diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -10,8377 +10,8377 @@
   "log_history": [
     {
       "epoch": 0.004173187271778821,
-      "grad_norm": 6.445867566745911,
+      "grad_norm": 6.4459224518547975,
       "learning_rate": 3.3333333333333335e-07,
       "loss": 1.0378,
       "step": 1
     },
     {
       "epoch": 0.008346374543557642,
-      "grad_norm": 6.349220661987718,
+      "grad_norm": 6.349157267563719,
       "learning_rate": 6.666666666666667e-07,
       "loss": 1.0303,
       "step": 2
     },
     {
       "epoch": 0.012519561815336464,
-      "grad_norm": 6.507565666699471,
+      "grad_norm": 6.514026566608604,
       "learning_rate": 1.0000000000000002e-06,
-      "loss": 1.0577,
+      "loss": 1.0579,
       "step": 3
     },
     {
       "epoch": 0.016692749087115284,
-      "grad_norm": 6.4116195705769305,
+      "grad_norm": 6.409347973181753,
       "learning_rate": 1.3333333333333334e-06,
-      "loss": 1.0509,
+      "loss": 1.0508,
       "step": 4
     },
     {
       "epoch": 0.020865936358894107,
-      "grad_norm": 6.142816048161353,
+      "grad_norm": 6.13903181752867,
       "learning_rate": 1.6666666666666667e-06,
-      "loss": 1.065,
+      "loss": 1.0649,
       "step": 5
     },
     {
       "epoch": 0.025039123630672927,
-      "grad_norm": 5.8495784314031525,
+      "grad_norm": 5.855586047883461,
       "learning_rate": 2.0000000000000003e-06,
-      "loss": 1.0591,
+      "loss": 1.0592,
       "step": 6
     },
     {
       "epoch": 0.029212310902451747,
-      "grad_norm": 4.644499431757354,
+      "grad_norm": 4.650159189654289,
       "learning_rate": 2.3333333333333336e-06,
-      "loss": 0.9879,
+      "loss": 0.9882,
       "step": 7
     },
     {
       "epoch": 0.03338549817423057,
-      "grad_norm": 4.3858829293213475,
+      "grad_norm": 4.383666347704108,
       "learning_rate": 2.666666666666667e-06,
       "loss": 0.9762,
       "step": 8
     },
     {
       "epoch": 0.03755868544600939,
-      "grad_norm": 2.345605300545527,
+      "grad_norm": 2.342245526100041,
       "learning_rate": 3e-06,
-      "loss": 0.9808,
+      "loss": 0.9807,
       "step": 9
     },
     {
       "epoch": 0.041731872717788214,
-      "grad_norm": 2.0680989366845397,
+      "grad_norm": 2.0690213571504197,
       "learning_rate": 3.3333333333333333e-06,
-      "loss": 0.943,
+      "loss": 0.9431,
       "step": 10
     },
     {
       "epoch": 0.04590505998956703,
-      "grad_norm": 2.244866250252733,
+      "grad_norm": 2.2442169305489816,
       "learning_rate": 3.6666666666666666e-06,
       "loss": 0.9603,
       "step": 11
     },
     {
       "epoch": 0.050078247261345854,
-      "grad_norm": 4.419657414203797,
+      "grad_norm": 4.4166094312848205,
       "learning_rate": 4.000000000000001e-06,
-      "loss": 0.9285,
+      "loss": 0.9284,
       "step": 12
     },
     {
       "epoch": 0.05425143453312467,
-      "grad_norm": 4.834711151064183,
+      "grad_norm": 4.832677621974598,
       "learning_rate": 4.333333333333334e-06,
-      "loss": 0.9622,
+      "loss": 0.9621,
       "step": 13
     },
     {
       "epoch": 0.058424621804903494,
-      "grad_norm": 4.509834025181203,
+      "grad_norm": 4.504720140666919,
       "learning_rate": 4.666666666666667e-06,
-      "loss": 0.925,
+      "loss": 0.9248,
       "step": 14
     },
     {
       "epoch": 0.06259780907668232,
-      "grad_norm": 4.152747266559372,
+      "grad_norm": 4.152216150844888,
       "learning_rate": 5e-06,
-      "loss": 0.9016,
+      "loss": 0.9017,
       "step": 15
     },
     {
       "epoch": 0.06677099634846113,
-      "grad_norm": 3.487588433075064,
+      "grad_norm": 3.4747746583570045,
       "learning_rate": 5.333333333333334e-06,
-      "loss": 0.902,
+      "loss": 0.9018,
       "step": 16
     },
     {
       "epoch": 0.07094418362023996,
-      "grad_norm": 3.1023171102897287,
+      "grad_norm": 3.0940105143479144,
       "learning_rate": 5.666666666666667e-06,
-      "loss": 0.8756,
+      "loss": 0.8755,
       "step": 17
     },
     {
       "epoch": 0.07511737089201878,
-      "grad_norm": 2.755892828840727,
+      "grad_norm": 2.7510947206466585,
       "learning_rate": 6e-06,
-      "loss": 0.8831,
+      "loss": 0.883,
       "step": 18
     },
     {
       "epoch": 0.0792905581637976,
-      "grad_norm": 2.239006201121862,
+      "grad_norm": 2.244018899861665,
       "learning_rate": 6.333333333333333e-06,
-      "loss": 0.8183,
+      "loss": 0.8184,
       "step": 19
     },
     {
       "epoch": 0.08346374543557643,
-      "grad_norm": 1.9553614471670977,
+      "grad_norm": 1.9574025401035455,
       "learning_rate": 6.666666666666667e-06,
-      "loss": 0.8162,
+      "loss": 0.8161,
       "step": 20
     },
     {
       "epoch": 0.08763693270735524,
-      "grad_norm": 1.9311567126653493,
+      "grad_norm": 1.9294400469418005,
       "learning_rate": 7e-06,
-      "loss": 0.8161,
+      "loss": 0.816,
       "step": 21
     },
     {
       "epoch": 0.09181011997913406,
-      "grad_norm": 2.144895652126078,
+      "grad_norm": 2.1475754313582236,
       "learning_rate": 7.333333333333333e-06,
-      "loss": 0.8312,
+      "loss": 0.8313,
       "step": 22
     },
     {
       "epoch": 0.09598330725091289,
-      "grad_norm": 1.9930976950107564,
+      "grad_norm": 1.992796543508962,
       "learning_rate": 7.666666666666667e-06,
-      "loss": 0.8089,
+      "loss": 0.8088,
       "step": 23
     },
     {
       "epoch": 0.10015649452269171,
-      "grad_norm": 2.011142215810346,
+      "grad_norm": 2.0070719172099505,
       "learning_rate": 8.000000000000001e-06,
-      "loss": 0.8161,
+      "loss": 0.8162,
       "step": 24
     },
     {
       "epoch": 0.10432968179447052,
-      "grad_norm": 1.575797867974249,
+      "grad_norm": 1.582450916763155,
       "learning_rate": 8.333333333333334e-06,
       "loss": 0.7889,
       "step": 25
     },
     {
       "epoch": 0.10850286906624934,
-      "grad_norm": 1.2921982968052508,
+      "grad_norm": 1.2953453523212095,
       "learning_rate": 8.666666666666668e-06,
       "loss": 0.7878,
       "step": 26
     },
     {
       "epoch": 0.11267605633802817,
-      "grad_norm": 1.5063573771692975,
+      "grad_norm": 1.5136548638414973,
       "learning_rate": 9e-06,
-      "loss": 0.7592,
+      "loss": 0.7593,
       "step": 27
     },
     {
       "epoch": 0.11684924360980699,
-      "grad_norm": 1.4770317161349518,
+      "grad_norm": 1.4669353484165313,
       "learning_rate": 9.333333333333334e-06,
-      "loss": 0.7758,
+      "loss": 0.7757,
       "step": 28
     },
     {
       "epoch": 0.1210224308815858,
-      "grad_norm": 1.0771421090710933,
+      "grad_norm": 1.0774620277082754,
       "learning_rate": 9.666666666666667e-06,
-      "loss": 0.8058,
+      "loss": 0.8059,
       "step": 29
     },
     {
       "epoch": 0.12519561815336464,
-      "grad_norm": 0.9045152255888053,
+      "grad_norm": 0.900863035403963,
       "learning_rate": 1e-05,
       "loss": 0.7606,
       "step": 30
     },
     {
       "epoch": 0.12936880542514345,
-      "grad_norm": 0.8696022006378192,
+      "grad_norm": 0.8813081955067554,
       "learning_rate": 1.0333333333333335e-05,
-      "loss": 0.7644,
+      "loss": 0.7645,
       "step": 31
     },
     {
       "epoch": 0.13354199269692227,
-      "grad_norm": 1.0105373105407278,
+      "grad_norm": 1.0096791709393576,
       "learning_rate": 1.0666666666666667e-05,
       "loss": 0.7453,
       "step": 32
     },
     {
       "epoch": 0.13771517996870108,
-      "grad_norm": 0.7847445077431984,
+      "grad_norm": 0.7780100064984126,
       "learning_rate": 1.1000000000000001e-05,
       "loss": 0.7601,
       "step": 33
     },
     {
       "epoch": 0.14188836724047993,
-      "grad_norm": 0.7250964785566338,
+      "grad_norm": 0.7254449537826935,
       "learning_rate": 1.1333333333333334e-05,
       "loss": 0.7406,
       "step": 34
     },
     {
       "epoch": 0.14606155451225875,
-      "grad_norm": 0.8450112690556413,
+      "grad_norm": 0.8447289946388474,
       "learning_rate": 1.1666666666666668e-05,
       "loss": 0.7528,
       "step": 35
     },
     {
       "epoch": 0.15023474178403756,
-      "grad_norm": 0.7662211896448476,
+      "grad_norm": 0.7582779882507266,
       "learning_rate": 1.2e-05,
       "loss": 0.753,
       "step": 36
     },
     {
       "epoch": 0.15440792905581638,
-      "grad_norm": 0.6279761643308388,
+      "grad_norm": 0.631137954003926,
       "learning_rate": 1.2333333333333334e-05,
       "loss": 0.7213,
       "step": 37
     },
     {
       "epoch": 0.1585811163275952,
-      "grad_norm": 0.7231288245723091,
+      "grad_norm": 0.7198079922194827,
       "learning_rate": 1.2666666666666667e-05,
-      "loss": 0.7404,
+      "loss": 0.7403,
       "step": 38
     },
     {
       "epoch": 0.162754303599374,
-      "grad_norm": 0.8812780791021576,
+      "grad_norm": 0.8643041541122825,
       "learning_rate": 1.3000000000000001e-05,
-      "loss": 0.7337,
+      "loss": 0.7336,
       "step": 39
     },
     {
       "epoch": 0.16692749087115286,
-      "grad_norm": 0.6552584176743845,
+      "grad_norm": 0.655002051615442,
       "learning_rate": 1.3333333333333333e-05,
-      "loss": 0.7276,
+      "loss": 0.7277,
       "step": 40
     },
     {
       "epoch": 0.17110067814293167,
-      "grad_norm": 0.7861218460721253,
+      "grad_norm": 0.775350626288524,
       "learning_rate": 1.3666666666666667e-05,
-      "loss": 0.7155,
+      "loss": 0.7154,
       "step": 41
     },
     {
       "epoch": 0.1752738654147105,
-      "grad_norm": 0.7194701620206855,
+      "grad_norm": 0.7181715577930341,
       "learning_rate": 1.4e-05,
       "loss": 0.7185,
       "step": 42
     },
     {
       "epoch": 0.1794470526864893,
-      "grad_norm": 0.8079569748310298,
+      "grad_norm": 0.8126237022699533,
       "learning_rate": 1.4333333333333334e-05,
       "loss": 0.7478,
       "step": 43
     },
     {
       "epoch": 0.18362023995826812,
-      "grad_norm": 0.8453209725532681,
+      "grad_norm": 0.83270935386456,
       "learning_rate": 1.4666666666666666e-05,
-      "loss": 0.7522,
+      "loss": 0.7521,
       "step": 44
     },
     {
       "epoch": 0.18779342723004694,
-      "grad_norm": 0.6922985891832245,
+      "grad_norm": 0.6957022760313389,
       "learning_rate": 1.5000000000000002e-05,
       "loss": 0.7334,
       "step": 45
     },
     {
       "epoch": 0.19196661450182578,
-      "grad_norm": 0.9963318666144858,
+      "grad_norm": 0.9746323272114665,
       "learning_rate": 1.5333333333333334e-05,
       "loss": 0.7258,
       "step": 46
     },
     {
       "epoch": 0.1961398017736046,
-      "grad_norm": 0.7089311128326758,
+      "grad_norm": 0.7133302431187958,
       "learning_rate": 1.5666666666666667e-05,
-      "loss": 0.7052,
+      "loss": 0.7051,
       "step": 47
     },
     {
       "epoch": 0.20031298904538342,
-      "grad_norm": 0.8894115855082984,
+      "grad_norm": 0.8698537728327713,
       "learning_rate": 1.6000000000000003e-05,
       "loss": 0.7408,
       "step": 48
     },
     {
       "epoch": 0.20448617631716223,
-      "grad_norm": 0.7782762230513075,
+      "grad_norm": 0.7911330820585855,
       "learning_rate": 1.6333333333333335e-05,
-      "loss": 0.7136,
+      "loss": 0.7137,
       "step": 49
     },
     {
       "epoch": 0.20865936358894105,
-      "grad_norm": 0.6338722552043631,
+      "grad_norm": 0.6347169570956186,
       "learning_rate": 1.6666666666666667e-05,
-      "loss": 0.7078,
+      "loss": 0.708,
       "step": 50
     },
     {
       "epoch": 0.21283255086071987,
-      "grad_norm": 0.6901213110575198,
+      "grad_norm": 0.6979437147980486,
       "learning_rate": 1.7e-05,
       "loss": 0.6946,
       "step": 51
     },
     {
       "epoch": 0.21700573813249868,
-      "grad_norm": 0.7535313971779523,
+      "grad_norm": 0.7716621972599521,
       "learning_rate": 1.7333333333333336e-05,
       "loss": 0.7415,
       "step": 52
     },
     {
       "epoch": 0.22117892540427753,
-      "grad_norm": 0.7117159701909843,
+      "grad_norm": 0.7130707895443592,
       "learning_rate": 1.7666666666666668e-05,
-      "loss": 0.6825,
+      "loss": 0.6827,
       "step": 53
     },
     {
       "epoch": 0.22535211267605634,
-      "grad_norm": 0.8952310455217533,
+      "grad_norm": 0.9314265829091966,
       "learning_rate": 1.8e-05,
-      "loss": 0.6814,
+      "loss": 0.6813,
       "step": 54
     },
     {
       "epoch": 0.22952529994783516,
-      "grad_norm": 0.680549710520827,
+      "grad_norm": 0.6754420627007269,
       "learning_rate": 1.8333333333333333e-05,
-      "loss": 0.6901,
+      "loss": 0.69,
       "step": 55
     },
     {
       "epoch": 0.23369848721961398,
-      "grad_norm": 0.8950412490867394,
+      "grad_norm": 0.9294464535802912,
       "learning_rate": 1.866666666666667e-05,
       "loss": 0.7401,
       "step": 56
     },
     {
       "epoch": 0.2378716744913928,
-      "grad_norm": 0.885908554949994,
+      "grad_norm": 0.8859549228340888,
       "learning_rate": 1.9e-05,
-      "loss": 0.6995,
+      "loss": 0.6994,
       "step": 57
     },
     {
       "epoch": 0.2420448617631716,
-      "grad_norm": 0.7754531992206622,
+      "grad_norm": 0.808288886051815,
       "learning_rate": 1.9333333333333333e-05,
-      "loss": 0.7015,
+      "loss": 0.7016,
       "step": 58
     },
     {
       "epoch": 0.24621804903495045,
-      "grad_norm": 0.7616875068091987,
+      "grad_norm": 0.7466889794444995,
       "learning_rate": 1.9666666666666666e-05,
-      "loss": 0.6685,
+      "loss": 0.6684,
       "step": 59
     },
     {
       "epoch": 0.25039123630672927,
-      "grad_norm": 0.6445370890699123,
+      "grad_norm": 0.6500484480047598,
       "learning_rate": 2e-05,
       "loss": 0.7043,
       "step": 60
     },
     {
       "epoch": 0.25456442357850806,
-      "grad_norm": 0.8448332773644573,
+      "grad_norm": 0.8258013251935666,
       "learning_rate": 2.0333333333333334e-05,
-      "loss": 0.7051,
+      "loss": 0.7052,
       "step": 61
     },
     {
       "epoch": 0.2587376108502869,
-      "grad_norm": 0.5928021010018815,
+      "grad_norm": 0.6004577519729952,
       "learning_rate": 2.066666666666667e-05,
-      "loss": 0.7042,
+      "loss": 0.7041,
       "step": 62
     },
     {
       "epoch": 0.26291079812206575,
-      "grad_norm": 0.7811482449744304,
+      "grad_norm": 0.7709590063735229,
       "learning_rate": 2.1000000000000002e-05,
       "loss": 0.6957,
       "step": 63
     },
     {
       "epoch": 0.26708398539384454,
-      "grad_norm": 0.6327717580228115,
+      "grad_norm": 0.6164833037067735,
       "learning_rate": 2.1333333333333335e-05,
-      "loss": 0.6837,
+      "loss": 0.6836,
       "step": 64
     },
     {
       "epoch": 0.2712571726656234,
-      "grad_norm": 0.6210800562446848,
+      "grad_norm": 0.6455502241481836,
       "learning_rate": 2.1666666666666667e-05,
-      "loss": 0.6587,
+      "loss": 0.6584,
       "step": 65
     },
     {
       "epoch": 0.27543035993740217,
-      "grad_norm": 0.7348100514175354,
+      "grad_norm": 0.7164216886846168,
       "learning_rate": 2.2000000000000003e-05,
-      "loss": 0.6874,
+      "loss": 0.6875,
       "step": 66
     },
     {
       "epoch": 0.279603547209181,
-      "grad_norm": 6.974337612398512,
+      "grad_norm": 0.7266966982776554,
       "learning_rate": 2.2333333333333335e-05,
-      "loss": 0.6943,
+      "loss": 0.6921,
       "step": 67
     },
     {
       "epoch": 0.28377673448095986,
-      "grad_norm": 1.1987450740745307,
+      "grad_norm": 0.7732350779799058,
       "learning_rate": 2.2666666666666668e-05,
-      "loss": 0.682,
+      "loss": 0.6778,
       "step": 68
     },
     {
       "epoch": 0.28794992175273865,
-      "grad_norm": 0.9443440319962901,
+      "grad_norm": 0.746508534248161,
       "learning_rate": 2.3e-05,
-      "loss": 0.6925,
+      "loss": 0.6894,
       "step": 69
     },
     {
       "epoch": 0.2921231090245175,
-      "grad_norm": 1.0945467908264084,
+      "grad_norm": 0.7867414646847366,
       "learning_rate": 2.3333333333333336e-05,
-      "loss": 0.6796,
+      "loss": 0.6783,
       "step": 70
     },
     {
       "epoch": 0.2962962962962963,
-      "grad_norm": 0.9746102541615308,
+      "grad_norm": 0.605778137034838,
       "learning_rate": 2.3666666666666668e-05,
-      "loss": 0.7037,
+      "loss": 0.7024,
       "step": 71
     },
     {
       "epoch": 0.3004694835680751,
-      "grad_norm": 0.9116912854037641,
+      "grad_norm": 0.688757844390044,
       "learning_rate": 2.4e-05,
-      "loss": 0.6834,
+      "loss": 0.6817,
       "step": 72
     },
     {
       "epoch": 0.3046426708398539,
-      "grad_norm": 0.816601373706202,
+      "grad_norm": 0.7296108335418863,
       "learning_rate": 2.4333333333333333e-05,
-      "loss": 0.6732,
+      "loss": 0.6723,
       "step": 73
     },
     {
       "epoch": 0.30881585811163276,
-      "grad_norm": 0.7983857073769415,
+      "grad_norm": 0.7836299209162253,
       "learning_rate": 2.466666666666667e-05,
       "loss": 0.6841,
       "step": 74
     },
     {
       "epoch": 0.3129890453834116,
-      "grad_norm": 0.8020664333303708,
+      "grad_norm": 0.8195324294774077,
       "learning_rate": 2.5e-05,
-      "loss": 0.6778,
+      "loss": 0.6774,
       "step": 75
     },
     {
       "epoch": 0.3171622326551904,
-      "grad_norm": 0.6215894300813539,
+      "grad_norm": 0.7081126560875635,
       "learning_rate": 2.5333333333333334e-05,
-      "loss": 0.6977,
+      "loss": 0.6981,
       "step": 76
     },
     {
       "epoch": 0.32133541992696923,
-      "grad_norm": 0.7250944228199042,
+      "grad_norm": 0.7527685190514966,
       "learning_rate": 2.566666666666667e-05,
-      "loss": 0.685,
+      "loss": 0.6848,
       "step": 77
     },
     {
       "epoch": 0.325508607198748,
-      "grad_norm": 0.7776877512356822,
+      "grad_norm": 0.6303435043304341,
       "learning_rate": 2.6000000000000002e-05,
-      "loss": 0.6537,
+      "loss": 0.6541,
       "step": 78
     },
     {
       "epoch": 0.32968179447052687,
-      "grad_norm": 0.732130293245925,
+      "grad_norm": 0.7732725641537613,
       "learning_rate": 2.6333333333333334e-05,
-      "loss": 0.6794,
+      "loss": 0.6796,
       "step": 79
     },
     {
       "epoch": 0.3338549817423057,
-      "grad_norm": 0.7014852987451703,
+      "grad_norm": 0.6954189787019706,
       "learning_rate": 2.6666666666666667e-05,
-      "loss": 0.6868,
+      "loss": 0.6867,
       "step": 80
     },
     {
       "epoch": 0.3380281690140845,
-      "grad_norm": 0.8677880096339924,
+      "grad_norm": 0.8400965120272477,
       "learning_rate": 2.7000000000000002e-05,
-      "loss": 0.7043,
+      "loss": 0.7041,
       "step": 81
     },
     {
       "epoch": 0.34220135628586335,
-      "grad_norm": 0.6674274265232697,
+      "grad_norm": 0.7687478601758556,
       "learning_rate": 2.7333333333333335e-05,
-      "loss": 0.6607,
+      "loss": 0.6615,
       "step": 82
     },
     {
       "epoch": 0.34637454355764213,
-      "grad_norm": 0.7685265641741991,
+      "grad_norm": 0.7691697431416561,
       "learning_rate": 2.7666666666666667e-05,
-      "loss": 0.6963,
+      "loss": 0.6968,
       "step": 83
     },
     {
       "epoch": 0.350547730829421,
-      "grad_norm": 0.7431458446229806,
+      "grad_norm": 0.7589559235840138,
       "learning_rate": 2.8e-05,
-      "loss": 0.639,
+      "loss": 0.6386,
       "step": 84
     },
     {
       "epoch": 0.35472091810119977,
-      "grad_norm": 0.623395886471966,
+      "grad_norm": 0.7810682969654213,
       "learning_rate": 2.833333333333334e-05,
-      "loss": 0.6731,
+      "loss": 0.6735,
       "step": 85
     },
     {
       "epoch": 0.3588941053729786,
-      "grad_norm": 0.9612666856896732,
+      "grad_norm": 0.8421206854195348,
       "learning_rate": 2.8666666666666668e-05,
-      "loss": 0.6491,
+      "loss": 0.6501,
       "step": 86
     },
     {
       "epoch": 0.36306729264475746,
-      "grad_norm": 0.908683777876259,
+      "grad_norm": 0.8728451699379745,
       "learning_rate": 2.9e-05,
-      "loss": 0.6467,
+      "loss": 0.6464,
       "step": 87
     },
     {
       "epoch": 0.36724047991653624,
-      "grad_norm": 0.6590462245554869,
+      "grad_norm": 0.855140737123505,
       "learning_rate": 2.9333333333333333e-05,
-      "loss": 0.6677,
+      "loss": 0.668,
       "step": 88
     },
     {
       "epoch": 0.3714136671883151,
-      "grad_norm": 0.75801570378802,
+      "grad_norm": 0.7367372807313465,
       "learning_rate": 2.9666666666666672e-05,
-      "loss": 0.6711,
+      "loss": 0.6717,
       "step": 89
     },
     {
       "epoch": 0.3755868544600939,
-      "grad_norm": 0.7972876198833736,
+      "grad_norm": 0.7462983195950743,
       "learning_rate": 3.0000000000000004e-05,
-      "loss": 0.6742,
+      "loss": 0.6737,
       "step": 90
     },
     {
       "epoch": 0.3797600417318727,
-      "grad_norm": 0.8944551152841265,
+      "grad_norm": 0.9237009423616097,
       "learning_rate": 3.0333333333333333e-05,
-      "loss": 0.667,
+      "loss": 0.6674,
       "step": 91
     },
     {
       "epoch": 0.38393322900365157,
-      "grad_norm": 0.7749784729215585,
+      "grad_norm": 0.9094444354447875,
       "learning_rate": 3.066666666666667e-05,
-      "loss": 0.6694,
+      "loss": 0.671,
       "step": 92
     },
     {
       "epoch": 0.38810641627543035,
-      "grad_norm": 0.8816144947259736,
+      "grad_norm": 0.9294634239648653,
       "learning_rate": 3.1e-05,
-      "loss": 0.6744,
+      "loss": 0.6748,
       "step": 93
     },
     {
       "epoch": 0.3922796035472092,
-      "grad_norm": 0.5910230015699827,
+      "grad_norm": 0.7917081526269463,
       "learning_rate": 3.1333333333333334e-05,
-      "loss": 0.6451,
+      "loss": 0.6456,
       "step": 94
     },
     {
       "epoch": 0.396452790818988,
-      "grad_norm": 0.8923123542202034,
+      "grad_norm": 0.7884602057931512,
       "learning_rate": 3.1666666666666666e-05,
-      "loss": 0.6565,
+      "loss": 0.6568,
       "step": 95
     },
     {
       "epoch": 0.40062597809076683,
-      "grad_norm": 0.771084402446723,
+      "grad_norm": 0.7516882630634099,
       "learning_rate": 3.2000000000000005e-05,
-      "loss": 0.6671,
+      "loss": 0.667,
       "step": 96
     },
     {
       "epoch": 0.4047991653625456,
-      "grad_norm": 0.8110369294252538,
+      "grad_norm": 0.7328373536910172,
       "learning_rate": 3.233333333333334e-05,
-      "loss": 0.645,
+      "loss": 0.6445,
       "step": 97
     },
     {
       "epoch": 0.40897235263432447,
-      "grad_norm": 0.7674572957725485,
+      "grad_norm": 0.7675785138446586,
       "learning_rate": 3.266666666666667e-05,
-      "loss": 0.6354,
+      "loss": 0.6361,
       "step": 98
     },
     {
       "epoch": 0.4131455399061033,
-      "grad_norm": 0.6721855773921498,
+      "grad_norm": 0.6957481900280716,
       "learning_rate": 3.3e-05,
-      "loss": 0.636,
+      "loss": 0.6354,
       "step": 99
     },
     {
       "epoch": 0.4173187271778821,
-      "grad_norm": 0.7676497692894325,
+      "grad_norm": 0.595488133955794,
       "learning_rate": 3.3333333333333335e-05,
-      "loss": 0.6896,
+      "loss": 0.688,
       "step": 100
     },
     {
       "epoch": 0.42149191444966094,
-      "grad_norm": 1.0389975750997646,
+      "grad_norm": 0.9745162775950517,
       "learning_rate": 3.366666666666667e-05,
-      "loss": 0.6916,
+      "loss": 0.6918,
       "step": 101
     },
     {
       "epoch": 0.42566510172143973,
-      "grad_norm": 0.9674451111564523,
+      "grad_norm": 1.0212678594186515,
       "learning_rate": 3.4e-05,
-      "loss": 0.6771,
+      "loss": 0.6762,
       "step": 102
     },
     {
       "epoch": 0.4298382889932186,
-      "grad_norm": 0.6889420438969273,
+      "grad_norm": 0.7516525122136645,
       "learning_rate": 3.433333333333333e-05,
-      "loss": 0.6365,
+      "loss": 0.6363,
       "step": 103
     },
     {
       "epoch": 0.43401147626499736,
-      "grad_norm": 0.6550007535911513,
+      "grad_norm": 0.7131604711122411,
       "learning_rate": 3.466666666666667e-05,
-      "loss": 0.6675,
+      "loss": 0.6674,
       "step": 104
     },
     {
       "epoch": 0.4381846635367762,
-      "grad_norm": 0.662973989407345,
+      "grad_norm": 0.7296595779308731,
       "learning_rate": 3.5000000000000004e-05,
-      "loss": 0.6767,
+      "loss": 0.676,
       "step": 105
     },
     {
       "epoch": 0.44235785080855505,
-      "grad_norm": 0.7211041564158813,
+      "grad_norm": 0.742803356921106,
       "learning_rate": 3.5333333333333336e-05,
       "loss": 0.6631,
       "step": 106
     },
     {
       "epoch": 0.44653103808033384,
-      "grad_norm": 0.8588290359583747,
+      "grad_norm": 0.9234187861191283,
       "learning_rate": 3.566666666666667e-05,
       "loss": 0.6783,
       "step": 107
     },
     {
       "epoch": 0.4507042253521127,
-      "grad_norm": 0.7866218586515221,
+      "grad_norm": 0.8252825804524461,
       "learning_rate": 3.6e-05,
-      "loss": 0.6513,
+      "loss": 0.6508,
       "step": 108
     },
     {
       "epoch": 0.4548774126238915,
-      "grad_norm": 0.6572777900524828,
+      "grad_norm": 0.5728522592791544,
       "learning_rate": 3.633333333333333e-05,
-      "loss": 0.6848,
+      "loss": 0.6839,
       "step": 109
     },
     {
       "epoch": 0.4590505998956703,
-      "grad_norm": 0.6970231554977443,
+      "grad_norm": 0.7566640714162131,
       "learning_rate": 3.6666666666666666e-05,
-      "loss": 0.6453,
+      "loss": 0.645,
       "step": 110
     },
     {
       "epoch": 0.46322378716744916,
-      "grad_norm": 0.7997816180795174,
+      "grad_norm": 0.872523923872913,
       "learning_rate": 3.7000000000000005e-05,
-      "loss": 0.6661,
+      "loss": 0.6663,
       "step": 111
     },
     {
       "epoch": 0.46739697443922795,
-      "grad_norm": 0.7788704910983766,
+      "grad_norm": 0.9901926960182357,
       "learning_rate": 3.733333333333334e-05,
-      "loss": 0.6566,
+      "loss": 0.6574,
       "step": 112
     },
     {
       "epoch": 0.4715701617110068,
-      "grad_norm": 0.7934291397764641,
+      "grad_norm": 0.8986681542538747,
       "learning_rate": 3.766666666666667e-05,
       "loss": 0.6491,
       "step": 113
     },
     {
       "epoch": 0.4757433489827856,
-      "grad_norm": 0.8157185359776757,
+      "grad_norm": 0.7602285162154537,
       "learning_rate": 3.8e-05,
-      "loss": 0.6566,
+      "loss": 0.6562,
       "step": 114
     },
     {
       "epoch": 0.47991653625456443,
-      "grad_norm": 0.8780934174547074,
+      "grad_norm": 0.7392216116240299,
       "learning_rate": 3.833333333333334e-05,
-      "loss": 0.6681,
+      "loss": 0.6667,
       "step": 115
     },
     {
       "epoch": 0.4840897235263432,
-      "grad_norm": 0.9934186919882819,
+      "grad_norm": 0.6537152899414258,
       "learning_rate": 3.866666666666667e-05,
-      "loss": 0.6377,
+      "loss": 0.635,
       "step": 116
     },
     {
       "epoch": 0.48826291079812206,
-      "grad_norm": 0.9359804582689443,
+      "grad_norm": 0.6999115894579441,
       "learning_rate": 3.9e-05,
-      "loss": 0.6487,
+      "loss": 0.6471,
       "step": 117
     },
     {
       "epoch": 0.4924360980699009,
-      "grad_norm": 0.8150139557847049,
+      "grad_norm": 0.8385139281046069,
       "learning_rate": 3.933333333333333e-05,
-      "loss": 0.6356,
+      "loss": 0.6345,
       "step": 118
     },
     {
       "epoch": 0.4966092853416797,
-      "grad_norm": 0.869989179503855,
+      "grad_norm": 0.8036615921231413,
       "learning_rate": 3.966666666666667e-05,
-      "loss": 0.6767,
+      "loss": 0.6758,
       "step": 119
     },
     {
       "epoch": 0.5007824726134585,
-      "grad_norm": 1.0548222884182141,
+      "grad_norm": 0.7857855740141692,
       "learning_rate": 4e-05,
-      "loss": 0.6882,
+      "loss": 0.6867,
       "step": 120
     },
     {
       "epoch": 0.5049556598852374,
-      "grad_norm": 0.9072953542445146,
+      "grad_norm": 0.7326227755404053,
       "learning_rate": 3.9999914595155206e-05,
-      "loss": 0.6607,
+      "loss": 0.6581,
       "step": 121
     },
     {
       "epoch": 0.5091288471570161,
-      "grad_norm": 0.792667635464213,
+      "grad_norm": 0.6728680902668054,
       "learning_rate": 3.999965838135022e-05,
-      "loss": 0.6375,
+      "loss": 0.6366,
       "step": 122
     },
     {
       "epoch": 0.513302034428795,
-      "grad_norm": 0.9141295502414879,
+      "grad_norm": 0.7576450807352835,
       "learning_rate": 3.999923136077322e-05,
-      "loss": 0.6718,
+      "loss": 0.6695,
       "step": 123
     },
     {
       "epoch": 0.5174752217005738,
-      "grad_norm": 1.1085787957524216,
+      "grad_norm": 0.9024582973130363,
       "learning_rate": 3.999863353707118e-05,
-      "loss": 0.6807,
+      "loss": 0.679,
       "step": 124
     },
     {
       "epoch": 0.5216484089723527,
-      "grad_norm": 0.6682050658471472,
+      "grad_norm": 0.6338203374743853,
       "learning_rate": 3.999786491534981e-05,
-      "loss": 0.6479,
+      "loss": 0.6465,
       "step": 125
     },
     {
       "epoch": 0.5258215962441315,
-      "grad_norm": 0.8859839417580586,
+      "grad_norm": 0.7044160447788558,
       "learning_rate": 3.9996925502173496e-05,
-      "loss": 0.6665,
+      "loss": 0.6642,
       "step": 126
     },
     {
       "epoch": 0.5299947835159102,
-      "grad_norm": 0.8124554187556665,
+      "grad_norm": 0.6299966392490217,
       "learning_rate": 3.999581530556529e-05,
-      "loss": 0.6486,
+      "loss": 0.6469,
       "step": 127
     },
     {
       "epoch": 0.5341679707876891,
-      "grad_norm": 0.7784074022399087,
+      "grad_norm": 0.7453798680223842,
       "learning_rate": 3.999453433500681e-05,
-      "loss": 0.6976,
+      "loss": 0.6968,
       "step": 128
     },
     {
       "epoch": 0.5383411580594679,
-      "grad_norm": 0.7608997858344617,
+      "grad_norm": 0.7943501350325215,
       "learning_rate": 3.999308260143816e-05,
-      "loss": 0.6642,
+      "loss": 0.6632,
       "step": 129
     },
     {
       "epoch": 0.5425143453312468,
-      "grad_norm": 0.724065260506949,
+      "grad_norm": 0.5884484112502454,
       "learning_rate": 3.9991460117257856e-05,
-      "loss": 0.6559,
+      "loss": 0.6545,
       "step": 130
     },
     {
       "epoch": 0.5466875326030256,
-      "grad_norm": 0.8113929310232064,
+      "grad_norm": 0.6160983354714633,
       "learning_rate": 3.998966689632269e-05,
-      "loss": 0.6556,
+      "loss": 0.6543,
       "step": 131
     },
     {
       "epoch": 0.5508607198748043,
-      "grad_norm": 0.66886048923594,
+      "grad_norm": 0.6218375885755708,
       "learning_rate": 3.998770295394765e-05,
-      "loss": 0.6555,
+      "loss": 0.6541,
       "step": 132
     },
     {
       "epoch": 0.5550339071465832,
-      "grad_norm": 0.815048817539772,
+      "grad_norm": 0.6685890369350261,
       "learning_rate": 3.9985568306905744e-05,
-      "loss": 0.6673,
+      "loss": 0.6655,
       "step": 133
     },
     {
       "epoch": 0.559207094418362,
-      "grad_norm": 0.6539691437870502,
+      "grad_norm": 0.5692188237572238,
       "learning_rate": 3.998326297342789e-05,
-      "loss": 0.627,
+      "loss": 0.6262,
       "step": 134
     },
     {
       "epoch": 0.5633802816901409,
-      "grad_norm": 0.825837634159562,
+      "grad_norm": 0.7893737490974674,
       "learning_rate": 3.998078697320277e-05,
-      "loss": 0.655,
+      "loss": 0.6539,
       "step": 135
     },
     {
       "epoch": 0.5675534689619197,
-      "grad_norm": 0.7358380636770038,
+      "grad_norm": 0.6864608897433194,
       "learning_rate": 3.9978140327376607e-05,
-      "loss": 0.6464,
+      "loss": 0.6454,
       "step": 136
     },
     {
       "epoch": 0.5717266562336984,
-      "grad_norm": 0.9913902957612841,
+      "grad_norm": 0.9301241064426671,
       "learning_rate": 3.997532305855305e-05,
-      "loss": 0.6649,
+      "loss": 0.6633,
       "step": 137
     },
     {
       "epoch": 0.5758998435054773,
-      "grad_norm": 1.0523421858080866,
+      "grad_norm": 0.9756482590442717,
       "learning_rate": 3.9972335190792924e-05,
-      "loss": 0.6492,
+      "loss": 0.6486,
       "step": 138
     },
     {
       "epoch": 0.5800730307772561,
-      "grad_norm": 0.8254864405588844,
+      "grad_norm": 0.782896028964664,
       "learning_rate": 3.9969176749614087e-05,
-      "loss": 0.6564,
+      "loss": 0.6555,
       "step": 139
     },
     {
       "epoch": 0.584246218049035,
-      "grad_norm": 0.8906778195535651,
+      "grad_norm": 0.773173491531386,
       "learning_rate": 3.996584776199114e-05,
-      "loss": 0.6444,
+      "loss": 0.6432,
       "step": 140
     },
     {
       "epoch": 0.5884194053208138,
-      "grad_norm": 0.9034523035644649,
+      "grad_norm": 0.6962122030055938,
       "learning_rate": 3.996234825635527e-05,
-      "loss": 0.6357,
+      "loss": 0.635,
       "step": 141
     },
     {
       "epoch": 0.5925925925925926,
-      "grad_norm": 0.8153065510943488,
+      "grad_norm": 0.8745062701358192,
       "learning_rate": 3.9958678262593936e-05,
-      "loss": 0.6522,
+      "loss": 0.652,
       "step": 142
     },
     {
       "epoch": 0.5967657798643714,
-      "grad_norm": 0.8538686521783709,
+      "grad_norm": 0.68857115468515,
       "learning_rate": 3.9954837812050656e-05,
-      "loss": 0.6647,
+      "loss": 0.6635,
       "step": 143
     },
     {
       "epoch": 0.6009389671361502,
-      "grad_norm": 1.1045464892598449,
+      "grad_norm": 0.9027030932731339,
       "learning_rate": 3.995082693752475e-05,
-      "loss": 0.6562,
+      "loss": 0.6549,
       "step": 144
     },
     {
       "epoch": 0.6051121544079291,
-      "grad_norm": 0.6456668231919385,
+      "grad_norm": 0.6130066209966619,
       "learning_rate": 3.9946645673271034e-05,
-      "loss": 0.6408,
+      "loss": 0.6403,
       "step": 145
     },
     {
       "epoch": 0.6092853416797078,
-      "grad_norm": 1.012334841661057,
+      "grad_norm": 0.9084242140250284,
       "learning_rate": 3.9942294054999525e-05,
-      "loss": 0.6393,
+      "loss": 0.6394,
       "step": 146
     },
     {
       "epoch": 0.6134585289514867,
-      "grad_norm": 0.6176665965476612,
+      "grad_norm": 0.5273990605412152,
       "learning_rate": 3.9937772119875144e-05,
-      "loss": 0.6227,
+      "loss": 0.6226,
       "step": 147
     },
     {
       "epoch": 0.6176317162232655,
-      "grad_norm": 0.8789897277760075,
+      "grad_norm": 0.7977795894410907,
       "learning_rate": 3.993307990651742e-05,
-      "loss": 0.6406,
+      "loss": 0.6395,
       "step": 148
     },
     {
       "epoch": 0.6218049034950444,
-      "grad_norm": 0.724972968872893,
+      "grad_norm": 0.8335545978572094,
       "learning_rate": 3.9928217455000114e-05,
-      "loss": 0.656,
+      "loss": 0.6557,
       "step": 149
     },
     {
       "epoch": 0.6259780907668232,
-      "grad_norm": 0.5929879699210887,
+      "grad_norm": 0.610198507275443,
       "learning_rate": 3.992318480685094e-05,
-      "loss": 0.677,
+      "loss": 0.6763,
       "step": 150
     },
     {
       "epoch": 0.6301512780386019,
-      "grad_norm": 0.6740299075007873,
+      "grad_norm": 0.684398740806449,
       "learning_rate": 3.9917982005051126e-05,
-      "loss": 0.6298,
+      "loss": 0.6296,
       "step": 151
     },
     {
       "epoch": 0.6343244653103808,
-      "grad_norm": 0.6428082358247227,
+      "grad_norm": 0.714348582668746,
       "learning_rate": 3.991260909403513e-05,
-      "loss": 0.6377,
+      "loss": 0.637,
       "step": 152
     },
     {
