.6008853912353516,
+      "learning_rate": 2.2070660169278166e-06,
+      "loss": 2.4823,
       "step": 938
     },
     {
       "epoch": 0.41803005008347244,
-      "grad_norm": 2.208151340484619,
-      "learning_rate": 2.2070660169278166e-06,
-      "loss": 1.9784,
+      "grad_norm": 2.338705539703369,
+      "learning_rate": 2.141909567452793e-06,
+      "loss": 2.008,
       "step": 939
     },
     {
       "epoch": 0.41847523650528656,
-      "grad_norm": 2.4797284603118896,
-      "learning_rate": 2.141909567452793e-06,
-      "loss": 2.1487,
+      "grad_norm": 2.024644613265991,
+      "learning_rate": 2.0777189378234143e-06,
+      "loss": 2.1419,
       "step": 940
     },
     {
       "epoch": 0.41892042292710074,
-      "grad_norm": 2.340677261352539,
-      "learning_rate": 2.0777189378234143e-06,
-      "loss": 1.695,
+      "grad_norm": 2.3318827152252197,
+      "learning_rate": 2.014494761575314e-06,
+      "loss": 1.6451,
       "step": 941
     },
     {
       "epoch": 0.41936560934891487,
-      "grad_norm": 2.2142953872680664,
-      "learning_rate": 2.014494761575314e-06,
-      "loss": 1.9991,
+      "grad_norm": 2.61426043510437,
+      "learning_rate": 1.9522376627055583e-06,
+      "loss": 2.037,
       "step": 942
     },
     {
       "epoch": 0.419810795770729,
-      "grad_norm": 2.083629846572876,
-      "learning_rate": 1.9522376627055583e-06,
-      "loss": 1.3691,
+      "grad_norm": 2.2980270385742188,
+      "learning_rate": 1.8909482556666024e-06,
+      "loss": 1.4046,
       "step": 943
     },
     {
       "epoch": 0.4202559821925431,
-      "grad_norm": 2.2179574966430664,
-      "learning_rate": 1.8909482556666024e-06,
-      "loss": 1.5776,
+      "grad_norm": 2.2436652183532715,
+      "learning_rate": 1.8306271453601199e-06,
+      "loss": 1.5081,
       "step": 944
     },
     {
       "epoch": 0.42070116861435725,
-      "grad_norm": 2.716609239578247,
-      "learning_rate": 1.8306271453601199e-06,
-      "loss": 1.6952,
+      "grad_norm": 2.862178325653076,
+      "learning_rate": 1.771274927131139e-06,
+      "loss": 1.6153,
       "step": 945
     },
     {
       "epoch": 0.42114635503617137,
-      "grad_norm": 3.5457558631896973,
-      "learning_rate": 1.771274927131139e-06,
-      "loss": 2.2282,
+      "grad_norm": 3.7336113452911377,
+      "learning_rate": 1.712892186762083e-06,
+      "loss": 2.3053,
       "step": 946
     },
     {
       "epoch": 0.42159154145798555,
-      "grad_norm": 2.859788656234741,
-      "learning_rate": 1.712892186762083e-06,
-      "loss": 2.4899,
+      "grad_norm": 2.882131814956665,
+      "learning_rate": 1.6554795004670388e-06,
+      "loss": 2.5382,
       "step": 947
     },
     {
       "epoch": 0.4220367278797997,
-      "grad_norm": 2.126269817352295,
-      "learning_rate": 1.6554795004670388e-06,
-      "loss": 1.7597,
+      "grad_norm": 2.3759119510650635,
+      "learning_rate": 1.5990374348860305e-06,
+      "loss": 1.7753,
       "step": 948
     },
     {
       "epoch": 0.4224819143016138,
-      "grad_norm": 11.28099250793457,
-      "learning_rate": 1.5990374348860305e-06,
-      "loss": 2.0871,
+      "grad_norm": 2.8023674488067627,
+      "learning_rate": 1.543566547079467e-06,
+      "loss": 2.0204,
       "step": 949
     },
     {
       "epoch": 0.42292710072342793,
-      "grad_norm": 2.5182433128356934,
-      "learning_rate": 1.543566547079467e-06,
-      "loss": 1.6924,
+      "grad_norm": 3.0776796340942383,
+      "learning_rate": 1.4890673845226133e-06,
+      "loss": 1.7093,
       "step": 950
     },
     {
       "epoch": 0.42337228714524205,
-      "grad_norm": 2.587977886199951,