       "epoch": 0.6384976525821596,
-      "grad_norm": 0.5586079069962967,
+      "grad_norm": 0.586429807544148,
       "learning_rate": 3.9907066119690224e-05,
-      "loss": 0.6214,
+      "loss": 0.6212,
       "step": 153
     },
     {
       "epoch": 0.6426708398539385,
-      "grad_norm": 0.7548742159959368,
+      "grad_norm": 0.7530285600566017,
       "learning_rate": 3.990135312935608e-05,
-      "loss": 0.6491,
+      "loss": 0.6484,
       "step": 154
     },
     {
       "epoch": 0.6468440271257173,
-      "grad_norm": 0.6573186891362943,
+      "grad_norm": 0.7569104187206357,
       "learning_rate": 3.989547017182442e-05,
-      "loss": 0.6512,
+      "loss": 0.6518,
       "step": 155
     },
     {
       "epoch": 0.651017214397496,
-      "grad_norm": 0.6294953515088532,
+      "grad_norm": 0.6436626121823218,
       "learning_rate": 3.9889417297338534e-05,
-      "loss": 0.6612,
+      "loss": 0.6614,
       "step": 156
     },
     {
       "epoch": 0.6551904016692749,
-      "grad_norm": 0.5375425450003001,
+      "grad_norm": 0.5863090545055207,
       "learning_rate": 3.9883194557592914e-05,
-      "loss": 0.6163,
+      "loss": 0.6167,
       "step": 157
     },
     {
       "epoch": 0.6593635889410537,
-      "grad_norm": 0.52649011788648,
+      "grad_norm": 0.5421389229194848,
       "learning_rate": 3.987680200573277e-05,
-      "loss": 0.6254,
+      "loss": 0.6253,
       "step": 158
     },
     {
       "epoch": 0.6635367762128326,
-      "grad_norm": 0.5309190114378901,
+      "grad_norm": 0.5873027784143924,
       "learning_rate": 3.987023969635358e-05,
-      "loss": 0.6387,
+      "loss": 0.6388,
       "step": 159
     },
     {
       "epoch": 0.6677099634846114,
-      "grad_norm": 0.6564713560746893,
+      "grad_norm": 0.7060361389263138,
       "learning_rate": 3.986350768550066e-05,
-      "loss": 0.6498,
+      "loss": 0.6497,
       "step": 160
     },
     {
       "epoch": 0.6718831507563902,
-      "grad_norm": 0.5665571674753245,
+      "grad_norm": 0.6679804968491524,
       "learning_rate": 3.985660603066865e-05,
-      "loss": 0.6361,
+      "loss": 0.6357,
       "step": 161
     },
     {
       "epoch": 0.676056338028169,
-      "grad_norm": 0.6525976834658077,
+      "grad_norm": 0.7455394144499289,
       "learning_rate": 3.984953479080101e-05,
-      "loss": 0.6428,
+      "loss": 0.6427,
       "step": 162
     },
     {
       "epoch": 0.6802295252999478,
-      "grad_norm": 0.5907253846363355,
+      "grad_norm": 0.5719022026319952,
       "learning_rate": 3.9842294026289565e-05,
-      "loss": 0.6358,
+      "loss": 0.6355,
       "step": 163
     },
     {
       "epoch": 0.6844027125717267,
-      "grad_norm": 0.5698166541354494,
+      "grad_norm": 0.5489597377362311,
       "learning_rate": 3.983488379897394e-05,
-      "loss": 0.6467,
+      "loss": 0.6457,
       "step": 164
     },
     {
       "epoch": 0.6885758998435054,
-      "grad_norm": 0.5459744460749723,
+      "grad_norm": 0.5416609524105679,
       "learning_rate": 3.982730417214107e-05,
-      "loss": 0.6509,
+      "loss": 0.6505,
       "step": 165
     },
     {
       "epoch": 0.6927490871152843,
-      "grad_norm": 0.6954586631298829,
+      "grad_norm": 0.7050463069078242,
       "learning_rate": 3.9819555210524656e-05,
-      "loss": 0.6495,
+      "loss": 0.6492,
       "step": 166
     },
     {
       "epoch": 0.6969222743870631,
-      "grad_norm": 0.6098474584884502,
+      "grad_norm": 0.6674705141224672,
       "learning_rate": 3.9811636980304564e-05,
-      "loss": 0.6364,
+      "loss": 0.6361,
       "step": 167
     },
     {
       "epoch": 0.701095461658842,
-      "grad_norm": 0.5166129239338109,
+      "grad_norm": 0.5924607244151662,
       "learning_rate": 3.9803549549106327e-05,
-      "loss": 0.6298,
+      "loss": 0.6297,
       "step": 168
     },
     {
       "epoch": 0.7052686489306208,
-      "grad_norm": 0.5725537407989815,
+      "grad_norm": 0.5580440366161854,
       "learning_rate": 3.9795292986000527e-05,
-      "loss": 0.6599,
+      "loss": 0.6592,
       "step": 169
     },
     {
       "epoch": 0.7094418362023995,
-      "grad_norm": 0.5319631475600647,
+      "grad_norm": 0.5870062697023997,
       "learning_rate": 3.978686736150221e-05,
-      "loss": 0.6166,
+      "loss": 0.6169,
       "step": 170
     },
     {
       "epoch": 0.7136150234741784,
-      "grad_norm": 0.6449671051266603,
+      "grad_norm": 0.6387148025657381,
       "learning_rate": 3.977827274757029e-05,
-      "loss": 0.6367,
+      "loss": 0.6365,
       "step": 171
     },
     {
       "epoch": 0.7177882107459572,
-      "grad_norm": 0.6281130210581456,
+      "grad_norm": 0.6612660109312574,
       "learning_rate": 3.976950921760692e-05,
-      "loss": 0.648,
+      "loss": 0.6479,
       "step": 172
     },
     {
       "epoch": 0.7219613980177361,
-      "grad_norm": 0.5689661739657118,
+      "grad_norm": 0.6447154717842158,
       "learning_rate": 3.9760576846456925e-05,
-      "loss": 0.6448,
+      "loss": 0.6446,
       "step": 173
     },
     {
       "epoch": 0.7261345852895149,
-      "grad_norm": 0.5699413346092014,
+      "grad_norm": 0.6264063207694949,
       "learning_rate": 3.975147571040706e-05,
-      "loss": 0.6104,
+      "loss": 0.6107,
       "step": 174
     },
     {
       "epoch": 0.7303077725612936,
-      "grad_norm": 0.5639070424409668,
+      "grad_norm": 0.5797917802202457,
       "learning_rate": 3.9742205887185434e-05,
-      "loss": 0.6435,
+      "loss": 0.6431,
       "step": 175
     },
     {
       "epoch": 0.7344809598330725,
-      "grad_norm": 0.6942622379693919,
+      "grad_norm": 0.6475876908656809,
       "learning_rate": 3.973276745596084e-05,
       "loss": 0.6426,
       "step": 176
     },
     {
       "epoch": 0.7386541471048513,
-      "grad_norm": 0.6143368941831149,
+      "grad_norm": 0.6460024647878482,
       "learning_rate": 3.9723160497342046e-05,
-      "loss": 0.6281,
+      "loss": 0.6282,
       "step": 177
     },
     {
       "epoch": 0.7428273343766302,
-      "grad_norm": 0.6336052386727556,
+      "grad_norm": 0.6018735931970374,
       "learning_rate": 3.971338509337713e-05,
-      "loss": 0.6424,
+      "loss": 0.642,
       "step": 178
     },
     {
       "epoch": 0.747000521648409,
-      "grad_norm": 0.6620612998135131,
+      "grad_norm": 0.5882567575717327,
       "learning_rate": 3.9703441327552786e-05,
-      "loss": 0.6076,
+      "loss": 0.6072,
       "step": 179
     },
     {
       "epoch": 0.7511737089201878,
-      "grad_norm": 0.5832436046262549,
+      "grad_norm": 0.549576475753751,
       "learning_rate": 3.9693329284793586e-05,
-      "loss": 0.6321,
+      "loss": 0.632,
       "step": 180
     },
     {
       "epoch": 0.7553468961919666,
-      "grad_norm": 0.814520064087668,
+      "grad_norm": 0.706560521045152,
       "learning_rate": 3.968304905146128e-05,
-      "loss": 0.6408,
+      "loss": 0.6409,
       "step": 181
     },
     {
       "epoch": 0.7595200834637454,
-      "grad_norm": 0.6269037102572034,
+      "grad_norm": 0.6074161768094577,
       "learning_rate": 3.9672600715354036e-05,
-      "loss": 0.6404,
+      "loss": 0.6399,
       "step": 182
     },
     {
       "epoch": 0.7636932707355243,
-      "grad_norm": 0.6231792285190983,
+      "grad_norm": 0.6051388533914291,
       "learning_rate": 3.96619843657057e-05,
-      "loss": 0.6103,
+      "loss": 0.6096,
       "step": 183
     },
     {
       "epoch": 0.7678664580073031,
-      "grad_norm": 0.6549458365107317,
+      "grad_norm": 0.675692296338123,
       "learning_rate": 3.965120009318505e-05,
-      "loss": 0.6565,
+      "loss": 0.6564,
       "step": 184
     },
     {
       "epoch": 0.7720396452790819,
-      "grad_norm": 0.5747328150569168,
+      "grad_norm": 0.606255560062987,
       "learning_rate": 3.964024798989501e-05,
-      "loss": 0.6222,
+      "loss": 0.6217,
       "step": 185
     },
     {
       "epoch": 0.7762128325508607,
-      "grad_norm": 0.525215487400015,
+      "grad_norm": 0.5216204125841473,
       "learning_rate": 3.962912814937183e-05,
-      "loss": 0.613,
+      "loss": 0.6129,
       "step": 186
     },
     {
       "epoch": 0.7803860198226396,
-      "grad_norm": 0.6217512205524626,
+      "grad_norm": 0.614270351974787,
       "learning_rate": 3.9617840666584336e-05,
-      "loss": 0.6439,
+      "loss": 0.6434,
       "step": 187
     },
     {
       "epoch": 0.7845592070944184,
-      "grad_norm": 0.5430416491885799,
+      "grad_norm": 0.5817464002432234,
       "learning_rate": 3.9606385637933106e-05,
-      "loss": 0.638,
+      "loss": 0.6377,
       "step": 188
     },
     {
       "epoch": 0.7887323943661971,
-      "grad_norm": 0.562697283336813,
+      "grad_norm": 0.5760504040336559,
       "learning_rate": 3.9594763161249636e-05,
-      "loss": 0.6422,
+      "loss": 0.6418,
       "step": 189
     },
     {
       "epoch": 0.792905581637976,
-      "grad_norm": 0.6545846885671794,
+      "grad_norm": 0.6750191741838109,
       "learning_rate": 3.95829733357955e-05,
-      "loss": 0.6511,
+      "loss": 0.6507,
       "step": 190
     },
     {
       "epoch": 0.7970787689097548,
-      "grad_norm": 0.592879517066043,
+      "grad_norm": 0.6608068974353737,
       "learning_rate": 3.957101626226153e-05,
       "loss": 0.6244,
       "step": 191
     },
     {
       "epoch": 0.8012519561815337,
-      "grad_norm": 0.5387646553719002,
+      "grad_norm": 0.5378033299034131,
       "learning_rate": 3.955889204276692e-05,
-      "loss": 0.6296,
+      "loss": 0.6287,
       "step": 192
     },
     {
       "epoch": 0.8054251434533125,
-      "grad_norm": 0.585433814146166,
+      "grad_norm": 0.5897821642510458,
       "learning_rate": 3.9546600780858375e-05,
-      "loss": 0.6317,
+      "loss": 0.6312,
       "step": 193
     },
     {
       "epoch": 0.8095983307250912,
-      "grad_norm": 0.5102484485477775,
+      "grad_norm": 0.5694866755811208,
       "learning_rate": 3.9534142581509226e-05,
-      "loss": 0.614,
+      "loss": 0.6133,
       "step": 194
     },
     {
       "epoch": 0.8137715179968701,
-      "grad_norm": 0.534901667599485,
+      "grad_norm": 0.488248648696434,
       "learning_rate": 3.952151755111855e-05,
-      "loss": 0.6044,
+      "loss": 0.604,
       "step": 195
     },
     {
       "epoch": 0.8179447052686489,
-      "grad_norm": 0.5321866401573603,
+      "grad_norm": 0.547850105018518,
       "learning_rate": 3.9508725797510195e-05,
-      "loss": 0.6052,
+      "loss": 0.605,
       "step": 196
     },
     {
       "epoch": 0.8221178925404278,
-      "grad_norm": 0.5409714584484635,
+      "grad_norm": 0.5252773917313027,
       "learning_rate": 3.9495767429931954e-05,
-      "loss": 0.6428,
+      "loss": 0.6423,
       "step": 197
     },
     {
       "epoch": 0.8262910798122066,
-      "grad_norm": 0.49524937791249923,
+      "grad_norm": 0.5117860136932925,
       "learning_rate": 3.948264255905456e-05,
-      "loss": 0.6079,
+      "loss": 0.6076,
       "step": 198
     },
     {
       "epoch": 0.8304642670839854,
-      "grad_norm": 0.5590138552868782,
+      "grad_norm": 0.5409477390053851,
       "learning_rate": 3.9469351296970766e-05,
-      "loss": 0.6306,
+      "loss": 0.6295,
       "step": 199
     },
     {
       "epoch": 0.8346374543557642,
-      "grad_norm": 0.5771567259517794,
+      "grad_norm": 0.5626782681858226,
       "learning_rate": 3.945589375719439e-05,
-      "loss": 0.632,
+      "loss": 0.6314,
       "step": 200
     },
     {
       "epoch": 0.838810641627543,
-      "grad_norm": 0.6120846652488096,
+      "grad_norm": 0.6027697314397016,
       "learning_rate": 3.9442270054659346e-05,
-      "loss": 0.6349,
+      "loss": 0.634,
       "step": 201
     },
     {
       "epoch": 0.8429838288993219,
-      "grad_norm": 0.5933895379854097,
+      "grad_norm": 0.5875615209207461,
       "learning_rate": 3.942848030571865e-05,
-      "loss": 0.632,
+      "loss": 0.6312,
       "step": 202
     },
     {
       "epoch": 0.8471570161711007,
-      "grad_norm": 0.5674838618990911,
+      "grad_norm": 0.5563644657724026,
       "learning_rate": 3.941452462814344e-05,
-      "loss": 0.6264,
+      "loss": 0.6259,
       "step": 203
     },
     {
       "epoch": 0.8513302034428795,
-      "grad_norm": 0.5950676294806098,
+      "grad_norm": 0.5827119674585158,
       "learning_rate": 3.940040314112196e-05,
-      "loss": 0.6193,
+      "loss": 0.6185,
       "step": 204
     },
     {
       "epoch": 0.8555033907146583,
-      "grad_norm": 0.540594964198133,
+      "grad_norm": 0.4874389548844733,
       "learning_rate": 3.938611596525855e-05,
-      "loss": 0.6383,
+      "loss": 0.638,
       "step": 205
     },
     {
       "epoch": 0.8596765779864372,
-      "grad_norm": 0.6059564592375596,
+      "grad_norm": 0.5997262388688029,
       "learning_rate": 3.9371663222572625e-05,
-      "loss": 0.6097,
+      "loss": 0.609,
       "step": 206
     },
     {
       "epoch": 0.863849765258216,
-      "grad_norm": 0.5014594880515366,
+      "grad_norm": 0.48209766956668454,
       "learning_rate": 3.93570450364976e-05,
-      "loss": 0.5909,
+      "loss": 0.5906,
       "step": 207
     },
     {
       "epoch": 0.8680229525299947,
-      "grad_norm": 0.6263219009273808,
+      "grad_norm": 0.6163963818980518,
       "learning_rate": 3.934226153187987e-05,
-      "loss": 0.6143,
+      "loss": 0.614,
       "step": 208
     },
     {
       "epoch": 0.8721961398017736,
-      "grad_norm": 0.5550580062627204,
+      "grad_norm": 0.5583517212157739,
       "learning_rate": 3.932731283497772e-05,
-      "loss": 0.6242,
+      "loss": 0.6237,
       "step": 209
     },
     {
       "epoch": 0.8763693270735524,
-      "grad_norm": 0.5291356522526318,
+      "grad_norm": 0.5487492073296804,
       "learning_rate": 3.931219907346028e-05,
-      "loss": 0.6112,
+      "loss": 0.6109,
       "step": 210
     },
     {
       "epoch": 0.8805425143453313,
-      "grad_norm": 0.5077568157800889,
+      "grad_norm": 0.4961358255313624,
       "learning_rate": 3.9296920376406366e-05,
-      "loss": 0.6136,
+      "loss": 0.6134,
       "step": 211
     },
     {
       "epoch": 0.8847157016171101,
-      "grad_norm": 0.5756405440779171,
+      "grad_norm": 0.6003867398734435,
       "learning_rate": 3.9281476874303486e-05,
-      "loss": 0.6358,
+      "loss": 0.6357,
       "step": 212
     },
     {
       "epoch": 0.8888888888888888,
-      "grad_norm": 0.4586792017714081,
+      "grad_norm": 0.4531563815593199,
       "learning_rate": 3.926586869904662e-05,
-      "loss": 0.5902,
+      "loss": 0.5898,
       "step": 213
     },
     {
       "epoch": 0.8930620761606677,
-      "grad_norm": 0.5532507041103961,
+      "grad_norm": 0.5656171949335493,
       "learning_rate": 3.9250095983937126e-05,
-      "loss": 0.609,
+      "loss": 0.6091,
       "step": 214
     },
     {
       "epoch": 0.8972352634324465,
-      "grad_norm": 0.5621085177533175,
+      "grad_norm": 0.5909638767221259,
       "learning_rate": 3.923415886368166e-05,
-      "loss": 0.6248,
+      "loss": 0.6246,
       "step": 215
     },
     {
       "epoch": 0.9014084507042254,
-      "grad_norm": 0.5709456049662635,
+      "grad_norm": 0.552888153549357,
       "learning_rate": 3.921805747439094e-05,
-      "loss": 0.5824,
+      "loss": 0.582,
       "step": 216
     },
     {
       "epoch": 0.9055816379760042,
-      "grad_norm": 0.5625057613805576,
+      "grad_norm": 0.6126728534481112,
       "learning_rate": 3.920179195357863e-05,
-      "loss": 0.5863,
+      "loss": 0.586,
       "step": 217
     },
     {
       "epoch": 0.909754825247783,
-      "grad_norm": 0.6594081329209324,
+      "grad_norm": 0.5977914521987667,
       "learning_rate": 3.918536244016016e-05,
-      "loss": 0.6094,
+      "loss": 0.6087,
       "step": 218
     },
     {
       "epoch": 0.9139280125195618,
-      "grad_norm": 0.6639553284817552,
+      "grad_norm": 0.7486482073576546,
       "learning_rate": 3.9168769074451525e-05,
-      "loss": 0.5956,
+      "loss": 0.5954,
       "step": 219
     },
     {
       "epoch": 0.9181011997913406,
-      "grad_norm": 0.6450384241406592,
+      "grad_norm": 0.5961160532375186,
       "learning_rate": 3.915201199816812e-05,
-      "loss": 0.6014,
+      "loss": 0.6012,
       "step": 220
     },
     {
       "epoch": 0.9222743870631195,
-      "grad_norm": 0.5573149395054288,
+      "grad_norm": 0.559760230907037,
       "learning_rate": 3.9135091354423494e-05,
-      "loss": 0.597,
+      "loss": 0.5966,
       "step": 221
     },
     {
       "epoch": 0.9264475743348983,
-      "grad_norm": 0.5945948853658436,
+      "grad_norm": 0.6338329650945789,
       "learning_rate": 3.9118007287728134e-05,
-      "loss": 0.6153,
+      "loss": 0.6149,
       "step": 222
     },
     {
       "epoch": 0.9306207616066771,
-      "grad_norm": 0.4753935844202169,
+      "grad_norm": 0.5015904893125557,
       "learning_rate": 3.9100759943988246e-05,
-      "loss": 0.6044,
+      "loss": 0.6046,
       "step": 223
     },
     {
       "epoch": 0.9347939488784559,
-      "grad_norm": 0.4661631993867029,
+      "grad_norm": 0.44860581914341924,
       "learning_rate": 3.9083349470504515e-05,
-      "loss": 0.5953,
+      "loss": 0.5948,
       "step": 224
     },
     {
       "epoch": 0.9389671361502347,
-      "grad_norm": 0.5476532680881577,
+      "grad_norm": 0.5065187294604524,
       "learning_rate": 3.9065776015970815e-05,
-      "loss": 0.6082,
+      "loss": 0.6077,
       "step": 225
     },
     {
       "epoch": 0.9431403234220136,
-      "grad_norm": 0.4795018946634012,
+      "grad_norm": 0.4923988042621045,
       "learning_rate": 3.904803973047295e-05,
-      "loss": 0.6148,
+      "loss": 0.6146,
       "step": 226
     },
     {
       "epoch": 0.9473135106937924,
-      "grad_norm": 0.5079231172457688,
+      "grad_norm": 0.49304437823493363,
       "learning_rate": 3.90301407654874e-05,
-      "loss": 0.6253,
+      "loss": 0.6247,
       "step": 227
     },
     {
       "epoch": 0.9514866979655712,
-      "grad_norm": 0.5465873047319323,
+      "grad_norm": 0.578524308784894,
       "learning_rate": 3.9012079273879994e-05,
-      "loss": 0.6399,
+      "loss": 0.64,
       "step": 228
     },
     {
       "epoch": 0.95565988523735,
-      "grad_norm": 0.5576956126327195,
+      "grad_norm": 0.5209926527509162,
       "learning_rate": 3.899385540990462e-05,
-      "loss": 0.6137,
+      "loss": 0.6134,
       "step": 229
     },
     {
       "epoch": 0.9598330725091289,
-      "grad_norm": 0.5868025893298928,
+      "grad_norm": 0.6388678269912473,
       "learning_rate": 3.897546932920191e-05,
-      "loss": 0.6319,
+      "loss": 0.6318,
       "step": 230
     },
     {
       "epoch": 0.9640062597809077,
-      "grad_norm": 0.4934290527059776,
+      "grad_norm": 0.4926883660540984,
       "learning_rate": 3.895692118879791e-05,
       "loss": 0.5848,
       "step": 231
     },
     {
       "epoch": 0.9681794470526864,
-      "grad_norm": 0.5918705450192724,
+      "grad_norm": 0.610559037610613,
       "learning_rate": 3.893821114710271e-05,
-      "loss": 0.6259,
+      "loss": 0.6256,
       "step": 232
     },
     {
       "epoch": 0.9723526343244653,
-      "grad_norm": 0.7877507135647431,
+      "grad_norm": 0.7652198551885145,
       "learning_rate": 3.891933936390914e-05,
-      "loss": 0.6243,
+      "loss": 0.6239,
       "step": 233
     },
     {
       "epoch": 0.9765258215962441,
-      "grad_norm": 0.5089693013538835,
+      "grad_norm": 0.5440283224136585,
       "learning_rate": 3.890030600039136e-05,
-      "loss": 0.6076,
+      "loss": 0.6075,
       "step": 234
     },
     {
       "epoch": 0.980699008868023,
-      "grad_norm": 0.6004426517703592,
+      "grad_norm": 0.5971453522527747,
       "learning_rate": 3.8881111219103516e-05,
       "loss": 0.6155,
       "step": 235
     },
     {
       "epoch": 0.9848721961398018,
-      "grad_norm": 0.48823182993879954,
+      "grad_norm": 0.4792335763640177,
       "learning_rate": 3.8861755183978346e-05,
-      "loss": 0.5946,
+      "loss": 0.5944,
       "step": 236
     },
     {
       "epoch": 0.9890453834115805,
-      "grad_norm": 0.628918555102249,
+      "grad_norm": 0.6081831041716331,
       "learning_rate": 3.8842238060325784e-05,
-      "loss": 0.6109,
+      "loss": 0.6099,
       "step": 237
     },
     {
       "epoch": 0.9932185706833594,
-      "grad_norm": 0.4862296623320978,
+      "grad_norm": 0.48234901111602896,
       "learning_rate": 3.882256001483149e-05,
-      "loss": 0.6067,
+      "loss": 0.6065,
       "step": 238
     },
     {
       "epoch": 0.9973917579551382,
-      "grad_norm": 0.5519509046270297,
+      "grad_norm": 0.5718605178871159,
       "learning_rate": 3.880272121555553e-05,
-      "loss": 0.6073,
+      "loss": 0.6069,
       "step": 239
     },
     {
       "epoch": 1.001564945226917,
-      "grad_norm": 0.6274542245502949,
+      "grad_norm": 0.6675407242338394,
       "learning_rate": 3.878272183193085e-05,
-      "loss": 0.5611,
+      "loss": 0.5606,
       "step": 240
     },
     {
       "epoch": 1.0057381324986958,
-      "grad_norm": 0.5363451118896134,
+      "grad_norm": 0.5475913347110059,
       "learning_rate": 3.876256203476188e-05,
-      "loss": 0.4993,
+      "loss": 0.4976,
       "step": 241
     },
     {
       "epoch": 1.0099113197704748,
-      "grad_norm": 0.7212298482211988,
+      "grad_norm": 0.7392652453883887,
       "learning_rate": 3.874224199622306e-05,
-      "loss": 0.5213,
+      "loss": 0.5204,
       "step": 242
     },
     {
       "epoch": 1.0140845070422535,
-      "grad_norm": 0.7428237616545396,
+      "grad_norm": 0.7871328206570581,
       "learning_rate": 3.8721761889857347e-05,
-      "loss": 0.5187,
+      "loss": 0.5183,
       "step": 243
     },
     {
       "epoch": 1.0182576943140322,
-      "grad_norm": 0.6452224783996009,
+      "grad_norm": 0.5957997813397626,
       "learning_rate": 3.870112189057478e-05,
-      "loss": 0.5014,
+      "loss": 0.5003,
       "step": 244
     },
     {
       "epoch": 1.0224308815858112,
-      "grad_norm": 0.8376232428822534,
+      "grad_norm": 0.8118557067374162,
       "learning_rate": 3.868032217465097e-05,
-      "loss": 0.5216,
+      "loss": 0.5217,
       "step": 245
     },
     {
       "epoch": 1.02660406885759,
-      "grad_norm": 0.7412211064004958,
+      "grad_norm": 0.7642767395917595,
       "learning_rate": 3.865936291972554e-05,
-      "loss": 0.5105,
+      "loss": 0.5095,
       "step": 246
     },
     {
       "epoch": 1.0307772561293689,
-      "grad_norm": 0.6032271833306786,
+      "grad_norm": 0.5262876765875568,
       "learning_rate": 3.863824430480069e-05,
-      "loss": 0.4988,
+      "loss": 0.4983,
       "step": 247
     },
     {
       "epoch": 1.0349504434011476,
-      "grad_norm": 0.7562821459856861,
+      "grad_norm": 0.7301362898552848,
       "learning_rate": 3.861696651023963e-05,
-      "loss": 0.4966,
+      "loss": 0.4964,
       "step": 248
     },
     {
       "epoch": 1.0391236306729263,
-      "grad_norm": 0.5314971511162915,
+      "grad_norm": 0.5605029503304007,
       "learning_rate": 3.859552971776503e-05,
-      "loss": 0.5139,
+      "loss": 0.5141,
       "step": 249
     },
     {
       "epoch": 1.0432968179447053,
-      "grad_norm": 0.7526728881761683,
+      "grad_norm": 0.6768686601850316,
       "learning_rate": 3.857393411045749e-05,
-      "loss": 0.4995,
+      "loss": 0.4978,
       "step": 250
     },
     {
       "epoch": 1.047470005216484,
-      "grad_norm": 0.6815902539720552,
+      "grad_norm": 0.6544347414719751,
       "learning_rate": 3.8552179872753946e-05,
-      "loss": 0.5206,
+      "loss": 0.5197,
       "step": 251
     },
     {
       "epoch": 1.051643192488263,
-      "grad_norm": 0.6625362514742431,
+      "grad_norm": 0.7013771676165595,
       "learning_rate": 3.853026719044614e-05,
       "loss": 0.5105,
       "step": 252
     },
     {
       "epoch": 1.0558163797600417,
-      "grad_norm": 0.58414830276389,
+      "grad_norm": 0.5586180653789645,
       "learning_rate": 3.850819625067899e-05,
-      "loss": 0.493,
+      "loss": 0.491,
       "step": 253
     },
     {
       "epoch": 1.0599895670318205,
-      "grad_norm": 0.8148180690060455,
+      "grad_norm": 0.7547765515784276,
       "learning_rate": 3.8485967241949005e-05,
-      "loss": 0.4906,
+      "loss": 0.4908,
       "step": 254
     },
     {
       "epoch": 1.0641627543035994,
-      "grad_norm": 0.5154136918632315,
+      "grad_norm": 0.6025665209663417,
       "learning_rate": 3.846358035410271e-05,
-      "loss": 0.5103,
+      "loss": 0.5109,
       "step": 255
     },
     {
       "epoch": 1.0683359415753781,
-      "grad_norm": 0.630643589810588,
+      "grad_norm": 0.5705792201090882,
       "learning_rate": 3.844103577833496e-05,
-      "loss": 0.4698,
+      "loss": 0.4694,
       "step": 256
     },
     {
       "epoch": 1.072509128847157,
-      "grad_norm": 0.671333533478983,
+      "grad_norm": 0.6718542764865053,
       "learning_rate": 3.841833370718735e-05,
-      "loss": 0.4974,
+      "loss": 0.4971,
       "step": 257
     },
     {
       "epoch": 1.0766823161189358,
-      "grad_norm": 0.6361178687331301,
+      "grad_norm": 0.6255614475681779,
       "learning_rate": 3.8395474334546576e-05,
-      "loss": 0.5268,
+      "loss": 0.5263,
       "step": 258
     },
     {
       "epoch": 1.0808555033907146,
-      "grad_norm": 0.8952977529341792,
+      "grad_norm": 0.7128891043014302,
       "learning_rate": 3.837245785564275e-05,
-      "loss": 0.5262,
+      "loss": 0.5252,
       "step": 259
     },
     {
       "epoch": 1.0850286906624935,
-      "grad_norm": 0.6807942586877037,
+      "grad_norm": 0.7532015657481717,
       "learning_rate": 3.834928446704775e-05,
-      "loss": 0.5015,
+      "loss": 0.5014,
       "step": 260
     },
     {
       "epoch": 1.0892018779342723,
-      "grad_norm": 0.5621745872493934,
+      "grad_norm": 0.5244264413112153,
       "learning_rate": 3.8325954366673556e-05,
-      "loss": 0.4839,
+      "loss": 0.4811,
       "step": 261
     },
     {
       "epoch": 1.0933750652060512,
-      "grad_norm": 0.8091065482299167,
+      "grad_norm": 0.7777530179679978,
       "learning_rate": 3.830246775377052e-05,
-      "loss": 0.5249,
+      "loss": 0.5244,
       "step": 262
     },
     {
       "epoch": 1.09754825247783,
-      "grad_norm": 0.53962748433483,
+      "grad_norm": 0.6697959890686898,
       "learning_rate": 3.827882482892568e-05,
-      "loss": 0.5029,
+      "loss": 0.5027,
       "step": 263
     },
     {
       "epoch": 1.1017214397496087,
-      "grad_norm": 0.8278867511137611,
+      "grad_norm": 0.7013202402249538,
       "learning_rate": 3.825502579406109e-05,
-      "loss": 0.4993,
+      "loss": 0.4987,
       "step": 264
     },
     {
       "epoch": 1.1058946270213876,
-      "grad_norm": 0.7501045278085758,
+      "grad_norm": 0.856575807112864,
       "learning_rate": 3.8231070852432035e-05,
-      "loss": 0.5067,
+      "loss": 0.5071,
       "step": 265
     },
     {
       "epoch": 1.1100678142931664,
-      "grad_norm": 0.5843904178987996,
+      "grad_norm": 0.6531190444148646,
       "learning_rate": 3.8206960208625305e-05,
-      "loss": 0.5032,
+      "loss": 0.5034,
       "step": 266
     },
     {
       "epoch": 1.1142410015649453,
-      "grad_norm": 0.6828386567972535,
+      "grad_norm": 0.6259209456830811,
       "learning_rate": 3.8182694068557505e-05,
-      "loss": 0.4946,
+      "loss": 0.4945,
       "step": 267
     },
     {
       "epoch": 1.118414188836724,
-      "grad_norm": 0.7048790926095468,
+      "grad_norm": 0.8274562726235793,
       "learning_rate": 3.81582726394732e-05,
-      "loss": 0.5099,
+      "loss": 0.5102,
       "step": 268
     },
     {
       "epoch": 1.1225873761085028,
-      "grad_norm": 0.5166777344625786,
+      "grad_norm": 0.5329270268250417,
       "learning_rate": 3.813369612994325e-05,
-      "loss": 0.4889,
+      "loss": 0.489,
       "step": 269
     },
     {
       "epoch": 1.1267605633802817,
-      "grad_norm": 0.6460516582573351,
+      "grad_norm": 0.696024243123464,
       "learning_rate": 3.810896474986294e-05,
-      "loss": 0.503,
+      "loss": 0.5027,
       "step": 270
     },
     {
       "epoch": 1.1309337506520605,
-      "grad_norm": 0.48954678289854975,
+      "grad_norm": 0.6026839594553245,
       "learning_rate": 3.8084078710450236e-05,
-      "loss": 0.5168,
+      "loss": 0.5174,
       "step": 271
     },
     {
       "epoch": 1.1351069379238394,
-      "grad_norm": 0.6071747691265882,
+      "grad_norm": 0.560652156491731,
       "learning_rate": 3.8059038224243984e-05,
       "loss": 0.5194,
       "step": 272
     },
     {
       "epoch": 1.1392801251956182,
-      "grad_norm": 0.4826780038931815,
+      "grad_norm": 0.5867810753559813,
       "learning_rate": 3.8033843505102056e-05,
-      "loss": 0.5052,
+      "loss": 0.5059,
       "step": 273
     },
     {
       "epoch": 1.143453312467397,
-      "grad_norm": 0.549084148125952,
+      "grad_norm": 0.5052291885618043,
       "learning_rate": 3.8008494768199554e-05,
-      "loss": 0.507,
+      "loss": 0.5076,
       "step": 274
     },
     {
       "epoch": 1.1476264997391759,
-      "grad_norm": 0.5687611327105874,
+      "grad_norm": 0.639507308098485,
       "learning_rate": 3.7982992230027e-05,
-      "loss": 0.504,
+      "loss": 0.5044,
       "step": 275
     },
     {
       "epoch": 1.1517996870109546,
-      "grad_norm": 0.4699880317166034,
+      "grad_norm": 0.5307906127265674,
       "learning_rate": 3.79573361083884e-05,
-      "loss": 0.5144,
+      "loss": 0.5156,
       "step": 276
     },
     {
       "epoch": 1.1559728742827335,
-      "grad_norm": 0.5038519877776471,
+      "grad_norm": 0.5462463846173543,
       "learning_rate": 3.793152662239947e-05,
-      "loss": 0.4858,
+      "loss": 0.4866,
       "step": 277
     },
     {
       "epoch": 1.1601460615545123,
-      "grad_norm": 0.4686349512343227,
+      "grad_norm": 0.45653791566988494,
       "learning_rate": 3.790556399248572e-05,
-      "loss": 0.5001,
+      "loss": 0.5008,
       "step": 278
     },
     {
       "epoch": 1.164319248826291,
-      "grad_norm": 0.6284548777027986,
+      "grad_norm": 0.623509620940565,
       "learning_rate": 3.78794484403806e-05,
       "loss": 0.5124,
       "step": 279
     },
     {
       "epoch": 1.16849243609807,
-      "grad_norm": 0.5197308389930859,
+      "grad_norm": 0.5211872749104363,
       "learning_rate": 3.785318018912357e-05,
-      "loss": 0.4964,
+      "loss": 0.4968,
       "step": 280
     },
     {
       "epoch": 1.1726656233698487,
-      "grad_norm": 0.5373410972528282,
+      "grad_norm": 0.5256474278090165,
       "learning_rate": 3.782675946305822e-05,
-      "loss": 0.4877,
+      "loss": 0.4883,
       "step": 281
     },
     {
       "epoch": 1.1768388106416277,
-      "grad_norm": 0.5147002401325995,
+      "grad_norm": 0.5346390485968523,
       "learning_rate": 3.780018648783035e-05,
-      "loss": 0.4905,
+      "loss": 0.4901,
       "step": 282
     },
     {
       "epoch": 1.1810119979134064,
-      "grad_norm": 0.5511397765368057,
+      "grad_norm": 0.5610102918659712,
       "learning_rate": 3.7773461490386045e-05,
-      "loss": 0.5077,
+      "loss": 0.508,
       "step": 283
     },
     {
       "epoch": 1.1851851851851851,
-      "grad_norm": 0.4913754365118303,
+      "grad_norm": 0.4998064445978452,
       "learning_rate": 3.774658469896973e-05,
-      "loss": 0.5112,
+      "loss": 0.5117,
       "step": 284
     },
     {
       "epoch": 1.189358372456964,
-      "grad_norm": 0.4909956357951322,
+      "grad_norm": 0.4932231066326169,
       "learning_rate": 3.7719556343122236e-05,
-      "loss": 0.5081,
+      "loss": 0.5083,
       "step": 285
     },
     {
       "epoch": 1.1935315597287428,
-      "grad_norm": 0.5439298807334857,
+      "grad_norm": 0.52222325465425,
       "learning_rate": 3.76923766536788e-05,
-      "loss": 0.5203,
+      "loss": 0.5202,
       "step": 286
     },
     {
       "epoch": 1.1977047470005218,
-      "grad_norm": 0.4873176950891612,
+      "grad_norm": 0.4942388148053565,
       "learning_rate": 3.766504586276714e-05,
-      "loss": 0.4928,
+      "loss": 0.493,
       "step": 287
     },
     {
       "epoch": 1.2018779342723005,
-      "grad_norm": 0.45361600260919854,
+      "grad_norm": 0.5021926583995827,
       "learning_rate": 3.763756420380547e-05,
-      "loss": 0.4832,
+      "loss": 0.4841,
       "step": 288
     },
     {
       "epoch": 1.2060511215440792,
-      "grad_norm": 0.5179655362303335,
+      "grad_norm": 0.5630253238528399,
       "learning_rate": 3.760993191150045e-05,
-      "loss": 0.487,
+      "loss": 0.4873,
       "step": 289
     },
     {
       "epoch": 1.2102243088158582,
-      "grad_norm": 0.4668182879427588,
+      "grad_norm": 0.46831133849950685,
       "learning_rate": 3.758214922184525e-05,
-      "loss": 0.5044,
+      "loss": 0.5045,
       "step": 290
     },
     {
       "epoch": 1.214397496087637,
-      "grad_norm": 0.4162488669093591,
+      "grad_norm": 0.48910931890224557,
       "learning_rate": 3.75542163721175e-05,
-      "loss": 0.5156,
+      "loss": 0.5157,
       "step": 291
     },
     {
       "epoch": 1.2185706833594159,
-      "grad_norm": 0.4726301957346717,
+      "grad_norm": 0.48714157388232676,
       "learning_rate": 3.7526133600877275e-05,
-      "loss": 0.5036,
+      "loss": 0.5034,
       "step": 292
     },
     {
       "epoch": 1.2227438706311946,
-      "grad_norm": 0.4708989904209839,
+      "grad_norm": 0.5444752525699794,
       "learning_rate": 3.749790114796504e-05,
-      "loss": 0.5067,
+      "loss": 0.5058,
       "step": 293
     },
     {
       "epoch": 1.2269170579029733,
-      "grad_norm": 0.4988340738708613,
+      "grad_norm": 0.48201671874113244,
       "learning_rate": 3.746951925449963e-05,
-      "loss": 0.5168,
+      "loss": 0.5163,
       "step": 294
     },
     {
       "epoch": 1.2310902451747523,
-      "grad_norm": 0.4786148783240318,
+      "grad_norm": 0.5229570123999121,
       "learning_rate": 3.744098816287616e-05,
-      "loss": 0.5026,
+      "loss": 0.5033,
       "step": 295
     },
     {
       "epoch": 1.235263432446531,
-      "grad_norm": 0.4314885193737111,
+      "grad_norm": 0.47014658580461255,
       "learning_rate": 3.741230811676396e-05,
-      "loss": 0.4941,
+      "loss": 0.4944,
       "step": 296
     },
     {
       "epoch": 1.2394366197183098,
-      "grad_norm": 0.49683353692386345,
+      "grad_norm": 0.5148050583163014,
       "learning_rate": 3.738347936110455e-05,
-      "loss": 0.5094,
+      "loss": 0.5082,
       "step": 297
     },
     {
       "epoch": 1.2436098069900887,
-      "grad_norm": 0.4928797025743165,
+      "grad_norm": 0.603823481205731,
       "learning_rate": 3.735450214210945e-05,
-      "loss": 0.514,
+      "loss": 0.5142,
       "step": 298
     },
     {
       "epoch": 1.2477829942618675,
-      "grad_norm": 0.5008799783918723,
+      "grad_norm": 0.5103006001216412,
       "learning_rate": 3.7325376707258146e-05,
       "loss": 0.5342,
       "step": 299
     },
     {
       "epoch": 1.2519561815336462,
-      "grad_norm": 0.48821696900832207,
+      "grad_norm": 0.5427041345699857,
       "learning_rate": 3.729610330529598e-05,
-      "loss": 0.504,
+      "loss": 0.5042,
       "step": 300
     },
     {
       "epoch": 1.2561293688054251,
-      "grad_norm": 0.5006940957598758,
+      "grad_norm": 0.4719299897168662,
       "learning_rate": 3.7266682186231975e-05,
-      "loss": 0.507,
+      "loss": 0.5068,
       "step": 301
     },
     {
       "epoch": 1.260302556077204,
-      "grad_norm": 0.5430387398836919,
+      "grad_norm": 0.5424207708711704,
       "learning_rate": 3.7237113601336735e-05,
-      "loss": 0.4908,
+      "loss": 0.4907,
       "step": 302
     },
     {
       "epoch": 1.2644757433489828,
-      "grad_norm": 0.4703723876649478,
+      "grad_norm": 0.49069011231257487,
       "learning_rate": 3.720739780314031e-05,
-      "loss": 0.4847,
+      "loss": 0.4848,
       "step": 303
     },
     {
       "epoch": 1.2686489306207616,
-      "grad_norm": 0.4818024070057611,
+      "grad_norm": 0.45882868499816754,
       "learning_rate": 3.717753504543001e-05,
-      "loss": 0.513,
+      "loss": 0.5123,
       "step": 304
     },
     {
       "epoch": 1.2728221178925403,
-      "grad_norm": 0.43921969829469215,
+      "grad_norm": 0.4718571291233123,
       "learning_rate": 3.7147525583248264e-05,
-      "loss": 0.5113,
+      "loss": 0.5121,
       "step": 305
     },
     {
       "epoch": 1.2769953051643192,
-      "grad_norm": 0.46561077600831974,
+      "grad_norm": 0.4566967248182435,
       "learning_rate": 3.711736967289041e-05,
-      "loss": 0.5063,
+      "loss": 0.5065,
       "step": 306
     },
     {
       "epoch": 1.2811684924360982,
-      "grad_norm": 0.48835959551503,
+      "grad_norm": 0.5194981943220417,
       "learning_rate": 3.708706757190253e-05,
-      "loss": 0.4854,
+      "loss": 0.4855,
       "step": 307
     },
     {
       "epoch": 1.285341679707877,
-      "grad_norm": 0.4500633049097859,
+      "grad_norm": 0.4554417784670908,
       "learning_rate": 3.7056619539079244e-05,
-      "loss": 0.48,
+      "loss": 0.4795,
       "step": 308
     },
     {
       "epoch": 1.2895148669796557,
-      "grad_norm": 0.4768836665452566,
+      "grad_norm": 0.4770562732072894,
       "learning_rate": 3.702602583446151e-05,
-      "loss": 0.5182,
+      "loss": 0.5172,
       "step": 309
     },
     {
       "epoch": 1.2936880542514344,
-      "grad_norm": 0.5727595152564913,
+      "grad_norm": 0.556421895141304,
       "learning_rate": 3.6995286719334385e-05,
-      "loss": 0.5252,
+      "loss": 0.5243,
       "step": 310
     },
     {
       "epoch": 1.2978612415232134,
-      "grad_norm": 0.46102080009018015,
+      "grad_norm": 0.47968427713892275,
       "learning_rate": 3.696440245622481e-05,
-      "loss": 0.4894,
+      "loss": 0.4901,
       "step": 311
     },
     {
       "epoch": 1.302034428794992,
-      "grad_norm": 0.4879595887626029,
+      "grad_norm": 0.4805946397215483,
       "learning_rate": 3.6933373308899344e-05,
-      "loss": 0.5033,
+      "loss": 0.5026,
       "step": 312
     },
     {
       "epoch": 1.306207616066771,
-      "grad_norm": 0.49067737230065245,
+      "grad_norm": 0.5012940340777688,
       "learning_rate": 3.690219954236195e-05,
       "loss": 0.5236,
       "step": 313
     },
     {
       "epoch": 1.3103808033385498,
-      "grad_norm": 0.46168228527806887,
+      "grad_norm": 0.46168288955770526,
       "learning_rate": 3.6870881422851696e-05,
-      "loss": 0.5048,
+      "loss": 0.505,
       "step": 314
     },
     {
       "epoch": 1.3145539906103285,
-      "grad_norm": 0.5010969929254614,
+      "grad_norm": 0.46903320077725685,
       "learning_rate": 3.6839419217840486e-05,
-      "loss": 0.5122,
+      "loss": 0.5123,
       "step": 315
     },
     {
       "epoch": 1.3187271778821075,
-      "grad_norm": 0.6379766147701686,
+      "grad_norm": 0.5911605776824526,
       "learning_rate": 3.68078131960308e-05,
-      "loss": 0.5183,
+      "loss": 0.5185,
       "step": 316
     },
     {
       "epoch": 1.3229003651538862,
-      "grad_norm": 0.4330152525599903,
+      "grad_norm": 0.44730006155614277,
       "learning_rate": 3.677606362735338e-05,
-      "loss": 0.5173,
+      "loss": 0.5169,
       "step": 317
     },
     {
       "epoch": 1.3270735524256652,
-      "grad_norm": 0.6495159699850558,
+      "grad_norm": 0.5868203649663003,
       "learning_rate": 3.674417078296491e-05,
-      "loss": 0.5219,
+      "loss": 0.52,
       "step": 318
     },
     {
       "epoch": 1.331246739697444,
-      "grad_norm": 0.48990454861821153,
+      "grad_norm": 0.49096147984021726,
       "learning_rate": 3.671213493524576e-05,
-      "loss": 0.51,
+      "loss": 0.5101,
       "step": 319
     },
     {
       "epoch": 1.3354199269692226,
-      "grad_norm": 0.5479899079754765,
+      "grad_norm": 0.5402280905560535,
       "learning_rate": 3.667995635779756e-05,
-      "loss": 0.5276,
+      "loss": 0.5264,
       "step": 320
     },
     {
       "epoch": 1.3395931142410016,
-      "grad_norm": 0.6216136381480168,
+      "grad_norm": 0.6000727326106444,
       "learning_rate": 3.664763532544097e-05,
-      "loss": 0.5124,
+      "loss": 0.5115,
       "step": 321
     },
     {
       "epoch": 1.3437663015127803,
-      "grad_norm": 0.45541970384715397,
+      "grad_norm": 0.46569009956519125,
       "learning_rate": 3.6615172114213257e-05,
-      "loss": 0.5044,
+      "loss": 0.5043,
       "step": 322
     },
     {
       "epoch": 1.3479394887845593,
-      "grad_norm": 0.5994059491731218,
+      "grad_norm": 0.5742352561969079,
       "learning_rate": 3.658256700136598e-05,
-      "loss": 0.505,
+      "loss": 0.5055,
       "step": 323
     },
     {
       "epoch": 1.352112676056338,
-      "grad_norm": 0.6274798457856241,
+      "grad_norm": 0.6173055484263121,
       "learning_rate": 3.654982026536259e-05,
-      "loss": 0.5183,
+      "loss": 0.5168,
       "step": 324
     },
     {
       "epoch": 1.3562858633281167,
-      "grad_norm": 0.4755061280155073,
+      "grad_norm": 0.4527704272622455,
       "learning_rate": 3.6516932185876085e-05,
-      "loss": 0.5087,
+      "loss": 0.5088,
       "step": 325
     },
     {
       "epoch": 1.3604590505998957,
-      "grad_norm": 0.6382938439686432,
+      "grad_norm": 0.6368518868627764,
       "learning_rate": 3.6483903043786595e-05,
-      "loss": 0.5114,
+      "loss": 0.5116,
       "step": 326
     },
     {
       "epoch": 1.3646322378716744,
-      "grad_norm": 0.44860789026950537,
+      "grad_norm": 0.43823142391921466,
       "learning_rate": 3.6450733121179e-05,
-      "loss": 0.5059,
+      "loss": 0.5053,
       "step": 327
     },
     {
       "epoch": 1.3688054251434534,
-      "grad_norm": 0.5719902919744003,
+      "grad_norm": 0.5573114027103476,
       "learning_rate": 3.64174227013405e-05,
-      "loss": 0.5112,
+      "loss": 0.5092,
       "step": 328
     },
     {
       "epoch": 1.3729786124152321,
-      "grad_norm": 0.5277943375673229,
+      "grad_norm": 0.5438187161588139,
       "learning_rate": 3.638397206875822e-05,
       "loss": 0.508,
       "step": 329
     },
     {
       "epoch": 1.3771517996870108,
-      "grad_norm": 0.5094079308701441,
+      "grad_norm": 0.47099236541411893,
       "learning_rate": 3.635038150911677e-05,
-      "loss": 0.4963,
+      "loss": 0.4952,
       "step": 330
     },
     {
       "epoch": 1.3813249869587898,
-      "grad_norm": 0.6136648569349284,
+      "grad_norm": 0.6208684824508853,
       "learning_rate": 3.6316651309295805e-05,
-      "loss": 0.5186,
+      "loss": 0.5178,
       "step": 331
     },
     {
       "epoch": 1.3854981742305685,
-      "grad_norm": 0.48873798711643984,
+      "grad_norm": 0.4618051450395384,
       "learning_rate": 3.628278175736758e-05,
-      "loss": 0.516,
+      "loss": 0.5154,
       "step": 332
     },
     {
       "epoch": 1.3896713615023475,
-      "grad_norm": 0.6489887405564713,
+      "grad_norm": 0.5872415946943532,
       "learning_rate": 3.624877314259446e-05,
-      "loss": 0.4983,
+      "loss": 0.4976,
       "step": 333
     },
     {
       "epoch": 1.3938445487741262,
-      "grad_norm": 0.4715519648283572,
+      "grad_norm": 0.46105169299374477,
       "learning_rate": 3.6214625755426516e-05,
-      "loss": 0.4867,
+      "loss": 0.4859,
       "step": 334
     },
     {
       "epoch": 1.398017736045905,
-      "grad_norm": 0.5552608835426576,
+      "grad_norm": 0.5107837006153852,
       "learning_rate": 3.6180339887498953e-05,
-      "loss": 0.5218,
+      "loss": 0.5212,
       "step": 335
     },
     {
       "epoch": 1.402190923317684,
-      "grad_norm": 0.4639902231157389,
+      "grad_norm": 0.4517742336627733,
       "learning_rate": 3.6145915831629704e-05,
       "loss": 0.5086,
       "step": 336
     },
     {
       "epoch": 1.4063641105894626,
-      "grad_norm": 0.4832360689544198,
+      "grad_norm": 0.4468473612988057,
       "learning_rate": 3.611135388181689e-05,
-      "loss": 0.5142,
+      "loss": 0.514,
       "step": 337
     },
     {
       "epoch": 1.4105372978612416,
-      "grad_norm": 0.48781555006062244,
+      "grad_norm": 0.46771019316499396,
       "learning_rate": 3.607665433323631e-05,
-      "loss": 0.5092,
+      "loss": 0.509,
       "step": 338
     },
     {
       "epoch": 1.4147104851330203,
-      "grad_norm": 0.5298288850895786,
+      "grad_norm": 0.5365062538475225,
       "learning_rate": 3.6041817482238894e-05,
-      "loss": 0.5193,
+      "loss": 0.518,
       "step": 339
     },
     {
       "epoch": 1.418883672404799,
-      "grad_norm": 0.4910091874753909,
+      "grad_norm": 0.4781248040422891,
       "learning_rate": 3.600684362634826e-05,
-      "loss": 0.5111,
+      "loss": 0.5104,
       "step": 340
     },
     {
       "epoch": 1.423056859676578,
-      "grad_norm": 0.45183541072144645,
+      "grad_norm": 0.4686270248002447,
       "learning_rate": 3.597173306425806e-05,
-      "loss": 0.478,
+      "loss": 0.4777,
       "step": 341
     },
     {
       "epoch": 1.4272300469483568,
-      "grad_norm": 0.45918216630456266,
+      "grad_norm": 0.4664162209402139,
       "learning_rate": 3.5936486095829505e-05,
-      "loss": 0.496,
+      "loss": 0.4961,
       "step": 342
     },
     {
       "epoch": 1.4314032342201357,
-      "grad_norm": 0.561292324421344,
+      "grad_norm": 0.5729708044369957,
       "learning_rate": 3.59011030220888e-05,
       "loss": 0.5249,
       "step": 343
     },
     {
       "epoch": 1.4355764214919144,
-      "grad_norm": 0.4728809887701109,
+      "grad_norm": 0.4727943757867601,
       "learning_rate": 3.586558414522452e-05,
-      "loss": 0.5027,
+      "loss": 0.5025,
       "step": 344
     },
     {
       "epoch": 1.4397496087636932,
-      "grad_norm": 0.5918338322132645,
+      "grad_norm": 0.5948766169904514,
       "learning_rate": 3.5829929768585086e-05,
-      "loss": 0.5054,
+      "loss": 0.5058,
       "step": 345
     },
     {
       "epoch": 1.4439227960354721,
-      "grad_norm": 0.483319698307783,
+      "grad_norm": 0.4856371264728164,
       "learning_rate": 3.579414019667615e-05,
       "loss": 0.5148,
       "step": 346
     },
     {
       "epoch": 1.4480959833072509,
-      "grad_norm": 0.53915248530229,
+      "grad_norm": 0.5260445680793728,
       "learning_rate": 3.5758215735157995e-05,
-      "loss": 0.499,
+      "loss": 0.4992,
       "step": 347
     },
     {
       "epoch": 1.4522691705790298,
-      "grad_norm": 0.4402784791794624,
+      "grad_norm": 0.44769778873368477,
       "learning_rate": 3.572215669084292e-05,
-      "loss": 0.5188,
+      "loss": 0.5183,
       "step": 348
     },
     {
       "epoch": 1.4564423578508086,
-      "grad_norm": 0.6009188159306901,
+      "grad_norm": 0.6107328291446028,
       "learning_rate": 3.5685963371692656e-05,
-      "loss": 0.5164,
+      "loss": 0.5165,
       "step": 349
     },
     {
       "epoch": 1.4606155451225873,
-      "grad_norm": 0.5326147178995722,
+      "grad_norm": 0.588027873822247,
       "learning_rate": 3.5649636086815656e-05,
-      "loss": 0.5156,
+      "loss": 0.5163,
       "step": 350
     },
     {
       "epoch": 1.4647887323943662,
-      "grad_norm": 0.5591140137554409,
+      "grad_norm": 0.5437654908862458,
       "learning_rate": 3.5613175146464544e-05,
-      "loss": 0.4856,
+      "loss": 0.4845,
       "step": 351
     },
     {
       "epoch": 1.468961919666145,
-      "grad_norm": 0.6347574829550262,
+      "grad_norm": 0.6820695214963703,
       "learning_rate": 3.557658086203342e-05,
-      "loss": 0.523,
+      "loss": 0.5233,
       "step": 352
     },
     {
       "epoch": 1.473135106937924,
-      "grad_norm": 0.4958102597250434,
+      "grad_norm": 0.5067984001452149,
       "learning_rate": 3.553985354605519e-05,
       "loss": 0.5045,
       "step": 353
     },
     {
       "epoch": 1.4773082942097027,
-      "grad_norm": 0.5145799312060604,
+      "grad_norm": 0.5277382085494268,
       "learning_rate": 3.5502993512198936e-05,
-      "loss": 0.5012,
+      "loss": 0.5016,
       "step": 354
     },
     {
       "epoch": 1.4814814814814814,
-      "grad_norm": 0.4672366649437367,
+      "grad_norm": 0.5013026484293635,
       "learning_rate": 3.546600107526721e-05,
-      "loss": 0.513,
+      "loss": 0.512,
       "step": 355
     },
     {
       "epoch": 1.4856546687532604,
-      "grad_norm": 0.4917858644759443,
+      "grad_norm": 0.5060147493295996,
       "learning_rate": 3.5428876551193324e-05,
-      "loss": 0.4868,
+      "loss": 0.4866,
       "step": 356
     },
     {
       "epoch": 1.489827856025039,
-      "grad_norm": 0.5048889128466759,
+      "grad_norm": 0.5245039683200435,
       "learning_rate": 3.539162025703873e-05,
-      "loss": 0.5072,
+      "loss": 0.5075,
       "step": 357
     },
     {
       "epoch": 1.494001043296818,
-      "grad_norm": 0.46252301237203036,
+      "grad_norm": 0.47336222129155275,
       "learning_rate": 3.5354232510990205e-05,
-      "loss": 0.5349,
+      "loss": 0.5348,
       "step": 358
     },
     {
       "epoch": 1.4981742305685968,
-      "grad_norm": 0.5759765163808653,
+      "grad_norm": 0.5622698452439008,
       "learning_rate": 3.531671363235722e-05,
-      "loss": 0.4951,
+      "loss": 0.4954,
       "step": 359
     },
     {
       "epoch": 1.5023474178403755,
-      "grad_norm": 0.515660343163818,
+      "grad_norm": 0.5137067823598883,
       "learning_rate": 3.527906394156919e-05,
-      "loss": 0.5201,
+      "loss": 0.5198,
       "step": 360
     },
     {
       "epoch": 1.5065206051121542,
-      "grad_norm": 0.5295633939565575,
+      "grad_norm": 0.510532103606806,
       "learning_rate": 3.524128376017269e-05,
-      "loss": 0.5217,
+      "loss": 0.5211,
       "step": 361
     },
     {
       "epoch": 1.5106937923839332,
-      "grad_norm": 0.47045229463349447,
+      "grad_norm": 0.47658188190983075,
       "learning_rate": 3.520337341082879e-05,
-      "loss": 0.4978,
+      "loss": 0.4982,
       "step": 362
     },
     {
       "epoch": 1.5148669796557122,
-      "grad_norm": 0.4887503669000644,
+      "grad_norm": 0.5024470190108087,
       "learning_rate": 3.516533321731024e-05,
-      "loss": 0.5063,
+      "loss": 0.5057,
       "step": 363
     },
     {
       "epoch": 1.5190401669274909,
-      "grad_norm": 0.421690495721393,
+      "grad_norm": 0.44126823436761964,
       "learning_rate": 3.5127163504498716e-05,
-      "loss": 0.5185,
+      "loss": 0.5176,
       "step": 364
     },
     {
       "epoch": 1.5232133541992696,
-      "grad_norm": 0.5120003075936974,
+      "grad_norm": 0.5206128899119382,
       "learning_rate": 3.508886459838206e-05,
-      "loss": 0.5382,
+      "loss": 0.5384,
       "step": 365
     },
     {
       "epoch": 1.5273865414710484,
-      "grad_norm": 0.44354004700379424,
+      "grad_norm": 0.5026195170322312,
       "learning_rate": 3.5050436826051485e-05,
-      "loss": 0.5256,
+      "loss": 0.5254,
       "step": 366
     },
     {
       "epoch": 1.5315597287428273,
-      "grad_norm": 0.47684234644794166,
+      "grad_norm": 0.48311097755562116,
       "learning_rate": 3.5011880515698786e-05,
       "loss": 0.4806,
       "step": 367
     },
     {
       "epoch": 1.5357329160146063,
-      "grad_norm": 0.4618317062495365,
+      "grad_norm": 0.5145035326690318,
       "learning_rate": 3.497319599661353e-05,
-      "loss": 0.5103,
+      "loss": 0.51,
       "step": 368
     },
     {
       "epoch": 1.539906103286385,
-      "grad_norm": 0.45239105420623754,
+      "grad_norm": 0.45506247266857464,
       "learning_rate": 3.493438359918026e-05,
-      "loss": 0.4887,
+      "loss": 0.4884,
       "step": 369
     },
     {
       "epoch": 1.5440792905581637,
-      "grad_norm": 0.5027286807739099,
+      "grad_norm": 0.5028346606491918,
       "learning_rate": 3.489544365487564e-05,
-      "loss": 0.5012,
+      "loss": 0.5008,
       "step": 370
     },
     {
       "epoch": 1.5482524778299425,
-      "grad_norm": 0.46977134475828625,
+      "grad_norm": 0.49653645559800746,
       "learning_rate": 3.485637649626567e-05,
-      "loss": 0.4977,
+      "loss": 0.4973,
       "step": 371
     },
     {
       "epoch": 1.5524256651017214,
-      "grad_norm": 0.4559220194637468,
+      "grad_norm": 0.43390075333042416,
       "learning_rate": 3.481718245700281e-05,
-      "loss": 0.4979,
+      "loss": 0.4977,
       "step": 372
     },
     {
       "epoch": 1.5565988523735004,
-      "grad_norm": 0.45536579127452353,
+      "grad_norm": 0.48675041070351444,
       "learning_rate": 3.4777861871823146e-05,
-      "loss": 0.4931,
+      "loss": 0.4933,
       "step": 373
     },
     {
       "epoch": 1.560772039645279,
-      "grad_norm": 0.44262603255294486,
+      "grad_norm": 0.41615809738822934,
       "learning_rate": 3.473841507654352e-05,
-      "loss": 0.4988,
+      "loss": 0.4987,
       "step": 374
     },
     {
       "epoch": 1.5649452269170578,
-      "grad_norm": 0.42085933981012347,
+      "grad_norm": 0.46837027249533864,
       "learning_rate": 3.469884240805869e-05,
-      "loss": 0.5088,
+      "loss": 0.509,
       "step": 375
     },
     {
       "epoch": 1.5691184141888366,
-      "grad_norm": 0.4225066695185206,
+      "grad_norm": 0.412910566643372,
       "learning_rate": 3.4659144204338386e-05,
-      "loss": 0.5006,
+      "loss": 0.4994,
       "step": 376
     },
     {
       "epoch": 1.5732916014606155,
-      "grad_norm": 0.4600371750758285,
+      "grad_norm": 0.475137570865477,
       "learning_rate": 3.461932080442453e-05,
-      "loss": 0.5031,
+      "loss": 0.5028,
       "step": 377
     },
     {
       "epoch": 1.5774647887323945,
-      "grad_norm": 0.4189463749528179,
+      "grad_norm": 0.42361240611761275,
       "learning_rate": 3.457937254842823e-05,
-      "loss": 0.5027,
+      "loss": 0.5031,
       "step": 378
     },
     {
       "epoch": 1.5816379760041732,
-      "grad_norm": 0.48564837254282367,
+      "grad_norm": 0.48238224722046535,
       "learning_rate": 3.453929977752697e-05,
-      "loss": 0.4961,
+      "loss": 0.4964,
       "step": 379
     },
     {
       "epoch": 1.585811163275952,
-      "grad_norm": 0.4311507310780022,
+      "grad_norm": 0.43329507440692416,
       "learning_rate": 3.449910283396161e-05,
-      "loss": 0.4965,
+      "loss": 0.497,
       "step": 380
     },
     {
       "epoch": 1.5899843505477307,
-      "grad_norm": 0.5117031694976955,
+      "grad_norm": 0.47174861777579097,
       "learning_rate": 3.445878206103352e-05,
-      "loss": 0.5073,
+      "loss": 0.5063,
       "step": 381
     },
     {
       "epoch": 1.5941575378195096,
-      "grad_norm": 0.43807585098111723,
+      "grad_norm": 0.44914480390327555,
       "learning_rate": 3.4418337803101655e-05,
-      "loss": 0.5057,
+      "loss": 0.5059,
       "step": 382
     },
     {
       "epoch": 1.5983307250912886,
-      "grad_norm": 0.5062006985678845,
+      "grad_norm": 0.4548890253877132,
       "learning_rate": 3.4377770405579554e-05,
-      "loss": 0.4793,
+      "loss": 0.4791,
       "step": 383
     },
     {
       "epoch": 1.6025039123630673,
-      "grad_norm": 0.4911877522984626,
+      "grad_norm": 0.5198306544395302,
       "learning_rate": 3.433708021493245e-05,
-      "loss": 0.5211,
+      "loss": 0.5205,
       "step": 384
     },
     {
       "epoch": 1.606677099634846,
-      "grad_norm": 0.46196345465169836,
+      "grad_norm": 0.4419682196355817,
       "learning_rate": 3.429626757867429e-05,
-      "loss": 0.4916,
+      "loss": 0.49,
       "step": 385
     },
     {
       "epoch": 1.6108502869066248,
-      "grad_norm": 0.44658498769027294,
+      "grad_norm": 0.44909076183755275,
       "learning_rate": 3.425533284536476e-05,
-      "loss": 0.5224,
+      "loss": 0.5198,
       "step": 386
     },
     {
       "epoch": 1.6150234741784038,
-      "grad_norm": 0.4392590995851261,
+      "grad_norm": 0.44365211499530155,
       "learning_rate": 3.4214276364606304e-05,
-      "loss": 0.4889,
+      "loss": 0.4881,
       "step": 387
     },
     {
       "epoch": 1.6191966614501827,
-      "grad_norm": 0.463053339604633,
+      "grad_norm": 0.4670461861021911,
       "learning_rate": 3.4173098487041174e-05,
-      "loss": 0.5159,
+      "loss": 0.5162,
       "step": 388
     },
     {
       "epoch": 1.6233698487219614,
-      "grad_norm": 0.38293138696843315,
+      "grad_norm": 0.41286974121688547,
       "learning_rate": 3.413179956434838e-05,
-      "loss": 0.4895,
+      "loss": 0.4903,
       "step": 389
     },
     {
       "epoch": 1.6275430359937402,
-      "grad_norm": 0.44358110294259073,
+      "grad_norm": 0.416108439397374,
       "learning_rate": 3.409037994924074e-05,
-      "loss": 0.5164,
+      "loss": 0.5161,
       "step": 390
     },
     {
       "epoch": 1.631716223265519,
-      "grad_norm": 0.4121979603171428,
+      "grad_norm": 0.462927748207203,
       "learning_rate": 3.404883999546183e-05,
-      "loss": 0.5282,
+      "loss": 0.5284,
       "step": 391
     },
     {
       "epoch": 1.6358894105372979,
-      "grad_norm": 0.4938661433512818,
+      "grad_norm": 0.4255055411182278,
       "learning_rate": 3.400718005778298e-05,
-      "loss": 0.4859,
+      "loss": 0.4855,
       "step": 392
     },
     {
       "epoch": 1.6400625978090768,
-      "grad_norm": 0.47984222384095987,
+      "grad_norm": 0.5192843816914167,
       "learning_rate": 3.3965400492000245e-05,
-      "loss": 0.5101,
+      "loss": 0.5107,
       "step": 393
     },
     {
       "epoch": 1.6442357850808555,
-      "grad_norm": 0.46095126445801976,
+      "grad_norm": 0.45428058991671794,
       "learning_rate": 3.392350165493136e-05,
       "loss": 0.5066,
       "step": 394
     },
     {
       "epoch": 1.6484089723526343,
-      "grad_norm": 0.4946940183717776,
+      "grad_norm": 0.48143298976734683,
       "learning_rate": 3.3881483904412685e-05,
-      "loss": 0.4949,
+      "loss": 0.4947,
       "step": 395
     },
     {
       "epoch": 1.652582159624413,
-      "grad_norm": 0.49119612096966564,
+      "grad_norm": 0.5104980232957163,
       "learning_rate": 3.383934759929616e-05,
-      "loss": 0.4979,
+      "loss": 0.4983,
       "step": 396
     },
     {
       "epoch": 1.656755346896192,
-      "grad_norm": 0.4567221757329101,
+      "grad_norm": 0.4442481977908017,
       "learning_rate": 3.3797093099446265e-05,
       "loss": 0.5174,
       "step": 397
     },
     {
       "epoch": 1.660928534167971,
-      "grad_norm": 0.4942486086613881,
+      "grad_norm": 0.4851760298508442,
       "learning_rate": 3.3754720765736885e-05,
-      "loss": 0.5155,
+      "loss": 0.5151,
       "step": 398
     },
     {
       "epoch": 1.6651017214397497,
-      "grad_norm": 0.49326653280337934,
+      "grad_norm": 0.5069685370264763,
       "learning_rate": 3.371223096004829e-05,
-      "loss": 0.5272,
+      "loss": 0.5277,
       "step": 399
     },
     {
       "epoch": 1.6692749087115284,
-      "grad_norm": 0.5069246037549443,
+      "grad_norm": 0.47458542560166106,
       "learning_rate": 3.3669624045264e-05,
       "loss": 0.5113,
       "step": 400
     },
     {
       "epoch": 1.6734480959833071,
-      "grad_norm": 0.45577928256350075,
+      "grad_norm": 0.5025562139998816,
       "learning_rate": 3.362690038526769e-05,
-      "loss": 0.49,
+      "loss": 0.4905,
       "step": 401
     },
     {
       "epoch": 1.677621283255086,
-      "grad_norm": 0.5001627960824105,
+      "grad_norm": 0.492782412001696,
       "learning_rate": 3.358406034494015e-05,
-      "loss": 0.5099,
+      "loss": 0.5102,
       "step": 402
     },
     {
       "epoch": 1.681794470526865,
-      "grad_norm": 0.41575098293202106,
+      "grad_norm": 0.45150765331910836,
       "learning_rate": 3.3541104290156056e-05,
       "loss": 0.5034,
       "step": 403
     },
     {
       "epoch": 1.6859676577986438,
-      "grad_norm": 0.4614801059415867,
+      "grad_norm": 0.46646815759924365,
       "learning_rate": 3.349803258778094e-05,
-      "loss": 0.4779,
+      "loss": 0.4782,
       "step": 404
     },
     {
       "epoch": 1.6901408450704225,
-      "grad_norm": 0.4363319574937817,
+      "grad_norm": 0.4226125155735945,
       "learning_rate": 3.3454845605668e-05,
-      "loss": 0.4931,
+      "loss": 0.4927,
       "step": 405
     },
     {
       "epoch": 1.6943140323422012,
-      "grad_norm": 0.46340008754079925,
+      "grad_norm": 0.45707608626457547,
       "learning_rate": 3.341154371265499e-05,
-      "loss": 0.4831,
+      "loss": 0.4834,
       "step": 406
     },
     {
       "epoch": 1.6984872196139802,
-      "grad_norm": 0.46516068223324,
+      "grad_norm": 0.4353007530215309,
       "learning_rate": 3.336812727856105e-05,
-      "loss": 0.4937,
+      "loss": 0.4934,
       "step": 407
     },
     {
       "epoch": 1.7026604068857591,
-      "grad_norm": 0.3844052948257072,
+      "grad_norm": 0.3843522541683085,
       "learning_rate": 3.332459667418357e-05,
-      "loss": 0.4939,
+      "loss": 0.4937,
       "step": 408
     },
     {
       "epoch": 1.7068335941575379,
-      "grad_norm": 0.48669465186112,
+      "grad_norm": 0.4568710619416651,
       "learning_rate": 3.3280952271294994e-05,
-      "loss": 0.5113,
+      "loss": 0.5114,
       "step": 409
     },
     {
       "epoch": 1.7110067814293166,
-      "grad_norm": 0.40816518217741615,
+      "grad_norm": 0.40154497525014626,
       "learning_rate": 3.323719444263967e-05,
-      "loss": 0.4835,
+      "loss": 0.4829,
       "step": 410
     },
     {
       "epoch": 1.7151799687010953,
-      "grad_norm": 0.4869054399140217,
+      "grad_norm": 0.44923072041599343,
       "learning_rate": 3.3193323561930645e-05,
-      "loss": 0.5071,
+      "loss": 0.5077,
       "step": 411
     },
     {
       "epoch": 1.7193531559728743,
-      "grad_norm": 0.43075292880740484,
+      "grad_norm": 0.40955601635996286,
       "learning_rate": 3.314934000384652e-05,
-      "loss": 0.5126,
+      "loss": 0.511,
       "step": 412
     },
     {
       "epoch": 1.7235263432446533,
-      "grad_norm": 0.41963972666647403,
+      "grad_norm": 0.4020894417399454,
       "learning_rate": 3.3105244144028164e-05,
       "loss": 0.5167,
       "step": 413
     },
     {
       "epoch": 1.727699530516432,
-      "grad_norm": 0.45536765727662243,
+      "grad_norm": 0.44662013999238853,
       "learning_rate": 3.30610363590756e-05,
-      "loss": 0.4881,
+      "loss": 0.4885,
       "step": 414
     },
     {
       "epoch": 1.7318727177882107,
-      "grad_norm": 0.39215945697801347,
+      "grad_norm": 0.39389830274457294,
       "learning_rate": 3.301671702654472e-05,
-      "loss": 0.4814,
+      "loss": 0.4812,
       "step": 415
     },
     {
       "epoch": 1.7360459050599895,
-      "grad_norm": 0.4553866396380636,
+      "grad_norm": 0.46000002660300476,
       "learning_rate": 3.297228652494409e-05,
-      "loss": 0.4826,
+      "loss": 0.4828,
       "step": 416
     },
     {
       "epoch": 1.7402190923317684,
-      "grad_norm": 0.46089662696883743,
+      "grad_norm": 0.47141731730171293,
       "learning_rate": 3.292774523373174e-05,
       "loss": 0.5054,
       "step": 417
     },
     {
       "epoch": 1.7443922796035474,
-      "grad_norm": 0.431258445924461,
+      "grad_norm": 0.461597282179614,
       "learning_rate": 3.288309353331186e-05,
       "loss": 0.5146,
       "step": 418
     },
     {
       "epoch": 1.748565466875326,
-      "grad_norm": 0.4367972352244524,
+      "grad_norm": 0.450149196238661,
       "learning_rate": 3.2838331805031604e-05,
-      "loss": 0.5028,
+      "loss": 0.5031,
       "step": 419
     },
     {
       "epoch": 1.7527386541471048,
-      "grad_norm": 0.38109329746957604,
+      "grad_norm": 0.39623019359370976,
       "learning_rate": 3.2793460431177827e-05,
-      "loss": 0.5041,
+      "loss": 0.5042,
       "step": 420
     },
     {
       "epoch": 1.7569118414188836,
-      "grad_norm": 0.44724239113407105,
+      "grad_norm": 0.45125250499654707,
       "learning_rate": 3.27484797949738e-05,
-      "loss": 0.4977,
+      "loss": 0.4979,
       "step": 421
     },
     {
       "epoch": 1.7610850286906625,
-      "grad_norm": 0.4367742160362723,
+      "grad_norm": 0.44338552743130183,
       "learning_rate": 3.270339028057594e-05,
-      "loss": 0.5056,
+      "loss": 0.5046,
       "step": 422
     },
     {
       "epoch": 1.7652582159624415,
-      "grad_norm": 0.4595617143186662,
+      "grad_norm": 0.46110188469569746,
       "learning_rate": 3.265819227307055e-05,
-      "loss": 0.507,
+      "loss": 0.5074,
       "step": 423
     },
     {
       "epoch": 1.7694314032342202,
-      "grad_norm": 0.46761722437264935,
+      "grad_norm": 0.47208170389761567,
       "learning_rate": 3.261288615847051e-05,
-      "loss": 0.4997,
+      "loss": 0.4992,
       "step": 424
     },
     {
       "epoch": 1.773604590505999,
-      "grad_norm": 0.4392976123280654,
+      "grad_norm": 0.4556245558923723,
       "learning_rate": 3.256747232371199e-05,
       "loss": 0.507,
       "step": 425
     },
     {
       "epoch": 1.7777777777777777,
-      "grad_norm": 0.42338027198064787,
+      "grad_norm": 0.4202357634070867,
       "learning_rate": 3.252195115665114e-05,
-      "loss": 0.5078,
+      "loss": 0.5068,
       "step": 426
     },
     {
       "epoch": 1.7819509650495566,
-      "grad_norm": 0.42764998590244296,
+      "grad_norm": 0.4561401852357617,
       "learning_rate": 3.247632304606079e-05,
-      "loss": 0.5119,
+      "loss": 0.5127,
       "step": 427
     },
     {
       "epoch": 1.7861241523213354,
-      "grad_norm": 0.4504558151540797,
+      "grad_norm": 0.42281343600978155,
       "learning_rate": 3.243058838162709e-05,
-      "loss": 0.506,
+      "loss": 0.5057,
       "step": 428
     },
     {
       "epoch": 1.7902973395931143,
-      "grad_norm": 0.47460162714166765,
+      "grad_norm": 0.47078581006024345,
       "learning_rate": 3.238474755394624e-05,
-      "loss": 0.509,
+      "loss": 0.5087,
       "step": 429
     },
     {
       "epoch": 1.794470526864893,
-      "grad_norm": 0.41949398225742485,
+      "grad_norm": 0.4174337801781674,
       "learning_rate": 3.233880095452113e-05,
-      "loss": 0.4929,
+      "loss": 0.4922,
       "step": 430
     },
     {
       "epoch": 1.7986437141366718,
-      "grad_norm": 0.42545409048156363,
+      "grad_norm": 0.4170215324063569,
       "learning_rate": 3.2292748975757975e-05,
-      "loss": 0.5051,
+      "loss": 0.5053,
       "step": 431
     },
     {
       "epoch": 1.8028169014084507,
-      "grad_norm": 0.45900370620370873,
+      "grad_norm": 0.44944181968653885,
       "learning_rate": 3.224659201096297e-05,
-      "loss": 0.5118,
+      "loss": 0.5108,
       "step": 432
     },
     {
       "epoch": 1.8069900886802295,
-      "grad_norm": 0.4363775055560789,
+      "grad_norm": 0.4116581090278689,
       "learning_rate": 3.220033045433898e-05,
-      "loss": 0.5,
+      "loss": 0.498,
       "step": 433
     },
     {
       "epoch": 1.8111632759520084,
-      "grad_norm": 0.4330498219060868,
+      "grad_norm": 0.43786020926475383,
       "learning_rate": 3.215396470098209e-05,
-      "loss": 0.5006,
+      "loss": 0.5011,
       "step": 434
     },
     {
       "epoch": 1.8153364632237872,
-      "grad_norm": 0.4022446450753603,
+      "grad_norm": 0.3869892009988906,
       "learning_rate": 3.2107495146878295e-05,
-      "loss": 0.5064,
+      "loss": 0.5058,
       "step": 435
     },
     {
       "epoch": 1.819509650495566,
-      "grad_norm": 0.43793996285041614,
+      "grad_norm": 0.4365905690488894,
       "learning_rate": 3.206092218890012e-05,
-      "loss": 0.5077,
+      "loss": 0.5078,
       "step": 436
     },
     {
       "epoch": 1.8236828377673449,
-      "grad_norm": 0.4085715464347046,
+      "grad_norm": 0.4063422757869833,
       "learning_rate": 3.201424622480318e-05,
-      "loss": 0.4882,
+      "loss": 0.4884,
       "step": 437
     },
     {
       "epoch": 1.8278560250391236,
-      "grad_norm": 0.44611605384794584,
+      "grad_norm": 0.4699537232834309,
       "learning_rate": 3.196746765322282e-05,
-      "loss": 0.5064,
+      "loss": 0.5068,
       "step": 438
     },
     {
       "epoch": 1.8320292123109025,
-      "grad_norm": 0.40261675938214003,
+      "grad_norm": 0.4056105701618899,
       "learning_rate": 3.1920586873670706e-05,
-      "loss": 0.5111,
+      "loss": 0.511,
       "step": 439
     },
     {
       "epoch": 1.8362023995826813,
-      "grad_norm": 0.5496572935759929,
+      "grad_norm": 0.5171472709606156,
       "learning_rate": 3.1873604286531415e-05,
-      "loss": 0.4983,
+      "loss": 0.498,
       "step": 440
     },
     {
       "epoch": 1.84037558685446,
-      "grad_norm": 0.4991041341822398,
+      "grad_norm": 0.5388935087728408,
       "learning_rate": 3.1826520293058986e-05,
       "loss": 0.5146,
       "step": 441
     },
     {
       "epoch": 1.844548774126239,
-      "grad_norm": 0.4171985737645811,
+      "grad_norm": 0.4238036226820878,
       "learning_rate": 3.177933529537356e-05,
-      "loss": 0.499,
+      "loss": 0.4989,
       "step": 442
     },
     {
       "epoch": 1.8487219613980177,
-      "grad_norm": 0.447102318137122,
+      "grad_norm": 0.4777227616751113,
       "learning_rate": 3.173204969645785e-05,
-      "loss": 0.5133,
+      "loss": 0.5137,
       "step": 443
     },
     {
       "epoch": 1.8528951486697967,
-      "grad_norm": 0.41127372540895973,
+      "grad_norm": 0.4120885871205803,
       "learning_rate": 3.16846639001538e-05,
-      "loss": 0.5019,
+      "loss": 0.5016,
       "step": 444
     },
     {
       "epoch": 1.8570683359415754,
-      "grad_norm": 0.5512832998253044,
+      "grad_norm": 0.6079582069893243,
       "learning_rate": 3.163717831115906e-05,
-      "loss": 0.513,
+      "loss": 0.5132,
       "step": 445
     },
     {
       "epoch": 1.8612415232133541,
-      "grad_norm": 0.3923543219383909,
+      "grad_norm": 0.3965637928526205,
       "learning_rate": 3.158959333502357e-05,
-      "loss": 0.5128,
+      "loss": 0.5132,
       "step": 446
     },
     {
       "epoch": 1.8654147104851329,
-      "grad_norm": 0.48958059242257523,
+      "grad_norm": 0.5011743926540937,
       "learning_rate": 3.1541909378146063e-05,
-      "loss": 0.4986,
+      "loss": 0.4985,
       "step": 447
     },
     {
       "epoch": 1.8695878977569118,
-      "grad_norm": 0.4039835426136765,
+      "grad_norm": 0.41432543627681434,
       "learning_rate": 3.149412684777066e-05,
-      "loss": 0.5085,
+      "loss": 0.5088,
       "step": 448
     },
     {
       "epoch": 1.8737610850286908,
-      "grad_norm": 0.49159788706609686,
+      "grad_norm": 0.4710360008955483,
       "learning_rate": 3.14462461519833e-05,
-      "loss": 0.4945,
+      "loss": 0.4942,
       "step": 449
     },
     {
       "epoch": 1.8779342723004695,
-      "grad_norm": 0.4618373208760923,
+      "grad_norm": 0.4676985936809986,
       "learning_rate": 3.139826769970833e-05,
-      "loss": 0.494,
+      "loss": 0.4941,
       "step": 450
     },
     {
       "epoch": 1.8821074595722482,
-      "grad_norm": 0.4616410292797367,
+      "grad_norm": 0.4494907641290245,
       "learning_rate": 3.135019190070497e-05,