-      "learning_rate": 1.4890673845226133e-06,
-      "loss": 2.0768,
+      "grad_norm": 2.533381938934326,
+      "learning_rate": 1.4355404851001952e-06,
+      "loss": 2.0766,
       "step": 951
     },
     {
       "epoch": 0.4238174735670562,
-      "grad_norm": 2.2390692234039307,
-      "learning_rate": 1.4355404851001952e-06,
-      "loss": 1.8168,
+      "grad_norm": 2.3900294303894043,
+      "learning_rate": 1.3829863771011253e-06,
+      "loss": 1.8321,
       "step": 952
     },
     {
       "epoch": 0.42426265998887036,
-      "grad_norm": 2.229036331176758,
-      "learning_rate": 1.3829863771011253e-06,
-      "loss": 1.4062,
+      "grad_norm": 2.0056207180023193,
+      "learning_rate": 1.3314055792131964e-06,
+      "loss": 1.3609,
       "step": 953
     },
     {
       "epoch": 0.4247078464106845,
-      "grad_norm": 2.124692440032959,
-      "learning_rate": 1.3314055792131964e-06,
-      "loss": 1.9792,
+      "grad_norm": 2.4295244216918945,
+      "learning_rate": 1.280798600518085e-06,
+      "loss": 1.9711,
       "step": 954
     },
     {
       "epoch": 0.4251530328324986,
-      "grad_norm": 9.60531997680664,
-      "learning_rate": 1.280798600518085e-06,
-      "loss": 2.1784,
+      "grad_norm": 3.231149911880493,
+      "learning_rate": 1.231165940486234e-06,
+      "loss": 2.1324,
       "step": 955
     },
     {
       "epoch": 0.42559821925431274,
-      "grad_norm": 3.135800838470459,
-      "learning_rate": 1.231165940486234e-06,
-      "loss": 2.4013,
+      "grad_norm": 3.490255355834961,
+      "learning_rate": 1.1825080889719563e-06,
+      "loss": 2.4046,
       "step": 956
     },
     {
       "epoch": 0.42604340567612686,
-      "grad_norm": 1.8015056848526,
-      "learning_rate": 1.1825080889719563e-06,
-      "loss": 2.0513,
+      "grad_norm": 1.7336961030960083,
+      "learning_rate": 1.134825526208605e-06,
+      "loss": 1.9998,
       "step": 957
     },
     {
       "epoch": 0.426488592097941,
-      "grad_norm": 2.0358049869537354,
-      "learning_rate": 1.134825526208605e-06,
-      "loss": 1.6947,
+      "grad_norm": 1.9206734895706177,
+      "learning_rate": 1.0881187228038215e-06,
+      "loss": 1.686,
       "step": 958
     },
     {
       "epoch": 0.42693377851975517,
-      "grad_norm": 2.1654505729675293,
-      "learning_rate": 1.0881187228038215e-06,
-      "loss": 1.5756,
+      "grad_norm": 2.213815450668335,
+      "learning_rate": 1.0423881397349068e-06,
+      "loss": 1.5822,
       "step": 959
     },
     {
       "epoch": 0.4273789649415693,
-      "grad_norm": 1.4749082326889038,
-      "learning_rate": 1.0423881397349068e-06,
-      "loss": 1.0004,
+      "grad_norm": 1.5032622814178467,
+      "learning_rate": 9.976342283442463e-07,
+      "loss": 0.9812,
       "step": 960
     },
     {
       "epoch": 0.4278241513633834,
-      "grad_norm": 2.168402671813965,
-      "learning_rate": 9.976342283442463e-07,
-      "loss": 1.8201,
+      "grad_norm": 2.7695553302764893,
+      "learning_rate": 9.538574303348813e-07,
+      "loss": 1.8789,
       "step": 961
     },
     {
       "epoch": 0.42826933778519755,
-      "grad_norm": 2.350822925567627,
-      "learning_rate": 9.538574303348813e-07,
-      "loss": 1.9771,
+      "grad_norm": 2.4367427825927734,
+      "learning_rate": 9.110581777661331e-07,
+      "loss": 2.0159,
       "step": 962
     },
     {
       "epoch": 0.42871452420701167,
-      "grad_norm": 3.4289748668670654,
-      "learning_rate": 9.110581777661331e-07,
-      "loss": 2.1341,
+      "grad_norm": 3.4534249305725098,
+      "learning_rate": 8.692368930493521e-07,
+      "loss": 2.1388,
       "step": 963
     },
     {
       "epoch": 0.4291597106288258,
-      "grad_norm": 2.5226845741271973,
-      "learning_rate": 8.692368930493521e-07,
-      "loss": 1.4104,
+      "grad_norm": 2.8300178050994873,
+      "learning_rate": 8.283939889437209e-07,
+      "loss": 1.3831,
       "step": 964
     },
     {
       "epoch": 0.42960489705064,
-      "grad_norm": 3.7586185932159424,
-      "learning_rate": 8.283939889437209e-07,
-      "loss": 2.83,
+      "grad_norm": 3.648935317993164,
+      "learning_rate": 7.885298685522235e-07,
+      "loss": 2.8245,
       "step": 965
     },
     {
       "epoch": 0.4300500834724541,
-      "grad_norm": 2.972186803817749,
-      "learning_rate": 7.885298685522235e-07,
-      "loss": 2.0237,
+      "grad_norm": 3.6127102375030518,
+      "learning_rate": 7.496449253176274e-07,
+      "loss": 2.0929,
       "step": 966
     },
     {
       "epoch": 0.43049526989426823,
-      "grad_norm": 2.904418468475342,
-      "learning_rate": 7.496449253176274e-07,
-      "loss": 1.6953,
+      "grad_norm": 3.364607810974121,
+      "learning_rate": 7.117395430186414e-07,
+      "loss": 1.702,
       "step": 967
     },
     {
       "epoch": 0.43094045631608235,
-      "grad_norm": 2.9558122158050537,
-      "learning_rate": 7.117395430186414e-07,
-      "loss": 3.0708,
+      "grad_norm": 2.9458580017089844,
+      "learning_rate": 6.748140957660631e-07,
+      "loss": 3.1025,
       "step": 968
     },
     {
       "epoch": 0.4313856427378965,
-      "grad_norm": 2.732140064239502,
-      "learning_rate": 6.748140957660631e-07,
-      "loss": 1.7392,
+      "grad_norm": 2.9718213081359863,
+      "learning_rate": 6.388689479991605e-07,
+      "loss": 1.7826,
       "step": 969
     },
     {
       "epoch": 0.4318308291597106,
-      "grad_norm": 2.7709476947784424,
-      "learning_rate": 6.388689479991605e-07,
-      "loss": 1.7855,
+      "grad_norm": 3.344043016433716,
+      "learning_rate": 6.039044544820404e-07,
+      "loss": 1.7272,
       "step": 970
     },
     {
       "epoch": 0.4322760155815248,
-      "grad_norm": 1.9800370931625366,
-      "learning_rate": 6.039044544820404e-07,
-      "loss": 0.8279,
+      "grad_norm": 2.1761598587036133,
+      "learning_rate": 5.699209603001076e-07,
+      "loss": 0.8196,
       "step": 971
     },
     {
       "epoch": 0.4327212020033389,
-      "grad_norm": 3.503222703933716,
-      "learning_rate": 5.699209603001076e-07,
-      "loss": 1.2507,
+      "grad_norm": 3.813335657119751,
+      "learning_rate": 5.369188008567672e-07,
+      "loss": 1.2445,
       "step": 972
     },
     {
       "epoch": 0.43316638842515304,
-      "grad_norm": 3.6716744899749756,
-      "learning_rate": 5.369188008567672e-07,
-      "loss": 1.5849,
+      "grad_norm": 3.2910823822021484,
+      "learning_rate": 5.048983018699827e-07,
+      "loss": 1.5288,
       "step": 973
     },
     {
       "epoch": 0.43361157484696716,
-      "grad_norm": 3.8361189365386963,
-      "learning_rate": 5.048983018699827e-07,
-      "loss": 2.7786,
+      "grad_norm": 3.7944729328155518,
+      "learning_rate": 4.738597793691679e-07,
+      "loss": 2.7908,
       "step": 974
     },
     {
       "epoch": 0.4340567612687813,
-      "grad_norm": 2.1607441902160645,
-      "learning_rate": 4.738597793691679e-07,
-      "loss": 1.6388,
+      "grad_norm": 2.3977789878845215,
+      "learning_rate": 4.438035396920004e-07,
+      "loss": 1.6361,
       "step": 975
     },
     {
       "epoch": 0.4345019476905954,
-      "grad_norm": 2.7278549671173096,
-      "learning_rate": 4.438035396920004e-07,
-      "loss": 2.5643,
+      "grad_norm": 2.7309062480926514,
+      "learning_rate": 4.1472987948143473e-07,