       "loss": 0.4923,
       "step": 451
     },
     {
       "epoch": 1.886280646844027,
-      "grad_norm": 0.46168212373562295,
+      "grad_norm": 0.4794209032623939,
       "learning_rate": 3.130201916556385e-05,
-      "loss": 0.4859,
+      "loss": 0.4865,
       "step": 452
     },
     {
       "epoch": 1.890453834115806,
-      "grad_norm": 0.4631733269583256,
+      "grad_norm": 0.4369175007530477,
       "learning_rate": 3.125374990570345e-05,
-      "loss": 0.4956,
+      "loss": 0.4957,
       "step": 453
     },
     {
       "epoch": 1.8946270213875849,
-      "grad_norm": 0.4383941912299457,
+      "grad_norm": 0.44752564186561883,
       "learning_rate": 3.120538453336665e-05,
-      "loss": 0.4942,
+      "loss": 0.4943,
       "step": 454
     },
     {
       "epoch": 1.8988002086593636,
-      "grad_norm": 0.4233225699914636,
+      "grad_norm": 0.41922701963838127,
       "learning_rate": 3.115692346161715e-05,
-      "loss": 0.4804,
+      "loss": 0.4807,
       "step": 455
     },
     {
       "epoch": 1.9029733959311423,
-      "grad_norm": 0.4281253202703246,
+      "grad_norm": 0.40648431489042314,
       "learning_rate": 3.1108367104335985e-05,
-      "loss": 0.5037,
+      "loss": 0.5039,
       "step": 456
     },
     {
       "epoch": 1.907146583202921,
-      "grad_norm": 0.44824327367909517,
+      "grad_norm": 0.44205281569087695,
       "learning_rate": 3.1059715876217973e-05,
-      "loss": 0.4899,
+      "loss": 0.4898,
       "step": 457
     },
     {
       "epoch": 1.9113197704747,
-      "grad_norm": 0.45154039339151814,
+      "grad_norm": 0.43864826957460284,
       "learning_rate": 3.101097019276817e-05,
-      "loss": 0.519,
+      "loss": 0.5192,
       "step": 458
     },
     {
       "epoch": 1.915492957746479,
-      "grad_norm": 0.3841861180448192,
+      "grad_norm": 0.4103611483852858,
       "learning_rate": 3.096213047029832e-05,
-      "loss": 0.4843,
+      "loss": 0.4847,
       "step": 459
     },
     {
       "epoch": 1.9196661450182577,
-      "grad_norm": 0.39950761347898894,
+      "grad_norm": 0.3866867669546823,
       "learning_rate": 3.091319712592333e-05,
-      "loss": 0.4912,
+      "loss": 0.4913,
       "step": 460
     },
     {
       "epoch": 1.9238393322900365,
-      "grad_norm": 0.4068470504022182,
+      "grad_norm": 0.4006686490173406,
       "learning_rate": 3.086417057755767e-05,
-      "loss": 0.476,
+      "loss": 0.4753,
       "step": 461
     },
     {
       "epoch": 1.9280125195618152,
-      "grad_norm": 0.49871228026093406,
+      "grad_norm": 0.5002367535479044,
       "learning_rate": 3.08150512439118e-05,
-      "loss": 0.4979,
+      "loss": 0.4974,
       "step": 462
     },
     {
       "epoch": 1.9321857068335941,
-      "grad_norm": 0.37618503650184054,
+      "grad_norm": 0.3802246781721408,
       "learning_rate": 3.076583954448864e-05,
       "loss": 0.5036,
       "step": 463
     },
     {
       "epoch": 1.936358894105373,
-      "grad_norm": 0.4710857723738476,
+      "grad_norm": 0.462957616765017,
       "learning_rate": 3.0716535899579936e-05,
       "loss": 0.4803,
       "step": 464
     },
     {
       "epoch": 1.9405320813771518,
-      "grad_norm": 0.41758306661480465,
+      "grad_norm": 0.42625583053025057,
       "learning_rate": 3.0667140730262706e-05,
-      "loss": 0.5179,
+      "loss": 0.5183,
       "step": 465
     },
     {
       "epoch": 1.9447052686489306,
-      "grad_norm": 0.44639583078459444,
+      "grad_norm": 0.4350240211215548,
       "learning_rate": 3.061765445839562e-05,
       "loss": 0.4884,
       "step": 466
     },
     {
       "epoch": 1.9488784559207093,
-      "grad_norm": 0.3856054506910291,
+      "grad_norm": 0.38939906653393147,
       "learning_rate": 3.0568077506615434e-05,
-      "loss": 0.4884,
+      "loss": 0.4885,
       "step": 467
     },
     {
       "epoch": 1.9530516431924883,
-      "grad_norm": 0.5015588444838592,
+      "grad_norm": 0.4975659604551595,
       "learning_rate": 3.051841029833332e-05,
-      "loss": 0.4949,
+      "loss": 0.495,
       "step": 468
     },
     {
       "epoch": 1.9572248304642672,
-      "grad_norm": 0.42059771791673217,
+      "grad_norm": 0.39746912524612193,
       "learning_rate": 3.04686532577313e-05,
-      "loss": 0.5,
+      "loss": 0.4998,
       "step": 469
     },
     {
       "epoch": 1.961398017736046,
-      "grad_norm": 0.49384177244218413,
+      "grad_norm": 0.48723167444947857,
       "learning_rate": 3.041880680975861e-05,
-      "loss": 0.4771,
+      "loss": 0.4772,
       "step": 470
     },
     {
       "epoch": 1.9655712050078247,
-      "grad_norm": 0.4512477765162067,
+      "grad_norm": 0.4365957487427277,
       "learning_rate": 3.0368871380128065e-05,
-      "loss": 0.4797,
+      "loss": 0.4782,
       "step": 471
     },
     {
       "epoch": 1.9697443922796034,
-      "grad_norm": 0.4843108870253368,
+      "grad_norm": 0.48364077045930803,
       "learning_rate": 3.0318847395312425e-05,
-      "loss": 0.4939,
+      "loss": 0.4942,
       "step": 472
     },
     {
       "epoch": 1.9739175795513824,
-      "grad_norm": 0.48799517316203755,
+      "grad_norm": 0.48271326671373127,
       "learning_rate": 3.0268735282540755e-05,
-      "loss": 0.4895,
+      "loss": 0.4894,
       "step": 473
     },
     {
       "epoch": 1.9780907668231613,
-      "grad_norm": 0.4360622175045534,
+      "grad_norm": 0.43764984275827873,
       "learning_rate": 3.021853546979478e-05,
       "loss": 0.4825,
       "step": 474
     },
     {
       "epoch": 1.98226395409494,
-      "grad_norm": 0.48477470376144355,
+      "grad_norm": 0.4848405600313535,
       "learning_rate": 3.0168248385805223e-05,
-      "loss": 0.4896,
+      "loss": 0.4895,
       "step": 475
     },
     {
       "epoch": 1.9864371413667188,
-      "grad_norm": 0.4649672928125755,
+      "grad_norm": 0.45058546530982824,
       "learning_rate": 3.0117874460048138e-05,
       "loss": 0.4828,
       "step": 476
     },
     {
       "epoch": 1.9906103286384975,
-      "grad_norm": 0.45671327298587755,
+      "grad_norm": 0.4506018534935378,
       "learning_rate": 3.0067414122741262e-05,
-      "loss": 0.4772,
+      "loss": 0.477,
       "step": 477
     },
     {
       "epoch": 1.9947835159102765,
-      "grad_norm": 0.4127488395959894,
+      "grad_norm": 0.41548055740423906,
       "learning_rate": 3.001686780484032e-05,
-      "loss": 0.4863,
+      "loss": 0.4867,
       "step": 478
     },
     {
       "epoch": 1.9989567031820554,
-      "grad_norm": 0.5247593821950686,
+      "grad_norm": 0.4995327358747625,
       "learning_rate": 2.9966235938035352e-05,
-      "loss": 0.4974,
+      "loss": 0.4973,
       "step": 479
     },
     {
       "epoch": 2.003129890453834,
-      "grad_norm": 0.5588714016568661,
+      "grad_norm": 0.5606594226988486,
       "learning_rate": 2.9915518954747038e-05,
-      "loss": 0.4053,
+      "loss": 0.4055,
       "step": 480
     },
     {
       "epoch": 2.007303077725613,
-      "grad_norm": 0.5454129881237091,
+      "grad_norm": 0.5237793998746713,
       "learning_rate": 2.9864717288122986e-05,
-      "loss": 0.3778,
+      "loss": 0.3777,
       "step": 481
     },
     {
       "epoch": 2.0114762649973916,
-      "grad_norm": 1.3539661441744257,
+      "grad_norm": 1.3911608378315576,
       "learning_rate": 2.9813831372034035e-05,
-      "loss": 0.3869,
+      "loss": 0.3875,
       "step": 482
     },
     {
       "epoch": 2.0156494522691704,
-      "grad_norm": 0.6231417829150485,
+      "grad_norm": 0.6014286947463934,
       "learning_rate": 2.976286164107057e-05,
-      "loss": 0.385,
+      "loss": 0.3846,
       "step": 483
     },
     {
       "epoch": 2.0198226395409495,
-      "grad_norm": 0.6041518516214323,
+      "grad_norm": 0.6113258270016576,
       "learning_rate": 2.9711808530538776e-05,
-      "loss": 0.3656,
+      "loss": 0.3661,
       "step": 484
     },
     {
       "epoch": 2.0239958268127283,
-      "grad_norm": 0.5425718430316419,
+      "grad_norm": 0.5331972801376571,
       "learning_rate": 2.966067247645696e-05,
-      "loss": 0.3687,
+      "loss": 0.3692,
       "step": 485
     },
     {
       "epoch": 2.028169014084507,
-      "grad_norm": 0.4682329194251342,
+      "grad_norm": 0.4768628880688289,
       "learning_rate": 2.9609453915551793e-05,
-      "loss": 0.3754,
+      "loss": 0.3756,
       "step": 486
     },
     {
       "epoch": 2.0323422013562857,
-      "grad_norm": 0.5385772464516133,
+      "grad_norm": 0.5401195891617366,
       "learning_rate": 2.9558153285254607e-05,
-      "loss": 0.3657,
+      "loss": 0.3662,
       "step": 487
     },
     {
       "epoch": 2.0365153886280645,
-      "grad_norm": 0.5275345840456208,
+      "grad_norm": 0.5308147187486485,
       "learning_rate": 2.9506771023697628e-05,
-      "loss": 0.3622,
+      "loss": 0.3626,
       "step": 488
     },
     {
       "epoch": 2.0406885758998436,
-      "grad_norm": 0.5181239161253565,
+      "grad_norm": 0.5447737341099776,
       "learning_rate": 2.9455307569710272e-05,
-      "loss": 0.3625,
+      "loss": 0.3631,
       "step": 489
     },
     {
       "epoch": 2.0448617631716224,
-      "grad_norm": 0.474365077751043,
+      "grad_norm": 0.4704644131707518,
       "learning_rate": 2.940376336281537e-05,
-      "loss": 0.3489,
+      "loss": 0.3486,
       "step": 490
     },
     {
       "epoch": 2.049034950443401,
-      "grad_norm": 0.6043161391629818,
+      "grad_norm": 0.6341513023869643,
       "learning_rate": 2.9352138843225413e-05,
-      "loss": 0.3523,
+      "loss": 0.3528,
       "step": 491
     },
     {
       "epoch": 2.05320813771518,
-      "grad_norm": 0.42930368997175306,
+      "grad_norm": 0.43156651827663284,
       "learning_rate": 2.9300434451838814e-05,
-      "loss": 0.3752,
+      "loss": 0.3759,
       "step": 492
     },
     {
       "epoch": 2.0573813249869586,
-      "grad_norm": 0.5213666535261181,
+      "grad_norm": 0.5431027042648565,
       "learning_rate": 2.9248650630236125e-05,
-      "loss": 0.3714,
+      "loss": 0.3719,
       "step": 493
     },
     {
       "epoch": 2.0615545122587378,
-      "grad_norm": 0.41078584081484,
+      "grad_norm": 0.4098808090675706,
       "learning_rate": 2.9196787820676273e-05,
-      "loss": 0.3638,
+      "loss": 0.3635,
       "step": 494
     },
     {
       "epoch": 2.0657276995305165,
-      "grad_norm": 0.49099297969874117,
+      "grad_norm": 0.4930317064445512,
       "learning_rate": 2.9144846466092773e-05,
-      "loss": 0.3665,
+      "loss": 0.3664,
       "step": 495
     },
     {
       "epoch": 2.0699008868022952,
-      "grad_norm": 0.4041585382056963,
+      "grad_norm": 0.4028047612180292,
       "learning_rate": 2.9092827010089966e-05,
-      "loss": 0.3503,
+      "loss": 0.3504,
       "step": 496
     },
     {
       "epoch": 2.074074074074074,
-      "grad_norm": 0.4454298915393295,
+      "grad_norm": 0.4492845020278998,
       "learning_rate": 2.9040729896939194e-05,
-      "loss": 0.3832,
+      "loss": 0.3831,
       "step": 497
     },
     {
       "epoch": 2.0782472613458527,
-      "grad_norm": 0.4340270090457894,
+      "grad_norm": 0.4088207403348072,
       "learning_rate": 2.898855557157506e-05,
-      "loss": 0.3563,
+      "loss": 0.3564,
       "step": 498
     },
     {
       "epoch": 2.082420448617632,
-      "grad_norm": 0.42309215838182007,
+      "grad_norm": 0.42571137052989716,
       "learning_rate": 2.8936304479591566e-05,
-      "loss": 0.3698,
+      "loss": 0.3699,
       "step": 499
     },
     {
       "epoch": 2.0865936358894106,
-      "grad_norm": 0.4723402125237257,
+      "grad_norm": 0.4488120127051875,
       "learning_rate": 2.8883977067238363e-05,
-      "loss": 0.3622,
+      "loss": 0.3619,
       "step": 500
     },
     {
       "epoch": 2.0907668231611893,
-      "grad_norm": 0.3652560586867096,
+      "grad_norm": 0.36892320530463035,
       "learning_rate": 2.8831573781416896e-05,
-      "loss": 0.3497,
+      "loss": 0.3499,
       "step": 501
     },
     {
       "epoch": 2.094940010432968,
-      "grad_norm": 0.49420026262590266,
+      "grad_norm": 0.47303904107617795,
       "learning_rate": 2.8779095069676622e-05,
-      "loss": 0.3785,
+      "loss": 0.3781,
       "step": 502
     },
     {
       "epoch": 2.099113197704747,
-      "grad_norm": 0.41481493110161133,
+      "grad_norm": 0.41094335570685786,
       "learning_rate": 2.8726541380211154e-05,
-      "loss": 0.3575,
+      "loss": 0.3577,
       "step": 503
     },
     {
       "epoch": 2.103286384976526,
-      "grad_norm": 0.401140142568712,
+      "grad_norm": 0.3955800028630068,
       "learning_rate": 2.867391316185447e-05,
-      "loss": 0.3415,
+      "loss": 0.3412,
       "step": 504
     },
     {
       "epoch": 2.1074595722483047,
-      "grad_norm": 0.3907499313082441,
+      "grad_norm": 0.38697071137276695,
       "learning_rate": 2.8621210864077053e-05,
-      "loss": 0.3505,
+      "loss": 0.3504,
       "step": 505
     },
     {
       "epoch": 2.1116327595200834,
-      "grad_norm": 0.403856703419521,
+      "grad_norm": 0.39860793437101216,
       "learning_rate": 2.8568434936982048e-05,
       "loss": 0.3482,
       "step": 506
     },
     {
       "epoch": 2.115805946791862,
-      "grad_norm": 0.41059670498163203,
+      "grad_norm": 0.4147073142723547,
       "learning_rate": 2.8515585831301456e-05,
       "loss": 0.3572,
       "step": 507
     },
     {
       "epoch": 2.119979134063641,
-      "grad_norm": 0.40324787461661543,
+      "grad_norm": 0.40042096552228806,
       "learning_rate": 2.8462663998392232e-05,
-      "loss": 0.3544,
+      "loss": 0.3546,
       "step": 508
     },
     {
       "epoch": 2.12415232133542,
-      "grad_norm": 0.41164859412594773,
+      "grad_norm": 0.4226011802521663,
       "learning_rate": 2.840966989023248e-05,
-      "loss": 0.3584,
+      "loss": 0.3587,
       "step": 509
     },
     {
       "epoch": 2.128325508607199,
-      "grad_norm": 0.3608310259465599,
+      "grad_norm": 0.3592969802443122,
       "learning_rate": 2.835660395941754e-05,
       "loss": 0.3809,
       "step": 510
     },
     {
       "epoch": 2.1324986958789776,
-      "grad_norm": 0.3791156362522083,
+      "grad_norm": 0.391324858922351,
       "learning_rate": 2.830346665915619e-05,
-      "loss": 0.3648,
+      "loss": 0.3651,
       "step": 511
     },
     {
       "epoch": 2.1366718831507563,
-      "grad_norm": 0.39655972896033287,
+      "grad_norm": 0.385400946247812,
       "learning_rate": 2.8250258443266712e-05,
-      "loss": 0.3553,
+      "loss": 0.3551,
       "step": 512
     },
     {
       "epoch": 2.140845070422535,
-      "grad_norm": 0.38547206654350263,
+      "grad_norm": 0.39474567022057705,
       "learning_rate": 2.8196979766173034e-05,
-      "loss": 0.3523,
+      "loss": 0.3522,
       "step": 513
     },
     {
       "epoch": 2.145018257694314,
-      "grad_norm": 0.4526507685476474,
+      "grad_norm": 0.4425713730752484,
       "learning_rate": 2.8143631082900894e-05,
-      "loss": 0.3722,
+      "loss": 0.3717,
       "step": 514
     },
     {
       "epoch": 2.149191444966093,
-      "grad_norm": 0.3762054780689094,
+      "grad_norm": 0.3917681973920762,
       "learning_rate": 2.8090212849073877e-05,
-      "loss": 0.3364,
+      "loss": 0.3369,
       "step": 515
     },
     {
       "epoch": 2.1533646322378717,
-      "grad_norm": 0.403082361426093,
+      "grad_norm": 0.41600562658232837,
       "learning_rate": 2.8036725520909586e-05,
-      "loss": 0.3722,
+      "loss": 0.3725,
       "step": 516
     },
     {
       "epoch": 2.1575378195096504,
-      "grad_norm": 0.38550087118205506,
+      "grad_norm": 0.38527510415860045,
       "learning_rate": 2.7983169555215716e-05,
-      "loss": 0.3664,
+      "loss": 0.3665,
       "step": 517
     },
     {
       "epoch": 2.161711006781429,
-      "grad_norm": 0.37766586544158887,
+      "grad_norm": 0.38351298108839976,
       "learning_rate": 2.7929545409386162e-05,
-      "loss": 0.351,
+      "loss": 0.3509,
       "step": 518
     },
     {
       "epoch": 2.1658841940532083,
-      "grad_norm": 0.4063816430486986,
+      "grad_norm": 0.41874738687020485,
       "learning_rate": 2.7875853541397107e-05,
-      "loss": 0.3724,
+      "loss": 0.3727,
       "step": 519
     },
     {
       "epoch": 2.170057381324987,
-      "grad_norm": 0.3860997250837504,
+      "grad_norm": 0.3829980993405463,
       "learning_rate": 2.782209440980312e-05,
-      "loss": 0.3513,
+      "loss": 0.3512,
       "step": 520
     },
     {
       "epoch": 2.174230568596766,
-      "grad_norm": 0.3868019278611356,
+      "grad_norm": 0.39343928041329834,
       "learning_rate": 2.776826847373323e-05,
-      "loss": 0.3461,
+      "loss": 0.3463,
       "step": 521
     },
     {
       "epoch": 2.1784037558685445,
-      "grad_norm": 0.3718911389460342,
+      "grad_norm": 0.37754679266127267,
       "learning_rate": 2.7714376192887e-05,
-      "loss": 0.3586,
+      "loss": 0.3591,
       "step": 522
     },
     {
       "epoch": 2.1825769431403232,
-      "grad_norm": 0.4306967365927621,
+      "grad_norm": 0.4227978588853828,
       "learning_rate": 2.7660418027530634e-05,
-      "loss": 0.3526,
+      "loss": 0.3517,
       "step": 523
     },
     {
       "epoch": 2.1867501304121024,
-      "grad_norm": 0.3841833149052425,
+      "grad_norm": 0.38394715378490435,
       "learning_rate": 2.7606394438492986e-05,
-      "loss": 0.3741,
+      "loss": 0.3743,
       "step": 524
     },
     {
       "epoch": 2.190923317683881,
-      "grad_norm": 0.3978201459670487,
+      "grad_norm": 0.3954510488805559,
       "learning_rate": 2.7552305887161693e-05,
-      "loss": 0.3679,
+      "loss": 0.3677,
       "step": 525
     },
     {
       "epoch": 2.19509650495566,
-      "grad_norm": 0.37894988072691815,
+      "grad_norm": 0.38970195628515064,
       "learning_rate": 2.7498152835479187e-05,
-      "loss": 0.3543,
+      "loss": 0.3552,
       "step": 526
     },
     {
       "epoch": 2.1992696922274386,
-      "grad_norm": 0.3993957513093158,
+      "grad_norm": 0.3852986882770263,
       "learning_rate": 2.7443935745938758e-05,
-      "loss": 0.3693,
+      "loss": 0.3694,
       "step": 527
     },
     {
       "epoch": 2.2034428794992174,
-      "grad_norm": 0.3899365310337407,
+      "grad_norm": 0.40126545445232537,
       "learning_rate": 2.738965508158063e-05,
-      "loss": 0.3514,
+      "loss": 0.3511,
       "step": 528
     },
     {
       "epoch": 2.2076160667709965,
-      "grad_norm": 0.4058898701830202,
+      "grad_norm": 0.39709305345454193,
       "learning_rate": 2.733531130598796e-05,
-      "loss": 0.3697,
+      "loss": 0.37,
       "step": 529
     },
     {
       "epoch": 2.2117892540427753,
-      "grad_norm": 0.4339080358113983,
+      "grad_norm": 0.4573853287492846,
       "learning_rate": 2.728090488328293e-05,
-      "loss": 0.3698,
+      "loss": 0.3703,
       "step": 530
     },
     {
       "epoch": 2.215962441314554,
-      "grad_norm": 0.3832709437444192,
+      "grad_norm": 0.3800206025512786,
       "learning_rate": 2.722643627812275e-05,
-      "loss": 0.3565,
+      "loss": 0.3564,
       "step": 531
     },
     {
       "epoch": 2.2201356285863327,
-      "grad_norm": 0.42267800558798874,
+      "grad_norm": 0.43497388691383904,
       "learning_rate": 2.7171905955695694e-05,
-      "loss": 0.3495,
+      "loss": 0.3498,
       "step": 532
     },
     {
       "epoch": 2.2243088158581115,
-      "grad_norm": 0.4041160856856104,
+      "grad_norm": 0.42311481119219957,
       "learning_rate": 2.711731438171713e-05,
-      "loss": 0.359,
+      "loss": 0.3597,
       "step": 533
     },
     {
       "epoch": 2.2284820031298906,
-      "grad_norm": 0.3710954566142956,
+      "grad_norm": 0.36674403898324176,
       "learning_rate": 2.7062662022425557e-05,
-      "loss": 0.3882,
+      "loss": 0.3884,
       "step": 534
     },
     {
       "epoch": 2.2326551904016694,
-      "grad_norm": 0.3869511697438778,
+      "grad_norm": 0.40382064789587785,
       "learning_rate": 2.700794934457859e-05,
-      "loss": 0.3714,
+      "loss": 0.3715,
       "step": 535
     },
     {
       "epoch": 2.236828377673448,
-      "grad_norm": 0.4390951198775265,
+      "grad_norm": 0.4395425179423991,
       "learning_rate": 2.695317681544902e-05,
-      "loss": 0.3779,
+      "loss": 0.3776,
       "step": 536
     },
     {
       "epoch": 2.241001564945227,
-      "grad_norm": 0.36290615122360276,
+      "grad_norm": 0.37978733065610054,
       "learning_rate": 2.6898344902820763e-05,
-      "loss": 0.373,
+      "loss": 0.3736,
       "step": 537
     },
     {
       "epoch": 2.2451747522170056,
-      "grad_norm": 0.39990857247016026,
+      "grad_norm": 0.39627427078919697,
       "learning_rate": 2.6843454074984932e-05,
-      "loss": 0.3878,
+      "loss": 0.3874,
       "step": 538
     },
     {
       "epoch": 2.2493479394887848,
-      "grad_norm": 0.42651620255557604,
+      "grad_norm": 0.4097383676738989,
       "learning_rate": 2.678850480073578e-05,
       "loss": 0.3701,
       "step": 539
     },
     {
       "epoch": 2.2535211267605635,
-      "grad_norm": 0.3843081727875576,
+      "grad_norm": 0.3854242215792716,
       "learning_rate": 2.673349754936675e-05,
-      "loss": 0.3603,
+      "loss": 0.3605,
       "step": 540
     },
     {
       "epoch": 2.257694314032342,
-      "grad_norm": 0.44434321637608326,
+      "grad_norm": 0.4220564460016855,
       "learning_rate": 2.6678432790666397e-05,
-      "loss": 0.3591,
+      "loss": 0.3594,
       "step": 541
     },
     {
       "epoch": 2.261867501304121,
-      "grad_norm": 0.3831737080284553,
+      "grad_norm": 0.3810046626222295,
       "learning_rate": 2.662331099491445e-05,
-      "loss": 0.3576,
+      "loss": 0.3568,
       "step": 542
     },
     {
       "epoch": 2.2660406885758997,
-      "grad_norm": 0.418317821240372,
+      "grad_norm": 0.41688165368006497,
       "learning_rate": 2.6568132632877744e-05,
-      "loss": 0.3734,
+      "loss": 0.3738,
       "step": 543
     },
     {
       "epoch": 2.270213875847679,
-      "grad_norm": 0.3990873922950322,
+      "grad_norm": 0.3887994007524032,
       "learning_rate": 2.651289817580623e-05,
-      "loss": 0.3654,
+      "loss": 0.3656,
       "step": 544
     },
     {
       "epoch": 2.2743870631194576,
-      "grad_norm": 0.38990441136752285,
+      "grad_norm": 0.3994340889307665,
       "learning_rate": 2.6457608095428925e-05,
-      "loss": 0.3654,
+      "loss": 0.3655,
       "step": 545
     },
     {
       "epoch": 2.2785602503912363,
-      "grad_norm": 0.38312325098253547,
+      "grad_norm": 0.38624989762376805,
       "learning_rate": 2.640226286394991e-05,
-      "loss": 0.3964,
+      "loss": 0.397,
       "step": 546
     },
     {
       "epoch": 2.282733437663015,
-      "grad_norm": 0.4054006085794088,
+      "grad_norm": 0.40988449987832615,
       "learning_rate": 2.6346862954044267e-05,
-      "loss": 0.3503,
+      "loss": 0.3505,
       "step": 547
     },
     {
       "epoch": 2.286906624934794,
-      "grad_norm": 0.3914282990079371,
+      "grad_norm": 0.39763056723013235,
       "learning_rate": 2.6291408838854074e-05,
-      "loss": 0.3508,
+      "loss": 0.3503,
       "step": 548
     },
     {
       "epoch": 2.291079812206573,
-      "grad_norm": 0.3850904322048971,
+      "grad_norm": 0.3919646441202639,
       "learning_rate": 2.6235900991984327e-05,
-      "loss": 0.3717,
+      "loss": 0.3714,
       "step": 549
     },
     {
       "epoch": 2.2952529994783517,
-      "grad_norm": 0.4498240460175877,
+      "grad_norm": 0.43387550110568074,
       "learning_rate": 2.618033988749895e-05,
-      "loss": 0.3679,
+      "loss": 0.3684,
       "step": 550
     },
     {
       "epoch": 2.2994261867501304,
-      "grad_norm": 0.35714957468528424,
+      "grad_norm": 0.36748065321980156,
       "learning_rate": 2.6124725999916684e-05,
-      "loss": 0.3672,
+      "loss": 0.3681,
       "step": 551
     },
     {
       "epoch": 2.303599374021909,
-      "grad_norm": 0.4782321646100373,
+      "grad_norm": 0.4646676929379771,
       "learning_rate": 2.6069059804207072e-05,
-      "loss": 0.3604,
+      "loss": 0.3607,
       "step": 552
     },
     {
       "epoch": 2.307772561293688,
-      "grad_norm": 0.3783567089821511,
+      "grad_norm": 0.3715511030786717,
       "learning_rate": 2.6013341775786396e-05,
-      "loss": 0.396,
+      "loss": 0.3961,
       "step": 553
     },
     {
       "epoch": 2.311945748565467,
-      "grad_norm": 0.41696750460385,
+      "grad_norm": 0.41164264620519486,
       "learning_rate": 2.5957572390513617e-05,
-      "loss": 0.3556,
+      "loss": 0.3553,
       "step": 554
     },
     {
       "epoch": 2.316118935837246,
-      "grad_norm": 0.375608638370001,
+      "grad_norm": 0.3733455470651702,
       "learning_rate": 2.5901752124686294e-05,
-      "loss": 0.3493,
+      "loss": 0.3487,
       "step": 555
     },
     {
       "epoch": 2.3202921231090246,
-      "grad_norm": 0.4181953294011339,
+      "grad_norm": 0.4061185558975713,
       "learning_rate": 2.5845881455036552e-05,
-      "loss": 0.3649,
+      "loss": 0.3648,
       "step": 556
     },
     {
       "epoch": 2.3244653103808033,
-      "grad_norm": 0.391372400945586,
+      "grad_norm": 0.39000899118046545,
       "learning_rate": 2.5789960858726968e-05,
-      "loss": 0.3674,
+      "loss": 0.3675,
       "step": 557
     },
     {
       "epoch": 2.328638497652582,
-      "grad_norm": 0.4198168178559077,
+      "grad_norm": 0.4243327191378882,
       "learning_rate": 2.5733990813346536e-05,
-      "loss": 0.3754,
+      "loss": 0.376,
       "step": 558
     },
     {
       "epoch": 2.332811684924361,
-      "grad_norm": 0.41707465847328457,
+      "grad_norm": 0.39489254687658626,
       "learning_rate": 2.5677971796906555e-05,
-      "loss": 0.3865,
+      "loss": 0.3871,
       "step": 559
     },
     {
       "epoch": 2.33698487219614,
-      "grad_norm": 0.3739110188893818,
+      "grad_norm": 0.378503176079661,
       "learning_rate": 2.5621904287836568e-05,
-      "loss": 0.367,
+      "loss": 0.3663,
       "step": 560
     },
     {
       "epoch": 2.3411580594679187,
-      "grad_norm": 0.3940534091361524,
+      "grad_norm": 0.39250655086874453,
       "learning_rate": 2.5565788764980265e-05,
-      "loss": 0.3624,
+      "loss": 0.3615,
       "step": 561
     },
     {
       "epoch": 2.3453312467396974,
-      "grad_norm": 0.35889792461142483,
+      "grad_norm": 0.36436967712453516,
       "learning_rate": 2.5509625707591394e-05,
-      "loss": 0.3663,
+      "loss": 0.3658,
       "step": 562
     },
     {
       "epoch": 2.349504434011476,
-      "grad_norm": 0.40052448379909006,
+      "grad_norm": 0.41344944272292744,
       "learning_rate": 2.5453415595329684e-05,
-      "loss": 0.3721,
+      "loss": 0.3728,
       "step": 563
     },
     {
       "epoch": 2.3536776212832553,
-      "grad_norm": 0.37689636513843067,
+      "grad_norm": 0.37759840510400666,
       "learning_rate": 2.539715890825672e-05,
-      "loss": 0.3615,
+      "loss": 0.3612,
       "step": 564
     },
     {
       "epoch": 2.357850808555034,
-      "grad_norm": 0.3926943375313602,
+      "grad_norm": 0.4034614253731671,
       "learning_rate": 2.5340856126831864e-05,
-      "loss": 0.3644,
+      "loss": 0.3643,
       "step": 565
     },
     {
       "epoch": 2.3620239958268128,
-      "grad_norm": 0.37546836429434155,
+      "grad_norm": 0.3890949879416151,
       "learning_rate": 2.528450773190814e-05,
-      "loss": 0.357,
+      "loss": 0.3567,
       "step": 566
     },
     {
       "epoch": 2.3661971830985915,
-      "grad_norm": 0.3992742496562248,
+      "grad_norm": 0.4010577083069116,
       "learning_rate": 2.5228114204728163e-05,
-      "loss": 0.3619,
+      "loss": 0.3621,
       "step": 567
     },
     {
       "epoch": 2.3703703703703702,
-      "grad_norm": 0.3961498687476965,
+      "grad_norm": 0.40248966467410935,
       "learning_rate": 2.5171676026919953e-05,
       "loss": 0.3713,
       "step": 568
     },
     {
       "epoch": 2.3745435576421494,
-      "grad_norm": 0.368233501258751,
+      "grad_norm": 0.3859188398500343,
       "learning_rate": 2.5115193680492908e-05,
-      "loss": 0.3674,
+      "loss": 0.3673,
       "step": 569
     },
     {
       "epoch": 2.378716744913928,
-      "grad_norm": 0.42157059821351517,
+      "grad_norm": 0.4149386744582473,
       "learning_rate": 2.5058667647833615e-05,
-      "loss": 0.3451,
+      "loss": 0.3452,
       "step": 570
     },
     {
       "epoch": 2.382889932185707,
-      "grad_norm": 0.39673841534437326,
+      "grad_norm": 0.3982819522221514,
       "learning_rate": 2.5002098411701797e-05,
-      "loss": 0.3723,
+      "loss": 0.372,
       "step": 571
     },
     {
       "epoch": 2.3870631194574856,
-      "grad_norm": 0.3833343005474658,
+      "grad_norm": 0.3882602940676942,
       "learning_rate": 2.494548645522612e-05,
-      "loss": 0.3625,
+      "loss": 0.363,
       "step": 572
     },
     {
       "epoch": 2.3912363067292643,
-      "grad_norm": 0.44451254841501814,
+      "grad_norm": 0.42927850735556794,
       "learning_rate": 2.4888832261900136e-05,
-      "loss": 0.355,
+      "loss": 0.3545,
       "step": 573
     },
     {
       "epoch": 2.3954094940010435,
-      "grad_norm": 0.38861927132457724,
+      "grad_norm": 0.39268838082287,
       "learning_rate": 2.4832136315578094e-05,
-      "loss": 0.3397,
+      "loss": 0.3398,
       "step": 574
     },
     {
       "epoch": 2.3995826812728223,
-      "grad_norm": 0.4255064682134055,
+      "grad_norm": 0.41804935180816016,
       "learning_rate": 2.4775399100470837e-05,
-      "loss": 0.3716,
+      "loss": 0.3722,
       "step": 575
     },
     {
       "epoch": 2.403755868544601,
-      "grad_norm": 0.47731397194631814,
+      "grad_norm": 0.46644616602423006,
       "learning_rate": 2.471862110114169e-05,
-      "loss": 0.3616,
+      "loss": 0.3617,
       "step": 576
     },
     {
       "epoch": 2.4079290558163797,
-      "grad_norm": 0.40031328304223135,
+      "grad_norm": 0.401805797819856,
       "learning_rate": 2.466180280250226e-05,
-      "loss": 0.3578,
+      "loss": 0.3577,
       "step": 577
     },
     {
       "epoch": 2.4121022430881585,
-      "grad_norm": 0.4400270936970402,
+      "grad_norm": 0.44727566826778437,
       "learning_rate": 2.4604944689808344e-05,
-      "loss": 0.3711,
+      "loss": 0.3712,
       "step": 578
     },
     {
       "epoch": 2.4162754303599376,
-      "grad_norm": 0.4113121072225045,
+      "grad_norm": 0.3979400364895845,
       "learning_rate": 2.4548047248655773e-05,
-      "loss": 0.3592,
+      "loss": 0.3585,
       "step": 579
     },
     {
       "epoch": 2.4204486176317164,
-      "grad_norm": 0.46213291821110014,
+      "grad_norm": 0.4848013007008637,
       "learning_rate": 2.449111096497627e-05,
-      "loss": 0.3688,
+      "loss": 0.3696,
       "step": 580
     },
     {
       "epoch": 2.424621804903495,
-      "grad_norm": 0.3996486886174434,
+      "grad_norm": 0.3969089989899458,
       "learning_rate": 2.443413632503327e-05,
       "loss": 0.3674,
       "step": 581
     },
     {
       "epoch": 2.428794992175274,
-      "grad_norm": 0.39146822308649354,
+      "grad_norm": 0.39473617240310976,
       "learning_rate": 2.4377123815417805e-05,
-      "loss": 0.386,
+      "loss": 0.3861,
       "step": 582
     },
     {
       "epoch": 2.4329681794470526,
-      "grad_norm": 0.41133315858808217,
+      "grad_norm": 0.4318276730872757,
       "learning_rate": 2.4320073923044326e-05,
-      "loss": 0.3824,
+      "loss": 0.3831,
       "step": 583
     },
     {
       "epoch": 2.4371413667188317,
-      "grad_norm": 0.4037601574862255,
+      "grad_norm": 0.4066990732619529,
       "learning_rate": 2.4262987135146555e-05,
-      "loss": 0.3707,
+      "loss": 0.3711,
       "step": 584
     },
     {
       "epoch": 2.4413145539906105,
-      "grad_norm": 0.3929087426717561,
+      "grad_norm": 0.41321294096672617,
       "learning_rate": 2.4205863939273328e-05,
-      "loss": 0.3669,
+      "loss": 0.3674,
       "step": 585
     },
     {
       "epoch": 2.445487741262389,
-      "grad_norm": 0.37883685255485,
+      "grad_norm": 0.38544793180038894,
       "learning_rate": 2.4148704823284395e-05,
-      "loss": 0.3593,
+      "loss": 0.3594,
       "step": 586
     },
     {
       "epoch": 2.449660928534168,
-      "grad_norm": 0.4109451045668216,
+      "grad_norm": 0.4124131172511295,
       "learning_rate": 2.4091510275346317e-05,
-      "loss": 0.3764,
+      "loss": 0.3766,
       "step": 587
     },
     {
       "epoch": 2.4538341158059467,
-      "grad_norm": 0.36786730357384584,
+      "grad_norm": 0.3801385155414366,
       "learning_rate": 2.4034280783928227e-05,
-      "loss": 0.3564,
+      "loss": 0.3567,
       "step": 588
     },
     {
       "epoch": 2.458007303077726,
-      "grad_norm": 0.38595289793563176,
+      "grad_norm": 0.3916594968221022,
       "learning_rate": 2.3977016837797718e-05,
-      "loss": 0.3592,
+      "loss": 0.3589,
       "step": 589
     },
     {
       "epoch": 2.4621804903495046,
-      "grad_norm": 0.41099406719974313,
+      "grad_norm": 0.41603101712646334,
       "learning_rate": 2.391971892601663e-05,
-      "loss": 0.3672,
+      "loss": 0.3679,
       "step": 590
     },
     {
       "epoch": 2.4663536776212833,
-      "grad_norm": 0.3818343314329391,
+      "grad_norm": 0.388463399721395,
       "learning_rate": 2.3862387537936894e-05,
-      "loss": 0.3617,
+      "loss": 0.3605,
       "step": 591
     },
     {
       "epoch": 2.470526864893062,
-      "grad_norm": 0.4178388737287968,
+      "grad_norm": 0.431222852105575,
       "learning_rate": 2.3805023163196326e-05,
-      "loss": 0.3649,
+      "loss": 0.3657,
       "step": 592
     },
     {
       "epoch": 2.474700052164841,
-      "grad_norm": 0.3717336068271807,
+      "grad_norm": 0.3702864697637371,
       "learning_rate": 2.37476262917145e-05,
-      "loss": 0.3766,
+      "loss": 0.377,
       "step": 593
     },
     {
       "epoch": 2.4788732394366195,
-      "grad_norm": 0.42885554072128773,
+      "grad_norm": 0.4315558528777625,
       "learning_rate": 2.3690197413688483e-05,
-      "loss": 0.3763,
+      "loss": 0.3769,
       "step": 594
     },
     {
       "epoch": 2.4830464267083987,
-      "grad_norm": 0.37132027646020377,
+      "grad_norm": 0.3669782752076605,
       "learning_rate": 2.363273701958873e-05,
       "loss": 0.3656,
       "step": 595
     },
     {
       "epoch": 2.4872196139801774,
-      "grad_norm": 0.3843886170339574,
+      "grad_norm": 0.3898231321679538,
       "learning_rate": 2.3575245600154845e-05,
-      "loss": 0.3843,
+      "loss": 0.3846,
       "step": 596
     },
     {
       "epoch": 2.491392801251956,
-      "grad_norm": 0.43986336799786974,
+      "grad_norm": 0.44353857646959477,
       "learning_rate": 2.35177236463914e-05,
-      "loss": 0.3851,
+      "loss": 0.3856,
       "step": 597
     },
     {
       "epoch": 2.495565988523735,
-      "grad_norm": 0.383850041309566,
+      "grad_norm": 0.381672637675844,
       "learning_rate": 2.3460171649563753e-05,
-      "loss": 0.3523,
+      "loss": 0.3525,
       "step": 598
     },
     {
       "epoch": 2.4997391757955136,
-      "grad_norm": 0.36181839009301886,
+      "grad_norm": 0.36898653513101676,
       "learning_rate": 2.3402590101193828e-05,
       "loss": 0.3507,
       "step": 599
     },
     {
       "epoch": 2.5039123630672924,
-      "grad_norm": 0.4252983259874796,
+      "grad_norm": 0.4264766508457069,
       "learning_rate": 2.3344979493055958e-05,
-      "loss": 0.3777,