+      "loss": 2.5606,
       "step": 976
     },
     {
       "epoch": 0.4349471341124096,
-      "grad_norm": 2.0202202796936035,
-      "learning_rate": 4.1472987948143473e-07,
-      "loss": 1.6711,
+      "grad_norm": 2.0864222049713135,
+      "learning_rate": 3.866390856827495e-07,
+      "loss": 1.65,
       "step": 977
     },
     {
       "epoch": 0.4353923205342237,
-      "grad_norm": 2.3101208209991455,
-      "learning_rate": 3.866390856827495e-07,
-      "loss": 1.8005,
+      "grad_norm": 2.458543300628662,
+      "learning_rate": 3.595314355407609e-07,
+      "loss": 1.8571,
       "step": 978
     },
     {
       "epoch": 0.43583750695603785,
-      "grad_norm": 1.7602624893188477,
-      "learning_rate": 3.595314355407609e-07,
-      "loss": 1.4123,
+      "grad_norm": 1.9558011293411255,
+      "learning_rate": 3.3340719659701313e-07,
+      "loss": 1.386,
       "step": 979
     },
     {
       "epoch": 0.43628269337785197,
-      "grad_norm": 2.8393287658691406,
-      "learning_rate": 3.3340719659701313e-07,
-      "loss": 1.9375,
+      "grad_norm": 1.8672804832458496,
+      "learning_rate": 3.0826662668720364e-07,
+      "loss": 1.9483,
       "step": 980
     },
     {
       "epoch": 0.4367278797996661,
-      "grad_norm": 2.671792507171631,
-      "learning_rate": 3.0826662668720364e-07,
-      "loss": 2.1422,
+      "grad_norm": 2.7956416606903076,
+      "learning_rate": 2.841099739386066e-07,
+      "loss": 2.1084,
       "step": 981
     },
     {
       "epoch": 0.4371730662214802,
-      "grad_norm": 2.798468828201294,
-      "learning_rate": 2.841099739386066e-07,
-      "loss": 1.1786,
+      "grad_norm": 2.68589448928833,
+      "learning_rate": 2.609374767676309e-07,
+      "loss": 1.1878,
       "step": 982
     },
     {
       "epoch": 0.4376182526432944,
-      "grad_norm": 2.4670653343200684,
-      "learning_rate": 2.609374767676309e-07,
-      "loss": 2.0231,
+      "grad_norm": 2.501877546310425,
+      "learning_rate": 2.387493638774774e-07,
+      "loss": 2.0479,
       "step": 983
     },
     {
       "epoch": 0.43806343906510853,
-      "grad_norm": 2.650512218475342,
-      "learning_rate": 2.387493638774774e-07,
-      "loss": 2.1189,
+      "grad_norm": 2.6020426750183105,
+      "learning_rate": 2.175458542558517e-07,
+      "loss": 2.17,
       "step": 984
     },
     {
       "epoch": 0.43850862548692265,
-      "grad_norm": 2.984624147415161,
-      "learning_rate": 2.175458542558517e-07,
-      "loss": 2.3214,
+      "grad_norm": 2.8752057552337646,
+      "learning_rate": 1.973271571728441e-07,
+      "loss": 2.3384,
       "step": 985
     },
     {
       "epoch": 0.4389538119087368,
-      "grad_norm": 2.806220769882202,
-      "learning_rate": 1.973271571728441e-07,
-      "loss": 2.8719,
+      "grad_norm": 2.631931781768799,
+      "learning_rate": 1.7809347217881966e-07,
+      "loss": 2.8829,
       "step": 986
     },
     {
       "epoch": 0.4393989983305509,
-      "grad_norm": 2.488363742828369,
-      "learning_rate": 1.7809347217881966e-07,
-      "loss": 2.2061,
+      "grad_norm": 2.9459500312805176,
+      "learning_rate": 1.598449891024978e-07,
+      "loss": 2.1954,
       "step": 987
     },
     {
       "epoch": 0.43984418475236503,
-      "grad_norm": 4.78094482421875,
-      "learning_rate": 1.598449891024978e-07,
-      "loss": 2.4017,
+      "grad_norm": 4.918516159057617,
+      "learning_rate": 1.425818880490315e-07,
+      "loss": 2.4102,
       "step": 988
     },
     {
       "epoch": 0.4402893711741792,
-      "grad_norm": 1.6316055059432983,
-      "learning_rate": 1.425818880490315e-07,
-      "loss": 1.8267,