+      "loss": 0.3781,
       "step": 600
     },
     {
       "epoch": 2.5080855503390715,
-      "grad_norm": 0.428522509095458,
+      "grad_norm": 0.41580112419463017,
       "learning_rate": 2.3287340317172638e-05,
       "loss": 0.363,
       "step": 601
     },
     {
       "epoch": 2.5122587376108503,
-      "grad_norm": 0.4538231283211154,
+      "grad_norm": 0.46784662322167064,
       "learning_rate": 2.3229673065810365e-05,
-      "loss": 0.3687,
+      "loss": 0.3691,
       "step": 602
     },
     {
       "epoch": 2.516431924882629,
-      "grad_norm": 0.38763035122663436,
+      "grad_norm": 0.38016343122242824,
       "learning_rate": 2.3171978231475403e-05,
-      "loss": 0.3544,
+      "loss": 0.3542,
       "step": 603
     },
     {
       "epoch": 2.520605112154408,
-      "grad_norm": 0.3921016688627311,
+      "grad_norm": 0.3928652984832863,
       "learning_rate": 2.311425630690958e-05,
-      "loss": 0.362,
+      "loss": 0.3614,
       "step": 604
     },
     {
       "epoch": 2.5247782994261865,
-      "grad_norm": 0.3698615814150281,
+      "grad_norm": 0.37036471328556103,
       "learning_rate": 2.3056507785086105e-05,
-      "loss": 0.3587,
+      "loss": 0.3586,
       "step": 605
     },
     {
       "epoch": 2.5289514866979657,
-      "grad_norm": 0.4073546521492878,
+      "grad_norm": 0.3986358144102633,
       "learning_rate": 2.2998733159205325e-05,
-      "loss": 0.3641,
+      "loss": 0.3637,
       "step": 606
     },
     {
       "epoch": 2.5331246739697444,
-      "grad_norm": 0.3862619769423595,
+      "grad_norm": 0.38893128877409056,
       "learning_rate": 2.2940932922690543e-05,
-      "loss": 0.3618,
+      "loss": 0.3615,
       "step": 607
     },
     {
       "epoch": 2.537297861241523,
-      "grad_norm": 0.35972152163761056,
+      "grad_norm": 0.3516606764357882,
       "learning_rate": 2.2883107569183776e-05,
-      "loss": 0.3743,
+      "loss": 0.3735,
       "step": 608
     },
     {
       "epoch": 2.5414710485133023,
-      "grad_norm": 0.4087989778800891,
+      "grad_norm": 0.40490866173229473,
       "learning_rate": 2.282525759254156e-05,
       "loss": 0.3903,
       "step": 609
     },
     {
       "epoch": 2.5456442357850806,
-      "grad_norm": 0.40558979032762105,
+      "grad_norm": 0.4021176562691965,
       "learning_rate": 2.2767383486830728e-05,
       "loss": 0.3665,
       "step": 610
     },
     {
       "epoch": 2.5498174230568598,
-      "grad_norm": 0.3775373696305449,
+      "grad_norm": 0.3726464324779564,
       "learning_rate": 2.2709485746324175e-05,
       "loss": 0.3587,
       "step": 611
     },
     {
       "epoch": 2.5539906103286385,
-      "grad_norm": 0.4078802463212513,
+      "grad_norm": 0.41447494767066057,
       "learning_rate": 2.265156486549666e-05,
-      "loss": 0.3784,
+      "loss": 0.3786,
       "step": 612
     },
     {
       "epoch": 2.5581637976004172,
-      "grad_norm": 0.41639680349464325,
+      "grad_norm": 0.4162522774962629,
       "learning_rate": 2.259362133902056e-05,
-      "loss": 0.3833,
+      "loss": 0.3837,
       "step": 613
     },
     {
       "epoch": 2.5623369848721964,
-      "grad_norm": 0.3811953210509229,
+      "grad_norm": 0.37895686838546944,
       "learning_rate": 2.2535655661761677e-05,
-      "loss": 0.3738,
+      "loss": 0.3735,
       "step": 614
     },
     {
       "epoch": 2.5665101721439747,
-      "grad_norm": 0.41403570517072036,
+      "grad_norm": 0.4190815315515346,
       "learning_rate": 2.247766832877496e-05,
-      "loss": 0.3664,
+      "loss": 0.3667,
       "step": 615
     },
     {
       "epoch": 2.570683359415754,
-      "grad_norm": 0.3970813010240881,
+      "grad_norm": 0.39672946355440253,
       "learning_rate": 2.241965983530034e-05,
-      "loss": 0.3564,
+      "loss": 0.3567,
       "step": 616
     },
     {
       "epoch": 2.5748565466875326,
-      "grad_norm": 0.34682796927818066,
+      "grad_norm": 0.3557957987792657,
       "learning_rate": 2.236163067675845e-05,
-      "loss": 0.3622,
+      "loss": 0.3624,
       "step": 617
     },
     {
       "epoch": 2.5790297339593113,
-      "grad_norm": 0.39284291816646477,
+      "grad_norm": 0.38807405661340333,
       "learning_rate": 2.230358134874642e-05,
-      "loss": 0.3681,
+      "loss": 0.3683,
       "step": 618
     },
     {
       "epoch": 2.5832029212310905,
-      "grad_norm": 0.3781405422181648,
+      "grad_norm": 0.37817633438320514,
       "learning_rate": 2.2245512347033625e-05,
       "loss": 0.3703,
       "step": 619
     },
     {
       "epoch": 2.587376108502869,
-      "grad_norm": 0.36338275095255373,
+      "grad_norm": 0.35840630304770754,
       "learning_rate": 2.2187424167557496e-05,
-      "loss": 0.3794,
+      "loss": 0.3799,
       "step": 620
     },
     {
       "epoch": 2.591549295774648,
-      "grad_norm": 0.37578674221765296,
+      "grad_norm": 0.37234819779805234,
       "learning_rate": 2.2129317306419198e-05,
-      "loss": 0.3553,
+      "loss": 0.3555,
       "step": 621
     },
     {
       "epoch": 2.5957224830464267,
-      "grad_norm": 0.364648028692508,
+      "grad_norm": 0.36482388973676755,
       "learning_rate": 2.2071192259879498e-05,
-      "loss": 0.368,
+      "loss": 0.3683,
       "step": 622
     },
     {
       "epoch": 2.5998956703182055,
-      "grad_norm": 0.38028060168400407,
+      "grad_norm": 0.37474079930650633,
       "learning_rate": 2.2013049524354453e-05,
-      "loss": 0.3534,
+      "loss": 0.3528,
       "step": 623
     },
     {
       "epoch": 2.604068857589984,
-      "grad_norm": 0.39075219539176836,
+      "grad_norm": 0.39251840745199673,
       "learning_rate": 2.1954889596411184e-05,
-      "loss": 0.3531,
+      "loss": 0.353,
       "step": 624
     },
     {
       "epoch": 2.608242044861763,
-      "grad_norm": 0.37626468610046376,
+      "grad_norm": 0.36853445253476425,
       "learning_rate": 2.1896712972763658e-05,
-      "loss": 0.3757,
+      "loss": 0.3759,
       "step": 625
     },
     {
       "epoch": 2.612415232133542,
-      "grad_norm": 0.3631753684648558,
+      "grad_norm": 0.369618731477996,
       "learning_rate": 2.183852015026842e-05,
-      "loss": 0.377,
+      "loss": 0.3767,
       "step": 626
     },
     {
       "epoch": 2.616588419405321,
-      "grad_norm": 0.3875500503641146,
+      "grad_norm": 0.38579149894574283,
       "learning_rate": 2.178031162592038e-05,
-      "loss": 0.3744,
+      "loss": 0.3745,
       "step": 627
     },
     {
       "epoch": 2.6207616066770996,
-      "grad_norm": 0.3671229318461734,
+      "grad_norm": 0.3656999203052834,
       "learning_rate": 2.1722087896848528e-05,
-      "loss": 0.3683,
+      "loss": 0.3686,
       "step": 628
     },
     {
       "epoch": 2.6249347939488783,
-      "grad_norm": 0.38219453208836063,
+      "grad_norm": 0.3766909600093165,
       "learning_rate": 2.1663849460311722e-05,
-      "loss": 0.3709,
+      "loss": 0.3708,
       "step": 629
     },
     {
       "epoch": 2.629107981220657,
-      "grad_norm": 0.408073083775421,
+      "grad_norm": 0.4006260235164137,
       "learning_rate": 2.1605596813694426e-05,
-      "loss": 0.3803,
+      "loss": 0.3804,
       "step": 630
     },
     {
       "epoch": 2.633281168492436,
-      "grad_norm": 0.3729105134945073,
+      "grad_norm": 0.3743088967507613,
       "learning_rate": 2.154733045450246e-05,
-      "loss": 0.3649,
+      "loss": 0.365,
       "step": 631
     },
     {
       "epoch": 2.637454355764215,
-      "grad_norm": 0.39085770994359953,
+      "grad_norm": 0.3853263550418572,
       "learning_rate": 2.1489050880358758e-05,
-      "loss": 0.3643,
+      "loss": 0.3648,
       "step": 632
     },
     {
       "epoch": 2.6416275430359937,
-      "grad_norm": 0.3664447312714141,
+      "grad_norm": 0.3749369195036436,
       "learning_rate": 2.143075858899913e-05,
-      "loss": 0.3672,
+      "loss": 0.3677,
       "step": 633
     },
     {
       "epoch": 2.6458007303077724,
-      "grad_norm": 0.35740063168622305,
+      "grad_norm": 0.3679876569694932,
       "learning_rate": 2.1372454078267975e-05,
-      "loss": 0.3703,
+      "loss": 0.3706,
       "step": 634
     },
     {
       "epoch": 2.649973917579551,
-      "grad_norm": 0.377792503284453,
+      "grad_norm": 0.392409686254139,
       "learning_rate": 2.131413784611406e-05,
-      "loss": 0.3733,
+      "loss": 0.3731,
       "step": 635
     },
     {
       "epoch": 2.6541471048513303,
-      "grad_norm": 0.3597852720741351,
+      "grad_norm": 0.36448683135574694,
       "learning_rate": 2.125581039058627e-05,
-      "loss": 0.3601,
+      "loss": 0.3605,
       "step": 636
     },
     {
       "epoch": 2.658320292123109,
-      "grad_norm": 0.4060714627756588,
+      "grad_norm": 0.40268288693165893,
       "learning_rate": 2.1197472209829328e-05,
-      "loss": 0.3822,
+      "loss": 0.3825,
       "step": 637
     },
     {
       "epoch": 2.662493479394888,
-      "grad_norm": 0.3775510896754153,
+      "grad_norm": 0.3832364960499003,
       "learning_rate": 2.113912380207957e-05,
-      "loss": 0.35,
+      "loss": 0.3502,
       "step": 638
     },
     {
       "epoch": 2.6666666666666665,
-      "grad_norm": 0.38417324555359916,
+      "grad_norm": 0.3899593645291949,
       "learning_rate": 2.1080765665660653e-05,
-      "loss": 0.3691,
+      "loss": 0.3689,
       "step": 639
     },
     {
       "epoch": 2.6708398539384453,
-      "grad_norm": 0.34911782147707876,
+      "grad_norm": 0.33956085716469847,
       "learning_rate": 2.1022398298979345e-05,
-      "loss": 0.3647,
+      "loss": 0.3642,
       "step": 640
     },
     {
       "epoch": 2.6750130412102244,
-      "grad_norm": 0.43601514430663374,
+      "grad_norm": 0.4330196386987855,
       "learning_rate": 2.096402220052123e-05,
       "loss": 0.376,
       "step": 641
     },
     {
       "epoch": 2.679186228482003,
-      "grad_norm": 0.3820463852414991,
+      "grad_norm": 0.3799835681997979,
       "learning_rate": 2.0905637868846478e-05,
       "loss": 0.3478,
       "step": 642
     },
     {
       "epoch": 2.683359415753782,
-      "grad_norm": 0.41102400832677854,
+      "grad_norm": 0.39738658384567416,
       "learning_rate": 2.0847245802585567e-05,
       "loss": 0.3612,
       "step": 643
     },
     {
       "epoch": 2.6875326030255606,
-      "grad_norm": 0.3813575042508285,
+      "grad_norm": 0.37678342085560784,
       "learning_rate": 2.078884650043502e-05,
-      "loss": 0.372,
+      "loss": 0.3716,
       "step": 644
     },
     {
       "epoch": 2.6917057902973394,
-      "grad_norm": 0.36898682604740835,
+      "grad_norm": 0.35574510425956074,
       "learning_rate": 2.0730440461153183e-05,
-      "loss": 0.371,
+      "loss": 0.3714,
       "step": 645
     },
     {
       "epoch": 2.6958789775691185,
-      "grad_norm": 0.3859606936935168,
+      "grad_norm": 0.39014874697386115,
       "learning_rate": 2.067202818355592e-05,
-      "loss": 0.3785,
+      "loss": 0.3787,
       "step": 646
     },
     {
       "epoch": 2.7000521648408973,
-      "grad_norm": 0.38497647547219854,
+      "grad_norm": 0.37342671641895336,
       "learning_rate": 2.061361016651239e-05,
-      "loss": 0.3677,
+      "loss": 0.3675,
       "step": 647
     },
     {
       "epoch": 2.704225352112676,
-      "grad_norm": 0.3653421314161833,
+      "grad_norm": 0.36737074258166824,
       "learning_rate": 2.0555186908940753e-05,
-      "loss": 0.351,
+      "loss": 0.3504,
       "step": 648
     },
     {
       "epoch": 2.7083985393844547,
-      "grad_norm": 0.39125438427377796,
+      "grad_norm": 0.3898373046949013,
       "learning_rate": 2.049675890980394e-05,
-      "loss": 0.36,
+      "loss": 0.3602,
       "step": 649
     },
     {
       "epoch": 2.7125717266562335,
-      "grad_norm": 0.36236470140930904,
+      "grad_norm": 0.3737858750307865,
       "learning_rate": 2.0438326668105364e-05,
-      "loss": 0.364,
+      "loss": 0.3644,
       "step": 650
     },
     {
       "epoch": 2.7167449139280127,
-      "grad_norm": 0.38643707077573186,
+      "grad_norm": 0.3794327235824899,
       "learning_rate": 2.0379890682884685e-05,
-      "loss": 0.3841,
+      "loss": 0.3846,
       "step": 651
     },
     {
       "epoch": 2.7209181011997914,
-      "grad_norm": 0.36638624236779455,
+      "grad_norm": 0.3689050480135403,
       "learning_rate": 2.0321451453213518e-05,
-      "loss": 0.3654,
+      "loss": 0.3659,
       "step": 652
     },
     {
       "epoch": 2.72509128847157,
-      "grad_norm": 0.378803915597188,
+      "grad_norm": 0.3702626640382188,
       "learning_rate": 2.026300947819121e-05,
-      "loss": 0.3684,
+      "loss": 0.368,
       "step": 653
     },
     {
       "epoch": 2.729264475743349,
-      "grad_norm": 0.36868955958638144,
+      "grad_norm": 0.35793596239729064,
       "learning_rate": 2.020456525694054e-05,
-      "loss": 0.3632,
+      "loss": 0.3633,
       "step": 654
     },
     {
       "epoch": 2.7334376630151276,
-      "grad_norm": 0.37356038467071667,
+      "grad_norm": 0.37569467910756593,
       "learning_rate": 2.014611928860346e-05,
-      "loss": 0.3648,
+      "loss": 0.3644,
       "step": 655
     },
     {
       "epoch": 2.7376108502869068,
-      "grad_norm": 0.36436146944948733,
+      "grad_norm": 0.351533197877585,
       "learning_rate": 2.008767207233687e-05,
-      "loss": 0.3712,
+      "loss": 0.3714,
       "step": 656
     },
     {
       "epoch": 2.7417840375586855,
-      "grad_norm": 0.37947636337870033,
+      "grad_norm": 0.38097207606951905,
       "learning_rate": 2.0029224107308304e-05,
-      "loss": 0.367,
+      "loss": 0.3675,
       "step": 657
     },
     {
       "epoch": 2.7459572248304642,
-      "grad_norm": 0.37598312587298516,
+      "grad_norm": 0.3673233385358431,
       "learning_rate": 1.99707758926917e-05,
-      "loss": 0.353,
+      "loss": 0.3531,
       "step": 658
     },
     {
       "epoch": 2.750130412102243,
-      "grad_norm": 0.37814691994940786,
+      "grad_norm": 0.380136921429998,
       "learning_rate": 1.9912327927663135e-05,
-      "loss": 0.3609,
+      "loss": 0.361,
       "step": 659
     },
     {
       "epoch": 2.7543035993740217,
-      "grad_norm": 0.3812294198458229,
+      "grad_norm": 0.37205665195123583,
       "learning_rate": 1.985388071139654e-05,
-      "loss": 0.3714,
+      "loss": 0.3716,
       "step": 660
     },
     {
       "epoch": 2.758476786645801,
-      "grad_norm": 0.37174344277091165,
+      "grad_norm": 0.3738741066015198,
       "learning_rate": 1.9795434743059467e-05,
-      "loss": 0.3678,
+      "loss": 0.3685,
       "step": 661
     },
     {
       "epoch": 2.7626499739175796,
-      "grad_norm": 0.3784743961661241,
+      "grad_norm": 0.3689800126503024,
       "learning_rate": 1.9736990521808793e-05,
       "loss": 0.3776,
       "step": 662
     },
     {
       "epoch": 2.7668231611893583,
-      "grad_norm": 0.39643887414710466,
+      "grad_norm": 0.39879922091620224,
       "learning_rate": 1.9678548546786486e-05,
-      "loss": 0.3723,
+      "loss": 0.3726,
       "step": 663
     },
     {
       "epoch": 2.770996348461137,
-      "grad_norm": 0.3710985214315389,
+      "grad_norm": 0.378999294681146,
       "learning_rate": 1.9620109317115325e-05,
-      "loss": 0.377,
+      "loss": 0.3765,
       "step": 664
     },
     {
       "epoch": 2.775169535732916,
-      "grad_norm": 0.35734558416091006,
+      "grad_norm": 0.35418608827780307,
       "learning_rate": 1.956167333189464e-05,
-      "loss": 0.351,
+      "loss": 0.3514,
       "step": 665
     },
     {
       "epoch": 2.779342723004695,
-      "grad_norm": 0.3828478622907984,
+      "grad_norm": 0.39713847335258956,
       "learning_rate": 1.9503241090196068e-05,
-      "loss": 0.354,
+      "loss": 0.3549,
       "step": 666
     },
     {
       "epoch": 2.7835159102764737,
-      "grad_norm": 0.3738399746108767,
+      "grad_norm": 0.3789011064016802,
       "learning_rate": 1.9444813091059253e-05,
-      "loss": 0.3683,
+      "loss": 0.3677,
       "step": 667
     },
     {
       "epoch": 2.7876890975482524,
-      "grad_norm": 0.37839878552716644,
+      "grad_norm": 0.38092090561265163,
       "learning_rate": 1.938638983348762e-05,
-      "loss": 0.3655,
+      "loss": 0.3659,
       "step": 668
     },
     {
       "epoch": 2.791862284820031,
-      "grad_norm": 0.35292660957372796,
+      "grad_norm": 0.36854426040366667,
       "learning_rate": 1.9327971816444086e-05,
       "loss": 0.3549,
       "step": 669
     },
     {
       "epoch": 2.79603547209181,
-      "grad_norm": 0.36396540303493063,
+      "grad_norm": 0.3533695252315581,
       "learning_rate": 1.9269559538846823e-05,
       "loss": 0.3911,
       "step": 670
     },
     {
       "epoch": 2.800208659363589,
-      "grad_norm": 0.3609361518726451,
+      "grad_norm": 0.3587911579256342,
       "learning_rate": 1.9211153499564984e-05,
-      "loss": 0.3696,
+      "loss": 0.3703,
       "step": 671
     },
     {
       "epoch": 2.804381846635368,
-      "grad_norm": 0.38276716564869806,
+      "grad_norm": 0.3802652353736712,
       "learning_rate": 1.9152754197414443e-05,
-      "loss": 0.3602,
+      "loss": 0.3608,
       "step": 672
     },
     {
       "epoch": 2.8085550339071466,
-      "grad_norm": 0.3516968526777236,
+      "grad_norm": 0.3377892433696322,
       "learning_rate": 1.909436213115353e-05,
-      "loss": 0.359,
+      "loss": 0.3588,
       "step": 673
     },
     {
       "epoch": 2.8127282211789253,
-      "grad_norm": 0.3742854111971573,
+      "grad_norm": 0.3732109780735204,
       "learning_rate": 1.9035977799478772e-05,
-      "loss": 0.3505,
+      "loss": 0.3501,
       "step": 674
     },
     {
       "epoch": 2.816901408450704,
-      "grad_norm": 0.4081054028288133,
+      "grad_norm": 0.39949876985035737,
       "learning_rate": 1.897760170102066e-05,
       "loss": 0.3597,
       "step": 675
     },
     {
       "epoch": 2.821074595722483,
-      "grad_norm": 0.3769607036889555,
+      "grad_norm": 0.36415092944746774,
       "learning_rate": 1.8919234334339354e-05,
-      "loss": 0.3762,
+      "loss": 0.3765,
       "step": 676
     },
     {
       "epoch": 2.825247782994262,
-      "grad_norm": 0.3644404453127881,
+      "grad_norm": 0.3807849662202544,
       "learning_rate": 1.886087619792044e-05,
-      "loss": 0.365,
+      "loss": 0.3657,
       "step": 677
     },
     {
       "epoch": 2.8294209702660407,
-      "grad_norm": 0.3955284240243983,
+      "grad_norm": 0.3797056915274904,
       "learning_rate": 1.880252779017067e-05,
-      "loss": 0.3555,
+      "loss": 0.3554,
       "step": 678
     },
     {
       "epoch": 2.8335941575378194,
-      "grad_norm": 0.35657786458733814,
+      "grad_norm": 0.3691545222487018,
       "learning_rate": 1.8744189609413733e-05,
-      "loss": 0.3746,
+      "loss": 0.3747,
       "step": 679
     },
     {
       "epoch": 2.837767344809598,
-      "grad_norm": 0.4290751473364993,
+      "grad_norm": 0.42512990849734744,
       "learning_rate": 1.8685862153885947e-05,
-      "loss": 0.3622,
+      "loss": 0.362,
       "step": 680
     },
     {
       "epoch": 2.8419405320813773,
-      "grad_norm": 0.3405809618972471,
+      "grad_norm": 0.34478648810851364,
       "learning_rate": 1.8627545921732038e-05,
-      "loss": 0.3535,
+      "loss": 0.3533,
       "step": 681
     },
     {
       "epoch": 2.846113719353156,
-      "grad_norm": 0.3724725742177897,
+      "grad_norm": 0.3764329570986726,
       "learning_rate": 1.856924141100088e-05,
-      "loss": 0.3748,
+      "loss": 0.3753,
       "step": 682
     },
     {
       "epoch": 2.850286906624935,
-      "grad_norm": 0.3530807374317114,
+      "grad_norm": 0.3400317126599088,
       "learning_rate": 1.8510949119641246e-05,
-      "loss": 0.3814,
+      "loss": 0.3808,
       "step": 683
     },
     {
       "epoch": 2.8544600938967135,
-      "grad_norm": 0.34612946024499613,
+      "grad_norm": 0.3442654890227704,
       "learning_rate": 1.8452669545497545e-05,
-      "loss": 0.354,
+      "loss": 0.3543,
       "step": 684
     },
     {
       "epoch": 2.8586332811684922,
-      "grad_norm": 0.3498203316829098,
+      "grad_norm": 0.3593986550938782,
       "learning_rate": 1.839440318630558e-05,
-      "loss": 0.3646,
+      "loss": 0.3643,
       "step": 685
     },
     {
       "epoch": 2.8628064684402714,
-      "grad_norm": 0.35776192586278743,
+      "grad_norm": 0.3557947821034801,
       "learning_rate": 1.833615053968828e-05,
       "loss": 0.3532,
       "step": 686
     },
     {
       "epoch": 2.86697965571205,
-      "grad_norm": 0.35042091659524294,
+      "grad_norm": 0.3540785418384583,
       "learning_rate": 1.8277912103151475e-05,
-      "loss": 0.3433,
+      "loss": 0.3432,
       "step": 687
     },
     {
       "epoch": 2.871152842983829,
-      "grad_norm": 0.3441898029835142,
+      "grad_norm": 0.3444869662562323,
       "learning_rate": 1.8219688374079624e-05,
-      "loss": 0.3679,
+      "loss": 0.3678,
       "step": 688
     },
     {
       "epoch": 2.8753260302556076,
-      "grad_norm": 0.3419949126423415,
+      "grad_norm": 0.34165422316036803,
       "learning_rate": 1.8161479849731582e-05,
       "loss": 0.3822,
       "step": 689
     },
     {
       "epoch": 2.8794992175273864,
-      "grad_norm": 0.35694202289084365,
+      "grad_norm": 0.3523944379831676,
       "learning_rate": 1.8103287027236352e-05,
-      "loss": 0.3507,
+      "loss": 0.3508,
       "step": 690
     },
     {
       "epoch": 2.8836724047991655,
-      "grad_norm": 0.34521846357174235,
+      "grad_norm": 0.3469827576282236,
       "learning_rate": 1.804511040358882e-05,
-      "loss": 0.3704,
+      "loss": 0.3697,
       "step": 691
     },
     {
       "epoch": 2.8878455920709443,
-      "grad_norm": 0.3349677371849335,
+      "grad_norm": 0.33877675464214985,
       "learning_rate": 1.7986950475645553e-05,
-      "loss": 0.3703,
+      "loss": 0.3695,
       "step": 692
     },
     {
       "epoch": 2.892018779342723,
-      "grad_norm": 0.34353335626498843,
+      "grad_norm": 0.35158623728039595,
       "learning_rate": 1.7928807740120506e-05,
-      "loss": 0.3756,
+      "loss": 0.3755,
       "step": 693
     },
     {
       "epoch": 2.8961919666145017,
-      "grad_norm": 0.3525278104680631,
+      "grad_norm": 0.352807261845618,
       "learning_rate": 1.787068269358081e-05,
-      "loss": 0.3748,
+      "loss": 0.3754,
       "step": 694
     },
     {
       "epoch": 2.9003651538862805,
-      "grad_norm": 0.3490121564933543,
+      "grad_norm": 0.35449027205408334,
       "learning_rate": 1.7812575832442518e-05,
-      "loss": 0.3688,
+      "loss": 0.3685,
       "step": 695
     },
     {
       "epoch": 2.9045383411580596,
-      "grad_norm": 0.322994876026181,
+      "grad_norm": 0.32907100649761745,
       "learning_rate": 1.775448765296637e-05,
-      "loss": 0.3686,
+      "loss": 0.3689,
       "step": 696
     },
     {
       "epoch": 2.9087115284298384,
-      "grad_norm": 0.34522668485534574,
+      "grad_norm": 0.3495486925739955,
       "learning_rate": 1.7696418651253585e-05,
-      "loss": 0.3602,
+      "loss": 0.3607,
       "step": 697
     },
     {
       "epoch": 2.912884715701617,
-      "grad_norm": 0.34597284483025825,
+      "grad_norm": 0.3500518697560427,
       "learning_rate": 1.7638369323241554e-05,
-      "loss": 0.3638,
+      "loss": 0.3642,
       "step": 698
     },
     {
       "epoch": 2.917057902973396,
-      "grad_norm": 0.3425871472877568,
+      "grad_norm": 0.34481675264850625,
       "learning_rate": 1.7580340164699666e-05,
-      "loss": 0.3761,
+      "loss": 0.376,
       "step": 699
     },
     {
       "epoch": 2.9212310902451746,
-      "grad_norm": 0.3549718972373015,
+      "grad_norm": 0.35202457526037984,
       "learning_rate": 1.752233167122504e-05,
-      "loss": 0.3576,
+      "loss": 0.3567,
       "step": 700
     },
     {
       "epoch": 2.9254042775169538,
-      "grad_norm": 0.3576698082011347,
+      "grad_norm": 0.3639521898173101,
       "learning_rate": 1.7464344338238327e-05,
-      "loss": 0.3503,
+      "loss": 0.3507,
       "step": 701
     },
     {
       "epoch": 2.9295774647887325,
-      "grad_norm": 0.3388144705080834,
+      "grad_norm": 0.3401588180885393,
       "learning_rate": 1.7406378660979443e-05,
-      "loss": 0.3553,
+      "loss": 0.3552,
       "step": 702
     },
     {
       "epoch": 2.933750652060511,
-      "grad_norm": 0.3389665602513851,
+      "grad_norm": 0.3470306181171091,
       "learning_rate": 1.7348435134503347e-05,
-      "loss": 0.3631,
+      "loss": 0.363,
       "step": 703
     },
     {
       "epoch": 2.93792383933229,
-      "grad_norm": 0.3474725385733189,
+      "grad_norm": 0.344767919810025,
       "learning_rate": 1.7290514253675832e-05,
-      "loss": 0.3729,
+      "loss": 0.3727,
       "step": 704
     },
     {
       "epoch": 2.9420970266040687,
-      "grad_norm": 0.35771584755338126,
+      "grad_norm": 0.3641647943678777,
       "learning_rate": 1.7232616513169272e-05,
-      "loss": 0.3592,
+      "loss": 0.3595,
       "step": 705
     },
     {
       "epoch": 2.946270213875848,
-      "grad_norm": 0.3462870291425408,
+      "grad_norm": 0.34560245193432954,
       "learning_rate": 1.717474240745844e-05,
-      "loss": 0.37,
+      "loss": 0.3703,
       "step": 706
     },
     {
       "epoch": 2.9504434011476266,
-      "grad_norm": 0.34564524742548247,
+      "grad_norm": 0.35092145250110474,
       "learning_rate": 1.711689243081623e-05,
-      "loss": 0.3588,
+      "loss": 0.3595,
       "step": 707
     },
     {
       "epoch": 2.9546165884194053,
-      "grad_norm": 0.35468462937590095,
+      "grad_norm": 0.35490413140460736,
       "learning_rate": 1.7059067077309467e-05,
-      "loss": 0.3701,
+      "loss": 0.3702,
       "step": 708
     },
     {
       "epoch": 2.958789775691184,
-      "grad_norm": 0.3466603942076293,
+      "grad_norm": 0.35194317807308506,
       "learning_rate": 1.7001266840794678e-05,
-      "loss": 0.3606,
+      "loss": 0.3605,
       "step": 709
     },
     {
       "epoch": 2.962962962962963,
-      "grad_norm": 0.36662040734798285,
+      "grad_norm": 0.365165358913193,
       "learning_rate": 1.69434922149139e-05,
-      "loss": 0.3618,
+      "loss": 0.3624,
       "step": 710
     },
     {
       "epoch": 2.967136150234742,
-      "grad_norm": 0.37042519574822785,
+      "grad_norm": 0.366617594164006,
       "learning_rate": 1.6885743693090422e-05,
-      "loss": 0.3509,
+      "loss": 0.351,
       "step": 711
     },
     {
       "epoch": 2.9713093375065207,
-      "grad_norm": 0.34524028877266866,
+      "grad_norm": 0.35160980043399026,
       "learning_rate": 1.6828021768524604e-05,
-      "loss": 0.3716,
+      "loss": 0.3717,
       "step": 712
     },
     {
       "epoch": 2.9754825247782994,
-      "grad_norm": 0.3330910561524174,
+      "grad_norm": 0.3325499642526304,
       "learning_rate": 1.6770326934189638e-05,
-      "loss": 0.3673,
+      "loss": 0.3679,
       "step": 713
     },
     {
       "epoch": 2.979655712050078,
-      "grad_norm": 0.3416799637540736,
+      "grad_norm": 0.3431312974243191,
       "learning_rate": 1.671265968282736e-05,
-      "loss": 0.3858,
+      "loss": 0.3863,
       "step": 714
     },
     {
       "epoch": 2.983828899321857,
-      "grad_norm": 0.32468722548074846,
+      "grad_norm": 0.32978477724071426,
       "learning_rate": 1.6655020506944046e-05,
-      "loss": 0.3598,
+      "loss": 0.3602,
       "step": 715
     },
     {
       "epoch": 2.988002086593636,
-      "grad_norm": 0.3453078752125995,
+      "grad_norm": 0.3470770528094728,
       "learning_rate": 1.659740989880618e-05,
-      "loss": 0.3535,
+      "loss": 0.3538,
       "step": 716
     },
     {
       "epoch": 2.992175273865415,
-      "grad_norm": 0.3388800637162424,
+      "grad_norm": 0.3416216896933235,
       "learning_rate": 1.653982835043626e-05,
-      "loss": 0.3565,
+      "loss": 0.3561,
       "step": 717
     },
     {
       "epoch": 2.9963484611371936,
-      "grad_norm": 0.3440742883765163,
+      "grad_norm": 0.35010781796470064,
       "learning_rate": 1.6482276353608602e-05,
-      "loss": 0.3823,
+      "loss": 0.383,
       "step": 718
     },
     {
       "epoch": 3.0005216484089723,
-      "grad_norm": 0.3543889482533494,
+      "grad_norm": 0.35480003007945177,
       "learning_rate": 1.642475439984516e-05,
-      "loss": 0.3346,
+      "loss": 0.335,
       "step": 719
     },
     {
       "epoch": 3.004694835680751,
-      "grad_norm": 0.5715896691266967,
+      "grad_norm": 0.5782467489622019,
       "learning_rate": 1.6367262980411273e-05,
-      "loss": 0.2534,
+      "loss": 0.2541,
       "step": 720
     },
     {
       "epoch": 3.00886802295253,
-      "grad_norm": 0.3707881189691564,
+      "grad_norm": 0.37817237108784724,
       "learning_rate": 1.6309802586311523e-05,
-      "loss": 0.259,
+      "loss": 0.2596,
       "step": 721
     },
     {
       "epoch": 3.013041210224309,
-      "grad_norm": 0.5604801888363965,
+      "grad_norm": 0.5504409730857783,
       "learning_rate": 1.6252373708285505e-05,
-      "loss": 0.2532,
+      "loss": 0.2529,
       "step": 722
     },
     {
       "epoch": 3.0172143974960877,
-      "grad_norm": 0.6956424221445986,
+      "grad_norm": 0.7068371529169682,
       "learning_rate": 1.6194976836803674e-05,
-      "loss": 0.2514,
+      "loss": 0.2517,
       "step": 723
     },
     {
       "epoch": 3.0213875847678664,
-      "grad_norm": 0.3882857101800422,
+      "grad_norm": 0.38140695604297026,
       "learning_rate": 1.6137612462063112e-05,
-      "loss": 0.2584,
+      "loss": 0.2586,
       "step": 724
     },
     {
       "epoch": 3.025560772039645,
-      "grad_norm": 0.5164821293766428,
+      "grad_norm": 0.5404237198737243,
       "learning_rate": 1.6080281073983375e-05,
-      "loss": 0.2594,
+      "loss": 0.2599,
       "step": 725
     },
     {
       "epoch": 3.0297339593114243,
-      "grad_norm": 0.5129370879688329,
+      "grad_norm": 0.5165220000211398,
       "learning_rate": 1.602298316220229e-05,
-      "loss": 0.2584,
+      "loss": 0.2585,
       "step": 726
     },
     {
       "epoch": 3.033907146583203,
-      "grad_norm": 0.39028303727206126,
+      "grad_norm": 0.3834808691976082,
       "learning_rate": 1.5965719216071776e-05,
-      "loss": 0.2638,
+      "loss": 0.2641,
       "step": 727
     },
     {
       "epoch": 3.0380803338549818,
-      "grad_norm": 0.4454728458991174,
+      "grad_norm": 0.4414277859758241,
       "learning_rate": 1.590848972465369e-05,
-      "loss": 0.2592,
+      "loss": 0.2599,
       "step": 728
     },
     {
       "epoch": 3.0422535211267605,
-      "grad_norm": 0.5182292325263265,
+      "grad_norm": 0.4945574087283716,
       "learning_rate": 1.5851295176715608e-05,
-      "loss": 0.2633,
+      "loss": 0.2635,
       "step": 729
     },
     {
       "epoch": 3.0464267083985392,
-      "grad_norm": 0.44265500073060654,
+      "grad_norm": 0.4547143078656569,
       "learning_rate": 1.5794136060726682e-05,
       "loss": 0.2608,
       "step": 730
     },
     {
       "epoch": 3.0505998956703184,
-      "grad_norm": 0.43581270853820403,
+      "grad_norm": 0.4196564926816924,
       "learning_rate": 1.573701286485344e-05,
       "loss": 0.254,
       "step": 731
     },
     {
       "epoch": 3.054773082942097,
-      "grad_norm": 0.41904646559891706,
+      "grad_norm": 0.39937888723472814,
       "learning_rate": 1.5679926076955677e-05,
-      "loss": 0.2359,
+      "loss": 0.2358,
       "step": 732
     },
     {
       "epoch": 3.058946270213876,
-      "grad_norm": 0.40196100766227705,
+      "grad_norm": 0.40325700901745176,
       "learning_rate": 1.5622876184582205e-05,
-      "loss": 0.255,
+      "loss": 0.2552,
       "step": 733
     },
     {
       "epoch": 3.0631194574856546,
-      "grad_norm": 0.38734575451331893,
+      "grad_norm": 0.3780706423525529,
       "learning_rate": 1.556586367496674e-05,
-      "loss": 0.2612,
+      "loss": 0.2613,
       "step": 734
     },
     {
       "epoch": 3.0672926447574334,
-      "grad_norm": 0.36544613932077685,
+      "grad_norm": 0.3620422861826096,
       "learning_rate": 1.5508889035023738e-05,
-      "loss": 0.2719,
+      "loss": 0.2723,
       "step": 735
     },
     {
       "epoch": 3.0714658320292125,
-      "grad_norm": 0.36056713901413406,
+      "grad_norm": 0.3553795475845748,
       "learning_rate": 1.5451952751344227e-05,
-      "loss": 0.2697,
+      "loss": 0.2699,
       "step": 736
     },
     {
       "epoch": 3.0756390193009913,
-      "grad_norm": 0.3661005349325296,
+      "grad_norm": 0.36353280083953776,
       "learning_rate": 1.539505531019166e-05,
-      "loss": 0.2539,
+      "loss": 0.2543,
       "step": 737
     },
     {
       "epoch": 3.07981220657277,
-      "grad_norm": 0.33633673527050023,
+      "grad_norm": 0.3450901887285021,
       "learning_rate": 1.533819719749775e-05,
-      "loss": 0.2623,
+      "loss": 0.2622,
       "step": 738
     },
     {
       "epoch": 3.0839853938445487,
-      "grad_norm": 0.3781134490843638,
+      "grad_norm": 0.3584936102209572,
       "learning_rate": 1.5281378898858318e-05,
-      "loss": 0.2224,
+      "loss": 0.2221,
       "step": 739
     },
     {
       "epoch": 3.0881585811163275,
-      "grad_norm": 0.3277206938339605,
+      "grad_norm": 0.32535222586689017,
       "learning_rate": 1.522460089952916e-05,
-      "loss": 0.2622,
+      "loss": 0.2626,
       "step": 740
     },
     {
       "epoch": 3.0923317683881066,
-      "grad_norm": 0.34734960977004437,
+      "grad_norm": 0.33664900452060037,
       "learning_rate": 1.5167863684421913e-05,
-      "loss": 0.2705,
+      "loss": 0.2706,
       "step": 741
     },
     {
       "epoch": 3.0965049556598854,
-      "grad_norm": 0.3254078427245206,
+      "grad_norm": 0.31586843752939436,
       "learning_rate": 1.511116773809987e-05,
-      "loss": 0.2669,
+      "loss": 0.2667,
       "step": 742
     },
     {
       "epoch": 3.100678142931664,
-      "grad_norm": 0.34462346301534946,
+      "grad_norm": 0.33460199906255683,
       "learning_rate": 1.5054513544773884e-05,
-      "loss": 0.2649,
+      "loss": 0.2656,
       "step": 743
     },
     {
       "epoch": 3.104851330203443,
-      "grad_norm": 0.3399148808880381,
+      "grad_norm": 0.3295245447664331,
       "learning_rate": 1.4997901588298214e-05,
-      "loss": 0.2361,
+      "loss": 0.2363,
       "step": 744
     },
     {
       "epoch": 3.1090245174752216,
-      "grad_norm": 0.32329856554850966,
+      "grad_norm": 0.3243862816608098,
       "learning_rate": 1.4941332352166385e-05,
-      "loss": 0.2488,
+      "loss": 0.2487,
       "step": 745
     },
     {
       "epoch": 3.1131977047470007,
-      "grad_norm": 0.3562925136749018,
+      "grad_norm": 0.35705004169593396,
       "learning_rate": 1.48848063195071e-05,
-      "loss": 0.245,
+      "loss": 0.2456,
       "step": 746
     },
     {
       "epoch": 3.1173708920187795,
-      "grad_norm": 0.33711870133757776,
+      "grad_norm": 0.3349788528606525,
       "learning_rate": 1.4828323973080054e-05,
-      "loss": 0.2503,
+      "loss": 0.2507,
       "step": 747
     },
     {
       "epoch": 3.121544079290558,
-      "grad_norm": 0.32013275126552065,
+      "grad_norm": 0.3304516234268912,
       "learning_rate": 1.4771885795271847e-05,
-      "loss": 0.243,
+      "loss": 0.2433,
       "step": 748
     },
     {
       "epoch": 3.125717266562337,
-      "grad_norm": 0.33519543712945055,
+      "grad_norm": 0.335700045866963,
       "learning_rate": 1.4715492268091858e-05,
-      "loss": 0.2688,
+      "loss": 0.269,
       "step": 749
     },
     {
       "epoch": 3.1298904538341157,
-      "grad_norm": 0.3225865788302496,
+      "grad_norm": 0.3381215085148102,
       "learning_rate": 1.4659143873168145e-05,
-      "loss": 0.2436,
+      "loss": 0.244,
       "step": 750
     },
     {
       "epoch": 3.134063641105895,
-      "grad_norm": 0.32537796487370424,
+      "grad_norm": 0.3256655330801169,
       "learning_rate": 1.460284109174329e-05,
-      "loss": 0.2528,
+      "loss": 0.2534,
       "step": 751
     },
     {
       "epoch": 3.1382368283776736,
-      "grad_norm": 0.31679090347314337,
+      "grad_norm": 0.31725777538602434,
       "learning_rate": 1.4546584404670323e-05,
-      "loss": 0.2545,
+      "loss": 0.2547,
       "step": 752
     },
     {
       "epoch": 3.1424100156494523,
-      "grad_norm": 0.3311844682186977,
+      "grad_norm": 0.3268310660496934,
       "learning_rate": 1.4490374292408613e-05,
-      "loss": 0.2685,
+      "loss": 0.2689,
       "step": 753
     },
     {
       "epoch": 3.146583202921231,
-      "grad_norm": 0.3355317761703804,
+      "grad_norm": 0.32642894007776535,
       "learning_rate": 1.443421123501974e-05,
-      "loss": 0.2369,
+      "loss": 0.2372,
       "step": 754
     },
     {
       "epoch": 3.15075639019301,
-      "grad_norm": 0.32924464333994147,
+      "grad_norm": 0.3389101601620327,
       "learning_rate": 1.4378095712163439e-05,
-      "loss": 0.2587,
+      "loss": 0.2592,
       "step": 755
     },
     {
       "epoch": 3.1549295774647885,
-      "grad_norm": 0.3192118579909607,
+      "grad_norm": 0.32134013970981284,
       "learning_rate": 1.4322028203093452e-05,
-      "loss": 0.2496,
+      "loss": 0.2493,
       "step": 756
     },
     {
       "epoch": 3.1591027647365677,
-      "grad_norm": 0.32199497615175726,
+      "grad_norm": 0.31501391815154517,
       "learning_rate": 1.4266009186653471e-05,
-      "loss": 0.2697,
+      "loss": 0.27,
       "step": 757
     },
     {
       "epoch": 3.1632759520083464,
-      "grad_norm": 0.33148871487709286,
+      "grad_norm": 0.33051844643050865,
       "learning_rate": 1.4210039141273032e-05,
-      "loss": 0.2434,
+      "loss": 0.2435,
       "step": 758
     },
     {
       "epoch": 3.167449139280125,
-      "grad_norm": 0.3572813101233042,
+      "grad_norm": 0.35611600865233356,
       "learning_rate": 1.4154118544963453e-05,
-      "loss": 0.2594,
+      "loss": 0.2598,
       "step": 759
     },
     {
       "epoch": 3.171622326551904,
-      "grad_norm": 0.33173979012914717,
+      "grad_norm": 0.33297415229719857,
       "learning_rate": 1.409824787531371e-05,
-      "loss": 0.262,
+      "loss": 0.2625,
       "step": 760
     },
     {
       "epoch": 3.1757955138236826,
-      "grad_norm": 0.3464947728926165,
+      "grad_norm": 0.34357274299478124,
       "learning_rate": 1.4042427609486393e-05,
-      "loss": 0.2796,
+      "loss": 0.2793,
       "step": 761
     },
     {
       "epoch": 3.179968701095462,
-      "grad_norm": 0.3207909630983132,
+      "grad_norm": 0.31615649095768844,
       "learning_rate": 1.3986658224213603e-05,
-      "loss": 0.2696,
+      "loss": 0.2694,
       "step": 762
     },
     {
       "epoch": 3.1841418883672405,
-      "grad_norm": 0.31766114878194496,
+      "grad_norm": 0.3282646573281619,
       "learning_rate": 1.393094019579293e-05,
-      "loss": 0.251,
+      "loss": 0.2511,
       "step": 763
     },
     {
       "epoch": 3.1883150756390193,
-      "grad_norm": 0.3385740372991529,
+      "grad_norm": 0.33007472837118584,
       "learning_rate": 1.3875274000083319e-05,
-      "loss": 0.2659,
+      "loss": 0.266,
       "step": 764
     },
     {
       "epoch": 3.192488262910798,
-      "grad_norm": 0.3270262691544981,
+      "grad_norm": 0.328498574219697,
       "learning_rate": 1.3819660112501054e-05,
-      "loss": 0.243,
+      "loss": 0.2433,
       "step": 765
     },
     {
       "epoch": 3.1966614501825767,
-      "grad_norm": 0.32704090692557697,
+      "grad_norm": 0.3337207350822446,
       "learning_rate": 1.3764099008015676e-05,
-      "loss": 0.2536,
+      "loss": 0.2539,
       "step": 766
     },
     {
       "epoch": 3.200834637454356,
-      "grad_norm": 0.3641585008678486,
+      "grad_norm": 0.3528586583031547,
       "learning_rate": 1.3708591161145934e-05,
-      "loss": 0.271,
+      "loss": 0.2716,
       "step": 767
     },
     {
       "epoch": 3.2050078247261347,
-      "grad_norm": 0.3110486924824008,
+      "grad_norm": 0.3171876905634677,
       "learning_rate": 1.365313704595574e-05,
-      "loss": 0.2381,
+      "loss": 0.2387,
       "step": 768
     },
     {
       "epoch": 3.2091810119979134,
-      "grad_norm": 0.3313720011413826,
+      "grad_norm": 0.3311721280594992,
       "learning_rate": 1.35977371360501e-05,
-      "loss": 0.2682,
+      "loss": 0.2683,
       "step": 769
     },
     {
       "epoch": 3.213354199269692,
-      "grad_norm": 0.3221158366107937,
+      "grad_norm": 0.3221570821215546,
       "learning_rate": 1.3542391904571082e-05,
-      "loss": 0.2505,
+      "loss": 0.2503,
       "step": 770
     },
     {
       "epoch": 3.217527386541471,
-      "grad_norm": 0.31193137331402243,
+      "grad_norm": 0.3199117078470472,
       "learning_rate": 1.348710182419377e-05,
-      "loss": 0.2477,
+      "loss": 0.2478,
       "step": 771
     },
     {
       "epoch": 3.22170057381325,
-      "grad_norm": 0.2973761734197825,
+      "grad_norm": 0.30158583192093713,
       "learning_rate": 1.3431867367122257e-05,
-      "loss": 0.2512,
+      "loss": 0.2511,
       "step": 772
     },
     {
       "epoch": 3.2258737610850288,
-      "grad_norm": 0.3189212677734855,
+      "grad_norm": 0.33027917961549247,
       "learning_rate": 1.3376689005085557e-05,
       "loss": 0.2474,
       "step": 773
     },
     {
       "epoch": 3.2300469483568075,
-      "grad_norm": 0.3110197069919759,
+      "grad_norm": 0.3066043079770959,
       "learning_rate": 1.332156720933361e-05,
-      "loss": 0.2669,
+      "loss": 0.2664,
       "step": 774
     },
     {
       "epoch": 3.2342201356285862,
-      "grad_norm": 0.32876648568812405,
+      "grad_norm": 0.3378549034438447,
       "learning_rate": 1.326650245063326e-05,
-      "loss": 0.2325,
+      "loss": 0.233,
       "step": 775
     },
     {
       "epoch": 3.238393322900365,
-      "grad_norm": 0.32525628709767784,
+      "grad_norm": 0.3334807902839644,
       "learning_rate": 1.3211495199264216e-05,
-      "loss": 0.2466,
+      "loss": 0.247,
       "step": 776
     },
     {
       "epoch": 3.242566510172144,
-      "grad_norm": 0.28820243178156824,
+      "grad_norm": 0.2906216257268021,
       "learning_rate": 1.3156545925015074e-05,
-      "loss": 0.2661,
+      "loss": 0.267,
       "step": 777
     },
     {
       "epoch": 3.246739697443923,
-      "grad_norm": 0.31609890961186576,
+      "grad_norm": 0.3199180246345434,
       "learning_rate": 1.3101655097179245e-05,
-      "loss": 0.2577,
+      "loss": 0.2578,
       "step": 778
     },
     {
       "epoch": 3.2509128847157016,
-      "grad_norm": 0.30747121984408027,
+      "grad_norm": 0.31565112170246745,
       "learning_rate": 1.3046823184550995e-05,
-      "loss": 0.2511,
+      "loss": 0.2517,
       "step": 779
     },
     {
       "epoch": 3.2550860719874803,
-      "grad_norm": 0.3085097222966809,
+      "grad_norm": 0.3098474716428039,
       "learning_rate": 1.2992050655421413e-05,
-      "loss": 0.2451,
+      "loss": 0.2453,
       "step": 780
     },
     {
       "epoch": 3.259259259259259,
-      "grad_norm": 0.31164522705596864,
+      "grad_norm": 0.316701535032436,
       "learning_rate": 1.2937337977574452e-05,
       "loss": 0.2548,
       "step": 781
     },
     {
       "epoch": 3.2634324465310383,
-      "grad_norm": 0.29814716927246426,
+      "grad_norm": 0.30545220368681264,
       "learning_rate": 1.2882685618282876e-05,
-      "loss": 0.2778,
+      "loss": 0.2786,
       "step": 782
     },
     {
       "epoch": 3.267605633802817,
-      "grad_norm": 0.32040417715958214,
+      "grad_norm": 0.32144323643565115,
       "learning_rate": 1.2828094044304316e-05,
-      "loss": 0.2457,
+      "loss": 0.2458,
       "step": 783
     },
     {
       "epoch": 3.2717788210745957,
-      "grad_norm": 0.30971484340787103,
+      "grad_norm": 0.31174684216433907,
       "learning_rate": 1.2773563721877258e-05,
-      "loss": 0.2757,
+      "loss": 0.2759,
       "step": 784
     },
     {
       "epoch": 3.2759520083463745,
-      "grad_norm": 0.29409144870312615,
+      "grad_norm": 0.2986060398202262,
       "learning_rate": 1.2719095116717069e-05,
-      "loss": 0.2485,
+      "loss": 0.2489,
       "step": 785
     },
     {
       "epoch": 3.280125195618153,
-      "grad_norm": 0.3108890886758001,
+      "grad_norm": 0.31744552825420225,
       "learning_rate": 1.2664688694012042e-05,
-      "loss": 0.247,
+      "loss": 0.2476,
       "step": 786
     },
     {
       "epoch": 3.2842983828899324,
-      "grad_norm": 0.31688866741473,
+      "grad_norm": 0.3204382587682495,
       "learning_rate": 1.2610344918419379e-05,
-      "loss": 0.2435,
+      "loss": 0.2438,
       "step": 787
     },
     {
       "epoch": 3.288471570161711,
-      "grad_norm": 0.3171386346072683,
+      "grad_norm": 0.3164242121596496,
       "learning_rate": 1.2556064254061249e-05,
       "loss": 0.2496,
       "step": 788
     },
     {
       "epoch": 3.29264475743349,
-      "grad_norm": 0.31651782205580503,
+      "grad_norm": 0.31970221714351194,
       "learning_rate": 1.2501847164520817e-05,
-      "loss": 0.2431,
+      "loss": 0.2433,
       "step": 789
     },
     {
       "epoch": 3.2968179447052686,
-      "grad_norm": 0.33052916122122644,
+      "grad_norm": 0.33970975361285016,
       "learning_rate": 1.2447694112838309e-05,
-      "loss": 0.2542,
+      "loss": 0.2544,
       "step": 790
     },
     {
       "epoch": 3.3009911319770473,
-      "grad_norm": 0.3289971887712876,
+      "grad_norm": 0.3356217282846709,
       "learning_rate": 1.239360556150702e-05,
-      "loss": 0.2504,
+      "loss": 0.251,
       "step": 791
     },
     {
       "epoch": 3.3051643192488265,
-      "grad_norm": 0.31895228646024315,
+      "grad_norm": 0.31848578051959753,
       "learning_rate": 1.2339581972469375e-05,
-      "loss": 0.2643,
+      "loss": 0.2644,
       "step": 792
     },
     {
       "epoch": 3.309337506520605,
-      "grad_norm": 0.3312841429746144,
+      "grad_norm": 0.3430721178039291,
       "learning_rate": 1.2285623807113e-05,
-      "loss": 0.246,
+      "loss": 0.2462,
       "step": 793
     },
     {
       "epoch": 3.313510693792384,
-      "grad_norm": 0.3068580419068359,
+      "grad_norm": 0.2990482039304521,
       "learning_rate": 1.2231731526266774e-05,
-      "loss": 0.2747,
+      "loss": 0.275,
       "step": 794
     },
     {
       "epoch": 3.3176838810641627,
-      "grad_norm": 0.3049174621485876,
+      "grad_norm": 0.3028941533902785,
       "learning_rate": 1.2177905590196884e-05,
-      "loss": 0.2347,
+      "loss": 0.2346,
       "step": 795
     },
     {
       "epoch": 3.3218570683359414,
-      "grad_norm": 0.3541760609482108,
+      "grad_norm": 0.35050911270905205,
       "learning_rate": 1.2124146458602896e-05,
-      "loss": 0.2578,
+      "loss": 0.2577,
       "step": 796
     },
     {
       "epoch": 3.3260302556077206,
-      "grad_norm": 0.31221904666219147,
+      "grad_norm": 0.31276218389393345,
       "learning_rate": 1.2070454590613844e-05,
-      "loss": 0.2586,
+      "loss": 0.2589,
       "step": 797
     },
     {
       "epoch": 3.3302034428794993,
-      "grad_norm": 0.31165803460654173,
+      "grad_norm": 0.3173025523989797,
       "learning_rate": 1.2016830444784287e-05,
-      "loss": 0.2514,
+      "loss": 0.2517,
       "step": 798
     },
     {
       "epoch": 3.334376630151278,
-      "grad_norm": 0.32995373849004084,
+      "grad_norm": 0.3247254897475166,
       "learning_rate": 1.1963274479090417e-05,
-      "loss": 0.2687,
+      "loss": 0.2688,
       "step": 799
     },
     {
       "epoch": 3.338549817423057,
-      "grad_norm": 0.32515635028869433,
+      "grad_norm": 0.3259347977969213,
       "learning_rate": 1.1909787150926128e-05,
-      "loss": 0.2511,
+      "loss": 0.2515,
       "step": 800
     },
     {
       "epoch": 3.3427230046948355,
-      "grad_norm": 0.34084979276726957,
+      "grad_norm": 0.34783686717059725,
       "learning_rate": 1.1856368917099114e-05,
-      "loss": 0.2521,
+      "loss": 0.2529,
       "step": 801
     },
     {
       "epoch": 3.3468961919666143,
-      "grad_norm": 0.3223473769469847,
+      "grad_norm": 0.32297871799337824,
       "learning_rate": 1.1803020233826965e-05,
-      "loss": 0.2494,
+      "loss": 0.2498,
       "step": 802
     },
     {
       "epoch": 3.3510693792383934,
-      "grad_norm": 0.31738703532444945,
+      "grad_norm": 0.3157105091786366,
       "learning_rate": 1.1749741556733297e-05,
       "loss": 0.2563,
       "step": 803
     },
     {
       "epoch": 3.355242566510172,
-      "grad_norm": 0.32972874498288685,
+      "grad_norm": 0.32304327391028914,
       "learning_rate": 1.1696533340843814e-05,
-      "loss": 0.2668,
+      "loss": 0.2673,
       "step": 804
     },
     {
       "epoch": 3.359415753781951,
-      "grad_norm": 0.32293486734095256,
+      "grad_norm": 0.31726988435390535,
       "learning_rate": 1.1643396040582468e-05,
       "loss": 0.2601,
       "step": 805
     },
     {
       "epoch": 3.3635889410537296,
-      "grad_norm": 0.313311708245681,
+      "grad_norm": 0.3109332198248973,
       "learning_rate": 1.1590330109767532e-05,
       "loss": 0.2564,
       "step": 806
     },
     {
       "epoch": 3.3677621283255084,
-      "grad_norm": 0.3213973269329586,
+      "grad_norm": 0.3198071135847881,
       "learning_rate": 1.153733600160777e-05,
-      "loss": 0.2583,
+      "loss": 0.2585,
       "step": 807
     },
     {
       "epoch": 3.3719353155972875,
-      "grad_norm": 0.29782294598761205,
+      "grad_norm": 0.29952673612449326,
       "learning_rate": 1.1484414168698547e-05,
-      "loss": 0.249,
+      "loss": 0.2491,
       "step": 808
     },
     {
       "epoch": 3.3761085028690663,
-      "grad_norm": 0.3089650611384535,
+      "grad_norm": 0.30712519121221393,
       "learning_rate": 1.1431565063017955e-05,
-      "loss": 0.2751,
+      "loss": 0.2753,
       "step": 809
     },
     {
       "epoch": 3.380281690140845,
-      "grad_norm": 0.3212185775874604,
+      "grad_norm": 0.3256570665352547,
       "learning_rate": 1.1378789135922954e-05,
-      "loss": 0.2468,
+      "loss": 0.2472,
       "step": 810
     },
     {
       "epoch": 3.3844548774126237,
-      "grad_norm": 0.3025013718875849,
+      "grad_norm": 0.30328499114193014,
       "learning_rate": 1.1326086838145528e-05,
-      "loss": 0.2552,
+      "loss": 0.2555,
       "step": 811
     },
     {
       "epoch": 3.3886280646844025,
-      "grad_norm": 0.3019904961961552,
+      "grad_norm": 0.3088893331332729,
       "learning_rate": 1.1273458619788844e-05,
       "loss": 0.263,
       "step": 812
     },
     {
       "epoch": 3.3928012519561817,
-      "grad_norm": 0.3090758299846271,
+      "grad_norm": 0.3101396549312823,
       "learning_rate": 1.1220904930323386e-05,
-      "loss": 0.2637,
+      "loss": 0.2641,
       "step": 813
     },
     {
       "epoch": 3.3969744392279604,
-      "grad_norm": 0.3115453709349569,
+      "grad_norm": 0.31542239926457977,
       "learning_rate": 1.1168426218583109e-05,
-      "loss": 0.2592,
+      "loss": 0.2595,
       "step": 814
     },
     {
       "epoch": 3.401147626499739,
-      "grad_norm": 0.29957848725797487,
+      "grad_norm": 0.2996588008892878,
       "learning_rate": 1.1116022932761648e-05,
-      "loss": 0.2402,
+      "loss": 0.24,
       "step": 815
     },
     {
       "epoch": 3.405320813771518,
-      "grad_norm": 0.3106383285211354,
+      "grad_norm": 0.3089093541347566,
       "learning_rate": 1.1063695520408442e-05,
-      "loss": 0.269,
+      "loss": 0.2694,
       "step": 816
     },
     {
       "epoch": 3.4094940010432966,
-      "grad_norm": 0.3234289878385872,
+      "grad_norm": 0.3265853688957117,
       "learning_rate": 1.1011444428424946e-05,
-      "loss": 0.2269,
+      "loss": 0.2271,
       "step": 817
     },
     {
       "epoch": 3.4136671883150758,
-      "grad_norm": 0.32444113327466845,
+      "grad_norm": 0.3243195368411595,
       "learning_rate": 1.0959270103060814e-05,
-      "loss": 0.2577,
+      "loss": 0.258,
       "step": 818
     },
     {
       "epoch": 3.4178403755868545,
-      "grad_norm": 0.3179390623994093,
+      "grad_norm": 0.32016281708944566,
       "learning_rate": 1.0907172989910046e-05,
-      "loss": 0.2454,
+      "loss": 0.246,
       "step": 819
     },
     {
       "epoch": 3.4220135628586332,
-      "grad_norm": 0.3095578723861612,
+      "grad_norm": 0.3119370846876167,
       "learning_rate": 1.085515353390723e-05,
       "loss": 0.2468,
       "step": 820
     },
     {
       "epoch": 3.426186750130412,
-      "grad_norm": 0.32398117055987696,
+      "grad_norm": 0.32045250448186424,
       "learning_rate": 1.0803212179323727e-05,
-      "loss": 0.2564,
+      "loss": 0.2567,
       "step": 821
     },
     {
       "epoch": 3.4303599374021907,
-      "grad_norm": 0.31116891642930766,
+      "grad_norm": 0.3094360717821285,
       "learning_rate": 1.0751349369763882e-05,
-      "loss": 0.2667,
+      "loss": 0.267,
       "step": 822
     },
     {
       "epoch": 3.43453312467397,
-      "grad_norm": 0.3097124981653213,
+      "grad_norm": 0.3144065498045969,
       "learning_rate": 1.0699565548161191e-05,
-      "loss": 0.2612,
+      "loss": 0.261,
       "step": 823
     },
     {
       "epoch": 3.4387063119457486,
-      "grad_norm": 0.3096203655811954,
+      "grad_norm": 0.3052951215187014,
       "learning_rate": 1.06478611567746e-05,
-      "loss": 0.2741,
+      "loss": 0.2746,
       "step": 824
     },
     {
       "epoch": 3.4428794992175273,
-      "grad_norm": 0.31481003774700006,
+      "grad_norm": 0.3129757731227562,
       "learning_rate": 1.0596236637184631e-05,
-      "loss": 0.2616,
+      "loss": 0.2621,
       "step": 825
     },
     {
       "epoch": 3.447052686489306,
-      "grad_norm": 0.31995688840516984,
+      "grad_norm": 0.3225312623515532,
       "learning_rate": 1.0544692430289731e-05,
-      "loss": 0.2602,
+      "loss": 0.2608,
       "step": 826
     },
     {
       "epoch": 3.451225873761085,
-      "grad_norm": 0.3188413406227095,
+      "grad_norm": 0.31772919495779556,
       "learning_rate": 1.0493228976302374e-05,
-      "loss": 0.2475,
+      "loss": 0.2478,
       "step": 827
     },
     {
       "epoch": 3.455399061032864,
-      "grad_norm": 0.30400699291440647,
+      "grad_norm": 0.3005959548500748,
       "learning_rate": 1.0441846714745403e-05,
       "loss": 0.2465,
       "step": 828
     },
     {
       "epoch": 3.4595722483046427,
-      "grad_norm": 0.3085611747432212,
+      "grad_norm": 0.30185949691601704,
       "learning_rate": 1.0390546084448211e-05,
-      "loss": 0.2674,
+      "loss": 0.2668,
       "step": 829
     },
     {
       "epoch": 3.4637454355764214,
-      "grad_norm": 0.3032261469361198,
+      "grad_norm": 0.2974978807673779,
       "learning_rate": 1.0339327523543043e-05,
-      "loss": 0.2685,
+      "loss": 0.2687,
       "step": 830
     },
     {
       "epoch": 3.4679186228482,
-      "grad_norm": 0.3074011636071174,
+      "grad_norm": 0.30648592230561406,
       "learning_rate": 1.0288191469461233e-05,
-      "loss": 0.2589,
+      "loss": 0.2593,
       "step": 831
     },
     {
       "epoch": 3.472091810119979,
-      "grad_norm": 0.3148861047432021,
+      "grad_norm": 0.31022888689532385,
       "learning_rate": 1.0237138358929438e-05,
-      "loss": 0.2476,
+      "loss": 0.248,
       "step": 832
     },
     {
       "epoch": 3.476264997391758,
-      "grad_norm": 0.32661935104608963,
+      "grad_norm": 0.3253555639917242,
       "learning_rate": 1.0186168627965966e-05,
-      "loss": 0.2552,
+      "loss": 0.2555,
       "step": 833
     },
     {
       "epoch": 3.480438184663537,
-      "grad_norm": 0.30845490907328227,
+      "grad_norm": 0.30998143917116844,
       "learning_rate": 1.0135282711877014e-05,
-      "loss": 0.254,
+      "loss": 0.2548,
       "step": 834
     },
     {
       "epoch": 3.4846113719353156,
-      "grad_norm": 0.31303595767734477,
+      "grad_norm": 0.3127592971557124,
       "learning_rate": 1.0084481045252965e-05,
-      "loss": 0.2706,
+      "loss": 0.2704,
       "step": 835
     },
     {
       "epoch": 3.4887845592070943,
-      "grad_norm": 0.3245990180365307,
+      "grad_norm": 0.32775092117255356,
       "learning_rate": 1.0033764061964648e-05,
-      "loss": 0.2452,
+      "loss": 0.2458,
       "step": 836
     },
     {
       "epoch": 3.492957746478873,
-      "grad_norm": 0.3021640492431065,
+      "grad_norm": 0.30116258089882547,
       "learning_rate": 9.98313219515969e-06,
-      "loss": 0.252,
+      "loss": 0.2522,
       "step": 837
     },
     {
       "epoch": 3.497130933750652,
-      "grad_norm": 0.30913156949949,
+      "grad_norm": 0.30501192949123374,
       "learning_rate": 9.932585877258738e-06,
-      "loss": 0.2583,
+      "loss": 0.2585,
       "step": 838
     },
     {
       "epoch": 3.501304121022431,
-      "grad_norm": 0.29861468542397096,
+      "grad_norm": 0.2980555826357034,
       "learning_rate": 9.882125539951865e-06,
       "loss": 0.2693,
       "step": 839
     },
     {
       "epoch": 3.5054773082942097,
-      "grad_norm": 0.3033654326350028,
+      "grad_norm": 0.30000712565815313,
       "learning_rate": 9.83175161419478e-06,
-      "loss": 0.2403,
+      "loss": 0.2399,
       "step": 840
     },
     {
       "epoch": 3.5096504955659884,
-      "grad_norm": 0.30821788568827496,
+      "grad_norm": 0.3095693958773187,
       "learning_rate": 9.781464530205226e-06,
-      "loss": 0.2614,
+      "loss": 0.2617,
       "step": 841
     },
     {
       "epoch": 3.513823682837767,
-      "grad_norm": 0.3085050807908234,
+      "grad_norm": 0.3035031523758809,
       "learning_rate": 9.73126471745925e-06,
-      "loss": 0.2647,
+      "loss": 0.2645,
       "step": 842
     },
     {
       "epoch": 3.5179968701095463,
-      "grad_norm": 0.30711698306956864,
+      "grad_norm": 0.3064534314169243,
       "learning_rate": 9.68115260468758e-06,
-      "loss": 0.2522,
+      "loss": 0.2521,
       "step": 843
     },
     {
       "epoch": 3.522170057381325,
-      "grad_norm": 0.31139702711686695,
+      "grad_norm": 0.31251772095938013,
       "learning_rate": 9.631128619871945e-06,
-      "loss": 0.248,
+      "loss": 0.2486,
       "step": 844
     },
     {
       "epoch": 3.526343244653104,
-      "grad_norm": 0.3051785210138254,
+      "grad_norm": 0.3025393980816354,
       "learning_rate": 9.581193190241398e-06,
-      "loss": 0.2428,
+      "loss": 0.2427,
       "step": 845
     },
     {
       "epoch": 3.5305164319248825,
-      "grad_norm": 0.3133186913208833,
+      "grad_norm": 0.32338575567675193,
       "learning_rate": 9.531346742268713e-06,
-      "loss": 0.2629,
+      "loss": 0.2631,
       "step": 846
     },
     {
       "epoch": 3.5346896191966612,
-      "grad_norm": 0.3211343912388273,
+      "grad_norm": 0.3222782775328403,
       "learning_rate": 9.481589701666682e-06,
-      "loss": 0.2731,
+      "loss": 0.2733,
       "step": 847
     },
     {
       "epoch": 3.5388628064684404,
-      "grad_norm": 0.31205479487281734,
+      "grad_norm": 0.3083069296120743,
       "learning_rate": 9.431922493384571e-06,
-      "loss": 0.2579,
+      "loss": 0.2582,
       "step": 848
     },
     {
       "epoch": 3.543035993740219,
-      "grad_norm": 0.3060513520995886,
+      "grad_norm": 0.30576051763004,
       "learning_rate": 9.382345541604376e-06,
-      "loss": 0.2539,
+      "loss": 0.2542,
       "step": 849
     },
     {
       "epoch": 3.547209181011998,
-      "grad_norm": 0.2964895387639281,
+      "grad_norm": 0.29714023866688777,
       "learning_rate": 9.332859269737303e-06,
-      "loss": 0.2757,
+      "loss": 0.2756,
       "step": 850
     },
     {
       "epoch": 3.5513823682837766,
-      "grad_norm": 0.31968429766649603,
+      "grad_norm": 0.3138673356672547,
       "learning_rate": 9.283464100420064e-06,
-      "loss": 0.2547,
+      "loss": 0.2548,
       "step": 851
     },
     {
       "epoch": 3.5555555555555554,
-      "grad_norm": 0.30138695541443394,
+      "grad_norm": 0.304872284918605,
       "learning_rate": 9.234160455511365e-06,
-      "loss": 0.2377,
+      "loss": 0.2386,
       "step": 852
     },
     {
       "epoch": 3.5597287428273345,
-      "grad_norm": 0.33246282958603696,
+      "grad_norm": 0.3382823662767099,
       "learning_rate": 9.184948756088201e-06,
-      "loss": 0.2474,
+      "loss": 0.2479,
       "step": 853
     },
     {
       "epoch": 3.5639019300991133,
-      "grad_norm": 0.3110935914373256,
+      "grad_norm": 0.3122912023152245,
       "learning_rate": 9.135829422442337e-06,
-      "loss": 0.2636,
+      "loss": 0.2637,
       "step": 854
     },
     {
       "epoch": 3.568075117370892,
-      "grad_norm": 0.29896161594904347,
+      "grad_norm": 0.29950239611537693,
       "learning_rate": 9.08680287407667e-06,
       "loss": 0.2609,
       "step": 855
     },
     {
       "epoch": 3.5722483046426707,
-      "grad_norm": 0.3256706165197879,
+      "grad_norm": 0.3307081775364636,
       "learning_rate": 9.03786952970168e-06,
-      "loss": 0.2679,
+      "loss": 0.2686,
       "step": 856
     },
     {
       "epoch": 3.5764214919144495,
-      "grad_norm": 0.30957332636601476,
+      "grad_norm": 0.31033886930276733,
       "learning_rate": 8.989029807231844e-06,
-      "loss": 0.2548,
+      "loss": 0.2547,
       "step": 857
     },
     {
       "epoch": 3.5805946791862286,
-      "grad_norm": 0.28015999263763436,
+      "grad_norm": 0.28013945742799296,
       "learning_rate": 8.940284123782037e-06,
-      "loss": 0.2608,
+      "loss": 0.2612,
       "step": 858
     },
     {
       "epoch": 3.5847678664580074,
-      "grad_norm": 0.3093535209795861,
+      "grad_norm": 0.3118887724159943,
       "learning_rate": 8.891632895664027e-06,
-      "loss": 0.2588,
+      "loss": 0.2594,
       "step": 859
     },
     {
       "epoch": 3.588941053729786,
-      "grad_norm": 0.31551111173299723,
+      "grad_norm": 0.3073244130890546,
       "learning_rate": 8.843076538382853e-06,
-      "loss": 0.2568,
+      "loss": 0.2572,
       "step": 860
     },
     {
       "epoch": 3.593114241001565,
-      "grad_norm": 0.29630911626627326,
+      "grad_norm": 0.2974541989551312,
       "learning_rate": 8.794615466633356e-06,
-      "loss": 0.2556,
+      "loss": 0.2554,
       "step": 861
     },
     {
       "epoch": 3.5972874282733436,
-      "grad_norm": 0.30329264276010187,
+      "grad_norm": 0.30313572080619733,
       "learning_rate": 8.74625009429655e-06,
-      "loss": 0.2369,
+      "loss": 0.2375,
       "step": 862
     },
     {
       "epoch": 3.6014606155451228,
-      "grad_norm": 0.30180059597310016,
+      "grad_norm": 0.2990675630503644,
       "learning_rate": 8.697980834436159e-06,
-      "loss": 0.2606,
+      "loss": 0.2604,
       "step": 863
     },
     {
       "epoch": 3.6056338028169015,
-      "grad_norm": 0.29562350639103896,
+      "grad_norm": 0.29774101618536586,
       "learning_rate": 8.649808099295031e-06,
-      "loss": 0.27,
+      "loss": 0.2704,
       "step": 864
     },
     {
       "epoch": 3.6098069900886802,
-      "grad_norm": 0.31150949568960834,
+      "grad_norm": 0.3189497115827565,
       "learning_rate": 8.601732300291674e-06,
-      "loss": 0.2429,
+      "loss": 0.2433,
       "step": 865
     },
     {
       "epoch": 3.613980177360459,
-      "grad_norm": 0.31786995102206206,
+      "grad_norm": 0.32017488799849925,
       "learning_rate": 8.553753848016702e-06,
-      "loss": 0.2446,
+      "loss": 0.2451,
       "step": 866
     },
     {
       "epoch": 3.6181533646322377,
-      "grad_norm": 0.3131311590206814,
+      "grad_norm": 0.31732484054811516,
       "learning_rate": 8.50587315222935e-06,
-      "loss": 0.2244,
+      "loss": 0.2246,
       "step": 867
     },
     {
       "epoch": 3.622326551904017,
-      "grad_norm": 0.3044418986884301,
+      "grad_norm": 0.31364129353653386,
       "learning_rate": 8.458090621853938e-06,
-      "loss": 0.2569,
+      "loss": 0.2573,
       "step": 868
     },
     {
       "epoch": 3.6264997391757956,
-      "grad_norm": 0.311328231677092,
+      "grad_norm": 0.30865402238159906,
       "learning_rate": 8.410406664976439e-06,
-      "loss": 0.2471,
+      "loss": 0.2472,
       "step": 869
     },
     {
       "epoch": 3.6306729264475743,
-      "grad_norm": 0.299579680139395,
+      "grad_norm": 0.2997719333802799,
       "learning_rate": 8.362821688840947e-06,
-      "loss": 0.2587,
+      "loss": 0.2592,
       "step": 870
     },
     {
       "epoch": 3.634846113719353,
-      "grad_norm": 0.30664818732477195,
+      "grad_norm": 0.3105979013103038,
       "learning_rate": 8.315336099846202e-06,
-      "loss": 0.2578,
+      "loss": 0.2581,
       "step": 871
     },
     {
       "epoch": 3.639019300991132,
-      "grad_norm": 0.31153039899710067,
+      "grad_norm": 0.31569139205536234,
       "learning_rate": 8.267950303542155e-06,
-      "loss": 0.2521,
+      "loss": 0.2524,
       "step": 872
     },
     {
       "epoch": 3.643192488262911,
-      "grad_norm": 0.29809240043797547,
+      "grad_norm": 0.29693619555107714,
       "learning_rate": 8.220664704626442e-06,
       "loss": 0.2679,
       "step": 873
     },
     {
       "epoch": 3.6473656755346897,
-      "grad_norm": 0.30063197158882693,
+      "grad_norm": 0.3017010358938564,
       "learning_rate": 8.173479706941014e-06,
-      "loss": 0.2457,
+      "loss": 0.2462,
       "step": 874
     },
     {
       "epoch": 3.6515388628064684,
-      "grad_norm": 0.30493242512197255,
+      "grad_norm": 0.30596840879187864,
       "learning_rate": 8.12639571346859e-06,
       "loss": 0.2282,
       "step": 875
     },
     {
       "epoch": 3.655712050078247,
-      "grad_norm": 0.3199685553442984,
+      "grad_norm": 0.3227411889071714,
       "learning_rate": 8.079413126329299e-06,
-      "loss": 0.2407,
+      "loss": 0.2405,
       "step": 876
     },
     {
       "epoch": 3.659885237350026,
-      "grad_norm": 0.30926384861195966,
+      "grad_norm": 0.31363357551896404,
       "learning_rate": 8.032532346777185e-06,
-      "loss": 0.2404,
+      "loss": 0.241,
       "step": 877
     },
     {
       "epoch": 3.664058424621805,
-      "grad_norm": 0.31000013685834826,
+      "grad_norm": 0.3081752880565644,
       "learning_rate": 7.985753775196822e-06,
       "loss": 0.236,
       "step": 878
     },
     {
       "epoch": 3.668231611893584,
-      "grad_norm": 0.30683353649649425,
+      "grad_norm": 0.3011564668175278,
       "learning_rate": 7.93907781109988e-06,
-      "loss": 0.2559,
+      "loss": 0.2556,
       "step": 879
     },
     {
       "epoch": 3.6724047991653626,
-      "grad_norm": 0.306379001123491,
+      "grad_norm": 0.31059210149617456,
       "learning_rate": 7.89250485312171e-06,
-      "loss": 0.2318,
+      "loss": 0.2328,
       "step": 880
     },
     {
       "epoch": 3.6765779864371413,
-      "grad_norm": 0.311404639456788,
+      "grad_norm": 0.3077643689169864,
       "learning_rate": 7.846035299017921e-06,
-      "loss": 0.2703,
+      "loss": 0.2706,
       "step": 881
     },
     {
       "epoch": 3.68075117370892,
-      "grad_norm": 0.297645756161866,
+      "grad_norm": 0.2954461075441741,
       "learning_rate": 7.799669545661026e-06,
-      "loss": 0.242,
+      "loss": 0.2422,
       "step": 882
     },
     {
       "epoch": 3.684924360980699,
-      "grad_norm": 0.28795522767463466,
+      "grad_norm": 0.2878062694927721,
       "learning_rate": 7.753407989037032e-06,
-      "loss": 0.2599,
+      "loss": 0.2604,
       "step": 883
     },
     {
       "epoch": 3.689097548252478,
-      "grad_norm": 0.3051881440806298,
+      "grad_norm": 0.3064746657009289,
       "learning_rate": 7.707251024242031e-06,
-      "loss": 0.2618,
+      "loss": 0.2621,
       "step": 884
     },
     {
       "epoch": 3.6932707355242567,
-      "grad_norm": 0.3063015354261532,
+      "grad_norm": 0.30918419109746864,
       "learning_rate": 7.661199045478874e-06,
       "loss": 0.2604,
       "step": 885
     },
     {
       "epoch": 3.6974439227960354,
-      "grad_norm": 0.3028629836193355,
+      "grad_norm": 0.29872672713086823,
       "learning_rate": 7.615252446053761e-06,
-      "loss": 0.2454,
+      "loss": 0.2453,
       "step": 886
     },
     {
       "epoch": 3.701617110067814,
-      "grad_norm": 0.3056014751756775,
+      "grad_norm": 0.30601010801663875,
       "learning_rate": 7.569411618372917e-06,
-      "loss": 0.2423,
+      "loss": 0.2424,
       "step": 887
     },
     {
       "epoch": 3.7057902973395933,
-      "grad_norm": 0.2861959045023739,
+      "grad_norm": 0.28869643054084687,
       "learning_rate": 7.523676953939218e-06,
-      "loss": 0.2473,
+      "loss": 0.2477,
       "step": 888
     },
     {
       "epoch": 3.709963484611372,
-      "grad_norm": 0.2897891651741676,
+      "grad_norm": 0.29430825595661064,
       "learning_rate": 7.478048843348866e-06,
-      "loss": 0.2588,
+      "loss": 0.2593,
       "step": 889
     },
     {
       "epoch": 3.7141366718831508,
-      "grad_norm": 0.3019447461377478,
+      "grad_norm": 0.30080215945944977,
       "learning_rate": 7.432527676288015e-06,
-      "loss": 0.2565,
+      "loss": 0.2568,
       "step": 890
     },
     {
       "epoch": 3.7183098591549295,
-      "grad_norm": 0.31435900335011224,
+      "grad_norm": 0.308453744749871,
       "learning_rate": 7.387113841529494e-06,
       "loss": 0.2436,
       "step": 891
     },
     {
       "epoch": 3.7224830464267082,
-      "grad_norm": 0.29380895318463673,
+      "grad_norm": 0.296204725273969,
       "learning_rate": 7.341807726929453e-06,
-      "loss": 0.2637,
+      "loss": 0.264,
       "step": 892
     },
     {
       "epoch": 3.7266562336984874,
-      "grad_norm": 0.2949687156855656,
+      "grad_norm": 0.3039573649370887,
       "learning_rate": 7.296609719424068e-06,
-      "loss": 0.2774,
+      "loss": 0.2779,
       "step": 893
     },
     {
       "epoch": 3.730829420970266,
-      "grad_norm": 0.2890699335641608,
+      "grad_norm": 0.29396708548352424,
       "learning_rate": 7.251520205026206e-06,
-      "loss": 0.2501,
+      "loss": 0.2506,
       "step": 894
     },
     {
       "epoch": 3.735002608242045,
-      "grad_norm": 0.3022047796058328,
+      "grad_norm": 0.292485313099948,
       "learning_rate": 7.206539568822179e-06,
       "loss": 0.2719,
       "step": 895
     },
     {
       "epoch": 3.7391757955138236,
-      "grad_norm": 0.2957292666265662,
+      "grad_norm": 0.3002004041694919,
       "learning_rate": 7.161668194968401e-06,
-      "loss": 0.2372,
+      "loss": 0.2377,
       "step": 896
     },
     {
       "epoch": 3.7433489827856024,
-      "grad_norm": 0.2917917390280248,
+      "grad_norm": 0.296975365557123,
       "learning_rate": 7.116906466688147e-06,
-      "loss": 0.264,
+      "loss": 0.2643,
       "step": 897
     },
     {
       "epoch": 3.7475221700573815,
-      "grad_norm": 0.2981171058077632,
+      "grad_norm": 0.3009961137903897,
       "learning_rate": 7.072254766268269e-06,
-      "loss": 0.2329,
+      "loss": 0.2326,
       "step": 898
     },
     {
       "epoch": 3.7516953573291603,
-      "grad_norm": 0.286569857987241,
+      "grad_norm": 0.2858084273657275,
       "learning_rate": 7.027713475055913e-06,