+      "grad_norm": 1.585867166519165,
+      "learning_rate": 1.2630433939825327e-07,
+      "loss": 1.8347,
       "step": 989
     },
     {
       "epoch": 0.44073455759599334,
-      "grad_norm": 2.0801150798797607,
-      "learning_rate": 1.2630433939825327e-07,
-      "loss": 1.6163,
+      "grad_norm": 2.0162644386291504,
+      "learning_rate": 1.1101250380300965e-07,
+      "loss": 1.5607,
       "step": 990
     },
     {
       "epoch": 0.44117974401780746,
-      "grad_norm": 1.5451546907424927,
-      "learning_rate": 1.1101250380300965e-07,
-      "loss": 1.3939,
+      "grad_norm": 1.688201904296875,
+      "learning_rate": 9.670653218752934e-08,
+      "loss": 1.392,
       "step": 991
     },
     {
       "epoch": 0.4416249304396216,
-      "grad_norm": 1.971220850944519,
-      "learning_rate": 9.670653218752934e-08,
-      "loss": 1.7467,
+      "grad_norm": 2.0100865364074707,
+      "learning_rate": 8.33865657459909e-08,
+      "loss": 1.7306,
       "step": 992
     },
     {
       "epoch": 0.4420701168614357,
-      "grad_norm": 3.593325138092041,
-      "learning_rate": 8.33865657459909e-08,
-      "loss": 2.2187,
+      "grad_norm": 3.2852158546447754,
+      "learning_rate": 7.105273594107953e-08,
+      "loss": 2.2937,
       "step": 993
     },
     {
       "epoch": 0.44251530328324984,
-      "grad_norm": 1.9523143768310547,
-      "learning_rate": 7.105273594107953e-08,
-      "loss": 1.2635,
+      "grad_norm": 2.951110601425171,
+      "learning_rate": 5.970516450271025e-08,
+      "loss": 1.2699,
       "step": 994
     },
     {
       "epoch": 0.442960489705064,
-      "grad_norm": 2.172438144683838,
-      "learning_rate": 5.970516450271025e-08,
-      "loss": 1.3526,
+      "grad_norm": 2.1301627159118652,
+      "learning_rate": 4.934396342684e-08,
+      "loss": 1.3962,
       "step": 995
     },
     {
       "epoch": 0.44340567612687815,
-      "grad_norm": 1.9198912382125854,
-      "learning_rate": 4.934396342684e-08,
-      "loss": 1.1006,
+      "grad_norm": 1.6765490770339966,
+      "learning_rate": 3.996923497434635e-08,
+      "loss": 1.046,
       "step": 996
     },
     {
       "epoch": 0.44385086254869227,
-      "grad_norm": 2.753998041152954,
-      "learning_rate": 3.996923497434635e-08,
-      "loss": 2.1173,
+      "grad_norm": 2.9391028881073,
+      "learning_rate": 3.1581071670006015e-08,
+      "loss": 2.1163,
       "step": 997
     },
     {
       "epoch": 0.4442960489705064,
-      "grad_norm": 2.4736664295196533,
-      "learning_rate": 3.1581071670006015e-08,
-      "loss": 2.3068,
+      "grad_norm": 2.6507630348205566,
+      "learning_rate": 2.417955630159563e-08,
+      "loss": 2.3596,
       "step": 998
     },
     {
       "epoch": 0.4447412353923205,
-      "grad_norm": 2.1527669429779053,
-      "learning_rate": 2.417955630159563e-08,
-      "loss": 1.4739,
+      "grad_norm": 2.1312270164489746,
+      "learning_rate": 1.7764761919103477e-08,
+      "loss": 1.4352,
       "step": 999
     },
     {
       "epoch": 0.44518642181413465,
-      "grad_norm": 2.12872052192688,
-      "learning_rate": 1.7764761919103477e-08,
-      "loss": 2.0056,
+      "grad_norm": 2.425968647003174,
+      "learning_rate": 1.2336751833941229e-08,
+      "loss": 1.9599,
       "step": 1000
     },
     {
       "epoch": 0.44518642181413465,
-      "eval_loss": 1.9295015335083008,
-      "eval_runtime": 133.475,
-      "eval_samples_per_second": 3.544,
-      "eval_steps_per_second": 3.544,
+      "eval_loss": 1.927238941192627,
+      "eval_runtime": 133.1586,
+      "eval_samples_per_second": 3.552,
+      "eval_steps_per_second": 3.552,
       "step": 1000
     }
   ],