-      "loss": 0.2528,
+      "loss": 0.2527,
       "step": 899
     },
     {
       "epoch": 3.755868544600939,
-      "grad_norm": 0.30334166530372547,
+      "grad_norm": 0.31066204884199383,
       "learning_rate": 6.98328297345529e-06,
-      "loss": 0.2665,
+      "loss": 0.2667,
       "step": 900
     },
     {
       "epoch": 3.7600417318727177,
-      "grad_norm": 0.3012115692785666,
+      "grad_norm": 0.3038925333441486,
       "learning_rate": 6.938963640924405e-06,
-      "loss": 0.2513,
+      "loss": 0.2515,
       "step": 901
     },
     {
       "epoch": 3.7642149191444965,
-      "grad_norm": 0.2868475495276251,
+      "grad_norm": 0.2918757147545174,
       "learning_rate": 6.894755855971842e-06,
-      "loss": 0.2588,
+      "loss": 0.2591,
       "step": 902
     },
     {
       "epoch": 3.7683881064162756,
-      "grad_norm": 0.30931384969857917,
+      "grad_norm": 0.31713171076347063,
       "learning_rate": 6.850659996153488e-06,
       "loss": 0.2738,
       "step": 903
     },
     {
       "epoch": 3.7725612936880544,
-      "grad_norm": 0.3032270869020924,
+      "grad_norm": 0.30845941868132654,
       "learning_rate": 6.806676438069359e-06,
-      "loss": 0.2344,
+      "loss": 0.2346,
       "step": 904
     },
     {
       "epoch": 3.776734480959833,
-      "grad_norm": 0.29790159449460646,
+      "grad_norm": 0.30416104863074106,
       "learning_rate": 6.762805557360335e-06,
-      "loss": 0.262,
+      "loss": 0.2622,
       "step": 905
     },
     {
       "epoch": 3.780907668231612,
-      "grad_norm": 0.287164811679991,
+      "grad_norm": 0.28628447605142154,
       "learning_rate": 6.7190477287050125e-06,
-      "loss": 0.2431,
+      "loss": 0.2434,
       "step": 906
     },
     {
       "epoch": 3.7850808555033906,
-      "grad_norm": 0.3110743270083782,
+      "grad_norm": 0.31708515359385525,
       "learning_rate": 6.6754033258164345e-06,
-      "loss": 0.2437,
+      "loss": 0.244,
       "step": 907
     },
     {
       "epoch": 3.7892540427751698,
-      "grad_norm": 0.30847394225374686,
+      "grad_norm": 0.30937397225135976,
       "learning_rate": 6.631872721438957e-06,
-      "loss": 0.2526,
+      "loss": 0.2531,
       "step": 908
     },
     {
       "epoch": 3.7934272300469485,
-      "grad_norm": 0.3178435094276417,
+      "grad_norm": 0.3123557646815316,
       "learning_rate": 6.588456287345015e-06,
       "loss": 0.2804,
       "step": 909
     },
     {
       "epoch": 3.797600417318727,
-      "grad_norm": 0.31775449662539434,
+      "grad_norm": 0.31796748210059,
       "learning_rate": 6.5451543943320005e-06,
-      "loss": 0.2421,
+      "loss": 0.2423,
       "step": 910
     },
     {
       "epoch": 3.801773604590506,
-      "grad_norm": 0.2856937870147108,
+      "grad_norm": 0.2900052838473373,
       "learning_rate": 6.501967412219065e-06,
-      "loss": 0.2724,
+      "loss": 0.2729,
       "step": 911
     },
     {
       "epoch": 3.8059467918622847,
-      "grad_norm": 0.3031870314583201,
+      "grad_norm": 0.30189649747499153,
       "learning_rate": 6.458895709843946e-06,
-      "loss": 0.2631,
+      "loss": 0.2628,
       "step": 912
     },
     {
       "epoch": 3.810119979134064,
-      "grad_norm": 0.3274159418043534,
+      "grad_norm": 0.32663285639887857,
       "learning_rate": 6.415939655059853e-06,
-      "loss": 0.2373,
+      "loss": 0.238,
       "step": 913
     },
     {
       "epoch": 3.8142931664058426,
-      "grad_norm": 0.30508026537185046,
+      "grad_norm": 0.30722695990602306,
       "learning_rate": 6.373099614732308e-06,
-      "loss": 0.2547,
+      "loss": 0.2551,
       "step": 914
     },
     {
       "epoch": 3.8184663536776213,
-      "grad_norm": 0.3037316120350648,
+      "grad_norm": 0.30696313150503113,
       "learning_rate": 6.330375954736014e-06,
-      "loss": 0.2465,
+      "loss": 0.2469,
       "step": 915
     },
     {
       "epoch": 3.8226395409494,
-      "grad_norm": 0.28916630362046175,
+      "grad_norm": 0.28759294364070026,
       "learning_rate": 6.287769039951716e-06,
       "loss": 0.2589,
       "step": 916
     },
     {
       "epoch": 3.826812728221179,
-      "grad_norm": 0.2918698971337681,
+      "grad_norm": 0.29087888453494115,
       "learning_rate": 6.2452792342631215e-06,
       "loss": 0.2623,
       "step": 917
     },
     {
       "epoch": 3.830985915492958,
-      "grad_norm": 0.3104249757568547,
+      "grad_norm": 0.31082673804909583,
       "learning_rate": 6.202906900553734e-06,
-      "loss": 0.2586,
+      "loss": 0.2589,
       "step": 918
     },
     {
       "epoch": 3.8351591027647367,
-      "grad_norm": 0.298280903222913,
+      "grad_norm": 0.29651912953715326,
       "learning_rate": 6.1606524007038415e-06,
-      "loss": 0.2402,
+      "loss": 0.2403,
       "step": 919
     },
     {
       "epoch": 3.8393322900365154,
-      "grad_norm": 0.2989609713540234,
+      "grad_norm": 0.2953670007970378,
       "learning_rate": 6.118516095587321e-06,
-      "loss": 0.2474,
+      "loss": 0.2481,
       "step": 920
     },
     {
       "epoch": 3.843505477308294,
-      "grad_norm": 0.29187173619899925,
+      "grad_norm": 0.2910434366187144,
       "learning_rate": 6.076498345068651e-06,
-      "loss": 0.251,
+      "loss": 0.2509,
       "step": 921
     },
     {
       "epoch": 3.847678664580073,
-      "grad_norm": 0.3208842024162722,
+      "grad_norm": 0.31478790999540485,
       "learning_rate": 6.034599507999759e-06,
       "loss": 0.2528,
       "step": 922
     },
     {
       "epoch": 3.851851851851852,
-      "grad_norm": 0.29155933444525006,
+      "grad_norm": 0.2975730471964458,
       "learning_rate": 5.992819942217021e-06,
-      "loss": 0.2408,
+      "loss": 0.2412,
       "step": 923
     },
     {
       "epoch": 3.856025039123631,
-      "grad_norm": 0.2931391199461599,
+      "grad_norm": 0.29130230493390313,
       "learning_rate": 5.951160004538177e-06,
-      "loss": 0.2656,
+      "loss": 0.2657,
       "step": 924
     },
     {
       "epoch": 3.8601982263954095,
-      "grad_norm": 0.2890709293701794,
+      "grad_norm": 0.2868217891028275,
       "learning_rate": 5.909620050759266e-06,
-      "loss": 0.2478,
+      "loss": 0.2477,
       "step": 925
     },
     {
       "epoch": 3.8643714136671883,
-      "grad_norm": 0.28850290392530703,
+      "grad_norm": 0.28670555701974704,
       "learning_rate": 5.86820043565163e-06,
-      "loss": 0.2488,
+      "loss": 0.2494,
       "step": 926
     },
     {
       "epoch": 3.868544600938967,
-      "grad_norm": 0.29292013357206964,
+      "grad_norm": 0.2958275160087685,
       "learning_rate": 5.826901512958829e-06,
-      "loss": 0.2536,
+      "loss": 0.2543,
       "step": 927
     },
     {
       "epoch": 3.872717788210746,
-      "grad_norm": 0.2861077199338979,
+      "grad_norm": 0.2899640450927271,
       "learning_rate": 5.785723635393701e-06,
-      "loss": 0.2499,
+      "loss": 0.2504,
       "step": 928
     },
     {
       "epoch": 3.876890975482525,
-      "grad_norm": 0.28368208519959087,
+      "grad_norm": 0.2855581312162716,
       "learning_rate": 5.744667154635246e-06,
-      "loss": 0.2556,
+      "loss": 0.256,
       "step": 929
     },
     {
       "epoch": 3.8810641627543037,
-      "grad_norm": 0.3039811077489683,
+      "grad_norm": 0.3033108668698368,
       "learning_rate": 5.703732421325716e-06,
       "loss": 0.2174,
       "step": 930
     },
     {
       "epoch": 3.8852373500260824,
-      "grad_norm": 0.30675867166521786,
+      "grad_norm": 0.3054322643267081,
       "learning_rate": 5.662919785067549e-06,
-      "loss": 0.2649,
+      "loss": 0.2645,
       "step": 931
     },
     {
       "epoch": 3.889410537297861,
-      "grad_norm": 0.2987933892784774,
+      "grad_norm": 0.2941465773751956,
       "learning_rate": 5.62222959442045e-06,
-      "loss": 0.2688,
+      "loss": 0.2682,
       "step": 932
     },
     {
       "epoch": 3.8935837245696403,
-      "grad_norm": 0.28992010840503735,
+      "grad_norm": 0.2938497077534761,
       "learning_rate": 5.581662196898348e-06,
-      "loss": 0.2508,
+      "loss": 0.2516,
       "step": 933
     },
     {
       "epoch": 3.897756911841419,
-      "grad_norm": 0.29766531122947393,
+      "grad_norm": 0.30333810015064977,
       "learning_rate": 5.541217938966483e-06,
-      "loss": 0.2526,
+      "loss": 0.2531,
       "step": 934
     },
     {
       "epoch": 3.9019300991131978,
-      "grad_norm": 0.2877865459089933,
+      "grad_norm": 0.29259581491625475,
       "learning_rate": 5.500897166038397e-06,
-      "loss": 0.2586,
+      "loss": 0.2591,
       "step": 935
     },
     {
       "epoch": 3.9061032863849765,
-      "grad_norm": 0.28823211520061287,
+      "grad_norm": 0.2914176726676219,
       "learning_rate": 5.460700222473034e-06,
-      "loss": 0.2392,
+      "loss": 0.2395,
       "step": 936
     },
     {
       "epoch": 3.9102764736567552,
-      "grad_norm": 0.2983311334647983,
+      "grad_norm": 0.29997033936078726,
       "learning_rate": 5.4206274515717735e-06,
-      "loss": 0.231,
+      "loss": 0.2309,
       "step": 937
     },
     {
       "epoch": 3.9144496609285344,
-      "grad_norm": 0.3033632675646967,
+      "grad_norm": 0.3037409326118684,
       "learning_rate": 5.38067919557548e-06,
-      "loss": 0.2465,
+      "loss": 0.2461,
       "step": 938
     },
     {
       "epoch": 3.918622848200313,
-      "grad_norm": 0.2885480231332167,
+      "grad_norm": 0.28870240088342475,
       "learning_rate": 5.340855795661626e-06,
-      "loss": 0.2553,
+      "loss": 0.2552,
       "step": 939
     },
     {
       "epoch": 3.922796035472092,
-      "grad_norm": 0.29206535574794484,
+      "grad_norm": 0.29495870444481387,
       "learning_rate": 5.3011575919413214e-06,
-      "loss": 0.2477,
+      "loss": 0.2478,
       "step": 940
     },
     {
       "epoch": 3.9269692227438706,
-      "grad_norm": 0.2957695834163288,
+      "grad_norm": 0.30099725844198094,
       "learning_rate": 5.261584923456484e-06,
-      "loss": 0.2459,
+      "loss": 0.2457,
       "step": 941
     },
     {
       "epoch": 3.9311424100156493,
-      "grad_norm": 0.3000585155116648,
+      "grad_norm": 0.29675395064811677,
       "learning_rate": 5.222138128176857e-06,
-      "loss": 0.231,
+      "loss": 0.2309,
       "step": 942
     },
     {
       "epoch": 3.9353155972874285,
-      "grad_norm": 0.3001707073082044,
+      "grad_norm": 0.2996679454406872,
       "learning_rate": 5.182817542997196e-06,
-      "loss": 0.2468,
+      "loss": 0.247,
       "step": 943
     },
     {
       "epoch": 3.9394887845592073,
-      "grad_norm": 0.2931022762295121,
+      "grad_norm": 0.29435560311722303,
       "learning_rate": 5.143623503734334e-06,
-      "loss": 0.2441,
+      "loss": 0.2442,
       "step": 944
     },
     {
       "epoch": 3.943661971830986,
-      "grad_norm": 0.3002361326287333,
+      "grad_norm": 0.29963782997617644,
       "learning_rate": 5.104556345124363e-06,
-      "loss": 0.2594,
+      "loss": 0.2593,
       "step": 945
     },
     {
       "epoch": 3.9478351591027647,
-      "grad_norm": 0.28492524977591194,
+      "grad_norm": 0.2836906726866129,
       "learning_rate": 5.065616400819746e-06,
-      "loss": 0.2509,
+      "loss": 0.2511,
       "step": 946
     },
     {
       "epoch": 3.9520083463745435,
-      "grad_norm": 0.27817987343729694,
+      "grad_norm": 0.2775586902619862,
       "learning_rate": 5.026804003386476e-06,
-      "loss": 0.2732,
+      "loss": 0.2735,
       "step": 947
     },
     {
       "epoch": 3.9561815336463226,
-      "grad_norm": 0.2752757137882095,
+      "grad_norm": 0.2788096414135683,
       "learning_rate": 4.988119484301219e-06,
-      "loss": 0.2689,
+      "loss": 0.2687,
       "step": 948
     },
     {
       "epoch": 3.960354720918101,
-      "grad_norm": 0.2923500934191291,
+      "grad_norm": 0.2915800562591813,
       "learning_rate": 4.949563173948517e-06,
-      "loss": 0.2594,
+      "loss": 0.26,
       "step": 949
     },
     {
       "epoch": 3.96452790818988,
-      "grad_norm": 0.286238720078625,
+      "grad_norm": 0.2843406238853573,
       "learning_rate": 4.911135401617948e-06,
-      "loss": 0.2532,
+      "loss": 0.2533,
       "step": 950
     },
     {
       "epoch": 3.968701095461659,
-      "grad_norm": 0.27455060497129585,
+      "grad_norm": 0.2755291615836495,
       "learning_rate": 4.872836495501292e-06,
       "loss": 0.2723,
       "step": 951
     },
     {
       "epoch": 3.9728742827334376,
-      "grad_norm": 0.28732438251523545,
+      "grad_norm": 0.2876039066447679,
       "learning_rate": 4.83466678268977e-06,
-      "loss": 0.2858,
+      "loss": 0.2857,
       "step": 952
     },
     {
       "epoch": 3.9770474700052167,
-      "grad_norm": 0.2871654055811332,
+      "grad_norm": 0.29254187326182163,
       "learning_rate": 4.796626589171211e-06,
-      "loss": 0.2576,
+      "loss": 0.258,
       "step": 953
     },
     {
       "epoch": 3.981220657276995,
-      "grad_norm": 0.2941859867456557,
+      "grad_norm": 0.289909782329283,
       "learning_rate": 4.7587162398273166e-06,
-      "loss": 0.2501,
+      "loss": 0.2505,
       "step": 954
     },
     {
       "epoch": 3.985393844548774,
-      "grad_norm": 0.2928551079336585,
+      "grad_norm": 0.2920627761301855,
       "learning_rate": 4.720936058430818e-06,
-      "loss": 0.2499,
+      "loss": 0.2497,
       "step": 955
     },
     {
       "epoch": 3.989567031820553,
-      "grad_norm": 0.29169382461446075,
+      "grad_norm": 0.290587091624112,
       "learning_rate": 4.683286367642785e-06,
       "loss": 0.2362,
       "step": 956
     },
     {
       "epoch": 3.9937402190923317,
-      "grad_norm": 0.2975971114416749,
+      "grad_norm": 0.2959611062638297,
       "learning_rate": 4.6457674890098025e-06,
-      "loss": 0.255,
+      "loss": 0.2548,
       "step": 957
     },
     {
       "epoch": 3.997913406364111,
-      "grad_norm": 0.29150173639727905,
+      "grad_norm": 0.28970228986791585,
       "learning_rate": 4.6083797429612776e-06,
-      "loss": 0.2382,
+      "loss": 0.238,
       "step": 958
     },
     {
       "epoch": 4.002086593635889,
-      "grad_norm": 0.3003791491717585,
+      "grad_norm": 0.299746302161212,
       "learning_rate": 4.571123448806673e-06,
-      "loss": 0.2324,
+      "loss": 0.2326,
       "step": 959
     },
     {
       "epoch": 4.006259780907668,
-      "grad_norm": 0.44100741802727744,
+      "grad_norm": 0.4503113069476192,
       "learning_rate": 4.533998924732801e-06,
-      "loss": 0.1976,
+      "loss": 0.1978,
       "step": 960
     },
     {
       "epoch": 4.010432968179447,
-      "grad_norm": 0.32557334124644915,
+      "grad_norm": 0.326044564349573,
       "learning_rate": 4.497006487801065e-06,
-      "loss": 0.1845,
+      "loss": 0.1847,
       "step": 961
     },
     {
       "epoch": 4.014606155451226,
-      "grad_norm": 0.24606702636257016,
+      "grad_norm": 0.23780919248673996,
       "learning_rate": 4.460146453944813e-06,
-      "loss": 0.192,
+      "loss": 0.1923,
       "step": 962
     },
     {
       "epoch": 4.018779342723005,
-      "grad_norm": 0.3425209750457978,
+      "grad_norm": 0.33565718987012405,
       "learning_rate": 4.423419137966592e-06,
-      "loss": 0.1836,
+      "loss": 0.184,
       "step": 963
     },
     {
       "epoch": 4.022952529994783,
-      "grad_norm": 0.448734433800044,
+      "grad_norm": 0.45388852210961117,
       "learning_rate": 4.386824853535463e-06,
-      "loss": 0.1924,
+      "loss": 0.193,
       "step": 964
     },
     {
       "epoch": 4.027125717266562,
-      "grad_norm": 0.4487510228297213,
+      "grad_norm": 0.4467068340357761,
       "learning_rate": 4.350363913184355e-06,
-      "loss": 0.2184,
+      "loss": 0.2188,
       "step": 965
     },
     {
       "epoch": 4.031298904538341,
-      "grad_norm": 0.3235366305706013,
+      "grad_norm": 0.33108786548799907,
       "learning_rate": 4.314036628307354e-06,
-      "loss": 0.1844,
+      "loss": 0.1848,
       "step": 966
     },
     {
       "epoch": 4.03547209181012,
-      "grad_norm": 0.3012078934862842,
+      "grad_norm": 0.29726386008499117,
       "learning_rate": 4.277843309157079e-06,
-      "loss": 0.1849,
+      "loss": 0.1848,
       "step": 967
     },
     {
       "epoch": 4.039645279081899,
-      "grad_norm": 0.32582624878201105,
+      "grad_norm": 0.32687393782948204,
       "learning_rate": 4.241784264842009e-06,
       "loss": 0.1976,
       "step": 968
     },
     {
       "epoch": 4.043818466353677,
-      "grad_norm": 0.35673777953344143,
+      "grad_norm": 0.3504215296674512,
       "learning_rate": 4.205859803323857e-06,
       "loss": 0.2067,
       "step": 969
     },
     {
       "epoch": 4.0479916536254565,
-      "grad_norm": 0.33536202280082295,
+      "grad_norm": 0.32138280345270165,
       "learning_rate": 4.17007023141492e-06,
-      "loss": 0.1795,
+      "loss": 0.1797,
       "step": 970
     },
     {
       "epoch": 4.052164840897235,
-      "grad_norm": 0.28879719715909064,
+      "grad_norm": 0.2840539569796602,
       "learning_rate": 4.1344158547754865e-06,
-      "loss": 0.1861,
+      "loss": 0.1864,
       "step": 971
     },
     {
       "epoch": 4.056338028169014,
-      "grad_norm": 0.2972131928790304,
+      "grad_norm": 0.2899936434751284,
       "learning_rate": 4.098896977911204e-06,
-      "loss": 0.1959,
+      "loss": 0.1962,
       "step": 972
     },
     {
       "epoch": 4.060511215440793,
-      "grad_norm": 0.29468924174682326,
+      "grad_norm": 0.2856113204001228,
       "learning_rate": 4.063513904170497e-06,
-      "loss": 0.1935,
+      "loss": 0.1937,
       "step": 973
     },
     {
       "epoch": 4.0646844027125715,
-      "grad_norm": 0.27391972961618005,
+      "grad_norm": 0.2712111307205563,
       "learning_rate": 4.028266935741949e-06,
-      "loss": 0.1957,
+      "loss": 0.1959,
       "step": 974
     },
     {
       "epoch": 4.068857589984351,
-      "grad_norm": 0.29282964240422005,
+      "grad_norm": 0.29786487031413417,
       "learning_rate": 3.993156373651752e-06,
-      "loss": 0.1658,
+      "loss": 0.1661,
       "step": 975
     },
     {
       "epoch": 4.073030777256129,
-      "grad_norm": 0.28954297168194515,
+      "grad_norm": 0.2915572423563808,
       "learning_rate": 3.958182517761111e-06,
-      "loss": 0.1906,
+      "loss": 0.1909,
       "step": 976
     },
     {
       "epoch": 4.077203964527908,
-      "grad_norm": 0.268227711876174,
+      "grad_norm": 0.2687984199935904,
       "learning_rate": 3.923345666763702e-06,
       "loss": 0.2256,
       "step": 977
     },
     {
       "epoch": 4.081377151799687,
-      "grad_norm": 0.26721506749946805,
+      "grad_norm": 0.2666715468288786,
       "learning_rate": 3.88864611818311e-06,
-      "loss": 0.1712,
+      "loss": 0.1717,
       "step": 978
     },
     {
       "epoch": 4.085550339071466,
-      "grad_norm": 0.2799194172615097,
+      "grad_norm": 0.2835960732211078,
       "learning_rate": 3.8540841683703e-06,
-      "loss": 0.1989,
+      "loss": 0.1991,
       "step": 979
     },
     {
       "epoch": 4.089723526343245,
-      "grad_norm": 0.2795880959873755,
+      "grad_norm": 0.2795818557428652,
       "learning_rate": 3.819660112501053e-06,
-      "loss": 0.1871,
+      "loss": 0.1875,
       "step": 980
     },
     {
       "epoch": 4.093896713615023,
-      "grad_norm": 0.2769416507069011,
+      "grad_norm": 0.2758404983136882,
       "learning_rate": 3.785374244573492e-06,
-      "loss": 0.2098,
+      "loss": 0.2101,
       "step": 981
     },
     {
       "epoch": 4.098069900886802,
-      "grad_norm": 0.2642985684885375,
+      "grad_norm": 0.2592960801500646,
       "learning_rate": 3.7512268574055434e-06,
-      "loss": 0.2015,
+      "loss": 0.2018,
       "step": 982
     },
     {
       "epoch": 4.102243088158581,
-      "grad_norm": 0.26794517570781246,
+      "grad_norm": 0.2680430068223741,
       "learning_rate": 3.717218242632428e-06,
-      "loss": 0.1875,
+      "loss": 0.1876,
       "step": 983
     },
     {
       "epoch": 4.10641627543036,
-      "grad_norm": 0.24733463857612767,
+      "grad_norm": 0.25536554189377836,
       "learning_rate": 3.683348690704196e-06,
-      "loss": 0.1882,
+      "loss": 0.1886,
       "step": 984
     },
     {
       "epoch": 4.110589462702139,
-      "grad_norm": 0.2516030300687953,
+      "grad_norm": 0.25314281409735295,
       "learning_rate": 3.649618490883233e-06,
-      "loss": 0.2078,
+      "loss": 0.208,
       "step": 985
     },
     {
       "epoch": 4.114762649973917,
-      "grad_norm": 0.27700940202272567,
+      "grad_norm": 0.27516484245117945,
       "learning_rate": 3.61602793124179e-06,
-      "loss": 0.2007,
+      "loss": 0.2012,
       "step": 986
     },
     {
       "epoch": 4.118935837245696,
-      "grad_norm": 0.25827446150794575,
+      "grad_norm": 0.2512796031295832,
       "learning_rate": 3.5825772986595086e-06,
-      "loss": 0.1743,
+      "loss": 0.1745,
       "step": 987
     },
     {
       "epoch": 4.1231090245174755,
-      "grad_norm": 0.2612323928144574,
+      "grad_norm": 0.25998630634965497,
       "learning_rate": 3.5492668788210116e-06,
-      "loss": 0.1915,
+      "loss": 0.1917,
       "step": 988
     },
     {
       "epoch": 4.127282211789254,
-      "grad_norm": 0.2685424288053998,
+      "grad_norm": 0.2682429964038558,
       "learning_rate": 3.5160969562134083e-06,
-      "loss": 0.2017,
+      "loss": 0.202,
       "step": 989
     },
     {
       "epoch": 4.131455399061033,
-      "grad_norm": 0.2536452628745547,
+      "grad_norm": 0.2539089050711405,
       "learning_rate": 3.483067814123917e-06,
       "loss": 0.1947,
       "step": 990
     },
     {
       "epoch": 4.135628586332811,
-      "grad_norm": 0.2556268721289269,
+      "grad_norm": 0.25532738562451107,
       "learning_rate": 3.450179734637409e-06,
-      "loss": 0.1772,
+      "loss": 0.1777,
       "step": 991
     },
     {
       "epoch": 4.1398017736045905,
-      "grad_norm": 0.25454295280536676,
+      "grad_norm": 0.2513514547808499,
       "learning_rate": 3.417432998634025e-06,
-      "loss": 0.197,
+      "loss": 0.1976,
       "step": 992
     },
     {
       "epoch": 4.14397496087637,
-      "grad_norm": 0.26995519279386576,
+      "grad_norm": 0.2597194303157173,
       "learning_rate": 3.3848278857867433e-06,
-      "loss": 0.1875,
+      "loss": 0.1879,
       "step": 993
     },
     {
       "epoch": 4.148148148148148,
-      "grad_norm": 0.27615096066234873,
+      "grad_norm": 0.2655918051971436,
       "learning_rate": 3.3523646745590323e-06,
-      "loss": 0.186,
+      "loss": 0.1861,
       "step": 994
     },
     {
       "epoch": 4.152321335419927,
-      "grad_norm": 0.2515034644548685,
+      "grad_norm": 0.24759982236121256,
       "learning_rate": 3.320043642202444e-06,
-      "loss": 0.2111,
+      "loss": 0.2116,
       "step": 995
     },
     {
       "epoch": 4.156494522691705,
-      "grad_norm": 0.2469039309609633,
+      "grad_norm": 0.2430720064533326,
       "learning_rate": 3.2878650647542475e-06,
-      "loss": 0.2009,
+      "loss": 0.2007,
       "step": 996
     },
     {
       "epoch": 4.160667709963485,
-      "grad_norm": 0.2502032018189144,
+      "grad_norm": 0.2534018426765107,
       "learning_rate": 3.2558292170350915e-06,
-      "loss": 0.1975,
+      "loss": 0.1979,
       "step": 997
     },
     {
       "epoch": 4.164840897235264,
-      "grad_norm": 0.2468158566109766,
+      "grad_norm": 0.24538471563098643,
       "learning_rate": 3.2239363726466232e-06,
-      "loss": 0.1986,
+      "loss": 0.1987,
       "step": 998
     },
     {
       "epoch": 4.169014084507042,
-      "grad_norm": 0.24467748138161668,
+      "grad_norm": 0.24425301578745964,
       "learning_rate": 3.1921868039692037e-06,
-      "loss": 0.1999,
+      "loss": 0.2004,
       "step": 999
     },
     {
       "epoch": 4.173187271778821,
-      "grad_norm": 0.25309258508407584,
+      "grad_norm": 0.2508482580705475,
       "learning_rate": 3.160580782159517e-06,
-      "loss": 0.1925,
+      "loss": 0.1924,
       "step": 1000
     },
     {
       "epoch": 4.1773604590505995,
-      "grad_norm": 0.2483131417131967,
+      "grad_norm": 0.24718582462356303,
       "learning_rate": 3.1291185771483135e-06,
-      "loss": 0.2116,
+      "loss": 0.212,
       "step": 1001
     },
     {
       "epoch": 4.181533646322379,
-      "grad_norm": 0.24718508658239122,
+      "grad_norm": 0.24710163437309832,
       "learning_rate": 3.097800457638054e-06,
-      "loss": 0.1857,
+      "loss": 0.186,
       "step": 1002
     },
     {
       "epoch": 4.185706833594158,
-      "grad_norm": 0.26098168736096716,
+      "grad_norm": 0.25613259647953507,
       "learning_rate": 3.066626691100658e-06,
-      "loss": 0.1799,
+      "loss": 0.1801,
       "step": 1003
     },
     {
       "epoch": 4.189880020865936,
-      "grad_norm": 0.24258219397438785,
+      "grad_norm": 0.24146612220714223,
       "learning_rate": 3.035597543775195e-06,
-      "loss": 0.1861,
+      "loss": 0.1867,
       "step": 1004
     },
     {
       "epoch": 4.194053208137715,
-      "grad_norm": 0.2653865365407629,
+      "grad_norm": 0.2613932161555007,
       "learning_rate": 3.004713280665621e-06,
-      "loss": 0.195,
+      "loss": 0.1951,
       "step": 1005
     },
     {
       "epoch": 4.198226395409494,
-      "grad_norm": 0.2574239161266361,
+      "grad_norm": 0.25771689669677944,
       "learning_rate": 2.973974165538498e-06,
-      "loss": 0.1933,
+      "loss": 0.1935,
       "step": 1006
     },
     {
       "epoch": 4.202399582681273,
-      "grad_norm": 0.2546240799963392,
+      "grad_norm": 0.2502543375119621,
       "learning_rate": 2.943380460920764e-06,
-      "loss": 0.1712,
+      "loss": 0.1708,
       "step": 1007
     },
     {
       "epoch": 4.206572769953052,
-      "grad_norm": 0.25568776414784855,
+      "grad_norm": 0.25610031650024007,
       "learning_rate": 2.912932428097481e-06,
-      "loss": 0.1835,
+      "loss": 0.1837,
       "step": 1008
     },
     {
       "epoch": 4.21074595722483,
-      "grad_norm": 0.2439965611540862,
+      "grad_norm": 0.2503899332212452,
       "learning_rate": 2.882630327109597e-06,
-      "loss": 0.2089,
+      "loss": 0.209,
       "step": 1009
     },
     {
       "epoch": 4.214919144496609,
-      "grad_norm": 0.24794124508540863,
+      "grad_norm": 0.24819688299778947,
       "learning_rate": 2.8524744167517427e-06,
-      "loss": 0.206,
+      "loss": 0.2061,
       "step": 1010
     },
     {
       "epoch": 4.219092331768388,
-      "grad_norm": 0.25251148995922174,
+      "grad_norm": 0.2573945675448645,
       "learning_rate": 2.8224649545699856e-06,
-      "loss": 0.192,
+      "loss": 0.1924,
       "step": 1011
     },
     {
       "epoch": 4.223265519040167,
-      "grad_norm": 0.2640340946571945,
+      "grad_norm": 0.2671706375603529,
       "learning_rate": 2.7926021968596928e-06,
-      "loss": 0.1829,
+      "loss": 0.183,
       "step": 1012
     },
     {
       "epoch": 4.227438706311946,
-      "grad_norm": 0.24163583313129144,
+      "grad_norm": 0.2496201007093637,
       "learning_rate": 2.7628863986632694e-06,
-      "loss": 0.2007,
+      "loss": 0.2012,
       "step": 1013
     },
     {
       "epoch": 4.231611893583724,
-      "grad_norm": 0.2452762643241012,
+      "grad_norm": 0.2408502352291878,
       "learning_rate": 2.7333178137680373e-06,
-      "loss": 0.1889,
+      "loss": 0.1892,
       "step": 1014
     },
     {
       "epoch": 4.2357850808555035,
-      "grad_norm": 0.2422259208235282,
+      "grad_norm": 0.2461128380769594,
       "learning_rate": 2.7038966947040247e-06,
-      "loss": 0.2086,
+      "loss": 0.2092,
       "step": 1015
     },
     {
       "epoch": 4.239958268127282,
-      "grad_norm": 0.25187047878383384,
+      "grad_norm": 0.2522856302431717,
       "learning_rate": 2.6746232927418535e-06,
-      "loss": 0.1892,
+      "loss": 0.1896,
       "step": 1016
     },
     {
       "epoch": 4.244131455399061,
-      "grad_norm": 0.24522364893263424,
+      "grad_norm": 0.24426548998637596,
       "learning_rate": 2.645497857890553e-06,
-      "loss": 0.1764,
+      "loss": 0.1768,
       "step": 1017
     },
     {
       "epoch": 4.24830464267084,
-      "grad_norm": 0.24288450809993806,
+      "grad_norm": 0.24307868147677103,
       "learning_rate": 2.6165206388954544e-06,
-      "loss": 0.1954,
+      "loss": 0.1955,
       "step": 1018
     },
     {
       "epoch": 4.2524778299426185,
-      "grad_norm": 0.2508878637306535,
+      "grad_norm": 0.2470276716528675,
       "learning_rate": 2.5876918832360364e-06,
-      "loss": 0.1693,
+      "loss": 0.1694,
       "step": 1019
     },
     {
       "epoch": 4.256651017214398,
-      "grad_norm": 0.253613679017836,
+      "grad_norm": 0.26265058025996063,
       "learning_rate": 2.559011837123846e-06,
-      "loss": 0.2071,
+      "loss": 0.2076,
       "step": 1020
     },
     {
       "epoch": 4.260824204486176,
-      "grad_norm": 0.25114269441486037,
+      "grad_norm": 0.25342390568901285,
       "learning_rate": 2.5304807455003768e-06,
-      "loss": 0.1939,
+      "loss": 0.1943,
       "step": 1021
     },
     {
       "epoch": 4.264997391757955,
-      "grad_norm": 0.24582293641821876,
+      "grad_norm": 0.24916756460201128,
       "learning_rate": 2.502098852034962e-06,
-      "loss": 0.1723,
+      "loss": 0.1724,
       "step": 1022
     },
     {
       "epoch": 4.269170579029734,
-      "grad_norm": 0.2552045555463069,
+      "grad_norm": 0.25403275517703616,
       "learning_rate": 2.473866399122733e-06,
-      "loss": 0.2103,
+      "loss": 0.2108,
       "step": 1023
     },
     {
       "epoch": 4.273343766301513,
-      "grad_norm": 0.24040416458209377,
+      "grad_norm": 0.23905328903811932,
       "learning_rate": 2.4457836278825007e-06,
       "loss": 0.1964,
       "step": 1024
     },
     {
       "epoch": 4.277516953573292,
-      "grad_norm": 0.2532068070819206,
+      "grad_norm": 0.24454866662030647,
       "learning_rate": 2.4178507781547577e-06,
       "loss": 0.1999,
       "step": 1025
     },
     {
       "epoch": 4.28169014084507,
-      "grad_norm": 0.2351852019273944,
+      "grad_norm": 0.23818275260345975,
       "learning_rate": 2.3900680884995577e-06,
-      "loss": 0.1844,
+      "loss": 0.1845,
       "step": 1026
     },
     {
       "epoch": 4.285863328116849,
-      "grad_norm": 0.23693712498266906,
+      "grad_norm": 0.23656549794531792,
       "learning_rate": 2.36243579619454e-06,
       "loss": 0.1884,
       "step": 1027
     },
     {
       "epoch": 4.290036515388628,
-      "grad_norm": 0.24591564377906516,
+      "grad_norm": 0.24487209446459487,
       "learning_rate": 2.334954137232861e-06,
-      "loss": 0.1795,
+      "loss": 0.18,
       "step": 1028
     },
     {
       "epoch": 4.294209702660407,
-      "grad_norm": 0.250574897723549,
+      "grad_norm": 0.2505619893219741,
       "learning_rate": 2.3076233463212063e-06,
-      "loss": 0.2041,
+      "loss": 0.2045,
       "step": 1029
     },
     {
       "epoch": 4.298382889932186,
-      "grad_norm": 0.24901721069404678,
+      "grad_norm": 0.24610959490819218,
       "learning_rate": 2.280443656877769e-06,
-      "loss": 0.1635,
+      "loss": 0.1638,
       "step": 1030
     },
     {
       "epoch": 4.302556077203964,
-      "grad_norm": 0.23642876503443205,
+      "grad_norm": 0.23933380250722203,
       "learning_rate": 2.2534153010302706e-06,
-      "loss": 0.1967,
+      "loss": 0.1971,
       "step": 1031
     },
     {
       "epoch": 4.306729264475743,
-      "grad_norm": 0.24706802912462783,
+      "grad_norm": 0.2444878643734393,
       "learning_rate": 2.226538509613958e-06,
       "loss": 0.1763,
       "step": 1032
     },
     {
       "epoch": 4.3109024517475225,
-      "grad_norm": 0.2491595729045565,
+      "grad_norm": 0.2511282942782746,
       "learning_rate": 2.199813512169655e-06,
-      "loss": 0.2068,
+      "loss": 0.2071,
       "step": 1033
     },
     {
       "epoch": 4.315075639019301,
-      "grad_norm": 0.2535084973663096,
+      "grad_norm": 0.24826489232202018,
       "learning_rate": 2.1732405369417896e-06,
-      "loss": 0.1961,
+      "loss": 0.1959,
       "step": 1034
     },
     {
       "epoch": 4.31924882629108,
-      "grad_norm": 0.24410407151634858,
+      "grad_norm": 0.24784785718328076,
       "learning_rate": 2.1468198108764373e-06,
-      "loss": 0.1854,
+      "loss": 0.1855,
       "step": 1035
     },
     {
       "epoch": 4.323422013562858,
-      "grad_norm": 0.2473958436456897,
+      "grad_norm": 0.24485569976147994,
       "learning_rate": 2.1205515596194058e-06,
       "loss": 0.2,
       "step": 1036
     },
     {
       "epoch": 4.3275952008346374,
-      "grad_norm": 0.25100416631840966,
+      "grad_norm": 0.2546663608683428,
       "learning_rate": 2.094436007514282e-06,
-      "loss": 0.1844,
+      "loss": 0.1846,
       "step": 1037
     },
     {
       "epoch": 4.331768388106417,
-      "grad_norm": 0.24338556129818617,
+      "grad_norm": 0.2451988394866516,
       "learning_rate": 2.068473377600535e-06,
-      "loss": 0.2014,
+      "loss": 0.202,
       "step": 1038
     },
     {
       "epoch": 4.335941575378195,
-      "grad_norm": 0.23890565914286865,
+      "grad_norm": 0.2366482495170551,
       "learning_rate": 2.0426638916116026e-06,
-      "loss": 0.1928,
+      "loss": 0.1931,
       "step": 1039
     },
     {
       "epoch": 4.340114762649974,
-      "grad_norm": 0.251603455789501,
+      "grad_norm": 0.2585078762101229,
       "learning_rate": 2.017007769973005e-06,
-      "loss": 0.182,
+      "loss": 0.1823,
       "step": 1040
     },
     {
       "epoch": 4.344287949921752,
-      "grad_norm": 0.24925388305896817,
+      "grad_norm": 0.2456216047320286,
       "learning_rate": 1.9915052318004435e-06,
-      "loss": 0.1942,
+      "loss": 0.1944,
       "step": 1041
     },
     {
       "epoch": 4.348461137193532,
-      "grad_norm": 0.2559940910457567,
+      "grad_norm": 0.25472492848762135,
       "learning_rate": 1.9661564948979505e-06,
-      "loss": 0.1843,
+      "loss": 0.1845,
       "step": 1042
     },
     {
       "epoch": 4.352634324465311,
-      "grad_norm": 0.25946890953086316,
+      "grad_norm": 0.25730009131272835,
       "learning_rate": 1.9409617757560206e-06,
       "loss": 0.212,
       "step": 1043
     },
     {
       "epoch": 4.356807511737089,
-      "grad_norm": 0.24739464513999532,
+      "grad_norm": 0.252941966700885,
       "learning_rate": 1.915921289549769e-06,
       "loss": 0.2045,
       "step": 1044
     },
     {
       "epoch": 4.360980699008868,
-      "grad_norm": 0.2363325131258638,
+      "grad_norm": 0.23543656553967415,
       "learning_rate": 1.8910352501370677e-06,
-      "loss": 0.1911,
+      "loss": 0.1912,
       "step": 1045
     },
     {
       "epoch": 4.3651538862806465,
-      "grad_norm": 0.2505492001703832,
+      "grad_norm": 0.24848245063832886,
       "learning_rate": 1.8663038700567604e-06,
-      "loss": 0.1927,
+      "loss": 0.193,
       "step": 1046
     },
     {
       "epoch": 4.369327073552426,
-      "grad_norm": 0.2514495773891022,
+      "grad_norm": 0.25251079888035405,
       "learning_rate": 1.8417273605268037e-06,
-      "loss": 0.1699,
+      "loss": 0.1703,
       "step": 1047
     },
     {
       "epoch": 4.373500260824205,
-      "grad_norm": 0.23612806807053585,
+      "grad_norm": 0.2377053066952455,
       "learning_rate": 1.8173059314425057e-06,
-      "loss": 0.2116,
+      "loss": 0.2119,
       "step": 1048
     },
     {
       "epoch": 4.377673448095983,
-      "grad_norm": 0.23557126692667318,
+      "grad_norm": 0.24079289450630167,
       "learning_rate": 1.7930397913747e-06,
-      "loss": 0.1991,
+      "loss": 0.1994,
       "step": 1049
     },
     {
       "epoch": 4.381846635367762,
-      "grad_norm": 0.2371250462538808,
+      "grad_norm": 0.2431422433712598,
       "learning_rate": 1.7689291475679748e-06,
-      "loss": 0.206,
+      "loss": 0.2065,
       "step": 1050
     },
     {
       "epoch": 4.386019822639541,
-      "grad_norm": 0.24581314657038514,
+      "grad_norm": 0.2450528558880145,
       "learning_rate": 1.7449742059389118e-06,
-      "loss": 0.2005,
+      "loss": 0.2009,
       "step": 1051
     },
     {
       "epoch": 4.39019300991132,
-      "grad_norm": 0.25088133959453385,
+      "grad_norm": 0.2518954882497055,
       "learning_rate": 1.7211751710743208e-06,
-      "loss": 0.1997,
+      "loss": 0.2,
       "step": 1052
     },
     {
       "epoch": 4.394366197183099,
-      "grad_norm": 0.2571990315970285,
+      "grad_norm": 0.2577686930410637,
       "learning_rate": 1.6975322462294897e-06,
-      "loss": 0.196,
+      "loss": 0.1959,
       "step": 1053
     },
     {
       "epoch": 4.398539384454877,
-      "grad_norm": 0.24229961567640554,
+      "grad_norm": 0.24696415181312203,
       "learning_rate": 1.6740456333264465e-06,
-      "loss": 0.1776,
+      "loss": 0.1781,
       "step": 1054
     },
     {
       "epoch": 4.402712571726656,
-      "grad_norm": 0.23691182017294535,
+      "grad_norm": 0.2405146853786115,
       "learning_rate": 1.6507155329522517e-06,
-      "loss": 0.1954,
+      "loss": 0.1957,
       "step": 1055
     },
     {
       "epoch": 4.406885758998435,
-      "grad_norm": 0.24093457622671619,
+      "grad_norm": 0.24317371504281304,
       "learning_rate": 1.6275421443572525e-06,
-      "loss": 0.1789,
+      "loss": 0.1791,
       "step": 1056
     },
     {
       "epoch": 4.411058946270214,
-      "grad_norm": 0.2581935263305873,
+      "grad_norm": 0.2507100297440804,
       "learning_rate": 1.604525665453429e-06,
-      "loss": 0.1895,
+      "loss": 0.1893,
       "step": 1057
     },
     {
       "epoch": 4.415232133541993,
-      "grad_norm": 0.2485284204245002,
+      "grad_norm": 0.24724619545394755,
       "learning_rate": 1.5816662928126535e-06,
       "loss": 0.189,
       "step": 1058
     },
     {
       "epoch": 4.419405320813771,
-      "grad_norm": 0.24144404233118977,
+      "grad_norm": 0.24330738458673748,
       "learning_rate": 1.5589642216650491e-06,
-      "loss": 0.1647,
+      "loss": 0.1651,
       "step": 1059
     },
     {
       "epoch": 4.4235785080855505,
-      "grad_norm": 0.23671874262973613,
+      "grad_norm": 0.23903102413416788,
       "learning_rate": 1.5364196458972957e-06,
-      "loss": 0.1892,
+      "loss": 0.1894,
       "step": 1060
     },
     {
       "epoch": 4.427751695357329,
-      "grad_norm": 0.24342489415701582,
+      "grad_norm": 0.26480590494758033,
       "learning_rate": 1.5140327580509961e-06,
-      "loss": 0.2095,
+      "loss": 0.2097,
       "step": 1061
     },
     {
       "epoch": 4.431924882629108,
-      "grad_norm": 0.2397291858032904,
+      "grad_norm": 0.24085018803041106,
       "learning_rate": 1.4918037493210191e-06,
-      "loss": 0.1945,
+      "loss": 0.1947,
       "step": 1062
     },
     {
       "epoch": 4.436098069900887,
-      "grad_norm": 0.2468073837993282,
+      "grad_norm": 0.24440831735880653,
       "learning_rate": 1.4697328095538655e-06,
-      "loss": 0.1851,
+      "loss": 0.1848,
       "step": 1063
     },
     {
       "epoch": 4.4402712571726655,
-      "grad_norm": 0.24886791913560588,
+      "grad_norm": 0.2510078978452769,
       "learning_rate": 1.447820127246058e-06,
       "loss": 0.1938,
       "step": 1064
     },
     {
       "epoch": 4.444444444444445,
-      "grad_norm": 0.24237745348251277,
+      "grad_norm": 0.24388495418305353,
       "learning_rate": 1.4260658895425162e-06,
-      "loss": 0.2254,
+      "loss": 0.2253,
       "step": 1065
     },
     {
       "epoch": 4.448617631716223,
-      "grad_norm": 0.24626243456358043,
+      "grad_norm": 0.24432119196955265,
       "learning_rate": 1.4044702822349731e-06,
-      "loss": 0.2035,
+      "loss": 0.2033,
       "step": 1066
     },
     {
       "epoch": 4.452790818988002,
-      "grad_norm": 0.2483655718263905,
+      "grad_norm": 0.25012555661544894,
       "learning_rate": 1.3830334897603747e-06,
-      "loss": 0.204,
+      "loss": 0.2044,
       "step": 1067
     },
     {
       "epoch": 4.456964006259781,
-      "grad_norm": 0.24376257830382864,
+      "grad_norm": 0.24058776168269047,
       "learning_rate": 1.3617556951993182e-06,
-      "loss": 0.1877,
+      "loss": 0.1879,
       "step": 1068
     },
     {
       "epoch": 4.46113719353156,
-      "grad_norm": 0.2486087096607852,
+      "grad_norm": 0.2489560127785182,
       "learning_rate": 1.3406370802744672e-06,
-      "loss": 0.18,
+      "loss": 0.1804,
       "step": 1069
     },
     {
       "epoch": 4.465310380803339,
-      "grad_norm": 0.25546638945116434,
+      "grad_norm": 0.24875745615210687,
       "learning_rate": 1.3196778253490417e-06,
-      "loss": 0.2028,
+      "loss": 0.2027,
       "step": 1070
     },
     {
       "epoch": 4.469483568075117,
-      "grad_norm": 0.23801780395862046,
+      "grad_norm": 0.23935177119749995,
       "learning_rate": 1.2988781094252189e-06,
-      "loss": 0.2027,
+      "loss": 0.2025,
       "step": 1071
     },
     {
       "epoch": 4.473656755346896,
-      "grad_norm": 0.2460920635065085,
+      "grad_norm": 0.24804229027813832,
       "learning_rate": 1.2782381101426578e-06,
-      "loss": 0.1785,
+      "loss": 0.1789,
       "step": 1072
     },
     {
       "epoch": 4.477829942618675,
-      "grad_norm": 0.24974744311202382,
+      "grad_norm": 0.24910398232402925,
       "learning_rate": 1.2577580037769476e-06,
-      "loss": 0.2015,
+      "loss": 0.2017,
       "step": 1073
     },
     {
       "epoch": 4.482003129890454,
-      "grad_norm": 0.24741411198639232,
+      "grad_norm": 0.2443132099471584,
       "learning_rate": 1.2374379652381196e-06,
-      "loss": 0.2145,
+      "loss": 0.2148,
       "step": 1074
     },
     {
       "epoch": 4.486176317162233,
-      "grad_norm": 0.25311297938358057,
+      "grad_norm": 0.2518057439102505,
       "learning_rate": 1.2172781680691515e-06,
-      "loss": 0.1914,
+      "loss": 0.1917,
       "step": 1075
     },
     {
       "epoch": 4.490349504434011,
-      "grad_norm": 0.24016849800448847,
+      "grad_norm": 0.24340303518636264,
       "learning_rate": 1.197278784444471e-06,
-      "loss": 0.2072,
+      "loss": 0.2075,
       "step": 1076
     },
     {
       "epoch": 4.49452269170579,
-      "grad_norm": 0.24014449452923373,
+      "grad_norm": 0.2390772875141554,
       "learning_rate": 1.1774399851685115e-06,
-      "loss": 0.1989,
+      "loss": 0.1993,
       "step": 1077
     },
     {
       "epoch": 4.4986958789775695,
-      "grad_norm": 0.240270218689764,
+      "grad_norm": 0.24219806611114617,
       "learning_rate": 1.15776193967422e-06,
-      "loss": 0.2083,
+      "loss": 0.2082,
       "step": 1078
     },
     {
       "epoch": 4.502869066249348,
-      "grad_norm": 0.24262681142268436,
+      "grad_norm": 0.24396978289642962,
       "learning_rate": 1.138244816021652e-06,
-      "loss": 0.1784,
+      "loss": 0.1787,
       "step": 1079
     },
     {
       "epoch": 4.507042253521127,
-      "grad_norm": 0.24390674910699023,
+      "grad_norm": 0.24355207359197745,
       "learning_rate": 1.1188887808964878e-06,
-      "loss": 0.1924,
+      "loss": 0.1928,
       "step": 1080
     },
     {
       "epoch": 4.511215440792905,
-      "grad_norm": 0.2455696084478399,
+      "grad_norm": 0.24983037077573841,
       "learning_rate": 1.099693999608653e-06,
-      "loss": 0.1773,
+      "loss": 0.1778,
       "step": 1081
     },
     {
       "epoch": 4.515388628064684,
-      "grad_norm": 0.2481049561764695,
+      "grad_norm": 0.2524499359613224,
       "learning_rate": 1.0806606360908712e-06,
       "loss": 0.2076,
       "step": 1082
     },
     {
       "epoch": 4.519561815336463,
-      "grad_norm": 0.2440955877428125,
+      "grad_norm": 0.24489631743337248,
       "learning_rate": 1.0617888528972965e-06,
-      "loss": 0.1975,
+      "loss": 0.1974,
       "step": 1083
     },
     {
       "epoch": 4.523735002608242,
-      "grad_norm": 0.2484524264664831,
+      "grad_norm": 0.24893112913736937,
       "learning_rate": 1.0430788112020918e-06,
-      "loss": 0.1973,
+      "loss": 0.1976,
       "step": 1084
     },
     {
       "epoch": 4.527908189880021,
-      "grad_norm": 0.24113643026635156,
+      "grad_norm": 0.24417610073392695,
       "learning_rate": 1.0245306707980873e-06,
-      "loss": 0.1912,
+      "loss": 0.1916,
       "step": 1085
     },
     {
       "epoch": 4.532081377151799,
-      "grad_norm": 0.24704419821617016,
+      "grad_norm": 0.2524750543820841,
       "learning_rate": 1.0061445900953814e-06,
-      "loss": 0.1928,
+      "loss": 0.193,
       "step": 1086
     },
     {
       "epoch": 4.5362545644235786,
-      "grad_norm": 0.2590881750169755,
+      "grad_norm": 0.255421185745401,
       "learning_rate": 9.879207261200107e-07,
-      "loss": 0.2043,
+      "loss": 0.2044,
       "step": 1087
     },
     {
       "epoch": 4.540427751695358,
-      "grad_norm": 0.2531828778633018,
+      "grad_norm": 0.25198834341544374,
       "learning_rate": 9.69859234512609e-07,
-      "loss": 0.1796,
+      "loss": 0.18,
       "step": 1088
     },
     {
       "epoch": 4.544600938967136,
-      "grad_norm": 0.23925304932005148,
+      "grad_norm": 0.23616327704921233,
       "learning_rate": 9.519602695270569e-07,
-      "loss": 0.2076,
+      "loss": 0.2074,
       "step": 1089
     },
     {
       "epoch": 4.548774126238915,
-      "grad_norm": 0.2477843144114964,
+      "grad_norm": 0.24315594750266786,
       "learning_rate": 9.342239840291967e-07,
-      "loss": 0.1957,
+      "loss": 0.1959,
       "step": 1090
     },
     {
       "epoch": 4.5529473135106935,
-      "grad_norm": 0.246631198979582,
+      "grad_norm": 0.24597462273554155,
       "learning_rate": 9.166505294954842e-07,
       "loss": 0.2079,
       "step": 1091
     },
     {
       "epoch": 4.557120500782473,
-      "grad_norm": 0.2558151734856077,
+      "grad_norm": 0.25309391416216936,
       "learning_rate": 8.992400560117542e-07,
-      "loss": 0.1823,
+      "loss": 0.1827,
       "step": 1092
     },
     {
       "epoch": 4.561293688054251,
-      "grad_norm": 0.23553223030845918,
+      "grad_norm": 0.2329553350589884,
       "learning_rate": 8.819927122718708e-07,
-      "loss": 0.1883,
+      "loss": 0.1882,
       "step": 1093
     },
     {
       "epoch": 4.56546687532603,
-      "grad_norm": 0.237615825357627,
+      "grad_norm": 0.23751691995350852,
       "learning_rate": 8.649086455765121e-07,
-      "loss": 0.1911,
+      "loss": 0.1912,
       "step": 1094
     },
     {
       "epoch": 4.569640062597809,
-      "grad_norm": 0.24507761030695296,
+      "grad_norm": 0.24492503958742146,
       "learning_rate": 8.479880018318831e-07,
-      "loss": 0.2091,
+      "loss": 0.2095,
       "step": 1095
     },
     {
       "epoch": 4.573813249869588,
-      "grad_norm": 0.24517400085795302,
+      "grad_norm": 0.2457142179098032,
       "learning_rate": 8.312309255484763e-07,
-      "loss": 0.1849,
+      "loss": 0.1851,
       "step": 1096
     },
     {
       "epoch": 4.577986437141367,
-      "grad_norm": 0.2339978906074357,
+      "grad_norm": 0.2360524538723885,
       "learning_rate": 8.14637559839846e-07,
-      "loss": 0.1924,
+      "loss": 0.1928,
       "step": 1097
     },
     {
       "epoch": 4.582159624413146,
-      "grad_norm": 0.2452184646221617,
+      "grad_norm": 0.24649876449016186,
       "learning_rate": 7.982080464213737e-07,
-      "loss": 0.1876,
+      "loss": 0.1877,
       "step": 1098
     },
     {
       "epoch": 4.586332811684924,
-      "grad_norm": 0.2492377253205176,
+      "grad_norm": 0.25095023559869306,
       "learning_rate": 7.819425256090607e-07,
-      "loss": 0.2116,
+      "loss": 0.2117,
       "step": 1099
     },
     {
       "epoch": 4.590505998956703,
-      "grad_norm": 0.24412935239594263,
+      "grad_norm": 0.248046557961123,
       "learning_rate": 7.658411363183393e-07,
-      "loss": 0.199,
+      "loss": 0.1997,
       "step": 1100
     },
     {
       "epoch": 4.594679186228482,
-      "grad_norm": 0.24737228244232362,
+      "grad_norm": 0.24395709505770813,
       "learning_rate": 7.499040160628724e-07,
-      "loss": 0.1897,
+      "loss": 0.1898,
       "step": 1101
     },
     {
       "epoch": 4.598852373500261,
-      "grad_norm": 0.24541265564782017,
+      "grad_norm": 0.25059562903325305,
       "learning_rate": 7.341313009533869e-07,
-      "loss": 0.202,
+      "loss": 0.2021,
       "step": 1102
     },
     {
       "epoch": 4.603025560772039,
-      "grad_norm": 0.25193834629616324,
+      "grad_norm": 0.25611688580278946,
       "learning_rate": 7.185231256965153e-07,
-      "loss": 0.1867,
+      "loss": 0.1874,
       "step": 1103
     },
     {
       "epoch": 4.607198748043818,
-      "grad_norm": 0.2428178545745535,
+      "grad_norm": 0.24232494737396135,
       "learning_rate": 7.030796235936299e-07,
-      "loss": 0.2083,
+      "loss": 0.2086,
       "step": 1104
     },
     {
       "epoch": 4.6113719353155975,
-      "grad_norm": 0.24246552598049126,
+      "grad_norm": 0.24299476488018787,
       "learning_rate": 6.878009265397301e-07,
-      "loss": 0.1952,
+      "loss": 0.1954,
       "step": 1105
     },
     {
       "epoch": 4.615545122587376,
-      "grad_norm": 0.24198942505911766,
+      "grad_norm": 0.24959627233854223,
       "learning_rate": 6.72687165022281e-07,
-      "loss": 0.196,
+      "loss": 0.1967,
       "step": 1106
     },
     {
       "epoch": 4.619718309859155,
-      "grad_norm": 0.2488080064213975,
+      "grad_norm": 0.2505565903836929,
       "learning_rate": 6.577384681201371e-07,
-      "loss": 0.184,
+      "loss": 0.1842,
       "step": 1107
     },
     {
       "epoch": 4.623891497130934,
-      "grad_norm": 0.23788010606890422,
+      "grad_norm": 0.23822176643569445,
       "learning_rate": 6.429549635024046e-07,
-      "loss": 0.1908,
+      "loss": 0.191,
       "step": 1108
     },
     {
       "epoch": 4.6280646844027125,
-      "grad_norm": 0.2546300440369056,
+      "grad_norm": 0.2568585006766818,
       "learning_rate": 6.283367774273785e-07,
-      "loss": 0.1783,
+      "loss": 0.1776,
       "step": 1109
     },
     {
       "epoch": 4.632237871674492,
-      "grad_norm": 0.24845416617149568,
+      "grad_norm": 0.24626390659590985,
       "learning_rate": 6.138840347414498e-07,
-      "loss": 0.1803,
+      "loss": 0.1807,
       "step": 1110
     },
     {
       "epoch": 4.63641105894627,
-      "grad_norm": 0.24959274331115516,
+      "grad_norm": 0.2502698243137087,
       "learning_rate": 5.995968588780465e-07,
       "loss": 0.1867,
       "step": 1111
     },
     {
       "epoch": 4.640584246218049,
-      "grad_norm": 0.2361945779017687,
+      "grad_norm": 0.23673825709030313,
       "learning_rate": 5.854753718565653e-07,
-      "loss": 0.2024,
+      "loss": 0.2026,
       "step": 1112
     },
     {
       "epoch": 4.644757433489827,
-      "grad_norm": 0.24539791740318556,
+      "grad_norm": 0.24366649526806947,
       "learning_rate": 5.715196942813528e-07,
-      "loss": 0.1845,
+      "loss": 0.1846,
       "step": 1113
     },
     {
       "epoch": 4.648930620761607,
-      "grad_norm": 0.24262965278155374,
+      "grad_norm": 0.23929695027341227,
       "learning_rate": 5.57729945340657e-07,
-      "loss": 0.197,
+      "loss": 0.1971,
       "step": 1114
     },
     {
       "epoch": 4.653103808033386,
-      "grad_norm": 0.23712205636531491,
+      "grad_norm": 0.23934685517161333,
       "learning_rate": 5.44106242805611e-07,
-      "loss": 0.2007,
+      "loss": 0.2012,
       "step": 1115
     },
     {
       "epoch": 4.657276995305164,
-      "grad_norm": 0.23851602176620923,
+      "grad_norm": 0.23895834769848806,
       "learning_rate": 5.306487030292374e-07,
-      "loss": 0.1929,
+      "loss": 0.1925,
       "step": 1116
     },
     {
       "epoch": 4.661450182576943,
-      "grad_norm": 0.23996684894572815,
+      "grad_norm": 0.24081300140292572,
       "learning_rate": 5.173574409454429e-07,
-      "loss": 0.2227,
+      "loss": 0.2228,
       "step": 1117
     },
     {
       "epoch": 4.665623369848722,
-      "grad_norm": 0.2817413812310611,
+      "grad_norm": 0.2402135741287411,
       "learning_rate": 5.042325700680461e-07,
-      "loss": 0.2193,
+      "loss": 0.2195,
       "step": 1118
     },
     {
       "epoch": 4.669796557120501,
-      "grad_norm": 0.24187928971014894,
+      "grad_norm": 0.24016585408911284,
       "learning_rate": 4.912742024898065e-07,
-      "loss": 0.1794,
+      "loss": 0.1797,
       "step": 1119
     },
     {
       "epoch": 4.67396974439228,
-      "grad_norm": 0.23177306390980404,
+      "grad_norm": 0.23416924997223898,
       "learning_rate": 4.784824488814588e-07,
-      "loss": 0.1924,
+      "loss": 0.1927,
       "step": 1120
     },
     {
       "epoch": 4.678142931664058,
-      "grad_norm": 0.2517484643592054,
+      "grad_norm": 0.2473291867792998,
       "learning_rate": 4.658574184907738e-07,
-      "loss": 0.2088,
+      "loss": 0.2087,
       "step": 1121
     },
     {
       "epoch": 4.682316118935837,
-      "grad_norm": 0.24265486519965998,
+      "grad_norm": 0.2456057604945064,
       "learning_rate": 4.533992191416303e-07,
-      "loss": 0.2137,
+      "loss": 0.214,
       "step": 1122
     },
     {
       "epoch": 4.686489306207616,
-      "grad_norm": 0.24838416200161784,
+      "grad_norm": 0.2494296955716854,
       "learning_rate": 4.4110795723308675e-07,
-      "loss": 0.2025,
+      "loss": 0.2027,
       "step": 1123
     },
     {
       "epoch": 4.690662493479395,
-      "grad_norm": 0.24119140575375062,
+      "grad_norm": 0.24228142109548245,
       "learning_rate": 4.289837377384753e-07,
-      "loss": 0.1958,
+      "loss": 0.1961,
       "step": 1124
     },
     {
       "epoch": 4.694835680751174,
-      "grad_norm": 0.24784362663475612,
+      "grad_norm": 0.2505745129363188,
       "learning_rate": 4.1702666420450064e-07,
-      "loss": 0.1726,
+      "loss": 0.1727,
       "step": 1125
     },
     {
       "epoch": 4.699008868022952,
-      "grad_norm": 0.23581330455632188,
+      "grad_norm": 0.2378200453904596,
       "learning_rate": 4.0523683875036914e-07,
-      "loss": 0.1926,
+      "loss": 0.1928,
       "step": 1126
     },
     {
       "epoch": 4.703182055294731,
-      "grad_norm": 0.249809073755895,
+      "grad_norm": 0.2481100866552395,
       "learning_rate": 3.936143620668964e-07,
-      "loss": 0.2125,
+      "loss": 0.213,
       "step": 1127
     },
     {
       "epoch": 4.707355242566511,
-      "grad_norm": 0.25045590923747385,
+      "grad_norm": 0.2522728077681205,
       "learning_rate": 3.821593334156659e-07,
-      "loss": 0.2112,
+      "loss": 0.2116,
       "step": 1128
     },
     {
       "epoch": 4.711528429838289,
-      "grad_norm": 0.250730945844985,
+      "grad_norm": 0.24142040295797687,
       "learning_rate": 3.708718506281761e-07,
-      "loss": 0.1785,
+      "loss": 0.179,
       "step": 1129
     },
     {
       "epoch": 4.715701617110068,
-      "grad_norm": 0.24530748920725884,
+      "grad_norm": 0.24799393176130138,
       "learning_rate": 3.597520101049945e-07,
-      "loss": 0.205,
+      "loss": 0.2051,
       "step": 1130
     },
     {
       "epoch": 4.719874804381846,
-      "grad_norm": 0.24635648679035527,
+      "grad_norm": 0.2452427641350734,
       "learning_rate": 3.4879990681494726e-07,
       "loss": 0.1789,
       "step": 1131
     },
     {
       "epoch": 4.7240479916536255,
-      "grad_norm": 0.24680691322052528,
+      "grad_norm": 0.24425145186966457,
       "learning_rate": 3.3801563429430196e-07,
-      "loss": 0.2052,
+      "loss": 0.2049,
       "step": 1132
     },
     {
       "epoch": 4.728221178925404,
-      "grad_norm": 0.2434249480680997,
+      "grad_norm": 0.24178711600461378,
       "learning_rate": 3.273992846459728e-07,
-      "loss": 0.1947,
+      "loss": 0.1948,
       "step": 1133
     },
     {
       "epoch": 4.732394366197183,
-      "grad_norm": 0.23693898017867174,
+      "grad_norm": 0.23891297716158552,
       "learning_rate": 3.1695094853872787e-07,
-      "loss": 0.2099,
+      "loss": 0.2102,
       "step": 1134
     },
     {
       "epoch": 4.736567553468962,
-      "grad_norm": 0.24592753921202246,
+      "grad_norm": 0.2456441459279307,
       "learning_rate": 3.0667071520641857e-07,
-      "loss": 0.2002,
+      "loss": 0.2004,
       "step": 1135
     },
     {
       "epoch": 4.7407407407407405,
-      "grad_norm": 0.2507163200104358,
+      "grad_norm": 0.2450171138597638,
       "learning_rate": 2.9655867244721805e-07,
-      "loss": 0.1896,
+      "loss": 0.1897,
       "step": 1136
     },
     {
       "epoch": 4.74491392801252,
-      "grad_norm": 0.2428692266924491,
+      "grad_norm": 0.23988983284372467,
       "learning_rate": 2.8661490662287516e-07,
-      "loss": 0.1908,
+      "loss": 0.1909,
       "step": 1137
     },
     {
       "epoch": 4.749087115284299,
-      "grad_norm": 0.2431048682085987,
+      "grad_norm": 0.24348866042617265,
       "learning_rate": 2.7683950265795955e-07,
-      "loss": 0.1817,
+      "loss": 0.1821,
       "step": 1138
     },
     {
       "epoch": 4.753260302556077,
-      "grad_norm": 0.24569820487530003,
+      "grad_norm": 0.24861368418994848,
       "learning_rate": 2.6723254403916655e-07,
-      "loss": 0.1833,
+      "loss": 0.1835,
       "step": 1139
     },
     {
       "epoch": 4.757433489827856,
-      "grad_norm": 0.249473758586037,
+      "grad_norm": 0.2515111588328722,
       "learning_rate": 2.577941128145689e-07,
-      "loss": 0.1922,
+      "loss": 0.1924,
       "step": 1140
     },
     {
       "epoch": 4.761606677099635,
-      "grad_norm": 0.2430910083850744,
+      "grad_norm": 0.23842697004471958,
       "learning_rate": 2.485242895929441e-07,
-      "loss": 0.2067,
+      "loss": 0.2068,
       "step": 1141
     },
     {
       "epoch": 4.765779864371414,
-      "grad_norm": 0.24652116384524908,
+      "grad_norm": 0.24898623422617133,
       "learning_rate": 2.394231535430791e-07,
-      "loss": 0.2039,
+      "loss": 0.2038,
       "step": 1142
     },
     {
       "epoch": 4.769953051643192,
-      "grad_norm": 0.23865705877844834,
+      "grad_norm": 0.2409737214064303,
       "learning_rate": 2.3049078239307575e-07,
-      "loss": 0.2,
+      "loss": 0.2003,
       "step": 1143
     },
     {
       "epoch": 4.774126238914971,
-      "grad_norm": 0.23768112246929932,
+      "grad_norm": 0.24436818377113675,
       "learning_rate": 2.217272524297176e-07,
-      "loss": 0.1876,
+      "loss": 0.1872,
       "step": 1144
     },
     {
       "epoch": 4.77829942618675,
-      "grad_norm": 0.22926950385930567,
+      "grad_norm": 0.22767262989897288,
       "learning_rate": 2.1313263849779498e-07,
-      "loss": 0.1997,
+      "loss": 0.2002,
       "step": 1145
     },
     {
       "epoch": 4.782472613458529,
-      "grad_norm": 0.24848428909602316,
+      "grad_norm": 0.2474200972301973,
       "learning_rate": 2.0470701399947667e-07,
-      "loss": 0.1917,
+      "loss": 0.1919,
       "step": 1146
     },
     {
       "epoch": 4.786645800730308,
-      "grad_norm": 0.23485611549166102,
+      "grad_norm": 0.23889435152299254,
       "learning_rate": 1.9645045089367488e-07,
-      "loss": 0.2053,
+      "loss": 0.2056,
       "step": 1147
     },
     {
       "epoch": 4.790818988002087,
-      "grad_norm": 0.24857826082342194,
+      "grad_norm": 0.2482252961015603,
       "learning_rate": 1.883630196954389e-07,
-      "loss": 0.1694,
+      "loss": 0.1698,
       "step": 1148
     },
     {
       "epoch": 4.794992175273865,
-      "grad_norm": 0.2464234016658559,
+      "grad_norm": 0.24550220696643527,
       "learning_rate": 1.8044478947534915e-07,
       "loss": 0.1913,
       "step": 1149
     },
     {
       "epoch": 4.7991653625456445,
-      "grad_norm": 0.24772118006223617,
+      "grad_norm": 0.24708047103075584,
       "learning_rate": 1.7269582785892858e-07,
-      "loss": 0.1669,
+      "loss": 0.1668,
       "step": 1150
     },
     {
       "epoch": 4.803338549817423,
-      "grad_norm": 0.23937638267539738,
+      "grad_norm": 0.2426595647686203,
       "learning_rate": 1.651162010260654e-07,
-      "loss": 0.1668,
+      "loss": 0.1671,
       "step": 1151
     },
     {
       "epoch": 4.807511737089202,
-      "grad_norm": 0.23224930655711892,
+      "grad_norm": 0.23023074264916457,
       "learning_rate": 1.577059737104447e-07,
-      "loss": 0.194,
+      "loss": 0.1937,
       "step": 1152
     },
     {
       "epoch": 4.81168492436098,
-      "grad_norm": 0.24476496643822518,
+      "grad_norm": 0.24677952560079566,
       "learning_rate": 1.5046520919899333e-07,
-      "loss": 0.1665,
+      "loss": 0.1672,
       "step": 1153
     },
     {
       "epoch": 4.8158581116327595,
-      "grad_norm": 0.24538659888135578,
+      "grad_norm": 0.25051342216332967,
       "learning_rate": 1.4339396933135353e-07,
-      "loss": 0.2288,
+      "loss": 0.2291,
       "step": 1154
     },
     {
       "epoch": 4.820031298904539,
-      "grad_norm": 0.24420570773797237,
+      "grad_norm": 0.24674189511981975,
       "learning_rate": 1.3649231449933686e-07,
-      "loss": 0.1945,
+      "loss": 0.195,
       "step": 1155
     },
     {
       "epoch": 4.824204486176317,
-      "grad_norm": 0.24510818399504766,
+      "grad_norm": 0.24772643664282484,
       "learning_rate": 1.297603036464201e-07,
-      "loss": 0.1926,
+      "loss": 0.1929,
       "step": 1156
     },
     {
       "epoch": 4.828377673448096,
-      "grad_norm": 0.24341067145233491,
+      "grad_norm": 0.24435880979989077,
       "learning_rate": 1.23197994267239e-07,
-      "loss": 0.1568,
+      "loss": 0.1573,
       "step": 1157
     },
     {
       "epoch": 4.832550860719875,
-      "grad_norm": 0.2397662604479901,
+      "grad_norm": 0.24365914250612938,
       "learning_rate": 1.1680544240709079e-07,
-      "loss": 0.1878,
+      "loss": 0.188,
       "step": 1158
     },
     {
       "epoch": 4.836724047991654,
-      "grad_norm": 0.25278161729249427,
+      "grad_norm": 0.25255306550995094,
       "learning_rate": 1.1058270266147031e-07,
-      "loss": 0.1903,
+      "loss": 0.1904,
       "step": 1159
     },
     {
       "epoch": 4.840897235263433,
-      "grad_norm": 0.2504196649935766,
+      "grad_norm": 0.24956577669564792,
       "learning_rate": 1.0452982817558577e-07,
-      "loss": 0.1798,
+      "loss": 0.1797,
       "step": 1160
     },
     {
       "epoch": 4.845070422535211,
-      "grad_norm": 0.23345799346688856,
+      "grad_norm": 0.2345096955485978,
       "learning_rate": 9.864687064392142e-08,
-      "loss": 0.1933,
+      "loss": 0.1937,
       "step": 1161
     },
     {
       "epoch": 4.84924360980699,
-      "grad_norm": 0.24702645565375708,
+      "grad_norm": 0.24878114350968558,
       "learning_rate": 9.293388030978012e-08,
-      "loss": 0.2093,
+      "loss": 0.2094,
       "step": 1162
     },
     {
       "epoch": 4.8534167970787685,
-      "grad_norm": 0.24181384776084783,
+      "grad_norm": 0.24335509998235116,
       "learning_rate": 8.739090596487032e-08,
-      "loss": 0.2032,
+      "loss": 0.2031,
       "step": 1163
     },
     {
       "epoch": 4.857589984350548,
-      "grad_norm": 0.24412622602532522,
+      "grad_norm": 0.24838890099023664,
       "learning_rate": 8.201799494887974e-08,
-      "loss": 0.1927,
+      "loss": 0.193,
       "step": 1164
     },
     {
       "epoch": 4.861763171622327,
-      "grad_norm": 0.24611156846075571,
+      "grad_norm": 0.24154965626696692,
       "learning_rate": 7.681519314906904e-08,
-      "loss": 0.1946,
+      "loss": 0.1948,
       "step": 1165
     },
     {
       "epoch": 4.865936358894105,
-      "grad_norm": 0.25213972039928206,
+      "grad_norm": 0.25371753263928276,
       "learning_rate": 7.178254499988546e-08,
-      "loss": 0.1823,
+      "loss": 0.1824,
       "step": 1166
     },
     {
       "epoch": 4.870109546165884,
-      "grad_norm": 0.25608653675862736,
+      "grad_norm": 0.2516905445438908,
       "learning_rate": 6.69200934825831e-08,
       "loss": 0.1994,
       "step": 1167
     },
     {
       "epoch": 4.8742827334376635,
-      "grad_norm": 0.2350447503947461,
+      "grad_norm": 0.23717653555767146,
       "learning_rate": 6.222788012485658e-08,
-      "loss": 0.1905,
+      "loss": 0.1906,
       "step": 1168
     },
     {
       "epoch": 4.878455920709442,
-      "grad_norm": 0.2506561831989489,
+      "grad_norm": 0.24704215624191175,
       "learning_rate": 5.77059450004791e-08,
-      "loss": 0.209,
+      "loss": 0.2094,
       "step": 1169
     },
     {
       "epoch": 4.882629107981221,
-      "grad_norm": 0.23787122228649948,
+      "grad_norm": 0.2374202008512172,
       "learning_rate": 5.335432672896712e-08,
-      "loss": 0.1795,
+      "loss": 0.1798,
       "step": 1170
     },
     {
       "epoch": 4.886802295252999,
-      "grad_norm": 0.2528574373720891,
+      "grad_norm": 0.25148035795868523,
       "learning_rate": 4.9173062475247336e-08,
-      "loss": 0.1714,
+      "loss": 0.172,
       "step": 1171
     },
     {
       "epoch": 4.890975482524778,
-      "grad_norm": 0.2450652135590409,
+      "grad_norm": 0.2451226451113643,
       "learning_rate": 4.5162187949345793e-08,
-      "loss": 0.1982,
+      "loss": 0.1987,
       "step": 1172
     },
     {
       "epoch": 4.895148669796557,
-      "grad_norm": 0.23727892198035,
+      "grad_norm": 0.2381157393301333,
       "learning_rate": 4.1321737406072594e-08,
-      "loss": 0.1997,
+      "loss": 0.2001,
       "step": 1173
     },
     {
       "epoch": 4.899321857068336,
-      "grad_norm": 0.252656803003026,
+      "grad_norm": 0.25173744377904844,
       "learning_rate": 3.765174364473545e-08,
       "loss": 0.2042,
       "step": 1174
     },
     {
       "epoch": 4.903495044340115,
-      "grad_norm": 0.23784235342016877,
+      "grad_norm": 0.24079657926333115,
       "learning_rate": 3.4152238008859915e-08,
-      "loss": 0.1977,
+      "loss": 0.1978,
       "step": 1175
     },
     {
       "epoch": 4.907668231611893,
-      "grad_norm": 0.2505847342626233,
+      "grad_norm": 0.2548594197861856,
       "learning_rate": 3.082325038591849e-08,
-      "loss": 0.199,
+      "loss": 0.1993,
       "step": 1176
     },
     {
       "epoch": 4.9118414188836725,
-      "grad_norm": 0.24026374530838937,
+      "grad_norm": 0.23721845083370927,
       "learning_rate": 2.7664809207079702e-08,
       "loss": 0.1862,
       "step": 1177
     },
     {
       "epoch": 4.916014606155452,
-      "grad_norm": 0.23625255296662617,
+      "grad_norm": 0.23930489303322974,
       "learning_rate": 2.4676941446959425e-08,
-      "loss": 0.1855,
+      "loss": 0.1859,
       "step": 1178
     },
     {
       "epoch": 4.92018779342723,
-      "grad_norm": 0.2500496699925114,
+      "grad_norm": 0.2541675667553618,
       "learning_rate": 2.1859672623398832e-08,
-      "loss": 0.1915,
+      "loss": 0.1918,
       "step": 1179
     },
     {
       "epoch": 4.924360980699009,
-      "grad_norm": 0.23163343272577014,
+      "grad_norm": 0.23625067278789924,
       "learning_rate": 1.921302679723569e-08,
-      "loss": 0.1823,
+      "loss": 0.1828,
       "step": 1180
     },
     {
       "epoch": 4.9285341679707875,
-      "grad_norm": 0.24551821476677813,
+      "grad_norm": 0.2471139385526376,
       "learning_rate": 1.673702657210896e-08,
-      "loss": 0.2002,
+      "loss": 0.2006,
       "step": 1181
     },
     {
       "epoch": 4.932707355242567,
-      "grad_norm": 0.24726571995315194,
+      "grad_norm": 0.24797637740275763,
       "learning_rate": 1.4431693094258959e-08,
-      "loss": 0.2128,
+      "loss": 0.2136,
       "step": 1182
     },
     {
       "epoch": 4.936880542514345,
-      "grad_norm": 0.22602108746664906,
+      "grad_norm": 0.22722140448655628,
       "learning_rate": 1.2297046052351936e-08,
-      "loss": 0.1849,
+      "loss": 0.1855,
       "step": 1183
     },
     {
       "epoch": 4.941053729786124,
-      "grad_norm": 0.2467789001929427,
+      "grad_norm": 0.24769526660636126,
       "learning_rate": 1.0333103677309108e-08,
-      "loss": 0.2082,
+      "loss": 0.2085,
       "step": 1184
     },
     {
       "epoch": 4.945226917057903,
-      "grad_norm": 0.23755656871736405,
+      "grad_norm": 0.24005372459083477,
       "learning_rate": 8.539882742146788e-09,
-      "loss": 0.1858,
+      "loss": 0.1861,
       "step": 1185
     },
     {
       "epoch": 4.949400104329682,
-      "grad_norm": 0.2466348269704945,
+      "grad_norm": 0.2464755553389869,
       "learning_rate": 6.917398561840926e-09,
-      "loss": 0.162,
+      "loss": 0.1622,
       "step": 1186
     },
     {
       "epoch": 4.953573291601461,
-      "grad_norm": 0.24312144079184667,
+      "grad_norm": 0.2385872660384093,
       "learning_rate": 5.465664993196118e-09,
-      "loss": 0.2007,
+      "loss": 0.201,
       "step": 1187
     },
     {
       "epoch": 4.957746478873239,
-      "grad_norm": 0.2477306786461515,
+      "grad_norm": 0.24710114004954792,
       "learning_rate": 4.184694434714587e-09,
-      "loss": 0.1838,
+      "loss": 0.184,
       "step": 1188
     },
     {
       "epoch": 4.961919666145018,
-      "grad_norm": 0.23602219442459352,
+      "grad_norm": 0.23591803513294088,
       "learning_rate": 3.0744978265095924e-09,
-      "loss": 0.1992,
+      "loss": 0.1994,
       "step": 1189
     },
     {
       "epoch": 4.966092853416797,
-      "grad_norm": 0.24614259538761116,
+      "grad_norm": 0.241937205638884,
       "learning_rate": 2.1350846501988487e-09,
-      "loss": 0.1815,
+      "loss": 0.1814,
       "step": 1190
     },
     {
       "epoch": 4.970266040688576,
-      "grad_norm": 0.2510518651909155,
+      "grad_norm": 0.23864904855407648,
       "learning_rate": 1.3664629288223652e-09,
       "loss": 0.1826,
       "step": 1191
     },
     {
       "epoch": 4.974439227960355,
-      "grad_norm": 0.23880555683838872,
+      "grad_norm": 0.23921346027730406,
       "learning_rate": 7.686392267847176e-10,
-      "loss": 0.1802,
+      "loss": 0.1804,
       "step": 1192
     },
     {
       "epoch": 4.978612415232133,
-      "grad_norm": 0.24295993062445956,
+      "grad_norm": 0.24112290071271028,
       "learning_rate": 3.416186497884333e-10,
-      "loss": 0.199,
+      "loss": 0.1992,
       "step": 1193
     },
     {
       "epoch": 4.982785602503912,
-      "grad_norm": 0.24914187146754355,
+      "grad_norm": 0.24303268306245385,
       "learning_rate": 8.540484479624411e-11,
-      "loss": 0.2115,
+      "loss": 0.2116,
       "step": 1194
     },
     {
       "epoch": 4.9869587897756915,
-      "grad_norm": 0.24675838136271278,
+      "grad_norm": 0.2538306538417598,
       "learning_rate": 0.0,
-      "loss": 0.1874,
+      "loss": 0.1876,
       "step": 1195
     },
     {
       "epoch": 4.9869587897756915,
       "step": 1195,
       "total_flos": 2.8011398523411825e+18,
-      "train_loss": 0.0,
-      "train_runtime": 4.4131,
-      "train_samples_per_second": 34739.752,
-      "train_steps_per_second": 270.785
+      "train_loss": 0.4013182716763668,
+      "train_runtime": 47374.7157,
+      "train_samples_per_second": 3.236,
+      "train_steps_per_second": 0.025
     }
   ],
   "logging_steps": 1,